diff --git "a/checkpoint-4000/trainer_state.json" "b/checkpoint-4000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-4000/trainer_state.json" @@ -0,0 +1,28034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0621348911311737, + "eval_steps": 500, + "global_step": 4000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0002655337227827934, + "grad_norm": 26.09913849342828, + "learning_rate": 0.0, + "loss": 3.2031, + "step": 1 + }, + { + "epoch": 0.0005310674455655868, + "grad_norm": 33.24418612316917, + "learning_rate": 2.6525198938992043e-08, + "loss": 3.4062, + "step": 2 + }, + { + "epoch": 0.0007966011683483802, + "grad_norm": 22.666174254080957, + "learning_rate": 5.3050397877984086e-08, + "loss": 3.5469, + "step": 3 + }, + { + "epoch": 0.0010621348911311736, + "grad_norm": 47.54041711728203, + "learning_rate": 7.957559681697613e-08, + "loss": 3.3906, + "step": 4 + }, + { + "epoch": 0.0013276686139139671, + "grad_norm": 28.74129018702557, + "learning_rate": 1.0610079575596817e-07, + "loss": 3.7344, + "step": 5 + }, + { + "epoch": 0.0015932023366967605, + "grad_norm": 34.823794266535295, + "learning_rate": 1.3262599469496022e-07, + "loss": 4.875, + "step": 6 + }, + { + "epoch": 0.0018587360594795538, + "grad_norm": 33.17874143897226, + "learning_rate": 1.5915119363395226e-07, + "loss": 4.8125, + "step": 7 + }, + { + "epoch": 0.002124269782262347, + "grad_norm": 22.916328083623917, + "learning_rate": 1.856763925729443e-07, + "loss": 3.5625, + "step": 8 + }, + { + "epoch": 0.002389803505045141, + "grad_norm": 22.722952034819503, + "learning_rate": 2.1220159151193635e-07, + "loss": 2.8281, + "step": 9 + }, + { + "epoch": 0.0026553372278279343, + "grad_norm": 71.13233477133232, + "learning_rate": 2.387267904509284e-07, + "loss": 4.2812, + "step": 10 + }, + { + "epoch": 0.0029208709506107276, + "grad_norm": 22.509369620011583, + "learning_rate": 2.6525198938992043e-07, + "loss": 2.6562, + "step": 11 + }, + { + "epoch": 0.003186404673393521, + "grad_norm": 36.02796517304586, + "learning_rate": 2.917771883289125e-07, + "loss": 3.9375, + "step": 12 + }, + { + "epoch": 0.0034519383961763143, + "grad_norm": 35.53736888706161, + "learning_rate": 3.183023872679045e-07, + "loss": 4.375, + "step": 13 + }, + { + "epoch": 0.0037174721189591076, + "grad_norm": 21.726341733924155, + "learning_rate": 3.4482758620689656e-07, + "loss": 2.8594, + "step": 14 + }, + { + "epoch": 0.003983005841741901, + "grad_norm": 26.02926274160067, + "learning_rate": 3.713527851458886e-07, + "loss": 3.3125, + "step": 15 + }, + { + "epoch": 0.004248539564524694, + "grad_norm": 33.549456836516455, + "learning_rate": 3.9787798408488065e-07, + "loss": 4.5, + "step": 16 + }, + { + "epoch": 0.004514073287307488, + "grad_norm": 5704.608112490134, + "learning_rate": 4.244031830238727e-07, + "loss": 2.6875, + "step": 17 + }, + { + "epoch": 0.004779607010090282, + "grad_norm": 46.28223578012096, + "learning_rate": 4.5092838196286473e-07, + "loss": 3.3906, + "step": 18 + }, + { + "epoch": 0.005045140732873075, + "grad_norm": 21.51924983534021, + "learning_rate": 4.774535809018568e-07, + "loss": 4.0938, + "step": 19 + }, + { + "epoch": 0.0053106744556558685, + "grad_norm": 27.91289735068091, + "learning_rate": 5.039787798408489e-07, + "loss": 3.4062, + "step": 20 + }, + { + "epoch": 0.0055762081784386614, + "grad_norm": 38.86707334268686, + "learning_rate": 5.305039787798409e-07, + "loss": 4.5938, + "step": 21 + }, + { + "epoch": 0.005841741901221455, + "grad_norm": 46.02296280217304, + "learning_rate": 5.570291777188329e-07, + "loss": 4.5312, + "step": 22 + }, + { + "epoch": 0.006107275624004248, + "grad_norm": 29.064306674372673, + "learning_rate": 5.83554376657825e-07, + "loss": 3.7031, + "step": 23 + }, + { + "epoch": 0.006372809346787042, + "grad_norm": 22.0053607091662, + "learning_rate": 6.10079575596817e-07, + "loss": 4.0938, + "step": 24 + }, + { + "epoch": 0.006638343069569836, + "grad_norm": 25.74196411791827, + "learning_rate": 6.36604774535809e-07, + "loss": 3.1875, + "step": 25 + }, + { + "epoch": 0.006903876792352629, + "grad_norm": 28.41314685895795, + "learning_rate": 6.631299734748012e-07, + "loss": 3.5938, + "step": 26 + }, + { + "epoch": 0.007169410515135422, + "grad_norm": 24.75234923466472, + "learning_rate": 6.896551724137931e-07, + "loss": 2.6875, + "step": 27 + }, + { + "epoch": 0.007434944237918215, + "grad_norm": 30.66371803716353, + "learning_rate": 7.161803713527853e-07, + "loss": 4.0312, + "step": 28 + }, + { + "epoch": 0.007700477960701009, + "grad_norm": 29.00163943943682, + "learning_rate": 7.427055702917772e-07, + "loss": 3.3594, + "step": 29 + }, + { + "epoch": 0.007966011683483803, + "grad_norm": 25.77157343246292, + "learning_rate": 7.692307692307694e-07, + "loss": 3.1875, + "step": 30 + }, + { + "epoch": 0.008231545406266597, + "grad_norm": 22.334502595274675, + "learning_rate": 7.957559681697613e-07, + "loss": 4.7188, + "step": 31 + }, + { + "epoch": 0.008497079129049389, + "grad_norm": 32.603443673328215, + "learning_rate": 8.222811671087534e-07, + "loss": 3.8906, + "step": 32 + }, + { + "epoch": 0.008762612851832182, + "grad_norm": 24.08930675302932, + "learning_rate": 8.488063660477454e-07, + "loss": 2.8906, + "step": 33 + }, + { + "epoch": 0.009028146574614976, + "grad_norm": 32.39366263050705, + "learning_rate": 8.753315649867375e-07, + "loss": 4.125, + "step": 34 + }, + { + "epoch": 0.00929368029739777, + "grad_norm": 29.971764947175103, + "learning_rate": 9.018567639257295e-07, + "loss": 3.2344, + "step": 35 + }, + { + "epoch": 0.009559214020180564, + "grad_norm": 33.19925342837082, + "learning_rate": 9.283819628647216e-07, + "loss": 4.25, + "step": 36 + }, + { + "epoch": 0.009824747742963356, + "grad_norm": 24.097206276546917, + "learning_rate": 9.549071618037136e-07, + "loss": 2.7031, + "step": 37 + }, + { + "epoch": 0.01009028146574615, + "grad_norm": 25.033091440150386, + "learning_rate": 9.814323607427057e-07, + "loss": 3.5469, + "step": 38 + }, + { + "epoch": 0.010355815188528943, + "grad_norm": 19.83584675026491, + "learning_rate": 1.0079575596816979e-06, + "loss": 3.8125, + "step": 39 + }, + { + "epoch": 0.010621348911311737, + "grad_norm": 22.843498610408417, + "learning_rate": 1.0344827586206898e-06, + "loss": 4.0, + "step": 40 + }, + { + "epoch": 0.01088688263409453, + "grad_norm": 27.23163955174518, + "learning_rate": 1.0610079575596817e-06, + "loss": 4.2188, + "step": 41 + }, + { + "epoch": 0.011152416356877323, + "grad_norm": 31.562228731897743, + "learning_rate": 1.0875331564986739e-06, + "loss": 4.375, + "step": 42 + }, + { + "epoch": 0.011417950079660117, + "grad_norm": 22.949386153635178, + "learning_rate": 1.1140583554376658e-06, + "loss": 3.5156, + "step": 43 + }, + { + "epoch": 0.01168348380244291, + "grad_norm": 88.31997625312175, + "learning_rate": 1.140583554376658e-06, + "loss": 2.7812, + "step": 44 + }, + { + "epoch": 0.011949017525225704, + "grad_norm": 14.729164979113321, + "learning_rate": 1.16710875331565e-06, + "loss": 2.2344, + "step": 45 + }, + { + "epoch": 0.012214551248008496, + "grad_norm": 26.318599853460583, + "learning_rate": 1.193633952254642e-06, + "loss": 3.8125, + "step": 46 + }, + { + "epoch": 0.01248008497079129, + "grad_norm": 36.435091476888246, + "learning_rate": 1.220159151193634e-06, + "loss": 3.5156, + "step": 47 + }, + { + "epoch": 0.012745618693574084, + "grad_norm": 21.353387655655027, + "learning_rate": 1.2466843501326261e-06, + "loss": 2.4219, + "step": 48 + }, + { + "epoch": 0.013011152416356878, + "grad_norm": 36.608824688888205, + "learning_rate": 1.273209549071618e-06, + "loss": 3.6562, + "step": 49 + }, + { + "epoch": 0.013276686139139671, + "grad_norm": 32.7469992104952, + "learning_rate": 1.29973474801061e-06, + "loss": 4.2812, + "step": 50 + }, + { + "epoch": 0.013542219861922463, + "grad_norm": 527.1528527718093, + "learning_rate": 1.3262599469496024e-06, + "loss": 2.4375, + "step": 51 + }, + { + "epoch": 0.013807753584705257, + "grad_norm": 27.990609419848134, + "learning_rate": 1.3527851458885943e-06, + "loss": 2.3906, + "step": 52 + }, + { + "epoch": 0.014073287307488051, + "grad_norm": 14.694389603614836, + "learning_rate": 1.3793103448275862e-06, + "loss": 3.2344, + "step": 53 + }, + { + "epoch": 0.014338821030270845, + "grad_norm": 20.735659157221477, + "learning_rate": 1.4058355437665782e-06, + "loss": 2.8281, + "step": 54 + }, + { + "epoch": 0.014604354753053638, + "grad_norm": 31.138348407935453, + "learning_rate": 1.4323607427055705e-06, + "loss": 3.8906, + "step": 55 + }, + { + "epoch": 0.01486988847583643, + "grad_norm": 29.62867209466913, + "learning_rate": 1.4588859416445625e-06, + "loss": 3.6875, + "step": 56 + }, + { + "epoch": 0.015135422198619224, + "grad_norm": 14.03880122581976, + "learning_rate": 1.4854111405835544e-06, + "loss": 2.1406, + "step": 57 + }, + { + "epoch": 0.015400955921402018, + "grad_norm": 33.23697742979803, + "learning_rate": 1.5119363395225464e-06, + "loss": 3.5, + "step": 58 + }, + { + "epoch": 0.015666489644184812, + "grad_norm": 29.453655249544546, + "learning_rate": 1.5384615384615387e-06, + "loss": 2.5312, + "step": 59 + }, + { + "epoch": 0.015932023366967606, + "grad_norm": 27.29098711329942, + "learning_rate": 1.5649867374005307e-06, + "loss": 2.6406, + "step": 60 + }, + { + "epoch": 0.0161975570897504, + "grad_norm": 28.33472292785293, + "learning_rate": 1.5915119363395226e-06, + "loss": 2.5, + "step": 61 + }, + { + "epoch": 0.016463090812533193, + "grad_norm": 16.521064455634754, + "learning_rate": 1.618037135278515e-06, + "loss": 2.4219, + "step": 62 + }, + { + "epoch": 0.016728624535315983, + "grad_norm": 13.384541738840948, + "learning_rate": 1.6445623342175069e-06, + "loss": 1.8672, + "step": 63 + }, + { + "epoch": 0.016994158258098777, + "grad_norm": 19.701703431707948, + "learning_rate": 1.6710875331564988e-06, + "loss": 2.4219, + "step": 64 + }, + { + "epoch": 0.01725969198088157, + "grad_norm": 32.143288806734, + "learning_rate": 1.6976127320954908e-06, + "loss": 2.6094, + "step": 65 + }, + { + "epoch": 0.017525225703664365, + "grad_norm": 30.63700955458724, + "learning_rate": 1.724137931034483e-06, + "loss": 2.8125, + "step": 66 + }, + { + "epoch": 0.01779075942644716, + "grad_norm": 18.61018137166964, + "learning_rate": 1.750663129973475e-06, + "loss": 2.125, + "step": 67 + }, + { + "epoch": 0.018056293149229952, + "grad_norm": 14.421577541639564, + "learning_rate": 1.777188328912467e-06, + "loss": 1.7578, + "step": 68 + }, + { + "epoch": 0.018321826872012746, + "grad_norm": 29.39406132029808, + "learning_rate": 1.803713527851459e-06, + "loss": 2.4375, + "step": 69 + }, + { + "epoch": 0.01858736059479554, + "grad_norm": 21.159681548926745, + "learning_rate": 1.830238726790451e-06, + "loss": 2.0938, + "step": 70 + }, + { + "epoch": 0.018852894317578334, + "grad_norm": 27.539564746628287, + "learning_rate": 1.8567639257294432e-06, + "loss": 3.0156, + "step": 71 + }, + { + "epoch": 0.019118428040361127, + "grad_norm": 24.190559734623005, + "learning_rate": 1.8832891246684352e-06, + "loss": 3.0, + "step": 72 + }, + { + "epoch": 0.019383961763143918, + "grad_norm": 26.96417513483311, + "learning_rate": 1.909814323607427e-06, + "loss": 3.0156, + "step": 73 + }, + { + "epoch": 0.01964949548592671, + "grad_norm": 24.89585126504584, + "learning_rate": 1.9363395225464193e-06, + "loss": 2.2344, + "step": 74 + }, + { + "epoch": 0.019915029208709505, + "grad_norm": 56.411148213900134, + "learning_rate": 1.9628647214854114e-06, + "loss": 2.1406, + "step": 75 + }, + { + "epoch": 0.0201805629314923, + "grad_norm": 26.867726567521345, + "learning_rate": 1.989389920424403e-06, + "loss": 1.5469, + "step": 76 + }, + { + "epoch": 0.020446096654275093, + "grad_norm": 25.736893839761308, + "learning_rate": 2.0159151193633957e-06, + "loss": 2.5312, + "step": 77 + }, + { + "epoch": 0.020711630377057887, + "grad_norm": 29.10383691197578, + "learning_rate": 2.0424403183023874e-06, + "loss": 2.0312, + "step": 78 + }, + { + "epoch": 0.02097716409984068, + "grad_norm": 16.34789260778447, + "learning_rate": 2.0689655172413796e-06, + "loss": 2.6719, + "step": 79 + }, + { + "epoch": 0.021242697822623474, + "grad_norm": 17.443495153034082, + "learning_rate": 2.0954907161803713e-06, + "loss": 0.7266, + "step": 80 + }, + { + "epoch": 0.021508231545406268, + "grad_norm": 25.598705341280716, + "learning_rate": 2.1220159151193635e-06, + "loss": 1.2109, + "step": 81 + }, + { + "epoch": 0.02177376526818906, + "grad_norm": 143.5401268253218, + "learning_rate": 2.1485411140583556e-06, + "loss": 1.6016, + "step": 82 + }, + { + "epoch": 0.022039298990971852, + "grad_norm": 26.389545474475735, + "learning_rate": 2.1750663129973478e-06, + "loss": 1.8438, + "step": 83 + }, + { + "epoch": 0.022304832713754646, + "grad_norm": 23.96871143922354, + "learning_rate": 2.2015915119363395e-06, + "loss": 0.9219, + "step": 84 + }, + { + "epoch": 0.02257036643653744, + "grad_norm": 38.19763740874266, + "learning_rate": 2.2281167108753316e-06, + "loss": 0.8359, + "step": 85 + }, + { + "epoch": 0.022835900159320233, + "grad_norm": 29.364774056584984, + "learning_rate": 2.2546419098143238e-06, + "loss": 0.6602, + "step": 86 + }, + { + "epoch": 0.023101433882103027, + "grad_norm": 26.845223667722713, + "learning_rate": 2.281167108753316e-06, + "loss": 0.8086, + "step": 87 + }, + { + "epoch": 0.02336696760488582, + "grad_norm": 28.831828873775432, + "learning_rate": 2.307692307692308e-06, + "loss": 1.4062, + "step": 88 + }, + { + "epoch": 0.023632501327668615, + "grad_norm": 24.13587932409368, + "learning_rate": 2.3342175066313e-06, + "loss": 1.3828, + "step": 89 + }, + { + "epoch": 0.02389803505045141, + "grad_norm": 24.636469804747414, + "learning_rate": 2.360742705570292e-06, + "loss": 1.0547, + "step": 90 + }, + { + "epoch": 0.024163568773234202, + "grad_norm": 10.515597346677467, + "learning_rate": 2.387267904509284e-06, + "loss": 0.0923, + "step": 91 + }, + { + "epoch": 0.024429102496016993, + "grad_norm": 22.66977819594809, + "learning_rate": 2.4137931034482762e-06, + "loss": 0.2832, + "step": 92 + }, + { + "epoch": 0.024694636218799786, + "grad_norm": 4.5274633525739185, + "learning_rate": 2.440318302387268e-06, + "loss": 0.0259, + "step": 93 + }, + { + "epoch": 0.02496016994158258, + "grad_norm": 9.36517984448498, + "learning_rate": 2.46684350132626e-06, + "loss": 0.0442, + "step": 94 + }, + { + "epoch": 0.025225703664365374, + "grad_norm": 11.897696771384094, + "learning_rate": 2.4933687002652523e-06, + "loss": 0.1138, + "step": 95 + }, + { + "epoch": 0.025491237387148168, + "grad_norm": 32.70173331678797, + "learning_rate": 2.5198938992042444e-06, + "loss": 6.4688, + "step": 96 + }, + { + "epoch": 0.02575677110993096, + "grad_norm": 0.031168975385739393, + "learning_rate": 2.546419098143236e-06, + "loss": 0.0003, + "step": 97 + }, + { + "epoch": 0.026022304832713755, + "grad_norm": 0.7313576810659675, + "learning_rate": 2.5729442970822283e-06, + "loss": 0.0075, + "step": 98 + }, + { + "epoch": 0.02628783855549655, + "grad_norm": 0.5685930769193894, + "learning_rate": 2.59946949602122e-06, + "loss": 0.0045, + "step": 99 + }, + { + "epoch": 0.026553372278279343, + "grad_norm": 4.772903853691814, + "learning_rate": 2.625994694960212e-06, + "loss": 0.2275, + "step": 100 + }, + { + "epoch": 0.026818906001062136, + "grad_norm": 5.102874825538071, + "learning_rate": 2.6525198938992047e-06, + "loss": 0.5703, + "step": 101 + }, + { + "epoch": 0.027084439723844927, + "grad_norm": 0.27402642551717465, + "learning_rate": 2.679045092838197e-06, + "loss": 0.0028, + "step": 102 + }, + { + "epoch": 0.02734997344662772, + "grad_norm": 4.6324372685319695, + "learning_rate": 2.7055702917771886e-06, + "loss": 0.1367, + "step": 103 + }, + { + "epoch": 0.027615507169410514, + "grad_norm": 0.038045626774934026, + "learning_rate": 2.7320954907161808e-06, + "loss": 0.0007, + "step": 104 + }, + { + "epoch": 0.027881040892193308, + "grad_norm": 0.6305735286653126, + "learning_rate": 2.7586206896551725e-06, + "loss": 0.0097, + "step": 105 + }, + { + "epoch": 0.028146574614976102, + "grad_norm": 0.019465351131816563, + "learning_rate": 2.7851458885941646e-06, + "loss": 0.0003, + "step": 106 + }, + { + "epoch": 0.028412108337758896, + "grad_norm": 0.03074317996513282, + "learning_rate": 2.8116710875331564e-06, + "loss": 0.0004, + "step": 107 + }, + { + "epoch": 0.02867764206054169, + "grad_norm": 4.503535246450842, + "learning_rate": 2.8381962864721485e-06, + "loss": 0.582, + "step": 108 + }, + { + "epoch": 0.028943175783324483, + "grad_norm": 4.8728912631764265, + "learning_rate": 2.864721485411141e-06, + "loss": 0.5469, + "step": 109 + }, + { + "epoch": 0.029208709506107277, + "grad_norm": 0.0095940645247496, + "learning_rate": 2.891246684350133e-06, + "loss": 0.0002, + "step": 110 + }, + { + "epoch": 0.02947424322889007, + "grad_norm": 5.105714132066514, + "learning_rate": 2.917771883289125e-06, + "loss": 0.7383, + "step": 111 + }, + { + "epoch": 0.02973977695167286, + "grad_norm": 0.18245285164952776, + "learning_rate": 2.944297082228117e-06, + "loss": 0.0027, + "step": 112 + }, + { + "epoch": 0.030005310674455655, + "grad_norm": 4.137453039540649, + "learning_rate": 2.970822281167109e-06, + "loss": 0.2158, + "step": 113 + }, + { + "epoch": 0.03027084439723845, + "grad_norm": 0.466146965565003, + "learning_rate": 2.997347480106101e-06, + "loss": 0.0085, + "step": 114 + }, + { + "epoch": 0.030536378120021242, + "grad_norm": 4.144598857842222, + "learning_rate": 3.0238726790450927e-06, + "loss": 0.4355, + "step": 115 + }, + { + "epoch": 0.030801911842804036, + "grad_norm": 0.07481190979762373, + "learning_rate": 3.0503978779840853e-06, + "loss": 0.0012, + "step": 116 + }, + { + "epoch": 0.03106744556558683, + "grad_norm": 4.864460397183156, + "learning_rate": 3.0769230769230774e-06, + "loss": 0.4883, + "step": 117 + }, + { + "epoch": 0.031332979288369624, + "grad_norm": 0.046629399654510524, + "learning_rate": 3.103448275862069e-06, + "loss": 0.0008, + "step": 118 + }, + { + "epoch": 0.031598513011152414, + "grad_norm": 0.9844013414494834, + "learning_rate": 3.1299734748010613e-06, + "loss": 0.0006, + "step": 119 + }, + { + "epoch": 0.03186404673393521, + "grad_norm": 2.8873700648345726, + "learning_rate": 3.1564986737400535e-06, + "loss": 0.0525, + "step": 120 + }, + { + "epoch": 0.032129580456718, + "grad_norm": 13.401562271488295, + "learning_rate": 3.183023872679045e-06, + "loss": 2.1094, + "step": 121 + }, + { + "epoch": 0.0323951141795008, + "grad_norm": 0.9753972945569677, + "learning_rate": 3.2095490716180373e-06, + "loss": 0.0198, + "step": 122 + }, + { + "epoch": 0.03266064790228359, + "grad_norm": 0.052264498037133865, + "learning_rate": 3.23607427055703e-06, + "loss": 0.0011, + "step": 123 + }, + { + "epoch": 0.032926181625066386, + "grad_norm": 0.0275297136646831, + "learning_rate": 3.2625994694960216e-06, + "loss": 0.0006, + "step": 124 + }, + { + "epoch": 0.03319171534784918, + "grad_norm": 0.1637629675751948, + "learning_rate": 3.2891246684350138e-06, + "loss": 0.0037, + "step": 125 + }, + { + "epoch": 0.03345724907063197, + "grad_norm": 4.78491469984592, + "learning_rate": 3.3156498673740055e-06, + "loss": 0.3535, + "step": 126 + }, + { + "epoch": 0.033722782793414764, + "grad_norm": 15.570720009635428, + "learning_rate": 3.3421750663129977e-06, + "loss": 2.7969, + "step": 127 + }, + { + "epoch": 0.033988316516197555, + "grad_norm": 4.505829037318201, + "learning_rate": 3.3687002652519894e-06, + "loss": 0.3105, + "step": 128 + }, + { + "epoch": 0.03425385023898035, + "grad_norm": 0.042401600725809944, + "learning_rate": 3.3952254641909815e-06, + "loss": 0.0009, + "step": 129 + }, + { + "epoch": 0.03451938396176314, + "grad_norm": 0.3272160377202937, + "learning_rate": 3.4217506631299737e-06, + "loss": 0.0074, + "step": 130 + }, + { + "epoch": 0.03478491768454594, + "grad_norm": 0.20468371949697137, + "learning_rate": 3.448275862068966e-06, + "loss": 0.0036, + "step": 131 + }, + { + "epoch": 0.03505045140732873, + "grad_norm": 4.559385816252051, + "learning_rate": 3.474801061007958e-06, + "loss": 0.2773, + "step": 132 + }, + { + "epoch": 0.03531598513011153, + "grad_norm": 0.44947961387998925, + "learning_rate": 3.50132625994695e-06, + "loss": 0.0093, + "step": 133 + }, + { + "epoch": 0.03558151885289432, + "grad_norm": 2.628443790962567, + "learning_rate": 3.527851458885942e-06, + "loss": 0.0767, + "step": 134 + }, + { + "epoch": 0.035847052575677114, + "grad_norm": 0.3344294580416105, + "learning_rate": 3.554376657824934e-06, + "loss": 0.0079, + "step": 135 + }, + { + "epoch": 0.036112586298459905, + "grad_norm": 6.1109142288253375, + "learning_rate": 3.5809018567639257e-06, + "loss": 0.1455, + "step": 136 + }, + { + "epoch": 0.036378120021242695, + "grad_norm": 0.19257798325675396, + "learning_rate": 3.607427055702918e-06, + "loss": 0.0043, + "step": 137 + }, + { + "epoch": 0.03664365374402549, + "grad_norm": 3.878928845456885, + "learning_rate": 3.6339522546419104e-06, + "loss": 0.1934, + "step": 138 + }, + { + "epoch": 0.03690918746680828, + "grad_norm": 1.5262893890482263, + "learning_rate": 3.660477453580902e-06, + "loss": 0.032, + "step": 139 + }, + { + "epoch": 0.03717472118959108, + "grad_norm": 0.19902989840213414, + "learning_rate": 3.6870026525198943e-06, + "loss": 0.0045, + "step": 140 + }, + { + "epoch": 0.03744025491237387, + "grad_norm": 0.5136742696527371, + "learning_rate": 3.7135278514588865e-06, + "loss": 0.0104, + "step": 141 + }, + { + "epoch": 0.03770578863515667, + "grad_norm": 3.86476349919558, + "learning_rate": 3.740053050397878e-06, + "loss": 0.1367, + "step": 142 + }, + { + "epoch": 0.03797132235793946, + "grad_norm": 0.796785610921372, + "learning_rate": 3.7665782493368703e-06, + "loss": 0.0194, + "step": 143 + }, + { + "epoch": 0.038236856080722255, + "grad_norm": 1.5271845218247893, + "learning_rate": 3.793103448275862e-06, + "loss": 0.0288, + "step": 144 + }, + { + "epoch": 0.038502389803505045, + "grad_norm": 2.427314819468491, + "learning_rate": 3.819628647214854e-06, + "loss": 0.0576, + "step": 145 + }, + { + "epoch": 0.038767923526287836, + "grad_norm": 4.117849401975011, + "learning_rate": 3.846153846153847e-06, + "loss": 0.2441, + "step": 146 + }, + { + "epoch": 0.03903345724907063, + "grad_norm": 0.15723067876063995, + "learning_rate": 3.8726790450928385e-06, + "loss": 0.0039, + "step": 147 + }, + { + "epoch": 0.03929899097185342, + "grad_norm": 3.7869596148640228, + "learning_rate": 3.89920424403183e-06, + "loss": 0.1641, + "step": 148 + }, + { + "epoch": 0.03956452469463622, + "grad_norm": 0.37297101582033887, + "learning_rate": 3.925729442970823e-06, + "loss": 0.0067, + "step": 149 + }, + { + "epoch": 0.03983005841741901, + "grad_norm": 3.6378474369832037, + "learning_rate": 3.9522546419098145e-06, + "loss": 0.0957, + "step": 150 + }, + { + "epoch": 0.04009559214020181, + "grad_norm": 0.43825791965553007, + "learning_rate": 3.978779840848806e-06, + "loss": 0.0105, + "step": 151 + }, + { + "epoch": 0.0403611258629846, + "grad_norm": 3.833802266035273, + "learning_rate": 4.005305039787799e-06, + "loss": 0.1934, + "step": 152 + }, + { + "epoch": 0.040626659585767395, + "grad_norm": 0.27648447133818854, + "learning_rate": 4.031830238726791e-06, + "loss": 0.0059, + "step": 153 + }, + { + "epoch": 0.040892193308550186, + "grad_norm": 0.5502618375049242, + "learning_rate": 4.058355437665783e-06, + "loss": 0.0117, + "step": 154 + }, + { + "epoch": 0.041157727031332976, + "grad_norm": 0.3759989484991633, + "learning_rate": 4.084880636604775e-06, + "loss": 0.0093, + "step": 155 + }, + { + "epoch": 0.04142326075411577, + "grad_norm": 0.15725269535355169, + "learning_rate": 4.111405835543767e-06, + "loss": 0.0033, + "step": 156 + }, + { + "epoch": 0.041688794476898564, + "grad_norm": 0.5545627730013464, + "learning_rate": 4.137931034482759e-06, + "loss": 0.0137, + "step": 157 + }, + { + "epoch": 0.04195432819968136, + "grad_norm": 0.3259056057700917, + "learning_rate": 4.164456233421751e-06, + "loss": 0.0081, + "step": 158 + }, + { + "epoch": 0.04221986192246415, + "grad_norm": 0.1865224020504324, + "learning_rate": 4.190981432360743e-06, + "loss": 0.0041, + "step": 159 + }, + { + "epoch": 0.04248539564524695, + "grad_norm": 1.8307778726247308, + "learning_rate": 4.217506631299735e-06, + "loss": 0.0532, + "step": 160 + }, + { + "epoch": 0.04275092936802974, + "grad_norm": 3.8899387815695117, + "learning_rate": 4.244031830238727e-06, + "loss": 0.1553, + "step": 161 + }, + { + "epoch": 0.043016463090812536, + "grad_norm": 0.35213110395365343, + "learning_rate": 4.2705570291777195e-06, + "loss": 0.0063, + "step": 162 + }, + { + "epoch": 0.043281996813595326, + "grad_norm": 0.12739793280337483, + "learning_rate": 4.297082228116711e-06, + "loss": 0.0029, + "step": 163 + }, + { + "epoch": 0.04354753053637812, + "grad_norm": 3.7917485475147767, + "learning_rate": 4.323607427055703e-06, + "loss": 0.2559, + "step": 164 + }, + { + "epoch": 0.043813064259160914, + "grad_norm": 4.222145018176052, + "learning_rate": 4.3501326259946955e-06, + "loss": 0.3105, + "step": 165 + }, + { + "epoch": 0.044078597981943704, + "grad_norm": 0.1510587971806312, + "learning_rate": 4.376657824933687e-06, + "loss": 0.0036, + "step": 166 + }, + { + "epoch": 0.0443441317047265, + "grad_norm": 0.3746365601473909, + "learning_rate": 4.403183023872679e-06, + "loss": 0.0091, + "step": 167 + }, + { + "epoch": 0.04460966542750929, + "grad_norm": 3.6742314366524482, + "learning_rate": 4.4297082228116715e-06, + "loss": 0.1641, + "step": 168 + }, + { + "epoch": 0.04487519915029209, + "grad_norm": 4.169049834925408, + "learning_rate": 4.456233421750663e-06, + "loss": 0.1367, + "step": 169 + }, + { + "epoch": 0.04514073287307488, + "grad_norm": 23.2584513485522, + "learning_rate": 4.482758620689656e-06, + "loss": 2.375, + "step": 170 + }, + { + "epoch": 0.045406266595857676, + "grad_norm": 3.6642006588717737, + "learning_rate": 4.5092838196286476e-06, + "loss": 0.1465, + "step": 171 + }, + { + "epoch": 0.04567180031864047, + "grad_norm": 0.2890573481827513, + "learning_rate": 4.535809018567639e-06, + "loss": 0.0065, + "step": 172 + }, + { + "epoch": 0.045937334041423264, + "grad_norm": 3.68276451135959, + "learning_rate": 4.562334217506632e-06, + "loss": 0.1836, + "step": 173 + }, + { + "epoch": 0.046202867764206054, + "grad_norm": 2.4177768079368986, + "learning_rate": 4.588859416445624e-06, + "loss": 0.0479, + "step": 174 + }, + { + "epoch": 0.046468401486988845, + "grad_norm": 17.645780333783485, + "learning_rate": 4.615384615384616e-06, + "loss": 1.9531, + "step": 175 + }, + { + "epoch": 0.04673393520977164, + "grad_norm": 0.9556851330908339, + "learning_rate": 4.641909814323608e-06, + "loss": 0.0231, + "step": 176 + }, + { + "epoch": 0.04699946893255443, + "grad_norm": 0.15825725091826115, + "learning_rate": 4.6684350132626e-06, + "loss": 0.0035, + "step": 177 + }, + { + "epoch": 0.04726500265533723, + "grad_norm": 5.0505715269377065, + "learning_rate": 4.694960212201592e-06, + "loss": 0.0219, + "step": 178 + }, + { + "epoch": 0.04753053637812002, + "grad_norm": 0.19812718466796467, + "learning_rate": 4.721485411140584e-06, + "loss": 0.0051, + "step": 179 + }, + { + "epoch": 0.04779607010090282, + "grad_norm": 3.6161442909421218, + "learning_rate": 4.748010610079576e-06, + "loss": 0.165, + "step": 180 + }, + { + "epoch": 0.04806160382368561, + "grad_norm": 0.322834845757156, + "learning_rate": 4.774535809018568e-06, + "loss": 0.0065, + "step": 181 + }, + { + "epoch": 0.048327137546468404, + "grad_norm": 5.4547556811669935, + "learning_rate": 4.80106100795756e-06, + "loss": 2.125, + "step": 182 + }, + { + "epoch": 0.048592671269251195, + "grad_norm": 3.584757373914313, + "learning_rate": 4.8275862068965525e-06, + "loss": 0.1553, + "step": 183 + }, + { + "epoch": 0.048858204992033985, + "grad_norm": 0.7190093535962121, + "learning_rate": 4.854111405835544e-06, + "loss": 0.0187, + "step": 184 + }, + { + "epoch": 0.04912373871481678, + "grad_norm": 8.285280730208004, + "learning_rate": 4.880636604774536e-06, + "loss": 2.0156, + "step": 185 + }, + { + "epoch": 0.04938927243759957, + "grad_norm": 0.1589100928383723, + "learning_rate": 4.9071618037135285e-06, + "loss": 0.0035, + "step": 186 + }, + { + "epoch": 0.04965480616038237, + "grad_norm": 1.9461569166969948, + "learning_rate": 4.93368700265252e-06, + "loss": 0.0479, + "step": 187 + }, + { + "epoch": 0.04992033988316516, + "grad_norm": 0.28568945003091645, + "learning_rate": 4.960212201591512e-06, + "loss": 0.0058, + "step": 188 + }, + { + "epoch": 0.05018587360594796, + "grad_norm": 0.19319543123307042, + "learning_rate": 4.9867374005305045e-06, + "loss": 0.0033, + "step": 189 + }, + { + "epoch": 0.05045140732873075, + "grad_norm": 4.037258353685831, + "learning_rate": 5.013262599469496e-06, + "loss": 0.0815, + "step": 190 + }, + { + "epoch": 0.050716941051513545, + "grad_norm": 1.3356221226261589, + "learning_rate": 5.039787798408489e-06, + "loss": 0.0354, + "step": 191 + }, + { + "epoch": 0.050982474774296335, + "grad_norm": 0.1480129388225318, + "learning_rate": 5.06631299734748e-06, + "loss": 0.0026, + "step": 192 + }, + { + "epoch": 0.05124800849707913, + "grad_norm": 3.5661629795250045, + "learning_rate": 5.092838196286472e-06, + "loss": 0.2441, + "step": 193 + }, + { + "epoch": 0.05151354221986192, + "grad_norm": 2.1523745771682172, + "learning_rate": 5.119363395225465e-06, + "loss": 0.0576, + "step": 194 + }, + { + "epoch": 0.05177907594264471, + "grad_norm": 0.9192144869318637, + "learning_rate": 5.145888594164457e-06, + "loss": 0.0208, + "step": 195 + }, + { + "epoch": 0.05204460966542751, + "grad_norm": 0.2690072757879695, + "learning_rate": 5.172413793103449e-06, + "loss": 0.0064, + "step": 196 + }, + { + "epoch": 0.0523101433882103, + "grad_norm": 2.7717636534246792, + "learning_rate": 5.19893899204244e-06, + "loss": 0.0884, + "step": 197 + }, + { + "epoch": 0.0525756771109931, + "grad_norm": 7.0745463915342635, + "learning_rate": 5.225464190981433e-06, + "loss": 2.0156, + "step": 198 + }, + { + "epoch": 0.05284121083377589, + "grad_norm": 0.19227286962350826, + "learning_rate": 5.251989389920424e-06, + "loss": 0.0045, + "step": 199 + }, + { + "epoch": 0.053106744556558685, + "grad_norm": 0.8537355801992382, + "learning_rate": 5.278514588859417e-06, + "loss": 0.0069, + "step": 200 + }, + { + "epoch": 0.053372278279341476, + "grad_norm": 0.11929159424700997, + "learning_rate": 5.3050397877984095e-06, + "loss": 0.0025, + "step": 201 + }, + { + "epoch": 0.05363781200212427, + "grad_norm": 0.07790410641989416, + "learning_rate": 5.331564986737401e-06, + "loss": 0.0013, + "step": 202 + }, + { + "epoch": 0.05390334572490706, + "grad_norm": 0.13268954955742412, + "learning_rate": 5.358090185676394e-06, + "loss": 0.0032, + "step": 203 + }, + { + "epoch": 0.054168879447689854, + "grad_norm": 4.439535441327748, + "learning_rate": 5.384615384615385e-06, + "loss": 0.1836, + "step": 204 + }, + { + "epoch": 0.05443441317047265, + "grad_norm": 0.2688219627795949, + "learning_rate": 5.411140583554377e-06, + "loss": 0.0047, + "step": 205 + }, + { + "epoch": 0.05469994689325544, + "grad_norm": 0.08483073451016193, + "learning_rate": 5.437665782493369e-06, + "loss": 0.0019, + "step": 206 + }, + { + "epoch": 0.05496548061603824, + "grad_norm": 4.028626588799805, + "learning_rate": 5.4641909814323615e-06, + "loss": 0.3203, + "step": 207 + }, + { + "epoch": 0.05523101433882103, + "grad_norm": 2.4610886737988866, + "learning_rate": 5.490716180371354e-06, + "loss": 0.0688, + "step": 208 + }, + { + "epoch": 0.055496548061603826, + "grad_norm": 1.133315536600772, + "learning_rate": 5.517241379310345e-06, + "loss": 0.0258, + "step": 209 + }, + { + "epoch": 0.055762081784386616, + "grad_norm": 0.7726546713585315, + "learning_rate": 5.5437665782493376e-06, + "loss": 0.0152, + "step": 210 + }, + { + "epoch": 0.05602761550716941, + "grad_norm": 0.38371963709194457, + "learning_rate": 5.570291777188329e-06, + "loss": 0.0092, + "step": 211 + }, + { + "epoch": 0.056293149229952204, + "grad_norm": 0.13225755055257674, + "learning_rate": 5.596816976127322e-06, + "loss": 0.0023, + "step": 212 + }, + { + "epoch": 0.056558682952734994, + "grad_norm": 0.08917614772102375, + "learning_rate": 5.623342175066313e-06, + "loss": 0.002, + "step": 213 + }, + { + "epoch": 0.05682421667551779, + "grad_norm": 4.854086836177229, + "learning_rate": 5.649867374005305e-06, + "loss": 0.1035, + "step": 214 + }, + { + "epoch": 0.05708975039830058, + "grad_norm": 0.04007293390414322, + "learning_rate": 5.676392572944297e-06, + "loss": 0.0007, + "step": 215 + }, + { + "epoch": 0.05735528412108338, + "grad_norm": 0.2848845451163088, + "learning_rate": 5.70291777188329e-06, + "loss": 0.0064, + "step": 216 + }, + { + "epoch": 0.05762081784386617, + "grad_norm": 4.686394489350751, + "learning_rate": 5.729442970822282e-06, + "loss": 0.2246, + "step": 217 + }, + { + "epoch": 0.057886351566648966, + "grad_norm": 0.11120455969442722, + "learning_rate": 5.755968169761274e-06, + "loss": 0.0029, + "step": 218 + }, + { + "epoch": 0.05815188528943176, + "grad_norm": 4.663386143892064, + "learning_rate": 5.782493368700266e-06, + "loss": 0.1641, + "step": 219 + }, + { + "epoch": 0.058417419012214554, + "grad_norm": 0.24512762915181402, + "learning_rate": 5.809018567639257e-06, + "loss": 0.0036, + "step": 220 + }, + { + "epoch": 0.058682952734997344, + "grad_norm": 4.183804090865918, + "learning_rate": 5.83554376657825e-06, + "loss": 0.2148, + "step": 221 + }, + { + "epoch": 0.05894848645778014, + "grad_norm": 0.06396150026208348, + "learning_rate": 5.862068965517242e-06, + "loss": 0.0012, + "step": 222 + }, + { + "epoch": 0.05921402018056293, + "grad_norm": 0.1942578634333766, + "learning_rate": 5.888594164456234e-06, + "loss": 0.0037, + "step": 223 + }, + { + "epoch": 0.05947955390334572, + "grad_norm": 0.07594022266578823, + "learning_rate": 5.915119363395227e-06, + "loss": 0.002, + "step": 224 + }, + { + "epoch": 0.05974508762612852, + "grad_norm": 0.35116426830013603, + "learning_rate": 5.941644562334218e-06, + "loss": 0.0074, + "step": 225 + }, + { + "epoch": 0.06001062134891131, + "grad_norm": 1.578167800096405, + "learning_rate": 5.96816976127321e-06, + "loss": 0.0356, + "step": 226 + }, + { + "epoch": 0.06027615507169411, + "grad_norm": 1.1285595351705588, + "learning_rate": 5.994694960212202e-06, + "loss": 0.0255, + "step": 227 + }, + { + "epoch": 0.0605416887944769, + "grad_norm": 4.241743427696131, + "learning_rate": 6.0212201591511945e-06, + "loss": 0.4551, + "step": 228 + }, + { + "epoch": 0.060807222517259694, + "grad_norm": 7.930922977922742, + "learning_rate": 6.0477453580901854e-06, + "loss": 1.9453, + "step": 229 + }, + { + "epoch": 0.061072756240042485, + "grad_norm": 4.1445314479670925, + "learning_rate": 6.074270557029178e-06, + "loss": 0.4766, + "step": 230 + }, + { + "epoch": 0.06133828996282528, + "grad_norm": 0.022757441023624224, + "learning_rate": 6.1007957559681706e-06, + "loss": 0.0006, + "step": 231 + }, + { + "epoch": 0.06160382368560807, + "grad_norm": 0.04301908301005633, + "learning_rate": 6.127320954907162e-06, + "loss": 0.0012, + "step": 232 + }, + { + "epoch": 0.06186935740839086, + "grad_norm": 3.9993964088934497, + "learning_rate": 6.153846153846155e-06, + "loss": 0.5352, + "step": 233 + }, + { + "epoch": 0.06213489113117366, + "grad_norm": 0.0384831919189867, + "learning_rate": 6.180371352785146e-06, + "loss": 0.0011, + "step": 234 + }, + { + "epoch": 0.06240042485395645, + "grad_norm": 0.11825509016660113, + "learning_rate": 6.206896551724138e-06, + "loss": 0.0033, + "step": 235 + }, + { + "epoch": 0.06266595857673925, + "grad_norm": 0.038756173266160945, + "learning_rate": 6.23342175066313e-06, + "loss": 0.0011, + "step": 236 + }, + { + "epoch": 0.06293149229952204, + "grad_norm": 0.0527621505367311, + "learning_rate": 6.259946949602123e-06, + "loss": 0.0016, + "step": 237 + }, + { + "epoch": 0.06319702602230483, + "grad_norm": 3.4032460542743235, + "learning_rate": 6.286472148541115e-06, + "loss": 0.4199, + "step": 238 + }, + { + "epoch": 0.06346255974508763, + "grad_norm": 1.5026047777466172, + "learning_rate": 6.312997347480107e-06, + "loss": 0.006, + "step": 239 + }, + { + "epoch": 0.06372809346787042, + "grad_norm": 0.09772198768152848, + "learning_rate": 6.339522546419099e-06, + "loss": 0.0028, + "step": 240 + }, + { + "epoch": 0.06399362719065321, + "grad_norm": 3.4974570386194848, + "learning_rate": 6.36604774535809e-06, + "loss": 0.3203, + "step": 241 + }, + { + "epoch": 0.064259160913436, + "grad_norm": 0.05852008470517513, + "learning_rate": 6.392572944297083e-06, + "loss": 0.0017, + "step": 242 + }, + { + "epoch": 0.0645246946362188, + "grad_norm": 0.6206245208833137, + "learning_rate": 6.419098143236075e-06, + "loss": 0.0187, + "step": 243 + }, + { + "epoch": 0.0647902283590016, + "grad_norm": 3.257984380379068, + "learning_rate": 6.445623342175067e-06, + "loss": 0.2988, + "step": 244 + }, + { + "epoch": 0.06505576208178439, + "grad_norm": 3.129808256412029, + "learning_rate": 6.47214854111406e-06, + "loss": 0.3672, + "step": 245 + }, + { + "epoch": 0.06532129580456718, + "grad_norm": 1.0556539855013083, + "learning_rate": 6.498673740053051e-06, + "loss": 0.0258, + "step": 246 + }, + { + "epoch": 0.06558682952734997, + "grad_norm": 2.2148968544822747, + "learning_rate": 6.525198938992043e-06, + "loss": 0.1367, + "step": 247 + }, + { + "epoch": 0.06585236325013277, + "grad_norm": 0.3150693176270268, + "learning_rate": 6.551724137931035e-06, + "loss": 0.0103, + "step": 248 + }, + { + "epoch": 0.06611789697291556, + "grad_norm": 0.9150893306492274, + "learning_rate": 6.5782493368700276e-06, + "loss": 0.0356, + "step": 249 + }, + { + "epoch": 0.06638343069569835, + "grad_norm": 0.28640145943412554, + "learning_rate": 6.6047745358090184e-06, + "loss": 0.0092, + "step": 250 + }, + { + "epoch": 0.06664896441848114, + "grad_norm": 0.463345992401992, + "learning_rate": 6.631299734748011e-06, + "loss": 0.0165, + "step": 251 + }, + { + "epoch": 0.06691449814126393, + "grad_norm": 1.1399589305877227, + "learning_rate": 6.657824933687003e-06, + "loss": 0.0393, + "step": 252 + }, + { + "epoch": 0.06718003186404674, + "grad_norm": 2.377572235801305, + "learning_rate": 6.684350132625995e-06, + "loss": 0.1641, + "step": 253 + }, + { + "epoch": 0.06744556558682953, + "grad_norm": 0.20240560295679277, + "learning_rate": 6.710875331564988e-06, + "loss": 0.0053, + "step": 254 + }, + { + "epoch": 0.06771109930961232, + "grad_norm": 0.5580660153168894, + "learning_rate": 6.737400530503979e-06, + "loss": 0.015, + "step": 255 + }, + { + "epoch": 0.06797663303239511, + "grad_norm": 0.34949012939902974, + "learning_rate": 6.763925729442971e-06, + "loss": 0.0104, + "step": 256 + }, + { + "epoch": 0.06824216675517791, + "grad_norm": 1.5003862236593613, + "learning_rate": 6.790450928381963e-06, + "loss": 0.0688, + "step": 257 + }, + { + "epoch": 0.0685077004779607, + "grad_norm": 3.1203129924267152, + "learning_rate": 6.816976127320956e-06, + "loss": 0.2246, + "step": 258 + }, + { + "epoch": 0.0687732342007435, + "grad_norm": 1.051643999026055, + "learning_rate": 6.843501326259947e-06, + "loss": 0.0391, + "step": 259 + }, + { + "epoch": 0.06903876792352628, + "grad_norm": 0.19030670002007702, + "learning_rate": 6.87002652519894e-06, + "loss": 0.0058, + "step": 260 + }, + { + "epoch": 0.06930430164630907, + "grad_norm": 0.6772208181131325, + "learning_rate": 6.896551724137932e-06, + "loss": 0.0206, + "step": 261 + }, + { + "epoch": 0.06956983536909188, + "grad_norm": 2.2887370072088564, + "learning_rate": 6.923076923076923e-06, + "loss": 0.1035, + "step": 262 + }, + { + "epoch": 0.06983536909187467, + "grad_norm": 0.6143154696309661, + "learning_rate": 6.949602122015916e-06, + "loss": 0.0187, + "step": 263 + }, + { + "epoch": 0.07010090281465746, + "grad_norm": 2.9212908220956946, + "learning_rate": 6.976127320954908e-06, + "loss": 0.1641, + "step": 264 + }, + { + "epoch": 0.07036643653744025, + "grad_norm": 0.05930635612845523, + "learning_rate": 7.0026525198939e-06, + "loss": 0.0016, + "step": 265 + }, + { + "epoch": 0.07063197026022305, + "grad_norm": 0.21269981419083292, + "learning_rate": 7.029177718832891e-06, + "loss": 0.0045, + "step": 266 + }, + { + "epoch": 0.07089750398300584, + "grad_norm": 3.3491904755168838, + "learning_rate": 7.055702917771884e-06, + "loss": 0.2461, + "step": 267 + }, + { + "epoch": 0.07116303770578863, + "grad_norm": 3.011992184448476, + "learning_rate": 7.082228116710876e-06, + "loss": 0.2246, + "step": 268 + }, + { + "epoch": 0.07142857142857142, + "grad_norm": 3.317414315449119, + "learning_rate": 7.108753315649868e-06, + "loss": 0.2344, + "step": 269 + }, + { + "epoch": 0.07169410515135423, + "grad_norm": 0.3112825686788188, + "learning_rate": 7.1352785145888606e-06, + "loss": 0.0093, + "step": 270 + }, + { + "epoch": 0.07195963887413702, + "grad_norm": 0.05440553438613368, + "learning_rate": 7.1618037135278515e-06, + "loss": 0.0017, + "step": 271 + }, + { + "epoch": 0.07222517259691981, + "grad_norm": 2.1971208118181416, + "learning_rate": 7.188328912466844e-06, + "loss": 0.1367, + "step": 272 + }, + { + "epoch": 0.0724907063197026, + "grad_norm": 12.646903729517486, + "learning_rate": 7.214854111405836e-06, + "loss": 1.625, + "step": 273 + }, + { + "epoch": 0.07275624004248539, + "grad_norm": 0.5134455047996075, + "learning_rate": 7.241379310344828e-06, + "loss": 0.0166, + "step": 274 + }, + { + "epoch": 0.0730217737652682, + "grad_norm": 0.7126267670034715, + "learning_rate": 7.267904509283821e-06, + "loss": 0.0288, + "step": 275 + }, + { + "epoch": 0.07328730748805098, + "grad_norm": 2.247754974205674, + "learning_rate": 7.294429708222812e-06, + "loss": 0.1367, + "step": 276 + }, + { + "epoch": 0.07355284121083377, + "grad_norm": 2.0961429578052546, + "learning_rate": 7.320954907161804e-06, + "loss": 0.0967, + "step": 277 + }, + { + "epoch": 0.07381837493361657, + "grad_norm": 1.5658452401665721, + "learning_rate": 7.347480106100796e-06, + "loss": 0.0757, + "step": 278 + }, + { + "epoch": 0.07408390865639937, + "grad_norm": 1.3474463383395376, + "learning_rate": 7.374005305039789e-06, + "loss": 0.0635, + "step": 279 + }, + { + "epoch": 0.07434944237918216, + "grad_norm": 0.9079733138650365, + "learning_rate": 7.40053050397878e-06, + "loss": 0.0354, + "step": 280 + }, + { + "epoch": 0.07461497610196495, + "grad_norm": 1.423025471379327, + "learning_rate": 7.427055702917773e-06, + "loss": 0.0579, + "step": 281 + }, + { + "epoch": 0.07488050982474774, + "grad_norm": 29.742858003866257, + "learning_rate": 7.453580901856765e-06, + "loss": 2.2188, + "step": 282 + }, + { + "epoch": 0.07514604354753053, + "grad_norm": 1.1822814883714525, + "learning_rate": 7.480106100795756e-06, + "loss": 0.0579, + "step": 283 + }, + { + "epoch": 0.07541157727031333, + "grad_norm": 1.183988881108207, + "learning_rate": 7.506631299734749e-06, + "loss": 0.053, + "step": 284 + }, + { + "epoch": 0.07567711099309612, + "grad_norm": 1.2925771262928738, + "learning_rate": 7.533156498673741e-06, + "loss": 0.0579, + "step": 285 + }, + { + "epoch": 0.07594264471587892, + "grad_norm": 0.6760998544572231, + "learning_rate": 7.559681697612733e-06, + "loss": 0.0208, + "step": 286 + }, + { + "epoch": 0.0762081784386617, + "grad_norm": 1.0282421815983582, + "learning_rate": 7.586206896551724e-06, + "loss": 0.0396, + "step": 287 + }, + { + "epoch": 0.07647371216144451, + "grad_norm": 0.4234422853011549, + "learning_rate": 7.612732095490717e-06, + "loss": 0.015, + "step": 288 + }, + { + "epoch": 0.0767392458842273, + "grad_norm": 0.39859873725327066, + "learning_rate": 7.639257294429708e-06, + "loss": 0.0135, + "step": 289 + }, + { + "epoch": 0.07700477960701009, + "grad_norm": 0.16222305800718434, + "learning_rate": 7.6657824933687e-06, + "loss": 0.0052, + "step": 290 + }, + { + "epoch": 0.07727031332979288, + "grad_norm": 0.11630227304441901, + "learning_rate": 7.692307692307694e-06, + "loss": 0.0038, + "step": 291 + }, + { + "epoch": 0.07753584705257567, + "grad_norm": 0.03372528156970035, + "learning_rate": 7.718832891246685e-06, + "loss": 0.0008, + "step": 292 + }, + { + "epoch": 0.07780138077535848, + "grad_norm": 0.051940974929390066, + "learning_rate": 7.745358090185677e-06, + "loss": 0.0015, + "step": 293 + }, + { + "epoch": 0.07806691449814127, + "grad_norm": 3.0783353845274295, + "learning_rate": 7.771883289124669e-06, + "loss": 0.2988, + "step": 294 + }, + { + "epoch": 0.07833244822092406, + "grad_norm": 3.19288706413644, + "learning_rate": 7.79840848806366e-06, + "loss": 1.5, + "step": 295 + }, + { + "epoch": 0.07859798194370685, + "grad_norm": 0.043571411993144044, + "learning_rate": 7.824933687002652e-06, + "loss": 0.0012, + "step": 296 + }, + { + "epoch": 0.07886351566648965, + "grad_norm": 0.022251985441288053, + "learning_rate": 7.851458885941646e-06, + "loss": 0.0006, + "step": 297 + }, + { + "epoch": 0.07912904938927244, + "grad_norm": 0.08604971258954734, + "learning_rate": 7.877984084880637e-06, + "loss": 0.002, + "step": 298 + }, + { + "epoch": 0.07939458311205523, + "grad_norm": 0.019663492930382675, + "learning_rate": 7.904509283819629e-06, + "loss": 0.0007, + "step": 299 + }, + { + "epoch": 0.07966011683483802, + "grad_norm": 4.060927451869871, + "learning_rate": 7.93103448275862e-06, + "loss": 0.5469, + "step": 300 + }, + { + "epoch": 0.07992565055762081, + "grad_norm": 3.7022259212796285, + "learning_rate": 7.957559681697613e-06, + "loss": 0.5, + "step": 301 + }, + { + "epoch": 0.08019118428040362, + "grad_norm": 0.009537116683502751, + "learning_rate": 7.984084880636606e-06, + "loss": 0.0002, + "step": 302 + }, + { + "epoch": 0.0804567180031864, + "grad_norm": 0.15972134226047616, + "learning_rate": 8.010610079575598e-06, + "loss": 0.0013, + "step": 303 + }, + { + "epoch": 0.0807222517259692, + "grad_norm": 3.667758151535006, + "learning_rate": 8.03713527851459e-06, + "loss": 0.3203, + "step": 304 + }, + { + "epoch": 0.08098778544875199, + "grad_norm": 0.11182970657148031, + "learning_rate": 8.063660477453583e-06, + "loss": 0.0036, + "step": 305 + }, + { + "epoch": 0.08125331917153479, + "grad_norm": 2.9682235846476854, + "learning_rate": 8.090185676392573e-06, + "loss": 0.3203, + "step": 306 + }, + { + "epoch": 0.08151885289431758, + "grad_norm": 2.6544777345452233, + "learning_rate": 8.116710875331566e-06, + "loss": 0.2148, + "step": 307 + }, + { + "epoch": 0.08178438661710037, + "grad_norm": 2.23551129108633, + "learning_rate": 8.143236074270558e-06, + "loss": 0.2129, + "step": 308 + }, + { + "epoch": 0.08204992033988316, + "grad_norm": 12.154843924500748, + "learning_rate": 8.16976127320955e-06, + "loss": 1.6797, + "step": 309 + }, + { + "epoch": 0.08231545406266595, + "grad_norm": 0.15211428151282252, + "learning_rate": 8.196286472148541e-06, + "loss": 0.006, + "step": 310 + }, + { + "epoch": 0.08258098778544876, + "grad_norm": 1.8572071183861802, + "learning_rate": 8.222811671087533e-06, + "loss": 0.1641, + "step": 311 + }, + { + "epoch": 0.08284652150823155, + "grad_norm": 0.04756323667504141, + "learning_rate": 8.249336870026527e-06, + "loss": 0.0015, + "step": 312 + }, + { + "epoch": 0.08311205523101434, + "grad_norm": 1.2740195810197956, + "learning_rate": 8.275862068965518e-06, + "loss": 0.0752, + "step": 313 + }, + { + "epoch": 0.08337758895379713, + "grad_norm": 2.2505964862303536, + "learning_rate": 8.30238726790451e-06, + "loss": 0.1553, + "step": 314 + }, + { + "epoch": 0.08364312267657993, + "grad_norm": 1.2815168496791627, + "learning_rate": 8.328912466843502e-06, + "loss": 0.0359, + "step": 315 + }, + { + "epoch": 0.08390865639936272, + "grad_norm": 0.6640260863661916, + "learning_rate": 8.355437665782494e-06, + "loss": 0.0356, + "step": 316 + }, + { + "epoch": 0.08417419012214551, + "grad_norm": 1.5201761587838876, + "learning_rate": 8.381962864721485e-06, + "loss": 0.1123, + "step": 317 + }, + { + "epoch": 0.0844397238449283, + "grad_norm": 8.490220248388562, + "learning_rate": 8.408488063660479e-06, + "loss": 1.625, + "step": 318 + }, + { + "epoch": 0.08470525756771109, + "grad_norm": 1.220932640130876, + "learning_rate": 8.43501326259947e-06, + "loss": 0.0762, + "step": 319 + }, + { + "epoch": 0.0849707912904939, + "grad_norm": 1.5989309966218423, + "learning_rate": 8.461538461538462e-06, + "loss": 0.1191, + "step": 320 + }, + { + "epoch": 0.08523632501327669, + "grad_norm": 0.9427180577076081, + "learning_rate": 8.488063660477454e-06, + "loss": 0.0583, + "step": 321 + }, + { + "epoch": 0.08550185873605948, + "grad_norm": 0.5733118835063594, + "learning_rate": 8.514588859416446e-06, + "loss": 0.0293, + "step": 322 + }, + { + "epoch": 0.08576739245884227, + "grad_norm": 0.14144397139040923, + "learning_rate": 8.541114058355439e-06, + "loss": 0.0066, + "step": 323 + }, + { + "epoch": 0.08603292618162507, + "grad_norm": 2.2159972152363077, + "learning_rate": 8.56763925729443e-06, + "loss": 0.2676, + "step": 324 + }, + { + "epoch": 0.08629845990440786, + "grad_norm": 1.6089684506296742, + "learning_rate": 8.594164456233422e-06, + "loss": 0.165, + "step": 325 + }, + { + "epoch": 0.08656399362719065, + "grad_norm": 1.907931611057086, + "learning_rate": 8.620689655172414e-06, + "loss": 0.2246, + "step": 326 + }, + { + "epoch": 0.08682952734997344, + "grad_norm": 1.3090038245570277, + "learning_rate": 8.647214854111406e-06, + "loss": 0.0894, + "step": 327 + }, + { + "epoch": 0.08709506107275625, + "grad_norm": 0.32857189533914255, + "learning_rate": 8.6737400530504e-06, + "loss": 0.0167, + "step": 328 + }, + { + "epoch": 0.08736059479553904, + "grad_norm": 1.6452882577007435, + "learning_rate": 8.700265251989391e-06, + "loss": 0.1934, + "step": 329 + }, + { + "epoch": 0.08762612851832183, + "grad_norm": 0.9569297269958028, + "learning_rate": 8.726790450928383e-06, + "loss": 0.0583, + "step": 330 + }, + { + "epoch": 0.08789166224110462, + "grad_norm": 1.5135689038875453, + "learning_rate": 8.753315649867374e-06, + "loss": 0.1641, + "step": 331 + }, + { + "epoch": 0.08815719596388741, + "grad_norm": 1.171616362989181, + "learning_rate": 8.779840848806366e-06, + "loss": 0.0967, + "step": 332 + }, + { + "epoch": 0.08842272968667021, + "grad_norm": 1.296129054455011, + "learning_rate": 8.806366047745358e-06, + "loss": 0.1279, + "step": 333 + }, + { + "epoch": 0.088688263409453, + "grad_norm": 1.7346960331999244, + "learning_rate": 8.832891246684351e-06, + "loss": 0.2031, + "step": 334 + }, + { + "epoch": 0.08895379713223579, + "grad_norm": 1.0461737141624319, + "learning_rate": 8.859416445623343e-06, + "loss": 0.0894, + "step": 335 + }, + { + "epoch": 0.08921933085501858, + "grad_norm": 0.6479647062091533, + "learning_rate": 8.885941644562335e-06, + "loss": 0.0439, + "step": 336 + }, + { + "epoch": 0.08948486457780139, + "grad_norm": 0.734572554709422, + "learning_rate": 8.912466843501327e-06, + "loss": 0.0479, + "step": 337 + }, + { + "epoch": 0.08975039830058418, + "grad_norm": 0.37460636172952666, + "learning_rate": 8.938992042440318e-06, + "loss": 0.0239, + "step": 338 + }, + { + "epoch": 0.09001593202336697, + "grad_norm": 0.7525268871585711, + "learning_rate": 8.965517241379312e-06, + "loss": 0.0579, + "step": 339 + }, + { + "epoch": 0.09028146574614976, + "grad_norm": 0.8611894174379208, + "learning_rate": 8.992042440318303e-06, + "loss": 0.0757, + "step": 340 + }, + { + "epoch": 0.09054699946893255, + "grad_norm": 0.7524042144808409, + "learning_rate": 9.018567639257295e-06, + "loss": 0.0579, + "step": 341 + }, + { + "epoch": 0.09081253319171535, + "grad_norm": 0.7389515011548652, + "learning_rate": 9.045092838196289e-06, + "loss": 0.0579, + "step": 342 + }, + { + "epoch": 0.09107806691449814, + "grad_norm": 0.717809617473176, + "learning_rate": 9.071618037135279e-06, + "loss": 0.0583, + "step": 343 + }, + { + "epoch": 0.09134360063728093, + "grad_norm": 3.9747880808948737, + "learning_rate": 9.098143236074272e-06, + "loss": 1.2812, + "step": 344 + }, + { + "epoch": 0.09160913436006372, + "grad_norm": 0.5440174756085783, + "learning_rate": 9.124668435013264e-06, + "loss": 0.0396, + "step": 345 + }, + { + "epoch": 0.09187466808284653, + "grad_norm": 0.5421286985756604, + "learning_rate": 9.151193633952255e-06, + "loss": 0.04, + "step": 346 + }, + { + "epoch": 0.09214020180562932, + "grad_norm": 0.7819761791319009, + "learning_rate": 9.177718832891247e-06, + "loss": 0.0698, + "step": 347 + }, + { + "epoch": 0.09240573552841211, + "grad_norm": 0.6347909735793549, + "learning_rate": 9.204244031830239e-06, + "loss": 0.0532, + "step": 348 + }, + { + "epoch": 0.0926712692511949, + "grad_norm": 6.09104951580698, + "learning_rate": 9.230769230769232e-06, + "loss": 1.3984, + "step": 349 + }, + { + "epoch": 0.09293680297397769, + "grad_norm": 1.041426014412066, + "learning_rate": 9.257294429708224e-06, + "loss": 0.1045, + "step": 350 + }, + { + "epoch": 0.0932023366967605, + "grad_norm": 0.47301291400697715, + "learning_rate": 9.283819628647216e-06, + "loss": 0.027, + "step": 351 + }, + { + "epoch": 0.09346787041954328, + "grad_norm": 0.8295019698863545, + "learning_rate": 9.310344827586207e-06, + "loss": 0.0693, + "step": 352 + }, + { + "epoch": 0.09373340414232607, + "grad_norm": 0.5101258233711066, + "learning_rate": 9.3368700265252e-06, + "loss": 0.0361, + "step": 353 + }, + { + "epoch": 0.09399893786510886, + "grad_norm": 0.846504832852156, + "learning_rate": 9.363395225464191e-06, + "loss": 0.0757, + "step": 354 + }, + { + "epoch": 0.09426447158789167, + "grad_norm": 1.228530334987593, + "learning_rate": 9.389920424403184e-06, + "loss": 0.0825, + "step": 355 + }, + { + "epoch": 0.09453000531067446, + "grad_norm": 1.0130610331102055, + "learning_rate": 9.416445623342176e-06, + "loss": 0.0903, + "step": 356 + }, + { + "epoch": 0.09479553903345725, + "grad_norm": 3.315689651222163, + "learning_rate": 9.442970822281168e-06, + "loss": 0.0698, + "step": 357 + }, + { + "epoch": 0.09506107275624004, + "grad_norm": 1.1215958152396903, + "learning_rate": 9.46949602122016e-06, + "loss": 0.0967, + "step": 358 + }, + { + "epoch": 0.09532660647902283, + "grad_norm": 0.9299426765923084, + "learning_rate": 9.496021220159151e-06, + "loss": 0.0767, + "step": 359 + }, + { + "epoch": 0.09559214020180563, + "grad_norm": 0.7265450766257645, + "learning_rate": 9.522546419098145e-06, + "loss": 0.0588, + "step": 360 + }, + { + "epoch": 0.09585767392458842, + "grad_norm": 0.8861616356173306, + "learning_rate": 9.549071618037136e-06, + "loss": 0.0894, + "step": 361 + }, + { + "epoch": 0.09612320764737121, + "grad_norm": 0.5801872302587202, + "learning_rate": 9.575596816976128e-06, + "loss": 0.0398, + "step": 362 + }, + { + "epoch": 0.096388741370154, + "grad_norm": 0.8463320496115545, + "learning_rate": 9.60212201591512e-06, + "loss": 0.0698, + "step": 363 + }, + { + "epoch": 0.09665427509293681, + "grad_norm": 1.0159728812303754, + "learning_rate": 9.628647214854112e-06, + "loss": 0.0825, + "step": 364 + }, + { + "epoch": 0.0969198088157196, + "grad_norm": 0.716278656041375, + "learning_rate": 9.655172413793105e-06, + "loss": 0.064, + "step": 365 + }, + { + "epoch": 0.09718534253850239, + "grad_norm": 0.9734713463730997, + "learning_rate": 9.681697612732097e-06, + "loss": 0.1035, + "step": 366 + }, + { + "epoch": 0.09745087626128518, + "grad_norm": 0.6573062869557463, + "learning_rate": 9.708222811671088e-06, + "loss": 0.0583, + "step": 367 + }, + { + "epoch": 0.09771640998406797, + "grad_norm": 0.4000747396309669, + "learning_rate": 9.73474801061008e-06, + "loss": 0.0259, + "step": 368 + }, + { + "epoch": 0.09798194370685077, + "grad_norm": 0.7259390646225368, + "learning_rate": 9.761273209549072e-06, + "loss": 0.0698, + "step": 369 + }, + { + "epoch": 0.09824747742963356, + "grad_norm": 0.5866962892418395, + "learning_rate": 9.787798408488064e-06, + "loss": 0.0439, + "step": 370 + }, + { + "epoch": 0.09851301115241635, + "grad_norm": 0.5114555343916287, + "learning_rate": 9.814323607427057e-06, + "loss": 0.0354, + "step": 371 + }, + { + "epoch": 0.09877854487519915, + "grad_norm": 0.6409404463637564, + "learning_rate": 9.840848806366049e-06, + "loss": 0.0525, + "step": 372 + }, + { + "epoch": 0.09904407859798195, + "grad_norm": 1.1219844119808053, + "learning_rate": 9.86737400530504e-06, + "loss": 0.1206, + "step": 373 + }, + { + "epoch": 0.09930961232076474, + "grad_norm": 0.553871104594828, + "learning_rate": 9.893899204244032e-06, + "loss": 0.0396, + "step": 374 + }, + { + "epoch": 0.09957514604354753, + "grad_norm": 0.9103969438290072, + "learning_rate": 9.920424403183024e-06, + "loss": 0.0825, + "step": 375 + }, + { + "epoch": 0.09984067976633032, + "grad_norm": 0.6220106264544292, + "learning_rate": 9.946949602122017e-06, + "loss": 0.053, + "step": 376 + }, + { + "epoch": 0.10010621348911311, + "grad_norm": 0.35674878941499344, + "learning_rate": 9.973474801061009e-06, + "loss": 0.0231, + "step": 377 + }, + { + "epoch": 0.10037174721189591, + "grad_norm": 0.829605031932029, + "learning_rate": 1e-05, + "loss": 0.0698, + "step": 378 + }, + { + "epoch": 0.1006372809346787, + "grad_norm": 5.884980972558981, + "learning_rate": 9.999999518029487e-06, + "loss": 1.0781, + "step": 379 + }, + { + "epoch": 0.1009028146574615, + "grad_norm": 0.5316843014168648, + "learning_rate": 9.999998072118034e-06, + "loss": 0.0396, + "step": 380 + }, + { + "epoch": 0.10116834838024429, + "grad_norm": 0.2781221112512736, + "learning_rate": 9.99999566226592e-06, + "loss": 0.0165, + "step": 381 + }, + { + "epoch": 0.10143388210302709, + "grad_norm": 0.48275929349374064, + "learning_rate": 9.999992288473615e-06, + "loss": 0.0393, + "step": 382 + }, + { + "epoch": 0.10169941582580988, + "grad_norm": 0.6278039868358899, + "learning_rate": 9.999987950741766e-06, + "loss": 0.0354, + "step": 383 + }, + { + "epoch": 0.10196494954859267, + "grad_norm": 0.6003466758282822, + "learning_rate": 9.99998264907121e-06, + "loss": 0.0361, + "step": 384 + }, + { + "epoch": 0.10223048327137546, + "grad_norm": 0.8912911423947107, + "learning_rate": 9.999976383462966e-06, + "loss": 0.0579, + "step": 385 + }, + { + "epoch": 0.10249601699415826, + "grad_norm": 0.7329391229020974, + "learning_rate": 9.999969153918246e-06, + "loss": 0.0435, + "step": 386 + }, + { + "epoch": 0.10276155071694106, + "grad_norm": 0.41829397558306736, + "learning_rate": 9.999960960438441e-06, + "loss": 0.0288, + "step": 387 + }, + { + "epoch": 0.10302708443972385, + "grad_norm": 0.9968145270135951, + "learning_rate": 9.999951803025134e-06, + "loss": 0.0757, + "step": 388 + }, + { + "epoch": 0.10329261816250664, + "grad_norm": 0.7676728389082504, + "learning_rate": 9.999941681680087e-06, + "loss": 0.053, + "step": 389 + }, + { + "epoch": 0.10355815188528943, + "grad_norm": 0.31126746105326625, + "learning_rate": 9.999930596405254e-06, + "loss": 0.0184, + "step": 390 + }, + { + "epoch": 0.10382368560807223, + "grad_norm": 0.5108289920792247, + "learning_rate": 9.999918547202769e-06, + "loss": 0.0354, + "step": 391 + }, + { + "epoch": 0.10408921933085502, + "grad_norm": 0.13796497678169306, + "learning_rate": 9.999905534074959e-06, + "loss": 0.0041, + "step": 392 + }, + { + "epoch": 0.10435475305363781, + "grad_norm": 0.7020072318414768, + "learning_rate": 9.999891557024329e-06, + "loss": 0.0435, + "step": 393 + }, + { + "epoch": 0.1046202867764206, + "grad_norm": 0.768122730327616, + "learning_rate": 9.999876616053574e-06, + "loss": 0.0435, + "step": 394 + }, + { + "epoch": 0.1048858204992034, + "grad_norm": 0.3543183385660938, + "learning_rate": 9.999860711165577e-06, + "loss": 0.0167, + "step": 395 + }, + { + "epoch": 0.1051513542219862, + "grad_norm": 0.23030281632455404, + "learning_rate": 9.999843842363402e-06, + "loss": 0.0115, + "step": 396 + }, + { + "epoch": 0.10541688794476899, + "grad_norm": 2.039085244924252, + "learning_rate": 9.999826009650302e-06, + "loss": 0.1279, + "step": 397 + }, + { + "epoch": 0.10568242166755178, + "grad_norm": 0.6863471545307963, + "learning_rate": 9.999807213029716e-06, + "loss": 0.0352, + "step": 398 + }, + { + "epoch": 0.10594795539033457, + "grad_norm": 0.2210748545563893, + "learning_rate": 9.999787452505265e-06, + "loss": 0.0103, + "step": 399 + }, + { + "epoch": 0.10621348911311737, + "grad_norm": 0.5710642294027447, + "learning_rate": 9.99976672808076e-06, + "loss": 0.0286, + "step": 400 + }, + { + "epoch": 0.10647902283590016, + "grad_norm": 1.681840306945433, + "learning_rate": 9.999745039760197e-06, + "loss": 0.1367, + "step": 401 + }, + { + "epoch": 0.10674455655868295, + "grad_norm": 0.5509992577306293, + "learning_rate": 9.999722387547759e-06, + "loss": 0.0076, + "step": 402 + }, + { + "epoch": 0.10701009028146574, + "grad_norm": 0.8731295387203799, + "learning_rate": 9.999698771447808e-06, + "loss": 0.0258, + "step": 403 + }, + { + "epoch": 0.10727562400424855, + "grad_norm": 0.09492425648826142, + "learning_rate": 9.9996741914649e-06, + "loss": 0.0025, + "step": 404 + }, + { + "epoch": 0.10754115772703134, + "grad_norm": 3.5100388797466575, + "learning_rate": 9.999648647603774e-06, + "loss": 0.332, + "step": 405 + }, + { + "epoch": 0.10780669144981413, + "grad_norm": 1.412486941835509, + "learning_rate": 9.999622139869356e-06, + "loss": 0.0688, + "step": 406 + }, + { + "epoch": 0.10807222517259692, + "grad_norm": 0.4019838803661139, + "learning_rate": 9.999594668266754e-06, + "loss": 0.0184, + "step": 407 + }, + { + "epoch": 0.10833775889537971, + "grad_norm": 0.10033143456094161, + "learning_rate": 9.999566232801263e-06, + "loss": 0.0006, + "step": 408 + }, + { + "epoch": 0.10860329261816251, + "grad_norm": 2.671317904601799, + "learning_rate": 9.999536833478367e-06, + "loss": 0.2031, + "step": 409 + }, + { + "epoch": 0.1088688263409453, + "grad_norm": 1.4428641692398059, + "learning_rate": 9.999506470303736e-06, + "loss": 0.0889, + "step": 410 + }, + { + "epoch": 0.10913436006372809, + "grad_norm": 0.8240991675135017, + "learning_rate": 9.999475143283219e-06, + "loss": 0.032, + "step": 411 + }, + { + "epoch": 0.10939989378651088, + "grad_norm": 0.6500544444722121, + "learning_rate": 9.999442852422858e-06, + "loss": 0.0286, + "step": 412 + }, + { + "epoch": 0.10966542750929369, + "grad_norm": 0.43616099112846474, + "learning_rate": 9.999409597728878e-06, + "loss": 0.0165, + "step": 413 + }, + { + "epoch": 0.10993096123207648, + "grad_norm": 0.4976424228594336, + "learning_rate": 9.99937537920769e-06, + "loss": 0.0229, + "step": 414 + }, + { + "epoch": 0.11019649495485927, + "grad_norm": 0.1461188597520504, + "learning_rate": 9.99934019686589e-06, + "loss": 0.0059, + "step": 415 + }, + { + "epoch": 0.11046202867764206, + "grad_norm": 2.5259359528371155, + "learning_rate": 9.999304050710264e-06, + "loss": 0.2461, + "step": 416 + }, + { + "epoch": 0.11072756240042485, + "grad_norm": 0.020285773595493026, + "learning_rate": 9.999266940747776e-06, + "loss": 0.0006, + "step": 417 + }, + { + "epoch": 0.11099309612320765, + "grad_norm": 2.126872840900898, + "learning_rate": 9.999228866985585e-06, + "loss": 0.1367, + "step": 418 + }, + { + "epoch": 0.11125862984599044, + "grad_norm": 0.10900902708813702, + "learning_rate": 9.999189829431029e-06, + "loss": 0.0035, + "step": 419 + }, + { + "epoch": 0.11152416356877323, + "grad_norm": 1.6722307761042388, + "learning_rate": 9.999149828091632e-06, + "loss": 0.1279, + "step": 420 + }, + { + "epoch": 0.11178969729155602, + "grad_norm": 0.3758020820014381, + "learning_rate": 9.99910886297511e-06, + "loss": 0.0164, + "step": 421 + }, + { + "epoch": 0.11205523101433883, + "grad_norm": 0.9309889664471976, + "learning_rate": 9.999066934089355e-06, + "loss": 0.0476, + "step": 422 + }, + { + "epoch": 0.11232076473712162, + "grad_norm": 1.9032139694898287, + "learning_rate": 9.999024041442455e-06, + "loss": 0.1641, + "step": 423 + }, + { + "epoch": 0.11258629845990441, + "grad_norm": 1.9393415053104626, + "learning_rate": 9.998980185042678e-06, + "loss": 0.1367, + "step": 424 + }, + { + "epoch": 0.1128518321826872, + "grad_norm": 1.7335978068394446, + "learning_rate": 9.99893536489848e-06, + "loss": 0.1553, + "step": 425 + }, + { + "epoch": 0.11311736590546999, + "grad_norm": 0.964411541181811, + "learning_rate": 9.998889581018498e-06, + "loss": 0.0525, + "step": 426 + }, + { + "epoch": 0.11338289962825279, + "grad_norm": 0.9732904594255166, + "learning_rate": 9.998842833411565e-06, + "loss": 0.0479, + "step": 427 + }, + { + "epoch": 0.11364843335103558, + "grad_norm": 0.3840885691796018, + "learning_rate": 9.998795122086687e-06, + "loss": 0.0184, + "step": 428 + }, + { + "epoch": 0.11391396707381837, + "grad_norm": 0.47672162986825717, + "learning_rate": 9.998746447053063e-06, + "loss": 0.0259, + "step": 429 + }, + { + "epoch": 0.11417950079660116, + "grad_norm": 0.06479878514865638, + "learning_rate": 9.998696808320083e-06, + "loss": 0.0018, + "step": 430 + }, + { + "epoch": 0.11444503451938397, + "grad_norm": 0.8617385798684972, + "learning_rate": 9.99864620589731e-06, + "loss": 0.0576, + "step": 431 + }, + { + "epoch": 0.11471056824216676, + "grad_norm": 0.25886899017260007, + "learning_rate": 9.998594639794502e-06, + "loss": 0.0095, + "step": 432 + }, + { + "epoch": 0.11497610196494955, + "grad_norm": 0.6784186872332681, + "learning_rate": 9.9985421100216e-06, + "loss": 0.0393, + "step": 433 + }, + { + "epoch": 0.11524163568773234, + "grad_norm": 1.7431553785848855, + "learning_rate": 9.998488616588733e-06, + "loss": 0.1455, + "step": 434 + }, + { + "epoch": 0.11550716941051513, + "grad_norm": 1.9060354513298985, + "learning_rate": 9.998434159506211e-06, + "loss": 0.1934, + "step": 435 + }, + { + "epoch": 0.11577270313329793, + "grad_norm": 0.19317549952322358, + "learning_rate": 9.998378738784536e-06, + "loss": 0.0092, + "step": 436 + }, + { + "epoch": 0.11603823685608072, + "grad_norm": 0.7688675078104924, + "learning_rate": 9.998322354434388e-06, + "loss": 0.0432, + "step": 437 + }, + { + "epoch": 0.11630377057886351, + "grad_norm": 0.7369207362136836, + "learning_rate": 9.998265006466642e-06, + "loss": 0.0432, + "step": 438 + }, + { + "epoch": 0.1165693043016463, + "grad_norm": 0.29436457021835755, + "learning_rate": 9.998206694892352e-06, + "loss": 0.0146, + "step": 439 + }, + { + "epoch": 0.11683483802442911, + "grad_norm": 0.9124118545569474, + "learning_rate": 9.998147419722759e-06, + "loss": 0.0525, + "step": 440 + }, + { + "epoch": 0.1171003717472119, + "grad_norm": 1.1218128409435468, + "learning_rate": 9.99808718096929e-06, + "loss": 0.0635, + "step": 441 + }, + { + "epoch": 0.11736590546999469, + "grad_norm": 1.149876905466913, + "learning_rate": 9.99802597864356e-06, + "loss": 0.0752, + "step": 442 + }, + { + "epoch": 0.11763143919277748, + "grad_norm": 1.0569853424545377, + "learning_rate": 9.997963812757368e-06, + "loss": 0.0635, + "step": 443 + }, + { + "epoch": 0.11789697291556028, + "grad_norm": 0.8351903364180481, + "learning_rate": 9.997900683322698e-06, + "loss": 0.0476, + "step": 444 + }, + { + "epoch": 0.11816250663834307, + "grad_norm": 1.7702924922542935, + "learning_rate": 9.997836590351721e-06, + "loss": 0.2246, + "step": 445 + }, + { + "epoch": 0.11842804036112586, + "grad_norm": 0.14841631560098084, + "learning_rate": 9.997771533856794e-06, + "loss": 0.0072, + "step": 446 + }, + { + "epoch": 0.11869357408390865, + "grad_norm": 0.2930240108885635, + "learning_rate": 9.997705513850458e-06, + "loss": 0.0165, + "step": 447 + }, + { + "epoch": 0.11895910780669144, + "grad_norm": 0.027324664896434907, + "learning_rate": 9.997638530345442e-06, + "loss": 0.001, + "step": 448 + }, + { + "epoch": 0.11922464152947425, + "grad_norm": 0.8050072457659047, + "learning_rate": 9.997570583354659e-06, + "loss": 0.0435, + "step": 449 + }, + { + "epoch": 0.11949017525225704, + "grad_norm": 0.8260959431214968, + "learning_rate": 9.997501672891208e-06, + "loss": 0.0579, + "step": 450 + }, + { + "epoch": 0.11975570897503983, + "grad_norm": 0.050172710801976175, + "learning_rate": 9.997431798968376e-06, + "loss": 0.0021, + "step": 451 + }, + { + "epoch": 0.12002124269782262, + "grad_norm": 0.3333884739733433, + "learning_rate": 9.99736096159963e-06, + "loss": 0.0147, + "step": 452 + }, + { + "epoch": 0.12028677642060542, + "grad_norm": 0.6853848215343389, + "learning_rate": 9.997289160798629e-06, + "loss": 0.0231, + "step": 453 + }, + { + "epoch": 0.12055231014338821, + "grad_norm": 6.935657752668328, + "learning_rate": 9.997216396579217e-06, + "loss": 1.2891, + "step": 454 + }, + { + "epoch": 0.120817843866171, + "grad_norm": 1.2388612554211353, + "learning_rate": 9.99714266895542e-06, + "loss": 0.0579, + "step": 455 + }, + { + "epoch": 0.1210833775889538, + "grad_norm": 0.8108343213894442, + "learning_rate": 9.997067977941453e-06, + "loss": 0.0354, + "step": 456 + }, + { + "epoch": 0.12134891131173658, + "grad_norm": 0.12285291127516768, + "learning_rate": 9.996992323551713e-06, + "loss": 0.0057, + "step": 457 + }, + { + "epoch": 0.12161444503451939, + "grad_norm": 1.8644514519390076, + "learning_rate": 9.996915705800786e-06, + "loss": 0.2031, + "step": 458 + }, + { + "epoch": 0.12187997875730218, + "grad_norm": 0.4192870410785925, + "learning_rate": 9.996838124703448e-06, + "loss": 0.0118, + "step": 459 + }, + { + "epoch": 0.12214551248008497, + "grad_norm": 13.085772336960922, + "learning_rate": 9.996759580274649e-06, + "loss": 1.1562, + "step": 460 + }, + { + "epoch": 0.12241104620286776, + "grad_norm": 1.9383018555391174, + "learning_rate": 9.996680072529536e-06, + "loss": 0.1455, + "step": 461 + }, + { + "epoch": 0.12267657992565056, + "grad_norm": 0.5114158494450113, + "learning_rate": 9.996599601483435e-06, + "loss": 0.0166, + "step": 462 + }, + { + "epoch": 0.12294211364843335, + "grad_norm": 2.071144791223387, + "learning_rate": 9.996518167151858e-06, + "loss": 0.3105, + "step": 463 + }, + { + "epoch": 0.12320764737121614, + "grad_norm": 0.3826668161577709, + "learning_rate": 9.996435769550509e-06, + "loss": 0.0184, + "step": 464 + }, + { + "epoch": 0.12347318109399893, + "grad_norm": 0.11580831376705496, + "learning_rate": 9.99635240869527e-06, + "loss": 0.0057, + "step": 465 + }, + { + "epoch": 0.12373871481678173, + "grad_norm": 2.0940806054901633, + "learning_rate": 9.996268084602213e-06, + "loss": 0.1836, + "step": 466 + }, + { + "epoch": 0.12400424853956453, + "grad_norm": 1.7889639735230736, + "learning_rate": 9.996182797287597e-06, + "loss": 0.1934, + "step": 467 + }, + { + "epoch": 0.12426978226234732, + "grad_norm": 0.274643361135419, + "learning_rate": 9.99609654676786e-06, + "loss": 0.0095, + "step": 468 + }, + { + "epoch": 0.12453531598513011, + "grad_norm": 0.37891937937045345, + "learning_rate": 9.996009333059633e-06, + "loss": 0.0206, + "step": 469 + }, + { + "epoch": 0.1248008497079129, + "grad_norm": 0.5048830191970928, + "learning_rate": 9.995921156179729e-06, + "loss": 0.0229, + "step": 470 + }, + { + "epoch": 0.1250663834306957, + "grad_norm": 2.793136832206918, + "learning_rate": 9.995832016145149e-06, + "loss": 0.2988, + "step": 471 + }, + { + "epoch": 0.1253319171534785, + "grad_norm": 0.9181078850272798, + "learning_rate": 9.995741912973073e-06, + "loss": 0.0479, + "step": 472 + }, + { + "epoch": 0.12559745087626129, + "grad_norm": 0.17214860190032075, + "learning_rate": 9.995650846680878e-06, + "loss": 0.0025, + "step": 473 + }, + { + "epoch": 0.12586298459904408, + "grad_norm": 0.1930329238705798, + "learning_rate": 9.995558817286117e-06, + "loss": 0.0092, + "step": 474 + }, + { + "epoch": 0.12612851832182687, + "grad_norm": 0.3890866531462164, + "learning_rate": 9.995465824806534e-06, + "loss": 0.0206, + "step": 475 + }, + { + "epoch": 0.12639405204460966, + "grad_norm": 0.294426458994664, + "learning_rate": 9.995371869260057e-06, + "loss": 0.0146, + "step": 476 + }, + { + "epoch": 0.12665958576739245, + "grad_norm": 0.49377766768440445, + "learning_rate": 9.995276950664796e-06, + "loss": 0.0234, + "step": 477 + }, + { + "epoch": 0.12692511949017526, + "grad_norm": 1.6403902226241507, + "learning_rate": 9.995181069039055e-06, + "loss": 0.1035, + "step": 478 + }, + { + "epoch": 0.12719065321295805, + "grad_norm": 2.3189750279643517, + "learning_rate": 9.995084224401316e-06, + "loss": 0.1934, + "step": 479 + }, + { + "epoch": 0.12745618693574085, + "grad_norm": 4.2411804627535385, + "learning_rate": 9.99498641677025e-06, + "loss": 0.7969, + "step": 480 + }, + { + "epoch": 0.12772172065852364, + "grad_norm": 0.6775891633342407, + "learning_rate": 9.994887646164712e-06, + "loss": 0.0354, + "step": 481 + }, + { + "epoch": 0.12798725438130643, + "grad_norm": 0.3451145426368071, + "learning_rate": 9.994787912603747e-06, + "loss": 0.0167, + "step": 482 + }, + { + "epoch": 0.12825278810408922, + "grad_norm": 0.7928434044000325, + "learning_rate": 9.99468721610658e-06, + "loss": 0.0396, + "step": 483 + }, + { + "epoch": 0.128518321826872, + "grad_norm": 0.35058719889346857, + "learning_rate": 9.994585556692624e-06, + "loss": 0.0188, + "step": 484 + }, + { + "epoch": 0.1287838555496548, + "grad_norm": 0.10110600298106223, + "learning_rate": 9.994482934381478e-06, + "loss": 0.0037, + "step": 485 + }, + { + "epoch": 0.1290493892724376, + "grad_norm": 0.8111935229536589, + "learning_rate": 9.994379349192927e-06, + "loss": 0.0393, + "step": 486 + }, + { + "epoch": 0.1293149229952204, + "grad_norm": 0.36191339865591954, + "learning_rate": 9.994274801146942e-06, + "loss": 0.0165, + "step": 487 + }, + { + "epoch": 0.1295804567180032, + "grad_norm": 1.4295506331809191, + "learning_rate": 9.994169290263675e-06, + "loss": 0.0688, + "step": 488 + }, + { + "epoch": 0.12984599044078599, + "grad_norm": 1.4477435816269884, + "learning_rate": 9.99406281656347e-06, + "loss": 0.1191, + "step": 489 + }, + { + "epoch": 0.13011152416356878, + "grad_norm": 0.43333547219941215, + "learning_rate": 9.993955380066856e-06, + "loss": 0.0168, + "step": 490 + }, + { + "epoch": 0.13037705788635157, + "grad_norm": 1.135520056624293, + "learning_rate": 9.993846980794542e-06, + "loss": 0.063, + "step": 491 + }, + { + "epoch": 0.13064259160913436, + "grad_norm": 0.043875501106062434, + "learning_rate": 9.993737618767426e-06, + "loss": 0.0015, + "step": 492 + }, + { + "epoch": 0.13090812533191715, + "grad_norm": 0.6090554091571819, + "learning_rate": 9.993627294006592e-06, + "loss": 0.0286, + "step": 493 + }, + { + "epoch": 0.13117365905469994, + "grad_norm": 1.2750886594774946, + "learning_rate": 9.993516006533311e-06, + "loss": 0.0576, + "step": 494 + }, + { + "epoch": 0.13143919277748273, + "grad_norm": 0.09866268692375627, + "learning_rate": 9.993403756369037e-06, + "loss": 0.0036, + "step": 495 + }, + { + "epoch": 0.13170472650026555, + "grad_norm": 2.5466564151997866, + "learning_rate": 9.993290543535412e-06, + "loss": 0.2148, + "step": 496 + }, + { + "epoch": 0.13197026022304834, + "grad_norm": 1.0356354392876912, + "learning_rate": 9.993176368054258e-06, + "loss": 0.0476, + "step": 497 + }, + { + "epoch": 0.13223579394583113, + "grad_norm": 0.1524377921395385, + "learning_rate": 9.993061229947591e-06, + "loss": 0.0072, + "step": 498 + }, + { + "epoch": 0.13250132766861392, + "grad_norm": 5.508400351087429, + "learning_rate": 9.992945129237607e-06, + "loss": 1.0156, + "step": 499 + }, + { + "epoch": 0.1327668613913967, + "grad_norm": 1.8391859755052025, + "learning_rate": 9.992828065946687e-06, + "loss": 0.2344, + "step": 500 + }, + { + "epoch": 0.1330323951141795, + "grad_norm": 0.2511312439828377, + "learning_rate": 9.992710040097402e-06, + "loss": 0.01, + "step": 501 + }, + { + "epoch": 0.1332979288369623, + "grad_norm": 1.6141514471649456, + "learning_rate": 9.992591051712503e-06, + "loss": 0.1738, + "step": 502 + }, + { + "epoch": 0.13356346255974508, + "grad_norm": 4.287684776700203, + "learning_rate": 9.992471100814934e-06, + "loss": 0.6641, + "step": 503 + }, + { + "epoch": 0.13382899628252787, + "grad_norm": 0.7498793642360017, + "learning_rate": 9.992350187427814e-06, + "loss": 0.0476, + "step": 504 + }, + { + "epoch": 0.13409453000531069, + "grad_norm": 0.7170932959418976, + "learning_rate": 9.992228311574462e-06, + "loss": 0.0393, + "step": 505 + }, + { + "epoch": 0.13436006372809348, + "grad_norm": 1.0630972502146814, + "learning_rate": 9.992105473278366e-06, + "loss": 0.0688, + "step": 506 + }, + { + "epoch": 0.13462559745087627, + "grad_norm": 1.605822286298563, + "learning_rate": 9.991981672563211e-06, + "loss": 0.0815, + "step": 507 + }, + { + "epoch": 0.13489113117365906, + "grad_norm": 0.606270866795599, + "learning_rate": 9.991856909452866e-06, + "loss": 0.032, + "step": 508 + }, + { + "epoch": 0.13515666489644185, + "grad_norm": 1.3621617997303943, + "learning_rate": 9.99173118397138e-06, + "loss": 0.1035, + "step": 509 + }, + { + "epoch": 0.13542219861922464, + "grad_norm": 0.12342567577614295, + "learning_rate": 9.991604496142997e-06, + "loss": 0.0064, + "step": 510 + }, + { + "epoch": 0.13568773234200743, + "grad_norm": 1.044690642843683, + "learning_rate": 9.991476845992137e-06, + "loss": 0.053, + "step": 511 + }, + { + "epoch": 0.13595326606479022, + "grad_norm": 0.35855152173267857, + "learning_rate": 9.991348233543408e-06, + "loss": 0.0206, + "step": 512 + }, + { + "epoch": 0.136218799787573, + "grad_norm": 1.57497690926948, + "learning_rate": 9.991218658821609e-06, + "loss": 0.2246, + "step": 513 + }, + { + "epoch": 0.13648433351035583, + "grad_norm": 0.1113998089521789, + "learning_rate": 9.991088121851716e-06, + "loss": 0.0057, + "step": 514 + }, + { + "epoch": 0.13674986723313862, + "grad_norm": 0.22281969208169422, + "learning_rate": 9.990956622658901e-06, + "loss": 0.0117, + "step": 515 + }, + { + "epoch": 0.1370154009559214, + "grad_norm": 0.2932759441130008, + "learning_rate": 9.990824161268509e-06, + "loss": 0.0148, + "step": 516 + }, + { + "epoch": 0.1372809346787042, + "grad_norm": 0.10622372878984855, + "learning_rate": 9.990690737706081e-06, + "loss": 0.0057, + "step": 517 + }, + { + "epoch": 0.137546468401487, + "grad_norm": 1.5597393946203308, + "learning_rate": 9.990556351997338e-06, + "loss": 0.1113, + "step": 518 + }, + { + "epoch": 0.13781200212426978, + "grad_norm": 1.9903611428670367, + "learning_rate": 9.99042100416819e-06, + "loss": 0.0703, + "step": 519 + }, + { + "epoch": 0.13807753584705257, + "grad_norm": 0.35832314011240834, + "learning_rate": 9.990284694244729e-06, + "loss": 0.0206, + "step": 520 + }, + { + "epoch": 0.13834306956983536, + "grad_norm": 5.2897686383169455, + "learning_rate": 9.990147422253234e-06, + "loss": 0.8125, + "step": 521 + }, + { + "epoch": 0.13860860329261815, + "grad_norm": 0.11713567466813415, + "learning_rate": 9.990009188220166e-06, + "loss": 0.0039, + "step": 522 + }, + { + "epoch": 0.13887413701540097, + "grad_norm": 0.5823075698755563, + "learning_rate": 9.989869992172183e-06, + "loss": 0.0242, + "step": 523 + }, + { + "epoch": 0.13913967073818376, + "grad_norm": 0.5765768333479371, + "learning_rate": 9.989729834136113e-06, + "loss": 0.0193, + "step": 524 + }, + { + "epoch": 0.13940520446096655, + "grad_norm": 0.14344215523221354, + "learning_rate": 9.989588714138977e-06, + "loss": 0.0039, + "step": 525 + }, + { + "epoch": 0.13967073818374934, + "grad_norm": 1.7019424437997808, + "learning_rate": 9.989446632207989e-06, + "loss": 0.0752, + "step": 526 + }, + { + "epoch": 0.13993627190653213, + "grad_norm": 0.2075958214948594, + "learning_rate": 9.989303588370531e-06, + "loss": 0.0117, + "step": 527 + }, + { + "epoch": 0.14020180562931492, + "grad_norm": 0.15118506962449466, + "learning_rate": 9.989159582654187e-06, + "loss": 0.0082, + "step": 528 + }, + { + "epoch": 0.1404673393520977, + "grad_norm": 0.07268392646364422, + "learning_rate": 9.989014615086716e-06, + "loss": 0.0035, + "step": 529 + }, + { + "epoch": 0.1407328730748805, + "grad_norm": 1.4906910982100858, + "learning_rate": 9.988868685696067e-06, + "loss": 0.2129, + "step": 530 + }, + { + "epoch": 0.14099840679766332, + "grad_norm": 1.4326095317287415, + "learning_rate": 9.988721794510374e-06, + "loss": 0.1836, + "step": 531 + }, + { + "epoch": 0.1412639405204461, + "grad_norm": 0.5130258898072058, + "learning_rate": 9.988573941557956e-06, + "loss": 0.0177, + "step": 532 + }, + { + "epoch": 0.1415294742432289, + "grad_norm": 1.3951302100850165, + "learning_rate": 9.988425126867317e-06, + "loss": 0.1641, + "step": 533 + }, + { + "epoch": 0.1417950079660117, + "grad_norm": 1.5620173297534108, + "learning_rate": 9.988275350467145e-06, + "loss": 0.082, + "step": 534 + }, + { + "epoch": 0.14206054168879448, + "grad_norm": 1.5140269849638728, + "learning_rate": 9.988124612386319e-06, + "loss": 0.1367, + "step": 535 + }, + { + "epoch": 0.14232607541157727, + "grad_norm": 0.6063242319852478, + "learning_rate": 9.987972912653894e-06, + "loss": 0.0393, + "step": 536 + }, + { + "epoch": 0.14259160913436006, + "grad_norm": 0.7374682676713821, + "learning_rate": 9.987820251299121e-06, + "loss": 0.0435, + "step": 537 + }, + { + "epoch": 0.14285714285714285, + "grad_norm": 1.39588526651874, + "learning_rate": 9.987666628351432e-06, + "loss": 0.1553, + "step": 538 + }, + { + "epoch": 0.14312267657992564, + "grad_norm": 2.2114571113292394, + "learning_rate": 9.987512043840437e-06, + "loss": 0.1113, + "step": 539 + }, + { + "epoch": 0.14338821030270846, + "grad_norm": 0.42061182715165285, + "learning_rate": 9.987356497795944e-06, + "loss": 0.0286, + "step": 540 + }, + { + "epoch": 0.14365374402549125, + "grad_norm": 0.9900996921832054, + "learning_rate": 9.987199990247938e-06, + "loss": 0.0752, + "step": 541 + }, + { + "epoch": 0.14391927774827404, + "grad_norm": 0.7148831432764048, + "learning_rate": 9.987042521226594e-06, + "loss": 0.0476, + "step": 542 + }, + { + "epoch": 0.14418481147105683, + "grad_norm": 1.5856605293259478, + "learning_rate": 9.986884090762266e-06, + "loss": 0.1123, + "step": 543 + }, + { + "epoch": 0.14445034519383962, + "grad_norm": 0.17251194671640718, + "learning_rate": 9.986724698885502e-06, + "loss": 0.0058, + "step": 544 + }, + { + "epoch": 0.1447158789166224, + "grad_norm": 0.7025948256764871, + "learning_rate": 9.986564345627027e-06, + "loss": 0.0322, + "step": 545 + }, + { + "epoch": 0.1449814126394052, + "grad_norm": 0.039152169713499795, + "learning_rate": 9.98640303101776e-06, + "loss": 0.0011, + "step": 546 + }, + { + "epoch": 0.145246946362188, + "grad_norm": 0.7346842884881487, + "learning_rate": 9.986240755088797e-06, + "loss": 0.033, + "step": 547 + }, + { + "epoch": 0.14551248008497078, + "grad_norm": 0.8712665193969032, + "learning_rate": 9.986077517871424e-06, + "loss": 0.0476, + "step": 548 + }, + { + "epoch": 0.1457780138077536, + "grad_norm": 0.4876598992779223, + "learning_rate": 9.98591331939711e-06, + "loss": 0.0234, + "step": 549 + }, + { + "epoch": 0.1460435475305364, + "grad_norm": 0.5723125307338557, + "learning_rate": 9.985748159697511e-06, + "loss": 0.0432, + "step": 550 + }, + { + "epoch": 0.14630908125331918, + "grad_norm": 5.026172323034893, + "learning_rate": 9.985582038804469e-06, + "loss": 0.7734, + "step": 551 + }, + { + "epoch": 0.14657461497610197, + "grad_norm": 0.691831344385777, + "learning_rate": 9.985414956750008e-06, + "loss": 0.0289, + "step": 552 + }, + { + "epoch": 0.14684014869888476, + "grad_norm": 0.5313434024030962, + "learning_rate": 9.985246913566343e-06, + "loss": 0.0121, + "step": 553 + }, + { + "epoch": 0.14710568242166755, + "grad_norm": 0.5022681233131463, + "learning_rate": 9.985077909285867e-06, + "loss": 0.0391, + "step": 554 + }, + { + "epoch": 0.14737121614445034, + "grad_norm": 2.8350298800583738, + "learning_rate": 9.984907943941164e-06, + "loss": 0.2988, + "step": 555 + }, + { + "epoch": 0.14763674986723313, + "grad_norm": 0.5845567847490889, + "learning_rate": 9.984737017565003e-06, + "loss": 0.0435, + "step": 556 + }, + { + "epoch": 0.14790228359001592, + "grad_norm": 5.065900056975599, + "learning_rate": 9.984565130190331e-06, + "loss": 0.3242, + "step": 557 + }, + { + "epoch": 0.14816781731279874, + "grad_norm": 0.9255981265488107, + "learning_rate": 9.984392281850293e-06, + "loss": 0.0889, + "step": 558 + }, + { + "epoch": 0.14843335103558153, + "grad_norm": 0.7464338462336892, + "learning_rate": 9.984218472578205e-06, + "loss": 0.0576, + "step": 559 + }, + { + "epoch": 0.14869888475836432, + "grad_norm": 0.9740861593787775, + "learning_rate": 9.984043702407581e-06, + "loss": 0.0957, + "step": 560 + }, + { + "epoch": 0.1489644184811471, + "grad_norm": 0.4188005366867957, + "learning_rate": 9.983867971372113e-06, + "loss": 0.0286, + "step": 561 + }, + { + "epoch": 0.1492299522039299, + "grad_norm": 3.331842808759764, + "learning_rate": 9.983691279505678e-06, + "loss": 0.3398, + "step": 562 + }, + { + "epoch": 0.1494954859267127, + "grad_norm": 4.35094959797938, + "learning_rate": 9.983513626842342e-06, + "loss": 0.6523, + "step": 563 + }, + { + "epoch": 0.14976101964949548, + "grad_norm": 0.5123642715371224, + "learning_rate": 9.983335013416354e-06, + "loss": 0.0391, + "step": 564 + }, + { + "epoch": 0.15002655337227827, + "grad_norm": 0.796553563912584, + "learning_rate": 9.983155439262148e-06, + "loss": 0.0579, + "step": 565 + }, + { + "epoch": 0.15029208709506106, + "grad_norm": 0.3439490402717059, + "learning_rate": 9.982974904414344e-06, + "loss": 0.0229, + "step": 566 + }, + { + "epoch": 0.15055762081784388, + "grad_norm": 0.4106869676123719, + "learning_rate": 9.982793408907747e-06, + "loss": 0.0186, + "step": 567 + }, + { + "epoch": 0.15082315454062667, + "grad_norm": 0.7143183560111053, + "learning_rate": 9.982610952777347e-06, + "loss": 0.0393, + "step": 568 + }, + { + "epoch": 0.15108868826340946, + "grad_norm": 0.6128505770192598, + "learning_rate": 9.982427536058321e-06, + "loss": 0.0476, + "step": 569 + }, + { + "epoch": 0.15135422198619225, + "grad_norm": 0.2585949344039441, + "learning_rate": 9.98224315878603e-06, + "loss": 0.0165, + "step": 570 + }, + { + "epoch": 0.15161975570897504, + "grad_norm": 0.11924071710544426, + "learning_rate": 9.982057820996015e-06, + "loss": 0.0039, + "step": 571 + }, + { + "epoch": 0.15188528943175783, + "grad_norm": 0.6551332258487033, + "learning_rate": 9.98187152272401e-06, + "loss": 0.0391, + "step": 572 + }, + { + "epoch": 0.15215082315454062, + "grad_norm": 2.691563544451101, + "learning_rate": 9.981684264005934e-06, + "loss": 0.2031, + "step": 573 + }, + { + "epoch": 0.1524163568773234, + "grad_norm": 1.360641707061635, + "learning_rate": 9.981496044877882e-06, + "loss": 0.1934, + "step": 574 + }, + { + "epoch": 0.1526818906001062, + "grad_norm": 0.43346474177116345, + "learning_rate": 9.981306865376146e-06, + "loss": 0.0139, + "step": 575 + }, + { + "epoch": 0.15294742432288902, + "grad_norm": 0.24328200960959998, + "learning_rate": 9.981116725537195e-06, + "loss": 0.0165, + "step": 576 + }, + { + "epoch": 0.1532129580456718, + "grad_norm": 1.0426671122461846, + "learning_rate": 9.980925625397686e-06, + "loss": 0.0576, + "step": 577 + }, + { + "epoch": 0.1534784917684546, + "grad_norm": 0.1999935342186825, + "learning_rate": 9.98073356499446e-06, + "loss": 0.0131, + "step": 578 + }, + { + "epoch": 0.1537440254912374, + "grad_norm": 0.7190524116140479, + "learning_rate": 9.980540544364546e-06, + "loss": 0.0525, + "step": 579 + }, + { + "epoch": 0.15400955921402018, + "grad_norm": 2.4567461753510256, + "learning_rate": 9.980346563545156e-06, + "loss": 0.1035, + "step": 580 + }, + { + "epoch": 0.15427509293680297, + "grad_norm": 1.3907008477488672, + "learning_rate": 9.980151622573684e-06, + "loss": 0.1279, + "step": 581 + }, + { + "epoch": 0.15454062665958576, + "grad_norm": 1.6662654026394559, + "learning_rate": 9.979955721487716e-06, + "loss": 0.0752, + "step": 582 + }, + { + "epoch": 0.15480616038236855, + "grad_norm": 0.2905768001459995, + "learning_rate": 9.97975886032502e-06, + "loss": 0.0104, + "step": 583 + }, + { + "epoch": 0.15507169410515134, + "grad_norm": 0.5722935674089386, + "learning_rate": 9.979561039123544e-06, + "loss": 0.0391, + "step": 584 + }, + { + "epoch": 0.15533722782793416, + "grad_norm": 3.9956006342859354, + "learning_rate": 9.979362257921428e-06, + "loss": 0.1699, + "step": 585 + }, + { + "epoch": 0.15560276155071695, + "grad_norm": 1.4427656836880338, + "learning_rate": 9.979162516756995e-06, + "loss": 0.0889, + "step": 586 + }, + { + "epoch": 0.15586829527349974, + "grad_norm": 1.105527580649903, + "learning_rate": 9.978961815668753e-06, + "loss": 0.0752, + "step": 587 + }, + { + "epoch": 0.15613382899628253, + "grad_norm": 0.1918918255795427, + "learning_rate": 9.978760154695393e-06, + "loss": 0.0059, + "step": 588 + }, + { + "epoch": 0.15639936271906532, + "grad_norm": 0.04571838620835277, + "learning_rate": 9.978557533875797e-06, + "loss": 0.0011, + "step": 589 + }, + { + "epoch": 0.1566648964418481, + "grad_norm": 1.9273246134899191, + "learning_rate": 9.978353953249023e-06, + "loss": 0.1641, + "step": 590 + }, + { + "epoch": 0.1569304301646309, + "grad_norm": 1.3054086665266786, + "learning_rate": 9.97814941285432e-06, + "loss": 0.0693, + "step": 591 + }, + { + "epoch": 0.1571959638874137, + "grad_norm": 0.030533978901150845, + "learning_rate": 9.977943912731126e-06, + "loss": 0.0007, + "step": 592 + }, + { + "epoch": 0.15746149761019648, + "grad_norm": 0.9348392424847526, + "learning_rate": 9.977737452919051e-06, + "loss": 0.0525, + "step": 593 + }, + { + "epoch": 0.1577270313329793, + "grad_norm": 0.6672615655023215, + "learning_rate": 9.977530033457906e-06, + "loss": 0.0317, + "step": 594 + }, + { + "epoch": 0.1579925650557621, + "grad_norm": 0.9167557805831286, + "learning_rate": 9.977321654387671e-06, + "loss": 0.0525, + "step": 595 + }, + { + "epoch": 0.15825809877854488, + "grad_norm": 0.5548577985701605, + "learning_rate": 9.977112315748525e-06, + "loss": 0.0391, + "step": 596 + }, + { + "epoch": 0.15852363250132767, + "grad_norm": 1.0741347246530948, + "learning_rate": 9.976902017580824e-06, + "loss": 0.0815, + "step": 597 + }, + { + "epoch": 0.15878916622411046, + "grad_norm": 0.1973076257512853, + "learning_rate": 9.976690759925112e-06, + "loss": 0.0058, + "step": 598 + }, + { + "epoch": 0.15905469994689325, + "grad_norm": 0.37899730498629314, + "learning_rate": 9.976478542822114e-06, + "loss": 0.0187, + "step": 599 + }, + { + "epoch": 0.15932023366967604, + "grad_norm": 0.030062306649064036, + "learning_rate": 9.976265366312746e-06, + "loss": 0.0008, + "step": 600 + }, + { + "epoch": 0.15958576739245883, + "grad_norm": 0.8706157736283716, + "learning_rate": 9.976051230438107e-06, + "loss": 0.063, + "step": 601 + }, + { + "epoch": 0.15985130111524162, + "grad_norm": 0.06252640600648844, + "learning_rate": 9.975836135239475e-06, + "loss": 0.0019, + "step": 602 + }, + { + "epoch": 0.16011683483802444, + "grad_norm": 1.020235806615748, + "learning_rate": 9.975620080758321e-06, + "loss": 0.0576, + "step": 603 + }, + { + "epoch": 0.16038236856080723, + "grad_norm": 0.6968794165000528, + "learning_rate": 9.975403067036298e-06, + "loss": 0.0432, + "step": 604 + }, + { + "epoch": 0.16064790228359002, + "grad_norm": 2.0675935959691527, + "learning_rate": 9.975185094115243e-06, + "loss": 0.1455, + "step": 605 + }, + { + "epoch": 0.1609134360063728, + "grad_norm": 0.07887764560075496, + "learning_rate": 9.97496616203718e-06, + "loss": 0.0014, + "step": 606 + }, + { + "epoch": 0.1611789697291556, + "grad_norm": 1.0038423253490598, + "learning_rate": 9.974746270844313e-06, + "loss": 0.0476, + "step": 607 + }, + { + "epoch": 0.1614445034519384, + "grad_norm": 0.015512665639925233, + "learning_rate": 9.974525420579039e-06, + "loss": 0.0003, + "step": 608 + }, + { + "epoch": 0.16171003717472118, + "grad_norm": 0.1729310979679677, + "learning_rate": 9.97430361128393e-06, + "loss": 0.0065, + "step": 609 + }, + { + "epoch": 0.16197557089750397, + "grad_norm": 0.9007630781441461, + "learning_rate": 9.974080843001755e-06, + "loss": 0.0576, + "step": 610 + }, + { + "epoch": 0.16224110462028676, + "grad_norm": 0.8783601833486903, + "learning_rate": 9.973857115775455e-06, + "loss": 0.0342, + "step": 611 + }, + { + "epoch": 0.16250663834306958, + "grad_norm": 1.7341324095089548, + "learning_rate": 9.973632429648165e-06, + "loss": 0.0752, + "step": 612 + }, + { + "epoch": 0.16277217206585237, + "grad_norm": 1.4666956044492996, + "learning_rate": 9.9734067846632e-06, + "loss": 0.0815, + "step": 613 + }, + { + "epoch": 0.16303770578863516, + "grad_norm": 0.8113376617255486, + "learning_rate": 9.973180180864064e-06, + "loss": 0.0317, + "step": 614 + }, + { + "epoch": 0.16330323951141795, + "grad_norm": 2.5241610625807303, + "learning_rate": 9.972952618294442e-06, + "loss": 0.1738, + "step": 615 + }, + { + "epoch": 0.16356877323420074, + "grad_norm": 0.008095688510017955, + "learning_rate": 9.972724096998206e-06, + "loss": 0.0001, + "step": 616 + }, + { + "epoch": 0.16383430695698353, + "grad_norm": 0.9981390757792195, + "learning_rate": 9.97249461701941e-06, + "loss": 0.0576, + "step": 617 + }, + { + "epoch": 0.16409984067976632, + "grad_norm": 1.3058310882090556, + "learning_rate": 9.9722641784023e-06, + "loss": 0.0688, + "step": 618 + }, + { + "epoch": 0.1643653744025491, + "grad_norm": 1.0447117296584805, + "learning_rate": 9.972032781191297e-06, + "loss": 0.0576, + "step": 619 + }, + { + "epoch": 0.1646309081253319, + "grad_norm": 0.0992909067061662, + "learning_rate": 9.971800425431014e-06, + "loss": 0.0025, + "step": 620 + }, + { + "epoch": 0.16489644184811472, + "grad_norm": 2.5678651091451097, + "learning_rate": 9.971567111166246e-06, + "loss": 0.2461, + "step": 621 + }, + { + "epoch": 0.1651619755708975, + "grad_norm": 1.3926560176517229, + "learning_rate": 9.971332838441973e-06, + "loss": 0.0525, + "step": 622 + }, + { + "epoch": 0.1654275092936803, + "grad_norm": 0.1312253383616017, + "learning_rate": 9.97109760730336e-06, + "loss": 0.004, + "step": 623 + }, + { + "epoch": 0.1656930430164631, + "grad_norm": 2.1426039577089244, + "learning_rate": 9.970861417795757e-06, + "loss": 0.1836, + "step": 624 + }, + { + "epoch": 0.16595857673924588, + "grad_norm": 1.4296002668699905, + "learning_rate": 9.970624269964697e-06, + "loss": 0.0889, + "step": 625 + }, + { + "epoch": 0.16622411046202867, + "grad_norm": 0.40810968379724855, + "learning_rate": 9.970386163855903e-06, + "loss": 0.0229, + "step": 626 + }, + { + "epoch": 0.16648964418481146, + "grad_norm": 0.026519673065559527, + "learning_rate": 9.970147099515277e-06, + "loss": 0.0006, + "step": 627 + }, + { + "epoch": 0.16675517790759425, + "grad_norm": 0.9111033300992362, + "learning_rate": 9.969907076988907e-06, + "loss": 0.0576, + "step": 628 + }, + { + "epoch": 0.16702071163037704, + "grad_norm": 1.84251401314855, + "learning_rate": 9.969666096323068e-06, + "loss": 0.1641, + "step": 629 + }, + { + "epoch": 0.16728624535315986, + "grad_norm": 1.1977915299686255, + "learning_rate": 9.969424157564215e-06, + "loss": 0.0752, + "step": 630 + }, + { + "epoch": 0.16755177907594265, + "grad_norm": 1.1596796848128883, + "learning_rate": 9.969181260758995e-06, + "loss": 0.0476, + "step": 631 + }, + { + "epoch": 0.16781731279872544, + "grad_norm": 0.7311624184389818, + "learning_rate": 9.968937405954235e-06, + "loss": 0.0576, + "step": 632 + }, + { + "epoch": 0.16808284652150823, + "grad_norm": 0.43066270261344336, + "learning_rate": 9.968692593196944e-06, + "loss": 0.0317, + "step": 633 + }, + { + "epoch": 0.16834838024429102, + "grad_norm": 1.0570009180451556, + "learning_rate": 9.968446822534324e-06, + "loss": 0.1191, + "step": 634 + }, + { + "epoch": 0.16861391396707381, + "grad_norm": 1.7875303182362978, + "learning_rate": 9.968200094013751e-06, + "loss": 0.1279, + "step": 635 + }, + { + "epoch": 0.1688794476898566, + "grad_norm": 0.8109838110303486, + "learning_rate": 9.967952407682796e-06, + "loss": 0.0525, + "step": 636 + }, + { + "epoch": 0.1691449814126394, + "grad_norm": 0.995056083014455, + "learning_rate": 9.967703763589208e-06, + "loss": 0.0889, + "step": 637 + }, + { + "epoch": 0.16941051513542218, + "grad_norm": 0.7317688073360701, + "learning_rate": 9.967454161780923e-06, + "loss": 0.0576, + "step": 638 + }, + { + "epoch": 0.169676048858205, + "grad_norm": 0.27467746292088824, + "learning_rate": 9.967203602306062e-06, + "loss": 0.0206, + "step": 639 + }, + { + "epoch": 0.1699415825809878, + "grad_norm": 1.0717423043804712, + "learning_rate": 9.966952085212929e-06, + "loss": 0.1455, + "step": 640 + }, + { + "epoch": 0.17020711630377058, + "grad_norm": 0.21472091966860316, + "learning_rate": 9.966699610550013e-06, + "loss": 0.0146, + "step": 641 + }, + { + "epoch": 0.17047265002655337, + "grad_norm": 0.5320762490185771, + "learning_rate": 9.96644617836599e-06, + "loss": 0.0216, + "step": 642 + }, + { + "epoch": 0.17073818374933616, + "grad_norm": 0.9624120570928735, + "learning_rate": 9.966191788709716e-06, + "loss": 0.1035, + "step": 643 + }, + { + "epoch": 0.17100371747211895, + "grad_norm": 0.6950471061549814, + "learning_rate": 9.965936441630237e-06, + "loss": 0.0693, + "step": 644 + }, + { + "epoch": 0.17126925119490174, + "grad_norm": 1.060844145170164, + "learning_rate": 9.965680137176778e-06, + "loss": 0.1553, + "step": 645 + }, + { + "epoch": 0.17153478491768454, + "grad_norm": 0.6493031275434652, + "learning_rate": 9.965422875398756e-06, + "loss": 0.0576, + "step": 646 + }, + { + "epoch": 0.17180031864046733, + "grad_norm": 0.2802358255478571, + "learning_rate": 9.965164656345765e-06, + "loss": 0.0206, + "step": 647 + }, + { + "epoch": 0.17206585236325014, + "grad_norm": 1.1889779284390927, + "learning_rate": 9.964905480067585e-06, + "loss": 0.1113, + "step": 648 + }, + { + "epoch": 0.17233138608603293, + "grad_norm": 0.48129568513721377, + "learning_rate": 9.964645346614186e-06, + "loss": 0.0432, + "step": 649 + }, + { + "epoch": 0.17259691980881572, + "grad_norm": 0.5955170121717723, + "learning_rate": 9.964384256035715e-06, + "loss": 0.0579, + "step": 650 + }, + { + "epoch": 0.17286245353159851, + "grad_norm": 0.4053980801589404, + "learning_rate": 9.964122208382511e-06, + "loss": 0.0354, + "step": 651 + }, + { + "epoch": 0.1731279872543813, + "grad_norm": 0.43450493299788406, + "learning_rate": 9.96385920370509e-06, + "loss": 0.0391, + "step": 652 + }, + { + "epoch": 0.1733935209771641, + "grad_norm": 0.4490492804675279, + "learning_rate": 9.963595242054158e-06, + "loss": 0.0354, + "step": 653 + }, + { + "epoch": 0.17365905469994689, + "grad_norm": 0.5527458067392451, + "learning_rate": 9.963330323480604e-06, + "loss": 0.0525, + "step": 654 + }, + { + "epoch": 0.17392458842272968, + "grad_norm": 0.5425679761751954, + "learning_rate": 9.9630644480355e-06, + "loss": 0.0525, + "step": 655 + }, + { + "epoch": 0.1741901221455125, + "grad_norm": 0.35110925559228934, + "learning_rate": 9.962797615770105e-06, + "loss": 0.0286, + "step": 656 + }, + { + "epoch": 0.17445565586829528, + "grad_norm": 0.40723902213459534, + "learning_rate": 9.96252982673586e-06, + "loss": 0.0354, + "step": 657 + }, + { + "epoch": 0.17472118959107807, + "grad_norm": 0.29867369752067024, + "learning_rate": 9.962261080984394e-06, + "loss": 0.0229, + "step": 658 + }, + { + "epoch": 0.17498672331386086, + "grad_norm": 0.30130627923135, + "learning_rate": 9.961991378567514e-06, + "loss": 0.0229, + "step": 659 + }, + { + "epoch": 0.17525225703664365, + "grad_norm": 0.4704399733543959, + "learning_rate": 9.961720719537217e-06, + "loss": 0.0432, + "step": 660 + }, + { + "epoch": 0.17551779075942645, + "grad_norm": 0.8375442861352078, + "learning_rate": 9.961449103945684e-06, + "loss": 0.0889, + "step": 661 + }, + { + "epoch": 0.17578332448220924, + "grad_norm": 0.13114805545544933, + "learning_rate": 9.96117653184528e-06, + "loss": 0.0045, + "step": 662 + }, + { + "epoch": 0.17604885820499203, + "grad_norm": 0.31507559879440783, + "learning_rate": 9.960903003288551e-06, + "loss": 0.0104, + "step": 663 + }, + { + "epoch": 0.17631439192777482, + "grad_norm": 0.38090689585552884, + "learning_rate": 9.960628518328231e-06, + "loss": 0.0231, + "step": 664 + }, + { + "epoch": 0.17657992565055763, + "grad_norm": 0.36367373348514975, + "learning_rate": 9.960353077017238e-06, + "loss": 0.0286, + "step": 665 + }, + { + "epoch": 0.17684545937334042, + "grad_norm": 0.14099382022025622, + "learning_rate": 9.960076679408675e-06, + "loss": 0.0092, + "step": 666 + }, + { + "epoch": 0.17711099309612321, + "grad_norm": 0.3729088508977205, + "learning_rate": 9.959799325555825e-06, + "loss": 0.0317, + "step": 667 + }, + { + "epoch": 0.177376526818906, + "grad_norm": 0.20249754798930547, + "learning_rate": 9.95952101551216e-06, + "loss": 0.0035, + "step": 668 + }, + { + "epoch": 0.1776420605416888, + "grad_norm": 0.728096668821041, + "learning_rate": 9.959241749331339e-06, + "loss": 0.0391, + "step": 669 + }, + { + "epoch": 0.17790759426447159, + "grad_norm": 0.38161937312578453, + "learning_rate": 9.958961527067194e-06, + "loss": 0.0165, + "step": 670 + }, + { + "epoch": 0.17817312798725438, + "grad_norm": 1.1222249017413026, + "learning_rate": 9.958680348773753e-06, + "loss": 0.1455, + "step": 671 + }, + { + "epoch": 0.17843866171003717, + "grad_norm": 1.069255220495418, + "learning_rate": 9.958398214505225e-06, + "loss": 0.1279, + "step": 672 + }, + { + "epoch": 0.17870419543281996, + "grad_norm": 0.6232120562558745, + "learning_rate": 9.958115124315996e-06, + "loss": 0.0432, + "step": 673 + }, + { + "epoch": 0.17896972915560277, + "grad_norm": 0.5800413383472506, + "learning_rate": 9.95783107826065e-06, + "loss": 0.0125, + "step": 674 + }, + { + "epoch": 0.17923526287838557, + "grad_norm": 0.1579792821012369, + "learning_rate": 9.957546076393944e-06, + "loss": 0.0103, + "step": 675 + }, + { + "epoch": 0.17950079660116836, + "grad_norm": 0.8480282976053962, + "learning_rate": 9.957260118770822e-06, + "loss": 0.0815, + "step": 676 + }, + { + "epoch": 0.17976633032395115, + "grad_norm": 0.9382679997758676, + "learning_rate": 9.956973205446416e-06, + "loss": 0.0957, + "step": 677 + }, + { + "epoch": 0.18003186404673394, + "grad_norm": 0.6567431263540028, + "learning_rate": 9.956685336476037e-06, + "loss": 0.0525, + "step": 678 + }, + { + "epoch": 0.18029739776951673, + "grad_norm": 0.9345447093648367, + "learning_rate": 9.956396511915182e-06, + "loss": 0.0688, + "step": 679 + }, + { + "epoch": 0.18056293149229952, + "grad_norm": 1.028643546453173, + "learning_rate": 9.956106731819538e-06, + "loss": 0.0752, + "step": 680 + }, + { + "epoch": 0.1808284652150823, + "grad_norm": 1.1200381609657348, + "learning_rate": 9.955815996244967e-06, + "loss": 0.0889, + "step": 681 + }, + { + "epoch": 0.1810939989378651, + "grad_norm": 0.49251673343344965, + "learning_rate": 9.955524305247518e-06, + "loss": 0.0391, + "step": 682 + }, + { + "epoch": 0.18135953266064792, + "grad_norm": 0.5170146835568467, + "learning_rate": 9.955231658883432e-06, + "loss": 0.0391, + "step": 683 + }, + { + "epoch": 0.1816250663834307, + "grad_norm": 0.7825780874484762, + "learning_rate": 9.954938057209121e-06, + "loss": 0.0525, + "step": 684 + }, + { + "epoch": 0.1818906001062135, + "grad_norm": 1.3290725760750859, + "learning_rate": 9.954643500281191e-06, + "loss": 0.1934, + "step": 685 + }, + { + "epoch": 0.1821561338289963, + "grad_norm": 1.2730693589730842, + "learning_rate": 9.95434798815643e-06, + "loss": 0.03, + "step": 686 + }, + { + "epoch": 0.18242166755177908, + "grad_norm": 0.29840206739057834, + "learning_rate": 9.954051520891808e-06, + "loss": 0.0229, + "step": 687 + }, + { + "epoch": 0.18268720127456187, + "grad_norm": 1.283727575547433, + "learning_rate": 9.95375409854448e-06, + "loss": 0.1836, + "step": 688 + }, + { + "epoch": 0.18295273499734466, + "grad_norm": 0.36148423987976935, + "learning_rate": 9.953455721171784e-06, + "loss": 0.0286, + "step": 689 + }, + { + "epoch": 0.18321826872012745, + "grad_norm": 0.26464506058473897, + "learning_rate": 9.953156388831246e-06, + "loss": 0.0184, + "step": 690 + }, + { + "epoch": 0.18348380244291024, + "grad_norm": 0.26815836434680423, + "learning_rate": 9.952856101580574e-06, + "loss": 0.0206, + "step": 691 + }, + { + "epoch": 0.18374933616569306, + "grad_norm": 0.9433969771302444, + "learning_rate": 9.952554859477661e-06, + "loss": 0.0815, + "step": 692 + }, + { + "epoch": 0.18401486988847585, + "grad_norm": 0.8371591636277368, + "learning_rate": 9.95225266258058e-06, + "loss": 0.0688, + "step": 693 + }, + { + "epoch": 0.18428040361125864, + "grad_norm": 0.31710782740299065, + "learning_rate": 9.95194951094759e-06, + "loss": 0.0255, + "step": 694 + }, + { + "epoch": 0.18454593733404143, + "grad_norm": 0.7792504139919215, + "learning_rate": 9.951645404637139e-06, + "loss": 0.063, + "step": 695 + }, + { + "epoch": 0.18481147105682422, + "grad_norm": 0.24864366834341872, + "learning_rate": 9.951340343707852e-06, + "loss": 0.0184, + "step": 696 + }, + { + "epoch": 0.185077004779607, + "grad_norm": 1.3198814057549204, + "learning_rate": 9.951034328218543e-06, + "loss": 0.0815, + "step": 697 + }, + { + "epoch": 0.1853425385023898, + "grad_norm": 0.1349697962445958, + "learning_rate": 9.950727358228207e-06, + "loss": 0.004, + "step": 698 + }, + { + "epoch": 0.1856080722251726, + "grad_norm": 0.7940800117641265, + "learning_rate": 9.950419433796026e-06, + "loss": 0.0525, + "step": 699 + }, + { + "epoch": 0.18587360594795538, + "grad_norm": 0.17975503032083975, + "learning_rate": 9.950110554981363e-06, + "loss": 0.0045, + "step": 700 + }, + { + "epoch": 0.1861391396707382, + "grad_norm": 0.48368320799159703, + "learning_rate": 9.949800721843765e-06, + "loss": 0.0391, + "step": 701 + }, + { + "epoch": 0.186404673393521, + "grad_norm": 0.31589438590765767, + "learning_rate": 9.949489934442966e-06, + "loss": 0.0255, + "step": 702 + }, + { + "epoch": 0.18667020711630378, + "grad_norm": 0.31533609255408135, + "learning_rate": 9.949178192838881e-06, + "loss": 0.0258, + "step": 703 + }, + { + "epoch": 0.18693574083908657, + "grad_norm": 1.3466582105948832, + "learning_rate": 9.94886549709161e-06, + "loss": 0.1279, + "step": 704 + }, + { + "epoch": 0.18720127456186936, + "grad_norm": 0.910289445658781, + "learning_rate": 9.948551847261439e-06, + "loss": 0.0815, + "step": 705 + }, + { + "epoch": 0.18746680828465215, + "grad_norm": 0.21688834388185138, + "learning_rate": 9.948237243408834e-06, + "loss": 0.0165, + "step": 706 + }, + { + "epoch": 0.18773234200743494, + "grad_norm": 0.021656248810182913, + "learning_rate": 9.947921685594448e-06, + "loss": 0.0005, + "step": 707 + }, + { + "epoch": 0.18799787573021773, + "grad_norm": 0.9878506170969464, + "learning_rate": 9.947605173879116e-06, + "loss": 0.1279, + "step": 708 + }, + { + "epoch": 0.18826340945300052, + "grad_norm": 1.1026344698990722, + "learning_rate": 9.947287708323858e-06, + "loss": 0.1553, + "step": 709 + }, + { + "epoch": 0.18852894317578334, + "grad_norm": 0.6327076367042769, + "learning_rate": 9.946969288989878e-06, + "loss": 0.0579, + "step": 710 + }, + { + "epoch": 0.18879447689856613, + "grad_norm": 0.972281811713341, + "learning_rate": 9.946649915938562e-06, + "loss": 0.1367, + "step": 711 + }, + { + "epoch": 0.18906001062134892, + "grad_norm": 0.9045454418723899, + "learning_rate": 9.946329589231484e-06, + "loss": 0.1113, + "step": 712 + }, + { + "epoch": 0.1893255443441317, + "grad_norm": 1.1652383753491669, + "learning_rate": 9.946008308930396e-06, + "loss": 0.1035, + "step": 713 + }, + { + "epoch": 0.1895910780669145, + "grad_norm": 0.5983398386283125, + "learning_rate": 9.94568607509724e-06, + "loss": 0.0525, + "step": 714 + }, + { + "epoch": 0.1898566117896973, + "grad_norm": 0.9406342065879315, + "learning_rate": 9.945362887794139e-06, + "loss": 0.1279, + "step": 715 + }, + { + "epoch": 0.19012214551248008, + "grad_norm": 1.0678169277527272, + "learning_rate": 9.945038747083396e-06, + "loss": 0.1455, + "step": 716 + }, + { + "epoch": 0.19038767923526287, + "grad_norm": 0.6152395770617992, + "learning_rate": 9.944713653027505e-06, + "loss": 0.063, + "step": 717 + }, + { + "epoch": 0.19065321295804566, + "grad_norm": 0.7306184341026286, + "learning_rate": 9.944387605689139e-06, + "loss": 0.0693, + "step": 718 + }, + { + "epoch": 0.19091874668082848, + "grad_norm": 0.32716061364977655, + "learning_rate": 9.944060605131157e-06, + "loss": 0.0165, + "step": 719 + }, + { + "epoch": 0.19118428040361127, + "grad_norm": 0.9594346998876806, + "learning_rate": 9.943732651416597e-06, + "loss": 0.0957, + "step": 720 + }, + { + "epoch": 0.19144981412639406, + "grad_norm": 0.8678737283379834, + "learning_rate": 9.94340374460869e-06, + "loss": 0.1035, + "step": 721 + }, + { + "epoch": 0.19171534784917685, + "grad_norm": 0.5586020989155928, + "learning_rate": 9.943073884770844e-06, + "loss": 0.0579, + "step": 722 + }, + { + "epoch": 0.19198088157195964, + "grad_norm": 0.6710124613449318, + "learning_rate": 9.94274307196665e-06, + "loss": 0.0752, + "step": 723 + }, + { + "epoch": 0.19224641529474243, + "grad_norm": 0.11750671730822675, + "learning_rate": 9.942411306259887e-06, + "loss": 0.0041, + "step": 724 + }, + { + "epoch": 0.19251194901752522, + "grad_norm": 0.9492120597858461, + "learning_rate": 9.942078587714514e-06, + "loss": 0.1113, + "step": 725 + }, + { + "epoch": 0.192777482740308, + "grad_norm": 0.7578093561638131, + "learning_rate": 9.941744916394675e-06, + "loss": 0.0889, + "step": 726 + }, + { + "epoch": 0.1930430164630908, + "grad_norm": 0.0703405751086318, + "learning_rate": 9.9414102923647e-06, + "loss": 0.0017, + "step": 727 + }, + { + "epoch": 0.19330855018587362, + "grad_norm": 0.2720572743378476, + "learning_rate": 9.941074715689097e-06, + "loss": 0.0132, + "step": 728 + }, + { + "epoch": 0.1935740839086564, + "grad_norm": 0.47926399744773224, + "learning_rate": 9.940738186432565e-06, + "loss": 0.0479, + "step": 729 + }, + { + "epoch": 0.1938396176314392, + "grad_norm": 0.4960442155304019, + "learning_rate": 9.94040070465998e-06, + "loss": 0.053, + "step": 730 + }, + { + "epoch": 0.194105151354222, + "grad_norm": 0.6099810000661499, + "learning_rate": 9.940062270436406e-06, + "loss": 0.0752, + "step": 731 + }, + { + "epoch": 0.19437068507700478, + "grad_norm": 0.5317125466849125, + "learning_rate": 9.93972288382709e-06, + "loss": 0.0579, + "step": 732 + }, + { + "epoch": 0.19463621879978757, + "grad_norm": 0.405350264279859, + "learning_rate": 9.93938254489746e-06, + "loss": 0.0359, + "step": 733 + }, + { + "epoch": 0.19490175252257036, + "grad_norm": 0.5550223239319415, + "learning_rate": 9.93904125371313e-06, + "loss": 0.063, + "step": 734 + }, + { + "epoch": 0.19516728624535315, + "grad_norm": 0.4548926771748096, + "learning_rate": 9.938699010339898e-06, + "loss": 0.0479, + "step": 735 + }, + { + "epoch": 0.19543281996813594, + "grad_norm": 0.4301648170034228, + "learning_rate": 9.93835581484374e-06, + "loss": 0.0432, + "step": 736 + }, + { + "epoch": 0.19569835369091876, + "grad_norm": 0.5759338926245092, + "learning_rate": 9.938011667290828e-06, + "loss": 0.0476, + "step": 737 + }, + { + "epoch": 0.19596388741370155, + "grad_norm": 0.47327836419648434, + "learning_rate": 9.9376665677475e-06, + "loss": 0.0525, + "step": 738 + }, + { + "epoch": 0.19622942113648434, + "grad_norm": 0.1587521869962895, + "learning_rate": 9.937320516280297e-06, + "loss": 0.0051, + "step": 739 + }, + { + "epoch": 0.19649495485926713, + "grad_norm": 0.29964993842166077, + "learning_rate": 9.936973512955927e-06, + "loss": 0.0132, + "step": 740 + }, + { + "epoch": 0.19676048858204992, + "grad_norm": 0.5192574350505638, + "learning_rate": 9.936625557841289e-06, + "loss": 0.0525, + "step": 741 + }, + { + "epoch": 0.1970260223048327, + "grad_norm": 0.6174706812099604, + "learning_rate": 9.936276651003465e-06, + "loss": 0.0693, + "step": 742 + }, + { + "epoch": 0.1972915560276155, + "grad_norm": 0.4060064868344745, + "learning_rate": 9.935926792509723e-06, + "loss": 0.0435, + "step": 743 + }, + { + "epoch": 0.1975570897503983, + "grad_norm": 0.6447420229580227, + "learning_rate": 9.935575982427507e-06, + "loss": 0.082, + "step": 744 + }, + { + "epoch": 0.19782262347318108, + "grad_norm": 0.8009583314059334, + "learning_rate": 9.935224220824452e-06, + "loss": 0.1191, + "step": 745 + }, + { + "epoch": 0.1980881571959639, + "grad_norm": 0.7682777474476555, + "learning_rate": 9.934871507768374e-06, + "loss": 0.0688, + "step": 746 + }, + { + "epoch": 0.1983536909187467, + "grad_norm": 0.27537017109972733, + "learning_rate": 9.934517843327269e-06, + "loss": 0.0104, + "step": 747 + }, + { + "epoch": 0.19861922464152948, + "grad_norm": 0.4708650125151908, + "learning_rate": 9.93416322756932e-06, + "loss": 0.0479, + "step": 748 + }, + { + "epoch": 0.19888475836431227, + "grad_norm": 0.706657435045738, + "learning_rate": 9.933807660562898e-06, + "loss": 0.082, + "step": 749 + }, + { + "epoch": 0.19915029208709506, + "grad_norm": 0.36488237415402874, + "learning_rate": 9.933451142376545e-06, + "loss": 0.0354, + "step": 750 + }, + { + "epoch": 0.19941582580987785, + "grad_norm": 0.4861196249654044, + "learning_rate": 9.933093673078994e-06, + "loss": 0.0525, + "step": 751 + }, + { + "epoch": 0.19968135953266064, + "grad_norm": 0.5291778145248267, + "learning_rate": 9.932735252739168e-06, + "loss": 0.0393, + "step": 752 + }, + { + "epoch": 0.19994689325544343, + "grad_norm": 0.35001230301241637, + "learning_rate": 9.93237588142616e-06, + "loss": 0.0354, + "step": 753 + }, + { + "epoch": 0.20021242697822622, + "grad_norm": 2.417388946512036, + "learning_rate": 9.932015559209255e-06, + "loss": 0.0277, + "step": 754 + }, + { + "epoch": 0.20047796070100904, + "grad_norm": 0.1251433570133762, + "learning_rate": 9.931654286157915e-06, + "loss": 0.0032, + "step": 755 + }, + { + "epoch": 0.20074349442379183, + "grad_norm": 0.209414281849796, + "learning_rate": 9.931292062341793e-06, + "loss": 0.0047, + "step": 756 + }, + { + "epoch": 0.20100902814657462, + "grad_norm": 0.008292667150507884, + "learning_rate": 9.930928887830721e-06, + "loss": 0.0002, + "step": 757 + }, + { + "epoch": 0.2012745618693574, + "grad_norm": 0.38647994443034445, + "learning_rate": 9.930564762694715e-06, + "loss": 0.0393, + "step": 758 + }, + { + "epoch": 0.2015400955921402, + "grad_norm": 0.6378261596904626, + "learning_rate": 9.930199687003972e-06, + "loss": 0.0579, + "step": 759 + }, + { + "epoch": 0.201805629314923, + "grad_norm": 2.3592667920352035, + "learning_rate": 9.929833660828878e-06, + "loss": 0.1455, + "step": 760 + }, + { + "epoch": 0.20207116303770578, + "grad_norm": 0.6334539242279383, + "learning_rate": 9.929466684239993e-06, + "loss": 0.0525, + "step": 761 + }, + { + "epoch": 0.20233669676048857, + "grad_norm": 0.30334797132909946, + "learning_rate": 9.929098757308071e-06, + "loss": 0.0286, + "step": 762 + }, + { + "epoch": 0.20260223048327136, + "grad_norm": 0.30034960277983547, + "learning_rate": 9.928729880104042e-06, + "loss": 0.0258, + "step": 763 + }, + { + "epoch": 0.20286776420605418, + "grad_norm": 0.7158437599898999, + "learning_rate": 9.928360052699019e-06, + "loss": 0.0579, + "step": 764 + }, + { + "epoch": 0.20313329792883697, + "grad_norm": 0.4427519026261093, + "learning_rate": 9.927989275164305e-06, + "loss": 0.0432, + "step": 765 + }, + { + "epoch": 0.20339883165161976, + "grad_norm": 1.223371374907059, + "learning_rate": 9.927617547571377e-06, + "loss": 0.0957, + "step": 766 + }, + { + "epoch": 0.20366436537440255, + "grad_norm": 0.8525259837165133, + "learning_rate": 9.927244869991905e-06, + "loss": 0.1191, + "step": 767 + }, + { + "epoch": 0.20392989909718534, + "grad_norm": 0.7529517092699473, + "learning_rate": 9.92687124249773e-06, + "loss": 0.0889, + "step": 768 + }, + { + "epoch": 0.20419543281996813, + "grad_norm": 0.4231343502487387, + "learning_rate": 9.926496665160887e-06, + "loss": 0.0393, + "step": 769 + }, + { + "epoch": 0.20446096654275092, + "grad_norm": 0.02781249909477673, + "learning_rate": 9.926121138053591e-06, + "loss": 0.0005, + "step": 770 + }, + { + "epoch": 0.2047265002655337, + "grad_norm": 0.32019858843907156, + "learning_rate": 9.925744661248238e-06, + "loss": 0.0286, + "step": 771 + }, + { + "epoch": 0.20499203398831653, + "grad_norm": 0.34210995299202096, + "learning_rate": 9.925367234817407e-06, + "loss": 0.0317, + "step": 772 + }, + { + "epoch": 0.20525756771109932, + "grad_norm": 0.29667952505177037, + "learning_rate": 9.924988858833862e-06, + "loss": 0.0258, + "step": 773 + }, + { + "epoch": 0.2055231014338821, + "grad_norm": 0.775170587796152, + "learning_rate": 9.924609533370551e-06, + "loss": 0.082, + "step": 774 + }, + { + "epoch": 0.2057886351566649, + "grad_norm": 0.8776207151240932, + "learning_rate": 9.924229258500601e-06, + "loss": 0.0688, + "step": 775 + }, + { + "epoch": 0.2060541688794477, + "grad_norm": 0.34921604699592496, + "learning_rate": 9.923848034297327e-06, + "loss": 0.0286, + "step": 776 + }, + { + "epoch": 0.20631970260223048, + "grad_norm": 0.7094770354343616, + "learning_rate": 9.923465860834224e-06, + "loss": 0.063, + "step": 777 + }, + { + "epoch": 0.20658523632501327, + "grad_norm": 0.8576553061781719, + "learning_rate": 9.923082738184969e-06, + "loss": 0.0688, + "step": 778 + }, + { + "epoch": 0.20685077004779606, + "grad_norm": 0.4617621570202573, + "learning_rate": 9.922698666423424e-06, + "loss": 0.0126, + "step": 779 + }, + { + "epoch": 0.20711630377057885, + "grad_norm": 0.2701329483492822, + "learning_rate": 9.922313645623634e-06, + "loss": 0.0205, + "step": 780 + }, + { + "epoch": 0.20738183749336167, + "grad_norm": 1.2246770118493904, + "learning_rate": 9.921927675859828e-06, + "loss": 0.0576, + "step": 781 + }, + { + "epoch": 0.20764737121614446, + "grad_norm": 0.29636608608865783, + "learning_rate": 9.921540757206414e-06, + "loss": 0.0258, + "step": 782 + }, + { + "epoch": 0.20791290493892725, + "grad_norm": 1.8125953283598475, + "learning_rate": 9.921152889737985e-06, + "loss": 0.1367, + "step": 783 + }, + { + "epoch": 0.20817843866171004, + "grad_norm": 0.03047349596838622, + "learning_rate": 9.920764073529319e-06, + "loss": 0.0006, + "step": 784 + }, + { + "epoch": 0.20844397238449283, + "grad_norm": 0.16957814720325154, + "learning_rate": 9.920374308655375e-06, + "loss": 0.0041, + "step": 785 + }, + { + "epoch": 0.20870950610727562, + "grad_norm": 0.7111174333404663, + "learning_rate": 9.919983595191295e-06, + "loss": 0.0165, + "step": 786 + }, + { + "epoch": 0.2089750398300584, + "grad_norm": 1.244139227465599, + "learning_rate": 9.919591933212403e-06, + "loss": 0.0815, + "step": 787 + }, + { + "epoch": 0.2092405735528412, + "grad_norm": 0.34737267129055766, + "learning_rate": 9.919199322794209e-06, + "loss": 0.0286, + "step": 788 + }, + { + "epoch": 0.209506107275624, + "grad_norm": 0.9063191327124598, + "learning_rate": 9.9188057640124e-06, + "loss": 0.0479, + "step": 789 + }, + { + "epoch": 0.2097716409984068, + "grad_norm": 1.1996030088378673, + "learning_rate": 9.918411256942852e-06, + "loss": 0.1279, + "step": 790 + }, + { + "epoch": 0.2100371747211896, + "grad_norm": 0.3272617386081918, + "learning_rate": 9.918015801661622e-06, + "loss": 0.0231, + "step": 791 + }, + { + "epoch": 0.2103027084439724, + "grad_norm": 0.20150941486275553, + "learning_rate": 9.91761939824495e-06, + "loss": 0.0146, + "step": 792 + }, + { + "epoch": 0.21056824216675518, + "grad_norm": 0.9406775259417158, + "learning_rate": 9.917222046769252e-06, + "loss": 0.1035, + "step": 793 + }, + { + "epoch": 0.21083377588953797, + "grad_norm": 0.33726088549905536, + "learning_rate": 9.916823747311138e-06, + "loss": 0.0255, + "step": 794 + }, + { + "epoch": 0.21109930961232076, + "grad_norm": 0.24491234727145536, + "learning_rate": 9.916424499947395e-06, + "loss": 0.0184, + "step": 795 + }, + { + "epoch": 0.21136484333510355, + "grad_norm": 0.1878113074555544, + "learning_rate": 9.91602430475499e-06, + "loss": 0.0115, + "step": 796 + }, + { + "epoch": 0.21163037705788634, + "grad_norm": 0.030677918403306813, + "learning_rate": 9.91562316181108e-06, + "loss": 0.0008, + "step": 797 + }, + { + "epoch": 0.21189591078066913, + "grad_norm": 0.8660378755286174, + "learning_rate": 9.915221071193e-06, + "loss": 0.0476, + "step": 798 + }, + { + "epoch": 0.21216144450345195, + "grad_norm": 0.8549555937928719, + "learning_rate": 9.914818032978263e-06, + "loss": 0.0525, + "step": 799 + }, + { + "epoch": 0.21242697822623474, + "grad_norm": 0.6372317187162225, + "learning_rate": 9.914414047244578e-06, + "loss": 0.0317, + "step": 800 + }, + { + "epoch": 0.21269251194901753, + "grad_norm": 0.9907320369841311, + "learning_rate": 9.914009114069824e-06, + "loss": 0.0525, + "step": 801 + }, + { + "epoch": 0.21295804567180032, + "grad_norm": 1.3053615357778507, + "learning_rate": 9.913603233532067e-06, + "loss": 0.0391, + "step": 802 + }, + { + "epoch": 0.2132235793945831, + "grad_norm": 0.5689635893757348, + "learning_rate": 9.913196405709557e-06, + "loss": 0.0286, + "step": 803 + }, + { + "epoch": 0.2134891131173659, + "grad_norm": 1.538445081246666, + "learning_rate": 9.912788630680727e-06, + "loss": 0.0525, + "step": 804 + }, + { + "epoch": 0.2137546468401487, + "grad_norm": 0.4053859965192937, + "learning_rate": 9.91237990852419e-06, + "loss": 0.0165, + "step": 805 + }, + { + "epoch": 0.21402018056293148, + "grad_norm": 0.12115693287671359, + "learning_rate": 9.911970239318743e-06, + "loss": 0.0072, + "step": 806 + }, + { + "epoch": 0.21428571428571427, + "grad_norm": 0.02612612601513698, + "learning_rate": 9.911559623143364e-06, + "loss": 0.0004, + "step": 807 + }, + { + "epoch": 0.2145512480084971, + "grad_norm": 0.155521917709447, + "learning_rate": 9.911148060077216e-06, + "loss": 0.0092, + "step": 808 + }, + { + "epoch": 0.21481678173127988, + "grad_norm": 0.800502691704114, + "learning_rate": 9.910735550199647e-06, + "loss": 0.0317, + "step": 809 + }, + { + "epoch": 0.21508231545406267, + "grad_norm": 1.7498059645077653, + "learning_rate": 9.910322093590177e-06, + "loss": 0.1035, + "step": 810 + }, + { + "epoch": 0.21534784917684546, + "grad_norm": 0.5768858949644357, + "learning_rate": 9.90990769032852e-06, + "loss": 0.0255, + "step": 811 + }, + { + "epoch": 0.21561338289962825, + "grad_norm": 0.17436803776687265, + "learning_rate": 9.909492340494567e-06, + "loss": 0.0064, + "step": 812 + }, + { + "epoch": 0.21587891662241104, + "grad_norm": 1.2847308699648239, + "learning_rate": 9.909076044168393e-06, + "loss": 0.0476, + "step": 813 + }, + { + "epoch": 0.21614445034519383, + "grad_norm": 0.20641696265719725, + "learning_rate": 9.908658801430256e-06, + "loss": 0.0115, + "step": 814 + }, + { + "epoch": 0.21640998406797662, + "grad_norm": 0.4594930508016625, + "learning_rate": 9.908240612360593e-06, + "loss": 0.0205, + "step": 815 + }, + { + "epoch": 0.21667551779075941, + "grad_norm": 1.6460520494176336, + "learning_rate": 9.907821477040028e-06, + "loss": 0.0391, + "step": 816 + }, + { + "epoch": 0.21694105151354223, + "grad_norm": 0.006952372103528881, + "learning_rate": 9.907401395549364e-06, + "loss": 0.0, + "step": 817 + }, + { + "epoch": 0.21720658523632502, + "grad_norm": 1.625155011661868, + "learning_rate": 9.90698036796959e-06, + "loss": 0.2344, + "step": 818 + }, + { + "epoch": 0.2174721189591078, + "grad_norm": 0.0352752424800895, + "learning_rate": 9.906558394381872e-06, + "loss": 0.0017, + "step": 819 + }, + { + "epoch": 0.2177376526818906, + "grad_norm": 0.20879590616740903, + "learning_rate": 9.906135474867562e-06, + "loss": 0.0103, + "step": 820 + }, + { + "epoch": 0.2180031864046734, + "grad_norm": 1.6599267611839752, + "learning_rate": 9.905711609508198e-06, + "loss": 0.2676, + "step": 821 + }, + { + "epoch": 0.21826872012745618, + "grad_norm": 3.9796641629384975, + "learning_rate": 9.90528679838549e-06, + "loss": 0.1553, + "step": 822 + }, + { + "epoch": 0.21853425385023897, + "grad_norm": 0.0648341014830216, + "learning_rate": 9.904861041581341e-06, + "loss": 0.0035, + "step": 823 + }, + { + "epoch": 0.21879978757302176, + "grad_norm": 11.724369911160045, + "learning_rate": 9.904434339177831e-06, + "loss": 0.4199, + "step": 824 + }, + { + "epoch": 0.21906532129580455, + "grad_norm": 1.7629568442469772, + "learning_rate": 9.904006691257224e-06, + "loss": 0.2559, + "step": 825 + }, + { + "epoch": 0.21933085501858737, + "grad_norm": 0.10426293175973483, + "learning_rate": 9.903578097901963e-06, + "loss": 0.0064, + "step": 826 + }, + { + "epoch": 0.21959638874137016, + "grad_norm": 1.4924171036944354, + "learning_rate": 9.903148559194677e-06, + "loss": 0.1641, + "step": 827 + }, + { + "epoch": 0.21986192246415295, + "grad_norm": 1.3050509840328823, + "learning_rate": 9.902718075218176e-06, + "loss": 0.1934, + "step": 828 + }, + { + "epoch": 0.22012745618693574, + "grad_norm": 1.235993047459142, + "learning_rate": 9.902286646055453e-06, + "loss": 0.0576, + "step": 829 + }, + { + "epoch": 0.22039298990971853, + "grad_norm": 0.24248121114396165, + "learning_rate": 9.90185427178968e-06, + "loss": 0.0165, + "step": 830 + }, + { + "epoch": 0.22065852363250132, + "grad_norm": 4.175312997922109, + "learning_rate": 9.901420952504217e-06, + "loss": 0.0576, + "step": 831 + }, + { + "epoch": 0.22092405735528411, + "grad_norm": 3.2184596570988093, + "learning_rate": 9.9009866882826e-06, + "loss": 0.1279, + "step": 832 + }, + { + "epoch": 0.2211895910780669, + "grad_norm": 1.0077714273038476, + "learning_rate": 9.900551479208553e-06, + "loss": 0.0576, + "step": 833 + }, + { + "epoch": 0.2214551248008497, + "grad_norm": 0.5299304251634451, + "learning_rate": 9.900115325365975e-06, + "loss": 0.0432, + "step": 834 + }, + { + "epoch": 0.2217206585236325, + "grad_norm": 0.25144714721412126, + "learning_rate": 9.899678226838956e-06, + "loss": 0.0206, + "step": 835 + }, + { + "epoch": 0.2219861922464153, + "grad_norm": 1.4383846378372553, + "learning_rate": 9.89924018371176e-06, + "loss": 0.1191, + "step": 836 + }, + { + "epoch": 0.2222517259691981, + "grad_norm": 0.6128666274034679, + "learning_rate": 9.898801196068839e-06, + "loss": 0.0391, + "step": 837 + }, + { + "epoch": 0.22251725969198088, + "grad_norm": 1.6083587605392458, + "learning_rate": 9.898361263994823e-06, + "loss": 0.1035, + "step": 838 + }, + { + "epoch": 0.22278279341476367, + "grad_norm": 0.33105381078233614, + "learning_rate": 9.897920387574525e-06, + "loss": 0.0286, + "step": 839 + }, + { + "epoch": 0.22304832713754646, + "grad_norm": 0.581750205311232, + "learning_rate": 9.897478566892942e-06, + "loss": 0.0317, + "step": 840 + }, + { + "epoch": 0.22331386086032926, + "grad_norm": 1.103725743553766, + "learning_rate": 9.897035802035253e-06, + "loss": 0.0354, + "step": 841 + }, + { + "epoch": 0.22357939458311205, + "grad_norm": 0.41623629278157465, + "learning_rate": 9.896592093086816e-06, + "loss": 0.0354, + "step": 842 + }, + { + "epoch": 0.22384492830589484, + "grad_norm": 0.22870378430920638, + "learning_rate": 9.896147440133173e-06, + "loss": 0.002, + "step": 843 + }, + { + "epoch": 0.22411046202867765, + "grad_norm": 0.026554099075059814, + "learning_rate": 9.89570184326005e-06, + "loss": 0.0005, + "step": 844 + }, + { + "epoch": 0.22437599575146044, + "grad_norm": 1.0737697123349998, + "learning_rate": 9.895255302553349e-06, + "loss": 0.1641, + "step": 845 + }, + { + "epoch": 0.22464152947424323, + "grad_norm": 1.05755659361386, + "learning_rate": 9.89480781809916e-06, + "loss": 0.0688, + "step": 846 + }, + { + "epoch": 0.22490706319702602, + "grad_norm": 0.8318386749828283, + "learning_rate": 9.894359389983753e-06, + "loss": 0.1191, + "step": 847 + }, + { + "epoch": 0.22517259691980882, + "grad_norm": 0.38706813856524885, + "learning_rate": 9.89391001829358e-06, + "loss": 0.0354, + "step": 848 + }, + { + "epoch": 0.2254381306425916, + "grad_norm": 0.06799335395718269, + "learning_rate": 9.893459703115272e-06, + "loss": 0.0017, + "step": 849 + }, + { + "epoch": 0.2257036643653744, + "grad_norm": 0.36794444497590895, + "learning_rate": 9.893008444535647e-06, + "loss": 0.0093, + "step": 850 + }, + { + "epoch": 0.22596919808815719, + "grad_norm": 0.5439922261966197, + "learning_rate": 9.892556242641702e-06, + "loss": 0.0117, + "step": 851 + }, + { + "epoch": 0.22623473181093998, + "grad_norm": 0.8380785010711241, + "learning_rate": 9.892103097520615e-06, + "loss": 0.063, + "step": 852 + }, + { + "epoch": 0.2265002655337228, + "grad_norm": 0.5431138700099414, + "learning_rate": 9.89164900925975e-06, + "loss": 0.0288, + "step": 853 + }, + { + "epoch": 0.22676579925650558, + "grad_norm": 0.8146895584075347, + "learning_rate": 9.891193977946645e-06, + "loss": 0.1113, + "step": 854 + }, + { + "epoch": 0.22703133297928837, + "grad_norm": 0.6831359882459529, + "learning_rate": 9.890738003669029e-06, + "loss": 0.0688, + "step": 855 + }, + { + "epoch": 0.22729686670207117, + "grad_norm": 0.4121057526788246, + "learning_rate": 9.890281086514806e-06, + "loss": 0.0391, + "step": 856 + }, + { + "epoch": 0.22756240042485396, + "grad_norm": 0.7258123968019558, + "learning_rate": 9.889823226572066e-06, + "loss": 0.0752, + "step": 857 + }, + { + "epoch": 0.22782793414763675, + "grad_norm": 0.8294544573466394, + "learning_rate": 9.889364423929076e-06, + "loss": 0.1035, + "step": 858 + }, + { + "epoch": 0.22809346787041954, + "grad_norm": 1.051757101289739, + "learning_rate": 9.888904678674292e-06, + "loss": 0.1113, + "step": 859 + }, + { + "epoch": 0.22835900159320233, + "grad_norm": 1.3061340829563401, + "learning_rate": 9.888443990896344e-06, + "loss": 0.1113, + "step": 860 + }, + { + "epoch": 0.22862453531598512, + "grad_norm": 0.8993459080630722, + "learning_rate": 9.887982360684049e-06, + "loss": 0.0889, + "step": 861 + }, + { + "epoch": 0.22889006903876793, + "grad_norm": 0.7347323587084027, + "learning_rate": 9.887519788126404e-06, + "loss": 0.0752, + "step": 862 + }, + { + "epoch": 0.22915560276155073, + "grad_norm": 0.5477598732831455, + "learning_rate": 9.887056273312584e-06, + "loss": 0.0354, + "step": 863 + }, + { + "epoch": 0.22942113648433352, + "grad_norm": 0.35585991259844624, + "learning_rate": 9.886591816331953e-06, + "loss": 0.0233, + "step": 864 + }, + { + "epoch": 0.2296866702071163, + "grad_norm": 1.3171584886462089, + "learning_rate": 9.886126417274054e-06, + "loss": 0.1191, + "step": 865 + }, + { + "epoch": 0.2299522039298991, + "grad_norm": 0.5959273314376112, + "learning_rate": 9.885660076228606e-06, + "loss": 0.0354, + "step": 866 + }, + { + "epoch": 0.2302177376526819, + "grad_norm": 0.5887513008564544, + "learning_rate": 9.885192793285516e-06, + "loss": 0.063, + "step": 867 + }, + { + "epoch": 0.23048327137546468, + "grad_norm": 1.2261572998536514, + "learning_rate": 9.884724568534873e-06, + "loss": 0.1191, + "step": 868 + }, + { + "epoch": 0.23074880509824747, + "grad_norm": 3.6206803166086385, + "learning_rate": 9.884255402066943e-06, + "loss": 0.2891, + "step": 869 + }, + { + "epoch": 0.23101433882103026, + "grad_norm": 0.8824142077915467, + "learning_rate": 9.883785293972175e-06, + "loss": 0.0957, + "step": 870 + }, + { + "epoch": 0.23127987254381308, + "grad_norm": 0.46423245513429523, + "learning_rate": 9.8833142443412e-06, + "loss": 0.0393, + "step": 871 + }, + { + "epoch": 0.23154540626659587, + "grad_norm": 0.5138918912335925, + "learning_rate": 9.882842253264835e-06, + "loss": 0.0525, + "step": 872 + }, + { + "epoch": 0.23181093998937866, + "grad_norm": 0.5843357342302304, + "learning_rate": 9.882369320834068e-06, + "loss": 0.0693, + "step": 873 + }, + { + "epoch": 0.23207647371216145, + "grad_norm": 0.2520466827681795, + "learning_rate": 9.88189544714008e-06, + "loss": 0.0074, + "step": 874 + }, + { + "epoch": 0.23234200743494424, + "grad_norm": 0.5794422421534507, + "learning_rate": 9.881420632274226e-06, + "loss": 0.0693, + "step": 875 + }, + { + "epoch": 0.23260754115772703, + "grad_norm": 0.44568917664701574, + "learning_rate": 9.880944876328045e-06, + "loss": 0.0354, + "step": 876 + }, + { + "epoch": 0.23287307488050982, + "grad_norm": 0.5250220184195629, + "learning_rate": 9.88046817939326e-06, + "loss": 0.0579, + "step": 877 + }, + { + "epoch": 0.2331386086032926, + "grad_norm": 0.5550066871107131, + "learning_rate": 9.879990541561766e-06, + "loss": 0.063, + "step": 878 + }, + { + "epoch": 0.2334041423260754, + "grad_norm": 0.4307639707627933, + "learning_rate": 9.879511962925651e-06, + "loss": 0.0476, + "step": 879 + }, + { + "epoch": 0.23366967604885822, + "grad_norm": 2.7902947643287472, + "learning_rate": 9.879032443577179e-06, + "loss": 0.2344, + "step": 880 + }, + { + "epoch": 0.233935209771641, + "grad_norm": 0.14563410670518806, + "learning_rate": 9.878551983608794e-06, + "loss": 0.0036, + "step": 881 + }, + { + "epoch": 0.2342007434944238, + "grad_norm": 0.8774501103628786, + "learning_rate": 9.878070583113123e-06, + "loss": 0.0957, + "step": 882 + }, + { + "epoch": 0.2344662772172066, + "grad_norm": 0.5560581059426071, + "learning_rate": 9.877588242182976e-06, + "loss": 0.063, + "step": 883 + }, + { + "epoch": 0.23473181093998938, + "grad_norm": 3.7081276425034457, + "learning_rate": 9.877104960911341e-06, + "loss": 0.3301, + "step": 884 + }, + { + "epoch": 0.23499734466277217, + "grad_norm": 0.22236979379444108, + "learning_rate": 9.87662073939139e-06, + "loss": 0.0074, + "step": 885 + }, + { + "epoch": 0.23526287838555496, + "grad_norm": 0.4092243618958261, + "learning_rate": 9.876135577716475e-06, + "loss": 0.0259, + "step": 886 + }, + { + "epoch": 0.23552841210833775, + "grad_norm": 0.37933414096989954, + "learning_rate": 9.875649475980129e-06, + "loss": 0.0432, + "step": 887 + }, + { + "epoch": 0.23579394583112057, + "grad_norm": 0.5651151674620623, + "learning_rate": 9.875162434276066e-06, + "loss": 0.0579, + "step": 888 + }, + { + "epoch": 0.23605947955390336, + "grad_norm": 0.5162559290907474, + "learning_rate": 9.874674452698184e-06, + "loss": 0.0479, + "step": 889 + }, + { + "epoch": 0.23632501327668615, + "grad_norm": 0.637236975984389, + "learning_rate": 9.874185531340561e-06, + "loss": 0.0889, + "step": 890 + }, + { + "epoch": 0.23659054699946894, + "grad_norm": 0.37992930731240715, + "learning_rate": 9.87369567029745e-06, + "loss": 0.0432, + "step": 891 + }, + { + "epoch": 0.23685608072225173, + "grad_norm": 0.4304890878920463, + "learning_rate": 9.873204869663296e-06, + "loss": 0.0479, + "step": 892 + }, + { + "epoch": 0.23712161444503452, + "grad_norm": 0.15235504129186367, + "learning_rate": 9.872713129532717e-06, + "loss": 0.0052, + "step": 893 + }, + { + "epoch": 0.2373871481678173, + "grad_norm": 0.6165775215284957, + "learning_rate": 9.872220450000514e-06, + "loss": 0.0393, + "step": 894 + }, + { + "epoch": 0.2376526818906001, + "grad_norm": 0.11125726497429037, + "learning_rate": 9.87172683116167e-06, + "loss": 0.0018, + "step": 895 + }, + { + "epoch": 0.2379182156133829, + "grad_norm": 0.3081309691181636, + "learning_rate": 9.871232273111351e-06, + "loss": 0.0317, + "step": 896 + }, + { + "epoch": 0.2381837493361657, + "grad_norm": 0.428579523646676, + "learning_rate": 9.8707367759449e-06, + "loss": 0.0479, + "step": 897 + }, + { + "epoch": 0.2384492830589485, + "grad_norm": 1.2174720808088506, + "learning_rate": 9.870240339757844e-06, + "loss": 0.0752, + "step": 898 + }, + { + "epoch": 0.2387148167817313, + "grad_norm": 0.3515404509638987, + "learning_rate": 9.869742964645889e-06, + "loss": 0.0393, + "step": 899 + }, + { + "epoch": 0.23898035050451408, + "grad_norm": 0.36157621557162617, + "learning_rate": 9.869244650704924e-06, + "loss": 0.0393, + "step": 900 + }, + { + "epoch": 0.23924588422729687, + "grad_norm": 0.4552592244911392, + "learning_rate": 9.868745398031018e-06, + "loss": 0.0479, + "step": 901 + }, + { + "epoch": 0.23951141795007966, + "grad_norm": 0.7564817410807659, + "learning_rate": 9.86824520672042e-06, + "loss": 0.1191, + "step": 902 + }, + { + "epoch": 0.23977695167286245, + "grad_norm": 0.3519836929818125, + "learning_rate": 9.867744076869564e-06, + "loss": 0.0354, + "step": 903 + }, + { + "epoch": 0.24004248539564524, + "grad_norm": 0.5911859305106711, + "learning_rate": 9.867242008575056e-06, + "loss": 0.0579, + "step": 904 + }, + { + "epoch": 0.24030801911842803, + "grad_norm": 0.7584559784767734, + "learning_rate": 9.866739001933695e-06, + "loss": 0.1113, + "step": 905 + }, + { + "epoch": 0.24057355284121085, + "grad_norm": 0.7691613201005841, + "learning_rate": 9.866235057042453e-06, + "loss": 0.1035, + "step": 906 + }, + { + "epoch": 0.24083908656399364, + "grad_norm": 0.7497833122031072, + "learning_rate": 9.865730173998482e-06, + "loss": 0.1191, + "step": 907 + }, + { + "epoch": 0.24110462028677643, + "grad_norm": 0.754572113206245, + "learning_rate": 9.86522435289912e-06, + "loss": 0.1279, + "step": 908 + }, + { + "epoch": 0.24137015400955922, + "grad_norm": 0.8672227298778474, + "learning_rate": 9.864717593841884e-06, + "loss": 0.0752, + "step": 909 + }, + { + "epoch": 0.241635687732342, + "grad_norm": 0.4009604044104567, + "learning_rate": 9.864209896924468e-06, + "loss": 0.0393, + "step": 910 + }, + { + "epoch": 0.2419012214551248, + "grad_norm": 0.2966195431778511, + "learning_rate": 9.863701262244753e-06, + "loss": 0.032, + "step": 911 + }, + { + "epoch": 0.2421667551779076, + "grad_norm": 0.5999702384639466, + "learning_rate": 9.863191689900796e-06, + "loss": 0.063, + "step": 912 + }, + { + "epoch": 0.24243228890069038, + "grad_norm": 0.5132953676012183, + "learning_rate": 9.862681179990838e-06, + "loss": 0.0231, + "step": 913 + }, + { + "epoch": 0.24269782262347317, + "grad_norm": 0.6668918417364565, + "learning_rate": 9.862169732613298e-06, + "loss": 0.0525, + "step": 914 + }, + { + "epoch": 0.242963356346256, + "grad_norm": 0.6465213387617673, + "learning_rate": 9.861657347866778e-06, + "loss": 0.0957, + "step": 915 + }, + { + "epoch": 0.24322889006903878, + "grad_norm": 0.2980696299642875, + "learning_rate": 9.861144025850058e-06, + "loss": 0.0317, + "step": 916 + }, + { + "epoch": 0.24349442379182157, + "grad_norm": 0.5561195146487935, + "learning_rate": 9.860629766662103e-06, + "loss": 0.0354, + "step": 917 + }, + { + "epoch": 0.24375995751460436, + "grad_norm": 0.8511473475660397, + "learning_rate": 9.860114570402055e-06, + "loss": 0.0354, + "step": 918 + }, + { + "epoch": 0.24402549123738715, + "grad_norm": 1.0193129893064767, + "learning_rate": 9.859598437169236e-06, + "loss": 0.082, + "step": 919 + }, + { + "epoch": 0.24429102496016994, + "grad_norm": 0.38753475869451304, + "learning_rate": 9.859081367063154e-06, + "loss": 0.0432, + "step": 920 + }, + { + "epoch": 0.24455655868295273, + "grad_norm": 1.2376062406421997, + "learning_rate": 9.858563360183491e-06, + "loss": 0.0815, + "step": 921 + }, + { + "epoch": 0.24482209240573552, + "grad_norm": 0.2614059783802485, + "learning_rate": 9.858044416630113e-06, + "loss": 0.0288, + "step": 922 + }, + { + "epoch": 0.2450876261285183, + "grad_norm": 0.7298755606796804, + "learning_rate": 9.857524536503067e-06, + "loss": 0.1191, + "step": 923 + }, + { + "epoch": 0.24535315985130113, + "grad_norm": 0.2655675767167442, + "learning_rate": 9.85700371990258e-06, + "loss": 0.0286, + "step": 924 + }, + { + "epoch": 0.24561869357408392, + "grad_norm": 0.6867605169602318, + "learning_rate": 9.85648196692906e-06, + "loss": 0.063, + "step": 925 + }, + { + "epoch": 0.2458842272968667, + "grad_norm": 0.7154487582123846, + "learning_rate": 9.855959277683092e-06, + "loss": 0.1113, + "step": 926 + }, + { + "epoch": 0.2461497610196495, + "grad_norm": 0.7662936924895251, + "learning_rate": 9.855435652265446e-06, + "loss": 0.0889, + "step": 927 + }, + { + "epoch": 0.2464152947424323, + "grad_norm": 0.9128535852883102, + "learning_rate": 9.854911090777071e-06, + "loss": 0.0815, + "step": 928 + }, + { + "epoch": 0.24668082846521508, + "grad_norm": 0.6423625754170631, + "learning_rate": 9.854385593319096e-06, + "loss": 0.082, + "step": 929 + }, + { + "epoch": 0.24694636218799787, + "grad_norm": 0.3506830449134971, + "learning_rate": 9.853859159992831e-06, + "loss": 0.0393, + "step": 930 + }, + { + "epoch": 0.24721189591078066, + "grad_norm": 0.6686174900920513, + "learning_rate": 9.853331790899765e-06, + "loss": 0.0957, + "step": 931 + }, + { + "epoch": 0.24747742963356345, + "grad_norm": 0.5803696610157401, + "learning_rate": 9.852803486141571e-06, + "loss": 0.0479, + "step": 932 + }, + { + "epoch": 0.24774296335634627, + "grad_norm": 0.6367144973315528, + "learning_rate": 9.852274245820097e-06, + "loss": 0.0889, + "step": 933 + }, + { + "epoch": 0.24800849707912906, + "grad_norm": 0.6019312344886419, + "learning_rate": 9.851744070037376e-06, + "loss": 0.082, + "step": 934 + }, + { + "epoch": 0.24827403080191185, + "grad_norm": 0.40743643859287837, + "learning_rate": 9.851212958895621e-06, + "loss": 0.032, + "step": 935 + }, + { + "epoch": 0.24853956452469464, + "grad_norm": 0.5172725409243906, + "learning_rate": 9.85068091249722e-06, + "loss": 0.063, + "step": 936 + }, + { + "epoch": 0.24880509824747743, + "grad_norm": 0.4756469334653875, + "learning_rate": 9.850147930944748e-06, + "loss": 0.0525, + "step": 937 + }, + { + "epoch": 0.24907063197026022, + "grad_norm": 0.4365306093001885, + "learning_rate": 9.849614014340959e-06, + "loss": 0.0432, + "step": 938 + }, + { + "epoch": 0.249336165693043, + "grad_norm": 0.6047425723994772, + "learning_rate": 9.849079162788782e-06, + "loss": 0.0752, + "step": 939 + }, + { + "epoch": 0.2496016994158258, + "grad_norm": 0.5680420397578201, + "learning_rate": 9.848543376391333e-06, + "loss": 0.0693, + "step": 940 + }, + { + "epoch": 0.2498672331386086, + "grad_norm": 0.8494400098616096, + "learning_rate": 9.848006655251904e-06, + "loss": 0.0393, + "step": 941 + }, + { + "epoch": 0.2501327668613914, + "grad_norm": 0.20707760476213186, + "learning_rate": 9.847468999473967e-06, + "loss": 0.0074, + "step": 942 + }, + { + "epoch": 0.25039830058417417, + "grad_norm": 0.5599934007032675, + "learning_rate": 9.846930409161179e-06, + "loss": 0.063, + "step": 943 + }, + { + "epoch": 0.250663834306957, + "grad_norm": 0.5738129646742892, + "learning_rate": 9.846390884417374e-06, + "loss": 0.0693, + "step": 944 + }, + { + "epoch": 0.25092936802973975, + "grad_norm": 0.3121394791058085, + "learning_rate": 9.845850425346563e-06, + "loss": 0.0093, + "step": 945 + }, + { + "epoch": 0.25119490175252257, + "grad_norm": 0.9012275305666327, + "learning_rate": 9.845309032052941e-06, + "loss": 0.1035, + "step": 946 + }, + { + "epoch": 0.2514604354753054, + "grad_norm": 0.8132551945926066, + "learning_rate": 9.844766704640883e-06, + "loss": 0.1035, + "step": 947 + }, + { + "epoch": 0.25172596919808815, + "grad_norm": 0.5059511213268292, + "learning_rate": 9.844223443214942e-06, + "loss": 0.0576, + "step": 948 + }, + { + "epoch": 0.25199150292087097, + "grad_norm": 0.5004469864929803, + "learning_rate": 9.843679247879856e-06, + "loss": 0.0579, + "step": 949 + }, + { + "epoch": 0.25225703664365373, + "grad_norm": 0.5325050038377581, + "learning_rate": 9.843134118740535e-06, + "loss": 0.063, + "step": 950 + }, + { + "epoch": 0.25252257036643655, + "grad_norm": 0.0018393961535569554, + "learning_rate": 9.842588055902076e-06, + "loss": 0.0, + "step": 951 + }, + { + "epoch": 0.2527881040892193, + "grad_norm": 0.29281234077102586, + "learning_rate": 9.842041059469752e-06, + "loss": 0.0117, + "step": 952 + }, + { + "epoch": 0.25305363781200213, + "grad_norm": 4.655974030022523, + "learning_rate": 9.841493129549018e-06, + "loss": 0.4082, + "step": 953 + }, + { + "epoch": 0.2533191715347849, + "grad_norm": 0.345197837145536, + "learning_rate": 9.840944266245511e-06, + "loss": 0.0354, + "step": 954 + }, + { + "epoch": 0.2535847052575677, + "grad_norm": 0.5150522007650085, + "learning_rate": 9.840394469665043e-06, + "loss": 0.032, + "step": 955 + }, + { + "epoch": 0.25385023898035053, + "grad_norm": 4.355092056512744, + "learning_rate": 9.839843739913606e-06, + "loss": 0.3438, + "step": 956 + }, + { + "epoch": 0.2541157727031333, + "grad_norm": 0.5533036225395148, + "learning_rate": 9.839292077097378e-06, + "loss": 0.063, + "step": 957 + }, + { + "epoch": 0.2543813064259161, + "grad_norm": 0.47245181913716905, + "learning_rate": 9.838739481322711e-06, + "loss": 0.032, + "step": 958 + }, + { + "epoch": 0.25464684014869887, + "grad_norm": 0.015777567489265145, + "learning_rate": 9.838185952696139e-06, + "loss": 0.0004, + "step": 959 + }, + { + "epoch": 0.2549123738714817, + "grad_norm": 0.3070632948499519, + "learning_rate": 9.837631491324379e-06, + "loss": 0.0317, + "step": 960 + }, + { + "epoch": 0.25517790759426445, + "grad_norm": 0.6417501675077917, + "learning_rate": 9.83707609731432e-06, + "loss": 0.0525, + "step": 961 + }, + { + "epoch": 0.25544344131704727, + "grad_norm": 1.103140251013234, + "learning_rate": 9.836519770773038e-06, + "loss": 0.0889, + "step": 962 + }, + { + "epoch": 0.25570897503983003, + "grad_norm": 0.36859782996784346, + "learning_rate": 9.835962511807786e-06, + "loss": 0.0391, + "step": 963 + }, + { + "epoch": 0.25597450876261285, + "grad_norm": 0.3995475012233262, + "learning_rate": 9.835404320525997e-06, + "loss": 0.0432, + "step": 964 + }, + { + "epoch": 0.25624004248539567, + "grad_norm": 0.6669653652071353, + "learning_rate": 9.834845197035282e-06, + "loss": 0.0752, + "step": 965 + }, + { + "epoch": 0.25650557620817843, + "grad_norm": 0.9463153272300644, + "learning_rate": 9.834285141443436e-06, + "loss": 0.082, + "step": 966 + }, + { + "epoch": 0.25677110993096125, + "grad_norm": 0.7472061020733514, + "learning_rate": 9.833724153858428e-06, + "loss": 0.0479, + "step": 967 + }, + { + "epoch": 0.257036643653744, + "grad_norm": 0.5964140946194706, + "learning_rate": 9.833162234388413e-06, + "loss": 0.0525, + "step": 968 + }, + { + "epoch": 0.25730217737652683, + "grad_norm": 0.69941976438153, + "learning_rate": 9.832599383141722e-06, + "loss": 0.0957, + "step": 969 + }, + { + "epoch": 0.2575677110993096, + "grad_norm": 0.6938979357435278, + "learning_rate": 9.832035600226863e-06, + "loss": 0.0957, + "step": 970 + }, + { + "epoch": 0.2578332448220924, + "grad_norm": 1.6505906761311266, + "learning_rate": 9.83147088575253e-06, + "loss": 0.0957, + "step": 971 + }, + { + "epoch": 0.2580987785448752, + "grad_norm": 0.2571270449075492, + "learning_rate": 9.830905239827592e-06, + "loss": 0.0258, + "step": 972 + }, + { + "epoch": 0.258364312267658, + "grad_norm": 0.17362195702763106, + "learning_rate": 9.830338662561098e-06, + "loss": 0.0059, + "step": 973 + }, + { + "epoch": 0.2586298459904408, + "grad_norm": 0.3252416871542705, + "learning_rate": 9.82977115406228e-06, + "loss": 0.0354, + "step": 974 + }, + { + "epoch": 0.2588953797132236, + "grad_norm": 0.7560895172752308, + "learning_rate": 9.829202714440544e-06, + "loss": 0.0752, + "step": 975 + }, + { + "epoch": 0.2591609134360064, + "grad_norm": 0.8515843549515478, + "learning_rate": 9.828633343805482e-06, + "loss": 0.032, + "step": 976 + }, + { + "epoch": 0.25942644715878915, + "grad_norm": 0.06839867591689461, + "learning_rate": 9.828063042266858e-06, + "loss": 0.0018, + "step": 977 + }, + { + "epoch": 0.25969198088157197, + "grad_norm": 0.2973877835283162, + "learning_rate": 9.827491809934621e-06, + "loss": 0.0317, + "step": 978 + }, + { + "epoch": 0.25995751460435473, + "grad_norm": 0.050237992088855266, + "learning_rate": 9.826919646918899e-06, + "loss": 0.0014, + "step": 979 + }, + { + "epoch": 0.26022304832713755, + "grad_norm": 0.6244940971022646, + "learning_rate": 9.826346553329997e-06, + "loss": 0.063, + "step": 980 + }, + { + "epoch": 0.2604885820499203, + "grad_norm": 0.3832845757482919, + "learning_rate": 9.825772529278402e-06, + "loss": 0.0393, + "step": 981 + }, + { + "epoch": 0.26075411577270313, + "grad_norm": 0.5607756019076436, + "learning_rate": 9.825197574874775e-06, + "loss": 0.0288, + "step": 982 + }, + { + "epoch": 0.26101964949548595, + "grad_norm": 0.3614253694825283, + "learning_rate": 9.824621690229966e-06, + "loss": 0.0354, + "step": 983 + }, + { + "epoch": 0.2612851832182687, + "grad_norm": 1.7457333721108772, + "learning_rate": 9.824044875454995e-06, + "loss": 0.0815, + "step": 984 + }, + { + "epoch": 0.26155071694105153, + "grad_norm": 0.39656420534134673, + "learning_rate": 9.823467130661068e-06, + "loss": 0.0354, + "step": 985 + }, + { + "epoch": 0.2618162506638343, + "grad_norm": 0.4441491434154354, + "learning_rate": 9.822888455959564e-06, + "loss": 0.0391, + "step": 986 + }, + { + "epoch": 0.2620817843866171, + "grad_norm": 0.9091935280621182, + "learning_rate": 9.822308851462046e-06, + "loss": 0.063, + "step": 987 + }, + { + "epoch": 0.2623473181093999, + "grad_norm": 0.2809293161710395, + "learning_rate": 9.821728317280257e-06, + "loss": 0.0258, + "step": 988 + }, + { + "epoch": 0.2626128518321827, + "grad_norm": 0.25690503502749795, + "learning_rate": 9.821146853526116e-06, + "loss": 0.0231, + "step": 989 + }, + { + "epoch": 0.26287838555496545, + "grad_norm": 0.9504240651916706, + "learning_rate": 9.820564460311719e-06, + "loss": 0.1035, + "step": 990 + }, + { + "epoch": 0.2631439192777483, + "grad_norm": 1.062595886433007, + "learning_rate": 9.81998113774935e-06, + "loss": 0.1553, + "step": 991 + }, + { + "epoch": 0.2634094530005311, + "grad_norm": 0.08306616186641885, + "learning_rate": 9.819396885951462e-06, + "loss": 0.0022, + "step": 992 + }, + { + "epoch": 0.26367498672331385, + "grad_norm": 0.00883581144917326, + "learning_rate": 9.818811705030696e-06, + "loss": 0.0002, + "step": 993 + }, + { + "epoch": 0.26394052044609667, + "grad_norm": 0.1998801649088699, + "learning_rate": 9.818225595099864e-06, + "loss": 0.0165, + "step": 994 + }, + { + "epoch": 0.26420605416887943, + "grad_norm": 0.8818331545317252, + "learning_rate": 9.817638556271964e-06, + "loss": 0.1367, + "step": 995 + }, + { + "epoch": 0.26447158789166225, + "grad_norm": 1.6642224007144948, + "learning_rate": 9.81705058866017e-06, + "loss": 0.0752, + "step": 996 + }, + { + "epoch": 0.264737121614445, + "grad_norm": 0.719955269890806, + "learning_rate": 9.816461692377833e-06, + "loss": 0.0317, + "step": 997 + }, + { + "epoch": 0.26500265533722783, + "grad_norm": 0.7209406456988616, + "learning_rate": 9.815871867538486e-06, + "loss": 0.0476, + "step": 998 + }, + { + "epoch": 0.2652681890600106, + "grad_norm": 0.814061894388583, + "learning_rate": 9.815281114255841e-06, + "loss": 0.063, + "step": 999 + }, + { + "epoch": 0.2655337227827934, + "grad_norm": 0.9966166060448677, + "learning_rate": 9.81468943264379e-06, + "loss": 0.1035, + "step": 1000 + }, + { + "epoch": 0.26579925650557623, + "grad_norm": 0.22419755071513356, + "learning_rate": 9.814096822816398e-06, + "loss": 0.0061, + "step": 1001 + }, + { + "epoch": 0.266064790228359, + "grad_norm": 0.2754600582043281, + "learning_rate": 9.813503284887916e-06, + "loss": 0.0258, + "step": 1002 + }, + { + "epoch": 0.2663303239511418, + "grad_norm": 1.1239681806680273, + "learning_rate": 9.812908818972771e-06, + "loss": 0.0889, + "step": 1003 + }, + { + "epoch": 0.2665958576739246, + "grad_norm": 0.5803488339467218, + "learning_rate": 9.81231342518557e-06, + "loss": 0.0476, + "step": 1004 + }, + { + "epoch": 0.2668613913967074, + "grad_norm": 0.35519796165827866, + "learning_rate": 9.811717103641096e-06, + "loss": 0.0286, + "step": 1005 + }, + { + "epoch": 0.26712692511949016, + "grad_norm": 2.462660544438473, + "learning_rate": 9.811119854454313e-06, + "loss": 0.0957, + "step": 1006 + }, + { + "epoch": 0.267392458842273, + "grad_norm": 0.0859060653931223, + "learning_rate": 9.810521677740363e-06, + "loss": 0.0022, + "step": 1007 + }, + { + "epoch": 0.26765799256505574, + "grad_norm": 0.21988261467431752, + "learning_rate": 9.80992257361457e-06, + "loss": 0.0206, + "step": 1008 + }, + { + "epoch": 0.26792352628783855, + "grad_norm": 0.37935970444692796, + "learning_rate": 9.809322542192432e-06, + "loss": 0.0317, + "step": 1009 + }, + { + "epoch": 0.26818906001062137, + "grad_norm": 0.3156766850538133, + "learning_rate": 9.808721583589628e-06, + "loss": 0.0286, + "step": 1010 + }, + { + "epoch": 0.26845459373340413, + "grad_norm": 0.17021294848589957, + "learning_rate": 9.808119697922016e-06, + "loss": 0.0146, + "step": 1011 + }, + { + "epoch": 0.26872012745618695, + "grad_norm": 1.3132137393142103, + "learning_rate": 9.807516885305634e-06, + "loss": 0.1641, + "step": 1012 + }, + { + "epoch": 0.2689856611789697, + "grad_norm": 0.14707242483307284, + "learning_rate": 9.806913145856697e-06, + "loss": 0.0115, + "step": 1013 + }, + { + "epoch": 0.26925119490175253, + "grad_norm": 3.511760120759758, + "learning_rate": 9.806308479691595e-06, + "loss": 0.2061, + "step": 1014 + }, + { + "epoch": 0.2695167286245353, + "grad_norm": 0.8061004291701489, + "learning_rate": 9.805702886926904e-06, + "loss": 0.0261, + "step": 1015 + }, + { + "epoch": 0.2697822623473181, + "grad_norm": 0.2234690978519963, + "learning_rate": 9.805096367679375e-06, + "loss": 0.0184, + "step": 1016 + }, + { + "epoch": 0.2700477960701009, + "grad_norm": 0.44975353193446077, + "learning_rate": 9.804488922065937e-06, + "loss": 0.0258, + "step": 1017 + }, + { + "epoch": 0.2703133297928837, + "grad_norm": 0.6109060230396508, + "learning_rate": 9.8038805502037e-06, + "loss": 0.0354, + "step": 1018 + }, + { + "epoch": 0.2705788635156665, + "grad_norm": 0.19828294757841264, + "learning_rate": 9.803271252209947e-06, + "loss": 0.0184, + "step": 1019 + }, + { + "epoch": 0.2708443972384493, + "grad_norm": 0.25772979651426625, + "learning_rate": 9.802661028202147e-06, + "loss": 0.0165, + "step": 1020 + }, + { + "epoch": 0.2711099309612321, + "grad_norm": 1.1839479936613346, + "learning_rate": 9.802049878297942e-06, + "loss": 0.0525, + "step": 1021 + }, + { + "epoch": 0.27137546468401486, + "grad_norm": 0.31430038721241105, + "learning_rate": 9.801437802615156e-06, + "loss": 0.0082, + "step": 1022 + }, + { + "epoch": 0.2716409984067977, + "grad_norm": 1.1023674078125263, + "learning_rate": 9.80082480127179e-06, + "loss": 0.1191, + "step": 1023 + }, + { + "epoch": 0.27190653212958044, + "grad_norm": 0.22162338231371304, + "learning_rate": 9.800210874386022e-06, + "loss": 0.0117, + "step": 1024 + }, + { + "epoch": 0.27217206585236325, + "grad_norm": 0.020666537436477367, + "learning_rate": 9.79959602207621e-06, + "loss": 0.0004, + "step": 1025 + }, + { + "epoch": 0.272437599575146, + "grad_norm": 0.1640864385317412, + "learning_rate": 9.798980244460892e-06, + "loss": 0.0131, + "step": 1026 + }, + { + "epoch": 0.27270313329792883, + "grad_norm": 1.33096980689375, + "learning_rate": 9.798363541658782e-06, + "loss": 0.1455, + "step": 1027 + }, + { + "epoch": 0.27296866702071165, + "grad_norm": 0.0030561000010345305, + "learning_rate": 9.797745913788772e-06, + "loss": 0.0001, + "step": 1028 + }, + { + "epoch": 0.2732342007434944, + "grad_norm": 0.3353151513139151, + "learning_rate": 9.797127360969933e-06, + "loss": 0.0093, + "step": 1029 + }, + { + "epoch": 0.27349973446627723, + "grad_norm": 0.24279565510695517, + "learning_rate": 9.796507883321519e-06, + "loss": 0.0165, + "step": 1030 + }, + { + "epoch": 0.27376526818906, + "grad_norm": 0.6409923619811838, + "learning_rate": 9.795887480962951e-06, + "loss": 0.0206, + "step": 1031 + }, + { + "epoch": 0.2740308019118428, + "grad_norm": 0.1865018425314411, + "learning_rate": 9.79526615401384e-06, + "loss": 0.0146, + "step": 1032 + }, + { + "epoch": 0.2742963356346256, + "grad_norm": 0.5691428064000503, + "learning_rate": 9.79464390259397e-06, + "loss": 0.0354, + "step": 1033 + }, + { + "epoch": 0.2745618693574084, + "grad_norm": 1.0718553071053665, + "learning_rate": 9.794020726823303e-06, + "loss": 0.1934, + "step": 1034 + }, + { + "epoch": 0.27482740308019116, + "grad_norm": 1.338016801862817, + "learning_rate": 9.79339662682198e-06, + "loss": 0.1279, + "step": 1035 + }, + { + "epoch": 0.275092936802974, + "grad_norm": 0.3434764844856958, + "learning_rate": 9.79277160271032e-06, + "loss": 0.0229, + "step": 1036 + }, + { + "epoch": 0.2753584705257568, + "grad_norm": 0.9536560542927063, + "learning_rate": 9.79214565460882e-06, + "loss": 0.1367, + "step": 1037 + }, + { + "epoch": 0.27562400424853956, + "grad_norm": 0.2777647682007733, + "learning_rate": 9.791518782638157e-06, + "loss": 0.0229, + "step": 1038 + }, + { + "epoch": 0.2758895379713224, + "grad_norm": 2.8846935746827778, + "learning_rate": 9.790890986919184e-06, + "loss": 0.1035, + "step": 1039 + }, + { + "epoch": 0.27615507169410514, + "grad_norm": 0.9622843736990357, + "learning_rate": 9.79026226757293e-06, + "loss": 0.1455, + "step": 1040 + }, + { + "epoch": 0.27642060541688795, + "grad_norm": 1.3517743387903762, + "learning_rate": 9.789632624720608e-06, + "loss": 0.1191, + "step": 1041 + }, + { + "epoch": 0.2766861391396707, + "grad_norm": 0.3315039166067069, + "learning_rate": 9.789002058483604e-06, + "loss": 0.0286, + "step": 1042 + }, + { + "epoch": 0.27695167286245354, + "grad_norm": 0.0228123965731073, + "learning_rate": 9.788370568983483e-06, + "loss": 0.0005, + "step": 1043 + }, + { + "epoch": 0.2772172065852363, + "grad_norm": 1.4788814320601762, + "learning_rate": 9.787738156341992e-06, + "loss": 0.0393, + "step": 1044 + }, + { + "epoch": 0.2774827403080191, + "grad_norm": 1.146212270628032, + "learning_rate": 9.78710482068105e-06, + "loss": 0.0649, + "step": 1045 + }, + { + "epoch": 0.27774827403080193, + "grad_norm": 0.5981452393985338, + "learning_rate": 9.786470562122758e-06, + "loss": 0.0391, + "step": 1046 + }, + { + "epoch": 0.2780138077535847, + "grad_norm": 0.10195539381598936, + "learning_rate": 9.78583538078939e-06, + "loss": 0.0026, + "step": 1047 + }, + { + "epoch": 0.2782793414763675, + "grad_norm": 0.3185752679243096, + "learning_rate": 9.785199276803408e-06, + "loss": 0.0317, + "step": 1048 + }, + { + "epoch": 0.2785448751991503, + "grad_norm": 0.3967640673554989, + "learning_rate": 9.78456225028744e-06, + "loss": 0.0432, + "step": 1049 + }, + { + "epoch": 0.2788104089219331, + "grad_norm": 0.7810124894163271, + "learning_rate": 9.783924301364297e-06, + "loss": 0.0184, + "step": 1050 + }, + { + "epoch": 0.27907594264471586, + "grad_norm": 0.08506478322426231, + "learning_rate": 9.783285430156972e-06, + "loss": 0.0022, + "step": 1051 + }, + { + "epoch": 0.2793414763674987, + "grad_norm": 0.7008776117228477, + "learning_rate": 9.782645636788628e-06, + "loss": 0.0476, + "step": 1052 + }, + { + "epoch": 0.27960701009028144, + "grad_norm": 0.3314725833848813, + "learning_rate": 9.782004921382612e-06, + "loss": 0.0354, + "step": 1053 + }, + { + "epoch": 0.27987254381306426, + "grad_norm": 0.03864368830561217, + "learning_rate": 9.781363284062446e-06, + "loss": 0.0008, + "step": 1054 + }, + { + "epoch": 0.2801380775358471, + "grad_norm": 0.5162549754187832, + "learning_rate": 9.78072072495183e-06, + "loss": 0.0479, + "step": 1055 + }, + { + "epoch": 0.28040361125862984, + "grad_norm": 0.5385016039999986, + "learning_rate": 9.780077244174642e-06, + "loss": 0.0184, + "step": 1056 + }, + { + "epoch": 0.28066914498141265, + "grad_norm": 0.39671754689268834, + "learning_rate": 9.779432841854937e-06, + "loss": 0.0391, + "step": 1057 + }, + { + "epoch": 0.2809346787041954, + "grad_norm": 0.3533088509471452, + "learning_rate": 9.778787518116948e-06, + "loss": 0.0354, + "step": 1058 + }, + { + "epoch": 0.28120021242697824, + "grad_norm": 0.8001393279845374, + "learning_rate": 9.778141273085085e-06, + "loss": 0.0752, + "step": 1059 + }, + { + "epoch": 0.281465746149761, + "grad_norm": 0.9390525477977365, + "learning_rate": 9.777494106883938e-06, + "loss": 0.063, + "step": 1060 + }, + { + "epoch": 0.2817312798725438, + "grad_norm": 0.09034297328274829, + "learning_rate": 9.776846019638273e-06, + "loss": 0.0015, + "step": 1061 + }, + { + "epoch": 0.28199681359532663, + "grad_norm": 0.3397054585962954, + "learning_rate": 9.776197011473034e-06, + "loss": 0.0073, + "step": 1062 + }, + { + "epoch": 0.2822623473181094, + "grad_norm": 0.8269003857531309, + "learning_rate": 9.775547082513339e-06, + "loss": 0.0391, + "step": 1063 + }, + { + "epoch": 0.2825278810408922, + "grad_norm": 11.321097522244969, + "learning_rate": 9.77489623288449e-06, + "loss": 0.3203, + "step": 1064 + }, + { + "epoch": 0.282793414763675, + "grad_norm": 0.3961581646443348, + "learning_rate": 9.774244462711962e-06, + "loss": 0.0354, + "step": 1065 + }, + { + "epoch": 0.2830589484864578, + "grad_norm": 0.9979585458391305, + "learning_rate": 9.773591772121409e-06, + "loss": 0.0815, + "step": 1066 + }, + { + "epoch": 0.28332448220924056, + "grad_norm": 0.24274949520810776, + "learning_rate": 9.77293816123866e-06, + "loss": 0.0051, + "step": 1067 + }, + { + "epoch": 0.2835900159320234, + "grad_norm": 5.344035497507486, + "learning_rate": 9.772283630189727e-06, + "loss": 0.2773, + "step": 1068 + }, + { + "epoch": 0.28385554965480614, + "grad_norm": 0.45994201443551747, + "learning_rate": 9.771628179100793e-06, + "loss": 0.0391, + "step": 1069 + }, + { + "epoch": 0.28412108337758896, + "grad_norm": 1.055834204920163, + "learning_rate": 9.770971808098221e-06, + "loss": 0.0525, + "step": 1070 + }, + { + "epoch": 0.2843866171003718, + "grad_norm": 0.5618926736078067, + "learning_rate": 9.770314517308554e-06, + "loss": 0.0317, + "step": 1071 + }, + { + "epoch": 0.28465215082315454, + "grad_norm": 0.24186740542409346, + "learning_rate": 9.76965630685851e-06, + "loss": 0.0206, + "step": 1072 + }, + { + "epoch": 0.28491768454593736, + "grad_norm": 0.9072361060971639, + "learning_rate": 9.768997176874982e-06, + "loss": 0.1191, + "step": 1073 + }, + { + "epoch": 0.2851832182687201, + "grad_norm": 1.1218511212461664, + "learning_rate": 9.768337127485044e-06, + "loss": 0.0576, + "step": 1074 + }, + { + "epoch": 0.28544875199150294, + "grad_norm": 0.2496818724866189, + "learning_rate": 9.767676158815944e-06, + "loss": 0.0073, + "step": 1075 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 1.5823043624366155, + "learning_rate": 9.76701427099511e-06, + "loss": 0.0688, + "step": 1076 + }, + { + "epoch": 0.2859798194370685, + "grad_norm": 2.327890956163168, + "learning_rate": 9.766351464150146e-06, + "loss": 0.1738, + "step": 1077 + }, + { + "epoch": 0.2862453531598513, + "grad_norm": 1.8266555516390197, + "learning_rate": 9.765687738408834e-06, + "loss": 0.0957, + "step": 1078 + }, + { + "epoch": 0.2865108868826341, + "grad_norm": 0.21974129050224078, + "learning_rate": 9.765023093899133e-06, + "loss": 0.0184, + "step": 1079 + }, + { + "epoch": 0.2867764206054169, + "grad_norm": 1.213711313563219, + "learning_rate": 9.764357530749178e-06, + "loss": 0.0576, + "step": 1080 + }, + { + "epoch": 0.2870419543281997, + "grad_norm": 2.075553937576423, + "learning_rate": 9.76369104908728e-06, + "loss": 0.0957, + "step": 1081 + }, + { + "epoch": 0.2873074880509825, + "grad_norm": 0.927208051761271, + "learning_rate": 9.76302364904193e-06, + "loss": 0.1455, + "step": 1082 + }, + { + "epoch": 0.28757302177376526, + "grad_norm": 0.7356955614292838, + "learning_rate": 9.762355330741797e-06, + "loss": 0.0231, + "step": 1083 + }, + { + "epoch": 0.2878385554965481, + "grad_norm": 0.7816526361217149, + "learning_rate": 9.761686094315721e-06, + "loss": 0.0432, + "step": 1084 + }, + { + "epoch": 0.28810408921933084, + "grad_norm": 1.7402446527684008, + "learning_rate": 9.761015939892727e-06, + "loss": 0.0957, + "step": 1085 + }, + { + "epoch": 0.28836962294211366, + "grad_norm": 0.05762993692685398, + "learning_rate": 9.760344867602008e-06, + "loss": 0.0003, + "step": 1086 + }, + { + "epoch": 0.2886351566648964, + "grad_norm": 0.024188141344319784, + "learning_rate": 9.759672877572943e-06, + "loss": 0.0005, + "step": 1087 + }, + { + "epoch": 0.28890069038767924, + "grad_norm": 0.13702148519441235, + "learning_rate": 9.758999969935083e-06, + "loss": 0.0017, + "step": 1088 + }, + { + "epoch": 0.28916622411046206, + "grad_norm": 0.23185659329270913, + "learning_rate": 9.758326144818155e-06, + "loss": 0.0206, + "step": 1089 + }, + { + "epoch": 0.2894317578332448, + "grad_norm": 2.020274635548338, + "learning_rate": 9.757651402352067e-06, + "loss": 0.1191, + "step": 1090 + }, + { + "epoch": 0.28969729155602764, + "grad_norm": 0.23791843579850122, + "learning_rate": 9.756975742666899e-06, + "loss": 0.0231, + "step": 1091 + }, + { + "epoch": 0.2899628252788104, + "grad_norm": 1.5926155722260418, + "learning_rate": 9.756299165892911e-06, + "loss": 0.0576, + "step": 1092 + }, + { + "epoch": 0.2902283590015932, + "grad_norm": 0.29759291019807677, + "learning_rate": 9.755621672160538e-06, + "loss": 0.0258, + "step": 1093 + }, + { + "epoch": 0.290493892724376, + "grad_norm": 0.3257843081640135, + "learning_rate": 9.754943261600396e-06, + "loss": 0.0206, + "step": 1094 + }, + { + "epoch": 0.2907594264471588, + "grad_norm": 0.24127331860074858, + "learning_rate": 9.754263934343272e-06, + "loss": 0.0206, + "step": 1095 + }, + { + "epoch": 0.29102496016994156, + "grad_norm": 0.05613263082076969, + "learning_rate": 9.753583690520132e-06, + "loss": 0.0007, + "step": 1096 + }, + { + "epoch": 0.2912904938927244, + "grad_norm": 0.15571605076216163, + "learning_rate": 9.75290253026212e-06, + "loss": 0.0131, + "step": 1097 + }, + { + "epoch": 0.2915560276155072, + "grad_norm": 0.9548153892509748, + "learning_rate": 9.752220453700556e-06, + "loss": 0.1455, + "step": 1098 + }, + { + "epoch": 0.29182156133828996, + "grad_norm": 0.03386757581041895, + "learning_rate": 9.751537460966935e-06, + "loss": 0.0008, + "step": 1099 + }, + { + "epoch": 0.2920870950610728, + "grad_norm": 1.0372337414427726, + "learning_rate": 9.750853552192931e-06, + "loss": 0.0889, + "step": 1100 + }, + { + "epoch": 0.29235262878385554, + "grad_norm": 0.7281567499028863, + "learning_rate": 9.750168727510394e-06, + "loss": 0.0286, + "step": 1101 + }, + { + "epoch": 0.29261816250663836, + "grad_norm": 0.6380207840024593, + "learning_rate": 9.749482987051349e-06, + "loss": 0.0391, + "step": 1102 + }, + { + "epoch": 0.2928836962294211, + "grad_norm": 0.9729212672078682, + "learning_rate": 9.748796330947998e-06, + "loss": 0.0889, + "step": 1103 + }, + { + "epoch": 0.29314922995220394, + "grad_norm": 1.0018172347011092, + "learning_rate": 9.748108759332722e-06, + "loss": 0.0815, + "step": 1104 + }, + { + "epoch": 0.2934147636749867, + "grad_norm": 0.004923265696314747, + "learning_rate": 9.747420272338078e-06, + "loss": 0.0001, + "step": 1105 + }, + { + "epoch": 0.2936802973977695, + "grad_norm": 0.022021697905026522, + "learning_rate": 9.746730870096794e-06, + "loss": 0.0005, + "step": 1106 + }, + { + "epoch": 0.29394583112055234, + "grad_norm": 1.6710028936610164, + "learning_rate": 9.74604055274178e-06, + "loss": 0.0815, + "step": 1107 + }, + { + "epoch": 0.2942113648433351, + "grad_norm": 1.3634452754585142, + "learning_rate": 9.745349320406123e-06, + "loss": 0.0432, + "step": 1108 + }, + { + "epoch": 0.2944768985661179, + "grad_norm": 0.8146412207177869, + "learning_rate": 9.744657173223082e-06, + "loss": 0.1113, + "step": 1109 + }, + { + "epoch": 0.2947424322889007, + "grad_norm": 1.2206960698761407, + "learning_rate": 9.743964111326098e-06, + "loss": 0.0688, + "step": 1110 + }, + { + "epoch": 0.2950079660116835, + "grad_norm": 0.042821624435204764, + "learning_rate": 9.743270134848782e-06, + "loss": 0.0004, + "step": 1111 + }, + { + "epoch": 0.29527349973446626, + "grad_norm": 0.6103965509662032, + "learning_rate": 9.742575243924925e-06, + "loss": 0.0206, + "step": 1112 + }, + { + "epoch": 0.2955390334572491, + "grad_norm": 0.31275659015640817, + "learning_rate": 9.741879438688495e-06, + "loss": 0.0317, + "step": 1113 + }, + { + "epoch": 0.29580456718003184, + "grad_norm": 0.7391124276417567, + "learning_rate": 9.741182719273636e-06, + "loss": 0.0815, + "step": 1114 + }, + { + "epoch": 0.29607010090281466, + "grad_norm": 1.8465376752624503, + "learning_rate": 9.740485085814665e-06, + "loss": 0.1113, + "step": 1115 + }, + { + "epoch": 0.2963356346255975, + "grad_norm": 0.8365414433585058, + "learning_rate": 9.739786538446076e-06, + "loss": 0.0576, + "step": 1116 + }, + { + "epoch": 0.29660116834838024, + "grad_norm": 0.7185562716195945, + "learning_rate": 9.739087077302547e-06, + "loss": 0.063, + "step": 1117 + }, + { + "epoch": 0.29686670207116306, + "grad_norm": 0.24881622042316806, + "learning_rate": 9.738386702518918e-06, + "loss": 0.0082, + "step": 1118 + }, + { + "epoch": 0.2971322357939458, + "grad_norm": 0.82204521512429, + "learning_rate": 9.737685414230218e-06, + "loss": 0.0476, + "step": 1119 + }, + { + "epoch": 0.29739776951672864, + "grad_norm": 0.4442702937110363, + "learning_rate": 9.736983212571646e-06, + "loss": 0.0476, + "step": 1120 + }, + { + "epoch": 0.2976633032395114, + "grad_norm": 0.017044609666170246, + "learning_rate": 9.73628009767858e-06, + "loss": 0.0004, + "step": 1121 + }, + { + "epoch": 0.2979288369622942, + "grad_norm": 1.2421240055318807, + "learning_rate": 9.73557606968657e-06, + "loss": 0.1035, + "step": 1122 + }, + { + "epoch": 0.298194370685077, + "grad_norm": 0.048110722657898286, + "learning_rate": 9.734871128731342e-06, + "loss": 0.0006, + "step": 1123 + }, + { + "epoch": 0.2984599044078598, + "grad_norm": 4.6600239173016185, + "learning_rate": 9.734165274948806e-06, + "loss": 0.5469, + "step": 1124 + }, + { + "epoch": 0.2987254381306426, + "grad_norm": 0.7244351120756174, + "learning_rate": 9.733458508475038e-06, + "loss": 0.0576, + "step": 1125 + }, + { + "epoch": 0.2989909718534254, + "grad_norm": 0.004283320379219228, + "learning_rate": 9.732750829446295e-06, + "loss": 0.0001, + "step": 1126 + }, + { + "epoch": 0.2992565055762082, + "grad_norm": 0.7629355667225856, + "learning_rate": 9.73204223799901e-06, + "loss": 0.1191, + "step": 1127 + }, + { + "epoch": 0.29952203929899096, + "grad_norm": 0.6208152245082953, + "learning_rate": 9.731332734269791e-06, + "loss": 0.0479, + "step": 1128 + }, + { + "epoch": 0.2997875730217738, + "grad_norm": 0.6038756411851064, + "learning_rate": 9.730622318395422e-06, + "loss": 0.0576, + "step": 1129 + }, + { + "epoch": 0.30005310674455654, + "grad_norm": 0.6098024815769895, + "learning_rate": 9.729910990512862e-06, + "loss": 0.0752, + "step": 1130 + }, + { + "epoch": 0.30031864046733936, + "grad_norm": 0.6672028744949045, + "learning_rate": 9.729198750759248e-06, + "loss": 0.0688, + "step": 1131 + }, + { + "epoch": 0.3005841741901221, + "grad_norm": 0.42173926069039175, + "learning_rate": 9.72848559927189e-06, + "loss": 0.0432, + "step": 1132 + }, + { + "epoch": 0.30084970791290494, + "grad_norm": 0.4224982111516302, + "learning_rate": 9.727771536188275e-06, + "loss": 0.0476, + "step": 1133 + }, + { + "epoch": 0.30111524163568776, + "grad_norm": 0.818705810028351, + "learning_rate": 9.727056561646067e-06, + "loss": 0.0815, + "step": 1134 + }, + { + "epoch": 0.3013807753584705, + "grad_norm": 0.4673281241576781, + "learning_rate": 9.726340675783105e-06, + "loss": 0.0391, + "step": 1135 + }, + { + "epoch": 0.30164630908125334, + "grad_norm": 0.6576605686901854, + "learning_rate": 9.725623878737402e-06, + "loss": 0.0752, + "step": 1136 + }, + { + "epoch": 0.3019118428040361, + "grad_norm": 0.5547502104537158, + "learning_rate": 9.724906170647151e-06, + "loss": 0.0479, + "step": 1137 + }, + { + "epoch": 0.3021773765268189, + "grad_norm": 0.35579660396826157, + "learning_rate": 9.724187551650712e-06, + "loss": 0.0132, + "step": 1138 + }, + { + "epoch": 0.3024429102496017, + "grad_norm": 0.4736433372724878, + "learning_rate": 9.723468021886632e-06, + "loss": 0.0525, + "step": 1139 + }, + { + "epoch": 0.3027084439723845, + "grad_norm": 0.02789314147339348, + "learning_rate": 9.722747581493625e-06, + "loss": 0.0007, + "step": 1140 + }, + { + "epoch": 0.30297397769516726, + "grad_norm": 0.21925197694495466, + "learning_rate": 9.722026230610584e-06, + "loss": 0.0206, + "step": 1141 + }, + { + "epoch": 0.3032395114179501, + "grad_norm": 0.010099703111864153, + "learning_rate": 9.721303969376576e-06, + "loss": 0.0003, + "step": 1142 + }, + { + "epoch": 0.3035050451407329, + "grad_norm": 0.31933908172577696, + "learning_rate": 9.720580797930845e-06, + "loss": 0.0354, + "step": 1143 + }, + { + "epoch": 0.30377057886351566, + "grad_norm": 1.1192764070316166, + "learning_rate": 9.719856716412813e-06, + "loss": 0.0889, + "step": 1144 + }, + { + "epoch": 0.3040361125862985, + "grad_norm": 0.6684300524852204, + "learning_rate": 9.71913172496207e-06, + "loss": 0.1113, + "step": 1145 + }, + { + "epoch": 0.30430164630908124, + "grad_norm": 0.29373443475884653, + "learning_rate": 9.718405823718386e-06, + "loss": 0.0317, + "step": 1146 + }, + { + "epoch": 0.30456718003186406, + "grad_norm": 0.3658446950988331, + "learning_rate": 9.71767901282171e-06, + "loss": 0.0432, + "step": 1147 + }, + { + "epoch": 0.3048327137546468, + "grad_norm": 0.567931719494275, + "learning_rate": 9.716951292412158e-06, + "loss": 0.0432, + "step": 1148 + }, + { + "epoch": 0.30509824747742964, + "grad_norm": 0.06610736970007457, + "learning_rate": 9.71622266263003e-06, + "loss": 0.0016, + "step": 1149 + }, + { + "epoch": 0.3053637812002124, + "grad_norm": 0.24553686436218464, + "learning_rate": 9.715493123615793e-06, + "loss": 0.0258, + "step": 1150 + }, + { + "epoch": 0.3056293149229952, + "grad_norm": 0.3986171599233714, + "learning_rate": 9.714762675510098e-06, + "loss": 0.0432, + "step": 1151 + }, + { + "epoch": 0.30589484864577804, + "grad_norm": 0.03893798946143774, + "learning_rate": 9.714031318453763e-06, + "loss": 0.001, + "step": 1152 + }, + { + "epoch": 0.3061603823685608, + "grad_norm": 0.30020768346470134, + "learning_rate": 9.713299052587788e-06, + "loss": 0.0066, + "step": 1153 + }, + { + "epoch": 0.3064259160913436, + "grad_norm": 0.7987344718538759, + "learning_rate": 9.712565878053344e-06, + "loss": 0.1191, + "step": 1154 + }, + { + "epoch": 0.3066914498141264, + "grad_norm": 0.2734737664568203, + "learning_rate": 9.711831794991777e-06, + "loss": 0.0286, + "step": 1155 + }, + { + "epoch": 0.3069569835369092, + "grad_norm": 0.9407886341613316, + "learning_rate": 9.711096803544612e-06, + "loss": 0.0206, + "step": 1156 + }, + { + "epoch": 0.30722251725969196, + "grad_norm": 0.2662920174573857, + "learning_rate": 9.710360903853545e-06, + "loss": 0.0286, + "step": 1157 + }, + { + "epoch": 0.3074880509824748, + "grad_norm": 0.40813353628480753, + "learning_rate": 9.70962409606045e-06, + "loss": 0.0432, + "step": 1158 + }, + { + "epoch": 0.30775358470525754, + "grad_norm": 0.33115218272803004, + "learning_rate": 9.708886380307373e-06, + "loss": 0.0286, + "step": 1159 + }, + { + "epoch": 0.30801911842804036, + "grad_norm": 2.752464726652913, + "learning_rate": 9.70814775673654e-06, + "loss": 0.1113, + "step": 1160 + }, + { + "epoch": 0.3082846521508232, + "grad_norm": 0.21100244178718114, + "learning_rate": 9.707408225490343e-06, + "loss": 0.0206, + "step": 1161 + }, + { + "epoch": 0.30855018587360594, + "grad_norm": 0.273597747914562, + "learning_rate": 9.706667786711362e-06, + "loss": 0.0286, + "step": 1162 + }, + { + "epoch": 0.30881571959638876, + "grad_norm": 0.5455974875060108, + "learning_rate": 9.705926440542342e-06, + "loss": 0.0476, + "step": 1163 + }, + { + "epoch": 0.3090812533191715, + "grad_norm": 0.04996096845556894, + "learning_rate": 9.705184187126205e-06, + "loss": 0.001, + "step": 1164 + }, + { + "epoch": 0.30934678704195434, + "grad_norm": 0.7661826955072533, + "learning_rate": 9.704441026606047e-06, + "loss": 0.0957, + "step": 1165 + }, + { + "epoch": 0.3096123207647371, + "grad_norm": 0.22898596466557175, + "learning_rate": 9.703696959125144e-06, + "loss": 0.0038, + "step": 1166 + }, + { + "epoch": 0.3098778544875199, + "grad_norm": 1.0310557558318094, + "learning_rate": 9.702951984826943e-06, + "loss": 0.1113, + "step": 1167 + }, + { + "epoch": 0.3101433882103027, + "grad_norm": 0.9506415865468816, + "learning_rate": 9.702206103855065e-06, + "loss": 0.1367, + "step": 1168 + }, + { + "epoch": 0.3104089219330855, + "grad_norm": 0.9519356526243661, + "learning_rate": 9.701459316353309e-06, + "loss": 0.0752, + "step": 1169 + }, + { + "epoch": 0.3106744556558683, + "grad_norm": 0.4403895621363803, + "learning_rate": 9.700711622465645e-06, + "loss": 0.0391, + "step": 1170 + }, + { + "epoch": 0.3109399893786511, + "grad_norm": 0.8734691695278628, + "learning_rate": 9.699963022336219e-06, + "loss": 0.0815, + "step": 1171 + }, + { + "epoch": 0.3112055231014339, + "grad_norm": 0.03692544587443828, + "learning_rate": 9.69921351610935e-06, + "loss": 0.001, + "step": 1172 + }, + { + "epoch": 0.31147105682421666, + "grad_norm": 0.38902223564103433, + "learning_rate": 9.698463103929542e-06, + "loss": 0.0354, + "step": 1173 + }, + { + "epoch": 0.3117365905469995, + "grad_norm": 0.4322873123614092, + "learning_rate": 9.697711785941459e-06, + "loss": 0.0391, + "step": 1174 + }, + { + "epoch": 0.31200212426978224, + "grad_norm": 1.04700194984681, + "learning_rate": 9.696959562289949e-06, + "loss": 0.1035, + "step": 1175 + }, + { + "epoch": 0.31226765799256506, + "grad_norm": 0.6455068902710267, + "learning_rate": 9.69620643312003e-06, + "loss": 0.0525, + "step": 1176 + }, + { + "epoch": 0.3125331917153478, + "grad_norm": 0.19469026077885143, + "learning_rate": 9.695452398576897e-06, + "loss": 0.0206, + "step": 1177 + }, + { + "epoch": 0.31279872543813064, + "grad_norm": 0.552300258388761, + "learning_rate": 9.69469745880592e-06, + "loss": 0.0576, + "step": 1178 + }, + { + "epoch": 0.31306425916091346, + "grad_norm": 0.15378094703192938, + "learning_rate": 9.693941613952642e-06, + "loss": 0.0146, + "step": 1179 + }, + { + "epoch": 0.3133297928836962, + "grad_norm": 0.7149303778962087, + "learning_rate": 9.693184864162781e-06, + "loss": 0.0525, + "step": 1180 + }, + { + "epoch": 0.31359532660647904, + "grad_norm": 12.606742773671819, + "learning_rate": 9.692427209582227e-06, + "loss": 0.3516, + "step": 1181 + }, + { + "epoch": 0.3138608603292618, + "grad_norm": 0.02382454479686282, + "learning_rate": 9.691668650357052e-06, + "loss": 0.0005, + "step": 1182 + }, + { + "epoch": 0.3141263940520446, + "grad_norm": 0.6509559366250702, + "learning_rate": 9.690909186633493e-06, + "loss": 0.063, + "step": 1183 + }, + { + "epoch": 0.3143919277748274, + "grad_norm": 0.22407883750706376, + "learning_rate": 9.690148818557966e-06, + "loss": 0.0231, + "step": 1184 + }, + { + "epoch": 0.3146574614976102, + "grad_norm": 1.194851766940795, + "learning_rate": 9.689387546277062e-06, + "loss": 0.0752, + "step": 1185 + }, + { + "epoch": 0.31492299522039296, + "grad_norm": 1.5889517299989575, + "learning_rate": 9.688625369937548e-06, + "loss": 0.1113, + "step": 1186 + }, + { + "epoch": 0.3151885289431758, + "grad_norm": 0.7528605787401368, + "learning_rate": 9.687862289686356e-06, + "loss": 0.1035, + "step": 1187 + }, + { + "epoch": 0.3154540626659586, + "grad_norm": 0.10046386842043101, + "learning_rate": 9.687098305670606e-06, + "loss": 0.0025, + "step": 1188 + }, + { + "epoch": 0.31571959638874136, + "grad_norm": 0.744431616154801, + "learning_rate": 9.68633341803758e-06, + "loss": 0.1113, + "step": 1189 + }, + { + "epoch": 0.3159851301115242, + "grad_norm": 0.3171470340086223, + "learning_rate": 9.685567626934742e-06, + "loss": 0.0354, + "step": 1190 + }, + { + "epoch": 0.31625066383430694, + "grad_norm": 0.24491006741762908, + "learning_rate": 9.684800932509726e-06, + "loss": 0.0258, + "step": 1191 + }, + { + "epoch": 0.31651619755708976, + "grad_norm": 0.057338171719011885, + "learning_rate": 9.684033334910342e-06, + "loss": 0.0014, + "step": 1192 + }, + { + "epoch": 0.3167817312798725, + "grad_norm": 0.2935840183769047, + "learning_rate": 9.683264834284575e-06, + "loss": 0.0258, + "step": 1193 + }, + { + "epoch": 0.31704726500265534, + "grad_norm": 0.7590487718984419, + "learning_rate": 9.682495430780581e-06, + "loss": 0.1455, + "step": 1194 + }, + { + "epoch": 0.3173127987254381, + "grad_norm": 1.1684076403623338, + "learning_rate": 9.681725124546695e-06, + "loss": 0.0693, + "step": 1195 + }, + { + "epoch": 0.3175783324482209, + "grad_norm": 1.6438208494368356, + "learning_rate": 9.680953915731418e-06, + "loss": 0.082, + "step": 1196 + }, + { + "epoch": 0.31784386617100374, + "grad_norm": 0.8695124690281735, + "learning_rate": 9.680181804483435e-06, + "loss": 0.0479, + "step": 1197 + }, + { + "epoch": 0.3181093998937865, + "grad_norm": 0.20326281659079895, + "learning_rate": 9.679408790951597e-06, + "loss": 0.0231, + "step": 1198 + }, + { + "epoch": 0.3183749336165693, + "grad_norm": 0.6573685494777957, + "learning_rate": 9.678634875284932e-06, + "loss": 0.0187, + "step": 1199 + }, + { + "epoch": 0.3186404673393521, + "grad_norm": 0.7687488235704806, + "learning_rate": 9.677860057632642e-06, + "loss": 0.1455, + "step": 1200 + }, + { + "epoch": 0.3189060010621349, + "grad_norm": 0.12038614362749626, + "learning_rate": 9.677084338144105e-06, + "loss": 0.0117, + "step": 1201 + }, + { + "epoch": 0.31917153478491767, + "grad_norm": 0.22290803237319895, + "learning_rate": 9.676307716968866e-06, + "loss": 0.0258, + "step": 1202 + }, + { + "epoch": 0.3194370685077005, + "grad_norm": 0.08514537747421477, + "learning_rate": 9.675530194256652e-06, + "loss": 0.0018, + "step": 1203 + }, + { + "epoch": 0.31970260223048325, + "grad_norm": 0.5579141610152237, + "learning_rate": 9.674751770157362e-06, + "loss": 0.0233, + "step": 1204 + }, + { + "epoch": 0.31996813595326606, + "grad_norm": 0.312483949561754, + "learning_rate": 9.673972444821063e-06, + "loss": 0.0354, + "step": 1205 + }, + { + "epoch": 0.3202336696760489, + "grad_norm": 0.22701164555191825, + "learning_rate": 9.673192218398e-06, + "loss": 0.0258, + "step": 1206 + }, + { + "epoch": 0.32049920339883164, + "grad_norm": 0.1487192071650038, + "learning_rate": 9.672411091038592e-06, + "loss": 0.0146, + "step": 1207 + }, + { + "epoch": 0.32076473712161446, + "grad_norm": 1.3761300654988238, + "learning_rate": 9.671629062893433e-06, + "loss": 0.1279, + "step": 1208 + }, + { + "epoch": 0.3210302708443972, + "grad_norm": 0.1704844264079489, + "learning_rate": 9.670846134113286e-06, + "loss": 0.0184, + "step": 1209 + }, + { + "epoch": 0.32129580456718004, + "grad_norm": 0.19819201079023002, + "learning_rate": 9.670062304849093e-06, + "loss": 0.0047, + "step": 1210 + }, + { + "epoch": 0.3215613382899628, + "grad_norm": 0.1838454976972731, + "learning_rate": 9.669277575251965e-06, + "loss": 0.0052, + "step": 1211 + }, + { + "epoch": 0.3218268720127456, + "grad_norm": 0.6558129007222472, + "learning_rate": 9.668491945473191e-06, + "loss": 0.0148, + "step": 1212 + }, + { + "epoch": 0.3220924057355284, + "grad_norm": 0.34799526026312055, + "learning_rate": 9.667705415664228e-06, + "loss": 0.0393, + "step": 1213 + }, + { + "epoch": 0.3223579394583112, + "grad_norm": 0.5171512089272212, + "learning_rate": 9.666917985976712e-06, + "loss": 0.0148, + "step": 1214 + }, + { + "epoch": 0.322623473181094, + "grad_norm": 0.2062886511175874, + "learning_rate": 9.66612965656245e-06, + "loss": 0.0231, + "step": 1215 + }, + { + "epoch": 0.3228890069038768, + "grad_norm": 0.7738586973681455, + "learning_rate": 9.665340427573422e-06, + "loss": 0.1641, + "step": 1216 + }, + { + "epoch": 0.3231545406266596, + "grad_norm": 0.03671678604671888, + "learning_rate": 9.664550299161782e-06, + "loss": 0.0008, + "step": 1217 + }, + { + "epoch": 0.32342007434944237, + "grad_norm": 0.19085387586302657, + "learning_rate": 9.663759271479858e-06, + "loss": 0.0206, + "step": 1218 + }, + { + "epoch": 0.3236856080722252, + "grad_norm": 3.034553850378646, + "learning_rate": 9.66296734468015e-06, + "loss": 0.1553, + "step": 1219 + }, + { + "epoch": 0.32395114179500795, + "grad_norm": 0.026411465752670175, + "learning_rate": 9.662174518915332e-06, + "loss": 0.0006, + "step": 1220 + }, + { + "epoch": 0.32421667551779076, + "grad_norm": 0.08856791563455327, + "learning_rate": 9.661380794338252e-06, + "loss": 0.002, + "step": 1221 + }, + { + "epoch": 0.3244822092405735, + "grad_norm": 0.04449604335631291, + "learning_rate": 9.660586171101931e-06, + "loss": 0.0008, + "step": 1222 + }, + { + "epoch": 0.32474774296335635, + "grad_norm": 0.8025875942279423, + "learning_rate": 9.659790649359563e-06, + "loss": 0.1738, + "step": 1223 + }, + { + "epoch": 0.32501327668613916, + "grad_norm": 0.13222478587189693, + "learning_rate": 9.658994229264514e-06, + "loss": 0.0131, + "step": 1224 + }, + { + "epoch": 0.3252788104089219, + "grad_norm": 0.6260229647329109, + "learning_rate": 9.658196910970327e-06, + "loss": 0.0889, + "step": 1225 + }, + { + "epoch": 0.32554434413170474, + "grad_norm": 0.7658967348933747, + "learning_rate": 9.657398694630713e-06, + "loss": 0.1553, + "step": 1226 + }, + { + "epoch": 0.3258098778544875, + "grad_norm": 0.48947788594768415, + "learning_rate": 9.65659958039956e-06, + "loss": 0.0288, + "step": 1227 + }, + { + "epoch": 0.3260754115772703, + "grad_norm": 0.8585792462640025, + "learning_rate": 9.655799568430926e-06, + "loss": 0.0688, + "step": 1228 + }, + { + "epoch": 0.3263409453000531, + "grad_norm": 0.6527231816101399, + "learning_rate": 9.654998658879046e-06, + "loss": 0.1113, + "step": 1229 + }, + { + "epoch": 0.3266064790228359, + "grad_norm": 0.6582417207082769, + "learning_rate": 9.654196851898325e-06, + "loss": 0.1191, + "step": 1230 + }, + { + "epoch": 0.32687201274561867, + "grad_norm": 0.6645575253832666, + "learning_rate": 9.653394147643343e-06, + "loss": 0.1113, + "step": 1231 + }, + { + "epoch": 0.3271375464684015, + "grad_norm": 0.27026297493457185, + "learning_rate": 9.65259054626885e-06, + "loss": 0.0317, + "step": 1232 + }, + { + "epoch": 0.3274030801911843, + "grad_norm": 0.6383288218728779, + "learning_rate": 9.651786047929772e-06, + "loss": 0.1035, + "step": 1233 + }, + { + "epoch": 0.32766861391396707, + "grad_norm": 0.011355216588088562, + "learning_rate": 9.650980652781208e-06, + "loss": 0.0003, + "step": 1234 + }, + { + "epoch": 0.3279341476367499, + "grad_norm": 0.5036691216178738, + "learning_rate": 9.650174360978427e-06, + "loss": 0.0525, + "step": 1235 + }, + { + "epoch": 0.32819968135953265, + "grad_norm": 0.48956110736136377, + "learning_rate": 9.649367172676873e-06, + "loss": 0.0317, + "step": 1236 + }, + { + "epoch": 0.32846521508231546, + "grad_norm": 0.6193502355457858, + "learning_rate": 9.648559088032162e-06, + "loss": 0.0957, + "step": 1237 + }, + { + "epoch": 0.3287307488050982, + "grad_norm": 0.7256539061385159, + "learning_rate": 9.647750107200082e-06, + "loss": 0.0889, + "step": 1238 + }, + { + "epoch": 0.32899628252788105, + "grad_norm": 0.5224109219073716, + "learning_rate": 9.646940230336599e-06, + "loss": 0.0045, + "step": 1239 + }, + { + "epoch": 0.3292618162506638, + "grad_norm": 0.2037818062231921, + "learning_rate": 9.646129457597844e-06, + "loss": 0.0023, + "step": 1240 + }, + { + "epoch": 0.3295273499734466, + "grad_norm": 0.4607652592533831, + "learning_rate": 9.645317789140125e-06, + "loss": 0.0579, + "step": 1241 + }, + { + "epoch": 0.32979288369622944, + "grad_norm": 0.9063414384859513, + "learning_rate": 9.644505225119922e-06, + "loss": 0.1113, + "step": 1242 + }, + { + "epoch": 0.3300584174190122, + "grad_norm": 0.2770486401552903, + "learning_rate": 9.64369176569389e-06, + "loss": 0.0187, + "step": 1243 + }, + { + "epoch": 0.330323951141795, + "grad_norm": 0.4514415661962752, + "learning_rate": 9.642877411018852e-06, + "loss": 0.0525, + "step": 1244 + }, + { + "epoch": 0.3305894848645778, + "grad_norm": 0.4388178911161381, + "learning_rate": 9.642062161251807e-06, + "loss": 0.0525, + "step": 1245 + }, + { + "epoch": 0.3308550185873606, + "grad_norm": 0.5887566574525436, + "learning_rate": 9.641246016549925e-06, + "loss": 0.0579, + "step": 1246 + }, + { + "epoch": 0.33112055231014337, + "grad_norm": 2.5823916037440813, + "learning_rate": 9.64042897707055e-06, + "loss": 0.2344, + "step": 1247 + }, + { + "epoch": 0.3313860860329262, + "grad_norm": 0.3156225772793237, + "learning_rate": 9.639611042971198e-06, + "loss": 0.0258, + "step": 1248 + }, + { + "epoch": 0.33165161975570895, + "grad_norm": 0.4480395609995155, + "learning_rate": 9.638792214409556e-06, + "loss": 0.0479, + "step": 1249 + }, + { + "epoch": 0.33191715347849177, + "grad_norm": 0.7209237690266596, + "learning_rate": 9.637972491543482e-06, + "loss": 0.0957, + "step": 1250 + }, + { + "epoch": 0.3321826872012746, + "grad_norm": 0.790962238735122, + "learning_rate": 9.637151874531014e-06, + "loss": 0.1035, + "step": 1251 + }, + { + "epoch": 0.33244822092405735, + "grad_norm": 0.26847764212042974, + "learning_rate": 9.636330363530352e-06, + "loss": 0.0206, + "step": 1252 + }, + { + "epoch": 0.33271375464684017, + "grad_norm": 0.13095952869121744, + "learning_rate": 9.63550795869988e-06, + "loss": 0.0042, + "step": 1253 + }, + { + "epoch": 0.33297928836962293, + "grad_norm": 0.4485281612188027, + "learning_rate": 9.63468466019814e-06, + "loss": 0.0479, + "step": 1254 + }, + { + "epoch": 0.33324482209240575, + "grad_norm": 6.678236601224771, + "learning_rate": 9.633860468183862e-06, + "loss": 0.6133, + "step": 1255 + }, + { + "epoch": 0.3335103558151885, + "grad_norm": 0.03640018770094486, + "learning_rate": 9.633035382815936e-06, + "loss": 0.0009, + "step": 1256 + }, + { + "epoch": 0.3337758895379713, + "grad_norm": 0.20995200600370745, + "learning_rate": 9.63220940425343e-06, + "loss": 0.0147, + "step": 1257 + }, + { + "epoch": 0.3340414232607541, + "grad_norm": 0.7744165649910679, + "learning_rate": 9.631382532655583e-06, + "loss": 0.1035, + "step": 1258 + }, + { + "epoch": 0.3343069569835369, + "grad_norm": 0.8521274871394714, + "learning_rate": 9.630554768181806e-06, + "loss": 0.1279, + "step": 1259 + }, + { + "epoch": 0.3345724907063197, + "grad_norm": 0.013869617394598415, + "learning_rate": 9.62972611099168e-06, + "loss": 0.0004, + "step": 1260 + }, + { + "epoch": 0.3348380244291025, + "grad_norm": 0.7984701653609472, + "learning_rate": 9.628896561244965e-06, + "loss": 0.1201, + "step": 1261 + }, + { + "epoch": 0.3351035581518853, + "grad_norm": 0.18442932803572457, + "learning_rate": 9.628066119101583e-06, + "loss": 0.0093, + "step": 1262 + }, + { + "epoch": 0.33536909187466807, + "grad_norm": 0.7728685034480455, + "learning_rate": 9.627234784721637e-06, + "loss": 0.1113, + "step": 1263 + }, + { + "epoch": 0.3356346255974509, + "grad_norm": 0.28335521981673617, + "learning_rate": 9.6264025582654e-06, + "loss": 0.0233, + "step": 1264 + }, + { + "epoch": 0.33590015932023365, + "grad_norm": 0.399075752307739, + "learning_rate": 9.62556943989331e-06, + "loss": 0.0479, + "step": 1265 + }, + { + "epoch": 0.33616569304301647, + "grad_norm": 0.12658474094636998, + "learning_rate": 9.624735429765988e-06, + "loss": 0.0041, + "step": 1266 + }, + { + "epoch": 0.33643122676579923, + "grad_norm": 0.5280828189652471, + "learning_rate": 9.623900528044216e-06, + "loss": 0.0752, + "step": 1267 + }, + { + "epoch": 0.33669676048858205, + "grad_norm": 0.24243489448797254, + "learning_rate": 9.623064734888959e-06, + "loss": 0.0066, + "step": 1268 + }, + { + "epoch": 0.33696229421136487, + "grad_norm": 0.4700632898354242, + "learning_rate": 9.622228050461345e-06, + "loss": 0.0693, + "step": 1269 + }, + { + "epoch": 0.33722782793414763, + "grad_norm": 0.41440456235100415, + "learning_rate": 9.621390474922675e-06, + "loss": 0.0579, + "step": 1270 + }, + { + "epoch": 0.33749336165693045, + "grad_norm": 0.7223449992115132, + "learning_rate": 9.620552008434426e-06, + "loss": 0.0752, + "step": 1271 + }, + { + "epoch": 0.3377588953797132, + "grad_norm": 0.4596889344479549, + "learning_rate": 9.619712651158243e-06, + "loss": 0.0693, + "step": 1272 + }, + { + "epoch": 0.338024429102496, + "grad_norm": 0.4441012257660843, + "learning_rate": 9.618872403255947e-06, + "loss": 0.063, + "step": 1273 + }, + { + "epoch": 0.3382899628252788, + "grad_norm": 0.1553935211134242, + "learning_rate": 9.618031264889529e-06, + "loss": 0.0066, + "step": 1274 + }, + { + "epoch": 0.3385554965480616, + "grad_norm": 0.42387440934012344, + "learning_rate": 9.617189236221143e-06, + "loss": 0.0579, + "step": 1275 + }, + { + "epoch": 0.33882103027084437, + "grad_norm": 0.47243312888409494, + "learning_rate": 9.61634631741313e-06, + "loss": 0.0752, + "step": 1276 + }, + { + "epoch": 0.3390865639936272, + "grad_norm": 0.06774040615620845, + "learning_rate": 9.61550250862799e-06, + "loss": 0.0018, + "step": 1277 + }, + { + "epoch": 0.33935209771641, + "grad_norm": 0.38367232223776365, + "learning_rate": 9.614657810028402e-06, + "loss": 0.0525, + "step": 1278 + }, + { + "epoch": 0.33961763143919277, + "grad_norm": 0.3392200068088174, + "learning_rate": 9.613812221777212e-06, + "loss": 0.0131, + "step": 1279 + }, + { + "epoch": 0.3398831651619756, + "grad_norm": 0.42002324638804434, + "learning_rate": 9.612965744037441e-06, + "loss": 0.0479, + "step": 1280 + }, + { + "epoch": 0.34014869888475835, + "grad_norm": 0.5274772223331019, + "learning_rate": 9.61211837697228e-06, + "loss": 0.0693, + "step": 1281 + }, + { + "epoch": 0.34041423260754117, + "grad_norm": 0.35860555791803894, + "learning_rate": 9.61127012074509e-06, + "loss": 0.0479, + "step": 1282 + }, + { + "epoch": 0.34067976633032393, + "grad_norm": 0.5038903692242176, + "learning_rate": 9.610420975519407e-06, + "loss": 0.0752, + "step": 1283 + }, + { + "epoch": 0.34094530005310675, + "grad_norm": 0.4033059759112738, + "learning_rate": 9.609570941458935e-06, + "loss": 0.0579, + "step": 1284 + }, + { + "epoch": 0.3412108337758895, + "grad_norm": 0.5660112789357162, + "learning_rate": 9.60872001872755e-06, + "loss": 0.0752, + "step": 1285 + }, + { + "epoch": 0.34147636749867233, + "grad_norm": 0.3263429089982916, + "learning_rate": 9.607868207489297e-06, + "loss": 0.0393, + "step": 1286 + }, + { + "epoch": 0.34174190122145515, + "grad_norm": 0.37153713665523197, + "learning_rate": 9.607015507908401e-06, + "loss": 0.0479, + "step": 1287 + }, + { + "epoch": 0.3420074349442379, + "grad_norm": 0.4755797521549427, + "learning_rate": 9.606161920149251e-06, + "loss": 0.0579, + "step": 1288 + }, + { + "epoch": 0.3422729686670207, + "grad_norm": 0.5767413183549783, + "learning_rate": 9.605307444376407e-06, + "loss": 0.063, + "step": 1289 + }, + { + "epoch": 0.3425385023898035, + "grad_norm": 0.06779770472600782, + "learning_rate": 9.604452080754601e-06, + "loss": 0.0017, + "step": 1290 + }, + { + "epoch": 0.3428040361125863, + "grad_norm": 0.5248146878342517, + "learning_rate": 9.603595829448737e-06, + "loss": 0.0815, + "step": 1291 + }, + { + "epoch": 0.34306956983536907, + "grad_norm": 3.5626821708062617, + "learning_rate": 9.602738690623895e-06, + "loss": 0.377, + "step": 1292 + }, + { + "epoch": 0.3433351035581519, + "grad_norm": 0.5707292470276522, + "learning_rate": 9.601880664445313e-06, + "loss": 0.0957, + "step": 1293 + }, + { + "epoch": 0.34360063728093465, + "grad_norm": 0.41750308884117965, + "learning_rate": 9.601021751078416e-06, + "loss": 0.0393, + "step": 1294 + }, + { + "epoch": 0.34386617100371747, + "grad_norm": 0.5159271503322392, + "learning_rate": 9.60016195068879e-06, + "loss": 0.063, + "step": 1295 + }, + { + "epoch": 0.3441317047265003, + "grad_norm": 0.5216179998375932, + "learning_rate": 9.599301263442194e-06, + "loss": 0.0688, + "step": 1296 + }, + { + "epoch": 0.34439723844928305, + "grad_norm": 0.30090609879974806, + "learning_rate": 9.598439689504556e-06, + "loss": 0.0391, + "step": 1297 + }, + { + "epoch": 0.34466277217206587, + "grad_norm": 0.3707745206401245, + "learning_rate": 9.597577229041981e-06, + "loss": 0.0525, + "step": 1298 + }, + { + "epoch": 0.34492830589484863, + "grad_norm": 2.103947714882032, + "learning_rate": 9.596713882220739e-06, + "loss": 0.1035, + "step": 1299 + }, + { + "epoch": 0.34519383961763145, + "grad_norm": 0.5141149862836718, + "learning_rate": 9.595849649207274e-06, + "loss": 0.0693, + "step": 1300 + }, + { + "epoch": 0.3454593733404142, + "grad_norm": 0.44396106651225886, + "learning_rate": 9.594984530168198e-06, + "loss": 0.0525, + "step": 1301 + }, + { + "epoch": 0.34572490706319703, + "grad_norm": 0.3455247284911814, + "learning_rate": 9.594118525270299e-06, + "loss": 0.0479, + "step": 1302 + }, + { + "epoch": 0.34599044078597985, + "grad_norm": 0.7585784664106814, + "learning_rate": 9.593251634680531e-06, + "loss": 0.0889, + "step": 1303 + }, + { + "epoch": 0.3462559745087626, + "grad_norm": 0.36123832598406336, + "learning_rate": 9.59238385856602e-06, + "loss": 0.0479, + "step": 1304 + }, + { + "epoch": 0.3465215082315454, + "grad_norm": 0.5097193438244858, + "learning_rate": 9.591515197094064e-06, + "loss": 0.082, + "step": 1305 + }, + { + "epoch": 0.3467870419543282, + "grad_norm": 0.2653818151867503, + "learning_rate": 9.59064565043213e-06, + "loss": 0.0354, + "step": 1306 + }, + { + "epoch": 0.347052575677111, + "grad_norm": 0.5864846744145519, + "learning_rate": 9.589775218747854e-06, + "loss": 0.0889, + "step": 1307 + }, + { + "epoch": 0.34731810939989377, + "grad_norm": 0.3781614336308014, + "learning_rate": 9.58890390220905e-06, + "loss": 0.0131, + "step": 1308 + }, + { + "epoch": 0.3475836431226766, + "grad_norm": 0.30092084572433875, + "learning_rate": 9.588031700983693e-06, + "loss": 0.0393, + "step": 1309 + }, + { + "epoch": 0.34784917684545935, + "grad_norm": 0.35904973427780035, + "learning_rate": 9.587158615239935e-06, + "loss": 0.0354, + "step": 1310 + }, + { + "epoch": 0.34811471056824217, + "grad_norm": 0.5600811056316272, + "learning_rate": 9.586284645146096e-06, + "loss": 0.0579, + "step": 1311 + }, + { + "epoch": 0.348380244291025, + "grad_norm": 0.05426252506203742, + "learning_rate": 9.58540979087067e-06, + "loss": 0.0015, + "step": 1312 + }, + { + "epoch": 0.34864577801380775, + "grad_norm": 0.5870410341108383, + "learning_rate": 9.584534052582314e-06, + "loss": 0.1035, + "step": 1313 + }, + { + "epoch": 0.34891131173659057, + "grad_norm": 0.6294228311893199, + "learning_rate": 9.583657430449862e-06, + "loss": 0.082, + "step": 1314 + }, + { + "epoch": 0.34917684545937333, + "grad_norm": 0.4895276277368698, + "learning_rate": 9.582779924642317e-06, + "loss": 0.0184, + "step": 1315 + }, + { + "epoch": 0.34944237918215615, + "grad_norm": 0.5733825038864722, + "learning_rate": 9.581901535328853e-06, + "loss": 0.082, + "step": 1316 + }, + { + "epoch": 0.3497079129049389, + "grad_norm": 0.32782198798301165, + "learning_rate": 9.58102226267881e-06, + "loss": 0.0146, + "step": 1317 + }, + { + "epoch": 0.34997344662772173, + "grad_norm": 0.590823967005489, + "learning_rate": 9.580142106861704e-06, + "loss": 0.0693, + "step": 1318 + }, + { + "epoch": 0.3502389803505045, + "grad_norm": 0.28663624606998783, + "learning_rate": 9.579261068047217e-06, + "loss": 0.0147, + "step": 1319 + }, + { + "epoch": 0.3505045140732873, + "grad_norm": 0.07002167741569795, + "learning_rate": 9.578379146405202e-06, + "loss": 0.002, + "step": 1320 + }, + { + "epoch": 0.35077004779607013, + "grad_norm": 0.3823299154936616, + "learning_rate": 9.577496342105686e-06, + "loss": 0.0476, + "step": 1321 + }, + { + "epoch": 0.3510355815188529, + "grad_norm": 0.41555476819705756, + "learning_rate": 9.576612655318863e-06, + "loss": 0.0184, + "step": 1322 + }, + { + "epoch": 0.3513011152416357, + "grad_norm": 0.6565947270509365, + "learning_rate": 9.575728086215093e-06, + "loss": 0.0479, + "step": 1323 + }, + { + "epoch": 0.35156664896441847, + "grad_norm": 0.28826966637792456, + "learning_rate": 9.574842634964917e-06, + "loss": 0.0354, + "step": 1324 + }, + { + "epoch": 0.3518321826872013, + "grad_norm": 0.47440584122085944, + "learning_rate": 9.573956301739033e-06, + "loss": 0.0576, + "step": 1325 + }, + { + "epoch": 0.35209771640998405, + "grad_norm": 0.013825970291321273, + "learning_rate": 9.573069086708322e-06, + "loss": 0.0003, + "step": 1326 + }, + { + "epoch": 0.35236325013276687, + "grad_norm": 0.308417343496391, + "learning_rate": 9.572180990043822e-06, + "loss": 0.0393, + "step": 1327 + }, + { + "epoch": 0.35262878385554963, + "grad_norm": 0.023751366893711805, + "learning_rate": 9.571292011916753e-06, + "loss": 0.0006, + "step": 1328 + }, + { + "epoch": 0.35289431757833245, + "grad_norm": 0.6910550885091158, + "learning_rate": 9.570402152498498e-06, + "loss": 0.0688, + "step": 1329 + }, + { + "epoch": 0.35315985130111527, + "grad_norm": 0.5584726753157215, + "learning_rate": 9.56951141196061e-06, + "loss": 0.063, + "step": 1330 + }, + { + "epoch": 0.35342538502389803, + "grad_norm": 0.3771730171037822, + "learning_rate": 9.568619790474815e-06, + "loss": 0.0432, + "step": 1331 + }, + { + "epoch": 0.35369091874668085, + "grad_norm": 0.006598573947351961, + "learning_rate": 9.567727288213005e-06, + "loss": 0.0001, + "step": 1332 + }, + { + "epoch": 0.3539564524694636, + "grad_norm": 0.24569141953841403, + "learning_rate": 9.566833905347245e-06, + "loss": 0.0286, + "step": 1333 + }, + { + "epoch": 0.35422198619224643, + "grad_norm": 0.31994068602155107, + "learning_rate": 9.56593964204977e-06, + "loss": 0.0354, + "step": 1334 + }, + { + "epoch": 0.3544875199150292, + "grad_norm": 0.48426314470272797, + "learning_rate": 9.565044498492984e-06, + "loss": 0.0476, + "step": 1335 + }, + { + "epoch": 0.354753053637812, + "grad_norm": 0.4284293731700466, + "learning_rate": 9.564148474849455e-06, + "loss": 0.0476, + "step": 1336 + }, + { + "epoch": 0.3550185873605948, + "grad_norm": 0.5223639882465799, + "learning_rate": 9.563251571291932e-06, + "loss": 0.0432, + "step": 1337 + }, + { + "epoch": 0.3552841210833776, + "grad_norm": 0.3682885994989024, + "learning_rate": 9.562353787993321e-06, + "loss": 0.0432, + "step": 1338 + }, + { + "epoch": 0.3555496548061604, + "grad_norm": 0.7276856913645094, + "learning_rate": 9.56145512512671e-06, + "loss": 0.0957, + "step": 1339 + }, + { + "epoch": 0.35581518852894317, + "grad_norm": 0.28021489029743807, + "learning_rate": 9.56055558286535e-06, + "loss": 0.0317, + "step": 1340 + }, + { + "epoch": 0.356080722251726, + "grad_norm": 0.01832087245829523, + "learning_rate": 9.559655161382658e-06, + "loss": 0.0004, + "step": 1341 + }, + { + "epoch": 0.35634625597450875, + "grad_norm": 0.914955086500517, + "learning_rate": 9.558753860852228e-06, + "loss": 0.0752, + "step": 1342 + }, + { + "epoch": 0.35661178969729157, + "grad_norm": 0.20459946602158102, + "learning_rate": 9.557851681447818e-06, + "loss": 0.0206, + "step": 1343 + }, + { + "epoch": 0.35687732342007433, + "grad_norm": 0.27576350826669555, + "learning_rate": 9.556948623343358e-06, + "loss": 0.0286, + "step": 1344 + }, + { + "epoch": 0.35714285714285715, + "grad_norm": 0.2761915334889806, + "learning_rate": 9.55604468671295e-06, + "loss": 0.0255, + "step": 1345 + }, + { + "epoch": 0.3574083908656399, + "grad_norm": 0.8233226661583315, + "learning_rate": 9.555139871730859e-06, + "loss": 0.0354, + "step": 1346 + }, + { + "epoch": 0.35767392458842273, + "grad_norm": 0.1926615669215923, + "learning_rate": 9.554234178571524e-06, + "loss": 0.0165, + "step": 1347 + }, + { + "epoch": 0.35793945831120555, + "grad_norm": 0.3098753295444041, + "learning_rate": 9.55332760740955e-06, + "loss": 0.0286, + "step": 1348 + }, + { + "epoch": 0.3582049920339883, + "grad_norm": 3.431526479191698, + "learning_rate": 9.552420158419715e-06, + "loss": 0.1191, + "step": 1349 + }, + { + "epoch": 0.35847052575677113, + "grad_norm": 0.0625932742877244, + "learning_rate": 9.551511831776966e-06, + "loss": 0.001, + "step": 1350 + }, + { + "epoch": 0.3587360594795539, + "grad_norm": 1.1478320124700105, + "learning_rate": 9.550602627656415e-06, + "loss": 0.0889, + "step": 1351 + }, + { + "epoch": 0.3590015932023367, + "grad_norm": 1.003238754359967, + "learning_rate": 9.549692546233348e-06, + "loss": 0.0688, + "step": 1352 + }, + { + "epoch": 0.3592671269251195, + "grad_norm": 0.25770729335411, + "learning_rate": 9.548781587683214e-06, + "loss": 0.0061, + "step": 1353 + }, + { + "epoch": 0.3595326606479023, + "grad_norm": 1.55958556462164, + "learning_rate": 9.54786975218164e-06, + "loss": 0.0688, + "step": 1354 + }, + { + "epoch": 0.35979819437068505, + "grad_norm": 1.2409731636678565, + "learning_rate": 9.546957039904414e-06, + "loss": 0.0688, + "step": 1355 + }, + { + "epoch": 0.36006372809346787, + "grad_norm": 0.032263196311481473, + "learning_rate": 9.546043451027498e-06, + "loss": 0.0007, + "step": 1356 + }, + { + "epoch": 0.3603292618162507, + "grad_norm": 0.4590699936620603, + "learning_rate": 9.54512898572702e-06, + "loss": 0.0205, + "step": 1357 + }, + { + "epoch": 0.36059479553903345, + "grad_norm": 6.157784939523085, + "learning_rate": 9.544213644179278e-06, + "loss": 0.4668, + "step": 1358 + }, + { + "epoch": 0.36086032926181627, + "grad_norm": 1.0445567828018743, + "learning_rate": 9.54329742656074e-06, + "loss": 0.1035, + "step": 1359 + }, + { + "epoch": 0.36112586298459903, + "grad_norm": 0.900975787190914, + "learning_rate": 9.54238033304804e-06, + "loss": 0.0525, + "step": 1360 + }, + { + "epoch": 0.36139139670738185, + "grad_norm": 2.1445461499571827, + "learning_rate": 9.541462363817986e-06, + "loss": 0.1279, + "step": 1361 + }, + { + "epoch": 0.3616569304301646, + "grad_norm": 2.6058708901982754, + "learning_rate": 9.540543519047547e-06, + "loss": 0.0889, + "step": 1362 + }, + { + "epoch": 0.36192246415294743, + "grad_norm": 0.30837506165256673, + "learning_rate": 9.53962379891387e-06, + "loss": 0.0255, + "step": 1363 + }, + { + "epoch": 0.3621879978757302, + "grad_norm": 0.6781227526654185, + "learning_rate": 9.538703203594266e-06, + "loss": 0.0476, + "step": 1364 + }, + { + "epoch": 0.362453531598513, + "grad_norm": 0.0014988796322645392, + "learning_rate": 9.53778173326621e-06, + "loss": 0.0, + "step": 1365 + }, + { + "epoch": 0.36271906532129583, + "grad_norm": 0.4848181673027792, + "learning_rate": 9.536859388107359e-06, + "loss": 0.0354, + "step": 1366 + }, + { + "epoch": 0.3629845990440786, + "grad_norm": 0.30798774656542016, + "learning_rate": 9.535936168295521e-06, + "loss": 0.0255, + "step": 1367 + }, + { + "epoch": 0.3632501327668614, + "grad_norm": 1.1103162493296537, + "learning_rate": 9.535012074008688e-06, + "loss": 0.1191, + "step": 1368 + }, + { + "epoch": 0.3635156664896442, + "grad_norm": 1.115968824289187, + "learning_rate": 9.53408710542501e-06, + "loss": 0.1934, + "step": 1369 + }, + { + "epoch": 0.363781200212427, + "grad_norm": 0.1558534145654595, + "learning_rate": 9.533161262722815e-06, + "loss": 0.0131, + "step": 1370 + }, + { + "epoch": 0.36404673393520975, + "grad_norm": 0.042070066792334465, + "learning_rate": 9.532234546080592e-06, + "loss": 0.0008, + "step": 1371 + }, + { + "epoch": 0.3643122676579926, + "grad_norm": 0.05020993746733037, + "learning_rate": 9.531306955677003e-06, + "loss": 0.0012, + "step": 1372 + }, + { + "epoch": 0.36457780138077533, + "grad_norm": 0.3478946473030989, + "learning_rate": 9.530378491690872e-06, + "loss": 0.0255, + "step": 1373 + }, + { + "epoch": 0.36484333510355815, + "grad_norm": 0.532671213666101, + "learning_rate": 9.5294491543012e-06, + "loss": 0.0286, + "step": 1374 + }, + { + "epoch": 0.36510886882634097, + "grad_norm": 0.0104645404151327, + "learning_rate": 9.52851894368715e-06, + "loss": 0.0002, + "step": 1375 + }, + { + "epoch": 0.36537440254912373, + "grad_norm": 0.6407476894876308, + "learning_rate": 9.527587860028055e-06, + "loss": 0.0165, + "step": 1376 + }, + { + "epoch": 0.36563993627190655, + "grad_norm": 1.683239750297336, + "learning_rate": 9.526655903503423e-06, + "loss": 0.0688, + "step": 1377 + }, + { + "epoch": 0.3659054699946893, + "grad_norm": 0.1956603056666467, + "learning_rate": 9.525723074292916e-06, + "loss": 0.0184, + "step": 1378 + }, + { + "epoch": 0.36617100371747213, + "grad_norm": 0.6108332993871155, + "learning_rate": 9.524789372576377e-06, + "loss": 0.0317, + "step": 1379 + }, + { + "epoch": 0.3664365374402549, + "grad_norm": 0.03952181334754172, + "learning_rate": 9.523854798533814e-06, + "loss": 0.0012, + "step": 1380 + }, + { + "epoch": 0.3667020711630377, + "grad_norm": 0.18831659774689863, + "learning_rate": 9.522919352345398e-06, + "loss": 0.0024, + "step": 1381 + }, + { + "epoch": 0.3669676048858205, + "grad_norm": 0.8786083910112925, + "learning_rate": 9.521983034191472e-06, + "loss": 0.0432, + "step": 1382 + }, + { + "epoch": 0.3672331386086033, + "grad_norm": 0.1730242308325762, + "learning_rate": 9.521045844252552e-06, + "loss": 0.0146, + "step": 1383 + }, + { + "epoch": 0.3674986723313861, + "grad_norm": 1.2876116546269887, + "learning_rate": 9.520107782709314e-06, + "loss": 0.0815, + "step": 1384 + }, + { + "epoch": 0.3677642060541689, + "grad_norm": 1.3082256972483228, + "learning_rate": 9.519168849742603e-06, + "loss": 0.0752, + "step": 1385 + }, + { + "epoch": 0.3680297397769517, + "grad_norm": 0.9335061419191978, + "learning_rate": 9.518229045533438e-06, + "loss": 0.0576, + "step": 1386 + }, + { + "epoch": 0.36829527349973445, + "grad_norm": 2.446950303059504, + "learning_rate": 9.517288370263e-06, + "loss": 0.2246, + "step": 1387 + }, + { + "epoch": 0.3685608072225173, + "grad_norm": 0.6333100630168161, + "learning_rate": 9.51634682411264e-06, + "loss": 0.0123, + "step": 1388 + }, + { + "epoch": 0.36882634094530004, + "grad_norm": 1.1503551358840165, + "learning_rate": 9.51540440726388e-06, + "loss": 0.1035, + "step": 1389 + }, + { + "epoch": 0.36909187466808285, + "grad_norm": 0.3449715903261567, + "learning_rate": 9.514461119898401e-06, + "loss": 0.0092, + "step": 1390 + }, + { + "epoch": 0.3693574083908656, + "grad_norm": 0.23906913907246705, + "learning_rate": 9.513516962198064e-06, + "loss": 0.0229, + "step": 1391 + }, + { + "epoch": 0.36962294211364843, + "grad_norm": 0.4848006955734177, + "learning_rate": 9.512571934344887e-06, + "loss": 0.0117, + "step": 1392 + }, + { + "epoch": 0.36988847583643125, + "grad_norm": 0.9490909420873752, + "learning_rate": 9.511626036521064e-06, + "loss": 0.0889, + "step": 1393 + }, + { + "epoch": 0.370154009559214, + "grad_norm": 3.92147853906124, + "learning_rate": 9.510679268908949e-06, + "loss": 0.1934, + "step": 1394 + }, + { + "epoch": 0.37041954328199683, + "grad_norm": 2.8735279348285796, + "learning_rate": 9.509731631691071e-06, + "loss": 0.2129, + "step": 1395 + }, + { + "epoch": 0.3706850770047796, + "grad_norm": 0.2436877641705072, + "learning_rate": 9.50878312505012e-06, + "loss": 0.0205, + "step": 1396 + }, + { + "epoch": 0.3709506107275624, + "grad_norm": 1.3855718841062967, + "learning_rate": 9.50783374916896e-06, + "loss": 0.063, + "step": 1397 + }, + { + "epoch": 0.3712161444503452, + "grad_norm": 0.39133070889265453, + "learning_rate": 9.506883504230618e-06, + "loss": 0.0354, + "step": 1398 + }, + { + "epoch": 0.371481678173128, + "grad_norm": 0.6442273431688003, + "learning_rate": 9.50593239041829e-06, + "loss": 0.0476, + "step": 1399 + }, + { + "epoch": 0.37174721189591076, + "grad_norm": 0.3654458756752601, + "learning_rate": 9.504980407915338e-06, + "loss": 0.0354, + "step": 1400 + }, + { + "epoch": 0.3720127456186936, + "grad_norm": 0.8586712209152337, + "learning_rate": 9.504027556905297e-06, + "loss": 0.1279, + "step": 1401 + }, + { + "epoch": 0.3722782793414764, + "grad_norm": 0.6773527175343205, + "learning_rate": 9.503073837571862e-06, + "loss": 0.0576, + "step": 1402 + }, + { + "epoch": 0.37254381306425915, + "grad_norm": 0.4186954158863423, + "learning_rate": 9.5021192500989e-06, + "loss": 0.0352, + "step": 1403 + }, + { + "epoch": 0.372809346787042, + "grad_norm": 1.9132653529788708, + "learning_rate": 9.501163794670445e-06, + "loss": 0.1738, + "step": 1404 + }, + { + "epoch": 0.37307488050982474, + "grad_norm": 2.2952198425920582, + "learning_rate": 9.500207471470695e-06, + "loss": 0.0815, + "step": 1405 + }, + { + "epoch": 0.37334041423260755, + "grad_norm": 0.40586858759644967, + "learning_rate": 9.49925028068402e-06, + "loss": 0.0165, + "step": 1406 + }, + { + "epoch": 0.3736059479553903, + "grad_norm": 0.18480641707554119, + "learning_rate": 9.498292222494955e-06, + "loss": 0.0035, + "step": 1407 + }, + { + "epoch": 0.37387148167817313, + "grad_norm": 0.7772486998466412, + "learning_rate": 9.497333297088202e-06, + "loss": 0.1191, + "step": 1408 + }, + { + "epoch": 0.3741370154009559, + "grad_norm": 0.2965999353034397, + "learning_rate": 9.496373504648632e-06, + "loss": 0.0286, + "step": 1409 + }, + { + "epoch": 0.3744025491237387, + "grad_norm": 2.294123343793471, + "learning_rate": 9.495412845361279e-06, + "loss": 0.1934, + "step": 1410 + }, + { + "epoch": 0.37466808284652153, + "grad_norm": 0.0026701934360799534, + "learning_rate": 9.494451319411347e-06, + "loss": 0.0001, + "step": 1411 + }, + { + "epoch": 0.3749336165693043, + "grad_norm": 1.4911265196857602, + "learning_rate": 9.493488926984208e-06, + "loss": 0.063, + "step": 1412 + }, + { + "epoch": 0.3751991502920871, + "grad_norm": 0.9596610809296374, + "learning_rate": 9.4925256682654e-06, + "loss": 0.0752, + "step": 1413 + }, + { + "epoch": 0.3754646840148699, + "grad_norm": 0.8806164654939721, + "learning_rate": 9.491561543440627e-06, + "loss": 0.0391, + "step": 1414 + }, + { + "epoch": 0.3757302177376527, + "grad_norm": 0.7690458874265063, + "learning_rate": 9.490596552695762e-06, + "loss": 0.1191, + "step": 1415 + }, + { + "epoch": 0.37599575146043546, + "grad_norm": 0.6872344076762994, + "learning_rate": 9.489630696216842e-06, + "loss": 0.0576, + "step": 1416 + }, + { + "epoch": 0.3762612851832183, + "grad_norm": 0.48727643854025365, + "learning_rate": 9.488663974190077e-06, + "loss": 0.0354, + "step": 1417 + }, + { + "epoch": 0.37652681890600104, + "grad_norm": 0.7530220920460418, + "learning_rate": 9.487696386801835e-06, + "loss": 0.0752, + "step": 1418 + }, + { + "epoch": 0.37679235262878386, + "grad_norm": 0.29673241857480187, + "learning_rate": 9.486727934238656e-06, + "loss": 0.0079, + "step": 1419 + }, + { + "epoch": 0.3770578863515667, + "grad_norm": 0.3529062224624686, + "learning_rate": 9.485758616687248e-06, + "loss": 0.0093, + "step": 1420 + }, + { + "epoch": 0.37732342007434944, + "grad_norm": 0.09031390787732817, + "learning_rate": 9.484788434334484e-06, + "loss": 0.0028, + "step": 1421 + }, + { + "epoch": 0.37758895379713225, + "grad_norm": 0.6189002563525746, + "learning_rate": 9.483817387367403e-06, + "loss": 0.0889, + "step": 1422 + }, + { + "epoch": 0.377854487519915, + "grad_norm": 0.8659301327268721, + "learning_rate": 9.482845475973212e-06, + "loss": 0.1113, + "step": 1423 + }, + { + "epoch": 0.37812002124269783, + "grad_norm": 0.025135957238805916, + "learning_rate": 9.481872700339282e-06, + "loss": 0.0007, + "step": 1424 + }, + { + "epoch": 0.3783855549654806, + "grad_norm": 0.20693260579313302, + "learning_rate": 9.480899060653154e-06, + "loss": 0.0105, + "step": 1425 + }, + { + "epoch": 0.3786510886882634, + "grad_norm": 3.127258712138675, + "learning_rate": 9.479924557102536e-06, + "loss": 0.2344, + "step": 1426 + }, + { + "epoch": 0.3789166224110462, + "grad_norm": 0.20096836647101268, + "learning_rate": 9.478949189875299e-06, + "loss": 0.004, + "step": 1427 + }, + { + "epoch": 0.379182156133829, + "grad_norm": 1.2506042183343944, + "learning_rate": 9.47797295915948e-06, + "loss": 0.1113, + "step": 1428 + }, + { + "epoch": 0.3794476898566118, + "grad_norm": 0.6735848636167625, + "learning_rate": 9.47699586514329e-06, + "loss": 0.0688, + "step": 1429 + }, + { + "epoch": 0.3797132235793946, + "grad_norm": 0.8729493010310873, + "learning_rate": 9.476017908015098e-06, + "loss": 0.0815, + "step": 1430 + }, + { + "epoch": 0.3799787573021774, + "grad_norm": 0.19365690895180718, + "learning_rate": 9.475039087963443e-06, + "loss": 0.0041, + "step": 1431 + }, + { + "epoch": 0.38024429102496016, + "grad_norm": 0.5438261709069115, + "learning_rate": 9.47405940517703e-06, + "loss": 0.063, + "step": 1432 + }, + { + "epoch": 0.380509824747743, + "grad_norm": 0.32752487349022863, + "learning_rate": 9.47307885984473e-06, + "loss": 0.0432, + "step": 1433 + }, + { + "epoch": 0.38077535847052574, + "grad_norm": 0.45526266402944837, + "learning_rate": 9.47209745215558e-06, + "loss": 0.032, + "step": 1434 + }, + { + "epoch": 0.38104089219330856, + "grad_norm": 1.0213144942650498, + "learning_rate": 9.471115182298786e-06, + "loss": 0.1035, + "step": 1435 + }, + { + "epoch": 0.3813064259160913, + "grad_norm": 0.4413563467345107, + "learning_rate": 9.470132050463716e-06, + "loss": 0.0354, + "step": 1436 + }, + { + "epoch": 0.38157195963887414, + "grad_norm": 0.0613716953521828, + "learning_rate": 9.469148056839907e-06, + "loss": 0.002, + "step": 1437 + }, + { + "epoch": 0.38183749336165695, + "grad_norm": 0.8153858333079927, + "learning_rate": 9.468163201617063e-06, + "loss": 0.082, + "step": 1438 + }, + { + "epoch": 0.3821030270844397, + "grad_norm": 0.5563381804784219, + "learning_rate": 9.467177484985048e-06, + "loss": 0.063, + "step": 1439 + }, + { + "epoch": 0.38236856080722254, + "grad_norm": 0.4245368835018735, + "learning_rate": 9.466190907133901e-06, + "loss": 0.0479, + "step": 1440 + }, + { + "epoch": 0.3826340945300053, + "grad_norm": 0.33400027625309914, + "learning_rate": 9.46520346825382e-06, + "loss": 0.0432, + "step": 1441 + }, + { + "epoch": 0.3828996282527881, + "grad_norm": 0.14631192270559776, + "learning_rate": 9.464215168535175e-06, + "loss": 0.0045, + "step": 1442 + }, + { + "epoch": 0.3831651619755709, + "grad_norm": 0.4620703244763112, + "learning_rate": 9.463226008168493e-06, + "loss": 0.0432, + "step": 1443 + }, + { + "epoch": 0.3834306956983537, + "grad_norm": 0.6625505596890234, + "learning_rate": 9.462235987344478e-06, + "loss": 0.0579, + "step": 1444 + }, + { + "epoch": 0.38369622942113646, + "grad_norm": 0.24691441798104835, + "learning_rate": 9.461245106253991e-06, + "loss": 0.0105, + "step": 1445 + }, + { + "epoch": 0.3839617631439193, + "grad_norm": 0.6337463070104612, + "learning_rate": 9.460253365088063e-06, + "loss": 0.0752, + "step": 1446 + }, + { + "epoch": 0.3842272968667021, + "grad_norm": 0.08857326498123558, + "learning_rate": 9.45926076403789e-06, + "loss": 0.0028, + "step": 1447 + }, + { + "epoch": 0.38449283058948486, + "grad_norm": 0.614060872702447, + "learning_rate": 9.458267303294833e-06, + "loss": 0.0957, + "step": 1448 + }, + { + "epoch": 0.3847583643122677, + "grad_norm": 0.34706341100621624, + "learning_rate": 9.457272983050421e-06, + "loss": 0.0432, + "step": 1449 + }, + { + "epoch": 0.38502389803505044, + "grad_norm": 0.5244496312054608, + "learning_rate": 9.456277803496345e-06, + "loss": 0.082, + "step": 1450 + }, + { + "epoch": 0.38528943175783326, + "grad_norm": 0.04965359509992197, + "learning_rate": 9.455281764824467e-06, + "loss": 0.0015, + "step": 1451 + }, + { + "epoch": 0.385554965480616, + "grad_norm": 0.4205660220211263, + "learning_rate": 9.45428486722681e-06, + "loss": 0.0286, + "step": 1452 + }, + { + "epoch": 0.38582049920339884, + "grad_norm": 0.5607004309206753, + "learning_rate": 9.453287110895564e-06, + "loss": 0.063, + "step": 1453 + }, + { + "epoch": 0.3860860329261816, + "grad_norm": 0.47148899218207774, + "learning_rate": 9.452288496023085e-06, + "loss": 0.0576, + "step": 1454 + }, + { + "epoch": 0.3863515666489644, + "grad_norm": 0.46004679938747756, + "learning_rate": 9.451289022801894e-06, + "loss": 0.0579, + "step": 1455 + }, + { + "epoch": 0.38661710037174724, + "grad_norm": 0.4292369995431107, + "learning_rate": 9.450288691424676e-06, + "loss": 0.0317, + "step": 1456 + }, + { + "epoch": 0.38688263409453, + "grad_norm": 0.9283721518542083, + "learning_rate": 9.449287502084284e-06, + "loss": 0.0957, + "step": 1457 + }, + { + "epoch": 0.3871481678173128, + "grad_norm": 1.199281966245374, + "learning_rate": 9.448285454973739e-06, + "loss": 0.063, + "step": 1458 + }, + { + "epoch": 0.3874137015400956, + "grad_norm": 1.5176795239281788, + "learning_rate": 9.447282550286217e-06, + "loss": 0.1191, + "step": 1459 + }, + { + "epoch": 0.3876792352628784, + "grad_norm": 0.574751228105574, + "learning_rate": 9.446278788215074e-06, + "loss": 0.0889, + "step": 1460 + }, + { + "epoch": 0.38794476898566116, + "grad_norm": 0.05157646996776102, + "learning_rate": 9.445274168953818e-06, + "loss": 0.0012, + "step": 1461 + }, + { + "epoch": 0.388210302708444, + "grad_norm": 0.7857880559484768, + "learning_rate": 9.444268692696129e-06, + "loss": 0.082, + "step": 1462 + }, + { + "epoch": 0.38847583643122674, + "grad_norm": 0.3007386860700484, + "learning_rate": 9.44326235963585e-06, + "loss": 0.0393, + "step": 1463 + }, + { + "epoch": 0.38874137015400956, + "grad_norm": 0.3055671787388409, + "learning_rate": 9.442255169966995e-06, + "loss": 0.0354, + "step": 1464 + }, + { + "epoch": 0.3890069038767924, + "grad_norm": 0.5568087984102646, + "learning_rate": 9.441247123883733e-06, + "loss": 0.0889, + "step": 1465 + }, + { + "epoch": 0.38927243759957514, + "grad_norm": 0.5895505413874539, + "learning_rate": 9.440238221580404e-06, + "loss": 0.1035, + "step": 1466 + }, + { + "epoch": 0.38953797132235796, + "grad_norm": 0.5616300389774914, + "learning_rate": 9.439228463251515e-06, + "loss": 0.0957, + "step": 1467 + }, + { + "epoch": 0.3898035050451407, + "grad_norm": 0.33063941694631, + "learning_rate": 9.438217849091732e-06, + "loss": 0.0393, + "step": 1468 + }, + { + "epoch": 0.39006903876792354, + "grad_norm": 0.2674438568163559, + "learning_rate": 9.437206379295895e-06, + "loss": 0.0354, + "step": 1469 + }, + { + "epoch": 0.3903345724907063, + "grad_norm": 0.7808160049243492, + "learning_rate": 9.436194054058998e-06, + "loss": 0.0752, + "step": 1470 + }, + { + "epoch": 0.3906001062134891, + "grad_norm": 0.5509891118998486, + "learning_rate": 9.435180873576208e-06, + "loss": 0.0957, + "step": 1471 + }, + { + "epoch": 0.3908656399362719, + "grad_norm": 0.6425294720673329, + "learning_rate": 9.434166838042854e-06, + "loss": 0.0889, + "step": 1472 + }, + { + "epoch": 0.3911311736590547, + "grad_norm": 1.8808615893246776, + "learning_rate": 9.433151947654428e-06, + "loss": 0.1279, + "step": 1473 + }, + { + "epoch": 0.3913967073818375, + "grad_norm": 0.6663733492038063, + "learning_rate": 9.432136202606594e-06, + "loss": 0.0184, + "step": 1474 + }, + { + "epoch": 0.3916622411046203, + "grad_norm": 0.5208712307214396, + "learning_rate": 9.431119603095169e-06, + "loss": 0.0889, + "step": 1475 + }, + { + "epoch": 0.3919277748274031, + "grad_norm": 0.2713853966362412, + "learning_rate": 9.430102149316147e-06, + "loss": 0.0354, + "step": 1476 + }, + { + "epoch": 0.39219330855018586, + "grad_norm": 0.5349973334569837, + "learning_rate": 9.429083841465679e-06, + "loss": 0.0957, + "step": 1477 + }, + { + "epoch": 0.3924588422729687, + "grad_norm": 0.1267971200504465, + "learning_rate": 9.428064679740081e-06, + "loss": 0.0025, + "step": 1478 + }, + { + "epoch": 0.39272437599575144, + "grad_norm": 0.5108936337689303, + "learning_rate": 9.427044664335837e-06, + "loss": 0.0815, + "step": 1479 + }, + { + "epoch": 0.39298990971853426, + "grad_norm": 0.33305136147342523, + "learning_rate": 9.426023795449595e-06, + "loss": 0.0479, + "step": 1480 + }, + { + "epoch": 0.393255443441317, + "grad_norm": 0.2984668226042693, + "learning_rate": 9.425002073278165e-06, + "loss": 0.0432, + "step": 1481 + }, + { + "epoch": 0.39352097716409984, + "grad_norm": 0.33394165373144796, + "learning_rate": 9.423979498018521e-06, + "loss": 0.0432, + "step": 1482 + }, + { + "epoch": 0.39378651088688266, + "grad_norm": 0.49293937391002435, + "learning_rate": 9.422956069867808e-06, + "loss": 0.0889, + "step": 1483 + }, + { + "epoch": 0.3940520446096654, + "grad_norm": 0.5210948201626876, + "learning_rate": 9.421931789023329e-06, + "loss": 0.0579, + "step": 1484 + }, + { + "epoch": 0.39431757833244824, + "grad_norm": 0.41304821586273616, + "learning_rate": 9.420906655682553e-06, + "loss": 0.0432, + "step": 1485 + }, + { + "epoch": 0.394583112055231, + "grad_norm": 0.46657739059291675, + "learning_rate": 9.41988067004311e-06, + "loss": 0.0354, + "step": 1486 + }, + { + "epoch": 0.3948486457780138, + "grad_norm": 0.026843605916078473, + "learning_rate": 9.418853832302806e-06, + "loss": 0.0007, + "step": 1487 + }, + { + "epoch": 0.3951141795007966, + "grad_norm": 0.2666054121708235, + "learning_rate": 9.417826142659596e-06, + "loss": 0.0103, + "step": 1488 + }, + { + "epoch": 0.3953797132235794, + "grad_norm": 0.3517104605858895, + "learning_rate": 9.416797601311611e-06, + "loss": 0.032, + "step": 1489 + }, + { + "epoch": 0.39564524694636216, + "grad_norm": 0.448265208584516, + "learning_rate": 9.415768208457139e-06, + "loss": 0.0693, + "step": 1490 + }, + { + "epoch": 0.395910780669145, + "grad_norm": 0.4432887319257767, + "learning_rate": 9.414737964294636e-06, + "loss": 0.063, + "step": 1491 + }, + { + "epoch": 0.3961763143919278, + "grad_norm": 0.15615182853960044, + "learning_rate": 9.41370686902272e-06, + "loss": 0.0057, + "step": 1492 + }, + { + "epoch": 0.39644184811471056, + "grad_norm": 0.40174854437847995, + "learning_rate": 9.412674922840175e-06, + "loss": 0.0432, + "step": 1493 + }, + { + "epoch": 0.3967073818374934, + "grad_norm": 0.5334231337886949, + "learning_rate": 9.411642125945945e-06, + "loss": 0.0752, + "step": 1494 + }, + { + "epoch": 0.39697291556027614, + "grad_norm": 0.39350262505624917, + "learning_rate": 9.410608478539146e-06, + "loss": 0.0525, + "step": 1495 + }, + { + "epoch": 0.39723844928305896, + "grad_norm": 0.5104750407099554, + "learning_rate": 9.409573980819052e-06, + "loss": 0.0432, + "step": 1496 + }, + { + "epoch": 0.3975039830058417, + "grad_norm": 0.018301303401147544, + "learning_rate": 9.408538632985098e-06, + "loss": 0.0005, + "step": 1497 + }, + { + "epoch": 0.39776951672862454, + "grad_norm": 2.925231221732229, + "learning_rate": 9.407502435236889e-06, + "loss": 0.0184, + "step": 1498 + }, + { + "epoch": 0.3980350504514073, + "grad_norm": 0.43976667471652925, + "learning_rate": 9.406465387774194e-06, + "loss": 0.032, + "step": 1499 + }, + { + "epoch": 0.3983005841741901, + "grad_norm": 0.004447437986864806, + "learning_rate": 9.405427490796941e-06, + "loss": 0.0001, + "step": 1500 + }, + { + "epoch": 0.39856611789697294, + "grad_norm": 0.22189544659424998, + "learning_rate": 9.404388744505225e-06, + "loss": 0.0147, + "step": 1501 + }, + { + "epoch": 0.3988316516197557, + "grad_norm": 0.5323862127877603, + "learning_rate": 9.403349149099304e-06, + "loss": 0.0354, + "step": 1502 + }, + { + "epoch": 0.3990971853425385, + "grad_norm": 0.7824556156322893, + "learning_rate": 9.4023087047796e-06, + "loss": 0.1035, + "step": 1503 + }, + { + "epoch": 0.3993627190653213, + "grad_norm": 0.537555019938912, + "learning_rate": 9.401267411746697e-06, + "loss": 0.0752, + "step": 1504 + }, + { + "epoch": 0.3996282527881041, + "grad_norm": 1.2599751022767143, + "learning_rate": 9.400225270201347e-06, + "loss": 0.1279, + "step": 1505 + }, + { + "epoch": 0.39989378651088686, + "grad_norm": 0.39943457727783255, + "learning_rate": 9.399182280344458e-06, + "loss": 0.0286, + "step": 1506 + }, + { + "epoch": 0.4001593202336697, + "grad_norm": 0.3951237763375591, + "learning_rate": 9.398138442377112e-06, + "loss": 0.0354, + "step": 1507 + }, + { + "epoch": 0.40042485395645244, + "grad_norm": 0.3348310768931224, + "learning_rate": 9.397093756500544e-06, + "loss": 0.0258, + "step": 1508 + }, + { + "epoch": 0.40069038767923526, + "grad_norm": 0.6183702558458398, + "learning_rate": 9.396048222916159e-06, + "loss": 0.0889, + "step": 1509 + }, + { + "epoch": 0.4009559214020181, + "grad_norm": 0.7389099287695277, + "learning_rate": 9.395001841825523e-06, + "loss": 0.0957, + "step": 1510 + }, + { + "epoch": 0.40122145512480084, + "grad_norm": 0.5072802900638348, + "learning_rate": 9.393954613430365e-06, + "loss": 0.0752, + "step": 1511 + }, + { + "epoch": 0.40148698884758366, + "grad_norm": 0.6807209999699108, + "learning_rate": 9.392906537932582e-06, + "loss": 0.0957, + "step": 1512 + }, + { + "epoch": 0.4017525225703664, + "grad_norm": 0.4704376229884201, + "learning_rate": 9.391857615534226e-06, + "loss": 0.0286, + "step": 1513 + }, + { + "epoch": 0.40201805629314924, + "grad_norm": 0.028093132307861573, + "learning_rate": 9.39080784643752e-06, + "loss": 0.0006, + "step": 1514 + }, + { + "epoch": 0.402283590015932, + "grad_norm": 0.5084581122752261, + "learning_rate": 9.389757230844845e-06, + "loss": 0.0688, + "step": 1515 + }, + { + "epoch": 0.4025491237387148, + "grad_norm": 0.006193390834703029, + "learning_rate": 9.388705768958748e-06, + "loss": 0.0001, + "step": 1516 + }, + { + "epoch": 0.4028146574614976, + "grad_norm": 0.46353104622161806, + "learning_rate": 9.38765346098194e-06, + "loss": 0.0525, + "step": 1517 + }, + { + "epoch": 0.4030801911842804, + "grad_norm": 0.49616262621142954, + "learning_rate": 9.386600307117293e-06, + "loss": 0.0393, + "step": 1518 + }, + { + "epoch": 0.4033457249070632, + "grad_norm": 0.0815583003071057, + "learning_rate": 9.38554630756784e-06, + "loss": 0.0022, + "step": 1519 + }, + { + "epoch": 0.403611258629846, + "grad_norm": 0.48994190511403873, + "learning_rate": 9.384491462536783e-06, + "loss": 0.0354, + "step": 1520 + }, + { + "epoch": 0.4038767923526288, + "grad_norm": 0.11251221678106954, + "learning_rate": 9.383435772227481e-06, + "loss": 0.004, + "step": 1521 + }, + { + "epoch": 0.40414232607541156, + "grad_norm": 0.5397658046795061, + "learning_rate": 9.382379236843459e-06, + "loss": 0.0688, + "step": 1522 + }, + { + "epoch": 0.4044078597981944, + "grad_norm": 0.09364212325756312, + "learning_rate": 9.381321856588407e-06, + "loss": 0.0032, + "step": 1523 + }, + { + "epoch": 0.40467339352097714, + "grad_norm": 0.9230122195682332, + "learning_rate": 9.380263631666175e-06, + "loss": 0.0815, + "step": 1524 + }, + { + "epoch": 0.40493892724375996, + "grad_norm": 0.9314503563752995, + "learning_rate": 9.379204562280775e-06, + "loss": 0.0889, + "step": 1525 + }, + { + "epoch": 0.4052044609665427, + "grad_norm": 0.60796669683741, + "learning_rate": 9.378144648636382e-06, + "loss": 0.0688, + "step": 1526 + }, + { + "epoch": 0.40546999468932554, + "grad_norm": 0.5554197315940151, + "learning_rate": 9.377083890937337e-06, + "loss": 0.0576, + "step": 1527 + }, + { + "epoch": 0.40573552841210836, + "grad_norm": 0.12196493290704465, + "learning_rate": 9.37602228938814e-06, + "loss": 0.0028, + "step": 1528 + }, + { + "epoch": 0.4060010621348911, + "grad_norm": 0.3126139063656892, + "learning_rate": 9.374959844193456e-06, + "loss": 0.0354, + "step": 1529 + }, + { + "epoch": 0.40626659585767394, + "grad_norm": 0.011854084112567173, + "learning_rate": 9.373896555558113e-06, + "loss": 0.0003, + "step": 1530 + }, + { + "epoch": 0.4065321295804567, + "grad_norm": 3.841192190633747, + "learning_rate": 9.372832423687099e-06, + "loss": 0.2031, + "step": 1531 + }, + { + "epoch": 0.4067976633032395, + "grad_norm": 0.6503926893484147, + "learning_rate": 9.371767448785566e-06, + "loss": 0.0184, + "step": 1532 + }, + { + "epoch": 0.4070631970260223, + "grad_norm": 0.8960971685517782, + "learning_rate": 9.37070163105883e-06, + "loss": 0.0815, + "step": 1533 + }, + { + "epoch": 0.4073287307488051, + "grad_norm": 0.011994467119951549, + "learning_rate": 9.369634970712365e-06, + "loss": 0.0003, + "step": 1534 + }, + { + "epoch": 0.4075942644715879, + "grad_norm": 0.00513544623691197, + "learning_rate": 9.368567467951815e-06, + "loss": 0.0001, + "step": 1535 + }, + { + "epoch": 0.4078597981943707, + "grad_norm": 0.043827008510423704, + "learning_rate": 9.367499122982979e-06, + "loss": 0.0009, + "step": 1536 + }, + { + "epoch": 0.4081253319171535, + "grad_norm": 0.46712106544993076, + "learning_rate": 9.366429936011822e-06, + "loss": 0.0476, + "step": 1537 + }, + { + "epoch": 0.40839086563993626, + "grad_norm": 0.07965123738499298, + "learning_rate": 9.365359907244469e-06, + "loss": 0.0019, + "step": 1538 + }, + { + "epoch": 0.4086563993627191, + "grad_norm": 0.5002938341280985, + "learning_rate": 9.364289036887214e-06, + "loss": 0.0476, + "step": 1539 + }, + { + "epoch": 0.40892193308550184, + "grad_norm": 0.5932213624042697, + "learning_rate": 9.3632173251465e-06, + "loss": 0.0286, + "step": 1540 + }, + { + "epoch": 0.40918746680828466, + "grad_norm": 0.01294158924164034, + "learning_rate": 9.362144772228949e-06, + "loss": 0.0004, + "step": 1541 + }, + { + "epoch": 0.4094530005310674, + "grad_norm": 0.6159693393016078, + "learning_rate": 9.361071378341332e-06, + "loss": 0.0231, + "step": 1542 + }, + { + "epoch": 0.40971853425385024, + "grad_norm": 0.09689935660402307, + "learning_rate": 9.359997143690586e-06, + "loss": 0.0019, + "step": 1543 + }, + { + "epoch": 0.40998406797663306, + "grad_norm": 0.3554406490463387, + "learning_rate": 9.358922068483813e-06, + "loss": 0.0391, + "step": 1544 + }, + { + "epoch": 0.4102496016994158, + "grad_norm": 4.665948195663397, + "learning_rate": 9.357846152928275e-06, + "loss": 0.3535, + "step": 1545 + }, + { + "epoch": 0.41051513542219864, + "grad_norm": 0.6643692186727801, + "learning_rate": 9.356769397231391e-06, + "loss": 0.0206, + "step": 1546 + }, + { + "epoch": 0.4107806691449814, + "grad_norm": 0.3417631602945585, + "learning_rate": 9.355691801600754e-06, + "loss": 0.0354, + "step": 1547 + }, + { + "epoch": 0.4110462028677642, + "grad_norm": 0.04476331486140047, + "learning_rate": 9.354613366244108e-06, + "loss": 0.0006, + "step": 1548 + }, + { + "epoch": 0.411311736590547, + "grad_norm": 0.6284446073342773, + "learning_rate": 9.35353409136936e-06, + "loss": 0.0815, + "step": 1549 + }, + { + "epoch": 0.4115772703133298, + "grad_norm": 0.3822928656858541, + "learning_rate": 9.352453977184588e-06, + "loss": 0.0391, + "step": 1550 + }, + { + "epoch": 0.41184280403611256, + "grad_norm": 0.6046689508647543, + "learning_rate": 9.35137302389802e-06, + "loss": 0.1035, + "step": 1551 + }, + { + "epoch": 0.4121083377588954, + "grad_norm": 0.00728328734429902, + "learning_rate": 9.350291231718051e-06, + "loss": 0.0001, + "step": 1552 + }, + { + "epoch": 0.4123738714816782, + "grad_norm": 0.41563519871352395, + "learning_rate": 9.349208600853242e-06, + "loss": 0.0131, + "step": 1553 + }, + { + "epoch": 0.41263940520446096, + "grad_norm": 0.8926182157973549, + "learning_rate": 9.348125131512309e-06, + "loss": 0.0688, + "step": 1554 + }, + { + "epoch": 0.4129049389272438, + "grad_norm": 0.682749649537968, + "learning_rate": 9.34704082390413e-06, + "loss": 0.0889, + "step": 1555 + }, + { + "epoch": 0.41317047265002654, + "grad_norm": 0.004689924901466316, + "learning_rate": 9.34595567823775e-06, + "loss": 0.0001, + "step": 1556 + }, + { + "epoch": 0.41343600637280936, + "grad_norm": 0.6858612644220904, + "learning_rate": 9.344869694722372e-06, + "loss": 0.0889, + "step": 1557 + }, + { + "epoch": 0.4137015400955921, + "grad_norm": 0.4520811232153325, + "learning_rate": 9.343782873567358e-06, + "loss": 0.0391, + "step": 1558 + }, + { + "epoch": 0.41396707381837494, + "grad_norm": 0.3326484721722159, + "learning_rate": 9.342695214982236e-06, + "loss": 0.0354, + "step": 1559 + }, + { + "epoch": 0.4142326075411577, + "grad_norm": 0.3853243104104061, + "learning_rate": 9.341606719176695e-06, + "loss": 0.0432, + "step": 1560 + }, + { + "epoch": 0.4144981412639405, + "grad_norm": 0.01786226839288024, + "learning_rate": 9.340517386360584e-06, + "loss": 0.0004, + "step": 1561 + }, + { + "epoch": 0.41476367498672334, + "grad_norm": 0.014836540197822963, + "learning_rate": 9.33942721674391e-06, + "loss": 0.0003, + "step": 1562 + }, + { + "epoch": 0.4150292087095061, + "grad_norm": 0.6491829250530613, + "learning_rate": 9.33833621053685e-06, + "loss": 0.063, + "step": 1563 + }, + { + "epoch": 0.4152947424322889, + "grad_norm": 0.0008502111840714258, + "learning_rate": 9.337244367949733e-06, + "loss": 0.0, + "step": 1564 + }, + { + "epoch": 0.4155602761550717, + "grad_norm": 0.4693791006428915, + "learning_rate": 9.336151689193055e-06, + "loss": 0.0476, + "step": 1565 + }, + { + "epoch": 0.4158258098778545, + "grad_norm": 0.638806505853796, + "learning_rate": 9.335058174477472e-06, + "loss": 0.0525, + "step": 1566 + }, + { + "epoch": 0.41609134360063726, + "grad_norm": 3.7713130277807108, + "learning_rate": 9.333963824013798e-06, + "loss": 0.2246, + "step": 1567 + }, + { + "epoch": 0.4163568773234201, + "grad_norm": 0.004494212269452434, + "learning_rate": 9.332868638013016e-06, + "loss": 0.0001, + "step": 1568 + }, + { + "epoch": 0.41662241104620285, + "grad_norm": 0.20743155158952067, + "learning_rate": 9.331772616686262e-06, + "loss": 0.0051, + "step": 1569 + }, + { + "epoch": 0.41688794476898566, + "grad_norm": 0.6985128357597653, + "learning_rate": 9.330675760244834e-06, + "loss": 0.0432, + "step": 1570 + }, + { + "epoch": 0.4171534784917685, + "grad_norm": 0.9089238790009025, + "learning_rate": 9.329578068900195e-06, + "loss": 0.063, + "step": 1571 + }, + { + "epoch": 0.41741901221455124, + "grad_norm": 2.064976993600294, + "learning_rate": 9.328479542863968e-06, + "loss": 0.1455, + "step": 1572 + }, + { + "epoch": 0.41768454593733406, + "grad_norm": 0.4081315774110765, + "learning_rate": 9.327380182347937e-06, + "loss": 0.0354, + "step": 1573 + }, + { + "epoch": 0.4179500796601168, + "grad_norm": 0.34553567048811684, + "learning_rate": 9.32627998756404e-06, + "loss": 0.0354, + "step": 1574 + }, + { + "epoch": 0.41821561338289964, + "grad_norm": 0.74457240073377, + "learning_rate": 9.325178958724387e-06, + "loss": 0.0476, + "step": 1575 + }, + { + "epoch": 0.4184811471056824, + "grad_norm": 0.02077584538656709, + "learning_rate": 9.32407709604124e-06, + "loss": 0.0005, + "step": 1576 + }, + { + "epoch": 0.4187466808284652, + "grad_norm": 1.5760043575275893, + "learning_rate": 9.322974399727029e-06, + "loss": 0.1191, + "step": 1577 + }, + { + "epoch": 0.419012214551248, + "grad_norm": 0.08166506720345244, + "learning_rate": 9.321870869994336e-06, + "loss": 0.0013, + "step": 1578 + }, + { + "epoch": 0.4192777482740308, + "grad_norm": 0.9348396362017879, + "learning_rate": 9.320766507055912e-06, + "loss": 0.0688, + "step": 1579 + }, + { + "epoch": 0.4195432819968136, + "grad_norm": 0.27331732617626087, + "learning_rate": 9.319661311124664e-06, + "loss": 0.0317, + "step": 1580 + }, + { + "epoch": 0.4198088157195964, + "grad_norm": 5.750235838182297, + "learning_rate": 9.318555282413662e-06, + "loss": 0.3809, + "step": 1581 + }, + { + "epoch": 0.4200743494423792, + "grad_norm": 0.6368103985830037, + "learning_rate": 9.317448421136134e-06, + "loss": 0.0476, + "step": 1582 + }, + { + "epoch": 0.42033988316516196, + "grad_norm": 0.2827804927558854, + "learning_rate": 9.316340727505469e-06, + "loss": 0.0317, + "step": 1583 + }, + { + "epoch": 0.4206054168879448, + "grad_norm": 0.30307381764984753, + "learning_rate": 9.315232201735217e-06, + "loss": 0.0064, + "step": 1584 + }, + { + "epoch": 0.42087095061072755, + "grad_norm": 0.21252100392930384, + "learning_rate": 9.314122844039094e-06, + "loss": 0.0231, + "step": 1585 + }, + { + "epoch": 0.42113648433351036, + "grad_norm": 1.2522860606912725, + "learning_rate": 9.313012654630963e-06, + "loss": 0.1035, + "step": 1586 + }, + { + "epoch": 0.4214020180562931, + "grad_norm": 0.022875048396212985, + "learning_rate": 9.311901633724861e-06, + "loss": 0.0006, + "step": 1587 + }, + { + "epoch": 0.42166755177907594, + "grad_norm": 0.9338167878973481, + "learning_rate": 9.310789781534978e-06, + "loss": 0.0286, + "step": 1588 + }, + { + "epoch": 0.42193308550185876, + "grad_norm": 0.6025789069506893, + "learning_rate": 9.309677098275667e-06, + "loss": 0.0476, + "step": 1589 + }, + { + "epoch": 0.4221986192246415, + "grad_norm": 0.15935459440059752, + "learning_rate": 9.308563584161439e-06, + "loss": 0.0146, + "step": 1590 + }, + { + "epoch": 0.42246415294742434, + "grad_norm": 0.09990679440548342, + "learning_rate": 9.307449239406967e-06, + "loss": 0.0021, + "step": 1591 + }, + { + "epoch": 0.4227296866702071, + "grad_norm": 0.4308256760120562, + "learning_rate": 9.306334064227082e-06, + "loss": 0.0206, + "step": 1592 + }, + { + "epoch": 0.4229952203929899, + "grad_norm": 0.17557875823654812, + "learning_rate": 9.305218058836778e-06, + "loss": 0.0165, + "step": 1593 + }, + { + "epoch": 0.4232607541157727, + "grad_norm": 0.8469708090699332, + "learning_rate": 9.304101223451209e-06, + "loss": 0.0391, + "step": 1594 + }, + { + "epoch": 0.4235262878385555, + "grad_norm": 0.9742488100637656, + "learning_rate": 9.302983558285686e-06, + "loss": 0.1738, + "step": 1595 + }, + { + "epoch": 0.42379182156133827, + "grad_norm": 0.3807711068184408, + "learning_rate": 9.301865063555681e-06, + "loss": 0.0255, + "step": 1596 + }, + { + "epoch": 0.4240573552841211, + "grad_norm": 0.1865814006843838, + "learning_rate": 9.30074573947683e-06, + "loss": 0.0146, + "step": 1597 + }, + { + "epoch": 0.4243228890069039, + "grad_norm": 0.9795085871771619, + "learning_rate": 9.299625586264918e-06, + "loss": 0.1455, + "step": 1598 + }, + { + "epoch": 0.42458842272968667, + "grad_norm": 0.8224575743387007, + "learning_rate": 9.298504604135906e-06, + "loss": 0.1641, + "step": 1599 + }, + { + "epoch": 0.4248539564524695, + "grad_norm": 0.08459635905989443, + "learning_rate": 9.297382793305902e-06, + "loss": 0.0017, + "step": 1600 + }, + { + "epoch": 0.42511949017525225, + "grad_norm": 0.9247975416587106, + "learning_rate": 9.29626015399118e-06, + "loss": 0.0525, + "step": 1601 + }, + { + "epoch": 0.42538502389803506, + "grad_norm": 0.9030669338926891, + "learning_rate": 9.295136686408165e-06, + "loss": 0.0815, + "step": 1602 + }, + { + "epoch": 0.4256505576208178, + "grad_norm": 0.176488322331327, + "learning_rate": 9.294012390773457e-06, + "loss": 0.0184, + "step": 1603 + }, + { + "epoch": 0.42591609134360064, + "grad_norm": 0.7575042176122994, + "learning_rate": 9.292887267303803e-06, + "loss": 0.1113, + "step": 1604 + }, + { + "epoch": 0.4261816250663834, + "grad_norm": 2.8198240780302513, + "learning_rate": 9.291761316216115e-06, + "loss": 0.0889, + "step": 1605 + }, + { + "epoch": 0.4264471587891662, + "grad_norm": 0.80505821260424, + "learning_rate": 9.290634537727459e-06, + "loss": 0.0889, + "step": 1606 + }, + { + "epoch": 0.42671269251194904, + "grad_norm": 0.4004451010986326, + "learning_rate": 9.28950693205507e-06, + "loss": 0.0092, + "step": 1607 + }, + { + "epoch": 0.4269782262347318, + "grad_norm": 0.05174174947474149, + "learning_rate": 9.288378499416334e-06, + "loss": 0.0013, + "step": 1608 + }, + { + "epoch": 0.4272437599575146, + "grad_norm": 0.2673693573498579, + "learning_rate": 9.287249240028799e-06, + "loss": 0.0258, + "step": 1609 + }, + { + "epoch": 0.4275092936802974, + "grad_norm": 0.20068759525709715, + "learning_rate": 9.286119154110173e-06, + "loss": 0.0051, + "step": 1610 + }, + { + "epoch": 0.4277748274030802, + "grad_norm": 0.2052076895631866, + "learning_rate": 9.284988241878326e-06, + "loss": 0.0206, + "step": 1611 + }, + { + "epoch": 0.42804036112586297, + "grad_norm": 0.1826457677627445, + "learning_rate": 9.283856503551283e-06, + "loss": 0.0165, + "step": 1612 + }, + { + "epoch": 0.4283058948486458, + "grad_norm": 0.016102516202645443, + "learning_rate": 9.282723939347227e-06, + "loss": 0.0003, + "step": 1613 + }, + { + "epoch": 0.42857142857142855, + "grad_norm": 0.06556414219099314, + "learning_rate": 9.281590549484508e-06, + "loss": 0.0015, + "step": 1614 + }, + { + "epoch": 0.42883696229421137, + "grad_norm": 0.6492833856406688, + "learning_rate": 9.280456334181626e-06, + "loss": 0.1035, + "step": 1615 + }, + { + "epoch": 0.4291024960169942, + "grad_norm": 0.26676720638479, + "learning_rate": 9.279321293657248e-06, + "loss": 0.0255, + "step": 1616 + }, + { + "epoch": 0.42936802973977695, + "grad_norm": 1.0528365853705146, + "learning_rate": 9.278185428130194e-06, + "loss": 0.0476, + "step": 1617 + }, + { + "epoch": 0.42963356346255976, + "grad_norm": 0.300356633549314, + "learning_rate": 9.277048737819445e-06, + "loss": 0.0255, + "step": 1618 + }, + { + "epoch": 0.4298990971853425, + "grad_norm": 0.28681150202389294, + "learning_rate": 9.275911222944143e-06, + "loss": 0.0317, + "step": 1619 + }, + { + "epoch": 0.43016463090812534, + "grad_norm": 0.41351911300181254, + "learning_rate": 9.274772883723587e-06, + "loss": 0.0317, + "step": 1620 + }, + { + "epoch": 0.4304301646309081, + "grad_norm": 0.3435709971804024, + "learning_rate": 9.273633720377235e-06, + "loss": 0.0354, + "step": 1621 + }, + { + "epoch": 0.4306956983536909, + "grad_norm": 0.006280834272583653, + "learning_rate": 9.272493733124706e-06, + "loss": 0.0002, + "step": 1622 + }, + { + "epoch": 0.4309612320764737, + "grad_norm": 0.31013270569482426, + "learning_rate": 9.271352922185772e-06, + "loss": 0.0286, + "step": 1623 + }, + { + "epoch": 0.4312267657992565, + "grad_norm": 0.003193536264216083, + "learning_rate": 9.270211287780374e-06, + "loss": 0.0001, + "step": 1624 + }, + { + "epoch": 0.4314922995220393, + "grad_norm": 1.2572420728217566, + "learning_rate": 9.269068830128599e-06, + "loss": 0.0815, + "step": 1625 + }, + { + "epoch": 0.4317578332448221, + "grad_norm": 0.16424497737231764, + "learning_rate": 9.267925549450703e-06, + "loss": 0.0092, + "step": 1626 + }, + { + "epoch": 0.4320233669676049, + "grad_norm": 0.0051496845292951345, + "learning_rate": 9.266781445967098e-06, + "loss": 0.0001, + "step": 1627 + }, + { + "epoch": 0.43228890069038767, + "grad_norm": 0.4131524580991801, + "learning_rate": 9.26563651989835e-06, + "loss": 0.0255, + "step": 1628 + }, + { + "epoch": 0.4325544344131705, + "grad_norm": 0.005905247669898074, + "learning_rate": 9.264490771465191e-06, + "loss": 0.0001, + "step": 1629 + }, + { + "epoch": 0.43281996813595325, + "grad_norm": 0.11239806675648503, + "learning_rate": 9.263344200888507e-06, + "loss": 0.005, + "step": 1630 + }, + { + "epoch": 0.43308550185873607, + "grad_norm": 1.9861947240668674, + "learning_rate": 9.26219680838934e-06, + "loss": 0.1279, + "step": 1631 + }, + { + "epoch": 0.43335103558151883, + "grad_norm": 0.1940044319795506, + "learning_rate": 9.2610485941889e-06, + "loss": 0.0131, + "step": 1632 + }, + { + "epoch": 0.43361656930430165, + "grad_norm": 0.2272907921942844, + "learning_rate": 9.259899558508543e-06, + "loss": 0.0205, + "step": 1633 + }, + { + "epoch": 0.43388210302708446, + "grad_norm": 0.4199007338504671, + "learning_rate": 9.258749701569793e-06, + "loss": 0.0092, + "step": 1634 + }, + { + "epoch": 0.4341476367498672, + "grad_norm": 0.5230506299864369, + "learning_rate": 9.257599023594326e-06, + "loss": 0.0352, + "step": 1635 + }, + { + "epoch": 0.43441317047265005, + "grad_norm": 1.7928341347867058, + "learning_rate": 9.256447524803983e-06, + "loss": 0.2129, + "step": 1636 + }, + { + "epoch": 0.4346787041954328, + "grad_norm": 1.0771613409897511, + "learning_rate": 9.255295205420757e-06, + "loss": 0.1035, + "step": 1637 + }, + { + "epoch": 0.4349442379182156, + "grad_norm": 0.23823363260187153, + "learning_rate": 9.254142065666802e-06, + "loss": 0.0229, + "step": 1638 + }, + { + "epoch": 0.4352097716409984, + "grad_norm": 1.7503348124699936, + "learning_rate": 9.252988105764428e-06, + "loss": 0.2129, + "step": 1639 + }, + { + "epoch": 0.4354753053637812, + "grad_norm": 0.002717147032072083, + "learning_rate": 9.251833325936108e-06, + "loss": 0.0001, + "step": 1640 + }, + { + "epoch": 0.43574083908656397, + "grad_norm": 0.43583660633350435, + "learning_rate": 9.25067772640447e-06, + "loss": 0.0229, + "step": 1641 + }, + { + "epoch": 0.4360063728093468, + "grad_norm": 0.01550132302821912, + "learning_rate": 9.249521307392294e-06, + "loss": 0.0003, + "step": 1642 + }, + { + "epoch": 0.4362719065321296, + "grad_norm": 0.21256823089610236, + "learning_rate": 9.248364069122531e-06, + "loss": 0.0146, + "step": 1643 + }, + { + "epoch": 0.43653744025491237, + "grad_norm": 0.2227015636760905, + "learning_rate": 9.247206011818281e-06, + "loss": 0.0229, + "step": 1644 + }, + { + "epoch": 0.4368029739776952, + "grad_norm": 0.3421213770839244, + "learning_rate": 9.246047135702802e-06, + "loss": 0.0286, + "step": 1645 + }, + { + "epoch": 0.43706850770047795, + "grad_norm": 0.2175367153596953, + "learning_rate": 9.244887440999515e-06, + "loss": 0.0056, + "step": 1646 + }, + { + "epoch": 0.43733404142326077, + "grad_norm": 0.2561588514828441, + "learning_rate": 9.24372692793199e-06, + "loss": 0.0255, + "step": 1647 + }, + { + "epoch": 0.43759957514604353, + "grad_norm": 0.5116398168106552, + "learning_rate": 9.242565596723965e-06, + "loss": 0.0103, + "step": 1648 + }, + { + "epoch": 0.43786510886882635, + "grad_norm": 0.004941866451653809, + "learning_rate": 9.241403447599329e-06, + "loss": 0.0001, + "step": 1649 + }, + { + "epoch": 0.4381306425916091, + "grad_norm": 0.07087433757924969, + "learning_rate": 9.24024048078213e-06, + "loss": 0.0015, + "step": 1650 + }, + { + "epoch": 0.4383961763143919, + "grad_norm": 0.33106687803013696, + "learning_rate": 9.239076696496577e-06, + "loss": 0.0354, + "step": 1651 + }, + { + "epoch": 0.43866171003717475, + "grad_norm": 0.0641176743240912, + "learning_rate": 9.237912094967031e-06, + "loss": 0.0012, + "step": 1652 + }, + { + "epoch": 0.4389272437599575, + "grad_norm": 0.30760602554788097, + "learning_rate": 9.236746676418016e-06, + "loss": 0.0255, + "step": 1653 + }, + { + "epoch": 0.4391927774827403, + "grad_norm": 0.1564547333112549, + "learning_rate": 9.235580441074208e-06, + "loss": 0.0129, + "step": 1654 + }, + { + "epoch": 0.4394583112055231, + "grad_norm": 1.3722315058379413, + "learning_rate": 9.234413389160446e-06, + "loss": 0.1191, + "step": 1655 + }, + { + "epoch": 0.4397238449283059, + "grad_norm": 1.2711619824610763, + "learning_rate": 9.233245520901723e-06, + "loss": 0.0815, + "step": 1656 + }, + { + "epoch": 0.43998937865108867, + "grad_norm": 0.8074581064566986, + "learning_rate": 9.23207683652319e-06, + "loss": 0.0317, + "step": 1657 + }, + { + "epoch": 0.4402549123738715, + "grad_norm": 0.18847549880372136, + "learning_rate": 9.230907336250155e-06, + "loss": 0.0165, + "step": 1658 + }, + { + "epoch": 0.44052044609665425, + "grad_norm": 0.3653202272308303, + "learning_rate": 9.229737020308084e-06, + "loss": 0.0092, + "step": 1659 + }, + { + "epoch": 0.44078597981943707, + "grad_norm": 5.255587093506221, + "learning_rate": 9.228565888922602e-06, + "loss": 0.3984, + "step": 1660 + }, + { + "epoch": 0.4410515135422199, + "grad_norm": 0.3264226155423392, + "learning_rate": 9.227393942319487e-06, + "loss": 0.0255, + "step": 1661 + }, + { + "epoch": 0.44131704726500265, + "grad_norm": 0.34863206865018187, + "learning_rate": 9.226221180724677e-06, + "loss": 0.0205, + "step": 1662 + }, + { + "epoch": 0.44158258098778547, + "grad_norm": 0.28323093506885416, + "learning_rate": 9.225047604364268e-06, + "loss": 0.0229, + "step": 1663 + }, + { + "epoch": 0.44184811471056823, + "grad_norm": 0.17631014929000818, + "learning_rate": 9.223873213464511e-06, + "loss": 0.0115, + "step": 1664 + }, + { + "epoch": 0.44211364843335105, + "grad_norm": 1.099027808353995, + "learning_rate": 9.222698008251814e-06, + "loss": 0.0432, + "step": 1665 + }, + { + "epoch": 0.4423791821561338, + "grad_norm": 0.10350955443306387, + "learning_rate": 9.221521988952742e-06, + "loss": 0.0021, + "step": 1666 + }, + { + "epoch": 0.44264471587891663, + "grad_norm": 0.5117982387799549, + "learning_rate": 9.22034515579402e-06, + "loss": 0.0317, + "step": 1667 + }, + { + "epoch": 0.4429102496016994, + "grad_norm": 0.17219084419699793, + "learning_rate": 9.219167509002526e-06, + "loss": 0.0092, + "step": 1668 + }, + { + "epoch": 0.4431757833244822, + "grad_norm": 1.6574596891900364, + "learning_rate": 9.217989048805297e-06, + "loss": 0.0957, + "step": 1669 + }, + { + "epoch": 0.443441317047265, + "grad_norm": 1.3273618450366584, + "learning_rate": 9.216809775429524e-06, + "loss": 0.1836, + "step": 1670 + }, + { + "epoch": 0.4437068507700478, + "grad_norm": 1.7850762788827483, + "learning_rate": 9.215629689102561e-06, + "loss": 0.1113, + "step": 1671 + }, + { + "epoch": 0.4439723844928306, + "grad_norm": 0.0017107214111774336, + "learning_rate": 9.214448790051913e-06, + "loss": 0.0, + "step": 1672 + }, + { + "epoch": 0.44423791821561337, + "grad_norm": 0.28213652008092643, + "learning_rate": 9.213267078505243e-06, + "loss": 0.0072, + "step": 1673 + }, + { + "epoch": 0.4445034519383962, + "grad_norm": 0.1888622989632113, + "learning_rate": 9.21208455469037e-06, + "loss": 0.0164, + "step": 1674 + }, + { + "epoch": 0.44476898566117895, + "grad_norm": 2.354678143727365, + "learning_rate": 9.210901218835273e-06, + "loss": 0.1367, + "step": 1675 + }, + { + "epoch": 0.44503451938396177, + "grad_norm": 0.1314357064594671, + "learning_rate": 9.209717071168083e-06, + "loss": 0.0092, + "step": 1676 + }, + { + "epoch": 0.44530005310674453, + "grad_norm": 2.3597528685967464, + "learning_rate": 9.208532111917092e-06, + "loss": 0.0957, + "step": 1677 + }, + { + "epoch": 0.44556558682952735, + "grad_norm": 0.015474073285452805, + "learning_rate": 9.207346341310744e-06, + "loss": 0.0003, + "step": 1678 + }, + { + "epoch": 0.44583112055231017, + "grad_norm": 1.151474305722994, + "learning_rate": 9.20615975957764e-06, + "loss": 0.0476, + "step": 1679 + }, + { + "epoch": 0.44609665427509293, + "grad_norm": 1.5488215701473362, + "learning_rate": 9.204972366946546e-06, + "loss": 0.2246, + "step": 1680 + }, + { + "epoch": 0.44636218799787575, + "grad_norm": 0.0018658514561948273, + "learning_rate": 9.203784163646368e-06, + "loss": 0.0, + "step": 1681 + }, + { + "epoch": 0.4466277217206585, + "grad_norm": 0.006199377876020424, + "learning_rate": 9.202595149906185e-06, + "loss": 0.0002, + "step": 1682 + }, + { + "epoch": 0.44689325544344133, + "grad_norm": 0.0004928732427429256, + "learning_rate": 9.201405325955222e-06, + "loss": 0.0, + "step": 1683 + }, + { + "epoch": 0.4471587891662241, + "grad_norm": 1.3116595608924697, + "learning_rate": 9.200214692022862e-06, + "loss": 0.1367, + "step": 1684 + }, + { + "epoch": 0.4474243228890069, + "grad_norm": 0.27607253424721656, + "learning_rate": 9.199023248338648e-06, + "loss": 0.0205, + "step": 1685 + }, + { + "epoch": 0.44768985661178967, + "grad_norm": 0.8567058953910064, + "learning_rate": 9.197830995132275e-06, + "loss": 0.0206, + "step": 1686 + }, + { + "epoch": 0.4479553903345725, + "grad_norm": 3.4163189556120814, + "learning_rate": 9.196637932633593e-06, + "loss": 0.1553, + "step": 1687 + }, + { + "epoch": 0.4482209240573553, + "grad_norm": 0.5354290085411026, + "learning_rate": 9.195444061072612e-06, + "loss": 0.0352, + "step": 1688 + }, + { + "epoch": 0.44848645778013807, + "grad_norm": 0.24446326873599736, + "learning_rate": 9.194249380679498e-06, + "loss": 0.0205, + "step": 1689 + }, + { + "epoch": 0.4487519915029209, + "grad_norm": 2.035095641007781, + "learning_rate": 9.19305389168457e-06, + "loss": 0.0688, + "step": 1690 + }, + { + "epoch": 0.44901752522570365, + "grad_norm": 0.0008471435364572292, + "learning_rate": 9.191857594318304e-06, + "loss": 0.0, + "step": 1691 + }, + { + "epoch": 0.44928305894848647, + "grad_norm": 3.829454052397207, + "learning_rate": 9.190660488811332e-06, + "loss": 0.5664, + "step": 1692 + }, + { + "epoch": 0.44954859267126923, + "grad_norm": 1.1644189483281735, + "learning_rate": 9.189462575394443e-06, + "loss": 0.1113, + "step": 1693 + }, + { + "epoch": 0.44981412639405205, + "grad_norm": 0.6076990567042442, + "learning_rate": 9.188263854298578e-06, + "loss": 0.0476, + "step": 1694 + }, + { + "epoch": 0.4500796601168348, + "grad_norm": 0.22821643461061178, + "learning_rate": 9.187064325754838e-06, + "loss": 0.0229, + "step": 1695 + }, + { + "epoch": 0.45034519383961763, + "grad_norm": 0.517920311097021, + "learning_rate": 9.185863989994479e-06, + "loss": 0.0432, + "step": 1696 + }, + { + "epoch": 0.45061072756240045, + "grad_norm": 1.6939168341092024, + "learning_rate": 9.18466284724891e-06, + "loss": 0.1035, + "step": 1697 + }, + { + "epoch": 0.4508762612851832, + "grad_norm": 0.2848964703780022, + "learning_rate": 9.183460897749697e-06, + "loss": 0.0229, + "step": 1698 + }, + { + "epoch": 0.45114179500796603, + "grad_norm": 0.7200198808392976, + "learning_rate": 9.182258141728563e-06, + "loss": 0.1455, + "step": 1699 + }, + { + "epoch": 0.4514073287307488, + "grad_norm": 0.6382191231542441, + "learning_rate": 9.181054579417384e-06, + "loss": 0.0525, + "step": 1700 + }, + { + "epoch": 0.4516728624535316, + "grad_norm": 0.7739371440317542, + "learning_rate": 9.179850211048193e-06, + "loss": 0.1279, + "step": 1701 + }, + { + "epoch": 0.45193839617631437, + "grad_norm": 0.05514363697004764, + "learning_rate": 9.178645036853177e-06, + "loss": 0.0017, + "step": 1702 + }, + { + "epoch": 0.4522039298990972, + "grad_norm": 0.7474045741866328, + "learning_rate": 9.177439057064684e-06, + "loss": 0.1279, + "step": 1703 + }, + { + "epoch": 0.45246946362187995, + "grad_norm": 0.25830195125237543, + "learning_rate": 9.176232271915206e-06, + "loss": 0.0255, + "step": 1704 + }, + { + "epoch": 0.45273499734466277, + "grad_norm": 0.08723998970372462, + "learning_rate": 9.175024681637401e-06, + "loss": 0.0031, + "step": 1705 + }, + { + "epoch": 0.4530005310674456, + "grad_norm": 0.7724681106268992, + "learning_rate": 9.173816286464077e-06, + "loss": 0.0576, + "step": 1706 + }, + { + "epoch": 0.45326606479022835, + "grad_norm": 0.2962473622329823, + "learning_rate": 9.172607086628197e-06, + "loss": 0.0286, + "step": 1707 + }, + { + "epoch": 0.45353159851301117, + "grad_norm": 0.5255810866203752, + "learning_rate": 9.171397082362885e-06, + "loss": 0.0286, + "step": 1708 + }, + { + "epoch": 0.45379713223579393, + "grad_norm": 0.527170789292327, + "learning_rate": 9.170186273901411e-06, + "loss": 0.0286, + "step": 1709 + }, + { + "epoch": 0.45406266595857675, + "grad_norm": 0.38467745191339814, + "learning_rate": 9.168974661477206e-06, + "loss": 0.0206, + "step": 1710 + }, + { + "epoch": 0.4543281996813595, + "grad_norm": 0.29047817939552817, + "learning_rate": 9.167762245323855e-06, + "loss": 0.0354, + "step": 1711 + }, + { + "epoch": 0.45459373340414233, + "grad_norm": 0.834962595717288, + "learning_rate": 9.166549025675098e-06, + "loss": 0.0576, + "step": 1712 + }, + { + "epoch": 0.4548592671269251, + "grad_norm": 0.8325104633372797, + "learning_rate": 9.165335002764828e-06, + "loss": 0.0889, + "step": 1713 + }, + { + "epoch": 0.4551248008497079, + "grad_norm": 0.29406829163167375, + "learning_rate": 9.164120176827097e-06, + "loss": 0.0317, + "step": 1714 + }, + { + "epoch": 0.45539033457249073, + "grad_norm": 0.8674015109840324, + "learning_rate": 9.162904548096103e-06, + "loss": 0.1191, + "step": 1715 + }, + { + "epoch": 0.4556558682952735, + "grad_norm": 0.2980629365474458, + "learning_rate": 9.16168811680621e-06, + "loss": 0.0354, + "step": 1716 + }, + { + "epoch": 0.4559214020180563, + "grad_norm": 0.236925754930231, + "learning_rate": 9.160470883191934e-06, + "loss": 0.0286, + "step": 1717 + }, + { + "epoch": 0.4561869357408391, + "grad_norm": 0.21775609808125498, + "learning_rate": 9.159252847487939e-06, + "loss": 0.0255, + "step": 1718 + }, + { + "epoch": 0.4564524694636219, + "grad_norm": 0.05357758947992938, + "learning_rate": 9.158034009929046e-06, + "loss": 0.0019, + "step": 1719 + }, + { + "epoch": 0.45671800318640465, + "grad_norm": 0.7295630149525557, + "learning_rate": 9.156814370750238e-06, + "loss": 0.0525, + "step": 1720 + }, + { + "epoch": 0.45698353690918747, + "grad_norm": 0.06168367786760944, + "learning_rate": 9.155593930186644e-06, + "loss": 0.0024, + "step": 1721 + }, + { + "epoch": 0.45724907063197023, + "grad_norm": 1.351984217373161, + "learning_rate": 9.15437268847355e-06, + "loss": 0.0815, + "step": 1722 + }, + { + "epoch": 0.45751460435475305, + "grad_norm": 0.10497376351174942, + "learning_rate": 9.153150645846399e-06, + "loss": 0.0036, + "step": 1723 + }, + { + "epoch": 0.45778013807753587, + "grad_norm": 0.18119641858710658, + "learning_rate": 9.151927802540783e-06, + "loss": 0.0045, + "step": 1724 + }, + { + "epoch": 0.45804567180031863, + "grad_norm": 0.9837813441272035, + "learning_rate": 9.150704158792456e-06, + "loss": 0.0688, + "step": 1725 + }, + { + "epoch": 0.45831120552310145, + "grad_norm": 0.8532756218779173, + "learning_rate": 9.14947971483732e-06, + "loss": 0.0576, + "step": 1726 + }, + { + "epoch": 0.4585767392458842, + "grad_norm": 0.6798270406461934, + "learning_rate": 9.148254470911434e-06, + "loss": 0.1113, + "step": 1727 + }, + { + "epoch": 0.45884227296866703, + "grad_norm": 0.33188951195468896, + "learning_rate": 9.14702842725101e-06, + "loss": 0.0317, + "step": 1728 + }, + { + "epoch": 0.4591078066914498, + "grad_norm": 0.9547380036001121, + "learning_rate": 9.145801584092415e-06, + "loss": 0.0688, + "step": 1729 + }, + { + "epoch": 0.4593733404142326, + "grad_norm": 1.121208129214251, + "learning_rate": 9.144573941672169e-06, + "loss": 0.0317, + "step": 1730 + }, + { + "epoch": 0.4596388741370154, + "grad_norm": 0.018532306902938646, + "learning_rate": 9.143345500226948e-06, + "loss": 0.0006, + "step": 1731 + }, + { + "epoch": 0.4599044078597982, + "grad_norm": 0.6486549833157388, + "learning_rate": 9.14211625999358e-06, + "loss": 0.1279, + "step": 1732 + }, + { + "epoch": 0.460169941582581, + "grad_norm": 0.2387504469567946, + "learning_rate": 9.14088622120905e-06, + "loss": 0.0082, + "step": 1733 + }, + { + "epoch": 0.4604354753053638, + "grad_norm": 2.1639487181238826, + "learning_rate": 9.139655384110494e-06, + "loss": 0.1641, + "step": 1734 + }, + { + "epoch": 0.4607010090281466, + "grad_norm": 0.6344741659386001, + "learning_rate": 9.1384237489352e-06, + "loss": 0.1113, + "step": 1735 + }, + { + "epoch": 0.46096654275092935, + "grad_norm": 0.6929723214303033, + "learning_rate": 9.13719131592062e-06, + "loss": 0.0889, + "step": 1736 + }, + { + "epoch": 0.46123207647371217, + "grad_norm": 0.38365443204668837, + "learning_rate": 9.135958085304345e-06, + "loss": 0.0432, + "step": 1737 + }, + { + "epoch": 0.46149761019649493, + "grad_norm": 0.4578266373131249, + "learning_rate": 9.134724057324132e-06, + "loss": 0.0476, + "step": 1738 + }, + { + "epoch": 0.46176314391927775, + "grad_norm": 0.3533084463962813, + "learning_rate": 9.133489232217882e-06, + "loss": 0.0103, + "step": 1739 + }, + { + "epoch": 0.4620286776420605, + "grad_norm": 0.3101849952354096, + "learning_rate": 9.13225361022366e-06, + "loss": 0.0391, + "step": 1740 + }, + { + "epoch": 0.46229421136484333, + "grad_norm": 1.3062414453586326, + "learning_rate": 9.131017191579679e-06, + "loss": 0.1035, + "step": 1741 + }, + { + "epoch": 0.46255974508762615, + "grad_norm": 0.28116099076278234, + "learning_rate": 9.129779976524303e-06, + "loss": 0.0354, + "step": 1742 + }, + { + "epoch": 0.4628252788104089, + "grad_norm": 0.31019250951818894, + "learning_rate": 9.128541965296051e-06, + "loss": 0.0391, + "step": 1743 + }, + { + "epoch": 0.46309081253319173, + "grad_norm": 1.032116360292696, + "learning_rate": 9.127303158133602e-06, + "loss": 0.0752, + "step": 1744 + }, + { + "epoch": 0.4633563462559745, + "grad_norm": 0.6052388800690558, + "learning_rate": 9.126063555275782e-06, + "loss": 0.0576, + "step": 1745 + }, + { + "epoch": 0.4636218799787573, + "grad_norm": 0.04310564970851126, + "learning_rate": 9.12482315696157e-06, + "loss": 0.0015, + "step": 1746 + }, + { + "epoch": 0.4638874137015401, + "grad_norm": 0.7380757773397368, + "learning_rate": 9.123581963430102e-06, + "loss": 0.063, + "step": 1747 + }, + { + "epoch": 0.4641529474243229, + "grad_norm": 0.8495163211366225, + "learning_rate": 9.122339974920664e-06, + "loss": 0.0688, + "step": 1748 + }, + { + "epoch": 0.46441848114710566, + "grad_norm": 0.17953121426797672, + "learning_rate": 9.121097191672697e-06, + "loss": 0.0082, + "step": 1749 + }, + { + "epoch": 0.4646840148698885, + "grad_norm": 0.6531547283200894, + "learning_rate": 9.119853613925796e-06, + "loss": 0.1367, + "step": 1750 + }, + { + "epoch": 0.4649495485926713, + "grad_norm": 1.3480427367731387, + "learning_rate": 9.118609241919707e-06, + "loss": 0.0688, + "step": 1751 + }, + { + "epoch": 0.46521508231545405, + "grad_norm": 0.5732194141376209, + "learning_rate": 9.11736407589433e-06, + "loss": 0.1191, + "step": 1752 + }, + { + "epoch": 0.46548061603823687, + "grad_norm": 0.615762764256231, + "learning_rate": 9.11611811608972e-06, + "loss": 0.1113, + "step": 1753 + }, + { + "epoch": 0.46574614976101963, + "grad_norm": 0.7635883205751591, + "learning_rate": 9.11487136274608e-06, + "loss": 0.0693, + "step": 1754 + }, + { + "epoch": 0.46601168348380245, + "grad_norm": 0.5733271972141396, + "learning_rate": 9.113623816103775e-06, + "loss": 0.0889, + "step": 1755 + }, + { + "epoch": 0.4662772172065852, + "grad_norm": 0.23390235066147239, + "learning_rate": 9.112375476403313e-06, + "loss": 0.0317, + "step": 1756 + }, + { + "epoch": 0.46654275092936803, + "grad_norm": 0.627686461965525, + "learning_rate": 9.111126343885359e-06, + "loss": 0.0752, + "step": 1757 + }, + { + "epoch": 0.4668082846521508, + "grad_norm": 1.1290944290448022, + "learning_rate": 9.109876418790733e-06, + "loss": 0.1113, + "step": 1758 + }, + { + "epoch": 0.4670738183749336, + "grad_norm": 0.5093788174002621, + "learning_rate": 9.108625701360403e-06, + "loss": 0.0957, + "step": 1759 + }, + { + "epoch": 0.46733935209771643, + "grad_norm": 0.5396898654005425, + "learning_rate": 9.107374191835496e-06, + "loss": 0.0752, + "step": 1760 + }, + { + "epoch": 0.4676048858204992, + "grad_norm": 0.31226491832256825, + "learning_rate": 9.106121890457289e-06, + "loss": 0.0432, + "step": 1761 + }, + { + "epoch": 0.467870419543282, + "grad_norm": 0.5714355485781667, + "learning_rate": 9.104868797467206e-06, + "loss": 0.0286, + "step": 1762 + }, + { + "epoch": 0.4681359532660648, + "grad_norm": 0.2957638989705724, + "learning_rate": 9.103614913106832e-06, + "loss": 0.0432, + "step": 1763 + }, + { + "epoch": 0.4684014869888476, + "grad_norm": 0.292588618562859, + "learning_rate": 9.1023602376179e-06, + "loss": 0.0432, + "step": 1764 + }, + { + "epoch": 0.46866702071163036, + "grad_norm": 0.33751587589467996, + "learning_rate": 9.101104771242297e-06, + "loss": 0.0525, + "step": 1765 + }, + { + "epoch": 0.4689325544344132, + "grad_norm": 0.394806774742472, + "learning_rate": 9.099848514222063e-06, + "loss": 0.0579, + "step": 1766 + }, + { + "epoch": 0.469198088157196, + "grad_norm": 0.30284921769372897, + "learning_rate": 9.098591466799388e-06, + "loss": 0.0479, + "step": 1767 + }, + { + "epoch": 0.46946362187997875, + "grad_norm": 0.06033720701590172, + "learning_rate": 9.097333629216618e-06, + "loss": 0.0019, + "step": 1768 + }, + { + "epoch": 0.46972915560276157, + "grad_norm": 0.6433340242166607, + "learning_rate": 9.096075001716245e-06, + "loss": 0.082, + "step": 1769 + }, + { + "epoch": 0.46999468932554433, + "grad_norm": 0.47188366036526186, + "learning_rate": 9.094815584540922e-06, + "loss": 0.063, + "step": 1770 + }, + { + "epoch": 0.47026022304832715, + "grad_norm": 0.48271838829532376, + "learning_rate": 9.093555377933449e-06, + "loss": 0.0957, + "step": 1771 + }, + { + "epoch": 0.4705257567711099, + "grad_norm": 0.47768086513551467, + "learning_rate": 9.092294382136778e-06, + "loss": 0.0889, + "step": 1772 + }, + { + "epoch": 0.47079129049389273, + "grad_norm": 0.28143225047541554, + "learning_rate": 9.091032597394012e-06, + "loss": 0.0432, + "step": 1773 + }, + { + "epoch": 0.4710568242166755, + "grad_norm": 0.3048274878976099, + "learning_rate": 9.089770023948413e-06, + "loss": 0.0132, + "step": 1774 + }, + { + "epoch": 0.4713223579394583, + "grad_norm": 0.4934823224302044, + "learning_rate": 9.088506662043387e-06, + "loss": 0.0688, + "step": 1775 + }, + { + "epoch": 0.47158789166224113, + "grad_norm": 0.4824507932270506, + "learning_rate": 9.087242511922497e-06, + "loss": 0.0889, + "step": 1776 + }, + { + "epoch": 0.4718534253850239, + "grad_norm": 2.06956586384691, + "learning_rate": 9.085977573829453e-06, + "loss": 0.1367, + "step": 1777 + }, + { + "epoch": 0.4721189591078067, + "grad_norm": 0.16024574451614348, + "learning_rate": 9.084711848008122e-06, + "loss": 0.0065, + "step": 1778 + }, + { + "epoch": 0.4723844928305895, + "grad_norm": 0.392020268202222, + "learning_rate": 9.083445334702525e-06, + "loss": 0.0317, + "step": 1779 + }, + { + "epoch": 0.4726500265533723, + "grad_norm": 0.2638470201441458, + "learning_rate": 9.082178034156823e-06, + "loss": 0.0391, + "step": 1780 + }, + { + "epoch": 0.47291556027615506, + "grad_norm": 0.35572403768500227, + "learning_rate": 9.080909946615344e-06, + "loss": 0.0525, + "step": 1781 + }, + { + "epoch": 0.4731810939989379, + "grad_norm": 0.1322872189240058, + "learning_rate": 9.079641072322555e-06, + "loss": 0.0045, + "step": 1782 + }, + { + "epoch": 0.47344662772172064, + "grad_norm": 0.4683993996526506, + "learning_rate": 9.078371411523085e-06, + "loss": 0.0889, + "step": 1783 + }, + { + "epoch": 0.47371216144450345, + "grad_norm": 0.05846428329547589, + "learning_rate": 9.077100964461705e-06, + "loss": 0.0017, + "step": 1784 + }, + { + "epoch": 0.4739776951672863, + "grad_norm": 0.4950840613938997, + "learning_rate": 9.075829731383342e-06, + "loss": 0.082, + "step": 1785 + }, + { + "epoch": 0.47424322889006904, + "grad_norm": 0.409989728413392, + "learning_rate": 9.07455771253308e-06, + "loss": 0.0525, + "step": 1786 + }, + { + "epoch": 0.47450876261285185, + "grad_norm": 0.540891443973615, + "learning_rate": 9.073284908156145e-06, + "loss": 0.0476, + "step": 1787 + }, + { + "epoch": 0.4747742963356346, + "grad_norm": 0.013696281162261802, + "learning_rate": 9.07201131849792e-06, + "loss": 0.0004, + "step": 1788 + }, + { + "epoch": 0.47503983005841743, + "grad_norm": 0.23943299662174408, + "learning_rate": 9.070736943803937e-06, + "loss": 0.0104, + "step": 1789 + }, + { + "epoch": 0.4753053637812002, + "grad_norm": 0.4895048851505992, + "learning_rate": 9.069461784319882e-06, + "loss": 0.0432, + "step": 1790 + }, + { + "epoch": 0.475570897503983, + "grad_norm": 0.10004712225856449, + "learning_rate": 9.068185840291588e-06, + "loss": 0.0025, + "step": 1791 + }, + { + "epoch": 0.4758364312267658, + "grad_norm": 0.159183886065426, + "learning_rate": 9.066909111965046e-06, + "loss": 0.0041, + "step": 1792 + }, + { + "epoch": 0.4761019649495486, + "grad_norm": 0.5751229116192297, + "learning_rate": 9.065631599586392e-06, + "loss": 0.0752, + "step": 1793 + }, + { + "epoch": 0.4763674986723314, + "grad_norm": 0.2988995897362944, + "learning_rate": 9.064353303401912e-06, + "loss": 0.0432, + "step": 1794 + }, + { + "epoch": 0.4766330323951142, + "grad_norm": 0.42552366259459906, + "learning_rate": 9.063074223658052e-06, + "loss": 0.0693, + "step": 1795 + }, + { + "epoch": 0.476898566117897, + "grad_norm": 1.4837334028705988, + "learning_rate": 9.061794360601401e-06, + "loss": 0.1113, + "step": 1796 + }, + { + "epoch": 0.47716409984067976, + "grad_norm": 0.6192662336206393, + "learning_rate": 9.060513714478701e-06, + "loss": 0.0752, + "step": 1797 + }, + { + "epoch": 0.4774296335634626, + "grad_norm": 0.321229007267076, + "learning_rate": 9.059232285536848e-06, + "loss": 0.0476, + "step": 1798 + }, + { + "epoch": 0.47769516728624534, + "grad_norm": 0.2890108799839838, + "learning_rate": 9.057950074022884e-06, + "loss": 0.0432, + "step": 1799 + }, + { + "epoch": 0.47796070100902815, + "grad_norm": 0.07149010496675735, + "learning_rate": 9.056667080184004e-06, + "loss": 0.0025, + "step": 1800 + }, + { + "epoch": 0.4782262347318109, + "grad_norm": 0.39389094255878604, + "learning_rate": 9.055383304267555e-06, + "loss": 0.0258, + "step": 1801 + }, + { + "epoch": 0.47849176845459374, + "grad_norm": 0.33439676094633747, + "learning_rate": 9.054098746521033e-06, + "loss": 0.0479, + "step": 1802 + }, + { + "epoch": 0.47875730217737655, + "grad_norm": 0.4026126537880819, + "learning_rate": 9.052813407192087e-06, + "loss": 0.0525, + "step": 1803 + }, + { + "epoch": 0.4790228359001593, + "grad_norm": 0.13383065541461073, + "learning_rate": 9.051527286528514e-06, + "loss": 0.004, + "step": 1804 + }, + { + "epoch": 0.47928836962294213, + "grad_norm": 0.45913557201091415, + "learning_rate": 9.050240384778265e-06, + "loss": 0.0525, + "step": 1805 + }, + { + "epoch": 0.4795539033457249, + "grad_norm": 0.32449255612491684, + "learning_rate": 9.048952702189437e-06, + "loss": 0.0131, + "step": 1806 + }, + { + "epoch": 0.4798194370685077, + "grad_norm": 0.23468227496683847, + "learning_rate": 9.047664239010281e-06, + "loss": 0.0317, + "step": 1807 + }, + { + "epoch": 0.4800849707912905, + "grad_norm": 0.4626085648207346, + "learning_rate": 9.046374995489197e-06, + "loss": 0.0576, + "step": 1808 + }, + { + "epoch": 0.4803505045140733, + "grad_norm": 0.2436457613182265, + "learning_rate": 9.045084971874738e-06, + "loss": 0.0354, + "step": 1809 + }, + { + "epoch": 0.48061603823685606, + "grad_norm": 0.6559558699245994, + "learning_rate": 9.043794168415603e-06, + "loss": 0.0206, + "step": 1810 + }, + { + "epoch": 0.4808815719596389, + "grad_norm": 1.0632918592533553, + "learning_rate": 9.042502585360644e-06, + "loss": 0.0815, + "step": 1811 + }, + { + "epoch": 0.4811471056824217, + "grad_norm": 0.24478371121705456, + "learning_rate": 9.041210222958863e-06, + "loss": 0.0286, + "step": 1812 + }, + { + "epoch": 0.48141263940520446, + "grad_norm": 1.4444380543418034, + "learning_rate": 9.039917081459415e-06, + "loss": 0.0986, + "step": 1813 + }, + { + "epoch": 0.4816781731279873, + "grad_norm": 0.7045925020545539, + "learning_rate": 9.038623161111597e-06, + "loss": 0.1035, + "step": 1814 + }, + { + "epoch": 0.48194370685077004, + "grad_norm": 0.6197386448287133, + "learning_rate": 9.037328462164866e-06, + "loss": 0.1113, + "step": 1815 + }, + { + "epoch": 0.48220924057355286, + "grad_norm": 0.1857729974346812, + "learning_rate": 9.036032984868825e-06, + "loss": 0.0206, + "step": 1816 + }, + { + "epoch": 0.4824747742963356, + "grad_norm": 0.24507656947520706, + "learning_rate": 9.034736729473223e-06, + "loss": 0.0286, + "step": 1817 + }, + { + "epoch": 0.48274030801911844, + "grad_norm": 0.18927032479525302, + "learning_rate": 9.033439696227966e-06, + "loss": 0.0258, + "step": 1818 + }, + { + "epoch": 0.4830058417419012, + "grad_norm": 0.06063697384987017, + "learning_rate": 9.032141885383106e-06, + "loss": 0.0017, + "step": 1819 + }, + { + "epoch": 0.483271375464684, + "grad_norm": 0.22908240400686025, + "learning_rate": 9.030843297188842e-06, + "loss": 0.0286, + "step": 1820 + }, + { + "epoch": 0.48353690918746683, + "grad_norm": 0.029382370073099087, + "learning_rate": 9.029543931895533e-06, + "loss": 0.0007, + "step": 1821 + }, + { + "epoch": 0.4838024429102496, + "grad_norm": 4.222452192211459, + "learning_rate": 9.028243789753676e-06, + "loss": 0.2129, + "step": 1822 + }, + { + "epoch": 0.4840679766330324, + "grad_norm": 0.2060789874561621, + "learning_rate": 9.026942871013927e-06, + "loss": 0.0057, + "step": 1823 + }, + { + "epoch": 0.4843335103558152, + "grad_norm": 0.6395860695177051, + "learning_rate": 9.025641175927084e-06, + "loss": 0.0525, + "step": 1824 + }, + { + "epoch": 0.484599044078598, + "grad_norm": 0.6417176946489609, + "learning_rate": 9.024338704744102e-06, + "loss": 0.0476, + "step": 1825 + }, + { + "epoch": 0.48486457780138076, + "grad_norm": 0.1480376723553451, + "learning_rate": 9.02303545771608e-06, + "loss": 0.0184, + "step": 1826 + }, + { + "epoch": 0.4851301115241636, + "grad_norm": 1.0502927779842128, + "learning_rate": 9.021731435094267e-06, + "loss": 0.1367, + "step": 1827 + }, + { + "epoch": 0.48539564524694634, + "grad_norm": 0.4100889619332376, + "learning_rate": 9.020426637130069e-06, + "loss": 0.0184, + "step": 1828 + }, + { + "epoch": 0.48566117896972916, + "grad_norm": 1.263025388192655, + "learning_rate": 9.01912106407503e-06, + "loss": 0.1035, + "step": 1829 + }, + { + "epoch": 0.485926712692512, + "grad_norm": 0.6821914603573469, + "learning_rate": 9.01781471618085e-06, + "loss": 0.1641, + "step": 1830 + }, + { + "epoch": 0.48619224641529474, + "grad_norm": 0.1830730022933736, + "learning_rate": 9.016507593699379e-06, + "loss": 0.0231, + "step": 1831 + }, + { + "epoch": 0.48645778013807756, + "grad_norm": 0.1500688928562703, + "learning_rate": 9.015199696882614e-06, + "loss": 0.0184, + "step": 1832 + }, + { + "epoch": 0.4867233138608603, + "grad_norm": 0.19372694562652476, + "learning_rate": 9.013891025982704e-06, + "loss": 0.0231, + "step": 1833 + }, + { + "epoch": 0.48698884758364314, + "grad_norm": 0.7571475485333456, + "learning_rate": 9.01258158125194e-06, + "loss": 0.1279, + "step": 1834 + }, + { + "epoch": 0.4872543813064259, + "grad_norm": 0.6523517733514138, + "learning_rate": 9.011271362942774e-06, + "loss": 0.1279, + "step": 1835 + }, + { + "epoch": 0.4875199150292087, + "grad_norm": 0.2681899167683771, + "learning_rate": 9.009960371307798e-06, + "loss": 0.0286, + "step": 1836 + }, + { + "epoch": 0.4877854487519915, + "grad_norm": 0.7058528797659482, + "learning_rate": 9.008648606599755e-06, + "loss": 0.1455, + "step": 1837 + }, + { + "epoch": 0.4880509824747743, + "grad_norm": 0.12442443775061221, + "learning_rate": 9.007336069071537e-06, + "loss": 0.0039, + "step": 1838 + }, + { + "epoch": 0.4883165161975571, + "grad_norm": 0.20140444385699593, + "learning_rate": 9.00602275897619e-06, + "loss": 0.0286, + "step": 1839 + }, + { + "epoch": 0.4885820499203399, + "grad_norm": 0.5373289060422746, + "learning_rate": 9.0047086765669e-06, + "loss": 0.1113, + "step": 1840 + }, + { + "epoch": 0.4888475836431227, + "grad_norm": 0.705010285755384, + "learning_rate": 9.003393822097005e-06, + "loss": 0.0576, + "step": 1841 + }, + { + "epoch": 0.48911311736590546, + "grad_norm": 0.20375285524703615, + "learning_rate": 9.002078195820002e-06, + "loss": 0.0286, + "step": 1842 + }, + { + "epoch": 0.4893786510886883, + "grad_norm": 0.2973123126251027, + "learning_rate": 9.000761797989521e-06, + "loss": 0.0391, + "step": 1843 + }, + { + "epoch": 0.48964418481147104, + "grad_norm": 0.5261967824579288, + "learning_rate": 8.99944462885935e-06, + "loss": 0.0391, + "step": 1844 + }, + { + "epoch": 0.48990971853425386, + "grad_norm": 0.641071008582391, + "learning_rate": 8.998126688683423e-06, + "loss": 0.0957, + "step": 1845 + }, + { + "epoch": 0.4901752522570366, + "grad_norm": 2.8556542247006456, + "learning_rate": 8.996807977715824e-06, + "loss": 0.1836, + "step": 1846 + }, + { + "epoch": 0.49044078597981944, + "grad_norm": 0.21588840632781164, + "learning_rate": 8.995488496210785e-06, + "loss": 0.0317, + "step": 1847 + }, + { + "epoch": 0.49070631970260226, + "grad_norm": 0.5200930577904636, + "learning_rate": 8.994168244422687e-06, + "loss": 0.0957, + "step": 1848 + }, + { + "epoch": 0.490971853425385, + "grad_norm": 0.5946617117384666, + "learning_rate": 8.992847222606059e-06, + "loss": 0.0815, + "step": 1849 + }, + { + "epoch": 0.49123738714816784, + "grad_norm": 0.5329049241163124, + "learning_rate": 8.991525431015575e-06, + "loss": 0.1035, + "step": 1850 + }, + { + "epoch": 0.4915029208709506, + "grad_norm": 0.13421039859666778, + "learning_rate": 8.990202869906067e-06, + "loss": 0.0038, + "step": 1851 + }, + { + "epoch": 0.4917684545937334, + "grad_norm": 0.21205384524729126, + "learning_rate": 8.988879539532502e-06, + "loss": 0.0035, + "step": 1852 + }, + { + "epoch": 0.4920339883165162, + "grad_norm": 0.3961762407408791, + "learning_rate": 8.987555440150009e-06, + "loss": 0.0576, + "step": 1853 + }, + { + "epoch": 0.492299522039299, + "grad_norm": 0.5124722957876611, + "learning_rate": 8.986230572013856e-06, + "loss": 0.0688, + "step": 1854 + }, + { + "epoch": 0.49256505576208176, + "grad_norm": 0.31830392157316406, + "learning_rate": 8.98490493537946e-06, + "loss": 0.0476, + "step": 1855 + }, + { + "epoch": 0.4928305894848646, + "grad_norm": 0.4475338780438215, + "learning_rate": 8.983578530502391e-06, + "loss": 0.0688, + "step": 1856 + }, + { + "epoch": 0.4930961232076474, + "grad_norm": 0.5131627119793176, + "learning_rate": 8.982251357638366e-06, + "loss": 0.0432, + "step": 1857 + }, + { + "epoch": 0.49336165693043016, + "grad_norm": 0.03746990293964691, + "learning_rate": 8.980923417043241e-06, + "loss": 0.001, + "step": 1858 + }, + { + "epoch": 0.493627190653213, + "grad_norm": 0.37701188868972424, + "learning_rate": 8.979594708973034e-06, + "loss": 0.0576, + "step": 1859 + }, + { + "epoch": 0.49389272437599574, + "grad_norm": 0.3246063141561153, + "learning_rate": 8.978265233683903e-06, + "loss": 0.0476, + "step": 1860 + }, + { + "epoch": 0.49415825809877856, + "grad_norm": 0.2529206787160113, + "learning_rate": 8.976934991432153e-06, + "loss": 0.0117, + "step": 1861 + }, + { + "epoch": 0.4944237918215613, + "grad_norm": 0.4707397561581439, + "learning_rate": 8.97560398247424e-06, + "loss": 0.0479, + "step": 1862 + }, + { + "epoch": 0.49468932554434414, + "grad_norm": 0.35709255896991166, + "learning_rate": 8.974272207066767e-06, + "loss": 0.0525, + "step": 1863 + }, + { + "epoch": 0.4949548592671269, + "grad_norm": 0.35165939869522234, + "learning_rate": 8.972939665466485e-06, + "loss": 0.0525, + "step": 1864 + }, + { + "epoch": 0.4952203929899097, + "grad_norm": 0.44447870911726006, + "learning_rate": 8.971606357930293e-06, + "loss": 0.0576, + "step": 1865 + }, + { + "epoch": 0.49548592671269254, + "grad_norm": 0.01898523611296749, + "learning_rate": 8.970272284715234e-06, + "loss": 0.0005, + "step": 1866 + }, + { + "epoch": 0.4957514604354753, + "grad_norm": 0.4586388691281053, + "learning_rate": 8.968937446078505e-06, + "loss": 0.0688, + "step": 1867 + }, + { + "epoch": 0.4960169941582581, + "grad_norm": 0.35720555993159125, + "learning_rate": 8.967601842277445e-06, + "loss": 0.0525, + "step": 1868 + }, + { + "epoch": 0.4962825278810409, + "grad_norm": 0.36644379984601816, + "learning_rate": 8.966265473569545e-06, + "loss": 0.0525, + "step": 1869 + }, + { + "epoch": 0.4965480616038237, + "grad_norm": 0.011010370117200359, + "learning_rate": 8.964928340212436e-06, + "loss": 0.0003, + "step": 1870 + }, + { + "epoch": 0.49681359532660646, + "grad_norm": 0.5993583089811542, + "learning_rate": 8.963590442463907e-06, + "loss": 0.082, + "step": 1871 + }, + { + "epoch": 0.4970791290493893, + "grad_norm": 0.4181666595587938, + "learning_rate": 8.962251780581888e-06, + "loss": 0.0354, + "step": 1872 + }, + { + "epoch": 0.49734466277217204, + "grad_norm": 0.4221668256111686, + "learning_rate": 8.960912354824455e-06, + "loss": 0.0393, + "step": 1873 + }, + { + "epoch": 0.49761019649495486, + "grad_norm": 0.3977116516433932, + "learning_rate": 8.959572165449834e-06, + "loss": 0.0576, + "step": 1874 + }, + { + "epoch": 0.4978757302177377, + "grad_norm": 0.010735836440220152, + "learning_rate": 8.9582312127164e-06, + "loss": 0.0003, + "step": 1875 + }, + { + "epoch": 0.49814126394052044, + "grad_norm": 0.27808107918242736, + "learning_rate": 8.95688949688267e-06, + "loss": 0.0354, + "step": 1876 + }, + { + "epoch": 0.49840679766330326, + "grad_norm": 0.0016704416964712005, + "learning_rate": 8.955547018207311e-06, + "loss": 0.0, + "step": 1877 + }, + { + "epoch": 0.498672331386086, + "grad_norm": 0.5686188793893922, + "learning_rate": 8.954203776949141e-06, + "loss": 0.0476, + "step": 1878 + }, + { + "epoch": 0.49893786510886884, + "grad_norm": 0.39056710727137667, + "learning_rate": 8.952859773367117e-06, + "loss": 0.0525, + "step": 1879 + }, + { + "epoch": 0.4992033988316516, + "grad_norm": 0.512461057922991, + "learning_rate": 8.951515007720348e-06, + "loss": 0.0525, + "step": 1880 + }, + { + "epoch": 0.4994689325544344, + "grad_norm": 0.07275683481894793, + "learning_rate": 8.950169480268089e-06, + "loss": 0.0022, + "step": 1881 + }, + { + "epoch": 0.4997344662772172, + "grad_norm": 0.5906991505102981, + "learning_rate": 8.948823191269742e-06, + "loss": 0.063, + "step": 1882 + }, + { + "epoch": 0.5, + "grad_norm": 0.6168155161242778, + "learning_rate": 8.947476140984856e-06, + "loss": 0.0688, + "step": 1883 + }, + { + "epoch": 0.5002655337227828, + "grad_norm": 0.5204598871312067, + "learning_rate": 8.946128329673126e-06, + "loss": 0.0752, + "step": 1884 + }, + { + "epoch": 0.5005310674455656, + "grad_norm": 1.8571138477453206, + "learning_rate": 8.944779757594397e-06, + "loss": 0.2129, + "step": 1885 + }, + { + "epoch": 0.5007966011683483, + "grad_norm": 0.49090367613182434, + "learning_rate": 8.943430425008653e-06, + "loss": 0.063, + "step": 1886 + }, + { + "epoch": 0.5010621348911312, + "grad_norm": 0.2501068284298111, + "learning_rate": 8.942080332176031e-06, + "loss": 0.0286, + "step": 1887 + }, + { + "epoch": 0.501327668613914, + "grad_norm": 0.19234786774639814, + "learning_rate": 8.940729479356814e-06, + "loss": 0.0073, + "step": 1888 + }, + { + "epoch": 0.5015932023366968, + "grad_norm": 1.5237924486409677, + "learning_rate": 8.93937786681143e-06, + "loss": 0.0889, + "step": 1889 + }, + { + "epoch": 0.5018587360594795, + "grad_norm": 0.5459929255095108, + "learning_rate": 8.938025494800454e-06, + "loss": 0.063, + "step": 1890 + }, + { + "epoch": 0.5021242697822623, + "grad_norm": 0.25938046401016135, + "learning_rate": 8.936672363584607e-06, + "loss": 0.0354, + "step": 1891 + }, + { + "epoch": 0.5023898035050451, + "grad_norm": 0.32388794224012313, + "learning_rate": 8.935318473424756e-06, + "loss": 0.0432, + "step": 1892 + }, + { + "epoch": 0.502655337227828, + "grad_norm": 0.526908944612307, + "learning_rate": 8.933963824581919e-06, + "loss": 0.0693, + "step": 1893 + }, + { + "epoch": 0.5029208709506108, + "grad_norm": 0.27900639395149873, + "learning_rate": 8.932608417317253e-06, + "loss": 0.0092, + "step": 1894 + }, + { + "epoch": 0.5031864046733935, + "grad_norm": 0.3445447813472986, + "learning_rate": 8.931252251892061e-06, + "loss": 0.0354, + "step": 1895 + }, + { + "epoch": 0.5034519383961763, + "grad_norm": 0.44933449962511285, + "learning_rate": 8.929895328567804e-06, + "loss": 0.0476, + "step": 1896 + }, + { + "epoch": 0.5037174721189591, + "grad_norm": 0.03865608508633661, + "learning_rate": 8.928537647606074e-06, + "loss": 0.001, + "step": 1897 + }, + { + "epoch": 0.5039830058417419, + "grad_norm": 0.20536476890220312, + "learning_rate": 8.92717920926862e-06, + "loss": 0.0258, + "step": 1898 + }, + { + "epoch": 0.5042485395645246, + "grad_norm": 0.2603825837594336, + "learning_rate": 8.92582001381733e-06, + "loss": 0.0317, + "step": 1899 + }, + { + "epoch": 0.5045140732873075, + "grad_norm": 0.664354927635177, + "learning_rate": 8.924460061514241e-06, + "loss": 0.1553, + "step": 1900 + }, + { + "epoch": 0.5047796070100903, + "grad_norm": 0.1417729671988901, + "learning_rate": 8.92309935262154e-06, + "loss": 0.0165, + "step": 1901 + }, + { + "epoch": 0.5050451407328731, + "grad_norm": 0.7573282597398386, + "learning_rate": 8.92173788740155e-06, + "loss": 0.0391, + "step": 1902 + }, + { + "epoch": 0.5053106744556559, + "grad_norm": 0.6651934597634159, + "learning_rate": 8.920375666116749e-06, + "loss": 0.1367, + "step": 1903 + }, + { + "epoch": 0.5055762081784386, + "grad_norm": 0.630368828497892, + "learning_rate": 8.919012689029756e-06, + "loss": 0.1279, + "step": 1904 + }, + { + "epoch": 0.5058417419012214, + "grad_norm": 0.011800849809816667, + "learning_rate": 8.917648956403338e-06, + "loss": 0.0003, + "step": 1905 + }, + { + "epoch": 0.5061072756240043, + "grad_norm": 0.1916335585910718, + "learning_rate": 8.916284468500404e-06, + "loss": 0.0231, + "step": 1906 + }, + { + "epoch": 0.5063728093467871, + "grad_norm": 0.1990089576514881, + "learning_rate": 8.914919225584014e-06, + "loss": 0.0258, + "step": 1907 + }, + { + "epoch": 0.5066383430695698, + "grad_norm": 0.42788748968525386, + "learning_rate": 8.913553227917366e-06, + "loss": 0.0165, + "step": 1908 + }, + { + "epoch": 0.5069038767923526, + "grad_norm": 0.6043691154555516, + "learning_rate": 8.912186475763815e-06, + "loss": 0.1113, + "step": 1909 + }, + { + "epoch": 0.5071694105151354, + "grad_norm": 0.20100736554436516, + "learning_rate": 8.910818969386852e-06, + "loss": 0.0258, + "step": 1910 + }, + { + "epoch": 0.5074349442379182, + "grad_norm": 0.2668002216539039, + "learning_rate": 8.909450709050115e-06, + "loss": 0.0067, + "step": 1911 + }, + { + "epoch": 0.5077004779607011, + "grad_norm": 0.25391486889680465, + "learning_rate": 8.908081695017388e-06, + "loss": 0.0286, + "step": 1912 + }, + { + "epoch": 0.5079660116834838, + "grad_norm": 0.10326740021263013, + "learning_rate": 8.906711927552601e-06, + "loss": 0.0022, + "step": 1913 + }, + { + "epoch": 0.5082315454062666, + "grad_norm": 0.7183992803108435, + "learning_rate": 8.905341406919832e-06, + "loss": 0.1455, + "step": 1914 + }, + { + "epoch": 0.5084970791290494, + "grad_norm": 0.24497092264223638, + "learning_rate": 8.903970133383297e-06, + "loss": 0.0286, + "step": 1915 + }, + { + "epoch": 0.5087626128518322, + "grad_norm": 0.5045822122027199, + "learning_rate": 8.902598107207364e-06, + "loss": 0.0476, + "step": 1916 + }, + { + "epoch": 0.5090281465746149, + "grad_norm": 0.21870695038083438, + "learning_rate": 8.901225328656543e-06, + "loss": 0.0258, + "step": 1917 + }, + { + "epoch": 0.5092936802973977, + "grad_norm": 0.006806934116766017, + "learning_rate": 8.89985179799549e-06, + "loss": 0.0002, + "step": 1918 + }, + { + "epoch": 0.5095592140201806, + "grad_norm": 0.06217815771623346, + "learning_rate": 8.898477515489003e-06, + "loss": 0.0013, + "step": 1919 + }, + { + "epoch": 0.5098247477429634, + "grad_norm": 0.15502279422567414, + "learning_rate": 8.897102481402031e-06, + "loss": 0.0036, + "step": 1920 + }, + { + "epoch": 0.5100902814657462, + "grad_norm": 0.2254494341774315, + "learning_rate": 8.895726695999663e-06, + "loss": 0.0286, + "step": 1921 + }, + { + "epoch": 0.5103558151885289, + "grad_norm": 4.048223290258338, + "learning_rate": 8.894350159547132e-06, + "loss": 0.2559, + "step": 1922 + }, + { + "epoch": 0.5106213489113117, + "grad_norm": 0.2224313022582158, + "learning_rate": 8.892972872309821e-06, + "loss": 0.0258, + "step": 1923 + }, + { + "epoch": 0.5108868826340945, + "grad_norm": 0.2649153994192695, + "learning_rate": 8.891594834553252e-06, + "loss": 0.0317, + "step": 1924 + }, + { + "epoch": 0.5111524163568774, + "grad_norm": 0.32381344159451636, + "learning_rate": 8.890216046543099e-06, + "loss": 0.0354, + "step": 1925 + }, + { + "epoch": 0.5114179500796601, + "grad_norm": 0.18911967304759786, + "learning_rate": 8.888836508545172e-06, + "loss": 0.0258, + "step": 1926 + }, + { + "epoch": 0.5116834838024429, + "grad_norm": 0.15794943866572278, + "learning_rate": 8.887456220825429e-06, + "loss": 0.0206, + "step": 1927 + }, + { + "epoch": 0.5119490175252257, + "grad_norm": 0.8298787508237971, + "learning_rate": 8.886075183649976e-06, + "loss": 0.0889, + "step": 1928 + }, + { + "epoch": 0.5122145512480085, + "grad_norm": 0.14098262352173152, + "learning_rate": 8.884693397285058e-06, + "loss": 0.0045, + "step": 1929 + }, + { + "epoch": 0.5124800849707913, + "grad_norm": 0.5112261546913817, + "learning_rate": 8.883310861997073e-06, + "loss": 0.0391, + "step": 1930 + }, + { + "epoch": 0.512745618693574, + "grad_norm": 0.0710276758168645, + "learning_rate": 8.881927578052549e-06, + "loss": 0.0014, + "step": 1931 + }, + { + "epoch": 0.5130111524163569, + "grad_norm": 0.7156148636827978, + "learning_rate": 8.880543545718173e-06, + "loss": 0.1191, + "step": 1932 + }, + { + "epoch": 0.5132766861391397, + "grad_norm": 0.7498314694324334, + "learning_rate": 8.879158765260767e-06, + "loss": 0.1279, + "step": 1933 + }, + { + "epoch": 0.5135422198619225, + "grad_norm": 0.16462904141113316, + "learning_rate": 8.877773236947302e-06, + "loss": 0.0206, + "step": 1934 + }, + { + "epoch": 0.5138077535847052, + "grad_norm": 0.8601863622602933, + "learning_rate": 8.876386961044892e-06, + "loss": 0.0752, + "step": 1935 + }, + { + "epoch": 0.514073287307488, + "grad_norm": 0.6362870475286894, + "learning_rate": 8.874999937820793e-06, + "loss": 0.1279, + "step": 1936 + }, + { + "epoch": 0.5143388210302708, + "grad_norm": 0.1624202878388607, + "learning_rate": 8.873612167542408e-06, + "loss": 0.0206, + "step": 1937 + }, + { + "epoch": 0.5146043547530537, + "grad_norm": 0.15596786935599946, + "learning_rate": 8.87222365047728e-06, + "loss": 0.0206, + "step": 1938 + }, + { + "epoch": 0.5148698884758365, + "grad_norm": 1.0578529242706238, + "learning_rate": 8.870834386893101e-06, + "loss": 0.0815, + "step": 1939 + }, + { + "epoch": 0.5151354221986192, + "grad_norm": 0.32309548985685693, + "learning_rate": 8.869444377057706e-06, + "loss": 0.0105, + "step": 1940 + }, + { + "epoch": 0.515400955921402, + "grad_norm": 0.6279202187077859, + "learning_rate": 8.868053621239072e-06, + "loss": 0.1279, + "step": 1941 + }, + { + "epoch": 0.5156664896441848, + "grad_norm": 0.29499114303717844, + "learning_rate": 8.866662119705317e-06, + "loss": 0.0391, + "step": 1942 + }, + { + "epoch": 0.5159320233669676, + "grad_norm": 0.22730083243000213, + "learning_rate": 8.865269872724708e-06, + "loss": 0.0286, + "step": 1943 + }, + { + "epoch": 0.5161975570897503, + "grad_norm": 0.8126774785685559, + "learning_rate": 8.863876880565656e-06, + "loss": 0.0286, + "step": 1944 + }, + { + "epoch": 0.5164630908125332, + "grad_norm": 0.5189987862432444, + "learning_rate": 8.862483143496713e-06, + "loss": 0.0286, + "step": 1945 + }, + { + "epoch": 0.516728624535316, + "grad_norm": 0.3079203367541235, + "learning_rate": 8.86108866178657e-06, + "loss": 0.0354, + "step": 1946 + }, + { + "epoch": 0.5169941582580988, + "grad_norm": 0.30146182888816936, + "learning_rate": 8.859693435704073e-06, + "loss": 0.0391, + "step": 1947 + }, + { + "epoch": 0.5172596919808816, + "grad_norm": 0.18766405767921446, + "learning_rate": 8.858297465518201e-06, + "loss": 0.0258, + "step": 1948 + }, + { + "epoch": 0.5175252257036643, + "grad_norm": 0.48345252694955393, + "learning_rate": 8.856900751498083e-06, + "loss": 0.0476, + "step": 1949 + }, + { + "epoch": 0.5177907594264471, + "grad_norm": 0.28999764767448305, + "learning_rate": 8.855503293912987e-06, + "loss": 0.0354, + "step": 1950 + }, + { + "epoch": 0.51805629314923, + "grad_norm": 0.301468091100063, + "learning_rate": 8.854105093032328e-06, + "loss": 0.0354, + "step": 1951 + }, + { + "epoch": 0.5183218268720128, + "grad_norm": 0.22535390470370426, + "learning_rate": 8.852706149125662e-06, + "loss": 0.0286, + "step": 1952 + }, + { + "epoch": 0.5185873605947955, + "grad_norm": 0.2907903784108349, + "learning_rate": 8.851306462462689e-06, + "loss": 0.0317, + "step": 1953 + }, + { + "epoch": 0.5188528943175783, + "grad_norm": 0.6566340180383853, + "learning_rate": 8.849906033313251e-06, + "loss": 0.0525, + "step": 1954 + }, + { + "epoch": 0.5191184280403611, + "grad_norm": 0.3414604558845798, + "learning_rate": 8.848504861947335e-06, + "loss": 0.0354, + "step": 1955 + }, + { + "epoch": 0.5193839617631439, + "grad_norm": 0.15095251988649963, + "learning_rate": 8.847102948635074e-06, + "loss": 0.0184, + "step": 1956 + }, + { + "epoch": 0.5196494954859268, + "grad_norm": 0.8850305174499467, + "learning_rate": 8.845700293646733e-06, + "loss": 0.063, + "step": 1957 + }, + { + "epoch": 0.5199150292087095, + "grad_norm": 0.18094733973006374, + "learning_rate": 8.844296897252733e-06, + "loss": 0.0039, + "step": 1958 + }, + { + "epoch": 0.5201805629314923, + "grad_norm": 0.0024453736726058376, + "learning_rate": 8.84289275972363e-06, + "loss": 0.0001, + "step": 1959 + }, + { + "epoch": 0.5204460966542751, + "grad_norm": 4.799487348660996, + "learning_rate": 8.841487881330126e-06, + "loss": 0.2891, + "step": 1960 + }, + { + "epoch": 0.5207116303770579, + "grad_norm": 0.00898819327183765, + "learning_rate": 8.840082262343064e-06, + "loss": 0.0002, + "step": 1961 + }, + { + "epoch": 0.5209771640998406, + "grad_norm": 0.11036124994527378, + "learning_rate": 8.838675903033432e-06, + "loss": 0.0117, + "step": 1962 + }, + { + "epoch": 0.5212426978226234, + "grad_norm": 0.17194701223020112, + "learning_rate": 8.837268803672358e-06, + "loss": 0.0184, + "step": 1963 + }, + { + "epoch": 0.5215082315454063, + "grad_norm": 0.19802383357405098, + "learning_rate": 8.835860964531115e-06, + "loss": 0.0206, + "step": 1964 + }, + { + "epoch": 0.5217737652681891, + "grad_norm": 0.11041893858226348, + "learning_rate": 8.834452385881121e-06, + "loss": 0.0117, + "step": 1965 + }, + { + "epoch": 0.5220392989909719, + "grad_norm": 0.6156090855902048, + "learning_rate": 8.833043067993926e-06, + "loss": 0.0233, + "step": 1966 + }, + { + "epoch": 0.5223048327137546, + "grad_norm": 0.14201533586884124, + "learning_rate": 8.831633011141237e-06, + "loss": 0.0146, + "step": 1967 + }, + { + "epoch": 0.5225703664365374, + "grad_norm": 0.09288200859356922, + "learning_rate": 8.83022221559489e-06, + "loss": 0.0092, + "step": 1968 + }, + { + "epoch": 0.5228359001593202, + "grad_norm": 0.9615613402120565, + "learning_rate": 8.828810681626876e-06, + "loss": 0.1455, + "step": 1969 + }, + { + "epoch": 0.5231014338821031, + "grad_norm": 1.3056914957728305, + "learning_rate": 8.827398409509318e-06, + "loss": 0.0476, + "step": 1970 + }, + { + "epoch": 0.5233669676048858, + "grad_norm": 0.9123728894775064, + "learning_rate": 8.825985399514488e-06, + "loss": 0.1738, + "step": 1971 + }, + { + "epoch": 0.5236325013276686, + "grad_norm": 0.5116325107103561, + "learning_rate": 8.824571651914796e-06, + "loss": 0.0354, + "step": 1972 + }, + { + "epoch": 0.5238980350504514, + "grad_norm": 1.0931368397378376, + "learning_rate": 8.823157166982794e-06, + "loss": 0.0889, + "step": 1973 + }, + { + "epoch": 0.5241635687732342, + "grad_norm": 1.5038247940641931, + "learning_rate": 8.821741944991182e-06, + "loss": 0.1191, + "step": 1974 + }, + { + "epoch": 0.524429102496017, + "grad_norm": 2.1410167922056185, + "learning_rate": 8.820325986212796e-06, + "loss": 0.0957, + "step": 1975 + }, + { + "epoch": 0.5246946362187997, + "grad_norm": 1.4124271244242217, + "learning_rate": 8.818909290920617e-06, + "loss": 0.063, + "step": 1976 + }, + { + "epoch": 0.5249601699415826, + "grad_norm": 0.044914449959208055, + "learning_rate": 8.817491859387765e-06, + "loss": 0.001, + "step": 1977 + }, + { + "epoch": 0.5252257036643654, + "grad_norm": 0.7192623272990704, + "learning_rate": 8.816073691887506e-06, + "loss": 0.1455, + "step": 1978 + }, + { + "epoch": 0.5254912373871482, + "grad_norm": 0.4995422114884727, + "learning_rate": 8.814654788693246e-06, + "loss": 0.0391, + "step": 1979 + }, + { + "epoch": 0.5257567711099309, + "grad_norm": 0.2947556592020414, + "learning_rate": 8.813235150078532e-06, + "loss": 0.0286, + "step": 1980 + }, + { + "epoch": 0.5260223048327137, + "grad_norm": 0.2865413479084726, + "learning_rate": 8.811814776317056e-06, + "loss": 0.0255, + "step": 1981 + }, + { + "epoch": 0.5262878385554965, + "grad_norm": 2.2341427898489776, + "learning_rate": 8.810393667682645e-06, + "loss": 0.0815, + "step": 1982 + }, + { + "epoch": 0.5265533722782794, + "grad_norm": 0.1433749071821932, + "learning_rate": 8.808971824449276e-06, + "loss": 0.0165, + "step": 1983 + }, + { + "epoch": 0.5268189060010622, + "grad_norm": 1.1251534229014521, + "learning_rate": 8.80754924689106e-06, + "loss": 0.0815, + "step": 1984 + }, + { + "epoch": 0.5270844397238449, + "grad_norm": 0.9046870543343221, + "learning_rate": 8.806125935282258e-06, + "loss": 0.0231, + "step": 1985 + }, + { + "epoch": 0.5273499734466277, + "grad_norm": 2.59520506547441, + "learning_rate": 8.804701889897261e-06, + "loss": 0.1641, + "step": 1986 + }, + { + "epoch": 0.5276155071694105, + "grad_norm": 1.1343653355959855, + "learning_rate": 8.803277111010613e-06, + "loss": 0.0688, + "step": 1987 + }, + { + "epoch": 0.5278810408921933, + "grad_norm": 1.0935958000724277, + "learning_rate": 8.801851598896996e-06, + "loss": 0.0293, + "step": 1988 + }, + { + "epoch": 0.528146574614976, + "grad_norm": 0.13718069103720956, + "learning_rate": 8.800425353831227e-06, + "loss": 0.0165, + "step": 1989 + }, + { + "epoch": 0.5284121083377589, + "grad_norm": 0.1382767894506031, + "learning_rate": 8.798998376088273e-06, + "loss": 0.0165, + "step": 1990 + }, + { + "epoch": 0.5286776420605417, + "grad_norm": 0.6992769860574649, + "learning_rate": 8.797570665943236e-06, + "loss": 0.1553, + "step": 1991 + }, + { + "epoch": 0.5289431757833245, + "grad_norm": 0.7959070902740847, + "learning_rate": 8.796142223671362e-06, + "loss": 0.1279, + "step": 1992 + }, + { + "epoch": 0.5292087095061073, + "grad_norm": 0.003383773068786995, + "learning_rate": 8.79471304954804e-06, + "loss": 0.0001, + "step": 1993 + }, + { + "epoch": 0.52947424322889, + "grad_norm": 1.6911270754687433, + "learning_rate": 8.793283143848796e-06, + "loss": 0.1035, + "step": 1994 + }, + { + "epoch": 0.5297397769516728, + "grad_norm": 0.25896993348139163, + "learning_rate": 8.791852506849301e-06, + "loss": 0.0258, + "step": 1995 + }, + { + "epoch": 0.5300053106744557, + "grad_norm": 0.1990072374637002, + "learning_rate": 8.790421138825362e-06, + "loss": 0.0229, + "step": 1996 + }, + { + "epoch": 0.5302708443972385, + "grad_norm": 0.12393114254257008, + "learning_rate": 8.788989040052932e-06, + "loss": 0.0146, + "step": 1997 + }, + { + "epoch": 0.5305363781200212, + "grad_norm": 0.1350871742151799, + "learning_rate": 8.787556210808101e-06, + "loss": 0.0165, + "step": 1998 + }, + { + "epoch": 0.530801911842804, + "grad_norm": 0.13938411153478383, + "learning_rate": 8.786122651367104e-06, + "loss": 0.0184, + "step": 1999 + }, + { + "epoch": 0.5310674455655868, + "grad_norm": 1.1760763074160063, + "learning_rate": 8.784688362006311e-06, + "loss": 0.1191, + "step": 2000 + }, + { + "epoch": 0.5313329792883696, + "grad_norm": 0.6616934500771192, + "learning_rate": 8.78325334300224e-06, + "loss": 0.1279, + "step": 2001 + }, + { + "epoch": 0.5315985130111525, + "grad_norm": 0.7731533530500331, + "learning_rate": 8.781817594631543e-06, + "loss": 0.1113, + "step": 2002 + }, + { + "epoch": 0.5318640467339352, + "grad_norm": 0.27309771991113013, + "learning_rate": 8.780381117171016e-06, + "loss": 0.0317, + "step": 2003 + }, + { + "epoch": 0.532129580456718, + "grad_norm": 0.677262816977555, + "learning_rate": 8.778943910897596e-06, + "loss": 0.0286, + "step": 2004 + }, + { + "epoch": 0.5323951141795008, + "grad_norm": 0.18989059432792887, + "learning_rate": 8.777505976088358e-06, + "loss": 0.0229, + "step": 2005 + }, + { + "epoch": 0.5326606479022836, + "grad_norm": 0.05337912083898583, + "learning_rate": 8.77606731302052e-06, + "loss": 0.0017, + "step": 2006 + }, + { + "epoch": 0.5329261816250663, + "grad_norm": 0.22329320874949543, + "learning_rate": 8.774627921971437e-06, + "loss": 0.0045, + "step": 2007 + }, + { + "epoch": 0.5331917153478491, + "grad_norm": 0.16455132622686192, + "learning_rate": 8.773187803218608e-06, + "loss": 0.0206, + "step": 2008 + }, + { + "epoch": 0.533457249070632, + "grad_norm": 0.033736099336577306, + "learning_rate": 8.771746957039673e-06, + "loss": 0.0007, + "step": 2009 + }, + { + "epoch": 0.5337227827934148, + "grad_norm": 0.8415678405292931, + "learning_rate": 8.770305383712407e-06, + "loss": 0.0957, + "step": 2010 + }, + { + "epoch": 0.5339883165161976, + "grad_norm": 0.22765177824118393, + "learning_rate": 8.76886308351473e-06, + "loss": 0.0258, + "step": 2011 + }, + { + "epoch": 0.5342538502389803, + "grad_norm": 0.7255653292332117, + "learning_rate": 8.7674200567247e-06, + "loss": 0.0432, + "step": 2012 + }, + { + "epoch": 0.5345193839617631, + "grad_norm": 0.10663870399531976, + "learning_rate": 8.765976303620517e-06, + "loss": 0.0024, + "step": 2013 + }, + { + "epoch": 0.534784917684546, + "grad_norm": 0.2157170543692871, + "learning_rate": 8.764531824480517e-06, + "loss": 0.0255, + "step": 2014 + }, + { + "epoch": 0.5350504514073288, + "grad_norm": 0.4013523554650016, + "learning_rate": 8.76308661958318e-06, + "loss": 0.0165, + "step": 2015 + }, + { + "epoch": 0.5353159851301115, + "grad_norm": 0.3385113937455839, + "learning_rate": 8.761640689207123e-06, + "loss": 0.0082, + "step": 2016 + }, + { + "epoch": 0.5355815188528943, + "grad_norm": 0.3134276462605137, + "learning_rate": 8.76019403363111e-06, + "loss": 0.0131, + "step": 2017 + }, + { + "epoch": 0.5358470525756771, + "grad_norm": 0.30575513730825443, + "learning_rate": 8.75874665313403e-06, + "loss": 0.0073, + "step": 2018 + }, + { + "epoch": 0.5361125862984599, + "grad_norm": 0.9262205830035705, + "learning_rate": 8.757298547994926e-06, + "loss": 0.0432, + "step": 2019 + }, + { + "epoch": 0.5363781200212427, + "grad_norm": 1.2366295546779627, + "learning_rate": 8.755849718492976e-06, + "loss": 0.0432, + "step": 2020 + }, + { + "epoch": 0.5366436537440255, + "grad_norm": 0.49152856889421803, + "learning_rate": 8.754400164907496e-06, + "loss": 0.0391, + "step": 2021 + }, + { + "epoch": 0.5369091874668083, + "grad_norm": 0.6153749556789703, + "learning_rate": 8.752949887517944e-06, + "loss": 0.0525, + "step": 2022 + }, + { + "epoch": 0.5371747211895911, + "grad_norm": 0.14657570392551136, + "learning_rate": 8.751498886603915e-06, + "loss": 0.0031, + "step": 2023 + }, + { + "epoch": 0.5374402549123739, + "grad_norm": 0.38713636673654483, + "learning_rate": 8.750047162445147e-06, + "loss": 0.0432, + "step": 2024 + }, + { + "epoch": 0.5377057886351566, + "grad_norm": 0.6870428668776961, + "learning_rate": 8.748594715321512e-06, + "loss": 0.0231, + "step": 2025 + }, + { + "epoch": 0.5379713223579394, + "grad_norm": 0.602183204441049, + "learning_rate": 8.747141545513027e-06, + "loss": 0.0432, + "step": 2026 + }, + { + "epoch": 0.5382368560807222, + "grad_norm": 1.4888276568449648, + "learning_rate": 8.745687653299845e-06, + "loss": 0.0752, + "step": 2027 + }, + { + "epoch": 0.5385023898035051, + "grad_norm": 0.22905710433753276, + "learning_rate": 8.744233038962262e-06, + "loss": 0.0229, + "step": 2028 + }, + { + "epoch": 0.5387679235262879, + "grad_norm": 0.7380308747793056, + "learning_rate": 8.742777702780706e-06, + "loss": 0.0229, + "step": 2029 + }, + { + "epoch": 0.5390334572490706, + "grad_norm": 1.2111164471948226, + "learning_rate": 8.741321645035752e-06, + "loss": 0.0815, + "step": 2030 + }, + { + "epoch": 0.5392989909718534, + "grad_norm": 1.13254890511383, + "learning_rate": 8.739864866008109e-06, + "loss": 0.0576, + "step": 2031 + }, + { + "epoch": 0.5395645246946362, + "grad_norm": 0.001744893988161318, + "learning_rate": 8.738407365978627e-06, + "loss": 0.0, + "step": 2032 + }, + { + "epoch": 0.539830058417419, + "grad_norm": 0.06121747725398596, + "learning_rate": 8.736949145228295e-06, + "loss": 0.001, + "step": 2033 + }, + { + "epoch": 0.5400955921402018, + "grad_norm": 0.6783861302917257, + "learning_rate": 8.735490204038242e-06, + "loss": 0.0255, + "step": 2034 + }, + { + "epoch": 0.5403611258629846, + "grad_norm": 4.932370224031872, + "learning_rate": 8.734030542689734e-06, + "loss": 0.2031, + "step": 2035 + }, + { + "epoch": 0.5406266595857674, + "grad_norm": 0.28349689325473554, + "learning_rate": 8.732570161464176e-06, + "loss": 0.0045, + "step": 2036 + }, + { + "epoch": 0.5408921933085502, + "grad_norm": 1.259517175334726, + "learning_rate": 8.731109060643112e-06, + "loss": 0.0957, + "step": 2037 + }, + { + "epoch": 0.541157727031333, + "grad_norm": 0.2881652220542512, + "learning_rate": 8.729647240508226e-06, + "loss": 0.0255, + "step": 2038 + }, + { + "epoch": 0.5414232607541157, + "grad_norm": 0.9087748131424184, + "learning_rate": 8.728184701341337e-06, + "loss": 0.0255, + "step": 2039 + }, + { + "epoch": 0.5416887944768985, + "grad_norm": 3.9657508815796123, + "learning_rate": 8.726721443424409e-06, + "loss": 0.6367, + "step": 2040 + }, + { + "epoch": 0.5419543281996814, + "grad_norm": 0.23733721141682648, + "learning_rate": 8.72525746703954e-06, + "loss": 0.0184, + "step": 2041 + }, + { + "epoch": 0.5422198619224642, + "grad_norm": 0.01624858899221418, + "learning_rate": 8.723792772468966e-06, + "loss": 0.0003, + "step": 2042 + }, + { + "epoch": 0.5424853956452469, + "grad_norm": 0.012896074699826891, + "learning_rate": 8.722327359995064e-06, + "loss": 0.0003, + "step": 2043 + }, + { + "epoch": 0.5427509293680297, + "grad_norm": 0.325368424549595, + "learning_rate": 8.720861229900347e-06, + "loss": 0.0064, + "step": 2044 + }, + { + "epoch": 0.5430164630908125, + "grad_norm": 2.4859883593066705, + "learning_rate": 8.71939438246747e-06, + "loss": 0.2031, + "step": 2045 + }, + { + "epoch": 0.5432819968135953, + "grad_norm": 0.30869148166609, + "learning_rate": 8.71792681797922e-06, + "loss": 0.0255, + "step": 2046 + }, + { + "epoch": 0.5435475305363782, + "grad_norm": 4.382545228036755, + "learning_rate": 8.71645853671853e-06, + "loss": 0.2988, + "step": 2047 + }, + { + "epoch": 0.5438130642591609, + "grad_norm": 2.5548527085433963, + "learning_rate": 8.714989538968464e-06, + "loss": 0.1641, + "step": 2048 + }, + { + "epoch": 0.5440785979819437, + "grad_norm": 0.0027537949623614687, + "learning_rate": 8.713519825012231e-06, + "loss": 0.0001, + "step": 2049 + }, + { + "epoch": 0.5443441317047265, + "grad_norm": 1.3038177577080179, + "learning_rate": 8.712049395133172e-06, + "loss": 0.0815, + "step": 2050 + }, + { + "epoch": 0.5446096654275093, + "grad_norm": 4.598790327635417, + "learning_rate": 8.71057824961477e-06, + "loss": 0.5781, + "step": 2051 + }, + { + "epoch": 0.544875199150292, + "grad_norm": 0.004905399729976801, + "learning_rate": 8.70910638874064e-06, + "loss": 0.0001, + "step": 2052 + }, + { + "epoch": 0.5451407328730749, + "grad_norm": 0.2592825010638339, + "learning_rate": 8.707633812794547e-06, + "loss": 0.0229, + "step": 2053 + }, + { + "epoch": 0.5454062665958577, + "grad_norm": 0.3820277888345581, + "learning_rate": 8.70616052206038e-06, + "loss": 0.0092, + "step": 2054 + }, + { + "epoch": 0.5456718003186405, + "grad_norm": 0.41100419560987933, + "learning_rate": 8.704686516822177e-06, + "loss": 0.0286, + "step": 2055 + }, + { + "epoch": 0.5459373340414233, + "grad_norm": 1.2656095819426663, + "learning_rate": 8.703211797364104e-06, + "loss": 0.0815, + "step": 2056 + }, + { + "epoch": 0.546202867764206, + "grad_norm": 0.42870667264919965, + "learning_rate": 8.701736363970474e-06, + "loss": 0.0184, + "step": 2057 + }, + { + "epoch": 0.5464684014869888, + "grad_norm": 1.0734083956006322, + "learning_rate": 8.700260216925728e-06, + "loss": 0.0688, + "step": 2058 + }, + { + "epoch": 0.5467339352097716, + "grad_norm": 0.1663957263533079, + "learning_rate": 8.698783356514455e-06, + "loss": 0.0165, + "step": 2059 + }, + { + "epoch": 0.5469994689325545, + "grad_norm": 0.41414126523670897, + "learning_rate": 8.697305783021374e-06, + "loss": 0.0131, + "step": 2060 + }, + { + "epoch": 0.5472650026553372, + "grad_norm": 0.5023614570515106, + "learning_rate": 8.695827496731343e-06, + "loss": 0.0354, + "step": 2061 + }, + { + "epoch": 0.54753053637812, + "grad_norm": 0.6393980791532451, + "learning_rate": 8.694348497929359e-06, + "loss": 0.0286, + "step": 2062 + }, + { + "epoch": 0.5477960701009028, + "grad_norm": 0.00788561980486304, + "learning_rate": 8.692868786900554e-06, + "loss": 0.0002, + "step": 2063 + }, + { + "epoch": 0.5480616038236856, + "grad_norm": 0.8424202170379343, + "learning_rate": 8.691388363930203e-06, + "loss": 0.1279, + "step": 2064 + }, + { + "epoch": 0.5483271375464684, + "grad_norm": 1.013948410030326, + "learning_rate": 8.689907229303708e-06, + "loss": 0.0688, + "step": 2065 + }, + { + "epoch": 0.5485926712692512, + "grad_norm": 0.2766437307140973, + "learning_rate": 8.68842538330662e-06, + "loss": 0.0229, + "step": 2066 + }, + { + "epoch": 0.548858204992034, + "grad_norm": 0.03899192665636356, + "learning_rate": 8.686942826224619e-06, + "loss": 0.0015, + "step": 2067 + }, + { + "epoch": 0.5491237387148168, + "grad_norm": 0.8867421054880262, + "learning_rate": 8.685459558343524e-06, + "loss": 0.0576, + "step": 2068 + }, + { + "epoch": 0.5493892724375996, + "grad_norm": 0.1233273132090829, + "learning_rate": 8.683975579949292e-06, + "loss": 0.0103, + "step": 2069 + }, + { + "epoch": 0.5496548061603823, + "grad_norm": 0.13672178667567692, + "learning_rate": 8.682490891328016e-06, + "loss": 0.0131, + "step": 2070 + }, + { + "epoch": 0.5499203398831651, + "grad_norm": 0.3141209250056832, + "learning_rate": 8.681005492765927e-06, + "loss": 0.0229, + "step": 2071 + }, + { + "epoch": 0.550185873605948, + "grad_norm": 0.861110394506879, + "learning_rate": 8.679519384549395e-06, + "loss": 0.0576, + "step": 2072 + }, + { + "epoch": 0.5504514073287308, + "grad_norm": 0.9913746488834646, + "learning_rate": 8.67803256696492e-06, + "loss": 0.0576, + "step": 2073 + }, + { + "epoch": 0.5507169410515136, + "grad_norm": 0.2627049070617153, + "learning_rate": 8.676545040299145e-06, + "loss": 0.0229, + "step": 2074 + }, + { + "epoch": 0.5509824747742963, + "grad_norm": 0.04664339351382908, + "learning_rate": 8.675056804838847e-06, + "loss": 0.0011, + "step": 2075 + }, + { + "epoch": 0.5512480084970791, + "grad_norm": 0.1373923823262272, + "learning_rate": 8.673567860870937e-06, + "loss": 0.0131, + "step": 2076 + }, + { + "epoch": 0.5515135422198619, + "grad_norm": 0.3233819127103764, + "learning_rate": 8.672078208682473e-06, + "loss": 0.0206, + "step": 2077 + }, + { + "epoch": 0.5517790759426447, + "grad_norm": 0.09966376082868723, + "learning_rate": 8.670587848560636e-06, + "loss": 0.0092, + "step": 2078 + }, + { + "epoch": 0.5520446096654275, + "grad_norm": 0.2549472728510907, + "learning_rate": 8.669096780792754e-06, + "loss": 0.0082, + "step": 2079 + }, + { + "epoch": 0.5523101433882103, + "grad_norm": 0.7710193109762611, + "learning_rate": 8.667605005666286e-06, + "loss": 0.0525, + "step": 2080 + }, + { + "epoch": 0.5525756771109931, + "grad_norm": 0.38075606771719833, + "learning_rate": 8.666112523468827e-06, + "loss": 0.0317, + "step": 2081 + }, + { + "epoch": 0.5528412108337759, + "grad_norm": 0.15386855761387094, + "learning_rate": 8.664619334488112e-06, + "loss": 0.0131, + "step": 2082 + }, + { + "epoch": 0.5531067445565587, + "grad_norm": 0.11117491845230147, + "learning_rate": 8.66312543901201e-06, + "loss": 0.0092, + "step": 2083 + }, + { + "epoch": 0.5533722782793414, + "grad_norm": 1.1195459170916962, + "learning_rate": 8.661630837328525e-06, + "loss": 0.2344, + "step": 2084 + }, + { + "epoch": 0.5536378120021243, + "grad_norm": 0.32352450410495975, + "learning_rate": 8.660135529725799e-06, + "loss": 0.0255, + "step": 2085 + }, + { + "epoch": 0.5539033457249071, + "grad_norm": 0.12323656291923073, + "learning_rate": 8.658639516492111e-06, + "loss": 0.0103, + "step": 2086 + }, + { + "epoch": 0.5541688794476899, + "grad_norm": 0.11415731216801826, + "learning_rate": 8.657142797915874e-06, + "loss": 0.0092, + "step": 2087 + }, + { + "epoch": 0.5544344131704726, + "grad_norm": 0.08186088709910397, + "learning_rate": 8.655645374285637e-06, + "loss": 0.0072, + "step": 2088 + }, + { + "epoch": 0.5546999468932554, + "grad_norm": 0.015454829814049737, + "learning_rate": 8.654147245890088e-06, + "loss": 0.0005, + "step": 2089 + }, + { + "epoch": 0.5549654806160382, + "grad_norm": 1.059692943955725, + "learning_rate": 8.652648413018044e-06, + "loss": 0.2246, + "step": 2090 + }, + { + "epoch": 0.555231014338821, + "grad_norm": 0.03690003839966559, + "learning_rate": 8.651148875958466e-06, + "loss": 0.0012, + "step": 2091 + }, + { + "epoch": 0.5554965480616039, + "grad_norm": 0.11719210561106906, + "learning_rate": 8.649648635000447e-06, + "loss": 0.0092, + "step": 2092 + }, + { + "epoch": 0.5557620817843866, + "grad_norm": 0.08489127541015637, + "learning_rate": 8.648147690433213e-06, + "loss": 0.0036, + "step": 2093 + }, + { + "epoch": 0.5560276155071694, + "grad_norm": 0.9434503374390149, + "learning_rate": 8.646646042546132e-06, + "loss": 0.1113, + "step": 2094 + }, + { + "epoch": 0.5562931492299522, + "grad_norm": 0.427301227161439, + "learning_rate": 8.6451436916287e-06, + "loss": 0.0391, + "step": 2095 + }, + { + "epoch": 0.556558682952735, + "grad_norm": 1.0224252017062094, + "learning_rate": 8.643640637970556e-06, + "loss": 0.063, + "step": 2096 + }, + { + "epoch": 0.5568242166755177, + "grad_norm": 0.1150187794129378, + "learning_rate": 8.642136881861472e-06, + "loss": 0.0045, + "step": 2097 + }, + { + "epoch": 0.5570897503983006, + "grad_norm": 0.01732138797421572, + "learning_rate": 8.640632423591349e-06, + "loss": 0.0005, + "step": 2098 + }, + { + "epoch": 0.5573552841210834, + "grad_norm": 0.004903477088047995, + "learning_rate": 8.639127263450234e-06, + "loss": 0.0001, + "step": 2099 + }, + { + "epoch": 0.5576208178438662, + "grad_norm": 0.12711273905747025, + "learning_rate": 8.6376214017283e-06, + "loss": 0.0115, + "step": 2100 + }, + { + "epoch": 0.557886351566649, + "grad_norm": 0.8577770773652588, + "learning_rate": 8.636114838715861e-06, + "loss": 0.1455, + "step": 2101 + }, + { + "epoch": 0.5581518852894317, + "grad_norm": 0.9126835598455695, + "learning_rate": 8.634607574703369e-06, + "loss": 0.0525, + "step": 2102 + }, + { + "epoch": 0.5584174190122145, + "grad_norm": 0.1320448286140073, + "learning_rate": 8.6330996099814e-06, + "loss": 0.0131, + "step": 2103 + }, + { + "epoch": 0.5586829527349974, + "grad_norm": 0.22776840834216264, + "learning_rate": 8.631590944840675e-06, + "loss": 0.0065, + "step": 2104 + }, + { + "epoch": 0.5589484864577802, + "grad_norm": 0.13201035714579087, + "learning_rate": 8.630081579572045e-06, + "loss": 0.0115, + "step": 2105 + }, + { + "epoch": 0.5592140201805629, + "grad_norm": 0.8421160097756943, + "learning_rate": 8.628571514466502e-06, + "loss": 0.1641, + "step": 2106 + }, + { + "epoch": 0.5594795539033457, + "grad_norm": 0.2790262747927227, + "learning_rate": 8.627060749815166e-06, + "loss": 0.0117, + "step": 2107 + }, + { + "epoch": 0.5597450876261285, + "grad_norm": 0.1729515717612833, + "learning_rate": 8.625549285909293e-06, + "loss": 0.0184, + "step": 2108 + }, + { + "epoch": 0.5600106213489113, + "grad_norm": 0.8945100123702472, + "learning_rate": 8.624037123040277e-06, + "loss": 0.2129, + "step": 2109 + }, + { + "epoch": 0.5602761550716941, + "grad_norm": 0.30409414787442957, + "learning_rate": 8.622524261499647e-06, + "loss": 0.0317, + "step": 2110 + }, + { + "epoch": 0.5605416887944769, + "grad_norm": 1.2600993505634053, + "learning_rate": 8.62101070157906e-06, + "loss": 0.0752, + "step": 2111 + }, + { + "epoch": 0.5608072225172597, + "grad_norm": 1.8640917424801848, + "learning_rate": 8.61949644357032e-06, + "loss": 0.2031, + "step": 2112 + }, + { + "epoch": 0.5610727562400425, + "grad_norm": 0.6593238623006484, + "learning_rate": 8.61798148776535e-06, + "loss": 0.0354, + "step": 2113 + }, + { + "epoch": 0.5613382899628253, + "grad_norm": 0.17561073733578697, + "learning_rate": 8.61646583445622e-06, + "loss": 0.0035, + "step": 2114 + }, + { + "epoch": 0.561603823685608, + "grad_norm": 0.2821405199749007, + "learning_rate": 8.61494948393513e-06, + "loss": 0.0317, + "step": 2115 + }, + { + "epoch": 0.5618693574083908, + "grad_norm": 0.1640700760499823, + "learning_rate": 8.613432436494415e-06, + "loss": 0.0206, + "step": 2116 + }, + { + "epoch": 0.5621348911311737, + "grad_norm": 0.6564138720838959, + "learning_rate": 8.61191469242654e-06, + "loss": 0.1035, + "step": 2117 + }, + { + "epoch": 0.5624004248539565, + "grad_norm": 0.5201171855272044, + "learning_rate": 8.610396252024113e-06, + "loss": 0.0258, + "step": 2118 + }, + { + "epoch": 0.5626659585767393, + "grad_norm": 0.007938281478945021, + "learning_rate": 8.608877115579869e-06, + "loss": 0.0003, + "step": 2119 + }, + { + "epoch": 0.562931492299522, + "grad_norm": 0.24162113659286918, + "learning_rate": 8.607357283386679e-06, + "loss": 0.0286, + "step": 2120 + }, + { + "epoch": 0.5631970260223048, + "grad_norm": 0.6330645535291699, + "learning_rate": 8.60583675573755e-06, + "loss": 0.1367, + "step": 2121 + }, + { + "epoch": 0.5634625597450876, + "grad_norm": 0.2450184582586985, + "learning_rate": 8.604315532925621e-06, + "loss": 0.0286, + "step": 2122 + }, + { + "epoch": 0.5637280934678705, + "grad_norm": 2.494398940429751, + "learning_rate": 8.602793615244168e-06, + "loss": 0.2988, + "step": 2123 + }, + { + "epoch": 0.5639936271906533, + "grad_norm": 0.2632842879772224, + "learning_rate": 8.601271002986595e-06, + "loss": 0.0354, + "step": 2124 + }, + { + "epoch": 0.564259160913436, + "grad_norm": 0.2164033572005904, + "learning_rate": 8.599747696446446e-06, + "loss": 0.0258, + "step": 2125 + }, + { + "epoch": 0.5645246946362188, + "grad_norm": 0.8791129482922673, + "learning_rate": 8.598223695917399e-06, + "loss": 0.063, + "step": 2126 + }, + { + "epoch": 0.5647902283590016, + "grad_norm": 0.7755228024230928, + "learning_rate": 8.596699001693257e-06, + "loss": 0.0889, + "step": 2127 + }, + { + "epoch": 0.5650557620817844, + "grad_norm": 0.5713169656759178, + "learning_rate": 8.595173614067966e-06, + "loss": 0.1191, + "step": 2128 + }, + { + "epoch": 0.5653212958045671, + "grad_norm": 0.18740816405332078, + "learning_rate": 8.593647533335607e-06, + "loss": 0.0258, + "step": 2129 + }, + { + "epoch": 0.56558682952735, + "grad_norm": 0.1864308174911974, + "learning_rate": 8.592120759790383e-06, + "loss": 0.0258, + "step": 2130 + }, + { + "epoch": 0.5658523632501328, + "grad_norm": 0.6821360941555628, + "learning_rate": 8.590593293726645e-06, + "loss": 0.0576, + "step": 2131 + }, + { + "epoch": 0.5661178969729156, + "grad_norm": 0.17495074237376623, + "learning_rate": 8.589065135438865e-06, + "loss": 0.0231, + "step": 2132 + }, + { + "epoch": 0.5663834306956984, + "grad_norm": 0.4377515188707458, + "learning_rate": 8.587536285221656e-06, + "loss": 0.032, + "step": 2133 + }, + { + "epoch": 0.5666489644184811, + "grad_norm": 0.579097271563448, + "learning_rate": 8.586006743369764e-06, + "loss": 0.1279, + "step": 2134 + }, + { + "epoch": 0.5669144981412639, + "grad_norm": 0.4253655518011528, + "learning_rate": 8.584476510178064e-06, + "loss": 0.0288, + "step": 2135 + }, + { + "epoch": 0.5671800318640468, + "grad_norm": 0.44830913561071484, + "learning_rate": 8.582945585941567e-06, + "loss": 0.0476, + "step": 2136 + }, + { + "epoch": 0.5674455655868296, + "grad_norm": 0.6481998360700997, + "learning_rate": 8.581413970955419e-06, + "loss": 0.0752, + "step": 2137 + }, + { + "epoch": 0.5677110993096123, + "grad_norm": 0.22662930010658439, + "learning_rate": 8.579881665514895e-06, + "loss": 0.0317, + "step": 2138 + }, + { + "epoch": 0.5679766330323951, + "grad_norm": 0.6425464798565367, + "learning_rate": 8.578348669915405e-06, + "loss": 0.0476, + "step": 2139 + }, + { + "epoch": 0.5682421667551779, + "grad_norm": 0.28726291694500367, + "learning_rate": 8.576814984452498e-06, + "loss": 0.0165, + "step": 2140 + }, + { + "epoch": 0.5685077004779607, + "grad_norm": 0.2344756723718212, + "learning_rate": 8.575280609421846e-06, + "loss": 0.0317, + "step": 2141 + }, + { + "epoch": 0.5687732342007435, + "grad_norm": 0.27673896137102727, + "learning_rate": 8.573745545119258e-06, + "loss": 0.0354, + "step": 2142 + }, + { + "epoch": 0.5690387679235263, + "grad_norm": 0.36067324570104875, + "learning_rate": 8.572209791840677e-06, + "loss": 0.0391, + "step": 2143 + }, + { + "epoch": 0.5693043016463091, + "grad_norm": 0.051997888485216644, + "learning_rate": 8.570673349882179e-06, + "loss": 0.0022, + "step": 2144 + }, + { + "epoch": 0.5695698353690919, + "grad_norm": 1.6191393419205136, + "learning_rate": 8.56913621953997e-06, + "loss": 0.1113, + "step": 2145 + }, + { + "epoch": 0.5698353690918747, + "grad_norm": 1.119824340190587, + "learning_rate": 8.567598401110394e-06, + "loss": 0.0815, + "step": 2146 + }, + { + "epoch": 0.5701009028146574, + "grad_norm": 0.19181446020051932, + "learning_rate": 8.566059894889923e-06, + "loss": 0.0073, + "step": 2147 + }, + { + "epoch": 0.5703664365374402, + "grad_norm": 0.61179439455525, + "learning_rate": 8.564520701175158e-06, + "loss": 0.0957, + "step": 2148 + }, + { + "epoch": 0.570631970260223, + "grad_norm": 0.639613191746644, + "learning_rate": 8.562980820262845e-06, + "loss": 0.1367, + "step": 2149 + }, + { + "epoch": 0.5708975039830059, + "grad_norm": 0.3691709216725773, + "learning_rate": 8.561440252449849e-06, + "loss": 0.0147, + "step": 2150 + }, + { + "epoch": 0.5711630377057887, + "grad_norm": 0.1049915464580976, + "learning_rate": 8.559898998033177e-06, + "loss": 0.0031, + "step": 2151 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 0.7435399976950415, + "learning_rate": 8.558357057309965e-06, + "loss": 0.0432, + "step": 2152 + }, + { + "epoch": 0.5716941051513542, + "grad_norm": 0.026922088502705003, + "learning_rate": 8.556814430577477e-06, + "loss": 0.0007, + "step": 2153 + }, + { + "epoch": 0.571959638874137, + "grad_norm": 0.1760746797840586, + "learning_rate": 8.555271118133115e-06, + "loss": 0.0229, + "step": 2154 + }, + { + "epoch": 0.5722251725969199, + "grad_norm": 0.5245042474519009, + "learning_rate": 8.553727120274412e-06, + "loss": 0.0957, + "step": 2155 + }, + { + "epoch": 0.5724907063197026, + "grad_norm": 0.2862448699385872, + "learning_rate": 8.552182437299033e-06, + "loss": 0.0354, + "step": 2156 + }, + { + "epoch": 0.5727562400424854, + "grad_norm": 1.0095982708225453, + "learning_rate": 8.550637069504774e-06, + "loss": 0.0752, + "step": 2157 + }, + { + "epoch": 0.5730217737652682, + "grad_norm": 2.0890779712717893, + "learning_rate": 8.549091017189566e-06, + "loss": 0.2559, + "step": 2158 + }, + { + "epoch": 0.573287307488051, + "grad_norm": 0.880394079582972, + "learning_rate": 8.547544280651464e-06, + "loss": 0.0815, + "step": 2159 + }, + { + "epoch": 0.5735528412108338, + "grad_norm": 0.7174456522388359, + "learning_rate": 8.545996860188668e-06, + "loss": 0.0525, + "step": 2160 + }, + { + "epoch": 0.5738183749336165, + "grad_norm": 0.3030739395204966, + "learning_rate": 8.544448756099495e-06, + "loss": 0.0432, + "step": 2161 + }, + { + "epoch": 0.5740839086563994, + "grad_norm": 0.7042300833775007, + "learning_rate": 8.542899968682408e-06, + "loss": 0.0354, + "step": 2162 + }, + { + "epoch": 0.5743494423791822, + "grad_norm": 0.21235492415396173, + "learning_rate": 8.54135049823599e-06, + "loss": 0.0286, + "step": 2163 + }, + { + "epoch": 0.574614976101965, + "grad_norm": 0.2177330854173654, + "learning_rate": 8.539800345058962e-06, + "loss": 0.0286, + "step": 2164 + }, + { + "epoch": 0.5748805098247477, + "grad_norm": 0.2746059066138724, + "learning_rate": 8.538249509450175e-06, + "loss": 0.0391, + "step": 2165 + }, + { + "epoch": 0.5751460435475305, + "grad_norm": 0.5279770663294744, + "learning_rate": 8.536697991708613e-06, + "loss": 0.0576, + "step": 2166 + }, + { + "epoch": 0.5754115772703133, + "grad_norm": 0.6218279587187799, + "learning_rate": 8.53514579213339e-06, + "loss": 0.032, + "step": 2167 + }, + { + "epoch": 0.5756771109930962, + "grad_norm": 0.5501404023356935, + "learning_rate": 8.533592911023751e-06, + "loss": 0.1191, + "step": 2168 + }, + { + "epoch": 0.575942644715879, + "grad_norm": 0.21915049760029717, + "learning_rate": 8.532039348679074e-06, + "loss": 0.0117, + "step": 2169 + }, + { + "epoch": 0.5762081784386617, + "grad_norm": 0.5271894798973, + "learning_rate": 8.530485105398867e-06, + "loss": 0.0391, + "step": 2170 + }, + { + "epoch": 0.5764737121614445, + "grad_norm": 0.6217404663425184, + "learning_rate": 8.52893018148277e-06, + "loss": 0.0752, + "step": 2171 + }, + { + "epoch": 0.5767392458842273, + "grad_norm": 0.2192797574894119, + "learning_rate": 8.527374577230553e-06, + "loss": 0.0317, + "step": 2172 + }, + { + "epoch": 0.5770047796070101, + "grad_norm": 0.1989895508357781, + "learning_rate": 8.52581829294212e-06, + "loss": 0.0286, + "step": 2173 + }, + { + "epoch": 0.5772703133297928, + "grad_norm": 1.2832949805865848, + "learning_rate": 8.524261328917502e-06, + "loss": 0.1367, + "step": 2174 + }, + { + "epoch": 0.5775358470525757, + "grad_norm": 0.6768681217867087, + "learning_rate": 8.522703685456866e-06, + "loss": 0.0576, + "step": 2175 + }, + { + "epoch": 0.5778013807753585, + "grad_norm": 0.3365562806686397, + "learning_rate": 8.521145362860505e-06, + "loss": 0.0432, + "step": 2176 + }, + { + "epoch": 0.5780669144981413, + "grad_norm": 0.19328080058808364, + "learning_rate": 8.519586361428847e-06, + "loss": 0.0258, + "step": 2177 + }, + { + "epoch": 0.5783324482209241, + "grad_norm": 0.23354351648232177, + "learning_rate": 8.518026681462448e-06, + "loss": 0.0082, + "step": 2178 + }, + { + "epoch": 0.5785979819437068, + "grad_norm": 0.19207994664524125, + "learning_rate": 8.516466323261995e-06, + "loss": 0.0258, + "step": 2179 + }, + { + "epoch": 0.5788635156664896, + "grad_norm": 0.5583292845399278, + "learning_rate": 8.51490528712831e-06, + "loss": 0.1035, + "step": 2180 + }, + { + "epoch": 0.5791290493892725, + "grad_norm": 0.15890903297353415, + "learning_rate": 8.513343573362338e-06, + "loss": 0.0231, + "step": 2181 + }, + { + "epoch": 0.5793945831120553, + "grad_norm": 0.5107066066900314, + "learning_rate": 8.511781182265161e-06, + "loss": 0.0579, + "step": 2182 + }, + { + "epoch": 0.579660116834838, + "grad_norm": 1.4585903468519719, + "learning_rate": 8.51021811413799e-06, + "loss": 0.1191, + "step": 2183 + }, + { + "epoch": 0.5799256505576208, + "grad_norm": 0.16712342777129816, + "learning_rate": 8.508654369282168e-06, + "loss": 0.0206, + "step": 2184 + }, + { + "epoch": 0.5801911842804036, + "grad_norm": 0.41489742139825686, + "learning_rate": 8.507089947999161e-06, + "loss": 0.0476, + "step": 2185 + }, + { + "epoch": 0.5804567180031864, + "grad_norm": 0.273137618895295, + "learning_rate": 8.505524850590576e-06, + "loss": 0.0092, + "step": 2186 + }, + { + "epoch": 0.5807222517259693, + "grad_norm": 0.2347200162315576, + "learning_rate": 8.503959077358143e-06, + "loss": 0.0286, + "step": 2187 + }, + { + "epoch": 0.580987785448752, + "grad_norm": 0.622438937823715, + "learning_rate": 8.502392628603728e-06, + "loss": 0.1455, + "step": 2188 + }, + { + "epoch": 0.5812533191715348, + "grad_norm": 0.08564336066669227, + "learning_rate": 8.500825504629318e-06, + "loss": 0.0036, + "step": 2189 + }, + { + "epoch": 0.5815188528943176, + "grad_norm": 0.5891436034256372, + "learning_rate": 8.49925770573704e-06, + "loss": 0.1455, + "step": 2190 + }, + { + "epoch": 0.5817843866171004, + "grad_norm": 0.8353500270934422, + "learning_rate": 8.497689232229146e-06, + "loss": 0.082, + "step": 2191 + }, + { + "epoch": 0.5820499203398831, + "grad_norm": 0.25497271956425765, + "learning_rate": 8.496120084408019e-06, + "loss": 0.0082, + "step": 2192 + }, + { + "epoch": 0.5823154540626659, + "grad_norm": 0.23965890480563298, + "learning_rate": 8.494550262576173e-06, + "loss": 0.0317, + "step": 2193 + }, + { + "epoch": 0.5825809877854488, + "grad_norm": 0.19777692069403172, + "learning_rate": 8.492979767036251e-06, + "loss": 0.0258, + "step": 2194 + }, + { + "epoch": 0.5828465215082316, + "grad_norm": 0.293448428931433, + "learning_rate": 8.491408598091026e-06, + "loss": 0.0147, + "step": 2195 + }, + { + "epoch": 0.5831120552310144, + "grad_norm": 0.19477847983449112, + "learning_rate": 8.4898367560434e-06, + "loss": 0.0258, + "step": 2196 + }, + { + "epoch": 0.5833775889537971, + "grad_norm": 0.1743458844990908, + "learning_rate": 8.488264241196407e-06, + "loss": 0.0231, + "step": 2197 + }, + { + "epoch": 0.5836431226765799, + "grad_norm": 0.18279184441758115, + "learning_rate": 8.48669105385321e-06, + "loss": 0.0258, + "step": 2198 + }, + { + "epoch": 0.5839086563993627, + "grad_norm": 0.599739227575356, + "learning_rate": 8.485117194317096e-06, + "loss": 0.1553, + "step": 2199 + }, + { + "epoch": 0.5841741901221456, + "grad_norm": 0.25109904569056296, + "learning_rate": 8.483542662891492e-06, + "loss": 0.0286, + "step": 2200 + }, + { + "epoch": 0.5844397238449283, + "grad_norm": 0.20109088486971746, + "learning_rate": 8.481967459879948e-06, + "loss": 0.0258, + "step": 2201 + }, + { + "epoch": 0.5847052575677111, + "grad_norm": 0.642469562172212, + "learning_rate": 8.480391585586143e-06, + "loss": 0.1035, + "step": 2202 + }, + { + "epoch": 0.5849707912904939, + "grad_norm": 0.5768178502674649, + "learning_rate": 8.478815040313886e-06, + "loss": 0.0479, + "step": 2203 + }, + { + "epoch": 0.5852363250132767, + "grad_norm": 0.6044126280560649, + "learning_rate": 8.47723782436712e-06, + "loss": 0.1455, + "step": 2204 + }, + { + "epoch": 0.5855018587360595, + "grad_norm": 0.23687678238533266, + "learning_rate": 8.475659938049912e-06, + "loss": 0.0288, + "step": 2205 + }, + { + "epoch": 0.5857673924588422, + "grad_norm": 0.6091641855112391, + "learning_rate": 8.47408138166646e-06, + "loss": 0.1035, + "step": 2206 + }, + { + "epoch": 0.5860329261816251, + "grad_norm": 0.616281223492897, + "learning_rate": 8.472502155521088e-06, + "loss": 0.1113, + "step": 2207 + }, + { + "epoch": 0.5862984599044079, + "grad_norm": 0.40423872635897945, + "learning_rate": 8.470922259918256e-06, + "loss": 0.0391, + "step": 2208 + }, + { + "epoch": 0.5865639936271907, + "grad_norm": 0.1736727031924769, + "learning_rate": 8.469341695162548e-06, + "loss": 0.0258, + "step": 2209 + }, + { + "epoch": 0.5868295273499734, + "grad_norm": 0.13539870344202742, + "learning_rate": 8.467760461558678e-06, + "loss": 0.0184, + "step": 2210 + }, + { + "epoch": 0.5870950610727562, + "grad_norm": 0.16002090679414568, + "learning_rate": 8.46617855941149e-06, + "loss": 0.0231, + "step": 2211 + }, + { + "epoch": 0.587360594795539, + "grad_norm": 0.19047823732685798, + "learning_rate": 8.464595989025954e-06, + "loss": 0.0286, + "step": 2212 + }, + { + "epoch": 0.5876261285183219, + "grad_norm": 0.15481092545082156, + "learning_rate": 8.463012750707176e-06, + "loss": 0.0206, + "step": 2213 + }, + { + "epoch": 0.5878916622411047, + "grad_norm": 0.8416456467094723, + "learning_rate": 8.46142884476038e-06, + "loss": 0.063, + "step": 2214 + }, + { + "epoch": 0.5881571959638874, + "grad_norm": 0.22093066258523367, + "learning_rate": 8.459844271490926e-06, + "loss": 0.0317, + "step": 2215 + }, + { + "epoch": 0.5884227296866702, + "grad_norm": 0.5080129690553327, + "learning_rate": 8.458259031204302e-06, + "loss": 0.0579, + "step": 2216 + }, + { + "epoch": 0.588688263409453, + "grad_norm": 0.25221536101882647, + "learning_rate": 8.456673124206122e-06, + "loss": 0.0354, + "step": 2217 + }, + { + "epoch": 0.5889537971322358, + "grad_norm": 0.14571763560262518, + "learning_rate": 8.455086550802134e-06, + "loss": 0.0206, + "step": 2218 + }, + { + "epoch": 0.5892193308550185, + "grad_norm": 0.28698494546924663, + "learning_rate": 8.453499311298205e-06, + "loss": 0.0391, + "step": 2219 + }, + { + "epoch": 0.5894848645778014, + "grad_norm": 0.14583361859520994, + "learning_rate": 8.45191140600034e-06, + "loss": 0.0206, + "step": 2220 + }, + { + "epoch": 0.5897503983005842, + "grad_norm": 0.14399595316768446, + "learning_rate": 8.450322835214669e-06, + "loss": 0.0206, + "step": 2221 + }, + { + "epoch": 0.590015932023367, + "grad_norm": 0.09343411049007874, + "learning_rate": 8.448733599247446e-06, + "loss": 0.0035, + "step": 2222 + }, + { + "epoch": 0.5902814657461498, + "grad_norm": 0.9133189350432392, + "learning_rate": 8.44714369840506e-06, + "loss": 0.0815, + "step": 2223 + }, + { + "epoch": 0.5905469994689325, + "grad_norm": 0.12625022328822275, + "learning_rate": 8.445553132994024e-06, + "loss": 0.0165, + "step": 2224 + }, + { + "epoch": 0.5908125331917153, + "grad_norm": 0.19043651961938599, + "learning_rate": 8.443961903320983e-06, + "loss": 0.0258, + "step": 2225 + }, + { + "epoch": 0.5910780669144982, + "grad_norm": 0.6085369857709387, + "learning_rate": 8.442370009692702e-06, + "loss": 0.1367, + "step": 2226 + }, + { + "epoch": 0.591343600637281, + "grad_norm": 0.11879484876974304, + "learning_rate": 8.440777452416082e-06, + "loss": 0.0165, + "step": 2227 + }, + { + "epoch": 0.5916091343600637, + "grad_norm": 0.6872949399843384, + "learning_rate": 8.43918423179815e-06, + "loss": 0.1367, + "step": 2228 + }, + { + "epoch": 0.5918746680828465, + "grad_norm": 0.18570232759964833, + "learning_rate": 8.43759034814606e-06, + "loss": 0.0231, + "step": 2229 + }, + { + "epoch": 0.5921402018056293, + "grad_norm": 0.1904395629784407, + "learning_rate": 8.435995801767091e-06, + "loss": 0.0206, + "step": 2230 + }, + { + "epoch": 0.5924057355284121, + "grad_norm": 0.5853967015932758, + "learning_rate": 8.434400592968659e-06, + "loss": 0.1553, + "step": 2231 + }, + { + "epoch": 0.592671269251195, + "grad_norm": 0.6287040917591198, + "learning_rate": 8.432804722058297e-06, + "loss": 0.1641, + "step": 2232 + }, + { + "epoch": 0.5929368029739777, + "grad_norm": 0.653631091162708, + "learning_rate": 8.43120818934367e-06, + "loss": 0.0889, + "step": 2233 + }, + { + "epoch": 0.5932023366967605, + "grad_norm": 0.6850913007908959, + "learning_rate": 8.42961099513257e-06, + "loss": 0.0148, + "step": 2234 + }, + { + "epoch": 0.5934678704195433, + "grad_norm": 0.6995956676184575, + "learning_rate": 8.42801313973292e-06, + "loss": 0.1035, + "step": 2235 + }, + { + "epoch": 0.5937334041423261, + "grad_norm": 0.29566651289459495, + "learning_rate": 8.426414623452766e-06, + "loss": 0.0354, + "step": 2236 + }, + { + "epoch": 0.5939989378651088, + "grad_norm": 0.22849451700328588, + "learning_rate": 8.424815446600283e-06, + "loss": 0.0317, + "step": 2237 + }, + { + "epoch": 0.5942644715878916, + "grad_norm": 0.17309800717753687, + "learning_rate": 8.423215609483776e-06, + "loss": 0.0258, + "step": 2238 + }, + { + "epoch": 0.5945300053106745, + "grad_norm": 0.5660889036167581, + "learning_rate": 8.42161511241167e-06, + "loss": 0.1367, + "step": 2239 + }, + { + "epoch": 0.5947955390334573, + "grad_norm": 0.5822588979724513, + "learning_rate": 8.420013955692525e-06, + "loss": 0.0889, + "step": 2240 + }, + { + "epoch": 0.5950610727562401, + "grad_norm": 0.04634852878719813, + "learning_rate": 8.418412139635026e-06, + "loss": 0.0018, + "step": 2241 + }, + { + "epoch": 0.5953266064790228, + "grad_norm": 0.8387543649127215, + "learning_rate": 8.416809664547983e-06, + "loss": 0.0752, + "step": 2242 + }, + { + "epoch": 0.5955921402018056, + "grad_norm": 0.1736573437651616, + "learning_rate": 8.415206530740331e-06, + "loss": 0.0258, + "step": 2243 + }, + { + "epoch": 0.5958576739245884, + "grad_norm": 0.5210090396241558, + "learning_rate": 8.413602738521141e-06, + "loss": 0.1279, + "step": 2244 + }, + { + "epoch": 0.5961232076473713, + "grad_norm": 0.20438778304783264, + "learning_rate": 8.411998288199604e-06, + "loss": 0.0258, + "step": 2245 + }, + { + "epoch": 0.596388741370154, + "grad_norm": 0.4701822838360082, + "learning_rate": 8.410393180085036e-06, + "loss": 0.1113, + "step": 2246 + }, + { + "epoch": 0.5966542750929368, + "grad_norm": 0.5599731810867489, + "learning_rate": 8.408787414486885e-06, + "loss": 0.0957, + "step": 2247 + }, + { + "epoch": 0.5969198088157196, + "grad_norm": 0.06536594043298097, + "learning_rate": 8.407180991714723e-06, + "loss": 0.0025, + "step": 2248 + }, + { + "epoch": 0.5971853425385024, + "grad_norm": 0.2463779894924957, + "learning_rate": 8.405573912078251e-06, + "loss": 0.0391, + "step": 2249 + }, + { + "epoch": 0.5974508762612852, + "grad_norm": 0.10228242005646583, + "learning_rate": 8.403966175887293e-06, + "loss": 0.0052, + "step": 2250 + }, + { + "epoch": 0.5977164099840679, + "grad_norm": 0.7761488948793263, + "learning_rate": 8.402357783451802e-06, + "loss": 0.0815, + "step": 2251 + }, + { + "epoch": 0.5979819437068508, + "grad_norm": 0.31228129276575595, + "learning_rate": 8.400748735081858e-06, + "loss": 0.0184, + "step": 2252 + }, + { + "epoch": 0.5982474774296336, + "grad_norm": 0.252578995679301, + "learning_rate": 8.399139031087665e-06, + "loss": 0.0393, + "step": 2253 + }, + { + "epoch": 0.5985130111524164, + "grad_norm": 0.03181331201020033, + "learning_rate": 8.397528671779557e-06, + "loss": 0.0012, + "step": 2254 + }, + { + "epoch": 0.5987785448751991, + "grad_norm": 0.26657621857496544, + "learning_rate": 8.395917657467992e-06, + "loss": 0.0432, + "step": 2255 + }, + { + "epoch": 0.5990440785979819, + "grad_norm": 0.17906719881263283, + "learning_rate": 8.394305988463552e-06, + "loss": 0.0082, + "step": 2256 + }, + { + "epoch": 0.5993096123207647, + "grad_norm": 0.4393542943754623, + "learning_rate": 8.392693665076949e-06, + "loss": 0.0432, + "step": 2257 + }, + { + "epoch": 0.5995751460435476, + "grad_norm": 0.5667520994534472, + "learning_rate": 8.391080687619023e-06, + "loss": 0.0432, + "step": 2258 + }, + { + "epoch": 0.5998406797663304, + "grad_norm": 0.0894530470851913, + "learning_rate": 8.389467056400734e-06, + "loss": 0.0036, + "step": 2259 + }, + { + "epoch": 0.6001062134891131, + "grad_norm": 0.20011922978304192, + "learning_rate": 8.387852771733169e-06, + "loss": 0.0317, + "step": 2260 + }, + { + "epoch": 0.6003717472118959, + "grad_norm": 0.4309381706686263, + "learning_rate": 8.386237833927549e-06, + "loss": 0.082, + "step": 2261 + }, + { + "epoch": 0.6006372809346787, + "grad_norm": 0.4637939531673208, + "learning_rate": 8.38462224329521e-06, + "loss": 0.0752, + "step": 2262 + }, + { + "epoch": 0.6009028146574615, + "grad_norm": 0.7587105503714413, + "learning_rate": 8.38300600014762e-06, + "loss": 0.0317, + "step": 2263 + }, + { + "epoch": 0.6011683483802442, + "grad_norm": 0.08918766275161935, + "learning_rate": 8.381389104796372e-06, + "loss": 0.004, + "step": 2264 + }, + { + "epoch": 0.6014338821030271, + "grad_norm": 0.468019724757281, + "learning_rate": 8.379771557553184e-06, + "loss": 0.0957, + "step": 2265 + }, + { + "epoch": 0.6016994158258099, + "grad_norm": 0.44806591049430067, + "learning_rate": 8.378153358729902e-06, + "loss": 0.0889, + "step": 2266 + }, + { + "epoch": 0.6019649495485927, + "grad_norm": 0.5271122724691344, + "learning_rate": 8.376534508638494e-06, + "loss": 0.063, + "step": 2267 + }, + { + "epoch": 0.6022304832713755, + "grad_norm": 0.10071896501145733, + "learning_rate": 8.374915007591053e-06, + "loss": 0.004, + "step": 2268 + }, + { + "epoch": 0.6024960169941582, + "grad_norm": 0.32084239132156583, + "learning_rate": 8.373294855899804e-06, + "loss": 0.0525, + "step": 2269 + }, + { + "epoch": 0.602761550716941, + "grad_norm": 0.18654452896183601, + "learning_rate": 8.37167405387709e-06, + "loss": 0.0073, + "step": 2270 + }, + { + "epoch": 0.6030270844397239, + "grad_norm": 0.3046879141242327, + "learning_rate": 8.370052601835385e-06, + "loss": 0.0476, + "step": 2271 + }, + { + "epoch": 0.6032926181625067, + "grad_norm": 0.3414683963401615, + "learning_rate": 8.368430500087283e-06, + "loss": 0.0525, + "step": 2272 + }, + { + "epoch": 0.6035581518852894, + "grad_norm": 0.3189001928680894, + "learning_rate": 8.366807748945509e-06, + "loss": 0.0525, + "step": 2273 + }, + { + "epoch": 0.6038236856080722, + "grad_norm": 0.4362972840162116, + "learning_rate": 8.365184348722907e-06, + "loss": 0.0576, + "step": 2274 + }, + { + "epoch": 0.604089219330855, + "grad_norm": 0.5370338346339294, + "learning_rate": 8.363560299732452e-06, + "loss": 0.0693, + "step": 2275 + }, + { + "epoch": 0.6043547530536378, + "grad_norm": 0.40550506124686464, + "learning_rate": 8.36193560228724e-06, + "loss": 0.063, + "step": 2276 + }, + { + "epoch": 0.6046202867764207, + "grad_norm": 0.02266383376611979, + "learning_rate": 8.360310256700496e-06, + "loss": 0.0008, + "step": 2277 + }, + { + "epoch": 0.6048858204992034, + "grad_norm": 0.18482353835063708, + "learning_rate": 8.358684263285566e-06, + "loss": 0.0093, + "step": 2278 + }, + { + "epoch": 0.6051513542219862, + "grad_norm": 0.7513238772641164, + "learning_rate": 8.35705762235592e-06, + "loss": 0.0752, + "step": 2279 + }, + { + "epoch": 0.605416887944769, + "grad_norm": 0.36547944478436245, + "learning_rate": 8.355430334225159e-06, + "loss": 0.0165, + "step": 2280 + }, + { + "epoch": 0.6056824216675518, + "grad_norm": 0.2140627847294915, + "learning_rate": 8.353802399207002e-06, + "loss": 0.0317, + "step": 2281 + }, + { + "epoch": 0.6059479553903345, + "grad_norm": 0.44382193770100115, + "learning_rate": 8.352173817615297e-06, + "loss": 0.0752, + "step": 2282 + }, + { + "epoch": 0.6062134891131173, + "grad_norm": 0.49723142268292964, + "learning_rate": 8.350544589764016e-06, + "loss": 0.063, + "step": 2283 + }, + { + "epoch": 0.6064790228359002, + "grad_norm": 0.9194236569451925, + "learning_rate": 8.348914715967254e-06, + "loss": 0.1191, + "step": 2284 + }, + { + "epoch": 0.606744556558683, + "grad_norm": 0.24813612518454406, + "learning_rate": 8.34728419653923e-06, + "loss": 0.0391, + "step": 2285 + }, + { + "epoch": 0.6070100902814658, + "grad_norm": 0.34688581028080623, + "learning_rate": 8.345653031794292e-06, + "loss": 0.0476, + "step": 2286 + }, + { + "epoch": 0.6072756240042485, + "grad_norm": 0.014976251503107323, + "learning_rate": 8.344021222046908e-06, + "loss": 0.0003, + "step": 2287 + }, + { + "epoch": 0.6075411577270313, + "grad_norm": 0.010239498051276927, + "learning_rate": 8.342388767611669e-06, + "loss": 0.0003, + "step": 2288 + }, + { + "epoch": 0.6078066914498141, + "grad_norm": 0.24797331251934185, + "learning_rate": 8.340755668803297e-06, + "loss": 0.0391, + "step": 2289 + }, + { + "epoch": 0.608072225172597, + "grad_norm": 0.39442360674937, + "learning_rate": 8.339121925936635e-06, + "loss": 0.063, + "step": 2290 + }, + { + "epoch": 0.6083377588953797, + "grad_norm": 0.4882557930299904, + "learning_rate": 8.337487539326642e-06, + "loss": 0.0957, + "step": 2291 + }, + { + "epoch": 0.6086032926181625, + "grad_norm": 0.24213981938791906, + "learning_rate": 8.335852509288419e-06, + "loss": 0.0391, + "step": 2292 + }, + { + "epoch": 0.6088688263409453, + "grad_norm": 0.405543557073391, + "learning_rate": 8.33421683613717e-06, + "loss": 0.0525, + "step": 2293 + }, + { + "epoch": 0.6091343600637281, + "grad_norm": 0.32266504598914586, + "learning_rate": 8.332580520188241e-06, + "loss": 0.0432, + "step": 2294 + }, + { + "epoch": 0.6093998937865109, + "grad_norm": 0.47791326335798107, + "learning_rate": 8.330943561757092e-06, + "loss": 0.1035, + "step": 2295 + }, + { + "epoch": 0.6096654275092936, + "grad_norm": 0.006265574433847926, + "learning_rate": 8.329305961159309e-06, + "loss": 0.0002, + "step": 2296 + }, + { + "epoch": 0.6099309612320765, + "grad_norm": 0.45121847028790313, + "learning_rate": 8.3276677187106e-06, + "loss": 0.0889, + "step": 2297 + }, + { + "epoch": 0.6101964949548593, + "grad_norm": 0.17145208056013195, + "learning_rate": 8.326028834726803e-06, + "loss": 0.0074, + "step": 2298 + }, + { + "epoch": 0.6104620286776421, + "grad_norm": 0.3932291465719769, + "learning_rate": 8.324389309523873e-06, + "loss": 0.0476, + "step": 2299 + }, + { + "epoch": 0.6107275624004248, + "grad_norm": 0.33105739943734536, + "learning_rate": 8.322749143417892e-06, + "loss": 0.0476, + "step": 2300 + }, + { + "epoch": 0.6109930961232076, + "grad_norm": 2.1885890649334465, + "learning_rate": 8.321108336725066e-06, + "loss": 0.1738, + "step": 2301 + }, + { + "epoch": 0.6112586298459904, + "grad_norm": 0.48775193895927893, + "learning_rate": 8.319466889761718e-06, + "loss": 0.0957, + "step": 2302 + }, + { + "epoch": 0.6115241635687733, + "grad_norm": 0.7877466710013121, + "learning_rate": 8.317824802844305e-06, + "loss": 0.1035, + "step": 2303 + }, + { + "epoch": 0.6117896972915561, + "grad_norm": 0.20592283381171456, + "learning_rate": 8.3161820762894e-06, + "loss": 0.0317, + "step": 2304 + }, + { + "epoch": 0.6120552310143388, + "grad_norm": 0.4716049405803976, + "learning_rate": 8.314538710413703e-06, + "loss": 0.0752, + "step": 2305 + }, + { + "epoch": 0.6123207647371216, + "grad_norm": 0.1928610939619178, + "learning_rate": 8.312894705534032e-06, + "loss": 0.0286, + "step": 2306 + }, + { + "epoch": 0.6125862984599044, + "grad_norm": 0.20049437341426857, + "learning_rate": 8.311250061967332e-06, + "loss": 0.0317, + "step": 2307 + }, + { + "epoch": 0.6128518321826872, + "grad_norm": 0.0048536170660810765, + "learning_rate": 8.309604780030676e-06, + "loss": 0.0001, + "step": 2308 + }, + { + "epoch": 0.61311736590547, + "grad_norm": 0.20288684277214483, + "learning_rate": 8.307958860041252e-06, + "loss": 0.0231, + "step": 2309 + }, + { + "epoch": 0.6133828996282528, + "grad_norm": 0.3336290833265773, + "learning_rate": 8.30631230231637e-06, + "loss": 0.0165, + "step": 2310 + }, + { + "epoch": 0.6136484333510356, + "grad_norm": 0.32902668206462005, + "learning_rate": 8.304665107173472e-06, + "loss": 0.0432, + "step": 2311 + }, + { + "epoch": 0.6139139670738184, + "grad_norm": 0.16120702595919442, + "learning_rate": 8.303017274930118e-06, + "loss": 0.0065, + "step": 2312 + }, + { + "epoch": 0.6141795007966012, + "grad_norm": 0.24888237536687158, + "learning_rate": 8.301368805903988e-06, + "loss": 0.0393, + "step": 2313 + }, + { + "epoch": 0.6144450345193839, + "grad_norm": 0.4778936575719, + "learning_rate": 8.299719700412887e-06, + "loss": 0.0957, + "step": 2314 + }, + { + "epoch": 0.6147105682421667, + "grad_norm": 0.46421659135889815, + "learning_rate": 8.298069958774744e-06, + "loss": 0.0889, + "step": 2315 + }, + { + "epoch": 0.6149761019649496, + "grad_norm": 0.2591944807949678, + "learning_rate": 8.296419581307612e-06, + "loss": 0.0393, + "step": 2316 + }, + { + "epoch": 0.6152416356877324, + "grad_norm": 0.2593281447418752, + "learning_rate": 8.294768568329663e-06, + "loss": 0.0391, + "step": 2317 + }, + { + "epoch": 0.6155071694105151, + "grad_norm": 0.2454039886131154, + "learning_rate": 8.29311692015919e-06, + "loss": 0.0393, + "step": 2318 + }, + { + "epoch": 0.6157727031332979, + "grad_norm": 1.0725614843473026, + "learning_rate": 8.291464637114615e-06, + "loss": 0.0889, + "step": 2319 + }, + { + "epoch": 0.6160382368560807, + "grad_norm": 0.5682068611390414, + "learning_rate": 8.289811719514475e-06, + "loss": 0.0688, + "step": 2320 + }, + { + "epoch": 0.6163037705788635, + "grad_norm": 0.061166479911426014, + "learning_rate": 8.288158167677438e-06, + "loss": 0.0019, + "step": 2321 + }, + { + "epoch": 0.6165693043016464, + "grad_norm": 0.33971581341465373, + "learning_rate": 8.286503981922284e-06, + "loss": 0.0525, + "step": 2322 + }, + { + "epoch": 0.6168348380244291, + "grad_norm": 0.5124635683300751, + "learning_rate": 8.284849162567925e-06, + "loss": 0.1279, + "step": 2323 + }, + { + "epoch": 0.6171003717472119, + "grad_norm": 0.5439010889294849, + "learning_rate": 8.283193709933385e-06, + "loss": 0.1367, + "step": 2324 + }, + { + "epoch": 0.6173659054699947, + "grad_norm": 0.4228613064274213, + "learning_rate": 8.281537624337823e-06, + "loss": 0.0752, + "step": 2325 + }, + { + "epoch": 0.6176314391927775, + "grad_norm": 0.40986802003962397, + "learning_rate": 8.279880906100509e-06, + "loss": 0.0206, + "step": 2326 + }, + { + "epoch": 0.6178969729155602, + "grad_norm": 0.19538534063919388, + "learning_rate": 8.278223555540839e-06, + "loss": 0.0082, + "step": 2327 + }, + { + "epoch": 0.618162506638343, + "grad_norm": 0.6747683629200573, + "learning_rate": 8.27656557297833e-06, + "loss": 0.0258, + "step": 2328 + }, + { + "epoch": 0.6184280403611259, + "grad_norm": 0.011110502129520684, + "learning_rate": 8.274906958732621e-06, + "loss": 0.0004, + "step": 2329 + }, + { + "epoch": 0.6186935740839087, + "grad_norm": 0.00968482341033898, + "learning_rate": 8.273247713123475e-06, + "loss": 0.0003, + "step": 2330 + }, + { + "epoch": 0.6189591078066915, + "grad_norm": 0.39803026751560056, + "learning_rate": 8.271587836470775e-06, + "loss": 0.0576, + "step": 2331 + }, + { + "epoch": 0.6192246415294742, + "grad_norm": 0.20216567324305207, + "learning_rate": 8.269927329094525e-06, + "loss": 0.0317, + "step": 2332 + }, + { + "epoch": 0.619490175252257, + "grad_norm": 0.482751336478894, + "learning_rate": 8.268266191314848e-06, + "loss": 0.0815, + "step": 2333 + }, + { + "epoch": 0.6197557089750398, + "grad_norm": 0.6348507704449461, + "learning_rate": 8.266604423451998e-06, + "loss": 0.082, + "step": 2334 + }, + { + "epoch": 0.6200212426978227, + "grad_norm": 0.4462307298833321, + "learning_rate": 8.26494202582634e-06, + "loss": 0.0957, + "step": 2335 + }, + { + "epoch": 0.6202867764206054, + "grad_norm": 0.22223260106113474, + "learning_rate": 8.263278998758368e-06, + "loss": 0.0354, + "step": 2336 + }, + { + "epoch": 0.6205523101433882, + "grad_norm": 0.5319176320043412, + "learning_rate": 8.261615342568691e-06, + "loss": 0.0693, + "step": 2337 + }, + { + "epoch": 0.620817843866171, + "grad_norm": 0.7897041254683038, + "learning_rate": 8.259951057578042e-06, + "loss": 0.0815, + "step": 2338 + }, + { + "epoch": 0.6210833775889538, + "grad_norm": 0.24789272502209345, + "learning_rate": 8.258286144107277e-06, + "loss": 0.0391, + "step": 2339 + }, + { + "epoch": 0.6213489113117366, + "grad_norm": 0.1962928039574505, + "learning_rate": 8.256620602477372e-06, + "loss": 0.0317, + "step": 2340 + }, + { + "epoch": 0.6216144450345193, + "grad_norm": 0.38062184874387445, + "learning_rate": 8.254954433009421e-06, + "loss": 0.0476, + "step": 2341 + }, + { + "epoch": 0.6218799787573022, + "grad_norm": 0.46836692014594117, + "learning_rate": 8.253287636024647e-06, + "loss": 0.1113, + "step": 2342 + }, + { + "epoch": 0.622145512480085, + "grad_norm": 0.5016670951914833, + "learning_rate": 8.251620211844384e-06, + "loss": 0.0889, + "step": 2343 + }, + { + "epoch": 0.6224110462028678, + "grad_norm": 0.27672384416956736, + "learning_rate": 8.249952160790094e-06, + "loss": 0.0432, + "step": 2344 + }, + { + "epoch": 0.6226765799256505, + "grad_norm": 2.2365272118296295, + "learning_rate": 8.248283483183355e-06, + "loss": 0.2461, + "step": 2345 + }, + { + "epoch": 0.6229421136484333, + "grad_norm": 0.010824866704553881, + "learning_rate": 8.246614179345874e-06, + "loss": 0.0004, + "step": 2346 + }, + { + "epoch": 0.6232076473712161, + "grad_norm": 0.46143064255751476, + "learning_rate": 8.244944249599468e-06, + "loss": 0.0693, + "step": 2347 + }, + { + "epoch": 0.623473181093999, + "grad_norm": 0.197913757831201, + "learning_rate": 8.243273694266078e-06, + "loss": 0.0317, + "step": 2348 + }, + { + "epoch": 0.6237387148167818, + "grad_norm": 0.18845701904230078, + "learning_rate": 8.241602513667775e-06, + "loss": 0.0317, + "step": 2349 + }, + { + "epoch": 0.6240042485395645, + "grad_norm": 0.01282290663905112, + "learning_rate": 8.239930708126738e-06, + "loss": 0.0004, + "step": 2350 + }, + { + "epoch": 0.6242697822623473, + "grad_norm": 16.351239706681206, + "learning_rate": 8.23825827796527e-06, + "loss": 0.1592, + "step": 2351 + }, + { + "epoch": 0.6245353159851301, + "grad_norm": 0.4638236760971655, + "learning_rate": 8.236585223505797e-06, + "loss": 0.0576, + "step": 2352 + }, + { + "epoch": 0.6248008497079129, + "grad_norm": 0.2768301789564024, + "learning_rate": 8.234911545070867e-06, + "loss": 0.0432, + "step": 2353 + }, + { + "epoch": 0.6250663834306956, + "grad_norm": 0.4471885131211778, + "learning_rate": 8.233237242983143e-06, + "loss": 0.0957, + "step": 2354 + }, + { + "epoch": 0.6253319171534785, + "grad_norm": 0.4690457707329022, + "learning_rate": 8.231562317565412e-06, + "loss": 0.1113, + "step": 2355 + }, + { + "epoch": 0.6255974508762613, + "grad_norm": 0.2117936916907193, + "learning_rate": 8.229886769140576e-06, + "loss": 0.0354, + "step": 2356 + }, + { + "epoch": 0.6258629845990441, + "grad_norm": 0.465788259161965, + "learning_rate": 8.228210598031667e-06, + "loss": 0.0957, + "step": 2357 + }, + { + "epoch": 0.6261285183218269, + "grad_norm": 0.47816690599129796, + "learning_rate": 8.226533804561828e-06, + "loss": 0.1191, + "step": 2358 + }, + { + "epoch": 0.6263940520446096, + "grad_norm": 0.45817966377340164, + "learning_rate": 8.224856389054323e-06, + "loss": 0.0752, + "step": 2359 + }, + { + "epoch": 0.6266595857673924, + "grad_norm": 0.4485603459973956, + "learning_rate": 8.223178351832539e-06, + "loss": 0.0889, + "step": 2360 + }, + { + "epoch": 0.6269251194901753, + "grad_norm": 0.24671501552023675, + "learning_rate": 8.221499693219983e-06, + "loss": 0.0393, + "step": 2361 + }, + { + "epoch": 0.6271906532129581, + "grad_norm": 0.44153040935284604, + "learning_rate": 8.219820413540282e-06, + "loss": 0.1035, + "step": 2362 + }, + { + "epoch": 0.6274561869357408, + "grad_norm": 0.22229038503291984, + "learning_rate": 8.218140513117179e-06, + "loss": 0.0391, + "step": 2363 + }, + { + "epoch": 0.6277217206585236, + "grad_norm": 0.21073179167941405, + "learning_rate": 8.216459992274538e-06, + "loss": 0.0354, + "step": 2364 + }, + { + "epoch": 0.6279872543813064, + "grad_norm": 0.4768461576690892, + "learning_rate": 8.214778851336347e-06, + "loss": 0.0576, + "step": 2365 + }, + { + "epoch": 0.6282527881040892, + "grad_norm": 0.24306794178792732, + "learning_rate": 8.213097090626707e-06, + "loss": 0.0391, + "step": 2366 + }, + { + "epoch": 0.6285183218268721, + "grad_norm": 0.22587702220375808, + "learning_rate": 8.211414710469844e-06, + "loss": 0.0391, + "step": 2367 + }, + { + "epoch": 0.6287838555496548, + "grad_norm": 0.34610716677320646, + "learning_rate": 8.2097317111901e-06, + "loss": 0.0525, + "step": 2368 + }, + { + "epoch": 0.6290493892724376, + "grad_norm": 0.2427284927431988, + "learning_rate": 8.208048093111936e-06, + "loss": 0.0432, + "step": 2369 + }, + { + "epoch": 0.6293149229952204, + "grad_norm": 0.22192022724435728, + "learning_rate": 8.206363856559935e-06, + "loss": 0.0391, + "step": 2370 + }, + { + "epoch": 0.6295804567180032, + "grad_norm": 0.4897512916877331, + "learning_rate": 8.204679001858799e-06, + "loss": 0.0693, + "step": 2371 + }, + { + "epoch": 0.6298459904407859, + "grad_norm": 0.46337682901056015, + "learning_rate": 8.202993529333347e-06, + "loss": 0.1113, + "step": 2372 + }, + { + "epoch": 0.6301115241635687, + "grad_norm": 0.2839984109594718, + "learning_rate": 8.201307439308518e-06, + "loss": 0.0479, + "step": 2373 + }, + { + "epoch": 0.6303770578863516, + "grad_norm": 0.1745841197189554, + "learning_rate": 8.199620732109372e-06, + "loss": 0.0082, + "step": 2374 + }, + { + "epoch": 0.6306425916091344, + "grad_norm": 0.004391893135969339, + "learning_rate": 8.197933408061084e-06, + "loss": 0.0001, + "step": 2375 + }, + { + "epoch": 0.6309081253319172, + "grad_norm": 1.2306109760090862, + "learning_rate": 8.19624546748895e-06, + "loss": 0.0889, + "step": 2376 + }, + { + "epoch": 0.6311736590546999, + "grad_norm": 0.3713908388581532, + "learning_rate": 8.194556910718387e-06, + "loss": 0.0579, + "step": 2377 + }, + { + "epoch": 0.6314391927774827, + "grad_norm": 0.24775797648691766, + "learning_rate": 8.192867738074927e-06, + "loss": 0.0432, + "step": 2378 + }, + { + "epoch": 0.6317047265002655, + "grad_norm": 0.20402971052931904, + "learning_rate": 8.191177949884225e-06, + "loss": 0.0354, + "step": 2379 + }, + { + "epoch": 0.6319702602230484, + "grad_norm": 0.4362281809419748, + "learning_rate": 8.18948754647205e-06, + "loss": 0.0957, + "step": 2380 + }, + { + "epoch": 0.6322357939458311, + "grad_norm": 0.28959527527546935, + "learning_rate": 8.187796528164293e-06, + "loss": 0.0476, + "step": 2381 + }, + { + "epoch": 0.6325013276686139, + "grad_norm": 0.09140151206442189, + "learning_rate": 8.186104895286963e-06, + "loss": 0.0035, + "step": 2382 + }, + { + "epoch": 0.6327668613913967, + "grad_norm": 0.19557327660362908, + "learning_rate": 8.184412648166183e-06, + "loss": 0.0317, + "step": 2383 + }, + { + "epoch": 0.6330323951141795, + "grad_norm": 0.44792122905298526, + "learning_rate": 8.182719787128203e-06, + "loss": 0.1113, + "step": 2384 + }, + { + "epoch": 0.6332979288369623, + "grad_norm": 0.16264109077612013, + "learning_rate": 8.181026312499383e-06, + "loss": 0.0065, + "step": 2385 + }, + { + "epoch": 0.633563462559745, + "grad_norm": 0.4174757303812451, + "learning_rate": 8.17933222460621e-06, + "loss": 0.0957, + "step": 2386 + }, + { + "epoch": 0.6338289962825279, + "grad_norm": 0.2844433170659035, + "learning_rate": 8.177637523775277e-06, + "loss": 0.0476, + "step": 2387 + }, + { + "epoch": 0.6340945300053107, + "grad_norm": 0.45102268091685427, + "learning_rate": 8.175942210333308e-06, + "loss": 0.0957, + "step": 2388 + }, + { + "epoch": 0.6343600637280935, + "grad_norm": 0.20134150219432534, + "learning_rate": 8.174246284607137e-06, + "loss": 0.0354, + "step": 2389 + }, + { + "epoch": 0.6346255974508762, + "grad_norm": 0.2030156976169986, + "learning_rate": 8.17254974692372e-06, + "loss": 0.0354, + "step": 2390 + }, + { + "epoch": 0.634891131173659, + "grad_norm": 0.2760159865347267, + "learning_rate": 8.170852597610129e-06, + "loss": 0.0476, + "step": 2391 + }, + { + "epoch": 0.6351566648964418, + "grad_norm": 0.2321461429834041, + "learning_rate": 8.16915483699355e-06, + "loss": 0.0391, + "step": 2392 + }, + { + "epoch": 0.6354221986192247, + "grad_norm": 0.19564761764998778, + "learning_rate": 8.1674564654013e-06, + "loss": 0.0317, + "step": 2393 + }, + { + "epoch": 0.6356877323420075, + "grad_norm": 0.22807901043563034, + "learning_rate": 8.165757483160798e-06, + "loss": 0.0354, + "step": 2394 + }, + { + "epoch": 0.6359532660647902, + "grad_norm": 0.40374718409756216, + "learning_rate": 8.16405789059959e-06, + "loss": 0.0957, + "step": 2395 + }, + { + "epoch": 0.636218799787573, + "grad_norm": 0.18828439448539036, + "learning_rate": 8.162357688045336e-06, + "loss": 0.0317, + "step": 2396 + }, + { + "epoch": 0.6364843335103558, + "grad_norm": 0.5508681551023428, + "learning_rate": 8.160656875825819e-06, + "loss": 0.0688, + "step": 2397 + }, + { + "epoch": 0.6367498672331386, + "grad_norm": 0.22943442453877513, + "learning_rate": 8.158955454268933e-06, + "loss": 0.0354, + "step": 2398 + }, + { + "epoch": 0.6370154009559214, + "grad_norm": 0.3184015763402584, + "learning_rate": 8.157253423702691e-06, + "loss": 0.0476, + "step": 2399 + }, + { + "epoch": 0.6372809346787042, + "grad_norm": 0.18844663466935352, + "learning_rate": 8.155550784455224e-06, + "loss": 0.0317, + "step": 2400 + }, + { + "epoch": 0.637546468401487, + "grad_norm": 0.29846025092636885, + "learning_rate": 8.153847536854783e-06, + "loss": 0.0165, + "step": 2401 + }, + { + "epoch": 0.6378120021242698, + "grad_norm": 2.35284057853226, + "learning_rate": 8.152143681229731e-06, + "loss": 0.1279, + "step": 2402 + }, + { + "epoch": 0.6380775358470526, + "grad_norm": 0.8354503289527497, + "learning_rate": 8.150439217908557e-06, + "loss": 0.063, + "step": 2403 + }, + { + "epoch": 0.6383430695698353, + "grad_norm": 0.20282594818458222, + "learning_rate": 8.148734147219854e-06, + "loss": 0.0317, + "step": 2404 + }, + { + "epoch": 0.6386086032926181, + "grad_norm": 0.20206675632195997, + "learning_rate": 8.147028469492346e-06, + "loss": 0.0317, + "step": 2405 + }, + { + "epoch": 0.638874137015401, + "grad_norm": 0.1891178850668181, + "learning_rate": 8.145322185054864e-06, + "loss": 0.0317, + "step": 2406 + }, + { + "epoch": 0.6391396707381838, + "grad_norm": 0.02594397956315926, + "learning_rate": 8.14361529423636e-06, + "loss": 0.0009, + "step": 2407 + }, + { + "epoch": 0.6394052044609665, + "grad_norm": 0.9499040690841225, + "learning_rate": 8.141907797365905e-06, + "loss": 0.0752, + "step": 2408 + }, + { + "epoch": 0.6396707381837493, + "grad_norm": 0.4294120442371818, + "learning_rate": 8.140199694772679e-06, + "loss": 0.0165, + "step": 2409 + }, + { + "epoch": 0.6399362719065321, + "grad_norm": 0.5108657047961442, + "learning_rate": 8.138490986785987e-06, + "loss": 0.0576, + "step": 2410 + }, + { + "epoch": 0.640201805629315, + "grad_norm": 0.16466473631208922, + "learning_rate": 8.13678167373525e-06, + "loss": 0.0229, + "step": 2411 + }, + { + "epoch": 0.6404673393520978, + "grad_norm": 0.3330742706061808, + "learning_rate": 8.13507175595e-06, + "loss": 0.0432, + "step": 2412 + }, + { + "epoch": 0.6407328730748805, + "grad_norm": 0.378847370562068, + "learning_rate": 8.133361233759891e-06, + "loss": 0.0432, + "step": 2413 + }, + { + "epoch": 0.6409984067976633, + "grad_norm": 0.2600353558677008, + "learning_rate": 8.13165010749469e-06, + "loss": 0.0072, + "step": 2414 + }, + { + "epoch": 0.6412639405204461, + "grad_norm": 0.29333077842226435, + "learning_rate": 8.129938377484284e-06, + "loss": 0.0064, + "step": 2415 + }, + { + "epoch": 0.6415294742432289, + "grad_norm": 0.5299335525891479, + "learning_rate": 8.12822604405867e-06, + "loss": 0.1279, + "step": 2416 + }, + { + "epoch": 0.6417950079660116, + "grad_norm": 0.17336858577260666, + "learning_rate": 8.126513107547971e-06, + "loss": 0.0231, + "step": 2417 + }, + { + "epoch": 0.6420605416887945, + "grad_norm": 0.5831842072683943, + "learning_rate": 8.124799568282418e-06, + "loss": 0.1367, + "step": 2418 + }, + { + "epoch": 0.6423260754115773, + "grad_norm": 0.13562484153020196, + "learning_rate": 8.123085426592363e-06, + "loss": 0.0206, + "step": 2419 + }, + { + "epoch": 0.6425916091343601, + "grad_norm": 0.2719928160261572, + "learning_rate": 8.12137068280827e-06, + "loss": 0.0082, + "step": 2420 + }, + { + "epoch": 0.6428571428571429, + "grad_norm": 0.9699416884676418, + "learning_rate": 8.119655337260721e-06, + "loss": 0.0957, + "step": 2421 + }, + { + "epoch": 0.6431226765799256, + "grad_norm": 0.152843068979682, + "learning_rate": 8.117939390280418e-06, + "loss": 0.0231, + "step": 2422 + }, + { + "epoch": 0.6433882103027084, + "grad_norm": 0.16211378881252245, + "learning_rate": 8.116222842198173e-06, + "loss": 0.0258, + "step": 2423 + }, + { + "epoch": 0.6436537440254912, + "grad_norm": 0.1639205560254292, + "learning_rate": 8.114505693344916e-06, + "loss": 0.0206, + "step": 2424 + }, + { + "epoch": 0.6439192777482741, + "grad_norm": 0.14428875393226245, + "learning_rate": 8.112787944051694e-06, + "loss": 0.0206, + "step": 2425 + }, + { + "epoch": 0.6441848114710568, + "grad_norm": 0.609707296664471, + "learning_rate": 8.111069594649668e-06, + "loss": 0.1279, + "step": 2426 + }, + { + "epoch": 0.6444503451938396, + "grad_norm": 0.1649677358814474, + "learning_rate": 8.109350645470116e-06, + "loss": 0.0035, + "step": 2427 + }, + { + "epoch": 0.6447158789166224, + "grad_norm": 0.5862630418546358, + "learning_rate": 8.107631096844431e-06, + "loss": 0.0476, + "step": 2428 + }, + { + "epoch": 0.6449814126394052, + "grad_norm": 0.2092197479267423, + "learning_rate": 8.105910949104121e-06, + "loss": 0.0286, + "step": 2429 + }, + { + "epoch": 0.645246946362188, + "grad_norm": 0.18859880230421508, + "learning_rate": 8.104190202580811e-06, + "loss": 0.0231, + "step": 2430 + }, + { + "epoch": 0.6455124800849708, + "grad_norm": 0.2835808719781997, + "learning_rate": 8.102468857606241e-06, + "loss": 0.0131, + "step": 2431 + }, + { + "epoch": 0.6457780138077536, + "grad_norm": 0.37105431499551944, + "learning_rate": 8.100746914512265e-06, + "loss": 0.0391, + "step": 2432 + }, + { + "epoch": 0.6460435475305364, + "grad_norm": 0.00911883912436982, + "learning_rate": 8.099024373630855e-06, + "loss": 0.0003, + "step": 2433 + }, + { + "epoch": 0.6463090812533192, + "grad_norm": 0.1851032562716545, + "learning_rate": 8.097301235294093e-06, + "loss": 0.0229, + "step": 2434 + }, + { + "epoch": 0.6465746149761019, + "grad_norm": 0.13013605072218193, + "learning_rate": 8.095577499834184e-06, + "loss": 0.0184, + "step": 2435 + }, + { + "epoch": 0.6468401486988847, + "grad_norm": 0.19029942024404486, + "learning_rate": 8.093853167583441e-06, + "loss": 0.0231, + "step": 2436 + }, + { + "epoch": 0.6471056824216675, + "grad_norm": 0.008503277311834938, + "learning_rate": 8.092128238874297e-06, + "loss": 0.0002, + "step": 2437 + }, + { + "epoch": 0.6473712161444504, + "grad_norm": 0.8826646461281974, + "learning_rate": 8.090402714039296e-06, + "loss": 0.1035, + "step": 2438 + }, + { + "epoch": 0.6476367498672332, + "grad_norm": 0.130882510723378, + "learning_rate": 8.0886765934111e-06, + "loss": 0.0184, + "step": 2439 + }, + { + "epoch": 0.6479022835900159, + "grad_norm": 1.3535906527255483, + "learning_rate": 8.086949877322486e-06, + "loss": 0.063, + "step": 2440 + }, + { + "epoch": 0.6481678173127987, + "grad_norm": 0.4121293197892748, + "learning_rate": 8.085222566106339e-06, + "loss": 0.0354, + "step": 2441 + }, + { + "epoch": 0.6484333510355815, + "grad_norm": 0.18830366088163383, + "learning_rate": 8.08349466009567e-06, + "loss": 0.0229, + "step": 2442 + }, + { + "epoch": 0.6486988847583643, + "grad_norm": 0.16680924139346118, + "learning_rate": 8.081766159623598e-06, + "loss": 0.0184, + "step": 2443 + }, + { + "epoch": 0.648964418481147, + "grad_norm": 0.4084781191271248, + "learning_rate": 8.080037065023355e-06, + "loss": 0.0354, + "step": 2444 + }, + { + "epoch": 0.6492299522039299, + "grad_norm": 5.238925921808123, + "learning_rate": 8.078307376628292e-06, + "loss": 0.5586, + "step": 2445 + }, + { + "epoch": 0.6494954859267127, + "grad_norm": 0.013936278880006165, + "learning_rate": 8.07657709477187e-06, + "loss": 0.0004, + "step": 2446 + }, + { + "epoch": 0.6497610196494955, + "grad_norm": 0.1395429328864441, + "learning_rate": 8.074846219787674e-06, + "loss": 0.0146, + "step": 2447 + }, + { + "epoch": 0.6500265533722783, + "grad_norm": 1.0650496927061919, + "learning_rate": 8.073114752009388e-06, + "loss": 0.1035, + "step": 2448 + }, + { + "epoch": 0.650292087095061, + "grad_norm": 0.10497343724109437, + "learning_rate": 8.071382691770821e-06, + "loss": 0.004, + "step": 2449 + }, + { + "epoch": 0.6505576208178439, + "grad_norm": 0.5695751393605243, + "learning_rate": 8.069650039405897e-06, + "loss": 0.0229, + "step": 2450 + }, + { + "epoch": 0.6508231545406267, + "grad_norm": 0.005958182138508997, + "learning_rate": 8.067916795248648e-06, + "loss": 0.0002, + "step": 2451 + }, + { + "epoch": 0.6510886882634095, + "grad_norm": 0.8943918463821318, + "learning_rate": 8.066182959633221e-06, + "loss": 0.0957, + "step": 2452 + }, + { + "epoch": 0.6513542219861922, + "grad_norm": 0.7564740145511355, + "learning_rate": 8.064448532893885e-06, + "loss": 0.1836, + "step": 2453 + }, + { + "epoch": 0.651619755708975, + "grad_norm": 0.14222559641670388, + "learning_rate": 8.062713515365014e-06, + "loss": 0.0184, + "step": 2454 + }, + { + "epoch": 0.6518852894317578, + "grad_norm": 0.4468206930244721, + "learning_rate": 8.060977907381095e-06, + "loss": 0.0206, + "step": 2455 + }, + { + "epoch": 0.6521508231545406, + "grad_norm": 0.13290279491751086, + "learning_rate": 8.059241709276738e-06, + "loss": 0.0027, + "step": 2456 + }, + { + "epoch": 0.6524163568773235, + "grad_norm": 1.0096898827859466, + "learning_rate": 8.057504921386661e-06, + "loss": 0.0354, + "step": 2457 + }, + { + "epoch": 0.6526818906001062, + "grad_norm": 0.1252796192114043, + "learning_rate": 8.055767544045694e-06, + "loss": 0.0165, + "step": 2458 + }, + { + "epoch": 0.652947424322889, + "grad_norm": 1.2062105359531898, + "learning_rate": 8.054029577588784e-06, + "loss": 0.0525, + "step": 2459 + }, + { + "epoch": 0.6532129580456718, + "grad_norm": 0.7982770345375926, + "learning_rate": 8.052291022350992e-06, + "loss": 0.1455, + "step": 2460 + }, + { + "epoch": 0.6534784917684546, + "grad_norm": 0.15230624461375658, + "learning_rate": 8.050551878667486e-06, + "loss": 0.0184, + "step": 2461 + }, + { + "epoch": 0.6537440254912373, + "grad_norm": 0.17124008317909933, + "learning_rate": 8.048812146873558e-06, + "loss": 0.0206, + "step": 2462 + }, + { + "epoch": 0.6540095592140202, + "grad_norm": 0.6799365873702063, + "learning_rate": 8.047071827304605e-06, + "loss": 0.1641, + "step": 2463 + }, + { + "epoch": 0.654275092936803, + "grad_norm": 0.2218502905044794, + "learning_rate": 8.04533092029614e-06, + "loss": 0.0255, + "step": 2464 + }, + { + "epoch": 0.6545406266595858, + "grad_norm": 0.6775412953084603, + "learning_rate": 8.04358942618379e-06, + "loss": 0.0432, + "step": 2465 + }, + { + "epoch": 0.6548061603823686, + "grad_norm": 0.8262956862812405, + "learning_rate": 8.041847345303296e-06, + "loss": 0.1035, + "step": 2466 + }, + { + "epoch": 0.6550716941051513, + "grad_norm": 1.710170102054662, + "learning_rate": 8.040104677990508e-06, + "loss": 0.1191, + "step": 2467 + }, + { + "epoch": 0.6553372278279341, + "grad_norm": 0.5331684422379329, + "learning_rate": 8.038361424581393e-06, + "loss": 0.1113, + "step": 2468 + }, + { + "epoch": 0.655602761550717, + "grad_norm": 0.051756961602491415, + "learning_rate": 8.036617585412029e-06, + "loss": 0.0015, + "step": 2469 + }, + { + "epoch": 0.6558682952734998, + "grad_norm": 0.3575356416416773, + "learning_rate": 8.034873160818608e-06, + "loss": 0.0391, + "step": 2470 + }, + { + "epoch": 0.6561338289962825, + "grad_norm": 0.7418845577654317, + "learning_rate": 8.033128151137437e-06, + "loss": 0.0576, + "step": 2471 + }, + { + "epoch": 0.6563993627190653, + "grad_norm": 0.5475151895114204, + "learning_rate": 8.031382556704927e-06, + "loss": 0.1191, + "step": 2472 + }, + { + "epoch": 0.6566648964418481, + "grad_norm": 0.7868808899482823, + "learning_rate": 8.029636377857615e-06, + "loss": 0.0525, + "step": 2473 + }, + { + "epoch": 0.6569304301646309, + "grad_norm": 0.031222580309087908, + "learning_rate": 8.027889614932138e-06, + "loss": 0.001, + "step": 2474 + }, + { + "epoch": 0.6571959638874137, + "grad_norm": 0.0023938766523944883, + "learning_rate": 8.026142268265256e-06, + "loss": 0.0001, + "step": 2475 + }, + { + "epoch": 0.6574614976101965, + "grad_norm": 0.5082204907290545, + "learning_rate": 8.024394338193834e-06, + "loss": 0.0231, + "step": 2476 + }, + { + "epoch": 0.6577270313329793, + "grad_norm": 0.2841401426726304, + "learning_rate": 8.022645825054853e-06, + "loss": 0.0391, + "step": 2477 + }, + { + "epoch": 0.6579925650557621, + "grad_norm": 0.1778476623013855, + "learning_rate": 8.020896729185406e-06, + "loss": 0.0258, + "step": 2478 + }, + { + "epoch": 0.6582580987785449, + "grad_norm": 0.6003444742766877, + "learning_rate": 8.019147050922697e-06, + "loss": 0.063, + "step": 2479 + }, + { + "epoch": 0.6585236325013276, + "grad_norm": 0.5391221602926725, + "learning_rate": 8.017396790604046e-06, + "loss": 0.0525, + "step": 2480 + }, + { + "epoch": 0.6587891662241104, + "grad_norm": 0.2558414281193221, + "learning_rate": 8.015645948566878e-06, + "loss": 0.0354, + "step": 2481 + }, + { + "epoch": 0.6590546999468933, + "grad_norm": 0.3430633850219204, + "learning_rate": 8.013894525148738e-06, + "loss": 0.0391, + "step": 2482 + }, + { + "epoch": 0.6593202336696761, + "grad_norm": 0.2280669290076067, + "learning_rate": 8.01214252068728e-06, + "loss": 0.0317, + "step": 2483 + }, + { + "epoch": 0.6595857673924589, + "grad_norm": 0.3287523649755011, + "learning_rate": 8.010389935520269e-06, + "loss": 0.0391, + "step": 2484 + }, + { + "epoch": 0.6598513011152416, + "grad_norm": 0.45223280240435687, + "learning_rate": 8.008636769985584e-06, + "loss": 0.0258, + "step": 2485 + }, + { + "epoch": 0.6601168348380244, + "grad_norm": 2.6262400091257017, + "learning_rate": 8.00688302442121e-06, + "loss": 0.1738, + "step": 2486 + }, + { + "epoch": 0.6603823685608072, + "grad_norm": 0.7361856205051326, + "learning_rate": 8.005128699165254e-06, + "loss": 0.0752, + "step": 2487 + }, + { + "epoch": 0.66064790228359, + "grad_norm": 0.2037951140406116, + "learning_rate": 8.003373794555927e-06, + "loss": 0.0064, + "step": 2488 + }, + { + "epoch": 0.6609134360063728, + "grad_norm": 1.933116255444121, + "learning_rate": 8.001618310931554e-06, + "loss": 0.1738, + "step": 2489 + }, + { + "epoch": 0.6611789697291556, + "grad_norm": 0.0022258349676849386, + "learning_rate": 7.999862248630569e-06, + "loss": 0.0001, + "step": 2490 + }, + { + "epoch": 0.6614445034519384, + "grad_norm": 0.21392769857053265, + "learning_rate": 7.998105607991526e-06, + "loss": 0.0317, + "step": 2491 + }, + { + "epoch": 0.6617100371747212, + "grad_norm": 0.48324900842751567, + "learning_rate": 7.99634838935308e-06, + "loss": 0.0525, + "step": 2492 + }, + { + "epoch": 0.661975570897504, + "grad_norm": 0.07633810882611429, + "learning_rate": 7.994590593054001e-06, + "loss": 0.0028, + "step": 2493 + }, + { + "epoch": 0.6622411046202867, + "grad_norm": 0.40052560457622993, + "learning_rate": 7.992832219433177e-06, + "loss": 0.0103, + "step": 2494 + }, + { + "epoch": 0.6625066383430696, + "grad_norm": 0.37318312412400484, + "learning_rate": 7.991073268829596e-06, + "loss": 0.0391, + "step": 2495 + }, + { + "epoch": 0.6627721720658524, + "grad_norm": 0.30707270343074616, + "learning_rate": 7.989313741582365e-06, + "loss": 0.0391, + "step": 2496 + }, + { + "epoch": 0.6630377057886352, + "grad_norm": 0.6986861949067361, + "learning_rate": 7.987553638030703e-06, + "loss": 0.0317, + "step": 2497 + }, + { + "epoch": 0.6633032395114179, + "grad_norm": 0.9700356456151883, + "learning_rate": 7.985792958513932e-06, + "loss": 0.0525, + "step": 2498 + }, + { + "epoch": 0.6635687732342007, + "grad_norm": 0.007614422122942421, + "learning_rate": 7.984031703371494e-06, + "loss": 0.0002, + "step": 2499 + }, + { + "epoch": 0.6638343069569835, + "grad_norm": 0.014724626259248861, + "learning_rate": 7.982269872942935e-06, + "loss": 0.0005, + "step": 2500 + }, + { + "epoch": 0.6640998406797664, + "grad_norm": 1.5296963212068422, + "learning_rate": 7.98050746756792e-06, + "loss": 0.1113, + "step": 2501 + }, + { + "epoch": 0.6643653744025492, + "grad_norm": 0.34743573798163596, + "learning_rate": 7.978744487586214e-06, + "loss": 0.0354, + "step": 2502 + }, + { + "epoch": 0.6646309081253319, + "grad_norm": 1.2957176882985888, + "learning_rate": 7.976980933337704e-06, + "loss": 0.0957, + "step": 2503 + }, + { + "epoch": 0.6648964418481147, + "grad_norm": 0.530402003953045, + "learning_rate": 7.975216805162379e-06, + "loss": 0.0476, + "step": 2504 + }, + { + "epoch": 0.6651619755708975, + "grad_norm": 0.2159838301413213, + "learning_rate": 7.973452103400345e-06, + "loss": 0.0286, + "step": 2505 + }, + { + "epoch": 0.6654275092936803, + "grad_norm": 0.58381324136832, + "learning_rate": 7.971686828391814e-06, + "loss": 0.1113, + "step": 2506 + }, + { + "epoch": 0.665693043016463, + "grad_norm": 0.7112116091478032, + "learning_rate": 7.969920980477109e-06, + "loss": 0.0525, + "step": 2507 + }, + { + "epoch": 0.6659585767392459, + "grad_norm": 0.1346181038929171, + "learning_rate": 7.968154559996667e-06, + "loss": 0.0184, + "step": 2508 + }, + { + "epoch": 0.6662241104620287, + "grad_norm": 0.36729101747451637, + "learning_rate": 7.96638756729103e-06, + "loss": 0.0103, + "step": 2509 + }, + { + "epoch": 0.6664896441848115, + "grad_norm": 0.24191590323286846, + "learning_rate": 7.964620002700857e-06, + "loss": 0.0286, + "step": 2510 + }, + { + "epoch": 0.6667551779075943, + "grad_norm": 0.13870626924469656, + "learning_rate": 7.962851866566912e-06, + "loss": 0.0184, + "step": 2511 + }, + { + "epoch": 0.667020711630377, + "grad_norm": 0.3484917362651463, + "learning_rate": 7.961083159230069e-06, + "loss": 0.0117, + "step": 2512 + }, + { + "epoch": 0.6672862453531598, + "grad_norm": 0.13731295839742647, + "learning_rate": 7.959313881031317e-06, + "loss": 0.0184, + "step": 2513 + }, + { + "epoch": 0.6675517790759427, + "grad_norm": 3.2324515678345236, + "learning_rate": 7.95754403231175e-06, + "loss": 0.3672, + "step": 2514 + }, + { + "epoch": 0.6678173127987255, + "grad_norm": 0.8027101054158648, + "learning_rate": 7.955773613412577e-06, + "loss": 0.1035, + "step": 2515 + }, + { + "epoch": 0.6680828465215082, + "grad_norm": 0.012276814395669461, + "learning_rate": 7.954002624675109e-06, + "loss": 0.0004, + "step": 2516 + }, + { + "epoch": 0.668348380244291, + "grad_norm": 0.6645136417504344, + "learning_rate": 7.952231066440774e-06, + "loss": 0.0432, + "step": 2517 + }, + { + "epoch": 0.6686139139670738, + "grad_norm": 0.020156802365212174, + "learning_rate": 7.95045893905111e-06, + "loss": 0.0006, + "step": 2518 + }, + { + "epoch": 0.6688794476898566, + "grad_norm": 0.13489267355193552, + "learning_rate": 7.948686242847757e-06, + "loss": 0.0184, + "step": 2519 + }, + { + "epoch": 0.6691449814126395, + "grad_norm": 0.18029719451530818, + "learning_rate": 7.946912978172475e-06, + "loss": 0.0231, + "step": 2520 + }, + { + "epoch": 0.6694105151354222, + "grad_norm": 0.13420055190099783, + "learning_rate": 7.945139145367125e-06, + "loss": 0.0184, + "step": 2521 + }, + { + "epoch": 0.669676048858205, + "grad_norm": 0.006569270064903438, + "learning_rate": 7.943364744773682e-06, + "loss": 0.0002, + "step": 2522 + }, + { + "epoch": 0.6699415825809878, + "grad_norm": 0.15121970442488306, + "learning_rate": 7.941589776734232e-06, + "loss": 0.0184, + "step": 2523 + }, + { + "epoch": 0.6702071163037706, + "grad_norm": 0.004603818239122557, + "learning_rate": 7.939814241590966e-06, + "loss": 0.0001, + "step": 2524 + }, + { + "epoch": 0.6704726500265533, + "grad_norm": 0.31491564303458147, + "learning_rate": 7.938038139686185e-06, + "loss": 0.0255, + "step": 2525 + }, + { + "epoch": 0.6707381837493361, + "grad_norm": 0.20646506110006732, + "learning_rate": 7.9362614713623e-06, + "loss": 0.0229, + "step": 2526 + }, + { + "epoch": 0.671003717472119, + "grad_norm": 0.07584957555895411, + "learning_rate": 7.934484236961837e-06, + "loss": 0.0015, + "step": 2527 + }, + { + "epoch": 0.6712692511949018, + "grad_norm": 0.7185035367483196, + "learning_rate": 7.93270643682742e-06, + "loss": 0.1738, + "step": 2528 + }, + { + "epoch": 0.6715347849176846, + "grad_norm": 0.11504967793452348, + "learning_rate": 7.93092807130179e-06, + "loss": 0.0131, + "step": 2529 + }, + { + "epoch": 0.6718003186404673, + "grad_norm": 0.11237662969120325, + "learning_rate": 7.929149140727796e-06, + "loss": 0.0146, + "step": 2530 + }, + { + "epoch": 0.6720658523632501, + "grad_norm": 0.10415644091405529, + "learning_rate": 7.927369645448393e-06, + "loss": 0.0131, + "step": 2531 + }, + { + "epoch": 0.6723313860860329, + "grad_norm": 0.3935392616871103, + "learning_rate": 7.925589585806647e-06, + "loss": 0.0104, + "step": 2532 + }, + { + "epoch": 0.6725969198088158, + "grad_norm": 0.09121156771253676, + "learning_rate": 7.923808962145733e-06, + "loss": 0.0103, + "step": 2533 + }, + { + "epoch": 0.6728624535315985, + "grad_norm": 1.0692451874978166, + "learning_rate": 7.922027774808935e-06, + "loss": 0.1113, + "step": 2534 + }, + { + "epoch": 0.6731279872543813, + "grad_norm": 0.9573688778411489, + "learning_rate": 7.920246024139644e-06, + "loss": 0.0576, + "step": 2535 + }, + { + "epoch": 0.6733935209771641, + "grad_norm": 0.059813247050768875, + "learning_rate": 7.91846371048136e-06, + "loss": 0.0019, + "step": 2536 + }, + { + "epoch": 0.6736590546999469, + "grad_norm": 0.6652100261639957, + "learning_rate": 7.916680834177694e-06, + "loss": 0.1738, + "step": 2537 + }, + { + "epoch": 0.6739245884227297, + "grad_norm": 0.13520882226006525, + "learning_rate": 7.914897395572362e-06, + "loss": 0.0165, + "step": 2538 + }, + { + "epoch": 0.6741901221455124, + "grad_norm": 0.9096640956241419, + "learning_rate": 7.91311339500919e-06, + "loss": 0.1113, + "step": 2539 + }, + { + "epoch": 0.6744556558682953, + "grad_norm": 0.33915274722269395, + "learning_rate": 7.91132883283211e-06, + "loss": 0.0317, + "step": 2540 + }, + { + "epoch": 0.6747211895910781, + "grad_norm": 0.005289165983345265, + "learning_rate": 7.90954370938517e-06, + "loss": 0.0001, + "step": 2541 + }, + { + "epoch": 0.6749867233138609, + "grad_norm": 17.60841829847521, + "learning_rate": 7.907758025012516e-06, + "loss": 0.9414, + "step": 2542 + }, + { + "epoch": 0.6752522570366436, + "grad_norm": 0.03091370275377362, + "learning_rate": 7.90597178005841e-06, + "loss": 0.0006, + "step": 2543 + }, + { + "epoch": 0.6755177907594264, + "grad_norm": 0.1865761648216731, + "learning_rate": 7.904184974867216e-06, + "loss": 0.0229, + "step": 2544 + }, + { + "epoch": 0.6757833244822092, + "grad_norm": 0.2789972234821362, + "learning_rate": 7.902397609783412e-06, + "loss": 0.0255, + "step": 2545 + }, + { + "epoch": 0.676048858204992, + "grad_norm": 0.7865666826616647, + "learning_rate": 7.90060968515158e-06, + "loss": 0.1113, + "step": 2546 + }, + { + "epoch": 0.6763143919277749, + "grad_norm": 0.6679300541605265, + "learning_rate": 7.898821201316408e-06, + "loss": 0.1455, + "step": 2547 + }, + { + "epoch": 0.6765799256505576, + "grad_norm": 0.23298557839281753, + "learning_rate": 7.897032158622698e-06, + "loss": 0.0229, + "step": 2548 + }, + { + "epoch": 0.6768454593733404, + "grad_norm": 0.020640745187170212, + "learning_rate": 7.895242557415354e-06, + "loss": 0.0007, + "step": 2549 + }, + { + "epoch": 0.6771109930961232, + "grad_norm": 0.13387843570835642, + "learning_rate": 7.893452398039391e-06, + "loss": 0.0039, + "step": 2550 + }, + { + "epoch": 0.677376526818906, + "grad_norm": 0.706475736265977, + "learning_rate": 7.891661680839932e-06, + "loss": 0.0525, + "step": 2551 + }, + { + "epoch": 0.6776420605416887, + "grad_norm": 0.6432417152852691, + "learning_rate": 7.889870406162205e-06, + "loss": 0.0476, + "step": 2552 + }, + { + "epoch": 0.6779075942644716, + "grad_norm": 0.6337768165849034, + "learning_rate": 7.888078574351547e-06, + "loss": 0.0525, + "step": 2553 + }, + { + "epoch": 0.6781731279872544, + "grad_norm": 2.2430583150365235, + "learning_rate": 7.886286185753401e-06, + "loss": 0.1191, + "step": 2554 + }, + { + "epoch": 0.6784386617100372, + "grad_norm": 0.5608499341852453, + "learning_rate": 7.88449324071332e-06, + "loss": 0.1191, + "step": 2555 + }, + { + "epoch": 0.67870419543282, + "grad_norm": 0.6111712799491892, + "learning_rate": 7.88269973957696e-06, + "loss": 0.0525, + "step": 2556 + }, + { + "epoch": 0.6789697291556027, + "grad_norm": 0.008023225339450253, + "learning_rate": 7.88090568269009e-06, + "loss": 0.0003, + "step": 2557 + }, + { + "epoch": 0.6792352628783855, + "grad_norm": 0.13890360468697963, + "learning_rate": 7.87911107039858e-06, + "loss": 0.0206, + "step": 2558 + }, + { + "epoch": 0.6795007966011684, + "grad_norm": 0.006396633874172496, + "learning_rate": 7.877315903048411e-06, + "loss": 0.0002, + "step": 2559 + }, + { + "epoch": 0.6797663303239512, + "grad_norm": 0.5272952841727334, + "learning_rate": 7.875520180985671e-06, + "loss": 0.1279, + "step": 2560 + }, + { + "epoch": 0.6800318640467339, + "grad_norm": 0.15215284739829107, + "learning_rate": 7.873723904556556e-06, + "loss": 0.0206, + "step": 2561 + }, + { + "epoch": 0.6802973977695167, + "grad_norm": 0.2685840828156918, + "learning_rate": 7.871927074107362e-06, + "loss": 0.0317, + "step": 2562 + }, + { + "epoch": 0.6805629314922995, + "grad_norm": 0.12454574365747811, + "learning_rate": 7.870129689984501e-06, + "loss": 0.0184, + "step": 2563 + }, + { + "epoch": 0.6808284652150823, + "grad_norm": 0.19406155805632785, + "learning_rate": 7.868331752534486e-06, + "loss": 0.0286, + "step": 2564 + }, + { + "epoch": 0.6810939989378652, + "grad_norm": 0.5857551690862067, + "learning_rate": 7.866533262103937e-06, + "loss": 0.1279, + "step": 2565 + }, + { + "epoch": 0.6813595326606479, + "grad_norm": 0.02346365222493494, + "learning_rate": 7.864734219039584e-06, + "loss": 0.0008, + "step": 2566 + }, + { + "epoch": 0.6816250663834307, + "grad_norm": 0.24255439688896133, + "learning_rate": 7.862934623688257e-06, + "loss": 0.0286, + "step": 2567 + }, + { + "epoch": 0.6818906001062135, + "grad_norm": 0.5255017101341883, + "learning_rate": 7.861134476396903e-06, + "loss": 0.0231, + "step": 2568 + }, + { + "epoch": 0.6821561338289963, + "grad_norm": 0.012736146961047403, + "learning_rate": 7.859333777512564e-06, + "loss": 0.0004, + "step": 2569 + }, + { + "epoch": 0.682421667551779, + "grad_norm": 0.9095361743993235, + "learning_rate": 7.857532527382396e-06, + "loss": 0.0815, + "step": 2570 + }, + { + "epoch": 0.6826872012745618, + "grad_norm": 0.49072357379397186, + "learning_rate": 7.855730726353658e-06, + "loss": 0.1191, + "step": 2571 + }, + { + "epoch": 0.6829527349973447, + "grad_norm": 0.7412126470540541, + "learning_rate": 7.853928374773716e-06, + "loss": 0.0576, + "step": 2572 + }, + { + "epoch": 0.6832182687201275, + "grad_norm": 0.4494084743162739, + "learning_rate": 7.852125472990042e-06, + "loss": 0.0354, + "step": 2573 + }, + { + "epoch": 0.6834838024429103, + "grad_norm": 0.5757809824518603, + "learning_rate": 7.850322021350216e-06, + "loss": 0.0432, + "step": 2574 + }, + { + "epoch": 0.683749336165693, + "grad_norm": 0.22852389238489448, + "learning_rate": 7.84851802020192e-06, + "loss": 0.0317, + "step": 2575 + }, + { + "epoch": 0.6840148698884758, + "grad_norm": 0.26575355386527516, + "learning_rate": 7.846713469892943e-06, + "loss": 0.0354, + "step": 2576 + }, + { + "epoch": 0.6842804036112586, + "grad_norm": 0.25113490741339156, + "learning_rate": 7.844908370771187e-06, + "loss": 0.0317, + "step": 2577 + }, + { + "epoch": 0.6845459373340415, + "grad_norm": 0.39357941940287294, + "learning_rate": 7.843102723184647e-06, + "loss": 0.0432, + "step": 2578 + }, + { + "epoch": 0.6848114710568242, + "grad_norm": 0.6717295350004934, + "learning_rate": 7.841296527481433e-06, + "loss": 0.0476, + "step": 2579 + }, + { + "epoch": 0.685077004779607, + "grad_norm": 0.1812351926764289, + "learning_rate": 7.839489784009759e-06, + "loss": 0.0255, + "step": 2580 + }, + { + "epoch": 0.6853425385023898, + "grad_norm": 0.8723411823136388, + "learning_rate": 7.837682493117944e-06, + "loss": 0.0815, + "step": 2581 + }, + { + "epoch": 0.6856080722251726, + "grad_norm": 0.1296932623359227, + "learning_rate": 7.835874655154409e-06, + "loss": 0.0206, + "step": 2582 + }, + { + "epoch": 0.6858736059479554, + "grad_norm": 0.5981434244492586, + "learning_rate": 7.83406627046769e-06, + "loss": 0.1279, + "step": 2583 + }, + { + "epoch": 0.6861391396707381, + "grad_norm": 0.39059476907440055, + "learning_rate": 7.832257339406417e-06, + "loss": 0.0184, + "step": 2584 + }, + { + "epoch": 0.686404673393521, + "grad_norm": 2.0655102860414924, + "learning_rate": 7.830447862319333e-06, + "loss": 0.0957, + "step": 2585 + }, + { + "epoch": 0.6866702071163038, + "grad_norm": 0.2908983749600404, + "learning_rate": 7.828637839555282e-06, + "loss": 0.0286, + "step": 2586 + }, + { + "epoch": 0.6869357408390866, + "grad_norm": 0.5809804664023751, + "learning_rate": 7.826827271463217e-06, + "loss": 0.1113, + "step": 2587 + }, + { + "epoch": 0.6872012745618693, + "grad_norm": 0.19958857836152052, + "learning_rate": 7.825016158392194e-06, + "loss": 0.0255, + "step": 2588 + }, + { + "epoch": 0.6874668082846521, + "grad_norm": 0.29631872762851064, + "learning_rate": 7.823204500691373e-06, + "loss": 0.0317, + "step": 2589 + }, + { + "epoch": 0.6877323420074349, + "grad_norm": 0.30229450012166925, + "learning_rate": 7.82139229871002e-06, + "loss": 0.0206, + "step": 2590 + }, + { + "epoch": 0.6879978757302178, + "grad_norm": 0.8077087883243114, + "learning_rate": 7.81957955279751e-06, + "loss": 0.1113, + "step": 2591 + }, + { + "epoch": 0.6882634094530006, + "grad_norm": 0.5280595157482699, + "learning_rate": 7.817766263303312e-06, + "loss": 0.1279, + "step": 2592 + }, + { + "epoch": 0.6885289431757833, + "grad_norm": 0.42379473969616366, + "learning_rate": 7.815952430577015e-06, + "loss": 0.0184, + "step": 2593 + }, + { + "epoch": 0.6887944768985661, + "grad_norm": 0.2741309719262226, + "learning_rate": 7.814138054968297e-06, + "loss": 0.0255, + "step": 2594 + }, + { + "epoch": 0.6890600106213489, + "grad_norm": 0.029443856659365122, + "learning_rate": 7.812323136826954e-06, + "loss": 0.0006, + "step": 2595 + }, + { + "epoch": 0.6893255443441317, + "grad_norm": 0.8098451232213809, + "learning_rate": 7.810507676502877e-06, + "loss": 0.0752, + "step": 2596 + }, + { + "epoch": 0.6895910780669146, + "grad_norm": 0.13890871318442746, + "learning_rate": 7.808691674346068e-06, + "loss": 0.0206, + "step": 2597 + }, + { + "epoch": 0.6898566117896973, + "grad_norm": 0.5656255216640846, + "learning_rate": 7.806875130706628e-06, + "loss": 0.1035, + "step": 2598 + }, + { + "epoch": 0.6901221455124801, + "grad_norm": 0.14849432718660888, + "learning_rate": 7.805058045934769e-06, + "loss": 0.0231, + "step": 2599 + }, + { + "epoch": 0.6903876792352629, + "grad_norm": 0.026839753528928107, + "learning_rate": 7.8032404203808e-06, + "loss": 0.0008, + "step": 2600 + }, + { + "epoch": 0.6906532129580457, + "grad_norm": 0.14828994683453836, + "learning_rate": 7.80142225439514e-06, + "loss": 0.0229, + "step": 2601 + }, + { + "epoch": 0.6909187466808284, + "grad_norm": 0.18245733867261665, + "learning_rate": 7.799603548328307e-06, + "loss": 0.0255, + "step": 2602 + }, + { + "epoch": 0.6911842804036112, + "grad_norm": 0.13107527377346984, + "learning_rate": 7.797784302530929e-06, + "loss": 0.0184, + "step": 2603 + }, + { + "epoch": 0.6914498141263941, + "grad_norm": 0.5601388641887275, + "learning_rate": 7.795964517353734e-06, + "loss": 0.1553, + "step": 2604 + }, + { + "epoch": 0.6917153478491769, + "grad_norm": 0.02722270966875372, + "learning_rate": 7.794144193147557e-06, + "loss": 0.0009, + "step": 2605 + }, + { + "epoch": 0.6919808815719597, + "grad_norm": 0.12719227925428875, + "learning_rate": 7.792323330263335e-06, + "loss": 0.0184, + "step": 2606 + }, + { + "epoch": 0.6922464152947424, + "grad_norm": 0.09183671445229785, + "learning_rate": 7.790501929052102e-06, + "loss": 0.002, + "step": 2607 + }, + { + "epoch": 0.6925119490175252, + "grad_norm": 0.1755189634331677, + "learning_rate": 7.788679989865011e-06, + "loss": 0.0255, + "step": 2608 + }, + { + "epoch": 0.692777482740308, + "grad_norm": 0.4933080991105523, + "learning_rate": 7.786857513053308e-06, + "loss": 0.1191, + "step": 2609 + }, + { + "epoch": 0.6930430164630909, + "grad_norm": 0.5656390035214872, + "learning_rate": 7.785034498968343e-06, + "loss": 0.1191, + "step": 2610 + }, + { + "epoch": 0.6933085501858736, + "grad_norm": 1.991877656255841, + "learning_rate": 7.783210947961574e-06, + "loss": 0.0815, + "step": 2611 + }, + { + "epoch": 0.6935740839086564, + "grad_norm": 0.5387468311096719, + "learning_rate": 7.78138686038456e-06, + "loss": 0.1191, + "step": 2612 + }, + { + "epoch": 0.6938396176314392, + "grad_norm": 0.5412513999748023, + "learning_rate": 7.77956223658896e-06, + "loss": 0.0149, + "step": 2613 + }, + { + "epoch": 0.694105151354222, + "grad_norm": 0.4976671358976493, + "learning_rate": 7.777737076926546e-06, + "loss": 0.1367, + "step": 2614 + }, + { + "epoch": 0.6943706850770048, + "grad_norm": 0.4970317946518801, + "learning_rate": 7.775911381749182e-06, + "loss": 0.1035, + "step": 2615 + }, + { + "epoch": 0.6946362187997875, + "grad_norm": 0.6177132039946754, + "learning_rate": 7.774085151408841e-06, + "loss": 0.0688, + "step": 2616 + }, + { + "epoch": 0.6949017525225704, + "grad_norm": 0.023614155884245598, + "learning_rate": 7.772258386257604e-06, + "loss": 0.0005, + "step": 2617 + }, + { + "epoch": 0.6951672862453532, + "grad_norm": 0.07088700172212423, + "learning_rate": 7.770431086647643e-06, + "loss": 0.0025, + "step": 2618 + }, + { + "epoch": 0.695432819968136, + "grad_norm": 0.010024093375749565, + "learning_rate": 7.768603252931243e-06, + "loss": 0.0003, + "step": 2619 + }, + { + "epoch": 0.6956983536909187, + "grad_norm": 0.22075379995274727, + "learning_rate": 7.766774885460788e-06, + "loss": 0.0391, + "step": 2620 + }, + { + "epoch": 0.6959638874137015, + "grad_norm": 0.6588844121258008, + "learning_rate": 7.764945984588767e-06, + "loss": 0.0432, + "step": 2621 + }, + { + "epoch": 0.6962294211364843, + "grad_norm": 0.3196668702324135, + "learning_rate": 7.763116550667771e-06, + "loss": 0.0476, + "step": 2622 + }, + { + "epoch": 0.6964949548592672, + "grad_norm": 0.23024507577457323, + "learning_rate": 7.76128658405049e-06, + "loss": 0.0073, + "step": 2623 + }, + { + "epoch": 0.69676048858205, + "grad_norm": 0.5136678868423971, + "learning_rate": 7.759456085089723e-06, + "loss": 0.0688, + "step": 2624 + }, + { + "epoch": 0.6970260223048327, + "grad_norm": 0.2787874389603621, + "learning_rate": 7.757625054138367e-06, + "loss": 0.0476, + "step": 2625 + }, + { + "epoch": 0.6972915560276155, + "grad_norm": 0.3557279409708689, + "learning_rate": 7.755793491549423e-06, + "loss": 0.0525, + "step": 2626 + }, + { + "epoch": 0.6975570897503983, + "grad_norm": 0.7293386235387719, + "learning_rate": 7.753961397675996e-06, + "loss": 0.0815, + "step": 2627 + }, + { + "epoch": 0.6978226234731811, + "grad_norm": 0.20618693938333246, + "learning_rate": 7.752128772871292e-06, + "loss": 0.0354, + "step": 2628 + }, + { + "epoch": 0.6980881571959638, + "grad_norm": 0.27494529156058145, + "learning_rate": 7.750295617488618e-06, + "loss": 0.0476, + "step": 2629 + }, + { + "epoch": 0.6983536909187467, + "grad_norm": 0.49066770827742867, + "learning_rate": 7.748461931881387e-06, + "loss": 0.0184, + "step": 2630 + }, + { + "epoch": 0.6986192246415295, + "grad_norm": 0.5833858940944898, + "learning_rate": 7.74662771640311e-06, + "loss": 0.0476, + "step": 2631 + }, + { + "epoch": 0.6988847583643123, + "grad_norm": 0.20861048977794938, + "learning_rate": 7.744792971407403e-06, + "loss": 0.0354, + "step": 2632 + }, + { + "epoch": 0.6991502920870951, + "grad_norm": 0.5925417691120017, + "learning_rate": 7.742957697247984e-06, + "loss": 0.0476, + "step": 2633 + }, + { + "epoch": 0.6994158258098778, + "grad_norm": 0.4286119110460914, + "learning_rate": 7.741121894278668e-06, + "loss": 0.1035, + "step": 2634 + }, + { + "epoch": 0.6996813595326606, + "grad_norm": 0.4442826736316304, + "learning_rate": 7.739285562853381e-06, + "loss": 0.0957, + "step": 2635 + }, + { + "epoch": 0.6999468932554435, + "grad_norm": 0.5095382483624094, + "learning_rate": 7.737448703326145e-06, + "loss": 0.0752, + "step": 2636 + }, + { + "epoch": 0.7002124269782263, + "grad_norm": 0.20505282395225904, + "learning_rate": 7.735611316051084e-06, + "loss": 0.0354, + "step": 2637 + }, + { + "epoch": 0.700477960701009, + "grad_norm": 0.401258602957699, + "learning_rate": 7.733773401382425e-06, + "loss": 0.0525, + "step": 2638 + }, + { + "epoch": 0.7007434944237918, + "grad_norm": 0.23090920920973987, + "learning_rate": 7.731934959674495e-06, + "loss": 0.0391, + "step": 2639 + }, + { + "epoch": 0.7010090281465746, + "grad_norm": 1.748391043734847, + "learning_rate": 7.730095991281725e-06, + "loss": 0.1738, + "step": 2640 + }, + { + "epoch": 0.7012745618693574, + "grad_norm": 0.23019013574527775, + "learning_rate": 7.728256496558649e-06, + "loss": 0.0391, + "step": 2641 + }, + { + "epoch": 0.7015400955921403, + "grad_norm": 0.46070090343648035, + "learning_rate": 7.726416475859894e-06, + "loss": 0.1035, + "step": 2642 + }, + { + "epoch": 0.701805629314923, + "grad_norm": 0.2350744963290744, + "learning_rate": 7.724575929540198e-06, + "loss": 0.0391, + "step": 2643 + }, + { + "epoch": 0.7020711630377058, + "grad_norm": 0.532425304856753, + "learning_rate": 7.722734857954396e-06, + "loss": 0.0576, + "step": 2644 + }, + { + "epoch": 0.7023366967604886, + "grad_norm": 0.5270642448347082, + "learning_rate": 7.720893261457428e-06, + "loss": 0.0206, + "step": 2645 + }, + { + "epoch": 0.7026022304832714, + "grad_norm": 0.5177346833645611, + "learning_rate": 7.719051140404327e-06, + "loss": 0.0889, + "step": 2646 + }, + { + "epoch": 0.7028677642060541, + "grad_norm": 0.20299284333726889, + "learning_rate": 7.717208495150234e-06, + "loss": 0.0354, + "step": 2647 + }, + { + "epoch": 0.7031332979288369, + "grad_norm": 1.4176997306096022, + "learning_rate": 7.71536532605039e-06, + "loss": 0.0957, + "step": 2648 + }, + { + "epoch": 0.7033988316516198, + "grad_norm": 0.41291088284746635, + "learning_rate": 7.713521633460138e-06, + "loss": 0.0525, + "step": 2649 + }, + { + "epoch": 0.7036643653744026, + "grad_norm": 1.226769408819419, + "learning_rate": 7.711677417734915e-06, + "loss": 0.1113, + "step": 2650 + }, + { + "epoch": 0.7039298990971854, + "grad_norm": 0.3297164221483673, + "learning_rate": 7.709832679230268e-06, + "loss": 0.0476, + "step": 2651 + }, + { + "epoch": 0.7041954328199681, + "grad_norm": 0.6622100907025877, + "learning_rate": 7.707987418301838e-06, + "loss": 0.0432, + "step": 2652 + }, + { + "epoch": 0.7044609665427509, + "grad_norm": 0.40251558797268133, + "learning_rate": 7.706141635305376e-06, + "loss": 0.0525, + "step": 2653 + }, + { + "epoch": 0.7047265002655337, + "grad_norm": 0.59157610342596, + "learning_rate": 7.70429533059672e-06, + "loss": 0.0576, + "step": 2654 + }, + { + "epoch": 0.7049920339883166, + "grad_norm": 0.023221331131184016, + "learning_rate": 7.702448504531818e-06, + "loss": 0.0006, + "step": 2655 + }, + { + "epoch": 0.7052575677110993, + "grad_norm": 0.8839263576153142, + "learning_rate": 7.700601157466718e-06, + "loss": 0.063, + "step": 2656 + }, + { + "epoch": 0.7055231014338821, + "grad_norm": 0.058379575728156614, + "learning_rate": 7.698753289757565e-06, + "loss": 0.0013, + "step": 2657 + }, + { + "epoch": 0.7057886351566649, + "grad_norm": 0.14525001558765177, + "learning_rate": 7.696904901760607e-06, + "loss": 0.0229, + "step": 2658 + }, + { + "epoch": 0.7060541688794477, + "grad_norm": 0.06713110591905418, + "learning_rate": 7.69505599383219e-06, + "loss": 0.0019, + "step": 2659 + }, + { + "epoch": 0.7063197026022305, + "grad_norm": 0.14213438112144283, + "learning_rate": 7.693206566328764e-06, + "loss": 0.0231, + "step": 2660 + }, + { + "epoch": 0.7065852363250132, + "grad_norm": 0.16047726283972039, + "learning_rate": 7.691356619606873e-06, + "loss": 0.0255, + "step": 2661 + }, + { + "epoch": 0.7068507700477961, + "grad_norm": 0.18623510810979663, + "learning_rate": 7.68950615402317e-06, + "loss": 0.0286, + "step": 2662 + }, + { + "epoch": 0.7071163037705789, + "grad_norm": 0.011228514736014899, + "learning_rate": 7.687655169934399e-06, + "loss": 0.0004, + "step": 2663 + }, + { + "epoch": 0.7073818374933617, + "grad_norm": 0.8774799228434651, + "learning_rate": 7.68580366769741e-06, + "loss": 0.0286, + "step": 2664 + }, + { + "epoch": 0.7076473712161444, + "grad_norm": 0.14067959118875095, + "learning_rate": 7.68395164766915e-06, + "loss": 0.0231, + "step": 2665 + }, + { + "epoch": 0.7079129049389272, + "grad_norm": 0.5157995810946556, + "learning_rate": 7.682099110206665e-06, + "loss": 0.1367, + "step": 2666 + }, + { + "epoch": 0.70817843866171, + "grad_norm": 0.5180006365680498, + "learning_rate": 7.680246055667106e-06, + "loss": 0.1279, + "step": 2667 + }, + { + "epoch": 0.7084439723844929, + "grad_norm": 0.2641640007270527, + "learning_rate": 7.678392484407718e-06, + "loss": 0.0092, + "step": 2668 + }, + { + "epoch": 0.7087095061072757, + "grad_norm": 0.11996519469140099, + "learning_rate": 7.676538396785846e-06, + "loss": 0.0184, + "step": 2669 + }, + { + "epoch": 0.7089750398300584, + "grad_norm": 0.2795175123956674, + "learning_rate": 7.674683793158938e-06, + "loss": 0.0032, + "step": 2670 + }, + { + "epoch": 0.7092405735528412, + "grad_norm": 0.1571200917447199, + "learning_rate": 7.672828673884541e-06, + "loss": 0.0258, + "step": 2671 + }, + { + "epoch": 0.709506107275624, + "grad_norm": 0.20336199311311373, + "learning_rate": 7.670973039320297e-06, + "loss": 0.0286, + "step": 2672 + }, + { + "epoch": 0.7097716409984068, + "grad_norm": 0.24844888213425997, + "learning_rate": 7.669116889823955e-06, + "loss": 0.0354, + "step": 2673 + }, + { + "epoch": 0.7100371747211895, + "grad_norm": 0.16767051088439403, + "learning_rate": 7.667260225753353e-06, + "loss": 0.0229, + "step": 2674 + }, + { + "epoch": 0.7103027084439724, + "grad_norm": 0.006744534699103974, + "learning_rate": 7.665403047466438e-06, + "loss": 0.0002, + "step": 2675 + }, + { + "epoch": 0.7105682421667552, + "grad_norm": 0.1555660374195956, + "learning_rate": 7.663545355321252e-06, + "loss": 0.0255, + "step": 2676 + }, + { + "epoch": 0.710833775889538, + "grad_norm": 0.5525927253925721, + "learning_rate": 7.661687149675934e-06, + "loss": 0.1367, + "step": 2677 + }, + { + "epoch": 0.7110993096123208, + "grad_norm": 0.14441213610546808, + "learning_rate": 7.659828430888726e-06, + "loss": 0.0231, + "step": 2678 + }, + { + "epoch": 0.7113648433351035, + "grad_norm": 0.6473219973422136, + "learning_rate": 7.657969199317964e-06, + "loss": 0.1191, + "step": 2679 + }, + { + "epoch": 0.7116303770578863, + "grad_norm": 0.011319335491774244, + "learning_rate": 7.656109455322092e-06, + "loss": 0.0004, + "step": 2680 + }, + { + "epoch": 0.7118959107806692, + "grad_norm": 0.010866134886274215, + "learning_rate": 7.65424919925964e-06, + "loss": 0.0002, + "step": 2681 + }, + { + "epoch": 0.712161444503452, + "grad_norm": 0.5967921985512348, + "learning_rate": 7.652388431489248e-06, + "loss": 0.1113, + "step": 2682 + }, + { + "epoch": 0.7124269782262347, + "grad_norm": 0.15087391504094505, + "learning_rate": 7.650527152369646e-06, + "loss": 0.0206, + "step": 2683 + }, + { + "epoch": 0.7126925119490175, + "grad_norm": 0.16895116757294115, + "learning_rate": 7.648665362259672e-06, + "loss": 0.0255, + "step": 2684 + }, + { + "epoch": 0.7129580456718003, + "grad_norm": 0.5675974491722848, + "learning_rate": 7.646803061518254e-06, + "loss": 0.1455, + "step": 2685 + }, + { + "epoch": 0.7132235793945831, + "grad_norm": 0.5647596868107823, + "learning_rate": 7.64494025050442e-06, + "loss": 0.1191, + "step": 2686 + }, + { + "epoch": 0.713489113117366, + "grad_norm": 0.4995756873480659, + "learning_rate": 7.6430769295773e-06, + "loss": 0.1279, + "step": 2687 + }, + { + "epoch": 0.7137546468401487, + "grad_norm": 0.3672241880277148, + "learning_rate": 7.64121309909612e-06, + "loss": 0.0391, + "step": 2688 + }, + { + "epoch": 0.7140201805629315, + "grad_norm": 0.19593687681328856, + "learning_rate": 7.639348759420207e-06, + "loss": 0.0286, + "step": 2689 + }, + { + "epoch": 0.7142857142857143, + "grad_norm": 0.5768227079240762, + "learning_rate": 7.63748391090898e-06, + "loss": 0.1035, + "step": 2690 + }, + { + "epoch": 0.7145512480084971, + "grad_norm": 0.39584647324038486, + "learning_rate": 7.635618553921962e-06, + "loss": 0.0432, + "step": 2691 + }, + { + "epoch": 0.7148167817312798, + "grad_norm": 0.18613267005731784, + "learning_rate": 7.63375268881877e-06, + "loss": 0.0317, + "step": 2692 + }, + { + "epoch": 0.7150823154540626, + "grad_norm": 0.008415169027936919, + "learning_rate": 7.631886315959121e-06, + "loss": 0.0002, + "step": 2693 + }, + { + "epoch": 0.7153478491768455, + "grad_norm": 0.36401181634248886, + "learning_rate": 7.630019435702832e-06, + "loss": 0.0391, + "step": 2694 + }, + { + "epoch": 0.7156133828996283, + "grad_norm": 0.10628893373443327, + "learning_rate": 7.628152048409814e-06, + "loss": 0.0035, + "step": 2695 + }, + { + "epoch": 0.7158789166224111, + "grad_norm": 0.8279777048989058, + "learning_rate": 7.626284154440078e-06, + "loss": 0.0258, + "step": 2696 + }, + { + "epoch": 0.7161444503451938, + "grad_norm": 0.5654985838880184, + "learning_rate": 7.624415754153729e-06, + "loss": 0.0231, + "step": 2697 + }, + { + "epoch": 0.7164099840679766, + "grad_norm": 0.45888723924718394, + "learning_rate": 7.622546847910977e-06, + "loss": 0.0957, + "step": 2698 + }, + { + "epoch": 0.7166755177907594, + "grad_norm": 0.474322812067065, + "learning_rate": 7.620677436072121e-06, + "loss": 0.1191, + "step": 2699 + }, + { + "epoch": 0.7169410515135423, + "grad_norm": 0.09808489791928204, + "learning_rate": 7.6188075189975644e-06, + "loss": 0.0023, + "step": 2700 + }, + { + "epoch": 0.717206585236325, + "grad_norm": 0.6343999408686629, + "learning_rate": 7.616937097047803e-06, + "loss": 0.0889, + "step": 2701 + }, + { + "epoch": 0.7174721189591078, + "grad_norm": 0.20305979257782597, + "learning_rate": 7.615066170583434e-06, + "loss": 0.0354, + "step": 2702 + }, + { + "epoch": 0.7177376526818906, + "grad_norm": 0.6260097099120551, + "learning_rate": 7.613194739965148e-06, + "loss": 0.0476, + "step": 2703 + }, + { + "epoch": 0.7180031864046734, + "grad_norm": 0.434264403873269, + "learning_rate": 7.611322805553738e-06, + "loss": 0.0889, + "step": 2704 + }, + { + "epoch": 0.7182687201274562, + "grad_norm": 0.039927693136070204, + "learning_rate": 7.609450367710087e-06, + "loss": 0.0009, + "step": 2705 + }, + { + "epoch": 0.718534253850239, + "grad_norm": 0.24970586840281603, + "learning_rate": 7.607577426795181e-06, + "loss": 0.0432, + "step": 2706 + }, + { + "epoch": 0.7187997875730218, + "grad_norm": 0.28036855986479214, + "learning_rate": 7.605703983170102e-06, + "loss": 0.0476, + "step": 2707 + }, + { + "epoch": 0.7190653212958046, + "grad_norm": 0.19509515727624133, + "learning_rate": 7.603830037196025e-06, + "loss": 0.0317, + "step": 2708 + }, + { + "epoch": 0.7193308550185874, + "grad_norm": 0.25430344092499496, + "learning_rate": 7.601955589234227e-06, + "loss": 0.0432, + "step": 2709 + }, + { + "epoch": 0.7195963887413701, + "grad_norm": 0.0036134043798567274, + "learning_rate": 7.600080639646077e-06, + "loss": 0.0001, + "step": 2710 + }, + { + "epoch": 0.7198619224641529, + "grad_norm": 0.4916283201554719, + "learning_rate": 7.5982051887930474e-06, + "loss": 0.0688, + "step": 2711 + }, + { + "epoch": 0.7201274561869357, + "grad_norm": 0.010633091364110748, + "learning_rate": 7.596329237036699e-06, + "loss": 0.0003, + "step": 2712 + }, + { + "epoch": 0.7203929899097186, + "grad_norm": 0.0023510959149679977, + "learning_rate": 7.5944527847386954e-06, + "loss": 0.0001, + "step": 2713 + }, + { + "epoch": 0.7206585236325014, + "grad_norm": 0.0008853267644760596, + "learning_rate": 7.5925758322607925e-06, + "loss": 0.0, + "step": 2714 + }, + { + "epoch": 0.7209240573552841, + "grad_norm": 0.39772817869286603, + "learning_rate": 7.590698379964845e-06, + "loss": 0.0815, + "step": 2715 + }, + { + "epoch": 0.7211895910780669, + "grad_norm": 0.5822216124324845, + "learning_rate": 7.588820428212805e-06, + "loss": 0.0476, + "step": 2716 + }, + { + "epoch": 0.7214551248008497, + "grad_norm": 0.004927685282107352, + "learning_rate": 7.58694197736672e-06, + "loss": 0.0001, + "step": 2717 + }, + { + "epoch": 0.7217206585236325, + "grad_norm": 0.8975099705217199, + "learning_rate": 7.58506302778873e-06, + "loss": 0.0815, + "step": 2718 + }, + { + "epoch": 0.7219861922464152, + "grad_norm": 0.7224995572702415, + "learning_rate": 7.583183579841078e-06, + "loss": 0.0258, + "step": 2719 + }, + { + "epoch": 0.7222517259691981, + "grad_norm": 0.23353276759847078, + "learning_rate": 7.581303633886097e-06, + "loss": 0.0391, + "step": 2720 + }, + { + "epoch": 0.7225172596919809, + "grad_norm": 2.666975891176335, + "learning_rate": 7.5794231902862195e-06, + "loss": 0.2988, + "step": 2721 + }, + { + "epoch": 0.7227827934147637, + "grad_norm": 0.28295768293920587, + "learning_rate": 7.5775422494039705e-06, + "loss": 0.0476, + "step": 2722 + }, + { + "epoch": 0.7230483271375465, + "grad_norm": 0.4381524130622177, + "learning_rate": 7.5756608116019766e-06, + "loss": 0.0476, + "step": 2723 + }, + { + "epoch": 0.7233138608603292, + "grad_norm": 9.717942824587892, + "learning_rate": 7.573778877242954e-06, + "loss": 0.4316, + "step": 2724 + }, + { + "epoch": 0.723579394583112, + "grad_norm": 0.40899089741591693, + "learning_rate": 7.571896446689721e-06, + "loss": 0.0889, + "step": 2725 + }, + { + "epoch": 0.7238449283058949, + "grad_norm": 1.295789465643148, + "learning_rate": 7.570013520305184e-06, + "loss": 0.0957, + "step": 2726 + }, + { + "epoch": 0.7241104620286777, + "grad_norm": 0.4112960920275529, + "learning_rate": 7.568130098452352e-06, + "loss": 0.0576, + "step": 2727 + }, + { + "epoch": 0.7243759957514604, + "grad_norm": 0.6506136032567195, + "learning_rate": 7.566246181494325e-06, + "loss": 0.0317, + "step": 2728 + }, + { + "epoch": 0.7246415294742432, + "grad_norm": 0.9317198447206394, + "learning_rate": 7.5643617697943e-06, + "loss": 0.0752, + "step": 2729 + }, + { + "epoch": 0.724907063197026, + "grad_norm": 0.20034810343290435, + "learning_rate": 7.56247686371557e-06, + "loss": 0.0317, + "step": 2730 + }, + { + "epoch": 0.7251725969198088, + "grad_norm": 0.41392125645699196, + "learning_rate": 7.5605914636215216e-06, + "loss": 0.0957, + "step": 2731 + }, + { + "epoch": 0.7254381306425917, + "grad_norm": 0.21828991435416753, + "learning_rate": 7.558705569875639e-06, + "loss": 0.0391, + "step": 2732 + }, + { + "epoch": 0.7257036643653744, + "grad_norm": 0.2333115794088071, + "learning_rate": 7.556819182841498e-06, + "loss": 0.0432, + "step": 2733 + }, + { + "epoch": 0.7259691980881572, + "grad_norm": 0.06610504820163955, + "learning_rate": 7.554932302882776e-06, + "loss": 0.0022, + "step": 2734 + }, + { + "epoch": 0.72623473181094, + "grad_norm": 0.00262623888277028, + "learning_rate": 7.553044930363237e-06, + "loss": 0.0001, + "step": 2735 + }, + { + "epoch": 0.7265002655337228, + "grad_norm": 0.4395975478833954, + "learning_rate": 7.551157065646747e-06, + "loss": 0.0957, + "step": 2736 + }, + { + "epoch": 0.7267657992565055, + "grad_norm": 0.6391809330118845, + "learning_rate": 7.549268709097262e-06, + "loss": 0.0752, + "step": 2737 + }, + { + "epoch": 0.7270313329792883, + "grad_norm": 0.2453814818990566, + "learning_rate": 7.5473798610788355e-06, + "loss": 0.0391, + "step": 2738 + }, + { + "epoch": 0.7272968667020712, + "grad_norm": 0.49520396768253455, + "learning_rate": 7.545490521955615e-06, + "loss": 0.063, + "step": 2739 + }, + { + "epoch": 0.727562400424854, + "grad_norm": 0.3349773174381591, + "learning_rate": 7.543600692091843e-06, + "loss": 0.0391, + "step": 2740 + }, + { + "epoch": 0.7278279341476368, + "grad_norm": 0.003769099000278558, + "learning_rate": 7.541710371851858e-06, + "loss": 0.0001, + "step": 2741 + }, + { + "epoch": 0.7280934678704195, + "grad_norm": 0.47841766086734994, + "learning_rate": 7.539819561600089e-06, + "loss": 0.0752, + "step": 2742 + }, + { + "epoch": 0.7283590015932023, + "grad_norm": 0.41209736161364247, + "learning_rate": 7.537928261701064e-06, + "loss": 0.0957, + "step": 2743 + }, + { + "epoch": 0.7286245353159851, + "grad_norm": 0.24857331638889915, + "learning_rate": 7.536036472519401e-06, + "loss": 0.0391, + "step": 2744 + }, + { + "epoch": 0.728890069038768, + "grad_norm": 0.4013596518437219, + "learning_rate": 7.534144194419817e-06, + "loss": 0.0576, + "step": 2745 + }, + { + "epoch": 0.7291556027615507, + "grad_norm": 0.2331854223123911, + "learning_rate": 7.532251427767119e-06, + "loss": 0.0391, + "step": 2746 + }, + { + "epoch": 0.7294211364843335, + "grad_norm": 0.25313810084990956, + "learning_rate": 7.530358172926211e-06, + "loss": 0.0432, + "step": 2747 + }, + { + "epoch": 0.7296866702071163, + "grad_norm": 0.3953485455537849, + "learning_rate": 7.5284644302620906e-06, + "loss": 0.0525, + "step": 2748 + }, + { + "epoch": 0.7299522039298991, + "grad_norm": 0.44215113588654426, + "learning_rate": 7.526570200139846e-06, + "loss": 0.0576, + "step": 2749 + }, + { + "epoch": 0.7302177376526819, + "grad_norm": 0.8783447362171677, + "learning_rate": 7.5246754829246685e-06, + "loss": 0.0206, + "step": 2750 + }, + { + "epoch": 0.7304832713754646, + "grad_norm": 1.2522110222840543, + "learning_rate": 7.5227802789818315e-06, + "loss": 0.0659, + "step": 2751 + }, + { + "epoch": 0.7307488050982475, + "grad_norm": 0.5193319702841309, + "learning_rate": 7.520884588676711e-06, + "loss": 0.1113, + "step": 2752 + }, + { + "epoch": 0.7310143388210303, + "grad_norm": 0.8712011079193801, + "learning_rate": 7.5189884123747725e-06, + "loss": 0.032, + "step": 2753 + }, + { + "epoch": 0.7312798725438131, + "grad_norm": 0.5655894111141081, + "learning_rate": 7.517091750441576e-06, + "loss": 0.063, + "step": 2754 + }, + { + "epoch": 0.7315454062665958, + "grad_norm": 0.4333168007453955, + "learning_rate": 7.515194603242777e-06, + "loss": 0.0957, + "step": 2755 + }, + { + "epoch": 0.7318109399893786, + "grad_norm": 0.5871227062661633, + "learning_rate": 7.513296971144124e-06, + "loss": 0.0576, + "step": 2756 + }, + { + "epoch": 0.7320764737121614, + "grad_norm": 0.025503086800507994, + "learning_rate": 7.511398854511455e-06, + "loss": 0.0005, + "step": 2757 + }, + { + "epoch": 0.7323420074349443, + "grad_norm": 0.04427058564643331, + "learning_rate": 7.509500253710706e-06, + "loss": 0.0013, + "step": 2758 + }, + { + "epoch": 0.7326075411577271, + "grad_norm": 0.3480723421970973, + "learning_rate": 7.5076011691079055e-06, + "loss": 0.0525, + "step": 2759 + }, + { + "epoch": 0.7328730748805098, + "grad_norm": 2.938031793623133, + "learning_rate": 7.505701601069173e-06, + "loss": 0.4434, + "step": 2760 + }, + { + "epoch": 0.7331386086032926, + "grad_norm": 0.9388804031645489, + "learning_rate": 7.503801549960726e-06, + "loss": 0.0688, + "step": 2761 + }, + { + "epoch": 0.7334041423260754, + "grad_norm": 0.4390417412233155, + "learning_rate": 7.501901016148867e-06, + "loss": 0.063, + "step": 2762 + }, + { + "epoch": 0.7336696760488582, + "grad_norm": 0.19622847405034344, + "learning_rate": 7.500000000000001e-06, + "loss": 0.0317, + "step": 2763 + }, + { + "epoch": 0.733935209771641, + "grad_norm": 0.2425271230814878, + "learning_rate": 7.4980985018806184e-06, + "loss": 0.0432, + "step": 2764 + }, + { + "epoch": 0.7342007434944238, + "grad_norm": 0.029397041848983584, + "learning_rate": 7.496196522157309e-06, + "loss": 0.0009, + "step": 2765 + }, + { + "epoch": 0.7344662772172066, + "grad_norm": 0.003234886394832873, + "learning_rate": 7.494294061196749e-06, + "loss": 0.0001, + "step": 2766 + }, + { + "epoch": 0.7347318109399894, + "grad_norm": 0.4737014595507085, + "learning_rate": 7.492391119365711e-06, + "loss": 0.0576, + "step": 2767 + }, + { + "epoch": 0.7349973446627722, + "grad_norm": 0.2659167981367224, + "learning_rate": 7.490487697031061e-06, + "loss": 0.0476, + "step": 2768 + }, + { + "epoch": 0.7352628783855549, + "grad_norm": 0.0358032798767517, + "learning_rate": 7.488583794559754e-06, + "loss": 0.0013, + "step": 2769 + }, + { + "epoch": 0.7355284121083377, + "grad_norm": 0.5916429849192902, + "learning_rate": 7.486679412318844e-06, + "loss": 0.0688, + "step": 2770 + }, + { + "epoch": 0.7357939458311206, + "grad_norm": 0.02719321454575265, + "learning_rate": 7.484774550675469e-06, + "loss": 0.0009, + "step": 2771 + }, + { + "epoch": 0.7360594795539034, + "grad_norm": 3.5689930891892967, + "learning_rate": 7.482869209996867e-06, + "loss": 0.1641, + "step": 2772 + }, + { + "epoch": 0.7363250132766861, + "grad_norm": 0.4458525072675684, + "learning_rate": 7.480963390650362e-06, + "loss": 0.0752, + "step": 2773 + }, + { + "epoch": 0.7365905469994689, + "grad_norm": 0.408602463820852, + "learning_rate": 7.479057093003378e-06, + "loss": 0.1035, + "step": 2774 + }, + { + "epoch": 0.7368560807222517, + "grad_norm": 0.5395624215803067, + "learning_rate": 7.477150317423423e-06, + "loss": 0.0693, + "step": 2775 + }, + { + "epoch": 0.7371216144450345, + "grad_norm": 0.2370112533000475, + "learning_rate": 7.475243064278104e-06, + "loss": 0.0432, + "step": 2776 + }, + { + "epoch": 0.7373871481678174, + "grad_norm": 0.5804310837563629, + "learning_rate": 7.473335333935112e-06, + "loss": 0.063, + "step": 2777 + }, + { + "epoch": 0.7376526818906001, + "grad_norm": 0.18417279828276742, + "learning_rate": 7.4714271267622395e-06, + "loss": 0.0354, + "step": 2778 + }, + { + "epoch": 0.7379182156133829, + "grad_norm": 0.4378722467910318, + "learning_rate": 7.469518443127364e-06, + "loss": 0.0889, + "step": 2779 + }, + { + "epoch": 0.7381837493361657, + "grad_norm": 0.4792190845518073, + "learning_rate": 7.467609283398459e-06, + "loss": 0.0258, + "step": 2780 + }, + { + "epoch": 0.7384492830589485, + "grad_norm": 0.2773001270308989, + "learning_rate": 7.465699647943586e-06, + "loss": 0.0476, + "step": 2781 + }, + { + "epoch": 0.7387148167817312, + "grad_norm": 0.37224569921836664, + "learning_rate": 7.463789537130901e-06, + "loss": 0.0476, + "step": 2782 + }, + { + "epoch": 0.738980350504514, + "grad_norm": 0.3923116421504141, + "learning_rate": 7.461878951328653e-06, + "loss": 0.0476, + "step": 2783 + }, + { + "epoch": 0.7392458842272969, + "grad_norm": 0.46458843955184786, + "learning_rate": 7.459967890905177e-06, + "loss": 0.0525, + "step": 2784 + }, + { + "epoch": 0.7395114179500797, + "grad_norm": 0.12032410519435113, + "learning_rate": 7.458056356228903e-06, + "loss": 0.0045, + "step": 2785 + }, + { + "epoch": 0.7397769516728625, + "grad_norm": 0.21407988444306555, + "learning_rate": 7.456144347668354e-06, + "loss": 0.0391, + "step": 2786 + }, + { + "epoch": 0.7400424853956452, + "grad_norm": 0.21201797691611676, + "learning_rate": 7.454231865592143e-06, + "loss": 0.0391, + "step": 2787 + }, + { + "epoch": 0.740308019118428, + "grad_norm": 0.24289775923774354, + "learning_rate": 7.452318910368973e-06, + "loss": 0.0432, + "step": 2788 + }, + { + "epoch": 0.7405735528412108, + "grad_norm": 0.2930353531141809, + "learning_rate": 7.4504054823676394e-06, + "loss": 0.0476, + "step": 2789 + }, + { + "epoch": 0.7408390865639937, + "grad_norm": 0.048155391371814744, + "learning_rate": 7.44849158195703e-06, + "loss": 0.0013, + "step": 2790 + }, + { + "epoch": 0.7411046202867764, + "grad_norm": 0.16293031727884294, + "learning_rate": 7.44657720950612e-06, + "loss": 0.0286, + "step": 2791 + }, + { + "epoch": 0.7413701540095592, + "grad_norm": 0.00628053217419269, + "learning_rate": 7.444662365383979e-06, + "loss": 0.0002, + "step": 2792 + }, + { + "epoch": 0.741635687732342, + "grad_norm": 0.228857492227635, + "learning_rate": 7.442747049959765e-06, + "loss": 0.0354, + "step": 2793 + }, + { + "epoch": 0.7419012214551248, + "grad_norm": 0.023366755121011988, + "learning_rate": 7.44083126360273e-06, + "loss": 0.0008, + "step": 2794 + }, + { + "epoch": 0.7421667551779076, + "grad_norm": 0.6557877711658092, + "learning_rate": 7.438915006682214e-06, + "loss": 0.0889, + "step": 2795 + }, + { + "epoch": 0.7424322889006904, + "grad_norm": 0.19934753887423692, + "learning_rate": 7.436998279567649e-06, + "loss": 0.0317, + "step": 2796 + }, + { + "epoch": 0.7426978226234732, + "grad_norm": 0.012922332241952057, + "learning_rate": 7.435081082628557e-06, + "loss": 0.0004, + "step": 2797 + }, + { + "epoch": 0.742963356346256, + "grad_norm": 1.9626574571147495, + "learning_rate": 7.433163416234552e-06, + "loss": 0.1641, + "step": 2798 + }, + { + "epoch": 0.7432288900690388, + "grad_norm": 0.020930515495406296, + "learning_rate": 7.4312452807553374e-06, + "loss": 0.0007, + "step": 2799 + }, + { + "epoch": 0.7434944237918215, + "grad_norm": 0.15147896003836883, + "learning_rate": 7.429326676560706e-06, + "loss": 0.0258, + "step": 2800 + }, + { + "epoch": 0.7437599575146043, + "grad_norm": 0.060357317533153716, + "learning_rate": 7.427407604020543e-06, + "loss": 0.0022, + "step": 2801 + }, + { + "epoch": 0.7440254912373871, + "grad_norm": 0.5185650048959446, + "learning_rate": 7.425488063504822e-06, + "loss": 0.1113, + "step": 2802 + }, + { + "epoch": 0.74429102496017, + "grad_norm": 0.20298155363983883, + "learning_rate": 7.42356805538361e-06, + "loss": 0.0317, + "step": 2803 + }, + { + "epoch": 0.7445565586829528, + "grad_norm": 0.04788201196025843, + "learning_rate": 7.421647580027058e-06, + "loss": 0.0017, + "step": 2804 + }, + { + "epoch": 0.7448220924057355, + "grad_norm": 0.158970679347593, + "learning_rate": 7.419726637805413e-06, + "loss": 0.0258, + "step": 2805 + }, + { + "epoch": 0.7450876261285183, + "grad_norm": 0.8081236442111026, + "learning_rate": 7.417805229089013e-06, + "loss": 0.063, + "step": 2806 + }, + { + "epoch": 0.7453531598513011, + "grad_norm": 0.1684577456349591, + "learning_rate": 7.415883354248276e-06, + "loss": 0.0286, + "step": 2807 + }, + { + "epoch": 0.745618693574084, + "grad_norm": 0.26656684659495494, + "learning_rate": 7.413961013653725e-06, + "loss": 0.0354, + "step": 2808 + }, + { + "epoch": 0.7458842272968667, + "grad_norm": 0.1812420365192952, + "learning_rate": 7.412038207675959e-06, + "loss": 0.0286, + "step": 2809 + }, + { + "epoch": 0.7461497610196495, + "grad_norm": 0.53682256190488, + "learning_rate": 7.4101149366856736e-06, + "loss": 0.1113, + "step": 2810 + }, + { + "epoch": 0.7464152947424323, + "grad_norm": 0.29334575241258665, + "learning_rate": 7.408191201053654e-06, + "loss": 0.0354, + "step": 2811 + }, + { + "epoch": 0.7466808284652151, + "grad_norm": 0.6540253294648104, + "learning_rate": 7.406267001150773e-06, + "loss": 0.0525, + "step": 2812 + }, + { + "epoch": 0.7469463621879979, + "grad_norm": 0.14605287172847867, + "learning_rate": 7.404342337347993e-06, + "loss": 0.0231, + "step": 2813 + }, + { + "epoch": 0.7472118959107806, + "grad_norm": 0.16682845997440088, + "learning_rate": 7.402417210016365e-06, + "loss": 0.0206, + "step": 2814 + }, + { + "epoch": 0.7474774296335635, + "grad_norm": 0.14355281090707867, + "learning_rate": 7.400491619527036e-06, + "loss": 0.0206, + "step": 2815 + }, + { + "epoch": 0.7477429633563463, + "grad_norm": 0.2635639116023051, + "learning_rate": 7.398565566251232e-06, + "loss": 0.0104, + "step": 2816 + }, + { + "epoch": 0.7480084970791291, + "grad_norm": 0.7242410224098833, + "learning_rate": 7.3966390505602755e-06, + "loss": 0.1279, + "step": 2817 + }, + { + "epoch": 0.7482740308019118, + "grad_norm": 0.11594214304627717, + "learning_rate": 7.394712072825576e-06, + "loss": 0.0165, + "step": 2818 + }, + { + "epoch": 0.7485395645246946, + "grad_norm": 0.05188530081119729, + "learning_rate": 7.392784633418632e-06, + "loss": 0.0013, + "step": 2819 + }, + { + "epoch": 0.7488050982474774, + "grad_norm": 0.14322025524767493, + "learning_rate": 7.39085673271103e-06, + "loss": 0.0231, + "step": 2820 + }, + { + "epoch": 0.7490706319702602, + "grad_norm": 0.3460088532572021, + "learning_rate": 7.3889283710744485e-06, + "loss": 0.0317, + "step": 2821 + }, + { + "epoch": 0.7493361656930431, + "grad_norm": 0.12482180328349493, + "learning_rate": 7.386999548880652e-06, + "loss": 0.0184, + "step": 2822 + }, + { + "epoch": 0.7496016994158258, + "grad_norm": 0.2651764903902242, + "learning_rate": 7.385070266501495e-06, + "loss": 0.0258, + "step": 2823 + }, + { + "epoch": 0.7498672331386086, + "grad_norm": 1.259378609167247, + "learning_rate": 7.3831405243089186e-06, + "loss": 0.0957, + "step": 2824 + }, + { + "epoch": 0.7501327668613914, + "grad_norm": 0.692956198979465, + "learning_rate": 7.381210322674956e-06, + "loss": 0.1035, + "step": 2825 + }, + { + "epoch": 0.7503983005841742, + "grad_norm": 0.026953390680846553, + "learning_rate": 7.379279661971728e-06, + "loss": 0.0006, + "step": 2826 + }, + { + "epoch": 0.7506638343069569, + "grad_norm": 0.16139886134789452, + "learning_rate": 7.3773485425714395e-06, + "loss": 0.0231, + "step": 2827 + }, + { + "epoch": 0.7509293680297398, + "grad_norm": 0.6285902597919525, + "learning_rate": 7.3754169648463924e-06, + "loss": 0.1641, + "step": 2828 + }, + { + "epoch": 0.7511949017525226, + "grad_norm": 0.3073207253039091, + "learning_rate": 7.37348492916897e-06, + "loss": 0.0117, + "step": 2829 + }, + { + "epoch": 0.7514604354753054, + "grad_norm": 0.13537958651498305, + "learning_rate": 7.371552435911646e-06, + "loss": 0.0184, + "step": 2830 + }, + { + "epoch": 0.7517259691980882, + "grad_norm": 0.5242031096495455, + "learning_rate": 7.36961948544698e-06, + "loss": 0.0206, + "step": 2831 + }, + { + "epoch": 0.7519915029208709, + "grad_norm": 0.6581575955119578, + "learning_rate": 7.367686078147626e-06, + "loss": 0.0576, + "step": 2832 + }, + { + "epoch": 0.7522570366436537, + "grad_norm": 0.6278492932988957, + "learning_rate": 7.365752214386321e-06, + "loss": 0.1553, + "step": 2833 + }, + { + "epoch": 0.7525225703664365, + "grad_norm": 3.598159480139266, + "learning_rate": 7.363817894535889e-06, + "loss": 0.2344, + "step": 2834 + }, + { + "epoch": 0.7527881040892194, + "grad_norm": 0.241102019886023, + "learning_rate": 7.361883118969248e-06, + "loss": 0.0317, + "step": 2835 + }, + { + "epoch": 0.7530536378120021, + "grad_norm": 0.5802688114195329, + "learning_rate": 7.3599478880593945e-06, + "loss": 0.1367, + "step": 2836 + }, + { + "epoch": 0.7533191715347849, + "grad_norm": 2.5110491527488805, + "learning_rate": 7.358012202179423e-06, + "loss": 0.2129, + "step": 2837 + }, + { + "epoch": 0.7535847052575677, + "grad_norm": 0.2056028695075356, + "learning_rate": 7.356076061702509e-06, + "loss": 0.0286, + "step": 2838 + }, + { + "epoch": 0.7538502389803505, + "grad_norm": 0.35848217654779274, + "learning_rate": 7.354139467001917e-06, + "loss": 0.0432, + "step": 2839 + }, + { + "epoch": 0.7541157727031333, + "grad_norm": 0.48256783300414496, + "learning_rate": 7.3522024184510025e-06, + "loss": 0.0432, + "step": 2840 + }, + { + "epoch": 0.754381306425916, + "grad_norm": 0.47980975942960147, + "learning_rate": 7.3502649164232e-06, + "loss": 0.1367, + "step": 2841 + }, + { + "epoch": 0.7546468401486989, + "grad_norm": 0.25630470892659724, + "learning_rate": 7.348326961292042e-06, + "loss": 0.0354, + "step": 2842 + }, + { + "epoch": 0.7549123738714817, + "grad_norm": 0.1657920619211148, + "learning_rate": 7.346388553431141e-06, + "loss": 0.0258, + "step": 2843 + }, + { + "epoch": 0.7551779075942645, + "grad_norm": 0.12340433023106545, + "learning_rate": 7.3444496932142e-06, + "loss": 0.0206, + "step": 2844 + }, + { + "epoch": 0.7554434413170472, + "grad_norm": 0.15867030398696375, + "learning_rate": 7.342510381015009e-06, + "loss": 0.0258, + "step": 2845 + }, + { + "epoch": 0.75570897503983, + "grad_norm": 0.45687395931971003, + "learning_rate": 7.340570617207444e-06, + "loss": 0.1279, + "step": 2846 + }, + { + "epoch": 0.7559745087626129, + "grad_norm": 0.1291270689450379, + "learning_rate": 7.338630402165469e-06, + "loss": 0.0206, + "step": 2847 + }, + { + "epoch": 0.7562400424853957, + "grad_norm": 0.20693539516262213, + "learning_rate": 7.336689736263134e-06, + "loss": 0.0317, + "step": 2848 + }, + { + "epoch": 0.7565055762081785, + "grad_norm": 0.13177441244115556, + "learning_rate": 7.3347486198745765e-06, + "loss": 0.0231, + "step": 2849 + }, + { + "epoch": 0.7567711099309612, + "grad_norm": 0.02717583875549814, + "learning_rate": 7.332807053374019e-06, + "loss": 0.0009, + "step": 2850 + }, + { + "epoch": 0.757036643653744, + "grad_norm": 0.19829545402439186, + "learning_rate": 7.330865037135776e-06, + "loss": 0.0317, + "step": 2851 + }, + { + "epoch": 0.7573021773765268, + "grad_norm": 0.012724002731151448, + "learning_rate": 7.328922571534244e-06, + "loss": 0.0004, + "step": 2852 + }, + { + "epoch": 0.7575677110993096, + "grad_norm": 0.15284325316893452, + "learning_rate": 7.326979656943907e-06, + "loss": 0.0206, + "step": 2853 + }, + { + "epoch": 0.7578332448220924, + "grad_norm": 0.6425774566966365, + "learning_rate": 7.325036293739335e-06, + "loss": 0.0354, + "step": 2854 + }, + { + "epoch": 0.7580987785448752, + "grad_norm": 0.657716711202906, + "learning_rate": 7.323092482295189e-06, + "loss": 0.0476, + "step": 2855 + }, + { + "epoch": 0.758364312267658, + "grad_norm": 0.4774457315284843, + "learning_rate": 7.3211482229862105e-06, + "loss": 0.1191, + "step": 2856 + }, + { + "epoch": 0.7586298459904408, + "grad_norm": 0.6382078436572101, + "learning_rate": 7.31920351618723e-06, + "loss": 0.0576, + "step": 2857 + }, + { + "epoch": 0.7588953797132236, + "grad_norm": 0.5283412342573214, + "learning_rate": 7.317258362273163e-06, + "loss": 0.1113, + "step": 2858 + }, + { + "epoch": 0.7591609134360063, + "grad_norm": 0.4752036040955733, + "learning_rate": 7.3153127616190135e-06, + "loss": 0.1113, + "step": 2859 + }, + { + "epoch": 0.7594264471587892, + "grad_norm": 0.3608925904506384, + "learning_rate": 7.31336671459987e-06, + "loss": 0.0432, + "step": 2860 + }, + { + "epoch": 0.759691980881572, + "grad_norm": 0.8818058458745188, + "learning_rate": 7.311420221590908e-06, + "loss": 0.1035, + "step": 2861 + }, + { + "epoch": 0.7599575146043548, + "grad_norm": 0.2218299455500649, + "learning_rate": 7.309473282967387e-06, + "loss": 0.0354, + "step": 2862 + }, + { + "epoch": 0.7602230483271375, + "grad_norm": 0.18696051200870123, + "learning_rate": 7.307525899104656e-06, + "loss": 0.0317, + "step": 2863 + }, + { + "epoch": 0.7604885820499203, + "grad_norm": 0.1377713040642783, + "learning_rate": 7.305578070378146e-06, + "loss": 0.0032, + "step": 2864 + }, + { + "epoch": 0.7607541157727031, + "grad_norm": 2.59326398834343, + "learning_rate": 7.303629797163376e-06, + "loss": 0.2031, + "step": 2865 + }, + { + "epoch": 0.761019649495486, + "grad_norm": 0.15845000559002115, + "learning_rate": 7.301681079835951e-06, + "loss": 0.0258, + "step": 2866 + }, + { + "epoch": 0.7612851832182688, + "grad_norm": 0.052770205661731046, + "learning_rate": 7.299731918771557e-06, + "loss": 0.0019, + "step": 2867 + }, + { + "epoch": 0.7615507169410515, + "grad_norm": 0.24152667743270337, + "learning_rate": 7.297782314345972e-06, + "loss": 0.0117, + "step": 2868 + }, + { + "epoch": 0.7618162506638343, + "grad_norm": 0.2268873408056939, + "learning_rate": 7.295832266935059e-06, + "loss": 0.0354, + "step": 2869 + }, + { + "epoch": 0.7620817843866171, + "grad_norm": 0.8115203089689542, + "learning_rate": 7.2938817769147606e-06, + "loss": 0.0752, + "step": 2870 + }, + { + "epoch": 0.7623473181093999, + "grad_norm": 0.17977093784633819, + "learning_rate": 7.291930844661109e-06, + "loss": 0.0317, + "step": 2871 + }, + { + "epoch": 0.7626128518321826, + "grad_norm": 0.5032054366779694, + "learning_rate": 7.289979470550222e-06, + "loss": 0.0815, + "step": 2872 + }, + { + "epoch": 0.7628783855549655, + "grad_norm": 0.44044711239312695, + "learning_rate": 7.288027654958301e-06, + "loss": 0.1191, + "step": 2873 + }, + { + "epoch": 0.7631439192777483, + "grad_norm": 0.49914234186607476, + "learning_rate": 7.286075398261633e-06, + "loss": 0.0432, + "step": 2874 + }, + { + "epoch": 0.7634094530005311, + "grad_norm": 0.1854274419328486, + "learning_rate": 7.284122700836589e-06, + "loss": 0.0317, + "step": 2875 + }, + { + "epoch": 0.7636749867233139, + "grad_norm": 2.9773814304041037, + "learning_rate": 7.282169563059627e-06, + "loss": 0.2461, + "step": 2876 + }, + { + "epoch": 0.7639405204460966, + "grad_norm": 0.5870033341617722, + "learning_rate": 7.280215985307289e-06, + "loss": 0.0525, + "step": 2877 + }, + { + "epoch": 0.7642060541688794, + "grad_norm": 0.19127127563907564, + "learning_rate": 7.278261967956203e-06, + "loss": 0.0286, + "step": 2878 + }, + { + "epoch": 0.7644715878916623, + "grad_norm": 0.8444518886403246, + "learning_rate": 7.276307511383078e-06, + "loss": 0.063, + "step": 2879 + }, + { + "epoch": 0.7647371216144451, + "grad_norm": 2.850318277814123, + "learning_rate": 7.274352615964713e-06, + "loss": 0.2559, + "step": 2880 + }, + { + "epoch": 0.7650026553372278, + "grad_norm": 0.3238687091034815, + "learning_rate": 7.272397282077986e-06, + "loss": 0.0391, + "step": 2881 + }, + { + "epoch": 0.7652681890600106, + "grad_norm": 0.14142541065938954, + "learning_rate": 7.2704415100998625e-06, + "loss": 0.0051, + "step": 2882 + }, + { + "epoch": 0.7655337227827934, + "grad_norm": 0.36146685369048176, + "learning_rate": 7.2684853004073945e-06, + "loss": 0.0432, + "step": 2883 + }, + { + "epoch": 0.7657992565055762, + "grad_norm": 0.20041311945341125, + "learning_rate": 7.2665286533777135e-06, + "loss": 0.0317, + "step": 2884 + }, + { + "epoch": 0.766064790228359, + "grad_norm": 0.5698213913839533, + "learning_rate": 7.264571569388039e-06, + "loss": 0.0525, + "step": 2885 + }, + { + "epoch": 0.7663303239511418, + "grad_norm": 0.17292280651134984, + "learning_rate": 7.262614048815676e-06, + "loss": 0.0286, + "step": 2886 + }, + { + "epoch": 0.7665958576739246, + "grad_norm": 0.4853008995909722, + "learning_rate": 7.26065609203801e-06, + "loss": 0.0258, + "step": 2887 + }, + { + "epoch": 0.7668613913967074, + "grad_norm": 0.017356961607042767, + "learning_rate": 7.258697699432511e-06, + "loss": 0.0006, + "step": 2888 + }, + { + "epoch": 0.7671269251194902, + "grad_norm": 0.17459281395542342, + "learning_rate": 7.256738871376733e-06, + "loss": 0.0286, + "step": 2889 + }, + { + "epoch": 0.7673924588422729, + "grad_norm": 0.5705614965167797, + "learning_rate": 7.2547796082483155e-06, + "loss": 0.1113, + "step": 2890 + }, + { + "epoch": 0.7676579925650557, + "grad_norm": 0.7709994652679434, + "learning_rate": 7.252819910424984e-06, + "loss": 0.0889, + "step": 2891 + }, + { + "epoch": 0.7679235262878386, + "grad_norm": 0.6567957495935665, + "learning_rate": 7.250859778284544e-06, + "loss": 0.0957, + "step": 2892 + }, + { + "epoch": 0.7681890600106214, + "grad_norm": 0.018785387429983837, + "learning_rate": 7.248899212204883e-06, + "loss": 0.0006, + "step": 2893 + }, + { + "epoch": 0.7684545937334042, + "grad_norm": 0.37828267860464776, + "learning_rate": 7.246938212563978e-06, + "loss": 0.0432, + "step": 2894 + }, + { + "epoch": 0.7687201274561869, + "grad_norm": 0.27512999387586706, + "learning_rate": 7.244976779739887e-06, + "loss": 0.0391, + "step": 2895 + }, + { + "epoch": 0.7689856611789697, + "grad_norm": 0.8365087686779479, + "learning_rate": 7.243014914110748e-06, + "loss": 0.0752, + "step": 2896 + }, + { + "epoch": 0.7692511949017525, + "grad_norm": 0.19062306016125935, + "learning_rate": 7.24105261605479e-06, + "loss": 0.0317, + "step": 2897 + }, + { + "epoch": 0.7695167286245354, + "grad_norm": 0.21536426733195435, + "learning_rate": 7.239089885950317e-06, + "loss": 0.0317, + "step": 2898 + }, + { + "epoch": 0.7697822623473181, + "grad_norm": 0.27494244089690895, + "learning_rate": 7.23712672417572e-06, + "loss": 0.0354, + "step": 2899 + }, + { + "epoch": 0.7700477960701009, + "grad_norm": 0.5905951220705504, + "learning_rate": 7.235163131109477e-06, + "loss": 0.0576, + "step": 2900 + }, + { + "epoch": 0.7703133297928837, + "grad_norm": 0.16398164182516856, + "learning_rate": 7.233199107130143e-06, + "loss": 0.0258, + "step": 2901 + }, + { + "epoch": 0.7705788635156665, + "grad_norm": 0.14868907573572718, + "learning_rate": 7.23123465261636e-06, + "loss": 0.0229, + "step": 2902 + }, + { + "epoch": 0.7708443972384493, + "grad_norm": 0.0668227086431703, + "learning_rate": 7.22926976794685e-06, + "loss": 0.0027, + "step": 2903 + }, + { + "epoch": 0.771109930961232, + "grad_norm": 0.12490439789722643, + "learning_rate": 7.227304453500421e-06, + "loss": 0.0039, + "step": 2904 + }, + { + "epoch": 0.7713754646840149, + "grad_norm": 0.1756744614125714, + "learning_rate": 7.2253387096559624e-06, + "loss": 0.0286, + "step": 2905 + }, + { + "epoch": 0.7716409984067977, + "grad_norm": 0.13448114469933353, + "learning_rate": 7.2233725367924455e-06, + "loss": 0.0206, + "step": 2906 + }, + { + "epoch": 0.7719065321295805, + "grad_norm": 0.13969460485900922, + "learning_rate": 7.2214059352889255e-06, + "loss": 0.0206, + "step": 2907 + }, + { + "epoch": 0.7721720658523632, + "grad_norm": 0.23636656047301416, + "learning_rate": 7.219438905524539e-06, + "loss": 0.0286, + "step": 2908 + }, + { + "epoch": 0.772437599575146, + "grad_norm": 0.24560068684401104, + "learning_rate": 7.21747144787851e-06, + "loss": 0.0317, + "step": 2909 + }, + { + "epoch": 0.7727031332979288, + "grad_norm": 0.6058270656607713, + "learning_rate": 7.215503562730137e-06, + "loss": 0.1367, + "step": 2910 + }, + { + "epoch": 0.7729686670207117, + "grad_norm": 0.18159088503174298, + "learning_rate": 7.2135352504588075e-06, + "loss": 0.0231, + "step": 2911 + }, + { + "epoch": 0.7732342007434945, + "grad_norm": 0.1548536389549544, + "learning_rate": 7.211566511443987e-06, + "loss": 0.0206, + "step": 2912 + }, + { + "epoch": 0.7734997344662772, + "grad_norm": 0.16636835555832918, + "learning_rate": 7.2095973460652265e-06, + "loss": 0.0255, + "step": 2913 + }, + { + "epoch": 0.77376526818906, + "grad_norm": 0.7660319815054392, + "learning_rate": 7.207627754702158e-06, + "loss": 0.0752, + "step": 2914 + }, + { + "epoch": 0.7740308019118428, + "grad_norm": 0.5882641889946315, + "learning_rate": 7.205657737734494e-06, + "loss": 0.1641, + "step": 2915 + }, + { + "epoch": 0.7742963356346256, + "grad_norm": 0.26612661931509535, + "learning_rate": 7.203687295542032e-06, + "loss": 0.0286, + "step": 2916 + }, + { + "epoch": 0.7745618693574083, + "grad_norm": 0.13671993817028616, + "learning_rate": 7.201716428504648e-06, + "loss": 0.0184, + "step": 2917 + }, + { + "epoch": 0.7748274030801912, + "grad_norm": 0.6397457970042721, + "learning_rate": 7.199745137002305e-06, + "loss": 0.1641, + "step": 2918 + }, + { + "epoch": 0.775092936802974, + "grad_norm": 0.15912374622827877, + "learning_rate": 7.197773421415043e-06, + "loss": 0.0231, + "step": 2919 + }, + { + "epoch": 0.7753584705257568, + "grad_norm": 0.6307542492456876, + "learning_rate": 7.1958012821229875e-06, + "loss": 0.1113, + "step": 2920 + }, + { + "epoch": 0.7756240042485396, + "grad_norm": 0.7335513883649463, + "learning_rate": 7.193828719506339e-06, + "loss": 0.0889, + "step": 2921 + }, + { + "epoch": 0.7758895379713223, + "grad_norm": 0.6007374902493993, + "learning_rate": 7.191855733945388e-06, + "loss": 0.1035, + "step": 2922 + }, + { + "epoch": 0.7761550716941051, + "grad_norm": 0.1480962315395715, + "learning_rate": 7.189882325820502e-06, + "loss": 0.0231, + "step": 2923 + }, + { + "epoch": 0.776420605416888, + "grad_norm": 0.18940608513779958, + "learning_rate": 7.187908495512132e-06, + "loss": 0.0255, + "step": 2924 + }, + { + "epoch": 0.7766861391396708, + "grad_norm": 0.036468166968755135, + "learning_rate": 7.185934243400806e-06, + "loss": 0.0013, + "step": 2925 + }, + { + "epoch": 0.7769516728624535, + "grad_norm": 0.1846619939812055, + "learning_rate": 7.183959569867139e-06, + "loss": 0.0286, + "step": 2926 + }, + { + "epoch": 0.7772172065852363, + "grad_norm": 0.18677638095875576, + "learning_rate": 7.1819844752918254e-06, + "loss": 0.0286, + "step": 2927 + }, + { + "epoch": 0.7774827403080191, + "grad_norm": 0.5057856617908502, + "learning_rate": 7.18000896005564e-06, + "loss": 0.1191, + "step": 2928 + }, + { + "epoch": 0.7777482740308019, + "grad_norm": 0.15379447760536835, + "learning_rate": 7.1780330245394356e-06, + "loss": 0.0231, + "step": 2929 + }, + { + "epoch": 0.7780138077535848, + "grad_norm": 0.535404559593995, + "learning_rate": 7.176056669124151e-06, + "loss": 0.1035, + "step": 2930 + }, + { + "epoch": 0.7782793414763675, + "grad_norm": 0.5885684385956611, + "learning_rate": 7.1740798941908075e-06, + "loss": 0.1113, + "step": 2931 + }, + { + "epoch": 0.7785448751991503, + "grad_norm": 1.2960479401687859, + "learning_rate": 7.1721027001205e-06, + "loss": 0.1367, + "step": 2932 + }, + { + "epoch": 0.7788104089219331, + "grad_norm": 0.16278696274439713, + "learning_rate": 7.17012508729441e-06, + "loss": 0.0255, + "step": 2933 + }, + { + "epoch": 0.7790759426447159, + "grad_norm": 0.6555468538162413, + "learning_rate": 7.1681470560937975e-06, + "loss": 0.063, + "step": 2934 + }, + { + "epoch": 0.7793414763674986, + "grad_norm": 0.6681252206363552, + "learning_rate": 7.166168606900004e-06, + "loss": 0.063, + "step": 2935 + }, + { + "epoch": 0.7796070100902814, + "grad_norm": 0.15339883381344643, + "learning_rate": 7.16418974009445e-06, + "loss": 0.0255, + "step": 2936 + }, + { + "epoch": 0.7798725438130643, + "grad_norm": 0.16143556475463686, + "learning_rate": 7.162210456058637e-06, + "loss": 0.0255, + "step": 2937 + }, + { + "epoch": 0.7801380775358471, + "grad_norm": 0.6409151246537623, + "learning_rate": 7.16023075517415e-06, + "loss": 0.063, + "step": 2938 + }, + { + "epoch": 0.7804036112586299, + "grad_norm": 0.36481025964273034, + "learning_rate": 7.15825063782265e-06, + "loss": 0.0476, + "step": 2939 + }, + { + "epoch": 0.7806691449814126, + "grad_norm": 0.5173488050665561, + "learning_rate": 7.156270104385883e-06, + "loss": 0.0525, + "step": 2940 + }, + { + "epoch": 0.7809346787041954, + "grad_norm": 0.545402761414387, + "learning_rate": 7.154289155245671e-06, + "loss": 0.0957, + "step": 2941 + }, + { + "epoch": 0.7812002124269782, + "grad_norm": 0.2742389550842906, + "learning_rate": 7.152307790783915e-06, + "loss": 0.0391, + "step": 2942 + }, + { + "epoch": 0.781465746149761, + "grad_norm": 0.46535355353496366, + "learning_rate": 7.1503260113826035e-06, + "loss": 0.0576, + "step": 2943 + }, + { + "epoch": 0.7817312798725438, + "grad_norm": 0.28067217268169, + "learning_rate": 7.148343817423796e-06, + "loss": 0.0131, + "step": 2944 + }, + { + "epoch": 0.7819968135953266, + "grad_norm": 0.02144069109536129, + "learning_rate": 7.146361209289639e-06, + "loss": 0.0008, + "step": 2945 + }, + { + "epoch": 0.7822623473181094, + "grad_norm": 0.28438066586009536, + "learning_rate": 7.144378187362354e-06, + "loss": 0.0146, + "step": 2946 + }, + { + "epoch": 0.7825278810408922, + "grad_norm": 0.263817437452264, + "learning_rate": 7.142394752024246e-06, + "loss": 0.0354, + "step": 2947 + }, + { + "epoch": 0.782793414763675, + "grad_norm": 1.3440697170730427, + "learning_rate": 7.140410903657695e-06, + "loss": 0.0957, + "step": 2948 + }, + { + "epoch": 0.7830589484864577, + "grad_norm": 0.9624109593359216, + "learning_rate": 7.138426642645167e-06, + "loss": 0.0889, + "step": 2949 + }, + { + "epoch": 0.7833244822092406, + "grad_norm": 0.47364686860005456, + "learning_rate": 7.136441969369203e-06, + "loss": 0.0889, + "step": 2950 + }, + { + "epoch": 0.7835900159320234, + "grad_norm": 0.39084757455633445, + "learning_rate": 7.134456884212424e-06, + "loss": 0.0476, + "step": 2951 + }, + { + "epoch": 0.7838555496548062, + "grad_norm": 0.16696576936454152, + "learning_rate": 7.132471387557533e-06, + "loss": 0.0258, + "step": 2952 + }, + { + "epoch": 0.7841210833775889, + "grad_norm": 0.552965242877744, + "learning_rate": 7.130485479787305e-06, + "loss": 0.0525, + "step": 2953 + }, + { + "epoch": 0.7843866171003717, + "grad_norm": 0.1912918857329157, + "learning_rate": 7.128499161284606e-06, + "loss": 0.0286, + "step": 2954 + }, + { + "epoch": 0.7846521508231545, + "grad_norm": 0.1466550664536916, + "learning_rate": 7.126512432432371e-06, + "loss": 0.0064, + "step": 2955 + }, + { + "epoch": 0.7849176845459374, + "grad_norm": 0.07248451903079635, + "learning_rate": 7.12452529361362e-06, + "loss": 0.0031, + "step": 2956 + }, + { + "epoch": 0.7851832182687202, + "grad_norm": 0.19509706165246218, + "learning_rate": 7.1225377452114465e-06, + "loss": 0.0317, + "step": 2957 + }, + { + "epoch": 0.7854487519915029, + "grad_norm": 0.5488476607363877, + "learning_rate": 7.12054978760903e-06, + "loss": 0.0957, + "step": 2958 + }, + { + "epoch": 0.7857142857142857, + "grad_norm": 0.28899650138716765, + "learning_rate": 7.118561421189625e-06, + "loss": 0.0105, + "step": 2959 + }, + { + "epoch": 0.7859798194370685, + "grad_norm": 0.48038262289836364, + "learning_rate": 7.116572646336563e-06, + "loss": 0.1035, + "step": 2960 + }, + { + "epoch": 0.7862453531598513, + "grad_norm": 0.3526916834420791, + "learning_rate": 7.1145834634332585e-06, + "loss": 0.0476, + "step": 2961 + }, + { + "epoch": 0.786510886882634, + "grad_norm": 0.03407279069444661, + "learning_rate": 7.1125938728632e-06, + "loss": 0.001, + "step": 2962 + }, + { + "epoch": 0.7867764206054169, + "grad_norm": 0.47644676683623466, + "learning_rate": 7.11060387500996e-06, + "loss": 0.0576, + "step": 2963 + }, + { + "epoch": 0.7870419543281997, + "grad_norm": 0.655599839004455, + "learning_rate": 7.108613470257184e-06, + "loss": 0.0688, + "step": 2964 + }, + { + "epoch": 0.7873074880509825, + "grad_norm": 0.7310969059023468, + "learning_rate": 7.1066226589886e-06, + "loss": 0.063, + "step": 2965 + }, + { + "epoch": 0.7875730217737653, + "grad_norm": 0.8490616447582846, + "learning_rate": 7.1046314415880125e-06, + "loss": 0.0688, + "step": 2966 + }, + { + "epoch": 0.787838555496548, + "grad_norm": 0.4952816328444134, + "learning_rate": 7.102639818439307e-06, + "loss": 0.1279, + "step": 2967 + }, + { + "epoch": 0.7881040892193308, + "grad_norm": 0.12125047854873192, + "learning_rate": 7.1006477899264405e-06, + "loss": 0.0036, + "step": 2968 + }, + { + "epoch": 0.7883696229421137, + "grad_norm": 0.47341887169317354, + "learning_rate": 7.0986553564334544e-06, + "loss": 0.1035, + "step": 2969 + }, + { + "epoch": 0.7886351566648965, + "grad_norm": 0.5438486797227732, + "learning_rate": 7.09666251834447e-06, + "loss": 0.0576, + "step": 2970 + }, + { + "epoch": 0.7889006903876792, + "grad_norm": 0.4600082312962437, + "learning_rate": 7.0946692760436755e-06, + "loss": 0.1191, + "step": 2971 + }, + { + "epoch": 0.789166224110462, + "grad_norm": 0.13859694012264917, + "learning_rate": 7.092675629915352e-06, + "loss": 0.0229, + "step": 2972 + }, + { + "epoch": 0.7894317578332448, + "grad_norm": 0.6085133841382123, + "learning_rate": 7.0906815803438465e-06, + "loss": 0.0752, + "step": 2973 + }, + { + "epoch": 0.7896972915560276, + "grad_norm": 0.29337370481380903, + "learning_rate": 7.088687127713591e-06, + "loss": 0.0432, + "step": 2974 + }, + { + "epoch": 0.7899628252788105, + "grad_norm": 0.19889810982002218, + "learning_rate": 7.08669227240909e-06, + "loss": 0.0354, + "step": 2975 + }, + { + "epoch": 0.7902283590015932, + "grad_norm": 0.46899951698914233, + "learning_rate": 7.08469701481493e-06, + "loss": 0.0476, + "step": 2976 + }, + { + "epoch": 0.790493892724376, + "grad_norm": 0.7680719620822224, + "learning_rate": 7.082701355315772e-06, + "loss": 0.0391, + "step": 2977 + }, + { + "epoch": 0.7907594264471588, + "grad_norm": 0.015761250295268286, + "learning_rate": 7.080705294296355e-06, + "loss": 0.0006, + "step": 2978 + }, + { + "epoch": 0.7910249601699416, + "grad_norm": 0.8202013166896612, + "learning_rate": 7.078708832141497e-06, + "loss": 0.0579, + "step": 2979 + }, + { + "epoch": 0.7912904938927243, + "grad_norm": 0.5513027173573466, + "learning_rate": 7.076711969236093e-06, + "loss": 0.0525, + "step": 2980 + }, + { + "epoch": 0.7915560276155071, + "grad_norm": 0.22482149945230054, + "learning_rate": 7.074714705965113e-06, + "loss": 0.0391, + "step": 2981 + }, + { + "epoch": 0.79182156133829, + "grad_norm": 0.18904600421014478, + "learning_rate": 7.072717042713608e-06, + "loss": 0.0317, + "step": 2982 + }, + { + "epoch": 0.7920870950610728, + "grad_norm": 1.3214970265084973, + "learning_rate": 7.070718979866702e-06, + "loss": 0.0752, + "step": 2983 + }, + { + "epoch": 0.7923526287838556, + "grad_norm": 0.19537700713929687, + "learning_rate": 7.068720517809599e-06, + "loss": 0.0317, + "step": 2984 + }, + { + "epoch": 0.7926181625066383, + "grad_norm": 0.012368511241780974, + "learning_rate": 7.0667216569275776e-06, + "loss": 0.0004, + "step": 2985 + }, + { + "epoch": 0.7928836962294211, + "grad_norm": 0.4233583256163125, + "learning_rate": 7.064722397605996e-06, + "loss": 0.0576, + "step": 2986 + }, + { + "epoch": 0.7931492299522039, + "grad_norm": 0.5908805128820646, + "learning_rate": 7.062722740230286e-06, + "loss": 0.063, + "step": 2987 + }, + { + "epoch": 0.7934147636749868, + "grad_norm": 0.9437226712928716, + "learning_rate": 7.060722685185961e-06, + "loss": 0.0815, + "step": 2988 + }, + { + "epoch": 0.7936802973977695, + "grad_norm": 0.4253965505855059, + "learning_rate": 7.058722232858604e-06, + "loss": 0.0258, + "step": 2989 + }, + { + "epoch": 0.7939458311205523, + "grad_norm": 0.12231640242354447, + "learning_rate": 7.056721383633883e-06, + "loss": 0.0045, + "step": 2990 + }, + { + "epoch": 0.7942113648433351, + "grad_norm": 0.423548902725689, + "learning_rate": 7.054720137897536e-06, + "loss": 0.0752, + "step": 2991 + }, + { + "epoch": 0.7944768985661179, + "grad_norm": 0.5284696377321977, + "learning_rate": 7.05271849603538e-06, + "loss": 0.0576, + "step": 2992 + }, + { + "epoch": 0.7947424322889007, + "grad_norm": 0.50003528163191, + "learning_rate": 7.050716458433305e-06, + "loss": 0.0576, + "step": 2993 + }, + { + "epoch": 0.7950079660116834, + "grad_norm": 1.5336070461053033, + "learning_rate": 7.048714025477286e-06, + "loss": 0.1191, + "step": 2994 + }, + { + "epoch": 0.7952734997344663, + "grad_norm": 0.1553670693396578, + "learning_rate": 7.046711197553363e-06, + "loss": 0.0229, + "step": 2995 + }, + { + "epoch": 0.7955390334572491, + "grad_norm": 0.48854875403118264, + "learning_rate": 7.04470797504766e-06, + "loss": 0.0752, + "step": 2996 + }, + { + "epoch": 0.7958045671800319, + "grad_norm": 2.1443128723646634, + "learning_rate": 7.042704358346375e-06, + "loss": 0.3203, + "step": 2997 + }, + { + "epoch": 0.7960701009028146, + "grad_norm": 0.18641325513771054, + "learning_rate": 7.040700347835781e-06, + "loss": 0.0317, + "step": 2998 + }, + { + "epoch": 0.7963356346255974, + "grad_norm": 0.14741215550952244, + "learning_rate": 7.038695943902228e-06, + "loss": 0.0255, + "step": 2999 + }, + { + "epoch": 0.7966011683483802, + "grad_norm": 0.8053739465959789, + "learning_rate": 7.036691146932142e-06, + "loss": 0.0576, + "step": 3000 + }, + { + "epoch": 0.7968667020711631, + "grad_norm": 0.5302984553969542, + "learning_rate": 7.034685957312022e-06, + "loss": 0.063, + "step": 3001 + }, + { + "epoch": 0.7971322357939459, + "grad_norm": 0.32862612818420966, + "learning_rate": 7.0326803754284454e-06, + "loss": 0.0432, + "step": 3002 + }, + { + "epoch": 0.7973977695167286, + "grad_norm": 0.020141815400440993, + "learning_rate": 7.030674401668068e-06, + "loss": 0.0007, + "step": 3003 + }, + { + "epoch": 0.7976633032395114, + "grad_norm": 0.15885360595794934, + "learning_rate": 7.028668036417614e-06, + "loss": 0.0255, + "step": 3004 + }, + { + "epoch": 0.7979288369622942, + "grad_norm": 0.19832612260923135, + "learning_rate": 7.026661280063889e-06, + "loss": 0.0317, + "step": 3005 + }, + { + "epoch": 0.798194370685077, + "grad_norm": 0.30141774337674915, + "learning_rate": 7.024654132993772e-06, + "loss": 0.0117, + "step": 3006 + }, + { + "epoch": 0.7984599044078597, + "grad_norm": 0.07715165958532787, + "learning_rate": 7.022646595594216e-06, + "loss": 0.0031, + "step": 3007 + }, + { + "epoch": 0.7987254381306426, + "grad_norm": 0.14328280172520452, + "learning_rate": 7.020638668252252e-06, + "loss": 0.0231, + "step": 3008 + }, + { + "epoch": 0.7989909718534254, + "grad_norm": 0.3178836457981987, + "learning_rate": 7.0186303513549836e-06, + "loss": 0.0391, + "step": 3009 + }, + { + "epoch": 0.7992565055762082, + "grad_norm": 0.025143683685695306, + "learning_rate": 7.016621645289591e-06, + "loss": 0.0008, + "step": 3010 + }, + { + "epoch": 0.799522039298991, + "grad_norm": 0.5122433415525655, + "learning_rate": 7.014612550443328e-06, + "loss": 0.1113, + "step": 3011 + }, + { + "epoch": 0.7997875730217737, + "grad_norm": 0.021135926814953465, + "learning_rate": 7.012603067203526e-06, + "loss": 0.0007, + "step": 3012 + }, + { + "epoch": 0.8000531067445565, + "grad_norm": 0.03479723395843193, + "learning_rate": 7.0105931959575905e-06, + "loss": 0.0013, + "step": 3013 + }, + { + "epoch": 0.8003186404673394, + "grad_norm": 0.5479003425960896, + "learning_rate": 7.008582937092998e-06, + "loss": 0.0752, + "step": 3014 + }, + { + "epoch": 0.8005841741901222, + "grad_norm": 0.020117648235620723, + "learning_rate": 7.006572290997304e-06, + "loss": 0.0007, + "step": 3015 + }, + { + "epoch": 0.8008497079129049, + "grad_norm": 0.12972170235146246, + "learning_rate": 7.00456125805814e-06, + "loss": 0.0184, + "step": 3016 + }, + { + "epoch": 0.8011152416356877, + "grad_norm": 0.184056277181454, + "learning_rate": 7.002549838663205e-06, + "loss": 0.0286, + "step": 3017 + }, + { + "epoch": 0.8013807753584705, + "grad_norm": 1.2567348650722587, + "learning_rate": 7.00053803320028e-06, + "loss": 0.1035, + "step": 3018 + }, + { + "epoch": 0.8016463090812533, + "grad_norm": 0.05387168119057167, + "learning_rate": 6.998525842057216e-06, + "loss": 0.0022, + "step": 3019 + }, + { + "epoch": 0.8019118428040362, + "grad_norm": 0.8621519554142695, + "learning_rate": 6.996513265621938e-06, + "loss": 0.0576, + "step": 3020 + }, + { + "epoch": 0.8021773765268189, + "grad_norm": 0.5458655815754319, + "learning_rate": 6.994500304282452e-06, + "loss": 0.0476, + "step": 3021 + }, + { + "epoch": 0.8024429102496017, + "grad_norm": 0.30113797385769164, + "learning_rate": 6.992486958426828e-06, + "loss": 0.0092, + "step": 3022 + }, + { + "epoch": 0.8027084439723845, + "grad_norm": 0.5985173819514868, + "learning_rate": 6.990473228443219e-06, + "loss": 0.0815, + "step": 3023 + }, + { + "epoch": 0.8029739776951673, + "grad_norm": 0.5438449287637052, + "learning_rate": 6.988459114719849e-06, + "loss": 0.1191, + "step": 3024 + }, + { + "epoch": 0.80323951141795, + "grad_norm": 0.203524044041484, + "learning_rate": 6.98644461764501e-06, + "loss": 0.0057, + "step": 3025 + }, + { + "epoch": 0.8035050451407328, + "grad_norm": 0.42850216351257353, + "learning_rate": 6.984429737607078e-06, + "loss": 0.0432, + "step": 3026 + }, + { + "epoch": 0.8037705788635157, + "grad_norm": 0.5767173342132778, + "learning_rate": 6.982414474994496e-06, + "loss": 0.1035, + "step": 3027 + }, + { + "epoch": 0.8040361125862985, + "grad_norm": 0.17502649298711, + "learning_rate": 6.980398830195785e-06, + "loss": 0.0255, + "step": 3028 + }, + { + "epoch": 0.8043016463090813, + "grad_norm": 0.005766243821583203, + "learning_rate": 6.9783828035995345e-06, + "loss": 0.0002, + "step": 3029 + }, + { + "epoch": 0.804567180031864, + "grad_norm": 0.23011324276074346, + "learning_rate": 6.976366395594414e-06, + "loss": 0.0354, + "step": 3030 + }, + { + "epoch": 0.8048327137546468, + "grad_norm": 0.5976616978457928, + "learning_rate": 6.9743496065691604e-06, + "loss": 0.1191, + "step": 3031 + }, + { + "epoch": 0.8050982474774296, + "grad_norm": 0.8692273227258032, + "learning_rate": 6.97233243691259e-06, + "loss": 0.0688, + "step": 3032 + }, + { + "epoch": 0.8053637812002125, + "grad_norm": 0.23690287244781913, + "learning_rate": 6.970314887013585e-06, + "loss": 0.0354, + "step": 3033 + }, + { + "epoch": 0.8056293149229952, + "grad_norm": 0.1256527182764814, + "learning_rate": 6.968296957261107e-06, + "loss": 0.0057, + "step": 3034 + }, + { + "epoch": 0.805894848645778, + "grad_norm": 0.6138955609847898, + "learning_rate": 6.9662786480441916e-06, + "loss": 0.0752, + "step": 3035 + }, + { + "epoch": 0.8061603823685608, + "grad_norm": 0.2415413655534569, + "learning_rate": 6.964259959751942e-06, + "loss": 0.0354, + "step": 3036 + }, + { + "epoch": 0.8064259160913436, + "grad_norm": 0.19048387939958525, + "learning_rate": 6.962240892773538e-06, + "loss": 0.0317, + "step": 3037 + }, + { + "epoch": 0.8066914498141264, + "grad_norm": 0.15457135822421353, + "learning_rate": 6.960221447498232e-06, + "loss": 0.0073, + "step": 3038 + }, + { + "epoch": 0.8069569835369091, + "grad_norm": 0.43913415602710754, + "learning_rate": 6.958201624315352e-06, + "loss": 0.0258, + "step": 3039 + }, + { + "epoch": 0.807222517259692, + "grad_norm": 0.24855562706826156, + "learning_rate": 6.9561814236142925e-06, + "loss": 0.0165, + "step": 3040 + }, + { + "epoch": 0.8074880509824748, + "grad_norm": 0.22073973219055812, + "learning_rate": 6.954160845784524e-06, + "loss": 0.0354, + "step": 3041 + }, + { + "epoch": 0.8077535847052576, + "grad_norm": 0.6539979716287775, + "learning_rate": 6.9521398912155935e-06, + "loss": 0.063, + "step": 3042 + }, + { + "epoch": 0.8080191184280403, + "grad_norm": 0.19841440578801994, + "learning_rate": 6.950118560297113e-06, + "loss": 0.0286, + "step": 3043 + }, + { + "epoch": 0.8082846521508231, + "grad_norm": 0.16684146681716522, + "learning_rate": 6.9480968534187756e-06, + "loss": 0.0073, + "step": 3044 + }, + { + "epoch": 0.8085501858736059, + "grad_norm": 0.035626425391478714, + "learning_rate": 6.94607477097034e-06, + "loss": 0.001, + "step": 3045 + }, + { + "epoch": 0.8088157195963888, + "grad_norm": 0.6789720971885709, + "learning_rate": 6.944052313341641e-06, + "loss": 0.0576, + "step": 3046 + }, + { + "epoch": 0.8090812533191716, + "grad_norm": 0.18625383126254888, + "learning_rate": 6.942029480922583e-06, + "loss": 0.0255, + "step": 3047 + }, + { + "epoch": 0.8093467870419543, + "grad_norm": 0.1758264482993835, + "learning_rate": 6.940006274103146e-06, + "loss": 0.0255, + "step": 3048 + }, + { + "epoch": 0.8096123207647371, + "grad_norm": 0.5327886792026215, + "learning_rate": 6.93798269327338e-06, + "loss": 0.1191, + "step": 3049 + }, + { + "epoch": 0.8098778544875199, + "grad_norm": 0.6229809437817544, + "learning_rate": 6.935958738823405e-06, + "loss": 0.0576, + "step": 3050 + }, + { + "epoch": 0.8101433882103027, + "grad_norm": 0.45940651839061775, + "learning_rate": 6.933934411143419e-06, + "loss": 0.0229, + "step": 3051 + }, + { + "epoch": 0.8104089219330854, + "grad_norm": 0.7238824667634822, + "learning_rate": 6.931909710623688e-06, + "loss": 0.0688, + "step": 3052 + }, + { + "epoch": 0.8106744556558683, + "grad_norm": 1.0445234458248163, + "learning_rate": 6.9298846376545485e-06, + "loss": 0.1035, + "step": 3053 + }, + { + "epoch": 0.8109399893786511, + "grad_norm": 1.6044922272189068, + "learning_rate": 6.927859192626412e-06, + "loss": 0.1367, + "step": 3054 + }, + { + "epoch": 0.8112055231014339, + "grad_norm": 0.21096744217579994, + "learning_rate": 6.92583337592976e-06, + "loss": 0.0286, + "step": 3055 + }, + { + "epoch": 0.8114710568242167, + "grad_norm": 2.340935463253369, + "learning_rate": 6.923807187955146e-06, + "loss": 0.1367, + "step": 3056 + }, + { + "epoch": 0.8117365905469994, + "grad_norm": 0.20026014253417043, + "learning_rate": 6.921780629093195e-06, + "loss": 0.0286, + "step": 3057 + }, + { + "epoch": 0.8120021242697822, + "grad_norm": 0.32873456227749953, + "learning_rate": 6.919753699734606e-06, + "loss": 0.0146, + "step": 3058 + }, + { + "epoch": 0.8122676579925651, + "grad_norm": 0.018825755924311082, + "learning_rate": 6.917726400270144e-06, + "loss": 0.0006, + "step": 3059 + }, + { + "epoch": 0.8125331917153479, + "grad_norm": 0.005601648667821216, + "learning_rate": 6.915698731090649e-06, + "loss": 0.0002, + "step": 3060 + }, + { + "epoch": 0.8127987254381307, + "grad_norm": 0.2995555787538075, + "learning_rate": 6.9136706925870315e-06, + "loss": 0.0432, + "step": 3061 + }, + { + "epoch": 0.8130642591609134, + "grad_norm": 1.414538287285168, + "learning_rate": 6.9116422851502755e-06, + "loss": 0.1035, + "step": 3062 + }, + { + "epoch": 0.8133297928836962, + "grad_norm": 0.012964608300811357, + "learning_rate": 6.909613509171432e-06, + "loss": 0.0004, + "step": 3063 + }, + { + "epoch": 0.813595326606479, + "grad_norm": 0.6660552817255484, + "learning_rate": 6.907584365041627e-06, + "loss": 0.0525, + "step": 3064 + }, + { + "epoch": 0.8138608603292619, + "grad_norm": 0.05358826167447712, + "learning_rate": 6.905554853152053e-06, + "loss": 0.0022, + "step": 3065 + }, + { + "epoch": 0.8141263940520446, + "grad_norm": 0.040385174127559445, + "learning_rate": 6.903524973893977e-06, + "loss": 0.0017, + "step": 3066 + }, + { + "epoch": 0.8143919277748274, + "grad_norm": 0.025199304555918196, + "learning_rate": 6.9014947276587375e-06, + "loss": 0.0006, + "step": 3067 + }, + { + "epoch": 0.8146574614976102, + "grad_norm": 0.011958409264480043, + "learning_rate": 6.899464114837739e-06, + "loss": 0.0004, + "step": 3068 + }, + { + "epoch": 0.814922995220393, + "grad_norm": 0.3601918647058819, + "learning_rate": 6.897433135822461e-06, + "loss": 0.0046, + "step": 3069 + }, + { + "epoch": 0.8151885289431758, + "grad_norm": 0.036477697072503416, + "learning_rate": 6.895401791004453e-06, + "loss": 0.0013, + "step": 3070 + }, + { + "epoch": 0.8154540626659585, + "grad_norm": 0.558701251227667, + "learning_rate": 6.893370080775335e-06, + "loss": 0.0255, + "step": 3071 + }, + { + "epoch": 0.8157195963887414, + "grad_norm": 0.2768977076423462, + "learning_rate": 6.891338005526796e-06, + "loss": 0.0317, + "step": 3072 + }, + { + "epoch": 0.8159851301115242, + "grad_norm": 0.2487420116846952, + "learning_rate": 6.889305565650595e-06, + "loss": 0.0117, + "step": 3073 + }, + { + "epoch": 0.816250663834307, + "grad_norm": 0.8269106312508646, + "learning_rate": 6.887272761538561e-06, + "loss": 0.0752, + "step": 3074 + }, + { + "epoch": 0.8165161975570897, + "grad_norm": 0.21473946197959184, + "learning_rate": 6.8852395935826e-06, + "loss": 0.0286, + "step": 3075 + }, + { + "epoch": 0.8167817312798725, + "grad_norm": 0.6814605105293484, + "learning_rate": 6.883206062174679e-06, + "loss": 0.0957, + "step": 3076 + }, + { + "epoch": 0.8170472650026553, + "grad_norm": 0.56964058697379, + "learning_rate": 6.881172167706839e-06, + "loss": 0.0525, + "step": 3077 + }, + { + "epoch": 0.8173127987254382, + "grad_norm": 0.09021202794060154, + "learning_rate": 6.879137910571191e-06, + "loss": 0.0009, + "step": 3078 + }, + { + "epoch": 0.817578332448221, + "grad_norm": 0.25097175884612427, + "learning_rate": 6.877103291159918e-06, + "loss": 0.0317, + "step": 3079 + }, + { + "epoch": 0.8178438661710037, + "grad_norm": 0.6222120235839559, + "learning_rate": 6.875068309865267e-06, + "loss": 0.0354, + "step": 3080 + }, + { + "epoch": 0.8181093998937865, + "grad_norm": 0.43902971421869624, + "learning_rate": 6.873032967079562e-06, + "loss": 0.0184, + "step": 3081 + }, + { + "epoch": 0.8183749336165693, + "grad_norm": 0.3111121193978235, + "learning_rate": 6.870997263195189e-06, + "loss": 0.0354, + "step": 3082 + }, + { + "epoch": 0.8186404673393521, + "grad_norm": 0.5384200020077577, + "learning_rate": 6.868961198604611e-06, + "loss": 0.1113, + "step": 3083 + }, + { + "epoch": 0.8189060010621348, + "grad_norm": 1.675421946165875, + "learning_rate": 6.866924773700357e-06, + "loss": 0.0889, + "step": 3084 + }, + { + "epoch": 0.8191715347849177, + "grad_norm": 0.01640929145166503, + "learning_rate": 6.864887988875024e-06, + "loss": 0.0004, + "step": 3085 + }, + { + "epoch": 0.8194370685077005, + "grad_norm": 0.23519681911205248, + "learning_rate": 6.862850844521282e-06, + "loss": 0.0286, + "step": 3086 + }, + { + "epoch": 0.8197026022304833, + "grad_norm": 0.5545733000060519, + "learning_rate": 6.860813341031867e-06, + "loss": 0.0957, + "step": 3087 + }, + { + "epoch": 0.8199681359532661, + "grad_norm": 1.0730803740470198, + "learning_rate": 6.858775478799586e-06, + "loss": 0.0752, + "step": 3088 + }, + { + "epoch": 0.8202336696760488, + "grad_norm": 0.7774570624349827, + "learning_rate": 6.856737258217315e-06, + "loss": 0.0525, + "step": 3089 + }, + { + "epoch": 0.8204992033988316, + "grad_norm": 0.6761604485490404, + "learning_rate": 6.854698679677999e-06, + "loss": 0.1279, + "step": 3090 + }, + { + "epoch": 0.8207647371216145, + "grad_norm": 0.7421475581724489, + "learning_rate": 6.8526597435746515e-06, + "loss": 0.1035, + "step": 3091 + }, + { + "epoch": 0.8210302708443973, + "grad_norm": 0.22063023851328456, + "learning_rate": 6.850620450300355e-06, + "loss": 0.0082, + "step": 3092 + }, + { + "epoch": 0.82129580456718, + "grad_norm": 0.8574523073523872, + "learning_rate": 6.848580800248263e-06, + "loss": 0.0688, + "step": 3093 + }, + { + "epoch": 0.8215613382899628, + "grad_norm": 0.003935795844182658, + "learning_rate": 6.846540793811594e-06, + "loss": 0.0001, + "step": 3094 + }, + { + "epoch": 0.8218268720127456, + "grad_norm": 0.2378222681927969, + "learning_rate": 6.844500431383637e-06, + "loss": 0.0317, + "step": 3095 + }, + { + "epoch": 0.8220924057355284, + "grad_norm": 0.5304906382466659, + "learning_rate": 6.842459713357752e-06, + "loss": 0.1113, + "step": 3096 + }, + { + "epoch": 0.8223579394583113, + "grad_norm": 1.1377493667962182, + "learning_rate": 6.840418640127363e-06, + "loss": 0.1035, + "step": 3097 + }, + { + "epoch": 0.822623473181094, + "grad_norm": 0.3253793868043899, + "learning_rate": 6.8383772120859674e-06, + "loss": 0.0391, + "step": 3098 + }, + { + "epoch": 0.8228890069038768, + "grad_norm": 0.7490381827884851, + "learning_rate": 6.836335429627126e-06, + "loss": 0.0688, + "step": 3099 + }, + { + "epoch": 0.8231545406266596, + "grad_norm": 0.38044025654402897, + "learning_rate": 6.834293293144471e-06, + "loss": 0.0525, + "step": 3100 + }, + { + "epoch": 0.8234200743494424, + "grad_norm": 0.4398262896483564, + "learning_rate": 6.832250803031701e-06, + "loss": 0.0525, + "step": 3101 + }, + { + "epoch": 0.8236856080722251, + "grad_norm": 0.46191153181424843, + "learning_rate": 6.830207959682589e-06, + "loss": 0.0525, + "step": 3102 + }, + { + "epoch": 0.823951141795008, + "grad_norm": 0.2849042654701407, + "learning_rate": 6.828164763490965e-06, + "loss": 0.0146, + "step": 3103 + }, + { + "epoch": 0.8242166755177908, + "grad_norm": 0.18502482410411553, + "learning_rate": 6.8261212148507385e-06, + "loss": 0.0045, + "step": 3104 + }, + { + "epoch": 0.8244822092405736, + "grad_norm": 0.6239249695598372, + "learning_rate": 6.824077314155877e-06, + "loss": 0.0576, + "step": 3105 + }, + { + "epoch": 0.8247477429633564, + "grad_norm": 0.6534779730016713, + "learning_rate": 6.822033061800423e-06, + "loss": 0.0815, + "step": 3106 + }, + { + "epoch": 0.8250132766861391, + "grad_norm": 0.2537237082234382, + "learning_rate": 6.819988458178483e-06, + "loss": 0.0354, + "step": 3107 + }, + { + "epoch": 0.8252788104089219, + "grad_norm": 0.03142861008804912, + "learning_rate": 6.8179435036842336e-06, + "loss": 0.0012, + "step": 3108 + }, + { + "epoch": 0.8255443441317047, + "grad_norm": 0.35843690695514047, + "learning_rate": 6.815898198711917e-06, + "loss": 0.0391, + "step": 3109 + }, + { + "epoch": 0.8258098778544876, + "grad_norm": 1.853896458771604, + "learning_rate": 6.813852543655844e-06, + "loss": 0.0889, + "step": 3110 + }, + { + "epoch": 0.8260754115772703, + "grad_norm": 0.31147081324844705, + "learning_rate": 6.811806538910394e-06, + "loss": 0.0391, + "step": 3111 + }, + { + "epoch": 0.8263409453000531, + "grad_norm": 0.039067829564739184, + "learning_rate": 6.80976018487001e-06, + "loss": 0.0012, + "step": 3112 + }, + { + "epoch": 0.8266064790228359, + "grad_norm": 0.5695275896143159, + "learning_rate": 6.807713481929208e-06, + "loss": 0.0476, + "step": 3113 + }, + { + "epoch": 0.8268720127456187, + "grad_norm": 0.9387839300076153, + "learning_rate": 6.805666430482565e-06, + "loss": 0.0688, + "step": 3114 + }, + { + "epoch": 0.8271375464684015, + "grad_norm": 0.627642519474533, + "learning_rate": 6.803619030924731e-06, + "loss": 0.0576, + "step": 3115 + }, + { + "epoch": 0.8274030801911842, + "grad_norm": 0.6838961206788357, + "learning_rate": 6.80157128365042e-06, + "loss": 0.0889, + "step": 3116 + }, + { + "epoch": 0.8276686139139671, + "grad_norm": 0.7468579775951218, + "learning_rate": 6.7995231890544126e-06, + "loss": 0.0286, + "step": 3117 + }, + { + "epoch": 0.8279341476367499, + "grad_norm": 0.5099408751834107, + "learning_rate": 6.797474747531559e-06, + "loss": 0.0432, + "step": 3118 + }, + { + "epoch": 0.8281996813595327, + "grad_norm": 0.47725851141825826, + "learning_rate": 6.795425959476771e-06, + "loss": 0.0432, + "step": 3119 + }, + { + "epoch": 0.8284652150823154, + "grad_norm": 0.4383617787561755, + "learning_rate": 6.793376825285034e-06, + "loss": 0.0391, + "step": 3120 + }, + { + "epoch": 0.8287307488050982, + "grad_norm": 0.44209859503369064, + "learning_rate": 6.791327345351398e-06, + "loss": 0.0184, + "step": 3121 + }, + { + "epoch": 0.828996282527881, + "grad_norm": 0.8190324245638593, + "learning_rate": 6.789277520070974e-06, + "loss": 0.063, + "step": 3122 + }, + { + "epoch": 0.8292618162506639, + "grad_norm": 0.3441015055032359, + "learning_rate": 6.787227349838946e-06, + "loss": 0.0391, + "step": 3123 + }, + { + "epoch": 0.8295273499734467, + "grad_norm": 1.0732539325353083, + "learning_rate": 6.785176835050565e-06, + "loss": 0.0815, + "step": 3124 + }, + { + "epoch": 0.8297928836962294, + "grad_norm": 0.7650130423289352, + "learning_rate": 6.783125976101144e-06, + "loss": 0.0815, + "step": 3125 + }, + { + "epoch": 0.8300584174190122, + "grad_norm": 3.0744893694845725, + "learning_rate": 6.781074773386065e-06, + "loss": 0.4434, + "step": 3126 + }, + { + "epoch": 0.830323951141795, + "grad_norm": 0.022161918350785268, + "learning_rate": 6.779023227300776e-06, + "loss": 0.0004, + "step": 3127 + }, + { + "epoch": 0.8305894848645778, + "grad_norm": 0.26171046191733366, + "learning_rate": 6.77697133824079e-06, + "loss": 0.0286, + "step": 3128 + }, + { + "epoch": 0.8308550185873605, + "grad_norm": 0.025948080554812705, + "learning_rate": 6.774919106601687e-06, + "loss": 0.0008, + "step": 3129 + }, + { + "epoch": 0.8311205523101434, + "grad_norm": 0.2255052262042786, + "learning_rate": 6.772866532779115e-06, + "loss": 0.0317, + "step": 3130 + }, + { + "epoch": 0.8313860860329262, + "grad_norm": 1.0205381888779612, + "learning_rate": 6.770813617168783e-06, + "loss": 0.0752, + "step": 3131 + }, + { + "epoch": 0.831651619755709, + "grad_norm": 0.8912415721330618, + "learning_rate": 6.768760360166471e-06, + "loss": 0.0432, + "step": 3132 + }, + { + "epoch": 0.8319171534784918, + "grad_norm": 0.7472942164325972, + "learning_rate": 6.766706762168021e-06, + "loss": 0.0576, + "step": 3133 + }, + { + "epoch": 0.8321826872012745, + "grad_norm": 0.35299953757238806, + "learning_rate": 6.7646528235693445e-06, + "loss": 0.0391, + "step": 3134 + }, + { + "epoch": 0.8324482209240573, + "grad_norm": 0.609367505717721, + "learning_rate": 6.762598544766417e-06, + "loss": 0.1035, + "step": 3135 + }, + { + "epoch": 0.8327137546468402, + "grad_norm": 0.4528094777298801, + "learning_rate": 6.7605439261552774e-06, + "loss": 0.0476, + "step": 3136 + }, + { + "epoch": 0.832979288369623, + "grad_norm": 0.878569486953114, + "learning_rate": 6.758488968132031e-06, + "loss": 0.0286, + "step": 3137 + }, + { + "epoch": 0.8332448220924057, + "grad_norm": 0.28546248013723374, + "learning_rate": 6.756433671092853e-06, + "loss": 0.0317, + "step": 3138 + }, + { + "epoch": 0.8335103558151885, + "grad_norm": 0.2122394769969129, + "learning_rate": 6.7543780354339775e-06, + "loss": 0.0286, + "step": 3139 + }, + { + "epoch": 0.8337758895379713, + "grad_norm": 0.522576364329861, + "learning_rate": 6.752322061551707e-06, + "loss": 0.0286, + "step": 3140 + }, + { + "epoch": 0.8340414232607541, + "grad_norm": 0.4110482587173931, + "learning_rate": 6.7502657498424096e-06, + "loss": 0.0391, + "step": 3141 + }, + { + "epoch": 0.834306956983537, + "grad_norm": 0.6128190842016135, + "learning_rate": 6.7482091007025186e-06, + "loss": 0.0354, + "step": 3142 + }, + { + "epoch": 0.8345724907063197, + "grad_norm": 0.6745000226643804, + "learning_rate": 6.746152114528531e-06, + "loss": 0.0688, + "step": 3143 + }, + { + "epoch": 0.8348380244291025, + "grad_norm": 0.8007507912367674, + "learning_rate": 6.7440947917170115e-06, + "loss": 0.0354, + "step": 3144 + }, + { + "epoch": 0.8351035581518853, + "grad_norm": 0.6698260331644375, + "learning_rate": 6.742037132664585e-06, + "loss": 0.0354, + "step": 3145 + }, + { + "epoch": 0.8353690918746681, + "grad_norm": 0.24987420158630502, + "learning_rate": 6.739979137767944e-06, + "loss": 0.0255, + "step": 3146 + }, + { + "epoch": 0.8356346255974508, + "grad_norm": 3.086918886391466, + "learning_rate": 6.737920807423845e-06, + "loss": 0.1279, + "step": 3147 + }, + { + "epoch": 0.8359001593202336, + "grad_norm": 4.15001563662398, + "learning_rate": 6.735862142029114e-06, + "loss": 0.4551, + "step": 3148 + }, + { + "epoch": 0.8361656930430165, + "grad_norm": 0.03053928725274191, + "learning_rate": 6.733803141980634e-06, + "loss": 0.0012, + "step": 3149 + }, + { + "epoch": 0.8364312267657993, + "grad_norm": 0.7762408060254922, + "learning_rate": 6.731743807675355e-06, + "loss": 0.0752, + "step": 3150 + }, + { + "epoch": 0.8366967604885821, + "grad_norm": 0.3762506038040812, + "learning_rate": 6.729684139510295e-06, + "loss": 0.0391, + "step": 3151 + }, + { + "epoch": 0.8369622942113648, + "grad_norm": 0.004072849880193395, + "learning_rate": 6.727624137882533e-06, + "loss": 0.0001, + "step": 3152 + }, + { + "epoch": 0.8372278279341476, + "grad_norm": 2.690251908046621, + "learning_rate": 6.7255638031892125e-06, + "loss": 0.2129, + "step": 3153 + }, + { + "epoch": 0.8374933616569304, + "grad_norm": 0.025482915226373515, + "learning_rate": 6.723503135827541e-06, + "loss": 0.001, + "step": 3154 + }, + { + "epoch": 0.8377588953797133, + "grad_norm": 1.4672009147487124, + "learning_rate": 6.721442136194791e-06, + "loss": 0.1279, + "step": 3155 + }, + { + "epoch": 0.838024429102496, + "grad_norm": 0.5503061633690576, + "learning_rate": 6.719380804688302e-06, + "loss": 0.0255, + "step": 3156 + }, + { + "epoch": 0.8382899628252788, + "grad_norm": 1.6922121375504098, + "learning_rate": 6.717319141705471e-06, + "loss": 0.1553, + "step": 3157 + }, + { + "epoch": 0.8385554965480616, + "grad_norm": 0.5761254230197108, + "learning_rate": 6.715257147643763e-06, + "loss": 0.0317, + "step": 3158 + }, + { + "epoch": 0.8388210302708444, + "grad_norm": 0.7168709359356019, + "learning_rate": 6.713194822900707e-06, + "loss": 0.0688, + "step": 3159 + }, + { + "epoch": 0.8390865639936272, + "grad_norm": 0.7640588243406803, + "learning_rate": 6.711132167873895e-06, + "loss": 0.0525, + "step": 3160 + }, + { + "epoch": 0.83935209771641, + "grad_norm": 1.3067799790283414, + "learning_rate": 6.709069182960981e-06, + "loss": 0.1035, + "step": 3161 + }, + { + "epoch": 0.8396176314391928, + "grad_norm": 0.1425475128935341, + "learning_rate": 6.707005868559686e-06, + "loss": 0.0057, + "step": 3162 + }, + { + "epoch": 0.8398831651619756, + "grad_norm": 0.07335456319070668, + "learning_rate": 6.704942225067791e-06, + "loss": 0.0027, + "step": 3163 + }, + { + "epoch": 0.8401486988847584, + "grad_norm": 0.01239576895692988, + "learning_rate": 6.702878252883143e-06, + "loss": 0.0004, + "step": 3164 + }, + { + "epoch": 0.8404142326075411, + "grad_norm": 0.17610923528762504, + "learning_rate": 6.700813952403651e-06, + "loss": 0.0229, + "step": 3165 + }, + { + "epoch": 0.8406797663303239, + "grad_norm": 0.4637515961165761, + "learning_rate": 6.698749324027289e-06, + "loss": 0.0432, + "step": 3166 + }, + { + "epoch": 0.8409453000531067, + "grad_norm": 2.1088755089611912, + "learning_rate": 6.696684368152092e-06, + "loss": 0.1191, + "step": 3167 + }, + { + "epoch": 0.8412108337758896, + "grad_norm": 0.031893499709736546, + "learning_rate": 6.694619085176159e-06, + "loss": 0.0012, + "step": 3168 + }, + { + "epoch": 0.8414763674986724, + "grad_norm": 0.028842649269623797, + "learning_rate": 6.692553475497654e-06, + "loss": 0.001, + "step": 3169 + }, + { + "epoch": 0.8417419012214551, + "grad_norm": 0.5352589942197944, + "learning_rate": 6.690487539514799e-06, + "loss": 0.0476, + "step": 3170 + }, + { + "epoch": 0.8420074349442379, + "grad_norm": 0.4519725773386277, + "learning_rate": 6.688421277625885e-06, + "loss": 0.0255, + "step": 3171 + }, + { + "epoch": 0.8422729686670207, + "grad_norm": 0.8144011596296882, + "learning_rate": 6.686354690229261e-06, + "loss": 0.0957, + "step": 3172 + }, + { + "epoch": 0.8425385023898035, + "grad_norm": 0.6006816915205485, + "learning_rate": 6.6842877777233404e-06, + "loss": 0.0354, + "step": 3173 + }, + { + "epoch": 0.8428040361125863, + "grad_norm": 0.48011251169373476, + "learning_rate": 6.6822205405066025e-06, + "loss": 0.0317, + "step": 3174 + }, + { + "epoch": 0.8430695698353691, + "grad_norm": 0.1255371136483073, + "learning_rate": 6.680152978977583e-06, + "loss": 0.0165, + "step": 3175 + }, + { + "epoch": 0.8433351035581519, + "grad_norm": 0.06955653937287123, + "learning_rate": 6.678085093534887e-06, + "loss": 0.0019, + "step": 3176 + }, + { + "epoch": 0.8436006372809347, + "grad_norm": 0.6739512549739441, + "learning_rate": 6.676016884577173e-06, + "loss": 0.0317, + "step": 3177 + }, + { + "epoch": 0.8438661710037175, + "grad_norm": 0.7101276844320112, + "learning_rate": 6.673948352503172e-06, + "loss": 0.0354, + "step": 3178 + }, + { + "epoch": 0.8441317047265002, + "grad_norm": 0.15708363260534108, + "learning_rate": 6.671879497711672e-06, + "loss": 0.0184, + "step": 3179 + }, + { + "epoch": 0.844397238449283, + "grad_norm": 0.9673453342806225, + "learning_rate": 6.6698103206015215e-06, + "loss": 0.0752, + "step": 3180 + }, + { + "epoch": 0.8446627721720659, + "grad_norm": 0.46402537049947107, + "learning_rate": 6.667740821571636e-06, + "loss": 0.0131, + "step": 3181 + }, + { + "epoch": 0.8449283058948487, + "grad_norm": 0.8009107746405942, + "learning_rate": 6.665671001020987e-06, + "loss": 0.063, + "step": 3182 + }, + { + "epoch": 0.8451938396176314, + "grad_norm": 0.554914628501058, + "learning_rate": 6.663600859348616e-06, + "loss": 0.0317, + "step": 3183 + }, + { + "epoch": 0.8454593733404142, + "grad_norm": 0.8080188231269724, + "learning_rate": 6.6615303969536196e-06, + "loss": 0.1191, + "step": 3184 + }, + { + "epoch": 0.845724907063197, + "grad_norm": 0.24909730346342457, + "learning_rate": 6.659459614235157e-06, + "loss": 0.0286, + "step": 3185 + }, + { + "epoch": 0.8459904407859798, + "grad_norm": 0.30614158655043927, + "learning_rate": 6.657388511592453e-06, + "loss": 0.0286, + "step": 3186 + }, + { + "epoch": 0.8462559745087627, + "grad_norm": 0.053944438907348166, + "learning_rate": 6.655317089424791e-06, + "loss": 0.0021, + "step": 3187 + }, + { + "epoch": 0.8465215082315454, + "grad_norm": 3.8955491773801327, + "learning_rate": 6.653245348131517e-06, + "loss": 0.2344, + "step": 3188 + }, + { + "epoch": 0.8467870419543282, + "grad_norm": 0.5444623125034657, + "learning_rate": 6.651173288112038e-06, + "loss": 0.0391, + "step": 3189 + }, + { + "epoch": 0.847052575677111, + "grad_norm": 3.171988295799826, + "learning_rate": 6.649100909765823e-06, + "loss": 0.2129, + "step": 3190 + }, + { + "epoch": 0.8473181093998938, + "grad_norm": 0.18757583785931756, + "learning_rate": 6.647028213492402e-06, + "loss": 0.0165, + "step": 3191 + }, + { + "epoch": 0.8475836431226765, + "grad_norm": 0.2188395478216061, + "learning_rate": 6.644955199691367e-06, + "loss": 0.0057, + "step": 3192 + }, + { + "epoch": 0.8478491768454594, + "grad_norm": 0.2580475326687884, + "learning_rate": 6.642881868762369e-06, + "loss": 0.0255, + "step": 3193 + }, + { + "epoch": 0.8481147105682422, + "grad_norm": 0.07654654152577416, + "learning_rate": 6.640808221105123e-06, + "loss": 0.0031, + "step": 3194 + }, + { + "epoch": 0.848380244291025, + "grad_norm": 1.8242248821573126, + "learning_rate": 6.638734257119402e-06, + "loss": 0.1641, + "step": 3195 + }, + { + "epoch": 0.8486457780138078, + "grad_norm": 0.24761518610691427, + "learning_rate": 6.636659977205044e-06, + "loss": 0.0229, + "step": 3196 + }, + { + "epoch": 0.8489113117365905, + "grad_norm": 0.3587057510086669, + "learning_rate": 6.634585381761946e-06, + "loss": 0.0184, + "step": 3197 + }, + { + "epoch": 0.8491768454593733, + "grad_norm": 0.7115013108219642, + "learning_rate": 6.632510471190065e-06, + "loss": 0.1553, + "step": 3198 + }, + { + "epoch": 0.8494423791821561, + "grad_norm": 1.049351682107704, + "learning_rate": 6.6304352458894176e-06, + "loss": 0.0576, + "step": 3199 + }, + { + "epoch": 0.849707912904939, + "grad_norm": 0.4479542858154741, + "learning_rate": 6.628359706260085e-06, + "loss": 0.0391, + "step": 3200 + }, + { + "epoch": 0.8499734466277217, + "grad_norm": 2.141219315175153, + "learning_rate": 6.6262838527022046e-06, + "loss": 0.1934, + "step": 3201 + }, + { + "epoch": 0.8502389803505045, + "grad_norm": 0.9732665925785687, + "learning_rate": 6.624207685615978e-06, + "loss": 0.0525, + "step": 3202 + }, + { + "epoch": 0.8505045140732873, + "grad_norm": 1.1417219242373788, + "learning_rate": 6.6221312054016665e-06, + "loss": 0.0815, + "step": 3203 + }, + { + "epoch": 0.8507700477960701, + "grad_norm": 0.26803551530131214, + "learning_rate": 6.620054412459588e-06, + "loss": 0.0255, + "step": 3204 + }, + { + "epoch": 0.851035581518853, + "grad_norm": 0.16279915945434512, + "learning_rate": 6.617977307190127e-06, + "loss": 0.0064, + "step": 3205 + }, + { + "epoch": 0.8513011152416357, + "grad_norm": 0.9295452638724799, + "learning_rate": 6.6158998899937225e-06, + "loss": 0.0688, + "step": 3206 + }, + { + "epoch": 0.8515666489644185, + "grad_norm": 0.19635952218677447, + "learning_rate": 6.613822161270877e-06, + "loss": 0.0258, + "step": 3207 + }, + { + "epoch": 0.8518321826872013, + "grad_norm": 6.88297992520138, + "learning_rate": 6.6117441214221545e-06, + "loss": 3.4062, + "step": 3208 + }, + { + "epoch": 0.8520977164099841, + "grad_norm": 0.8291226787475618, + "learning_rate": 6.60966577084817e-06, + "loss": 0.063, + "step": 3209 + }, + { + "epoch": 0.8523632501327668, + "grad_norm": 0.12794987208417472, + "learning_rate": 6.607587109949612e-06, + "loss": 0.0045, + "step": 3210 + }, + { + "epoch": 0.8526287838555496, + "grad_norm": 0.44453377586873516, + "learning_rate": 6.605508139127218e-06, + "loss": 0.0391, + "step": 3211 + }, + { + "epoch": 0.8528943175783325, + "grad_norm": 0.42246252378882915, + "learning_rate": 6.603428858781791e-06, + "loss": 0.0354, + "step": 3212 + }, + { + "epoch": 0.8531598513011153, + "grad_norm": 0.1399369738246371, + "learning_rate": 6.601349269314188e-06, + "loss": 0.0165, + "step": 3213 + }, + { + "epoch": 0.8534253850238981, + "grad_norm": 0.1272210649705708, + "learning_rate": 6.599269371125334e-06, + "loss": 0.0146, + "step": 3214 + }, + { + "epoch": 0.8536909187466808, + "grad_norm": 0.9107149473182585, + "learning_rate": 6.5971891646162065e-06, + "loss": 0.0576, + "step": 3215 + }, + { + "epoch": 0.8539564524694636, + "grad_norm": 0.16435397706688415, + "learning_rate": 6.595108650187847e-06, + "loss": 0.0057, + "step": 3216 + }, + { + "epoch": 0.8542219861922464, + "grad_norm": 1.6401193789128157, + "learning_rate": 6.593027828241351e-06, + "loss": 0.1113, + "step": 3217 + }, + { + "epoch": 0.8544875199150292, + "grad_norm": 1.098830943084169, + "learning_rate": 6.590946699177876e-06, + "loss": 0.0576, + "step": 3218 + }, + { + "epoch": 0.854753053637812, + "grad_norm": 0.8055493531170383, + "learning_rate": 6.588865263398643e-06, + "loss": 0.0476, + "step": 3219 + }, + { + "epoch": 0.8550185873605948, + "grad_norm": 0.6301900092568686, + "learning_rate": 6.586783521304926e-06, + "loss": 0.1455, + "step": 3220 + }, + { + "epoch": 0.8552841210833776, + "grad_norm": 0.23150481671629922, + "learning_rate": 6.5847014732980594e-06, + "loss": 0.0258, + "step": 3221 + }, + { + "epoch": 0.8555496548061604, + "grad_norm": 0.8906812516152857, + "learning_rate": 6.582619119779439e-06, + "loss": 0.1113, + "step": 3222 + }, + { + "epoch": 0.8558151885289432, + "grad_norm": 1.4977799284236537, + "learning_rate": 6.5805364611505186e-06, + "loss": 0.1279, + "step": 3223 + }, + { + "epoch": 0.8560807222517259, + "grad_norm": 0.6581186250412577, + "learning_rate": 6.578453497812808e-06, + "loss": 0.1455, + "step": 3224 + }, + { + "epoch": 0.8563462559745088, + "grad_norm": 0.7191091619477337, + "learning_rate": 6.576370230167879e-06, + "loss": 0.1367, + "step": 3225 + }, + { + "epoch": 0.8566117896972916, + "grad_norm": 0.8638305189657841, + "learning_rate": 6.574286658617363e-06, + "loss": 0.0525, + "step": 3226 + }, + { + "epoch": 0.8568773234200744, + "grad_norm": 0.6577935556472292, + "learning_rate": 6.572202783562943e-06, + "loss": 0.1455, + "step": 3227 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 0.17257469044605947, + "learning_rate": 6.5701186054063704e-06, + "loss": 0.0057, + "step": 3228 + }, + { + "epoch": 0.8574083908656399, + "grad_norm": 0.46938986874827243, + "learning_rate": 6.568034124549449e-06, + "loss": 0.0165, + "step": 3229 + }, + { + "epoch": 0.8576739245884227, + "grad_norm": 0.16315356885989296, + "learning_rate": 6.565949341394041e-06, + "loss": 0.0255, + "step": 3230 + }, + { + "epoch": 0.8579394583112055, + "grad_norm": 0.24432795826056708, + "learning_rate": 6.56386425634207e-06, + "loss": 0.0317, + "step": 3231 + }, + { + "epoch": 0.8582049920339884, + "grad_norm": 0.24461026179530254, + "learning_rate": 6.561778869795513e-06, + "loss": 0.0317, + "step": 3232 + }, + { + "epoch": 0.8584705257567711, + "grad_norm": 1.1323792615832808, + "learning_rate": 6.559693182156411e-06, + "loss": 0.0525, + "step": 3233 + }, + { + "epoch": 0.8587360594795539, + "grad_norm": 0.07855000060787923, + "learning_rate": 6.557607193826856e-06, + "loss": 0.0035, + "step": 3234 + }, + { + "epoch": 0.8590015932023367, + "grad_norm": 0.26079508432031945, + "learning_rate": 6.5555209052090055e-06, + "loss": 0.0092, + "step": 3235 + }, + { + "epoch": 0.8592671269251195, + "grad_norm": 0.25794265307379477, + "learning_rate": 6.5534343167050696e-06, + "loss": 0.0317, + "step": 3236 + }, + { + "epoch": 0.8595326606479022, + "grad_norm": 0.5906172767825625, + "learning_rate": 6.551347428717318e-06, + "loss": 0.1113, + "step": 3237 + }, + { + "epoch": 0.859798194370685, + "grad_norm": 0.7133931104069032, + "learning_rate": 6.5492602416480795e-06, + "loss": 0.063, + "step": 3238 + }, + { + "epoch": 0.8600637280934679, + "grad_norm": 0.012482317570715346, + "learning_rate": 6.547172755899739e-06, + "loss": 0.0004, + "step": 3239 + }, + { + "epoch": 0.8603292618162507, + "grad_norm": 0.6090701100096286, + "learning_rate": 6.545084971874738e-06, + "loss": 0.0576, + "step": 3240 + }, + { + "epoch": 0.8605947955390335, + "grad_norm": 0.008150999610717878, + "learning_rate": 6.542996889975577e-06, + "loss": 0.0002, + "step": 3241 + }, + { + "epoch": 0.8608603292618162, + "grad_norm": 0.5954237369566643, + "learning_rate": 6.540908510604813e-06, + "loss": 0.1279, + "step": 3242 + }, + { + "epoch": 0.861125862984599, + "grad_norm": 0.1865774170805672, + "learning_rate": 6.538819834165062e-06, + "loss": 0.0255, + "step": 3243 + }, + { + "epoch": 0.8613913967073819, + "grad_norm": 1.8366051377187302, + "learning_rate": 6.5367308610589955e-06, + "loss": 0.0889, + "step": 3244 + }, + { + "epoch": 0.8616569304301647, + "grad_norm": 0.17143814823574377, + "learning_rate": 6.5346415916893415e-06, + "loss": 0.0255, + "step": 3245 + }, + { + "epoch": 0.8619224641529474, + "grad_norm": 0.43981923627730096, + "learning_rate": 6.53255202645889e-06, + "loss": 0.0432, + "step": 3246 + }, + { + "epoch": 0.8621879978757302, + "grad_norm": 1.1644488336637033, + "learning_rate": 6.530462165770482e-06, + "loss": 0.0752, + "step": 3247 + }, + { + "epoch": 0.862453531598513, + "grad_norm": 0.7830218834200647, + "learning_rate": 6.5283720100270194e-06, + "loss": 0.0757, + "step": 3248 + }, + { + "epoch": 0.8627190653212958, + "grad_norm": 0.4963750620619639, + "learning_rate": 6.526281559631457e-06, + "loss": 0.0889, + "step": 3249 + }, + { + "epoch": 0.8629845990440786, + "grad_norm": 1.3108882806450404, + "learning_rate": 6.524190814986811e-06, + "loss": 0.1367, + "step": 3250 + }, + { + "epoch": 0.8632501327668614, + "grad_norm": 0.69872732806686, + "learning_rate": 6.522099776496154e-06, + "loss": 0.0476, + "step": 3251 + }, + { + "epoch": 0.8635156664896442, + "grad_norm": 0.5597431305272228, + "learning_rate": 6.52000844456261e-06, + "loss": 0.1113, + "step": 3252 + }, + { + "epoch": 0.863781200212427, + "grad_norm": 0.4354024232566979, + "learning_rate": 6.517916819589365e-06, + "loss": 0.0432, + "step": 3253 + }, + { + "epoch": 0.8640467339352098, + "grad_norm": 0.027259200145566514, + "learning_rate": 6.51582490197966e-06, + "loss": 0.0007, + "step": 3254 + }, + { + "epoch": 0.8643122676579925, + "grad_norm": 0.6656906770031343, + "learning_rate": 6.5137326921367895e-06, + "loss": 0.0432, + "step": 3255 + }, + { + "epoch": 0.8645778013807753, + "grad_norm": 0.5539487707554477, + "learning_rate": 6.511640190464111e-06, + "loss": 0.0889, + "step": 3256 + }, + { + "epoch": 0.8648433351035582, + "grad_norm": 0.687228899664994, + "learning_rate": 6.509547397365031e-06, + "loss": 0.0576, + "step": 3257 + }, + { + "epoch": 0.865108868826341, + "grad_norm": 0.2136520061331435, + "learning_rate": 6.507454313243016e-06, + "loss": 0.0354, + "step": 3258 + }, + { + "epoch": 0.8653744025491238, + "grad_norm": 0.23171222416509865, + "learning_rate": 6.505360938501588e-06, + "loss": 0.0391, + "step": 3259 + }, + { + "epoch": 0.8656399362719065, + "grad_norm": 0.4594024885086515, + "learning_rate": 6.503267273544326e-06, + "loss": 0.1035, + "step": 3260 + }, + { + "epoch": 0.8659054699946893, + "grad_norm": 0.48885731080788136, + "learning_rate": 6.501173318774863e-06, + "loss": 0.0525, + "step": 3261 + }, + { + "epoch": 0.8661710037174721, + "grad_norm": 0.26673166444501073, + "learning_rate": 6.499079074596887e-06, + "loss": 0.0146, + "step": 3262 + }, + { + "epoch": 0.866436537440255, + "grad_norm": 0.44375727625178163, + "learning_rate": 6.496984541414148e-06, + "loss": 0.0889, + "step": 3263 + }, + { + "epoch": 0.8667020711630377, + "grad_norm": 0.28667168026036893, + "learning_rate": 6.494889719630444e-06, + "loss": 0.0432, + "step": 3264 + }, + { + "epoch": 0.8669676048858205, + "grad_norm": 1.4390802218679875, + "learning_rate": 6.492794609649632e-06, + "loss": 0.1113, + "step": 3265 + }, + { + "epoch": 0.8672331386086033, + "grad_norm": 1.5636089711532934, + "learning_rate": 6.490699211875625e-06, + "loss": 0.1455, + "step": 3266 + }, + { + "epoch": 0.8674986723313861, + "grad_norm": 0.24531994294810155, + "learning_rate": 6.488603526712391e-06, + "loss": 0.0432, + "step": 3267 + }, + { + "epoch": 0.8677642060541689, + "grad_norm": 0.42432410420185457, + "learning_rate": 6.486507554563953e-06, + "loss": 0.0525, + "step": 3268 + }, + { + "epoch": 0.8680297397769516, + "grad_norm": 0.7360150600177398, + "learning_rate": 6.484411295834392e-06, + "loss": 0.0752, + "step": 3269 + }, + { + "epoch": 0.8682952734997345, + "grad_norm": 0.6821938959978506, + "learning_rate": 6.482314750927839e-06, + "loss": 0.0688, + "step": 3270 + }, + { + "epoch": 0.8685608072225173, + "grad_norm": 1.1711482275012044, + "learning_rate": 6.480217920248485e-06, + "loss": 0.1191, + "step": 3271 + }, + { + "epoch": 0.8688263409453001, + "grad_norm": 0.798247437625764, + "learning_rate": 6.478120804200573e-06, + "loss": 0.0815, + "step": 3272 + }, + { + "epoch": 0.8690918746680828, + "grad_norm": 0.4543173831281007, + "learning_rate": 6.476023403188403e-06, + "loss": 0.0688, + "step": 3273 + }, + { + "epoch": 0.8693574083908656, + "grad_norm": 0.5397033337210341, + "learning_rate": 6.473925717616329e-06, + "loss": 0.0688, + "step": 3274 + }, + { + "epoch": 0.8696229421136484, + "grad_norm": 0.42345382800911124, + "learning_rate": 6.471827747888759e-06, + "loss": 0.0688, + "step": 3275 + }, + { + "epoch": 0.8698884758364313, + "grad_norm": 0.5470754707447224, + "learning_rate": 6.4697294944101585e-06, + "loss": 0.0476, + "step": 3276 + }, + { + "epoch": 0.8701540095592141, + "grad_norm": 0.440906964358512, + "learning_rate": 6.467630957585046e-06, + "loss": 0.0576, + "step": 3277 + }, + { + "epoch": 0.8704195432819968, + "grad_norm": 0.21419471428860137, + "learning_rate": 6.4655321378179935e-06, + "loss": 0.0391, + "step": 3278 + }, + { + "epoch": 0.8706850770047796, + "grad_norm": 0.23713854753225969, + "learning_rate": 6.463433035513628e-06, + "loss": 0.0072, + "step": 3279 + }, + { + "epoch": 0.8709506107275624, + "grad_norm": 0.173169585215048, + "learning_rate": 6.461333651076634e-06, + "loss": 0.0317, + "step": 3280 + }, + { + "epoch": 0.8712161444503452, + "grad_norm": 0.17555422399363732, + "learning_rate": 6.459233984911748e-06, + "loss": 0.0317, + "step": 3281 + }, + { + "epoch": 0.8714816781731279, + "grad_norm": 0.29772566723410554, + "learning_rate": 6.457134037423758e-06, + "loss": 0.0432, + "step": 3282 + }, + { + "epoch": 0.8717472118959108, + "grad_norm": 0.19453971987243454, + "learning_rate": 6.455033809017513e-06, + "loss": 0.0317, + "step": 3283 + }, + { + "epoch": 0.8720127456186936, + "grad_norm": 0.6033204279885397, + "learning_rate": 6.4529333000979074e-06, + "loss": 0.0889, + "step": 3284 + }, + { + "epoch": 0.8722782793414764, + "grad_norm": 0.3684431085916947, + "learning_rate": 6.450832511069898e-06, + "loss": 0.0432, + "step": 3285 + }, + { + "epoch": 0.8725438130642592, + "grad_norm": 0.7312523563841785, + "learning_rate": 6.448731442338491e-06, + "loss": 0.0525, + "step": 3286 + }, + { + "epoch": 0.8728093467870419, + "grad_norm": 0.16913062192706313, + "learning_rate": 6.44663009430875e-06, + "loss": 0.0286, + "step": 3287 + }, + { + "epoch": 0.8730748805098247, + "grad_norm": 0.4158475721007959, + "learning_rate": 6.444528467385789e-06, + "loss": 0.1113, + "step": 3288 + }, + { + "epoch": 0.8733404142326076, + "grad_norm": 0.16401731313830273, + "learning_rate": 6.442426561974774e-06, + "loss": 0.0286, + "step": 3289 + }, + { + "epoch": 0.8736059479553904, + "grad_norm": 0.1086997764169726, + "learning_rate": 6.44032437848093e-06, + "loss": 0.0045, + "step": 3290 + }, + { + "epoch": 0.8738714816781731, + "grad_norm": 0.6222199868612331, + "learning_rate": 6.438221917309533e-06, + "loss": 0.0286, + "step": 3291 + }, + { + "epoch": 0.8741370154009559, + "grad_norm": 0.9452660074974694, + "learning_rate": 6.4361191788659125e-06, + "loss": 0.0525, + "step": 3292 + }, + { + "epoch": 0.8744025491237387, + "grad_norm": 0.43957282763478517, + "learning_rate": 6.434016163555452e-06, + "loss": 0.1035, + "step": 3293 + }, + { + "epoch": 0.8746680828465215, + "grad_norm": 0.46308811320934234, + "learning_rate": 6.431912871783587e-06, + "loss": 0.1035, + "step": 3294 + }, + { + "epoch": 0.8749336165693044, + "grad_norm": 0.013673713399171403, + "learning_rate": 6.42980930395581e-06, + "loss": 0.0004, + "step": 3295 + }, + { + "epoch": 0.8751991502920871, + "grad_norm": 0.15081473226938324, + "learning_rate": 6.4277054604776616e-06, + "loss": 0.0258, + "step": 3296 + }, + { + "epoch": 0.8754646840148699, + "grad_norm": 1.0480367128446855, + "learning_rate": 6.425601341754738e-06, + "loss": 0.1113, + "step": 3297 + }, + { + "epoch": 0.8757302177376527, + "grad_norm": 0.5006703602269802, + "learning_rate": 6.423496948192689e-06, + "loss": 0.0957, + "step": 3298 + }, + { + "epoch": 0.8759957514604355, + "grad_norm": 2.3363828501114097, + "learning_rate": 6.421392280197216e-06, + "loss": 0.2559, + "step": 3299 + }, + { + "epoch": 0.8762612851832182, + "grad_norm": 0.7029521296325448, + "learning_rate": 6.419287338174077e-06, + "loss": 0.0576, + "step": 3300 + }, + { + "epoch": 0.876526818906001, + "grad_norm": 0.45191331498694903, + "learning_rate": 6.417182122529077e-06, + "loss": 0.0688, + "step": 3301 + }, + { + "epoch": 0.8767923526287839, + "grad_norm": 0.19963055476736272, + "learning_rate": 6.4150766336680795e-06, + "loss": 0.0354, + "step": 3302 + }, + { + "epoch": 0.8770578863515667, + "grad_norm": 0.6538835486175559, + "learning_rate": 6.412970871996995e-06, + "loss": 0.1191, + "step": 3303 + }, + { + "epoch": 0.8773234200743495, + "grad_norm": 0.06572701642875628, + "learning_rate": 6.4108648379217905e-06, + "loss": 0.0025, + "step": 3304 + }, + { + "epoch": 0.8775889537971322, + "grad_norm": 0.011360450833469192, + "learning_rate": 6.4087585318484864e-06, + "loss": 0.0003, + "step": 3305 + }, + { + "epoch": 0.877854487519915, + "grad_norm": 0.8549716182950936, + "learning_rate": 6.406651954183151e-06, + "loss": 0.0835, + "step": 3306 + }, + { + "epoch": 0.8781200212426978, + "grad_norm": 0.3449944056046003, + "learning_rate": 6.404545105331909e-06, + "loss": 0.0576, + "step": 3307 + }, + { + "epoch": 0.8783855549654807, + "grad_norm": 1.2765972907125898, + "learning_rate": 6.402437985700936e-06, + "loss": 0.0889, + "step": 3308 + }, + { + "epoch": 0.8786510886882634, + "grad_norm": 0.41076597632311446, + "learning_rate": 6.400330595696458e-06, + "loss": 0.0957, + "step": 3309 + }, + { + "epoch": 0.8789166224110462, + "grad_norm": 0.4385630627786458, + "learning_rate": 6.3982229357247585e-06, + "loss": 0.0889, + "step": 3310 + }, + { + "epoch": 0.879182156133829, + "grad_norm": 0.2584978089076774, + "learning_rate": 6.3961150061921664e-06, + "loss": 0.0476, + "step": 3311 + }, + { + "epoch": 0.8794476898566118, + "grad_norm": 1.4199143757769417, + "learning_rate": 6.394006807505068e-06, + "loss": 0.1641, + "step": 3312 + }, + { + "epoch": 0.8797132235793946, + "grad_norm": 0.26789714753189203, + "learning_rate": 6.391898340069896e-06, + "loss": 0.0476, + "step": 3313 + }, + { + "epoch": 0.8799787573021773, + "grad_norm": 0.19469842536179863, + "learning_rate": 6.389789604293141e-06, + "loss": 0.0354, + "step": 3314 + }, + { + "epoch": 0.8802442910249602, + "grad_norm": 0.4450091903230278, + "learning_rate": 6.387680600581342e-06, + "loss": 0.0476, + "step": 3315 + }, + { + "epoch": 0.880509824747743, + "grad_norm": 0.14000910419487178, + "learning_rate": 6.3855713293410874e-06, + "loss": 0.0064, + "step": 3316 + }, + { + "epoch": 0.8807753584705258, + "grad_norm": 0.5654529236853104, + "learning_rate": 6.383461790979023e-06, + "loss": 0.0317, + "step": 3317 + }, + { + "epoch": 0.8810408921933085, + "grad_norm": 0.0695671847885732, + "learning_rate": 6.381351985901842e-06, + "loss": 0.0027, + "step": 3318 + }, + { + "epoch": 0.8813064259160913, + "grad_norm": 0.1870648863948819, + "learning_rate": 6.3792419145162895e-06, + "loss": 0.0354, + "step": 3319 + }, + { + "epoch": 0.8815719596388741, + "grad_norm": 0.39294656469894834, + "learning_rate": 6.377131577229164e-06, + "loss": 0.1035, + "step": 3320 + }, + { + "epoch": 0.881837493361657, + "grad_norm": 0.20899723026997796, + "learning_rate": 6.3750209744473105e-06, + "loss": 0.0103, + "step": 3321 + }, + { + "epoch": 0.8821030270844398, + "grad_norm": 0.40989021098921186, + "learning_rate": 6.37291010657763e-06, + "loss": 0.0889, + "step": 3322 + }, + { + "epoch": 0.8823685608072225, + "grad_norm": 0.4276479882944912, + "learning_rate": 6.370798974027073e-06, + "loss": 0.0815, + "step": 3323 + }, + { + "epoch": 0.8826340945300053, + "grad_norm": 0.4550786893387324, + "learning_rate": 6.3686875772026415e-06, + "loss": 0.063, + "step": 3324 + }, + { + "epoch": 0.8828996282527881, + "grad_norm": 0.5143493806068249, + "learning_rate": 6.366575916511385e-06, + "loss": 0.0752, + "step": 3325 + }, + { + "epoch": 0.8831651619755709, + "grad_norm": 0.6414916731035794, + "learning_rate": 6.364463992360411e-06, + "loss": 0.0752, + "step": 3326 + }, + { + "epoch": 0.8834306956983536, + "grad_norm": 0.38725052652653363, + "learning_rate": 6.362351805156871e-06, + "loss": 0.0889, + "step": 3327 + }, + { + "epoch": 0.8836962294211365, + "grad_norm": 0.44278788489065524, + "learning_rate": 6.360239355307972e-06, + "loss": 0.0286, + "step": 3328 + }, + { + "epoch": 0.8839617631439193, + "grad_norm": 0.2928541896779108, + "learning_rate": 6.3581266432209675e-06, + "loss": 0.0476, + "step": 3329 + }, + { + "epoch": 0.8842272968667021, + "grad_norm": 0.38933961657871224, + "learning_rate": 6.356013669303162e-06, + "loss": 0.0815, + "step": 3330 + }, + { + "epoch": 0.8844928305894849, + "grad_norm": 0.33912581407910086, + "learning_rate": 6.353900433961915e-06, + "loss": 0.0576, + "step": 3331 + }, + { + "epoch": 0.8847583643122676, + "grad_norm": 0.42709743125738914, + "learning_rate": 6.351786937604632e-06, + "loss": 0.0576, + "step": 3332 + }, + { + "epoch": 0.8850238980350504, + "grad_norm": 0.34933134363313856, + "learning_rate": 6.349673180638769e-06, + "loss": 0.0525, + "step": 3333 + }, + { + "epoch": 0.8852894317578333, + "grad_norm": 0.29501898050392794, + "learning_rate": 6.347559163471836e-06, + "loss": 0.0525, + "step": 3334 + }, + { + "epoch": 0.8855549654806161, + "grad_norm": 0.27055076750205703, + "learning_rate": 6.345444886511389e-06, + "loss": 0.0147, + "step": 3335 + }, + { + "epoch": 0.8858204992033988, + "grad_norm": 0.4116154407404248, + "learning_rate": 6.343330350165037e-06, + "loss": 0.0354, + "step": 3336 + }, + { + "epoch": 0.8860860329261816, + "grad_norm": 0.2453390138634745, + "learning_rate": 6.341215554840435e-06, + "loss": 0.0391, + "step": 3337 + }, + { + "epoch": 0.8863515666489644, + "grad_norm": 0.831956602243331, + "learning_rate": 6.3391005009452936e-06, + "loss": 0.1191, + "step": 3338 + }, + { + "epoch": 0.8866171003717472, + "grad_norm": 0.32779055539253976, + "learning_rate": 6.336985188887367e-06, + "loss": 0.0525, + "step": 3339 + }, + { + "epoch": 0.88688263409453, + "grad_norm": 0.32625824036164164, + "learning_rate": 6.334869619074465e-06, + "loss": 0.0576, + "step": 3340 + }, + { + "epoch": 0.8871481678173128, + "grad_norm": 0.17544827150635783, + "learning_rate": 6.3327537919144444e-06, + "loss": 0.0103, + "step": 3341 + }, + { + "epoch": 0.8874137015400956, + "grad_norm": 0.32338471354685416, + "learning_rate": 6.3306377078152104e-06, + "loss": 0.0576, + "step": 3342 + }, + { + "epoch": 0.8876792352628784, + "grad_norm": 0.3929385009421306, + "learning_rate": 6.3285213671847215e-06, + "loss": 0.0752, + "step": 3343 + }, + { + "epoch": 0.8879447689856612, + "grad_norm": 0.4386600058747594, + "learning_rate": 6.326404770430979e-06, + "loss": 0.0525, + "step": 3344 + }, + { + "epoch": 0.8882103027084439, + "grad_norm": 0.3652890199388593, + "learning_rate": 6.324287917962042e-06, + "loss": 0.0286, + "step": 3345 + }, + { + "epoch": 0.8884758364312267, + "grad_norm": 0.5785223570314608, + "learning_rate": 6.322170810186013e-06, + "loss": 0.0391, + "step": 3346 + }, + { + "epoch": 0.8887413701540096, + "grad_norm": 0.309008123856303, + "learning_rate": 6.320053447511044e-06, + "loss": 0.0354, + "step": 3347 + }, + { + "epoch": 0.8890069038767924, + "grad_norm": 0.22693343309296524, + "learning_rate": 6.3179358303453386e-06, + "loss": 0.0391, + "step": 3348 + }, + { + "epoch": 0.8892724375995752, + "grad_norm": 0.4962450987077813, + "learning_rate": 6.315817959097151e-06, + "loss": 0.0576, + "step": 3349 + }, + { + "epoch": 0.8895379713223579, + "grad_norm": 0.23953374116093432, + "learning_rate": 6.313699834174777e-06, + "loss": 0.0391, + "step": 3350 + }, + { + "epoch": 0.8898035050451407, + "grad_norm": 2.561173409001289, + "learning_rate": 6.31158145598657e-06, + "loss": 0.2344, + "step": 3351 + }, + { + "epoch": 0.8900690387679235, + "grad_norm": 0.00909595239300903, + "learning_rate": 6.309462824940926e-06, + "loss": 0.0003, + "step": 3352 + }, + { + "epoch": 0.8903345724907064, + "grad_norm": 0.31831215657168166, + "learning_rate": 6.307343941446295e-06, + "loss": 0.0525, + "step": 3353 + }, + { + "epoch": 0.8906001062134891, + "grad_norm": 0.46194486102429155, + "learning_rate": 6.30522480591117e-06, + "loss": 0.0286, + "step": 3354 + }, + { + "epoch": 0.8908656399362719, + "grad_norm": 0.3895614360509502, + "learning_rate": 6.303105418744095e-06, + "loss": 0.0476, + "step": 3355 + }, + { + "epoch": 0.8911311736590547, + "grad_norm": 0.006909945690377405, + "learning_rate": 6.300985780353664e-06, + "loss": 0.0002, + "step": 3356 + }, + { + "epoch": 0.8913967073818375, + "grad_norm": 0.36813541799498656, + "learning_rate": 6.298865891148518e-06, + "loss": 0.063, + "step": 3357 + }, + { + "epoch": 0.8916622411046203, + "grad_norm": 0.4104472563200188, + "learning_rate": 6.296745751537349e-06, + "loss": 0.0752, + "step": 3358 + }, + { + "epoch": 0.891927774827403, + "grad_norm": 0.30637433343333986, + "learning_rate": 6.294625361928891e-06, + "loss": 0.0476, + "step": 3359 + }, + { + "epoch": 0.8921933085501859, + "grad_norm": 0.5314610693001266, + "learning_rate": 6.292504722731933e-06, + "loss": 0.063, + "step": 3360 + }, + { + "epoch": 0.8924588422729687, + "grad_norm": 0.34360172854533594, + "learning_rate": 6.290383834355307e-06, + "loss": 0.0576, + "step": 3361 + }, + { + "epoch": 0.8927243759957515, + "grad_norm": 0.13180720540967106, + "learning_rate": 6.2882626972078976e-06, + "loss": 0.0064, + "step": 3362 + }, + { + "epoch": 0.8929899097185342, + "grad_norm": 0.05302168509486064, + "learning_rate": 6.286141311698634e-06, + "loss": 0.0014, + "step": 3363 + }, + { + "epoch": 0.893255443441317, + "grad_norm": 0.8508518442871725, + "learning_rate": 6.284019678236492e-06, + "loss": 0.1035, + "step": 3364 + }, + { + "epoch": 0.8935209771640998, + "grad_norm": 0.3654550943655776, + "learning_rate": 6.281897797230502e-06, + "loss": 0.0286, + "step": 3365 + }, + { + "epoch": 0.8937865108868827, + "grad_norm": 0.3560818754057503, + "learning_rate": 6.279775669089734e-06, + "loss": 0.0576, + "step": 3366 + }, + { + "epoch": 0.8940520446096655, + "grad_norm": 0.25460935938572987, + "learning_rate": 6.27765329422331e-06, + "loss": 0.0432, + "step": 3367 + }, + { + "epoch": 0.8943175783324482, + "grad_norm": 0.4260190590411832, + "learning_rate": 6.275530673040402e-06, + "loss": 0.0752, + "step": 3368 + }, + { + "epoch": 0.894583112055231, + "grad_norm": 0.3011751983631827, + "learning_rate": 6.273407805950221e-06, + "loss": 0.0476, + "step": 3369 + }, + { + "epoch": 0.8948486457780138, + "grad_norm": 0.3364992726232293, + "learning_rate": 6.2712846933620344e-06, + "loss": 0.0317, + "step": 3370 + }, + { + "epoch": 0.8951141795007966, + "grad_norm": 0.2999707670552433, + "learning_rate": 6.269161335685152e-06, + "loss": 0.0476, + "step": 3371 + }, + { + "epoch": 0.8953797132235793, + "grad_norm": 0.443720813661731, + "learning_rate": 6.267037733328933e-06, + "loss": 0.0889, + "step": 3372 + }, + { + "epoch": 0.8956452469463622, + "grad_norm": 0.45907935323635557, + "learning_rate": 6.264913886702783e-06, + "loss": 0.063, + "step": 3373 + }, + { + "epoch": 0.895910780669145, + "grad_norm": 0.26047840032572134, + "learning_rate": 6.262789796216152e-06, + "loss": 0.0432, + "step": 3374 + }, + { + "epoch": 0.8961763143919278, + "grad_norm": 0.009206020326249774, + "learning_rate": 6.260665462278544e-06, + "loss": 0.0003, + "step": 3375 + }, + { + "epoch": 0.8964418481147106, + "grad_norm": 0.40065499499297064, + "learning_rate": 6.258540885299501e-06, + "loss": 0.0576, + "step": 3376 + }, + { + "epoch": 0.8967073818374933, + "grad_norm": 0.34670805062787113, + "learning_rate": 6.256416065688618e-06, + "loss": 0.0476, + "step": 3377 + }, + { + "epoch": 0.8969729155602761, + "grad_norm": 0.4112965384416454, + "learning_rate": 6.254291003855537e-06, + "loss": 0.0752, + "step": 3378 + }, + { + "epoch": 0.897238449283059, + "grad_norm": 1.3605437965603657, + "learning_rate": 6.252165700209942e-06, + "loss": 0.1191, + "step": 3379 + }, + { + "epoch": 0.8975039830058418, + "grad_norm": 0.3860865227542549, + "learning_rate": 6.250040155161569e-06, + "loss": 0.0317, + "step": 3380 + }, + { + "epoch": 0.8977695167286245, + "grad_norm": 0.6507980782696453, + "learning_rate": 6.247914369120198e-06, + "loss": 0.0889, + "step": 3381 + }, + { + "epoch": 0.8980350504514073, + "grad_norm": 0.004289057567750268, + "learning_rate": 6.245788342495653e-06, + "loss": 0.0001, + "step": 3382 + }, + { + "epoch": 0.8983005841741901, + "grad_norm": 0.41270474091668763, + "learning_rate": 6.243662075697809e-06, + "loss": 0.0889, + "step": 3383 + }, + { + "epoch": 0.8985661178969729, + "grad_norm": 0.01866320333409826, + "learning_rate": 6.241535569136584e-06, + "loss": 0.0007, + "step": 3384 + }, + { + "epoch": 0.8988316516197558, + "grad_norm": 1.2793161509856161, + "learning_rate": 6.239408823221944e-06, + "loss": 0.1562, + "step": 3385 + }, + { + "epoch": 0.8990971853425385, + "grad_norm": 0.3489002016581621, + "learning_rate": 6.237281838363901e-06, + "loss": 0.0576, + "step": 3386 + }, + { + "epoch": 0.8993627190653213, + "grad_norm": 0.4171727773232605, + "learning_rate": 6.235154614972511e-06, + "loss": 0.0576, + "step": 3387 + }, + { + "epoch": 0.8996282527881041, + "grad_norm": 0.14129597978333727, + "learning_rate": 6.233027153457878e-06, + "loss": 0.0064, + "step": 3388 + }, + { + "epoch": 0.8998937865108869, + "grad_norm": 0.35637718652767775, + "learning_rate": 6.230899454230152e-06, + "loss": 0.0525, + "step": 3389 + }, + { + "epoch": 0.9001593202336696, + "grad_norm": 0.4101868351443795, + "learning_rate": 6.22877151769953e-06, + "loss": 0.0576, + "step": 3390 + }, + { + "epoch": 0.9004248539564524, + "grad_norm": 0.2233940019817123, + "learning_rate": 6.226643344276251e-06, + "loss": 0.0354, + "step": 3391 + }, + { + "epoch": 0.9006903876792353, + "grad_norm": 0.2070839114404063, + "learning_rate": 6.224514934370601e-06, + "loss": 0.0064, + "step": 3392 + }, + { + "epoch": 0.9009559214020181, + "grad_norm": 0.4209368148957864, + "learning_rate": 6.222386288392914e-06, + "loss": 0.0432, + "step": 3393 + }, + { + "epoch": 0.9012214551248009, + "grad_norm": 0.010810171154702881, + "learning_rate": 6.220257406753568e-06, + "loss": 0.0003, + "step": 3394 + }, + { + "epoch": 0.9014869888475836, + "grad_norm": 0.3727355778801294, + "learning_rate": 6.2181282898629845e-06, + "loss": 0.0815, + "step": 3395 + }, + { + "epoch": 0.9017525225703664, + "grad_norm": 0.7951692595350958, + "learning_rate": 6.2159989381316345e-06, + "loss": 0.0889, + "step": 3396 + }, + { + "epoch": 0.9020180562931492, + "grad_norm": 0.1742913524844309, + "learning_rate": 6.2138693519700274e-06, + "loss": 0.0286, + "step": 3397 + }, + { + "epoch": 0.9022835900159321, + "grad_norm": 0.5535012932645996, + "learning_rate": 6.211739531788729e-06, + "loss": 0.0815, + "step": 3398 + }, + { + "epoch": 0.9025491237387148, + "grad_norm": 1.2459968576206453, + "learning_rate": 6.209609477998339e-06, + "loss": 0.1113, + "step": 3399 + }, + { + "epoch": 0.9028146574614976, + "grad_norm": 0.557973090065412, + "learning_rate": 6.207479191009509e-06, + "loss": 0.0957, + "step": 3400 + }, + { + "epoch": 0.9030801911842804, + "grad_norm": 0.21496419482161805, + "learning_rate": 6.20534867123293e-06, + "loss": 0.0354, + "step": 3401 + }, + { + "epoch": 0.9033457249070632, + "grad_norm": 0.22272639763431404, + "learning_rate": 6.203217919079343e-06, + "loss": 0.0391, + "step": 3402 + }, + { + "epoch": 0.903611258629846, + "grad_norm": 0.4393088137647846, + "learning_rate": 6.201086934959532e-06, + "loss": 0.0391, + "step": 3403 + }, + { + "epoch": 0.9038767923526287, + "grad_norm": 0.11332028570928786, + "learning_rate": 6.1989557192843255e-06, + "loss": 0.0051, + "step": 3404 + }, + { + "epoch": 0.9041423260754116, + "grad_norm": 0.47938443637823436, + "learning_rate": 6.196824272464596e-06, + "loss": 0.063, + "step": 3405 + }, + { + "epoch": 0.9044078597981944, + "grad_norm": 0.11296619118694176, + "learning_rate": 6.194692594911263e-06, + "loss": 0.0045, + "step": 3406 + }, + { + "epoch": 0.9046733935209772, + "grad_norm": 0.4263481380020443, + "learning_rate": 6.192560687035288e-06, + "loss": 0.0815, + "step": 3407 + }, + { + "epoch": 0.9049389272437599, + "grad_norm": 0.007533654311269684, + "learning_rate": 6.190428549247678e-06, + "loss": 0.0002, + "step": 3408 + }, + { + "epoch": 0.9052044609665427, + "grad_norm": 0.2020457747747357, + "learning_rate": 6.1882961819594835e-06, + "loss": 0.0354, + "step": 3409 + }, + { + "epoch": 0.9054699946893255, + "grad_norm": 0.45097415891788933, + "learning_rate": 6.186163585581798e-06, + "loss": 0.1113, + "step": 3410 + }, + { + "epoch": 0.9057355284121084, + "grad_norm": 0.49379403371802477, + "learning_rate": 6.184030760525763e-06, + "loss": 0.0688, + "step": 3411 + }, + { + "epoch": 0.9060010621348912, + "grad_norm": 0.39597671398866896, + "learning_rate": 6.181897707202562e-06, + "loss": 0.0525, + "step": 3412 + }, + { + "epoch": 0.9062665958576739, + "grad_norm": 0.4751839898537747, + "learning_rate": 6.179764426023422e-06, + "loss": 0.0576, + "step": 3413 + }, + { + "epoch": 0.9065321295804567, + "grad_norm": 0.21390728327396213, + "learning_rate": 6.177630917399616e-06, + "loss": 0.0391, + "step": 3414 + }, + { + "epoch": 0.9067976633032395, + "grad_norm": 0.24004498794777443, + "learning_rate": 6.175497181742458e-06, + "loss": 0.0092, + "step": 3415 + }, + { + "epoch": 0.9070631970260223, + "grad_norm": 0.7292959151407421, + "learning_rate": 6.173363219463306e-06, + "loss": 0.0815, + "step": 3416 + }, + { + "epoch": 0.907328730748805, + "grad_norm": 0.48825983332731254, + "learning_rate": 6.171229030973564e-06, + "loss": 0.0752, + "step": 3417 + }, + { + "epoch": 0.9075942644715879, + "grad_norm": 0.21571700311232203, + "learning_rate": 6.169094616684679e-06, + "loss": 0.0354, + "step": 3418 + }, + { + "epoch": 0.9078597981943707, + "grad_norm": 0.49126523803114835, + "learning_rate": 6.166959977008139e-06, + "loss": 0.0688, + "step": 3419 + }, + { + "epoch": 0.9081253319171535, + "grad_norm": 1.1561015216960921, + "learning_rate": 6.164825112355478e-06, + "loss": 0.0957, + "step": 3420 + }, + { + "epoch": 0.9083908656399363, + "grad_norm": 0.25363762377157295, + "learning_rate": 6.162690023138275e-06, + "loss": 0.0432, + "step": 3421 + }, + { + "epoch": 0.908656399362719, + "grad_norm": 0.356850972896345, + "learning_rate": 6.160554709768146e-06, + "loss": 0.0317, + "step": 3422 + }, + { + "epoch": 0.9089219330855018, + "grad_norm": 0.37363099910146885, + "learning_rate": 6.158419172656759e-06, + "loss": 0.0688, + "step": 3423 + }, + { + "epoch": 0.9091874668082847, + "grad_norm": 0.16266933592156554, + "learning_rate": 6.156283412215816e-06, + "loss": 0.0286, + "step": 3424 + }, + { + "epoch": 0.9094530005310675, + "grad_norm": 0.19784410791399792, + "learning_rate": 6.154147428857069e-06, + "loss": 0.0072, + "step": 3425 + }, + { + "epoch": 0.9097185342538502, + "grad_norm": 0.3760002541925043, + "learning_rate": 6.152011222992309e-06, + "loss": 0.0432, + "step": 3426 + }, + { + "epoch": 0.909984067976633, + "grad_norm": 0.46410133185238184, + "learning_rate": 6.149874795033373e-06, + "loss": 0.063, + "step": 3427 + }, + { + "epoch": 0.9102496016994158, + "grad_norm": 0.1950843496026646, + "learning_rate": 6.147738145392137e-06, + "loss": 0.0317, + "step": 3428 + }, + { + "epoch": 0.9105151354221986, + "grad_norm": 0.5621722776885398, + "learning_rate": 6.145601274480521e-06, + "loss": 0.063, + "step": 3429 + }, + { + "epoch": 0.9107806691449815, + "grad_norm": 0.43608891323173327, + "learning_rate": 6.143464182710493e-06, + "loss": 0.0576, + "step": 3430 + }, + { + "epoch": 0.9110462028677642, + "grad_norm": 0.006821202872897568, + "learning_rate": 6.141326870494057e-06, + "loss": 0.0002, + "step": 3431 + }, + { + "epoch": 0.911311736590547, + "grad_norm": 0.2046037410511334, + "learning_rate": 6.139189338243261e-06, + "loss": 0.0072, + "step": 3432 + }, + { + "epoch": 0.9115772703133298, + "grad_norm": 0.31823293188354573, + "learning_rate": 6.137051586370194e-06, + "loss": 0.0476, + "step": 3433 + }, + { + "epoch": 0.9118428040361126, + "grad_norm": 0.4590670570372276, + "learning_rate": 6.134913615286993e-06, + "loss": 0.0957, + "step": 3434 + }, + { + "epoch": 0.9121083377588953, + "grad_norm": 0.2381376951950056, + "learning_rate": 6.132775425405831e-06, + "loss": 0.0391, + "step": 3435 + }, + { + "epoch": 0.9123738714816781, + "grad_norm": 0.55527390286308, + "learning_rate": 6.130637017138929e-06, + "loss": 0.0815, + "step": 3436 + }, + { + "epoch": 0.912639405204461, + "grad_norm": 0.2149073993303838, + "learning_rate": 6.128498390898543e-06, + "loss": 0.0317, + "step": 3437 + }, + { + "epoch": 0.9129049389272438, + "grad_norm": 0.4308488704053223, + "learning_rate": 6.126359547096975e-06, + "loss": 0.0317, + "step": 3438 + }, + { + "epoch": 0.9131704726500266, + "grad_norm": 0.561469353875624, + "learning_rate": 6.124220486146573e-06, + "loss": 0.0688, + "step": 3439 + }, + { + "epoch": 0.9134360063728093, + "grad_norm": 0.4260453261642916, + "learning_rate": 6.12208120845972e-06, + "loss": 0.0957, + "step": 3440 + }, + { + "epoch": 0.9137015400955921, + "grad_norm": 0.23843229994576592, + "learning_rate": 6.1199417144488436e-06, + "loss": 0.0317, + "step": 3441 + }, + { + "epoch": 0.9139670738183749, + "grad_norm": 0.19044318379954026, + "learning_rate": 6.117802004526412e-06, + "loss": 0.0317, + "step": 3442 + }, + { + "epoch": 0.9142326075411578, + "grad_norm": 0.36986008885809224, + "learning_rate": 6.115662079104938e-06, + "loss": 0.0525, + "step": 3443 + }, + { + "epoch": 0.9144981412639405, + "grad_norm": 0.35750469788444195, + "learning_rate": 6.113521938596973e-06, + "loss": 0.0525, + "step": 3444 + }, + { + "epoch": 0.9147636749867233, + "grad_norm": 0.3451353151921243, + "learning_rate": 6.111381583415112e-06, + "loss": 0.0576, + "step": 3445 + }, + { + "epoch": 0.9150292087095061, + "grad_norm": 0.4644654142738302, + "learning_rate": 6.109241013971988e-06, + "loss": 0.0576, + "step": 3446 + }, + { + "epoch": 0.9152947424322889, + "grad_norm": 0.26067443296051357, + "learning_rate": 6.10710023068028e-06, + "loss": 0.0391, + "step": 3447 + }, + { + "epoch": 0.9155602761550717, + "grad_norm": 0.07855998523996609, + "learning_rate": 6.104959233952703e-06, + "loss": 0.0031, + "step": 3448 + }, + { + "epoch": 0.9158258098778544, + "grad_norm": 0.1147957054456189, + "learning_rate": 6.102818024202019e-06, + "loss": 0.005, + "step": 3449 + }, + { + "epoch": 0.9160913436006373, + "grad_norm": 0.5931501318425644, + "learning_rate": 6.100676601841026e-06, + "loss": 0.0576, + "step": 3450 + }, + { + "epoch": 0.9163568773234201, + "grad_norm": 0.4969632286347464, + "learning_rate": 6.098534967282563e-06, + "loss": 0.063, + "step": 3451 + }, + { + "epoch": 0.9166224110462029, + "grad_norm": 0.4839310622530124, + "learning_rate": 6.0963931209395165e-06, + "loss": 0.0957, + "step": 3452 + }, + { + "epoch": 0.9168879447689856, + "grad_norm": 0.36253760604400953, + "learning_rate": 6.094251063224808e-06, + "loss": 0.0476, + "step": 3453 + }, + { + "epoch": 0.9171534784917684, + "grad_norm": 0.11187822944877983, + "learning_rate": 6.092108794551399e-06, + "loss": 0.0031, + "step": 3454 + }, + { + "epoch": 0.9174190122145512, + "grad_norm": 0.8249711763784169, + "learning_rate": 6.0899663153322944e-06, + "loss": 0.0889, + "step": 3455 + }, + { + "epoch": 0.9176845459373341, + "grad_norm": 0.4890074994693259, + "learning_rate": 6.08782362598054e-06, + "loss": 0.0391, + "step": 3456 + }, + { + "epoch": 0.9179500796601169, + "grad_norm": 0.16154767305854273, + "learning_rate": 6.0856807269092185e-06, + "loss": 0.0229, + "step": 3457 + }, + { + "epoch": 0.9182156133828996, + "grad_norm": 0.45082218824739045, + "learning_rate": 6.083537618531459e-06, + "loss": 0.0889, + "step": 3458 + }, + { + "epoch": 0.9184811471056824, + "grad_norm": 0.25040609907539324, + "learning_rate": 6.081394301260426e-06, + "loss": 0.0354, + "step": 3459 + }, + { + "epoch": 0.9187466808284652, + "grad_norm": 0.1933647214630779, + "learning_rate": 6.079250775509324e-06, + "loss": 0.0317, + "step": 3460 + }, + { + "epoch": 0.919012214551248, + "grad_norm": 0.466442212769545, + "learning_rate": 6.077107041691401e-06, + "loss": 0.0476, + "step": 3461 + }, + { + "epoch": 0.9192777482740307, + "grad_norm": 0.5034702877622166, + "learning_rate": 6.074963100219945e-06, + "loss": 0.1191, + "step": 3462 + }, + { + "epoch": 0.9195432819968136, + "grad_norm": 0.5285573662531591, + "learning_rate": 6.07281895150828e-06, + "loss": 0.0576, + "step": 3463 + }, + { + "epoch": 0.9198088157195964, + "grad_norm": 0.0219764212317078, + "learning_rate": 6.070674595969775e-06, + "loss": 0.0008, + "step": 3464 + }, + { + "epoch": 0.9200743494423792, + "grad_norm": 0.48772242714130337, + "learning_rate": 6.068530034017836e-06, + "loss": 0.0815, + "step": 3465 + }, + { + "epoch": 0.920339883165162, + "grad_norm": 0.2804432477346251, + "learning_rate": 6.066385266065907e-06, + "loss": 0.0476, + "step": 3466 + }, + { + "epoch": 0.9206054168879447, + "grad_norm": 0.3383814900955449, + "learning_rate": 6.064240292527476e-06, + "loss": 0.0476, + "step": 3467 + }, + { + "epoch": 0.9208709506107275, + "grad_norm": 0.019507681326121844, + "learning_rate": 6.062095113816069e-06, + "loss": 0.0007, + "step": 3468 + }, + { + "epoch": 0.9211364843335104, + "grad_norm": 0.4591145779711353, + "learning_rate": 6.059949730345248e-06, + "loss": 0.0815, + "step": 3469 + }, + { + "epoch": 0.9214020180562932, + "grad_norm": 0.4820935364667389, + "learning_rate": 6.057804142528621e-06, + "loss": 0.063, + "step": 3470 + }, + { + "epoch": 0.9216675517790759, + "grad_norm": 0.46712423696252486, + "learning_rate": 6.055658350779831e-06, + "loss": 0.0576, + "step": 3471 + }, + { + "epoch": 0.9219330855018587, + "grad_norm": 0.2562905962605505, + "learning_rate": 6.053512355512564e-06, + "loss": 0.0131, + "step": 3472 + }, + { + "epoch": 0.9221986192246415, + "grad_norm": 0.442112227932835, + "learning_rate": 6.0513661571405365e-06, + "loss": 0.0957, + "step": 3473 + }, + { + "epoch": 0.9224641529474243, + "grad_norm": 0.19319483031622223, + "learning_rate": 6.049219756077514e-06, + "loss": 0.0317, + "step": 3474 + }, + { + "epoch": 0.9227296866702072, + "grad_norm": 0.3787818513529901, + "learning_rate": 6.047073152737298e-06, + "loss": 0.0525, + "step": 3475 + }, + { + "epoch": 0.9229952203929899, + "grad_norm": 0.005275964681136279, + "learning_rate": 6.0449263475337285e-06, + "loss": 0.0001, + "step": 3476 + }, + { + "epoch": 0.9232607541157727, + "grad_norm": 0.4563042010037955, + "learning_rate": 6.0427793408806805e-06, + "loss": 0.0476, + "step": 3477 + }, + { + "epoch": 0.9235262878385555, + "grad_norm": 0.20374423572428957, + "learning_rate": 6.040632133192074e-06, + "loss": 0.0317, + "step": 3478 + }, + { + "epoch": 0.9237918215613383, + "grad_norm": 0.5044308204761224, + "learning_rate": 6.038484724881867e-06, + "loss": 0.1113, + "step": 3479 + }, + { + "epoch": 0.924057355284121, + "grad_norm": 0.33828337955648, + "learning_rate": 6.036337116364053e-06, + "loss": 0.0354, + "step": 3480 + }, + { + "epoch": 0.9243228890069038, + "grad_norm": 0.04857502694856827, + "learning_rate": 6.034189308052665e-06, + "loss": 0.0019, + "step": 3481 + }, + { + "epoch": 0.9245884227296867, + "grad_norm": 0.2309070604476939, + "learning_rate": 6.032041300361773e-06, + "loss": 0.0354, + "step": 3482 + }, + { + "epoch": 0.9248539564524695, + "grad_norm": 0.4551674901838645, + "learning_rate": 6.029893093705492e-06, + "loss": 0.0889, + "step": 3483 + }, + { + "epoch": 0.9251194901752523, + "grad_norm": 0.2797822158634693, + "learning_rate": 6.027744688497969e-06, + "loss": 0.0432, + "step": 3484 + }, + { + "epoch": 0.925385023898035, + "grad_norm": 0.2010317418095399, + "learning_rate": 6.025596085153392e-06, + "loss": 0.0286, + "step": 3485 + }, + { + "epoch": 0.9256505576208178, + "grad_norm": 0.40052462885441625, + "learning_rate": 6.023447284085985e-06, + "loss": 0.0229, + "step": 3486 + }, + { + "epoch": 0.9259160913436006, + "grad_norm": 0.3613656358033218, + "learning_rate": 6.0212982857100125e-06, + "loss": 0.0476, + "step": 3487 + }, + { + "epoch": 0.9261816250663835, + "grad_norm": 0.5901824063553496, + "learning_rate": 6.019149090439774e-06, + "loss": 0.0957, + "step": 3488 + }, + { + "epoch": 0.9264471587891662, + "grad_norm": 0.4539682395291598, + "learning_rate": 6.0169996986896116e-06, + "loss": 0.0815, + "step": 3489 + }, + { + "epoch": 0.926712692511949, + "grad_norm": 2.357086606078313, + "learning_rate": 6.014850110873902e-06, + "loss": 0.2246, + "step": 3490 + }, + { + "epoch": 0.9269782262347318, + "grad_norm": 0.42715630611427663, + "learning_rate": 6.0127003274070595e-06, + "loss": 0.0258, + "step": 3491 + }, + { + "epoch": 0.9272437599575146, + "grad_norm": 1.8658909155706167, + "learning_rate": 6.0105503487035375e-06, + "loss": 0.2031, + "step": 3492 + }, + { + "epoch": 0.9275092936802974, + "grad_norm": 0.46148473309270044, + "learning_rate": 6.008400175177827e-06, + "loss": 0.0688, + "step": 3493 + }, + { + "epoch": 0.9277748274030801, + "grad_norm": 0.07172948487007977, + "learning_rate": 6.006249807244456e-06, + "loss": 0.0019, + "step": 3494 + }, + { + "epoch": 0.928040361125863, + "grad_norm": 0.25945424497919034, + "learning_rate": 6.004099245317989e-06, + "loss": 0.0391, + "step": 3495 + }, + { + "epoch": 0.9283058948486458, + "grad_norm": 0.23904981316033103, + "learning_rate": 6.001948489813031e-06, + "loss": 0.0286, + "step": 3496 + }, + { + "epoch": 0.9285714285714286, + "grad_norm": 0.276595088898044, + "learning_rate": 5.99979754114422e-06, + "loss": 0.0432, + "step": 3497 + }, + { + "epoch": 0.9288369622942113, + "grad_norm": 0.04597253362318379, + "learning_rate": 5.997646399726236e-06, + "loss": 0.0017, + "step": 3498 + }, + { + "epoch": 0.9291024960169941, + "grad_norm": 2.0640126495510827, + "learning_rate": 5.995495065973791e-06, + "loss": 0.1836, + "step": 3499 + }, + { + "epoch": 0.929368029739777, + "grad_norm": 0.9735775218694223, + "learning_rate": 5.993343540301637e-06, + "loss": 0.1113, + "step": 3500 + }, + { + "epoch": 0.9296335634625598, + "grad_norm": 0.24385798168987846, + "learning_rate": 5.991191823124565e-06, + "loss": 0.0354, + "step": 3501 + }, + { + "epoch": 0.9298990971853426, + "grad_norm": 0.18321368932190932, + "learning_rate": 5.989039914857399e-06, + "loss": 0.0286, + "step": 3502 + }, + { + "epoch": 0.9301646309081253, + "grad_norm": 0.013937979191544736, + "learning_rate": 5.986887815915004e-06, + "loss": 0.0005, + "step": 3503 + }, + { + "epoch": 0.9304301646309081, + "grad_norm": 0.019236270672119696, + "learning_rate": 5.984735526712275e-06, + "loss": 0.0006, + "step": 3504 + }, + { + "epoch": 0.9306956983536909, + "grad_norm": 0.22381461790167695, + "learning_rate": 5.982583047664151e-06, + "loss": 0.0286, + "step": 3505 + }, + { + "epoch": 0.9309612320764737, + "grad_norm": 0.00689914090906262, + "learning_rate": 5.980430379185604e-06, + "loss": 0.0002, + "step": 3506 + }, + { + "epoch": 0.9312267657992565, + "grad_norm": 0.4586392356058838, + "learning_rate": 5.978277521691643e-06, + "loss": 0.063, + "step": 3507 + }, + { + "epoch": 0.9314922995220393, + "grad_norm": 1.0100548034856884, + "learning_rate": 5.976124475597314e-06, + "loss": 0.063, + "step": 3508 + }, + { + "epoch": 0.9317578332448221, + "grad_norm": 0.44703247069965524, + "learning_rate": 5.9739712413176976e-06, + "loss": 0.0576, + "step": 3509 + }, + { + "epoch": 0.9320233669676049, + "grad_norm": 0.44921265922739456, + "learning_rate": 5.971817819267914e-06, + "loss": 0.0525, + "step": 3510 + }, + { + "epoch": 0.9322889006903877, + "grad_norm": 0.4154440147030062, + "learning_rate": 5.9696642098631154e-06, + "loss": 0.0391, + "step": 3511 + }, + { + "epoch": 0.9325544344131704, + "grad_norm": 0.3252796040576017, + "learning_rate": 5.967510413518495e-06, + "loss": 0.0391, + "step": 3512 + }, + { + "epoch": 0.9328199681359532, + "grad_norm": 0.5752651661234184, + "learning_rate": 5.965356430649276e-06, + "loss": 0.0476, + "step": 3513 + }, + { + "epoch": 0.9330855018587361, + "grad_norm": 0.3258264331741429, + "learning_rate": 5.963202261670722e-06, + "loss": 0.0206, + "step": 3514 + }, + { + "epoch": 0.9333510355815189, + "grad_norm": 0.2317167433851736, + "learning_rate": 5.961047906998132e-06, + "loss": 0.0286, + "step": 3515 + }, + { + "epoch": 0.9336165693043016, + "grad_norm": 0.14563140184977325, + "learning_rate": 5.958893367046841e-06, + "loss": 0.0205, + "step": 3516 + }, + { + "epoch": 0.9338821030270844, + "grad_norm": 0.5819670821811292, + "learning_rate": 5.956738642232217e-06, + "loss": 0.1455, + "step": 3517 + }, + { + "epoch": 0.9341476367498672, + "grad_norm": 0.30920435352520664, + "learning_rate": 5.954583732969667e-06, + "loss": 0.0317, + "step": 3518 + }, + { + "epoch": 0.93441317047265, + "grad_norm": 0.21700788095165485, + "learning_rate": 5.952428639674632e-06, + "loss": 0.0286, + "step": 3519 + }, + { + "epoch": 0.9346787041954329, + "grad_norm": 0.6947071231695563, + "learning_rate": 5.950273362762587e-06, + "loss": 0.1113, + "step": 3520 + }, + { + "epoch": 0.9349442379182156, + "grad_norm": 0.6922343188438913, + "learning_rate": 5.948117902649046e-06, + "loss": 0.1035, + "step": 3521 + }, + { + "epoch": 0.9352097716409984, + "grad_norm": 0.5324703085144005, + "learning_rate": 5.945962259749556e-06, + "loss": 0.1279, + "step": 3522 + }, + { + "epoch": 0.9354753053637812, + "grad_norm": 0.14959138576085362, + "learning_rate": 5.943806434479697e-06, + "loss": 0.0045, + "step": 3523 + }, + { + "epoch": 0.935740839086564, + "grad_norm": 0.16481096861122918, + "learning_rate": 5.941650427255089e-06, + "loss": 0.0045, + "step": 3524 + }, + { + "epoch": 0.9360063728093467, + "grad_norm": 0.6054384525173877, + "learning_rate": 5.939494238491386e-06, + "loss": 0.1367, + "step": 3525 + }, + { + "epoch": 0.9362719065321295, + "grad_norm": 0.13793573184301736, + "learning_rate": 5.937337868604273e-06, + "loss": 0.0206, + "step": 3526 + }, + { + "epoch": 0.9365374402549124, + "grad_norm": 1.4083770185300826, + "learning_rate": 5.935181318009473e-06, + "loss": 0.0815, + "step": 3527 + }, + { + "epoch": 0.9368029739776952, + "grad_norm": 0.1417793256641535, + "learning_rate": 5.933024587122745e-06, + "loss": 0.0229, + "step": 3528 + }, + { + "epoch": 0.937068507700478, + "grad_norm": 0.928586174534263, + "learning_rate": 5.930867676359882e-06, + "loss": 0.0889, + "step": 3529 + }, + { + "epoch": 0.9373340414232607, + "grad_norm": 0.2611388052433496, + "learning_rate": 5.9287105861367075e-06, + "loss": 0.0391, + "step": 3530 + }, + { + "epoch": 0.9375995751460435, + "grad_norm": 1.0919741803565677, + "learning_rate": 5.926553316869087e-06, + "loss": 0.0815, + "step": 3531 + }, + { + "epoch": 0.9378651088688263, + "grad_norm": 0.19410917515339404, + "learning_rate": 5.924395868972912e-06, + "loss": 0.0354, + "step": 3532 + }, + { + "epoch": 0.9381306425916092, + "grad_norm": 0.13142548438469778, + "learning_rate": 5.922238242864118e-06, + "loss": 0.0206, + "step": 3533 + }, + { + "epoch": 0.938396176314392, + "grad_norm": 0.2744695725921158, + "learning_rate": 5.9200804389586675e-06, + "loss": 0.0391, + "step": 3534 + }, + { + "epoch": 0.9386617100371747, + "grad_norm": 0.6078406564154273, + "learning_rate": 5.917922457672559e-06, + "loss": 0.1191, + "step": 3535 + }, + { + "epoch": 0.9389272437599575, + "grad_norm": 0.21355653127736224, + "learning_rate": 5.915764299421827e-06, + "loss": 0.0092, + "step": 3536 + }, + { + "epoch": 0.9391927774827403, + "grad_norm": 0.16077964722784296, + "learning_rate": 5.9136059646225375e-06, + "loss": 0.0229, + "step": 3537 + }, + { + "epoch": 0.9394583112055231, + "grad_norm": 0.14275019363203106, + "learning_rate": 5.911447453690794e-06, + "loss": 0.0229, + "step": 3538 + }, + { + "epoch": 0.9397238449283059, + "grad_norm": 0.7240245574113348, + "learning_rate": 5.909288767042731e-06, + "loss": 0.0258, + "step": 3539 + }, + { + "epoch": 0.9399893786510887, + "grad_norm": 0.5201537897643087, + "learning_rate": 5.907129905094517e-06, + "loss": 0.0752, + "step": 3540 + }, + { + "epoch": 0.9402549123738715, + "grad_norm": 0.025259191688958304, + "learning_rate": 5.904970868262357e-06, + "loss": 0.0007, + "step": 3541 + }, + { + "epoch": 0.9405204460966543, + "grad_norm": 0.2041709912826682, + "learning_rate": 5.902811656962485e-06, + "loss": 0.0286, + "step": 3542 + }, + { + "epoch": 0.9407859798194371, + "grad_norm": 0.2247677451795941, + "learning_rate": 5.900652271611174e-06, + "loss": 0.0391, + "step": 3543 + }, + { + "epoch": 0.9410515135422198, + "grad_norm": 0.12953149555265855, + "learning_rate": 5.898492712624729e-06, + "loss": 0.0184, + "step": 3544 + }, + { + "epoch": 0.9413170472650026, + "grad_norm": 0.35807203441784785, + "learning_rate": 5.896332980419483e-06, + "loss": 0.0117, + "step": 3545 + }, + { + "epoch": 0.9415825809877855, + "grad_norm": 0.19027996570238656, + "learning_rate": 5.894173075411812e-06, + "loss": 0.0286, + "step": 3546 + }, + { + "epoch": 0.9418481147105683, + "grad_norm": 0.4355621748884647, + "learning_rate": 5.892012998018116e-06, + "loss": 0.1035, + "step": 3547 + }, + { + "epoch": 0.942113648433351, + "grad_norm": 0.008866818762037574, + "learning_rate": 5.889852748654836e-06, + "loss": 0.0002, + "step": 3548 + }, + { + "epoch": 0.9423791821561338, + "grad_norm": 1.8642112040242895, + "learning_rate": 5.88769232773844e-06, + "loss": 0.0391, + "step": 3549 + }, + { + "epoch": 0.9426447158789166, + "grad_norm": 1.2489167612290633, + "learning_rate": 5.885531735685431e-06, + "loss": 0.1279, + "step": 3550 + }, + { + "epoch": 0.9429102496016994, + "grad_norm": 0.9698939416309471, + "learning_rate": 5.883370972912349e-06, + "loss": 0.0889, + "step": 3551 + }, + { + "epoch": 0.9431757833244823, + "grad_norm": 0.47455425835582066, + "learning_rate": 5.881210039835763e-06, + "loss": 0.0957, + "step": 3552 + }, + { + "epoch": 0.943441317047265, + "grad_norm": 0.044285590409362655, + "learning_rate": 5.879048936872273e-06, + "loss": 0.0012, + "step": 3553 + }, + { + "epoch": 0.9437068507700478, + "grad_norm": 0.15830404600812403, + "learning_rate": 5.876887664438515e-06, + "loss": 0.0255, + "step": 3554 + }, + { + "epoch": 0.9439723844928306, + "grad_norm": 0.5053049959368788, + "learning_rate": 5.874726222951157e-06, + "loss": 0.1279, + "step": 3555 + }, + { + "epoch": 0.9442379182156134, + "grad_norm": 0.4513891523061316, + "learning_rate": 5.8725646128269e-06, + "loss": 0.0957, + "step": 3556 + }, + { + "epoch": 0.9445034519383961, + "grad_norm": 0.40506247111380944, + "learning_rate": 5.870402834482477e-06, + "loss": 0.0815, + "step": 3557 + }, + { + "epoch": 0.944768985661179, + "grad_norm": 0.5456890428976771, + "learning_rate": 5.8682408883346535e-06, + "loss": 0.0815, + "step": 3558 + }, + { + "epoch": 0.9450345193839618, + "grad_norm": 0.12150605952988766, + "learning_rate": 5.866078774800225e-06, + "loss": 0.0165, + "step": 3559 + }, + { + "epoch": 0.9453000531067446, + "grad_norm": 0.514951297623655, + "learning_rate": 5.863916494296025e-06, + "loss": 0.0889, + "step": 3560 + }, + { + "epoch": 0.9455655868295274, + "grad_norm": 0.1835924675611359, + "learning_rate": 5.861754047238913e-06, + "loss": 0.0286, + "step": 3561 + }, + { + "epoch": 0.9458311205523101, + "grad_norm": 0.010257804788460204, + "learning_rate": 5.859591434045784e-06, + "loss": 0.0003, + "step": 3562 + }, + { + "epoch": 0.9460966542750929, + "grad_norm": 0.1422767767926214, + "learning_rate": 5.857428655133562e-06, + "loss": 0.0229, + "step": 3563 + }, + { + "epoch": 0.9463621879978757, + "grad_norm": 0.162873433925966, + "learning_rate": 5.855265710919211e-06, + "loss": 0.0255, + "step": 3564 + }, + { + "epoch": 0.9466277217206586, + "grad_norm": 0.5040944975118017, + "learning_rate": 5.8531026018197165e-06, + "loss": 0.0206, + "step": 3565 + }, + { + "epoch": 0.9468932554434413, + "grad_norm": 0.24624202867409065, + "learning_rate": 5.850939328252103e-06, + "loss": 0.0391, + "step": 3566 + }, + { + "epoch": 0.9471587891662241, + "grad_norm": 0.4251585052239867, + "learning_rate": 5.8487758906334225e-06, + "loss": 0.0258, + "step": 3567 + }, + { + "epoch": 0.9474243228890069, + "grad_norm": 0.7390472395581791, + "learning_rate": 5.84661228938076e-06, + "loss": 0.0576, + "step": 3568 + }, + { + "epoch": 0.9476898566117897, + "grad_norm": 0.1895817603050598, + "learning_rate": 5.8444485249112335e-06, + "loss": 0.0354, + "step": 3569 + }, + { + "epoch": 0.9479553903345725, + "grad_norm": 0.1471316336055243, + "learning_rate": 5.842284597641991e-06, + "loss": 0.0031, + "step": 3570 + }, + { + "epoch": 0.9482209240573553, + "grad_norm": 0.27931225161895057, + "learning_rate": 5.840120507990211e-06, + "loss": 0.0476, + "step": 3571 + }, + { + "epoch": 0.9484864577801381, + "grad_norm": 0.16562615867958405, + "learning_rate": 5.837956256373105e-06, + "loss": 0.0255, + "step": 3572 + }, + { + "epoch": 0.9487519915029209, + "grad_norm": 0.22406069525703046, + "learning_rate": 5.835791843207916e-06, + "loss": 0.0391, + "step": 3573 + }, + { + "epoch": 0.9490175252257037, + "grad_norm": 0.5268965555087725, + "learning_rate": 5.833627268911919e-06, + "loss": 0.0752, + "step": 3574 + }, + { + "epoch": 0.9492830589484864, + "grad_norm": 0.05838339258423595, + "learning_rate": 5.831462533902415e-06, + "loss": 0.0014, + "step": 3575 + }, + { + "epoch": 0.9495485926712692, + "grad_norm": 0.23153533569135928, + "learning_rate": 5.829297638596739e-06, + "loss": 0.0092, + "step": 3576 + }, + { + "epoch": 0.949814126394052, + "grad_norm": 0.4917740973504414, + "learning_rate": 5.827132583412261e-06, + "loss": 0.0889, + "step": 3577 + }, + { + "epoch": 0.9500796601168349, + "grad_norm": 0.16883676192265118, + "learning_rate": 5.824967368766375e-06, + "loss": 0.0255, + "step": 3578 + }, + { + "epoch": 0.9503451938396177, + "grad_norm": 0.13051429763527322, + "learning_rate": 5.82280199507651e-06, + "loss": 0.0184, + "step": 3579 + }, + { + "epoch": 0.9506107275624004, + "grad_norm": 0.576758993411103, + "learning_rate": 5.820636462760125e-06, + "loss": 0.1191, + "step": 3580 + }, + { + "epoch": 0.9508762612851832, + "grad_norm": 0.4386795661869916, + "learning_rate": 5.818470772234707e-06, + "loss": 0.0815, + "step": 3581 + }, + { + "epoch": 0.951141795007966, + "grad_norm": 0.5021461869479189, + "learning_rate": 5.816304923917778e-06, + "loss": 0.1191, + "step": 3582 + }, + { + "epoch": 0.9514073287307488, + "grad_norm": 0.5333784005873679, + "learning_rate": 5.814138918226887e-06, + "loss": 0.0576, + "step": 3583 + }, + { + "epoch": 0.9516728624535316, + "grad_norm": 0.28190547430108814, + "learning_rate": 5.811972755579616e-06, + "loss": 0.0476, + "step": 3584 + }, + { + "epoch": 0.9519383961763144, + "grad_norm": 0.538884954185595, + "learning_rate": 5.809806436393572e-06, + "loss": 0.0576, + "step": 3585 + }, + { + "epoch": 0.9522039298990972, + "grad_norm": 0.36106918335823984, + "learning_rate": 5.807639961086398e-06, + "loss": 0.0131, + "step": 3586 + }, + { + "epoch": 0.95246946362188, + "grad_norm": 0.21198045513886093, + "learning_rate": 5.805473330075766e-06, + "loss": 0.0354, + "step": 3587 + }, + { + "epoch": 0.9527349973446628, + "grad_norm": 0.5218348791471842, + "learning_rate": 5.803306543779374e-06, + "loss": 0.0889, + "step": 3588 + }, + { + "epoch": 0.9530005310674455, + "grad_norm": 0.4362245892536183, + "learning_rate": 5.801139602614956e-06, + "loss": 0.0576, + "step": 3589 + }, + { + "epoch": 0.9532660647902284, + "grad_norm": 0.36604821714285785, + "learning_rate": 5.798972507000269e-06, + "loss": 0.0391, + "step": 3590 + }, + { + "epoch": 0.9535315985130112, + "grad_norm": 0.027261074520874555, + "learning_rate": 5.796805257353109e-06, + "loss": 0.0009, + "step": 3591 + }, + { + "epoch": 0.953797132235794, + "grad_norm": 0.0037963858548896146, + "learning_rate": 5.7946378540912905e-06, + "loss": 0.0001, + "step": 3592 + }, + { + "epoch": 0.9540626659585767, + "grad_norm": 0.3869588641166267, + "learning_rate": 5.792470297632666e-06, + "loss": 0.0525, + "step": 3593 + }, + { + "epoch": 0.9543281996813595, + "grad_norm": 0.4724955887097785, + "learning_rate": 5.7903025883951145e-06, + "loss": 0.0103, + "step": 3594 + }, + { + "epoch": 0.9545937334041423, + "grad_norm": 1.5136860119713154, + "learning_rate": 5.788134726796545e-06, + "loss": 0.0688, + "step": 3595 + }, + { + "epoch": 0.9548592671269251, + "grad_norm": 0.4277340308502645, + "learning_rate": 5.785966713254894e-06, + "loss": 0.0889, + "step": 3596 + }, + { + "epoch": 0.955124800849708, + "grad_norm": 0.171879159388591, + "learning_rate": 5.783798548188132e-06, + "loss": 0.0229, + "step": 3597 + }, + { + "epoch": 0.9553903345724907, + "grad_norm": 0.3310202717770771, + "learning_rate": 5.781630232014253e-06, + "loss": 0.0525, + "step": 3598 + }, + { + "epoch": 0.9556558682952735, + "grad_norm": 0.3361714224055187, + "learning_rate": 5.779461765151286e-06, + "loss": 0.0525, + "step": 3599 + }, + { + "epoch": 0.9559214020180563, + "grad_norm": 0.24963517743183233, + "learning_rate": 5.77729314801728e-06, + "loss": 0.0352, + "step": 3600 + }, + { + "epoch": 0.9561869357408391, + "grad_norm": 0.5560287339220119, + "learning_rate": 5.775124381030325e-06, + "loss": 0.0688, + "step": 3601 + }, + { + "epoch": 0.9564524694636218, + "grad_norm": 0.004199994689870284, + "learning_rate": 5.7729554646085295e-06, + "loss": 0.0001, + "step": 3602 + }, + { + "epoch": 0.9567180031864047, + "grad_norm": 0.02023488802183647, + "learning_rate": 5.770786399170038e-06, + "loss": 0.0007, + "step": 3603 + }, + { + "epoch": 0.9569835369091875, + "grad_norm": 0.6230084432351841, + "learning_rate": 5.768617185133017e-06, + "loss": 0.0815, + "step": 3604 + }, + { + "epoch": 0.9572490706319703, + "grad_norm": 0.21352910077171497, + "learning_rate": 5.766447822915669e-06, + "loss": 0.0317, + "step": 3605 + }, + { + "epoch": 0.9575146043547531, + "grad_norm": 0.5353366116538953, + "learning_rate": 5.764278312936221e-06, + "loss": 0.1113, + "step": 3606 + }, + { + "epoch": 0.9577801380775358, + "grad_norm": 0.45098903208658, + "learning_rate": 5.762108655612928e-06, + "loss": 0.0889, + "step": 3607 + }, + { + "epoch": 0.9580456718003186, + "grad_norm": 0.5822026001858617, + "learning_rate": 5.759938851364074e-06, + "loss": 0.0286, + "step": 3608 + }, + { + "epoch": 0.9583112055231015, + "grad_norm": 0.22060586326838053, + "learning_rate": 5.757768900607972e-06, + "loss": 0.0317, + "step": 3609 + }, + { + "epoch": 0.9585767392458843, + "grad_norm": 0.3166631588489053, + "learning_rate": 5.755598803762963e-06, + "loss": 0.0476, + "step": 3610 + }, + { + "epoch": 0.958842272968667, + "grad_norm": 0.004068249886425086, + "learning_rate": 5.753428561247416e-06, + "loss": 0.0001, + "step": 3611 + }, + { + "epoch": 0.9591078066914498, + "grad_norm": 0.3948575674743605, + "learning_rate": 5.751258173479729e-06, + "loss": 0.0752, + "step": 3612 + }, + { + "epoch": 0.9593733404142326, + "grad_norm": 0.9276325524508592, + "learning_rate": 5.749087640878325e-06, + "loss": 0.0957, + "step": 3613 + }, + { + "epoch": 0.9596388741370154, + "grad_norm": 0.4230163053746998, + "learning_rate": 5.746916963861659e-06, + "loss": 0.0525, + "step": 3614 + }, + { + "epoch": 0.9599044078597982, + "grad_norm": 0.6406946242250046, + "learning_rate": 5.744746142848212e-06, + "loss": 0.0525, + "step": 3615 + }, + { + "epoch": 0.960169941582581, + "grad_norm": 0.4391160992861725, + "learning_rate": 5.742575178256493e-06, + "loss": 0.0525, + "step": 3616 + }, + { + "epoch": 0.9604354753053638, + "grad_norm": 0.48853965191857235, + "learning_rate": 5.740404070505036e-06, + "loss": 0.0476, + "step": 3617 + }, + { + "epoch": 0.9607010090281466, + "grad_norm": 0.4207783920359763, + "learning_rate": 5.738232820012407e-06, + "loss": 0.063, + "step": 3618 + }, + { + "epoch": 0.9609665427509294, + "grad_norm": 0.5612163493574303, + "learning_rate": 5.736061427197197e-06, + "loss": 0.0255, + "step": 3619 + }, + { + "epoch": 0.9612320764737121, + "grad_norm": 1.2298144402830893, + "learning_rate": 5.733889892478025e-06, + "loss": 0.1279, + "step": 3620 + }, + { + "epoch": 0.9614976101964949, + "grad_norm": 0.31935587519417447, + "learning_rate": 5.731718216273536e-06, + "loss": 0.0432, + "step": 3621 + }, + { + "epoch": 0.9617631439192778, + "grad_norm": 0.42591651031960104, + "learning_rate": 5.729546399002405e-06, + "loss": 0.0752, + "step": 3622 + }, + { + "epoch": 0.9620286776420606, + "grad_norm": 0.39920955192212426, + "learning_rate": 5.727374441083333e-06, + "loss": 0.0354, + "step": 3623 + }, + { + "epoch": 0.9622942113648434, + "grad_norm": 0.5082968855225545, + "learning_rate": 5.725202342935049e-06, + "loss": 0.0317, + "step": 3624 + }, + { + "epoch": 0.9625597450876261, + "grad_norm": 0.24127096306209905, + "learning_rate": 5.723030104976305e-06, + "loss": 0.0391, + "step": 3625 + }, + { + "epoch": 0.9628252788104089, + "grad_norm": 0.04095118919276792, + "learning_rate": 5.720857727625882e-06, + "loss": 0.0013, + "step": 3626 + }, + { + "epoch": 0.9630908125331917, + "grad_norm": 0.27323213753795456, + "learning_rate": 5.718685211302592e-06, + "loss": 0.0432, + "step": 3627 + }, + { + "epoch": 0.9633563462559745, + "grad_norm": 0.23088982699954425, + "learning_rate": 5.716512556425271e-06, + "loss": 0.0354, + "step": 3628 + }, + { + "epoch": 0.9636218799787573, + "grad_norm": 0.21461181825975714, + "learning_rate": 5.714339763412779e-06, + "loss": 0.0317, + "step": 3629 + }, + { + "epoch": 0.9638874137015401, + "grad_norm": 0.07173080521512337, + "learning_rate": 5.7121668326840055e-06, + "loss": 0.0024, + "step": 3630 + }, + { + "epoch": 0.9641529474243229, + "grad_norm": 0.8639282466695275, + "learning_rate": 5.709993764657865e-06, + "loss": 0.0576, + "step": 3631 + }, + { + "epoch": 0.9644184811471057, + "grad_norm": 0.003688935393390043, + "learning_rate": 5.707820559753301e-06, + "loss": 0.0001, + "step": 3632 + }, + { + "epoch": 0.9646840148698885, + "grad_norm": 0.3617595075336269, + "learning_rate": 5.705647218389281e-06, + "loss": 0.0476, + "step": 3633 + }, + { + "epoch": 0.9649495485926712, + "grad_norm": 0.23347403843515865, + "learning_rate": 5.703473740984799e-06, + "loss": 0.0354, + "step": 3634 + }, + { + "epoch": 0.965215082315454, + "grad_norm": 0.7297844776768874, + "learning_rate": 5.701300127958875e-06, + "loss": 0.063, + "step": 3635 + }, + { + "epoch": 0.9654806160382369, + "grad_norm": 0.368753673584919, + "learning_rate": 5.6991263797305594e-06, + "loss": 0.0391, + "step": 3636 + }, + { + "epoch": 0.9657461497610197, + "grad_norm": 0.3712998363774496, + "learning_rate": 5.696952496718924e-06, + "loss": 0.0576, + "step": 3637 + }, + { + "epoch": 0.9660116834838024, + "grad_norm": 0.21447394214324036, + "learning_rate": 5.694778479343066e-06, + "loss": 0.0255, + "step": 3638 + }, + { + "epoch": 0.9662772172065852, + "grad_norm": 0.9400012945676488, + "learning_rate": 5.692604328022111e-06, + "loss": 0.0688, + "step": 3639 + }, + { + "epoch": 0.966542750929368, + "grad_norm": 0.14527846265686906, + "learning_rate": 5.690430043175211e-06, + "loss": 0.0184, + "step": 3640 + }, + { + "epoch": 0.9668082846521509, + "grad_norm": 0.26411548778215144, + "learning_rate": 5.688255625221541e-06, + "loss": 0.0255, + "step": 3641 + }, + { + "epoch": 0.9670738183749337, + "grad_norm": 0.17055570150378818, + "learning_rate": 5.686081074580304e-06, + "loss": 0.0229, + "step": 3642 + }, + { + "epoch": 0.9673393520977164, + "grad_norm": 0.0038505762641045708, + "learning_rate": 5.683906391670727e-06, + "loss": 0.0001, + "step": 3643 + }, + { + "epoch": 0.9676048858204992, + "grad_norm": 0.16245707063745415, + "learning_rate": 5.6817315769120626e-06, + "loss": 0.0205, + "step": 3644 + }, + { + "epoch": 0.967870419543282, + "grad_norm": 0.15383274387272294, + "learning_rate": 5.679556630723592e-06, + "loss": 0.0205, + "step": 3645 + }, + { + "epoch": 0.9681359532660648, + "grad_norm": 0.0006055403396439249, + "learning_rate": 5.677381553524617e-06, + "loss": 0.0, + "step": 3646 + }, + { + "epoch": 0.9684014869888475, + "grad_norm": 0.7926106392845446, + "learning_rate": 5.675206345734467e-06, + "loss": 0.0432, + "step": 3647 + }, + { + "epoch": 0.9686670207116304, + "grad_norm": 0.45646149614315307, + "learning_rate": 5.673031007772498e-06, + "loss": 0.0688, + "step": 3648 + }, + { + "epoch": 0.9689325544344132, + "grad_norm": 0.11527800357734554, + "learning_rate": 5.670855540058088e-06, + "loss": 0.0129, + "step": 3649 + }, + { + "epoch": 0.969198088157196, + "grad_norm": 0.23289862588103916, + "learning_rate": 5.668679943010642e-06, + "loss": 0.0286, + "step": 3650 + }, + { + "epoch": 0.9694636218799788, + "grad_norm": 0.703526070290471, + "learning_rate": 5.666504217049589e-06, + "loss": 0.0752, + "step": 3651 + }, + { + "epoch": 0.9697291556027615, + "grad_norm": 0.3614707504476666, + "learning_rate": 5.664328362594384e-06, + "loss": 0.0286, + "step": 3652 + }, + { + "epoch": 0.9699946893255443, + "grad_norm": 0.24148871398944463, + "learning_rate": 5.662152380064504e-06, + "loss": 0.0229, + "step": 3653 + }, + { + "epoch": 0.9702602230483272, + "grad_norm": 0.22055841337448734, + "learning_rate": 5.659976269879456e-06, + "loss": 0.0205, + "step": 3654 + }, + { + "epoch": 0.97052575677111, + "grad_norm": 0.9730664986733953, + "learning_rate": 5.657800032458767e-06, + "loss": 0.1553, + "step": 3655 + }, + { + "epoch": 0.9707912904938927, + "grad_norm": 0.363672745754524, + "learning_rate": 5.655623668221991e-06, + "loss": 0.0286, + "step": 3656 + }, + { + "epoch": 0.9710568242166755, + "grad_norm": 0.11292775787654695, + "learning_rate": 5.653447177588703e-06, + "loss": 0.0115, + "step": 3657 + }, + { + "epoch": 0.9713223579394583, + "grad_norm": 0.052051068270945015, + "learning_rate": 5.651270560978504e-06, + "loss": 0.0039, + "step": 3658 + }, + { + "epoch": 0.9715878916622411, + "grad_norm": 0.10293656918945553, + "learning_rate": 5.649093818811023e-06, + "loss": 0.0103, + "step": 3659 + }, + { + "epoch": 0.971853425385024, + "grad_norm": 0.3051005343451974, + "learning_rate": 5.646916951505909e-06, + "loss": 0.0117, + "step": 3660 + }, + { + "epoch": 0.9721189591078067, + "grad_norm": 0.14883940302865978, + "learning_rate": 5.644739959482838e-06, + "loss": 0.0146, + "step": 3661 + }, + { + "epoch": 0.9723844928305895, + "grad_norm": 1.4325430744676855, + "learning_rate": 5.6425628431615054e-06, + "loss": 0.063, + "step": 3662 + }, + { + "epoch": 0.9726500265533723, + "grad_norm": 0.20211867853491797, + "learning_rate": 5.640385602961634e-06, + "loss": 0.0184, + "step": 3663 + }, + { + "epoch": 0.9729155602761551, + "grad_norm": 0.17363920554460457, + "learning_rate": 5.638208239302975e-06, + "loss": 0.0129, + "step": 3664 + }, + { + "epoch": 0.9731810939989378, + "grad_norm": 4.148392375976074, + "learning_rate": 5.636030752605292e-06, + "loss": 0.3105, + "step": 3665 + }, + { + "epoch": 0.9734466277217206, + "grad_norm": 0.146142375132244, + "learning_rate": 5.6338531432883795e-06, + "loss": 0.0115, + "step": 3666 + }, + { + "epoch": 0.9737121614445035, + "grad_norm": 0.9306454565974007, + "learning_rate": 5.6316754117720575e-06, + "loss": 0.1836, + "step": 3667 + }, + { + "epoch": 0.9739776951672863, + "grad_norm": 0.9912134285992179, + "learning_rate": 5.629497558476167e-06, + "loss": 0.1934, + "step": 3668 + }, + { + "epoch": 0.9742432288900691, + "grad_norm": 0.0444477676145146, + "learning_rate": 5.627319583820571e-06, + "loss": 0.0017, + "step": 3669 + }, + { + "epoch": 0.9745087626128518, + "grad_norm": 0.11208163440166442, + "learning_rate": 5.625141488225158e-06, + "loss": 0.0031, + "step": 3670 + }, + { + "epoch": 0.9747742963356346, + "grad_norm": 1.4555592616728488, + "learning_rate": 5.622963272109838e-06, + "loss": 0.0815, + "step": 3671 + }, + { + "epoch": 0.9750398300584174, + "grad_norm": 0.12189606583314123, + "learning_rate": 5.620784935894548e-06, + "loss": 0.0115, + "step": 3672 + }, + { + "epoch": 0.9753053637812003, + "grad_norm": 0.28752550212186684, + "learning_rate": 5.618606479999242e-06, + "loss": 0.0255, + "step": 3673 + }, + { + "epoch": 0.975570897503983, + "grad_norm": 0.8669978915657431, + "learning_rate": 5.616427904843904e-06, + "loss": 0.1641, + "step": 3674 + }, + { + "epoch": 0.9758364312267658, + "grad_norm": 0.14062140218831734, + "learning_rate": 5.614249210848534e-06, + "loss": 0.0115, + "step": 3675 + }, + { + "epoch": 0.9761019649495486, + "grad_norm": 0.003087782467878347, + "learning_rate": 5.612070398433161e-06, + "loss": 0.0, + "step": 3676 + }, + { + "epoch": 0.9763674986723314, + "grad_norm": 0.5840021914000452, + "learning_rate": 5.609891468017835e-06, + "loss": 0.0432, + "step": 3677 + }, + { + "epoch": 0.9766330323951142, + "grad_norm": 2.771082293195368, + "learning_rate": 5.607712420022627e-06, + "loss": 0.2344, + "step": 3678 + }, + { + "epoch": 0.9768985661178969, + "grad_norm": 1.4064901797093847, + "learning_rate": 5.6055332548676325e-06, + "loss": 0.0752, + "step": 3679 + }, + { + "epoch": 0.9771640998406798, + "grad_norm": 0.609869908158293, + "learning_rate": 5.603353972972967e-06, + "loss": 0.1035, + "step": 3680 + }, + { + "epoch": 0.9774296335634626, + "grad_norm": 0.341735416510148, + "learning_rate": 5.6011745747587706e-06, + "loss": 0.0354, + "step": 3681 + }, + { + "epoch": 0.9776951672862454, + "grad_norm": 0.22433774353910363, + "learning_rate": 5.598995060645207e-06, + "loss": 0.0286, + "step": 3682 + }, + { + "epoch": 0.9779607010090281, + "grad_norm": 0.3474295002238389, + "learning_rate": 5.596815431052461e-06, + "loss": 0.0391, + "step": 3683 + }, + { + "epoch": 0.9782262347318109, + "grad_norm": 0.747625362893198, + "learning_rate": 5.594635686400737e-06, + "loss": 0.1367, + "step": 3684 + }, + { + "epoch": 0.9784917684545937, + "grad_norm": 0.005384135017925143, + "learning_rate": 5.592455827110267e-06, + "loss": 0.0002, + "step": 3685 + }, + { + "epoch": 0.9787573021773766, + "grad_norm": 0.048612725747091766, + "learning_rate": 5.590275853601301e-06, + "loss": 0.001, + "step": 3686 + }, + { + "epoch": 0.9790228359001594, + "grad_norm": 0.14890273319028954, + "learning_rate": 5.588095766294112e-06, + "loss": 0.0184, + "step": 3687 + }, + { + "epoch": 0.9792883696229421, + "grad_norm": 0.8976296012789594, + "learning_rate": 5.5859155656089944e-06, + "loss": 0.0576, + "step": 3688 + }, + { + "epoch": 0.9795539033457249, + "grad_norm": 0.36093616267675654, + "learning_rate": 5.583735251966267e-06, + "loss": 0.0317, + "step": 3689 + }, + { + "epoch": 0.9798194370685077, + "grad_norm": 0.3077169832559515, + "learning_rate": 5.581554825786268e-06, + "loss": 0.0432, + "step": 3690 + }, + { + "epoch": 0.9800849707912905, + "grad_norm": 0.45804001785307535, + "learning_rate": 5.579374287489355e-06, + "loss": 0.0432, + "step": 3691 + }, + { + "epoch": 0.9803505045140732, + "grad_norm": 1.0084882607909862, + "learning_rate": 5.577193637495914e-06, + "loss": 0.0688, + "step": 3692 + }, + { + "epoch": 0.9806160382368561, + "grad_norm": 0.26079360077065494, + "learning_rate": 5.575012876226346e-06, + "loss": 0.0317, + "step": 3693 + }, + { + "epoch": 0.9808815719596389, + "grad_norm": 0.23944228690217625, + "learning_rate": 5.572832004101079e-06, + "loss": 0.0317, + "step": 3694 + }, + { + "epoch": 0.9811471056824217, + "grad_norm": 0.19164782799714325, + "learning_rate": 5.570651021540556e-06, + "loss": 0.0184, + "step": 3695 + }, + { + "epoch": 0.9814126394052045, + "grad_norm": 0.6033608441019069, + "learning_rate": 5.5684699289652486e-06, + "loss": 0.0432, + "step": 3696 + }, + { + "epoch": 0.9816781731279872, + "grad_norm": 0.4163256813920979, + "learning_rate": 5.566288726795642e-06, + "loss": 0.0525, + "step": 3697 + }, + { + "epoch": 0.98194370685077, + "grad_norm": 0.21299614146581108, + "learning_rate": 5.564107415452247e-06, + "loss": 0.0286, + "step": 3698 + }, + { + "epoch": 0.9822092405735529, + "grad_norm": 0.014609210643025195, + "learning_rate": 5.561925995355595e-06, + "loss": 0.0004, + "step": 3699 + }, + { + "epoch": 0.9824747742963357, + "grad_norm": 0.393442899417264, + "learning_rate": 5.559744466926239e-06, + "loss": 0.0476, + "step": 3700 + }, + { + "epoch": 0.9827403080191184, + "grad_norm": 0.7603231825571393, + "learning_rate": 5.557562830584752e-06, + "loss": 0.0432, + "step": 3701 + }, + { + "epoch": 0.9830058417419012, + "grad_norm": 0.5915667926871345, + "learning_rate": 5.555381086751725e-06, + "loss": 0.1113, + "step": 3702 + }, + { + "epoch": 0.983271375464684, + "grad_norm": 0.3143478710792181, + "learning_rate": 5.553199235847777e-06, + "loss": 0.0229, + "step": 3703 + }, + { + "epoch": 0.9835369091874668, + "grad_norm": 0.23396467192232726, + "learning_rate": 5.55101727829354e-06, + "loss": 0.0103, + "step": 3704 + }, + { + "epoch": 0.9838024429102497, + "grad_norm": 0.19990315097450354, + "learning_rate": 5.54883521450967e-06, + "loss": 0.0255, + "step": 3705 + }, + { + "epoch": 0.9840679766330324, + "grad_norm": 1.4009435123641103, + "learning_rate": 5.546653044916845e-06, + "loss": 0.1035, + "step": 3706 + }, + { + "epoch": 0.9843335103558152, + "grad_norm": 1.3262362500752496, + "learning_rate": 5.544470769935757e-06, + "loss": 0.0815, + "step": 3707 + }, + { + "epoch": 0.984599044078598, + "grad_norm": 0.5125349458269568, + "learning_rate": 5.542288389987128e-06, + "loss": 0.0354, + "step": 3708 + }, + { + "epoch": 0.9848645778013808, + "grad_norm": 0.10568890500877494, + "learning_rate": 5.540105905491693e-06, + "loss": 0.0115, + "step": 3709 + }, + { + "epoch": 0.9851301115241635, + "grad_norm": 0.19877497386860007, + "learning_rate": 5.53792331687021e-06, + "loss": 0.0255, + "step": 3710 + }, + { + "epoch": 0.9853956452469463, + "grad_norm": 0.8943979357329969, + "learning_rate": 5.535740624543454e-06, + "loss": 0.1934, + "step": 3711 + }, + { + "epoch": 0.9856611789697292, + "grad_norm": 0.010772117126646585, + "learning_rate": 5.533557828932226e-06, + "loss": 0.0003, + "step": 3712 + }, + { + "epoch": 0.985926712692512, + "grad_norm": 0.00469914951212219, + "learning_rate": 5.531374930457341e-06, + "loss": 0.0001, + "step": 3713 + }, + { + "epoch": 0.9861922464152948, + "grad_norm": 0.7893820984761565, + "learning_rate": 5.5291919295396356e-06, + "loss": 0.0815, + "step": 3714 + }, + { + "epoch": 0.9864577801380775, + "grad_norm": 0.19112504119171422, + "learning_rate": 5.527008826599968e-06, + "loss": 0.0206, + "step": 3715 + }, + { + "epoch": 0.9867233138608603, + "grad_norm": 0.6268537775916571, + "learning_rate": 5.524825622059212e-06, + "loss": 0.1113, + "step": 3716 + }, + { + "epoch": 0.9869888475836431, + "grad_norm": 0.6973408162239882, + "learning_rate": 5.522642316338268e-06, + "loss": 0.1455, + "step": 3717 + }, + { + "epoch": 0.987254381306426, + "grad_norm": 0.7996711464470367, + "learning_rate": 5.5204589098580484e-06, + "loss": 0.0354, + "step": 3718 + }, + { + "epoch": 0.9875199150292087, + "grad_norm": 0.8029601578520161, + "learning_rate": 5.51827540303949e-06, + "loss": 0.0576, + "step": 3719 + }, + { + "epoch": 0.9877854487519915, + "grad_norm": 0.002442037184537762, + "learning_rate": 5.516091796303547e-06, + "loss": 0.0001, + "step": 3720 + }, + { + "epoch": 0.9880509824747743, + "grad_norm": 0.6844514270167829, + "learning_rate": 5.513908090071192e-06, + "loss": 0.0525, + "step": 3721 + }, + { + "epoch": 0.9883165161975571, + "grad_norm": 0.19858577230492813, + "learning_rate": 5.511724284763417e-06, + "loss": 0.0184, + "step": 3722 + }, + { + "epoch": 0.9885820499203399, + "grad_norm": 0.5479958920065439, + "learning_rate": 5.509540380801236e-06, + "loss": 0.1035, + "step": 3723 + }, + { + "epoch": 0.9888475836431226, + "grad_norm": 2.920697350588586, + "learning_rate": 5.507356378605679e-06, + "loss": 0.1367, + "step": 3724 + }, + { + "epoch": 0.9891131173659055, + "grad_norm": 0.22115661936395586, + "learning_rate": 5.505172278597794e-06, + "loss": 0.0082, + "step": 3725 + }, + { + "epoch": 0.9893786510886883, + "grad_norm": 0.002622279222874992, + "learning_rate": 5.5029880811986546e-06, + "loss": 0.0001, + "step": 3726 + }, + { + "epoch": 0.9896441848114711, + "grad_norm": 0.24829172354682366, + "learning_rate": 5.500803786829345e-06, + "loss": 0.0317, + "step": 3727 + }, + { + "epoch": 0.9899097185342538, + "grad_norm": 0.015726937028719905, + "learning_rate": 5.4986193959109716e-06, + "loss": 0.0004, + "step": 3728 + }, + { + "epoch": 0.9901752522570366, + "grad_norm": 0.2336391110051149, + "learning_rate": 5.4964349088646584e-06, + "loss": 0.0354, + "step": 3729 + }, + { + "epoch": 0.9904407859798194, + "grad_norm": 0.18317480862973168, + "learning_rate": 5.49425032611155e-06, + "loss": 0.0255, + "step": 3730 + }, + { + "epoch": 0.9907063197026023, + "grad_norm": 0.45158078664479184, + "learning_rate": 5.492065648072808e-06, + "loss": 0.0391, + "step": 3731 + }, + { + "epoch": 0.9909718534253851, + "grad_norm": 0.7036218372397953, + "learning_rate": 5.489880875169612e-06, + "loss": 0.1836, + "step": 3732 + }, + { + "epoch": 0.9912373871481678, + "grad_norm": 1.0341916805707945, + "learning_rate": 5.4876960078231615e-06, + "loss": 0.0576, + "step": 3733 + }, + { + "epoch": 0.9915029208709506, + "grad_norm": 0.11963223889768665, + "learning_rate": 5.485511046454671e-06, + "loss": 0.0146, + "step": 3734 + }, + { + "epoch": 0.9917684545937334, + "grad_norm": 0.23578038518293518, + "learning_rate": 5.483325991485378e-06, + "loss": 0.0354, + "step": 3735 + }, + { + "epoch": 0.9920339883165162, + "grad_norm": 0.7805231528927336, + "learning_rate": 5.4811408433365355e-06, + "loss": 0.0317, + "step": 3736 + }, + { + "epoch": 0.9922995220392989, + "grad_norm": 0.6590195188375937, + "learning_rate": 5.478955602429412e-06, + "loss": 0.0432, + "step": 3737 + }, + { + "epoch": 0.9925650557620818, + "grad_norm": 0.2884661106791668, + "learning_rate": 5.4767702691852955e-06, + "loss": 0.0082, + "step": 3738 + }, + { + "epoch": 0.9928305894848646, + "grad_norm": 0.5795916723942305, + "learning_rate": 5.474584844025495e-06, + "loss": 0.0391, + "step": 3739 + }, + { + "epoch": 0.9930961232076474, + "grad_norm": 0.19570653962945442, + "learning_rate": 5.472399327371335e-06, + "loss": 0.0255, + "step": 3740 + }, + { + "epoch": 0.9933616569304302, + "grad_norm": 0.47437718713362304, + "learning_rate": 5.470213719644155e-06, + "loss": 0.0131, + "step": 3741 + }, + { + "epoch": 0.9936271906532129, + "grad_norm": 0.2590782349889445, + "learning_rate": 5.468028021265316e-06, + "loss": 0.0255, + "step": 3742 + }, + { + "epoch": 0.9938927243759957, + "grad_norm": 0.6621932539397766, + "learning_rate": 5.465842232656194e-06, + "loss": 0.1279, + "step": 3743 + }, + { + "epoch": 0.9941582580987786, + "grad_norm": 0.0023276005347458123, + "learning_rate": 5.463656354238184e-06, + "loss": 0.0001, + "step": 3744 + }, + { + "epoch": 0.9944237918215614, + "grad_norm": 1.140629768252437, + "learning_rate": 5.461470386432697e-06, + "loss": 0.063, + "step": 3745 + }, + { + "epoch": 0.9946893255443441, + "grad_norm": 0.7684371956572211, + "learning_rate": 5.459284329661162e-06, + "loss": 0.1191, + "step": 3746 + }, + { + "epoch": 0.9949548592671269, + "grad_norm": 0.0016930682043989349, + "learning_rate": 5.4570981843450245e-06, + "loss": 0.0, + "step": 3747 + }, + { + "epoch": 0.9952203929899097, + "grad_norm": 1.0482585426051265, + "learning_rate": 5.454911950905748e-06, + "loss": 0.0752, + "step": 3748 + }, + { + "epoch": 0.9954859267126925, + "grad_norm": 0.46483370406649005, + "learning_rate": 5.452725629764813e-06, + "loss": 0.0354, + "step": 3749 + }, + { + "epoch": 0.9957514604354754, + "grad_norm": 0.9939184296673645, + "learning_rate": 5.450539221343715e-06, + "loss": 0.0752, + "step": 3750 + }, + { + "epoch": 0.9960169941582581, + "grad_norm": 0.007011783756214607, + "learning_rate": 5.448352726063971e-06, + "loss": 0.0002, + "step": 3751 + }, + { + "epoch": 0.9962825278810409, + "grad_norm": 0.34791823977212416, + "learning_rate": 5.446166144347108e-06, + "loss": 0.0103, + "step": 3752 + }, + { + "epoch": 0.9965480616038237, + "grad_norm": 0.912298431767587, + "learning_rate": 5.443979476614674e-06, + "loss": 0.0476, + "step": 3753 + }, + { + "epoch": 0.9968135953266065, + "grad_norm": 0.5012510920756169, + "learning_rate": 5.441792723288235e-06, + "loss": 0.0815, + "step": 3754 + }, + { + "epoch": 0.9970791290493892, + "grad_norm": 0.4786327963204643, + "learning_rate": 5.439605884789367e-06, + "loss": 0.0889, + "step": 3755 + }, + { + "epoch": 0.997344662772172, + "grad_norm": 0.21496778800831168, + "learning_rate": 5.4374189615396696e-06, + "loss": 0.0317, + "step": 3756 + }, + { + "epoch": 0.9976101964949549, + "grad_norm": 0.0006071564621294956, + "learning_rate": 5.435231953960756e-06, + "loss": 0.0, + "step": 3757 + }, + { + "epoch": 0.9978757302177377, + "grad_norm": 0.30583630809181617, + "learning_rate": 5.433044862474255e-06, + "loss": 0.0354, + "step": 3758 + }, + { + "epoch": 0.9981412639405205, + "grad_norm": 0.17841791497544582, + "learning_rate": 5.4308576875018114e-06, + "loss": 0.0229, + "step": 3759 + }, + { + "epoch": 0.9984067976633032, + "grad_norm": 2.4781550521502687, + "learning_rate": 5.4286704294650876e-06, + "loss": 0.4316, + "step": 3760 + }, + { + "epoch": 0.998672331386086, + "grad_norm": 0.5696209113775326, + "learning_rate": 5.42648308878576e-06, + "loss": 0.0525, + "step": 3761 + }, + { + "epoch": 0.9989378651088688, + "grad_norm": 0.002903259214047079, + "learning_rate": 5.424295665885523e-06, + "loss": 0.0001, + "step": 3762 + }, + { + "epoch": 0.9992033988316517, + "grad_norm": 0.002810382019861746, + "learning_rate": 5.422108161186086e-06, + "loss": 0.0001, + "step": 3763 + }, + { + "epoch": 0.9994689325544344, + "grad_norm": 0.03831954712788271, + "learning_rate": 5.419920575109173e-06, + "loss": 0.0012, + "step": 3764 + }, + { + "epoch": 0.9997344662772172, + "grad_norm": 0.883574271317008, + "learning_rate": 5.417732908076525e-06, + "loss": 0.0576, + "step": 3765 + }, + { + "epoch": 1.0, + "grad_norm": 0.7124596374923314, + "learning_rate": 5.415545160509901e-06, + "loss": 0.0752, + "step": 3766 + }, + { + "epoch": 1.0002655337227828, + "grad_norm": 0.21050699491626956, + "learning_rate": 5.413357332831069e-06, + "loss": 0.0286, + "step": 3767 + }, + { + "epoch": 1.0005310674455656, + "grad_norm": 0.22865553953701728, + "learning_rate": 5.411169425461822e-06, + "loss": 0.0255, + "step": 3768 + }, + { + "epoch": 1.0007966011683485, + "grad_norm": 0.5074410613579378, + "learning_rate": 5.4089814388239555e-06, + "loss": 0.0258, + "step": 3769 + }, + { + "epoch": 1.0010621348911313, + "grad_norm": 0.8738158112690851, + "learning_rate": 5.406793373339292e-06, + "loss": 0.063, + "step": 3770 + }, + { + "epoch": 1.0013276686139139, + "grad_norm": 0.8756455398918709, + "learning_rate": 5.4046052294296635e-06, + "loss": 0.0391, + "step": 3771 + }, + { + "epoch": 1.0015932023366967, + "grad_norm": 1.0006465872830141, + "learning_rate": 5.40241700751692e-06, + "loss": 0.0752, + "step": 3772 + }, + { + "epoch": 1.0018587360594795, + "grad_norm": 0.02494828408282636, + "learning_rate": 5.400228708022922e-06, + "loss": 0.0008, + "step": 3773 + }, + { + "epoch": 1.0021242697822623, + "grad_norm": 0.7940453904627217, + "learning_rate": 5.398040331369549e-06, + "loss": 0.1035, + "step": 3774 + }, + { + "epoch": 1.0023898035050451, + "grad_norm": 1.5310199462730643, + "learning_rate": 5.395851877978696e-06, + "loss": 0.1279, + "step": 3775 + }, + { + "epoch": 1.002655337227828, + "grad_norm": 0.2077572484010964, + "learning_rate": 5.39366334827227e-06, + "loss": 0.0072, + "step": 3776 + }, + { + "epoch": 1.0029208709506108, + "grad_norm": 0.44818499352334584, + "learning_rate": 5.391474742672193e-06, + "loss": 0.0476, + "step": 3777 + }, + { + "epoch": 1.0031864046733936, + "grad_norm": 0.01092388784396091, + "learning_rate": 5.389286061600402e-06, + "loss": 0.0004, + "step": 3778 + }, + { + "epoch": 1.0034519383961764, + "grad_norm": 0.9820547557235545, + "learning_rate": 5.3870973054788475e-06, + "loss": 0.0525, + "step": 3779 + }, + { + "epoch": 1.003717472118959, + "grad_norm": 0.20063672666743312, + "learning_rate": 5.384908474729501e-06, + "loss": 0.0286, + "step": 3780 + }, + { + "epoch": 1.0039830058417418, + "grad_norm": 1.307468341787533, + "learning_rate": 5.382719569774338e-06, + "loss": 0.1113, + "step": 3781 + }, + { + "epoch": 1.0042485395645246, + "grad_norm": 0.09910892464058764, + "learning_rate": 5.380530591035358e-06, + "loss": 0.0035, + "step": 3782 + }, + { + "epoch": 1.0045140732873075, + "grad_norm": 0.4589619282878988, + "learning_rate": 5.3783415389345665e-06, + "loss": 0.0476, + "step": 3783 + }, + { + "epoch": 1.0047796070100903, + "grad_norm": 0.24587409721400108, + "learning_rate": 5.3761524138939894e-06, + "loss": 0.0317, + "step": 3784 + }, + { + "epoch": 1.005045140732873, + "grad_norm": 0.37927113897157316, + "learning_rate": 5.3739632163356615e-06, + "loss": 0.0131, + "step": 3785 + }, + { + "epoch": 1.005310674455656, + "grad_norm": 0.9482506816499944, + "learning_rate": 5.371773946681638e-06, + "loss": 0.0815, + "step": 3786 + }, + { + "epoch": 1.0055762081784387, + "grad_norm": 0.7564055332732849, + "learning_rate": 5.36958460535398e-06, + "loss": 0.063, + "step": 3787 + }, + { + "epoch": 1.0058417419012216, + "grad_norm": 1.4692829046698042, + "learning_rate": 5.367395192774769e-06, + "loss": 0.1279, + "step": 3788 + }, + { + "epoch": 1.0061072756240041, + "grad_norm": 0.0020758057825133198, + "learning_rate": 5.365205709366099e-06, + "loss": 0.0001, + "step": 3789 + }, + { + "epoch": 1.006372809346787, + "grad_norm": 0.8643616472504103, + "learning_rate": 5.363016155550076e-06, + "loss": 0.0391, + "step": 3790 + }, + { + "epoch": 1.0066383430695698, + "grad_norm": 0.8531241428735703, + "learning_rate": 5.36082653174882e-06, + "loss": 0.0525, + "step": 3791 + }, + { + "epoch": 1.0069038767923526, + "grad_norm": 1.3432980371463943, + "learning_rate": 5.358636838384462e-06, + "loss": 0.1455, + "step": 3792 + }, + { + "epoch": 1.0071694105151354, + "grad_norm": 1.3091439571514856, + "learning_rate": 5.356447075879153e-06, + "loss": 0.0815, + "step": 3793 + }, + { + "epoch": 1.0074349442379182, + "grad_norm": 0.876642566395881, + "learning_rate": 5.354257244655051e-06, + "loss": 0.0815, + "step": 3794 + }, + { + "epoch": 1.007700477960701, + "grad_norm": 0.5975133919646292, + "learning_rate": 5.3520673451343305e-06, + "loss": 0.063, + "step": 3795 + }, + { + "epoch": 1.0079660116834839, + "grad_norm": 0.6852507665678661, + "learning_rate": 5.349877377739179e-06, + "loss": 0.0444, + "step": 3796 + }, + { + "epoch": 1.0082315454062667, + "grad_norm": 0.8358979553690543, + "learning_rate": 5.347687342891793e-06, + "loss": 0.063, + "step": 3797 + }, + { + "epoch": 1.0084970791290493, + "grad_norm": 0.4940553759454101, + "learning_rate": 5.34549724101439e-06, + "loss": 0.0317, + "step": 3798 + }, + { + "epoch": 1.008762612851832, + "grad_norm": 0.38669737247854225, + "learning_rate": 5.3433070725291935e-06, + "loss": 0.0391, + "step": 3799 + }, + { + "epoch": 1.009028146574615, + "grad_norm": 0.26138766412828146, + "learning_rate": 5.341116837858442e-06, + "loss": 0.0117, + "step": 3800 + }, + { + "epoch": 1.0092936802973977, + "grad_norm": 0.8763258196111736, + "learning_rate": 5.338926537424387e-06, + "loss": 0.1035, + "step": 3801 + }, + { + "epoch": 1.0095592140201806, + "grad_norm": 1.121772571365035, + "learning_rate": 5.336736171649294e-06, + "loss": 0.0752, + "step": 3802 + }, + { + "epoch": 1.0098247477429634, + "grad_norm": 0.1840304034062003, + "learning_rate": 5.334545740955437e-06, + "loss": 0.0205, + "step": 3803 + }, + { + "epoch": 1.0100902814657462, + "grad_norm": 0.39602567793168897, + "learning_rate": 5.332355245765107e-06, + "loss": 0.0525, + "step": 3804 + }, + { + "epoch": 1.010355815188529, + "grad_norm": 0.27318002458473084, + "learning_rate": 5.330164686500606e-06, + "loss": 0.0117, + "step": 3805 + }, + { + "epoch": 1.0106213489113118, + "grad_norm": 0.649703103316738, + "learning_rate": 5.327974063584247e-06, + "loss": 0.0229, + "step": 3806 + }, + { + "epoch": 1.0108868826340944, + "grad_norm": 0.6941390251878886, + "learning_rate": 5.325783377438357e-06, + "loss": 0.1553, + "step": 3807 + }, + { + "epoch": 1.0111524163568772, + "grad_norm": 0.9039237867439225, + "learning_rate": 5.323592628485275e-06, + "loss": 0.0688, + "step": 3808 + }, + { + "epoch": 1.01141795007966, + "grad_norm": 0.4405776591777784, + "learning_rate": 5.3214018171473496e-06, + "loss": 0.0258, + "step": 3809 + }, + { + "epoch": 1.0116834838024429, + "grad_norm": 0.16860368132796316, + "learning_rate": 5.319210943846943e-06, + "loss": 0.0206, + "step": 3810 + }, + { + "epoch": 1.0119490175252257, + "grad_norm": 0.46946123119381444, + "learning_rate": 5.3170200090064325e-06, + "loss": 0.0432, + "step": 3811 + }, + { + "epoch": 1.0122145512480085, + "grad_norm": 0.5596888944072114, + "learning_rate": 5.314829013048204e-06, + "loss": 0.0432, + "step": 3812 + }, + { + "epoch": 1.0124800849707913, + "grad_norm": 0.3659964573386736, + "learning_rate": 5.312637956394654e-06, + "loss": 0.0391, + "step": 3813 + }, + { + "epoch": 1.0127456186935742, + "grad_norm": 0.4414821093268419, + "learning_rate": 5.3104468394681926e-06, + "loss": 0.0317, + "step": 3814 + }, + { + "epoch": 1.013011152416357, + "grad_norm": 0.09118536177092872, + "learning_rate": 5.308255662691243e-06, + "loss": 0.0039, + "step": 3815 + }, + { + "epoch": 1.0132766861391396, + "grad_norm": 0.9518153957788452, + "learning_rate": 5.306064426486237e-06, + "loss": 0.0576, + "step": 3816 + }, + { + "epoch": 1.0135422198619224, + "grad_norm": 0.33586862600495404, + "learning_rate": 5.30387313127562e-06, + "loss": 0.0432, + "step": 3817 + }, + { + "epoch": 1.0138077535847052, + "grad_norm": 0.18750386273239655, + "learning_rate": 5.301681777481846e-06, + "loss": 0.0229, + "step": 3818 + }, + { + "epoch": 1.014073287307488, + "grad_norm": 0.644380037242849, + "learning_rate": 5.2994903655273825e-06, + "loss": 0.0889, + "step": 3819 + }, + { + "epoch": 1.0143388210302708, + "grad_norm": 1.420496507089352, + "learning_rate": 5.2972988958347095e-06, + "loss": 0.0815, + "step": 3820 + }, + { + "epoch": 1.0146043547530537, + "grad_norm": 0.014457210763818232, + "learning_rate": 5.295107368826316e-06, + "loss": 0.0005, + "step": 3821 + }, + { + "epoch": 1.0148698884758365, + "grad_norm": 0.626872956662359, + "learning_rate": 5.292915784924703e-06, + "loss": 0.1191, + "step": 3822 + }, + { + "epoch": 1.0151354221986193, + "grad_norm": 0.29471960947470066, + "learning_rate": 5.290724144552379e-06, + "loss": 0.0255, + "step": 3823 + }, + { + "epoch": 1.0154009559214021, + "grad_norm": 0.1755861759665334, + "learning_rate": 5.288532448131871e-06, + "loss": 0.0206, + "step": 3824 + }, + { + "epoch": 1.0156664896441847, + "grad_norm": 0.8492775011100269, + "learning_rate": 5.286340696085709e-06, + "loss": 0.063, + "step": 3825 + }, + { + "epoch": 1.0159320233669675, + "grad_norm": 0.3261419649875987, + "learning_rate": 5.284148888836438e-06, + "loss": 0.0286, + "step": 3826 + }, + { + "epoch": 1.0161975570897503, + "grad_norm": 0.26311093531847035, + "learning_rate": 5.2819570268066136e-06, + "loss": 0.0286, + "step": 3827 + }, + { + "epoch": 1.0164630908125332, + "grad_norm": 0.003495377534466825, + "learning_rate": 5.2797651104187965e-06, + "loss": 0.0001, + "step": 3828 + }, + { + "epoch": 1.016728624535316, + "grad_norm": 0.4754621123014767, + "learning_rate": 5.277573140095569e-06, + "loss": 0.0391, + "step": 3829 + }, + { + "epoch": 1.0169941582580988, + "grad_norm": 0.02071403555014752, + "learning_rate": 5.2753811162595114e-06, + "loss": 0.0006, + "step": 3830 + }, + { + "epoch": 1.0172596919808816, + "grad_norm": 0.20645805944485318, + "learning_rate": 5.273189039333224e-06, + "loss": 0.0229, + "step": 3831 + }, + { + "epoch": 1.0175252257036644, + "grad_norm": 0.27780044934226716, + "learning_rate": 5.270996909739311e-06, + "loss": 0.0092, + "step": 3832 + }, + { + "epoch": 1.0177907594264473, + "grad_norm": 0.9052819842702935, + "learning_rate": 5.26880472790039e-06, + "loss": 0.063, + "step": 3833 + }, + { + "epoch": 1.0180562931492299, + "grad_norm": 0.15824206586605163, + "learning_rate": 5.266612494239088e-06, + "loss": 0.0184, + "step": 3834 + }, + { + "epoch": 1.0183218268720127, + "grad_norm": 0.29373995129838915, + "learning_rate": 5.264420209178043e-06, + "loss": 0.0317, + "step": 3835 + }, + { + "epoch": 1.0185873605947955, + "grad_norm": 0.12428148196168284, + "learning_rate": 5.262227873139898e-06, + "loss": 0.0131, + "step": 3836 + }, + { + "epoch": 1.0188528943175783, + "grad_norm": 1.0060804300424908, + "learning_rate": 5.260035486547311e-06, + "loss": 0.0432, + "step": 3837 + }, + { + "epoch": 1.0191184280403611, + "grad_norm": 0.8762522629592433, + "learning_rate": 5.257843049822951e-06, + "loss": 0.1191, + "step": 3838 + }, + { + "epoch": 1.019383961763144, + "grad_norm": 0.7249857924750114, + "learning_rate": 5.255650563389492e-06, + "loss": 0.1279, + "step": 3839 + }, + { + "epoch": 1.0196494954859268, + "grad_norm": 0.16696681737635566, + "learning_rate": 5.253458027669619e-06, + "loss": 0.0184, + "step": 3840 + }, + { + "epoch": 1.0199150292087096, + "grad_norm": 0.6289212822206072, + "learning_rate": 5.251265443086027e-06, + "loss": 0.0354, + "step": 3841 + }, + { + "epoch": 1.0201805629314924, + "grad_norm": 1.7770550855777398, + "learning_rate": 5.249072810061422e-06, + "loss": 0.0815, + "step": 3842 + }, + { + "epoch": 1.020446096654275, + "grad_norm": 0.9074047730236781, + "learning_rate": 5.246880129018515e-06, + "loss": 0.0525, + "step": 3843 + }, + { + "epoch": 1.0207116303770578, + "grad_norm": 0.010936229256195298, + "learning_rate": 5.244687400380033e-06, + "loss": 0.0003, + "step": 3844 + }, + { + "epoch": 1.0209771640998406, + "grad_norm": 0.8376123823142112, + "learning_rate": 5.242494624568705e-06, + "loss": 0.0165, + "step": 3845 + }, + { + "epoch": 1.0212426978226234, + "grad_norm": 0.18497417255688325, + "learning_rate": 5.240301802007273e-06, + "loss": 0.0184, + "step": 3846 + }, + { + "epoch": 1.0215082315454063, + "grad_norm": 0.9712520687066858, + "learning_rate": 5.238108933118487e-06, + "loss": 0.063, + "step": 3847 + }, + { + "epoch": 1.021773765268189, + "grad_norm": 0.000543771891241758, + "learning_rate": 5.235916018325109e-06, + "loss": 0.0, + "step": 3848 + }, + { + "epoch": 1.022039298990972, + "grad_norm": 0.8170343144840189, + "learning_rate": 5.233723058049905e-06, + "loss": 0.1367, + "step": 3849 + }, + { + "epoch": 1.0223048327137547, + "grad_norm": 0.47650507343641635, + "learning_rate": 5.231530052715651e-06, + "loss": 0.0317, + "step": 3850 + }, + { + "epoch": 1.0225703664365375, + "grad_norm": 0.17553121807132122, + "learning_rate": 5.229337002745132e-06, + "loss": 0.0165, + "step": 3851 + }, + { + "epoch": 1.0228359001593201, + "grad_norm": 0.8311217040130671, + "learning_rate": 5.227143908561146e-06, + "loss": 0.0432, + "step": 3852 + }, + { + "epoch": 1.023101433882103, + "grad_norm": 0.22002867753980834, + "learning_rate": 5.224950770586491e-06, + "loss": 0.0255, + "step": 3853 + }, + { + "epoch": 1.0233669676048858, + "grad_norm": 0.9329928186228303, + "learning_rate": 5.222757589243982e-06, + "loss": 0.0391, + "step": 3854 + }, + { + "epoch": 1.0236325013276686, + "grad_norm": 1.3888154623581277, + "learning_rate": 5.220564364956438e-06, + "loss": 0.0957, + "step": 3855 + }, + { + "epoch": 1.0238980350504514, + "grad_norm": 0.7719234891229885, + "learning_rate": 5.218371098146684e-06, + "loss": 0.1455, + "step": 3856 + }, + { + "epoch": 1.0241635687732342, + "grad_norm": 0.7417043643756472, + "learning_rate": 5.216177789237559e-06, + "loss": 0.0432, + "step": 3857 + }, + { + "epoch": 1.024429102496017, + "grad_norm": 0.0016057983105514549, + "learning_rate": 5.2139844386519046e-06, + "loss": 0.0, + "step": 3858 + }, + { + "epoch": 1.0246946362187999, + "grad_norm": 0.610816556396215, + "learning_rate": 5.211791046812573e-06, + "loss": 0.0432, + "step": 3859 + }, + { + "epoch": 1.0249601699415827, + "grad_norm": 0.3008508296843735, + "learning_rate": 5.209597614142428e-06, + "loss": 0.0286, + "step": 3860 + }, + { + "epoch": 1.0252257036643653, + "grad_norm": 0.007420285779009959, + "learning_rate": 5.207404141064334e-06, + "loss": 0.0002, + "step": 3861 + }, + { + "epoch": 1.025491237387148, + "grad_norm": 0.6210101875894289, + "learning_rate": 5.205210628001167e-06, + "loss": 0.0082, + "step": 3862 + }, + { + "epoch": 1.025756771109931, + "grad_norm": 0.14339278051977042, + "learning_rate": 5.203017075375813e-06, + "loss": 0.0146, + "step": 3863 + }, + { + "epoch": 1.0260223048327137, + "grad_norm": 0.5093612632221544, + "learning_rate": 5.20082348361116e-06, + "loss": 0.0476, + "step": 3864 + }, + { + "epoch": 1.0262878385554965, + "grad_norm": 0.4290420162094184, + "learning_rate": 5.198629853130108e-06, + "loss": 0.0354, + "step": 3865 + }, + { + "epoch": 1.0265533722782794, + "grad_norm": 0.008316870852760205, + "learning_rate": 5.196436184355563e-06, + "loss": 0.0002, + "step": 3866 + }, + { + "epoch": 1.0268189060010622, + "grad_norm": 0.05074378231510025, + "learning_rate": 5.194242477710439e-06, + "loss": 0.0015, + "step": 3867 + }, + { + "epoch": 1.027084439723845, + "grad_norm": 0.20471837253782607, + "learning_rate": 5.192048733617654e-06, + "loss": 0.0255, + "step": 3868 + }, + { + "epoch": 1.0273499734466278, + "grad_norm": 0.005525171846192723, + "learning_rate": 5.189854952500139e-06, + "loss": 0.0002, + "step": 3869 + }, + { + "epoch": 1.0276155071694104, + "grad_norm": 0.21078293439993118, + "learning_rate": 5.187661134780829e-06, + "loss": 0.0205, + "step": 3870 + }, + { + "epoch": 1.0278810408921932, + "grad_norm": 3.371825934571778, + "learning_rate": 5.1854672808826656e-06, + "loss": 0.0957, + "step": 3871 + }, + { + "epoch": 1.028146574614976, + "grad_norm": 0.1965496605045824, + "learning_rate": 5.183273391228598e-06, + "loss": 0.0184, + "step": 3872 + }, + { + "epoch": 1.0284121083377589, + "grad_norm": 0.21055206333270785, + "learning_rate": 5.181079466241582e-06, + "loss": 0.0165, + "step": 3873 + }, + { + "epoch": 1.0286776420605417, + "grad_norm": 0.5955790191374454, + "learning_rate": 5.178885506344581e-06, + "loss": 0.0229, + "step": 3874 + }, + { + "epoch": 1.0289431757833245, + "grad_norm": 1.157600517619345, + "learning_rate": 5.1766915119605635e-06, + "loss": 0.063, + "step": 3875 + }, + { + "epoch": 1.0292087095061073, + "grad_norm": 0.25048049264215533, + "learning_rate": 5.174497483512506e-06, + "loss": 0.0229, + "step": 3876 + }, + { + "epoch": 1.0294742432288901, + "grad_norm": 0.8561095563649015, + "learning_rate": 5.17230342142339e-06, + "loss": 0.0317, + "step": 3877 + }, + { + "epoch": 1.029739776951673, + "grad_norm": 1.401885585902608, + "learning_rate": 5.1701093261162095e-06, + "loss": 0.0576, + "step": 3878 + }, + { + "epoch": 1.0300053106744556, + "grad_norm": 0.5630642311318185, + "learning_rate": 5.167915198013956e-06, + "loss": 0.0391, + "step": 3879 + }, + { + "epoch": 1.0302708443972384, + "grad_norm": 0.1768915001768712, + "learning_rate": 5.165721037539634e-06, + "loss": 0.0146, + "step": 3880 + }, + { + "epoch": 1.0305363781200212, + "grad_norm": 1.130136974920396, + "learning_rate": 5.163526845116249e-06, + "loss": 0.063, + "step": 3881 + }, + { + "epoch": 1.030801911842804, + "grad_norm": 0.20422766315148758, + "learning_rate": 5.161332621166817e-06, + "loss": 0.0184, + "step": 3882 + }, + { + "epoch": 1.0310674455655868, + "grad_norm": 0.868737092654465, + "learning_rate": 5.159138366114358e-06, + "loss": 0.1553, + "step": 3883 + }, + { + "epoch": 1.0313329792883696, + "grad_norm": 0.23270817668496274, + "learning_rate": 5.1569440803818996e-06, + "loss": 0.0205, + "step": 3884 + }, + { + "epoch": 1.0315985130111525, + "grad_norm": 0.06864806136707152, + "learning_rate": 5.154749764392474e-06, + "loss": 0.0056, + "step": 3885 + }, + { + "epoch": 1.0318640467339353, + "grad_norm": 0.6789604784721732, + "learning_rate": 5.1525554185691175e-06, + "loss": 0.0165, + "step": 3886 + }, + { + "epoch": 1.032129580456718, + "grad_norm": 0.014311447641845988, + "learning_rate": 5.150361043334876e-06, + "loss": 0.0003, + "step": 3887 + }, + { + "epoch": 1.0323951141795007, + "grad_norm": 0.4061434646904656, + "learning_rate": 5.148166639112799e-06, + "loss": 0.0317, + "step": 3888 + }, + { + "epoch": 1.0326606479022835, + "grad_norm": 0.14723473128947823, + "learning_rate": 5.145972206325942e-06, + "loss": 0.0115, + "step": 3889 + }, + { + "epoch": 1.0329261816250663, + "grad_norm": 0.13653592427124006, + "learning_rate": 5.143777745397365e-06, + "loss": 0.0103, + "step": 3890 + }, + { + "epoch": 1.0331917153478491, + "grad_norm": 0.4097233726506065, + "learning_rate": 5.1415832567501335e-06, + "loss": 0.0286, + "step": 3891 + }, + { + "epoch": 1.033457249070632, + "grad_norm": 0.572321674199075, + "learning_rate": 5.1393887408073195e-06, + "loss": 0.0184, + "step": 3892 + }, + { + "epoch": 1.0337227827934148, + "grad_norm": 0.002578691435307613, + "learning_rate": 5.1371941979920015e-06, + "loss": 0.0001, + "step": 3893 + }, + { + "epoch": 1.0339883165161976, + "grad_norm": 1.132190915638641, + "learning_rate": 5.13499962872726e-06, + "loss": 0.1279, + "step": 3894 + }, + { + "epoch": 1.0342538502389804, + "grad_norm": 0.19920600838829475, + "learning_rate": 5.132805033436182e-06, + "loss": 0.0129, + "step": 3895 + }, + { + "epoch": 1.0345193839617632, + "grad_norm": 0.005481239886738112, + "learning_rate": 5.13061041254186e-06, + "loss": 0.0002, + "step": 3896 + }, + { + "epoch": 1.0347849176845458, + "grad_norm": 3.012636312175512, + "learning_rate": 5.128415766467392e-06, + "loss": 0.1367, + "step": 3897 + }, + { + "epoch": 1.0350504514073287, + "grad_norm": 1.0900970420410285, + "learning_rate": 5.126221095635877e-06, + "loss": 0.2031, + "step": 3898 + }, + { + "epoch": 1.0353159851301115, + "grad_norm": 0.45593330374622143, + "learning_rate": 5.124026400470423e-06, + "loss": 0.0286, + "step": 3899 + }, + { + "epoch": 1.0355815188528943, + "grad_norm": 0.3284270875477546, + "learning_rate": 5.121831681394143e-06, + "loss": 0.0229, + "step": 3900 + }, + { + "epoch": 1.035847052575677, + "grad_norm": 0.08286253683110838, + "learning_rate": 5.119636938830151e-06, + "loss": 0.0019, + "step": 3901 + }, + { + "epoch": 1.03611258629846, + "grad_norm": 3.6244909647115113, + "learning_rate": 5.117442173201569e-06, + "loss": 0.1553, + "step": 3902 + }, + { + "epoch": 1.0363781200212427, + "grad_norm": 0.17019013511945463, + "learning_rate": 5.11524738493152e-06, + "loss": 0.0146, + "step": 3903 + }, + { + "epoch": 1.0366436537440256, + "grad_norm": 0.8082129151013328, + "learning_rate": 5.113052574443136e-06, + "loss": 0.1455, + "step": 3904 + }, + { + "epoch": 1.0369091874668084, + "grad_norm": 1.2765568154281313, + "learning_rate": 5.110857742159547e-06, + "loss": 0.0525, + "step": 3905 + }, + { + "epoch": 1.037174721189591, + "grad_norm": 0.011569285523818161, + "learning_rate": 5.108662888503895e-06, + "loss": 0.0003, + "step": 3906 + }, + { + "epoch": 1.0374402549123738, + "grad_norm": 0.12745734417875482, + "learning_rate": 5.106468013899319e-06, + "loss": 0.0115, + "step": 3907 + }, + { + "epoch": 1.0377057886351566, + "grad_norm": 0.999963169218087, + "learning_rate": 5.104273118768964e-06, + "loss": 0.063, + "step": 3908 + }, + { + "epoch": 1.0379713223579394, + "grad_norm": 0.17852805699286778, + "learning_rate": 5.102078203535982e-06, + "loss": 0.0164, + "step": 3909 + }, + { + "epoch": 1.0382368560807222, + "grad_norm": 0.2613852806176485, + "learning_rate": 5.099883268623526e-06, + "loss": 0.0205, + "step": 3910 + }, + { + "epoch": 1.038502389803505, + "grad_norm": 0.47480350095864804, + "learning_rate": 5.097688314454755e-06, + "loss": 0.0317, + "step": 3911 + }, + { + "epoch": 1.0387679235262879, + "grad_norm": 0.9345214941157061, + "learning_rate": 5.095493341452828e-06, + "loss": 0.1738, + "step": 3912 + }, + { + "epoch": 1.0390334572490707, + "grad_norm": 0.000661359351510568, + "learning_rate": 5.093298350040913e-06, + "loss": 0.0, + "step": 3913 + }, + { + "epoch": 1.0392989909718535, + "grad_norm": 0.8246731819746077, + "learning_rate": 5.091103340642175e-06, + "loss": 0.0815, + "step": 3914 + }, + { + "epoch": 1.0395645246946361, + "grad_norm": 0.10997176945135709, + "learning_rate": 5.088908313679788e-06, + "loss": 0.0092, + "step": 3915 + }, + { + "epoch": 1.039830058417419, + "grad_norm": 0.9768069549319981, + "learning_rate": 5.0867132695769265e-06, + "loss": 0.0525, + "step": 3916 + }, + { + "epoch": 1.0400955921402018, + "grad_norm": 0.1592753398312262, + "learning_rate": 5.084518208756769e-06, + "loss": 0.0164, + "step": 3917 + }, + { + "epoch": 1.0403611258629846, + "grad_norm": 4.4007985865226855, + "learning_rate": 5.082323131642496e-06, + "loss": 0.1191, + "step": 3918 + }, + { + "epoch": 1.0406266595857674, + "grad_norm": 0.13122181178066158, + "learning_rate": 5.080128038657296e-06, + "loss": 0.0129, + "step": 3919 + }, + { + "epoch": 1.0408921933085502, + "grad_norm": 2.139055450557411, + "learning_rate": 5.077932930224356e-06, + "loss": 0.1113, + "step": 3920 + }, + { + "epoch": 1.041157727031333, + "grad_norm": 0.7358936927859386, + "learning_rate": 5.075737806766866e-06, + "loss": 0.0576, + "step": 3921 + }, + { + "epoch": 1.0414232607541158, + "grad_norm": 0.0009215771538169852, + "learning_rate": 5.073542668708018e-06, + "loss": 0.0, + "step": 3922 + }, + { + "epoch": 1.0416887944768987, + "grad_norm": 2.591976984812403, + "learning_rate": 5.071347516471012e-06, + "loss": 0.1367, + "step": 3923 + }, + { + "epoch": 1.0419543281996813, + "grad_norm": 0.0059001007808243, + "learning_rate": 5.069152350479048e-06, + "loss": 0.0001, + "step": 3924 + }, + { + "epoch": 1.042219861922464, + "grad_norm": 0.1781050338600249, + "learning_rate": 5.066957171155324e-06, + "loss": 0.0165, + "step": 3925 + }, + { + "epoch": 1.042485395645247, + "grad_norm": 0.3109780256172115, + "learning_rate": 5.064761978923048e-06, + "loss": 0.0352, + "step": 3926 + }, + { + "epoch": 1.0427509293680297, + "grad_norm": 0.9611739473350313, + "learning_rate": 5.062566774205426e-06, + "loss": 0.1113, + "step": 3927 + }, + { + "epoch": 1.0430164630908125, + "grad_norm": 0.8395551228840635, + "learning_rate": 5.060371557425669e-06, + "loss": 0.0391, + "step": 3928 + }, + { + "epoch": 1.0432819968135953, + "grad_norm": 0.13880465707453873, + "learning_rate": 5.0581763290069865e-06, + "loss": 0.0146, + "step": 3929 + }, + { + "epoch": 1.0435475305363782, + "grad_norm": 0.3861856309301203, + "learning_rate": 5.0559810893725936e-06, + "loss": 0.0184, + "step": 3930 + }, + { + "epoch": 1.043813064259161, + "grad_norm": 0.4325123131197369, + "learning_rate": 5.0537858389457075e-06, + "loss": 0.0205, + "step": 3931 + }, + { + "epoch": 1.0440785979819438, + "grad_norm": 0.4309789146651493, + "learning_rate": 5.051590578149546e-06, + "loss": 0.0317, + "step": 3932 + }, + { + "epoch": 1.0443441317047264, + "grad_norm": 0.0018602926385112687, + "learning_rate": 5.049395307407329e-06, + "loss": 0.0001, + "step": 3933 + }, + { + "epoch": 1.0446096654275092, + "grad_norm": 0.7701484807634638, + "learning_rate": 5.0472000271422794e-06, + "loss": 0.1279, + "step": 3934 + }, + { + "epoch": 1.044875199150292, + "grad_norm": 0.7510210906618583, + "learning_rate": 5.0450047377776225e-06, + "loss": 0.0317, + "step": 3935 + }, + { + "epoch": 1.0451407328730749, + "grad_norm": 0.0035621840501628313, + "learning_rate": 5.042809439736581e-06, + "loss": 0.0001, + "step": 3936 + }, + { + "epoch": 1.0454062665958577, + "grad_norm": 1.1145965320514017, + "learning_rate": 5.040614133442387e-06, + "loss": 0.0576, + "step": 3937 + }, + { + "epoch": 1.0456718003186405, + "grad_norm": 0.26868177829115636, + "learning_rate": 5.038418819318265e-06, + "loss": 0.0255, + "step": 3938 + }, + { + "epoch": 1.0459373340414233, + "grad_norm": 0.7143276275809338, + "learning_rate": 5.036223497787448e-06, + "loss": 0.1367, + "step": 3939 + }, + { + "epoch": 1.0462028677642061, + "grad_norm": 0.1409031732231826, + "learning_rate": 5.034028169273168e-06, + "loss": 0.0146, + "step": 3940 + }, + { + "epoch": 1.046468401486989, + "grad_norm": 0.02485407302385219, + "learning_rate": 5.031832834198658e-06, + "loss": 0.0005, + "step": 3941 + }, + { + "epoch": 1.0467339352097715, + "grad_norm": 0.1944152575598485, + "learning_rate": 5.029637492987153e-06, + "loss": 0.0229, + "step": 3942 + }, + { + "epoch": 1.0469994689325544, + "grad_norm": 0.0011765203014942663, + "learning_rate": 5.027442146061889e-06, + "loss": 0.0, + "step": 3943 + }, + { + "epoch": 1.0472650026553372, + "grad_norm": 1.2246272593803174, + "learning_rate": 5.025246793846103e-06, + "loss": 0.0391, + "step": 3944 + }, + { + "epoch": 1.04753053637812, + "grad_norm": 0.023140015916557802, + "learning_rate": 5.023051436763034e-06, + "loss": 0.0007, + "step": 3945 + }, + { + "epoch": 1.0477960701009028, + "grad_norm": 1.3380817810929808, + "learning_rate": 5.0208560752359195e-06, + "loss": 0.0432, + "step": 3946 + }, + { + "epoch": 1.0480616038236856, + "grad_norm": 0.13597716123385897, + "learning_rate": 5.018660709687999e-06, + "loss": 0.0129, + "step": 3947 + }, + { + "epoch": 1.0483271375464684, + "grad_norm": 0.0013626061778893937, + "learning_rate": 5.016465340542514e-06, + "loss": 0.0, + "step": 3948 + }, + { + "epoch": 1.0485926712692513, + "grad_norm": 0.7735839844901511, + "learning_rate": 5.0142699682227035e-06, + "loss": 0.0229, + "step": 3949 + }, + { + "epoch": 1.048858204992034, + "grad_norm": 0.47162806319460476, + "learning_rate": 5.012074593151814e-06, + "loss": 0.0432, + "step": 3950 + }, + { + "epoch": 1.0491237387148167, + "grad_norm": 0.0015779927343305105, + "learning_rate": 5.009879215753085e-06, + "loss": 0.0, + "step": 3951 + }, + { + "epoch": 1.0493892724375995, + "grad_norm": 0.002623563831475705, + "learning_rate": 5.007683836449761e-06, + "loss": 0.0001, + "step": 3952 + }, + { + "epoch": 1.0496548061603823, + "grad_norm": 0.44095254607001877, + "learning_rate": 5.005488455665082e-06, + "loss": 0.0391, + "step": 3953 + }, + { + "epoch": 1.0499203398831651, + "grad_norm": 0.10583630959708755, + "learning_rate": 5.003293073822293e-06, + "loss": 0.0092, + "step": 3954 + }, + { + "epoch": 1.050185873605948, + "grad_norm": 0.10477553178399741, + "learning_rate": 5.001097691344639e-06, + "loss": 0.0092, + "step": 3955 + }, + { + "epoch": 1.0504514073287308, + "grad_norm": 2.3418924846656757, + "learning_rate": 4.998902308655362e-06, + "loss": 0.2344, + "step": 3956 + }, + { + "epoch": 1.0507169410515136, + "grad_norm": 0.22031745803656774, + "learning_rate": 4.996706926177708e-06, + "loss": 0.0255, + "step": 3957 + }, + { + "epoch": 1.0509824747742964, + "grad_norm": 0.008919057783529843, + "learning_rate": 4.99451154433492e-06, + "loss": 0.0002, + "step": 3958 + }, + { + "epoch": 1.0512480084970792, + "grad_norm": 1.134394441196696, + "learning_rate": 4.992316163550241e-06, + "loss": 0.0889, + "step": 3959 + }, + { + "epoch": 1.0515135422198618, + "grad_norm": 0.9777303924860541, + "learning_rate": 4.990120784246917e-06, + "loss": 0.1279, + "step": 3960 + }, + { + "epoch": 1.0517790759426446, + "grad_norm": 0.3245599342549157, + "learning_rate": 4.987925406848188e-06, + "loss": 0.0286, + "step": 3961 + }, + { + "epoch": 1.0520446096654275, + "grad_norm": 0.20109912647813233, + "learning_rate": 4.985730031777297e-06, + "loss": 0.0164, + "step": 3962 + }, + { + "epoch": 1.0523101433882103, + "grad_norm": 0.7848346887060025, + "learning_rate": 4.983534659457489e-06, + "loss": 0.0815, + "step": 3963 + }, + { + "epoch": 1.052575677110993, + "grad_norm": 4.175773749771835, + "learning_rate": 4.981339290312004e-06, + "loss": 0.4766, + "step": 3964 + }, + { + "epoch": 1.052841210833776, + "grad_norm": 0.1293653485377906, + "learning_rate": 4.979143924764081e-06, + "loss": 0.0129, + "step": 3965 + }, + { + "epoch": 1.0531067445565587, + "grad_norm": 0.08387864344347365, + "learning_rate": 4.976948563236967e-06, + "loss": 0.0072, + "step": 3966 + }, + { + "epoch": 1.0533722782793415, + "grad_norm": 0.20452500329868045, + "learning_rate": 4.9747532061538975e-06, + "loss": 0.0146, + "step": 3967 + }, + { + "epoch": 1.0536378120021244, + "grad_norm": 0.10159938512262978, + "learning_rate": 4.972557853938112e-06, + "loss": 0.0092, + "step": 3968 + }, + { + "epoch": 1.053903345724907, + "grad_norm": 0.17687690205055934, + "learning_rate": 4.970362507012848e-06, + "loss": 0.0184, + "step": 3969 + }, + { + "epoch": 1.0541688794476898, + "grad_norm": 0.7632907195216834, + "learning_rate": 4.9681671658013435e-06, + "loss": 0.1455, + "step": 3970 + }, + { + "epoch": 1.0544344131704726, + "grad_norm": 0.008705073314158452, + "learning_rate": 4.965971830726834e-06, + "loss": 0.0003, + "step": 3971 + }, + { + "epoch": 1.0546999468932554, + "grad_norm": 0.12787929421305408, + "learning_rate": 4.963776502212555e-06, + "loss": 0.0129, + "step": 3972 + }, + { + "epoch": 1.0549654806160382, + "grad_norm": 0.8616450062938196, + "learning_rate": 4.961581180681738e-06, + "loss": 0.1738, + "step": 3973 + }, + { + "epoch": 1.055231014338821, + "grad_norm": 0.6192040672481679, + "learning_rate": 4.959385866557615e-06, + "loss": 0.0476, + "step": 3974 + }, + { + "epoch": 1.0554965480616039, + "grad_norm": 2.069364547740031, + "learning_rate": 4.95719056026342e-06, + "loss": 0.1191, + "step": 3975 + }, + { + "epoch": 1.0557620817843867, + "grad_norm": 0.04115511888678888, + "learning_rate": 4.95499526222238e-06, + "loss": 0.0013, + "step": 3976 + }, + { + "epoch": 1.0560276155071695, + "grad_norm": 0.8535580114843936, + "learning_rate": 4.952799972857721e-06, + "loss": 0.0432, + "step": 3977 + }, + { + "epoch": 1.056293149229952, + "grad_norm": 0.0021201084404281908, + "learning_rate": 4.9506046925926725e-06, + "loss": 0.0001, + "step": 3978 + }, + { + "epoch": 1.056558682952735, + "grad_norm": 0.4703657159403737, + "learning_rate": 4.948409421850456e-06, + "loss": 0.0255, + "step": 3979 + }, + { + "epoch": 1.0568242166755177, + "grad_norm": 2.058474969859397, + "learning_rate": 4.946214161054295e-06, + "loss": 0.1279, + "step": 3980 + }, + { + "epoch": 1.0570897503983006, + "grad_norm": 0.0011401655663284185, + "learning_rate": 4.944018910627408e-06, + "loss": 0.0, + "step": 3981 + }, + { + "epoch": 1.0573552841210834, + "grad_norm": 0.2166035843286793, + "learning_rate": 4.941823670993016e-06, + "loss": 0.0229, + "step": 3982 + }, + { + "epoch": 1.0576208178438662, + "grad_norm": 3.4228364561592888, + "learning_rate": 4.939628442574333e-06, + "loss": 0.2891, + "step": 3983 + }, + { + "epoch": 1.057886351566649, + "grad_norm": 0.16076322493558576, + "learning_rate": 4.9374332257945754e-06, + "loss": 0.0205, + "step": 3984 + }, + { + "epoch": 1.0581518852894318, + "grad_norm": 2.0745298774054484, + "learning_rate": 4.935238021076953e-06, + "loss": 0.1641, + "step": 3985 + }, + { + "epoch": 1.0584174190122146, + "grad_norm": 0.14438184017746963, + "learning_rate": 4.933042828844677e-06, + "loss": 0.0184, + "step": 3986 + }, + { + "epoch": 1.0586829527349972, + "grad_norm": 0.06598617206907567, + "learning_rate": 4.930847649520955e-06, + "loss": 0.0027, + "step": 3987 + }, + { + "epoch": 1.05894848645778, + "grad_norm": 0.2895451306487173, + "learning_rate": 4.928652483528989e-06, + "loss": 0.0255, + "step": 3988 + }, + { + "epoch": 1.0592140201805629, + "grad_norm": 0.1853829688687134, + "learning_rate": 4.926457331291983e-06, + "loss": 0.0184, + "step": 3989 + }, + { + "epoch": 1.0594795539033457, + "grad_norm": 0.20132503634546797, + "learning_rate": 4.9242621932331365e-06, + "loss": 0.0205, + "step": 3990 + }, + { + "epoch": 1.0597450876261285, + "grad_norm": 0.18360998531789777, + "learning_rate": 4.922067069775646e-06, + "loss": 0.0184, + "step": 3991 + }, + { + "epoch": 1.0600106213489113, + "grad_norm": 2.701538633592593, + "learning_rate": 4.919871961342704e-06, + "loss": 0.0957, + "step": 3992 + }, + { + "epoch": 1.0602761550716941, + "grad_norm": 0.27720295474550904, + "learning_rate": 4.917676868357504e-06, + "loss": 0.0286, + "step": 3993 + }, + { + "epoch": 1.060541688794477, + "grad_norm": 0.6923653209602763, + "learning_rate": 4.915481791243233e-06, + "loss": 0.1455, + "step": 3994 + }, + { + "epoch": 1.0608072225172598, + "grad_norm": 3.9435757276474903, + "learning_rate": 4.913286730423075e-06, + "loss": 0.5, + "step": 3995 + }, + { + "epoch": 1.0610727562400424, + "grad_norm": 0.8108605401714497, + "learning_rate": 4.911091686320213e-06, + "loss": 0.1934, + "step": 3996 + }, + { + "epoch": 1.0613382899628252, + "grad_norm": 0.10522958290419485, + "learning_rate": 4.908896659357826e-06, + "loss": 0.0115, + "step": 3997 + }, + { + "epoch": 1.061603823685608, + "grad_norm": 0.18405079565354707, + "learning_rate": 4.906701649959089e-06, + "loss": 0.0229, + "step": 3998 + }, + { + "epoch": 1.0618693574083908, + "grad_norm": 0.7980740544222001, + "learning_rate": 4.904506658547173e-06, + "loss": 0.1738, + "step": 3999 + }, + { + "epoch": 1.0621348911311737, + "grad_norm": 0.14773231917697055, + "learning_rate": 4.9023116855452465e-06, + "loss": 0.0165, + "step": 4000 + } + ], + "logging_steps": 1.0, + "max_steps": 7532, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 2000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 762398638080000.0, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}