| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1535, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03257328990228013, | |
| "grad_norm": 9.89926528930664, | |
| "learning_rate": 2.5974025974025972e-05, | |
| "loss": 2.0334, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06514657980456026, | |
| "grad_norm": 7.768723964691162, | |
| "learning_rate": 5.1948051948051944e-05, | |
| "loss": 0.5571, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09771986970684039, | |
| "grad_norm": 1.5563685894012451, | |
| "learning_rate": 7.792207792207793e-05, | |
| "loss": 0.1878, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13029315960912052, | |
| "grad_norm": 1.674676537513733, | |
| "learning_rate": 0.00010389610389610389, | |
| "loss": 0.1257, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16286644951140064, | |
| "grad_norm": 1.1463338136672974, | |
| "learning_rate": 0.00012987012987012987, | |
| "loss": 0.1011, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19543973941368079, | |
| "grad_norm": 1.0946776866912842, | |
| "learning_rate": 0.00015584415584415587, | |
| "loss": 0.0972, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2280130293159609, | |
| "grad_norm": 0.7055681347846985, | |
| "learning_rate": 0.00018181818181818183, | |
| "loss": 0.0724, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.26058631921824105, | |
| "grad_norm": 0.51561439037323, | |
| "learning_rate": 0.0001999979107245606, | |
| "loss": 0.0641, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2931596091205212, | |
| "grad_norm": 1.2206898927688599, | |
| "learning_rate": 0.0001999607704786645, | |
| "loss": 0.0727, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3257328990228013, | |
| "grad_norm": 0.7886886596679688, | |
| "learning_rate": 0.00019987722173728587, | |
| "loss": 0.0735, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3583061889250814, | |
| "grad_norm": 0.4768007695674896, | |
| "learning_rate": 0.00019974730328935534, | |
| "loss": 0.0582, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.39087947882736157, | |
| "grad_norm": 0.5169155597686768, | |
| "learning_rate": 0.00019957107545173192, | |
| "loss": 0.0593, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4234527687296417, | |
| "grad_norm": 0.359069287776947, | |
| "learning_rate": 0.00019934862004120004, | |
| "loss": 0.0358, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4560260586319218, | |
| "grad_norm": 0.6432995200157166, | |
| "learning_rate": 0.00019908004033648453, | |
| "loss": 0.0382, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.48859934853420195, | |
| "grad_norm": 1.358920693397522, | |
| "learning_rate": 0.00019876546103030196, | |
| "loss": 0.0519, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5211726384364821, | |
| "grad_norm": 0.8731968402862549, | |
| "learning_rate": 0.00019840502817146966, | |
| "loss": 0.0582, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5537459283387622, | |
| "grad_norm": 0.8558807969093323, | |
| "learning_rate": 0.00019799890909710013, | |
| "loss": 0.0383, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5863192182410424, | |
| "grad_norm": 0.5326903462409973, | |
| "learning_rate": 0.00019754729235491207, | |
| "loss": 0.0326, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6188925081433225, | |
| "grad_norm": 0.434563547372818, | |
| "learning_rate": 0.0001970503876156937, | |
| "loss": 0.0351, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6514657980456026, | |
| "grad_norm": 0.778796374797821, | |
| "learning_rate": 0.00019650842557595967, | |
| "loss": 0.0362, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6840390879478827, | |
| "grad_norm": 0.49492743611335754, | |
| "learning_rate": 0.00019592165785084603, | |
| "loss": 0.0363, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7166123778501629, | |
| "grad_norm": 0.640010416507721, | |
| "learning_rate": 0.00019529035685729391, | |
| "loss": 0.0403, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.749185667752443, | |
| "grad_norm": 0.573532223701477, | |
| "learning_rate": 0.00019461481568757506, | |
| "loss": 0.0394, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7817589576547231, | |
| "grad_norm": 0.3103967308998108, | |
| "learning_rate": 0.00019389534797321884, | |
| "loss": 0.0318, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8143322475570033, | |
| "grad_norm": 0.7902340292930603, | |
| "learning_rate": 0.00019313228773940345, | |
| "loss": 0.0355, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8469055374592834, | |
| "grad_norm": 0.38804200291633606, | |
| "learning_rate": 0.00019232598924987903, | |
| "loss": 0.0353, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8794788273615635, | |
| "grad_norm": 0.540195643901825, | |
| "learning_rate": 0.0001914768268424946, | |
| "loss": 0.0322, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9120521172638436, | |
| "grad_norm": 0.6464592218399048, | |
| "learning_rate": 0.00019058519475540538, | |
| "loss": 0.0369, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9446254071661238, | |
| "grad_norm": 0.36654534935951233, | |
| "learning_rate": 0.00018965150694404094, | |
| "loss": 0.0258, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9771986970684039, | |
| "grad_norm": 0.6431544423103333, | |
| "learning_rate": 0.00018867619688891937, | |
| "loss": 0.0317, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.009771986970684, | |
| "grad_norm": 0.47796347737312317, | |
| "learning_rate": 0.0001876597173943965, | |
| "loss": 0.0256, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0423452768729642, | |
| "grad_norm": 0.6077843904495239, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 0.0299, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0749185667752443, | |
| "grad_norm": 0.5036475658416748, | |
| "learning_rate": 0.00018550515665355247, | |
| "loss": 0.0341, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1074918566775245, | |
| "grad_norm": 0.7570111751556396, | |
| "learning_rate": 0.000184368075698865, | |
| "loss": 0.0259, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1400651465798046, | |
| "grad_norm": 0.3445906341075897, | |
| "learning_rate": 0.00018319182542364117, | |
| "loss": 0.0253, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1726384364820848, | |
| "grad_norm": 0.43921205401420593, | |
| "learning_rate": 0.00018197695192216702, | |
| "loss": 0.027, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.205211726384365, | |
| "grad_norm": 0.390203595161438, | |
| "learning_rate": 0.00018072401922022117, | |
| "loss": 0.0196, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.237785016286645, | |
| "grad_norm": 0.3478577733039856, | |
| "learning_rate": 0.0001794336090132164, | |
| "loss": 0.0321, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2703583061889252, | |
| "grad_norm": 0.6172159910202026, | |
| "learning_rate": 0.00017810632039613736, | |
| "loss": 0.0398, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3029315960912053, | |
| "grad_norm": 0.6841723322868347, | |
| "learning_rate": 0.00017674276958540072, | |
| "loss": 0.0339, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3355048859934853, | |
| "grad_norm": 0.26006555557250977, | |
| "learning_rate": 0.00017534358963276607, | |
| "loss": 0.0293, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3680781758957654, | |
| "grad_norm": 0.4724225699901581, | |
| "learning_rate": 0.00017390943013143083, | |
| "loss": 0.0346, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4006514657980456, | |
| "grad_norm": 0.660692572593689, | |
| "learning_rate": 0.0001724409569144455, | |
| "loss": 0.0299, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4332247557003257, | |
| "grad_norm": 0.5522527098655701, | |
| "learning_rate": 0.0001709388517455893, | |
| "loss": 0.0355, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4657980456026058, | |
| "grad_norm": 0.8123102188110352, | |
| "learning_rate": 0.00016940381200284972, | |
| "loss": 0.0303, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.498371335504886, | |
| "grad_norm": 1.0083576440811157, | |
| "learning_rate": 0.0001678365503546528, | |
| "loss": 0.0279, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5309446254071661, | |
| "grad_norm": 0.4633665680885315, | |
| "learning_rate": 0.0001662377944289948, | |
| "loss": 0.0299, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5635179153094463, | |
| "grad_norm": 0.42941364645957947, | |
| "learning_rate": 0.0001646082864756282, | |
| "loss": 0.0218, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.5960912052117264, | |
| "grad_norm": 0.2960715591907501, | |
| "learning_rate": 0.00016294878302145987, | |
| "loss": 0.025, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.6286644951140063, | |
| "grad_norm": 0.4294084310531616, | |
| "learning_rate": 0.0001612600545193203, | |
| "loss": 0.0175, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6612377850162865, | |
| "grad_norm": 0.5202115178108215, | |
| "learning_rate": 0.00015954288499026782, | |
| "loss": 0.0249, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.6938110749185666, | |
| "grad_norm": 0.3897187113761902, | |
| "learning_rate": 0.0001577980716595934, | |
| "loss": 0.0178, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7263843648208468, | |
| "grad_norm": 0.33642974495887756, | |
| "learning_rate": 0.00015602642458669528, | |
| "loss": 0.0259, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.758957654723127, | |
| "grad_norm": 0.37540072202682495, | |
| "learning_rate": 0.0001542287662889948, | |
| "loss": 0.0261, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.791530944625407, | |
| "grad_norm": 0.36353737115859985, | |
| "learning_rate": 0.00015240593136006897, | |
| "loss": 0.0194, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8241042345276872, | |
| "grad_norm": 0.29128432273864746, | |
| "learning_rate": 0.0001505587660821759, | |
| "loss": 0.0201, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8566775244299674, | |
| "grad_norm": 0.5382766723632812, | |
| "learning_rate": 0.0001486881280333539, | |
| "loss": 0.0259, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8892508143322475, | |
| "grad_norm": 0.3132719397544861, | |
| "learning_rate": 0.00014679488568927616, | |
| "loss": 0.0246, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9218241042345277, | |
| "grad_norm": 0.27731725573539734, | |
| "learning_rate": 0.00014487991802004623, | |
| "loss": 0.0176, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9543973941368078, | |
| "grad_norm": 0.45612412691116333, | |
| "learning_rate": 0.0001429441140821209, | |
| "loss": 0.0218, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.986970684039088, | |
| "grad_norm": 0.36944088339805603, | |
| "learning_rate": 0.00014098837260555084, | |
| "loss": 0.023, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.019543973941368, | |
| "grad_norm": 0.33426433801651, | |
| "learning_rate": 0.0001390136015767295, | |
| "loss": 0.0288, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0521172638436482, | |
| "grad_norm": 0.4996889531612396, | |
| "learning_rate": 0.00013702071781684517, | |
| "loss": 0.0171, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0846905537459284, | |
| "grad_norm": 0.34143733978271484, | |
| "learning_rate": 0.00013501064655623094, | |
| "loss": 0.0171, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.1172638436482085, | |
| "grad_norm": 0.33777856826782227, | |
| "learning_rate": 0.00013298432100481079, | |
| "loss": 0.0189, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.1498371335504887, | |
| "grad_norm": 0.6728042960166931, | |
| "learning_rate": 0.0001309426819188409, | |
| "loss": 0.0243, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.182410423452769, | |
| "grad_norm": 0.4677506685256958, | |
| "learning_rate": 0.0001288866771641474, | |
| "loss": 0.0129, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.214983713355049, | |
| "grad_norm": 0.48820221424102783, | |
| "learning_rate": 0.00012681726127606376, | |
| "loss": 0.0276, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.247557003257329, | |
| "grad_norm": 0.3881731927394867, | |
| "learning_rate": 0.000124735395016271, | |
| "loss": 0.0158, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.2801302931596092, | |
| "grad_norm": 0.27123430371284485, | |
| "learning_rate": 0.00012264204492674815, | |
| "loss": 0.0167, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3127035830618894, | |
| "grad_norm": 0.2786751091480255, | |
| "learning_rate": 0.0001205381828810382, | |
| "loss": 0.0194, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.3452768729641695, | |
| "grad_norm": 0.23343567550182343, | |
| "learning_rate": 0.00011842478563303952, | |
| "loss": 0.0155, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.3778501628664497, | |
| "grad_norm": 0.22697576880455017, | |
| "learning_rate": 0.00011630283436353098, | |
| "loss": 0.0185, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.41042345276873, | |
| "grad_norm": 0.38111135363578796, | |
| "learning_rate": 0.00011417331422464205, | |
| "loss": 0.0151, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.44299674267101, | |
| "grad_norm": 0.3761187493801117, | |
| "learning_rate": 0.00011203721388247923, | |
| "loss": 0.0185, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.47557003257329, | |
| "grad_norm": 0.29886382818222046, | |
| "learning_rate": 0.00010989552505812072, | |
| "loss": 0.0093, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.5081433224755703, | |
| "grad_norm": 0.308753103017807, | |
| "learning_rate": 0.0001077492420671931, | |
| "loss": 0.0182, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.5407166123778504, | |
| "grad_norm": 0.26204925775527954, | |
| "learning_rate": 0.00010559936135824322, | |
| "loss": 0.0192, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.5732899022801305, | |
| "grad_norm": 0.433326780796051, | |
| "learning_rate": 0.00010344688105012005, | |
| "loss": 0.0154, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.6058631921824107, | |
| "grad_norm": 0.24838528037071228, | |
| "learning_rate": 0.00010129280046858086, | |
| "loss": 0.0175, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.6384364820846904, | |
| "grad_norm": 0.2929915487766266, | |
| "learning_rate": 9.913811968233716e-05, | |
| "loss": 0.0129, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.6710097719869705, | |
| "grad_norm": 0.30677530169487, | |
| "learning_rate": 9.69838390387558e-05, | |
| "loss": 0.0315, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.7035830618892507, | |
| "grad_norm": 0.34507474303245544, | |
| "learning_rate": 9.483095869943055e-05, | |
| "loss": 0.0151, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.736156351791531, | |
| "grad_norm": 0.2926924526691437, | |
| "learning_rate": 9.268047817583998e-05, | |
| "loss": 0.0112, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.768729641693811, | |
| "grad_norm": 0.3242110311985016, | |
| "learning_rate": 9.053339586530723e-05, | |
| "loss": 0.0133, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.801302931596091, | |
| "grad_norm": 0.274239718914032, | |
| "learning_rate": 8.839070858747697e-05, | |
| "loss": 0.02, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.8338762214983713, | |
| "grad_norm": 0.186782106757164, | |
| "learning_rate": 8.625341112152487e-05, | |
| "loss": 0.0121, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.8664495114006514, | |
| "grad_norm": 0.1265726238489151, | |
| "learning_rate": 8.412249574431428e-05, | |
| "loss": 0.0134, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.8990228013029316, | |
| "grad_norm": 0.3958815634250641, | |
| "learning_rate": 8.199895176971488e-05, | |
| "loss": 0.0113, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.9315960912052117, | |
| "grad_norm": 0.19703464210033417, | |
| "learning_rate": 7.988376508929676e-05, | |
| "loss": 0.0144, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.964169381107492, | |
| "grad_norm": 0.2459426373243332, | |
| "learning_rate": 7.777791771461332e-05, | |
| "loss": 0.0138, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.996742671009772, | |
| "grad_norm": 0.2412751019001007, | |
| "learning_rate": 7.568238732128585e-05, | |
| "loss": 0.0227, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.029315960912052, | |
| "grad_norm": 0.38352710008621216, | |
| "learning_rate": 7.359814679510065e-05, | |
| "loss": 0.0245, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.0618892508143323, | |
| "grad_norm": 0.33591127395629883, | |
| "learning_rate": 7.152616378033042e-05, | |
| "loss": 0.0091, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.0944625407166124, | |
| "grad_norm": 0.2634866237640381, | |
| "learning_rate": 6.94674002304887e-05, | |
| "loss": 0.01, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.1270358306188926, | |
| "grad_norm": 0.17260074615478516, | |
| "learning_rate": 6.742281196172663e-05, | |
| "loss": 0.0156, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.1596091205211727, | |
| "grad_norm": 0.20153683423995972, | |
| "learning_rate": 6.539334820907888e-05, | |
| "loss": 0.0091, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.192182410423453, | |
| "grad_norm": 0.4081243872642517, | |
| "learning_rate": 6.337995118576521e-05, | |
| "loss": 0.0101, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.224755700325733, | |
| "grad_norm": 0.24244199693202972, | |
| "learning_rate": 6.138355564575169e-05, | |
| "loss": 0.0103, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.257328990228013, | |
| "grad_norm": 0.17980413138866425, | |
| "learning_rate": 5.940508844977537e-05, | |
| "loss": 0.0111, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2899022801302933, | |
| "grad_norm": 0.13628683984279633, | |
| "learning_rate": 5.744546813503328e-05, | |
| "loss": 0.0115, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.3224755700325734, | |
| "grad_norm": 0.408488005399704, | |
| "learning_rate": 5.550560448873575e-05, | |
| "loss": 0.0146, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.3550488599348536, | |
| "grad_norm": 0.23400142788887024, | |
| "learning_rate": 5.358639812572244e-05, | |
| "loss": 0.0114, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.3876221498371337, | |
| "grad_norm": 0.2922910749912262, | |
| "learning_rate": 5.168874007033615e-05, | |
| "loss": 0.0094, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.420195439739414, | |
| "grad_norm": 0.20555023849010468, | |
| "learning_rate": 4.9813511342749805e-05, | |
| "loss": 0.0122, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.4527687296416936, | |
| "grad_norm": 0.17685680091381073, | |
| "learning_rate": 4.7961582549937675e-05, | |
| "loss": 0.0182, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.4853420195439737, | |
| "grad_norm": 0.14065317809581757, | |
| "learning_rate": 4.6133813481481246e-05, | |
| "loss": 0.0097, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.517915309446254, | |
| "grad_norm": 0.34961211681365967, | |
| "learning_rate": 4.433105271039721e-05, | |
| "loss": 0.012, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.550488599348534, | |
| "grad_norm": 0.38357603549957275, | |
| "learning_rate": 4.255413719917294e-05, | |
| "loss": 0.0102, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.583061889250814, | |
| "grad_norm": 0.25049689412117004, | |
| "learning_rate": 4.080389191119241e-05, | |
| "loss": 0.0113, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.6156351791530943, | |
| "grad_norm": 0.23103076219558716, | |
| "learning_rate": 3.9081129427732774e-05, | |
| "loss": 0.0096, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.6482084690553744, | |
| "grad_norm": 0.5483614206314087, | |
| "learning_rate": 3.7386649570709644e-05, | |
| "loss": 0.0113, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.6807817589576546, | |
| "grad_norm": 0.14837254583835602, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 0.0078, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.7133550488599347, | |
| "grad_norm": 0.1685313582420349, | |
| "learning_rate": 3.408567100493787e-05, | |
| "loss": 0.0098, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.745928338762215, | |
| "grad_norm": 0.6023198366165161, | |
| "learning_rate": 3.248070483188426e-05, | |
| "loss": 0.0102, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.778501628664495, | |
| "grad_norm": 0.11718238145112991, | |
| "learning_rate": 3.090708564515124e-05, | |
| "loss": 0.0065, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.811074918566775, | |
| "grad_norm": 0.23708564043045044, | |
| "learning_rate": 2.936554402433087e-05, | |
| "loss": 0.0122, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.8436482084690553, | |
| "grad_norm": 0.4252511262893677, | |
| "learning_rate": 2.7856795656457257e-05, | |
| "loss": 0.0098, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.8762214983713354, | |
| "grad_norm": 0.3115089237689972, | |
| "learning_rate": 2.6381541003736486e-05, | |
| "loss": 0.0105, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.9087947882736156, | |
| "grad_norm": 0.39385488629341125, | |
| "learning_rate": 2.494046497834518e-05, | |
| "loss": 0.0114, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.9413680781758957, | |
| "grad_norm": 0.29438289999961853, | |
| "learning_rate": 2.3534236624448302e-05, | |
| "loss": 0.0148, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.973941368078176, | |
| "grad_norm": 0.25224828720092773, | |
| "learning_rate": 2.2163508807583998e-05, | |
| "loss": 0.0114, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.006514657980456, | |
| "grad_norm": 0.2634766697883606, | |
| "learning_rate": 2.082891791155954e-05, | |
| "loss": 0.0091, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.039087947882736, | |
| "grad_norm": 0.4244247376918793, | |
| "learning_rate": 1.9531083542999317e-05, | |
| "loss": 0.012, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.071661237785016, | |
| "grad_norm": 0.24010491371154785, | |
| "learning_rate": 1.8270608243681953e-05, | |
| "loss": 0.0078, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.1042345276872965, | |
| "grad_norm": 0.2347911149263382, | |
| "learning_rate": 1.7048077210799772e-05, | |
| "loss": 0.0061, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.136807817589577, | |
| "grad_norm": 0.16366265714168549, | |
| "learning_rate": 1.5864058025271246e-05, | |
| "loss": 0.0083, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.169381107491857, | |
| "grad_norm": 0.12379926443099976, | |
| "learning_rate": 1.47191003882317e-05, | |
| "loss": 0.0116, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.201954397394137, | |
| "grad_norm": 0.2692829370498657, | |
| "learning_rate": 1.3613735865825305e-05, | |
| "loss": 0.0084, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.234527687296417, | |
| "grad_norm": 0.180028036236763, | |
| "learning_rate": 1.2548477642416256e-05, | |
| "loss": 0.0073, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.267100977198697, | |
| "grad_norm": 0.1352684646844864, | |
| "learning_rate": 1.1523820282334219e-05, | |
| "loss": 0.0086, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.299674267100977, | |
| "grad_norm": 0.2455337941646576, | |
| "learning_rate": 1.0540239500264516e-05, | |
| "loss": 0.0096, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.3322475570032575, | |
| "grad_norm": 0.23969142138957977, | |
| "learning_rate": 9.598191940389256e-06, | |
| "loss": 0.013, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.364820846905538, | |
| "grad_norm": 0.37968650460243225, | |
| "learning_rate": 8.698114964382598e-06, | |
| "loss": 0.0141, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.397394136807818, | |
| "grad_norm": 0.07284022867679596, | |
| "learning_rate": 7.840426448358085e-06, | |
| "loss": 0.0069, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.429967426710098, | |
| "grad_norm": 0.25828060507774353, | |
| "learning_rate": 7.025524588862542e-06, | |
| "loss": 0.0147, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.462540716612378, | |
| "grad_norm": 0.1627136766910553, | |
| "learning_rate": 6.253787718006498e-06, | |
| "loss": 0.0102, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.495114006514658, | |
| "grad_norm": 0.20821145176887512, | |
| "learning_rate": 5.525574127817046e-06, | |
| "loss": 0.0091, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.527687296416938, | |
| "grad_norm": 0.1315605640411377, | |
| "learning_rate": 4.841221903894633e-06, | |
| "loss": 0.011, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.5602605863192185, | |
| "grad_norm": 0.10593540966510773, | |
| "learning_rate": 4.20104876845111e-06, | |
| "loss": 0.0057, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.592833876221499, | |
| "grad_norm": 0.20290839672088623, | |
| "learning_rate": 3.605351932801693e-06, | |
| "loss": 0.0151, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.625407166123779, | |
| "grad_norm": 0.1976938247680664, | |
| "learning_rate": 3.0544079593795573e-06, | |
| "loss": 0.0119, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.657980456026059, | |
| "grad_norm": 0.22302116453647614, | |
| "learning_rate": 2.548472633337007e-06, | |
| "loss": 0.008, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.690553745928339, | |
| "grad_norm": 0.1248151957988739, | |
| "learning_rate": 2.0877808437928637e-06, | |
| "loss": 0.0078, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.723127035830619, | |
| "grad_norm": 0.20839862525463104, | |
| "learning_rate": 1.6725464747811447e-06, | |
| "loss": 0.0132, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.755700325732899, | |
| "grad_norm": 0.2530406713485718, | |
| "learning_rate": 1.3029623059517493e-06, | |
| "loss": 0.0075, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.7882736156351795, | |
| "grad_norm": 0.11152759194374084, | |
| "learning_rate": 9.791999230692629e-07, | |
| "loss": 0.0064, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.82084690553746, | |
| "grad_norm": 0.20677945017814636, | |
| "learning_rate": 7.014096383512802e-07, | |
| "loss": 0.0083, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.85342019543974, | |
| "grad_norm": 0.186812624335289, | |
| "learning_rate": 4.6972042068341714e-07, | |
| "loss": 0.0074, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.88599348534202, | |
| "grad_norm": 0.07393082976341248, | |
| "learning_rate": 2.8423983574328295e-07, | |
| "loss": 0.0075, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.918566775244299, | |
| "grad_norm": 0.06852855533361435, | |
| "learning_rate": 1.4505399606130621e-07, | |
| "loss": 0.0073, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.95114006514658, | |
| "grad_norm": 0.1197730079293251, | |
| "learning_rate": 5.2227521041470216e-08, | |
| "loss": 0.0057, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.9837133550488595, | |
| "grad_norm": 0.2546689212322235, | |
| "learning_rate": 5.803506960722072e-09, | |
| "loss": 0.0097, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1535, | |
| "total_flos": 5.19646304846448e+16, | |
| "train_loss": 0.04019785498296205, | |
| "train_runtime": 734.2052, | |
| "train_samples_per_second": 33.451, | |
| "train_steps_per_second": 2.091 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1535, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 5.19646304846448e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |