| { | |
| "best_metric": 2.5859532356262207, | |
| "best_model_checkpoint": "./FT_models/[LDH]0224_docs_chatml/checkpoint-500", | |
| "epoch": 52.23880597014925, | |
| "eval_steps": 500, | |
| "global_step": 3500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14925373134328357, | |
| "grad_norm": 0.7784017324447632, | |
| "learning_rate": 0.0001999995055317446, | |
| "loss": 2.1851, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 0.5991038084030151, | |
| "learning_rate": 0.0001999955498150411, | |
| "loss": 1.5526, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.44776119402985076, | |
| "grad_norm": 0.6780698895454407, | |
| "learning_rate": 0.00019998763853811184, | |
| "loss": 1.405, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 0.7235689759254456, | |
| "learning_rate": 0.00019997577201390606, | |
| "loss": 1.3309, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "grad_norm": 0.7904847860336304, | |
| "learning_rate": 0.0001999599507118322, | |
| "loss": 1.1353, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8955223880597015, | |
| "grad_norm": 0.7685543894767761, | |
| "learning_rate": 0.00019994017525773913, | |
| "loss": 1.0295, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.044776119402985, | |
| "grad_norm": 0.963201642036438, | |
| "learning_rate": 0.0001999164464338918, | |
| "loss": 0.9815, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1940298507462686, | |
| "grad_norm": 1.0725153684616089, | |
| "learning_rate": 0.0001998887651789398, | |
| "loss": 0.7019, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3432835820895521, | |
| "grad_norm": 1.2167037725448608, | |
| "learning_rate": 0.0001998571325878806, | |
| "loss": 0.639, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "grad_norm": 1.256081461906433, | |
| "learning_rate": 0.00019982154991201608, | |
| "loss": 0.5614, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6417910447761193, | |
| "grad_norm": 0.7639100551605225, | |
| "learning_rate": 0.00019978201855890308, | |
| "loss": 0.5479, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.7910447761194028, | |
| "grad_norm": 1.0223500728607178, | |
| "learning_rate": 0.00019973854009229763, | |
| "loss": 0.3933, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.9402985074626866, | |
| "grad_norm": 1.0072954893112183, | |
| "learning_rate": 0.00019969111623209323, | |
| "loss": 0.4029, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.08955223880597, | |
| "grad_norm": 1.0809686183929443, | |
| "learning_rate": 0.00019963974885425266, | |
| "loss": 0.3913, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.2388059701492535, | |
| "grad_norm": 0.4724240005016327, | |
| "learning_rate": 0.00019958443999073397, | |
| "loss": 0.2301, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.388059701492537, | |
| "grad_norm": 0.6505578756332397, | |
| "learning_rate": 0.00019952519182940993, | |
| "loss": 0.2533, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.5373134328358207, | |
| "grad_norm": 0.39939674735069275, | |
| "learning_rate": 0.0001994620067139815, | |
| "loss": 0.1551, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.6865671641791042, | |
| "grad_norm": 0.26736512780189514, | |
| "learning_rate": 0.00019939488714388524, | |
| "loss": 0.2489, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.835820895522388, | |
| "grad_norm": 0.7464708089828491, | |
| "learning_rate": 0.00019932383577419432, | |
| "loss": 0.2118, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "grad_norm": 0.348090261220932, | |
| "learning_rate": 0.0001992488554155135, | |
| "loss": 0.2277, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.1343283582089554, | |
| "grad_norm": 0.4606758654117584, | |
| "learning_rate": 0.0001991699490338681, | |
| "loss": 0.1216, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.283582089552239, | |
| "grad_norm": 0.5326898694038391, | |
| "learning_rate": 0.00019908711975058637, | |
| "loss": 0.0946, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.4328358208955225, | |
| "grad_norm": 0.5780682563781738, | |
| "learning_rate": 0.00019900037084217637, | |
| "loss": 0.0982, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.582089552238806, | |
| "grad_norm": 0.5874966382980347, | |
| "learning_rate": 0.00019890970574019617, | |
| "loss": 0.1244, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.7313432835820897, | |
| "grad_norm": 0.5044897198677063, | |
| "learning_rate": 0.00019881512803111796, | |
| "loss": 0.1127, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.8805970149253732, | |
| "grad_norm": 0.45550045371055603, | |
| "learning_rate": 0.00019871664145618657, | |
| "loss": 0.1211, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.029850746268656, | |
| "grad_norm": 0.3092655539512634, | |
| "learning_rate": 0.00019861424991127115, | |
| "loss": 0.1156, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.17910447761194, | |
| "grad_norm": 0.4061109721660614, | |
| "learning_rate": 0.00019850795744671116, | |
| "loss": 0.0695, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.3283582089552235, | |
| "grad_norm": 0.36566805839538574, | |
| "learning_rate": 0.00019839776826715614, | |
| "loss": 0.0677, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 4.477611940298507, | |
| "grad_norm": 0.29778343439102173, | |
| "learning_rate": 0.00019828368673139947, | |
| "loss": 0.0445, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.6268656716417915, | |
| "grad_norm": 0.18671129643917084, | |
| "learning_rate": 0.00019816571735220583, | |
| "loss": 0.0644, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 4.776119402985074, | |
| "grad_norm": 0.3877570629119873, | |
| "learning_rate": 0.0001980438647961327, | |
| "loss": 0.0912, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.925373134328359, | |
| "grad_norm": 0.5179444551467896, | |
| "learning_rate": 0.00019791813388334581, | |
| "loss": 0.0607, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.074626865671641, | |
| "grad_norm": 0.15815891325473785, | |
| "learning_rate": 0.00019778852958742853, | |
| "loss": 0.0542, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 5.223880597014926, | |
| "grad_norm": 0.42946234345436096, | |
| "learning_rate": 0.00019765505703518496, | |
| "loss": 0.0563, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 5.373134328358209, | |
| "grad_norm": 0.5582062005996704, | |
| "learning_rate": 0.00019751772150643722, | |
| "loss": 0.0408, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 5.522388059701493, | |
| "grad_norm": 0.2707679569721222, | |
| "learning_rate": 0.0001973765284338167, | |
| "loss": 0.0395, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 5.6716417910447765, | |
| "grad_norm": 0.5061952471733093, | |
| "learning_rate": 0.00019723148340254892, | |
| "loss": 0.0501, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 5.82089552238806, | |
| "grad_norm": 0.31659457087516785, | |
| "learning_rate": 0.0001970825921502328, | |
| "loss": 0.049, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 5.970149253731344, | |
| "grad_norm": 0.2836756110191345, | |
| "learning_rate": 0.00019692986056661356, | |
| "loss": 0.0483, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.119402985074627, | |
| "grad_norm": 0.1509529948234558, | |
| "learning_rate": 0.0001967732946933499, | |
| "loss": 0.0394, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 6.268656716417911, | |
| "grad_norm": 0.2609387934207916, | |
| "learning_rate": 0.00019661290072377482, | |
| "loss": 0.0388, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 6.417910447761194, | |
| "grad_norm": 0.5987827777862549, | |
| "learning_rate": 0.0001964486850026507, | |
| "loss": 0.0383, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 6.567164179104478, | |
| "grad_norm": 0.4549473822116852, | |
| "learning_rate": 0.00019628065402591845, | |
| "loss": 0.038, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 6.7164179104477615, | |
| "grad_norm": 0.2026955932378769, | |
| "learning_rate": 0.0001961088144404403, | |
| "loss": 0.0384, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 6.865671641791045, | |
| "grad_norm": 0.21201670169830322, | |
| "learning_rate": 0.00019593317304373705, | |
| "loss": 0.0347, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 7.014925373134329, | |
| "grad_norm": 0.0839223638176918, | |
| "learning_rate": 0.00019575373678371909, | |
| "loss": 0.0458, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 7.164179104477612, | |
| "grad_norm": 0.38603273034095764, | |
| "learning_rate": 0.0001955705127584117, | |
| "loss": 0.0366, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 7.313432835820896, | |
| "grad_norm": 0.22731441259384155, | |
| "learning_rate": 0.00019538350821567404, | |
| "loss": 0.0324, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 7.462686567164179, | |
| "grad_norm": 0.3230607509613037, | |
| "learning_rate": 0.00019519273055291266, | |
| "loss": 0.0321, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.462686567164179, | |
| "eval_loss": 2.5859532356262207, | |
| "eval_runtime": 34.8624, | |
| "eval_samples_per_second": 9.093, | |
| "eval_steps_per_second": 4.561, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.611940298507463, | |
| "grad_norm": 0.1323220282793045, | |
| "learning_rate": 0.00019499818731678873, | |
| "loss": 0.0271, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 7.7611940298507465, | |
| "grad_norm": 0.17675843834877014, | |
| "learning_rate": 0.00019479988620291956, | |
| "loss": 0.0356, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 7.91044776119403, | |
| "grad_norm": 0.5695576071739197, | |
| "learning_rate": 0.00019459783505557424, | |
| "loss": 0.0377, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 8.059701492537313, | |
| "grad_norm": 0.14659564197063446, | |
| "learning_rate": 0.0001943920418673633, | |
| "loss": 0.0295, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 8.208955223880597, | |
| "grad_norm": 0.3027053475379944, | |
| "learning_rate": 0.0001941825147789225, | |
| "loss": 0.0283, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 8.35820895522388, | |
| "grad_norm": 0.3131660521030426, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 0.0354, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 8.507462686567164, | |
| "grad_norm": 0.2541915476322174, | |
| "learning_rate": 0.00019375229220208276, | |
| "loss": 0.0305, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 8.656716417910447, | |
| "grad_norm": 0.15617048740386963, | |
| "learning_rate": 0.0001935316137321543, | |
| "loss": 0.0374, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 8.805970149253731, | |
| "grad_norm": 0.8228221535682678, | |
| "learning_rate": 0.00019330723539826375, | |
| "loss": 0.0332, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 8.955223880597014, | |
| "grad_norm": 0.29615482687950134, | |
| "learning_rate": 0.0001930791660762262, | |
| "loss": 0.0319, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.104477611940299, | |
| "grad_norm": 0.222768172621727, | |
| "learning_rate": 0.0001928474147878626, | |
| "loss": 0.0279, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 9.253731343283581, | |
| "grad_norm": 0.1895366758108139, | |
| "learning_rate": 0.0001926119907006426, | |
| "loss": 0.0337, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 9.402985074626866, | |
| "grad_norm": 0.19570475816726685, | |
| "learning_rate": 0.00019237290312732226, | |
| "loss": 0.0276, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 9.552238805970148, | |
| "grad_norm": 0.5487334728240967, | |
| "learning_rate": 0.0001921301615255754, | |
| "loss": 0.0346, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 9.701492537313433, | |
| "grad_norm": 0.17226798832416534, | |
| "learning_rate": 0.00019188377549761963, | |
| "loss": 0.0256, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 9.850746268656717, | |
| "grad_norm": 0.2766994833946228, | |
| "learning_rate": 0.00019163375478983632, | |
| "loss": 0.0296, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.25644323229789734, | |
| "learning_rate": 0.00019138010929238534, | |
| "loss": 0.0326, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 10.149253731343283, | |
| "grad_norm": 0.3244101405143738, | |
| "learning_rate": 0.0001911228490388136, | |
| "loss": 0.0262, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 10.298507462686567, | |
| "grad_norm": 0.24429234862327576, | |
| "learning_rate": 0.00019086198420565823, | |
| "loss": 0.0321, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 10.447761194029852, | |
| "grad_norm": 0.15597452223300934, | |
| "learning_rate": 0.000190597525112044, | |
| "loss": 0.0314, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 10.597014925373134, | |
| "grad_norm": 0.22565604746341705, | |
| "learning_rate": 0.00019032948221927524, | |
| "loss": 0.028, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 10.746268656716419, | |
| "grad_norm": 0.1876748502254486, | |
| "learning_rate": 0.00019005786613042185, | |
| "loss": 0.0292, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 10.895522388059701, | |
| "grad_norm": 0.33821797370910645, | |
| "learning_rate": 0.00018978268758989991, | |
| "loss": 0.0289, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 11.044776119402986, | |
| "grad_norm": 0.1624440997838974, | |
| "learning_rate": 0.00018950395748304678, | |
| "loss": 0.0306, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 11.194029850746269, | |
| "grad_norm": 0.17152149975299835, | |
| "learning_rate": 0.0001892216868356904, | |
| "loss": 0.0293, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 11.343283582089553, | |
| "grad_norm": 0.22814153134822845, | |
| "learning_rate": 0.00018893588681371303, | |
| "loss": 0.0294, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 11.492537313432836, | |
| "grad_norm": 0.3319993317127228, | |
| "learning_rate": 0.00018864656872260985, | |
| "loss": 0.0282, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 11.64179104477612, | |
| "grad_norm": 0.1329302042722702, | |
| "learning_rate": 0.00018835374400704154, | |
| "loss": 0.029, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 11.791044776119403, | |
| "grad_norm": 0.2312331646680832, | |
| "learning_rate": 0.00018805742425038145, | |
| "loss": 0.0313, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 11.940298507462687, | |
| "grad_norm": 0.11133890599012375, | |
| "learning_rate": 0.00018775762117425777, | |
| "loss": 0.0291, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 12.08955223880597, | |
| "grad_norm": 0.42977747321128845, | |
| "learning_rate": 0.00018745434663808942, | |
| "loss": 0.0293, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 12.238805970149254, | |
| "grad_norm": 0.4400336742401123, | |
| "learning_rate": 0.00018714761263861728, | |
| "loss": 0.0258, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 12.388059701492537, | |
| "grad_norm": 0.3886387348175049, | |
| "learning_rate": 0.00018683743130942928, | |
| "loss": 0.03, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 12.537313432835822, | |
| "grad_norm": 0.1791338473558426, | |
| "learning_rate": 0.00018652381492048083, | |
| "loss": 0.0259, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 12.686567164179104, | |
| "grad_norm": 0.17430442571640015, | |
| "learning_rate": 0.00018620677587760916, | |
| "loss": 0.0282, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 12.835820895522389, | |
| "grad_norm": 0.3342001140117645, | |
| "learning_rate": 0.00018588632672204264, | |
| "loss": 0.0301, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 12.985074626865671, | |
| "grad_norm": 0.2671709358692169, | |
| "learning_rate": 0.00018556248012990468, | |
| "loss": 0.0269, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 13.134328358208956, | |
| "grad_norm": 0.27415236830711365, | |
| "learning_rate": 0.0001852352489117124, | |
| "loss": 0.026, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 13.283582089552239, | |
| "grad_norm": 0.13104292750358582, | |
| "learning_rate": 0.0001849046460118698, | |
| "loss": 0.0228, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 13.432835820895523, | |
| "grad_norm": 0.0972188338637352, | |
| "learning_rate": 0.00018457068450815562, | |
| "loss": 0.0224, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 13.582089552238806, | |
| "grad_norm": 0.36645978689193726, | |
| "learning_rate": 0.00018423337761120618, | |
| "loss": 0.023, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 13.73134328358209, | |
| "grad_norm": 0.09311423450708389, | |
| "learning_rate": 0.00018389273866399275, | |
| "loss": 0.0249, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 13.880597014925373, | |
| "grad_norm": 0.26524588465690613, | |
| "learning_rate": 0.00018354878114129367, | |
| "loss": 0.0283, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 14.029850746268657, | |
| "grad_norm": 0.09386450052261353, | |
| "learning_rate": 0.00018320151864916135, | |
| "loss": 0.0248, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 14.17910447761194, | |
| "grad_norm": 0.15628236532211304, | |
| "learning_rate": 0.00018285096492438424, | |
| "loss": 0.0242, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 14.328358208955224, | |
| "grad_norm": 0.09542356431484222, | |
| "learning_rate": 0.00018249713383394303, | |
| "loss": 0.0227, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 14.477611940298507, | |
| "grad_norm": 0.07725339382886887, | |
| "learning_rate": 0.00018214003937446253, | |
| "loss": 0.0238, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 14.626865671641792, | |
| "grad_norm": 0.10331985354423523, | |
| "learning_rate": 0.0001817796956716578, | |
| "loss": 0.0269, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 14.776119402985074, | |
| "grad_norm": 0.15463414788246155, | |
| "learning_rate": 0.00018141611697977529, | |
| "loss": 0.0217, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 14.925373134328359, | |
| "grad_norm": 0.14817017316818237, | |
| "learning_rate": 0.0001810493176810292, | |
| "loss": 0.025, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 14.925373134328359, | |
| "eval_loss": 2.6268842220306396, | |
| "eval_runtime": 34.7235, | |
| "eval_samples_per_second": 9.129, | |
| "eval_steps_per_second": 4.579, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 15.074626865671641, | |
| "grad_norm": 0.12652496993541718, | |
| "learning_rate": 0.00018067931228503246, | |
| "loss": 0.0209, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 15.223880597014926, | |
| "grad_norm": 0.10728372633457184, | |
| "learning_rate": 0.00018030611542822257, | |
| "loss": 0.0222, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 15.373134328358208, | |
| "grad_norm": 0.06331785768270493, | |
| "learning_rate": 0.00017992974187328305, | |
| "loss": 0.0217, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 15.522388059701493, | |
| "grad_norm": 0.21516793966293335, | |
| "learning_rate": 0.000179550206508559, | |
| "loss": 0.022, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 15.671641791044776, | |
| "grad_norm": 0.06852179765701294, | |
| "learning_rate": 0.00017916752434746856, | |
| "loss": 0.0213, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 15.82089552238806, | |
| "grad_norm": 0.07877414673566818, | |
| "learning_rate": 0.00017878171052790868, | |
| "loss": 0.0244, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 15.970149253731343, | |
| "grad_norm": 0.28540125489234924, | |
| "learning_rate": 0.00017839278031165658, | |
| "loss": 0.0244, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 16.119402985074625, | |
| "grad_norm": 0.060050830245018005, | |
| "learning_rate": 0.00017800074908376584, | |
| "loss": 0.0197, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 16.26865671641791, | |
| "grad_norm": 0.14403647184371948, | |
| "learning_rate": 0.0001776056323519579, | |
| "loss": 0.0197, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 16.417910447761194, | |
| "grad_norm": 0.14129780232906342, | |
| "learning_rate": 0.00017720744574600863, | |
| "loss": 0.0235, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 16.567164179104477, | |
| "grad_norm": 0.14378951489925385, | |
| "learning_rate": 0.00017680620501712996, | |
| "loss": 0.0213, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 16.71641791044776, | |
| "grad_norm": 0.09062501788139343, | |
| "learning_rate": 0.00017640192603734692, | |
| "loss": 0.0247, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 16.865671641791046, | |
| "grad_norm": 0.33645904064178467, | |
| "learning_rate": 0.00017599462479886974, | |
| "loss": 0.0236, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 17.01492537313433, | |
| "grad_norm": 0.10567416250705719, | |
| "learning_rate": 0.00017558431741346122, | |
| "loss": 0.0227, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 17.16417910447761, | |
| "grad_norm": 0.3064155578613281, | |
| "learning_rate": 0.00017517102011179933, | |
| "loss": 0.0238, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 17.313432835820894, | |
| "grad_norm": 0.256085067987442, | |
| "learning_rate": 0.00017475474924283536, | |
| "loss": 0.021, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 17.46268656716418, | |
| "grad_norm": 0.8420864343643188, | |
| "learning_rate": 0.000174335521273147, | |
| "loss": 0.0241, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 17.611940298507463, | |
| "grad_norm": 0.21292446553707123, | |
| "learning_rate": 0.00017391335278628712, | |
| "loss": 0.0251, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 17.761194029850746, | |
| "grad_norm": 0.14111167192459106, | |
| "learning_rate": 0.0001734882604821276, | |
| "loss": 0.0251, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 17.91044776119403, | |
| "grad_norm": 0.2401103377342224, | |
| "learning_rate": 0.00017306026117619889, | |
| "loss": 0.0294, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 18.059701492537314, | |
| "grad_norm": 0.19608861207962036, | |
| "learning_rate": 0.00017262937179902472, | |
| "loss": 0.025, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 18.208955223880597, | |
| "grad_norm": 0.10760471224784851, | |
| "learning_rate": 0.00017219560939545246, | |
| "loss": 0.0252, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 18.35820895522388, | |
| "grad_norm": 0.1591477394104004, | |
| "learning_rate": 0.0001717589911239788, | |
| "loss": 0.0225, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 18.507462686567163, | |
| "grad_norm": 0.21945133805274963, | |
| "learning_rate": 0.00017131953425607104, | |
| "loss": 0.0234, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 18.65671641791045, | |
| "grad_norm": 0.147572860121727, | |
| "learning_rate": 0.00017087725617548385, | |
| "loss": 0.0216, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 18.80597014925373, | |
| "grad_norm": 0.19251006841659546, | |
| "learning_rate": 0.00017043217437757164, | |
| "loss": 0.0258, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 18.955223880597014, | |
| "grad_norm": 0.3675619959831238, | |
| "learning_rate": 0.00016998430646859654, | |
| "loss": 0.0263, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 19.104477611940297, | |
| "grad_norm": 0.2582189440727234, | |
| "learning_rate": 0.00016953367016503182, | |
| "loss": 0.0258, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 19.253731343283583, | |
| "grad_norm": 0.22689688205718994, | |
| "learning_rate": 0.00016908028329286112, | |
| "loss": 0.0226, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 19.402985074626866, | |
| "grad_norm": 0.24671746790409088, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 0.0224, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 19.55223880597015, | |
| "grad_norm": 0.20800915360450745, | |
| "learning_rate": 0.00016816532968995328, | |
| "loss": 0.0276, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 19.701492537313435, | |
| "grad_norm": 0.27109411358833313, | |
| "learning_rate": 0.00016770379915236766, | |
| "loss": 0.0345, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 19.850746268656717, | |
| "grad_norm": 0.14396627247333527, | |
| "learning_rate": 0.00016723959043104728, | |
| "loss": 0.0244, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.3481923043727875, | |
| "learning_rate": 0.00016677272188886483, | |
| "loss": 0.0256, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 20.149253731343283, | |
| "grad_norm": 0.4062567353248596, | |
| "learning_rate": 0.00016630321199390867, | |
| "loss": 0.0195, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 20.298507462686565, | |
| "grad_norm": 0.41222649812698364, | |
| "learning_rate": 0.00016583107931875192, | |
| "loss": 0.023, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 20.44776119402985, | |
| "grad_norm": 0.15970966219902039, | |
| "learning_rate": 0.00016535634253971794, | |
| "loss": 0.024, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 20.597014925373134, | |
| "grad_norm": 0.2472175806760788, | |
| "learning_rate": 0.00016487902043614173, | |
| "loss": 0.0259, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 20.746268656716417, | |
| "grad_norm": 0.30030208826065063, | |
| "learning_rate": 0.00016439913188962685, | |
| "loss": 0.0287, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 20.895522388059703, | |
| "grad_norm": 0.17114506661891937, | |
| "learning_rate": 0.0001639166958832985, | |
| "loss": 0.0279, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 21.044776119402986, | |
| "grad_norm": 0.22200174629688263, | |
| "learning_rate": 0.00016343173150105278, | |
| "loss": 0.0254, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 21.19402985074627, | |
| "grad_norm": 0.3260650336742401, | |
| "learning_rate": 0.0001629442579268016, | |
| "loss": 0.0252, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 21.34328358208955, | |
| "grad_norm": 0.2409895658493042, | |
| "learning_rate": 0.0001624542944437139, | |
| "loss": 0.0255, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 21.492537313432837, | |
| "grad_norm": 0.37012478709220886, | |
| "learning_rate": 0.00016196186043345288, | |
| "loss": 0.0293, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 21.64179104477612, | |
| "grad_norm": 0.1385307013988495, | |
| "learning_rate": 0.00016146697537540924, | |
| "loss": 0.0273, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 21.791044776119403, | |
| "grad_norm": 0.33346429467201233, | |
| "learning_rate": 0.0001609696588459307, | |
| "loss": 0.0294, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 21.940298507462686, | |
| "grad_norm": 0.14864549040794373, | |
| "learning_rate": 0.00016046993051754756, | |
| "loss": 0.0269, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 22.08955223880597, | |
| "grad_norm": 0.11181981861591339, | |
| "learning_rate": 0.0001599678101581945, | |
| "loss": 0.022, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 22.238805970149254, | |
| "grad_norm": 0.18521477282047272, | |
| "learning_rate": 0.00015946331763042867, | |
| "loss": 0.0189, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 22.388059701492537, | |
| "grad_norm": 0.07250799983739853, | |
| "learning_rate": 0.00015895647289064396, | |
| "loss": 0.0191, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 22.388059701492537, | |
| "eval_loss": 2.901611566543579, | |
| "eval_runtime": 34.7441, | |
| "eval_samples_per_second": 9.124, | |
| "eval_steps_per_second": 4.576, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 22.53731343283582, | |
| "grad_norm": 0.11993291229009628, | |
| "learning_rate": 0.0001584472959882815, | |
| "loss": 0.0232, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 22.686567164179106, | |
| "grad_norm": 0.08708442002534866, | |
| "learning_rate": 0.0001579358070650367, | |
| "loss": 0.0203, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 22.83582089552239, | |
| "grad_norm": 0.17798146605491638, | |
| "learning_rate": 0.00015742202635406235, | |
| "loss": 0.0278, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 22.98507462686567, | |
| "grad_norm": 0.33858776092529297, | |
| "learning_rate": 0.0001569059741791684, | |
| "loss": 0.0266, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 23.134328358208954, | |
| "grad_norm": 0.14069178700447083, | |
| "learning_rate": 0.0001563876709540178, | |
| "loss": 0.0235, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 23.28358208955224, | |
| "grad_norm": 0.23493413627147675, | |
| "learning_rate": 0.00015586713718131922, | |
| "loss": 0.0225, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 23.432835820895523, | |
| "grad_norm": 0.23869164288043976, | |
| "learning_rate": 0.0001553443934520159, | |
| "loss": 0.0207, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 23.582089552238806, | |
| "grad_norm": 0.04802766814827919, | |
| "learning_rate": 0.00015481946044447099, | |
| "loss": 0.0208, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 23.73134328358209, | |
| "grad_norm": 0.10397443175315857, | |
| "learning_rate": 0.00015429235892364994, | |
| "loss": 0.0218, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 23.880597014925375, | |
| "grad_norm": 0.06602863222360611, | |
| "learning_rate": 0.00015376310974029873, | |
| "loss": 0.0191, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 24.029850746268657, | |
| "grad_norm": 0.12586766481399536, | |
| "learning_rate": 0.0001532317338301192, | |
| "loss": 0.0222, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 24.17910447761194, | |
| "grad_norm": 0.058740176260471344, | |
| "learning_rate": 0.00015269825221294098, | |
| "loss": 0.0194, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 24.328358208955223, | |
| "grad_norm": 0.046583324670791626, | |
| "learning_rate": 0.0001521626859918898, | |
| "loss": 0.0192, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 24.47761194029851, | |
| "grad_norm": 0.07122834771871567, | |
| "learning_rate": 0.00015162505635255287, | |
| "loss": 0.0191, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 24.62686567164179, | |
| "grad_norm": 0.17352773249149323, | |
| "learning_rate": 0.0001510853845621409, | |
| "loss": 0.0215, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 24.776119402985074, | |
| "grad_norm": 0.039071228355169296, | |
| "learning_rate": 0.00015054369196864644, | |
| "loss": 0.0194, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 24.925373134328357, | |
| "grad_norm": 0.0495145358145237, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.0196, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 25.074626865671643, | |
| "grad_norm": 0.06830393522977829, | |
| "learning_rate": 0.0001494543301632219, | |
| "loss": 0.0196, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 25.223880597014926, | |
| "grad_norm": 0.046369269490242004, | |
| "learning_rate": 0.0001489067040435717, | |
| "loss": 0.0203, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 25.37313432835821, | |
| "grad_norm": 0.08401994407176971, | |
| "learning_rate": 0.00014835714330369446, | |
| "loss": 0.0177, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 25.52238805970149, | |
| "grad_norm": 0.15754491090774536, | |
| "learning_rate": 0.0001478056696827636, | |
| "loss": 0.0183, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 25.671641791044777, | |
| "grad_norm": 0.03660280629992485, | |
| "learning_rate": 0.00014725230499562119, | |
| "loss": 0.0187, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 25.82089552238806, | |
| "grad_norm": 0.03743986785411835, | |
| "learning_rate": 0.00014669707113191483, | |
| "loss": 0.0177, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 25.970149253731343, | |
| "grad_norm": 0.061036206781864166, | |
| "learning_rate": 0.00014613999005523174, | |
| "loss": 0.0216, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 26.119402985074625, | |
| "grad_norm": 0.052410393953323364, | |
| "learning_rate": 0.00014558108380223012, | |
| "loss": 0.0179, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 26.26865671641791, | |
| "grad_norm": 0.04502878338098526, | |
| "learning_rate": 0.00014502037448176734, | |
| "loss": 0.0173, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 26.417910447761194, | |
| "grad_norm": 0.09054642915725708, | |
| "learning_rate": 0.00014445788427402528, | |
| "loss": 0.0175, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 26.567164179104477, | |
| "grad_norm": 0.04439815133810043, | |
| "learning_rate": 0.00014389363542963306, | |
| "loss": 0.0166, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 26.71641791044776, | |
| "grad_norm": 0.057578206062316895, | |
| "learning_rate": 0.00014332765026878687, | |
| "loss": 0.0204, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 26.865671641791046, | |
| "grad_norm": 0.04316519573330879, | |
| "learning_rate": 0.00014275995118036693, | |
| "loss": 0.0204, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 27.01492537313433, | |
| "grad_norm": 0.044371455907821655, | |
| "learning_rate": 0.00014219056062105193, | |
| "loss": 0.0188, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 27.16417910447761, | |
| "grad_norm": 0.0417649932205677, | |
| "learning_rate": 0.00014161950111443077, | |
| "loss": 0.0167, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 27.313432835820894, | |
| "grad_norm": 0.08106902241706848, | |
| "learning_rate": 0.0001410467952501114, | |
| "loss": 0.0196, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 27.46268656716418, | |
| "grad_norm": 0.1356726437807083, | |
| "learning_rate": 0.00014047246568282736, | |
| "loss": 0.0214, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 27.611940298507463, | |
| "grad_norm": 0.11032121628522873, | |
| "learning_rate": 0.00013989653513154165, | |
| "loss": 0.0188, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 27.761194029850746, | |
| "grad_norm": 0.1375930905342102, | |
| "learning_rate": 0.0001393190263785479, | |
| "loss": 0.0217, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 27.91044776119403, | |
| "grad_norm": 0.12069140374660492, | |
| "learning_rate": 0.00013873996226856933, | |
| "loss": 0.0198, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 28.059701492537314, | |
| "grad_norm": 0.043185122311115265, | |
| "learning_rate": 0.00013815936570785487, | |
| "loss": 0.0182, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 28.208955223880597, | |
| "grad_norm": 0.14993228018283844, | |
| "learning_rate": 0.00013757725966327322, | |
| "loss": 0.0167, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 28.35820895522388, | |
| "grad_norm": 0.0337139368057251, | |
| "learning_rate": 0.00013699366716140435, | |
| "loss": 0.0186, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 28.507462686567163, | |
| "grad_norm": 0.26397907733917236, | |
| "learning_rate": 0.0001364086112876284, | |
| "loss": 0.0186, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 28.65671641791045, | |
| "grad_norm": 0.06083005666732788, | |
| "learning_rate": 0.00013582211518521273, | |
| "loss": 0.0195, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 28.80597014925373, | |
| "grad_norm": 0.10358510911464691, | |
| "learning_rate": 0.00013523420205439646, | |
| "loss": 0.0186, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 28.955223880597014, | |
| "grad_norm": 0.06939724832773209, | |
| "learning_rate": 0.00013464489515147238, | |
| "loss": 0.0201, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 29.104477611940297, | |
| "grad_norm": 0.038348495960235596, | |
| "learning_rate": 0.00013405421778786737, | |
| "loss": 0.0198, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 29.253731343283583, | |
| "grad_norm": 0.045545510947704315, | |
| "learning_rate": 0.00013346219332922016, | |
| "loss": 0.0199, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 29.402985074626866, | |
| "grad_norm": 0.03619709983468056, | |
| "learning_rate": 0.0001328688451944569, | |
| "loss": 0.018, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 29.55223880597015, | |
| "grad_norm": 0.03463749587535858, | |
| "learning_rate": 0.00013227419685486492, | |
| "loss": 0.0192, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 29.701492537313435, | |
| "grad_norm": 0.16788923740386963, | |
| "learning_rate": 0.0001316782718331643, | |
| "loss": 0.0209, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 29.850746268656717, | |
| "grad_norm": 0.04790572449564934, | |
| "learning_rate": 0.00013108109370257712, | |
| "loss": 0.0211, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 29.850746268656717, | |
| "eval_loss": 2.8224031925201416, | |
| "eval_runtime": 34.7119, | |
| "eval_samples_per_second": 9.132, | |
| "eval_steps_per_second": 4.581, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.33791643381118774, | |
| "learning_rate": 0.00013048268608589533, | |
| "loss": 0.0198, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 30.149253731343283, | |
| "grad_norm": 0.04068003594875336, | |
| "learning_rate": 0.00012988307265454597, | |
| "loss": 0.0192, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 30.298507462686565, | |
| "grad_norm": 0.05943215265870094, | |
| "learning_rate": 0.00012928227712765504, | |
| "loss": 0.0181, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 30.44776119402985, | |
| "grad_norm": 0.18460267782211304, | |
| "learning_rate": 0.00012868032327110904, | |
| "loss": 0.0208, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 30.597014925373134, | |
| "grad_norm": 0.061664972454309464, | |
| "learning_rate": 0.00012807723489661495, | |
| "loss": 0.0205, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 30.746268656716417, | |
| "grad_norm": 0.36015012860298157, | |
| "learning_rate": 0.0001274730358607583, | |
| "loss": 0.0184, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 30.895522388059703, | |
| "grad_norm": 0.1974068284034729, | |
| "learning_rate": 0.00012686775006405946, | |
| "loss": 0.0196, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 31.044776119402986, | |
| "grad_norm": 0.12781082093715668, | |
| "learning_rate": 0.0001262614014500282, | |
| "loss": 0.021, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 31.19402985074627, | |
| "grad_norm": 0.22529159486293793, | |
| "learning_rate": 0.00012565401400421651, | |
| "loss": 0.018, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 31.34328358208955, | |
| "grad_norm": 0.0831318125128746, | |
| "learning_rate": 0.00012504561175326985, | |
| "loss": 0.0194, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 31.492537313432837, | |
| "grad_norm": 0.07306008040904999, | |
| "learning_rate": 0.0001244362187639767, | |
| "loss": 0.0183, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 31.64179104477612, | |
| "grad_norm": 0.08519799262285233, | |
| "learning_rate": 0.0001238258591423165, | |
| "loss": 0.0184, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 31.791044776119403, | |
| "grad_norm": 0.061566926538944244, | |
| "learning_rate": 0.00012321455703250616, | |
| "loss": 0.0198, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 31.940298507462686, | |
| "grad_norm": 0.04921621084213257, | |
| "learning_rate": 0.0001226023366160449, | |
| "loss": 0.0192, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 32.08955223880597, | |
| "grad_norm": 0.0568234883248806, | |
| "learning_rate": 0.00012198922211075778, | |
| "loss": 0.0186, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 32.23880597014925, | |
| "grad_norm": 0.09815705567598343, | |
| "learning_rate": 0.00012137523776983757, | |
| "loss": 0.0175, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 32.38805970149254, | |
| "grad_norm": 0.18607860803604126, | |
| "learning_rate": 0.00012076040788088554, | |
| "loss": 0.0178, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 32.53731343283582, | |
| "grad_norm": 0.1101093739271164, | |
| "learning_rate": 0.00012014475676495052, | |
| "loss": 0.0179, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 32.6865671641791, | |
| "grad_norm": 0.03449343517422676, | |
| "learning_rate": 0.000119528308775567, | |
| "loss": 0.0206, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 32.83582089552239, | |
| "grad_norm": 0.042853228747844696, | |
| "learning_rate": 0.00011891108829779165, | |
| "loss": 0.0191, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 32.985074626865675, | |
| "grad_norm": 0.048137255012989044, | |
| "learning_rate": 0.00011829311974723867, | |
| "loss": 0.0182, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 33.134328358208954, | |
| "grad_norm": 0.19318771362304688, | |
| "learning_rate": 0.00011767442756911417, | |
| "loss": 0.0178, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 33.28358208955224, | |
| "grad_norm": 0.08087719976902008, | |
| "learning_rate": 0.00011705503623724898, | |
| "loss": 0.0177, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 33.43283582089552, | |
| "grad_norm": 0.04939524829387665, | |
| "learning_rate": 0.00011643497025313061, | |
| "loss": 0.0194, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 33.582089552238806, | |
| "grad_norm": 0.04750213399529457, | |
| "learning_rate": 0.0001158142541449341, | |
| "loss": 0.0183, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 33.73134328358209, | |
| "grad_norm": 0.04823266714811325, | |
| "learning_rate": 0.0001151929124665516, | |
| "loss": 0.0172, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 33.88059701492537, | |
| "grad_norm": 0.03645576909184456, | |
| "learning_rate": 0.00011457096979662114, | |
| "loss": 0.018, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 34.02985074626866, | |
| "grad_norm": 0.025920415297150612, | |
| "learning_rate": 0.00011394845073755455, | |
| "loss": 0.0178, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 34.17910447761194, | |
| "grad_norm": 0.038914065808057785, | |
| "learning_rate": 0.00011332537991456398, | |
| "loss": 0.0168, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 34.32835820895522, | |
| "grad_norm": 0.07641326636075974, | |
| "learning_rate": 0.00011270178197468789, | |
| "loss": 0.0194, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 34.47761194029851, | |
| "grad_norm": 0.08722823858261108, | |
| "learning_rate": 0.00011207768158581613, | |
| "loss": 0.0199, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 34.62686567164179, | |
| "grad_norm": 0.4223034679889679, | |
| "learning_rate": 0.00011145310343571411, | |
| "loss": 0.02, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 34.776119402985074, | |
| "grad_norm": 0.12001374363899231, | |
| "learning_rate": 0.0001108280722310462, | |
| "loss": 0.0173, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 34.92537313432836, | |
| "grad_norm": 0.14997516572475433, | |
| "learning_rate": 0.00011020261269639842, | |
| "loss": 0.0195, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 35.07462686567164, | |
| "grad_norm": 0.052674125880002975, | |
| "learning_rate": 0.00010957674957330042, | |
| "loss": 0.018, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 35.223880597014926, | |
| "grad_norm": 0.06538953632116318, | |
| "learning_rate": 0.00010895050761924668, | |
| "loss": 0.0168, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 35.37313432835821, | |
| "grad_norm": 0.08561014384031296, | |
| "learning_rate": 0.00010832391160671729, | |
| "loss": 0.0177, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 35.52238805970149, | |
| "grad_norm": 0.057601574808359146, | |
| "learning_rate": 0.00010769698632219794, | |
| "loss": 0.0176, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 35.67164179104478, | |
| "grad_norm": 0.051563702523708344, | |
| "learning_rate": 0.00010706975656519946, | |
| "loss": 0.0196, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 35.82089552238806, | |
| "grad_norm": 0.034905366599559784, | |
| "learning_rate": 0.00010644224714727681, | |
| "loss": 0.0183, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 35.97014925373134, | |
| "grad_norm": 0.053671374917030334, | |
| "learning_rate": 0.00010581448289104758, | |
| "loss": 0.0179, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 36.11940298507463, | |
| "grad_norm": 0.06900997459888458, | |
| "learning_rate": 0.00010518648862921012, | |
| "loss": 0.0187, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 36.26865671641791, | |
| "grad_norm": 0.12105145305395126, | |
| "learning_rate": 0.00010455828920356115, | |
| "loss": 0.0164, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 36.417910447761194, | |
| "grad_norm": 0.04216486215591431, | |
| "learning_rate": 0.00010392990946401313, | |
| "loss": 0.0187, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 36.56716417910448, | |
| "grad_norm": 0.03389362245798111, | |
| "learning_rate": 0.00010330137426761135, | |
| "loss": 0.02, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 36.71641791044776, | |
| "grad_norm": 0.06512041389942169, | |
| "learning_rate": 0.00010267270847755048, | |
| "loss": 0.0181, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 36.865671641791046, | |
| "grad_norm": 0.052388746291399, | |
| "learning_rate": 0.00010204393696219117, | |
| "loss": 0.0171, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 37.014925373134325, | |
| "grad_norm": 0.16947126388549805, | |
| "learning_rate": 0.00010141508459407623, | |
| "loss": 0.0205, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 37.16417910447761, | |
| "grad_norm": 0.03826919198036194, | |
| "learning_rate": 0.00010078617624894684, | |
| "loss": 0.0168, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 37.3134328358209, | |
| "grad_norm": 0.033569592982530594, | |
| "learning_rate": 0.00010015723680475846, | |
| "loss": 0.0162, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 37.3134328358209, | |
| "eval_loss": 2.9721479415893555, | |
| "eval_runtime": 34.6726, | |
| "eval_samples_per_second": 9.143, | |
| "eval_steps_per_second": 4.586, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 37.46268656716418, | |
| "grad_norm": 0.03183290734887123, | |
| "learning_rate": 9.95282911406968e-05, | |
| "loss": 0.0175, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 37.61194029850746, | |
| "grad_norm": 0.03882099688053131, | |
| "learning_rate": 9.889936413619356e-05, | |
| "loss": 0.0162, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 37.76119402985075, | |
| "grad_norm": 0.08923082053661346, | |
| "learning_rate": 9.827048066994225e-05, | |
| "loss": 0.0206, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 37.91044776119403, | |
| "grad_norm": 0.07512518018484116, | |
| "learning_rate": 9.764166561891432e-05, | |
| "loss": 0.018, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 38.059701492537314, | |
| "grad_norm": 0.07171567529439926, | |
| "learning_rate": 9.70129438573747e-05, | |
| "loss": 0.0189, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 38.208955223880594, | |
| "grad_norm": 0.041823286563158035, | |
| "learning_rate": 9.63843402558981e-05, | |
| "loss": 0.0182, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 38.35820895522388, | |
| "grad_norm": 0.042579639703035355, | |
| "learning_rate": 9.57558796803852e-05, | |
| "loss": 0.015, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 38.507462686567166, | |
| "grad_norm": 0.032053157687187195, | |
| "learning_rate": 9.512758699107879e-05, | |
| "loss": 0.0192, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 38.656716417910445, | |
| "grad_norm": 0.03747331723570824, | |
| "learning_rate": 9.449948704158071e-05, | |
| "loss": 0.0185, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 38.80597014925373, | |
| "grad_norm": 0.03378698602318764, | |
| "learning_rate": 9.38716046778684e-05, | |
| "loss": 0.0177, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 38.95522388059702, | |
| "grad_norm": 0.0920565128326416, | |
| "learning_rate": 9.324396473731217e-05, | |
| "loss": 0.0186, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 39.1044776119403, | |
| "grad_norm": 0.10099617391824722, | |
| "learning_rate": 9.261659204769284e-05, | |
| "loss": 0.017, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 39.25373134328358, | |
| "grad_norm": 0.040173519402742386, | |
| "learning_rate": 9.198951142621929e-05, | |
| "loss": 0.0175, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 39.40298507462686, | |
| "grad_norm": 0.04502606391906738, | |
| "learning_rate": 9.136274767854716e-05, | |
| "loss": 0.0181, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 39.55223880597015, | |
| "grad_norm": 0.039172179996967316, | |
| "learning_rate": 9.07363255977973e-05, | |
| "loss": 0.018, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 39.701492537313435, | |
| "grad_norm": 0.05952875688672066, | |
| "learning_rate": 9.011026996357503e-05, | |
| "loss": 0.0171, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 39.850746268656714, | |
| "grad_norm": 0.03593125194311142, | |
| "learning_rate": 8.948460554099018e-05, | |
| "loss": 0.017, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.041005708277225494, | |
| "learning_rate": 8.885935707967716e-05, | |
| "loss": 0.0164, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 40.149253731343286, | |
| "grad_norm": 0.03663647174835205, | |
| "learning_rate": 8.823454931281616e-05, | |
| "loss": 0.0177, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 40.298507462686565, | |
| "grad_norm": 0.034003522247076035, | |
| "learning_rate": 8.76102069561545e-05, | |
| "loss": 0.0166, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 40.44776119402985, | |
| "grad_norm": 0.06096246466040611, | |
| "learning_rate": 8.698635470702923e-05, | |
| "loss": 0.0166, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 40.59701492537313, | |
| "grad_norm": 0.03656260296702385, | |
| "learning_rate": 8.636301724339004e-05, | |
| "loss": 0.0162, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 40.74626865671642, | |
| "grad_norm": 0.03990943357348442, | |
| "learning_rate": 8.574021922282292e-05, | |
| "loss": 0.018, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 40.8955223880597, | |
| "grad_norm": 0.03584331274032593, | |
| "learning_rate": 8.511798528157512e-05, | |
| "loss": 0.0184, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 41.04477611940298, | |
| "grad_norm": 0.027507085353136063, | |
| "learning_rate": 8.449634003358022e-05, | |
| "loss": 0.0163, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 41.19402985074627, | |
| "grad_norm": 0.03266240283846855, | |
| "learning_rate": 8.387530806948476e-05, | |
| "loss": 0.0175, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 41.343283582089555, | |
| "grad_norm": 0.041966021060943604, | |
| "learning_rate": 8.325491395567541e-05, | |
| "loss": 0.0175, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 41.492537313432834, | |
| "grad_norm": 0.03868953138589859, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 0.0179, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 41.64179104477612, | |
| "grad_norm": 0.03475317731499672, | |
| "learning_rate": 8.201613741733203e-05, | |
| "loss": 0.0159, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 41.791044776119406, | |
| "grad_norm": 0.04897564649581909, | |
| "learning_rate": 8.13978039955308e-05, | |
| "loss": 0.0183, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 41.940298507462686, | |
| "grad_norm": 0.03372865915298462, | |
| "learning_rate": 8.078020642754274e-05, | |
| "loss": 0.0169, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 42.08955223880597, | |
| "grad_norm": 0.046989791095256805, | |
| "learning_rate": 8.016336914389874e-05, | |
| "loss": 0.0174, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 42.23880597014925, | |
| "grad_norm": 0.03214934468269348, | |
| "learning_rate": 7.954731654505491e-05, | |
| "loss": 0.0146, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 42.38805970149254, | |
| "grad_norm": 0.05004828795790672, | |
| "learning_rate": 7.89320730004274e-05, | |
| "loss": 0.0182, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 42.53731343283582, | |
| "grad_norm": 0.031027644872665405, | |
| "learning_rate": 7.831766284742807e-05, | |
| "loss": 0.0156, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 42.6865671641791, | |
| "grad_norm": 0.04323369264602661, | |
| "learning_rate": 7.77041103905023e-05, | |
| "loss": 0.0177, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 42.83582089552239, | |
| "grad_norm": 0.03035310097038746, | |
| "learning_rate": 7.709143990016702e-05, | |
| "loss": 0.0164, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 42.985074626865675, | |
| "grad_norm": 0.03815029561519623, | |
| "learning_rate": 7.6479675612051e-05, | |
| "loss": 0.0205, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 43.134328358208954, | |
| "grad_norm": 0.039352428168058395, | |
| "learning_rate": 7.586884172593609e-05, | |
| "loss": 0.0169, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 43.28358208955224, | |
| "grad_norm": 0.031735971570014954, | |
| "learning_rate": 7.525896240479976e-05, | |
| "loss": 0.0159, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 43.43283582089552, | |
| "grad_norm": 0.03945886343717575, | |
| "learning_rate": 7.465006177385953e-05, | |
| "loss": 0.0166, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 43.582089552238806, | |
| "grad_norm": 0.030156582593917847, | |
| "learning_rate": 7.404216391961847e-05, | |
| "loss": 0.0176, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 43.73134328358209, | |
| "grad_norm": 0.03693369776010513, | |
| "learning_rate": 7.343529288891239e-05, | |
| "loss": 0.0159, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 43.88059701492537, | |
| "grad_norm": 0.03338786959648132, | |
| "learning_rate": 7.282947268795877e-05, | |
| "loss": 0.0171, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 44.02985074626866, | |
| "grad_norm": 0.021122202277183533, | |
| "learning_rate": 7.222472728140695e-05, | |
| "loss": 0.0178, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 44.17910447761194, | |
| "grad_norm": 0.02877069264650345, | |
| "learning_rate": 7.162108059139032e-05, | |
| "loss": 0.0186, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 44.32835820895522, | |
| "grad_norm": 0.07340731471776962, | |
| "learning_rate": 7.101855649657991e-05, | |
| "loss": 0.0187, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 44.47761194029851, | |
| "grad_norm": 0.04397398233413696, | |
| "learning_rate": 7.041717883123977e-05, | |
| "loss": 0.0169, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 44.62686567164179, | |
| "grad_norm": 0.03213610127568245, | |
| "learning_rate": 6.981697138428434e-05, | |
| "loss": 0.017, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 44.776119402985074, | |
| "grad_norm": 0.024713682010769844, | |
| "learning_rate": 6.921795789833723e-05, | |
| "loss": 0.016, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 44.776119402985074, | |
| "eval_loss": 3.2444493770599365, | |
| "eval_runtime": 34.6924, | |
| "eval_samples_per_second": 9.137, | |
| "eval_steps_per_second": 4.583, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 44.92537313432836, | |
| "grad_norm": 0.028246300294995308, | |
| "learning_rate": 6.862016206879216e-05, | |
| "loss": 0.017, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 45.07462686567164, | |
| "grad_norm": 0.03457006812095642, | |
| "learning_rate": 6.802360754287547e-05, | |
| "loss": 0.0162, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 45.223880597014926, | |
| "grad_norm": 0.04756162688136101, | |
| "learning_rate": 6.742831791871096e-05, | |
| "loss": 0.0176, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 45.37313432835821, | |
| "grad_norm": 0.03616981953382492, | |
| "learning_rate": 6.683431674438612e-05, | |
| "loss": 0.0153, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 45.52238805970149, | |
| "grad_norm": 0.03674984350800514, | |
| "learning_rate": 6.624162751702076e-05, | |
| "loss": 0.0155, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 45.67164179104478, | |
| "grad_norm": 0.026007067412137985, | |
| "learning_rate": 6.565027368183769e-05, | |
| "loss": 0.0167, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 45.82089552238806, | |
| "grad_norm": 0.027320370078086853, | |
| "learning_rate": 6.506027863123492e-05, | |
| "loss": 0.0183, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 45.97014925373134, | |
| "grad_norm": 0.037796761840581894, | |
| "learning_rate": 6.447166570386063e-05, | |
| "loss": 0.0184, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 46.11940298507463, | |
| "grad_norm": 0.03350226208567619, | |
| "learning_rate": 6.388445818368991e-05, | |
| "loss": 0.0156, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 46.26865671641791, | |
| "grad_norm": 0.026382336392998695, | |
| "learning_rate": 6.329867929910347e-05, | |
| "loss": 0.0156, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 46.417910447761194, | |
| "grad_norm": 0.03873557224869728, | |
| "learning_rate": 6.271435222196916e-05, | |
| "loss": 0.021, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 46.56716417910448, | |
| "grad_norm": 0.039839617908000946, | |
| "learning_rate": 6.213150006672499e-05, | |
| "loss": 0.0172, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 46.71641791044776, | |
| "grad_norm": 0.03648209199309349, | |
| "learning_rate": 6.15501458894651e-05, | |
| "loss": 0.0161, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 46.865671641791046, | |
| "grad_norm": 0.04052448272705078, | |
| "learning_rate": 6.097031268702746e-05, | |
| "loss": 0.0178, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 47.014925373134325, | |
| "grad_norm": 0.03359508886933327, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 0.0172, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 47.16417910447761, | |
| "grad_norm": 0.03347504511475563, | |
| "learning_rate": 5.981530089223489e-05, | |
| "loss": 0.0162, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 47.3134328358209, | |
| "grad_norm": 0.03658764436841011, | |
| "learning_rate": 5.924016798910037e-05, | |
| "loss": 0.0164, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 47.46268656716418, | |
| "grad_norm": 0.02819441817700863, | |
| "learning_rate": 5.866664743742162e-05, | |
| "loss": 0.0169, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 47.61194029850746, | |
| "grad_norm": 0.03421681374311447, | |
| "learning_rate": 5.809476192415905e-05, | |
| "loss": 0.0185, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 47.76119402985075, | |
| "grad_norm": 0.050075121223926544, | |
| "learning_rate": 5.752453407159522e-05, | |
| "loss": 0.0178, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 47.91044776119403, | |
| "grad_norm": 0.030316824093461037, | |
| "learning_rate": 5.69559864364402e-05, | |
| "loss": 0.0151, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 48.059701492537314, | |
| "grad_norm": 0.053834062069654465, | |
| "learning_rate": 5.6389141508938903e-05, | |
| "loss": 0.0173, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 48.208955223880594, | |
| "grad_norm": 0.036416202783584595, | |
| "learning_rate": 5.5824021711981686e-05, | |
| "loss": 0.0158, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 48.35820895522388, | |
| "grad_norm": 0.030989298596978188, | |
| "learning_rate": 5.5260649400217326e-05, | |
| "loss": 0.017, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 48.507462686567166, | |
| "grad_norm": 0.023621903732419014, | |
| "learning_rate": 5.469904685916861e-05, | |
| "loss": 0.0202, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 48.656716417910445, | |
| "grad_norm": 0.03500565141439438, | |
| "learning_rate": 5.4139236304350935e-05, | |
| "loss": 0.0178, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 48.80597014925373, | |
| "grad_norm": 0.040891725569963455, | |
| "learning_rate": 5.3581239880393375e-05, | |
| "loss": 0.0179, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 48.95522388059702, | |
| "grad_norm": 0.027273844927549362, | |
| "learning_rate": 5.302507966016295e-05, | |
| "loss": 0.014, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 49.1044776119403, | |
| "grad_norm": 0.02311500906944275, | |
| "learning_rate": 5.247077764389099e-05, | |
| "loss": 0.0167, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 49.25373134328358, | |
| "grad_norm": 0.052333466708660126, | |
| "learning_rate": 5.191835575830352e-05, | |
| "loss": 0.0177, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 49.40298507462686, | |
| "grad_norm": 0.042159441858530045, | |
| "learning_rate": 5.136783585575336e-05, | |
| "loss": 0.0162, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 49.55223880597015, | |
| "grad_norm": 0.025089124217629433, | |
| "learning_rate": 5.081923971335582e-05, | |
| "loss": 0.0166, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 49.701492537313435, | |
| "grad_norm": 0.03574312478303909, | |
| "learning_rate": 5.0272589032127594e-05, | |
| "loss": 0.0168, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 49.850746268656714, | |
| "grad_norm": 0.035489972680807114, | |
| "learning_rate": 4.972790543612783e-05, | |
| "loss": 0.0149, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 0.04449532926082611, | |
| "learning_rate": 4.918521047160308e-05, | |
| "loss": 0.0185, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 50.149253731343286, | |
| "grad_norm": 0.037168972194194794, | |
| "learning_rate": 4.864452560613485e-05, | |
| "loss": 0.0159, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 50.298507462686565, | |
| "grad_norm": 0.032931167632341385, | |
| "learning_rate": 4.810587222779043e-05, | |
| "loss": 0.0165, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 50.44776119402985, | |
| "grad_norm": 0.08401083201169968, | |
| "learning_rate": 4.756927164427685e-05, | |
| "loss": 0.0186, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 50.59701492537313, | |
| "grad_norm": 0.03464508801698685, | |
| "learning_rate": 4.703474508209793e-05, | |
| "loss": 0.0156, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 50.74626865671642, | |
| "grad_norm": 0.03466491773724556, | |
| "learning_rate": 4.650231368571486e-05, | |
| "loss": 0.0173, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 50.8955223880597, | |
| "grad_norm": 0.030173856765031815, | |
| "learning_rate": 4.597199851670932e-05, | |
| "loss": 0.0161, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 51.04477611940298, | |
| "grad_norm": 0.029572507366538048, | |
| "learning_rate": 4.54438205529508e-05, | |
| "loss": 0.0162, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 51.19402985074627, | |
| "grad_norm": 0.032448384910821915, | |
| "learning_rate": 4.491780068776663e-05, | |
| "loss": 0.0173, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 51.343283582089555, | |
| "grad_norm": 0.029228707775473595, | |
| "learning_rate": 4.4393959729115244e-05, | |
| "loss": 0.0182, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 51.492537313432834, | |
| "grad_norm": 0.03576023131608963, | |
| "learning_rate": 4.387231839876349e-05, | |
| "loss": 0.0156, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 51.64179104477612, | |
| "grad_norm": 0.02814898081123829, | |
| "learning_rate": 4.335289733146665e-05, | |
| "loss": 0.0163, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 51.791044776119406, | |
| "grad_norm": 0.05810336023569107, | |
| "learning_rate": 4.283571707415214e-05, | |
| "loss": 0.0172, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 51.940298507462686, | |
| "grad_norm": 0.03793029487133026, | |
| "learning_rate": 4.2320798085107036e-05, | |
| "loss": 0.0177, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 52.08955223880597, | |
| "grad_norm": 0.02555203065276146, | |
| "learning_rate": 4.18081607331685e-05, | |
| "loss": 0.0168, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 52.23880597014925, | |
| "grad_norm": 0.028642071411013603, | |
| "learning_rate": 4.129782529691815e-05, | |
| "loss": 0.0167, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 52.23880597014925, | |
| "eval_loss": 3.4037022590637207, | |
| "eval_runtime": 34.709, | |
| "eval_samples_per_second": 9.133, | |
| "eval_steps_per_second": 4.581, | |
| "step": 3500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 75, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.873632111370404e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |