{ "best_global_step": 45657, "best_metric": 0.8766032193098018, "best_model_checkpoint": "output/QA-DeBERTa-MeanPooling-binary/checkpoint-45657", "epoch": 3.0, "eval_steps": 5073, "global_step": 50721, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0005914709883480216, "grad_norm": 3.4899961948394775, "learning_rate": 3.54796320630749e-08, "loss": 0.7046, "step": 10 }, { "epoch": 0.0011829419766960431, "grad_norm": 3.580930709838867, "learning_rate": 7.490144546649146e-08, "loss": 0.6823, "step": 20 }, { "epoch": 0.0017744129650440646, "grad_norm": 2.794832944869995, "learning_rate": 1.1432325886990803e-07, "loss": 0.6714, "step": 30 }, { "epoch": 0.0023658839533920862, "grad_norm": 3.255413293838501, "learning_rate": 1.537450722733246e-07, "loss": 0.7061, "step": 40 }, { "epoch": 0.0029573549417401075, "grad_norm": 4.7144060134887695, "learning_rate": 1.9316688567674116e-07, "loss": 0.7046, "step": 50 }, { "epoch": 0.003548825930088129, "grad_norm": 2.8437976837158203, "learning_rate": 2.325886990801577e-07, "loss": 0.6832, "step": 60 }, { "epoch": 0.004140296918436151, "grad_norm": 2.9447689056396484, "learning_rate": 2.7201051248357425e-07, "loss": 0.6945, "step": 70 }, { "epoch": 0.0047317679067841725, "grad_norm": 3.6161584854125977, "learning_rate": 3.114323258869908e-07, "loss": 0.6941, "step": 80 }, { "epoch": 0.005323238895132194, "grad_norm": 3.2070276737213135, "learning_rate": 3.5085413929040734e-07, "loss": 0.6976, "step": 90 }, { "epoch": 0.005914709883480215, "grad_norm": 4.326408863067627, "learning_rate": 3.902759526938239e-07, "loss": 0.706, "step": 100 }, { "epoch": 0.006506180871828237, "grad_norm": 5.369222640991211, "learning_rate": 4.296977660972405e-07, "loss": 0.6927, "step": 110 }, { "epoch": 0.007097651860176258, "grad_norm": 3.566425085067749, "learning_rate": 4.69119579500657e-07, "loss": 0.686, "step": 120 }, { "epoch": 0.00768912284852428, "grad_norm": 6.7012529373168945, "learning_rate": 5.085413929040735e-07, "loss": 0.7096, "step": 130 }, { "epoch": 0.008280593836872302, "grad_norm": 3.0854568481445312, "learning_rate": 5.479632063074902e-07, "loss": 0.6941, "step": 140 }, { "epoch": 0.008872064825220323, "grad_norm": 5.390612602233887, "learning_rate": 5.873850197109068e-07, "loss": 0.7075, "step": 150 }, { "epoch": 0.009463535813568345, "grad_norm": 2.7556567192077637, "learning_rate": 6.268068331143233e-07, "loss": 0.6763, "step": 160 }, { "epoch": 0.010055006801916367, "grad_norm": 3.142559051513672, "learning_rate": 6.662286465177398e-07, "loss": 0.6807, "step": 170 }, { "epoch": 0.010646477790264388, "grad_norm": 3.0090959072113037, "learning_rate": 7.056504599211564e-07, "loss": 0.6986, "step": 180 }, { "epoch": 0.011237948778612408, "grad_norm": 4.610124111175537, "learning_rate": 7.450722733245729e-07, "loss": 0.6942, "step": 190 }, { "epoch": 0.01182941976696043, "grad_norm": 8.813920974731445, "learning_rate": 7.844940867279895e-07, "loss": 0.7106, "step": 200 }, { "epoch": 0.012420890755308452, "grad_norm": 3.7932097911834717, "learning_rate": 8.239159001314061e-07, "loss": 0.6772, "step": 210 }, { "epoch": 0.013012361743656473, "grad_norm": 3.004957914352417, "learning_rate": 8.633377135348226e-07, "loss": 0.6743, "step": 220 }, { "epoch": 0.013603832732004495, "grad_norm": 3.0236282348632812, "learning_rate": 9.027595269382391e-07, "loss": 0.6846, "step": 230 }, { "epoch": 0.014195303720352517, "grad_norm": 3.353036403656006, "learning_rate": 9.421813403416557e-07, "loss": 0.6727, "step": 240 }, { "epoch": 0.014786774708700538, "grad_norm": 3.3610880374908447, "learning_rate": 9.816031537450722e-07, "loss": 0.6689, "step": 250 }, { "epoch": 0.01537824569704856, "grad_norm": 2.8647947311401367, "learning_rate": 1.0210249671484888e-06, "loss": 0.6864, "step": 260 }, { "epoch": 0.01596971668539658, "grad_norm": 3.1844050884246826, "learning_rate": 1.0604467805519054e-06, "loss": 0.6822, "step": 270 }, { "epoch": 0.016561187673744603, "grad_norm": 2.991396903991699, "learning_rate": 1.0998685939553218e-06, "loss": 0.6461, "step": 280 }, { "epoch": 0.017152658662092625, "grad_norm": 3.1962642669677734, "learning_rate": 1.1392904073587384e-06, "loss": 0.6637, "step": 290 }, { "epoch": 0.017744129650440647, "grad_norm": 5.148901462554932, "learning_rate": 1.178712220762155e-06, "loss": 0.6652, "step": 300 }, { "epoch": 0.01833560063878867, "grad_norm": 4.323608875274658, "learning_rate": 1.2181340341655715e-06, "loss": 0.62, "step": 310 }, { "epoch": 0.01892707162713669, "grad_norm": 3.058164119720459, "learning_rate": 1.2575558475689883e-06, "loss": 0.6228, "step": 320 }, { "epoch": 0.01951854261548471, "grad_norm": 4.159371376037598, "learning_rate": 1.296977660972405e-06, "loss": 0.6173, "step": 330 }, { "epoch": 0.020110013603832733, "grad_norm": 5.1191511154174805, "learning_rate": 1.3363994743758213e-06, "loss": 0.5838, "step": 340 }, { "epoch": 0.020701484592180755, "grad_norm": 4.162530899047852, "learning_rate": 1.375821287779238e-06, "loss": 0.5824, "step": 350 }, { "epoch": 0.021292955580528777, "grad_norm": 5.657260417938232, "learning_rate": 1.4152431011826544e-06, "loss": 0.5155, "step": 360 }, { "epoch": 0.0218844265688768, "grad_norm": 4.746448516845703, "learning_rate": 1.454664914586071e-06, "loss": 0.5116, "step": 370 }, { "epoch": 0.022475897557224817, "grad_norm": 6.091861248016357, "learning_rate": 1.4940867279894876e-06, "loss": 0.4975, "step": 380 }, { "epoch": 0.023067368545572838, "grad_norm": 3.715581178665161, "learning_rate": 1.533508541392904e-06, "loss": 0.4749, "step": 390 }, { "epoch": 0.02365883953392086, "grad_norm": 5.297259330749512, "learning_rate": 1.5729303547963205e-06, "loss": 0.4881, "step": 400 }, { "epoch": 0.02425031052226888, "grad_norm": 7.30882453918457, "learning_rate": 1.612352168199737e-06, "loss": 0.4131, "step": 410 }, { "epoch": 0.024841781510616903, "grad_norm": 5.098006725311279, "learning_rate": 1.6517739816031537e-06, "loss": 0.4673, "step": 420 }, { "epoch": 0.025433252498964925, "grad_norm": 7.409234046936035, "learning_rate": 1.6911957950065705e-06, "loss": 0.4428, "step": 430 }, { "epoch": 0.026024723487312947, "grad_norm": 6.832000732421875, "learning_rate": 1.7306176084099872e-06, "loss": 0.4688, "step": 440 }, { "epoch": 0.026616194475660968, "grad_norm": 5.658520698547363, "learning_rate": 1.7700394218134036e-06, "loss": 0.4475, "step": 450 }, { "epoch": 0.02720766546400899, "grad_norm": 8.078325271606445, "learning_rate": 1.8094612352168202e-06, "loss": 0.4639, "step": 460 }, { "epoch": 0.02779913645235701, "grad_norm": 8.99971866607666, "learning_rate": 1.8488830486202368e-06, "loss": 0.4785, "step": 470 }, { "epoch": 0.028390607440705033, "grad_norm": 5.38008975982666, "learning_rate": 1.8883048620236532e-06, "loss": 0.4488, "step": 480 }, { "epoch": 0.028982078429053055, "grad_norm": 20.354536056518555, "learning_rate": 1.92772667542707e-06, "loss": 0.4513, "step": 490 }, { "epoch": 0.029573549417401077, "grad_norm": 4.146787643432617, "learning_rate": 1.9671484888304865e-06, "loss": 0.4585, "step": 500 }, { "epoch": 0.030165020405749098, "grad_norm": 4.377186298370361, "learning_rate": 2.006570302233903e-06, "loss": 0.4834, "step": 510 }, { "epoch": 0.03075649139409712, "grad_norm": 6.287274360656738, "learning_rate": 2.0459921156373193e-06, "loss": 0.4604, "step": 520 }, { "epoch": 0.03134796238244514, "grad_norm": 3.84550404548645, "learning_rate": 2.085413929040736e-06, "loss": 0.4304, "step": 530 }, { "epoch": 0.03193943337079316, "grad_norm": 5.306344032287598, "learning_rate": 2.1248357424441526e-06, "loss": 0.393, "step": 540 }, { "epoch": 0.03253090435914118, "grad_norm": 6.244089126586914, "learning_rate": 2.164257555847569e-06, "loss": 0.4417, "step": 550 }, { "epoch": 0.03312237534748921, "grad_norm": 8.557563781738281, "learning_rate": 2.203679369250986e-06, "loss": 0.4048, "step": 560 }, { "epoch": 0.033713846335837225, "grad_norm": 8.98536491394043, "learning_rate": 2.2431011826544024e-06, "loss": 0.3665, "step": 570 }, { "epoch": 0.03430531732418525, "grad_norm": 7.016035556793213, "learning_rate": 2.2825229960578186e-06, "loss": 0.3933, "step": 580 }, { "epoch": 0.03489678831253327, "grad_norm": 9.608597755432129, "learning_rate": 2.3219448094612352e-06, "loss": 0.3979, "step": 590 }, { "epoch": 0.03548825930088129, "grad_norm": 8.200888633728027, "learning_rate": 2.361366622864652e-06, "loss": 0.4474, "step": 600 }, { "epoch": 0.03607973028922931, "grad_norm": 7.2876715660095215, "learning_rate": 2.4007884362680685e-06, "loss": 0.4558, "step": 610 }, { "epoch": 0.03667120127757734, "grad_norm": 5.957475662231445, "learning_rate": 2.440210249671485e-06, "loss": 0.3965, "step": 620 }, { "epoch": 0.037262672265925355, "grad_norm": 5.001860618591309, "learning_rate": 2.4796320630749017e-06, "loss": 0.4718, "step": 630 }, { "epoch": 0.03785414325427338, "grad_norm": 9.620041847229004, "learning_rate": 2.519053876478318e-06, "loss": 0.4372, "step": 640 }, { "epoch": 0.0384456142426214, "grad_norm": 6.821552753448486, "learning_rate": 2.5584756898817346e-06, "loss": 0.4348, "step": 650 }, { "epoch": 0.03903708523096942, "grad_norm": 6.374046802520752, "learning_rate": 2.597897503285151e-06, "loss": 0.441, "step": 660 }, { "epoch": 0.03962855621931744, "grad_norm": 8.545299530029297, "learning_rate": 2.637319316688568e-06, "loss": 0.3923, "step": 670 }, { "epoch": 0.04022002720766547, "grad_norm": 7.049600124359131, "learning_rate": 2.6767411300919844e-06, "loss": 0.4011, "step": 680 }, { "epoch": 0.040811498196013485, "grad_norm": 6.139647006988525, "learning_rate": 2.7161629434954006e-06, "loss": 0.3889, "step": 690 }, { "epoch": 0.04140296918436151, "grad_norm": 4.314230918884277, "learning_rate": 2.7555847568988173e-06, "loss": 0.3805, "step": 700 }, { "epoch": 0.04199444017270953, "grad_norm": 13.20296859741211, "learning_rate": 2.795006570302234e-06, "loss": 0.3953, "step": 710 }, { "epoch": 0.04258591116105755, "grad_norm": 9.031804084777832, "learning_rate": 2.8344283837056505e-06, "loss": 0.435, "step": 720 }, { "epoch": 0.04317738214940557, "grad_norm": 7.628100872039795, "learning_rate": 2.873850197109067e-06, "loss": 0.4614, "step": 730 }, { "epoch": 0.0437688531377536, "grad_norm": 5.438625812530518, "learning_rate": 2.9132720105124838e-06, "loss": 0.4165, "step": 740 }, { "epoch": 0.044360324126101615, "grad_norm": 7.7397565841674805, "learning_rate": 2.9526938239159e-06, "loss": 0.403, "step": 750 }, { "epoch": 0.04495179511444963, "grad_norm": 6.702985763549805, "learning_rate": 2.9921156373193166e-06, "loss": 0.4577, "step": 760 }, { "epoch": 0.04554326610279766, "grad_norm": 6.685020446777344, "learning_rate": 3.0315374507227336e-06, "loss": 0.4282, "step": 770 }, { "epoch": 0.046134737091145676, "grad_norm": 5.472206115722656, "learning_rate": 3.0709592641261503e-06, "loss": 0.4371, "step": 780 }, { "epoch": 0.0467262080794937, "grad_norm": 5.869134426116943, "learning_rate": 3.110381077529567e-06, "loss": 0.4116, "step": 790 }, { "epoch": 0.04731767906784172, "grad_norm": 4.651507377624512, "learning_rate": 3.149802890932983e-06, "loss": 0.4809, "step": 800 }, { "epoch": 0.047909150056189745, "grad_norm": 3.889069080352783, "learning_rate": 3.1892247043363997e-06, "loss": 0.4642, "step": 810 }, { "epoch": 0.04850062104453776, "grad_norm": 4.131644248962402, "learning_rate": 3.2286465177398163e-06, "loss": 0.3646, "step": 820 }, { "epoch": 0.04909209203288579, "grad_norm": 8.528861045837402, "learning_rate": 3.268068331143233e-06, "loss": 0.4585, "step": 830 }, { "epoch": 0.049683563021233806, "grad_norm": 4.892425537109375, "learning_rate": 3.3074901445466496e-06, "loss": 0.4156, "step": 840 }, { "epoch": 0.05027503400958183, "grad_norm": 4.803565502166748, "learning_rate": 3.346911957950066e-06, "loss": 0.4052, "step": 850 }, { "epoch": 0.05086650499792985, "grad_norm": 4.584690093994141, "learning_rate": 3.3863337713534824e-06, "loss": 0.3263, "step": 860 }, { "epoch": 0.051457975986277875, "grad_norm": 5.978686332702637, "learning_rate": 3.425755584756899e-06, "loss": 0.4088, "step": 870 }, { "epoch": 0.05204944697462589, "grad_norm": 10.72308349609375, "learning_rate": 3.4651773981603156e-06, "loss": 0.4309, "step": 880 }, { "epoch": 0.05264091796297392, "grad_norm": 5.273160934448242, "learning_rate": 3.5045992115637323e-06, "loss": 0.3736, "step": 890 }, { "epoch": 0.053232388951321936, "grad_norm": 5.966947078704834, "learning_rate": 3.544021024967149e-06, "loss": 0.4009, "step": 900 }, { "epoch": 0.05382385993966996, "grad_norm": 9.172127723693848, "learning_rate": 3.583442838370565e-06, "loss": 0.381, "step": 910 }, { "epoch": 0.05441533092801798, "grad_norm": 8.974810600280762, "learning_rate": 3.6228646517739817e-06, "loss": 0.3887, "step": 920 }, { "epoch": 0.055006801916366005, "grad_norm": 6.651805877685547, "learning_rate": 3.6622864651773983e-06, "loss": 0.4119, "step": 930 }, { "epoch": 0.05559827290471402, "grad_norm": 9.609299659729004, "learning_rate": 3.701708278580815e-06, "loss": 0.4426, "step": 940 }, { "epoch": 0.05618974389306205, "grad_norm": 5.562663555145264, "learning_rate": 3.7411300919842316e-06, "loss": 0.4375, "step": 950 }, { "epoch": 0.056781214881410066, "grad_norm": 12.02171802520752, "learning_rate": 3.780551905387648e-06, "loss": 0.422, "step": 960 }, { "epoch": 0.05737268586975809, "grad_norm": 4.127640247344971, "learning_rate": 3.819973718791065e-06, "loss": 0.3963, "step": 970 }, { "epoch": 0.05796415685810611, "grad_norm": 5.606810569763184, "learning_rate": 3.8593955321944815e-06, "loss": 0.4124, "step": 980 }, { "epoch": 0.05855562784645413, "grad_norm": 9.747590065002441, "learning_rate": 3.898817345597898e-06, "loss": 0.3459, "step": 990 }, { "epoch": 0.05914709883480215, "grad_norm": 7.6824750900268555, "learning_rate": 3.938239159001314e-06, "loss": 0.38, "step": 1000 }, { "epoch": 0.05973856982315017, "grad_norm": 4.980546951293945, "learning_rate": 3.9776609724047305e-06, "loss": 0.3719, "step": 1010 }, { "epoch": 0.060330040811498196, "grad_norm": 4.772561073303223, "learning_rate": 4.017082785808147e-06, "loss": 0.4044, "step": 1020 }, { "epoch": 0.060921511799846215, "grad_norm": 5.81464958190918, "learning_rate": 4.056504599211564e-06, "loss": 0.449, "step": 1030 }, { "epoch": 0.06151298278819424, "grad_norm": 8.63952350616455, "learning_rate": 4.09592641261498e-06, "loss": 0.4481, "step": 1040 }, { "epoch": 0.06210445377654226, "grad_norm": 4.700294017791748, "learning_rate": 4.135348226018397e-06, "loss": 0.3446, "step": 1050 }, { "epoch": 0.06269592476489028, "grad_norm": 4.8281354904174805, "learning_rate": 4.174770039421814e-06, "loss": 0.4305, "step": 1060 }, { "epoch": 0.0632873957532383, "grad_norm": 6.409106731414795, "learning_rate": 4.21419185282523e-06, "loss": 0.4062, "step": 1070 }, { "epoch": 0.06387886674158633, "grad_norm": 6.680933952331543, "learning_rate": 4.253613666228647e-06, "loss": 0.3767, "step": 1080 }, { "epoch": 0.06447033772993435, "grad_norm": 4.746706962585449, "learning_rate": 4.2930354796320635e-06, "loss": 0.4022, "step": 1090 }, { "epoch": 0.06506180871828236, "grad_norm": 7.483639717102051, "learning_rate": 4.33245729303548e-06, "loss": 0.3442, "step": 1100 }, { "epoch": 0.06565327970663039, "grad_norm": 4.240472316741943, "learning_rate": 4.371879106438897e-06, "loss": 0.3723, "step": 1110 }, { "epoch": 0.06624475069497841, "grad_norm": 8.767252922058105, "learning_rate": 4.4113009198423125e-06, "loss": 0.4123, "step": 1120 }, { "epoch": 0.06683622168332644, "grad_norm": 6.967141151428223, "learning_rate": 4.450722733245729e-06, "loss": 0.398, "step": 1130 }, { "epoch": 0.06742769267167445, "grad_norm": 6.145914077758789, "learning_rate": 4.490144546649146e-06, "loss": 0.3799, "step": 1140 }, { "epoch": 0.06801916366002247, "grad_norm": 7.687303066253662, "learning_rate": 4.529566360052562e-06, "loss": 0.4001, "step": 1150 }, { "epoch": 0.0686106346483705, "grad_norm": 4.149770736694336, "learning_rate": 4.568988173455979e-06, "loss": 0.3645, "step": 1160 }, { "epoch": 0.06920210563671852, "grad_norm": 4.779444217681885, "learning_rate": 4.608409986859396e-06, "loss": 0.3942, "step": 1170 }, { "epoch": 0.06979357662506654, "grad_norm": 4.6843132972717285, "learning_rate": 4.647831800262812e-06, "loss": 0.3988, "step": 1180 }, { "epoch": 0.07038504761341456, "grad_norm": 7.65682315826416, "learning_rate": 4.687253613666229e-06, "loss": 0.3755, "step": 1190 }, { "epoch": 0.07097651860176259, "grad_norm": 7.211122035980225, "learning_rate": 4.7266754270696455e-06, "loss": 0.3835, "step": 1200 }, { "epoch": 0.07156798959011061, "grad_norm": 5.304022789001465, "learning_rate": 4.766097240473062e-06, "loss": 0.3663, "step": 1210 }, { "epoch": 0.07215946057845862, "grad_norm": 7.462334156036377, "learning_rate": 4.805519053876479e-06, "loss": 0.4088, "step": 1220 }, { "epoch": 0.07275093156680665, "grad_norm": 8.520455360412598, "learning_rate": 4.8449408672798945e-06, "loss": 0.3909, "step": 1230 }, { "epoch": 0.07334240255515467, "grad_norm": 5.808437347412109, "learning_rate": 4.884362680683311e-06, "loss": 0.3982, "step": 1240 }, { "epoch": 0.07393387354350268, "grad_norm": 4.8766937255859375, "learning_rate": 4.923784494086728e-06, "loss": 0.3844, "step": 1250 }, { "epoch": 0.07452534453185071, "grad_norm": 4.98210334777832, "learning_rate": 4.963206307490144e-06, "loss": 0.4442, "step": 1260 }, { "epoch": 0.07511681552019873, "grad_norm": 3.691650152206421, "learning_rate": 5.002628120893561e-06, "loss": 0.3703, "step": 1270 }, { "epoch": 0.07570828650854676, "grad_norm": 6.813695907592773, "learning_rate": 5.042049934296978e-06, "loss": 0.3704, "step": 1280 }, { "epoch": 0.07629975749689477, "grad_norm": 8.51002311706543, "learning_rate": 5.081471747700394e-06, "loss": 0.3722, "step": 1290 }, { "epoch": 0.0768912284852428, "grad_norm": 6.425374507904053, "learning_rate": 5.120893561103811e-06, "loss": 0.3557, "step": 1300 }, { "epoch": 0.07748269947359082, "grad_norm": 4.950004577636719, "learning_rate": 5.1603153745072275e-06, "loss": 0.36, "step": 1310 }, { "epoch": 0.07807417046193885, "grad_norm": 7.337320804595947, "learning_rate": 5.199737187910644e-06, "loss": 0.3916, "step": 1320 }, { "epoch": 0.07866564145028686, "grad_norm": 4.00466775894165, "learning_rate": 5.239159001314061e-06, "loss": 0.3979, "step": 1330 }, { "epoch": 0.07925711243863488, "grad_norm": 5.359980583190918, "learning_rate": 5.278580814717477e-06, "loss": 0.3557, "step": 1340 }, { "epoch": 0.07984858342698291, "grad_norm": 4.687009334564209, "learning_rate": 5.318002628120893e-06, "loss": 0.3695, "step": 1350 }, { "epoch": 0.08044005441533093, "grad_norm": 5.6446967124938965, "learning_rate": 5.35742444152431e-06, "loss": 0.3153, "step": 1360 }, { "epoch": 0.08103152540367894, "grad_norm": 6.959810733795166, "learning_rate": 5.396846254927726e-06, "loss": 0.3481, "step": 1370 }, { "epoch": 0.08162299639202697, "grad_norm": 6.431433200836182, "learning_rate": 5.436268068331143e-06, "loss": 0.3979, "step": 1380 }, { "epoch": 0.082214467380375, "grad_norm": 4.417128562927246, "learning_rate": 5.47568988173456e-06, "loss": 0.4018, "step": 1390 }, { "epoch": 0.08280593836872302, "grad_norm": 9.306402206420898, "learning_rate": 5.515111695137976e-06, "loss": 0.2843, "step": 1400 }, { "epoch": 0.08339740935707103, "grad_norm": 4.698502540588379, "learning_rate": 5.554533508541393e-06, "loss": 0.3917, "step": 1410 }, { "epoch": 0.08398888034541906, "grad_norm": 6.421928882598877, "learning_rate": 5.5939553219448095e-06, "loss": 0.4216, "step": 1420 }, { "epoch": 0.08458035133376708, "grad_norm": 6.615372180938721, "learning_rate": 5.633377135348226e-06, "loss": 0.4052, "step": 1430 }, { "epoch": 0.0851718223221151, "grad_norm": 4.75895881652832, "learning_rate": 5.672798948751643e-06, "loss": 0.4411, "step": 1440 }, { "epoch": 0.08576329331046312, "grad_norm": 7.480281829833984, "learning_rate": 5.712220762155059e-06, "loss": 0.3828, "step": 1450 }, { "epoch": 0.08635476429881114, "grad_norm": 4.139289855957031, "learning_rate": 5.751642575558476e-06, "loss": 0.3748, "step": 1460 }, { "epoch": 0.08694623528715917, "grad_norm": 7.469866752624512, "learning_rate": 5.791064388961892e-06, "loss": 0.4568, "step": 1470 }, { "epoch": 0.0875377062755072, "grad_norm": 5.682164669036865, "learning_rate": 5.830486202365308e-06, "loss": 0.3783, "step": 1480 }, { "epoch": 0.0881291772638552, "grad_norm": 5.343912601470947, "learning_rate": 5.869908015768725e-06, "loss": 0.4157, "step": 1490 }, { "epoch": 0.08872064825220323, "grad_norm": 6.916319370269775, "learning_rate": 5.909329829172142e-06, "loss": 0.3795, "step": 1500 }, { "epoch": 0.08931211924055125, "grad_norm": 4.257317066192627, "learning_rate": 5.948751642575558e-06, "loss": 0.3698, "step": 1510 }, { "epoch": 0.08990359022889927, "grad_norm": 3.654940366744995, "learning_rate": 5.988173455978975e-06, "loss": 0.3578, "step": 1520 }, { "epoch": 0.09049506121724729, "grad_norm": 7.610441207885742, "learning_rate": 5.9999997003084304e-06, "loss": 0.4285, "step": 1530 }, { "epoch": 0.09108653220559532, "grad_norm": 3.565256118774414, "learning_rate": 5.9999982324315e-06, "loss": 0.3076, "step": 1540 }, { "epoch": 0.09167800319394334, "grad_norm": 7.7041802406311035, "learning_rate": 5.999995541324415e-06, "loss": 0.3288, "step": 1550 }, { "epoch": 0.09226947418229135, "grad_norm": 4.205705642700195, "learning_rate": 5.999991626988273e-06, "loss": 0.3172, "step": 1560 }, { "epoch": 0.09286094517063938, "grad_norm": 7.665149211883545, "learning_rate": 5.999986489424671e-06, "loss": 0.3585, "step": 1570 }, { "epoch": 0.0934524161589874, "grad_norm": 4.530918598175049, "learning_rate": 5.999980128635703e-06, "loss": 0.4553, "step": 1580 }, { "epoch": 0.09404388714733543, "grad_norm": 4.176219940185547, "learning_rate": 5.999972544623963e-06, "loss": 0.4002, "step": 1590 }, { "epoch": 0.09463535813568344, "grad_norm": 7.610601425170898, "learning_rate": 5.999963737392543e-06, "loss": 0.3231, "step": 1600 }, { "epoch": 0.09522682912403146, "grad_norm": 4.856690883636475, "learning_rate": 5.999953706945034e-06, "loss": 0.4056, "step": 1610 }, { "epoch": 0.09581830011237949, "grad_norm": 8.173506736755371, "learning_rate": 5.999942453285527e-06, "loss": 0.4145, "step": 1620 }, { "epoch": 0.09640977110072751, "grad_norm": 4.1468024253845215, "learning_rate": 5.999929976418609e-06, "loss": 0.4414, "step": 1630 }, { "epoch": 0.09700124208907553, "grad_norm": 5.958793640136719, "learning_rate": 5.999916276349368e-06, "loss": 0.3643, "step": 1640 }, { "epoch": 0.09759271307742355, "grad_norm": 10.282833099365234, "learning_rate": 5.9999013530833905e-06, "loss": 0.3524, "step": 1650 }, { "epoch": 0.09818418406577158, "grad_norm": 6.332075119018555, "learning_rate": 5.99988520662676e-06, "loss": 0.3857, "step": 1660 }, { "epoch": 0.0987756550541196, "grad_norm": 8.943772315979004, "learning_rate": 5.999867836986061e-06, "loss": 0.4218, "step": 1670 }, { "epoch": 0.09936712604246761, "grad_norm": 6.539673805236816, "learning_rate": 5.9998492441683775e-06, "loss": 0.3558, "step": 1680 }, { "epoch": 0.09995859703081564, "grad_norm": 4.802232265472412, "learning_rate": 5.999829428181288e-06, "loss": 0.3293, "step": 1690 }, { "epoch": 0.10055006801916366, "grad_norm": 8.28378677368164, "learning_rate": 5.999808389032874e-06, "loss": 0.39, "step": 1700 }, { "epoch": 0.10114153900751169, "grad_norm": 3.566438913345337, "learning_rate": 5.999786126731712e-06, "loss": 0.3361, "step": 1710 }, { "epoch": 0.1017330099958597, "grad_norm": 4.262284755706787, "learning_rate": 5.9997626412868825e-06, "loss": 0.3773, "step": 1720 }, { "epoch": 0.10232448098420772, "grad_norm": 5.294562339782715, "learning_rate": 5.99973793270796e-06, "loss": 0.3837, "step": 1730 }, { "epoch": 0.10291595197255575, "grad_norm": 6.596896648406982, "learning_rate": 5.999712001005018e-06, "loss": 0.3677, "step": 1740 }, { "epoch": 0.10350742296090376, "grad_norm": 8.13280963897705, "learning_rate": 5.9996848461886315e-06, "loss": 0.337, "step": 1750 }, { "epoch": 0.10409889394925179, "grad_norm": 4.2672624588012695, "learning_rate": 5.999656468269872e-06, "loss": 0.3879, "step": 1760 }, { "epoch": 0.10469036493759981, "grad_norm": 4.591782569885254, "learning_rate": 5.999626867260311e-06, "loss": 0.4098, "step": 1770 }, { "epoch": 0.10528183592594784, "grad_norm": 4.3288350105285645, "learning_rate": 5.999596043172018e-06, "loss": 0.4126, "step": 1780 }, { "epoch": 0.10587330691429585, "grad_norm": 5.757929801940918, "learning_rate": 5.999563996017561e-06, "loss": 0.3687, "step": 1790 }, { "epoch": 0.10646477790264387, "grad_norm": 7.001318454742432, "learning_rate": 5.999530725810008e-06, "loss": 0.4376, "step": 1800 }, { "epoch": 0.1070562488909919, "grad_norm": 4.472096920013428, "learning_rate": 5.999496232562923e-06, "loss": 0.3787, "step": 1810 }, { "epoch": 0.10764771987933992, "grad_norm": 5.558248519897461, "learning_rate": 5.999460516290371e-06, "loss": 0.3184, "step": 1820 }, { "epoch": 0.10823919086768793, "grad_norm": 3.1399147510528564, "learning_rate": 5.999423577006916e-06, "loss": 0.4335, "step": 1830 }, { "epoch": 0.10883066185603596, "grad_norm": 5.9244537353515625, "learning_rate": 5.9993854147276194e-06, "loss": 0.3371, "step": 1840 }, { "epoch": 0.10942213284438398, "grad_norm": 6.128269672393799, "learning_rate": 5.999346029468041e-06, "loss": 0.3659, "step": 1850 }, { "epoch": 0.11001360383273201, "grad_norm": 3.1777572631835938, "learning_rate": 5.99930542124424e-06, "loss": 0.3603, "step": 1860 }, { "epoch": 0.11060507482108002, "grad_norm": 4.25728702545166, "learning_rate": 5.999263590072774e-06, "loss": 0.3004, "step": 1870 }, { "epoch": 0.11119654580942805, "grad_norm": 8.999167442321777, "learning_rate": 5.999220535970701e-06, "loss": 0.384, "step": 1880 }, { "epoch": 0.11178801679777607, "grad_norm": 4.269046306610107, "learning_rate": 5.999176258955574e-06, "loss": 0.3466, "step": 1890 }, { "epoch": 0.1123794877861241, "grad_norm": 5.9434404373168945, "learning_rate": 5.999130759045448e-06, "loss": 0.3726, "step": 1900 }, { "epoch": 0.11297095877447211, "grad_norm": 7.025379657745361, "learning_rate": 5.9990840362588744e-06, "loss": 0.3991, "step": 1910 }, { "epoch": 0.11356242976282013, "grad_norm": 3.620330333709717, "learning_rate": 5.999036090614905e-06, "loss": 0.3389, "step": 1920 }, { "epoch": 0.11415390075116816, "grad_norm": 6.221548557281494, "learning_rate": 5.998986922133088e-06, "loss": 0.3749, "step": 1930 }, { "epoch": 0.11474537173951618, "grad_norm": 3.3750925064086914, "learning_rate": 5.998936530833474e-06, "loss": 0.2855, "step": 1940 }, { "epoch": 0.1153368427278642, "grad_norm": 5.545888900756836, "learning_rate": 5.998884916736606e-06, "loss": 0.3129, "step": 1950 }, { "epoch": 0.11592831371621222, "grad_norm": 6.760555744171143, "learning_rate": 5.998832079863533e-06, "loss": 0.3823, "step": 1960 }, { "epoch": 0.11651978470456024, "grad_norm": 4.505759239196777, "learning_rate": 5.9987780202357966e-06, "loss": 0.4425, "step": 1970 }, { "epoch": 0.11711125569290826, "grad_norm": 4.5961127281188965, "learning_rate": 5.99872273787544e-06, "loss": 0.4256, "step": 1980 }, { "epoch": 0.11770272668125628, "grad_norm": 5.0527167320251465, "learning_rate": 5.998666232805006e-06, "loss": 0.419, "step": 1990 }, { "epoch": 0.1182941976696043, "grad_norm": 6.028763771057129, "learning_rate": 5.998608505047531e-06, "loss": 0.349, "step": 2000 }, { "epoch": 0.11888566865795233, "grad_norm": 4.7811408042907715, "learning_rate": 5.998549554626554e-06, "loss": 0.3844, "step": 2010 }, { "epoch": 0.11947713964630034, "grad_norm": 5.331112384796143, "learning_rate": 5.9984893815661125e-06, "loss": 0.4025, "step": 2020 }, { "epoch": 0.12006861063464837, "grad_norm": 3.754223346710205, "learning_rate": 5.998427985890742e-06, "loss": 0.3914, "step": 2030 }, { "epoch": 0.12066008162299639, "grad_norm": 3.8598806858062744, "learning_rate": 5.998365367625475e-06, "loss": 0.3564, "step": 2040 }, { "epoch": 0.12125155261134442, "grad_norm": 5.016531944274902, "learning_rate": 5.998301526795844e-06, "loss": 0.3612, "step": 2050 }, { "epoch": 0.12184302359969243, "grad_norm": 3.50274395942688, "learning_rate": 5.998236463427881e-06, "loss": 0.3689, "step": 2060 }, { "epoch": 0.12243449458804045, "grad_norm": 3.133310079574585, "learning_rate": 5.9981701775481135e-06, "loss": 0.3723, "step": 2070 }, { "epoch": 0.12302596557638848, "grad_norm": 4.097148418426514, "learning_rate": 5.9981026691835695e-06, "loss": 0.4427, "step": 2080 }, { "epoch": 0.1236174365647365, "grad_norm": 5.01294469833374, "learning_rate": 5.998033938361776e-06, "loss": 0.3265, "step": 2090 }, { "epoch": 0.12420890755308452, "grad_norm": 5.9389872550964355, "learning_rate": 5.9979639851107565e-06, "loss": 0.3925, "step": 2100 }, { "epoch": 0.12480037854143254, "grad_norm": 3.5741353034973145, "learning_rate": 5.9978928094590345e-06, "loss": 0.3617, "step": 2110 }, { "epoch": 0.12539184952978055, "grad_norm": 3.2828152179718018, "learning_rate": 5.997820411435632e-06, "loss": 0.3566, "step": 2120 }, { "epoch": 0.12598332051812858, "grad_norm": 3.855929374694824, "learning_rate": 5.997746791070068e-06, "loss": 0.2999, "step": 2130 }, { "epoch": 0.1265747915064766, "grad_norm": 5.36519193649292, "learning_rate": 5.9976719483923606e-06, "loss": 0.4022, "step": 2140 }, { "epoch": 0.12716626249482463, "grad_norm": 5.975133895874023, "learning_rate": 5.9975958834330265e-06, "loss": 0.365, "step": 2150 }, { "epoch": 0.12775773348317265, "grad_norm": 3.9919333457946777, "learning_rate": 5.997518596223081e-06, "loss": 0.3987, "step": 2160 }, { "epoch": 0.12834920447152068, "grad_norm": 4.0615129470825195, "learning_rate": 5.997440086794038e-06, "loss": 0.3939, "step": 2170 }, { "epoch": 0.1289406754598687, "grad_norm": 7.158664226531982, "learning_rate": 5.997360355177908e-06, "loss": 0.3252, "step": 2180 }, { "epoch": 0.12953214644821673, "grad_norm": 4.854522705078125, "learning_rate": 5.997279401407202e-06, "loss": 0.3884, "step": 2190 }, { "epoch": 0.13012361743656473, "grad_norm": 6.501551628112793, "learning_rate": 5.997197225514928e-06, "loss": 0.3852, "step": 2200 }, { "epoch": 0.13071508842491275, "grad_norm": 7.240449905395508, "learning_rate": 5.997113827534593e-06, "loss": 0.3812, "step": 2210 }, { "epoch": 0.13130655941326078, "grad_norm": 3.442182779312134, "learning_rate": 5.997029207500202e-06, "loss": 0.3748, "step": 2220 }, { "epoch": 0.1318980304016088, "grad_norm": 4.921692848205566, "learning_rate": 5.996943365446257e-06, "loss": 0.3703, "step": 2230 }, { "epoch": 0.13248950138995683, "grad_norm": 4.713640213012695, "learning_rate": 5.9968563014077616e-06, "loss": 0.4215, "step": 2240 }, { "epoch": 0.13308097237830485, "grad_norm": 4.514056205749512, "learning_rate": 5.996768015420214e-06, "loss": 0.4034, "step": 2250 }, { "epoch": 0.13367244336665288, "grad_norm": 4.399463176727295, "learning_rate": 5.996678507519613e-06, "loss": 0.3327, "step": 2260 }, { "epoch": 0.13426391435500087, "grad_norm": 5.061159133911133, "learning_rate": 5.996587777742455e-06, "loss": 0.3071, "step": 2270 }, { "epoch": 0.1348553853433489, "grad_norm": 7.690370559692383, "learning_rate": 5.9964958261257336e-06, "loss": 0.2632, "step": 2280 }, { "epoch": 0.13544685633169692, "grad_norm": 4.480180740356445, "learning_rate": 5.996402652706943e-06, "loss": 0.4383, "step": 2290 }, { "epoch": 0.13603832732004495, "grad_norm": 7.156957149505615, "learning_rate": 5.996308257524072e-06, "loss": 0.3285, "step": 2300 }, { "epoch": 0.13662979830839297, "grad_norm": 3.920637607574463, "learning_rate": 5.996212640615612e-06, "loss": 0.3545, "step": 2310 }, { "epoch": 0.137221269296741, "grad_norm": 3.8787219524383545, "learning_rate": 5.9961158020205494e-06, "loss": 0.3311, "step": 2320 }, { "epoch": 0.13781274028508902, "grad_norm": 3.810554027557373, "learning_rate": 5.996017741778368e-06, "loss": 0.4119, "step": 2330 }, { "epoch": 0.13840421127343705, "grad_norm": 3.7553300857543945, "learning_rate": 5.995918459929053e-06, "loss": 0.371, "step": 2340 }, { "epoch": 0.13899568226178505, "grad_norm": 4.101539611816406, "learning_rate": 5.995817956513084e-06, "loss": 0.3462, "step": 2350 }, { "epoch": 0.13958715325013307, "grad_norm": 3.166151285171509, "learning_rate": 5.9957162315714425e-06, "loss": 0.3505, "step": 2360 }, { "epoch": 0.1401786242384811, "grad_norm": 3.485133171081543, "learning_rate": 5.995613285145607e-06, "loss": 0.3281, "step": 2370 }, { "epoch": 0.14077009522682912, "grad_norm": 5.145594120025635, "learning_rate": 5.99550911727755e-06, "loss": 0.351, "step": 2380 }, { "epoch": 0.14136156621517715, "grad_norm": 4.108577728271484, "learning_rate": 5.995403728009748e-06, "loss": 0.3513, "step": 2390 }, { "epoch": 0.14195303720352517, "grad_norm": 5.4304890632629395, "learning_rate": 5.995297117385173e-06, "loss": 0.4498, "step": 2400 }, { "epoch": 0.1425445081918732, "grad_norm": 3.799926519393921, "learning_rate": 5.9951892854472925e-06, "loss": 0.3988, "step": 2410 }, { "epoch": 0.14313597918022122, "grad_norm": 3.932490587234497, "learning_rate": 5.995080232240075e-06, "loss": 0.3922, "step": 2420 }, { "epoch": 0.14372745016856922, "grad_norm": 4.788163185119629, "learning_rate": 5.994969957807988e-06, "loss": 0.3672, "step": 2430 }, { "epoch": 0.14431892115691725, "grad_norm": 3.5540826320648193, "learning_rate": 5.994858462195993e-06, "loss": 0.3939, "step": 2440 }, { "epoch": 0.14491039214526527, "grad_norm": 5.575675010681152, "learning_rate": 5.994745745449554e-06, "loss": 0.2755, "step": 2450 }, { "epoch": 0.1455018631336133, "grad_norm": 12.945412635803223, "learning_rate": 5.9946318076146275e-06, "loss": 0.3584, "step": 2460 }, { "epoch": 0.14609333412196132, "grad_norm": 8.080669403076172, "learning_rate": 5.9945166487376736e-06, "loss": 0.368, "step": 2470 }, { "epoch": 0.14668480511030935, "grad_norm": 4.546182155609131, "learning_rate": 5.994400268865647e-06, "loss": 0.362, "step": 2480 }, { "epoch": 0.14727627609865737, "grad_norm": 5.923382759094238, "learning_rate": 5.994282668045999e-06, "loss": 0.411, "step": 2490 }, { "epoch": 0.14786774708700537, "grad_norm": 6.3221330642700195, "learning_rate": 5.994163846326683e-06, "loss": 0.3771, "step": 2500 }, { "epoch": 0.1484592180753534, "grad_norm": 4.627146244049072, "learning_rate": 5.994043803756146e-06, "loss": 0.3814, "step": 2510 }, { "epoch": 0.14905068906370142, "grad_norm": 5.227594375610352, "learning_rate": 5.993922540383336e-06, "loss": 0.3513, "step": 2520 }, { "epoch": 0.14964216005204944, "grad_norm": 5.245380878448486, "learning_rate": 5.9938000562576965e-06, "loss": 0.3565, "step": 2530 }, { "epoch": 0.15023363104039747, "grad_norm": 3.5842931270599365, "learning_rate": 5.993676351429171e-06, "loss": 0.3196, "step": 2540 }, { "epoch": 0.1508251020287455, "grad_norm": 6.634786605834961, "learning_rate": 5.993551425948197e-06, "loss": 0.3444, "step": 2550 }, { "epoch": 0.15141657301709352, "grad_norm": 7.34418249130249, "learning_rate": 5.993425279865713e-06, "loss": 0.3324, "step": 2560 }, { "epoch": 0.15200804400544154, "grad_norm": 4.076575756072998, "learning_rate": 5.993297913233155e-06, "loss": 0.3849, "step": 2570 }, { "epoch": 0.15259951499378954, "grad_norm": 3.8893473148345947, "learning_rate": 5.993169326102457e-06, "loss": 0.3377, "step": 2580 }, { "epoch": 0.15319098598213757, "grad_norm": 4.390764236450195, "learning_rate": 5.993039518526046e-06, "loss": 0.3485, "step": 2590 }, { "epoch": 0.1537824569704856, "grad_norm": 7.095729827880859, "learning_rate": 5.992908490556853e-06, "loss": 0.3327, "step": 2600 }, { "epoch": 0.15437392795883362, "grad_norm": 8.7472505569458, "learning_rate": 5.992776242248303e-06, "loss": 0.4164, "step": 2610 }, { "epoch": 0.15496539894718164, "grad_norm": 3.4556796550750732, "learning_rate": 5.9926427736543195e-06, "loss": 0.368, "step": 2620 }, { "epoch": 0.15555686993552967, "grad_norm": 8.038222312927246, "learning_rate": 5.992508084829324e-06, "loss": 0.3074, "step": 2630 }, { "epoch": 0.1561483409238777, "grad_norm": 5.242167949676514, "learning_rate": 5.992372175828235e-06, "loss": 0.3035, "step": 2640 }, { "epoch": 0.15673981191222572, "grad_norm": 7.868963241577148, "learning_rate": 5.9922350467064656e-06, "loss": 0.3796, "step": 2650 }, { "epoch": 0.15733128290057372, "grad_norm": 4.002470016479492, "learning_rate": 5.992096697519934e-06, "loss": 0.4122, "step": 2660 }, { "epoch": 0.15792275388892174, "grad_norm": 4.216087341308594, "learning_rate": 5.991957128325048e-06, "loss": 0.3785, "step": 2670 }, { "epoch": 0.15851422487726977, "grad_norm": 3.819345712661743, "learning_rate": 5.991816339178718e-06, "loss": 0.2895, "step": 2680 }, { "epoch": 0.1591056958656178, "grad_norm": 4.6164445877075195, "learning_rate": 5.991674330138348e-06, "loss": 0.2736, "step": 2690 }, { "epoch": 0.15969716685396582, "grad_norm": 14.11606216430664, "learning_rate": 5.991531101261843e-06, "loss": 0.4384, "step": 2700 }, { "epoch": 0.16028863784231384, "grad_norm": 5.359839916229248, "learning_rate": 5.9913866526076015e-06, "loss": 0.3731, "step": 2710 }, { "epoch": 0.16088010883066187, "grad_norm": 3.1515097618103027, "learning_rate": 5.991240984234523e-06, "loss": 0.2942, "step": 2720 }, { "epoch": 0.16147157981900986, "grad_norm": 3.8857176303863525, "learning_rate": 5.991094096202004e-06, "loss": 0.4224, "step": 2730 }, { "epoch": 0.1620630508073579, "grad_norm": 2.779806613922119, "learning_rate": 5.990945988569935e-06, "loss": 0.3473, "step": 2740 }, { "epoch": 0.16265452179570591, "grad_norm": 6.004514694213867, "learning_rate": 5.990796661398707e-06, "loss": 0.2905, "step": 2750 }, { "epoch": 0.16324599278405394, "grad_norm": 4.208150386810303, "learning_rate": 5.990646114749207e-06, "loss": 0.3682, "step": 2760 }, { "epoch": 0.16383746377240196, "grad_norm": 3.3078043460845947, "learning_rate": 5.9904943486828186e-06, "loss": 0.3843, "step": 2770 }, { "epoch": 0.16442893476075, "grad_norm": 5.013100624084473, "learning_rate": 5.990341363261424e-06, "loss": 0.4641, "step": 2780 }, { "epoch": 0.16502040574909801, "grad_norm": 3.3766603469848633, "learning_rate": 5.9901871585474036e-06, "loss": 0.3673, "step": 2790 }, { "epoch": 0.16561187673744604, "grad_norm": 6.937079906463623, "learning_rate": 5.9900317346036325e-06, "loss": 0.3658, "step": 2800 }, { "epoch": 0.16620334772579404, "grad_norm": 7.280210494995117, "learning_rate": 5.989875091493483e-06, "loss": 0.3664, "step": 2810 }, { "epoch": 0.16679481871414206, "grad_norm": 5.245065689086914, "learning_rate": 5.989717229280825e-06, "loss": 0.4042, "step": 2820 }, { "epoch": 0.1673862897024901, "grad_norm": 4.642737865447998, "learning_rate": 5.989558148030026e-06, "loss": 0.3352, "step": 2830 }, { "epoch": 0.1679777606908381, "grad_norm": 3.237586498260498, "learning_rate": 5.989397847805953e-06, "loss": 0.3172, "step": 2840 }, { "epoch": 0.16856923167918614, "grad_norm": 3.705432653427124, "learning_rate": 5.989236328673964e-06, "loss": 0.3781, "step": 2850 }, { "epoch": 0.16916070266753416, "grad_norm": 4.139902591705322, "learning_rate": 5.989073590699919e-06, "loss": 0.4025, "step": 2860 }, { "epoch": 0.1697521736558822, "grad_norm": 2.9744670391082764, "learning_rate": 5.988909633950173e-06, "loss": 0.369, "step": 2870 }, { "epoch": 0.1703436446442302, "grad_norm": 6.039011478424072, "learning_rate": 5.988744458491578e-06, "loss": 0.3981, "step": 2880 }, { "epoch": 0.1709351156325782, "grad_norm": 3.389691114425659, "learning_rate": 5.988578064391484e-06, "loss": 0.4219, "step": 2890 }, { "epoch": 0.17152658662092624, "grad_norm": 8.219846725463867, "learning_rate": 5.988410451717737e-06, "loss": 0.3363, "step": 2900 }, { "epoch": 0.17211805760927426, "grad_norm": 4.396327972412109, "learning_rate": 5.988241620538679e-06, "loss": 0.3809, "step": 2910 }, { "epoch": 0.17270952859762229, "grad_norm": 6.202813148498535, "learning_rate": 5.988071570923152e-06, "loss": 0.3598, "step": 2920 }, { "epoch": 0.1733009995859703, "grad_norm": 3.573610782623291, "learning_rate": 5.98790030294049e-06, "loss": 0.3679, "step": 2930 }, { "epoch": 0.17389247057431834, "grad_norm": 2.897308111190796, "learning_rate": 5.987727816660528e-06, "loss": 0.3795, "step": 2940 }, { "epoch": 0.17448394156266636, "grad_norm": 10.100503921508789, "learning_rate": 5.987554112153596e-06, "loss": 0.3293, "step": 2950 }, { "epoch": 0.1750754125510144, "grad_norm": 2.565941333770752, "learning_rate": 5.9873791894905204e-06, "loss": 0.3185, "step": 2960 }, { "epoch": 0.17566688353936238, "grad_norm": 6.899318218231201, "learning_rate": 5.987203048742626e-06, "loss": 0.3484, "step": 2970 }, { "epoch": 0.1762583545277104, "grad_norm": 3.9865939617156982, "learning_rate": 5.9870256899817314e-06, "loss": 0.3183, "step": 2980 }, { "epoch": 0.17684982551605843, "grad_norm": 2.862142324447632, "learning_rate": 5.986847113280155e-06, "loss": 0.3614, "step": 2990 }, { "epoch": 0.17744129650440646, "grad_norm": 4.334918022155762, "learning_rate": 5.98666731871071e-06, "loss": 0.3395, "step": 3000 }, { "epoch": 0.17803276749275448, "grad_norm": 5.696463108062744, "learning_rate": 5.986486306346705e-06, "loss": 0.3513, "step": 3010 }, { "epoch": 0.1786242384811025, "grad_norm": 3.5400445461273193, "learning_rate": 5.986304076261949e-06, "loss": 0.3471, "step": 3020 }, { "epoch": 0.17921570946945053, "grad_norm": 4.904641628265381, "learning_rate": 5.986120628530742e-06, "loss": 0.3593, "step": 3030 }, { "epoch": 0.17980718045779853, "grad_norm": 3.678229570388794, "learning_rate": 5.985935963227887e-06, "loss": 0.3185, "step": 3040 }, { "epoch": 0.18039865144614656, "grad_norm": 4.233956813812256, "learning_rate": 5.985750080428679e-06, "loss": 0.3989, "step": 3050 }, { "epoch": 0.18099012243449458, "grad_norm": 4.262417793273926, "learning_rate": 5.9855629802089095e-06, "loss": 0.3701, "step": 3060 }, { "epoch": 0.1815815934228426, "grad_norm": 2.822638750076294, "learning_rate": 5.985374662644869e-06, "loss": 0.3314, "step": 3070 }, { "epoch": 0.18217306441119063, "grad_norm": 4.370901107788086, "learning_rate": 5.985185127813341e-06, "loss": 0.3861, "step": 3080 }, { "epoch": 0.18276453539953866, "grad_norm": 4.10318660736084, "learning_rate": 5.984994375791608e-06, "loss": 0.3926, "step": 3090 }, { "epoch": 0.18335600638788668, "grad_norm": 7.48127555847168, "learning_rate": 5.984802406657449e-06, "loss": 0.3947, "step": 3100 }, { "epoch": 0.1839474773762347, "grad_norm": 3.4195802211761475, "learning_rate": 5.984609220489136e-06, "loss": 0.3592, "step": 3110 }, { "epoch": 0.1845389483645827, "grad_norm": 2.8071131706237793, "learning_rate": 5.984414817365441e-06, "loss": 0.4535, "step": 3120 }, { "epoch": 0.18513041935293073, "grad_norm": 3.711695432662964, "learning_rate": 5.9842191973656306e-06, "loss": 0.3678, "step": 3130 }, { "epoch": 0.18572189034127876, "grad_norm": 3.293994665145874, "learning_rate": 5.984022360569467e-06, "loss": 0.367, "step": 3140 }, { "epoch": 0.18631336132962678, "grad_norm": 8.707414627075195, "learning_rate": 5.983824307057209e-06, "loss": 0.3441, "step": 3150 }, { "epoch": 0.1869048323179748, "grad_norm": 2.5944008827209473, "learning_rate": 5.983625036909612e-06, "loss": 0.4228, "step": 3160 }, { "epoch": 0.18749630330632283, "grad_norm": 2.9264960289001465, "learning_rate": 5.983424550207928e-06, "loss": 0.3794, "step": 3170 }, { "epoch": 0.18808777429467086, "grad_norm": 3.2377493381500244, "learning_rate": 5.983222847033902e-06, "loss": 0.4292, "step": 3180 }, { "epoch": 0.18867924528301888, "grad_norm": 4.601253032684326, "learning_rate": 5.983019927469779e-06, "loss": 0.3481, "step": 3190 }, { "epoch": 0.18927071627136688, "grad_norm": 5.114628791809082, "learning_rate": 5.982815791598298e-06, "loss": 0.376, "step": 3200 }, { "epoch": 0.1898621872597149, "grad_norm": 3.3446176052093506, "learning_rate": 5.982610439502694e-06, "loss": 0.3717, "step": 3210 }, { "epoch": 0.19045365824806293, "grad_norm": 3.63680362701416, "learning_rate": 5.9824038712666975e-06, "loss": 0.3699, "step": 3220 }, { "epoch": 0.19104512923641095, "grad_norm": 2.9219088554382324, "learning_rate": 5.982196086974536e-06, "loss": 0.3355, "step": 3230 }, { "epoch": 0.19163660022475898, "grad_norm": 4.42433500289917, "learning_rate": 5.9819870867109304e-06, "loss": 0.3286, "step": 3240 }, { "epoch": 0.192228071213107, "grad_norm": 7.516807556152344, "learning_rate": 5.981776870561103e-06, "loss": 0.3612, "step": 3250 }, { "epoch": 0.19281954220145503, "grad_norm": 5.8899664878845215, "learning_rate": 5.981565438610764e-06, "loss": 0.3633, "step": 3260 }, { "epoch": 0.19341101318980303, "grad_norm": 4.401045799255371, "learning_rate": 5.981352790946126e-06, "loss": 0.3355, "step": 3270 }, { "epoch": 0.19400248417815105, "grad_norm": 3.632904291152954, "learning_rate": 5.981138927653894e-06, "loss": 0.3633, "step": 3280 }, { "epoch": 0.19459395516649908, "grad_norm": 2.992751359939575, "learning_rate": 5.980923848821269e-06, "loss": 0.3478, "step": 3290 }, { "epoch": 0.1951854261548471, "grad_norm": 7.537950038909912, "learning_rate": 5.980707554535949e-06, "loss": 0.322, "step": 3300 }, { "epoch": 0.19577689714319513, "grad_norm": 3.891007900238037, "learning_rate": 5.980490044886125e-06, "loss": 0.3337, "step": 3310 }, { "epoch": 0.19636836813154315, "grad_norm": 5.460606575012207, "learning_rate": 5.980271319960487e-06, "loss": 0.3507, "step": 3320 }, { "epoch": 0.19695983911989118, "grad_norm": 3.802039861679077, "learning_rate": 5.980051379848217e-06, "loss": 0.3592, "step": 3330 }, { "epoch": 0.1975513101082392, "grad_norm": 3.393437147140503, "learning_rate": 5.979830224638996e-06, "loss": 0.3752, "step": 3340 }, { "epoch": 0.1981427810965872, "grad_norm": 4.4404802322387695, "learning_rate": 5.9796078544229975e-06, "loss": 0.3335, "step": 3350 }, { "epoch": 0.19873425208493523, "grad_norm": 3.323720932006836, "learning_rate": 5.979384269290891e-06, "loss": 0.3378, "step": 3360 }, { "epoch": 0.19932572307328325, "grad_norm": 6.091846466064453, "learning_rate": 5.979159469333843e-06, "loss": 0.3388, "step": 3370 }, { "epoch": 0.19991719406163128, "grad_norm": 4.964151382446289, "learning_rate": 5.978933454643515e-06, "loss": 0.4027, "step": 3380 }, { "epoch": 0.2005086650499793, "grad_norm": 6.872981071472168, "learning_rate": 5.9787062253120606e-06, "loss": 0.3631, "step": 3390 }, { "epoch": 0.20110013603832733, "grad_norm": 5.55396842956543, "learning_rate": 5.978477781432133e-06, "loss": 0.309, "step": 3400 }, { "epoch": 0.20169160702667535, "grad_norm": 4.817401885986328, "learning_rate": 5.978248123096878e-06, "loss": 0.3219, "step": 3410 }, { "epoch": 0.20228307801502338, "grad_norm": 5.830477714538574, "learning_rate": 5.978017250399937e-06, "loss": 0.3617, "step": 3420 }, { "epoch": 0.20287454900337137, "grad_norm": 3.3920204639434814, "learning_rate": 5.977785163435448e-06, "loss": 0.3683, "step": 3430 }, { "epoch": 0.2034660199917194, "grad_norm": 4.148220062255859, "learning_rate": 5.977551862298042e-06, "loss": 0.3231, "step": 3440 }, { "epoch": 0.20405749098006742, "grad_norm": 4.6151885986328125, "learning_rate": 5.9773173470828465e-06, "loss": 0.3131, "step": 3450 }, { "epoch": 0.20464896196841545, "grad_norm": 5.061943531036377, "learning_rate": 5.977081617885484e-06, "loss": 0.413, "step": 3460 }, { "epoch": 0.20524043295676347, "grad_norm": 3.122866630554199, "learning_rate": 5.97684467480207e-06, "loss": 0.3602, "step": 3470 }, { "epoch": 0.2058319039451115, "grad_norm": 3.5772242546081543, "learning_rate": 5.9766065179292175e-06, "loss": 0.3174, "step": 3480 }, { "epoch": 0.20642337493345952, "grad_norm": 4.413961410522461, "learning_rate": 5.976367147364034e-06, "loss": 0.4348, "step": 3490 }, { "epoch": 0.20701484592180752, "grad_norm": 8.122231483459473, "learning_rate": 5.976126563204121e-06, "loss": 0.3223, "step": 3500 }, { "epoch": 0.20760631691015555, "grad_norm": 5.135698318481445, "learning_rate": 5.975884765547575e-06, "loss": 0.3226, "step": 3510 }, { "epoch": 0.20819778789850357, "grad_norm": 5.4160284996032715, "learning_rate": 5.975641754492987e-06, "loss": 0.3828, "step": 3520 }, { "epoch": 0.2087892588868516, "grad_norm": 3.1204588413238525, "learning_rate": 5.975397530139444e-06, "loss": 0.4224, "step": 3530 }, { "epoch": 0.20938072987519962, "grad_norm": 4.48161506652832, "learning_rate": 5.975152092586527e-06, "loss": 0.3514, "step": 3540 }, { "epoch": 0.20997220086354765, "grad_norm": 5.510709285736084, "learning_rate": 5.974905441934311e-06, "loss": 0.2704, "step": 3550 }, { "epoch": 0.21056367185189567, "grad_norm": 7.9629926681518555, "learning_rate": 5.974657578283366e-06, "loss": 0.4204, "step": 3560 }, { "epoch": 0.2111551428402437, "grad_norm": 4.331998825073242, "learning_rate": 5.974408501734758e-06, "loss": 0.4292, "step": 3570 }, { "epoch": 0.2117466138285917, "grad_norm": 2.791213035583496, "learning_rate": 5.974158212390045e-06, "loss": 0.3692, "step": 3580 }, { "epoch": 0.21233808481693972, "grad_norm": 3.1580724716186523, "learning_rate": 5.973906710351282e-06, "loss": 0.3638, "step": 3590 }, { "epoch": 0.21292955580528775, "grad_norm": 11.021209716796875, "learning_rate": 5.973653995721016e-06, "loss": 0.3168, "step": 3600 }, { "epoch": 0.21352102679363577, "grad_norm": 4.118621349334717, "learning_rate": 5.973400068602291e-06, "loss": 0.3146, "step": 3610 }, { "epoch": 0.2141124977819838, "grad_norm": 3.567999839782715, "learning_rate": 5.973144929098644e-06, "loss": 0.3491, "step": 3620 }, { "epoch": 0.21470396877033182, "grad_norm": 4.998449802398682, "learning_rate": 5.9728885773141065e-06, "loss": 0.3651, "step": 3630 }, { "epoch": 0.21529543975867985, "grad_norm": 3.6098792552948, "learning_rate": 5.972631013353203e-06, "loss": 0.3759, "step": 3640 }, { "epoch": 0.21588691074702787, "grad_norm": 7.993710041046143, "learning_rate": 5.972372237320956e-06, "loss": 0.3682, "step": 3650 }, { "epoch": 0.21647838173537587, "grad_norm": 3.5091257095336914, "learning_rate": 5.972112249322879e-06, "loss": 0.3223, "step": 3660 }, { "epoch": 0.2170698527237239, "grad_norm": 3.8190629482269287, "learning_rate": 5.971851049464978e-06, "loss": 0.3885, "step": 3670 }, { "epoch": 0.21766132371207192, "grad_norm": 4.369227409362793, "learning_rate": 5.971588637853758e-06, "loss": 0.3334, "step": 3680 }, { "epoch": 0.21825279470041994, "grad_norm": 3.988248109817505, "learning_rate": 5.971325014596216e-06, "loss": 0.309, "step": 3690 }, { "epoch": 0.21884426568876797, "grad_norm": 6.8458333015441895, "learning_rate": 5.971060179799841e-06, "loss": 0.4087, "step": 3700 }, { "epoch": 0.219435736677116, "grad_norm": 3.7141480445861816, "learning_rate": 5.970794133572619e-06, "loss": 0.3858, "step": 3710 }, { "epoch": 0.22002720766546402, "grad_norm": 3.6203761100769043, "learning_rate": 5.970526876023029e-06, "loss": 0.3918, "step": 3720 }, { "epoch": 0.22061867865381202, "grad_norm": 3.0205190181732178, "learning_rate": 5.9702584072600415e-06, "loss": 0.3124, "step": 3730 }, { "epoch": 0.22121014964216004, "grad_norm": 4.7661943435668945, "learning_rate": 5.969988727393123e-06, "loss": 0.3735, "step": 3740 }, { "epoch": 0.22180162063050807, "grad_norm": 8.174528121948242, "learning_rate": 5.969717836532237e-06, "loss": 0.374, "step": 3750 }, { "epoch": 0.2223930916188561, "grad_norm": 3.9241738319396973, "learning_rate": 5.969445734787835e-06, "loss": 0.336, "step": 3760 }, { "epoch": 0.22298456260720412, "grad_norm": 4.192067623138428, "learning_rate": 5.9691724222708655e-06, "loss": 0.3854, "step": 3770 }, { "epoch": 0.22357603359555214, "grad_norm": 2.1740684509277344, "learning_rate": 5.968897899092769e-06, "loss": 0.2678, "step": 3780 }, { "epoch": 0.22416750458390017, "grad_norm": 3.169905424118042, "learning_rate": 5.968622165365482e-06, "loss": 0.3338, "step": 3790 }, { "epoch": 0.2247589755722482, "grad_norm": 4.324224472045898, "learning_rate": 5.968345221201431e-06, "loss": 0.3316, "step": 3800 }, { "epoch": 0.2253504465605962, "grad_norm": 5.064943313598633, "learning_rate": 5.968067066713541e-06, "loss": 0.3834, "step": 3810 }, { "epoch": 0.22594191754894422, "grad_norm": 4.90156888961792, "learning_rate": 5.967787702015225e-06, "loss": 0.268, "step": 3820 }, { "epoch": 0.22653338853729224, "grad_norm": 4.608811855316162, "learning_rate": 5.967507127220394e-06, "loss": 0.3308, "step": 3830 }, { "epoch": 0.22712485952564027, "grad_norm": 6.308506011962891, "learning_rate": 5.967225342443451e-06, "loss": 0.3844, "step": 3840 }, { "epoch": 0.2277163305139883, "grad_norm": 8.869704246520996, "learning_rate": 5.966942347799291e-06, "loss": 0.3773, "step": 3850 }, { "epoch": 0.22830780150233632, "grad_norm": 4.420282363891602, "learning_rate": 5.9666581434033015e-06, "loss": 0.2986, "step": 3860 }, { "epoch": 0.22889927249068434, "grad_norm": 4.571893692016602, "learning_rate": 5.966372729371369e-06, "loss": 0.3869, "step": 3870 }, { "epoch": 0.22949074347903237, "grad_norm": 3.273925542831421, "learning_rate": 5.9660861058198655e-06, "loss": 0.3364, "step": 3880 }, { "epoch": 0.23008221446738036, "grad_norm": 2.500347852706909, "learning_rate": 5.965798272865662e-06, "loss": 0.3186, "step": 3890 }, { "epoch": 0.2306736854557284, "grad_norm": 6.365070819854736, "learning_rate": 5.96550923062612e-06, "loss": 0.3498, "step": 3900 }, { "epoch": 0.23126515644407641, "grad_norm": 4.007701396942139, "learning_rate": 5.965218979219095e-06, "loss": 0.3637, "step": 3910 }, { "epoch": 0.23185662743242444, "grad_norm": 5.6600518226623535, "learning_rate": 5.964927518762935e-06, "loss": 0.4022, "step": 3920 }, { "epoch": 0.23244809842077246, "grad_norm": 3.3195571899414062, "learning_rate": 5.964634849376479e-06, "loss": 0.3688, "step": 3930 }, { "epoch": 0.2330395694091205, "grad_norm": 5.083430290222168, "learning_rate": 5.964340971179066e-06, "loss": 0.325, "step": 3940 }, { "epoch": 0.23363104039746851, "grad_norm": 5.04158353805542, "learning_rate": 5.964045884290518e-06, "loss": 0.3853, "step": 3950 }, { "epoch": 0.2342225113858165, "grad_norm": 3.5427329540252686, "learning_rate": 5.963749588831158e-06, "loss": 0.3155, "step": 3960 }, { "epoch": 0.23481398237416454, "grad_norm": 8.604192733764648, "learning_rate": 5.963452084921796e-06, "loss": 0.3339, "step": 3970 }, { "epoch": 0.23540545336251256, "grad_norm": 9.696724891662598, "learning_rate": 5.96315337268374e-06, "loss": 0.3187, "step": 3980 }, { "epoch": 0.2359969243508606, "grad_norm": 3.0684001445770264, "learning_rate": 5.9628534522387865e-06, "loss": 0.321, "step": 3990 }, { "epoch": 0.2365883953392086, "grad_norm": 7.240175724029541, "learning_rate": 5.962552323709226e-06, "loss": 0.3195, "step": 4000 }, { "epoch": 0.23717986632755664, "grad_norm": 6.191490173339844, "learning_rate": 5.962249987217843e-06, "loss": 0.3627, "step": 4010 }, { "epoch": 0.23777133731590466, "grad_norm": 6.271635055541992, "learning_rate": 5.961946442887911e-06, "loss": 0.4102, "step": 4020 }, { "epoch": 0.2383628083042527, "grad_norm": 2.741194009780884, "learning_rate": 5.961641690843201e-06, "loss": 0.3831, "step": 4030 }, { "epoch": 0.23895427929260069, "grad_norm": 2.1166470050811768, "learning_rate": 5.9613357312079704e-06, "loss": 0.3682, "step": 4040 }, { "epoch": 0.2395457502809487, "grad_norm": 4.0290703773498535, "learning_rate": 5.961028564106976e-06, "loss": 0.2945, "step": 4050 }, { "epoch": 0.24013722126929674, "grad_norm": 5.187795162200928, "learning_rate": 5.960720189665461e-06, "loss": 0.3201, "step": 4060 }, { "epoch": 0.24072869225764476, "grad_norm": 2.9597175121307373, "learning_rate": 5.960410608009164e-06, "loss": 0.3886, "step": 4070 }, { "epoch": 0.24132016324599279, "grad_norm": 3.556403160095215, "learning_rate": 5.960099819264315e-06, "loss": 0.3816, "step": 4080 }, { "epoch": 0.2419116342343408, "grad_norm": 4.0555925369262695, "learning_rate": 5.9597878235576345e-06, "loss": 0.3174, "step": 4090 }, { "epoch": 0.24250310522268884, "grad_norm": 5.130129814147949, "learning_rate": 5.959474621016339e-06, "loss": 0.3679, "step": 4100 }, { "epoch": 0.24309457621103686, "grad_norm": 3.3061435222625732, "learning_rate": 5.959160211768133e-06, "loss": 0.3353, "step": 4110 }, { "epoch": 0.24368604719938486, "grad_norm": 4.538347244262695, "learning_rate": 5.958844595941215e-06, "loss": 0.3651, "step": 4120 }, { "epoch": 0.24427751818773288, "grad_norm": 2.4151105880737305, "learning_rate": 5.958527773664277e-06, "loss": 0.3443, "step": 4130 }, { "epoch": 0.2448689891760809, "grad_norm": 4.200124263763428, "learning_rate": 5.9582097450665e-06, "loss": 0.3945, "step": 4140 }, { "epoch": 0.24546046016442893, "grad_norm": 5.971325874328613, "learning_rate": 5.957890510277558e-06, "loss": 0.299, "step": 4150 }, { "epoch": 0.24605193115277696, "grad_norm": 3.6739721298217773, "learning_rate": 5.957570069427618e-06, "loss": 0.3074, "step": 4160 }, { "epoch": 0.24664340214112498, "grad_norm": 4.97865629196167, "learning_rate": 5.9572484226473365e-06, "loss": 0.3243, "step": 4170 }, { "epoch": 0.247234873129473, "grad_norm": 4.596795558929443, "learning_rate": 5.956925570067864e-06, "loss": 0.3851, "step": 4180 }, { "epoch": 0.24782634411782103, "grad_norm": 3.573253870010376, "learning_rate": 5.956601511820841e-06, "loss": 0.3357, "step": 4190 }, { "epoch": 0.24841781510616903, "grad_norm": 7.978744029998779, "learning_rate": 5.9562762480384e-06, "loss": 0.3154, "step": 4200 }, { "epoch": 0.24900928609451706, "grad_norm": 2.634296417236328, "learning_rate": 5.955949778853165e-06, "loss": 0.3877, "step": 4210 }, { "epoch": 0.24960075708286508, "grad_norm": 4.621628761291504, "learning_rate": 5.955622104398253e-06, "loss": 0.3018, "step": 4220 }, { "epoch": 0.2501922280712131, "grad_norm": 4.847694396972656, "learning_rate": 5.95529322480727e-06, "loss": 0.3519, "step": 4230 }, { "epoch": 0.2507836990595611, "grad_norm": 3.701289653778076, "learning_rate": 5.954963140214315e-06, "loss": 0.3328, "step": 4240 }, { "epoch": 0.25137517004790916, "grad_norm": 5.6982808113098145, "learning_rate": 5.954631850753978e-06, "loss": 0.368, "step": 4250 }, { "epoch": 0.25196664103625716, "grad_norm": 4.358921051025391, "learning_rate": 5.954299356561339e-06, "loss": 0.3058, "step": 4260 }, { "epoch": 0.2525581120246052, "grad_norm": 4.9247145652771, "learning_rate": 5.9539656577719735e-06, "loss": 0.3541, "step": 4270 }, { "epoch": 0.2531495830129532, "grad_norm": 3.0930957794189453, "learning_rate": 5.953630754521941e-06, "loss": 0.3768, "step": 4280 }, { "epoch": 0.25374105400130126, "grad_norm": 4.562338829040527, "learning_rate": 5.953294646947798e-06, "loss": 0.3342, "step": 4290 }, { "epoch": 0.25433252498964926, "grad_norm": 4.343782424926758, "learning_rate": 5.9529573351865915e-06, "loss": 0.3184, "step": 4300 }, { "epoch": 0.25492399597799725, "grad_norm": 2.4002926349639893, "learning_rate": 5.952618819375857e-06, "loss": 0.3784, "step": 4310 }, { "epoch": 0.2555154669663453, "grad_norm": 3.2639970779418945, "learning_rate": 5.952279099653621e-06, "loss": 0.3735, "step": 4320 }, { "epoch": 0.2561069379546933, "grad_norm": 2.8424172401428223, "learning_rate": 5.951938176158404e-06, "loss": 0.3456, "step": 4330 }, { "epoch": 0.25669840894304136, "grad_norm": 5.7685627937316895, "learning_rate": 5.9515960490292145e-06, "loss": 0.3262, "step": 4340 }, { "epoch": 0.25728987993138935, "grad_norm": 4.788645267486572, "learning_rate": 5.9512527184055525e-06, "loss": 0.3405, "step": 4350 }, { "epoch": 0.2578813509197374, "grad_norm": 2.526440143585205, "learning_rate": 5.95090818442741e-06, "loss": 0.3373, "step": 4360 }, { "epoch": 0.2584728219080854, "grad_norm": 3.0996029376983643, "learning_rate": 5.950562447235266e-06, "loss": 0.3414, "step": 4370 }, { "epoch": 0.25906429289643346, "grad_norm": 3.5387279987335205, "learning_rate": 5.950215506970096e-06, "loss": 0.3626, "step": 4380 }, { "epoch": 0.25965576388478145, "grad_norm": 3.621683120727539, "learning_rate": 5.9498673637733604e-06, "loss": 0.3473, "step": 4390 }, { "epoch": 0.26024723487312945, "grad_norm": 4.956582546234131, "learning_rate": 5.949518017787013e-06, "loss": 0.3536, "step": 4400 }, { "epoch": 0.2608387058614775, "grad_norm": 4.221882343292236, "learning_rate": 5.949167469153497e-06, "loss": 0.3967, "step": 4410 }, { "epoch": 0.2614301768498255, "grad_norm": 2.615689754486084, "learning_rate": 5.948815718015748e-06, "loss": 0.3925, "step": 4420 }, { "epoch": 0.26202164783817355, "grad_norm": 3.9598543643951416, "learning_rate": 5.948462764517188e-06, "loss": 0.4069, "step": 4430 }, { "epoch": 0.26261311882652155, "grad_norm": 3.3983371257781982, "learning_rate": 5.9481086088017325e-06, "loss": 0.3659, "step": 4440 }, { "epoch": 0.2632045898148696, "grad_norm": 10.501864433288574, "learning_rate": 5.947753251013786e-06, "loss": 0.3423, "step": 4450 }, { "epoch": 0.2637960608032176, "grad_norm": 5.012192726135254, "learning_rate": 5.9473966912982445e-06, "loss": 0.3119, "step": 4460 }, { "epoch": 0.2643875317915656, "grad_norm": 3.664231777191162, "learning_rate": 5.947038929800492e-06, "loss": 0.3095, "step": 4470 }, { "epoch": 0.26497900277991365, "grad_norm": 4.538515090942383, "learning_rate": 5.946679966666404e-06, "loss": 0.388, "step": 4480 }, { "epoch": 0.26557047376826165, "grad_norm": 3.031019449234009, "learning_rate": 5.946319802042343e-06, "loss": 0.3022, "step": 4490 }, { "epoch": 0.2661619447566097, "grad_norm": 5.032029151916504, "learning_rate": 5.945958436075168e-06, "loss": 0.359, "step": 4500 }, { "epoch": 0.2667534157449577, "grad_norm": 4.7050065994262695, "learning_rate": 5.945595868912221e-06, "loss": 0.3461, "step": 4510 }, { "epoch": 0.26734488673330575, "grad_norm": 3.0268406867980957, "learning_rate": 5.945232100701337e-06, "loss": 0.2982, "step": 4520 }, { "epoch": 0.26793635772165375, "grad_norm": 4.2078351974487305, "learning_rate": 5.944867131590839e-06, "loss": 0.3123, "step": 4530 }, { "epoch": 0.26852782871000175, "grad_norm": 3.3568105697631836, "learning_rate": 5.944500961729543e-06, "loss": 0.3717, "step": 4540 }, { "epoch": 0.2691192996983498, "grad_norm": 3.9498302936553955, "learning_rate": 5.944133591266751e-06, "loss": 0.4319, "step": 4550 }, { "epoch": 0.2697107706866978, "grad_norm": 2.9369874000549316, "learning_rate": 5.943765020352257e-06, "loss": 0.377, "step": 4560 }, { "epoch": 0.27030224167504585, "grad_norm": 2.54371976852417, "learning_rate": 5.943395249136342e-06, "loss": 0.3844, "step": 4570 }, { "epoch": 0.27089371266339385, "grad_norm": 3.2625250816345215, "learning_rate": 5.94302427776978e-06, "loss": 0.3598, "step": 4580 }, { "epoch": 0.2714851836517419, "grad_norm": 4.234705924987793, "learning_rate": 5.9426521064038285e-06, "loss": 0.2697, "step": 4590 }, { "epoch": 0.2720766546400899, "grad_norm": 6.613722801208496, "learning_rate": 5.942278735190242e-06, "loss": 0.3752, "step": 4600 }, { "epoch": 0.27266812562843795, "grad_norm": 4.917906284332275, "learning_rate": 5.941904164281259e-06, "loss": 0.3489, "step": 4610 }, { "epoch": 0.27325959661678595, "grad_norm": 3.987262010574341, "learning_rate": 5.941528393829607e-06, "loss": 0.3782, "step": 4620 }, { "epoch": 0.27385106760513395, "grad_norm": 6.0949015617370605, "learning_rate": 5.941151423988506e-06, "loss": 0.346, "step": 4630 }, { "epoch": 0.274442538593482, "grad_norm": 3.4891269207000732, "learning_rate": 5.94077325491166e-06, "loss": 0.2941, "step": 4640 }, { "epoch": 0.27503400958183, "grad_norm": 6.606314182281494, "learning_rate": 5.940393886753269e-06, "loss": 0.3746, "step": 4650 }, { "epoch": 0.27562548057017805, "grad_norm": 3.410438060760498, "learning_rate": 5.9400133196680155e-06, "loss": 0.3166, "step": 4660 }, { "epoch": 0.27621695155852605, "grad_norm": 4.037724494934082, "learning_rate": 5.939631553811074e-06, "loss": 0.3903, "step": 4670 }, { "epoch": 0.2768084225468741, "grad_norm": 2.7326676845550537, "learning_rate": 5.9392485893381065e-06, "loss": 0.4385, "step": 4680 }, { "epoch": 0.2773998935352221, "grad_norm": 4.113360404968262, "learning_rate": 5.938864426405266e-06, "loss": 0.312, "step": 4690 }, { "epoch": 0.2779913645235701, "grad_norm": 6.021586894989014, "learning_rate": 5.9384790651691895e-06, "loss": 0.3877, "step": 4700 }, { "epoch": 0.27858283551191815, "grad_norm": 5.689541339874268, "learning_rate": 5.938092505787008e-06, "loss": 0.3487, "step": 4710 }, { "epoch": 0.27917430650026615, "grad_norm": 4.1705217361450195, "learning_rate": 5.9377047484163385e-06, "loss": 0.3259, "step": 4720 }, { "epoch": 0.2797657774886142, "grad_norm": 3.1582753658294678, "learning_rate": 5.937315793215285e-06, "loss": 0.3145, "step": 4730 }, { "epoch": 0.2803572484769622, "grad_norm": 4.719818115234375, "learning_rate": 5.936925640342443e-06, "loss": 0.3183, "step": 4740 }, { "epoch": 0.28094871946531025, "grad_norm": 3.582045078277588, "learning_rate": 5.936534289956895e-06, "loss": 0.3873, "step": 4750 }, { "epoch": 0.28154019045365825, "grad_norm": 3.191992998123169, "learning_rate": 5.93614174221821e-06, "loss": 0.3656, "step": 4760 }, { "epoch": 0.28213166144200624, "grad_norm": 2.426666736602783, "learning_rate": 5.935747997286449e-06, "loss": 0.3498, "step": 4770 }, { "epoch": 0.2827231324303543, "grad_norm": 2.704719066619873, "learning_rate": 5.935353055322157e-06, "loss": 0.3266, "step": 4780 }, { "epoch": 0.2833146034187023, "grad_norm": 4.314831256866455, "learning_rate": 5.93495691648637e-06, "loss": 0.3676, "step": 4790 }, { "epoch": 0.28390607440705035, "grad_norm": 5.099610328674316, "learning_rate": 5.934559580940611e-06, "loss": 0.399, "step": 4800 }, { "epoch": 0.28449754539539834, "grad_norm": 2.4174766540527344, "learning_rate": 5.9341610488468915e-06, "loss": 0.3416, "step": 4810 }, { "epoch": 0.2850890163837464, "grad_norm": 3.0553388595581055, "learning_rate": 5.933761320367709e-06, "loss": 0.3809, "step": 4820 }, { "epoch": 0.2856804873720944, "grad_norm": 1.987639307975769, "learning_rate": 5.933360395666053e-06, "loss": 0.3734, "step": 4830 }, { "epoch": 0.28627195836044245, "grad_norm": 2.2430434226989746, "learning_rate": 5.932958274905394e-06, "loss": 0.3812, "step": 4840 }, { "epoch": 0.28686342934879044, "grad_norm": 4.301815509796143, "learning_rate": 5.932554958249698e-06, "loss": 0.3442, "step": 4850 }, { "epoch": 0.28745490033713844, "grad_norm": 3.654698133468628, "learning_rate": 5.932150445863412e-06, "loss": 0.3069, "step": 4860 }, { "epoch": 0.2880463713254865, "grad_norm": 3.7789466381073, "learning_rate": 5.931744737911475e-06, "loss": 0.3737, "step": 4870 }, { "epoch": 0.2886378423138345, "grad_norm": 3.074875831604004, "learning_rate": 5.931337834559311e-06, "loss": 0.3126, "step": 4880 }, { "epoch": 0.28922931330218254, "grad_norm": 3.927415132522583, "learning_rate": 5.930929735972833e-06, "loss": 0.3507, "step": 4890 }, { "epoch": 0.28982078429053054, "grad_norm": 3.7327239513397217, "learning_rate": 5.93052044231844e-06, "loss": 0.3711, "step": 4900 }, { "epoch": 0.2904122552788786, "grad_norm": 3.515810489654541, "learning_rate": 5.930109953763019e-06, "loss": 0.3556, "step": 4910 }, { "epoch": 0.2910037262672266, "grad_norm": 2.8134331703186035, "learning_rate": 5.929698270473945e-06, "loss": 0.3249, "step": 4920 }, { "epoch": 0.2915951972555746, "grad_norm": 4.292950630187988, "learning_rate": 5.929285392619077e-06, "loss": 0.324, "step": 4930 }, { "epoch": 0.29218666824392264, "grad_norm": 4.489079475402832, "learning_rate": 5.9288713203667655e-06, "loss": 0.3372, "step": 4940 }, { "epoch": 0.29277813923227064, "grad_norm": 5.930366039276123, "learning_rate": 5.928456053885845e-06, "loss": 0.4058, "step": 4950 }, { "epoch": 0.2933696102206187, "grad_norm": 3.3764851093292236, "learning_rate": 5.9280395933456375e-06, "loss": 0.3514, "step": 4960 }, { "epoch": 0.2939610812089667, "grad_norm": 3.184710741043091, "learning_rate": 5.927621938915952e-06, "loss": 0.3584, "step": 4970 }, { "epoch": 0.29455255219731474, "grad_norm": 2.81225323677063, "learning_rate": 5.9272030907670855e-06, "loss": 0.3329, "step": 4980 }, { "epoch": 0.29514402318566274, "grad_norm": 2.4070491790771484, "learning_rate": 5.9267830490698205e-06, "loss": 0.3355, "step": 4990 }, { "epoch": 0.29573549417401074, "grad_norm": 4.582977294921875, "learning_rate": 5.926361813995425e-06, "loss": 0.3297, "step": 5000 }, { "epoch": 0.2963269651623588, "grad_norm": 6.714651584625244, "learning_rate": 5.925939385715656e-06, "loss": 0.279, "step": 5010 }, { "epoch": 0.2969184361507068, "grad_norm": 3.769761562347412, "learning_rate": 5.925515764402755e-06, "loss": 0.3721, "step": 5020 }, { "epoch": 0.29750990713905484, "grad_norm": 3.4236230850219727, "learning_rate": 5.9250909502294526e-06, "loss": 0.3029, "step": 5030 }, { "epoch": 0.29810137812740284, "grad_norm": 2.9467926025390625, "learning_rate": 5.924664943368963e-06, "loss": 0.3594, "step": 5040 }, { "epoch": 0.2986928491157509, "grad_norm": 4.920210838317871, "learning_rate": 5.924237743994987e-06, "loss": 0.3545, "step": 5050 }, { "epoch": 0.2992843201040989, "grad_norm": 3.7383148670196533, "learning_rate": 5.923809352281714e-06, "loss": 0.3496, "step": 5060 }, { "epoch": 0.29987579109244694, "grad_norm": 3.9579849243164062, "learning_rate": 5.923379768403817e-06, "loss": 0.3568, "step": 5070 }, { "epoch": 0.3000532323889513, "eval_accuracy": 0.8499683933859001, "eval_loss": 0.34088680148124695, "eval_runtime": 195.3506, "eval_safe_aucpr": 0.9069012942611618, "eval_safe_f1": 0.8260861181280009, "eval_safe_fpr": 0.11288152820543447, "eval_safe_precision": 0.8501349420542943, "eval_safe_recall": 0.8033604620635337, "eval_samples_per_second": 307.724, "eval_steps_per_second": 38.469, "eval_unsafe_aucpr": 0.9479658645801321, "eval_unsafe_f1": 0.8680834881305854, "eval_unsafe_fpr": 0.19663953793646555, "eval_unsafe_precision": 0.849848215819921, "eval_unsafe_recall": 0.8871184717945652, "step": 5073 }, { "epoch": 0.30046726208079494, "grad_norm": 3.9788262844085693, "learning_rate": 5.922948992536456e-06, "loss": 0.4177, "step": 5080 }, { "epoch": 0.30105873306914294, "grad_norm": 4.025001049041748, "learning_rate": 5.9225170248552775e-06, "loss": 0.3225, "step": 5090 }, { "epoch": 0.301650204057491, "grad_norm": 5.060151100158691, "learning_rate": 5.922083865536414e-06, "loss": 0.3558, "step": 5100 }, { "epoch": 0.302241675045839, "grad_norm": 3.667146682739258, "learning_rate": 5.921649514756483e-06, "loss": 0.3219, "step": 5110 }, { "epoch": 0.30283314603418704, "grad_norm": 3.5262818336486816, "learning_rate": 5.921213972692587e-06, "loss": 0.2832, "step": 5120 }, { "epoch": 0.30342461702253504, "grad_norm": 6.8113932609558105, "learning_rate": 5.920777239522318e-06, "loss": 0.4091, "step": 5130 }, { "epoch": 0.3040160880108831, "grad_norm": 3.2919890880584717, "learning_rate": 5.920339315423749e-06, "loss": 0.3279, "step": 5140 }, { "epoch": 0.3046075589992311, "grad_norm": 7.52615213394165, "learning_rate": 5.919900200575442e-06, "loss": 0.3308, "step": 5150 }, { "epoch": 0.3051990299875791, "grad_norm": 5.625899791717529, "learning_rate": 5.919459895156443e-06, "loss": 0.351, "step": 5160 }, { "epoch": 0.30579050097592714, "grad_norm": 3.5103766918182373, "learning_rate": 5.919018399346284e-06, "loss": 0.3893, "step": 5170 }, { "epoch": 0.30638197196427513, "grad_norm": 2.8448448181152344, "learning_rate": 5.918575713324981e-06, "loss": 0.3104, "step": 5180 }, { "epoch": 0.3069734429526232, "grad_norm": 5.322744846343994, "learning_rate": 5.918131837273038e-06, "loss": 0.315, "step": 5190 }, { "epoch": 0.3075649139409712, "grad_norm": 5.191041469573975, "learning_rate": 5.917686771371441e-06, "loss": 0.3219, "step": 5200 }, { "epoch": 0.30815638492931924, "grad_norm": 2.720644235610962, "learning_rate": 5.9172405158016646e-06, "loss": 0.3441, "step": 5210 }, { "epoch": 0.30874785591766724, "grad_norm": 12.33743667602539, "learning_rate": 5.916793070745665e-06, "loss": 0.3801, "step": 5220 }, { "epoch": 0.30933932690601523, "grad_norm": 4.348657131195068, "learning_rate": 5.916344436385885e-06, "loss": 0.3029, "step": 5230 }, { "epoch": 0.3099307978943633, "grad_norm": 2.528801679611206, "learning_rate": 5.915894612905254e-06, "loss": 0.3711, "step": 5240 }, { "epoch": 0.3105222688827113, "grad_norm": 4.697808742523193, "learning_rate": 5.915443600487184e-06, "loss": 0.3397, "step": 5250 }, { "epoch": 0.31111373987105934, "grad_norm": 4.518587112426758, "learning_rate": 5.9149913993155705e-06, "loss": 0.3394, "step": 5260 }, { "epoch": 0.31170521085940733, "grad_norm": 5.199132442474365, "learning_rate": 5.914538009574799e-06, "loss": 0.3499, "step": 5270 }, { "epoch": 0.3122966818477554, "grad_norm": 4.121863842010498, "learning_rate": 5.914083431449734e-06, "loss": 0.3582, "step": 5280 }, { "epoch": 0.3128881528361034, "grad_norm": 1.6972838640213013, "learning_rate": 5.913627665125728e-06, "loss": 0.2851, "step": 5290 }, { "epoch": 0.31347962382445144, "grad_norm": 9.634567260742188, "learning_rate": 5.913170710788615e-06, "loss": 0.3794, "step": 5300 }, { "epoch": 0.31407109481279943, "grad_norm": 4.62807559967041, "learning_rate": 5.912712568624718e-06, "loss": 0.3211, "step": 5310 }, { "epoch": 0.31466256580114743, "grad_norm": 6.242055416107178, "learning_rate": 5.912253238820839e-06, "loss": 0.3855, "step": 5320 }, { "epoch": 0.3152540367894955, "grad_norm": 4.210643768310547, "learning_rate": 5.911792721564268e-06, "loss": 0.3317, "step": 5330 }, { "epoch": 0.3158455077778435, "grad_norm": 3.8272929191589355, "learning_rate": 5.911331017042778e-06, "loss": 0.3727, "step": 5340 }, { "epoch": 0.31643697876619153, "grad_norm": 4.428116321563721, "learning_rate": 5.910868125444626e-06, "loss": 0.3646, "step": 5350 }, { "epoch": 0.31702844975453953, "grad_norm": 4.637458324432373, "learning_rate": 5.910404046958553e-06, "loss": 0.3643, "step": 5360 }, { "epoch": 0.3176199207428876, "grad_norm": 3.5912392139434814, "learning_rate": 5.909938781773783e-06, "loss": 0.315, "step": 5370 }, { "epoch": 0.3182113917312356, "grad_norm": 3.8700063228607178, "learning_rate": 5.909472330080028e-06, "loss": 0.3238, "step": 5380 }, { "epoch": 0.3188028627195836, "grad_norm": 4.393209934234619, "learning_rate": 5.9090046920674765e-06, "loss": 0.3668, "step": 5390 }, { "epoch": 0.31939433370793163, "grad_norm": 3.342196226119995, "learning_rate": 5.908535867926807e-06, "loss": 0.3089, "step": 5400 }, { "epoch": 0.31998580469627963, "grad_norm": 3.430628776550293, "learning_rate": 5.908065857849181e-06, "loss": 0.3135, "step": 5410 }, { "epoch": 0.3205772756846277, "grad_norm": 6.852587699890137, "learning_rate": 5.9075946620262405e-06, "loss": 0.3916, "step": 5420 }, { "epoch": 0.3211687466729757, "grad_norm": 3.816279888153076, "learning_rate": 5.907122280650113e-06, "loss": 0.3794, "step": 5430 }, { "epoch": 0.32176021766132373, "grad_norm": 2.9171488285064697, "learning_rate": 5.9066487139134085e-06, "loss": 0.3752, "step": 5440 }, { "epoch": 0.32235168864967173, "grad_norm": 4.847435474395752, "learning_rate": 5.9061739620092205e-06, "loss": 0.366, "step": 5450 }, { "epoch": 0.32294315963801973, "grad_norm": 4.161375999450684, "learning_rate": 5.9056980251311275e-06, "loss": 0.3239, "step": 5460 }, { "epoch": 0.3235346306263678, "grad_norm": 5.4873175621032715, "learning_rate": 5.905220903473188e-06, "loss": 0.4342, "step": 5470 }, { "epoch": 0.3241261016147158, "grad_norm": 6.188025951385498, "learning_rate": 5.904742597229947e-06, "loss": 0.3701, "step": 5480 }, { "epoch": 0.32471757260306383, "grad_norm": 4.6967034339904785, "learning_rate": 5.90426310659643e-06, "loss": 0.3174, "step": 5490 }, { "epoch": 0.32530904359141183, "grad_norm": 3.154341459274292, "learning_rate": 5.903782431768147e-06, "loss": 0.2867, "step": 5500 }, { "epoch": 0.3259005145797599, "grad_norm": 5.039037704467773, "learning_rate": 5.903300572941088e-06, "loss": 0.3012, "step": 5510 }, { "epoch": 0.3264919855681079, "grad_norm": 4.017624378204346, "learning_rate": 5.90281753031173e-06, "loss": 0.3552, "step": 5520 }, { "epoch": 0.32708345655645593, "grad_norm": 3.357999324798584, "learning_rate": 5.902333304077029e-06, "loss": 0.395, "step": 5530 }, { "epoch": 0.32767492754480393, "grad_norm": 4.182588577270508, "learning_rate": 5.9018478944344265e-06, "loss": 0.3327, "step": 5540 }, { "epoch": 0.3282663985331519, "grad_norm": 4.013810157775879, "learning_rate": 5.901361301581845e-06, "loss": 0.3698, "step": 5550 }, { "epoch": 0.3288578695215, "grad_norm": 4.19458532333374, "learning_rate": 5.9008735257176894e-06, "loss": 0.3323, "step": 5560 }, { "epoch": 0.329449340509848, "grad_norm": 4.9410552978515625, "learning_rate": 5.900384567040847e-06, "loss": 0.3632, "step": 5570 }, { "epoch": 0.33004081149819603, "grad_norm": 3.385518789291382, "learning_rate": 5.899894425750687e-06, "loss": 0.4044, "step": 5580 }, { "epoch": 0.330632282486544, "grad_norm": 3.1909101009368896, "learning_rate": 5.899403102047063e-06, "loss": 0.3822, "step": 5590 }, { "epoch": 0.3312237534748921, "grad_norm": 5.03477668762207, "learning_rate": 5.898910596130308e-06, "loss": 0.3605, "step": 5600 }, { "epoch": 0.3318152244632401, "grad_norm": 3.72001576423645, "learning_rate": 5.898416908201238e-06, "loss": 0.2902, "step": 5610 }, { "epoch": 0.3324066954515881, "grad_norm": 4.320418357849121, "learning_rate": 5.8979220384611506e-06, "loss": 0.3118, "step": 5620 }, { "epoch": 0.3329981664399361, "grad_norm": 3.1242363452911377, "learning_rate": 5.897425987111829e-06, "loss": 0.3357, "step": 5630 }, { "epoch": 0.3335896374282841, "grad_norm": 2.418677806854248, "learning_rate": 5.89692875435553e-06, "loss": 0.4097, "step": 5640 }, { "epoch": 0.3341811084166322, "grad_norm": 3.738186836242676, "learning_rate": 5.8964303403950015e-06, "loss": 0.3076, "step": 5650 }, { "epoch": 0.3347725794049802, "grad_norm": 4.3074235916137695, "learning_rate": 5.8959307454334655e-06, "loss": 0.3594, "step": 5660 }, { "epoch": 0.33536405039332823, "grad_norm": 2.405959367752075, "learning_rate": 5.8954299696746305e-06, "loss": 0.3173, "step": 5670 }, { "epoch": 0.3359555213816762, "grad_norm": 3.0738086700439453, "learning_rate": 5.894928013322683e-06, "loss": 0.3152, "step": 5680 }, { "epoch": 0.3365469923700242, "grad_norm": 4.173305988311768, "learning_rate": 5.894424876582295e-06, "loss": 0.3205, "step": 5690 }, { "epoch": 0.3371384633583723, "grad_norm": 5.547184467315674, "learning_rate": 5.893920559658616e-06, "loss": 0.3216, "step": 5700 }, { "epoch": 0.3377299343467203, "grad_norm": 4.8621978759765625, "learning_rate": 5.893415062757277e-06, "loss": 0.2683, "step": 5710 }, { "epoch": 0.3383214053350683, "grad_norm": 5.318309307098389, "learning_rate": 5.892908386084393e-06, "loss": 0.3853, "step": 5720 }, { "epoch": 0.3389128763234163, "grad_norm": 4.587392330169678, "learning_rate": 5.892400529846556e-06, "loss": 0.2864, "step": 5730 }, { "epoch": 0.3395043473117644, "grad_norm": 4.846919059753418, "learning_rate": 5.891891494250843e-06, "loss": 0.3148, "step": 5740 }, { "epoch": 0.3400958183001124, "grad_norm": 6.083952903747559, "learning_rate": 5.891381279504808e-06, "loss": 0.3693, "step": 5750 }, { "epoch": 0.3406872892884604, "grad_norm": 3.69539475440979, "learning_rate": 5.890869885816491e-06, "loss": 0.3175, "step": 5760 }, { "epoch": 0.3412787602768084, "grad_norm": 2.805887460708618, "learning_rate": 5.890357313394407e-06, "loss": 0.3532, "step": 5770 }, { "epoch": 0.3418702312651564, "grad_norm": 3.5093283653259277, "learning_rate": 5.889843562447556e-06, "loss": 0.2727, "step": 5780 }, { "epoch": 0.3424617022535045, "grad_norm": 2.9825711250305176, "learning_rate": 5.889328633185414e-06, "loss": 0.3475, "step": 5790 }, { "epoch": 0.34305317324185247, "grad_norm": 4.78188943862915, "learning_rate": 5.888812525817942e-06, "loss": 0.2936, "step": 5800 }, { "epoch": 0.3436446442302005, "grad_norm": 3.4061827659606934, "learning_rate": 5.88829524055558e-06, "loss": 0.4088, "step": 5810 }, { "epoch": 0.3442361152185485, "grad_norm": 4.728603363037109, "learning_rate": 5.887776777609245e-06, "loss": 0.3509, "step": 5820 }, { "epoch": 0.3448275862068966, "grad_norm": 2.8141345977783203, "learning_rate": 5.88725713719034e-06, "loss": 0.3716, "step": 5830 }, { "epoch": 0.34541905719524457, "grad_norm": 2.323753595352173, "learning_rate": 5.886736319510744e-06, "loss": 0.3413, "step": 5840 }, { "epoch": 0.34601052818359257, "grad_norm": 3.3416576385498047, "learning_rate": 5.886214324782816e-06, "loss": 0.3222, "step": 5850 }, { "epoch": 0.3466019991719406, "grad_norm": 4.863747596740723, "learning_rate": 5.885691153219398e-06, "loss": 0.4073, "step": 5860 }, { "epoch": 0.3471934701602886, "grad_norm": 3.3804848194122314, "learning_rate": 5.88516680503381e-06, "loss": 0.4006, "step": 5870 }, { "epoch": 0.3477849411486367, "grad_norm": 5.0079545974731445, "learning_rate": 5.884641280439848e-06, "loss": 0.3535, "step": 5880 }, { "epoch": 0.34837641213698467, "grad_norm": 3.8826191425323486, "learning_rate": 5.884114579651796e-06, "loss": 0.3443, "step": 5890 }, { "epoch": 0.3489678831253327, "grad_norm": 6.46497106552124, "learning_rate": 5.88358670288441e-06, "loss": 0.3111, "step": 5900 }, { "epoch": 0.3495593541136807, "grad_norm": 3.9818084239959717, "learning_rate": 5.883057650352929e-06, "loss": 0.3699, "step": 5910 }, { "epoch": 0.3501508251020288, "grad_norm": 3.6611902713775635, "learning_rate": 5.8825274222730705e-06, "loss": 0.3701, "step": 5920 }, { "epoch": 0.35074229609037677, "grad_norm": 3.002556562423706, "learning_rate": 5.881996018861032e-06, "loss": 0.3104, "step": 5930 }, { "epoch": 0.35133376707872477, "grad_norm": 4.159395217895508, "learning_rate": 5.88146344033349e-06, "loss": 0.3592, "step": 5940 }, { "epoch": 0.3519252380670728, "grad_norm": 3.6778438091278076, "learning_rate": 5.880929686907601e-06, "loss": 0.2871, "step": 5950 }, { "epoch": 0.3525167090554208, "grad_norm": 4.151058673858643, "learning_rate": 5.880394758800997e-06, "loss": 0.3501, "step": 5960 }, { "epoch": 0.35310818004376887, "grad_norm": 4.144840240478516, "learning_rate": 5.879858656231793e-06, "loss": 0.4319, "step": 5970 }, { "epoch": 0.35369965103211687, "grad_norm": 4.845097064971924, "learning_rate": 5.879321379418582e-06, "loss": 0.3243, "step": 5980 }, { "epoch": 0.3542911220204649, "grad_norm": 3.0042994022369385, "learning_rate": 5.878782928580434e-06, "loss": 0.3956, "step": 5990 }, { "epoch": 0.3548825930088129, "grad_norm": 3.341571092605591, "learning_rate": 5.8782433039369006e-06, "loss": 0.3865, "step": 6000 }, { "epoch": 0.3554740639971609, "grad_norm": 4.159686088562012, "learning_rate": 5.877702505708008e-06, "loss": 0.36, "step": 6010 }, { "epoch": 0.35606553498550897, "grad_norm": 3.634249448776245, "learning_rate": 5.877160534114265e-06, "loss": 0.3422, "step": 6020 }, { "epoch": 0.35665700597385697, "grad_norm": 3.7531983852386475, "learning_rate": 5.876617389376656e-06, "loss": 0.3271, "step": 6030 }, { "epoch": 0.357248476962205, "grad_norm": 2.7075181007385254, "learning_rate": 5.8760730717166456e-06, "loss": 0.3582, "step": 6040 }, { "epoch": 0.357839947950553, "grad_norm": 5.431192874908447, "learning_rate": 5.875527581356175e-06, "loss": 0.3549, "step": 6050 }, { "epoch": 0.35843141893890107, "grad_norm": 2.69948148727417, "learning_rate": 5.874980918517665e-06, "loss": 0.3304, "step": 6060 }, { "epoch": 0.35902288992724907, "grad_norm": 3.393376588821411, "learning_rate": 5.874433083424014e-06, "loss": 0.3646, "step": 6070 }, { "epoch": 0.35961436091559706, "grad_norm": 4.417396068572998, "learning_rate": 5.873884076298599e-06, "loss": 0.3422, "step": 6080 }, { "epoch": 0.3602058319039451, "grad_norm": 2.8824353218078613, "learning_rate": 5.8733338973652726e-06, "loss": 0.3109, "step": 6090 }, { "epoch": 0.3607973028922931, "grad_norm": 6.361566543579102, "learning_rate": 5.872782546848366e-06, "loss": 0.3368, "step": 6100 }, { "epoch": 0.36138877388064117, "grad_norm": 5.078776836395264, "learning_rate": 5.872230024972692e-06, "loss": 0.3947, "step": 6110 }, { "epoch": 0.36198024486898916, "grad_norm": 3.5751898288726807, "learning_rate": 5.871676331963535e-06, "loss": 0.3555, "step": 6120 }, { "epoch": 0.3625717158573372, "grad_norm": 2.813054084777832, "learning_rate": 5.871121468046662e-06, "loss": 0.3264, "step": 6130 }, { "epoch": 0.3631631868456852, "grad_norm": 3.0372908115386963, "learning_rate": 5.870565433448314e-06, "loss": 0.3381, "step": 6140 }, { "epoch": 0.36375465783403327, "grad_norm": 7.198215961456299, "learning_rate": 5.87000822839521e-06, "loss": 0.3235, "step": 6150 }, { "epoch": 0.36434612882238127, "grad_norm": 4.296572685241699, "learning_rate": 5.869449853114548e-06, "loss": 0.376, "step": 6160 }, { "epoch": 0.36493759981072926, "grad_norm": 2.1846959590911865, "learning_rate": 5.868890307834001e-06, "loss": 0.3576, "step": 6170 }, { "epoch": 0.3655290707990773, "grad_norm": 3.6156129837036133, "learning_rate": 5.868329592781721e-06, "loss": 0.3989, "step": 6180 }, { "epoch": 0.3661205417874253, "grad_norm": 2.79758882522583, "learning_rate": 5.867767708186335e-06, "loss": 0.3769, "step": 6190 }, { "epoch": 0.36671201277577337, "grad_norm": 5.314364910125732, "learning_rate": 5.867204654276948e-06, "loss": 0.4051, "step": 6200 }, { "epoch": 0.36730348376412136, "grad_norm": 3.619029998779297, "learning_rate": 5.866640431283143e-06, "loss": 0.3768, "step": 6210 }, { "epoch": 0.3678949547524694, "grad_norm": 4.152932643890381, "learning_rate": 5.866075039434977e-06, "loss": 0.3726, "step": 6220 }, { "epoch": 0.3684864257408174, "grad_norm": 3.1290571689605713, "learning_rate": 5.865508478962985e-06, "loss": 0.3901, "step": 6230 }, { "epoch": 0.3690778967291654, "grad_norm": 3.5193254947662354, "learning_rate": 5.864940750098178e-06, "loss": 0.3909, "step": 6240 }, { "epoch": 0.36966936771751346, "grad_norm": 3.677345037460327, "learning_rate": 5.864371853072046e-06, "loss": 0.3139, "step": 6250 }, { "epoch": 0.37026083870586146, "grad_norm": 3.848825216293335, "learning_rate": 5.863801788116551e-06, "loss": 0.2888, "step": 6260 }, { "epoch": 0.3708523096942095, "grad_norm": 3.439356803894043, "learning_rate": 5.863230555464133e-06, "loss": 0.3544, "step": 6270 }, { "epoch": 0.3714437806825575, "grad_norm": 3.553623676300049, "learning_rate": 5.862658155347711e-06, "loss": 0.2578, "step": 6280 }, { "epoch": 0.37203525167090556, "grad_norm": 3.3316032886505127, "learning_rate": 5.862084588000676e-06, "loss": 0.3031, "step": 6290 }, { "epoch": 0.37262672265925356, "grad_norm": 3.4183731079101562, "learning_rate": 5.861509853656894e-06, "loss": 0.2961, "step": 6300 }, { "epoch": 0.37321819364760156, "grad_norm": 7.368538856506348, "learning_rate": 5.860933952550714e-06, "loss": 0.3409, "step": 6310 }, { "epoch": 0.3738096646359496, "grad_norm": 3.6013472080230713, "learning_rate": 5.860356884916953e-06, "loss": 0.3494, "step": 6320 }, { "epoch": 0.3744011356242976, "grad_norm": 4.5222487449646, "learning_rate": 5.859778650990907e-06, "loss": 0.353, "step": 6330 }, { "epoch": 0.37499260661264566, "grad_norm": 3.539823055267334, "learning_rate": 5.859199251008347e-06, "loss": 0.3151, "step": 6340 }, { "epoch": 0.37558407760099366, "grad_norm": 4.358543395996094, "learning_rate": 5.8586186852055216e-06, "loss": 0.2955, "step": 6350 }, { "epoch": 0.3761755485893417, "grad_norm": 2.708253860473633, "learning_rate": 5.85803695381915e-06, "loss": 0.3357, "step": 6360 }, { "epoch": 0.3767670195776897, "grad_norm": 2.8288822174072266, "learning_rate": 5.857454057086431e-06, "loss": 0.3702, "step": 6370 }, { "epoch": 0.37735849056603776, "grad_norm": 3.867382526397705, "learning_rate": 5.856869995245038e-06, "loss": 0.3437, "step": 6380 }, { "epoch": 0.37794996155438576, "grad_norm": 2.254829168319702, "learning_rate": 5.856284768533116e-06, "loss": 0.2997, "step": 6390 }, { "epoch": 0.37854143254273376, "grad_norm": 4.928765773773193, "learning_rate": 5.855698377189288e-06, "loss": 0.304, "step": 6400 }, { "epoch": 0.3791329035310818, "grad_norm": 3.240772247314453, "learning_rate": 5.855110821452654e-06, "loss": 0.3286, "step": 6410 }, { "epoch": 0.3797243745194298, "grad_norm": 3.2659780979156494, "learning_rate": 5.854522101562783e-06, "loss": 0.3032, "step": 6420 }, { "epoch": 0.38031584550777786, "grad_norm": 1.8397643566131592, "learning_rate": 5.853932217759724e-06, "loss": 0.3487, "step": 6430 }, { "epoch": 0.38090731649612586, "grad_norm": 3.1508238315582275, "learning_rate": 5.853341170283996e-06, "loss": 0.3426, "step": 6440 }, { "epoch": 0.3814987874844739, "grad_norm": 4.046362400054932, "learning_rate": 5.852748959376598e-06, "loss": 0.3924, "step": 6450 }, { "epoch": 0.3820902584728219, "grad_norm": 3.722952365875244, "learning_rate": 5.852155585278996e-06, "loss": 0.3604, "step": 6460 }, { "epoch": 0.3826817294611699, "grad_norm": 2.4273293018341064, "learning_rate": 5.851561048233139e-06, "loss": 0.3302, "step": 6470 }, { "epoch": 0.38327320044951796, "grad_norm": 4.294403553009033, "learning_rate": 5.850965348481443e-06, "loss": 0.4083, "step": 6480 }, { "epoch": 0.38386467143786596, "grad_norm": 4.042994022369385, "learning_rate": 5.850368486266802e-06, "loss": 0.3255, "step": 6490 }, { "epoch": 0.384456142426214, "grad_norm": 4.579367637634277, "learning_rate": 5.8497704618325815e-06, "loss": 0.3059, "step": 6500 }, { "epoch": 0.385047613414562, "grad_norm": 3.80751371383667, "learning_rate": 5.849171275422623e-06, "loss": 0.3227, "step": 6510 }, { "epoch": 0.38563908440291006, "grad_norm": 2.7446677684783936, "learning_rate": 5.848570927281241e-06, "loss": 0.2966, "step": 6520 }, { "epoch": 0.38623055539125806, "grad_norm": 3.460026502609253, "learning_rate": 5.847969417653222e-06, "loss": 0.3531, "step": 6530 }, { "epoch": 0.38682202637960605, "grad_norm": 4.988143444061279, "learning_rate": 5.84736674678383e-06, "loss": 0.2855, "step": 6540 }, { "epoch": 0.3874134973679541, "grad_norm": 4.74730920791626, "learning_rate": 5.8467629149188e-06, "loss": 0.3984, "step": 6550 }, { "epoch": 0.3880049683563021, "grad_norm": 2.73442006111145, "learning_rate": 5.846157922304339e-06, "loss": 0.4189, "step": 6560 }, { "epoch": 0.38859643934465016, "grad_norm": 3.966217517852783, "learning_rate": 5.845551769187129e-06, "loss": 0.3487, "step": 6570 }, { "epoch": 0.38918791033299815, "grad_norm": 3.2719802856445312, "learning_rate": 5.844944455814328e-06, "loss": 0.3495, "step": 6580 }, { "epoch": 0.3897793813213462, "grad_norm": 3.63812255859375, "learning_rate": 5.84433598243356e-06, "loss": 0.3157, "step": 6590 }, { "epoch": 0.3903708523096942, "grad_norm": 5.844852447509766, "learning_rate": 5.84372634929293e-06, "loss": 0.3628, "step": 6600 }, { "epoch": 0.39096232329804226, "grad_norm": 3.142457962036133, "learning_rate": 5.843115556641009e-06, "loss": 0.3352, "step": 6610 }, { "epoch": 0.39155379428639026, "grad_norm": 2.573864459991455, "learning_rate": 5.842503604726844e-06, "loss": 0.3957, "step": 6620 }, { "epoch": 0.39214526527473825, "grad_norm": 3.454551935195923, "learning_rate": 5.8418904937999575e-06, "loss": 0.3379, "step": 6630 }, { "epoch": 0.3927367362630863, "grad_norm": 2.2886226177215576, "learning_rate": 5.841276224110338e-06, "loss": 0.327, "step": 6640 }, { "epoch": 0.3933282072514343, "grad_norm": 4.825653553009033, "learning_rate": 5.840660795908452e-06, "loss": 0.4036, "step": 6650 }, { "epoch": 0.39391967823978236, "grad_norm": 2.4120705127716064, "learning_rate": 5.840044209445237e-06, "loss": 0.3683, "step": 6660 }, { "epoch": 0.39451114922813035, "grad_norm": 2.863158702850342, "learning_rate": 5.8394264649721006e-06, "loss": 0.3391, "step": 6670 }, { "epoch": 0.3951026202164784, "grad_norm": 2.5085294246673584, "learning_rate": 5.838807562740925e-06, "loss": 0.2921, "step": 6680 }, { "epoch": 0.3956940912048264, "grad_norm": 2.6497418880462646, "learning_rate": 5.838187503004063e-06, "loss": 0.338, "step": 6690 }, { "epoch": 0.3962855621931744, "grad_norm": 5.180778503417969, "learning_rate": 5.83756628601434e-06, "loss": 0.3241, "step": 6700 }, { "epoch": 0.39687703318152245, "grad_norm": 3.3266122341156006, "learning_rate": 5.8369439120250536e-06, "loss": 0.3787, "step": 6710 }, { "epoch": 0.39746850416987045, "grad_norm": 3.426403760910034, "learning_rate": 5.836320381289973e-06, "loss": 0.2572, "step": 6720 }, { "epoch": 0.3980599751582185, "grad_norm": 4.1935200691223145, "learning_rate": 5.83569569406334e-06, "loss": 0.3549, "step": 6730 }, { "epoch": 0.3986514461465665, "grad_norm": 5.015183448791504, "learning_rate": 5.835069850599865e-06, "loss": 0.2616, "step": 6740 }, { "epoch": 0.39924291713491455, "grad_norm": 5.879730701446533, "learning_rate": 5.834442851154733e-06, "loss": 0.364, "step": 6750 }, { "epoch": 0.39983438812326255, "grad_norm": 3.0781447887420654, "learning_rate": 5.833814695983596e-06, "loss": 0.3683, "step": 6760 }, { "epoch": 0.40042585911161055, "grad_norm": 3.1903834342956543, "learning_rate": 5.833185385342585e-06, "loss": 0.3201, "step": 6770 }, { "epoch": 0.4010173300999586, "grad_norm": 7.266666889190674, "learning_rate": 5.832554919488294e-06, "loss": 0.3322, "step": 6780 }, { "epoch": 0.4016088010883066, "grad_norm": 2.893882989883423, "learning_rate": 5.831923298677793e-06, "loss": 0.3771, "step": 6790 }, { "epoch": 0.40220027207665465, "grad_norm": 4.754356861114502, "learning_rate": 5.83129052316862e-06, "loss": 0.3058, "step": 6800 }, { "epoch": 0.40279174306500265, "grad_norm": 3.94140362739563, "learning_rate": 5.830656593218786e-06, "loss": 0.3205, "step": 6810 }, { "epoch": 0.4033832140533507, "grad_norm": 4.626865386962891, "learning_rate": 5.8300215090867735e-06, "loss": 0.3926, "step": 6820 }, { "epoch": 0.4039746850416987, "grad_norm": 3.1377575397491455, "learning_rate": 5.82938527103153e-06, "loss": 0.3427, "step": 6830 }, { "epoch": 0.40456615603004675, "grad_norm": 4.025646209716797, "learning_rate": 5.828747879312481e-06, "loss": 0.3978, "step": 6840 }, { "epoch": 0.40515762701839475, "grad_norm": 3.615919351577759, "learning_rate": 5.828109334189518e-06, "loss": 0.304, "step": 6850 }, { "epoch": 0.40574909800674275, "grad_norm": 2.724827527999878, "learning_rate": 5.827469635923004e-06, "loss": 0.3162, "step": 6860 }, { "epoch": 0.4063405689950908, "grad_norm": 2.4393653869628906, "learning_rate": 5.826828784773771e-06, "loss": 0.3215, "step": 6870 }, { "epoch": 0.4069320399834388, "grad_norm": 3.286824941635132, "learning_rate": 5.826186781003123e-06, "loss": 0.3176, "step": 6880 }, { "epoch": 0.40752351097178685, "grad_norm": 3.9131083488464355, "learning_rate": 5.82554362487283e-06, "loss": 0.3285, "step": 6890 }, { "epoch": 0.40811498196013485, "grad_norm": 3.8140616416931152, "learning_rate": 5.824899316645139e-06, "loss": 0.3486, "step": 6900 }, { "epoch": 0.4087064529484829, "grad_norm": 3.503833532333374, "learning_rate": 5.824253856582759e-06, "loss": 0.3066, "step": 6910 }, { "epoch": 0.4092979239368309, "grad_norm": 1.9727652072906494, "learning_rate": 5.823607244948876e-06, "loss": 0.3546, "step": 6920 }, { "epoch": 0.4098893949251789, "grad_norm": 2.628587007522583, "learning_rate": 5.82295948200714e-06, "loss": 0.3263, "step": 6930 }, { "epoch": 0.41048086591352695, "grad_norm": 2.374650239944458, "learning_rate": 5.82231056802167e-06, "loss": 0.2988, "step": 6940 }, { "epoch": 0.41107233690187495, "grad_norm": 4.1091532707214355, "learning_rate": 5.821660503257059e-06, "loss": 0.3141, "step": 6950 }, { "epoch": 0.411663807890223, "grad_norm": 4.625342845916748, "learning_rate": 5.821009287978367e-06, "loss": 0.3654, "step": 6960 }, { "epoch": 0.412255278878571, "grad_norm": 2.7168726921081543, "learning_rate": 5.820356922451121e-06, "loss": 0.329, "step": 6970 }, { "epoch": 0.41284674986691905, "grad_norm": 2.289541244506836, "learning_rate": 5.8197034069413215e-06, "loss": 0.3585, "step": 6980 }, { "epoch": 0.41343822085526705, "grad_norm": 6.639915943145752, "learning_rate": 5.819048741715432e-06, "loss": 0.3077, "step": 6990 }, { "epoch": 0.41402969184361504, "grad_norm": 6.101804256439209, "learning_rate": 5.818392927040391e-06, "loss": 0.3419, "step": 7000 }, { "epoch": 0.4146211628319631, "grad_norm": 4.220304012298584, "learning_rate": 5.8177359631836015e-06, "loss": 0.3229, "step": 7010 }, { "epoch": 0.4152126338203111, "grad_norm": 4.547323226928711, "learning_rate": 5.817077850412936e-06, "loss": 0.3256, "step": 7020 }, { "epoch": 0.41580410480865915, "grad_norm": 4.403805255889893, "learning_rate": 5.816418588996737e-06, "loss": 0.3388, "step": 7030 }, { "epoch": 0.41639557579700714, "grad_norm": 2.00107479095459, "learning_rate": 5.815758179203812e-06, "loss": 0.3627, "step": 7040 }, { "epoch": 0.4169870467853552, "grad_norm": 3.0551116466522217, "learning_rate": 5.815096621303441e-06, "loss": 0.3312, "step": 7050 }, { "epoch": 0.4175785177737032, "grad_norm": 3.3686106204986572, "learning_rate": 5.814433915565369e-06, "loss": 0.2857, "step": 7060 }, { "epoch": 0.41816998876205125, "grad_norm": 3.465566635131836, "learning_rate": 5.81377006225981e-06, "loss": 0.3189, "step": 7070 }, { "epoch": 0.41876145975039925, "grad_norm": 2.7168960571289062, "learning_rate": 5.8131050616574466e-06, "loss": 0.4101, "step": 7080 }, { "epoch": 0.41935293073874724, "grad_norm": 2.9596924781799316, "learning_rate": 5.812438914029429e-06, "loss": 0.3498, "step": 7090 }, { "epoch": 0.4199444017270953, "grad_norm": 5.308952331542969, "learning_rate": 5.8117716196473725e-06, "loss": 0.3021, "step": 7100 }, { "epoch": 0.4205358727154433, "grad_norm": 6.6363911628723145, "learning_rate": 5.8111031787833645e-06, "loss": 0.3833, "step": 7110 }, { "epoch": 0.42112734370379135, "grad_norm": 2.676182270050049, "learning_rate": 5.810433591709956e-06, "loss": 0.3545, "step": 7120 }, { "epoch": 0.42171881469213934, "grad_norm": 2.5218212604522705, "learning_rate": 5.809762858700167e-06, "loss": 0.361, "step": 7130 }, { "epoch": 0.4223102856804874, "grad_norm": 2.402456283569336, "learning_rate": 5.809090980027485e-06, "loss": 0.3519, "step": 7140 }, { "epoch": 0.4229017566688354, "grad_norm": 3.3567137718200684, "learning_rate": 5.808417955965864e-06, "loss": 0.3831, "step": 7150 }, { "epoch": 0.4234932276571834, "grad_norm": 1.9695286750793457, "learning_rate": 5.807743786789726e-06, "loss": 0.317, "step": 7160 }, { "epoch": 0.42408469864553144, "grad_norm": 3.6184346675872803, "learning_rate": 5.807068472773959e-06, "loss": 0.3783, "step": 7170 }, { "epoch": 0.42467616963387944, "grad_norm": 3.6566648483276367, "learning_rate": 5.806392014193916e-06, "loss": 0.4007, "step": 7180 }, { "epoch": 0.4252676406222275, "grad_norm": 2.234712600708008, "learning_rate": 5.805714411325422e-06, "loss": 0.3595, "step": 7190 }, { "epoch": 0.4258591116105755, "grad_norm": 6.07689094543457, "learning_rate": 5.805035664444763e-06, "loss": 0.3138, "step": 7200 }, { "epoch": 0.42645058259892354, "grad_norm": 3.8110804557800293, "learning_rate": 5.804355773828694e-06, "loss": 0.3294, "step": 7210 }, { "epoch": 0.42704205358727154, "grad_norm": 3.2936794757843018, "learning_rate": 5.803674739754436e-06, "loss": 0.3463, "step": 7220 }, { "epoch": 0.42763352457561954, "grad_norm": 4.606428146362305, "learning_rate": 5.8029925624996775e-06, "loss": 0.3661, "step": 7230 }, { "epoch": 0.4282249955639676, "grad_norm": 3.2391107082366943, "learning_rate": 5.80230924234257e-06, "loss": 0.2732, "step": 7240 }, { "epoch": 0.4288164665523156, "grad_norm": 4.212532997131348, "learning_rate": 5.801624779561735e-06, "loss": 0.3598, "step": 7250 }, { "epoch": 0.42940793754066364, "grad_norm": 2.8464412689208984, "learning_rate": 5.8009391744362565e-06, "loss": 0.327, "step": 7260 }, { "epoch": 0.42999940852901164, "grad_norm": 2.991882085800171, "learning_rate": 5.800252427245686e-06, "loss": 0.3015, "step": 7270 }, { "epoch": 0.4305908795173597, "grad_norm": 6.02423095703125, "learning_rate": 5.79956453827004e-06, "loss": 0.3333, "step": 7280 }, { "epoch": 0.4311823505057077, "grad_norm": 4.377829551696777, "learning_rate": 5.798875507789801e-06, "loss": 0.353, "step": 7290 }, { "epoch": 0.43177382149405574, "grad_norm": 3.295513391494751, "learning_rate": 5.798185336085917e-06, "loss": 0.3219, "step": 7300 }, { "epoch": 0.43236529248240374, "grad_norm": 2.898085355758667, "learning_rate": 5.797494023439801e-06, "loss": 0.3146, "step": 7310 }, { "epoch": 0.43295676347075174, "grad_norm": 4.5826921463012695, "learning_rate": 5.796801570133331e-06, "loss": 0.2671, "step": 7320 }, { "epoch": 0.4335482344590998, "grad_norm": 3.325389862060547, "learning_rate": 5.796107976448852e-06, "loss": 0.3237, "step": 7330 }, { "epoch": 0.4341397054474478, "grad_norm": 3.5083162784576416, "learning_rate": 5.7954132426691696e-06, "loss": 0.3827, "step": 7340 }, { "epoch": 0.43473117643579584, "grad_norm": 4.219829082489014, "learning_rate": 5.794717369077559e-06, "loss": 0.2884, "step": 7350 }, { "epoch": 0.43532264742414384, "grad_norm": 6.353860378265381, "learning_rate": 5.794020355957758e-06, "loss": 0.3197, "step": 7360 }, { "epoch": 0.4359141184124919, "grad_norm": 2.1638848781585693, "learning_rate": 5.7933222035939686e-06, "loss": 0.3632, "step": 7370 }, { "epoch": 0.4365055894008399, "grad_norm": 3.7190568447113037, "learning_rate": 5.792622912270859e-06, "loss": 0.375, "step": 7380 }, { "epoch": 0.4370970603891879, "grad_norm": 3.5217368602752686, "learning_rate": 5.79192248227356e-06, "loss": 0.2822, "step": 7390 }, { "epoch": 0.43768853137753594, "grad_norm": 3.692201614379883, "learning_rate": 5.791220913887667e-06, "loss": 0.369, "step": 7400 }, { "epoch": 0.43828000236588394, "grad_norm": 1.7712736129760742, "learning_rate": 5.790518207399243e-06, "loss": 0.3603, "step": 7410 }, { "epoch": 0.438871473354232, "grad_norm": 3.7341463565826416, "learning_rate": 5.789814363094807e-06, "loss": 0.3487, "step": 7420 }, { "epoch": 0.43946294434258, "grad_norm": 3.114025115966797, "learning_rate": 5.789109381261351e-06, "loss": 0.3257, "step": 7430 }, { "epoch": 0.44005441533092804, "grad_norm": 3.027696132659912, "learning_rate": 5.788403262186326e-06, "loss": 0.3063, "step": 7440 }, { "epoch": 0.44064588631927604, "grad_norm": 5.573663234710693, "learning_rate": 5.787696006157647e-06, "loss": 0.3833, "step": 7450 }, { "epoch": 0.44123735730762403, "grad_norm": 2.9096481800079346, "learning_rate": 5.7869876134636935e-06, "loss": 0.3556, "step": 7460 }, { "epoch": 0.4418288282959721, "grad_norm": 4.283048629760742, "learning_rate": 5.7862780843933076e-06, "loss": 0.3834, "step": 7470 }, { "epoch": 0.4424202992843201, "grad_norm": 2.9317421913146973, "learning_rate": 5.785567419235796e-06, "loss": 0.2914, "step": 7480 }, { "epoch": 0.44301177027266814, "grad_norm": 3.593651294708252, "learning_rate": 5.784855618280927e-06, "loss": 0.3455, "step": 7490 }, { "epoch": 0.44360324126101613, "grad_norm": 4.599911689758301, "learning_rate": 5.784142681818935e-06, "loss": 0.3277, "step": 7500 }, { "epoch": 0.4441947122493642, "grad_norm": 2.9873995780944824, "learning_rate": 5.7834286101405115e-06, "loss": 0.274, "step": 7510 }, { "epoch": 0.4447861832377122, "grad_norm": 4.855149269104004, "learning_rate": 5.782713403536818e-06, "loss": 0.2863, "step": 7520 }, { "epoch": 0.44537765422606024, "grad_norm": 2.67348313331604, "learning_rate": 5.781997062299475e-06, "loss": 0.3501, "step": 7530 }, { "epoch": 0.44596912521440824, "grad_norm": 3.2093939781188965, "learning_rate": 5.7812795867205645e-06, "loss": 0.3867, "step": 7540 }, { "epoch": 0.44656059620275623, "grad_norm": 2.9433445930480957, "learning_rate": 5.780560977092634e-06, "loss": 0.2858, "step": 7550 }, { "epoch": 0.4471520671911043, "grad_norm": 3.306851625442505, "learning_rate": 5.7798412337086905e-06, "loss": 0.3473, "step": 7560 }, { "epoch": 0.4477435381794523, "grad_norm": 2.996220111846924, "learning_rate": 5.779120356862207e-06, "loss": 0.3436, "step": 7570 }, { "epoch": 0.44833500916780034, "grad_norm": 4.114177703857422, "learning_rate": 5.778398346847114e-06, "loss": 0.3765, "step": 7580 }, { "epoch": 0.44892648015614833, "grad_norm": 3.146385908126831, "learning_rate": 5.7776752039578085e-06, "loss": 0.3401, "step": 7590 }, { "epoch": 0.4495179511444964, "grad_norm": 3.8336877822875977, "learning_rate": 5.776950928489146e-06, "loss": 0.2829, "step": 7600 }, { "epoch": 0.4501094221328444, "grad_norm": 4.353850364685059, "learning_rate": 5.776225520736446e-06, "loss": 0.3259, "step": 7610 }, { "epoch": 0.4507008931211924, "grad_norm": 2.8415286540985107, "learning_rate": 5.775498980995489e-06, "loss": 0.3535, "step": 7620 }, { "epoch": 0.45129236410954043, "grad_norm": 4.668543338775635, "learning_rate": 5.7747713095625165e-06, "loss": 0.4083, "step": 7630 }, { "epoch": 0.45188383509788843, "grad_norm": 3.5147628784179688, "learning_rate": 5.774042506734232e-06, "loss": 0.3528, "step": 7640 }, { "epoch": 0.4524753060862365, "grad_norm": 3.7026398181915283, "learning_rate": 5.773312572807801e-06, "loss": 0.3116, "step": 7650 }, { "epoch": 0.4530667770745845, "grad_norm": 3.1766555309295654, "learning_rate": 5.772581508080848e-06, "loss": 0.2662, "step": 7660 }, { "epoch": 0.45365824806293253, "grad_norm": 4.383542537689209, "learning_rate": 5.77184931285146e-06, "loss": 0.3462, "step": 7670 }, { "epoch": 0.45424971905128053, "grad_norm": 2.6459243297576904, "learning_rate": 5.771115987418186e-06, "loss": 0.4235, "step": 7680 }, { "epoch": 0.45484119003962853, "grad_norm": 4.974571228027344, "learning_rate": 5.770381532080035e-06, "loss": 0.3411, "step": 7690 }, { "epoch": 0.4554326610279766, "grad_norm": 4.866667747497559, "learning_rate": 5.7696459471364745e-06, "loss": 0.321, "step": 7700 }, { "epoch": 0.4560241320163246, "grad_norm": 3.011040210723877, "learning_rate": 5.768909232887437e-06, "loss": 0.3144, "step": 7710 }, { "epoch": 0.45661560300467263, "grad_norm": 3.8132565021514893, "learning_rate": 5.768171389633311e-06, "loss": 0.3268, "step": 7720 }, { "epoch": 0.45720707399302063, "grad_norm": 3.9491372108459473, "learning_rate": 5.767432417674949e-06, "loss": 0.3158, "step": 7730 }, { "epoch": 0.4577985449813687, "grad_norm": 3.673891067504883, "learning_rate": 5.766692317313661e-06, "loss": 0.2876, "step": 7740 }, { "epoch": 0.4583900159697167, "grad_norm": 4.644852161407471, "learning_rate": 5.765951088851219e-06, "loss": 0.3236, "step": 7750 }, { "epoch": 0.45898148695806473, "grad_norm": 5.046876907348633, "learning_rate": 5.765208732589854e-06, "loss": 0.3365, "step": 7760 }, { "epoch": 0.45957295794641273, "grad_norm": 3.079955577850342, "learning_rate": 5.764465248832255e-06, "loss": 0.3312, "step": 7770 }, { "epoch": 0.4601644289347607, "grad_norm": 2.117067337036133, "learning_rate": 5.7637206378815756e-06, "loss": 0.3166, "step": 7780 }, { "epoch": 0.4607558999231088, "grad_norm": 3.0251824855804443, "learning_rate": 5.762974900041426e-06, "loss": 0.346, "step": 7790 }, { "epoch": 0.4613473709114568, "grad_norm": 4.223757743835449, "learning_rate": 5.762228035615875e-06, "loss": 0.3096, "step": 7800 }, { "epoch": 0.46193884189980483, "grad_norm": 4.2156171798706055, "learning_rate": 5.7614800449094515e-06, "loss": 0.3033, "step": 7810 }, { "epoch": 0.46253031288815283, "grad_norm": 3.1276566982269287, "learning_rate": 5.760730928227145e-06, "loss": 0.3438, "step": 7820 }, { "epoch": 0.4631217838765009, "grad_norm": 5.443521499633789, "learning_rate": 5.7599806858744015e-06, "loss": 0.2587, "step": 7830 }, { "epoch": 0.4637132548648489, "grad_norm": 4.462831497192383, "learning_rate": 5.7592293181571285e-06, "loss": 0.3475, "step": 7840 }, { "epoch": 0.4643047258531969, "grad_norm": 6.173925876617432, "learning_rate": 5.758476825381693e-06, "loss": 0.3451, "step": 7850 }, { "epoch": 0.46489619684154493, "grad_norm": 5.888710975646973, "learning_rate": 5.757723207854916e-06, "loss": 0.3424, "step": 7860 }, { "epoch": 0.4654876678298929, "grad_norm": 4.642391204833984, "learning_rate": 5.756968465884083e-06, "loss": 0.2854, "step": 7870 }, { "epoch": 0.466079138818241, "grad_norm": 2.299492597579956, "learning_rate": 5.756212599776933e-06, "loss": 0.3634, "step": 7880 }, { "epoch": 0.466670609806589, "grad_norm": 4.848480701446533, "learning_rate": 5.7554556098416674e-06, "loss": 0.287, "step": 7890 }, { "epoch": 0.46726208079493703, "grad_norm": 4.397586345672607, "learning_rate": 5.754697496386943e-06, "loss": 0.3265, "step": 7900 }, { "epoch": 0.467853551783285, "grad_norm": 1.9073652029037476, "learning_rate": 5.753938259721876e-06, "loss": 0.2967, "step": 7910 }, { "epoch": 0.468445022771633, "grad_norm": 3.066622495651245, "learning_rate": 5.75317790015604e-06, "loss": 0.3455, "step": 7920 }, { "epoch": 0.4690364937599811, "grad_norm": 2.5614609718322754, "learning_rate": 5.752416417999467e-06, "loss": 0.3096, "step": 7930 }, { "epoch": 0.4696279647483291, "grad_norm": 3.6001811027526855, "learning_rate": 5.751653813562647e-06, "loss": 0.3524, "step": 7940 }, { "epoch": 0.4702194357366771, "grad_norm": 4.280065536499023, "learning_rate": 5.750890087156526e-06, "loss": 0.3143, "step": 7950 }, { "epoch": 0.4708109067250251, "grad_norm": 4.157315731048584, "learning_rate": 5.7501252390925095e-06, "loss": 0.3629, "step": 7960 }, { "epoch": 0.4714023777133732, "grad_norm": 6.01833438873291, "learning_rate": 5.7493592696824594e-06, "loss": 0.3422, "step": 7970 }, { "epoch": 0.4719938487017212, "grad_norm": 3.745678424835205, "learning_rate": 5.748592179238694e-06, "loss": 0.301, "step": 7980 }, { "epoch": 0.4725853196900692, "grad_norm": 2.7259521484375, "learning_rate": 5.747823968073991e-06, "loss": 0.3365, "step": 7990 }, { "epoch": 0.4731767906784172, "grad_norm": 2.5477945804595947, "learning_rate": 5.747054636501581e-06, "loss": 0.3564, "step": 8000 }, { "epoch": 0.4737682616667652, "grad_norm": 3.284111261367798, "learning_rate": 5.746284184835158e-06, "loss": 0.3598, "step": 8010 }, { "epoch": 0.4743597326551133, "grad_norm": 2.538360834121704, "learning_rate": 5.7455126133888646e-06, "loss": 0.3188, "step": 8020 }, { "epoch": 0.4749512036434613, "grad_norm": 2.7571372985839844, "learning_rate": 5.744739922477307e-06, "loss": 0.3215, "step": 8030 }, { "epoch": 0.4755426746318093, "grad_norm": 6.068333625793457, "learning_rate": 5.7439661124155445e-06, "loss": 0.2784, "step": 8040 }, { "epoch": 0.4761341456201573, "grad_norm": 6.192043781280518, "learning_rate": 5.743191183519092e-06, "loss": 0.3, "step": 8050 }, { "epoch": 0.4767256166085054, "grad_norm": 2.831573247909546, "learning_rate": 5.742415136103923e-06, "loss": 0.3754, "step": 8060 }, { "epoch": 0.4773170875968534, "grad_norm": 3.875117063522339, "learning_rate": 5.741637970486465e-06, "loss": 0.3491, "step": 8070 }, { "epoch": 0.47790855858520137, "grad_norm": 2.3252830505371094, "learning_rate": 5.740859686983603e-06, "loss": 0.3491, "step": 8080 }, { "epoch": 0.4785000295735494, "grad_norm": 3.244638204574585, "learning_rate": 5.740080285912678e-06, "loss": 0.3551, "step": 8090 }, { "epoch": 0.4790915005618974, "grad_norm": 4.240131378173828, "learning_rate": 5.739299767591482e-06, "loss": 0.2993, "step": 8100 }, { "epoch": 0.4796829715502455, "grad_norm": 4.574280261993408, "learning_rate": 5.738518132338271e-06, "loss": 0.3135, "step": 8110 }, { "epoch": 0.48027444253859347, "grad_norm": 2.845818519592285, "learning_rate": 5.73773538047175e-06, "loss": 0.3136, "step": 8120 }, { "epoch": 0.4808659135269415, "grad_norm": 2.630843162536621, "learning_rate": 5.736951512311081e-06, "loss": 0.3188, "step": 8130 }, { "epoch": 0.4814573845152895, "grad_norm": 3.6751608848571777, "learning_rate": 5.736166528175881e-06, "loss": 0.3026, "step": 8140 }, { "epoch": 0.4820488555036375, "grad_norm": 3.434178113937378, "learning_rate": 5.735380428386222e-06, "loss": 0.371, "step": 8150 }, { "epoch": 0.48264032649198557, "grad_norm": 2.652557134628296, "learning_rate": 5.7345932132626335e-06, "loss": 0.2903, "step": 8160 }, { "epoch": 0.48323179748033357, "grad_norm": 2.633223295211792, "learning_rate": 5.733804883126094e-06, "loss": 0.3073, "step": 8170 }, { "epoch": 0.4838232684686816, "grad_norm": 2.5601820945739746, "learning_rate": 5.7330154382980416e-06, "loss": 0.293, "step": 8180 }, { "epoch": 0.4844147394570296, "grad_norm": 3.0220184326171875, "learning_rate": 5.732224879100368e-06, "loss": 0.3224, "step": 8190 }, { "epoch": 0.4850062104453777, "grad_norm": 4.992455959320068, "learning_rate": 5.7314332058554185e-06, "loss": 0.3031, "step": 8200 }, { "epoch": 0.48559768143372567, "grad_norm": 4.856472492218018, "learning_rate": 5.730640418885992e-06, "loss": 0.3868, "step": 8210 }, { "epoch": 0.4861891524220737, "grad_norm": 1.880338430404663, "learning_rate": 5.729846518515343e-06, "loss": 0.3234, "step": 8220 }, { "epoch": 0.4867806234104217, "grad_norm": 2.6706185340881348, "learning_rate": 5.7290515050671795e-06, "loss": 0.3769, "step": 8230 }, { "epoch": 0.4873720943987697, "grad_norm": 4.6470627784729, "learning_rate": 5.728255378865663e-06, "loss": 0.2854, "step": 8240 }, { "epoch": 0.48796356538711777, "grad_norm": 2.969381093978882, "learning_rate": 5.727458140235408e-06, "loss": 0.2909, "step": 8250 }, { "epoch": 0.48855503637546577, "grad_norm": 4.436374664306641, "learning_rate": 5.7266597895014845e-06, "loss": 0.3973, "step": 8260 }, { "epoch": 0.4891465073638138, "grad_norm": 2.7214248180389404, "learning_rate": 5.725860326989414e-06, "loss": 0.3993, "step": 8270 }, { "epoch": 0.4897379783521618, "grad_norm": 4.143650531768799, "learning_rate": 5.725059753025174e-06, "loss": 0.3358, "step": 8280 }, { "epoch": 0.49032944934050987, "grad_norm": 2.619661331176758, "learning_rate": 5.724258067935191e-06, "loss": 0.3411, "step": 8290 }, { "epoch": 0.49092092032885787, "grad_norm": 2.7867414951324463, "learning_rate": 5.7234552720463495e-06, "loss": 0.2894, "step": 8300 }, { "epoch": 0.49151239131720587, "grad_norm": 2.960881233215332, "learning_rate": 5.722651365685982e-06, "loss": 0.2933, "step": 8310 }, { "epoch": 0.4921038623055539, "grad_norm": 2.515216827392578, "learning_rate": 5.721846349181879e-06, "loss": 0.3082, "step": 8320 }, { "epoch": 0.4926953332939019, "grad_norm": 5.965352535247803, "learning_rate": 5.721040222862277e-06, "loss": 0.3112, "step": 8330 }, { "epoch": 0.49328680428224997, "grad_norm": 2.4769303798675537, "learning_rate": 5.720232987055873e-06, "loss": 0.2461, "step": 8340 }, { "epoch": 0.49387827527059797, "grad_norm": 4.6433634757995605, "learning_rate": 5.719424642091809e-06, "loss": 0.3301, "step": 8350 }, { "epoch": 0.494469746258946, "grad_norm": 2.6571619510650635, "learning_rate": 5.718615188299685e-06, "loss": 0.3371, "step": 8360 }, { "epoch": 0.495061217247294, "grad_norm": 3.7736318111419678, "learning_rate": 5.71780462600955e-06, "loss": 0.3255, "step": 8370 }, { "epoch": 0.49565268823564207, "grad_norm": 2.5228450298309326, "learning_rate": 5.7169929555519026e-06, "loss": 0.3455, "step": 8380 }, { "epoch": 0.49624415922399007, "grad_norm": 3.0344960689544678, "learning_rate": 5.7161801772577004e-06, "loss": 0.3594, "step": 8390 }, { "epoch": 0.49683563021233806, "grad_norm": 6.785481929779053, "learning_rate": 5.715366291458347e-06, "loss": 0.2763, "step": 8400 }, { "epoch": 0.4974271012006861, "grad_norm": 4.3936872482299805, "learning_rate": 5.7145512984856985e-06, "loss": 0.3307, "step": 8410 }, { "epoch": 0.4980185721890341, "grad_norm": 5.039597511291504, "learning_rate": 5.713735198672063e-06, "loss": 0.316, "step": 8420 }, { "epoch": 0.49861004317738217, "grad_norm": 3.0283915996551514, "learning_rate": 5.7129179923502016e-06, "loss": 0.2861, "step": 8430 }, { "epoch": 0.49920151416573016, "grad_norm": 4.075348854064941, "learning_rate": 5.712099679853323e-06, "loss": 0.3226, "step": 8440 }, { "epoch": 0.4997929851540782, "grad_norm": 4.022360324859619, "learning_rate": 5.711280261515091e-06, "loss": 0.3218, "step": 8450 }, { "epoch": 0.5003844561424262, "grad_norm": 5.088761329650879, "learning_rate": 5.710459737669617e-06, "loss": 0.3046, "step": 8460 }, { "epoch": 0.5009759271307742, "grad_norm": 3.508687734603882, "learning_rate": 5.709638108651464e-06, "loss": 0.3092, "step": 8470 }, { "epoch": 0.5015673981191222, "grad_norm": 2.441190242767334, "learning_rate": 5.708815374795647e-06, "loss": 0.4157, "step": 8480 }, { "epoch": 0.5021588691074703, "grad_norm": 3.919224262237549, "learning_rate": 5.70799153643763e-06, "loss": 0.3183, "step": 8490 }, { "epoch": 0.5027503400958183, "grad_norm": 3.801893949508667, "learning_rate": 5.707166593913327e-06, "loss": 0.3163, "step": 8500 }, { "epoch": 0.5033418110841663, "grad_norm": 5.043663501739502, "learning_rate": 5.706340547559104e-06, "loss": 0.2835, "step": 8510 }, { "epoch": 0.5039332820725143, "grad_norm": 4.417965888977051, "learning_rate": 5.705513397711776e-06, "loss": 0.2996, "step": 8520 }, { "epoch": 0.5045247530608624, "grad_norm": 3.1816375255584717, "learning_rate": 5.704685144708608e-06, "loss": 0.3877, "step": 8530 }, { "epoch": 0.5051162240492104, "grad_norm": 3.3934953212738037, "learning_rate": 5.703855788887316e-06, "loss": 0.4351, "step": 8540 }, { "epoch": 0.5057076950375584, "grad_norm": 2.8155338764190674, "learning_rate": 5.7030253305860626e-06, "loss": 0.3474, "step": 8550 }, { "epoch": 0.5062991660259064, "grad_norm": 3.624669075012207, "learning_rate": 5.702193770143462e-06, "loss": 0.3462, "step": 8560 }, { "epoch": 0.5068906370142544, "grad_norm": 3.2609169483184814, "learning_rate": 5.701361107898579e-06, "loss": 0.2925, "step": 8570 }, { "epoch": 0.5074821080026025, "grad_norm": 1.9543650150299072, "learning_rate": 5.700527344190926e-06, "loss": 0.2973, "step": 8580 }, { "epoch": 0.5080735789909505, "grad_norm": 2.260852336883545, "learning_rate": 5.699692479360464e-06, "loss": 0.3978, "step": 8590 }, { "epoch": 0.5086650499792985, "grad_norm": 3.807457447052002, "learning_rate": 5.6988565137476056e-06, "loss": 0.3178, "step": 8600 }, { "epoch": 0.5092565209676465, "grad_norm": 4.427671432495117, "learning_rate": 5.698019447693209e-06, "loss": 0.3797, "step": 8610 }, { "epoch": 0.5098479919559945, "grad_norm": 5.111141681671143, "learning_rate": 5.697181281538582e-06, "loss": 0.3526, "step": 8620 }, { "epoch": 0.5104394629443426, "grad_norm": 2.6560378074645996, "learning_rate": 5.696342015625483e-06, "loss": 0.3102, "step": 8630 }, { "epoch": 0.5110309339326906, "grad_norm": 3.423405647277832, "learning_rate": 5.695501650296116e-06, "loss": 0.3264, "step": 8640 }, { "epoch": 0.5116224049210386, "grad_norm": 5.373915195465088, "learning_rate": 5.694660185893135e-06, "loss": 0.396, "step": 8650 }, { "epoch": 0.5122138759093866, "grad_norm": 2.2734556198120117, "learning_rate": 5.693817622759643e-06, "loss": 0.3164, "step": 8660 }, { "epoch": 0.5128053468977347, "grad_norm": 2.470900058746338, "learning_rate": 5.692973961239188e-06, "loss": 0.2946, "step": 8670 }, { "epoch": 0.5133968178860827, "grad_norm": 4.406801223754883, "learning_rate": 5.6921292016757684e-06, "loss": 0.3677, "step": 8680 }, { "epoch": 0.5139882888744307, "grad_norm": 1.5799778699874878, "learning_rate": 5.6912833444138296e-06, "loss": 0.333, "step": 8690 }, { "epoch": 0.5145797598627787, "grad_norm": 4.611819267272949, "learning_rate": 5.6904363897982645e-06, "loss": 0.2834, "step": 8700 }, { "epoch": 0.5151712308511267, "grad_norm": 3.068695068359375, "learning_rate": 5.689588338174413e-06, "loss": 0.2683, "step": 8710 }, { "epoch": 0.5157627018394748, "grad_norm": 3.901982069015503, "learning_rate": 5.688739189888063e-06, "loss": 0.3112, "step": 8720 }, { "epoch": 0.5163541728278228, "grad_norm": 4.245181083679199, "learning_rate": 5.68788894528545e-06, "loss": 0.4143, "step": 8730 }, { "epoch": 0.5169456438161708, "grad_norm": 1.4601598978042603, "learning_rate": 5.687037604713254e-06, "loss": 0.3198, "step": 8740 }, { "epoch": 0.5175371148045188, "grad_norm": 3.584897756576538, "learning_rate": 5.686185168518606e-06, "loss": 0.3441, "step": 8750 }, { "epoch": 0.5181285857928669, "grad_norm": 2.860246419906616, "learning_rate": 5.68533163704908e-06, "loss": 0.3365, "step": 8760 }, { "epoch": 0.5187200567812149, "grad_norm": 2.68734073638916, "learning_rate": 5.684477010652698e-06, "loss": 0.3313, "step": 8770 }, { "epoch": 0.5193115277695629, "grad_norm": 4.5146098136901855, "learning_rate": 5.683621289677929e-06, "loss": 0.3511, "step": 8780 }, { "epoch": 0.5199029987579109, "grad_norm": 2.0588924884796143, "learning_rate": 5.682764474473687e-06, "loss": 0.3036, "step": 8790 }, { "epoch": 0.5204944697462589, "grad_norm": 3.8813276290893555, "learning_rate": 5.681906565389334e-06, "loss": 0.3179, "step": 8800 }, { "epoch": 0.521085940734607, "grad_norm": 2.2539222240448, "learning_rate": 5.681047562774677e-06, "loss": 0.2813, "step": 8810 }, { "epoch": 0.521677411722955, "grad_norm": 8.845205307006836, "learning_rate": 5.680187466979967e-06, "loss": 0.2819, "step": 8820 }, { "epoch": 0.522268882711303, "grad_norm": 4.425050258636475, "learning_rate": 5.679326278355904e-06, "loss": 0.3523, "step": 8830 }, { "epoch": 0.522860353699651, "grad_norm": 2.6416115760803223, "learning_rate": 5.678463997253633e-06, "loss": 0.3017, "step": 8840 }, { "epoch": 0.523451824687999, "grad_norm": 7.102107524871826, "learning_rate": 5.677600624024742e-06, "loss": 0.3049, "step": 8850 }, { "epoch": 0.5240432956763471, "grad_norm": 2.359229803085327, "learning_rate": 5.676736159021267e-06, "loss": 0.2886, "step": 8860 }, { "epoch": 0.5246347666646951, "grad_norm": 3.473363161087036, "learning_rate": 5.675870602595687e-06, "loss": 0.3576, "step": 8870 }, { "epoch": 0.5252262376530431, "grad_norm": 3.157467842102051, "learning_rate": 5.675003955100928e-06, "loss": 0.3624, "step": 8880 }, { "epoch": 0.5258177086413911, "grad_norm": 3.6484527587890625, "learning_rate": 5.674136216890359e-06, "loss": 0.3717, "step": 8890 }, { "epoch": 0.5264091796297392, "grad_norm": 3.4994003772735596, "learning_rate": 5.673267388317797e-06, "loss": 0.2915, "step": 8900 }, { "epoch": 0.5270006506180872, "grad_norm": 2.372403383255005, "learning_rate": 5.672397469737499e-06, "loss": 0.3389, "step": 8910 }, { "epoch": 0.5275921216064352, "grad_norm": 4.096915245056152, "learning_rate": 5.67152646150417e-06, "loss": 0.3673, "step": 8920 }, { "epoch": 0.5281835925947832, "grad_norm": 3.0323922634124756, "learning_rate": 5.6706543639729574e-06, "loss": 0.3003, "step": 8930 }, { "epoch": 0.5287750635831312, "grad_norm": 3.375513792037964, "learning_rate": 5.669781177499453e-06, "loss": 0.2973, "step": 8940 }, { "epoch": 0.5293665345714793, "grad_norm": 5.360010147094727, "learning_rate": 5.668906902439695e-06, "loss": 0.3985, "step": 8950 }, { "epoch": 0.5299580055598273, "grad_norm": 3.0715577602386475, "learning_rate": 5.668031539150161e-06, "loss": 0.3264, "step": 8960 }, { "epoch": 0.5305494765481753, "grad_norm": 2.881575584411621, "learning_rate": 5.667155087987776e-06, "loss": 0.3894, "step": 8970 }, { "epoch": 0.5311409475365233, "grad_norm": 2.4999589920043945, "learning_rate": 5.666277549309908e-06, "loss": 0.3363, "step": 8980 }, { "epoch": 0.5317324185248714, "grad_norm": 3.3840770721435547, "learning_rate": 5.665398923474366e-06, "loss": 0.3587, "step": 8990 }, { "epoch": 0.5323238895132194, "grad_norm": 3.7031211853027344, "learning_rate": 5.664519210839406e-06, "loss": 0.3133, "step": 9000 }, { "epoch": 0.5329153605015674, "grad_norm": 3.049877643585205, "learning_rate": 5.663638411763725e-06, "loss": 0.3622, "step": 9010 }, { "epoch": 0.5335068314899154, "grad_norm": 3.178436756134033, "learning_rate": 5.6627565266064615e-06, "loss": 0.343, "step": 9020 }, { "epoch": 0.5340983024782634, "grad_norm": 3.2570478916168213, "learning_rate": 5.6618735557272e-06, "loss": 0.3869, "step": 9030 }, { "epoch": 0.5346897734666115, "grad_norm": 2.5773653984069824, "learning_rate": 5.6609894994859655e-06, "loss": 0.323, "step": 9040 }, { "epoch": 0.5352812444549595, "grad_norm": 5.289306163787842, "learning_rate": 5.660104358243227e-06, "loss": 0.3117, "step": 9050 }, { "epoch": 0.5358727154433075, "grad_norm": 3.7207729816436768, "learning_rate": 5.659218132359895e-06, "loss": 0.3144, "step": 9060 }, { "epoch": 0.5364641864316555, "grad_norm": 3.0170862674713135, "learning_rate": 5.658330822197322e-06, "loss": 0.3265, "step": 9070 }, { "epoch": 0.5370556574200035, "grad_norm": 2.76920485496521, "learning_rate": 5.657442428117304e-06, "loss": 0.318, "step": 9080 }, { "epoch": 0.5376471284083516, "grad_norm": 3.2434937953948975, "learning_rate": 5.656552950482078e-06, "loss": 0.3167, "step": 9090 }, { "epoch": 0.5382385993966996, "grad_norm": 3.947178363800049, "learning_rate": 5.655662389654321e-06, "loss": 0.3372, "step": 9100 }, { "epoch": 0.5388300703850476, "grad_norm": 2.4691669940948486, "learning_rate": 5.654770745997156e-06, "loss": 0.3322, "step": 9110 }, { "epoch": 0.5394215413733956, "grad_norm": 2.9352169036865234, "learning_rate": 5.6538780198741424e-06, "loss": 0.2959, "step": 9120 }, { "epoch": 0.5400130123617437, "grad_norm": 3.5789918899536133, "learning_rate": 5.652984211649285e-06, "loss": 0.3799, "step": 9130 }, { "epoch": 0.5406044833500917, "grad_norm": 1.9224169254302979, "learning_rate": 5.6520893216870295e-06, "loss": 0.3307, "step": 9140 }, { "epoch": 0.5411959543384397, "grad_norm": 6.82618522644043, "learning_rate": 5.65119335035226e-06, "loss": 0.3339, "step": 9150 }, { "epoch": 0.5417874253267877, "grad_norm": 3.322726249694824, "learning_rate": 5.650296298010303e-06, "loss": 0.3327, "step": 9160 }, { "epoch": 0.5423788963151357, "grad_norm": 2.866572856903076, "learning_rate": 5.649398165026928e-06, "loss": 0.364, "step": 9170 }, { "epoch": 0.5429703673034838, "grad_norm": 2.9290504455566406, "learning_rate": 5.64849895176834e-06, "loss": 0.3255, "step": 9180 }, { "epoch": 0.5435618382918318, "grad_norm": 4.672257900238037, "learning_rate": 5.64759865860119e-06, "loss": 0.3633, "step": 9190 }, { "epoch": 0.5441533092801798, "grad_norm": 4.921064853668213, "learning_rate": 5.646697285892564e-06, "loss": 0.36, "step": 9200 }, { "epoch": 0.5447447802685278, "grad_norm": 2.6018097400665283, "learning_rate": 5.645794834009994e-06, "loss": 0.3142, "step": 9210 }, { "epoch": 0.5453362512568759, "grad_norm": 3.0425655841827393, "learning_rate": 5.644891303321447e-06, "loss": 0.4088, "step": 9220 }, { "epoch": 0.5459277222452239, "grad_norm": 2.4031784534454346, "learning_rate": 5.643986694195333e-06, "loss": 0.3975, "step": 9230 }, { "epoch": 0.5465191932335719, "grad_norm": 2.4593586921691895, "learning_rate": 5.6430810070005e-06, "loss": 0.3187, "step": 9240 }, { "epoch": 0.5471106642219199, "grad_norm": 4.086849212646484, "learning_rate": 5.642174242106236e-06, "loss": 0.3261, "step": 9250 }, { "epoch": 0.5477021352102679, "grad_norm": 2.7902615070343018, "learning_rate": 5.64126639988227e-06, "loss": 0.2984, "step": 9260 }, { "epoch": 0.548293606198616, "grad_norm": 2.3847525119781494, "learning_rate": 5.640357480698768e-06, "loss": 0.3652, "step": 9270 }, { "epoch": 0.548885077186964, "grad_norm": 4.116032123565674, "learning_rate": 5.639447484926334e-06, "loss": 0.3154, "step": 9280 }, { "epoch": 0.549476548175312, "grad_norm": 3.4112722873687744, "learning_rate": 5.638536412936016e-06, "loss": 0.3857, "step": 9290 }, { "epoch": 0.55006801916366, "grad_norm": 5.422371864318848, "learning_rate": 5.637624265099296e-06, "loss": 0.2885, "step": 9300 }, { "epoch": 0.550659490152008, "grad_norm": 3.392624616622925, "learning_rate": 5.636711041788098e-06, "loss": 0.344, "step": 9310 }, { "epoch": 0.5512509611403561, "grad_norm": 3.6207938194274902, "learning_rate": 5.635796743374782e-06, "loss": 0.3349, "step": 9320 }, { "epoch": 0.5518424321287041, "grad_norm": 2.8290629386901855, "learning_rate": 5.634881370232146e-06, "loss": 0.3769, "step": 9330 }, { "epoch": 0.5524339031170521, "grad_norm": 2.58441162109375, "learning_rate": 5.633964922733431e-06, "loss": 0.3102, "step": 9340 }, { "epoch": 0.5530253741054001, "grad_norm": 5.559804916381836, "learning_rate": 5.6330474012523095e-06, "loss": 0.2738, "step": 9350 }, { "epoch": 0.5536168450937482, "grad_norm": 3.903137683868408, "learning_rate": 5.632128806162896e-06, "loss": 0.3107, "step": 9360 }, { "epoch": 0.5542083160820962, "grad_norm": 3.253185749053955, "learning_rate": 5.631209137839742e-06, "loss": 0.3561, "step": 9370 }, { "epoch": 0.5547997870704442, "grad_norm": 2.7363429069519043, "learning_rate": 5.630288396657835e-06, "loss": 0.305, "step": 9380 }, { "epoch": 0.5553912580587922, "grad_norm": 1.845442771911621, "learning_rate": 5.629366582992605e-06, "loss": 0.3599, "step": 9390 }, { "epoch": 0.5559827290471402, "grad_norm": 3.2555348873138428, "learning_rate": 5.628443697219913e-06, "loss": 0.3399, "step": 9400 }, { "epoch": 0.5565742000354883, "grad_norm": 1.9976693391799927, "learning_rate": 5.627519739716059e-06, "loss": 0.374, "step": 9410 }, { "epoch": 0.5571656710238363, "grad_norm": 2.4063730239868164, "learning_rate": 5.626594710857783e-06, "loss": 0.3174, "step": 9420 }, { "epoch": 0.5577571420121843, "grad_norm": 2.985078811645508, "learning_rate": 5.6256686110222574e-06, "loss": 0.3534, "step": 9430 }, { "epoch": 0.5583486130005323, "grad_norm": 4.213679790496826, "learning_rate": 5.624741440587095e-06, "loss": 0.3261, "step": 9440 }, { "epoch": 0.5589400839888804, "grad_norm": 4.61176872253418, "learning_rate": 5.623813199930342e-06, "loss": 0.3679, "step": 9450 }, { "epoch": 0.5595315549772284, "grad_norm": 2.424420118331909, "learning_rate": 5.622883889430486e-06, "loss": 0.3757, "step": 9460 }, { "epoch": 0.5601230259655764, "grad_norm": 3.168846368789673, "learning_rate": 5.6219535094664445e-06, "loss": 0.3324, "step": 9470 }, { "epoch": 0.5607144969539244, "grad_norm": 2.3702917098999023, "learning_rate": 5.6210220604175746e-06, "loss": 0.3535, "step": 9480 }, { "epoch": 0.5613059679422724, "grad_norm": 2.86739182472229, "learning_rate": 5.620089542663668e-06, "loss": 0.3328, "step": 9490 }, { "epoch": 0.5618974389306205, "grad_norm": 5.887235641479492, "learning_rate": 5.619155956584955e-06, "loss": 0.3024, "step": 9500 }, { "epoch": 0.5624889099189685, "grad_norm": 4.7069172859191895, "learning_rate": 5.618221302562098e-06, "loss": 0.3767, "step": 9510 }, { "epoch": 0.5630803809073165, "grad_norm": 3.0073657035827637, "learning_rate": 5.617285580976196e-06, "loss": 0.3963, "step": 9520 }, { "epoch": 0.5636718518956645, "grad_norm": 4.093173027038574, "learning_rate": 5.616348792208784e-06, "loss": 0.3091, "step": 9530 }, { "epoch": 0.5642633228840125, "grad_norm": 3.712144374847412, "learning_rate": 5.615410936641832e-06, "loss": 0.3035, "step": 9540 }, { "epoch": 0.5648547938723606, "grad_norm": 3.1291909217834473, "learning_rate": 5.614472014657743e-06, "loss": 0.3607, "step": 9550 }, { "epoch": 0.5654462648607086, "grad_norm": 4.3665852546691895, "learning_rate": 5.613532026639359e-06, "loss": 0.3076, "step": 9560 }, { "epoch": 0.5660377358490566, "grad_norm": 3.1885993480682373, "learning_rate": 5.612590972969951e-06, "loss": 0.3363, "step": 9570 }, { "epoch": 0.5666292068374046, "grad_norm": 2.1787796020507812, "learning_rate": 5.61164885403323e-06, "loss": 0.3141, "step": 9580 }, { "epoch": 0.5672206778257527, "grad_norm": 2.3418548107147217, "learning_rate": 5.61070567021334e-06, "loss": 0.3044, "step": 9590 }, { "epoch": 0.5678121488141007, "grad_norm": 4.426952838897705, "learning_rate": 5.6097614218948545e-06, "loss": 0.3057, "step": 9600 }, { "epoch": 0.5684036198024487, "grad_norm": 3.607754707336426, "learning_rate": 5.608816109462787e-06, "loss": 0.3084, "step": 9610 }, { "epoch": 0.5689950907907967, "grad_norm": 2.414097547531128, "learning_rate": 5.607869733302583e-06, "loss": 0.379, "step": 9620 }, { "epoch": 0.5695865617791447, "grad_norm": 2.3879356384277344, "learning_rate": 5.606922293800121e-06, "loss": 0.314, "step": 9630 }, { "epoch": 0.5701780327674928, "grad_norm": 3.068188190460205, "learning_rate": 5.605973791341712e-06, "loss": 0.3674, "step": 9640 }, { "epoch": 0.5707695037558408, "grad_norm": 4.431291580200195, "learning_rate": 5.605024226314104e-06, "loss": 0.3301, "step": 9650 }, { "epoch": 0.5713609747441888, "grad_norm": 2.534173011779785, "learning_rate": 5.604073599104474e-06, "loss": 0.353, "step": 9660 }, { "epoch": 0.5719524457325368, "grad_norm": 2.751615524291992, "learning_rate": 5.603121910100436e-06, "loss": 0.2512, "step": 9670 }, { "epoch": 0.5725439167208849, "grad_norm": 3.3634519577026367, "learning_rate": 5.6021691596900345e-06, "loss": 0.2981, "step": 9680 }, { "epoch": 0.5731353877092329, "grad_norm": 4.197679042816162, "learning_rate": 5.601215348261747e-06, "loss": 0.2833, "step": 9690 }, { "epoch": 0.5737268586975809, "grad_norm": 5.56403923034668, "learning_rate": 5.6002604762044845e-06, "loss": 0.3152, "step": 9700 }, { "epoch": 0.5743183296859289, "grad_norm": 5.678090572357178, "learning_rate": 5.599304543907588e-06, "loss": 0.3703, "step": 9710 }, { "epoch": 0.5749098006742769, "grad_norm": 2.374371290206909, "learning_rate": 5.598347551760838e-06, "loss": 0.3567, "step": 9720 }, { "epoch": 0.575501271662625, "grad_norm": 2.6103241443634033, "learning_rate": 5.5973895001544366e-06, "loss": 0.3309, "step": 9730 }, { "epoch": 0.576092742650973, "grad_norm": 3.835602045059204, "learning_rate": 5.5964303894790246e-06, "loss": 0.3171, "step": 9740 }, { "epoch": 0.576684213639321, "grad_norm": 5.30594539642334, "learning_rate": 5.595470220125675e-06, "loss": 0.3363, "step": 9750 }, { "epoch": 0.577275684627669, "grad_norm": 2.3856585025787354, "learning_rate": 5.5945089924858886e-06, "loss": 0.339, "step": 9760 }, { "epoch": 0.577867155616017, "grad_norm": 2.9386606216430664, "learning_rate": 5.593546706951601e-06, "loss": 0.3112, "step": 9770 }, { "epoch": 0.5784586266043651, "grad_norm": 2.9721317291259766, "learning_rate": 5.592583363915178e-06, "loss": 0.3649, "step": 9780 }, { "epoch": 0.5790500975927131, "grad_norm": 2.926286458969116, "learning_rate": 5.591618963769417e-06, "loss": 0.2712, "step": 9790 }, { "epoch": 0.5796415685810611, "grad_norm": 5.985281944274902, "learning_rate": 5.590653506907544e-06, "loss": 0.2841, "step": 9800 }, { "epoch": 0.5802330395694091, "grad_norm": 5.744395732879639, "learning_rate": 5.589686993723221e-06, "loss": 0.3578, "step": 9810 }, { "epoch": 0.5808245105577572, "grad_norm": 2.7726478576660156, "learning_rate": 5.588719424610535e-06, "loss": 0.2947, "step": 9820 }, { "epoch": 0.5814159815461052, "grad_norm": 4.838591575622559, "learning_rate": 5.5877507999640065e-06, "loss": 0.3005, "step": 9830 }, { "epoch": 0.5820074525344532, "grad_norm": 2.9504199028015137, "learning_rate": 5.586781120178587e-06, "loss": 0.3289, "step": 9840 }, { "epoch": 0.5825989235228012, "grad_norm": 12.143149375915527, "learning_rate": 5.585810385649657e-06, "loss": 0.3183, "step": 9850 }, { "epoch": 0.5831903945111492, "grad_norm": 3.883408784866333, "learning_rate": 5.584838596773028e-06, "loss": 0.3057, "step": 9860 }, { "epoch": 0.5837818654994973, "grad_norm": 2.844644784927368, "learning_rate": 5.583865753944938e-06, "loss": 0.3145, "step": 9870 }, { "epoch": 0.5843733364878453, "grad_norm": 3.0679590702056885, "learning_rate": 5.5828918575620605e-06, "loss": 0.3007, "step": 9880 }, { "epoch": 0.5849648074761933, "grad_norm": 4.724521636962891, "learning_rate": 5.581916908021493e-06, "loss": 0.3871, "step": 9890 }, { "epoch": 0.5855562784645413, "grad_norm": 2.688218593597412, "learning_rate": 5.580940905720767e-06, "loss": 0.3135, "step": 9900 }, { "epoch": 0.5861477494528894, "grad_norm": 4.261818885803223, "learning_rate": 5.579963851057841e-06, "loss": 0.3694, "step": 9910 }, { "epoch": 0.5867392204412374, "grad_norm": 2.8821072578430176, "learning_rate": 5.578985744431102e-06, "loss": 0.3115, "step": 9920 }, { "epoch": 0.5873306914295854, "grad_norm": 2.6980929374694824, "learning_rate": 5.578006586239366e-06, "loss": 0.307, "step": 9930 }, { "epoch": 0.5879221624179334, "grad_norm": 2.7986397743225098, "learning_rate": 5.577026376881882e-06, "loss": 0.34, "step": 9940 }, { "epoch": 0.5885136334062814, "grad_norm": 3.9016635417938232, "learning_rate": 5.576045116758319e-06, "loss": 0.3144, "step": 9950 }, { "epoch": 0.5891051043946295, "grad_norm": 2.693300724029541, "learning_rate": 5.575062806268783e-06, "loss": 0.2904, "step": 9960 }, { "epoch": 0.5896965753829775, "grad_norm": 2.317556619644165, "learning_rate": 5.574079445813805e-06, "loss": 0.3146, "step": 9970 }, { "epoch": 0.5902880463713255, "grad_norm": 3.1623189449310303, "learning_rate": 5.573095035794342e-06, "loss": 0.3697, "step": 9980 }, { "epoch": 0.5908795173596735, "grad_norm": 2.704840898513794, "learning_rate": 5.5721095766117835e-06, "loss": 0.2967, "step": 9990 }, { "epoch": 0.5914709883480215, "grad_norm": 3.521183490753174, "learning_rate": 5.5711230686679415e-06, "loss": 0.3566, "step": 10000 }, { "epoch": 0.5920624593363696, "grad_norm": 2.2813234329223633, "learning_rate": 5.570135512365059e-06, "loss": 0.2903, "step": 10010 }, { "epoch": 0.5926539303247176, "grad_norm": 2.3359432220458984, "learning_rate": 5.569146908105806e-06, "loss": 0.3093, "step": 10020 }, { "epoch": 0.5932454013130656, "grad_norm": 2.9333574771881104, "learning_rate": 5.568157256293282e-06, "loss": 0.3426, "step": 10030 }, { "epoch": 0.5938368723014136, "grad_norm": 3.1349570751190186, "learning_rate": 5.567166557331007e-06, "loss": 0.294, "step": 10040 }, { "epoch": 0.5944283432897617, "grad_norm": 6.040511608123779, "learning_rate": 5.566174811622935e-06, "loss": 0.3054, "step": 10050 }, { "epoch": 0.5950198142781097, "grad_norm": 2.276334285736084, "learning_rate": 5.565182019573443e-06, "loss": 0.3224, "step": 10060 }, { "epoch": 0.5956112852664577, "grad_norm": 3.380671262741089, "learning_rate": 5.5641881815873354e-06, "loss": 0.3439, "step": 10070 }, { "epoch": 0.5962027562548057, "grad_norm": 3.4548346996307373, "learning_rate": 5.563193298069845e-06, "loss": 0.3166, "step": 10080 }, { "epoch": 0.5967942272431537, "grad_norm": 3.8056654930114746, "learning_rate": 5.562197369426627e-06, "loss": 0.303, "step": 10090 }, { "epoch": 0.5973856982315018, "grad_norm": 3.8026905059814453, "learning_rate": 5.561200396063766e-06, "loss": 0.3273, "step": 10100 }, { "epoch": 0.5979771692198498, "grad_norm": 2.45173716545105, "learning_rate": 5.560202378387772e-06, "loss": 0.3146, "step": 10110 }, { "epoch": 0.5985686402081978, "grad_norm": 2.1250839233398438, "learning_rate": 5.559203316805579e-06, "loss": 0.3328, "step": 10120 }, { "epoch": 0.5991601111965458, "grad_norm": 3.7460782527923584, "learning_rate": 5.55820321172455e-06, "loss": 0.3219, "step": 10130 }, { "epoch": 0.5997515821848939, "grad_norm": 3.9143710136413574, "learning_rate": 5.557202063552468e-06, "loss": 0.3161, "step": 10140 }, { "epoch": 0.6001064647779026, "eval_accuracy": 0.859150946534917, "eval_loss": 0.32434797286987305, "eval_runtime": 197.7983, "eval_safe_aucpr": 0.9135243921440691, "eval_safe_f1": 0.8420424229987128, "eval_safe_fpr": 0.1306986338226058, "eval_safe_precision": 0.837713437268003, "eval_safe_recall": 0.8464163822525598, "eval_samples_per_second": 303.916, "eval_steps_per_second": 37.993, "eval_unsafe_aucpr": 0.9518693260499573, "eval_unsafe_f1": 0.8729155722326454, "eval_unsafe_fpr": 0.1535836177474397, "eval_unsafe_precision": 0.8765599565925122, "eval_unsafe_recall": 0.8693013661773938, "step": 10146 }, { "epoch": 0.6003430531732419, "grad_norm": 5.367057800292969, "learning_rate": 5.55619987269755e-06, "loss": 0.3939, "step": 10150 }, { "epoch": 0.6009345241615899, "grad_norm": 2.3232765197753906, "learning_rate": 5.5551966395684285e-06, "loss": 0.3329, "step": 10160 }, { "epoch": 0.6015259951499379, "grad_norm": 3.1666789054870605, "learning_rate": 5.554192364574166e-06, "loss": 0.3231, "step": 10170 }, { "epoch": 0.6021174661382859, "grad_norm": 4.899357795715332, "learning_rate": 5.553187048124251e-06, "loss": 0.3481, "step": 10180 }, { "epoch": 0.602708937126634, "grad_norm": 2.679821491241455, "learning_rate": 5.552180690628594e-06, "loss": 0.3472, "step": 10190 }, { "epoch": 0.603300408114982, "grad_norm": 3.003375291824341, "learning_rate": 5.551173292497531e-06, "loss": 0.3312, "step": 10200 }, { "epoch": 0.60389187910333, "grad_norm": 3.362159490585327, "learning_rate": 5.550164854141822e-06, "loss": 0.4345, "step": 10210 }, { "epoch": 0.604483350091678, "grad_norm": 4.024406909942627, "learning_rate": 5.549155375972651e-06, "loss": 0.3516, "step": 10220 }, { "epoch": 0.605074821080026, "grad_norm": 4.7620368003845215, "learning_rate": 5.548144858401628e-06, "loss": 0.3787, "step": 10230 }, { "epoch": 0.6056662920683741, "grad_norm": 3.2564823627471924, "learning_rate": 5.547133301840782e-06, "loss": 0.3077, "step": 10240 }, { "epoch": 0.6062577630567221, "grad_norm": 3.5173559188842773, "learning_rate": 5.54612070670257e-06, "loss": 0.3653, "step": 10250 }, { "epoch": 0.6068492340450701, "grad_norm": 2.3945281505584717, "learning_rate": 5.5451070733998735e-06, "loss": 0.3938, "step": 10260 }, { "epoch": 0.6074407050334181, "grad_norm": 3.5916104316711426, "learning_rate": 5.544092402345992e-06, "loss": 0.3597, "step": 10270 }, { "epoch": 0.6080321760217662, "grad_norm": 3.4283447265625, "learning_rate": 5.543076693954652e-06, "loss": 0.3131, "step": 10280 }, { "epoch": 0.6086236470101142, "grad_norm": 3.570405960083008, "learning_rate": 5.542059948640004e-06, "loss": 0.3968, "step": 10290 }, { "epoch": 0.6092151179984622, "grad_norm": 5.12440299987793, "learning_rate": 5.5410421668166175e-06, "loss": 0.381, "step": 10300 }, { "epoch": 0.6098065889868102, "grad_norm": 2.4499878883361816, "learning_rate": 5.540023348899485e-06, "loss": 0.3614, "step": 10310 }, { "epoch": 0.6103980599751582, "grad_norm": 2.861220598220825, "learning_rate": 5.539003495304027e-06, "loss": 0.3816, "step": 10320 }, { "epoch": 0.6109895309635063, "grad_norm": 2.666935682296753, "learning_rate": 5.537982606446079e-06, "loss": 0.3338, "step": 10330 }, { "epoch": 0.6115810019518543, "grad_norm": 3.1984241008758545, "learning_rate": 5.536960682741904e-06, "loss": 0.2965, "step": 10340 }, { "epoch": 0.6121724729402023, "grad_norm": 4.581541538238525, "learning_rate": 5.535937724608184e-06, "loss": 0.3548, "step": 10350 }, { "epoch": 0.6127639439285503, "grad_norm": 4.027710914611816, "learning_rate": 5.534913732462023e-06, "loss": 0.32, "step": 10360 }, { "epoch": 0.6133554149168984, "grad_norm": 2.5709972381591797, "learning_rate": 5.533888706720949e-06, "loss": 0.2979, "step": 10370 }, { "epoch": 0.6139468859052464, "grad_norm": 3.139954090118408, "learning_rate": 5.532862647802907e-06, "loss": 0.3459, "step": 10380 }, { "epoch": 0.6145383568935944, "grad_norm": 2.555962085723877, "learning_rate": 5.531835556126269e-06, "loss": 0.2779, "step": 10390 }, { "epoch": 0.6151298278819424, "grad_norm": 4.06376314163208, "learning_rate": 5.530807432109823e-06, "loss": 0.3492, "step": 10400 }, { "epoch": 0.6157212988702904, "grad_norm": 2.4832115173339844, "learning_rate": 5.5297782761727815e-06, "loss": 0.4165, "step": 10410 }, { "epoch": 0.6163127698586385, "grad_norm": 2.3903310298919678, "learning_rate": 5.528748088734775e-06, "loss": 0.3026, "step": 10420 }, { "epoch": 0.6169042408469865, "grad_norm": 1.62978994846344, "learning_rate": 5.527716870215857e-06, "loss": 0.3308, "step": 10430 }, { "epoch": 0.6174957118353345, "grad_norm": 2.996938705444336, "learning_rate": 5.5266846210364995e-06, "loss": 0.321, "step": 10440 }, { "epoch": 0.6180871828236825, "grad_norm": 3.8461251258850098, "learning_rate": 5.525651341617595e-06, "loss": 0.365, "step": 10450 }, { "epoch": 0.6186786538120305, "grad_norm": 5.652013778686523, "learning_rate": 5.524617032380458e-06, "loss": 0.2797, "step": 10460 }, { "epoch": 0.6192701248003786, "grad_norm": 2.507563352584839, "learning_rate": 5.523581693746821e-06, "loss": 0.3301, "step": 10470 }, { "epoch": 0.6198615957887266, "grad_norm": 3.630125045776367, "learning_rate": 5.522545326138836e-06, "loss": 0.3688, "step": 10480 }, { "epoch": 0.6204530667770746, "grad_norm": 2.502631425857544, "learning_rate": 5.521507929979075e-06, "loss": 0.3348, "step": 10490 }, { "epoch": 0.6210445377654226, "grad_norm": 3.1283771991729736, "learning_rate": 5.520469505690531e-06, "loss": 0.3254, "step": 10500 }, { "epoch": 0.6216360087537707, "grad_norm": 2.328892230987549, "learning_rate": 5.519430053696615e-06, "loss": 0.3495, "step": 10510 }, { "epoch": 0.6222274797421187, "grad_norm": 2.0433945655822754, "learning_rate": 5.518389574421156e-06, "loss": 0.301, "step": 10520 }, { "epoch": 0.6228189507304667, "grad_norm": 3.0234525203704834, "learning_rate": 5.517348068288402e-06, "loss": 0.2777, "step": 10530 }, { "epoch": 0.6234104217188147, "grad_norm": 3.067286968231201, "learning_rate": 5.5163055357230225e-06, "loss": 0.356, "step": 10540 }, { "epoch": 0.6240018927071627, "grad_norm": 2.769198417663574, "learning_rate": 5.515261977150103e-06, "loss": 0.3484, "step": 10550 }, { "epoch": 0.6245933636955108, "grad_norm": 3.1559340953826904, "learning_rate": 5.514217392995147e-06, "loss": 0.3104, "step": 10560 }, { "epoch": 0.6251848346838588, "grad_norm": 3.1331470012664795, "learning_rate": 5.513171783684076e-06, "loss": 0.2893, "step": 10570 }, { "epoch": 0.6257763056722068, "grad_norm": 2.736424207687378, "learning_rate": 5.512125149643234e-06, "loss": 0.3257, "step": 10580 }, { "epoch": 0.6263677766605548, "grad_norm": 3.401714324951172, "learning_rate": 5.511077491299376e-06, "loss": 0.3774, "step": 10590 }, { "epoch": 0.6269592476489029, "grad_norm": 3.0849132537841797, "learning_rate": 5.510028809079681e-06, "loss": 0.3421, "step": 10600 }, { "epoch": 0.6275507186372509, "grad_norm": 3.3780689239501953, "learning_rate": 5.5089791034117395e-06, "loss": 0.3069, "step": 10610 }, { "epoch": 0.6281421896255989, "grad_norm": 2.7801876068115234, "learning_rate": 5.5079283747235635e-06, "loss": 0.3117, "step": 10620 }, { "epoch": 0.6287336606139469, "grad_norm": 3.366401433944702, "learning_rate": 5.506876623443582e-06, "loss": 0.3972, "step": 10630 }, { "epoch": 0.6293251316022949, "grad_norm": 2.867943525314331, "learning_rate": 5.5058238500006384e-06, "loss": 0.3133, "step": 10640 }, { "epoch": 0.629916602590643, "grad_norm": 3.8413968086242676, "learning_rate": 5.504770054823995e-06, "loss": 0.3129, "step": 10650 }, { "epoch": 0.630508073578991, "grad_norm": 2.6922152042388916, "learning_rate": 5.5037152383433295e-06, "loss": 0.3456, "step": 10660 }, { "epoch": 0.631099544567339, "grad_norm": 1.944676399230957, "learning_rate": 5.5026594009887364e-06, "loss": 0.2938, "step": 10670 }, { "epoch": 0.631691015555687, "grad_norm": 2.952179193496704, "learning_rate": 5.501602543190729e-06, "loss": 0.3408, "step": 10680 }, { "epoch": 0.632282486544035, "grad_norm": 4.193296432495117, "learning_rate": 5.500544665380231e-06, "loss": 0.3362, "step": 10690 }, { "epoch": 0.6328739575323831, "grad_norm": 3.17592191696167, "learning_rate": 5.4994857679885876e-06, "loss": 0.3213, "step": 10700 }, { "epoch": 0.6334654285207311, "grad_norm": 2.7673869132995605, "learning_rate": 5.498425851447558e-06, "loss": 0.3436, "step": 10710 }, { "epoch": 0.6340568995090791, "grad_norm": 3.661531448364258, "learning_rate": 5.497364916189314e-06, "loss": 0.349, "step": 10720 }, { "epoch": 0.6346483704974271, "grad_norm": 2.321941614151001, "learning_rate": 5.496302962646447e-06, "loss": 0.4112, "step": 10730 }, { "epoch": 0.6352398414857752, "grad_norm": 2.632530927658081, "learning_rate": 5.49523999125196e-06, "loss": 0.357, "step": 10740 }, { "epoch": 0.6358313124741232, "grad_norm": 4.727316856384277, "learning_rate": 5.4941760024392745e-06, "loss": 0.3204, "step": 10750 }, { "epoch": 0.6364227834624712, "grad_norm": 3.3030500411987305, "learning_rate": 5.493110996642225e-06, "loss": 0.2953, "step": 10760 }, { "epoch": 0.6370142544508192, "grad_norm": 4.595522880554199, "learning_rate": 5.492044974295061e-06, "loss": 0.3486, "step": 10770 }, { "epoch": 0.6376057254391672, "grad_norm": 2.5073657035827637, "learning_rate": 5.490977935832445e-06, "loss": 0.3501, "step": 10780 }, { "epoch": 0.6381971964275153, "grad_norm": 3.0581376552581787, "learning_rate": 5.489909881689457e-06, "loss": 0.3443, "step": 10790 }, { "epoch": 0.6387886674158633, "grad_norm": 3.064723253250122, "learning_rate": 5.488840812301587e-06, "loss": 0.3379, "step": 10800 }, { "epoch": 0.6393801384042113, "grad_norm": 3.249323606491089, "learning_rate": 5.487770728104743e-06, "loss": 0.3457, "step": 10810 }, { "epoch": 0.6399716093925593, "grad_norm": 3.7704920768737793, "learning_rate": 5.486699629535244e-06, "loss": 0.3421, "step": 10820 }, { "epoch": 0.6405630803809074, "grad_norm": 3.2521049976348877, "learning_rate": 5.485627517029825e-06, "loss": 0.3302, "step": 10830 }, { "epoch": 0.6411545513692554, "grad_norm": 3.8313465118408203, "learning_rate": 5.4845543910256314e-06, "loss": 0.3563, "step": 10840 }, { "epoch": 0.6417460223576034, "grad_norm": 3.272324800491333, "learning_rate": 5.483480251960224e-06, "loss": 0.3431, "step": 10850 }, { "epoch": 0.6423374933459514, "grad_norm": 2.9159798622131348, "learning_rate": 5.482405100271577e-06, "loss": 0.3524, "step": 10860 }, { "epoch": 0.6429289643342994, "grad_norm": 2.4529192447662354, "learning_rate": 5.481328936398075e-06, "loss": 0.3251, "step": 10870 }, { "epoch": 0.6435204353226475, "grad_norm": 2.3804736137390137, "learning_rate": 5.480251760778518e-06, "loss": 0.288, "step": 10880 }, { "epoch": 0.6441119063109955, "grad_norm": 3.73533034324646, "learning_rate": 5.479173573852118e-06, "loss": 0.352, "step": 10890 }, { "epoch": 0.6447033772993435, "grad_norm": 6.60170841217041, "learning_rate": 5.4780943760584995e-06, "loss": 0.3439, "step": 10900 }, { "epoch": 0.6452948482876915, "grad_norm": 3.3231446743011475, "learning_rate": 5.477014167837696e-06, "loss": 0.3245, "step": 10910 }, { "epoch": 0.6458863192760395, "grad_norm": 2.615600824356079, "learning_rate": 5.4759329496301565e-06, "loss": 0.3369, "step": 10920 }, { "epoch": 0.6464777902643876, "grad_norm": 4.326155662536621, "learning_rate": 5.474850721876741e-06, "loss": 0.3611, "step": 10930 }, { "epoch": 0.6470692612527356, "grad_norm": 3.128275156021118, "learning_rate": 5.473767485018722e-06, "loss": 0.2891, "step": 10940 }, { "epoch": 0.6476607322410836, "grad_norm": 4.730691432952881, "learning_rate": 5.472683239497781e-06, "loss": 0.3151, "step": 10950 }, { "epoch": 0.6482522032294316, "grad_norm": 3.730571746826172, "learning_rate": 5.471597985756013e-06, "loss": 0.3633, "step": 10960 }, { "epoch": 0.6488436742177797, "grad_norm": 3.381258487701416, "learning_rate": 5.4705117242359225e-06, "loss": 0.3028, "step": 10970 }, { "epoch": 0.6494351452061277, "grad_norm": 4.794594764709473, "learning_rate": 5.469424455380427e-06, "loss": 0.3235, "step": 10980 }, { "epoch": 0.6500266161944757, "grad_norm": 2.5049755573272705, "learning_rate": 5.468336179632853e-06, "loss": 0.3632, "step": 10990 }, { "epoch": 0.6506180871828237, "grad_norm": 4.9052228927612305, "learning_rate": 5.467246897436937e-06, "loss": 0.3533, "step": 11000 }, { "epoch": 0.6512095581711717, "grad_norm": 3.098860502243042, "learning_rate": 5.466156609236827e-06, "loss": 0.3321, "step": 11010 }, { "epoch": 0.6518010291595198, "grad_norm": 2.9118754863739014, "learning_rate": 5.465065315477083e-06, "loss": 0.3462, "step": 11020 }, { "epoch": 0.6523925001478678, "grad_norm": 3.2587502002716064, "learning_rate": 5.463973016602671e-06, "loss": 0.2876, "step": 11030 }, { "epoch": 0.6529839711362158, "grad_norm": 5.329993724822998, "learning_rate": 5.462879713058969e-06, "loss": 0.2723, "step": 11040 }, { "epoch": 0.6535754421245638, "grad_norm": 2.778127908706665, "learning_rate": 5.461785405291766e-06, "loss": 0.2817, "step": 11050 }, { "epoch": 0.6541669131129119, "grad_norm": 2.403733730316162, "learning_rate": 5.460690093747257e-06, "loss": 0.3084, "step": 11060 }, { "epoch": 0.6547583841012599, "grad_norm": 4.59122371673584, "learning_rate": 5.459593778872051e-06, "loss": 0.2877, "step": 11070 }, { "epoch": 0.6553498550896079, "grad_norm": 3.250215768814087, "learning_rate": 5.458496461113161e-06, "loss": 0.3315, "step": 11080 }, { "epoch": 0.6559413260779559, "grad_norm": 3.7262184619903564, "learning_rate": 5.457398140918012e-06, "loss": 0.37, "step": 11090 }, { "epoch": 0.6565327970663039, "grad_norm": 6.55772590637207, "learning_rate": 5.456298818734438e-06, "loss": 0.394, "step": 11100 }, { "epoch": 0.657124268054652, "grad_norm": 3.423121213912964, "learning_rate": 5.455198495010679e-06, "loss": 0.3556, "step": 11110 }, { "epoch": 0.657715739043, "grad_norm": 3.0040409564971924, "learning_rate": 5.4540971701953865e-06, "loss": 0.2858, "step": 11120 }, { "epoch": 0.658307210031348, "grad_norm": 3.9907591342926025, "learning_rate": 5.452994844737618e-06, "loss": 0.3649, "step": 11130 }, { "epoch": 0.658898681019696, "grad_norm": 1.5305800437927246, "learning_rate": 5.451891519086839e-06, "loss": 0.2954, "step": 11140 }, { "epoch": 0.659490152008044, "grad_norm": 10.817150115966797, "learning_rate": 5.450787193692925e-06, "loss": 0.3318, "step": 11150 }, { "epoch": 0.6600816229963921, "grad_norm": 3.1212785243988037, "learning_rate": 5.449681869006156e-06, "loss": 0.3106, "step": 11160 }, { "epoch": 0.6606730939847401, "grad_norm": 3.419796943664551, "learning_rate": 5.448575545477222e-06, "loss": 0.3078, "step": 11170 }, { "epoch": 0.661264564973088, "grad_norm": 3.488070011138916, "learning_rate": 5.447468223557221e-06, "loss": 0.327, "step": 11180 }, { "epoch": 0.661856035961436, "grad_norm": 1.797451138496399, "learning_rate": 5.446359903697652e-06, "loss": 0.3512, "step": 11190 }, { "epoch": 0.6624475069497842, "grad_norm": 3.032282590866089, "learning_rate": 5.44525058635043e-06, "loss": 0.3138, "step": 11200 }, { "epoch": 0.6630389779381322, "grad_norm": 3.106802225112915, "learning_rate": 5.44414027196787e-06, "loss": 0.349, "step": 11210 }, { "epoch": 0.6636304489264802, "grad_norm": 1.8143367767333984, "learning_rate": 5.443028961002695e-06, "loss": 0.2965, "step": 11220 }, { "epoch": 0.6642219199148282, "grad_norm": 2.5006580352783203, "learning_rate": 5.441916653908036e-06, "loss": 0.3308, "step": 11230 }, { "epoch": 0.6648133909031761, "grad_norm": 3.149103879928589, "learning_rate": 5.440803351137429e-06, "loss": 0.3239, "step": 11240 }, { "epoch": 0.6654048618915243, "grad_norm": 5.481205940246582, "learning_rate": 5.439689053144816e-06, "loss": 0.387, "step": 11250 }, { "epoch": 0.6659963328798723, "grad_norm": 2.4086861610412598, "learning_rate": 5.438573760384546e-06, "loss": 0.3138, "step": 11260 }, { "epoch": 0.6665878038682203, "grad_norm": 3.3968355655670166, "learning_rate": 5.4374574733113696e-06, "loss": 0.3094, "step": 11270 }, { "epoch": 0.6671792748565682, "grad_norm": 2.593656063079834, "learning_rate": 5.436340192380449e-06, "loss": 0.3352, "step": 11280 }, { "epoch": 0.6677707458449164, "grad_norm": 1.9375946521759033, "learning_rate": 5.435221918047346e-06, "loss": 0.2686, "step": 11290 }, { "epoch": 0.6683622168332644, "grad_norm": 5.416504859924316, "learning_rate": 5.434102650768031e-06, "loss": 0.3555, "step": 11300 }, { "epoch": 0.6689536878216124, "grad_norm": 2.8142006397247314, "learning_rate": 5.432982390998877e-06, "loss": 0.2723, "step": 11310 }, { "epoch": 0.6695451588099604, "grad_norm": 2.0731546878814697, "learning_rate": 5.431861139196664e-06, "loss": 0.3553, "step": 11320 }, { "epoch": 0.6701366297983083, "grad_norm": 2.688056468963623, "learning_rate": 5.4307388958185766e-06, "loss": 0.2782, "step": 11330 }, { "epoch": 0.6707281007866565, "grad_norm": 3.6605935096740723, "learning_rate": 5.429615661322199e-06, "loss": 0.3161, "step": 11340 }, { "epoch": 0.6713195717750045, "grad_norm": 3.6980228424072266, "learning_rate": 5.428491436165525e-06, "loss": 0.2666, "step": 11350 }, { "epoch": 0.6719110427633525, "grad_norm": 4.189696311950684, "learning_rate": 5.427366220806949e-06, "loss": 0.3168, "step": 11360 }, { "epoch": 0.6725025137517004, "grad_norm": 3.5657267570495605, "learning_rate": 5.4262400157052725e-06, "loss": 0.3178, "step": 11370 }, { "epoch": 0.6730939847400484, "grad_norm": 1.9687237739562988, "learning_rate": 5.425112821319696e-06, "loss": 0.269, "step": 11380 }, { "epoch": 0.6736854557283966, "grad_norm": 2.4728927612304688, "learning_rate": 5.423984638109828e-06, "loss": 0.3258, "step": 11390 }, { "epoch": 0.6742769267167446, "grad_norm": 2.90541410446167, "learning_rate": 5.4228554665356766e-06, "loss": 0.3216, "step": 11400 }, { "epoch": 0.6748683977050925, "grad_norm": 3.2047619819641113, "learning_rate": 5.421725307057655e-06, "loss": 0.2979, "step": 11410 }, { "epoch": 0.6754598686934405, "grad_norm": 5.287949085235596, "learning_rate": 5.4205941601365775e-06, "loss": 0.3626, "step": 11420 }, { "epoch": 0.6760513396817887, "grad_norm": 3.233625888824463, "learning_rate": 5.419462026233663e-06, "loss": 0.3079, "step": 11430 }, { "epoch": 0.6766428106701367, "grad_norm": 3.482001543045044, "learning_rate": 5.418328905810532e-06, "loss": 0.3477, "step": 11440 }, { "epoch": 0.6772342816584846, "grad_norm": 4.905953407287598, "learning_rate": 5.417194799329206e-06, "loss": 0.3166, "step": 11450 }, { "epoch": 0.6778257526468326, "grad_norm": 3.5122032165527344, "learning_rate": 5.416059707252112e-06, "loss": 0.3131, "step": 11460 }, { "epoch": 0.6784172236351806, "grad_norm": 2.819847345352173, "learning_rate": 5.414923630042074e-06, "loss": 0.3282, "step": 11470 }, { "epoch": 0.6790086946235288, "grad_norm": 3.79093861579895, "learning_rate": 5.413786568162321e-06, "loss": 0.4104, "step": 11480 }, { "epoch": 0.6796001656118767, "grad_norm": 3.7301347255706787, "learning_rate": 5.412648522076484e-06, "loss": 0.3299, "step": 11490 }, { "epoch": 0.6801916366002247, "grad_norm": 4.639546871185303, "learning_rate": 5.411509492248592e-06, "loss": 0.3792, "step": 11500 }, { "epoch": 0.6807831075885727, "grad_norm": 2.890782117843628, "learning_rate": 5.410369479143079e-06, "loss": 0.4055, "step": 11510 }, { "epoch": 0.6813745785769209, "grad_norm": 2.778917074203491, "learning_rate": 5.409228483224776e-06, "loss": 0.3395, "step": 11520 }, { "epoch": 0.6819660495652689, "grad_norm": 3.490252733230591, "learning_rate": 5.408086504958919e-06, "loss": 0.3636, "step": 11530 }, { "epoch": 0.6825575205536168, "grad_norm": 3.5161986351013184, "learning_rate": 5.406943544811142e-06, "loss": 0.3264, "step": 11540 }, { "epoch": 0.6831489915419648, "grad_norm": 3.4243030548095703, "learning_rate": 5.405799603247479e-06, "loss": 0.2821, "step": 11550 }, { "epoch": 0.6837404625303128, "grad_norm": 2.86574387550354, "learning_rate": 5.404654680734364e-06, "loss": 0.3516, "step": 11560 }, { "epoch": 0.684331933518661, "grad_norm": 2.8214430809020996, "learning_rate": 5.403508777738633e-06, "loss": 0.2527, "step": 11570 }, { "epoch": 0.684923404507009, "grad_norm": 2.1371984481811523, "learning_rate": 5.402361894727521e-06, "loss": 0.2764, "step": 11580 }, { "epoch": 0.685514875495357, "grad_norm": 2.885420083999634, "learning_rate": 5.401214032168662e-06, "loss": 0.4384, "step": 11590 }, { "epoch": 0.6861063464837049, "grad_norm": 4.237430572509766, "learning_rate": 5.400065190530089e-06, "loss": 0.385, "step": 11600 }, { "epoch": 0.6866978174720529, "grad_norm": 3.335634231567383, "learning_rate": 5.3989153702802355e-06, "loss": 0.3246, "step": 11610 }, { "epoch": 0.687289288460401, "grad_norm": 2.773118019104004, "learning_rate": 5.397764571887933e-06, "loss": 0.2786, "step": 11620 }, { "epoch": 0.687880759448749, "grad_norm": 3.4220504760742188, "learning_rate": 5.396612795822412e-06, "loss": 0.3302, "step": 11630 }, { "epoch": 0.688472230437097, "grad_norm": 2.877349853515625, "learning_rate": 5.395460042553303e-06, "loss": 0.3052, "step": 11640 }, { "epoch": 0.689063701425445, "grad_norm": 3.984241008758545, "learning_rate": 5.394306312550633e-06, "loss": 0.3247, "step": 11650 }, { "epoch": 0.6896551724137931, "grad_norm": 5.585387229919434, "learning_rate": 5.393151606284828e-06, "loss": 0.3332, "step": 11660 }, { "epoch": 0.6902466434021411, "grad_norm": 2.5017781257629395, "learning_rate": 5.391995924226711e-06, "loss": 0.3168, "step": 11670 }, { "epoch": 0.6908381143904891, "grad_norm": 2.874974012374878, "learning_rate": 5.3908392668475065e-06, "loss": 0.3852, "step": 11680 }, { "epoch": 0.6914295853788371, "grad_norm": 2.1212170124053955, "learning_rate": 5.389681634618832e-06, "loss": 0.3175, "step": 11690 }, { "epoch": 0.6920210563671851, "grad_norm": 3.7862958908081055, "learning_rate": 5.388523028012705e-06, "loss": 0.3401, "step": 11700 }, { "epoch": 0.6926125273555332, "grad_norm": 4.624659538269043, "learning_rate": 5.387363447501541e-06, "loss": 0.3774, "step": 11710 }, { "epoch": 0.6932039983438812, "grad_norm": 2.368448257446289, "learning_rate": 5.386202893558152e-06, "loss": 0.2555, "step": 11720 }, { "epoch": 0.6937954693322292, "grad_norm": 3.7900173664093018, "learning_rate": 5.385041366655744e-06, "loss": 0.3153, "step": 11730 }, { "epoch": 0.6943869403205772, "grad_norm": 3.1193134784698486, "learning_rate": 5.383878867267924e-06, "loss": 0.3928, "step": 11740 }, { "epoch": 0.6949784113089253, "grad_norm": 4.336760520935059, "learning_rate": 5.382715395868694e-06, "loss": 0.2831, "step": 11750 }, { "epoch": 0.6955698822972733, "grad_norm": 2.4827752113342285, "learning_rate": 5.3815509529324505e-06, "loss": 0.2989, "step": 11760 }, { "epoch": 0.6961613532856213, "grad_norm": 4.220667839050293, "learning_rate": 5.380385538933989e-06, "loss": 0.3222, "step": 11770 }, { "epoch": 0.6967528242739693, "grad_norm": 3.325425863265991, "learning_rate": 5.379219154348501e-06, "loss": 0.2897, "step": 11780 }, { "epoch": 0.6973442952623173, "grad_norm": 3.5478882789611816, "learning_rate": 5.378051799651569e-06, "loss": 0.308, "step": 11790 }, { "epoch": 0.6979357662506654, "grad_norm": 3.7686927318573, "learning_rate": 5.376883475319175e-06, "loss": 0.3013, "step": 11800 }, { "epoch": 0.6985272372390134, "grad_norm": 3.0364975929260254, "learning_rate": 5.3757141818276974e-06, "loss": 0.2775, "step": 11810 }, { "epoch": 0.6991187082273614, "grad_norm": 2.87695050239563, "learning_rate": 5.374543919653909e-06, "loss": 0.3103, "step": 11820 }, { "epoch": 0.6997101792157094, "grad_norm": 2.613064765930176, "learning_rate": 5.3733726892749744e-06, "loss": 0.3691, "step": 11830 }, { "epoch": 0.7003016502040575, "grad_norm": 2.5890941619873047, "learning_rate": 5.372200491168456e-06, "loss": 0.2982, "step": 11840 }, { "epoch": 0.7008931211924055, "grad_norm": 4.796480178833008, "learning_rate": 5.371027325812312e-06, "loss": 0.3521, "step": 11850 }, { "epoch": 0.7014845921807535, "grad_norm": 1.9401779174804688, "learning_rate": 5.369853193684889e-06, "loss": 0.2975, "step": 11860 }, { "epoch": 0.7020760631691015, "grad_norm": 5.196966648101807, "learning_rate": 5.368678095264936e-06, "loss": 0.2846, "step": 11870 }, { "epoch": 0.7026675341574495, "grad_norm": 5.613885402679443, "learning_rate": 5.36750203103159e-06, "loss": 0.3461, "step": 11880 }, { "epoch": 0.7032590051457976, "grad_norm": 5.4519219398498535, "learning_rate": 5.366325001464384e-06, "loss": 0.2746, "step": 11890 }, { "epoch": 0.7038504761341456, "grad_norm": 5.337847709655762, "learning_rate": 5.365147007043243e-06, "loss": 0.3336, "step": 11900 }, { "epoch": 0.7044419471224936, "grad_norm": 3.229902744293213, "learning_rate": 5.36396804824849e-06, "loss": 0.3102, "step": 11910 }, { "epoch": 0.7050334181108416, "grad_norm": 2.9857823848724365, "learning_rate": 5.362788125560835e-06, "loss": 0.3425, "step": 11920 }, { "epoch": 0.7056248890991896, "grad_norm": 2.590820074081421, "learning_rate": 5.361607239461385e-06, "loss": 0.3654, "step": 11930 }, { "epoch": 0.7062163600875377, "grad_norm": 3.20982027053833, "learning_rate": 5.360425390431638e-06, "loss": 0.3375, "step": 11940 }, { "epoch": 0.7068078310758857, "grad_norm": 5.209718704223633, "learning_rate": 5.359242578953486e-06, "loss": 0.3562, "step": 11950 }, { "epoch": 0.7073993020642337, "grad_norm": 2.328248977661133, "learning_rate": 5.358058805509213e-06, "loss": 0.3255, "step": 11960 }, { "epoch": 0.7079907730525817, "grad_norm": 2.8075504302978516, "learning_rate": 5.356874070581495e-06, "loss": 0.2824, "step": 11970 }, { "epoch": 0.7085822440409298, "grad_norm": 3.317308187484741, "learning_rate": 5.355688374653401e-06, "loss": 0.3017, "step": 11980 }, { "epoch": 0.7091737150292778, "grad_norm": 3.500256061553955, "learning_rate": 5.354501718208389e-06, "loss": 0.2978, "step": 11990 }, { "epoch": 0.7097651860176258, "grad_norm": 3.228241205215454, "learning_rate": 5.353314101730312e-06, "loss": 0.2783, "step": 12000 }, { "epoch": 0.7103566570059738, "grad_norm": 3.6374614238739014, "learning_rate": 5.352125525703413e-06, "loss": 0.3519, "step": 12010 }, { "epoch": 0.7109481279943218, "grad_norm": 2.181389808654785, "learning_rate": 5.350935990612325e-06, "loss": 0.3141, "step": 12020 }, { "epoch": 0.7115395989826699, "grad_norm": 3.557527542114258, "learning_rate": 5.349745496942076e-06, "loss": 0.3751, "step": 12030 }, { "epoch": 0.7121310699710179, "grad_norm": 2.654552936553955, "learning_rate": 5.3485540451780785e-06, "loss": 0.3494, "step": 12040 }, { "epoch": 0.7127225409593659, "grad_norm": 3.097775459289551, "learning_rate": 5.347361635806142e-06, "loss": 0.3291, "step": 12050 }, { "epoch": 0.7133140119477139, "grad_norm": 3.323918104171753, "learning_rate": 5.346168269312464e-06, "loss": 0.3495, "step": 12060 }, { "epoch": 0.713905482936062, "grad_norm": 4.0073723793029785, "learning_rate": 5.344973946183631e-06, "loss": 0.3252, "step": 12070 }, { "epoch": 0.71449695392441, "grad_norm": 3.6085305213928223, "learning_rate": 5.343778666906621e-06, "loss": 0.345, "step": 12080 }, { "epoch": 0.715088424912758, "grad_norm": 2.277613401412964, "learning_rate": 5.342582431968801e-06, "loss": 0.31, "step": 12090 }, { "epoch": 0.715679895901106, "grad_norm": 4.799652576446533, "learning_rate": 5.341385241857929e-06, "loss": 0.291, "step": 12100 }, { "epoch": 0.716271366889454, "grad_norm": 5.189013957977295, "learning_rate": 5.3401870970621504e-06, "loss": 0.342, "step": 12110 }, { "epoch": 0.7168628378778021, "grad_norm": 4.2283525466918945, "learning_rate": 5.338987998070003e-06, "loss": 0.3391, "step": 12120 }, { "epoch": 0.7174543088661501, "grad_norm": 2.2030458450317383, "learning_rate": 5.33778794537041e-06, "loss": 0.3848, "step": 12130 }, { "epoch": 0.7180457798544981, "grad_norm": 2.7663912773132324, "learning_rate": 5.336586939452687e-06, "loss": 0.3313, "step": 12140 }, { "epoch": 0.7186372508428461, "grad_norm": 3.7383508682250977, "learning_rate": 5.335384980806534e-06, "loss": 0.3157, "step": 12150 }, { "epoch": 0.7192287218311941, "grad_norm": 3.333766460418701, "learning_rate": 5.334182069922044e-06, "loss": 0.3392, "step": 12160 }, { "epoch": 0.7198201928195422, "grad_norm": 4.4491190910339355, "learning_rate": 5.332978207289696e-06, "loss": 0.3413, "step": 12170 }, { "epoch": 0.7204116638078902, "grad_norm": 2.3313779830932617, "learning_rate": 5.331773393400357e-06, "loss": 0.3588, "step": 12180 }, { "epoch": 0.7210031347962382, "grad_norm": 2.4235215187072754, "learning_rate": 5.330567628745281e-06, "loss": 0.271, "step": 12190 }, { "epoch": 0.7215946057845862, "grad_norm": 4.509303092956543, "learning_rate": 5.329360913816114e-06, "loss": 0.3375, "step": 12200 }, { "epoch": 0.7221860767729343, "grad_norm": 3.5397791862487793, "learning_rate": 5.328153249104883e-06, "loss": 0.3163, "step": 12210 }, { "epoch": 0.7227775477612823, "grad_norm": 3.5372581481933594, "learning_rate": 5.326944635104006e-06, "loss": 0.3207, "step": 12220 }, { "epoch": 0.7233690187496303, "grad_norm": 3.0586233139038086, "learning_rate": 5.32573507230629e-06, "loss": 0.3414, "step": 12230 }, { "epoch": 0.7239604897379783, "grad_norm": 2.979261875152588, "learning_rate": 5.324524561204925e-06, "loss": 0.3337, "step": 12240 }, { "epoch": 0.7245519607263263, "grad_norm": 3.3083345890045166, "learning_rate": 5.32331310229349e-06, "loss": 0.3067, "step": 12250 }, { "epoch": 0.7251434317146744, "grad_norm": 3.4951634407043457, "learning_rate": 5.322100696065948e-06, "loss": 0.2822, "step": 12260 }, { "epoch": 0.7257349027030224, "grad_norm": 5.5618133544921875, "learning_rate": 5.320887343016651e-06, "loss": 0.3279, "step": 12270 }, { "epoch": 0.7263263736913704, "grad_norm": 3.36031174659729, "learning_rate": 5.3196730436403355e-06, "loss": 0.3515, "step": 12280 }, { "epoch": 0.7269178446797184, "grad_norm": 3.1831934452056885, "learning_rate": 5.318457798432125e-06, "loss": 0.364, "step": 12290 }, { "epoch": 0.7275093156680665, "grad_norm": 3.9166629314422607, "learning_rate": 5.317241607887527e-06, "loss": 0.352, "step": 12300 }, { "epoch": 0.7281007866564145, "grad_norm": 3.934251308441162, "learning_rate": 5.316024472502437e-06, "loss": 0.4004, "step": 12310 }, { "epoch": 0.7286922576447625, "grad_norm": 2.1537046432495117, "learning_rate": 5.3148063927731325e-06, "loss": 0.3182, "step": 12320 }, { "epoch": 0.7292837286331105, "grad_norm": 1.9447267055511475, "learning_rate": 5.313587369196279e-06, "loss": 0.2811, "step": 12330 }, { "epoch": 0.7298751996214585, "grad_norm": 3.0786426067352295, "learning_rate": 5.312367402268925e-06, "loss": 0.3505, "step": 12340 }, { "epoch": 0.7304666706098066, "grad_norm": 4.542599201202393, "learning_rate": 5.311146492488504e-06, "loss": 0.3643, "step": 12350 }, { "epoch": 0.7310581415981546, "grad_norm": 4.029390811920166, "learning_rate": 5.309924640352835e-06, "loss": 0.346, "step": 12360 }, { "epoch": 0.7316496125865026, "grad_norm": 2.1521260738372803, "learning_rate": 5.30870184636012e-06, "loss": 0.3227, "step": 12370 }, { "epoch": 0.7322410835748506, "grad_norm": 2.384255886077881, "learning_rate": 5.307478111008945e-06, "loss": 0.3275, "step": 12380 }, { "epoch": 0.7328325545631986, "grad_norm": 4.094067573547363, "learning_rate": 5.306253434798281e-06, "loss": 0.3431, "step": 12390 }, { "epoch": 0.7334240255515467, "grad_norm": 4.671746253967285, "learning_rate": 5.305027818227481e-06, "loss": 0.3268, "step": 12400 }, { "epoch": 0.7340154965398947, "grad_norm": 2.7899911403656006, "learning_rate": 5.303801261796283e-06, "loss": 0.3318, "step": 12410 }, { "epoch": 0.7346069675282427, "grad_norm": 3.1717417240142822, "learning_rate": 5.302573766004808e-06, "loss": 0.2752, "step": 12420 }, { "epoch": 0.7351984385165907, "grad_norm": 2.080063581466675, "learning_rate": 5.30134533135356e-06, "loss": 0.2711, "step": 12430 }, { "epoch": 0.7357899095049388, "grad_norm": 2.492462158203125, "learning_rate": 5.300115958343422e-06, "loss": 0.2717, "step": 12440 }, { "epoch": 0.7363813804932868, "grad_norm": 2.7198286056518555, "learning_rate": 5.298885647475668e-06, "loss": 0.3052, "step": 12450 }, { "epoch": 0.7369728514816348, "grad_norm": 3.8123819828033447, "learning_rate": 5.297654399251945e-06, "loss": 0.3352, "step": 12460 }, { "epoch": 0.7375643224699828, "grad_norm": 3.746631622314453, "learning_rate": 5.29642221417429e-06, "loss": 0.4096, "step": 12470 }, { "epoch": 0.7381557934583308, "grad_norm": 2.39164662361145, "learning_rate": 5.295189092745116e-06, "loss": 0.3, "step": 12480 }, { "epoch": 0.7387472644466789, "grad_norm": 3.462897777557373, "learning_rate": 5.293955035467222e-06, "loss": 0.3134, "step": 12490 }, { "epoch": 0.7393387354350269, "grad_norm": 4.820435047149658, "learning_rate": 5.292720042843786e-06, "loss": 0.3204, "step": 12500 }, { "epoch": 0.7399302064233749, "grad_norm": 2.2853124141693115, "learning_rate": 5.29148411537837e-06, "loss": 0.3732, "step": 12510 }, { "epoch": 0.7405216774117229, "grad_norm": 2.4368956089019775, "learning_rate": 5.290247253574914e-06, "loss": 0.3654, "step": 12520 }, { "epoch": 0.741113148400071, "grad_norm": 2.4372434616088867, "learning_rate": 5.2890094579377404e-06, "loss": 0.3004, "step": 12530 }, { "epoch": 0.741704619388419, "grad_norm": 3.7763593196868896, "learning_rate": 5.287770728971554e-06, "loss": 0.3633, "step": 12540 }, { "epoch": 0.742296090376767, "grad_norm": 6.455133438110352, "learning_rate": 5.286531067181436e-06, "loss": 0.336, "step": 12550 }, { "epoch": 0.742887561365115, "grad_norm": 4.735551834106445, "learning_rate": 5.285290473072854e-06, "loss": 0.3267, "step": 12560 }, { "epoch": 0.743479032353463, "grad_norm": 3.152033567428589, "learning_rate": 5.28404894715165e-06, "loss": 0.3269, "step": 12570 }, { "epoch": 0.7440705033418111, "grad_norm": 3.255662441253662, "learning_rate": 5.282806489924048e-06, "loss": 0.2583, "step": 12580 }, { "epoch": 0.7446619743301591, "grad_norm": 3.6491901874542236, "learning_rate": 5.281563101896655e-06, "loss": 0.4389, "step": 12590 }, { "epoch": 0.7452534453185071, "grad_norm": 5.109926223754883, "learning_rate": 5.28031878357645e-06, "loss": 0.3351, "step": 12600 }, { "epoch": 0.7458449163068551, "grad_norm": 3.657804489135742, "learning_rate": 5.2790735354708e-06, "loss": 0.3035, "step": 12610 }, { "epoch": 0.7464363872952031, "grad_norm": 3.4089672565460205, "learning_rate": 5.277827358087444e-06, "loss": 0.3619, "step": 12620 }, { "epoch": 0.7470278582835512, "grad_norm": 2.086312770843506, "learning_rate": 5.276580251934505e-06, "loss": 0.3224, "step": 12630 }, { "epoch": 0.7476193292718992, "grad_norm": 3.1427431106567383, "learning_rate": 5.275332217520482e-06, "loss": 0.3052, "step": 12640 }, { "epoch": 0.7482108002602472, "grad_norm": 5.159964084625244, "learning_rate": 5.274083255354252e-06, "loss": 0.2773, "step": 12650 }, { "epoch": 0.7488022712485952, "grad_norm": 4.193582534790039, "learning_rate": 5.272833365945072e-06, "loss": 0.3593, "step": 12660 }, { "epoch": 0.7493937422369433, "grad_norm": 2.0380265712738037, "learning_rate": 5.271582549802577e-06, "loss": 0.3376, "step": 12670 }, { "epoch": 0.7499852132252913, "grad_norm": 2.2413878440856934, "learning_rate": 5.27033080743678e-06, "loss": 0.2783, "step": 12680 }, { "epoch": 0.7505766842136393, "grad_norm": 3.8818986415863037, "learning_rate": 5.2690781393580684e-06, "loss": 0.3646, "step": 12690 }, { "epoch": 0.7511681552019873, "grad_norm": 10.003957748413086, "learning_rate": 5.267824546077213e-06, "loss": 0.3254, "step": 12700 }, { "epoch": 0.7517596261903353, "grad_norm": 3.1408584117889404, "learning_rate": 5.266570028105355e-06, "loss": 0.3347, "step": 12710 }, { "epoch": 0.7523510971786834, "grad_norm": 2.7851176261901855, "learning_rate": 5.265314585954017e-06, "loss": 0.2904, "step": 12720 }, { "epoch": 0.7529425681670314, "grad_norm": 2.728928565979004, "learning_rate": 5.2640582201351e-06, "loss": 0.3308, "step": 12730 }, { "epoch": 0.7535340391553794, "grad_norm": 2.835986614227295, "learning_rate": 5.262800931160876e-06, "loss": 0.3684, "step": 12740 }, { "epoch": 0.7541255101437274, "grad_norm": 3.447253942489624, "learning_rate": 5.261542719543999e-06, "loss": 0.3257, "step": 12750 }, { "epoch": 0.7547169811320755, "grad_norm": 2.2112700939178467, "learning_rate": 5.2602835857974946e-06, "loss": 0.3301, "step": 12760 }, { "epoch": 0.7553084521204235, "grad_norm": 2.6507296562194824, "learning_rate": 5.259023530434769e-06, "loss": 0.3578, "step": 12770 }, { "epoch": 0.7558999231087715, "grad_norm": 2.42807674407959, "learning_rate": 5.2577625539695985e-06, "loss": 0.3592, "step": 12780 }, { "epoch": 0.7564913940971195, "grad_norm": 3.2352101802825928, "learning_rate": 5.256500656916143e-06, "loss": 0.2796, "step": 12790 }, { "epoch": 0.7570828650854675, "grad_norm": 6.792921543121338, "learning_rate": 5.255237839788927e-06, "loss": 0.3351, "step": 12800 }, { "epoch": 0.7576743360738156, "grad_norm": 3.093815803527832, "learning_rate": 5.253974103102861e-06, "loss": 0.3475, "step": 12810 }, { "epoch": 0.7582658070621636, "grad_norm": 3.6442697048187256, "learning_rate": 5.252709447373224e-06, "loss": 0.3551, "step": 12820 }, { "epoch": 0.7588572780505116, "grad_norm": 3.0458221435546875, "learning_rate": 5.25144387311567e-06, "loss": 0.3035, "step": 12830 }, { "epoch": 0.7594487490388596, "grad_norm": 2.160700798034668, "learning_rate": 5.25017738084623e-06, "loss": 0.3776, "step": 12840 }, { "epoch": 0.7600402200272076, "grad_norm": 3.839510202407837, "learning_rate": 5.24890997108131e-06, "loss": 0.3033, "step": 12850 }, { "epoch": 0.7606316910155557, "grad_norm": 2.71405291557312, "learning_rate": 5.247641644337684e-06, "loss": 0.351, "step": 12860 }, { "epoch": 0.7612231620039037, "grad_norm": 3.9637839794158936, "learning_rate": 5.246372401132508e-06, "loss": 0.3289, "step": 12870 }, { "epoch": 0.7618146329922517, "grad_norm": 2.4480831623077393, "learning_rate": 5.245102241983306e-06, "loss": 0.3644, "step": 12880 }, { "epoch": 0.7624061039805997, "grad_norm": 3.1675217151641846, "learning_rate": 5.2438311674079766e-06, "loss": 0.3418, "step": 12890 }, { "epoch": 0.7629975749689478, "grad_norm": 3.705007553100586, "learning_rate": 5.242559177924794e-06, "loss": 0.3121, "step": 12900 }, { "epoch": 0.7635890459572958, "grad_norm": 2.203047513961792, "learning_rate": 5.2412862740524035e-06, "loss": 0.3477, "step": 12910 }, { "epoch": 0.7641805169456438, "grad_norm": 4.609606742858887, "learning_rate": 5.2400124563098236e-06, "loss": 0.4127, "step": 12920 }, { "epoch": 0.7647719879339918, "grad_norm": 2.402787208557129, "learning_rate": 5.238737725216443e-06, "loss": 0.3306, "step": 12930 }, { "epoch": 0.7653634589223398, "grad_norm": 2.287241220474243, "learning_rate": 5.23746208129203e-06, "loss": 0.2749, "step": 12940 }, { "epoch": 0.7659549299106879, "grad_norm": 2.1896770000457764, "learning_rate": 5.2361855250567155e-06, "loss": 0.2825, "step": 12950 }, { "epoch": 0.7665464008990359, "grad_norm": 3.3045871257781982, "learning_rate": 5.23490805703101e-06, "loss": 0.2753, "step": 12960 }, { "epoch": 0.7671378718873839, "grad_norm": 2.6399848461151123, "learning_rate": 5.233629677735791e-06, "loss": 0.283, "step": 12970 }, { "epoch": 0.7677293428757319, "grad_norm": 3.206524610519409, "learning_rate": 5.232350387692312e-06, "loss": 0.3066, "step": 12980 }, { "epoch": 0.76832081386408, "grad_norm": 4.996457576751709, "learning_rate": 5.231070187422193e-06, "loss": 0.3171, "step": 12990 }, { "epoch": 0.768912284852428, "grad_norm": 5.871002197265625, "learning_rate": 5.229789077447428e-06, "loss": 0.4273, "step": 13000 }, { "epoch": 0.769503755840776, "grad_norm": 2.2439823150634766, "learning_rate": 5.228507058290382e-06, "loss": 0.3652, "step": 13010 }, { "epoch": 0.770095226829124, "grad_norm": 3.308121681213379, "learning_rate": 5.227224130473791e-06, "loss": 0.2896, "step": 13020 }, { "epoch": 0.770686697817472, "grad_norm": 3.856656074523926, "learning_rate": 5.2259402945207585e-06, "loss": 0.3303, "step": 13030 }, { "epoch": 0.7712781688058201, "grad_norm": 2.9920666217803955, "learning_rate": 5.224655550954762e-06, "loss": 0.3246, "step": 13040 }, { "epoch": 0.7718696397941681, "grad_norm": 4.6950201988220215, "learning_rate": 5.2233699002996476e-06, "loss": 0.3468, "step": 13050 }, { "epoch": 0.7724611107825161, "grad_norm": 2.537060022354126, "learning_rate": 5.22208334307963e-06, "loss": 0.3634, "step": 13060 }, { "epoch": 0.7730525817708641, "grad_norm": 2.251004219055176, "learning_rate": 5.220795879819295e-06, "loss": 0.2926, "step": 13070 }, { "epoch": 0.7736440527592121, "grad_norm": 2.45076584815979, "learning_rate": 5.219507511043599e-06, "loss": 0.3218, "step": 13080 }, { "epoch": 0.7742355237475602, "grad_norm": 2.810533046722412, "learning_rate": 5.2182182372778635e-06, "loss": 0.316, "step": 13090 }, { "epoch": 0.7748269947359082, "grad_norm": 4.232554912567139, "learning_rate": 5.216928059047785e-06, "loss": 0.2649, "step": 13100 }, { "epoch": 0.7754184657242562, "grad_norm": 3.0219340324401855, "learning_rate": 5.215636976879422e-06, "loss": 0.3667, "step": 13110 }, { "epoch": 0.7760099367126042, "grad_norm": 2.7820751667022705, "learning_rate": 5.214344991299207e-06, "loss": 0.2923, "step": 13120 }, { "epoch": 0.7766014077009523, "grad_norm": 4.003213405609131, "learning_rate": 5.213052102833938e-06, "loss": 0.2967, "step": 13130 }, { "epoch": 0.7771928786893003, "grad_norm": 4.858673572540283, "learning_rate": 5.211758312010783e-06, "loss": 0.3263, "step": 13140 }, { "epoch": 0.7777843496776483, "grad_norm": 3.2959988117218018, "learning_rate": 5.210463619357276e-06, "loss": 0.3186, "step": 13150 }, { "epoch": 0.7783758206659963, "grad_norm": 3.1034111976623535, "learning_rate": 5.209168025401321e-06, "loss": 0.3723, "step": 13160 }, { "epoch": 0.7789672916543443, "grad_norm": 5.402548313140869, "learning_rate": 5.207871530671188e-06, "loss": 0.3438, "step": 13170 }, { "epoch": 0.7795587626426924, "grad_norm": 4.2821221351623535, "learning_rate": 5.2065741356955124e-06, "loss": 0.3709, "step": 13180 }, { "epoch": 0.7801502336310404, "grad_norm": 1.479868769645691, "learning_rate": 5.205275841003301e-06, "loss": 0.2942, "step": 13190 }, { "epoch": 0.7807417046193884, "grad_norm": 5.280636310577393, "learning_rate": 5.203976647123924e-06, "loss": 0.3272, "step": 13200 }, { "epoch": 0.7813331756077364, "grad_norm": 3.28999400138855, "learning_rate": 5.202676554587119e-06, "loss": 0.3058, "step": 13210 }, { "epoch": 0.7819246465960845, "grad_norm": 2.9875168800354004, "learning_rate": 5.201375563922993e-06, "loss": 0.3193, "step": 13220 }, { "epoch": 0.7825161175844325, "grad_norm": 3.8022680282592773, "learning_rate": 5.200073675662013e-06, "loss": 0.2998, "step": 13230 }, { "epoch": 0.7831075885727805, "grad_norm": 3.6042051315307617, "learning_rate": 5.198770890335018e-06, "loss": 0.3571, "step": 13240 }, { "epoch": 0.7836990595611285, "grad_norm": 4.403831958770752, "learning_rate": 5.19746720847321e-06, "loss": 0.2898, "step": 13250 }, { "epoch": 0.7842905305494765, "grad_norm": 2.6784770488739014, "learning_rate": 5.1961626306081575e-06, "loss": 0.3283, "step": 13260 }, { "epoch": 0.7848820015378246, "grad_norm": 3.043395757675171, "learning_rate": 5.194857157271792e-06, "loss": 0.3087, "step": 13270 }, { "epoch": 0.7854734725261726, "grad_norm": 5.086563587188721, "learning_rate": 5.193550788996414e-06, "loss": 0.3551, "step": 13280 }, { "epoch": 0.7860649435145206, "grad_norm": 2.5777673721313477, "learning_rate": 5.192243526314686e-06, "loss": 0.2938, "step": 13290 }, { "epoch": 0.7866564145028686, "grad_norm": 5.048000335693359, "learning_rate": 5.190935369759635e-06, "loss": 0.321, "step": 13300 }, { "epoch": 0.7872478854912166, "grad_norm": 3.078775644302368, "learning_rate": 5.189626319864654e-06, "loss": 0.3096, "step": 13310 }, { "epoch": 0.7878393564795647, "grad_norm": 2.8679120540618896, "learning_rate": 5.188316377163502e-06, "loss": 0.2779, "step": 13320 }, { "epoch": 0.7884308274679127, "grad_norm": 3.0509657859802246, "learning_rate": 5.1870055421902974e-06, "loss": 0.324, "step": 13330 }, { "epoch": 0.7890222984562607, "grad_norm": 3.491056203842163, "learning_rate": 5.185693815479525e-06, "loss": 0.312, "step": 13340 }, { "epoch": 0.7896137694446087, "grad_norm": 3.975297689437866, "learning_rate": 5.184381197566033e-06, "loss": 0.2889, "step": 13350 }, { "epoch": 0.7902052404329568, "grad_norm": 2.335360527038574, "learning_rate": 5.183067688985034e-06, "loss": 0.2905, "step": 13360 }, { "epoch": 0.7907967114213048, "grad_norm": 2.7606074810028076, "learning_rate": 5.181753290272102e-06, "loss": 0.3801, "step": 13370 }, { "epoch": 0.7913881824096528, "grad_norm": 3.249248504638672, "learning_rate": 5.180438001963176e-06, "loss": 0.3443, "step": 13380 }, { "epoch": 0.7919796533980008, "grad_norm": 3.6370067596435547, "learning_rate": 5.179121824594553e-06, "loss": 0.3553, "step": 13390 }, { "epoch": 0.7925711243863488, "grad_norm": 4.121008396148682, "learning_rate": 5.1778047587029e-06, "loss": 0.3169, "step": 13400 }, { "epoch": 0.7931625953746969, "grad_norm": 3.620272636413574, "learning_rate": 5.17648680482524e-06, "loss": 0.3278, "step": 13410 }, { "epoch": 0.7937540663630449, "grad_norm": 3.696549654006958, "learning_rate": 5.17516796349896e-06, "loss": 0.3494, "step": 13420 }, { "epoch": 0.7943455373513929, "grad_norm": 2.5770206451416016, "learning_rate": 5.17384823526181e-06, "loss": 0.3632, "step": 13430 }, { "epoch": 0.7949370083397409, "grad_norm": 2.1182029247283936, "learning_rate": 5.172527620651902e-06, "loss": 0.3302, "step": 13440 }, { "epoch": 0.795528479328089, "grad_norm": 3.9199745655059814, "learning_rate": 5.171206120207705e-06, "loss": 0.2976, "step": 13450 }, { "epoch": 0.796119950316437, "grad_norm": 2.131645441055298, "learning_rate": 5.169883734468055e-06, "loss": 0.3081, "step": 13460 }, { "epoch": 0.796711421304785, "grad_norm": 3.5527892112731934, "learning_rate": 5.168560463972145e-06, "loss": 0.2655, "step": 13470 }, { "epoch": 0.797302892293133, "grad_norm": 3.612787961959839, "learning_rate": 5.167236309259531e-06, "loss": 0.3579, "step": 13480 }, { "epoch": 0.797894363281481, "grad_norm": 2.415775775909424, "learning_rate": 5.1659112708701285e-06, "loss": 0.3429, "step": 13490 }, { "epoch": 0.7984858342698291, "grad_norm": 3.2057440280914307, "learning_rate": 5.164585349344213e-06, "loss": 0.3006, "step": 13500 }, { "epoch": 0.7990773052581771, "grad_norm": 2.4629061222076416, "learning_rate": 5.163258545222421e-06, "loss": 0.3246, "step": 13510 }, { "epoch": 0.7996687762465251, "grad_norm": 2.0737671852111816, "learning_rate": 5.161930859045748e-06, "loss": 0.3126, "step": 13520 }, { "epoch": 0.8002602472348731, "grad_norm": 2.0249853134155273, "learning_rate": 5.160602291355549e-06, "loss": 0.3378, "step": 13530 }, { "epoch": 0.8008517182232211, "grad_norm": 3.2189083099365234, "learning_rate": 5.1592728426935415e-06, "loss": 0.3078, "step": 13540 }, { "epoch": 0.8014431892115692, "grad_norm": 3.11150860786438, "learning_rate": 5.157942513601796e-06, "loss": 0.3504, "step": 13550 }, { "epoch": 0.8020346601999172, "grad_norm": 3.5194919109344482, "learning_rate": 5.15661130462275e-06, "loss": 0.3341, "step": 13560 }, { "epoch": 0.8026261311882652, "grad_norm": 4.16337251663208, "learning_rate": 5.155279216299192e-06, "loss": 0.3391, "step": 13570 }, { "epoch": 0.8032176021766132, "grad_norm": 1.6421558856964111, "learning_rate": 5.153946249174274e-06, "loss": 0.2897, "step": 13580 }, { "epoch": 0.8038090731649613, "grad_norm": 3.2226152420043945, "learning_rate": 5.152612403791504e-06, "loss": 0.3586, "step": 13590 }, { "epoch": 0.8044005441533093, "grad_norm": 4.544743537902832, "learning_rate": 5.151277680694749e-06, "loss": 0.34, "step": 13600 }, { "epoch": 0.8049920151416573, "grad_norm": 2.445129632949829, "learning_rate": 5.149942080428235e-06, "loss": 0.3283, "step": 13610 }, { "epoch": 0.8055834861300053, "grad_norm": 3.7136645317077637, "learning_rate": 5.148605603536545e-06, "loss": 0.3667, "step": 13620 }, { "epoch": 0.8061749571183533, "grad_norm": 2.2281336784362793, "learning_rate": 5.147268250564616e-06, "loss": 0.3182, "step": 13630 }, { "epoch": 0.8067664281067014, "grad_norm": 2.29670786857605, "learning_rate": 5.145930022057747e-06, "loss": 0.3786, "step": 13640 }, { "epoch": 0.8073578990950494, "grad_norm": 3.5225377082824707, "learning_rate": 5.144590918561593e-06, "loss": 0.2826, "step": 13650 }, { "epoch": 0.8079493700833974, "grad_norm": 3.4138247966766357, "learning_rate": 5.143250940622164e-06, "loss": 0.3251, "step": 13660 }, { "epoch": 0.8085408410717454, "grad_norm": 3.6082794666290283, "learning_rate": 5.141910088785827e-06, "loss": 0.3055, "step": 13670 }, { "epoch": 0.8091323120600935, "grad_norm": 2.155980110168457, "learning_rate": 5.140568363599306e-06, "loss": 0.3266, "step": 13680 }, { "epoch": 0.8097237830484415, "grad_norm": 2.235934257507324, "learning_rate": 5.1392257656096825e-06, "loss": 0.2488, "step": 13690 }, { "epoch": 0.8103152540367895, "grad_norm": 4.20320463180542, "learning_rate": 5.137882295364389e-06, "loss": 0.2759, "step": 13700 }, { "epoch": 0.8109067250251375, "grad_norm": 3.7266271114349365, "learning_rate": 5.136537953411221e-06, "loss": 0.2782, "step": 13710 }, { "epoch": 0.8114981960134855, "grad_norm": 2.8277409076690674, "learning_rate": 5.135192740298321e-06, "loss": 0.3431, "step": 13720 }, { "epoch": 0.8120896670018336, "grad_norm": 2.5839099884033203, "learning_rate": 5.133846656574194e-06, "loss": 0.2633, "step": 13730 }, { "epoch": 0.8126811379901816, "grad_norm": 2.6259243488311768, "learning_rate": 5.132499702787696e-06, "loss": 0.3753, "step": 13740 }, { "epoch": 0.8132726089785296, "grad_norm": 4.49720573425293, "learning_rate": 5.13115187948804e-06, "loss": 0.3003, "step": 13750 }, { "epoch": 0.8138640799668776, "grad_norm": 2.0331649780273438, "learning_rate": 5.129803187224791e-06, "loss": 0.3197, "step": 13760 }, { "epoch": 0.8144555509552256, "grad_norm": 2.6688144207000732, "learning_rate": 5.12845362654787e-06, "loss": 0.2943, "step": 13770 }, { "epoch": 0.8150470219435737, "grad_norm": 3.046826124191284, "learning_rate": 5.127103198007552e-06, "loss": 0.3954, "step": 13780 }, { "epoch": 0.8156384929319217, "grad_norm": 2.6698033809661865, "learning_rate": 5.1257519021544644e-06, "loss": 0.3606, "step": 13790 }, { "epoch": 0.8162299639202697, "grad_norm": 2.9641878604888916, "learning_rate": 5.124399739539593e-06, "loss": 0.3522, "step": 13800 }, { "epoch": 0.8168214349086177, "grad_norm": 2.5664725303649902, "learning_rate": 5.123046710714269e-06, "loss": 0.3344, "step": 13810 }, { "epoch": 0.8174129058969658, "grad_norm": 3.4095869064331055, "learning_rate": 5.121692816230185e-06, "loss": 0.3331, "step": 13820 }, { "epoch": 0.8180043768853138, "grad_norm": 2.25933575630188, "learning_rate": 5.12033805663938e-06, "loss": 0.2898, "step": 13830 }, { "epoch": 0.8185958478736618, "grad_norm": 3.4331119060516357, "learning_rate": 5.118982432494251e-06, "loss": 0.3119, "step": 13840 }, { "epoch": 0.8191873188620098, "grad_norm": 2.8013863563537598, "learning_rate": 5.117625944347543e-06, "loss": 0.3246, "step": 13850 }, { "epoch": 0.8197787898503578, "grad_norm": 3.3788537979125977, "learning_rate": 5.116268592752356e-06, "loss": 0.3527, "step": 13860 }, { "epoch": 0.8203702608387059, "grad_norm": 2.168949604034424, "learning_rate": 5.114910378262142e-06, "loss": 0.3625, "step": 13870 }, { "epoch": 0.8209617318270539, "grad_norm": 2.470757246017456, "learning_rate": 5.113551301430704e-06, "loss": 0.2796, "step": 13880 }, { "epoch": 0.8215532028154019, "grad_norm": 2.8372480869293213, "learning_rate": 5.1121913628121985e-06, "loss": 0.5014, "step": 13890 }, { "epoch": 0.8221446738037499, "grad_norm": 3.5244507789611816, "learning_rate": 5.110830562961129e-06, "loss": 0.2917, "step": 13900 }, { "epoch": 0.822736144792098, "grad_norm": 2.987863063812256, "learning_rate": 5.109468902432354e-06, "loss": 0.3152, "step": 13910 }, { "epoch": 0.823327615780446, "grad_norm": 4.159335613250732, "learning_rate": 5.1081063817810826e-06, "loss": 0.3509, "step": 13920 }, { "epoch": 0.823919086768794, "grad_norm": 2.291978359222412, "learning_rate": 5.106743001562874e-06, "loss": 0.3236, "step": 13930 }, { "epoch": 0.824510557757142, "grad_norm": 5.336324691772461, "learning_rate": 5.105378762333636e-06, "loss": 0.3106, "step": 13940 }, { "epoch": 0.82510202874549, "grad_norm": 2.520005941390991, "learning_rate": 5.10401366464963e-06, "loss": 0.3018, "step": 13950 }, { "epoch": 0.8256934997338381, "grad_norm": 4.3336358070373535, "learning_rate": 5.1026477090674656e-06, "loss": 0.3489, "step": 13960 }, { "epoch": 0.8262849707221861, "grad_norm": 3.328267812728882, "learning_rate": 5.101280896144102e-06, "loss": 0.3707, "step": 13970 }, { "epoch": 0.8268764417105341, "grad_norm": 2.5179057121276855, "learning_rate": 5.09991322643685e-06, "loss": 0.3299, "step": 13980 }, { "epoch": 0.8274679126988821, "grad_norm": 3.438192129135132, "learning_rate": 5.0985447005033675e-06, "loss": 0.3023, "step": 13990 }, { "epoch": 0.8280593836872301, "grad_norm": 4.702442646026611, "learning_rate": 5.097175318901661e-06, "loss": 0.3275, "step": 14000 }, { "epoch": 0.8286508546755782, "grad_norm": 2.333428144454956, "learning_rate": 5.095805082190089e-06, "loss": 0.3167, "step": 14010 }, { "epoch": 0.8292423256639262, "grad_norm": 2.9614479541778564, "learning_rate": 5.094433990927354e-06, "loss": 0.3369, "step": 14020 }, { "epoch": 0.8298337966522742, "grad_norm": 3.3690786361694336, "learning_rate": 5.093062045672512e-06, "loss": 0.2858, "step": 14030 }, { "epoch": 0.8304252676406222, "grad_norm": 3.3415915966033936, "learning_rate": 5.091689246984965e-06, "loss": 0.3067, "step": 14040 }, { "epoch": 0.8310167386289703, "grad_norm": 3.7526943683624268, "learning_rate": 5.090315595424461e-06, "loss": 0.3654, "step": 14050 }, { "epoch": 0.8316082096173183, "grad_norm": 3.317004919052124, "learning_rate": 5.0889410915511006e-06, "loss": 0.3768, "step": 14060 }, { "epoch": 0.8321996806056663, "grad_norm": 2.5512049198150635, "learning_rate": 5.087565735925326e-06, "loss": 0.3893, "step": 14070 }, { "epoch": 0.8327911515940143, "grad_norm": 2.022367238998413, "learning_rate": 5.086189529107932e-06, "loss": 0.3502, "step": 14080 }, { "epoch": 0.8333826225823623, "grad_norm": 3.1597325801849365, "learning_rate": 5.084812471660056e-06, "loss": 0.3538, "step": 14090 }, { "epoch": 0.8339740935707104, "grad_norm": 3.700450897216797, "learning_rate": 5.0834345641431874e-06, "loss": 0.3142, "step": 14100 }, { "epoch": 0.8345655645590584, "grad_norm": 4.872536659240723, "learning_rate": 5.082055807119157e-06, "loss": 0.37, "step": 14110 }, { "epoch": 0.8351570355474064, "grad_norm": 2.880089521408081, "learning_rate": 5.080676201150143e-06, "loss": 0.3284, "step": 14120 }, { "epoch": 0.8357485065357544, "grad_norm": 1.912048101425171, "learning_rate": 5.0792957467986735e-06, "loss": 0.2625, "step": 14130 }, { "epoch": 0.8363399775241025, "grad_norm": 2.727766513824463, "learning_rate": 5.07791444462762e-06, "loss": 0.3345, "step": 14140 }, { "epoch": 0.8369314485124505, "grad_norm": 3.5660438537597656, "learning_rate": 5.0765322952001975e-06, "loss": 0.3226, "step": 14150 }, { "epoch": 0.8375229195007985, "grad_norm": 2.8701486587524414, "learning_rate": 5.07514929907997e-06, "loss": 0.3079, "step": 14160 }, { "epoch": 0.8381143904891465, "grad_norm": 2.5898361206054688, "learning_rate": 5.073765456830845e-06, "loss": 0.3214, "step": 14170 }, { "epoch": 0.8387058614774945, "grad_norm": 3.130897283554077, "learning_rate": 5.072380769017075e-06, "loss": 0.2755, "step": 14180 }, { "epoch": 0.8392973324658426, "grad_norm": 4.4185638427734375, "learning_rate": 5.0709952362032594e-06, "loss": 0.3442, "step": 14190 }, { "epoch": 0.8398888034541906, "grad_norm": 3.865841865539551, "learning_rate": 5.069608858954338e-06, "loss": 0.3253, "step": 14200 }, { "epoch": 0.8404802744425386, "grad_norm": 3.323648691177368, "learning_rate": 5.068221637835599e-06, "loss": 0.294, "step": 14210 }, { "epoch": 0.8410717454308866, "grad_norm": 2.6683194637298584, "learning_rate": 5.066833573412673e-06, "loss": 0.324, "step": 14220 }, { "epoch": 0.8416632164192346, "grad_norm": 4.430836200714111, "learning_rate": 5.065444666251532e-06, "loss": 0.3293, "step": 14230 }, { "epoch": 0.8422546874075827, "grad_norm": 1.900251030921936, "learning_rate": 5.064054916918497e-06, "loss": 0.3059, "step": 14240 }, { "epoch": 0.8428461583959307, "grad_norm": 3.730832576751709, "learning_rate": 5.062664325980227e-06, "loss": 0.3444, "step": 14250 }, { "epoch": 0.8434376293842787, "grad_norm": 2.8013036251068115, "learning_rate": 5.06127289400373e-06, "loss": 0.3225, "step": 14260 }, { "epoch": 0.8440291003726267, "grad_norm": 2.473881959915161, "learning_rate": 5.059880621556349e-06, "loss": 0.2901, "step": 14270 }, { "epoch": 0.8446205713609748, "grad_norm": 5.923923015594482, "learning_rate": 5.058487509205778e-06, "loss": 0.4225, "step": 14280 }, { "epoch": 0.8452120423493228, "grad_norm": 1.8776572942733765, "learning_rate": 5.057093557520047e-06, "loss": 0.3064, "step": 14290 }, { "epoch": 0.8458035133376708, "grad_norm": 4.994443416595459, "learning_rate": 5.055698767067533e-06, "loss": 0.2607, "step": 14300 }, { "epoch": 0.8463949843260188, "grad_norm": 3.0978431701660156, "learning_rate": 5.054303138416952e-06, "loss": 0.325, "step": 14310 }, { "epoch": 0.8469864553143668, "grad_norm": 3.247818946838379, "learning_rate": 5.052906672137363e-06, "loss": 0.3316, "step": 14320 }, { "epoch": 0.8475779263027149, "grad_norm": 3.963541030883789, "learning_rate": 5.051509368798165e-06, "loss": 0.3378, "step": 14330 }, { "epoch": 0.8481693972910629, "grad_norm": 2.363816499710083, "learning_rate": 5.050111228969101e-06, "loss": 0.3256, "step": 14340 }, { "epoch": 0.8487608682794109, "grad_norm": 5.362619876861572, "learning_rate": 5.048712253220253e-06, "loss": 0.341, "step": 14350 }, { "epoch": 0.8493523392677589, "grad_norm": 4.195221424102783, "learning_rate": 5.047312442122045e-06, "loss": 0.3649, "step": 14360 }, { "epoch": 0.849943810256107, "grad_norm": 4.0399041175842285, "learning_rate": 5.045911796245241e-06, "loss": 0.2876, "step": 14370 }, { "epoch": 0.850535281244455, "grad_norm": 1.971643328666687, "learning_rate": 5.044510316160943e-06, "loss": 0.3466, "step": 14380 }, { "epoch": 0.851126752232803, "grad_norm": 2.553417205810547, "learning_rate": 5.0431080024405995e-06, "loss": 0.3028, "step": 14390 }, { "epoch": 0.851718223221151, "grad_norm": 3.676905870437622, "learning_rate": 5.041704855655993e-06, "loss": 0.2948, "step": 14400 }, { "epoch": 0.852309694209499, "grad_norm": 3.6025543212890625, "learning_rate": 5.040300876379246e-06, "loss": 0.2583, "step": 14410 }, { "epoch": 0.8529011651978471, "grad_norm": 2.7310335636138916, "learning_rate": 5.038896065182826e-06, "loss": 0.3242, "step": 14420 }, { "epoch": 0.8534926361861951, "grad_norm": 3.0755257606506348, "learning_rate": 5.0374904226395326e-06, "loss": 0.3121, "step": 14430 }, { "epoch": 0.8540841071745431, "grad_norm": 2.6045331954956055, "learning_rate": 5.036083949322508e-06, "loss": 0.338, "step": 14440 }, { "epoch": 0.8546755781628911, "grad_norm": 3.976710081100464, "learning_rate": 5.0346766458052346e-06, "loss": 0.2978, "step": 14450 }, { "epoch": 0.8552670491512391, "grad_norm": 2.166835069656372, "learning_rate": 5.033268512661529e-06, "loss": 0.3243, "step": 14460 }, { "epoch": 0.8558585201395872, "grad_norm": 2.6948862075805664, "learning_rate": 5.0318595504655506e-06, "loss": 0.3164, "step": 14470 }, { "epoch": 0.8564499911279352, "grad_norm": 3.890519380569458, "learning_rate": 5.030449759791793e-06, "loss": 0.3333, "step": 14480 }, { "epoch": 0.8570414621162832, "grad_norm": 3.347702741622925, "learning_rate": 5.0290391412150905e-06, "loss": 0.395, "step": 14490 }, { "epoch": 0.8576329331046312, "grad_norm": 4.296267986297607, "learning_rate": 5.027627695310614e-06, "loss": 0.3458, "step": 14500 }, { "epoch": 0.8582244040929793, "grad_norm": 2.42610764503479, "learning_rate": 5.026215422653871e-06, "loss": 0.3213, "step": 14510 }, { "epoch": 0.8588158750813273, "grad_norm": 3.000802516937256, "learning_rate": 5.024802323820706e-06, "loss": 0.3129, "step": 14520 }, { "epoch": 0.8594073460696753, "grad_norm": 3.924447774887085, "learning_rate": 5.0233883993873024e-06, "loss": 0.3379, "step": 14530 }, { "epoch": 0.8599988170580233, "grad_norm": 2.056272029876709, "learning_rate": 5.021973649930178e-06, "loss": 0.3511, "step": 14540 }, { "epoch": 0.8605902880463713, "grad_norm": 2.9029440879821777, "learning_rate": 5.020558076026189e-06, "loss": 0.3464, "step": 14550 }, { "epoch": 0.8611817590347194, "grad_norm": 3.0985467433929443, "learning_rate": 5.019141678252525e-06, "loss": 0.3396, "step": 14560 }, { "epoch": 0.8617732300230674, "grad_norm": 2.039576768875122, "learning_rate": 5.0177244571867135e-06, "loss": 0.3231, "step": 14570 }, { "epoch": 0.8623647010114154, "grad_norm": 2.7131338119506836, "learning_rate": 5.016306413406619e-06, "loss": 0.284, "step": 14580 }, { "epoch": 0.8629561719997634, "grad_norm": 3.423962354660034, "learning_rate": 5.014887547490438e-06, "loss": 0.2832, "step": 14590 }, { "epoch": 0.8635476429881115, "grad_norm": 2.974243640899658, "learning_rate": 5.013467860016704e-06, "loss": 0.3682, "step": 14600 }, { "epoch": 0.8641391139764595, "grad_norm": 4.154688835144043, "learning_rate": 5.0120473515642865e-06, "loss": 0.2859, "step": 14610 }, { "epoch": 0.8647305849648075, "grad_norm": 2.5508038997650146, "learning_rate": 5.010626022712389e-06, "loss": 0.338, "step": 14620 }, { "epoch": 0.8653220559531555, "grad_norm": 3.480614423751831, "learning_rate": 5.009203874040548e-06, "loss": 0.2973, "step": 14630 }, { "epoch": 0.8659135269415035, "grad_norm": 3.014747381210327, "learning_rate": 5.007780906128638e-06, "loss": 0.3316, "step": 14640 }, { "epoch": 0.8665049979298516, "grad_norm": 4.795420169830322, "learning_rate": 5.006357119556862e-06, "loss": 0.3344, "step": 14650 }, { "epoch": 0.8670964689181996, "grad_norm": 3.8607006072998047, "learning_rate": 5.004932514905761e-06, "loss": 0.334, "step": 14660 }, { "epoch": 0.8676879399065476, "grad_norm": 3.6889281272888184, "learning_rate": 5.003507092756209e-06, "loss": 0.3388, "step": 14670 }, { "epoch": 0.8682794108948956, "grad_norm": 2.7314839363098145, "learning_rate": 5.0020808536894135e-06, "loss": 0.2546, "step": 14680 }, { "epoch": 0.8688708818832436, "grad_norm": 1.742049217224121, "learning_rate": 5.000653798286912e-06, "loss": 0.2883, "step": 14690 }, { "epoch": 0.8694623528715917, "grad_norm": 5.243233680725098, "learning_rate": 4.999225927130579e-06, "loss": 0.3232, "step": 14700 }, { "epoch": 0.8700538238599397, "grad_norm": 4.112339019775391, "learning_rate": 4.99779724080262e-06, "loss": 0.3594, "step": 14710 }, { "epoch": 0.8706452948482877, "grad_norm": 3.4033620357513428, "learning_rate": 4.996367739885571e-06, "loss": 0.3171, "step": 14720 }, { "epoch": 0.8712367658366357, "grad_norm": 3.983520984649658, "learning_rate": 4.9949374249623044e-06, "loss": 0.3268, "step": 14730 }, { "epoch": 0.8718282368249838, "grad_norm": 3.0166523456573486, "learning_rate": 4.993506296616021e-06, "loss": 0.3302, "step": 14740 }, { "epoch": 0.8724197078133318, "grad_norm": 2.9674293994903564, "learning_rate": 4.992074355430253e-06, "loss": 0.3588, "step": 14750 }, { "epoch": 0.8730111788016798, "grad_norm": 2.9442312717437744, "learning_rate": 4.9906416019888675e-06, "loss": 0.3359, "step": 14760 }, { "epoch": 0.8736026497900278, "grad_norm": 3.2128243446350098, "learning_rate": 4.989208036876059e-06, "loss": 0.3438, "step": 14770 }, { "epoch": 0.8741941207783758, "grad_norm": 3.586866617202759, "learning_rate": 4.987773660676354e-06, "loss": 0.2705, "step": 14780 }, { "epoch": 0.8747855917667239, "grad_norm": 4.579881191253662, "learning_rate": 4.986338473974611e-06, "loss": 0.3544, "step": 14790 }, { "epoch": 0.8753770627550719, "grad_norm": 4.406407356262207, "learning_rate": 4.984902477356019e-06, "loss": 0.3985, "step": 14800 }, { "epoch": 0.8759685337434199, "grad_norm": 3.7108261585235596, "learning_rate": 4.983465671406096e-06, "loss": 0.3157, "step": 14810 }, { "epoch": 0.8765600047317679, "grad_norm": 4.5111284255981445, "learning_rate": 4.9820280567106895e-06, "loss": 0.3681, "step": 14820 }, { "epoch": 0.877151475720116, "grad_norm": 4.712976455688477, "learning_rate": 4.98058963385598e-06, "loss": 0.3105, "step": 14830 }, { "epoch": 0.877742946708464, "grad_norm": 3.434842348098755, "learning_rate": 4.979150403428472e-06, "loss": 0.3478, "step": 14840 }, { "epoch": 0.878334417696812, "grad_norm": 3.093197822570801, "learning_rate": 4.977710366015006e-06, "loss": 0.4058, "step": 14850 }, { "epoch": 0.87892588868516, "grad_norm": 2.774444580078125, "learning_rate": 4.976269522202745e-06, "loss": 0.3133, "step": 14860 }, { "epoch": 0.879517359673508, "grad_norm": 3.55328631401062, "learning_rate": 4.974827872579186e-06, "loss": 0.3643, "step": 14870 }, { "epoch": 0.8801088306618561, "grad_norm": 2.405346393585205, "learning_rate": 4.973385417732152e-06, "loss": 0.3252, "step": 14880 }, { "epoch": 0.8807003016502041, "grad_norm": 3.0000803470611572, "learning_rate": 4.971942158249795e-06, "loss": 0.396, "step": 14890 }, { "epoch": 0.8812917726385521, "grad_norm": 2.544468879699707, "learning_rate": 4.970498094720593e-06, "loss": 0.336, "step": 14900 }, { "epoch": 0.8818832436269001, "grad_norm": 2.8891336917877197, "learning_rate": 4.9690532277333555e-06, "loss": 0.3138, "step": 14910 }, { "epoch": 0.8824747146152481, "grad_norm": 3.010514974594116, "learning_rate": 4.967607557877217e-06, "loss": 0.2794, "step": 14920 }, { "epoch": 0.8830661856035962, "grad_norm": 2.0123350620269775, "learning_rate": 4.966161085741641e-06, "loss": 0.3422, "step": 14930 }, { "epoch": 0.8836576565919442, "grad_norm": 2.6486976146698, "learning_rate": 4.964713811916416e-06, "loss": 0.3395, "step": 14940 }, { "epoch": 0.8842491275802922, "grad_norm": 3.0548129081726074, "learning_rate": 4.96326573699166e-06, "loss": 0.3268, "step": 14950 }, { "epoch": 0.8848405985686402, "grad_norm": 3.6414635181427, "learning_rate": 4.961816861557817e-06, "loss": 0.3257, "step": 14960 }, { "epoch": 0.8854320695569883, "grad_norm": 2.86342191696167, "learning_rate": 4.960367186205655e-06, "loss": 0.3064, "step": 14970 }, { "epoch": 0.8860235405453363, "grad_norm": 3.4242985248565674, "learning_rate": 4.9589167115262694e-06, "loss": 0.3268, "step": 14980 }, { "epoch": 0.8866150115336843, "grad_norm": 1.8852741718292236, "learning_rate": 4.957465438111083e-06, "loss": 0.3352, "step": 14990 }, { "epoch": 0.8872064825220323, "grad_norm": 5.230129718780518, "learning_rate": 4.956013366551843e-06, "loss": 0.3446, "step": 15000 }, { "epoch": 0.8877979535103803, "grad_norm": 2.478756904602051, "learning_rate": 4.954560497440624e-06, "loss": 0.3238, "step": 15010 }, { "epoch": 0.8883894244987284, "grad_norm": 2.023585319519043, "learning_rate": 4.953106831369822e-06, "loss": 0.3599, "step": 15020 }, { "epoch": 0.8889808954870764, "grad_norm": 3.219345808029175, "learning_rate": 4.95165236893216e-06, "loss": 0.329, "step": 15030 }, { "epoch": 0.8895723664754244, "grad_norm": 3.7435340881347656, "learning_rate": 4.950197110720688e-06, "loss": 0.289, "step": 15040 }, { "epoch": 0.8901638374637724, "grad_norm": 4.18199348449707, "learning_rate": 4.948741057328776e-06, "loss": 0.3064, "step": 15050 }, { "epoch": 0.8907553084521205, "grad_norm": 6.682132720947266, "learning_rate": 4.947284209350121e-06, "loss": 0.33, "step": 15060 }, { "epoch": 0.8913467794404685, "grad_norm": 4.005115032196045, "learning_rate": 4.945826567378745e-06, "loss": 0.3044, "step": 15070 }, { "epoch": 0.8919382504288165, "grad_norm": 2.2128050327301025, "learning_rate": 4.94436813200899e-06, "loss": 0.3146, "step": 15080 }, { "epoch": 0.8925297214171645, "grad_norm": 3.2593445777893066, "learning_rate": 4.942908903835525e-06, "loss": 0.3394, "step": 15090 }, { "epoch": 0.8931211924055125, "grad_norm": 2.867985725402832, "learning_rate": 4.9414488834533416e-06, "loss": 0.3715, "step": 15100 }, { "epoch": 0.8937126633938606, "grad_norm": 2.8322536945343018, "learning_rate": 4.939988071457752e-06, "loss": 0.3169, "step": 15110 }, { "epoch": 0.8943041343822086, "grad_norm": 3.008557081222534, "learning_rate": 4.938526468444393e-06, "loss": 0.2856, "step": 15120 }, { "epoch": 0.8948956053705566, "grad_norm": 2.691032648086548, "learning_rate": 4.937064075009227e-06, "loss": 0.2764, "step": 15130 }, { "epoch": 0.8954870763589046, "grad_norm": 2.365212917327881, "learning_rate": 4.935600891748532e-06, "loss": 0.3333, "step": 15140 }, { "epoch": 0.8960785473472526, "grad_norm": 3.7591607570648193, "learning_rate": 4.934136919258913e-06, "loss": 0.3015, "step": 15150 }, { "epoch": 0.8966700183356007, "grad_norm": 1.9893885850906372, "learning_rate": 4.932672158137297e-06, "loss": 0.3508, "step": 15160 }, { "epoch": 0.8972614893239487, "grad_norm": 2.0009050369262695, "learning_rate": 4.931206608980927e-06, "loss": 0.3569, "step": 15170 }, { "epoch": 0.8978529603122967, "grad_norm": 2.9199464321136475, "learning_rate": 4.929740272387376e-06, "loss": 0.3287, "step": 15180 }, { "epoch": 0.8984444313006447, "grad_norm": 5.0569868087768555, "learning_rate": 4.92827314895453e-06, "loss": 0.2803, "step": 15190 }, { "epoch": 0.8990359022889928, "grad_norm": 3.728503942489624, "learning_rate": 4.926805239280601e-06, "loss": 0.3215, "step": 15200 }, { "epoch": 0.8996273732773408, "grad_norm": 3.765122175216675, "learning_rate": 4.925336543964119e-06, "loss": 0.304, "step": 15210 }, { "epoch": 0.900159697166854, "eval_accuracy": 0.8609308979605417, "eval_loss": 0.3203848898410797, "eval_runtime": 199.2873, "eval_safe_aucpr": 0.9169491189959273, "eval_safe_f1": 0.8421509761715947, "eval_safe_fpr": 0.11951810110310568, "eval_safe_precision": 0.8479790106087683, "eval_safe_recall": 0.8364025053444849, "eval_samples_per_second": 301.645, "eval_steps_per_second": 37.709, "eval_unsafe_aucpr": 0.9536892539716878, "eval_unsafe_f1": 0.8757173014598757, "eval_unsafe_fpr": 0.16359749465551454, "eval_unsafe_precision": 0.8710039923111046, "eval_unsafe_recall": 0.880481898896894, "step": 15219 }, { "epoch": 0.9002188442656888, "grad_norm": 3.3620963096618652, "learning_rate": 4.923867063603936e-06, "loss": 0.3316, "step": 15220 }, { "epoch": 0.9008103152540368, "grad_norm": 3.323751449584961, "learning_rate": 4.922396798799222e-06, "loss": 0.2835, "step": 15230 }, { "epoch": 0.9014017862423848, "grad_norm": 2.7786684036254883, "learning_rate": 4.92092575014947e-06, "loss": 0.3735, "step": 15240 }, { "epoch": 0.9019932572307329, "grad_norm": 5.640398025512695, "learning_rate": 4.9194539182544885e-06, "loss": 0.237, "step": 15250 }, { "epoch": 0.9025847282190809, "grad_norm": 5.0095319747924805, "learning_rate": 4.917981303714409e-06, "loss": 0.3396, "step": 15260 }, { "epoch": 0.9031761992074289, "grad_norm": 3.3858425617218018, "learning_rate": 4.91650790712968e-06, "loss": 0.3482, "step": 15270 }, { "epoch": 0.9037676701957769, "grad_norm": 1.9781662225723267, "learning_rate": 4.915033729101071e-06, "loss": 0.3381, "step": 15280 }, { "epoch": 0.904359141184125, "grad_norm": 3.1703758239746094, "learning_rate": 4.913558770229666e-06, "loss": 0.3625, "step": 15290 }, { "epoch": 0.904950612172473, "grad_norm": 3.432389974594116, "learning_rate": 4.912083031116872e-06, "loss": 0.3304, "step": 15300 }, { "epoch": 0.905542083160821, "grad_norm": 2.134312152862549, "learning_rate": 4.910606512364412e-06, "loss": 0.268, "step": 15310 }, { "epoch": 0.906133554149169, "grad_norm": 2.346173048019409, "learning_rate": 4.909129214574326e-06, "loss": 0.3417, "step": 15320 }, { "epoch": 0.906725025137517, "grad_norm": 2.6945130825042725, "learning_rate": 4.907651138348974e-06, "loss": 0.3415, "step": 15330 }, { "epoch": 0.9073164961258651, "grad_norm": 2.0993399620056152, "learning_rate": 4.9061722842910324e-06, "loss": 0.3298, "step": 15340 }, { "epoch": 0.9079079671142131, "grad_norm": 3.187474489212036, "learning_rate": 4.904692653003494e-06, "loss": 0.3202, "step": 15350 }, { "epoch": 0.9084994381025611, "grad_norm": 3.4601986408233643, "learning_rate": 4.90321224508967e-06, "loss": 0.3504, "step": 15360 }, { "epoch": 0.9090909090909091, "grad_norm": 4.12734317779541, "learning_rate": 4.901731061153184e-06, "loss": 0.3347, "step": 15370 }, { "epoch": 0.9096823800792571, "grad_norm": 3.2159409523010254, "learning_rate": 4.900249101797983e-06, "loss": 0.3273, "step": 15380 }, { "epoch": 0.9102738510676052, "grad_norm": 2.612872362136841, "learning_rate": 4.898766367628325e-06, "loss": 0.3225, "step": 15390 }, { "epoch": 0.9108653220559532, "grad_norm": 4.750239849090576, "learning_rate": 4.897282859248785e-06, "loss": 0.3156, "step": 15400 }, { "epoch": 0.9114567930443012, "grad_norm": 1.738272786140442, "learning_rate": 4.895798577264255e-06, "loss": 0.3162, "step": 15410 }, { "epoch": 0.9120482640326492, "grad_norm": 1.4551767110824585, "learning_rate": 4.894313522279941e-06, "loss": 0.2662, "step": 15420 }, { "epoch": 0.9126397350209973, "grad_norm": 2.128450393676758, "learning_rate": 4.892827694901364e-06, "loss": 0.2693, "step": 15430 }, { "epoch": 0.9132312060093453, "grad_norm": 3.3926517963409424, "learning_rate": 4.891341095734361e-06, "loss": 0.3103, "step": 15440 }, { "epoch": 0.9138226769976933, "grad_norm": 3.775010585784912, "learning_rate": 4.889853725385085e-06, "loss": 0.2869, "step": 15450 }, { "epoch": 0.9144141479860413, "grad_norm": 3.6521806716918945, "learning_rate": 4.88836558446e-06, "loss": 0.3886, "step": 15460 }, { "epoch": 0.9150056189743893, "grad_norm": 1.9084830284118652, "learning_rate": 4.886876673565887e-06, "loss": 0.2989, "step": 15470 }, { "epoch": 0.9155970899627374, "grad_norm": 1.9643568992614746, "learning_rate": 4.885386993309839e-06, "loss": 0.3224, "step": 15480 }, { "epoch": 0.9161885609510854, "grad_norm": 3.025437593460083, "learning_rate": 4.883896544299264e-06, "loss": 0.3366, "step": 15490 }, { "epoch": 0.9167800319394334, "grad_norm": 2.8364274501800537, "learning_rate": 4.8824053271418835e-06, "loss": 0.3446, "step": 15500 }, { "epoch": 0.9173715029277814, "grad_norm": 2.4385151863098145, "learning_rate": 4.880913342445731e-06, "loss": 0.2846, "step": 15510 }, { "epoch": 0.9179629739161295, "grad_norm": 2.4848251342773438, "learning_rate": 4.879420590819154e-06, "loss": 0.325, "step": 15520 }, { "epoch": 0.9185544449044775, "grad_norm": 4.780979156494141, "learning_rate": 4.877927072870814e-06, "loss": 0.3839, "step": 15530 }, { "epoch": 0.9191459158928255, "grad_norm": 3.2323784828186035, "learning_rate": 4.87643278920968e-06, "loss": 0.3465, "step": 15540 }, { "epoch": 0.9197373868811735, "grad_norm": 3.8760085105895996, "learning_rate": 4.874937740445039e-06, "loss": 0.2887, "step": 15550 }, { "epoch": 0.9203288578695215, "grad_norm": 6.14196252822876, "learning_rate": 4.873441927186489e-06, "loss": 0.2858, "step": 15560 }, { "epoch": 0.9209203288578696, "grad_norm": 3.8795981407165527, "learning_rate": 4.871945350043934e-06, "loss": 0.3406, "step": 15570 }, { "epoch": 0.9215117998462176, "grad_norm": 3.5772979259490967, "learning_rate": 4.870448009627599e-06, "loss": 0.3052, "step": 15580 }, { "epoch": 0.9221032708345656, "grad_norm": 1.813433051109314, "learning_rate": 4.868949906548011e-06, "loss": 0.3363, "step": 15590 }, { "epoch": 0.9226947418229136, "grad_norm": 4.101635932922363, "learning_rate": 4.8674510414160145e-06, "loss": 0.3165, "step": 15600 }, { "epoch": 0.9232862128112616, "grad_norm": 4.070963382720947, "learning_rate": 4.8659514148427595e-06, "loss": 0.3438, "step": 15610 }, { "epoch": 0.9238776837996097, "grad_norm": 2.753176212310791, "learning_rate": 4.864451027439713e-06, "loss": 0.3396, "step": 15620 }, { "epoch": 0.9244691547879577, "grad_norm": 2.5864503383636475, "learning_rate": 4.862949879818643e-06, "loss": 0.2683, "step": 15630 }, { "epoch": 0.9250606257763057, "grad_norm": 3.027881145477295, "learning_rate": 4.861447972591638e-06, "loss": 0.3465, "step": 15640 }, { "epoch": 0.9256520967646537, "grad_norm": 2.8669588565826416, "learning_rate": 4.859945306371089e-06, "loss": 0.3587, "step": 15650 }, { "epoch": 0.9262435677530018, "grad_norm": 3.2465553283691406, "learning_rate": 4.8584418817696985e-06, "loss": 0.3325, "step": 15660 }, { "epoch": 0.9268350387413498, "grad_norm": 4.2276716232299805, "learning_rate": 4.8569376994004785e-06, "loss": 0.3936, "step": 15670 }, { "epoch": 0.9274265097296978, "grad_norm": 1.9386107921600342, "learning_rate": 4.8554327598767495e-06, "loss": 0.3225, "step": 15680 }, { "epoch": 0.9280179807180458, "grad_norm": 3.2198352813720703, "learning_rate": 4.853927063812141e-06, "loss": 0.274, "step": 15690 }, { "epoch": 0.9286094517063938, "grad_norm": 5.719334125518799, "learning_rate": 4.852420611820593e-06, "loss": 0.2945, "step": 15700 }, { "epoch": 0.9292009226947419, "grad_norm": 3.2260587215423584, "learning_rate": 4.850913404516348e-06, "loss": 0.3185, "step": 15710 }, { "epoch": 0.9297923936830899, "grad_norm": 3.8232107162475586, "learning_rate": 4.849405442513962e-06, "loss": 0.3651, "step": 15720 }, { "epoch": 0.9303838646714379, "grad_norm": 2.7706685066223145, "learning_rate": 4.847896726428298e-06, "loss": 0.3256, "step": 15730 }, { "epoch": 0.9309753356597859, "grad_norm": 2.5927600860595703, "learning_rate": 4.8463872568745245e-06, "loss": 0.3259, "step": 15740 }, { "epoch": 0.931566806648134, "grad_norm": 3.898158550262451, "learning_rate": 4.844877034468117e-06, "loss": 0.3553, "step": 15750 }, { "epoch": 0.932158277636482, "grad_norm": 3.71560001373291, "learning_rate": 4.843366059824861e-06, "loss": 0.2955, "step": 15760 }, { "epoch": 0.93274974862483, "grad_norm": 2.5170793533325195, "learning_rate": 4.8418543335608455e-06, "loss": 0.2983, "step": 15770 }, { "epoch": 0.933341219613178, "grad_norm": 4.619828224182129, "learning_rate": 4.840341856292467e-06, "loss": 0.3922, "step": 15780 }, { "epoch": 0.933932690601526, "grad_norm": 3.7232306003570557, "learning_rate": 4.8388286286364295e-06, "loss": 0.3489, "step": 15790 }, { "epoch": 0.9345241615898741, "grad_norm": 2.7852883338928223, "learning_rate": 4.837314651209741e-06, "loss": 0.2912, "step": 15800 }, { "epoch": 0.9351156325782221, "grad_norm": 2.6773109436035156, "learning_rate": 4.835799924629717e-06, "loss": 0.2515, "step": 15810 }, { "epoch": 0.93570710356657, "grad_norm": 2.7732625007629395, "learning_rate": 4.834284449513977e-06, "loss": 0.3826, "step": 15820 }, { "epoch": 0.936298574554918, "grad_norm": 3.0243752002716064, "learning_rate": 4.832768226480447e-06, "loss": 0.3159, "step": 15830 }, { "epoch": 0.936890045543266, "grad_norm": 4.05096435546875, "learning_rate": 4.831251256147357e-06, "loss": 0.3171, "step": 15840 }, { "epoch": 0.9374815165316142, "grad_norm": 4.104322910308838, "learning_rate": 4.82973353913324e-06, "loss": 0.3386, "step": 15850 }, { "epoch": 0.9380729875199622, "grad_norm": 2.678168535232544, "learning_rate": 4.828215076056938e-06, "loss": 0.3124, "step": 15860 }, { "epoch": 0.9386644585083102, "grad_norm": 4.327289581298828, "learning_rate": 4.826695867537593e-06, "loss": 0.2917, "step": 15870 }, { "epoch": 0.9392559294966581, "grad_norm": 2.44892954826355, "learning_rate": 4.8251759141946554e-06, "loss": 0.326, "step": 15880 }, { "epoch": 0.9398474004850063, "grad_norm": 4.20181131362915, "learning_rate": 4.823655216647873e-06, "loss": 0.31, "step": 15890 }, { "epoch": 0.9404388714733543, "grad_norm": 3.7402830123901367, "learning_rate": 4.822133775517303e-06, "loss": 0.345, "step": 15900 }, { "epoch": 0.9410303424617023, "grad_norm": 2.843750238418579, "learning_rate": 4.820611591423302e-06, "loss": 0.3843, "step": 15910 }, { "epoch": 0.9416218134500502, "grad_norm": 2.598572015762329, "learning_rate": 4.819088664986531e-06, "loss": 0.3186, "step": 15920 }, { "epoch": 0.9422132844383982, "grad_norm": 2.0697689056396484, "learning_rate": 4.817564996827954e-06, "loss": 0.2876, "step": 15930 }, { "epoch": 0.9428047554267464, "grad_norm": 2.580213785171509, "learning_rate": 4.816040587568836e-06, "loss": 0.2892, "step": 15940 }, { "epoch": 0.9433962264150944, "grad_norm": 2.782437801361084, "learning_rate": 4.814515437830747e-06, "loss": 0.3547, "step": 15950 }, { "epoch": 0.9439876974034423, "grad_norm": 4.653677940368652, "learning_rate": 4.812989548235556e-06, "loss": 0.2973, "step": 15960 }, { "epoch": 0.9445791683917903, "grad_norm": 2.9328560829162598, "learning_rate": 4.811462919405434e-06, "loss": 0.3596, "step": 15970 }, { "epoch": 0.9451706393801385, "grad_norm": 2.5789759159088135, "learning_rate": 4.809935551962855e-06, "loss": 0.3622, "step": 15980 }, { "epoch": 0.9457621103684865, "grad_norm": 2.378706216812134, "learning_rate": 4.808407446530593e-06, "loss": 0.2893, "step": 15990 }, { "epoch": 0.9463535813568345, "grad_norm": 4.059628009796143, "learning_rate": 4.806878603731724e-06, "loss": 0.3411, "step": 16000 }, { "epoch": 0.9469450523451824, "grad_norm": 2.968478202819824, "learning_rate": 4.805349024189623e-06, "loss": 0.2636, "step": 16010 }, { "epoch": 0.9475365233335304, "grad_norm": 3.127575159072876, "learning_rate": 4.803818708527967e-06, "loss": 0.2751, "step": 16020 }, { "epoch": 0.9481279943218786, "grad_norm": 1.8899832963943481, "learning_rate": 4.802287657370732e-06, "loss": 0.3085, "step": 16030 }, { "epoch": 0.9487194653102266, "grad_norm": 3.1122567653656006, "learning_rate": 4.800755871342195e-06, "loss": 0.3078, "step": 16040 }, { "epoch": 0.9493109362985745, "grad_norm": 3.151719093322754, "learning_rate": 4.79922335106693e-06, "loss": 0.3537, "step": 16050 }, { "epoch": 0.9499024072869225, "grad_norm": 1.919867753982544, "learning_rate": 4.797690097169816e-06, "loss": 0.3363, "step": 16060 }, { "epoch": 0.9504938782752705, "grad_norm": 2.507673740386963, "learning_rate": 4.796156110276024e-06, "loss": 0.3134, "step": 16070 }, { "epoch": 0.9510853492636187, "grad_norm": 2.0464775562286377, "learning_rate": 4.794621391011029e-06, "loss": 0.3042, "step": 16080 }, { "epoch": 0.9516768202519666, "grad_norm": 2.363704204559326, "learning_rate": 4.793085940000602e-06, "loss": 0.2902, "step": 16090 }, { "epoch": 0.9522682912403146, "grad_norm": 3.7071692943573, "learning_rate": 4.791549757870815e-06, "loss": 0.3366, "step": 16100 }, { "epoch": 0.9528597622286626, "grad_norm": 2.817458391189575, "learning_rate": 4.790012845248036e-06, "loss": 0.298, "step": 16110 }, { "epoch": 0.9534512332170108, "grad_norm": 2.9618260860443115, "learning_rate": 4.78847520275893e-06, "loss": 0.3116, "step": 16120 }, { "epoch": 0.9540427042053587, "grad_norm": 2.6378321647644043, "learning_rate": 4.786936831030461e-06, "loss": 0.3437, "step": 16130 }, { "epoch": 0.9546341751937067, "grad_norm": 3.2677690982818604, "learning_rate": 4.785397730689893e-06, "loss": 0.3139, "step": 16140 }, { "epoch": 0.9552256461820547, "grad_norm": 3.645596981048584, "learning_rate": 4.7838579023647815e-06, "loss": 0.3235, "step": 16150 }, { "epoch": 0.9558171171704027, "grad_norm": 2.7554931640625, "learning_rate": 4.782317346682983e-06, "loss": 0.3426, "step": 16160 }, { "epoch": 0.9564085881587508, "grad_norm": 2.507883310317993, "learning_rate": 4.7807760642726485e-06, "loss": 0.3386, "step": 16170 }, { "epoch": 0.9570000591470988, "grad_norm": 4.208895206451416, "learning_rate": 4.779234055762227e-06, "loss": 0.3509, "step": 16180 }, { "epoch": 0.9575915301354468, "grad_norm": 2.0495245456695557, "learning_rate": 4.7776913217804614e-06, "loss": 0.3243, "step": 16190 }, { "epoch": 0.9581830011237948, "grad_norm": 3.4831180572509766, "learning_rate": 4.776147862956393e-06, "loss": 0.3064, "step": 16200 }, { "epoch": 0.958774472112143, "grad_norm": 3.307873249053955, "learning_rate": 4.7746036799193555e-06, "loss": 0.2966, "step": 16210 }, { "epoch": 0.959365943100491, "grad_norm": 3.0764095783233643, "learning_rate": 4.773058773298982e-06, "loss": 0.3182, "step": 16220 }, { "epoch": 0.959957414088839, "grad_norm": 4.538529396057129, "learning_rate": 4.771513143725196e-06, "loss": 0.3268, "step": 16230 }, { "epoch": 0.9605488850771869, "grad_norm": 2.6974246501922607, "learning_rate": 4.769966791828219e-06, "loss": 0.2708, "step": 16240 }, { "epoch": 0.9611403560655349, "grad_norm": 2.4175212383270264, "learning_rate": 4.768419718238568e-06, "loss": 0.3119, "step": 16250 }, { "epoch": 0.961731827053883, "grad_norm": 4.780592918395996, "learning_rate": 4.7668719235870505e-06, "loss": 0.3211, "step": 16260 }, { "epoch": 0.962323298042231, "grad_norm": 2.8652796745300293, "learning_rate": 4.765323408504769e-06, "loss": 0.2979, "step": 16270 }, { "epoch": 0.962914769030579, "grad_norm": 2.6609671115875244, "learning_rate": 4.763774173623123e-06, "loss": 0.3016, "step": 16280 }, { "epoch": 0.963506240018927, "grad_norm": 3.33339786529541, "learning_rate": 4.762224219573802e-06, "loss": 0.3549, "step": 16290 }, { "epoch": 0.964097711007275, "grad_norm": 5.644134521484375, "learning_rate": 4.7606735469887885e-06, "loss": 0.3468, "step": 16300 }, { "epoch": 0.9646891819956231, "grad_norm": 2.31178879737854, "learning_rate": 4.759122156500361e-06, "loss": 0.31, "step": 16310 }, { "epoch": 0.9652806529839711, "grad_norm": 2.5142459869384766, "learning_rate": 4.75757004874109e-06, "loss": 0.3221, "step": 16320 }, { "epoch": 0.9658721239723191, "grad_norm": 2.3319013118743896, "learning_rate": 4.756017224343835e-06, "loss": 0.271, "step": 16330 }, { "epoch": 0.9664635949606671, "grad_norm": 3.1083741188049316, "learning_rate": 4.754463683941753e-06, "loss": 0.2366, "step": 16340 }, { "epoch": 0.9670550659490152, "grad_norm": 3.5621519088745117, "learning_rate": 4.752909428168288e-06, "loss": 0.2785, "step": 16350 }, { "epoch": 0.9676465369373632, "grad_norm": 5.374129295349121, "learning_rate": 4.751354457657179e-06, "loss": 0.3138, "step": 16360 }, { "epoch": 0.9682380079257112, "grad_norm": 5.732593059539795, "learning_rate": 4.7497987730424545e-06, "loss": 0.4036, "step": 16370 }, { "epoch": 0.9688294789140592, "grad_norm": 1.5983829498291016, "learning_rate": 4.748242374958436e-06, "loss": 0.3109, "step": 16380 }, { "epoch": 0.9694209499024072, "grad_norm": 1.8374629020690918, "learning_rate": 4.746685264039733e-06, "loss": 0.3224, "step": 16390 }, { "epoch": 0.9700124208907553, "grad_norm": 3.793588161468506, "learning_rate": 4.74512744092125e-06, "loss": 0.334, "step": 16400 }, { "epoch": 0.9706038918791033, "grad_norm": 2.5068726539611816, "learning_rate": 4.743568906238178e-06, "loss": 0.2663, "step": 16410 }, { "epoch": 0.9711953628674513, "grad_norm": 2.800074338912964, "learning_rate": 4.742009660626001e-06, "loss": 0.3296, "step": 16420 }, { "epoch": 0.9717868338557993, "grad_norm": 4.174688339233398, "learning_rate": 4.740449704720489e-06, "loss": 0.3098, "step": 16430 }, { "epoch": 0.9723783048441474, "grad_norm": 3.072627305984497, "learning_rate": 4.738889039157707e-06, "loss": 0.347, "step": 16440 }, { "epoch": 0.9729697758324954, "grad_norm": 3.7549943923950195, "learning_rate": 4.737327664574004e-06, "loss": 0.3978, "step": 16450 }, { "epoch": 0.9735612468208434, "grad_norm": 3.065549850463867, "learning_rate": 4.735765581606022e-06, "loss": 0.3183, "step": 16460 }, { "epoch": 0.9741527178091914, "grad_norm": 3.1468026638031006, "learning_rate": 4.734202790890689e-06, "loss": 0.3353, "step": 16470 }, { "epoch": 0.9747441887975394, "grad_norm": 2.3557288646698, "learning_rate": 4.7326392930652244e-06, "loss": 0.3209, "step": 16480 }, { "epoch": 0.9753356597858875, "grad_norm": 2.538161277770996, "learning_rate": 4.731075088767133e-06, "loss": 0.3049, "step": 16490 }, { "epoch": 0.9759271307742355, "grad_norm": 3.096086263656616, "learning_rate": 4.72951017863421e-06, "loss": 0.2432, "step": 16500 }, { "epoch": 0.9765186017625835, "grad_norm": 4.013457298278809, "learning_rate": 4.727944563304538e-06, "loss": 0.3621, "step": 16510 }, { "epoch": 0.9771100727509315, "grad_norm": 4.206152439117432, "learning_rate": 4.726378243416485e-06, "loss": 0.3176, "step": 16520 }, { "epoch": 0.9777015437392796, "grad_norm": 1.9044580459594727, "learning_rate": 4.724811219608711e-06, "loss": 0.3459, "step": 16530 }, { "epoch": 0.9782930147276276, "grad_norm": 1.7666800022125244, "learning_rate": 4.723243492520157e-06, "loss": 0.2686, "step": 16540 }, { "epoch": 0.9788844857159756, "grad_norm": 9.541177749633789, "learning_rate": 4.721675062790055e-06, "loss": 0.2993, "step": 16550 }, { "epoch": 0.9794759567043236, "grad_norm": 2.843207359313965, "learning_rate": 4.720105931057922e-06, "loss": 0.2836, "step": 16560 }, { "epoch": 0.9800674276926716, "grad_norm": 2.5295791625976562, "learning_rate": 4.718536097963562e-06, "loss": 0.3126, "step": 16570 }, { "epoch": 0.9806588986810197, "grad_norm": 2.749816656112671, "learning_rate": 4.7169655641470635e-06, "loss": 0.3273, "step": 16580 }, { "epoch": 0.9812503696693677, "grad_norm": 3.200345754623413, "learning_rate": 4.715394330248802e-06, "loss": 0.3749, "step": 16590 }, { "epoch": 0.9818418406577157, "grad_norm": 3.692337989807129, "learning_rate": 4.713822396909438e-06, "loss": 0.3006, "step": 16600 }, { "epoch": 0.9824333116460637, "grad_norm": 2.195751667022705, "learning_rate": 4.7122497647699174e-06, "loss": 0.32, "step": 16610 }, { "epoch": 0.9830247826344117, "grad_norm": 2.5720691680908203, "learning_rate": 4.710676434471472e-06, "loss": 0.3856, "step": 16620 }, { "epoch": 0.9836162536227598, "grad_norm": 2.857969045639038, "learning_rate": 4.709102406655615e-06, "loss": 0.2899, "step": 16630 }, { "epoch": 0.9842077246111078, "grad_norm": 1.8571844100952148, "learning_rate": 4.707527681964148e-06, "loss": 0.2824, "step": 16640 }, { "epoch": 0.9847991955994558, "grad_norm": 6.957851886749268, "learning_rate": 4.705952261039154e-06, "loss": 0.3303, "step": 16650 }, { "epoch": 0.9853906665878038, "grad_norm": 2.395742893218994, "learning_rate": 4.7043761445230014e-06, "loss": 0.2831, "step": 16660 }, { "epoch": 0.9859821375761519, "grad_norm": 5.644003391265869, "learning_rate": 4.702799333058342e-06, "loss": 0.3639, "step": 16670 }, { "epoch": 0.9865736085644999, "grad_norm": 3.6825220584869385, "learning_rate": 4.7012218272881094e-06, "loss": 0.3134, "step": 16680 }, { "epoch": 0.9871650795528479, "grad_norm": 2.247535228729248, "learning_rate": 4.699643627855523e-06, "loss": 0.3298, "step": 16690 }, { "epoch": 0.9877565505411959, "grad_norm": 3.169652223587036, "learning_rate": 4.698064735404083e-06, "loss": 0.3056, "step": 16700 }, { "epoch": 0.9883480215295439, "grad_norm": 2.4838671684265137, "learning_rate": 4.696485150577571e-06, "loss": 0.3409, "step": 16710 }, { "epoch": 0.988939492517892, "grad_norm": 4.018945217132568, "learning_rate": 4.694904874020055e-06, "loss": 0.3237, "step": 16720 }, { "epoch": 0.98953096350624, "grad_norm": 2.575408935546875, "learning_rate": 4.693323906375881e-06, "loss": 0.2635, "step": 16730 }, { "epoch": 0.990122434494588, "grad_norm": 2.280029535293579, "learning_rate": 4.6917422482896805e-06, "loss": 0.2724, "step": 16740 }, { "epoch": 0.990713905482936, "grad_norm": 2.656015634536743, "learning_rate": 4.690159900406363e-06, "loss": 0.2931, "step": 16750 }, { "epoch": 0.9913053764712841, "grad_norm": 2.2961363792419434, "learning_rate": 4.688576863371121e-06, "loss": 0.3148, "step": 16760 }, { "epoch": 0.9918968474596321, "grad_norm": 2.4203603267669678, "learning_rate": 4.686993137829427e-06, "loss": 0.3371, "step": 16770 }, { "epoch": 0.9924883184479801, "grad_norm": 2.4906296730041504, "learning_rate": 4.685408724427037e-06, "loss": 0.3465, "step": 16780 }, { "epoch": 0.9930797894363281, "grad_norm": 3.2459092140197754, "learning_rate": 4.683823623809984e-06, "loss": 0.3226, "step": 16790 }, { "epoch": 0.9936712604246761, "grad_norm": 4.291164875030518, "learning_rate": 4.682237836624583e-06, "loss": 0.2885, "step": 16800 }, { "epoch": 0.9942627314130242, "grad_norm": 2.986771821975708, "learning_rate": 4.680651363517428e-06, "loss": 0.3368, "step": 16810 }, { "epoch": 0.9948542024013722, "grad_norm": 1.5785176753997803, "learning_rate": 4.679064205135394e-06, "loss": 0.3343, "step": 16820 }, { "epoch": 0.9954456733897202, "grad_norm": 2.4919722080230713, "learning_rate": 4.6774763621256354e-06, "loss": 0.2751, "step": 16830 }, { "epoch": 0.9960371443780682, "grad_norm": 3.1233842372894287, "learning_rate": 4.675887835135583e-06, "loss": 0.3242, "step": 16840 }, { "epoch": 0.9966286153664162, "grad_norm": 8.381209373474121, "learning_rate": 4.67429862481295e-06, "loss": 0.3026, "step": 16850 }, { "epoch": 0.9972200863547643, "grad_norm": 3.1535539627075195, "learning_rate": 4.672708731805728e-06, "loss": 0.2941, "step": 16860 }, { "epoch": 0.9978115573431123, "grad_norm": 2.2888896465301514, "learning_rate": 4.671118156762183e-06, "loss": 0.3356, "step": 16870 }, { "epoch": 0.9984030283314603, "grad_norm": 2.4113214015960693, "learning_rate": 4.6695269003308635e-06, "loss": 0.2957, "step": 16880 }, { "epoch": 0.9989944993198083, "grad_norm": 4.239777088165283, "learning_rate": 4.667934963160595e-06, "loss": 0.3136, "step": 16890 }, { "epoch": 0.9995859703081564, "grad_norm": 3.335905075073242, "learning_rate": 4.666342345900477e-06, "loss": 0.348, "step": 16900 }, { "epoch": 1.0001774412965043, "grad_norm": 2.2869250774383545, "learning_rate": 4.664749049199891e-06, "loss": 0.3823, "step": 16910 }, { "epoch": 1.0007689122848524, "grad_norm": 2.0591390132904053, "learning_rate": 4.663155073708493e-06, "loss": 0.2591, "step": 16920 }, { "epoch": 1.0013603832732005, "grad_norm": 2.2811458110809326, "learning_rate": 4.661560420076216e-06, "loss": 0.2765, "step": 16930 }, { "epoch": 1.0019518542615484, "grad_norm": 4.233405113220215, "learning_rate": 4.659965088953272e-06, "loss": 0.2805, "step": 16940 }, { "epoch": 1.0025433252498965, "grad_norm": 2.654818534851074, "learning_rate": 4.658369080990143e-06, "loss": 0.2474, "step": 16950 }, { "epoch": 1.0031347962382444, "grad_norm": 2.7204737663269043, "learning_rate": 4.6567723968375964e-06, "loss": 0.2253, "step": 16960 }, { "epoch": 1.0037262672265925, "grad_norm": 5.10129976272583, "learning_rate": 4.655175037146665e-06, "loss": 0.3348, "step": 16970 }, { "epoch": 1.0043177382149406, "grad_norm": 3.2961466312408447, "learning_rate": 4.653577002568665e-06, "loss": 0.2813, "step": 16980 }, { "epoch": 1.0049092092032885, "grad_norm": 2.9351718425750732, "learning_rate": 4.651978293755184e-06, "loss": 0.3139, "step": 16990 }, { "epoch": 1.0055006801916366, "grad_norm": 3.013028144836426, "learning_rate": 4.6503789113580855e-06, "loss": 0.3449, "step": 17000 }, { "epoch": 1.0060921511799845, "grad_norm": 3.264341354370117, "learning_rate": 4.648778856029506e-06, "loss": 0.3337, "step": 17010 }, { "epoch": 1.0066836221683326, "grad_norm": 2.4180314540863037, "learning_rate": 4.647178128421861e-06, "loss": 0.2361, "step": 17020 }, { "epoch": 1.0072750931566807, "grad_norm": 3.6979732513427734, "learning_rate": 4.645576729187834e-06, "loss": 0.3241, "step": 17030 }, { "epoch": 1.0078665641450286, "grad_norm": 1.6972005367279053, "learning_rate": 4.643974658980386e-06, "loss": 0.279, "step": 17040 }, { "epoch": 1.0084580351333767, "grad_norm": 4.8941569328308105, "learning_rate": 4.642371918452752e-06, "loss": 0.3182, "step": 17050 }, { "epoch": 1.0090495061217248, "grad_norm": 4.517833232879639, "learning_rate": 4.6407685082584384e-06, "loss": 0.2888, "step": 17060 }, { "epoch": 1.0096409771100727, "grad_norm": 2.2167015075683594, "learning_rate": 4.639164429051225e-06, "loss": 0.2493, "step": 17070 }, { "epoch": 1.0102324480984208, "grad_norm": 6.6228532791137695, "learning_rate": 4.637559681485166e-06, "loss": 0.3062, "step": 17080 }, { "epoch": 1.0108239190867687, "grad_norm": 3.6505212783813477, "learning_rate": 4.6359542662145866e-06, "loss": 0.2476, "step": 17090 }, { "epoch": 1.0114153900751168, "grad_norm": 2.938830852508545, "learning_rate": 4.634348183894085e-06, "loss": 0.3504, "step": 17100 }, { "epoch": 1.012006861063465, "grad_norm": 2.9572458267211914, "learning_rate": 4.63274143517853e-06, "loss": 0.2793, "step": 17110 }, { "epoch": 1.0125983320518128, "grad_norm": 3.1707093715667725, "learning_rate": 4.6311340207230635e-06, "loss": 0.2504, "step": 17120 }, { "epoch": 1.013189803040161, "grad_norm": 1.774966835975647, "learning_rate": 4.629525941183099e-06, "loss": 0.3248, "step": 17130 }, { "epoch": 1.0137812740285088, "grad_norm": 3.367788076400757, "learning_rate": 4.62791719721432e-06, "loss": 0.3165, "step": 17140 }, { "epoch": 1.014372745016857, "grad_norm": 3.095541477203369, "learning_rate": 4.626307789472683e-06, "loss": 0.2952, "step": 17150 }, { "epoch": 1.014964216005205, "grad_norm": 3.513197183609009, "learning_rate": 4.624697718614413e-06, "loss": 0.3257, "step": 17160 }, { "epoch": 1.015555686993553, "grad_norm": 2.1466736793518066, "learning_rate": 4.6230869852960045e-06, "loss": 0.2936, "step": 17170 }, { "epoch": 1.016147157981901, "grad_norm": 2.309603452682495, "learning_rate": 4.621475590174226e-06, "loss": 0.3197, "step": 17180 }, { "epoch": 1.016738628970249, "grad_norm": 3.160773515701294, "learning_rate": 4.619863533906112e-06, "loss": 0.2726, "step": 17190 }, { "epoch": 1.017330099958597, "grad_norm": 2.5248355865478516, "learning_rate": 4.6182508171489695e-06, "loss": 0.251, "step": 17200 }, { "epoch": 1.0179215709469451, "grad_norm": 7.549440860748291, "learning_rate": 4.616637440560373e-06, "loss": 0.2956, "step": 17210 }, { "epoch": 1.018513041935293, "grad_norm": 3.602994203567505, "learning_rate": 4.6150234047981656e-06, "loss": 0.2563, "step": 17220 }, { "epoch": 1.0191045129236411, "grad_norm": 1.4920613765716553, "learning_rate": 4.613408710520461e-06, "loss": 0.2682, "step": 17230 }, { "epoch": 1.019695983911989, "grad_norm": 3.8239567279815674, "learning_rate": 4.611793358385642e-06, "loss": 0.3039, "step": 17240 }, { "epoch": 1.0202874549003371, "grad_norm": 3.6404476165771484, "learning_rate": 4.6101773490523545e-06, "loss": 0.3434, "step": 17250 }, { "epoch": 1.0208789258886852, "grad_norm": 3.293882131576538, "learning_rate": 4.6085606831795185e-06, "loss": 0.2897, "step": 17260 }, { "epoch": 1.0214703968770331, "grad_norm": 3.933575391769409, "learning_rate": 4.606943361426318e-06, "loss": 0.3428, "step": 17270 }, { "epoch": 1.0220618678653812, "grad_norm": 2.175508975982666, "learning_rate": 4.605325384452207e-06, "loss": 0.2556, "step": 17280 }, { "epoch": 1.0226533388537293, "grad_norm": 2.756922483444214, "learning_rate": 4.603706752916904e-06, "loss": 0.2706, "step": 17290 }, { "epoch": 1.0232448098420772, "grad_norm": 2.665309429168701, "learning_rate": 4.602087467480398e-06, "loss": 0.2783, "step": 17300 }, { "epoch": 1.0238362808304253, "grad_norm": 5.930862903594971, "learning_rate": 4.60046752880294e-06, "loss": 0.363, "step": 17310 }, { "epoch": 1.0244277518187732, "grad_norm": 2.9444563388824463, "learning_rate": 4.59884693754505e-06, "loss": 0.2825, "step": 17320 }, { "epoch": 1.0250192228071213, "grad_norm": 3.9739718437194824, "learning_rate": 4.597225694367514e-06, "loss": 0.3036, "step": 17330 }, { "epoch": 1.0256106937954694, "grad_norm": 4.221290588378906, "learning_rate": 4.5956037999313835e-06, "loss": 0.3012, "step": 17340 }, { "epoch": 1.0262021647838173, "grad_norm": 5.280589580535889, "learning_rate": 4.5939812548979756e-06, "loss": 0.3062, "step": 17350 }, { "epoch": 1.0267936357721654, "grad_norm": 3.4688491821289062, "learning_rate": 4.592358059928872e-06, "loss": 0.3216, "step": 17360 }, { "epoch": 1.0273851067605133, "grad_norm": 3.713205337524414, "learning_rate": 4.590734215685923e-06, "loss": 0.2943, "step": 17370 }, { "epoch": 1.0279765777488614, "grad_norm": 3.937586784362793, "learning_rate": 4.589109722831238e-06, "loss": 0.3558, "step": 17380 }, { "epoch": 1.0285680487372095, "grad_norm": 2.5275957584381104, "learning_rate": 4.587484582027194e-06, "loss": 0.2618, "step": 17390 }, { "epoch": 1.0291595197255574, "grad_norm": 3.431504487991333, "learning_rate": 4.585858793936432e-06, "loss": 0.3142, "step": 17400 }, { "epoch": 1.0297509907139055, "grad_norm": 3.811033010482788, "learning_rate": 4.584232359221857e-06, "loss": 0.2827, "step": 17410 }, { "epoch": 1.0303424617022534, "grad_norm": 6.508264541625977, "learning_rate": 4.582605278546638e-06, "loss": 0.2723, "step": 17420 }, { "epoch": 1.0309339326906015, "grad_norm": 2.9867000579833984, "learning_rate": 4.580977552574204e-06, "loss": 0.3007, "step": 17430 }, { "epoch": 1.0315254036789496, "grad_norm": 3.9488792419433594, "learning_rate": 4.579349181968253e-06, "loss": 0.2673, "step": 17440 }, { "epoch": 1.0321168746672975, "grad_norm": 2.495467185974121, "learning_rate": 4.5777201673927415e-06, "loss": 0.2665, "step": 17450 }, { "epoch": 1.0327083456556456, "grad_norm": 4.302645206451416, "learning_rate": 4.576090509511889e-06, "loss": 0.2848, "step": 17460 }, { "epoch": 1.0332998166439935, "grad_norm": 2.913517475128174, "learning_rate": 4.57446020899018e-06, "loss": 0.3531, "step": 17470 }, { "epoch": 1.0338912876323416, "grad_norm": 2.067399501800537, "learning_rate": 4.572829266492358e-06, "loss": 0.2721, "step": 17480 }, { "epoch": 1.0344827586206897, "grad_norm": 3.9263219833374023, "learning_rate": 4.5711976826834305e-06, "loss": 0.3346, "step": 17490 }, { "epoch": 1.0350742296090376, "grad_norm": 3.0830204486846924, "learning_rate": 4.5695654582286634e-06, "loss": 0.2205, "step": 17500 }, { "epoch": 1.0356657005973857, "grad_norm": 5.203652858734131, "learning_rate": 4.567932593793587e-06, "loss": 0.2698, "step": 17510 }, { "epoch": 1.0362571715857338, "grad_norm": 3.967395067214966, "learning_rate": 4.566299090043991e-06, "loss": 0.2895, "step": 17520 }, { "epoch": 1.0368486425740817, "grad_norm": 4.103964805603027, "learning_rate": 4.564664947645927e-06, "loss": 0.3678, "step": 17530 }, { "epoch": 1.0374401135624298, "grad_norm": 5.013073444366455, "learning_rate": 4.563030167265705e-06, "loss": 0.4063, "step": 17540 }, { "epoch": 1.0380315845507777, "grad_norm": 2.639113664627075, "learning_rate": 4.561394749569897e-06, "loss": 0.245, "step": 17550 }, { "epoch": 1.0386230555391258, "grad_norm": 3.7462215423583984, "learning_rate": 4.559758695225334e-06, "loss": 0.3332, "step": 17560 }, { "epoch": 1.039214526527474, "grad_norm": 2.4227960109710693, "learning_rate": 4.558122004899106e-06, "loss": 0.3727, "step": 17570 }, { "epoch": 1.0398059975158218, "grad_norm": 2.1172192096710205, "learning_rate": 4.556484679258565e-06, "loss": 0.2714, "step": 17580 }, { "epoch": 1.04039746850417, "grad_norm": 5.363768100738525, "learning_rate": 4.5548467189713175e-06, "loss": 0.3335, "step": 17590 }, { "epoch": 1.0409889394925178, "grad_norm": 4.175081253051758, "learning_rate": 4.553208124705233e-06, "loss": 0.2643, "step": 17600 }, { "epoch": 1.041580410480866, "grad_norm": 4.635672092437744, "learning_rate": 4.551568897128437e-06, "loss": 0.3532, "step": 17610 }, { "epoch": 1.042171881469214, "grad_norm": 3.34712553024292, "learning_rate": 4.549929036909316e-06, "loss": 0.3304, "step": 17620 }, { "epoch": 1.042763352457562, "grad_norm": 2.4211509227752686, "learning_rate": 4.548288544716511e-06, "loss": 0.2721, "step": 17630 }, { "epoch": 1.04335482344591, "grad_norm": 3.171177387237549, "learning_rate": 4.546647421218922e-06, "loss": 0.2806, "step": 17640 }, { "epoch": 1.043946294434258, "grad_norm": 3.4869742393493652, "learning_rate": 4.545005667085706e-06, "loss": 0.2734, "step": 17650 }, { "epoch": 1.044537765422606, "grad_norm": 3.010958433151245, "learning_rate": 4.543363282986281e-06, "loss": 0.2812, "step": 17660 }, { "epoch": 1.0451292364109541, "grad_norm": 3.6773736476898193, "learning_rate": 4.541720269590314e-06, "loss": 0.3324, "step": 17670 }, { "epoch": 1.045720707399302, "grad_norm": 1.7617377042770386, "learning_rate": 4.540076627567737e-06, "loss": 0.2736, "step": 17680 }, { "epoch": 1.0463121783876501, "grad_norm": 3.8207931518554688, "learning_rate": 4.5384323575887345e-06, "loss": 0.2507, "step": 17690 }, { "epoch": 1.046903649375998, "grad_norm": 5.009983062744141, "learning_rate": 4.536787460323744e-06, "loss": 0.278, "step": 17700 }, { "epoch": 1.047495120364346, "grad_norm": 3.22523832321167, "learning_rate": 4.535141936443465e-06, "loss": 0.3308, "step": 17710 }, { "epoch": 1.0480865913526942, "grad_norm": 3.6126708984375, "learning_rate": 4.533495786618847e-06, "loss": 0.269, "step": 17720 }, { "epoch": 1.048678062341042, "grad_norm": 5.55758810043335, "learning_rate": 4.5318490115210985e-06, "loss": 0.3065, "step": 17730 }, { "epoch": 1.0492695333293902, "grad_norm": 2.7169156074523926, "learning_rate": 4.530201611821682e-06, "loss": 0.2843, "step": 17740 }, { "epoch": 1.0498610043177383, "grad_norm": 3.066561222076416, "learning_rate": 4.528553588192313e-06, "loss": 0.2652, "step": 17750 }, { "epoch": 1.0504524753060862, "grad_norm": 2.0885939598083496, "learning_rate": 4.526904941304965e-06, "loss": 0.2398, "step": 17760 }, { "epoch": 1.0510439462944343, "grad_norm": 5.098671913146973, "learning_rate": 4.525255671831859e-06, "loss": 0.3019, "step": 17770 }, { "epoch": 1.0516354172827822, "grad_norm": 4.448405742645264, "learning_rate": 4.523605780445476e-06, "loss": 0.3202, "step": 17780 }, { "epoch": 1.0522268882711303, "grad_norm": 2.1770710945129395, "learning_rate": 4.52195526781855e-06, "loss": 0.3163, "step": 17790 }, { "epoch": 1.0528183592594784, "grad_norm": 2.172612190246582, "learning_rate": 4.520304134624067e-06, "loss": 0.2841, "step": 17800 }, { "epoch": 1.0534098302478263, "grad_norm": 2.704414129257202, "learning_rate": 4.518652381535265e-06, "loss": 0.2935, "step": 17810 }, { "epoch": 1.0540013012361744, "grad_norm": 3.4816391468048096, "learning_rate": 4.517000009225635e-06, "loss": 0.2996, "step": 17820 }, { "epoch": 1.0545927722245223, "grad_norm": 2.916141986846924, "learning_rate": 4.5153470183689225e-06, "loss": 0.2401, "step": 17830 }, { "epoch": 1.0551842432128704, "grad_norm": 2.4444034099578857, "learning_rate": 4.513693409639125e-06, "loss": 0.2807, "step": 17840 }, { "epoch": 1.0557757142012185, "grad_norm": 2.135671377182007, "learning_rate": 4.512039183710488e-06, "loss": 0.232, "step": 17850 }, { "epoch": 1.0563671851895664, "grad_norm": 5.2215962409973145, "learning_rate": 4.510384341257515e-06, "loss": 0.2728, "step": 17860 }, { "epoch": 1.0569586561779145, "grad_norm": 3.5986924171447754, "learning_rate": 4.508728882954955e-06, "loss": 0.2934, "step": 17870 }, { "epoch": 1.0575501271662624, "grad_norm": 5.2210540771484375, "learning_rate": 4.507072809477812e-06, "loss": 0.2757, "step": 17880 }, { "epoch": 1.0581415981546105, "grad_norm": 3.3177006244659424, "learning_rate": 4.505416121501337e-06, "loss": 0.3018, "step": 17890 }, { "epoch": 1.0587330691429586, "grad_norm": 3.1965601444244385, "learning_rate": 4.503758819701037e-06, "loss": 0.2789, "step": 17900 }, { "epoch": 1.0593245401313065, "grad_norm": 2.431068181991577, "learning_rate": 4.502100904752664e-06, "loss": 0.3608, "step": 17910 }, { "epoch": 1.0599160111196546, "grad_norm": 4.359749794006348, "learning_rate": 4.500442377332223e-06, "loss": 0.3363, "step": 17920 }, { "epoch": 1.0605074821080025, "grad_norm": 3.443697690963745, "learning_rate": 4.498783238115969e-06, "loss": 0.2936, "step": 17930 }, { "epoch": 1.0610989530963506, "grad_norm": 3.146984100341797, "learning_rate": 4.497123487780405e-06, "loss": 0.2588, "step": 17940 }, { "epoch": 1.0616904240846987, "grad_norm": 2.7617788314819336, "learning_rate": 4.495463127002282e-06, "loss": 0.3086, "step": 17950 }, { "epoch": 1.0622818950730466, "grad_norm": 2.2403647899627686, "learning_rate": 4.493802156458604e-06, "loss": 0.2886, "step": 17960 }, { "epoch": 1.0628733660613947, "grad_norm": 3.6546711921691895, "learning_rate": 4.492140576826619e-06, "loss": 0.3115, "step": 17970 }, { "epoch": 1.0634648370497426, "grad_norm": 3.2465174198150635, "learning_rate": 4.490478388783827e-06, "loss": 0.3165, "step": 17980 }, { "epoch": 1.0640563080380907, "grad_norm": 4.148971080780029, "learning_rate": 4.4888155930079725e-06, "loss": 0.2738, "step": 17990 }, { "epoch": 1.0646477790264388, "grad_norm": 2.6192190647125244, "learning_rate": 4.487152190177052e-06, "loss": 0.2729, "step": 18000 }, { "epoch": 1.0652392500147867, "grad_norm": 5.104258060455322, "learning_rate": 4.4854881809693085e-06, "loss": 0.3244, "step": 18010 }, { "epoch": 1.0658307210031348, "grad_norm": 1.984011173248291, "learning_rate": 4.483823566063227e-06, "loss": 0.2944, "step": 18020 }, { "epoch": 1.066422191991483, "grad_norm": 3.2251977920532227, "learning_rate": 4.482158346137547e-06, "loss": 0.2639, "step": 18030 }, { "epoch": 1.0670136629798308, "grad_norm": 3.3736367225646973, "learning_rate": 4.48049252187125e-06, "loss": 0.2878, "step": 18040 }, { "epoch": 1.067605133968179, "grad_norm": 4.276313781738281, "learning_rate": 4.478826093943566e-06, "loss": 0.2741, "step": 18050 }, { "epoch": 1.0681966049565268, "grad_norm": 2.8170254230499268, "learning_rate": 4.477159063033969e-06, "loss": 0.3136, "step": 18060 }, { "epoch": 1.068788075944875, "grad_norm": 2.4296038150787354, "learning_rate": 4.475491429822182e-06, "loss": 0.2327, "step": 18070 }, { "epoch": 1.069379546933223, "grad_norm": 3.5622520446777344, "learning_rate": 4.47382319498817e-06, "loss": 0.3761, "step": 18080 }, { "epoch": 1.069971017921571, "grad_norm": 4.229092597961426, "learning_rate": 4.472154359212147e-06, "loss": 0.2717, "step": 18090 }, { "epoch": 1.070562488909919, "grad_norm": 2.8972411155700684, "learning_rate": 4.4704849231745695e-06, "loss": 0.2745, "step": 18100 }, { "epoch": 1.071153959898267, "grad_norm": 3.340649127960205, "learning_rate": 4.468814887556138e-06, "loss": 0.3361, "step": 18110 }, { "epoch": 1.071745430886615, "grad_norm": 2.9985849857330322, "learning_rate": 4.4671442530378e-06, "loss": 0.2902, "step": 18120 }, { "epoch": 1.072336901874963, "grad_norm": 2.486987352371216, "learning_rate": 4.4654730203007465e-06, "loss": 0.2704, "step": 18130 }, { "epoch": 1.072928372863311, "grad_norm": 4.419274806976318, "learning_rate": 4.463801190026411e-06, "loss": 0.3282, "step": 18140 }, { "epoch": 1.073519843851659, "grad_norm": 2.526017904281616, "learning_rate": 4.4621287628964725e-06, "loss": 0.3259, "step": 18150 }, { "epoch": 1.0741113148400072, "grad_norm": 4.1205315589904785, "learning_rate": 4.460455739592852e-06, "loss": 0.2975, "step": 18160 }, { "epoch": 1.074702785828355, "grad_norm": 4.789926528930664, "learning_rate": 4.458782120797715e-06, "loss": 0.3688, "step": 18170 }, { "epoch": 1.0752942568167032, "grad_norm": 4.178130626678467, "learning_rate": 4.457107907193468e-06, "loss": 0.291, "step": 18180 }, { "epoch": 1.075885727805051, "grad_norm": 2.1490769386291504, "learning_rate": 4.455433099462761e-06, "loss": 0.279, "step": 18190 }, { "epoch": 1.0764771987933992, "grad_norm": 3.9377498626708984, "learning_rate": 4.453757698288486e-06, "loss": 0.2661, "step": 18200 }, { "epoch": 1.077068669781747, "grad_norm": 3.0502893924713135, "learning_rate": 4.4520817043537775e-06, "loss": 0.2676, "step": 18210 }, { "epoch": 1.0776601407700952, "grad_norm": 3.432637929916382, "learning_rate": 4.450405118342011e-06, "loss": 0.2894, "step": 18220 }, { "epoch": 1.0782516117584433, "grad_norm": 3.2230567932128906, "learning_rate": 4.448727940936804e-06, "loss": 0.3318, "step": 18230 }, { "epoch": 1.0788430827467912, "grad_norm": 3.406329393386841, "learning_rate": 4.447050172822015e-06, "loss": 0.2959, "step": 18240 }, { "epoch": 1.0794345537351393, "grad_norm": 2.617551803588867, "learning_rate": 4.445371814681743e-06, "loss": 0.2823, "step": 18250 }, { "epoch": 1.0800260247234874, "grad_norm": 2.3511312007904053, "learning_rate": 4.443692867200328e-06, "loss": 0.2821, "step": 18260 }, { "epoch": 1.0806174957118353, "grad_norm": 1.8826947212219238, "learning_rate": 4.44201333106235e-06, "loss": 0.2534, "step": 18270 }, { "epoch": 1.0812089667001834, "grad_norm": 3.553863525390625, "learning_rate": 4.44033320695263e-06, "loss": 0.2451, "step": 18280 }, { "epoch": 1.0818004376885313, "grad_norm": 3.816248893737793, "learning_rate": 4.438652495556227e-06, "loss": 0.3149, "step": 18290 }, { "epoch": 1.0823919086768794, "grad_norm": 3.4486119747161865, "learning_rate": 4.43697119755844e-06, "loss": 0.2749, "step": 18300 }, { "epoch": 1.0829833796652275, "grad_norm": 3.5407631397247314, "learning_rate": 4.4352893136448074e-06, "loss": 0.3373, "step": 18310 }, { "epoch": 1.0835748506535754, "grad_norm": 2.3404979705810547, "learning_rate": 4.433606844501109e-06, "loss": 0.299, "step": 18320 }, { "epoch": 1.0841663216419235, "grad_norm": 3.1772663593292236, "learning_rate": 4.431923790813358e-06, "loss": 0.2777, "step": 18330 }, { "epoch": 1.0847577926302714, "grad_norm": 2.6304008960723877, "learning_rate": 4.4302401532678105e-06, "loss": 0.2669, "step": 18340 }, { "epoch": 1.0853492636186195, "grad_norm": 3.550168991088867, "learning_rate": 4.428555932550958e-06, "loss": 0.3185, "step": 18350 }, { "epoch": 1.0859407346069676, "grad_norm": 3.379551887512207, "learning_rate": 4.426871129349531e-06, "loss": 0.3342, "step": 18360 }, { "epoch": 1.0865322055953155, "grad_norm": 3.114323616027832, "learning_rate": 4.425185744350498e-06, "loss": 0.3428, "step": 18370 }, { "epoch": 1.0871236765836636, "grad_norm": 2.2027018070220947, "learning_rate": 4.423499778241063e-06, "loss": 0.2566, "step": 18380 }, { "epoch": 1.0877151475720117, "grad_norm": 4.013097763061523, "learning_rate": 4.421813231708669e-06, "loss": 0.2963, "step": 18390 }, { "epoch": 1.0883066185603596, "grad_norm": 5.384995460510254, "learning_rate": 4.4201261054409945e-06, "loss": 0.3638, "step": 18400 }, { "epoch": 1.0888980895487077, "grad_norm": 3.6668903827667236, "learning_rate": 4.418438400125953e-06, "loss": 0.3776, "step": 18410 }, { "epoch": 1.0894895605370556, "grad_norm": 3.1524922847747803, "learning_rate": 4.416750116451697e-06, "loss": 0.3024, "step": 18420 }, { "epoch": 1.0900810315254037, "grad_norm": 2.9014382362365723, "learning_rate": 4.415061255106613e-06, "loss": 0.3428, "step": 18430 }, { "epoch": 1.0906725025137516, "grad_norm": 2.4488279819488525, "learning_rate": 4.413371816779323e-06, "loss": 0.3153, "step": 18440 }, { "epoch": 1.0912639735020997, "grad_norm": 3.4289705753326416, "learning_rate": 4.411681802158684e-06, "loss": 0.3166, "step": 18450 }, { "epoch": 1.0918554444904478, "grad_norm": 5.2176337242126465, "learning_rate": 4.4099912119337925e-06, "loss": 0.3099, "step": 18460 }, { "epoch": 1.0924469154787957, "grad_norm": 2.7946949005126953, "learning_rate": 4.408300046793972e-06, "loss": 0.2786, "step": 18470 }, { "epoch": 1.0930383864671438, "grad_norm": 7.622854709625244, "learning_rate": 4.406608307428784e-06, "loss": 0.2997, "step": 18480 }, { "epoch": 1.093629857455492, "grad_norm": 4.088396072387695, "learning_rate": 4.404915994528028e-06, "loss": 0.3141, "step": 18490 }, { "epoch": 1.0942213284438398, "grad_norm": 2.929938793182373, "learning_rate": 4.403223108781731e-06, "loss": 0.3093, "step": 18500 }, { "epoch": 1.094812799432188, "grad_norm": 3.328360080718994, "learning_rate": 4.401529650880158e-06, "loss": 0.2867, "step": 18510 }, { "epoch": 1.0954042704205358, "grad_norm": 5.31728982925415, "learning_rate": 4.3998356215138036e-06, "loss": 0.3186, "step": 18520 }, { "epoch": 1.095995741408884, "grad_norm": 3.8908891677856445, "learning_rate": 4.398141021373399e-06, "loss": 0.347, "step": 18530 }, { "epoch": 1.096587212397232, "grad_norm": 3.8909800052642822, "learning_rate": 4.396445851149906e-06, "loss": 0.3255, "step": 18540 }, { "epoch": 1.09717868338558, "grad_norm": 2.0855040550231934, "learning_rate": 4.39475011153452e-06, "loss": 0.3157, "step": 18550 }, { "epoch": 1.097770154373928, "grad_norm": 3.0681354999542236, "learning_rate": 4.393053803218667e-06, "loss": 0.2858, "step": 18560 }, { "epoch": 1.0983616253622759, "grad_norm": 2.96596360206604, "learning_rate": 4.391356926894007e-06, "loss": 0.3172, "step": 18570 }, { "epoch": 1.098953096350624, "grad_norm": 2.706883668899536, "learning_rate": 4.38965948325243e-06, "loss": 0.3372, "step": 18580 }, { "epoch": 1.099544567338972, "grad_norm": 1.542153239250183, "learning_rate": 4.387961472986058e-06, "loss": 0.2757, "step": 18590 }, { "epoch": 1.10013603832732, "grad_norm": 3.473844528198242, "learning_rate": 4.386262896787244e-06, "loss": 0.2724, "step": 18600 }, { "epoch": 1.100727509315668, "grad_norm": 3.059429407119751, "learning_rate": 4.384563755348573e-06, "loss": 0.2688, "step": 18610 }, { "epoch": 1.1013189803040162, "grad_norm": 4.8419108390808105, "learning_rate": 4.382864049362855e-06, "loss": 0.3194, "step": 18620 }, { "epoch": 1.101910451292364, "grad_norm": 3.291043519973755, "learning_rate": 4.381163779523138e-06, "loss": 0.2766, "step": 18630 }, { "epoch": 1.1025019222807122, "grad_norm": 5.178318977355957, "learning_rate": 4.379462946522694e-06, "loss": 0.2698, "step": 18640 }, { "epoch": 1.10309339326906, "grad_norm": 4.011436462402344, "learning_rate": 4.37776155105503e-06, "loss": 0.2937, "step": 18650 }, { "epoch": 1.1036848642574082, "grad_norm": 3.1692566871643066, "learning_rate": 4.376059593813875e-06, "loss": 0.3283, "step": 18660 }, { "epoch": 1.104276335245756, "grad_norm": 3.0238075256347656, "learning_rate": 4.374357075493193e-06, "loss": 0.286, "step": 18670 }, { "epoch": 1.1048678062341042, "grad_norm": 3.0190954208374023, "learning_rate": 4.372653996787176e-06, "loss": 0.2644, "step": 18680 }, { "epoch": 1.1054592772224523, "grad_norm": 4.5057454109191895, "learning_rate": 4.370950358390242e-06, "loss": 0.2733, "step": 18690 }, { "epoch": 1.1060507482108002, "grad_norm": 2.4388248920440674, "learning_rate": 4.36924616099704e-06, "loss": 0.2957, "step": 18700 }, { "epoch": 1.1066422191991483, "grad_norm": 1.6961699724197388, "learning_rate": 4.367541405302444e-06, "loss": 0.2798, "step": 18710 }, { "epoch": 1.1072336901874964, "grad_norm": 2.853496551513672, "learning_rate": 4.36583609200156e-06, "loss": 0.2487, "step": 18720 }, { "epoch": 1.1078251611758443, "grad_norm": 3.791062116622925, "learning_rate": 4.364130221789716e-06, "loss": 0.3348, "step": 18730 }, { "epoch": 1.1084166321641924, "grad_norm": 4.19989538192749, "learning_rate": 4.362423795362471e-06, "loss": 0.2892, "step": 18740 }, { "epoch": 1.1090081031525403, "grad_norm": 3.0246570110321045, "learning_rate": 4.360716813415608e-06, "loss": 0.3488, "step": 18750 }, { "epoch": 1.1095995741408884, "grad_norm": 2.5639591217041016, "learning_rate": 4.359009276645139e-06, "loss": 0.2993, "step": 18760 }, { "epoch": 1.1101910451292365, "grad_norm": 4.130743503570557, "learning_rate": 4.357301185747302e-06, "loss": 0.3298, "step": 18770 }, { "epoch": 1.1107825161175844, "grad_norm": 2.182361602783203, "learning_rate": 4.35559254141856e-06, "loss": 0.2821, "step": 18780 }, { "epoch": 1.1113739871059325, "grad_norm": 2.669750690460205, "learning_rate": 4.3538833443556e-06, "loss": 0.3194, "step": 18790 }, { "epoch": 1.1119654580942804, "grad_norm": 3.2874128818511963, "learning_rate": 4.352173595255338e-06, "loss": 0.2704, "step": 18800 }, { "epoch": 1.1125569290826285, "grad_norm": 3.2499310970306396, "learning_rate": 4.350463294814913e-06, "loss": 0.2576, "step": 18810 }, { "epoch": 1.1131484000709766, "grad_norm": 1.9718321561813354, "learning_rate": 4.348752443731688e-06, "loss": 0.2432, "step": 18820 }, { "epoch": 1.1137398710593245, "grad_norm": 2.09666109085083, "learning_rate": 4.347041042703254e-06, "loss": 0.2929, "step": 18830 }, { "epoch": 1.1143313420476726, "grad_norm": 4.442263603210449, "learning_rate": 4.345329092427421e-06, "loss": 0.309, "step": 18840 }, { "epoch": 1.1149228130360207, "grad_norm": 5.097445964813232, "learning_rate": 4.343616593602228e-06, "loss": 0.2454, "step": 18850 }, { "epoch": 1.1155142840243686, "grad_norm": 4.605345249176025, "learning_rate": 4.341903546925935e-06, "loss": 0.3295, "step": 18860 }, { "epoch": 1.1161057550127167, "grad_norm": 2.5792229175567627, "learning_rate": 4.340189953097024e-06, "loss": 0.2396, "step": 18870 }, { "epoch": 1.1166972260010646, "grad_norm": 3.5387349128723145, "learning_rate": 4.3384758128142045e-06, "loss": 0.3251, "step": 18880 }, { "epoch": 1.1172886969894127, "grad_norm": 2.6988067626953125, "learning_rate": 4.336761126776405e-06, "loss": 0.2785, "step": 18890 }, { "epoch": 1.1178801679777606, "grad_norm": 2.2510008811950684, "learning_rate": 4.335045895682779e-06, "loss": 0.2302, "step": 18900 }, { "epoch": 1.1184716389661087, "grad_norm": 5.479905128479004, "learning_rate": 4.333330120232699e-06, "loss": 0.2967, "step": 18910 }, { "epoch": 1.1190631099544568, "grad_norm": 2.396517753601074, "learning_rate": 4.331613801125764e-06, "loss": 0.3617, "step": 18920 }, { "epoch": 1.1196545809428047, "grad_norm": 3.024648904800415, "learning_rate": 4.32989693906179e-06, "loss": 0.3, "step": 18930 }, { "epoch": 1.1202460519311528, "grad_norm": 2.7009592056274414, "learning_rate": 4.328179534740817e-06, "loss": 0.2646, "step": 18940 }, { "epoch": 1.120837522919501, "grad_norm": 3.83956241607666, "learning_rate": 4.326461588863106e-06, "loss": 0.3307, "step": 18950 }, { "epoch": 1.1214289939078488, "grad_norm": 3.0033228397369385, "learning_rate": 4.324743102129138e-06, "loss": 0.307, "step": 18960 }, { "epoch": 1.122020464896197, "grad_norm": 3.678713083267212, "learning_rate": 4.3230240752396164e-06, "loss": 0.3639, "step": 18970 }, { "epoch": 1.1226119358845448, "grad_norm": 4.043980598449707, "learning_rate": 4.321304508895463e-06, "loss": 0.3159, "step": 18980 }, { "epoch": 1.1232034068728929, "grad_norm": 2.6200673580169678, "learning_rate": 4.319584403797818e-06, "loss": 0.2856, "step": 18990 }, { "epoch": 1.123794877861241, "grad_norm": 2.4232752323150635, "learning_rate": 4.317863760648046e-06, "loss": 0.2731, "step": 19000 }, { "epoch": 1.1243863488495889, "grad_norm": 4.450223922729492, "learning_rate": 4.316142580147727e-06, "loss": 0.2879, "step": 19010 }, { "epoch": 1.124977819837937, "grad_norm": 2.601515769958496, "learning_rate": 4.314420862998661e-06, "loss": 0.33, "step": 19020 }, { "epoch": 1.1255692908262849, "grad_norm": 2.2870609760284424, "learning_rate": 4.312698609902869e-06, "loss": 0.2976, "step": 19030 }, { "epoch": 1.126160761814633, "grad_norm": 3.142299175262451, "learning_rate": 4.310975821562587e-06, "loss": 0.306, "step": 19040 }, { "epoch": 1.126752232802981, "grad_norm": 3.390784978866577, "learning_rate": 4.30925249868027e-06, "loss": 0.3106, "step": 19050 }, { "epoch": 1.127343703791329, "grad_norm": 4.461346626281738, "learning_rate": 4.307528641958594e-06, "loss": 0.3172, "step": 19060 }, { "epoch": 1.127935174779677, "grad_norm": 2.631408929824829, "learning_rate": 4.305804252100451e-06, "loss": 0.3126, "step": 19070 }, { "epoch": 1.1285266457680252, "grad_norm": 2.2730295658111572, "learning_rate": 4.304079329808947e-06, "loss": 0.2887, "step": 19080 }, { "epoch": 1.129118116756373, "grad_norm": 3.524040699005127, "learning_rate": 4.302353875787411e-06, "loss": 0.2567, "step": 19090 }, { "epoch": 1.1297095877447212, "grad_norm": 1.6717512607574463, "learning_rate": 4.3006278907393845e-06, "loss": 0.3061, "step": 19100 }, { "epoch": 1.130301058733069, "grad_norm": 2.2473487854003906, "learning_rate": 4.298901375368628e-06, "loss": 0.3226, "step": 19110 }, { "epoch": 1.1308925297214172, "grad_norm": 3.505798101425171, "learning_rate": 4.297174330379115e-06, "loss": 0.2882, "step": 19120 }, { "epoch": 1.131484000709765, "grad_norm": 2.6387484073638916, "learning_rate": 4.295446756475039e-06, "loss": 0.2858, "step": 19130 }, { "epoch": 1.1320754716981132, "grad_norm": 4.004335880279541, "learning_rate": 4.293718654360807e-06, "loss": 0.2976, "step": 19140 }, { "epoch": 1.1326669426864613, "grad_norm": 3.0942773818969727, "learning_rate": 4.291990024741041e-06, "loss": 0.2594, "step": 19150 }, { "epoch": 1.1332584136748092, "grad_norm": 4.031425952911377, "learning_rate": 4.290260868320579e-06, "loss": 0.2399, "step": 19160 }, { "epoch": 1.1338498846631573, "grad_norm": 3.6472721099853516, "learning_rate": 4.288531185804475e-06, "loss": 0.218, "step": 19170 }, { "epoch": 1.1344413556515054, "grad_norm": 3.539475202560425, "learning_rate": 4.286800977897993e-06, "loss": 0.3564, "step": 19180 }, { "epoch": 1.1350328266398533, "grad_norm": 2.6184306144714355, "learning_rate": 4.285070245306615e-06, "loss": 0.2739, "step": 19190 }, { "epoch": 1.1356242976282014, "grad_norm": 3.2064406871795654, "learning_rate": 4.2833389887360374e-06, "loss": 0.2919, "step": 19200 }, { "epoch": 1.1362157686165493, "grad_norm": 2.380692481994629, "learning_rate": 4.281607208892169e-06, "loss": 0.2349, "step": 19210 }, { "epoch": 1.1368072396048974, "grad_norm": 2.625814914703369, "learning_rate": 4.279874906481131e-06, "loss": 0.3209, "step": 19220 }, { "epoch": 1.1373987105932455, "grad_norm": 1.903015375137329, "learning_rate": 4.278142082209261e-06, "loss": 0.2196, "step": 19230 }, { "epoch": 1.1379901815815934, "grad_norm": 2.3114216327667236, "learning_rate": 4.276408736783105e-06, "loss": 0.3439, "step": 19240 }, { "epoch": 1.1385816525699415, "grad_norm": 3.674055814743042, "learning_rate": 4.274674870909422e-06, "loss": 0.2832, "step": 19250 }, { "epoch": 1.1391731235582894, "grad_norm": 2.242675542831421, "learning_rate": 4.272940485295188e-06, "loss": 0.273, "step": 19260 }, { "epoch": 1.1397645945466375, "grad_norm": 2.7175843715667725, "learning_rate": 4.271205580647586e-06, "loss": 0.2751, "step": 19270 }, { "epoch": 1.1403560655349856, "grad_norm": 4.477102756500244, "learning_rate": 4.269470157674012e-06, "loss": 0.3165, "step": 19280 }, { "epoch": 1.1409475365233335, "grad_norm": 3.406383514404297, "learning_rate": 4.267734217082074e-06, "loss": 0.2984, "step": 19290 }, { "epoch": 1.1415390075116816, "grad_norm": 2.1384739875793457, "learning_rate": 4.265997759579593e-06, "loss": 0.2596, "step": 19300 }, { "epoch": 1.1421304785000297, "grad_norm": 2.9027628898620605, "learning_rate": 4.264260785874595e-06, "loss": 0.3134, "step": 19310 }, { "epoch": 1.1427219494883776, "grad_norm": 2.3354344367980957, "learning_rate": 4.262523296675321e-06, "loss": 0.2955, "step": 19320 }, { "epoch": 1.1433134204767257, "grad_norm": 4.422817707061768, "learning_rate": 4.260785292690221e-06, "loss": 0.2701, "step": 19330 }, { "epoch": 1.1439048914650736, "grad_norm": 3.8003275394439697, "learning_rate": 4.259046774627956e-06, "loss": 0.3084, "step": 19340 }, { "epoch": 1.1444963624534217, "grad_norm": 2.2730629444122314, "learning_rate": 4.2573077431973965e-06, "loss": 0.2858, "step": 19350 }, { "epoch": 1.1450878334417696, "grad_norm": 2.8439552783966064, "learning_rate": 4.25556819910762e-06, "loss": 0.2771, "step": 19360 }, { "epoch": 1.1456793044301177, "grad_norm": 4.391366481781006, "learning_rate": 4.253828143067913e-06, "loss": 0.2761, "step": 19370 }, { "epoch": 1.1462707754184658, "grad_norm": 3.955793857574463, "learning_rate": 4.252087575787776e-06, "loss": 0.293, "step": 19380 }, { "epoch": 1.1468622464068137, "grad_norm": 3.819779396057129, "learning_rate": 4.25034649797691e-06, "loss": 0.3164, "step": 19390 }, { "epoch": 1.1474537173951618, "grad_norm": 2.5986738204956055, "learning_rate": 4.248604910345232e-06, "loss": 0.2948, "step": 19400 }, { "epoch": 1.1480451883835099, "grad_norm": 1.917783498764038, "learning_rate": 4.246862813602861e-06, "loss": 0.2942, "step": 19410 }, { "epoch": 1.1486366593718578, "grad_norm": 2.817293405532837, "learning_rate": 4.245120208460128e-06, "loss": 0.2838, "step": 19420 }, { "epoch": 1.1492281303602059, "grad_norm": 2.629216194152832, "learning_rate": 4.243377095627567e-06, "loss": 0.2596, "step": 19430 }, { "epoch": 1.1498196013485538, "grad_norm": 3.148768186569214, "learning_rate": 4.241633475815921e-06, "loss": 0.3065, "step": 19440 }, { "epoch": 1.1504110723369019, "grad_norm": 2.8173105716705322, "learning_rate": 4.239889349736142e-06, "loss": 0.3247, "step": 19450 }, { "epoch": 1.15100254332525, "grad_norm": 2.8522443771362305, "learning_rate": 4.238144718099384e-06, "loss": 0.2705, "step": 19460 }, { "epoch": 1.1515940143135979, "grad_norm": 3.2155890464782715, "learning_rate": 4.2363995816170125e-06, "loss": 0.3124, "step": 19470 }, { "epoch": 1.152185485301946, "grad_norm": 4.815730571746826, "learning_rate": 4.234653941000592e-06, "loss": 0.291, "step": 19480 }, { "epoch": 1.1527769562902939, "grad_norm": 5.894669532775879, "learning_rate": 4.2329077969619e-06, "loss": 0.2648, "step": 19490 }, { "epoch": 1.153368427278642, "grad_norm": 4.52048397064209, "learning_rate": 4.231161150212912e-06, "loss": 0.3306, "step": 19500 }, { "epoch": 1.15395989826699, "grad_norm": 2.700951337814331, "learning_rate": 4.229414001465814e-06, "loss": 0.3092, "step": 19510 }, { "epoch": 1.154551369255338, "grad_norm": 3.0089612007141113, "learning_rate": 4.227666351432994e-06, "loss": 0.3218, "step": 19520 }, { "epoch": 1.155142840243686, "grad_norm": 2.9668242931365967, "learning_rate": 4.225918200827047e-06, "loss": 0.3981, "step": 19530 }, { "epoch": 1.1557343112320342, "grad_norm": 3.2057981491088867, "learning_rate": 4.2241695503607674e-06, "loss": 0.3209, "step": 19540 }, { "epoch": 1.156325782220382, "grad_norm": 3.564058542251587, "learning_rate": 4.22242040074716e-06, "loss": 0.3172, "step": 19550 }, { "epoch": 1.1569172532087302, "grad_norm": 2.0039002895355225, "learning_rate": 4.220670752699426e-06, "loss": 0.2932, "step": 19560 }, { "epoch": 1.157508724197078, "grad_norm": 4.055542469024658, "learning_rate": 4.218920606930975e-06, "loss": 0.2527, "step": 19570 }, { "epoch": 1.1581001951854262, "grad_norm": 3.3519890308380127, "learning_rate": 4.217169964155417e-06, "loss": 0.2943, "step": 19580 }, { "epoch": 1.158691666173774, "grad_norm": 3.300375461578369, "learning_rate": 4.215418825086565e-06, "loss": 0.3394, "step": 19590 }, { "epoch": 1.1592831371621222, "grad_norm": 3.347454309463501, "learning_rate": 4.213667190438437e-06, "loss": 0.2278, "step": 19600 }, { "epoch": 1.1598746081504703, "grad_norm": 3.8394901752471924, "learning_rate": 4.211915060925248e-06, "loss": 0.3858, "step": 19610 }, { "epoch": 1.1604660791388182, "grad_norm": 3.550410032272339, "learning_rate": 4.21016243726142e-06, "loss": 0.2877, "step": 19620 }, { "epoch": 1.1610575501271663, "grad_norm": 3.767458438873291, "learning_rate": 4.2084093201615725e-06, "loss": 0.2449, "step": 19630 }, { "epoch": 1.1616490211155144, "grad_norm": 3.544475555419922, "learning_rate": 4.206655710340529e-06, "loss": 0.3302, "step": 19640 }, { "epoch": 1.1622404921038623, "grad_norm": 4.2046637535095215, "learning_rate": 4.204901608513311e-06, "loss": 0.317, "step": 19650 }, { "epoch": 1.1628319630922104, "grad_norm": 3.732708692550659, "learning_rate": 4.203147015395145e-06, "loss": 0.2878, "step": 19660 }, { "epoch": 1.1634234340805583, "grad_norm": 2.4466331005096436, "learning_rate": 4.201391931701453e-06, "loss": 0.2936, "step": 19670 }, { "epoch": 1.1640149050689064, "grad_norm": 2.0265417098999023, "learning_rate": 4.199636358147859e-06, "loss": 0.3289, "step": 19680 }, { "epoch": 1.1646063760572545, "grad_norm": 3.5031895637512207, "learning_rate": 4.197880295450191e-06, "loss": 0.3104, "step": 19690 }, { "epoch": 1.1651978470456024, "grad_norm": 3.5693209171295166, "learning_rate": 4.196123744324465e-06, "loss": 0.3018, "step": 19700 }, { "epoch": 1.1657893180339505, "grad_norm": 3.20611834526062, "learning_rate": 4.19436670548691e-06, "loss": 0.348, "step": 19710 }, { "epoch": 1.1663807890222984, "grad_norm": 4.072464466094971, "learning_rate": 4.192609179653944e-06, "loss": 0.3273, "step": 19720 }, { "epoch": 1.1669722600106465, "grad_norm": 3.1917338371276855, "learning_rate": 4.190851167542189e-06, "loss": 0.2769, "step": 19730 }, { "epoch": 1.1675637309989946, "grad_norm": 3.4168622493743896, "learning_rate": 4.189092669868463e-06, "loss": 0.2906, "step": 19740 }, { "epoch": 1.1681552019873425, "grad_norm": 1.3425867557525635, "learning_rate": 4.187333687349779e-06, "loss": 0.2457, "step": 19750 }, { "epoch": 1.1687466729756906, "grad_norm": 4.954278945922852, "learning_rate": 4.1855742207033544e-06, "loss": 0.3446, "step": 19760 }, { "epoch": 1.1693381439640387, "grad_norm": 3.384310483932495, "learning_rate": 4.1838142706466e-06, "loss": 0.3629, "step": 19770 }, { "epoch": 1.1699296149523866, "grad_norm": 2.001405954360962, "learning_rate": 4.182053837897123e-06, "loss": 0.2976, "step": 19780 }, { "epoch": 1.1705210859407347, "grad_norm": 4.344552993774414, "learning_rate": 4.180292923172731e-06, "loss": 0.3451, "step": 19790 }, { "epoch": 1.1711125569290826, "grad_norm": 3.9154059886932373, "learning_rate": 4.178531527191422e-06, "loss": 0.3248, "step": 19800 }, { "epoch": 1.1717040279174307, "grad_norm": 2.014502763748169, "learning_rate": 4.1767696506713975e-06, "loss": 0.2779, "step": 19810 }, { "epoch": 1.1722954989057786, "grad_norm": 4.08761739730835, "learning_rate": 4.1750072943310495e-06, "loss": 0.3012, "step": 19820 }, { "epoch": 1.1728869698941267, "grad_norm": 2.96844220161438, "learning_rate": 4.173244458888968e-06, "loss": 0.2468, "step": 19830 }, { "epoch": 1.1734784408824748, "grad_norm": 3.2615513801574707, "learning_rate": 4.171481145063938e-06, "loss": 0.2613, "step": 19840 }, { "epoch": 1.1740699118708227, "grad_norm": 2.4595625400543213, "learning_rate": 4.169717353574939e-06, "loss": 0.2512, "step": 19850 }, { "epoch": 1.1746613828591708, "grad_norm": 4.755798816680908, "learning_rate": 4.167953085141146e-06, "loss": 0.3652, "step": 19860 }, { "epoch": 1.1752528538475189, "grad_norm": 3.915400981903076, "learning_rate": 4.16618834048193e-06, "loss": 0.34, "step": 19870 }, { "epoch": 1.1758443248358668, "grad_norm": 2.433000087738037, "learning_rate": 4.164423120316852e-06, "loss": 0.312, "step": 19880 }, { "epoch": 1.1764357958242149, "grad_norm": 2.8739559650421143, "learning_rate": 4.162657425365668e-06, "loss": 0.3001, "step": 19890 }, { "epoch": 1.1770272668125628, "grad_norm": 2.3134453296661377, "learning_rate": 4.160891256348334e-06, "loss": 0.242, "step": 19900 }, { "epoch": 1.1776187378009109, "grad_norm": 4.393862724304199, "learning_rate": 4.159124613984989e-06, "loss": 0.2792, "step": 19910 }, { "epoch": 1.178210208789259, "grad_norm": 4.257948398590088, "learning_rate": 4.157357498995973e-06, "loss": 0.2342, "step": 19920 }, { "epoch": 1.1788016797776069, "grad_norm": 2.262514352798462, "learning_rate": 4.1555899121018145e-06, "loss": 0.3072, "step": 19930 }, { "epoch": 1.179393150765955, "grad_norm": 3.2887322902679443, "learning_rate": 4.153821854023237e-06, "loss": 0.3149, "step": 19940 }, { "epoch": 1.1799846217543029, "grad_norm": 6.912112236022949, "learning_rate": 4.152053325481154e-06, "loss": 0.2988, "step": 19950 }, { "epoch": 1.180576092742651, "grad_norm": 2.254776954650879, "learning_rate": 4.150284327196672e-06, "loss": 0.319, "step": 19960 }, { "epoch": 1.181167563730999, "grad_norm": 3.782851219177246, "learning_rate": 4.148514859891089e-06, "loss": 0.2791, "step": 19970 }, { "epoch": 1.181759034719347, "grad_norm": 2.5606906414031982, "learning_rate": 4.146744924285893e-06, "loss": 0.2679, "step": 19980 }, { "epoch": 1.182350505707695, "grad_norm": 3.643106460571289, "learning_rate": 4.144974521102765e-06, "loss": 0.2428, "step": 19990 }, { "epoch": 1.1829419766960432, "grad_norm": 2.768740177154541, "learning_rate": 4.143203651063576e-06, "loss": 0.2831, "step": 20000 }, { "epoch": 1.183533447684391, "grad_norm": 4.883973121643066, "learning_rate": 4.1414323148903875e-06, "loss": 0.3397, "step": 20010 }, { "epoch": 1.1841249186727392, "grad_norm": 2.8383262157440186, "learning_rate": 4.1396605133054474e-06, "loss": 0.3421, "step": 20020 }, { "epoch": 1.184716389661087, "grad_norm": 2.8787240982055664, "learning_rate": 4.137888247031199e-06, "loss": 0.2565, "step": 20030 }, { "epoch": 1.1853078606494352, "grad_norm": 2.0451982021331787, "learning_rate": 4.136115516790272e-06, "loss": 0.2654, "step": 20040 }, { "epoch": 1.185899331637783, "grad_norm": 3.1647186279296875, "learning_rate": 4.134342323305486e-06, "loss": 0.2583, "step": 20050 }, { "epoch": 1.1864908026261312, "grad_norm": 3.7619566917419434, "learning_rate": 4.132568667299849e-06, "loss": 0.2839, "step": 20060 }, { "epoch": 1.1870822736144793, "grad_norm": 5.380703449249268, "learning_rate": 4.130794549496559e-06, "loss": 0.3251, "step": 20070 }, { "epoch": 1.1876737446028272, "grad_norm": 3.421454906463623, "learning_rate": 4.129019970618999e-06, "loss": 0.2797, "step": 20080 }, { "epoch": 1.1882652155911753, "grad_norm": 2.928265333175659, "learning_rate": 4.127244931390744e-06, "loss": 0.3204, "step": 20090 }, { "epoch": 1.1888566865795234, "grad_norm": 2.505737781524658, "learning_rate": 4.125469432535554e-06, "loss": 0.2946, "step": 20100 }, { "epoch": 1.1894481575678713, "grad_norm": 2.5254628658294678, "learning_rate": 4.12369347477738e-06, "loss": 0.2866, "step": 20110 }, { "epoch": 1.1900396285562194, "grad_norm": 3.1585745811462402, "learning_rate": 4.121917058840353e-06, "loss": 0.2831, "step": 20120 }, { "epoch": 1.1906310995445673, "grad_norm": 2.3900585174560547, "learning_rate": 4.1201401854488e-06, "loss": 0.3051, "step": 20130 }, { "epoch": 1.1912225705329154, "grad_norm": 5.00233268737793, "learning_rate": 4.118362855327226e-06, "loss": 0.2501, "step": 20140 }, { "epoch": 1.1918140415212635, "grad_norm": 3.194591522216797, "learning_rate": 4.116585069200327e-06, "loss": 0.2742, "step": 20150 }, { "epoch": 1.1924055125096114, "grad_norm": 3.2211551666259766, "learning_rate": 4.114806827792985e-06, "loss": 0.237, "step": 20160 }, { "epoch": 1.1929969834979595, "grad_norm": 2.7687718868255615, "learning_rate": 4.113028131830265e-06, "loss": 0.3576, "step": 20170 }, { "epoch": 1.1935884544863073, "grad_norm": 3.3792688846588135, "learning_rate": 4.111248982037422e-06, "loss": 0.2999, "step": 20180 }, { "epoch": 1.1941799254746555, "grad_norm": 2.35890531539917, "learning_rate": 4.109469379139889e-06, "loss": 0.2354, "step": 20190 }, { "epoch": 1.1947713964630036, "grad_norm": 2.596205711364746, "learning_rate": 4.107689323863291e-06, "loss": 0.3346, "step": 20200 }, { "epoch": 1.1953628674513515, "grad_norm": 2.103747844696045, "learning_rate": 4.105908816933433e-06, "loss": 0.3202, "step": 20210 }, { "epoch": 1.1959543384396996, "grad_norm": 3.6850767135620117, "learning_rate": 4.104127859076304e-06, "loss": 0.2788, "step": 20220 }, { "epoch": 1.1965458094280477, "grad_norm": 7.547243595123291, "learning_rate": 4.10234645101808e-06, "loss": 0.2946, "step": 20230 }, { "epoch": 1.1971372804163956, "grad_norm": 3.6698873043060303, "learning_rate": 4.100564593485118e-06, "loss": 0.2785, "step": 20240 }, { "epoch": 1.1977287514047437, "grad_norm": 3.325680732727051, "learning_rate": 4.098782287203958e-06, "loss": 0.2994, "step": 20250 }, { "epoch": 1.1983202223930915, "grad_norm": 2.9590067863464355, "learning_rate": 4.096999532901327e-06, "loss": 0.3586, "step": 20260 }, { "epoch": 1.1989116933814397, "grad_norm": 2.846670150756836, "learning_rate": 4.09521633130413e-06, "loss": 0.284, "step": 20270 }, { "epoch": 1.1995031643697875, "grad_norm": 3.6687681674957275, "learning_rate": 4.093432683139456e-06, "loss": 0.2707, "step": 20280 }, { "epoch": 1.2000946353581357, "grad_norm": 1.7186319828033447, "learning_rate": 4.091648589134575e-06, "loss": 0.3128, "step": 20290 }, { "epoch": 1.2002129295558053, "eval_accuracy": 0.8627441195062714, "eval_loss": 0.3310759365558624, "eval_runtime": 200.818, "eval_safe_aucpr": 0.919105096789151, "eval_safe_f1": 0.849255503790993, "eval_safe_fpr": 0.14439030223311666, "eval_safe_precision": 0.8279424337418068, "eval_safe_recall": 0.8716948580429809, "eval_samples_per_second": 299.346, "eval_steps_per_second": 37.422, "eval_unsafe_aucpr": 0.9531596403695981, "eval_unsafe_f1": 0.8740170705266211, "eval_unsafe_fpr": 0.12830514195701861, "eval_unsafe_precision": 0.893233880531802, "eval_unsafe_recall": 0.8556096977668829, "step": 20292 }, { "epoch": 1.2006861063464838, "grad_norm": 4.1715288162231445, "learning_rate": 4.089864050016943e-06, "loss": 0.3109, "step": 20300 }, { "epoch": 1.2012775773348316, "grad_norm": 3.9092912673950195, "learning_rate": 4.0880790665141905e-06, "loss": 0.2345, "step": 20310 }, { "epoch": 1.2018690483231798, "grad_norm": 3.2306067943573, "learning_rate": 4.0862936393541365e-06, "loss": 0.2684, "step": 20320 }, { "epoch": 1.2024605193115279, "grad_norm": 1.8747831583023071, "learning_rate": 4.084507769264777e-06, "loss": 0.3157, "step": 20330 }, { "epoch": 1.2030519902998758, "grad_norm": 3.042409658432007, "learning_rate": 4.082721456974287e-06, "loss": 0.3679, "step": 20340 }, { "epoch": 1.2036434612882239, "grad_norm": 2.9119303226470947, "learning_rate": 4.080934703211026e-06, "loss": 0.279, "step": 20350 }, { "epoch": 1.2042349322765717, "grad_norm": 3.012389659881592, "learning_rate": 4.079147508703531e-06, "loss": 0.2651, "step": 20360 }, { "epoch": 1.2048264032649199, "grad_norm": 4.5980916023254395, "learning_rate": 4.077359874180518e-06, "loss": 0.2589, "step": 20370 }, { "epoch": 1.205417874253268, "grad_norm": 4.228001117706299, "learning_rate": 4.075571800370885e-06, "loss": 0.2831, "step": 20380 }, { "epoch": 1.2060093452416158, "grad_norm": 2.842913866043091, "learning_rate": 4.0737832880037085e-06, "loss": 0.2677, "step": 20390 }, { "epoch": 1.206600816229964, "grad_norm": 2.411369562149048, "learning_rate": 4.071994337808241e-06, "loss": 0.2503, "step": 20400 }, { "epoch": 1.2071922872183118, "grad_norm": 2.9672956466674805, "learning_rate": 4.070204950513916e-06, "loss": 0.2997, "step": 20410 }, { "epoch": 1.20778375820666, "grad_norm": 2.3697445392608643, "learning_rate": 4.068415126850346e-06, "loss": 0.2961, "step": 20420 }, { "epoch": 1.208375229195008, "grad_norm": 3.6388251781463623, "learning_rate": 4.066624867547318e-06, "loss": 0.3024, "step": 20430 }, { "epoch": 1.208966700183356, "grad_norm": 3.510526418685913, "learning_rate": 4.0648341733348014e-06, "loss": 0.2768, "step": 20440 }, { "epoch": 1.209558171171704, "grad_norm": 2.514333724975586, "learning_rate": 4.063043044942938e-06, "loss": 0.2687, "step": 20450 }, { "epoch": 1.2101496421600522, "grad_norm": 3.403837203979492, "learning_rate": 4.061251483102051e-06, "loss": 0.3519, "step": 20460 }, { "epoch": 1.2107411131484, "grad_norm": 3.142629384994507, "learning_rate": 4.0594594885426375e-06, "loss": 0.2936, "step": 20470 }, { "epoch": 1.2113325841367482, "grad_norm": 3.3574156761169434, "learning_rate": 4.05766706199537e-06, "loss": 0.3103, "step": 20480 }, { "epoch": 1.211924055125096, "grad_norm": 2.508437395095825, "learning_rate": 4.055874204191102e-06, "loss": 0.3344, "step": 20490 }, { "epoch": 1.2125155261134442, "grad_norm": 2.484335422515869, "learning_rate": 4.054080915860858e-06, "loss": 0.3134, "step": 20500 }, { "epoch": 1.213106997101792, "grad_norm": 2.890902042388916, "learning_rate": 4.052287197735842e-06, "loss": 0.3374, "step": 20510 }, { "epoch": 1.2136984680901401, "grad_norm": 3.590841054916382, "learning_rate": 4.0504930505474285e-06, "loss": 0.3504, "step": 20520 }, { "epoch": 1.2142899390784883, "grad_norm": 2.0149879455566406, "learning_rate": 4.048698475027171e-06, "loss": 0.3018, "step": 20530 }, { "epoch": 1.2148814100668361, "grad_norm": 2.9346234798431396, "learning_rate": 4.046903471906795e-06, "loss": 0.2841, "step": 20540 }, { "epoch": 1.2154728810551843, "grad_norm": 3.102548837661743, "learning_rate": 4.0451080419182035e-06, "loss": 0.3256, "step": 20550 }, { "epoch": 1.2160643520435324, "grad_norm": 3.3685483932495117, "learning_rate": 4.0433121857934695e-06, "loss": 0.2977, "step": 20560 }, { "epoch": 1.2166558230318802, "grad_norm": 2.2467429637908936, "learning_rate": 4.041515904264844e-06, "loss": 0.2397, "step": 20570 }, { "epoch": 1.2172472940202284, "grad_norm": 2.8709728717803955, "learning_rate": 4.039719198064748e-06, "loss": 0.2779, "step": 20580 }, { "epoch": 1.2178387650085762, "grad_norm": 2.390738010406494, "learning_rate": 4.037922067925777e-06, "loss": 0.2878, "step": 20590 }, { "epoch": 1.2184302359969243, "grad_norm": 2.7928378582000732, "learning_rate": 4.0361245145807e-06, "loss": 0.2427, "step": 20600 }, { "epoch": 1.2190217069852725, "grad_norm": 2.3413987159729004, "learning_rate": 4.034326538762457e-06, "loss": 0.2629, "step": 20610 }, { "epoch": 1.2196131779736203, "grad_norm": 3.022162914276123, "learning_rate": 4.032528141204162e-06, "loss": 0.2842, "step": 20620 }, { "epoch": 1.2202046489619685, "grad_norm": 5.002121925354004, "learning_rate": 4.0307293226391e-06, "loss": 0.2992, "step": 20630 }, { "epoch": 1.2207961199503163, "grad_norm": 2.910151958465576, "learning_rate": 4.028930083800729e-06, "loss": 0.2895, "step": 20640 }, { "epoch": 1.2213875909386644, "grad_norm": 1.9784057140350342, "learning_rate": 4.027130425422674e-06, "loss": 0.2869, "step": 20650 }, { "epoch": 1.2219790619270126, "grad_norm": 3.1606860160827637, "learning_rate": 4.025330348238737e-06, "loss": 0.3198, "step": 20660 }, { "epoch": 1.2225705329153604, "grad_norm": 2.3709053993225098, "learning_rate": 4.023529852982887e-06, "loss": 0.3584, "step": 20670 }, { "epoch": 1.2231620039037086, "grad_norm": 1.879734992980957, "learning_rate": 4.0217289403892645e-06, "loss": 0.2835, "step": 20680 }, { "epoch": 1.2237534748920567, "grad_norm": 4.057839393615723, "learning_rate": 4.01992761119218e-06, "loss": 0.3117, "step": 20690 }, { "epoch": 1.2243449458804045, "grad_norm": 3.37515926361084, "learning_rate": 4.018125866126114e-06, "loss": 0.2844, "step": 20700 }, { "epoch": 1.2249364168687527, "grad_norm": 4.256083011627197, "learning_rate": 4.016323705925718e-06, "loss": 0.3551, "step": 20710 }, { "epoch": 1.2255278878571005, "grad_norm": 3.573638677597046, "learning_rate": 4.014521131325809e-06, "loss": 0.2796, "step": 20720 }, { "epoch": 1.2261193588454486, "grad_norm": 2.827806234359741, "learning_rate": 4.012718143061376e-06, "loss": 0.2872, "step": 20730 }, { "epoch": 1.2267108298337965, "grad_norm": 3.245790481567383, "learning_rate": 4.010914741867578e-06, "loss": 0.2946, "step": 20740 }, { "epoch": 1.2273023008221446, "grad_norm": 2.2304093837738037, "learning_rate": 4.0091109284797376e-06, "loss": 0.2813, "step": 20750 }, { "epoch": 1.2278937718104928, "grad_norm": 3.8928170204162598, "learning_rate": 4.007306703633349e-06, "loss": 0.3102, "step": 20760 }, { "epoch": 1.2284852427988406, "grad_norm": 3.0669772624969482, "learning_rate": 4.005502068064074e-06, "loss": 0.383, "step": 20770 }, { "epoch": 1.2290767137871887, "grad_norm": 2.400132417678833, "learning_rate": 4.003697022507741e-06, "loss": 0.295, "step": 20780 }, { "epoch": 1.2296681847755369, "grad_norm": 4.6859450340271, "learning_rate": 4.0018915677003465e-06, "loss": 0.3468, "step": 20790 }, { "epoch": 1.2302596557638847, "grad_norm": 2.200775623321533, "learning_rate": 4.000085704378052e-06, "loss": 0.2882, "step": 20800 }, { "epoch": 1.2308511267522328, "grad_norm": 1.9656991958618164, "learning_rate": 3.998279433277186e-06, "loss": 0.3019, "step": 20810 }, { "epoch": 1.2314425977405807, "grad_norm": 4.176045894622803, "learning_rate": 3.996472755134246e-06, "loss": 0.3113, "step": 20820 }, { "epoch": 1.2320340687289288, "grad_norm": 2.2928171157836914, "learning_rate": 3.994665670685894e-06, "loss": 0.2969, "step": 20830 }, { "epoch": 1.232625539717277, "grad_norm": 3.806422472000122, "learning_rate": 3.992858180668954e-06, "loss": 0.3239, "step": 20840 }, { "epoch": 1.2332170107056248, "grad_norm": 2.9722955226898193, "learning_rate": 3.9910502858204215e-06, "loss": 0.2977, "step": 20850 }, { "epoch": 1.233808481693973, "grad_norm": 3.4820804595947266, "learning_rate": 3.989241986877453e-06, "loss": 0.2741, "step": 20860 }, { "epoch": 1.2343999526823208, "grad_norm": 4.019099712371826, "learning_rate": 3.987433284577371e-06, "loss": 0.2906, "step": 20870 }, { "epoch": 1.234991423670669, "grad_norm": 2.3220839500427246, "learning_rate": 3.985624179657662e-06, "loss": 0.3072, "step": 20880 }, { "epoch": 1.235582894659017, "grad_norm": 3.44693922996521, "learning_rate": 3.9838146728559765e-06, "loss": 0.2734, "step": 20890 }, { "epoch": 1.236174365647365, "grad_norm": 3.8457822799682617, "learning_rate": 3.982004764910131e-06, "loss": 0.2458, "step": 20900 }, { "epoch": 1.236765836635713, "grad_norm": 6.166121006011963, "learning_rate": 3.980194456558102e-06, "loss": 0.2407, "step": 20910 }, { "epoch": 1.2373573076240612, "grad_norm": 3.4304347038269043, "learning_rate": 3.9783837485380334e-06, "loss": 0.3083, "step": 20920 }, { "epoch": 1.237948778612409, "grad_norm": 4.261125564575195, "learning_rate": 3.9765726415882275e-06, "loss": 0.336, "step": 20930 }, { "epoch": 1.2385402496007571, "grad_norm": 3.9217276573181152, "learning_rate": 3.974761136447153e-06, "loss": 0.2564, "step": 20940 }, { "epoch": 1.239131720589105, "grad_norm": 1.7338852882385254, "learning_rate": 3.97294923385344e-06, "loss": 0.2426, "step": 20950 }, { "epoch": 1.2397231915774531, "grad_norm": 3.293468713760376, "learning_rate": 3.971136934545879e-06, "loss": 0.3166, "step": 20960 }, { "epoch": 1.240314662565801, "grad_norm": 4.378649711608887, "learning_rate": 3.969324239263425e-06, "loss": 0.3139, "step": 20970 }, { "epoch": 1.2409061335541491, "grad_norm": 3.0796139240264893, "learning_rate": 3.9675111487451915e-06, "loss": 0.3177, "step": 20980 }, { "epoch": 1.2414976045424972, "grad_norm": 3.7481329441070557, "learning_rate": 3.965697663730455e-06, "loss": 0.3413, "step": 20990 }, { "epoch": 1.2420890755308451, "grad_norm": 8.165270805358887, "learning_rate": 3.9638837849586515e-06, "loss": 0.3267, "step": 21000 }, { "epoch": 1.2426805465191932, "grad_norm": 2.160054922103882, "learning_rate": 3.962069513169381e-06, "loss": 0.2098, "step": 21010 }, { "epoch": 1.2432720175075413, "grad_norm": 2.8127756118774414, "learning_rate": 3.960254849102399e-06, "loss": 0.3175, "step": 21020 }, { "epoch": 1.2438634884958892, "grad_norm": 4.616055488586426, "learning_rate": 3.958439793497625e-06, "loss": 0.318, "step": 21030 }, { "epoch": 1.2444549594842373, "grad_norm": 3.6967434883117676, "learning_rate": 3.956624347095134e-06, "loss": 0.3493, "step": 21040 }, { "epoch": 1.2450464304725852, "grad_norm": 1.6208961009979248, "learning_rate": 3.954808510635165e-06, "loss": 0.2618, "step": 21050 }, { "epoch": 1.2456379014609333, "grad_norm": 4.28981876373291, "learning_rate": 3.952992284858113e-06, "loss": 0.3135, "step": 21060 }, { "epoch": 1.2462293724492814, "grad_norm": 3.690321207046509, "learning_rate": 3.9511756705045305e-06, "loss": 0.3195, "step": 21070 }, { "epoch": 1.2468208434376293, "grad_norm": 3.9370062351226807, "learning_rate": 3.949358668315134e-06, "loss": 0.2738, "step": 21080 }, { "epoch": 1.2474123144259774, "grad_norm": 3.132798433303833, "learning_rate": 3.9475412790307924e-06, "loss": 0.3596, "step": 21090 }, { "epoch": 1.2480037854143253, "grad_norm": 2.170522689819336, "learning_rate": 3.945723503392536e-06, "loss": 0.2819, "step": 21100 }, { "epoch": 1.2485952564026734, "grad_norm": 2.5800418853759766, "learning_rate": 3.943905342141548e-06, "loss": 0.2648, "step": 21110 }, { "epoch": 1.2491867273910215, "grad_norm": 3.112859010696411, "learning_rate": 3.942086796019175e-06, "loss": 0.244, "step": 21120 }, { "epoch": 1.2497781983793694, "grad_norm": 3.8415579795837402, "learning_rate": 3.940267865766917e-06, "loss": 0.298, "step": 21130 }, { "epoch": 1.2503696693677175, "grad_norm": 3.426593065261841, "learning_rate": 3.938448552126431e-06, "loss": 0.2817, "step": 21140 }, { "epoch": 1.2509611403560656, "grad_norm": 1.9889700412750244, "learning_rate": 3.936628855839531e-06, "loss": 0.2736, "step": 21150 }, { "epoch": 1.2515526113444135, "grad_norm": 4.387689590454102, "learning_rate": 3.934808777648186e-06, "loss": 0.2614, "step": 21160 }, { "epoch": 1.2521440823327616, "grad_norm": 3.0832467079162598, "learning_rate": 3.932988318294521e-06, "loss": 0.2692, "step": 21170 }, { "epoch": 1.2527355533211095, "grad_norm": 3.0305347442626953, "learning_rate": 3.931167478520817e-06, "loss": 0.357, "step": 21180 }, { "epoch": 1.2533270243094576, "grad_norm": 3.8542888164520264, "learning_rate": 3.929346259069512e-06, "loss": 0.272, "step": 21190 }, { "epoch": 1.2539184952978055, "grad_norm": 3.2654943466186523, "learning_rate": 3.927524660683192e-06, "loss": 0.2277, "step": 21200 }, { "epoch": 1.2545099662861536, "grad_norm": 3.2665648460388184, "learning_rate": 3.9257026841046056e-06, "loss": 0.3078, "step": 21210 }, { "epoch": 1.2551014372745017, "grad_norm": 4.333670616149902, "learning_rate": 3.923880330076652e-06, "loss": 0.2553, "step": 21220 }, { "epoch": 1.2556929082628496, "grad_norm": 6.1278910636901855, "learning_rate": 3.922057599342383e-06, "loss": 0.2819, "step": 21230 }, { "epoch": 1.2562843792511977, "grad_norm": 3.777696371078491, "learning_rate": 3.920234492645006e-06, "loss": 0.2397, "step": 21240 }, { "epoch": 1.2568758502395458, "grad_norm": 2.9328508377075195, "learning_rate": 3.91841101072788e-06, "loss": 0.2517, "step": 21250 }, { "epoch": 1.2574673212278937, "grad_norm": 2.7502729892730713, "learning_rate": 3.916587154334521e-06, "loss": 0.3292, "step": 21260 }, { "epoch": 1.2580587922162418, "grad_norm": 3.2547831535339355, "learning_rate": 3.914762924208593e-06, "loss": 0.2891, "step": 21270 }, { "epoch": 1.25865026320459, "grad_norm": 4.079752445220947, "learning_rate": 3.912938321093915e-06, "loss": 0.3143, "step": 21280 }, { "epoch": 1.2592417341929378, "grad_norm": 2.500560760498047, "learning_rate": 3.911113345734456e-06, "loss": 0.224, "step": 21290 }, { "epoch": 1.2598332051812857, "grad_norm": 2.2639212608337402, "learning_rate": 3.909287998874338e-06, "loss": 0.2371, "step": 21300 }, { "epoch": 1.2604246761696338, "grad_norm": 7.98052978515625, "learning_rate": 3.907462281257835e-06, "loss": 0.3178, "step": 21310 }, { "epoch": 1.261016147157982, "grad_norm": 6.584686756134033, "learning_rate": 3.905636193629373e-06, "loss": 0.3205, "step": 21320 }, { "epoch": 1.2616076181463298, "grad_norm": 1.8793987035751343, "learning_rate": 3.903809736733527e-06, "loss": 0.3009, "step": 21330 }, { "epoch": 1.262199089134678, "grad_norm": 2.123574733734131, "learning_rate": 3.901982911315021e-06, "loss": 0.2786, "step": 21340 }, { "epoch": 1.262790560123026, "grad_norm": 2.049868583679199, "learning_rate": 3.900155718118735e-06, "loss": 0.2959, "step": 21350 }, { "epoch": 1.263382031111374, "grad_norm": 2.5558571815490723, "learning_rate": 3.898328157889692e-06, "loss": 0.2577, "step": 21360 }, { "epoch": 1.263973502099722, "grad_norm": 4.744325160980225, "learning_rate": 3.896500231373071e-06, "loss": 0.3148, "step": 21370 }, { "epoch": 1.2645649730880701, "grad_norm": 2.235379457473755, "learning_rate": 3.894671939314195e-06, "loss": 0.3031, "step": 21380 }, { "epoch": 1.265156444076418, "grad_norm": 3.530984878540039, "learning_rate": 3.89284328245854e-06, "loss": 0.3064, "step": 21390 }, { "epoch": 1.2657479150647661, "grad_norm": 2.614243507385254, "learning_rate": 3.891014261551728e-06, "loss": 0.3308, "step": 21400 }, { "epoch": 1.266339386053114, "grad_norm": 2.6370081901550293, "learning_rate": 3.8891848773395315e-06, "loss": 0.3137, "step": 21410 }, { "epoch": 1.2669308570414621, "grad_norm": 5.0104780197143555, "learning_rate": 3.88735513056787e-06, "loss": 0.25, "step": 21420 }, { "epoch": 1.26752232802981, "grad_norm": 2.807814121246338, "learning_rate": 3.885525021982811e-06, "loss": 0.3401, "step": 21430 }, { "epoch": 1.2681137990181581, "grad_norm": 4.293516635894775, "learning_rate": 3.883694552330569e-06, "loss": 0.3528, "step": 21440 }, { "epoch": 1.2687052700065062, "grad_norm": 1.869515061378479, "learning_rate": 3.8818637223575076e-06, "loss": 0.2869, "step": 21450 }, { "epoch": 1.2692967409948541, "grad_norm": 1.85780668258667, "learning_rate": 3.880032532810135e-06, "loss": 0.2431, "step": 21460 }, { "epoch": 1.2698882119832022, "grad_norm": 3.640456438064575, "learning_rate": 3.878200984435107e-06, "loss": 0.2831, "step": 21470 }, { "epoch": 1.2704796829715503, "grad_norm": 2.369586944580078, "learning_rate": 3.876369077979228e-06, "loss": 0.2899, "step": 21480 }, { "epoch": 1.2710711539598982, "grad_norm": 5.658085823059082, "learning_rate": 3.874536814189443e-06, "loss": 0.2419, "step": 21490 }, { "epoch": 1.2716626249482463, "grad_norm": 3.2098231315612793, "learning_rate": 3.872704193812847e-06, "loss": 0.3565, "step": 21500 }, { "epoch": 1.2722540959365944, "grad_norm": 2.0693209171295166, "learning_rate": 3.870871217596681e-06, "loss": 0.2132, "step": 21510 }, { "epoch": 1.2728455669249423, "grad_norm": 2.9947829246520996, "learning_rate": 3.869037886288327e-06, "loss": 0.3274, "step": 21520 }, { "epoch": 1.2734370379132902, "grad_norm": 1.6450546979904175, "learning_rate": 3.8672042006353145e-06, "loss": 0.2685, "step": 21530 }, { "epoch": 1.2740285089016383, "grad_norm": 3.1976187229156494, "learning_rate": 3.865370161385319e-06, "loss": 0.2996, "step": 21540 }, { "epoch": 1.2746199798899864, "grad_norm": 2.650824785232544, "learning_rate": 3.8635357692861565e-06, "loss": 0.3042, "step": 21550 }, { "epoch": 1.2752114508783343, "grad_norm": 3.819765329360962, "learning_rate": 3.86170102508579e-06, "loss": 0.2856, "step": 21560 }, { "epoch": 1.2758029218666824, "grad_norm": 3.3977904319763184, "learning_rate": 3.8598659295323226e-06, "loss": 0.2903, "step": 21570 }, { "epoch": 1.2763943928550305, "grad_norm": 5.072076320648193, "learning_rate": 3.858030483374004e-06, "loss": 0.3025, "step": 21580 }, { "epoch": 1.2769858638433784, "grad_norm": 2.499817132949829, "learning_rate": 3.856194687359226e-06, "loss": 0.2244, "step": 21590 }, { "epoch": 1.2775773348317265, "grad_norm": 4.147136688232422, "learning_rate": 3.8543585422365234e-06, "loss": 0.3536, "step": 21600 }, { "epoch": 1.2781688058200746, "grad_norm": 3.648061513900757, "learning_rate": 3.852522048754572e-06, "loss": 0.3857, "step": 21610 }, { "epoch": 1.2787602768084225, "grad_norm": 2.4555137157440186, "learning_rate": 3.8506852076621905e-06, "loss": 0.291, "step": 21620 }, { "epoch": 1.2793517477967706, "grad_norm": 2.813035011291504, "learning_rate": 3.848848019708337e-06, "loss": 0.3256, "step": 21630 }, { "epoch": 1.2799432187851185, "grad_norm": 2.8028221130371094, "learning_rate": 3.847010485642116e-06, "loss": 0.2614, "step": 21640 }, { "epoch": 1.2805346897734666, "grad_norm": 3.6106221675872803, "learning_rate": 3.84517260621277e-06, "loss": 0.3495, "step": 21650 }, { "epoch": 1.2811261607618145, "grad_norm": 3.5101559162139893, "learning_rate": 3.843334382169681e-06, "loss": 0.3051, "step": 21660 }, { "epoch": 1.2817176317501626, "grad_norm": 3.908879518508911, "learning_rate": 3.841495814262374e-06, "loss": 0.2702, "step": 21670 }, { "epoch": 1.2823091027385107, "grad_norm": 2.400029182434082, "learning_rate": 3.839656903240513e-06, "loss": 0.3112, "step": 21680 }, { "epoch": 1.2829005737268586, "grad_norm": 2.0085830688476562, "learning_rate": 3.837817649853903e-06, "loss": 0.3002, "step": 21690 }, { "epoch": 1.2834920447152067, "grad_norm": 2.745906114578247, "learning_rate": 3.835978054852487e-06, "loss": 0.2604, "step": 21700 }, { "epoch": 1.2840835157035548, "grad_norm": 2.451981544494629, "learning_rate": 3.834138118986349e-06, "loss": 0.3262, "step": 21710 }, { "epoch": 1.2846749866919027, "grad_norm": 3.557530641555786, "learning_rate": 3.832297843005711e-06, "loss": 0.3202, "step": 21720 }, { "epoch": 1.2852664576802508, "grad_norm": 3.366117238998413, "learning_rate": 3.830457227660934e-06, "loss": 0.2398, "step": 21730 }, { "epoch": 1.285857928668599, "grad_norm": 3.8764920234680176, "learning_rate": 3.828616273702517e-06, "loss": 0.3127, "step": 21740 }, { "epoch": 1.2864493996569468, "grad_norm": 2.906102418899536, "learning_rate": 3.826774981881095e-06, "loss": 0.3224, "step": 21750 }, { "epoch": 1.2870408706452947, "grad_norm": 2.3216962814331055, "learning_rate": 3.8249333529474464e-06, "loss": 0.2802, "step": 21760 }, { "epoch": 1.2876323416336428, "grad_norm": 4.813936710357666, "learning_rate": 3.823091387652481e-06, "loss": 0.3608, "step": 21770 }, { "epoch": 1.288223812621991, "grad_norm": 3.182711124420166, "learning_rate": 3.82124908674725e-06, "loss": 0.3029, "step": 21780 }, { "epoch": 1.2888152836103388, "grad_norm": 2.873163938522339, "learning_rate": 3.81940645098294e-06, "loss": 0.316, "step": 21790 }, { "epoch": 1.289406754598687, "grad_norm": 2.6190385818481445, "learning_rate": 3.817563481110873e-06, "loss": 0.3542, "step": 21800 }, { "epoch": 1.289998225587035, "grad_norm": 4.860269546508789, "learning_rate": 3.815720177882509e-06, "loss": 0.2794, "step": 21810 }, { "epoch": 1.290589696575383, "grad_norm": 3.2856242656707764, "learning_rate": 3.8138765420494424e-06, "loss": 0.2235, "step": 21820 }, { "epoch": 1.291181167563731, "grad_norm": 4.217858791351318, "learning_rate": 3.8120325743634054e-06, "loss": 0.2699, "step": 21830 }, { "epoch": 1.2917726385520791, "grad_norm": 2.5087671279907227, "learning_rate": 3.8101882755762614e-06, "loss": 0.2849, "step": 21840 }, { "epoch": 1.292364109540427, "grad_norm": 5.778660774230957, "learning_rate": 3.808343646440014e-06, "loss": 0.3002, "step": 21850 }, { "epoch": 1.2929555805287751, "grad_norm": 3.4364278316497803, "learning_rate": 3.8064986877067976e-06, "loss": 0.307, "step": 21860 }, { "epoch": 1.293547051517123, "grad_norm": 2.752471923828125, "learning_rate": 3.8046534001288834e-06, "loss": 0.2689, "step": 21870 }, { "epoch": 1.2941385225054711, "grad_norm": 2.979738712310791, "learning_rate": 3.802807784458675e-06, "loss": 0.2894, "step": 21880 }, { "epoch": 1.294729993493819, "grad_norm": 2.2372138500213623, "learning_rate": 3.80096184144871e-06, "loss": 0.2501, "step": 21890 }, { "epoch": 1.2953214644821671, "grad_norm": 4.639530658721924, "learning_rate": 3.7991155718516603e-06, "loss": 0.2645, "step": 21900 }, { "epoch": 1.2959129354705152, "grad_norm": 3.6660993099212646, "learning_rate": 3.7972689764203305e-06, "loss": 0.2814, "step": 21910 }, { "epoch": 1.2965044064588631, "grad_norm": 3.5634775161743164, "learning_rate": 3.7954220559076587e-06, "loss": 0.2945, "step": 21920 }, { "epoch": 1.2970958774472112, "grad_norm": 3.326815605163574, "learning_rate": 3.7935748110667146e-06, "loss": 0.2756, "step": 21930 }, { "epoch": 1.2976873484355593, "grad_norm": 2.981699228286743, "learning_rate": 3.791727242650701e-06, "loss": 0.2716, "step": 21940 }, { "epoch": 1.2982788194239072, "grad_norm": 2.674781322479248, "learning_rate": 3.7898793514129507e-06, "loss": 0.2815, "step": 21950 }, { "epoch": 1.2988702904122553, "grad_norm": 5.285210609436035, "learning_rate": 3.78803113810693e-06, "loss": 0.3434, "step": 21960 }, { "epoch": 1.2994617614006034, "grad_norm": 2.579340696334839, "learning_rate": 3.786182603486237e-06, "loss": 0.2685, "step": 21970 }, { "epoch": 1.3000532323889513, "grad_norm": 2.907378673553467, "learning_rate": 3.7843337483045998e-06, "loss": 0.3195, "step": 21980 }, { "epoch": 1.3006447033772992, "grad_norm": 3.9031286239624023, "learning_rate": 3.782484573315877e-06, "loss": 0.2832, "step": 21990 }, { "epoch": 1.3012361743656473, "grad_norm": 3.367434501647949, "learning_rate": 3.780635079274057e-06, "loss": 0.2207, "step": 22000 }, { "epoch": 1.3018276453539954, "grad_norm": 4.000437259674072, "learning_rate": 3.778785266933261e-06, "loss": 0.3441, "step": 22010 }, { "epoch": 1.3024191163423433, "grad_norm": 3.865381956100464, "learning_rate": 3.7769351370477363e-06, "loss": 0.2607, "step": 22020 }, { "epoch": 1.3030105873306914, "grad_norm": 2.6188056468963623, "learning_rate": 3.775084690371863e-06, "loss": 0.3055, "step": 22030 }, { "epoch": 1.3036020583190395, "grad_norm": 3.063939332962036, "learning_rate": 3.7732339276601483e-06, "loss": 0.3388, "step": 22040 }, { "epoch": 1.3041935293073874, "grad_norm": 3.1212127208709717, "learning_rate": 3.77138284966723e-06, "loss": 0.279, "step": 22050 }, { "epoch": 1.3047850002957355, "grad_norm": 5.029014587402344, "learning_rate": 3.7695314571478718e-06, "loss": 0.2731, "step": 22060 }, { "epoch": 1.3053764712840836, "grad_norm": 4.77285623550415, "learning_rate": 3.7676797508569678e-06, "loss": 0.3318, "step": 22070 }, { "epoch": 1.3059679422724315, "grad_norm": 2.756594181060791, "learning_rate": 3.7658277315495392e-06, "loss": 0.2456, "step": 22080 }, { "epoch": 1.3065594132607796, "grad_norm": 2.550023317337036, "learning_rate": 3.763975399980735e-06, "loss": 0.2531, "step": 22090 }, { "epoch": 1.3071508842491275, "grad_norm": 3.0384979248046875, "learning_rate": 3.7621227569058326e-06, "loss": 0.3299, "step": 22100 }, { "epoch": 1.3077423552374756, "grad_norm": 4.583970546722412, "learning_rate": 3.7602698030802336e-06, "loss": 0.2473, "step": 22110 }, { "epoch": 1.3083338262258235, "grad_norm": 4.294337272644043, "learning_rate": 3.7584165392594706e-06, "loss": 0.2736, "step": 22120 }, { "epoch": 1.3089252972141716, "grad_norm": 2.3386988639831543, "learning_rate": 3.756562966199198e-06, "loss": 0.2711, "step": 22130 }, { "epoch": 1.3095167682025197, "grad_norm": 2.6501238346099854, "learning_rate": 3.754709084655199e-06, "loss": 0.2615, "step": 22140 }, { "epoch": 1.3101082391908676, "grad_norm": 3.4176464080810547, "learning_rate": 3.752854895383383e-06, "loss": 0.3295, "step": 22150 }, { "epoch": 1.3106997101792157, "grad_norm": 3.7159180641174316, "learning_rate": 3.751000399139782e-06, "loss": 0.2459, "step": 22160 }, { "epoch": 1.3112911811675638, "grad_norm": 2.925718069076538, "learning_rate": 3.7491455966805563e-06, "loss": 0.2988, "step": 22170 }, { "epoch": 1.3118826521559117, "grad_norm": 2.936918258666992, "learning_rate": 3.747290488761989e-06, "loss": 0.3345, "step": 22180 }, { "epoch": 1.3124741231442598, "grad_norm": 1.9616867303848267, "learning_rate": 3.74543507614049e-06, "loss": 0.2319, "step": 22190 }, { "epoch": 1.313065594132608, "grad_norm": 2.929548501968384, "learning_rate": 3.7435793595725903e-06, "loss": 0.2541, "step": 22200 }, { "epoch": 1.3136570651209558, "grad_norm": 2.2553255558013916, "learning_rate": 3.7417233398149462e-06, "loss": 0.2536, "step": 22210 }, { "epoch": 1.3142485361093037, "grad_norm": 5.058745861053467, "learning_rate": 3.73986701762434e-06, "loss": 0.3543, "step": 22220 }, { "epoch": 1.3148400070976518, "grad_norm": 4.706000804901123, "learning_rate": 3.738010393757674e-06, "loss": 0.3075, "step": 22230 }, { "epoch": 1.315431478086, "grad_norm": 2.941375970840454, "learning_rate": 3.7361534689719742e-06, "loss": 0.3516, "step": 22240 }, { "epoch": 1.3160229490743478, "grad_norm": 2.026909351348877, "learning_rate": 3.734296244024392e-06, "loss": 0.2701, "step": 22250 }, { "epoch": 1.316614420062696, "grad_norm": 4.363923072814941, "learning_rate": 3.7324387196721964e-06, "loss": 0.2783, "step": 22260 }, { "epoch": 1.317205891051044, "grad_norm": 2.4912502765655518, "learning_rate": 3.7305808966727822e-06, "loss": 0.3099, "step": 22270 }, { "epoch": 1.317797362039392, "grad_norm": 2.528306722640991, "learning_rate": 3.728722775783665e-06, "loss": 0.3558, "step": 22280 }, { "epoch": 1.31838883302774, "grad_norm": 1.9128966331481934, "learning_rate": 3.7268643577624817e-06, "loss": 0.3163, "step": 22290 }, { "epoch": 1.3189803040160881, "grad_norm": 2.6160943508148193, "learning_rate": 3.7250056433669907e-06, "loss": 0.2884, "step": 22300 }, { "epoch": 1.319571775004436, "grad_norm": 2.0996994972229004, "learning_rate": 3.72314663335507e-06, "loss": 0.2907, "step": 22310 }, { "epoch": 1.3201632459927841, "grad_norm": 2.878767728805542, "learning_rate": 3.72128732848472e-06, "loss": 0.3557, "step": 22320 }, { "epoch": 1.320754716981132, "grad_norm": 3.4718880653381348, "learning_rate": 3.7194277295140605e-06, "loss": 0.299, "step": 22330 }, { "epoch": 1.3213461879694801, "grad_norm": 3.1585848331451416, "learning_rate": 3.71756783720133e-06, "loss": 0.3252, "step": 22340 }, { "epoch": 1.321937658957828, "grad_norm": 2.412949323654175, "learning_rate": 3.715707652304889e-06, "loss": 0.2989, "step": 22350 }, { "epoch": 1.322529129946176, "grad_norm": 3.739098072052002, "learning_rate": 3.7138471755832144e-06, "loss": 0.2742, "step": 22360 }, { "epoch": 1.3231206009345242, "grad_norm": 4.154986381530762, "learning_rate": 3.7119864077949067e-06, "loss": 0.2598, "step": 22370 }, { "epoch": 1.323712071922872, "grad_norm": 2.8838348388671875, "learning_rate": 3.710125349698679e-06, "loss": 0.2793, "step": 22380 }, { "epoch": 1.3243035429112202, "grad_norm": 2.5399975776672363, "learning_rate": 3.708264002053368e-06, "loss": 0.2871, "step": 22390 }, { "epoch": 1.3248950138995683, "grad_norm": 2.41507887840271, "learning_rate": 3.7064023656179243e-06, "loss": 0.2793, "step": 22400 }, { "epoch": 1.3254864848879162, "grad_norm": 4.0087714195251465, "learning_rate": 3.70454044115142e-06, "loss": 0.3325, "step": 22410 }, { "epoch": 1.3260779558762643, "grad_norm": 1.5216121673583984, "learning_rate": 3.7026782294130423e-06, "loss": 0.2874, "step": 22420 }, { "epoch": 1.3266694268646124, "grad_norm": 3.551537275314331, "learning_rate": 3.7008157311620956e-06, "loss": 0.3971, "step": 22430 }, { "epoch": 1.3272608978529603, "grad_norm": 2.441359043121338, "learning_rate": 3.6989529471580036e-06, "loss": 0.3186, "step": 22440 }, { "epoch": 1.3278523688413084, "grad_norm": 3.846930742263794, "learning_rate": 3.6970898781603023e-06, "loss": 0.3291, "step": 22450 }, { "epoch": 1.3284438398296563, "grad_norm": 2.529151439666748, "learning_rate": 3.695226524928647e-06, "loss": 0.27, "step": 22460 }, { "epoch": 1.3290353108180044, "grad_norm": 3.046563148498535, "learning_rate": 3.693362888222809e-06, "loss": 0.3051, "step": 22470 }, { "epoch": 1.3296267818063523, "grad_norm": 3.172122001647949, "learning_rate": 3.6914989688026736e-06, "loss": 0.2965, "step": 22480 }, { "epoch": 1.3302182527947004, "grad_norm": 2.377866506576538, "learning_rate": 3.689634767428242e-06, "loss": 0.3065, "step": 22490 }, { "epoch": 1.3308097237830485, "grad_norm": 1.9088189601898193, "learning_rate": 3.68777028485963e-06, "loss": 0.3091, "step": 22500 }, { "epoch": 1.3314011947713964, "grad_norm": 4.123067855834961, "learning_rate": 3.6859055218570696e-06, "loss": 0.3567, "step": 22510 }, { "epoch": 1.3319926657597445, "grad_norm": 2.7718167304992676, "learning_rate": 3.6840404791809053e-06, "loss": 0.3088, "step": 22520 }, { "epoch": 1.3325841367480926, "grad_norm": 3.529362201690674, "learning_rate": 3.682175157591596e-06, "loss": 0.2671, "step": 22530 }, { "epoch": 1.3331756077364405, "grad_norm": 2.5574049949645996, "learning_rate": 3.6803095578497146e-06, "loss": 0.304, "step": 22540 }, { "epoch": 1.3337670787247886, "grad_norm": 2.940222978591919, "learning_rate": 3.678443680715949e-06, "loss": 0.3005, "step": 22550 }, { "epoch": 1.3343585497131365, "grad_norm": 2.120893716812134, "learning_rate": 3.676577526951097e-06, "loss": 0.2911, "step": 22560 }, { "epoch": 1.3349500207014846, "grad_norm": 3.2421882152557373, "learning_rate": 3.674711097316073e-06, "loss": 0.2674, "step": 22570 }, { "epoch": 1.3355414916898325, "grad_norm": 7.643344879150391, "learning_rate": 3.6728443925718996e-06, "loss": 0.346, "step": 22580 }, { "epoch": 1.3361329626781806, "grad_norm": 5.303382873535156, "learning_rate": 3.6709774134797148e-06, "loss": 0.3081, "step": 22590 }, { "epoch": 1.3367244336665287, "grad_norm": 3.5546212196350098, "learning_rate": 3.669110160800767e-06, "loss": 0.3041, "step": 22600 }, { "epoch": 1.3373159046548766, "grad_norm": 3.9058494567871094, "learning_rate": 3.6672426352964166e-06, "loss": 0.2802, "step": 22610 }, { "epoch": 1.3379073756432247, "grad_norm": 2.0741171836853027, "learning_rate": 3.665374837728135e-06, "loss": 0.283, "step": 22620 }, { "epoch": 1.3384988466315728, "grad_norm": 3.357982635498047, "learning_rate": 3.6635067688575062e-06, "loss": 0.2623, "step": 22630 }, { "epoch": 1.3390903176199207, "grad_norm": 4.587502956390381, "learning_rate": 3.6616384294462222e-06, "loss": 0.3029, "step": 22640 }, { "epoch": 1.3396817886082688, "grad_norm": 3.810856819152832, "learning_rate": 3.659769820256086e-06, "loss": 0.3301, "step": 22650 }, { "epoch": 1.340273259596617, "grad_norm": 1.9686088562011719, "learning_rate": 3.657900942049012e-06, "loss": 0.2617, "step": 22660 }, { "epoch": 1.3408647305849648, "grad_norm": 3.4506256580352783, "learning_rate": 3.656031795587023e-06, "loss": 0.2792, "step": 22670 }, { "epoch": 1.341456201573313, "grad_norm": 2.386646270751953, "learning_rate": 3.654162381632251e-06, "loss": 0.2791, "step": 22680 }, { "epoch": 1.3420476725616608, "grad_norm": 4.156431198120117, "learning_rate": 3.6522927009469404e-06, "loss": 0.3785, "step": 22690 }, { "epoch": 1.342639143550009, "grad_norm": 2.9131693840026855, "learning_rate": 3.650422754293437e-06, "loss": 0.3188, "step": 22700 }, { "epoch": 1.3432306145383568, "grad_norm": 3.122441053390503, "learning_rate": 3.648552542434204e-06, "loss": 0.3042, "step": 22710 }, { "epoch": 1.343822085526705, "grad_norm": 5.3659443855285645, "learning_rate": 3.646682066131805e-06, "loss": 0.3423, "step": 22720 }, { "epoch": 1.344413556515053, "grad_norm": 2.0462825298309326, "learning_rate": 3.644811326148917e-06, "loss": 0.2609, "step": 22730 }, { "epoch": 1.345005027503401, "grad_norm": 3.077315092086792, "learning_rate": 3.642940323248321e-06, "loss": 0.2881, "step": 22740 }, { "epoch": 1.345596498491749, "grad_norm": 3.8373465538024902, "learning_rate": 3.641069058192907e-06, "loss": 0.2616, "step": 22750 }, { "epoch": 1.3461879694800971, "grad_norm": 2.5266363620758057, "learning_rate": 3.639197531745672e-06, "loss": 0.2877, "step": 22760 }, { "epoch": 1.346779440468445, "grad_norm": 2.2643446922302246, "learning_rate": 3.6373257446697177e-06, "loss": 0.3003, "step": 22770 }, { "epoch": 1.347370911456793, "grad_norm": 2.0503945350646973, "learning_rate": 3.635453697728254e-06, "loss": 0.2599, "step": 22780 }, { "epoch": 1.347962382445141, "grad_norm": 4.253026485443115, "learning_rate": 3.6335813916845964e-06, "loss": 0.2986, "step": 22790 }, { "epoch": 1.348553853433489, "grad_norm": 3.1561009883880615, "learning_rate": 3.6317088273021665e-06, "loss": 0.2786, "step": 22800 }, { "epoch": 1.349145324421837, "grad_norm": 4.438661575317383, "learning_rate": 3.6298360053444887e-06, "loss": 0.2787, "step": 22810 }, { "epoch": 1.349736795410185, "grad_norm": 2.267580270767212, "learning_rate": 3.627962926575194e-06, "loss": 0.3095, "step": 22820 }, { "epoch": 1.3503282663985332, "grad_norm": 3.4157958030700684, "learning_rate": 3.6260895917580207e-06, "loss": 0.3499, "step": 22830 }, { "epoch": 1.350919737386881, "grad_norm": 2.7633323669433594, "learning_rate": 3.6242160016568074e-06, "loss": 0.3393, "step": 22840 }, { "epoch": 1.3515112083752292, "grad_norm": 2.4705123901367188, "learning_rate": 3.6223421570354987e-06, "loss": 0.2496, "step": 22850 }, { "epoch": 1.3521026793635773, "grad_norm": 5.0784430503845215, "learning_rate": 3.6204680586581423e-06, "loss": 0.3129, "step": 22860 }, { "epoch": 1.3526941503519252, "grad_norm": 2.9196112155914307, "learning_rate": 3.6185937072888917e-06, "loss": 0.2976, "step": 22870 }, { "epoch": 1.3532856213402733, "grad_norm": 4.817223072052002, "learning_rate": 3.6167191036919996e-06, "loss": 0.2956, "step": 22880 }, { "epoch": 1.3538770923286214, "grad_norm": 3.4544837474823, "learning_rate": 3.6148442486318257e-06, "loss": 0.2936, "step": 22890 }, { "epoch": 1.3544685633169693, "grad_norm": 3.6580164432525635, "learning_rate": 3.6129691428728285e-06, "loss": 0.3185, "step": 22900 }, { "epoch": 1.3550600343053174, "grad_norm": 3.951563835144043, "learning_rate": 3.6110937871795715e-06, "loss": 0.2707, "step": 22910 }, { "epoch": 1.3556515052936653, "grad_norm": 2.135103225708008, "learning_rate": 3.6092181823167177e-06, "loss": 0.3005, "step": 22920 }, { "epoch": 1.3562429762820134, "grad_norm": 3.126335382461548, "learning_rate": 3.607342329049034e-06, "loss": 0.3133, "step": 22930 }, { "epoch": 1.3568344472703613, "grad_norm": 3.696439266204834, "learning_rate": 3.6054662281413884e-06, "loss": 0.3244, "step": 22940 }, { "epoch": 1.3574259182587094, "grad_norm": 2.477107524871826, "learning_rate": 3.603589880358747e-06, "loss": 0.2861, "step": 22950 }, { "epoch": 1.3580173892470575, "grad_norm": 2.1409316062927246, "learning_rate": 3.601713286466181e-06, "loss": 0.2903, "step": 22960 }, { "epoch": 1.3586088602354054, "grad_norm": 3.660029411315918, "learning_rate": 3.5998364472288577e-06, "loss": 0.3199, "step": 22970 }, { "epoch": 1.3592003312237535, "grad_norm": 2.7576258182525635, "learning_rate": 3.5979593634120474e-06, "loss": 0.2837, "step": 22980 }, { "epoch": 1.3597918022121016, "grad_norm": 2.1882920265197754, "learning_rate": 3.5960820357811185e-06, "loss": 0.2742, "step": 22990 }, { "epoch": 1.3603832732004495, "grad_norm": 2.8169257640838623, "learning_rate": 3.594204465101539e-06, "loss": 0.2313, "step": 23000 }, { "epoch": 1.3609747441887976, "grad_norm": 4.394028663635254, "learning_rate": 3.5923266521388787e-06, "loss": 0.2744, "step": 23010 }, { "epoch": 1.3615662151771455, "grad_norm": 3.2297236919403076, "learning_rate": 3.590448597658801e-06, "loss": 0.2623, "step": 23020 }, { "epoch": 1.3621576861654936, "grad_norm": 2.2910640239715576, "learning_rate": 3.588570302427072e-06, "loss": 0.3483, "step": 23030 }, { "epoch": 1.3627491571538415, "grad_norm": 3.072711944580078, "learning_rate": 3.586691767209555e-06, "loss": 0.2672, "step": 23040 }, { "epoch": 1.3633406281421896, "grad_norm": 4.047763824462891, "learning_rate": 3.584812992772209e-06, "loss": 0.269, "step": 23050 }, { "epoch": 1.3639320991305377, "grad_norm": 3.167573928833008, "learning_rate": 3.5829339798810936e-06, "loss": 0.3439, "step": 23060 }, { "epoch": 1.3645235701188856, "grad_norm": 6.05312967300415, "learning_rate": 3.581054729302364e-06, "loss": 0.3015, "step": 23070 }, { "epoch": 1.3651150411072337, "grad_norm": 4.076580047607422, "learning_rate": 3.5791752418022734e-06, "loss": 0.2901, "step": 23080 }, { "epoch": 1.3657065120955818, "grad_norm": 2.7116143703460693, "learning_rate": 3.577295518147171e-06, "loss": 0.3118, "step": 23090 }, { "epoch": 1.3662979830839297, "grad_norm": 3.9120914936065674, "learning_rate": 3.5754155591035e-06, "loss": 0.3155, "step": 23100 }, { "epoch": 1.3668894540722778, "grad_norm": 2.38331937789917, "learning_rate": 3.573535365437804e-06, "loss": 0.3165, "step": 23110 }, { "epoch": 1.367480925060626, "grad_norm": 2.8862955570220947, "learning_rate": 3.571654937916719e-06, "loss": 0.2482, "step": 23120 }, { "epoch": 1.3680723960489738, "grad_norm": 3.563281297683716, "learning_rate": 3.569774277306977e-06, "loss": 0.3662, "step": 23130 }, { "epoch": 1.368663867037322, "grad_norm": 2.8603341579437256, "learning_rate": 3.567893384375406e-06, "loss": 0.2625, "step": 23140 }, { "epoch": 1.3692553380256698, "grad_norm": 4.3483452796936035, "learning_rate": 3.5660122598889276e-06, "loss": 0.298, "step": 23150 }, { "epoch": 1.369846809014018, "grad_norm": 2.688302993774414, "learning_rate": 3.5641309046145594e-06, "loss": 0.2616, "step": 23160 }, { "epoch": 1.3704382800023658, "grad_norm": 2.425929307937622, "learning_rate": 3.5622493193194104e-06, "loss": 0.2499, "step": 23170 }, { "epoch": 1.371029750990714, "grad_norm": 4.0015974044799805, "learning_rate": 3.560367504770686e-06, "loss": 0.2292, "step": 23180 }, { "epoch": 1.371621221979062, "grad_norm": 7.751138210296631, "learning_rate": 3.558485461735685e-06, "loss": 0.3362, "step": 23190 }, { "epoch": 1.3722126929674099, "grad_norm": 5.0387187004089355, "learning_rate": 3.556603190981796e-06, "loss": 0.2862, "step": 23200 }, { "epoch": 1.372804163955758, "grad_norm": 3.9164981842041016, "learning_rate": 3.5547206932765055e-06, "loss": 0.2595, "step": 23210 }, { "epoch": 1.373395634944106, "grad_norm": 2.8991079330444336, "learning_rate": 3.552837969387388e-06, "loss": 0.2789, "step": 23220 }, { "epoch": 1.373987105932454, "grad_norm": 2.811629295349121, "learning_rate": 3.550955020082113e-06, "loss": 0.2853, "step": 23230 }, { "epoch": 1.374578576920802, "grad_norm": 3.1573474407196045, "learning_rate": 3.549071846128441e-06, "loss": 0.305, "step": 23240 }, { "epoch": 1.37517004790915, "grad_norm": 2.673616886138916, "learning_rate": 3.547188448294224e-06, "loss": 0.282, "step": 23250 }, { "epoch": 1.375761518897498, "grad_norm": 2.4725401401519775, "learning_rate": 3.545304827347406e-06, "loss": 0.2949, "step": 23260 }, { "epoch": 1.376352989885846, "grad_norm": 3.581662178039551, "learning_rate": 3.5434209840560203e-06, "loss": 0.2881, "step": 23270 }, { "epoch": 1.376944460874194, "grad_norm": 3.0942764282226562, "learning_rate": 3.541536919188194e-06, "loss": 0.3352, "step": 23280 }, { "epoch": 1.3775359318625422, "grad_norm": 2.603532075881958, "learning_rate": 3.53965263351214e-06, "loss": 0.3214, "step": 23290 }, { "epoch": 1.37812740285089, "grad_norm": 2.499866008758545, "learning_rate": 3.5377681277961647e-06, "loss": 0.2713, "step": 23300 }, { "epoch": 1.3787188738392382, "grad_norm": 3.000901937484741, "learning_rate": 3.5358834028086645e-06, "loss": 0.2987, "step": 23310 }, { "epoch": 1.3793103448275863, "grad_norm": 4.450767993927002, "learning_rate": 3.5339984593181227e-06, "loss": 0.3529, "step": 23320 }, { "epoch": 1.3799018158159342, "grad_norm": 2.5086233615875244, "learning_rate": 3.5321132980931133e-06, "loss": 0.2793, "step": 23330 }, { "epoch": 1.3804932868042823, "grad_norm": 3.0932140350341797, "learning_rate": 3.5302279199022996e-06, "loss": 0.2661, "step": 23340 }, { "epoch": 1.3810847577926304, "grad_norm": 3.024855613708496, "learning_rate": 3.5283423255144326e-06, "loss": 0.3161, "step": 23350 }, { "epoch": 1.3816762287809783, "grad_norm": 3.725867986679077, "learning_rate": 3.5264565156983494e-06, "loss": 0.2909, "step": 23360 }, { "epoch": 1.3822676997693264, "grad_norm": 2.6003992557525635, "learning_rate": 3.5245704912229775e-06, "loss": 0.2918, "step": 23370 }, { "epoch": 1.3828591707576743, "grad_norm": 1.9379987716674805, "learning_rate": 3.5226842528573335e-06, "loss": 0.2312, "step": 23380 }, { "epoch": 1.3834506417460224, "grad_norm": 2.9778971672058105, "learning_rate": 3.520797801370516e-06, "loss": 0.2924, "step": 23390 }, { "epoch": 1.3840421127343703, "grad_norm": 4.030172824859619, "learning_rate": 3.518911137531716e-06, "loss": 0.321, "step": 23400 }, { "epoch": 1.3846335837227184, "grad_norm": 3.651186943054199, "learning_rate": 3.517024262110208e-06, "loss": 0.3101, "step": 23410 }, { "epoch": 1.3852250547110665, "grad_norm": 3.018885850906372, "learning_rate": 3.5151371758753534e-06, "loss": 0.2761, "step": 23420 }, { "epoch": 1.3858165256994144, "grad_norm": 2.9568378925323486, "learning_rate": 3.5132498795965986e-06, "loss": 0.335, "step": 23430 }, { "epoch": 1.3864079966877625, "grad_norm": 2.624765157699585, "learning_rate": 3.5113623740434787e-06, "loss": 0.2705, "step": 23440 }, { "epoch": 1.3869994676761106, "grad_norm": 1.6034772396087646, "learning_rate": 3.509474659985609e-06, "loss": 0.341, "step": 23450 }, { "epoch": 1.3875909386644585, "grad_norm": 2.9370994567871094, "learning_rate": 3.507586738192696e-06, "loss": 0.2571, "step": 23460 }, { "epoch": 1.3881824096528066, "grad_norm": 2.905796527862549, "learning_rate": 3.5056986094345255e-06, "loss": 0.3695, "step": 23470 }, { "epoch": 1.3887738806411545, "grad_norm": 2.6796176433563232, "learning_rate": 3.503810274480972e-06, "loss": 0.2794, "step": 23480 }, { "epoch": 1.3893653516295026, "grad_norm": 2.618013620376587, "learning_rate": 3.50192173410199e-06, "loss": 0.3147, "step": 23490 }, { "epoch": 1.3899568226178505, "grad_norm": 3.1757397651672363, "learning_rate": 3.5000329890676204e-06, "loss": 0.2296, "step": 23500 }, { "epoch": 1.3905482936061986, "grad_norm": 3.8114967346191406, "learning_rate": 3.498144040147988e-06, "loss": 0.358, "step": 23510 }, { "epoch": 1.3911397645945467, "grad_norm": 2.6287240982055664, "learning_rate": 3.4962548881132984e-06, "loss": 0.3026, "step": 23520 }, { "epoch": 1.3917312355828946, "grad_norm": 4.042983531951904, "learning_rate": 3.4943655337338433e-06, "loss": 0.3383, "step": 23530 }, { "epoch": 1.3923227065712427, "grad_norm": 4.259223461151123, "learning_rate": 3.492475977779992e-06, "loss": 0.3205, "step": 23540 }, { "epoch": 1.3929141775595908, "grad_norm": 2.5575482845306396, "learning_rate": 3.490586221022201e-06, "loss": 0.3144, "step": 23550 }, { "epoch": 1.3935056485479387, "grad_norm": 3.546043634414673, "learning_rate": 3.488696264231006e-06, "loss": 0.2325, "step": 23560 }, { "epoch": 1.3940971195362868, "grad_norm": 2.0744524002075195, "learning_rate": 3.4868061081770247e-06, "loss": 0.2872, "step": 23570 }, { "epoch": 1.394688590524635, "grad_norm": 3.3039872646331787, "learning_rate": 3.484915753630956e-06, "loss": 0.3143, "step": 23580 }, { "epoch": 1.3952800615129828, "grad_norm": 3.19909930229187, "learning_rate": 3.4830252013635793e-06, "loss": 0.3121, "step": 23590 }, { "epoch": 1.395871532501331, "grad_norm": 2.9345788955688477, "learning_rate": 3.4811344521457575e-06, "loss": 0.2403, "step": 23600 }, { "epoch": 1.3964630034896788, "grad_norm": 3.2892870903015137, "learning_rate": 3.47924350674843e-06, "loss": 0.338, "step": 23610 }, { "epoch": 1.3970544744780269, "grad_norm": 3.0093865394592285, "learning_rate": 3.477352365942617e-06, "loss": 0.3086, "step": 23620 }, { "epoch": 1.3976459454663748, "grad_norm": 2.750732183456421, "learning_rate": 3.47546103049942e-06, "loss": 0.2837, "step": 23630 }, { "epoch": 1.3982374164547229, "grad_norm": 2.5504019260406494, "learning_rate": 3.473569501190019e-06, "loss": 0.2815, "step": 23640 }, { "epoch": 1.398828887443071, "grad_norm": 4.193896293640137, "learning_rate": 3.4716777787856737e-06, "loss": 0.3066, "step": 23650 }, { "epoch": 1.3994203584314189, "grad_norm": 2.182359218597412, "learning_rate": 3.469785864057721e-06, "loss": 0.3298, "step": 23660 }, { "epoch": 1.400011829419767, "grad_norm": 3.227673292160034, "learning_rate": 3.4678937577775775e-06, "loss": 0.3404, "step": 23670 }, { "epoch": 1.400603300408115, "grad_norm": 3.3914074897766113, "learning_rate": 3.466001460716736e-06, "loss": 0.3102, "step": 23680 }, { "epoch": 1.401194771396463, "grad_norm": 1.9398612976074219, "learning_rate": 3.46410897364677e-06, "loss": 0.2951, "step": 23690 }, { "epoch": 1.401786242384811, "grad_norm": 2.7833731174468994, "learning_rate": 3.4622162973393292e-06, "loss": 0.3193, "step": 23700 }, { "epoch": 1.402377713373159, "grad_norm": 2.616562604904175, "learning_rate": 3.4603234325661386e-06, "loss": 0.2949, "step": 23710 }, { "epoch": 1.402969184361507, "grad_norm": 2.311020851135254, "learning_rate": 3.458430380099005e-06, "loss": 0.2873, "step": 23720 }, { "epoch": 1.403560655349855, "grad_norm": 2.5328657627105713, "learning_rate": 3.4565371407098055e-06, "loss": 0.3218, "step": 23730 }, { "epoch": 1.404152126338203, "grad_norm": 2.501277446746826, "learning_rate": 3.454643715170497e-06, "loss": 0.2496, "step": 23740 }, { "epoch": 1.4047435973265512, "grad_norm": 5.135712623596191, "learning_rate": 3.452750104253112e-06, "loss": 0.3127, "step": 23750 }, { "epoch": 1.405335068314899, "grad_norm": 3.9131999015808105, "learning_rate": 3.4508563087297585e-06, "loss": 0.3257, "step": 23760 }, { "epoch": 1.4059265393032472, "grad_norm": 3.581350564956665, "learning_rate": 3.4489623293726203e-06, "loss": 0.2649, "step": 23770 }, { "epoch": 1.4065180102915953, "grad_norm": 3.609428882598877, "learning_rate": 3.447068166953954e-06, "loss": 0.2694, "step": 23780 }, { "epoch": 1.4071094812799432, "grad_norm": 3.108393907546997, "learning_rate": 3.4451738222460916e-06, "loss": 0.3267, "step": 23790 }, { "epoch": 1.4077009522682913, "grad_norm": 2.687380790710449, "learning_rate": 3.443279296021443e-06, "loss": 0.259, "step": 23800 }, { "epoch": 1.4082924232566394, "grad_norm": 3.9408364295959473, "learning_rate": 3.441384589052486e-06, "loss": 0.3318, "step": 23810 }, { "epoch": 1.4088838942449873, "grad_norm": 2.908442497253418, "learning_rate": 3.439489702111777e-06, "loss": 0.3033, "step": 23820 }, { "epoch": 1.4094753652333354, "grad_norm": 2.7655560970306396, "learning_rate": 3.437594635971944e-06, "loss": 0.2695, "step": 23830 }, { "epoch": 1.4100668362216833, "grad_norm": 3.4648916721343994, "learning_rate": 3.4356993914056877e-06, "loss": 0.2654, "step": 23840 }, { "epoch": 1.4106583072100314, "grad_norm": 4.475517749786377, "learning_rate": 3.433803969185782e-06, "loss": 0.3437, "step": 23850 }, { "epoch": 1.4112497781983793, "grad_norm": 3.9880638122558594, "learning_rate": 3.431908370085074e-06, "loss": 0.2495, "step": 23860 }, { "epoch": 1.4118412491867274, "grad_norm": 3.002122402191162, "learning_rate": 3.4300125948764817e-06, "loss": 0.2628, "step": 23870 }, { "epoch": 1.4124327201750755, "grad_norm": 4.649486541748047, "learning_rate": 3.4281166443329946e-06, "loss": 0.3473, "step": 23880 }, { "epoch": 1.4130241911634234, "grad_norm": 4.467259407043457, "learning_rate": 3.4262205192276756e-06, "loss": 0.2361, "step": 23890 }, { "epoch": 1.4136156621517715, "grad_norm": 2.858870029449463, "learning_rate": 3.4243242203336566e-06, "loss": 0.2927, "step": 23900 }, { "epoch": 1.4142071331401196, "grad_norm": 2.1929268836975098, "learning_rate": 3.422427748424142e-06, "loss": 0.2936, "step": 23910 }, { "epoch": 1.4147986041284675, "grad_norm": 4.640931606292725, "learning_rate": 3.420531104272407e-06, "loss": 0.3039, "step": 23920 }, { "epoch": 1.4153900751168156, "grad_norm": 2.1518630981445312, "learning_rate": 3.418634288651793e-06, "loss": 0.2767, "step": 23930 }, { "epoch": 1.4159815461051635, "grad_norm": 4.90189790725708, "learning_rate": 3.4167373023357184e-06, "loss": 0.2937, "step": 23940 }, { "epoch": 1.4165730170935116, "grad_norm": 3.8895742893218994, "learning_rate": 3.4148401460976646e-06, "loss": 0.2792, "step": 23950 }, { "epoch": 1.4171644880818595, "grad_norm": 3.240851879119873, "learning_rate": 3.412942820711186e-06, "loss": 0.2769, "step": 23960 }, { "epoch": 1.4177559590702076, "grad_norm": 2.027639150619507, "learning_rate": 3.411045326949905e-06, "loss": 0.2555, "step": 23970 }, { "epoch": 1.4183474300585557, "grad_norm": 2.8879992961883545, "learning_rate": 3.409147665587513e-06, "loss": 0.3389, "step": 23980 }, { "epoch": 1.4189389010469036, "grad_norm": 3.4499967098236084, "learning_rate": 3.4072498373977684e-06, "loss": 0.2468, "step": 23990 }, { "epoch": 1.4195303720352517, "grad_norm": 3.3029332160949707, "learning_rate": 3.4053518431544988e-06, "loss": 0.285, "step": 24000 }, { "epoch": 1.4201218430235998, "grad_norm": 3.3327255249023438, "learning_rate": 3.403453683631599e-06, "loss": 0.3188, "step": 24010 }, { "epoch": 1.4207133140119477, "grad_norm": 2.280890703201294, "learning_rate": 3.4015553596030317e-06, "loss": 0.2396, "step": 24020 }, { "epoch": 1.4213047850002958, "grad_norm": 1.9334347248077393, "learning_rate": 3.399656871842827e-06, "loss": 0.2797, "step": 24030 }, { "epoch": 1.4218962559886439, "grad_norm": 1.5013970136642456, "learning_rate": 3.39775822112508e-06, "loss": 0.2811, "step": 24040 }, { "epoch": 1.4224877269769918, "grad_norm": 4.488505840301514, "learning_rate": 3.395859408223955e-06, "loss": 0.2887, "step": 24050 }, { "epoch": 1.4230791979653399, "grad_norm": 2.712599515914917, "learning_rate": 3.3939604339136796e-06, "loss": 0.3116, "step": 24060 }, { "epoch": 1.4236706689536878, "grad_norm": 3.878270387649536, "learning_rate": 3.392061298968548e-06, "loss": 0.2795, "step": 24070 }, { "epoch": 1.4242621399420359, "grad_norm": 1.8882231712341309, "learning_rate": 3.3901620041629217e-06, "loss": 0.3399, "step": 24080 }, { "epoch": 1.4248536109303838, "grad_norm": 3.24355149269104, "learning_rate": 3.3882625502712264e-06, "loss": 0.3377, "step": 24090 }, { "epoch": 1.4254450819187319, "grad_norm": 2.965482473373413, "learning_rate": 3.386362938067951e-06, "loss": 0.3388, "step": 24100 }, { "epoch": 1.42603655290708, "grad_norm": 3.896446704864502, "learning_rate": 3.38446316832765e-06, "loss": 0.338, "step": 24110 }, { "epoch": 1.4266280238954279, "grad_norm": 2.4138667583465576, "learning_rate": 3.382563241824944e-06, "loss": 0.2566, "step": 24120 }, { "epoch": 1.427219494883776, "grad_norm": 2.9879329204559326, "learning_rate": 3.3806631593345144e-06, "loss": 0.3149, "step": 24130 }, { "epoch": 1.427810965872124, "grad_norm": 3.659269094467163, "learning_rate": 3.3787629216311084e-06, "loss": 0.3305, "step": 24140 }, { "epoch": 1.428402436860472, "grad_norm": 2.3761813640594482, "learning_rate": 3.3768625294895356e-06, "loss": 0.2827, "step": 24150 }, { "epoch": 1.42899390784882, "grad_norm": 2.653759479522705, "learning_rate": 3.374961983684669e-06, "loss": 0.3098, "step": 24160 }, { "epoch": 1.429585378837168, "grad_norm": 2.8180007934570312, "learning_rate": 3.373061284991444e-06, "loss": 0.3027, "step": 24170 }, { "epoch": 1.430176849825516, "grad_norm": 3.6903202533721924, "learning_rate": 3.371160434184859e-06, "loss": 0.2565, "step": 24180 }, { "epoch": 1.430768320813864, "grad_norm": 4.086744785308838, "learning_rate": 3.3692594320399735e-06, "loss": 0.3251, "step": 24190 }, { "epoch": 1.431359791802212, "grad_norm": 5.361542701721191, "learning_rate": 3.3673582793319074e-06, "loss": 0.2654, "step": 24200 }, { "epoch": 1.4319512627905602, "grad_norm": 4.204805374145508, "learning_rate": 3.3654569768358453e-06, "loss": 0.2541, "step": 24210 }, { "epoch": 1.432542733778908, "grad_norm": 1.476047158241272, "learning_rate": 3.3635555253270307e-06, "loss": 0.2705, "step": 24220 }, { "epoch": 1.4331342047672562, "grad_norm": 4.319033145904541, "learning_rate": 3.3616539255807685e-06, "loss": 0.2688, "step": 24230 }, { "epoch": 1.4337256757556043, "grad_norm": 4.236924648284912, "learning_rate": 3.3597521783724234e-06, "loss": 0.31, "step": 24240 }, { "epoch": 1.4343171467439522, "grad_norm": 5.594067573547363, "learning_rate": 3.3578502844774214e-06, "loss": 0.3631, "step": 24250 }, { "epoch": 1.4349086177323003, "grad_norm": 2.279604196548462, "learning_rate": 3.355948244671247e-06, "loss": 0.3696, "step": 24260 }, { "epoch": 1.4355000887206484, "grad_norm": 3.400949001312256, "learning_rate": 3.3540460597294448e-06, "loss": 0.2381, "step": 24270 }, { "epoch": 1.4360915597089963, "grad_norm": 4.588635444641113, "learning_rate": 3.3521437304276196e-06, "loss": 0.2949, "step": 24280 }, { "epoch": 1.4366830306973444, "grad_norm": 2.9724364280700684, "learning_rate": 3.3502412575414325e-06, "loss": 0.242, "step": 24290 }, { "epoch": 1.4372745016856923, "grad_norm": 3.463711977005005, "learning_rate": 3.348338641846606e-06, "loss": 0.3179, "step": 24300 }, { "epoch": 1.4378659726740404, "grad_norm": 3.0538134574890137, "learning_rate": 3.3464358841189188e-06, "loss": 0.2433, "step": 24310 }, { "epoch": 1.4384574436623883, "grad_norm": 4.499630928039551, "learning_rate": 3.344532985134209e-06, "loss": 0.3128, "step": 24320 }, { "epoch": 1.4390489146507364, "grad_norm": 1.86431884765625, "learning_rate": 3.34262994566837e-06, "loss": 0.2753, "step": 24330 }, { "epoch": 1.4396403856390845, "grad_norm": 2.0793840885162354, "learning_rate": 3.3407267664973557e-06, "loss": 0.2631, "step": 24340 }, { "epoch": 1.4402318566274324, "grad_norm": 3.4913864135742188, "learning_rate": 3.3388234483971742e-06, "loss": 0.2383, "step": 24350 }, { "epoch": 1.4408233276157805, "grad_norm": 2.980727195739746, "learning_rate": 3.3369199921438913e-06, "loss": 0.3296, "step": 24360 }, { "epoch": 1.4414147986041286, "grad_norm": 4.388894557952881, "learning_rate": 3.33501639851363e-06, "loss": 0.2695, "step": 24370 }, { "epoch": 1.4420062695924765, "grad_norm": 1.8671890497207642, "learning_rate": 3.3331126682825678e-06, "loss": 0.3049, "step": 24380 }, { "epoch": 1.4425977405808246, "grad_norm": 2.478510856628418, "learning_rate": 3.331208802226939e-06, "loss": 0.309, "step": 24390 }, { "epoch": 1.4431892115691725, "grad_norm": 2.4237418174743652, "learning_rate": 3.3293048011230315e-06, "loss": 0.3266, "step": 24400 }, { "epoch": 1.4437806825575206, "grad_norm": 3.300830841064453, "learning_rate": 3.327400665747192e-06, "loss": 0.3254, "step": 24410 }, { "epoch": 1.4443721535458685, "grad_norm": 3.0246567726135254, "learning_rate": 3.325496396875818e-06, "loss": 0.3279, "step": 24420 }, { "epoch": 1.4449636245342166, "grad_norm": 2.8438000679016113, "learning_rate": 3.3235919952853625e-06, "loss": 0.2777, "step": 24430 }, { "epoch": 1.4455550955225647, "grad_norm": 3.1746134757995605, "learning_rate": 3.321687461752335e-06, "loss": 0.3282, "step": 24440 }, { "epoch": 1.4461465665109126, "grad_norm": 2.973914861679077, "learning_rate": 3.319782797053295e-06, "loss": 0.3104, "step": 24450 }, { "epoch": 1.4467380374992607, "grad_norm": 3.608588695526123, "learning_rate": 3.3178780019648584e-06, "loss": 0.2649, "step": 24460 }, { "epoch": 1.4473295084876088, "grad_norm": 4.235956192016602, "learning_rate": 3.315973077263693e-06, "loss": 0.295, "step": 24470 }, { "epoch": 1.4479209794759567, "grad_norm": 3.9041945934295654, "learning_rate": 3.314068023726519e-06, "loss": 0.2558, "step": 24480 }, { "epoch": 1.4485124504643048, "grad_norm": 2.2557003498077393, "learning_rate": 3.3121628421301114e-06, "loss": 0.2063, "step": 24490 }, { "epoch": 1.4491039214526529, "grad_norm": 4.850674152374268, "learning_rate": 3.310257533251295e-06, "loss": 0.314, "step": 24500 }, { "epoch": 1.4496953924410008, "grad_norm": 3.550273895263672, "learning_rate": 3.3083520978669473e-06, "loss": 0.3265, "step": 24510 }, { "epoch": 1.4502868634293489, "grad_norm": 3.052180290222168, "learning_rate": 3.306446536753998e-06, "loss": 0.352, "step": 24520 }, { "epoch": 1.4508783344176968, "grad_norm": 3.0732529163360596, "learning_rate": 3.3045408506894256e-06, "loss": 0.267, "step": 24530 }, { "epoch": 1.4514698054060449, "grad_norm": 2.2239227294921875, "learning_rate": 3.3026350404502632e-06, "loss": 0.2858, "step": 24540 }, { "epoch": 1.4520612763943928, "grad_norm": 2.4022932052612305, "learning_rate": 3.300729106813593e-06, "loss": 0.2394, "step": 24550 }, { "epoch": 1.4526527473827409, "grad_norm": 2.7648427486419678, "learning_rate": 3.298823050556546e-06, "loss": 0.294, "step": 24560 }, { "epoch": 1.453244218371089, "grad_norm": 3.948390245437622, "learning_rate": 3.2969168724563063e-06, "loss": 0.2898, "step": 24570 }, { "epoch": 1.4538356893594369, "grad_norm": 4.4489850997924805, "learning_rate": 3.2950105732901036e-06, "loss": 0.3126, "step": 24580 }, { "epoch": 1.454427160347785, "grad_norm": 2.509122371673584, "learning_rate": 3.293104153835221e-06, "loss": 0.3055, "step": 24590 }, { "epoch": 1.455018631336133, "grad_norm": 3.5321943759918213, "learning_rate": 3.291197614868987e-06, "loss": 0.273, "step": 24600 }, { "epoch": 1.455610102324481, "grad_norm": 3.5609169006347656, "learning_rate": 3.289290957168783e-06, "loss": 0.2905, "step": 24610 }, { "epoch": 1.456201573312829, "grad_norm": 2.409616708755493, "learning_rate": 3.2873841815120363e-06, "loss": 0.2347, "step": 24620 }, { "epoch": 1.456793044301177, "grad_norm": 2.457463502883911, "learning_rate": 3.28547728867622e-06, "loss": 0.299, "step": 24630 }, { "epoch": 1.457384515289525, "grad_norm": 2.119673728942871, "learning_rate": 3.2835702794388605e-06, "loss": 0.2635, "step": 24640 }, { "epoch": 1.457975986277873, "grad_norm": 2.7702152729034424, "learning_rate": 3.281663154577527e-06, "loss": 0.3318, "step": 24650 }, { "epoch": 1.458567457266221, "grad_norm": 2.9877395629882812, "learning_rate": 3.2797559148698375e-06, "loss": 0.259, "step": 24660 }, { "epoch": 1.4591589282545692, "grad_norm": 1.8747104406356812, "learning_rate": 3.277848561093457e-06, "loss": 0.2849, "step": 24670 }, { "epoch": 1.459750399242917, "grad_norm": 3.448568344116211, "learning_rate": 3.2759410940260976e-06, "loss": 0.2353, "step": 24680 }, { "epoch": 1.4603418702312652, "grad_norm": 2.1943087577819824, "learning_rate": 3.2740335144455166e-06, "loss": 0.3172, "step": 24690 }, { "epoch": 1.4609333412196133, "grad_norm": 3.748281478881836, "learning_rate": 3.272125823129517e-06, "loss": 0.3051, "step": 24700 }, { "epoch": 1.4615248122079612, "grad_norm": 2.4545881748199463, "learning_rate": 3.2702180208559473e-06, "loss": 0.2287, "step": 24710 }, { "epoch": 1.4621162831963093, "grad_norm": 3.419858932495117, "learning_rate": 3.268310108402702e-06, "loss": 0.283, "step": 24720 }, { "epoch": 1.4627077541846574, "grad_norm": 3.4365646839141846, "learning_rate": 3.2664020865477215e-06, "loss": 0.3199, "step": 24730 }, { "epoch": 1.4632992251730053, "grad_norm": 3.701984167098999, "learning_rate": 3.2644939560689874e-06, "loss": 0.2699, "step": 24740 }, { "epoch": 1.4638906961613534, "grad_norm": 3.902087688446045, "learning_rate": 3.2625857177445293e-06, "loss": 0.2367, "step": 24750 }, { "epoch": 1.4644821671497013, "grad_norm": 3.0171916484832764, "learning_rate": 3.2606773723524195e-06, "loss": 0.3332, "step": 24760 }, { "epoch": 1.4650736381380494, "grad_norm": 2.206313133239746, "learning_rate": 3.2587689206707725e-06, "loss": 0.2478, "step": 24770 }, { "epoch": 1.4656651091263972, "grad_norm": 4.121060848236084, "learning_rate": 3.2568603634777473e-06, "loss": 0.3643, "step": 24780 }, { "epoch": 1.4662565801147454, "grad_norm": 2.034881353378296, "learning_rate": 3.2549517015515466e-06, "loss": 0.3115, "step": 24790 }, { "epoch": 1.4668480511030935, "grad_norm": 5.241061687469482, "learning_rate": 3.253042935670415e-06, "loss": 0.3172, "step": 24800 }, { "epoch": 1.4674395220914414, "grad_norm": 2.7933850288391113, "learning_rate": 3.25113406661264e-06, "loss": 0.2689, "step": 24810 }, { "epoch": 1.4680309930797895, "grad_norm": 2.0501909255981445, "learning_rate": 3.24922509515655e-06, "loss": 0.2927, "step": 24820 }, { "epoch": 1.4686224640681376, "grad_norm": 2.129985809326172, "learning_rate": 3.2473160220805165e-06, "loss": 0.3665, "step": 24830 }, { "epoch": 1.4692139350564855, "grad_norm": 3.0011391639709473, "learning_rate": 3.245406848162952e-06, "loss": 0.3033, "step": 24840 }, { "epoch": 1.4698054060448336, "grad_norm": 2.7639293670654297, "learning_rate": 3.2434975741823093e-06, "loss": 0.2921, "step": 24850 }, { "epoch": 1.4703968770331814, "grad_norm": 2.8749167919158936, "learning_rate": 3.2415882009170826e-06, "loss": 0.3029, "step": 24860 }, { "epoch": 1.4709883480215296, "grad_norm": 2.996591806411743, "learning_rate": 3.239678729145807e-06, "loss": 0.3535, "step": 24870 }, { "epoch": 1.4715798190098774, "grad_norm": 2.1575162410736084, "learning_rate": 3.237769159647058e-06, "loss": 0.2857, "step": 24880 }, { "epoch": 1.4721712899982256, "grad_norm": 2.2610299587249756, "learning_rate": 3.23585949319945e-06, "loss": 0.3172, "step": 24890 }, { "epoch": 1.4727627609865737, "grad_norm": 1.549115777015686, "learning_rate": 3.233949730581638e-06, "loss": 0.2954, "step": 24900 }, { "epoch": 1.4733542319749215, "grad_norm": 4.604971885681152, "learning_rate": 3.232039872572314e-06, "loss": 0.2875, "step": 24910 }, { "epoch": 1.4739457029632697, "grad_norm": 4.8294243812561035, "learning_rate": 3.230129919950211e-06, "loss": 0.2771, "step": 24920 }, { "epoch": 1.4745371739516178, "grad_norm": 3.3236725330352783, "learning_rate": 3.2282198734941e-06, "loss": 0.2907, "step": 24930 }, { "epoch": 1.4751286449399656, "grad_norm": 4.126614093780518, "learning_rate": 3.226309733982791e-06, "loss": 0.2982, "step": 24940 }, { "epoch": 1.4757201159283138, "grad_norm": 4.825358867645264, "learning_rate": 3.2243995021951315e-06, "loss": 0.303, "step": 24950 }, { "epoch": 1.4763115869166619, "grad_norm": 4.482806205749512, "learning_rate": 3.2224891789100056e-06, "loss": 0.276, "step": 24960 }, { "epoch": 1.4769030579050098, "grad_norm": 3.3490383625030518, "learning_rate": 3.2205787649063343e-06, "loss": 0.285, "step": 24970 }, { "epoch": 1.4774945288933579, "grad_norm": 4.6165056228637695, "learning_rate": 3.218668260963078e-06, "loss": 0.3198, "step": 24980 }, { "epoch": 1.4780859998817057, "grad_norm": 3.750896692276001, "learning_rate": 3.216757667859232e-06, "loss": 0.2851, "step": 24990 }, { "epoch": 1.4786774708700539, "grad_norm": 2.3222575187683105, "learning_rate": 3.2148469863738296e-06, "loss": 0.2634, "step": 25000 }, { "epoch": 1.4792689418584017, "grad_norm": 4.879354000091553, "learning_rate": 3.2129362172859377e-06, "loss": 0.2279, "step": 25010 }, { "epoch": 1.4798604128467499, "grad_norm": 3.582110643386841, "learning_rate": 3.211025361374661e-06, "loss": 0.2516, "step": 25020 }, { "epoch": 1.480451883835098, "grad_norm": 3.381779193878174, "learning_rate": 3.2091144194191384e-06, "loss": 0.3187, "step": 25030 }, { "epoch": 1.4810433548234458, "grad_norm": 2.5687050819396973, "learning_rate": 3.207203392198544e-06, "loss": 0.3132, "step": 25040 }, { "epoch": 1.481634825811794, "grad_norm": 3.4716734886169434, "learning_rate": 3.2052922804920884e-06, "loss": 0.2749, "step": 25050 }, { "epoch": 1.482226296800142, "grad_norm": 2.49271559715271, "learning_rate": 3.2033810850790134e-06, "loss": 0.2669, "step": 25060 }, { "epoch": 1.48281776778849, "grad_norm": 3.500342607498169, "learning_rate": 3.2014698067385976e-06, "loss": 0.3121, "step": 25070 }, { "epoch": 1.483409238776838, "grad_norm": 3.5275461673736572, "learning_rate": 3.199558446250152e-06, "loss": 0.2996, "step": 25080 }, { "epoch": 1.484000709765186, "grad_norm": 2.1417596340179443, "learning_rate": 3.1976470043930236e-06, "loss": 0.2927, "step": 25090 }, { "epoch": 1.484592180753534, "grad_norm": 3.553389549255371, "learning_rate": 3.195735481946588e-06, "loss": 0.2691, "step": 25100 }, { "epoch": 1.485183651741882, "grad_norm": 2.7141189575195312, "learning_rate": 3.193823879690258e-06, "loss": 0.2976, "step": 25110 }, { "epoch": 1.48577512273023, "grad_norm": 3.175114393234253, "learning_rate": 3.1919121984034764e-06, "loss": 0.3206, "step": 25120 }, { "epoch": 1.4863665937185782, "grad_norm": 1.8652920722961426, "learning_rate": 3.190000438865719e-06, "loss": 0.2546, "step": 25130 }, { "epoch": 1.486958064706926, "grad_norm": 2.854996681213379, "learning_rate": 3.188088601856495e-06, "loss": 0.2922, "step": 25140 }, { "epoch": 1.4875495356952741, "grad_norm": 3.1072490215301514, "learning_rate": 3.1861766881553414e-06, "loss": 0.3255, "step": 25150 }, { "epoch": 1.4881410066836223, "grad_norm": 2.6069047451019287, "learning_rate": 3.184264698541832e-06, "loss": 0.2904, "step": 25160 }, { "epoch": 1.4887324776719701, "grad_norm": 4.189115524291992, "learning_rate": 3.1823526337955645e-06, "loss": 0.3072, "step": 25170 }, { "epoch": 1.4893239486603183, "grad_norm": 3.7973787784576416, "learning_rate": 3.1804404946961736e-06, "loss": 0.3315, "step": 25180 }, { "epoch": 1.4899154196486664, "grad_norm": 2.308440923690796, "learning_rate": 3.178528282023322e-06, "loss": 0.2649, "step": 25190 }, { "epoch": 1.4905068906370142, "grad_norm": 2.4473860263824463, "learning_rate": 3.1766159965567003e-06, "loss": 0.2092, "step": 25200 }, { "epoch": 1.4910983616253624, "grad_norm": 3.6097702980041504, "learning_rate": 3.1747036390760335e-06, "loss": 0.2703, "step": 25210 }, { "epoch": 1.4916898326137102, "grad_norm": 6.961747646331787, "learning_rate": 3.1727912103610703e-06, "loss": 0.3365, "step": 25220 }, { "epoch": 1.4922813036020584, "grad_norm": 2.3891475200653076, "learning_rate": 3.1708787111915933e-06, "loss": 0.305, "step": 25230 }, { "epoch": 1.4928727745904062, "grad_norm": 5.904917240142822, "learning_rate": 3.1689661423474113e-06, "loss": 0.3649, "step": 25240 }, { "epoch": 1.4934642455787543, "grad_norm": 3.513176441192627, "learning_rate": 3.1670535046083615e-06, "loss": 0.2891, "step": 25250 }, { "epoch": 1.4940557165671025, "grad_norm": 2.5412213802337646, "learning_rate": 3.1651407987543103e-06, "loss": 0.3086, "step": 25260 }, { "epoch": 1.4946471875554503, "grad_norm": 3.824460506439209, "learning_rate": 3.1632280255651514e-06, "loss": 0.2662, "step": 25270 }, { "epoch": 1.4952386585437984, "grad_norm": 1.6095994710922241, "learning_rate": 3.1613151858208063e-06, "loss": 0.3108, "step": 25280 }, { "epoch": 1.4958301295321466, "grad_norm": 1.5731308460235596, "learning_rate": 3.159402280301222e-06, "loss": 0.2295, "step": 25290 }, { "epoch": 1.4964216005204944, "grad_norm": 1.6426239013671875, "learning_rate": 3.157489309786374e-06, "loss": 0.2494, "step": 25300 }, { "epoch": 1.4970130715088426, "grad_norm": 2.835155963897705, "learning_rate": 3.1555762750562632e-06, "loss": 0.3754, "step": 25310 }, { "epoch": 1.4976045424971904, "grad_norm": 5.530085563659668, "learning_rate": 3.1536631768909184e-06, "loss": 0.2647, "step": 25320 }, { "epoch": 1.4981960134855385, "grad_norm": 2.9295380115509033, "learning_rate": 3.1517500160703935e-06, "loss": 0.2887, "step": 25330 }, { "epoch": 1.4987874844738864, "grad_norm": 5.166153907775879, "learning_rate": 3.149836793374767e-06, "loss": 0.3384, "step": 25340 }, { "epoch": 1.4993789554622345, "grad_norm": 2.9524030685424805, "learning_rate": 3.1479235095841427e-06, "loss": 0.2817, "step": 25350 }, { "epoch": 1.4999704264505827, "grad_norm": 2.9684805870056152, "learning_rate": 3.1460101654786507e-06, "loss": 0.3198, "step": 25360 }, { "epoch": 1.5002661619447566, "eval_accuracy": 0.8623615131250624, "eval_loss": 0.3146374821662903, "eval_runtime": 158.541, "eval_safe_aucpr": 0.922519859460732, "eval_safe_f1": 0.8461509854964671, "eval_safe_fpr": 0.13045947804250954, "eval_safe_precision": 0.8390677434819486, "eval_safe_recall": 0.8533548362899899, "eval_samples_per_second": 379.17, "eval_steps_per_second": 47.401, "eval_unsafe_aucpr": 0.9555193772259687, "eval_unsafe_f1": 0.8754815795810258, "eval_unsafe_fpr": 0.14664516371000957, "eval_unsafe_precision": 0.8815043791859866, "eval_unsafe_recall": 0.8695405219574901, "step": 25365 }, { "epoch": 1.5005618974389305, "grad_norm": 2.5346572399139404, "learning_rate": 3.144096761838444e-06, "loss": 0.2787, "step": 25370 }, { "epoch": 1.5011533684272786, "grad_norm": 2.199615001678467, "learning_rate": 3.1421832994437024e-06, "loss": 0.3027, "step": 25380 }, { "epoch": 1.5017448394156268, "grad_norm": 2.84907865524292, "learning_rate": 3.1402697790746266e-06, "loss": 0.332, "step": 25390 }, { "epoch": 1.5023363104039746, "grad_norm": 2.242475748062134, "learning_rate": 3.1383562015114416e-06, "loss": 0.3004, "step": 25400 }, { "epoch": 1.5029277813923227, "grad_norm": 2.9061338901519775, "learning_rate": 3.1364425675343983e-06, "loss": 0.3274, "step": 25410 }, { "epoch": 1.5035192523806709, "grad_norm": 1.7527225017547607, "learning_rate": 3.134528877923768e-06, "loss": 0.3306, "step": 25420 }, { "epoch": 1.5041107233690187, "grad_norm": 2.937824249267578, "learning_rate": 3.132615133459845e-06, "loss": 0.3274, "step": 25430 }, { "epoch": 1.5047021943573666, "grad_norm": 3.763871192932129, "learning_rate": 3.1307013349229465e-06, "loss": 0.288, "step": 25440 }, { "epoch": 1.5052936653457147, "grad_norm": 6.027100086212158, "learning_rate": 3.1287874830934125e-06, "loss": 0.3206, "step": 25450 }, { "epoch": 1.5058851363340628, "grad_norm": 3.8608033657073975, "learning_rate": 3.1268735787516034e-06, "loss": 0.2742, "step": 25460 }, { "epoch": 1.5064766073224107, "grad_norm": 4.195398807525635, "learning_rate": 3.1249596226779012e-06, "loss": 0.3189, "step": 25470 }, { "epoch": 1.5070680783107588, "grad_norm": 3.4592270851135254, "learning_rate": 3.1230456156527096e-06, "loss": 0.3371, "step": 25480 }, { "epoch": 1.507659549299107, "grad_norm": 2.60465145111084, "learning_rate": 3.121131558456453e-06, "loss": 0.3047, "step": 25490 }, { "epoch": 1.5082510202874548, "grad_norm": 2.208045721054077, "learning_rate": 3.119217451869574e-06, "loss": 0.2575, "step": 25500 }, { "epoch": 1.508842491275803, "grad_norm": 3.9322116374969482, "learning_rate": 3.1173032966725407e-06, "loss": 0.2856, "step": 25510 }, { "epoch": 1.509433962264151, "grad_norm": 3.396644115447998, "learning_rate": 3.115389093645835e-06, "loss": 0.3235, "step": 25520 }, { "epoch": 1.510025433252499, "grad_norm": 3.8228366374969482, "learning_rate": 3.1134748435699632e-06, "loss": 0.2474, "step": 25530 }, { "epoch": 1.5106169042408468, "grad_norm": 4.083652973175049, "learning_rate": 3.1115605472254464e-06, "loss": 0.2768, "step": 25540 }, { "epoch": 1.5112083752291952, "grad_norm": 4.496576309204102, "learning_rate": 3.1096462053928277e-06, "loss": 0.3066, "step": 25550 }, { "epoch": 1.511799846217543, "grad_norm": 2.8662455081939697, "learning_rate": 3.1077318188526678e-06, "loss": 0.3309, "step": 25560 }, { "epoch": 1.512391317205891, "grad_norm": 3.247518301010132, "learning_rate": 3.105817388385546e-06, "loss": 0.3146, "step": 25570 }, { "epoch": 1.512982788194239, "grad_norm": 2.8356926441192627, "learning_rate": 3.1039029147720576e-06, "loss": 0.2763, "step": 25580 }, { "epoch": 1.5135742591825871, "grad_norm": 3.803041934967041, "learning_rate": 3.101988398792819e-06, "loss": 0.2793, "step": 25590 }, { "epoch": 1.514165730170935, "grad_norm": 3.7863290309906006, "learning_rate": 3.100073841228462e-06, "loss": 0.245, "step": 25600 }, { "epoch": 1.5147572011592831, "grad_norm": 1.9435148239135742, "learning_rate": 3.0981592428596326e-06, "loss": 0.2599, "step": 25610 }, { "epoch": 1.5153486721476312, "grad_norm": 3.4606521129608154, "learning_rate": 3.096244604466998e-06, "loss": 0.2711, "step": 25620 }, { "epoch": 1.5159401431359791, "grad_norm": 3.339477062225342, "learning_rate": 3.0943299268312397e-06, "loss": 0.2756, "step": 25630 }, { "epoch": 1.5165316141243272, "grad_norm": 3.769590139389038, "learning_rate": 3.092415210733055e-06, "loss": 0.3094, "step": 25640 }, { "epoch": 1.5171230851126754, "grad_norm": 2.901398181915283, "learning_rate": 3.0905004569531576e-06, "loss": 0.2883, "step": 25650 }, { "epoch": 1.5177145561010232, "grad_norm": 4.210694789886475, "learning_rate": 3.0885856662722756e-06, "loss": 0.3482, "step": 25660 }, { "epoch": 1.5183060270893711, "grad_norm": 2.442542791366577, "learning_rate": 3.086670839471153e-06, "loss": 0.2917, "step": 25670 }, { "epoch": 1.5188974980777192, "grad_norm": 3.1890923976898193, "learning_rate": 3.084755977330548e-06, "loss": 0.2527, "step": 25680 }, { "epoch": 1.5194889690660673, "grad_norm": 3.091140031814575, "learning_rate": 3.082841080631233e-06, "loss": 0.3147, "step": 25690 }, { "epoch": 1.5200804400544152, "grad_norm": 3.257516860961914, "learning_rate": 3.0809261501539967e-06, "loss": 0.3231, "step": 25700 }, { "epoch": 1.5206719110427633, "grad_norm": 3.2833664417266846, "learning_rate": 3.079011186679638e-06, "loss": 0.3102, "step": 25710 }, { "epoch": 1.5212633820311114, "grad_norm": 2.664778470993042, "learning_rate": 3.0770961909889716e-06, "loss": 0.3071, "step": 25720 }, { "epoch": 1.5218548530194593, "grad_norm": 2.6665494441986084, "learning_rate": 3.075181163862825e-06, "loss": 0.2601, "step": 25730 }, { "epoch": 1.5224463240078074, "grad_norm": 2.3728246688842773, "learning_rate": 3.0732661060820385e-06, "loss": 0.279, "step": 25740 }, { "epoch": 1.5230377949961555, "grad_norm": 2.2177798748016357, "learning_rate": 3.0713510184274636e-06, "loss": 0.3196, "step": 25750 }, { "epoch": 1.5236292659845034, "grad_norm": 2.226431369781494, "learning_rate": 3.0694359016799653e-06, "loss": 0.2352, "step": 25760 }, { "epoch": 1.5242207369728513, "grad_norm": 3.1366467475891113, "learning_rate": 3.0675207566204217e-06, "loss": 0.2623, "step": 25770 }, { "epoch": 1.5248122079611997, "grad_norm": 5.250776290893555, "learning_rate": 3.0656055840297193e-06, "loss": 0.3344, "step": 25780 }, { "epoch": 1.5254036789495475, "grad_norm": 4.341691970825195, "learning_rate": 3.06369038468876e-06, "loss": 0.315, "step": 25790 }, { "epoch": 1.5259951499378954, "grad_norm": 3.8148510456085205, "learning_rate": 3.0617751593784504e-06, "loss": 0.2517, "step": 25800 }, { "epoch": 1.5265866209262435, "grad_norm": 2.50771164894104, "learning_rate": 3.0598599088797144e-06, "loss": 0.2354, "step": 25810 }, { "epoch": 1.5271780919145916, "grad_norm": 3.7260053157806396, "learning_rate": 3.0579446339734823e-06, "loss": 0.2735, "step": 25820 }, { "epoch": 1.5277695629029395, "grad_norm": 3.4159319400787354, "learning_rate": 3.0560293354406946e-06, "loss": 0.2624, "step": 25830 }, { "epoch": 1.5283610338912876, "grad_norm": 4.505009174346924, "learning_rate": 3.054114014062302e-06, "loss": 0.27, "step": 25840 }, { "epoch": 1.5289525048796357, "grad_norm": 2.284076452255249, "learning_rate": 3.052198670619267e-06, "loss": 0.2648, "step": 25850 }, { "epoch": 1.5295439758679836, "grad_norm": 1.6729878187179565, "learning_rate": 3.0502833058925548e-06, "loss": 0.3321, "step": 25860 }, { "epoch": 1.5301354468563317, "grad_norm": 3.070261001586914, "learning_rate": 3.048367920663146e-06, "loss": 0.2718, "step": 25870 }, { "epoch": 1.5307269178446798, "grad_norm": 5.534438610076904, "learning_rate": 3.046452515712026e-06, "loss": 0.2863, "step": 25880 }, { "epoch": 1.5313183888330277, "grad_norm": 3.8497025966644287, "learning_rate": 3.0445370918201886e-06, "loss": 0.3254, "step": 25890 }, { "epoch": 1.5319098598213756, "grad_norm": 2.8933048248291016, "learning_rate": 3.0426216497686358e-06, "loss": 0.2867, "step": 25900 }, { "epoch": 1.532501330809724, "grad_norm": 2.1970372200012207, "learning_rate": 3.040706190338378e-06, "loss": 0.3087, "step": 25910 }, { "epoch": 1.5330928017980718, "grad_norm": 2.3027572631835938, "learning_rate": 3.0387907143104304e-06, "loss": 0.2284, "step": 25920 }, { "epoch": 1.5336842727864197, "grad_norm": 2.725386142730713, "learning_rate": 3.036875222465816e-06, "loss": 0.3023, "step": 25930 }, { "epoch": 1.5342757437747678, "grad_norm": 4.081414222717285, "learning_rate": 3.034959715585564e-06, "loss": 0.2944, "step": 25940 }, { "epoch": 1.534867214763116, "grad_norm": 3.7029733657836914, "learning_rate": 3.0330441944507113e-06, "loss": 0.269, "step": 25950 }, { "epoch": 1.5354586857514638, "grad_norm": 3.6880617141723633, "learning_rate": 3.0311286598422993e-06, "loss": 0.2539, "step": 25960 }, { "epoch": 1.536050156739812, "grad_norm": 2.90102219581604, "learning_rate": 3.029213112541374e-06, "loss": 0.3187, "step": 25970 }, { "epoch": 1.53664162772816, "grad_norm": 5.123143196105957, "learning_rate": 3.0272975533289894e-06, "loss": 0.2926, "step": 25980 }, { "epoch": 1.537233098716508, "grad_norm": 3.1495792865753174, "learning_rate": 3.0253819829862012e-06, "loss": 0.3062, "step": 25990 }, { "epoch": 1.5378245697048558, "grad_norm": 2.3517379760742188, "learning_rate": 3.0234664022940706e-06, "loss": 0.2861, "step": 26000 }, { "epoch": 1.5384160406932041, "grad_norm": 4.222979545593262, "learning_rate": 3.0215508120336644e-06, "loss": 0.2379, "step": 26010 }, { "epoch": 1.539007511681552, "grad_norm": 3.812103748321533, "learning_rate": 3.0196352129860523e-06, "loss": 0.2229, "step": 26020 }, { "epoch": 1.5395989826699, "grad_norm": 4.253983020782471, "learning_rate": 3.0177196059323065e-06, "loss": 0.2335, "step": 26030 }, { "epoch": 1.540190453658248, "grad_norm": 4.096487522125244, "learning_rate": 3.0158039916535047e-06, "loss": 0.371, "step": 26040 }, { "epoch": 1.5407819246465961, "grad_norm": 3.9204699993133545, "learning_rate": 3.013888370930727e-06, "loss": 0.3194, "step": 26050 }, { "epoch": 1.541373395634944, "grad_norm": 3.3249051570892334, "learning_rate": 3.011972744545053e-06, "loss": 0.2565, "step": 26060 }, { "epoch": 1.5419648666232921, "grad_norm": 1.9052549600601196, "learning_rate": 3.0100571132775697e-06, "loss": 0.3004, "step": 26070 }, { "epoch": 1.5425563376116402, "grad_norm": 2.4889862537384033, "learning_rate": 3.0081414779093623e-06, "loss": 0.278, "step": 26080 }, { "epoch": 1.5431478085999881, "grad_norm": 2.8260884284973145, "learning_rate": 3.0062258392215194e-06, "loss": 0.2533, "step": 26090 }, { "epoch": 1.5437392795883362, "grad_norm": 2.978153705596924, "learning_rate": 3.0043101979951293e-06, "loss": 0.2629, "step": 26100 }, { "epoch": 1.5443307505766843, "grad_norm": 3.3211867809295654, "learning_rate": 3.002394555011285e-06, "loss": 0.2784, "step": 26110 }, { "epoch": 1.5449222215650322, "grad_norm": 3.406031608581543, "learning_rate": 3.0004789110510754e-06, "loss": 0.2673, "step": 26120 }, { "epoch": 1.5455136925533801, "grad_norm": 2.6304101943969727, "learning_rate": 2.998563266895593e-06, "loss": 0.3209, "step": 26130 }, { "epoch": 1.5461051635417284, "grad_norm": 4.139064311981201, "learning_rate": 2.9966476233259284e-06, "loss": 0.308, "step": 26140 }, { "epoch": 1.5466966345300763, "grad_norm": 4.783566474914551, "learning_rate": 2.994731981123175e-06, "loss": 0.2668, "step": 26150 }, { "epoch": 1.5472881055184242, "grad_norm": 4.021961688995361, "learning_rate": 2.992816341068423e-06, "loss": 0.2492, "step": 26160 }, { "epoch": 1.5478795765067723, "grad_norm": 5.277831554412842, "learning_rate": 2.9909007039427614e-06, "loss": 0.2342, "step": 26170 }, { "epoch": 1.5484710474951204, "grad_norm": 5.949741840362549, "learning_rate": 2.988985070527281e-06, "loss": 0.2451, "step": 26180 }, { "epoch": 1.5490625184834683, "grad_norm": 2.754152536392212, "learning_rate": 2.9870694416030673e-06, "loss": 0.2864, "step": 26190 }, { "epoch": 1.5496539894718164, "grad_norm": 3.238816261291504, "learning_rate": 2.985153817951206e-06, "loss": 0.3443, "step": 26200 }, { "epoch": 1.5502454604601645, "grad_norm": 2.8574118614196777, "learning_rate": 2.98323820035278e-06, "loss": 0.2488, "step": 26210 }, { "epoch": 1.5508369314485124, "grad_norm": 4.752385139465332, "learning_rate": 2.981322589588872e-06, "loss": 0.2486, "step": 26220 }, { "epoch": 1.5514284024368603, "grad_norm": 2.0920796394348145, "learning_rate": 2.9794069864405583e-06, "loss": 0.2641, "step": 26230 }, { "epoch": 1.5520198734252086, "grad_norm": 2.0783212184906006, "learning_rate": 2.9774913916889153e-06, "loss": 0.311, "step": 26240 }, { "epoch": 1.5526113444135565, "grad_norm": 3.8802833557128906, "learning_rate": 2.975575806115014e-06, "loss": 0.2652, "step": 26250 }, { "epoch": 1.5532028154019044, "grad_norm": 3.290278434753418, "learning_rate": 2.9736602304999213e-06, "loss": 0.27, "step": 26260 }, { "epoch": 1.5537942863902525, "grad_norm": 2.0454154014587402, "learning_rate": 2.971744665624702e-06, "loss": 0.314, "step": 26270 }, { "epoch": 1.5543857573786006, "grad_norm": 2.4045791625976562, "learning_rate": 2.969829112270415e-06, "loss": 0.3065, "step": 26280 }, { "epoch": 1.5549772283669485, "grad_norm": 4.768416881561279, "learning_rate": 2.9679135712181154e-06, "loss": 0.2933, "step": 26290 }, { "epoch": 1.5555686993552966, "grad_norm": 4.040069580078125, "learning_rate": 2.9659980432488526e-06, "loss": 0.3059, "step": 26300 }, { "epoch": 1.5561601703436447, "grad_norm": 3.215203285217285, "learning_rate": 2.9640825291436705e-06, "loss": 0.3407, "step": 26310 }, { "epoch": 1.5567516413319926, "grad_norm": 2.818946123123169, "learning_rate": 2.9621670296836085e-06, "loss": 0.2645, "step": 26320 }, { "epoch": 1.5573431123203407, "grad_norm": 2.553252696990967, "learning_rate": 2.960251545649699e-06, "loss": 0.3181, "step": 26330 }, { "epoch": 1.5579345833086888, "grad_norm": 2.1011476516723633, "learning_rate": 2.958336077822968e-06, "loss": 0.275, "step": 26340 }, { "epoch": 1.5585260542970367, "grad_norm": 3.215480327606201, "learning_rate": 2.956420626984436e-06, "loss": 0.263, "step": 26350 }, { "epoch": 1.5591175252853846, "grad_norm": 2.2534332275390625, "learning_rate": 2.9545051939151157e-06, "loss": 0.3, "step": 26360 }, { "epoch": 1.559708996273733, "grad_norm": 3.697725534439087, "learning_rate": 2.9525897793960122e-06, "loss": 0.2923, "step": 26370 }, { "epoch": 1.5603004672620808, "grad_norm": 2.356848955154419, "learning_rate": 2.950674384208124e-06, "loss": 0.2861, "step": 26380 }, { "epoch": 1.5608919382504287, "grad_norm": 4.679379463195801, "learning_rate": 2.9487590091324407e-06, "loss": 0.3392, "step": 26390 }, { "epoch": 1.5614834092387768, "grad_norm": 2.9393928050994873, "learning_rate": 2.9468436549499467e-06, "loss": 0.2578, "step": 26400 }, { "epoch": 1.562074880227125, "grad_norm": 3.4031412601470947, "learning_rate": 2.9449283224416116e-06, "loss": 0.2901, "step": 26410 }, { "epoch": 1.5626663512154728, "grad_norm": 2.0396687984466553, "learning_rate": 2.9430130123884025e-06, "loss": 0.2841, "step": 26420 }, { "epoch": 1.563257822203821, "grad_norm": 3.337575674057007, "learning_rate": 2.941097725571274e-06, "loss": 0.2874, "step": 26430 }, { "epoch": 1.563849293192169, "grad_norm": 2.9310286045074463, "learning_rate": 2.939182462771173e-06, "loss": 0.2526, "step": 26440 }, { "epoch": 1.564440764180517, "grad_norm": 5.641634464263916, "learning_rate": 2.9372672247690347e-06, "loss": 0.2643, "step": 26450 }, { "epoch": 1.5650322351688648, "grad_norm": 1.4400808811187744, "learning_rate": 2.9353520123457854e-06, "loss": 0.2493, "step": 26460 }, { "epoch": 1.5656237061572131, "grad_norm": 2.4059717655181885, "learning_rate": 2.933436826282341e-06, "loss": 0.2785, "step": 26470 }, { "epoch": 1.566215177145561, "grad_norm": 3.791339874267578, "learning_rate": 2.9315216673596063e-06, "loss": 0.3131, "step": 26480 }, { "epoch": 1.566806648133909, "grad_norm": 1.9973297119140625, "learning_rate": 2.929606536358475e-06, "loss": 0.2562, "step": 26490 }, { "epoch": 1.567398119122257, "grad_norm": 2.816432237625122, "learning_rate": 2.9276914340598306e-06, "loss": 0.3009, "step": 26500 }, { "epoch": 1.5679895901106051, "grad_norm": 1.8645859956741333, "learning_rate": 2.9257763612445423e-06, "loss": 0.2806, "step": 26510 }, { "epoch": 1.568581061098953, "grad_norm": 3.8732948303222656, "learning_rate": 2.9238613186934694e-06, "loss": 0.3559, "step": 26520 }, { "epoch": 1.5691725320873011, "grad_norm": 3.320103168487549, "learning_rate": 2.9219463071874584e-06, "loss": 0.2819, "step": 26530 }, { "epoch": 1.5697640030756492, "grad_norm": 5.329129219055176, "learning_rate": 2.9200313275073433e-06, "loss": 0.2975, "step": 26540 }, { "epoch": 1.5703554740639971, "grad_norm": 1.980336308479309, "learning_rate": 2.9181163804339445e-06, "loss": 0.2804, "step": 26550 }, { "epoch": 1.5709469450523452, "grad_norm": 3.8186604976654053, "learning_rate": 2.91620146674807e-06, "loss": 0.3304, "step": 26560 }, { "epoch": 1.5715384160406933, "grad_norm": 1.364706039428711, "learning_rate": 2.9142865872305123e-06, "loss": 0.2633, "step": 26570 }, { "epoch": 1.5721298870290412, "grad_norm": 3.439218282699585, "learning_rate": 2.912371742662053e-06, "loss": 0.2445, "step": 26580 }, { "epoch": 1.572721358017389, "grad_norm": 3.4876255989074707, "learning_rate": 2.9104569338234565e-06, "loss": 0.3177, "step": 26590 }, { "epoch": 1.5733128290057374, "grad_norm": 1.3355575799942017, "learning_rate": 2.908542161495474e-06, "loss": 0.2738, "step": 26600 }, { "epoch": 1.5739042999940853, "grad_norm": 4.281816482543945, "learning_rate": 2.9066274264588445e-06, "loss": 0.3129, "step": 26610 }, { "epoch": 1.5744957709824332, "grad_norm": 2.405324935913086, "learning_rate": 2.9047127294942846e-06, "loss": 0.2666, "step": 26620 }, { "epoch": 1.5750872419707813, "grad_norm": 3.136601448059082, "learning_rate": 2.9027980713825017e-06, "loss": 0.2523, "step": 26630 }, { "epoch": 1.5756787129591294, "grad_norm": 3.204272508621216, "learning_rate": 2.9008834529041866e-06, "loss": 0.3093, "step": 26640 }, { "epoch": 1.5762701839474773, "grad_norm": 3.8411288261413574, "learning_rate": 2.8989688748400107e-06, "loss": 0.2843, "step": 26650 }, { "epoch": 1.5768616549358254, "grad_norm": 3.1713101863861084, "learning_rate": 2.897054337970632e-06, "loss": 0.2711, "step": 26660 }, { "epoch": 1.5774531259241735, "grad_norm": 2.637556314468384, "learning_rate": 2.895139843076691e-06, "loss": 0.2492, "step": 26670 }, { "epoch": 1.5780445969125214, "grad_norm": 2.3727879524230957, "learning_rate": 2.89322539093881e-06, "loss": 0.3255, "step": 26680 }, { "epoch": 1.5786360679008693, "grad_norm": 2.738128423690796, "learning_rate": 2.891310982337596e-06, "loss": 0.2814, "step": 26690 }, { "epoch": 1.5792275388892176, "grad_norm": 2.664134979248047, "learning_rate": 2.8893966180536354e-06, "loss": 0.3507, "step": 26700 }, { "epoch": 1.5798190098775655, "grad_norm": 3.3364064693450928, "learning_rate": 2.8874822988674992e-06, "loss": 0.3013, "step": 26710 }, { "epoch": 1.5804104808659134, "grad_norm": 3.9307382106781006, "learning_rate": 2.8855680255597375e-06, "loss": 0.3108, "step": 26720 }, { "epoch": 1.5810019518542615, "grad_norm": 3.4089064598083496, "learning_rate": 2.8836537989108843e-06, "loss": 0.2428, "step": 26730 }, { "epoch": 1.5815934228426096, "grad_norm": 3.667955160140991, "learning_rate": 2.881739619701453e-06, "loss": 0.2951, "step": 26740 }, { "epoch": 1.5821848938309575, "grad_norm": 3.6474263668060303, "learning_rate": 2.8798254887119385e-06, "loss": 0.2656, "step": 26750 }, { "epoch": 1.5827763648193056, "grad_norm": 3.107556104660034, "learning_rate": 2.877911406722814e-06, "loss": 0.3081, "step": 26760 }, { "epoch": 1.5833678358076537, "grad_norm": 2.328822374343872, "learning_rate": 2.875997374514535e-06, "loss": 0.3229, "step": 26770 }, { "epoch": 1.5839593067960016, "grad_norm": 2.7758066654205322, "learning_rate": 2.874083392867536e-06, "loss": 0.3313, "step": 26780 }, { "epoch": 1.5845507777843497, "grad_norm": 3.4121406078338623, "learning_rate": 2.8721694625622306e-06, "loss": 0.2856, "step": 26790 }, { "epoch": 1.5851422487726978, "grad_norm": 3.7363169193267822, "learning_rate": 2.8702555843790126e-06, "loss": 0.2524, "step": 26800 }, { "epoch": 1.5857337197610457, "grad_norm": 2.9959628582000732, "learning_rate": 2.868341759098253e-06, "loss": 0.2281, "step": 26810 }, { "epoch": 1.5863251907493936, "grad_norm": 2.858116626739502, "learning_rate": 2.8664279875003015e-06, "loss": 0.3598, "step": 26820 }, { "epoch": 1.586916661737742, "grad_norm": 2.563568592071533, "learning_rate": 2.8645142703654877e-06, "loss": 0.2652, "step": 26830 }, { "epoch": 1.5875081327260898, "grad_norm": 3.35103702545166, "learning_rate": 2.862600608474115e-06, "loss": 0.3316, "step": 26840 }, { "epoch": 1.5880996037144377, "grad_norm": 2.7326977252960205, "learning_rate": 2.8606870026064684e-06, "loss": 0.2687, "step": 26850 }, { "epoch": 1.5886910747027858, "grad_norm": 2.6076114177703857, "learning_rate": 2.8587734535428086e-06, "loss": 0.2909, "step": 26860 }, { "epoch": 1.589282545691134, "grad_norm": 3.2033491134643555, "learning_rate": 2.8568599620633724e-06, "loss": 0.3707, "step": 26870 }, { "epoch": 1.5898740166794818, "grad_norm": 3.73589825630188, "learning_rate": 2.8549465289483743e-06, "loss": 0.3048, "step": 26880 }, { "epoch": 1.59046548766783, "grad_norm": 2.7579526901245117, "learning_rate": 2.853033154978005e-06, "loss": 0.3385, "step": 26890 }, { "epoch": 1.591056958656178, "grad_norm": 2.3465874195098877, "learning_rate": 2.8511198409324288e-06, "loss": 0.3311, "step": 26900 }, { "epoch": 1.591648429644526, "grad_norm": 2.7534573078155518, "learning_rate": 2.849206587591789e-06, "loss": 0.2435, "step": 26910 }, { "epoch": 1.5922399006328738, "grad_norm": 4.660729885101318, "learning_rate": 2.847293395736201e-06, "loss": 0.3807, "step": 26920 }, { "epoch": 1.5928313716212221, "grad_norm": 1.9461309909820557, "learning_rate": 2.845380266145758e-06, "loss": 0.2401, "step": 26930 }, { "epoch": 1.59342284260957, "grad_norm": 1.7222020626068115, "learning_rate": 2.8434671996005253e-06, "loss": 0.301, "step": 26940 }, { "epoch": 1.594014313597918, "grad_norm": 7.641035556793213, "learning_rate": 2.841554196880545e-06, "loss": 0.2702, "step": 26950 }, { "epoch": 1.594605784586266, "grad_norm": 4.6164116859436035, "learning_rate": 2.83964125876583e-06, "loss": 0.3278, "step": 26960 }, { "epoch": 1.5951972555746141, "grad_norm": 2.430551528930664, "learning_rate": 2.8377283860363693e-06, "loss": 0.2347, "step": 26970 }, { "epoch": 1.595788726562962, "grad_norm": 3.4573874473571777, "learning_rate": 2.835815579472125e-06, "loss": 0.3458, "step": 26980 }, { "epoch": 1.59638019755131, "grad_norm": 1.802105188369751, "learning_rate": 2.8339028398530315e-06, "loss": 0.2655, "step": 26990 }, { "epoch": 1.5969716685396582, "grad_norm": 3.309400796890259, "learning_rate": 2.8319901679589965e-06, "loss": 0.3259, "step": 27000 }, { "epoch": 1.597563139528006, "grad_norm": 3.8488335609436035, "learning_rate": 2.830077564569899e-06, "loss": 0.3057, "step": 27010 }, { "epoch": 1.5981546105163542, "grad_norm": 2.49363112449646, "learning_rate": 2.828165030465591e-06, "loss": 0.2882, "step": 27020 }, { "epoch": 1.5987460815047023, "grad_norm": 2.994127035140991, "learning_rate": 2.826252566425897e-06, "loss": 0.2989, "step": 27030 }, { "epoch": 1.5993375524930502, "grad_norm": 4.4712347984313965, "learning_rate": 2.824340173230612e-06, "loss": 0.3107, "step": 27040 }, { "epoch": 1.599929023481398, "grad_norm": 3.2980940341949463, "learning_rate": 2.8224278516595005e-06, "loss": 0.2771, "step": 27050 }, { "epoch": 1.6005204944697464, "grad_norm": 3.0468709468841553, "learning_rate": 2.8205156024923006e-06, "loss": 0.2665, "step": 27060 }, { "epoch": 1.6011119654580943, "grad_norm": 3.3822851181030273, "learning_rate": 2.8186034265087195e-06, "loss": 0.3416, "step": 27070 }, { "epoch": 1.6017034364464422, "grad_norm": 2.0378475189208984, "learning_rate": 2.816691324488435e-06, "loss": 0.3202, "step": 27080 }, { "epoch": 1.6022949074347903, "grad_norm": 1.8889912366867065, "learning_rate": 2.8147792972110945e-06, "loss": 0.3087, "step": 27090 }, { "epoch": 1.6028863784231384, "grad_norm": 1.8607791662216187, "learning_rate": 2.812867345456314e-06, "loss": 0.2962, "step": 27100 }, { "epoch": 1.6034778494114863, "grad_norm": 4.5775556564331055, "learning_rate": 2.8109554700036807e-06, "loss": 0.3802, "step": 27110 }, { "epoch": 1.6040693203998344, "grad_norm": 2.183790683746338, "learning_rate": 2.809043671632749e-06, "loss": 0.2883, "step": 27120 }, { "epoch": 1.6046607913881825, "grad_norm": 2.1895320415496826, "learning_rate": 2.807131951123043e-06, "loss": 0.3157, "step": 27130 }, { "epoch": 1.6052522623765304, "grad_norm": 2.854404926300049, "learning_rate": 2.805220309254056e-06, "loss": 0.3038, "step": 27140 }, { "epoch": 1.6058437333648783, "grad_norm": 3.0374186038970947, "learning_rate": 2.8033087468052442e-06, "loss": 0.3201, "step": 27150 }, { "epoch": 1.6064352043532266, "grad_norm": 3.459061861038208, "learning_rate": 2.8013972645560375e-06, "loss": 0.2979, "step": 27160 }, { "epoch": 1.6070266753415745, "grad_norm": 2.2276439666748047, "learning_rate": 2.7994858632858297e-06, "loss": 0.3064, "step": 27170 }, { "epoch": 1.6076181463299224, "grad_norm": 3.0571961402893066, "learning_rate": 2.7975745437739833e-06, "loss": 0.2687, "step": 27180 }, { "epoch": 1.6082096173182705, "grad_norm": 3.8238208293914795, "learning_rate": 2.7956633067998264e-06, "loss": 0.3697, "step": 27190 }, { "epoch": 1.6088010883066186, "grad_norm": 2.733386754989624, "learning_rate": 2.7937521531426533e-06, "loss": 0.2084, "step": 27200 }, { "epoch": 1.6093925592949665, "grad_norm": 3.31535005569458, "learning_rate": 2.791841083581725e-06, "loss": 0.2833, "step": 27210 }, { "epoch": 1.6099840302833146, "grad_norm": 2.1498162746429443, "learning_rate": 2.7899300988962675e-06, "loss": 0.2554, "step": 27220 }, { "epoch": 1.6105755012716627, "grad_norm": 2.460080862045288, "learning_rate": 2.788019199865473e-06, "loss": 0.274, "step": 27230 }, { "epoch": 1.6111669722600106, "grad_norm": 4.135615348815918, "learning_rate": 2.7861083872684984e-06, "loss": 0.2627, "step": 27240 }, { "epoch": 1.6117584432483587, "grad_norm": 3.7820069789886475, "learning_rate": 2.784197661884466e-06, "loss": 0.2987, "step": 27250 }, { "epoch": 1.6123499142367068, "grad_norm": 2.952387571334839, "learning_rate": 2.782287024492461e-06, "loss": 0.2714, "step": 27260 }, { "epoch": 1.6129413852250547, "grad_norm": 3.341758966445923, "learning_rate": 2.780376475871533e-06, "loss": 0.3355, "step": 27270 }, { "epoch": 1.6135328562134026, "grad_norm": 3.838465690612793, "learning_rate": 2.778466016800698e-06, "loss": 0.3308, "step": 27280 }, { "epoch": 1.614124327201751, "grad_norm": 6.473362445831299, "learning_rate": 2.7765556480589313e-06, "loss": 0.3282, "step": 27290 }, { "epoch": 1.6147157981900988, "grad_norm": 3.7356629371643066, "learning_rate": 2.7746453704251747e-06, "loss": 0.2771, "step": 27300 }, { "epoch": 1.6153072691784467, "grad_norm": 3.2335052490234375, "learning_rate": 2.772735184678332e-06, "loss": 0.3133, "step": 27310 }, { "epoch": 1.6158987401667948, "grad_norm": 3.662656784057617, "learning_rate": 2.7708250915972686e-06, "loss": 0.2685, "step": 27320 }, { "epoch": 1.616490211155143, "grad_norm": 3.4927430152893066, "learning_rate": 2.768915091960813e-06, "loss": 0.3119, "step": 27330 }, { "epoch": 1.6170816821434908, "grad_norm": 4.100736618041992, "learning_rate": 2.767005186547756e-06, "loss": 0.3316, "step": 27340 }, { "epoch": 1.617673153131839, "grad_norm": 4.451988220214844, "learning_rate": 2.7650953761368485e-06, "loss": 0.2547, "step": 27350 }, { "epoch": 1.618264624120187, "grad_norm": 2.15630841255188, "learning_rate": 2.7631856615068036e-06, "loss": 0.2727, "step": 27360 }, { "epoch": 1.618856095108535, "grad_norm": 4.191481590270996, "learning_rate": 2.761276043436296e-06, "loss": 0.2812, "step": 27370 }, { "epoch": 1.6194475660968828, "grad_norm": 4.749541759490967, "learning_rate": 2.7593665227039594e-06, "loss": 0.3437, "step": 27380 }, { "epoch": 1.6200390370852311, "grad_norm": 3.4758248329162598, "learning_rate": 2.7574571000883897e-06, "loss": 0.3295, "step": 27390 }, { "epoch": 1.620630508073579, "grad_norm": 2.8413403034210205, "learning_rate": 2.7555477763681402e-06, "loss": 0.2638, "step": 27400 }, { "epoch": 1.6212219790619269, "grad_norm": 3.8673043251037598, "learning_rate": 2.753638552321727e-06, "loss": 0.3011, "step": 27410 }, { "epoch": 1.621813450050275, "grad_norm": 2.847066879272461, "learning_rate": 2.7517294287276236e-06, "loss": 0.2594, "step": 27420 }, { "epoch": 1.622404921038623, "grad_norm": 3.3939545154571533, "learning_rate": 2.7498204063642627e-06, "loss": 0.3287, "step": 27430 }, { "epoch": 1.622996392026971, "grad_norm": 4.162200450897217, "learning_rate": 2.7479114860100364e-06, "loss": 0.3271, "step": 27440 }, { "epoch": 1.623587863015319, "grad_norm": 2.305328369140625, "learning_rate": 2.7460026684432947e-06, "loss": 0.2298, "step": 27450 }, { "epoch": 1.6241793340036672, "grad_norm": 2.7248098850250244, "learning_rate": 2.7440939544423456e-06, "loss": 0.2701, "step": 27460 }, { "epoch": 1.624770804992015, "grad_norm": 2.766946315765381, "learning_rate": 2.7421853447854553e-06, "loss": 0.2996, "step": 27470 }, { "epoch": 1.6253622759803632, "grad_norm": 1.4986969232559204, "learning_rate": 2.7402768402508472e-06, "loss": 0.3029, "step": 27480 }, { "epoch": 1.6259537469687113, "grad_norm": 3.513193368911743, "learning_rate": 2.7383684416167015e-06, "loss": 0.2889, "step": 27490 }, { "epoch": 1.6265452179570592, "grad_norm": 3.030067205429077, "learning_rate": 2.7364601496611555e-06, "loss": 0.2952, "step": 27500 }, { "epoch": 1.627136688945407, "grad_norm": 3.94041109085083, "learning_rate": 2.734551965162303e-06, "loss": 0.3578, "step": 27510 }, { "epoch": 1.6277281599337554, "grad_norm": 2.6413278579711914, "learning_rate": 2.732643888898195e-06, "loss": 0.272, "step": 27520 }, { "epoch": 1.6283196309221033, "grad_norm": 3.4660983085632324, "learning_rate": 2.7307359216468373e-06, "loss": 0.3212, "step": 27530 }, { "epoch": 1.6289111019104512, "grad_norm": 2.927711009979248, "learning_rate": 2.7288280641861903e-06, "loss": 0.3241, "step": 27540 }, { "epoch": 1.6295025728987993, "grad_norm": 5.16730260848999, "learning_rate": 2.726920317294171e-06, "loss": 0.2665, "step": 27550 }, { "epoch": 1.6300940438871474, "grad_norm": 2.7573959827423096, "learning_rate": 2.725012681748652e-06, "loss": 0.2998, "step": 27560 }, { "epoch": 1.6306855148754953, "grad_norm": 2.620607614517212, "learning_rate": 2.723105158327459e-06, "loss": 0.2723, "step": 27570 }, { "epoch": 1.6312769858638434, "grad_norm": 2.83890962600708, "learning_rate": 2.721197747808372e-06, "loss": 0.3225, "step": 27580 }, { "epoch": 1.6318684568521915, "grad_norm": 3.094550848007202, "learning_rate": 2.719290450969127e-06, "loss": 0.2763, "step": 27590 }, { "epoch": 1.6324599278405394, "grad_norm": 2.4281954765319824, "learning_rate": 2.71738326858741e-06, "loss": 0.2452, "step": 27600 }, { "epoch": 1.6330513988288873, "grad_norm": 1.5173667669296265, "learning_rate": 2.7154762014408638e-06, "loss": 0.1835, "step": 27610 }, { "epoch": 1.6336428698172356, "grad_norm": 2.0448296070098877, "learning_rate": 2.7135692503070823e-06, "loss": 0.3548, "step": 27620 }, { "epoch": 1.6342343408055835, "grad_norm": 3.5531628131866455, "learning_rate": 2.7116624159636134e-06, "loss": 0.2714, "step": 27630 }, { "epoch": 1.6348258117939314, "grad_norm": 1.8311761617660522, "learning_rate": 2.709755699187956e-06, "loss": 0.3234, "step": 27640 }, { "epoch": 1.6354172827822795, "grad_norm": 1.9411275386810303, "learning_rate": 2.7078491007575625e-06, "loss": 0.2544, "step": 27650 }, { "epoch": 1.6360087537706276, "grad_norm": 3.2396626472473145, "learning_rate": 2.7059426214498348e-06, "loss": 0.2714, "step": 27660 }, { "epoch": 1.6366002247589755, "grad_norm": 5.029738903045654, "learning_rate": 2.704036262042129e-06, "loss": 0.3016, "step": 27670 }, { "epoch": 1.6371916957473236, "grad_norm": 2.515908718109131, "learning_rate": 2.7021300233117514e-06, "loss": 0.2812, "step": 27680 }, { "epoch": 1.6377831667356717, "grad_norm": 4.265142917633057, "learning_rate": 2.700223906035957e-06, "loss": 0.3296, "step": 27690 }, { "epoch": 1.6383746377240196, "grad_norm": 1.8427376747131348, "learning_rate": 2.6983179109919535e-06, "loss": 0.2648, "step": 27700 }, { "epoch": 1.6389661087123677, "grad_norm": 3.621833324432373, "learning_rate": 2.696412038956899e-06, "loss": 0.3273, "step": 27710 }, { "epoch": 1.6395575797007158, "grad_norm": 2.4883251190185547, "learning_rate": 2.6945062907078997e-06, "loss": 0.2606, "step": 27720 }, { "epoch": 1.6401490506890637, "grad_norm": 2.1267805099487305, "learning_rate": 2.6926006670220134e-06, "loss": 0.2556, "step": 27730 }, { "epoch": 1.6407405216774116, "grad_norm": 2.3596689701080322, "learning_rate": 2.690695168676245e-06, "loss": 0.2925, "step": 27740 }, { "epoch": 1.64133199266576, "grad_norm": 4.1664137840271, "learning_rate": 2.6887897964475496e-06, "loss": 0.2643, "step": 27750 }, { "epoch": 1.6419234636541078, "grad_norm": 3.2956583499908447, "learning_rate": 2.686884551112831e-06, "loss": 0.2611, "step": 27760 }, { "epoch": 1.6425149346424557, "grad_norm": 2.829350233078003, "learning_rate": 2.68497943344894e-06, "loss": 0.2925, "step": 27770 }, { "epoch": 1.6431064056308038, "grad_norm": 3.319758176803589, "learning_rate": 2.683074444232678e-06, "loss": 0.2646, "step": 27780 }, { "epoch": 1.643697876619152, "grad_norm": 3.2266314029693604, "learning_rate": 2.68116958424079e-06, "loss": 0.237, "step": 27790 }, { "epoch": 1.6442893476074998, "grad_norm": 3.1519296169281006, "learning_rate": 2.679264854249971e-06, "loss": 0.2696, "step": 27800 }, { "epoch": 1.644880818595848, "grad_norm": 4.047759056091309, "learning_rate": 2.6773602550368636e-06, "loss": 0.3298, "step": 27810 }, { "epoch": 1.645472289584196, "grad_norm": 3.0397236347198486, "learning_rate": 2.6754557873780552e-06, "loss": 0.2181, "step": 27820 }, { "epoch": 1.6460637605725439, "grad_norm": 2.7599780559539795, "learning_rate": 2.67355145205008e-06, "loss": 0.3001, "step": 27830 }, { "epoch": 1.6466552315608918, "grad_norm": 4.2884840965271, "learning_rate": 2.6716472498294207e-06, "loss": 0.3167, "step": 27840 }, { "epoch": 1.64724670254924, "grad_norm": 4.387386322021484, "learning_rate": 2.6697431814925004e-06, "loss": 0.2882, "step": 27850 }, { "epoch": 1.647838173537588, "grad_norm": 5.223574161529541, "learning_rate": 2.6678392478156926e-06, "loss": 0.2786, "step": 27860 }, { "epoch": 1.6484296445259359, "grad_norm": 2.962374210357666, "learning_rate": 2.6659354495753144e-06, "loss": 0.2913, "step": 27870 }, { "epoch": 1.649021115514284, "grad_norm": 2.138209581375122, "learning_rate": 2.6640317875476267e-06, "loss": 0.2512, "step": 27880 }, { "epoch": 1.649612586502632, "grad_norm": 3.8437178134918213, "learning_rate": 2.6621282625088357e-06, "loss": 0.3251, "step": 27890 }, { "epoch": 1.65020405749098, "grad_norm": 3.9747731685638428, "learning_rate": 2.6602248752350932e-06, "loss": 0.2688, "step": 27900 }, { "epoch": 1.650795528479328, "grad_norm": 5.141392230987549, "learning_rate": 2.6583216265024904e-06, "loss": 0.3116, "step": 27910 }, { "epoch": 1.6513869994676762, "grad_norm": 2.391909122467041, "learning_rate": 2.6564185170870663e-06, "loss": 0.3607, "step": 27920 }, { "epoch": 1.651978470456024, "grad_norm": 1.9404668807983398, "learning_rate": 2.6545155477648016e-06, "loss": 0.3085, "step": 27930 }, { "epoch": 1.6525699414443722, "grad_norm": 2.056077718734741, "learning_rate": 2.6526127193116195e-06, "loss": 0.314, "step": 27940 }, { "epoch": 1.6531614124327203, "grad_norm": 4.330619812011719, "learning_rate": 2.650710032503386e-06, "loss": 0.3521, "step": 27950 }, { "epoch": 1.6537528834210682, "grad_norm": 4.6042962074279785, "learning_rate": 2.6488074881159105e-06, "loss": 0.2817, "step": 27960 }, { "epoch": 1.654344354409416, "grad_norm": 3.385756015777588, "learning_rate": 2.6469050869249427e-06, "loss": 0.2721, "step": 27970 }, { "epoch": 1.6549358253977644, "grad_norm": 3.4646520614624023, "learning_rate": 2.6450028297061753e-06, "loss": 0.3337, "step": 27980 }, { "epoch": 1.6555272963861123, "grad_norm": 2.108407735824585, "learning_rate": 2.6431007172352402e-06, "loss": 0.2834, "step": 27990 }, { "epoch": 1.6561187673744602, "grad_norm": 4.257559776306152, "learning_rate": 2.6411987502877122e-06, "loss": 0.3327, "step": 28000 }, { "epoch": 1.6567102383628083, "grad_norm": 5.00366735458374, "learning_rate": 2.639296929639107e-06, "loss": 0.3029, "step": 28010 }, { "epoch": 1.6573017093511564, "grad_norm": 2.1731512546539307, "learning_rate": 2.6373952560648786e-06, "loss": 0.2962, "step": 28020 }, { "epoch": 1.6578931803395043, "grad_norm": 2.8183798789978027, "learning_rate": 2.6354937303404233e-06, "loss": 0.2893, "step": 28030 }, { "epoch": 1.6584846513278524, "grad_norm": 2.492478132247925, "learning_rate": 2.633592353241076e-06, "loss": 0.2851, "step": 28040 }, { "epoch": 1.6590761223162005, "grad_norm": 3.4954121112823486, "learning_rate": 2.6316911255421104e-06, "loss": 0.2787, "step": 28050 }, { "epoch": 1.6596675933045484, "grad_norm": 4.532485008239746, "learning_rate": 2.6297900480187403e-06, "loss": 0.2882, "step": 28060 }, { "epoch": 1.6602590642928963, "grad_norm": 2.485513687133789, "learning_rate": 2.627889121446118e-06, "loss": 0.2424, "step": 28070 }, { "epoch": 1.6608505352812446, "grad_norm": 1.709669589996338, "learning_rate": 2.625988346599334e-06, "loss": 0.2749, "step": 28080 }, { "epoch": 1.6614420062695925, "grad_norm": 2.4433538913726807, "learning_rate": 2.6240877242534182e-06, "loss": 0.3018, "step": 28090 }, { "epoch": 1.6620334772579404, "grad_norm": 3.529161214828491, "learning_rate": 2.6221872551833357e-06, "loss": 0.2948, "step": 28100 }, { "epoch": 1.6626249482462885, "grad_norm": 3.8142662048339844, "learning_rate": 2.6202869401639917e-06, "loss": 0.3097, "step": 28110 }, { "epoch": 1.6632164192346366, "grad_norm": 3.301605463027954, "learning_rate": 2.6183867799702273e-06, "loss": 0.3092, "step": 28120 }, { "epoch": 1.6638078902229845, "grad_norm": 2.9797041416168213, "learning_rate": 2.61648677537682e-06, "loss": 0.2709, "step": 28130 }, { "epoch": 1.6643993612113326, "grad_norm": 3.6533310413360596, "learning_rate": 2.6145869271584846e-06, "loss": 0.3052, "step": 28140 }, { "epoch": 1.6649908321996807, "grad_norm": 3.3977978229522705, "learning_rate": 2.6126872360898734e-06, "loss": 0.2898, "step": 28150 }, { "epoch": 1.6655823031880286, "grad_norm": 3.525543212890625, "learning_rate": 2.6107877029455714e-06, "loss": 0.2945, "step": 28160 }, { "epoch": 1.6661737741763767, "grad_norm": 2.271230459213257, "learning_rate": 2.6088883285001032e-06, "loss": 0.287, "step": 28170 }, { "epoch": 1.6667652451647248, "grad_norm": 1.8311883211135864, "learning_rate": 2.6069891135279264e-06, "loss": 0.2707, "step": 28180 }, { "epoch": 1.6673567161530727, "grad_norm": 2.9287168979644775, "learning_rate": 2.6050900588034315e-06, "loss": 0.334, "step": 28190 }, { "epoch": 1.6679481871414206, "grad_norm": 2.627790689468384, "learning_rate": 2.6031911651009485e-06, "loss": 0.3117, "step": 28200 }, { "epoch": 1.668539658129769, "grad_norm": 3.992520809173584, "learning_rate": 2.6012924331947375e-06, "loss": 0.2882, "step": 28210 }, { "epoch": 1.6691311291181168, "grad_norm": 1.9149889945983887, "learning_rate": 2.5993938638589955e-06, "loss": 0.2678, "step": 28220 }, { "epoch": 1.6697226001064647, "grad_norm": 2.632460832595825, "learning_rate": 2.5974954578678522e-06, "loss": 0.249, "step": 28230 }, { "epoch": 1.6703140710948128, "grad_norm": 3.870338201522827, "learning_rate": 2.595597215995369e-06, "loss": 0.3184, "step": 28240 }, { "epoch": 1.6709055420831609, "grad_norm": 1.8554291725158691, "learning_rate": 2.5936991390155433e-06, "loss": 0.2769, "step": 28250 }, { "epoch": 1.6714970130715088, "grad_norm": 4.410476207733154, "learning_rate": 2.591801227702303e-06, "loss": 0.3389, "step": 28260 }, { "epoch": 1.6720884840598569, "grad_norm": 3.2427327632904053, "learning_rate": 2.58990348282951e-06, "loss": 0.3081, "step": 28270 }, { "epoch": 1.672679955048205, "grad_norm": 3.3633992671966553, "learning_rate": 2.5880059051709577e-06, "loss": 0.2729, "step": 28280 }, { "epoch": 1.6732714260365529, "grad_norm": 4.113663673400879, "learning_rate": 2.5861084955003717e-06, "loss": 0.3372, "step": 28290 }, { "epoch": 1.6738628970249008, "grad_norm": 1.5956708192825317, "learning_rate": 2.584211254591407e-06, "loss": 0.2887, "step": 28300 }, { "epoch": 1.674454368013249, "grad_norm": 2.4642512798309326, "learning_rate": 2.582314183217653e-06, "loss": 0.2809, "step": 28310 }, { "epoch": 1.675045839001597, "grad_norm": 3.369577407836914, "learning_rate": 2.5804172821526274e-06, "loss": 0.3196, "step": 28320 }, { "epoch": 1.6756373099899449, "grad_norm": 2.433457374572754, "learning_rate": 2.578520552169782e-06, "loss": 0.2796, "step": 28330 }, { "epoch": 1.676228780978293, "grad_norm": 2.501300811767578, "learning_rate": 2.5766239940424927e-06, "loss": 0.2561, "step": 28340 }, { "epoch": 1.676820251966641, "grad_norm": 1.5711085796356201, "learning_rate": 2.5747276085440707e-06, "loss": 0.2616, "step": 28350 }, { "epoch": 1.677411722954989, "grad_norm": 3.025919198989868, "learning_rate": 2.5728313964477557e-06, "loss": 0.2639, "step": 28360 }, { "epoch": 1.678003193943337, "grad_norm": 3.3951218128204346, "learning_rate": 2.570935358526716e-06, "loss": 0.2699, "step": 28370 }, { "epoch": 1.6785946649316852, "grad_norm": 4.270242691040039, "learning_rate": 2.5690394955540473e-06, "loss": 0.3193, "step": 28380 }, { "epoch": 1.679186135920033, "grad_norm": 3.4924771785736084, "learning_rate": 2.567143808302778e-06, "loss": 0.2816, "step": 28390 }, { "epoch": 1.6797776069083812, "grad_norm": 3.620030164718628, "learning_rate": 2.565248297545861e-06, "loss": 0.2578, "step": 28400 }, { "epoch": 1.6803690778967293, "grad_norm": 3.987103223800659, "learning_rate": 2.5633529640561786e-06, "loss": 0.2751, "step": 28410 }, { "epoch": 1.6809605488850772, "grad_norm": 4.091274261474609, "learning_rate": 2.5614578086065415e-06, "loss": 0.2717, "step": 28420 }, { "epoch": 1.681552019873425, "grad_norm": 4.102645397186279, "learning_rate": 2.559562831969688e-06, "loss": 0.3703, "step": 28430 }, { "epoch": 1.6821434908617734, "grad_norm": 4.565437316894531, "learning_rate": 2.557668034918281e-06, "loss": 0.2931, "step": 28440 }, { "epoch": 1.6827349618501213, "grad_norm": 2.581843137741089, "learning_rate": 2.555773418224913e-06, "loss": 0.2611, "step": 28450 }, { "epoch": 1.6833264328384692, "grad_norm": 1.8109240531921387, "learning_rate": 2.553878982662102e-06, "loss": 0.2204, "step": 28460 }, { "epoch": 1.6839179038268173, "grad_norm": 3.9404261112213135, "learning_rate": 2.5519847290022915e-06, "loss": 0.2902, "step": 28470 }, { "epoch": 1.6845093748151654, "grad_norm": 2.110699415206909, "learning_rate": 2.5500906580178522e-06, "loss": 0.3083, "step": 28480 }, { "epoch": 1.6851008458035133, "grad_norm": 3.9401917457580566, "learning_rate": 2.5481967704810782e-06, "loss": 0.3322, "step": 28490 }, { "epoch": 1.6856923167918614, "grad_norm": 2.29423451423645, "learning_rate": 2.546303067164191e-06, "loss": 0.286, "step": 28500 }, { "epoch": 1.6862837877802095, "grad_norm": 1.5438156127929688, "learning_rate": 2.5444095488393356e-06, "loss": 0.3143, "step": 28510 }, { "epoch": 1.6868752587685574, "grad_norm": 1.6219987869262695, "learning_rate": 2.5425162162785825e-06, "loss": 0.261, "step": 28520 }, { "epoch": 1.6874667297569053, "grad_norm": 3.0713467597961426, "learning_rate": 2.540623070253925e-06, "loss": 0.3251, "step": 28530 }, { "epoch": 1.6880582007452536, "grad_norm": 4.40298318862915, "learning_rate": 2.538730111537284e-06, "loss": 0.2709, "step": 28540 }, { "epoch": 1.6886496717336015, "grad_norm": 3.524562120437622, "learning_rate": 2.536837340900498e-06, "loss": 0.267, "step": 28550 }, { "epoch": 1.6892411427219494, "grad_norm": 4.982982158660889, "learning_rate": 2.5349447591153336e-06, "loss": 0.2826, "step": 28560 }, { "epoch": 1.6898326137102975, "grad_norm": 3.5969202518463135, "learning_rate": 2.533052366953479e-06, "loss": 0.2999, "step": 28570 }, { "epoch": 1.6904240846986456, "grad_norm": 2.992112636566162, "learning_rate": 2.5311601651865444e-06, "loss": 0.3072, "step": 28580 }, { "epoch": 1.6910155556869935, "grad_norm": 1.7316465377807617, "learning_rate": 2.529268154586064e-06, "loss": 0.3162, "step": 28590 }, { "epoch": 1.6916070266753416, "grad_norm": 3.812333106994629, "learning_rate": 2.5273763359234925e-06, "loss": 0.2755, "step": 28600 }, { "epoch": 1.6921984976636897, "grad_norm": 4.319789409637451, "learning_rate": 2.525484709970207e-06, "loss": 0.2947, "step": 28610 }, { "epoch": 1.6927899686520376, "grad_norm": 3.9880471229553223, "learning_rate": 2.5235932774975064e-06, "loss": 0.2833, "step": 28620 }, { "epoch": 1.6933814396403857, "grad_norm": 3.436004161834717, "learning_rate": 2.521702039276609e-06, "loss": 0.2503, "step": 28630 }, { "epoch": 1.6939729106287338, "grad_norm": 2.3767380714416504, "learning_rate": 2.519810996078656e-06, "loss": 0.2968, "step": 28640 }, { "epoch": 1.6945643816170817, "grad_norm": 2.655344009399414, "learning_rate": 2.517920148674708e-06, "loss": 0.2885, "step": 28650 }, { "epoch": 1.6951558526054296, "grad_norm": 3.428213119506836, "learning_rate": 2.516029497835746e-06, "loss": 0.2789, "step": 28660 }, { "epoch": 1.6957473235937779, "grad_norm": 1.9561350345611572, "learning_rate": 2.5141390443326713e-06, "loss": 0.3492, "step": 28670 }, { "epoch": 1.6963387945821258, "grad_norm": 2.44429087638855, "learning_rate": 2.512248788936304e-06, "loss": 0.2682, "step": 28680 }, { "epoch": 1.6969302655704737, "grad_norm": 3.539679765701294, "learning_rate": 2.510358732417383e-06, "loss": 0.3033, "step": 28690 }, { "epoch": 1.6975217365588218, "grad_norm": 3.2544243335723877, "learning_rate": 2.5084688755465676e-06, "loss": 0.2769, "step": 28700 }, { "epoch": 1.6981132075471699, "grad_norm": 1.378494381904602, "learning_rate": 2.506579219094435e-06, "loss": 0.2949, "step": 28710 }, { "epoch": 1.6987046785355178, "grad_norm": 3.459803581237793, "learning_rate": 2.504689763831479e-06, "loss": 0.2314, "step": 28720 }, { "epoch": 1.6992961495238659, "grad_norm": 4.009194374084473, "learning_rate": 2.5028005105281152e-06, "loss": 0.2622, "step": 28730 }, { "epoch": 1.699887620512214, "grad_norm": 3.0859313011169434, "learning_rate": 2.5009114599546737e-06, "loss": 0.2836, "step": 28740 }, { "epoch": 1.7004790915005619, "grad_norm": 3.183560609817505, "learning_rate": 2.499022612881402e-06, "loss": 0.3042, "step": 28750 }, { "epoch": 1.7010705624889098, "grad_norm": 3.1072540283203125, "learning_rate": 2.497133970078468e-06, "loss": 0.2752, "step": 28760 }, { "epoch": 1.701662033477258, "grad_norm": 1.9240405559539795, "learning_rate": 2.49524553231595e-06, "loss": 0.2961, "step": 28770 }, { "epoch": 1.702253504465606, "grad_norm": 4.1512451171875, "learning_rate": 2.4933573003638483e-06, "loss": 0.3538, "step": 28780 }, { "epoch": 1.7028449754539539, "grad_norm": 2.95644211769104, "learning_rate": 2.4914692749920775e-06, "loss": 0.2475, "step": 28790 }, { "epoch": 1.703436446442302, "grad_norm": 3.6679043769836426, "learning_rate": 2.489581456970468e-06, "loss": 0.2963, "step": 28800 }, { "epoch": 1.70402791743065, "grad_norm": 3.896171808242798, "learning_rate": 2.4876938470687652e-06, "loss": 0.2945, "step": 28810 }, { "epoch": 1.704619388418998, "grad_norm": 3.1912474632263184, "learning_rate": 2.4858064460566312e-06, "loss": 0.2611, "step": 28820 }, { "epoch": 1.705210859407346, "grad_norm": 3.1870057582855225, "learning_rate": 2.4839192547036402e-06, "loss": 0.3169, "step": 28830 }, { "epoch": 1.7058023303956942, "grad_norm": 3.1682639122009277, "learning_rate": 2.482032273779283e-06, "loss": 0.2746, "step": 28840 }, { "epoch": 1.706393801384042, "grad_norm": 2.886059284210205, "learning_rate": 2.4801455040529638e-06, "loss": 0.2822, "step": 28850 }, { "epoch": 1.7069852723723902, "grad_norm": 3.706965446472168, "learning_rate": 2.4782589462940013e-06, "loss": 0.3739, "step": 28860 }, { "epoch": 1.7075767433607383, "grad_norm": 3.7082066535949707, "learning_rate": 2.4763726012716275e-06, "loss": 0.2728, "step": 28870 }, { "epoch": 1.7081682143490862, "grad_norm": 3.328829050064087, "learning_rate": 2.4744864697549877e-06, "loss": 0.2418, "step": 28880 }, { "epoch": 1.708759685337434, "grad_norm": 1.9696592092514038, "learning_rate": 2.472600552513139e-06, "loss": 0.255, "step": 28890 }, { "epoch": 1.7093511563257824, "grad_norm": 4.096414089202881, "learning_rate": 2.470714850315053e-06, "loss": 0.3335, "step": 28900 }, { "epoch": 1.7099426273141303, "grad_norm": 4.327395915985107, "learning_rate": 2.4688293639296126e-06, "loss": 0.2818, "step": 28910 }, { "epoch": 1.7105340983024782, "grad_norm": 3.8213493824005127, "learning_rate": 2.466944094125612e-06, "loss": 0.2804, "step": 28920 }, { "epoch": 1.7111255692908263, "grad_norm": 2.697415351867676, "learning_rate": 2.4650590416717605e-06, "loss": 0.25, "step": 28930 }, { "epoch": 1.7117170402791744, "grad_norm": 3.675581455230713, "learning_rate": 2.4631742073366733e-06, "loss": 0.2663, "step": 28940 }, { "epoch": 1.7123085112675223, "grad_norm": 3.2869863510131836, "learning_rate": 2.4612895918888803e-06, "loss": 0.298, "step": 28950 }, { "epoch": 1.7128999822558704, "grad_norm": 3.152714252471924, "learning_rate": 2.4594051960968226e-06, "loss": 0.3294, "step": 28960 }, { "epoch": 1.7134914532442185, "grad_norm": 3.74511456489563, "learning_rate": 2.4575210207288505e-06, "loss": 0.2938, "step": 28970 }, { "epoch": 1.7140829242325664, "grad_norm": 5.124378204345703, "learning_rate": 2.455637066553223e-06, "loss": 0.3094, "step": 28980 }, { "epoch": 1.7146743952209142, "grad_norm": 2.1031429767608643, "learning_rate": 2.453753334338111e-06, "loss": 0.2615, "step": 28990 }, { "epoch": 1.7152658662092626, "grad_norm": 2.5339677333831787, "learning_rate": 2.4518698248515946e-06, "loss": 0.2458, "step": 29000 }, { "epoch": 1.7158573371976105, "grad_norm": 3.1010892391204834, "learning_rate": 2.4499865388616632e-06, "loss": 0.296, "step": 29010 }, { "epoch": 1.7164488081859584, "grad_norm": 3.6202127933502197, "learning_rate": 2.448103477136214e-06, "loss": 0.2702, "step": 29020 }, { "epoch": 1.7170402791743065, "grad_norm": 2.707664728164673, "learning_rate": 2.446220640443053e-06, "loss": 0.3, "step": 29030 }, { "epoch": 1.7176317501626546, "grad_norm": 2.9525976181030273, "learning_rate": 2.4443380295498957e-06, "loss": 0.3295, "step": 29040 }, { "epoch": 1.7182232211510025, "grad_norm": 2.5064849853515625, "learning_rate": 2.4424556452243645e-06, "loss": 0.3178, "step": 29050 }, { "epoch": 1.7188146921393506, "grad_norm": 2.8300437927246094, "learning_rate": 2.44057348823399e-06, "loss": 0.274, "step": 29060 }, { "epoch": 1.7194061631276987, "grad_norm": 2.102660894393921, "learning_rate": 2.438691559346209e-06, "loss": 0.2624, "step": 29070 }, { "epoch": 1.7199976341160466, "grad_norm": 2.96524977684021, "learning_rate": 2.4368098593283663e-06, "loss": 0.2949, "step": 29080 }, { "epoch": 1.7205891051043947, "grad_norm": 2.6221256256103516, "learning_rate": 2.434928388947713e-06, "loss": 0.2602, "step": 29090 }, { "epoch": 1.7211805760927428, "grad_norm": 3.418853998184204, "learning_rate": 2.4330471489714064e-06, "loss": 0.3302, "step": 29100 }, { "epoch": 1.7217720470810907, "grad_norm": 2.59032940864563, "learning_rate": 2.4311661401665103e-06, "loss": 0.2529, "step": 29110 }, { "epoch": 1.7223635180694385, "grad_norm": 6.3017191886901855, "learning_rate": 2.4292853632999946e-06, "loss": 0.3065, "step": 29120 }, { "epoch": 1.7229549890577869, "grad_norm": 4.82774019241333, "learning_rate": 2.427404819138734e-06, "loss": 0.2638, "step": 29130 }, { "epoch": 1.7235464600461348, "grad_norm": 3.493204355239868, "learning_rate": 2.4255245084495075e-06, "loss": 0.2691, "step": 29140 }, { "epoch": 1.7241379310344827, "grad_norm": 2.226330041885376, "learning_rate": 2.4236444319990003e-06, "loss": 0.2633, "step": 29150 }, { "epoch": 1.7247294020228308, "grad_norm": 3.344409942626953, "learning_rate": 2.4217645905538015e-06, "loss": 0.2924, "step": 29160 }, { "epoch": 1.7253208730111789, "grad_norm": 2.465085983276367, "learning_rate": 2.419884984880405e-06, "loss": 0.3135, "step": 29170 }, { "epoch": 1.7259123439995268, "grad_norm": 2.091506242752075, "learning_rate": 2.4180056157452074e-06, "loss": 0.2861, "step": 29180 }, { "epoch": 1.7265038149878749, "grad_norm": 4.238035678863525, "learning_rate": 2.416126483914511e-06, "loss": 0.294, "step": 29190 }, { "epoch": 1.727095285976223, "grad_norm": 4.713216781616211, "learning_rate": 2.414247590154517e-06, "loss": 0.2605, "step": 29200 }, { "epoch": 1.7276867569645709, "grad_norm": 4.343008995056152, "learning_rate": 2.412368935231335e-06, "loss": 0.3011, "step": 29210 }, { "epoch": 1.7282782279529187, "grad_norm": 2.692474365234375, "learning_rate": 2.410490519910972e-06, "loss": 0.2899, "step": 29220 }, { "epoch": 1.728869698941267, "grad_norm": 3.1082050800323486, "learning_rate": 2.408612344959342e-06, "loss": 0.2798, "step": 29230 }, { "epoch": 1.729461169929615, "grad_norm": 2.671060562133789, "learning_rate": 2.4067344111422577e-06, "loss": 0.2654, "step": 29240 }, { "epoch": 1.7300526409179628, "grad_norm": 3.5150094032287598, "learning_rate": 2.4048567192254356e-06, "loss": 0.2613, "step": 29250 }, { "epoch": 1.730644111906311, "grad_norm": 2.7770981788635254, "learning_rate": 2.402979269974492e-06, "loss": 0.293, "step": 29260 }, { "epoch": 1.731235582894659, "grad_norm": 3.4728918075561523, "learning_rate": 2.4011020641549457e-06, "loss": 0.3479, "step": 29270 }, { "epoch": 1.731827053883007, "grad_norm": 2.643918752670288, "learning_rate": 2.3992251025322137e-06, "loss": 0.2422, "step": 29280 }, { "epoch": 1.732418524871355, "grad_norm": 1.8970947265625, "learning_rate": 2.3973483858716164e-06, "loss": 0.3022, "step": 29290 }, { "epoch": 1.7330099958597032, "grad_norm": 3.219374895095825, "learning_rate": 2.395471914938373e-06, "loss": 0.2369, "step": 29300 }, { "epoch": 1.733601466848051, "grad_norm": 6.299752712249756, "learning_rate": 2.393595690497602e-06, "loss": 0.2342, "step": 29310 }, { "epoch": 1.7341929378363992, "grad_norm": 4.149672508239746, "learning_rate": 2.3917197133143226e-06, "loss": 0.2897, "step": 29320 }, { "epoch": 1.7347844088247473, "grad_norm": 2.3528900146484375, "learning_rate": 2.389843984153452e-06, "loss": 0.2899, "step": 29330 }, { "epoch": 1.7353758798130952, "grad_norm": 2.019463300704956, "learning_rate": 2.3879685037798074e-06, "loss": 0.2469, "step": 29340 }, { "epoch": 1.735967350801443, "grad_norm": 1.6167330741882324, "learning_rate": 2.386093272958103e-06, "loss": 0.213, "step": 29350 }, { "epoch": 1.7365588217897914, "grad_norm": 3.4094157218933105, "learning_rate": 2.3842182924529532e-06, "loss": 0.3575, "step": 29360 }, { "epoch": 1.7371502927781393, "grad_norm": 4.93418550491333, "learning_rate": 2.382343563028869e-06, "loss": 0.3345, "step": 29370 }, { "epoch": 1.7377417637664871, "grad_norm": 3.754167079925537, "learning_rate": 2.3804690854502597e-06, "loss": 0.3256, "step": 29380 }, { "epoch": 1.7383332347548353, "grad_norm": 3.265349864959717, "learning_rate": 2.3785948604814306e-06, "loss": 0.2868, "step": 29390 }, { "epoch": 1.7389247057431834, "grad_norm": 2.6782124042510986, "learning_rate": 2.3767208888865864e-06, "loss": 0.2308, "step": 29400 }, { "epoch": 1.7395161767315312, "grad_norm": 2.4394161701202393, "learning_rate": 2.3748471714298267e-06, "loss": 0.2473, "step": 29410 }, { "epoch": 1.7401076477198794, "grad_norm": 3.125577688217163, "learning_rate": 2.3729737088751463e-06, "loss": 0.3056, "step": 29420 }, { "epoch": 1.7406991187082275, "grad_norm": 2.5684547424316406, "learning_rate": 2.371100501986439e-06, "loss": 0.2697, "step": 29430 }, { "epoch": 1.7412905896965754, "grad_norm": 3.3113081455230713, "learning_rate": 2.3692275515274925e-06, "loss": 0.2543, "step": 29440 }, { "epoch": 1.7418820606849232, "grad_norm": 4.145740032196045, "learning_rate": 2.36735485826199e-06, "loss": 0.335, "step": 29450 }, { "epoch": 1.7424735316732716, "grad_norm": 3.372393846511841, "learning_rate": 2.3654824229535122e-06, "loss": 0.2758, "step": 29460 }, { "epoch": 1.7430650026616195, "grad_norm": 2.589458703994751, "learning_rate": 2.363610246365531e-06, "loss": 0.3121, "step": 29470 }, { "epoch": 1.7436564736499673, "grad_norm": 3.6435904502868652, "learning_rate": 2.361738329261414e-06, "loss": 0.2998, "step": 29480 }, { "epoch": 1.7442479446383155, "grad_norm": 3.6443562507629395, "learning_rate": 2.3598666724044246e-06, "loss": 0.2633, "step": 29490 }, { "epoch": 1.7448394156266636, "grad_norm": 4.106545925140381, "learning_rate": 2.357995276557719e-06, "loss": 0.3048, "step": 29500 }, { "epoch": 1.7454308866150114, "grad_norm": 3.219784736633301, "learning_rate": 2.356124142484346e-06, "loss": 0.3033, "step": 29510 }, { "epoch": 1.7460223576033596, "grad_norm": 3.6249139308929443, "learning_rate": 2.35425327094725e-06, "loss": 0.2579, "step": 29520 }, { "epoch": 1.7466138285917077, "grad_norm": 2.8021368980407715, "learning_rate": 2.3523826627092653e-06, "loss": 0.3712, "step": 29530 }, { "epoch": 1.7472052995800555, "grad_norm": 3.4311647415161133, "learning_rate": 2.350512318533121e-06, "loss": 0.2804, "step": 29540 }, { "epoch": 1.7477967705684037, "grad_norm": 3.862765312194824, "learning_rate": 2.348642239181439e-06, "loss": 0.2973, "step": 29550 }, { "epoch": 1.7483882415567518, "grad_norm": 3.880857229232788, "learning_rate": 2.346772425416731e-06, "loss": 0.2281, "step": 29560 }, { "epoch": 1.7489797125450997, "grad_norm": 3.4554073810577393, "learning_rate": 2.344902878001402e-06, "loss": 0.2456, "step": 29570 }, { "epoch": 1.7495711835334475, "grad_norm": 2.720250129699707, "learning_rate": 2.343033597697749e-06, "loss": 0.273, "step": 29580 }, { "epoch": 1.7501626545217959, "grad_norm": 3.604426860809326, "learning_rate": 2.341164585267958e-06, "loss": 0.303, "step": 29590 }, { "epoch": 1.7507541255101438, "grad_norm": 4.899552345275879, "learning_rate": 2.3392958414741066e-06, "loss": 0.309, "step": 29600 }, { "epoch": 1.7513455964984916, "grad_norm": 3.591465473175049, "learning_rate": 2.337427367078164e-06, "loss": 0.3009, "step": 29610 }, { "epoch": 1.7519370674868397, "grad_norm": 1.9740208387374878, "learning_rate": 2.3355591628419897e-06, "loss": 0.2936, "step": 29620 }, { "epoch": 1.7525285384751879, "grad_norm": 2.841811180114746, "learning_rate": 2.3336912295273297e-06, "loss": 0.2954, "step": 29630 }, { "epoch": 1.7531200094635357, "grad_norm": 3.652050256729126, "learning_rate": 2.331823567895823e-06, "loss": 0.2933, "step": 29640 }, { "epoch": 1.7537114804518839, "grad_norm": 3.4950790405273438, "learning_rate": 2.329956178708997e-06, "loss": 0.2695, "step": 29650 }, { "epoch": 1.754302951440232, "grad_norm": 4.056483745574951, "learning_rate": 2.3280890627282676e-06, "loss": 0.3167, "step": 29660 }, { "epoch": 1.7548944224285798, "grad_norm": 3.698537826538086, "learning_rate": 2.32622222071494e-06, "loss": 0.2781, "step": 29670 }, { "epoch": 1.7554858934169277, "grad_norm": 5.147281646728516, "learning_rate": 2.324355653430205e-06, "loss": 0.3183, "step": 29680 }, { "epoch": 1.756077364405276, "grad_norm": 2.9327213764190674, "learning_rate": 2.3224893616351456e-06, "loss": 0.2811, "step": 29690 }, { "epoch": 1.756668835393624, "grad_norm": 2.863604784011841, "learning_rate": 2.32062334609073e-06, "loss": 0.2764, "step": 29700 }, { "epoch": 1.7572603063819718, "grad_norm": 2.17572021484375, "learning_rate": 2.3187576075578146e-06, "loss": 0.2697, "step": 29710 }, { "epoch": 1.75785177737032, "grad_norm": 2.355447769165039, "learning_rate": 2.3168921467971405e-06, "loss": 0.2555, "step": 29720 }, { "epoch": 1.758443248358668, "grad_norm": 4.320287704467773, "learning_rate": 2.315026964569339e-06, "loss": 0.3602, "step": 29730 }, { "epoch": 1.759034719347016, "grad_norm": 2.3165433406829834, "learning_rate": 2.313162061634926e-06, "loss": 0.3185, "step": 29740 }, { "epoch": 1.759626190335364, "grad_norm": 2.842963218688965, "learning_rate": 2.3112974387543036e-06, "loss": 0.3312, "step": 29750 }, { "epoch": 1.7602176613237122, "grad_norm": 2.768099546432495, "learning_rate": 2.30943309668776e-06, "loss": 0.3378, "step": 29760 }, { "epoch": 1.76080913231206, "grad_norm": 3.9438934326171875, "learning_rate": 2.3075690361954694e-06, "loss": 0.3348, "step": 29770 }, { "epoch": 1.7614006033004082, "grad_norm": 2.028867483139038, "learning_rate": 2.3057052580374893e-06, "loss": 0.2476, "step": 29780 }, { "epoch": 1.7619920742887563, "grad_norm": 2.259089946746826, "learning_rate": 2.303841762973764e-06, "loss": 0.2919, "step": 29790 }, { "epoch": 1.7625835452771041, "grad_norm": 2.0420799255371094, "learning_rate": 2.301978551764122e-06, "loss": 0.2641, "step": 29800 }, { "epoch": 1.763175016265452, "grad_norm": 4.046327590942383, "learning_rate": 2.3001156251682754e-06, "loss": 0.3311, "step": 29810 }, { "epoch": 1.7637664872538004, "grad_norm": 1.7474772930145264, "learning_rate": 2.298252983945821e-06, "loss": 0.257, "step": 29820 }, { "epoch": 1.7643579582421482, "grad_norm": 2.6133594512939453, "learning_rate": 2.29639062885624e-06, "loss": 0.327, "step": 29830 }, { "epoch": 1.7649494292304961, "grad_norm": 1.9355981349945068, "learning_rate": 2.2945285606588926e-06, "loss": 0.3055, "step": 29840 }, { "epoch": 1.7655409002188442, "grad_norm": 3.278437376022339, "learning_rate": 2.292666780113028e-06, "loss": 0.2776, "step": 29850 }, { "epoch": 1.7661323712071924, "grad_norm": 3.607677936553955, "learning_rate": 2.2908052879777737e-06, "loss": 0.2648, "step": 29860 }, { "epoch": 1.7667238421955402, "grad_norm": 4.103071689605713, "learning_rate": 2.2889440850121426e-06, "loss": 0.3224, "step": 29870 }, { "epoch": 1.7673153131838883, "grad_norm": 3.6330056190490723, "learning_rate": 2.287083171975027e-06, "loss": 0.3197, "step": 29880 }, { "epoch": 1.7679067841722365, "grad_norm": 3.780383586883545, "learning_rate": 2.285222549625204e-06, "loss": 0.3102, "step": 29890 }, { "epoch": 1.7684982551605843, "grad_norm": 1.8640971183776855, "learning_rate": 2.283362218721329e-06, "loss": 0.3225, "step": 29900 }, { "epoch": 1.7690897261489322, "grad_norm": 4.575825214385986, "learning_rate": 2.2815021800219415e-06, "loss": 0.3301, "step": 29910 }, { "epoch": 1.7696811971372806, "grad_norm": 3.329057216644287, "learning_rate": 2.279642434285459e-06, "loss": 0.2642, "step": 29920 }, { "epoch": 1.7702726681256284, "grad_norm": 4.214750289916992, "learning_rate": 2.2777829822701814e-06, "loss": 0.3294, "step": 29930 }, { "epoch": 1.7708641391139763, "grad_norm": 3.0933175086975098, "learning_rate": 2.2759238247342887e-06, "loss": 0.2823, "step": 29940 }, { "epoch": 1.7714556101023244, "grad_norm": 2.778177499771118, "learning_rate": 2.27406496243584e-06, "loss": 0.2488, "step": 29950 }, { "epoch": 1.7720470810906725, "grad_norm": 2.3520700931549072, "learning_rate": 2.2722063961327753e-06, "loss": 0.2934, "step": 29960 }, { "epoch": 1.7726385520790204, "grad_norm": 2.8480405807495117, "learning_rate": 2.2703481265829137e-06, "loss": 0.2601, "step": 29970 }, { "epoch": 1.7732300230673685, "grad_norm": 3.5802254676818848, "learning_rate": 2.268490154543951e-06, "loss": 0.3012, "step": 29980 }, { "epoch": 1.7738214940557167, "grad_norm": 2.3734898567199707, "learning_rate": 2.266632480773464e-06, "loss": 0.2981, "step": 29990 }, { "epoch": 1.7744129650440645, "grad_norm": 1.5414009094238281, "learning_rate": 2.264775106028908e-06, "loss": 0.2089, "step": 30000 }, { "epoch": 1.7750044360324126, "grad_norm": 2.9623701572418213, "learning_rate": 2.2629180310676155e-06, "loss": 0.2587, "step": 30010 }, { "epoch": 1.7755959070207608, "grad_norm": 2.9688005447387695, "learning_rate": 2.2610612566467972e-06, "loss": 0.2455, "step": 30020 }, { "epoch": 1.7761873780091086, "grad_norm": 2.933403491973877, "learning_rate": 2.25920478352354e-06, "loss": 0.2857, "step": 30030 }, { "epoch": 1.7767788489974565, "grad_norm": 3.815114974975586, "learning_rate": 2.25734861245481e-06, "loss": 0.2864, "step": 30040 }, { "epoch": 1.7773703199858049, "grad_norm": 3.014711618423462, "learning_rate": 2.2554927441974495e-06, "loss": 0.2831, "step": 30050 }, { "epoch": 1.7779617909741527, "grad_norm": 3.708507537841797, "learning_rate": 2.2536371795081755e-06, "loss": 0.3087, "step": 30060 }, { "epoch": 1.7785532619625006, "grad_norm": 5.29056978225708, "learning_rate": 2.251781919143583e-06, "loss": 0.2646, "step": 30070 }, { "epoch": 1.7791447329508487, "grad_norm": 3.03767728805542, "learning_rate": 2.2499269638601432e-06, "loss": 0.2998, "step": 30080 }, { "epoch": 1.7797362039391968, "grad_norm": 2.8139336109161377, "learning_rate": 2.2480723144142024e-06, "loss": 0.2932, "step": 30090 }, { "epoch": 1.7803276749275447, "grad_norm": 4.393672943115234, "learning_rate": 2.2462179715619812e-06, "loss": 0.3536, "step": 30100 }, { "epoch": 1.7809191459158928, "grad_norm": 3.3109993934631348, "learning_rate": 2.244363936059578e-06, "loss": 0.2634, "step": 30110 }, { "epoch": 1.781510616904241, "grad_norm": 3.501690149307251, "learning_rate": 2.242510208662961e-06, "loss": 0.3393, "step": 30120 }, { "epoch": 1.7821020878925888, "grad_norm": 1.939081072807312, "learning_rate": 2.240656790127978e-06, "loss": 0.3182, "step": 30130 }, { "epoch": 1.7826935588809367, "grad_norm": 3.7422070503234863, "learning_rate": 2.2388036812103477e-06, "loss": 0.3372, "step": 30140 }, { "epoch": 1.783285029869285, "grad_norm": 2.3048229217529297, "learning_rate": 2.2369508826656636e-06, "loss": 0.2804, "step": 30150 }, { "epoch": 1.783876500857633, "grad_norm": 2.935603141784668, "learning_rate": 2.235098395249394e-06, "loss": 0.2546, "step": 30160 }, { "epoch": 1.7844679718459808, "grad_norm": 3.4439032077789307, "learning_rate": 2.233246219716876e-06, "loss": 0.2548, "step": 30170 }, { "epoch": 1.785059442834329, "grad_norm": 3.6633079051971436, "learning_rate": 2.2313943568233234e-06, "loss": 0.3334, "step": 30180 }, { "epoch": 1.785650913822677, "grad_norm": 2.5521652698516846, "learning_rate": 2.2295428073238228e-06, "loss": 0.3161, "step": 30190 }, { "epoch": 1.786242384811025, "grad_norm": 3.4835686683654785, "learning_rate": 2.22769157197333e-06, "loss": 0.335, "step": 30200 }, { "epoch": 1.786833855799373, "grad_norm": 3.5359132289886475, "learning_rate": 2.2258406515266755e-06, "loss": 0.2454, "step": 30210 }, { "epoch": 1.7874253267877211, "grad_norm": 3.1973226070404053, "learning_rate": 2.2239900467385605e-06, "loss": 0.3377, "step": 30220 }, { "epoch": 1.788016797776069, "grad_norm": 3.1075003147125244, "learning_rate": 2.2221397583635566e-06, "loss": 0.2277, "step": 30230 }, { "epoch": 1.7886082687644171, "grad_norm": 3.146876811981201, "learning_rate": 2.220289787156107e-06, "loss": 0.2904, "step": 30240 }, { "epoch": 1.7891997397527653, "grad_norm": 1.5939435958862305, "learning_rate": 2.2184401338705254e-06, "loss": 0.2418, "step": 30250 }, { "epoch": 1.7897912107411131, "grad_norm": 1.886264681816101, "learning_rate": 2.2165907992609982e-06, "loss": 0.2632, "step": 30260 }, { "epoch": 1.790382681729461, "grad_norm": 3.564622402191162, "learning_rate": 2.2147417840815774e-06, "loss": 0.2839, "step": 30270 }, { "epoch": 1.7909741527178094, "grad_norm": 3.0356664657592773, "learning_rate": 2.212893089086188e-06, "loss": 0.319, "step": 30280 }, { "epoch": 1.7915656237061572, "grad_norm": 3.152024984359741, "learning_rate": 2.2110447150286236e-06, "loss": 0.2629, "step": 30290 }, { "epoch": 1.7921570946945051, "grad_norm": 3.69534969329834, "learning_rate": 2.2091966626625474e-06, "loss": 0.247, "step": 30300 }, { "epoch": 1.7927485656828532, "grad_norm": 2.8451879024505615, "learning_rate": 2.2073489327414897e-06, "loss": 0.2832, "step": 30310 }, { "epoch": 1.7933400366712013, "grad_norm": 6.298223495483398, "learning_rate": 2.205501526018852e-06, "loss": 0.3259, "step": 30320 }, { "epoch": 1.7939315076595492, "grad_norm": 3.1582601070404053, "learning_rate": 2.2036544432479018e-06, "loss": 0.293, "step": 30330 }, { "epoch": 1.7945229786478973, "grad_norm": 2.531764268875122, "learning_rate": 2.2018076851817755e-06, "loss": 0.283, "step": 30340 }, { "epoch": 1.7951144496362454, "grad_norm": 2.4230453968048096, "learning_rate": 2.199961252573477e-06, "loss": 0.2702, "step": 30350 }, { "epoch": 1.7957059206245933, "grad_norm": 2.4228334426879883, "learning_rate": 2.1981151461758778e-06, "loss": 0.2952, "step": 30360 }, { "epoch": 1.7962973916129412, "grad_norm": 2.8123292922973633, "learning_rate": 2.1962693667417155e-06, "loss": 0.3382, "step": 30370 }, { "epoch": 1.7968888626012895, "grad_norm": 2.3525164127349854, "learning_rate": 2.1944239150235947e-06, "loss": 0.311, "step": 30380 }, { "epoch": 1.7974803335896374, "grad_norm": 3.1808724403381348, "learning_rate": 2.192578791773987e-06, "loss": 0.2528, "step": 30390 }, { "epoch": 1.7980718045779853, "grad_norm": 2.925367832183838, "learning_rate": 2.1907339977452298e-06, "loss": 0.2815, "step": 30400 }, { "epoch": 1.7986632755663334, "grad_norm": 4.159297943115234, "learning_rate": 2.188889533689526e-06, "loss": 0.3045, "step": 30410 }, { "epoch": 1.7992547465546815, "grad_norm": 2.79315185546875, "learning_rate": 2.1870454003589437e-06, "loss": 0.2572, "step": 30420 }, { "epoch": 1.7998462175430294, "grad_norm": 4.891920566558838, "learning_rate": 2.1852015985054164e-06, "loss": 0.2852, "step": 30430 }, { "epoch": 1.800319394333708, "eval_accuracy": 0.8647236916525268, "eval_loss": 0.3173646628856659, "eval_runtime": 158.912, "eval_safe_aucpr": 0.9239692010628959, "eval_safe_f1": 0.8511168070303918, "eval_safe_fpr": 0.14089264894920886, "eval_safe_precision": 0.8314196802231999, "eval_safe_recall": 0.8717698683568991, "eval_samples_per_second": 378.285, "eval_steps_per_second": 47.29, "eval_unsafe_aucpr": 0.9567548849415166, "eval_unsafe_f1": 0.8760517010120717, "eval_unsafe_fpr": 0.12823013164310046, "eval_unsafe_precision": 0.8936778928382623, "eval_unsafe_recall": 0.8591073510507907, "step": 30438 }, { "epoch": 1.8004376885313775, "grad_norm": 2.004011392593384, "learning_rate": 2.183358128880743e-06, "loss": 0.2379, "step": 30440 }, { "epoch": 1.8010291595197256, "grad_norm": 2.915527582168579, "learning_rate": 2.1815149922365864e-06, "loss": 0.2392, "step": 30450 }, { "epoch": 1.8016206305080735, "grad_norm": 3.6409828662872314, "learning_rate": 2.179672189324474e-06, "loss": 0.3259, "step": 30460 }, { "epoch": 1.8022121014964216, "grad_norm": 3.2273142337799072, "learning_rate": 2.177829720895797e-06, "loss": 0.2865, "step": 30470 }, { "epoch": 1.8028035724847697, "grad_norm": 3.511944055557251, "learning_rate": 2.17598758770181e-06, "loss": 0.3056, "step": 30480 }, { "epoch": 1.8033950434731176, "grad_norm": 2.3051671981811523, "learning_rate": 2.1741457904936304e-06, "loss": 0.2844, "step": 30490 }, { "epoch": 1.8039865144614655, "grad_norm": 3.492067337036133, "learning_rate": 2.17230433002224e-06, "loss": 0.2563, "step": 30500 }, { "epoch": 1.8045779854498138, "grad_norm": 3.2795605659484863, "learning_rate": 2.1704632070384822e-06, "loss": 0.2677, "step": 30510 }, { "epoch": 1.8051694564381617, "grad_norm": 3.1838953495025635, "learning_rate": 2.1686224222930633e-06, "loss": 0.259, "step": 30520 }, { "epoch": 1.8057609274265096, "grad_norm": 1.7665600776672363, "learning_rate": 2.1667819765365513e-06, "loss": 0.3455, "step": 30530 }, { "epoch": 1.8063523984148577, "grad_norm": 2.4395759105682373, "learning_rate": 2.164941870519377e-06, "loss": 0.2668, "step": 30540 }, { "epoch": 1.8069438694032058, "grad_norm": 4.8868231773376465, "learning_rate": 2.1631021049918316e-06, "loss": 0.2867, "step": 30550 }, { "epoch": 1.8075353403915537, "grad_norm": 2.8763577938079834, "learning_rate": 2.161262680704067e-06, "loss": 0.2565, "step": 30560 }, { "epoch": 1.8081268113799018, "grad_norm": 2.3862383365631104, "learning_rate": 2.1594235984060975e-06, "loss": 0.2271, "step": 30570 }, { "epoch": 1.80871828236825, "grad_norm": 3.7026405334472656, "learning_rate": 2.1575848588477968e-06, "loss": 0.3012, "step": 30580 }, { "epoch": 1.8093097533565978, "grad_norm": 3.2478723526000977, "learning_rate": 2.1557464627788995e-06, "loss": 0.3618, "step": 30590 }, { "epoch": 1.809901224344946, "grad_norm": 1.9452104568481445, "learning_rate": 2.153908410949e-06, "loss": 0.2611, "step": 30600 }, { "epoch": 1.810492695333294, "grad_norm": 2.4545698165893555, "learning_rate": 2.152070704107552e-06, "loss": 0.215, "step": 30610 }, { "epoch": 1.811084166321642, "grad_norm": 2.884474754333496, "learning_rate": 2.150233343003869e-06, "loss": 0.2563, "step": 30620 }, { "epoch": 1.8116756373099898, "grad_norm": 3.09951114654541, "learning_rate": 2.1483963283871223e-06, "loss": 0.3077, "step": 30630 }, { "epoch": 1.812267108298338, "grad_norm": 4.532462120056152, "learning_rate": 2.146559661006344e-06, "loss": 0.3112, "step": 30640 }, { "epoch": 1.812858579286686, "grad_norm": 3.5505104064941406, "learning_rate": 2.144723341610423e-06, "loss": 0.3054, "step": 30650 }, { "epoch": 1.813450050275034, "grad_norm": 3.5509190559387207, "learning_rate": 2.142887370948107e-06, "loss": 0.2786, "step": 30660 }, { "epoch": 1.814041521263382, "grad_norm": 3.093444347381592, "learning_rate": 2.1410517497680017e-06, "loss": 0.2865, "step": 30670 }, { "epoch": 1.8146329922517301, "grad_norm": 2.851574420928955, "learning_rate": 2.1392164788185693e-06, "loss": 0.2586, "step": 30680 }, { "epoch": 1.815224463240078, "grad_norm": 2.9984617233276367, "learning_rate": 2.137381558848131e-06, "loss": 0.2608, "step": 30690 }, { "epoch": 1.8158159342284261, "grad_norm": 5.632620811462402, "learning_rate": 2.135546990604862e-06, "loss": 0.3203, "step": 30700 }, { "epoch": 1.8164074052167742, "grad_norm": 2.517502784729004, "learning_rate": 2.133712774836796e-06, "loss": 0.2819, "step": 30710 }, { "epoch": 1.8169988762051221, "grad_norm": 3.7751059532165527, "learning_rate": 2.131878912291823e-06, "loss": 0.2494, "step": 30720 }, { "epoch": 1.81759034719347, "grad_norm": 3.3555994033813477, "learning_rate": 2.13004540371769e-06, "loss": 0.3185, "step": 30730 }, { "epoch": 1.8181818181818183, "grad_norm": 4.241505146026611, "learning_rate": 2.1282122498619976e-06, "loss": 0.3348, "step": 30740 }, { "epoch": 1.8187732891701662, "grad_norm": 4.937350749969482, "learning_rate": 2.126379451472203e-06, "loss": 0.2684, "step": 30750 }, { "epoch": 1.8193647601585141, "grad_norm": 3.3576791286468506, "learning_rate": 2.124547009295617e-06, "loss": 0.2622, "step": 30760 }, { "epoch": 1.8199562311468622, "grad_norm": 2.545168161392212, "learning_rate": 2.122714924079407e-06, "loss": 0.3074, "step": 30770 }, { "epoch": 1.8205477021352103, "grad_norm": 3.9872965812683105, "learning_rate": 2.1208831965705935e-06, "loss": 0.287, "step": 30780 }, { "epoch": 1.8211391731235582, "grad_norm": 4.043783664703369, "learning_rate": 2.1190518275160527e-06, "loss": 0.2929, "step": 30790 }, { "epoch": 1.8217306441119063, "grad_norm": 1.963836669921875, "learning_rate": 2.1172208176625126e-06, "loss": 0.2344, "step": 30800 }, { "epoch": 1.8223221151002544, "grad_norm": 5.607038974761963, "learning_rate": 2.115390167756557e-06, "loss": 0.3583, "step": 30810 }, { "epoch": 1.8229135860886023, "grad_norm": 3.511023998260498, "learning_rate": 2.1135598785446205e-06, "loss": 0.2535, "step": 30820 }, { "epoch": 1.8235050570769504, "grad_norm": 3.5397469997406006, "learning_rate": 2.111729950772992e-06, "loss": 0.34, "step": 30830 }, { "epoch": 1.8240965280652985, "grad_norm": 2.10614275932312, "learning_rate": 2.109900385187814e-06, "loss": 0.2687, "step": 30840 }, { "epoch": 1.8246879990536464, "grad_norm": 5.029149055480957, "learning_rate": 2.108071182535079e-06, "loss": 0.2966, "step": 30850 }, { "epoch": 1.8252794700419943, "grad_norm": 3.1837117671966553, "learning_rate": 2.1062423435606336e-06, "loss": 0.2562, "step": 30860 }, { "epoch": 1.8258709410303424, "grad_norm": 3.919114112854004, "learning_rate": 2.104413869010174e-06, "loss": 0.3319, "step": 30870 }, { "epoch": 1.8264624120186905, "grad_norm": 4.5236124992370605, "learning_rate": 2.1025857596292503e-06, "loss": 0.3054, "step": 30880 }, { "epoch": 1.8270538830070384, "grad_norm": 3.1304757595062256, "learning_rate": 2.100758016163262e-06, "loss": 0.3, "step": 30890 }, { "epoch": 1.8276453539953865, "grad_norm": 6.786297798156738, "learning_rate": 2.0989306393574596e-06, "loss": 0.2681, "step": 30900 }, { "epoch": 1.8282368249837346, "grad_norm": 1.9765396118164062, "learning_rate": 2.0971036299569458e-06, "loss": 0.2599, "step": 30910 }, { "epoch": 1.8288282959720825, "grad_norm": 3.5474934577941895, "learning_rate": 2.09527698870667e-06, "loss": 0.2879, "step": 30920 }, { "epoch": 1.8294197669604306, "grad_norm": 2.1633598804473877, "learning_rate": 2.093450716351434e-06, "loss": 0.2935, "step": 30930 }, { "epoch": 1.8300112379487787, "grad_norm": 4.486394882202148, "learning_rate": 2.0916248136358894e-06, "loss": 0.3697, "step": 30940 }, { "epoch": 1.8306027089371266, "grad_norm": 3.862847328186035, "learning_rate": 2.089799281304536e-06, "loss": 0.2726, "step": 30950 }, { "epoch": 1.8311941799254745, "grad_norm": 3.1322460174560547, "learning_rate": 2.087974120101723e-06, "loss": 0.3101, "step": 30960 }, { "epoch": 1.8317856509138228, "grad_norm": 3.4639782905578613, "learning_rate": 2.0861493307716477e-06, "loss": 0.2401, "step": 30970 }, { "epoch": 1.8323771219021707, "grad_norm": 2.6901116371154785, "learning_rate": 2.0843249140583572e-06, "loss": 0.3212, "step": 30980 }, { "epoch": 1.8329685928905186, "grad_norm": 4.174993515014648, "learning_rate": 2.0825008707057453e-06, "loss": 0.3248, "step": 30990 }, { "epoch": 1.8335600638788667, "grad_norm": 3.0649912357330322, "learning_rate": 2.0806772014575554e-06, "loss": 0.2995, "step": 31000 }, { "epoch": 1.8341515348672148, "grad_norm": 1.9445427656173706, "learning_rate": 2.078853907057375e-06, "loss": 0.2836, "step": 31010 }, { "epoch": 1.8347430058555627, "grad_norm": 2.355973720550537, "learning_rate": 2.077030988248641e-06, "loss": 0.2521, "step": 31020 }, { "epoch": 1.8353344768439108, "grad_norm": 3.5535314083099365, "learning_rate": 2.0752084457746376e-06, "loss": 0.3132, "step": 31030 }, { "epoch": 1.835925947832259, "grad_norm": 4.102923393249512, "learning_rate": 2.073386280378495e-06, "loss": 0.2977, "step": 31040 }, { "epoch": 1.8365174188206068, "grad_norm": 1.9148259162902832, "learning_rate": 2.0715644928031894e-06, "loss": 0.2993, "step": 31050 }, { "epoch": 1.837108889808955, "grad_norm": 2.9977431297302246, "learning_rate": 2.0697430837915438e-06, "loss": 0.3078, "step": 31060 }, { "epoch": 1.837700360797303, "grad_norm": 2.186070203781128, "learning_rate": 2.0679220540862245e-06, "loss": 0.2649, "step": 31070 }, { "epoch": 1.838291831785651, "grad_norm": 2.788227081298828, "learning_rate": 2.0661014044297456e-06, "loss": 0.2994, "step": 31080 }, { "epoch": 1.8388833027739988, "grad_norm": 3.0577144622802734, "learning_rate": 2.064281135564465e-06, "loss": 0.2895, "step": 31090 }, { "epoch": 1.839474773762347, "grad_norm": 1.9512826204299927, "learning_rate": 2.0624612482325865e-06, "loss": 0.2577, "step": 31100 }, { "epoch": 1.840066244750695, "grad_norm": 2.2741053104400635, "learning_rate": 2.0606417431761567e-06, "loss": 0.2751, "step": 31110 }, { "epoch": 1.840657715739043, "grad_norm": 2.888453722000122, "learning_rate": 2.0588226211370684e-06, "loss": 0.298, "step": 31120 }, { "epoch": 1.841249186727391, "grad_norm": 4.163794994354248, "learning_rate": 2.0570038828570557e-06, "loss": 0.3703, "step": 31130 }, { "epoch": 1.8418406577157391, "grad_norm": 3.158825635910034, "learning_rate": 2.055185529077699e-06, "loss": 0.3207, "step": 31140 }, { "epoch": 1.842432128704087, "grad_norm": 2.039567232131958, "learning_rate": 2.0533675605404186e-06, "loss": 0.2225, "step": 31150 }, { "epoch": 1.8430235996924351, "grad_norm": 3.9955031871795654, "learning_rate": 2.05154997798648e-06, "loss": 0.3178, "step": 31160 }, { "epoch": 1.8436150706807832, "grad_norm": 5.12684965133667, "learning_rate": 2.049732782156992e-06, "loss": 0.3037, "step": 31170 }, { "epoch": 1.8442065416691311, "grad_norm": 2.8764138221740723, "learning_rate": 2.0479159737929035e-06, "loss": 0.3178, "step": 31180 }, { "epoch": 1.844798012657479, "grad_norm": 2.3843352794647217, "learning_rate": 2.0460995536350076e-06, "loss": 0.2596, "step": 31190 }, { "epoch": 1.8453894836458273, "grad_norm": 6.128424644470215, "learning_rate": 2.0442835224239377e-06, "loss": 0.2604, "step": 31200 }, { "epoch": 1.8459809546341752, "grad_norm": 3.2804348468780518, "learning_rate": 2.042467880900168e-06, "loss": 0.2949, "step": 31210 }, { "epoch": 1.846572425622523, "grad_norm": 2.3971011638641357, "learning_rate": 2.040652629804016e-06, "loss": 0.3173, "step": 31220 }, { "epoch": 1.8471638966108712, "grad_norm": 2.9632067680358887, "learning_rate": 2.0388377698756376e-06, "loss": 0.2589, "step": 31230 }, { "epoch": 1.8477553675992193, "grad_norm": 4.5638556480407715, "learning_rate": 2.0370233018550313e-06, "loss": 0.3045, "step": 31240 }, { "epoch": 1.8483468385875672, "grad_norm": 1.7635035514831543, "learning_rate": 2.0352092264820352e-06, "loss": 0.252, "step": 31250 }, { "epoch": 1.8489383095759153, "grad_norm": 3.499422073364258, "learning_rate": 2.033395544496326e-06, "loss": 0.3127, "step": 31260 }, { "epoch": 1.8495297805642634, "grad_norm": 5.35884952545166, "learning_rate": 2.031582256637421e-06, "loss": 0.3266, "step": 31270 }, { "epoch": 1.8501212515526113, "grad_norm": 5.605325698852539, "learning_rate": 2.029769363644677e-06, "loss": 0.3027, "step": 31280 }, { "epoch": 1.8507127225409594, "grad_norm": 3.666318893432617, "learning_rate": 2.0279568662572904e-06, "loss": 0.2978, "step": 31290 }, { "epoch": 1.8513041935293075, "grad_norm": 2.989314556121826, "learning_rate": 2.026144765214294e-06, "loss": 0.315, "step": 31300 }, { "epoch": 1.8518956645176554, "grad_norm": 2.9791505336761475, "learning_rate": 2.0243330612545616e-06, "loss": 0.3047, "step": 31310 }, { "epoch": 1.8524871355060033, "grad_norm": 2.5527632236480713, "learning_rate": 2.0225217551168032e-06, "loss": 0.3204, "step": 31320 }, { "epoch": 1.8530786064943514, "grad_norm": 1.9773222208023071, "learning_rate": 2.0207108475395672e-06, "loss": 0.2572, "step": 31330 }, { "epoch": 1.8536700774826995, "grad_norm": 3.4737794399261475, "learning_rate": 2.018900339261242e-06, "loss": 0.277, "step": 31340 }, { "epoch": 1.8542615484710474, "grad_norm": 2.930225133895874, "learning_rate": 2.017090231020047e-06, "loss": 0.2843, "step": 31350 }, { "epoch": 1.8548530194593955, "grad_norm": 3.663126230239868, "learning_rate": 2.0152805235540444e-06, "loss": 0.3116, "step": 31360 }, { "epoch": 1.8554444904477436, "grad_norm": 3.278526544570923, "learning_rate": 2.0134712176011303e-06, "loss": 0.2813, "step": 31370 }, { "epoch": 1.8560359614360915, "grad_norm": 3.8184893131256104, "learning_rate": 2.0116623138990384e-06, "loss": 0.2638, "step": 31380 }, { "epoch": 1.8566274324244396, "grad_norm": 3.2004120349884033, "learning_rate": 2.009853813185338e-06, "loss": 0.2894, "step": 31390 }, { "epoch": 1.8572189034127877, "grad_norm": 3.3134307861328125, "learning_rate": 2.008045716197432e-06, "loss": 0.2943, "step": 31400 }, { "epoch": 1.8578103744011356, "grad_norm": 4.3683929443359375, "learning_rate": 2.0062380236725614e-06, "loss": 0.2931, "step": 31410 }, { "epoch": 1.8584018453894835, "grad_norm": 5.062314987182617, "learning_rate": 2.0044307363478014e-06, "loss": 0.2967, "step": 31420 }, { "epoch": 1.8589933163778318, "grad_norm": 2.063014268875122, "learning_rate": 2.002623854960062e-06, "loss": 0.2325, "step": 31430 }, { "epoch": 1.8595847873661797, "grad_norm": 3.6610076427459717, "learning_rate": 2.000817380246087e-06, "loss": 0.3645, "step": 31440 }, { "epoch": 1.8601762583545276, "grad_norm": 4.455199241638184, "learning_rate": 1.999011312942456e-06, "loss": 0.3472, "step": 31450 }, { "epoch": 1.8607677293428757, "grad_norm": 2.8034026622772217, "learning_rate": 1.9972056537855805e-06, "loss": 0.2607, "step": 31460 }, { "epoch": 1.8613592003312238, "grad_norm": 3.907402515411377, "learning_rate": 1.995400403511706e-06, "loss": 0.3277, "step": 31470 }, { "epoch": 1.8619506713195717, "grad_norm": 3.154264450073242, "learning_rate": 1.993595562856913e-06, "loss": 0.2841, "step": 31480 }, { "epoch": 1.8625421423079198, "grad_norm": 2.472649097442627, "learning_rate": 1.991791132557113e-06, "loss": 0.321, "step": 31490 }, { "epoch": 1.863133613296268, "grad_norm": 3.5905492305755615, "learning_rate": 1.989987113348052e-06, "loss": 0.3032, "step": 31500 }, { "epoch": 1.8637250842846158, "grad_norm": 2.6059205532073975, "learning_rate": 1.9881835059653063e-06, "loss": 0.3187, "step": 31510 }, { "epoch": 1.864316555272964, "grad_norm": 3.2364015579223633, "learning_rate": 1.9863803111442853e-06, "loss": 0.2971, "step": 31520 }, { "epoch": 1.864908026261312, "grad_norm": 1.9458703994750977, "learning_rate": 1.9845775296202303e-06, "loss": 0.3274, "step": 31530 }, { "epoch": 1.86549949724966, "grad_norm": 4.399162292480469, "learning_rate": 1.9827751621282143e-06, "loss": 0.2719, "step": 31540 }, { "epoch": 1.8660909682380078, "grad_norm": 2.894235610961914, "learning_rate": 1.9809732094031427e-06, "loss": 0.2641, "step": 31550 }, { "epoch": 1.866682439226356, "grad_norm": 2.1352837085723877, "learning_rate": 1.979171672179747e-06, "loss": 0.2842, "step": 31560 }, { "epoch": 1.867273910214704, "grad_norm": 2.0233354568481445, "learning_rate": 1.977370551192594e-06, "loss": 0.2814, "step": 31570 }, { "epoch": 1.867865381203052, "grad_norm": 1.926650047302246, "learning_rate": 1.9755698471760793e-06, "loss": 0.2982, "step": 31580 }, { "epoch": 1.8684568521914, "grad_norm": 4.03793478012085, "learning_rate": 1.973769560864429e-06, "loss": 0.2712, "step": 31590 }, { "epoch": 1.8690483231797481, "grad_norm": 4.255593776702881, "learning_rate": 1.971969692991698e-06, "loss": 0.2846, "step": 31600 }, { "epoch": 1.869639794168096, "grad_norm": 2.38187837600708, "learning_rate": 1.97017024429177e-06, "loss": 0.2435, "step": 31610 }, { "epoch": 1.8702312651564441, "grad_norm": 4.487394332885742, "learning_rate": 1.9683712154983603e-06, "loss": 0.3336, "step": 31620 }, { "epoch": 1.8708227361447922, "grad_norm": 4.889803409576416, "learning_rate": 1.9665726073450103e-06, "loss": 0.3277, "step": 31630 }, { "epoch": 1.87141420713314, "grad_norm": 2.687617778778076, "learning_rate": 1.9647744205650927e-06, "loss": 0.3748, "step": 31640 }, { "epoch": 1.872005678121488, "grad_norm": 3.1245534420013428, "learning_rate": 1.962976655891804e-06, "loss": 0.3104, "step": 31650 }, { "epoch": 1.8725971491098363, "grad_norm": 4.1550726890563965, "learning_rate": 1.961179314058173e-06, "loss": 0.2292, "step": 31660 }, { "epoch": 1.8731886200981842, "grad_norm": 4.165633678436279, "learning_rate": 1.959382395797054e-06, "loss": 0.3023, "step": 31670 }, { "epoch": 1.873780091086532, "grad_norm": 3.022974967956543, "learning_rate": 1.9575859018411292e-06, "loss": 0.3087, "step": 31680 }, { "epoch": 1.8743715620748802, "grad_norm": 2.726952075958252, "learning_rate": 1.9557898329229076e-06, "loss": 0.3145, "step": 31690 }, { "epoch": 1.8749630330632283, "grad_norm": 2.1634654998779297, "learning_rate": 1.953994189774725e-06, "loss": 0.2484, "step": 31700 }, { "epoch": 1.8755545040515762, "grad_norm": 2.509061098098755, "learning_rate": 1.9521989731287428e-06, "loss": 0.2793, "step": 31710 }, { "epoch": 1.8761459750399243, "grad_norm": 3.6500680446624756, "learning_rate": 1.9504041837169495e-06, "loss": 0.3742, "step": 31720 }, { "epoch": 1.8767374460282724, "grad_norm": 2.595262289047241, "learning_rate": 1.9486098222711582e-06, "loss": 0.291, "step": 31730 }, { "epoch": 1.8773289170166203, "grad_norm": 3.749293565750122, "learning_rate": 1.94681588952301e-06, "loss": 0.2777, "step": 31740 }, { "epoch": 1.8779203880049684, "grad_norm": 3.676224946975708, "learning_rate": 1.945022386203968e-06, "loss": 0.3247, "step": 31750 }, { "epoch": 1.8785118589933165, "grad_norm": 2.5477230548858643, "learning_rate": 1.943229313045324e-06, "loss": 0.2864, "step": 31760 }, { "epoch": 1.8791033299816644, "grad_norm": 3.2881815433502197, "learning_rate": 1.9414366707781895e-06, "loss": 0.3175, "step": 31770 }, { "epoch": 1.8796948009700123, "grad_norm": 3.97786545753479, "learning_rate": 1.939644460133504e-06, "loss": 0.3002, "step": 31780 }, { "epoch": 1.8802862719583604, "grad_norm": 4.9924116134643555, "learning_rate": 1.9378526818420303e-06, "loss": 0.2339, "step": 31790 }, { "epoch": 1.8808777429467085, "grad_norm": 3.6717939376831055, "learning_rate": 1.9360613366343535e-06, "loss": 0.2991, "step": 31800 }, { "epoch": 1.8814692139350564, "grad_norm": 3.277723550796509, "learning_rate": 1.9342704252408843e-06, "loss": 0.2617, "step": 31810 }, { "epoch": 1.8820606849234045, "grad_norm": 1.4790536165237427, "learning_rate": 1.932479948391855e-06, "loss": 0.2273, "step": 31820 }, { "epoch": 1.8826521559117526, "grad_norm": 5.371345043182373, "learning_rate": 1.9306899068173216e-06, "loss": 0.3011, "step": 31830 }, { "epoch": 1.8832436269001005, "grad_norm": 2.62030291557312, "learning_rate": 1.928900301247162e-06, "loss": 0.288, "step": 31840 }, { "epoch": 1.8838350978884486, "grad_norm": 3.394092082977295, "learning_rate": 1.927111132411075e-06, "loss": 0.2785, "step": 31850 }, { "epoch": 1.8844265688767967, "grad_norm": 2.7590911388397217, "learning_rate": 1.9253224010385847e-06, "loss": 0.3263, "step": 31860 }, { "epoch": 1.8850180398651446, "grad_norm": 5.192659854888916, "learning_rate": 1.9235341078590338e-06, "loss": 0.3079, "step": 31870 }, { "epoch": 1.8856095108534925, "grad_norm": 3.052724599838257, "learning_rate": 1.9217462536015883e-06, "loss": 0.307, "step": 31880 }, { "epoch": 1.8862009818418408, "grad_norm": 2.4261600971221924, "learning_rate": 1.9199588389952334e-06, "loss": 0.285, "step": 31890 }, { "epoch": 1.8867924528301887, "grad_norm": 2.4742259979248047, "learning_rate": 1.918171864768777e-06, "loss": 0.2451, "step": 31900 }, { "epoch": 1.8873839238185366, "grad_norm": 3.677093029022217, "learning_rate": 1.9163853316508464e-06, "loss": 0.3129, "step": 31910 }, { "epoch": 1.8879753948068847, "grad_norm": 2.901416301727295, "learning_rate": 1.9145992403698875e-06, "loss": 0.2863, "step": 31920 }, { "epoch": 1.8885668657952328, "grad_norm": 2.8962149620056152, "learning_rate": 1.9128135916541695e-06, "loss": 0.2357, "step": 31930 }, { "epoch": 1.8891583367835807, "grad_norm": 2.0826499462127686, "learning_rate": 1.9110283862317777e-06, "loss": 0.2669, "step": 31940 }, { "epoch": 1.8897498077719288, "grad_norm": 4.296623229980469, "learning_rate": 1.9092436248306206e-06, "loss": 0.3355, "step": 31950 }, { "epoch": 1.890341278760277, "grad_norm": 3.4737863540649414, "learning_rate": 1.907459308178421e-06, "loss": 0.2847, "step": 31960 }, { "epoch": 1.8909327497486248, "grad_norm": 3.075895309448242, "learning_rate": 1.9056754370027228e-06, "loss": 0.2475, "step": 31970 }, { "epoch": 1.891524220736973, "grad_norm": 3.981438159942627, "learning_rate": 1.9038920120308908e-06, "loss": 0.2494, "step": 31980 }, { "epoch": 1.892115691725321, "grad_norm": 4.040311336517334, "learning_rate": 1.902109033990101e-06, "loss": 0.3051, "step": 31990 }, { "epoch": 1.892707162713669, "grad_norm": 3.691366672515869, "learning_rate": 1.9003265036073541e-06, "loss": 0.2615, "step": 32000 }, { "epoch": 1.8932986337020168, "grad_norm": 6.534249305725098, "learning_rate": 1.8985444216094643e-06, "loss": 0.301, "step": 32010 }, { "epoch": 1.893890104690365, "grad_norm": 3.1061577796936035, "learning_rate": 1.8967627887230651e-06, "loss": 0.2945, "step": 32020 }, { "epoch": 1.894481575678713, "grad_norm": 3.1206414699554443, "learning_rate": 1.8949816056746054e-06, "loss": 0.2658, "step": 32030 }, { "epoch": 1.895073046667061, "grad_norm": 4.203948497772217, "learning_rate": 1.8932008731903518e-06, "loss": 0.2992, "step": 32040 }, { "epoch": 1.895664517655409, "grad_norm": 1.9868812561035156, "learning_rate": 1.8914205919963857e-06, "loss": 0.3033, "step": 32050 }, { "epoch": 1.896255988643757, "grad_norm": 2.7313077449798584, "learning_rate": 1.8896407628186056e-06, "loss": 0.2535, "step": 32060 }, { "epoch": 1.896847459632105, "grad_norm": 4.143831253051758, "learning_rate": 1.8878613863827264e-06, "loss": 0.2444, "step": 32070 }, { "epoch": 1.897438930620453, "grad_norm": 2.6430420875549316, "learning_rate": 1.8860824634142767e-06, "loss": 0.2949, "step": 32080 }, { "epoch": 1.8980304016088012, "grad_norm": 2.522223949432373, "learning_rate": 1.8843039946386016e-06, "loss": 0.2488, "step": 32090 }, { "epoch": 1.898621872597149, "grad_norm": 3.182310104370117, "learning_rate": 1.8825259807808593e-06, "loss": 0.3414, "step": 32100 }, { "epoch": 1.899213343585497, "grad_norm": 2.5020272731781006, "learning_rate": 1.8807484225660246e-06, "loss": 0.3193, "step": 32110 }, { "epoch": 1.8998048145738453, "grad_norm": 2.9907422065734863, "learning_rate": 1.878971320718885e-06, "loss": 0.3052, "step": 32120 }, { "epoch": 1.9003962855621932, "grad_norm": 2.412153959274292, "learning_rate": 1.877194675964043e-06, "loss": 0.3026, "step": 32130 }, { "epoch": 1.900987756550541, "grad_norm": 1.960724115371704, "learning_rate": 1.8754184890259136e-06, "loss": 0.2844, "step": 32140 }, { "epoch": 1.9015792275388892, "grad_norm": 2.6538753509521484, "learning_rate": 1.8736427606287268e-06, "loss": 0.2735, "step": 32150 }, { "epoch": 1.9021706985272373, "grad_norm": 2.971649408340454, "learning_rate": 1.8718674914965226e-06, "loss": 0.3135, "step": 32160 }, { "epoch": 1.9027621695155852, "grad_norm": 2.1408679485321045, "learning_rate": 1.870092682353157e-06, "loss": 0.2446, "step": 32170 }, { "epoch": 1.9033536405039333, "grad_norm": 2.6269965171813965, "learning_rate": 1.8683183339222972e-06, "loss": 0.267, "step": 32180 }, { "epoch": 1.9039451114922814, "grad_norm": 3.31681227684021, "learning_rate": 1.8665444469274229e-06, "loss": 0.3238, "step": 32190 }, { "epoch": 1.9045365824806293, "grad_norm": 4.002077579498291, "learning_rate": 1.8647710220918235e-06, "loss": 0.2322, "step": 32200 }, { "epoch": 1.9051280534689774, "grad_norm": 3.2302963733673096, "learning_rate": 1.8629980601386025e-06, "loss": 0.2423, "step": 32210 }, { "epoch": 1.9057195244573255, "grad_norm": 3.526888847351074, "learning_rate": 1.8612255617906743e-06, "loss": 0.2727, "step": 32220 }, { "epoch": 1.9063109954456734, "grad_norm": 2.8004140853881836, "learning_rate": 1.8594535277707636e-06, "loss": 0.3105, "step": 32230 }, { "epoch": 1.9069024664340213, "grad_norm": 3.138695478439331, "learning_rate": 1.8576819588014056e-06, "loss": 0.2961, "step": 32240 }, { "epoch": 1.9074939374223694, "grad_norm": 2.9063775539398193, "learning_rate": 1.8559108556049465e-06, "loss": 0.2782, "step": 32250 }, { "epoch": 1.9080854084107175, "grad_norm": 2.180849552154541, "learning_rate": 1.8541402189035419e-06, "loss": 0.3343, "step": 32260 }, { "epoch": 1.9086768793990654, "grad_norm": 3.141425132751465, "learning_rate": 1.8523700494191584e-06, "loss": 0.2942, "step": 32270 }, { "epoch": 1.9092683503874135, "grad_norm": 3.142690896987915, "learning_rate": 1.8506003478735707e-06, "loss": 0.3079, "step": 32280 }, { "epoch": 1.9098598213757616, "grad_norm": 2.786173105239868, "learning_rate": 1.8488311149883644e-06, "loss": 0.2713, "step": 32290 }, { "epoch": 1.9104512923641095, "grad_norm": 1.8409886360168457, "learning_rate": 1.8470623514849316e-06, "loss": 0.2492, "step": 32300 }, { "epoch": 1.9110427633524576, "grad_norm": 4.697363376617432, "learning_rate": 1.8452940580844748e-06, "loss": 0.3318, "step": 32310 }, { "epoch": 1.9116342343408057, "grad_norm": 2.206634044647217, "learning_rate": 1.8435262355080042e-06, "loss": 0.2568, "step": 32320 }, { "epoch": 1.9122257053291536, "grad_norm": 3.854714870452881, "learning_rate": 1.8417588844763387e-06, "loss": 0.2222, "step": 32330 }, { "epoch": 1.9128171763175015, "grad_norm": 4.488344669342041, "learning_rate": 1.8399920057101044e-06, "loss": 0.3035, "step": 32340 }, { "epoch": 1.9134086473058498, "grad_norm": 2.8494679927825928, "learning_rate": 1.8382255999297347e-06, "loss": 0.2704, "step": 32350 }, { "epoch": 1.9140001182941977, "grad_norm": 4.278850078582764, "learning_rate": 1.8364596678554694e-06, "loss": 0.2805, "step": 32360 }, { "epoch": 1.9145915892825456, "grad_norm": 2.9465174674987793, "learning_rate": 1.8346942102073572e-06, "loss": 0.2346, "step": 32370 }, { "epoch": 1.9151830602708937, "grad_norm": 3.3017914295196533, "learning_rate": 1.832929227705252e-06, "loss": 0.2966, "step": 32380 }, { "epoch": 1.9157745312592418, "grad_norm": 3.1500072479248047, "learning_rate": 1.8311647210688138e-06, "loss": 0.2801, "step": 32390 }, { "epoch": 1.9163660022475897, "grad_norm": 3.8669486045837402, "learning_rate": 1.8294006910175095e-06, "loss": 0.2125, "step": 32400 }, { "epoch": 1.9169574732359378, "grad_norm": 4.101596832275391, "learning_rate": 1.827637138270611e-06, "loss": 0.3302, "step": 32410 }, { "epoch": 1.917548944224286, "grad_norm": 2.826262950897217, "learning_rate": 1.8258740635471947e-06, "loss": 0.2967, "step": 32420 }, { "epoch": 1.9181404152126338, "grad_norm": 2.8086063861846924, "learning_rate": 1.8241114675661443e-06, "loss": 0.308, "step": 32430 }, { "epoch": 1.918731886200982, "grad_norm": 3.1073496341705322, "learning_rate": 1.8223493510461462e-06, "loss": 0.3105, "step": 32440 }, { "epoch": 1.91932335718933, "grad_norm": 3.8665199279785156, "learning_rate": 1.8205877147056925e-06, "loss": 0.3324, "step": 32450 }, { "epoch": 1.919914828177678, "grad_norm": 3.8022096157073975, "learning_rate": 1.818826559263079e-06, "loss": 0.3252, "step": 32460 }, { "epoch": 1.9205062991660258, "grad_norm": 2.5773684978485107, "learning_rate": 1.8170658854364064e-06, "loss": 0.2623, "step": 32470 }, { "epoch": 1.9210977701543739, "grad_norm": 1.8380590677261353, "learning_rate": 1.815305693943578e-06, "loss": 0.2452, "step": 32480 }, { "epoch": 1.921689241142722, "grad_norm": 2.017090320587158, "learning_rate": 1.8135459855022996e-06, "loss": 0.2826, "step": 32490 }, { "epoch": 1.9222807121310699, "grad_norm": 2.0691826343536377, "learning_rate": 1.811786760830082e-06, "loss": 0.2625, "step": 32500 }, { "epoch": 1.922872183119418, "grad_norm": 3.6818923950195312, "learning_rate": 1.8100280206442374e-06, "loss": 0.262, "step": 32510 }, { "epoch": 1.923463654107766, "grad_norm": 1.3418902158737183, "learning_rate": 1.8082697656618813e-06, "loss": 0.2487, "step": 32520 }, { "epoch": 1.924055125096114, "grad_norm": 2.015840768814087, "learning_rate": 1.806511996599931e-06, "loss": 0.2835, "step": 32530 }, { "epoch": 1.924646596084462, "grad_norm": 1.7782065868377686, "learning_rate": 1.8047547141751058e-06, "loss": 0.3023, "step": 32540 }, { "epoch": 1.9252380670728102, "grad_norm": 3.3886911869049072, "learning_rate": 1.8029979191039255e-06, "loss": 0.2836, "step": 32550 }, { "epoch": 1.925829538061158, "grad_norm": 1.6183972358703613, "learning_rate": 1.8012416121027127e-06, "loss": 0.2972, "step": 32560 }, { "epoch": 1.926421009049506, "grad_norm": 2.2255470752716064, "learning_rate": 1.7994857938875907e-06, "loss": 0.2794, "step": 32570 }, { "epoch": 1.9270124800378543, "grad_norm": 4.072093963623047, "learning_rate": 1.7977304651744823e-06, "loss": 0.3388, "step": 32580 }, { "epoch": 1.9276039510262022, "grad_norm": 1.4841912984848022, "learning_rate": 1.7959756266791126e-06, "loss": 0.2341, "step": 32590 }, { "epoch": 1.92819542201455, "grad_norm": 2.7013027667999268, "learning_rate": 1.7942212791170054e-06, "loss": 0.2933, "step": 32600 }, { "epoch": 1.9287868930028982, "grad_norm": 3.135075330734253, "learning_rate": 1.792467423203484e-06, "loss": 0.2635, "step": 32610 }, { "epoch": 1.9293783639912463, "grad_norm": 4.388999938964844, "learning_rate": 1.7907140596536747e-06, "loss": 0.2844, "step": 32620 }, { "epoch": 1.9299698349795942, "grad_norm": 3.1513991355895996, "learning_rate": 1.7889611891824965e-06, "loss": 0.2564, "step": 32630 }, { "epoch": 1.9305613059679423, "grad_norm": 2.6619293689727783, "learning_rate": 1.7872088125046724e-06, "loss": 0.3106, "step": 32640 }, { "epoch": 1.9311527769562904, "grad_norm": 1.8346278667449951, "learning_rate": 1.785456930334724e-06, "loss": 0.3497, "step": 32650 }, { "epoch": 1.9317442479446383, "grad_norm": 2.3105647563934326, "learning_rate": 1.7837055433869692e-06, "loss": 0.3087, "step": 32660 }, { "epoch": 1.9323357189329864, "grad_norm": 4.881100177764893, "learning_rate": 1.7819546523755247e-06, "loss": 0.3162, "step": 32670 }, { "epoch": 1.9329271899213345, "grad_norm": 1.8921427726745605, "learning_rate": 1.780204258014307e-06, "loss": 0.2774, "step": 32680 }, { "epoch": 1.9335186609096824, "grad_norm": 4.568484306335449, "learning_rate": 1.7784543610170253e-06, "loss": 0.32, "step": 32690 }, { "epoch": 1.9341101318980303, "grad_norm": 2.7630958557128906, "learning_rate": 1.7767049620971908e-06, "loss": 0.2737, "step": 32700 }, { "epoch": 1.9347016028863784, "grad_norm": 3.1720943450927734, "learning_rate": 1.7749560619681096e-06, "loss": 0.2733, "step": 32710 }, { "epoch": 1.9352930738747265, "grad_norm": 3.6933844089508057, "learning_rate": 1.7732076613428838e-06, "loss": 0.2643, "step": 32720 }, { "epoch": 1.9358845448630744, "grad_norm": 5.440460681915283, "learning_rate": 1.7714597609344131e-06, "loss": 0.2838, "step": 32730 }, { "epoch": 1.9364760158514225, "grad_norm": 2.560786247253418, "learning_rate": 1.7697123614553935e-06, "loss": 0.2711, "step": 32740 }, { "epoch": 1.9370674868397706, "grad_norm": 5.695807933807373, "learning_rate": 1.7679654636183154e-06, "loss": 0.3159, "step": 32750 }, { "epoch": 1.9376589578281185, "grad_norm": 2.660168170928955, "learning_rate": 1.7662190681354646e-06, "loss": 0.2635, "step": 32760 }, { "epoch": 1.9382504288164666, "grad_norm": 2.706683874130249, "learning_rate": 1.7644731757189232e-06, "loss": 0.2421, "step": 32770 }, { "epoch": 1.9388418998048147, "grad_norm": 2.460068941116333, "learning_rate": 1.7627277870805676e-06, "loss": 0.2717, "step": 32780 }, { "epoch": 1.9394333707931626, "grad_norm": 3.8146889209747314, "learning_rate": 1.76098290293207e-06, "loss": 0.2652, "step": 32790 }, { "epoch": 1.9400248417815105, "grad_norm": 2.3496005535125732, "learning_rate": 1.7592385239848942e-06, "loss": 0.2446, "step": 32800 }, { "epoch": 1.9406163127698588, "grad_norm": 3.772310972213745, "learning_rate": 1.7574946509503002e-06, "loss": 0.3093, "step": 32810 }, { "epoch": 1.9412077837582067, "grad_norm": 3.1131224632263184, "learning_rate": 1.7557512845393411e-06, "loss": 0.2904, "step": 32820 }, { "epoch": 1.9417992547465546, "grad_norm": 5.269538879394531, "learning_rate": 1.7540084254628635e-06, "loss": 0.2895, "step": 32830 }, { "epoch": 1.9423907257349027, "grad_norm": 4.602481842041016, "learning_rate": 1.7522660744315083e-06, "loss": 0.2801, "step": 32840 }, { "epoch": 1.9429821967232508, "grad_norm": 2.915858745574951, "learning_rate": 1.7505242321557065e-06, "loss": 0.2644, "step": 32850 }, { "epoch": 1.9435736677115987, "grad_norm": 5.287514686584473, "learning_rate": 1.748782899345682e-06, "loss": 0.2569, "step": 32860 }, { "epoch": 1.9441651386999468, "grad_norm": 2.8256423473358154, "learning_rate": 1.7470420767114546e-06, "loss": 0.3033, "step": 32870 }, { "epoch": 1.944756609688295, "grad_norm": 1.651275396347046, "learning_rate": 1.7453017649628315e-06, "loss": 0.2794, "step": 32880 }, { "epoch": 1.9453480806766428, "grad_norm": 2.2410366535186768, "learning_rate": 1.7435619648094169e-06, "loss": 0.2748, "step": 32890 }, { "epoch": 1.9459395516649909, "grad_norm": 4.635104179382324, "learning_rate": 1.7418226769606005e-06, "loss": 0.3125, "step": 32900 }, { "epoch": 1.946531022653339, "grad_norm": 2.549508571624756, "learning_rate": 1.7400839021255653e-06, "loss": 0.2983, "step": 32910 }, { "epoch": 1.9471224936416869, "grad_norm": 4.450366497039795, "learning_rate": 1.7383456410132883e-06, "loss": 0.2854, "step": 32920 }, { "epoch": 1.9477139646300348, "grad_norm": 3.7574708461761475, "learning_rate": 1.7366078943325317e-06, "loss": 0.2915, "step": 32930 }, { "epoch": 1.9483054356183829, "grad_norm": 2.2803006172180176, "learning_rate": 1.7348706627918534e-06, "loss": 0.2601, "step": 32940 }, { "epoch": 1.948896906606731, "grad_norm": 2.0237295627593994, "learning_rate": 1.7331339470995954e-06, "loss": 0.2315, "step": 32950 }, { "epoch": 1.9494883775950789, "grad_norm": 2.3520874977111816, "learning_rate": 1.7313977479638954e-06, "loss": 0.3089, "step": 32960 }, { "epoch": 1.950079848583427, "grad_norm": 3.661684989929199, "learning_rate": 1.7296620660926762e-06, "loss": 0.2832, "step": 32970 }, { "epoch": 1.950671319571775, "grad_norm": 2.7146947383880615, "learning_rate": 1.7279269021936497e-06, "loss": 0.2767, "step": 32980 }, { "epoch": 1.951262790560123, "grad_norm": 2.8290927410125732, "learning_rate": 1.7261922569743206e-06, "loss": 0.2496, "step": 32990 }, { "epoch": 1.951854261548471, "grad_norm": 3.7328455448150635, "learning_rate": 1.7244581311419767e-06, "loss": 0.2417, "step": 33000 }, { "epoch": 1.9524457325368192, "grad_norm": 3.714980125427246, "learning_rate": 1.7227245254036995e-06, "loss": 0.3227, "step": 33010 }, { "epoch": 1.953037203525167, "grad_norm": 3.1426167488098145, "learning_rate": 1.7209914404663539e-06, "loss": 0.299, "step": 33020 }, { "epoch": 1.953628674513515, "grad_norm": 5.084488391876221, "learning_rate": 1.7192588770365937e-06, "loss": 0.2606, "step": 33030 }, { "epoch": 1.9542201455018633, "grad_norm": 3.0850493907928467, "learning_rate": 1.717526835820863e-06, "loss": 0.271, "step": 33040 }, { "epoch": 1.9548116164902112, "grad_norm": 1.9520325660705566, "learning_rate": 1.7157953175253883e-06, "loss": 0.3119, "step": 33050 }, { "epoch": 1.955403087478559, "grad_norm": 2.6476337909698486, "learning_rate": 1.7140643228561866e-06, "loss": 0.3463, "step": 33060 }, { "epoch": 1.9559945584669072, "grad_norm": 2.255103588104248, "learning_rate": 1.7123338525190579e-06, "loss": 0.2875, "step": 33070 }, { "epoch": 1.9565860294552553, "grad_norm": 2.492579221725464, "learning_rate": 1.7106039072195927e-06, "loss": 0.3326, "step": 33080 }, { "epoch": 1.9571775004436032, "grad_norm": 2.8534772396087646, "learning_rate": 1.7088744876631632e-06, "loss": 0.2887, "step": 33090 }, { "epoch": 1.9577689714319513, "grad_norm": 3.1626343727111816, "learning_rate": 1.7071455945549319e-06, "loss": 0.2782, "step": 33100 }, { "epoch": 1.9583604424202994, "grad_norm": 3.609473466873169, "learning_rate": 1.705417228599842e-06, "loss": 0.2464, "step": 33110 }, { "epoch": 1.9589519134086473, "grad_norm": 2.9888312816619873, "learning_rate": 1.703689390502623e-06, "loss": 0.2323, "step": 33120 }, { "epoch": 1.9595433843969954, "grad_norm": 2.3723981380462646, "learning_rate": 1.7019620809677915e-06, "loss": 0.2767, "step": 33130 }, { "epoch": 1.9601348553853435, "grad_norm": 3.4475393295288086, "learning_rate": 1.700235300699646e-06, "loss": 0.3318, "step": 33140 }, { "epoch": 1.9607263263736914, "grad_norm": 2.860732078552246, "learning_rate": 1.6985090504022711e-06, "loss": 0.2997, "step": 33150 }, { "epoch": 1.9613177973620393, "grad_norm": 3.453784227371216, "learning_rate": 1.696783330779534e-06, "loss": 0.2914, "step": 33160 }, { "epoch": 1.9619092683503874, "grad_norm": 2.229827880859375, "learning_rate": 1.6950581425350843e-06, "loss": 0.3279, "step": 33170 }, { "epoch": 1.9625007393387355, "grad_norm": 2.713175058364868, "learning_rate": 1.6933334863723592e-06, "loss": 0.2608, "step": 33180 }, { "epoch": 1.9630922103270834, "grad_norm": 3.240230083465576, "learning_rate": 1.6916093629945733e-06, "loss": 0.2733, "step": 33190 }, { "epoch": 1.9636836813154315, "grad_norm": 3.828423500061035, "learning_rate": 1.68988577310473e-06, "loss": 0.304, "step": 33200 }, { "epoch": 1.9642751523037796, "grad_norm": 3.9174647331237793, "learning_rate": 1.688162717405611e-06, "loss": 0.327, "step": 33210 }, { "epoch": 1.9648666232921275, "grad_norm": 2.7972121238708496, "learning_rate": 1.686440196599779e-06, "loss": 0.2552, "step": 33220 }, { "epoch": 1.9654580942804756, "grad_norm": 3.9784910678863525, "learning_rate": 1.684718211389584e-06, "loss": 0.3045, "step": 33230 }, { "epoch": 1.9660495652688237, "grad_norm": 3.2796807289123535, "learning_rate": 1.682996762477152e-06, "loss": 0.2967, "step": 33240 }, { "epoch": 1.9666410362571716, "grad_norm": 1.748595952987671, "learning_rate": 1.6812758505643954e-06, "loss": 0.3262, "step": 33250 }, { "epoch": 1.9672325072455195, "grad_norm": 3.3644702434539795, "learning_rate": 1.6795554763530027e-06, "loss": 0.3407, "step": 33260 }, { "epoch": 1.9678239782338678, "grad_norm": 2.875129461288452, "learning_rate": 1.6778356405444488e-06, "loss": 0.2456, "step": 33270 }, { "epoch": 1.9684154492222157, "grad_norm": 2.4148762226104736, "learning_rate": 1.676116343839981e-06, "loss": 0.2834, "step": 33280 }, { "epoch": 1.9690069202105636, "grad_norm": 2.6161048412323, "learning_rate": 1.6743975869406351e-06, "loss": 0.3059, "step": 33290 }, { "epoch": 1.9695983911989117, "grad_norm": 3.0488433837890625, "learning_rate": 1.6726793705472229e-06, "loss": 0.2948, "step": 33300 }, { "epoch": 1.9701898621872598, "grad_norm": 2.346245050430298, "learning_rate": 1.6709616953603338e-06, "loss": 0.2967, "step": 33310 }, { "epoch": 1.9707813331756077, "grad_norm": 2.277008533477783, "learning_rate": 1.6692445620803418e-06, "loss": 0.3009, "step": 33320 }, { "epoch": 1.9713728041639558, "grad_norm": 2.723493814468384, "learning_rate": 1.6675279714073952e-06, "loss": 0.2755, "step": 33330 }, { "epoch": 1.9719642751523039, "grad_norm": 3.933779239654541, "learning_rate": 1.6658119240414246e-06, "loss": 0.3025, "step": 33340 }, { "epoch": 1.9725557461406518, "grad_norm": 2.6207730770111084, "learning_rate": 1.6640964206821358e-06, "loss": 0.3079, "step": 33350 }, { "epoch": 1.9731472171289999, "grad_norm": 4.120064735412598, "learning_rate": 1.6623814620290145e-06, "loss": 0.3105, "step": 33360 }, { "epoch": 1.973738688117348, "grad_norm": 2.930260419845581, "learning_rate": 1.6606670487813254e-06, "loss": 0.2391, "step": 33370 }, { "epoch": 1.9743301591056959, "grad_norm": 3.487510919570923, "learning_rate": 1.6589531816381078e-06, "loss": 0.2481, "step": 33380 }, { "epoch": 1.9749216300940438, "grad_norm": 4.478091239929199, "learning_rate": 1.657239861298182e-06, "loss": 0.341, "step": 33390 }, { "epoch": 1.9755131010823919, "grad_norm": 4.631832599639893, "learning_rate": 1.6555270884601412e-06, "loss": 0.2391, "step": 33400 }, { "epoch": 1.97610457207074, "grad_norm": 3.3568170070648193, "learning_rate": 1.6538148638223603e-06, "loss": 0.3022, "step": 33410 }, { "epoch": 1.9766960430590879, "grad_norm": 3.2074687480926514, "learning_rate": 1.652103188082987e-06, "loss": 0.2473, "step": 33420 }, { "epoch": 1.977287514047436, "grad_norm": 5.482704162597656, "learning_rate": 1.6503920619399438e-06, "loss": 0.3214, "step": 33430 }, { "epoch": 1.977878985035784, "grad_norm": 3.2199549674987793, "learning_rate": 1.648681486090935e-06, "loss": 0.3309, "step": 33440 }, { "epoch": 1.978470456024132, "grad_norm": 1.8878720998764038, "learning_rate": 1.6469714612334342e-06, "loss": 0.2514, "step": 33450 }, { "epoch": 1.97906192701248, "grad_norm": 3.0656206607818604, "learning_rate": 1.6452619880646952e-06, "loss": 0.3161, "step": 33460 }, { "epoch": 1.9796533980008282, "grad_norm": 2.754995822906494, "learning_rate": 1.6435530672817445e-06, "loss": 0.2725, "step": 33470 }, { "epoch": 1.980244868989176, "grad_norm": 2.4647018909454346, "learning_rate": 1.641844699581383e-06, "loss": 0.2306, "step": 33480 }, { "epoch": 1.980836339977524, "grad_norm": 2.8450000286102295, "learning_rate": 1.6401368856601868e-06, "loss": 0.3362, "step": 33490 }, { "epoch": 1.9814278109658723, "grad_norm": 3.655280590057373, "learning_rate": 1.6384296262145051e-06, "loss": 0.37, "step": 33500 }, { "epoch": 1.9820192819542202, "grad_norm": 4.301876068115234, "learning_rate": 1.636722921940464e-06, "loss": 0.2499, "step": 33510 }, { "epoch": 1.982610752942568, "grad_norm": 3.0211007595062256, "learning_rate": 1.6350167735339596e-06, "loss": 0.2391, "step": 33520 }, { "epoch": 1.9832022239309162, "grad_norm": 2.347785234451294, "learning_rate": 1.6333111816906646e-06, "loss": 0.3536, "step": 33530 }, { "epoch": 1.9837936949192643, "grad_norm": 3.6185805797576904, "learning_rate": 1.6316061471060218e-06, "loss": 0.287, "step": 33540 }, { "epoch": 1.9843851659076122, "grad_norm": 2.7091281414031982, "learning_rate": 1.6299016704752498e-06, "loss": 0.2704, "step": 33550 }, { "epoch": 1.9849766368959603, "grad_norm": 3.739119529724121, "learning_rate": 1.6281977524933372e-06, "loss": 0.2791, "step": 33560 }, { "epoch": 1.9855681078843084, "grad_norm": 1.7627215385437012, "learning_rate": 1.6264943938550445e-06, "loss": 0.2642, "step": 33570 }, { "epoch": 1.9861595788726563, "grad_norm": 2.471867799758911, "learning_rate": 1.6247915952549074e-06, "loss": 0.2756, "step": 33580 }, { "epoch": 1.9867510498610044, "grad_norm": 2.7878284454345703, "learning_rate": 1.6230893573872294e-06, "loss": 0.2921, "step": 33590 }, { "epoch": 1.9873425208493525, "grad_norm": 3.827418565750122, "learning_rate": 1.6213876809460886e-06, "loss": 0.255, "step": 33600 }, { "epoch": 1.9879339918377004, "grad_norm": 2.5560555458068848, "learning_rate": 1.6196865666253321e-06, "loss": 0.2549, "step": 33610 }, { "epoch": 1.9885254628260483, "grad_norm": 2.258436679840088, "learning_rate": 1.6179860151185778e-06, "loss": 0.2451, "step": 33620 }, { "epoch": 1.9891169338143964, "grad_norm": 2.136476755142212, "learning_rate": 1.6162860271192162e-06, "loss": 0.2764, "step": 33630 }, { "epoch": 1.9897084048027445, "grad_norm": 3.2596466541290283, "learning_rate": 1.6145866033204044e-06, "loss": 0.2555, "step": 33640 }, { "epoch": 1.9902998757910924, "grad_norm": 3.149590492248535, "learning_rate": 1.6128877444150744e-06, "loss": 0.2623, "step": 33650 }, { "epoch": 1.9908913467794405, "grad_norm": 4.882331848144531, "learning_rate": 1.6111894510959236e-06, "loss": 0.3224, "step": 33660 }, { "epoch": 1.9914828177677886, "grad_norm": 3.295163869857788, "learning_rate": 1.609491724055419e-06, "loss": 0.3132, "step": 33670 }, { "epoch": 1.9920742887561365, "grad_norm": 3.6489195823669434, "learning_rate": 1.6077945639858e-06, "loss": 0.2782, "step": 33680 }, { "epoch": 1.9926657597444846, "grad_norm": 2.8715507984161377, "learning_rate": 1.6060979715790714e-06, "loss": 0.3201, "step": 33690 }, { "epoch": 1.9932572307328327, "grad_norm": 3.801738739013672, "learning_rate": 1.6044019475270108e-06, "loss": 0.2826, "step": 33700 }, { "epoch": 1.9938487017211806, "grad_norm": 2.3693079948425293, "learning_rate": 1.6027064925211567e-06, "loss": 0.2857, "step": 33710 }, { "epoch": 1.9944401727095284, "grad_norm": 5.765575408935547, "learning_rate": 1.6010116072528237e-06, "loss": 0.284, "step": 33720 }, { "epoch": 1.9950316436978768, "grad_norm": 4.078457832336426, "learning_rate": 1.599317292413088e-06, "loss": 0.3326, "step": 33730 }, { "epoch": 1.9956231146862247, "grad_norm": 2.4788217544555664, "learning_rate": 1.5976235486927974e-06, "loss": 0.3005, "step": 33740 }, { "epoch": 1.9962145856745726, "grad_norm": 2.834920644760132, "learning_rate": 1.5959303767825648e-06, "loss": 0.2502, "step": 33750 }, { "epoch": 1.9968060566629207, "grad_norm": 3.195190191268921, "learning_rate": 1.594237777372769e-06, "loss": 0.2717, "step": 33760 }, { "epoch": 1.9973975276512688, "grad_norm": 2.0378942489624023, "learning_rate": 1.5925457511535584e-06, "loss": 0.2767, "step": 33770 }, { "epoch": 1.9979889986396167, "grad_norm": 4.769126892089844, "learning_rate": 1.5908542988148445e-06, "loss": 0.3131, "step": 33780 }, { "epoch": 1.9985804696279648, "grad_norm": 4.193257808685303, "learning_rate": 1.5891634210463074e-06, "loss": 0.3162, "step": 33790 }, { "epoch": 1.9991719406163129, "grad_norm": 3.7497923374176025, "learning_rate": 1.587473118537391e-06, "loss": 0.3231, "step": 33800 }, { "epoch": 1.9997634116046608, "grad_norm": 2.569728136062622, "learning_rate": 1.5857833919773048e-06, "loss": 0.2416, "step": 33810 }, { "epoch": 2.0003548825930086, "grad_norm": 2.4185872077941895, "learning_rate": 1.5840942420550259e-06, "loss": 0.2898, "step": 33820 }, { "epoch": 2.000946353581357, "grad_norm": 2.6887428760528564, "learning_rate": 1.5824056694592918e-06, "loss": 0.2372, "step": 33830 }, { "epoch": 2.001537824569705, "grad_norm": 3.0200839042663574, "learning_rate": 1.5807176748786098e-06, "loss": 0.2449, "step": 33840 }, { "epoch": 2.0021292955580527, "grad_norm": 3.214160203933716, "learning_rate": 1.5790302590012466e-06, "loss": 0.2574, "step": 33850 }, { "epoch": 2.002720766546401, "grad_norm": 2.104814291000366, "learning_rate": 1.5773434225152377e-06, "loss": 0.2203, "step": 33860 }, { "epoch": 2.003312237534749, "grad_norm": 4.020064830780029, "learning_rate": 1.5756571661083788e-06, "loss": 0.3027, "step": 33870 }, { "epoch": 2.003903708523097, "grad_norm": 3.4237897396087646, "learning_rate": 1.5739714904682286e-06, "loss": 0.2661, "step": 33880 }, { "epoch": 2.004495179511445, "grad_norm": 2.283681631088257, "learning_rate": 1.5722863962821136e-06, "loss": 0.2276, "step": 33890 }, { "epoch": 2.005086650499793, "grad_norm": 3.3959097862243652, "learning_rate": 1.5706018842371172e-06, "loss": 0.289, "step": 33900 }, { "epoch": 2.005678121488141, "grad_norm": 3.7945120334625244, "learning_rate": 1.5689179550200924e-06, "loss": 0.2463, "step": 33910 }, { "epoch": 2.006269592476489, "grad_norm": 3.8028461933135986, "learning_rate": 1.5672346093176457e-06, "loss": 0.2401, "step": 33920 }, { "epoch": 2.006861063464837, "grad_norm": 2.026705026626587, "learning_rate": 1.5655518478161546e-06, "loss": 0.2095, "step": 33930 }, { "epoch": 2.007452534453185, "grad_norm": 5.246816158294678, "learning_rate": 1.5638696712017522e-06, "loss": 0.2495, "step": 33940 }, { "epoch": 2.008044005441533, "grad_norm": 3.4206578731536865, "learning_rate": 1.5621880801603353e-06, "loss": 0.2389, "step": 33950 }, { "epoch": 2.0086354764298813, "grad_norm": 3.141270399093628, "learning_rate": 1.560507075377563e-06, "loss": 0.2429, "step": 33960 }, { "epoch": 2.009226947418229, "grad_norm": 4.810205459594727, "learning_rate": 1.558826657538853e-06, "loss": 0.2802, "step": 33970 }, { "epoch": 2.009818418406577, "grad_norm": 4.109495639801025, "learning_rate": 1.5571468273293866e-06, "loss": 0.2577, "step": 33980 }, { "epoch": 2.0104098893949254, "grad_norm": 3.5354013442993164, "learning_rate": 1.5554675854341031e-06, "loss": 0.2395, "step": 33990 }, { "epoch": 2.0110013603832733, "grad_norm": 2.700627088546753, "learning_rate": 1.5537889325377015e-06, "loss": 0.3138, "step": 34000 }, { "epoch": 2.011592831371621, "grad_norm": 3.2241928577423096, "learning_rate": 1.5521108693246443e-06, "loss": 0.3048, "step": 34010 }, { "epoch": 2.012184302359969, "grad_norm": 2.0670039653778076, "learning_rate": 1.5504333964791484e-06, "loss": 0.2648, "step": 34020 }, { "epoch": 2.0127757733483174, "grad_norm": 2.453742027282715, "learning_rate": 1.548756514685195e-06, "loss": 0.2249, "step": 34030 }, { "epoch": 2.0133672443366653, "grad_norm": 3.3937525749206543, "learning_rate": 1.5470802246265194e-06, "loss": 0.2388, "step": 34040 }, { "epoch": 2.013958715325013, "grad_norm": 3.981480836868286, "learning_rate": 1.5454045269866212e-06, "loss": 0.2853, "step": 34050 }, { "epoch": 2.0145501863133615, "grad_norm": 3.656019687652588, "learning_rate": 1.543729422448754e-06, "loss": 0.2242, "step": 34060 }, { "epoch": 2.0151416573017094, "grad_norm": 2.798781156539917, "learning_rate": 1.5420549116959297e-06, "loss": 0.2562, "step": 34070 }, { "epoch": 2.0157331282900572, "grad_norm": 2.8310909271240234, "learning_rate": 1.5403809954109214e-06, "loss": 0.2516, "step": 34080 }, { "epoch": 2.0163245992784056, "grad_norm": 2.733079195022583, "learning_rate": 1.5387076742762563e-06, "loss": 0.2921, "step": 34090 }, { "epoch": 2.0169160702667535, "grad_norm": 2.4602696895599365, "learning_rate": 1.537034948974222e-06, "loss": 0.3009, "step": 34100 }, { "epoch": 2.0175075412551013, "grad_norm": 4.419715881347656, "learning_rate": 1.5353628201868605e-06, "loss": 0.2874, "step": 34110 }, { "epoch": 2.0180990122434497, "grad_norm": 1.6635695695877075, "learning_rate": 1.5336912885959709e-06, "loss": 0.224, "step": 34120 }, { "epoch": 2.0186904832317976, "grad_norm": 4.443561553955078, "learning_rate": 1.5320203548831128e-06, "loss": 0.2361, "step": 34130 }, { "epoch": 2.0192819542201454, "grad_norm": 3.72198748588562, "learning_rate": 1.5303500197295942e-06, "loss": 0.3157, "step": 34140 }, { "epoch": 2.0198734252084933, "grad_norm": 3.228640079498291, "learning_rate": 1.528680283816487e-06, "loss": 0.2631, "step": 34150 }, { "epoch": 2.0204648961968417, "grad_norm": 4.169002532958984, "learning_rate": 1.527011147824613e-06, "loss": 0.2422, "step": 34160 }, { "epoch": 2.0210563671851896, "grad_norm": 2.2134854793548584, "learning_rate": 1.5253426124345542e-06, "loss": 0.2701, "step": 34170 }, { "epoch": 2.0216478381735374, "grad_norm": 1.7443538904190063, "learning_rate": 1.523674678326643e-06, "loss": 0.2767, "step": 34180 }, { "epoch": 2.0222393091618858, "grad_norm": 4.515565872192383, "learning_rate": 1.522007346180971e-06, "loss": 0.2566, "step": 34190 }, { "epoch": 2.0228307801502337, "grad_norm": 3.450299024581909, "learning_rate": 1.5203406166773812e-06, "loss": 0.2677, "step": 34200 }, { "epoch": 2.0234222511385815, "grad_norm": 2.231497287750244, "learning_rate": 1.5186744904954708e-06, "loss": 0.261, "step": 34210 }, { "epoch": 2.02401372212693, "grad_norm": 2.816800117492676, "learning_rate": 1.5170089683145944e-06, "loss": 0.27, "step": 34220 }, { "epoch": 2.0246051931152778, "grad_norm": 2.7367515563964844, "learning_rate": 1.515344050813856e-06, "loss": 0.3091, "step": 34230 }, { "epoch": 2.0251966641036256, "grad_norm": 3.7192044258117676, "learning_rate": 1.513679738672117e-06, "loss": 0.2653, "step": 34240 }, { "epoch": 2.0257881350919735, "grad_norm": 2.9983370304107666, "learning_rate": 1.5120160325679892e-06, "loss": 0.2523, "step": 34250 }, { "epoch": 2.026379606080322, "grad_norm": 3.6733181476593018, "learning_rate": 1.510352933179837e-06, "loss": 0.2528, "step": 34260 }, { "epoch": 2.0269710770686697, "grad_norm": 1.6978791952133179, "learning_rate": 1.5086904411857802e-06, "loss": 0.2178, "step": 34270 }, { "epoch": 2.0275625480570176, "grad_norm": 2.995387315750122, "learning_rate": 1.507028557263688e-06, "loss": 0.2632, "step": 34280 }, { "epoch": 2.028154019045366, "grad_norm": 2.3857250213623047, "learning_rate": 1.505367282091185e-06, "loss": 0.2976, "step": 34290 }, { "epoch": 2.028745490033714, "grad_norm": 4.599531650543213, "learning_rate": 1.503706616345644e-06, "loss": 0.3034, "step": 34300 }, { "epoch": 2.0293369610220617, "grad_norm": 2.020186185836792, "learning_rate": 1.50204656070419e-06, "loss": 0.2197, "step": 34310 }, { "epoch": 2.02992843201041, "grad_norm": 4.37985897064209, "learning_rate": 1.5003871158437029e-06, "loss": 0.2435, "step": 34320 }, { "epoch": 2.030519902998758, "grad_norm": 3.904498338699341, "learning_rate": 1.4987282824408072e-06, "loss": 0.2741, "step": 34330 }, { "epoch": 2.031111373987106, "grad_norm": 4.386171817779541, "learning_rate": 1.4970700611718864e-06, "loss": 0.2655, "step": 34340 }, { "epoch": 2.031702844975454, "grad_norm": 2.734095811843872, "learning_rate": 1.4954124527130643e-06, "loss": 0.226, "step": 34350 }, { "epoch": 2.032294315963802, "grad_norm": 5.061698913574219, "learning_rate": 1.4937554577402239e-06, "loss": 0.2536, "step": 34360 }, { "epoch": 2.03288578695215, "grad_norm": 2.889474391937256, "learning_rate": 1.4920990769289915e-06, "loss": 0.2128, "step": 34370 }, { "epoch": 2.033477257940498, "grad_norm": 4.5274739265441895, "learning_rate": 1.4904433109547494e-06, "loss": 0.2448, "step": 34380 }, { "epoch": 2.034068728928846, "grad_norm": 3.6230978965759277, "learning_rate": 1.4887881604926232e-06, "loss": 0.2673, "step": 34390 }, { "epoch": 2.034660199917194, "grad_norm": 3.4641048908233643, "learning_rate": 1.4871336262174892e-06, "loss": 0.3058, "step": 34400 }, { "epoch": 2.035251670905542, "grad_norm": 2.973914861679077, "learning_rate": 1.4854797088039756e-06, "loss": 0.2382, "step": 34410 }, { "epoch": 2.0358431418938903, "grad_norm": 11.089842796325684, "learning_rate": 1.4838264089264545e-06, "loss": 0.2863, "step": 34420 }, { "epoch": 2.036434612882238, "grad_norm": 3.1760175228118896, "learning_rate": 1.4821737272590505e-06, "loss": 0.26, "step": 34430 }, { "epoch": 2.037026083870586, "grad_norm": 2.6012015342712402, "learning_rate": 1.4805216644756326e-06, "loss": 0.2463, "step": 34440 }, { "epoch": 2.0376175548589344, "grad_norm": 2.9633493423461914, "learning_rate": 1.4788702212498182e-06, "loss": 0.2849, "step": 34450 }, { "epoch": 2.0382090258472823, "grad_norm": 2.9149410724639893, "learning_rate": 1.4772193982549752e-06, "loss": 0.2997, "step": 34460 }, { "epoch": 2.03880049683563, "grad_norm": 2.1206603050231934, "learning_rate": 1.4755691961642132e-06, "loss": 0.2805, "step": 34470 }, { "epoch": 2.039391967823978, "grad_norm": 5.097540855407715, "learning_rate": 1.4739196156503943e-06, "loss": 0.298, "step": 34480 }, { "epoch": 2.0399834388123264, "grad_norm": 3.252506971359253, "learning_rate": 1.4722706573861217e-06, "loss": 0.2855, "step": 34490 }, { "epoch": 2.0405749098006742, "grad_norm": 3.9117658138275146, "learning_rate": 1.4706223220437508e-06, "loss": 0.2763, "step": 34500 }, { "epoch": 2.041166380789022, "grad_norm": 1.5414575338363647, "learning_rate": 1.4689746102953778e-06, "loss": 0.2249, "step": 34510 }, { "epoch": 2.0417578517773705, "grad_norm": 3.446925401687622, "learning_rate": 1.4673275228128459e-06, "loss": 0.2744, "step": 34520 }, { "epoch": 2.0423493227657183, "grad_norm": 2.9129555225372314, "learning_rate": 1.4656810602677464e-06, "loss": 0.232, "step": 34530 }, { "epoch": 2.0429407937540662, "grad_norm": 2.575098991394043, "learning_rate": 1.4640352233314123e-06, "loss": 0.285, "step": 34540 }, { "epoch": 2.0435322647424146, "grad_norm": 3.3006341457366943, "learning_rate": 1.462390012674925e-06, "loss": 0.2806, "step": 34550 }, { "epoch": 2.0441237357307624, "grad_norm": 3.4367082118988037, "learning_rate": 1.4607454289691076e-06, "loss": 0.2832, "step": 34560 }, { "epoch": 2.0447152067191103, "grad_norm": 4.238278388977051, "learning_rate": 1.4591014728845285e-06, "loss": 0.2416, "step": 34570 }, { "epoch": 2.0453066777074587, "grad_norm": 3.1564691066741943, "learning_rate": 1.4574581450915007e-06, "loss": 0.2707, "step": 34580 }, { "epoch": 2.0458981486958066, "grad_norm": 3.543006658554077, "learning_rate": 1.4558154462600792e-06, "loss": 0.2067, "step": 34590 }, { "epoch": 2.0464896196841544, "grad_norm": 4.601242542266846, "learning_rate": 1.4541733770600661e-06, "loss": 0.2555, "step": 34600 }, { "epoch": 2.0470810906725023, "grad_norm": 3.1182124614715576, "learning_rate": 1.4525319381610024e-06, "loss": 0.322, "step": 34610 }, { "epoch": 2.0476725616608507, "grad_norm": 3.277829885482788, "learning_rate": 1.4508911302321772e-06, "loss": 0.2672, "step": 34620 }, { "epoch": 2.0482640326491985, "grad_norm": 2.4594600200653076, "learning_rate": 1.4492509539426165e-06, "loss": 0.2867, "step": 34630 }, { "epoch": 2.0488555036375464, "grad_norm": 3.650374174118042, "learning_rate": 1.447611409961094e-06, "loss": 0.2245, "step": 34640 }, { "epoch": 2.0494469746258948, "grad_norm": 2.9155936241149902, "learning_rate": 1.4459724989561227e-06, "loss": 0.2397, "step": 34650 }, { "epoch": 2.0500384456142426, "grad_norm": 4.5966620445251465, "learning_rate": 1.444334221595957e-06, "loss": 0.2738, "step": 34660 }, { "epoch": 2.0506299166025905, "grad_norm": 2.6858458518981934, "learning_rate": 1.4426965785485958e-06, "loss": 0.2805, "step": 34670 }, { "epoch": 2.051221387590939, "grad_norm": 3.644690752029419, "learning_rate": 1.4410595704817753e-06, "loss": 0.1944, "step": 34680 }, { "epoch": 2.0518128585792867, "grad_norm": 3.400470495223999, "learning_rate": 1.439423198062978e-06, "loss": 0.2084, "step": 34690 }, { "epoch": 2.0524043295676346, "grad_norm": 2.410771608352661, "learning_rate": 1.4377874619594227e-06, "loss": 0.2698, "step": 34700 }, { "epoch": 2.0529958005559825, "grad_norm": 3.045090675354004, "learning_rate": 1.4361523628380693e-06, "loss": 0.2024, "step": 34710 }, { "epoch": 2.053587271544331, "grad_norm": 2.2303292751312256, "learning_rate": 1.4345179013656214e-06, "loss": 0.2553, "step": 34720 }, { "epoch": 2.0541787425326787, "grad_norm": 3.487971782684326, "learning_rate": 1.4328840782085183e-06, "loss": 0.2636, "step": 34730 }, { "epoch": 2.0547702135210266, "grad_norm": 2.8958706855773926, "learning_rate": 1.4312508940329423e-06, "loss": 0.2357, "step": 34740 }, { "epoch": 2.055361684509375, "grad_norm": 3.3809475898742676, "learning_rate": 1.4296183495048136e-06, "loss": 0.224, "step": 34750 }, { "epoch": 2.055953155497723, "grad_norm": 2.4259307384490967, "learning_rate": 1.4279864452897903e-06, "loss": 0.2473, "step": 34760 }, { "epoch": 2.0565446264860707, "grad_norm": 4.792229175567627, "learning_rate": 1.4263551820532734e-06, "loss": 0.3176, "step": 34770 }, { "epoch": 2.057136097474419, "grad_norm": 3.0800349712371826, "learning_rate": 1.4247245604603989e-06, "loss": 0.3077, "step": 34780 }, { "epoch": 2.057727568462767, "grad_norm": 3.6339664459228516, "learning_rate": 1.4230945811760424e-06, "loss": 0.2672, "step": 34790 }, { "epoch": 2.058319039451115, "grad_norm": 2.0235559940338135, "learning_rate": 1.421465244864816e-06, "loss": 0.2582, "step": 34800 }, { "epoch": 2.058910510439463, "grad_norm": 3.172271251678467, "learning_rate": 1.419836552191074e-06, "loss": 0.2194, "step": 34810 }, { "epoch": 2.059501981427811, "grad_norm": 1.9033660888671875, "learning_rate": 1.4182085038189028e-06, "loss": 0.2241, "step": 34820 }, { "epoch": 2.060093452416159, "grad_norm": 3.209027051925659, "learning_rate": 1.4165811004121315e-06, "loss": 0.2164, "step": 34830 }, { "epoch": 2.060684923404507, "grad_norm": 2.9743170738220215, "learning_rate": 1.414954342634322e-06, "loss": 0.3275, "step": 34840 }, { "epoch": 2.061276394392855, "grad_norm": 3.8588364124298096, "learning_rate": 1.413328231148773e-06, "loss": 0.3184, "step": 34850 }, { "epoch": 2.061867865381203, "grad_norm": 3.724870443344116, "learning_rate": 1.4117027666185236e-06, "loss": 0.2594, "step": 34860 }, { "epoch": 2.062459336369551, "grad_norm": 3.1082258224487305, "learning_rate": 1.410077949706345e-06, "loss": 0.295, "step": 34870 }, { "epoch": 2.0630508073578993, "grad_norm": 4.538603782653809, "learning_rate": 1.4084537810747472e-06, "loss": 0.2196, "step": 34880 }, { "epoch": 2.063642278346247, "grad_norm": 4.086660861968994, "learning_rate": 1.4068302613859736e-06, "loss": 0.2411, "step": 34890 }, { "epoch": 2.064233749334595, "grad_norm": 3.8851053714752197, "learning_rate": 1.4052073913020036e-06, "loss": 0.3064, "step": 34900 }, { "epoch": 2.0648252203229434, "grad_norm": 2.958585739135742, "learning_rate": 1.4035851714845538e-06, "loss": 0.2445, "step": 34910 }, { "epoch": 2.0654166913112912, "grad_norm": 3.651308298110962, "learning_rate": 1.4019636025950718e-06, "loss": 0.2641, "step": 34920 }, { "epoch": 2.066008162299639, "grad_norm": 4.329830646514893, "learning_rate": 1.4003426852947445e-06, "loss": 0.3101, "step": 34930 }, { "epoch": 2.066599633287987, "grad_norm": 4.009402751922607, "learning_rate": 1.398722420244488e-06, "loss": 0.2893, "step": 34940 }, { "epoch": 2.0671911042763353, "grad_norm": 2.2943334579467773, "learning_rate": 1.3971028081049577e-06, "loss": 0.2486, "step": 34950 }, { "epoch": 2.0677825752646832, "grad_norm": 1.9065144062042236, "learning_rate": 1.3954838495365384e-06, "loss": 0.2587, "step": 34960 }, { "epoch": 2.068374046253031, "grad_norm": 3.6934940814971924, "learning_rate": 1.3938655451993496e-06, "loss": 0.3034, "step": 34970 }, { "epoch": 2.0689655172413794, "grad_norm": 3.941056489944458, "learning_rate": 1.392247895753246e-06, "loss": 0.2979, "step": 34980 }, { "epoch": 2.0695569882297273, "grad_norm": 4.196787357330322, "learning_rate": 1.390630901857814e-06, "loss": 0.2247, "step": 34990 }, { "epoch": 2.070148459218075, "grad_norm": 3.100497245788574, "learning_rate": 1.3890145641723715e-06, "loss": 0.2377, "step": 35000 }, { "epoch": 2.0707399302064236, "grad_norm": 0.9917346239089966, "learning_rate": 1.3873988833559691e-06, "loss": 0.2482, "step": 35010 }, { "epoch": 2.0713314011947714, "grad_norm": 3.916677236557007, "learning_rate": 1.385783860067393e-06, "loss": 0.3089, "step": 35020 }, { "epoch": 2.0719228721831193, "grad_norm": 1.7510710954666138, "learning_rate": 1.3841694949651567e-06, "loss": 0.2149, "step": 35030 }, { "epoch": 2.0725143431714677, "grad_norm": 4.039798259735107, "learning_rate": 1.3825557887075067e-06, "loss": 0.2808, "step": 35040 }, { "epoch": 2.0731058141598155, "grad_norm": 1.8532016277313232, "learning_rate": 1.3809427419524235e-06, "loss": 0.2148, "step": 35050 }, { "epoch": 2.0736972851481634, "grad_norm": 3.709458351135254, "learning_rate": 1.3793303553576145e-06, "loss": 0.2708, "step": 35060 }, { "epoch": 2.0742887561365113, "grad_norm": 5.710624694824219, "learning_rate": 1.3777186295805224e-06, "loss": 0.2851, "step": 35070 }, { "epoch": 2.0748802271248596, "grad_norm": 3.3741378784179688, "learning_rate": 1.3761075652783158e-06, "loss": 0.1934, "step": 35080 }, { "epoch": 2.0754716981132075, "grad_norm": 5.03251838684082, "learning_rate": 1.374497163107898e-06, "loss": 0.2468, "step": 35090 }, { "epoch": 2.0760631691015554, "grad_norm": 3.9133996963500977, "learning_rate": 1.3728874237258992e-06, "loss": 0.2851, "step": 35100 }, { "epoch": 2.0766546400899037, "grad_norm": 3.36279296875, "learning_rate": 1.371278347788679e-06, "loss": 0.1855, "step": 35110 }, { "epoch": 2.0772461110782516, "grad_norm": 4.086946964263916, "learning_rate": 1.3696699359523307e-06, "loss": 0.2629, "step": 35120 }, { "epoch": 2.0778375820665995, "grad_norm": 3.960537910461426, "learning_rate": 1.368062188872671e-06, "loss": 0.2552, "step": 35130 }, { "epoch": 2.078429053054948, "grad_norm": 3.475950002670288, "learning_rate": 1.3664551072052512e-06, "loss": 0.2435, "step": 35140 }, { "epoch": 2.0790205240432957, "grad_norm": 2.3106210231781006, "learning_rate": 1.3648486916053477e-06, "loss": 0.2617, "step": 35150 }, { "epoch": 2.0796119950316436, "grad_norm": 2.6220285892486572, "learning_rate": 1.3632429427279643e-06, "loss": 0.1976, "step": 35160 }, { "epoch": 2.0802034660199915, "grad_norm": 2.8127663135528564, "learning_rate": 1.361637861227838e-06, "loss": 0.2573, "step": 35170 }, { "epoch": 2.08079493700834, "grad_norm": 6.42233943939209, "learning_rate": 1.3600334477594277e-06, "loss": 0.2879, "step": 35180 }, { "epoch": 2.0813864079966877, "grad_norm": 2.374314069747925, "learning_rate": 1.3584297029769254e-06, "loss": 0.2679, "step": 35190 }, { "epoch": 2.0819778789850356, "grad_norm": 2.829287528991699, "learning_rate": 1.3568266275342463e-06, "loss": 0.2383, "step": 35200 }, { "epoch": 2.082569349973384, "grad_norm": 2.6599221229553223, "learning_rate": 1.3552242220850344e-06, "loss": 0.1968, "step": 35210 }, { "epoch": 2.083160820961732, "grad_norm": 4.325389862060547, "learning_rate": 1.3536224872826604e-06, "loss": 0.246, "step": 35220 }, { "epoch": 2.0837522919500797, "grad_norm": 6.37714147567749, "learning_rate": 1.3520214237802203e-06, "loss": 0.2635, "step": 35230 }, { "epoch": 2.084343762938428, "grad_norm": 5.474202632904053, "learning_rate": 1.3504210322305398e-06, "loss": 0.2598, "step": 35240 }, { "epoch": 2.084935233926776, "grad_norm": 1.923708438873291, "learning_rate": 1.3488213132861657e-06, "loss": 0.2527, "step": 35250 }, { "epoch": 2.085526704915124, "grad_norm": 4.474984169006348, "learning_rate": 1.347222267599376e-06, "loss": 0.2435, "step": 35260 }, { "epoch": 2.086118175903472, "grad_norm": 5.093284606933594, "learning_rate": 1.345623895822169e-06, "loss": 0.2804, "step": 35270 }, { "epoch": 2.08670964689182, "grad_norm": 5.213742733001709, "learning_rate": 1.3440261986062726e-06, "loss": 0.3175, "step": 35280 }, { "epoch": 2.087301117880168, "grad_norm": 2.343261241912842, "learning_rate": 1.3424291766031366e-06, "loss": 0.2204, "step": 35290 }, { "epoch": 2.087892588868516, "grad_norm": 13.719453811645508, "learning_rate": 1.3408328304639356e-06, "loss": 0.2885, "step": 35300 }, { "epoch": 2.088484059856864, "grad_norm": 2.723346471786499, "learning_rate": 1.339237160839572e-06, "loss": 0.2019, "step": 35310 }, { "epoch": 2.089075530845212, "grad_norm": 3.6287217140197754, "learning_rate": 1.3376421683806672e-06, "loss": 0.3347, "step": 35320 }, { "epoch": 2.08966700183356, "grad_norm": 3.1115264892578125, "learning_rate": 1.3360478537375714e-06, "loss": 0.2283, "step": 35330 }, { "epoch": 2.0902584728219082, "grad_norm": 2.886557102203369, "learning_rate": 1.3344542175603556e-06, "loss": 0.2655, "step": 35340 }, { "epoch": 2.090849943810256, "grad_norm": 4.388082504272461, "learning_rate": 1.3328612604988135e-06, "loss": 0.2454, "step": 35350 }, { "epoch": 2.091441414798604, "grad_norm": 4.307737827301025, "learning_rate": 1.331268983202465e-06, "loss": 0.2336, "step": 35360 }, { "epoch": 2.0920328857869523, "grad_norm": 5.300798416137695, "learning_rate": 1.3296773863205495e-06, "loss": 0.2924, "step": 35370 }, { "epoch": 2.0926243567753002, "grad_norm": 3.4763495922088623, "learning_rate": 1.328086470502032e-06, "loss": 0.2395, "step": 35380 }, { "epoch": 2.093215827763648, "grad_norm": 4.046213626861572, "learning_rate": 1.3264962363955977e-06, "loss": 0.2932, "step": 35390 }, { "epoch": 2.093807298751996, "grad_norm": 3.6347787380218506, "learning_rate": 1.3249066846496532e-06, "loss": 0.2547, "step": 35400 }, { "epoch": 2.0943987697403443, "grad_norm": 3.999889850616455, "learning_rate": 1.3233178159123303e-06, "loss": 0.2369, "step": 35410 }, { "epoch": 2.094990240728692, "grad_norm": 5.412294864654541, "learning_rate": 1.3217296308314786e-06, "loss": 0.2178, "step": 35420 }, { "epoch": 2.09558171171704, "grad_norm": 3.587629556655884, "learning_rate": 1.3201421300546707e-06, "loss": 0.2317, "step": 35430 }, { "epoch": 2.0961731827053884, "grad_norm": 4.471014022827148, "learning_rate": 1.3185553142291994e-06, "loss": 0.2715, "step": 35440 }, { "epoch": 2.0967646536937363, "grad_norm": 2.481844663619995, "learning_rate": 1.3169691840020802e-06, "loss": 0.26, "step": 35450 }, { "epoch": 2.097356124682084, "grad_norm": 4.655752182006836, "learning_rate": 1.3153837400200453e-06, "loss": 0.2764, "step": 35460 }, { "epoch": 2.0979475956704325, "grad_norm": 3.67112398147583, "learning_rate": 1.313798982929552e-06, "loss": 0.2464, "step": 35470 }, { "epoch": 2.0985390666587804, "grad_norm": 3.6378026008605957, "learning_rate": 1.312214913376774e-06, "loss": 0.2438, "step": 35480 }, { "epoch": 2.0991305376471283, "grad_norm": 2.884944438934326, "learning_rate": 1.3106315320076038e-06, "loss": 0.2622, "step": 35490 }, { "epoch": 2.0997220086354766, "grad_norm": 3.8647265434265137, "learning_rate": 1.3090488394676578e-06, "loss": 0.3227, "step": 35500 }, { "epoch": 2.1003134796238245, "grad_norm": 3.5330440998077393, "learning_rate": 1.3074668364022663e-06, "loss": 0.2694, "step": 35510 }, { "epoch": 2.1003726267226592, "eval_accuracy": 0.8642246398509499, "eval_loss": 0.32597148418426514, "eval_runtime": 159.6252, "eval_safe_aucpr": 0.9230834222688074, "eval_safe_f1": 0.8497146013625483, "eval_safe_fpr": 0.13670742279752432, "eval_safe_precision": 0.8345932650920533, "eval_safe_recall": 0.8653939916738551, "eval_samples_per_second": 376.595, "eval_steps_per_second": 47.079, "eval_unsafe_aucpr": 0.9565719310824816, "eval_unsafe_f1": 0.876179495737128, "eval_unsafe_fpr": 0.13460600832614436, "eval_unsafe_precision": 0.8894569870945883, "eval_unsafe_recall": 0.8632925772024752, "step": 35511 }, { "epoch": 2.1009049506121724, "grad_norm": 3.1730713844299316, "learning_rate": 1.3058855234564832e-06, "loss": 0.2517, "step": 35520 }, { "epoch": 2.1014964216005203, "grad_norm": 4.2889838218688965, "learning_rate": 1.3043049012750778e-06, "loss": 0.2658, "step": 35530 }, { "epoch": 2.1020878925888686, "grad_norm": 2.0397629737854004, "learning_rate": 1.3027249705025379e-06, "loss": 0.2664, "step": 35540 }, { "epoch": 2.1026793635772165, "grad_norm": 3.336432456970215, "learning_rate": 1.3011457317830715e-06, "loss": 0.2435, "step": 35550 }, { "epoch": 2.1032708345655644, "grad_norm": 3.0176641941070557, "learning_rate": 1.2995671857606018e-06, "loss": 0.2251, "step": 35560 }, { "epoch": 2.1038623055539127, "grad_norm": 3.428039312362671, "learning_rate": 1.2979893330787726e-06, "loss": 0.3232, "step": 35570 }, { "epoch": 2.1044537765422606, "grad_norm": 3.5257010459899902, "learning_rate": 1.296412174380941e-06, "loss": 0.3104, "step": 35580 }, { "epoch": 2.1050452475306085, "grad_norm": 3.412670612335205, "learning_rate": 1.2948357103101862e-06, "loss": 0.3381, "step": 35590 }, { "epoch": 2.105636718518957, "grad_norm": 3.6397242546081543, "learning_rate": 1.2932599415093e-06, "loss": 0.2402, "step": 35600 }, { "epoch": 2.1062281895073047, "grad_norm": 4.095355987548828, "learning_rate": 1.2916848686207905e-06, "loss": 0.2618, "step": 35610 }, { "epoch": 2.1068196604956526, "grad_norm": 3.0006766319274902, "learning_rate": 1.2901104922868866e-06, "loss": 0.265, "step": 35620 }, { "epoch": 2.1074111314840005, "grad_norm": 1.9666520357131958, "learning_rate": 1.2885368131495285e-06, "loss": 0.2203, "step": 35630 }, { "epoch": 2.108002602472349, "grad_norm": 2.979280471801758, "learning_rate": 1.286963831850374e-06, "loss": 0.2091, "step": 35640 }, { "epoch": 2.1085940734606967, "grad_norm": 2.0937297344207764, "learning_rate": 1.2853915490307953e-06, "loss": 0.242, "step": 35650 }, { "epoch": 2.1091855444490446, "grad_norm": 2.776287317276001, "learning_rate": 1.2838199653318825e-06, "loss": 0.3161, "step": 35660 }, { "epoch": 2.109777015437393, "grad_norm": 2.748450756072998, "learning_rate": 1.282249081394438e-06, "loss": 0.2156, "step": 35670 }, { "epoch": 2.110368486425741, "grad_norm": 4.882239818572998, "learning_rate": 1.280678897858979e-06, "loss": 0.3072, "step": 35680 }, { "epoch": 2.1109599574140887, "grad_norm": 2.1799936294555664, "learning_rate": 1.2791094153657394e-06, "loss": 0.2993, "step": 35690 }, { "epoch": 2.111551428402437, "grad_norm": 2.053194046020508, "learning_rate": 1.2775406345546637e-06, "loss": 0.2532, "step": 35700 }, { "epoch": 2.112142899390785, "grad_norm": 3.0923871994018555, "learning_rate": 1.2759725560654146e-06, "loss": 0.2933, "step": 35710 }, { "epoch": 2.112734370379133, "grad_norm": 2.9900288581848145, "learning_rate": 1.2744051805373636e-06, "loss": 0.2301, "step": 35720 }, { "epoch": 2.113325841367481, "grad_norm": 4.0050530433654785, "learning_rate": 1.2728385086096006e-06, "loss": 0.2464, "step": 35730 }, { "epoch": 2.113917312355829, "grad_norm": 6.32888650894165, "learning_rate": 1.2712725409209246e-06, "loss": 0.3479, "step": 35740 }, { "epoch": 2.114508783344177, "grad_norm": 1.8969404697418213, "learning_rate": 1.2697072781098481e-06, "loss": 0.2578, "step": 35750 }, { "epoch": 2.115100254332525, "grad_norm": 4.6414337158203125, "learning_rate": 1.2681427208145998e-06, "loss": 0.259, "step": 35760 }, { "epoch": 2.115691725320873, "grad_norm": 3.5302579402923584, "learning_rate": 1.2665788696731148e-06, "loss": 0.2503, "step": 35770 }, { "epoch": 2.116283196309221, "grad_norm": 8.487517356872559, "learning_rate": 1.265015725323046e-06, "loss": 0.3214, "step": 35780 }, { "epoch": 2.116874667297569, "grad_norm": 3.7677955627441406, "learning_rate": 1.2634532884017543e-06, "loss": 0.2241, "step": 35790 }, { "epoch": 2.1174661382859172, "grad_norm": 2.862657070159912, "learning_rate": 1.2618915595463129e-06, "loss": 0.2387, "step": 35800 }, { "epoch": 2.118057609274265, "grad_norm": 4.551316738128662, "learning_rate": 1.2603305393935086e-06, "loss": 0.2809, "step": 35810 }, { "epoch": 2.118649080262613, "grad_norm": 2.417414903640747, "learning_rate": 1.2587702285798348e-06, "loss": 0.2861, "step": 35820 }, { "epoch": 2.1192405512509613, "grad_norm": 1.7181646823883057, "learning_rate": 1.257210627741501e-06, "loss": 0.2527, "step": 35830 }, { "epoch": 2.119832022239309, "grad_norm": 4.647910118103027, "learning_rate": 1.255651737514423e-06, "loss": 0.3041, "step": 35840 }, { "epoch": 2.120423493227657, "grad_norm": 3.0320234298706055, "learning_rate": 1.254093558534229e-06, "loss": 0.3097, "step": 35850 }, { "epoch": 2.121014964216005, "grad_norm": 2.882004499435425, "learning_rate": 1.2525360914362548e-06, "loss": 0.2337, "step": 35860 }, { "epoch": 2.1216064352043533, "grad_norm": 2.6081957817077637, "learning_rate": 1.2509793368555503e-06, "loss": 0.31, "step": 35870 }, { "epoch": 2.122197906192701, "grad_norm": 6.614410400390625, "learning_rate": 1.249423295426871e-06, "loss": 0.2952, "step": 35880 }, { "epoch": 2.122789377181049, "grad_norm": 4.254317283630371, "learning_rate": 1.2478679677846817e-06, "loss": 0.2987, "step": 35890 }, { "epoch": 2.1233808481693974, "grad_norm": 3.191835403442383, "learning_rate": 1.2463133545631594e-06, "loss": 0.25, "step": 35900 }, { "epoch": 2.1239723191577453, "grad_norm": 2.929776668548584, "learning_rate": 1.244759456396186e-06, "loss": 0.199, "step": 35910 }, { "epoch": 2.124563790146093, "grad_norm": 6.07573127746582, "learning_rate": 1.243206273917355e-06, "loss": 0.3144, "step": 35920 }, { "epoch": 2.1251552611344415, "grad_norm": 2.9897992610931396, "learning_rate": 1.2416538077599666e-06, "loss": 0.2492, "step": 35930 }, { "epoch": 2.1257467321227894, "grad_norm": 4.041589260101318, "learning_rate": 1.2401020585570271e-06, "loss": 0.2436, "step": 35940 }, { "epoch": 2.1263382031111373, "grad_norm": 2.4372191429138184, "learning_rate": 1.2385510269412549e-06, "loss": 0.2657, "step": 35950 }, { "epoch": 2.126929674099485, "grad_norm": 2.007127046585083, "learning_rate": 1.2370007135450707e-06, "loss": 0.2451, "step": 35960 }, { "epoch": 2.1275211450878335, "grad_norm": 4.954339504241943, "learning_rate": 1.2354511190006074e-06, "loss": 0.2672, "step": 35970 }, { "epoch": 2.1281126160761814, "grad_norm": 3.151386022567749, "learning_rate": 1.233902243939701e-06, "loss": 0.2581, "step": 35980 }, { "epoch": 2.1287040870645293, "grad_norm": 3.2401909828186035, "learning_rate": 1.232354088993895e-06, "loss": 0.258, "step": 35990 }, { "epoch": 2.1292955580528776, "grad_norm": 3.1167514324188232, "learning_rate": 1.2308066547944408e-06, "loss": 0.3138, "step": 36000 }, { "epoch": 2.1298870290412255, "grad_norm": 2.700554847717285, "learning_rate": 1.2292599419722933e-06, "loss": 0.2272, "step": 36010 }, { "epoch": 2.1304785000295734, "grad_norm": 3.062814474105835, "learning_rate": 1.2277139511581171e-06, "loss": 0.3216, "step": 36020 }, { "epoch": 2.1310699710179217, "grad_norm": 4.606808185577393, "learning_rate": 1.2261686829822775e-06, "loss": 0.2636, "step": 36030 }, { "epoch": 2.1316614420062696, "grad_norm": 2.9336135387420654, "learning_rate": 1.2246241380748504e-06, "loss": 0.2418, "step": 36040 }, { "epoch": 2.1322529129946175, "grad_norm": 3.2143325805664062, "learning_rate": 1.2230803170656125e-06, "loss": 0.2752, "step": 36050 }, { "epoch": 2.132844383982966, "grad_norm": 6.079376220703125, "learning_rate": 1.2215372205840472e-06, "loss": 0.2498, "step": 36060 }, { "epoch": 2.1334358549713137, "grad_norm": 4.09152889251709, "learning_rate": 1.2199948492593427e-06, "loss": 0.3147, "step": 36070 }, { "epoch": 2.1340273259596616, "grad_norm": 3.13192081451416, "learning_rate": 1.218453203720389e-06, "loss": 0.3088, "step": 36080 }, { "epoch": 2.1346187969480095, "grad_norm": 2.3775713443756104, "learning_rate": 1.2169122845957852e-06, "loss": 0.2665, "step": 36090 }, { "epoch": 2.135210267936358, "grad_norm": 2.6690096855163574, "learning_rate": 1.215372092513829e-06, "loss": 0.2742, "step": 36100 }, { "epoch": 2.1358017389247057, "grad_norm": 3.2272703647613525, "learning_rate": 1.213832628102526e-06, "loss": 0.2777, "step": 36110 }, { "epoch": 2.1363932099130536, "grad_norm": 2.690793037414551, "learning_rate": 1.212293891989582e-06, "loss": 0.2519, "step": 36120 }, { "epoch": 2.136984680901402, "grad_norm": 5.692530632019043, "learning_rate": 1.2107558848024053e-06, "loss": 0.2638, "step": 36130 }, { "epoch": 2.13757615188975, "grad_norm": 2.4734420776367188, "learning_rate": 1.2092186071681111e-06, "loss": 0.238, "step": 36140 }, { "epoch": 2.1381676228780977, "grad_norm": 2.2316982746124268, "learning_rate": 1.207682059713512e-06, "loss": 0.2483, "step": 36150 }, { "epoch": 2.138759093866446, "grad_norm": 4.146926403045654, "learning_rate": 1.2061462430651285e-06, "loss": 0.2772, "step": 36160 }, { "epoch": 2.139350564854794, "grad_norm": 3.602257490158081, "learning_rate": 1.2046111578491772e-06, "loss": 0.2216, "step": 36170 }, { "epoch": 2.139942035843142, "grad_norm": 1.4924331903457642, "learning_rate": 1.2030768046915811e-06, "loss": 0.2374, "step": 36180 }, { "epoch": 2.14053350683149, "grad_norm": 2.5124239921569824, "learning_rate": 1.2015431842179628e-06, "loss": 0.2519, "step": 36190 }, { "epoch": 2.141124977819838, "grad_norm": 2.889117956161499, "learning_rate": 1.2000102970536442e-06, "loss": 0.2884, "step": 36200 }, { "epoch": 2.141716448808186, "grad_norm": 2.1339261531829834, "learning_rate": 1.1984781438236532e-06, "loss": 0.2432, "step": 36210 }, { "epoch": 2.142307919796534, "grad_norm": 3.0326895713806152, "learning_rate": 1.1969467251527132e-06, "loss": 0.2925, "step": 36220 }, { "epoch": 2.142899390784882, "grad_norm": 2.2904157638549805, "learning_rate": 1.1954160416652521e-06, "loss": 0.2212, "step": 36230 }, { "epoch": 2.14349086177323, "grad_norm": 1.9657909870147705, "learning_rate": 1.193886093985396e-06, "loss": 0.2041, "step": 36240 }, { "epoch": 2.144082332761578, "grad_norm": 5.282613754272461, "learning_rate": 1.1923568827369694e-06, "loss": 0.2442, "step": 36250 }, { "epoch": 2.144673803749926, "grad_norm": 1.7047197818756104, "learning_rate": 1.190828408543501e-06, "loss": 0.223, "step": 36260 }, { "epoch": 2.145265274738274, "grad_norm": 3.8736841678619385, "learning_rate": 1.1893006720282156e-06, "loss": 0.2848, "step": 36270 }, { "epoch": 2.145856745726622, "grad_norm": 3.9920437335968018, "learning_rate": 1.1877736738140376e-06, "loss": 0.2257, "step": 36280 }, { "epoch": 2.1464482167149703, "grad_norm": 4.330691814422607, "learning_rate": 1.18624741452359e-06, "loss": 0.2768, "step": 36290 }, { "epoch": 2.147039687703318, "grad_norm": 2.4402482509613037, "learning_rate": 1.1847218947791971e-06, "loss": 0.263, "step": 36300 }, { "epoch": 2.147631158691666, "grad_norm": 4.285919666290283, "learning_rate": 1.1831971152028792e-06, "loss": 0.2977, "step": 36310 }, { "epoch": 2.1482226296800144, "grad_norm": 4.373694896697998, "learning_rate": 1.1816730764163542e-06, "loss": 0.2925, "step": 36320 }, { "epoch": 2.1488141006683623, "grad_norm": 3.47001051902771, "learning_rate": 1.180149779041041e-06, "loss": 0.2269, "step": 36330 }, { "epoch": 2.14940557165671, "grad_norm": 5.136878967285156, "learning_rate": 1.178627223698053e-06, "loss": 0.318, "step": 36340 }, { "epoch": 2.149997042645058, "grad_norm": 3.337109327316284, "learning_rate": 1.1771054110082037e-06, "loss": 0.2669, "step": 36350 }, { "epoch": 2.1505885136334064, "grad_norm": 3.933936595916748, "learning_rate": 1.1755843415920011e-06, "loss": 0.2392, "step": 36360 }, { "epoch": 2.1511799846217543, "grad_norm": 6.806023120880127, "learning_rate": 1.1740640160696535e-06, "loss": 0.2267, "step": 36370 }, { "epoch": 2.151771455610102, "grad_norm": 5.302818298339844, "learning_rate": 1.1725444350610627e-06, "loss": 0.2386, "step": 36380 }, { "epoch": 2.1523629265984505, "grad_norm": 3.9712860584259033, "learning_rate": 1.171025599185827e-06, "loss": 0.2514, "step": 36390 }, { "epoch": 2.1529543975867984, "grad_norm": 3.883683204650879, "learning_rate": 1.1695075090632446e-06, "loss": 0.256, "step": 36400 }, { "epoch": 2.1535458685751463, "grad_norm": 3.1203293800354004, "learning_rate": 1.1679901653123047e-06, "loss": 0.2285, "step": 36410 }, { "epoch": 2.154137339563494, "grad_norm": 3.0994677543640137, "learning_rate": 1.1664735685516967e-06, "loss": 0.2118, "step": 36420 }, { "epoch": 2.1547288105518425, "grad_norm": 4.155055046081543, "learning_rate": 1.164957719399802e-06, "loss": 0.2677, "step": 36430 }, { "epoch": 2.1553202815401904, "grad_norm": 3.0268216133117676, "learning_rate": 1.1634426184746978e-06, "loss": 0.2614, "step": 36440 }, { "epoch": 2.1559117525285383, "grad_norm": 2.8034310340881348, "learning_rate": 1.1619282663941586e-06, "loss": 0.2285, "step": 36450 }, { "epoch": 2.1565032235168866, "grad_norm": 4.631202697753906, "learning_rate": 1.16041466377565e-06, "loss": 0.2968, "step": 36460 }, { "epoch": 2.1570946945052345, "grad_norm": 4.340625762939453, "learning_rate": 1.1589018112363353e-06, "loss": 0.2486, "step": 36470 }, { "epoch": 2.1576861654935824, "grad_norm": 3.05794358253479, "learning_rate": 1.1573897093930688e-06, "loss": 0.2595, "step": 36480 }, { "epoch": 2.1582776364819307, "grad_norm": 2.225684642791748, "learning_rate": 1.155878358862404e-06, "loss": 0.2778, "step": 36490 }, { "epoch": 2.1588691074702786, "grad_norm": 4.191442489624023, "learning_rate": 1.1543677602605794e-06, "loss": 0.2657, "step": 36500 }, { "epoch": 2.1594605784586265, "grad_norm": 2.5723934173583984, "learning_rate": 1.1528579142035355e-06, "loss": 0.2566, "step": 36510 }, { "epoch": 2.160052049446975, "grad_norm": 4.5175065994262695, "learning_rate": 1.1513488213069017e-06, "loss": 0.2615, "step": 36520 }, { "epoch": 2.1606435204353227, "grad_norm": 3.007746934890747, "learning_rate": 1.149840482185999e-06, "loss": 0.2703, "step": 36530 }, { "epoch": 2.1612349914236706, "grad_norm": 2.597292423248291, "learning_rate": 1.1483328974558463e-06, "loss": 0.2604, "step": 36540 }, { "epoch": 2.1618264624120185, "grad_norm": 3.198725461959839, "learning_rate": 1.1468260677311489e-06, "loss": 0.2557, "step": 36550 }, { "epoch": 2.162417933400367, "grad_norm": 4.182612419128418, "learning_rate": 1.145319993626309e-06, "loss": 0.1867, "step": 36560 }, { "epoch": 2.1630094043887147, "grad_norm": 5.516971588134766, "learning_rate": 1.1438146757554182e-06, "loss": 0.2467, "step": 36570 }, { "epoch": 2.1636008753770626, "grad_norm": 2.8767342567443848, "learning_rate": 1.1423101147322586e-06, "loss": 0.207, "step": 36580 }, { "epoch": 2.164192346365411, "grad_norm": 2.728006601333618, "learning_rate": 1.140806311170308e-06, "loss": 0.2872, "step": 36590 }, { "epoch": 2.164783817353759, "grad_norm": 2.5797433853149414, "learning_rate": 1.1393032656827303e-06, "loss": 0.3391, "step": 36600 }, { "epoch": 2.1653752883421067, "grad_norm": 2.759378433227539, "learning_rate": 1.137800978882385e-06, "loss": 0.1935, "step": 36610 }, { "epoch": 2.165966759330455, "grad_norm": 2.252959966659546, "learning_rate": 1.136299451381819e-06, "loss": 0.2428, "step": 36620 }, { "epoch": 2.166558230318803, "grad_norm": 2.6755452156066895, "learning_rate": 1.134798683793269e-06, "loss": 0.2561, "step": 36630 }, { "epoch": 2.167149701307151, "grad_norm": 3.5991575717926025, "learning_rate": 1.133298676728666e-06, "loss": 0.2765, "step": 36640 }, { "epoch": 2.167741172295499, "grad_norm": 3.443843364715576, "learning_rate": 1.1317994307996257e-06, "loss": 0.2758, "step": 36650 }, { "epoch": 2.168332643283847, "grad_norm": 3.2752530574798584, "learning_rate": 1.1303009466174591e-06, "loss": 0.2053, "step": 36660 }, { "epoch": 2.168924114272195, "grad_norm": 2.0055923461914062, "learning_rate": 1.1288032247931607e-06, "loss": 0.2718, "step": 36670 }, { "epoch": 2.169515585260543, "grad_norm": 4.40498161315918, "learning_rate": 1.1273062659374194e-06, "loss": 0.3116, "step": 36680 }, { "epoch": 2.170107056248891, "grad_norm": 2.0252113342285156, "learning_rate": 1.1258100706606094e-06, "loss": 0.2157, "step": 36690 }, { "epoch": 2.170698527237239, "grad_norm": 3.1471431255340576, "learning_rate": 1.124314639572795e-06, "loss": 0.2087, "step": 36700 }, { "epoch": 2.171289998225587, "grad_norm": 2.335298776626587, "learning_rate": 1.1228199732837285e-06, "loss": 0.252, "step": 36710 }, { "epoch": 2.171881469213935, "grad_norm": 1.8817201852798462, "learning_rate": 1.1213260724028498e-06, "loss": 0.2244, "step": 36720 }, { "epoch": 2.172472940202283, "grad_norm": 8.582755088806152, "learning_rate": 1.1198329375392894e-06, "loss": 0.2559, "step": 36730 }, { "epoch": 2.173064411190631, "grad_norm": 5.125313758850098, "learning_rate": 1.118340569301861e-06, "loss": 0.244, "step": 36740 }, { "epoch": 2.1736558821789793, "grad_norm": 3.125760078430176, "learning_rate": 1.1168489682990708e-06, "loss": 0.2857, "step": 36750 }, { "epoch": 2.174247353167327, "grad_norm": 4.232656955718994, "learning_rate": 1.1153581351391089e-06, "loss": 0.2877, "step": 36760 }, { "epoch": 2.174838824155675, "grad_norm": 4.781519412994385, "learning_rate": 1.1138680704298515e-06, "loss": 0.2656, "step": 36770 }, { "epoch": 2.1754302951440234, "grad_norm": 2.8000741004943848, "learning_rate": 1.1123787747788648e-06, "loss": 0.2134, "step": 36780 }, { "epoch": 2.1760217661323713, "grad_norm": 5.285323619842529, "learning_rate": 1.1108902487933978e-06, "loss": 0.2732, "step": 36790 }, { "epoch": 2.176613237120719, "grad_norm": 3.2760133743286133, "learning_rate": 1.1094024930803895e-06, "loss": 0.2538, "step": 36800 }, { "epoch": 2.177204708109067, "grad_norm": 4.573400497436523, "learning_rate": 1.1079155082464613e-06, "loss": 0.2688, "step": 36810 }, { "epoch": 2.1777961790974154, "grad_norm": 2.1399343013763428, "learning_rate": 1.106429294897923e-06, "loss": 0.2232, "step": 36820 }, { "epoch": 2.1783876500857633, "grad_norm": 4.90778923034668, "learning_rate": 1.104943853640768e-06, "loss": 0.2591, "step": 36830 }, { "epoch": 2.178979121074111, "grad_norm": 4.4315056800842285, "learning_rate": 1.1034591850806748e-06, "loss": 0.2882, "step": 36840 }, { "epoch": 2.1795705920624595, "grad_norm": 3.7913317680358887, "learning_rate": 1.101975289823009e-06, "loss": 0.2441, "step": 36850 }, { "epoch": 2.1801620630508074, "grad_norm": 2.3566739559173584, "learning_rate": 1.1004921684728177e-06, "loss": 0.2273, "step": 36860 }, { "epoch": 2.1807535340391553, "grad_norm": 4.391221523284912, "learning_rate": 1.099009821634836e-06, "loss": 0.2482, "step": 36870 }, { "epoch": 2.181345005027503, "grad_norm": 4.727273464202881, "learning_rate": 1.0975282499134805e-06, "loss": 0.2521, "step": 36880 }, { "epoch": 2.1819364760158515, "grad_norm": 2.8825886249542236, "learning_rate": 1.0960474539128515e-06, "loss": 0.2599, "step": 36890 }, { "epoch": 2.1825279470041994, "grad_norm": 3.7720019817352295, "learning_rate": 1.0945674342367367e-06, "loss": 0.2965, "step": 36900 }, { "epoch": 2.1831194179925473, "grad_norm": 2.7214670181274414, "learning_rate": 1.0930881914886019e-06, "loss": 0.2808, "step": 36910 }, { "epoch": 2.1837108889808956, "grad_norm": 3.254199981689453, "learning_rate": 1.0916097262716022e-06, "loss": 0.2664, "step": 36920 }, { "epoch": 2.1843023599692435, "grad_norm": 5.114208221435547, "learning_rate": 1.0901320391885684e-06, "loss": 0.2602, "step": 36930 }, { "epoch": 2.1848938309575914, "grad_norm": 4.601163387298584, "learning_rate": 1.0886551308420212e-06, "loss": 0.2736, "step": 36940 }, { "epoch": 2.1854853019459397, "grad_norm": 4.222206115722656, "learning_rate": 1.0871790018341582e-06, "loss": 0.236, "step": 36950 }, { "epoch": 2.1860767729342876, "grad_norm": 4.065635681152344, "learning_rate": 1.0857036527668638e-06, "loss": 0.2189, "step": 36960 }, { "epoch": 2.1866682439226355, "grad_norm": 4.21702241897583, "learning_rate": 1.0842290842417012e-06, "loss": 0.2455, "step": 36970 }, { "epoch": 2.187259714910984, "grad_norm": 3.3090338706970215, "learning_rate": 1.0827552968599151e-06, "loss": 0.2431, "step": 36980 }, { "epoch": 2.1878511858993317, "grad_norm": 2.7334418296813965, "learning_rate": 1.0812822912224348e-06, "loss": 0.2463, "step": 36990 }, { "epoch": 2.1884426568876796, "grad_norm": 6.457273960113525, "learning_rate": 1.079810067929868e-06, "loss": 0.2743, "step": 37000 }, { "epoch": 2.1890341278760275, "grad_norm": 5.484560489654541, "learning_rate": 1.0783386275825047e-06, "loss": 0.3313, "step": 37010 }, { "epoch": 2.189625598864376, "grad_norm": 2.8202574253082275, "learning_rate": 1.0768679707803154e-06, "loss": 0.3098, "step": 37020 }, { "epoch": 2.1902170698527237, "grad_norm": 3.217043399810791, "learning_rate": 1.07539809812295e-06, "loss": 0.264, "step": 37030 }, { "epoch": 2.1908085408410716, "grad_norm": 4.284571647644043, "learning_rate": 1.0739290102097419e-06, "loss": 0.3092, "step": 37040 }, { "epoch": 2.19140001182942, "grad_norm": 3.105205774307251, "learning_rate": 1.0724607076396994e-06, "loss": 0.2462, "step": 37050 }, { "epoch": 2.191991482817768, "grad_norm": 2.4904112815856934, "learning_rate": 1.0709931910115162e-06, "loss": 0.2158, "step": 37060 }, { "epoch": 2.1925829538061157, "grad_norm": 5.854325771331787, "learning_rate": 1.0695264609235617e-06, "loss": 0.2878, "step": 37070 }, { "epoch": 2.193174424794464, "grad_norm": 3.0288994312286377, "learning_rate": 1.0680605179738848e-06, "loss": 0.2498, "step": 37080 }, { "epoch": 2.193765895782812, "grad_norm": 4.083075046539307, "learning_rate": 1.0665953627602163e-06, "loss": 0.268, "step": 37090 }, { "epoch": 2.19435736677116, "grad_norm": 2.7761616706848145, "learning_rate": 1.065130995879962e-06, "loss": 0.2425, "step": 37100 }, { "epoch": 2.194948837759508, "grad_norm": 4.400742053985596, "learning_rate": 1.0636674179302103e-06, "loss": 0.2773, "step": 37110 }, { "epoch": 2.195540308747856, "grad_norm": 2.2056894302368164, "learning_rate": 1.0622046295077235e-06, "loss": 0.2578, "step": 37120 }, { "epoch": 2.196131779736204, "grad_norm": 2.8998610973358154, "learning_rate": 1.0607426312089472e-06, "loss": 0.2174, "step": 37130 }, { "epoch": 2.1967232507245518, "grad_norm": 4.667211532592773, "learning_rate": 1.059281423629998e-06, "loss": 0.2697, "step": 37140 }, { "epoch": 2.1973147217129, "grad_norm": 3.4235341548919678, "learning_rate": 1.0578210073666773e-06, "loss": 0.2934, "step": 37150 }, { "epoch": 2.197906192701248, "grad_norm": 2.5756945610046387, "learning_rate": 1.056361383014459e-06, "loss": 0.1929, "step": 37160 }, { "epoch": 2.198497663689596, "grad_norm": 3.700007915496826, "learning_rate": 1.0549025511684949e-06, "loss": 0.2411, "step": 37170 }, { "epoch": 2.199089134677944, "grad_norm": 5.368896484375, "learning_rate": 1.0534445124236163e-06, "loss": 0.2844, "step": 37180 }, { "epoch": 2.199680605666292, "grad_norm": 3.348568916320801, "learning_rate": 1.0519872673743273e-06, "loss": 0.3062, "step": 37190 }, { "epoch": 2.20027207665464, "grad_norm": 2.6480906009674072, "learning_rate": 1.0505308166148122e-06, "loss": 0.2937, "step": 37200 }, { "epoch": 2.2008635476429883, "grad_norm": 4.947358131408691, "learning_rate": 1.0490751607389288e-06, "loss": 0.2681, "step": 37210 }, { "epoch": 2.201455018631336, "grad_norm": 6.6441168785095215, "learning_rate": 1.0476203003402098e-06, "loss": 0.2477, "step": 37220 }, { "epoch": 2.202046489619684, "grad_norm": 4.554011344909668, "learning_rate": 1.046166236011868e-06, "loss": 0.2801, "step": 37230 }, { "epoch": 2.2026379606080324, "grad_norm": 6.26849889755249, "learning_rate": 1.0447129683467863e-06, "loss": 0.2802, "step": 37240 }, { "epoch": 2.2032294315963803, "grad_norm": 4.6942009925842285, "learning_rate": 1.0432604979375276e-06, "loss": 0.2837, "step": 37250 }, { "epoch": 2.203820902584728, "grad_norm": 4.25251579284668, "learning_rate": 1.0418088253763256e-06, "loss": 0.2551, "step": 37260 }, { "epoch": 2.204412373573076, "grad_norm": 2.6061129570007324, "learning_rate": 1.0403579512550923e-06, "loss": 0.2894, "step": 37270 }, { "epoch": 2.2050038445614244, "grad_norm": 3.534987449645996, "learning_rate": 1.0389078761654115e-06, "loss": 0.278, "step": 37280 }, { "epoch": 2.2055953155497723, "grad_norm": 3.0556087493896484, "learning_rate": 1.0374586006985406e-06, "loss": 0.1751, "step": 37290 }, { "epoch": 2.20618678653812, "grad_norm": 4.074889659881592, "learning_rate": 1.0360101254454153e-06, "loss": 0.2657, "step": 37300 }, { "epoch": 2.2067782575264685, "grad_norm": 3.1298835277557373, "learning_rate": 1.0345624509966397e-06, "loss": 0.2051, "step": 37310 }, { "epoch": 2.2073697285148164, "grad_norm": 2.8761913776397705, "learning_rate": 1.0331155779424954e-06, "loss": 0.2545, "step": 37320 }, { "epoch": 2.2079611995031643, "grad_norm": 3.695089101791382, "learning_rate": 1.0316695068729354e-06, "loss": 0.3277, "step": 37330 }, { "epoch": 2.208552670491512, "grad_norm": 4.653964996337891, "learning_rate": 1.0302242383775846e-06, "loss": 0.2591, "step": 37340 }, { "epoch": 2.2091441414798605, "grad_norm": 3.3654401302337646, "learning_rate": 1.028779773045745e-06, "loss": 0.2149, "step": 37350 }, { "epoch": 2.2097356124682084, "grad_norm": 2.4335293769836426, "learning_rate": 1.0273361114663843e-06, "loss": 0.1971, "step": 37360 }, { "epoch": 2.2103270834565563, "grad_norm": 3.257554054260254, "learning_rate": 1.0258932542281488e-06, "loss": 0.2532, "step": 37370 }, { "epoch": 2.2109185544449046, "grad_norm": 6.745285987854004, "learning_rate": 1.024451201919353e-06, "loss": 0.2745, "step": 37380 }, { "epoch": 2.2115100254332525, "grad_norm": 3.31268572807312, "learning_rate": 1.0230099551279863e-06, "loss": 0.2331, "step": 37390 }, { "epoch": 2.2121014964216004, "grad_norm": 4.428475856781006, "learning_rate": 1.021569514441706e-06, "loss": 0.2829, "step": 37400 }, { "epoch": 2.2126929674099487, "grad_norm": 3.116645574569702, "learning_rate": 1.0201298804478443e-06, "loss": 0.2403, "step": 37410 }, { "epoch": 2.2132844383982966, "grad_norm": 3.3279595375061035, "learning_rate": 1.018691053733402e-06, "loss": 0.2552, "step": 37420 }, { "epoch": 2.2138759093866445, "grad_norm": 2.42789626121521, "learning_rate": 1.0172530348850502e-06, "loss": 0.216, "step": 37430 }, { "epoch": 2.214467380374993, "grad_norm": 2.927405595779419, "learning_rate": 1.0158158244891345e-06, "loss": 0.2367, "step": 37440 }, { "epoch": 2.2150588513633407, "grad_norm": 4.5184645652771, "learning_rate": 1.0143794231316657e-06, "loss": 0.2681, "step": 37450 }, { "epoch": 2.2156503223516886, "grad_norm": 3.733344078063965, "learning_rate": 1.0129438313983297e-06, "loss": 0.2026, "step": 37460 }, { "epoch": 2.2162417933400365, "grad_norm": 3.2809629440307617, "learning_rate": 1.0115090498744786e-06, "loss": 0.2392, "step": 37470 }, { "epoch": 2.216833264328385, "grad_norm": 3.8687705993652344, "learning_rate": 1.010075079145135e-06, "loss": 0.2119, "step": 37480 }, { "epoch": 2.2174247353167327, "grad_norm": 2.6168735027313232, "learning_rate": 1.0086419197949924e-06, "loss": 0.256, "step": 37490 }, { "epoch": 2.2180162063050806, "grad_norm": 3.40051531791687, "learning_rate": 1.007209572408411e-06, "loss": 0.2389, "step": 37500 }, { "epoch": 2.218607677293429, "grad_norm": 4.069873332977295, "learning_rate": 1.0057780375694234e-06, "loss": 0.2395, "step": 37510 }, { "epoch": 2.219199148281777, "grad_norm": 3.359464168548584, "learning_rate": 1.004347315861727e-06, "loss": 0.2931, "step": 37520 }, { "epoch": 2.2197906192701247, "grad_norm": 3.1278889179229736, "learning_rate": 1.0029174078686898e-06, "loss": 0.2826, "step": 37530 }, { "epoch": 2.220382090258473, "grad_norm": 4.502227306365967, "learning_rate": 1.0014883141733483e-06, "loss": 0.2895, "step": 37540 }, { "epoch": 2.220973561246821, "grad_norm": 2.529566526412964, "learning_rate": 1.000060035358405e-06, "loss": 0.298, "step": 37550 }, { "epoch": 2.2215650322351688, "grad_norm": 4.128578186035156, "learning_rate": 9.986325720062349e-07, "loss": 0.1922, "step": 37560 }, { "epoch": 2.222156503223517, "grad_norm": 2.6690332889556885, "learning_rate": 9.972059246988723e-07, "loss": 0.2863, "step": 37570 }, { "epoch": 2.222747974211865, "grad_norm": 3.257185220718384, "learning_rate": 9.957800940180268e-07, "loss": 0.2296, "step": 37580 }, { "epoch": 2.223339445200213, "grad_norm": 4.5664753913879395, "learning_rate": 9.9435508054507e-07, "loss": 0.2914, "step": 37590 }, { "epoch": 2.2239309161885608, "grad_norm": 5.329639434814453, "learning_rate": 9.929308848610438e-07, "loss": 0.2849, "step": 37600 }, { "epoch": 2.224522387176909, "grad_norm": 2.9468538761138916, "learning_rate": 9.91507507546654e-07, "loss": 0.2621, "step": 37610 }, { "epoch": 2.225113858165257, "grad_norm": 4.447097301483154, "learning_rate": 9.900849491822724e-07, "loss": 0.2929, "step": 37620 }, { "epoch": 2.225705329153605, "grad_norm": 4.8824238777160645, "learning_rate": 9.886632103479405e-07, "loss": 0.2675, "step": 37630 }, { "epoch": 2.226296800141953, "grad_norm": 5.58901309967041, "learning_rate": 9.872422916233612e-07, "loss": 0.2623, "step": 37640 }, { "epoch": 2.226888271130301, "grad_norm": 2.7247695922851562, "learning_rate": 9.858221935879072e-07, "loss": 0.2595, "step": 37650 }, { "epoch": 2.227479742118649, "grad_norm": 5.758363246917725, "learning_rate": 9.844029168206126e-07, "loss": 0.2488, "step": 37660 }, { "epoch": 2.2280712131069973, "grad_norm": 4.400136470794678, "learning_rate": 9.829844619001781e-07, "loss": 0.2649, "step": 37670 }, { "epoch": 2.228662684095345, "grad_norm": 2.452852964401245, "learning_rate": 9.815668294049719e-07, "loss": 0.2611, "step": 37680 }, { "epoch": 2.229254155083693, "grad_norm": 2.206658124923706, "learning_rate": 9.80150019913022e-07, "loss": 0.2679, "step": 37690 }, { "epoch": 2.2298456260720414, "grad_norm": 4.041728973388672, "learning_rate": 9.787340340020255e-07, "loss": 0.3397, "step": 37700 }, { "epoch": 2.2304370970603893, "grad_norm": 2.97286319732666, "learning_rate": 9.773188722493402e-07, "loss": 0.2317, "step": 37710 }, { "epoch": 2.231028568048737, "grad_norm": 1.870058536529541, "learning_rate": 9.759045352319908e-07, "loss": 0.2883, "step": 37720 }, { "epoch": 2.231620039037085, "grad_norm": 2.331120252609253, "learning_rate": 9.744910235266634e-07, "loss": 0.2532, "step": 37730 }, { "epoch": 2.2322115100254334, "grad_norm": 3.3212499618530273, "learning_rate": 9.730783377097068e-07, "loss": 0.281, "step": 37740 }, { "epoch": 2.2328029810137813, "grad_norm": 5.793781757354736, "learning_rate": 9.716664783571378e-07, "loss": 0.2752, "step": 37750 }, { "epoch": 2.233394452002129, "grad_norm": 3.7849225997924805, "learning_rate": 9.7025544604463e-07, "loss": 0.2631, "step": 37760 }, { "epoch": 2.2339859229904775, "grad_norm": 5.008594512939453, "learning_rate": 9.688452413475252e-07, "loss": 0.2659, "step": 37770 }, { "epoch": 2.2345773939788254, "grad_norm": 5.734516620635986, "learning_rate": 9.674358648408245e-07, "loss": 0.2831, "step": 37780 }, { "epoch": 2.2351688649671733, "grad_norm": 2.4048244953155518, "learning_rate": 9.660273170991923e-07, "loss": 0.263, "step": 37790 }, { "epoch": 2.235760335955521, "grad_norm": 2.383942127227783, "learning_rate": 9.646195986969544e-07, "loss": 0.2861, "step": 37800 }, { "epoch": 2.2363518069438695, "grad_norm": 2.507859945297241, "learning_rate": 9.632127102080985e-07, "loss": 0.2428, "step": 37810 }, { "epoch": 2.2369432779322174, "grad_norm": 3.320664882659912, "learning_rate": 9.61806652206276e-07, "loss": 0.2566, "step": 37820 }, { "epoch": 2.2375347489205653, "grad_norm": 3.1137170791625977, "learning_rate": 9.604014252647966e-07, "loss": 0.278, "step": 37830 }, { "epoch": 2.2381262199089136, "grad_norm": 3.0619499683380127, "learning_rate": 9.589970299566343e-07, "loss": 0.3075, "step": 37840 }, { "epoch": 2.2387176908972615, "grad_norm": 3.0354244709014893, "learning_rate": 9.575934668544216e-07, "loss": 0.2518, "step": 37850 }, { "epoch": 2.2393091618856094, "grad_norm": 2.7719967365264893, "learning_rate": 9.561907365304511e-07, "loss": 0.2366, "step": 37860 }, { "epoch": 2.2399006328739577, "grad_norm": 2.7793359756469727, "learning_rate": 9.547888395566794e-07, "loss": 0.2352, "step": 37870 }, { "epoch": 2.2404921038623056, "grad_norm": 5.436713695526123, "learning_rate": 9.533877765047186e-07, "loss": 0.2301, "step": 37880 }, { "epoch": 2.2410835748506535, "grad_norm": 5.47089958190918, "learning_rate": 9.519875479458457e-07, "loss": 0.2818, "step": 37890 }, { "epoch": 2.241675045839002, "grad_norm": 4.189024925231934, "learning_rate": 9.505881544509929e-07, "loss": 0.2748, "step": 37900 }, { "epoch": 2.2422665168273497, "grad_norm": 4.135555744171143, "learning_rate": 9.491895965907556e-07, "loss": 0.211, "step": 37910 }, { "epoch": 2.2428579878156976, "grad_norm": 3.3642194271087646, "learning_rate": 9.477918749353864e-07, "loss": 0.264, "step": 37920 }, { "epoch": 2.2434494588040454, "grad_norm": 5.1051411628723145, "learning_rate": 9.46394990054796e-07, "loss": 0.2369, "step": 37930 }, { "epoch": 2.244040929792394, "grad_norm": 1.7307960987091064, "learning_rate": 9.449989425185574e-07, "loss": 0.2034, "step": 37940 }, { "epoch": 2.2446324007807417, "grad_norm": 3.1094441413879395, "learning_rate": 9.436037328958981e-07, "loss": 0.2499, "step": 37950 }, { "epoch": 2.2452238717690896, "grad_norm": 4.134902000427246, "learning_rate": 9.422093617557081e-07, "loss": 0.2271, "step": 37960 }, { "epoch": 2.245815342757438, "grad_norm": 1.611288070678711, "learning_rate": 9.408158296665322e-07, "loss": 0.2272, "step": 37970 }, { "epoch": 2.2464068137457858, "grad_norm": 3.1497855186462402, "learning_rate": 9.394231371965731e-07, "loss": 0.302, "step": 37980 }, { "epoch": 2.2469982847341337, "grad_norm": 3.588629961013794, "learning_rate": 9.380312849136952e-07, "loss": 0.2896, "step": 37990 }, { "epoch": 2.247589755722482, "grad_norm": 5.68558931350708, "learning_rate": 9.36640273385414e-07, "loss": 0.2859, "step": 38000 }, { "epoch": 2.24818122671083, "grad_norm": 4.131667613983154, "learning_rate": 9.35250103178908e-07, "loss": 0.2311, "step": 38010 }, { "epoch": 2.2487726976991778, "grad_norm": 2.3817996978759766, "learning_rate": 9.338607748610082e-07, "loss": 0.2271, "step": 38020 }, { "epoch": 2.249364168687526, "grad_norm": 3.3191120624542236, "learning_rate": 9.324722889982066e-07, "loss": 0.2783, "step": 38030 }, { "epoch": 2.249955639675874, "grad_norm": 2.6765143871307373, "learning_rate": 9.310846461566477e-07, "loss": 0.2706, "step": 38040 }, { "epoch": 2.250547110664222, "grad_norm": 2.8983516693115234, "learning_rate": 9.296978469021359e-07, "loss": 0.2313, "step": 38050 }, { "epoch": 2.2511385816525697, "grad_norm": 4.273983955383301, "learning_rate": 9.283118918001283e-07, "loss": 0.222, "step": 38060 }, { "epoch": 2.251730052640918, "grad_norm": 3.3369576930999756, "learning_rate": 9.269267814157386e-07, "loss": 0.3292, "step": 38070 }, { "epoch": 2.252321523629266, "grad_norm": 5.717566967010498, "learning_rate": 9.255425163137386e-07, "loss": 0.2507, "step": 38080 }, { "epoch": 2.252912994617614, "grad_norm": 3.9518823623657227, "learning_rate": 9.241590970585519e-07, "loss": 0.2732, "step": 38090 }, { "epoch": 2.253504465605962, "grad_norm": 4.572654724121094, "learning_rate": 9.227765242142603e-07, "loss": 0.2563, "step": 38100 }, { "epoch": 2.25409593659431, "grad_norm": 4.460939407348633, "learning_rate": 9.213947983445987e-07, "loss": 0.3087, "step": 38110 }, { "epoch": 2.254687407582658, "grad_norm": 3.531325340270996, "learning_rate": 9.200139200129556e-07, "loss": 0.2324, "step": 38120 }, { "epoch": 2.2552788785710063, "grad_norm": 2.2046666145324707, "learning_rate": 9.186338897823775e-07, "loss": 0.213, "step": 38130 }, { "epoch": 2.255870349559354, "grad_norm": 4.646945953369141, "learning_rate": 9.17254708215561e-07, "loss": 0.2963, "step": 38140 }, { "epoch": 2.256461820547702, "grad_norm": 3.7578182220458984, "learning_rate": 9.158763758748603e-07, "loss": 0.2726, "step": 38150 }, { "epoch": 2.2570532915360504, "grad_norm": 3.648653030395508, "learning_rate": 9.144988933222811e-07, "loss": 0.2654, "step": 38160 }, { "epoch": 2.2576447625243983, "grad_norm": 3.275817394256592, "learning_rate": 9.131222611194815e-07, "loss": 0.2258, "step": 38170 }, { "epoch": 2.258236233512746, "grad_norm": 2.4535868167877197, "learning_rate": 9.117464798277769e-07, "loss": 0.2666, "step": 38180 }, { "epoch": 2.258827704501094, "grad_norm": 3.447112798690796, "learning_rate": 9.103715500081313e-07, "loss": 0.283, "step": 38190 }, { "epoch": 2.2594191754894424, "grad_norm": 2.829606771469116, "learning_rate": 9.089974722211655e-07, "loss": 0.258, "step": 38200 }, { "epoch": 2.2600106464777903, "grad_norm": 4.233550548553467, "learning_rate": 9.076242470271504e-07, "loss": 0.2487, "step": 38210 }, { "epoch": 2.260602117466138, "grad_norm": 3.636887550354004, "learning_rate": 9.062518749860093e-07, "loss": 0.2851, "step": 38220 }, { "epoch": 2.2611935884544865, "grad_norm": 2.660994052886963, "learning_rate": 9.048803566573172e-07, "loss": 0.2611, "step": 38230 }, { "epoch": 2.2617850594428344, "grad_norm": 5.768449306488037, "learning_rate": 9.035096926003038e-07, "loss": 0.2924, "step": 38240 }, { "epoch": 2.2623765304311823, "grad_norm": 3.617582321166992, "learning_rate": 9.021398833738481e-07, "loss": 0.1984, "step": 38250 }, { "epoch": 2.26296800141953, "grad_norm": 2.4449682235717773, "learning_rate": 9.007709295364797e-07, "loss": 0.2348, "step": 38260 }, { "epoch": 2.2635594724078785, "grad_norm": 3.722778797149658, "learning_rate": 8.994028316463832e-07, "loss": 0.2668, "step": 38270 }, { "epoch": 2.2641509433962264, "grad_norm": 4.389488697052002, "learning_rate": 8.980355902613893e-07, "loss": 0.2382, "step": 38280 }, { "epoch": 2.2647424143845742, "grad_norm": 9.638482093811035, "learning_rate": 8.966692059389844e-07, "loss": 0.3371, "step": 38290 }, { "epoch": 2.2653338853729226, "grad_norm": 2.2616798877716064, "learning_rate": 8.953036792363016e-07, "loss": 0.2427, "step": 38300 }, { "epoch": 2.2659253563612705, "grad_norm": 3.8150265216827393, "learning_rate": 8.939390107101251e-07, "loss": 0.2337, "step": 38310 }, { "epoch": 2.2665168273496183, "grad_norm": 2.480182409286499, "learning_rate": 8.925752009168916e-07, "loss": 0.2566, "step": 38320 }, { "epoch": 2.2671082983379667, "grad_norm": 6.42069149017334, "learning_rate": 8.912122504126838e-07, "loss": 0.2334, "step": 38330 }, { "epoch": 2.2676997693263146, "grad_norm": 3.4451873302459717, "learning_rate": 8.898501597532379e-07, "loss": 0.2098, "step": 38340 }, { "epoch": 2.2682912403146624, "grad_norm": 2.0235471725463867, "learning_rate": 8.884889294939363e-07, "loss": 0.2085, "step": 38350 }, { "epoch": 2.268882711303011, "grad_norm": 3.497314691543579, "learning_rate": 8.871285601898134e-07, "loss": 0.203, "step": 38360 }, { "epoch": 2.2694741822913587, "grad_norm": 5.919647216796875, "learning_rate": 8.857690523955503e-07, "loss": 0.2719, "step": 38370 }, { "epoch": 2.2700656532797066, "grad_norm": 4.199284076690674, "learning_rate": 8.844104066654765e-07, "loss": 0.2619, "step": 38380 }, { "epoch": 2.2706571242680544, "grad_norm": 3.2986741065979004, "learning_rate": 8.830526235535735e-07, "loss": 0.258, "step": 38390 }, { "epoch": 2.2712485952564028, "grad_norm": 2.1485307216644287, "learning_rate": 8.816957036134663e-07, "loss": 0.2302, "step": 38400 }, { "epoch": 2.2718400662447507, "grad_norm": 2.2774763107299805, "learning_rate": 8.803396473984323e-07, "loss": 0.2042, "step": 38410 }, { "epoch": 2.2724315372330985, "grad_norm": 6.0407209396362305, "learning_rate": 8.789844554613942e-07, "loss": 0.2997, "step": 38420 }, { "epoch": 2.273023008221447, "grad_norm": 5.766561031341553, "learning_rate": 8.776301283549229e-07, "loss": 0.3221, "step": 38430 }, { "epoch": 2.2736144792097948, "grad_norm": 2.358400344848633, "learning_rate": 8.76276666631236e-07, "loss": 0.2246, "step": 38440 }, { "epoch": 2.2742059501981426, "grad_norm": 5.337082862854004, "learning_rate": 8.749240708421983e-07, "loss": 0.2977, "step": 38450 }, { "epoch": 2.274797421186491, "grad_norm": 5.947932720184326, "learning_rate": 8.735723415393243e-07, "loss": 0.2577, "step": 38460 }, { "epoch": 2.275388892174839, "grad_norm": 3.8235297203063965, "learning_rate": 8.722214792737709e-07, "loss": 0.2443, "step": 38470 }, { "epoch": 2.2759803631631867, "grad_norm": 3.7083797454833984, "learning_rate": 8.708714845963454e-07, "loss": 0.229, "step": 38480 }, { "epoch": 2.276571834151535, "grad_norm": 3.818988561630249, "learning_rate": 8.695223580574975e-07, "loss": 0.2776, "step": 38490 }, { "epoch": 2.277163305139883, "grad_norm": 2.508059501647949, "learning_rate": 8.681741002073275e-07, "loss": 0.2051, "step": 38500 }, { "epoch": 2.277754776128231, "grad_norm": 3.804154634475708, "learning_rate": 8.668267115955769e-07, "loss": 0.2478, "step": 38510 }, { "epoch": 2.2783462471165787, "grad_norm": 3.2037599086761475, "learning_rate": 8.654801927716348e-07, "loss": 0.2699, "step": 38520 }, { "epoch": 2.278937718104927, "grad_norm": 3.0820019245147705, "learning_rate": 8.64134544284537e-07, "loss": 0.2683, "step": 38530 }, { "epoch": 2.279529189093275, "grad_norm": 4.113240718841553, "learning_rate": 8.627897666829619e-07, "loss": 0.2841, "step": 38540 }, { "epoch": 2.280120660081623, "grad_norm": 2.0944135189056396, "learning_rate": 8.614458605152347e-07, "loss": 0.2028, "step": 38550 }, { "epoch": 2.280712131069971, "grad_norm": 5.583799362182617, "learning_rate": 8.601028263293248e-07, "loss": 0.2399, "step": 38560 }, { "epoch": 2.281303602058319, "grad_norm": 4.506545543670654, "learning_rate": 8.587606646728445e-07, "loss": 0.2813, "step": 38570 }, { "epoch": 2.281895073046667, "grad_norm": 7.466597080230713, "learning_rate": 8.574193760930532e-07, "loss": 0.2583, "step": 38580 }, { "epoch": 2.2824865440350153, "grad_norm": 5.142943382263184, "learning_rate": 8.560789611368514e-07, "loss": 0.2288, "step": 38590 }, { "epoch": 2.283078015023363, "grad_norm": 5.004454135894775, "learning_rate": 8.547394203507861e-07, "loss": 0.2571, "step": 38600 }, { "epoch": 2.283669486011711, "grad_norm": 4.224496364593506, "learning_rate": 8.534007542810462e-07, "loss": 0.1922, "step": 38610 }, { "epoch": 2.2842609570000594, "grad_norm": 2.4224095344543457, "learning_rate": 8.520629634734628e-07, "loss": 0.2513, "step": 38620 }, { "epoch": 2.2848524279884073, "grad_norm": 4.971318244934082, "learning_rate": 8.507260484735138e-07, "loss": 0.2269, "step": 38630 }, { "epoch": 2.285443898976755, "grad_norm": 4.937352657318115, "learning_rate": 8.493900098263165e-07, "loss": 0.3218, "step": 38640 }, { "epoch": 2.286035369965103, "grad_norm": 4.819253921508789, "learning_rate": 8.480548480766323e-07, "loss": 0.2786, "step": 38650 }, { "epoch": 2.2866268409534514, "grad_norm": 2.360435724258423, "learning_rate": 8.467205637688641e-07, "loss": 0.2132, "step": 38660 }, { "epoch": 2.2872183119417993, "grad_norm": 2.7517735958099365, "learning_rate": 8.453871574470594e-07, "loss": 0.336, "step": 38670 }, { "epoch": 2.287809782930147, "grad_norm": 3.9534943103790283, "learning_rate": 8.440546296549044e-07, "loss": 0.2238, "step": 38680 }, { "epoch": 2.2884012539184955, "grad_norm": 6.889859199523926, "learning_rate": 8.427229809357309e-07, "loss": 0.2338, "step": 38690 }, { "epoch": 2.2889927249068434, "grad_norm": 4.11316442489624, "learning_rate": 8.413922118325087e-07, "loss": 0.2447, "step": 38700 }, { "epoch": 2.2895841958951912, "grad_norm": 3.853228807449341, "learning_rate": 8.400623228878499e-07, "loss": 0.2917, "step": 38710 }, { "epoch": 2.290175666883539, "grad_norm": 5.51610803604126, "learning_rate": 8.3873331464401e-07, "loss": 0.2671, "step": 38720 }, { "epoch": 2.2907671378718875, "grad_norm": 3.690131187438965, "learning_rate": 8.37405187642882e-07, "loss": 0.2662, "step": 38730 }, { "epoch": 2.2913586088602353, "grad_norm": 2.390136480331421, "learning_rate": 8.360779424260029e-07, "loss": 0.1965, "step": 38740 }, { "epoch": 2.2919500798485837, "grad_norm": 1.8128976821899414, "learning_rate": 8.347515795345475e-07, "loss": 0.2189, "step": 38750 }, { "epoch": 2.2925415508369316, "grad_norm": 3.323129653930664, "learning_rate": 8.334260995093306e-07, "loss": 0.2425, "step": 38760 }, { "epoch": 2.2931330218252794, "grad_norm": 4.397581100463867, "learning_rate": 8.321015028908108e-07, "loss": 0.2537, "step": 38770 }, { "epoch": 2.2937244928136273, "grad_norm": 3.543300151824951, "learning_rate": 8.307777902190814e-07, "loss": 0.2946, "step": 38780 }, { "epoch": 2.2943159638019757, "grad_norm": 2.7931854724884033, "learning_rate": 8.2945496203388e-07, "loss": 0.2415, "step": 38790 }, { "epoch": 2.2949074347903236, "grad_norm": 3.72454571723938, "learning_rate": 8.281330188745793e-07, "loss": 0.2791, "step": 38800 }, { "epoch": 2.2954989057786714, "grad_norm": 3.422755479812622, "learning_rate": 8.268119612801952e-07, "loss": 0.2294, "step": 38810 }, { "epoch": 2.2960903767670198, "grad_norm": 3.6425538063049316, "learning_rate": 8.254917897893797e-07, "loss": 0.3042, "step": 38820 }, { "epoch": 2.2966818477553677, "grad_norm": 3.2233664989471436, "learning_rate": 8.241725049404234e-07, "loss": 0.3429, "step": 38830 }, { "epoch": 2.2972733187437155, "grad_norm": 2.876453399658203, "learning_rate": 8.228541072712578e-07, "loss": 0.2009, "step": 38840 }, { "epoch": 2.2978647897320634, "grad_norm": 3.0825133323669434, "learning_rate": 8.215365973194501e-07, "loss": 0.2421, "step": 38850 }, { "epoch": 2.2984562607204118, "grad_norm": 6.340720176696777, "learning_rate": 8.202199756222073e-07, "loss": 0.2473, "step": 38860 }, { "epoch": 2.2990477317087596, "grad_norm": 3.206899404525757, "learning_rate": 8.18904242716372e-07, "loss": 0.2504, "step": 38870 }, { "epoch": 2.2996392026971075, "grad_norm": 3.249828577041626, "learning_rate": 8.17589399138428e-07, "loss": 0.2929, "step": 38880 }, { "epoch": 2.300230673685456, "grad_norm": 4.320089817047119, "learning_rate": 8.162754454244932e-07, "loss": 0.2742, "step": 38890 }, { "epoch": 2.3008221446738037, "grad_norm": 3.9549453258514404, "learning_rate": 8.149623821103232e-07, "loss": 0.2701, "step": 38900 }, { "epoch": 2.3014136156621516, "grad_norm": 3.776624917984009, "learning_rate": 8.13650209731313e-07, "loss": 0.2187, "step": 38910 }, { "epoch": 2.3020050866505, "grad_norm": 4.320415019989014, "learning_rate": 8.123389288224907e-07, "loss": 0.2535, "step": 38920 }, { "epoch": 2.302596557638848, "grad_norm": 5.307142734527588, "learning_rate": 8.110285399185245e-07, "loss": 0.2244, "step": 38930 }, { "epoch": 2.3031880286271957, "grad_norm": 3.118116855621338, "learning_rate": 8.097190435537151e-07, "loss": 0.2133, "step": 38940 }, { "epoch": 2.303779499615544, "grad_norm": 7.813197135925293, "learning_rate": 8.084104402620035e-07, "loss": 0.2282, "step": 38950 }, { "epoch": 2.304370970603892, "grad_norm": 3.4657514095306396, "learning_rate": 8.07102730576963e-07, "loss": 0.2583, "step": 38960 }, { "epoch": 2.30496244159224, "grad_norm": 2.9814751148223877, "learning_rate": 8.057959150318035e-07, "loss": 0.3189, "step": 38970 }, { "epoch": 2.3055539125805877, "grad_norm": 3.096785068511963, "learning_rate": 8.04489994159372e-07, "loss": 0.2671, "step": 38980 }, { "epoch": 2.306145383568936, "grad_norm": 3.8899760246276855, "learning_rate": 8.03184968492148e-07, "loss": 0.2459, "step": 38990 }, { "epoch": 2.306736854557284, "grad_norm": 3.5818417072296143, "learning_rate": 8.018808385622488e-07, "loss": 0.2809, "step": 39000 }, { "epoch": 2.307328325545632, "grad_norm": 2.9815616607666016, "learning_rate": 8.005776049014248e-07, "loss": 0.2992, "step": 39010 }, { "epoch": 2.30791979653398, "grad_norm": 4.152881145477295, "learning_rate": 7.992752680410598e-07, "loss": 0.2753, "step": 39020 }, { "epoch": 2.308511267522328, "grad_norm": 2.775160789489746, "learning_rate": 7.979738285121752e-07, "loss": 0.2201, "step": 39030 }, { "epoch": 2.309102738510676, "grad_norm": 3.799024820327759, "learning_rate": 7.966732868454229e-07, "loss": 0.2346, "step": 39040 }, { "epoch": 2.3096942094990243, "grad_norm": 5.729594707489014, "learning_rate": 7.953736435710923e-07, "loss": 0.3426, "step": 39050 }, { "epoch": 2.310285680487372, "grad_norm": 4.153153896331787, "learning_rate": 7.940748992191041e-07, "loss": 0.2911, "step": 39060 }, { "epoch": 2.31087715147572, "grad_norm": 5.8571577072143555, "learning_rate": 7.927770543190123e-07, "loss": 0.2815, "step": 39070 }, { "epoch": 2.3114686224640684, "grad_norm": 3.839905023574829, "learning_rate": 7.914801094000057e-07, "loss": 0.2472, "step": 39080 }, { "epoch": 2.3120600934524163, "grad_norm": 2.1390321254730225, "learning_rate": 7.901840649909037e-07, "loss": 0.2073, "step": 39090 }, { "epoch": 2.312651564440764, "grad_norm": 3.1226375102996826, "learning_rate": 7.888889216201623e-07, "loss": 0.2737, "step": 39100 }, { "epoch": 2.313243035429112, "grad_norm": 3.3796281814575195, "learning_rate": 7.875946798158663e-07, "loss": 0.2802, "step": 39110 }, { "epoch": 2.3138345064174604, "grad_norm": 5.701668739318848, "learning_rate": 7.863013401057361e-07, "loss": 0.3041, "step": 39120 }, { "epoch": 2.3144259774058082, "grad_norm": 3.979395627975464, "learning_rate": 7.85008903017121e-07, "loss": 0.2087, "step": 39130 }, { "epoch": 2.315017448394156, "grad_norm": 4.239561080932617, "learning_rate": 7.837173690770057e-07, "loss": 0.2798, "step": 39140 }, { "epoch": 2.3156089193825045, "grad_norm": 4.389551639556885, "learning_rate": 7.82426738812004e-07, "loss": 0.2692, "step": 39150 }, { "epoch": 2.3162003903708523, "grad_norm": 2.962977647781372, "learning_rate": 7.811370127483615e-07, "loss": 0.2797, "step": 39160 }, { "epoch": 2.3167918613592002, "grad_norm": 6.393438816070557, "learning_rate": 7.798481914119574e-07, "loss": 0.2519, "step": 39170 }, { "epoch": 2.317383332347548, "grad_norm": 2.4234910011291504, "learning_rate": 7.785602753282988e-07, "loss": 0.2633, "step": 39180 }, { "epoch": 2.3179748033358964, "grad_norm": 5.964151859283447, "learning_rate": 7.772732650225265e-07, "loss": 0.2705, "step": 39190 }, { "epoch": 2.3185662743242443, "grad_norm": 2.3044683933258057, "learning_rate": 7.759871610194105e-07, "loss": 0.2054, "step": 39200 }, { "epoch": 2.3191577453125927, "grad_norm": 3.1530232429504395, "learning_rate": 7.747019638433506e-07, "loss": 0.2188, "step": 39210 }, { "epoch": 2.3197492163009406, "grad_norm": 7.681234359741211, "learning_rate": 7.73417674018379e-07, "loss": 0.2919, "step": 39220 }, { "epoch": 2.3203406872892884, "grad_norm": 3.1806180477142334, "learning_rate": 7.721342920681553e-07, "loss": 0.2983, "step": 39230 }, { "epoch": 2.3209321582776363, "grad_norm": 3.449035406112671, "learning_rate": 7.708518185159718e-07, "loss": 0.2884, "step": 39240 }, { "epoch": 2.3215236292659847, "grad_norm": 5.162156581878662, "learning_rate": 7.695702538847486e-07, "loss": 0.2755, "step": 39250 }, { "epoch": 2.3221151002543325, "grad_norm": 2.643190383911133, "learning_rate": 7.682895986970338e-07, "loss": 0.2343, "step": 39260 }, { "epoch": 2.3227065712426804, "grad_norm": 4.150285720825195, "learning_rate": 7.670098534750088e-07, "loss": 0.314, "step": 39270 }, { "epoch": 2.3232980422310288, "grad_norm": 3.6082494258880615, "learning_rate": 7.657310187404804e-07, "loss": 0.2375, "step": 39280 }, { "epoch": 2.3238895132193766, "grad_norm": 3.6247074604034424, "learning_rate": 7.644530950148855e-07, "loss": 0.2647, "step": 39290 }, { "epoch": 2.3244809842077245, "grad_norm": 4.550276279449463, "learning_rate": 7.631760828192883e-07, "loss": 0.2767, "step": 39300 }, { "epoch": 2.3250724551960724, "grad_norm": 2.825698137283325, "learning_rate": 7.618999826743844e-07, "loss": 0.2433, "step": 39310 }, { "epoch": 2.3256639261844207, "grad_norm": 4.146150588989258, "learning_rate": 7.606247951004933e-07, "loss": 0.2579, "step": 39320 }, { "epoch": 2.3262553971727686, "grad_norm": 2.3185057640075684, "learning_rate": 7.593505206175667e-07, "loss": 0.2913, "step": 39330 }, { "epoch": 2.3268468681611165, "grad_norm": 2.428525686264038, "learning_rate": 7.58077159745181e-07, "loss": 0.2884, "step": 39340 }, { "epoch": 2.327438339149465, "grad_norm": 4.735589504241943, "learning_rate": 7.568047130025399e-07, "loss": 0.3055, "step": 39350 }, { "epoch": 2.3280298101378127, "grad_norm": 2.6237077713012695, "learning_rate": 7.555331809084772e-07, "loss": 0.2157, "step": 39360 }, { "epoch": 2.3286212811261606, "grad_norm": 3.4673194885253906, "learning_rate": 7.542625639814505e-07, "loss": 0.2516, "step": 39370 }, { "epoch": 2.329212752114509, "grad_norm": 2.908780574798584, "learning_rate": 7.529928627395474e-07, "loss": 0.2746, "step": 39380 }, { "epoch": 2.329804223102857, "grad_norm": 3.916877508163452, "learning_rate": 7.517240777004795e-07, "loss": 0.263, "step": 39390 }, { "epoch": 2.3303956940912047, "grad_norm": 2.427394390106201, "learning_rate": 7.504562093815852e-07, "loss": 0.2492, "step": 39400 }, { "epoch": 2.330987165079553, "grad_norm": 3.6045961380004883, "learning_rate": 7.491892582998311e-07, "loss": 0.2142, "step": 39410 }, { "epoch": 2.331578636067901, "grad_norm": 10.539131164550781, "learning_rate": 7.47923224971807e-07, "loss": 0.2426, "step": 39420 }, { "epoch": 2.332170107056249, "grad_norm": 5.162057876586914, "learning_rate": 7.466581099137312e-07, "loss": 0.3249, "step": 39430 }, { "epoch": 2.3327615780445967, "grad_norm": 2.9747581481933594, "learning_rate": 7.453939136414454e-07, "loss": 0.2166, "step": 39440 }, { "epoch": 2.333353049032945, "grad_norm": 2.6437861919403076, "learning_rate": 7.441306366704186e-07, "loss": 0.3145, "step": 39450 }, { "epoch": 2.333944520021293, "grad_norm": 3.800621747970581, "learning_rate": 7.428682795157433e-07, "loss": 0.2764, "step": 39460 }, { "epoch": 2.334535991009641, "grad_norm": 2.5157716274261475, "learning_rate": 7.416068426921365e-07, "loss": 0.3677, "step": 39470 }, { "epoch": 2.335127461997989, "grad_norm": 2.8500843048095703, "learning_rate": 7.40346326713943e-07, "loss": 0.2906, "step": 39480 }, { "epoch": 2.335718932986337, "grad_norm": 2.205350875854492, "learning_rate": 7.390867320951285e-07, "loss": 0.269, "step": 39490 }, { "epoch": 2.336310403974685, "grad_norm": 3.5816617012023926, "learning_rate": 7.378280593492871e-07, "loss": 0.2418, "step": 39500 }, { "epoch": 2.3369018749630333, "grad_norm": 3.9407732486724854, "learning_rate": 7.365703089896312e-07, "loss": 0.2413, "step": 39510 }, { "epoch": 2.337493345951381, "grad_norm": 4.268648624420166, "learning_rate": 7.353134815290028e-07, "loss": 0.2308, "step": 39520 }, { "epoch": 2.338084816939729, "grad_norm": 1.8810911178588867, "learning_rate": 7.34057577479865e-07, "loss": 0.2485, "step": 39530 }, { "epoch": 2.3386762879280774, "grad_norm": 2.711543321609497, "learning_rate": 7.328025973543032e-07, "loss": 0.2148, "step": 39540 }, { "epoch": 2.3392677589164252, "grad_norm": 3.518251657485962, "learning_rate": 7.315485416640297e-07, "loss": 0.261, "step": 39550 }, { "epoch": 2.339859229904773, "grad_norm": 4.703413963317871, "learning_rate": 7.302954109203759e-07, "loss": 0.2911, "step": 39560 }, { "epoch": 2.340450700893121, "grad_norm": 2.733342170715332, "learning_rate": 7.290432056342993e-07, "loss": 0.3019, "step": 39570 }, { "epoch": 2.3410421718814693, "grad_norm": 2.5200600624084473, "learning_rate": 7.277919263163772e-07, "loss": 0.2634, "step": 39580 }, { "epoch": 2.3416336428698172, "grad_norm": 5.3481059074401855, "learning_rate": 7.265415734768124e-07, "loss": 0.2482, "step": 39590 }, { "epoch": 2.342225113858165, "grad_norm": 2.9391236305236816, "learning_rate": 7.252921476254275e-07, "loss": 0.2247, "step": 39600 }, { "epoch": 2.3428165848465135, "grad_norm": 3.1728515625, "learning_rate": 7.24043649271667e-07, "loss": 0.2383, "step": 39610 }, { "epoch": 2.3434080558348613, "grad_norm": 4.22908878326416, "learning_rate": 7.227960789245999e-07, "loss": 0.2594, "step": 39620 }, { "epoch": 2.3439995268232092, "grad_norm": 5.961239337921143, "learning_rate": 7.215494370929136e-07, "loss": 0.2839, "step": 39630 }, { "epoch": 2.344590997811557, "grad_norm": 3.2716872692108154, "learning_rate": 7.203037242849194e-07, "loss": 0.2458, "step": 39640 }, { "epoch": 2.3451824687999054, "grad_norm": 3.826338529586792, "learning_rate": 7.190589410085486e-07, "loss": 0.2961, "step": 39650 }, { "epoch": 2.3457739397882533, "grad_norm": 5.076517581939697, "learning_rate": 7.178150877713522e-07, "loss": 0.2305, "step": 39660 }, { "epoch": 2.3463654107766017, "grad_norm": 2.475449800491333, "learning_rate": 7.165721650805054e-07, "loss": 0.2456, "step": 39670 }, { "epoch": 2.3469568817649495, "grad_norm": 3.7047476768493652, "learning_rate": 7.153301734428005e-07, "loss": 0.2198, "step": 39680 }, { "epoch": 2.3475483527532974, "grad_norm": 1.914905071258545, "learning_rate": 7.14089113364653e-07, "loss": 0.1859, "step": 39690 }, { "epoch": 2.3481398237416453, "grad_norm": 1.8115465641021729, "learning_rate": 7.128489853520968e-07, "loss": 0.1965, "step": 39700 }, { "epoch": 2.3487312947299936, "grad_norm": 3.7795777320861816, "learning_rate": 7.11609789910786e-07, "loss": 0.2331, "step": 39710 }, { "epoch": 2.3493227657183415, "grad_norm": 3.562927484512329, "learning_rate": 7.103715275459936e-07, "loss": 0.2356, "step": 39720 }, { "epoch": 2.3499142367066894, "grad_norm": 3.4521522521972656, "learning_rate": 7.09134198762615e-07, "loss": 0.27, "step": 39730 }, { "epoch": 2.3505057076950377, "grad_norm": 4.644891262054443, "learning_rate": 7.078978040651627e-07, "loss": 0.3042, "step": 39740 }, { "epoch": 2.3510971786833856, "grad_norm": 3.276338577270508, "learning_rate": 7.066623439577671e-07, "loss": 0.2739, "step": 39750 }, { "epoch": 2.3516886496717335, "grad_norm": 3.243729829788208, "learning_rate": 7.054278189441815e-07, "loss": 0.2385, "step": 39760 }, { "epoch": 2.3522801206600814, "grad_norm": 3.908015727996826, "learning_rate": 7.041942295277737e-07, "loss": 0.1916, "step": 39770 }, { "epoch": 2.3528715916484297, "grad_norm": 4.218501567840576, "learning_rate": 7.029615762115336e-07, "loss": 0.3146, "step": 39780 }, { "epoch": 2.3534630626367776, "grad_norm": 2.311314344406128, "learning_rate": 7.017298594980668e-07, "loss": 0.3052, "step": 39790 }, { "epoch": 2.3540545336251255, "grad_norm": 5.309919357299805, "learning_rate": 7.004990798895973e-07, "loss": 0.2361, "step": 39800 }, { "epoch": 2.354646004613474, "grad_norm": 2.7411253452301025, "learning_rate": 6.992692378879692e-07, "loss": 0.1811, "step": 39810 }, { "epoch": 2.3552374756018217, "grad_norm": 3.332235336303711, "learning_rate": 6.980403339946409e-07, "loss": 0.2735, "step": 39820 }, { "epoch": 2.3558289465901696, "grad_norm": 2.256992816925049, "learning_rate": 6.968123687106925e-07, "loss": 0.2355, "step": 39830 }, { "epoch": 2.356420417578518, "grad_norm": 4.955102443695068, "learning_rate": 6.955853425368171e-07, "loss": 0.3416, "step": 39840 }, { "epoch": 2.357011888566866, "grad_norm": 5.571427345275879, "learning_rate": 6.943592559733269e-07, "loss": 0.2738, "step": 39850 }, { "epoch": 2.3576033595552137, "grad_norm": 3.034194231033325, "learning_rate": 6.931341095201523e-07, "loss": 0.2354, "step": 39860 }, { "epoch": 2.358194830543562, "grad_norm": 4.714970111846924, "learning_rate": 6.919099036768367e-07, "loss": 0.2734, "step": 39870 }, { "epoch": 2.35878630153191, "grad_norm": 3.382848024368286, "learning_rate": 6.906866389425448e-07, "loss": 0.3031, "step": 39880 }, { "epoch": 2.359377772520258, "grad_norm": 3.170109510421753, "learning_rate": 6.894643158160526e-07, "loss": 0.2375, "step": 39890 }, { "epoch": 2.3599692435086057, "grad_norm": 2.729538679122925, "learning_rate": 6.882429347957569e-07, "loss": 0.2511, "step": 39900 }, { "epoch": 2.360560714496954, "grad_norm": 2.258570432662964, "learning_rate": 6.870224963796664e-07, "loss": 0.2675, "step": 39910 }, { "epoch": 2.361152185485302, "grad_norm": 4.0116190910339355, "learning_rate": 6.85803001065407e-07, "loss": 0.2688, "step": 39920 }, { "epoch": 2.36174365647365, "grad_norm": 4.837122440338135, "learning_rate": 6.84584449350222e-07, "loss": 0.3358, "step": 39930 }, { "epoch": 2.362335127461998, "grad_norm": 3.6081879138946533, "learning_rate": 6.833668417309652e-07, "loss": 0.2545, "step": 39940 }, { "epoch": 2.362926598450346, "grad_norm": 3.6264195442199707, "learning_rate": 6.821501787041107e-07, "loss": 0.238, "step": 39950 }, { "epoch": 2.363518069438694, "grad_norm": 4.636666297912598, "learning_rate": 6.809344607657436e-07, "loss": 0.2732, "step": 39960 }, { "epoch": 2.3641095404270422, "grad_norm": 2.920416831970215, "learning_rate": 6.797196884115665e-07, "loss": 0.2857, "step": 39970 }, { "epoch": 2.36470101141539, "grad_norm": 3.612913131713867, "learning_rate": 6.785058621368944e-07, "loss": 0.2433, "step": 39980 }, { "epoch": 2.365292482403738, "grad_norm": 2.8355019092559814, "learning_rate": 6.772929824366563e-07, "loss": 0.2725, "step": 39990 }, { "epoch": 2.3658839533920863, "grad_norm": 3.132934331893921, "learning_rate": 6.760810498053979e-07, "loss": 0.2013, "step": 40000 }, { "epoch": 2.3664754243804342, "grad_norm": 2.1254963874816895, "learning_rate": 6.748700647372756e-07, "loss": 0.2518, "step": 40010 }, { "epoch": 2.367066895368782, "grad_norm": 1.9041781425476074, "learning_rate": 6.73660027726062e-07, "loss": 0.2163, "step": 40020 }, { "epoch": 2.36765836635713, "grad_norm": 3.474200487136841, "learning_rate": 6.724509392651409e-07, "loss": 0.24, "step": 40030 }, { "epoch": 2.3682498373454783, "grad_norm": 5.27243185043335, "learning_rate": 6.712427998475115e-07, "loss": 0.1904, "step": 40040 }, { "epoch": 2.3688413083338262, "grad_norm": 5.52693510055542, "learning_rate": 6.700356099657843e-07, "loss": 0.2203, "step": 40050 }, { "epoch": 2.369432779322174, "grad_norm": 7.930360794067383, "learning_rate": 6.688293701121826e-07, "loss": 0.2557, "step": 40060 }, { "epoch": 2.3700242503105224, "grad_norm": 2.254854440689087, "learning_rate": 6.676240807785447e-07, "loss": 0.2489, "step": 40070 }, { "epoch": 2.3706157212988703, "grad_norm": 5.242307186126709, "learning_rate": 6.664197424563176e-07, "loss": 0.2892, "step": 40080 }, { "epoch": 2.371207192287218, "grad_norm": 2.8623275756835938, "learning_rate": 6.652163556365648e-07, "loss": 0.278, "step": 40090 }, { "epoch": 2.371798663275566, "grad_norm": 3.6795027256011963, "learning_rate": 6.640139208099582e-07, "loss": 0.3088, "step": 40100 }, { "epoch": 2.3723901342639144, "grad_norm": 4.104439735412598, "learning_rate": 6.628124384667825e-07, "loss": 0.2681, "step": 40110 }, { "epoch": 2.3729816052522623, "grad_norm": 3.889435291290283, "learning_rate": 6.61611909096936e-07, "loss": 0.2722, "step": 40120 }, { "epoch": 2.3735730762406106, "grad_norm": 5.229700088500977, "learning_rate": 6.604123331899252e-07, "loss": 0.2555, "step": 40130 }, { "epoch": 2.3741645472289585, "grad_norm": 3.1272168159484863, "learning_rate": 6.592137112348721e-07, "loss": 0.2188, "step": 40140 }, { "epoch": 2.3747560182173064, "grad_norm": 3.0629892349243164, "learning_rate": 6.58016043720504e-07, "loss": 0.2848, "step": 40150 }, { "epoch": 2.3753474892056543, "grad_norm": 6.229086875915527, "learning_rate": 6.568193311351643e-07, "loss": 0.2299, "step": 40160 }, { "epoch": 2.3759389601940026, "grad_norm": 4.195915699005127, "learning_rate": 6.556235739668043e-07, "loss": 0.3369, "step": 40170 }, { "epoch": 2.3765304311823505, "grad_norm": 3.269291877746582, "learning_rate": 6.544287727029854e-07, "loss": 0.2225, "step": 40180 }, { "epoch": 2.3771219021706984, "grad_norm": 6.792934894561768, "learning_rate": 6.532349278308823e-07, "loss": 0.2349, "step": 40190 }, { "epoch": 2.3777133731590467, "grad_norm": 3.3215179443359375, "learning_rate": 6.520420398372751e-07, "loss": 0.2746, "step": 40200 }, { "epoch": 2.3783048441473946, "grad_norm": 4.489322662353516, "learning_rate": 6.508501092085588e-07, "loss": 0.2583, "step": 40210 }, { "epoch": 2.3788963151357425, "grad_norm": 2.4938642978668213, "learning_rate": 6.496591364307337e-07, "loss": 0.2734, "step": 40220 }, { "epoch": 2.3794877861240904, "grad_norm": 2.8039908409118652, "learning_rate": 6.484691219894124e-07, "loss": 0.2053, "step": 40230 }, { "epoch": 2.3800792571124387, "grad_norm": 3.1838552951812744, "learning_rate": 6.472800663698159e-07, "loss": 0.3032, "step": 40240 }, { "epoch": 2.3806707281007866, "grad_norm": 3.523519992828369, "learning_rate": 6.460919700567721e-07, "loss": 0.2225, "step": 40250 }, { "epoch": 2.3812621990891345, "grad_norm": 2.9932126998901367, "learning_rate": 6.449048335347221e-07, "loss": 0.2379, "step": 40260 }, { "epoch": 2.381853670077483, "grad_norm": 2.7773289680480957, "learning_rate": 6.437186572877115e-07, "loss": 0.1996, "step": 40270 }, { "epoch": 2.3824451410658307, "grad_norm": 3.75354266166687, "learning_rate": 6.425334417993976e-07, "loss": 0.3218, "step": 40280 }, { "epoch": 2.3830366120541786, "grad_norm": 3.6789486408233643, "learning_rate": 6.413491875530436e-07, "loss": 0.2933, "step": 40290 }, { "epoch": 2.383628083042527, "grad_norm": 3.4139766693115234, "learning_rate": 6.401658950315211e-07, "loss": 0.2415, "step": 40300 }, { "epoch": 2.384219554030875, "grad_norm": 3.0244030952453613, "learning_rate": 6.389835647173115e-07, "loss": 0.246, "step": 40310 }, { "epoch": 2.3848110250192227, "grad_norm": 3.73664927482605, "learning_rate": 6.378021970925006e-07, "loss": 0.2603, "step": 40320 }, { "epoch": 2.385402496007571, "grad_norm": 4.402113914489746, "learning_rate": 6.366217926387854e-07, "loss": 0.254, "step": 40330 }, { "epoch": 2.385993966995919, "grad_norm": 4.778438568115234, "learning_rate": 6.354423518374666e-07, "loss": 0.2256, "step": 40340 }, { "epoch": 2.386585437984267, "grad_norm": 2.851933479309082, "learning_rate": 6.342638751694562e-07, "loss": 0.2702, "step": 40350 }, { "epoch": 2.3871769089726147, "grad_norm": 4.719964504241943, "learning_rate": 6.330863631152671e-07, "loss": 0.2146, "step": 40360 }, { "epoch": 2.387768379960963, "grad_norm": 6.916764259338379, "learning_rate": 6.319098161550252e-07, "loss": 0.2526, "step": 40370 }, { "epoch": 2.388359850949311, "grad_norm": 5.903205871582031, "learning_rate": 6.307342347684586e-07, "loss": 0.3156, "step": 40380 }, { "epoch": 2.388951321937659, "grad_norm": 2.961794376373291, "learning_rate": 6.29559619434903e-07, "loss": 0.2214, "step": 40390 }, { "epoch": 2.389542792926007, "grad_norm": 3.112874746322632, "learning_rate": 6.28385970633301e-07, "loss": 0.2216, "step": 40400 }, { "epoch": 2.390134263914355, "grad_norm": 4.942294120788574, "learning_rate": 6.272132888421998e-07, "loss": 0.2721, "step": 40410 }, { "epoch": 2.390725734902703, "grad_norm": 4.022857666015625, "learning_rate": 6.260415745397539e-07, "loss": 0.2137, "step": 40420 }, { "epoch": 2.3913172058910512, "grad_norm": 7.746889114379883, "learning_rate": 6.248708282037219e-07, "loss": 0.2428, "step": 40430 }, { "epoch": 2.391908676879399, "grad_norm": 4.2930378913879395, "learning_rate": 6.237010503114671e-07, "loss": 0.2652, "step": 40440 }, { "epoch": 2.392500147867747, "grad_norm": 1.752265453338623, "learning_rate": 6.225322413399605e-07, "loss": 0.244, "step": 40450 }, { "epoch": 2.3930916188560953, "grad_norm": 5.610169887542725, "learning_rate": 6.213644017657751e-07, "loss": 0.2862, "step": 40460 }, { "epoch": 2.3936830898444432, "grad_norm": 1.6037452220916748, "learning_rate": 6.201975320650916e-07, "loss": 0.2826, "step": 40470 }, { "epoch": 2.394274560832791, "grad_norm": 2.900513172149658, "learning_rate": 6.190316327136929e-07, "loss": 0.2326, "step": 40480 }, { "epoch": 2.394866031821139, "grad_norm": 3.0545849800109863, "learning_rate": 6.178667041869662e-07, "loss": 0.2297, "step": 40490 }, { "epoch": 2.3954575028094873, "grad_norm": 3.8431460857391357, "learning_rate": 6.167027469599056e-07, "loss": 0.2835, "step": 40500 }, { "epoch": 2.396048973797835, "grad_norm": 4.873370170593262, "learning_rate": 6.15539761507105e-07, "loss": 0.2266, "step": 40510 }, { "epoch": 2.396640444786183, "grad_norm": 3.765383005142212, "learning_rate": 6.143777483027662e-07, "loss": 0.3223, "step": 40520 }, { "epoch": 2.3972319157745314, "grad_norm": 5.3396992683410645, "learning_rate": 6.132167078206914e-07, "loss": 0.2904, "step": 40530 }, { "epoch": 2.3978233867628793, "grad_norm": 1.9037436246871948, "learning_rate": 6.120566405342883e-07, "loss": 0.2153, "step": 40540 }, { "epoch": 2.398414857751227, "grad_norm": 3.404733419418335, "learning_rate": 6.10897546916567e-07, "loss": 0.2407, "step": 40550 }, { "epoch": 2.399006328739575, "grad_norm": 3.889543294906616, "learning_rate": 6.097394274401392e-07, "loss": 0.2779, "step": 40560 }, { "epoch": 2.3995977997279234, "grad_norm": 3.508162021636963, "learning_rate": 6.085822825772234e-07, "loss": 0.2822, "step": 40570 }, { "epoch": 2.4001892707162713, "grad_norm": 3.6957945823669434, "learning_rate": 6.074261127996346e-07, "loss": 0.3003, "step": 40580 }, { "epoch": 2.4004258591116105, "eval_accuracy": 0.8639584788901088, "eval_loss": 0.3264763653278351, "eval_runtime": 158.1007, "eval_safe_aucpr": 0.9238952245214203, "eval_safe_f1": 0.8484320557491289, "eval_safe_fpr": 0.13165525694299085, "eval_safe_precision": 0.8386399443080643, "eval_safe_recall": 0.858455537636425, "eval_samples_per_second": 380.226, "eval_steps_per_second": 47.533, "eval_unsafe_aucpr": 0.9567608679603048, "eval_unsafe_f1": 0.8765994688556253, "eval_unsafe_fpr": 0.14154446236357446, "eval_unsafe_precision": 0.8850126443435605, "eval_unsafe_recall": 0.8683447430570087, "step": 40584 }, { "epoch": 2.4007807417046196, "grad_norm": 5.855945587158203, "learning_rate": 6.06270918578796e-07, "loss": 0.2719, "step": 40590 }, { "epoch": 2.4013722126929675, "grad_norm": 3.7241933345794678, "learning_rate": 6.051167003857296e-07, "loss": 0.3103, "step": 40600 }, { "epoch": 2.4019636836813154, "grad_norm": 2.1036441326141357, "learning_rate": 6.039634586910617e-07, "loss": 0.23, "step": 40610 }, { "epoch": 2.4025551546696633, "grad_norm": 4.457276821136475, "learning_rate": 6.028111939650182e-07, "loss": 0.2722, "step": 40620 }, { "epoch": 2.4031466256580116, "grad_norm": 3.469268560409546, "learning_rate": 6.016599066774274e-07, "loss": 0.2389, "step": 40630 }, { "epoch": 2.4037380966463595, "grad_norm": 5.841517448425293, "learning_rate": 6.005095972977207e-07, "loss": 0.3033, "step": 40640 }, { "epoch": 2.4043295676347074, "grad_norm": 1.5159083604812622, "learning_rate": 5.993602662949281e-07, "loss": 0.2582, "step": 40650 }, { "epoch": 2.4049210386230557, "grad_norm": 4.089616775512695, "learning_rate": 5.982119141376832e-07, "loss": 0.2603, "step": 40660 }, { "epoch": 2.4055125096114036, "grad_norm": 3.204653739929199, "learning_rate": 5.970645412942176e-07, "loss": 0.2936, "step": 40670 }, { "epoch": 2.4061039805997515, "grad_norm": 4.318663597106934, "learning_rate": 5.959181482323676e-07, "loss": 0.236, "step": 40680 }, { "epoch": 2.4066954515880994, "grad_norm": 3.168856143951416, "learning_rate": 5.947727354195658e-07, "loss": 0.2322, "step": 40690 }, { "epoch": 2.4072869225764477, "grad_norm": 4.362775802612305, "learning_rate": 5.93628303322847e-07, "loss": 0.2096, "step": 40700 }, { "epoch": 2.4078783935647956, "grad_norm": 4.025397300720215, "learning_rate": 5.924848524088476e-07, "loss": 0.2865, "step": 40710 }, { "epoch": 2.4084698645531435, "grad_norm": 2.2874057292938232, "learning_rate": 5.913423831438004e-07, "loss": 0.2268, "step": 40720 }, { "epoch": 2.409061335541492, "grad_norm": 4.214907169342041, "learning_rate": 5.902008959935416e-07, "loss": 0.2158, "step": 40730 }, { "epoch": 2.4096528065298397, "grad_norm": 4.386917591094971, "learning_rate": 5.890603914235051e-07, "loss": 0.2716, "step": 40740 }, { "epoch": 2.4102442775181876, "grad_norm": 5.048923969268799, "learning_rate": 5.87920869898723e-07, "loss": 0.3183, "step": 40750 }, { "epoch": 2.410835748506536, "grad_norm": 4.779331684112549, "learning_rate": 5.867823318838299e-07, "loss": 0.2762, "step": 40760 }, { "epoch": 2.411427219494884, "grad_norm": 5.07138204574585, "learning_rate": 5.856447778430556e-07, "loss": 0.2476, "step": 40770 }, { "epoch": 2.4120186904832317, "grad_norm": 8.250643730163574, "learning_rate": 5.845082082402327e-07, "loss": 0.2741, "step": 40780 }, { "epoch": 2.41261016147158, "grad_norm": 2.593803882598877, "learning_rate": 5.833726235387874e-07, "loss": 0.2328, "step": 40790 }, { "epoch": 2.413201632459928, "grad_norm": 5.172274589538574, "learning_rate": 5.822380242017495e-07, "loss": 0.2849, "step": 40800 }, { "epoch": 2.413793103448276, "grad_norm": 3.315138339996338, "learning_rate": 5.811044106917425e-07, "loss": 0.2795, "step": 40810 }, { "epoch": 2.4143845744366237, "grad_norm": 2.635082960128784, "learning_rate": 5.799717834709924e-07, "loss": 0.2437, "step": 40820 }, { "epoch": 2.414976045424972, "grad_norm": 5.6733078956604, "learning_rate": 5.788401430013196e-07, "loss": 0.2667, "step": 40830 }, { "epoch": 2.41556751641332, "grad_norm": 3.1584837436676025, "learning_rate": 5.777094897441425e-07, "loss": 0.29, "step": 40840 }, { "epoch": 2.416158987401668, "grad_norm": 2.314190626144409, "learning_rate": 5.765798241604794e-07, "loss": 0.3014, "step": 40850 }, { "epoch": 2.416750458390016, "grad_norm": 3.7700846195220947, "learning_rate": 5.754511467109424e-07, "loss": 0.2662, "step": 40860 }, { "epoch": 2.417341929378364, "grad_norm": 3.8314294815063477, "learning_rate": 5.743234578557446e-07, "loss": 0.308, "step": 40870 }, { "epoch": 2.417933400366712, "grad_norm": 3.930920362472534, "learning_rate": 5.73196758054693e-07, "loss": 0.2176, "step": 40880 }, { "epoch": 2.4185248713550602, "grad_norm": 2.581385850906372, "learning_rate": 5.720710477671911e-07, "loss": 0.2087, "step": 40890 }, { "epoch": 2.419116342343408, "grad_norm": 5.201752185821533, "learning_rate": 5.709463274522421e-07, "loss": 0.2435, "step": 40900 }, { "epoch": 2.419707813331756, "grad_norm": 1.6197470426559448, "learning_rate": 5.698225975684416e-07, "loss": 0.2072, "step": 40910 }, { "epoch": 2.4202992843201043, "grad_norm": 3.9352316856384277, "learning_rate": 5.686998585739851e-07, "loss": 0.2239, "step": 40920 }, { "epoch": 2.420890755308452, "grad_norm": 3.010437488555908, "learning_rate": 5.675781109266615e-07, "loss": 0.2689, "step": 40930 }, { "epoch": 2.4214822262968, "grad_norm": 3.3179516792297363, "learning_rate": 5.664573550838555e-07, "loss": 0.2481, "step": 40940 }, { "epoch": 2.422073697285148, "grad_norm": 1.694847822189331, "learning_rate": 5.653375915025495e-07, "loss": 0.2241, "step": 40950 }, { "epoch": 2.4226651682734963, "grad_norm": 3.6984310150146484, "learning_rate": 5.642188206393189e-07, "loss": 0.2568, "step": 40960 }, { "epoch": 2.423256639261844, "grad_norm": 4.049103736877441, "learning_rate": 5.631010429503361e-07, "loss": 0.2465, "step": 40970 }, { "epoch": 2.423848110250192, "grad_norm": 3.3819093704223633, "learning_rate": 5.619842588913668e-07, "loss": 0.2717, "step": 40980 }, { "epoch": 2.4244395812385404, "grad_norm": 6.756174087524414, "learning_rate": 5.608684689177744e-07, "loss": 0.3019, "step": 40990 }, { "epoch": 2.4250310522268883, "grad_norm": 2.8454997539520264, "learning_rate": 5.597536734845139e-07, "loss": 0.2265, "step": 41000 }, { "epoch": 2.425622523215236, "grad_norm": 5.395732402801514, "learning_rate": 5.586398730461359e-07, "loss": 0.2457, "step": 41010 }, { "epoch": 2.426213994203584, "grad_norm": 3.4455909729003906, "learning_rate": 5.575270680567859e-07, "loss": 0.2421, "step": 41020 }, { "epoch": 2.4268054651919324, "grad_norm": 3.025534152984619, "learning_rate": 5.564152589702021e-07, "loss": 0.2584, "step": 41030 }, { "epoch": 2.4273969361802803, "grad_norm": 4.004396438598633, "learning_rate": 5.553044462397193e-07, "loss": 0.2203, "step": 41040 }, { "epoch": 2.4279884071686286, "grad_norm": 2.707599639892578, "learning_rate": 5.541946303182624e-07, "loss": 0.2623, "step": 41050 }, { "epoch": 2.4285798781569765, "grad_norm": 4.548275470733643, "learning_rate": 5.530858116583531e-07, "loss": 0.2366, "step": 41060 }, { "epoch": 2.4291713491453244, "grad_norm": 4.756258964538574, "learning_rate": 5.519779907121053e-07, "loss": 0.2933, "step": 41070 }, { "epoch": 2.4297628201336723, "grad_norm": 4.560119152069092, "learning_rate": 5.508711679312241e-07, "loss": 0.2867, "step": 41080 }, { "epoch": 2.4303542911220206, "grad_norm": 3.7374637126922607, "learning_rate": 5.49765343767012e-07, "loss": 0.2504, "step": 41090 }, { "epoch": 2.4309457621103685, "grad_norm": 6.690953254699707, "learning_rate": 5.486605186703593e-07, "loss": 0.2583, "step": 41100 }, { "epoch": 2.4315372330987164, "grad_norm": 4.285192966461182, "learning_rate": 5.475566930917534e-07, "loss": 0.2393, "step": 41110 }, { "epoch": 2.4321287040870647, "grad_norm": 2.9288501739501953, "learning_rate": 5.464538674812704e-07, "loss": 0.2574, "step": 41120 }, { "epoch": 2.4327201750754126, "grad_norm": 2.7828688621520996, "learning_rate": 5.453520422885821e-07, "loss": 0.2834, "step": 41130 }, { "epoch": 2.4333116460637605, "grad_norm": 3.9414408206939697, "learning_rate": 5.442512179629501e-07, "loss": 0.3371, "step": 41140 }, { "epoch": 2.4339031170521084, "grad_norm": 3.751842498779297, "learning_rate": 5.431513949532274e-07, "loss": 0.3045, "step": 41150 }, { "epoch": 2.4344945880404567, "grad_norm": 8.37027645111084, "learning_rate": 5.420525737078614e-07, "loss": 0.2085, "step": 41160 }, { "epoch": 2.4350860590288046, "grad_norm": 3.202549934387207, "learning_rate": 5.409547546748882e-07, "loss": 0.2516, "step": 41170 }, { "epoch": 2.4356775300171525, "grad_norm": 2.543661117553711, "learning_rate": 5.398579383019376e-07, "loss": 0.2643, "step": 41180 }, { "epoch": 2.436269001005501, "grad_norm": 3.4571149349212646, "learning_rate": 5.38762125036229e-07, "loss": 0.2756, "step": 41190 }, { "epoch": 2.4368604719938487, "grad_norm": 5.299675941467285, "learning_rate": 5.376673153245726e-07, "loss": 0.2714, "step": 41200 }, { "epoch": 2.4374519429821966, "grad_norm": 4.064559459686279, "learning_rate": 5.365735096133723e-07, "loss": 0.2435, "step": 41210 }, { "epoch": 2.438043413970545, "grad_norm": 5.029878616333008, "learning_rate": 5.354807083486173e-07, "loss": 0.2816, "step": 41220 }, { "epoch": 2.438634884958893, "grad_norm": 2.6937899589538574, "learning_rate": 5.343889119758925e-07, "loss": 0.194, "step": 41230 }, { "epoch": 2.4392263559472407, "grad_norm": 3.5077743530273438, "learning_rate": 5.332981209403693e-07, "loss": 0.279, "step": 41240 }, { "epoch": 2.439817826935589, "grad_norm": 2.7880759239196777, "learning_rate": 5.322083356868128e-07, "loss": 0.2416, "step": 41250 }, { "epoch": 2.440409297923937, "grad_norm": 4.660515785217285, "learning_rate": 5.311195566595743e-07, "loss": 0.2311, "step": 41260 }, { "epoch": 2.441000768912285, "grad_norm": 5.368011951446533, "learning_rate": 5.300317843025978e-07, "loss": 0.2608, "step": 41270 }, { "epoch": 2.4415922399006327, "grad_norm": 3.1949782371520996, "learning_rate": 5.289450190594151e-07, "loss": 0.2229, "step": 41280 }, { "epoch": 2.442183710888981, "grad_norm": 5.195900917053223, "learning_rate": 5.278592613731466e-07, "loss": 0.2475, "step": 41290 }, { "epoch": 2.442775181877329, "grad_norm": 3.799980640411377, "learning_rate": 5.267745116865051e-07, "loss": 0.2786, "step": 41300 }, { "epoch": 2.443366652865677, "grad_norm": 3.0978424549102783, "learning_rate": 5.256907704417883e-07, "loss": 0.2369, "step": 41310 }, { "epoch": 2.443958123854025, "grad_norm": 3.696993827819824, "learning_rate": 5.246080380808867e-07, "loss": 0.2737, "step": 41320 }, { "epoch": 2.444549594842373, "grad_norm": 4.137643337249756, "learning_rate": 5.235263150452768e-07, "loss": 0.2994, "step": 41330 }, { "epoch": 2.445141065830721, "grad_norm": 2.598311424255371, "learning_rate": 5.224456017760234e-07, "loss": 0.2428, "step": 41340 }, { "epoch": 2.445732536819069, "grad_norm": 3.0384521484375, "learning_rate": 5.213658987137817e-07, "loss": 0.3492, "step": 41350 }, { "epoch": 2.446324007807417, "grad_norm": 2.323138952255249, "learning_rate": 5.202872062987926e-07, "loss": 0.2596, "step": 41360 }, { "epoch": 2.446915478795765, "grad_norm": 4.683264255523682, "learning_rate": 5.192095249708875e-07, "loss": 0.2629, "step": 41370 }, { "epoch": 2.4475069497841133, "grad_norm": 3.8818061351776123, "learning_rate": 5.18132855169483e-07, "loss": 0.2535, "step": 41380 }, { "epoch": 2.448098420772461, "grad_norm": 5.0063581466674805, "learning_rate": 5.170571973335842e-07, "loss": 0.2358, "step": 41390 }, { "epoch": 2.448689891760809, "grad_norm": 2.305802345275879, "learning_rate": 5.159825519017846e-07, "loss": 0.2788, "step": 41400 }, { "epoch": 2.449281362749157, "grad_norm": 2.4624195098876953, "learning_rate": 5.14908919312263e-07, "loss": 0.3122, "step": 41410 }, { "epoch": 2.4498728337375053, "grad_norm": 3.50945782661438, "learning_rate": 5.138363000027874e-07, "loss": 0.2679, "step": 41420 }, { "epoch": 2.450464304725853, "grad_norm": 4.021756172180176, "learning_rate": 5.127646944107111e-07, "loss": 0.2792, "step": 41430 }, { "epoch": 2.451055775714201, "grad_norm": 4.643605709075928, "learning_rate": 5.116941029729743e-07, "loss": 0.2633, "step": 41440 }, { "epoch": 2.4516472467025494, "grad_norm": 1.7948857545852661, "learning_rate": 5.106245261261027e-07, "loss": 0.2425, "step": 41450 }, { "epoch": 2.4522387176908973, "grad_norm": 1.9347032308578491, "learning_rate": 5.095559643062114e-07, "loss": 0.2615, "step": 41460 }, { "epoch": 2.452830188679245, "grad_norm": 3.622103452682495, "learning_rate": 5.084884179489987e-07, "loss": 0.2693, "step": 41470 }, { "epoch": 2.453421659667593, "grad_norm": 3.6812849044799805, "learning_rate": 5.074218874897494e-07, "loss": 0.2516, "step": 41480 }, { "epoch": 2.4540131306559414, "grad_norm": 5.605101108551025, "learning_rate": 5.063563733633356e-07, "loss": 0.3084, "step": 41490 }, { "epoch": 2.4546046016442893, "grad_norm": 4.8234429359436035, "learning_rate": 5.052918760042125e-07, "loss": 0.2862, "step": 41500 }, { "epoch": 2.4551960726326376, "grad_norm": 2.6035728454589844, "learning_rate": 5.042283958464237e-07, "loss": 0.2379, "step": 41510 }, { "epoch": 2.4557875436209855, "grad_norm": 2.9586660861968994, "learning_rate": 5.031659333235957e-07, "loss": 0.2765, "step": 41520 }, { "epoch": 2.4563790146093334, "grad_norm": 3.74186372756958, "learning_rate": 5.021044888689399e-07, "loss": 0.3452, "step": 41530 }, { "epoch": 2.4569704855976813, "grad_norm": 2.551831007003784, "learning_rate": 5.010440629152553e-07, "loss": 0.2308, "step": 41540 }, { "epoch": 2.4575619565860296, "grad_norm": 4.911281108856201, "learning_rate": 4.999846558949221e-07, "loss": 0.2486, "step": 41550 }, { "epoch": 2.4581534275743775, "grad_norm": 5.93736457824707, "learning_rate": 4.989262682399083e-07, "loss": 0.2442, "step": 41560 }, { "epoch": 2.4587448985627254, "grad_norm": 2.122197389602661, "learning_rate": 4.978689003817631e-07, "loss": 0.2914, "step": 41570 }, { "epoch": 2.4593363695510737, "grad_norm": 2.0283641815185547, "learning_rate": 4.968125527516234e-07, "loss": 0.2329, "step": 41580 }, { "epoch": 2.4599278405394216, "grad_norm": 3.1489851474761963, "learning_rate": 4.957572257802072e-07, "loss": 0.217, "step": 41590 }, { "epoch": 2.4605193115277695, "grad_norm": 1.3882060050964355, "learning_rate": 4.947029198978164e-07, "loss": 0.2329, "step": 41600 }, { "epoch": 2.4611107825161174, "grad_norm": 1.643277883529663, "learning_rate": 4.936496355343393e-07, "loss": 0.1996, "step": 41610 }, { "epoch": 2.4617022535044657, "grad_norm": 2.494767427444458, "learning_rate": 4.925973731192441e-07, "loss": 0.3097, "step": 41620 }, { "epoch": 2.4622937244928136, "grad_norm": 4.898434638977051, "learning_rate": 4.91546133081586e-07, "loss": 0.2382, "step": 41630 }, { "epoch": 2.4628851954811615, "grad_norm": 5.123208045959473, "learning_rate": 4.904959158500004e-07, "loss": 0.2928, "step": 41640 }, { "epoch": 2.46347666646951, "grad_norm": 3.30802321434021, "learning_rate": 4.894467218527068e-07, "loss": 0.2588, "step": 41650 }, { "epoch": 2.4640681374578577, "grad_norm": 2.865478992462158, "learning_rate": 4.883985515175073e-07, "loss": 0.2281, "step": 41660 }, { "epoch": 2.4646596084462056, "grad_norm": 2.8032190799713135, "learning_rate": 4.873514052717859e-07, "loss": 0.2565, "step": 41670 }, { "epoch": 2.465251079434554, "grad_norm": 5.27036714553833, "learning_rate": 4.863052835425113e-07, "loss": 0.2249, "step": 41680 }, { "epoch": 2.465842550422902, "grad_norm": 3.7820818424224854, "learning_rate": 4.852601867562314e-07, "loss": 0.2566, "step": 41690 }, { "epoch": 2.4664340214112497, "grad_norm": 3.8379952907562256, "learning_rate": 4.842161153390795e-07, "loss": 0.2489, "step": 41700 }, { "epoch": 2.467025492399598, "grad_norm": 4.6055145263671875, "learning_rate": 4.83173069716768e-07, "loss": 0.2403, "step": 41710 }, { "epoch": 2.467616963387946, "grad_norm": 4.645397186279297, "learning_rate": 4.821310503145917e-07, "loss": 0.2627, "step": 41720 }, { "epoch": 2.468208434376294, "grad_norm": 6.889851093292236, "learning_rate": 4.810900575574284e-07, "loss": 0.2242, "step": 41730 }, { "epoch": 2.4687999053646417, "grad_norm": 6.107765197753906, "learning_rate": 4.800500918697352e-07, "loss": 0.2497, "step": 41740 }, { "epoch": 2.46939137635299, "grad_norm": 2.5591585636138916, "learning_rate": 4.790111536755527e-07, "loss": 0.2284, "step": 41750 }, { "epoch": 2.469982847341338, "grad_norm": 3.9893343448638916, "learning_rate": 4.779732433985003e-07, "loss": 0.19, "step": 41760 }, { "epoch": 2.4705743183296858, "grad_norm": 4.762939929962158, "learning_rate": 4.7693636146178e-07, "loss": 0.21, "step": 41770 }, { "epoch": 2.471165789318034, "grad_norm": 4.598576545715332, "learning_rate": 4.759005082881738e-07, "loss": 0.3358, "step": 41780 }, { "epoch": 2.471757260306382, "grad_norm": 1.4369169473648071, "learning_rate": 4.748656843000431e-07, "loss": 0.2158, "step": 41790 }, { "epoch": 2.47234873129473, "grad_norm": 3.4487617015838623, "learning_rate": 4.73831889919332e-07, "loss": 0.2628, "step": 41800 }, { "epoch": 2.472940202283078, "grad_norm": 4.175044536590576, "learning_rate": 4.727991255675628e-07, "loss": 0.1907, "step": 41810 }, { "epoch": 2.473531673271426, "grad_norm": 4.044718265533447, "learning_rate": 4.717673916658395e-07, "loss": 0.2507, "step": 41820 }, { "epoch": 2.474123144259774, "grad_norm": 2.283686637878418, "learning_rate": 4.707366886348446e-07, "loss": 0.2536, "step": 41830 }, { "epoch": 2.4747146152481223, "grad_norm": 4.309886455535889, "learning_rate": 4.697070168948401e-07, "loss": 0.2869, "step": 41840 }, { "epoch": 2.47530608623647, "grad_norm": 2.518158197402954, "learning_rate": 4.6867837686566897e-07, "loss": 0.2487, "step": 41850 }, { "epoch": 2.475897557224818, "grad_norm": 2.395402431488037, "learning_rate": 4.6765076896675207e-07, "loss": 0.2194, "step": 41860 }, { "epoch": 2.476489028213166, "grad_norm": 3.456690549850464, "learning_rate": 4.666241936170903e-07, "loss": 0.2279, "step": 41870 }, { "epoch": 2.4770804992015143, "grad_norm": 6.503537178039551, "learning_rate": 4.6559865123526215e-07, "loss": 0.3286, "step": 41880 }, { "epoch": 2.477671970189862, "grad_norm": 5.228196144104004, "learning_rate": 4.6457414223942776e-07, "loss": 0.2828, "step": 41890 }, { "epoch": 2.47826344117821, "grad_norm": 2.870009183883667, "learning_rate": 4.635506670473223e-07, "loss": 0.2356, "step": 41900 }, { "epoch": 2.4788549121665584, "grad_norm": 3.6249797344207764, "learning_rate": 4.62528226076263e-07, "loss": 0.2611, "step": 41910 }, { "epoch": 2.4794463831549063, "grad_norm": 4.040245532989502, "learning_rate": 4.61506819743143e-07, "loss": 0.2958, "step": 41920 }, { "epoch": 2.480037854143254, "grad_norm": 3.1008236408233643, "learning_rate": 4.6048644846443324e-07, "loss": 0.2739, "step": 41930 }, { "epoch": 2.480629325131602, "grad_norm": 4.7979888916015625, "learning_rate": 4.594671126561851e-07, "loss": 0.267, "step": 41940 }, { "epoch": 2.4812207961199504, "grad_norm": 5.5165324211120605, "learning_rate": 4.5844881273402495e-07, "loss": 0.2379, "step": 41950 }, { "epoch": 2.4818122671082983, "grad_norm": 2.9270107746124268, "learning_rate": 4.574315491131592e-07, "loss": 0.231, "step": 41960 }, { "epoch": 2.4824037380966466, "grad_norm": 4.2566399574279785, "learning_rate": 4.5641532220837045e-07, "loss": 0.2927, "step": 41970 }, { "epoch": 2.4829952090849945, "grad_norm": 9.795827865600586, "learning_rate": 4.554001324340176e-07, "loss": 0.2445, "step": 41980 }, { "epoch": 2.4835866800733424, "grad_norm": 3.5268349647521973, "learning_rate": 4.543859802040391e-07, "loss": 0.298, "step": 41990 }, { "epoch": 2.4841781510616903, "grad_norm": 8.03066635131836, "learning_rate": 4.5337286593194827e-07, "loss": 0.2688, "step": 42000 }, { "epoch": 2.4847696220500386, "grad_norm": 3.9399755001068115, "learning_rate": 4.523607900308366e-07, "loss": 0.2674, "step": 42010 }, { "epoch": 2.4853610930383865, "grad_norm": 4.496291637420654, "learning_rate": 4.5134975291337175e-07, "loss": 0.3073, "step": 42020 }, { "epoch": 2.4859525640267344, "grad_norm": 5.357113361358643, "learning_rate": 4.5033975499179623e-07, "loss": 0.3169, "step": 42030 }, { "epoch": 2.4865440350150827, "grad_norm": 2.658388137817383, "learning_rate": 4.4933079667793197e-07, "loss": 0.2447, "step": 42040 }, { "epoch": 2.4871355060034306, "grad_norm": 2.6241769790649414, "learning_rate": 4.483228783831739e-07, "loss": 0.2784, "step": 42050 }, { "epoch": 2.4877269769917785, "grad_norm": 3.3541321754455566, "learning_rate": 4.473160005184955e-07, "loss": 0.2897, "step": 42060 }, { "epoch": 2.4883184479801264, "grad_norm": 3.286064863204956, "learning_rate": 4.463101634944445e-07, "loss": 0.221, "step": 42070 }, { "epoch": 2.4889099189684747, "grad_norm": 3.036853551864624, "learning_rate": 4.453053677211438e-07, "loss": 0.2503, "step": 42080 }, { "epoch": 2.4895013899568226, "grad_norm": 3.5903353691101074, "learning_rate": 4.443016136082927e-07, "loss": 0.284, "step": 42090 }, { "epoch": 2.4900928609451705, "grad_norm": 3.1972861289978027, "learning_rate": 4.432989015651664e-07, "loss": 0.1923, "step": 42100 }, { "epoch": 2.490684331933519, "grad_norm": 3.005530834197998, "learning_rate": 4.4229723200061357e-07, "loss": 0.3, "step": 42110 }, { "epoch": 2.4912758029218667, "grad_norm": 3.7625038623809814, "learning_rate": 4.4129660532305814e-07, "loss": 0.2668, "step": 42120 }, { "epoch": 2.4918672739102146, "grad_norm": 5.828659534454346, "learning_rate": 4.402970219405005e-07, "loss": 0.292, "step": 42130 }, { "epoch": 2.492458744898563, "grad_norm": 2.786600351333618, "learning_rate": 4.3929848226051316e-07, "loss": 0.2474, "step": 42140 }, { "epoch": 2.493050215886911, "grad_norm": 5.535369396209717, "learning_rate": 4.383009866902452e-07, "loss": 0.245, "step": 42150 }, { "epoch": 2.4936416868752587, "grad_norm": 2.1756527423858643, "learning_rate": 4.373045356364193e-07, "loss": 0.2014, "step": 42160 }, { "epoch": 2.494233157863607, "grad_norm": 2.8561794757843018, "learning_rate": 4.3630912950533063e-07, "loss": 0.2781, "step": 42170 }, { "epoch": 2.494824628851955, "grad_norm": 2.078529119491577, "learning_rate": 4.3531476870285135e-07, "loss": 0.2297, "step": 42180 }, { "epoch": 2.4954160998403028, "grad_norm": 3.3922388553619385, "learning_rate": 4.343214536344241e-07, "loss": 0.2378, "step": 42190 }, { "epoch": 2.4960075708286507, "grad_norm": 4.049450874328613, "learning_rate": 4.3332918470506857e-07, "loss": 0.334, "step": 42200 }, { "epoch": 2.496599041816999, "grad_norm": 4.090813159942627, "learning_rate": 4.323379623193741e-07, "loss": 0.2535, "step": 42210 }, { "epoch": 2.497190512805347, "grad_norm": 4.718582630157471, "learning_rate": 4.313477868815073e-07, "loss": 0.284, "step": 42220 }, { "epoch": 2.4977819837936948, "grad_norm": 4.583869934082031, "learning_rate": 4.3035865879520475e-07, "loss": 0.2824, "step": 42230 }, { "epoch": 2.498373454782043, "grad_norm": 2.4432008266448975, "learning_rate": 4.293705784637767e-07, "loss": 0.2642, "step": 42240 }, { "epoch": 2.498964925770391, "grad_norm": 3.1839451789855957, "learning_rate": 4.2838354629010823e-07, "loss": 0.2469, "step": 42250 }, { "epoch": 2.499556396758739, "grad_norm": 5.678066730499268, "learning_rate": 4.273975626766534e-07, "loss": 0.2384, "step": 42260 }, { "epoch": 2.5001478677470867, "grad_norm": 3.161421060562134, "learning_rate": 4.2641262802544235e-07, "loss": 0.2371, "step": 42270 }, { "epoch": 2.500739338735435, "grad_norm": 3.3236770629882812, "learning_rate": 4.254287427380754e-07, "loss": 0.2564, "step": 42280 }, { "epoch": 2.501330809723783, "grad_norm": 3.477947950363159, "learning_rate": 4.2444590721572573e-07, "loss": 0.309, "step": 42290 }, { "epoch": 2.5019222807121313, "grad_norm": 6.887374401092529, "learning_rate": 4.234641218591374e-07, "loss": 0.2378, "step": 42300 }, { "epoch": 2.502513751700479, "grad_norm": 3.5843002796173096, "learning_rate": 4.224833870686272e-07, "loss": 0.2055, "step": 42310 }, { "epoch": 2.503105222688827, "grad_norm": 5.580773830413818, "learning_rate": 4.2150370324408493e-07, "loss": 0.2968, "step": 42320 }, { "epoch": 2.503696693677175, "grad_norm": 3.2768256664276123, "learning_rate": 4.205250707849684e-07, "loss": 0.2622, "step": 42330 }, { "epoch": 2.5042881646655233, "grad_norm": 3.139451503753662, "learning_rate": 4.195474900903107e-07, "loss": 0.2787, "step": 42340 }, { "epoch": 2.504879635653871, "grad_norm": 4.498793125152588, "learning_rate": 4.185709615587123e-07, "loss": 0.2827, "step": 42350 }, { "epoch": 2.505471106642219, "grad_norm": 5.6750383377075195, "learning_rate": 4.1759548558834815e-07, "loss": 0.2591, "step": 42360 }, { "epoch": 2.5060625776305674, "grad_norm": 3.1615114212036133, "learning_rate": 4.1662106257696197e-07, "loss": 0.2739, "step": 42370 }, { "epoch": 2.5066540486189153, "grad_norm": 5.027472019195557, "learning_rate": 4.1564769292186697e-07, "loss": 0.2624, "step": 42380 }, { "epoch": 2.507245519607263, "grad_norm": 2.266758441925049, "learning_rate": 4.146753770199507e-07, "loss": 0.2577, "step": 42390 }, { "epoch": 2.507836990595611, "grad_norm": 2.8389694690704346, "learning_rate": 4.1370411526766707e-07, "loss": 0.2412, "step": 42400 }, { "epoch": 2.5084284615839594, "grad_norm": 5.728036880493164, "learning_rate": 4.127339080610428e-07, "loss": 0.2843, "step": 42410 }, { "epoch": 2.5090199325723073, "grad_norm": 3.4502952098846436, "learning_rate": 4.1176475579567365e-07, "loss": 0.229, "step": 42420 }, { "epoch": 2.5096114035606556, "grad_norm": 4.7375311851501465, "learning_rate": 4.1079665886672434e-07, "loss": 0.261, "step": 42430 }, { "epoch": 2.5102028745490035, "grad_norm": 7.34950590133667, "learning_rate": 4.0982961766893156e-07, "loss": 0.321, "step": 42440 }, { "epoch": 2.5107943455373514, "grad_norm": 2.7612783908843994, "learning_rate": 4.088636325965989e-07, "loss": 0.2113, "step": 42450 }, { "epoch": 2.5113858165256993, "grad_norm": 4.922420024871826, "learning_rate": 4.0789870404360187e-07, "loss": 0.2681, "step": 42460 }, { "epoch": 2.5119772875140476, "grad_norm": 2.7045764923095703, "learning_rate": 4.0693483240338346e-07, "loss": 0.2337, "step": 42470 }, { "epoch": 2.5125687585023955, "grad_norm": 3.8351714611053467, "learning_rate": 4.0597201806895535e-07, "loss": 0.2028, "step": 42480 }, { "epoch": 2.5131602294907434, "grad_norm": 5.3030290603637695, "learning_rate": 4.050102614329001e-07, "loss": 0.2516, "step": 42490 }, { "epoch": 2.5137517004790917, "grad_norm": 3.035473346710205, "learning_rate": 4.040495628873674e-07, "loss": 0.2819, "step": 42500 }, { "epoch": 2.5143431714674396, "grad_norm": 2.4334938526153564, "learning_rate": 4.030899228240762e-07, "loss": 0.2125, "step": 42510 }, { "epoch": 2.5149346424557875, "grad_norm": 5.497757434844971, "learning_rate": 4.0213134163431254e-07, "loss": 0.2607, "step": 42520 }, { "epoch": 2.5155261134441353, "grad_norm": 3.187373638153076, "learning_rate": 4.011738197089331e-07, "loss": 0.2132, "step": 42530 }, { "epoch": 2.5161175844324837, "grad_norm": 4.248727321624756, "learning_rate": 4.0021735743836053e-07, "loss": 0.2427, "step": 42540 }, { "epoch": 2.5167090554208316, "grad_norm": 5.665658950805664, "learning_rate": 3.992619552125869e-07, "loss": 0.2989, "step": 42550 }, { "epoch": 2.51730052640918, "grad_norm": 5.403797149658203, "learning_rate": 3.983076134211713e-07, "loss": 0.2375, "step": 42560 }, { "epoch": 2.517891997397528, "grad_norm": 3.0099711418151855, "learning_rate": 3.973543324532394e-07, "loss": 0.2551, "step": 42570 }, { "epoch": 2.5184834683858757, "grad_norm": 4.293970108032227, "learning_rate": 3.9640211269748697e-07, "loss": 0.2666, "step": 42580 }, { "epoch": 2.5190749393742236, "grad_norm": 3.4270918369293213, "learning_rate": 3.954509545421744e-07, "loss": 0.2585, "step": 42590 }, { "epoch": 2.5196664103625714, "grad_norm": 4.593798637390137, "learning_rate": 3.945008583751312e-07, "loss": 0.2665, "step": 42600 }, { "epoch": 2.5202578813509198, "grad_norm": 5.428871154785156, "learning_rate": 3.9355182458375274e-07, "loss": 0.2232, "step": 42610 }, { "epoch": 2.5208493523392677, "grad_norm": 3.229423999786377, "learning_rate": 3.926038535550005e-07, "loss": 0.2831, "step": 42620 }, { "epoch": 2.521440823327616, "grad_norm": 3.9587326049804688, "learning_rate": 3.916569456754057e-07, "loss": 0.2623, "step": 42630 }, { "epoch": 2.522032294315964, "grad_norm": 5.715728282928467, "learning_rate": 3.9071110133106167e-07, "loss": 0.262, "step": 42640 }, { "epoch": 2.5226237653043118, "grad_norm": 5.587085247039795, "learning_rate": 3.897663209076324e-07, "loss": 0.2845, "step": 42650 }, { "epoch": 2.5232152362926596, "grad_norm": 4.441544055938721, "learning_rate": 3.888226047903447e-07, "loss": 0.254, "step": 42660 }, { "epoch": 2.523806707281008, "grad_norm": 3.137707233428955, "learning_rate": 3.878799533639937e-07, "loss": 0.2931, "step": 42670 }, { "epoch": 2.524398178269356, "grad_norm": 4.277061462402344, "learning_rate": 3.869383670129394e-07, "loss": 0.2676, "step": 42680 }, { "epoch": 2.5249896492577037, "grad_norm": 3.865785837173462, "learning_rate": 3.8599784612110656e-07, "loss": 0.241, "step": 42690 }, { "epoch": 2.525581120246052, "grad_norm": 3.8568387031555176, "learning_rate": 3.8505839107198813e-07, "loss": 0.2632, "step": 42700 }, { "epoch": 2.5261725912344, "grad_norm": 1.9897544384002686, "learning_rate": 3.841200022486394e-07, "loss": 0.2473, "step": 42710 }, { "epoch": 2.526764062222748, "grad_norm": 2.7865512371063232, "learning_rate": 3.831826800336846e-07, "loss": 0.2324, "step": 42720 }, { "epoch": 2.5273555332110957, "grad_norm": 3.8217689990997314, "learning_rate": 3.822464248093084e-07, "loss": 0.2083, "step": 42730 }, { "epoch": 2.527947004199444, "grad_norm": 4.141240119934082, "learning_rate": 3.8131123695726433e-07, "loss": 0.2755, "step": 42740 }, { "epoch": 2.528538475187792, "grad_norm": 4.832301139831543, "learning_rate": 3.803771168588691e-07, "loss": 0.2625, "step": 42750 }, { "epoch": 2.5291299461761403, "grad_norm": 2.5303781032562256, "learning_rate": 3.794440648950034e-07, "loss": 0.2662, "step": 42760 }, { "epoch": 2.529721417164488, "grad_norm": 2.8743176460266113, "learning_rate": 3.7851208144611483e-07, "loss": 0.2224, "step": 42770 }, { "epoch": 2.530312888152836, "grad_norm": 4.7465901374816895, "learning_rate": 3.775811668922118e-07, "loss": 0.2706, "step": 42780 }, { "epoch": 2.530904359141184, "grad_norm": 6.319700241088867, "learning_rate": 3.766513216128703e-07, "loss": 0.2493, "step": 42790 }, { "epoch": 2.5314958301295323, "grad_norm": 4.467331886291504, "learning_rate": 3.7572254598722775e-07, "loss": 0.2917, "step": 42800 }, { "epoch": 2.53208730111788, "grad_norm": 4.058661460876465, "learning_rate": 3.747948403939876e-07, "loss": 0.2691, "step": 42810 }, { "epoch": 2.532678772106228, "grad_norm": 2.3174314498901367, "learning_rate": 3.738682052114155e-07, "loss": 0.2775, "step": 42820 }, { "epoch": 2.5332702430945764, "grad_norm": 3.2949447631835938, "learning_rate": 3.7294264081734e-07, "loss": 0.2797, "step": 42830 }, { "epoch": 2.5338617140829243, "grad_norm": 3.6501541137695312, "learning_rate": 3.7201814758915543e-07, "loss": 0.2274, "step": 42840 }, { "epoch": 2.534453185071272, "grad_norm": 2.356595754623413, "learning_rate": 3.71094725903817e-07, "loss": 0.2094, "step": 42850 }, { "epoch": 2.53504465605962, "grad_norm": 3.0048797130584717, "learning_rate": 3.7017237613784516e-07, "loss": 0.2546, "step": 42860 }, { "epoch": 2.5356361270479684, "grad_norm": 7.170106887817383, "learning_rate": 3.692510986673214e-07, "loss": 0.3421, "step": 42870 }, { "epoch": 2.5362275980363163, "grad_norm": 4.40541934967041, "learning_rate": 3.683308938678903e-07, "loss": 0.2918, "step": 42880 }, { "epoch": 2.5368190690246646, "grad_norm": 4.103330135345459, "learning_rate": 3.674117621147607e-07, "loss": 0.272, "step": 42890 }, { "epoch": 2.5374105400130125, "grad_norm": 3.703587055206299, "learning_rate": 3.664937037827014e-07, "loss": 0.2713, "step": 42900 }, { "epoch": 2.5380020110013604, "grad_norm": 4.223320484161377, "learning_rate": 3.6557671924604606e-07, "loss": 0.1927, "step": 42910 }, { "epoch": 2.5385934819897082, "grad_norm": 1.835263729095459, "learning_rate": 3.646608088786886e-07, "loss": 0.2525, "step": 42920 }, { "epoch": 2.5391849529780566, "grad_norm": 2.956939935684204, "learning_rate": 3.637459730540859e-07, "loss": 0.2598, "step": 42930 }, { "epoch": 2.5397764239664045, "grad_norm": 3.728458881378174, "learning_rate": 3.628322121452559e-07, "loss": 0.3194, "step": 42940 }, { "epoch": 2.5403678949547523, "grad_norm": 2.9048237800598145, "learning_rate": 3.6191952652477887e-07, "loss": 0.2986, "step": 42950 }, { "epoch": 2.5409593659431007, "grad_norm": 3.8490967750549316, "learning_rate": 3.6100791656479727e-07, "loss": 0.2509, "step": 42960 }, { "epoch": 2.5415508369314486, "grad_norm": 4.343085765838623, "learning_rate": 3.600973826370133e-07, "loss": 0.3263, "step": 42970 }, { "epoch": 2.5421423079197965, "grad_norm": 3.5647072792053223, "learning_rate": 3.591879251126924e-07, "loss": 0.2613, "step": 42980 }, { "epoch": 2.5427337789081443, "grad_norm": 2.781684160232544, "learning_rate": 3.582795443626592e-07, "loss": 0.2343, "step": 42990 }, { "epoch": 2.5433252498964927, "grad_norm": 2.6549789905548096, "learning_rate": 3.573722407573008e-07, "loss": 0.241, "step": 43000 }, { "epoch": 2.5439167208848406, "grad_norm": 2.1007533073425293, "learning_rate": 3.564660146665643e-07, "loss": 0.2052, "step": 43010 }, { "epoch": 2.544508191873189, "grad_norm": 2.644181489944458, "learning_rate": 3.55560866459957e-07, "loss": 0.2403, "step": 43020 }, { "epoch": 2.5450996628615368, "grad_norm": 3.8100028038024902, "learning_rate": 3.5465679650654815e-07, "loss": 0.3112, "step": 43030 }, { "epoch": 2.5456911338498847, "grad_norm": 2.72637939453125, "learning_rate": 3.537538051749658e-07, "loss": 0.2043, "step": 43040 }, { "epoch": 2.5462826048382325, "grad_norm": 2.0585601329803467, "learning_rate": 3.528518928333998e-07, "loss": 0.2224, "step": 43050 }, { "epoch": 2.5468740758265804, "grad_norm": 3.1916239261627197, "learning_rate": 3.519510598495983e-07, "loss": 0.2467, "step": 43060 }, { "epoch": 2.5474655468149288, "grad_norm": 1.9193655252456665, "learning_rate": 3.510513065908705e-07, "loss": 0.2693, "step": 43070 }, { "epoch": 2.5480570178032766, "grad_norm": 5.574139595031738, "learning_rate": 3.501526334240852e-07, "loss": 0.2672, "step": 43080 }, { "epoch": 2.548648488791625, "grad_norm": 2.835552453994751, "learning_rate": 3.4925504071567005e-07, "loss": 0.2649, "step": 43090 }, { "epoch": 2.549239959779973, "grad_norm": 10.421663284301758, "learning_rate": 3.483585288316139e-07, "loss": 0.2876, "step": 43100 }, { "epoch": 2.5498314307683208, "grad_norm": 4.061481952667236, "learning_rate": 3.47463098137463e-07, "loss": 0.2409, "step": 43110 }, { "epoch": 2.5504229017566686, "grad_norm": 2.038593053817749, "learning_rate": 3.465687489983234e-07, "loss": 0.2552, "step": 43120 }, { "epoch": 2.551014372745017, "grad_norm": 3.2889702320098877, "learning_rate": 3.4567548177886053e-07, "loss": 0.2343, "step": 43130 }, { "epoch": 2.551605843733365, "grad_norm": 5.074699401855469, "learning_rate": 3.447832968432979e-07, "loss": 0.2667, "step": 43140 }, { "epoch": 2.5521973147217127, "grad_norm": 3.086184024810791, "learning_rate": 3.438921945554201e-07, "loss": 0.3108, "step": 43150 }, { "epoch": 2.552788785710061, "grad_norm": 3.9150595664978027, "learning_rate": 3.4300217527856546e-07, "loss": 0.2472, "step": 43160 }, { "epoch": 2.553380256698409, "grad_norm": 2.1583170890808105, "learning_rate": 3.421132393756361e-07, "loss": 0.3256, "step": 43170 }, { "epoch": 2.553971727686757, "grad_norm": 2.571340322494507, "learning_rate": 3.41225387209088e-07, "loss": 0.2668, "step": 43180 }, { "epoch": 2.5545631986751047, "grad_norm": 4.319618225097656, "learning_rate": 3.4033861914093936e-07, "loss": 0.2476, "step": 43190 }, { "epoch": 2.555154669663453, "grad_norm": 3.2042899131774902, "learning_rate": 3.394529355327628e-07, "loss": 0.2135, "step": 43200 }, { "epoch": 2.555746140651801, "grad_norm": 2.3420073986053467, "learning_rate": 3.385683367456899e-07, "loss": 0.1955, "step": 43210 }, { "epoch": 2.5563376116401493, "grad_norm": 3.0815303325653076, "learning_rate": 3.376848231404114e-07, "loss": 0.3107, "step": 43220 }, { "epoch": 2.556929082628497, "grad_norm": 5.36102294921875, "learning_rate": 3.368023950771731e-07, "loss": 0.2891, "step": 43230 }, { "epoch": 2.557520553616845, "grad_norm": 6.79210090637207, "learning_rate": 3.359210529157807e-07, "loss": 0.2669, "step": 43240 }, { "epoch": 2.558112024605193, "grad_norm": 4.280198574066162, "learning_rate": 3.350407970155952e-07, "loss": 0.2632, "step": 43250 }, { "epoch": 2.5587034955935413, "grad_norm": 4.189841270446777, "learning_rate": 3.341616277355346e-07, "loss": 0.2198, "step": 43260 }, { "epoch": 2.559294966581889, "grad_norm": 3.0976717472076416, "learning_rate": 3.332835454340758e-07, "loss": 0.2379, "step": 43270 }, { "epoch": 2.559886437570237, "grad_norm": 3.0150458812713623, "learning_rate": 3.3240655046925003e-07, "loss": 0.2816, "step": 43280 }, { "epoch": 2.5604779085585854, "grad_norm": 2.204524517059326, "learning_rate": 3.3153064319864803e-07, "loss": 0.3029, "step": 43290 }, { "epoch": 2.5610693795469333, "grad_norm": 3.5494492053985596, "learning_rate": 3.306558239794136e-07, "loss": 0.2661, "step": 43300 }, { "epoch": 2.561660850535281, "grad_norm": 4.869757652282715, "learning_rate": 3.2978209316825023e-07, "loss": 0.2244, "step": 43310 }, { "epoch": 2.562252321523629, "grad_norm": 2.322106122970581, "learning_rate": 3.289094511214156e-07, "loss": 0.2353, "step": 43320 }, { "epoch": 2.5628437925119774, "grad_norm": 5.868742942810059, "learning_rate": 3.280378981947226e-07, "loss": 0.2836, "step": 43330 }, { "epoch": 2.5634352635003252, "grad_norm": 8.468486785888672, "learning_rate": 3.2716743474354383e-07, "loss": 0.2992, "step": 43340 }, { "epoch": 2.5640267344886736, "grad_norm": 4.671648025512695, "learning_rate": 3.262980611228029e-07, "loss": 0.2938, "step": 43350 }, { "epoch": 2.5646182054770215, "grad_norm": 4.500870704650879, "learning_rate": 3.254297776869837e-07, "loss": 0.3099, "step": 43360 }, { "epoch": 2.5652096764653693, "grad_norm": 2.8290672302246094, "learning_rate": 3.245625847901206e-07, "loss": 0.2482, "step": 43370 }, { "epoch": 2.5658011474537172, "grad_norm": 2.2031612396240234, "learning_rate": 3.236964827858081e-07, "loss": 0.231, "step": 43380 }, { "epoch": 2.5663926184420656, "grad_norm": 2.5978238582611084, "learning_rate": 3.2283147202719307e-07, "loss": 0.3585, "step": 43390 }, { "epoch": 2.5669840894304135, "grad_norm": 2.978260040283203, "learning_rate": 3.2196755286697743e-07, "loss": 0.2586, "step": 43400 }, { "epoch": 2.5675755604187613, "grad_norm": 6.029501914978027, "learning_rate": 3.2110472565741967e-07, "loss": 0.246, "step": 43410 }, { "epoch": 2.5681670314071097, "grad_norm": 6.123772144317627, "learning_rate": 3.202429907503314e-07, "loss": 0.2507, "step": 43420 }, { "epoch": 2.5687585023954576, "grad_norm": 4.796741962432861, "learning_rate": 3.193823484970805e-07, "loss": 0.254, "step": 43430 }, { "epoch": 2.5693499733838054, "grad_norm": 3.2879045009613037, "learning_rate": 3.185227992485873e-07, "loss": 0.235, "step": 43440 }, { "epoch": 2.5699414443721533, "grad_norm": 4.384537696838379, "learning_rate": 3.1766434335532813e-07, "loss": 0.2588, "step": 43450 }, { "epoch": 2.5705329153605017, "grad_norm": 4.067159175872803, "learning_rate": 3.16806981167333e-07, "loss": 0.2438, "step": 43460 }, { "epoch": 2.5711243863488495, "grad_norm": 2.5319347381591797, "learning_rate": 3.159507130341852e-07, "loss": 0.2119, "step": 43470 }, { "epoch": 2.571715857337198, "grad_norm": 2.737128257751465, "learning_rate": 3.150955393050235e-07, "loss": 0.2178, "step": 43480 }, { "epoch": 2.5723073283255458, "grad_norm": 3.9195072650909424, "learning_rate": 3.142414603285384e-07, "loss": 0.3628, "step": 43490 }, { "epoch": 2.5728987993138936, "grad_norm": 3.9120264053344727, "learning_rate": 3.1338847645297655e-07, "loss": 0.2125, "step": 43500 }, { "epoch": 2.5734902703022415, "grad_norm": 3.454498291015625, "learning_rate": 3.125365880261357e-07, "loss": 0.246, "step": 43510 }, { "epoch": 2.5740817412905894, "grad_norm": 3.9292876720428467, "learning_rate": 3.1168579539536777e-07, "loss": 0.2753, "step": 43520 }, { "epoch": 2.5746732122789378, "grad_norm": 2.560147285461426, "learning_rate": 3.108360989075789e-07, "loss": 0.2639, "step": 43530 }, { "epoch": 2.5752646832672856, "grad_norm": 3.326657772064209, "learning_rate": 3.099874989092266e-07, "loss": 0.2571, "step": 43540 }, { "epoch": 2.575856154255634, "grad_norm": 2.847184419631958, "learning_rate": 3.091399957463231e-07, "loss": 0.2674, "step": 43550 }, { "epoch": 2.576447625243982, "grad_norm": 2.374274253845215, "learning_rate": 3.082935897644317e-07, "loss": 0.2465, "step": 43560 }, { "epoch": 2.5770390962323297, "grad_norm": 2.7023329734802246, "learning_rate": 3.074482813086685e-07, "loss": 0.2613, "step": 43570 }, { "epoch": 2.5776305672206776, "grad_norm": 2.686950206756592, "learning_rate": 3.06604070723704e-07, "loss": 0.3197, "step": 43580 }, { "epoch": 2.578222038209026, "grad_norm": 4.625484943389893, "learning_rate": 3.0576095835375874e-07, "loss": 0.2572, "step": 43590 }, { "epoch": 2.578813509197374, "grad_norm": 3.1710703372955322, "learning_rate": 3.049189445426067e-07, "loss": 0.2598, "step": 43600 }, { "epoch": 2.5794049801857217, "grad_norm": 2.24458646774292, "learning_rate": 3.04078029633573e-07, "loss": 0.252, "step": 43610 }, { "epoch": 2.57999645117407, "grad_norm": 7.17633581161499, "learning_rate": 3.0323821396953634e-07, "loss": 0.2384, "step": 43620 }, { "epoch": 2.580587922162418, "grad_norm": 4.985254764556885, "learning_rate": 3.0239949789292544e-07, "loss": 0.2365, "step": 43630 }, { "epoch": 2.581179393150766, "grad_norm": 4.248702049255371, "learning_rate": 3.015618817457221e-07, "loss": 0.2721, "step": 43640 }, { "epoch": 2.5817708641391137, "grad_norm": 3.447258234024048, "learning_rate": 3.0072536586945877e-07, "loss": 0.2489, "step": 43650 }, { "epoch": 2.582362335127462, "grad_norm": 2.812027931213379, "learning_rate": 2.9988995060521864e-07, "loss": 0.235, "step": 43660 }, { "epoch": 2.58295380611581, "grad_norm": 4.061874866485596, "learning_rate": 2.990556362936385e-07, "loss": 0.2483, "step": 43670 }, { "epoch": 2.5835452771041583, "grad_norm": 2.139474630355835, "learning_rate": 2.98222423274903e-07, "loss": 0.2913, "step": 43680 }, { "epoch": 2.584136748092506, "grad_norm": 3.028388738632202, "learning_rate": 2.9739031188875065e-07, "loss": 0.2678, "step": 43690 }, { "epoch": 2.584728219080854, "grad_norm": 4.462337493896484, "learning_rate": 2.965593024744694e-07, "loss": 0.2541, "step": 43700 }, { "epoch": 2.585319690069202, "grad_norm": 5.911144733428955, "learning_rate": 2.957293953708975e-07, "loss": 0.2821, "step": 43710 }, { "epoch": 2.5859111610575503, "grad_norm": 4.102914333343506, "learning_rate": 2.9490059091642464e-07, "loss": 0.2452, "step": 43720 }, { "epoch": 2.586502632045898, "grad_norm": 3.815560817718506, "learning_rate": 2.9407288944899044e-07, "loss": 0.2908, "step": 43730 }, { "epoch": 2.587094103034246, "grad_norm": 3.6812942028045654, "learning_rate": 2.9324629130608504e-07, "loss": 0.2601, "step": 43740 }, { "epoch": 2.5876855740225944, "grad_norm": 3.95977783203125, "learning_rate": 2.924207968247484e-07, "loss": 0.2694, "step": 43750 }, { "epoch": 2.5882770450109422, "grad_norm": 6.202737331390381, "learning_rate": 2.9159640634157104e-07, "loss": 0.2894, "step": 43760 }, { "epoch": 2.58886851599929, "grad_norm": 4.048576831817627, "learning_rate": 2.9077312019269263e-07, "loss": 0.2605, "step": 43770 }, { "epoch": 2.589459986987638, "grad_norm": 3.099811553955078, "learning_rate": 2.899509387138024e-07, "loss": 0.2555, "step": 43780 }, { "epoch": 2.5900514579759863, "grad_norm": 3.247433662414551, "learning_rate": 2.891298622401408e-07, "loss": 0.2846, "step": 43790 }, { "epoch": 2.5906429289643342, "grad_norm": 3.8579418659210205, "learning_rate": 2.8830989110649506e-07, "loss": 0.2916, "step": 43800 }, { "epoch": 2.5912343999526826, "grad_norm": 3.3244640827178955, "learning_rate": 2.874910256472041e-07, "loss": 0.2675, "step": 43810 }, { "epoch": 2.5918258709410305, "grad_norm": 2.262117862701416, "learning_rate": 2.866732661961544e-07, "loss": 0.2478, "step": 43820 }, { "epoch": 2.5924173419293783, "grad_norm": 3.4460232257843018, "learning_rate": 2.8585661308678333e-07, "loss": 0.3535, "step": 43830 }, { "epoch": 2.5930088129177262, "grad_norm": 3.53004789352417, "learning_rate": 2.850410666520752e-07, "loss": 0.2698, "step": 43840 }, { "epoch": 2.5936002839060746, "grad_norm": 2.517448663711548, "learning_rate": 2.842266272245636e-07, "loss": 0.2691, "step": 43850 }, { "epoch": 2.5941917548944224, "grad_norm": 5.605027675628662, "learning_rate": 2.834132951363316e-07, "loss": 0.2948, "step": 43860 }, { "epoch": 2.5947832258827703, "grad_norm": 2.774348735809326, "learning_rate": 2.8260107071900965e-07, "loss": 0.2585, "step": 43870 }, { "epoch": 2.5953746968711187, "grad_norm": 4.83784818649292, "learning_rate": 2.8178995430377775e-07, "loss": 0.272, "step": 43880 }, { "epoch": 2.5959661678594665, "grad_norm": 6.22749662399292, "learning_rate": 2.8097994622136267e-07, "loss": 0.2778, "step": 43890 }, { "epoch": 2.5965576388478144, "grad_norm": 4.239101886749268, "learning_rate": 2.80171046802041e-07, "loss": 0.2599, "step": 43900 }, { "epoch": 2.5971491098361623, "grad_norm": 3.123016595840454, "learning_rate": 2.7936325637563544e-07, "loss": 0.2312, "step": 43910 }, { "epoch": 2.5977405808245106, "grad_norm": 2.969872236251831, "learning_rate": 2.7855657527151757e-07, "loss": 0.2841, "step": 43920 }, { "epoch": 2.5983320518128585, "grad_norm": 4.279462814331055, "learning_rate": 2.7775100381860694e-07, "loss": 0.2621, "step": 43930 }, { "epoch": 2.598923522801207, "grad_norm": 4.6590046882629395, "learning_rate": 2.7694654234536954e-07, "loss": 0.261, "step": 43940 }, { "epoch": 2.5995149937895548, "grad_norm": 5.411220073699951, "learning_rate": 2.7614319117982024e-07, "loss": 0.2866, "step": 43950 }, { "epoch": 2.6001064647779026, "grad_norm": 5.0671844482421875, "learning_rate": 2.753409506495197e-07, "loss": 0.2742, "step": 43960 }, { "epoch": 2.6006979357662505, "grad_norm": 4.651954650878906, "learning_rate": 2.7453982108157595e-07, "loss": 0.2141, "step": 43970 }, { "epoch": 2.6012894067545984, "grad_norm": 3.748746156692505, "learning_rate": 2.7373980280264533e-07, "loss": 0.2949, "step": 43980 }, { "epoch": 2.6018808777429467, "grad_norm": 4.419342517852783, "learning_rate": 2.729408961389296e-07, "loss": 0.2728, "step": 43990 }, { "epoch": 2.6024723487312946, "grad_norm": 4.834417819976807, "learning_rate": 2.721431014161788e-07, "loss": 0.2546, "step": 44000 }, { "epoch": 2.603063819719643, "grad_norm": 2.8409459590911865, "learning_rate": 2.7134641895968783e-07, "loss": 0.2875, "step": 44010 }, { "epoch": 2.603655290707991, "grad_norm": 4.522036552429199, "learning_rate": 2.705508490942994e-07, "loss": 0.2464, "step": 44020 }, { "epoch": 2.6042467616963387, "grad_norm": 4.687526226043701, "learning_rate": 2.6975639214440137e-07, "loss": 0.26, "step": 44030 }, { "epoch": 2.6048382326846866, "grad_norm": 4.083695411682129, "learning_rate": 2.689630484339284e-07, "loss": 0.2379, "step": 44040 }, { "epoch": 2.605429703673035, "grad_norm": 3.5442006587982178, "learning_rate": 2.6817081828636194e-07, "loss": 0.2877, "step": 44050 }, { "epoch": 2.606021174661383, "grad_norm": 5.782157897949219, "learning_rate": 2.6737970202472826e-07, "loss": 0.2716, "step": 44060 }, { "epoch": 2.6066126456497307, "grad_norm": 4.074464797973633, "learning_rate": 2.665896999716011e-07, "loss": 0.229, "step": 44070 }, { "epoch": 2.607204116638079, "grad_norm": 3.6927013397216797, "learning_rate": 2.658008124490974e-07, "loss": 0.2444, "step": 44080 }, { "epoch": 2.607795587626427, "grad_norm": 3.2187113761901855, "learning_rate": 2.650130397788818e-07, "loss": 0.2199, "step": 44090 }, { "epoch": 2.608387058614775, "grad_norm": 2.1298961639404297, "learning_rate": 2.6422638228216344e-07, "loss": 0.2519, "step": 44100 }, { "epoch": 2.6089785296031227, "grad_norm": 2.830772638320923, "learning_rate": 2.634408402796964e-07, "loss": 0.1909, "step": 44110 }, { "epoch": 2.609570000591471, "grad_norm": 4.611769676208496, "learning_rate": 2.6265641409178143e-07, "loss": 0.2419, "step": 44120 }, { "epoch": 2.610161471579819, "grad_norm": 3.697809934616089, "learning_rate": 2.6187310403826237e-07, "loss": 0.2244, "step": 44130 }, { "epoch": 2.6107529425681673, "grad_norm": 5.137073516845703, "learning_rate": 2.610909104385294e-07, "loss": 0.2649, "step": 44140 }, { "epoch": 2.611344413556515, "grad_norm": 5.111652374267578, "learning_rate": 2.6030983361151703e-07, "loss": 0.2642, "step": 44150 }, { "epoch": 2.611935884544863, "grad_norm": 5.793638229370117, "learning_rate": 2.595298738757036e-07, "loss": 0.2548, "step": 44160 }, { "epoch": 2.612527355533211, "grad_norm": 3.402808427810669, "learning_rate": 2.5875103154911393e-07, "loss": 0.2607, "step": 44170 }, { "epoch": 2.6131188265215592, "grad_norm": 4.984960556030273, "learning_rate": 2.5797330694931477e-07, "loss": 0.2567, "step": 44180 }, { "epoch": 2.613710297509907, "grad_norm": 2.7546041011810303, "learning_rate": 2.571967003934194e-07, "loss": 0.2588, "step": 44190 }, { "epoch": 2.614301768498255, "grad_norm": 3.259373188018799, "learning_rate": 2.564212121980831e-07, "loss": 0.2589, "step": 44200 }, { "epoch": 2.6148932394866033, "grad_norm": 4.311593532562256, "learning_rate": 2.556468426795078e-07, "loss": 0.2204, "step": 44210 }, { "epoch": 2.6154847104749512, "grad_norm": 5.045707702636719, "learning_rate": 2.5487359215343643e-07, "loss": 0.2716, "step": 44220 }, { "epoch": 2.616076181463299, "grad_norm": 6.881606578826904, "learning_rate": 2.541014609351575e-07, "loss": 0.2806, "step": 44230 }, { "epoch": 2.616667652451647, "grad_norm": 5.606700420379639, "learning_rate": 2.5333044933950255e-07, "loss": 0.2567, "step": 44240 }, { "epoch": 2.6172591234399953, "grad_norm": 2.9487905502319336, "learning_rate": 2.525605576808461e-07, "loss": 0.2481, "step": 44250 }, { "epoch": 2.6178505944283432, "grad_norm": 3.6016781330108643, "learning_rate": 2.5179178627310727e-07, "loss": 0.301, "step": 44260 }, { "epoch": 2.6184420654166916, "grad_norm": 1.584152102470398, "learning_rate": 2.5102413542974714e-07, "loss": 0.2553, "step": 44270 }, { "epoch": 2.6190335364050394, "grad_norm": 3.082733631134033, "learning_rate": 2.502576054637712e-07, "loss": 0.2938, "step": 44280 }, { "epoch": 2.6196250073933873, "grad_norm": 2.6137619018554688, "learning_rate": 2.494921966877265e-07, "loss": 0.2362, "step": 44290 }, { "epoch": 2.620216478381735, "grad_norm": 3.3068535327911377, "learning_rate": 2.487279094137037e-07, "loss": 0.2141, "step": 44300 }, { "epoch": 2.6208079493700835, "grad_norm": 6.637497901916504, "learning_rate": 2.479647439533362e-07, "loss": 0.2447, "step": 44310 }, { "epoch": 2.6213994203584314, "grad_norm": 3.1730916500091553, "learning_rate": 2.4720270061779946e-07, "loss": 0.2825, "step": 44320 }, { "epoch": 2.6219908913467793, "grad_norm": 3.9365811347961426, "learning_rate": 2.464417797178129e-07, "loss": 0.2419, "step": 44330 }, { "epoch": 2.6225823623351276, "grad_norm": 3.771808624267578, "learning_rate": 2.456819815636363e-07, "loss": 0.2864, "step": 44340 }, { "epoch": 2.6231738333234755, "grad_norm": 3.5315871238708496, "learning_rate": 2.449233064650719e-07, "loss": 0.3058, "step": 44350 }, { "epoch": 2.6237653043118234, "grad_norm": 5.331015586853027, "learning_rate": 2.441657547314662e-07, "loss": 0.2654, "step": 44360 }, { "epoch": 2.6243567753001713, "grad_norm": 5.902303218841553, "learning_rate": 2.4340932667170446e-07, "loss": 0.2525, "step": 44370 }, { "epoch": 2.6249482462885196, "grad_norm": 1.9643657207489014, "learning_rate": 2.4265402259421664e-07, "loss": 0.2451, "step": 44380 }, { "epoch": 2.6255397172768675, "grad_norm": 5.202276229858398, "learning_rate": 2.418998428069724e-07, "loss": 0.246, "step": 44390 }, { "epoch": 2.626131188265216, "grad_norm": 4.488577842712402, "learning_rate": 2.4114678761748443e-07, "loss": 0.2581, "step": 44400 }, { "epoch": 2.6267226592535637, "grad_norm": 3.5549864768981934, "learning_rate": 2.4039485733280574e-07, "loss": 0.1903, "step": 44410 }, { "epoch": 2.6273141302419116, "grad_norm": 6.709818363189697, "learning_rate": 2.3964405225953076e-07, "loss": 0.3028, "step": 44420 }, { "epoch": 2.6279056012302595, "grad_norm": 4.896917819976807, "learning_rate": 2.38894372703797e-07, "loss": 0.2286, "step": 44430 }, { "epoch": 2.6284970722186074, "grad_norm": 3.9423537254333496, "learning_rate": 2.3814581897127918e-07, "loss": 0.2603, "step": 44440 }, { "epoch": 2.6290885432069557, "grad_norm": 4.026264667510986, "learning_rate": 2.373983913671972e-07, "loss": 0.2412, "step": 44450 }, { "epoch": 2.6296800141953036, "grad_norm": 2.723006248474121, "learning_rate": 2.3665209019630862e-07, "loss": 0.198, "step": 44460 }, { "epoch": 2.630271485183652, "grad_norm": 2.921541213989258, "learning_rate": 2.3590691576291435e-07, "loss": 0.2707, "step": 44470 }, { "epoch": 2.630862956172, "grad_norm": 4.012599468231201, "learning_rate": 2.3516286837085334e-07, "loss": 0.2314, "step": 44480 }, { "epoch": 2.6314544271603477, "grad_norm": 4.5963521003723145, "learning_rate": 2.3441994832350655e-07, "loss": 0.2475, "step": 44490 }, { "epoch": 2.6320458981486956, "grad_norm": 4.110540866851807, "learning_rate": 2.3367815592379503e-07, "loss": 0.2494, "step": 44500 }, { "epoch": 2.632637369137044, "grad_norm": 5.249846458435059, "learning_rate": 2.3293749147417952e-07, "loss": 0.2736, "step": 44510 }, { "epoch": 2.633228840125392, "grad_norm": 4.838911533355713, "learning_rate": 2.321979552766621e-07, "loss": 0.3268, "step": 44520 }, { "epoch": 2.6338203111137397, "grad_norm": 4.27505350112915, "learning_rate": 2.3145954763278253e-07, "loss": 0.2404, "step": 44530 }, { "epoch": 2.634411782102088, "grad_norm": 2.845940589904785, "learning_rate": 2.3072226884362336e-07, "loss": 0.252, "step": 44540 }, { "epoch": 2.635003253090436, "grad_norm": 3.070873737335205, "learning_rate": 2.2998611920980473e-07, "loss": 0.2772, "step": 44550 }, { "epoch": 2.635594724078784, "grad_norm": 4.578070163726807, "learning_rate": 2.2925109903148623e-07, "loss": 0.2453, "step": 44560 }, { "epoch": 2.6361861950671317, "grad_norm": 5.063473224639893, "learning_rate": 2.2851720860836878e-07, "loss": 0.2562, "step": 44570 }, { "epoch": 2.63677766605548, "grad_norm": 4.834127426147461, "learning_rate": 2.2778444823969035e-07, "loss": 0.2549, "step": 44580 }, { "epoch": 2.637369137043828, "grad_norm": 3.573396921157837, "learning_rate": 2.2705281822423063e-07, "loss": 0.2463, "step": 44590 }, { "epoch": 2.6379606080321762, "grad_norm": 1.859602689743042, "learning_rate": 2.263223188603063e-07, "loss": 0.2084, "step": 44600 }, { "epoch": 2.638552079020524, "grad_norm": 5.057963848114014, "learning_rate": 2.2559295044577343e-07, "loss": 0.2747, "step": 44610 }, { "epoch": 2.639143550008872, "grad_norm": 3.316497325897217, "learning_rate": 2.2486471327802815e-07, "loss": 0.2436, "step": 44620 }, { "epoch": 2.63973502099722, "grad_norm": 3.140310287475586, "learning_rate": 2.2413760765400427e-07, "loss": 0.2307, "step": 44630 }, { "epoch": 2.6403264919855682, "grad_norm": 4.016728401184082, "learning_rate": 2.2341163387017427e-07, "loss": 0.2451, "step": 44640 }, { "epoch": 2.640917962973916, "grad_norm": 3.1732141971588135, "learning_rate": 2.2268679222255007e-07, "loss": 0.274, "step": 44650 }, { "epoch": 2.641509433962264, "grad_norm": 5.316905498504639, "learning_rate": 2.2196308300668023e-07, "loss": 0.2691, "step": 44660 }, { "epoch": 2.6421009049506123, "grad_norm": 3.5040924549102783, "learning_rate": 2.212405065176527e-07, "loss": 0.291, "step": 44670 }, { "epoch": 2.6426923759389602, "grad_norm": 5.473266124725342, "learning_rate": 2.2051906305009417e-07, "loss": 0.2466, "step": 44680 }, { "epoch": 2.643283846927308, "grad_norm": 3.0526630878448486, "learning_rate": 2.1979875289816832e-07, "loss": 0.2686, "step": 44690 }, { "epoch": 2.643875317915656, "grad_norm": 4.202913761138916, "learning_rate": 2.1907957635557617e-07, "loss": 0.2588, "step": 44700 }, { "epoch": 2.6444667889040043, "grad_norm": 2.84163236618042, "learning_rate": 2.1836153371555856e-07, "loss": 0.2074, "step": 44710 }, { "epoch": 2.645058259892352, "grad_norm": 2.733293294906616, "learning_rate": 2.1764462527089191e-07, "loss": 0.2048, "step": 44720 }, { "epoch": 2.6456497308807005, "grad_norm": 3.9113991260528564, "learning_rate": 2.1692885131389173e-07, "loss": 0.2412, "step": 44730 }, { "epoch": 2.6462412018690484, "grad_norm": 3.462825298309326, "learning_rate": 2.1621421213641024e-07, "loss": 0.2808, "step": 44740 }, { "epoch": 2.6468326728573963, "grad_norm": 4.181183815002441, "learning_rate": 2.1550070802983636e-07, "loss": 0.2555, "step": 44750 }, { "epoch": 2.647424143845744, "grad_norm": 6.072109699249268, "learning_rate": 2.1478833928509767e-07, "loss": 0.2264, "step": 44760 }, { "epoch": 2.6480156148340925, "grad_norm": 6.231730937957764, "learning_rate": 2.140771061926572e-07, "loss": 0.3586, "step": 44770 }, { "epoch": 2.6486070858224404, "grad_norm": 5.26189661026001, "learning_rate": 2.1336700904251628e-07, "loss": 0.2256, "step": 44780 }, { "epoch": 2.6491985568107883, "grad_norm": 4.036810874938965, "learning_rate": 2.1265804812421263e-07, "loss": 0.2295, "step": 44790 }, { "epoch": 2.6497900277991366, "grad_norm": 3.672102928161621, "learning_rate": 2.119502237268197e-07, "loss": 0.2703, "step": 44800 }, { "epoch": 2.6503814987874845, "grad_norm": 3.82915997505188, "learning_rate": 2.1124353613894932e-07, "loss": 0.2365, "step": 44810 }, { "epoch": 2.6509729697758324, "grad_norm": 4.0126166343688965, "learning_rate": 2.1053798564874793e-07, "loss": 0.2609, "step": 44820 }, { "epoch": 2.6515644407641803, "grad_norm": 4.35971736907959, "learning_rate": 2.098335725439001e-07, "loss": 0.222, "step": 44830 }, { "epoch": 2.6521559117525286, "grad_norm": 4.292269706726074, "learning_rate": 2.091302971116251e-07, "loss": 0.2521, "step": 44840 }, { "epoch": 2.6527473827408765, "grad_norm": 7.066451549530029, "learning_rate": 2.084281596386799e-07, "loss": 0.2657, "step": 44850 }, { "epoch": 2.653338853729225, "grad_norm": 5.290751934051514, "learning_rate": 2.077271604113561e-07, "loss": 0.2597, "step": 44860 }, { "epoch": 2.6539303247175727, "grad_norm": 4.759531021118164, "learning_rate": 2.070272997154814e-07, "loss": 0.2382, "step": 44870 }, { "epoch": 2.6545217957059206, "grad_norm": 5.830311298370361, "learning_rate": 2.0632857783641957e-07, "loss": 0.2799, "step": 44880 }, { "epoch": 2.6551132666942685, "grad_norm": 5.650075435638428, "learning_rate": 2.0563099505907002e-07, "loss": 0.232, "step": 44890 }, { "epoch": 2.655704737682617, "grad_norm": 3.0949907302856445, "learning_rate": 2.049345516678679e-07, "loss": 0.2497, "step": 44900 }, { "epoch": 2.6562962086709647, "grad_norm": 4.856547832489014, "learning_rate": 2.042392479467834e-07, "loss": 0.2293, "step": 44910 }, { "epoch": 2.6568876796593126, "grad_norm": 3.8841681480407715, "learning_rate": 2.0354508417932238e-07, "loss": 0.2245, "step": 44920 }, { "epoch": 2.657479150647661, "grad_norm": 4.667036533355713, "learning_rate": 2.028520606485258e-07, "loss": 0.2479, "step": 44930 }, { "epoch": 2.658070621636009, "grad_norm": 5.3721394538879395, "learning_rate": 2.021601776369686e-07, "loss": 0.3013, "step": 44940 }, { "epoch": 2.6586620926243567, "grad_norm": 6.168529987335205, "learning_rate": 2.014694354267631e-07, "loss": 0.3258, "step": 44950 }, { "epoch": 2.6592535636127046, "grad_norm": 3.2870147228240967, "learning_rate": 2.0077983429955371e-07, "loss": 0.3001, "step": 44960 }, { "epoch": 2.659845034601053, "grad_norm": 2.528377056121826, "learning_rate": 2.0009137453652215e-07, "loss": 0.2274, "step": 44970 }, { "epoch": 2.660436505589401, "grad_norm": 4.109161376953125, "learning_rate": 1.9940405641838255e-07, "loss": 0.266, "step": 44980 }, { "epoch": 2.6610279765777487, "grad_norm": 2.013798952102661, "learning_rate": 1.9871788022538507e-07, "loss": 0.2458, "step": 44990 }, { "epoch": 2.661619447566097, "grad_norm": 5.000362873077393, "learning_rate": 1.980328462373132e-07, "loss": 0.2539, "step": 45000 }, { "epoch": 2.662210918554445, "grad_norm": 6.547689437866211, "learning_rate": 1.9734895473348524e-07, "loss": 0.3001, "step": 45010 }, { "epoch": 2.662802389542793, "grad_norm": 4.427182197570801, "learning_rate": 1.9666620599275443e-07, "loss": 0.2638, "step": 45020 }, { "epoch": 2.6633938605311407, "grad_norm": 2.98930287361145, "learning_rate": 1.9598460029350573e-07, "loss": 0.278, "step": 45030 }, { "epoch": 2.663985331519489, "grad_norm": 3.4594004154205322, "learning_rate": 1.9530413791366086e-07, "loss": 0.3035, "step": 45040 }, { "epoch": 2.664576802507837, "grad_norm": 2.5798635482788086, "learning_rate": 1.9462481913067354e-07, "loss": 0.3017, "step": 45050 }, { "epoch": 2.6651682734961852, "grad_norm": 3.7909088134765625, "learning_rate": 1.9394664422153119e-07, "loss": 0.2387, "step": 45060 }, { "epoch": 2.665759744484533, "grad_norm": 3.5400571823120117, "learning_rate": 1.932696134627563e-07, "loss": 0.2827, "step": 45070 }, { "epoch": 2.666351215472881, "grad_norm": 4.466935157775879, "learning_rate": 1.9259372713040302e-07, "loss": 0.234, "step": 45080 }, { "epoch": 2.666942686461229, "grad_norm": 1.8230594396591187, "learning_rate": 1.9191898550006026e-07, "loss": 0.2696, "step": 45090 }, { "epoch": 2.6675341574495772, "grad_norm": 3.431318759918213, "learning_rate": 1.912453888468486e-07, "loss": 0.2327, "step": 45100 }, { "epoch": 2.668125628437925, "grad_norm": 2.73157000541687, "learning_rate": 1.9057293744542438e-07, "loss": 0.2032, "step": 45110 }, { "epoch": 2.668717099426273, "grad_norm": 3.91200852394104, "learning_rate": 1.8990163156997354e-07, "loss": 0.2413, "step": 45120 }, { "epoch": 2.6693085704146213, "grad_norm": 3.315077781677246, "learning_rate": 1.892314714942185e-07, "loss": 0.2661, "step": 45130 }, { "epoch": 2.669900041402969, "grad_norm": 3.618009567260742, "learning_rate": 1.8856245749141165e-07, "loss": 0.2873, "step": 45140 }, { "epoch": 2.670491512391317, "grad_norm": 2.395414113998413, "learning_rate": 1.878945898343395e-07, "loss": 0.2122, "step": 45150 }, { "epoch": 2.671082983379665, "grad_norm": 4.191140174865723, "learning_rate": 1.8722786879532083e-07, "loss": 0.2996, "step": 45160 }, { "epoch": 2.6716744543680133, "grad_norm": 2.08585524559021, "learning_rate": 1.865622946462069e-07, "loss": 0.2218, "step": 45170 }, { "epoch": 2.672265925356361, "grad_norm": 5.027773857116699, "learning_rate": 1.8589786765838157e-07, "loss": 0.2813, "step": 45180 }, { "epoch": 2.6728573963447095, "grad_norm": 4.3387017250061035, "learning_rate": 1.852345881027605e-07, "loss": 0.2526, "step": 45190 }, { "epoch": 2.6734488673330574, "grad_norm": 4.273444652557373, "learning_rate": 1.8457245624979103e-07, "loss": 0.2903, "step": 45200 }, { "epoch": 2.6740403383214053, "grad_norm": 2.931481122970581, "learning_rate": 1.839114723694545e-07, "loss": 0.2688, "step": 45210 }, { "epoch": 2.674631809309753, "grad_norm": 3.3569629192352295, "learning_rate": 1.8325163673126166e-07, "loss": 0.2617, "step": 45220 }, { "epoch": 2.6752232802981015, "grad_norm": 2.769390344619751, "learning_rate": 1.8259294960425766e-07, "loss": 0.2204, "step": 45230 }, { "epoch": 2.6758147512864494, "grad_norm": 4.40372896194458, "learning_rate": 1.819354112570166e-07, "loss": 0.2686, "step": 45240 }, { "epoch": 2.6764062222747973, "grad_norm": 3.888953924179077, "learning_rate": 1.8127902195764612e-07, "loss": 0.3338, "step": 45250 }, { "epoch": 2.6769976932631456, "grad_norm": 5.831239700317383, "learning_rate": 1.8062378197378504e-07, "loss": 0.2804, "step": 45260 }, { "epoch": 2.6775891642514935, "grad_norm": 2.9289631843566895, "learning_rate": 1.7996969157260235e-07, "loss": 0.3447, "step": 45270 }, { "epoch": 2.6781806352398414, "grad_norm": 3.334141969680786, "learning_rate": 1.793167510208007e-07, "loss": 0.285, "step": 45280 }, { "epoch": 2.6787721062281893, "grad_norm": 1.9552618265151978, "learning_rate": 1.7866496058461145e-07, "loss": 0.2226, "step": 45290 }, { "epoch": 2.6793635772165376, "grad_norm": 3.0807981491088867, "learning_rate": 1.7801432052979837e-07, "loss": 0.2646, "step": 45300 }, { "epoch": 2.6799550482048855, "grad_norm": 3.9034245014190674, "learning_rate": 1.7736483112165492e-07, "loss": 0.2516, "step": 45310 }, { "epoch": 2.680546519193234, "grad_norm": 6.174076557159424, "learning_rate": 1.7671649262500755e-07, "loss": 0.2676, "step": 45320 }, { "epoch": 2.6811379901815817, "grad_norm": 2.389742374420166, "learning_rate": 1.760693053042115e-07, "loss": 0.2398, "step": 45330 }, { "epoch": 2.6817294611699296, "grad_norm": 2.7416796684265137, "learning_rate": 1.7542326942315266e-07, "loss": 0.2747, "step": 45340 }, { "epoch": 2.6823209321582775, "grad_norm": 2.8825161457061768, "learning_rate": 1.7477838524524936e-07, "loss": 0.2358, "step": 45350 }, { "epoch": 2.682912403146626, "grad_norm": 5.530859470367432, "learning_rate": 1.7413465303344788e-07, "loss": 0.2786, "step": 45360 }, { "epoch": 2.6835038741349737, "grad_norm": 4.79443883895874, "learning_rate": 1.7349207305022697e-07, "loss": 0.2452, "step": 45370 }, { "epoch": 2.6840953451233216, "grad_norm": 4.350762844085693, "learning_rate": 1.72850645557594e-07, "loss": 0.2921, "step": 45380 }, { "epoch": 2.68468681611167, "grad_norm": 5.69695520401001, "learning_rate": 1.7221037081708612e-07, "loss": 0.3009, "step": 45390 }, { "epoch": 2.685278287100018, "grad_norm": 5.852113246917725, "learning_rate": 1.7157124908977284e-07, "loss": 0.3126, "step": 45400 }, { "epoch": 2.6858697580883657, "grad_norm": 2.6636974811553955, "learning_rate": 1.709332806362507e-07, "loss": 0.3241, "step": 45410 }, { "epoch": 2.6864612290767136, "grad_norm": 3.509650468826294, "learning_rate": 1.7029646571664792e-07, "loss": 0.2337, "step": 45420 }, { "epoch": 2.687052700065062, "grad_norm": 5.477014064788818, "learning_rate": 1.6966080459062117e-07, "loss": 0.2827, "step": 45430 }, { "epoch": 2.68764417105341, "grad_norm": 11.121209144592285, "learning_rate": 1.6902629751735842e-07, "loss": 0.3112, "step": 45440 }, { "epoch": 2.6882356420417577, "grad_norm": 3.6565871238708496, "learning_rate": 1.6839294475557475e-07, "loss": 0.2572, "step": 45450 }, { "epoch": 2.688827113030106, "grad_norm": 2.8973357677459717, "learning_rate": 1.6776074656351558e-07, "loss": 0.2862, "step": 45460 }, { "epoch": 2.689418584018454, "grad_norm": 3.777581214904785, "learning_rate": 1.671297031989567e-07, "loss": 0.2357, "step": 45470 }, { "epoch": 2.690010055006802, "grad_norm": 3.5167946815490723, "learning_rate": 1.6649981491920098e-07, "loss": 0.2716, "step": 45480 }, { "epoch": 2.6906015259951497, "grad_norm": 3.623290777206421, "learning_rate": 1.6587108198108226e-07, "loss": 0.2605, "step": 45490 }, { "epoch": 2.691192996983498, "grad_norm": 6.169553756713867, "learning_rate": 1.6524350464096182e-07, "loss": 0.2493, "step": 45500 }, { "epoch": 2.691784467971846, "grad_norm": 3.3098795413970947, "learning_rate": 1.6461708315473067e-07, "loss": 0.2355, "step": 45510 }, { "epoch": 2.6923759389601942, "grad_norm": 5.325136661529541, "learning_rate": 1.6399181777780815e-07, "loss": 0.269, "step": 45520 }, { "epoch": 2.692967409948542, "grad_norm": 3.287424325942993, "learning_rate": 1.6336770876514128e-07, "loss": 0.2927, "step": 45530 }, { "epoch": 2.69355888093689, "grad_norm": 5.615536689758301, "learning_rate": 1.6274475637120788e-07, "loss": 0.2809, "step": 45540 }, { "epoch": 2.694150351925238, "grad_norm": 2.423311948776245, "learning_rate": 1.6212296085001177e-07, "loss": 0.2526, "step": 45550 }, { "epoch": 2.694741822913586, "grad_norm": 3.8062832355499268, "learning_rate": 1.6150232245508678e-07, "loss": 0.2281, "step": 45560 }, { "epoch": 2.695333293901934, "grad_norm": 5.652602672576904, "learning_rate": 1.608828414394945e-07, "loss": 0.2952, "step": 45570 }, { "epoch": 2.695924764890282, "grad_norm": 3.9833972454071045, "learning_rate": 1.6026451805582321e-07, "loss": 0.2621, "step": 45580 }, { "epoch": 2.6965162358786303, "grad_norm": 3.111898183822632, "learning_rate": 1.5964735255619157e-07, "loss": 0.2007, "step": 45590 }, { "epoch": 2.697107706866978, "grad_norm": 3.2126002311706543, "learning_rate": 1.590313451922436e-07, "loss": 0.22, "step": 45600 }, { "epoch": 2.697699177855326, "grad_norm": 4.571125030517578, "learning_rate": 1.5841649621515408e-07, "loss": 0.274, "step": 45610 }, { "epoch": 2.698290648843674, "grad_norm": 5.184834003448486, "learning_rate": 1.5780280587562212e-07, "loss": 0.3342, "step": 45620 }, { "epoch": 2.6988821198320223, "grad_norm": 4.123230457305908, "learning_rate": 1.571902744238769e-07, "loss": 0.3137, "step": 45630 }, { "epoch": 2.69947359082037, "grad_norm": 2.416278839111328, "learning_rate": 1.565789021096743e-07, "loss": 0.2017, "step": 45640 }, { "epoch": 2.7000650618087185, "grad_norm": 4.139973163604736, "learning_rate": 1.559686891822969e-07, "loss": 0.2723, "step": 45650 }, { "epoch": 2.700479091500562, "eval_accuracy": 0.8644408956316332, "eval_loss": 0.3245636820793152, "eval_runtime": 159.2033, "eval_safe_aucpr": 0.9240851391305119, "eval_safe_f1": 0.8496189263503663, "eval_safe_fpr": 0.13470449313921815, "eval_safe_precision": 0.8363002252415898, "eval_safe_recall": 0.8633687131980647, "eval_samples_per_second": 377.593, "eval_steps_per_second": 47.204, "eval_unsafe_aucpr": 0.9566907477401958, "eval_unsafe_f1": 0.8766032193098018, "eval_unsafe_fpr": 0.13663128680193476, "eval_unsafe_precision": 0.8882103841904996, "eval_unsafe_recall": 0.8652955068607815, "step": 45657 }, { "epoch": 2.7006565327970664, "grad_norm": 2.4504244327545166, "learning_rate": 1.5535963589055568e-07, "loss": 0.1866, "step": 45660 }, { "epoch": 2.7012480037854143, "grad_norm": 7.920657634735107, "learning_rate": 1.5475174248278768e-07, "loss": 0.2078, "step": 45670 }, { "epoch": 2.701839474773762, "grad_norm": 2.840508222579956, "learning_rate": 1.5414500920685824e-07, "loss": 0.2705, "step": 45680 }, { "epoch": 2.7024309457621105, "grad_norm": 3.6188178062438965, "learning_rate": 1.5353943631015888e-07, "loss": 0.2551, "step": 45690 }, { "epoch": 2.7030224167504584, "grad_norm": 1.3388850688934326, "learning_rate": 1.5293502403960736e-07, "loss": 0.2591, "step": 45700 }, { "epoch": 2.7036138877388063, "grad_norm": 3.0736324787139893, "learning_rate": 1.5233177264164954e-07, "loss": 0.2806, "step": 45710 }, { "epoch": 2.7042053587271546, "grad_norm": 4.609429836273193, "learning_rate": 1.5172968236225736e-07, "loss": 0.2492, "step": 45720 }, { "epoch": 2.7047968297155025, "grad_norm": 3.5754406452178955, "learning_rate": 1.511287534469291e-07, "loss": 0.285, "step": 45730 }, { "epoch": 2.7053883007038504, "grad_norm": 3.9386234283447266, "learning_rate": 1.5052898614068943e-07, "loss": 0.3031, "step": 45740 }, { "epoch": 2.7059797716921983, "grad_norm": 1.7094557285308838, "learning_rate": 1.4993038068809039e-07, "loss": 0.275, "step": 45750 }, { "epoch": 2.7065712426805466, "grad_norm": 4.086227893829346, "learning_rate": 1.4933293733320875e-07, "loss": 0.2552, "step": 45760 }, { "epoch": 2.7071627136688945, "grad_norm": 3.0530383586883545, "learning_rate": 1.4873665631964894e-07, "loss": 0.213, "step": 45770 }, { "epoch": 2.707754184657243, "grad_norm": 4.555500030517578, "learning_rate": 1.4814153789054018e-07, "loss": 0.3064, "step": 45780 }, { "epoch": 2.7083456556455907, "grad_norm": 3.2705211639404297, "learning_rate": 1.4754758228853805e-07, "loss": 0.2199, "step": 45790 }, { "epoch": 2.7089371266339386, "grad_norm": 3.4241671562194824, "learning_rate": 1.469547897558251e-07, "loss": 0.2332, "step": 45800 }, { "epoch": 2.7095285976222865, "grad_norm": 4.491323471069336, "learning_rate": 1.4636316053410769e-07, "loss": 0.2431, "step": 45810 }, { "epoch": 2.710120068610635, "grad_norm": 4.687567234039307, "learning_rate": 1.4577269486461951e-07, "loss": 0.281, "step": 45820 }, { "epoch": 2.7107115395989827, "grad_norm": 4.525064945220947, "learning_rate": 1.4518339298811934e-07, "loss": 0.2498, "step": 45830 }, { "epoch": 2.7113030105873306, "grad_norm": 7.036793231964111, "learning_rate": 1.4459525514489026e-07, "loss": 0.2272, "step": 45840 }, { "epoch": 2.711894481575679, "grad_norm": 5.354616165161133, "learning_rate": 1.4400828157474276e-07, "loss": 0.2528, "step": 45850 }, { "epoch": 2.712485952564027, "grad_norm": 1.9384233951568604, "learning_rate": 1.434224725170108e-07, "loss": 0.2974, "step": 45860 }, { "epoch": 2.7130774235523747, "grad_norm": 3.5631940364837646, "learning_rate": 1.428378282105549e-07, "loss": 0.2504, "step": 45870 }, { "epoch": 2.7136688945407226, "grad_norm": 9.022805213928223, "learning_rate": 1.422543488937601e-07, "loss": 0.2939, "step": 45880 }, { "epoch": 2.714260365529071, "grad_norm": 4.654193878173828, "learning_rate": 1.416720348045354e-07, "loss": 0.2978, "step": 45890 }, { "epoch": 2.714851836517419, "grad_norm": 3.0619897842407227, "learning_rate": 1.4109088618031684e-07, "loss": 0.2715, "step": 45900 }, { "epoch": 2.7154433075057667, "grad_norm": 2.30802059173584, "learning_rate": 1.4051090325806325e-07, "loss": 0.2339, "step": 45910 }, { "epoch": 2.716034778494115, "grad_norm": 4.471658706665039, "learning_rate": 1.399320862742598e-07, "loss": 0.279, "step": 45920 }, { "epoch": 2.716626249482463, "grad_norm": 3.7004733085632324, "learning_rate": 1.39354435464914e-07, "loss": 0.2268, "step": 45930 }, { "epoch": 2.717217720470811, "grad_norm": 4.506265163421631, "learning_rate": 1.3877795106556147e-07, "loss": 0.2895, "step": 45940 }, { "epoch": 2.7178091914591587, "grad_norm": 5.87864875793457, "learning_rate": 1.3820263331125782e-07, "loss": 0.2453, "step": 45950 }, { "epoch": 2.718400662447507, "grad_norm": 3.019174098968506, "learning_rate": 1.3762848243658643e-07, "loss": 0.2655, "step": 45960 }, { "epoch": 2.718992133435855, "grad_norm": 3.3315727710723877, "learning_rate": 1.370554986756537e-07, "loss": 0.278, "step": 45970 }, { "epoch": 2.719583604424203, "grad_norm": 4.354169845581055, "learning_rate": 1.3648368226208907e-07, "loss": 0.2428, "step": 45980 }, { "epoch": 2.720175075412551, "grad_norm": 3.283424139022827, "learning_rate": 1.3591303342904804e-07, "loss": 0.2707, "step": 45990 }, { "epoch": 2.720766546400899, "grad_norm": 5.71893835067749, "learning_rate": 1.353435524092088e-07, "loss": 0.335, "step": 46000 }, { "epoch": 2.721358017389247, "grad_norm": 4.947049140930176, "learning_rate": 1.3477523943477365e-07, "loss": 0.2789, "step": 46010 }, { "epoch": 2.721949488377595, "grad_norm": 3.0629851818084717, "learning_rate": 1.3420809473746852e-07, "loss": 0.2274, "step": 46020 }, { "epoch": 2.722540959365943, "grad_norm": 2.061479330062866, "learning_rate": 1.3364211854854246e-07, "loss": 0.268, "step": 46030 }, { "epoch": 2.723132430354291, "grad_norm": 2.2547411918640137, "learning_rate": 1.3307731109876985e-07, "loss": 0.3201, "step": 46040 }, { "epoch": 2.7237239013426393, "grad_norm": 3.0115389823913574, "learning_rate": 1.3251367261844617e-07, "loss": 0.2926, "step": 46050 }, { "epoch": 2.724315372330987, "grad_norm": 3.489244222640991, "learning_rate": 1.3195120333739219e-07, "loss": 0.2603, "step": 46060 }, { "epoch": 2.724906843319335, "grad_norm": 3.8022310733795166, "learning_rate": 1.3138990348495083e-07, "loss": 0.3062, "step": 46070 }, { "epoch": 2.725498314307683, "grad_norm": 2.035160779953003, "learning_rate": 1.3082977328998868e-07, "loss": 0.2805, "step": 46080 }, { "epoch": 2.7260897852960313, "grad_norm": 3.149550437927246, "learning_rate": 1.3027081298089538e-07, "loss": 0.2534, "step": 46090 }, { "epoch": 2.726681256284379, "grad_norm": 3.786999225616455, "learning_rate": 1.297130227855826e-07, "loss": 0.291, "step": 46100 }, { "epoch": 2.7272727272727275, "grad_norm": 5.550275802612305, "learning_rate": 1.291564029314868e-07, "loss": 0.2456, "step": 46110 }, { "epoch": 2.7278641982610754, "grad_norm": 1.999353051185608, "learning_rate": 1.2860095364556567e-07, "loss": 0.2622, "step": 46120 }, { "epoch": 2.7284556692494233, "grad_norm": 5.6928229331970215, "learning_rate": 1.2804667515430012e-07, "loss": 0.2309, "step": 46130 }, { "epoch": 2.729047140237771, "grad_norm": 2.86706805229187, "learning_rate": 1.2749356768369403e-07, "loss": 0.2853, "step": 46140 }, { "epoch": 2.7296386112261195, "grad_norm": 3.5665810108184814, "learning_rate": 1.269416314592733e-07, "loss": 0.2846, "step": 46150 }, { "epoch": 2.7302300822144674, "grad_norm": 3.400937557220459, "learning_rate": 1.263908667060859e-07, "loss": 0.1984, "step": 46160 }, { "epoch": 2.7308215532028153, "grad_norm": 3.1500587463378906, "learning_rate": 1.2584127364870292e-07, "loss": 0.2473, "step": 46170 }, { "epoch": 2.7314130241911636, "grad_norm": 3.3622846603393555, "learning_rate": 1.2529285251121768e-07, "loss": 0.2733, "step": 46180 }, { "epoch": 2.7320044951795115, "grad_norm": 3.9610867500305176, "learning_rate": 1.2474560351724508e-07, "loss": 0.266, "step": 46190 }, { "epoch": 2.7325959661678594, "grad_norm": 6.87916374206543, "learning_rate": 1.2419952688992287e-07, "loss": 0.2566, "step": 46200 }, { "epoch": 2.7331874371562073, "grad_norm": 4.183213710784912, "learning_rate": 1.2365462285190933e-07, "loss": 0.2109, "step": 46210 }, { "epoch": 2.7337789081445556, "grad_norm": 5.947887897491455, "learning_rate": 1.2311089162538703e-07, "loss": 0.3185, "step": 46220 }, { "epoch": 2.7343703791329035, "grad_norm": 3.8040337562561035, "learning_rate": 1.2256833343205799e-07, "loss": 0.2349, "step": 46230 }, { "epoch": 2.734961850121252, "grad_norm": 5.191075801849365, "learning_rate": 1.2202694849314656e-07, "loss": 0.2318, "step": 46240 }, { "epoch": 2.7355533211095997, "grad_norm": 3.679682493209839, "learning_rate": 1.2148673702940018e-07, "loss": 0.2521, "step": 46250 }, { "epoch": 2.7361447920979476, "grad_norm": 6.693512916564941, "learning_rate": 1.2094769926108528e-07, "loss": 0.2209, "step": 46260 }, { "epoch": 2.7367362630862955, "grad_norm": 1.5912039279937744, "learning_rate": 1.2040983540799212e-07, "loss": 0.2414, "step": 46270 }, { "epoch": 2.737327734074644, "grad_norm": 2.907270669937134, "learning_rate": 1.1987314568943087e-07, "loss": 0.1915, "step": 46280 }, { "epoch": 2.7379192050629917, "grad_norm": 5.458493232727051, "learning_rate": 1.1933763032423316e-07, "loss": 0.2626, "step": 46290 }, { "epoch": 2.7385106760513396, "grad_norm": 1.6969331502914429, "learning_rate": 1.1880328953075203e-07, "loss": 0.2125, "step": 46300 }, { "epoch": 2.739102147039688, "grad_norm": 3.4726321697235107, "learning_rate": 1.1827012352686183e-07, "loss": 0.2467, "step": 46310 }, { "epoch": 2.739693618028036, "grad_norm": 4.012031078338623, "learning_rate": 1.1773813252995735e-07, "loss": 0.274, "step": 46320 }, { "epoch": 2.7402850890163837, "grad_norm": 4.776124000549316, "learning_rate": 1.1720731675695473e-07, "loss": 0.2524, "step": 46330 }, { "epoch": 2.7408765600047316, "grad_norm": 3.320279121398926, "learning_rate": 1.1667767642429017e-07, "loss": 0.2335, "step": 46340 }, { "epoch": 2.74146803099308, "grad_norm": 4.133067607879639, "learning_rate": 1.1614921174792193e-07, "loss": 0.2676, "step": 46350 }, { "epoch": 2.742059501981428, "grad_norm": 4.979892730712891, "learning_rate": 1.156219229433273e-07, "loss": 0.2633, "step": 46360 }, { "epoch": 2.7426509729697757, "grad_norm": 6.017746448516846, "learning_rate": 1.1509581022550596e-07, "loss": 0.2438, "step": 46370 }, { "epoch": 2.743242443958124, "grad_norm": 3.406108856201172, "learning_rate": 1.1457087380897568e-07, "loss": 0.2652, "step": 46380 }, { "epoch": 2.743833914946472, "grad_norm": 3.6981794834136963, "learning_rate": 1.1404711390777723e-07, "loss": 0.2673, "step": 46390 }, { "epoch": 2.7444253859348198, "grad_norm": 4.575705528259277, "learning_rate": 1.1352453073546943e-07, "loss": 0.3239, "step": 46400 }, { "epoch": 2.7450168569231677, "grad_norm": 7.1377949714660645, "learning_rate": 1.1300312450513283e-07, "loss": 0.2651, "step": 46410 }, { "epoch": 2.745608327911516, "grad_norm": 2.5254080295562744, "learning_rate": 1.124828954293674e-07, "loss": 0.2705, "step": 46420 }, { "epoch": 2.746199798899864, "grad_norm": 4.114309310913086, "learning_rate": 1.1196384372029278e-07, "loss": 0.2952, "step": 46430 }, { "epoch": 2.746791269888212, "grad_norm": 3.0383975505828857, "learning_rate": 1.1144596958954933e-07, "loss": 0.2741, "step": 46440 }, { "epoch": 2.74738274087656, "grad_norm": 3.9882442951202393, "learning_rate": 1.1092927324829683e-07, "loss": 0.301, "step": 46450 }, { "epoch": 2.747974211864908, "grad_norm": 3.1956162452697754, "learning_rate": 1.1041375490721539e-07, "loss": 0.278, "step": 46460 }, { "epoch": 2.748565682853256, "grad_norm": 2.589803695678711, "learning_rate": 1.0989941477650389e-07, "loss": 0.26, "step": 46470 }, { "epoch": 2.749157153841604, "grad_norm": 2.198286294937134, "learning_rate": 1.093862530658809e-07, "loss": 0.2098, "step": 46480 }, { "epoch": 2.749748624829952, "grad_norm": 4.006079196929932, "learning_rate": 1.0887426998458539e-07, "loss": 0.3159, "step": 46490 }, { "epoch": 2.7503400958183, "grad_norm": 4.182543754577637, "learning_rate": 1.08363465741375e-07, "loss": 0.2485, "step": 46500 }, { "epoch": 2.7509315668066483, "grad_norm": 2.5165693759918213, "learning_rate": 1.0785384054452718e-07, "loss": 0.2212, "step": 46510 }, { "epoch": 2.751523037794996, "grad_norm": 3.719153642654419, "learning_rate": 1.0734539460183767e-07, "loss": 0.2275, "step": 46520 }, { "epoch": 2.752114508783344, "grad_norm": 2.8510377407073975, "learning_rate": 1.0683812812062299e-07, "loss": 0.2564, "step": 46530 }, { "epoch": 2.752705979771692, "grad_norm": 4.73987340927124, "learning_rate": 1.0633204130771735e-07, "loss": 0.2661, "step": 46540 }, { "epoch": 2.7532974507600403, "grad_norm": 3.300299644470215, "learning_rate": 1.0582713436947366e-07, "loss": 0.2859, "step": 46550 }, { "epoch": 2.753888921748388, "grad_norm": 3.0253050327301025, "learning_rate": 1.0532340751176594e-07, "loss": 0.2573, "step": 46560 }, { "epoch": 2.7544803927367365, "grad_norm": 3.4387991428375244, "learning_rate": 1.0482086093998455e-07, "loss": 0.2508, "step": 46570 }, { "epoch": 2.7550718637250844, "grad_norm": 2.346327543258667, "learning_rate": 1.0431949485904058e-07, "loss": 0.2783, "step": 46580 }, { "epoch": 2.7556633347134323, "grad_norm": 4.367478370666504, "learning_rate": 1.0381930947336183e-07, "loss": 0.2956, "step": 46590 }, { "epoch": 2.75625480570178, "grad_norm": 10.321924209594727, "learning_rate": 1.0332030498689616e-07, "loss": 0.3329, "step": 46600 }, { "epoch": 2.7568462766901285, "grad_norm": 7.567683219909668, "learning_rate": 1.028224816031098e-07, "loss": 0.2849, "step": 46610 }, { "epoch": 2.7574377476784764, "grad_norm": 6.301678657531738, "learning_rate": 1.0232583952498608e-07, "loss": 0.2152, "step": 46620 }, { "epoch": 2.7580292186668243, "grad_norm": 4.5629425048828125, "learning_rate": 1.0183037895502867e-07, "loss": 0.263, "step": 46630 }, { "epoch": 2.7586206896551726, "grad_norm": 4.684072017669678, "learning_rate": 1.0133610009525795e-07, "loss": 0.3028, "step": 46640 }, { "epoch": 2.7592121606435205, "grad_norm": 3.39241361618042, "learning_rate": 1.008430031472134e-07, "loss": 0.2706, "step": 46650 }, { "epoch": 2.7598036316318684, "grad_norm": 3.954214572906494, "learning_rate": 1.003510883119515e-07, "loss": 0.2757, "step": 46660 }, { "epoch": 2.7603951026202163, "grad_norm": 2.8545358180999756, "learning_rate": 9.986035579004781e-08, "loss": 0.2468, "step": 46670 }, { "epoch": 2.7609865736085646, "grad_norm": 3.4868884086608887, "learning_rate": 9.937080578159563e-08, "loss": 0.2756, "step": 46680 }, { "epoch": 2.7615780445969125, "grad_norm": 3.6846301555633545, "learning_rate": 9.888243848620493e-08, "loss": 0.3077, "step": 46690 }, { "epoch": 2.762169515585261, "grad_norm": 4.584513187408447, "learning_rate": 9.839525410300542e-08, "loss": 0.2671, "step": 46700 }, { "epoch": 2.7627609865736087, "grad_norm": 4.984211444854736, "learning_rate": 9.79092528306429e-08, "loss": 0.2783, "step": 46710 }, { "epoch": 2.7633524575619566, "grad_norm": 4.285290241241455, "learning_rate": 9.742443486728148e-08, "loss": 0.2668, "step": 46720 }, { "epoch": 2.7639439285503045, "grad_norm": 4.839413166046143, "learning_rate": 9.694080041060238e-08, "loss": 0.2713, "step": 46730 }, { "epoch": 2.764535399538653, "grad_norm": 7.886335849761963, "learning_rate": 9.645834965780454e-08, "loss": 0.2436, "step": 46740 }, { "epoch": 2.7651268705270007, "grad_norm": 2.6550040245056152, "learning_rate": 9.597708280560424e-08, "loss": 0.3023, "step": 46750 }, { "epoch": 2.7657183415153486, "grad_norm": 4.213726997375488, "learning_rate": 9.549700005023488e-08, "loss": 0.216, "step": 46760 }, { "epoch": 2.766309812503697, "grad_norm": 2.930363416671753, "learning_rate": 9.501810158744717e-08, "loss": 0.2498, "step": 46770 }, { "epoch": 2.766901283492045, "grad_norm": 4.89617919921875, "learning_rate": 9.454038761250927e-08, "loss": 0.2012, "step": 46780 }, { "epoch": 2.7674927544803927, "grad_norm": 2.884829521179199, "learning_rate": 9.406385832020536e-08, "loss": 0.2698, "step": 46790 }, { "epoch": 2.7680842254687406, "grad_norm": 3.981612205505371, "learning_rate": 9.358851390483802e-08, "loss": 0.2365, "step": 46800 }, { "epoch": 2.768675696457089, "grad_norm": 5.186490058898926, "learning_rate": 9.311435456022487e-08, "loss": 0.2394, "step": 46810 }, { "epoch": 2.7692671674454368, "grad_norm": 6.2554731369018555, "learning_rate": 9.264138047970195e-08, "loss": 0.2495, "step": 46820 }, { "epoch": 2.7698586384337847, "grad_norm": 2.530890941619873, "learning_rate": 9.216959185612128e-08, "loss": 0.2944, "step": 46830 }, { "epoch": 2.770450109422133, "grad_norm": 2.9633100032806396, "learning_rate": 9.169898888185236e-08, "loss": 0.3058, "step": 46840 }, { "epoch": 2.771041580410481, "grad_norm": 7.091713905334473, "learning_rate": 9.122957174877933e-08, "loss": 0.2946, "step": 46850 }, { "epoch": 2.7716330513988288, "grad_norm": 3.6731410026550293, "learning_rate": 9.07613406483051e-08, "loss": 0.2804, "step": 46860 }, { "epoch": 2.7722245223871766, "grad_norm": 2.682908058166504, "learning_rate": 9.02942957713473e-08, "loss": 0.2198, "step": 46870 }, { "epoch": 2.772815993375525, "grad_norm": 3.676765203475952, "learning_rate": 8.982843730834056e-08, "loss": 0.1978, "step": 46880 }, { "epoch": 2.773407464363873, "grad_norm": 4.30279541015625, "learning_rate": 8.93637654492363e-08, "loss": 0.2284, "step": 46890 }, { "epoch": 2.773998935352221, "grad_norm": 5.329656600952148, "learning_rate": 8.890028038350029e-08, "loss": 0.3065, "step": 46900 }, { "epoch": 2.774590406340569, "grad_norm": 4.306096076965332, "learning_rate": 8.843798230011701e-08, "loss": 0.2355, "step": 46910 }, { "epoch": 2.775181877328917, "grad_norm": 2.5274698734283447, "learning_rate": 8.79768713875847e-08, "loss": 0.2419, "step": 46920 }, { "epoch": 2.775773348317265, "grad_norm": 5.7604169845581055, "learning_rate": 8.751694783391861e-08, "loss": 0.2507, "step": 46930 }, { "epoch": 2.776364819305613, "grad_norm": 2.2257113456726074, "learning_rate": 8.705821182664941e-08, "loss": 0.2318, "step": 46940 }, { "epoch": 2.776956290293961, "grad_norm": 2.754539728164673, "learning_rate": 8.660066355282414e-08, "loss": 0.2562, "step": 46950 }, { "epoch": 2.777547761282309, "grad_norm": 4.7821760177612305, "learning_rate": 8.614430319900524e-08, "loss": 0.2833, "step": 46960 }, { "epoch": 2.7781392322706573, "grad_norm": 3.117271900177002, "learning_rate": 8.56891309512705e-08, "loss": 0.2737, "step": 46970 }, { "epoch": 2.778730703259005, "grad_norm": 6.104854583740234, "learning_rate": 8.523514699521318e-08, "loss": 0.3322, "step": 46980 }, { "epoch": 2.779322174247353, "grad_norm": 4.608172416687012, "learning_rate": 8.478235151594316e-08, "loss": 0.2501, "step": 46990 }, { "epoch": 2.779913645235701, "grad_norm": 3.5192599296569824, "learning_rate": 8.433074469808411e-08, "loss": 0.2387, "step": 47000 }, { "epoch": 2.7805051162240493, "grad_norm": 5.065816879272461, "learning_rate": 8.388032672577673e-08, "loss": 0.2482, "step": 47010 }, { "epoch": 2.781096587212397, "grad_norm": 2.813706398010254, "learning_rate": 8.343109778267477e-08, "loss": 0.2026, "step": 47020 }, { "epoch": 2.7816880582007455, "grad_norm": 5.40609884262085, "learning_rate": 8.298305805194972e-08, "loss": 0.2735, "step": 47030 }, { "epoch": 2.7822795291890934, "grad_norm": 3.7408692836761475, "learning_rate": 8.253620771628612e-08, "loss": 0.2512, "step": 47040 }, { "epoch": 2.7828710001774413, "grad_norm": 2.8259804248809814, "learning_rate": 8.209054695788487e-08, "loss": 0.2993, "step": 47050 }, { "epoch": 2.783462471165789, "grad_norm": 4.330702304840088, "learning_rate": 8.164607595846096e-08, "loss": 0.2391, "step": 47060 }, { "epoch": 2.7840539421541375, "grad_norm": 5.134322166442871, "learning_rate": 8.120279489924476e-08, "loss": 0.2451, "step": 47070 }, { "epoch": 2.7846454131424854, "grad_norm": 2.3226969242095947, "learning_rate": 8.076070396098134e-08, "loss": 0.3117, "step": 47080 }, { "epoch": 2.7852368841308333, "grad_norm": 9.823625564575195, "learning_rate": 8.031980332393018e-08, "loss": 0.2909, "step": 47090 }, { "epoch": 2.7858283551191816, "grad_norm": 3.928090810775757, "learning_rate": 7.988009316786616e-08, "loss": 0.2681, "step": 47100 }, { "epoch": 2.7864198261075295, "grad_norm": 4.623139381408691, "learning_rate": 7.944157367207816e-08, "loss": 0.2343, "step": 47110 }, { "epoch": 2.7870112970958774, "grad_norm": 4.773643970489502, "learning_rate": 7.900424501536919e-08, "loss": 0.237, "step": 47120 }, { "epoch": 2.7876027680842252, "grad_norm": 5.02002477645874, "learning_rate": 7.85681073760579e-08, "loss": 0.2477, "step": 47130 }, { "epoch": 2.7881942390725736, "grad_norm": 3.9996728897094727, "learning_rate": 7.813316093197608e-08, "loss": 0.2303, "step": 47140 }, { "epoch": 2.7887857100609215, "grad_norm": 4.032493591308594, "learning_rate": 7.769940586047119e-08, "loss": 0.2103, "step": 47150 }, { "epoch": 2.78937718104927, "grad_norm": 2.531291961669922, "learning_rate": 7.726684233840309e-08, "loss": 0.2452, "step": 47160 }, { "epoch": 2.7899686520376177, "grad_norm": 2.907489061355591, "learning_rate": 7.683547054214734e-08, "loss": 0.3105, "step": 47170 }, { "epoch": 2.7905601230259656, "grad_norm": 6.736056804656982, "learning_rate": 7.640529064759328e-08, "loss": 0.2913, "step": 47180 }, { "epoch": 2.7911515940143135, "grad_norm": 2.9723761081695557, "learning_rate": 7.597630283014323e-08, "loss": 0.1986, "step": 47190 }, { "epoch": 2.791743065002662, "grad_norm": 6.648223876953125, "learning_rate": 7.554850726471496e-08, "loss": 0.2474, "step": 47200 }, { "epoch": 2.7923345359910097, "grad_norm": 1.9831478595733643, "learning_rate": 7.512190412573894e-08, "loss": 0.2234, "step": 47210 }, { "epoch": 2.7929260069793576, "grad_norm": 4.6721906661987305, "learning_rate": 7.469649358716035e-08, "loss": 0.2546, "step": 47220 }, { "epoch": 2.793517477967706, "grad_norm": 2.034217119216919, "learning_rate": 7.427227582243745e-08, "loss": 0.2574, "step": 47230 }, { "epoch": 2.7941089489560538, "grad_norm": 2.920809745788574, "learning_rate": 7.384925100454187e-08, "loss": 0.3051, "step": 47240 }, { "epoch": 2.7947004199444017, "grad_norm": 1.9711273908615112, "learning_rate": 7.342741930595997e-08, "loss": 0.2244, "step": 47250 }, { "epoch": 2.7952918909327495, "grad_norm": 5.780262470245361, "learning_rate": 7.300678089869017e-08, "loss": 0.2455, "step": 47260 }, { "epoch": 2.795883361921098, "grad_norm": 5.922795295715332, "learning_rate": 7.258733595424594e-08, "loss": 0.3587, "step": 47270 }, { "epoch": 2.7964748329094458, "grad_norm": 5.002440452575684, "learning_rate": 7.21690846436528e-08, "loss": 0.3124, "step": 47280 }, { "epoch": 2.7970663038977936, "grad_norm": 4.982954978942871, "learning_rate": 7.175202713744999e-08, "loss": 0.2735, "step": 47290 }, { "epoch": 2.797657774886142, "grad_norm": 3.2688188552856445, "learning_rate": 7.133616360569017e-08, "loss": 0.255, "step": 47300 }, { "epoch": 2.79824924587449, "grad_norm": 4.355690956115723, "learning_rate": 7.092149421793936e-08, "loss": 0.2942, "step": 47310 }, { "epoch": 2.7988407168628378, "grad_norm": 3.478090763092041, "learning_rate": 7.050801914327598e-08, "loss": 0.2608, "step": 47320 }, { "epoch": 2.7994321878511856, "grad_norm": 2.9335737228393555, "learning_rate": 7.00957385502915e-08, "loss": 0.2217, "step": 47330 }, { "epoch": 2.800023658839534, "grad_norm": 4.4626545906066895, "learning_rate": 6.968465260709145e-08, "loss": 0.285, "step": 47340 }, { "epoch": 2.800615129827882, "grad_norm": 3.798767328262329, "learning_rate": 6.927476148129308e-08, "loss": 0.2902, "step": 47350 }, { "epoch": 2.80120660081623, "grad_norm": 3.8039891719818115, "learning_rate": 6.886606534002737e-08, "loss": 0.252, "step": 47360 }, { "epoch": 2.801798071804578, "grad_norm": 3.2709286212921143, "learning_rate": 6.845856434993702e-08, "loss": 0.2499, "step": 47370 }, { "epoch": 2.802389542792926, "grad_norm": 3.581393241882324, "learning_rate": 6.805225867717779e-08, "loss": 0.2026, "step": 47380 }, { "epoch": 2.802981013781274, "grad_norm": 2.508173942565918, "learning_rate": 6.764714848741916e-08, "loss": 0.2367, "step": 47390 }, { "epoch": 2.803572484769622, "grad_norm": 3.976816415786743, "learning_rate": 6.724323394584131e-08, "loss": 0.2847, "step": 47400 }, { "epoch": 2.80416395575797, "grad_norm": 3.521280288696289, "learning_rate": 6.684051521713819e-08, "loss": 0.2498, "step": 47410 }, { "epoch": 2.804755426746318, "grad_norm": 6.9953203201293945, "learning_rate": 6.643899246551577e-08, "loss": 0.3134, "step": 47420 }, { "epoch": 2.8053468977346663, "grad_norm": 2.6335866451263428, "learning_rate": 6.603866585469243e-08, "loss": 0.2661, "step": 47430 }, { "epoch": 2.805938368723014, "grad_norm": 2.7099921703338623, "learning_rate": 6.563953554789858e-08, "loss": 0.2224, "step": 47440 }, { "epoch": 2.806529839711362, "grad_norm": 3.8198046684265137, "learning_rate": 6.52416017078774e-08, "loss": 0.294, "step": 47450 }, { "epoch": 2.80712131069971, "grad_norm": 2.1408045291900635, "learning_rate": 6.484486449688376e-08, "loss": 0.2252, "step": 47460 }, { "epoch": 2.8077127816880583, "grad_norm": 3.7051033973693848, "learning_rate": 6.444932407668457e-08, "loss": 0.2334, "step": 47470 }, { "epoch": 2.808304252676406, "grad_norm": 5.153872489929199, "learning_rate": 6.405498060855886e-08, "loss": 0.2504, "step": 47480 }, { "epoch": 2.8088957236647545, "grad_norm": 4.207669258117676, "learning_rate": 6.366183425329797e-08, "loss": 0.2419, "step": 47490 }, { "epoch": 2.8094871946531024, "grad_norm": 7.79500675201416, "learning_rate": 6.326988517120468e-08, "loss": 0.3268, "step": 47500 }, { "epoch": 2.8100786656414503, "grad_norm": 5.113680839538574, "learning_rate": 6.287913352209385e-08, "loss": 0.253, "step": 47510 }, { "epoch": 2.810670136629798, "grad_norm": 2.809093952178955, "learning_rate": 6.248957946529133e-08, "loss": 0.2373, "step": 47520 }, { "epoch": 2.8112616076181465, "grad_norm": 2.6164252758026123, "learning_rate": 6.21012231596364e-08, "loss": 0.2149, "step": 47530 }, { "epoch": 2.8118530786064944, "grad_norm": 4.521169662475586, "learning_rate": 6.171406476347774e-08, "loss": 0.2709, "step": 47540 }, { "epoch": 2.8124445495948422, "grad_norm": 3.705968141555786, "learning_rate": 6.132810443467773e-08, "loss": 0.2636, "step": 47550 }, { "epoch": 2.8130360205831906, "grad_norm": 5.022098541259766, "learning_rate": 6.09433423306085e-08, "loss": 0.2069, "step": 47560 }, { "epoch": 2.8136274915715385, "grad_norm": 4.214235782623291, "learning_rate": 6.055977860815454e-08, "loss": 0.3313, "step": 47570 }, { "epoch": 2.8142189625598864, "grad_norm": 8.042247772216797, "learning_rate": 6.017741342371174e-08, "loss": 0.2939, "step": 47580 }, { "epoch": 2.8148104335482342, "grad_norm": 6.325862407684326, "learning_rate": 5.979624693318675e-08, "loss": 0.2624, "step": 47590 }, { "epoch": 2.8154019045365826, "grad_norm": 3.089198589324951, "learning_rate": 5.9416279291998224e-08, "loss": 0.2471, "step": 47600 }, { "epoch": 2.8159933755249305, "grad_norm": 3.480525493621826, "learning_rate": 5.90375106550749e-08, "loss": 0.2118, "step": 47610 }, { "epoch": 2.816584846513279, "grad_norm": 3.8028037548065186, "learning_rate": 5.8659941176857913e-08, "loss": 0.2717, "step": 47620 }, { "epoch": 2.8171763175016267, "grad_norm": 5.803852558135986, "learning_rate": 5.828357101129844e-08, "loss": 0.2626, "step": 47630 }, { "epoch": 2.8177677884899746, "grad_norm": 3.345600128173828, "learning_rate": 5.790840031185907e-08, "loss": 0.2569, "step": 47640 }, { "epoch": 2.8183592594783224, "grad_norm": 3.4861133098602295, "learning_rate": 5.753442923151342e-08, "loss": 0.2181, "step": 47650 }, { "epoch": 2.8189507304666708, "grad_norm": 4.942575931549072, "learning_rate": 5.716165792274552e-08, "loss": 0.2912, "step": 47660 }, { "epoch": 2.8195422014550187, "grad_norm": 4.808716297149658, "learning_rate": 5.679008653755113e-08, "loss": 0.2574, "step": 47670 }, { "epoch": 2.8201336724433665, "grad_norm": 3.5134410858154297, "learning_rate": 5.641971522743505e-08, "loss": 0.1713, "step": 47680 }, { "epoch": 2.820725143431715, "grad_norm": 2.5106632709503174, "learning_rate": 5.605054414341449e-08, "loss": 0.2367, "step": 47690 }, { "epoch": 2.8213166144200628, "grad_norm": 2.7739040851593018, "learning_rate": 5.5682573436017015e-08, "loss": 0.2303, "step": 47700 }, { "epoch": 2.8219080854084106, "grad_norm": 4.766023635864258, "learning_rate": 5.5315803255279296e-08, "loss": 0.2327, "step": 47710 }, { "epoch": 2.8224995563967585, "grad_norm": 3.031858444213867, "learning_rate": 5.4950233750750365e-08, "loss": 0.2873, "step": 47720 }, { "epoch": 2.823091027385107, "grad_norm": 2.830396890640259, "learning_rate": 5.4585865071488327e-08, "loss": 0.2489, "step": 47730 }, { "epoch": 2.8236824983734548, "grad_norm": 3.863086223602295, "learning_rate": 5.422269736606267e-08, "loss": 0.2548, "step": 47740 }, { "epoch": 2.8242739693618026, "grad_norm": 4.553227424621582, "learning_rate": 5.386073078255227e-08, "loss": 0.2538, "step": 47750 }, { "epoch": 2.824865440350151, "grad_norm": 2.337498188018799, "learning_rate": 5.349996546854741e-08, "loss": 0.2884, "step": 47760 }, { "epoch": 2.825456911338499, "grad_norm": 2.6622493267059326, "learning_rate": 5.314040157114708e-08, "loss": 0.2093, "step": 47770 }, { "epoch": 2.8260483823268467, "grad_norm": 3.3279104232788086, "learning_rate": 5.278203923696134e-08, "loss": 0.2407, "step": 47780 }, { "epoch": 2.8266398533151946, "grad_norm": 4.1217217445373535, "learning_rate": 5.2424878612110314e-08, "loss": 0.296, "step": 47790 }, { "epoch": 2.827231324303543, "grad_norm": 3.7885632514953613, "learning_rate": 5.206891984222384e-08, "loss": 0.2894, "step": 47800 }, { "epoch": 2.827822795291891, "grad_norm": 3.326192855834961, "learning_rate": 5.171416307244181e-08, "loss": 0.1716, "step": 47810 }, { "epoch": 2.828414266280239, "grad_norm": 5.010356426239014, "learning_rate": 5.136060844741452e-08, "loss": 0.3041, "step": 47820 }, { "epoch": 2.829005737268587, "grad_norm": 2.5019352436065674, "learning_rate": 5.1008256111300663e-08, "loss": 0.258, "step": 47830 }, { "epoch": 2.829597208256935, "grad_norm": 5.3263840675354, "learning_rate": 5.065710620777064e-08, "loss": 0.2945, "step": 47840 }, { "epoch": 2.830188679245283, "grad_norm": 4.045004844665527, "learning_rate": 5.030715888000292e-08, "loss": 0.2278, "step": 47850 }, { "epoch": 2.830780150233631, "grad_norm": 1.7805956602096558, "learning_rate": 4.9958414270686705e-08, "loss": 0.218, "step": 47860 }, { "epoch": 2.831371621221979, "grad_norm": 5.339155197143555, "learning_rate": 4.961087252202024e-08, "loss": 0.2799, "step": 47870 }, { "epoch": 2.831963092210327, "grad_norm": 3.3169655799865723, "learning_rate": 4.9264533775711494e-08, "loss": 0.2566, "step": 47880 }, { "epoch": 2.8325545631986753, "grad_norm": 3.260118246078491, "learning_rate": 4.8919398172977856e-08, "loss": 0.2173, "step": 47890 }, { "epoch": 2.833146034187023, "grad_norm": 5.0732035636901855, "learning_rate": 4.8575465854545755e-08, "loss": 0.2979, "step": 47900 }, { "epoch": 2.833737505175371, "grad_norm": 7.644023895263672, "learning_rate": 4.8232736960652e-08, "loss": 0.3417, "step": 47910 }, { "epoch": 2.834328976163719, "grad_norm": 3.8174028396606445, "learning_rate": 4.789121163104182e-08, "loss": 0.2419, "step": 47920 }, { "epoch": 2.8349204471520673, "grad_norm": 3.13362979888916, "learning_rate": 4.755089000497015e-08, "loss": 0.3017, "step": 47930 }, { "epoch": 2.835511918140415, "grad_norm": 2.944751501083374, "learning_rate": 4.7211772221200655e-08, "loss": 0.301, "step": 47940 }, { "epoch": 2.8361033891287635, "grad_norm": 3.724365711212158, "learning_rate": 4.687385841800673e-08, "loss": 0.3172, "step": 47950 }, { "epoch": 2.8366948601171114, "grad_norm": 5.540832042694092, "learning_rate": 4.653714873317083e-08, "loss": 0.2425, "step": 47960 }, { "epoch": 2.8372863311054592, "grad_norm": 7.233078479766846, "learning_rate": 4.620164330398346e-08, "loss": 0.2979, "step": 47970 }, { "epoch": 2.837877802093807, "grad_norm": 5.335787296295166, "learning_rate": 4.586734226724554e-08, "loss": 0.2571, "step": 47980 }, { "epoch": 2.8384692730821555, "grad_norm": 4.060981750488281, "learning_rate": 4.553424575926568e-08, "loss": 0.2954, "step": 47990 }, { "epoch": 2.8390607440705034, "grad_norm": 3.5804920196533203, "learning_rate": 4.520235391586192e-08, "loss": 0.2382, "step": 48000 }, { "epoch": 2.8396522150588512, "grad_norm": 4.6500654220581055, "learning_rate": 4.487166687236166e-08, "loss": 0.2707, "step": 48010 }, { "epoch": 2.8402436860471996, "grad_norm": 5.176040172576904, "learning_rate": 4.454218476359939e-08, "loss": 0.2401, "step": 48020 }, { "epoch": 2.8408351570355475, "grad_norm": 3.8175621032714844, "learning_rate": 4.421390772392031e-08, "loss": 0.222, "step": 48030 }, { "epoch": 2.8414266280238953, "grad_norm": 2.454313278198242, "learning_rate": 4.388683588717668e-08, "loss": 0.2365, "step": 48040 }, { "epoch": 2.8420180990122432, "grad_norm": 6.1503729820251465, "learning_rate": 4.356096938673015e-08, "loss": 0.2606, "step": 48050 }, { "epoch": 2.8426095700005916, "grad_norm": 5.2874436378479, "learning_rate": 4.323630835545045e-08, "loss": 0.2329, "step": 48060 }, { "epoch": 2.8432010409889394, "grad_norm": 4.666669845581055, "learning_rate": 4.291285292571667e-08, "loss": 0.2843, "step": 48070 }, { "epoch": 2.8437925119772878, "grad_norm": 4.239163398742676, "learning_rate": 4.2590603229415323e-08, "loss": 0.2689, "step": 48080 }, { "epoch": 2.8443839829656357, "grad_norm": 1.4676053524017334, "learning_rate": 4.226955939794197e-08, "loss": 0.2177, "step": 48090 }, { "epoch": 2.8449754539539835, "grad_norm": 3.9211370944976807, "learning_rate": 4.194972156219956e-08, "loss": 0.2678, "step": 48100 }, { "epoch": 2.8455669249423314, "grad_norm": 3.8484745025634766, "learning_rate": 4.1631089852600114e-08, "loss": 0.2508, "step": 48110 }, { "epoch": 2.8461583959306798, "grad_norm": 2.4095206260681152, "learning_rate": 4.1313664399064036e-08, "loss": 0.2514, "step": 48120 }, { "epoch": 2.8467498669190276, "grad_norm": 4.714374542236328, "learning_rate": 4.0997445331019125e-08, "loss": 0.2603, "step": 48130 }, { "epoch": 2.8473413379073755, "grad_norm": 3.3755950927734375, "learning_rate": 4.0682432777402245e-08, "loss": 0.2839, "step": 48140 }, { "epoch": 2.847932808895724, "grad_norm": 8.794307708740234, "learning_rate": 4.036862686665732e-08, "loss": 0.2698, "step": 48150 }, { "epoch": 2.8485242798840718, "grad_norm": 2.866725206375122, "learning_rate": 4.005602772673633e-08, "loss": 0.246, "step": 48160 }, { "epoch": 2.8491157508724196, "grad_norm": 4.089855194091797, "learning_rate": 3.9744635485100655e-08, "loss": 0.2625, "step": 48170 }, { "epoch": 2.8497072218607675, "grad_norm": 4.4511213302612305, "learning_rate": 3.943445026871739e-08, "loss": 0.2348, "step": 48180 }, { "epoch": 2.850298692849116, "grad_norm": 2.4683420658111572, "learning_rate": 3.912547220406337e-08, "loss": 0.26, "step": 48190 }, { "epoch": 2.8508901638374637, "grad_norm": 2.9435653686523438, "learning_rate": 3.8817701417122154e-08, "loss": 0.341, "step": 48200 }, { "epoch": 2.8514816348258116, "grad_norm": 3.4206111431121826, "learning_rate": 3.851113803338502e-08, "loss": 0.2188, "step": 48210 }, { "epoch": 2.85207310581416, "grad_norm": 2.7732670307159424, "learning_rate": 3.820578217785198e-08, "loss": 0.2678, "step": 48220 }, { "epoch": 2.852664576802508, "grad_norm": 2.8875656127929688, "learning_rate": 3.7901633975029105e-08, "loss": 0.2496, "step": 48230 }, { "epoch": 2.8532560477908557, "grad_norm": 4.466434001922607, "learning_rate": 3.759869354893153e-08, "loss": 0.2489, "step": 48240 }, { "epoch": 2.8538475187792036, "grad_norm": 4.119154930114746, "learning_rate": 3.729696102308045e-08, "loss": 0.2668, "step": 48250 }, { "epoch": 2.854438989767552, "grad_norm": 3.581247091293335, "learning_rate": 3.699643652050644e-08, "loss": 0.2531, "step": 48260 }, { "epoch": 2.8550304607559, "grad_norm": 3.651606798171997, "learning_rate": 3.669712016374616e-08, "loss": 0.2349, "step": 48270 }, { "epoch": 2.855621931744248, "grad_norm": 4.57949161529541, "learning_rate": 3.6399012074843306e-08, "loss": 0.3498, "step": 48280 }, { "epoch": 2.856213402732596, "grad_norm": 3.02182936668396, "learning_rate": 3.6102112375349995e-08, "loss": 0.2861, "step": 48290 }, { "epoch": 2.856804873720944, "grad_norm": 4.872827529907227, "learning_rate": 3.580642118632538e-08, "loss": 0.2807, "step": 48300 }, { "epoch": 2.857396344709292, "grad_norm": 2.2758867740631104, "learning_rate": 3.551193862833535e-08, "loss": 0.2459, "step": 48310 }, { "epoch": 2.85798781569764, "grad_norm": 6.352958679199219, "learning_rate": 3.521866482145319e-08, "loss": 0.2148, "step": 48320 }, { "epoch": 2.858579286685988, "grad_norm": 4.713232040405273, "learning_rate": 3.492659988525992e-08, "loss": 0.2201, "step": 48330 }, { "epoch": 2.859170757674336, "grad_norm": 3.572964668273926, "learning_rate": 3.463574393884295e-08, "loss": 0.2947, "step": 48340 }, { "epoch": 2.8597622286626843, "grad_norm": 2.7628166675567627, "learning_rate": 3.4346097100796394e-08, "loss": 0.2378, "step": 48350 }, { "epoch": 2.860353699651032, "grad_norm": 1.500471830368042, "learning_rate": 3.405765948922279e-08, "loss": 0.2057, "step": 48360 }, { "epoch": 2.86094517063938, "grad_norm": 4.91864538192749, "learning_rate": 3.377043122173007e-08, "loss": 0.2263, "step": 48370 }, { "epoch": 2.861536641627728, "grad_norm": 2.100034475326538, "learning_rate": 3.3484412415433876e-08, "loss": 0.296, "step": 48380 }, { "epoch": 2.8621281126160762, "grad_norm": 3.525707483291626, "learning_rate": 3.3199603186956585e-08, "loss": 0.2627, "step": 48390 }, { "epoch": 2.862719583604424, "grad_norm": 3.8177990913391113, "learning_rate": 3.2916003652427637e-08, "loss": 0.2529, "step": 48400 }, { "epoch": 2.8633110545927725, "grad_norm": 4.441494941711426, "learning_rate": 3.263361392748288e-08, "loss": 0.2371, "step": 48410 }, { "epoch": 2.8639025255811204, "grad_norm": 2.6629796028137207, "learning_rate": 3.235243412726418e-08, "loss": 0.1973, "step": 48420 }, { "epoch": 2.8644939965694682, "grad_norm": 0.7687374949455261, "learning_rate": 3.207246436642186e-08, "loss": 0.233, "step": 48430 }, { "epoch": 2.865085467557816, "grad_norm": 3.9647083282470703, "learning_rate": 3.179370475911092e-08, "loss": 0.2434, "step": 48440 }, { "epoch": 2.8656769385461645, "grad_norm": 4.634439945220947, "learning_rate": 3.1516155418994776e-08, "loss": 0.2906, "step": 48450 }, { "epoch": 2.8662684095345123, "grad_norm": 3.1007838249206543, "learning_rate": 3.1239816459241234e-08, "loss": 0.2131, "step": 48460 }, { "epoch": 2.8668598805228602, "grad_norm": 2.35159969329834, "learning_rate": 3.0964687992526495e-08, "loss": 0.281, "step": 48470 }, { "epoch": 2.8674513515112086, "grad_norm": 5.649021148681641, "learning_rate": 3.0690770131032496e-08, "loss": 0.3022, "step": 48480 }, { "epoch": 2.8680428224995564, "grad_norm": 2.434258460998535, "learning_rate": 3.0418062986447226e-08, "loss": 0.2464, "step": 48490 }, { "epoch": 2.8686342934879043, "grad_norm": 4.16245174407959, "learning_rate": 3.014656666996507e-08, "loss": 0.273, "step": 48500 }, { "epoch": 2.869225764476252, "grad_norm": 5.609617233276367, "learning_rate": 2.9876281292287475e-08, "loss": 0.2786, "step": 48510 }, { "epoch": 2.8698172354646005, "grad_norm": 4.105457782745361, "learning_rate": 2.9607206963621293e-08, "loss": 0.2926, "step": 48520 }, { "epoch": 2.8704087064529484, "grad_norm": 4.791525363922119, "learning_rate": 2.933934379367975e-08, "loss": 0.239, "step": 48530 }, { "epoch": 2.8710001774412968, "grad_norm": 3.1950299739837646, "learning_rate": 2.907269189168249e-08, "loss": 0.266, "step": 48540 }, { "epoch": 2.8715916484296446, "grad_norm": 5.3316650390625, "learning_rate": 2.88072513663552e-08, "loss": 0.2341, "step": 48550 }, { "epoch": 2.8721831194179925, "grad_norm": 3.4637057781219482, "learning_rate": 2.8543022325929315e-08, "loss": 0.1997, "step": 48560 }, { "epoch": 2.8727745904063404, "grad_norm": 4.4794206619262695, "learning_rate": 2.8280004878142972e-08, "loss": 0.2517, "step": 48570 }, { "epoch": 2.8733660613946888, "grad_norm": 1.5951039791107178, "learning_rate": 2.8018199130239065e-08, "loss": 0.232, "step": 48580 }, { "epoch": 2.8739575323830366, "grad_norm": 2.6408705711364746, "learning_rate": 2.7757605188968193e-08, "loss": 0.218, "step": 48590 }, { "epoch": 2.8745490033713845, "grad_norm": 3.7295191287994385, "learning_rate": 2.7498223160585367e-08, "loss": 0.2786, "step": 48600 }, { "epoch": 2.875140474359733, "grad_norm": 4.401564598083496, "learning_rate": 2.7240053150851653e-08, "loss": 0.2119, "step": 48610 }, { "epoch": 2.8757319453480807, "grad_norm": 3.9310224056243896, "learning_rate": 2.698309526503484e-08, "loss": 0.3553, "step": 48620 }, { "epoch": 2.8763234163364286, "grad_norm": 2.6296112537384033, "learning_rate": 2.672734960790746e-08, "loss": 0.2751, "step": 48630 }, { "epoch": 2.8769148873247765, "grad_norm": 2.4326608180999756, "learning_rate": 2.6472816283748757e-08, "loss": 0.2707, "step": 48640 }, { "epoch": 2.877506358313125, "grad_norm": 5.087028503417969, "learning_rate": 2.6219495396342386e-08, "loss": 0.2093, "step": 48650 }, { "epoch": 2.8780978293014727, "grad_norm": 3.297581672668457, "learning_rate": 2.5967387048978388e-08, "loss": 0.2888, "step": 48660 }, { "epoch": 2.8786893002898206, "grad_norm": 3.7772138118743896, "learning_rate": 2.571649134445253e-08, "loss": 0.2718, "step": 48670 }, { "epoch": 2.879280771278169, "grad_norm": 2.5672731399536133, "learning_rate": 2.5466808385065986e-08, "loss": 0.2356, "step": 48680 }, { "epoch": 2.879872242266517, "grad_norm": 2.971139430999756, "learning_rate": 2.521833827262532e-08, "loss": 0.2551, "step": 48690 }, { "epoch": 2.8804637132548647, "grad_norm": 3.2585670948028564, "learning_rate": 2.4971081108442818e-08, "loss": 0.2049, "step": 48700 }, { "epoch": 2.8810551842432126, "grad_norm": 4.797037601470947, "learning_rate": 2.4725036993335505e-08, "loss": 0.2033, "step": 48710 }, { "epoch": 2.881646655231561, "grad_norm": 4.131491184234619, "learning_rate": 2.448020602762713e-08, "loss": 0.3326, "step": 48720 }, { "epoch": 2.882238126219909, "grad_norm": 3.851044178009033, "learning_rate": 2.423658831114517e-08, "loss": 0.2066, "step": 48730 }, { "epoch": 2.882829597208257, "grad_norm": 4.600578784942627, "learning_rate": 2.3994183943223503e-08, "loss": 0.2715, "step": 48740 }, { "epoch": 2.883421068196605, "grad_norm": 4.057844638824463, "learning_rate": 2.375299302270073e-08, "loss": 0.233, "step": 48750 }, { "epoch": 2.884012539184953, "grad_norm": 2.9227712154388428, "learning_rate": 2.351301564792152e-08, "loss": 0.3206, "step": 48760 }, { "epoch": 2.884604010173301, "grad_norm": 2.720989942550659, "learning_rate": 2.32742519167346e-08, "loss": 0.283, "step": 48770 }, { "epoch": 2.885195481161649, "grad_norm": 2.6006016731262207, "learning_rate": 2.303670192649443e-08, "loss": 0.1811, "step": 48780 }, { "epoch": 2.885786952149997, "grad_norm": 2.5524327754974365, "learning_rate": 2.280036577406053e-08, "loss": 0.2906, "step": 48790 }, { "epoch": 2.886378423138345, "grad_norm": 3.3799147605895996, "learning_rate": 2.2565243555797475e-08, "loss": 0.2513, "step": 48800 }, { "epoch": 2.8869698941266932, "grad_norm": 4.508274078369141, "learning_rate": 2.2331335367574925e-08, "loss": 0.2409, "step": 48810 }, { "epoch": 2.887561365115041, "grad_norm": 3.2114298343658447, "learning_rate": 2.2098641304767242e-08, "loss": 0.2468, "step": 48820 }, { "epoch": 2.888152836103389, "grad_norm": 2.9786016941070557, "learning_rate": 2.1867161462254204e-08, "loss": 0.2422, "step": 48830 }, { "epoch": 2.888744307091737, "grad_norm": 4.181690692901611, "learning_rate": 2.1636895934420307e-08, "loss": 0.2816, "step": 48840 }, { "epoch": 2.8893357780800852, "grad_norm": 5.489808082580566, "learning_rate": 2.1407844815154454e-08, "loss": 0.1887, "step": 48850 }, { "epoch": 2.889927249068433, "grad_norm": 3.3795948028564453, "learning_rate": 2.1180008197851264e-08, "loss": 0.2567, "step": 48860 }, { "epoch": 2.8905187200567815, "grad_norm": 4.580235481262207, "learning_rate": 2.0953386175409095e-08, "loss": 0.2695, "step": 48870 }, { "epoch": 2.8911101910451293, "grad_norm": 5.584447860717773, "learning_rate": 2.0727978840232032e-08, "loss": 0.2437, "step": 48880 }, { "epoch": 2.8917016620334772, "grad_norm": 3.9861011505126953, "learning_rate": 2.0503786284228553e-08, "loss": 0.2525, "step": 48890 }, { "epoch": 2.892293133021825, "grad_norm": 5.618991851806641, "learning_rate": 2.028080859881154e-08, "loss": 0.2377, "step": 48900 }, { "epoch": 2.8928846040101734, "grad_norm": 5.56109094619751, "learning_rate": 2.0059045874898596e-08, "loss": 0.3064, "step": 48910 }, { "epoch": 2.8934760749985213, "grad_norm": 4.277801036834717, "learning_rate": 1.9838498202912392e-08, "loss": 0.2674, "step": 48920 }, { "epoch": 2.894067545986869, "grad_norm": 4.159215450286865, "learning_rate": 1.961916567277966e-08, "loss": 0.2235, "step": 48930 }, { "epoch": 2.8946590169752175, "grad_norm": 1.8050291538238525, "learning_rate": 1.940104837393186e-08, "loss": 0.2953, "step": 48940 }, { "epoch": 2.8952504879635654, "grad_norm": 2.319861888885498, "learning_rate": 1.918414639530486e-08, "loss": 0.192, "step": 48950 }, { "epoch": 2.8958419589519133, "grad_norm": 3.28787899017334, "learning_rate": 1.8968459825339235e-08, "loss": 0.222, "step": 48960 }, { "epoch": 2.896433429940261, "grad_norm": 4.92340612411499, "learning_rate": 1.8753988751979644e-08, "loss": 0.2771, "step": 48970 }, { "epoch": 2.8970249009286095, "grad_norm": 6.290781497955322, "learning_rate": 1.8540733262675136e-08, "loss": 0.2817, "step": 48980 }, { "epoch": 2.8976163719169574, "grad_norm": 4.010510444641113, "learning_rate": 1.8328693444379484e-08, "loss": 0.3324, "step": 48990 }, { "epoch": 2.8982078429053058, "grad_norm": 2.4777870178222656, "learning_rate": 1.8117869383550867e-08, "loss": 0.266, "step": 49000 }, { "epoch": 2.8987993138936536, "grad_norm": 3.2084295749664307, "learning_rate": 1.790826116615085e-08, "loss": 0.2568, "step": 49010 }, { "epoch": 2.8993907848820015, "grad_norm": 2.725067615509033, "learning_rate": 1.769986887764641e-08, "loss": 0.2403, "step": 49020 }, { "epoch": 2.8999822558703494, "grad_norm": 3.527355194091797, "learning_rate": 1.7492692603007898e-08, "loss": 0.257, "step": 49030 }, { "epoch": 2.9005737268586977, "grad_norm": 3.172919273376465, "learning_rate": 1.7286732426710084e-08, "loss": 0.211, "step": 49040 }, { "epoch": 2.9011651978470456, "grad_norm": 1.828147053718567, "learning_rate": 1.708198843273212e-08, "loss": 0.273, "step": 49050 }, { "epoch": 2.9017566688353935, "grad_norm": 5.276821613311768, "learning_rate": 1.6878460704556564e-08, "loss": 0.2466, "step": 49060 }, { "epoch": 2.902348139823742, "grad_norm": 5.716615200042725, "learning_rate": 1.6676149325171364e-08, "loss": 0.3198, "step": 49070 }, { "epoch": 2.9029396108120897, "grad_norm": 4.612838268280029, "learning_rate": 1.647505437706687e-08, "loss": 0.2285, "step": 49080 }, { "epoch": 2.9035310818004376, "grad_norm": 4.792753219604492, "learning_rate": 1.6275175942238817e-08, "loss": 0.2372, "step": 49090 }, { "epoch": 2.9041225527887855, "grad_norm": 2.002694606781006, "learning_rate": 1.607651410218636e-08, "loss": 0.2343, "step": 49100 }, { "epoch": 2.904714023777134, "grad_norm": 3.265979290008545, "learning_rate": 1.5879068937912023e-08, "loss": 0.2328, "step": 49110 }, { "epoch": 2.9053054947654817, "grad_norm": 5.2703857421875, "learning_rate": 1.5682840529923413e-08, "loss": 0.272, "step": 49120 }, { "epoch": 2.9058969657538296, "grad_norm": 2.5312702655792236, "learning_rate": 1.54878289582312e-08, "loss": 0.2269, "step": 49130 }, { "epoch": 2.906488436742178, "grad_norm": 5.3589630126953125, "learning_rate": 1.52940343023501e-08, "loss": 0.2735, "step": 49140 }, { "epoch": 2.907079907730526, "grad_norm": 2.913470506668091, "learning_rate": 1.5101456641298584e-08, "loss": 0.2489, "step": 49150 }, { "epoch": 2.9076713787188737, "grad_norm": 3.582841157913208, "learning_rate": 1.4910096053599497e-08, "loss": 0.2345, "step": 49160 }, { "epoch": 2.9082628497072216, "grad_norm": 4.425666809082031, "learning_rate": 1.4719952617277765e-08, "loss": 0.2682, "step": 49170 }, { "epoch": 2.90885432069557, "grad_norm": 3.407349109649658, "learning_rate": 1.4531026409864367e-08, "loss": 0.2943, "step": 49180 }, { "epoch": 2.909445791683918, "grad_norm": 2.7980008125305176, "learning_rate": 1.4343317508392017e-08, "loss": 0.258, "step": 49190 }, { "epoch": 2.910037262672266, "grad_norm": 2.708268880844116, "learning_rate": 1.4156825989398158e-08, "loss": 0.2176, "step": 49200 }, { "epoch": 2.910628733660614, "grad_norm": 4.416399002075195, "learning_rate": 1.3971551928923299e-08, "loss": 0.3022, "step": 49210 }, { "epoch": 2.911220204648962, "grad_norm": 4.611632347106934, "learning_rate": 1.3787495402511675e-08, "loss": 0.2438, "step": 49220 }, { "epoch": 2.91181167563731, "grad_norm": 2.6185996532440186, "learning_rate": 1.3604656485211587e-08, "loss": 0.2582, "step": 49230 }, { "epoch": 2.912403146625658, "grad_norm": 3.5791373252868652, "learning_rate": 1.34230352515744e-08, "loss": 0.2452, "step": 49240 }, { "epoch": 2.912994617614006, "grad_norm": 3.464632749557495, "learning_rate": 1.3242631775654546e-08, "loss": 0.2596, "step": 49250 }, { "epoch": 2.913586088602354, "grad_norm": 8.848947525024414, "learning_rate": 1.3063446131010514e-08, "loss": 0.2723, "step": 49260 }, { "epoch": 2.9141775595907022, "grad_norm": 3.2963571548461914, "learning_rate": 1.2885478390704531e-08, "loss": 0.2226, "step": 49270 }, { "epoch": 2.91476903057905, "grad_norm": 3.0662755966186523, "learning_rate": 1.2708728627301213e-08, "loss": 0.2956, "step": 49280 }, { "epoch": 2.915360501567398, "grad_norm": 2.833822727203369, "learning_rate": 1.2533196912869915e-08, "loss": 0.2236, "step": 49290 }, { "epoch": 2.915951972555746, "grad_norm": 3.650702953338623, "learning_rate": 1.2358883318981718e-08, "loss": 0.2648, "step": 49300 }, { "epoch": 2.9165434435440942, "grad_norm": 3.4277637004852295, "learning_rate": 1.2185787916712432e-08, "loss": 0.1973, "step": 49310 }, { "epoch": 2.917134914532442, "grad_norm": 3.437443733215332, "learning_rate": 1.2013910776640602e-08, "loss": 0.2661, "step": 49320 }, { "epoch": 2.9177263855207904, "grad_norm": 6.167128086090088, "learning_rate": 1.1843251968847835e-08, "loss": 0.3707, "step": 49330 }, { "epoch": 2.9183178565091383, "grad_norm": 6.807882785797119, "learning_rate": 1.1673811562919135e-08, "loss": 0.3004, "step": 49340 }, { "epoch": 2.918909327497486, "grad_norm": 4.551494121551514, "learning_rate": 1.1505589627942903e-08, "loss": 0.307, "step": 49350 }, { "epoch": 2.919500798485834, "grad_norm": 3.629563093185425, "learning_rate": 1.1338586232510605e-08, "loss": 0.2207, "step": 49360 }, { "epoch": 2.9200922694741824, "grad_norm": 5.470210075378418, "learning_rate": 1.117280144471644e-08, "loss": 0.2486, "step": 49370 }, { "epoch": 2.9206837404625303, "grad_norm": 4.664397716522217, "learning_rate": 1.1008235332158667e-08, "loss": 0.2432, "step": 49380 }, { "epoch": 2.921275211450878, "grad_norm": 5.797998905181885, "learning_rate": 1.0844887961937611e-08, "loss": 0.2569, "step": 49390 }, { "epoch": 2.9218666824392265, "grad_norm": 3.867750644683838, "learning_rate": 1.0682759400657327e-08, "loss": 0.224, "step": 49400 }, { "epoch": 2.9224581534275744, "grad_norm": 3.42917537689209, "learning_rate": 1.0521849714424602e-08, "loss": 0.2376, "step": 49410 }, { "epoch": 2.9230496244159223, "grad_norm": 4.248804092407227, "learning_rate": 1.036215896884929e-08, "loss": 0.283, "step": 49420 }, { "epoch": 2.92364109540427, "grad_norm": 2.4484753608703613, "learning_rate": 1.0203687229044633e-08, "loss": 0.2571, "step": 49430 }, { "epoch": 2.9242325663926185, "grad_norm": 4.17326545715332, "learning_rate": 1.004643455962595e-08, "loss": 0.2749, "step": 49440 }, { "epoch": 2.9248240373809664, "grad_norm": 6.163310527801514, "learning_rate": 9.89040102471228e-09, "loss": 0.2809, "step": 49450 }, { "epoch": 2.9254155083693147, "grad_norm": 5.07047176361084, "learning_rate": 9.735586687925402e-09, "loss": 0.2316, "step": 49460 }, { "epoch": 2.9260069793576626, "grad_norm": 3.316983938217163, "learning_rate": 9.581991612389817e-09, "loss": 0.2594, "step": 49470 }, { "epoch": 2.9265984503460105, "grad_norm": 1.9770519733428955, "learning_rate": 9.429615860732766e-09, "loss": 0.2525, "step": 49480 }, { "epoch": 2.9271899213343584, "grad_norm": 3.248347043991089, "learning_rate": 9.278459495084545e-09, "loss": 0.2504, "step": 49490 }, { "epoch": 2.9277813923227067, "grad_norm": 3.1970884799957275, "learning_rate": 9.128522577078524e-09, "loss": 0.2551, "step": 49500 }, { "epoch": 2.9283728633110546, "grad_norm": 3.055202007293701, "learning_rate": 8.97980516785013e-09, "loss": 0.2274, "step": 49510 }, { "epoch": 2.9289643342994025, "grad_norm": 6.733698844909668, "learning_rate": 8.832307328038191e-09, "loss": 0.2784, "step": 49520 }, { "epoch": 2.929555805287751, "grad_norm": 3.108008623123169, "learning_rate": 8.6860291177836e-09, "loss": 0.333, "step": 49530 }, { "epoch": 2.9301472762760987, "grad_norm": 5.27102518081665, "learning_rate": 8.54097059673098e-09, "loss": 0.3054, "step": 49540 }, { "epoch": 2.9307387472644466, "grad_norm": 4.97764778137207, "learning_rate": 8.397131824026349e-09, "loss": 0.2853, "step": 49550 }, { "epoch": 2.9313302182527945, "grad_norm": 4.593897342681885, "learning_rate": 8.254512858319795e-09, "loss": 0.2494, "step": 49560 }, { "epoch": 2.931921689241143, "grad_norm": 3.461805820465088, "learning_rate": 8.113113757762802e-09, "loss": 0.2751, "step": 49570 }, { "epoch": 2.9325131602294907, "grad_norm": 6.719745635986328, "learning_rate": 7.972934580009917e-09, "loss": 0.2345, "step": 49580 }, { "epoch": 2.933104631217839, "grad_norm": 3.028684139251709, "learning_rate": 7.833975382218417e-09, "loss": 0.2695, "step": 49590 }, { "epoch": 2.933696102206187, "grad_norm": 5.779026031494141, "learning_rate": 7.69623622104798e-09, "loss": 0.3075, "step": 49600 }, { "epoch": 2.934287573194535, "grad_norm": 3.0350780487060547, "learning_rate": 7.559717152661349e-09, "loss": 0.263, "step": 49610 }, { "epoch": 2.9348790441828827, "grad_norm": 6.06341028213501, "learning_rate": 7.42441823272233e-09, "loss": 0.2965, "step": 49620 }, { "epoch": 2.9354705151712306, "grad_norm": 4.113712310791016, "learning_rate": 7.290339516399125e-09, "loss": 0.2839, "step": 49630 }, { "epoch": 2.936061986159579, "grad_norm": 4.4378461837768555, "learning_rate": 7.157481058361004e-09, "loss": 0.2267, "step": 49640 }, { "epoch": 2.936653457147927, "grad_norm": 3.0468294620513916, "learning_rate": 7.0258429127803e-09, "loss": 0.2188, "step": 49650 }, { "epoch": 2.937244928136275, "grad_norm": 3.7370378971099854, "learning_rate": 6.8954251333317454e-09, "loss": 0.2002, "step": 49660 }, { "epoch": 2.937836399124623, "grad_norm": 3.3617024421691895, "learning_rate": 6.7662277731918024e-09, "loss": 0.2855, "step": 49670 }, { "epoch": 2.938427870112971, "grad_norm": 3.2461414337158203, "learning_rate": 6.638250885040665e-09, "loss": 0.2442, "step": 49680 }, { "epoch": 2.939019341101319, "grad_norm": 3.797248125076294, "learning_rate": 6.511494521059591e-09, "loss": 0.2832, "step": 49690 }, { "epoch": 2.939610812089667, "grad_norm": 5.686367511749268, "learning_rate": 6.385958732932573e-09, "loss": 0.3175, "step": 49700 }, { "epoch": 2.940202283078015, "grad_norm": 4.533515453338623, "learning_rate": 6.2616435718466645e-09, "loss": 0.2777, "step": 49710 }, { "epoch": 2.940793754066363, "grad_norm": 4.502671718597412, "learning_rate": 6.138549088489653e-09, "loss": 0.2332, "step": 49720 }, { "epoch": 2.9413852250547112, "grad_norm": 4.742866516113281, "learning_rate": 6.01667533305339e-09, "loss": 0.223, "step": 49730 }, { "epoch": 2.941976696043059, "grad_norm": 1.8322235345840454, "learning_rate": 5.89602235523079e-09, "loss": 0.2424, "step": 49740 }, { "epoch": 2.942568167031407, "grad_norm": 3.9277637004852295, "learning_rate": 5.776590204217169e-09, "loss": 0.237, "step": 49750 }, { "epoch": 2.943159638019755, "grad_norm": 2.1633999347686768, "learning_rate": 5.65837892871024e-09, "loss": 0.2188, "step": 49760 }, { "epoch": 2.943751109008103, "grad_norm": 3.4798521995544434, "learning_rate": 5.541388576910111e-09, "loss": 0.2286, "step": 49770 }, { "epoch": 2.944342579996451, "grad_norm": 2.738286256790161, "learning_rate": 5.42561919651896e-09, "loss": 0.2741, "step": 49780 }, { "epoch": 2.9449340509847994, "grad_norm": 3.5201451778411865, "learning_rate": 5.311070834740695e-09, "loss": 0.2962, "step": 49790 }, { "epoch": 2.9455255219731473, "grad_norm": 3.570971727371216, "learning_rate": 5.197743538281952e-09, "loss": 0.256, "step": 49800 }, { "epoch": 2.946116992961495, "grad_norm": 3.449470281600952, "learning_rate": 5.085637353350769e-09, "loss": 0.2454, "step": 49810 }, { "epoch": 2.946708463949843, "grad_norm": 6.2066731452941895, "learning_rate": 4.9747523256582496e-09, "loss": 0.248, "step": 49820 }, { "epoch": 2.9472999349381914, "grad_norm": 4.44912576675415, "learning_rate": 4.865088500416892e-09, "loss": 0.2556, "step": 49830 }, { "epoch": 2.9478914059265393, "grad_norm": 5.945096969604492, "learning_rate": 4.756645922341263e-09, "loss": 0.3457, "step": 49840 }, { "epoch": 2.948482876914887, "grad_norm": 5.783085823059082, "learning_rate": 4.649424635647992e-09, "loss": 0.3208, "step": 49850 }, { "epoch": 2.9490743479032355, "grad_norm": 2.8426263332366943, "learning_rate": 4.5434246840564406e-09, "loss": 0.2568, "step": 49860 }, { "epoch": 2.9496658188915834, "grad_norm": 3.4660215377807617, "learning_rate": 4.438646110787037e-09, "loss": 0.2722, "step": 49870 }, { "epoch": 2.9502572898799313, "grad_norm": 4.417478084564209, "learning_rate": 4.3350889585626054e-09, "loss": 0.2402, "step": 49880 }, { "epoch": 2.950848760868279, "grad_norm": 5.763789653778076, "learning_rate": 4.232753269607703e-09, "loss": 0.2892, "step": 49890 }, { "epoch": 2.9514402318566275, "grad_norm": 4.46579647064209, "learning_rate": 4.131639085649619e-09, "loss": 0.267, "step": 49900 }, { "epoch": 2.9520317028449754, "grad_norm": 2.8542819023132324, "learning_rate": 4.031746447916706e-09, "loss": 0.2834, "step": 49910 }, { "epoch": 2.9526231738333237, "grad_norm": 3.5997719764709473, "learning_rate": 3.933075397139385e-09, "loss": 0.2949, "step": 49920 }, { "epoch": 2.9532146448216716, "grad_norm": 6.654110908508301, "learning_rate": 3.835625973550138e-09, "loss": 0.213, "step": 49930 }, { "epoch": 2.9538061158100195, "grad_norm": 4.282567024230957, "learning_rate": 3.739398216883849e-09, "loss": 0.2967, "step": 49940 }, { "epoch": 2.9543975867983674, "grad_norm": 1.7725270986557007, "learning_rate": 3.6443921663761317e-09, "loss": 0.2263, "step": 49950 }, { "epoch": 2.9549890577867157, "grad_norm": 3.3107001781463623, "learning_rate": 3.5506078607656644e-09, "loss": 0.2669, "step": 49960 }, { "epoch": 2.9555805287750636, "grad_norm": 2.29089617729187, "learning_rate": 3.4580453382918596e-09, "loss": 0.2182, "step": 49970 }, { "epoch": 2.9561719997634115, "grad_norm": 3.7116708755493164, "learning_rate": 3.36670463669686e-09, "loss": 0.2071, "step": 49980 }, { "epoch": 2.95676347075176, "grad_norm": 3.182521343231201, "learning_rate": 3.276585793224207e-09, "loss": 0.2536, "step": 49990 }, { "epoch": 2.9573549417401077, "grad_norm": 3.5540640354156494, "learning_rate": 3.1876888446191742e-09, "loss": 0.2272, "step": 50000 }, { "epoch": 2.9579464127284556, "grad_norm": 2.8526535034179688, "learning_rate": 3.1000138271291e-09, "loss": 0.2368, "step": 50010 }, { "epoch": 2.9585378837168035, "grad_norm": 3.9488837718963623, "learning_rate": 3.0135607765027216e-09, "loss": 0.3589, "step": 50020 }, { "epoch": 2.959129354705152, "grad_norm": 2.6230924129486084, "learning_rate": 2.9283297279908418e-09, "loss": 0.2945, "step": 50030 }, { "epoch": 2.9597208256934997, "grad_norm": 3.620832920074463, "learning_rate": 2.8443207163456607e-09, "loss": 0.2396, "step": 50040 }, { "epoch": 2.960312296681848, "grad_norm": 2.9387154579162598, "learning_rate": 2.7615337758217783e-09, "loss": 0.238, "step": 50050 }, { "epoch": 2.960903767670196, "grad_norm": 3.1175551414489746, "learning_rate": 2.6799689401745265e-09, "loss": 0.2441, "step": 50060 }, { "epoch": 2.961495238658544, "grad_norm": 4.812291145324707, "learning_rate": 2.5996262426616346e-09, "loss": 0.2398, "step": 50070 }, { "epoch": 2.9620867096468917, "grad_norm": 3.9079227447509766, "learning_rate": 2.520505716042565e-09, "loss": 0.2549, "step": 50080 }, { "epoch": 2.9626781806352396, "grad_norm": 3.586601972579956, "learning_rate": 2.4426073925778445e-09, "loss": 0.3059, "step": 50090 }, { "epoch": 2.963269651623588, "grad_norm": 2.8848013877868652, "learning_rate": 2.3659313040304e-09, "loss": 0.2271, "step": 50100 }, { "epoch": 2.963861122611936, "grad_norm": 3.77620005607605, "learning_rate": 2.2904774816638884e-09, "loss": 0.2995, "step": 50110 }, { "epoch": 2.964452593600284, "grad_norm": 2.2785184383392334, "learning_rate": 2.2162459562447e-09, "loss": 0.2286, "step": 50120 }, { "epoch": 2.965044064588632, "grad_norm": 2.7883834838867188, "learning_rate": 2.143236758039957e-09, "loss": 0.2168, "step": 50130 }, { "epoch": 2.96563553557698, "grad_norm": 2.992920398712158, "learning_rate": 2.071449916818846e-09, "loss": 0.2577, "step": 50140 }, { "epoch": 2.966227006565328, "grad_norm": 3.154301166534424, "learning_rate": 2.0008854618516204e-09, "loss": 0.2717, "step": 50150 }, { "epoch": 2.966818477553676, "grad_norm": 6.537745475769043, "learning_rate": 1.931543421910931e-09, "loss": 0.2512, "step": 50160 }, { "epoch": 2.967409948542024, "grad_norm": 3.392179489135742, "learning_rate": 1.8634238252701606e-09, "loss": 0.2309, "step": 50170 }, { "epoch": 2.968001419530372, "grad_norm": 3.8439688682556152, "learning_rate": 1.796526699705092e-09, "loss": 0.2727, "step": 50180 }, { "epoch": 2.96859289051872, "grad_norm": 2.7584471702575684, "learning_rate": 1.7308520724922393e-09, "loss": 0.2408, "step": 50190 }, { "epoch": 2.969184361507068, "grad_norm": 1.8631590604782104, "learning_rate": 1.6663999704101817e-09, "loss": 0.2337, "step": 50200 }, { "epoch": 2.969775832495416, "grad_norm": 35.87821578979492, "learning_rate": 1.6031704197388975e-09, "loss": 0.2427, "step": 50210 }, { "epoch": 2.970367303483764, "grad_norm": 4.782816410064697, "learning_rate": 1.5411634462597634e-09, "loss": 0.2922, "step": 50220 }, { "epoch": 2.970958774472112, "grad_norm": 4.868821144104004, "learning_rate": 1.4803790752555557e-09, "loss": 0.3007, "step": 50230 }, { "epoch": 2.97155024546046, "grad_norm": 4.788413047790527, "learning_rate": 1.420817331510782e-09, "loss": 0.243, "step": 50240 }, { "epoch": 2.9721417164488084, "grad_norm": 3.5484254360198975, "learning_rate": 1.362478239311682e-09, "loss": 0.2607, "step": 50250 }, { "epoch": 2.9727331874371563, "grad_norm": 6.093918800354004, "learning_rate": 1.305361822445228e-09, "loss": 0.2176, "step": 50260 }, { "epoch": 2.973324658425504, "grad_norm": 3.1653754711151123, "learning_rate": 1.2494681042007905e-09, "loss": 0.2435, "step": 50270 }, { "epoch": 2.973916129413852, "grad_norm": 5.099849700927734, "learning_rate": 1.1947971073681396e-09, "loss": 0.2926, "step": 50280 }, { "epoch": 2.9745076004022004, "grad_norm": 5.385120868682861, "learning_rate": 1.1413488542391104e-09, "loss": 0.2371, "step": 50290 }, { "epoch": 2.9750990713905483, "grad_norm": 2.698096513748169, "learning_rate": 1.0891233666069367e-09, "loss": 0.2648, "step": 50300 }, { "epoch": 2.975690542378896, "grad_norm": 2.299778461456299, "learning_rate": 1.0381206657665844e-09, "loss": 0.2309, "step": 50310 }, { "epoch": 2.9762820133672445, "grad_norm": 6.18850564956665, "learning_rate": 9.883407725137517e-10, "loss": 0.281, "step": 50320 }, { "epoch": 2.9768734843555924, "grad_norm": 1.9693735837936401, "learning_rate": 9.397837071458692e-10, "loss": 0.2179, "step": 50330 }, { "epoch": 2.9774649553439403, "grad_norm": 3.351940155029297, "learning_rate": 8.924494894617662e-10, "loss": 0.3123, "step": 50340 }, { "epoch": 2.978056426332288, "grad_norm": 2.708603858947754, "learning_rate": 8.463381387620039e-10, "loss": 0.214, "step": 50350 }, { "epoch": 2.9786478973206365, "grad_norm": 3.8800885677337646, "learning_rate": 8.01449673847543e-10, "loss": 0.2525, "step": 50360 }, { "epoch": 2.9792393683089844, "grad_norm": 5.233798503875732, "learning_rate": 7.577841130217422e-10, "loss": 0.2717, "step": 50370 }, { "epoch": 2.9798308392973327, "grad_norm": 2.826780319213867, "learning_rate": 7.153414740893593e-10, "loss": 0.2491, "step": 50380 }, { "epoch": 2.9804223102856806, "grad_norm": 5.171389102935791, "learning_rate": 6.741217743552186e-10, "loss": 0.2542, "step": 50390 }, { "epoch": 2.9810137812740285, "grad_norm": 5.448873043060303, "learning_rate": 6.341250306272084e-10, "loss": 0.2844, "step": 50400 }, { "epoch": 2.9816052522623764, "grad_norm": 4.0103278160095215, "learning_rate": 5.953512592132837e-10, "loss": 0.3131, "step": 50410 }, { "epoch": 2.9821967232507247, "grad_norm": 3.7892439365386963, "learning_rate": 5.578004759234645e-10, "loss": 0.2965, "step": 50420 }, { "epoch": 2.9827881942390726, "grad_norm": 4.210599422454834, "learning_rate": 5.214726960685035e-10, "loss": 0.2463, "step": 50430 }, { "epoch": 2.9833796652274205, "grad_norm": 5.991377830505371, "learning_rate": 4.863679344612182e-10, "loss": 0.3256, "step": 50440 }, { "epoch": 2.983971136215769, "grad_norm": 5.6980366706848145, "learning_rate": 4.52486205414826e-10, "loss": 0.2974, "step": 50450 }, { "epoch": 2.9845626072041167, "grad_norm": 4.8856892585754395, "learning_rate": 4.198275227449422e-10, "loss": 0.3056, "step": 50460 }, { "epoch": 2.9851540781924646, "grad_norm": 3.7360291481018066, "learning_rate": 3.8839189976791476e-10, "loss": 0.29, "step": 50470 }, { "epoch": 2.9857455491808125, "grad_norm": 5.990756034851074, "learning_rate": 3.581793493008245e-10, "loss": 0.2643, "step": 50480 }, { "epoch": 2.986337020169161, "grad_norm": 3.931535243988037, "learning_rate": 3.291898836634832e-10, "loss": 0.2445, "step": 50490 }, { "epoch": 2.9869284911575087, "grad_norm": 3.4403486251831055, "learning_rate": 3.0142351467543627e-10, "loss": 0.2669, "step": 50500 }, { "epoch": 2.987519962145857, "grad_norm": 2.8949670791625977, "learning_rate": 2.7488025365829395e-10, "loss": 0.2285, "step": 50510 }, { "epoch": 2.988111433134205, "grad_norm": 3.0129897594451904, "learning_rate": 2.495601114353985e-10, "loss": 0.2817, "step": 50520 }, { "epoch": 2.988702904122553, "grad_norm": 2.5115346908569336, "learning_rate": 2.254630983308248e-10, "loss": 0.2603, "step": 50530 }, { "epoch": 2.9892943751109007, "grad_norm": 4.246535301208496, "learning_rate": 2.0258922416938052e-10, "loss": 0.2141, "step": 50540 }, { "epoch": 2.9898858460992486, "grad_norm": 2.876964807510376, "learning_rate": 1.8093849827827136e-10, "loss": 0.2413, "step": 50550 }, { "epoch": 2.990477317087597, "grad_norm": 3.9823317527770996, "learning_rate": 1.6051092948543566e-10, "loss": 0.2092, "step": 50560 }, { "epoch": 2.991068788075945, "grad_norm": 2.7155141830444336, "learning_rate": 1.413065261198776e-10, "loss": 0.2486, "step": 50570 }, { "epoch": 2.991660259064293, "grad_norm": 4.5380072593688965, "learning_rate": 1.2332529601200014e-10, "loss": 0.266, "step": 50580 }, { "epoch": 2.992251730052641, "grad_norm": 5.906786918640137, "learning_rate": 1.0656724649360517e-10, "loss": 0.2523, "step": 50590 }, { "epoch": 2.992843201040989, "grad_norm": 3.2416796684265137, "learning_rate": 9.10323843978933e-11, "loss": 0.2766, "step": 50600 }, { "epoch": 2.9934346720293368, "grad_norm": 3.2858381271362305, "learning_rate": 7.672071605879793e-11, "loss": 0.2703, "step": 50610 }, { "epoch": 2.994026143017685, "grad_norm": 3.1679322719573975, "learning_rate": 6.363224731231742e-11, "loss": 0.239, "step": 50620 }, { "epoch": 2.994617614006033, "grad_norm": 2.4633986949920654, "learning_rate": 5.17669834945167e-11, "loss": 0.2579, "step": 50630 }, { "epoch": 2.995209084994381, "grad_norm": 5.359203338623047, "learning_rate": 4.1124929443858704e-11, "loss": 0.288, "step": 50640 }, { "epoch": 2.995800555982729, "grad_norm": 4.129180908203125, "learning_rate": 3.1706089499539125e-11, "loss": 0.2853, "step": 50650 }, { "epoch": 2.996392026971077, "grad_norm": 4.443981647491455, "learning_rate": 2.35104675018194e-11, "loss": 0.2988, "step": 50660 }, { "epoch": 2.996983497959425, "grad_norm": 4.936062335968018, "learning_rate": 1.6538066792359808e-11, "loss": 0.2584, "step": 50670 }, { "epoch": 2.997574968947773, "grad_norm": 3.828183650970459, "learning_rate": 1.078889021421947e-11, "loss": 0.3251, "step": 50680 }, { "epoch": 2.998166439936121, "grad_norm": 3.6964921951293945, "learning_rate": 6.2629401118563525e-12, "loss": 0.2543, "step": 50690 }, { "epoch": 2.998757910924469, "grad_norm": 3.526721239089966, "learning_rate": 2.9602183304611174e-12, "loss": 0.3139, "step": 50700 }, { "epoch": 2.9993493819128174, "grad_norm": 2.554339647293091, "learning_rate": 8.807262166232733e-13, "loss": 0.2887, "step": 50710 }, { "epoch": 2.9999408529011653, "grad_norm": 3.675246238708496, "learning_rate": 2.4464618331165868e-14, "loss": 0.2389, "step": 50720 }, { "epoch": 3.0, "step": 50721, "total_flos": 3.4165989094226784e+17, "train_loss": 0.3003270911840429, "train_runtime": 28686.1001, "train_samples_per_second": 56.58, "train_steps_per_second": 1.768 } ], "logging_steps": 10, "max_steps": 50721, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 5073, "stateful_callbacks": { "EarlyStoppingCallback": { "args": { "early_stopping_patience": 4, "early_stopping_threshold": 0.0 }, "attributes": { "early_stopping_patience_counter": 0 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.4165989094226784e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }