| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9998077107970387, | |
| "eval_steps": 500, | |
| "global_step": 2600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007691568118450149, | |
| "grad_norm": 0.8310319185256958, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 2.3717, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.015383136236900298, | |
| "grad_norm": 0.8349233269691467, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 2.2925, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02307470435535045, | |
| "grad_norm": 0.7424416542053223, | |
| "learning_rate": 4.992229992229993e-05, | |
| "loss": 2.0978, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.030766272473800597, | |
| "grad_norm": 0.7559643387794495, | |
| "learning_rate": 4.972804972804973e-05, | |
| "loss": 1.9267, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03845784059225075, | |
| "grad_norm": 0.6887438297271729, | |
| "learning_rate": 4.9533799533799534e-05, | |
| "loss": 1.7711, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0461494087107009, | |
| "grad_norm": 0.8163133263587952, | |
| "learning_rate": 4.9339549339549344e-05, | |
| "loss": 1.7121, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.053840976829151045, | |
| "grad_norm": 0.979810357093811, | |
| "learning_rate": 4.9145299145299147e-05, | |
| "loss": 1.6597, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.061532544947601193, | |
| "grad_norm": 0.9440365433692932, | |
| "learning_rate": 4.8951048951048956e-05, | |
| "loss": 1.6236, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06922411306605133, | |
| "grad_norm": 0.8523435592651367, | |
| "learning_rate": 4.875679875679876e-05, | |
| "loss": 1.616, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0769156811845015, | |
| "grad_norm": 1.1356542110443115, | |
| "learning_rate": 4.856254856254856e-05, | |
| "loss": 1.58, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08460724930295165, | |
| "grad_norm": 1.1632951498031616, | |
| "learning_rate": 4.836829836829837e-05, | |
| "loss": 1.5241, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0922988174214018, | |
| "grad_norm": 1.1739879846572876, | |
| "learning_rate": 4.8174048174048175e-05, | |
| "loss": 1.4737, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09999038553985194, | |
| "grad_norm": 1.1852024793624878, | |
| "learning_rate": 4.797979797979798e-05, | |
| "loss": 1.5002, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.10768195365830209, | |
| "grad_norm": 1.2705507278442383, | |
| "learning_rate": 4.778554778554779e-05, | |
| "loss": 1.4457, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11537352177675224, | |
| "grad_norm": 1.3838472366333008, | |
| "learning_rate": 4.75912975912976e-05, | |
| "loss": 1.4372, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12306508989520239, | |
| "grad_norm": 1.463762640953064, | |
| "learning_rate": 4.73970473970474e-05, | |
| "loss": 1.4695, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.13075665801365252, | |
| "grad_norm": 1.4812041521072388, | |
| "learning_rate": 4.7202797202797204e-05, | |
| "loss": 1.465, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.13844822613210267, | |
| "grad_norm": 1.3840124607086182, | |
| "learning_rate": 4.700854700854701e-05, | |
| "loss": 1.4121, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.14613979425055282, | |
| "grad_norm": 1.5955114364624023, | |
| "learning_rate": 4.681429681429682e-05, | |
| "loss": 1.4064, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.153831362369003, | |
| "grad_norm": 1.4825975894927979, | |
| "learning_rate": 4.662004662004662e-05, | |
| "loss": 1.3869, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.16152293048745314, | |
| "grad_norm": 1.5520519018173218, | |
| "learning_rate": 4.642579642579643e-05, | |
| "loss": 1.3914, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1692144986059033, | |
| "grad_norm": 1.3405165672302246, | |
| "learning_rate": 4.623154623154623e-05, | |
| "loss": 1.378, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.17690606672435344, | |
| "grad_norm": 1.4172375202178955, | |
| "learning_rate": 4.603729603729604e-05, | |
| "loss": 1.3843, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.1845976348428036, | |
| "grad_norm": 1.447364330291748, | |
| "learning_rate": 4.5843045843045846e-05, | |
| "loss": 1.3959, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.19228920296125374, | |
| "grad_norm": 1.5292952060699463, | |
| "learning_rate": 4.564879564879565e-05, | |
| "loss": 1.3926, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.19998077107970388, | |
| "grad_norm": 1.5983524322509766, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 1.3721, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.20767233919815403, | |
| "grad_norm": 1.520230770111084, | |
| "learning_rate": 4.526029526029526e-05, | |
| "loss": 1.362, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.21536390731660418, | |
| "grad_norm": 1.5406700372695923, | |
| "learning_rate": 4.506604506604507e-05, | |
| "loss": 1.36, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.22305547543505433, | |
| "grad_norm": 1.5974327325820923, | |
| "learning_rate": 4.4871794871794874e-05, | |
| "loss": 1.3326, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.23074704355350448, | |
| "grad_norm": 1.5478413105010986, | |
| "learning_rate": 4.467754467754468e-05, | |
| "loss": 1.3277, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.23843861167195463, | |
| "grad_norm": 1.537761926651001, | |
| "learning_rate": 4.448329448329449e-05, | |
| "loss": 1.3329, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.24613017979040477, | |
| "grad_norm": 1.5494816303253174, | |
| "learning_rate": 4.428904428904429e-05, | |
| "loss": 1.2966, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2538217479088549, | |
| "grad_norm": 1.6187357902526855, | |
| "learning_rate": 4.4094794094794093e-05, | |
| "loss": 1.3152, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.26151331602730504, | |
| "grad_norm": 1.9419796466827393, | |
| "learning_rate": 4.39005439005439e-05, | |
| "loss": 1.3223, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.2692048841457552, | |
| "grad_norm": 1.6638212203979492, | |
| "learning_rate": 4.370629370629371e-05, | |
| "loss": 1.3366, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.27689645226420534, | |
| "grad_norm": 1.8047728538513184, | |
| "learning_rate": 4.3512043512043516e-05, | |
| "loss": 1.3156, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2845880203826555, | |
| "grad_norm": 1.6696109771728516, | |
| "learning_rate": 4.331779331779332e-05, | |
| "loss": 1.3229, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.29227958850110564, | |
| "grad_norm": 1.6470463275909424, | |
| "learning_rate": 4.312354312354312e-05, | |
| "loss": 1.2842, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.2999711566195558, | |
| "grad_norm": 1.6966880559921265, | |
| "learning_rate": 4.292929292929293e-05, | |
| "loss": 1.2962, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.307662724738006, | |
| "grad_norm": 2.1080636978149414, | |
| "learning_rate": 4.2735042735042735e-05, | |
| "loss": 1.2865, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3153542928564561, | |
| "grad_norm": 1.9488563537597656, | |
| "learning_rate": 4.254079254079254e-05, | |
| "loss": 1.2932, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3230458609749063, | |
| "grad_norm": 1.8715245723724365, | |
| "learning_rate": 4.234654234654235e-05, | |
| "loss": 1.2751, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.3307374290933564, | |
| "grad_norm": 1.6135368347167969, | |
| "learning_rate": 4.215229215229216e-05, | |
| "loss": 1.2923, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3384289972118066, | |
| "grad_norm": 2.0071396827697754, | |
| "learning_rate": 4.195804195804196e-05, | |
| "loss": 1.3004, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3461205653302567, | |
| "grad_norm": 1.954253077507019, | |
| "learning_rate": 4.1763791763791764e-05, | |
| "loss": 1.2898, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.3538121334487069, | |
| "grad_norm": 2.44484806060791, | |
| "learning_rate": 4.1569541569541574e-05, | |
| "loss": 1.2873, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.361503701567157, | |
| "grad_norm": 1.7962288856506348, | |
| "learning_rate": 4.1375291375291377e-05, | |
| "loss": 1.2697, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.3691952696856072, | |
| "grad_norm": 1.6677371263504028, | |
| "learning_rate": 4.1181041181041186e-05, | |
| "loss": 1.2602, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.3768868378040573, | |
| "grad_norm": 1.6148051023483276, | |
| "learning_rate": 4.098679098679099e-05, | |
| "loss": 1.2889, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.38457840592250747, | |
| "grad_norm": 2.0414624214172363, | |
| "learning_rate": 4.079254079254079e-05, | |
| "loss": 1.2861, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3922699740409576, | |
| "grad_norm": 1.5257327556610107, | |
| "learning_rate": 4.05982905982906e-05, | |
| "loss": 1.2399, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.39996154215940777, | |
| "grad_norm": 1.9141441583633423, | |
| "learning_rate": 4.0404040404040405e-05, | |
| "loss": 1.2775, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4076531102778579, | |
| "grad_norm": 1.686346411705017, | |
| "learning_rate": 4.020979020979021e-05, | |
| "loss": 1.2471, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.41534467839630806, | |
| "grad_norm": 1.7622222900390625, | |
| "learning_rate": 4.001554001554002e-05, | |
| "loss": 1.2707, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4230362465147582, | |
| "grad_norm": 1.8795199394226074, | |
| "learning_rate": 3.982128982128983e-05, | |
| "loss": 1.2862, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.43072781463320836, | |
| "grad_norm": 1.8214608430862427, | |
| "learning_rate": 3.962703962703963e-05, | |
| "loss": 1.2553, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.4384193827516585, | |
| "grad_norm": 1.7057331800460815, | |
| "learning_rate": 3.9432789432789434e-05, | |
| "loss": 1.2684, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.44611095087010866, | |
| "grad_norm": 1.9549400806427002, | |
| "learning_rate": 3.923853923853924e-05, | |
| "loss": 1.2512, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.4538025189885588, | |
| "grad_norm": 1.906495451927185, | |
| "learning_rate": 3.904428904428905e-05, | |
| "loss": 1.2524, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.46149408710700895, | |
| "grad_norm": 1.76618230342865, | |
| "learning_rate": 3.885003885003885e-05, | |
| "loss": 1.2613, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4691856552254591, | |
| "grad_norm": 2.038341522216797, | |
| "learning_rate": 3.865578865578865e-05, | |
| "loss": 1.2266, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.47687722334390925, | |
| "grad_norm": 1.7204527854919434, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 1.2478, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.48456879146235937, | |
| "grad_norm": 1.7129229307174683, | |
| "learning_rate": 3.826728826728827e-05, | |
| "loss": 1.2207, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.49226035958080955, | |
| "grad_norm": 1.8858224153518677, | |
| "learning_rate": 3.8073038073038076e-05, | |
| "loss": 1.2219, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.49995192769925967, | |
| "grad_norm": 1.8016833066940308, | |
| "learning_rate": 3.787878787878788e-05, | |
| "loss": 1.2586, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5076434958177098, | |
| "grad_norm": 1.7311338186264038, | |
| "learning_rate": 3.768453768453769e-05, | |
| "loss": 1.2072, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.51533506393616, | |
| "grad_norm": 1.7362291812896729, | |
| "learning_rate": 3.749028749028749e-05, | |
| "loss": 1.229, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5230266320546101, | |
| "grad_norm": 2.358015537261963, | |
| "learning_rate": 3.72960372960373e-05, | |
| "loss": 1.196, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5307182001730603, | |
| "grad_norm": 1.9326188564300537, | |
| "learning_rate": 3.7101787101787104e-05, | |
| "loss": 1.2491, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5384097682915104, | |
| "grad_norm": 1.9867770671844482, | |
| "learning_rate": 3.690753690753691e-05, | |
| "loss": 1.2383, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5461013364099606, | |
| "grad_norm": 2.026604175567627, | |
| "learning_rate": 3.671328671328672e-05, | |
| "loss": 1.2172, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.5537929045284107, | |
| "grad_norm": 2.0612032413482666, | |
| "learning_rate": 3.651903651903652e-05, | |
| "loss": 1.2607, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.5614844726468609, | |
| "grad_norm": 1.8911861181259155, | |
| "learning_rate": 3.6324786324786323e-05, | |
| "loss": 1.2352, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.569176040765311, | |
| "grad_norm": 1.8968251943588257, | |
| "learning_rate": 3.613053613053613e-05, | |
| "loss": 1.2274, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5768676088837612, | |
| "grad_norm": 1.8784927129745483, | |
| "learning_rate": 3.593628593628594e-05, | |
| "loss": 1.2138, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5845591770022113, | |
| "grad_norm": 2.070425033569336, | |
| "learning_rate": 3.5742035742035746e-05, | |
| "loss": 1.2343, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.5922507451206614, | |
| "grad_norm": 2.00534987449646, | |
| "learning_rate": 3.554778554778555e-05, | |
| "loss": 1.2484, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.5999423132391116, | |
| "grad_norm": 2.018569231033325, | |
| "learning_rate": 3.535353535353535e-05, | |
| "loss": 1.1964, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.6076338813575618, | |
| "grad_norm": 2.2748711109161377, | |
| "learning_rate": 3.515928515928516e-05, | |
| "loss": 1.2415, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.615325449476012, | |
| "grad_norm": 1.8931610584259033, | |
| "learning_rate": 3.4965034965034965e-05, | |
| "loss": 1.2515, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.623017017594462, | |
| "grad_norm": 1.8468350172042847, | |
| "learning_rate": 3.477078477078477e-05, | |
| "loss": 1.2381, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6307085857129122, | |
| "grad_norm": 2.2593376636505127, | |
| "learning_rate": 3.457653457653458e-05, | |
| "loss": 1.2556, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.6384001538313624, | |
| "grad_norm": 1.9469683170318604, | |
| "learning_rate": 3.438228438228439e-05, | |
| "loss": 1.2071, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.6460917219498126, | |
| "grad_norm": 1.769471526145935, | |
| "learning_rate": 3.418803418803419e-05, | |
| "loss": 1.2112, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.6537832900682626, | |
| "grad_norm": 2.112982988357544, | |
| "learning_rate": 3.3993783993783994e-05, | |
| "loss": 1.2471, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.6614748581867128, | |
| "grad_norm": 1.8016294240951538, | |
| "learning_rate": 3.37995337995338e-05, | |
| "loss": 1.2145, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.669166426305163, | |
| "grad_norm": 1.9988363981246948, | |
| "learning_rate": 3.3605283605283607e-05, | |
| "loss": 1.191, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.6768579944236132, | |
| "grad_norm": 1.8531577587127686, | |
| "learning_rate": 3.341103341103341e-05, | |
| "loss": 1.1722, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.6845495625420632, | |
| "grad_norm": 1.9101706743240356, | |
| "learning_rate": 3.321678321678322e-05, | |
| "loss": 1.1738, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.6922411306605134, | |
| "grad_norm": 2.064260482788086, | |
| "learning_rate": 3.302253302253302e-05, | |
| "loss": 1.2045, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6999326987789636, | |
| "grad_norm": 1.895975947380066, | |
| "learning_rate": 3.282828282828283e-05, | |
| "loss": 1.1815, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.7076242668974138, | |
| "grad_norm": 1.6610194444656372, | |
| "learning_rate": 3.2634032634032635e-05, | |
| "loss": 1.2059, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.7153158350158638, | |
| "grad_norm": 1.8297659158706665, | |
| "learning_rate": 3.243978243978244e-05, | |
| "loss": 1.2034, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.723007403134314, | |
| "grad_norm": 2.0258054733276367, | |
| "learning_rate": 3.224553224553225e-05, | |
| "loss": 1.2051, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.7306989712527642, | |
| "grad_norm": 1.9012644290924072, | |
| "learning_rate": 3.205128205128206e-05, | |
| "loss": 1.2072, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.7383905393712143, | |
| "grad_norm": 1.8051044940948486, | |
| "learning_rate": 3.185703185703186e-05, | |
| "loss": 1.2149, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.7460821074896644, | |
| "grad_norm": 2.0245988368988037, | |
| "learning_rate": 3.1662781662781664e-05, | |
| "loss": 1.1886, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.7537736756081146, | |
| "grad_norm": 1.9541068077087402, | |
| "learning_rate": 3.146853146853147e-05, | |
| "loss": 1.2094, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.7614652437265648, | |
| "grad_norm": 1.8421745300292969, | |
| "learning_rate": 3.127428127428128e-05, | |
| "loss": 1.2041, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.7691568118450149, | |
| "grad_norm": 1.866947889328003, | |
| "learning_rate": 3.108003108003108e-05, | |
| "loss": 1.2103, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.776848379963465, | |
| "grad_norm": 2.1929476261138916, | |
| "learning_rate": 3.088578088578088e-05, | |
| "loss": 1.2058, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.7845399480819152, | |
| "grad_norm": 2.0555646419525146, | |
| "learning_rate": 3.069153069153069e-05, | |
| "loss": 1.1894, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.7922315162003654, | |
| "grad_norm": 2.0418717861175537, | |
| "learning_rate": 3.04972804972805e-05, | |
| "loss": 1.1942, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.7999230843188155, | |
| "grad_norm": 2.218815326690674, | |
| "learning_rate": 3.0303030303030306e-05, | |
| "loss": 1.1676, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.8076146524372656, | |
| "grad_norm": 1.9924787282943726, | |
| "learning_rate": 3.010878010878011e-05, | |
| "loss": 1.1938, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8153062205557158, | |
| "grad_norm": 1.7555562257766724, | |
| "learning_rate": 2.9914529914529915e-05, | |
| "loss": 1.1737, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.822997788674166, | |
| "grad_norm": 1.8218282461166382, | |
| "learning_rate": 2.972027972027972e-05, | |
| "loss": 1.1635, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.8306893567926161, | |
| "grad_norm": 1.9514037370681763, | |
| "learning_rate": 2.9526029526029525e-05, | |
| "loss": 1.1751, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.8383809249110662, | |
| "grad_norm": 1.995100498199463, | |
| "learning_rate": 2.9331779331779334e-05, | |
| "loss": 1.1705, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.8460724930295164, | |
| "grad_norm": 1.9724220037460327, | |
| "learning_rate": 2.913752913752914e-05, | |
| "loss": 1.2067, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8537640611479665, | |
| "grad_norm": 2.083712577819824, | |
| "learning_rate": 2.8943278943278944e-05, | |
| "loss": 1.1644, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.8614556292664167, | |
| "grad_norm": 2.1601104736328125, | |
| "learning_rate": 2.874902874902875e-05, | |
| "loss": 1.1772, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.8691471973848668, | |
| "grad_norm": 2.0567638874053955, | |
| "learning_rate": 2.8554778554778557e-05, | |
| "loss": 1.1846, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.876838765503317, | |
| "grad_norm": 2.4114835262298584, | |
| "learning_rate": 2.836052836052836e-05, | |
| "loss": 1.1658, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.8845303336217671, | |
| "grad_norm": 1.685823917388916, | |
| "learning_rate": 2.8166278166278166e-05, | |
| "loss": 1.1858, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.8922219017402173, | |
| "grad_norm": 2.006232738494873, | |
| "learning_rate": 2.7972027972027976e-05, | |
| "loss": 1.1781, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.8999134698586674, | |
| "grad_norm": 2.089742660522461, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 1.1569, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.9076050379771176, | |
| "grad_norm": 1.8699461221694946, | |
| "learning_rate": 2.7583527583527586e-05, | |
| "loss": 1.1653, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.9152966060955677, | |
| "grad_norm": 1.997068166732788, | |
| "learning_rate": 2.738927738927739e-05, | |
| "loss": 1.2116, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.9229881742140179, | |
| "grad_norm": 2.135075807571411, | |
| "learning_rate": 2.7195027195027195e-05, | |
| "loss": 1.2062, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9306797423324681, | |
| "grad_norm": 2.3898837566375732, | |
| "learning_rate": 2.7000777000777e-05, | |
| "loss": 1.1802, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.9383713104509182, | |
| "grad_norm": 1.8470486402511597, | |
| "learning_rate": 2.680652680652681e-05, | |
| "loss": 1.1388, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.9460628785693683, | |
| "grad_norm": 1.909771203994751, | |
| "learning_rate": 2.6612276612276614e-05, | |
| "loss": 1.1456, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.9537544466878185, | |
| "grad_norm": 1.75010085105896, | |
| "learning_rate": 2.641802641802642e-05, | |
| "loss": 1.1792, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.9614460148062687, | |
| "grad_norm": 1.835392951965332, | |
| "learning_rate": 2.6223776223776224e-05, | |
| "loss": 1.1666, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.9691375829247187, | |
| "grad_norm": 1.74484384059906, | |
| "learning_rate": 2.602952602952603e-05, | |
| "loss": 1.184, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.9768291510431689, | |
| "grad_norm": 1.9739990234375, | |
| "learning_rate": 2.5835275835275837e-05, | |
| "loss": 1.1342, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.9845207191616191, | |
| "grad_norm": 1.9165340662002563, | |
| "learning_rate": 2.564102564102564e-05, | |
| "loss": 1.1852, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.9922122872800693, | |
| "grad_norm": 2.0717194080352783, | |
| "learning_rate": 2.544677544677545e-05, | |
| "loss": 1.1611, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.9999038553985193, | |
| "grad_norm": 1.7079836130142212, | |
| "learning_rate": 2.5252525252525256e-05, | |
| "loss": 1.1498, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.0075954235169695, | |
| "grad_norm": 1.904863953590393, | |
| "learning_rate": 2.505827505827506e-05, | |
| "loss": 1.252, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.0152869916354197, | |
| "grad_norm": 1.7407102584838867, | |
| "learning_rate": 2.4864024864024865e-05, | |
| "loss": 1.1177, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.0229785597538699, | |
| "grad_norm": 2.13275146484375, | |
| "learning_rate": 2.4669774669774672e-05, | |
| "loss": 1.1409, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.03067012787232, | |
| "grad_norm": 1.8755226135253906, | |
| "learning_rate": 2.4475524475524478e-05, | |
| "loss": 1.1262, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.0383616959907702, | |
| "grad_norm": 2.0891833305358887, | |
| "learning_rate": 2.428127428127428e-05, | |
| "loss": 1.1126, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.0460532641092202, | |
| "grad_norm": 2.072007894515991, | |
| "learning_rate": 2.4087024087024088e-05, | |
| "loss": 1.1098, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.0537448322276703, | |
| "grad_norm": 1.8499518632888794, | |
| "learning_rate": 2.3892773892773894e-05, | |
| "loss": 1.1199, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.0614364003461205, | |
| "grad_norm": 1.9085497856140137, | |
| "learning_rate": 2.36985236985237e-05, | |
| "loss": 1.1178, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.0691279684645707, | |
| "grad_norm": 1.980193018913269, | |
| "learning_rate": 2.3504273504273504e-05, | |
| "loss": 1.1405, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.0768195365830209, | |
| "grad_norm": 1.961827039718628, | |
| "learning_rate": 2.331002331002331e-05, | |
| "loss": 1.1294, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.084511104701471, | |
| "grad_norm": 2.014617443084717, | |
| "learning_rate": 2.3115773115773116e-05, | |
| "loss": 1.1186, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.0922026728199212, | |
| "grad_norm": 2.0065386295318604, | |
| "learning_rate": 2.2921522921522923e-05, | |
| "loss": 1.0977, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.0998942409383714, | |
| "grad_norm": 1.9613438844680786, | |
| "learning_rate": 2.272727272727273e-05, | |
| "loss": 1.1649, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.1075858090568214, | |
| "grad_norm": 2.0321905612945557, | |
| "learning_rate": 2.2533022533022536e-05, | |
| "loss": 1.1437, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.1152773771752715, | |
| "grad_norm": 1.954882264137268, | |
| "learning_rate": 2.233877233877234e-05, | |
| "loss": 1.1166, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.1229689452937217, | |
| "grad_norm": 2.1001336574554443, | |
| "learning_rate": 2.2144522144522145e-05, | |
| "loss": 1.1222, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.1306605134121719, | |
| "grad_norm": 1.9967658519744873, | |
| "learning_rate": 2.195027195027195e-05, | |
| "loss": 1.113, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.138352081530622, | |
| "grad_norm": 2.000359296798706, | |
| "learning_rate": 2.1756021756021758e-05, | |
| "loss": 1.1019, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.1460436496490722, | |
| "grad_norm": 1.7127162218093872, | |
| "learning_rate": 2.156177156177156e-05, | |
| "loss": 1.1077, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.1537352177675224, | |
| "grad_norm": 2.044915199279785, | |
| "learning_rate": 2.1367521367521368e-05, | |
| "loss": 1.1493, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.1614267858859726, | |
| "grad_norm": 1.906977653503418, | |
| "learning_rate": 2.1173271173271174e-05, | |
| "loss": 1.0935, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.1691183540044228, | |
| "grad_norm": 2.004258871078491, | |
| "learning_rate": 2.097902097902098e-05, | |
| "loss": 1.1548, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.1768099221228727, | |
| "grad_norm": 2.028057813644409, | |
| "learning_rate": 2.0784770784770787e-05, | |
| "loss": 1.12, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.184501490241323, | |
| "grad_norm": 2.0909903049468994, | |
| "learning_rate": 2.0590520590520593e-05, | |
| "loss": 1.1347, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.192193058359773, | |
| "grad_norm": 2.309083938598633, | |
| "learning_rate": 2.0396270396270396e-05, | |
| "loss": 1.1467, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.1998846264782232, | |
| "grad_norm": 2.195953845977783, | |
| "learning_rate": 2.0202020202020203e-05, | |
| "loss": 1.1166, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.2075761945966734, | |
| "grad_norm": 1.8063645362854004, | |
| "learning_rate": 2.000777000777001e-05, | |
| "loss": 1.0918, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.2152677627151236, | |
| "grad_norm": 2.0312583446502686, | |
| "learning_rate": 1.9813519813519816e-05, | |
| "loss": 1.1409, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.2229593308335738, | |
| "grad_norm": 2.091669797897339, | |
| "learning_rate": 1.961926961926962e-05, | |
| "loss": 1.0938, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.2306508989520237, | |
| "grad_norm": 1.7335410118103027, | |
| "learning_rate": 1.9425019425019425e-05, | |
| "loss": 1.1085, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.238342467070474, | |
| "grad_norm": 2.0697057247161865, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 1.1355, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.246034035188924, | |
| "grad_norm": 2.063481569290161, | |
| "learning_rate": 1.9036519036519038e-05, | |
| "loss": 1.1069, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.2537256033073743, | |
| "grad_norm": 2.1378798484802246, | |
| "learning_rate": 1.8842268842268844e-05, | |
| "loss": 1.1273, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.2614171714258244, | |
| "grad_norm": 2.2633299827575684, | |
| "learning_rate": 1.864801864801865e-05, | |
| "loss": 1.1173, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.2691087395442746, | |
| "grad_norm": 1.8550769090652466, | |
| "learning_rate": 1.8453768453768454e-05, | |
| "loss": 1.0902, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.2768003076627248, | |
| "grad_norm": 1.6822105646133423, | |
| "learning_rate": 1.825951825951826e-05, | |
| "loss": 1.0986, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.284491875781175, | |
| "grad_norm": 1.8236192464828491, | |
| "learning_rate": 1.8065268065268067e-05, | |
| "loss": 1.1102, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.2921834438996251, | |
| "grad_norm": 2.2287065982818604, | |
| "learning_rate": 1.7871017871017873e-05, | |
| "loss": 1.1302, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.299875012018075, | |
| "grad_norm": 2.2197864055633545, | |
| "learning_rate": 1.7676767676767676e-05, | |
| "loss": 1.0902, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.3075665801365253, | |
| "grad_norm": 2.1204230785369873, | |
| "learning_rate": 1.7482517482517483e-05, | |
| "loss": 1.1228, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.3152581482549754, | |
| "grad_norm": 1.8001035451889038, | |
| "learning_rate": 1.728826728826729e-05, | |
| "loss": 1.1183, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.3229497163734256, | |
| "grad_norm": 1.9405324459075928, | |
| "learning_rate": 1.7094017094017095e-05, | |
| "loss": 1.1377, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.3306412844918758, | |
| "grad_norm": 1.9750436544418335, | |
| "learning_rate": 1.68997668997669e-05, | |
| "loss": 1.1051, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.338332852610326, | |
| "grad_norm": 1.7894169092178345, | |
| "learning_rate": 1.6705516705516705e-05, | |
| "loss": 1.0972, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.3460244207287761, | |
| "grad_norm": 1.7701419591903687, | |
| "learning_rate": 1.651126651126651e-05, | |
| "loss": 1.0815, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.353715988847226, | |
| "grad_norm": 2.0538313388824463, | |
| "learning_rate": 1.6317016317016318e-05, | |
| "loss": 1.1072, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.3614075569656765, | |
| "grad_norm": 2.0709080696105957, | |
| "learning_rate": 1.6122766122766124e-05, | |
| "loss": 1.0918, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.3690991250841265, | |
| "grad_norm": 2.1939444541931152, | |
| "learning_rate": 1.592851592851593e-05, | |
| "loss": 1.0957, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.3767906932025766, | |
| "grad_norm": 1.5446499586105347, | |
| "learning_rate": 1.5734265734265734e-05, | |
| "loss": 1.1032, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.3844822613210268, | |
| "grad_norm": 1.971351981163025, | |
| "learning_rate": 1.554001554001554e-05, | |
| "loss": 1.1319, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.392173829439477, | |
| "grad_norm": 1.9268254041671753, | |
| "learning_rate": 1.5345765345765346e-05, | |
| "loss": 1.1242, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.3998653975579272, | |
| "grad_norm": 2.0808980464935303, | |
| "learning_rate": 1.5151515151515153e-05, | |
| "loss": 1.1173, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.4075569656763773, | |
| "grad_norm": 2.044785976409912, | |
| "learning_rate": 1.4957264957264958e-05, | |
| "loss": 1.1481, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.4152485337948275, | |
| "grad_norm": 1.690045714378357, | |
| "learning_rate": 1.4763014763014762e-05, | |
| "loss": 1.1239, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.4229401019132775, | |
| "grad_norm": 1.950405478477478, | |
| "learning_rate": 1.456876456876457e-05, | |
| "loss": 1.1306, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.4306316700317276, | |
| "grad_norm": 2.236177921295166, | |
| "learning_rate": 1.4374514374514375e-05, | |
| "loss": 1.0963, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.4383232381501778, | |
| "grad_norm": 2.0903525352478027, | |
| "learning_rate": 1.418026418026418e-05, | |
| "loss": 1.1024, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.446014806268628, | |
| "grad_norm": 2.220693349838257, | |
| "learning_rate": 1.3986013986013988e-05, | |
| "loss": 1.1132, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.4537063743870782, | |
| "grad_norm": 1.9859932661056519, | |
| "learning_rate": 1.3791763791763793e-05, | |
| "loss": 1.1163, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.4613979425055283, | |
| "grad_norm": 2.0174367427825928, | |
| "learning_rate": 1.3597513597513598e-05, | |
| "loss": 1.0975, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.4690895106239785, | |
| "grad_norm": 2.1779918670654297, | |
| "learning_rate": 1.3403263403263406e-05, | |
| "loss": 1.1076, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.4767810787424285, | |
| "grad_norm": 2.0282936096191406, | |
| "learning_rate": 1.320901320901321e-05, | |
| "loss": 1.1014, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.4844726468608789, | |
| "grad_norm": 2.0938243865966797, | |
| "learning_rate": 1.3014763014763015e-05, | |
| "loss": 1.1028, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.4921642149793288, | |
| "grad_norm": 2.0279526710510254, | |
| "learning_rate": 1.282051282051282e-05, | |
| "loss": 1.0922, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.499855783097779, | |
| "grad_norm": 1.9357686042785645, | |
| "learning_rate": 1.2626262626262628e-05, | |
| "loss": 1.0896, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.5075473512162292, | |
| "grad_norm": 2.1684436798095703, | |
| "learning_rate": 1.2432012432012433e-05, | |
| "loss": 1.1063, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.5152389193346794, | |
| "grad_norm": 1.9890198707580566, | |
| "learning_rate": 1.2237762237762239e-05, | |
| "loss": 1.0927, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.5229304874531295, | |
| "grad_norm": 1.9705474376678467, | |
| "learning_rate": 1.2043512043512044e-05, | |
| "loss": 1.1135, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.5306220555715797, | |
| "grad_norm": 2.2323312759399414, | |
| "learning_rate": 1.184926184926185e-05, | |
| "loss": 1.1035, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.5383136236900299, | |
| "grad_norm": 2.031677007675171, | |
| "learning_rate": 1.1655011655011655e-05, | |
| "loss": 1.0881, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.5460051918084798, | |
| "grad_norm": 1.9229755401611328, | |
| "learning_rate": 1.1460761460761461e-05, | |
| "loss": 1.1678, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.5536967599269302, | |
| "grad_norm": 2.2385263442993164, | |
| "learning_rate": 1.1266511266511268e-05, | |
| "loss": 1.0922, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.5613883280453802, | |
| "grad_norm": 2.14032244682312, | |
| "learning_rate": 1.1072261072261073e-05, | |
| "loss": 1.0948, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.5690798961638304, | |
| "grad_norm": 2.2739627361297607, | |
| "learning_rate": 1.0878010878010879e-05, | |
| "loss": 1.1033, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.5767714642822805, | |
| "grad_norm": 2.117405891418457, | |
| "learning_rate": 1.0683760683760684e-05, | |
| "loss": 1.1144, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.5844630324007307, | |
| "grad_norm": 1.917315125465393, | |
| "learning_rate": 1.048951048951049e-05, | |
| "loss": 1.1462, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.592154600519181, | |
| "grad_norm": 1.9057689905166626, | |
| "learning_rate": 1.0295260295260297e-05, | |
| "loss": 1.1243, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.5998461686376308, | |
| "grad_norm": 2.0241281986236572, | |
| "learning_rate": 1.0101010101010101e-05, | |
| "loss": 1.1061, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.6075377367560812, | |
| "grad_norm": 2.33536958694458, | |
| "learning_rate": 9.906759906759908e-06, | |
| "loss": 1.0989, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.6152293048745312, | |
| "grad_norm": 2.1172101497650146, | |
| "learning_rate": 9.712509712509713e-06, | |
| "loss": 1.0936, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.6229208729929816, | |
| "grad_norm": 1.8319240808486938, | |
| "learning_rate": 9.518259518259519e-06, | |
| "loss": 1.1099, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.6306124411114316, | |
| "grad_norm": 1.9867093563079834, | |
| "learning_rate": 9.324009324009325e-06, | |
| "loss": 1.0983, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.6383040092298817, | |
| "grad_norm": 1.9787898063659668, | |
| "learning_rate": 9.12975912975913e-06, | |
| "loss": 1.1042, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.645995577348332, | |
| "grad_norm": 1.9842088222503662, | |
| "learning_rate": 8.935508935508937e-06, | |
| "loss": 1.1094, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.653687145466782, | |
| "grad_norm": 1.9378868341445923, | |
| "learning_rate": 8.741258741258741e-06, | |
| "loss": 1.0995, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.6613787135852323, | |
| "grad_norm": 1.8845783472061157, | |
| "learning_rate": 8.547008547008548e-06, | |
| "loss": 1.0884, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.6690702817036822, | |
| "grad_norm": 1.9667729139328003, | |
| "learning_rate": 8.352758352758352e-06, | |
| "loss": 1.0916, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.6767618498221326, | |
| "grad_norm": 2.1440649032592773, | |
| "learning_rate": 8.158508158508159e-06, | |
| "loss": 1.1166, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.6844534179405826, | |
| "grad_norm": 2.1377923488616943, | |
| "learning_rate": 7.964257964257965e-06, | |
| "loss": 1.0733, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.6921449860590327, | |
| "grad_norm": 1.9222036600112915, | |
| "learning_rate": 7.77000777000777e-06, | |
| "loss": 1.1064, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.699836554177483, | |
| "grad_norm": 1.9658797979354858, | |
| "learning_rate": 7.5757575757575764e-06, | |
| "loss": 1.1212, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.707528122295933, | |
| "grad_norm": 1.9604525566101074, | |
| "learning_rate": 7.381507381507381e-06, | |
| "loss": 1.094, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.7152196904143833, | |
| "grad_norm": 2.1394431591033936, | |
| "learning_rate": 7.187257187257188e-06, | |
| "loss": 1.0974, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.7229112585328332, | |
| "grad_norm": 1.819686770439148, | |
| "learning_rate": 6.993006993006994e-06, | |
| "loss": 1.0815, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.7306028266512836, | |
| "grad_norm": 1.8422613143920898, | |
| "learning_rate": 6.798756798756799e-06, | |
| "loss": 1.1034, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.7382943947697336, | |
| "grad_norm": 1.9406566619873047, | |
| "learning_rate": 6.604506604506605e-06, | |
| "loss": 1.0835, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.745985962888184, | |
| "grad_norm": 1.831289291381836, | |
| "learning_rate": 6.41025641025641e-06, | |
| "loss": 1.0857, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.753677531006634, | |
| "grad_norm": 1.942730188369751, | |
| "learning_rate": 6.216006216006216e-06, | |
| "loss": 1.1112, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.761369099125084, | |
| "grad_norm": 1.8144134283065796, | |
| "learning_rate": 6.021756021756022e-06, | |
| "loss": 1.0873, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.7690606672435343, | |
| "grad_norm": 1.7852503061294556, | |
| "learning_rate": 5.8275058275058275e-06, | |
| "loss": 1.0756, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.7767522353619845, | |
| "grad_norm": 2.232545852661133, | |
| "learning_rate": 5.633255633255634e-06, | |
| "loss": 1.114, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.7844438034804346, | |
| "grad_norm": 2.143216133117676, | |
| "learning_rate": 5.4390054390054395e-06, | |
| "loss": 1.1015, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.7921353715988846, | |
| "grad_norm": 1.8443506956100464, | |
| "learning_rate": 5.244755244755245e-06, | |
| "loss": 1.1006, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.799826939717335, | |
| "grad_norm": 1.7747691869735718, | |
| "learning_rate": 5.050505050505051e-06, | |
| "loss": 1.099, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.807518507835785, | |
| "grad_norm": 1.8745768070220947, | |
| "learning_rate": 4.856254856254856e-06, | |
| "loss": 1.0979, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.8152100759542353, | |
| "grad_norm": 1.9972470998764038, | |
| "learning_rate": 4.662004662004663e-06, | |
| "loss": 1.1083, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.8229016440726853, | |
| "grad_norm": 1.8246738910675049, | |
| "learning_rate": 4.467754467754468e-06, | |
| "loss": 1.1061, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.8305932121911355, | |
| "grad_norm": 1.9446369409561157, | |
| "learning_rate": 4.273504273504274e-06, | |
| "loss": 1.0843, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.8382847803095856, | |
| "grad_norm": 1.9822577238082886, | |
| "learning_rate": 4.079254079254079e-06, | |
| "loss": 1.0722, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.8459763484280358, | |
| "grad_norm": 2.3684942722320557, | |
| "learning_rate": 3.885003885003885e-06, | |
| "loss": 1.1354, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.853667916546486, | |
| "grad_norm": 1.8989230394363403, | |
| "learning_rate": 3.6907536907536906e-06, | |
| "loss": 1.0931, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.861359484664936, | |
| "grad_norm": 1.952711582183838, | |
| "learning_rate": 3.496503496503497e-06, | |
| "loss": 1.0823, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.8690510527833863, | |
| "grad_norm": 2.154554605484009, | |
| "learning_rate": 3.3022533022533026e-06, | |
| "loss": 1.0846, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.8767426209018363, | |
| "grad_norm": 1.8646643161773682, | |
| "learning_rate": 3.108003108003108e-06, | |
| "loss": 1.0961, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.8844341890202865, | |
| "grad_norm": 1.9604412317276, | |
| "learning_rate": 2.9137529137529138e-06, | |
| "loss": 1.0823, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.8921257571387367, | |
| "grad_norm": 1.7040070295333862, | |
| "learning_rate": 2.7195027195027198e-06, | |
| "loss": 1.0831, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.8998173252571868, | |
| "grad_norm": 2.3262603282928467, | |
| "learning_rate": 2.5252525252525253e-06, | |
| "loss": 1.0668, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.907508893375637, | |
| "grad_norm": 2.0159196853637695, | |
| "learning_rate": 2.3310023310023313e-06, | |
| "loss": 1.094, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.915200461494087, | |
| "grad_norm": 1.9653328657150269, | |
| "learning_rate": 2.136752136752137e-06, | |
| "loss": 1.0677, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.9228920296125374, | |
| "grad_norm": 1.7489328384399414, | |
| "learning_rate": 1.9425019425019425e-06, | |
| "loss": 1.0947, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.9305835977309873, | |
| "grad_norm": 1.9971283674240112, | |
| "learning_rate": 1.7482517482517485e-06, | |
| "loss": 1.135, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 1.9382751658494377, | |
| "grad_norm": 1.9700243473052979, | |
| "learning_rate": 1.554001554001554e-06, | |
| "loss": 1.0783, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.9459667339678877, | |
| "grad_norm": 1.8461562395095825, | |
| "learning_rate": 1.3597513597513599e-06, | |
| "loss": 1.113, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 1.9536583020863378, | |
| "grad_norm": 1.6744381189346313, | |
| "learning_rate": 1.1655011655011657e-06, | |
| "loss": 1.0829, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 1.961349870204788, | |
| "grad_norm": 1.9034210443496704, | |
| "learning_rate": 9.712509712509713e-07, | |
| "loss": 1.0781, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.9690414383232382, | |
| "grad_norm": 1.98823881149292, | |
| "learning_rate": 7.77000777000777e-07, | |
| "loss": 1.1031, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 1.9767330064416884, | |
| "grad_norm": 1.905478596687317, | |
| "learning_rate": 5.827505827505828e-07, | |
| "loss": 1.081, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 1.9844245745601383, | |
| "grad_norm": 1.8001110553741455, | |
| "learning_rate": 3.885003885003885e-07, | |
| "loss": 1.0793, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.9921161426785887, | |
| "grad_norm": 2.089235305786133, | |
| "learning_rate": 1.9425019425019426e-07, | |
| "loss": 1.1, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 1.9998077107970387, | |
| "grad_norm": 2.018970489501953, | |
| "learning_rate": 0.0, | |
| "loss": 1.0801, | |
| "step": 2600 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.1771710700767724e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |