| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.990881458966565, | |
| "eval_steps": 500, | |
| "global_step": 1642, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.030395136778115502, | |
| "grad_norm": 10.756904602050781, | |
| "learning_rate": 2.4096385542168677e-05, | |
| "loss": 1.6027, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.060790273556231005, | |
| "grad_norm": 9.451399803161621, | |
| "learning_rate": 4.8192771084337354e-05, | |
| "loss": 0.4287, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0911854103343465, | |
| "grad_norm": 1.1107724905014038, | |
| "learning_rate": 7.228915662650602e-05, | |
| "loss": 0.0996, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12158054711246201, | |
| "grad_norm": 0.6546974778175354, | |
| "learning_rate": 9.638554216867471e-05, | |
| "loss": 0.0405, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1519756838905775, | |
| "grad_norm": 0.4774630665779114, | |
| "learning_rate": 0.0001204819277108434, | |
| "loss": 0.0249, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.182370820668693, | |
| "grad_norm": 1.0936344861984253, | |
| "learning_rate": 0.00014457831325301204, | |
| "loss": 0.027, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2127659574468085, | |
| "grad_norm": 0.6960570216178894, | |
| "learning_rate": 0.00016867469879518074, | |
| "loss": 0.0187, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.24316109422492402, | |
| "grad_norm": 0.45391130447387695, | |
| "learning_rate": 0.00019277108433734942, | |
| "loss": 0.0143, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2735562310030395, | |
| "grad_norm": 0.3450222611427307, | |
| "learning_rate": 0.0001999900512985548, | |
| "loss": 0.0122, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.303951367781155, | |
| "grad_norm": 0.8266738653182983, | |
| "learning_rate": 0.00019994132773027597, | |
| "loss": 0.0186, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3343465045592705, | |
| "grad_norm": 0.4097825586795807, | |
| "learning_rate": 0.00019985202174272956, | |
| "loss": 0.014, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.364741641337386, | |
| "grad_norm": 0.6124298572540283, | |
| "learning_rate": 0.00019972216959972274, | |
| "loss": 0.0156, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3951367781155015, | |
| "grad_norm": 0.5567657351493835, | |
| "learning_rate": 0.00019955182402933334, | |
| "loss": 0.0078, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.425531914893617, | |
| "grad_norm": 0.6027739644050598, | |
| "learning_rate": 0.00019934105420249908, | |
| "loss": 0.0089, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.45592705167173253, | |
| "grad_norm": 0.5931612849235535, | |
| "learning_rate": 0.00019908994570492993, | |
| "loss": 0.0143, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.48632218844984804, | |
| "grad_norm": 0.44545474648475647, | |
| "learning_rate": 0.00019879860050235469, | |
| "loss": 0.0123, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5167173252279635, | |
| "grad_norm": 0.29651954770088196, | |
| "learning_rate": 0.0001984671368991169, | |
| "loss": 0.008, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.547112462006079, | |
| "grad_norm": 0.19712242484092712, | |
| "learning_rate": 0.0001980956894901356, | |
| "loss": 0.006, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5775075987841946, | |
| "grad_norm": 0.4244577884674072, | |
| "learning_rate": 0.00019768440910625162, | |
| "loss": 0.0096, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.60790273556231, | |
| "grad_norm": 0.5452653169631958, | |
| "learning_rate": 0.00019723346275298052, | |
| "loss": 0.0128, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6382978723404256, | |
| "grad_norm": 0.5842880010604858, | |
| "learning_rate": 0.00019674303354269833, | |
| "loss": 0.0115, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.668693009118541, | |
| "grad_norm": 0.5381531715393066, | |
| "learning_rate": 0.00019621332062028617, | |
| "loss": 0.0102, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6990881458966566, | |
| "grad_norm": 0.7361140847206116, | |
| "learning_rate": 0.00019564453908226515, | |
| "loss": 0.0098, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.729483282674772, | |
| "grad_norm": 0.45423439145088196, | |
| "learning_rate": 0.00019503691988945367, | |
| "loss": 0.0097, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7598784194528876, | |
| "grad_norm": 0.38471490144729614, | |
| "learning_rate": 0.000194390709773183, | |
| "loss": 0.0117, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.790273556231003, | |
| "grad_norm": 0.36030113697052, | |
| "learning_rate": 0.0001937061711351089, | |
| "loss": 0.009, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8206686930091185, | |
| "grad_norm": 0.3028634786605835, | |
| "learning_rate": 0.00019298358194066016, | |
| "loss": 0.0069, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 0.278566837310791, | |
| "learning_rate": 0.0001922232356061672, | |
| "loss": 0.011, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.8814589665653495, | |
| "grad_norm": 0.3821546137332916, | |
| "learning_rate": 0.00019142544087971693, | |
| "loss": 0.0054, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9118541033434651, | |
| "grad_norm": 0.26465708017349243, | |
| "learning_rate": 0.00019059052171578155, | |
| "loss": 0.0077, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9422492401215805, | |
| "grad_norm": 0.4308549463748932, | |
| "learning_rate": 0.00018971881714367295, | |
| "loss": 0.0057, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.9726443768996961, | |
| "grad_norm": 0.226839080452919, | |
| "learning_rate": 0.0001888106811298755, | |
| "loss": 0.0077, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0030395136778116, | |
| "grad_norm": 0.39627695083618164, | |
| "learning_rate": 0.00018786648243431363, | |
| "loss": 0.0115, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.033434650455927, | |
| "grad_norm": 0.6822514533996582, | |
| "learning_rate": 0.00018688660446061235, | |
| "loss": 0.0103, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.0638297872340425, | |
| "grad_norm": 0.34325116872787476, | |
| "learning_rate": 0.00018587144510041128, | |
| "loss": 0.0095, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.094224924012158, | |
| "grad_norm": 1.1588419675827026, | |
| "learning_rate": 0.00018482141657179594, | |
| "loss": 0.0199, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.1246200607902737, | |
| "grad_norm": 0.5164496898651123, | |
| "learning_rate": 0.00018373694525191138, | |
| "loss": 0.0113, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.155015197568389, | |
| "grad_norm": 0.38176482915878296, | |
| "learning_rate": 0.00018261847150382644, | |
| "loss": 0.0109, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.1854103343465046, | |
| "grad_norm": 0.43293264508247375, | |
| "learning_rate": 0.000181466449497719, | |
| "loss": 0.0126, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.21580547112462, | |
| "grad_norm": 0.2756524980068207, | |
| "learning_rate": 0.00018028134702645425, | |
| "loss": 0.0057, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2462006079027357, | |
| "grad_norm": 0.3666749894618988, | |
| "learning_rate": 0.00017906364531563185, | |
| "loss": 0.0047, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.2765957446808511, | |
| "grad_norm": 0.3208726942539215, | |
| "learning_rate": 0.00017781383882817811, | |
| "loss": 0.0153, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.3069908814589666, | |
| "grad_norm": 0.36565855145454407, | |
| "learning_rate": 0.00017653243506356332, | |
| "loss": 0.0078, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.337386018237082, | |
| "grad_norm": 0.30531516671180725, | |
| "learning_rate": 0.00017521995435172504, | |
| "loss": 0.0086, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.3677811550151975, | |
| "grad_norm": 0.4445202648639679, | |
| "learning_rate": 0.00017387692964178198, | |
| "loss": 0.0046, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.3981762917933132, | |
| "grad_norm": 0.30582693219184875, | |
| "learning_rate": 0.00017250390628562303, | |
| "loss": 0.0044, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.22891288995742798, | |
| "learning_rate": 0.00017110144181646072, | |
| "loss": 0.0084, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.458966565349544, | |
| "grad_norm": 0.34802454710006714, | |
| "learning_rate": 0.00016967010572243758, | |
| "loss": 0.0053, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.4893617021276595, | |
| "grad_norm": 0.4866830110549927, | |
| "learning_rate": 0.00016821047921537858, | |
| "loss": 0.0082, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.5197568389057752, | |
| "grad_norm": 0.11244069784879684, | |
| "learning_rate": 0.0001667231549947828, | |
| "loss": 0.0043, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.5501519756838906, | |
| "grad_norm": 0.37716448307037354, | |
| "learning_rate": 0.00016520873700715045, | |
| "loss": 0.0075, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.580547112462006, | |
| "grad_norm": 0.3906088173389435, | |
| "learning_rate": 0.00016366784020074282, | |
| "loss": 0.0051, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.6109422492401215, | |
| "grad_norm": 0.400751531124115, | |
| "learning_rate": 0.00016210109027587494, | |
| "loss": 0.0066, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.641337386018237, | |
| "grad_norm": 0.3523465096950531, | |
| "learning_rate": 0.00016050912343084216, | |
| "loss": 0.0076, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.6717325227963524, | |
| "grad_norm": 0.1771322637796402, | |
| "learning_rate": 0.00015889258610358398, | |
| "loss": 0.0045, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.702127659574468, | |
| "grad_norm": 0.1138119027018547, | |
| "learning_rate": 0.00015725213470918977, | |
| "loss": 0.0058, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.7325227963525835, | |
| "grad_norm": 0.32440638542175293, | |
| "learning_rate": 0.00015558843537335338, | |
| "loss": 0.0086, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.7629179331306992, | |
| "grad_norm": 0.35322776436805725, | |
| "learning_rate": 0.0001539021636618844, | |
| "loss": 0.0063, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.7933130699088147, | |
| "grad_norm": 0.36127084493637085, | |
| "learning_rate": 0.0001521940043063863, | |
| "loss": 0.0058, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.8237082066869301, | |
| "grad_norm": 0.3828403949737549, | |
| "learning_rate": 0.00015046465092621278, | |
| "loss": 0.0066, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.8541033434650456, | |
| "grad_norm": 0.2598365247249603, | |
| "learning_rate": 0.00014871480574681477, | |
| "loss": 0.0098, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.884498480243161, | |
| "grad_norm": 0.42567044496536255, | |
| "learning_rate": 0.00014694517931459317, | |
| "loss": 0.0102, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.9148936170212765, | |
| "grad_norm": 0.30201858282089233, | |
| "learning_rate": 0.00014515649020837277, | |
| "loss": 0.0059, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.9452887537993921, | |
| "grad_norm": 0.2485564947128296, | |
| "learning_rate": 0.00014334946474761412, | |
| "loss": 0.0056, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.9756838905775076, | |
| "grad_norm": 0.23254001140594482, | |
| "learning_rate": 0.0001415248366974826, | |
| "loss": 0.0048, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.0060790273556233, | |
| "grad_norm": 0.2745867967605591, | |
| "learning_rate": 0.00013968334697089406, | |
| "loss": 0.0035, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.0364741641337387, | |
| "grad_norm": 0.21013915538787842, | |
| "learning_rate": 0.0001378257433276578, | |
| "loss": 0.0027, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.066869300911854, | |
| "grad_norm": 0.11735199391841888, | |
| "learning_rate": 0.00013595278007083933, | |
| "loss": 0.0063, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.0972644376899696, | |
| "grad_norm": 0.14921429753303528, | |
| "learning_rate": 0.00013406521774046636, | |
| "loss": 0.0023, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "grad_norm": 0.12081817537546158, | |
| "learning_rate": 0.000132163822804702, | |
| "loss": 0.0019, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.1580547112462005, | |
| "grad_norm": 0.21969059109687805, | |
| "learning_rate": 0.00013024936734861087, | |
| "loss": 0.0051, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.188449848024316, | |
| "grad_norm": 0.15178360044956207, | |
| "learning_rate": 0.00012832262876064427, | |
| "loss": 0.003, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.2188449848024314, | |
| "grad_norm": 0.10473588109016418, | |
| "learning_rate": 0.00012638438941697206, | |
| "loss": 0.0042, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.2492401215805473, | |
| "grad_norm": 0.12606075406074524, | |
| "learning_rate": 0.0001244354363637889, | |
| "loss": 0.0035, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.2796352583586628, | |
| "grad_norm": 0.17029185593128204, | |
| "learning_rate": 0.0001224765609977246, | |
| "loss": 0.0029, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.310030395136778, | |
| "grad_norm": 0.35497593879699707, | |
| "learning_rate": 0.00012050855874448737, | |
| "loss": 0.0032, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.3404255319148937, | |
| "grad_norm": 0.17530225217342377, | |
| "learning_rate": 0.00011853222873587167, | |
| "loss": 0.0039, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.370820668693009, | |
| "grad_norm": 0.26195624470710754, | |
| "learning_rate": 0.00011654837348526044, | |
| "loss": 0.0063, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.4012158054711246, | |
| "grad_norm": 0.19774211943149567, | |
| "learning_rate": 0.00011455779856175488, | |
| "loss": 0.0019, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.43161094224924, | |
| "grad_norm": 0.23176009953022003, | |
| "learning_rate": 0.00011256131226306288, | |
| "loss": 0.0033, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.4620060790273555, | |
| "grad_norm": 0.21269629895687103, | |
| "learning_rate": 0.00011055972528727973, | |
| "loss": 0.002, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.4924012158054714, | |
| "grad_norm": 0.15158076584339142, | |
| "learning_rate": 0.00010855385040369419, | |
| "loss": 0.0093, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.522796352583587, | |
| "grad_norm": 0.16666357219219208, | |
| "learning_rate": 0.00010654450212275324, | |
| "loss": 0.0045, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.5531914893617023, | |
| "grad_norm": 0.20873229205608368, | |
| "learning_rate": 0.00010453249636532007, | |
| "loss": 0.0027, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.5835866261398177, | |
| "grad_norm": 0.20907987654209137, | |
| "learning_rate": 0.00010251865013135931, | |
| "loss": 0.0027, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.613981762917933, | |
| "grad_norm": 0.4642738997936249, | |
| "learning_rate": 0.00010050378116818391, | |
| "loss": 0.0047, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.6443768996960486, | |
| "grad_norm": 0.16256296634674072, | |
| "learning_rate": 9.848870763839877e-05, | |
| "loss": 0.0032, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.674772036474164, | |
| "grad_norm": 0.23262229561805725, | |
| "learning_rate": 9.64742477876756e-05, | |
| "loss": 0.0046, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.7051671732522795, | |
| "grad_norm": 0.255894273519516, | |
| "learning_rate": 9.44612196124941e-05, | |
| "loss": 0.0049, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.735562310030395, | |
| "grad_norm": 0.1491995006799698, | |
| "learning_rate": 9.245044052798435e-05, | |
| "loss": 0.0034, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.7659574468085104, | |
| "grad_norm": 0.18624486029148102, | |
| "learning_rate": 9.044272703600505e-05, | |
| "loss": 0.0031, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.7963525835866263, | |
| "grad_norm": 0.3413298428058624, | |
| "learning_rate": 8.843889439359308e-05, | |
| "loss": 0.0091, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.8267477203647418, | |
| "grad_norm": 0.4524446725845337, | |
| "learning_rate": 8.643975628191802e-05, | |
| "loss": 0.0084, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.2205473631620407, | |
| "learning_rate": 8.444612447587683e-05, | |
| "loss": 0.0038, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.8875379939209727, | |
| "grad_norm": 0.3354128301143646, | |
| "learning_rate": 8.245880851446255e-05, | |
| "loss": 0.0039, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.917933130699088, | |
| "grad_norm": 0.40203210711479187, | |
| "learning_rate": 8.047861537204107e-05, | |
| "loss": 0.0069, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.9483282674772036, | |
| "grad_norm": 0.16645453870296478, | |
| "learning_rate": 7.850634913066887e-05, | |
| "loss": 0.003, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.978723404255319, | |
| "grad_norm": 0.1760583221912384, | |
| "learning_rate": 7.654281065358575e-05, | |
| "loss": 0.0038, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.0091185410334345, | |
| "grad_norm": 0.09641872346401215, | |
| "learning_rate": 7.458879726001431e-05, | |
| "loss": 0.002, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.0395136778115504, | |
| "grad_norm": 0.3135606050491333, | |
| "learning_rate": 7.264510240139836e-05, | |
| "loss": 0.0052, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.069908814589666, | |
| "grad_norm": 0.1596423238515854, | |
| "learning_rate": 7.071251533921235e-05, | |
| "loss": 0.0035, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.1003039513677813, | |
| "grad_norm": 0.09552032500505447, | |
| "learning_rate": 6.879182082447185e-05, | |
| "loss": 0.0023, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.1306990881458967, | |
| "grad_norm": 0.14375700056552887, | |
| "learning_rate": 6.688379877907548e-05, | |
| "loss": 0.0034, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.161094224924012, | |
| "grad_norm": 0.18640564382076263, | |
| "learning_rate": 6.498922397910822e-05, | |
| "loss": 0.0021, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.1914893617021276, | |
| "grad_norm": 0.15173585712909698, | |
| "learning_rate": 6.310886574023362e-05, | |
| "loss": 0.0019, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.221884498480243, | |
| "grad_norm": 0.17078043520450592, | |
| "learning_rate": 6.124348760530383e-05, | |
| "loss": 0.0064, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.2522796352583585, | |
| "grad_norm": 0.11285611987113953, | |
| "learning_rate": 5.9393847034313544e-05, | |
| "loss": 0.0022, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.282674772036474, | |
| "grad_norm": 0.20150907337665558, | |
| "learning_rate": 5.756069509682368e-05, | |
| "loss": 0.0021, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.31306990881459, | |
| "grad_norm": 0.23782743513584137, | |
| "learning_rate": 5.574477616698061e-05, | |
| "loss": 0.0014, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.3434650455927053, | |
| "grad_norm": 0.05953775346279144, | |
| "learning_rate": 5.3946827621253514e-05, | |
| "loss": 0.0025, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.3738601823708207, | |
| "grad_norm": 0.09737365692853928, | |
| "learning_rate": 5.2167579539013456e-05, | |
| "loss": 0.0016, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.404255319148936, | |
| "grad_norm": 0.10946661978960037, | |
| "learning_rate": 5.0407754406075926e-05, | |
| "loss": 0.0018, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.4346504559270516, | |
| "grad_norm": 0.18973323702812195, | |
| "learning_rate": 4.866806682132611e-05, | |
| "loss": 0.0025, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.465045592705167, | |
| "grad_norm": 0.10450422018766403, | |
| "learning_rate": 4.694922320654727e-05, | |
| "loss": 0.0009, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.4954407294832825, | |
| "grad_norm": 0.1521192193031311, | |
| "learning_rate": 4.5251921519569404e-05, | |
| "loss": 0.0026, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.5258358662613984, | |
| "grad_norm": 0.11275038123130798, | |
| "learning_rate": 4.35768509708548e-05, | |
| "loss": 0.0014, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.556231003039514, | |
| "grad_norm": 0.209175705909729, | |
| "learning_rate": 4.1924691743635504e-05, | |
| "loss": 0.0025, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.5866261398176293, | |
| "grad_norm": 0.17129185795783997, | |
| "learning_rate": 4.029611471771646e-05, | |
| "loss": 0.0015, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.617021276595745, | |
| "grad_norm": 0.08248832076787949, | |
| "learning_rate": 3.869178119705648e-05, | |
| "loss": 0.0024, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.6474164133738602, | |
| "grad_norm": 0.18126089870929718, | |
| "learning_rate": 3.711234264123747e-05, | |
| "loss": 0.0028, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.6778115501519757, | |
| "grad_norm": 0.2583954632282257, | |
| "learning_rate": 3.555844040093129e-05, | |
| "loss": 0.0041, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.708206686930091, | |
| "grad_norm": 0.16434016823768616, | |
| "learning_rate": 3.403070545747107e-05, | |
| "loss": 0.0028, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.7386018237082066, | |
| "grad_norm": 0.13008223474025726, | |
| "learning_rate": 3.252975816663375e-05, | |
| "loss": 0.0015, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.768996960486322, | |
| "grad_norm": 0.0734616294503212, | |
| "learning_rate": 3.1056208006736634e-05, | |
| "loss": 0.0016, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.7993920972644375, | |
| "grad_norm": 0.06645731627941132, | |
| "learning_rate": 2.9610653331151216e-05, | |
| "loss": 0.0016, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.829787234042553, | |
| "grad_norm": 0.24428494274616241, | |
| "learning_rate": 2.8193681125334393e-05, | |
| "loss": 0.002, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.860182370820669, | |
| "grad_norm": 0.22978048026561737, | |
| "learning_rate": 2.6805866768475663e-05, | |
| "loss": 0.0017, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.8905775075987843, | |
| "grad_norm": 0.1229507103562355, | |
| "learning_rate": 2.5447773799857244e-05, | |
| "loss": 0.0018, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 3.9209726443768997, | |
| "grad_norm": 0.2323068082332611, | |
| "learning_rate": 2.4119953690022025e-05, | |
| "loss": 0.0017, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.951367781155015, | |
| "grad_norm": 0.23123906552791595, | |
| "learning_rate": 2.2822945616841963e-05, | |
| "loss": 0.002, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.9817629179331306, | |
| "grad_norm": 0.10087133944034576, | |
| "learning_rate": 2.1557276246578307e-05, | |
| "loss": 0.0018, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.0121580547112465, | |
| "grad_norm": 0.058749064803123474, | |
| "learning_rate": 2.032345952002218e-05, | |
| "loss": 0.0011, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.042553191489362, | |
| "grad_norm": 0.15570074319839478, | |
| "learning_rate": 1.9121996443802482e-05, | |
| "loss": 0.004, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.072948328267477, | |
| "grad_norm": 0.2478976845741272, | |
| "learning_rate": 1.7953374886946006e-05, | |
| "loss": 0.0036, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.103343465045593, | |
| "grad_norm": 0.08812441676855087, | |
| "learning_rate": 1.681806938277205e-05, | |
| "loss": 0.0015, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.133738601823708, | |
| "grad_norm": 0.0631280392408371, | |
| "learning_rate": 1.5716540936202363e-05, | |
| "loss": 0.0013, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.164133738601824, | |
| "grad_norm": 0.061599250882864, | |
| "learning_rate": 1.4649236836564263e-05, | |
| "loss": 0.0078, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.194528875379939, | |
| "grad_norm": 0.04393768310546875, | |
| "learning_rate": 1.361659047596332e-05, | |
| "loss": 0.0021, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.224924012158055, | |
| "grad_norm": 0.0476597435772419, | |
| "learning_rate": 1.2619021173299051e-05, | |
| "loss": 0.0014, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.25531914893617, | |
| "grad_norm": 0.07678095251321793, | |
| "learning_rate": 1.1656934003995302e-05, | |
| "loss": 0.0027, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.1129755899310112, | |
| "learning_rate": 1.0730719635514296e-05, | |
| "loss": 0.0053, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.316109422492401, | |
| "grad_norm": 0.13822700083255768, | |
| "learning_rate": 9.840754168721289e-06, | |
| "loss": 0.002, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.3465045592705165, | |
| "grad_norm": 0.10752403736114502, | |
| "learning_rate": 8.987398985164108e-06, | |
| "loss": 0.0024, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.376899696048632, | |
| "grad_norm": 0.06812784075737, | |
| "learning_rate": 8.171000600329682e-06, | |
| "loss": 0.001, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.407294832826747, | |
| "grad_norm": 0.20128777623176575, | |
| "learning_rate": 7.391890522937139e-06, | |
| "loss": 0.0015, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.437689969604863, | |
| "grad_norm": 0.0681811273097992, | |
| "learning_rate": 6.6503851203245205e-06, | |
| "loss": 0.0073, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.468085106382979, | |
| "grad_norm": 0.04031313210725784, | |
| "learning_rate": 5.946785489983941e-06, | |
| "loss": 0.001, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.498480243161095, | |
| "grad_norm": 0.16159360110759735, | |
| "learning_rate": 5.2813773372971995e-06, | |
| "loss": 0.002, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.52887537993921, | |
| "grad_norm": 0.051200274378061295, | |
| "learning_rate": 4.654430859521519e-06, | |
| "loss": 0.0008, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.5592705167173255, | |
| "grad_norm": 0.05586516112089157, | |
| "learning_rate": 4.066200636072604e-06, | |
| "loss": 0.0024, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.589665653495441, | |
| "grad_norm": 0.22293174266815186, | |
| "learning_rate": 3.5169255251495283e-06, | |
| "loss": 0.0012, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.620060790273556, | |
| "grad_norm": 0.06470826268196106, | |
| "learning_rate": 3.006828566743358e-06, | |
| "loss": 0.0009, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.650455927051672, | |
| "grad_norm": 0.11062423884868622, | |
| "learning_rate": 2.536116892069007e-06, | |
| "loss": 0.0017, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 4.680851063829787, | |
| "grad_norm": 0.08456968516111374, | |
| "learning_rate": 2.1049816394570486e-06, | |
| "loss": 0.0006, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 4.711246200607903, | |
| "grad_norm": 0.0969931110739708, | |
| "learning_rate": 1.7135978767395588e-06, | |
| "loss": 0.0037, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 4.741641337386018, | |
| "grad_norm": 0.0655050203204155, | |
| "learning_rate": 1.3621245301617014e-06, | |
| "loss": 0.001, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 4.772036474164134, | |
| "grad_norm": 0.038389451801776886, | |
| "learning_rate": 1.0507043198477617e-06, | |
| "loss": 0.0007, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 4.802431610942249, | |
| "grad_norm": 0.027961203828454018, | |
| "learning_rate": 7.794637018477824e-07, | |
| "loss": 0.0013, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 4.832826747720365, | |
| "grad_norm": 0.24096284806728363, | |
| "learning_rate": 5.485128167885933e-07, | |
| "loss": 0.003, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 4.86322188449848, | |
| "grad_norm": 0.031791090965270996, | |
| "learning_rate": 3.579454451498099e-07, | |
| "loss": 0.0023, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.8936170212765955, | |
| "grad_norm": 0.028634727001190186, | |
| "learning_rate": 2.0783896918310508e-07, | |
| "loss": 0.001, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 4.924012158054711, | |
| "grad_norm": 0.11568836122751236, | |
| "learning_rate": 9.82543414901782e-08, | |
| "loss": 0.0038, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 4.954407294832826, | |
| "grad_norm": 0.13896380364894867, | |
| "learning_rate": 2.9236060272186395e-08, | |
| "loss": 0.005, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 4.984802431610943, | |
| "grad_norm": 0.02793799713253975, | |
| "learning_rate": 8.121512607317528e-10, | |
| "loss": 0.0016, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.990881458966565, | |
| "step": 1642, | |
| "total_flos": 5.715020951091302e+16, | |
| "train_loss": 0.018644058636483826, | |
| "train_runtime": 735.9075, | |
| "train_samples_per_second": 35.7, | |
| "train_steps_per_second": 2.231 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1642, | |
| "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": 5.715020951091302e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |