{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 6.0, "eval_steps": 500, "global_step": 1590, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003780718336483932, "grad_norm": 18.518301010131836, "learning_rate": 0.0, "loss": 7.5825, "step": 1 }, { "epoch": 0.007561436672967864, "grad_norm": 16.119892120361328, "learning_rate": 3.1446540880503146e-08, "loss": 7.5404, "step": 2 }, { "epoch": 0.011342155009451797, "grad_norm": 75.88004302978516, "learning_rate": 6.289308176100629e-08, "loss": 7.5524, "step": 3 }, { "epoch": 0.015122873345935728, "grad_norm": 20.45383644104004, "learning_rate": 9.433962264150944e-08, "loss": 7.5742, "step": 4 }, { "epoch": 0.01890359168241966, "grad_norm": 19.595623016357422, "learning_rate": 1.2578616352201258e-07, "loss": 7.5086, "step": 5 }, { "epoch": 0.022684310018903593, "grad_norm": 32.997398376464844, "learning_rate": 1.5723270440251575e-07, "loss": 7.519, "step": 6 }, { "epoch": 0.026465028355387523, "grad_norm": 24.622861862182617, "learning_rate": 1.886792452830189e-07, "loss": 7.472, "step": 7 }, { "epoch": 0.030245746691871456, "grad_norm": 34.02603530883789, "learning_rate": 2.2012578616352203e-07, "loss": 7.5547, "step": 8 }, { "epoch": 0.034026465028355386, "grad_norm": 24.93535041809082, "learning_rate": 2.5157232704402517e-07, "loss": 7.5492, "step": 9 }, { "epoch": 0.03780718336483932, "grad_norm": 29.976503372192383, "learning_rate": 2.8301886792452833e-07, "loss": 7.5392, "step": 10 }, { "epoch": 0.04158790170132325, "grad_norm": 19.242721557617188, "learning_rate": 3.144654088050315e-07, "loss": 7.5246, "step": 11 }, { "epoch": 0.045368620037807186, "grad_norm": 22.243343353271484, "learning_rate": 3.459119496855346e-07, "loss": 7.537, "step": 12 }, { "epoch": 0.04914933837429111, "grad_norm": 119.47364044189453, "learning_rate": 3.773584905660378e-07, "loss": 7.4558, "step": 13 }, { "epoch": 0.052930056710775046, "grad_norm": 17.34946060180664, "learning_rate": 4.088050314465409e-07, "loss": 7.5581, "step": 14 }, { "epoch": 0.05671077504725898, "grad_norm": 24.03491973876953, "learning_rate": 4.4025157232704405e-07, "loss": 7.5126, "step": 15 }, { "epoch": 0.06049149338374291, "grad_norm": 77.42933654785156, "learning_rate": 4.716981132075472e-07, "loss": 7.5016, "step": 16 }, { "epoch": 0.06427221172022685, "grad_norm": 19.134775161743164, "learning_rate": 5.031446540880503e-07, "loss": 7.5656, "step": 17 }, { "epoch": 0.06805293005671077, "grad_norm": 43.67958450317383, "learning_rate": 5.345911949685535e-07, "loss": 7.4018, "step": 18 }, { "epoch": 0.07183364839319471, "grad_norm": 15.906808853149414, "learning_rate": 5.660377358490567e-07, "loss": 7.4636, "step": 19 }, { "epoch": 0.07561436672967864, "grad_norm": 33.41653060913086, "learning_rate": 5.974842767295598e-07, "loss": 7.4655, "step": 20 }, { "epoch": 0.07939508506616257, "grad_norm": 189.6513214111328, "learning_rate": 6.28930817610063e-07, "loss": 7.5318, "step": 21 }, { "epoch": 0.0831758034026465, "grad_norm": 69.53518676757812, "learning_rate": 6.603773584905661e-07, "loss": 7.4457, "step": 22 }, { "epoch": 0.08695652173913043, "grad_norm": 40.8270378112793, "learning_rate": 6.918238993710692e-07, "loss": 7.4464, "step": 23 }, { "epoch": 0.09073724007561437, "grad_norm": 27.530742645263672, "learning_rate": 7.232704402515723e-07, "loss": 7.4257, "step": 24 }, { "epoch": 0.0945179584120983, "grad_norm": 21.862537384033203, "learning_rate": 7.547169811320755e-07, "loss": 7.474, "step": 25 }, { "epoch": 0.09829867674858223, "grad_norm": 60.54854965209961, "learning_rate": 7.861635220125787e-07, "loss": 7.377, "step": 26 }, { "epoch": 0.10207939508506617, "grad_norm": 66.68376159667969, "learning_rate": 8.176100628930818e-07, "loss": 7.3615, "step": 27 }, { "epoch": 0.10586011342155009, "grad_norm": 16.780244827270508, "learning_rate": 8.490566037735849e-07, "loss": 7.3869, "step": 28 }, { "epoch": 0.10964083175803403, "grad_norm": 15.466861724853516, "learning_rate": 8.805031446540881e-07, "loss": 7.4175, "step": 29 }, { "epoch": 0.11342155009451796, "grad_norm": 23.016958236694336, "learning_rate": 9.119496855345912e-07, "loss": 7.3557, "step": 30 }, { "epoch": 0.11720226843100189, "grad_norm": 43.27973556518555, "learning_rate": 9.433962264150944e-07, "loss": 7.3206, "step": 31 }, { "epoch": 0.12098298676748583, "grad_norm": 21.37535285949707, "learning_rate": 9.748427672955975e-07, "loss": 7.3672, "step": 32 }, { "epoch": 0.12476370510396975, "grad_norm": 54.55345153808594, "learning_rate": 1.0062893081761007e-06, "loss": 7.2563, "step": 33 }, { "epoch": 0.1285444234404537, "grad_norm": 16.139854431152344, "learning_rate": 1.037735849056604e-06, "loss": 7.3057, "step": 34 }, { "epoch": 0.1323251417769376, "grad_norm": 44.60958480834961, "learning_rate": 1.069182389937107e-06, "loss": 7.2398, "step": 35 }, { "epoch": 0.13610586011342155, "grad_norm": 52.11935043334961, "learning_rate": 1.1006289308176102e-06, "loss": 7.261, "step": 36 }, { "epoch": 0.13988657844990549, "grad_norm": 44.808494567871094, "learning_rate": 1.1320754716981133e-06, "loss": 7.207, "step": 37 }, { "epoch": 0.14366729678638943, "grad_norm": 17.9429988861084, "learning_rate": 1.1635220125786164e-06, "loss": 7.2579, "step": 38 }, { "epoch": 0.14744801512287334, "grad_norm": 16.833396911621094, "learning_rate": 1.1949685534591195e-06, "loss": 7.1611, "step": 39 }, { "epoch": 0.15122873345935728, "grad_norm": 22.69009017944336, "learning_rate": 1.2264150943396227e-06, "loss": 7.1372, "step": 40 }, { "epoch": 0.15500945179584122, "grad_norm": 17.776609420776367, "learning_rate": 1.257861635220126e-06, "loss": 7.1193, "step": 41 }, { "epoch": 0.15879017013232513, "grad_norm": 37.8000602722168, "learning_rate": 1.289308176100629e-06, "loss": 7.1182, "step": 42 }, { "epoch": 0.16257088846880907, "grad_norm": 16.039527893066406, "learning_rate": 1.3207547169811322e-06, "loss": 7.0696, "step": 43 }, { "epoch": 0.166351606805293, "grad_norm": 14.535687446594238, "learning_rate": 1.3522012578616353e-06, "loss": 6.9287, "step": 44 }, { "epoch": 0.17013232514177692, "grad_norm": 30.682804107666016, "learning_rate": 1.3836477987421384e-06, "loss": 6.8904, "step": 45 }, { "epoch": 0.17391304347826086, "grad_norm": 28.5286808013916, "learning_rate": 1.4150943396226415e-06, "loss": 6.8815, "step": 46 }, { "epoch": 0.1776937618147448, "grad_norm": 22.324094772338867, "learning_rate": 1.4465408805031447e-06, "loss": 6.821, "step": 47 }, { "epoch": 0.18147448015122875, "grad_norm": 17.17375373840332, "learning_rate": 1.4779874213836478e-06, "loss": 6.7409, "step": 48 }, { "epoch": 0.18525519848771266, "grad_norm": 40.4192008972168, "learning_rate": 1.509433962264151e-06, "loss": 6.6866, "step": 49 }, { "epoch": 0.1890359168241966, "grad_norm": 167.84400939941406, "learning_rate": 1.5408805031446542e-06, "loss": 6.6029, "step": 50 }, { "epoch": 0.19281663516068054, "grad_norm": 38.26666259765625, "learning_rate": 1.5723270440251573e-06, "loss": 6.523, "step": 51 }, { "epoch": 0.19659735349716445, "grad_norm": 16.243309020996094, "learning_rate": 1.6037735849056604e-06, "loss": 6.5247, "step": 52 }, { "epoch": 0.2003780718336484, "grad_norm": 31.232200622558594, "learning_rate": 1.6352201257861635e-06, "loss": 6.443, "step": 53 }, { "epoch": 0.20415879017013233, "grad_norm": 21.760366439819336, "learning_rate": 1.6666666666666667e-06, "loss": 6.3139, "step": 54 }, { "epoch": 0.20793950850661624, "grad_norm": 33.58915710449219, "learning_rate": 1.6981132075471698e-06, "loss": 6.268, "step": 55 }, { "epoch": 0.21172022684310018, "grad_norm": 22.62858772277832, "learning_rate": 1.7295597484276729e-06, "loss": 6.2796, "step": 56 }, { "epoch": 0.21550094517958412, "grad_norm": 21.4224910736084, "learning_rate": 1.7610062893081762e-06, "loss": 6.2031, "step": 57 }, { "epoch": 0.21928166351606806, "grad_norm": 56.09198760986328, "learning_rate": 1.7924528301886793e-06, "loss": 6.047, "step": 58 }, { "epoch": 0.22306238185255198, "grad_norm": 22.593782424926758, "learning_rate": 1.8238993710691824e-06, "loss": 6.0111, "step": 59 }, { "epoch": 0.22684310018903592, "grad_norm": 13.674504280090332, "learning_rate": 1.8553459119496855e-06, "loss": 6.0051, "step": 60 }, { "epoch": 0.23062381852551986, "grad_norm": 14.918376922607422, "learning_rate": 1.8867924528301889e-06, "loss": 5.8901, "step": 61 }, { "epoch": 0.23440453686200377, "grad_norm": 17.77597427368164, "learning_rate": 1.918238993710692e-06, "loss": 5.7772, "step": 62 }, { "epoch": 0.2381852551984877, "grad_norm": 14.99068546295166, "learning_rate": 1.949685534591195e-06, "loss": 5.7429, "step": 63 }, { "epoch": 0.24196597353497165, "grad_norm": 13.07969856262207, "learning_rate": 1.981132075471698e-06, "loss": 5.6662, "step": 64 }, { "epoch": 0.24574669187145556, "grad_norm": 15.897534370422363, "learning_rate": 2.0125786163522013e-06, "loss": 5.5693, "step": 65 }, { "epoch": 0.2495274102079395, "grad_norm": 121.32555389404297, "learning_rate": 2.044025157232705e-06, "loss": 5.4538, "step": 66 }, { "epoch": 0.2533081285444234, "grad_norm": 21.261945724487305, "learning_rate": 2.075471698113208e-06, "loss": 5.3612, "step": 67 }, { "epoch": 0.2570888468809074, "grad_norm": 111.38444519042969, "learning_rate": 2.106918238993711e-06, "loss": 5.194, "step": 68 }, { "epoch": 0.2608695652173913, "grad_norm": 14.083466529846191, "learning_rate": 2.138364779874214e-06, "loss": 5.089, "step": 69 }, { "epoch": 0.2646502835538752, "grad_norm": 13.966374397277832, "learning_rate": 2.1698113207547173e-06, "loss": 5.0088, "step": 70 }, { "epoch": 0.2684310018903592, "grad_norm": 22.73938751220703, "learning_rate": 2.2012578616352204e-06, "loss": 4.9012, "step": 71 }, { "epoch": 0.2722117202268431, "grad_norm": 14.570304870605469, "learning_rate": 2.2327044025157235e-06, "loss": 4.8024, "step": 72 }, { "epoch": 0.27599243856332706, "grad_norm": 10.787749290466309, "learning_rate": 2.2641509433962266e-06, "loss": 4.7959, "step": 73 }, { "epoch": 0.27977315689981097, "grad_norm": 138.98147583007812, "learning_rate": 2.2955974842767298e-06, "loss": 4.6746, "step": 74 }, { "epoch": 0.2835538752362949, "grad_norm": 12.275903701782227, "learning_rate": 2.327044025157233e-06, "loss": 4.5545, "step": 75 }, { "epoch": 0.28733459357277885, "grad_norm": 11.80927848815918, "learning_rate": 2.358490566037736e-06, "loss": 4.4783, "step": 76 }, { "epoch": 0.29111531190926276, "grad_norm": 12.5546293258667, "learning_rate": 2.389937106918239e-06, "loss": 4.3771, "step": 77 }, { "epoch": 0.2948960302457467, "grad_norm": 10.4177827835083, "learning_rate": 2.421383647798742e-06, "loss": 4.268, "step": 78 }, { "epoch": 0.29867674858223064, "grad_norm": 21.497817993164062, "learning_rate": 2.4528301886792453e-06, "loss": 4.1602, "step": 79 }, { "epoch": 0.30245746691871456, "grad_norm": 10.733284950256348, "learning_rate": 2.4842767295597484e-06, "loss": 4.0907, "step": 80 }, { "epoch": 0.30623818525519847, "grad_norm": 9.1729097366333, "learning_rate": 2.515723270440252e-06, "loss": 4.0039, "step": 81 }, { "epoch": 0.31001890359168244, "grad_norm": 7.811710357666016, "learning_rate": 2.547169811320755e-06, "loss": 3.8463, "step": 82 }, { "epoch": 0.31379962192816635, "grad_norm": 8.925241470336914, "learning_rate": 2.578616352201258e-06, "loss": 3.7781, "step": 83 }, { "epoch": 0.31758034026465026, "grad_norm": 7.611026763916016, "learning_rate": 2.6100628930817613e-06, "loss": 3.6956, "step": 84 }, { "epoch": 0.32136105860113423, "grad_norm": 10.380526542663574, "learning_rate": 2.6415094339622644e-06, "loss": 3.5492, "step": 85 }, { "epoch": 0.32514177693761814, "grad_norm": 8.114707946777344, "learning_rate": 2.6729559748427675e-06, "loss": 3.4574, "step": 86 }, { "epoch": 0.32892249527410206, "grad_norm": 9.994078636169434, "learning_rate": 2.7044025157232706e-06, "loss": 3.2612, "step": 87 }, { "epoch": 0.332703213610586, "grad_norm": 7.271584987640381, "learning_rate": 2.7358490566037738e-06, "loss": 3.1956, "step": 88 }, { "epoch": 0.33648393194706994, "grad_norm": 7.082602500915527, "learning_rate": 2.767295597484277e-06, "loss": 3.0726, "step": 89 }, { "epoch": 0.34026465028355385, "grad_norm": 6.304625034332275, "learning_rate": 2.79874213836478e-06, "loss": 2.9696, "step": 90 }, { "epoch": 0.3440453686200378, "grad_norm": 5.38296365737915, "learning_rate": 2.830188679245283e-06, "loss": 2.8606, "step": 91 }, { "epoch": 0.34782608695652173, "grad_norm": 5.451972961425781, "learning_rate": 2.861635220125786e-06, "loss": 2.7711, "step": 92 }, { "epoch": 0.3516068052930057, "grad_norm": 5.667670726776123, "learning_rate": 2.8930817610062893e-06, "loss": 2.7252, "step": 93 }, { "epoch": 0.3553875236294896, "grad_norm": 8.441760063171387, "learning_rate": 2.9245283018867924e-06, "loss": 2.6017, "step": 94 }, { "epoch": 0.3591682419659735, "grad_norm": 4.929771423339844, "learning_rate": 2.9559748427672955e-06, "loss": 2.4397, "step": 95 }, { "epoch": 0.3629489603024575, "grad_norm": 4.556966304779053, "learning_rate": 2.987421383647799e-06, "loss": 2.3587, "step": 96 }, { "epoch": 0.3667296786389414, "grad_norm": 4.929013252258301, "learning_rate": 3.018867924528302e-06, "loss": 2.2931, "step": 97 }, { "epoch": 0.3705103969754253, "grad_norm": 4.905407905578613, "learning_rate": 3.0503144654088053e-06, "loss": 2.2011, "step": 98 }, { "epoch": 0.3742911153119093, "grad_norm": 4.2658562660217285, "learning_rate": 3.0817610062893084e-06, "loss": 2.0455, "step": 99 }, { "epoch": 0.3780718336483932, "grad_norm": 4.6085028648376465, "learning_rate": 3.1132075471698115e-06, "loss": 1.9807, "step": 100 }, { "epoch": 0.3818525519848771, "grad_norm": 4.2887773513793945, "learning_rate": 3.1446540880503146e-06, "loss": 1.8445, "step": 101 }, { "epoch": 0.3856332703213611, "grad_norm": 4.113661766052246, "learning_rate": 3.1761006289308178e-06, "loss": 1.7745, "step": 102 }, { "epoch": 0.389413988657845, "grad_norm": 3.933649778366089, "learning_rate": 3.207547169811321e-06, "loss": 1.68, "step": 103 }, { "epoch": 0.3931947069943289, "grad_norm": 3.656949520111084, "learning_rate": 3.238993710691824e-06, "loss": 1.6325, "step": 104 }, { "epoch": 0.39697542533081287, "grad_norm": 4.945619583129883, "learning_rate": 3.270440251572327e-06, "loss": 1.5492, "step": 105 }, { "epoch": 0.4007561436672968, "grad_norm": 3.693225145339966, "learning_rate": 3.30188679245283e-06, "loss": 1.4417, "step": 106 }, { "epoch": 0.4045368620037807, "grad_norm": 3.863574266433716, "learning_rate": 3.3333333333333333e-06, "loss": 1.402, "step": 107 }, { "epoch": 0.40831758034026466, "grad_norm": 3.174236536026001, "learning_rate": 3.3647798742138364e-06, "loss": 1.3154, "step": 108 }, { "epoch": 0.4120982986767486, "grad_norm": 3.3694076538085938, "learning_rate": 3.3962264150943395e-06, "loss": 1.2421, "step": 109 }, { "epoch": 0.4158790170132325, "grad_norm": 4.764991760253906, "learning_rate": 3.4276729559748427e-06, "loss": 1.1771, "step": 110 }, { "epoch": 0.41965973534971646, "grad_norm": 7.065610885620117, "learning_rate": 3.4591194968553458e-06, "loss": 1.1174, "step": 111 }, { "epoch": 0.42344045368620037, "grad_norm": 3.8380560874938965, "learning_rate": 3.4905660377358493e-06, "loss": 1.0765, "step": 112 }, { "epoch": 0.42722117202268434, "grad_norm": 4.32865571975708, "learning_rate": 3.5220125786163524e-06, "loss": 1.0041, "step": 113 }, { "epoch": 0.43100189035916825, "grad_norm": 2.5833635330200195, "learning_rate": 3.5534591194968555e-06, "loss": 1.0019, "step": 114 }, { "epoch": 0.43478260869565216, "grad_norm": 2.5374109745025635, "learning_rate": 3.5849056603773586e-06, "loss": 0.9325, "step": 115 }, { "epoch": 0.43856332703213613, "grad_norm": 2.5228700637817383, "learning_rate": 3.6163522012578618e-06, "loss": 0.9002, "step": 116 }, { "epoch": 0.44234404536862004, "grad_norm": 2.3162992000579834, "learning_rate": 3.647798742138365e-06, "loss": 0.8339, "step": 117 }, { "epoch": 0.44612476370510395, "grad_norm": 5.359051704406738, "learning_rate": 3.679245283018868e-06, "loss": 0.7762, "step": 118 }, { "epoch": 0.4499054820415879, "grad_norm": 2.14282488822937, "learning_rate": 3.710691823899371e-06, "loss": 0.7618, "step": 119 }, { "epoch": 0.45368620037807184, "grad_norm": 4.327908992767334, "learning_rate": 3.742138364779874e-06, "loss": 0.7086, "step": 120 }, { "epoch": 0.45746691871455575, "grad_norm": 1.910068392753601, "learning_rate": 3.7735849056603777e-06, "loss": 0.6768, "step": 121 }, { "epoch": 0.4612476370510397, "grad_norm": 2.157931089401245, "learning_rate": 3.8050314465408813e-06, "loss": 0.6522, "step": 122 }, { "epoch": 0.46502835538752363, "grad_norm": 2.8408241271972656, "learning_rate": 3.836477987421384e-06, "loss": 0.612, "step": 123 }, { "epoch": 0.46880907372400754, "grad_norm": 2.4224166870117188, "learning_rate": 3.8679245283018875e-06, "loss": 0.6153, "step": 124 }, { "epoch": 0.4725897920604915, "grad_norm": 2.827714681625366, "learning_rate": 3.89937106918239e-06, "loss": 0.5322, "step": 125 }, { "epoch": 0.4763705103969754, "grad_norm": 2.496584415435791, "learning_rate": 3.930817610062894e-06, "loss": 0.5237, "step": 126 }, { "epoch": 0.48015122873345933, "grad_norm": 1.7752296924591064, "learning_rate": 3.962264150943396e-06, "loss": 0.5096, "step": 127 }, { "epoch": 0.4839319470699433, "grad_norm": 1.72308349609375, "learning_rate": 3.9937106918239e-06, "loss": 0.4851, "step": 128 }, { "epoch": 0.4877126654064272, "grad_norm": 1.6874773502349854, "learning_rate": 4.025157232704403e-06, "loss": 0.4686, "step": 129 }, { "epoch": 0.4914933837429111, "grad_norm": 1.6524617671966553, "learning_rate": 4.056603773584906e-06, "loss": 0.4313, "step": 130 }, { "epoch": 0.4952741020793951, "grad_norm": 2.207883834838867, "learning_rate": 4.08805031446541e-06, "loss": 0.4363, "step": 131 }, { "epoch": 0.499054820415879, "grad_norm": 1.421894907951355, "learning_rate": 4.119496855345912e-06, "loss": 0.3946, "step": 132 }, { "epoch": 0.502835538752363, "grad_norm": 1.6512476205825806, "learning_rate": 4.150943396226416e-06, "loss": 0.3995, "step": 133 }, { "epoch": 0.5066162570888468, "grad_norm": 4.216320514678955, "learning_rate": 4.182389937106919e-06, "loss": 0.3511, "step": 134 }, { "epoch": 0.5103969754253308, "grad_norm": 1.8675678968429565, "learning_rate": 4.213836477987422e-06, "loss": 0.3483, "step": 135 }, { "epoch": 0.5141776937618148, "grad_norm": 2.213804006576538, "learning_rate": 4.245283018867925e-06, "loss": 0.3362, "step": 136 }, { "epoch": 0.5179584120982986, "grad_norm": 2.8736376762390137, "learning_rate": 4.276729559748428e-06, "loss": 0.3062, "step": 137 }, { "epoch": 0.5217391304347826, "grad_norm": 2.3165109157562256, "learning_rate": 4.308176100628931e-06, "loss": 0.2813, "step": 138 }, { "epoch": 0.5255198487712666, "grad_norm": 2.413364887237549, "learning_rate": 4.339622641509435e-06, "loss": 0.2461, "step": 139 }, { "epoch": 0.5293005671077504, "grad_norm": 1.9445260763168335, "learning_rate": 4.371069182389937e-06, "loss": 0.2029, "step": 140 }, { "epoch": 0.5330812854442344, "grad_norm": 3.132275342941284, "learning_rate": 4.402515723270441e-06, "loss": 0.175, "step": 141 }, { "epoch": 0.5368620037807184, "grad_norm": 2.3779661655426025, "learning_rate": 4.4339622641509435e-06, "loss": 0.1658, "step": 142 }, { "epoch": 0.5406427221172023, "grad_norm": 1.8241318464279175, "learning_rate": 4.465408805031447e-06, "loss": 0.1326, "step": 143 }, { "epoch": 0.5444234404536862, "grad_norm": 1.5344020128250122, "learning_rate": 4.49685534591195e-06, "loss": 0.1054, "step": 144 }, { "epoch": 0.5482041587901701, "grad_norm": 1.6879894733428955, "learning_rate": 4.528301886792453e-06, "loss": 0.114, "step": 145 }, { "epoch": 0.5519848771266541, "grad_norm": 1.603250503540039, "learning_rate": 4.559748427672957e-06, "loss": 0.0814, "step": 146 }, { "epoch": 0.555765595463138, "grad_norm": 1.6196143627166748, "learning_rate": 4.5911949685534595e-06, "loss": 0.072, "step": 147 }, { "epoch": 0.5595463137996219, "grad_norm": 1.2664083242416382, "learning_rate": 4.622641509433963e-06, "loss": 0.0623, "step": 148 }, { "epoch": 0.5633270321361059, "grad_norm": 1.2467018365859985, "learning_rate": 4.654088050314466e-06, "loss": 0.073, "step": 149 }, { "epoch": 0.5671077504725898, "grad_norm": 1.374241828918457, "learning_rate": 4.685534591194969e-06, "loss": 0.0665, "step": 150 }, { "epoch": 0.5708884688090737, "grad_norm": 1.0149943828582764, "learning_rate": 4.716981132075472e-06, "loss": 0.0461, "step": 151 }, { "epoch": 0.5746691871455577, "grad_norm": 0.880368173122406, "learning_rate": 4.7484276729559755e-06, "loss": 0.0362, "step": 152 }, { "epoch": 0.5784499054820416, "grad_norm": 2.355700731277466, "learning_rate": 4.779874213836478e-06, "loss": 0.0356, "step": 153 }, { "epoch": 0.5822306238185255, "grad_norm": 0.8400501012802124, "learning_rate": 4.811320754716982e-06, "loss": 0.0334, "step": 154 }, { "epoch": 0.5860113421550095, "grad_norm": 1.005251169204712, "learning_rate": 4.842767295597484e-06, "loss": 0.0275, "step": 155 }, { "epoch": 0.5897920604914934, "grad_norm": 1.0081993341445923, "learning_rate": 4.874213836477988e-06, "loss": 0.0225, "step": 156 }, { "epoch": 0.5935727788279773, "grad_norm": 0.5543925166130066, "learning_rate": 4.905660377358491e-06, "loss": 0.0217, "step": 157 }, { "epoch": 0.5973534971644613, "grad_norm": 0.9431848526000977, "learning_rate": 4.937106918238994e-06, "loss": 0.0221, "step": 158 }, { "epoch": 0.6011342155009451, "grad_norm": 0.4760805666446686, "learning_rate": 4.968553459119497e-06, "loss": 0.0159, "step": 159 }, { "epoch": 0.6049149338374291, "grad_norm": 0.6997100114822388, "learning_rate": 5e-06, "loss": 0.0275, "step": 160 }, { "epoch": 0.6086956521739131, "grad_norm": 0.7088273763656616, "learning_rate": 4.999993975370883e-06, "loss": 0.0155, "step": 161 }, { "epoch": 0.6124763705103969, "grad_norm": 0.5313673615455627, "learning_rate": 4.999975901512567e-06, "loss": 0.0152, "step": 162 }, { "epoch": 0.6162570888468809, "grad_norm": 1.7096068859100342, "learning_rate": 4.999945778512164e-06, "loss": 0.0155, "step": 163 }, { "epoch": 0.6200378071833649, "grad_norm": 0.6380202770233154, "learning_rate": 4.999903606514858e-06, "loss": 0.0167, "step": 164 }, { "epoch": 0.6238185255198487, "grad_norm": 0.9574909210205078, "learning_rate": 4.999849385723904e-06, "loss": 0.0182, "step": 165 }, { "epoch": 0.6275992438563327, "grad_norm": 0.5092853307723999, "learning_rate": 4.9997831164006315e-06, "loss": 0.0167, "step": 166 }, { "epoch": 0.6313799621928167, "grad_norm": 0.5826191306114197, "learning_rate": 4.999704798864439e-06, "loss": 0.011, "step": 167 }, { "epoch": 0.6351606805293005, "grad_norm": 2.9003238677978516, "learning_rate": 4.9996144334927924e-06, "loss": 0.0212, "step": 168 }, { "epoch": 0.6389413988657845, "grad_norm": 1.0899226665496826, "learning_rate": 4.9995120207212275e-06, "loss": 0.0159, "step": 169 }, { "epoch": 0.6427221172022685, "grad_norm": 0.6725714206695557, "learning_rate": 4.999397561043343e-06, "loss": 0.0155, "step": 170 }, { "epoch": 0.6465028355387523, "grad_norm": 0.6329910755157471, "learning_rate": 4.9992710550108e-06, "loss": 0.0183, "step": 171 }, { "epoch": 0.6502835538752363, "grad_norm": 2.138045072555542, "learning_rate": 4.999132503233321e-06, "loss": 0.0154, "step": 172 }, { "epoch": 0.6540642722117203, "grad_norm": 0.4029994606971741, "learning_rate": 4.998981906378684e-06, "loss": 0.0117, "step": 173 }, { "epoch": 0.6578449905482041, "grad_norm": 0.4679683446884155, "learning_rate": 4.998819265172721e-06, "loss": 0.0144, "step": 174 }, { "epoch": 0.6616257088846881, "grad_norm": 1.7001372575759888, "learning_rate": 4.9986445803993144e-06, "loss": 0.0118, "step": 175 }, { "epoch": 0.665406427221172, "grad_norm": 0.5954533815383911, "learning_rate": 4.998457852900393e-06, "loss": 0.0192, "step": 176 }, { "epoch": 0.6691871455576559, "grad_norm": 0.39165517687797546, "learning_rate": 4.998259083575929e-06, "loss": 0.0147, "step": 177 }, { "epoch": 0.6729678638941399, "grad_norm": 0.4655926525592804, "learning_rate": 4.99804827338393e-06, "loss": 0.0116, "step": 178 }, { "epoch": 0.6767485822306238, "grad_norm": 0.48431676626205444, "learning_rate": 4.997825423340439e-06, "loss": 0.0117, "step": 179 }, { "epoch": 0.6805293005671077, "grad_norm": 0.33966612815856934, "learning_rate": 4.997590534519528e-06, "loss": 0.0103, "step": 180 }, { "epoch": 0.6843100189035917, "grad_norm": 0.5036893486976624, "learning_rate": 4.99734360805329e-06, "loss": 0.0145, "step": 181 }, { "epoch": 0.6880907372400756, "grad_norm": 0.7529216408729553, "learning_rate": 4.997084645131837e-06, "loss": 0.0119, "step": 182 }, { "epoch": 0.6918714555765595, "grad_norm": 0.4617402255535126, "learning_rate": 4.996813647003296e-06, "loss": 0.0112, "step": 183 }, { "epoch": 0.6956521739130435, "grad_norm": 0.3071569502353668, "learning_rate": 4.9965306149737955e-06, "loss": 0.0118, "step": 184 }, { "epoch": 0.6994328922495274, "grad_norm": 1.0627433061599731, "learning_rate": 4.996235550407467e-06, "loss": 0.0189, "step": 185 }, { "epoch": 0.7032136105860114, "grad_norm": 0.5708284378051758, "learning_rate": 4.995928454726433e-06, "loss": 0.011, "step": 186 }, { "epoch": 0.7069943289224953, "grad_norm": 0.7609148621559143, "learning_rate": 4.9956093294108046e-06, "loss": 0.0185, "step": 187 }, { "epoch": 0.7107750472589792, "grad_norm": 0.9582763314247131, "learning_rate": 4.995278175998669e-06, "loss": 0.018, "step": 188 }, { "epoch": 0.7145557655954632, "grad_norm": 0.6773777604103088, "learning_rate": 4.994934996086091e-06, "loss": 0.0175, "step": 189 }, { "epoch": 0.718336483931947, "grad_norm": 0.27355310320854187, "learning_rate": 4.994579791327093e-06, "loss": 0.0081, "step": 190 }, { "epoch": 0.722117202268431, "grad_norm": 0.3438358008861542, "learning_rate": 4.994212563433658e-06, "loss": 0.0119, "step": 191 }, { "epoch": 0.725897920604915, "grad_norm": 0.44883400201797485, "learning_rate": 4.993833314175716e-06, "loss": 0.0138, "step": 192 }, { "epoch": 0.7296786389413988, "grad_norm": 1.1573532819747925, "learning_rate": 4.993442045381133e-06, "loss": 0.0121, "step": 193 }, { "epoch": 0.7334593572778828, "grad_norm": 0.3534741997718811, "learning_rate": 4.993038758935712e-06, "loss": 0.0107, "step": 194 }, { "epoch": 0.7372400756143668, "grad_norm": 0.32773202657699585, "learning_rate": 4.9926234567831716e-06, "loss": 0.0104, "step": 195 }, { "epoch": 0.7410207939508506, "grad_norm": 0.45407310128211975, "learning_rate": 4.9921961409251465e-06, "loss": 0.0179, "step": 196 }, { "epoch": 0.7448015122873346, "grad_norm": 0.3127530515193939, "learning_rate": 4.991756813421171e-06, "loss": 0.0094, "step": 197 }, { "epoch": 0.7485822306238186, "grad_norm": 0.2898798882961273, "learning_rate": 4.991305476388674e-06, "loss": 0.0093, "step": 198 }, { "epoch": 0.7523629489603024, "grad_norm": 0.23214803636074066, "learning_rate": 4.990842132002967e-06, "loss": 0.0094, "step": 199 }, { "epoch": 0.7561436672967864, "grad_norm": 0.3595448434352875, "learning_rate": 4.99036678249723e-06, "loss": 0.011, "step": 200 }, { "epoch": 0.7599243856332704, "grad_norm": 0.5961601734161377, "learning_rate": 4.98987943016251e-06, "loss": 0.0126, "step": 201 }, { "epoch": 0.7637051039697542, "grad_norm": 0.4877797067165375, "learning_rate": 4.9893800773476965e-06, "loss": 0.0116, "step": 202 }, { "epoch": 0.7674858223062382, "grad_norm": 0.37965238094329834, "learning_rate": 4.988868726459526e-06, "loss": 0.0146, "step": 203 }, { "epoch": 0.7712665406427222, "grad_norm": 0.303009569644928, "learning_rate": 4.988345379962556e-06, "loss": 0.0088, "step": 204 }, { "epoch": 0.775047258979206, "grad_norm": 0.2572024166584015, "learning_rate": 4.987810040379161e-06, "loss": 0.0096, "step": 205 }, { "epoch": 0.77882797731569, "grad_norm": 0.4183861017227173, "learning_rate": 4.987262710289519e-06, "loss": 0.0108, "step": 206 }, { "epoch": 0.782608695652174, "grad_norm": 0.3773617744445801, "learning_rate": 4.9867033923316e-06, "loss": 0.0132, "step": 207 }, { "epoch": 0.7863894139886578, "grad_norm": 0.31557539105415344, "learning_rate": 4.98613208920115e-06, "loss": 0.0126, "step": 208 }, { "epoch": 0.7901701323251418, "grad_norm": 0.32191208004951477, "learning_rate": 4.98554880365168e-06, "loss": 0.0097, "step": 209 }, { "epoch": 0.7939508506616257, "grad_norm": 0.4623849391937256, "learning_rate": 4.984953538494453e-06, "loss": 0.0089, "step": 210 }, { "epoch": 0.7977315689981096, "grad_norm": 0.32853055000305176, "learning_rate": 4.984346296598472e-06, "loss": 0.0083, "step": 211 }, { "epoch": 0.8015122873345936, "grad_norm": 0.4608067572116852, "learning_rate": 4.983727080890461e-06, "loss": 0.0075, "step": 212 }, { "epoch": 0.8052930056710775, "grad_norm": 0.33871349692344666, "learning_rate": 4.983095894354858e-06, "loss": 0.009, "step": 213 }, { "epoch": 0.8090737240075614, "grad_norm": 0.4419747591018677, "learning_rate": 4.982452740033793e-06, "loss": 0.0138, "step": 214 }, { "epoch": 0.8128544423440454, "grad_norm": 1.1355232000350952, "learning_rate": 4.981797621027079e-06, "loss": 0.0133, "step": 215 }, { "epoch": 0.8166351606805293, "grad_norm": 0.41319403052330017, "learning_rate": 4.981130540492197e-06, "loss": 0.0141, "step": 216 }, { "epoch": 0.8204158790170132, "grad_norm": 0.7314413785934448, "learning_rate": 4.9804515016442754e-06, "loss": 0.0088, "step": 217 }, { "epoch": 0.8241965973534972, "grad_norm": 0.3597504794597626, "learning_rate": 4.97976050775608e-06, "loss": 0.0092, "step": 218 }, { "epoch": 0.8279773156899811, "grad_norm": 0.2793259918689728, "learning_rate": 4.979057562157998e-06, "loss": 0.0088, "step": 219 }, { "epoch": 0.831758034026465, "grad_norm": 0.31024208664894104, "learning_rate": 4.978342668238018e-06, "loss": 0.0096, "step": 220 }, { "epoch": 0.8355387523629489, "grad_norm": 0.48921728134155273, "learning_rate": 4.977615829441716e-06, "loss": 0.0081, "step": 221 }, { "epoch": 0.8393194706994329, "grad_norm": 0.42166244983673096, "learning_rate": 4.97687704927224e-06, "loss": 0.0075, "step": 222 }, { "epoch": 0.8431001890359168, "grad_norm": 0.36283212900161743, "learning_rate": 4.97612633129029e-06, "loss": 0.0073, "step": 223 }, { "epoch": 0.8468809073724007, "grad_norm": 0.5259446501731873, "learning_rate": 4.975363679114104e-06, "loss": 0.0111, "step": 224 }, { "epoch": 0.8506616257088847, "grad_norm": 0.4186738133430481, "learning_rate": 4.974589096419441e-06, "loss": 0.0092, "step": 225 }, { "epoch": 0.8544423440453687, "grad_norm": 0.5796731114387512, "learning_rate": 4.973802586939558e-06, "loss": 0.0114, "step": 226 }, { "epoch": 0.8582230623818525, "grad_norm": 0.31079840660095215, "learning_rate": 4.973004154465198e-06, "loss": 0.0086, "step": 227 }, { "epoch": 0.8620037807183365, "grad_norm": 0.6040017604827881, "learning_rate": 4.9721938028445695e-06, "loss": 0.0113, "step": 228 }, { "epoch": 0.8657844990548205, "grad_norm": 0.5101784467697144, "learning_rate": 4.971371535983325e-06, "loss": 0.0103, "step": 229 }, { "epoch": 0.8695652173913043, "grad_norm": 0.6561850905418396, "learning_rate": 4.970537357844549e-06, "loss": 0.0111, "step": 230 }, { "epoch": 0.8733459357277883, "grad_norm": 0.6086797714233398, "learning_rate": 4.96969127244873e-06, "loss": 0.0096, "step": 231 }, { "epoch": 0.8771266540642723, "grad_norm": 0.45817309617996216, "learning_rate": 4.96883328387375e-06, "loss": 0.009, "step": 232 }, { "epoch": 0.8809073724007561, "grad_norm": 0.43313246965408325, "learning_rate": 4.9679633962548615e-06, "loss": 0.0094, "step": 233 }, { "epoch": 0.8846880907372401, "grad_norm": 0.46806737780570984, "learning_rate": 4.96708161378466e-06, "loss": 0.0078, "step": 234 }, { "epoch": 0.888468809073724, "grad_norm": 0.7021424174308777, "learning_rate": 4.96618794071308e-06, "loss": 0.0124, "step": 235 }, { "epoch": 0.8922495274102079, "grad_norm": 0.46001574397087097, "learning_rate": 4.965282381347357e-06, "loss": 0.0091, "step": 236 }, { "epoch": 0.8960302457466919, "grad_norm": 0.7153300046920776, "learning_rate": 4.964364940052021e-06, "loss": 0.0108, "step": 237 }, { "epoch": 0.8998109640831758, "grad_norm": 0.3781479299068451, "learning_rate": 4.963435621248865e-06, "loss": 0.0117, "step": 238 }, { "epoch": 0.9035916824196597, "grad_norm": 0.7988784909248352, "learning_rate": 4.962494429416931e-06, "loss": 0.0115, "step": 239 }, { "epoch": 0.9073724007561437, "grad_norm": 0.6492530703544617, "learning_rate": 4.9615413690924844e-06, "loss": 0.0132, "step": 240 }, { "epoch": 0.9111531190926276, "grad_norm": 0.4843565821647644, "learning_rate": 4.960576444868992e-06, "loss": 0.0104, "step": 241 }, { "epoch": 0.9149338374291115, "grad_norm": 0.46952396631240845, "learning_rate": 4.9595996613971046e-06, "loss": 0.0156, "step": 242 }, { "epoch": 0.9187145557655955, "grad_norm": 0.3070496618747711, "learning_rate": 4.958611023384626e-06, "loss": 0.0072, "step": 243 }, { "epoch": 0.9224952741020794, "grad_norm": 0.3375200629234314, "learning_rate": 4.9576105355965e-06, "loss": 0.0082, "step": 244 }, { "epoch": 0.9262759924385633, "grad_norm": 0.6639671325683594, "learning_rate": 4.956598202854778e-06, "loss": 0.0079, "step": 245 }, { "epoch": 0.9300567107750473, "grad_norm": 0.32887664437294006, "learning_rate": 4.955574030038608e-06, "loss": 0.0056, "step": 246 }, { "epoch": 0.9338374291115312, "grad_norm": 0.33802103996276855, "learning_rate": 4.954538022084197e-06, "loss": 0.0079, "step": 247 }, { "epoch": 0.9376181474480151, "grad_norm": 0.9058698415756226, "learning_rate": 4.953490183984795e-06, "loss": 0.0069, "step": 248 }, { "epoch": 0.941398865784499, "grad_norm": 0.7622438073158264, "learning_rate": 4.9524305207906705e-06, "loss": 0.0075, "step": 249 }, { "epoch": 0.945179584120983, "grad_norm": 0.4123559296131134, "learning_rate": 4.951359037609088e-06, "loss": 0.0078, "step": 250 }, { "epoch": 0.9489603024574669, "grad_norm": 0.5265935063362122, "learning_rate": 4.950275739604277e-06, "loss": 0.0088, "step": 251 }, { "epoch": 0.9527410207939508, "grad_norm": 0.25473034381866455, "learning_rate": 4.9491806319974115e-06, "loss": 0.0065, "step": 252 }, { "epoch": 0.9565217391304348, "grad_norm": 0.5168606042861938, "learning_rate": 4.948073720066587e-06, "loss": 0.0111, "step": 253 }, { "epoch": 0.9603024574669187, "grad_norm": 0.5597686171531677, "learning_rate": 4.94695500914679e-06, "loss": 0.0069, "step": 254 }, { "epoch": 0.9640831758034026, "grad_norm": 0.9759865999221802, "learning_rate": 4.945824504629875e-06, "loss": 0.0058, "step": 255 }, { "epoch": 0.9678638941398866, "grad_norm": 0.6512097716331482, "learning_rate": 4.944682211964538e-06, "loss": 0.0075, "step": 256 }, { "epoch": 0.9716446124763705, "grad_norm": 0.335421085357666, "learning_rate": 4.943528136656292e-06, "loss": 0.0058, "step": 257 }, { "epoch": 0.9754253308128544, "grad_norm": 0.4521774649620056, "learning_rate": 4.9423622842674366e-06, "loss": 0.008, "step": 258 }, { "epoch": 0.9792060491493384, "grad_norm": 0.6024395227432251, "learning_rate": 4.941184660417034e-06, "loss": 0.0137, "step": 259 }, { "epoch": 0.9829867674858223, "grad_norm": 0.656848132610321, "learning_rate": 4.939995270780883e-06, "loss": 0.0142, "step": 260 }, { "epoch": 0.9867674858223062, "grad_norm": 0.4404570162296295, "learning_rate": 4.938794121091488e-06, "loss": 0.012, "step": 261 }, { "epoch": 0.9905482041587902, "grad_norm": 0.26803073287010193, "learning_rate": 4.937581217138033e-06, "loss": 0.0045, "step": 262 }, { "epoch": 0.994328922495274, "grad_norm": 0.3339914381504059, "learning_rate": 4.936356564766358e-06, "loss": 0.0055, "step": 263 }, { "epoch": 0.998109640831758, "grad_norm": 0.28216224908828735, "learning_rate": 4.935120169878922e-06, "loss": 0.0051, "step": 264 }, { "epoch": 1.0, "grad_norm": 0.3504510521888733, "learning_rate": 4.933872038434782e-06, "loss": 0.0059, "step": 265 }, { "epoch": 1.003780718336484, "grad_norm": 0.6130300164222717, "learning_rate": 4.93261217644956e-06, "loss": 0.0096, "step": 266 }, { "epoch": 1.007561436672968, "grad_norm": 0.3444952070713043, "learning_rate": 4.931340589995419e-06, "loss": 0.0063, "step": 267 }, { "epoch": 1.011342155009452, "grad_norm": 0.6962162256240845, "learning_rate": 4.930057285201028e-06, "loss": 0.0103, "step": 268 }, { "epoch": 1.0151228733459357, "grad_norm": 0.6356163024902344, "learning_rate": 4.928762268251533e-06, "loss": 0.016, "step": 269 }, { "epoch": 1.0189035916824196, "grad_norm": 0.39561784267425537, "learning_rate": 4.9274555453885355e-06, "loss": 0.0068, "step": 270 }, { "epoch": 1.0226843100189036, "grad_norm": 0.4642280638217926, "learning_rate": 4.926137122910049e-06, "loss": 0.0092, "step": 271 }, { "epoch": 1.0264650283553876, "grad_norm": 0.2640933692455292, "learning_rate": 4.924807007170479e-06, "loss": 0.0052, "step": 272 }, { "epoch": 1.0302457466918715, "grad_norm": 0.32941174507141113, "learning_rate": 4.923465204580589e-06, "loss": 0.0044, "step": 273 }, { "epoch": 1.0340264650283555, "grad_norm": 0.42055749893188477, "learning_rate": 4.922111721607471e-06, "loss": 0.0067, "step": 274 }, { "epoch": 1.0378071833648392, "grad_norm": 0.5835608243942261, "learning_rate": 4.920746564774509e-06, "loss": 0.0112, "step": 275 }, { "epoch": 1.0415879017013232, "grad_norm": 0.43108245730400085, "learning_rate": 4.9193697406613555e-06, "loss": 0.0094, "step": 276 }, { "epoch": 1.0453686200378072, "grad_norm": 0.6540151834487915, "learning_rate": 4.917981255903893e-06, "loss": 0.01, "step": 277 }, { "epoch": 1.0491493383742911, "grad_norm": 0.4299629032611847, "learning_rate": 4.9165811171942065e-06, "loss": 0.0056, "step": 278 }, { "epoch": 1.052930056710775, "grad_norm": 0.6015660762786865, "learning_rate": 4.915169331280549e-06, "loss": 0.0073, "step": 279 }, { "epoch": 1.056710775047259, "grad_norm": 0.4806179702281952, "learning_rate": 4.9137459049673105e-06, "loss": 0.0102, "step": 280 }, { "epoch": 1.0604914933837428, "grad_norm": 0.5540146827697754, "learning_rate": 4.912310845114983e-06, "loss": 0.006, "step": 281 }, { "epoch": 1.0642722117202268, "grad_norm": 0.39366039633750916, "learning_rate": 4.910864158640129e-06, "loss": 0.0046, "step": 282 }, { "epoch": 1.0680529300567108, "grad_norm": 0.3995983302593231, "learning_rate": 4.909405852515348e-06, "loss": 0.0069, "step": 283 }, { "epoch": 1.0718336483931947, "grad_norm": 0.41662174463272095, "learning_rate": 4.907935933769243e-06, "loss": 0.005, "step": 284 }, { "epoch": 1.0756143667296787, "grad_norm": 0.6686727404594421, "learning_rate": 4.9064544094863854e-06, "loss": 0.0058, "step": 285 }, { "epoch": 1.0793950850661627, "grad_norm": 0.35762885212898254, "learning_rate": 4.904961286807285e-06, "loss": 0.0052, "step": 286 }, { "epoch": 1.0831758034026464, "grad_norm": 0.46720755100250244, "learning_rate": 4.903456572928348e-06, "loss": 0.0075, "step": 287 }, { "epoch": 1.0869565217391304, "grad_norm": 0.34636828303337097, "learning_rate": 4.90194027510185e-06, "loss": 0.0028, "step": 288 }, { "epoch": 1.0907372400756143, "grad_norm": 0.5153427124023438, "learning_rate": 4.900412400635895e-06, "loss": 0.0086, "step": 289 }, { "epoch": 1.0945179584120983, "grad_norm": 0.43430694937705994, "learning_rate": 4.898872956894386e-06, "loss": 0.0053, "step": 290 }, { "epoch": 1.0982986767485823, "grad_norm": 0.5609577894210815, "learning_rate": 4.897321951296985e-06, "loss": 0.0086, "step": 291 }, { "epoch": 1.1020793950850662, "grad_norm": 0.43855106830596924, "learning_rate": 4.8957593913190785e-06, "loss": 0.005, "step": 292 }, { "epoch": 1.10586011342155, "grad_norm": 0.6675010323524475, "learning_rate": 4.894185284491742e-06, "loss": 0.0065, "step": 293 }, { "epoch": 1.109640831758034, "grad_norm": 0.2379831224679947, "learning_rate": 4.892599638401704e-06, "loss": 0.004, "step": 294 }, { "epoch": 1.113421550094518, "grad_norm": 0.3421691656112671, "learning_rate": 4.891002460691306e-06, "loss": 0.007, "step": 295 }, { "epoch": 1.117202268431002, "grad_norm": 0.5493922829627991, "learning_rate": 4.889393759058474e-06, "loss": 0.0079, "step": 296 }, { "epoch": 1.1209829867674859, "grad_norm": 0.5068092346191406, "learning_rate": 4.887773541256669e-06, "loss": 0.0101, "step": 297 }, { "epoch": 1.1247637051039698, "grad_norm": 0.5580309629440308, "learning_rate": 4.886141815094864e-06, "loss": 0.0063, "step": 298 }, { "epoch": 1.1285444234404536, "grad_norm": 0.8997318148612976, "learning_rate": 4.884498588437491e-06, "loss": 0.0094, "step": 299 }, { "epoch": 1.1323251417769375, "grad_norm": 0.38125643134117126, "learning_rate": 4.882843869204417e-06, "loss": 0.0046, "step": 300 }, { "epoch": 1.1361058601134215, "grad_norm": 0.7842485308647156, "learning_rate": 4.8811776653708985e-06, "loss": 0.0074, "step": 301 }, { "epoch": 1.1398865784499055, "grad_norm": 0.5279122591018677, "learning_rate": 4.8794999849675416e-06, "loss": 0.0071, "step": 302 }, { "epoch": 1.1436672967863895, "grad_norm": 0.9195687174797058, "learning_rate": 4.877810836080269e-06, "loss": 0.0032, "step": 303 }, { "epoch": 1.1474480151228734, "grad_norm": 0.5626787543296814, "learning_rate": 4.876110226850278e-06, "loss": 0.0077, "step": 304 }, { "epoch": 1.1512287334593574, "grad_norm": 0.3872135281562805, "learning_rate": 4.874398165473999e-06, "loss": 0.0044, "step": 305 }, { "epoch": 1.1550094517958411, "grad_norm": 0.7638955116271973, "learning_rate": 4.872674660203062e-06, "loss": 0.0051, "step": 306 }, { "epoch": 1.158790170132325, "grad_norm": 0.18408967554569244, "learning_rate": 4.870939719344248e-06, "loss": 0.0027, "step": 307 }, { "epoch": 1.162570888468809, "grad_norm": 0.2827281355857849, "learning_rate": 4.869193351259459e-06, "loss": 0.0026, "step": 308 }, { "epoch": 1.166351606805293, "grad_norm": 0.4914953410625458, "learning_rate": 4.867435564365671e-06, "loss": 0.0044, "step": 309 }, { "epoch": 1.170132325141777, "grad_norm": 0.2728334963321686, "learning_rate": 4.8656663671348954e-06, "loss": 0.0029, "step": 310 }, { "epoch": 1.1739130434782608, "grad_norm": 0.3587643504142761, "learning_rate": 4.863885768094138e-06, "loss": 0.0031, "step": 311 }, { "epoch": 1.1776937618147447, "grad_norm": 0.6220464110374451, "learning_rate": 4.862093775825356e-06, "loss": 0.0056, "step": 312 }, { "epoch": 1.1814744801512287, "grad_norm": 0.6655571460723877, "learning_rate": 4.860290398965423e-06, "loss": 0.0071, "step": 313 }, { "epoch": 1.1852551984877127, "grad_norm": 0.5949417948722839, "learning_rate": 4.858475646206079e-06, "loss": 0.0092, "step": 314 }, { "epoch": 1.1890359168241966, "grad_norm": 1.0792511701583862, "learning_rate": 4.856649526293894e-06, "loss": 0.0052, "step": 315 }, { "epoch": 1.1928166351606806, "grad_norm": 0.5649266839027405, "learning_rate": 4.854812048030225e-06, "loss": 0.0073, "step": 316 }, { "epoch": 1.1965973534971646, "grad_norm": 0.374542236328125, "learning_rate": 4.852963220271171e-06, "loss": 0.0038, "step": 317 }, { "epoch": 1.2003780718336483, "grad_norm": 0.5689668655395508, "learning_rate": 4.851103051927533e-06, "loss": 0.0097, "step": 318 }, { "epoch": 1.2041587901701323, "grad_norm": 0.5911904573440552, "learning_rate": 4.849231551964771e-06, "loss": 0.0069, "step": 319 }, { "epoch": 1.2079395085066162, "grad_norm": 0.2607024312019348, "learning_rate": 4.84734872940296e-06, "loss": 0.0052, "step": 320 }, { "epoch": 1.2117202268431002, "grad_norm": 0.3516089916229248, "learning_rate": 4.845454593316747e-06, "loss": 0.0052, "step": 321 }, { "epoch": 1.2155009451795842, "grad_norm": 0.5216485261917114, "learning_rate": 4.843549152835303e-06, "loss": 0.0052, "step": 322 }, { "epoch": 1.2192816635160681, "grad_norm": 0.34397995471954346, "learning_rate": 4.841632417142288e-06, "loss": 0.0054, "step": 323 }, { "epoch": 1.223062381852552, "grad_norm": 0.42554324865341187, "learning_rate": 4.839704395475798e-06, "loss": 0.0057, "step": 324 }, { "epoch": 1.2268431001890359, "grad_norm": 0.4021840989589691, "learning_rate": 4.837765097128326e-06, "loss": 0.0068, "step": 325 }, { "epoch": 1.2306238185255198, "grad_norm": 0.27495795488357544, "learning_rate": 4.8358145314467165e-06, "loss": 0.0044, "step": 326 }, { "epoch": 1.2344045368620038, "grad_norm": 0.22830204665660858, "learning_rate": 4.833852707832114e-06, "loss": 0.0027, "step": 327 }, { "epoch": 1.2381852551984878, "grad_norm": 0.5374839901924133, "learning_rate": 4.831879635739929e-06, "loss": 0.004, "step": 328 }, { "epoch": 1.2419659735349717, "grad_norm": 0.3192364275455475, "learning_rate": 4.829895324679782e-06, "loss": 0.0044, "step": 329 }, { "epoch": 1.2457466918714555, "grad_norm": 0.41506636142730713, "learning_rate": 4.827899784215463e-06, "loss": 0.003, "step": 330 }, { "epoch": 1.2495274102079394, "grad_norm": 0.3320922255516052, "learning_rate": 4.825893023964886e-06, "loss": 0.0038, "step": 331 }, { "epoch": 1.2533081285444234, "grad_norm": 0.2081664651632309, "learning_rate": 4.823875053600041e-06, "loss": 0.0025, "step": 332 }, { "epoch": 1.2570888468809074, "grad_norm": 0.8778406977653503, "learning_rate": 4.821845882846945e-06, "loss": 0.0059, "step": 333 }, { "epoch": 1.2608695652173914, "grad_norm": 2.7111332416534424, "learning_rate": 4.819805521485599e-06, "loss": 0.0052, "step": 334 }, { "epoch": 1.264650283553875, "grad_norm": 0.4180721342563629, "learning_rate": 4.81775397934994e-06, "loss": 0.0046, "step": 335 }, { "epoch": 1.268431001890359, "grad_norm": 0.49803799390792847, "learning_rate": 4.815691266327791e-06, "loss": 0.0068, "step": 336 }, { "epoch": 1.272211720226843, "grad_norm": 0.32105717062950134, "learning_rate": 4.813617392360819e-06, "loss": 0.0035, "step": 337 }, { "epoch": 1.275992438563327, "grad_norm": 0.18351660668849945, "learning_rate": 4.8115323674444795e-06, "loss": 0.0015, "step": 338 }, { "epoch": 1.279773156899811, "grad_norm": 0.1784365177154541, "learning_rate": 4.809436201627974e-06, "loss": 0.0028, "step": 339 }, { "epoch": 1.283553875236295, "grad_norm": 0.44830846786499023, "learning_rate": 4.807328905014201e-06, "loss": 0.0046, "step": 340 }, { "epoch": 1.287334593572779, "grad_norm": 0.29059353470802307, "learning_rate": 4.805210487759704e-06, "loss": 0.0036, "step": 341 }, { "epoch": 1.2911153119092629, "grad_norm": 0.8145375847816467, "learning_rate": 4.8030809600746255e-06, "loss": 0.0033, "step": 342 }, { "epoch": 1.2948960302457466, "grad_norm": 0.27350112795829773, "learning_rate": 4.800940332222656e-06, "loss": 0.0042, "step": 343 }, { "epoch": 1.2986767485822306, "grad_norm": 0.13680307567119598, "learning_rate": 4.798788614520991e-06, "loss": 0.0011, "step": 344 }, { "epoch": 1.3024574669187146, "grad_norm": 0.5477114915847778, "learning_rate": 4.796625817340266e-06, "loss": 0.0094, "step": 345 }, { "epoch": 1.3062381852551985, "grad_norm": 0.4699587821960449, "learning_rate": 4.794451951104525e-06, "loss": 0.0042, "step": 346 }, { "epoch": 1.3100189035916825, "grad_norm": 0.38969650864601135, "learning_rate": 4.792267026291157e-06, "loss": 0.0063, "step": 347 }, { "epoch": 1.3137996219281662, "grad_norm": 1.8919707536697388, "learning_rate": 4.790071053430852e-06, "loss": 0.0021, "step": 348 }, { "epoch": 1.3175803402646502, "grad_norm": 0.4610402286052704, "learning_rate": 4.7878640431075466e-06, "loss": 0.0045, "step": 349 }, { "epoch": 1.3213610586011342, "grad_norm": 0.7579343914985657, "learning_rate": 4.785646005958377e-06, "loss": 0.0074, "step": 350 }, { "epoch": 1.3251417769376181, "grad_norm": 0.3805699348449707, "learning_rate": 4.783416952673623e-06, "loss": 0.0042, "step": 351 }, { "epoch": 1.3289224952741021, "grad_norm": 4.160273551940918, "learning_rate": 4.7811768939966605e-06, "loss": 0.0046, "step": 352 }, { "epoch": 1.332703213610586, "grad_norm": 0.2657456398010254, "learning_rate": 4.778925840723909e-06, "loss": 0.0024, "step": 353 }, { "epoch": 1.33648393194707, "grad_norm": 0.6165263056755066, "learning_rate": 4.776663803704775e-06, "loss": 0.0067, "step": 354 }, { "epoch": 1.3402646502835538, "grad_norm": 0.38582417368888855, "learning_rate": 4.774390793841607e-06, "loss": 0.0027, "step": 355 }, { "epoch": 1.3440453686200378, "grad_norm": 0.5040014982223511, "learning_rate": 4.772106822089639e-06, "loss": 0.0071, "step": 356 }, { "epoch": 1.3478260869565217, "grad_norm": 0.8076869249343872, "learning_rate": 4.769811899456935e-06, "loss": 0.0075, "step": 357 }, { "epoch": 1.3516068052930057, "grad_norm": 0.29268744587898254, "learning_rate": 4.767506037004344e-06, "loss": 0.003, "step": 358 }, { "epoch": 1.3553875236294897, "grad_norm": 0.5552521347999573, "learning_rate": 4.765189245845436e-06, "loss": 0.0027, "step": 359 }, { "epoch": 1.3591682419659734, "grad_norm": 0.27705779671669006, "learning_rate": 4.762861537146458e-06, "loss": 0.0025, "step": 360 }, { "epoch": 1.3629489603024574, "grad_norm": 0.46618857979774475, "learning_rate": 4.760522922126276e-06, "loss": 0.0048, "step": 361 }, { "epoch": 1.3667296786389413, "grad_norm": 0.3646470606327057, "learning_rate": 4.75817341205632e-06, "loss": 0.0065, "step": 362 }, { "epoch": 1.3705103969754253, "grad_norm": 0.3433074355125427, "learning_rate": 4.755813018260532e-06, "loss": 0.0037, "step": 363 }, { "epoch": 1.3742911153119093, "grad_norm": 0.3142302632331848, "learning_rate": 4.753441752115309e-06, "loss": 0.0035, "step": 364 }, { "epoch": 1.3780718336483933, "grad_norm": 0.5433574318885803, "learning_rate": 4.75105962504945e-06, "loss": 0.0059, "step": 365 }, { "epoch": 1.3818525519848772, "grad_norm": 1.089618444442749, "learning_rate": 4.748666648544101e-06, "loss": 0.0032, "step": 366 }, { "epoch": 1.3856332703213612, "grad_norm": 0.3445349931716919, "learning_rate": 4.7462628341327e-06, "loss": 0.0062, "step": 367 }, { "epoch": 1.389413988657845, "grad_norm": 0.5173162817955017, "learning_rate": 4.743848193400918e-06, "loss": 0.0062, "step": 368 }, { "epoch": 1.393194706994329, "grad_norm": 4.956406116485596, "learning_rate": 4.741422737986606e-06, "loss": 0.0074, "step": 369 }, { "epoch": 1.3969754253308129, "grad_norm": 0.9225440621376038, "learning_rate": 4.73898647957974e-06, "loss": 0.0108, "step": 370 }, { "epoch": 1.4007561436672968, "grad_norm": 0.33078140020370483, "learning_rate": 4.736539429922365e-06, "loss": 0.0031, "step": 371 }, { "epoch": 1.4045368620037806, "grad_norm": 0.23648452758789062, "learning_rate": 4.734081600808531e-06, "loss": 0.0017, "step": 372 }, { "epoch": 1.4083175803402646, "grad_norm": 0.5308226943016052, "learning_rate": 4.731613004084247e-06, "loss": 0.0056, "step": 373 }, { "epoch": 1.4120982986767485, "grad_norm": 0.39204317331314087, "learning_rate": 4.7291336516474165e-06, "loss": 0.0032, "step": 374 }, { "epoch": 1.4158790170132325, "grad_norm": 0.5415582656860352, "learning_rate": 4.726643555447782e-06, "loss": 0.0025, "step": 375 }, { "epoch": 1.4196597353497165, "grad_norm": 0.6073048710823059, "learning_rate": 4.724142727486869e-06, "loss": 0.0046, "step": 376 }, { "epoch": 1.4234404536862004, "grad_norm": 0.3404282331466675, "learning_rate": 4.721631179817925e-06, "loss": 0.0032, "step": 377 }, { "epoch": 1.4272211720226844, "grad_norm": 0.3819228708744049, "learning_rate": 4.719108924545866e-06, "loss": 0.0042, "step": 378 }, { "epoch": 1.4310018903591684, "grad_norm": 0.2740926444530487, "learning_rate": 4.716575973827214e-06, "loss": 0.003, "step": 379 }, { "epoch": 1.434782608695652, "grad_norm": 0.32899898290634155, "learning_rate": 4.714032339870041e-06, "loss": 0.0025, "step": 380 }, { "epoch": 1.438563327032136, "grad_norm": 0.6806893944740295, "learning_rate": 4.711478034933904e-06, "loss": 0.0081, "step": 381 }, { "epoch": 1.44234404536862, "grad_norm": 0.3089682459831238, "learning_rate": 4.708913071329798e-06, "loss": 0.0035, "step": 382 }, { "epoch": 1.446124763705104, "grad_norm": 0.44838741421699524, "learning_rate": 4.706337461420087e-06, "loss": 0.0045, "step": 383 }, { "epoch": 1.449905482041588, "grad_norm": 0.6625654697418213, "learning_rate": 4.703751217618444e-06, "loss": 0.0099, "step": 384 }, { "epoch": 1.4536862003780717, "grad_norm": 0.3470344543457031, "learning_rate": 4.7011543523898e-06, "loss": 0.0034, "step": 385 }, { "epoch": 1.4574669187145557, "grad_norm": 0.31737810373306274, "learning_rate": 4.698546878250272e-06, "loss": 0.0029, "step": 386 }, { "epoch": 1.4612476370510397, "grad_norm": 0.6312741637229919, "learning_rate": 4.695928807767113e-06, "loss": 0.0057, "step": 387 }, { "epoch": 1.4650283553875236, "grad_norm": 0.6768957376480103, "learning_rate": 4.693300153558647e-06, "loss": 0.0058, "step": 388 }, { "epoch": 1.4688090737240076, "grad_norm": 0.5521048307418823, "learning_rate": 4.690660928294204e-06, "loss": 0.0061, "step": 389 }, { "epoch": 1.4725897920604916, "grad_norm": 0.888009250164032, "learning_rate": 4.68801114469407e-06, "loss": 0.0102, "step": 390 }, { "epoch": 1.4763705103969755, "grad_norm": 0.3135982155799866, "learning_rate": 4.685350815529415e-06, "loss": 0.0021, "step": 391 }, { "epoch": 1.4801512287334593, "grad_norm": 0.6620728969573975, "learning_rate": 4.682679953622234e-06, "loss": 0.0028, "step": 392 }, { "epoch": 1.4839319470699432, "grad_norm": 0.631092369556427, "learning_rate": 4.679998571845293e-06, "loss": 0.0067, "step": 393 }, { "epoch": 1.4877126654064272, "grad_norm": 0.2744283974170685, "learning_rate": 4.677306683122054e-06, "loss": 0.0031, "step": 394 }, { "epoch": 1.4914933837429112, "grad_norm": 6.616799354553223, "learning_rate": 4.674604300426621e-06, "loss": 0.0061, "step": 395 }, { "epoch": 1.4952741020793952, "grad_norm": 0.8005297183990479, "learning_rate": 4.671891436783679e-06, "loss": 0.0129, "step": 396 }, { "epoch": 1.499054820415879, "grad_norm": 0.5362093448638916, "learning_rate": 4.669168105268425e-06, "loss": 0.005, "step": 397 }, { "epoch": 1.5028355387523629, "grad_norm": 0.28220292925834656, "learning_rate": 4.666434319006508e-06, "loss": 0.0027, "step": 398 }, { "epoch": 1.5066162570888468, "grad_norm": 0.5086351037025452, "learning_rate": 4.663690091173968e-06, "loss": 0.0044, "step": 399 }, { "epoch": 1.5103969754253308, "grad_norm": 0.29584789276123047, "learning_rate": 4.660935434997168e-06, "loss": 0.0036, "step": 400 }, { "epoch": 1.5141776937618148, "grad_norm": 0.41253986954689026, "learning_rate": 4.658170363752734e-06, "loss": 0.0056, "step": 401 }, { "epoch": 1.5179584120982987, "grad_norm": 0.3914569020271301, "learning_rate": 4.655394890767489e-06, "loss": 0.0036, "step": 402 }, { "epoch": 1.5217391304347827, "grad_norm": 0.4907969832420349, "learning_rate": 4.652609029418389e-06, "loss": 0.0041, "step": 403 }, { "epoch": 1.5255198487712667, "grad_norm": 1.597739815711975, "learning_rate": 4.649812793132459e-06, "loss": 0.0042, "step": 404 }, { "epoch": 1.5293005671077504, "grad_norm": 0.39329442381858826, "learning_rate": 4.647006195386728e-06, "loss": 0.004, "step": 405 }, { "epoch": 1.5330812854442344, "grad_norm": 0.5135226249694824, "learning_rate": 4.644189249708167e-06, "loss": 0.0048, "step": 406 }, { "epoch": 1.5368620037807184, "grad_norm": 0.24599513411521912, "learning_rate": 4.641361969673614e-06, "loss": 0.002, "step": 407 }, { "epoch": 1.5406427221172023, "grad_norm": 0.7556615471839905, "learning_rate": 4.638524368909723e-06, "loss": 0.006, "step": 408 }, { "epoch": 1.544423440453686, "grad_norm": 0.47028326988220215, "learning_rate": 4.635676461092886e-06, "loss": 0.0027, "step": 409 }, { "epoch": 1.54820415879017, "grad_norm": 0.28152233362197876, "learning_rate": 4.632818259949175e-06, "loss": 0.0028, "step": 410 }, { "epoch": 1.551984877126654, "grad_norm": 0.2820948660373688, "learning_rate": 4.6299497792542714e-06, "loss": 0.0019, "step": 411 }, { "epoch": 1.555765595463138, "grad_norm": 0.25290393829345703, "learning_rate": 4.627071032833401e-06, "loss": 0.0026, "step": 412 }, { "epoch": 1.559546313799622, "grad_norm": 0.6501352190971375, "learning_rate": 4.624182034561266e-06, "loss": 0.0036, "step": 413 }, { "epoch": 1.563327032136106, "grad_norm": 0.5022422671318054, "learning_rate": 4.621282798361983e-06, "loss": 0.002, "step": 414 }, { "epoch": 1.5671077504725899, "grad_norm": 0.29338476061820984, "learning_rate": 4.618373338209008e-06, "loss": 0.0033, "step": 415 }, { "epoch": 1.5708884688090738, "grad_norm": 1.2304599285125732, "learning_rate": 4.615453668125077e-06, "loss": 0.0032, "step": 416 }, { "epoch": 1.5746691871455578, "grad_norm": 0.30708685517311096, "learning_rate": 4.612523802182134e-06, "loss": 0.0035, "step": 417 }, { "epoch": 1.5784499054820416, "grad_norm": 0.2716330289840698, "learning_rate": 4.609583754501263e-06, "loss": 0.0021, "step": 418 }, { "epoch": 1.5822306238185255, "grad_norm": 0.4459162652492523, "learning_rate": 4.6066335392526215e-06, "loss": 0.003, "step": 419 }, { "epoch": 1.5860113421550095, "grad_norm": 0.33895647525787354, "learning_rate": 4.603673170655372e-06, "loss": 0.0029, "step": 420 }, { "epoch": 1.5897920604914932, "grad_norm": 0.562910795211792, "learning_rate": 4.600702662977611e-06, "loss": 0.0076, "step": 421 }, { "epoch": 1.5935727788279772, "grad_norm": 0.3354416787624359, "learning_rate": 4.597722030536306e-06, "loss": 0.0028, "step": 422 }, { "epoch": 1.5973534971644612, "grad_norm": 0.21339942514896393, "learning_rate": 4.5947312876972215e-06, "loss": 0.0015, "step": 423 }, { "epoch": 1.6011342155009451, "grad_norm": 0.2937569320201874, "learning_rate": 4.591730448874849e-06, "loss": 0.0026, "step": 424 }, { "epoch": 1.6049149338374291, "grad_norm": 0.7903040051460266, "learning_rate": 4.588719528532342e-06, "loss": 0.0024, "step": 425 }, { "epoch": 1.608695652173913, "grad_norm": 0.30319225788116455, "learning_rate": 4.585698541181443e-06, "loss": 0.0037, "step": 426 }, { "epoch": 1.612476370510397, "grad_norm": 0.4225195348262787, "learning_rate": 4.582667501382414e-06, "loss": 0.0043, "step": 427 }, { "epoch": 1.616257088846881, "grad_norm": 1.0661922693252563, "learning_rate": 4.579626423743969e-06, "loss": 0.0052, "step": 428 }, { "epoch": 1.620037807183365, "grad_norm": 0.3146645426750183, "learning_rate": 4.576575322923199e-06, "loss": 0.0021, "step": 429 }, { "epoch": 1.6238185255198487, "grad_norm": 0.2542162537574768, "learning_rate": 4.573514213625505e-06, "loss": 0.0019, "step": 430 }, { "epoch": 1.6275992438563327, "grad_norm": 0.9462196826934814, "learning_rate": 4.570443110604525e-06, "loss": 0.0066, "step": 431 }, { "epoch": 1.6313799621928167, "grad_norm": 0.2747437059879303, "learning_rate": 4.567362028662066e-06, "loss": 0.0017, "step": 432 }, { "epoch": 1.6351606805293004, "grad_norm": 0.3337705731391907, "learning_rate": 4.564270982648026e-06, "loss": 0.0015, "step": 433 }, { "epoch": 1.6389413988657844, "grad_norm": 0.6268848180770874, "learning_rate": 4.561169987460332e-06, "loss": 0.0035, "step": 434 }, { "epoch": 1.6427221172022684, "grad_norm": 0.6272996664047241, "learning_rate": 4.55805905804486e-06, "loss": 0.0085, "step": 435 }, { "epoch": 1.6465028355387523, "grad_norm": 0.537577211856842, "learning_rate": 4.554938209395366e-06, "loss": 0.0103, "step": 436 }, { "epoch": 1.6502835538752363, "grad_norm": 0.6211628913879395, "learning_rate": 4.551807456553416e-06, "loss": 0.0043, "step": 437 }, { "epoch": 1.6540642722117203, "grad_norm": 0.3048252761363983, "learning_rate": 4.548666814608309e-06, "loss": 0.0025, "step": 438 }, { "epoch": 1.6578449905482042, "grad_norm": 0.5037123560905457, "learning_rate": 4.545516298697006e-06, "loss": 0.0061, "step": 439 }, { "epoch": 1.6616257088846882, "grad_norm": 0.918006420135498, "learning_rate": 4.542355924004061e-06, "loss": 0.0046, "step": 440 }, { "epoch": 1.6654064272211722, "grad_norm": 0.9184623956680298, "learning_rate": 4.5391857057615426e-06, "loss": 0.009, "step": 441 }, { "epoch": 1.669187145557656, "grad_norm": 0.5904452204704285, "learning_rate": 4.53600565924896e-06, "loss": 0.0042, "step": 442 }, { "epoch": 1.6729678638941399, "grad_norm": 0.5156521201133728, "learning_rate": 4.532815799793196e-06, "loss": 0.0062, "step": 443 }, { "epoch": 1.6767485822306238, "grad_norm": 0.18715080618858337, "learning_rate": 4.529616142768424e-06, "loss": 0.0019, "step": 444 }, { "epoch": 1.6805293005671076, "grad_norm": 0.3885091245174408, "learning_rate": 4.526406703596043e-06, "loss": 0.0047, "step": 445 }, { "epoch": 1.6843100189035916, "grad_norm": 0.17866213619709015, "learning_rate": 4.523187497744599e-06, "loss": 0.0015, "step": 446 }, { "epoch": 1.6880907372400755, "grad_norm": 0.3022480309009552, "learning_rate": 4.519958540729708e-06, "loss": 0.0026, "step": 447 }, { "epoch": 1.6918714555765595, "grad_norm": 0.7824174761772156, "learning_rate": 4.516719848113983e-06, "loss": 0.004, "step": 448 }, { "epoch": 1.6956521739130435, "grad_norm": 0.4345873296260834, "learning_rate": 4.5134714355069635e-06, "loss": 0.0082, "step": 449 }, { "epoch": 1.6994328922495274, "grad_norm": 0.38643717765808105, "learning_rate": 4.510213318565034e-06, "loss": 0.003, "step": 450 }, { "epoch": 1.7032136105860114, "grad_norm": 1.2097886800765991, "learning_rate": 4.506945512991352e-06, "loss": 0.0065, "step": 451 }, { "epoch": 1.7069943289224954, "grad_norm": 0.32429248094558716, "learning_rate": 4.503668034535768e-06, "loss": 0.0033, "step": 452 }, { "epoch": 1.7107750472589793, "grad_norm": 0.182096928358078, "learning_rate": 4.500380898994761e-06, "loss": 0.002, "step": 453 }, { "epoch": 1.7145557655954633, "grad_norm": 0.1907053142786026, "learning_rate": 4.4970841222113435e-06, "loss": 0.0012, "step": 454 }, { "epoch": 1.718336483931947, "grad_norm": 0.2800076901912689, "learning_rate": 4.4937777200750045e-06, "loss": 0.0017, "step": 455 }, { "epoch": 1.722117202268431, "grad_norm": 0.2724849283695221, "learning_rate": 4.490461708521621e-06, "loss": 0.0094, "step": 456 }, { "epoch": 1.725897920604915, "grad_norm": 0.2867650091648102, "learning_rate": 4.487136103533383e-06, "loss": 0.0028, "step": 457 }, { "epoch": 1.7296786389413987, "grad_norm": 0.2651146650314331, "learning_rate": 4.483800921138722e-06, "loss": 0.0024, "step": 458 }, { "epoch": 1.7334593572778827, "grad_norm": 0.2688697278499603, "learning_rate": 4.4804561774122275e-06, "loss": 0.0024, "step": 459 }, { "epoch": 1.7372400756143667, "grad_norm": 0.28717854619026184, "learning_rate": 4.4771018884745705e-06, "loss": 0.0025, "step": 460 }, { "epoch": 1.7410207939508506, "grad_norm": 0.6200689673423767, "learning_rate": 4.473738070492428e-06, "loss": 0.0042, "step": 461 }, { "epoch": 1.7448015122873346, "grad_norm": 0.41647911071777344, "learning_rate": 4.470364739678407e-06, "loss": 0.0043, "step": 462 }, { "epoch": 1.7485822306238186, "grad_norm": 0.6958944797515869, "learning_rate": 4.466981912290959e-06, "loss": 0.0031, "step": 463 }, { "epoch": 1.7523629489603025, "grad_norm": 0.4741610586643219, "learning_rate": 4.46358960463431e-06, "loss": 0.0053, "step": 464 }, { "epoch": 1.7561436672967865, "grad_norm": 0.38444384932518005, "learning_rate": 4.4601878330583744e-06, "loss": 0.002, "step": 465 }, { "epoch": 1.7599243856332705, "grad_norm": 0.18333545327186584, "learning_rate": 4.456776613958683e-06, "loss": 0.0017, "step": 466 }, { "epoch": 1.7637051039697542, "grad_norm": 0.2698777914047241, "learning_rate": 4.453355963776301e-06, "loss": 0.0016, "step": 467 }, { "epoch": 1.7674858223062382, "grad_norm": 0.7263327836990356, "learning_rate": 4.449925898997745e-06, "loss": 0.0102, "step": 468 }, { "epoch": 1.7712665406427222, "grad_norm": 0.3550947904586792, "learning_rate": 4.4464864361549105e-06, "loss": 0.0023, "step": 469 }, { "epoch": 1.775047258979206, "grad_norm": 0.5708559155464172, "learning_rate": 4.443037591824988e-06, "loss": 0.0019, "step": 470 }, { "epoch": 1.7788279773156899, "grad_norm": 0.08885316550731659, "learning_rate": 4.439579382630384e-06, "loss": 0.0007, "step": 471 }, { "epoch": 1.7826086956521738, "grad_norm": 0.6298693418502808, "learning_rate": 4.43611182523864e-06, "loss": 0.0048, "step": 472 }, { "epoch": 1.7863894139886578, "grad_norm": 0.18471179902553558, "learning_rate": 4.432634936362354e-06, "loss": 0.0011, "step": 473 }, { "epoch": 1.7901701323251418, "grad_norm": 0.5520258545875549, "learning_rate": 4.429148732759101e-06, "loss": 0.0055, "step": 474 }, { "epoch": 1.7939508506616257, "grad_norm": 0.30655577778816223, "learning_rate": 4.425653231231344e-06, "loss": 0.0031, "step": 475 }, { "epoch": 1.7977315689981097, "grad_norm": 0.38472461700439453, "learning_rate": 4.4221484486263664e-06, "loss": 0.0019, "step": 476 }, { "epoch": 1.8015122873345937, "grad_norm": 0.38701826333999634, "learning_rate": 4.418634401836179e-06, "loss": 0.0074, "step": 477 }, { "epoch": 1.8052930056710776, "grad_norm": 0.46479108929634094, "learning_rate": 4.415111107797445e-06, "loss": 0.0016, "step": 478 }, { "epoch": 1.8090737240075614, "grad_norm": 0.3576240837574005, "learning_rate": 4.411578583491397e-06, "loss": 0.0033, "step": 479 }, { "epoch": 1.8128544423440454, "grad_norm": 0.63249272108078, "learning_rate": 4.408036845943753e-06, "loss": 0.0053, "step": 480 }, { "epoch": 1.8166351606805293, "grad_norm": 0.5044552683830261, "learning_rate": 4.404485912224638e-06, "loss": 0.0054, "step": 481 }, { "epoch": 1.820415879017013, "grad_norm": 0.17272500693798065, "learning_rate": 4.400925799448498e-06, "loss": 0.0012, "step": 482 }, { "epoch": 1.824196597353497, "grad_norm": 0.4603140652179718, "learning_rate": 4.39735652477402e-06, "loss": 0.0039, "step": 483 }, { "epoch": 1.827977315689981, "grad_norm": 0.18995338678359985, "learning_rate": 4.393778105404051e-06, "loss": 0.0009, "step": 484 }, { "epoch": 1.831758034026465, "grad_norm": 0.30568787455558777, "learning_rate": 4.390190558585508e-06, "loss": 0.0028, "step": 485 }, { "epoch": 1.835538752362949, "grad_norm": 0.32309314608573914, "learning_rate": 4.386593901609304e-06, "loss": 0.0024, "step": 486 }, { "epoch": 1.839319470699433, "grad_norm": 0.2775632441043854, "learning_rate": 4.382988151810258e-06, "loss": 0.0039, "step": 487 }, { "epoch": 1.8431001890359169, "grad_norm": 0.2362305074930191, "learning_rate": 4.379373326567015e-06, "loss": 0.0021, "step": 488 }, { "epoch": 1.8468809073724008, "grad_norm": 0.33779892325401306, "learning_rate": 4.375749443301957e-06, "loss": 0.0025, "step": 489 }, { "epoch": 1.8506616257088848, "grad_norm": 0.2156297266483307, "learning_rate": 4.372116519481131e-06, "loss": 0.0019, "step": 490 }, { "epoch": 1.8544423440453688, "grad_norm": 0.37446320056915283, "learning_rate": 4.3684745726141474e-06, "loss": 0.0023, "step": 491 }, { "epoch": 1.8582230623818525, "grad_norm": 0.8646283149719238, "learning_rate": 4.364823620254113e-06, "loss": 0.0025, "step": 492 }, { "epoch": 1.8620037807183365, "grad_norm": 0.34199878573417664, "learning_rate": 4.361163679997532e-06, "loss": 0.0079, "step": 493 }, { "epoch": 1.8657844990548205, "grad_norm": 0.5534374117851257, "learning_rate": 4.357494769484232e-06, "loss": 0.0009, "step": 494 }, { "epoch": 1.8695652173913042, "grad_norm": 0.1436002403497696, "learning_rate": 4.3538169063972735e-06, "loss": 0.0012, "step": 495 }, { "epoch": 1.8733459357277882, "grad_norm": 0.7071740031242371, "learning_rate": 4.350130108462864e-06, "loss": 0.0028, "step": 496 }, { "epoch": 1.8771266540642721, "grad_norm": 0.297133207321167, "learning_rate": 4.346434393450277e-06, "loss": 0.0026, "step": 497 }, { "epoch": 1.8809073724007561, "grad_norm": 0.10467611253261566, "learning_rate": 4.342729779171762e-06, "loss": 0.0009, "step": 498 }, { "epoch": 1.88468809073724, "grad_norm": 0.2588742673397064, "learning_rate": 4.33901628348246e-06, "loss": 0.0019, "step": 499 }, { "epoch": 1.888468809073724, "grad_norm": 0.5257592797279358, "learning_rate": 4.335293924280318e-06, "loss": 0.0062, "step": 500 }, { "epoch": 1.892249527410208, "grad_norm": 0.34946784377098083, "learning_rate": 4.331562719506005e-06, "loss": 0.003, "step": 501 }, { "epoch": 1.896030245746692, "grad_norm": 0.4206084609031677, "learning_rate": 4.327822687142818e-06, "loss": 0.0027, "step": 502 }, { "epoch": 1.899810964083176, "grad_norm": 0.27836281061172485, "learning_rate": 4.324073845216606e-06, "loss": 0.0016, "step": 503 }, { "epoch": 1.9035916824196597, "grad_norm": 0.4917697608470917, "learning_rate": 4.320316211795674e-06, "loss": 0.0036, "step": 504 }, { "epoch": 1.9073724007561437, "grad_norm": 0.3069250285625458, "learning_rate": 4.3165498049907e-06, "loss": 0.0113, "step": 505 }, { "epoch": 1.9111531190926276, "grad_norm": 0.683668315410614, "learning_rate": 4.3127746429546465e-06, "loss": 0.0043, "step": 506 }, { "epoch": 1.9149338374291114, "grad_norm": 0.7328367829322815, "learning_rate": 4.308990743882675e-06, "loss": 0.0073, "step": 507 }, { "epoch": 1.9187145557655954, "grad_norm": 0.5277181267738342, "learning_rate": 4.305198126012057e-06, "loss": 0.0055, "step": 508 }, { "epoch": 1.9224952741020793, "grad_norm": 0.5691857933998108, "learning_rate": 4.301396807622084e-06, "loss": 0.0069, "step": 509 }, { "epoch": 1.9262759924385633, "grad_norm": 0.806879997253418, "learning_rate": 4.297586807033983e-06, "loss": 0.0087, "step": 510 }, { "epoch": 1.9300567107750473, "grad_norm": 0.32039013504981995, "learning_rate": 4.293768142610828e-06, "loss": 0.0019, "step": 511 }, { "epoch": 1.9338374291115312, "grad_norm": 0.34749680757522583, "learning_rate": 4.2899408327574475e-06, "loss": 0.0016, "step": 512 }, { "epoch": 1.9376181474480152, "grad_norm": 0.5303663015365601, "learning_rate": 4.2861048959203386e-06, "loss": 0.0088, "step": 513 }, { "epoch": 1.9413988657844992, "grad_norm": 0.20052987337112427, "learning_rate": 4.282260350587579e-06, "loss": 0.0021, "step": 514 }, { "epoch": 1.9451795841209831, "grad_norm": 0.8037423491477966, "learning_rate": 4.278407215288738e-06, "loss": 0.0105, "step": 515 }, { "epoch": 1.9489603024574669, "grad_norm": 0.28701499104499817, "learning_rate": 4.274545508594783e-06, "loss": 0.0015, "step": 516 }, { "epoch": 1.9527410207939508, "grad_norm": 1.585216760635376, "learning_rate": 4.270675249117995e-06, "loss": 0.0033, "step": 517 }, { "epoch": 1.9565217391304348, "grad_norm": 0.42489415407180786, "learning_rate": 4.2667964555118755e-06, "loss": 0.0043, "step": 518 }, { "epoch": 1.9603024574669186, "grad_norm": 0.4090639650821686, "learning_rate": 4.26290914647106e-06, "loss": 0.0056, "step": 519 }, { "epoch": 1.9640831758034025, "grad_norm": 1.2218302488327026, "learning_rate": 4.259013340731224e-06, "loss": 0.0058, "step": 520 }, { "epoch": 1.9678638941398865, "grad_norm": 0.19953207671642303, "learning_rate": 4.255109057068995e-06, "loss": 0.001, "step": 521 }, { "epoch": 1.9716446124763705, "grad_norm": 0.6282517910003662, "learning_rate": 4.251196314301864e-06, "loss": 0.004, "step": 522 }, { "epoch": 1.9754253308128544, "grad_norm": 0.4743454158306122, "learning_rate": 4.2472751312880866e-06, "loss": 0.0087, "step": 523 }, { "epoch": 1.9792060491493384, "grad_norm": 0.7538068890571594, "learning_rate": 4.243345526926605e-06, "loss": 0.0026, "step": 524 }, { "epoch": 1.9829867674858224, "grad_norm": 0.3533986508846283, "learning_rate": 4.239407520156943e-06, "loss": 0.0055, "step": 525 }, { "epoch": 1.9867674858223063, "grad_norm": 0.17272821068763733, "learning_rate": 4.235461129959128e-06, "loss": 0.0025, "step": 526 }, { "epoch": 1.9905482041587903, "grad_norm": 0.23916426301002502, "learning_rate": 4.231506375353588e-06, "loss": 0.0012, "step": 527 }, { "epoch": 1.994328922495274, "grad_norm": 1.7950382232666016, "learning_rate": 4.227543275401067e-06, "loss": 0.0021, "step": 528 }, { "epoch": 1.998109640831758, "grad_norm": 0.5973864197731018, "learning_rate": 4.22357184920253e-06, "loss": 0.0048, "step": 529 }, { "epoch": 2.0, "grad_norm": 0.496730774641037, "learning_rate": 4.219592115899075e-06, "loss": 0.0031, "step": 530 }, { "epoch": 2.003780718336484, "grad_norm": 0.23886018991470337, "learning_rate": 4.215604094671835e-06, "loss": 0.0022, "step": 531 }, { "epoch": 2.007561436672968, "grad_norm": 0.7827872633934021, "learning_rate": 4.211607804741888e-06, "loss": 0.0018, "step": 532 }, { "epoch": 2.011342155009452, "grad_norm": 0.4427229166030884, "learning_rate": 4.207603265370166e-06, "loss": 0.0024, "step": 533 }, { "epoch": 2.015122873345936, "grad_norm": 0.29099422693252563, "learning_rate": 4.2035904958573615e-06, "loss": 0.0023, "step": 534 }, { "epoch": 2.01890359168242, "grad_norm": 0.13313233852386475, "learning_rate": 4.1995695155438325e-06, "loss": 0.0007, "step": 535 }, { "epoch": 2.022684310018904, "grad_norm": 0.1648566722869873, "learning_rate": 4.195540343809511e-06, "loss": 0.0013, "step": 536 }, { "epoch": 2.0264650283553873, "grad_norm": 0.40962958335876465, "learning_rate": 4.19150300007381e-06, "loss": 0.0009, "step": 537 }, { "epoch": 2.0302457466918713, "grad_norm": 0.21508942544460297, "learning_rate": 4.187457503795526e-06, "loss": 0.002, "step": 538 }, { "epoch": 2.0340264650283553, "grad_norm": 0.5949055552482605, "learning_rate": 4.183403874472754e-06, "loss": 0.009, "step": 539 }, { "epoch": 2.0378071833648392, "grad_norm": 0.2826746106147766, "learning_rate": 4.179342131642783e-06, "loss": 0.0014, "step": 540 }, { "epoch": 2.041587901701323, "grad_norm": 1.0113091468811035, "learning_rate": 4.175272294882008e-06, "loss": 0.0053, "step": 541 }, { "epoch": 2.045368620037807, "grad_norm": 0.1866329461336136, "learning_rate": 4.1711943838058346e-06, "loss": 0.0009, "step": 542 }, { "epoch": 2.049149338374291, "grad_norm": 0.539692223072052, "learning_rate": 4.167108418068585e-06, "loss": 0.0016, "step": 543 }, { "epoch": 2.052930056710775, "grad_norm": 0.5694279670715332, "learning_rate": 4.163014417363402e-06, "loss": 0.0024, "step": 544 }, { "epoch": 2.056710775047259, "grad_norm": 0.19736915826797485, "learning_rate": 4.158912401422152e-06, "loss": 0.0014, "step": 545 }, { "epoch": 2.060491493383743, "grad_norm": 0.2381485253572464, "learning_rate": 4.154802390015338e-06, "loss": 0.0015, "step": 546 }, { "epoch": 2.064272211720227, "grad_norm": 0.12526023387908936, "learning_rate": 4.150684402951994e-06, "loss": 0.0007, "step": 547 }, { "epoch": 2.068052930056711, "grad_norm": 0.26598063111305237, "learning_rate": 4.146558460079595e-06, "loss": 0.0021, "step": 548 }, { "epoch": 2.0718336483931945, "grad_norm": 0.3488152325153351, "learning_rate": 4.142424581283962e-06, "loss": 0.0015, "step": 549 }, { "epoch": 2.0756143667296785, "grad_norm": 0.3600805103778839, "learning_rate": 4.138282786489165e-06, "loss": 0.0016, "step": 550 }, { "epoch": 2.0793950850661624, "grad_norm": 0.16840273141860962, "learning_rate": 4.134133095657424e-06, "loss": 0.0011, "step": 551 }, { "epoch": 2.0831758034026464, "grad_norm": 2.7653298377990723, "learning_rate": 4.129975528789021e-06, "loss": 0.0006, "step": 552 }, { "epoch": 2.0869565217391304, "grad_norm": 0.8231133818626404, "learning_rate": 4.125810105922192e-06, "loss": 0.004, "step": 553 }, { "epoch": 2.0907372400756143, "grad_norm": 0.6421529650688171, "learning_rate": 4.121636847133039e-06, "loss": 0.0059, "step": 554 }, { "epoch": 2.0945179584120983, "grad_norm": 0.3676932156085968, "learning_rate": 4.117455772535433e-06, "loss": 0.0023, "step": 555 }, { "epoch": 2.0982986767485823, "grad_norm": 0.210940420627594, "learning_rate": 4.113266902280914e-06, "loss": 0.001, "step": 556 }, { "epoch": 2.1020793950850662, "grad_norm": 0.5619054436683655, "learning_rate": 4.109070256558589e-06, "loss": 0.0041, "step": 557 }, { "epoch": 2.10586011342155, "grad_norm": 0.3032905161380768, "learning_rate": 4.10486585559505e-06, "loss": 0.0016, "step": 558 }, { "epoch": 2.109640831758034, "grad_norm": 0.33789193630218506, "learning_rate": 4.100653719654259e-06, "loss": 0.0016, "step": 559 }, { "epoch": 2.113421550094518, "grad_norm": 0.329702228307724, "learning_rate": 4.096433869037464e-06, "loss": 0.0035, "step": 560 }, { "epoch": 2.1172022684310017, "grad_norm": 0.05710732191801071, "learning_rate": 4.0922063240830914e-06, "loss": 0.0005, "step": 561 }, { "epoch": 2.1209829867674856, "grad_norm": 0.5501359701156616, "learning_rate": 4.087971105166654e-06, "loss": 0.0036, "step": 562 }, { "epoch": 2.1247637051039696, "grad_norm": 0.21680495142936707, "learning_rate": 4.08372823270065e-06, "loss": 0.0009, "step": 563 }, { "epoch": 2.1285444234404536, "grad_norm": 0.12248539179563522, "learning_rate": 4.079477727134466e-06, "loss": 0.0006, "step": 564 }, { "epoch": 2.1323251417769375, "grad_norm": 0.23461495339870453, "learning_rate": 4.075219608954279e-06, "loss": 0.0012, "step": 565 }, { "epoch": 2.1361058601134215, "grad_norm": 0.6181909441947937, "learning_rate": 4.070953898682952e-06, "loss": 0.0061, "step": 566 }, { "epoch": 2.1398865784499055, "grad_norm": 0.2439657747745514, "learning_rate": 4.066680616879946e-06, "loss": 0.0009, "step": 567 }, { "epoch": 2.1436672967863895, "grad_norm": 0.47801685333251953, "learning_rate": 4.062399784141209e-06, "loss": 0.0019, "step": 568 }, { "epoch": 2.1474480151228734, "grad_norm": 0.31439632177352905, "learning_rate": 4.058111421099087e-06, "loss": 0.0021, "step": 569 }, { "epoch": 2.1512287334593574, "grad_norm": 0.05819639191031456, "learning_rate": 4.0538155484222156e-06, "loss": 0.0004, "step": 570 }, { "epoch": 2.1550094517958414, "grad_norm": 0.29354599118232727, "learning_rate": 4.0495121868154275e-06, "loss": 0.0013, "step": 571 }, { "epoch": 2.1587901701323253, "grad_norm": 0.7409812211990356, "learning_rate": 4.045201357019648e-06, "loss": 0.0024, "step": 572 }, { "epoch": 2.1625708884688093, "grad_norm": 1.6672428846359253, "learning_rate": 4.040883079811799e-06, "loss": 0.0013, "step": 573 }, { "epoch": 2.166351606805293, "grad_norm": 0.2385888546705246, "learning_rate": 4.036557376004694e-06, "loss": 0.0021, "step": 574 }, { "epoch": 2.170132325141777, "grad_norm": 0.3318484425544739, "learning_rate": 4.032224266446943e-06, "loss": 0.0042, "step": 575 }, { "epoch": 2.1739130434782608, "grad_norm": 0.2983977198600769, "learning_rate": 4.027883772022847e-06, "loss": 0.0012, "step": 576 }, { "epoch": 2.1776937618147447, "grad_norm": 0.3238692879676819, "learning_rate": 4.0235359136523024e-06, "loss": 0.0015, "step": 577 }, { "epoch": 2.1814744801512287, "grad_norm": 0.4926668405532837, "learning_rate": 4.0191807122906965e-06, "loss": 0.0014, "step": 578 }, { "epoch": 2.1852551984877127, "grad_norm": 0.5650737881660461, "learning_rate": 4.014818188928808e-06, "loss": 0.0066, "step": 579 }, { "epoch": 2.1890359168241966, "grad_norm": 0.22845497727394104, "learning_rate": 4.010448364592703e-06, "loss": 0.0011, "step": 580 }, { "epoch": 2.1928166351606806, "grad_norm": 0.08073125034570694, "learning_rate": 4.00607126034364e-06, "loss": 0.0004, "step": 581 }, { "epoch": 2.1965973534971646, "grad_norm": 0.23095814883708954, "learning_rate": 4.0016868972779636e-06, "loss": 0.0013, "step": 582 }, { "epoch": 2.2003780718336485, "grad_norm": 0.11648385226726532, "learning_rate": 3.997295296527001e-06, "loss": 0.0006, "step": 583 }, { "epoch": 2.2041587901701325, "grad_norm": 0.4050494432449341, "learning_rate": 3.992896479256966e-06, "loss": 0.0031, "step": 584 }, { "epoch": 2.207939508506616, "grad_norm": 0.6414379477500916, "learning_rate": 3.988490466668852e-06, "loss": 0.0029, "step": 585 }, { "epoch": 2.2117202268431, "grad_norm": 0.2900727689266205, "learning_rate": 3.984077279998333e-06, "loss": 0.0011, "step": 586 }, { "epoch": 2.215500945179584, "grad_norm": 0.19610387086868286, "learning_rate": 3.979656940515659e-06, "loss": 0.0009, "step": 587 }, { "epoch": 2.219281663516068, "grad_norm": 0.3006332814693451, "learning_rate": 3.975229469525555e-06, "loss": 0.0027, "step": 588 }, { "epoch": 2.223062381852552, "grad_norm": 0.40080544352531433, "learning_rate": 3.970794888367116e-06, "loss": 0.0018, "step": 589 }, { "epoch": 2.226843100189036, "grad_norm": 0.24024249613285065, "learning_rate": 3.96635321841371e-06, "loss": 0.0018, "step": 590 }, { "epoch": 2.23062381852552, "grad_norm": 0.23088271915912628, "learning_rate": 3.961904481072867e-06, "loss": 0.0014, "step": 591 }, { "epoch": 2.234404536862004, "grad_norm": 0.19720318913459778, "learning_rate": 3.957448697786181e-06, "loss": 0.0013, "step": 592 }, { "epoch": 2.2381852551984878, "grad_norm": 0.245695099234581, "learning_rate": 3.952985890029205e-06, "loss": 0.0017, "step": 593 }, { "epoch": 2.2419659735349717, "grad_norm": 0.5289967656135559, "learning_rate": 3.948516079311349e-06, "loss": 0.0043, "step": 594 }, { "epoch": 2.2457466918714557, "grad_norm": 0.17461897432804108, "learning_rate": 3.9440392871757735e-06, "loss": 0.0008, "step": 595 }, { "epoch": 2.2495274102079397, "grad_norm": 0.31308841705322266, "learning_rate": 3.93955553519929e-06, "loss": 0.0025, "step": 596 }, { "epoch": 2.2533081285444236, "grad_norm": 0.34215158224105835, "learning_rate": 3.935064844992251e-06, "loss": 0.0011, "step": 597 }, { "epoch": 2.257088846880907, "grad_norm": 0.14022661745548248, "learning_rate": 3.930567238198451e-06, "loss": 0.0005, "step": 598 }, { "epoch": 2.260869565217391, "grad_norm": 0.1508352905511856, "learning_rate": 3.926062736495021e-06, "loss": 0.0009, "step": 599 }, { "epoch": 2.264650283553875, "grad_norm": 0.46096035838127136, "learning_rate": 3.921551361592323e-06, "loss": 0.0031, "step": 600 }, { "epoch": 2.268431001890359, "grad_norm": 0.037727996706962585, "learning_rate": 3.917033135233845e-06, "loss": 0.0004, "step": 601 }, { "epoch": 2.272211720226843, "grad_norm": 0.4264487624168396, "learning_rate": 3.912508079196097e-06, "loss": 0.0032, "step": 602 }, { "epoch": 2.275992438563327, "grad_norm": 0.5846503376960754, "learning_rate": 3.907976215288507e-06, "loss": 0.0032, "step": 603 }, { "epoch": 2.279773156899811, "grad_norm": 0.10320373624563217, "learning_rate": 3.903437565353314e-06, "loss": 0.0004, "step": 604 }, { "epoch": 2.283553875236295, "grad_norm": 0.30576497316360474, "learning_rate": 3.8988921512654644e-06, "loss": 0.0014, "step": 605 }, { "epoch": 2.287334593572779, "grad_norm": 0.26759201288223267, "learning_rate": 3.8943399949325055e-06, "loss": 0.0022, "step": 606 }, { "epoch": 2.291115311909263, "grad_norm": 0.8713322877883911, "learning_rate": 3.88978111829448e-06, "loss": 0.003, "step": 607 }, { "epoch": 2.294896030245747, "grad_norm": 0.2531176805496216, "learning_rate": 3.885215543323821e-06, "loss": 0.0007, "step": 608 }, { "epoch": 2.298676748582231, "grad_norm": 1.2418513298034668, "learning_rate": 3.880643292025245e-06, "loss": 0.0043, "step": 609 }, { "epoch": 2.302457466918715, "grad_norm": 0.2262953668832779, "learning_rate": 3.876064386435646e-06, "loss": 0.0011, "step": 610 }, { "epoch": 2.3062381852551983, "grad_norm": 0.36947792768478394, "learning_rate": 3.871478848623992e-06, "loss": 0.0011, "step": 611 }, { "epoch": 2.3100189035916823, "grad_norm": 0.26794055104255676, "learning_rate": 3.8668867006912126e-06, "loss": 0.0047, "step": 612 }, { "epoch": 2.3137996219281662, "grad_norm": 0.3170675039291382, "learning_rate": 3.862287964770099e-06, "loss": 0.0022, "step": 613 }, { "epoch": 2.31758034026465, "grad_norm": 0.5809718370437622, "learning_rate": 3.857682663025195e-06, "loss": 0.0029, "step": 614 }, { "epoch": 2.321361058601134, "grad_norm": 0.4641650915145874, "learning_rate": 3.853070817652688e-06, "loss": 0.0041, "step": 615 }, { "epoch": 2.325141776937618, "grad_norm": 0.3503637909889221, "learning_rate": 3.848452450880303e-06, "loss": 0.0033, "step": 616 }, { "epoch": 2.328922495274102, "grad_norm": 0.5029269456863403, "learning_rate": 3.8438275849671994e-06, "loss": 0.006, "step": 617 }, { "epoch": 2.332703213610586, "grad_norm": 0.21303796768188477, "learning_rate": 3.839196242203859e-06, "loss": 0.0009, "step": 618 }, { "epoch": 2.33648393194707, "grad_norm": 0.3005557358264923, "learning_rate": 3.834558444911978e-06, "loss": 0.0021, "step": 619 }, { "epoch": 2.340264650283554, "grad_norm": 0.31283143162727356, "learning_rate": 3.829914215444363e-06, "loss": 0.0015, "step": 620 }, { "epoch": 2.344045368620038, "grad_norm": 1.1231919527053833, "learning_rate": 3.825263576184823e-06, "loss": 0.0041, "step": 621 }, { "epoch": 2.3478260869565215, "grad_norm": 0.4406910836696625, "learning_rate": 3.82060654954806e-06, "loss": 0.0015, "step": 622 }, { "epoch": 2.351606805293006, "grad_norm": 0.34202250838279724, "learning_rate": 3.8159431579795605e-06, "loss": 0.0034, "step": 623 }, { "epoch": 2.3553875236294894, "grad_norm": 0.3115376830101013, "learning_rate": 3.8112734239554884e-06, "loss": 0.0014, "step": 624 }, { "epoch": 2.3591682419659734, "grad_norm": 0.46241116523742676, "learning_rate": 3.806597369982574e-06, "loss": 0.0035, "step": 625 }, { "epoch": 2.3629489603024574, "grad_norm": 0.28592902421951294, "learning_rate": 3.8019150185980125e-06, "loss": 0.0028, "step": 626 }, { "epoch": 2.3667296786389413, "grad_norm": 0.2742108106613159, "learning_rate": 3.7972263923693477e-06, "loss": 0.0007, "step": 627 }, { "epoch": 2.3705103969754253, "grad_norm": 0.19143778085708618, "learning_rate": 3.792531513894365e-06, "loss": 0.0008, "step": 628 }, { "epoch": 2.3742911153119093, "grad_norm": 0.26212093234062195, "learning_rate": 3.7878304058009883e-06, "loss": 0.0009, "step": 629 }, { "epoch": 2.3780718336483933, "grad_norm": 0.45981907844543457, "learning_rate": 3.7831230907471626e-06, "loss": 0.0018, "step": 630 }, { "epoch": 2.381852551984877, "grad_norm": 0.0761369839310646, "learning_rate": 3.778409591420749e-06, "loss": 0.0003, "step": 631 }, { "epoch": 2.385633270321361, "grad_norm": 1.1732951402664185, "learning_rate": 3.7736899305394178e-06, "loss": 0.0055, "step": 632 }, { "epoch": 2.389413988657845, "grad_norm": 0.35270413756370544, "learning_rate": 3.768964130850532e-06, "loss": 0.0017, "step": 633 }, { "epoch": 2.393194706994329, "grad_norm": 0.09166122227907181, "learning_rate": 3.7642322151310453e-06, "loss": 0.0005, "step": 634 }, { "epoch": 2.3969754253308126, "grad_norm": 0.07523413002490997, "learning_rate": 3.7594942061873863e-06, "loss": 0.0005, "step": 635 }, { "epoch": 2.4007561436672966, "grad_norm": 0.37303024530410767, "learning_rate": 3.754750126855353e-06, "loss": 0.0017, "step": 636 }, { "epoch": 2.4045368620037806, "grad_norm": 0.7728914022445679, "learning_rate": 3.7500000000000005e-06, "loss": 0.0051, "step": 637 }, { "epoch": 2.4083175803402646, "grad_norm": 0.7667577266693115, "learning_rate": 3.7452438485155305e-06, "loss": 0.0038, "step": 638 }, { "epoch": 2.4120982986767485, "grad_norm": 0.13398635387420654, "learning_rate": 3.740481695325181e-06, "loss": 0.0003, "step": 639 }, { "epoch": 2.4158790170132325, "grad_norm": 0.27832144498825073, "learning_rate": 3.7357135633811198e-06, "loss": 0.0007, "step": 640 }, { "epoch": 2.4196597353497165, "grad_norm": 0.47430655360221863, "learning_rate": 3.7309394756643265e-06, "loss": 0.0024, "step": 641 }, { "epoch": 2.4234404536862004, "grad_norm": 0.23291529715061188, "learning_rate": 3.7261594551844863e-06, "loss": 0.001, "step": 642 }, { "epoch": 2.4272211720226844, "grad_norm": 0.24835488200187683, "learning_rate": 3.7213735249798826e-06, "loss": 0.0014, "step": 643 }, { "epoch": 2.4310018903591684, "grad_norm": 0.7222422361373901, "learning_rate": 3.716581708117276e-06, "loss": 0.0043, "step": 644 }, { "epoch": 2.4347826086956523, "grad_norm": 0.8068685531616211, "learning_rate": 3.711784027691805e-06, "loss": 0.0072, "step": 645 }, { "epoch": 2.4385633270321363, "grad_norm": 0.31845006346702576, "learning_rate": 3.7069805068268626e-06, "loss": 0.0008, "step": 646 }, { "epoch": 2.4423440453686203, "grad_norm": 1.0978343486785889, "learning_rate": 3.7021711686739963e-06, "loss": 0.0052, "step": 647 }, { "epoch": 2.446124763705104, "grad_norm": 0.16923770308494568, "learning_rate": 3.697356036412788e-06, "loss": 0.001, "step": 648 }, { "epoch": 2.4499054820415878, "grad_norm": 0.5710031986236572, "learning_rate": 3.6925351332507477e-06, "loss": 0.0035, "step": 649 }, { "epoch": 2.4536862003780717, "grad_norm": 0.11478786915540695, "learning_rate": 3.6877084824231962e-06, "loss": 0.0006, "step": 650 }, { "epoch": 2.4574669187145557, "grad_norm": 0.49666333198547363, "learning_rate": 3.6828761071931605e-06, "loss": 0.0024, "step": 651 }, { "epoch": 2.4612476370510397, "grad_norm": 0.21035583317279816, "learning_rate": 3.6780380308512544e-06, "loss": 0.0009, "step": 652 }, { "epoch": 2.4650283553875236, "grad_norm": 0.5114822387695312, "learning_rate": 3.6731942767155703e-06, "loss": 0.0031, "step": 653 }, { "epoch": 2.4688090737240076, "grad_norm": 0.15796847641468048, "learning_rate": 3.668344868131567e-06, "loss": 0.0005, "step": 654 }, { "epoch": 2.4725897920604916, "grad_norm": 0.3661459982395172, "learning_rate": 3.6634898284719533e-06, "loss": 0.0027, "step": 655 }, { "epoch": 2.4763705103969755, "grad_norm": 0.43741801381111145, "learning_rate": 3.6586291811365814e-06, "loss": 0.0022, "step": 656 }, { "epoch": 2.4801512287334595, "grad_norm": 0.1882481873035431, "learning_rate": 3.653762949552328e-06, "loss": 0.0013, "step": 657 }, { "epoch": 2.4839319470699435, "grad_norm": 0.08593080937862396, "learning_rate": 3.648891157172986e-06, "loss": 0.0004, "step": 658 }, { "epoch": 2.487712665406427, "grad_norm": 0.10869196802377701, "learning_rate": 3.6440138274791505e-06, "loss": 0.0005, "step": 659 }, { "epoch": 2.491493383742911, "grad_norm": 0.6629149913787842, "learning_rate": 3.6391309839781014e-06, "loss": 0.0073, "step": 660 }, { "epoch": 2.495274102079395, "grad_norm": 0.10865134745836258, "learning_rate": 3.6342426502036972e-06, "loss": 0.0007, "step": 661 }, { "epoch": 2.499054820415879, "grad_norm": 0.1243506446480751, "learning_rate": 3.6293488497162555e-06, "loss": 0.0006, "step": 662 }, { "epoch": 2.502835538752363, "grad_norm": 0.086354099214077, "learning_rate": 3.6244496061024416e-06, "loss": 0.0006, "step": 663 }, { "epoch": 2.506616257088847, "grad_norm": 0.19867630302906036, "learning_rate": 3.6195449429751585e-06, "loss": 0.001, "step": 664 }, { "epoch": 2.510396975425331, "grad_norm": 0.4615210294723511, "learning_rate": 3.614634883973425e-06, "loss": 0.0019, "step": 665 }, { "epoch": 2.5141776937618148, "grad_norm": 0.07232984155416489, "learning_rate": 3.6097194527622697e-06, "loss": 0.0005, "step": 666 }, { "epoch": 2.5179584120982987, "grad_norm": 0.26847028732299805, "learning_rate": 3.6047986730326132e-06, "loss": 0.0023, "step": 667 }, { "epoch": 2.5217391304347827, "grad_norm": 0.14426179230213165, "learning_rate": 3.5998725685011526e-06, "loss": 0.001, "step": 668 }, { "epoch": 2.5255198487712667, "grad_norm": 0.259517639875412, "learning_rate": 3.594941162910251e-06, "loss": 0.0012, "step": 669 }, { "epoch": 2.52930056710775, "grad_norm": 0.23957423865795135, "learning_rate": 3.59000448002782e-06, "loss": 0.0015, "step": 670 }, { "epoch": 2.5330812854442346, "grad_norm": 0.30308109521865845, "learning_rate": 3.585062543647205e-06, "loss": 0.0013, "step": 671 }, { "epoch": 2.536862003780718, "grad_norm": 0.7398681640625, "learning_rate": 3.580115377587075e-06, "loss": 0.0091, "step": 672 }, { "epoch": 2.5406427221172025, "grad_norm": 0.24520723521709442, "learning_rate": 3.5751630056913017e-06, "loss": 0.0011, "step": 673 }, { "epoch": 2.544423440453686, "grad_norm": 0.5432690382003784, "learning_rate": 3.5702054518288477e-06, "loss": 0.0065, "step": 674 }, { "epoch": 2.54820415879017, "grad_norm": 0.38682425022125244, "learning_rate": 3.5652427398936527e-06, "loss": 0.0022, "step": 675 }, { "epoch": 2.551984877126654, "grad_norm": 0.1622408777475357, "learning_rate": 3.560274893804515e-06, "loss": 0.0008, "step": 676 }, { "epoch": 2.555765595463138, "grad_norm": 0.35491493344306946, "learning_rate": 3.55530193750498e-06, "loss": 0.0047, "step": 677 }, { "epoch": 2.559546313799622, "grad_norm": 0.45866596698760986, "learning_rate": 3.5503238949632202e-06, "loss": 0.0025, "step": 678 }, { "epoch": 2.563327032136106, "grad_norm": 0.15724623203277588, "learning_rate": 3.5453407901719232e-06, "loss": 0.0006, "step": 679 }, { "epoch": 2.56710775047259, "grad_norm": 0.33184540271759033, "learning_rate": 3.5403526471481777e-06, "loss": 0.0012, "step": 680 }, { "epoch": 2.570888468809074, "grad_norm": 0.5639476180076599, "learning_rate": 3.535359489933351e-06, "loss": 0.0022, "step": 681 }, { "epoch": 2.574669187145558, "grad_norm": 0.09735933691263199, "learning_rate": 3.5303613425929805e-06, "loss": 0.0005, "step": 682 }, { "epoch": 2.5784499054820413, "grad_norm": 0.4168652892112732, "learning_rate": 3.5253582292166525e-06, "loss": 0.0036, "step": 683 }, { "epoch": 2.5822306238185257, "grad_norm": 0.28641799092292786, "learning_rate": 3.5203501739178904e-06, "loss": 0.0011, "step": 684 }, { "epoch": 2.5860113421550093, "grad_norm": 0.148334339261055, "learning_rate": 3.515337200834034e-06, "loss": 0.0006, "step": 685 }, { "epoch": 2.5897920604914932, "grad_norm": 1.0006016492843628, "learning_rate": 3.510319334126126e-06, "loss": 0.0019, "step": 686 }, { "epoch": 2.593572778827977, "grad_norm": 0.15660707652568817, "learning_rate": 3.5052965979787952e-06, "loss": 0.0008, "step": 687 }, { "epoch": 2.597353497164461, "grad_norm": 0.3002500534057617, "learning_rate": 3.50026901660014e-06, "loss": 0.0015, "step": 688 }, { "epoch": 2.601134215500945, "grad_norm": 0.3913247585296631, "learning_rate": 3.4952366142216097e-06, "loss": 0.0013, "step": 689 }, { "epoch": 2.604914933837429, "grad_norm": 0.2940994203090668, "learning_rate": 3.4901994150978926e-06, "loss": 0.0014, "step": 690 }, { "epoch": 2.608695652173913, "grad_norm": 0.6625819802284241, "learning_rate": 3.4851574435067925e-06, "loss": 0.0023, "step": 691 }, { "epoch": 2.612476370510397, "grad_norm": 1.3197686672210693, "learning_rate": 3.480110723749116e-06, "loss": 0.0029, "step": 692 }, { "epoch": 2.616257088846881, "grad_norm": 0.051980722695589066, "learning_rate": 3.4750592801485565e-06, "loss": 0.0004, "step": 693 }, { "epoch": 2.620037807183365, "grad_norm": 0.4224025011062622, "learning_rate": 3.470003137051573e-06, "loss": 0.0016, "step": 694 }, { "epoch": 2.623818525519849, "grad_norm": 0.4571843445301056, "learning_rate": 3.4649423188272743e-06, "loss": 0.0041, "step": 695 }, { "epoch": 2.6275992438563325, "grad_norm": 0.17115405201911926, "learning_rate": 3.459876849867303e-06, "loss": 0.0011, "step": 696 }, { "epoch": 2.631379962192817, "grad_norm": 0.04624168574810028, "learning_rate": 3.454806754585716e-06, "loss": 0.0002, "step": 697 }, { "epoch": 2.6351606805293004, "grad_norm": 0.19578991830348969, "learning_rate": 3.4497320574188696e-06, "loss": 0.001, "step": 698 }, { "epoch": 2.6389413988657844, "grad_norm": 0.04589693248271942, "learning_rate": 3.4446527828252975e-06, "loss": 0.0002, "step": 699 }, { "epoch": 2.6427221172022684, "grad_norm": 0.047621648758649826, "learning_rate": 3.4395689552855956e-06, "loss": 0.0002, "step": 700 }, { "epoch": 2.6465028355387523, "grad_norm": 0.22325396537780762, "learning_rate": 3.4344805993023056e-06, "loss": 0.0004, "step": 701 }, { "epoch": 2.6502835538752363, "grad_norm": 0.4162357449531555, "learning_rate": 3.429387739399793e-06, "loss": 0.0027, "step": 702 }, { "epoch": 2.6540642722117203, "grad_norm": 0.10015284270048141, "learning_rate": 3.4242904001241313e-06, "loss": 0.0006, "step": 703 }, { "epoch": 2.6578449905482042, "grad_norm": 0.44711926579475403, "learning_rate": 3.4191886060429837e-06, "loss": 0.0023, "step": 704 }, { "epoch": 2.661625708884688, "grad_norm": 0.21799851953983307, "learning_rate": 3.4140823817454826e-06, "loss": 0.0007, "step": 705 }, { "epoch": 2.665406427221172, "grad_norm": 0.3622993230819702, "learning_rate": 3.4089717518421168e-06, "loss": 0.0059, "step": 706 }, { "epoch": 2.6691871455576557, "grad_norm": 0.13052959740161896, "learning_rate": 3.403856740964604e-06, "loss": 0.0007, "step": 707 }, { "epoch": 2.67296786389414, "grad_norm": 0.5214902758598328, "learning_rate": 3.3987373737657793e-06, "loss": 0.0035, "step": 708 }, { "epoch": 2.6767485822306236, "grad_norm": 0.39454877376556396, "learning_rate": 3.393613674919473e-06, "loss": 0.0016, "step": 709 }, { "epoch": 2.6805293005671076, "grad_norm": 0.12668050825595856, "learning_rate": 3.388485669120395e-06, "loss": 0.0005, "step": 710 }, { "epoch": 2.6843100189035916, "grad_norm": 0.2751777172088623, "learning_rate": 3.3833533810840107e-06, "loss": 0.0009, "step": 711 }, { "epoch": 2.6880907372400755, "grad_norm": 0.4737643599510193, "learning_rate": 3.3782168355464263e-06, "loss": 0.0017, "step": 712 }, { "epoch": 2.6918714555765595, "grad_norm": 0.18790681660175323, "learning_rate": 3.3730760572642656e-06, "loss": 0.0016, "step": 713 }, { "epoch": 2.6956521739130435, "grad_norm": 0.25024181604385376, "learning_rate": 3.367931071014557e-06, "loss": 0.0012, "step": 714 }, { "epoch": 2.6994328922495274, "grad_norm": 0.18225879967212677, "learning_rate": 3.3627819015946062e-06, "loss": 0.0012, "step": 715 }, { "epoch": 2.7032136105860114, "grad_norm": 0.5009677410125732, "learning_rate": 3.3576285738218816e-06, "loss": 0.004, "step": 716 }, { "epoch": 2.7069943289224954, "grad_norm": 0.3747783601284027, "learning_rate": 3.3524711125338954e-06, "loss": 0.0026, "step": 717 }, { "epoch": 2.7107750472589793, "grad_norm": 1.0328412055969238, "learning_rate": 3.3473095425880795e-06, "loss": 0.0092, "step": 718 }, { "epoch": 2.7145557655954633, "grad_norm": 0.6312348246574402, "learning_rate": 3.3421438888616702e-06, "loss": 0.0035, "step": 719 }, { "epoch": 2.718336483931947, "grad_norm": 0.3027622103691101, "learning_rate": 3.336974176251586e-06, "loss": 0.0014, "step": 720 }, { "epoch": 2.7221172022684312, "grad_norm": 0.0532967671751976, "learning_rate": 3.331800429674308e-06, "loss": 0.0002, "step": 721 }, { "epoch": 2.7258979206049148, "grad_norm": 0.09550432115793228, "learning_rate": 3.3266226740657585e-06, "loss": 0.0005, "step": 722 }, { "epoch": 2.7296786389413987, "grad_norm": 0.1936490684747696, "learning_rate": 3.3214409343811844e-06, "loss": 0.0009, "step": 723 }, { "epoch": 2.7334593572778827, "grad_norm": 0.34296971559524536, "learning_rate": 3.3162552355950324e-06, "loss": 0.0019, "step": 724 }, { "epoch": 2.7372400756143667, "grad_norm": 0.5238574743270874, "learning_rate": 3.3110656027008333e-06, "loss": 0.0018, "step": 725 }, { "epoch": 2.7410207939508506, "grad_norm": 0.12622769176959991, "learning_rate": 3.3058720607110772e-06, "loss": 0.0004, "step": 726 }, { "epoch": 2.7448015122873346, "grad_norm": 1.7755227088928223, "learning_rate": 3.300674634657094e-06, "loss": 0.0025, "step": 727 }, { "epoch": 2.7485822306238186, "grad_norm": 0.39883720874786377, "learning_rate": 3.295473349588938e-06, "loss": 0.0031, "step": 728 }, { "epoch": 2.7523629489603025, "grad_norm": 0.4000013470649719, "learning_rate": 3.2902682305752593e-06, "loss": 0.0027, "step": 729 }, { "epoch": 2.7561436672967865, "grad_norm": 0.44501301646232605, "learning_rate": 3.2850593027031852e-06, "loss": 0.0058, "step": 730 }, { "epoch": 2.7599243856332705, "grad_norm": 0.1018594279885292, "learning_rate": 3.2798465910782053e-06, "loss": 0.0008, "step": 731 }, { "epoch": 2.7637051039697544, "grad_norm": 0.22918225824832916, "learning_rate": 3.2746301208240398e-06, "loss": 0.0007, "step": 732 }, { "epoch": 2.767485822306238, "grad_norm": 0.22460971772670746, "learning_rate": 3.269409917082531e-06, "loss": 0.0046, "step": 733 }, { "epoch": 2.7712665406427224, "grad_norm": 0.027669047936797142, "learning_rate": 3.2641860050135097e-06, "loss": 0.0001, "step": 734 }, { "epoch": 2.775047258979206, "grad_norm": 0.17839884757995605, "learning_rate": 3.2589584097946823e-06, "loss": 0.001, "step": 735 }, { "epoch": 2.77882797731569, "grad_norm": 0.3255077302455902, "learning_rate": 3.253727156621508e-06, "loss": 0.0024, "step": 736 }, { "epoch": 2.782608695652174, "grad_norm": 0.04910814017057419, "learning_rate": 3.248492270707074e-06, "loss": 0.0003, "step": 737 }, { "epoch": 2.786389413988658, "grad_norm": 0.052469439804553986, "learning_rate": 3.2432537772819767e-06, "loss": 0.0003, "step": 738 }, { "epoch": 2.7901701323251418, "grad_norm": 0.8889063000679016, "learning_rate": 3.2380117015942014e-06, "loss": 0.0026, "step": 739 }, { "epoch": 2.7939508506616257, "grad_norm": 0.10238675773143768, "learning_rate": 3.2327660689089967e-06, "loss": 0.0006, "step": 740 }, { "epoch": 2.7977315689981097, "grad_norm": 0.4949992895126343, "learning_rate": 3.2275169045087563e-06, "loss": 0.0022, "step": 741 }, { "epoch": 2.8015122873345937, "grad_norm": 0.2096443474292755, "learning_rate": 3.2222642336928943e-06, "loss": 0.001, "step": 742 }, { "epoch": 2.8052930056710776, "grad_norm": 0.5116773247718811, "learning_rate": 3.217008081777726e-06, "loss": 0.005, "step": 743 }, { "epoch": 2.809073724007561, "grad_norm": 0.644146740436554, "learning_rate": 3.2117484740963445e-06, "loss": 0.0022, "step": 744 }, { "epoch": 2.8128544423440456, "grad_norm": 0.5464627146720886, "learning_rate": 3.2064854359984976e-06, "loss": 0.002, "step": 745 }, { "epoch": 2.816635160680529, "grad_norm": 0.56934654712677, "learning_rate": 3.201218992850468e-06, "loss": 0.0007, "step": 746 }, { "epoch": 2.820415879017013, "grad_norm": 0.7829223275184631, "learning_rate": 3.195949170034948e-06, "loss": 0.0048, "step": 747 }, { "epoch": 2.824196597353497, "grad_norm": 0.335639089345932, "learning_rate": 3.190675992950921e-06, "loss": 0.0008, "step": 748 }, { "epoch": 2.827977315689981, "grad_norm": 0.321363240480423, "learning_rate": 3.1853994870135364e-06, "loss": 0.002, "step": 749 }, { "epoch": 2.831758034026465, "grad_norm": 0.4974987506866455, "learning_rate": 3.180119677653986e-06, "loss": 0.0061, "step": 750 }, { "epoch": 2.835538752362949, "grad_norm": 0.03324643149971962, "learning_rate": 3.1748365903193844e-06, "loss": 0.0002, "step": 751 }, { "epoch": 2.839319470699433, "grad_norm": 0.3876146674156189, "learning_rate": 3.1695502504726468e-06, "loss": 0.004, "step": 752 }, { "epoch": 2.843100189035917, "grad_norm": 0.38967448472976685, "learning_rate": 3.1642606835923608e-06, "loss": 0.0027, "step": 753 }, { "epoch": 2.846880907372401, "grad_norm": 0.06486915051937103, "learning_rate": 3.1589679151726693e-06, "loss": 0.0002, "step": 754 }, { "epoch": 2.850661625708885, "grad_norm": 0.4168080985546112, "learning_rate": 3.1536719707231473e-06, "loss": 0.0031, "step": 755 }, { "epoch": 2.854442344045369, "grad_norm": 0.23055201768875122, "learning_rate": 3.1483728757686745e-06, "loss": 0.001, "step": 756 }, { "epoch": 2.8582230623818523, "grad_norm": 0.04968590661883354, "learning_rate": 3.143070655849317e-06, "loss": 0.0003, "step": 757 }, { "epoch": 2.8620037807183367, "grad_norm": 0.5554412007331848, "learning_rate": 3.137765336520201e-06, "loss": 0.0063, "step": 758 }, { "epoch": 2.8657844990548202, "grad_norm": 0.19094473123550415, "learning_rate": 3.1324569433513915e-06, "loss": 0.0011, "step": 759 }, { "epoch": 2.869565217391304, "grad_norm": 0.7190486192703247, "learning_rate": 3.1271455019277686e-06, "loss": 0.0049, "step": 760 }, { "epoch": 2.873345935727788, "grad_norm": 0.5520555377006531, "learning_rate": 3.1218310378489046e-06, "loss": 0.0017, "step": 761 }, { "epoch": 2.877126654064272, "grad_norm": 0.09796418994665146, "learning_rate": 3.116513576728939e-06, "loss": 0.0006, "step": 762 }, { "epoch": 2.880907372400756, "grad_norm": 0.18798820674419403, "learning_rate": 3.111193144196457e-06, "loss": 0.0013, "step": 763 }, { "epoch": 2.88468809073724, "grad_norm": 0.48156529664993286, "learning_rate": 3.1058697658943644e-06, "loss": 0.0012, "step": 764 }, { "epoch": 2.888468809073724, "grad_norm": 0.03461313247680664, "learning_rate": 3.100543467479766e-06, "loss": 0.0002, "step": 765 }, { "epoch": 2.892249527410208, "grad_norm": 0.3334561884403229, "learning_rate": 3.095214274623839e-06, "loss": 0.0022, "step": 766 }, { "epoch": 2.896030245746692, "grad_norm": 0.198794424533844, "learning_rate": 3.089882213011711e-06, "loss": 0.0012, "step": 767 }, { "epoch": 2.899810964083176, "grad_norm": 0.3802403509616852, "learning_rate": 3.0845473083423393e-06, "loss": 0.0013, "step": 768 }, { "epoch": 2.90359168241966, "grad_norm": 0.2745780944824219, "learning_rate": 3.0792095863283794e-06, "loss": 0.0023, "step": 769 }, { "epoch": 2.9073724007561434, "grad_norm": 0.12996114790439606, "learning_rate": 3.0738690726960683e-06, "loss": 0.001, "step": 770 }, { "epoch": 2.911153119092628, "grad_norm": 0.2239939421415329, "learning_rate": 3.068525793185097e-06, "loss": 0.0012, "step": 771 }, { "epoch": 2.9149338374291114, "grad_norm": 1.0930694341659546, "learning_rate": 3.0631797735484877e-06, "loss": 0.0013, "step": 772 }, { "epoch": 2.9187145557655954, "grad_norm": 0.179174542427063, "learning_rate": 3.057831039552469e-06, "loss": 0.0012, "step": 773 }, { "epoch": 2.9224952741020793, "grad_norm": 0.3305235505104065, "learning_rate": 3.0524796169763513e-06, "loss": 0.0017, "step": 774 }, { "epoch": 2.9262759924385633, "grad_norm": 0.24241851270198822, "learning_rate": 3.047125531612404e-06, "loss": 0.0006, "step": 775 }, { "epoch": 2.9300567107750473, "grad_norm": 0.2687188684940338, "learning_rate": 3.0417688092657296e-06, "loss": 0.0013, "step": 776 }, { "epoch": 2.9338374291115312, "grad_norm": 0.3197347819805145, "learning_rate": 3.03640947575414e-06, "loss": 0.0027, "step": 777 }, { "epoch": 2.937618147448015, "grad_norm": 0.6880815625190735, "learning_rate": 3.0310475569080346e-06, "loss": 0.004, "step": 778 }, { "epoch": 2.941398865784499, "grad_norm": 0.09015991538763046, "learning_rate": 3.0256830785702682e-06, "loss": 0.0005, "step": 779 }, { "epoch": 2.945179584120983, "grad_norm": 0.04749230667948723, "learning_rate": 3.020316066596037e-06, "loss": 0.0003, "step": 780 }, { "epoch": 2.9489603024574667, "grad_norm": 0.31878674030303955, "learning_rate": 3.014946546852746e-06, "loss": 0.0023, "step": 781 }, { "epoch": 2.952741020793951, "grad_norm": 0.30465060472488403, "learning_rate": 3.009574545219887e-06, "loss": 0.0017, "step": 782 }, { "epoch": 2.9565217391304346, "grad_norm": 0.5834708213806152, "learning_rate": 3.0042000875889136e-06, "loss": 0.0064, "step": 783 }, { "epoch": 2.9603024574669186, "grad_norm": 0.10668514668941498, "learning_rate": 2.998823199863118e-06, "loss": 0.0005, "step": 784 }, { "epoch": 2.9640831758034025, "grad_norm": 0.12301790714263916, "learning_rate": 2.993443907957502e-06, "loss": 0.0008, "step": 785 }, { "epoch": 2.9678638941398865, "grad_norm": 0.10479345917701721, "learning_rate": 2.988062237798657e-06, "loss": 0.0005, "step": 786 }, { "epoch": 2.9716446124763705, "grad_norm": 0.21987728774547577, "learning_rate": 2.982678215324638e-06, "loss": 0.0043, "step": 787 }, { "epoch": 2.9754253308128544, "grad_norm": 0.01853783242404461, "learning_rate": 2.9772918664848334e-06, "loss": 0.0001, "step": 788 }, { "epoch": 2.9792060491493384, "grad_norm": 0.06683525443077087, "learning_rate": 2.9719032172398484e-06, "loss": 0.0004, "step": 789 }, { "epoch": 2.9829867674858224, "grad_norm": 0.44726628065109253, "learning_rate": 2.9665122935613727e-06, "loss": 0.0039, "step": 790 }, { "epoch": 2.9867674858223063, "grad_norm": 0.08658285439014435, "learning_rate": 2.961119121432059e-06, "loss": 0.0004, "step": 791 }, { "epoch": 2.9905482041587903, "grad_norm": 0.04073711857199669, "learning_rate": 2.955723726845397e-06, "loss": 0.0002, "step": 792 }, { "epoch": 2.9943289224952743, "grad_norm": 0.4395197033882141, "learning_rate": 2.950326135805587e-06, "loss": 0.0028, "step": 793 }, { "epoch": 2.998109640831758, "grad_norm": 0.06479670107364655, "learning_rate": 2.944926374327418e-06, "loss": 0.0003, "step": 794 }, { "epoch": 3.0, "grad_norm": 0.8987540602684021, "learning_rate": 2.9395244684361366e-06, "loss": 0.0016, "step": 795 }, { "epoch": 3.003780718336484, "grad_norm": 0.17659112811088562, "learning_rate": 2.9341204441673267e-06, "loss": 0.001, "step": 796 }, { "epoch": 3.007561436672968, "grad_norm": 0.06824270635843277, "learning_rate": 2.928714327566781e-06, "loss": 0.0005, "step": 797 }, { "epoch": 3.011342155009452, "grad_norm": 0.11952294409275055, "learning_rate": 2.92330614469038e-06, "loss": 0.0009, "step": 798 }, { "epoch": 3.015122873345936, "grad_norm": 0.37841182947158813, "learning_rate": 2.917895921603958e-06, "loss": 0.0016, "step": 799 }, { "epoch": 3.01890359168242, "grad_norm": 0.24612008035182953, "learning_rate": 2.9124836843831873e-06, "loss": 0.0008, "step": 800 }, { "epoch": 3.022684310018904, "grad_norm": 0.09307125955820084, "learning_rate": 2.9070694591134435e-06, "loss": 0.0005, "step": 801 }, { "epoch": 3.0264650283553873, "grad_norm": 0.154510959982872, "learning_rate": 2.901653271889687e-06, "loss": 0.0008, "step": 802 }, { "epoch": 3.0302457466918713, "grad_norm": 0.07291584461927414, "learning_rate": 2.896235148816333e-06, "loss": 0.0005, "step": 803 }, { "epoch": 3.0340264650283553, "grad_norm": 0.20764315128326416, "learning_rate": 2.8908151160071266e-06, "loss": 0.0011, "step": 804 }, { "epoch": 3.0378071833648392, "grad_norm": 0.08480814844369888, "learning_rate": 2.885393199585019e-06, "loss": 0.0004, "step": 805 }, { "epoch": 3.041587901701323, "grad_norm": 0.15913569927215576, "learning_rate": 2.879969425682037e-06, "loss": 0.0006, "step": 806 }, { "epoch": 3.045368620037807, "grad_norm": 0.0319024957716465, "learning_rate": 2.8745438204391623e-06, "loss": 0.0003, "step": 807 }, { "epoch": 3.049149338374291, "grad_norm": 0.24450285732746124, "learning_rate": 2.8691164100062035e-06, "loss": 0.0008, "step": 808 }, { "epoch": 3.052930056710775, "grad_norm": 0.4530331492424011, "learning_rate": 2.8636872205416666e-06, "loss": 0.0028, "step": 809 }, { "epoch": 3.056710775047259, "grad_norm": 0.3292180895805359, "learning_rate": 2.8582562782126356e-06, "loss": 0.001, "step": 810 }, { "epoch": 3.060491493383743, "grad_norm": 0.2506657540798187, "learning_rate": 2.8528236091946406e-06, "loss": 0.0015, "step": 811 }, { "epoch": 3.064272211720227, "grad_norm": 0.2139052301645279, "learning_rate": 2.847389239671533e-06, "loss": 0.0016, "step": 812 }, { "epoch": 3.068052930056711, "grad_norm": 0.31210198998451233, "learning_rate": 2.8419531958353636e-06, "loss": 0.0017, "step": 813 }, { "epoch": 3.0718336483931945, "grad_norm": 0.5276247262954712, "learning_rate": 2.836515503886249e-06, "loss": 0.0013, "step": 814 }, { "epoch": 3.0756143667296785, "grad_norm": 0.13789638876914978, "learning_rate": 2.831076190032252e-06, "loss": 0.0004, "step": 815 }, { "epoch": 3.0793950850661624, "grad_norm": 0.26674172282218933, "learning_rate": 2.825635280489252e-06, "loss": 0.0012, "step": 816 }, { "epoch": 3.0831758034026464, "grad_norm": 0.6654260158538818, "learning_rate": 2.820192801480817e-06, "loss": 0.0007, "step": 817 }, { "epoch": 3.0869565217391304, "grad_norm": 0.1329723447561264, "learning_rate": 2.8147487792380835e-06, "loss": 0.0003, "step": 818 }, { "epoch": 3.0907372400756143, "grad_norm": 0.2627672255039215, "learning_rate": 2.809303239999621e-06, "loss": 0.001, "step": 819 }, { "epoch": 3.0945179584120983, "grad_norm": 0.28436851501464844, "learning_rate": 2.8038562100113135e-06, "loss": 0.0017, "step": 820 }, { "epoch": 3.0982986767485823, "grad_norm": 0.25146713852882385, "learning_rate": 2.7984077155262305e-06, "loss": 0.0011, "step": 821 }, { "epoch": 3.1020793950850662, "grad_norm": 0.1561456173658371, "learning_rate": 2.7929577828044972e-06, "loss": 0.0003, "step": 822 }, { "epoch": 3.10586011342155, "grad_norm": 0.5860417485237122, "learning_rate": 2.787506438113173e-06, "loss": 0.0039, "step": 823 }, { "epoch": 3.109640831758034, "grad_norm": 0.3568424582481384, "learning_rate": 2.7820537077261236e-06, "loss": 0.0037, "step": 824 }, { "epoch": 3.113421550094518, "grad_norm": 0.05923071876168251, "learning_rate": 2.7765996179238885e-06, "loss": 0.0003, "step": 825 }, { "epoch": 3.1172022684310017, "grad_norm": 0.1132260411977768, "learning_rate": 2.771144194993564e-06, "loss": 0.0007, "step": 826 }, { "epoch": 3.1209829867674856, "grad_norm": 0.10238923132419586, "learning_rate": 2.7656874652286703e-06, "loss": 0.0003, "step": 827 }, { "epoch": 3.1247637051039696, "grad_norm": 0.05917969346046448, "learning_rate": 2.7602294549290242e-06, "loss": 0.0003, "step": 828 }, { "epoch": 3.1285444234404536, "grad_norm": 0.5441000461578369, "learning_rate": 2.754770190400618e-06, "loss": 0.0019, "step": 829 }, { "epoch": 3.1323251417769375, "grad_norm": 0.22579485177993774, "learning_rate": 2.7493096979554858e-06, "loss": 0.0007, "step": 830 }, { "epoch": 3.1361058601134215, "grad_norm": 0.319571852684021, "learning_rate": 2.7438480039115807e-06, "loss": 0.005, "step": 831 }, { "epoch": 3.1398865784499055, "grad_norm": 0.2539418935775757, "learning_rate": 2.7383851345926478e-06, "loss": 0.001, "step": 832 }, { "epoch": 3.1436672967863895, "grad_norm": 0.13025632500648499, "learning_rate": 2.732921116328097e-06, "loss": 0.0003, "step": 833 }, { "epoch": 3.1474480151228734, "grad_norm": 0.1206442266702652, "learning_rate": 2.727455975452874e-06, "loss": 0.0004, "step": 834 }, { "epoch": 3.1512287334593574, "grad_norm": 0.5282520651817322, "learning_rate": 2.721989738307337e-06, "loss": 0.0006, "step": 835 }, { "epoch": 3.1550094517958414, "grad_norm": 0.08641062676906586, "learning_rate": 2.7165224312371276e-06, "loss": 0.0003, "step": 836 }, { "epoch": 3.1587901701323253, "grad_norm": 0.1159621998667717, "learning_rate": 2.711054080593043e-06, "loss": 0.0005, "step": 837 }, { "epoch": 3.1625708884688093, "grad_norm": 0.35232606530189514, "learning_rate": 2.705584712730911e-06, "loss": 0.0024, "step": 838 }, { "epoch": 3.166351606805293, "grad_norm": 0.07070118933916092, "learning_rate": 2.700114354011461e-06, "loss": 0.0003, "step": 839 }, { "epoch": 3.170132325141777, "grad_norm": 0.3711482286453247, "learning_rate": 2.694643030800201e-06, "loss": 0.0029, "step": 840 }, { "epoch": 3.1739130434782608, "grad_norm": 0.23828861117362976, "learning_rate": 2.6891707694672833e-06, "loss": 0.0004, "step": 841 }, { "epoch": 3.1776937618147447, "grad_norm": 0.7307355999946594, "learning_rate": 2.6836975963873846e-06, "loss": 0.0043, "step": 842 }, { "epoch": 3.1814744801512287, "grad_norm": 0.15122373402118683, "learning_rate": 2.6782235379395766e-06, "loss": 0.0005, "step": 843 }, { "epoch": 3.1852551984877127, "grad_norm": 0.2806379199028015, "learning_rate": 2.672748620507195e-06, "loss": 0.0011, "step": 844 }, { "epoch": 3.1890359168241966, "grad_norm": 0.14764343202114105, "learning_rate": 2.6672728704777184e-06, "loss": 0.0022, "step": 845 }, { "epoch": 3.1928166351606806, "grad_norm": 0.13294583559036255, "learning_rate": 2.6617963142426373e-06, "loss": 0.0005, "step": 846 }, { "epoch": 3.1965973534971646, "grad_norm": 0.18426863849163055, "learning_rate": 2.656318978197327e-06, "loss": 0.0008, "step": 847 }, { "epoch": 3.2003780718336485, "grad_norm": 0.11875985562801361, "learning_rate": 2.650840888740923e-06, "loss": 0.0005, "step": 848 }, { "epoch": 3.2041587901701325, "grad_norm": 0.07550536841154099, "learning_rate": 2.6453620722761897e-06, "loss": 0.0003, "step": 849 }, { "epoch": 3.207939508506616, "grad_norm": 0.7213388681411743, "learning_rate": 2.6398825552093982e-06, "loss": 0.0022, "step": 850 }, { "epoch": 3.2117202268431, "grad_norm": 0.23421509563922882, "learning_rate": 2.634402363950195e-06, "loss": 0.0013, "step": 851 }, { "epoch": 3.215500945179584, "grad_norm": 0.7003602385520935, "learning_rate": 2.6289215249114753e-06, "loss": 0.004, "step": 852 }, { "epoch": 3.219281663516068, "grad_norm": 0.0813993588089943, "learning_rate": 2.6234400645092576e-06, "loss": 0.0002, "step": 853 }, { "epoch": 3.223062381852552, "grad_norm": 0.2848401665687561, "learning_rate": 2.6179580091625543e-06, "loss": 0.0011, "step": 854 }, { "epoch": 3.226843100189036, "grad_norm": 0.7367457151412964, "learning_rate": 2.6124753852932457e-06, "loss": 0.0033, "step": 855 }, { "epoch": 3.23062381852552, "grad_norm": 0.06620590388774872, "learning_rate": 2.6069922193259523e-06, "loss": 0.0002, "step": 856 }, { "epoch": 3.234404536862004, "grad_norm": 0.20744609832763672, "learning_rate": 2.6015085376879063e-06, "loss": 0.0009, "step": 857 }, { "epoch": 3.2381852551984878, "grad_norm": 0.9347597360610962, "learning_rate": 2.596024366808827e-06, "loss": 0.002, "step": 858 }, { "epoch": 3.2419659735349717, "grad_norm": 0.37228289246559143, "learning_rate": 2.590539733120791e-06, "loss": 0.002, "step": 859 }, { "epoch": 3.2457466918714557, "grad_norm": 0.084204763174057, "learning_rate": 2.5850546630581048e-06, "loss": 0.0004, "step": 860 }, { "epoch": 3.2495274102079397, "grad_norm": 0.18583443760871887, "learning_rate": 2.579569183057179e-06, "loss": 0.0006, "step": 861 }, { "epoch": 3.2533081285444236, "grad_norm": 0.1771743893623352, "learning_rate": 2.5740833195563996e-06, "loss": 0.0004, "step": 862 }, { "epoch": 3.257088846880907, "grad_norm": 0.4152580201625824, "learning_rate": 2.5685970989960008e-06, "loss": 0.0014, "step": 863 }, { "epoch": 3.260869565217391, "grad_norm": 0.03394977003335953, "learning_rate": 2.5631105478179385e-06, "loss": 0.0001, "step": 864 }, { "epoch": 3.264650283553875, "grad_norm": 0.014686000533401966, "learning_rate": 2.55762369246576e-06, "loss": 0.0001, "step": 865 }, { "epoch": 3.268431001890359, "grad_norm": 0.1269843876361847, "learning_rate": 2.552136559384482e-06, "loss": 0.0006, "step": 866 }, { "epoch": 3.272211720226843, "grad_norm": 0.0958874449133873, "learning_rate": 2.5466491750204564e-06, "loss": 0.0005, "step": 867 }, { "epoch": 3.275992438563327, "grad_norm": 0.01814945414662361, "learning_rate": 2.541161565821248e-06, "loss": 0.0001, "step": 868 }, { "epoch": 3.279773156899811, "grad_norm": 0.06434807926416397, "learning_rate": 2.535673758235506e-06, "loss": 0.0002, "step": 869 }, { "epoch": 3.283553875236295, "grad_norm": 0.7588925957679749, "learning_rate": 2.5301857787128344e-06, "loss": 0.0023, "step": 870 }, { "epoch": 3.287334593572779, "grad_norm": 0.13211266696453094, "learning_rate": 2.5246976537036646e-06, "loss": 0.0005, "step": 871 }, { "epoch": 3.291115311909263, "grad_norm": 0.28094759583473206, "learning_rate": 2.5192094096591323e-06, "loss": 0.0021, "step": 872 }, { "epoch": 3.294896030245747, "grad_norm": 0.1784561425447464, "learning_rate": 2.5137210730309445e-06, "loss": 0.0006, "step": 873 }, { "epoch": 3.298676748582231, "grad_norm": 0.036693841218948364, "learning_rate": 2.508232670271257e-06, "loss": 0.0002, "step": 874 }, { "epoch": 3.302457466918715, "grad_norm": 0.2820994257926941, "learning_rate": 2.502744227832541e-06, "loss": 0.0022, "step": 875 }, { "epoch": 3.3062381852551983, "grad_norm": 0.09304822981357574, "learning_rate": 2.49725577216746e-06, "loss": 0.0004, "step": 876 }, { "epoch": 3.3100189035916823, "grad_norm": 0.06585922092199326, "learning_rate": 2.4917673297287443e-06, "loss": 0.0004, "step": 877 }, { "epoch": 3.3137996219281662, "grad_norm": 0.2266349196434021, "learning_rate": 2.486278926969056e-06, "loss": 0.0012, "step": 878 }, { "epoch": 3.31758034026465, "grad_norm": 0.4772503674030304, "learning_rate": 2.480790590340869e-06, "loss": 0.0015, "step": 879 }, { "epoch": 3.321361058601134, "grad_norm": 0.05545281618833542, "learning_rate": 2.4753023462963363e-06, "loss": 0.0003, "step": 880 }, { "epoch": 3.325141776937618, "grad_norm": 0.16695745289325714, "learning_rate": 2.4698142212871665e-06, "loss": 0.0015, "step": 881 }, { "epoch": 3.328922495274102, "grad_norm": 0.03031153976917267, "learning_rate": 2.4643262417644943e-06, "loss": 0.0001, "step": 882 }, { "epoch": 3.332703213610586, "grad_norm": 0.07497631758451462, "learning_rate": 2.458838434178752e-06, "loss": 0.0002, "step": 883 }, { "epoch": 3.33648393194707, "grad_norm": 0.24848903715610504, "learning_rate": 2.4533508249795445e-06, "loss": 0.0021, "step": 884 }, { "epoch": 3.340264650283554, "grad_norm": 0.07428223639726639, "learning_rate": 2.4478634406155188e-06, "loss": 0.0003, "step": 885 }, { "epoch": 3.344045368620038, "grad_norm": 0.37231776118278503, "learning_rate": 2.4423763075342404e-06, "loss": 0.0015, "step": 886 }, { "epoch": 3.3478260869565215, "grad_norm": 0.039045173674821854, "learning_rate": 2.4368894521820623e-06, "loss": 0.0002, "step": 887 }, { "epoch": 3.351606805293006, "grad_norm": 0.06308232247829437, "learning_rate": 2.4314029010040005e-06, "loss": 0.0002, "step": 888 }, { "epoch": 3.3553875236294894, "grad_norm": 0.06569451838731766, "learning_rate": 2.4259166804436008e-06, "loss": 0.0002, "step": 889 }, { "epoch": 3.3591682419659734, "grad_norm": 0.10555455833673477, "learning_rate": 2.420430816942821e-06, "loss": 0.0003, "step": 890 }, { "epoch": 3.3629489603024574, "grad_norm": 0.15522027015686035, "learning_rate": 2.414945336941896e-06, "loss": 0.0006, "step": 891 }, { "epoch": 3.3667296786389413, "grad_norm": 0.36748412251472473, "learning_rate": 2.4094602668792092e-06, "loss": 0.0013, "step": 892 }, { "epoch": 3.3705103969754253, "grad_norm": 0.05719131976366043, "learning_rate": 2.403975633191174e-06, "loss": 0.0002, "step": 893 }, { "epoch": 3.3742911153119093, "grad_norm": 0.5186507105827332, "learning_rate": 2.3984914623120945e-06, "loss": 0.0014, "step": 894 }, { "epoch": 3.3780718336483933, "grad_norm": 0.1489984691143036, "learning_rate": 2.393007780674049e-06, "loss": 0.0007, "step": 895 }, { "epoch": 3.381852551984877, "grad_norm": 0.3767106533050537, "learning_rate": 2.387524614706755e-06, "loss": 0.0013, "step": 896 }, { "epoch": 3.385633270321361, "grad_norm": 0.10746520757675171, "learning_rate": 2.3820419908374465e-06, "loss": 0.0004, "step": 897 }, { "epoch": 3.389413988657845, "grad_norm": 0.05779121071100235, "learning_rate": 2.376559935490743e-06, "loss": 0.0002, "step": 898 }, { "epoch": 3.393194706994329, "grad_norm": 0.12964102625846863, "learning_rate": 2.3710784750885247e-06, "loss": 0.0003, "step": 899 }, { "epoch": 3.3969754253308126, "grad_norm": 0.20334431529045105, "learning_rate": 2.3655976360498053e-06, "loss": 0.0003, "step": 900 }, { "epoch": 3.4007561436672966, "grad_norm": 0.2376059740781784, "learning_rate": 2.3601174447906017e-06, "loss": 0.0023, "step": 901 }, { "epoch": 3.4045368620037806, "grad_norm": 0.34268078207969666, "learning_rate": 2.3546379277238107e-06, "loss": 0.0008, "step": 902 }, { "epoch": 3.4083175803402646, "grad_norm": 0.1896052211523056, "learning_rate": 2.3491591112590776e-06, "loss": 0.0006, "step": 903 }, { "epoch": 3.4120982986767485, "grad_norm": 0.6880627870559692, "learning_rate": 2.3436810218026744e-06, "loss": 0.0006, "step": 904 }, { "epoch": 3.4158790170132325, "grad_norm": 0.7180553674697876, "learning_rate": 2.3382036857573635e-06, "loss": 0.0011, "step": 905 }, { "epoch": 3.4196597353497165, "grad_norm": 0.708527684211731, "learning_rate": 2.3327271295222824e-06, "loss": 0.0048, "step": 906 }, { "epoch": 3.4234404536862004, "grad_norm": 0.30467963218688965, "learning_rate": 2.3272513794928055e-06, "loss": 0.002, "step": 907 }, { "epoch": 3.4272211720226844, "grad_norm": 0.024917298927903175, "learning_rate": 2.3217764620604234e-06, "loss": 0.0001, "step": 908 }, { "epoch": 3.4310018903591684, "grad_norm": 0.03179596737027168, "learning_rate": 2.316302403612616e-06, "loss": 0.0001, "step": 909 }, { "epoch": 3.4347826086956523, "grad_norm": 0.2751535475254059, "learning_rate": 2.3108292305327176e-06, "loss": 0.0012, "step": 910 }, { "epoch": 3.4385633270321363, "grad_norm": 0.26875996589660645, "learning_rate": 2.3053569691998003e-06, "loss": 0.0008, "step": 911 }, { "epoch": 3.4423440453686203, "grad_norm": 0.2487889677286148, "learning_rate": 2.2998856459885397e-06, "loss": 0.0011, "step": 912 }, { "epoch": 3.446124763705104, "grad_norm": 0.3235238492488861, "learning_rate": 2.29441528726909e-06, "loss": 0.0013, "step": 913 }, { "epoch": 3.4499054820415878, "grad_norm": 0.08037448674440384, "learning_rate": 2.288945919406958e-06, "loss": 0.0004, "step": 914 }, { "epoch": 3.4536862003780717, "grad_norm": 0.14094021916389465, "learning_rate": 2.2834775687628732e-06, "loss": 0.0005, "step": 915 }, { "epoch": 3.4574669187145557, "grad_norm": 0.24907419085502625, "learning_rate": 2.278010261692663e-06, "loss": 0.0006, "step": 916 }, { "epoch": 3.4612476370510397, "grad_norm": 0.06824265420436859, "learning_rate": 2.272544024547126e-06, "loss": 0.0002, "step": 917 }, { "epoch": 3.4650283553875236, "grad_norm": 0.22965072095394135, "learning_rate": 2.267078883671904e-06, "loss": 0.0008, "step": 918 }, { "epoch": 3.4688090737240076, "grad_norm": 0.3709386885166168, "learning_rate": 2.2616148654073527e-06, "loss": 0.0009, "step": 919 }, { "epoch": 3.4725897920604916, "grad_norm": 0.24136681854724884, "learning_rate": 2.2561519960884205e-06, "loss": 0.0006, "step": 920 }, { "epoch": 3.4763705103969755, "grad_norm": 0.3291138708591461, "learning_rate": 2.2506903020445146e-06, "loss": 0.0049, "step": 921 }, { "epoch": 3.4801512287334595, "grad_norm": 0.14625725150108337, "learning_rate": 2.245229809599383e-06, "loss": 0.0003, "step": 922 }, { "epoch": 3.4839319470699435, "grad_norm": 0.3351140022277832, "learning_rate": 2.2397705450709766e-06, "loss": 0.0009, "step": 923 }, { "epoch": 3.487712665406427, "grad_norm": 0.2638176381587982, "learning_rate": 2.2343125347713305e-06, "loss": 0.0006, "step": 924 }, { "epoch": 3.491493383742911, "grad_norm": 0.5429024696350098, "learning_rate": 2.2288558050064366e-06, "loss": 0.0006, "step": 925 }, { "epoch": 3.495274102079395, "grad_norm": 0.8680464625358582, "learning_rate": 2.2234003820761123e-06, "loss": 0.0004, "step": 926 }, { "epoch": 3.499054820415879, "grad_norm": 0.7402521967887878, "learning_rate": 2.2179462922738777e-06, "loss": 0.0018, "step": 927 }, { "epoch": 3.502835538752363, "grad_norm": 0.12084399908781052, "learning_rate": 2.212493561886827e-06, "loss": 0.0003, "step": 928 }, { "epoch": 3.506616257088847, "grad_norm": 0.02428770810365677, "learning_rate": 2.207042217195503e-06, "loss": 0.0001, "step": 929 }, { "epoch": 3.510396975425331, "grad_norm": 0.19316118955612183, "learning_rate": 2.2015922844737703e-06, "loss": 0.0008, "step": 930 }, { "epoch": 3.5141776937618148, "grad_norm": 1.0070908069610596, "learning_rate": 2.196143789988687e-06, "loss": 0.0034, "step": 931 }, { "epoch": 3.5179584120982987, "grad_norm": 0.3398049771785736, "learning_rate": 2.1906967600003796e-06, "loss": 0.0013, "step": 932 }, { "epoch": 3.5217391304347827, "grad_norm": 0.02249988168478012, "learning_rate": 2.185251220761917e-06, "loss": 0.0002, "step": 933 }, { "epoch": 3.5255198487712667, "grad_norm": 0.10283073782920837, "learning_rate": 2.1798071985191833e-06, "loss": 0.0002, "step": 934 }, { "epoch": 3.52930056710775, "grad_norm": 2.55683970451355, "learning_rate": 2.174364719510748e-06, "loss": 0.0016, "step": 935 }, { "epoch": 3.5330812854442346, "grad_norm": 0.06272849440574646, "learning_rate": 2.1689238099677482e-06, "loss": 0.0002, "step": 936 }, { "epoch": 3.536862003780718, "grad_norm": 0.09796170145273209, "learning_rate": 2.1634844961137517e-06, "loss": 0.0003, "step": 937 }, { "epoch": 3.5406427221172025, "grad_norm": 0.1700635552406311, "learning_rate": 2.1580468041646377e-06, "loss": 0.0009, "step": 938 }, { "epoch": 3.544423440453686, "grad_norm": 0.25389400124549866, "learning_rate": 2.1526107603284675e-06, "loss": 0.001, "step": 939 }, { "epoch": 3.54820415879017, "grad_norm": 0.13603591918945312, "learning_rate": 2.147176390805361e-06, "loss": 0.0005, "step": 940 }, { "epoch": 3.551984877126654, "grad_norm": 0.05718974024057388, "learning_rate": 2.1417437217873653e-06, "loss": 0.0002, "step": 941 }, { "epoch": 3.555765595463138, "grad_norm": 0.7403815984725952, "learning_rate": 2.1363127794583334e-06, "loss": 0.004, "step": 942 }, { "epoch": 3.559546313799622, "grad_norm": 0.48677870631217957, "learning_rate": 2.1308835899937974e-06, "loss": 0.0019, "step": 943 }, { "epoch": 3.563327032136106, "grad_norm": 0.35012948513031006, "learning_rate": 2.1254561795608377e-06, "loss": 0.001, "step": 944 }, { "epoch": 3.56710775047259, "grad_norm": 0.5069158673286438, "learning_rate": 2.1200305743179637e-06, "loss": 0.0029, "step": 945 }, { "epoch": 3.570888468809074, "grad_norm": 0.06929867714643478, "learning_rate": 2.1146068004149815e-06, "loss": 0.0002, "step": 946 }, { "epoch": 3.574669187145558, "grad_norm": 0.14685168862342834, "learning_rate": 2.109184883992874e-06, "loss": 0.0009, "step": 947 }, { "epoch": 3.5784499054820413, "grad_norm": 0.041475772857666016, "learning_rate": 2.1037648511836674e-06, "loss": 0.0002, "step": 948 }, { "epoch": 3.5822306238185257, "grad_norm": 0.30418428778648376, "learning_rate": 2.098346728110314e-06, "loss": 0.001, "step": 949 }, { "epoch": 3.5860113421550093, "grad_norm": 0.10565707087516785, "learning_rate": 2.0929305408865573e-06, "loss": 0.0004, "step": 950 }, { "epoch": 3.5897920604914932, "grad_norm": 0.29487159848213196, "learning_rate": 2.0875163156168127e-06, "loss": 0.0016, "step": 951 }, { "epoch": 3.593572778827977, "grad_norm": 0.2743481695652008, "learning_rate": 2.0821040783960422e-06, "loss": 0.0009, "step": 952 }, { "epoch": 3.597353497164461, "grad_norm": 0.8704341650009155, "learning_rate": 2.0766938553096204e-06, "loss": 0.0003, "step": 953 }, { "epoch": 3.601134215500945, "grad_norm": 0.11365697532892227, "learning_rate": 2.0712856724332193e-06, "loss": 0.0006, "step": 954 }, { "epoch": 3.604914933837429, "grad_norm": 0.018890921026468277, "learning_rate": 2.0658795558326745e-06, "loss": 0.0001, "step": 955 }, { "epoch": 3.608695652173913, "grad_norm": 0.20384187996387482, "learning_rate": 2.0604755315638646e-06, "loss": 0.0003, "step": 956 }, { "epoch": 3.612476370510397, "grad_norm": 0.12472031265497208, "learning_rate": 2.0550736256725827e-06, "loss": 0.0005, "step": 957 }, { "epoch": 3.616257088846881, "grad_norm": 0.039687398821115494, "learning_rate": 2.0496738641944132e-06, "loss": 0.0001, "step": 958 }, { "epoch": 3.620037807183365, "grad_norm": 0.34768256545066833, "learning_rate": 2.0442762731546033e-06, "loss": 0.002, "step": 959 }, { "epoch": 3.623818525519849, "grad_norm": 0.2582302689552307, "learning_rate": 2.038880878567941e-06, "loss": 0.0018, "step": 960 }, { "epoch": 3.6275992438563325, "grad_norm": 0.03289654478430748, "learning_rate": 2.0334877064386277e-06, "loss": 0.0001, "step": 961 }, { "epoch": 3.631379962192817, "grad_norm": 0.10774717479944229, "learning_rate": 2.028096782760152e-06, "loss": 0.0003, "step": 962 }, { "epoch": 3.6351606805293004, "grad_norm": 0.04548002406954765, "learning_rate": 2.022708133515168e-06, "loss": 0.0002, "step": 963 }, { "epoch": 3.6389413988657844, "grad_norm": 0.22105950117111206, "learning_rate": 2.017321784675363e-06, "loss": 0.0002, "step": 964 }, { "epoch": 3.6427221172022684, "grad_norm": 0.373381108045578, "learning_rate": 2.011937762201344e-06, "loss": 0.0025, "step": 965 }, { "epoch": 3.6465028355387523, "grad_norm": 0.019999120384454727, "learning_rate": 2.0065560920424994e-06, "loss": 0.0001, "step": 966 }, { "epoch": 3.6502835538752363, "grad_norm": 0.053666889667510986, "learning_rate": 2.0011768001368836e-06, "loss": 0.0002, "step": 967 }, { "epoch": 3.6540642722117203, "grad_norm": 0.5051717758178711, "learning_rate": 1.9957999124110872e-06, "loss": 0.0016, "step": 968 }, { "epoch": 3.6578449905482042, "grad_norm": 0.2403508871793747, "learning_rate": 1.9904254547801134e-06, "loss": 0.0008, "step": 969 }, { "epoch": 3.661625708884688, "grad_norm": 0.6390255093574524, "learning_rate": 1.9850534531472544e-06, "loss": 0.0018, "step": 970 }, { "epoch": 3.665406427221172, "grad_norm": 0.1011655256152153, "learning_rate": 1.979683933403963e-06, "loss": 0.0005, "step": 971 }, { "epoch": 3.6691871455576557, "grad_norm": 0.15766626596450806, "learning_rate": 1.9743169214297326e-06, "loss": 0.0008, "step": 972 }, { "epoch": 3.67296786389414, "grad_norm": 0.4769021272659302, "learning_rate": 1.9689524430919667e-06, "loss": 0.0025, "step": 973 }, { "epoch": 3.6767485822306236, "grad_norm": 0.10289421677589417, "learning_rate": 1.9635905242458607e-06, "loss": 0.0005, "step": 974 }, { "epoch": 3.6805293005671076, "grad_norm": 0.5679051280021667, "learning_rate": 1.958231190734271e-06, "loss": 0.0024, "step": 975 }, { "epoch": 3.6843100189035916, "grad_norm": 0.0657648965716362, "learning_rate": 1.9528744683875973e-06, "loss": 0.0004, "step": 976 }, { "epoch": 3.6880907372400755, "grad_norm": 1.5473754405975342, "learning_rate": 1.947520383023649e-06, "loss": 0.0007, "step": 977 }, { "epoch": 3.6918714555765595, "grad_norm": 0.36639949679374695, "learning_rate": 1.942168960447531e-06, "loss": 0.001, "step": 978 }, { "epoch": 3.6956521739130435, "grad_norm": 0.12034526467323303, "learning_rate": 1.9368202264515127e-06, "loss": 0.0004, "step": 979 }, { "epoch": 3.6994328922495274, "grad_norm": 0.19727705419063568, "learning_rate": 1.931474206814903e-06, "loss": 0.0007, "step": 980 }, { "epoch": 3.7032136105860114, "grad_norm": 0.6420516967773438, "learning_rate": 1.926130927303933e-06, "loss": 0.0033, "step": 981 }, { "epoch": 3.7069943289224954, "grad_norm": 0.02208607830107212, "learning_rate": 1.9207904136716215e-06, "loss": 0.0001, "step": 982 }, { "epoch": 3.7107750472589793, "grad_norm": 0.33331099152565, "learning_rate": 1.915452691657662e-06, "loss": 0.0025, "step": 983 }, { "epoch": 3.7145557655954633, "grad_norm": 0.5736892223358154, "learning_rate": 1.9101177869882893e-06, "loss": 0.0035, "step": 984 }, { "epoch": 3.718336483931947, "grad_norm": 0.4493763744831085, "learning_rate": 1.9047857253761622e-06, "loss": 0.0047, "step": 985 }, { "epoch": 3.7221172022684312, "grad_norm": 0.11402516812086105, "learning_rate": 1.8994565325202346e-06, "loss": 0.0003, "step": 986 }, { "epoch": 3.7258979206049148, "grad_norm": 0.19264838099479675, "learning_rate": 1.8941302341056356e-06, "loss": 0.0007, "step": 987 }, { "epoch": 3.7296786389413987, "grad_norm": 0.5359593629837036, "learning_rate": 1.8888068558035435e-06, "loss": 0.0021, "step": 988 }, { "epoch": 3.7334593572778827, "grad_norm": 0.049667809158563614, "learning_rate": 1.8834864232710614e-06, "loss": 0.0001, "step": 989 }, { "epoch": 3.7372400756143667, "grad_norm": 0.1993870735168457, "learning_rate": 1.8781689621510962e-06, "loss": 0.0008, "step": 990 }, { "epoch": 3.7410207939508506, "grad_norm": 0.009530361741781235, "learning_rate": 1.8728544980722318e-06, "loss": 0.0001, "step": 991 }, { "epoch": 3.7448015122873346, "grad_norm": 0.20599816739559174, "learning_rate": 1.8675430566486102e-06, "loss": 0.0007, "step": 992 }, { "epoch": 3.7485822306238186, "grad_norm": 5.536590099334717, "learning_rate": 1.8622346634798e-06, "loss": 0.0015, "step": 993 }, { "epoch": 3.7523629489603025, "grad_norm": 0.10816589742898941, "learning_rate": 1.8569293441506847e-06, "loss": 0.0003, "step": 994 }, { "epoch": 3.7561436672967865, "grad_norm": 0.020855804905295372, "learning_rate": 1.8516271242313267e-06, "loss": 0.0001, "step": 995 }, { "epoch": 3.7599243856332705, "grad_norm": 0.05966445803642273, "learning_rate": 1.8463280292768529e-06, "loss": 0.0002, "step": 996 }, { "epoch": 3.7637051039697544, "grad_norm": 0.28455814719200134, "learning_rate": 1.8410320848273315e-06, "loss": 0.0008, "step": 997 }, { "epoch": 3.767485822306238, "grad_norm": 0.5421791076660156, "learning_rate": 1.8357393164076403e-06, "loss": 0.0019, "step": 998 }, { "epoch": 3.7712665406427224, "grad_norm": 2.5801000595092773, "learning_rate": 1.8304497495273543e-06, "loss": 0.0024, "step": 999 }, { "epoch": 3.775047258979206, "grad_norm": 0.34670326113700867, "learning_rate": 1.8251634096806158e-06, "loss": 0.0019, "step": 1000 }, { "epoch": 3.77882797731569, "grad_norm": 0.6210291385650635, "learning_rate": 1.8198803223460148e-06, "loss": 0.0023, "step": 1001 }, { "epoch": 3.782608695652174, "grad_norm": 0.49758532643318176, "learning_rate": 1.8146005129864644e-06, "loss": 0.0033, "step": 1002 }, { "epoch": 3.786389413988658, "grad_norm": 0.05057361349463463, "learning_rate": 1.8093240070490791e-06, "loss": 0.0002, "step": 1003 }, { "epoch": 3.7901701323251418, "grad_norm": 0.8116357922554016, "learning_rate": 1.8040508299650524e-06, "loss": 0.0036, "step": 1004 }, { "epoch": 3.7939508506616257, "grad_norm": 0.2625158131122589, "learning_rate": 1.7987810071495327e-06, "loss": 0.0008, "step": 1005 }, { "epoch": 3.7977315689981097, "grad_norm": 0.34827500581741333, "learning_rate": 1.793514564001503e-06, "loss": 0.0028, "step": 1006 }, { "epoch": 3.8015122873345937, "grad_norm": 0.053880367428064346, "learning_rate": 1.788251525903656e-06, "loss": 0.0002, "step": 1007 }, { "epoch": 3.8052930056710776, "grad_norm": 0.22675637900829315, "learning_rate": 1.7829919182222752e-06, "loss": 0.0014, "step": 1008 }, { "epoch": 3.809073724007561, "grad_norm": 0.40593239665031433, "learning_rate": 1.7777357663071063e-06, "loss": 0.0035, "step": 1009 }, { "epoch": 3.8128544423440456, "grad_norm": 0.6634936332702637, "learning_rate": 1.7724830954912454e-06, "loss": 0.0016, "step": 1010 }, { "epoch": 3.816635160680529, "grad_norm": 0.3126247823238373, "learning_rate": 1.767233931091004e-06, "loss": 0.0008, "step": 1011 }, { "epoch": 3.820415879017013, "grad_norm": 0.271623432636261, "learning_rate": 1.7619882984057988e-06, "loss": 0.0014, "step": 1012 }, { "epoch": 3.824196597353497, "grad_norm": 0.4311028718948364, "learning_rate": 1.7567462227180241e-06, "loss": 0.0027, "step": 1013 }, { "epoch": 3.827977315689981, "grad_norm": 0.09972798079252243, "learning_rate": 1.7515077292929267e-06, "loss": 0.0006, "step": 1014 }, { "epoch": 3.831758034026465, "grad_norm": 0.13386182487010956, "learning_rate": 1.746272843378493e-06, "loss": 0.0004, "step": 1015 }, { "epoch": 3.835538752362949, "grad_norm": 0.5869089961051941, "learning_rate": 1.7410415902053179e-06, "loss": 0.0017, "step": 1016 }, { "epoch": 3.839319470699433, "grad_norm": 0.22970812022686005, "learning_rate": 1.7358139949864913e-06, "loss": 0.0009, "step": 1017 }, { "epoch": 3.843100189035917, "grad_norm": 0.11757975071668625, "learning_rate": 1.73059008291747e-06, "loss": 0.0004, "step": 1018 }, { "epoch": 3.846880907372401, "grad_norm": 1.447178840637207, "learning_rate": 1.7253698791759604e-06, "loss": 0.0017, "step": 1019 }, { "epoch": 3.850661625708885, "grad_norm": 0.49450984597206116, "learning_rate": 1.720153408921796e-06, "loss": 0.0037, "step": 1020 }, { "epoch": 3.854442344045369, "grad_norm": 0.21628008782863617, "learning_rate": 1.7149406972968148e-06, "loss": 0.001, "step": 1021 }, { "epoch": 3.8582230623818523, "grad_norm": 0.49233701825141907, "learning_rate": 1.7097317694247418e-06, "loss": 0.0045, "step": 1022 }, { "epoch": 3.8620037807183367, "grad_norm": 0.016612421721220016, "learning_rate": 1.704526650411062e-06, "loss": 0.0001, "step": 1023 }, { "epoch": 3.8657844990548202, "grad_norm": 0.12304850667715073, "learning_rate": 1.6993253653429064e-06, "loss": 0.0004, "step": 1024 }, { "epoch": 3.869565217391304, "grad_norm": 0.2783352732658386, "learning_rate": 1.6941279392889238e-06, "loss": 0.0043, "step": 1025 }, { "epoch": 3.873345935727788, "grad_norm": 0.2207585871219635, "learning_rate": 1.6889343972991682e-06, "loss": 0.0009, "step": 1026 }, { "epoch": 3.877126654064272, "grad_norm": 0.13213740289211273, "learning_rate": 1.6837447644049682e-06, "loss": 0.0006, "step": 1027 }, { "epoch": 3.880907372400756, "grad_norm": 0.05475637689232826, "learning_rate": 1.678559065618817e-06, "loss": 0.0003, "step": 1028 }, { "epoch": 3.88468809073724, "grad_norm": 0.0370832197368145, "learning_rate": 1.6733773259342424e-06, "loss": 0.0002, "step": 1029 }, { "epoch": 3.888468809073724, "grad_norm": 0.0499420128762722, "learning_rate": 1.6681995703256925e-06, "loss": 0.0002, "step": 1030 }, { "epoch": 3.892249527410208, "grad_norm": 0.664375901222229, "learning_rate": 1.6630258237484144e-06, "loss": 0.0007, "step": 1031 }, { "epoch": 3.896030245746692, "grad_norm": 0.04301537573337555, "learning_rate": 1.6578561111383298e-06, "loss": 0.0002, "step": 1032 }, { "epoch": 3.899810964083176, "grad_norm": 0.07999247312545776, "learning_rate": 1.6526904574119213e-06, "loss": 0.0004, "step": 1033 }, { "epoch": 3.90359168241966, "grad_norm": 0.7571427226066589, "learning_rate": 1.6475288874661053e-06, "loss": 0.0024, "step": 1034 }, { "epoch": 3.9073724007561434, "grad_norm": 0.02368680201470852, "learning_rate": 1.6423714261781196e-06, "loss": 0.0001, "step": 1035 }, { "epoch": 3.911153119092628, "grad_norm": 0.3289349377155304, "learning_rate": 1.6372180984053946e-06, "loss": 0.0006, "step": 1036 }, { "epoch": 3.9149338374291114, "grad_norm": 0.41246291995048523, "learning_rate": 1.632068928985444e-06, "loss": 0.0015, "step": 1037 }, { "epoch": 3.9187145557655954, "grad_norm": 0.4724266231060028, "learning_rate": 1.6269239427357348e-06, "loss": 0.0018, "step": 1038 }, { "epoch": 3.9224952741020793, "grad_norm": 0.18117232620716095, "learning_rate": 1.6217831644535743e-06, "loss": 0.0009, "step": 1039 }, { "epoch": 3.9262759924385633, "grad_norm": 0.45204848051071167, "learning_rate": 1.6166466189159901e-06, "loss": 0.0039, "step": 1040 }, { "epoch": 3.9300567107750473, "grad_norm": 0.09278274327516556, "learning_rate": 1.6115143308796054e-06, "loss": 0.0004, "step": 1041 }, { "epoch": 3.9338374291115312, "grad_norm": 0.07383128255605698, "learning_rate": 1.6063863250805279e-06, "loss": 0.0004, "step": 1042 }, { "epoch": 3.937618147448015, "grad_norm": 0.15998311340808868, "learning_rate": 1.6012626262342217e-06, "loss": 0.0005, "step": 1043 }, { "epoch": 3.941398865784499, "grad_norm": 0.3451702892780304, "learning_rate": 1.5961432590353973e-06, "loss": 0.0039, "step": 1044 }, { "epoch": 3.945179584120983, "grad_norm": 0.13052286207675934, "learning_rate": 1.5910282481578843e-06, "loss": 0.0002, "step": 1045 }, { "epoch": 3.9489603024574667, "grad_norm": 0.049013491719961166, "learning_rate": 1.5859176182545178e-06, "loss": 0.0003, "step": 1046 }, { "epoch": 3.952741020793951, "grad_norm": 0.9534972310066223, "learning_rate": 1.5808113939570175e-06, "loss": 0.0019, "step": 1047 }, { "epoch": 3.9565217391304346, "grad_norm": 0.7160435914993286, "learning_rate": 1.5757095998758693e-06, "loss": 0.0015, "step": 1048 }, { "epoch": 3.9603024574669186, "grad_norm": 0.3361499607563019, "learning_rate": 1.5706122606002078e-06, "loss": 0.001, "step": 1049 }, { "epoch": 3.9640831758034025, "grad_norm": 0.19443443417549133, "learning_rate": 1.5655194006976948e-06, "loss": 0.0018, "step": 1050 }, { "epoch": 3.9678638941398865, "grad_norm": 0.15781041979789734, "learning_rate": 1.5604310447144052e-06, "loss": 0.0008, "step": 1051 }, { "epoch": 3.9716446124763705, "grad_norm": 0.16736602783203125, "learning_rate": 1.5553472171747031e-06, "loss": 0.0007, "step": 1052 }, { "epoch": 3.9754253308128544, "grad_norm": 0.14986012876033783, "learning_rate": 1.5502679425811317e-06, "loss": 0.0003, "step": 1053 }, { "epoch": 3.9792060491493384, "grad_norm": 0.0438813790678978, "learning_rate": 1.5451932454142847e-06, "loss": 0.0002, "step": 1054 }, { "epoch": 3.9829867674858224, "grad_norm": 0.14121653139591217, "learning_rate": 1.5401231501326985e-06, "loss": 0.0007, "step": 1055 }, { "epoch": 3.9867674858223063, "grad_norm": 0.09703072160482407, "learning_rate": 1.535057681172727e-06, "loss": 0.0005, "step": 1056 }, { "epoch": 3.9905482041587903, "grad_norm": 0.18912367522716522, "learning_rate": 1.5299968629484277e-06, "loss": 0.0005, "step": 1057 }, { "epoch": 3.9943289224952743, "grad_norm": 0.08505091816186905, "learning_rate": 1.524940719851444e-06, "loss": 0.0003, "step": 1058 }, { "epoch": 3.998109640831758, "grad_norm": 0.024707768112421036, "learning_rate": 1.5198892762508843e-06, "loss": 0.0001, "step": 1059 }, { "epoch": 4.0, "grad_norm": 0.05147272348403931, "learning_rate": 1.5148425564932085e-06, "loss": 0.0001, "step": 1060 }, { "epoch": 4.0037807183364835, "grad_norm": 0.24065309762954712, "learning_rate": 1.509800584902108e-06, "loss": 0.0019, "step": 1061 }, { "epoch": 4.007561436672968, "grad_norm": 0.041670165956020355, "learning_rate": 1.504763385778391e-06, "loss": 0.0002, "step": 1062 }, { "epoch": 4.0113421550094515, "grad_norm": 0.1158413365483284, "learning_rate": 1.4997309833998608e-06, "loss": 0.0003, "step": 1063 }, { "epoch": 4.015122873345936, "grad_norm": 0.16893485188484192, "learning_rate": 1.4947034020212062e-06, "loss": 0.0008, "step": 1064 }, { "epoch": 4.018903591682419, "grad_norm": 0.23386147618293762, "learning_rate": 1.4896806658738747e-06, "loss": 0.0039, "step": 1065 }, { "epoch": 4.022684310018904, "grad_norm": 0.12124040722846985, "learning_rate": 1.4846627991659662e-06, "loss": 0.0005, "step": 1066 }, { "epoch": 4.026465028355387, "grad_norm": 0.1346861720085144, "learning_rate": 1.4796498260821104e-06, "loss": 0.0003, "step": 1067 }, { "epoch": 4.030245746691872, "grad_norm": 0.31912630796432495, "learning_rate": 1.4746417707833472e-06, "loss": 0.0023, "step": 1068 }, { "epoch": 4.034026465028355, "grad_norm": 0.18380481004714966, "learning_rate": 1.4696386574070203e-06, "loss": 0.0006, "step": 1069 }, { "epoch": 4.03780718336484, "grad_norm": 0.08836250752210617, "learning_rate": 1.4646405100666498e-06, "loss": 0.0005, "step": 1070 }, { "epoch": 4.041587901701323, "grad_norm": 0.1820770651102066, "learning_rate": 1.4596473528518236e-06, "loss": 0.0004, "step": 1071 }, { "epoch": 4.045368620037808, "grad_norm": 0.02899593859910965, "learning_rate": 1.4546592098280772e-06, "loss": 0.0001, "step": 1072 }, { "epoch": 4.049149338374291, "grad_norm": 0.16005228459835052, "learning_rate": 1.449676105036781e-06, "loss": 0.0006, "step": 1073 }, { "epoch": 4.052930056710775, "grad_norm": 0.29831090569496155, "learning_rate": 1.4446980624950201e-06, "loss": 0.0011, "step": 1074 }, { "epoch": 4.056710775047259, "grad_norm": 0.3436811566352844, "learning_rate": 1.4397251061954848e-06, "loss": 0.0016, "step": 1075 }, { "epoch": 4.060491493383743, "grad_norm": 0.3190964460372925, "learning_rate": 1.4347572601063477e-06, "loss": 0.001, "step": 1076 }, { "epoch": 4.064272211720227, "grad_norm": 0.27734869718551636, "learning_rate": 1.4297945481711527e-06, "loss": 0.0012, "step": 1077 }, { "epoch": 4.0680529300567105, "grad_norm": 0.10421045869588852, "learning_rate": 1.4248369943086997e-06, "loss": 0.0002, "step": 1078 }, { "epoch": 4.071833648393195, "grad_norm": 0.3754621744155884, "learning_rate": 1.4198846224129255e-06, "loss": 0.0015, "step": 1079 }, { "epoch": 4.0756143667296785, "grad_norm": 0.15679197013378143, "learning_rate": 1.4149374563527957e-06, "loss": 0.0008, "step": 1080 }, { "epoch": 4.079395085066163, "grad_norm": 0.5258352756500244, "learning_rate": 1.4099955199721816e-06, "loss": 0.0006, "step": 1081 }, { "epoch": 4.083175803402646, "grad_norm": 0.09583219885826111, "learning_rate": 1.4050588370897493e-06, "loss": 0.0004, "step": 1082 }, { "epoch": 4.086956521739131, "grad_norm": 0.21159407496452332, "learning_rate": 1.4001274314988476e-06, "loss": 0.0012, "step": 1083 }, { "epoch": 4.090737240075614, "grad_norm": 0.06300465762615204, "learning_rate": 1.3952013269673872e-06, "loss": 0.0004, "step": 1084 }, { "epoch": 4.094517958412098, "grad_norm": 0.2787025570869446, "learning_rate": 1.3902805472377307e-06, "loss": 0.0016, "step": 1085 }, { "epoch": 4.098298676748582, "grad_norm": 0.015505858696997166, "learning_rate": 1.3853651160265752e-06, "loss": 0.0001, "step": 1086 }, { "epoch": 4.102079395085066, "grad_norm": 0.12251166254281998, "learning_rate": 1.3804550570248431e-06, "loss": 0.0005, "step": 1087 }, { "epoch": 4.10586011342155, "grad_norm": 0.350531667470932, "learning_rate": 1.375550393897559e-06, "loss": 0.0008, "step": 1088 }, { "epoch": 4.109640831758034, "grad_norm": 0.15398718416690826, "learning_rate": 1.370651150283746e-06, "loss": 0.0006, "step": 1089 }, { "epoch": 4.113421550094518, "grad_norm": 0.39199334383010864, "learning_rate": 1.3657573497963042e-06, "loss": 0.0004, "step": 1090 }, { "epoch": 4.117202268431002, "grad_norm": 0.5439186692237854, "learning_rate": 1.360869016021899e-06, "loss": 0.0004, "step": 1091 }, { "epoch": 4.120982986767486, "grad_norm": 0.012271784245967865, "learning_rate": 1.3559861725208501e-06, "loss": 0.0001, "step": 1092 }, { "epoch": 4.12476370510397, "grad_norm": 0.26115530729293823, "learning_rate": 1.3511088428270141e-06, "loss": 0.0012, "step": 1093 }, { "epoch": 4.128544423440454, "grad_norm": 0.28641173243522644, "learning_rate": 1.3462370504476729e-06, "loss": 0.0012, "step": 1094 }, { "epoch": 4.1323251417769375, "grad_norm": 0.3195199966430664, "learning_rate": 1.3413708188634192e-06, "loss": 0.0007, "step": 1095 }, { "epoch": 4.136105860113422, "grad_norm": 0.10783225297927856, "learning_rate": 1.3365101715280473e-06, "loss": 0.0004, "step": 1096 }, { "epoch": 4.1398865784499055, "grad_norm": 0.09313753992319107, "learning_rate": 1.3316551318684342e-06, "loss": 0.0004, "step": 1097 }, { "epoch": 4.143667296786389, "grad_norm": 0.01211604755371809, "learning_rate": 1.3268057232844305e-06, "loss": 0.0001, "step": 1098 }, { "epoch": 4.147448015122873, "grad_norm": 0.23705434799194336, "learning_rate": 1.321961969148746e-06, "loss": 0.0008, "step": 1099 }, { "epoch": 4.151228733459357, "grad_norm": 0.06034567952156067, "learning_rate": 1.3171238928068403e-06, "loss": 0.0002, "step": 1100 }, { "epoch": 4.155009451795841, "grad_norm": 0.01622333563864231, "learning_rate": 1.3122915175768044e-06, "loss": 0.0001, "step": 1101 }, { "epoch": 4.158790170132325, "grad_norm": 0.34421801567077637, "learning_rate": 1.307464866749253e-06, "loss": 0.0003, "step": 1102 }, { "epoch": 4.162570888468809, "grad_norm": 0.014288971200585365, "learning_rate": 1.3026439635872132e-06, "loss": 0.0001, "step": 1103 }, { "epoch": 4.166351606805293, "grad_norm": 0.03465963155031204, "learning_rate": 1.2978288313260045e-06, "loss": 0.0002, "step": 1104 }, { "epoch": 4.170132325141777, "grad_norm": 0.2304038256406784, "learning_rate": 1.2930194931731382e-06, "loss": 0.001, "step": 1105 }, { "epoch": 4.173913043478261, "grad_norm": 0.026781270280480385, "learning_rate": 1.2882159723081967e-06, "loss": 0.0001, "step": 1106 }, { "epoch": 4.177693761814745, "grad_norm": 0.3735407292842865, "learning_rate": 1.2834182918827248e-06, "loss": 0.0026, "step": 1107 }, { "epoch": 4.181474480151229, "grad_norm": 0.10482420772314072, "learning_rate": 1.2786264750201182e-06, "loss": 0.0004, "step": 1108 }, { "epoch": 4.185255198487713, "grad_norm": 0.1388075202703476, "learning_rate": 1.2738405448155139e-06, "loss": 0.0004, "step": 1109 }, { "epoch": 4.189035916824197, "grad_norm": 0.2576792538166046, "learning_rate": 1.269060524335675e-06, "loss": 0.0009, "step": 1110 }, { "epoch": 4.19281663516068, "grad_norm": 0.13617852330207825, "learning_rate": 1.2642864366188806e-06, "loss": 0.0006, "step": 1111 }, { "epoch": 4.196597353497165, "grad_norm": 0.14368586242198944, "learning_rate": 1.2595183046748187e-06, "loss": 0.0005, "step": 1112 }, { "epoch": 4.200378071833648, "grad_norm": 0.05967310443520546, "learning_rate": 1.2547561514844706e-06, "loss": 0.0003, "step": 1113 }, { "epoch": 4.2041587901701325, "grad_norm": 0.2563086152076721, "learning_rate": 1.2500000000000007e-06, "loss": 0.0008, "step": 1114 }, { "epoch": 4.207939508506616, "grad_norm": 2.2195889949798584, "learning_rate": 1.245249873144647e-06, "loss": 0.0016, "step": 1115 }, { "epoch": 4.2117202268431, "grad_norm": 0.4171058237552643, "learning_rate": 1.2405057938126152e-06, "loss": 0.001, "step": 1116 }, { "epoch": 4.215500945179584, "grad_norm": 0.28222987055778503, "learning_rate": 1.235767784868956e-06, "loss": 0.0009, "step": 1117 }, { "epoch": 4.219281663516068, "grad_norm": 0.11864659935235977, "learning_rate": 1.231035869149468e-06, "loss": 0.0004, "step": 1118 }, { "epoch": 4.223062381852552, "grad_norm": 0.18131184577941895, "learning_rate": 1.2263100694605837e-06, "loss": 0.0006, "step": 1119 }, { "epoch": 4.226843100189036, "grad_norm": 0.03620271012187004, "learning_rate": 1.221590408579251e-06, "loss": 0.0002, "step": 1120 }, { "epoch": 4.23062381852552, "grad_norm": 0.43053460121154785, "learning_rate": 1.2168769092528382e-06, "loss": 0.0006, "step": 1121 }, { "epoch": 4.234404536862003, "grad_norm": 0.11408156156539917, "learning_rate": 1.2121695941990126e-06, "loss": 0.0003, "step": 1122 }, { "epoch": 4.238185255198488, "grad_norm": 0.025387326255440712, "learning_rate": 1.207468486105636e-06, "loss": 0.0001, "step": 1123 }, { "epoch": 4.241965973534971, "grad_norm": 0.3768988251686096, "learning_rate": 1.2027736076306533e-06, "loss": 0.0012, "step": 1124 }, { "epoch": 4.245746691871456, "grad_norm": 0.10813384503126144, "learning_rate": 1.1980849814019881e-06, "loss": 0.0003, "step": 1125 }, { "epoch": 4.249527410207939, "grad_norm": 0.03732176125049591, "learning_rate": 1.1934026300174265e-06, "loss": 0.0001, "step": 1126 }, { "epoch": 4.253308128544424, "grad_norm": 0.023056620731949806, "learning_rate": 1.1887265760445122e-06, "loss": 0.0001, "step": 1127 }, { "epoch": 4.257088846880907, "grad_norm": 0.018509745597839355, "learning_rate": 1.1840568420204393e-06, "loss": 0.0001, "step": 1128 }, { "epoch": 4.260869565217392, "grad_norm": 0.9386374354362488, "learning_rate": 1.17939345045194e-06, "loss": 0.0048, "step": 1129 }, { "epoch": 4.264650283553875, "grad_norm": 0.09051921963691711, "learning_rate": 1.1747364238151774e-06, "loss": 0.0005, "step": 1130 }, { "epoch": 4.2684310018903595, "grad_norm": 0.1713877171278, "learning_rate": 1.1700857845556374e-06, "loss": 0.0006, "step": 1131 }, { "epoch": 4.272211720226843, "grad_norm": 0.06043676659464836, "learning_rate": 1.1654415550880245e-06, "loss": 0.0002, "step": 1132 }, { "epoch": 4.2759924385633274, "grad_norm": 0.01333721075206995, "learning_rate": 1.1608037577961423e-06, "loss": 0.0001, "step": 1133 }, { "epoch": 4.279773156899811, "grad_norm": 0.4935985505580902, "learning_rate": 1.156172415032801e-06, "loss": 0.0018, "step": 1134 }, { "epoch": 4.2835538752362945, "grad_norm": 0.14692044258117676, "learning_rate": 1.1515475491196975e-06, "loss": 0.0005, "step": 1135 }, { "epoch": 4.287334593572779, "grad_norm": 0.3206477165222168, "learning_rate": 1.1469291823473127e-06, "loss": 0.0012, "step": 1136 }, { "epoch": 4.291115311909262, "grad_norm": 0.041015662252902985, "learning_rate": 1.1423173369748056e-06, "loss": 0.0001, "step": 1137 }, { "epoch": 4.294896030245747, "grad_norm": 0.009320144541561604, "learning_rate": 1.1377120352299015e-06, "loss": 0.0001, "step": 1138 }, { "epoch": 4.29867674858223, "grad_norm": 0.061102405190467834, "learning_rate": 1.1331132993087887e-06, "loss": 0.0001, "step": 1139 }, { "epoch": 4.302457466918715, "grad_norm": 0.20246689021587372, "learning_rate": 1.1285211513760088e-06, "loss": 0.001, "step": 1140 }, { "epoch": 4.306238185255198, "grad_norm": 0.0354740135371685, "learning_rate": 1.1239356135643544e-06, "loss": 0.0001, "step": 1141 }, { "epoch": 4.310018903591683, "grad_norm": 0.08336465805768967, "learning_rate": 1.119356707974756e-06, "loss": 0.0003, "step": 1142 }, { "epoch": 4.313799621928166, "grad_norm": 0.034681014716625214, "learning_rate": 1.11478445667618e-06, "loss": 0.0002, "step": 1143 }, { "epoch": 4.317580340264651, "grad_norm": 0.10348597913980484, "learning_rate": 1.1102188817055202e-06, "loss": 0.0004, "step": 1144 }, { "epoch": 4.321361058601134, "grad_norm": 0.11162035167217255, "learning_rate": 1.105660005067495e-06, "loss": 0.0004, "step": 1145 }, { "epoch": 4.325141776937619, "grad_norm": 0.2652709484100342, "learning_rate": 1.1011078487345362e-06, "loss": 0.0017, "step": 1146 }, { "epoch": 4.328922495274102, "grad_norm": 0.1814330667257309, "learning_rate": 1.096562434646686e-06, "loss": 0.0009, "step": 1147 }, { "epoch": 4.332703213610586, "grad_norm": 0.23978620767593384, "learning_rate": 1.0920237847114945e-06, "loss": 0.0005, "step": 1148 }, { "epoch": 4.33648393194707, "grad_norm": 0.1966646909713745, "learning_rate": 1.0874919208039036e-06, "loss": 0.0014, "step": 1149 }, { "epoch": 4.340264650283554, "grad_norm": 0.06779544800519943, "learning_rate": 1.0829668647661558e-06, "loss": 0.0002, "step": 1150 }, { "epoch": 4.344045368620038, "grad_norm": 0.08684524893760681, "learning_rate": 1.0784486384076778e-06, "loss": 0.0002, "step": 1151 }, { "epoch": 4.3478260869565215, "grad_norm": 0.25578299164772034, "learning_rate": 1.0739372635049796e-06, "loss": 0.0006, "step": 1152 }, { "epoch": 4.351606805293006, "grad_norm": 0.2866862714290619, "learning_rate": 1.0694327618015492e-06, "loss": 0.0007, "step": 1153 }, { "epoch": 4.355387523629489, "grad_norm": 0.07343002408742905, "learning_rate": 1.06493515500775e-06, "loss": 0.0003, "step": 1154 }, { "epoch": 4.359168241965974, "grad_norm": 0.4368140995502472, "learning_rate": 1.0604444648007114e-06, "loss": 0.0011, "step": 1155 }, { "epoch": 4.362948960302457, "grad_norm": 0.07969985157251358, "learning_rate": 1.0559607128242267e-06, "loss": 0.0003, "step": 1156 }, { "epoch": 4.366729678638942, "grad_norm": 0.028882203623652458, "learning_rate": 1.0514839206886519e-06, "loss": 0.0001, "step": 1157 }, { "epoch": 4.370510396975425, "grad_norm": 0.11747908592224121, "learning_rate": 1.0470141099707958e-06, "loss": 0.0003, "step": 1158 }, { "epoch": 4.374291115311909, "grad_norm": 0.20348979532718658, "learning_rate": 1.0425513022138203e-06, "loss": 0.0005, "step": 1159 }, { "epoch": 4.378071833648393, "grad_norm": 0.0517929345369339, "learning_rate": 1.0380955189271336e-06, "loss": 0.0001, "step": 1160 }, { "epoch": 4.381852551984877, "grad_norm": 0.15096187591552734, "learning_rate": 1.0336467815862905e-06, "loss": 0.0014, "step": 1161 }, { "epoch": 4.385633270321361, "grad_norm": 0.27733010053634644, "learning_rate": 1.0292051116328844e-06, "loss": 0.0007, "step": 1162 }, { "epoch": 4.389413988657845, "grad_norm": 0.13852417469024658, "learning_rate": 1.0247705304744458e-06, "loss": 0.0005, "step": 1163 }, { "epoch": 4.393194706994329, "grad_norm": 0.0745406225323677, "learning_rate": 1.0203430594843426e-06, "loss": 0.0002, "step": 1164 }, { "epoch": 4.396975425330813, "grad_norm": 0.1584874838590622, "learning_rate": 1.0159227200016677e-06, "loss": 0.0006, "step": 1165 }, { "epoch": 4.400756143667297, "grad_norm": 0.07365313917398453, "learning_rate": 1.0115095333311487e-06, "loss": 0.0003, "step": 1166 }, { "epoch": 4.404536862003781, "grad_norm": 0.34414687752723694, "learning_rate": 1.0071035207430352e-06, "loss": 0.0021, "step": 1167 }, { "epoch": 4.408317580340265, "grad_norm": 0.11103616654872894, "learning_rate": 1.002704703473e-06, "loss": 0.0005, "step": 1168 }, { "epoch": 4.4120982986767485, "grad_norm": 0.3156139850616455, "learning_rate": 9.983131027220373e-07, "loss": 0.0012, "step": 1169 }, { "epoch": 4.415879017013232, "grad_norm": 0.6131001710891724, "learning_rate": 9.939287396563604e-07, "loss": 0.004, "step": 1170 }, { "epoch": 4.4196597353497165, "grad_norm": 0.035383161157369614, "learning_rate": 9.895516354072981e-07, "loss": 0.0001, "step": 1171 }, { "epoch": 4.4234404536862, "grad_norm": 0.07956578582525253, "learning_rate": 9.85181811071193e-07, "loss": 0.0003, "step": 1172 }, { "epoch": 4.427221172022684, "grad_norm": 0.04703371599316597, "learning_rate": 9.808192877093039e-07, "loss": 0.0002, "step": 1173 }, { "epoch": 4.431001890359168, "grad_norm": 0.030056219547986984, "learning_rate": 9.764640863476982e-07, "loss": 0.0001, "step": 1174 }, { "epoch": 4.434782608695652, "grad_norm": 0.4825742244720459, "learning_rate": 9.72116227977154e-07, "loss": 0.0003, "step": 1175 }, { "epoch": 4.438563327032136, "grad_norm": 0.10369419306516647, "learning_rate": 9.677757335530577e-07, "loss": 0.0003, "step": 1176 }, { "epoch": 4.44234404536862, "grad_norm": 0.25386276841163635, "learning_rate": 9.634426239953074e-07, "loss": 0.0008, "step": 1177 }, { "epoch": 4.446124763705104, "grad_norm": 0.14689001441001892, "learning_rate": 9.59116920188202e-07, "loss": 0.0003, "step": 1178 }, { "epoch": 4.449905482041588, "grad_norm": 0.42591592669487, "learning_rate": 9.547986429803518e-07, "loss": 0.0021, "step": 1179 }, { "epoch": 4.453686200378072, "grad_norm": 0.24819211661815643, "learning_rate": 9.504878131845738e-07, "loss": 0.0003, "step": 1180 }, { "epoch": 4.457466918714556, "grad_norm": 0.020263489335775375, "learning_rate": 9.461844515777849e-07, "loss": 0.0001, "step": 1181 }, { "epoch": 4.46124763705104, "grad_norm": 0.18784894049167633, "learning_rate": 9.418885789009139e-07, "loss": 0.0007, "step": 1182 }, { "epoch": 4.465028355387524, "grad_norm": 0.18127810955047607, "learning_rate": 9.376002158587915e-07, "loss": 0.0004, "step": 1183 }, { "epoch": 4.468809073724008, "grad_norm": 0.30408334732055664, "learning_rate": 9.333193831200557e-07, "loss": 0.0009, "step": 1184 }, { "epoch": 4.472589792060491, "grad_norm": 0.23862513899803162, "learning_rate": 9.290461013170485e-07, "loss": 0.0008, "step": 1185 }, { "epoch": 4.4763705103969755, "grad_norm": 0.017677878960967064, "learning_rate": 9.247803910457226e-07, "loss": 0.0001, "step": 1186 }, { "epoch": 4.480151228733459, "grad_norm": 0.23461668193340302, "learning_rate": 9.205222728655344e-07, "loss": 0.0006, "step": 1187 }, { "epoch": 4.4839319470699435, "grad_norm": 0.08011229336261749, "learning_rate": 9.162717672993499e-07, "loss": 0.0003, "step": 1188 }, { "epoch": 4.487712665406427, "grad_norm": 0.10889612138271332, "learning_rate": 9.120288948333464e-07, "loss": 0.0003, "step": 1189 }, { "epoch": 4.491493383742911, "grad_norm": 0.04668443277478218, "learning_rate": 9.07793675916909e-07, "loss": 0.0002, "step": 1190 }, { "epoch": 4.495274102079395, "grad_norm": 0.02013458125293255, "learning_rate": 9.035661309625366e-07, "loss": 0.0001, "step": 1191 }, { "epoch": 4.499054820415879, "grad_norm": 0.05026013031601906, "learning_rate": 8.993462803457403e-07, "loss": 0.0002, "step": 1192 }, { "epoch": 4.502835538752363, "grad_norm": 0.2004999816417694, "learning_rate": 8.951341444049513e-07, "loss": 0.0008, "step": 1193 }, { "epoch": 4.506616257088847, "grad_norm": 0.2229345291852951, "learning_rate": 8.90929743441411e-07, "loss": 0.0008, "step": 1194 }, { "epoch": 4.510396975425331, "grad_norm": 0.8019787073135376, "learning_rate": 8.867330977190877e-07, "loss": 0.001, "step": 1195 }, { "epoch": 4.514177693761814, "grad_norm": 0.0796135812997818, "learning_rate": 8.825442274645671e-07, "loss": 0.0003, "step": 1196 }, { "epoch": 4.517958412098299, "grad_norm": 0.12469209730625153, "learning_rate": 8.783631528669607e-07, "loss": 0.0003, "step": 1197 }, { "epoch": 4.521739130434782, "grad_norm": 0.6032193899154663, "learning_rate": 8.741898940778088e-07, "loss": 0.004, "step": 1198 }, { "epoch": 4.525519848771267, "grad_norm": 0.09719254821538925, "learning_rate": 8.700244712109798e-07, "loss": 0.0004, "step": 1199 }, { "epoch": 4.52930056710775, "grad_norm": 0.21795785427093506, "learning_rate": 8.658669043425763e-07, "loss": 0.0007, "step": 1200 }, { "epoch": 4.533081285444235, "grad_norm": 0.29704999923706055, "learning_rate": 8.617172135108356e-07, "loss": 0.0004, "step": 1201 }, { "epoch": 4.536862003780718, "grad_norm": 0.0031759943813085556, "learning_rate": 8.575754187160385e-07, "loss": 0.0, "step": 1202 }, { "epoch": 4.5406427221172025, "grad_norm": 0.4515712261199951, "learning_rate": 8.53441539920406e-07, "loss": 0.0003, "step": 1203 }, { "epoch": 4.544423440453686, "grad_norm": 0.15835940837860107, "learning_rate": 8.493155970480074e-07, "loss": 0.0004, "step": 1204 }, { "epoch": 4.5482041587901705, "grad_norm": 0.14874432981014252, "learning_rate": 8.45197609984662e-07, "loss": 0.0004, "step": 1205 }, { "epoch": 4.551984877126654, "grad_norm": 0.07101760059595108, "learning_rate": 8.410875985778477e-07, "loss": 0.0002, "step": 1206 }, { "epoch": 4.5557655954631375, "grad_norm": 0.08789049834012985, "learning_rate": 8.369855826365988e-07, "loss": 0.0002, "step": 1207 }, { "epoch": 4.559546313799622, "grad_norm": 0.022374188527464867, "learning_rate": 8.328915819314149e-07, "loss": 0.0001, "step": 1208 }, { "epoch": 4.563327032136106, "grad_norm": 0.6488557457923889, "learning_rate": 8.288056161941663e-07, "loss": 0.0023, "step": 1209 }, { "epoch": 4.56710775047259, "grad_norm": 0.004378811921924353, "learning_rate": 8.247277051179925e-07, "loss": 0.0001, "step": 1210 }, { "epoch": 4.570888468809073, "grad_norm": 0.05439801514148712, "learning_rate": 8.206578683572178e-07, "loss": 0.0002, "step": 1211 }, { "epoch": 4.574669187145558, "grad_norm": 0.08567692339420319, "learning_rate": 8.165961255272467e-07, "loss": 0.0004, "step": 1212 }, { "epoch": 4.578449905482041, "grad_norm": 0.19837629795074463, "learning_rate": 8.125424962044742e-07, "loss": 0.0005, "step": 1213 }, { "epoch": 4.582230623818526, "grad_norm": 0.1341010332107544, "learning_rate": 8.084969999261907e-07, "loss": 0.0007, "step": 1214 }, { "epoch": 4.586011342155009, "grad_norm": 0.10602247714996338, "learning_rate": 8.044596561904894e-07, "loss": 0.0003, "step": 1215 }, { "epoch": 4.589792060491494, "grad_norm": 0.0057820482179522514, "learning_rate": 8.004304844561684e-07, "loss": 0.0, "step": 1216 }, { "epoch": 4.593572778827977, "grad_norm": 0.018521621823310852, "learning_rate": 7.964095041426387e-07, "loss": 0.0001, "step": 1217 }, { "epoch": 4.597353497164462, "grad_norm": 0.31507861614227295, "learning_rate": 7.923967346298345e-07, "loss": 0.0008, "step": 1218 }, { "epoch": 4.601134215500945, "grad_norm": 0.01326778158545494, "learning_rate": 7.883921952581131e-07, "loss": 0.0001, "step": 1219 }, { "epoch": 4.60491493383743, "grad_norm": 0.12220294773578644, "learning_rate": 7.843959053281663e-07, "loss": 0.0002, "step": 1220 }, { "epoch": 4.608695652173913, "grad_norm": 0.1671430617570877, "learning_rate": 7.804078841009252e-07, "loss": 0.0003, "step": 1221 }, { "epoch": 4.612476370510397, "grad_norm": 0.72855144739151, "learning_rate": 7.764281507974711e-07, "loss": 0.0009, "step": 1222 }, { "epoch": 4.616257088846881, "grad_norm": 0.47011008858680725, "learning_rate": 7.724567245989342e-07, "loss": 0.0042, "step": 1223 }, { "epoch": 4.6200378071833645, "grad_norm": 0.24560636281967163, "learning_rate": 7.684936246464125e-07, "loss": 0.002, "step": 1224 }, { "epoch": 4.623818525519849, "grad_norm": 0.04877178370952606, "learning_rate": 7.645388700408732e-07, "loss": 0.0002, "step": 1225 }, { "epoch": 4.6275992438563325, "grad_norm": 0.14510324597358704, "learning_rate": 7.60592479843057e-07, "loss": 0.0002, "step": 1226 }, { "epoch": 4.631379962192817, "grad_norm": 0.32103654742240906, "learning_rate": 7.566544730733961e-07, "loss": 0.0008, "step": 1227 }, { "epoch": 4.6351606805293, "grad_norm": 0.22647596895694733, "learning_rate": 7.527248687119137e-07, "loss": 0.001, "step": 1228 }, { "epoch": 4.638941398865785, "grad_norm": 0.24789294600486755, "learning_rate": 7.488036856981373e-07, "loss": 0.0009, "step": 1229 }, { "epoch": 4.642722117202268, "grad_norm": 0.06908345222473145, "learning_rate": 7.448909429310047e-07, "loss": 0.0002, "step": 1230 }, { "epoch": 4.646502835538753, "grad_norm": 0.32550469040870667, "learning_rate": 7.409866592687768e-07, "loss": 0.0019, "step": 1231 }, { "epoch": 4.650283553875236, "grad_norm": 0.33530768752098083, "learning_rate": 7.370908535289409e-07, "loss": 0.0021, "step": 1232 }, { "epoch": 4.65406427221172, "grad_norm": 0.0069428663700819016, "learning_rate": 7.332035444881247e-07, "loss": 0.0, "step": 1233 }, { "epoch": 4.657844990548204, "grad_norm": 0.9020087122917175, "learning_rate": 7.293247508820059e-07, "loss": 0.0022, "step": 1234 }, { "epoch": 4.661625708884688, "grad_norm": 0.3032546937465668, "learning_rate": 7.254544914052178e-07, "loss": 0.0005, "step": 1235 }, { "epoch": 4.665406427221172, "grad_norm": 0.09237213432788849, "learning_rate": 7.215927847112628e-07, "loss": 0.0003, "step": 1236 }, { "epoch": 4.669187145557656, "grad_norm": 0.018140338361263275, "learning_rate": 7.177396494124206e-07, "loss": 0.0001, "step": 1237 }, { "epoch": 4.67296786389414, "grad_norm": 0.05702436715364456, "learning_rate": 7.138951040796627e-07, "loss": 0.0002, "step": 1238 }, { "epoch": 4.676748582230624, "grad_norm": 0.03389357402920723, "learning_rate": 7.100591672425533e-07, "loss": 0.0001, "step": 1239 }, { "epoch": 4.680529300567108, "grad_norm": 0.5182787775993347, "learning_rate": 7.062318573891716e-07, "loss": 0.0023, "step": 1240 }, { "epoch": 4.6843100189035916, "grad_norm": 0.17741823196411133, "learning_rate": 7.024131929660172e-07, "loss": 0.0005, "step": 1241 }, { "epoch": 4.688090737240076, "grad_norm": 0.5899286270141602, "learning_rate": 6.986031923779166e-07, "loss": 0.0018, "step": 1242 }, { "epoch": 4.6918714555765595, "grad_norm": 0.19821827113628387, "learning_rate": 6.948018739879439e-07, "loss": 0.0006, "step": 1243 }, { "epoch": 4.695652173913043, "grad_norm": 0.022515807300806046, "learning_rate": 6.910092561173256e-07, "loss": 0.0001, "step": 1244 }, { "epoch": 4.699432892249527, "grad_norm": 0.30378711223602295, "learning_rate": 6.872253570453546e-07, "loss": 0.0009, "step": 1245 }, { "epoch": 4.703213610586012, "grad_norm": 0.03955739364027977, "learning_rate": 6.834501950093006e-07, "loss": 0.0002, "step": 1246 }, { "epoch": 4.706994328922495, "grad_norm": 0.03287004306912422, "learning_rate": 6.796837882043261e-07, "loss": 0.0001, "step": 1247 }, { "epoch": 4.710775047258979, "grad_norm": 0.11705687642097473, "learning_rate": 6.75926154783394e-07, "loss": 0.0004, "step": 1248 }, { "epoch": 4.714555765595463, "grad_norm": 0.04283098503947258, "learning_rate": 6.721773128571812e-07, "loss": 0.0001, "step": 1249 }, { "epoch": 4.718336483931947, "grad_norm": 0.08523938804864883, "learning_rate": 6.684372804939954e-07, "loss": 0.0002, "step": 1250 }, { "epoch": 4.722117202268431, "grad_norm": 0.05707685276865959, "learning_rate": 6.647060757196818e-07, "loss": 0.0002, "step": 1251 }, { "epoch": 4.725897920604915, "grad_norm": 0.07423222810029984, "learning_rate": 6.609837165175408e-07, "loss": 0.0002, "step": 1252 }, { "epoch": 4.729678638941399, "grad_norm": 0.5939041376113892, "learning_rate": 6.57270220828238e-07, "loss": 0.0005, "step": 1253 }, { "epoch": 4.733459357277883, "grad_norm": 0.28552377223968506, "learning_rate": 6.535656065497237e-07, "loss": 0.0006, "step": 1254 }, { "epoch": 4.737240075614367, "grad_norm": 0.38389602303504944, "learning_rate": 6.49869891537136e-07, "loss": 0.0015, "step": 1255 }, { "epoch": 4.741020793950851, "grad_norm": 0.01133513543754816, "learning_rate": 6.461830936027272e-07, "loss": 0.0, "step": 1256 }, { "epoch": 4.744801512287335, "grad_norm": 0.021422209218144417, "learning_rate": 6.425052305157683e-07, "loss": 0.0001, "step": 1257 }, { "epoch": 4.748582230623819, "grad_norm": 0.6229876279830933, "learning_rate": 6.38836320002468e-07, "loss": 0.0018, "step": 1258 }, { "epoch": 4.752362948960302, "grad_norm": 0.34762510657310486, "learning_rate": 6.351763797458879e-07, "loss": 0.0008, "step": 1259 }, { "epoch": 4.7561436672967865, "grad_norm": 0.04826958850026131, "learning_rate": 6.315254273858525e-07, "loss": 0.0002, "step": 1260 }, { "epoch": 4.75992438563327, "grad_norm": 0.1443760246038437, "learning_rate": 6.278834805188699e-07, "loss": 0.0005, "step": 1261 }, { "epoch": 4.763705103969754, "grad_norm": 0.08828070759773254, "learning_rate": 6.242505566980423e-07, "loss": 0.0002, "step": 1262 }, { "epoch": 4.767485822306238, "grad_norm": 0.13958247005939484, "learning_rate": 6.20626673432986e-07, "loss": 0.0002, "step": 1263 }, { "epoch": 4.771266540642722, "grad_norm": 0.10381146520376205, "learning_rate": 6.170118481897422e-07, "loss": 0.0003, "step": 1264 }, { "epoch": 4.775047258979206, "grad_norm": 0.3253827393054962, "learning_rate": 6.134060983906964e-07, "loss": 0.0012, "step": 1265 }, { "epoch": 4.77882797731569, "grad_norm": 0.0819317176938057, "learning_rate": 6.09809441414492e-07, "loss": 0.0004, "step": 1266 }, { "epoch": 4.782608695652174, "grad_norm": 0.2494785040616989, "learning_rate": 6.062218945959497e-07, "loss": 0.001, "step": 1267 }, { "epoch": 4.786389413988658, "grad_norm": 0.02512381225824356, "learning_rate": 6.026434752259802e-07, "loss": 0.0001, "step": 1268 }, { "epoch": 4.790170132325142, "grad_norm": 0.14734607934951782, "learning_rate": 5.990742005515024e-07, "loss": 0.0005, "step": 1269 }, { "epoch": 4.793950850661625, "grad_norm": 0.22312569618225098, "learning_rate": 5.955140877753635e-07, "loss": 0.0007, "step": 1270 }, { "epoch": 4.79773156899811, "grad_norm": 0.2650485038757324, "learning_rate": 5.919631540562473e-07, "loss": 0.0008, "step": 1271 }, { "epoch": 4.801512287334593, "grad_norm": 0.2240552008152008, "learning_rate": 5.884214165086036e-07, "loss": 0.0009, "step": 1272 }, { "epoch": 4.805293005671078, "grad_norm": 0.009949145838618279, "learning_rate": 5.848888922025553e-07, "loss": 0.0001, "step": 1273 }, { "epoch": 4.809073724007561, "grad_norm": 0.034081898629665375, "learning_rate": 5.813655981638217e-07, "loss": 0.0001, "step": 1274 }, { "epoch": 4.812854442344046, "grad_norm": 0.25449320673942566, "learning_rate": 5.778515513736338e-07, "loss": 0.0011, "step": 1275 }, { "epoch": 4.816635160680529, "grad_norm": 0.09371744841337204, "learning_rate": 5.743467687686563e-07, "loss": 0.0003, "step": 1276 }, { "epoch": 4.8204158790170135, "grad_norm": 0.07984409481287003, "learning_rate": 5.708512672409002e-07, "loss": 0.0004, "step": 1277 }, { "epoch": 4.824196597353497, "grad_norm": 0.4380701184272766, "learning_rate": 5.673650636376457e-07, "loss": 0.0025, "step": 1278 }, { "epoch": 4.8279773156899815, "grad_norm": 0.11228739470243454, "learning_rate": 5.638881747613603e-07, "loss": 0.0005, "step": 1279 }, { "epoch": 4.831758034026465, "grad_norm": 0.32348689436912537, "learning_rate": 5.604206173696169e-07, "loss": 0.0006, "step": 1280 }, { "epoch": 4.8355387523629485, "grad_norm": 0.06818557530641556, "learning_rate": 5.56962408175013e-07, "loss": 0.0002, "step": 1281 }, { "epoch": 4.839319470699433, "grad_norm": 0.014409010298550129, "learning_rate": 5.5351356384509e-07, "loss": 0.0001, "step": 1282 }, { "epoch": 4.843100189035916, "grad_norm": 1.157328724861145, "learning_rate": 5.500741010022564e-07, "loss": 0.0007, "step": 1283 }, { "epoch": 4.846880907372401, "grad_norm": 0.16351604461669922, "learning_rate": 5.466440362237e-07, "loss": 0.0004, "step": 1284 }, { "epoch": 4.850661625708884, "grad_norm": 0.06235453113913536, "learning_rate": 5.432233860413172e-07, "loss": 0.0003, "step": 1285 }, { "epoch": 4.854442344045369, "grad_norm": 0.18696486949920654, "learning_rate": 5.398121669416262e-07, "loss": 0.0003, "step": 1286 }, { "epoch": 4.858223062381852, "grad_norm": 0.22089411318302155, "learning_rate": 5.364103953656905e-07, "loss": 0.0007, "step": 1287 }, { "epoch": 4.862003780718337, "grad_norm": 0.33909812569618225, "learning_rate": 5.33018087709041e-07, "loss": 0.0009, "step": 1288 }, { "epoch": 4.86578449905482, "grad_norm": 0.02142472006380558, "learning_rate": 5.296352603215935e-07, "loss": 0.0001, "step": 1289 }, { "epoch": 4.869565217391305, "grad_norm": 0.1449928879737854, "learning_rate": 5.262619295075721e-07, "loss": 0.0004, "step": 1290 }, { "epoch": 4.873345935727788, "grad_norm": 0.40573325753211975, "learning_rate": 5.2289811152543e-07, "loss": 0.0013, "step": 1291 }, { "epoch": 4.877126654064273, "grad_norm": 0.022667789831757545, "learning_rate": 5.195438225877738e-07, "loss": 0.0001, "step": 1292 }, { "epoch": 4.880907372400756, "grad_norm": 0.056100767105817795, "learning_rate": 5.161990788612781e-07, "loss": 0.0002, "step": 1293 }, { "epoch": 4.8846880907372405, "grad_norm": 0.11975480616092682, "learning_rate": 5.128638964666166e-07, "loss": 0.0004, "step": 1294 }, { "epoch": 4.888468809073724, "grad_norm": 0.03958946466445923, "learning_rate": 5.095382914783803e-07, "loss": 0.0001, "step": 1295 }, { "epoch": 4.892249527410208, "grad_norm": 0.7900720834732056, "learning_rate": 5.062222799249958e-07, "loss": 0.0006, "step": 1296 }, { "epoch": 4.896030245746692, "grad_norm": 0.051875967532396317, "learning_rate": 5.02915877788657e-07, "loss": 0.0002, "step": 1297 }, { "epoch": 4.8998109640831755, "grad_norm": 0.5205870270729065, "learning_rate": 4.996191010052404e-07, "loss": 0.0024, "step": 1298 }, { "epoch": 4.90359168241966, "grad_norm": 0.41348710656166077, "learning_rate": 4.963319654642321e-07, "loss": 0.0021, "step": 1299 }, { "epoch": 4.9073724007561434, "grad_norm": 0.01838637888431549, "learning_rate": 4.93054487008649e-07, "loss": 0.0001, "step": 1300 }, { "epoch": 4.911153119092628, "grad_norm": 0.00513014942407608, "learning_rate": 4.897866814349667e-07, "loss": 0.0, "step": 1301 }, { "epoch": 4.914933837429111, "grad_norm": 0.13928453624248505, "learning_rate": 4.865285644930373e-07, "loss": 0.0005, "step": 1302 }, { "epoch": 4.918714555765596, "grad_norm": 0.03825385123491287, "learning_rate": 4.832801518860175e-07, "loss": 0.0001, "step": 1303 }, { "epoch": 4.922495274102079, "grad_norm": 0.09964524954557419, "learning_rate": 4.800414592702934e-07, "loss": 0.0002, "step": 1304 }, { "epoch": 4.926275992438564, "grad_norm": 0.08674181252717972, "learning_rate": 4.7681250225540184e-07, "loss": 0.0003, "step": 1305 }, { "epoch": 4.930056710775047, "grad_norm": 0.20237761735916138, "learning_rate": 4.7359329640395746e-07, "loss": 0.0002, "step": 1306 }, { "epoch": 4.933837429111531, "grad_norm": 0.04080205783247948, "learning_rate": 4.7038385723157666e-07, "loss": 0.0001, "step": 1307 }, { "epoch": 4.937618147448015, "grad_norm": 0.10876837372779846, "learning_rate": 4.671842002068061e-07, "loss": 0.0003, "step": 1308 }, { "epoch": 4.941398865784499, "grad_norm": 0.4733971655368805, "learning_rate": 4.639943407510408e-07, "loss": 0.0013, "step": 1309 }, { "epoch": 4.945179584120983, "grad_norm": 0.5352734923362732, "learning_rate": 4.608142942384586e-07, "loss": 0.0016, "step": 1310 }, { "epoch": 4.948960302457467, "grad_norm": 0.09120377153158188, "learning_rate": 4.576440759959394e-07, "loss": 0.0001, "step": 1311 }, { "epoch": 4.952741020793951, "grad_norm": 0.13324621319770813, "learning_rate": 4.54483701302994e-07, "loss": 0.0004, "step": 1312 }, { "epoch": 4.956521739130435, "grad_norm": 0.17279818654060364, "learning_rate": 4.513331853916922e-07, "loss": 0.0005, "step": 1313 }, { "epoch": 4.960302457466919, "grad_norm": 0.29556113481521606, "learning_rate": 4.4819254344658505e-07, "loss": 0.0013, "step": 1314 }, { "epoch": 4.9640831758034025, "grad_norm": 0.48007726669311523, "learning_rate": 4.450617906046348e-07, "loss": 0.0006, "step": 1315 }, { "epoch": 4.967863894139887, "grad_norm": 0.05618850886821747, "learning_rate": 4.419409419551407e-07, "loss": 0.0002, "step": 1316 }, { "epoch": 4.9716446124763705, "grad_norm": 0.004624408669769764, "learning_rate": 4.388300125396686e-07, "loss": 0.0, "step": 1317 }, { "epoch": 4.975425330812854, "grad_norm": 0.23210549354553223, "learning_rate": 4.3572901735197457e-07, "loss": 0.0017, "step": 1318 }, { "epoch": 4.979206049149338, "grad_norm": 0.061289653182029724, "learning_rate": 4.32637971337935e-07, "loss": 0.0002, "step": 1319 }, { "epoch": 4.982986767485822, "grad_norm": 0.02157505601644516, "learning_rate": 4.2955688939547516e-07, "loss": 0.0001, "step": 1320 }, { "epoch": 4.986767485822306, "grad_norm": 0.2549765110015869, "learning_rate": 4.264857863744956e-07, "loss": 0.0007, "step": 1321 }, { "epoch": 4.99054820415879, "grad_norm": 0.02391950413584709, "learning_rate": 4.2342467707680163e-07, "loss": 0.0001, "step": 1322 }, { "epoch": 4.994328922495274, "grad_norm": 0.3912001848220825, "learning_rate": 4.2037357625603127e-07, "loss": 0.0017, "step": 1323 }, { "epoch": 4.998109640831758, "grad_norm": 0.6139587759971619, "learning_rate": 4.1733249861758697e-07, "loss": 0.0012, "step": 1324 }, { "epoch": 5.0, "grad_norm": 0.021029487252235413, "learning_rate": 4.143014588185582e-07, "loss": 0.0, "step": 1325 }, { "epoch": 5.0037807183364835, "grad_norm": 0.03371495380997658, "learning_rate": 4.1128047146765936e-07, "loss": 0.0001, "step": 1326 }, { "epoch": 5.007561436672968, "grad_norm": 0.08553440123796463, "learning_rate": 4.082695511251522e-07, "loss": 0.0003, "step": 1327 }, { "epoch": 5.0113421550094515, "grad_norm": 0.059370994567871094, "learning_rate": 4.0526871230277904e-07, "loss": 0.0001, "step": 1328 }, { "epoch": 5.015122873345936, "grad_norm": 0.23729994893074036, "learning_rate": 4.022779694636941e-07, "loss": 0.0004, "step": 1329 }, { "epoch": 5.018903591682419, "grad_norm": 0.04285334795713425, "learning_rate": 3.992973370223896e-07, "loss": 0.0002, "step": 1330 }, { "epoch": 5.022684310018904, "grad_norm": 0.016388334333896637, "learning_rate": 3.963268293446296e-07, "loss": 0.0001, "step": 1331 }, { "epoch": 5.026465028355387, "grad_norm": 0.2529665529727936, "learning_rate": 3.933664607473789e-07, "loss": 0.0007, "step": 1332 }, { "epoch": 5.030245746691872, "grad_norm": 0.06798732280731201, "learning_rate": 3.904162454987373e-07, "loss": 0.0002, "step": 1333 }, { "epoch": 5.034026465028355, "grad_norm": 0.03351491689682007, "learning_rate": 3.874761978178665e-07, "loss": 0.0001, "step": 1334 }, { "epoch": 5.03780718336484, "grad_norm": 0.11256009340286255, "learning_rate": 3.845463318749235e-07, "loss": 0.0002, "step": 1335 }, { "epoch": 5.041587901701323, "grad_norm": 0.125478133559227, "learning_rate": 3.8162666179099253e-07, "loss": 0.0003, "step": 1336 }, { "epoch": 5.045368620037808, "grad_norm": 0.025008127093315125, "learning_rate": 3.7871720163801836e-07, "loss": 0.0001, "step": 1337 }, { "epoch": 5.049149338374291, "grad_norm": 0.17265991866588593, "learning_rate": 3.7581796543873476e-07, "loss": 0.0006, "step": 1338 }, { "epoch": 5.052930056710775, "grad_norm": 0.33894410729408264, "learning_rate": 3.729289671665998e-07, "loss": 0.0005, "step": 1339 }, { "epoch": 5.056710775047259, "grad_norm": 0.015326871536672115, "learning_rate": 3.700502207457293e-07, "loss": 0.0001, "step": 1340 }, { "epoch": 5.060491493383743, "grad_norm": 0.1767851859331131, "learning_rate": 3.6718174005082515e-07, "loss": 0.0006, "step": 1341 }, { "epoch": 5.064272211720227, "grad_norm": 0.59239661693573, "learning_rate": 3.6432353890711426e-07, "loss": 0.0015, "step": 1342 }, { "epoch": 5.0680529300567105, "grad_norm": 0.26718172430992126, "learning_rate": 3.614756310902781e-07, "loss": 0.0009, "step": 1343 }, { "epoch": 5.071833648393195, "grad_norm": 0.13824322819709778, "learning_rate": 3.58638030326387e-07, "loss": 0.0007, "step": 1344 }, { "epoch": 5.0756143667296785, "grad_norm": 0.049522772431373596, "learning_rate": 3.5581075029183427e-07, "loss": 0.0002, "step": 1345 }, { "epoch": 5.079395085066163, "grad_norm": 0.31123489141464233, "learning_rate": 3.529938046132717e-07, "loss": 0.0006, "step": 1346 }, { "epoch": 5.083175803402646, "grad_norm": 0.029472079128026962, "learning_rate": 3.501872068675413e-07, "loss": 0.0001, "step": 1347 }, { "epoch": 5.086956521739131, "grad_norm": 0.05294061824679375, "learning_rate": 3.4739097058161116e-07, "loss": 0.0001, "step": 1348 }, { "epoch": 5.090737240075614, "grad_norm": 0.04228921979665756, "learning_rate": 3.446051092325112e-07, "loss": 0.0001, "step": 1349 }, { "epoch": 5.094517958412098, "grad_norm": 0.12975500524044037, "learning_rate": 3.4182963624726616e-07, "loss": 0.0005, "step": 1350 }, { "epoch": 5.098298676748582, "grad_norm": 0.054230622947216034, "learning_rate": 3.390645650028324e-07, "loss": 0.0003, "step": 1351 }, { "epoch": 5.102079395085066, "grad_norm": 0.0159421656280756, "learning_rate": 3.3630990882603215e-07, "loss": 0.0001, "step": 1352 }, { "epoch": 5.10586011342155, "grad_norm": 0.03157084807753563, "learning_rate": 3.3356568099349287e-07, "loss": 0.0001, "step": 1353 }, { "epoch": 5.109640831758034, "grad_norm": 0.010646462440490723, "learning_rate": 3.3083189473157584e-07, "loss": 0.0001, "step": 1354 }, { "epoch": 5.113421550094518, "grad_norm": 0.2389862835407257, "learning_rate": 3.281085632163211e-07, "loss": 0.0013, "step": 1355 }, { "epoch": 5.117202268431002, "grad_norm": 0.04302045330405235, "learning_rate": 3.253956995733798e-07, "loss": 0.0001, "step": 1356 }, { "epoch": 5.120982986767486, "grad_norm": 0.04390261322259903, "learning_rate": 3.2269331687794695e-07, "loss": 0.0001, "step": 1357 }, { "epoch": 5.12476370510397, "grad_norm": 0.1536787450313568, "learning_rate": 3.200014281547076e-07, "loss": 0.0003, "step": 1358 }, { "epoch": 5.128544423440454, "grad_norm": 0.5643743276596069, "learning_rate": 3.1732004637776594e-07, "loss": 0.0008, "step": 1359 }, { "epoch": 5.1323251417769375, "grad_norm": 0.03174169734120369, "learning_rate": 3.1464918447058623e-07, "loss": 0.0001, "step": 1360 }, { "epoch": 5.136105860113422, "grad_norm": 0.12881946563720703, "learning_rate": 3.1198885530593025e-07, "loss": 0.0002, "step": 1361 }, { "epoch": 5.1398865784499055, "grad_norm": 0.13744963705539703, "learning_rate": 3.0933907170579646e-07, "loss": 0.0006, "step": 1362 }, { "epoch": 5.143667296786389, "grad_norm": 0.04274684563279152, "learning_rate": 3.0669984644135455e-07, "loss": 0.0001, "step": 1363 }, { "epoch": 5.147448015122873, "grad_norm": 0.3116391599178314, "learning_rate": 3.040711922328871e-07, "loss": 0.0014, "step": 1364 }, { "epoch": 5.151228733459357, "grad_norm": 0.23607707023620605, "learning_rate": 3.014531217497282e-07, "loss": 0.0007, "step": 1365 }, { "epoch": 5.155009451795841, "grad_norm": 0.055339179933071136, "learning_rate": 2.9884564761020083e-07, "loss": 0.0001, "step": 1366 }, { "epoch": 5.158790170132325, "grad_norm": 0.11408417671918869, "learning_rate": 2.9624878238155605e-07, "loss": 0.0003, "step": 1367 }, { "epoch": 5.162570888468809, "grad_norm": 0.12866030633449554, "learning_rate": 2.9366253857991326e-07, "loss": 0.0004, "step": 1368 }, { "epoch": 5.166351606805293, "grad_norm": 0.08845337480306625, "learning_rate": 2.910869286702023e-07, "loss": 0.0002, "step": 1369 }, { "epoch": 5.170132325141777, "grad_norm": 0.15770214796066284, "learning_rate": 2.8852196506609644e-07, "loss": 0.0002, "step": 1370 }, { "epoch": 5.173913043478261, "grad_norm": 0.1258392333984375, "learning_rate": 2.8596766012996047e-07, "loss": 0.0003, "step": 1371 }, { "epoch": 5.177693761814745, "grad_norm": 0.012674415484070778, "learning_rate": 2.8342402617278606e-07, "loss": 0.0001, "step": 1372 }, { "epoch": 5.181474480151229, "grad_norm": 0.005511714145541191, "learning_rate": 2.808910754541336e-07, "loss": 0.0001, "step": 1373 }, { "epoch": 5.185255198487713, "grad_norm": 0.2423507422208786, "learning_rate": 2.78368820182075e-07, "loss": 0.0007, "step": 1374 }, { "epoch": 5.189035916824197, "grad_norm": 0.09085964411497116, "learning_rate": 2.7585727251313196e-07, "loss": 0.0005, "step": 1375 }, { "epoch": 5.19281663516068, "grad_norm": 0.29729926586151123, "learning_rate": 2.733564445522191e-07, "loss": 0.0011, "step": 1376 }, { "epoch": 5.196597353497165, "grad_norm": 0.008104115724563599, "learning_rate": 2.7086634835258416e-07, "loss": 0.0001, "step": 1377 }, { "epoch": 5.200378071833648, "grad_norm": 0.01323778461664915, "learning_rate": 2.683869959157534e-07, "loss": 0.0001, "step": 1378 }, { "epoch": 5.2041587901701325, "grad_norm": 0.06875714659690857, "learning_rate": 2.6591839919146963e-07, "loss": 0.0003, "step": 1379 }, { "epoch": 5.207939508506616, "grad_norm": 0.07459014654159546, "learning_rate": 2.6346057007763603e-07, "loss": 0.0002, "step": 1380 }, { "epoch": 5.2117202268431, "grad_norm": 0.03255518525838852, "learning_rate": 2.610135204202599e-07, "loss": 0.0001, "step": 1381 }, { "epoch": 5.215500945179584, "grad_norm": 0.5607032775878906, "learning_rate": 2.585772620133947e-07, "loss": 0.0005, "step": 1382 }, { "epoch": 5.219281663516068, "grad_norm": 0.08286086469888687, "learning_rate": 2.561518065990834e-07, "loss": 0.0002, "step": 1383 }, { "epoch": 5.223062381852552, "grad_norm": 0.0021001130808144808, "learning_rate": 2.5373716586730047e-07, "loss": 0.0, "step": 1384 }, { "epoch": 5.226843100189036, "grad_norm": 0.11454706639051437, "learning_rate": 2.513333514558994e-07, "loss": 0.0009, "step": 1385 }, { "epoch": 5.23062381852552, "grad_norm": 0.01842721365392208, "learning_rate": 2.4894037495055036e-07, "loss": 0.0001, "step": 1386 }, { "epoch": 5.234404536862003, "grad_norm": 0.035587288439273834, "learning_rate": 2.465582478846917e-07, "loss": 0.0001, "step": 1387 }, { "epoch": 5.238185255198488, "grad_norm": 0.08331210911273956, "learning_rate": 2.4418698173946873e-07, "loss": 0.0002, "step": 1388 }, { "epoch": 5.241965973534971, "grad_norm": 0.4966404438018799, "learning_rate": 2.418265879436804e-07, "loss": 0.0014, "step": 1389 }, { "epoch": 5.245746691871456, "grad_norm": 0.3974553048610687, "learning_rate": 2.39477077873724e-07, "loss": 0.0015, "step": 1390 }, { "epoch": 5.249527410207939, "grad_norm": 0.057664141058921814, "learning_rate": 2.3713846285354226e-07, "loss": 0.0002, "step": 1391 }, { "epoch": 5.253308128544424, "grad_norm": 0.04705015569925308, "learning_rate": 2.3481075415456494e-07, "loss": 0.0002, "step": 1392 }, { "epoch": 5.257088846880907, "grad_norm": 0.08599993586540222, "learning_rate": 2.3249396299565685e-07, "loss": 0.0003, "step": 1393 }, { "epoch": 5.260869565217392, "grad_norm": 0.4913703203201294, "learning_rate": 2.3018810054306505e-07, "loss": 0.0017, "step": 1394 }, { "epoch": 5.264650283553875, "grad_norm": 0.20572242140769958, "learning_rate": 2.2789317791036186e-07, "loss": 0.0005, "step": 1395 }, { "epoch": 5.2684310018903595, "grad_norm": 0.009350297041237354, "learning_rate": 2.2560920615839339e-07, "loss": 0.0001, "step": 1396 }, { "epoch": 5.272211720226843, "grad_norm": 0.12118514627218246, "learning_rate": 2.2333619629522519e-07, "loss": 0.0005, "step": 1397 }, { "epoch": 5.2759924385633274, "grad_norm": 0.06716551631689072, "learning_rate": 2.2107415927609177e-07, "loss": 0.0002, "step": 1398 }, { "epoch": 5.279773156899811, "grad_norm": 0.03946450725197792, "learning_rate": 2.1882310600333968e-07, "loss": 0.0001, "step": 1399 }, { "epoch": 5.2835538752362945, "grad_norm": 0.33042624592781067, "learning_rate": 2.1658304732637719e-07, "loss": 0.0026, "step": 1400 }, { "epoch": 5.287334593572779, "grad_norm": 0.12655547261238098, "learning_rate": 2.1435399404162393e-07, "loss": 0.0005, "step": 1401 }, { "epoch": 5.291115311909262, "grad_norm": 0.03512566536664963, "learning_rate": 2.1213595689245386e-07, "loss": 0.0001, "step": 1402 }, { "epoch": 5.294896030245747, "grad_norm": 0.21356050670146942, "learning_rate": 2.0992894656914898e-07, "loss": 0.001, "step": 1403 }, { "epoch": 5.29867674858223, "grad_norm": 0.12144608795642853, "learning_rate": 2.0773297370884376e-07, "loss": 0.0003, "step": 1404 }, { "epoch": 5.302457466918715, "grad_norm": 0.11224925518035889, "learning_rate": 2.0554804889547587e-07, "loss": 0.0001, "step": 1405 }, { "epoch": 5.306238185255198, "grad_norm": 0.3938990831375122, "learning_rate": 2.0337418265973435e-07, "loss": 0.0008, "step": 1406 }, { "epoch": 5.310018903591683, "grad_norm": 0.09306345134973526, "learning_rate": 2.0121138547901014e-07, "loss": 0.0003, "step": 1407 }, { "epoch": 5.313799621928166, "grad_norm": 0.16920205950737, "learning_rate": 1.990596677773435e-07, "loss": 0.0005, "step": 1408 }, { "epoch": 5.317580340264651, "grad_norm": 0.2709536552429199, "learning_rate": 1.9691903992537508e-07, "loss": 0.0003, "step": 1409 }, { "epoch": 5.321361058601134, "grad_norm": 0.6464247703552246, "learning_rate": 1.947895122402968e-07, "loss": 0.0006, "step": 1410 }, { "epoch": 5.325141776937619, "grad_norm": 0.29522690176963806, "learning_rate": 1.9267109498579962e-07, "loss": 0.0007, "step": 1411 }, { "epoch": 5.328922495274102, "grad_norm": 0.04720759764313698, "learning_rate": 1.9056379837202626e-07, "loss": 0.0001, "step": 1412 }, { "epoch": 5.332703213610586, "grad_norm": 0.0806199163198471, "learning_rate": 1.8846763255552098e-07, "loss": 0.0003, "step": 1413 }, { "epoch": 5.33648393194707, "grad_norm": 0.06869063526391983, "learning_rate": 1.8638260763918198e-07, "loss": 0.0001, "step": 1414 }, { "epoch": 5.340264650283554, "grad_norm": 0.07782938331365585, "learning_rate": 1.8430873367220908e-07, "loss": 0.0003, "step": 1415 }, { "epoch": 5.344045368620038, "grad_norm": 0.2248040735721588, "learning_rate": 1.8224602065006035e-07, "loss": 0.0009, "step": 1416 }, { "epoch": 5.3478260869565215, "grad_norm": 0.31432026624679565, "learning_rate": 1.8019447851440165e-07, "loss": 0.0007, "step": 1417 }, { "epoch": 5.351606805293006, "grad_norm": 0.20966413617134094, "learning_rate": 1.781541171530554e-07, "loss": 0.0032, "step": 1418 }, { "epoch": 5.355387523629489, "grad_norm": 0.23049578070640564, "learning_rate": 1.7612494639995926e-07, "loss": 0.0007, "step": 1419 }, { "epoch": 5.359168241965974, "grad_norm": 0.007607457228004932, "learning_rate": 1.7410697603511383e-07, "loss": 0.0001, "step": 1420 }, { "epoch": 5.362948960302457, "grad_norm": 0.4492802321910858, "learning_rate": 1.721002157845375e-07, "loss": 0.0014, "step": 1421 }, { "epoch": 5.366729678638942, "grad_norm": 0.2642250061035156, "learning_rate": 1.7010467532021902e-07, "loss": 0.0004, "step": 1422 }, { "epoch": 5.370510396975425, "grad_norm": 0.2687743604183197, "learning_rate": 1.6812036426007177e-07, "loss": 0.0002, "step": 1423 }, { "epoch": 5.374291115311909, "grad_norm": 0.027900557965040207, "learning_rate": 1.661472921678864e-07, "loss": 0.0001, "step": 1424 }, { "epoch": 5.378071833648393, "grad_norm": 0.14470556378364563, "learning_rate": 1.64185468553284e-07, "loss": 0.0003, "step": 1425 }, { "epoch": 5.381852551984877, "grad_norm": 0.12733840942382812, "learning_rate": 1.6223490287167354e-07, "loss": 0.0003, "step": 1426 }, { "epoch": 5.385633270321361, "grad_norm": 0.09421952068805695, "learning_rate": 1.602956045242024e-07, "loss": 0.0001, "step": 1427 }, { "epoch": 5.389413988657845, "grad_norm": 0.09196946024894714, "learning_rate": 1.5836758285771303e-07, "loss": 0.0003, "step": 1428 }, { "epoch": 5.393194706994329, "grad_norm": 0.07771217823028564, "learning_rate": 1.5645084716469778e-07, "loss": 0.0002, "step": 1429 }, { "epoch": 5.396975425330813, "grad_norm": 0.013214212842285633, "learning_rate": 1.5454540668325442e-07, "loss": 0.0001, "step": 1430 }, { "epoch": 5.400756143667297, "grad_norm": 0.003943422343581915, "learning_rate": 1.526512705970401e-07, "loss": 0.0, "step": 1431 }, { "epoch": 5.404536862003781, "grad_norm": 0.10806100070476532, "learning_rate": 1.507684480352292e-07, "loss": 0.0002, "step": 1432 }, { "epoch": 5.408317580340265, "grad_norm": 0.36559733748435974, "learning_rate": 1.488969480724678e-07, "loss": 0.0005, "step": 1433 }, { "epoch": 5.4120982986767485, "grad_norm": 0.16127647459506989, "learning_rate": 1.4703677972882996e-07, "loss": 0.0005, "step": 1434 }, { "epoch": 5.415879017013232, "grad_norm": 0.22948214411735535, "learning_rate": 1.4518795196977575e-07, "loss": 0.0005, "step": 1435 }, { "epoch": 5.4196597353497165, "grad_norm": 0.03453262895345688, "learning_rate": 1.4335047370610616e-07, "loss": 0.0001, "step": 1436 }, { "epoch": 5.4234404536862, "grad_norm": 0.24430623650550842, "learning_rate": 1.415243537939215e-07, "loss": 0.0007, "step": 1437 }, { "epoch": 5.427221172022684, "grad_norm": 0.2824579179286957, "learning_rate": 1.3970960103457722e-07, "loss": 0.0047, "step": 1438 }, { "epoch": 5.431001890359168, "grad_norm": 0.021669110283255577, "learning_rate": 1.37906224174644e-07, "loss": 0.0001, "step": 1439 }, { "epoch": 5.434782608695652, "grad_norm": 0.15396416187286377, "learning_rate": 1.3611423190586308e-07, "loss": 0.0004, "step": 1440 }, { "epoch": 5.438563327032136, "grad_norm": 1.791243553161621, "learning_rate": 1.3433363286510514e-07, "loss": 0.0022, "step": 1441 }, { "epoch": 5.44234404536862, "grad_norm": 0.2876231074333191, "learning_rate": 1.325644356343289e-07, "loss": 0.0003, "step": 1442 }, { "epoch": 5.446124763705104, "grad_norm": 0.12883572280406952, "learning_rate": 1.3080664874054127e-07, "loss": 0.0002, "step": 1443 }, { "epoch": 5.449905482041588, "grad_norm": 0.08326122164726257, "learning_rate": 1.2906028065575255e-07, "loss": 0.0003, "step": 1444 }, { "epoch": 5.453686200378072, "grad_norm": 0.08425617963075638, "learning_rate": 1.2732533979693902e-07, "loss": 0.0002, "step": 1445 }, { "epoch": 5.457466918714556, "grad_norm": 0.4331730604171753, "learning_rate": 1.256018345260013e-07, "loss": 0.001, "step": 1446 }, { "epoch": 5.46124763705104, "grad_norm": 0.11338285356760025, "learning_rate": 1.238897731497224e-07, "loss": 0.0003, "step": 1447 }, { "epoch": 5.465028355387524, "grad_norm": 0.2402881532907486, "learning_rate": 1.2218916391973118e-07, "loss": 0.0007, "step": 1448 }, { "epoch": 5.468809073724008, "grad_norm": 0.04669580236077309, "learning_rate": 1.2050001503245894e-07, "loss": 0.0002, "step": 1449 }, { "epoch": 5.472589792060491, "grad_norm": 0.0060251131653785706, "learning_rate": 1.1882233462910259e-07, "loss": 0.0001, "step": 1450 }, { "epoch": 5.4763705103969755, "grad_norm": 0.029734700918197632, "learning_rate": 1.1715613079558319e-07, "loss": 0.0002, "step": 1451 }, { "epoch": 5.480151228733459, "grad_norm": 0.07456950843334198, "learning_rate": 1.1550141156250944e-07, "loss": 0.0002, "step": 1452 }, { "epoch": 5.4839319470699435, "grad_norm": 0.03908802196383476, "learning_rate": 1.1385818490513733e-07, "loss": 0.0001, "step": 1453 }, { "epoch": 5.487712665406427, "grad_norm": 0.14960382878780365, "learning_rate": 1.1222645874333082e-07, "loss": 0.0006, "step": 1454 }, { "epoch": 5.491493383742911, "grad_norm": 0.3117847740650177, "learning_rate": 1.1060624094152678e-07, "loss": 0.0008, "step": 1455 }, { "epoch": 5.495274102079395, "grad_norm": 0.595193088054657, "learning_rate": 1.0899753930869395e-07, "loss": 0.0013, "step": 1456 }, { "epoch": 5.499054820415879, "grad_norm": 0.16845948994159698, "learning_rate": 1.0740036159829714e-07, "loss": 0.0005, "step": 1457 }, { "epoch": 5.502835538752363, "grad_norm": 0.05613188445568085, "learning_rate": 1.0581471550825812e-07, "loss": 0.0003, "step": 1458 }, { "epoch": 5.506616257088847, "grad_norm": 0.6277530193328857, "learning_rate": 1.0424060868092195e-07, "loss": 0.0007, "step": 1459 }, { "epoch": 5.510396975425331, "grad_norm": 0.10682839155197144, "learning_rate": 1.0267804870301546e-07, "loss": 0.0004, "step": 1460 }, { "epoch": 5.514177693761814, "grad_norm": 0.061431948095560074, "learning_rate": 1.011270431056141e-07, "loss": 0.0003, "step": 1461 }, { "epoch": 5.517958412098299, "grad_norm": 0.03574751690030098, "learning_rate": 9.958759936410572e-08, "loss": 0.0001, "step": 1462 }, { "epoch": 5.521739130434782, "grad_norm": 0.06222573295235634, "learning_rate": 9.805972489815103e-08, "loss": 0.0002, "step": 1463 }, { "epoch": 5.525519848771267, "grad_norm": 0.03182777762413025, "learning_rate": 9.654342707165254e-08, "loss": 0.0001, "step": 1464 }, { "epoch": 5.52930056710775, "grad_norm": 0.23973198235034943, "learning_rate": 9.503871319271552e-08, "loss": 0.0006, "step": 1465 }, { "epoch": 5.533081285444235, "grad_norm": 0.016300465911626816, "learning_rate": 9.354559051361457e-08, "loss": 0.0001, "step": 1466 }, { "epoch": 5.536862003780718, "grad_norm": 0.07304215431213379, "learning_rate": 9.20640662307576e-08, "loss": 0.0001, "step": 1467 }, { "epoch": 5.5406427221172025, "grad_norm": 0.20647095143795013, "learning_rate": 9.059414748465278e-08, "loss": 0.0007, "step": 1468 }, { "epoch": 5.544423440453686, "grad_norm": 0.29599592089653015, "learning_rate": 8.913584135987197e-08, "loss": 0.0009, "step": 1469 }, { "epoch": 5.5482041587901705, "grad_norm": 0.09809119999408722, "learning_rate": 8.768915488501728e-08, "loss": 0.0005, "step": 1470 }, { "epoch": 5.551984877126654, "grad_norm": 0.06101972237229347, "learning_rate": 8.625409503268956e-08, "loss": 0.0002, "step": 1471 }, { "epoch": 5.5557655954631375, "grad_norm": 0.09933365881443024, "learning_rate": 8.483066871945083e-08, "loss": 0.0002, "step": 1472 }, { "epoch": 5.559546313799622, "grad_norm": 0.1317649781703949, "learning_rate": 8.341888280579386e-08, "loss": 0.0003, "step": 1473 }, { "epoch": 5.563327032136106, "grad_norm": 0.032011978328228, "learning_rate": 8.201874409610732e-08, "loss": 0.0001, "step": 1474 }, { "epoch": 5.56710775047259, "grad_norm": 0.074247807264328, "learning_rate": 8.063025933864515e-08, "loss": 0.0003, "step": 1475 }, { "epoch": 5.570888468809073, "grad_norm": 0.01774515025317669, "learning_rate": 7.92534352254909e-08, "loss": 0.0001, "step": 1476 }, { "epoch": 5.574669187145558, "grad_norm": 0.12577727437019348, "learning_rate": 7.788827839252888e-08, "loss": 0.0003, "step": 1477 }, { "epoch": 5.578449905482041, "grad_norm": 0.10297442227602005, "learning_rate": 7.653479541941039e-08, "loss": 0.0004, "step": 1478 }, { "epoch": 5.582230623818526, "grad_norm": 0.7751023173332214, "learning_rate": 7.519299282952086e-08, "loss": 0.0025, "step": 1479 }, { "epoch": 5.586011342155009, "grad_norm": 0.0045355260372161865, "learning_rate": 7.38628770899516e-08, "loss": 0.0, "step": 1480 }, { "epoch": 5.589792060491494, "grad_norm": 0.26779571175575256, "learning_rate": 7.254445461146509e-08, "loss": 0.0008, "step": 1481 }, { "epoch": 5.593572778827977, "grad_norm": 0.09517142176628113, "learning_rate": 7.123773174846694e-08, "loss": 0.0004, "step": 1482 }, { "epoch": 5.597353497164462, "grad_norm": 0.18995550274848938, "learning_rate": 6.994271479897313e-08, "loss": 0.0006, "step": 1483 }, { "epoch": 5.601134215500945, "grad_norm": 0.031352896243333817, "learning_rate": 6.865941000458148e-08, "loss": 0.0001, "step": 1484 }, { "epoch": 5.60491493383743, "grad_norm": 0.1497746706008911, "learning_rate": 6.738782355044048e-08, "loss": 0.0004, "step": 1485 }, { "epoch": 5.608695652173913, "grad_norm": 0.06485506892204285, "learning_rate": 6.612796156521912e-08, "loss": 0.0003, "step": 1486 }, { "epoch": 5.612476370510397, "grad_norm": 0.18997794389724731, "learning_rate": 6.487983012107873e-08, "loss": 0.0006, "step": 1487 }, { "epoch": 5.616257088846881, "grad_norm": 0.09351346641778946, "learning_rate": 6.364343523364264e-08, "loss": 0.0002, "step": 1488 }, { "epoch": 5.6200378071833645, "grad_norm": 0.12646786868572235, "learning_rate": 6.241878286196684e-08, "loss": 0.0003, "step": 1489 }, { "epoch": 5.623818525519849, "grad_norm": 0.24480198323726654, "learning_rate": 6.120587890851293e-08, "loss": 0.0001, "step": 1490 }, { "epoch": 5.6275992438563325, "grad_norm": 0.005080193281173706, "learning_rate": 6.000472921911776e-08, "loss": 0.0001, "step": 1491 }, { "epoch": 5.631379962192817, "grad_norm": 0.2879715859889984, "learning_rate": 5.8815339582966316e-08, "loss": 0.0008, "step": 1492 }, { "epoch": 5.6351606805293, "grad_norm": 0.13152775168418884, "learning_rate": 5.763771573256416e-08, "loss": 0.0005, "step": 1493 }, { "epoch": 5.638941398865785, "grad_norm": 0.11099958419799805, "learning_rate": 5.6471863343708636e-08, "loss": 0.0002, "step": 1494 }, { "epoch": 5.642722117202268, "grad_norm": 0.00900343619287014, "learning_rate": 5.531778803546217e-08, "loss": 0.0001, "step": 1495 }, { "epoch": 5.646502835538753, "grad_norm": 0.052634697407484055, "learning_rate": 5.4175495370125674e-08, "loss": 0.0002, "step": 1496 }, { "epoch": 5.650283553875236, "grad_norm": 0.09383983910083771, "learning_rate": 5.304499085321074e-08, "loss": 0.0001, "step": 1497 }, { "epoch": 5.65406427221172, "grad_norm": 0.06193653121590614, "learning_rate": 5.192627993341359e-08, "loss": 0.0002, "step": 1498 }, { "epoch": 5.657844990548204, "grad_norm": 0.16941288113594055, "learning_rate": 5.081936800258897e-08, "loss": 0.0003, "step": 1499 }, { "epoch": 5.661625708884688, "grad_norm": 0.1040022000670433, "learning_rate": 4.972426039572431e-08, "loss": 0.0003, "step": 1500 }, { "epoch": 5.665406427221172, "grad_norm": 0.055065304040908813, "learning_rate": 4.864096239091287e-08, "loss": 0.0002, "step": 1501 }, { "epoch": 5.669187145557656, "grad_norm": 0.05536071956157684, "learning_rate": 4.7569479209329804e-08, "loss": 0.0001, "step": 1502 }, { "epoch": 5.67296786389414, "grad_norm": 0.40613073110580444, "learning_rate": 4.65098160152061e-08, "loss": 0.0012, "step": 1503 }, { "epoch": 5.676748582230624, "grad_norm": 0.27516570687294006, "learning_rate": 4.5461977915803866e-08, "loss": 0.0015, "step": 1504 }, { "epoch": 5.680529300567108, "grad_norm": 0.02240636572241783, "learning_rate": 4.442596996139192e-08, "loss": 0.0001, "step": 1505 }, { "epoch": 5.6843100189035916, "grad_norm": 0.011783379130065441, "learning_rate": 4.3401797145221356e-08, "loss": 0.0001, "step": 1506 }, { "epoch": 5.688090737240076, "grad_norm": 0.11582031846046448, "learning_rate": 4.238946440350111e-08, "loss": 0.0003, "step": 1507 }, { "epoch": 5.6918714555765595, "grad_norm": 0.15164542198181152, "learning_rate": 4.1388976615374664e-08, "loss": 0.0002, "step": 1508 }, { "epoch": 5.695652173913043, "grad_norm": 0.22043666243553162, "learning_rate": 4.040033860289644e-08, "loss": 0.0004, "step": 1509 }, { "epoch": 5.699432892249527, "grad_norm": 0.6590160131454468, "learning_rate": 3.942355513100793e-08, "loss": 0.0012, "step": 1510 }, { "epoch": 5.703213610586012, "grad_norm": 0.276409775018692, "learning_rate": 3.8458630907516336e-08, "loss": 0.0006, "step": 1511 }, { "epoch": 5.706994328922495, "grad_norm": 0.03431544452905655, "learning_rate": 3.7505570583069304e-08, "loss": 0.0001, "step": 1512 }, { "epoch": 5.710775047258979, "grad_norm": 0.2288655787706375, "learning_rate": 3.656437875113522e-08, "loss": 0.0005, "step": 1513 }, { "epoch": 5.714555765595463, "grad_norm": 0.17267882823944092, "learning_rate": 3.563505994797961e-08, "loss": 0.0005, "step": 1514 }, { "epoch": 5.718336483931947, "grad_norm": 0.031439099460840225, "learning_rate": 3.4717618652642947e-08, "loss": 0.0001, "step": 1515 }, { "epoch": 5.722117202268431, "grad_norm": 0.2682400047779083, "learning_rate": 3.381205928692094e-08, "loss": 0.0005, "step": 1516 }, { "epoch": 5.725897920604915, "grad_norm": 0.15354733169078827, "learning_rate": 3.2918386215339824e-08, "loss": 0.0002, "step": 1517 }, { "epoch": 5.729678638941399, "grad_norm": 0.028868496417999268, "learning_rate": 3.203660374513945e-08, "loss": 0.0001, "step": 1518 }, { "epoch": 5.733459357277883, "grad_norm": 0.07966670393943787, "learning_rate": 3.1166716126249664e-08, "loss": 0.0002, "step": 1519 }, { "epoch": 5.737240075614367, "grad_norm": 0.07197248190641403, "learning_rate": 3.030872755127034e-08, "loss": 0.0002, "step": 1520 }, { "epoch": 5.741020793950851, "grad_norm": 0.026676548644900322, "learning_rate": 2.9462642155451972e-08, "loss": 0.0001, "step": 1521 }, { "epoch": 5.744801512287335, "grad_norm": 0.053479839116334915, "learning_rate": 2.8628464016675073e-08, "loss": 0.0002, "step": 1522 }, { "epoch": 5.748582230623819, "grad_norm": 0.05728183314204216, "learning_rate": 2.7806197155431093e-08, "loss": 0.0002, "step": 1523 }, { "epoch": 5.752362948960302, "grad_norm": 0.015395663678646088, "learning_rate": 2.6995845534802113e-08, "loss": 0.0001, "step": 1524 }, { "epoch": 5.7561436672967865, "grad_norm": 0.03682807460427284, "learning_rate": 2.6197413060442268e-08, "loss": 0.0001, "step": 1525 }, { "epoch": 5.75992438563327, "grad_norm": 0.10636741667985916, "learning_rate": 2.5410903580559698e-08, "loss": 0.0004, "step": 1526 }, { "epoch": 5.763705103969754, "grad_norm": 0.3648461401462555, "learning_rate": 2.4636320885896294e-08, "loss": 0.0017, "step": 1527 }, { "epoch": 5.767485822306238, "grad_norm": 0.0164412222802639, "learning_rate": 2.387366870971103e-08, "loss": 0.0001, "step": 1528 }, { "epoch": 5.771266540642722, "grad_norm": 0.06743861734867096, "learning_rate": 2.3122950727761383e-08, "loss": 0.0001, "step": 1529 }, { "epoch": 5.775047258979206, "grad_norm": 0.19245493412017822, "learning_rate": 2.2384170558284456e-08, "loss": 0.0003, "step": 1530 }, { "epoch": 5.77882797731569, "grad_norm": 0.8845333456993103, "learning_rate": 2.165733176198198e-08, "loss": 0.0026, "step": 1531 }, { "epoch": 5.782608695652174, "grad_norm": 0.03313414752483368, "learning_rate": 2.0942437842001727e-08, "loss": 0.0001, "step": 1532 }, { "epoch": 5.786389413988658, "grad_norm": 0.1397186666727066, "learning_rate": 2.0239492243919467e-08, "loss": 0.0004, "step": 1533 }, { "epoch": 5.790170132325142, "grad_norm": 0.3315114676952362, "learning_rate": 1.9548498355725086e-08, "loss": 0.0004, "step": 1534 }, { "epoch": 5.793950850661625, "grad_norm": 0.022824617102742195, "learning_rate": 1.8869459507803722e-08, "loss": 0.0001, "step": 1535 }, { "epoch": 5.79773156899811, "grad_norm": 0.038835592567920685, "learning_rate": 1.8202378972921043e-08, "loss": 0.0002, "step": 1536 }, { "epoch": 5.801512287334593, "grad_norm": 0.053325552493333817, "learning_rate": 1.7547259966207708e-08, "loss": 0.0001, "step": 1537 }, { "epoch": 5.805293005671078, "grad_norm": 0.1744251549243927, "learning_rate": 1.6904105645142443e-08, "loss": 0.0005, "step": 1538 }, { "epoch": 5.809073724007561, "grad_norm": 0.05694884434342384, "learning_rate": 1.6272919109538986e-08, "loss": 0.0001, "step": 1539 }, { "epoch": 5.812854442344046, "grad_norm": 0.3757440447807312, "learning_rate": 1.565370340152833e-08, "loss": 0.0015, "step": 1540 }, { "epoch": 5.816635160680529, "grad_norm": 0.16067902743816376, "learning_rate": 1.504646150554706e-08, "loss": 0.0003, "step": 1541 }, { "epoch": 5.8204158790170135, "grad_norm": 0.16800613701343536, "learning_rate": 1.4451196348320706e-08, "loss": 0.0005, "step": 1542 }, { "epoch": 5.824196597353497, "grad_norm": 0.010517935268580914, "learning_rate": 1.3867910798850692e-08, "loss": 0.0001, "step": 1543 }, { "epoch": 5.8279773156899815, "grad_norm": 0.06694194674491882, "learning_rate": 1.3296607668400186e-08, "loss": 0.0002, "step": 1544 }, { "epoch": 5.831758034026465, "grad_norm": 0.3510046601295471, "learning_rate": 1.2737289710481048e-08, "loss": 0.0024, "step": 1545 }, { "epoch": 5.8355387523629485, "grad_norm": 0.06765420734882355, "learning_rate": 1.2189959620839686e-08, "loss": 0.0002, "step": 1546 }, { "epoch": 5.839319470699433, "grad_norm": 0.15534651279449463, "learning_rate": 1.1654620037445109e-08, "loss": 0.0003, "step": 1547 }, { "epoch": 5.843100189035916, "grad_norm": 0.19730833172798157, "learning_rate": 1.1131273540474497e-08, "loss": 0.0003, "step": 1548 }, { "epoch": 5.846880907372401, "grad_norm": 0.4626521170139313, "learning_rate": 1.0619922652303494e-08, "loss": 0.003, "step": 1549 }, { "epoch": 5.850661625708884, "grad_norm": 0.04452673718333244, "learning_rate": 1.0120569837491212e-08, "loss": 0.0002, "step": 1550 }, { "epoch": 5.854442344045369, "grad_norm": 0.4129607081413269, "learning_rate": 9.633217502770242e-09, "loss": 0.0017, "step": 1551 }, { "epoch": 5.858223062381852, "grad_norm": 0.3019474744796753, "learning_rate": 9.157867997034165e-09, "loss": 0.0005, "step": 1552 }, { "epoch": 5.862003780718337, "grad_norm": 0.009345407597720623, "learning_rate": 8.694523611326445e-09, "loss": 0.0, "step": 1553 }, { "epoch": 5.86578449905482, "grad_norm": 0.22389154136180878, "learning_rate": 8.243186578829609e-09, "loss": 0.0009, "step": 1554 }, { "epoch": 5.869565217391305, "grad_norm": 0.35236331820487976, "learning_rate": 7.803859074854425e-09, "loss": 0.0005, "step": 1555 }, { "epoch": 5.873345935727788, "grad_norm": 0.04430253058671951, "learning_rate": 7.376543216828791e-09, "loss": 0.0001, "step": 1556 }, { "epoch": 5.877126654064273, "grad_norm": 0.13414061069488525, "learning_rate": 6.9612410642888595e-09, "loss": 0.0004, "step": 1557 }, { "epoch": 5.880907372400756, "grad_norm": 0.3142542243003845, "learning_rate": 6.557954618867102e-09, "loss": 0.0012, "step": 1558 }, { "epoch": 5.8846880907372405, "grad_norm": 0.19400948286056519, "learning_rate": 6.16668582428509e-09, "loss": 0.0005, "step": 1559 }, { "epoch": 5.888468809073724, "grad_norm": 0.11932040005922318, "learning_rate": 5.7874365663421195e-09, "loss": 0.0004, "step": 1560 }, { "epoch": 5.892249527410208, "grad_norm": 0.10091670602560043, "learning_rate": 5.420208672907157e-09, "loss": 0.0004, "step": 1561 }, { "epoch": 5.896030245746692, "grad_norm": 0.04721083864569664, "learning_rate": 5.065003913909406e-09, "loss": 0.0001, "step": 1562 }, { "epoch": 5.8998109640831755, "grad_norm": 0.23722729086875916, "learning_rate": 4.7218240013308105e-09, "loss": 0.0013, "step": 1563 }, { "epoch": 5.90359168241966, "grad_norm": 0.15456841886043549, "learning_rate": 4.390670589196622e-09, "loss": 0.0004, "step": 1564 }, { "epoch": 5.9073724007561434, "grad_norm": 0.02825787477195263, "learning_rate": 4.071545273567623e-09, "loss": 0.0001, "step": 1565 }, { "epoch": 5.911153119092628, "grad_norm": 0.012501970864832401, "learning_rate": 3.764449592533748e-09, "loss": 0.0001, "step": 1566 }, { "epoch": 5.914933837429111, "grad_norm": 0.08930414170026779, "learning_rate": 3.4693850262046415e-09, "loss": 0.0003, "step": 1567 }, { "epoch": 5.918714555765596, "grad_norm": 0.017230011522769928, "learning_rate": 3.1863529967041117e-09, "loss": 0.0001, "step": 1568 }, { "epoch": 5.922495274102079, "grad_norm": 0.015877196565270424, "learning_rate": 2.9153548681626344e-09, "loss": 0.0001, "step": 1569 }, { "epoch": 5.926275992438564, "grad_norm": 0.1481255292892456, "learning_rate": 2.65639194671069e-09, "loss": 0.0004, "step": 1570 }, { "epoch": 5.930056710775047, "grad_norm": 0.009568752720952034, "learning_rate": 2.409465480472939e-09, "loss": 0.0001, "step": 1571 }, { "epoch": 5.933837429111531, "grad_norm": 0.07046500593423843, "learning_rate": 2.1745766595612785e-09, "loss": 0.0002, "step": 1572 }, { "epoch": 5.937618147448015, "grad_norm": 0.015811152756214142, "learning_rate": 1.951726616070404e-09, "loss": 0.0001, "step": 1573 }, { "epoch": 5.941398865784499, "grad_norm": 0.04811807721853256, "learning_rate": 1.740916424071426e-09, "loss": 0.0001, "step": 1574 }, { "epoch": 5.945179584120983, "grad_norm": 0.09774316102266312, "learning_rate": 1.5421470996065946e-09, "loss": 0.0003, "step": 1575 }, { "epoch": 5.948960302457467, "grad_norm": 0.2202201783657074, "learning_rate": 1.3554196006854148e-09, "loss": 0.0007, "step": 1576 }, { "epoch": 5.952741020793951, "grad_norm": 0.07608017325401306, "learning_rate": 1.1807348272790952e-09, "loss": 0.0003, "step": 1577 }, { "epoch": 5.956521739130435, "grad_norm": 0.8428947925567627, "learning_rate": 1.018093621316385e-09, "loss": 0.0007, "step": 1578 }, { "epoch": 5.960302457466919, "grad_norm": 0.19594945013523102, "learning_rate": 8.674967666791323e-10, "loss": 0.0004, "step": 1579 }, { "epoch": 5.9640831758034025, "grad_norm": 0.010538334026932716, "learning_rate": 7.289449892000644e-10, "loss": 0.0001, "step": 1580 }, { "epoch": 5.967863894139887, "grad_norm": 0.006928558461368084, "learning_rate": 6.02438956657514e-10, "loss": 0.0001, "step": 1581 }, { "epoch": 5.9716446124763705, "grad_norm": 0.10099567472934723, "learning_rate": 4.87979278772921e-10, "loss": 0.0005, "step": 1582 }, { "epoch": 5.975425330812854, "grad_norm": 0.05212395638227463, "learning_rate": 3.855665072080572e-10, "loss": 0.0002, "step": 1583 }, { "epoch": 5.979206049149338, "grad_norm": 0.41635531187057495, "learning_rate": 2.952011355619733e-10, "loss": 0.0013, "step": 1584 }, { "epoch": 5.982986767485822, "grad_norm": 0.2869757115840912, "learning_rate": 2.1688359936905547e-10, "loss": 0.0007, "step": 1585 }, { "epoch": 5.986767485822306, "grad_norm": 0.3929155766963959, "learning_rate": 1.506142760965279e-10, "loss": 0.0011, "step": 1586 }, { "epoch": 5.99054820415879, "grad_norm": 0.06228722631931305, "learning_rate": 9.639348514278724e-11, "loss": 0.0001, "step": 1587 }, { "epoch": 5.994328922495274, "grad_norm": 0.15704721212387085, "learning_rate": 5.422148783629233e-11, "loss": 0.0004, "step": 1588 }, { "epoch": 5.998109640831758, "grad_norm": 0.3572693467140198, "learning_rate": 2.4098487433066308e-11, "loss": 0.0025, "step": 1589 }, { "epoch": 6.0, "grad_norm": 0.1746542900800705, "learning_rate": 6.024629117529212e-12, "loss": 0.0004, "step": 1590 } ], "logging_steps": 1, "max_steps": 1590, "num_input_tokens_seen": 0, "num_train_epochs": 6, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.298364001621348e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }