| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1914, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001567398119122257, | |
| "grad_norm": 0.5552594661712646, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 3.3921, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01567398119122257, | |
| "grad_norm": 0.660531222820282, | |
| "learning_rate": 3.4482758620689657e-05, | |
| "loss": 3.5922, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03134796238244514, | |
| "grad_norm": 0.5598097443580627, | |
| "learning_rate": 6.896551724137931e-05, | |
| "loss": 3.619, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.047021943573667714, | |
| "grad_norm": 0.5996779799461365, | |
| "learning_rate": 0.00010344827586206898, | |
| "loss": 3.4824, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06269592476489028, | |
| "grad_norm": 0.8172075152397156, | |
| "learning_rate": 0.00013793103448275863, | |
| "loss": 3.3997, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07836990595611286, | |
| "grad_norm": 1.6439019441604614, | |
| "learning_rate": 0.00017241379310344826, | |
| "loss": 3.2485, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09404388714733543, | |
| "grad_norm": 0.28180772066116333, | |
| "learning_rate": 0.00019999942697524717, | |
| "loss": 2.9975, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.109717868338558, | |
| "grad_norm": 0.3870689868927002, | |
| "learning_rate": 0.00019997937179843937, | |
| "loss": 3.0446, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12539184952978055, | |
| "grad_norm": 0.575548529624939, | |
| "learning_rate": 0.00019993067195079803, | |
| "loss": 3.0178, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14106583072100312, | |
| "grad_norm": 0.781446635723114, | |
| "learning_rate": 0.00019985334138511237, | |
| "loss": 3.0394, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15673981191222572, | |
| "grad_norm": 1.725696325302124, | |
| "learning_rate": 0.00019974740225703878, | |
| "loss": 3.0751, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 0.2845414876937866, | |
| "learning_rate": 0.00019961288491875278, | |
| "loss": 2.9291, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.18808777429467086, | |
| "grad_norm": 0.36810651421546936, | |
| "learning_rate": 0.00019944982791025333, | |
| "loss": 2.9491, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.20376175548589343, | |
| "grad_norm": 0.5454439520835876, | |
| "learning_rate": 0.00019925827794832056, | |
| "loss": 3.0337, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.219435736677116, | |
| "grad_norm": 0.6503669619560242, | |
| "learning_rate": 0.00019903828991313138, | |
| "loss": 3.0246, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.23510971786833856, | |
| "grad_norm": 1.4392451047897339, | |
| "learning_rate": 0.00019878992683253582, | |
| "loss": 3.0232, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2507836990595611, | |
| "grad_norm": 0.2859440743923187, | |
| "learning_rate": 0.00019851325986399934, | |
| "loss": 2.8955, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2664576802507837, | |
| "grad_norm": 0.44268104434013367, | |
| "learning_rate": 0.0001982083682742156, | |
| "loss": 2.9338, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.28213166144200624, | |
| "grad_norm": 0.5128395557403564, | |
| "learning_rate": 0.00019787533941639638, | |
| "loss": 3.0089, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.29780564263322884, | |
| "grad_norm": 0.7328920364379883, | |
| "learning_rate": 0.00019751426870524407, | |
| "loss": 3.0157, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.31347962382445144, | |
| "grad_norm": 1.5265012979507446, | |
| "learning_rate": 0.000197125259589615, | |
| "loss": 2.9007, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.329153605015674, | |
| "grad_norm": 0.2766813635826111, | |
| "learning_rate": 0.0001967084235228807, | |
| "loss": 2.8275, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 0.36695396900177, | |
| "learning_rate": 0.00019626387993099579, | |
| "loss": 2.9158, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3605015673981191, | |
| "grad_norm": 0.5359162092208862, | |
| "learning_rate": 0.00019579175617828187, | |
| "loss": 2.9465, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3761755485893417, | |
| "grad_norm": 0.6529833674430847, | |
| "learning_rate": 0.0001952921875309368, | |
| "loss": 2.981, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.39184952978056425, | |
| "grad_norm": 1.5314627885818481, | |
| "learning_rate": 0.00019476531711828027, | |
| "loss": 2.9737, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.40752351097178685, | |
| "grad_norm": 0.2949506342411041, | |
| "learning_rate": 0.00019421129589174618, | |
| "loss": 2.8208, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4231974921630094, | |
| "grad_norm": 0.39567869901657104, | |
| "learning_rate": 0.00019363028258163447, | |
| "loss": 2.8557, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.438871473354232, | |
| "grad_norm": 0.5587254166603088, | |
| "learning_rate": 0.00019302244365163376, | |
| "loss": 2.9494, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.7218978404998779, | |
| "learning_rate": 0.0001923879532511287, | |
| "loss": 2.9742, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4702194357366771, | |
| "grad_norm": 1.4482598304748535, | |
| "learning_rate": 0.0001917269931653049, | |
| "loss": 2.8646, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.48589341692789967, | |
| "grad_norm": 0.2901701033115387, | |
| "learning_rate": 0.00019103975276306678, | |
| "loss": 2.7788, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5015673981191222, | |
| "grad_norm": 0.4310539960861206, | |
| "learning_rate": 0.00019032642894278192, | |
| "loss": 2.8655, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 0.5589954853057861, | |
| "learning_rate": 0.0001895872260758688, | |
| "loss": 2.914, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5329153605015674, | |
| "grad_norm": 0.7243526577949524, | |
| "learning_rate": 0.00018882235594824308, | |
| "loss": 2.9191, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.54858934169279, | |
| "grad_norm": 1.4200222492218018, | |
| "learning_rate": 0.00018803203769963967, | |
| "loss": 2.8128, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5642633228840125, | |
| "grad_norm": 0.26594147086143494, | |
| "learning_rate": 0.000187216497760828, | |
| "loss": 2.762, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5799373040752351, | |
| "grad_norm": 0.3894258439540863, | |
| "learning_rate": 0.00018637596978873835, | |
| "loss": 2.9077, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5956112852664577, | |
| "grad_norm": 0.5348561406135559, | |
| "learning_rate": 0.00018551069459951758, | |
| "loss": 2.9292, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6112852664576802, | |
| "grad_norm": 0.746507465839386, | |
| "learning_rate": 0.00018462092009953408, | |
| "loss": 2.8795, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6269592476489029, | |
| "grad_norm": 1.5225753784179688, | |
| "learning_rate": 0.0001837069012143511, | |
| "loss": 2.8263, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6426332288401254, | |
| "grad_norm": 0.26905450224876404, | |
| "learning_rate": 0.00018276889981568906, | |
| "loss": 2.7218, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.658307210031348, | |
| "grad_norm": 0.3912515342235565, | |
| "learning_rate": 0.00018180718464639787, | |
| "loss": 2.819, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6739811912225705, | |
| "grad_norm": 0.5661373138427734, | |
| "learning_rate": 0.00018082203124346045, | |
| "loss": 2.8772, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 0.805776059627533, | |
| "learning_rate": 0.0001798137218590498, | |
| "loss": 2.9562, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7053291536050157, | |
| "grad_norm": 1.4879732131958008, | |
| "learning_rate": 0.00017878254537966216, | |
| "loss": 2.7925, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7210031347962382, | |
| "grad_norm": 0.28762391209602356, | |
| "learning_rate": 0.00017772879724334937, | |
| "loss": 2.8006, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7366771159874608, | |
| "grad_norm": 0.41769474744796753, | |
| "learning_rate": 0.00017665277935507398, | |
| "loss": 2.8148, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7523510971786834, | |
| "grad_norm": 0.633671760559082, | |
| "learning_rate": 0.00017555480000021198, | |
| "loss": 2.8461, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.768025078369906, | |
| "grad_norm": 0.816681444644928, | |
| "learning_rate": 0.00017443517375622704, | |
| "loss": 2.8826, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7836990595611285, | |
| "grad_norm": 1.3913438320159912, | |
| "learning_rate": 0.00017329422140254235, | |
| "loss": 2.7449, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.799373040752351, | |
| "grad_norm": 0.30519339442253113, | |
| "learning_rate": 0.0001721322698286354, | |
| "loss": 2.7933, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8150470219435737, | |
| "grad_norm": 0.3911365270614624, | |
| "learning_rate": 0.0001709496519403823, | |
| "loss": 2.8433, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8307210031347962, | |
| "grad_norm": 0.5197113752365112, | |
| "learning_rate": 0.00016974670656467824, | |
| "loss": 2.828, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8463949843260188, | |
| "grad_norm": 0.7463776469230652, | |
| "learning_rate": 0.00016852377835236166, | |
| "loss": 2.9549, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 1.6616039276123047, | |
| "learning_rate": 0.00016728121767946977, | |
| "loss": 2.8843, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.877742946708464, | |
| "grad_norm": 0.2525625228881836, | |
| "learning_rate": 0.00016601938054685385, | |
| "loss": 2.7715, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8934169278996865, | |
| "grad_norm": 0.4268622100353241, | |
| "learning_rate": 0.00016473862847818277, | |
| "loss": 2.8196, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.5538944005966187, | |
| "learning_rate": 0.00016343932841636456, | |
| "loss": 2.8619, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9247648902821317, | |
| "grad_norm": 0.7213597297668457, | |
| "learning_rate": 0.00016212185261841499, | |
| "loss": 2.9276, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.9404388714733543, | |
| "grad_norm": 1.6927486658096313, | |
| "learning_rate": 0.00016078657854880376, | |
| "loss": 2.7446, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9561128526645768, | |
| "grad_norm": 0.3090650141239166, | |
| "learning_rate": 0.000159433888771309, | |
| "loss": 2.747, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9717868338557993, | |
| "grad_norm": 0.4454885423183441, | |
| "learning_rate": 0.00015806417083941002, | |
| "loss": 2.8622, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.987460815047022, | |
| "grad_norm": 0.8089830279350281, | |
| "learning_rate": 0.00015667781718525157, | |
| "loss": 2.8568, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.0031347962382444, | |
| "grad_norm": 0.21911662817001343, | |
| "learning_rate": 0.00015527522500720934, | |
| "loss": 2.7012, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.0188087774294672, | |
| "grad_norm": 0.3196061849594116, | |
| "learning_rate": 0.00015385679615609042, | |
| "loss": 2.7127, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0344827586206897, | |
| "grad_norm": 0.4679659605026245, | |
| "learning_rate": 0.00015242293702000086, | |
| "loss": 2.7103, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0501567398119123, | |
| "grad_norm": 0.6506277322769165, | |
| "learning_rate": 0.00015097405840791276, | |
| "loss": 2.6762, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0658307210031348, | |
| "grad_norm": 0.9151834845542908, | |
| "learning_rate": 0.00014951057543196566, | |
| "loss": 2.6503, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0815047021943573, | |
| "grad_norm": 0.2834145724773407, | |
| "learning_rate": 0.00014803290738853395, | |
| "loss": 2.5103, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.09717868338558, | |
| "grad_norm": 0.3199128210544586, | |
| "learning_rate": 0.00014654147763809637, | |
| "loss": 2.7147, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1128526645768024, | |
| "grad_norm": 0.5001178979873657, | |
| "learning_rate": 0.00014503671348394057, | |
| "loss": 2.7124, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1285266457680252, | |
| "grad_norm": 0.6373796463012695, | |
| "learning_rate": 0.0001435190460497384, | |
| "loss": 2.7012, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1442006269592477, | |
| "grad_norm": 0.9262835383415222, | |
| "learning_rate": 0.00014198891015602646, | |
| "loss": 2.6493, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1598746081504703, | |
| "grad_norm": 0.2955109775066376, | |
| "learning_rate": 0.00014044674419562734, | |
| "loss": 2.5023, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1755485893416928, | |
| "grad_norm": 0.34623944759368896, | |
| "learning_rate": 0.0001388929900080476, | |
| "loss": 2.6849, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1912225705329154, | |
| "grad_norm": 0.43281346559524536, | |
| "learning_rate": 0.00013732809275288828, | |
| "loss": 2.6655, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.206896551724138, | |
| "grad_norm": 0.6403496265411377, | |
| "learning_rate": 0.000135752500782304, | |
| "loss": 2.6557, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2225705329153604, | |
| "grad_norm": 0.8644494414329529, | |
| "learning_rate": 0.00013416666551254748, | |
| "loss": 2.6162, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.238244514106583, | |
| "grad_norm": 0.2673655152320862, | |
| "learning_rate": 0.00013257104129463614, | |
| "loss": 2.4991, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2539184952978055, | |
| "grad_norm": 0.3414609432220459, | |
| "learning_rate": 0.00013096608528417788, | |
| "loss": 2.6638, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2695924764890283, | |
| "grad_norm": 0.47588247060775757, | |
| "learning_rate": 0.00012935225731039348, | |
| "loss": 2.6465, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2852664576802508, | |
| "grad_norm": 0.6843693852424622, | |
| "learning_rate": 0.00012773001974437267, | |
| "loss": 2.6697, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.3009404388714734, | |
| "grad_norm": 0.9213519096374512, | |
| "learning_rate": 0.0001260998373666022, | |
| "loss": 2.6691, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.316614420062696, | |
| "grad_norm": 0.26475438475608826, | |
| "learning_rate": 0.0001244621772338036, | |
| "loss": 2.5109, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3322884012539185, | |
| "grad_norm": 0.3754674196243286, | |
| "learning_rate": 0.0001228175085451186, | |
| "loss": 2.6699, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.347962382445141, | |
| "grad_norm": 0.49352237582206726, | |
| "learning_rate": 0.00012116630250768097, | |
| "loss": 2.6889, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.7134298086166382, | |
| "learning_rate": 0.00011950903220161285, | |
| "loss": 2.6502, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 0.9475263953208923, | |
| "learning_rate": 0.00011784617244448451, | |
| "loss": 2.6062, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3949843260188088, | |
| "grad_norm": 0.2767506539821625, | |
| "learning_rate": 0.0001161781996552765, | |
| "loss": 2.4584, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.4106583072100314, | |
| "grad_norm": 0.38563597202301025, | |
| "learning_rate": 0.00011450559171788269, | |
| "loss": 2.6751, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.426332288401254, | |
| "grad_norm": 0.466782808303833, | |
| "learning_rate": 0.00011282882784419398, | |
| "loss": 2.6181, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.4420062695924765, | |
| "grad_norm": 0.7600008845329285, | |
| "learning_rate": 0.00011114838843680095, | |
| "loss": 2.6729, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.457680250783699, | |
| "grad_norm": 0.956167459487915, | |
| "learning_rate": 0.0001094647549513561, | |
| "loss": 2.5486, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.4733542319749215, | |
| "grad_norm": 0.257744699716568, | |
| "learning_rate": 0.00010777840975863383, | |
| "loss": 2.4545, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.489028213166144, | |
| "grad_norm": 0.347033828496933, | |
| "learning_rate": 0.00010608983600632831, | |
| "loss": 2.6313, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.5047021943573666, | |
| "grad_norm": 0.5105902552604675, | |
| "learning_rate": 0.00010439951748062912, | |
| "loss": 2.623, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.5203761755485894, | |
| "grad_norm": 0.7116357684135437, | |
| "learning_rate": 0.00010270793846761347, | |
| "loss": 2.6493, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.536050156739812, | |
| "grad_norm": 0.9367708563804626, | |
| "learning_rate": 0.00010101558361449552, | |
| "loss": 2.6081, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5517241379310345, | |
| "grad_norm": 0.25939515233039856, | |
| "learning_rate": 9.932293779077216e-05, | |
| "loss": 2.4289, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.567398119122257, | |
| "grad_norm": 0.3675819933414459, | |
| "learning_rate": 9.763048594930502e-05, | |
| "loss": 2.6617, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5830721003134798, | |
| "grad_norm": 0.47927770018577576, | |
| "learning_rate": 9.59387129873787e-05, | |
| "loss": 2.661, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.5987460815047023, | |
| "grad_norm": 0.6929473876953125, | |
| "learning_rate": 9.42481036077749e-05, | |
| "loss": 2.6103, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.6144200626959249, | |
| "grad_norm": 0.9767149090766907, | |
| "learning_rate": 9.255914217990211e-05, | |
| "loss": 2.5943, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6300940438871474, | |
| "grad_norm": 0.24275819957256317, | |
| "learning_rate": 9.08723126010212e-05, | |
| "loss": 2.4676, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.64576802507837, | |
| "grad_norm": 0.3554080128669739, | |
| "learning_rate": 8.918809815760585e-05, | |
| "loss": 2.6107, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6614420062695925, | |
| "grad_norm": 0.5181464552879333, | |
| "learning_rate": 8.750698138687827e-05, | |
| "loss": 2.617, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.677115987460815, | |
| "grad_norm": 0.7479755282402039, | |
| "learning_rate": 8.582944393855941e-05, | |
| "loss": 2.6401, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6927899686520376, | |
| "grad_norm": 0.9864292740821838, | |
| "learning_rate": 8.415596643687363e-05, | |
| "loss": 2.6327, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.70846394984326, | |
| "grad_norm": 0.24126508831977844, | |
| "learning_rate": 8.248702834284693e-05, | |
| "loss": 2.4569, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 0.3491465449333191, | |
| "learning_rate": 8.082310781693865e-05, | |
| "loss": 2.6473, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7398119122257052, | |
| "grad_norm": 0.5092320442199707, | |
| "learning_rate": 7.916468158204576e-05, | |
| "loss": 2.6587, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7554858934169277, | |
| "grad_norm": 0.6617609262466431, | |
| "learning_rate": 7.751222478691884e-05, | |
| "loss": 2.6862, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.7711598746081505, | |
| "grad_norm": 1.0016579627990723, | |
| "learning_rate": 7.586621087002945e-05, | |
| "loss": 2.6226, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.786833855799373, | |
| "grad_norm": 0.2429930418729782, | |
| "learning_rate": 7.422711142392695e-05, | |
| "loss": 2.5479, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.8025078369905956, | |
| "grad_norm": 0.396308034658432, | |
| "learning_rate": 7.259539606012478e-05, | |
| "loss": 2.6005, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.5580389499664307, | |
| "learning_rate": 7.097153227455379e-05, | |
| "loss": 2.6714, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8338557993730409, | |
| "grad_norm": 0.7043899893760681, | |
| "learning_rate": 6.93559853136221e-05, | |
| "loss": 2.6417, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8495297805642634, | |
| "grad_norm": 0.9616154432296753, | |
| "learning_rate": 6.774921804091934e-05, | |
| "loss": 2.5572, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.865203761755486, | |
| "grad_norm": 0.2324633002281189, | |
| "learning_rate": 6.615169080460331e-05, | |
| "loss": 2.4474, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.8808777429467085, | |
| "grad_norm": 0.3579946756362915, | |
| "learning_rate": 6.456386130550782e-05, | |
| "loss": 2.6281, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.896551724137931, | |
| "grad_norm": 0.5167599320411682, | |
| "learning_rate": 6.298618446600856e-05, | |
| "loss": 2.6397, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.9122257053291536, | |
| "grad_norm": 0.7201157808303833, | |
| "learning_rate": 6.141911229968533e-05, | |
| "loss": 2.6389, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.9278996865203761, | |
| "grad_norm": 1.1355105638504028, | |
| "learning_rate": 5.9863093781817394e-05, | |
| "loss": 2.6135, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.9435736677115987, | |
| "grad_norm": 0.2417139708995819, | |
| "learning_rate": 5.831857472074956e-05, | |
| "loss": 2.456, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.9592476489028212, | |
| "grad_norm": 0.4348711669445038, | |
| "learning_rate": 5.6785997630165435e-05, | |
| "loss": 2.6756, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9749216300940438, | |
| "grad_norm": 0.6851107478141785, | |
| "learning_rate": 5.526580160230476e-05, | |
| "loss": 2.6338, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9905956112852663, | |
| "grad_norm": 1.0602185726165771, | |
| "learning_rate": 5.375842218216076e-05, | |
| "loss": 2.5925, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.006269592476489, | |
| "grad_norm": 0.23286418616771698, | |
| "learning_rate": 5.226429124269423e-05, | |
| "loss": 2.4199, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.0219435736677114, | |
| "grad_norm": 0.41238126158714294, | |
| "learning_rate": 5.078383686109926e-05, | |
| "loss": 2.5441, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.0376175548589344, | |
| "grad_norm": 0.5980674624443054, | |
| "learning_rate": 4.931748319615656e-05, | |
| "loss": 2.5002, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.053291536050157, | |
| "grad_norm": 0.8170694708824158, | |
| "learning_rate": 4.786565036670972e-05, | |
| "loss": 2.4228, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.0689655172413794, | |
| "grad_norm": 1.1204980611801147, | |
| "learning_rate": 4.642875433129854e-05, | |
| "loss": 2.3089, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.084639498432602, | |
| "grad_norm": 0.28670328855514526, | |
| "learning_rate": 4.500720676898452e-05, | |
| "loss": 2.2163, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.1003134796238245, | |
| "grad_norm": 0.45365315675735474, | |
| "learning_rate": 4.36014149614026e-05, | |
| "loss": 2.5476, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.115987460815047, | |
| "grad_norm": 0.589374840259552, | |
| "learning_rate": 4.221178167607226e-05, | |
| "loss": 2.4646, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.1316614420062696, | |
| "grad_norm": 0.9127137064933777, | |
| "learning_rate": 4.083870505100263e-05, | |
| "loss": 2.3819, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.147335423197492, | |
| "grad_norm": 1.167529582977295, | |
| "learning_rate": 3.948257848062351e-05, | |
| "loss": 2.3445, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.1630094043887147, | |
| "grad_norm": 0.26559045910835266, | |
| "learning_rate": 3.8143790503075403e-05, | |
| "loss": 2.228, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.1786833855799372, | |
| "grad_norm": 0.5015429258346558, | |
| "learning_rate": 3.6822724688891416e-05, | |
| "loss": 2.5579, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.19435736677116, | |
| "grad_norm": 0.7187398672103882, | |
| "learning_rate": 3.551975953110177e-05, | |
| "loss": 2.4841, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.2100313479623823, | |
| "grad_norm": 0.872721791267395, | |
| "learning_rate": 3.423526833679355e-05, | |
| "loss": 2.418, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.225705329153605, | |
| "grad_norm": 1.3087424039840698, | |
| "learning_rate": 3.296961912015598e-05, | |
| "loss": 2.2849, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.2413793103448274, | |
| "grad_norm": 0.31001710891723633, | |
| "learning_rate": 3.172317449704216e-05, | |
| "loss": 2.2705, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.2570532915360504, | |
| "grad_norm": 0.47308361530303955, | |
| "learning_rate": 3.0496291581077673e-05, | |
| "loss": 2.5207, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 0.6600085496902466, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 2.4747, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.2884012539184955, | |
| "grad_norm": 0.8648523092269897, | |
| "learning_rate": 2.8102611201675598e-05, | |
| "loss": 2.4407, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.304075235109718, | |
| "grad_norm": 1.2338064908981323, | |
| "learning_rate": 2.6936499541572445e-05, | |
| "loss": 2.3066, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.3197492163009406, | |
| "grad_norm": 0.2991880476474762, | |
| "learning_rate": 2.5791320998800695e-05, | |
| "loss": 2.2581, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.335423197492163, | |
| "grad_norm": 0.48960086703300476, | |
| "learning_rate": 2.4667403673665623e-05, | |
| "loss": 2.5032, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.3510971786833856, | |
| "grad_norm": 0.6952352523803711, | |
| "learning_rate": 2.3565069575010035e-05, | |
| "loss": 2.4815, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.366771159874608, | |
| "grad_norm": 0.9776033163070679, | |
| "learning_rate": 2.248463452795705e-05, | |
| "loss": 2.4033, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.3824451410658307, | |
| "grad_norm": 1.2295323610305786, | |
| "learning_rate": 2.142640808342429e-05, | |
| "loss": 2.2546, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.3981191222570533, | |
| "grad_norm": 0.2598329782485962, | |
| "learning_rate": 2.0390693429435627e-05, | |
| "loss": 2.2122, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.413793103448276, | |
| "grad_norm": 0.4593777358531952, | |
| "learning_rate": 1.9377787304256302e-05, | |
| "loss": 2.518, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.4294670846394983, | |
| "grad_norm": 0.6363881826400757, | |
| "learning_rate": 1.838797991137543e-05, | |
| "loss": 2.4564, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.445141065830721, | |
| "grad_norm": 0.9114318490028381, | |
| "learning_rate": 1.742155483636123e-05, | |
| "loss": 2.4088, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.4608150470219434, | |
| "grad_norm": 1.3736695051193237, | |
| "learning_rate": 1.6478788965611993e-05, | |
| "loss": 2.3423, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.476489028213166, | |
| "grad_norm": 0.2641979157924652, | |
| "learning_rate": 1.555995240702648e-05, | |
| "loss": 2.1546, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.492163009404389, | |
| "grad_norm": 0.5061512589454651, | |
| "learning_rate": 1.4665308412616596e-05, | |
| "loss": 2.5098, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.507836990595611, | |
| "grad_norm": 0.6692546606063843, | |
| "learning_rate": 1.3795113303084006e-05, | |
| "loss": 2.4528, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.523510971786834, | |
| "grad_norm": 0.9549630880355835, | |
| "learning_rate": 1.2949616394382802e-05, | |
| "loss": 2.4189, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.5391849529780566, | |
| "grad_norm": 1.450711727142334, | |
| "learning_rate": 1.2129059926289166e-05, | |
| "loss": 2.2557, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.554858934169279, | |
| "grad_norm": 0.2581474781036377, | |
| "learning_rate": 1.1333678992998043e-05, | |
| "loss": 2.2043, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.5705329153605017, | |
| "grad_norm": 0.4815433621406555, | |
| "learning_rate": 1.0563701475767462e-05, | |
| "loss": 2.5434, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.586206896551724, | |
| "grad_norm": 0.6861922144889832, | |
| "learning_rate": 9.819347977629202e-06, | |
| "loss": 2.4619, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.6018808777429467, | |
| "grad_norm": 0.9293055534362793, | |
| "learning_rate": 9.100831760184625e-06, | |
| "loss": 2.4539, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.6175548589341693, | |
| "grad_norm": 1.2631028890609741, | |
| "learning_rate": 8.408358682504147e-06, | |
| "loss": 2.2262, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.633228840125392, | |
| "grad_norm": 0.3051944673061371, | |
| "learning_rate": 7.742127142147337e-06, | |
| "loss": 2.2413, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.6489028213166144, | |
| "grad_norm": 0.5327479243278503, | |
| "learning_rate": 7.102328018320858e-06, | |
| "loss": 2.5487, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.664576802507837, | |
| "grad_norm": 0.6830099821090698, | |
| "learning_rate": 6.489144617190501e-06, | |
| "loss": 2.492, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6802507836990594, | |
| "grad_norm": 1.006087303161621, | |
| "learning_rate": 5.902752619362861e-06, | |
| "loss": 2.4549, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.695924764890282, | |
| "grad_norm": 1.3168909549713135, | |
| "learning_rate": 5.343320029551868e-06, | |
| "loss": 2.2644, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.7115987460815045, | |
| "grad_norm": 0.29884764552116394, | |
| "learning_rate": 4.811007128444445e-06, | |
| "loss": 2.3007, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.4936012029647827, | |
| "learning_rate": 4.305966426779118e-06, | |
| "loss": 2.502, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.7429467084639496, | |
| "grad_norm": 0.689198911190033, | |
| "learning_rate": 3.828342621650882e-06, | |
| "loss": 2.4674, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.7586206896551726, | |
| "grad_norm": 0.9643833041191101, | |
| "learning_rate": 3.3782725550545625e-06, | |
| "loss": 2.4213, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.774294670846395, | |
| "grad_norm": 1.3325724601745605, | |
| "learning_rate": 2.9558851746788517e-06, | |
| "loss": 2.2601, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.7899686520376177, | |
| "grad_norm": 0.2799229621887207, | |
| "learning_rate": 2.561301496962043e-06, | |
| "loss": 2.2256, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.80564263322884, | |
| "grad_norm": 0.5209997296333313, | |
| "learning_rate": 2.194634572420029e-06, | |
| "loss": 2.5132, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.8213166144200628, | |
| "grad_norm": 0.6882754564285278, | |
| "learning_rate": 1.8559894532568122e-06, | |
| "loss": 2.428, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.8369905956112853, | |
| "grad_norm": 0.9445529580116272, | |
| "learning_rate": 1.545463163266303e-06, | |
| "loss": 2.4341, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.852664576802508, | |
| "grad_norm": 1.3681193590164185, | |
| "learning_rate": 1.263144670034555e-06, | |
| "loss": 2.2468, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.8683385579937304, | |
| "grad_norm": 0.2578428089618683, | |
| "learning_rate": 1.0091148594499666e-06, | |
| "loss": 2.2084, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.884012539184953, | |
| "grad_norm": 0.5128017663955688, | |
| "learning_rate": 7.83446512529007e-07, | |
| "loss": 2.5086, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.8996865203761755, | |
| "grad_norm": 0.7125623226165771, | |
| "learning_rate": 5.862042845640403e-07, | |
| "loss": 2.4374, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.915360501567398, | |
| "grad_norm": 1.0510890483856201, | |
| "learning_rate": 4.174446865992332e-07, | |
| "loss": 2.4102, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.9310344827586206, | |
| "grad_norm": 1.4845973253250122, | |
| "learning_rate": 2.7721606923978293e-07, | |
| "loss": 2.2825, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.946708463949843, | |
| "grad_norm": 0.3054816722869873, | |
| "learning_rate": 1.6555860879919892e-07, | |
| "loss": 2.1791, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.962382445141066, | |
| "grad_norm": 0.6139517426490784, | |
| "learning_rate": 8.250429578855467e-08, | |
| "loss": 2.5106, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.978056426332288, | |
| "grad_norm": 0.9643613696098328, | |
| "learning_rate": 2.8076925751008286e-08, | |
| "loss": 2.4253, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.993730407523511, | |
| "grad_norm": 1.5458600521087646, | |
| "learning_rate": 2.292092444255989e-09, | |
| "loss": 2.2764, | |
| "step": 1910 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1914, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.5314350752283034e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |