ISEF / checkpoint-4361 /trainer_state.json
diddmstjr's picture
text_only_model_projector_ft
52d664c verified
Raw
History Blame Contribute Delete
71.5 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 4361,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002293052052281587,
"grad_norm": 18.875,
"learning_rate": 1.9958725063058934e-05,
"loss": 1.7556,
"step": 10
},
{
"epoch": 0.004586104104563174,
"grad_norm": 19.875,
"learning_rate": 1.99128640220133e-05,
"loss": 1.1973,
"step": 20
},
{
"epoch": 0.00687915615684476,
"grad_norm": 15.6875,
"learning_rate": 1.986700298096767e-05,
"loss": 1.2111,
"step": 30
},
{
"epoch": 0.009172208209126348,
"grad_norm": 15.5625,
"learning_rate": 1.9821141939922037e-05,
"loss": 1.2078,
"step": 40
},
{
"epoch": 0.011465260261407933,
"grad_norm": 11.125,
"learning_rate": 1.9775280898876404e-05,
"loss": 1.1955,
"step": 50
},
{
"epoch": 0.01375831231368952,
"grad_norm": 15.875,
"learning_rate": 1.9729419857830774e-05,
"loss": 1.2044,
"step": 60
},
{
"epoch": 0.016051364365971106,
"grad_norm": 17.875,
"learning_rate": 1.968355881678514e-05,
"loss": 1.1926,
"step": 70
},
{
"epoch": 0.018344416418252695,
"grad_norm": 14.5,
"learning_rate": 1.963769777573951e-05,
"loss": 1.2002,
"step": 80
},
{
"epoch": 0.02063746847053428,
"grad_norm": 11.8125,
"learning_rate": 1.9591836734693877e-05,
"loss": 1.1892,
"step": 90
},
{
"epoch": 0.022930520522815866,
"grad_norm": 11.5,
"learning_rate": 1.9545975693648247e-05,
"loss": 1.19,
"step": 100
},
{
"epoch": 0.025223572575097455,
"grad_norm": 16.0,
"learning_rate": 1.9500114652602614e-05,
"loss": 1.1949,
"step": 110
},
{
"epoch": 0.02751662462737904,
"grad_norm": 16.75,
"learning_rate": 1.9454253611556984e-05,
"loss": 1.1921,
"step": 120
},
{
"epoch": 0.029809676679660627,
"grad_norm": 14.125,
"learning_rate": 1.940839257051135e-05,
"loss": 1.1958,
"step": 130
},
{
"epoch": 0.03210272873194221,
"grad_norm": 11.5625,
"learning_rate": 1.936253152946572e-05,
"loss": 1.1852,
"step": 140
},
{
"epoch": 0.034395780784223805,
"grad_norm": 18.25,
"learning_rate": 1.9316670488420087e-05,
"loss": 1.187,
"step": 150
},
{
"epoch": 0.03668883283650539,
"grad_norm": 15.9375,
"learning_rate": 1.9270809447374457e-05,
"loss": 1.1867,
"step": 160
},
{
"epoch": 0.038981884888786976,
"grad_norm": 19.75,
"learning_rate": 1.9224948406328824e-05,
"loss": 1.2015,
"step": 170
},
{
"epoch": 0.04127493694106856,
"grad_norm": 11.625,
"learning_rate": 1.9179087365283194e-05,
"loss": 1.2059,
"step": 180
},
{
"epoch": 0.04356798899335015,
"grad_norm": 16.375,
"learning_rate": 1.913322632423756e-05,
"loss": 1.1962,
"step": 190
},
{
"epoch": 0.04586104104563173,
"grad_norm": 13.0,
"learning_rate": 1.908736528319193e-05,
"loss": 1.1955,
"step": 200
},
{
"epoch": 0.048154093097913325,
"grad_norm": 11.4375,
"learning_rate": 1.9041504242146297e-05,
"loss": 1.1766,
"step": 210
},
{
"epoch": 0.05044714515019491,
"grad_norm": 11.5625,
"learning_rate": 1.8995643201100668e-05,
"loss": 1.1857,
"step": 220
},
{
"epoch": 0.052740197202476496,
"grad_norm": 17.875,
"learning_rate": 1.8949782160055034e-05,
"loss": 1.1981,
"step": 230
},
{
"epoch": 0.05503324925475808,
"grad_norm": 14.25,
"learning_rate": 1.8903921119009404e-05,
"loss": 1.2041,
"step": 240
},
{
"epoch": 0.05732630130703967,
"grad_norm": 16.0,
"learning_rate": 1.885806007796377e-05,
"loss": 1.1989,
"step": 250
},
{
"epoch": 0.05961935335932125,
"grad_norm": 17.25,
"learning_rate": 1.881219903691814e-05,
"loss": 1.196,
"step": 260
},
{
"epoch": 0.061912405411602846,
"grad_norm": 13.3125,
"learning_rate": 1.8766337995872508e-05,
"loss": 1.1797,
"step": 270
},
{
"epoch": 0.06420545746388442,
"grad_norm": 13.3125,
"learning_rate": 1.8720476954826878e-05,
"loss": 1.1974,
"step": 280
},
{
"epoch": 0.06649850951616602,
"grad_norm": 10.375,
"learning_rate": 1.8674615913781244e-05,
"loss": 1.1929,
"step": 290
},
{
"epoch": 0.06879156156844761,
"grad_norm": 15.125,
"learning_rate": 1.8628754872735614e-05,
"loss": 1.1838,
"step": 300
},
{
"epoch": 0.0710846136207292,
"grad_norm": 8.75,
"learning_rate": 1.858289383168998e-05,
"loss": 1.1515,
"step": 310
},
{
"epoch": 0.07337766567301078,
"grad_norm": 10.9375,
"learning_rate": 1.8537032790644348e-05,
"loss": 1.2107,
"step": 320
},
{
"epoch": 0.07567071772529237,
"grad_norm": 12.5625,
"learning_rate": 1.8491171749598718e-05,
"loss": 1.1874,
"step": 330
},
{
"epoch": 0.07796376977757395,
"grad_norm": 9.6875,
"learning_rate": 1.8445310708553084e-05,
"loss": 1.1988,
"step": 340
},
{
"epoch": 0.08025682182985554,
"grad_norm": 10.5625,
"learning_rate": 1.8399449667507454e-05,
"loss": 1.1998,
"step": 350
},
{
"epoch": 0.08254987388213712,
"grad_norm": 9.1875,
"learning_rate": 1.835358862646182e-05,
"loss": 1.1957,
"step": 360
},
{
"epoch": 0.08484292593441871,
"grad_norm": 10.1875,
"learning_rate": 1.830772758541619e-05,
"loss": 1.204,
"step": 370
},
{
"epoch": 0.0871359779867003,
"grad_norm": 17.125,
"learning_rate": 1.8261866544370558e-05,
"loss": 1.191,
"step": 380
},
{
"epoch": 0.08942903003898188,
"grad_norm": 12.1875,
"learning_rate": 1.8216005503324928e-05,
"loss": 1.19,
"step": 390
},
{
"epoch": 0.09172208209126347,
"grad_norm": 9.9375,
"learning_rate": 1.8170144462279294e-05,
"loss": 1.1729,
"step": 400
},
{
"epoch": 0.09401513414354506,
"grad_norm": 12.875,
"learning_rate": 1.8124283421233664e-05,
"loss": 1.201,
"step": 410
},
{
"epoch": 0.09630818619582665,
"grad_norm": 15.1875,
"learning_rate": 1.807842238018803e-05,
"loss": 1.1834,
"step": 420
},
{
"epoch": 0.09860123824810824,
"grad_norm": 10.125,
"learning_rate": 1.80325613391424e-05,
"loss": 1.1984,
"step": 430
},
{
"epoch": 0.10089429030038982,
"grad_norm": 12.875,
"learning_rate": 1.7986700298096768e-05,
"loss": 1.1877,
"step": 440
},
{
"epoch": 0.10318734235267141,
"grad_norm": 14.8125,
"learning_rate": 1.7940839257051138e-05,
"loss": 1.179,
"step": 450
},
{
"epoch": 0.10548039440495299,
"grad_norm": 17.75,
"learning_rate": 1.7894978216005504e-05,
"loss": 1.1861,
"step": 460
},
{
"epoch": 0.10777344645723458,
"grad_norm": 8.375,
"learning_rate": 1.7849117174959874e-05,
"loss": 1.1912,
"step": 470
},
{
"epoch": 0.11006649850951616,
"grad_norm": 8.9375,
"learning_rate": 1.780325613391424e-05,
"loss": 1.1795,
"step": 480
},
{
"epoch": 0.11235955056179775,
"grad_norm": 11.5625,
"learning_rate": 1.775739509286861e-05,
"loss": 1.1827,
"step": 490
},
{
"epoch": 0.11465260261407934,
"grad_norm": 11.625,
"learning_rate": 1.7711534051822978e-05,
"loss": 1.1813,
"step": 500
},
{
"epoch": 0.11694565466636092,
"grad_norm": 11.6875,
"learning_rate": 1.7665673010777348e-05,
"loss": 1.1836,
"step": 510
},
{
"epoch": 0.1192387067186425,
"grad_norm": 7.84375,
"learning_rate": 1.7619811969731714e-05,
"loss": 1.171,
"step": 520
},
{
"epoch": 0.1215317587709241,
"grad_norm": 11.9375,
"learning_rate": 1.7573950928686084e-05,
"loss": 1.1835,
"step": 530
},
{
"epoch": 0.12382481082320569,
"grad_norm": 14.9375,
"learning_rate": 1.752808988764045e-05,
"loss": 1.1887,
"step": 540
},
{
"epoch": 0.12611786287548726,
"grad_norm": 13.3125,
"learning_rate": 1.748222884659482e-05,
"loss": 1.1737,
"step": 550
},
{
"epoch": 0.12841091492776885,
"grad_norm": 13.3125,
"learning_rate": 1.7436367805549188e-05,
"loss": 1.1871,
"step": 560
},
{
"epoch": 0.13070396698005043,
"grad_norm": 10.5625,
"learning_rate": 1.7390506764503558e-05,
"loss": 1.1873,
"step": 570
},
{
"epoch": 0.13299701903233205,
"grad_norm": 8.1875,
"learning_rate": 1.7344645723457925e-05,
"loss": 1.1946,
"step": 580
},
{
"epoch": 0.13529007108461363,
"grad_norm": 13.9375,
"learning_rate": 1.7298784682412295e-05,
"loss": 1.1838,
"step": 590
},
{
"epoch": 0.13758312313689522,
"grad_norm": 12.8125,
"learning_rate": 1.725292364136666e-05,
"loss": 1.1873,
"step": 600
},
{
"epoch": 0.1398761751891768,
"grad_norm": 11.3125,
"learning_rate": 1.7207062600321028e-05,
"loss": 1.1802,
"step": 610
},
{
"epoch": 0.1421692272414584,
"grad_norm": 15.25,
"learning_rate": 1.7161201559275398e-05,
"loss": 1.1806,
"step": 620
},
{
"epoch": 0.14446227929373998,
"grad_norm": 10.1875,
"learning_rate": 1.7115340518229765e-05,
"loss": 1.195,
"step": 630
},
{
"epoch": 0.14675533134602156,
"grad_norm": 9.875,
"learning_rate": 1.7069479477184135e-05,
"loss": 1.1835,
"step": 640
},
{
"epoch": 0.14904838339830315,
"grad_norm": 11.6875,
"learning_rate": 1.70236184361385e-05,
"loss": 1.1884,
"step": 650
},
{
"epoch": 0.15134143545058473,
"grad_norm": 12.0,
"learning_rate": 1.697775739509287e-05,
"loss": 1.1811,
"step": 660
},
{
"epoch": 0.15363448750286632,
"grad_norm": 9.875,
"learning_rate": 1.6931896354047238e-05,
"loss": 1.195,
"step": 670
},
{
"epoch": 0.1559275395551479,
"grad_norm": 11.0,
"learning_rate": 1.6886035313001608e-05,
"loss": 1.1827,
"step": 680
},
{
"epoch": 0.1582205916074295,
"grad_norm": 11.1875,
"learning_rate": 1.6840174271955975e-05,
"loss": 1.1868,
"step": 690
},
{
"epoch": 0.16051364365971107,
"grad_norm": 7.84375,
"learning_rate": 1.6794313230910345e-05,
"loss": 1.1813,
"step": 700
},
{
"epoch": 0.16280669571199266,
"grad_norm": 13.0625,
"learning_rate": 1.674845218986471e-05,
"loss": 1.1856,
"step": 710
},
{
"epoch": 0.16509974776427425,
"grad_norm": 11.6875,
"learning_rate": 1.670259114881908e-05,
"loss": 1.1826,
"step": 720
},
{
"epoch": 0.16739279981655583,
"grad_norm": 11.625,
"learning_rate": 1.6656730107773448e-05,
"loss": 1.1712,
"step": 730
},
{
"epoch": 0.16968585186883742,
"grad_norm": 10.4375,
"learning_rate": 1.6610869066727818e-05,
"loss": 1.1701,
"step": 740
},
{
"epoch": 0.171978903921119,
"grad_norm": 9.6875,
"learning_rate": 1.6565008025682185e-05,
"loss": 1.1905,
"step": 750
},
{
"epoch": 0.1742719559734006,
"grad_norm": 12.25,
"learning_rate": 1.651914698463655e-05,
"loss": 1.1887,
"step": 760
},
{
"epoch": 0.17656500802568217,
"grad_norm": 7.9375,
"learning_rate": 1.647328594359092e-05,
"loss": 1.2096,
"step": 770
},
{
"epoch": 0.17885806007796376,
"grad_norm": 13.375,
"learning_rate": 1.6427424902545288e-05,
"loss": 1.1932,
"step": 780
},
{
"epoch": 0.18115111213024535,
"grad_norm": 10.8125,
"learning_rate": 1.6381563861499658e-05,
"loss": 1.1798,
"step": 790
},
{
"epoch": 0.18344416418252693,
"grad_norm": 12.625,
"learning_rate": 1.6335702820454025e-05,
"loss": 1.178,
"step": 800
},
{
"epoch": 0.18573721623480852,
"grad_norm": 12.3125,
"learning_rate": 1.628984177940839e-05,
"loss": 1.1875,
"step": 810
},
{
"epoch": 0.18803026828709013,
"grad_norm": 9.3125,
"learning_rate": 1.624398073836276e-05,
"loss": 1.1791,
"step": 820
},
{
"epoch": 0.19032332033937172,
"grad_norm": 14.6875,
"learning_rate": 1.6198119697317128e-05,
"loss": 1.1764,
"step": 830
},
{
"epoch": 0.1926163723916533,
"grad_norm": 10.625,
"learning_rate": 1.6152258656271498e-05,
"loss": 1.1783,
"step": 840
},
{
"epoch": 0.1949094244439349,
"grad_norm": 12.875,
"learning_rate": 1.6106397615225865e-05,
"loss": 1.1882,
"step": 850
},
{
"epoch": 0.19720247649621647,
"grad_norm": 14.9375,
"learning_rate": 1.6060536574180235e-05,
"loss": 1.1877,
"step": 860
},
{
"epoch": 0.19949552854849806,
"grad_norm": 9.75,
"learning_rate": 1.60146755331346e-05,
"loss": 1.1942,
"step": 870
},
{
"epoch": 0.20178858060077964,
"grad_norm": 12.1875,
"learning_rate": 1.596881449208897e-05,
"loss": 1.1855,
"step": 880
},
{
"epoch": 0.20408163265306123,
"grad_norm": 14.4375,
"learning_rate": 1.5922953451043338e-05,
"loss": 1.1707,
"step": 890
},
{
"epoch": 0.20637468470534281,
"grad_norm": 12.5625,
"learning_rate": 1.5877092409997708e-05,
"loss": 1.1809,
"step": 900
},
{
"epoch": 0.2086677367576244,
"grad_norm": 11.0,
"learning_rate": 1.5831231368952075e-05,
"loss": 1.1929,
"step": 910
},
{
"epoch": 0.21096078880990599,
"grad_norm": 7.4375,
"learning_rate": 1.5785370327906445e-05,
"loss": 1.2096,
"step": 920
},
{
"epoch": 0.21325384086218757,
"grad_norm": 7.0,
"learning_rate": 1.573950928686081e-05,
"loss": 1.1921,
"step": 930
},
{
"epoch": 0.21554689291446916,
"grad_norm": 10.9375,
"learning_rate": 1.569364824581518e-05,
"loss": 1.1849,
"step": 940
},
{
"epoch": 0.21783994496675074,
"grad_norm": 11.25,
"learning_rate": 1.5647787204769548e-05,
"loss": 1.1984,
"step": 950
},
{
"epoch": 0.22013299701903233,
"grad_norm": 12.0625,
"learning_rate": 1.5601926163723918e-05,
"loss": 1.1942,
"step": 960
},
{
"epoch": 0.2224260490713139,
"grad_norm": 10.4375,
"learning_rate": 1.5556065122678285e-05,
"loss": 1.1818,
"step": 970
},
{
"epoch": 0.2247191011235955,
"grad_norm": 12.8125,
"learning_rate": 1.5510204081632655e-05,
"loss": 1.1945,
"step": 980
},
{
"epoch": 0.22701215317587708,
"grad_norm": 6.875,
"learning_rate": 1.546434304058702e-05,
"loss": 1.1643,
"step": 990
},
{
"epoch": 0.22930520522815867,
"grad_norm": 13.25,
"learning_rate": 1.541848199954139e-05,
"loss": 1.2059,
"step": 1000
},
{
"epoch": 0.23159825728044026,
"grad_norm": 10.5,
"learning_rate": 1.5372620958495758e-05,
"loss": 1.1918,
"step": 1010
},
{
"epoch": 0.23389130933272184,
"grad_norm": 6.9375,
"learning_rate": 1.5326759917450128e-05,
"loss": 1.1879,
"step": 1020
},
{
"epoch": 0.23618436138500343,
"grad_norm": 9.25,
"learning_rate": 1.5280898876404495e-05,
"loss": 1.1619,
"step": 1030
},
{
"epoch": 0.238477413437285,
"grad_norm": 8.4375,
"learning_rate": 1.5235037835358863e-05,
"loss": 1.1709,
"step": 1040
},
{
"epoch": 0.24077046548956663,
"grad_norm": 12.625,
"learning_rate": 1.5189176794313232e-05,
"loss": 1.1922,
"step": 1050
},
{
"epoch": 0.2430635175418482,
"grad_norm": 8.0625,
"learning_rate": 1.51433157532676e-05,
"loss": 1.1695,
"step": 1060
},
{
"epoch": 0.2453565695941298,
"grad_norm": 11.3125,
"learning_rate": 1.5097454712221968e-05,
"loss": 1.1987,
"step": 1070
},
{
"epoch": 0.24764962164641138,
"grad_norm": 9.0625,
"learning_rate": 1.5051593671176337e-05,
"loss": 1.1807,
"step": 1080
},
{
"epoch": 0.24994267369869297,
"grad_norm": 12.4375,
"learning_rate": 1.5005732630130705e-05,
"loss": 1.1839,
"step": 1090
},
{
"epoch": 0.2522357257509745,
"grad_norm": 13.0,
"learning_rate": 1.4959871589085073e-05,
"loss": 1.1795,
"step": 1100
},
{
"epoch": 0.2545287778032561,
"grad_norm": 13.125,
"learning_rate": 1.4914010548039442e-05,
"loss": 1.1739,
"step": 1110
},
{
"epoch": 0.2568218298555377,
"grad_norm": 12.3125,
"learning_rate": 1.486814950699381e-05,
"loss": 1.1936,
"step": 1120
},
{
"epoch": 0.2591148819078193,
"grad_norm": 7.03125,
"learning_rate": 1.4822288465948178e-05,
"loss": 1.1775,
"step": 1130
},
{
"epoch": 0.26140793396010087,
"grad_norm": 14.75,
"learning_rate": 1.4776427424902547e-05,
"loss": 1.1958,
"step": 1140
},
{
"epoch": 0.26370098601238245,
"grad_norm": 11.375,
"learning_rate": 1.4730566383856915e-05,
"loss": 1.18,
"step": 1150
},
{
"epoch": 0.2659940380646641,
"grad_norm": 12.3125,
"learning_rate": 1.4684705342811283e-05,
"loss": 1.1812,
"step": 1160
},
{
"epoch": 0.2682870901169457,
"grad_norm": 11.1875,
"learning_rate": 1.4638844301765652e-05,
"loss": 1.1657,
"step": 1170
},
{
"epoch": 0.27058014216922727,
"grad_norm": 11.9375,
"learning_rate": 1.459298326072002e-05,
"loss": 1.1862,
"step": 1180
},
{
"epoch": 0.27287319422150885,
"grad_norm": 9.9375,
"learning_rate": 1.4547122219674388e-05,
"loss": 1.1869,
"step": 1190
},
{
"epoch": 0.27516624627379044,
"grad_norm": 8.9375,
"learning_rate": 1.4501261178628757e-05,
"loss": 1.1881,
"step": 1200
},
{
"epoch": 0.277459298326072,
"grad_norm": 11.625,
"learning_rate": 1.4455400137583125e-05,
"loss": 1.1951,
"step": 1210
},
{
"epoch": 0.2797523503783536,
"grad_norm": 9.375,
"learning_rate": 1.4409539096537493e-05,
"loss": 1.1654,
"step": 1220
},
{
"epoch": 0.2820454024306352,
"grad_norm": 10.6875,
"learning_rate": 1.436367805549186e-05,
"loss": 1.1646,
"step": 1230
},
{
"epoch": 0.2843384544829168,
"grad_norm": 9.125,
"learning_rate": 1.4317817014446228e-05,
"loss": 1.1662,
"step": 1240
},
{
"epoch": 0.28663150653519837,
"grad_norm": 10.8125,
"learning_rate": 1.4271955973400597e-05,
"loss": 1.1961,
"step": 1250
},
{
"epoch": 0.28892455858747995,
"grad_norm": 8.6875,
"learning_rate": 1.4226094932354965e-05,
"loss": 1.179,
"step": 1260
},
{
"epoch": 0.29121761063976154,
"grad_norm": 11.0625,
"learning_rate": 1.4180233891309333e-05,
"loss": 1.2015,
"step": 1270
},
{
"epoch": 0.2935106626920431,
"grad_norm": 8.25,
"learning_rate": 1.4134372850263702e-05,
"loss": 1.176,
"step": 1280
},
{
"epoch": 0.2958037147443247,
"grad_norm": 8.25,
"learning_rate": 1.408851180921807e-05,
"loss": 1.1626,
"step": 1290
},
{
"epoch": 0.2980967667966063,
"grad_norm": 8.8125,
"learning_rate": 1.4042650768172439e-05,
"loss": 1.178,
"step": 1300
},
{
"epoch": 0.3003898188488879,
"grad_norm": 11.25,
"learning_rate": 1.3996789727126807e-05,
"loss": 1.1763,
"step": 1310
},
{
"epoch": 0.30268287090116947,
"grad_norm": 13.0,
"learning_rate": 1.3950928686081175e-05,
"loss": 1.1889,
"step": 1320
},
{
"epoch": 0.30497592295345105,
"grad_norm": 9.8125,
"learning_rate": 1.3905067645035544e-05,
"loss": 1.2124,
"step": 1330
},
{
"epoch": 0.30726897500573264,
"grad_norm": 8.375,
"learning_rate": 1.3859206603989912e-05,
"loss": 1.1906,
"step": 1340
},
{
"epoch": 0.3095620270580142,
"grad_norm": 7.3125,
"learning_rate": 1.381334556294428e-05,
"loss": 1.1758,
"step": 1350
},
{
"epoch": 0.3118550791102958,
"grad_norm": 12.8125,
"learning_rate": 1.3767484521898649e-05,
"loss": 1.2033,
"step": 1360
},
{
"epoch": 0.3141481311625774,
"grad_norm": 9.0625,
"learning_rate": 1.3721623480853017e-05,
"loss": 1.1986,
"step": 1370
},
{
"epoch": 0.316441183214859,
"grad_norm": 11.75,
"learning_rate": 1.3675762439807385e-05,
"loss": 1.17,
"step": 1380
},
{
"epoch": 0.31873423526714056,
"grad_norm": 11.875,
"learning_rate": 1.3629901398761754e-05,
"loss": 1.1687,
"step": 1390
},
{
"epoch": 0.32102728731942215,
"grad_norm": 7.375,
"learning_rate": 1.3584040357716122e-05,
"loss": 1.1815,
"step": 1400
},
{
"epoch": 0.32332033937170374,
"grad_norm": 8.1875,
"learning_rate": 1.353817931667049e-05,
"loss": 1.1864,
"step": 1410
},
{
"epoch": 0.3256133914239853,
"grad_norm": 14.125,
"learning_rate": 1.3492318275624859e-05,
"loss": 1.1832,
"step": 1420
},
{
"epoch": 0.3279064434762669,
"grad_norm": 7.59375,
"learning_rate": 1.3446457234579227e-05,
"loss": 1.189,
"step": 1430
},
{
"epoch": 0.3301994955285485,
"grad_norm": 10.8125,
"learning_rate": 1.3400596193533595e-05,
"loss": 1.1734,
"step": 1440
},
{
"epoch": 0.3324925475808301,
"grad_norm": 10.6875,
"learning_rate": 1.3354735152487964e-05,
"loss": 1.1807,
"step": 1450
},
{
"epoch": 0.33478559963311166,
"grad_norm": 10.375,
"learning_rate": 1.3308874111442332e-05,
"loss": 1.1752,
"step": 1460
},
{
"epoch": 0.33707865168539325,
"grad_norm": 11.375,
"learning_rate": 1.32630130703967e-05,
"loss": 1.1578,
"step": 1470
},
{
"epoch": 0.33937170373767483,
"grad_norm": 10.0,
"learning_rate": 1.3217152029351069e-05,
"loss": 1.177,
"step": 1480
},
{
"epoch": 0.3416647557899564,
"grad_norm": 12.4375,
"learning_rate": 1.3171290988305437e-05,
"loss": 1.1644,
"step": 1490
},
{
"epoch": 0.343957807842238,
"grad_norm": 9.625,
"learning_rate": 1.3125429947259805e-05,
"loss": 1.167,
"step": 1500
},
{
"epoch": 0.3462508598945196,
"grad_norm": 11.0,
"learning_rate": 1.3079568906214172e-05,
"loss": 1.1817,
"step": 1510
},
{
"epoch": 0.3485439119468012,
"grad_norm": 8.1875,
"learning_rate": 1.303370786516854e-05,
"loss": 1.1822,
"step": 1520
},
{
"epoch": 0.35083696399908276,
"grad_norm": 8.625,
"learning_rate": 1.2987846824122909e-05,
"loss": 1.1676,
"step": 1530
},
{
"epoch": 0.35313001605136435,
"grad_norm": 6.71875,
"learning_rate": 1.2941985783077277e-05,
"loss": 1.1792,
"step": 1540
},
{
"epoch": 0.35542306810364593,
"grad_norm": 9.0625,
"learning_rate": 1.2896124742031645e-05,
"loss": 1.1726,
"step": 1550
},
{
"epoch": 0.3577161201559275,
"grad_norm": 10.1875,
"learning_rate": 1.2850263700986014e-05,
"loss": 1.1728,
"step": 1560
},
{
"epoch": 0.3600091722082091,
"grad_norm": 12.3125,
"learning_rate": 1.2804402659940382e-05,
"loss": 1.1825,
"step": 1570
},
{
"epoch": 0.3623022242604907,
"grad_norm": 7.125,
"learning_rate": 1.275854161889475e-05,
"loss": 1.159,
"step": 1580
},
{
"epoch": 0.3645952763127723,
"grad_norm": 8.8125,
"learning_rate": 1.2712680577849119e-05,
"loss": 1.1728,
"step": 1590
},
{
"epoch": 0.36688832836505386,
"grad_norm": 11.6875,
"learning_rate": 1.2666819536803487e-05,
"loss": 1.1976,
"step": 1600
},
{
"epoch": 0.36918138041733545,
"grad_norm": 9.875,
"learning_rate": 1.2620958495757855e-05,
"loss": 1.1806,
"step": 1610
},
{
"epoch": 0.37147443246961703,
"grad_norm": 9.8125,
"learning_rate": 1.2575097454712224e-05,
"loss": 1.1829,
"step": 1620
},
{
"epoch": 0.3737674845218987,
"grad_norm": 12.625,
"learning_rate": 1.2529236413666592e-05,
"loss": 1.169,
"step": 1630
},
{
"epoch": 0.37606053657418026,
"grad_norm": 7.625,
"learning_rate": 1.248337537262096e-05,
"loss": 1.1923,
"step": 1640
},
{
"epoch": 0.37835358862646185,
"grad_norm": 10.75,
"learning_rate": 1.2437514331575329e-05,
"loss": 1.1879,
"step": 1650
},
{
"epoch": 0.38064664067874343,
"grad_norm": 8.875,
"learning_rate": 1.2391653290529697e-05,
"loss": 1.1742,
"step": 1660
},
{
"epoch": 0.382939692731025,
"grad_norm": 10.5,
"learning_rate": 1.2345792249484066e-05,
"loss": 1.1725,
"step": 1670
},
{
"epoch": 0.3852327447833066,
"grad_norm": 8.9375,
"learning_rate": 1.2299931208438432e-05,
"loss": 1.1836,
"step": 1680
},
{
"epoch": 0.3875257968355882,
"grad_norm": 8.5,
"learning_rate": 1.22540701673928e-05,
"loss": 1.1698,
"step": 1690
},
{
"epoch": 0.3898188488878698,
"grad_norm": 10.4375,
"learning_rate": 1.2208209126347169e-05,
"loss": 1.1747,
"step": 1700
},
{
"epoch": 0.39211190094015136,
"grad_norm": 11.375,
"learning_rate": 1.2162348085301536e-05,
"loss": 1.1947,
"step": 1710
},
{
"epoch": 0.39440495299243294,
"grad_norm": 9.0625,
"learning_rate": 1.2116487044255904e-05,
"loss": 1.1847,
"step": 1720
},
{
"epoch": 0.39669800504471453,
"grad_norm": 12.0,
"learning_rate": 1.2070626003210272e-05,
"loss": 1.1897,
"step": 1730
},
{
"epoch": 0.3989910570969961,
"grad_norm": 10.6875,
"learning_rate": 1.202476496216464e-05,
"loss": 1.1668,
"step": 1740
},
{
"epoch": 0.4012841091492777,
"grad_norm": 6.8125,
"learning_rate": 1.1978903921119009e-05,
"loss": 1.1617,
"step": 1750
},
{
"epoch": 0.4035771612015593,
"grad_norm": 14.1875,
"learning_rate": 1.1933042880073377e-05,
"loss": 1.1748,
"step": 1760
},
{
"epoch": 0.4058702132538409,
"grad_norm": 11.1875,
"learning_rate": 1.1887181839027746e-05,
"loss": 1.1836,
"step": 1770
},
{
"epoch": 0.40816326530612246,
"grad_norm": 7.90625,
"learning_rate": 1.1841320797982114e-05,
"loss": 1.1731,
"step": 1780
},
{
"epoch": 0.41045631735840404,
"grad_norm": 13.6875,
"learning_rate": 1.1795459756936482e-05,
"loss": 1.1769,
"step": 1790
},
{
"epoch": 0.41274936941068563,
"grad_norm": 10.75,
"learning_rate": 1.174959871589085e-05,
"loss": 1.181,
"step": 1800
},
{
"epoch": 0.4150424214629672,
"grad_norm": 7.28125,
"learning_rate": 1.1703737674845219e-05,
"loss": 1.1851,
"step": 1810
},
{
"epoch": 0.4173354735152488,
"grad_norm": 11.25,
"learning_rate": 1.1657876633799587e-05,
"loss": 1.1791,
"step": 1820
},
{
"epoch": 0.4196285255675304,
"grad_norm": 8.6875,
"learning_rate": 1.1612015592753956e-05,
"loss": 1.1773,
"step": 1830
},
{
"epoch": 0.42192157761981197,
"grad_norm": 11.3125,
"learning_rate": 1.1566154551708324e-05,
"loss": 1.1644,
"step": 1840
},
{
"epoch": 0.42421462967209356,
"grad_norm": 10.0,
"learning_rate": 1.1520293510662692e-05,
"loss": 1.1703,
"step": 1850
},
{
"epoch": 0.42650768172437514,
"grad_norm": 9.4375,
"learning_rate": 1.147443246961706e-05,
"loss": 1.1696,
"step": 1860
},
{
"epoch": 0.42880073377665673,
"grad_norm": 11.625,
"learning_rate": 1.1428571428571429e-05,
"loss": 1.1809,
"step": 1870
},
{
"epoch": 0.4310937858289383,
"grad_norm": 10.0,
"learning_rate": 1.1382710387525797e-05,
"loss": 1.1865,
"step": 1880
},
{
"epoch": 0.4333868378812199,
"grad_norm": 13.0625,
"learning_rate": 1.1336849346480166e-05,
"loss": 1.1864,
"step": 1890
},
{
"epoch": 0.4356798899335015,
"grad_norm": 9.1875,
"learning_rate": 1.1290988305434534e-05,
"loss": 1.1812,
"step": 1900
},
{
"epoch": 0.43797294198578307,
"grad_norm": 11.1875,
"learning_rate": 1.1245127264388902e-05,
"loss": 1.1623,
"step": 1910
},
{
"epoch": 0.44026599403806466,
"grad_norm": 8.9375,
"learning_rate": 1.119926622334327e-05,
"loss": 1.1655,
"step": 1920
},
{
"epoch": 0.44255904609034624,
"grad_norm": 9.4375,
"learning_rate": 1.1153405182297639e-05,
"loss": 1.1528,
"step": 1930
},
{
"epoch": 0.4448520981426278,
"grad_norm": 12.4375,
"learning_rate": 1.1107544141252007e-05,
"loss": 1.1836,
"step": 1940
},
{
"epoch": 0.4471451501949094,
"grad_norm": 10.375,
"learning_rate": 1.1061683100206376e-05,
"loss": 1.1571,
"step": 1950
},
{
"epoch": 0.449438202247191,
"grad_norm": 11.8125,
"learning_rate": 1.1015822059160744e-05,
"loss": 1.1732,
"step": 1960
},
{
"epoch": 0.4517312542994726,
"grad_norm": 14.3125,
"learning_rate": 1.0969961018115112e-05,
"loss": 1.1716,
"step": 1970
},
{
"epoch": 0.45402430635175417,
"grad_norm": 14.1875,
"learning_rate": 1.092409997706948e-05,
"loss": 1.1583,
"step": 1980
},
{
"epoch": 0.45631735840403576,
"grad_norm": 10.125,
"learning_rate": 1.0878238936023847e-05,
"loss": 1.1839,
"step": 1990
},
{
"epoch": 0.45861041045631734,
"grad_norm": 14.4375,
"learning_rate": 1.0832377894978216e-05,
"loss": 1.1666,
"step": 2000
},
{
"epoch": 0.4609034625085989,
"grad_norm": 13.25,
"learning_rate": 1.0786516853932584e-05,
"loss": 1.1816,
"step": 2010
},
{
"epoch": 0.4631965145608805,
"grad_norm": 9.75,
"learning_rate": 1.0740655812886952e-05,
"loss": 1.1702,
"step": 2020
},
{
"epoch": 0.4654895666131621,
"grad_norm": 9.25,
"learning_rate": 1.069479477184132e-05,
"loss": 1.1636,
"step": 2030
},
{
"epoch": 0.4677826186654437,
"grad_norm": 7.9375,
"learning_rate": 1.064893373079569e-05,
"loss": 1.1645,
"step": 2040
},
{
"epoch": 0.47007567071772527,
"grad_norm": 12.0625,
"learning_rate": 1.0603072689750058e-05,
"loss": 1.169,
"step": 2050
},
{
"epoch": 0.47236872277000685,
"grad_norm": 11.5625,
"learning_rate": 1.0557211648704426e-05,
"loss": 1.1595,
"step": 2060
},
{
"epoch": 0.47466177482228844,
"grad_norm": 15.0,
"learning_rate": 1.0511350607658794e-05,
"loss": 1.16,
"step": 2070
},
{
"epoch": 0.47695482687457,
"grad_norm": 9.25,
"learning_rate": 1.0465489566613163e-05,
"loss": 1.1707,
"step": 2080
},
{
"epoch": 0.47924787892685167,
"grad_norm": 9.75,
"learning_rate": 1.0419628525567531e-05,
"loss": 1.152,
"step": 2090
},
{
"epoch": 0.48154093097913325,
"grad_norm": 12.125,
"learning_rate": 1.03737674845219e-05,
"loss": 1.1793,
"step": 2100
},
{
"epoch": 0.48383398303141484,
"grad_norm": 12.8125,
"learning_rate": 1.0327906443476268e-05,
"loss": 1.1921,
"step": 2110
},
{
"epoch": 0.4861270350836964,
"grad_norm": 13.9375,
"learning_rate": 1.0282045402430636e-05,
"loss": 1.1683,
"step": 2120
},
{
"epoch": 0.488420087135978,
"grad_norm": 11.0625,
"learning_rate": 1.0236184361385004e-05,
"loss": 1.1755,
"step": 2130
},
{
"epoch": 0.4907131391882596,
"grad_norm": 11.375,
"learning_rate": 1.0190323320339373e-05,
"loss": 1.1825,
"step": 2140
},
{
"epoch": 0.4930061912405412,
"grad_norm": 9.8125,
"learning_rate": 1.0144462279293741e-05,
"loss": 1.1531,
"step": 2150
},
{
"epoch": 0.49529924329282277,
"grad_norm": 11.375,
"learning_rate": 1.009860123824811e-05,
"loss": 1.1867,
"step": 2160
},
{
"epoch": 0.49759229534510435,
"grad_norm": 8.625,
"learning_rate": 1.0052740197202478e-05,
"loss": 1.1564,
"step": 2170
},
{
"epoch": 0.49988534739738594,
"grad_norm": 6.46875,
"learning_rate": 1.0006879156156846e-05,
"loss": 1.1859,
"step": 2180
},
{
"epoch": 0.5021783994496675,
"grad_norm": 15.25,
"learning_rate": 9.961018115111214e-06,
"loss": 1.1779,
"step": 2190
},
{
"epoch": 0.504471451501949,
"grad_norm": 9.5,
"learning_rate": 9.915157074065583e-06,
"loss": 1.1663,
"step": 2200
},
{
"epoch": 0.5067645035542306,
"grad_norm": 9.5,
"learning_rate": 9.869296033019951e-06,
"loss": 1.1526,
"step": 2210
},
{
"epoch": 0.5090575556065122,
"grad_norm": 7.6875,
"learning_rate": 9.82343499197432e-06,
"loss": 1.179,
"step": 2220
},
{
"epoch": 0.5113506076587938,
"grad_norm": 11.125,
"learning_rate": 9.777573950928688e-06,
"loss": 1.1688,
"step": 2230
},
{
"epoch": 0.5136436597110754,
"grad_norm": 10.1875,
"learning_rate": 9.731712909883056e-06,
"loss": 1.159,
"step": 2240
},
{
"epoch": 0.515936711763357,
"grad_norm": 13.4375,
"learning_rate": 9.685851868837424e-06,
"loss": 1.153,
"step": 2250
},
{
"epoch": 0.5182297638156386,
"grad_norm": 12.875,
"learning_rate": 9.639990827791793e-06,
"loss": 1.1774,
"step": 2260
},
{
"epoch": 0.5205228158679202,
"grad_norm": 9.9375,
"learning_rate": 9.594129786746161e-06,
"loss": 1.1825,
"step": 2270
},
{
"epoch": 0.5228158679202017,
"grad_norm": 6.375,
"learning_rate": 9.548268745700528e-06,
"loss": 1.1784,
"step": 2280
},
{
"epoch": 0.5251089199724833,
"grad_norm": 14.375,
"learning_rate": 9.502407704654896e-06,
"loss": 1.1542,
"step": 2290
},
{
"epoch": 0.5274019720247649,
"grad_norm": 13.0,
"learning_rate": 9.456546663609264e-06,
"loss": 1.1705,
"step": 2300
},
{
"epoch": 0.5296950240770465,
"grad_norm": 15.6875,
"learning_rate": 9.410685622563633e-06,
"loss": 1.1906,
"step": 2310
},
{
"epoch": 0.5319880761293282,
"grad_norm": 9.1875,
"learning_rate": 9.364824581518001e-06,
"loss": 1.1732,
"step": 2320
},
{
"epoch": 0.5342811281816098,
"grad_norm": 11.4375,
"learning_rate": 9.31896354047237e-06,
"loss": 1.1994,
"step": 2330
},
{
"epoch": 0.5365741802338914,
"grad_norm": 9.8125,
"learning_rate": 9.273102499426738e-06,
"loss": 1.1633,
"step": 2340
},
{
"epoch": 0.538867232286173,
"grad_norm": 8.9375,
"learning_rate": 9.227241458381106e-06,
"loss": 1.1683,
"step": 2350
},
{
"epoch": 0.5411602843384545,
"grad_norm": 12.0,
"learning_rate": 9.181380417335474e-06,
"loss": 1.1565,
"step": 2360
},
{
"epoch": 0.5434533363907361,
"grad_norm": 10.0625,
"learning_rate": 9.135519376289841e-06,
"loss": 1.1726,
"step": 2370
},
{
"epoch": 0.5457463884430177,
"grad_norm": 13.9375,
"learning_rate": 9.08965833524421e-06,
"loss": 1.1582,
"step": 2380
},
{
"epoch": 0.5480394404952993,
"grad_norm": 7.90625,
"learning_rate": 9.043797294198578e-06,
"loss": 1.1647,
"step": 2390
},
{
"epoch": 0.5503324925475809,
"grad_norm": 16.0,
"learning_rate": 8.997936253152946e-06,
"loss": 1.1838,
"step": 2400
},
{
"epoch": 0.5526255445998625,
"grad_norm": 11.75,
"learning_rate": 8.952075212107315e-06,
"loss": 1.1699,
"step": 2410
},
{
"epoch": 0.554918596652144,
"grad_norm": 12.3125,
"learning_rate": 8.906214171061683e-06,
"loss": 1.1693,
"step": 2420
},
{
"epoch": 0.5572116487044256,
"grad_norm": 8.625,
"learning_rate": 8.860353130016051e-06,
"loss": 1.1793,
"step": 2430
},
{
"epoch": 0.5595047007567072,
"grad_norm": 10.1875,
"learning_rate": 8.81449208897042e-06,
"loss": 1.1799,
"step": 2440
},
{
"epoch": 0.5617977528089888,
"grad_norm": 7.625,
"learning_rate": 8.768631047924788e-06,
"loss": 1.1503,
"step": 2450
},
{
"epoch": 0.5640908048612704,
"grad_norm": 11.125,
"learning_rate": 8.722770006879156e-06,
"loss": 1.178,
"step": 2460
},
{
"epoch": 0.566383856913552,
"grad_norm": 11.375,
"learning_rate": 8.676908965833525e-06,
"loss": 1.149,
"step": 2470
},
{
"epoch": 0.5686769089658336,
"grad_norm": 12.1875,
"learning_rate": 8.631047924787893e-06,
"loss": 1.1781,
"step": 2480
},
{
"epoch": 0.5709699610181151,
"grad_norm": 11.4375,
"learning_rate": 8.585186883742261e-06,
"loss": 1.1664,
"step": 2490
},
{
"epoch": 0.5732630130703967,
"grad_norm": 9.625,
"learning_rate": 8.53932584269663e-06,
"loss": 1.1448,
"step": 2500
},
{
"epoch": 0.5755560651226783,
"grad_norm": 11.0625,
"learning_rate": 8.493464801650998e-06,
"loss": 1.1537,
"step": 2510
},
{
"epoch": 0.5778491171749599,
"grad_norm": 10.1875,
"learning_rate": 8.447603760605366e-06,
"loss": 1.1583,
"step": 2520
},
{
"epoch": 0.5801421692272415,
"grad_norm": 9.25,
"learning_rate": 8.401742719559735e-06,
"loss": 1.1794,
"step": 2530
},
{
"epoch": 0.5824352212795231,
"grad_norm": 10.8125,
"learning_rate": 8.355881678514103e-06,
"loss": 1.1558,
"step": 2540
},
{
"epoch": 0.5847282733318047,
"grad_norm": 9.9375,
"learning_rate": 8.310020637468471e-06,
"loss": 1.1693,
"step": 2550
},
{
"epoch": 0.5870213253840862,
"grad_norm": 13.6875,
"learning_rate": 8.26415959642284e-06,
"loss": 1.1685,
"step": 2560
},
{
"epoch": 0.5893143774363678,
"grad_norm": 9.1875,
"learning_rate": 8.218298555377208e-06,
"loss": 1.1697,
"step": 2570
},
{
"epoch": 0.5916074294886494,
"grad_norm": 15.0,
"learning_rate": 8.172437514331576e-06,
"loss": 1.1509,
"step": 2580
},
{
"epoch": 0.593900481540931,
"grad_norm": 11.75,
"learning_rate": 8.126576473285945e-06,
"loss": 1.1541,
"step": 2590
},
{
"epoch": 0.5961935335932126,
"grad_norm": 10.375,
"learning_rate": 8.080715432240313e-06,
"loss": 1.1517,
"step": 2600
},
{
"epoch": 0.5984865856454942,
"grad_norm": 13.3125,
"learning_rate": 8.034854391194681e-06,
"loss": 1.178,
"step": 2610
},
{
"epoch": 0.6007796376977758,
"grad_norm": 12.6875,
"learning_rate": 7.98899335014905e-06,
"loss": 1.159,
"step": 2620
},
{
"epoch": 0.6030726897500573,
"grad_norm": 12.75,
"learning_rate": 7.943132309103418e-06,
"loss": 1.1604,
"step": 2630
},
{
"epoch": 0.6053657418023389,
"grad_norm": 7.0625,
"learning_rate": 7.897271268057786e-06,
"loss": 1.1696,
"step": 2640
},
{
"epoch": 0.6076587938546205,
"grad_norm": 13.0625,
"learning_rate": 7.851410227012155e-06,
"loss": 1.1713,
"step": 2650
},
{
"epoch": 0.6099518459069021,
"grad_norm": 8.8125,
"learning_rate": 7.805549185966521e-06,
"loss": 1.1678,
"step": 2660
},
{
"epoch": 0.6122448979591837,
"grad_norm": 12.75,
"learning_rate": 7.75968814492089e-06,
"loss": 1.168,
"step": 2670
},
{
"epoch": 0.6145379500114653,
"grad_norm": 7.09375,
"learning_rate": 7.713827103875258e-06,
"loss": 1.1519,
"step": 2680
},
{
"epoch": 0.6168310020637469,
"grad_norm": 10.375,
"learning_rate": 7.667966062829626e-06,
"loss": 1.1651,
"step": 2690
},
{
"epoch": 0.6191240541160284,
"grad_norm": 11.25,
"learning_rate": 7.622105021783996e-06,
"loss": 1.1303,
"step": 2700
},
{
"epoch": 0.62141710616831,
"grad_norm": 13.1875,
"learning_rate": 7.576243980738364e-06,
"loss": 1.1462,
"step": 2710
},
{
"epoch": 0.6237101582205916,
"grad_norm": 10.0,
"learning_rate": 7.530382939692732e-06,
"loss": 1.1786,
"step": 2720
},
{
"epoch": 0.6260032102728732,
"grad_norm": 11.6875,
"learning_rate": 7.4845218986471e-06,
"loss": 1.179,
"step": 2730
},
{
"epoch": 0.6282962623251548,
"grad_norm": 12.3125,
"learning_rate": 7.438660857601468e-06,
"loss": 1.1708,
"step": 2740
},
{
"epoch": 0.6305893143774364,
"grad_norm": 11.625,
"learning_rate": 7.3927998165558365e-06,
"loss": 1.1582,
"step": 2750
},
{
"epoch": 0.632882366429718,
"grad_norm": 15.3125,
"learning_rate": 7.346938775510205e-06,
"loss": 1.1622,
"step": 2760
},
{
"epoch": 0.6351754184819995,
"grad_norm": 13.75,
"learning_rate": 7.301077734464573e-06,
"loss": 1.1626,
"step": 2770
},
{
"epoch": 0.6374684705342811,
"grad_norm": 9.0625,
"learning_rate": 7.2552166934189415e-06,
"loss": 1.1477,
"step": 2780
},
{
"epoch": 0.6397615225865627,
"grad_norm": 8.75,
"learning_rate": 7.20935565237331e-06,
"loss": 1.1482,
"step": 2790
},
{
"epoch": 0.6420545746388443,
"grad_norm": 9.375,
"learning_rate": 7.163494611327678e-06,
"loss": 1.1293,
"step": 2800
},
{
"epoch": 0.6443476266911259,
"grad_norm": 10.3125,
"learning_rate": 7.1176335702820466e-06,
"loss": 1.1391,
"step": 2810
},
{
"epoch": 0.6466406787434075,
"grad_norm": 12.9375,
"learning_rate": 7.071772529236415e-06,
"loss": 1.153,
"step": 2820
},
{
"epoch": 0.6489337307956891,
"grad_norm": 16.875,
"learning_rate": 7.025911488190783e-06,
"loss": 1.1551,
"step": 2830
},
{
"epoch": 0.6512267828479706,
"grad_norm": 12.0625,
"learning_rate": 6.98005044714515e-06,
"loss": 1.1548,
"step": 2840
},
{
"epoch": 0.6535198349002522,
"grad_norm": 14.0625,
"learning_rate": 6.934189406099518e-06,
"loss": 1.1771,
"step": 2850
},
{
"epoch": 0.6558128869525338,
"grad_norm": 14.9375,
"learning_rate": 6.888328365053887e-06,
"loss": 1.159,
"step": 2860
},
{
"epoch": 0.6581059390048154,
"grad_norm": 12.0,
"learning_rate": 6.842467324008255e-06,
"loss": 1.1413,
"step": 2870
},
{
"epoch": 0.660398991057097,
"grad_norm": 11.3125,
"learning_rate": 6.796606282962623e-06,
"loss": 1.1717,
"step": 2880
},
{
"epoch": 0.6626920431093786,
"grad_norm": 11.0625,
"learning_rate": 6.750745241916992e-06,
"loss": 1.1267,
"step": 2890
},
{
"epoch": 0.6649850951616602,
"grad_norm": 10.9375,
"learning_rate": 6.70488420087136e-06,
"loss": 1.121,
"step": 2900
},
{
"epoch": 0.6672781472139417,
"grad_norm": 8.6875,
"learning_rate": 6.659023159825728e-06,
"loss": 1.1593,
"step": 2910
},
{
"epoch": 0.6695711992662233,
"grad_norm": 11.0,
"learning_rate": 6.613162118780097e-06,
"loss": 1.1764,
"step": 2920
},
{
"epoch": 0.6718642513185049,
"grad_norm": 8.25,
"learning_rate": 6.567301077734465e-06,
"loss": 1.1719,
"step": 2930
},
{
"epoch": 0.6741573033707865,
"grad_norm": 12.4375,
"learning_rate": 6.521440036688833e-06,
"loss": 1.154,
"step": 2940
},
{
"epoch": 0.6764503554230681,
"grad_norm": 10.1875,
"learning_rate": 6.475578995643202e-06,
"loss": 1.1754,
"step": 2950
},
{
"epoch": 0.6787434074753497,
"grad_norm": 11.75,
"learning_rate": 6.42971795459757e-06,
"loss": 1.1703,
"step": 2960
},
{
"epoch": 0.6810364595276313,
"grad_norm": 9.875,
"learning_rate": 6.3838569135519375e-06,
"loss": 1.1494,
"step": 2970
},
{
"epoch": 0.6833295115799128,
"grad_norm": 13.0625,
"learning_rate": 6.337995872506306e-06,
"loss": 1.1638,
"step": 2980
},
{
"epoch": 0.6856225636321944,
"grad_norm": 10.75,
"learning_rate": 6.292134831460674e-06,
"loss": 1.1403,
"step": 2990
},
{
"epoch": 0.687915615684476,
"grad_norm": 10.9375,
"learning_rate": 6.2462737904150426e-06,
"loss": 1.1696,
"step": 3000
},
{
"epoch": 0.6902086677367576,
"grad_norm": 9.625,
"learning_rate": 6.200412749369411e-06,
"loss": 1.1577,
"step": 3010
},
{
"epoch": 0.6925017197890392,
"grad_norm": 8.4375,
"learning_rate": 6.154551708323779e-06,
"loss": 1.1644,
"step": 3020
},
{
"epoch": 0.6947947718413208,
"grad_norm": 9.4375,
"learning_rate": 6.108690667278148e-06,
"loss": 1.1587,
"step": 3030
},
{
"epoch": 0.6970878238936024,
"grad_norm": 9.9375,
"learning_rate": 6.062829626232516e-06,
"loss": 1.1618,
"step": 3040
},
{
"epoch": 0.6993808759458839,
"grad_norm": 9.5625,
"learning_rate": 6.016968585186884e-06,
"loss": 1.1613,
"step": 3050
},
{
"epoch": 0.7016739279981655,
"grad_norm": 12.6875,
"learning_rate": 5.971107544141253e-06,
"loss": 1.1492,
"step": 3060
},
{
"epoch": 0.7039669800504471,
"grad_norm": 8.875,
"learning_rate": 5.925246503095621e-06,
"loss": 1.1495,
"step": 3070
},
{
"epoch": 0.7062600321027287,
"grad_norm": 12.75,
"learning_rate": 5.879385462049989e-06,
"loss": 1.149,
"step": 3080
},
{
"epoch": 0.7085530841550103,
"grad_norm": 9.8125,
"learning_rate": 5.833524421004358e-06,
"loss": 1.1616,
"step": 3090
},
{
"epoch": 0.7108461362072919,
"grad_norm": 11.3125,
"learning_rate": 5.787663379958726e-06,
"loss": 1.1694,
"step": 3100
},
{
"epoch": 0.7131391882595735,
"grad_norm": 7.90625,
"learning_rate": 5.7418023389130935e-06,
"loss": 1.1384,
"step": 3110
},
{
"epoch": 0.715432240311855,
"grad_norm": 10.0,
"learning_rate": 5.695941297867462e-06,
"loss": 1.156,
"step": 3120
},
{
"epoch": 0.7177252923641366,
"grad_norm": 14.6875,
"learning_rate": 5.65008025682183e-06,
"loss": 1.1736,
"step": 3130
},
{
"epoch": 0.7200183444164182,
"grad_norm": 11.4375,
"learning_rate": 5.6042192157761985e-06,
"loss": 1.1293,
"step": 3140
},
{
"epoch": 0.7223113964686998,
"grad_norm": 12.875,
"learning_rate": 5.558358174730567e-06,
"loss": 1.1067,
"step": 3150
},
{
"epoch": 0.7246044485209814,
"grad_norm": 11.6875,
"learning_rate": 5.512497133684935e-06,
"loss": 1.1516,
"step": 3160
},
{
"epoch": 0.726897500573263,
"grad_norm": 10.5625,
"learning_rate": 5.4666360926393036e-06,
"loss": 1.1532,
"step": 3170
},
{
"epoch": 0.7291905526255446,
"grad_norm": 9.6875,
"learning_rate": 5.420775051593672e-06,
"loss": 1.1647,
"step": 3180
},
{
"epoch": 0.7314836046778261,
"grad_norm": 12.1875,
"learning_rate": 5.37491401054804e-06,
"loss": 1.1914,
"step": 3190
},
{
"epoch": 0.7337766567301077,
"grad_norm": 13.25,
"learning_rate": 5.329052969502409e-06,
"loss": 1.148,
"step": 3200
},
{
"epoch": 0.7360697087823893,
"grad_norm": 8.375,
"learning_rate": 5.283191928456777e-06,
"loss": 1.1612,
"step": 3210
},
{
"epoch": 0.7383627608346709,
"grad_norm": 8.75,
"learning_rate": 5.237330887411145e-06,
"loss": 1.1402,
"step": 3220
},
{
"epoch": 0.7406558128869525,
"grad_norm": 11.5,
"learning_rate": 5.191469846365514e-06,
"loss": 1.1737,
"step": 3230
},
{
"epoch": 0.7429488649392341,
"grad_norm": 11.625,
"learning_rate": 5.145608805319882e-06,
"loss": 1.1854,
"step": 3240
},
{
"epoch": 0.7452419169915158,
"grad_norm": 11.125,
"learning_rate": 5.0997477642742495e-06,
"loss": 1.1611,
"step": 3250
},
{
"epoch": 0.7475349690437973,
"grad_norm": 9.6875,
"learning_rate": 5.053886723228618e-06,
"loss": 1.1477,
"step": 3260
},
{
"epoch": 0.7498280210960789,
"grad_norm": 10.75,
"learning_rate": 5.008025682182986e-06,
"loss": 1.1608,
"step": 3270
},
{
"epoch": 0.7521210731483605,
"grad_norm": 14.125,
"learning_rate": 4.9621646411373545e-06,
"loss": 1.1658,
"step": 3280
},
{
"epoch": 0.7544141252006421,
"grad_norm": 9.5625,
"learning_rate": 4.916303600091723e-06,
"loss": 1.15,
"step": 3290
},
{
"epoch": 0.7567071772529237,
"grad_norm": 13.625,
"learning_rate": 4.87044255904609e-06,
"loss": 1.1616,
"step": 3300
},
{
"epoch": 0.7590002293052053,
"grad_norm": 12.625,
"learning_rate": 4.824581518000459e-06,
"loss": 1.1545,
"step": 3310
},
{
"epoch": 0.7612932813574869,
"grad_norm": 12.75,
"learning_rate": 4.778720476954827e-06,
"loss": 1.1637,
"step": 3320
},
{
"epoch": 0.7635863334097684,
"grad_norm": 10.0625,
"learning_rate": 4.732859435909195e-06,
"loss": 1.1497,
"step": 3330
},
{
"epoch": 0.76587938546205,
"grad_norm": 11.3125,
"learning_rate": 4.686998394863564e-06,
"loss": 1.1739,
"step": 3340
},
{
"epoch": 0.7681724375143316,
"grad_norm": 12.75,
"learning_rate": 4.641137353817932e-06,
"loss": 1.157,
"step": 3350
},
{
"epoch": 0.7704654895666132,
"grad_norm": 7.90625,
"learning_rate": 4.5952763127723e-06,
"loss": 1.1537,
"step": 3360
},
{
"epoch": 0.7727585416188948,
"grad_norm": 10.625,
"learning_rate": 4.549415271726669e-06,
"loss": 1.1519,
"step": 3370
},
{
"epoch": 0.7750515936711764,
"grad_norm": 11.6875,
"learning_rate": 4.503554230681037e-06,
"loss": 1.1483,
"step": 3380
},
{
"epoch": 0.777344645723458,
"grad_norm": 11.75,
"learning_rate": 4.4576931896354054e-06,
"loss": 1.1423,
"step": 3390
},
{
"epoch": 0.7796376977757395,
"grad_norm": 13.9375,
"learning_rate": 4.411832148589774e-06,
"loss": 1.1684,
"step": 3400
},
{
"epoch": 0.7819307498280211,
"grad_norm": 15.9375,
"learning_rate": 4.365971107544142e-06,
"loss": 1.1695,
"step": 3410
},
{
"epoch": 0.7842238018803027,
"grad_norm": 13.625,
"learning_rate": 4.32011006649851e-06,
"loss": 1.1625,
"step": 3420
},
{
"epoch": 0.7865168539325843,
"grad_norm": 13.375,
"learning_rate": 4.274249025452878e-06,
"loss": 1.1753,
"step": 3430
},
{
"epoch": 0.7888099059848659,
"grad_norm": 11.25,
"learning_rate": 4.228387984407246e-06,
"loss": 1.1755,
"step": 3440
},
{
"epoch": 0.7911029580371475,
"grad_norm": 9.375,
"learning_rate": 4.182526943361615e-06,
"loss": 1.1578,
"step": 3450
},
{
"epoch": 0.7933960100894291,
"grad_norm": 9.125,
"learning_rate": 4.136665902315983e-06,
"loss": 1.1278,
"step": 3460
},
{
"epoch": 0.7956890621417106,
"grad_norm": 10.625,
"learning_rate": 4.090804861270351e-06,
"loss": 1.1695,
"step": 3470
},
{
"epoch": 0.7979821141939922,
"grad_norm": 9.8125,
"learning_rate": 4.04494382022472e-06,
"loss": 1.1471,
"step": 3480
},
{
"epoch": 0.8002751662462738,
"grad_norm": 10.5,
"learning_rate": 3.999082779179087e-06,
"loss": 1.1484,
"step": 3490
},
{
"epoch": 0.8025682182985554,
"grad_norm": 14.125,
"learning_rate": 3.9532217381334555e-06,
"loss": 1.173,
"step": 3500
},
{
"epoch": 0.804861270350837,
"grad_norm": 8.125,
"learning_rate": 3.907360697087824e-06,
"loss": 1.1756,
"step": 3510
},
{
"epoch": 0.8071543224031186,
"grad_norm": 9.125,
"learning_rate": 3.861499656042192e-06,
"loss": 1.1492,
"step": 3520
},
{
"epoch": 0.8094473744554002,
"grad_norm": 8.3125,
"learning_rate": 3.8156386149965606e-06,
"loss": 1.1502,
"step": 3530
},
{
"epoch": 0.8117404265076817,
"grad_norm": 9.625,
"learning_rate": 3.769777573950929e-06,
"loss": 1.1442,
"step": 3540
},
{
"epoch": 0.8140334785599633,
"grad_norm": 11.75,
"learning_rate": 3.7239165329052972e-06,
"loss": 1.1612,
"step": 3550
},
{
"epoch": 0.8163265306122449,
"grad_norm": 10.5,
"learning_rate": 3.6780554918596656e-06,
"loss": 1.1628,
"step": 3560
},
{
"epoch": 0.8186195826645265,
"grad_norm": 10.75,
"learning_rate": 3.632194450814034e-06,
"loss": 1.1583,
"step": 3570
},
{
"epoch": 0.8209126347168081,
"grad_norm": 14.5,
"learning_rate": 3.5863334097684023e-06,
"loss": 1.1596,
"step": 3580
},
{
"epoch": 0.8232056867690897,
"grad_norm": 10.0,
"learning_rate": 3.5404723687227706e-06,
"loss": 1.1573,
"step": 3590
},
{
"epoch": 0.8254987388213713,
"grad_norm": 12.4375,
"learning_rate": 3.4946113276771385e-06,
"loss": 1.1575,
"step": 3600
},
{
"epoch": 0.8277917908736528,
"grad_norm": 10.625,
"learning_rate": 3.448750286631507e-06,
"loss": 1.136,
"step": 3610
},
{
"epoch": 0.8300848429259344,
"grad_norm": 10.875,
"learning_rate": 3.4028892455858752e-06,
"loss": 1.1493,
"step": 3620
},
{
"epoch": 0.832377894978216,
"grad_norm": 17.0,
"learning_rate": 3.3570282045402436e-06,
"loss": 1.1689,
"step": 3630
},
{
"epoch": 0.8346709470304976,
"grad_norm": 8.9375,
"learning_rate": 3.311167163494612e-06,
"loss": 1.1406,
"step": 3640
},
{
"epoch": 0.8369639990827792,
"grad_norm": 9.625,
"learning_rate": 3.2653061224489794e-06,
"loss": 1.1438,
"step": 3650
},
{
"epoch": 0.8392570511350608,
"grad_norm": 12.125,
"learning_rate": 3.2194450814033478e-06,
"loss": 1.1721,
"step": 3660
},
{
"epoch": 0.8415501031873424,
"grad_norm": 8.5,
"learning_rate": 3.173584040357716e-06,
"loss": 1.1498,
"step": 3670
},
{
"epoch": 0.8438431552396239,
"grad_norm": 11.625,
"learning_rate": 3.1277229993120844e-06,
"loss": 1.1531,
"step": 3680
},
{
"epoch": 0.8461362072919055,
"grad_norm": 10.0625,
"learning_rate": 3.081861958266453e-06,
"loss": 1.1489,
"step": 3690
},
{
"epoch": 0.8484292593441871,
"grad_norm": 14.375,
"learning_rate": 3.036000917220821e-06,
"loss": 1.133,
"step": 3700
},
{
"epoch": 0.8507223113964687,
"grad_norm": 9.3125,
"learning_rate": 2.9901398761751895e-06,
"loss": 1.1565,
"step": 3710
},
{
"epoch": 0.8530153634487503,
"grad_norm": 10.0,
"learning_rate": 2.9442788351295574e-06,
"loss": 1.162,
"step": 3720
},
{
"epoch": 0.8553084155010319,
"grad_norm": 13.9375,
"learning_rate": 2.8984177940839257e-06,
"loss": 1.1472,
"step": 3730
},
{
"epoch": 0.8576014675533135,
"grad_norm": 11.25,
"learning_rate": 2.852556753038294e-06,
"loss": 1.1293,
"step": 3740
},
{
"epoch": 0.859894519605595,
"grad_norm": 10.4375,
"learning_rate": 2.8066957119926624e-06,
"loss": 1.1667,
"step": 3750
},
{
"epoch": 0.8621875716578766,
"grad_norm": 11.0625,
"learning_rate": 2.7608346709470308e-06,
"loss": 1.148,
"step": 3760
},
{
"epoch": 0.8644806237101582,
"grad_norm": 12.6875,
"learning_rate": 2.714973629901399e-06,
"loss": 1.1476,
"step": 3770
},
{
"epoch": 0.8667736757624398,
"grad_norm": 15.0,
"learning_rate": 2.6691125888557675e-06,
"loss": 1.1556,
"step": 3780
},
{
"epoch": 0.8690667278147214,
"grad_norm": 8.0,
"learning_rate": 2.6232515478101354e-06,
"loss": 1.1458,
"step": 3790
},
{
"epoch": 0.871359779867003,
"grad_norm": 13.9375,
"learning_rate": 2.5773905067645037e-06,
"loss": 1.1335,
"step": 3800
},
{
"epoch": 0.8736528319192846,
"grad_norm": 11.75,
"learning_rate": 2.531529465718872e-06,
"loss": 1.1534,
"step": 3810
},
{
"epoch": 0.8759458839715661,
"grad_norm": 11.9375,
"learning_rate": 2.48566842467324e-06,
"loss": 1.1266,
"step": 3820
},
{
"epoch": 0.8782389360238477,
"grad_norm": 9.875,
"learning_rate": 2.4398073836276083e-06,
"loss": 1.1467,
"step": 3830
},
{
"epoch": 0.8805319880761293,
"grad_norm": 12.1875,
"learning_rate": 2.3939463425819767e-06,
"loss": 1.1495,
"step": 3840
},
{
"epoch": 0.8828250401284109,
"grad_norm": 12.0625,
"learning_rate": 2.348085301536345e-06,
"loss": 1.1471,
"step": 3850
},
{
"epoch": 0.8851180921806925,
"grad_norm": 11.5625,
"learning_rate": 2.3022242604907134e-06,
"loss": 1.143,
"step": 3860
},
{
"epoch": 0.8874111442329741,
"grad_norm": 13.25,
"learning_rate": 2.2563632194450817e-06,
"loss": 1.158,
"step": 3870
},
{
"epoch": 0.8897041962852557,
"grad_norm": 13.4375,
"learning_rate": 2.21050217839945e-06,
"loss": 1.1639,
"step": 3880
},
{
"epoch": 0.8919972483375372,
"grad_norm": 16.375,
"learning_rate": 2.164641137353818e-06,
"loss": 1.1484,
"step": 3890
},
{
"epoch": 0.8942903003898188,
"grad_norm": 13.0625,
"learning_rate": 2.1187800963081863e-06,
"loss": 1.1343,
"step": 3900
},
{
"epoch": 0.8965833524421004,
"grad_norm": 10.0625,
"learning_rate": 2.0729190552625547e-06,
"loss": 1.1346,
"step": 3910
},
{
"epoch": 0.898876404494382,
"grad_norm": 13.0625,
"learning_rate": 2.027058014216923e-06,
"loss": 1.1496,
"step": 3920
},
{
"epoch": 0.9011694565466636,
"grad_norm": 12.1875,
"learning_rate": 1.9811969731712914e-06,
"loss": 1.1345,
"step": 3930
},
{
"epoch": 0.9034625085989452,
"grad_norm": 10.6875,
"learning_rate": 1.9353359321256593e-06,
"loss": 1.1551,
"step": 3940
},
{
"epoch": 0.9057555606512268,
"grad_norm": 10.5625,
"learning_rate": 1.8894748910800276e-06,
"loss": 1.1252,
"step": 3950
},
{
"epoch": 0.9080486127035083,
"grad_norm": 15.0,
"learning_rate": 1.843613850034396e-06,
"loss": 1.1214,
"step": 3960
},
{
"epoch": 0.9103416647557899,
"grad_norm": 11.375,
"learning_rate": 1.797752808988764e-06,
"loss": 1.1846,
"step": 3970
},
{
"epoch": 0.9126347168080715,
"grad_norm": 10.125,
"learning_rate": 1.7518917679431324e-06,
"loss": 1.168,
"step": 3980
},
{
"epoch": 0.9149277688603531,
"grad_norm": 11.0,
"learning_rate": 1.7060307268975008e-06,
"loss": 1.1499,
"step": 3990
},
{
"epoch": 0.9172208209126347,
"grad_norm": 15.0625,
"learning_rate": 1.660169685851869e-06,
"loss": 1.1486,
"step": 4000
},
{
"epoch": 0.9195138729649163,
"grad_norm": 14.875,
"learning_rate": 1.6143086448062373e-06,
"loss": 1.1458,
"step": 4010
},
{
"epoch": 0.9218069250171979,
"grad_norm": 10.25,
"learning_rate": 1.5684476037606056e-06,
"loss": 1.1349,
"step": 4020
},
{
"epoch": 0.9240999770694794,
"grad_norm": 12.75,
"learning_rate": 1.522586562714974e-06,
"loss": 1.1826,
"step": 4030
},
{
"epoch": 0.926393029121761,
"grad_norm": 10.0625,
"learning_rate": 1.476725521669342e-06,
"loss": 1.1549,
"step": 4040
},
{
"epoch": 0.9286860811740426,
"grad_norm": 14.0,
"learning_rate": 1.4308644806237102e-06,
"loss": 1.1659,
"step": 4050
},
{
"epoch": 0.9309791332263242,
"grad_norm": 13.6875,
"learning_rate": 1.3850034395780783e-06,
"loss": 1.1705,
"step": 4060
},
{
"epoch": 0.9332721852786058,
"grad_norm": 9.0,
"learning_rate": 1.3391423985324467e-06,
"loss": 1.1297,
"step": 4070
},
{
"epoch": 0.9355652373308874,
"grad_norm": 11.4375,
"learning_rate": 1.293281357486815e-06,
"loss": 1.1488,
"step": 4080
},
{
"epoch": 0.937858289383169,
"grad_norm": 9.6875,
"learning_rate": 1.2474203164411834e-06,
"loss": 1.1374,
"step": 4090
},
{
"epoch": 0.9401513414354505,
"grad_norm": 10.9375,
"learning_rate": 1.2015592753955515e-06,
"loss": 1.1393,
"step": 4100
},
{
"epoch": 0.9424443934877321,
"grad_norm": 14.5,
"learning_rate": 1.1556982343499199e-06,
"loss": 1.1513,
"step": 4110
},
{
"epoch": 0.9447374455400137,
"grad_norm": 8.9375,
"learning_rate": 1.1098371933042882e-06,
"loss": 1.139,
"step": 4120
},
{
"epoch": 0.9470304975922953,
"grad_norm": 11.5625,
"learning_rate": 1.0639761522586563e-06,
"loss": 1.1548,
"step": 4130
},
{
"epoch": 0.9493235496445769,
"grad_norm": 11.3125,
"learning_rate": 1.0181151112130245e-06,
"loss": 1.1737,
"step": 4140
},
{
"epoch": 0.9516166016968585,
"grad_norm": 13.6875,
"learning_rate": 9.722540701673928e-07,
"loss": 1.1784,
"step": 4150
},
{
"epoch": 0.95390965374914,
"grad_norm": 11.75,
"learning_rate": 9.263930291217611e-07,
"loss": 1.1478,
"step": 4160
},
{
"epoch": 0.9562027058014216,
"grad_norm": 15.6875,
"learning_rate": 8.805319880761294e-07,
"loss": 1.1353,
"step": 4170
},
{
"epoch": 0.9584957578537033,
"grad_norm": 12.625,
"learning_rate": 8.346709470304977e-07,
"loss": 1.145,
"step": 4180
},
{
"epoch": 0.9607888099059849,
"grad_norm": 10.0625,
"learning_rate": 7.88809905984866e-07,
"loss": 1.1473,
"step": 4190
},
{
"epoch": 0.9630818619582665,
"grad_norm": 13.375,
"learning_rate": 7.429488649392341e-07,
"loss": 1.1293,
"step": 4200
},
{
"epoch": 0.9653749140105481,
"grad_norm": 10.0625,
"learning_rate": 6.970878238936024e-07,
"loss": 1.1409,
"step": 4210
},
{
"epoch": 0.9676679660628297,
"grad_norm": 9.625,
"learning_rate": 6.512267828479707e-07,
"loss": 1.1391,
"step": 4220
},
{
"epoch": 0.9699610181151113,
"grad_norm": 9.3125,
"learning_rate": 6.05365741802339e-07,
"loss": 1.1364,
"step": 4230
},
{
"epoch": 0.9722540701673928,
"grad_norm": 11.25,
"learning_rate": 5.595047007567072e-07,
"loss": 1.1412,
"step": 4240
},
{
"epoch": 0.9745471222196744,
"grad_norm": 12.4375,
"learning_rate": 5.136436597110755e-07,
"loss": 1.146,
"step": 4250
},
{
"epoch": 0.976840174271956,
"grad_norm": 10.875,
"learning_rate": 4.6778261866544374e-07,
"loss": 1.1279,
"step": 4260
},
{
"epoch": 0.9791332263242376,
"grad_norm": 16.375,
"learning_rate": 4.21921577619812e-07,
"loss": 1.1603,
"step": 4270
},
{
"epoch": 0.9814262783765192,
"grad_norm": 11.0625,
"learning_rate": 3.7606053657418027e-07,
"loss": 1.1666,
"step": 4280
},
{
"epoch": 0.9837193304288008,
"grad_norm": 9.375,
"learning_rate": 3.3019949552854856e-07,
"loss": 1.1609,
"step": 4290
},
{
"epoch": 0.9860123824810824,
"grad_norm": 14.5625,
"learning_rate": 2.843384544829168e-07,
"loss": 1.1507,
"step": 4300
},
{
"epoch": 0.988305434533364,
"grad_norm": 10.1875,
"learning_rate": 2.3847741343728504e-07,
"loss": 1.1535,
"step": 4310
},
{
"epoch": 0.9905984865856455,
"grad_norm": 12.3125,
"learning_rate": 1.926163723916533e-07,
"loss": 1.1835,
"step": 4320
},
{
"epoch": 0.9928915386379271,
"grad_norm": 9.75,
"learning_rate": 1.4675533134602156e-07,
"loss": 1.1578,
"step": 4330
},
{
"epoch": 0.9951845906902087,
"grad_norm": 11.3125,
"learning_rate": 1.0089429030038983e-07,
"loss": 1.1447,
"step": 4340
},
{
"epoch": 0.9974776427424903,
"grad_norm": 11.875,
"learning_rate": 5.5033249254758087e-08,
"loss": 1.1568,
"step": 4350
},
{
"epoch": 0.9997706947947719,
"grad_norm": 11.375,
"learning_rate": 9.172208209126348e-09,
"loss": 1.15,
"step": 4360
}
],
"logging_steps": 10,
"max_steps": 4361,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7.745762607278223e+17,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}