diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,3682 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.0, + "eval_steps": 500, + "global_step": 456, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0043907793633369925, + "grad_norm": 2.833900695415164, + "learning_rate": 4.347826086956522e-07, + "loss": 0.6168212890625, + "step": 1, + "token_acc": 0.8299740080322707 + }, + { + "epoch": 0.008781558726673985, + "grad_norm": 2.441954178715713, + "learning_rate": 8.695652173913044e-07, + "loss": 0.5771484375, + "step": 2, + "token_acc": 0.8402379919383847 + }, + { + "epoch": 0.013172338090010977, + "grad_norm": 2.3400168784617796, + "learning_rate": 1.3043478260869566e-06, + "loss": 0.5506591796875, + "step": 3, + "token_acc": 0.8475516568344045 + }, + { + "epoch": 0.01756311745334797, + "grad_norm": 2.380407017832379, + "learning_rate": 1.7391304347826088e-06, + "loss": 0.5848388671875, + "step": 4, + "token_acc": 0.8378425120730101 + }, + { + "epoch": 0.021953896816684963, + "grad_norm": 1.7638106380932321, + "learning_rate": 2.173913043478261e-06, + "loss": 0.5223388671875, + "step": 5, + "token_acc": 0.8545390186376319 + }, + { + "epoch": 0.026344676180021953, + "grad_norm": 1.211433347249036, + "learning_rate": 2.6086956521739132e-06, + "loss": 0.5751953125, + "step": 6, + "token_acc": 0.8389151935417897 + }, + { + "epoch": 0.030735455543358946, + "grad_norm": 1.0146058725955927, + "learning_rate": 3.043478260869566e-06, + "loss": 0.5172119140625, + "step": 7, + "token_acc": 0.8548223173710081 + }, + { + "epoch": 0.03512623490669594, + "grad_norm": 1.0748297263345705, + "learning_rate": 3.4782608695652175e-06, + "loss": 0.5316162109375, + "step": 8, + "token_acc": 0.8489027039951763 + }, + { + "epoch": 0.03951701427003293, + "grad_norm": 1.0857562596183432, + "learning_rate": 3.91304347826087e-06, + "loss": 0.535400390625, + "step": 9, + "token_acc": 0.8491097827546807 + }, + { + "epoch": 0.043907793633369926, + "grad_norm": 0.9224300194653032, + "learning_rate": 4.347826086956522e-06, + "loss": 0.4791259765625, + "step": 10, + "token_acc": 0.8638291338521461 + }, + { + "epoch": 0.048298572996706916, + "grad_norm": 1.0296199866637192, + "learning_rate": 4.782608695652174e-06, + "loss": 0.50146484375, + "step": 11, + "token_acc": 0.8564573991865367 + }, + { + "epoch": 0.052689352360043906, + "grad_norm": 0.9873486692422354, + "learning_rate": 5.2173913043478265e-06, + "loss": 0.5025634765625, + "step": 12, + "token_acc": 0.8529394427633906 + }, + { + "epoch": 0.0570801317233809, + "grad_norm": 0.8185948156277064, + "learning_rate": 5.652173913043479e-06, + "loss": 0.517333984375, + "step": 13, + "token_acc": 0.8504350836336171 + }, + { + "epoch": 0.06147091108671789, + "grad_norm": 0.6631444876743735, + "learning_rate": 6.086956521739132e-06, + "loss": 0.4774169921875, + "step": 14, + "token_acc": 0.8615918882424461 + }, + { + "epoch": 0.06586169045005488, + "grad_norm": 0.7794489431842011, + "learning_rate": 6.521739130434783e-06, + "loss": 0.4833984375, + "step": 15, + "token_acc": 0.8593623551327664 + }, + { + "epoch": 0.07025246981339188, + "grad_norm": 0.7597840066367508, + "learning_rate": 6.956521739130435e-06, + "loss": 0.5009765625, + "step": 16, + "token_acc": 0.8542362145969853 + }, + { + "epoch": 0.07464324917672886, + "grad_norm": 0.6397542926095154, + "learning_rate": 7.391304347826087e-06, + "loss": 0.4658203125, + "step": 17, + "token_acc": 0.8663710059193312 + }, + { + "epoch": 0.07903402854006586, + "grad_norm": 0.6460169235950904, + "learning_rate": 7.82608695652174e-06, + "loss": 0.492431640625, + "step": 18, + "token_acc": 0.8560206344178469 + }, + { + "epoch": 0.08342480790340286, + "grad_norm": 0.5555957965562055, + "learning_rate": 8.260869565217392e-06, + "loss": 0.477783203125, + "step": 19, + "token_acc": 0.8599260421013252 + }, + { + "epoch": 0.08781558726673985, + "grad_norm": 0.6027995894426754, + "learning_rate": 8.695652173913044e-06, + "loss": 0.46710205078125, + "step": 20, + "token_acc": 0.8625439077051981 + }, + { + "epoch": 0.09220636663007684, + "grad_norm": 0.5687893939186547, + "learning_rate": 9.130434782608697e-06, + "loss": 0.4857177734375, + "step": 21, + "token_acc": 0.8567808460493117 + }, + { + "epoch": 0.09659714599341383, + "grad_norm": 0.46547299222293903, + "learning_rate": 9.565217391304349e-06, + "loss": 0.451416015625, + "step": 22, + "token_acc": 0.8687051892909547 + }, + { + "epoch": 0.10098792535675083, + "grad_norm": 0.5093205577984702, + "learning_rate": 1e-05, + "loss": 0.443115234375, + "step": 23, + "token_acc": 0.8681511754829399 + }, + { + "epoch": 0.10537870472008781, + "grad_norm": 0.43975124277897326, + "learning_rate": 9.999868398131282e-06, + "loss": 0.43994140625, + "step": 24, + "token_acc": 0.8702387489069204 + }, + { + "epoch": 0.10976948408342481, + "grad_norm": 0.45089180449271005, + "learning_rate": 9.999473599452746e-06, + "loss": 0.4444580078125, + "step": 25, + "token_acc": 0.8698691544595688 + }, + { + "epoch": 0.1141602634467618, + "grad_norm": 0.43879224523868243, + "learning_rate": 9.99881562474689e-06, + "loss": 0.4208984375, + "step": 26, + "token_acc": 0.8750299328244403 + }, + { + "epoch": 0.11855104281009879, + "grad_norm": 0.4316116408140785, + "learning_rate": 9.997894508649995e-06, + "loss": 0.4478759765625, + "step": 27, + "token_acc": 0.8676176445929599 + }, + { + "epoch": 0.12294182217343579, + "grad_norm": 0.4268412361226033, + "learning_rate": 9.996710299650302e-06, + "loss": 0.4429931640625, + "step": 28, + "token_acc": 0.8691655605967833 + }, + { + "epoch": 0.12733260153677278, + "grad_norm": 0.40644251545911186, + "learning_rate": 9.995263060085456e-06, + "loss": 0.405029296875, + "step": 29, + "token_acc": 0.8785781306123047 + }, + { + "epoch": 0.13172338090010977, + "grad_norm": 0.4358630834831772, + "learning_rate": 9.99355286613923e-06, + "loss": 0.4681396484375, + "step": 30, + "token_acc": 0.8614033803048956 + }, + { + "epoch": 0.13611416026344675, + "grad_norm": 0.5006055484811447, + "learning_rate": 9.991579807837511e-06, + "loss": 0.4735107421875, + "step": 31, + "token_acc": 0.8569515531626238 + }, + { + "epoch": 0.14050493962678376, + "grad_norm": 0.4339503914133248, + "learning_rate": 9.989343989043563e-06, + "loss": 0.43487548828125, + "step": 32, + "token_acc": 0.8693686707945142 + }, + { + "epoch": 0.14489571899012074, + "grad_norm": 0.40763103283199237, + "learning_rate": 9.98684552745256e-06, + "loss": 0.415771484375, + "step": 33, + "token_acc": 0.8742195510606712 + }, + { + "epoch": 0.14928649835345773, + "grad_norm": 0.427575395649408, + "learning_rate": 9.984084554585387e-06, + "loss": 0.434814453125, + "step": 34, + "token_acc": 0.8698706700750862 + }, + { + "epoch": 0.15367727771679474, + "grad_norm": 0.40322122738955996, + "learning_rate": 9.98106121578172e-06, + "loss": 0.4114990234375, + "step": 35, + "token_acc": 0.8744793955523036 + }, + { + "epoch": 0.15806805708013172, + "grad_norm": 0.37400365026886323, + "learning_rate": 9.977775670192373e-06, + "loss": 0.4068603515625, + "step": 36, + "token_acc": 0.8780122698049075 + }, + { + "epoch": 0.16245883644346873, + "grad_norm": 0.368282839817705, + "learning_rate": 9.97422809077092e-06, + "loss": 0.4403076171875, + "step": 37, + "token_acc": 0.8691931492703077 + }, + { + "epoch": 0.1668496158068057, + "grad_norm": 0.40472633033080596, + "learning_rate": 9.970418664264596e-06, + "loss": 0.42376708984375, + "step": 38, + "token_acc": 0.8731970977885961 + }, + { + "epoch": 0.1712403951701427, + "grad_norm": 0.3536787769901149, + "learning_rate": 9.966347591204459e-06, + "loss": 0.40185546875, + "step": 39, + "token_acc": 0.8780035917624731 + }, + { + "epoch": 0.1756311745334797, + "grad_norm": 0.3953310789443251, + "learning_rate": 9.962015085894838e-06, + "loss": 0.412109375, + "step": 40, + "token_acc": 0.8752016328884643 + }, + { + "epoch": 0.1800219538968167, + "grad_norm": 0.4148084674386976, + "learning_rate": 9.957421376402053e-06, + "loss": 0.42657470703125, + "step": 41, + "token_acc": 0.8705481901069846 + }, + { + "epoch": 0.18441273326015367, + "grad_norm": 0.37789346575572064, + "learning_rate": 9.9525667045424e-06, + "loss": 0.42022705078125, + "step": 42, + "token_acc": 0.8736403899467848 + }, + { + "epoch": 0.18880351262349068, + "grad_norm": 0.3667318215801622, + "learning_rate": 9.94745132586944e-06, + "loss": 0.406982421875, + "step": 43, + "token_acc": 0.876900068934092 + }, + { + "epoch": 0.19319429198682767, + "grad_norm": 0.41464715087306936, + "learning_rate": 9.942075509660527e-06, + "loss": 0.4429931640625, + "step": 44, + "token_acc": 0.8660984334018896 + }, + { + "epoch": 0.19758507135016465, + "grad_norm": 0.3984507277058535, + "learning_rate": 9.936439538902644e-06, + "loss": 0.41485595703125, + "step": 45, + "token_acc": 0.8746052349855475 + }, + { + "epoch": 0.20197585071350166, + "grad_norm": 0.3563059511829739, + "learning_rate": 9.93054371027751e-06, + "loss": 0.42730712890625, + "step": 46, + "token_acc": 0.8729698510081998 + }, + { + "epoch": 0.20636663007683864, + "grad_norm": 0.4016584355027647, + "learning_rate": 9.924388334145943e-06, + "loss": 0.4156494140625, + "step": 47, + "token_acc": 0.8750582994258512 + }, + { + "epoch": 0.21075740944017562, + "grad_norm": 0.3983586904345504, + "learning_rate": 9.917973734531549e-06, + "loss": 0.429931640625, + "step": 48, + "token_acc": 0.870179892896137 + }, + { + "epoch": 0.21514818880351264, + "grad_norm": 0.3643013846346374, + "learning_rate": 9.911300249103646e-06, + "loss": 0.38763427734375, + "step": 49, + "token_acc": 0.8829403403914137 + }, + { + "epoch": 0.21953896816684962, + "grad_norm": 0.37469353915411313, + "learning_rate": 9.904368229159494e-06, + "loss": 0.4312744140625, + "step": 50, + "token_acc": 0.870172455916827 + }, + { + "epoch": 0.2239297475301866, + "grad_norm": 0.3627070621832165, + "learning_rate": 9.897178039605803e-06, + "loss": 0.3935546875, + "step": 51, + "token_acc": 0.8809188065186905 + }, + { + "epoch": 0.2283205268935236, + "grad_norm": 0.396086671715509, + "learning_rate": 9.889730058939529e-06, + "loss": 0.4736328125, + "step": 52, + "token_acc": 0.8598151554777367 + }, + { + "epoch": 0.2327113062568606, + "grad_norm": 0.37766425012501803, + "learning_rate": 9.88202467922794e-06, + "loss": 0.39007568359375, + "step": 53, + "token_acc": 0.880743890775099 + }, + { + "epoch": 0.23710208562019758, + "grad_norm": 0.4341034390676532, + "learning_rate": 9.874062306087983e-06, + "loss": 0.48052978515625, + "step": 54, + "token_acc": 0.8605994944023113 + }, + { + "epoch": 0.2414928649835346, + "grad_norm": 0.3697701194722883, + "learning_rate": 9.865843358664933e-06, + "loss": 0.40374755859375, + "step": 55, + "token_acc": 0.8786038905474529 + }, + { + "epoch": 0.24588364434687157, + "grad_norm": 0.38785842426375305, + "learning_rate": 9.857368269610325e-06, + "loss": 0.37890625, + "step": 56, + "token_acc": 0.8835334883145686 + }, + { + "epoch": 0.2502744237102086, + "grad_norm": 0.41653734477290827, + "learning_rate": 9.848637485059183e-06, + "loss": 0.4058837890625, + "step": 57, + "token_acc": 0.8766154925985425 + }, + { + "epoch": 0.25466520307354557, + "grad_norm": 0.3957651287576482, + "learning_rate": 9.83965146460653e-06, + "loss": 0.4259033203125, + "step": 58, + "token_acc": 0.8705599895649054 + }, + { + "epoch": 0.25905598243688255, + "grad_norm": 0.41213136511886145, + "learning_rate": 9.830410681283203e-06, + "loss": 0.41693115234375, + "step": 59, + "token_acc": 0.8748202120166441 + }, + { + "epoch": 0.26344676180021953, + "grad_norm": 0.3808997874258946, + "learning_rate": 9.820915621530939e-06, + "loss": 0.40032958984375, + "step": 60, + "token_acc": 0.878064430260808 + }, + { + "epoch": 0.2678375411635565, + "grad_norm": 0.34694502133664223, + "learning_rate": 9.811166785176785e-06, + "loss": 0.39154052734375, + "step": 61, + "token_acc": 0.8813461234940287 + }, + { + "epoch": 0.2722283205268935, + "grad_norm": 0.3867337659427633, + "learning_rate": 9.80116468540677e-06, + "loss": 0.40081787109375, + "step": 62, + "token_acc": 0.8786373357651632 + }, + { + "epoch": 0.27661909989023054, + "grad_norm": 0.3863950377574671, + "learning_rate": 9.790909848738907e-06, + "loss": 0.39849853515625, + "step": 63, + "token_acc": 0.8796994213840694 + }, + { + "epoch": 0.2810098792535675, + "grad_norm": 0.3776439635090338, + "learning_rate": 9.780402814995458e-06, + "loss": 0.402099609375, + "step": 64, + "token_acc": 0.8788000518572756 + }, + { + "epoch": 0.2854006586169045, + "grad_norm": 0.41515630824304595, + "learning_rate": 9.76964413727454e-06, + "loss": 0.4017333984375, + "step": 65, + "token_acc": 0.8765627154349158 + }, + { + "epoch": 0.2897914379802415, + "grad_norm": 0.35183225407905605, + "learning_rate": 9.758634381920982e-06, + "loss": 0.380126953125, + "step": 66, + "token_acc": 0.8842076941642893 + }, + { + "epoch": 0.29418221734357847, + "grad_norm": 0.44297721995987493, + "learning_rate": 9.747374128496541e-06, + "loss": 0.4088134765625, + "step": 67, + "token_acc": 0.8751857321674607 + }, + { + "epoch": 0.29857299670691545, + "grad_norm": 0.37931729187838137, + "learning_rate": 9.735863969749373e-06, + "loss": 0.3839111328125, + "step": 68, + "token_acc": 0.8825879497953658 + }, + { + "epoch": 0.3029637760702525, + "grad_norm": 0.39424410720211267, + "learning_rate": 9.724104511582838e-06, + "loss": 0.3985595703125, + "step": 69, + "token_acc": 0.8775368384478103 + }, + { + "epoch": 0.30735455543358947, + "grad_norm": 0.418704027987836, + "learning_rate": 9.712096373023603e-06, + "loss": 0.4119873046875, + "step": 70, + "token_acc": 0.8738080900414597 + }, + { + "epoch": 0.31174533479692645, + "grad_norm": 0.381156542405727, + "learning_rate": 9.699840186189061e-06, + "loss": 0.40582275390625, + "step": 71, + "token_acc": 0.8782644158643692 + }, + { + "epoch": 0.31613611416026344, + "grad_norm": 0.40118752065949825, + "learning_rate": 9.687336596254045e-06, + "loss": 0.400634765625, + "step": 72, + "token_acc": 0.8780178199420579 + }, + { + "epoch": 0.3205268935236004, + "grad_norm": 0.35733629669262923, + "learning_rate": 9.674586261416874e-06, + "loss": 0.41162109375, + "step": 73, + "token_acc": 0.8753492148975917 + }, + { + "epoch": 0.32491767288693746, + "grad_norm": 0.3456260403324577, + "learning_rate": 9.66158985286471e-06, + "loss": 0.39501953125, + "step": 74, + "token_acc": 0.8803660773771422 + }, + { + "epoch": 0.32930845225027444, + "grad_norm": 0.36425476383529126, + "learning_rate": 9.648348054738208e-06, + "loss": 0.3846435546875, + "step": 75, + "token_acc": 0.8830634004075755 + }, + { + "epoch": 0.3336992316136114, + "grad_norm": 0.3782416134794376, + "learning_rate": 9.634861564095525e-06, + "loss": 0.3890380859375, + "step": 76, + "token_acc": 0.8794278043109145 + }, + { + "epoch": 0.3380900109769484, + "grad_norm": 0.3551769340083096, + "learning_rate": 9.621131090875603e-06, + "loss": 0.3880615234375, + "step": 77, + "token_acc": 0.8825253450858896 + }, + { + "epoch": 0.3424807903402854, + "grad_norm": 0.3585167918667367, + "learning_rate": 9.607157357860823e-06, + "loss": 0.38714599609375, + "step": 78, + "token_acc": 0.8821463111628309 + }, + { + "epoch": 0.3468715697036224, + "grad_norm": 0.3362331147837606, + "learning_rate": 9.59294110063893e-06, + "loss": 0.38916015625, + "step": 79, + "token_acc": 0.8817058995514544 + }, + { + "epoch": 0.3512623490669594, + "grad_norm": 0.3763003424816892, + "learning_rate": 9.578483067564335e-06, + "loss": 0.39990234375, + "step": 80, + "token_acc": 0.8777995931650056 + }, + { + "epoch": 0.3556531284302964, + "grad_norm": 0.37078466401656457, + "learning_rate": 9.563784019718704e-06, + "loss": 0.3997802734375, + "step": 81, + "token_acc": 0.8785774295372237 + }, + { + "epoch": 0.3600439077936334, + "grad_norm": 0.34346869451831064, + "learning_rate": 9.548844730870903e-06, + "loss": 0.39178466796875, + "step": 82, + "token_acc": 0.88060131256661 + }, + { + "epoch": 0.36443468715697036, + "grad_norm": 0.40809336341807895, + "learning_rate": 9.533665987436262e-06, + "loss": 0.41259765625, + "step": 83, + "token_acc": 0.8734696553855953 + }, + { + "epoch": 0.36882546652030734, + "grad_norm": 0.3670168129309835, + "learning_rate": 9.518248588435185e-06, + "loss": 0.39404296875, + "step": 84, + "token_acc": 0.878865753468321 + }, + { + "epoch": 0.3732162458836443, + "grad_norm": 0.4234497007384386, + "learning_rate": 9.502593345451078e-06, + "loss": 0.39080810546875, + "step": 85, + "token_acc": 0.8812611243285717 + }, + { + "epoch": 0.37760702524698136, + "grad_norm": 0.4190337029590078, + "learning_rate": 9.486701082587635e-06, + "loss": 0.44061279296875, + "step": 86, + "token_acc": 0.8672058574385492 + }, + { + "epoch": 0.38199780461031835, + "grad_norm": 0.38441974581357985, + "learning_rate": 9.470572636425451e-06, + "loss": 0.37774658203125, + "step": 87, + "token_acc": 0.8834160287255358 + }, + { + "epoch": 0.38638858397365533, + "grad_norm": 0.3845038406186474, + "learning_rate": 9.454208855977986e-06, + "loss": 0.40692138671875, + "step": 88, + "token_acc": 0.8772729239690163 + }, + { + "epoch": 0.3907793633369923, + "grad_norm": 0.3980068297336857, + "learning_rate": 9.437610602646878e-06, + "loss": 0.38934326171875, + "step": 89, + "token_acc": 0.8812292716986169 + }, + { + "epoch": 0.3951701427003293, + "grad_norm": 0.3515321052605919, + "learning_rate": 9.420778750176588e-06, + "loss": 0.3980712890625, + "step": 90, + "token_acc": 0.880305903243086 + }, + { + "epoch": 0.3995609220636663, + "grad_norm": 0.4150437186372563, + "learning_rate": 9.403714184608411e-06, + "loss": 0.41363525390625, + "step": 91, + "token_acc": 0.8744749752235499 + }, + { + "epoch": 0.4039517014270033, + "grad_norm": 0.3644452606552959, + "learning_rate": 9.386417804233836e-06, + "loss": 0.39398193359375, + "step": 92, + "token_acc": 0.8787620934599563 + }, + { + "epoch": 0.4083424807903403, + "grad_norm": 0.4160904416711138, + "learning_rate": 9.36889051954725e-06, + "loss": 0.40155029296875, + "step": 93, + "token_acc": 0.875743910024713 + }, + { + "epoch": 0.4127332601536773, + "grad_norm": 0.36391609863319885, + "learning_rate": 9.351133253198027e-06, + "loss": 0.3848876953125, + "step": 94, + "token_acc": 0.8823650794820584 + }, + { + "epoch": 0.41712403951701427, + "grad_norm": 0.3658547226007181, + "learning_rate": 9.333146939941938e-06, + "loss": 0.4993896484375, + "step": 95, + "token_acc": 0.8619243851157109 + }, + { + "epoch": 0.42151481888035125, + "grad_norm": 0.3849167416364751, + "learning_rate": 9.314932526591956e-06, + "loss": 0.3837890625, + "step": 96, + "token_acc": 0.881912430617248 + }, + { + "epoch": 0.42590559824368823, + "grad_norm": 0.33470552310184915, + "learning_rate": 9.296490971968416e-06, + "loss": 0.4061279296875, + "step": 97, + "token_acc": 0.8772653313604807 + }, + { + "epoch": 0.43029637760702527, + "grad_norm": 0.3615805211697039, + "learning_rate": 9.277823246848537e-06, + "loss": 0.41302490234375, + "step": 98, + "token_acc": 0.874873878812435 + }, + { + "epoch": 0.43468715697036225, + "grad_norm": 0.404823301395682, + "learning_rate": 9.258930333915325e-06, + "loss": 0.3902587890625, + "step": 99, + "token_acc": 0.8795714993290125 + }, + { + "epoch": 0.43907793633369924, + "grad_norm": 0.349751013844174, + "learning_rate": 9.23981322770584e-06, + "loss": 0.3941650390625, + "step": 100, + "token_acc": 0.8788965124760538 + }, + { + "epoch": 0.4434687156970362, + "grad_norm": 0.37824751509711396, + "learning_rate": 9.220472934558838e-06, + "loss": 0.4033203125, + "step": 101, + "token_acc": 0.8775104581880655 + }, + { + "epoch": 0.4478594950603732, + "grad_norm": 0.3193466487997388, + "learning_rate": 9.20091047256181e-06, + "loss": 0.35516357421875, + "step": 102, + "token_acc": 0.8912133045064126 + }, + { + "epoch": 0.4522502744237102, + "grad_norm": 0.34191092849697585, + "learning_rate": 9.181126871497378e-06, + "loss": 0.3431396484375, + "step": 103, + "token_acc": 0.8959809324296724 + }, + { + "epoch": 0.4566410537870472, + "grad_norm": 0.37045247002325327, + "learning_rate": 9.161123172789091e-06, + "loss": 0.384033203125, + "step": 104, + "token_acc": 0.8812417570069138 + }, + { + "epoch": 0.4610318331503842, + "grad_norm": 0.38912337453284385, + "learning_rate": 9.140900429446601e-06, + "loss": 0.40545654296875, + "step": 105, + "token_acc": 0.8761549038076369 + }, + { + "epoch": 0.4654226125137212, + "grad_norm": 0.4088916859694267, + "learning_rate": 9.120459706010233e-06, + "loss": 0.39544677734375, + "step": 106, + "token_acc": 0.8785151421508035 + }, + { + "epoch": 0.4698133918770582, + "grad_norm": 0.4028383560960719, + "learning_rate": 9.099802078494947e-06, + "loss": 0.400390625, + "step": 107, + "token_acc": 0.8766336371460729 + }, + { + "epoch": 0.47420417124039516, + "grad_norm": 0.37361603754710926, + "learning_rate": 9.0789286343337e-06, + "loss": 0.41094970703125, + "step": 108, + "token_acc": 0.874300627166199 + }, + { + "epoch": 0.47859495060373214, + "grad_norm": 0.35129460752742203, + "learning_rate": 9.057840472320192e-06, + "loss": 0.3841552734375, + "step": 109, + "token_acc": 0.8817200742169932 + }, + { + "epoch": 0.4829857299670692, + "grad_norm": 0.3428342088483489, + "learning_rate": 9.036538702551037e-06, + "loss": 0.36383056640625, + "step": 110, + "token_acc": 0.8875094041599138 + }, + { + "epoch": 0.48737650933040616, + "grad_norm": 0.34882689675216083, + "learning_rate": 9.015024446367315e-06, + "loss": 0.35845947265625, + "step": 111, + "token_acc": 0.88992498510808 + }, + { + "epoch": 0.49176728869374314, + "grad_norm": 0.33455190906763604, + "learning_rate": 8.993298836295556e-06, + "loss": 0.373046875, + "step": 112, + "token_acc": 0.8847770512096015 + }, + { + "epoch": 0.4961580680570801, + "grad_norm": 0.3959796808814928, + "learning_rate": 8.971363015988115e-06, + "loss": 0.4105224609375, + "step": 113, + "token_acc": 0.8738999192211215 + }, + { + "epoch": 0.5005488474204172, + "grad_norm": 0.33657519969330657, + "learning_rate": 8.949218140162965e-06, + "loss": 0.37725830078125, + "step": 114, + "token_acc": 0.8840684593222664 + }, + { + "epoch": 0.5049396267837541, + "grad_norm": 0.354805294682572, + "learning_rate": 8.926865374542928e-06, + "loss": 0.39300537109375, + "step": 115, + "token_acc": 0.8799238573189989 + }, + { + "epoch": 0.5093304061470911, + "grad_norm": 0.38568933953225865, + "learning_rate": 8.904305895794292e-06, + "loss": 0.406494140625, + "step": 116, + "token_acc": 0.8751395506651003 + }, + { + "epoch": 0.5137211855104281, + "grad_norm": 0.3330582874374485, + "learning_rate": 8.88154089146488e-06, + "loss": 0.377197265625, + "step": 117, + "token_acc": 0.8845054326929923 + }, + { + "epoch": 0.5181119648737651, + "grad_norm": 0.3470360235478156, + "learning_rate": 8.858571559921539e-06, + "loss": 0.38043212890625, + "step": 118, + "token_acc": 0.8828714799586802 + }, + { + "epoch": 0.5225027442371021, + "grad_norm": 0.3693022990438559, + "learning_rate": 8.835399110287046e-06, + "loss": 0.40362548828125, + "step": 119, + "token_acc": 0.8770892591496617 + }, + { + "epoch": 0.5268935236004391, + "grad_norm": 0.506814198712941, + "learning_rate": 8.812024762376477e-06, + "loss": 0.3729248046875, + "step": 120, + "token_acc": 0.8869120919881306 + }, + { + "epoch": 0.531284302963776, + "grad_norm": 0.3906138083115678, + "learning_rate": 8.788449746632976e-06, + "loss": 0.4075927734375, + "step": 121, + "token_acc": 0.8753834433837508 + }, + { + "epoch": 0.535675082327113, + "grad_norm": 0.3427861857538148, + "learning_rate": 8.764675304062992e-06, + "loss": 0.38037109375, + "step": 122, + "token_acc": 0.8822804682887565 + }, + { + "epoch": 0.54006586169045, + "grad_norm": 0.37005172163753364, + "learning_rate": 8.740702686170955e-06, + "loss": 0.353271484375, + "step": 123, + "token_acc": 0.8895579714862314 + }, + { + "epoch": 0.544456641053787, + "grad_norm": 0.3166454664099373, + "learning_rate": 8.71653315489339e-06, + "loss": 0.36798095703125, + "step": 124, + "token_acc": 0.8880577636981947 + }, + { + "epoch": 0.5488474204171241, + "grad_norm": 0.3473262076024377, + "learning_rate": 8.692167982532487e-06, + "loss": 0.3809814453125, + "step": 125, + "token_acc": 0.8847568204299533 + }, + { + "epoch": 0.5532381997804611, + "grad_norm": 0.356492202684863, + "learning_rate": 8.667608451689135e-06, + "loss": 0.4140625, + "step": 126, + "token_acc": 0.874559601744392 + }, + { + "epoch": 0.557628979143798, + "grad_norm": 0.3802001343376944, + "learning_rate": 8.642855855195394e-06, + "loss": 0.391845703125, + "step": 127, + "token_acc": 0.879835788975938 + }, + { + "epoch": 0.562019758507135, + "grad_norm": 0.33254820908020294, + "learning_rate": 8.617911496046446e-06, + "loss": 0.35955810546875, + "step": 128, + "token_acc": 0.888526902532168 + }, + { + "epoch": 0.566410537870472, + "grad_norm": 0.3752050004414951, + "learning_rate": 8.592776687332003e-06, + "loss": 0.4024658203125, + "step": 129, + "token_acc": 0.8762640719579115 + }, + { + "epoch": 0.570801317233809, + "grad_norm": 0.3432362851011395, + "learning_rate": 8.567452752167183e-06, + "loss": 0.357177734375, + "step": 130, + "token_acc": 0.8892355694227769 + }, + { + "epoch": 0.575192096597146, + "grad_norm": 0.33808726850202697, + "learning_rate": 8.54194102362286e-06, + "loss": 0.3665771484375, + "step": 131, + "token_acc": 0.8875122671648726 + }, + { + "epoch": 0.579582875960483, + "grad_norm": 0.37261912231511524, + "learning_rate": 8.516242844655498e-06, + "loss": 0.38726806640625, + "step": 132, + "token_acc": 0.8803020435714727 + }, + { + "epoch": 0.58397365532382, + "grad_norm": 0.33634527235461154, + "learning_rate": 8.490359568036446e-06, + "loss": 0.36956787109375, + "step": 133, + "token_acc": 0.8856262970882335 + }, + { + "epoch": 0.5883644346871569, + "grad_norm": 0.34468604715571666, + "learning_rate": 8.464292556280734e-06, + "loss": 0.38287353515625, + "step": 134, + "token_acc": 0.8824359742031745 + }, + { + "epoch": 0.5927552140504939, + "grad_norm": 0.3349650332247326, + "learning_rate": 8.43804318157534e-06, + "loss": 0.3526611328125, + "step": 135, + "token_acc": 0.8907932305506437 + }, + { + "epoch": 0.5971459934138309, + "grad_norm": 0.3415955731785513, + "learning_rate": 8.411612825706976e-06, + "loss": 0.375732421875, + "step": 136, + "token_acc": 0.8851655394080413 + }, + { + "epoch": 0.601536772777168, + "grad_norm": 0.399098480751626, + "learning_rate": 8.385002879989328e-06, + "loss": 0.408203125, + "step": 137, + "token_acc": 0.8738623786595984 + }, + { + "epoch": 0.605927552140505, + "grad_norm": 0.3649375419019348, + "learning_rate": 8.35821474518983e-06, + "loss": 0.39898681640625, + "step": 138, + "token_acc": 0.877580248777501 + }, + { + "epoch": 0.610318331503842, + "grad_norm": 0.3629655126758596, + "learning_rate": 8.331249831455921e-06, + "loss": 0.3973388671875, + "step": 139, + "token_acc": 0.8785421267889452 + }, + { + "epoch": 0.6147091108671789, + "grad_norm": 0.3616039889633607, + "learning_rate": 8.304109558240817e-06, + "loss": 0.37713623046875, + "step": 140, + "token_acc": 0.8842137320386112 + }, + { + "epoch": 0.6190998902305159, + "grad_norm": 0.37391368917180245, + "learning_rate": 8.276795354228785e-06, + "loss": 0.3895263671875, + "step": 141, + "token_acc": 0.8796410921731552 + }, + { + "epoch": 0.6234906695938529, + "grad_norm": 0.3311214436901146, + "learning_rate": 8.249308657259943e-06, + "loss": 0.367431640625, + "step": 142, + "token_acc": 0.887939618837817 + }, + { + "epoch": 0.6278814489571899, + "grad_norm": 0.324470329512847, + "learning_rate": 8.221650914254566e-06, + "loss": 0.38189697265625, + "step": 143, + "token_acc": 0.8833656311616217 + }, + { + "epoch": 0.6322722283205269, + "grad_norm": 0.3857696046293104, + "learning_rate": 8.193823581136919e-06, + "loss": 0.444580078125, + "step": 144, + "token_acc": 0.8655548088069931 + }, + { + "epoch": 0.6366630076838639, + "grad_norm": 0.4018240884485351, + "learning_rate": 8.165828122758615e-06, + "loss": 0.38043212890625, + "step": 145, + "token_acc": 0.8830743081657076 + }, + { + "epoch": 0.6410537870472008, + "grad_norm": 0.33372413990990984, + "learning_rate": 8.137666012821514e-06, + "loss": 0.38275146484375, + "step": 146, + "token_acc": 0.8836489499783672 + }, + { + "epoch": 0.6454445664105378, + "grad_norm": 0.3809702451433533, + "learning_rate": 8.109338733800132e-06, + "loss": 0.38555908203125, + "step": 147, + "token_acc": 0.8817689926423516 + }, + { + "epoch": 0.6498353457738749, + "grad_norm": 0.36947662510117957, + "learning_rate": 8.080847776863609e-06, + "loss": 0.4063720703125, + "step": 148, + "token_acc": 0.8752523769868135 + }, + { + "epoch": 0.6542261251372119, + "grad_norm": 0.3520775835686237, + "learning_rate": 8.052194641797217e-06, + "loss": 0.42059326171875, + "step": 149, + "token_acc": 0.8716511030084401 + }, + { + "epoch": 0.6586169045005489, + "grad_norm": 0.3627703170166138, + "learning_rate": 8.023380836923404e-06, + "loss": 0.38970947265625, + "step": 150, + "token_acc": 0.8809665907706646 + }, + { + "epoch": 0.6630076838638859, + "grad_norm": 0.37526919496711153, + "learning_rate": 7.994407879022397e-06, + "loss": 0.37677001953125, + "step": 151, + "token_acc": 0.8841504205924117 + }, + { + "epoch": 0.6673984632272228, + "grad_norm": 0.33686419489923275, + "learning_rate": 7.965277293252354e-06, + "loss": 0.3992919921875, + "step": 152, + "token_acc": 0.8784468693013606 + }, + { + "epoch": 0.6717892425905598, + "grad_norm": 0.36585972291630886, + "learning_rate": 7.935990613069087e-06, + "loss": 0.3931884765625, + "step": 153, + "token_acc": 0.8791216339989009 + }, + { + "epoch": 0.6761800219538968, + "grad_norm": 0.39031334799687256, + "learning_rate": 7.90654938014533e-06, + "loss": 0.3990478515625, + "step": 154, + "token_acc": 0.8776981358334242 + }, + { + "epoch": 0.6805708013172338, + "grad_norm": 0.32457845503536664, + "learning_rate": 7.876955144289594e-06, + "loss": 0.3712158203125, + "step": 155, + "token_acc": 0.8861263313983978 + }, + { + "epoch": 0.6849615806805708, + "grad_norm": 0.3345059774293838, + "learning_rate": 7.847209463364574e-06, + "loss": 0.3612060546875, + "step": 156, + "token_acc": 0.8889989642058881 + }, + { + "epoch": 0.6893523600439078, + "grad_norm": 0.39032458418186533, + "learning_rate": 7.817313903205148e-06, + "loss": 0.41949462890625, + "step": 157, + "token_acc": 0.8716605851843228 + }, + { + "epoch": 0.6937431394072447, + "grad_norm": 0.32720394654080537, + "learning_rate": 7.78727003753595e-06, + "loss": 0.3970947265625, + "step": 158, + "token_acc": 0.8792898355165782 + }, + { + "epoch": 0.6981339187705817, + "grad_norm": 0.4315930935523265, + "learning_rate": 7.757079447888529e-06, + "loss": 0.3895263671875, + "step": 159, + "token_acc": 0.8802383807186157 + }, + { + "epoch": 0.7025246981339188, + "grad_norm": 0.31181801076565785, + "learning_rate": 7.726743723518087e-06, + "loss": 0.3509521484375, + "step": 160, + "token_acc": 0.8916374731608091 + }, + { + "epoch": 0.7069154774972558, + "grad_norm": 0.3840632553730295, + "learning_rate": 7.696264461319831e-06, + "loss": 0.40155029296875, + "step": 161, + "token_acc": 0.8779857793578492 + }, + { + "epoch": 0.7113062568605928, + "grad_norm": 0.37983904315138733, + "learning_rate": 7.66564326574491e-06, + "loss": 0.380859375, + "step": 162, + "token_acc": 0.8814475623971476 + }, + { + "epoch": 0.7156970362239298, + "grad_norm": 0.3619797805093762, + "learning_rate": 7.634881748715941e-06, + "loss": 0.3818359375, + "step": 163, + "token_acc": 0.88148122448006 + }, + { + "epoch": 0.7200878155872668, + "grad_norm": 0.31621965896919957, + "learning_rate": 7.60398152954218e-06, + "loss": 0.3565673828125, + "step": 164, + "token_acc": 0.8898631743479333 + }, + { + "epoch": 0.7244785949506037, + "grad_norm": 0.4585628653480396, + "learning_rate": 7.572944234834261e-06, + "loss": 0.3770751953125, + "step": 165, + "token_acc": 0.8835862390799688 + }, + { + "epoch": 0.7288693743139407, + "grad_norm": 0.3415613865633226, + "learning_rate": 7.541771498418575e-06, + "loss": 0.3719482421875, + "step": 166, + "token_acc": 0.8849516768544523 + }, + { + "epoch": 0.7332601536772777, + "grad_norm": 0.35619766971258054, + "learning_rate": 7.510464961251271e-06, + "loss": 0.3702392578125, + "step": 167, + "token_acc": 0.8849932987100017 + }, + { + "epoch": 0.7376509330406147, + "grad_norm": 0.32400495581439437, + "learning_rate": 7.479026271331864e-06, + "loss": 0.354736328125, + "step": 168, + "token_acc": 0.8893855026159803 + }, + { + "epoch": 0.7420417124039517, + "grad_norm": 0.36420285487752785, + "learning_rate": 7.447457083616494e-06, + "loss": 0.37969970703125, + "step": 169, + "token_acc": 0.8824101587510055 + }, + { + "epoch": 0.7464324917672887, + "grad_norm": 0.3790739453786472, + "learning_rate": 7.415759059930799e-06, + "loss": 0.38427734375, + "step": 170, + "token_acc": 0.8806128850210843 + }, + { + "epoch": 0.7508232711306256, + "grad_norm": 0.342673058015813, + "learning_rate": 7.383933868882438e-06, + "loss": 0.38458251953125, + "step": 171, + "token_acc": 0.8827557781934625 + }, + { + "epoch": 0.7552140504939627, + "grad_norm": 0.4376034436917223, + "learning_rate": 7.351983185773259e-06, + "loss": 0.38787841796875, + "step": 172, + "token_acc": 0.8803101856706032 + }, + { + "epoch": 0.7596048298572997, + "grad_norm": 0.38074486562009263, + "learning_rate": 7.319908692511103e-06, + "loss": 0.3804931640625, + "step": 173, + "token_acc": 0.881339015205142 + }, + { + "epoch": 0.7639956092206367, + "grad_norm": 0.3627656511721131, + "learning_rate": 7.2877120775212685e-06, + "loss": 0.3790283203125, + "step": 174, + "token_acc": 0.8826412805931554 + }, + { + "epoch": 0.7683863885839737, + "grad_norm": 0.32526897125202175, + "learning_rate": 7.255395035657639e-06, + "loss": 0.40692138671875, + "step": 175, + "token_acc": 0.8751592079337259 + }, + { + "epoch": 0.7727771679473107, + "grad_norm": 0.3531493263881796, + "learning_rate": 7.222959268113452e-06, + "loss": 0.37255859375, + "step": 176, + "token_acc": 0.8840336064080796 + }, + { + "epoch": 0.7771679473106476, + "grad_norm": 0.34388743318528936, + "learning_rate": 7.190406482331757e-06, + "loss": 0.35540771484375, + "step": 177, + "token_acc": 0.8885860314175941 + }, + { + "epoch": 0.7815587266739846, + "grad_norm": 0.3400578760073584, + "learning_rate": 7.157738391915531e-06, + "loss": 0.3719482421875, + "step": 178, + "token_acc": 0.8851536340828222 + }, + { + "epoch": 0.7859495060373216, + "grad_norm": 0.3065770409213107, + "learning_rate": 7.124956716537471e-06, + "loss": 0.34527587890625, + "step": 179, + "token_acc": 0.8942501657520913 + }, + { + "epoch": 0.7903402854006586, + "grad_norm": 0.31872574714144547, + "learning_rate": 7.0920631818494745e-06, + "loss": 0.38250732421875, + "step": 180, + "token_acc": 0.8825593696421405 + }, + { + "epoch": 0.7947310647639956, + "grad_norm": 0.3530750555537168, + "learning_rate": 7.059059519391794e-06, + "loss": 0.39556884765625, + "step": 181, + "token_acc": 0.8784444327436114 + }, + { + "epoch": 0.7991218441273326, + "grad_norm": 0.36873765684022797, + "learning_rate": 7.0259474665018915e-06, + "loss": 0.39208984375, + "step": 182, + "token_acc": 0.8789131915403404 + }, + { + "epoch": 0.8035126234906695, + "grad_norm": 0.32458737719879455, + "learning_rate": 6.992728766222982e-06, + "loss": 0.37786865234375, + "step": 183, + "token_acc": 0.8829038830639858 + }, + { + "epoch": 0.8079034028540066, + "grad_norm": 0.3523657409863041, + "learning_rate": 6.959405167212278e-06, + "loss": 0.39154052734375, + "step": 184, + "token_acc": 0.878811179987066 + }, + { + "epoch": 0.8122941822173436, + "grad_norm": 0.3399194905692937, + "learning_rate": 6.925978423648941e-06, + "loss": 0.3775634765625, + "step": 185, + "token_acc": 0.8836490787751949 + }, + { + "epoch": 0.8166849615806806, + "grad_norm": 0.3202766027815532, + "learning_rate": 6.892450295141737e-06, + "loss": 0.3599853515625, + "step": 186, + "token_acc": 0.8901854169926459 + }, + { + "epoch": 0.8210757409440176, + "grad_norm": 0.31947441341487925, + "learning_rate": 6.858822546636417e-06, + "loss": 0.38568115234375, + "step": 187, + "token_acc": 0.8821333233773264 + }, + { + "epoch": 0.8254665203073546, + "grad_norm": 0.34132217767891093, + "learning_rate": 6.825096948322791e-06, + "loss": 0.370361328125, + "step": 188, + "token_acc": 0.8857913419030761 + }, + { + "epoch": 0.8298572996706916, + "grad_norm": 0.3678327095868294, + "learning_rate": 6.7912752755415716e-06, + "loss": 0.42767333984375, + "step": 189, + "token_acc": 0.869923790794662 + }, + { + "epoch": 0.8342480790340285, + "grad_norm": 0.3179637335071831, + "learning_rate": 6.757359308690889e-06, + "loss": 0.37274169921875, + "step": 190, + "token_acc": 0.8865945912044817 + }, + { + "epoch": 0.8386388583973655, + "grad_norm": 0.31582344943005664, + "learning_rate": 6.723350833132596e-06, + "loss": 0.3643798828125, + "step": 191, + "token_acc": 0.8865578127966585 + }, + { + "epoch": 0.8430296377607025, + "grad_norm": 0.3474116956399158, + "learning_rate": 6.689251639098261e-06, + "loss": 0.37677001953125, + "step": 192, + "token_acc": 0.8840790683420165 + }, + { + "epoch": 0.8474204171240395, + "grad_norm": 0.3575274878096281, + "learning_rate": 6.65506352159495e-06, + "loss": 0.36041259765625, + "step": 193, + "token_acc": 0.8874090708202439 + }, + { + "epoch": 0.8518111964873765, + "grad_norm": 0.340671631772466, + "learning_rate": 6.620788280310722e-06, + "loss": 0.361572265625, + "step": 194, + "token_acc": 0.887554323438835 + }, + { + "epoch": 0.8562019758507134, + "grad_norm": 0.36093128612228575, + "learning_rate": 6.586427719519901e-06, + "loss": 0.4222412109375, + "step": 195, + "token_acc": 0.8723851983602273 + }, + { + "epoch": 0.8605927552140505, + "grad_norm": 0.36238446930397766, + "learning_rate": 6.551983647988089e-06, + "loss": 0.39404296875, + "step": 196, + "token_acc": 0.8777444583752941 + }, + { + "epoch": 0.8649835345773875, + "grad_norm": 0.3353170393091822, + "learning_rate": 6.517457878876958e-06, + "loss": 0.36285400390625, + "step": 197, + "token_acc": 0.8872218963831867 + }, + { + "epoch": 0.8693743139407245, + "grad_norm": 0.3426656529570822, + "learning_rate": 6.4828522296488014e-06, + "loss": 0.373779296875, + "step": 198, + "token_acc": 0.8858460530573277 + }, + { + "epoch": 0.8737650933040615, + "grad_norm": 0.31142201108082124, + "learning_rate": 6.448168521970865e-06, + "loss": 0.35833740234375, + "step": 199, + "token_acc": 0.8897007572084821 + }, + { + "epoch": 0.8781558726673985, + "grad_norm": 0.3415603924832066, + "learning_rate": 6.41340858161944e-06, + "loss": 0.36865234375, + "step": 200, + "token_acc": 0.886851209736584 + }, + { + "epoch": 0.8825466520307355, + "grad_norm": 0.3602063170833425, + "learning_rate": 6.378574238383776e-06, + "loss": 0.3909912109375, + "step": 201, + "token_acc": 0.8777218002087004 + }, + { + "epoch": 0.8869374313940724, + "grad_norm": 0.3280058188462181, + "learning_rate": 6.343667325969736e-06, + "loss": 0.353271484375, + "step": 202, + "token_acc": 0.8911195594728677 + }, + { + "epoch": 0.8913282107574094, + "grad_norm": 0.38274409915348256, + "learning_rate": 6.3086896819032814e-06, + "loss": 0.4415283203125, + "step": 203, + "token_acc": 0.8650215632279404 + }, + { + "epoch": 0.8957189901207464, + "grad_norm": 0.33725833063346167, + "learning_rate": 6.273643147433743e-06, + "loss": 0.37359619140625, + "step": 204, + "token_acc": 0.8841945374251076 + }, + { + "epoch": 0.9001097694840834, + "grad_norm": 0.31954237097775756, + "learning_rate": 6.238529567436892e-06, + "loss": 0.368408203125, + "step": 205, + "token_acc": 0.8868830927805617 + }, + { + "epoch": 0.9045005488474204, + "grad_norm": 0.3535581136616006, + "learning_rate": 6.203350790317825e-06, + "loss": 0.3941650390625, + "step": 206, + "token_acc": 0.8795915644722664 + }, + { + "epoch": 0.9088913282107574, + "grad_norm": 0.3789748553818223, + "learning_rate": 6.168108667913666e-06, + "loss": 0.375244140625, + "step": 207, + "token_acc": 0.8825434868088453 + }, + { + "epoch": 0.9132821075740944, + "grad_norm": 0.340578316793545, + "learning_rate": 6.1328050553960804e-06, + "loss": 0.3594970703125, + "step": 208, + "token_acc": 0.8872000756527196 + }, + { + "epoch": 0.9176728869374314, + "grad_norm": 0.37590204046096487, + "learning_rate": 6.0974418111736235e-06, + "loss": 0.364501953125, + "step": 209, + "token_acc": 0.8857336130262178 + }, + { + "epoch": 0.9220636663007684, + "grad_norm": 0.3265257835179892, + "learning_rate": 6.0620207967939e-06, + "loss": 0.36517333984375, + "step": 210, + "token_acc": 0.8864980013426096 + }, + { + "epoch": 0.9264544456641054, + "grad_norm": 0.3247028591880143, + "learning_rate": 6.026543876845586e-06, + "loss": 0.36480712890625, + "step": 211, + "token_acc": 0.8870534119854157 + }, + { + "epoch": 0.9308452250274424, + "grad_norm": 0.3696471996964635, + "learning_rate": 5.9910129188602665e-06, + "loss": 0.367919921875, + "step": 212, + "token_acc": 0.8857886233166047 + }, + { + "epoch": 0.9352360043907794, + "grad_norm": 0.3432343371525426, + "learning_rate": 5.955429793214129e-06, + "loss": 0.38818359375, + "step": 213, + "token_acc": 0.8796593213691741 + }, + { + "epoch": 0.9396267837541163, + "grad_norm": 0.3298648448953858, + "learning_rate": 5.919796373029504e-06, + "loss": 0.37548828125, + "step": 214, + "token_acc": 0.8848114804720147 + }, + { + "epoch": 0.9440175631174533, + "grad_norm": 0.3388738734674106, + "learning_rate": 5.8841145340762665e-06, + "loss": 0.3521728515625, + "step": 215, + "token_acc": 0.8905765179455674 + }, + { + "epoch": 0.9484083424807903, + "grad_norm": 0.34211047050118515, + "learning_rate": 5.8483861546730915e-06, + "loss": 0.3751220703125, + "step": 216, + "token_acc": 0.8849980537815334 + }, + { + "epoch": 0.9527991218441273, + "grad_norm": 0.31547049057865456, + "learning_rate": 5.812613115588575e-06, + "loss": 0.3609619140625, + "step": 217, + "token_acc": 0.8887366033621789 + }, + { + "epoch": 0.9571899012074643, + "grad_norm": 0.38549813725977156, + "learning_rate": 5.776797299942236e-06, + "loss": 0.387451171875, + "step": 218, + "token_acc": 0.8790944373213566 + }, + { + "epoch": 0.9615806805708014, + "grad_norm": 0.3224837065818038, + "learning_rate": 5.740940593105383e-06, + "loss": 0.37103271484375, + "step": 219, + "token_acc": 0.8852090858139245 + }, + { + "epoch": 0.9659714599341384, + "grad_norm": 0.37142561967800986, + "learning_rate": 5.705044882601862e-06, + "loss": 0.382568359375, + "step": 220, + "token_acc": 0.8823479648623807 + }, + { + "epoch": 0.9703622392974753, + "grad_norm": 0.4233998401399281, + "learning_rate": 5.6691120580087126e-06, + "loss": 0.3818359375, + "step": 221, + "token_acc": 0.8819167916429642 + }, + { + "epoch": 0.9747530186608123, + "grad_norm": 0.3049406743219966, + "learning_rate": 5.6331440108566735e-06, + "loss": 0.36798095703125, + "step": 222, + "token_acc": 0.88659642693541 + }, + { + "epoch": 0.9791437980241493, + "grad_norm": 0.3603156529875038, + "learning_rate": 5.597142634530639e-06, + "loss": 0.4005126953125, + "step": 223, + "token_acc": 0.8761579179532422 + }, + { + "epoch": 0.9835345773874863, + "grad_norm": 0.3473380674907868, + "learning_rate": 5.561109824169962e-06, + "loss": 0.3885498046875, + "step": 224, + "token_acc": 0.8802932363013699 + }, + { + "epoch": 0.9879253567508233, + "grad_norm": 0.3455081187919915, + "learning_rate": 5.525047476568722e-06, + "loss": 0.38543701171875, + "step": 225, + "token_acc": 0.8810951291558197 + }, + { + "epoch": 0.9923161361141603, + "grad_norm": 0.32814174469662283, + "learning_rate": 5.488957490075846e-06, + "loss": 0.3759765625, + "step": 226, + "token_acc": 0.8845346517812964 + }, + { + "epoch": 0.9967069154774972, + "grad_norm": 0.42259044420354114, + "learning_rate": 5.452841764495203e-06, + "loss": 0.36700439453125, + "step": 227, + "token_acc": 0.8856456066407621 + }, + { + "epoch": 1.0, + "grad_norm": 0.3420661875520739, + "learning_rate": 5.416702200985585e-06, + "loss": 0.34716796875, + "step": 228, + "token_acc": 0.893556253707406 + }, + { + "epoch": 1.004390779363337, + "grad_norm": 0.3526762978167876, + "learning_rate": 5.380540701960627e-06, + "loss": 0.4190673828125, + "step": 229, + "token_acc": 0.8712556290100469 + }, + { + "epoch": 1.008781558726674, + "grad_norm": 0.3185399320403508, + "learning_rate": 5.344359170988668e-06, + "loss": 0.390625, + "step": 230, + "token_acc": 0.8797135790383168 + }, + { + "epoch": 1.013172338090011, + "grad_norm": 0.33869348744361016, + "learning_rate": 5.308159512692544e-06, + "loss": 0.36761474609375, + "step": 231, + "token_acc": 0.8865497449871445 + }, + { + "epoch": 1.017563117453348, + "grad_norm": 0.3665451005015509, + "learning_rate": 5.2719436326493255e-06, + "loss": 0.39532470703125, + "step": 232, + "token_acc": 0.8776601641663304 + }, + { + "epoch": 1.021953896816685, + "grad_norm": 0.3353621295614313, + "learning_rate": 5.235713437290012e-06, + "loss": 0.348876953125, + "step": 233, + "token_acc": 0.8918691769231061 + }, + { + "epoch": 1.026344676180022, + "grad_norm": 0.33722465426571885, + "learning_rate": 5.199470833799164e-06, + "loss": 0.4034423828125, + "step": 234, + "token_acc": 0.8756343022964155 + }, + { + "epoch": 1.030735455543359, + "grad_norm": 0.3388353883798304, + "learning_rate": 5.1632177300145255e-06, + "loss": 0.35125732421875, + "step": 235, + "token_acc": 0.8902197267134511 + }, + { + "epoch": 1.0351262349066959, + "grad_norm": 0.3345585217471032, + "learning_rate": 5.126956034326573e-06, + "loss": 0.365478515625, + "step": 236, + "token_acc": 0.8860912143547733 + }, + { + "epoch": 1.039517014270033, + "grad_norm": 0.33922284859129886, + "learning_rate": 5.090687655578078e-06, + "loss": 0.37237548828125, + "step": 237, + "token_acc": 0.8849442628822134 + }, + { + "epoch": 1.0439077936333698, + "grad_norm": 0.3278620545376965, + "learning_rate": 5.054414502963605e-06, + "loss": 0.32977294921875, + "step": 238, + "token_acc": 0.8974845872173027 + }, + { + "epoch": 1.048298572996707, + "grad_norm": 0.3164606607512339, + "learning_rate": 5.0181384859290215e-06, + "loss": 0.35595703125, + "step": 239, + "token_acc": 0.8891893450077274 + }, + { + "epoch": 1.0526893523600438, + "grad_norm": 0.3326063982801495, + "learning_rate": 4.981861514070979e-06, + "loss": 0.3563232421875, + "step": 240, + "token_acc": 0.8868982591268126 + }, + { + "epoch": 1.057080131723381, + "grad_norm": 0.35606349945181254, + "learning_rate": 4.945585497036396e-06, + "loss": 0.3707275390625, + "step": 241, + "token_acc": 0.8838458201174759 + }, + { + "epoch": 1.061470911086718, + "grad_norm": 0.32716162695624673, + "learning_rate": 4.909312344421923e-06, + "loss": 0.345947265625, + "step": 242, + "token_acc": 0.8912522785026517 + }, + { + "epoch": 1.0658616904500549, + "grad_norm": 0.32495072680974807, + "learning_rate": 4.873043965673427e-06, + "loss": 0.35052490234375, + "step": 243, + "token_acc": 0.8896871325831031 + }, + { + "epoch": 1.070252469813392, + "grad_norm": 0.35135879994609154, + "learning_rate": 4.836782269985475e-06, + "loss": 0.3651123046875, + "step": 244, + "token_acc": 0.8849511835891829 + }, + { + "epoch": 1.0746432491767288, + "grad_norm": 0.299533289637608, + "learning_rate": 4.800529166200837e-06, + "loss": 0.3544921875, + "step": 245, + "token_acc": 0.8907510908376973 + }, + { + "epoch": 1.079034028540066, + "grad_norm": 0.34907033095336737, + "learning_rate": 4.76428656270999e-06, + "loss": 0.3638916015625, + "step": 246, + "token_acc": 0.8850038463278881 + }, + { + "epoch": 1.0834248079034028, + "grad_norm": 0.3437591237664625, + "learning_rate": 4.7280563673506745e-06, + "loss": 0.35601806640625, + "step": 247, + "token_acc": 0.887521505962059 + }, + { + "epoch": 1.0878155872667399, + "grad_norm": 0.344376045357089, + "learning_rate": 4.6918404873074574e-06, + "loss": 0.348175048828125, + "step": 248, + "token_acc": 0.8899573544734835 + }, + { + "epoch": 1.0922063666300768, + "grad_norm": 0.38757166642463897, + "learning_rate": 4.655640829011335e-06, + "loss": 0.3636474609375, + "step": 249, + "token_acc": 0.8847563630952511 + }, + { + "epoch": 1.0965971459934138, + "grad_norm": 0.33891244441691515, + "learning_rate": 4.619459298039373e-06, + "loss": 0.3475341796875, + "step": 250, + "token_acc": 0.8917179199306426 + }, + { + "epoch": 1.1009879253567507, + "grad_norm": 0.32303379197467086, + "learning_rate": 4.5832977990144165e-06, + "loss": 0.33807373046875, + "step": 251, + "token_acc": 0.8926653345856346 + }, + { + "epoch": 1.1053787047200878, + "grad_norm": 0.3361295073502542, + "learning_rate": 4.547158235504797e-06, + "loss": 0.33905029296875, + "step": 252, + "token_acc": 0.8929544668822321 + }, + { + "epoch": 1.109769484083425, + "grad_norm": 0.3337900379095731, + "learning_rate": 4.511042509924157e-06, + "loss": 0.34722900390625, + "step": 253, + "token_acc": 0.8913863807887104 + }, + { + "epoch": 1.1141602634467618, + "grad_norm": 0.3212794640436032, + "learning_rate": 4.47495252343128e-06, + "loss": 0.32049560546875, + "step": 254, + "token_acc": 0.8989152764532649 + }, + { + "epoch": 1.1185510428100989, + "grad_norm": 0.353961566610381, + "learning_rate": 4.438890175830039e-06, + "loss": 0.34942626953125, + "step": 255, + "token_acc": 0.8902083607807441 + }, + { + "epoch": 1.1229418221734357, + "grad_norm": 0.3356793152337837, + "learning_rate": 4.402857365469364e-06, + "loss": 0.34796142578125, + "step": 256, + "token_acc": 0.8908204439745496 + }, + { + "epoch": 1.1273326015367728, + "grad_norm": 0.3046809319102677, + "learning_rate": 4.366855989143326e-06, + "loss": 0.3170166015625, + "step": 257, + "token_acc": 0.8992992178508858 + }, + { + "epoch": 1.1317233809001097, + "grad_norm": 0.3954255039255281, + "learning_rate": 4.330887941991288e-06, + "loss": 0.3665771484375, + "step": 258, + "token_acc": 0.8842385160708647 + }, + { + "epoch": 1.1361141602634468, + "grad_norm": 0.39739742832544567, + "learning_rate": 4.294955117398139e-06, + "loss": 0.36370849609375, + "step": 259, + "token_acc": 0.8829140826597989 + }, + { + "epoch": 1.1405049396267837, + "grad_norm": 0.368752927536437, + "learning_rate": 4.259059406894619e-06, + "loss": 0.3394775390625, + "step": 260, + "token_acc": 0.8913581389965859 + }, + { + "epoch": 1.1448957189901208, + "grad_norm": 0.3401168495318414, + "learning_rate": 4.223202700057765e-06, + "loss": 0.32916259765625, + "step": 261, + "token_acc": 0.894392052408034 + }, + { + "epoch": 1.1492864983534576, + "grad_norm": 0.37526207167560366, + "learning_rate": 4.187386884411426e-06, + "loss": 0.342529296875, + "step": 262, + "token_acc": 0.8909135039324492 + }, + { + "epoch": 1.1536772777167947, + "grad_norm": 0.35889769900995294, + "learning_rate": 4.151613845326912e-06, + "loss": 0.322998046875, + "step": 263, + "token_acc": 0.8951741891732121 + }, + { + "epoch": 1.1580680570801318, + "grad_norm": 0.32113645997062235, + "learning_rate": 4.115885465923734e-06, + "loss": 0.32476806640625, + "step": 264, + "token_acc": 0.8972362963864571 + }, + { + "epoch": 1.1624588364434687, + "grad_norm": 0.3395360059813307, + "learning_rate": 4.080203626970498e-06, + "loss": 0.35357666015625, + "step": 265, + "token_acc": 0.8887321473803168 + }, + { + "epoch": 1.1668496158068058, + "grad_norm": 0.32715139341297833, + "learning_rate": 4.044570206785874e-06, + "loss": 0.34271240234375, + "step": 266, + "token_acc": 0.8909816337724649 + }, + { + "epoch": 1.1712403951701427, + "grad_norm": 0.3524832743836693, + "learning_rate": 4.008987081139734e-06, + "loss": 0.321533203125, + "step": 267, + "token_acc": 0.8969505309220949 + }, + { + "epoch": 1.1756311745334798, + "grad_norm": 0.34764249107265743, + "learning_rate": 3.973456123154415e-06, + "loss": 0.32476806640625, + "step": 268, + "token_acc": 0.8958763258207247 + }, + { + "epoch": 1.1800219538968166, + "grad_norm": 0.356152162638529, + "learning_rate": 3.937979203206103e-06, + "loss": 0.33734130859375, + "step": 269, + "token_acc": 0.8916322440415142 + }, + { + "epoch": 1.1844127332601537, + "grad_norm": 0.3291475777565464, + "learning_rate": 3.90255818882638e-06, + "loss": 0.34295654296875, + "step": 270, + "token_acc": 0.8912276254554348 + }, + { + "epoch": 1.1888035126234906, + "grad_norm": 0.34742649793974, + "learning_rate": 3.86719494460392e-06, + "loss": 0.32952880859375, + "step": 271, + "token_acc": 0.8949748215308042 + }, + { + "epoch": 1.1931942919868277, + "grad_norm": 0.344795628023517, + "learning_rate": 3.8318913320863355e-06, + "loss": 0.3546142578125, + "step": 272, + "token_acc": 0.8867931419388564 + }, + { + "epoch": 1.1975850713501646, + "grad_norm": 0.40362302120251126, + "learning_rate": 3.7966492096821773e-06, + "loss": 0.3326416015625, + "step": 273, + "token_acc": 0.8935762587874246 + }, + { + "epoch": 1.2019758507135017, + "grad_norm": 0.3784151436195393, + "learning_rate": 3.761470432563109e-06, + "loss": 0.354705810546875, + "step": 274, + "token_acc": 0.8888853291755882 + }, + { + "epoch": 1.2063666300768388, + "grad_norm": 0.3348345571823369, + "learning_rate": 3.7263568525662574e-06, + "loss": 0.33758544921875, + "step": 275, + "token_acc": 0.8925461168561734 + }, + { + "epoch": 1.2107574094401756, + "grad_norm": 0.3602090169898457, + "learning_rate": 3.691310318096719e-06, + "loss": 0.34918212890625, + "step": 276, + "token_acc": 0.8888169182509835 + }, + { + "epoch": 1.2151481888035127, + "grad_norm": 0.335730786585322, + "learning_rate": 3.6563326740302664e-06, + "loss": 0.314697265625, + "step": 277, + "token_acc": 0.9000328000291555 + }, + { + "epoch": 1.2195389681668496, + "grad_norm": 0.34253616383542923, + "learning_rate": 3.621425761616224e-06, + "loss": 0.35162353515625, + "step": 278, + "token_acc": 0.8880379277829491 + }, + { + "epoch": 1.2239297475301867, + "grad_norm": 0.34485406703569277, + "learning_rate": 3.5865914183805606e-06, + "loss": 0.32269287109375, + "step": 279, + "token_acc": 0.8974534551960927 + }, + { + "epoch": 1.2283205268935236, + "grad_norm": 0.35371293367929946, + "learning_rate": 3.5518314780291384e-06, + "loss": 0.3890380859375, + "step": 280, + "token_acc": 0.8783802026560067 + }, + { + "epoch": 1.2327113062568607, + "grad_norm": 0.31585946387965186, + "learning_rate": 3.517147770351199e-06, + "loss": 0.31927490234375, + "step": 281, + "token_acc": 0.8972110764353297 + }, + { + "epoch": 1.2371020856201975, + "grad_norm": 0.3764775136478381, + "learning_rate": 3.4825421211230437e-06, + "loss": 0.39239501953125, + "step": 282, + "token_acc": 0.8806873877296413 + }, + { + "epoch": 1.2414928649835346, + "grad_norm": 0.3460580716590409, + "learning_rate": 3.448016352011914e-06, + "loss": 0.3333740234375, + "step": 283, + "token_acc": 0.8943163018882275 + }, + { + "epoch": 1.2458836443468715, + "grad_norm": 0.3302255596745639, + "learning_rate": 3.4135722804801004e-06, + "loss": 0.30987548828125, + "step": 284, + "token_acc": 0.8998882709670116 + }, + { + "epoch": 1.2502744237102086, + "grad_norm": 0.3377273476413208, + "learning_rate": 3.379211719689278e-06, + "loss": 0.327880859375, + "step": 285, + "token_acc": 0.895182894797073 + }, + { + "epoch": 1.2546652030735457, + "grad_norm": 0.3473400775282267, + "learning_rate": 3.3449364784050515e-06, + "loss": 0.34979248046875, + "step": 286, + "token_acc": 0.8884178603164768 + }, + { + "epoch": 1.2590559824368825, + "grad_norm": 0.3612953797753178, + "learning_rate": 3.310748360901741e-06, + "loss": 0.34393310546875, + "step": 287, + "token_acc": 0.8910696916426456 + }, + { + "epoch": 1.2634467618002194, + "grad_norm": 0.34667855549309917, + "learning_rate": 3.276649166867406e-06, + "loss": 0.3321533203125, + "step": 288, + "token_acc": 0.8937538364352589 + }, + { + "epoch": 1.2678375411635565, + "grad_norm": 0.309905575308465, + "learning_rate": 3.242640691309111e-06, + "loss": 0.32720947265625, + "step": 289, + "token_acc": 0.8956456166809884 + }, + { + "epoch": 1.2722283205268936, + "grad_norm": 0.35771643178607937, + "learning_rate": 3.20872472445843e-06, + "loss": 0.329345703125, + "step": 290, + "token_acc": 0.8946723833625464 + }, + { + "epoch": 1.2766190998902305, + "grad_norm": 0.33183470119337916, + "learning_rate": 3.1749030516772084e-06, + "loss": 0.327392578125, + "step": 291, + "token_acc": 0.8957716205854872 + }, + { + "epoch": 1.2810098792535676, + "grad_norm": 0.31382638222408177, + "learning_rate": 3.1411774533635854e-06, + "loss": 0.33636474609375, + "step": 292, + "token_acc": 0.8938404877505446 + }, + { + "epoch": 1.2854006586169044, + "grad_norm": 0.3622219963442776, + "learning_rate": 3.1075497048582635e-06, + "loss": 0.327880859375, + "step": 293, + "token_acc": 0.8939353870425999 + }, + { + "epoch": 1.2897914379802415, + "grad_norm": 0.3152928659418184, + "learning_rate": 3.0740215763510617e-06, + "loss": 0.3114013671875, + "step": 294, + "token_acc": 0.9000977300513973 + }, + { + "epoch": 1.2941822173435784, + "grad_norm": 0.3553271178575454, + "learning_rate": 3.0405948327877233e-06, + "loss": 0.33306884765625, + "step": 295, + "token_acc": 0.8928606806127135 + }, + { + "epoch": 1.2985729967069155, + "grad_norm": 0.317297999441236, + "learning_rate": 3.00727123377702e-06, + "loss": 0.31988525390625, + "step": 296, + "token_acc": 0.8976165673732301 + }, + { + "epoch": 1.3029637760702526, + "grad_norm": 0.3510073260163728, + "learning_rate": 2.9740525334981105e-06, + "loss": 0.325439453125, + "step": 297, + "token_acc": 0.8945735629564617 + }, + { + "epoch": 1.3073545554335895, + "grad_norm": 0.34502937409150547, + "learning_rate": 2.9409404806082077e-06, + "loss": 0.338623046875, + "step": 298, + "token_acc": 0.8913558639153999 + }, + { + "epoch": 1.3117453347969263, + "grad_norm": 0.3444307091901023, + "learning_rate": 2.9079368181505263e-06, + "loss": 0.334228515625, + "step": 299, + "token_acc": 0.894660621432837 + }, + { + "epoch": 1.3161361141602634, + "grad_norm": 0.36685070270301506, + "learning_rate": 2.8750432834625312e-06, + "loss": 0.3323974609375, + "step": 300, + "token_acc": 0.8939670932358318 + }, + { + "epoch": 1.3205268935236005, + "grad_norm": 0.34665711527427445, + "learning_rate": 2.84226160808447e-06, + "loss": 0.34234619140625, + "step": 301, + "token_acc": 0.8911080630774668 + }, + { + "epoch": 1.3249176728869374, + "grad_norm": 0.32737189500283015, + "learning_rate": 2.809593517668243e-06, + "loss": 0.328125, + "step": 302, + "token_acc": 0.8956877294282094 + }, + { + "epoch": 1.3293084522502745, + "grad_norm": 0.34393396439779755, + "learning_rate": 2.777040731886549e-06, + "loss": 0.3206787109375, + "step": 303, + "token_acc": 0.8972659318997501 + }, + { + "epoch": 1.3336992316136114, + "grad_norm": 0.33269075563606537, + "learning_rate": 2.744604964342364e-06, + "loss": 0.32196044921875, + "step": 304, + "token_acc": 0.8957547445988107 + }, + { + "epoch": 1.3380900109769485, + "grad_norm": 0.3262552091418682, + "learning_rate": 2.7122879224787315e-06, + "loss": 0.32373046875, + "step": 305, + "token_acc": 0.8973069482625707 + }, + { + "epoch": 1.3424807903402853, + "grad_norm": 0.33544708546828367, + "learning_rate": 2.6800913074888984e-06, + "loss": 0.32244873046875, + "step": 306, + "token_acc": 0.8970129185963313 + }, + { + "epoch": 1.3468715697036224, + "grad_norm": 0.3068061592526234, + "learning_rate": 2.648016814226742e-06, + "loss": 0.3287353515625, + "step": 307, + "token_acc": 0.8953868973336852 + }, + { + "epoch": 1.3512623490669595, + "grad_norm": 0.35060578968767275, + "learning_rate": 2.616066131117563e-06, + "loss": 0.3338623046875, + "step": 308, + "token_acc": 0.892730062690899 + }, + { + "epoch": 1.3556531284302964, + "grad_norm": 0.3371212700359974, + "learning_rate": 2.5842409400692026e-06, + "loss": 0.3309326171875, + "step": 309, + "token_acc": 0.8943601262615738 + }, + { + "epoch": 1.3600439077936333, + "grad_norm": 0.3230161963407697, + "learning_rate": 2.552542916383507e-06, + "loss": 0.33050537109375, + "step": 310, + "token_acc": 0.8945870650487071 + }, + { + "epoch": 1.3644346871569704, + "grad_norm": 0.3881520121988701, + "learning_rate": 2.5209737286681367e-06, + "loss": 0.336669921875, + "step": 311, + "token_acc": 0.8914918928032822 + }, + { + "epoch": 1.3688254665203075, + "grad_norm": 0.3274352369059981, + "learning_rate": 2.4895350387487304e-06, + "loss": 0.3272705078125, + "step": 312, + "token_acc": 0.8948985202361884 + }, + { + "epoch": 1.3732162458836443, + "grad_norm": 0.32912953601491696, + "learning_rate": 2.4582285015814263e-06, + "loss": 0.32586669921875, + "step": 313, + "token_acc": 0.896020144335993 + }, + { + "epoch": 1.3776070252469814, + "grad_norm": 0.37903034721157913, + "learning_rate": 2.427055765165741e-06, + "loss": 0.35980224609375, + "step": 314, + "token_acc": 0.8853262030509947 + }, + { + "epoch": 1.3819978046103183, + "grad_norm": 0.3447441431371463, + "learning_rate": 2.396018470457821e-06, + "loss": 0.31353759765625, + "step": 315, + "token_acc": 0.8986164043264642 + }, + { + "epoch": 1.3863885839736554, + "grad_norm": 0.34906602556480076, + "learning_rate": 2.3651182512840604e-06, + "loss": 0.33709716796875, + "step": 316, + "token_acc": 0.8933294387468086 + }, + { + "epoch": 1.3907793633369923, + "grad_norm": 0.3278105252006223, + "learning_rate": 2.3343567342550933e-06, + "loss": 0.32659912109375, + "step": 317, + "token_acc": 0.8956145845098623 + }, + { + "epoch": 1.3951701427003294, + "grad_norm": 0.32630280385837857, + "learning_rate": 2.3037355386801683e-06, + "loss": 0.33660888671875, + "step": 318, + "token_acc": 0.8941868374861218 + }, + { + "epoch": 1.3995609220636662, + "grad_norm": 0.39149392629647267, + "learning_rate": 2.2732562764819157e-06, + "loss": 0.34185791015625, + "step": 319, + "token_acc": 0.8912802626640808 + }, + { + "epoch": 1.4039517014270033, + "grad_norm": 0.3365088550939114, + "learning_rate": 2.242920552111473e-06, + "loss": 0.33001708984375, + "step": 320, + "token_acc": 0.8936313816098664 + }, + { + "epoch": 1.4083424807903402, + "grad_norm": 0.35064050650739736, + "learning_rate": 2.212729962464051e-06, + "loss": 0.3333740234375, + "step": 321, + "token_acc": 0.8919439167493234 + }, + { + "epoch": 1.4127332601536773, + "grad_norm": 0.32899381373645326, + "learning_rate": 2.182686096794852e-06, + "loss": 0.3253173828125, + "step": 322, + "token_acc": 0.8956434103956784 + }, + { + "epoch": 1.4171240395170144, + "grad_norm": 0.3414271967170536, + "learning_rate": 2.1527905366354292e-06, + "loss": 0.435791015625, + "step": 323, + "token_acc": 0.8756033009115767 + }, + { + "epoch": 1.4215148188803512, + "grad_norm": 0.35099050188199654, + "learning_rate": 2.1230448557104087e-06, + "loss": 0.31524658203125, + "step": 324, + "token_acc": 0.8985438426286719 + }, + { + "epoch": 1.4259055982436881, + "grad_norm": 0.32002163766716885, + "learning_rate": 2.093450619854671e-06, + "loss": 0.349609375, + "step": 325, + "token_acc": 0.8895163438997129 + }, + { + "epoch": 1.4302963776070252, + "grad_norm": 0.3237970999566587, + "learning_rate": 2.064009386930915e-06, + "loss": 0.34820556640625, + "step": 326, + "token_acc": 0.8896111892334215 + }, + { + "epoch": 1.4346871569703623, + "grad_norm": 0.3517155890476678, + "learning_rate": 2.0347227067476478e-06, + "loss": 0.32562255859375, + "step": 327, + "token_acc": 0.8950105622325046 + }, + { + "epoch": 1.4390779363336992, + "grad_norm": 0.32561829823435806, + "learning_rate": 2.005592120977606e-06, + "loss": 0.332275390625, + "step": 328, + "token_acc": 0.8935006692903918 + }, + { + "epoch": 1.4434687156970363, + "grad_norm": 0.35317318467760783, + "learning_rate": 1.9766191630765964e-06, + "loss": 0.3367919921875, + "step": 329, + "token_acc": 0.8930185979344394 + }, + { + "epoch": 1.4478594950603731, + "grad_norm": 0.30076628679605677, + "learning_rate": 1.9478053582027826e-06, + "loss": 0.30316162109375, + "step": 330, + "token_acc": 0.9033544387575009 + }, + { + "epoch": 1.4522502744237102, + "grad_norm": 0.29803366911037665, + "learning_rate": 1.919152223136391e-06, + "loss": 0.29071044921875, + "step": 331, + "token_acc": 0.9080499399453981 + }, + { + "epoch": 1.4566410537870471, + "grad_norm": 0.31440029376550305, + "learning_rate": 1.8906612661998698e-06, + "loss": 0.32464599609375, + "step": 332, + "token_acc": 0.8953354269796491 + }, + { + "epoch": 1.4610318331503842, + "grad_norm": 0.3420264812794626, + "learning_rate": 1.8623339871784869e-06, + "loss": 0.3382568359375, + "step": 333, + "token_acc": 0.8920696785648554 + }, + { + "epoch": 1.4654226125137213, + "grad_norm": 0.3395263811275314, + "learning_rate": 1.8341718772413852e-06, + "loss": 0.33270263671875, + "step": 334, + "token_acc": 0.8928632742755116 + }, + { + "epoch": 1.4698133918770582, + "grad_norm": 0.3580272934709791, + "learning_rate": 1.8061764188630831e-06, + "loss": 0.33453369140625, + "step": 335, + "token_acc": 0.8917485048440845 + }, + { + "epoch": 1.474204171240395, + "grad_norm": 0.3544028793357353, + "learning_rate": 1.7783490857454354e-06, + "loss": 0.34503173828125, + "step": 336, + "token_acc": 0.8893608681300544 + }, + { + "epoch": 1.4785949506037321, + "grad_norm": 0.33934434336633224, + "learning_rate": 1.750691342740058e-06, + "loss": 0.3255615234375, + "step": 337, + "token_acc": 0.8956289461848922 + }, + { + "epoch": 1.4829857299670692, + "grad_norm": 0.311157147542992, + "learning_rate": 1.7232046457712164e-06, + "loss": 0.31121826171875, + "step": 338, + "token_acc": 0.8997806631757427 + }, + { + "epoch": 1.487376509330406, + "grad_norm": 0.31060720949060305, + "learning_rate": 1.6958904417591853e-06, + "loss": 0.302978515625, + "step": 339, + "token_acc": 0.9030642018435692 + }, + { + "epoch": 1.4917672886937432, + "grad_norm": 0.31680783960377595, + "learning_rate": 1.668750168544081e-06, + "loss": 0.31939697265625, + "step": 340, + "token_acc": 0.8974534233919073 + }, + { + "epoch": 1.49615806805708, + "grad_norm": 0.35214507098758113, + "learning_rate": 1.641785254810172e-06, + "loss": 0.34417724609375, + "step": 341, + "token_acc": 0.8891002217503283 + }, + { + "epoch": 1.5005488474204172, + "grad_norm": 0.3064286764090022, + "learning_rate": 1.6149971200106723e-06, + "loss": 0.32403564453125, + "step": 342, + "token_acc": 0.8965564549320895 + }, + { + "epoch": 1.504939626783754, + "grad_norm": 0.3390141744530287, + "learning_rate": 1.5883871742930257e-06, + "loss": 0.3338623046875, + "step": 343, + "token_acc": 0.8933097878348695 + }, + { + "epoch": 1.5093304061470911, + "grad_norm": 0.3211934499221605, + "learning_rate": 1.561956818424661e-06, + "loss": 0.34765625, + "step": 344, + "token_acc": 0.8888465271038403 + }, + { + "epoch": 1.5137211855104282, + "grad_norm": 0.30787984012661906, + "learning_rate": 1.5357074437192688e-06, + "loss": 0.32342529296875, + "step": 345, + "token_acc": 0.8970173589849479 + }, + { + "epoch": 1.518111964873765, + "grad_norm": 0.3230910486286158, + "learning_rate": 1.5096404319635533e-06, + "loss": 0.321533203125, + "step": 346, + "token_acc": 0.8965351810190307 + }, + { + "epoch": 1.522502744237102, + "grad_norm": 0.33417908219974973, + "learning_rate": 1.483757155344503e-06, + "loss": 0.3433837890625, + "step": 347, + "token_acc": 0.8908140033456329 + }, + { + "epoch": 1.526893523600439, + "grad_norm": 0.3185138344297658, + "learning_rate": 1.4580589763771413e-06, + "loss": 0.3184814453125, + "step": 348, + "token_acc": 0.8991635756676558 + }, + { + "epoch": 1.5312843029637762, + "grad_norm": 0.36687139913335864, + "learning_rate": 1.432547247832819e-06, + "loss": 0.33880615234375, + "step": 349, + "token_acc": 0.8914219448293634 + }, + { + "epoch": 1.535675082327113, + "grad_norm": 0.3121656092094008, + "learning_rate": 1.4072233126679985e-06, + "loss": 0.32598876953125, + "step": 350, + "token_acc": 0.8951324014890778 + }, + { + "epoch": 1.54006586169045, + "grad_norm": 0.31448626553122083, + "learning_rate": 1.3820885039535564e-06, + "loss": 0.3004150390625, + "step": 351, + "token_acc": 0.9019656109628279 + }, + { + "epoch": 1.544456641053787, + "grad_norm": 0.29880386591493907, + "learning_rate": 1.3571441448046086e-06, + "loss": 0.31536865234375, + "step": 352, + "token_acc": 0.9001778253729766 + }, + { + "epoch": 1.548847420417124, + "grad_norm": 0.3166310885575997, + "learning_rate": 1.3323915483108662e-06, + "loss": 0.324462890625, + "step": 353, + "token_acc": 0.898118293778741 + }, + { + "epoch": 1.553238199780461, + "grad_norm": 0.35151841989920424, + "learning_rate": 1.3078320174675141e-06, + "loss": 0.35040283203125, + "step": 354, + "token_acc": 0.8890206514150418 + }, + { + "epoch": 1.557628979143798, + "grad_norm": 0.33479685910226337, + "learning_rate": 1.2834668451066118e-06, + "loss": 0.33172607421875, + "step": 355, + "token_acc": 0.8938539080719176 + }, + { + "epoch": 1.5620197585071351, + "grad_norm": 0.30928796916886575, + "learning_rate": 1.259297313829046e-06, + "loss": 0.3077392578125, + "step": 356, + "token_acc": 0.9004423980861475 + }, + { + "epoch": 1.566410537870472, + "grad_norm": 0.34920620061200947, + "learning_rate": 1.2353246959370086e-06, + "loss": 0.3406982421875, + "step": 357, + "token_acc": 0.8906262374837359 + }, + { + "epoch": 1.570801317233809, + "grad_norm": 0.32542342214290737, + "learning_rate": 1.2115502533670253e-06, + "loss": 0.3040771484375, + "step": 358, + "token_acc": 0.9017812464520144 + }, + { + "epoch": 1.575192096597146, + "grad_norm": 0.3088306425812143, + "learning_rate": 1.1879752376235231e-06, + "loss": 0.31365966796875, + "step": 359, + "token_acc": 0.8998575177235227 + }, + { + "epoch": 1.579582875960483, + "grad_norm": 0.3673988268314337, + "learning_rate": 1.1646008897129546e-06, + "loss": 0.3232421875, + "step": 360, + "token_acc": 0.8960839865568269 + }, + { + "epoch": 1.58397365532382, + "grad_norm": 0.38954586735694496, + "learning_rate": 1.1414284400784643e-06, + "loss": 0.316162109375, + "step": 361, + "token_acc": 0.8983161436387649 + }, + { + "epoch": 1.5883644346871568, + "grad_norm": 0.3338851804634168, + "learning_rate": 1.118459108535122e-06, + "loss": 0.32855224609375, + "step": 362, + "token_acc": 0.8951910712357544 + }, + { + "epoch": 1.592755214050494, + "grad_norm": 0.31354249264446227, + "learning_rate": 1.0956941042057106e-06, + "loss": 0.30206298828125, + "step": 363, + "token_acc": 0.9030254446357413 + }, + { + "epoch": 1.597145993413831, + "grad_norm": 0.32617845412106616, + "learning_rate": 1.0731346254570735e-06, + "loss": 0.32171630859375, + "step": 364, + "token_acc": 0.8975390925403742 + }, + { + "epoch": 1.6015367727771679, + "grad_norm": 0.3693973221493379, + "learning_rate": 1.0507818598370355e-06, + "loss": 0.3443603515625, + "step": 365, + "token_acc": 0.888619439372684 + }, + { + "epoch": 1.605927552140505, + "grad_norm": 0.3341196055937999, + "learning_rate": 1.0286369840118859e-06, + "loss": 0.3385009765625, + "step": 366, + "token_acc": 0.8918811806446743 + }, + { + "epoch": 1.610318331503842, + "grad_norm": 0.34605954547206286, + "learning_rate": 1.006701163704445e-06, + "loss": 0.339599609375, + "step": 367, + "token_acc": 0.8921929642200683 + }, + { + "epoch": 1.614709110867179, + "grad_norm": 0.3214712819943197, + "learning_rate": 9.849755536326866e-07, + "loss": 0.32080078125, + "step": 368, + "token_acc": 0.8973029365828966 + }, + { + "epoch": 1.6190998902305158, + "grad_norm": 0.33765450186529306, + "learning_rate": 9.63461297448966e-07, + "loss": 0.33447265625, + "step": 369, + "token_acc": 0.8924900716032999 + }, + { + "epoch": 1.623490669593853, + "grad_norm": 0.32487660178259903, + "learning_rate": 9.421595276798084e-07, + "loss": 0.31402587890625, + "step": 370, + "token_acc": 0.9003300316619045 + }, + { + "epoch": 1.62788144895719, + "grad_norm": 0.31734405029950935, + "learning_rate": 9.210713656663023e-07, + "loss": 0.334136962890625, + "step": 371, + "token_acc": 0.8945020351862978 + }, + { + "epoch": 1.6322722283205269, + "grad_norm": 0.36552587976973605, + "learning_rate": 9.001979215050544e-07, + "loss": 0.37933349609375, + "step": 372, + "token_acc": 0.8805883798673687 + }, + { + "epoch": 1.6366630076838637, + "grad_norm": 0.33850050157465783, + "learning_rate": 8.795402939897679e-07, + "loss": 0.3240966796875, + "step": 373, + "token_acc": 0.8961853403592756 + }, + { + "epoch": 1.6410537870472008, + "grad_norm": 0.3304963750478896, + "learning_rate": 8.590995705533994e-07, + "loss": 0.33135986328125, + "step": 374, + "token_acc": 0.8949704286813404 + }, + { + "epoch": 1.645444566410538, + "grad_norm": 0.344128572070614, + "learning_rate": 8.388768272109105e-07, + "loss": 0.32818603515625, + "step": 375, + "token_acc": 0.8955371784865519 + }, + { + "epoch": 1.6498353457738748, + "grad_norm": 0.34883527846917517, + "learning_rate": 8.188731285026219e-07, + "loss": 0.3492431640625, + "step": 376, + "token_acc": 0.8886745390313118 + }, + { + "epoch": 1.654226125137212, + "grad_norm": 0.3925879050862546, + "learning_rate": 7.99089527438191e-07, + "loss": 0.3648681640625, + "step": 377, + "token_acc": 0.8843120912345157 + }, + { + "epoch": 1.658616904500549, + "grad_norm": 0.3267473216090058, + "learning_rate": 7.795270654411635e-07, + "loss": 0.3353271484375, + "step": 378, + "token_acc": 0.8933586229167774 + }, + { + "epoch": 1.6630076838638859, + "grad_norm": 0.3505350578881492, + "learning_rate": 7.601867722941642e-07, + "loss": 0.31964111328125, + "step": 379, + "token_acc": 0.8975450696383639 + }, + { + "epoch": 1.6673984632272227, + "grad_norm": 0.3140296646528115, + "learning_rate": 7.410696660846761e-07, + "loss": 0.348388671875, + "step": 380, + "token_acc": 0.8905270618461562 + }, + { + "epoch": 1.6717892425905598, + "grad_norm": 0.33890201394504826, + "learning_rate": 7.22176753151464e-07, + "loss": 0.33770751953125, + "step": 381, + "token_acc": 0.8921922096954155 + }, + { + "epoch": 1.676180021953897, + "grad_norm": 0.34754729201367335, + "learning_rate": 7.035090280315854e-07, + "loss": 0.3411865234375, + "step": 382, + "token_acc": 0.8912410516370508 + }, + { + "epoch": 1.6805708013172338, + "grad_norm": 0.30817240468755525, + "learning_rate": 6.850674734080454e-07, + "loss": 0.3236083984375, + "step": 383, + "token_acc": 0.897392357908823 + }, + { + "epoch": 1.6849615806805707, + "grad_norm": 0.3133130749995393, + "learning_rate": 6.66853060058063e-07, + "loss": 0.3126220703125, + "step": 384, + "token_acc": 0.9002761491017165 + }, + { + "epoch": 1.6893523600439078, + "grad_norm": 0.3528327881645742, + "learning_rate": 6.488667468019727e-07, + "loss": 0.361083984375, + "step": 385, + "token_acc": 0.8851972838730512 + }, + { + "epoch": 1.6937431394072449, + "grad_norm": 0.3125258347085141, + "learning_rate": 6.31109480452749e-07, + "loss": 0.34759521484375, + "step": 386, + "token_acc": 0.8904842245197799 + }, + { + "epoch": 1.6981339187705817, + "grad_norm": 0.3354760865119189, + "learning_rate": 6.135821957661658e-07, + "loss": 0.33538818359375, + "step": 387, + "token_acc": 0.8933355095618039 + }, + { + "epoch": 1.7025246981339188, + "grad_norm": 0.2980443483920128, + "learning_rate": 5.962858153915896e-07, + "loss": 0.3055419921875, + "step": 388, + "token_acc": 0.9020604964779447 + }, + { + "epoch": 1.706915477497256, + "grad_norm": 0.3475742282240845, + "learning_rate": 5.792212498234134e-07, + "loss": 0.34375, + "step": 389, + "token_acc": 0.8913870681035441 + }, + { + "epoch": 1.7113062568605928, + "grad_norm": 0.3382744660227279, + "learning_rate": 5.623893973531225e-07, + "loss": 0.32763671875, + "step": 390, + "token_acc": 0.8944369685957213 + }, + { + "epoch": 1.7156970362239297, + "grad_norm": 0.36406944654219947, + "learning_rate": 5.457911440220154e-07, + "loss": 0.33001708984375, + "step": 391, + "token_acc": 0.8939367983710608 + }, + { + "epoch": 1.7200878155872668, + "grad_norm": 0.2973145447250047, + "learning_rate": 5.294273635745517e-07, + "loss": 0.31414794921875, + "step": 392, + "token_acc": 0.8997379322005241 + }, + { + "epoch": 1.7244785949506038, + "grad_norm": 0.31937049306043686, + "learning_rate": 5.132989174123659e-07, + "loss": 0.33099365234375, + "step": 393, + "token_acc": 0.8942759065257196 + }, + { + "epoch": 1.7288693743139407, + "grad_norm": 0.3195244152084989, + "learning_rate": 4.97406654548922e-07, + "loss": 0.323974609375, + "step": 394, + "token_acc": 0.8963840244506857 + }, + { + "epoch": 1.7332601536772776, + "grad_norm": 0.32918330351597275, + "learning_rate": 4.817514115648164e-07, + "loss": 0.32049560546875, + "step": 395, + "token_acc": 0.8964964818227509 + }, + { + "epoch": 1.7376509330406147, + "grad_norm": 0.30823870393244884, + "learning_rate": 4.663340125637389e-07, + "loss": 0.31060791015625, + "step": 396, + "token_acc": 0.8997806321920347 + }, + { + "epoch": 1.7420417124039518, + "grad_norm": 0.33331387895596026, + "learning_rate": 4.511552691290988e-07, + "loss": 0.32708740234375, + "step": 397, + "token_acc": 0.894811041977903 + }, + { + "epoch": 1.7464324917672887, + "grad_norm": 0.3657346964297189, + "learning_rate": 4.362159802812971e-07, + "loss": 0.32757568359375, + "step": 398, + "token_acc": 0.894116022835441 + }, + { + "epoch": 1.7508232711306255, + "grad_norm": 0.3375902606396153, + "learning_rate": 4.215169324356666e-07, + "loss": 0.3341064453125, + "step": 399, + "token_acc": 0.8944647443010685 + }, + { + "epoch": 1.7552140504939628, + "grad_norm": 0.3490777079667427, + "learning_rate": 4.070588993610697e-07, + "loss": 0.33544921875, + "step": 400, + "token_acc": 0.8928451446826995 + }, + { + "epoch": 1.7596048298572997, + "grad_norm": 0.3454508780019287, + "learning_rate": 3.928426421391773e-07, + "loss": 0.32806396484375, + "step": 401, + "token_acc": 0.8937952449717322 + }, + { + "epoch": 1.7639956092206366, + "grad_norm": 0.3432807805381615, + "learning_rate": 3.7886890912439633e-07, + "loss": 0.3271484375, + "step": 402, + "token_acc": 0.8948156115198462 + }, + { + "epoch": 1.7683863885839737, + "grad_norm": 0.32460523818298503, + "learning_rate": 3.651384359044774e-07, + "loss": 0.36053466796875, + "step": 403, + "token_acc": 0.8857531124924252 + }, + { + "epoch": 1.7727771679473108, + "grad_norm": 0.32862364677559325, + "learning_rate": 3.516519452617922e-07, + "loss": 0.327392578125, + "step": 404, + "token_acc": 0.8948030658624465 + }, + { + "epoch": 1.7771679473106476, + "grad_norm": 0.3215013561453796, + "learning_rate": 3.3841014713529184e-07, + "loss": 0.30902099609375, + "step": 405, + "token_acc": 0.8998735514791382 + }, + { + "epoch": 1.7815587266739845, + "grad_norm": 0.3134191902573514, + "learning_rate": 3.254137385831263e-07, + "loss": 0.326904296875, + "step": 406, + "token_acc": 0.8955593501674187 + }, + { + "epoch": 1.7859495060373216, + "grad_norm": 0.29496479459085717, + "learning_rate": 3.1266340374595693e-07, + "loss": 0.306915283203125, + "step": 407, + "token_acc": 0.9027193922893538 + }, + { + "epoch": 1.7903402854006587, + "grad_norm": 0.3259220503055919, + "learning_rate": 3.0015981381094073e-07, + "loss": 0.33740234375, + "step": 408, + "token_acc": 0.8932714405573998 + }, + { + "epoch": 1.7947310647639956, + "grad_norm": 0.3462558528034366, + "learning_rate": 2.8790362697639685e-07, + "loss": 0.3414306640625, + "step": 409, + "token_acc": 0.8913687342038753 + }, + { + "epoch": 1.7991218441273324, + "grad_norm": 0.3605113445687066, + "learning_rate": 2.7589548841716274e-07, + "loss": 0.3402099609375, + "step": 410, + "token_acc": 0.89065803514579 + }, + { + "epoch": 1.8035126234906695, + "grad_norm": 0.3264923577051289, + "learning_rate": 2.64136030250628e-07, + "loss": 0.333251953125, + "step": 411, + "token_acc": 0.8939395383322247 + }, + { + "epoch": 1.8079034028540066, + "grad_norm": 0.329246741520078, + "learning_rate": 2.526258715034602e-07, + "loss": 0.3447265625, + "step": 412, + "token_acc": 0.8896532279688955 + }, + { + "epoch": 1.8122941822173435, + "grad_norm": 0.3203671989051025, + "learning_rate": 2.4136561807901916e-07, + "loss": 0.3319091796875, + "step": 413, + "token_acc": 0.8944627231083568 + }, + { + "epoch": 1.8166849615806806, + "grad_norm": 0.3128688524191343, + "learning_rate": 2.3035586272546207e-07, + "loss": 0.3179931640625, + "step": 414, + "token_acc": 0.8994386637458927 + }, + { + "epoch": 1.8210757409440177, + "grad_norm": 0.3162543682462319, + "learning_rate": 2.1959718500454196e-07, + "loss": 0.34197998046875, + "step": 415, + "token_acc": 0.8922324478709701 + }, + { + "epoch": 1.8254665203073546, + "grad_norm": 0.31628023882532796, + "learning_rate": 2.0909015126109488e-07, + "loss": 0.3271484375, + "step": 416, + "token_acc": 0.896125079935866 + }, + { + "epoch": 1.8298572996706914, + "grad_norm": 0.35061974607883883, + "learning_rate": 1.988353145932298e-07, + "loss": 0.37640380859375, + "step": 417, + "token_acc": 0.8815436502633716 + }, + { + "epoch": 1.8342480790340285, + "grad_norm": 0.31617503605918285, + "learning_rate": 1.8883321482321583e-07, + "loss": 0.3310546875, + "step": 418, + "token_acc": 0.8962308874283595 + }, + { + "epoch": 1.8386388583973656, + "grad_norm": 0.3169484487134376, + "learning_rate": 1.7908437846906158e-07, + "loss": 0.32293701171875, + "step": 419, + "token_acc": 0.8963603569394342 + }, + { + "epoch": 1.8430296377607025, + "grad_norm": 0.3365985480902347, + "learning_rate": 1.6958931871679908e-07, + "loss": 0.33087158203125, + "step": 420, + "token_acc": 0.894784686617356 + }, + { + "epoch": 1.8474204171240394, + "grad_norm": 0.3420749242889434, + "learning_rate": 1.603485353934703e-07, + "loss": 0.312255859375, + "step": 421, + "token_acc": 0.899052783716958 + }, + { + "epoch": 1.8518111964873765, + "grad_norm": 0.34636887044367853, + "learning_rate": 1.5136251494081822e-07, + "loss": 0.31878662109375, + "step": 422, + "token_acc": 0.8978319799856161 + }, + { + "epoch": 1.8562019758507136, + "grad_norm": 0.34035225117826334, + "learning_rate": 1.4263173038967627e-07, + "loss": 0.37469482421875, + "step": 423, + "token_acc": 0.882996524649247 + }, + { + "epoch": 1.8605927552140504, + "grad_norm": 0.3455548802919666, + "learning_rate": 1.3415664133506812e-07, + "loss": 0.3470458984375, + "step": 424, + "token_acc": 0.888893873092701 + }, + { + "epoch": 1.8649835345773875, + "grad_norm": 0.31846607420181183, + "learning_rate": 1.2593769391201827e-07, + "loss": 0.322265625, + "step": 425, + "token_acc": 0.8966021505376344 + }, + { + "epoch": 1.8693743139407246, + "grad_norm": 0.3366284106732347, + "learning_rate": 1.1797532077206187e-07, + "loss": 0.33251953125, + "step": 426, + "token_acc": 0.8953254265318595 + }, + { + "epoch": 1.8737650933040615, + "grad_norm": 0.3009000303918809, + "learning_rate": 1.1026994106047296e-07, + "loss": 0.3204345703125, + "step": 427, + "token_acc": 0.8987217647586135 + }, + { + "epoch": 1.8781558726673984, + "grad_norm": 0.3270006697398233, + "learning_rate": 1.0282196039419823e-07, + "loss": 0.327392578125, + "step": 428, + "token_acc": 0.8962514769366211 + }, + { + "epoch": 1.8825466520307355, + "grad_norm": 0.34482835758960784, + "learning_rate": 9.5631770840508e-08, + "loss": 0.34466552734375, + "step": 429, + "token_acc": 0.8888076029756629 + }, + { + "epoch": 1.8869374313940726, + "grad_norm": 0.2985600447147717, + "learning_rate": 8.869975089635552e-08, + "loss": 0.31842041015625, + "step": 430, + "token_acc": 0.8991015565064698 + }, + { + "epoch": 1.8913282107574094, + "grad_norm": 0.3662613501530102, + "learning_rate": 8.202626546845172e-08, + "loss": 0.3890380859375, + "step": 431, + "token_acc": 0.8769567149745507 + }, + { + "epoch": 1.8957189901207463, + "grad_norm": 0.3136762743080743, + "learning_rate": 7.561166585405789e-08, + "loss": 0.3333740234375, + "step": 432, + "token_acc": 0.8939009149833037 + }, + { + "epoch": 1.9001097694840834, + "grad_norm": 0.2989801839581692, + "learning_rate": 6.945628972249208e-08, + "loss": 0.3311767578125, + "step": 433, + "token_acc": 0.8955504875298014 + }, + { + "epoch": 1.9045005488474205, + "grad_norm": 0.33986553610150577, + "learning_rate": 6.356046109735614e-08, + "loss": 0.3505859375, + "step": 434, + "token_acc": 0.8896977782338755 + }, + { + "epoch": 1.9088913282107574, + "grad_norm": 0.3585350716098544, + "learning_rate": 5.7924490339474335e-08, + "loss": 0.33050537109375, + "step": 435, + "token_acc": 0.8932045424757055 + }, + { + "epoch": 1.9132821075740944, + "grad_norm": 0.33318455426689786, + "learning_rate": 5.2548674130561974e-08, + "loss": 0.31787109375, + "step": 436, + "token_acc": 0.8970736152543393 + }, + { + "epoch": 1.9176728869374315, + "grad_norm": 0.3560122835652356, + "learning_rate": 4.743329545760122e-08, + "loss": 0.32080078125, + "step": 437, + "token_acc": 0.8963951960241702 + }, + { + "epoch": 1.9220636663007684, + "grad_norm": 0.31085098069734834, + "learning_rate": 4.2578623597949174e-08, + "loss": 0.32952880859375, + "step": 438, + "token_acc": 0.8948359117539363 + }, + { + "epoch": 1.9264544456641053, + "grad_norm": 0.31421688424510547, + "learning_rate": 3.7984914105162474e-08, + "loss": 0.32672119140625, + "step": 439, + "token_acc": 0.896141181272641 + }, + { + "epoch": 1.9308452250274424, + "grad_norm": 0.3396314942583363, + "learning_rate": 3.365240879554144e-08, + "loss": 0.32476806640625, + "step": 440, + "token_acc": 0.89614173015128 + }, + { + "epoch": 1.9352360043907795, + "grad_norm": 0.33770968758923126, + "learning_rate": 2.9581335735404672e-08, + "loss": 0.34759521484375, + "step": 441, + "token_acc": 0.888780297517056 + }, + { + "epoch": 1.9396267837541163, + "grad_norm": 0.3359526828470558, + "learning_rate": 2.577190922908035e-08, + "loss": 0.336669921875, + "step": 442, + "token_acc": 0.8932492221730838 + }, + { + "epoch": 1.9440175631174532, + "grad_norm": 0.30809999333074994, + "learning_rate": 2.222432980762912e-08, + "loss": 0.31591796875, + "step": 443, + "token_acc": 0.8989644887272107 + }, + { + "epoch": 1.9484083424807903, + "grad_norm": 0.32588061865422974, + "learning_rate": 1.8938784218281435e-08, + "loss": 0.3359375, + "step": 444, + "token_acc": 0.8939156473905226 + }, + { + "epoch": 1.9527991218441274, + "grad_norm": 0.316204595673142, + "learning_rate": 1.5915445414613208e-08, + "loss": 0.32470703125, + "step": 445, + "token_acc": 0.8972734617593521 + }, + { + "epoch": 1.9571899012074643, + "grad_norm": 0.3534293579497171, + "learning_rate": 1.3154472547440289e-08, + "loss": 0.34326171875, + "step": 446, + "token_acc": 0.8898358889984115 + }, + { + "epoch": 1.9615806805708014, + "grad_norm": 0.3222323910725636, + "learning_rate": 1.0656010956437979e-08, + "loss": 0.33538818359375, + "step": 447, + "token_acc": 0.89324441088554 + }, + { + "epoch": 1.9659714599341385, + "grad_norm": 0.33723294622312994, + "learning_rate": 8.42019216249046e-09, + "loss": 0.33941650390625, + "step": 448, + "token_acc": 0.8923053003662836 + }, + { + "epoch": 1.9703622392974753, + "grad_norm": 0.34061189298308664, + "learning_rate": 6.447133860771893e-09, + "loss": 0.34271240234375, + "step": 449, + "token_acc": 0.8913180258916412 + }, + { + "epoch": 1.9747530186608122, + "grad_norm": 0.31254186799137346, + "learning_rate": 4.736939914545824e-09, + "loss": 0.3350830078125, + "step": 450, + "token_acc": 0.8941951442968392 + }, + { + "epoch": 1.9791437980241493, + "grad_norm": 0.3494134810547919, + "learning_rate": 3.289700349698999e-09, + "loss": 0.3580322265625, + "step": 451, + "token_acc": 0.8864843405381562 + }, + { + "epoch": 1.9835345773874864, + "grad_norm": 0.34301867099975397, + "learning_rate": 2.1054913500051512e-09, + "loss": 0.3482666015625, + "step": 452, + "token_acc": 0.8892958047945205 + }, + { + "epoch": 1.9879253567508233, + "grad_norm": 0.3549677040304002, + "learning_rate": 1.1843752531104368e-09, + "loss": 0.3450927734375, + "step": 453, + "token_acc": 0.8906645805394129 + }, + { + "epoch": 1.9923161361141601, + "grad_norm": 0.3298335503597009, + "learning_rate": 5.264005472549461e-10, + "loss": 0.3409423828125, + "step": 454, + "token_acc": 0.8928574189757545 + }, + { + "epoch": 1.9967069154774972, + "grad_norm": 0.33394556946336496, + "learning_rate": 1.316018687191889e-10, + "loss": 0.32904052734375, + "step": 455, + "token_acc": 0.8945393287321052 + }, + { + "epoch": 2.0, + "grad_norm": 0.3495909530257807, + "learning_rate": 0.0, + "loss": 0.3092041015625, + "step": 456, + "token_acc": 0.9025272608768885 + } + ], + "logging_steps": 1, + "max_steps": 456, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 50, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.1510127542665216e+16, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}