{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 3267, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0009182736455463728, "grad_norm": 1.8071929216384888, "learning_rate": 6.134969325153375e-08, "loss": 0.6175, "step": 1 }, { "epoch": 0.0018365472910927456, "grad_norm": 1.8376882076263428, "learning_rate": 1.226993865030675e-07, "loss": 0.6167, "step": 2 }, { "epoch": 0.0027548209366391185, "grad_norm": 2.1415135860443115, "learning_rate": 1.8404907975460125e-07, "loss": 0.3544, "step": 3 }, { "epoch": 0.0036730945821854912, "grad_norm": 1.2032291889190674, "learning_rate": 2.45398773006135e-07, "loss": 0.4794, "step": 4 }, { "epoch": 0.004591368227731864, "grad_norm": 1.5761905908584595, "learning_rate": 3.0674846625766876e-07, "loss": 0.4177, "step": 5 }, { "epoch": 0.005509641873278237, "grad_norm": 1.3120619058609009, "learning_rate": 3.680981595092025e-07, "loss": 0.331, "step": 6 }, { "epoch": 0.006427915518824609, "grad_norm": 1.60676908493042, "learning_rate": 4.2944785276073624e-07, "loss": 0.4466, "step": 7 }, { "epoch": 0.0073461891643709825, "grad_norm": 2.0586845874786377, "learning_rate": 4.9079754601227e-07, "loss": 0.4633, "step": 8 }, { "epoch": 0.008264462809917356, "grad_norm": 2.492260217666626, "learning_rate": 5.521472392638038e-07, "loss": 0.621, "step": 9 }, { "epoch": 0.009182736455463728, "grad_norm": 2.713146209716797, "learning_rate": 6.134969325153375e-07, "loss": 0.4282, "step": 10 }, { "epoch": 0.010101010101010102, "grad_norm": 1.5842885971069336, "learning_rate": 6.748466257668713e-07, "loss": 0.7242, "step": 11 }, { "epoch": 0.011019283746556474, "grad_norm": 2.057081460952759, "learning_rate": 7.36196319018405e-07, "loss": 0.6834, "step": 12 }, { "epoch": 0.011937557392102846, "grad_norm": 1.9753230810165405, "learning_rate": 7.975460122699387e-07, "loss": 0.6395, "step": 13 }, { "epoch": 0.012855831037649219, "grad_norm": 1.3050103187561035, "learning_rate": 8.588957055214725e-07, "loss": 0.4354, "step": 14 }, { "epoch": 0.013774104683195593, "grad_norm": 1.6500200033187866, "learning_rate": 9.202453987730062e-07, "loss": 0.6328, "step": 15 }, { "epoch": 0.014692378328741965, "grad_norm": 3.730330228805542, "learning_rate": 9.8159509202454e-07, "loss": 0.4787, "step": 16 }, { "epoch": 0.015610651974288337, "grad_norm": 1.0558382272720337, "learning_rate": 1.0429447852760736e-06, "loss": 0.341, "step": 17 }, { "epoch": 0.01652892561983471, "grad_norm": 4.048964500427246, "learning_rate": 1.1042944785276075e-06, "loss": 0.6144, "step": 18 }, { "epoch": 0.017447199265381085, "grad_norm": 2.6209568977355957, "learning_rate": 1.165644171779141e-06, "loss": 0.6811, "step": 19 }, { "epoch": 0.018365472910927456, "grad_norm": 2.157410144805908, "learning_rate": 1.226993865030675e-06, "loss": 0.3874, "step": 20 }, { "epoch": 0.01928374655647383, "grad_norm": 3.25909686088562, "learning_rate": 1.2883435582822088e-06, "loss": 0.7502, "step": 21 }, { "epoch": 0.020202020202020204, "grad_norm": 2.355311870574951, "learning_rate": 1.3496932515337425e-06, "loss": 0.5833, "step": 22 }, { "epoch": 0.021120293847566574, "grad_norm": 1.1741214990615845, "learning_rate": 1.411042944785276e-06, "loss": 0.2835, "step": 23 }, { "epoch": 0.02203856749311295, "grad_norm": 1.2681844234466553, "learning_rate": 1.47239263803681e-06, "loss": 0.3504, "step": 24 }, { "epoch": 0.02295684113865932, "grad_norm": 1.79082190990448, "learning_rate": 1.5337423312883437e-06, "loss": 0.5301, "step": 25 }, { "epoch": 0.023875114784205693, "grad_norm": 2.2695329189300537, "learning_rate": 1.5950920245398775e-06, "loss": 0.518, "step": 26 }, { "epoch": 0.024793388429752067, "grad_norm": 1.9395643472671509, "learning_rate": 1.656441717791411e-06, "loss": 0.632, "step": 27 }, { "epoch": 0.025711662075298437, "grad_norm": 1.9656764268875122, "learning_rate": 1.717791411042945e-06, "loss": 0.816, "step": 28 }, { "epoch": 0.02662993572084481, "grad_norm": 1.400386929512024, "learning_rate": 1.7791411042944787e-06, "loss": 0.3399, "step": 29 }, { "epoch": 0.027548209366391185, "grad_norm": 1.8908504247665405, "learning_rate": 1.8404907975460124e-06, "loss": 0.5133, "step": 30 }, { "epoch": 0.028466483011937556, "grad_norm": 1.0727101564407349, "learning_rate": 1.9018404907975464e-06, "loss": 0.3374, "step": 31 }, { "epoch": 0.02938475665748393, "grad_norm": 5.055539608001709, "learning_rate": 1.96319018404908e-06, "loss": 0.7697, "step": 32 }, { "epoch": 0.030303030303030304, "grad_norm": 3.1050431728363037, "learning_rate": 2.0245398773006137e-06, "loss": 0.9296, "step": 33 }, { "epoch": 0.031221303948576674, "grad_norm": 3.2649054527282715, "learning_rate": 2.085889570552147e-06, "loss": 0.9977, "step": 34 }, { "epoch": 0.03213957759412305, "grad_norm": 1.9716732501983643, "learning_rate": 2.147239263803681e-06, "loss": 0.6219, "step": 35 }, { "epoch": 0.03305785123966942, "grad_norm": 4.022078990936279, "learning_rate": 2.208588957055215e-06, "loss": 0.5615, "step": 36 }, { "epoch": 0.03397612488521579, "grad_norm": 1.894551157951355, "learning_rate": 2.2699386503067486e-06, "loss": 0.5217, "step": 37 }, { "epoch": 0.03489439853076217, "grad_norm": 2.1989734172821045, "learning_rate": 2.331288343558282e-06, "loss": 0.6993, "step": 38 }, { "epoch": 0.03581267217630854, "grad_norm": 1.384602665901184, "learning_rate": 2.392638036809816e-06, "loss": 0.7016, "step": 39 }, { "epoch": 0.03673094582185491, "grad_norm": 1.663246750831604, "learning_rate": 2.45398773006135e-06, "loss": 0.5655, "step": 40 }, { "epoch": 0.03764921946740129, "grad_norm": 1.3926371335983276, "learning_rate": 2.5153374233128836e-06, "loss": 0.3273, "step": 41 }, { "epoch": 0.03856749311294766, "grad_norm": 4.090561866760254, "learning_rate": 2.5766871165644175e-06, "loss": 0.8029, "step": 42 }, { "epoch": 0.03948576675849403, "grad_norm": 2.4800820350646973, "learning_rate": 2.638036809815951e-06, "loss": 0.6126, "step": 43 }, { "epoch": 0.04040404040404041, "grad_norm": 1.3543769121170044, "learning_rate": 2.699386503067485e-06, "loss": 0.5926, "step": 44 }, { "epoch": 0.04132231404958678, "grad_norm": 1.4166005849838257, "learning_rate": 2.760736196319019e-06, "loss": 0.5343, "step": 45 }, { "epoch": 0.04224058769513315, "grad_norm": 1.8947725296020508, "learning_rate": 2.822085889570552e-06, "loss": 0.5408, "step": 46 }, { "epoch": 0.04315886134067952, "grad_norm": 2.2579703330993652, "learning_rate": 2.883435582822086e-06, "loss": 0.3853, "step": 47 }, { "epoch": 0.0440771349862259, "grad_norm": 1.6457964181900024, "learning_rate": 2.94478527607362e-06, "loss": 0.4669, "step": 48 }, { "epoch": 0.04499540863177227, "grad_norm": 3.173774242401123, "learning_rate": 3.0061349693251535e-06, "loss": 0.6037, "step": 49 }, { "epoch": 0.04591368227731864, "grad_norm": 1.3746641874313354, "learning_rate": 3.0674846625766875e-06, "loss": 0.3714, "step": 50 }, { "epoch": 0.046831955922865015, "grad_norm": 2.309722900390625, "learning_rate": 3.1288343558282214e-06, "loss": 0.8246, "step": 51 }, { "epoch": 0.047750229568411386, "grad_norm": 1.6517795324325562, "learning_rate": 3.190184049079755e-06, "loss": 0.6211, "step": 52 }, { "epoch": 0.048668503213957756, "grad_norm": 4.159029960632324, "learning_rate": 3.251533742331289e-06, "loss": 0.7755, "step": 53 }, { "epoch": 0.049586776859504134, "grad_norm": 4.4690446853637695, "learning_rate": 3.312883435582822e-06, "loss": 0.6495, "step": 54 }, { "epoch": 0.050505050505050504, "grad_norm": 1.8065283298492432, "learning_rate": 3.374233128834356e-06, "loss": 0.5862, "step": 55 }, { "epoch": 0.051423324150596875, "grad_norm": 3.043789863586426, "learning_rate": 3.43558282208589e-06, "loss": 0.5344, "step": 56 }, { "epoch": 0.05234159779614325, "grad_norm": 0.9541704058647156, "learning_rate": 3.4969325153374234e-06, "loss": 0.2314, "step": 57 }, { "epoch": 0.05325987144168962, "grad_norm": 1.9889920949935913, "learning_rate": 3.5582822085889574e-06, "loss": 0.4671, "step": 58 }, { "epoch": 0.05417814508723599, "grad_norm": 2.0764195919036865, "learning_rate": 3.6196319018404913e-06, "loss": 0.5909, "step": 59 }, { "epoch": 0.05509641873278237, "grad_norm": 2.5211448669433594, "learning_rate": 3.680981595092025e-06, "loss": 0.4815, "step": 60 }, { "epoch": 0.05601469237832874, "grad_norm": 2.1269280910491943, "learning_rate": 3.742331288343559e-06, "loss": 0.5728, "step": 61 }, { "epoch": 0.05693296602387511, "grad_norm": 1.920554518699646, "learning_rate": 3.8036809815950928e-06, "loss": 0.7018, "step": 62 }, { "epoch": 0.05785123966942149, "grad_norm": 3.4384260177612305, "learning_rate": 3.865030674846626e-06, "loss": 0.6774, "step": 63 }, { "epoch": 0.05876951331496786, "grad_norm": 3.1624772548675537, "learning_rate": 3.92638036809816e-06, "loss": 0.6764, "step": 64 }, { "epoch": 0.05968778696051423, "grad_norm": 2.4577443599700928, "learning_rate": 3.987730061349693e-06, "loss": 0.3985, "step": 65 }, { "epoch": 0.06060606060606061, "grad_norm": 1.8537696599960327, "learning_rate": 4.049079754601227e-06, "loss": 0.4033, "step": 66 }, { "epoch": 0.06152433425160698, "grad_norm": 5.230608940124512, "learning_rate": 4.110429447852761e-06, "loss": 0.754, "step": 67 }, { "epoch": 0.06244260789715335, "grad_norm": 2.530317544937134, "learning_rate": 4.171779141104294e-06, "loss": 0.3059, "step": 68 }, { "epoch": 0.06336088154269973, "grad_norm": 1.935369849205017, "learning_rate": 4.233128834355829e-06, "loss": 0.4776, "step": 69 }, { "epoch": 0.0642791551882461, "grad_norm": 2.665940761566162, "learning_rate": 4.294478527607362e-06, "loss": 0.6086, "step": 70 }, { "epoch": 0.06519742883379247, "grad_norm": 1.8953174352645874, "learning_rate": 4.355828220858896e-06, "loss": 0.548, "step": 71 }, { "epoch": 0.06611570247933884, "grad_norm": 4.126149654388428, "learning_rate": 4.41717791411043e-06, "loss": 0.7517, "step": 72 }, { "epoch": 0.06703397612488522, "grad_norm": 3.315345287322998, "learning_rate": 4.478527607361964e-06, "loss": 0.5569, "step": 73 }, { "epoch": 0.06795224977043159, "grad_norm": 4.32432746887207, "learning_rate": 4.539877300613497e-06, "loss": 0.3771, "step": 74 }, { "epoch": 0.06887052341597796, "grad_norm": 2.588759183883667, "learning_rate": 4.601226993865031e-06, "loss": 0.5486, "step": 75 }, { "epoch": 0.06978879706152434, "grad_norm": 5.809013366699219, "learning_rate": 4.662576687116564e-06, "loss": 0.7192, "step": 76 }, { "epoch": 0.0707070707070707, "grad_norm": 2.1375041007995605, "learning_rate": 4.723926380368099e-06, "loss": 0.6602, "step": 77 }, { "epoch": 0.07162534435261708, "grad_norm": 2.2996957302093506, "learning_rate": 4.785276073619632e-06, "loss": 0.5594, "step": 78 }, { "epoch": 0.07254361799816346, "grad_norm": 1.4927068948745728, "learning_rate": 4.846625766871166e-06, "loss": 0.3498, "step": 79 }, { "epoch": 0.07346189164370982, "grad_norm": 1.8580626249313354, "learning_rate": 4.9079754601227e-06, "loss": 0.3062, "step": 80 }, { "epoch": 0.0743801652892562, "grad_norm": 1.7916381359100342, "learning_rate": 4.969325153374234e-06, "loss": 0.4137, "step": 81 }, { "epoch": 0.07529843893480258, "grad_norm": 3.3792197704315186, "learning_rate": 5.030674846625767e-06, "loss": 0.5416, "step": 82 }, { "epoch": 0.07621671258034894, "grad_norm": 5.268410682678223, "learning_rate": 5.092024539877301e-06, "loss": 0.3984, "step": 83 }, { "epoch": 0.07713498622589532, "grad_norm": 2.9354114532470703, "learning_rate": 5.153374233128835e-06, "loss": 0.4329, "step": 84 }, { "epoch": 0.0780532598714417, "grad_norm": 2.829068422317505, "learning_rate": 5.214723926380369e-06, "loss": 0.4088, "step": 85 }, { "epoch": 0.07897153351698806, "grad_norm": 1.6574037075042725, "learning_rate": 5.276073619631902e-06, "loss": 0.456, "step": 86 }, { "epoch": 0.07988980716253444, "grad_norm": 2.3795013427734375, "learning_rate": 5.3374233128834365e-06, "loss": 0.6362, "step": 87 }, { "epoch": 0.08080808080808081, "grad_norm": 1.4017529487609863, "learning_rate": 5.39877300613497e-06, "loss": 0.4186, "step": 88 }, { "epoch": 0.08172635445362718, "grad_norm": 3.832167387008667, "learning_rate": 5.460122699386503e-06, "loss": 0.5266, "step": 89 }, { "epoch": 0.08264462809917356, "grad_norm": 2.9205362796783447, "learning_rate": 5.521472392638038e-06, "loss": 0.3751, "step": 90 }, { "epoch": 0.08356290174471992, "grad_norm": 2.0804810523986816, "learning_rate": 5.582822085889571e-06, "loss": 0.4563, "step": 91 }, { "epoch": 0.0844811753902663, "grad_norm": 5.194464206695557, "learning_rate": 5.644171779141104e-06, "loss": 0.416, "step": 92 }, { "epoch": 0.08539944903581267, "grad_norm": 1.7892860174179077, "learning_rate": 5.7055214723926385e-06, "loss": 0.4358, "step": 93 }, { "epoch": 0.08631772268135904, "grad_norm": 2.607855796813965, "learning_rate": 5.766871165644172e-06, "loss": 0.3799, "step": 94 }, { "epoch": 0.08723599632690542, "grad_norm": 4.317014694213867, "learning_rate": 5.828220858895706e-06, "loss": 0.2596, "step": 95 }, { "epoch": 0.0881542699724518, "grad_norm": 1.760987639427185, "learning_rate": 5.88957055214724e-06, "loss": 0.1852, "step": 96 }, { "epoch": 0.08907254361799816, "grad_norm": 2.362788438796997, "learning_rate": 5.9509202453987735e-06, "loss": 0.3995, "step": 97 }, { "epoch": 0.08999081726354453, "grad_norm": 1.9631767272949219, "learning_rate": 6.012269938650307e-06, "loss": 0.3713, "step": 98 }, { "epoch": 0.09090909090909091, "grad_norm": 1.5262316465377808, "learning_rate": 6.073619631901841e-06, "loss": 0.3726, "step": 99 }, { "epoch": 0.09182736455463728, "grad_norm": 4.732651233673096, "learning_rate": 6.134969325153375e-06, "loss": 0.182, "step": 100 }, { "epoch": 0.09274563820018365, "grad_norm": 2.1742968559265137, "learning_rate": 6.1963190184049085e-06, "loss": 0.532, "step": 101 }, { "epoch": 0.09366391184573003, "grad_norm": 1.4860295057296753, "learning_rate": 6.257668711656443e-06, "loss": 0.2129, "step": 102 }, { "epoch": 0.0945821854912764, "grad_norm": 2.5295426845550537, "learning_rate": 6.319018404907976e-06, "loss": 0.5016, "step": 103 }, { "epoch": 0.09550045913682277, "grad_norm": 1.684767484664917, "learning_rate": 6.38036809815951e-06, "loss": 0.2486, "step": 104 }, { "epoch": 0.09641873278236915, "grad_norm": 2.4607913494110107, "learning_rate": 6.441717791411043e-06, "loss": 0.3307, "step": 105 }, { "epoch": 0.09733700642791551, "grad_norm": 3.824033737182617, "learning_rate": 6.503067484662578e-06, "loss": 0.3437, "step": 106 }, { "epoch": 0.09825528007346189, "grad_norm": 2.288320779800415, "learning_rate": 6.5644171779141105e-06, "loss": 0.4909, "step": 107 }, { "epoch": 0.09917355371900827, "grad_norm": 3.717694044113159, "learning_rate": 6.625766871165644e-06, "loss": 0.4517, "step": 108 }, { "epoch": 0.10009182736455463, "grad_norm": 1.2853457927703857, "learning_rate": 6.687116564417178e-06, "loss": 0.1437, "step": 109 }, { "epoch": 0.10101010101010101, "grad_norm": 1.8578802347183228, "learning_rate": 6.748466257668712e-06, "loss": 0.2047, "step": 110 }, { "epoch": 0.10192837465564739, "grad_norm": 2.2640042304992676, "learning_rate": 6.8098159509202454e-06, "loss": 0.292, "step": 111 }, { "epoch": 0.10284664830119375, "grad_norm": 2.5107994079589844, "learning_rate": 6.87116564417178e-06, "loss": 0.2709, "step": 112 }, { "epoch": 0.10376492194674013, "grad_norm": 1.0788021087646484, "learning_rate": 6.932515337423313e-06, "loss": 0.1201, "step": 113 }, { "epoch": 0.1046831955922865, "grad_norm": 1.632118821144104, "learning_rate": 6.993865030674847e-06, "loss": 0.332, "step": 114 }, { "epoch": 0.10560146923783287, "grad_norm": 8.019002914428711, "learning_rate": 7.055214723926381e-06, "loss": 0.8762, "step": 115 }, { "epoch": 0.10651974288337925, "grad_norm": 2.9700589179992676, "learning_rate": 7.116564417177915e-06, "loss": 0.1621, "step": 116 }, { "epoch": 0.10743801652892562, "grad_norm": 3.4083149433135986, "learning_rate": 7.177914110429448e-06, "loss": 0.2588, "step": 117 }, { "epoch": 0.10835629017447199, "grad_norm": 2.2250559329986572, "learning_rate": 7.239263803680983e-06, "loss": 0.325, "step": 118 }, { "epoch": 0.10927456382001836, "grad_norm": 2.2234904766082764, "learning_rate": 7.300613496932516e-06, "loss": 0.4514, "step": 119 }, { "epoch": 0.11019283746556474, "grad_norm": 1.7380170822143555, "learning_rate": 7.36196319018405e-06, "loss": 0.0997, "step": 120 }, { "epoch": 0.1111111111111111, "grad_norm": 1.342515230178833, "learning_rate": 7.423312883435584e-06, "loss": 0.1809, "step": 121 }, { "epoch": 0.11202938475665748, "grad_norm": 3.6385371685028076, "learning_rate": 7.484662576687118e-06, "loss": 0.1829, "step": 122 }, { "epoch": 0.11294765840220386, "grad_norm": 0.8947508931159973, "learning_rate": 7.54601226993865e-06, "loss": 0.0335, "step": 123 }, { "epoch": 0.11386593204775022, "grad_norm": 1.8219997882843018, "learning_rate": 7.6073619631901856e-06, "loss": 0.3791, "step": 124 }, { "epoch": 0.1147842056932966, "grad_norm": 2.4506819248199463, "learning_rate": 7.668711656441718e-06, "loss": 0.5225, "step": 125 }, { "epoch": 0.11570247933884298, "grad_norm": 2.1354904174804688, "learning_rate": 7.730061349693252e-06, "loss": 0.1222, "step": 126 }, { "epoch": 0.11662075298438934, "grad_norm": 2.823265552520752, "learning_rate": 7.791411042944787e-06, "loss": 0.1821, "step": 127 }, { "epoch": 0.11753902662993572, "grad_norm": 1.3198742866516113, "learning_rate": 7.85276073619632e-06, "loss": 0.2933, "step": 128 }, { "epoch": 0.1184573002754821, "grad_norm": 2.3982014656066895, "learning_rate": 7.914110429447854e-06, "loss": 0.2277, "step": 129 }, { "epoch": 0.11937557392102846, "grad_norm": 2.0608410835266113, "learning_rate": 7.975460122699386e-06, "loss": 0.5264, "step": 130 }, { "epoch": 0.12029384756657484, "grad_norm": 0.7628580331802368, "learning_rate": 8.036809815950921e-06, "loss": 0.1653, "step": 131 }, { "epoch": 0.12121212121212122, "grad_norm": 1.6852771043777466, "learning_rate": 8.098159509202455e-06, "loss": 0.0599, "step": 132 }, { "epoch": 0.12213039485766758, "grad_norm": 2.674825429916382, "learning_rate": 8.159509202453988e-06, "loss": 0.2363, "step": 133 }, { "epoch": 0.12304866850321396, "grad_norm": 1.9686322212219238, "learning_rate": 8.220858895705522e-06, "loss": 0.2692, "step": 134 }, { "epoch": 0.12396694214876033, "grad_norm": 2.0790648460388184, "learning_rate": 8.282208588957055e-06, "loss": 0.2441, "step": 135 }, { "epoch": 0.1248852157943067, "grad_norm": 1.0941556692123413, "learning_rate": 8.343558282208589e-06, "loss": 0.092, "step": 136 }, { "epoch": 0.1258034894398531, "grad_norm": 0.9221530556678772, "learning_rate": 8.404907975460124e-06, "loss": 0.048, "step": 137 }, { "epoch": 0.12672176308539945, "grad_norm": 1.5687788724899292, "learning_rate": 8.466257668711658e-06, "loss": 0.2515, "step": 138 }, { "epoch": 0.12764003673094582, "grad_norm": 1.2704660892486572, "learning_rate": 8.527607361963191e-06, "loss": 0.1033, "step": 139 }, { "epoch": 0.1285583103764922, "grad_norm": 1.7672405242919922, "learning_rate": 8.588957055214725e-06, "loss": 0.2143, "step": 140 }, { "epoch": 0.12947658402203857, "grad_norm": 1.732448935508728, "learning_rate": 8.650306748466258e-06, "loss": 0.0993, "step": 141 }, { "epoch": 0.13039485766758493, "grad_norm": 1.1615676879882812, "learning_rate": 8.711656441717792e-06, "loss": 0.0996, "step": 142 }, { "epoch": 0.13131313131313133, "grad_norm": 1.3651612997055054, "learning_rate": 8.773006134969327e-06, "loss": 0.2665, "step": 143 }, { "epoch": 0.1322314049586777, "grad_norm": 2.677142858505249, "learning_rate": 8.83435582822086e-06, "loss": 0.287, "step": 144 }, { "epoch": 0.13314967860422405, "grad_norm": 3.3156652450561523, "learning_rate": 8.895705521472394e-06, "loss": 0.3568, "step": 145 }, { "epoch": 0.13406795224977044, "grad_norm": 1.211480736732483, "learning_rate": 8.957055214723927e-06, "loss": 0.0472, "step": 146 }, { "epoch": 0.1349862258953168, "grad_norm": 1.2335317134857178, "learning_rate": 9.018404907975461e-06, "loss": 0.2334, "step": 147 }, { "epoch": 0.13590449954086317, "grad_norm": 1.0041553974151611, "learning_rate": 9.079754601226994e-06, "loss": 0.0618, "step": 148 }, { "epoch": 0.13682277318640956, "grad_norm": 1.0807467699050903, "learning_rate": 9.14110429447853e-06, "loss": 0.0565, "step": 149 }, { "epoch": 0.13774104683195593, "grad_norm": 1.9716891050338745, "learning_rate": 9.202453987730062e-06, "loss": 0.1754, "step": 150 }, { "epoch": 0.1386593204775023, "grad_norm": 5.320068836212158, "learning_rate": 9.263803680981595e-06, "loss": 0.344, "step": 151 }, { "epoch": 0.13957759412304868, "grad_norm": 1.2444785833358765, "learning_rate": 9.325153374233129e-06, "loss": 0.2464, "step": 152 }, { "epoch": 0.14049586776859505, "grad_norm": 1.4572241306304932, "learning_rate": 9.386503067484664e-06, "loss": 0.1144, "step": 153 }, { "epoch": 0.1414141414141414, "grad_norm": 2.0562398433685303, "learning_rate": 9.447852760736197e-06, "loss": 0.3371, "step": 154 }, { "epoch": 0.1423324150596878, "grad_norm": 0.8066809177398682, "learning_rate": 9.509202453987731e-06, "loss": 0.0403, "step": 155 }, { "epoch": 0.14325068870523416, "grad_norm": 1.1875790357589722, "learning_rate": 9.570552147239264e-06, "loss": 0.0658, "step": 156 }, { "epoch": 0.14416896235078053, "grad_norm": 1.1361182928085327, "learning_rate": 9.631901840490798e-06, "loss": 0.1587, "step": 157 }, { "epoch": 0.14508723599632692, "grad_norm": 1.807503342628479, "learning_rate": 9.693251533742331e-06, "loss": 0.314, "step": 158 }, { "epoch": 0.14600550964187328, "grad_norm": 1.0723178386688232, "learning_rate": 9.754601226993867e-06, "loss": 0.0205, "step": 159 }, { "epoch": 0.14692378328741965, "grad_norm": 1.449602723121643, "learning_rate": 9.8159509202454e-06, "loss": 0.1006, "step": 160 }, { "epoch": 0.14784205693296604, "grad_norm": 1.2368807792663574, "learning_rate": 9.877300613496934e-06, "loss": 0.1443, "step": 161 }, { "epoch": 0.1487603305785124, "grad_norm": 0.8863618969917297, "learning_rate": 9.938650306748467e-06, "loss": 0.2069, "step": 162 }, { "epoch": 0.14967860422405876, "grad_norm": 1.1009668111801147, "learning_rate": 1e-05, "loss": 0.0289, "step": 163 }, { "epoch": 0.15059687786960516, "grad_norm": 2.15852689743042, "learning_rate": 1.0061349693251534e-05, "loss": 0.152, "step": 164 }, { "epoch": 0.15151515151515152, "grad_norm": 0.8894155621528625, "learning_rate": 1.0122699386503068e-05, "loss": 0.0436, "step": 165 }, { "epoch": 0.15243342516069788, "grad_norm": 1.938822865486145, "learning_rate": 1.0184049079754601e-05, "loss": 0.2343, "step": 166 }, { "epoch": 0.15335169880624427, "grad_norm": 2.079414129257202, "learning_rate": 1.0245398773006137e-05, "loss": 0.261, "step": 167 }, { "epoch": 0.15426997245179064, "grad_norm": 1.536110758781433, "learning_rate": 1.030674846625767e-05, "loss": 0.0371, "step": 168 }, { "epoch": 0.155188246097337, "grad_norm": 3.430079936981201, "learning_rate": 1.0368098159509204e-05, "loss": 0.1834, "step": 169 }, { "epoch": 0.1561065197428834, "grad_norm": 1.932553768157959, "learning_rate": 1.0429447852760737e-05, "loss": 0.0783, "step": 170 }, { "epoch": 0.15702479338842976, "grad_norm": 1.1011507511138916, "learning_rate": 1.049079754601227e-05, "loss": 0.1095, "step": 171 }, { "epoch": 0.15794306703397612, "grad_norm": 1.4907087087631226, "learning_rate": 1.0552147239263804e-05, "loss": 0.2321, "step": 172 }, { "epoch": 0.1588613406795225, "grad_norm": 0.9360774755477905, "learning_rate": 1.061349693251534e-05, "loss": 0.0638, "step": 173 }, { "epoch": 0.15977961432506887, "grad_norm": 1.8636996746063232, "learning_rate": 1.0674846625766873e-05, "loss": 0.2116, "step": 174 }, { "epoch": 0.16069788797061524, "grad_norm": 2.9026846885681152, "learning_rate": 1.0736196319018407e-05, "loss": 0.2827, "step": 175 }, { "epoch": 0.16161616161616163, "grad_norm": 2.139031410217285, "learning_rate": 1.079754601226994e-05, "loss": 0.2997, "step": 176 }, { "epoch": 0.162534435261708, "grad_norm": 1.0117506980895996, "learning_rate": 1.0858895705521472e-05, "loss": 0.0829, "step": 177 }, { "epoch": 0.16345270890725436, "grad_norm": 1.6674681901931763, "learning_rate": 1.0920245398773005e-05, "loss": 0.0543, "step": 178 }, { "epoch": 0.16437098255280075, "grad_norm": 1.1443489789962769, "learning_rate": 1.0981595092024542e-05, "loss": 0.0801, "step": 179 }, { "epoch": 0.1652892561983471, "grad_norm": 1.5915380716323853, "learning_rate": 1.1042944785276076e-05, "loss": 0.1977, "step": 180 }, { "epoch": 0.16620752984389348, "grad_norm": 2.4907948970794678, "learning_rate": 1.1104294478527608e-05, "loss": 0.1139, "step": 181 }, { "epoch": 0.16712580348943984, "grad_norm": 1.4267793893814087, "learning_rate": 1.1165644171779141e-05, "loss": 0.2605, "step": 182 }, { "epoch": 0.16804407713498623, "grad_norm": 1.550787091255188, "learning_rate": 1.1226993865030675e-05, "loss": 0.0825, "step": 183 }, { "epoch": 0.1689623507805326, "grad_norm": 1.1739060878753662, "learning_rate": 1.1288343558282208e-05, "loss": 0.0617, "step": 184 }, { "epoch": 0.16988062442607896, "grad_norm": 4.908758640289307, "learning_rate": 1.1349693251533744e-05, "loss": 0.175, "step": 185 }, { "epoch": 0.17079889807162535, "grad_norm": 1.2631374597549438, "learning_rate": 1.1411042944785277e-05, "loss": 0.0502, "step": 186 }, { "epoch": 0.1717171717171717, "grad_norm": 3.026979923248291, "learning_rate": 1.147239263803681e-05, "loss": 0.3393, "step": 187 }, { "epoch": 0.17263544536271808, "grad_norm": 0.8136230111122131, "learning_rate": 1.1533742331288344e-05, "loss": 0.0678, "step": 188 }, { "epoch": 0.17355371900826447, "grad_norm": 2.0517139434814453, "learning_rate": 1.1595092024539878e-05, "loss": 0.2835, "step": 189 }, { "epoch": 0.17447199265381083, "grad_norm": 2.396540880203247, "learning_rate": 1.1656441717791411e-05, "loss": 0.3226, "step": 190 }, { "epoch": 0.1753902662993572, "grad_norm": 1.247757911682129, "learning_rate": 1.1717791411042945e-05, "loss": 0.0559, "step": 191 }, { "epoch": 0.1763085399449036, "grad_norm": 1.4913488626480103, "learning_rate": 1.177914110429448e-05, "loss": 0.1894, "step": 192 }, { "epoch": 0.17722681359044995, "grad_norm": 3.3776721954345703, "learning_rate": 1.1840490797546013e-05, "loss": 0.3119, "step": 193 }, { "epoch": 0.1781450872359963, "grad_norm": 1.195646047592163, "learning_rate": 1.1901840490797547e-05, "loss": 0.1112, "step": 194 }, { "epoch": 0.1790633608815427, "grad_norm": 0.9839358329772949, "learning_rate": 1.196319018404908e-05, "loss": 0.0265, "step": 195 }, { "epoch": 0.17998163452708907, "grad_norm": 1.4071948528289795, "learning_rate": 1.2024539877300614e-05, "loss": 0.1428, "step": 196 }, { "epoch": 0.18089990817263543, "grad_norm": 2.4342496395111084, "learning_rate": 1.2085889570552148e-05, "loss": 0.1021, "step": 197 }, { "epoch": 0.18181818181818182, "grad_norm": 0.5633125305175781, "learning_rate": 1.2147239263803683e-05, "loss": 0.0293, "step": 198 }, { "epoch": 0.1827364554637282, "grad_norm": 1.315462589263916, "learning_rate": 1.2208588957055216e-05, "loss": 0.1987, "step": 199 }, { "epoch": 0.18365472910927455, "grad_norm": 1.228236198425293, "learning_rate": 1.226993865030675e-05, "loss": 0.1665, "step": 200 }, { "epoch": 0.18457300275482094, "grad_norm": 2.7400388717651367, "learning_rate": 1.2331288343558283e-05, "loss": 0.0733, "step": 201 }, { "epoch": 0.1854912764003673, "grad_norm": 2.858956813812256, "learning_rate": 1.2392638036809817e-05, "loss": 0.388, "step": 202 }, { "epoch": 0.18640955004591367, "grad_norm": 0.9802287220954895, "learning_rate": 1.245398773006135e-05, "loss": 0.0801, "step": 203 }, { "epoch": 0.18732782369146006, "grad_norm": 0.6114208698272705, "learning_rate": 1.2515337423312886e-05, "loss": 0.019, "step": 204 }, { "epoch": 0.18824609733700642, "grad_norm": 2.9587295055389404, "learning_rate": 1.257668711656442e-05, "loss": 0.1656, "step": 205 }, { "epoch": 0.1891643709825528, "grad_norm": 1.2445155382156372, "learning_rate": 1.2638036809815953e-05, "loss": 0.1205, "step": 206 }, { "epoch": 0.19008264462809918, "grad_norm": 0.30377689003944397, "learning_rate": 1.2699386503067486e-05, "loss": 0.0097, "step": 207 }, { "epoch": 0.19100091827364554, "grad_norm": 1.9758464097976685, "learning_rate": 1.276073619631902e-05, "loss": 0.2104, "step": 208 }, { "epoch": 0.1919191919191919, "grad_norm": 2.0401060581207275, "learning_rate": 1.2822085889570552e-05, "loss": 0.0522, "step": 209 }, { "epoch": 0.1928374655647383, "grad_norm": 1.6133114099502563, "learning_rate": 1.2883435582822085e-05, "loss": 0.0751, "step": 210 }, { "epoch": 0.19375573921028466, "grad_norm": 2.6802468299865723, "learning_rate": 1.2944785276073622e-05, "loss": 0.2754, "step": 211 }, { "epoch": 0.19467401285583102, "grad_norm": 1.544118046760559, "learning_rate": 1.3006134969325156e-05, "loss": 0.1, "step": 212 }, { "epoch": 0.19559228650137742, "grad_norm": 2.455885171890259, "learning_rate": 1.3067484662576687e-05, "loss": 0.1405, "step": 213 }, { "epoch": 0.19651056014692378, "grad_norm": 0.8160684704780579, "learning_rate": 1.3128834355828221e-05, "loss": 0.0268, "step": 214 }, { "epoch": 0.19742883379247014, "grad_norm": 1.2557436227798462, "learning_rate": 1.3190184049079754e-05, "loss": 0.0724, "step": 215 }, { "epoch": 0.19834710743801653, "grad_norm": 1.79561448097229, "learning_rate": 1.3251533742331288e-05, "loss": 0.1224, "step": 216 }, { "epoch": 0.1992653810835629, "grad_norm": 2.1937143802642822, "learning_rate": 1.3312883435582823e-05, "loss": 0.1384, "step": 217 }, { "epoch": 0.20018365472910926, "grad_norm": 1.5160998106002808, "learning_rate": 1.3374233128834357e-05, "loss": 0.1341, "step": 218 }, { "epoch": 0.20110192837465565, "grad_norm": 2.3276143074035645, "learning_rate": 1.343558282208589e-05, "loss": 0.064, "step": 219 }, { "epoch": 0.20202020202020202, "grad_norm": 0.6172122955322266, "learning_rate": 1.3496932515337424e-05, "loss": 0.0195, "step": 220 }, { "epoch": 0.20293847566574838, "grad_norm": 2.1404776573181152, "learning_rate": 1.3558282208588957e-05, "loss": 0.1344, "step": 221 }, { "epoch": 0.20385674931129477, "grad_norm": 0.5099356174468994, "learning_rate": 1.3619631901840491e-05, "loss": 0.0195, "step": 222 }, { "epoch": 0.20477502295684114, "grad_norm": 1.3010395765304565, "learning_rate": 1.3680981595092026e-05, "loss": 0.0708, "step": 223 }, { "epoch": 0.2056932966023875, "grad_norm": 1.205059289932251, "learning_rate": 1.374233128834356e-05, "loss": 0.3432, "step": 224 }, { "epoch": 0.2066115702479339, "grad_norm": 1.4143084287643433, "learning_rate": 1.3803680981595093e-05, "loss": 0.0803, "step": 225 }, { "epoch": 0.20752984389348025, "grad_norm": 0.4269416332244873, "learning_rate": 1.3865030674846627e-05, "loss": 0.0113, "step": 226 }, { "epoch": 0.20844811753902662, "grad_norm": 1.873276948928833, "learning_rate": 1.392638036809816e-05, "loss": 0.1496, "step": 227 }, { "epoch": 0.209366391184573, "grad_norm": 4.959446907043457, "learning_rate": 1.3987730061349694e-05, "loss": 0.2261, "step": 228 }, { "epoch": 0.21028466483011937, "grad_norm": 1.704219102859497, "learning_rate": 1.4049079754601229e-05, "loss": 0.049, "step": 229 }, { "epoch": 0.21120293847566574, "grad_norm": 0.7962740063667297, "learning_rate": 1.4110429447852763e-05, "loss": 0.0955, "step": 230 }, { "epoch": 0.21212121212121213, "grad_norm": 1.2131562232971191, "learning_rate": 1.4171779141104296e-05, "loss": 0.0968, "step": 231 }, { "epoch": 0.2130394857667585, "grad_norm": 2.1266703605651855, "learning_rate": 1.423312883435583e-05, "loss": 0.031, "step": 232 }, { "epoch": 0.21395775941230485, "grad_norm": 1.7848790884017944, "learning_rate": 1.4294478527607363e-05, "loss": 0.0505, "step": 233 }, { "epoch": 0.21487603305785125, "grad_norm": 4.735909461975098, "learning_rate": 1.4355828220858897e-05, "loss": 0.1547, "step": 234 }, { "epoch": 0.2157943067033976, "grad_norm": 1.6043524742126465, "learning_rate": 1.441717791411043e-05, "loss": 0.201, "step": 235 }, { "epoch": 0.21671258034894397, "grad_norm": 0.23088207840919495, "learning_rate": 1.4478527607361965e-05, "loss": 0.0035, "step": 236 }, { "epoch": 0.21763085399449036, "grad_norm": 1.9821141958236694, "learning_rate": 1.4539877300613499e-05, "loss": 0.1945, "step": 237 }, { "epoch": 0.21854912764003673, "grad_norm": 3.4085285663604736, "learning_rate": 1.4601226993865032e-05, "loss": 0.2345, "step": 238 }, { "epoch": 0.2194674012855831, "grad_norm": 1.768754243850708, "learning_rate": 1.4662576687116566e-05, "loss": 0.0691, "step": 239 }, { "epoch": 0.22038567493112948, "grad_norm": 0.9548588395118713, "learning_rate": 1.47239263803681e-05, "loss": 0.1884, "step": 240 }, { "epoch": 0.22130394857667585, "grad_norm": 1.538268804550171, "learning_rate": 1.4785276073619631e-05, "loss": 0.1277, "step": 241 }, { "epoch": 0.2222222222222222, "grad_norm": 0.920171856880188, "learning_rate": 1.4846625766871168e-05, "loss": 0.0303, "step": 242 }, { "epoch": 0.2231404958677686, "grad_norm": 2.9227821826934814, "learning_rate": 1.4907975460122702e-05, "loss": 0.2024, "step": 243 }, { "epoch": 0.22405876951331496, "grad_norm": 0.9292974472045898, "learning_rate": 1.4969325153374235e-05, "loss": 0.0199, "step": 244 }, { "epoch": 0.22497704315886133, "grad_norm": 0.35771840810775757, "learning_rate": 1.5030674846625767e-05, "loss": 0.0139, "step": 245 }, { "epoch": 0.22589531680440772, "grad_norm": 2.507694721221924, "learning_rate": 1.50920245398773e-05, "loss": 0.2227, "step": 246 }, { "epoch": 0.22681359044995408, "grad_norm": 0.28483450412750244, "learning_rate": 1.5153374233128834e-05, "loss": 0.005, "step": 247 }, { "epoch": 0.22773186409550045, "grad_norm": 4.1794257164001465, "learning_rate": 1.5214723926380371e-05, "loss": 0.0691, "step": 248 }, { "epoch": 0.22865013774104684, "grad_norm": 4.1826171875, "learning_rate": 1.5276073619631903e-05, "loss": 0.0561, "step": 249 }, { "epoch": 0.2295684113865932, "grad_norm": 0.5465260744094849, "learning_rate": 1.5337423312883436e-05, "loss": 0.0208, "step": 250 }, { "epoch": 0.23048668503213957, "grad_norm": 1.328572154045105, "learning_rate": 1.539877300613497e-05, "loss": 0.0539, "step": 251 }, { "epoch": 0.23140495867768596, "grad_norm": 1.896087884902954, "learning_rate": 1.5460122699386504e-05, "loss": 0.2329, "step": 252 }, { "epoch": 0.23232323232323232, "grad_norm": 2.318634271621704, "learning_rate": 1.5521472392638037e-05, "loss": 0.1004, "step": 253 }, { "epoch": 0.23324150596877868, "grad_norm": 0.877866804599762, "learning_rate": 1.5582822085889574e-05, "loss": 0.0201, "step": 254 }, { "epoch": 0.23415977961432508, "grad_norm": 0.730012059211731, "learning_rate": 1.5644171779141108e-05, "loss": 0.0285, "step": 255 }, { "epoch": 0.23507805325987144, "grad_norm": 1.8603830337524414, "learning_rate": 1.570552147239264e-05, "loss": 0.0923, "step": 256 }, { "epoch": 0.2359963269054178, "grad_norm": 0.8288235068321228, "learning_rate": 1.5766871165644175e-05, "loss": 0.0342, "step": 257 }, { "epoch": 0.2369146005509642, "grad_norm": 1.471210241317749, "learning_rate": 1.5828220858895708e-05, "loss": 0.0992, "step": 258 }, { "epoch": 0.23783287419651056, "grad_norm": 1.2569507360458374, "learning_rate": 1.5889570552147238e-05, "loss": 0.1061, "step": 259 }, { "epoch": 0.23875114784205692, "grad_norm": 1.174983263015747, "learning_rate": 1.5950920245398772e-05, "loss": 0.1038, "step": 260 }, { "epoch": 0.2396694214876033, "grad_norm": 0.3272751569747925, "learning_rate": 1.601226993865031e-05, "loss": 0.0108, "step": 261 }, { "epoch": 0.24058769513314968, "grad_norm": 1.4782977104187012, "learning_rate": 1.6073619631901842e-05, "loss": 0.0623, "step": 262 }, { "epoch": 0.24150596877869604, "grad_norm": 5.842593193054199, "learning_rate": 1.6134969325153376e-05, "loss": 0.4375, "step": 263 }, { "epoch": 0.24242424242424243, "grad_norm": 1.7520140409469604, "learning_rate": 1.619631901840491e-05, "loss": 0.1028, "step": 264 }, { "epoch": 0.2433425160697888, "grad_norm": 3.1242735385894775, "learning_rate": 1.6257668711656443e-05, "loss": 0.1606, "step": 265 }, { "epoch": 0.24426078971533516, "grad_norm": 0.8759829998016357, "learning_rate": 1.6319018404907976e-05, "loss": 0.0924, "step": 266 }, { "epoch": 0.24517906336088155, "grad_norm": 1.8450413942337036, "learning_rate": 1.638036809815951e-05, "loss": 0.1176, "step": 267 }, { "epoch": 0.2460973370064279, "grad_norm": 1.9202346801757812, "learning_rate": 1.6441717791411043e-05, "loss": 0.0429, "step": 268 }, { "epoch": 0.24701561065197428, "grad_norm": 2.1405327320098877, "learning_rate": 1.6503067484662577e-05, "loss": 0.1768, "step": 269 }, { "epoch": 0.24793388429752067, "grad_norm": 1.392878532409668, "learning_rate": 1.656441717791411e-05, "loss": 0.0766, "step": 270 }, { "epoch": 0.24885215794306703, "grad_norm": 2.2452633380889893, "learning_rate": 1.6625766871165644e-05, "loss": 0.0833, "step": 271 }, { "epoch": 0.2497704315886134, "grad_norm": 1.400245189666748, "learning_rate": 1.6687116564417178e-05, "loss": 0.1745, "step": 272 }, { "epoch": 0.25068870523415976, "grad_norm": 2.8018417358398438, "learning_rate": 1.6748466257668714e-05, "loss": 0.0665, "step": 273 }, { "epoch": 0.2516069788797062, "grad_norm": 1.14453125, "learning_rate": 1.6809815950920248e-05, "loss": 0.0277, "step": 274 }, { "epoch": 0.25252525252525254, "grad_norm": 1.1324659585952759, "learning_rate": 1.687116564417178e-05, "loss": 0.1981, "step": 275 }, { "epoch": 0.2534435261707989, "grad_norm": 1.0482170581817627, "learning_rate": 1.6932515337423315e-05, "loss": 0.1148, "step": 276 }, { "epoch": 0.25436179981634527, "grad_norm": 1.8526408672332764, "learning_rate": 1.699386503067485e-05, "loss": 0.0993, "step": 277 }, { "epoch": 0.25528007346189163, "grad_norm": 1.07634437084198, "learning_rate": 1.7055214723926382e-05, "loss": 0.031, "step": 278 }, { "epoch": 0.256198347107438, "grad_norm": 0.38415610790252686, "learning_rate": 1.7116564417177916e-05, "loss": 0.0076, "step": 279 }, { "epoch": 0.2571166207529844, "grad_norm": 1.2956031560897827, "learning_rate": 1.717791411042945e-05, "loss": 0.1355, "step": 280 }, { "epoch": 0.2580348943985308, "grad_norm": 1.6450817584991455, "learning_rate": 1.7239263803680983e-05, "loss": 0.1052, "step": 281 }, { "epoch": 0.25895316804407714, "grad_norm": 1.5900222063064575, "learning_rate": 1.7300613496932516e-05, "loss": 0.1067, "step": 282 }, { "epoch": 0.2598714416896235, "grad_norm": 3.2554874420166016, "learning_rate": 1.736196319018405e-05, "loss": 0.0847, "step": 283 }, { "epoch": 0.26078971533516987, "grad_norm": 1.7889176607131958, "learning_rate": 1.7423312883435583e-05, "loss": 0.0772, "step": 284 }, { "epoch": 0.26170798898071623, "grad_norm": 1.1897656917572021, "learning_rate": 1.7484662576687117e-05, "loss": 0.0914, "step": 285 }, { "epoch": 0.26262626262626265, "grad_norm": 1.3643805980682373, "learning_rate": 1.7546012269938654e-05, "loss": 0.1311, "step": 286 }, { "epoch": 0.263544536271809, "grad_norm": 1.1704901456832886, "learning_rate": 1.7607361963190187e-05, "loss": 0.1595, "step": 287 }, { "epoch": 0.2644628099173554, "grad_norm": 2.5353832244873047, "learning_rate": 1.766871165644172e-05, "loss": 0.1533, "step": 288 }, { "epoch": 0.26538108356290174, "grad_norm": 0.947003185749054, "learning_rate": 1.7730061349693254e-05, "loss": 0.0194, "step": 289 }, { "epoch": 0.2662993572084481, "grad_norm": 0.5042552351951599, "learning_rate": 1.7791411042944788e-05, "loss": 0.0155, "step": 290 }, { "epoch": 0.26721763085399447, "grad_norm": 1.331705093383789, "learning_rate": 1.7852760736196318e-05, "loss": 0.0696, "step": 291 }, { "epoch": 0.2681359044995409, "grad_norm": 0.9892377257347107, "learning_rate": 1.7914110429447855e-05, "loss": 0.0622, "step": 292 }, { "epoch": 0.26905417814508725, "grad_norm": 1.3816404342651367, "learning_rate": 1.797546012269939e-05, "loss": 0.26, "step": 293 }, { "epoch": 0.2699724517906336, "grad_norm": 2.210120916366577, "learning_rate": 1.8036809815950922e-05, "loss": 0.2242, "step": 294 }, { "epoch": 0.27089072543618, "grad_norm": 1.5082335472106934, "learning_rate": 1.8098159509202455e-05, "loss": 0.1537, "step": 295 }, { "epoch": 0.27180899908172634, "grad_norm": 1.9463592767715454, "learning_rate": 1.815950920245399e-05, "loss": 0.185, "step": 296 }, { "epoch": 0.2727272727272727, "grad_norm": 1.9531348943710327, "learning_rate": 1.8220858895705523e-05, "loss": 0.3141, "step": 297 }, { "epoch": 0.2736455463728191, "grad_norm": 1.5043789148330688, "learning_rate": 1.828220858895706e-05, "loss": 0.1085, "step": 298 }, { "epoch": 0.2745638200183655, "grad_norm": 2.1161186695098877, "learning_rate": 1.834355828220859e-05, "loss": 0.1315, "step": 299 }, { "epoch": 0.27548209366391185, "grad_norm": 3.104872941970825, "learning_rate": 1.8404907975460123e-05, "loss": 0.1145, "step": 300 }, { "epoch": 0.2764003673094582, "grad_norm": 2.9949631690979004, "learning_rate": 1.8466257668711657e-05, "loss": 0.3024, "step": 301 }, { "epoch": 0.2773186409550046, "grad_norm": 4.017205238342285, "learning_rate": 1.852760736196319e-05, "loss": 0.0921, "step": 302 }, { "epoch": 0.27823691460055094, "grad_norm": 1.249814748764038, "learning_rate": 1.8588957055214724e-05, "loss": 0.1119, "step": 303 }, { "epoch": 0.27915518824609736, "grad_norm": 3.9489002227783203, "learning_rate": 1.8650306748466257e-05, "loss": 0.3915, "step": 304 }, { "epoch": 0.2800734618916437, "grad_norm": 2.1979916095733643, "learning_rate": 1.8711656441717794e-05, "loss": 0.1127, "step": 305 }, { "epoch": 0.2809917355371901, "grad_norm": 6.637365818023682, "learning_rate": 1.8773006134969328e-05, "loss": 0.0274, "step": 306 }, { "epoch": 0.28191000918273645, "grad_norm": 1.9082177877426147, "learning_rate": 1.883435582822086e-05, "loss": 0.0524, "step": 307 }, { "epoch": 0.2828282828282828, "grad_norm": 1.0823543071746826, "learning_rate": 1.8895705521472395e-05, "loss": 0.0509, "step": 308 }, { "epoch": 0.2837465564738292, "grad_norm": 1.2804851531982422, "learning_rate": 1.8957055214723928e-05, "loss": 0.1314, "step": 309 }, { "epoch": 0.2846648301193756, "grad_norm": 3.347489356994629, "learning_rate": 1.9018404907975462e-05, "loss": 0.6058, "step": 310 }, { "epoch": 0.28558310376492196, "grad_norm": 2.011793375015259, "learning_rate": 1.9079754601226995e-05, "loss": 0.0872, "step": 311 }, { "epoch": 0.2865013774104683, "grad_norm": 0.974129319190979, "learning_rate": 1.914110429447853e-05, "loss": 0.0354, "step": 312 }, { "epoch": 0.2874196510560147, "grad_norm": 0.6963267922401428, "learning_rate": 1.9202453987730062e-05, "loss": 0.0215, "step": 313 }, { "epoch": 0.28833792470156105, "grad_norm": 0.5707560181617737, "learning_rate": 1.9263803680981596e-05, "loss": 0.0151, "step": 314 }, { "epoch": 0.2892561983471074, "grad_norm": 0.6265159845352173, "learning_rate": 1.932515337423313e-05, "loss": 0.0155, "step": 315 }, { "epoch": 0.29017447199265384, "grad_norm": 1.488086462020874, "learning_rate": 1.9386503067484663e-05, "loss": 0.0196, "step": 316 }, { "epoch": 0.2910927456382002, "grad_norm": 1.577979564666748, "learning_rate": 1.94478527607362e-05, "loss": 0.0309, "step": 317 }, { "epoch": 0.29201101928374656, "grad_norm": 1.94890558719635, "learning_rate": 1.9509202453987733e-05, "loss": 0.2715, "step": 318 }, { "epoch": 0.29292929292929293, "grad_norm": 1.2272343635559082, "learning_rate": 1.9570552147239267e-05, "loss": 0.0488, "step": 319 }, { "epoch": 0.2938475665748393, "grad_norm": 1.7968394756317139, "learning_rate": 1.96319018404908e-05, "loss": 0.1944, "step": 320 }, { "epoch": 0.29476584022038566, "grad_norm": 2.5640666484832764, "learning_rate": 1.9693251533742334e-05, "loss": 0.0292, "step": 321 }, { "epoch": 0.2956841138659321, "grad_norm": 6.013891220092773, "learning_rate": 1.9754601226993868e-05, "loss": 0.1842, "step": 322 }, { "epoch": 0.29660238751147844, "grad_norm": 2.0934150218963623, "learning_rate": 1.98159509202454e-05, "loss": 0.4021, "step": 323 }, { "epoch": 0.2975206611570248, "grad_norm": 1.732842206954956, "learning_rate": 1.9877300613496935e-05, "loss": 0.116, "step": 324 }, { "epoch": 0.29843893480257117, "grad_norm": 0.4566493034362793, "learning_rate": 1.9938650306748468e-05, "loss": 0.0109, "step": 325 }, { "epoch": 0.29935720844811753, "grad_norm": 0.8532238006591797, "learning_rate": 2e-05, "loss": 0.0959, "step": 326 }, { "epoch": 0.3002754820936639, "grad_norm": 2.632111072540283, "learning_rate": 1.999999429468458e-05, "loss": 0.3, "step": 327 }, { "epoch": 0.3011937557392103, "grad_norm": 0.12246864289045334, "learning_rate": 1.9999977178744835e-05, "loss": 0.0028, "step": 328 }, { "epoch": 0.3021120293847567, "grad_norm": 1.2515559196472168, "learning_rate": 1.9999948652200295e-05, "loss": 0.0615, "step": 329 }, { "epoch": 0.30303030303030304, "grad_norm": 1.2689533233642578, "learning_rate": 1.9999908715083503e-05, "loss": 0.123, "step": 330 }, { "epoch": 0.3039485766758494, "grad_norm": 0.29163703322410583, "learning_rate": 1.999985736744004e-05, "loss": 0.0082, "step": 331 }, { "epoch": 0.30486685032139577, "grad_norm": 1.325951337814331, "learning_rate": 1.999979460932849e-05, "loss": 0.0568, "step": 332 }, { "epoch": 0.30578512396694213, "grad_norm": 1.2607991695404053, "learning_rate": 1.9999720440820468e-05, "loss": 0.0314, "step": 333 }, { "epoch": 0.30670339761248855, "grad_norm": 1.9455457925796509, "learning_rate": 1.99996348620006e-05, "loss": 0.0845, "step": 334 }, { "epoch": 0.3076216712580349, "grad_norm": 0.8968974351882935, "learning_rate": 1.9999537872966543e-05, "loss": 0.0285, "step": 335 }, { "epoch": 0.3085399449035813, "grad_norm": 3.4470160007476807, "learning_rate": 1.9999429473828967e-05, "loss": 0.0896, "step": 336 }, { "epoch": 0.30945821854912764, "grad_norm": 2.761850595474243, "learning_rate": 1.9999309664711556e-05, "loss": 0.362, "step": 337 }, { "epoch": 0.310376492194674, "grad_norm": 1.980118989944458, "learning_rate": 1.9999178445751024e-05, "loss": 0.1256, "step": 338 }, { "epoch": 0.31129476584022037, "grad_norm": 1.5449047088623047, "learning_rate": 1.99990358170971e-05, "loss": 0.0412, "step": 339 }, { "epoch": 0.3122130394857668, "grad_norm": 1.7459899187088013, "learning_rate": 1.9998881778912532e-05, "loss": 0.0409, "step": 340 }, { "epoch": 0.31313131313131315, "grad_norm": 0.13255417346954346, "learning_rate": 1.9998716331373088e-05, "loss": 0.0031, "step": 341 }, { "epoch": 0.3140495867768595, "grad_norm": 2.596161365509033, "learning_rate": 1.9998539474667555e-05, "loss": 0.1176, "step": 342 }, { "epoch": 0.3149678604224059, "grad_norm": 5.408975124359131, "learning_rate": 1.9998351208997734e-05, "loss": 0.4208, "step": 343 }, { "epoch": 0.31588613406795224, "grad_norm": 1.1937506198883057, "learning_rate": 1.999815153457845e-05, "loss": 0.0442, "step": 344 }, { "epoch": 0.3168044077134986, "grad_norm": 3.565005302429199, "learning_rate": 1.9997940451637547e-05, "loss": 0.2507, "step": 345 }, { "epoch": 0.317722681359045, "grad_norm": 1.3836617469787598, "learning_rate": 1.999771796041588e-05, "loss": 0.1641, "step": 346 }, { "epoch": 0.3186409550045914, "grad_norm": 1.587546944618225, "learning_rate": 1.9997484061167325e-05, "loss": 0.1777, "step": 347 }, { "epoch": 0.31955922865013775, "grad_norm": 2.6962673664093018, "learning_rate": 1.999723875415878e-05, "loss": 0.1226, "step": 348 }, { "epoch": 0.3204775022956841, "grad_norm": 2.5454933643341064, "learning_rate": 1.999698203967015e-05, "loss": 0.1801, "step": 349 }, { "epoch": 0.3213957759412305, "grad_norm": 3.6735851764678955, "learning_rate": 1.999671391799437e-05, "loss": 0.1829, "step": 350 }, { "epoch": 0.32231404958677684, "grad_norm": 2.3582255840301514, "learning_rate": 1.9996434389437377e-05, "loss": 0.248, "step": 351 }, { "epoch": 0.32323232323232326, "grad_norm": 1.2499151229858398, "learning_rate": 1.9996143454318133e-05, "loss": 0.0358, "step": 352 }, { "epoch": 0.3241505968778696, "grad_norm": 1.4912818670272827, "learning_rate": 1.9995841112968613e-05, "loss": 0.0962, "step": 353 }, { "epoch": 0.325068870523416, "grad_norm": 0.655762791633606, "learning_rate": 1.999552736573381e-05, "loss": 0.1011, "step": 354 }, { "epoch": 0.32598714416896235, "grad_norm": 1.160790205001831, "learning_rate": 1.9995202212971723e-05, "loss": 0.0699, "step": 355 }, { "epoch": 0.3269054178145087, "grad_norm": 5.155803680419922, "learning_rate": 1.9994865655053382e-05, "loss": 0.2121, "step": 356 }, { "epoch": 0.3278236914600551, "grad_norm": 1.5815479755401611, "learning_rate": 1.9994517692362813e-05, "loss": 0.1546, "step": 357 }, { "epoch": 0.3287419651056015, "grad_norm": 1.9940189123153687, "learning_rate": 1.9994158325297065e-05, "loss": 0.275, "step": 358 }, { "epoch": 0.32966023875114786, "grad_norm": 0.7288774847984314, "learning_rate": 1.9993787554266196e-05, "loss": 0.0173, "step": 359 }, { "epoch": 0.3305785123966942, "grad_norm": 1.6438087224960327, "learning_rate": 1.999340537969329e-05, "loss": 0.0727, "step": 360 }, { "epoch": 0.3314967860422406, "grad_norm": 2.6688013076782227, "learning_rate": 1.999301180201442e-05, "loss": 0.1452, "step": 361 }, { "epoch": 0.33241505968778695, "grad_norm": 1.0706363916397095, "learning_rate": 1.9992606821678685e-05, "loss": 0.0437, "step": 362 }, { "epoch": 0.3333333333333333, "grad_norm": 6.374241828918457, "learning_rate": 1.9992190439148196e-05, "loss": 0.0895, "step": 363 }, { "epoch": 0.3342516069788797, "grad_norm": 2.5880768299102783, "learning_rate": 1.9991762654898075e-05, "loss": 0.2164, "step": 364 }, { "epoch": 0.3351698806244261, "grad_norm": 0.7571276426315308, "learning_rate": 1.9991323469416443e-05, "loss": 0.0372, "step": 365 }, { "epoch": 0.33608815426997246, "grad_norm": 2.0681042671203613, "learning_rate": 1.9990872883204444e-05, "loss": 0.1563, "step": 366 }, { "epoch": 0.3370064279155188, "grad_norm": 1.7404201030731201, "learning_rate": 1.9990410896776226e-05, "loss": 0.1063, "step": 367 }, { "epoch": 0.3379247015610652, "grad_norm": 0.48120829463005066, "learning_rate": 1.998993751065894e-05, "loss": 0.0716, "step": 368 }, { "epoch": 0.33884297520661155, "grad_norm": 1.156941294670105, "learning_rate": 1.998945272539275e-05, "loss": 0.0812, "step": 369 }, { "epoch": 0.3397612488521579, "grad_norm": 1.4767205715179443, "learning_rate": 1.998895654153083e-05, "loss": 0.0836, "step": 370 }, { "epoch": 0.34067952249770433, "grad_norm": 1.5283399820327759, "learning_rate": 1.998844895963936e-05, "loss": 0.1932, "step": 371 }, { "epoch": 0.3415977961432507, "grad_norm": 0.5636906623840332, "learning_rate": 1.998792998029751e-05, "loss": 0.0871, "step": 372 }, { "epoch": 0.34251606978879706, "grad_norm": 0.08499777317047119, "learning_rate": 1.998739960409748e-05, "loss": 0.0032, "step": 373 }, { "epoch": 0.3434343434343434, "grad_norm": 1.9133425951004028, "learning_rate": 1.9986857831644458e-05, "loss": 0.0432, "step": 374 }, { "epoch": 0.3443526170798898, "grad_norm": 0.8006404638290405, "learning_rate": 1.9986304663556642e-05, "loss": 0.0317, "step": 375 }, { "epoch": 0.34527089072543615, "grad_norm": 0.8622770309448242, "learning_rate": 1.9985740100465233e-05, "loss": 0.0187, "step": 376 }, { "epoch": 0.34618916437098257, "grad_norm": 1.4154953956604004, "learning_rate": 1.9985164143014433e-05, "loss": 0.0868, "step": 377 }, { "epoch": 0.34710743801652894, "grad_norm": 2.094325065612793, "learning_rate": 1.9984576791861442e-05, "loss": 0.1311, "step": 378 }, { "epoch": 0.3480257116620753, "grad_norm": 1.0866682529449463, "learning_rate": 1.998397804767647e-05, "loss": 0.0444, "step": 379 }, { "epoch": 0.34894398530762166, "grad_norm": 0.8058256506919861, "learning_rate": 1.9983367911142715e-05, "loss": 0.0507, "step": 380 }, { "epoch": 0.349862258953168, "grad_norm": 1.2305113077163696, "learning_rate": 1.9982746382956386e-05, "loss": 0.1386, "step": 381 }, { "epoch": 0.3507805325987144, "grad_norm": 1.750862956047058, "learning_rate": 1.9982113463826685e-05, "loss": 0.167, "step": 382 }, { "epoch": 0.3516988062442608, "grad_norm": 3.7469239234924316, "learning_rate": 1.9981469154475818e-05, "loss": 0.1596, "step": 383 }, { "epoch": 0.3526170798898072, "grad_norm": 1.6888744831085205, "learning_rate": 1.998081345563897e-05, "loss": 0.1138, "step": 384 }, { "epoch": 0.35353535353535354, "grad_norm": 2.110262393951416, "learning_rate": 1.9980146368064345e-05, "loss": 0.2305, "step": 385 }, { "epoch": 0.3544536271808999, "grad_norm": 0.8024462461471558, "learning_rate": 1.9979467892513128e-05, "loss": 0.0363, "step": 386 }, { "epoch": 0.35537190082644626, "grad_norm": 1.3497532606124878, "learning_rate": 1.9978778029759506e-05, "loss": 0.0396, "step": 387 }, { "epoch": 0.3562901744719926, "grad_norm": 2.7912485599517822, "learning_rate": 1.997807678059065e-05, "loss": 0.2367, "step": 388 }, { "epoch": 0.35720844811753905, "grad_norm": 1.8525277376174927, "learning_rate": 1.9977364145806738e-05, "loss": 0.0471, "step": 389 }, { "epoch": 0.3581267217630854, "grad_norm": 1.8696492910385132, "learning_rate": 1.9976640126220924e-05, "loss": 0.149, "step": 390 }, { "epoch": 0.3590449954086318, "grad_norm": 2.0907773971557617, "learning_rate": 1.997590472265936e-05, "loss": 0.0346, "step": 391 }, { "epoch": 0.35996326905417814, "grad_norm": 2.186912775039673, "learning_rate": 1.9975157935961192e-05, "loss": 0.1548, "step": 392 }, { "epoch": 0.3608815426997245, "grad_norm": 1.9128203392028809, "learning_rate": 1.9974399766978546e-05, "loss": 0.1216, "step": 393 }, { "epoch": 0.36179981634527086, "grad_norm": 0.8682861328125, "learning_rate": 1.9973630216576547e-05, "loss": 0.0332, "step": 394 }, { "epoch": 0.3627180899908173, "grad_norm": 2.256568431854248, "learning_rate": 1.9972849285633296e-05, "loss": 0.0548, "step": 395 }, { "epoch": 0.36363636363636365, "grad_norm": 1.4579709768295288, "learning_rate": 1.9972056975039884e-05, "loss": 0.0872, "step": 396 }, { "epoch": 0.36455463728191, "grad_norm": 1.3660660982131958, "learning_rate": 1.997125328570039e-05, "loss": 0.111, "step": 397 }, { "epoch": 0.3654729109274564, "grad_norm": 1.78518807888031, "learning_rate": 1.9970438218531873e-05, "loss": 0.1508, "step": 398 }, { "epoch": 0.36639118457300274, "grad_norm": 0.1357259452342987, "learning_rate": 1.9969611774464378e-05, "loss": 0.0037, "step": 399 }, { "epoch": 0.3673094582185491, "grad_norm": 1.8301008939743042, "learning_rate": 1.9968773954440923e-05, "loss": 0.084, "step": 400 }, { "epoch": 0.3682277318640955, "grad_norm": 1.1644885540008545, "learning_rate": 1.9967924759417524e-05, "loss": 0.0534, "step": 401 }, { "epoch": 0.3691460055096419, "grad_norm": 1.2698159217834473, "learning_rate": 1.9967064190363156e-05, "loss": 0.0702, "step": 402 }, { "epoch": 0.37006427915518825, "grad_norm": 2.118833065032959, "learning_rate": 1.996619224825979e-05, "loss": 0.1683, "step": 403 }, { "epoch": 0.3709825528007346, "grad_norm": 1.298014760017395, "learning_rate": 1.9965308934102366e-05, "loss": 0.0215, "step": 404 }, { "epoch": 0.371900826446281, "grad_norm": 1.7084227800369263, "learning_rate": 1.9964414248898795e-05, "loss": 0.2878, "step": 405 }, { "epoch": 0.37281910009182734, "grad_norm": 2.0923380851745605, "learning_rate": 1.9963508193669975e-05, "loss": 0.0694, "step": 406 }, { "epoch": 0.37373737373737376, "grad_norm": 2.1242690086364746, "learning_rate": 1.996259076944977e-05, "loss": 0.1172, "step": 407 }, { "epoch": 0.3746556473829201, "grad_norm": 3.3362133502960205, "learning_rate": 1.996166197728502e-05, "loss": 0.0398, "step": 408 }, { "epoch": 0.3755739210284665, "grad_norm": 1.5325989723205566, "learning_rate": 1.9960721818235536e-05, "loss": 0.0649, "step": 409 }, { "epoch": 0.37649219467401285, "grad_norm": 2.177088975906372, "learning_rate": 1.9959770293374096e-05, "loss": 0.1718, "step": 410 }, { "epoch": 0.3774104683195592, "grad_norm": 1.2872308492660522, "learning_rate": 1.9958807403786452e-05, "loss": 0.0632, "step": 411 }, { "epoch": 0.3783287419651056, "grad_norm": 1.225250482559204, "learning_rate": 1.9957833150571325e-05, "loss": 0.1916, "step": 412 }, { "epoch": 0.379247015610652, "grad_norm": 1.358361005783081, "learning_rate": 1.995684753484039e-05, "loss": 0.0748, "step": 413 }, { "epoch": 0.38016528925619836, "grad_norm": 1.081457495689392, "learning_rate": 1.9955850557718307e-05, "loss": 0.0324, "step": 414 }, { "epoch": 0.3810835629017447, "grad_norm": 0.7105931639671326, "learning_rate": 1.9954842220342684e-05, "loss": 0.0197, "step": 415 }, { "epoch": 0.3820018365472911, "grad_norm": 0.33286502957344055, "learning_rate": 1.99538225238641e-05, "loss": 0.0085, "step": 416 }, { "epoch": 0.38292011019283745, "grad_norm": 1.854170322418213, "learning_rate": 1.995279146944609e-05, "loss": 0.1624, "step": 417 }, { "epoch": 0.3838383838383838, "grad_norm": 1.8389472961425781, "learning_rate": 1.995174905826516e-05, "loss": 0.0912, "step": 418 }, { "epoch": 0.38475665748393023, "grad_norm": 1.4987330436706543, "learning_rate": 1.995069529151075e-05, "loss": 0.0919, "step": 419 }, { "epoch": 0.3856749311294766, "grad_norm": 1.5254359245300293, "learning_rate": 1.9949630170385294e-05, "loss": 0.1855, "step": 420 }, { "epoch": 0.38659320477502296, "grad_norm": 1.1976407766342163, "learning_rate": 1.994855369610415e-05, "loss": 0.0588, "step": 421 }, { "epoch": 0.3875114784205693, "grad_norm": 1.6197240352630615, "learning_rate": 1.9947465869895646e-05, "loss": 0.1256, "step": 422 }, { "epoch": 0.3884297520661157, "grad_norm": 1.3569403886795044, "learning_rate": 1.9946366693001066e-05, "loss": 0.134, "step": 423 }, { "epoch": 0.38934802571166205, "grad_norm": 0.6141202449798584, "learning_rate": 1.994525616667463e-05, "loss": 0.0614, "step": 424 }, { "epoch": 0.39026629935720847, "grad_norm": 1.0418928861618042, "learning_rate": 1.9944134292183528e-05, "loss": 0.0395, "step": 425 }, { "epoch": 0.39118457300275483, "grad_norm": 2.168886184692383, "learning_rate": 1.9943001070807883e-05, "loss": 0.1033, "step": 426 }, { "epoch": 0.3921028466483012, "grad_norm": 0.10323487222194672, "learning_rate": 1.9941856503840777e-05, "loss": 0.0037, "step": 427 }, { "epoch": 0.39302112029384756, "grad_norm": 2.3121557235717773, "learning_rate": 1.9940700592588228e-05, "loss": 0.1178, "step": 428 }, { "epoch": 0.3939393939393939, "grad_norm": 1.4852752685546875, "learning_rate": 1.9939533338369206e-05, "loss": 0.0426, "step": 429 }, { "epoch": 0.3948576675849403, "grad_norm": 1.9060066938400269, "learning_rate": 1.9938354742515624e-05, "loss": 0.2793, "step": 430 }, { "epoch": 0.3957759412304867, "grad_norm": 0.37687253952026367, "learning_rate": 1.9937164806372328e-05, "loss": 0.0088, "step": 431 }, { "epoch": 0.39669421487603307, "grad_norm": 2.4034130573272705, "learning_rate": 1.993596353129712e-05, "loss": 0.1828, "step": 432 }, { "epoch": 0.39761248852157943, "grad_norm": 1.3113515377044678, "learning_rate": 1.9934750918660724e-05, "loss": 0.0674, "step": 433 }, { "epoch": 0.3985307621671258, "grad_norm": 0.06838607043027878, "learning_rate": 1.9933526969846805e-05, "loss": 0.0023, "step": 434 }, { "epoch": 0.39944903581267216, "grad_norm": 2.140943765640259, "learning_rate": 1.9932291686251973e-05, "loss": 0.1005, "step": 435 }, { "epoch": 0.4003673094582185, "grad_norm": 2.0392792224884033, "learning_rate": 1.993104506928576e-05, "loss": 0.1726, "step": 436 }, { "epoch": 0.40128558310376494, "grad_norm": 1.6239908933639526, "learning_rate": 1.9929787120370634e-05, "loss": 0.0946, "step": 437 }, { "epoch": 0.4022038567493113, "grad_norm": 1.239485263824463, "learning_rate": 1.9928517840941998e-05, "loss": 0.0227, "step": 438 }, { "epoch": 0.40312213039485767, "grad_norm": 0.116143599152565, "learning_rate": 1.9927237232448174e-05, "loss": 0.0034, "step": 439 }, { "epoch": 0.40404040404040403, "grad_norm": 1.6612412929534912, "learning_rate": 1.9925945296350423e-05, "loss": 0.1059, "step": 440 }, { "epoch": 0.4049586776859504, "grad_norm": 1.667924404144287, "learning_rate": 1.9924642034122923e-05, "loss": 0.043, "step": 441 }, { "epoch": 0.40587695133149676, "grad_norm": 1.193485975265503, "learning_rate": 1.992332744725278e-05, "loss": 0.0392, "step": 442 }, { "epoch": 0.4067952249770432, "grad_norm": 0.4670940339565277, "learning_rate": 1.9922001537240018e-05, "loss": 0.0059, "step": 443 }, { "epoch": 0.40771349862258954, "grad_norm": 1.4763669967651367, "learning_rate": 1.992066430559758e-05, "loss": 0.0733, "step": 444 }, { "epoch": 0.4086317722681359, "grad_norm": 0.1560511738061905, "learning_rate": 1.9919315753851343e-05, "loss": 0.0043, "step": 445 }, { "epoch": 0.40955004591368227, "grad_norm": 1.974105715751648, "learning_rate": 1.9917955883540082e-05, "loss": 0.2384, "step": 446 }, { "epoch": 0.41046831955922863, "grad_norm": 2.9267778396606445, "learning_rate": 1.9916584696215497e-05, "loss": 0.2992, "step": 447 }, { "epoch": 0.411386593204775, "grad_norm": 1.1731144189834595, "learning_rate": 1.9915202193442194e-05, "loss": 0.0472, "step": 448 }, { "epoch": 0.4123048668503214, "grad_norm": 2.0003442764282227, "learning_rate": 1.9913808376797705e-05, "loss": 0.3291, "step": 449 }, { "epoch": 0.4132231404958678, "grad_norm": 2.258904457092285, "learning_rate": 1.9912403247872456e-05, "loss": 0.1409, "step": 450 }, { "epoch": 0.41414141414141414, "grad_norm": 0.871258556842804, "learning_rate": 1.991098680826979e-05, "loss": 0.0195, "step": 451 }, { "epoch": 0.4150596877869605, "grad_norm": 1.3280644416809082, "learning_rate": 1.990955905960595e-05, "loss": 0.0639, "step": 452 }, { "epoch": 0.41597796143250687, "grad_norm": 0.6393201351165771, "learning_rate": 1.9908120003510096e-05, "loss": 0.0166, "step": 453 }, { "epoch": 0.41689623507805323, "grad_norm": 0.5868614912033081, "learning_rate": 1.9906669641624275e-05, "loss": 0.015, "step": 454 }, { "epoch": 0.41781450872359965, "grad_norm": 1.1606813669204712, "learning_rate": 1.9905207975603444e-05, "loss": 0.0527, "step": 455 }, { "epoch": 0.418732782369146, "grad_norm": 0.949892520904541, "learning_rate": 1.9903735007115456e-05, "loss": 0.0808, "step": 456 }, { "epoch": 0.4196510560146924, "grad_norm": 0.17449747025966644, "learning_rate": 1.9902250737841057e-05, "loss": 0.005, "step": 457 }, { "epoch": 0.42056932966023874, "grad_norm": 3.039234161376953, "learning_rate": 1.99007551694739e-05, "loss": 0.3965, "step": 458 }, { "epoch": 0.4214876033057851, "grad_norm": 1.2705671787261963, "learning_rate": 1.9899248303720515e-05, "loss": 0.1558, "step": 459 }, { "epoch": 0.42240587695133147, "grad_norm": 0.4812701344490051, "learning_rate": 1.9897730142300332e-05, "loss": 0.019, "step": 460 }, { "epoch": 0.4233241505968779, "grad_norm": 1.5014787912368774, "learning_rate": 1.9896200686945673e-05, "loss": 0.0805, "step": 461 }, { "epoch": 0.42424242424242425, "grad_norm": 1.3016610145568848, "learning_rate": 1.989465993940174e-05, "loss": 0.0844, "step": 462 }, { "epoch": 0.4251606978879706, "grad_norm": 0.71735680103302, "learning_rate": 1.9893107901426625e-05, "loss": 0.0365, "step": 463 }, { "epoch": 0.426078971533517, "grad_norm": 2.3299524784088135, "learning_rate": 1.98915445747913e-05, "loss": 0.2935, "step": 464 }, { "epoch": 0.42699724517906334, "grad_norm": 0.2388847917318344, "learning_rate": 1.988996996127962e-05, "loss": 0.0057, "step": 465 }, { "epoch": 0.4279155188246097, "grad_norm": 0.49840596318244934, "learning_rate": 1.9888384062688316e-05, "loss": 0.0114, "step": 466 }, { "epoch": 0.4288337924701561, "grad_norm": 0.5390764474868774, "learning_rate": 1.9886786880827e-05, "loss": 0.0208, "step": 467 }, { "epoch": 0.4297520661157025, "grad_norm": 1.7297059297561646, "learning_rate": 1.988517841751816e-05, "loss": 0.143, "step": 468 }, { "epoch": 0.43067033976124885, "grad_norm": 1.6472265720367432, "learning_rate": 1.9883558674597153e-05, "loss": 0.0604, "step": 469 }, { "epoch": 0.4315886134067952, "grad_norm": 1.2907642126083374, "learning_rate": 1.98819276539122e-05, "loss": 0.2078, "step": 470 }, { "epoch": 0.4325068870523416, "grad_norm": 0.4959701895713806, "learning_rate": 1.9880285357324412e-05, "loss": 0.017, "step": 471 }, { "epoch": 0.43342516069788795, "grad_norm": 1.8797651529312134, "learning_rate": 1.9878631786707743e-05, "loss": 0.1962, "step": 472 }, { "epoch": 0.43434343434343436, "grad_norm": 1.4257732629776, "learning_rate": 1.9876966943949024e-05, "loss": 0.1472, "step": 473 }, { "epoch": 0.43526170798898073, "grad_norm": 0.14491939544677734, "learning_rate": 1.9875290830947946e-05, "loss": 0.0052, "step": 474 }, { "epoch": 0.4361799816345271, "grad_norm": 1.2971880435943604, "learning_rate": 1.987360344961706e-05, "loss": 0.1228, "step": 475 }, { "epoch": 0.43709825528007346, "grad_norm": 1.7007063627243042, "learning_rate": 1.9871904801881774e-05, "loss": 0.1033, "step": 476 }, { "epoch": 0.4380165289256198, "grad_norm": 1.0301835536956787, "learning_rate": 1.9870194889680353e-05, "loss": 0.0973, "step": 477 }, { "epoch": 0.4389348025711662, "grad_norm": 1.5715056657791138, "learning_rate": 1.9868473714963913e-05, "loss": 0.1671, "step": 478 }, { "epoch": 0.4398530762167126, "grad_norm": 1.6641765832901, "learning_rate": 1.986674127969642e-05, "loss": 0.1523, "step": 479 }, { "epoch": 0.44077134986225897, "grad_norm": 0.2920869290828705, "learning_rate": 1.9864997585854704e-05, "loss": 0.0089, "step": 480 }, { "epoch": 0.44168962350780533, "grad_norm": 2.2901127338409424, "learning_rate": 1.9863242635428418e-05, "loss": 0.3528, "step": 481 }, { "epoch": 0.4426078971533517, "grad_norm": 1.2778034210205078, "learning_rate": 1.9861476430420074e-05, "loss": 0.0517, "step": 482 }, { "epoch": 0.44352617079889806, "grad_norm": 0.9196149110794067, "learning_rate": 1.9859698972845024e-05, "loss": 0.0587, "step": 483 }, { "epoch": 0.4444444444444444, "grad_norm": 1.762027382850647, "learning_rate": 1.985791026473146e-05, "loss": 0.054, "step": 484 }, { "epoch": 0.44536271808999084, "grad_norm": 0.9298413991928101, "learning_rate": 1.985611030812041e-05, "loss": 0.0301, "step": 485 }, { "epoch": 0.4462809917355372, "grad_norm": 0.7052368521690369, "learning_rate": 1.9854299105065738e-05, "loss": 0.026, "step": 486 }, { "epoch": 0.44719926538108357, "grad_norm": 5.175395965576172, "learning_rate": 1.9852476657634142e-05, "loss": 0.4786, "step": 487 }, { "epoch": 0.44811753902662993, "grad_norm": 2.867482900619507, "learning_rate": 1.985064296790515e-05, "loss": 0.3147, "step": 488 }, { "epoch": 0.4490358126721763, "grad_norm": 1.254047155380249, "learning_rate": 1.984879803797111e-05, "loss": 0.034, "step": 489 }, { "epoch": 0.44995408631772266, "grad_norm": 0.3437318205833435, "learning_rate": 1.9846941869937216e-05, "loss": 0.0091, "step": 490 }, { "epoch": 0.4508723599632691, "grad_norm": 0.8414430618286133, "learning_rate": 1.9845074465921463e-05, "loss": 0.0506, "step": 491 }, { "epoch": 0.45179063360881544, "grad_norm": 0.056748196482658386, "learning_rate": 1.9843195828054682e-05, "loss": 0.0016, "step": 492 }, { "epoch": 0.4527089072543618, "grad_norm": 1.6541780233383179, "learning_rate": 1.9841305958480517e-05, "loss": 0.0862, "step": 493 }, { "epoch": 0.45362718089990817, "grad_norm": 1.4461182355880737, "learning_rate": 1.9839404859355426e-05, "loss": 0.0644, "step": 494 }, { "epoch": 0.45454545454545453, "grad_norm": 2.4127674102783203, "learning_rate": 1.9837492532848684e-05, "loss": 0.2114, "step": 495 }, { "epoch": 0.4554637281910009, "grad_norm": 1.4115185737609863, "learning_rate": 1.9835568981142376e-05, "loss": 0.1046, "step": 496 }, { "epoch": 0.4563820018365473, "grad_norm": 0.13092227280139923, "learning_rate": 1.9833634206431395e-05, "loss": 0.0042, "step": 497 }, { "epoch": 0.4573002754820937, "grad_norm": 0.2289186418056488, "learning_rate": 1.983168821092344e-05, "loss": 0.0068, "step": 498 }, { "epoch": 0.45821854912764004, "grad_norm": 2.153229236602783, "learning_rate": 1.982973099683902e-05, "loss": 0.1156, "step": 499 }, { "epoch": 0.4591368227731864, "grad_norm": 0.8030224442481995, "learning_rate": 1.9827762566411436e-05, "loss": 0.0193, "step": 500 }, { "epoch": 0.46005509641873277, "grad_norm": 3.0752031803131104, "learning_rate": 1.982578292188679e-05, "loss": 0.1276, "step": 501 }, { "epoch": 0.46097337006427913, "grad_norm": 4.997933864593506, "learning_rate": 1.9823792065523985e-05, "loss": 0.2338, "step": 502 }, { "epoch": 0.46189164370982555, "grad_norm": 1.4143433570861816, "learning_rate": 1.982178999959471e-05, "loss": 0.1355, "step": 503 }, { "epoch": 0.4628099173553719, "grad_norm": 0.23707588016986847, "learning_rate": 1.9819776726383452e-05, "loss": 0.0065, "step": 504 }, { "epoch": 0.4637281910009183, "grad_norm": 1.4886257648468018, "learning_rate": 1.981775224818748e-05, "loss": 0.0827, "step": 505 }, { "epoch": 0.46464646464646464, "grad_norm": 0.4402680993080139, "learning_rate": 1.9815716567316855e-05, "loss": 0.0122, "step": 506 }, { "epoch": 0.465564738292011, "grad_norm": 1.7015588283538818, "learning_rate": 1.981366968609441e-05, "loss": 0.2791, "step": 507 }, { "epoch": 0.46648301193755737, "grad_norm": 0.15042413771152496, "learning_rate": 1.981161160685577e-05, "loss": 0.005, "step": 508 }, { "epoch": 0.4674012855831038, "grad_norm": 0.36762022972106934, "learning_rate": 1.9809542331949338e-05, "loss": 0.0077, "step": 509 }, { "epoch": 0.46831955922865015, "grad_norm": 0.2697368264198303, "learning_rate": 1.980746186373628e-05, "loss": 0.0091, "step": 510 }, { "epoch": 0.4692378328741965, "grad_norm": 2.2874295711517334, "learning_rate": 1.9805370204590545e-05, "loss": 0.0982, "step": 511 }, { "epoch": 0.4701561065197429, "grad_norm": 0.9545071125030518, "learning_rate": 1.9803267356898844e-05, "loss": 0.0694, "step": 512 }, { "epoch": 0.47107438016528924, "grad_norm": 0.08303436636924744, "learning_rate": 1.9801153323060667e-05, "loss": 0.0026, "step": 513 }, { "epoch": 0.4719926538108356, "grad_norm": 0.3449619710445404, "learning_rate": 1.979902810548825e-05, "loss": 0.0133, "step": 514 }, { "epoch": 0.472910927456382, "grad_norm": 1.9265220165252686, "learning_rate": 1.979689170660661e-05, "loss": 0.0401, "step": 515 }, { "epoch": 0.4738292011019284, "grad_norm": 0.8582781553268433, "learning_rate": 1.9794744128853507e-05, "loss": 0.0414, "step": 516 }, { "epoch": 0.47474747474747475, "grad_norm": 1.8246479034423828, "learning_rate": 1.9792585374679464e-05, "loss": 0.1099, "step": 517 }, { "epoch": 0.4756657483930211, "grad_norm": 1.6095668077468872, "learning_rate": 1.9790415446547754e-05, "loss": 0.13, "step": 518 }, { "epoch": 0.4765840220385675, "grad_norm": 1.1178349256515503, "learning_rate": 1.9788234346934403e-05, "loss": 0.0519, "step": 519 }, { "epoch": 0.47750229568411384, "grad_norm": 2.026629686355591, "learning_rate": 1.9786042078328184e-05, "loss": 0.3092, "step": 520 }, { "epoch": 0.47842056932966026, "grad_norm": 1.3625456094741821, "learning_rate": 1.9783838643230616e-05, "loss": 0.0253, "step": 521 }, { "epoch": 0.4793388429752066, "grad_norm": 0.15079079568386078, "learning_rate": 1.9781624044155952e-05, "loss": 0.0039, "step": 522 }, { "epoch": 0.480257116620753, "grad_norm": 0.14605781435966492, "learning_rate": 1.9779398283631195e-05, "loss": 0.0028, "step": 523 }, { "epoch": 0.48117539026629935, "grad_norm": 1.1828484535217285, "learning_rate": 1.9777161364196074e-05, "loss": 0.1174, "step": 524 }, { "epoch": 0.4820936639118457, "grad_norm": 0.0851387158036232, "learning_rate": 1.9774913288403055e-05, "loss": 0.003, "step": 525 }, { "epoch": 0.4830119375573921, "grad_norm": 0.04864233732223511, "learning_rate": 1.9772654058817336e-05, "loss": 0.0019, "step": 526 }, { "epoch": 0.4839302112029385, "grad_norm": 1.0542397499084473, "learning_rate": 1.9770383678016843e-05, "loss": 0.0538, "step": 527 }, { "epoch": 0.48484848484848486, "grad_norm": 2.3739333152770996, "learning_rate": 1.9768102148592222e-05, "loss": 0.12, "step": 528 }, { "epoch": 0.4857667584940312, "grad_norm": 1.2538129091262817, "learning_rate": 1.976580947314684e-05, "loss": 0.1499, "step": 529 }, { "epoch": 0.4866850321395776, "grad_norm": 0.06536607444286346, "learning_rate": 1.9763505654296782e-05, "loss": 0.0026, "step": 530 }, { "epoch": 0.48760330578512395, "grad_norm": 0.09087325632572174, "learning_rate": 1.9761190694670855e-05, "loss": 0.0028, "step": 531 }, { "epoch": 0.4885215794306703, "grad_norm": 0.9555211663246155, "learning_rate": 1.9758864596910575e-05, "loss": 0.0504, "step": 532 }, { "epoch": 0.48943985307621674, "grad_norm": 0.9447014331817627, "learning_rate": 1.975652736367016e-05, "loss": 0.0216, "step": 533 }, { "epoch": 0.4903581267217631, "grad_norm": 3.1546592712402344, "learning_rate": 1.975417899761655e-05, "loss": 0.1973, "step": 534 }, { "epoch": 0.49127640036730946, "grad_norm": 1.6641873121261597, "learning_rate": 1.9751819501429374e-05, "loss": 0.093, "step": 535 }, { "epoch": 0.4921946740128558, "grad_norm": 0.7381713390350342, "learning_rate": 1.9749448877800964e-05, "loss": 0.0288, "step": 536 }, { "epoch": 0.4931129476584022, "grad_norm": 5.151943206787109, "learning_rate": 1.974706712943635e-05, "loss": 0.2115, "step": 537 }, { "epoch": 0.49403122130394855, "grad_norm": 1.4554216861724854, "learning_rate": 1.9744674259053263e-05, "loss": 0.1362, "step": 538 }, { "epoch": 0.494949494949495, "grad_norm": 0.7117225527763367, "learning_rate": 1.9742270269382112e-05, "loss": 0.0245, "step": 539 }, { "epoch": 0.49586776859504134, "grad_norm": 1.846510410308838, "learning_rate": 1.973985516316601e-05, "loss": 0.1728, "step": 540 }, { "epoch": 0.4967860422405877, "grad_norm": 1.3253047466278076, "learning_rate": 1.973742894316074e-05, "loss": 0.0515, "step": 541 }, { "epoch": 0.49770431588613406, "grad_norm": 0.16445739567279816, "learning_rate": 1.973499161213477e-05, "loss": 0.0035, "step": 542 }, { "epoch": 0.4986225895316804, "grad_norm": 1.2315943241119385, "learning_rate": 1.9732543172869252e-05, "loss": 0.1296, "step": 543 }, { "epoch": 0.4995408631772268, "grad_norm": 1.011060118675232, "learning_rate": 1.9730083628158007e-05, "loss": 0.0751, "step": 544 }, { "epoch": 0.5004591368227732, "grad_norm": 1.995467185974121, "learning_rate": 1.9727612980807534e-05, "loss": 0.0821, "step": 545 }, { "epoch": 0.5013774104683195, "grad_norm": 1.3598464727401733, "learning_rate": 1.9725131233636993e-05, "loss": 0.1139, "step": 546 }, { "epoch": 0.5022956841138659, "grad_norm": 1.7011747360229492, "learning_rate": 1.9722638389478218e-05, "loss": 0.1463, "step": 547 }, { "epoch": 0.5032139577594124, "grad_norm": 2.456599712371826, "learning_rate": 1.9720134451175702e-05, "loss": 0.1258, "step": 548 }, { "epoch": 0.5041322314049587, "grad_norm": 1.4699056148529053, "learning_rate": 1.9717619421586595e-05, "loss": 0.0283, "step": 549 }, { "epoch": 0.5050505050505051, "grad_norm": 0.15069150924682617, "learning_rate": 1.9715093303580702e-05, "loss": 0.0046, "step": 550 }, { "epoch": 0.5059687786960514, "grad_norm": 2.5119874477386475, "learning_rate": 1.9712556100040484e-05, "loss": 0.1294, "step": 551 }, { "epoch": 0.5068870523415978, "grad_norm": 0.4345417618751526, "learning_rate": 1.9710007813861055e-05, "loss": 0.0089, "step": 552 }, { "epoch": 0.5078053259871441, "grad_norm": 2.3631367683410645, "learning_rate": 1.9707448447950167e-05, "loss": 0.0812, "step": 553 }, { "epoch": 0.5087235996326905, "grad_norm": 1.0422861576080322, "learning_rate": 1.9704878005228214e-05, "loss": 0.0452, "step": 554 }, { "epoch": 0.509641873278237, "grad_norm": 2.5723071098327637, "learning_rate": 1.970229648862824e-05, "loss": 0.1262, "step": 555 }, { "epoch": 0.5105601469237833, "grad_norm": 1.4065065383911133, "learning_rate": 1.9699703901095915e-05, "loss": 0.0815, "step": 556 }, { "epoch": 0.5114784205693297, "grad_norm": 2.0104610919952393, "learning_rate": 1.969710024558954e-05, "loss": 0.2113, "step": 557 }, { "epoch": 0.512396694214876, "grad_norm": 1.028111457824707, "learning_rate": 1.9694485525080066e-05, "loss": 0.0332, "step": 558 }, { "epoch": 0.5133149678604224, "grad_norm": 0.5628036856651306, "learning_rate": 1.9691859742551035e-05, "loss": 0.012, "step": 559 }, { "epoch": 0.5142332415059688, "grad_norm": 1.6373012065887451, "learning_rate": 1.9689222900998644e-05, "loss": 0.4852, "step": 560 }, { "epoch": 0.5151515151515151, "grad_norm": 0.847821056842804, "learning_rate": 1.9686575003431687e-05, "loss": 0.0677, "step": 561 }, { "epoch": 0.5160697887970616, "grad_norm": 0.09126312285661697, "learning_rate": 1.9683916052871587e-05, "loss": 0.0039, "step": 562 }, { "epoch": 0.5169880624426079, "grad_norm": 0.18187028169631958, "learning_rate": 1.9681246052352372e-05, "loss": 0.0056, "step": 563 }, { "epoch": 0.5179063360881543, "grad_norm": 1.2133458852767944, "learning_rate": 1.967856500492068e-05, "loss": 0.039, "step": 564 }, { "epoch": 0.5188246097337006, "grad_norm": 0.48039376735687256, "learning_rate": 1.9675872913635764e-05, "loss": 0.0146, "step": 565 }, { "epoch": 0.519742883379247, "grad_norm": 2.2071475982666016, "learning_rate": 1.967316978156946e-05, "loss": 0.1088, "step": 566 }, { "epoch": 0.5206611570247934, "grad_norm": 0.5345866084098816, "learning_rate": 1.9670455611806214e-05, "loss": 0.0185, "step": 567 }, { "epoch": 0.5215794306703397, "grad_norm": 1.0209088325500488, "learning_rate": 1.9667730407443065e-05, "loss": 0.1018, "step": 568 }, { "epoch": 0.5224977043158862, "grad_norm": 1.1715754270553589, "learning_rate": 1.966499417158965e-05, "loss": 0.1206, "step": 569 }, { "epoch": 0.5234159779614325, "grad_norm": 3.837494373321533, "learning_rate": 1.966224690736818e-05, "loss": 0.1854, "step": 570 }, { "epoch": 0.5243342516069789, "grad_norm": 1.696754813194275, "learning_rate": 1.9659488617913457e-05, "loss": 0.0665, "step": 571 }, { "epoch": 0.5252525252525253, "grad_norm": 1.6204071044921875, "learning_rate": 1.965671930637286e-05, "loss": 0.09, "step": 572 }, { "epoch": 0.5261707988980716, "grad_norm": 0.4269339442253113, "learning_rate": 1.9653938975906356e-05, "loss": 0.0118, "step": 573 }, { "epoch": 0.527089072543618, "grad_norm": 2.255819320678711, "learning_rate": 1.9651147629686476e-05, "loss": 0.2011, "step": 574 }, { "epoch": 0.5280073461891643, "grad_norm": 1.7299877405166626, "learning_rate": 1.9648345270898317e-05, "loss": 0.179, "step": 575 }, { "epoch": 0.5289256198347108, "grad_norm": 1.1816836595535278, "learning_rate": 1.964553190273955e-05, "loss": 0.0858, "step": 576 }, { "epoch": 0.5298438934802571, "grad_norm": 0.6362840533256531, "learning_rate": 1.964270752842041e-05, "loss": 0.079, "step": 577 }, { "epoch": 0.5307621671258035, "grad_norm": 0.18778753280639648, "learning_rate": 1.963987215116368e-05, "loss": 0.0069, "step": 578 }, { "epoch": 0.5316804407713499, "grad_norm": 0.7993911504745483, "learning_rate": 1.9637025774204702e-05, "loss": 0.0387, "step": 579 }, { "epoch": 0.5325987144168962, "grad_norm": 0.20114266872406006, "learning_rate": 1.963416840079138e-05, "loss": 0.006, "step": 580 }, { "epoch": 0.5335169880624426, "grad_norm": 1.8525736331939697, "learning_rate": 1.9631300034184155e-05, "loss": 0.1582, "step": 581 }, { "epoch": 0.5344352617079889, "grad_norm": 0.15104453265666962, "learning_rate": 1.9628420677656012e-05, "loss": 0.0049, "step": 582 }, { "epoch": 0.5353535353535354, "grad_norm": 1.1583020687103271, "learning_rate": 1.9625530334492473e-05, "loss": 0.0402, "step": 583 }, { "epoch": 0.5362718089990818, "grad_norm": 0.6747578382492065, "learning_rate": 1.9622629007991614e-05, "loss": 0.0199, "step": 584 }, { "epoch": 0.5371900826446281, "grad_norm": 0.7891210913658142, "learning_rate": 1.961971670146402e-05, "loss": 0.0322, "step": 585 }, { "epoch": 0.5381083562901745, "grad_norm": 0.41837063431739807, "learning_rate": 1.961679341823283e-05, "loss": 0.0089, "step": 586 }, { "epoch": 0.5390266299357208, "grad_norm": 1.0988554954528809, "learning_rate": 1.9613859161633683e-05, "loss": 0.0286, "step": 587 }, { "epoch": 0.5399449035812672, "grad_norm": 2.40497088432312, "learning_rate": 1.9610913935014752e-05, "loss": 0.0655, "step": 588 }, { "epoch": 0.5408631772268135, "grad_norm": 0.9067246913909912, "learning_rate": 1.9607957741736728e-05, "loss": 0.0359, "step": 589 }, { "epoch": 0.54178145087236, "grad_norm": 0.9768886566162109, "learning_rate": 1.960499058517282e-05, "loss": 0.1452, "step": 590 }, { "epoch": 0.5426997245179064, "grad_norm": 0.47527778148651123, "learning_rate": 1.9602012468708733e-05, "loss": 0.0213, "step": 591 }, { "epoch": 0.5436179981634527, "grad_norm": 0.5622909665107727, "learning_rate": 1.959902339574269e-05, "loss": 0.039, "step": 592 }, { "epoch": 0.5445362718089991, "grad_norm": 1.079635500907898, "learning_rate": 1.9596023369685405e-05, "loss": 0.012, "step": 593 }, { "epoch": 0.5454545454545454, "grad_norm": 0.07651326805353165, "learning_rate": 1.9593012393960106e-05, "loss": 0.0021, "step": 594 }, { "epoch": 0.5463728191000918, "grad_norm": 2.094458818435669, "learning_rate": 1.95899904720025e-05, "loss": 0.0619, "step": 595 }, { "epoch": 0.5472910927456383, "grad_norm": 0.5298919677734375, "learning_rate": 1.9586957607260796e-05, "loss": 0.0179, "step": 596 }, { "epoch": 0.5482093663911846, "grad_norm": 1.6802196502685547, "learning_rate": 1.958391380319568e-05, "loss": 0.1501, "step": 597 }, { "epoch": 0.549127640036731, "grad_norm": 1.4646730422973633, "learning_rate": 1.9580859063280326e-05, "loss": 0.027, "step": 598 }, { "epoch": 0.5500459136822773, "grad_norm": 0.4926934540271759, "learning_rate": 1.9577793391000385e-05, "loss": 0.0087, "step": 599 }, { "epoch": 0.5509641873278237, "grad_norm": 0.3937261402606964, "learning_rate": 1.957471678985398e-05, "loss": 0.0104, "step": 600 }, { "epoch": 0.55188246097337, "grad_norm": 1.4782140254974365, "learning_rate": 1.957162926335171e-05, "loss": 0.0343, "step": 601 }, { "epoch": 0.5528007346189164, "grad_norm": 1.8491870164871216, "learning_rate": 1.9568530815016633e-05, "loss": 0.0909, "step": 602 }, { "epoch": 0.5537190082644629, "grad_norm": 1.6810264587402344, "learning_rate": 1.956542144838428e-05, "loss": 0.1031, "step": 603 }, { "epoch": 0.5546372819100092, "grad_norm": 3.613898277282715, "learning_rate": 1.9562301167002632e-05, "loss": 0.2311, "step": 604 }, { "epoch": 0.5555555555555556, "grad_norm": 5.023265838623047, "learning_rate": 1.955916997443213e-05, "loss": 0.2378, "step": 605 }, { "epoch": 0.5564738292011019, "grad_norm": 1.3591153621673584, "learning_rate": 1.955602787424565e-05, "loss": 0.1694, "step": 606 }, { "epoch": 0.5573921028466483, "grad_norm": 0.07303488999605179, "learning_rate": 1.9552874870028543e-05, "loss": 0.0022, "step": 607 }, { "epoch": 0.5583103764921947, "grad_norm": 0.3176625370979309, "learning_rate": 1.9549710965378576e-05, "loss": 0.0075, "step": 608 }, { "epoch": 0.559228650137741, "grad_norm": 0.3238816261291504, "learning_rate": 1.9546536163905966e-05, "loss": 0.026, "step": 609 }, { "epoch": 0.5601469237832875, "grad_norm": 1.0448073148727417, "learning_rate": 1.9543350469233366e-05, "loss": 0.0463, "step": 610 }, { "epoch": 0.5610651974288338, "grad_norm": 0.4612521827220917, "learning_rate": 1.9540153884995844e-05, "loss": 0.0072, "step": 611 }, { "epoch": 0.5619834710743802, "grad_norm": 2.2090866565704346, "learning_rate": 1.9536946414840913e-05, "loss": 0.0604, "step": 612 }, { "epoch": 0.5629017447199265, "grad_norm": 0.2630915343761444, "learning_rate": 1.9533728062428496e-05, "loss": 0.0037, "step": 613 }, { "epoch": 0.5638200183654729, "grad_norm": 2.531892776489258, "learning_rate": 1.9530498831430936e-05, "loss": 0.0831, "step": 614 }, { "epoch": 0.5647382920110193, "grad_norm": 1.944661021232605, "learning_rate": 1.952725872553299e-05, "loss": 0.1742, "step": 615 }, { "epoch": 0.5656565656565656, "grad_norm": 2.2934865951538086, "learning_rate": 1.9524007748431825e-05, "loss": 0.0764, "step": 616 }, { "epoch": 0.566574839302112, "grad_norm": 1.1092978715896606, "learning_rate": 1.9520745903837007e-05, "loss": 0.046, "step": 617 }, { "epoch": 0.5674931129476584, "grad_norm": 2.133833646774292, "learning_rate": 1.9517473195470512e-05, "loss": 0.1301, "step": 618 }, { "epoch": 0.5684113865932048, "grad_norm": 1.6055550575256348, "learning_rate": 1.95141896270667e-05, "loss": 0.0627, "step": 619 }, { "epoch": 0.5693296602387512, "grad_norm": 1.6075007915496826, "learning_rate": 1.9510895202372335e-05, "loss": 0.0938, "step": 620 }, { "epoch": 0.5702479338842975, "grad_norm": 2.925813913345337, "learning_rate": 1.950758992514656e-05, "loss": 0.0571, "step": 621 }, { "epoch": 0.5711662075298439, "grad_norm": 1.392244577407837, "learning_rate": 1.9504273799160907e-05, "loss": 0.0537, "step": 622 }, { "epoch": 0.5720844811753902, "grad_norm": 0.31894245743751526, "learning_rate": 1.9500946828199282e-05, "loss": 0.0084, "step": 623 }, { "epoch": 0.5730027548209367, "grad_norm": 0.12603555619716644, "learning_rate": 1.9497609016057972e-05, "loss": 0.003, "step": 624 }, { "epoch": 0.573921028466483, "grad_norm": 0.9948137402534485, "learning_rate": 1.949426036654563e-05, "loss": 0.0222, "step": 625 }, { "epoch": 0.5748393021120294, "grad_norm": 3.185605049133301, "learning_rate": 1.9490900883483275e-05, "loss": 0.2445, "step": 626 }, { "epoch": 0.5757575757575758, "grad_norm": 1.837626576423645, "learning_rate": 1.9487530570704293e-05, "loss": 0.2004, "step": 627 }, { "epoch": 0.5766758494031221, "grad_norm": 1.535634160041809, "learning_rate": 1.948414943205442e-05, "loss": 0.0598, "step": 628 }, { "epoch": 0.5775941230486685, "grad_norm": 0.16773362457752228, "learning_rate": 1.9480757471391745e-05, "loss": 0.0024, "step": 629 }, { "epoch": 0.5785123966942148, "grad_norm": 1.2377078533172607, "learning_rate": 1.947735469258672e-05, "loss": 0.0393, "step": 630 }, { "epoch": 0.5794306703397613, "grad_norm": 1.1349104642868042, "learning_rate": 1.947394109952212e-05, "loss": 0.0998, "step": 631 }, { "epoch": 0.5803489439853077, "grad_norm": 0.8528761267662048, "learning_rate": 1.9470516696093075e-05, "loss": 0.0208, "step": 632 }, { "epoch": 0.581267217630854, "grad_norm": 1.0262727737426758, "learning_rate": 1.9467081486207046e-05, "loss": 0.0826, "step": 633 }, { "epoch": 0.5821854912764004, "grad_norm": 1.363182544708252, "learning_rate": 1.9463635473783817e-05, "loss": 0.0477, "step": 634 }, { "epoch": 0.5831037649219467, "grad_norm": 1.0040497779846191, "learning_rate": 1.9460178662755516e-05, "loss": 0.0446, "step": 635 }, { "epoch": 0.5840220385674931, "grad_norm": 1.075750470161438, "learning_rate": 1.9456711057066574e-05, "loss": 0.0291, "step": 636 }, { "epoch": 0.5849403122130394, "grad_norm": 2.6203129291534424, "learning_rate": 1.9453232660673755e-05, "loss": 0.4601, "step": 637 }, { "epoch": 0.5858585858585859, "grad_norm": 1.3162869215011597, "learning_rate": 1.9449743477546122e-05, "loss": 0.1533, "step": 638 }, { "epoch": 0.5867768595041323, "grad_norm": 2.7481396198272705, "learning_rate": 1.9446243511665056e-05, "loss": 0.1457, "step": 639 }, { "epoch": 0.5876951331496786, "grad_norm": 1.808953881263733, "learning_rate": 1.944273276702424e-05, "loss": 0.1075, "step": 640 }, { "epoch": 0.588613406795225, "grad_norm": 0.2805784046649933, "learning_rate": 1.9439211247629654e-05, "loss": 0.0073, "step": 641 }, { "epoch": 0.5895316804407713, "grad_norm": 1.5811619758605957, "learning_rate": 1.9435678957499573e-05, "loss": 0.0586, "step": 642 }, { "epoch": 0.5904499540863177, "grad_norm": 0.8168129324913025, "learning_rate": 1.9432135900664565e-05, "loss": 0.063, "step": 643 }, { "epoch": 0.5913682277318641, "grad_norm": 0.1437254697084427, "learning_rate": 1.9428582081167477e-05, "loss": 0.0049, "step": 644 }, { "epoch": 0.5922865013774105, "grad_norm": 0.16943904757499695, "learning_rate": 1.9425017503063446e-05, "loss": 0.0054, "step": 645 }, { "epoch": 0.5932047750229569, "grad_norm": 1.2495930194854736, "learning_rate": 1.942144217041988e-05, "loss": 0.0734, "step": 646 }, { "epoch": 0.5941230486685032, "grad_norm": 2.1677286624908447, "learning_rate": 1.9417856087316455e-05, "loss": 0.1534, "step": 647 }, { "epoch": 0.5950413223140496, "grad_norm": 1.3407752513885498, "learning_rate": 1.9414259257845123e-05, "loss": 0.0561, "step": 648 }, { "epoch": 0.5959595959595959, "grad_norm": 1.4936124086380005, "learning_rate": 1.941065168611009e-05, "loss": 0.0733, "step": 649 }, { "epoch": 0.5968778696051423, "grad_norm": 1.0455621480941772, "learning_rate": 1.9407033376227823e-05, "loss": 0.0477, "step": 650 }, { "epoch": 0.5977961432506887, "grad_norm": 0.3708937466144562, "learning_rate": 1.9403404332327043e-05, "loss": 0.0078, "step": 651 }, { "epoch": 0.5987144168962351, "grad_norm": 0.9694789052009583, "learning_rate": 1.9399764558548723e-05, "loss": 0.066, "step": 652 }, { "epoch": 0.5996326905417815, "grad_norm": 0.946877121925354, "learning_rate": 1.9396114059046064e-05, "loss": 0.0379, "step": 653 }, { "epoch": 0.6005509641873278, "grad_norm": 1.5393356084823608, "learning_rate": 1.939245283798452e-05, "loss": 0.1857, "step": 654 }, { "epoch": 0.6014692378328742, "grad_norm": 0.8740437626838684, "learning_rate": 1.9388780899541785e-05, "loss": 0.0236, "step": 655 }, { "epoch": 0.6023875114784206, "grad_norm": 0.34330007433891296, "learning_rate": 1.9385098247907756e-05, "loss": 0.0111, "step": 656 }, { "epoch": 0.6033057851239669, "grad_norm": 0.4051339626312256, "learning_rate": 1.938140488728458e-05, "loss": 0.0132, "step": 657 }, { "epoch": 0.6042240587695134, "grad_norm": 1.438993215560913, "learning_rate": 1.937770082188662e-05, "loss": 0.1591, "step": 658 }, { "epoch": 0.6051423324150597, "grad_norm": 2.1236953735351562, "learning_rate": 1.9373986055940434e-05, "loss": 0.1256, "step": 659 }, { "epoch": 0.6060606060606061, "grad_norm": 0.3216019868850708, "learning_rate": 1.9370260593684817e-05, "loss": 0.0069, "step": 660 }, { "epoch": 0.6069788797061524, "grad_norm": 0.47244158387184143, "learning_rate": 1.936652443937075e-05, "loss": 0.0114, "step": 661 }, { "epoch": 0.6078971533516988, "grad_norm": 2.7887747287750244, "learning_rate": 1.936277759726142e-05, "loss": 0.215, "step": 662 }, { "epoch": 0.6088154269972452, "grad_norm": 1.2485493421554565, "learning_rate": 1.9359020071632218e-05, "loss": 0.0891, "step": 663 }, { "epoch": 0.6097337006427915, "grad_norm": 1.130566120147705, "learning_rate": 1.935525186677071e-05, "loss": 0.0598, "step": 664 }, { "epoch": 0.610651974288338, "grad_norm": 1.9977930784225464, "learning_rate": 1.9351472986976657e-05, "loss": 0.2201, "step": 665 }, { "epoch": 0.6115702479338843, "grad_norm": 1.617570400238037, "learning_rate": 1.9347683436562e-05, "loss": 0.1055, "step": 666 }, { "epoch": 0.6124885215794307, "grad_norm": 1.8524110317230225, "learning_rate": 1.9343883219850854e-05, "loss": 0.0547, "step": 667 }, { "epoch": 0.6134067952249771, "grad_norm": 1.338852047920227, "learning_rate": 1.934007234117951e-05, "loss": 0.0463, "step": 668 }, { "epoch": 0.6143250688705234, "grad_norm": 0.9554400444030762, "learning_rate": 1.9336250804896415e-05, "loss": 0.0365, "step": 669 }, { "epoch": 0.6152433425160698, "grad_norm": 1.9949162006378174, "learning_rate": 1.9332418615362186e-05, "loss": 0.1737, "step": 670 }, { "epoch": 0.6161616161616161, "grad_norm": 2.126028299331665, "learning_rate": 1.9328575776949595e-05, "loss": 0.0983, "step": 671 }, { "epoch": 0.6170798898071626, "grad_norm": 0.5846743583679199, "learning_rate": 1.932472229404356e-05, "loss": 0.0177, "step": 672 }, { "epoch": 0.6179981634527089, "grad_norm": 3.031480550765991, "learning_rate": 1.932085817104115e-05, "loss": 0.2119, "step": 673 }, { "epoch": 0.6189164370982553, "grad_norm": 1.390633463859558, "learning_rate": 1.931698341235157e-05, "loss": 0.0339, "step": 674 }, { "epoch": 0.6198347107438017, "grad_norm": 1.3791753053665161, "learning_rate": 1.931309802239617e-05, "loss": 0.0757, "step": 675 }, { "epoch": 0.620752984389348, "grad_norm": 1.1882648468017578, "learning_rate": 1.930920200560842e-05, "loss": 0.0607, "step": 676 }, { "epoch": 0.6216712580348944, "grad_norm": 1.0269478559494019, "learning_rate": 1.930529536643392e-05, "loss": 0.0549, "step": 677 }, { "epoch": 0.6225895316804407, "grad_norm": 1.4285205602645874, "learning_rate": 1.9301378109330396e-05, "loss": 0.0981, "step": 678 }, { "epoch": 0.6235078053259872, "grad_norm": 0.457570880651474, "learning_rate": 1.9297450238767684e-05, "loss": 0.0112, "step": 679 }, { "epoch": 0.6244260789715336, "grad_norm": 1.7735077142715454, "learning_rate": 1.929351175922773e-05, "loss": 0.1076, "step": 680 }, { "epoch": 0.6253443526170799, "grad_norm": 0.2730364203453064, "learning_rate": 1.928956267520459e-05, "loss": 0.006, "step": 681 }, { "epoch": 0.6262626262626263, "grad_norm": 0.10474833846092224, "learning_rate": 1.9285602991204415e-05, "loss": 0.003, "step": 682 }, { "epoch": 0.6271808999081726, "grad_norm": 1.7206974029541016, "learning_rate": 1.928163271174546e-05, "loss": 0.1656, "step": 683 }, { "epoch": 0.628099173553719, "grad_norm": 0.038157884031534195, "learning_rate": 1.927765184135806e-05, "loss": 0.0014, "step": 684 }, { "epoch": 0.6290174471992653, "grad_norm": 1.7900052070617676, "learning_rate": 1.927366038458464e-05, "loss": 0.0527, "step": 685 }, { "epoch": 0.6299357208448118, "grad_norm": 0.9853944778442383, "learning_rate": 1.9269658345979706e-05, "loss": 0.019, "step": 686 }, { "epoch": 0.6308539944903582, "grad_norm": 1.1626765727996826, "learning_rate": 1.926564573010983e-05, "loss": 0.1175, "step": 687 }, { "epoch": 0.6317722681359045, "grad_norm": 2.9818778038024902, "learning_rate": 1.9261622541553665e-05, "loss": 0.2471, "step": 688 }, { "epoch": 0.6326905417814509, "grad_norm": 0.25173455476760864, "learning_rate": 1.9257588784901926e-05, "loss": 0.0037, "step": 689 }, { "epoch": 0.6336088154269972, "grad_norm": 0.08437729626893997, "learning_rate": 1.9253544464757378e-05, "loss": 0.0023, "step": 690 }, { "epoch": 0.6345270890725436, "grad_norm": 2.3047845363616943, "learning_rate": 1.924948958573485e-05, "loss": 0.2006, "step": 691 }, { "epoch": 0.63544536271809, "grad_norm": 0.9033703207969666, "learning_rate": 1.9245424152461207e-05, "loss": 0.0446, "step": 692 }, { "epoch": 0.6363636363636364, "grad_norm": 0.07256557792425156, "learning_rate": 1.9241348169575372e-05, "loss": 0.0022, "step": 693 }, { "epoch": 0.6372819100091828, "grad_norm": 1.2655662298202515, "learning_rate": 1.92372616417283e-05, "loss": 0.1181, "step": 694 }, { "epoch": 0.6382001836547291, "grad_norm": 2.271008253097534, "learning_rate": 1.923316457358297e-05, "loss": 0.1673, "step": 695 }, { "epoch": 0.6391184573002755, "grad_norm": 1.1038682460784912, "learning_rate": 1.92290569698144e-05, "loss": 0.0583, "step": 696 }, { "epoch": 0.6400367309458218, "grad_norm": 1.3737159967422485, "learning_rate": 1.9224938835109627e-05, "loss": 0.1169, "step": 697 }, { "epoch": 0.6409550045913682, "grad_norm": 1.175825834274292, "learning_rate": 1.9220810174167695e-05, "loss": 0.0837, "step": 698 }, { "epoch": 0.6418732782369146, "grad_norm": 0.43475446105003357, "learning_rate": 1.9216670991699673e-05, "loss": 0.0145, "step": 699 }, { "epoch": 0.642791551882461, "grad_norm": 0.8955399990081787, "learning_rate": 1.921252129242863e-05, "loss": 0.0535, "step": 700 }, { "epoch": 0.6437098255280074, "grad_norm": 0.08064457774162292, "learning_rate": 1.920836108108963e-05, "loss": 0.0028, "step": 701 }, { "epoch": 0.6446280991735537, "grad_norm": 0.23636838793754578, "learning_rate": 1.9204190362429736e-05, "loss": 0.0099, "step": 702 }, { "epoch": 0.6455463728191001, "grad_norm": 1.9284014701843262, "learning_rate": 1.9200009141208006e-05, "loss": 0.141, "step": 703 }, { "epoch": 0.6464646464646465, "grad_norm": 1.9867844581604004, "learning_rate": 1.919581742219547e-05, "loss": 0.142, "step": 704 }, { "epoch": 0.6473829201101928, "grad_norm": 0.9472944140434265, "learning_rate": 1.9191615210175158e-05, "loss": 0.0452, "step": 705 }, { "epoch": 0.6483011937557392, "grad_norm": 2.8020215034484863, "learning_rate": 1.9187402509942045e-05, "loss": 0.1335, "step": 706 }, { "epoch": 0.6492194674012856, "grad_norm": 1.4216312170028687, "learning_rate": 1.918317932630309e-05, "loss": 0.0567, "step": 707 }, { "epoch": 0.650137741046832, "grad_norm": 0.5007367134094238, "learning_rate": 1.917894566407722e-05, "loss": 0.0154, "step": 708 }, { "epoch": 0.6510560146923783, "grad_norm": 1.7372359037399292, "learning_rate": 1.9174701528095302e-05, "loss": 0.0486, "step": 709 }, { "epoch": 0.6519742883379247, "grad_norm": 1.6262034177780151, "learning_rate": 1.9170446923200167e-05, "loss": 0.1116, "step": 710 }, { "epoch": 0.6528925619834711, "grad_norm": 0.6465626955032349, "learning_rate": 1.9166181854246586e-05, "loss": 0.0136, "step": 711 }, { "epoch": 0.6538108356290174, "grad_norm": 1.134250283241272, "learning_rate": 1.916190632610127e-05, "loss": 0.0423, "step": 712 }, { "epoch": 0.6547291092745638, "grad_norm": 0.4924454689025879, "learning_rate": 1.9157620343642872e-05, "loss": 0.0122, "step": 713 }, { "epoch": 0.6556473829201102, "grad_norm": 2.2086753845214844, "learning_rate": 1.9153323911761963e-05, "loss": 0.2314, "step": 714 }, { "epoch": 0.6565656565656566, "grad_norm": 1.6239358186721802, "learning_rate": 1.9149017035361046e-05, "loss": 0.1192, "step": 715 }, { "epoch": 0.657483930211203, "grad_norm": 1.7272869348526, "learning_rate": 1.9144699719354535e-05, "loss": 0.098, "step": 716 }, { "epoch": 0.6584022038567493, "grad_norm": 1.8614451885223389, "learning_rate": 1.9140371968668767e-05, "loss": 0.1456, "step": 717 }, { "epoch": 0.6593204775022957, "grad_norm": 0.82757967710495, "learning_rate": 1.913603378824197e-05, "loss": 0.0384, "step": 718 }, { "epoch": 0.660238751147842, "grad_norm": 0.9287965893745422, "learning_rate": 1.9131685183024288e-05, "loss": 0.0635, "step": 719 }, { "epoch": 0.6611570247933884, "grad_norm": 1.3210560083389282, "learning_rate": 1.9127326157977747e-05, "loss": 0.0587, "step": 720 }, { "epoch": 0.6620752984389348, "grad_norm": 1.2859951257705688, "learning_rate": 1.9122956718076277e-05, "loss": 0.0989, "step": 721 }, { "epoch": 0.6629935720844812, "grad_norm": 2.0840871334075928, "learning_rate": 1.911857686830568e-05, "loss": 0.0524, "step": 722 }, { "epoch": 0.6639118457300276, "grad_norm": 0.6828410625457764, "learning_rate": 1.9114186613663646e-05, "loss": 0.0156, "step": 723 }, { "epoch": 0.6648301193755739, "grad_norm": 1.1442022323608398, "learning_rate": 1.9109785959159727e-05, "loss": 0.015, "step": 724 }, { "epoch": 0.6657483930211203, "grad_norm": 1.1192210912704468, "learning_rate": 1.910537490981535e-05, "loss": 0.1071, "step": 725 }, { "epoch": 0.6666666666666666, "grad_norm": 1.887071967124939, "learning_rate": 1.9100953470663795e-05, "loss": 0.0783, "step": 726 }, { "epoch": 0.667584940312213, "grad_norm": 0.1044616848230362, "learning_rate": 1.909652164675021e-05, "loss": 0.0028, "step": 727 }, { "epoch": 0.6685032139577594, "grad_norm": 0.46700188517570496, "learning_rate": 1.9092079443131587e-05, "loss": 0.0121, "step": 728 }, { "epoch": 0.6694214876033058, "grad_norm": 0.6661792397499084, "learning_rate": 1.9087626864876755e-05, "loss": 0.0256, "step": 729 }, { "epoch": 0.6703397612488522, "grad_norm": 0.27669039368629456, "learning_rate": 1.908316391706639e-05, "loss": 0.0055, "step": 730 }, { "epoch": 0.6712580348943985, "grad_norm": 0.4949319064617157, "learning_rate": 1.9078690604792993e-05, "loss": 0.0122, "step": 731 }, { "epoch": 0.6721763085399449, "grad_norm": 1.2253912687301636, "learning_rate": 1.9074206933160898e-05, "loss": 0.0339, "step": 732 }, { "epoch": 0.6730945821854912, "grad_norm": 3.531479597091675, "learning_rate": 1.9069712907286257e-05, "loss": 0.1939, "step": 733 }, { "epoch": 0.6740128558310376, "grad_norm": 0.9280268549919128, "learning_rate": 1.9065208532297043e-05, "loss": 0.0241, "step": 734 }, { "epoch": 0.6749311294765841, "grad_norm": 0.13837631046772003, "learning_rate": 1.906069381333302e-05, "loss": 0.0036, "step": 735 }, { "epoch": 0.6758494031221304, "grad_norm": 0.6809249520301819, "learning_rate": 1.9056168755545777e-05, "loss": 0.0181, "step": 736 }, { "epoch": 0.6767676767676768, "grad_norm": 0.5183512568473816, "learning_rate": 1.9051633364098684e-05, "loss": 0.0185, "step": 737 }, { "epoch": 0.6776859504132231, "grad_norm": 0.9101403951644897, "learning_rate": 1.9047087644166914e-05, "loss": 0.0229, "step": 738 }, { "epoch": 0.6786042240587695, "grad_norm": 1.7349629402160645, "learning_rate": 1.904253160093742e-05, "loss": 0.1175, "step": 739 }, { "epoch": 0.6795224977043158, "grad_norm": 3.0038318634033203, "learning_rate": 1.9037965239608928e-05, "loss": 0.1777, "step": 740 }, { "epoch": 0.6804407713498623, "grad_norm": 0.13713420927524567, "learning_rate": 1.903338856539195e-05, "loss": 0.0031, "step": 741 }, { "epoch": 0.6813590449954087, "grad_norm": 2.2927701473236084, "learning_rate": 1.9028801583508762e-05, "loss": 0.204, "step": 742 }, { "epoch": 0.682277318640955, "grad_norm": 0.3198506832122803, "learning_rate": 1.9024204299193393e-05, "loss": 0.0081, "step": 743 }, { "epoch": 0.6831955922865014, "grad_norm": 0.2552512586116791, "learning_rate": 1.9019596717691642e-05, "loss": 0.0097, "step": 744 }, { "epoch": 0.6841138659320477, "grad_norm": 2.460127830505371, "learning_rate": 1.9014978844261042e-05, "loss": 0.1491, "step": 745 }, { "epoch": 0.6850321395775941, "grad_norm": 4.284700393676758, "learning_rate": 1.9010350684170883e-05, "loss": 0.155, "step": 746 }, { "epoch": 0.6859504132231405, "grad_norm": 2.954174757003784, "learning_rate": 1.9005712242702186e-05, "loss": 0.0723, "step": 747 }, { "epoch": 0.6868686868686869, "grad_norm": 3.387244462966919, "learning_rate": 1.900106352514771e-05, "loss": 0.1552, "step": 748 }, { "epoch": 0.6877869605142333, "grad_norm": 0.8756685853004456, "learning_rate": 1.899640453681193e-05, "loss": 0.0241, "step": 749 }, { "epoch": 0.6887052341597796, "grad_norm": 3.072763442993164, "learning_rate": 1.8991735283011047e-05, "loss": 0.1444, "step": 750 }, { "epoch": 0.689623507805326, "grad_norm": 1.1019874811172485, "learning_rate": 1.8987055769072973e-05, "loss": 0.0284, "step": 751 }, { "epoch": 0.6905417814508723, "grad_norm": 0.7862335443496704, "learning_rate": 1.898236600033733e-05, "loss": 0.0312, "step": 752 }, { "epoch": 0.6914600550964187, "grad_norm": 2.314314126968384, "learning_rate": 1.8977665982155435e-05, "loss": 0.0736, "step": 753 }, { "epoch": 0.6923783287419651, "grad_norm": 1.164863109588623, "learning_rate": 1.8972955719890313e-05, "loss": 0.0792, "step": 754 }, { "epoch": 0.6932966023875115, "grad_norm": 0.1360074132680893, "learning_rate": 1.896823521891667e-05, "loss": 0.0032, "step": 755 }, { "epoch": 0.6942148760330579, "grad_norm": 1.0688225030899048, "learning_rate": 1.8963504484620885e-05, "loss": 0.0306, "step": 756 }, { "epoch": 0.6951331496786042, "grad_norm": 0.6444321274757385, "learning_rate": 1.895876352240104e-05, "loss": 0.0155, "step": 757 }, { "epoch": 0.6960514233241506, "grad_norm": 2.659153461456299, "learning_rate": 1.8954012337666857e-05, "loss": 0.0502, "step": 758 }, { "epoch": 0.696969696969697, "grad_norm": 0.52004075050354, "learning_rate": 1.8949250935839745e-05, "loss": 0.0217, "step": 759 }, { "epoch": 0.6978879706152433, "grad_norm": 0.2477065771818161, "learning_rate": 1.8944479322352768e-05, "loss": 0.005, "step": 760 }, { "epoch": 0.6988062442607897, "grad_norm": 1.5473657846450806, "learning_rate": 1.8939697502650628e-05, "loss": 0.0161, "step": 761 }, { "epoch": 0.699724517906336, "grad_norm": 0.025443483144044876, "learning_rate": 1.8934905482189688e-05, "loss": 0.0008, "step": 762 }, { "epoch": 0.7006427915518825, "grad_norm": 0.20634689927101135, "learning_rate": 1.893010326643795e-05, "loss": 0.0038, "step": 763 }, { "epoch": 0.7015610651974288, "grad_norm": 0.8498860001564026, "learning_rate": 1.8925290860875037e-05, "loss": 0.0504, "step": 764 }, { "epoch": 0.7024793388429752, "grad_norm": 1.8241487741470337, "learning_rate": 1.892046827099221e-05, "loss": 0.1891, "step": 765 }, { "epoch": 0.7033976124885216, "grad_norm": 1.3340779542922974, "learning_rate": 1.8915635502292352e-05, "loss": 0.0586, "step": 766 }, { "epoch": 0.7043158861340679, "grad_norm": 0.44023260474205017, "learning_rate": 1.8910792560289953e-05, "loss": 0.009, "step": 767 }, { "epoch": 0.7052341597796143, "grad_norm": 4.375004291534424, "learning_rate": 1.8905939450511117e-05, "loss": 0.0916, "step": 768 }, { "epoch": 0.7061524334251607, "grad_norm": 1.837843656539917, "learning_rate": 1.890107617849355e-05, "loss": 0.1309, "step": 769 }, { "epoch": 0.7070707070707071, "grad_norm": 0.4709227681159973, "learning_rate": 1.8896202749786547e-05, "loss": 0.0048, "step": 770 }, { "epoch": 0.7079889807162535, "grad_norm": 0.09414304047822952, "learning_rate": 1.8891319169951003e-05, "loss": 0.0022, "step": 771 }, { "epoch": 0.7089072543617998, "grad_norm": 0.7529767751693726, "learning_rate": 1.8886425444559385e-05, "loss": 0.0228, "step": 772 }, { "epoch": 0.7098255280073462, "grad_norm": 1.9495905637741089, "learning_rate": 1.8881521579195752e-05, "loss": 0.181, "step": 773 }, { "epoch": 0.7107438016528925, "grad_norm": 1.1566153764724731, "learning_rate": 1.887660757945571e-05, "loss": 0.0135, "step": 774 }, { "epoch": 0.711662075298439, "grad_norm": 0.211246058344841, "learning_rate": 1.8871683450946458e-05, "loss": 0.0036, "step": 775 }, { "epoch": 0.7125803489439853, "grad_norm": 1.3777227401733398, "learning_rate": 1.886674919928672e-05, "loss": 0.0617, "step": 776 }, { "epoch": 0.7134986225895317, "grad_norm": 2.225163221359253, "learning_rate": 1.886180483010681e-05, "loss": 0.1379, "step": 777 }, { "epoch": 0.7144168962350781, "grad_norm": 1.160119891166687, "learning_rate": 1.8856850349048544e-05, "loss": 0.0533, "step": 778 }, { "epoch": 0.7153351698806244, "grad_norm": 0.5785722136497498, "learning_rate": 1.885188576176531e-05, "loss": 0.0151, "step": 779 }, { "epoch": 0.7162534435261708, "grad_norm": 3.85102915763855, "learning_rate": 1.884691107392201e-05, "loss": 0.8417, "step": 780 }, { "epoch": 0.7171717171717171, "grad_norm": 2.0654191970825195, "learning_rate": 1.8841926291195078e-05, "loss": 0.1111, "step": 781 }, { "epoch": 0.7180899908172635, "grad_norm": 1.9693912267684937, "learning_rate": 1.883693141927247e-05, "loss": 0.1106, "step": 782 }, { "epoch": 0.71900826446281, "grad_norm": 1.4405615329742432, "learning_rate": 1.8831926463853642e-05, "loss": 0.0634, "step": 783 }, { "epoch": 0.7199265381083563, "grad_norm": 0.39214181900024414, "learning_rate": 1.882691143064957e-05, "loss": 0.017, "step": 784 }, { "epoch": 0.7208448117539027, "grad_norm": 3.4594438076019287, "learning_rate": 1.8821886325382718e-05, "loss": 0.1739, "step": 785 }, { "epoch": 0.721763085399449, "grad_norm": 0.49304667115211487, "learning_rate": 1.8816851153787053e-05, "loss": 0.0091, "step": 786 }, { "epoch": 0.7226813590449954, "grad_norm": 1.2737538814544678, "learning_rate": 1.8811805921608024e-05, "loss": 0.1372, "step": 787 }, { "epoch": 0.7235996326905417, "grad_norm": 1.1017963886260986, "learning_rate": 1.8806750634602553e-05, "loss": 0.0641, "step": 788 }, { "epoch": 0.7245179063360881, "grad_norm": 1.7834010124206543, "learning_rate": 1.8801685298539048e-05, "loss": 0.1359, "step": 789 }, { "epoch": 0.7254361799816346, "grad_norm": 0.9019016027450562, "learning_rate": 1.879660991919737e-05, "loss": 0.0201, "step": 790 }, { "epoch": 0.7263544536271809, "grad_norm": 3.1717987060546875, "learning_rate": 1.8791524502368858e-05, "loss": 0.1556, "step": 791 }, { "epoch": 0.7272727272727273, "grad_norm": 2.6397202014923096, "learning_rate": 1.8786429053856282e-05, "loss": 0.1659, "step": 792 }, { "epoch": 0.7281910009182736, "grad_norm": 1.5660406351089478, "learning_rate": 1.8781323579473873e-05, "loss": 0.3124, "step": 793 }, { "epoch": 0.72910927456382, "grad_norm": 0.12215866148471832, "learning_rate": 1.8776208085047302e-05, "loss": 0.0034, "step": 794 }, { "epoch": 0.7300275482093664, "grad_norm": 1.3989065885543823, "learning_rate": 1.8771082576413674e-05, "loss": 0.0472, "step": 795 }, { "epoch": 0.7309458218549127, "grad_norm": 0.6318830251693726, "learning_rate": 1.876594705942151e-05, "loss": 0.0141, "step": 796 }, { "epoch": 0.7318640955004592, "grad_norm": 1.7975674867630005, "learning_rate": 1.8760801539930762e-05, "loss": 0.1401, "step": 797 }, { "epoch": 0.7327823691460055, "grad_norm": 1.3481440544128418, "learning_rate": 1.8755646023812796e-05, "loss": 0.0571, "step": 798 }, { "epoch": 0.7337006427915519, "grad_norm": 1.3416045904159546, "learning_rate": 1.875048051695038e-05, "loss": 0.0847, "step": 799 }, { "epoch": 0.7346189164370982, "grad_norm": 2.2551276683807373, "learning_rate": 1.8745305025237674e-05, "loss": 0.1843, "step": 800 }, { "epoch": 0.7355371900826446, "grad_norm": 0.9506853222846985, "learning_rate": 1.8740119554580252e-05, "loss": 0.0245, "step": 801 }, { "epoch": 0.736455463728191, "grad_norm": 1.6878376007080078, "learning_rate": 1.8734924110895056e-05, "loss": 0.0912, "step": 802 }, { "epoch": 0.7373737373737373, "grad_norm": 1.1176332235336304, "learning_rate": 1.872971870011042e-05, "loss": 0.1185, "step": 803 }, { "epoch": 0.7382920110192838, "grad_norm": 0.5398158431053162, "learning_rate": 1.8724503328166044e-05, "loss": 0.01, "step": 804 }, { "epoch": 0.7392102846648301, "grad_norm": 1.2089167833328247, "learning_rate": 1.8719278001012996e-05, "loss": 0.0553, "step": 805 }, { "epoch": 0.7401285583103765, "grad_norm": 1.458974003791809, "learning_rate": 1.87140427246137e-05, "loss": 0.0841, "step": 806 }, { "epoch": 0.7410468319559229, "grad_norm": 1.385846734046936, "learning_rate": 1.870879750494195e-05, "loss": 0.1654, "step": 807 }, { "epoch": 0.7419651056014692, "grad_norm": 0.34972456097602844, "learning_rate": 1.8703542347982856e-05, "loss": 0.0106, "step": 808 }, { "epoch": 0.7428833792470156, "grad_norm": 1.9664890766143799, "learning_rate": 1.869827725973289e-05, "loss": 0.1302, "step": 809 }, { "epoch": 0.743801652892562, "grad_norm": 1.3836567401885986, "learning_rate": 1.8693002246199858e-05, "loss": 0.0504, "step": 810 }, { "epoch": 0.7447199265381084, "grad_norm": 0.2944009602069855, "learning_rate": 1.868771731340287e-05, "loss": 0.0115, "step": 811 }, { "epoch": 0.7456382001836547, "grad_norm": 0.2606390416622162, "learning_rate": 1.8682422467372376e-05, "loss": 0.0063, "step": 812 }, { "epoch": 0.7465564738292011, "grad_norm": 1.7450973987579346, "learning_rate": 1.8677117714150128e-05, "loss": 0.0775, "step": 813 }, { "epoch": 0.7474747474747475, "grad_norm": 0.17461243271827698, "learning_rate": 1.867180305978918e-05, "loss": 0.004, "step": 814 }, { "epoch": 0.7483930211202938, "grad_norm": 1.8070703744888306, "learning_rate": 1.8666478510353893e-05, "loss": 0.2017, "step": 815 }, { "epoch": 0.7493112947658402, "grad_norm": 1.631805419921875, "learning_rate": 1.866114407191991e-05, "loss": 0.099, "step": 816 }, { "epoch": 0.7502295684113865, "grad_norm": 2.072880506515503, "learning_rate": 1.8655799750574166e-05, "loss": 0.1511, "step": 817 }, { "epoch": 0.751147842056933, "grad_norm": 0.24457967281341553, "learning_rate": 1.8650445552414868e-05, "loss": 0.0059, "step": 818 }, { "epoch": 0.7520661157024794, "grad_norm": 0.6207495927810669, "learning_rate": 1.864508148355149e-05, "loss": 0.0249, "step": 819 }, { "epoch": 0.7529843893480257, "grad_norm": 2.3262102603912354, "learning_rate": 1.8639707550104775e-05, "loss": 0.2827, "step": 820 }, { "epoch": 0.7539026629935721, "grad_norm": 1.6815576553344727, "learning_rate": 1.8634323758206718e-05, "loss": 0.1234, "step": 821 }, { "epoch": 0.7548209366391184, "grad_norm": 1.8384088277816772, "learning_rate": 1.862893011400057e-05, "loss": 0.0426, "step": 822 }, { "epoch": 0.7557392102846648, "grad_norm": 0.9627880454063416, "learning_rate": 1.8623526623640818e-05, "loss": 0.0722, "step": 823 }, { "epoch": 0.7566574839302111, "grad_norm": 0.9714153409004211, "learning_rate": 1.861811329329318e-05, "loss": 0.0904, "step": 824 }, { "epoch": 0.7575757575757576, "grad_norm": 1.4170223474502563, "learning_rate": 1.8612690129134616e-05, "loss": 0.1268, "step": 825 }, { "epoch": 0.758494031221304, "grad_norm": 0.2788683772087097, "learning_rate": 1.860725713735329e-05, "loss": 0.008, "step": 826 }, { "epoch": 0.7594123048668503, "grad_norm": 2.7739124298095703, "learning_rate": 1.8601814324148596e-05, "loss": 0.0573, "step": 827 }, { "epoch": 0.7603305785123967, "grad_norm": 0.4976068437099457, "learning_rate": 1.859636169573112e-05, "loss": 0.0129, "step": 828 }, { "epoch": 0.761248852157943, "grad_norm": 1.9394763708114624, "learning_rate": 1.8590899258322663e-05, "loss": 0.0542, "step": 829 }, { "epoch": 0.7621671258034894, "grad_norm": 0.5482538342475891, "learning_rate": 1.8585427018156203e-05, "loss": 0.0184, "step": 830 }, { "epoch": 0.7630853994490359, "grad_norm": 1.1924070119857788, "learning_rate": 1.8579944981475916e-05, "loss": 0.022, "step": 831 }, { "epoch": 0.7640036730945822, "grad_norm": 0.6716696619987488, "learning_rate": 1.8574453154537148e-05, "loss": 0.0238, "step": 832 }, { "epoch": 0.7649219467401286, "grad_norm": 0.21539214253425598, "learning_rate": 1.8568951543606424e-05, "loss": 0.0057, "step": 833 }, { "epoch": 0.7658402203856749, "grad_norm": 1.8140908479690552, "learning_rate": 1.8563440154961427e-05, "loss": 0.0761, "step": 834 }, { "epoch": 0.7667584940312213, "grad_norm": 0.17721927165985107, "learning_rate": 1.8557918994890998e-05, "loss": 0.0061, "step": 835 }, { "epoch": 0.7676767676767676, "grad_norm": 1.3231558799743652, "learning_rate": 1.855238806969513e-05, "loss": 0.0452, "step": 836 }, { "epoch": 0.768595041322314, "grad_norm": 0.8054507970809937, "learning_rate": 1.8546847385684962e-05, "loss": 0.0222, "step": 837 }, { "epoch": 0.7695133149678605, "grad_norm": 1.3089964389801025, "learning_rate": 1.8541296949182753e-05, "loss": 0.0343, "step": 838 }, { "epoch": 0.7704315886134068, "grad_norm": 3.2877535820007324, "learning_rate": 1.8535736766521915e-05, "loss": 0.1674, "step": 839 }, { "epoch": 0.7713498622589532, "grad_norm": 1.6836572885513306, "learning_rate": 1.8530166844046957e-05, "loss": 0.0717, "step": 840 }, { "epoch": 0.7722681359044995, "grad_norm": 0.22562739253044128, "learning_rate": 1.8524587188113515e-05, "loss": 0.0039, "step": 841 }, { "epoch": 0.7731864095500459, "grad_norm": 0.44364291429519653, "learning_rate": 1.8518997805088328e-05, "loss": 0.0144, "step": 842 }, { "epoch": 0.7741046831955923, "grad_norm": 1.6867196559906006, "learning_rate": 1.8513398701349236e-05, "loss": 0.0357, "step": 843 }, { "epoch": 0.7750229568411386, "grad_norm": 0.9813880324363708, "learning_rate": 1.850778988328517e-05, "loss": 0.0555, "step": 844 }, { "epoch": 0.7759412304866851, "grad_norm": 0.6549972891807556, "learning_rate": 1.8502171357296144e-05, "loss": 0.0218, "step": 845 }, { "epoch": 0.7768595041322314, "grad_norm": 2.977707624435425, "learning_rate": 1.849654312979325e-05, "loss": 0.1322, "step": 846 }, { "epoch": 0.7777777777777778, "grad_norm": 0.8461600542068481, "learning_rate": 1.8490905207198648e-05, "loss": 0.0268, "step": 847 }, { "epoch": 0.7786960514233241, "grad_norm": 2.5170679092407227, "learning_rate": 1.848525759594557e-05, "loss": 0.1393, "step": 848 }, { "epoch": 0.7796143250688705, "grad_norm": 0.4039227068424225, "learning_rate": 1.8479600302478293e-05, "loss": 0.0136, "step": 849 }, { "epoch": 0.7805325987144169, "grad_norm": 2.1216890811920166, "learning_rate": 1.8473933333252146e-05, "loss": 0.1483, "step": 850 }, { "epoch": 0.7814508723599632, "grad_norm": 1.2559642791748047, "learning_rate": 1.8468256694733496e-05, "loss": 0.1141, "step": 851 }, { "epoch": 0.7823691460055097, "grad_norm": 1.061733365058899, "learning_rate": 1.846257039339975e-05, "loss": 0.0828, "step": 852 }, { "epoch": 0.783287419651056, "grad_norm": 2.566311836242676, "learning_rate": 1.8456874435739337e-05, "loss": 0.133, "step": 853 }, { "epoch": 0.7842056932966024, "grad_norm": 3.4098916053771973, "learning_rate": 1.8451168828251695e-05, "loss": 0.0694, "step": 854 }, { "epoch": 0.7851239669421488, "grad_norm": 1.3004955053329468, "learning_rate": 1.8445453577447293e-05, "loss": 0.1197, "step": 855 }, { "epoch": 0.7860422405876951, "grad_norm": 1.1627532243728638, "learning_rate": 1.843972868984759e-05, "loss": 0.0367, "step": 856 }, { "epoch": 0.7869605142332415, "grad_norm": 1.7457060813903809, "learning_rate": 1.8433994171985036e-05, "loss": 0.1693, "step": 857 }, { "epoch": 0.7878787878787878, "grad_norm": 1.40994131565094, "learning_rate": 1.842825003040309e-05, "loss": 0.0394, "step": 858 }, { "epoch": 0.7887970615243343, "grad_norm": 1.2037501335144043, "learning_rate": 1.8422496271656173e-05, "loss": 0.038, "step": 859 }, { "epoch": 0.7897153351698806, "grad_norm": 1.3405145406723022, "learning_rate": 1.8416732902309684e-05, "loss": 0.0491, "step": 860 }, { "epoch": 0.790633608815427, "grad_norm": 2.6144192218780518, "learning_rate": 1.8410959928939998e-05, "loss": 0.1389, "step": 861 }, { "epoch": 0.7915518824609734, "grad_norm": 1.0645402669906616, "learning_rate": 1.8405177358134436e-05, "loss": 0.0593, "step": 862 }, { "epoch": 0.7924701561065197, "grad_norm": 0.2787421643733978, "learning_rate": 1.839938519649128e-05, "loss": 0.0072, "step": 863 }, { "epoch": 0.7933884297520661, "grad_norm": 2.097607135772705, "learning_rate": 1.839358345061975e-05, "loss": 0.1244, "step": 864 }, { "epoch": 0.7943067033976124, "grad_norm": 2.5458250045776367, "learning_rate": 1.838777212714e-05, "loss": 0.1314, "step": 865 }, { "epoch": 0.7952249770431589, "grad_norm": 1.4508072137832642, "learning_rate": 1.838195123268313e-05, "loss": 0.1097, "step": 866 }, { "epoch": 0.7961432506887053, "grad_norm": 2.435711622238159, "learning_rate": 1.837612077389113e-05, "loss": 0.1038, "step": 867 }, { "epoch": 0.7970615243342516, "grad_norm": 0.8004124164581299, "learning_rate": 1.837028075741693e-05, "loss": 0.0196, "step": 868 }, { "epoch": 0.797979797979798, "grad_norm": 0.9533159136772156, "learning_rate": 1.8364431189924365e-05, "loss": 0.0857, "step": 869 }, { "epoch": 0.7988980716253443, "grad_norm": 1.2293213605880737, "learning_rate": 1.8358572078088144e-05, "loss": 0.0587, "step": 870 }, { "epoch": 0.7998163452708907, "grad_norm": 2.7401678562164307, "learning_rate": 1.83527034285939e-05, "loss": 0.0624, "step": 871 }, { "epoch": 0.800734618916437, "grad_norm": 4.748978137969971, "learning_rate": 1.834682524813812e-05, "loss": 0.386, "step": 872 }, { "epoch": 0.8016528925619835, "grad_norm": 0.7893763780593872, "learning_rate": 1.8340937543428184e-05, "loss": 0.0118, "step": 873 }, { "epoch": 0.8025711662075299, "grad_norm": 0.28215694427490234, "learning_rate": 1.8335040321182333e-05, "loss": 0.0063, "step": 874 }, { "epoch": 0.8034894398530762, "grad_norm": 0.8345978856086731, "learning_rate": 1.8329133588129667e-05, "loss": 0.0535, "step": 875 }, { "epoch": 0.8044077134986226, "grad_norm": 1.7630425691604614, "learning_rate": 1.8323217351010146e-05, "loss": 0.1389, "step": 876 }, { "epoch": 0.8053259871441689, "grad_norm": 1.4005361795425415, "learning_rate": 1.831729161657457e-05, "loss": 0.0913, "step": 877 }, { "epoch": 0.8062442607897153, "grad_norm": 1.3438661098480225, "learning_rate": 1.831135639158457e-05, "loss": 0.05, "step": 878 }, { "epoch": 0.8071625344352618, "grad_norm": 1.4552983045578003, "learning_rate": 1.8305411682812618e-05, "loss": 0.1304, "step": 879 }, { "epoch": 0.8080808080808081, "grad_norm": 0.48621565103530884, "learning_rate": 1.8299457497042002e-05, "loss": 0.017, "step": 880 }, { "epoch": 0.8089990817263545, "grad_norm": 0.5654535293579102, "learning_rate": 1.8293493841066818e-05, "loss": 0.0131, "step": 881 }, { "epoch": 0.8099173553719008, "grad_norm": 1.0698201656341553, "learning_rate": 1.8287520721691975e-05, "loss": 0.1522, "step": 882 }, { "epoch": 0.8108356290174472, "grad_norm": 1.4909547567367554, "learning_rate": 1.8281538145733186e-05, "loss": 0.1844, "step": 883 }, { "epoch": 0.8117539026629935, "grad_norm": 1.5333360433578491, "learning_rate": 1.827554612001694e-05, "loss": 0.0996, "step": 884 }, { "epoch": 0.8126721763085399, "grad_norm": 1.515586018562317, "learning_rate": 1.8269544651380516e-05, "loss": 0.0742, "step": 885 }, { "epoch": 0.8135904499540864, "grad_norm": 1.4595680236816406, "learning_rate": 1.826353374667197e-05, "loss": 0.0918, "step": 886 }, { "epoch": 0.8145087235996327, "grad_norm": 0.193825826048851, "learning_rate": 1.825751341275013e-05, "loss": 0.0074, "step": 887 }, { "epoch": 0.8154269972451791, "grad_norm": 0.9326702952384949, "learning_rate": 1.825148365648457e-05, "loss": 0.0237, "step": 888 }, { "epoch": 0.8163452708907254, "grad_norm": 0.3451515734195709, "learning_rate": 1.824544448475562e-05, "loss": 0.0159, "step": 889 }, { "epoch": 0.8172635445362718, "grad_norm": 2.5906782150268555, "learning_rate": 1.823939590445436e-05, "loss": 0.2269, "step": 890 }, { "epoch": 0.8181818181818182, "grad_norm": 0.226533442735672, "learning_rate": 1.8233337922482602e-05, "loss": 0.0073, "step": 891 }, { "epoch": 0.8191000918273645, "grad_norm": 1.466554045677185, "learning_rate": 1.8227270545752888e-05, "loss": 0.1098, "step": 892 }, { "epoch": 0.820018365472911, "grad_norm": 0.24480152130126953, "learning_rate": 1.8221193781188474e-05, "loss": 0.0057, "step": 893 }, { "epoch": 0.8209366391184573, "grad_norm": 0.12010310590267181, "learning_rate": 1.8215107635723326e-05, "loss": 0.0047, "step": 894 }, { "epoch": 0.8218549127640037, "grad_norm": 0.861146867275238, "learning_rate": 1.820901211630213e-05, "loss": 0.0554, "step": 895 }, { "epoch": 0.82277318640955, "grad_norm": 1.1884863376617432, "learning_rate": 1.8202907229880257e-05, "loss": 0.0452, "step": 896 }, { "epoch": 0.8236914600550964, "grad_norm": 1.2261710166931152, "learning_rate": 1.8196792983423758e-05, "loss": 0.1002, "step": 897 }, { "epoch": 0.8246097337006428, "grad_norm": 2.0717923641204834, "learning_rate": 1.8190669383909386e-05, "loss": 0.189, "step": 898 }, { "epoch": 0.8255280073461891, "grad_norm": 0.8087487816810608, "learning_rate": 1.8184536438324544e-05, "loss": 0.0182, "step": 899 }, { "epoch": 0.8264462809917356, "grad_norm": 0.5492955446243286, "learning_rate": 1.817839415366732e-05, "loss": 0.0155, "step": 900 }, { "epoch": 0.8273645546372819, "grad_norm": 1.1123809814453125, "learning_rate": 1.817224253694644e-05, "loss": 0.055, "step": 901 }, { "epoch": 0.8282828282828283, "grad_norm": 0.3334474265575409, "learning_rate": 1.816608159518129e-05, "loss": 0.0073, "step": 902 }, { "epoch": 0.8292011019283747, "grad_norm": 0.5147704482078552, "learning_rate": 1.815991133540189e-05, "loss": 0.0156, "step": 903 }, { "epoch": 0.830119375573921, "grad_norm": 2.113401174545288, "learning_rate": 1.8153731764648907e-05, "loss": 0.3785, "step": 904 }, { "epoch": 0.8310376492194674, "grad_norm": 0.08946482837200165, "learning_rate": 1.8147542889973605e-05, "loss": 0.0038, "step": 905 }, { "epoch": 0.8319559228650137, "grad_norm": 1.5212361812591553, "learning_rate": 1.814134471843789e-05, "loss": 0.0533, "step": 906 }, { "epoch": 0.8328741965105602, "grad_norm": 1.2291266918182373, "learning_rate": 1.8135137257114266e-05, "loss": 0.0852, "step": 907 }, { "epoch": 0.8337924701561065, "grad_norm": 1.5695605278015137, "learning_rate": 1.8128920513085834e-05, "loss": 0.0546, "step": 908 }, { "epoch": 0.8347107438016529, "grad_norm": 2.726864814758301, "learning_rate": 1.81226944934463e-05, "loss": 0.1589, "step": 909 }, { "epoch": 0.8356290174471993, "grad_norm": 2.160820245742798, "learning_rate": 1.8116459205299932e-05, "loss": 0.1431, "step": 910 }, { "epoch": 0.8365472910927456, "grad_norm": 0.5509659647941589, "learning_rate": 1.8110214655761598e-05, "loss": 0.0169, "step": 911 }, { "epoch": 0.837465564738292, "grad_norm": 1.1949433088302612, "learning_rate": 1.8103960851956718e-05, "loss": 0.0699, "step": 912 }, { "epoch": 0.8383838383838383, "grad_norm": 0.40145841240882874, "learning_rate": 1.8097697801021277e-05, "loss": 0.0194, "step": 913 }, { "epoch": 0.8393021120293848, "grad_norm": 0.9899370074272156, "learning_rate": 1.809142551010181e-05, "loss": 0.0522, "step": 914 }, { "epoch": 0.8402203856749312, "grad_norm": 1.680017352104187, "learning_rate": 1.8085143986355405e-05, "loss": 0.0745, "step": 915 }, { "epoch": 0.8411386593204775, "grad_norm": 1.1087539196014404, "learning_rate": 1.8078853236949666e-05, "loss": 0.0195, "step": 916 }, { "epoch": 0.8420569329660239, "grad_norm": 1.9482910633087158, "learning_rate": 1.807255326906274e-05, "loss": 0.1066, "step": 917 }, { "epoch": 0.8429752066115702, "grad_norm": 0.1378219723701477, "learning_rate": 1.8066244089883288e-05, "loss": 0.0047, "step": 918 }, { "epoch": 0.8438934802571166, "grad_norm": 0.5931695699691772, "learning_rate": 1.805992570661048e-05, "loss": 0.0179, "step": 919 }, { "epoch": 0.8448117539026629, "grad_norm": 0.6738635897636414, "learning_rate": 1.8053598126453993e-05, "loss": 0.0267, "step": 920 }, { "epoch": 0.8457300275482094, "grad_norm": 1.0406776666641235, "learning_rate": 1.804726135663399e-05, "loss": 0.0287, "step": 921 }, { "epoch": 0.8466483011937558, "grad_norm": 2.4274826049804688, "learning_rate": 1.804091540438113e-05, "loss": 0.2228, "step": 922 }, { "epoch": 0.8475665748393021, "grad_norm": 0.9579838514328003, "learning_rate": 1.803456027693654e-05, "loss": 0.0832, "step": 923 }, { "epoch": 0.8484848484848485, "grad_norm": 1.083612084388733, "learning_rate": 1.8028195981551823e-05, "loss": 0.0825, "step": 924 }, { "epoch": 0.8494031221303948, "grad_norm": 1.7974944114685059, "learning_rate": 1.8021822525489043e-05, "loss": 0.1507, "step": 925 }, { "epoch": 0.8503213957759412, "grad_norm": 1.0680992603302002, "learning_rate": 1.8015439916020715e-05, "loss": 0.0823, "step": 926 }, { "epoch": 0.8512396694214877, "grad_norm": 0.8491564989089966, "learning_rate": 1.80090481604298e-05, "loss": 0.0188, "step": 927 }, { "epoch": 0.852157943067034, "grad_norm": 1.6451951265335083, "learning_rate": 1.8002647266009696e-05, "loss": 0.1188, "step": 928 }, { "epoch": 0.8530762167125804, "grad_norm": 1.5458258390426636, "learning_rate": 1.799623724006422e-05, "loss": 0.0436, "step": 929 }, { "epoch": 0.8539944903581267, "grad_norm": 0.3100503981113434, "learning_rate": 1.7989818089907622e-05, "loss": 0.0093, "step": 930 }, { "epoch": 0.8549127640036731, "grad_norm": 1.3783217668533325, "learning_rate": 1.7983389822864557e-05, "loss": 0.1038, "step": 931 }, { "epoch": 0.8558310376492194, "grad_norm": 0.6798468828201294, "learning_rate": 1.797695244627008e-05, "loss": 0.0383, "step": 932 }, { "epoch": 0.8567493112947658, "grad_norm": 0.9993956685066223, "learning_rate": 1.7970505967469645e-05, "loss": 0.1685, "step": 933 }, { "epoch": 0.8576675849403123, "grad_norm": 0.42667651176452637, "learning_rate": 1.7964050393819095e-05, "loss": 0.0157, "step": 934 }, { "epoch": 0.8585858585858586, "grad_norm": 0.6041220426559448, "learning_rate": 1.7957585732684642e-05, "loss": 0.0197, "step": 935 }, { "epoch": 0.859504132231405, "grad_norm": 0.9719440340995789, "learning_rate": 1.7951111991442875e-05, "loss": 0.0551, "step": 936 }, { "epoch": 0.8604224058769513, "grad_norm": 0.8458185791969299, "learning_rate": 1.7944629177480738e-05, "loss": 0.0385, "step": 937 }, { "epoch": 0.8613406795224977, "grad_norm": 1.3269394636154175, "learning_rate": 1.793813729819553e-05, "loss": 0.0792, "step": 938 }, { "epoch": 0.8622589531680441, "grad_norm": 1.7172402143478394, "learning_rate": 1.7931636360994903e-05, "loss": 0.1974, "step": 939 }, { "epoch": 0.8631772268135904, "grad_norm": 0.33468616008758545, "learning_rate": 1.792512637329683e-05, "loss": 0.0105, "step": 940 }, { "epoch": 0.8640955004591369, "grad_norm": 2.490328311920166, "learning_rate": 1.7918607342529615e-05, "loss": 0.2181, "step": 941 }, { "epoch": 0.8650137741046832, "grad_norm": 1.8517637252807617, "learning_rate": 1.791207927613189e-05, "loss": 0.2245, "step": 942 }, { "epoch": 0.8659320477502296, "grad_norm": 2.4156107902526855, "learning_rate": 1.7905542181552586e-05, "loss": 0.0902, "step": 943 }, { "epoch": 0.8668503213957759, "grad_norm": 0.47747623920440674, "learning_rate": 1.789899606625094e-05, "loss": 0.0167, "step": 944 }, { "epoch": 0.8677685950413223, "grad_norm": 0.8178057074546814, "learning_rate": 1.7892440937696485e-05, "loss": 0.0253, "step": 945 }, { "epoch": 0.8686868686868687, "grad_norm": 1.5786612033843994, "learning_rate": 1.788587680336904e-05, "loss": 0.0777, "step": 946 }, { "epoch": 0.869605142332415, "grad_norm": 1.9165605306625366, "learning_rate": 1.7879303670758687e-05, "loss": 0.2, "step": 947 }, { "epoch": 0.8705234159779615, "grad_norm": 2.3003029823303223, "learning_rate": 1.7872721547365792e-05, "loss": 0.1624, "step": 948 }, { "epoch": 0.8714416896235078, "grad_norm": 1.3562381267547607, "learning_rate": 1.786613044070097e-05, "loss": 0.0472, "step": 949 }, { "epoch": 0.8723599632690542, "grad_norm": 0.6948869824409485, "learning_rate": 1.785953035828509e-05, "loss": 0.0247, "step": 950 }, { "epoch": 0.8732782369146006, "grad_norm": 0.06592235714197159, "learning_rate": 1.7852921307649262e-05, "loss": 0.0022, "step": 951 }, { "epoch": 0.8741965105601469, "grad_norm": 1.1386051177978516, "learning_rate": 1.7846303296334833e-05, "loss": 0.1259, "step": 952 }, { "epoch": 0.8751147842056933, "grad_norm": 0.3553920388221741, "learning_rate": 1.7839676331893366e-05, "loss": 0.0124, "step": 953 }, { "epoch": 0.8760330578512396, "grad_norm": 0.06424497812986374, "learning_rate": 1.783304042188665e-05, "loss": 0.0021, "step": 954 }, { "epoch": 0.8769513314967861, "grad_norm": 0.5649924278259277, "learning_rate": 1.782639557388667e-05, "loss": 0.0258, "step": 955 }, { "epoch": 0.8778696051423324, "grad_norm": 1.0281494855880737, "learning_rate": 1.7819741795475628e-05, "loss": 0.0493, "step": 956 }, { "epoch": 0.8787878787878788, "grad_norm": 0.4365397095680237, "learning_rate": 1.7813079094245897e-05, "loss": 0.015, "step": 957 }, { "epoch": 0.8797061524334252, "grad_norm": 1.9280999898910522, "learning_rate": 1.7806407477800042e-05, "loss": 0.1154, "step": 958 }, { "epoch": 0.8806244260789715, "grad_norm": 0.18862737715244293, "learning_rate": 1.7799726953750797e-05, "loss": 0.0042, "step": 959 }, { "epoch": 0.8815426997245179, "grad_norm": 0.10989876091480255, "learning_rate": 1.7793037529721058e-05, "loss": 0.0037, "step": 960 }, { "epoch": 0.8824609733700642, "grad_norm": 3.2657902240753174, "learning_rate": 1.778633921334389e-05, "loss": 0.4812, "step": 961 }, { "epoch": 0.8833792470156107, "grad_norm": 1.0967702865600586, "learning_rate": 1.7779632012262485e-05, "loss": 0.0691, "step": 962 }, { "epoch": 0.8842975206611571, "grad_norm": 1.1379897594451904, "learning_rate": 1.7772915934130182e-05, "loss": 0.0549, "step": 963 }, { "epoch": 0.8852157943067034, "grad_norm": 1.023267388343811, "learning_rate": 1.7766190986610457e-05, "loss": 0.0232, "step": 964 }, { "epoch": 0.8861340679522498, "grad_norm": 0.6641501188278198, "learning_rate": 1.77594571773769e-05, "loss": 0.023, "step": 965 }, { "epoch": 0.8870523415977961, "grad_norm": 1.1474347114562988, "learning_rate": 1.7752714514113202e-05, "loss": 0.0729, "step": 966 }, { "epoch": 0.8879706152433425, "grad_norm": 2.310032606124878, "learning_rate": 1.7745963004513176e-05, "loss": 0.0565, "step": 967 }, { "epoch": 0.8888888888888888, "grad_norm": 1.5633028745651245, "learning_rate": 1.7739202656280718e-05, "loss": 0.0949, "step": 968 }, { "epoch": 0.8898071625344353, "grad_norm": 1.0546313524246216, "learning_rate": 1.773243347712981e-05, "loss": 0.0793, "step": 969 }, { "epoch": 0.8907254361799817, "grad_norm": 3.661121368408203, "learning_rate": 1.7725655474784514e-05, "loss": 0.6676, "step": 970 }, { "epoch": 0.891643709825528, "grad_norm": 0.11959222704172134, "learning_rate": 1.7718868656978957e-05, "loss": 0.0039, "step": 971 }, { "epoch": 0.8925619834710744, "grad_norm": 0.6278905272483826, "learning_rate": 1.7712073031457332e-05, "loss": 0.0378, "step": 972 }, { "epoch": 0.8934802571166207, "grad_norm": 1.9301574230194092, "learning_rate": 1.7705268605973866e-05, "loss": 0.1242, "step": 973 }, { "epoch": 0.8943985307621671, "grad_norm": 1.1021806001663208, "learning_rate": 1.7698455388292846e-05, "loss": 0.0278, "step": 974 }, { "epoch": 0.8953168044077136, "grad_norm": 0.831920325756073, "learning_rate": 1.769163338618858e-05, "loss": 0.0377, "step": 975 }, { "epoch": 0.8962350780532599, "grad_norm": 2.250180244445801, "learning_rate": 1.7684802607445405e-05, "loss": 0.1946, "step": 976 }, { "epoch": 0.8971533516988063, "grad_norm": 0.3181585669517517, "learning_rate": 1.7677963059857665e-05, "loss": 0.0108, "step": 977 }, { "epoch": 0.8980716253443526, "grad_norm": 0.2845122516155243, "learning_rate": 1.7671114751229722e-05, "loss": 0.0119, "step": 978 }, { "epoch": 0.898989898989899, "grad_norm": 1.931889295578003, "learning_rate": 1.7664257689375923e-05, "loss": 0.1545, "step": 979 }, { "epoch": 0.8999081726354453, "grad_norm": 0.12999482452869415, "learning_rate": 1.7657391882120613e-05, "loss": 0.0031, "step": 980 }, { "epoch": 0.9008264462809917, "grad_norm": 0.7030467391014099, "learning_rate": 1.765051733729811e-05, "loss": 0.0129, "step": 981 }, { "epoch": 0.9017447199265382, "grad_norm": 2.0100812911987305, "learning_rate": 1.7643634062752698e-05, "loss": 0.1267, "step": 982 }, { "epoch": 0.9026629935720845, "grad_norm": 1.7277295589447021, "learning_rate": 1.7636742066338636e-05, "loss": 0.1152, "step": 983 }, { "epoch": 0.9035812672176309, "grad_norm": 0.7424048781394958, "learning_rate": 1.7629841355920118e-05, "loss": 0.0227, "step": 984 }, { "epoch": 0.9044995408631772, "grad_norm": 1.2281509637832642, "learning_rate": 1.7622931939371294e-05, "loss": 0.0965, "step": 985 }, { "epoch": 0.9054178145087236, "grad_norm": 0.07850968092679977, "learning_rate": 1.7616013824576245e-05, "loss": 0.0025, "step": 986 }, { "epoch": 0.90633608815427, "grad_norm": 1.1314661502838135, "learning_rate": 1.7609087019428976e-05, "loss": 0.044, "step": 987 }, { "epoch": 0.9072543617998163, "grad_norm": 1.4295257329940796, "learning_rate": 1.7602151531833405e-05, "loss": 0.267, "step": 988 }, { "epoch": 0.9081726354453628, "grad_norm": 1.7959898710250854, "learning_rate": 1.759520736970337e-05, "loss": 0.1366, "step": 989 }, { "epoch": 0.9090909090909091, "grad_norm": 0.6427844762802124, "learning_rate": 1.758825454096259e-05, "loss": 0.0592, "step": 990 }, { "epoch": 0.9100091827364555, "grad_norm": 0.46331170201301575, "learning_rate": 1.7581293053544678e-05, "loss": 0.0093, "step": 991 }, { "epoch": 0.9109274563820018, "grad_norm": 3.592252016067505, "learning_rate": 1.7574322915393143e-05, "loss": 0.2936, "step": 992 }, { "epoch": 0.9118457300275482, "grad_norm": 1.1294585466384888, "learning_rate": 1.756734413446134e-05, "loss": 0.0361, "step": 993 }, { "epoch": 0.9127640036730946, "grad_norm": 1.6938261985778809, "learning_rate": 1.7560356718712506e-05, "loss": 0.0416, "step": 994 }, { "epoch": 0.9136822773186409, "grad_norm": 1.2837735414505005, "learning_rate": 1.755336067611972e-05, "loss": 0.0715, "step": 995 }, { "epoch": 0.9146005509641874, "grad_norm": 0.06934461742639542, "learning_rate": 1.754635601466591e-05, "loss": 0.0024, "step": 996 }, { "epoch": 0.9155188246097337, "grad_norm": 0.86185622215271, "learning_rate": 1.753934274234383e-05, "loss": 0.0221, "step": 997 }, { "epoch": 0.9164370982552801, "grad_norm": 0.5301318764686584, "learning_rate": 1.7532320867156075e-05, "loss": 0.0204, "step": 998 }, { "epoch": 0.9173553719008265, "grad_norm": 0.7848318815231323, "learning_rate": 1.752529039711504e-05, "loss": 0.0391, "step": 999 }, { "epoch": 0.9182736455463728, "grad_norm": 0.13388220965862274, "learning_rate": 1.751825134024294e-05, "loss": 0.0046, "step": 1000 }, { "epoch": 0.9191919191919192, "grad_norm": 0.32933199405670166, "learning_rate": 1.7511203704571785e-05, "loss": 0.0076, "step": 1001 }, { "epoch": 0.9201101928374655, "grad_norm": 1.2730234861373901, "learning_rate": 1.7504147498143365e-05, "loss": 0.1021, "step": 1002 }, { "epoch": 0.921028466483012, "grad_norm": 1.131553053855896, "learning_rate": 1.749708272900926e-05, "loss": 0.0338, "step": 1003 }, { "epoch": 0.9219467401285583, "grad_norm": 1.2142157554626465, "learning_rate": 1.7490009405230816e-05, "loss": 0.0856, "step": 1004 }, { "epoch": 0.9228650137741047, "grad_norm": 1.7515645027160645, "learning_rate": 1.7482927534879147e-05, "loss": 0.0914, "step": 1005 }, { "epoch": 0.9237832874196511, "grad_norm": 0.6229833960533142, "learning_rate": 1.7475837126035105e-05, "loss": 0.0112, "step": 1006 }, { "epoch": 0.9247015610651974, "grad_norm": 1.0549362897872925, "learning_rate": 1.74687381867893e-05, "loss": 0.1042, "step": 1007 }, { "epoch": 0.9256198347107438, "grad_norm": 0.39923763275146484, "learning_rate": 1.746163072524207e-05, "loss": 0.0114, "step": 1008 }, { "epoch": 0.9265381083562901, "grad_norm": 0.20008127391338348, "learning_rate": 1.745451474950347e-05, "loss": 0.0069, "step": 1009 }, { "epoch": 0.9274563820018366, "grad_norm": 1.5875968933105469, "learning_rate": 1.744739026769329e-05, "loss": 0.2231, "step": 1010 }, { "epoch": 0.928374655647383, "grad_norm": 1.8158200979232788, "learning_rate": 1.7440257287941e-05, "loss": 0.1022, "step": 1011 }, { "epoch": 0.9292929292929293, "grad_norm": 1.6675087213516235, "learning_rate": 1.7433115818385787e-05, "loss": 0.15, "step": 1012 }, { "epoch": 0.9302112029384757, "grad_norm": 1.9676135778427124, "learning_rate": 1.7425965867176512e-05, "loss": 0.0611, "step": 1013 }, { "epoch": 0.931129476584022, "grad_norm": 0.9018652439117432, "learning_rate": 1.7418807442471732e-05, "loss": 0.1564, "step": 1014 }, { "epoch": 0.9320477502295684, "grad_norm": 1.9369357824325562, "learning_rate": 1.7411640552439646e-05, "loss": 0.1675, "step": 1015 }, { "epoch": 0.9329660238751147, "grad_norm": 1.5509675741195679, "learning_rate": 1.740446520525814e-05, "loss": 0.1971, "step": 1016 }, { "epoch": 0.9338842975206612, "grad_norm": 0.7508147358894348, "learning_rate": 1.7397281409114737e-05, "loss": 0.0198, "step": 1017 }, { "epoch": 0.9348025711662076, "grad_norm": 1.7474889755249023, "learning_rate": 1.7390089172206594e-05, "loss": 0.1121, "step": 1018 }, { "epoch": 0.9357208448117539, "grad_norm": 0.102937713265419, "learning_rate": 1.7382888502740512e-05, "loss": 0.0035, "step": 1019 }, { "epoch": 0.9366391184573003, "grad_norm": 0.5365200638771057, "learning_rate": 1.7375679408932907e-05, "loss": 0.0197, "step": 1020 }, { "epoch": 0.9375573921028466, "grad_norm": 1.296120524406433, "learning_rate": 1.736846189900981e-05, "loss": 0.095, "step": 1021 }, { "epoch": 0.938475665748393, "grad_norm": 1.2771378755569458, "learning_rate": 1.7361235981206857e-05, "loss": 0.1164, "step": 1022 }, { "epoch": 0.9393939393939394, "grad_norm": 2.9245638847351074, "learning_rate": 1.7354001663769278e-05, "loss": 0.1966, "step": 1023 }, { "epoch": 0.9403122130394858, "grad_norm": 0.6236991882324219, "learning_rate": 1.734675895495188e-05, "loss": 0.0294, "step": 1024 }, { "epoch": 0.9412304866850322, "grad_norm": 0.834666907787323, "learning_rate": 1.7339507863019056e-05, "loss": 0.0444, "step": 1025 }, { "epoch": 0.9421487603305785, "grad_norm": 0.1769048273563385, "learning_rate": 1.733224839624475e-05, "loss": 0.0055, "step": 1026 }, { "epoch": 0.9430670339761249, "grad_norm": 1.243240237236023, "learning_rate": 1.7324980562912485e-05, "loss": 0.1297, "step": 1027 }, { "epoch": 0.9439853076216712, "grad_norm": 0.13904732465744019, "learning_rate": 1.731770437131531e-05, "loss": 0.0043, "step": 1028 }, { "epoch": 0.9449035812672176, "grad_norm": 1.2471349239349365, "learning_rate": 1.7310419829755818e-05, "loss": 0.0656, "step": 1029 }, { "epoch": 0.945821854912764, "grad_norm": 2.5635924339294434, "learning_rate": 1.730312694654613e-05, "loss": 0.165, "step": 1030 }, { "epoch": 0.9467401285583104, "grad_norm": 0.4595795273780823, "learning_rate": 1.7295825730007888e-05, "loss": 0.0165, "step": 1031 }, { "epoch": 0.9476584022038568, "grad_norm": 3.341477155685425, "learning_rate": 1.7288516188472243e-05, "loss": 0.4276, "step": 1032 }, { "epoch": 0.9485766758494031, "grad_norm": 0.34425023198127747, "learning_rate": 1.7281198330279835e-05, "loss": 0.0135, "step": 1033 }, { "epoch": 0.9494949494949495, "grad_norm": 1.1570574045181274, "learning_rate": 1.727387216378081e-05, "loss": 0.1396, "step": 1034 }, { "epoch": 0.9504132231404959, "grad_norm": 2.7273833751678467, "learning_rate": 1.7266537697334788e-05, "loss": 0.0921, "step": 1035 }, { "epoch": 0.9513314967860422, "grad_norm": 1.5292619466781616, "learning_rate": 1.7259194939310844e-05, "loss": 0.0436, "step": 1036 }, { "epoch": 0.9522497704315886, "grad_norm": 0.7050167918205261, "learning_rate": 1.7251843898087544e-05, "loss": 0.0346, "step": 1037 }, { "epoch": 0.953168044077135, "grad_norm": 1.5786025524139404, "learning_rate": 1.724448458205288e-05, "loss": 0.0597, "step": 1038 }, { "epoch": 0.9540863177226814, "grad_norm": 1.9303077459335327, "learning_rate": 1.7237116999604295e-05, "loss": 0.1406, "step": 1039 }, { "epoch": 0.9550045913682277, "grad_norm": 1.8483085632324219, "learning_rate": 1.7229741159148676e-05, "loss": 0.195, "step": 1040 }, { "epoch": 0.9559228650137741, "grad_norm": 1.2204796075820923, "learning_rate": 1.722235706910231e-05, "loss": 0.0535, "step": 1041 }, { "epoch": 0.9568411386593205, "grad_norm": 0.30978772044181824, "learning_rate": 1.7214964737890918e-05, "loss": 0.0087, "step": 1042 }, { "epoch": 0.9577594123048668, "grad_norm": 1.3891596794128418, "learning_rate": 1.7207564173949613e-05, "loss": 0.1109, "step": 1043 }, { "epoch": 0.9586776859504132, "grad_norm": 0.5974666476249695, "learning_rate": 1.7200155385722902e-05, "loss": 0.013, "step": 1044 }, { "epoch": 0.9595959595959596, "grad_norm": 0.8836641311645508, "learning_rate": 1.7192738381664687e-05, "loss": 0.0244, "step": 1045 }, { "epoch": 0.960514233241506, "grad_norm": 0.16914936900138855, "learning_rate": 1.7185313170238232e-05, "loss": 0.0065, "step": 1046 }, { "epoch": 0.9614325068870524, "grad_norm": 1.628206491470337, "learning_rate": 1.7177879759916174e-05, "loss": 0.0693, "step": 1047 }, { "epoch": 0.9623507805325987, "grad_norm": 0.7997005581855774, "learning_rate": 1.71704381591805e-05, "loss": 0.0714, "step": 1048 }, { "epoch": 0.9632690541781451, "grad_norm": 1.2557110786437988, "learning_rate": 1.7162988376522552e-05, "loss": 0.109, "step": 1049 }, { "epoch": 0.9641873278236914, "grad_norm": 0.9064063429832458, "learning_rate": 1.7155530420442995e-05, "loss": 0.1016, "step": 1050 }, { "epoch": 0.9651056014692379, "grad_norm": 1.754573106765747, "learning_rate": 1.7148064299451833e-05, "loss": 0.1447, "step": 1051 }, { "epoch": 0.9660238751147842, "grad_norm": 1.4491684436798096, "learning_rate": 1.714059002206838e-05, "loss": 0.1627, "step": 1052 }, { "epoch": 0.9669421487603306, "grad_norm": 0.9786409139633179, "learning_rate": 1.7133107596821254e-05, "loss": 0.0265, "step": 1053 }, { "epoch": 0.967860422405877, "grad_norm": 0.42748063802719116, "learning_rate": 1.7125617032248375e-05, "loss": 0.0102, "step": 1054 }, { "epoch": 0.9687786960514233, "grad_norm": 1.7376692295074463, "learning_rate": 1.7118118336896957e-05, "loss": 0.105, "step": 1055 }, { "epoch": 0.9696969696969697, "grad_norm": 2.076237916946411, "learning_rate": 1.7110611519323474e-05, "loss": 0.0988, "step": 1056 }, { "epoch": 0.970615243342516, "grad_norm": 1.0685756206512451, "learning_rate": 1.7103096588093686e-05, "loss": 0.0586, "step": 1057 }, { "epoch": 0.9715335169880625, "grad_norm": 1.4383774995803833, "learning_rate": 1.70955735517826e-05, "loss": 0.2064, "step": 1058 }, { "epoch": 0.9724517906336089, "grad_norm": 1.2383801937103271, "learning_rate": 1.7088042418974478e-05, "loss": 0.0622, "step": 1059 }, { "epoch": 0.9733700642791552, "grad_norm": 3.276230573654175, "learning_rate": 1.7080503198262815e-05, "loss": 0.1565, "step": 1060 }, { "epoch": 0.9742883379247016, "grad_norm": 1.1337790489196777, "learning_rate": 1.7072955898250337e-05, "loss": 0.0693, "step": 1061 }, { "epoch": 0.9752066115702479, "grad_norm": 0.6572016477584839, "learning_rate": 1.7065400527548993e-05, "loss": 0.0275, "step": 1062 }, { "epoch": 0.9761248852157943, "grad_norm": 0.437177449464798, "learning_rate": 1.7057837094779933e-05, "loss": 0.0099, "step": 1063 }, { "epoch": 0.9770431588613406, "grad_norm": 0.44774582982063293, "learning_rate": 1.7050265608573514e-05, "loss": 0.0218, "step": 1064 }, { "epoch": 0.977961432506887, "grad_norm": 1.1672191619873047, "learning_rate": 1.704268607756928e-05, "loss": 0.0528, "step": 1065 }, { "epoch": 0.9788797061524335, "grad_norm": 0.7144413590431213, "learning_rate": 1.7035098510415952e-05, "loss": 0.0155, "step": 1066 }, { "epoch": 0.9797979797979798, "grad_norm": 0.5753883123397827, "learning_rate": 1.702750291577142e-05, "loss": 0.0367, "step": 1067 }, { "epoch": 0.9807162534435262, "grad_norm": 0.7442868947982788, "learning_rate": 1.7019899302302742e-05, "loss": 0.0599, "step": 1068 }, { "epoch": 0.9816345270890725, "grad_norm": 0.21477794647216797, "learning_rate": 1.701228767868612e-05, "loss": 0.0051, "step": 1069 }, { "epoch": 0.9825528007346189, "grad_norm": 1.7491014003753662, "learning_rate": 1.700466805360689e-05, "loss": 0.1679, "step": 1070 }, { "epoch": 0.9834710743801653, "grad_norm": 0.8826123476028442, "learning_rate": 1.6997040435759538e-05, "loss": 0.0233, "step": 1071 }, { "epoch": 0.9843893480257117, "grad_norm": 1.1712584495544434, "learning_rate": 1.6989404833847644e-05, "loss": 0.103, "step": 1072 }, { "epoch": 0.9853076216712581, "grad_norm": 1.4755264520645142, "learning_rate": 1.698176125658392e-05, "loss": 0.1515, "step": 1073 }, { "epoch": 0.9862258953168044, "grad_norm": 0.550084114074707, "learning_rate": 1.6974109712690163e-05, "loss": 0.0266, "step": 1074 }, { "epoch": 0.9871441689623508, "grad_norm": 0.18700267374515533, "learning_rate": 1.6966450210897273e-05, "loss": 0.0063, "step": 1075 }, { "epoch": 0.9880624426078971, "grad_norm": 0.8584281802177429, "learning_rate": 1.6958782759945223e-05, "loss": 0.0266, "step": 1076 }, { "epoch": 0.9889807162534435, "grad_norm": 1.518373966217041, "learning_rate": 1.6951107368583057e-05, "loss": 0.0966, "step": 1077 }, { "epoch": 0.98989898989899, "grad_norm": 1.2497800588607788, "learning_rate": 1.6943424045568882e-05, "loss": 0.0224, "step": 1078 }, { "epoch": 0.9908172635445363, "grad_norm": 0.3272691071033478, "learning_rate": 1.6935732799669855e-05, "loss": 0.0115, "step": 1079 }, { "epoch": 0.9917355371900827, "grad_norm": 0.3955132067203522, "learning_rate": 1.692803363966217e-05, "loss": 0.0138, "step": 1080 }, { "epoch": 0.992653810835629, "grad_norm": 0.7558044791221619, "learning_rate": 1.6920326574331055e-05, "loss": 0.0176, "step": 1081 }, { "epoch": 0.9935720844811754, "grad_norm": 0.7923538684844971, "learning_rate": 1.691261161247076e-05, "loss": 0.0311, "step": 1082 }, { "epoch": 0.9944903581267218, "grad_norm": 0.8200954794883728, "learning_rate": 1.6904888762884542e-05, "loss": 0.0481, "step": 1083 }, { "epoch": 0.9954086317722681, "grad_norm": 1.6505489349365234, "learning_rate": 1.689715803438466e-05, "loss": 0.0958, "step": 1084 }, { "epoch": 0.9963269054178145, "grad_norm": 0.12002972513437271, "learning_rate": 1.688941943579236e-05, "loss": 0.0036, "step": 1085 }, { "epoch": 0.9972451790633609, "grad_norm": 0.8169644474983215, "learning_rate": 1.6881672975937875e-05, "loss": 0.0477, "step": 1086 }, { "epoch": 0.9981634527089073, "grad_norm": 0.7680754661560059, "learning_rate": 1.68739186636604e-05, "loss": 0.0638, "step": 1087 }, { "epoch": 0.9990817263544536, "grad_norm": 1.7113102674484253, "learning_rate": 1.68661565078081e-05, "loss": 0.0739, "step": 1088 }, { "epoch": 1.0, "grad_norm": 0.9866504073143005, "learning_rate": 1.685838651723808e-05, "loss": 0.0397, "step": 1089 }, { "epoch": 1.0009182736455464, "grad_norm": 0.6785283088684082, "learning_rate": 1.6850608700816395e-05, "loss": 0.02, "step": 1090 }, { "epoch": 1.0018365472910928, "grad_norm": 0.10186060518026352, "learning_rate": 1.684282306741802e-05, "loss": 0.0028, "step": 1091 }, { "epoch": 1.002754820936639, "grad_norm": 0.7490137815475464, "learning_rate": 1.683502962592685e-05, "loss": 0.0311, "step": 1092 }, { "epoch": 1.0036730945821855, "grad_norm": 1.7826306819915771, "learning_rate": 1.6827228385235697e-05, "loss": 0.065, "step": 1093 }, { "epoch": 1.0045913682277319, "grad_norm": 0.6346295475959778, "learning_rate": 1.6819419354246274e-05, "loss": 0.0245, "step": 1094 }, { "epoch": 1.0055096418732783, "grad_norm": 0.3759898841381073, "learning_rate": 1.681160254186917e-05, "loss": 0.0108, "step": 1095 }, { "epoch": 1.0064279155188247, "grad_norm": 0.23324760794639587, "learning_rate": 1.680377795702387e-05, "loss": 0.0048, "step": 1096 }, { "epoch": 1.007346189164371, "grad_norm": 0.11575739830732346, "learning_rate": 1.679594560863871e-05, "loss": 0.0024, "step": 1097 }, { "epoch": 1.0082644628099173, "grad_norm": 0.14304469525814056, "learning_rate": 1.6788105505650893e-05, "loss": 0.0026, "step": 1098 }, { "epoch": 1.0091827364554637, "grad_norm": 0.23645029962062836, "learning_rate": 1.6780257657006477e-05, "loss": 0.005, "step": 1099 }, { "epoch": 1.0101010101010102, "grad_norm": 1.5328387022018433, "learning_rate": 1.6772402071660354e-05, "loss": 0.069, "step": 1100 }, { "epoch": 1.0110192837465566, "grad_norm": 1.1664130687713623, "learning_rate": 1.6764538758576235e-05, "loss": 0.0514, "step": 1101 }, { "epoch": 1.0119375573921028, "grad_norm": 0.5177730321884155, "learning_rate": 1.675666772672666e-05, "loss": 0.011, "step": 1102 }, { "epoch": 1.0128558310376492, "grad_norm": 1.0807702541351318, "learning_rate": 1.6748788985092977e-05, "loss": 0.0638, "step": 1103 }, { "epoch": 1.0137741046831956, "grad_norm": 1.652345895767212, "learning_rate": 1.6740902542665322e-05, "loss": 0.141, "step": 1104 }, { "epoch": 1.014692378328742, "grad_norm": 1.0169775485992432, "learning_rate": 1.673300840844262e-05, "loss": 0.0771, "step": 1105 }, { "epoch": 1.0156106519742882, "grad_norm": 0.4996049106121063, "learning_rate": 1.6725106591432583e-05, "loss": 0.0218, "step": 1106 }, { "epoch": 1.0165289256198347, "grad_norm": 1.0688492059707642, "learning_rate": 1.671719710065168e-05, "loss": 0.0267, "step": 1107 }, { "epoch": 1.017447199265381, "grad_norm": 1.4121403694152832, "learning_rate": 1.670927994512514e-05, "loss": 0.1039, "step": 1108 }, { "epoch": 1.0183654729109275, "grad_norm": 1.1169379949569702, "learning_rate": 1.6701355133886936e-05, "loss": 0.0921, "step": 1109 }, { "epoch": 1.019283746556474, "grad_norm": 0.9637802243232727, "learning_rate": 1.669342267597978e-05, "loss": 0.0145, "step": 1110 }, { "epoch": 1.02020202020202, "grad_norm": 2.290038585662842, "learning_rate": 1.66854825804551e-05, "loss": 0.3003, "step": 1111 }, { "epoch": 1.0211202938475665, "grad_norm": 1.4926031827926636, "learning_rate": 1.6677534856373055e-05, "loss": 0.1642, "step": 1112 }, { "epoch": 1.022038567493113, "grad_norm": 0.8571377992630005, "learning_rate": 1.6669579512802494e-05, "loss": 0.0995, "step": 1113 }, { "epoch": 1.0229568411386594, "grad_norm": 1.1651657819747925, "learning_rate": 1.6661616558820965e-05, "loss": 0.0441, "step": 1114 }, { "epoch": 1.0238751147842058, "grad_norm": 0.14629371464252472, "learning_rate": 1.6653646003514704e-05, "loss": 0.0033, "step": 1115 }, { "epoch": 1.024793388429752, "grad_norm": 0.0402643084526062, "learning_rate": 1.6645667855978613e-05, "loss": 0.0011, "step": 1116 }, { "epoch": 1.0257116620752984, "grad_norm": 0.9911729097366333, "learning_rate": 1.6637682125316267e-05, "loss": 0.0721, "step": 1117 }, { "epoch": 1.0266299357208448, "grad_norm": 1.84031081199646, "learning_rate": 1.6629688820639887e-05, "loss": 0.0917, "step": 1118 }, { "epoch": 1.0275482093663912, "grad_norm": 0.624131441116333, "learning_rate": 1.6621687951070336e-05, "loss": 0.0206, "step": 1119 }, { "epoch": 1.0284664830119377, "grad_norm": 1.3018771409988403, "learning_rate": 1.6613679525737113e-05, "loss": 0.1795, "step": 1120 }, { "epoch": 1.0293847566574839, "grad_norm": 2.6654350757598877, "learning_rate": 1.6605663553778335e-05, "loss": 0.1874, "step": 1121 }, { "epoch": 1.0303030303030303, "grad_norm": 1.053051233291626, "learning_rate": 1.6597640044340733e-05, "loss": 0.0867, "step": 1122 }, { "epoch": 1.0312213039485767, "grad_norm": 0.44362232089042664, "learning_rate": 1.6589609006579636e-05, "loss": 0.01, "step": 1123 }, { "epoch": 1.0321395775941231, "grad_norm": 2.272336959838867, "learning_rate": 1.6581570449658965e-05, "loss": 0.0607, "step": 1124 }, { "epoch": 1.0330578512396693, "grad_norm": 1.1743338108062744, "learning_rate": 1.657352438275122e-05, "loss": 0.0541, "step": 1125 }, { "epoch": 1.0339761248852157, "grad_norm": 3.6155831813812256, "learning_rate": 1.6565470815037474e-05, "loss": 0.3623, "step": 1126 }, { "epoch": 1.0348943985307621, "grad_norm": 0.3061164617538452, "learning_rate": 1.6557409755707352e-05, "loss": 0.0092, "step": 1127 }, { "epoch": 1.0358126721763086, "grad_norm": 1.415618896484375, "learning_rate": 1.6549341213959035e-05, "loss": 0.0877, "step": 1128 }, { "epoch": 1.036730945821855, "grad_norm": 0.3136952817440033, "learning_rate": 1.6541265198999235e-05, "loss": 0.0076, "step": 1129 }, { "epoch": 1.0376492194674012, "grad_norm": 1.9680007696151733, "learning_rate": 1.6533181720043194e-05, "loss": 0.0543, "step": 1130 }, { "epoch": 1.0385674931129476, "grad_norm": 2.6887810230255127, "learning_rate": 1.6525090786314673e-05, "loss": 0.1818, "step": 1131 }, { "epoch": 1.039485766758494, "grad_norm": 0.2735365033149719, "learning_rate": 1.651699240704594e-05, "loss": 0.0059, "step": 1132 }, { "epoch": 1.0404040404040404, "grad_norm": 0.5878204107284546, "learning_rate": 1.650888659147775e-05, "loss": 0.0143, "step": 1133 }, { "epoch": 1.0413223140495869, "grad_norm": 0.30697107315063477, "learning_rate": 1.6500773348859358e-05, "loss": 0.0084, "step": 1134 }, { "epoch": 1.042240587695133, "grad_norm": 1.6768163442611694, "learning_rate": 1.6492652688448482e-05, "loss": 0.202, "step": 1135 }, { "epoch": 1.0431588613406795, "grad_norm": 1.6352733373641968, "learning_rate": 1.6484524619511305e-05, "loss": 0.1264, "step": 1136 }, { "epoch": 1.044077134986226, "grad_norm": 0.4612835645675659, "learning_rate": 1.647638915132247e-05, "loss": 0.0085, "step": 1137 }, { "epoch": 1.0449954086317723, "grad_norm": 1.404921293258667, "learning_rate": 1.6468246293165057e-05, "loss": 0.075, "step": 1138 }, { "epoch": 1.0459136822773187, "grad_norm": 0.9463897943496704, "learning_rate": 1.6460096054330588e-05, "loss": 0.0601, "step": 1139 }, { "epoch": 1.046831955922865, "grad_norm": 0.9130315184593201, "learning_rate": 1.6451938444118986e-05, "loss": 0.0304, "step": 1140 }, { "epoch": 1.0477502295684114, "grad_norm": 0.2698214650154114, "learning_rate": 1.6443773471838612e-05, "loss": 0.0062, "step": 1141 }, { "epoch": 1.0486685032139578, "grad_norm": 0.6721444129943848, "learning_rate": 1.643560114680621e-05, "loss": 0.0141, "step": 1142 }, { "epoch": 1.0495867768595042, "grad_norm": 1.9571260213851929, "learning_rate": 1.6427421478346914e-05, "loss": 0.0917, "step": 1143 }, { "epoch": 1.0505050505050506, "grad_norm": 0.9289267063140869, "learning_rate": 1.641923447579425e-05, "loss": 0.0361, "step": 1144 }, { "epoch": 1.0514233241505968, "grad_norm": 0.7171139717102051, "learning_rate": 1.6411040148490095e-05, "loss": 0.022, "step": 1145 }, { "epoch": 1.0523415977961432, "grad_norm": 1.5529870986938477, "learning_rate": 1.6402838505784703e-05, "loss": 0.1437, "step": 1146 }, { "epoch": 1.0532598714416896, "grad_norm": 0.09928872436285019, "learning_rate": 1.6394629557036655e-05, "loss": 0.0031, "step": 1147 }, { "epoch": 1.054178145087236, "grad_norm": 1.9085804224014282, "learning_rate": 1.638641331161289e-05, "loss": 0.0976, "step": 1148 }, { "epoch": 1.0550964187327825, "grad_norm": 2.495861053466797, "learning_rate": 1.6378189778888657e-05, "loss": 0.1463, "step": 1149 }, { "epoch": 1.0560146923783287, "grad_norm": 0.38919195532798767, "learning_rate": 1.6369958968247525e-05, "loss": 0.0078, "step": 1150 }, { "epoch": 1.056932966023875, "grad_norm": 0.7475545406341553, "learning_rate": 1.636172088908137e-05, "loss": 0.019, "step": 1151 }, { "epoch": 1.0578512396694215, "grad_norm": 1.914889931678772, "learning_rate": 1.635347555079036e-05, "loss": 0.0968, "step": 1152 }, { "epoch": 1.058769513314968, "grad_norm": 0.19092686474323273, "learning_rate": 1.6345222962782945e-05, "loss": 0.0048, "step": 1153 }, { "epoch": 1.0596877869605141, "grad_norm": 1.610000491142273, "learning_rate": 1.633696313447585e-05, "loss": 0.0785, "step": 1154 }, { "epoch": 1.0606060606060606, "grad_norm": 0.12122561037540436, "learning_rate": 1.632869607529406e-05, "loss": 0.0038, "step": 1155 }, { "epoch": 1.061524334251607, "grad_norm": 0.9204964637756348, "learning_rate": 1.6320421794670805e-05, "loss": 0.0357, "step": 1156 }, { "epoch": 1.0624426078971534, "grad_norm": 1.0925873517990112, "learning_rate": 1.631214030204757e-05, "loss": 0.0989, "step": 1157 }, { "epoch": 1.0633608815426998, "grad_norm": 0.9016194343566895, "learning_rate": 1.6303851606874056e-05, "loss": 0.0326, "step": 1158 }, { "epoch": 1.064279155188246, "grad_norm": 1.3771699666976929, "learning_rate": 1.629555571860819e-05, "loss": 0.0551, "step": 1159 }, { "epoch": 1.0651974288337924, "grad_norm": 0.4862889051437378, "learning_rate": 1.6287252646716104e-05, "loss": 0.0061, "step": 1160 }, { "epoch": 1.0661157024793388, "grad_norm": 1.388861894607544, "learning_rate": 1.627894240067212e-05, "loss": 0.1034, "step": 1161 }, { "epoch": 1.0670339761248853, "grad_norm": 1.5621111392974854, "learning_rate": 1.6270624989958758e-05, "loss": 0.1245, "step": 1162 }, { "epoch": 1.0679522497704317, "grad_norm": 2.1273229122161865, "learning_rate": 1.626230042406671e-05, "loss": 0.1555, "step": 1163 }, { "epoch": 1.0688705234159779, "grad_norm": 0.19466054439544678, "learning_rate": 1.6253968712494828e-05, "loss": 0.006, "step": 1164 }, { "epoch": 1.0697887970615243, "grad_norm": 1.5801881551742554, "learning_rate": 1.624562986475012e-05, "loss": 0.158, "step": 1165 }, { "epoch": 1.0707070707070707, "grad_norm": 1.1622947454452515, "learning_rate": 1.623728389034774e-05, "loss": 0.0359, "step": 1166 }, { "epoch": 1.0716253443526171, "grad_norm": 0.5168084502220154, "learning_rate": 1.622893079881097e-05, "loss": 0.0094, "step": 1167 }, { "epoch": 1.0725436179981636, "grad_norm": 0.9405928254127502, "learning_rate": 1.6220570599671215e-05, "loss": 0.0593, "step": 1168 }, { "epoch": 1.0734618916437098, "grad_norm": 0.8439944386482239, "learning_rate": 1.621220330246799e-05, "loss": 0.0523, "step": 1169 }, { "epoch": 1.0743801652892562, "grad_norm": 1.019848346710205, "learning_rate": 1.6203828916748904e-05, "loss": 0.0179, "step": 1170 }, { "epoch": 1.0752984389348026, "grad_norm": 1.3474767208099365, "learning_rate": 1.6195447452069666e-05, "loss": 0.0345, "step": 1171 }, { "epoch": 1.076216712580349, "grad_norm": 1.9247620105743408, "learning_rate": 1.6187058917994052e-05, "loss": 0.1061, "step": 1172 }, { "epoch": 1.0771349862258952, "grad_norm": 0.6798341274261475, "learning_rate": 1.617866332409391e-05, "loss": 0.045, "step": 1173 }, { "epoch": 1.0780532598714416, "grad_norm": 0.27019450068473816, "learning_rate": 1.6170260679949143e-05, "loss": 0.0053, "step": 1174 }, { "epoch": 1.078971533516988, "grad_norm": 2.53371524810791, "learning_rate": 1.6161850995147695e-05, "loss": 0.2548, "step": 1175 }, { "epoch": 1.0798898071625345, "grad_norm": 1.0742439031600952, "learning_rate": 1.615343427928555e-05, "loss": 0.0551, "step": 1176 }, { "epoch": 1.0808080808080809, "grad_norm": 1.2951029539108276, "learning_rate": 1.6145010541966712e-05, "loss": 0.0488, "step": 1177 }, { "epoch": 1.081726354453627, "grad_norm": 1.774095892906189, "learning_rate": 1.6136579792803194e-05, "loss": 0.0442, "step": 1178 }, { "epoch": 1.0826446280991735, "grad_norm": 1.557079792022705, "learning_rate": 1.6128142041415012e-05, "loss": 0.0697, "step": 1179 }, { "epoch": 1.08356290174472, "grad_norm": 0.026672707870602608, "learning_rate": 1.611969729743018e-05, "loss": 0.0009, "step": 1180 }, { "epoch": 1.0844811753902663, "grad_norm": 1.4366753101348877, "learning_rate": 1.611124557048467e-05, "loss": 0.202, "step": 1181 }, { "epoch": 1.0853994490358128, "grad_norm": 0.21122939884662628, "learning_rate": 1.6102786870222456e-05, "loss": 0.0061, "step": 1182 }, { "epoch": 1.086317722681359, "grad_norm": 1.1685709953308105, "learning_rate": 1.6094321206295427e-05, "loss": 0.1023, "step": 1183 }, { "epoch": 1.0872359963269054, "grad_norm": 0.5266413688659668, "learning_rate": 1.6085848588363453e-05, "loss": 0.0197, "step": 1184 }, { "epoch": 1.0881542699724518, "grad_norm": 0.5491708517074585, "learning_rate": 1.607736902609432e-05, "loss": 0.0158, "step": 1185 }, { "epoch": 1.0890725436179982, "grad_norm": 1.3290923833847046, "learning_rate": 1.6068882529163744e-05, "loss": 0.0354, "step": 1186 }, { "epoch": 1.0899908172635446, "grad_norm": 0.26192110776901245, "learning_rate": 1.6060389107255357e-05, "loss": 0.0075, "step": 1187 }, { "epoch": 1.0909090909090908, "grad_norm": 1.8586928844451904, "learning_rate": 1.6051888770060683e-05, "loss": 0.0882, "step": 1188 }, { "epoch": 1.0918273645546372, "grad_norm": 0.45045021176338196, "learning_rate": 1.604338152727915e-05, "loss": 0.0192, "step": 1189 }, { "epoch": 1.0927456382001837, "grad_norm": 0.783062219619751, "learning_rate": 1.6034867388618053e-05, "loss": 0.0326, "step": 1190 }, { "epoch": 1.09366391184573, "grad_norm": 0.5641867518424988, "learning_rate": 1.6026346363792565e-05, "loss": 0.018, "step": 1191 }, { "epoch": 1.0945821854912765, "grad_norm": 1.6470648050308228, "learning_rate": 1.6017818462525712e-05, "loss": 0.0692, "step": 1192 }, { "epoch": 1.0955004591368227, "grad_norm": 0.44569963216781616, "learning_rate": 1.6009283694548365e-05, "loss": 0.033, "step": 1193 }, { "epoch": 1.0964187327823691, "grad_norm": 1.186705231666565, "learning_rate": 1.6000742069599236e-05, "loss": 0.0876, "step": 1194 }, { "epoch": 1.0973370064279155, "grad_norm": 0.4294511675834656, "learning_rate": 1.5992193597424857e-05, "loss": 0.02, "step": 1195 }, { "epoch": 1.098255280073462, "grad_norm": 1.8398621082305908, "learning_rate": 1.598363828777957e-05, "loss": 0.1087, "step": 1196 }, { "epoch": 1.0991735537190084, "grad_norm": 1.7107799053192139, "learning_rate": 1.597507615042553e-05, "loss": 0.0926, "step": 1197 }, { "epoch": 1.1000918273645546, "grad_norm": 0.19297198951244354, "learning_rate": 1.596650719513267e-05, "loss": 0.0067, "step": 1198 }, { "epoch": 1.101010101010101, "grad_norm": 0.9415361881256104, "learning_rate": 1.595793143167871e-05, "loss": 0.0164, "step": 1199 }, { "epoch": 1.1019283746556474, "grad_norm": 1.2781774997711182, "learning_rate": 1.594934886984914e-05, "loss": 0.1277, "step": 1200 }, { "epoch": 1.1028466483011938, "grad_norm": 1.749854326248169, "learning_rate": 1.59407595194372e-05, "loss": 0.0802, "step": 1201 }, { "epoch": 1.10376492194674, "grad_norm": 0.5708281993865967, "learning_rate": 1.5932163390243886e-05, "loss": 0.0165, "step": 1202 }, { "epoch": 1.1046831955922864, "grad_norm": 2.0087082386016846, "learning_rate": 1.5923560492077917e-05, "loss": 0.1139, "step": 1203 }, { "epoch": 1.1056014692378329, "grad_norm": 1.694793939590454, "learning_rate": 1.5914950834755753e-05, "loss": 0.0965, "step": 1204 }, { "epoch": 1.1065197428833793, "grad_norm": 0.7105377912521362, "learning_rate": 1.5906334428101544e-05, "loss": 0.0204, "step": 1205 }, { "epoch": 1.1074380165289257, "grad_norm": 1.3004225492477417, "learning_rate": 1.589771128194716e-05, "loss": 0.1212, "step": 1206 }, { "epoch": 1.108356290174472, "grad_norm": 1.784205675125122, "learning_rate": 1.5889081406132154e-05, "loss": 0.2487, "step": 1207 }, { "epoch": 1.1092745638200183, "grad_norm": 1.7664217948913574, "learning_rate": 1.5880444810503755e-05, "loss": 0.0467, "step": 1208 }, { "epoch": 1.1101928374655647, "grad_norm": 0.7393720746040344, "learning_rate": 1.587180150491687e-05, "loss": 0.0341, "step": 1209 }, { "epoch": 1.1111111111111112, "grad_norm": 0.9522415995597839, "learning_rate": 1.5863151499234053e-05, "loss": 0.0566, "step": 1210 }, { "epoch": 1.1120293847566576, "grad_norm": 2.1601030826568604, "learning_rate": 1.5854494803325504e-05, "loss": 0.0948, "step": 1211 }, { "epoch": 1.1129476584022038, "grad_norm": 0.3965552747249603, "learning_rate": 1.5845831427069063e-05, "loss": 0.0108, "step": 1212 }, { "epoch": 1.1138659320477502, "grad_norm": 0.5010169744491577, "learning_rate": 1.5837161380350185e-05, "loss": 0.014, "step": 1213 }, { "epoch": 1.1147842056932966, "grad_norm": 0.8076397776603699, "learning_rate": 1.5828484673061943e-05, "loss": 0.0295, "step": 1214 }, { "epoch": 1.115702479338843, "grad_norm": 1.5302069187164307, "learning_rate": 1.5819801315105004e-05, "loss": 0.0599, "step": 1215 }, { "epoch": 1.1166207529843892, "grad_norm": 1.1491806507110596, "learning_rate": 1.5811111316387635e-05, "loss": 0.0636, "step": 1216 }, { "epoch": 1.1175390266299357, "grad_norm": 0.821139931678772, "learning_rate": 1.5802414686825662e-05, "loss": 0.0543, "step": 1217 }, { "epoch": 1.118457300275482, "grad_norm": 1.384337067604065, "learning_rate": 1.5793711436342495e-05, "loss": 0.0954, "step": 1218 }, { "epoch": 1.1193755739210285, "grad_norm": 0.13052840530872345, "learning_rate": 1.5785001574869093e-05, "loss": 0.0057, "step": 1219 }, { "epoch": 1.120293847566575, "grad_norm": 0.6518274545669556, "learning_rate": 1.577628511234395e-05, "loss": 0.033, "step": 1220 }, { "epoch": 1.121212121212121, "grad_norm": 2.414296865463257, "learning_rate": 1.576756205871311e-05, "loss": 0.1439, "step": 1221 }, { "epoch": 1.1221303948576675, "grad_norm": 0.5947332978248596, "learning_rate": 1.5758832423930114e-05, "loss": 0.0192, "step": 1222 }, { "epoch": 1.123048668503214, "grad_norm": 1.9316643476486206, "learning_rate": 1.575009621795604e-05, "loss": 0.0588, "step": 1223 }, { "epoch": 1.1239669421487604, "grad_norm": 2.8702473640441895, "learning_rate": 1.574135345075944e-05, "loss": 0.2071, "step": 1224 }, { "epoch": 1.1248852157943068, "grad_norm": 0.5003494024276733, "learning_rate": 1.573260413231637e-05, "loss": 0.0125, "step": 1225 }, { "epoch": 1.125803489439853, "grad_norm": 1.8631782531738281, "learning_rate": 1.572384827261035e-05, "loss": 0.1275, "step": 1226 }, { "epoch": 1.1267217630853994, "grad_norm": 0.324327677488327, "learning_rate": 1.5715085881632366e-05, "loss": 0.0084, "step": 1227 }, { "epoch": 1.1276400367309458, "grad_norm": 0.40160006284713745, "learning_rate": 1.5706316969380866e-05, "loss": 0.0127, "step": 1228 }, { "epoch": 1.1285583103764922, "grad_norm": 0.5431967973709106, "learning_rate": 1.5697541545861725e-05, "loss": 0.0192, "step": 1229 }, { "epoch": 1.1294765840220387, "grad_norm": 0.31535184383392334, "learning_rate": 1.568875962108826e-05, "loss": 0.0064, "step": 1230 }, { "epoch": 1.1303948576675849, "grad_norm": 0.4392637312412262, "learning_rate": 1.5679971205081198e-05, "loss": 0.007, "step": 1231 }, { "epoch": 1.1313131313131313, "grad_norm": 1.1030960083007812, "learning_rate": 1.5671176307868678e-05, "loss": 0.0775, "step": 1232 }, { "epoch": 1.1322314049586777, "grad_norm": 0.3381805717945099, "learning_rate": 1.566237493948623e-05, "loss": 0.0159, "step": 1233 }, { "epoch": 1.133149678604224, "grad_norm": 0.8241015672683716, "learning_rate": 1.5653567109976774e-05, "loss": 0.0172, "step": 1234 }, { "epoch": 1.1340679522497705, "grad_norm": 0.7968249320983887, "learning_rate": 1.5644752829390598e-05, "loss": 0.0476, "step": 1235 }, { "epoch": 1.1349862258953167, "grad_norm": 1.2564327716827393, "learning_rate": 1.563593210778535e-05, "loss": 0.0933, "step": 1236 }, { "epoch": 1.1359044995408631, "grad_norm": 2.4569337368011475, "learning_rate": 1.562710495522603e-05, "loss": 0.2006, "step": 1237 }, { "epoch": 1.1368227731864096, "grad_norm": 0.46071097254753113, "learning_rate": 1.5618271381784983e-05, "loss": 0.0086, "step": 1238 }, { "epoch": 1.137741046831956, "grad_norm": 1.3570448160171509, "learning_rate": 1.560943139754186e-05, "loss": 0.1404, "step": 1239 }, { "epoch": 1.1386593204775024, "grad_norm": 1.4522700309753418, "learning_rate": 1.5600585012583654e-05, "loss": 0.1405, "step": 1240 }, { "epoch": 1.1395775941230486, "grad_norm": 1.0360761880874634, "learning_rate": 1.559173223700464e-05, "loss": 0.0656, "step": 1241 }, { "epoch": 1.140495867768595, "grad_norm": 0.5533316731452942, "learning_rate": 1.5582873080906397e-05, "loss": 0.0152, "step": 1242 }, { "epoch": 1.1414141414141414, "grad_norm": 2.2243471145629883, "learning_rate": 1.557400755439778e-05, "loss": 0.0913, "step": 1243 }, { "epoch": 1.1423324150596879, "grad_norm": 0.17527642846107483, "learning_rate": 1.5565135667594916e-05, "loss": 0.0046, "step": 1244 }, { "epoch": 1.1432506887052343, "grad_norm": 0.9135543704032898, "learning_rate": 1.5556257430621184e-05, "loss": 0.0389, "step": 1245 }, { "epoch": 1.1441689623507805, "grad_norm": 0.6444950103759766, "learning_rate": 1.5547372853607213e-05, "loss": 0.015, "step": 1246 }, { "epoch": 1.145087235996327, "grad_norm": 3.1034069061279297, "learning_rate": 1.553848194669087e-05, "loss": 0.5074, "step": 1247 }, { "epoch": 1.1460055096418733, "grad_norm": 0.9352930784225464, "learning_rate": 1.552958472001724e-05, "loss": 0.0193, "step": 1248 }, { "epoch": 1.1469237832874197, "grad_norm": 1.1065837144851685, "learning_rate": 1.552068118373861e-05, "loss": 0.0202, "step": 1249 }, { "epoch": 1.1478420569329661, "grad_norm": 1.8526664972305298, "learning_rate": 1.5511771348014488e-05, "loss": 0.105, "step": 1250 }, { "epoch": 1.1487603305785123, "grad_norm": 2.553651809692383, "learning_rate": 1.5502855223011555e-05, "loss": 0.1166, "step": 1251 }, { "epoch": 1.1496786042240588, "grad_norm": 0.06300089508295059, "learning_rate": 1.5493932818903673e-05, "loss": 0.0018, "step": 1252 }, { "epoch": 1.1505968778696052, "grad_norm": 0.33692821860313416, "learning_rate": 1.5485004145871866e-05, "loss": 0.0087, "step": 1253 }, { "epoch": 1.1515151515151516, "grad_norm": 0.12767106294631958, "learning_rate": 1.5476069214104312e-05, "loss": 0.0036, "step": 1254 }, { "epoch": 1.1524334251606978, "grad_norm": 1.0627505779266357, "learning_rate": 1.5467128033796336e-05, "loss": 0.0171, "step": 1255 }, { "epoch": 1.1533516988062442, "grad_norm": 0.8958908319473267, "learning_rate": 1.5458180615150386e-05, "loss": 0.0556, "step": 1256 }, { "epoch": 1.1542699724517906, "grad_norm": 1.1682360172271729, "learning_rate": 1.5449226968376033e-05, "loss": 0.068, "step": 1257 }, { "epoch": 1.155188246097337, "grad_norm": 0.39344894886016846, "learning_rate": 1.5440267103689946e-05, "loss": 0.0109, "step": 1258 }, { "epoch": 1.1561065197428835, "grad_norm": 1.6611500978469849, "learning_rate": 1.5431301031315908e-05, "loss": 0.0679, "step": 1259 }, { "epoch": 1.1570247933884297, "grad_norm": 0.0273414496332407, "learning_rate": 1.5422328761484764e-05, "loss": 0.0009, "step": 1260 }, { "epoch": 1.157943067033976, "grad_norm": 0.2257087528705597, "learning_rate": 1.541335030443444e-05, "loss": 0.0048, "step": 1261 }, { "epoch": 1.1588613406795225, "grad_norm": 0.7080309391021729, "learning_rate": 1.5404365670409927e-05, "loss": 0.0306, "step": 1262 }, { "epoch": 1.159779614325069, "grad_norm": 2.19899320602417, "learning_rate": 1.539537486966326e-05, "loss": 0.0629, "step": 1263 }, { "epoch": 1.1606978879706151, "grad_norm": 1.460584044456482, "learning_rate": 1.53863779124535e-05, "loss": 0.056, "step": 1264 }, { "epoch": 1.1616161616161615, "grad_norm": 0.894615888595581, "learning_rate": 1.537737480904675e-05, "loss": 0.0374, "step": 1265 }, { "epoch": 1.162534435261708, "grad_norm": 1.4662567377090454, "learning_rate": 1.536836556971612e-05, "loss": 0.0502, "step": 1266 }, { "epoch": 1.1634527089072544, "grad_norm": 0.6907074451446533, "learning_rate": 1.5359350204741718e-05, "loss": 0.0411, "step": 1267 }, { "epoch": 1.1643709825528008, "grad_norm": 0.9768441319465637, "learning_rate": 1.5350328724410645e-05, "loss": 0.0368, "step": 1268 }, { "epoch": 1.165289256198347, "grad_norm": 0.6083048582077026, "learning_rate": 1.5341301139016976e-05, "loss": 0.0066, "step": 1269 }, { "epoch": 1.1662075298438934, "grad_norm": 0.9275221824645996, "learning_rate": 1.533226745886176e-05, "loss": 0.0121, "step": 1270 }, { "epoch": 1.1671258034894398, "grad_norm": 1.4892922639846802, "learning_rate": 1.5323227694252994e-05, "loss": 0.1637, "step": 1271 }, { "epoch": 1.1680440771349863, "grad_norm": 1.0824357271194458, "learning_rate": 1.5314181855505623e-05, "loss": 0.0728, "step": 1272 }, { "epoch": 1.1689623507805327, "grad_norm": 0.047079432755708694, "learning_rate": 1.530512995294151e-05, "loss": 0.0014, "step": 1273 }, { "epoch": 1.1698806244260789, "grad_norm": 0.06344927102327347, "learning_rate": 1.5296071996889455e-05, "loss": 0.0017, "step": 1274 }, { "epoch": 1.1707988980716253, "grad_norm": 0.09926442056894302, "learning_rate": 1.5287007997685157e-05, "loss": 0.0024, "step": 1275 }, { "epoch": 1.1717171717171717, "grad_norm": 1.4974980354309082, "learning_rate": 1.527793796567121e-05, "loss": 0.0767, "step": 1276 }, { "epoch": 1.1726354453627181, "grad_norm": 2.1709251403808594, "learning_rate": 1.526886191119709e-05, "loss": 0.0268, "step": 1277 }, { "epoch": 1.1735537190082646, "grad_norm": 1.0085610151290894, "learning_rate": 1.5259779844619152e-05, "loss": 0.0549, "step": 1278 }, { "epoch": 1.1744719926538107, "grad_norm": 0.056669313460588455, "learning_rate": 1.5250691776300605e-05, "loss": 0.0014, "step": 1279 }, { "epoch": 1.1753902662993572, "grad_norm": 1.0459644794464111, "learning_rate": 1.5241597716611507e-05, "loss": 0.0922, "step": 1280 }, { "epoch": 1.1763085399449036, "grad_norm": 0.8088529706001282, "learning_rate": 1.5232497675928755e-05, "loss": 0.0238, "step": 1281 }, { "epoch": 1.17722681359045, "grad_norm": 0.9174458384513855, "learning_rate": 1.5223391664636067e-05, "loss": 0.0209, "step": 1282 }, { "epoch": 1.1781450872359964, "grad_norm": 1.0632041692733765, "learning_rate": 1.5214279693123981e-05, "loss": 0.0524, "step": 1283 }, { "epoch": 1.1790633608815426, "grad_norm": 1.2634763717651367, "learning_rate": 1.520516177178983e-05, "loss": 0.081, "step": 1284 }, { "epoch": 1.179981634527089, "grad_norm": 0.11434893310070038, "learning_rate": 1.5196037911037733e-05, "loss": 0.0035, "step": 1285 }, { "epoch": 1.1808999081726355, "grad_norm": 0.11867304891347885, "learning_rate": 1.5186908121278596e-05, "loss": 0.0033, "step": 1286 }, { "epoch": 1.1818181818181819, "grad_norm": 0.0740370973944664, "learning_rate": 1.5177772412930084e-05, "loss": 0.002, "step": 1287 }, { "epoch": 1.1827364554637283, "grad_norm": 2.351283550262451, "learning_rate": 1.5168630796416614e-05, "loss": 0.1402, "step": 1288 }, { "epoch": 1.1836547291092745, "grad_norm": 1.5682311058044434, "learning_rate": 1.515948328216935e-05, "loss": 0.0429, "step": 1289 }, { "epoch": 1.184573002754821, "grad_norm": 1.7619497776031494, "learning_rate": 1.515032988062618e-05, "loss": 0.0912, "step": 1290 }, { "epoch": 1.1854912764003673, "grad_norm": 1.4277276992797852, "learning_rate": 1.5141170602231716e-05, "loss": 0.0733, "step": 1291 }, { "epoch": 1.1864095500459138, "grad_norm": 0.025811122730374336, "learning_rate": 1.5132005457437267e-05, "loss": 0.0008, "step": 1292 }, { "epoch": 1.1873278236914602, "grad_norm": 0.07539168000221252, "learning_rate": 1.512283445670085e-05, "loss": 0.002, "step": 1293 }, { "epoch": 1.1882460973370064, "grad_norm": 0.37523579597473145, "learning_rate": 1.5113657610487145e-05, "loss": 0.0141, "step": 1294 }, { "epoch": 1.1891643709825528, "grad_norm": 1.5776855945587158, "learning_rate": 1.510447492926752e-05, "loss": 0.1294, "step": 1295 }, { "epoch": 1.1900826446280992, "grad_norm": 1.1566779613494873, "learning_rate": 1.5095286423519993e-05, "loss": 0.0397, "step": 1296 }, { "epoch": 1.1910009182736456, "grad_norm": 0.02321074903011322, "learning_rate": 1.5086092103729226e-05, "loss": 0.0007, "step": 1297 }, { "epoch": 1.1919191919191918, "grad_norm": 0.039550382643938065, "learning_rate": 1.507689198038652e-05, "loss": 0.0012, "step": 1298 }, { "epoch": 1.1928374655647382, "grad_norm": 0.7146385312080383, "learning_rate": 1.5067686063989793e-05, "loss": 0.0119, "step": 1299 }, { "epoch": 1.1937557392102847, "grad_norm": 1.0617892742156982, "learning_rate": 1.5058474365043579e-05, "loss": 0.0333, "step": 1300 }, { "epoch": 1.194674012855831, "grad_norm": 0.09871313720941544, "learning_rate": 1.5049256894059008e-05, "loss": 0.0028, "step": 1301 }, { "epoch": 1.1955922865013775, "grad_norm": 0.45368796586990356, "learning_rate": 1.5040033661553794e-05, "loss": 0.0178, "step": 1302 }, { "epoch": 1.1965105601469237, "grad_norm": 1.026053786277771, "learning_rate": 1.5030804678052228e-05, "loss": 0.022, "step": 1303 }, { "epoch": 1.1974288337924701, "grad_norm": 0.748936116695404, "learning_rate": 1.502156995408516e-05, "loss": 0.0163, "step": 1304 }, { "epoch": 1.1983471074380165, "grad_norm": 3.059931755065918, "learning_rate": 1.5012329500189999e-05, "loss": 0.1487, "step": 1305 }, { "epoch": 1.199265381083563, "grad_norm": 0.15573690831661224, "learning_rate": 1.5003083326910681e-05, "loss": 0.0035, "step": 1306 }, { "epoch": 1.2001836547291092, "grad_norm": 1.5742480754852295, "learning_rate": 1.4993831444797673e-05, "loss": 0.1711, "step": 1307 }, { "epoch": 1.2011019283746556, "grad_norm": 1.2625138759613037, "learning_rate": 1.4984573864407958e-05, "loss": 0.0508, "step": 1308 }, { "epoch": 1.202020202020202, "grad_norm": 1.171500325202942, "learning_rate": 1.4975310596305014e-05, "loss": 0.025, "step": 1309 }, { "epoch": 1.2029384756657484, "grad_norm": 0.17074929177761078, "learning_rate": 1.4966041651058825e-05, "loss": 0.0049, "step": 1310 }, { "epoch": 1.2038567493112948, "grad_norm": 1.2583304643630981, "learning_rate": 1.4956767039245832e-05, "loss": 0.1008, "step": 1311 }, { "epoch": 1.204775022956841, "grad_norm": 0.05326935648918152, "learning_rate": 1.4947486771448955e-05, "loss": 0.0015, "step": 1312 }, { "epoch": 1.2056932966023874, "grad_norm": 1.4588842391967773, "learning_rate": 1.4938200858257567e-05, "loss": 0.0567, "step": 1313 }, { "epoch": 1.2066115702479339, "grad_norm": 1.3217723369598389, "learning_rate": 1.492890931026748e-05, "loss": 0.0353, "step": 1314 }, { "epoch": 1.2075298438934803, "grad_norm": 0.7128495573997498, "learning_rate": 1.4919612138080934e-05, "loss": 0.0798, "step": 1315 }, { "epoch": 1.2084481175390267, "grad_norm": 0.7111722826957703, "learning_rate": 1.4910309352306595e-05, "loss": 0.0185, "step": 1316 }, { "epoch": 1.209366391184573, "grad_norm": 0.18263565003871918, "learning_rate": 1.4901000963559522e-05, "loss": 0.0032, "step": 1317 }, { "epoch": 1.2102846648301193, "grad_norm": 1.0827876329421997, "learning_rate": 1.4891686982461176e-05, "loss": 0.0212, "step": 1318 }, { "epoch": 1.2112029384756657, "grad_norm": 0.7115151286125183, "learning_rate": 1.4882367419639397e-05, "loss": 0.0175, "step": 1319 }, { "epoch": 1.2121212121212122, "grad_norm": 1.6152924299240112, "learning_rate": 1.4873042285728394e-05, "loss": 0.063, "step": 1320 }, { "epoch": 1.2130394857667586, "grad_norm": 1.2398662567138672, "learning_rate": 1.4863711591368732e-05, "loss": 0.0567, "step": 1321 }, { "epoch": 1.2139577594123048, "grad_norm": 0.3149197995662689, "learning_rate": 1.4854375347207326e-05, "loss": 0.0041, "step": 1322 }, { "epoch": 1.2148760330578512, "grad_norm": 0.0354398638010025, "learning_rate": 1.4845033563897414e-05, "loss": 0.0011, "step": 1323 }, { "epoch": 1.2157943067033976, "grad_norm": 0.5944275259971619, "learning_rate": 1.4835686252098565e-05, "loss": 0.0074, "step": 1324 }, { "epoch": 1.216712580348944, "grad_norm": 1.5587214231491089, "learning_rate": 1.4826333422476648e-05, "loss": 0.0321, "step": 1325 }, { "epoch": 1.2176308539944904, "grad_norm": 1.26503586769104, "learning_rate": 1.4816975085703833e-05, "loss": 0.0308, "step": 1326 }, { "epoch": 1.2185491276400366, "grad_norm": 2.2985401153564453, "learning_rate": 1.4807611252458572e-05, "loss": 0.1577, "step": 1327 }, { "epoch": 1.219467401285583, "grad_norm": 1.0657761096954346, "learning_rate": 1.479824193342559e-05, "loss": 0.0228, "step": 1328 }, { "epoch": 1.2203856749311295, "grad_norm": 1.494015097618103, "learning_rate": 1.478886713929587e-05, "loss": 0.0273, "step": 1329 }, { "epoch": 1.221303948576676, "grad_norm": 0.3253692388534546, "learning_rate": 1.4779486880766647e-05, "loss": 0.0092, "step": 1330 }, { "epoch": 1.2222222222222223, "grad_norm": 1.535876750946045, "learning_rate": 1.4770101168541385e-05, "loss": 0.0992, "step": 1331 }, { "epoch": 1.2231404958677685, "grad_norm": 1.6007992029190063, "learning_rate": 1.4760710013329771e-05, "loss": 0.0763, "step": 1332 }, { "epoch": 1.224058769513315, "grad_norm": 0.16145643591880798, "learning_rate": 1.475131342584771e-05, "loss": 0.0029, "step": 1333 }, { "epoch": 1.2249770431588614, "grad_norm": 0.4277931749820709, "learning_rate": 1.4741911416817299e-05, "loss": 0.0118, "step": 1334 }, { "epoch": 1.2258953168044078, "grad_norm": 1.2851041555404663, "learning_rate": 1.4732503996966823e-05, "loss": 0.0696, "step": 1335 }, { "epoch": 1.2268135904499542, "grad_norm": 1.1918232440948486, "learning_rate": 1.4723091177030747e-05, "loss": 0.1295, "step": 1336 }, { "epoch": 1.2277318640955004, "grad_norm": 1.7181370258331299, "learning_rate": 1.4713672967749684e-05, "loss": 0.0757, "step": 1337 }, { "epoch": 1.2286501377410468, "grad_norm": 0.5153246521949768, "learning_rate": 1.4704249379870407e-05, "loss": 0.0208, "step": 1338 }, { "epoch": 1.2295684113865932, "grad_norm": 0.8718439936637878, "learning_rate": 1.4694820424145828e-05, "loss": 0.0174, "step": 1339 }, { "epoch": 1.2304866850321396, "grad_norm": 0.06026957184076309, "learning_rate": 1.4685386111334978e-05, "loss": 0.0016, "step": 1340 }, { "epoch": 1.231404958677686, "grad_norm": 0.9866961240768433, "learning_rate": 1.4675946452203005e-05, "loss": 0.0457, "step": 1341 }, { "epoch": 1.2323232323232323, "grad_norm": 1.145094394683838, "learning_rate": 1.4666501457521154e-05, "loss": 0.0837, "step": 1342 }, { "epoch": 1.2332415059687787, "grad_norm": 1.749701976776123, "learning_rate": 1.4657051138066755e-05, "loss": 0.0889, "step": 1343 }, { "epoch": 1.234159779614325, "grad_norm": 2.237783432006836, "learning_rate": 1.4647595504623229e-05, "loss": 0.3536, "step": 1344 }, { "epoch": 1.2350780532598715, "grad_norm": 3.2181859016418457, "learning_rate": 1.463813456798004e-05, "loss": 0.3937, "step": 1345 }, { "epoch": 1.2359963269054177, "grad_norm": 0.16558580100536346, "learning_rate": 1.4628668338932721e-05, "loss": 0.0035, "step": 1346 }, { "epoch": 1.2369146005509641, "grad_norm": 0.9569987058639526, "learning_rate": 1.461919682828283e-05, "loss": 0.0532, "step": 1347 }, { "epoch": 1.2378328741965106, "grad_norm": 1.4198359251022339, "learning_rate": 1.4609720046837966e-05, "loss": 0.05, "step": 1348 }, { "epoch": 1.238751147842057, "grad_norm": 1.212015151977539, "learning_rate": 1.460023800541173e-05, "loss": 0.0568, "step": 1349 }, { "epoch": 1.2396694214876034, "grad_norm": 1.1781357526779175, "learning_rate": 1.4590750714823728e-05, "loss": 0.0192, "step": 1350 }, { "epoch": 1.2405876951331496, "grad_norm": 0.8566604852676392, "learning_rate": 1.4581258185899558e-05, "loss": 0.0813, "step": 1351 }, { "epoch": 1.241505968778696, "grad_norm": 0.7922446131706238, "learning_rate": 1.4571760429470794e-05, "loss": 0.0462, "step": 1352 }, { "epoch": 1.2424242424242424, "grad_norm": 1.0106927156448364, "learning_rate": 1.4562257456374979e-05, "loss": 0.0243, "step": 1353 }, { "epoch": 1.2433425160697889, "grad_norm": 0.5323588848114014, "learning_rate": 1.4552749277455602e-05, "loss": 0.0186, "step": 1354 }, { "epoch": 1.244260789715335, "grad_norm": 0.9507684707641602, "learning_rate": 1.4543235903562094e-05, "loss": 0.0623, "step": 1355 }, { "epoch": 1.2451790633608815, "grad_norm": 1.1500047445297241, "learning_rate": 1.4533717345549816e-05, "loss": 0.0269, "step": 1356 }, { "epoch": 1.2460973370064279, "grad_norm": 1.102471947669983, "learning_rate": 1.452419361428004e-05, "loss": 0.071, "step": 1357 }, { "epoch": 1.2470156106519743, "grad_norm": 0.2313786894083023, "learning_rate": 1.4514664720619951e-05, "loss": 0.0071, "step": 1358 }, { "epoch": 1.2479338842975207, "grad_norm": 1.0613731145858765, "learning_rate": 1.4505130675442612e-05, "loss": 0.0385, "step": 1359 }, { "epoch": 1.248852157943067, "grad_norm": 0.6685549020767212, "learning_rate": 1.4495591489626973e-05, "loss": 0.0145, "step": 1360 }, { "epoch": 1.2497704315886133, "grad_norm": 1.2858562469482422, "learning_rate": 1.4486047174057844e-05, "loss": 0.0328, "step": 1361 }, { "epoch": 1.2506887052341598, "grad_norm": 0.35447919368743896, "learning_rate": 1.4476497739625894e-05, "loss": 0.0078, "step": 1362 }, { "epoch": 1.2516069788797062, "grad_norm": 1.6132014989852905, "learning_rate": 1.446694319722763e-05, "loss": 0.0772, "step": 1363 }, { "epoch": 1.2525252525252526, "grad_norm": 0.6720917224884033, "learning_rate": 1.4457383557765385e-05, "loss": 0.0158, "step": 1364 }, { "epoch": 1.2534435261707988, "grad_norm": 0.06080632284283638, "learning_rate": 1.444781883214731e-05, "loss": 0.0019, "step": 1365 }, { "epoch": 1.2543617998163452, "grad_norm": 1.435562014579773, "learning_rate": 1.4438249031287365e-05, "loss": 0.0686, "step": 1366 }, { "epoch": 1.2552800734618916, "grad_norm": 0.39553844928741455, "learning_rate": 1.4428674166105294e-05, "loss": 0.0089, "step": 1367 }, { "epoch": 1.256198347107438, "grad_norm": 1.1331499814987183, "learning_rate": 1.4419094247526621e-05, "loss": 0.0363, "step": 1368 }, { "epoch": 1.2571166207529845, "grad_norm": 0.7735754251480103, "learning_rate": 1.4409509286482641e-05, "loss": 0.0094, "step": 1369 }, { "epoch": 1.2580348943985307, "grad_norm": 1.7218488454818726, "learning_rate": 1.4399919293910396e-05, "loss": 0.1031, "step": 1370 }, { "epoch": 1.258953168044077, "grad_norm": 0.44980621337890625, "learning_rate": 1.4390324280752674e-05, "loss": 0.0067, "step": 1371 }, { "epoch": 1.2598714416896235, "grad_norm": 0.1809144765138626, "learning_rate": 1.4380724257957985e-05, "loss": 0.0033, "step": 1372 }, { "epoch": 1.26078971533517, "grad_norm": 1.3895292282104492, "learning_rate": 1.4371119236480571e-05, "loss": 0.0638, "step": 1373 }, { "epoch": 1.2617079889807163, "grad_norm": 2.4137563705444336, "learning_rate": 1.4361509227280357e-05, "loss": 0.3338, "step": 1374 }, { "epoch": 1.2626262626262625, "grad_norm": 0.02335980534553528, "learning_rate": 1.4351894241322977e-05, "loss": 0.0007, "step": 1375 }, { "epoch": 1.263544536271809, "grad_norm": 1.971006155014038, "learning_rate": 1.4342274289579731e-05, "loss": 0.174, "step": 1376 }, { "epoch": 1.2644628099173554, "grad_norm": 2.0516486167907715, "learning_rate": 1.4332649383027597e-05, "loss": 0.0418, "step": 1377 }, { "epoch": 1.2653810835629018, "grad_norm": 1.5072158575057983, "learning_rate": 1.4323019532649198e-05, "loss": 0.1634, "step": 1378 }, { "epoch": 1.2662993572084482, "grad_norm": 1.6795438528060913, "learning_rate": 1.4313384749432797e-05, "loss": 0.1266, "step": 1379 }, { "epoch": 1.2672176308539944, "grad_norm": 1.440539836883545, "learning_rate": 1.4303745044372293e-05, "loss": 0.042, "step": 1380 }, { "epoch": 1.2681359044995408, "grad_norm": 1.0741921663284302, "learning_rate": 1.4294100428467197e-05, "loss": 0.0559, "step": 1381 }, { "epoch": 1.2690541781450873, "grad_norm": 1.027987003326416, "learning_rate": 1.4284450912722624e-05, "loss": 0.024, "step": 1382 }, { "epoch": 1.2699724517906337, "grad_norm": 0.36295393109321594, "learning_rate": 1.4274796508149282e-05, "loss": 0.0094, "step": 1383 }, { "epoch": 1.27089072543618, "grad_norm": 0.761314868927002, "learning_rate": 1.4265137225763452e-05, "loss": 0.0233, "step": 1384 }, { "epoch": 1.2718089990817263, "grad_norm": 1.3624578714370728, "learning_rate": 1.425547307658699e-05, "loss": 0.1077, "step": 1385 }, { "epoch": 1.2727272727272727, "grad_norm": 0.3581523299217224, "learning_rate": 1.4245804071647294e-05, "loss": 0.0076, "step": 1386 }, { "epoch": 1.2736455463728191, "grad_norm": 0.7298873066902161, "learning_rate": 1.4236130221977311e-05, "loss": 0.0218, "step": 1387 }, { "epoch": 1.2745638200183655, "grad_norm": 0.5934001803398132, "learning_rate": 1.4226451538615513e-05, "loss": 0.0121, "step": 1388 }, { "epoch": 1.275482093663912, "grad_norm": 1.467621088027954, "learning_rate": 1.421676803260589e-05, "loss": 0.0806, "step": 1389 }, { "epoch": 1.2764003673094582, "grad_norm": 0.3423660099506378, "learning_rate": 1.4207079714997933e-05, "loss": 0.0083, "step": 1390 }, { "epoch": 1.2773186409550046, "grad_norm": 0.3791164755821228, "learning_rate": 1.4197386596846624e-05, "loss": 0.0065, "step": 1391 }, { "epoch": 1.278236914600551, "grad_norm": 0.48991885781288147, "learning_rate": 1.4187688689212417e-05, "loss": 0.0089, "step": 1392 }, { "epoch": 1.2791551882460974, "grad_norm": 0.8918730020523071, "learning_rate": 1.4177986003161244e-05, "loss": 0.0617, "step": 1393 }, { "epoch": 1.2800734618916438, "grad_norm": 0.9387685060501099, "learning_rate": 1.4168278549764479e-05, "loss": 0.1246, "step": 1394 }, { "epoch": 1.28099173553719, "grad_norm": 0.329987496137619, "learning_rate": 1.4158566340098937e-05, "loss": 0.0027, "step": 1395 }, { "epoch": 1.2819100091827365, "grad_norm": 0.8850637674331665, "learning_rate": 1.414884938524686e-05, "loss": 0.032, "step": 1396 }, { "epoch": 1.2828282828282829, "grad_norm": 0.976370632648468, "learning_rate": 1.4139127696295913e-05, "loss": 0.0351, "step": 1397 }, { "epoch": 1.283746556473829, "grad_norm": 8.030200004577637, "learning_rate": 1.412940128433915e-05, "loss": 0.7436, "step": 1398 }, { "epoch": 1.2846648301193757, "grad_norm": 5.73181676864624, "learning_rate": 1.4119670160475028e-05, "loss": 0.1487, "step": 1399 }, { "epoch": 1.285583103764922, "grad_norm": 0.2607089877128601, "learning_rate": 1.4109934335807365e-05, "loss": 0.0082, "step": 1400 }, { "epoch": 1.2865013774104683, "grad_norm": 0.6773238778114319, "learning_rate": 1.4100193821445356e-05, "loss": 0.0405, "step": 1401 }, { "epoch": 1.2874196510560147, "grad_norm": 0.825866162776947, "learning_rate": 1.4090448628503542e-05, "loss": 0.0156, "step": 1402 }, { "epoch": 1.288337924701561, "grad_norm": 1.571028709411621, "learning_rate": 1.40806987681018e-05, "loss": 0.0888, "step": 1403 }, { "epoch": 1.2892561983471074, "grad_norm": 1.0329324007034302, "learning_rate": 1.407094425136534e-05, "loss": 0.0304, "step": 1404 }, { "epoch": 1.2901744719926538, "grad_norm": 0.795371949672699, "learning_rate": 1.4061185089424676e-05, "loss": 0.0413, "step": 1405 }, { "epoch": 1.2910927456382002, "grad_norm": 1.536001443862915, "learning_rate": 1.4051421293415635e-05, "loss": 0.0178, "step": 1406 }, { "epoch": 1.2920110192837466, "grad_norm": 0.7221614718437195, "learning_rate": 1.4041652874479317e-05, "loss": 0.0157, "step": 1407 }, { "epoch": 1.2929292929292928, "grad_norm": 0.07364149391651154, "learning_rate": 1.4031879843762104e-05, "loss": 0.0026, "step": 1408 }, { "epoch": 1.2938475665748392, "grad_norm": 0.07770667225122452, "learning_rate": 1.4022102212415647e-05, "loss": 0.0025, "step": 1409 }, { "epoch": 1.2947658402203857, "grad_norm": 1.278053641319275, "learning_rate": 1.4012319991596835e-05, "loss": 0.0251, "step": 1410 }, { "epoch": 1.295684113865932, "grad_norm": 0.6270511746406555, "learning_rate": 1.40025331924678e-05, "loss": 0.016, "step": 1411 }, { "epoch": 1.2966023875114785, "grad_norm": 0.7099285125732422, "learning_rate": 1.3992741826195899e-05, "loss": 0.0755, "step": 1412 }, { "epoch": 1.2975206611570247, "grad_norm": 0.9976291656494141, "learning_rate": 1.3982945903953694e-05, "loss": 0.0677, "step": 1413 }, { "epoch": 1.298438934802571, "grad_norm": 0.9596277475357056, "learning_rate": 1.3973145436918957e-05, "loss": 0.062, "step": 1414 }, { "epoch": 1.2993572084481175, "grad_norm": 1.7740943431854248, "learning_rate": 1.3963340436274635e-05, "loss": 0.1282, "step": 1415 }, { "epoch": 1.300275482093664, "grad_norm": 0.9065020084381104, "learning_rate": 1.395353091320885e-05, "loss": 0.0551, "step": 1416 }, { "epoch": 1.3011937557392104, "grad_norm": 0.8288874626159668, "learning_rate": 1.3943716878914892e-05, "loss": 0.0227, "step": 1417 }, { "epoch": 1.3021120293847566, "grad_norm": 1.205163598060608, "learning_rate": 1.3933898344591192e-05, "loss": 0.1126, "step": 1418 }, { "epoch": 1.303030303030303, "grad_norm": 2.7789723873138428, "learning_rate": 1.3924075321441313e-05, "loss": 0.0867, "step": 1419 }, { "epoch": 1.3039485766758494, "grad_norm": 0.8977410197257996, "learning_rate": 1.391424782067395e-05, "loss": 0.0564, "step": 1420 }, { "epoch": 1.3048668503213958, "grad_norm": 1.0026594400405884, "learning_rate": 1.3904415853502898e-05, "loss": 0.0211, "step": 1421 }, { "epoch": 1.3057851239669422, "grad_norm": 1.6274348497390747, "learning_rate": 1.3894579431147048e-05, "loss": 0.1282, "step": 1422 }, { "epoch": 1.3067033976124884, "grad_norm": 1.0977287292480469, "learning_rate": 1.3884738564830385e-05, "loss": 0.037, "step": 1423 }, { "epoch": 1.3076216712580349, "grad_norm": 2.0556561946868896, "learning_rate": 1.3874893265781957e-05, "loss": 0.1382, "step": 1424 }, { "epoch": 1.3085399449035813, "grad_norm": 2.8433923721313477, "learning_rate": 1.3865043545235865e-05, "loss": 0.1217, "step": 1425 }, { "epoch": 1.3094582185491277, "grad_norm": 0.2416485846042633, "learning_rate": 1.3855189414431272e-05, "loss": 0.0074, "step": 1426 }, { "epoch": 1.3103764921946741, "grad_norm": 0.5300748944282532, "learning_rate": 1.3845330884612353e-05, "loss": 0.013, "step": 1427 }, { "epoch": 1.3112947658402203, "grad_norm": 1.6921361684799194, "learning_rate": 1.383546796702832e-05, "loss": 0.1241, "step": 1428 }, { "epoch": 1.3122130394857667, "grad_norm": 4.56161642074585, "learning_rate": 1.3825600672933379e-05, "loss": 0.1439, "step": 1429 }, { "epoch": 1.3131313131313131, "grad_norm": 1.1459616422653198, "learning_rate": 1.3815729013586739e-05, "loss": 0.0175, "step": 1430 }, { "epoch": 1.3140495867768596, "grad_norm": 1.2235965728759766, "learning_rate": 1.3805853000252584e-05, "loss": 0.2708, "step": 1431 }, { "epoch": 1.314967860422406, "grad_norm": 1.8148595094680786, "learning_rate": 1.3795972644200066e-05, "loss": 0.12, "step": 1432 }, { "epoch": 1.3158861340679522, "grad_norm": 0.3302179276943207, "learning_rate": 1.37860879567033e-05, "loss": 0.0037, "step": 1433 }, { "epoch": 1.3168044077134986, "grad_norm": 1.2210602760314941, "learning_rate": 1.377619894904133e-05, "loss": 0.042, "step": 1434 }, { "epoch": 1.317722681359045, "grad_norm": 3.1871652603149414, "learning_rate": 1.3766305632498148e-05, "loss": 0.1855, "step": 1435 }, { "epoch": 1.3186409550045914, "grad_norm": 0.6383305788040161, "learning_rate": 1.3756408018362644e-05, "loss": 0.0214, "step": 1436 }, { "epoch": 1.3195592286501379, "grad_norm": 0.37105199694633484, "learning_rate": 1.3746506117928622e-05, "loss": 0.0141, "step": 1437 }, { "epoch": 1.320477502295684, "grad_norm": 1.5616554021835327, "learning_rate": 1.3736599942494777e-05, "loss": 0.0349, "step": 1438 }, { "epoch": 1.3213957759412305, "grad_norm": 0.13334499299526215, "learning_rate": 1.3726689503364679e-05, "loss": 0.0033, "step": 1439 }, { "epoch": 1.322314049586777, "grad_norm": 0.4598747193813324, "learning_rate": 1.3716774811846763e-05, "loss": 0.0347, "step": 1440 }, { "epoch": 1.3232323232323233, "grad_norm": 1.2357678413391113, "learning_rate": 1.3706855879254317e-05, "loss": 0.0958, "step": 1441 }, { "epoch": 1.3241505968778697, "grad_norm": 1.5633628368377686, "learning_rate": 1.3696932716905473e-05, "loss": 0.22, "step": 1442 }, { "epoch": 1.325068870523416, "grad_norm": 1.0772192478179932, "learning_rate": 1.3687005336123185e-05, "loss": 0.0391, "step": 1443 }, { "epoch": 1.3259871441689624, "grad_norm": 0.11936847120523453, "learning_rate": 1.3677073748235215e-05, "loss": 0.003, "step": 1444 }, { "epoch": 1.3269054178145088, "grad_norm": 0.1269206553697586, "learning_rate": 1.3667137964574135e-05, "loss": 0.0038, "step": 1445 }, { "epoch": 1.327823691460055, "grad_norm": 2.996716022491455, "learning_rate": 1.36571979964773e-05, "loss": 0.1795, "step": 1446 }, { "epoch": 1.3287419651056016, "grad_norm": 2.1550629138946533, "learning_rate": 1.3647253855286843e-05, "loss": 0.0894, "step": 1447 }, { "epoch": 1.3296602387511478, "grad_norm": 1.9352715015411377, "learning_rate": 1.3637305552349656e-05, "loss": 0.0812, "step": 1448 }, { "epoch": 1.3305785123966942, "grad_norm": 0.6135302782058716, "learning_rate": 1.3627353099017375e-05, "loss": 0.0185, "step": 1449 }, { "epoch": 1.3314967860422406, "grad_norm": 1.2556521892547607, "learning_rate": 1.3617396506646386e-05, "loss": 0.0663, "step": 1450 }, { "epoch": 1.3324150596877868, "grad_norm": 0.20395106077194214, "learning_rate": 1.3607435786597779e-05, "loss": 0.0048, "step": 1451 }, { "epoch": 1.3333333333333333, "grad_norm": 0.34823986887931824, "learning_rate": 1.3597470950237372e-05, "loss": 0.01, "step": 1452 }, { "epoch": 1.3342516069788797, "grad_norm": 0.3579889237880707, "learning_rate": 1.3587502008935666e-05, "loss": 0.0093, "step": 1453 }, { "epoch": 1.335169880624426, "grad_norm": 1.1027183532714844, "learning_rate": 1.357752897406786e-05, "loss": 0.0347, "step": 1454 }, { "epoch": 1.3360881542699725, "grad_norm": 1.011323094367981, "learning_rate": 1.3567551857013807e-05, "loss": 0.0331, "step": 1455 }, { "epoch": 1.3370064279155187, "grad_norm": 3.222738265991211, "learning_rate": 1.3557570669158029e-05, "loss": 0.1364, "step": 1456 }, { "epoch": 1.3379247015610651, "grad_norm": 0.861265242099762, "learning_rate": 1.3547585421889696e-05, "loss": 0.073, "step": 1457 }, { "epoch": 1.3388429752066116, "grad_norm": 0.8662389516830444, "learning_rate": 1.3537596126602599e-05, "loss": 0.0248, "step": 1458 }, { "epoch": 1.339761248852158, "grad_norm": 1.0932942628860474, "learning_rate": 1.3527602794695159e-05, "loss": 0.0773, "step": 1459 }, { "epoch": 1.3406795224977044, "grad_norm": 1.6843175888061523, "learning_rate": 1.3517605437570396e-05, "loss": 0.0548, "step": 1460 }, { "epoch": 1.3415977961432506, "grad_norm": 1.462658405303955, "learning_rate": 1.3507604066635923e-05, "loss": 0.3648, "step": 1461 }, { "epoch": 1.342516069788797, "grad_norm": 0.8938584923744202, "learning_rate": 1.3497598693303939e-05, "loss": 0.0321, "step": 1462 }, { "epoch": 1.3434343434343434, "grad_norm": 1.2755552530288696, "learning_rate": 1.3487589328991203e-05, "loss": 0.0383, "step": 1463 }, { "epoch": 1.3443526170798898, "grad_norm": 3.0948970317840576, "learning_rate": 1.3477575985119034e-05, "loss": 0.1347, "step": 1464 }, { "epoch": 1.3452708907254363, "grad_norm": 2.3230063915252686, "learning_rate": 1.3467558673113286e-05, "loss": 0.1218, "step": 1465 }, { "epoch": 1.3461891643709825, "grad_norm": 2.186903953552246, "learning_rate": 1.3457537404404346e-05, "loss": 0.1225, "step": 1466 }, { "epoch": 1.3471074380165289, "grad_norm": 1.5007251501083374, "learning_rate": 1.3447512190427112e-05, "loss": 0.0354, "step": 1467 }, { "epoch": 1.3480257116620753, "grad_norm": 2.0396759510040283, "learning_rate": 1.3437483042620988e-05, "loss": 0.2147, "step": 1468 }, { "epoch": 1.3489439853076217, "grad_norm": 0.29022401571273804, "learning_rate": 1.3427449972429865e-05, "loss": 0.0098, "step": 1469 }, { "epoch": 1.3498622589531681, "grad_norm": 1.3758379220962524, "learning_rate": 1.3417412991302103e-05, "loss": 0.0381, "step": 1470 }, { "epoch": 1.3507805325987143, "grad_norm": 1.7509135007858276, "learning_rate": 1.3407372110690538e-05, "loss": 0.1097, "step": 1471 }, { "epoch": 1.3516988062442608, "grad_norm": 1.0362128019332886, "learning_rate": 1.3397327342052445e-05, "loss": 0.0266, "step": 1472 }, { "epoch": 1.3526170798898072, "grad_norm": 1.1638273000717163, "learning_rate": 1.3387278696849535e-05, "loss": 0.0434, "step": 1473 }, { "epoch": 1.3535353535353536, "grad_norm": 0.6686537265777588, "learning_rate": 1.3377226186547954e-05, "loss": 0.0108, "step": 1474 }, { "epoch": 1.3544536271809, "grad_norm": 1.1910710334777832, "learning_rate": 1.3367169822618242e-05, "loss": 0.0476, "step": 1475 }, { "epoch": 1.3553719008264462, "grad_norm": 0.31856080889701843, "learning_rate": 1.3357109616535352e-05, "loss": 0.0088, "step": 1476 }, { "epoch": 1.3562901744719926, "grad_norm": 3.465343475341797, "learning_rate": 1.334704557977861e-05, "loss": 0.2469, "step": 1477 }, { "epoch": 1.357208448117539, "grad_norm": 1.1648708581924438, "learning_rate": 1.3336977723831717e-05, "loss": 0.1219, "step": 1478 }, { "epoch": 1.3581267217630855, "grad_norm": 1.1484390497207642, "learning_rate": 1.3326906060182732e-05, "loss": 0.058, "step": 1479 }, { "epoch": 1.3590449954086319, "grad_norm": 1.155448317527771, "learning_rate": 1.3316830600324057e-05, "loss": 0.0392, "step": 1480 }, { "epoch": 1.359963269054178, "grad_norm": 1.7666198015213013, "learning_rate": 1.330675135575243e-05, "loss": 0.1102, "step": 1481 }, { "epoch": 1.3608815426997245, "grad_norm": 1.2074227333068848, "learning_rate": 1.3296668337968904e-05, "loss": 0.0536, "step": 1482 }, { "epoch": 1.361799816345271, "grad_norm": 0.7518617510795593, "learning_rate": 1.328658155847884e-05, "loss": 0.0332, "step": 1483 }, { "epoch": 1.3627180899908173, "grad_norm": 0.05703407898545265, "learning_rate": 1.3276491028791888e-05, "loss": 0.0017, "step": 1484 }, { "epoch": 1.3636363636363638, "grad_norm": 1.5222060680389404, "learning_rate": 1.3266396760421976e-05, "loss": 0.0476, "step": 1485 }, { "epoch": 1.36455463728191, "grad_norm": 0.3774162828922272, "learning_rate": 1.3256298764887302e-05, "loss": 0.0125, "step": 1486 }, { "epoch": 1.3654729109274564, "grad_norm": 1.1229931116104126, "learning_rate": 1.324619705371032e-05, "loss": 0.0251, "step": 1487 }, { "epoch": 1.3663911845730028, "grad_norm": 0.6298224329948425, "learning_rate": 1.323609163841772e-05, "loss": 0.0219, "step": 1488 }, { "epoch": 1.367309458218549, "grad_norm": 0.182538241147995, "learning_rate": 1.3225982530540416e-05, "loss": 0.0063, "step": 1489 }, { "epoch": 1.3682277318640956, "grad_norm": 0.1673356145620346, "learning_rate": 1.3215869741613534e-05, "loss": 0.0041, "step": 1490 }, { "epoch": 1.3691460055096418, "grad_norm": 0.19133666157722473, "learning_rate": 1.320575328317641e-05, "loss": 0.0063, "step": 1491 }, { "epoch": 1.3700642791551882, "grad_norm": 1.274795651435852, "learning_rate": 1.3195633166772557e-05, "loss": 0.0657, "step": 1492 }, { "epoch": 1.3709825528007347, "grad_norm": 1.4993994235992432, "learning_rate": 1.318550940394967e-05, "loss": 0.0709, "step": 1493 }, { "epoch": 1.3719008264462809, "grad_norm": 1.2915112972259521, "learning_rate": 1.3175382006259599e-05, "loss": 0.0752, "step": 1494 }, { "epoch": 1.3728191000918273, "grad_norm": 0.7298415899276733, "learning_rate": 1.3165250985258344e-05, "loss": 0.0242, "step": 1495 }, { "epoch": 1.3737373737373737, "grad_norm": 0.12022161483764648, "learning_rate": 1.3155116352506038e-05, "loss": 0.0038, "step": 1496 }, { "epoch": 1.3746556473829201, "grad_norm": 0.472030907869339, "learning_rate": 1.3144978119566937e-05, "loss": 0.0187, "step": 1497 }, { "epoch": 1.3755739210284665, "grad_norm": 1.5011793375015259, "learning_rate": 1.3134836298009408e-05, "loss": 0.0775, "step": 1498 }, { "epoch": 1.3764921946740127, "grad_norm": 0.49107256531715393, "learning_rate": 1.3124690899405903e-05, "loss": 0.0179, "step": 1499 }, { "epoch": 1.3774104683195592, "grad_norm": 0.3146115839481354, "learning_rate": 1.3114541935332966e-05, "loss": 0.0072, "step": 1500 }, { "epoch": 1.3783287419651056, "grad_norm": 0.7482264637947083, "learning_rate": 1.3104389417371201e-05, "loss": 0.0148, "step": 1501 }, { "epoch": 1.379247015610652, "grad_norm": 1.608660340309143, "learning_rate": 1.3094233357105277e-05, "loss": 0.2261, "step": 1502 }, { "epoch": 1.3801652892561984, "grad_norm": 0.7975808382034302, "learning_rate": 1.3084073766123894e-05, "loss": 0.0141, "step": 1503 }, { "epoch": 1.3810835629017446, "grad_norm": 1.2092190980911255, "learning_rate": 1.307391065601979e-05, "loss": 0.0252, "step": 1504 }, { "epoch": 1.382001836547291, "grad_norm": 1.8259412050247192, "learning_rate": 1.3063744038389715e-05, "loss": 0.0438, "step": 1505 }, { "epoch": 1.3829201101928374, "grad_norm": 1.9100322723388672, "learning_rate": 1.3053573924834422e-05, "loss": 0.1989, "step": 1506 }, { "epoch": 1.3838383838383839, "grad_norm": 2.6719939708709717, "learning_rate": 1.3043400326958646e-05, "loss": 0.2136, "step": 1507 }, { "epoch": 1.3847566574839303, "grad_norm": 1.1718413829803467, "learning_rate": 1.3033223256371109e-05, "loss": 0.082, "step": 1508 }, { "epoch": 1.3856749311294765, "grad_norm": 3.9917428493499756, "learning_rate": 1.3023042724684487e-05, "loss": 0.256, "step": 1509 }, { "epoch": 1.386593204775023, "grad_norm": 0.15129457414150238, "learning_rate": 1.3012858743515415e-05, "loss": 0.0047, "step": 1510 }, { "epoch": 1.3875114784205693, "grad_norm": 1.1269807815551758, "learning_rate": 1.3002671324484453e-05, "loss": 0.0587, "step": 1511 }, { "epoch": 1.3884297520661157, "grad_norm": 0.31384220719337463, "learning_rate": 1.299248047921609e-05, "loss": 0.0086, "step": 1512 }, { "epoch": 1.3893480257116622, "grad_norm": 3.8617441654205322, "learning_rate": 1.2982286219338724e-05, "loss": 0.1997, "step": 1513 }, { "epoch": 1.3902662993572084, "grad_norm": 1.2505799531936646, "learning_rate": 1.2972088556484645e-05, "loss": 0.0325, "step": 1514 }, { "epoch": 1.3911845730027548, "grad_norm": 0.4550495445728302, "learning_rate": 1.2961887502290036e-05, "loss": 0.0112, "step": 1515 }, { "epoch": 1.3921028466483012, "grad_norm": 0.0885041356086731, "learning_rate": 1.2951683068394941e-05, "loss": 0.0026, "step": 1516 }, { "epoch": 1.3930211202938476, "grad_norm": 0.37975966930389404, "learning_rate": 1.2941475266443258e-05, "loss": 0.005, "step": 1517 }, { "epoch": 1.393939393939394, "grad_norm": 0.28421857953071594, "learning_rate": 1.293126410808274e-05, "loss": 0.0087, "step": 1518 }, { "epoch": 1.3948576675849402, "grad_norm": 0.6830288171768188, "learning_rate": 1.2921049604964957e-05, "loss": 0.0162, "step": 1519 }, { "epoch": 1.3957759412304866, "grad_norm": 1.8712999820709229, "learning_rate": 1.2910831768745304e-05, "loss": 0.1258, "step": 1520 }, { "epoch": 1.396694214876033, "grad_norm": 1.8727294206619263, "learning_rate": 1.2900610611082977e-05, "loss": 0.1031, "step": 1521 }, { "epoch": 1.3976124885215795, "grad_norm": 1.1437268257141113, "learning_rate": 1.2890386143640965e-05, "loss": 0.0676, "step": 1522 }, { "epoch": 1.398530762167126, "grad_norm": 0.412086546421051, "learning_rate": 1.2880158378086023e-05, "loss": 0.0061, "step": 1523 }, { "epoch": 1.399449035812672, "grad_norm": 0.06430686265230179, "learning_rate": 1.286992732608868e-05, "loss": 0.0018, "step": 1524 }, { "epoch": 1.4003673094582185, "grad_norm": 0.1637493371963501, "learning_rate": 1.2859692999323215e-05, "loss": 0.0046, "step": 1525 }, { "epoch": 1.401285583103765, "grad_norm": 0.8898007869720459, "learning_rate": 1.2849455409467636e-05, "loss": 0.0305, "step": 1526 }, { "epoch": 1.4022038567493114, "grad_norm": 2.1258695125579834, "learning_rate": 1.283921456820368e-05, "loss": 0.0514, "step": 1527 }, { "epoch": 1.4031221303948578, "grad_norm": 2.0626821517944336, "learning_rate": 1.2828970487216794e-05, "loss": 0.0756, "step": 1528 }, { "epoch": 1.404040404040404, "grad_norm": 0.43485671281814575, "learning_rate": 1.281872317819612e-05, "loss": 0.0086, "step": 1529 }, { "epoch": 1.4049586776859504, "grad_norm": 0.5193171501159668, "learning_rate": 1.2808472652834486e-05, "loss": 0.0132, "step": 1530 }, { "epoch": 1.4058769513314968, "grad_norm": 2.6670925617218018, "learning_rate": 1.279821892282838e-05, "loss": 0.0359, "step": 1531 }, { "epoch": 1.4067952249770432, "grad_norm": 2.080864191055298, "learning_rate": 1.2787961999877968e-05, "loss": 0.19, "step": 1532 }, { "epoch": 1.4077134986225897, "grad_norm": 0.24805764853954315, "learning_rate": 1.2777701895687034e-05, "loss": 0.0068, "step": 1533 }, { "epoch": 1.4086317722681359, "grad_norm": 1.4307785034179688, "learning_rate": 1.2767438621963013e-05, "loss": 0.188, "step": 1534 }, { "epoch": 1.4095500459136823, "grad_norm": 1.5302629470825195, "learning_rate": 1.275717219041694e-05, "loss": 0.0545, "step": 1535 }, { "epoch": 1.4104683195592287, "grad_norm": 0.17209069430828094, "learning_rate": 1.274690261276347e-05, "loss": 0.0063, "step": 1536 }, { "epoch": 1.4113865932047749, "grad_norm": 0.13780298829078674, "learning_rate": 1.2736629900720832e-05, "loss": 0.0029, "step": 1537 }, { "epoch": 1.4123048668503215, "grad_norm": 0.31114983558654785, "learning_rate": 1.2726354066010838e-05, "loss": 0.0087, "step": 1538 }, { "epoch": 1.4132231404958677, "grad_norm": 0.04411287233233452, "learning_rate": 1.2716075120358869e-05, "loss": 0.0013, "step": 1539 }, { "epoch": 1.4141414141414141, "grad_norm": 1.0890171527862549, "learning_rate": 1.2705793075493845e-05, "loss": 0.0218, "step": 1540 }, { "epoch": 1.4150596877869606, "grad_norm": 5.735299587249756, "learning_rate": 1.269550794314823e-05, "loss": 0.1739, "step": 1541 }, { "epoch": 1.4159779614325068, "grad_norm": 0.6252123713493347, "learning_rate": 1.2685219735058014e-05, "loss": 0.0164, "step": 1542 }, { "epoch": 1.4168962350780532, "grad_norm": 0.5476959347724915, "learning_rate": 1.267492846296268e-05, "loss": 0.0258, "step": 1543 }, { "epoch": 1.4178145087235996, "grad_norm": 0.8396297097206116, "learning_rate": 1.2664634138605231e-05, "loss": 0.0343, "step": 1544 }, { "epoch": 1.418732782369146, "grad_norm": 0.2872767746448517, "learning_rate": 1.2654336773732133e-05, "loss": 0.0074, "step": 1545 }, { "epoch": 1.4196510560146924, "grad_norm": 0.8232660889625549, "learning_rate": 1.264403638009333e-05, "loss": 0.0502, "step": 1546 }, { "epoch": 1.4205693296602386, "grad_norm": 0.13369010388851166, "learning_rate": 1.2633732969442218e-05, "loss": 0.0031, "step": 1547 }, { "epoch": 1.421487603305785, "grad_norm": 1.673848032951355, "learning_rate": 1.2623426553535648e-05, "loss": 0.0306, "step": 1548 }, { "epoch": 1.4224058769513315, "grad_norm": 1.3026249408721924, "learning_rate": 1.2613117144133884e-05, "loss": 0.1183, "step": 1549 }, { "epoch": 1.423324150596878, "grad_norm": 2.0804195404052734, "learning_rate": 1.2602804753000611e-05, "loss": 0.1127, "step": 1550 }, { "epoch": 1.4242424242424243, "grad_norm": 0.0686287134885788, "learning_rate": 1.2592489391902922e-05, "loss": 0.0024, "step": 1551 }, { "epoch": 1.4251606978879705, "grad_norm": 0.046727944165468216, "learning_rate": 1.2582171072611291e-05, "loss": 0.0014, "step": 1552 }, { "epoch": 1.426078971533517, "grad_norm": 3.1876981258392334, "learning_rate": 1.2571849806899574e-05, "loss": 0.0482, "step": 1553 }, { "epoch": 1.4269972451790633, "grad_norm": 1.1826698780059814, "learning_rate": 1.2561525606544986e-05, "loss": 0.1357, "step": 1554 }, { "epoch": 1.4279155188246098, "grad_norm": 0.05360499396920204, "learning_rate": 1.2551198483328091e-05, "loss": 0.0015, "step": 1555 }, { "epoch": 1.4288337924701562, "grad_norm": 0.036801133304834366, "learning_rate": 1.2540868449032785e-05, "loss": 0.0011, "step": 1556 }, { "epoch": 1.4297520661157024, "grad_norm": 0.211741104722023, "learning_rate": 1.2530535515446291e-05, "loss": 0.0052, "step": 1557 }, { "epoch": 1.4306703397612488, "grad_norm": 0.07676536589860916, "learning_rate": 1.2520199694359143e-05, "loss": 0.0024, "step": 1558 }, { "epoch": 1.4315886134067952, "grad_norm": 1.6685951948165894, "learning_rate": 1.2509860997565158e-05, "loss": 0.0631, "step": 1559 }, { "epoch": 1.4325068870523416, "grad_norm": 0.726445198059082, "learning_rate": 1.249951943686144e-05, "loss": 0.0402, "step": 1560 }, { "epoch": 1.433425160697888, "grad_norm": 0.9606001973152161, "learning_rate": 1.248917502404837e-05, "loss": 0.0371, "step": 1561 }, { "epoch": 1.4343434343434343, "grad_norm": 0.08557160198688507, "learning_rate": 1.2478827770929567e-05, "loss": 0.003, "step": 1562 }, { "epoch": 1.4352617079889807, "grad_norm": 1.32611882686615, "learning_rate": 1.2468477689311907e-05, "loss": 0.025, "step": 1563 }, { "epoch": 1.436179981634527, "grad_norm": 1.6499688625335693, "learning_rate": 1.2458124791005481e-05, "loss": 0.1356, "step": 1564 }, { "epoch": 1.4370982552800735, "grad_norm": 1.0992649793624878, "learning_rate": 1.24477690878236e-05, "loss": 0.0204, "step": 1565 }, { "epoch": 1.43801652892562, "grad_norm": 0.5392690300941467, "learning_rate": 1.2437410591582777e-05, "loss": 0.0128, "step": 1566 }, { "epoch": 1.4389348025711661, "grad_norm": 0.3189356327056885, "learning_rate": 1.2427049314102708e-05, "loss": 0.0115, "step": 1567 }, { "epoch": 1.4398530762167125, "grad_norm": 0.5092428922653198, "learning_rate": 1.2416685267206264e-05, "loss": 0.0096, "step": 1568 }, { "epoch": 1.440771349862259, "grad_norm": 0.5477570295333862, "learning_rate": 1.2406318462719475e-05, "loss": 0.0115, "step": 1569 }, { "epoch": 1.4416896235078054, "grad_norm": 0.5139508247375488, "learning_rate": 1.2395948912471524e-05, "loss": 0.0117, "step": 1570 }, { "epoch": 1.4426078971533518, "grad_norm": 1.110296368598938, "learning_rate": 1.2385576628294715e-05, "loss": 0.0538, "step": 1571 }, { "epoch": 1.443526170798898, "grad_norm": 1.0437670946121216, "learning_rate": 1.2375201622024483e-05, "loss": 0.0366, "step": 1572 }, { "epoch": 1.4444444444444444, "grad_norm": 0.19889909029006958, "learning_rate": 1.2364823905499366e-05, "loss": 0.0049, "step": 1573 }, { "epoch": 1.4453627180899908, "grad_norm": 0.563510537147522, "learning_rate": 1.2354443490560986e-05, "loss": 0.0127, "step": 1574 }, { "epoch": 1.4462809917355373, "grad_norm": 1.5119891166687012, "learning_rate": 1.234406038905406e-05, "loss": 0.1149, "step": 1575 }, { "epoch": 1.4471992653810837, "grad_norm": 2.267784595489502, "learning_rate": 1.2333674612826356e-05, "loss": 0.1373, "step": 1576 }, { "epoch": 1.4481175390266299, "grad_norm": 0.8701643943786621, "learning_rate": 1.2323286173728703e-05, "loss": 0.0283, "step": 1577 }, { "epoch": 1.4490358126721763, "grad_norm": 3.272038698196411, "learning_rate": 1.2312895083614966e-05, "loss": 0.2206, "step": 1578 }, { "epoch": 1.4499540863177227, "grad_norm": 0.9827746748924255, "learning_rate": 1.230250135434203e-05, "loss": 0.0302, "step": 1579 }, { "epoch": 1.4508723599632691, "grad_norm": 0.28160080313682556, "learning_rate": 1.2292104997769799e-05, "loss": 0.0073, "step": 1580 }, { "epoch": 1.4517906336088156, "grad_norm": 0.9369149208068848, "learning_rate": 1.2281706025761168e-05, "loss": 0.0355, "step": 1581 }, { "epoch": 1.4527089072543617, "grad_norm": 1.6857761144638062, "learning_rate": 1.2271304450182026e-05, "loss": 0.1152, "step": 1582 }, { "epoch": 1.4536271808999082, "grad_norm": 0.1839858740568161, "learning_rate": 1.2260900282901227e-05, "loss": 0.0062, "step": 1583 }, { "epoch": 1.4545454545454546, "grad_norm": 1.2992037534713745, "learning_rate": 1.2250493535790574e-05, "loss": 0.0338, "step": 1584 }, { "epoch": 1.4554637281910008, "grad_norm": 0.5904789566993713, "learning_rate": 1.224008422072483e-05, "loss": 0.0213, "step": 1585 }, { "epoch": 1.4563820018365474, "grad_norm": 0.36617568135261536, "learning_rate": 1.2229672349581674e-05, "loss": 0.0092, "step": 1586 }, { "epoch": 1.4573002754820936, "grad_norm": 0.10227639973163605, "learning_rate": 1.2219257934241715e-05, "loss": 0.0025, "step": 1587 }, { "epoch": 1.45821854912764, "grad_norm": 1.0734188556671143, "learning_rate": 1.2208840986588451e-05, "loss": 0.0447, "step": 1588 }, { "epoch": 1.4591368227731865, "grad_norm": 1.8984320163726807, "learning_rate": 1.219842151850828e-05, "loss": 0.295, "step": 1589 }, { "epoch": 1.4600550964187327, "grad_norm": 1.886082410812378, "learning_rate": 1.2187999541890476e-05, "loss": 0.0958, "step": 1590 }, { "epoch": 1.460973370064279, "grad_norm": 0.4929989278316498, "learning_rate": 1.2177575068627162e-05, "loss": 0.0146, "step": 1591 }, { "epoch": 1.4618916437098255, "grad_norm": 1.0957071781158447, "learning_rate": 1.2167148110613327e-05, "loss": 0.1114, "step": 1592 }, { "epoch": 1.462809917355372, "grad_norm": 0.04555021598935127, "learning_rate": 1.2156718679746785e-05, "loss": 0.0014, "step": 1593 }, { "epoch": 1.4637281910009183, "grad_norm": 0.7964405417442322, "learning_rate": 1.2146286787928178e-05, "loss": 0.0204, "step": 1594 }, { "epoch": 1.4646464646464645, "grad_norm": 0.36438480019569397, "learning_rate": 1.213585244706095e-05, "loss": 0.0108, "step": 1595 }, { "epoch": 1.465564738292011, "grad_norm": 2.4392409324645996, "learning_rate": 1.2125415669051339e-05, "loss": 0.1159, "step": 1596 }, { "epoch": 1.4664830119375574, "grad_norm": 0.23458242416381836, "learning_rate": 1.2114976465808374e-05, "loss": 0.0049, "step": 1597 }, { "epoch": 1.4674012855831038, "grad_norm": 1.3301581144332886, "learning_rate": 1.210453484924384e-05, "loss": 0.0761, "step": 1598 }, { "epoch": 1.4683195592286502, "grad_norm": 1.4806665182113647, "learning_rate": 1.2094090831272279e-05, "loss": 0.098, "step": 1599 }, { "epoch": 1.4692378328741964, "grad_norm": 1.662431001663208, "learning_rate": 1.2083644423810978e-05, "loss": 0.0274, "step": 1600 }, { "epoch": 1.4701561065197428, "grad_norm": 0.38247764110565186, "learning_rate": 1.2073195638779944e-05, "loss": 0.0123, "step": 1601 }, { "epoch": 1.4710743801652892, "grad_norm": 1.6496295928955078, "learning_rate": 1.2062744488101905e-05, "loss": 0.0755, "step": 1602 }, { "epoch": 1.4719926538108357, "grad_norm": 1.3451918363571167, "learning_rate": 1.2052290983702275e-05, "loss": 0.1224, "step": 1603 }, { "epoch": 1.472910927456382, "grad_norm": 0.8697177767753601, "learning_rate": 1.204183513750917e-05, "loss": 0.027, "step": 1604 }, { "epoch": 1.4738292011019283, "grad_norm": 0.6521785259246826, "learning_rate": 1.2031376961453362e-05, "loss": 0.0127, "step": 1605 }, { "epoch": 1.4747474747474747, "grad_norm": 0.16043150424957275, "learning_rate": 1.2020916467468298e-05, "loss": 0.0032, "step": 1606 }, { "epoch": 1.4756657483930211, "grad_norm": 0.5871933102607727, "learning_rate": 1.2010453667490058e-05, "loss": 0.0185, "step": 1607 }, { "epoch": 1.4765840220385675, "grad_norm": 1.1675398349761963, "learning_rate": 1.1999988573457352e-05, "loss": 0.073, "step": 1608 }, { "epoch": 1.477502295684114, "grad_norm": 0.7734425663948059, "learning_rate": 1.1989521197311522e-05, "loss": 0.0244, "step": 1609 }, { "epoch": 1.4784205693296602, "grad_norm": 1.2245303392410278, "learning_rate": 1.1979051550996497e-05, "loss": 0.0294, "step": 1610 }, { "epoch": 1.4793388429752066, "grad_norm": 1.532114863395691, "learning_rate": 1.1968579646458807e-05, "loss": 0.0961, "step": 1611 }, { "epoch": 1.480257116620753, "grad_norm": 6.909714698791504, "learning_rate": 1.1958105495647557e-05, "loss": 0.0936, "step": 1612 }, { "epoch": 1.4811753902662994, "grad_norm": 0.9206078052520752, "learning_rate": 1.1947629110514408e-05, "loss": 0.0285, "step": 1613 }, { "epoch": 1.4820936639118458, "grad_norm": 1.2203686237335205, "learning_rate": 1.1937150503013583e-05, "loss": 0.0314, "step": 1614 }, { "epoch": 1.483011937557392, "grad_norm": 0.08844926953315735, "learning_rate": 1.1926669685101827e-05, "loss": 0.0031, "step": 1615 }, { "epoch": 1.4839302112029384, "grad_norm": 1.0796210765838623, "learning_rate": 1.1916186668738425e-05, "loss": 0.0293, "step": 1616 }, { "epoch": 1.4848484848484849, "grad_norm": 1.5738489627838135, "learning_rate": 1.1905701465885151e-05, "loss": 0.1354, "step": 1617 }, { "epoch": 1.4857667584940313, "grad_norm": 0.9895362854003906, "learning_rate": 1.1895214088506284e-05, "loss": 0.0191, "step": 1618 }, { "epoch": 1.4866850321395777, "grad_norm": 0.1404476910829544, "learning_rate": 1.1884724548568587e-05, "loss": 0.0034, "step": 1619 }, { "epoch": 1.487603305785124, "grad_norm": 0.09938345849514008, "learning_rate": 1.1874232858041285e-05, "loss": 0.0027, "step": 1620 }, { "epoch": 1.4885215794306703, "grad_norm": 1.354430913925171, "learning_rate": 1.1863739028896056e-05, "loss": 0.0231, "step": 1621 }, { "epoch": 1.4894398530762167, "grad_norm": 0.7588155269622803, "learning_rate": 1.1853243073107025e-05, "loss": 0.0136, "step": 1622 }, { "epoch": 1.4903581267217632, "grad_norm": 1.329035758972168, "learning_rate": 1.184274500265074e-05, "loss": 0.0619, "step": 1623 }, { "epoch": 1.4912764003673096, "grad_norm": 1.1537206172943115, "learning_rate": 1.1832244829506157e-05, "loss": 0.1525, "step": 1624 }, { "epoch": 1.4921946740128558, "grad_norm": 0.4227020740509033, "learning_rate": 1.1821742565654639e-05, "loss": 0.0074, "step": 1625 }, { "epoch": 1.4931129476584022, "grad_norm": 0.264200896024704, "learning_rate": 1.181123822307993e-05, "loss": 0.0059, "step": 1626 }, { "epoch": 1.4940312213039486, "grad_norm": 0.049308959394693375, "learning_rate": 1.1800731813768151e-05, "loss": 0.0011, "step": 1627 }, { "epoch": 1.494949494949495, "grad_norm": 0.9174440503120422, "learning_rate": 1.1790223349707775e-05, "loss": 0.0329, "step": 1628 }, { "epoch": 1.4958677685950414, "grad_norm": 1.0988399982452393, "learning_rate": 1.177971284288962e-05, "loss": 0.034, "step": 1629 }, { "epoch": 1.4967860422405876, "grad_norm": 1.1897692680358887, "learning_rate": 1.1769200305306845e-05, "loss": 0.1158, "step": 1630 }, { "epoch": 1.497704315886134, "grad_norm": 0.4628237783908844, "learning_rate": 1.1758685748954913e-05, "loss": 0.0122, "step": 1631 }, { "epoch": 1.4986225895316805, "grad_norm": 0.2701522707939148, "learning_rate": 1.1748169185831596e-05, "loss": 0.0051, "step": 1632 }, { "epoch": 1.4995408631772267, "grad_norm": 0.12728597223758698, "learning_rate": 1.1737650627936957e-05, "loss": 0.0032, "step": 1633 }, { "epoch": 1.5004591368227733, "grad_norm": 0.3675557076931, "learning_rate": 1.172713008727333e-05, "loss": 0.0061, "step": 1634 }, { "epoch": 1.5013774104683195, "grad_norm": 1.345911979675293, "learning_rate": 1.1716607575845327e-05, "loss": 0.0423, "step": 1635 }, { "epoch": 1.502295684113866, "grad_norm": 0.050724610686302185, "learning_rate": 1.1706083105659787e-05, "loss": 0.0011, "step": 1636 }, { "epoch": 1.5032139577594124, "grad_norm": 2.1647963523864746, "learning_rate": 1.1695556688725796e-05, "loss": 0.1235, "step": 1637 }, { "epoch": 1.5041322314049586, "grad_norm": 2.3303332328796387, "learning_rate": 1.1685028337054663e-05, "loss": 0.1858, "step": 1638 }, { "epoch": 1.5050505050505052, "grad_norm": 0.04621339589357376, "learning_rate": 1.1674498062659895e-05, "loss": 0.0012, "step": 1639 }, { "epoch": 1.5059687786960514, "grad_norm": 0.1317455917596817, "learning_rate": 1.1663965877557209e-05, "loss": 0.0033, "step": 1640 }, { "epoch": 1.5068870523415978, "grad_norm": 0.5862374901771545, "learning_rate": 1.1653431793764488e-05, "loss": 0.0094, "step": 1641 }, { "epoch": 1.5078053259871442, "grad_norm": 3.6082942485809326, "learning_rate": 1.164289582330178e-05, "loss": 0.1606, "step": 1642 }, { "epoch": 1.5087235996326904, "grad_norm": 0.0805804654955864, "learning_rate": 1.1632357978191303e-05, "loss": 0.002, "step": 1643 }, { "epoch": 1.509641873278237, "grad_norm": 0.33015772700309753, "learning_rate": 1.1621818270457395e-05, "loss": 0.0041, "step": 1644 }, { "epoch": 1.5105601469237833, "grad_norm": 0.3248351812362671, "learning_rate": 1.1611276712126531e-05, "loss": 0.0071, "step": 1645 }, { "epoch": 1.5114784205693297, "grad_norm": 1.6816513538360596, "learning_rate": 1.160073331522729e-05, "loss": 0.0205, "step": 1646 }, { "epoch": 1.512396694214876, "grad_norm": 1.0930410623550415, "learning_rate": 1.159018809179036e-05, "loss": 0.0334, "step": 1647 }, { "epoch": 1.5133149678604223, "grad_norm": 0.2954522371292114, "learning_rate": 1.1579641053848502e-05, "loss": 0.0059, "step": 1648 }, { "epoch": 1.514233241505969, "grad_norm": 1.3748897314071655, "learning_rate": 1.156909221343655e-05, "loss": 0.14, "step": 1649 }, { "epoch": 1.5151515151515151, "grad_norm": 1.7905101776123047, "learning_rate": 1.15585415825914e-05, "loss": 0.0693, "step": 1650 }, { "epoch": 1.5160697887970616, "grad_norm": 2.2719004154205322, "learning_rate": 1.1547989173351986e-05, "loss": 0.0702, "step": 1651 }, { "epoch": 1.516988062442608, "grad_norm": 0.6427391767501831, "learning_rate": 1.153743499775927e-05, "loss": 0.0199, "step": 1652 }, { "epoch": 1.5179063360881542, "grad_norm": 0.37752383947372437, "learning_rate": 1.1526879067856239e-05, "loss": 0.0139, "step": 1653 }, { "epoch": 1.5188246097337006, "grad_norm": 0.30428823828697205, "learning_rate": 1.1516321395687864e-05, "loss": 0.0066, "step": 1654 }, { "epoch": 1.519742883379247, "grad_norm": 2.2893428802490234, "learning_rate": 1.1505761993301125e-05, "loss": 0.1412, "step": 1655 }, { "epoch": 1.5206611570247934, "grad_norm": 3.0771963596343994, "learning_rate": 1.149520087274496e-05, "loss": 0.2063, "step": 1656 }, { "epoch": 1.5215794306703398, "grad_norm": 1.2860441207885742, "learning_rate": 1.148463804607028e-05, "loss": 0.1023, "step": 1657 }, { "epoch": 1.522497704315886, "grad_norm": 1.2257939577102661, "learning_rate": 1.1474073525329929e-05, "loss": 0.083, "step": 1658 }, { "epoch": 1.5234159779614325, "grad_norm": 0.31156492233276367, "learning_rate": 1.14635073225787e-05, "loss": 0.0088, "step": 1659 }, { "epoch": 1.5243342516069789, "grad_norm": 0.5757427215576172, "learning_rate": 1.145293944987329e-05, "loss": 0.0069, "step": 1660 }, { "epoch": 1.5252525252525253, "grad_norm": 0.6866268515586853, "learning_rate": 1.1442369919272309e-05, "loss": 0.0202, "step": 1661 }, { "epoch": 1.5261707988980717, "grad_norm": 0.5849758386611938, "learning_rate": 1.1431798742836262e-05, "loss": 0.0254, "step": 1662 }, { "epoch": 1.527089072543618, "grad_norm": 3.425710916519165, "learning_rate": 1.1421225932627524e-05, "loss": 0.3221, "step": 1663 }, { "epoch": 1.5280073461891643, "grad_norm": 1.2247107028961182, "learning_rate": 1.1410651500710345e-05, "loss": 0.0709, "step": 1664 }, { "epoch": 1.5289256198347108, "grad_norm": 1.1974455118179321, "learning_rate": 1.1400075459150812e-05, "loss": 0.0814, "step": 1665 }, { "epoch": 1.529843893480257, "grad_norm": 1.1653634309768677, "learning_rate": 1.1389497820016859e-05, "loss": 0.0442, "step": 1666 }, { "epoch": 1.5307621671258036, "grad_norm": 1.575628638267517, "learning_rate": 1.1378918595378239e-05, "loss": 0.0565, "step": 1667 }, { "epoch": 1.5316804407713498, "grad_norm": 1.9796161651611328, "learning_rate": 1.1368337797306513e-05, "loss": 0.0865, "step": 1668 }, { "epoch": 1.5325987144168962, "grad_norm": 0.3961884379386902, "learning_rate": 1.135775543787504e-05, "loss": 0.015, "step": 1669 }, { "epoch": 1.5335169880624426, "grad_norm": 1.2871453762054443, "learning_rate": 1.1347171529158963e-05, "loss": 0.0849, "step": 1670 }, { "epoch": 1.5344352617079888, "grad_norm": 0.11286300420761108, "learning_rate": 1.1336586083235184e-05, "loss": 0.0034, "step": 1671 }, { "epoch": 1.5353535353535355, "grad_norm": 1.2782742977142334, "learning_rate": 1.1325999112182367e-05, "loss": 0.0629, "step": 1672 }, { "epoch": 1.5362718089990817, "grad_norm": 1.2194396257400513, "learning_rate": 1.1315410628080915e-05, "loss": 0.0346, "step": 1673 }, { "epoch": 1.537190082644628, "grad_norm": 2.08431339263916, "learning_rate": 1.1304820643012955e-05, "loss": 0.0837, "step": 1674 }, { "epoch": 1.5381083562901745, "grad_norm": 0.0943789854645729, "learning_rate": 1.1294229169062327e-05, "loss": 0.0026, "step": 1675 }, { "epoch": 1.5390266299357207, "grad_norm": 1.9883638620376587, "learning_rate": 1.1283636218314575e-05, "loss": 0.0525, "step": 1676 }, { "epoch": 1.5399449035812673, "grad_norm": 0.055516231805086136, "learning_rate": 1.127304180285692e-05, "loss": 0.0017, "step": 1677 }, { "epoch": 1.5408631772268135, "grad_norm": 0.9948946237564087, "learning_rate": 1.1262445934778258e-05, "loss": 0.0325, "step": 1678 }, { "epoch": 1.54178145087236, "grad_norm": 0.16287299990653992, "learning_rate": 1.1251848626169148e-05, "loss": 0.0043, "step": 1679 }, { "epoch": 1.5426997245179064, "grad_norm": 1.0054166316986084, "learning_rate": 1.1241249889121781e-05, "loss": 0.043, "step": 1680 }, { "epoch": 1.5436179981634526, "grad_norm": 1.3409167528152466, "learning_rate": 1.123064973572999e-05, "loss": 0.0521, "step": 1681 }, { "epoch": 1.5445362718089992, "grad_norm": 3.1624839305877686, "learning_rate": 1.1220048178089215e-05, "loss": 0.5394, "step": 1682 }, { "epoch": 1.5454545454545454, "grad_norm": 0.9077445268630981, "learning_rate": 1.1209445228296506e-05, "loss": 0.0366, "step": 1683 }, { "epoch": 1.5463728191000918, "grad_norm": 1.0106688737869263, "learning_rate": 1.1198840898450493e-05, "loss": 0.0528, "step": 1684 }, { "epoch": 1.5472910927456383, "grad_norm": 0.4115508496761322, "learning_rate": 1.1188235200651387e-05, "loss": 0.0116, "step": 1685 }, { "epoch": 1.5482093663911844, "grad_norm": 0.1855420023202896, "learning_rate": 1.1177628147000961e-05, "loss": 0.0053, "step": 1686 }, { "epoch": 1.549127640036731, "grad_norm": 0.311124324798584, "learning_rate": 1.1167019749602527e-05, "loss": 0.006, "step": 1687 }, { "epoch": 1.5500459136822773, "grad_norm": 1.2261589765548706, "learning_rate": 1.1156410020560943e-05, "loss": 0.0426, "step": 1688 }, { "epoch": 1.5509641873278237, "grad_norm": 1.9252090454101562, "learning_rate": 1.1145798971982572e-05, "loss": 0.0979, "step": 1689 }, { "epoch": 1.5518824609733701, "grad_norm": 0.596693217754364, "learning_rate": 1.1135186615975293e-05, "loss": 0.0366, "step": 1690 }, { "epoch": 1.5528007346189163, "grad_norm": 1.3246040344238281, "learning_rate": 1.1124572964648475e-05, "loss": 0.0321, "step": 1691 }, { "epoch": 1.553719008264463, "grad_norm": 0.5857438445091248, "learning_rate": 1.1113958030112959e-05, "loss": 0.0123, "step": 1692 }, { "epoch": 1.5546372819100092, "grad_norm": 1.1647073030471802, "learning_rate": 1.1103341824481059e-05, "loss": 0.1291, "step": 1693 }, { "epoch": 1.5555555555555556, "grad_norm": 0.8950112462043762, "learning_rate": 1.1092724359866537e-05, "loss": 0.0287, "step": 1694 }, { "epoch": 1.556473829201102, "grad_norm": 0.4775436520576477, "learning_rate": 1.1082105648384584e-05, "loss": 0.0055, "step": 1695 }, { "epoch": 1.5573921028466482, "grad_norm": 0.21644887328147888, "learning_rate": 1.1071485702151825e-05, "loss": 0.004, "step": 1696 }, { "epoch": 1.5583103764921948, "grad_norm": 1.3247164487838745, "learning_rate": 1.1060864533286286e-05, "loss": 0.0706, "step": 1697 }, { "epoch": 1.559228650137741, "grad_norm": 0.4288880228996277, "learning_rate": 1.105024215390739e-05, "loss": 0.0174, "step": 1698 }, { "epoch": 1.5601469237832875, "grad_norm": 3.335099458694458, "learning_rate": 1.1039618576135943e-05, "loss": 0.2799, "step": 1699 }, { "epoch": 1.5610651974288339, "grad_norm": 2.0215675830841064, "learning_rate": 1.102899381209412e-05, "loss": 0.1582, "step": 1700 }, { "epoch": 1.56198347107438, "grad_norm": 1.4136135578155518, "learning_rate": 1.1018367873905444e-05, "loss": 0.0472, "step": 1701 }, { "epoch": 1.5629017447199265, "grad_norm": 0.6038053631782532, "learning_rate": 1.1007740773694779e-05, "loss": 0.008, "step": 1702 }, { "epoch": 1.563820018365473, "grad_norm": 0.7760810256004333, "learning_rate": 1.0997112523588322e-05, "loss": 0.0311, "step": 1703 }, { "epoch": 1.5647382920110193, "grad_norm": 0.33101221919059753, "learning_rate": 1.0986483135713573e-05, "loss": 0.0093, "step": 1704 }, { "epoch": 1.5656565656565657, "grad_norm": 1.100105881690979, "learning_rate": 1.0975852622199337e-05, "loss": 0.0473, "step": 1705 }, { "epoch": 1.566574839302112, "grad_norm": 1.0492099523544312, "learning_rate": 1.0965220995175695e-05, "loss": 0.0186, "step": 1706 }, { "epoch": 1.5674931129476584, "grad_norm": 1.4313894510269165, "learning_rate": 1.0954588266774008e-05, "loss": 0.0537, "step": 1707 }, { "epoch": 1.5684113865932048, "grad_norm": 1.1799801588058472, "learning_rate": 1.0943954449126894e-05, "loss": 0.0797, "step": 1708 }, { "epoch": 1.5693296602387512, "grad_norm": 1.923316478729248, "learning_rate": 1.09333195543682e-05, "loss": 0.1879, "step": 1709 }, { "epoch": 1.5702479338842976, "grad_norm": 0.6903000473976135, "learning_rate": 1.092268359463302e-05, "loss": 0.0233, "step": 1710 }, { "epoch": 1.5711662075298438, "grad_norm": 1.1974974870681763, "learning_rate": 1.0912046582057652e-05, "loss": 0.0522, "step": 1711 }, { "epoch": 1.5720844811753902, "grad_norm": 1.5694881677627563, "learning_rate": 1.0901408528779598e-05, "loss": 0.0259, "step": 1712 }, { "epoch": 1.5730027548209367, "grad_norm": 1.8803975582122803, "learning_rate": 1.089076944693755e-05, "loss": 0.0753, "step": 1713 }, { "epoch": 1.5739210284664829, "grad_norm": 1.8632904291152954, "learning_rate": 1.0880129348671369e-05, "loss": 0.1446, "step": 1714 }, { "epoch": 1.5748393021120295, "grad_norm": 1.0607258081436157, "learning_rate": 1.086948824612208e-05, "loss": 0.0555, "step": 1715 }, { "epoch": 1.5757575757575757, "grad_norm": 0.1868482083082199, "learning_rate": 1.0858846151431848e-05, "loss": 0.0036, "step": 1716 }, { "epoch": 1.576675849403122, "grad_norm": 0.28285402059555054, "learning_rate": 1.0848203076743982e-05, "loss": 0.0079, "step": 1717 }, { "epoch": 1.5775941230486685, "grad_norm": 1.5691953897476196, "learning_rate": 1.0837559034202898e-05, "loss": 0.0623, "step": 1718 }, { "epoch": 1.5785123966942147, "grad_norm": 0.35567742586135864, "learning_rate": 1.0826914035954117e-05, "loss": 0.007, "step": 1719 }, { "epoch": 1.5794306703397614, "grad_norm": 1.3767414093017578, "learning_rate": 1.0816268094144257e-05, "loss": 0.2354, "step": 1720 }, { "epoch": 1.5803489439853076, "grad_norm": 0.28843259811401367, "learning_rate": 1.080562122092101e-05, "loss": 0.0084, "step": 1721 }, { "epoch": 1.581267217630854, "grad_norm": 2.1033146381378174, "learning_rate": 1.0794973428433126e-05, "loss": 0.1211, "step": 1722 }, { "epoch": 1.5821854912764004, "grad_norm": 1.1314795017242432, "learning_rate": 1.0784324728830411e-05, "loss": 0.1555, "step": 1723 }, { "epoch": 1.5831037649219466, "grad_norm": 0.16588358581066132, "learning_rate": 1.0773675134263702e-05, "loss": 0.0045, "step": 1724 }, { "epoch": 1.5840220385674932, "grad_norm": 1.0117363929748535, "learning_rate": 1.076302465688486e-05, "loss": 0.0354, "step": 1725 }, { "epoch": 1.5849403122130394, "grad_norm": 1.200469970703125, "learning_rate": 1.0752373308846749e-05, "loss": 0.0477, "step": 1726 }, { "epoch": 1.5858585858585859, "grad_norm": 1.7063164710998535, "learning_rate": 1.0741721102303228e-05, "loss": 0.1036, "step": 1727 }, { "epoch": 1.5867768595041323, "grad_norm": 1.4133715629577637, "learning_rate": 1.0731068049409142e-05, "loss": 0.1001, "step": 1728 }, { "epoch": 1.5876951331496785, "grad_norm": 0.5926108956336975, "learning_rate": 1.0720414162320291e-05, "loss": 0.0192, "step": 1729 }, { "epoch": 1.5886134067952251, "grad_norm": 0.8882378935813904, "learning_rate": 1.0709759453193435e-05, "loss": 0.0221, "step": 1730 }, { "epoch": 1.5895316804407713, "grad_norm": 0.20611409842967987, "learning_rate": 1.0699103934186268e-05, "loss": 0.0043, "step": 1731 }, { "epoch": 1.5904499540863177, "grad_norm": 0.3286760449409485, "learning_rate": 1.068844761745741e-05, "loss": 0.0098, "step": 1732 }, { "epoch": 1.5913682277318641, "grad_norm": 0.7659929394721985, "learning_rate": 1.0677790515166393e-05, "loss": 0.0382, "step": 1733 }, { "epoch": 1.5922865013774103, "grad_norm": 1.3289642333984375, "learning_rate": 1.0667132639473638e-05, "loss": 0.0455, "step": 1734 }, { "epoch": 1.593204775022957, "grad_norm": 0.3661649823188782, "learning_rate": 1.065647400254046e-05, "loss": 0.0085, "step": 1735 }, { "epoch": 1.5941230486685032, "grad_norm": 1.8290855884552002, "learning_rate": 1.0645814616529027e-05, "loss": 0.0359, "step": 1736 }, { "epoch": 1.5950413223140496, "grad_norm": 0.7277882695198059, "learning_rate": 1.063515449360238e-05, "loss": 0.0127, "step": 1737 }, { "epoch": 1.595959595959596, "grad_norm": 1.0978567600250244, "learning_rate": 1.0624493645924385e-05, "loss": 0.0187, "step": 1738 }, { "epoch": 1.5968778696051422, "grad_norm": 0.44673699140548706, "learning_rate": 1.0613832085659748e-05, "loss": 0.012, "step": 1739 }, { "epoch": 1.5977961432506889, "grad_norm": 1.4570921659469604, "learning_rate": 1.0603169824973974e-05, "loss": 0.0768, "step": 1740 }, { "epoch": 1.598714416896235, "grad_norm": 1.662911295890808, "learning_rate": 1.0592506876033384e-05, "loss": 0.0318, "step": 1741 }, { "epoch": 1.5996326905417815, "grad_norm": 0.24951447546482086, "learning_rate": 1.0581843251005068e-05, "loss": 0.0065, "step": 1742 }, { "epoch": 1.600550964187328, "grad_norm": 1.33821439743042, "learning_rate": 1.0571178962056899e-05, "loss": 0.0628, "step": 1743 }, { "epoch": 1.601469237832874, "grad_norm": 0.43109047412872314, "learning_rate": 1.0560514021357502e-05, "loss": 0.0068, "step": 1744 }, { "epoch": 1.6023875114784207, "grad_norm": 0.17178532481193542, "learning_rate": 1.0549848441076244e-05, "loss": 0.0041, "step": 1745 }, { "epoch": 1.603305785123967, "grad_norm": 1.6422852277755737, "learning_rate": 1.0539182233383233e-05, "loss": 0.1358, "step": 1746 }, { "epoch": 1.6042240587695134, "grad_norm": 3.67649245262146, "learning_rate": 1.052851541044928e-05, "loss": 0.2011, "step": 1747 }, { "epoch": 1.6051423324150598, "grad_norm": 0.4297116696834564, "learning_rate": 1.05178479844459e-05, "loss": 0.0087, "step": 1748 }, { "epoch": 1.606060606060606, "grad_norm": 1.3101879358291626, "learning_rate": 1.0507179967545304e-05, "loss": 0.0658, "step": 1749 }, { "epoch": 1.6069788797061524, "grad_norm": 0.4793664515018463, "learning_rate": 1.0496511371920368e-05, "loss": 0.0165, "step": 1750 }, { "epoch": 1.6078971533516988, "grad_norm": 2.7769615650177, "learning_rate": 1.0485842209744638e-05, "loss": 0.1113, "step": 1751 }, { "epoch": 1.6088154269972452, "grad_norm": 0.30461251735687256, "learning_rate": 1.0475172493192295e-05, "loss": 0.0069, "step": 1752 }, { "epoch": 1.6097337006427916, "grad_norm": 0.16963638365268707, "learning_rate": 1.0464502234438162e-05, "loss": 0.0036, "step": 1753 }, { "epoch": 1.6106519742883378, "grad_norm": 0.4945947229862213, "learning_rate": 1.045383144565768e-05, "loss": 0.0148, "step": 1754 }, { "epoch": 1.6115702479338843, "grad_norm": 4.765690326690674, "learning_rate": 1.0443160139026886e-05, "loss": 0.1104, "step": 1755 }, { "epoch": 1.6124885215794307, "grad_norm": 0.15735794603824615, "learning_rate": 1.0432488326722418e-05, "loss": 0.0048, "step": 1756 }, { "epoch": 1.613406795224977, "grad_norm": 3.093502998352051, "learning_rate": 1.0421816020921486e-05, "loss": 0.0454, "step": 1757 }, { "epoch": 1.6143250688705235, "grad_norm": 0.1500793844461441, "learning_rate": 1.0411143233801866e-05, "loss": 0.0039, "step": 1758 }, { "epoch": 1.6152433425160697, "grad_norm": 0.16832010447978973, "learning_rate": 1.0400469977541881e-05, "loss": 0.0043, "step": 1759 }, { "epoch": 1.6161616161616161, "grad_norm": 0.43627703189849854, "learning_rate": 1.0389796264320384e-05, "loss": 0.0126, "step": 1760 }, { "epoch": 1.6170798898071626, "grad_norm": 0.29146772623062134, "learning_rate": 1.037912210631676e-05, "loss": 0.0057, "step": 1761 }, { "epoch": 1.6179981634527087, "grad_norm": 0.909965455532074, "learning_rate": 1.03684475157109e-05, "loss": 0.0155, "step": 1762 }, { "epoch": 1.6189164370982554, "grad_norm": 2.2290828227996826, "learning_rate": 1.0357772504683178e-05, "loss": 0.3533, "step": 1763 }, { "epoch": 1.6198347107438016, "grad_norm": 0.2263786643743515, "learning_rate": 1.034709708541446e-05, "loss": 0.0035, "step": 1764 }, { "epoch": 1.620752984389348, "grad_norm": 1.1935417652130127, "learning_rate": 1.0336421270086071e-05, "loss": 0.0717, "step": 1765 }, { "epoch": 1.6216712580348944, "grad_norm": 0.6949930191040039, "learning_rate": 1.0325745070879793e-05, "loss": 0.0233, "step": 1766 }, { "epoch": 1.6225895316804406, "grad_norm": 0.352042555809021, "learning_rate": 1.0315068499977836e-05, "loss": 0.009, "step": 1767 }, { "epoch": 1.6235078053259873, "grad_norm": 0.5318616032600403, "learning_rate": 1.0304391569562846e-05, "loss": 0.0155, "step": 1768 }, { "epoch": 1.6244260789715335, "grad_norm": 1.7381056547164917, "learning_rate": 1.0293714291817876e-05, "loss": 0.0633, "step": 1769 }, { "epoch": 1.6253443526170799, "grad_norm": 0.7981970310211182, "learning_rate": 1.0283036678926365e-05, "loss": 0.0115, "step": 1770 }, { "epoch": 1.6262626262626263, "grad_norm": 1.9158682823181152, "learning_rate": 1.0272358743072152e-05, "loss": 0.1256, "step": 1771 }, { "epoch": 1.6271808999081725, "grad_norm": 2.0862817764282227, "learning_rate": 1.0261680496439433e-05, "loss": 0.127, "step": 1772 }, { "epoch": 1.6280991735537191, "grad_norm": 0.8581390976905823, "learning_rate": 1.0251001951212755e-05, "loss": 0.0516, "step": 1773 }, { "epoch": 1.6290174471992653, "grad_norm": 2.5112762451171875, "learning_rate": 1.024032311957702e-05, "loss": 0.1216, "step": 1774 }, { "epoch": 1.6299357208448118, "grad_norm": 0.9531700015068054, "learning_rate": 1.0229644013717446e-05, "loss": 0.0229, "step": 1775 }, { "epoch": 1.6308539944903582, "grad_norm": 0.8895351886749268, "learning_rate": 1.0218964645819562e-05, "loss": 0.0191, "step": 1776 }, { "epoch": 1.6317722681359044, "grad_norm": 3.6883482933044434, "learning_rate": 1.0208285028069207e-05, "loss": 0.2053, "step": 1777 }, { "epoch": 1.632690541781451, "grad_norm": 0.9191868901252747, "learning_rate": 1.0197605172652493e-05, "loss": 0.014, "step": 1778 }, { "epoch": 1.6336088154269972, "grad_norm": 1.8399053812026978, "learning_rate": 1.0186925091755812e-05, "loss": 0.161, "step": 1779 }, { "epoch": 1.6345270890725436, "grad_norm": 1.5125656127929688, "learning_rate": 1.0176244797565811e-05, "loss": 0.0288, "step": 1780 }, { "epoch": 1.63544536271809, "grad_norm": 0.9421117305755615, "learning_rate": 1.0165564302269375e-05, "loss": 0.0249, "step": 1781 }, { "epoch": 1.6363636363636362, "grad_norm": 1.3347094058990479, "learning_rate": 1.0154883618053628e-05, "loss": 0.044, "step": 1782 }, { "epoch": 1.6372819100091829, "grad_norm": 1.341504693031311, "learning_rate": 1.0144202757105903e-05, "loss": 0.1006, "step": 1783 }, { "epoch": 1.638200183654729, "grad_norm": 4.851644992828369, "learning_rate": 1.0133521731613734e-05, "loss": 0.1073, "step": 1784 }, { "epoch": 1.6391184573002755, "grad_norm": 0.9429187774658203, "learning_rate": 1.0122840553764845e-05, "loss": 0.0704, "step": 1785 }, { "epoch": 1.640036730945822, "grad_norm": 2.223780632019043, "learning_rate": 1.0112159235747137e-05, "loss": 0.1076, "step": 1786 }, { "epoch": 1.6409550045913681, "grad_norm": 0.5160186290740967, "learning_rate": 1.0101477789748663e-05, "loss": 0.0183, "step": 1787 }, { "epoch": 1.6418732782369148, "grad_norm": 0.7691090106964111, "learning_rate": 1.0090796227957633e-05, "loss": 0.0187, "step": 1788 }, { "epoch": 1.642791551882461, "grad_norm": 0.6348578929901123, "learning_rate": 1.0080114562562375e-05, "loss": 0.0054, "step": 1789 }, { "epoch": 1.6437098255280074, "grad_norm": 1.0306012630462646, "learning_rate": 1.0069432805751351e-05, "loss": 0.0264, "step": 1790 }, { "epoch": 1.6446280991735538, "grad_norm": 0.16869811713695526, "learning_rate": 1.005875096971311e-05, "loss": 0.004, "step": 1791 }, { "epoch": 1.6455463728191, "grad_norm": 2.4766786098480225, "learning_rate": 1.0048069066636312e-05, "loss": 0.0341, "step": 1792 }, { "epoch": 1.6464646464646466, "grad_norm": 1.650105595588684, "learning_rate": 1.0037387108709675e-05, "loss": 0.0924, "step": 1793 }, { "epoch": 1.6473829201101928, "grad_norm": 1.9203547239303589, "learning_rate": 1.0026705108121988e-05, "loss": 0.0715, "step": 1794 }, { "epoch": 1.6483011937557392, "grad_norm": 0.40670347213745117, "learning_rate": 1.0016023077062088e-05, "loss": 0.0108, "step": 1795 }, { "epoch": 1.6492194674012857, "grad_norm": 3.8184659481048584, "learning_rate": 1.0005341027718846e-05, "loss": 0.2564, "step": 1796 }, { "epoch": 1.6501377410468319, "grad_norm": 0.1021365225315094, "learning_rate": 9.994658972281156e-06, "loss": 0.0037, "step": 1797 }, { "epoch": 1.6510560146923783, "grad_norm": 2.5041258335113525, "learning_rate": 9.983976922937917e-06, "loss": 0.0867, "step": 1798 }, { "epoch": 1.6519742883379247, "grad_norm": 0.4828931391239166, "learning_rate": 9.973294891878012e-06, "loss": 0.0092, "step": 1799 }, { "epoch": 1.6528925619834711, "grad_norm": 0.42067793011665344, "learning_rate": 9.962612891290327e-06, "loss": 0.0095, "step": 1800 }, { "epoch": 1.6538108356290175, "grad_norm": 0.8106966018676758, "learning_rate": 9.95193093336369e-06, "loss": 0.0223, "step": 1801 }, { "epoch": 1.6547291092745637, "grad_norm": 1.5916285514831543, "learning_rate": 9.941249030286891e-06, "loss": 0.0493, "step": 1802 }, { "epoch": 1.6556473829201102, "grad_norm": 0.6096727848052979, "learning_rate": 9.930567194248655e-06, "loss": 0.0214, "step": 1803 }, { "epoch": 1.6565656565656566, "grad_norm": 2.2408483028411865, "learning_rate": 9.919885437437625e-06, "loss": 0.0767, "step": 1804 }, { "epoch": 1.657483930211203, "grad_norm": 0.5433112382888794, "learning_rate": 9.909203772042369e-06, "loss": 0.0087, "step": 1805 }, { "epoch": 1.6584022038567494, "grad_norm": 2.042544364929199, "learning_rate": 9.898522210251338e-06, "loss": 0.0855, "step": 1806 }, { "epoch": 1.6593204775022956, "grad_norm": 1.8381526470184326, "learning_rate": 9.887840764252868e-06, "loss": 0.1306, "step": 1807 }, { "epoch": 1.660238751147842, "grad_norm": 1.1948528289794922, "learning_rate": 9.87715944623516e-06, "loss": 0.0415, "step": 1808 }, { "epoch": 1.6611570247933884, "grad_norm": 2.2584304809570312, "learning_rate": 9.866478268386268e-06, "loss": 0.1094, "step": 1809 }, { "epoch": 1.6620752984389346, "grad_norm": 0.03934241458773613, "learning_rate": 9.855797242894099e-06, "loss": 0.0013, "step": 1810 }, { "epoch": 1.6629935720844813, "grad_norm": 1.0973306894302368, "learning_rate": 9.845116381946374e-06, "loss": 0.0169, "step": 1811 }, { "epoch": 1.6639118457300275, "grad_norm": 1.8814115524291992, "learning_rate": 9.834435697730628e-06, "loss": 0.0452, "step": 1812 }, { "epoch": 1.664830119375574, "grad_norm": 0.9947444200515747, "learning_rate": 9.823755202434194e-06, "loss": 0.0447, "step": 1813 }, { "epoch": 1.6657483930211203, "grad_norm": 0.03637301176786423, "learning_rate": 9.81307490824419e-06, "loss": 0.0009, "step": 1814 }, { "epoch": 1.6666666666666665, "grad_norm": 2.53145170211792, "learning_rate": 9.802394827347508e-06, "loss": 0.0908, "step": 1815 }, { "epoch": 1.6675849403122132, "grad_norm": 0.8128257393836975, "learning_rate": 9.791714971930798e-06, "loss": 0.0143, "step": 1816 }, { "epoch": 1.6685032139577594, "grad_norm": 0.8279877305030823, "learning_rate": 9.781035354180441e-06, "loss": 0.0069, "step": 1817 }, { "epoch": 1.6694214876033058, "grad_norm": 1.1408828496932983, "learning_rate": 9.77035598628256e-06, "loss": 0.0598, "step": 1818 }, { "epoch": 1.6703397612488522, "grad_norm": 1.149680495262146, "learning_rate": 9.759676880422983e-06, "loss": 0.121, "step": 1819 }, { "epoch": 1.6712580348943984, "grad_norm": 2.068449020385742, "learning_rate": 9.748998048787248e-06, "loss": 0.1834, "step": 1820 }, { "epoch": 1.672176308539945, "grad_norm": 1.209964632987976, "learning_rate": 9.73831950356057e-06, "loss": 0.0701, "step": 1821 }, { "epoch": 1.6730945821854912, "grad_norm": 0.6803509593009949, "learning_rate": 9.72764125692785e-06, "loss": 0.0286, "step": 1822 }, { "epoch": 1.6740128558310376, "grad_norm": 0.5691112875938416, "learning_rate": 9.716963321073637e-06, "loss": 0.0199, "step": 1823 }, { "epoch": 1.674931129476584, "grad_norm": 0.9980854988098145, "learning_rate": 9.706285708182129e-06, "loss": 0.0304, "step": 1824 }, { "epoch": 1.6758494031221303, "grad_norm": 0.8141520023345947, "learning_rate": 9.695608430437156e-06, "loss": 0.0218, "step": 1825 }, { "epoch": 1.676767676767677, "grad_norm": 0.06649437546730042, "learning_rate": 9.684931500022167e-06, "loss": 0.0024, "step": 1826 }, { "epoch": 1.677685950413223, "grad_norm": 1.8353791236877441, "learning_rate": 9.67425492912021e-06, "loss": 0.091, "step": 1827 }, { "epoch": 1.6786042240587695, "grad_norm": 0.26440978050231934, "learning_rate": 9.663578729913929e-06, "loss": 0.0085, "step": 1828 }, { "epoch": 1.679522497704316, "grad_norm": 1.2231405973434448, "learning_rate": 9.65290291458554e-06, "loss": 0.0945, "step": 1829 }, { "epoch": 1.6804407713498621, "grad_norm": 0.440205454826355, "learning_rate": 9.642227495316823e-06, "loss": 0.0135, "step": 1830 }, { "epoch": 1.6813590449954088, "grad_norm": 0.17247043550014496, "learning_rate": 9.631552484289104e-06, "loss": 0.0061, "step": 1831 }, { "epoch": 1.682277318640955, "grad_norm": 1.1240766048431396, "learning_rate": 9.620877893683243e-06, "loss": 0.0424, "step": 1832 }, { "epoch": 1.6831955922865014, "grad_norm": 0.90940260887146, "learning_rate": 9.610203735679617e-06, "loss": 0.0424, "step": 1833 }, { "epoch": 1.6841138659320478, "grad_norm": 0.48056161403656006, "learning_rate": 9.599530022458124e-06, "loss": 0.0146, "step": 1834 }, { "epoch": 1.685032139577594, "grad_norm": 0.6797727346420288, "learning_rate": 9.588856766198137e-06, "loss": 0.0144, "step": 1835 }, { "epoch": 1.6859504132231407, "grad_norm": 0.8310089707374573, "learning_rate": 9.578183979078515e-06, "loss": 0.0102, "step": 1836 }, { "epoch": 1.6868686868686869, "grad_norm": 0.025881526991724968, "learning_rate": 9.567511673277585e-06, "loss": 0.0008, "step": 1837 }, { "epoch": 1.6877869605142333, "grad_norm": 0.057864077389240265, "learning_rate": 9.556839860973114e-06, "loss": 0.0014, "step": 1838 }, { "epoch": 1.6887052341597797, "grad_norm": 0.7779200077056885, "learning_rate": 9.546168554342323e-06, "loss": 0.0228, "step": 1839 }, { "epoch": 1.6896235078053259, "grad_norm": 1.2333074808120728, "learning_rate": 9.53549776556184e-06, "loss": 0.0375, "step": 1840 }, { "epoch": 1.6905417814508723, "grad_norm": 0.11093039065599442, "learning_rate": 9.524827506807708e-06, "loss": 0.0026, "step": 1841 }, { "epoch": 1.6914600550964187, "grad_norm": 2.7494990825653076, "learning_rate": 9.514157790255367e-06, "loss": 0.3336, "step": 1842 }, { "epoch": 1.6923783287419651, "grad_norm": 2.1518893241882324, "learning_rate": 9.503488628079632e-06, "loss": 0.1, "step": 1843 }, { "epoch": 1.6932966023875116, "grad_norm": 0.08421065658330917, "learning_rate": 9.492820032454699e-06, "loss": 0.0028, "step": 1844 }, { "epoch": 1.6942148760330578, "grad_norm": 2.193918228149414, "learning_rate": 9.482152015554102e-06, "loss": 0.1139, "step": 1845 }, { "epoch": 1.6951331496786042, "grad_norm": 2.004753828048706, "learning_rate": 9.471484589550725e-06, "loss": 0.1995, "step": 1846 }, { "epoch": 1.6960514233241506, "grad_norm": 0.21469789743423462, "learning_rate": 9.460817766616772e-06, "loss": 0.0064, "step": 1847 }, { "epoch": 1.696969696969697, "grad_norm": 0.9991695880889893, "learning_rate": 9.450151558923754e-06, "loss": 0.0209, "step": 1848 }, { "epoch": 1.6978879706152434, "grad_norm": 0.4832303822040558, "learning_rate": 9.439485978642501e-06, "loss": 0.0072, "step": 1849 }, { "epoch": 1.6988062442607896, "grad_norm": 0.5144418478012085, "learning_rate": 9.428821037943104e-06, "loss": 0.0087, "step": 1850 }, { "epoch": 1.699724517906336, "grad_norm": 1.5869849920272827, "learning_rate": 9.418156748994937e-06, "loss": 0.0409, "step": 1851 }, { "epoch": 1.7006427915518825, "grad_norm": 0.37864604592323303, "learning_rate": 9.407493123966618e-06, "loss": 0.0112, "step": 1852 }, { "epoch": 1.7015610651974287, "grad_norm": 1.1724663972854614, "learning_rate": 9.396830175026026e-06, "loss": 0.0678, "step": 1853 }, { "epoch": 1.7024793388429753, "grad_norm": 0.03273358941078186, "learning_rate": 9.386167914340257e-06, "loss": 0.001, "step": 1854 }, { "epoch": 1.7033976124885215, "grad_norm": 0.37831756472587585, "learning_rate": 9.375506354075618e-06, "loss": 0.0108, "step": 1855 }, { "epoch": 1.704315886134068, "grad_norm": 0.6558111310005188, "learning_rate": 9.364845506397625e-06, "loss": 0.0222, "step": 1856 }, { "epoch": 1.7052341597796143, "grad_norm": 1.2949291467666626, "learning_rate": 9.354185383470973e-06, "loss": 0.037, "step": 1857 }, { "epoch": 1.7061524334251605, "grad_norm": 0.14035268127918243, "learning_rate": 9.343525997459544e-06, "loss": 0.0043, "step": 1858 }, { "epoch": 1.7070707070707072, "grad_norm": 0.7551899552345276, "learning_rate": 9.332867360526363e-06, "loss": 0.0208, "step": 1859 }, { "epoch": 1.7079889807162534, "grad_norm": 1.6735656261444092, "learning_rate": 9.32220948483361e-06, "loss": 0.1443, "step": 1860 }, { "epoch": 1.7089072543617998, "grad_norm": 0.08380790799856186, "learning_rate": 9.311552382542593e-06, "loss": 0.002, "step": 1861 }, { "epoch": 1.7098255280073462, "grad_norm": 2.804563045501709, "learning_rate": 9.300896065813732e-06, "loss": 0.1375, "step": 1862 }, { "epoch": 1.7107438016528924, "grad_norm": 0.199548602104187, "learning_rate": 9.290240546806567e-06, "loss": 0.0053, "step": 1863 }, { "epoch": 1.711662075298439, "grad_norm": 1.367158055305481, "learning_rate": 9.279585837679712e-06, "loss": 0.0726, "step": 1864 }, { "epoch": 1.7125803489439853, "grad_norm": 2.303842067718506, "learning_rate": 9.268931950590862e-06, "loss": 0.1585, "step": 1865 }, { "epoch": 1.7134986225895317, "grad_norm": 1.440865159034729, "learning_rate": 9.258278897696775e-06, "loss": 0.0686, "step": 1866 }, { "epoch": 1.714416896235078, "grad_norm": 2.47525954246521, "learning_rate": 9.247626691153256e-06, "loss": 0.1712, "step": 1867 }, { "epoch": 1.7153351698806243, "grad_norm": 0.23828962445259094, "learning_rate": 9.236975343115142e-06, "loss": 0.0047, "step": 1868 }, { "epoch": 1.716253443526171, "grad_norm": 0.3413209319114685, "learning_rate": 9.2263248657363e-06, "loss": 0.0058, "step": 1869 }, { "epoch": 1.7171717171717171, "grad_norm": 1.5757755041122437, "learning_rate": 9.215675271169592e-06, "loss": 0.0886, "step": 1870 }, { "epoch": 1.7180899908172635, "grad_norm": 2.337773323059082, "learning_rate": 9.20502657156688e-06, "loss": 0.0764, "step": 1871 }, { "epoch": 1.71900826446281, "grad_norm": 0.3174985349178314, "learning_rate": 9.194378779078993e-06, "loss": 0.006, "step": 1872 }, { "epoch": 1.7199265381083562, "grad_norm": 2.127171516418457, "learning_rate": 9.183731905855746e-06, "loss": 0.0975, "step": 1873 }, { "epoch": 1.7208448117539028, "grad_norm": 1.2394150495529175, "learning_rate": 9.173085964045884e-06, "loss": 0.1187, "step": 1874 }, { "epoch": 1.721763085399449, "grad_norm": 3.230731248855591, "learning_rate": 9.162440965797105e-06, "loss": 0.2536, "step": 1875 }, { "epoch": 1.7226813590449954, "grad_norm": 2.13912034034729, "learning_rate": 9.151796923256021e-06, "loss": 0.1503, "step": 1876 }, { "epoch": 1.7235996326905418, "grad_norm": 1.181375503540039, "learning_rate": 9.141153848568153e-06, "loss": 0.0895, "step": 1877 }, { "epoch": 1.724517906336088, "grad_norm": 2.773712635040283, "learning_rate": 9.130511753877924e-06, "loss": 0.137, "step": 1878 }, { "epoch": 1.7254361799816347, "grad_norm": 0.996084988117218, "learning_rate": 9.119870651328634e-06, "loss": 0.0797, "step": 1879 }, { "epoch": 1.7263544536271809, "grad_norm": 0.9319124221801758, "learning_rate": 9.109230553062454e-06, "loss": 0.0417, "step": 1880 }, { "epoch": 1.7272727272727273, "grad_norm": 1.9992011785507202, "learning_rate": 9.098591471220402e-06, "loss": 0.2573, "step": 1881 }, { "epoch": 1.7281910009182737, "grad_norm": 0.09376855194568634, "learning_rate": 9.08795341794235e-06, "loss": 0.0024, "step": 1882 }, { "epoch": 1.72910927456382, "grad_norm": 0.23316670954227448, "learning_rate": 9.07731640536698e-06, "loss": 0.0065, "step": 1883 }, { "epoch": 1.7300275482093666, "grad_norm": 0.8093835711479187, "learning_rate": 9.066680445631802e-06, "loss": 0.0318, "step": 1884 }, { "epoch": 1.7309458218549127, "grad_norm": 0.8325884938240051, "learning_rate": 9.056045550873111e-06, "loss": 0.015, "step": 1885 }, { "epoch": 1.7318640955004592, "grad_norm": 1.3844232559204102, "learning_rate": 9.045411733225992e-06, "loss": 0.079, "step": 1886 }, { "epoch": 1.7327823691460056, "grad_norm": 0.059901174157857895, "learning_rate": 9.034779004824306e-06, "loss": 0.0017, "step": 1887 }, { "epoch": 1.7337006427915518, "grad_norm": 0.16548766195774078, "learning_rate": 9.024147377800668e-06, "loss": 0.007, "step": 1888 }, { "epoch": 1.7346189164370982, "grad_norm": 0.3275957405567169, "learning_rate": 9.01351686428643e-06, "loss": 0.0074, "step": 1889 }, { "epoch": 1.7355371900826446, "grad_norm": 2.025322675704956, "learning_rate": 9.002887476411681e-06, "loss": 0.1297, "step": 1890 }, { "epoch": 1.736455463728191, "grad_norm": 0.9297202229499817, "learning_rate": 8.99225922630522e-06, "loss": 0.0259, "step": 1891 }, { "epoch": 1.7373737373737375, "grad_norm": 0.45863422751426697, "learning_rate": 8.981632126094559e-06, "loss": 0.0096, "step": 1892 }, { "epoch": 1.7382920110192837, "grad_norm": 0.5913201570510864, "learning_rate": 8.971006187905881e-06, "loss": 0.0148, "step": 1893 }, { "epoch": 1.73921028466483, "grad_norm": 1.3957091569900513, "learning_rate": 8.960381423864058e-06, "loss": 0.0515, "step": 1894 }, { "epoch": 1.7401285583103765, "grad_norm": 1.744573712348938, "learning_rate": 8.949757846092615e-06, "loss": 0.065, "step": 1895 }, { "epoch": 1.741046831955923, "grad_norm": 0.6096499562263489, "learning_rate": 8.939135466713715e-06, "loss": 0.0062, "step": 1896 }, { "epoch": 1.7419651056014693, "grad_norm": 4.103384017944336, "learning_rate": 8.928514297848176e-06, "loss": 0.3488, "step": 1897 }, { "epoch": 1.7428833792470155, "grad_norm": 2.345700740814209, "learning_rate": 8.917894351615418e-06, "loss": 0.2028, "step": 1898 }, { "epoch": 1.743801652892562, "grad_norm": 1.0437172651290894, "learning_rate": 8.907275640133468e-06, "loss": 0.0303, "step": 1899 }, { "epoch": 1.7447199265381084, "grad_norm": 0.0679100975394249, "learning_rate": 8.896658175518944e-06, "loss": 0.002, "step": 1900 }, { "epoch": 1.7456382001836546, "grad_norm": 1.7762070894241333, "learning_rate": 8.886041969887041e-06, "loss": 0.041, "step": 1901 }, { "epoch": 1.7465564738292012, "grad_norm": 1.66452157497406, "learning_rate": 8.875427035351529e-06, "loss": 0.0696, "step": 1902 }, { "epoch": 1.7474747474747474, "grad_norm": 1.3901065587997437, "learning_rate": 8.86481338402471e-06, "loss": 0.0735, "step": 1903 }, { "epoch": 1.7483930211202938, "grad_norm": 0.22653169929981232, "learning_rate": 8.854201028017433e-06, "loss": 0.0061, "step": 1904 }, { "epoch": 1.7493112947658402, "grad_norm": 0.18256448209285736, "learning_rate": 8.843589979439058e-06, "loss": 0.0037, "step": 1905 }, { "epoch": 1.7502295684113864, "grad_norm": 0.23818093538284302, "learning_rate": 8.832980250397473e-06, "loss": 0.0068, "step": 1906 }, { "epoch": 1.751147842056933, "grad_norm": 0.7092468738555908, "learning_rate": 8.82237185299904e-06, "loss": 0.0212, "step": 1907 }, { "epoch": 1.7520661157024793, "grad_norm": 1.2277172803878784, "learning_rate": 8.811764799348615e-06, "loss": 0.061, "step": 1908 }, { "epoch": 1.7529843893480257, "grad_norm": 0.16623364388942719, "learning_rate": 8.801159101549512e-06, "loss": 0.0044, "step": 1909 }, { "epoch": 1.7539026629935721, "grad_norm": 0.2956843078136444, "learning_rate": 8.790554771703496e-06, "loss": 0.007, "step": 1910 }, { "epoch": 1.7548209366391183, "grad_norm": 0.09338343143463135, "learning_rate": 8.779951821910787e-06, "loss": 0.0024, "step": 1911 }, { "epoch": 1.755739210284665, "grad_norm": 1.5891085863113403, "learning_rate": 8.769350264270012e-06, "loss": 0.0923, "step": 1912 }, { "epoch": 1.7566574839302111, "grad_norm": 0.2662743926048279, "learning_rate": 8.758750110878222e-06, "loss": 0.0059, "step": 1913 }, { "epoch": 1.7575757575757576, "grad_norm": 1.6645762920379639, "learning_rate": 8.748151373830857e-06, "loss": 0.0492, "step": 1914 }, { "epoch": 1.758494031221304, "grad_norm": 0.16139045357704163, "learning_rate": 8.737554065221744e-06, "loss": 0.0045, "step": 1915 }, { "epoch": 1.7594123048668502, "grad_norm": 0.03848852589726448, "learning_rate": 8.726958197143083e-06, "loss": 0.0012, "step": 1916 }, { "epoch": 1.7603305785123968, "grad_norm": 1.2606416940689087, "learning_rate": 8.716363781685428e-06, "loss": 0.097, "step": 1917 }, { "epoch": 1.761248852157943, "grad_norm": 0.47651875019073486, "learning_rate": 8.705770830937675e-06, "loss": 0.009, "step": 1918 }, { "epoch": 1.7621671258034894, "grad_norm": 0.09705628454685211, "learning_rate": 8.695179356987049e-06, "loss": 0.0038, "step": 1919 }, { "epoch": 1.7630853994490359, "grad_norm": 1.7799177169799805, "learning_rate": 8.684589371919087e-06, "loss": 0.0925, "step": 1920 }, { "epoch": 1.764003673094582, "grad_norm": 1.578963279724121, "learning_rate": 8.674000887817638e-06, "loss": 0.1804, "step": 1921 }, { "epoch": 1.7649219467401287, "grad_norm": 0.6966227889060974, "learning_rate": 8.663413916764819e-06, "loss": 0.0169, "step": 1922 }, { "epoch": 1.765840220385675, "grad_norm": 2.2828238010406494, "learning_rate": 8.65282847084104e-06, "loss": 0.0872, "step": 1923 }, { "epoch": 1.7667584940312213, "grad_norm": 3.1307992935180664, "learning_rate": 8.642244562124962e-06, "loss": 0.1253, "step": 1924 }, { "epoch": 1.7676767676767677, "grad_norm": 0.7003632187843323, "learning_rate": 8.631662202693489e-06, "loss": 0.0264, "step": 1925 }, { "epoch": 1.768595041322314, "grad_norm": 1.6621145009994507, "learning_rate": 8.621081404621765e-06, "loss": 0.0781, "step": 1926 }, { "epoch": 1.7695133149678606, "grad_norm": 1.3620069026947021, "learning_rate": 8.610502179983145e-06, "loss": 0.0475, "step": 1927 }, { "epoch": 1.7704315886134068, "grad_norm": 1.1037182807922363, "learning_rate": 8.59992454084919e-06, "loss": 0.0754, "step": 1928 }, { "epoch": 1.7713498622589532, "grad_norm": 0.6177394390106201, "learning_rate": 8.589348499289658e-06, "loss": 0.0131, "step": 1929 }, { "epoch": 1.7722681359044996, "grad_norm": 2.5246846675872803, "learning_rate": 8.578774067372474e-06, "loss": 0.061, "step": 1930 }, { "epoch": 1.7731864095500458, "grad_norm": 1.3525375127792358, "learning_rate": 8.56820125716374e-06, "loss": 0.1265, "step": 1931 }, { "epoch": 1.7741046831955924, "grad_norm": 3.2416880130767822, "learning_rate": 8.557630080727694e-06, "loss": 0.0609, "step": 1932 }, { "epoch": 1.7750229568411386, "grad_norm": 0.4194507598876953, "learning_rate": 8.547060550126715e-06, "loss": 0.0056, "step": 1933 }, { "epoch": 1.775941230486685, "grad_norm": 0.31495994329452515, "learning_rate": 8.536492677421301e-06, "loss": 0.0112, "step": 1934 }, { "epoch": 1.7768595041322315, "grad_norm": 0.6241561770439148, "learning_rate": 8.525926474670073e-06, "loss": 0.0147, "step": 1935 }, { "epoch": 1.7777777777777777, "grad_norm": 1.261527419090271, "learning_rate": 8.515361953929723e-06, "loss": 0.0617, "step": 1936 }, { "epoch": 1.778696051423324, "grad_norm": 0.32504552602767944, "learning_rate": 8.504799127255043e-06, "loss": 0.0103, "step": 1937 }, { "epoch": 1.7796143250688705, "grad_norm": 1.9505643844604492, "learning_rate": 8.49423800669888e-06, "loss": 0.0633, "step": 1938 }, { "epoch": 1.780532598714417, "grad_norm": 0.8821505308151245, "learning_rate": 8.483678604312137e-06, "loss": 0.0225, "step": 1939 }, { "epoch": 1.7814508723599634, "grad_norm": 3.5006396770477295, "learning_rate": 8.473120932143766e-06, "loss": 0.0457, "step": 1940 }, { "epoch": 1.7823691460055096, "grad_norm": 1.312477469444275, "learning_rate": 8.462565002240733e-06, "loss": 0.0402, "step": 1941 }, { "epoch": 1.783287419651056, "grad_norm": 1.4448729753494263, "learning_rate": 8.452010826648019e-06, "loss": 0.116, "step": 1942 }, { "epoch": 1.7842056932966024, "grad_norm": 3.006804943084717, "learning_rate": 8.441458417408604e-06, "loss": 0.204, "step": 1943 }, { "epoch": 1.7851239669421488, "grad_norm": 2.670544147491455, "learning_rate": 8.43090778656345e-06, "loss": 0.1746, "step": 1944 }, { "epoch": 1.7860422405876952, "grad_norm": 0.5436847805976868, "learning_rate": 8.4203589461515e-06, "loss": 0.0147, "step": 1945 }, { "epoch": 1.7869605142332414, "grad_norm": 0.5676372051239014, "learning_rate": 8.409811908209641e-06, "loss": 0.015, "step": 1946 }, { "epoch": 1.7878787878787878, "grad_norm": 0.6405561566352844, "learning_rate": 8.399266684772712e-06, "loss": 0.0175, "step": 1947 }, { "epoch": 1.7887970615243343, "grad_norm": 1.4908260107040405, "learning_rate": 8.388723287873474e-06, "loss": 0.0679, "step": 1948 }, { "epoch": 1.7897153351698805, "grad_norm": 0.9289946556091309, "learning_rate": 8.378181729542607e-06, "loss": 0.0273, "step": 1949 }, { "epoch": 1.790633608815427, "grad_norm": 0.1536717712879181, "learning_rate": 8.3676420218087e-06, "loss": 0.0053, "step": 1950 }, { "epoch": 1.7915518824609733, "grad_norm": 2.2505316734313965, "learning_rate": 8.357104176698223e-06, "loss": 0.0814, "step": 1951 }, { "epoch": 1.7924701561065197, "grad_norm": 0.10282382369041443, "learning_rate": 8.346568206235519e-06, "loss": 0.0029, "step": 1952 }, { "epoch": 1.7933884297520661, "grad_norm": 0.8414713144302368, "learning_rate": 8.336034122442796e-06, "loss": 0.0137, "step": 1953 }, { "epoch": 1.7943067033976123, "grad_norm": 1.7413733005523682, "learning_rate": 8.325501937340105e-06, "loss": 0.1634, "step": 1954 }, { "epoch": 1.795224977043159, "grad_norm": 0.656270444393158, "learning_rate": 8.314971662945342e-06, "loss": 0.0305, "step": 1955 }, { "epoch": 1.7961432506887052, "grad_norm": 1.6164568662643433, "learning_rate": 8.304443311274209e-06, "loss": 0.0224, "step": 1956 }, { "epoch": 1.7970615243342516, "grad_norm": 0.0480070635676384, "learning_rate": 8.29391689434022e-06, "loss": 0.0016, "step": 1957 }, { "epoch": 1.797979797979798, "grad_norm": 2.542511463165283, "learning_rate": 8.28339242415468e-06, "loss": 0.2732, "step": 1958 }, { "epoch": 1.7988980716253442, "grad_norm": 1.9929057359695435, "learning_rate": 8.272869912726668e-06, "loss": 0.1134, "step": 1959 }, { "epoch": 1.7998163452708908, "grad_norm": 0.08823496848344803, "learning_rate": 8.262349372063047e-06, "loss": 0.0024, "step": 1960 }, { "epoch": 1.800734618916437, "grad_norm": 1.2352956533432007, "learning_rate": 8.251830814168408e-06, "loss": 0.1053, "step": 1961 }, { "epoch": 1.8016528925619835, "grad_norm": 1.3229180574417114, "learning_rate": 8.241314251045092e-06, "loss": 0.0554, "step": 1962 }, { "epoch": 1.8025711662075299, "grad_norm": 1.7665026187896729, "learning_rate": 8.230799694693155e-06, "loss": 0.0956, "step": 1963 }, { "epoch": 1.803489439853076, "grad_norm": 1.0615016222000122, "learning_rate": 8.22028715711038e-06, "loss": 0.0407, "step": 1964 }, { "epoch": 1.8044077134986227, "grad_norm": 0.5939935445785522, "learning_rate": 8.209776650292228e-06, "loss": 0.0157, "step": 1965 }, { "epoch": 1.805325987144169, "grad_norm": 0.515291154384613, "learning_rate": 8.199268186231852e-06, "loss": 0.0146, "step": 1966 }, { "epoch": 1.8062442607897153, "grad_norm": 1.234570026397705, "learning_rate": 8.188761776920073e-06, "loss": 0.0739, "step": 1967 }, { "epoch": 1.8071625344352618, "grad_norm": 1.983948826789856, "learning_rate": 8.178257434345365e-06, "loss": 0.1149, "step": 1968 }, { "epoch": 1.808080808080808, "grad_norm": 0.7337007522583008, "learning_rate": 8.167755170493845e-06, "loss": 0.0279, "step": 1969 }, { "epoch": 1.8089990817263546, "grad_norm": 1.9738246202468872, "learning_rate": 8.157254997349264e-06, "loss": 0.0853, "step": 1970 }, { "epoch": 1.8099173553719008, "grad_norm": 1.6931456327438354, "learning_rate": 8.146756926892976e-06, "loss": 0.0346, "step": 1971 }, { "epoch": 1.8108356290174472, "grad_norm": 0.09421636164188385, "learning_rate": 8.136260971103946e-06, "loss": 0.0026, "step": 1972 }, { "epoch": 1.8117539026629936, "grad_norm": 0.2024223804473877, "learning_rate": 8.125767141958718e-06, "loss": 0.0055, "step": 1973 }, { "epoch": 1.8126721763085398, "grad_norm": 0.40461841225624084, "learning_rate": 8.115275451431415e-06, "loss": 0.0184, "step": 1974 }, { "epoch": 1.8135904499540865, "grad_norm": 1.4996925592422485, "learning_rate": 8.10478591149372e-06, "loss": 0.061, "step": 1975 }, { "epoch": 1.8145087235996327, "grad_norm": 0.29058635234832764, "learning_rate": 8.094298534114852e-06, "loss": 0.0054, "step": 1976 }, { "epoch": 1.815426997245179, "grad_norm": 2.329068660736084, "learning_rate": 8.083813331261578e-06, "loss": 0.075, "step": 1977 }, { "epoch": 1.8163452708907255, "grad_norm": 1.2829796075820923, "learning_rate": 8.073330314898172e-06, "loss": 0.1409, "step": 1978 }, { "epoch": 1.8172635445362717, "grad_norm": 1.445169448852539, "learning_rate": 8.06284949698642e-06, "loss": 0.0846, "step": 1979 }, { "epoch": 1.8181818181818183, "grad_norm": 1.9105759859085083, "learning_rate": 8.052370889485594e-06, "loss": 0.0453, "step": 1980 }, { "epoch": 1.8191000918273645, "grad_norm": 0.47017332911491394, "learning_rate": 8.041894504352448e-06, "loss": 0.0086, "step": 1981 }, { "epoch": 1.820018365472911, "grad_norm": 1.7531079053878784, "learning_rate": 8.031420353541195e-06, "loss": 0.1241, "step": 1982 }, { "epoch": 1.8209366391184574, "grad_norm": 1.6215425729751587, "learning_rate": 8.020948449003503e-06, "loss": 0.1949, "step": 1983 }, { "epoch": 1.8218549127640036, "grad_norm": 1.19618558883667, "learning_rate": 8.01047880268848e-06, "loss": 0.0696, "step": 1984 }, { "epoch": 1.82277318640955, "grad_norm": 0.07511110603809357, "learning_rate": 8.00001142654265e-06, "loss": 0.0016, "step": 1985 }, { "epoch": 1.8236914600550964, "grad_norm": 2.126824140548706, "learning_rate": 7.989546332509947e-06, "loss": 0.1042, "step": 1986 }, { "epoch": 1.8246097337006428, "grad_norm": 1.4427239894866943, "learning_rate": 7.979083532531702e-06, "loss": 0.0786, "step": 1987 }, { "epoch": 1.8255280073461893, "grad_norm": 1.2253066301345825, "learning_rate": 7.968623038546638e-06, "loss": 0.0568, "step": 1988 }, { "epoch": 1.8264462809917354, "grad_norm": 0.3413166403770447, "learning_rate": 7.958164862490834e-06, "loss": 0.0042, "step": 1989 }, { "epoch": 1.8273645546372819, "grad_norm": 0.20297518372535706, "learning_rate": 7.947709016297729e-06, "loss": 0.0072, "step": 1990 }, { "epoch": 1.8282828282828283, "grad_norm": 0.09956304728984833, "learning_rate": 7.937255511898102e-06, "loss": 0.0025, "step": 1991 }, { "epoch": 1.8292011019283747, "grad_norm": 1.1128499507904053, "learning_rate": 7.926804361220056e-06, "loss": 0.0227, "step": 1992 }, { "epoch": 1.8301193755739211, "grad_norm": 1.3692017793655396, "learning_rate": 7.916355576189024e-06, "loss": 0.047, "step": 1993 }, { "epoch": 1.8310376492194673, "grad_norm": 2.383944034576416, "learning_rate": 7.905909168727723e-06, "loss": 0.1433, "step": 1994 }, { "epoch": 1.8319559228650137, "grad_norm": 0.25518375635147095, "learning_rate": 7.895465150756166e-06, "loss": 0.0071, "step": 1995 }, { "epoch": 1.8328741965105602, "grad_norm": 0.9924852252006531, "learning_rate": 7.885023534191631e-06, "loss": 0.0276, "step": 1996 }, { "epoch": 1.8337924701561064, "grad_norm": 0.8239807486534119, "learning_rate": 7.874584330948661e-06, "loss": 0.0214, "step": 1997 }, { "epoch": 1.834710743801653, "grad_norm": 1.7210898399353027, "learning_rate": 7.864147552939052e-06, "loss": 0.0518, "step": 1998 }, { "epoch": 1.8356290174471992, "grad_norm": 0.41743823885917664, "learning_rate": 7.853713212071825e-06, "loss": 0.0159, "step": 1999 }, { "epoch": 1.8365472910927456, "grad_norm": 1.013236403465271, "learning_rate": 7.843281320253218e-06, "loss": 0.0889, "step": 2000 }, { "epoch": 1.837465564738292, "grad_norm": 0.6579924821853638, "learning_rate": 7.832851889386678e-06, "loss": 0.0171, "step": 2001 }, { "epoch": 1.8383838383838382, "grad_norm": 2.4705898761749268, "learning_rate": 7.822424931372838e-06, "loss": 0.0555, "step": 2002 }, { "epoch": 1.8393021120293849, "grad_norm": 1.438650369644165, "learning_rate": 7.812000458109527e-06, "loss": 0.0508, "step": 2003 }, { "epoch": 1.840220385674931, "grad_norm": 2.9840829372406006, "learning_rate": 7.801578481491721e-06, "loss": 0.086, "step": 2004 }, { "epoch": 1.8411386593204775, "grad_norm": 1.2166460752487183, "learning_rate": 7.79115901341155e-06, "loss": 0.0361, "step": 2005 }, { "epoch": 1.842056932966024, "grad_norm": 0.8396257758140564, "learning_rate": 7.78074206575829e-06, "loss": 0.0345, "step": 2006 }, { "epoch": 1.84297520661157, "grad_norm": 1.2402950525283813, "learning_rate": 7.770327650418326e-06, "loss": 0.0474, "step": 2007 }, { "epoch": 1.8438934802571167, "grad_norm": 3.738618850708008, "learning_rate": 7.759915779275173e-06, "loss": 0.0969, "step": 2008 }, { "epoch": 1.844811753902663, "grad_norm": 3.0763535499572754, "learning_rate": 7.749506464209428e-06, "loss": 0.0378, "step": 2009 }, { "epoch": 1.8457300275482094, "grad_norm": 1.8393634557724, "learning_rate": 7.73909971709878e-06, "loss": 0.1837, "step": 2010 }, { "epoch": 1.8466483011937558, "grad_norm": 0.274227499961853, "learning_rate": 7.728695549817977e-06, "loss": 0.0083, "step": 2011 }, { "epoch": 1.847566574839302, "grad_norm": 0.6241540312767029, "learning_rate": 7.718293974238832e-06, "loss": 0.0162, "step": 2012 }, { "epoch": 1.8484848484848486, "grad_norm": 0.7866369485855103, "learning_rate": 7.707895002230205e-06, "loss": 0.0263, "step": 2013 }, { "epoch": 1.8494031221303948, "grad_norm": 1.6238106489181519, "learning_rate": 7.697498645657974e-06, "loss": 0.1199, "step": 2014 }, { "epoch": 1.8503213957759412, "grad_norm": 0.20126470923423767, "learning_rate": 7.68710491638504e-06, "loss": 0.0031, "step": 2015 }, { "epoch": 1.8512396694214877, "grad_norm": 1.9415687322616577, "learning_rate": 7.6767138262713e-06, "loss": 0.0235, "step": 2016 }, { "epoch": 1.8521579430670339, "grad_norm": 0.18885232508182526, "learning_rate": 7.666325387173646e-06, "loss": 0.0039, "step": 2017 }, { "epoch": 1.8530762167125805, "grad_norm": 0.2007027566432953, "learning_rate": 7.655939610945943e-06, "loss": 0.0052, "step": 2018 }, { "epoch": 1.8539944903581267, "grad_norm": 2.770440101623535, "learning_rate": 7.645556509439016e-06, "loss": 0.1197, "step": 2019 }, { "epoch": 1.854912764003673, "grad_norm": 0.766726553440094, "learning_rate": 7.63517609450064e-06, "loss": 0.0097, "step": 2020 }, { "epoch": 1.8558310376492195, "grad_norm": 2.0974855422973633, "learning_rate": 7.62479837797552e-06, "loss": 0.0777, "step": 2021 }, { "epoch": 1.8567493112947657, "grad_norm": 0.9071203470230103, "learning_rate": 7.614423371705288e-06, "loss": 0.0413, "step": 2022 }, { "epoch": 1.8576675849403124, "grad_norm": 0.9079411625862122, "learning_rate": 7.604051087528479e-06, "loss": 0.0135, "step": 2023 }, { "epoch": 1.8585858585858586, "grad_norm": 0.18213620781898499, "learning_rate": 7.593681537280527e-06, "loss": 0.0053, "step": 2024 }, { "epoch": 1.859504132231405, "grad_norm": 0.8660861253738403, "learning_rate": 7.5833147327937406e-06, "loss": 0.0176, "step": 2025 }, { "epoch": 1.8604224058769514, "grad_norm": 0.8693211674690247, "learning_rate": 7.572950685897295e-06, "loss": 0.0151, "step": 2026 }, { "epoch": 1.8613406795224976, "grad_norm": 0.2985236346721649, "learning_rate": 7.5625894084172255e-06, "loss": 0.008, "step": 2027 }, { "epoch": 1.8622589531680442, "grad_norm": 1.2316169738769531, "learning_rate": 7.552230912176403e-06, "loss": 0.0325, "step": 2028 }, { "epoch": 1.8631772268135904, "grad_norm": 2.9183363914489746, "learning_rate": 7.541875208994521e-06, "loss": 0.2925, "step": 2029 }, { "epoch": 1.8640955004591369, "grad_norm": 0.07988791912794113, "learning_rate": 7.531522310688096e-06, "loss": 0.0025, "step": 2030 }, { "epoch": 1.8650137741046833, "grad_norm": 0.06463775783777237, "learning_rate": 7.521172229070433e-06, "loss": 0.002, "step": 2031 }, { "epoch": 1.8659320477502295, "grad_norm": 0.5702720284461975, "learning_rate": 7.510824975951634e-06, "loss": 0.009, "step": 2032 }, { "epoch": 1.866850321395776, "grad_norm": 0.4105806350708008, "learning_rate": 7.500480563138562e-06, "loss": 0.0091, "step": 2033 }, { "epoch": 1.8677685950413223, "grad_norm": 0.7005887627601624, "learning_rate": 7.4901390024348476e-06, "loss": 0.0108, "step": 2034 }, { "epoch": 1.8686868686868687, "grad_norm": 1.183706521987915, "learning_rate": 7.479800305640863e-06, "loss": 0.0627, "step": 2035 }, { "epoch": 1.8696051423324151, "grad_norm": 2.29952335357666, "learning_rate": 7.469464484553708e-06, "loss": 0.0973, "step": 2036 }, { "epoch": 1.8705234159779613, "grad_norm": 1.3787825107574463, "learning_rate": 7.459131550967216e-06, "loss": 0.0757, "step": 2037 }, { "epoch": 1.8714416896235078, "grad_norm": 0.1916346400976181, "learning_rate": 7.448801516671913e-06, "loss": 0.0048, "step": 2038 }, { "epoch": 1.8723599632690542, "grad_norm": 2.326059579849243, "learning_rate": 7.438474393455017e-06, "loss": 0.0484, "step": 2039 }, { "epoch": 1.8732782369146006, "grad_norm": 0.10259797424077988, "learning_rate": 7.4281501931004254e-06, "loss": 0.0022, "step": 2040 }, { "epoch": 1.874196510560147, "grad_norm": 0.30620619654655457, "learning_rate": 7.41782892738871e-06, "loss": 0.006, "step": 2041 }, { "epoch": 1.8751147842056932, "grad_norm": 2.504824638366699, "learning_rate": 7.407510608097081e-06, "loss": 0.1097, "step": 2042 }, { "epoch": 1.8760330578512396, "grad_norm": 0.9450139403343201, "learning_rate": 7.397195246999391e-06, "loss": 0.017, "step": 2043 }, { "epoch": 1.876951331496786, "grad_norm": 0.7712787389755249, "learning_rate": 7.386882855866121e-06, "loss": 0.0328, "step": 2044 }, { "epoch": 1.8778696051423323, "grad_norm": 0.3856699466705322, "learning_rate": 7.376573446464352e-06, "loss": 0.0113, "step": 2045 }, { "epoch": 1.878787878787879, "grad_norm": 2.512347936630249, "learning_rate": 7.366267030557781e-06, "loss": 0.0898, "step": 2046 }, { "epoch": 1.879706152433425, "grad_norm": 1.8299163579940796, "learning_rate": 7.355963619906674e-06, "loss": 0.0693, "step": 2047 }, { "epoch": 1.8806244260789715, "grad_norm": 0.0643567368388176, "learning_rate": 7.345663226267872e-06, "loss": 0.0017, "step": 2048 }, { "epoch": 1.881542699724518, "grad_norm": 0.12511347234249115, "learning_rate": 7.335365861394774e-06, "loss": 0.0036, "step": 2049 }, { "epoch": 1.8824609733700641, "grad_norm": 2.324901819229126, "learning_rate": 7.3250715370373185e-06, "loss": 0.0826, "step": 2050 }, { "epoch": 1.8833792470156108, "grad_norm": 0.09418908506631851, "learning_rate": 7.31478026494199e-06, "loss": 0.002, "step": 2051 }, { "epoch": 1.884297520661157, "grad_norm": 0.8457493782043457, "learning_rate": 7.3044920568517705e-06, "loss": 0.0074, "step": 2052 }, { "epoch": 1.8852157943067034, "grad_norm": 0.9803299307823181, "learning_rate": 7.294206924506159e-06, "loss": 0.0607, "step": 2053 }, { "epoch": 1.8861340679522498, "grad_norm": 0.0478668138384819, "learning_rate": 7.283924879641137e-06, "loss": 0.0016, "step": 2054 }, { "epoch": 1.887052341597796, "grad_norm": 0.18924634158611298, "learning_rate": 7.273645933989163e-06, "loss": 0.0052, "step": 2055 }, { "epoch": 1.8879706152433426, "grad_norm": 0.06043931841850281, "learning_rate": 7.263370099279173e-06, "loss": 0.0017, "step": 2056 }, { "epoch": 1.8888888888888888, "grad_norm": 1.0764315128326416, "learning_rate": 7.253097387236534e-06, "loss": 0.0578, "step": 2057 }, { "epoch": 1.8898071625344353, "grad_norm": 2.487588405609131, "learning_rate": 7.242827809583062e-06, "loss": 0.1132, "step": 2058 }, { "epoch": 1.8907254361799817, "grad_norm": 1.5013446807861328, "learning_rate": 7.232561378036993e-06, "loss": 0.1788, "step": 2059 }, { "epoch": 1.8916437098255279, "grad_norm": 0.10621438920497894, "learning_rate": 7.222298104312966e-06, "loss": 0.0035, "step": 2060 }, { "epoch": 1.8925619834710745, "grad_norm": 0.11578827351331711, "learning_rate": 7.212038000122035e-06, "loss": 0.0041, "step": 2061 }, { "epoch": 1.8934802571166207, "grad_norm": 2.0818071365356445, "learning_rate": 7.201781077171621e-06, "loss": 0.1334, "step": 2062 }, { "epoch": 1.8943985307621671, "grad_norm": 0.31853488087654114, "learning_rate": 7.191527347165521e-06, "loss": 0.0133, "step": 2063 }, { "epoch": 1.8953168044077136, "grad_norm": 1.6686853170394897, "learning_rate": 7.181276821803885e-06, "loss": 0.0699, "step": 2064 }, { "epoch": 1.8962350780532597, "grad_norm": 1.663484811782837, "learning_rate": 7.171029512783207e-06, "loss": 0.0941, "step": 2065 }, { "epoch": 1.8971533516988064, "grad_norm": 0.41326290369033813, "learning_rate": 7.160785431796322e-06, "loss": 0.0126, "step": 2066 }, { "epoch": 1.8980716253443526, "grad_norm": 1.122972011566162, "learning_rate": 7.1505445905323676e-06, "loss": 0.0189, "step": 2067 }, { "epoch": 1.898989898989899, "grad_norm": 0.8169239163398743, "learning_rate": 7.1403070006767895e-06, "loss": 0.0319, "step": 2068 }, { "epoch": 1.8999081726354454, "grad_norm": 1.2804996967315674, "learning_rate": 7.130072673911321e-06, "loss": 0.0304, "step": 2069 }, { "epoch": 1.9008264462809916, "grad_norm": 0.07489041984081268, "learning_rate": 7.119841621913981e-06, "loss": 0.0017, "step": 2070 }, { "epoch": 1.9017447199265383, "grad_norm": 1.3600448369979858, "learning_rate": 7.109613856359039e-06, "loss": 0.0415, "step": 2071 }, { "epoch": 1.9026629935720845, "grad_norm": 0.1946638524532318, "learning_rate": 7.099389388917024e-06, "loss": 0.0066, "step": 2072 }, { "epoch": 1.9035812672176309, "grad_norm": 1.8841586112976074, "learning_rate": 7.0891682312547e-06, "loss": 0.0531, "step": 2073 }, { "epoch": 1.9044995408631773, "grad_norm": 2.4795892238616943, "learning_rate": 7.078950395035044e-06, "loss": 0.1084, "step": 2074 }, { "epoch": 1.9054178145087235, "grad_norm": 3.0060341358184814, "learning_rate": 7.068735891917264e-06, "loss": 0.1161, "step": 2075 }, { "epoch": 1.9063360881542701, "grad_norm": 0.12741070985794067, "learning_rate": 7.0585247335567455e-06, "loss": 0.0034, "step": 2076 }, { "epoch": 1.9072543617998163, "grad_norm": 0.11244813352823257, "learning_rate": 7.048316931605062e-06, "loss": 0.0033, "step": 2077 }, { "epoch": 1.9081726354453628, "grad_norm": 2.4892804622650146, "learning_rate": 7.0381124977099656e-06, "loss": 0.0566, "step": 2078 }, { "epoch": 1.9090909090909092, "grad_norm": 0.7030420303344727, "learning_rate": 7.027911443515354e-06, "loss": 0.0361, "step": 2079 }, { "epoch": 1.9100091827364554, "grad_norm": 0.17325621843338013, "learning_rate": 7.01771378066128e-06, "loss": 0.0032, "step": 2080 }, { "epoch": 1.9109274563820018, "grad_norm": 1.2973414659500122, "learning_rate": 7.007519520783913e-06, "loss": 0.0642, "step": 2081 }, { "epoch": 1.9118457300275482, "grad_norm": 0.38100308179855347, "learning_rate": 6.9973286755155515e-06, "loss": 0.0083, "step": 2082 }, { "epoch": 1.9127640036730946, "grad_norm": 1.8109495639801025, "learning_rate": 6.987141256484588e-06, "loss": 0.0497, "step": 2083 }, { "epoch": 1.913682277318641, "grad_norm": 0.4843994081020355, "learning_rate": 6.9769572753155126e-06, "loss": 0.0116, "step": 2084 }, { "epoch": 1.9146005509641872, "grad_norm": 0.343118280172348, "learning_rate": 6.966776743628894e-06, "loss": 0.0058, "step": 2085 }, { "epoch": 1.9155188246097337, "grad_norm": 1.6198872327804565, "learning_rate": 6.956599673041358e-06, "loss": 0.0635, "step": 2086 }, { "epoch": 1.91643709825528, "grad_norm": 0.7020811438560486, "learning_rate": 6.946426075165585e-06, "loss": 0.0268, "step": 2087 }, { "epoch": 1.9173553719008265, "grad_norm": 1.8222291469573975, "learning_rate": 6.936255961610289e-06, "loss": 0.0624, "step": 2088 }, { "epoch": 1.918273645546373, "grad_norm": 3.1112844944000244, "learning_rate": 6.926089343980209e-06, "loss": 0.0451, "step": 2089 }, { "epoch": 1.9191919191919191, "grad_norm": 0.2204694300889969, "learning_rate": 6.915926233876107e-06, "loss": 0.0045, "step": 2090 }, { "epoch": 1.9201101928374655, "grad_norm": 1.6323190927505493, "learning_rate": 6.905766642894727e-06, "loss": 0.0334, "step": 2091 }, { "epoch": 1.921028466483012, "grad_norm": 1.2104750871658325, "learning_rate": 6.895610582628802e-06, "loss": 0.0544, "step": 2092 }, { "epoch": 1.9219467401285582, "grad_norm": 1.5693645477294922, "learning_rate": 6.885458064667041e-06, "loss": 0.0301, "step": 2093 }, { "epoch": 1.9228650137741048, "grad_norm": 1.171659231185913, "learning_rate": 6.875309100594098e-06, "loss": 0.0581, "step": 2094 }, { "epoch": 1.923783287419651, "grad_norm": 2.5417490005493164, "learning_rate": 6.865163701990596e-06, "loss": 0.2362, "step": 2095 }, { "epoch": 1.9247015610651974, "grad_norm": 0.3172614574432373, "learning_rate": 6.855021880433064e-06, "loss": 0.0072, "step": 2096 }, { "epoch": 1.9256198347107438, "grad_norm": 1.0319018363952637, "learning_rate": 6.844883647493965e-06, "loss": 0.0397, "step": 2097 }, { "epoch": 1.92653810835629, "grad_norm": 1.7878608703613281, "learning_rate": 6.834749014741657e-06, "loss": 0.1334, "step": 2098 }, { "epoch": 1.9274563820018367, "grad_norm": 0.0629027783870697, "learning_rate": 6.824617993740402e-06, "loss": 0.0022, "step": 2099 }, { "epoch": 1.9283746556473829, "grad_norm": 1.3675612211227417, "learning_rate": 6.8144905960503315e-06, "loss": 0.0583, "step": 2100 }, { "epoch": 1.9292929292929293, "grad_norm": 2.536353826522827, "learning_rate": 6.804366833227445e-06, "loss": 0.1818, "step": 2101 }, { "epoch": 1.9302112029384757, "grad_norm": 0.15887629985809326, "learning_rate": 6.7942467168235944e-06, "loss": 0.0029, "step": 2102 }, { "epoch": 1.931129476584022, "grad_norm": 0.42315155267715454, "learning_rate": 6.784130258386467e-06, "loss": 0.0176, "step": 2103 }, { "epoch": 1.9320477502295685, "grad_norm": 3.181713104248047, "learning_rate": 6.774017469459587e-06, "loss": 0.2942, "step": 2104 }, { "epoch": 1.9329660238751147, "grad_norm": 1.1815608739852905, "learning_rate": 6.763908361582282e-06, "loss": 0.0629, "step": 2105 }, { "epoch": 1.9338842975206612, "grad_norm": 1.0976064205169678, "learning_rate": 6.75380294628968e-06, "loss": 0.0365, "step": 2106 }, { "epoch": 1.9348025711662076, "grad_norm": 1.9788634777069092, "learning_rate": 6.743701235112701e-06, "loss": 0.0566, "step": 2107 }, { "epoch": 1.9357208448117538, "grad_norm": 0.34314829111099243, "learning_rate": 6.733603239578026e-06, "loss": 0.0101, "step": 2108 }, { "epoch": 1.9366391184573004, "grad_norm": 2.6075282096862793, "learning_rate": 6.7235089712081165e-06, "loss": 0.0897, "step": 2109 }, { "epoch": 1.9375573921028466, "grad_norm": 0.16788259148597717, "learning_rate": 6.7134184415211635e-06, "loss": 0.0046, "step": 2110 }, { "epoch": 1.938475665748393, "grad_norm": 1.1367123126983643, "learning_rate": 6.7033316620310985e-06, "loss": 0.0814, "step": 2111 }, { "epoch": 1.9393939393939394, "grad_norm": 2.257824659347534, "learning_rate": 6.693248644247574e-06, "loss": 0.1011, "step": 2112 }, { "epoch": 1.9403122130394856, "grad_norm": 2.3508434295654297, "learning_rate": 6.683169399675944e-06, "loss": 0.0483, "step": 2113 }, { "epoch": 1.9412304866850323, "grad_norm": 0.171247661113739, "learning_rate": 6.6730939398172714e-06, "loss": 0.0054, "step": 2114 }, { "epoch": 1.9421487603305785, "grad_norm": 0.608050525188446, "learning_rate": 6.663022276168287e-06, "loss": 0.0128, "step": 2115 }, { "epoch": 1.943067033976125, "grad_norm": 0.25301969051361084, "learning_rate": 6.652954420221394e-06, "loss": 0.0065, "step": 2116 }, { "epoch": 1.9439853076216713, "grad_norm": 0.31126999855041504, "learning_rate": 6.642890383464653e-06, "loss": 0.008, "step": 2117 }, { "epoch": 1.9449035812672175, "grad_norm": 1.712687611579895, "learning_rate": 6.632830177381757e-06, "loss": 0.0276, "step": 2118 }, { "epoch": 1.9458218549127642, "grad_norm": 1.0572303533554077, "learning_rate": 6.622773813452048e-06, "loss": 0.033, "step": 2119 }, { "epoch": 1.9467401285583104, "grad_norm": 1.615032434463501, "learning_rate": 6.612721303150467e-06, "loss": 0.0666, "step": 2120 }, { "epoch": 1.9476584022038568, "grad_norm": 0.646139919757843, "learning_rate": 6.602672657947559e-06, "loss": 0.0129, "step": 2121 }, { "epoch": 1.9485766758494032, "grad_norm": 1.1124688386917114, "learning_rate": 6.592627889309464e-06, "loss": 0.0512, "step": 2122 }, { "epoch": 1.9494949494949494, "grad_norm": 0.08152138441801071, "learning_rate": 6.582587008697898e-06, "loss": 0.0018, "step": 2123 }, { "epoch": 1.950413223140496, "grad_norm": 1.9765936136245728, "learning_rate": 6.572550027570137e-06, "loss": 0.173, "step": 2124 }, { "epoch": 1.9513314967860422, "grad_norm": 0.36117249727249146, "learning_rate": 6.562516957379012e-06, "loss": 0.0105, "step": 2125 }, { "epoch": 1.9522497704315886, "grad_norm": 0.5704871416091919, "learning_rate": 6.5524878095728895e-06, "loss": 0.0119, "step": 2126 }, { "epoch": 1.953168044077135, "grad_norm": 0.9174497723579407, "learning_rate": 6.542462595595655e-06, "loss": 0.0357, "step": 2127 }, { "epoch": 1.9540863177226813, "grad_norm": 0.06658988445997238, "learning_rate": 6.532441326886716e-06, "loss": 0.0016, "step": 2128 }, { "epoch": 1.9550045913682277, "grad_norm": 1.9793950319290161, "learning_rate": 6.522424014880971e-06, "loss": 0.0828, "step": 2129 }, { "epoch": 1.955922865013774, "grad_norm": 2.313861846923828, "learning_rate": 6.512410671008801e-06, "loss": 0.1522, "step": 2130 }, { "epoch": 1.9568411386593205, "grad_norm": 1.5600991249084473, "learning_rate": 6.502401306696065e-06, "loss": 0.0304, "step": 2131 }, { "epoch": 1.957759412304867, "grad_norm": 0.7640740275382996, "learning_rate": 6.492395933364077e-06, "loss": 0.0201, "step": 2132 }, { "epoch": 1.9586776859504131, "grad_norm": 0.30501800775527954, "learning_rate": 6.482394562429607e-06, "loss": 0.0075, "step": 2133 }, { "epoch": 1.9595959595959596, "grad_norm": 0.839054524898529, "learning_rate": 6.472397205304844e-06, "loss": 0.0237, "step": 2134 }, { "epoch": 1.960514233241506, "grad_norm": 1.8250335454940796, "learning_rate": 6.4624038733974034e-06, "loss": 0.0271, "step": 2135 }, { "epoch": 1.9614325068870524, "grad_norm": 0.4306468069553375, "learning_rate": 6.452414578110309e-06, "loss": 0.0114, "step": 2136 }, { "epoch": 1.9623507805325988, "grad_norm": 1.8252085447311401, "learning_rate": 6.442429330841972e-06, "loss": 0.0945, "step": 2137 }, { "epoch": 1.963269054178145, "grad_norm": 0.015572148375213146, "learning_rate": 6.432448142986196e-06, "loss": 0.0005, "step": 2138 }, { "epoch": 1.9641873278236914, "grad_norm": 0.10909423232078552, "learning_rate": 6.422471025932145e-06, "loss": 0.003, "step": 2139 }, { "epoch": 1.9651056014692379, "grad_norm": 1.811461091041565, "learning_rate": 6.412497991064337e-06, "loss": 0.0624, "step": 2140 }, { "epoch": 1.966023875114784, "grad_norm": 1.3465752601623535, "learning_rate": 6.402529049762634e-06, "loss": 0.026, "step": 2141 }, { "epoch": 1.9669421487603307, "grad_norm": 0.4730839133262634, "learning_rate": 6.392564213402221e-06, "loss": 0.0107, "step": 2142 }, { "epoch": 1.9678604224058769, "grad_norm": 1.6159652471542358, "learning_rate": 6.382603493353619e-06, "loss": 0.031, "step": 2143 }, { "epoch": 1.9687786960514233, "grad_norm": 0.23660066723823547, "learning_rate": 6.372646900982626e-06, "loss": 0.0065, "step": 2144 }, { "epoch": 1.9696969696969697, "grad_norm": 0.5095412135124207, "learning_rate": 6.3626944476503485e-06, "loss": 0.0151, "step": 2145 }, { "epoch": 1.970615243342516, "grad_norm": 0.43246912956237793, "learning_rate": 6.352746144713161e-06, "loss": 0.013, "step": 2146 }, { "epoch": 1.9715335169880626, "grad_norm": 1.2010942697525024, "learning_rate": 6.3428020035227e-06, "loss": 0.0247, "step": 2147 }, { "epoch": 1.9724517906336088, "grad_norm": 0.19087471067905426, "learning_rate": 6.3328620354258666e-06, "loss": 0.0037, "step": 2148 }, { "epoch": 1.9733700642791552, "grad_norm": 0.27245479822158813, "learning_rate": 6.322926251764788e-06, "loss": 0.0052, "step": 2149 }, { "epoch": 1.9742883379247016, "grad_norm": 0.4757018983364105, "learning_rate": 6.31299466387682e-06, "loss": 0.0138, "step": 2150 }, { "epoch": 1.9752066115702478, "grad_norm": 0.34683090448379517, "learning_rate": 6.303067283094526e-06, "loss": 0.009, "step": 2151 }, { "epoch": 1.9761248852157944, "grad_norm": 0.39308103919029236, "learning_rate": 6.293144120745682e-06, "loss": 0.0159, "step": 2152 }, { "epoch": 1.9770431588613406, "grad_norm": 0.9858159422874451, "learning_rate": 6.28322518815324e-06, "loss": 0.0452, "step": 2153 }, { "epoch": 1.977961432506887, "grad_norm": 0.2661818861961365, "learning_rate": 6.273310496635325e-06, "loss": 0.0069, "step": 2154 }, { "epoch": 1.9788797061524335, "grad_norm": 3.453636407852173, "learning_rate": 6.263400057505228e-06, "loss": 0.1105, "step": 2155 }, { "epoch": 1.9797979797979797, "grad_norm": 1.463174819946289, "learning_rate": 6.2534938820713794e-06, "loss": 0.0614, "step": 2156 }, { "epoch": 1.9807162534435263, "grad_norm": 2.0949974060058594, "learning_rate": 6.2435919816373595e-06, "loss": 0.071, "step": 2157 }, { "epoch": 1.9816345270890725, "grad_norm": 0.6053652763366699, "learning_rate": 6.233694367501856e-06, "loss": 0.017, "step": 2158 }, { "epoch": 1.982552800734619, "grad_norm": 0.6706982254981995, "learning_rate": 6.2238010509586725e-06, "loss": 0.0199, "step": 2159 }, { "epoch": 1.9834710743801653, "grad_norm": 2.493802070617676, "learning_rate": 6.213912043296706e-06, "loss": 0.0526, "step": 2160 }, { "epoch": 1.9843893480257115, "grad_norm": 0.4530322551727295, "learning_rate": 6.204027355799935e-06, "loss": 0.015, "step": 2161 }, { "epoch": 1.9853076216712582, "grad_norm": 0.11862383782863617, "learning_rate": 6.194146999747419e-06, "loss": 0.0016, "step": 2162 }, { "epoch": 1.9862258953168044, "grad_norm": 2.328131914138794, "learning_rate": 6.184270986413263e-06, "loss": 0.1496, "step": 2163 }, { "epoch": 1.9871441689623508, "grad_norm": 2.1780295372009277, "learning_rate": 6.174399327066624e-06, "loss": 0.1515, "step": 2164 }, { "epoch": 1.9880624426078972, "grad_norm": 1.1135084629058838, "learning_rate": 6.164532032971684e-06, "loss": 0.0112, "step": 2165 }, { "epoch": 1.9889807162534434, "grad_norm": 1.8339381217956543, "learning_rate": 6.154669115387647e-06, "loss": 0.0465, "step": 2166 }, { "epoch": 1.98989898989899, "grad_norm": 1.7856981754302979, "learning_rate": 6.1448105855687316e-06, "loss": 0.0575, "step": 2167 }, { "epoch": 1.9908172635445363, "grad_norm": 0.14215636253356934, "learning_rate": 6.134956454764136e-06, "loss": 0.0031, "step": 2168 }, { "epoch": 1.9917355371900827, "grad_norm": 0.6412303447723389, "learning_rate": 6.125106734218049e-06, "loss": 0.0173, "step": 2169 }, { "epoch": 1.992653810835629, "grad_norm": 3.37412428855896, "learning_rate": 6.11526143516962e-06, "loss": 0.6319, "step": 2170 }, { "epoch": 1.9935720844811753, "grad_norm": 2.645092248916626, "learning_rate": 6.1054205688529555e-06, "loss": 0.0654, "step": 2171 }, { "epoch": 1.994490358126722, "grad_norm": 2.39790940284729, "learning_rate": 6.095584146497106e-06, "loss": 0.0336, "step": 2172 }, { "epoch": 1.9954086317722681, "grad_norm": 0.8441511392593384, "learning_rate": 6.085752179326054e-06, "loss": 0.0249, "step": 2173 }, { "epoch": 1.9963269054178145, "grad_norm": 1.2189452648162842, "learning_rate": 6.075924678558689e-06, "loss": 0.0322, "step": 2174 }, { "epoch": 1.997245179063361, "grad_norm": 1.360795259475708, "learning_rate": 6.06610165540881e-06, "loss": 0.0417, "step": 2175 }, { "epoch": 1.9981634527089072, "grad_norm": 1.6724344491958618, "learning_rate": 6.056283121085109e-06, "loss": 0.0656, "step": 2176 }, { "epoch": 1.9990817263544536, "grad_norm": 0.1868937909603119, "learning_rate": 6.046469086791152e-06, "loss": 0.0049, "step": 2177 }, { "epoch": 2.0, "grad_norm": 0.2796348035335541, "learning_rate": 6.036659563725368e-06, "loss": 0.006, "step": 2178 }, { "epoch": 2.000918273645546, "grad_norm": 0.10434655100107193, "learning_rate": 6.026854563081046e-06, "loss": 0.0019, "step": 2179 }, { "epoch": 2.001836547291093, "grad_norm": 0.25287681818008423, "learning_rate": 6.017054096046306e-06, "loss": 0.004, "step": 2180 }, { "epoch": 2.002754820936639, "grad_norm": 0.3407348096370697, "learning_rate": 6.007258173804104e-06, "loss": 0.0084, "step": 2181 }, { "epoch": 2.0036730945821857, "grad_norm": 0.14312580227851868, "learning_rate": 5.997466807532203e-06, "loss": 0.0038, "step": 2182 }, { "epoch": 2.004591368227732, "grad_norm": 0.045787371695041656, "learning_rate": 5.98768000840317e-06, "loss": 0.001, "step": 2183 }, { "epoch": 2.005509641873278, "grad_norm": 1.2801481485366821, "learning_rate": 5.977897787584356e-06, "loss": 0.0314, "step": 2184 }, { "epoch": 2.0064279155188247, "grad_norm": 0.11737214028835297, "learning_rate": 5.9681201562378955e-06, "loss": 0.0024, "step": 2185 }, { "epoch": 2.007346189164371, "grad_norm": 1.588784098625183, "learning_rate": 5.958347125520686e-06, "loss": 0.0884, "step": 2186 }, { "epoch": 2.0082644628099175, "grad_norm": 1.0077193975448608, "learning_rate": 5.9485787065843695e-06, "loss": 0.0383, "step": 2187 }, { "epoch": 2.0091827364554637, "grad_norm": 1.5550599098205566, "learning_rate": 5.938814910575326e-06, "loss": 0.0648, "step": 2188 }, { "epoch": 2.01010101010101, "grad_norm": 0.5805178880691528, "learning_rate": 5.9290557486346645e-06, "loss": 0.0168, "step": 2189 }, { "epoch": 2.0110192837465566, "grad_norm": 1.142361044883728, "learning_rate": 5.9193012318982e-06, "loss": 0.0428, "step": 2190 }, { "epoch": 2.011937557392103, "grad_norm": 0.18860410153865814, "learning_rate": 5.9095513714964604e-06, "loss": 0.0038, "step": 2191 }, { "epoch": 2.0128558310376494, "grad_norm": 0.0928327888250351, "learning_rate": 5.899806178554646e-06, "loss": 0.0021, "step": 2192 }, { "epoch": 2.0137741046831956, "grad_norm": 0.2421255260705948, "learning_rate": 5.890065664192638e-06, "loss": 0.0067, "step": 2193 }, { "epoch": 2.014692378328742, "grad_norm": 0.41788625717163086, "learning_rate": 5.880329839524976e-06, "loss": 0.0084, "step": 2194 }, { "epoch": 2.0156106519742885, "grad_norm": 0.32163575291633606, "learning_rate": 5.870598715660849e-06, "loss": 0.0085, "step": 2195 }, { "epoch": 2.0165289256198347, "grad_norm": 0.22137615084648132, "learning_rate": 5.8608723037040894e-06, "loss": 0.0056, "step": 2196 }, { "epoch": 2.0174471992653813, "grad_norm": 0.20752303302288055, "learning_rate": 5.8511506147531425e-06, "loss": 0.0068, "step": 2197 }, { "epoch": 2.0183654729109275, "grad_norm": 1.3526684045791626, "learning_rate": 5.841433659901068e-06, "loss": 0.0971, "step": 2198 }, { "epoch": 2.0192837465564737, "grad_norm": 0.12442323565483093, "learning_rate": 5.831721450235526e-06, "loss": 0.0035, "step": 2199 }, { "epoch": 2.0202020202020203, "grad_norm": 1.750939965248108, "learning_rate": 5.822013996838759e-06, "loss": 0.0506, "step": 2200 }, { "epoch": 2.0211202938475665, "grad_norm": 0.28097644448280334, "learning_rate": 5.8123113107875865e-06, "loss": 0.0069, "step": 2201 }, { "epoch": 2.022038567493113, "grad_norm": 1.6677228212356567, "learning_rate": 5.802613403153379e-06, "loss": 0.0547, "step": 2202 }, { "epoch": 2.0229568411386594, "grad_norm": 0.3437405526638031, "learning_rate": 5.792920285002068e-06, "loss": 0.0089, "step": 2203 }, { "epoch": 2.0238751147842056, "grad_norm": 0.13200826942920685, "learning_rate": 5.783231967394112e-06, "loss": 0.0028, "step": 2204 }, { "epoch": 2.024793388429752, "grad_norm": 0.9323151111602783, "learning_rate": 5.7735484613844905e-06, "loss": 0.0731, "step": 2205 }, { "epoch": 2.0257116620752984, "grad_norm": 1.5939714908599854, "learning_rate": 5.763869778022694e-06, "loss": 0.1115, "step": 2206 }, { "epoch": 2.0266299357208446, "grad_norm": 0.056002844125032425, "learning_rate": 5.7541959283527125e-06, "loss": 0.0013, "step": 2207 }, { "epoch": 2.0275482093663912, "grad_norm": 0.3963686525821686, "learning_rate": 5.744526923413013e-06, "loss": 0.0071, "step": 2208 }, { "epoch": 2.0284664830119374, "grad_norm": 0.3142702877521515, "learning_rate": 5.734862774236549e-06, "loss": 0.0073, "step": 2209 }, { "epoch": 2.029384756657484, "grad_norm": 0.20771349966526031, "learning_rate": 5.725203491850722e-06, "loss": 0.0044, "step": 2210 }, { "epoch": 2.0303030303030303, "grad_norm": 1.170409917831421, "learning_rate": 5.715549087277379e-06, "loss": 0.0264, "step": 2211 }, { "epoch": 2.0312213039485765, "grad_norm": 0.3745550513267517, "learning_rate": 5.705899571532808e-06, "loss": 0.0088, "step": 2212 }, { "epoch": 2.032139577594123, "grad_norm": 1.2890238761901855, "learning_rate": 5.6962549556277134e-06, "loss": 0.0548, "step": 2213 }, { "epoch": 2.0330578512396693, "grad_norm": 0.7098655700683594, "learning_rate": 5.686615250567207e-06, "loss": 0.0133, "step": 2214 }, { "epoch": 2.033976124885216, "grad_norm": 0.5373619794845581, "learning_rate": 5.676980467350808e-06, "loss": 0.0072, "step": 2215 }, { "epoch": 2.034894398530762, "grad_norm": 1.0057272911071777, "learning_rate": 5.667350616972407e-06, "loss": 0.023, "step": 2216 }, { "epoch": 2.0358126721763083, "grad_norm": 0.47574615478515625, "learning_rate": 5.6577257104202735e-06, "loss": 0.0092, "step": 2217 }, { "epoch": 2.036730945821855, "grad_norm": 0.8104116916656494, "learning_rate": 5.648105758677029e-06, "loss": 0.0296, "step": 2218 }, { "epoch": 2.037649219467401, "grad_norm": 0.054813072085380554, "learning_rate": 5.638490772719646e-06, "loss": 0.0011, "step": 2219 }, { "epoch": 2.038567493112948, "grad_norm": 1.1233893632888794, "learning_rate": 5.6288807635194355e-06, "loss": 0.0558, "step": 2220 }, { "epoch": 2.039485766758494, "grad_norm": 0.699813961982727, "learning_rate": 5.619275742042019e-06, "loss": 0.013, "step": 2221 }, { "epoch": 2.04040404040404, "grad_norm": 0.08380339294672012, "learning_rate": 5.609675719247334e-06, "loss": 0.0022, "step": 2222 }, { "epoch": 2.041322314049587, "grad_norm": 0.06374821066856384, "learning_rate": 5.60008070608961e-06, "loss": 0.0014, "step": 2223 }, { "epoch": 2.042240587695133, "grad_norm": 1.1818922758102417, "learning_rate": 5.590490713517362e-06, "loss": 0.0645, "step": 2224 }, { "epoch": 2.0431588613406797, "grad_norm": 3.380676746368408, "learning_rate": 5.58090575247338e-06, "loss": 0.2498, "step": 2225 }, { "epoch": 2.044077134986226, "grad_norm": 3.6140034198760986, "learning_rate": 5.571325833894709e-06, "loss": 0.5901, "step": 2226 }, { "epoch": 2.044995408631772, "grad_norm": 0.09461050480604172, "learning_rate": 5.561750968712638e-06, "loss": 0.0018, "step": 2227 }, { "epoch": 2.0459136822773187, "grad_norm": 0.4000767767429352, "learning_rate": 5.552181167852695e-06, "loss": 0.0083, "step": 2228 }, { "epoch": 2.046831955922865, "grad_norm": 0.24144043028354645, "learning_rate": 5.542616442234618e-06, "loss": 0.0082, "step": 2229 }, { "epoch": 2.0477502295684116, "grad_norm": 0.6362507939338684, "learning_rate": 5.533056802772374e-06, "loss": 0.0124, "step": 2230 }, { "epoch": 2.0486685032139578, "grad_norm": 1.00215744972229, "learning_rate": 5.52350226037411e-06, "loss": 0.0183, "step": 2231 }, { "epoch": 2.049586776859504, "grad_norm": 0.6746169924736023, "learning_rate": 5.513952825942159e-06, "loss": 0.0264, "step": 2232 }, { "epoch": 2.0505050505050506, "grad_norm": 0.08102639764547348, "learning_rate": 5.504408510373029e-06, "loss": 0.0013, "step": 2233 }, { "epoch": 2.051423324150597, "grad_norm": 0.4489283859729767, "learning_rate": 5.494869324557391e-06, "loss": 0.0087, "step": 2234 }, { "epoch": 2.0523415977961434, "grad_norm": 1.164612889289856, "learning_rate": 5.485335279380052e-06, "loss": 0.0452, "step": 2235 }, { "epoch": 2.0532598714416896, "grad_norm": 0.7597163319587708, "learning_rate": 5.475806385719962e-06, "loss": 0.0142, "step": 2236 }, { "epoch": 2.054178145087236, "grad_norm": 0.2588871419429779, "learning_rate": 5.46628265445019e-06, "loss": 0.0059, "step": 2237 }, { "epoch": 2.0550964187327825, "grad_norm": 1.6339621543884277, "learning_rate": 5.456764096437908e-06, "loss": 0.0998, "step": 2238 }, { "epoch": 2.0560146923783287, "grad_norm": 0.3298821449279785, "learning_rate": 5.447250722544402e-06, "loss": 0.0084, "step": 2239 }, { "epoch": 2.0569329660238753, "grad_norm": 0.06803268194198608, "learning_rate": 5.437742543625024e-06, "loss": 0.0014, "step": 2240 }, { "epoch": 2.0578512396694215, "grad_norm": 1.772585153579712, "learning_rate": 5.428239570529208e-06, "loss": 0.0589, "step": 2241 }, { "epoch": 2.0587695133149677, "grad_norm": 0.38087910413742065, "learning_rate": 5.418741814100447e-06, "loss": 0.0092, "step": 2242 }, { "epoch": 2.0596877869605144, "grad_norm": 0.5190092325210571, "learning_rate": 5.409249285176275e-06, "loss": 0.013, "step": 2243 }, { "epoch": 2.0606060606060606, "grad_norm": 0.34029868245124817, "learning_rate": 5.399761994588273e-06, "loss": 0.0094, "step": 2244 }, { "epoch": 2.061524334251607, "grad_norm": 0.04259571433067322, "learning_rate": 5.390279953162037e-06, "loss": 0.0009, "step": 2245 }, { "epoch": 2.0624426078971534, "grad_norm": 1.0722955465316772, "learning_rate": 5.380803171717172e-06, "loss": 0.0255, "step": 2246 }, { "epoch": 2.0633608815426996, "grad_norm": 1.6727955341339111, "learning_rate": 5.371331661067284e-06, "loss": 0.0702, "step": 2247 }, { "epoch": 2.0642791551882462, "grad_norm": 0.8358317613601685, "learning_rate": 5.361865432019961e-06, "loss": 0.0232, "step": 2248 }, { "epoch": 2.0651974288337924, "grad_norm": 2.2334063053131104, "learning_rate": 5.352404495376775e-06, "loss": 0.0659, "step": 2249 }, { "epoch": 2.0661157024793386, "grad_norm": 0.7159305214881897, "learning_rate": 5.342948861933246e-06, "loss": 0.0162, "step": 2250 }, { "epoch": 2.0670339761248853, "grad_norm": 1.612284779548645, "learning_rate": 5.333498542478852e-06, "loss": 0.1575, "step": 2251 }, { "epoch": 2.0679522497704315, "grad_norm": 1.4297980070114136, "learning_rate": 5.324053547796999e-06, "loss": 0.0289, "step": 2252 }, { "epoch": 2.068870523415978, "grad_norm": 0.4325422942638397, "learning_rate": 5.314613888665022e-06, "loss": 0.0103, "step": 2253 }, { "epoch": 2.0697887970615243, "grad_norm": 0.06030826270580292, "learning_rate": 5.305179575854173e-06, "loss": 0.001, "step": 2254 }, { "epoch": 2.0707070707070705, "grad_norm": 1.381588101387024, "learning_rate": 5.2957506201295946e-06, "loss": 0.0579, "step": 2255 }, { "epoch": 2.071625344352617, "grad_norm": 0.9456705451011658, "learning_rate": 5.2863270322503214e-06, "loss": 0.0343, "step": 2256 }, { "epoch": 2.0725436179981633, "grad_norm": 0.24636121094226837, "learning_rate": 5.276908822969256e-06, "loss": 0.0051, "step": 2257 }, { "epoch": 2.07346189164371, "grad_norm": 1.1309165954589844, "learning_rate": 5.2674960030331766e-06, "loss": 0.0751, "step": 2258 }, { "epoch": 2.074380165289256, "grad_norm": 0.5227562189102173, "learning_rate": 5.258088583182704e-06, "loss": 0.0061, "step": 2259 }, { "epoch": 2.0752984389348024, "grad_norm": 0.8285462260246277, "learning_rate": 5.248686574152293e-06, "loss": 0.0211, "step": 2260 }, { "epoch": 2.076216712580349, "grad_norm": 0.12175718694925308, "learning_rate": 5.239289986670233e-06, "loss": 0.0031, "step": 2261 }, { "epoch": 2.077134986225895, "grad_norm": 1.0064828395843506, "learning_rate": 5.229898831458618e-06, "loss": 0.028, "step": 2262 }, { "epoch": 2.078053259871442, "grad_norm": 1.6101104021072388, "learning_rate": 5.220513119233354e-06, "loss": 0.0491, "step": 2263 }, { "epoch": 2.078971533516988, "grad_norm": 0.10207781940698624, "learning_rate": 5.211132860704131e-06, "loss": 0.0025, "step": 2264 }, { "epoch": 2.0798898071625342, "grad_norm": 0.40594008564949036, "learning_rate": 5.201758066574414e-06, "loss": 0.0067, "step": 2265 }, { "epoch": 2.080808080808081, "grad_norm": 0.997552752494812, "learning_rate": 5.192388747541432e-06, "loss": 0.0495, "step": 2266 }, { "epoch": 2.081726354453627, "grad_norm": 1.0655392408370972, "learning_rate": 5.183024914296169e-06, "loss": 0.0746, "step": 2267 }, { "epoch": 2.0826446280991737, "grad_norm": 1.0579078197479248, "learning_rate": 5.173666577523354e-06, "loss": 0.0204, "step": 2268 }, { "epoch": 2.08356290174472, "grad_norm": 1.395107626914978, "learning_rate": 5.1643137479014386e-06, "loss": 0.0182, "step": 2269 }, { "epoch": 2.084481175390266, "grad_norm": 1.17180597782135, "learning_rate": 5.154966436102589e-06, "loss": 0.0394, "step": 2270 }, { "epoch": 2.0853994490358128, "grad_norm": 0.2431759238243103, "learning_rate": 5.145624652792679e-06, "loss": 0.0049, "step": 2271 }, { "epoch": 2.086317722681359, "grad_norm": 0.178292378783226, "learning_rate": 5.136288408631268e-06, "loss": 0.0041, "step": 2272 }, { "epoch": 2.0872359963269056, "grad_norm": 0.22816675901412964, "learning_rate": 5.126957714271608e-06, "loss": 0.0042, "step": 2273 }, { "epoch": 2.088154269972452, "grad_norm": 0.28758785128593445, "learning_rate": 5.1176325803606064e-06, "loss": 0.005, "step": 2274 }, { "epoch": 2.089072543617998, "grad_norm": 1.0377463102340698, "learning_rate": 5.108313017538828e-06, "loss": 0.0377, "step": 2275 }, { "epoch": 2.0899908172635446, "grad_norm": 2.2243616580963135, "learning_rate": 5.098999036440482e-06, "loss": 0.0342, "step": 2276 }, { "epoch": 2.090909090909091, "grad_norm": 1.2942243814468384, "learning_rate": 5.089690647693406e-06, "loss": 0.0214, "step": 2277 }, { "epoch": 2.0918273645546375, "grad_norm": 0.14863908290863037, "learning_rate": 5.080387861919065e-06, "loss": 0.0031, "step": 2278 }, { "epoch": 2.0927456382001837, "grad_norm": 0.054995786398649216, "learning_rate": 5.071090689732521e-06, "loss": 0.0015, "step": 2279 }, { "epoch": 2.09366391184573, "grad_norm": 0.7771134972572327, "learning_rate": 5.061799141742434e-06, "loss": 0.0079, "step": 2280 }, { "epoch": 2.0945821854912765, "grad_norm": 0.12749284505844116, "learning_rate": 5.052513228551048e-06, "loss": 0.0024, "step": 2281 }, { "epoch": 2.0955004591368227, "grad_norm": 1.4899041652679443, "learning_rate": 5.043232960754169e-06, "loss": 0.0712, "step": 2282 }, { "epoch": 2.0964187327823693, "grad_norm": 1.2852073907852173, "learning_rate": 5.033958348941177e-06, "loss": 0.0268, "step": 2283 }, { "epoch": 2.0973370064279155, "grad_norm": 1.2812281847000122, "learning_rate": 5.024689403694985e-06, "loss": 0.0565, "step": 2284 }, { "epoch": 2.0982552800734617, "grad_norm": 0.37708449363708496, "learning_rate": 5.0154261355920455e-06, "loss": 0.0074, "step": 2285 }, { "epoch": 2.0991735537190084, "grad_norm": 0.10309432446956635, "learning_rate": 5.0061685552023265e-06, "loss": 0.0018, "step": 2286 }, { "epoch": 2.1000918273645546, "grad_norm": 0.18390938639640808, "learning_rate": 4.996916673089319e-06, "loss": 0.005, "step": 2287 }, { "epoch": 2.101010101010101, "grad_norm": 2.858041286468506, "learning_rate": 4.987670499810001e-06, "loss": 0.0822, "step": 2288 }, { "epoch": 2.1019283746556474, "grad_norm": 1.0878427028656006, "learning_rate": 4.97843004591484e-06, "loss": 0.0238, "step": 2289 }, { "epoch": 2.1028466483011936, "grad_norm": 0.12825508415699005, "learning_rate": 4.969195321947774e-06, "loss": 0.0023, "step": 2290 }, { "epoch": 2.1037649219467403, "grad_norm": 0.45835599303245544, "learning_rate": 4.959966338446206e-06, "loss": 0.0149, "step": 2291 }, { "epoch": 2.1046831955922864, "grad_norm": 2.825117826461792, "learning_rate": 4.9507431059409925e-06, "loss": 0.0858, "step": 2292 }, { "epoch": 2.1056014692378326, "grad_norm": 0.9273111820220947, "learning_rate": 4.941525634956421e-06, "loss": 0.0201, "step": 2293 }, { "epoch": 2.1065197428833793, "grad_norm": 1.9458352327346802, "learning_rate": 4.932313936010209e-06, "loss": 0.1512, "step": 2294 }, { "epoch": 2.1074380165289255, "grad_norm": 1.230194091796875, "learning_rate": 4.923108019613483e-06, "loss": 0.0783, "step": 2295 }, { "epoch": 2.108356290174472, "grad_norm": 0.6702539920806885, "learning_rate": 4.9139078962707765e-06, "loss": 0.0195, "step": 2296 }, { "epoch": 2.1092745638200183, "grad_norm": 0.3930825889110565, "learning_rate": 4.904713576480007e-06, "loss": 0.0077, "step": 2297 }, { "epoch": 2.110192837465565, "grad_norm": 0.9272628426551819, "learning_rate": 4.89552507073248e-06, "loss": 0.0217, "step": 2298 }, { "epoch": 2.111111111111111, "grad_norm": 0.25333306193351746, "learning_rate": 4.886342389512856e-06, "loss": 0.005, "step": 2299 }, { "epoch": 2.1120293847566574, "grad_norm": 1.0743340253829956, "learning_rate": 4.877165543299154e-06, "loss": 0.0807, "step": 2300 }, { "epoch": 2.112947658402204, "grad_norm": 1.2217533588409424, "learning_rate": 4.867994542562735e-06, "loss": 0.0411, "step": 2301 }, { "epoch": 2.11386593204775, "grad_norm": 3.4082953929901123, "learning_rate": 4.85882939776829e-06, "loss": 0.1483, "step": 2302 }, { "epoch": 2.1147842056932964, "grad_norm": 0.2680063545703888, "learning_rate": 4.849670119373822e-06, "loss": 0.0077, "step": 2303 }, { "epoch": 2.115702479338843, "grad_norm": 0.6700600981712341, "learning_rate": 4.840516717830652e-06, "loss": 0.007, "step": 2304 }, { "epoch": 2.1166207529843892, "grad_norm": 0.15074005722999573, "learning_rate": 4.831369203583388e-06, "loss": 0.0026, "step": 2305 }, { "epoch": 2.117539026629936, "grad_norm": 0.10300244390964508, "learning_rate": 4.82222758706992e-06, "loss": 0.0018, "step": 2306 }, { "epoch": 2.118457300275482, "grad_norm": 1.830506443977356, "learning_rate": 4.813091878721408e-06, "loss": 0.0808, "step": 2307 }, { "epoch": 2.1193755739210283, "grad_norm": 0.6274886131286621, "learning_rate": 4.803962088962271e-06, "loss": 0.0121, "step": 2308 }, { "epoch": 2.120293847566575, "grad_norm": 1.4498733282089233, "learning_rate": 4.794838228210172e-06, "loss": 0.0727, "step": 2309 }, { "epoch": 2.121212121212121, "grad_norm": 0.061251141130924225, "learning_rate": 4.7857203068760205e-06, "loss": 0.0011, "step": 2310 }, { "epoch": 2.1221303948576677, "grad_norm": 0.3291700482368469, "learning_rate": 4.776608335363935e-06, "loss": 0.0052, "step": 2311 }, { "epoch": 2.123048668503214, "grad_norm": 1.709920883178711, "learning_rate": 4.767502324071249e-06, "loss": 0.0203, "step": 2312 }, { "epoch": 2.12396694214876, "grad_norm": 1.2096740007400513, "learning_rate": 4.7584022833884976e-06, "loss": 0.0758, "step": 2313 }, { "epoch": 2.124885215794307, "grad_norm": 0.3426104784011841, "learning_rate": 4.7493082236994005e-06, "loss": 0.005, "step": 2314 }, { "epoch": 2.125803489439853, "grad_norm": 1.0054347515106201, "learning_rate": 4.74022015538085e-06, "loss": 0.0253, "step": 2315 }, { "epoch": 2.1267217630853996, "grad_norm": 0.1656254678964615, "learning_rate": 4.731138088802911e-06, "loss": 0.0037, "step": 2316 }, { "epoch": 2.127640036730946, "grad_norm": 1.4126497507095337, "learning_rate": 4.722062034328794e-06, "loss": 0.1063, "step": 2317 }, { "epoch": 2.128558310376492, "grad_norm": 0.2782822549343109, "learning_rate": 4.712992002314846e-06, "loss": 0.0037, "step": 2318 }, { "epoch": 2.1294765840220387, "grad_norm": 0.09124380350112915, "learning_rate": 4.70392800311055e-06, "loss": 0.002, "step": 2319 }, { "epoch": 2.130394857667585, "grad_norm": 0.10553082823753357, "learning_rate": 4.694870047058493e-06, "loss": 0.0027, "step": 2320 }, { "epoch": 2.1313131313131315, "grad_norm": 2.0747780799865723, "learning_rate": 4.685818144494384e-06, "loss": 0.2181, "step": 2321 }, { "epoch": 2.1322314049586777, "grad_norm": 0.051261018961668015, "learning_rate": 4.676772305747011e-06, "loss": 0.001, "step": 2322 }, { "epoch": 2.133149678604224, "grad_norm": 0.04028478264808655, "learning_rate": 4.6677325411382455e-06, "loss": 0.001, "step": 2323 }, { "epoch": 2.1340679522497705, "grad_norm": 0.2693515717983246, "learning_rate": 4.65869886098303e-06, "loss": 0.0038, "step": 2324 }, { "epoch": 2.1349862258953167, "grad_norm": 1.8984618186950684, "learning_rate": 4.649671275589359e-06, "loss": 0.1062, "step": 2325 }, { "epoch": 2.1359044995408634, "grad_norm": 2.0793392658233643, "learning_rate": 4.640649795258286e-06, "loss": 0.0583, "step": 2326 }, { "epoch": 2.1368227731864096, "grad_norm": 1.5104140043258667, "learning_rate": 4.631634430283884e-06, "loss": 0.0483, "step": 2327 }, { "epoch": 2.1377410468319558, "grad_norm": 1.1665524244308472, "learning_rate": 4.622625190953255e-06, "loss": 0.0217, "step": 2328 }, { "epoch": 2.1386593204775024, "grad_norm": 0.7507249116897583, "learning_rate": 4.613622087546507e-06, "loss": 0.02, "step": 2329 }, { "epoch": 2.1395775941230486, "grad_norm": 0.7719230651855469, "learning_rate": 4.604625130336746e-06, "loss": 0.0165, "step": 2330 }, { "epoch": 2.1404958677685952, "grad_norm": 0.07237987965345383, "learning_rate": 4.595634329590075e-06, "loss": 0.0014, "step": 2331 }, { "epoch": 2.1414141414141414, "grad_norm": 0.340585857629776, "learning_rate": 4.586649695565563e-06, "loss": 0.0084, "step": 2332 }, { "epoch": 2.1423324150596876, "grad_norm": 1.9663859605789185, "learning_rate": 4.577671238515241e-06, "loss": 0.1142, "step": 2333 }, { "epoch": 2.1432506887052343, "grad_norm": 0.21870839595794678, "learning_rate": 4.568698968684097e-06, "loss": 0.0053, "step": 2334 }, { "epoch": 2.1441689623507805, "grad_norm": 2.0747690200805664, "learning_rate": 4.559732896310054e-06, "loss": 0.088, "step": 2335 }, { "epoch": 2.145087235996327, "grad_norm": 1.1275554895401, "learning_rate": 4.550773031623972e-06, "loss": 0.0342, "step": 2336 }, { "epoch": 2.1460055096418733, "grad_norm": 0.8312899470329285, "learning_rate": 4.541819384849618e-06, "loss": 0.0124, "step": 2337 }, { "epoch": 2.1469237832874195, "grad_norm": 1.9338234663009644, "learning_rate": 4.5328719662036695e-06, "loss": 0.0886, "step": 2338 }, { "epoch": 2.147842056932966, "grad_norm": 0.8676115274429321, "learning_rate": 4.523930785895689e-06, "loss": 0.0107, "step": 2339 }, { "epoch": 2.1487603305785123, "grad_norm": 0.24489907920360565, "learning_rate": 4.514995854128137e-06, "loss": 0.0045, "step": 2340 }, { "epoch": 2.149678604224059, "grad_norm": 1.9321372509002686, "learning_rate": 4.506067181096331e-06, "loss": 0.0489, "step": 2341 }, { "epoch": 2.150596877869605, "grad_norm": 0.8259553909301758, "learning_rate": 4.497144776988448e-06, "loss": 0.0226, "step": 2342 }, { "epoch": 2.1515151515151514, "grad_norm": 1.1399952173233032, "learning_rate": 4.4882286519855156e-06, "loss": 0.0648, "step": 2343 }, { "epoch": 2.152433425160698, "grad_norm": 2.794127941131592, "learning_rate": 4.47931881626139e-06, "loss": 0.1805, "step": 2344 }, { "epoch": 2.153351698806244, "grad_norm": 0.13637632131576538, "learning_rate": 4.4704152799827645e-06, "loss": 0.0025, "step": 2345 }, { "epoch": 2.1542699724517904, "grad_norm": 2.3013103008270264, "learning_rate": 4.461518053309132e-06, "loss": 0.1261, "step": 2346 }, { "epoch": 2.155188246097337, "grad_norm": 0.3690642714500427, "learning_rate": 4.452627146392788e-06, "loss": 0.0087, "step": 2347 }, { "epoch": 2.1561065197428833, "grad_norm": 0.043110717087984085, "learning_rate": 4.4437425693788195e-06, "loss": 0.0009, "step": 2348 }, { "epoch": 2.15702479338843, "grad_norm": 2.5191810131073, "learning_rate": 4.434864332405085e-06, "loss": 0.0812, "step": 2349 }, { "epoch": 2.157943067033976, "grad_norm": 3.7510275840759277, "learning_rate": 4.425992445602221e-06, "loss": 0.105, "step": 2350 }, { "epoch": 2.1588613406795227, "grad_norm": 0.42147496342658997, "learning_rate": 4.417126919093605e-06, "loss": 0.0062, "step": 2351 }, { "epoch": 2.159779614325069, "grad_norm": 0.28600630164146423, "learning_rate": 4.408267762995363e-06, "loss": 0.0055, "step": 2352 }, { "epoch": 2.160697887970615, "grad_norm": 1.6876825094223022, "learning_rate": 4.3994149874163514e-06, "loss": 0.0667, "step": 2353 }, { "epoch": 2.1616161616161618, "grad_norm": 2.8767223358154297, "learning_rate": 4.390568602458141e-06, "loss": 0.2958, "step": 2354 }, { "epoch": 2.162534435261708, "grad_norm": 0.6954145431518555, "learning_rate": 4.381728618215022e-06, "loss": 0.0101, "step": 2355 }, { "epoch": 2.163452708907254, "grad_norm": 1.0215270519256592, "learning_rate": 4.3728950447739724e-06, "loss": 0.0227, "step": 2356 }, { "epoch": 2.164370982552801, "grad_norm": 1.7390638589859009, "learning_rate": 4.364067892214655e-06, "loss": 0.1635, "step": 2357 }, { "epoch": 2.165289256198347, "grad_norm": 2.126894950866699, "learning_rate": 4.355247170609408e-06, "loss": 0.1532, "step": 2358 }, { "epoch": 2.1662075298438936, "grad_norm": 3.1040496826171875, "learning_rate": 4.3464328900232274e-06, "loss": 0.2354, "step": 2359 }, { "epoch": 2.16712580348944, "grad_norm": 0.19557468593120575, "learning_rate": 4.3376250605137725e-06, "loss": 0.005, "step": 2360 }, { "epoch": 2.168044077134986, "grad_norm": 2.7327542304992676, "learning_rate": 4.3288236921313265e-06, "loss": 0.0647, "step": 2361 }, { "epoch": 2.1689623507805327, "grad_norm": 0.6060188412666321, "learning_rate": 4.3200287949188065e-06, "loss": 0.0115, "step": 2362 }, { "epoch": 2.169880624426079, "grad_norm": 0.030372226610779762, "learning_rate": 4.3112403789117455e-06, "loss": 0.0007, "step": 2363 }, { "epoch": 2.1707988980716255, "grad_norm": 0.883770227432251, "learning_rate": 4.3024584541382774e-06, "loss": 0.0196, "step": 2364 }, { "epoch": 2.1717171717171717, "grad_norm": 0.8529956936836243, "learning_rate": 4.293683030619138e-06, "loss": 0.0289, "step": 2365 }, { "epoch": 2.172635445362718, "grad_norm": 0.09572714567184448, "learning_rate": 4.284914118367637e-06, "loss": 0.0023, "step": 2366 }, { "epoch": 2.1735537190082646, "grad_norm": 0.15946859121322632, "learning_rate": 4.276151727389656e-06, "loss": 0.0036, "step": 2367 }, { "epoch": 2.1744719926538107, "grad_norm": 0.15879037976264954, "learning_rate": 4.267395867683631e-06, "loss": 0.0035, "step": 2368 }, { "epoch": 2.1753902662993574, "grad_norm": 2.307878017425537, "learning_rate": 4.25864654924056e-06, "loss": 0.079, "step": 2369 }, { "epoch": 2.1763085399449036, "grad_norm": 0.04854109138250351, "learning_rate": 4.249903782043963e-06, "loss": 0.0008, "step": 2370 }, { "epoch": 2.17722681359045, "grad_norm": 0.5023438930511475, "learning_rate": 4.241167576069886e-06, "loss": 0.0062, "step": 2371 }, { "epoch": 2.1781450872359964, "grad_norm": 1.14212167263031, "learning_rate": 4.232437941286896e-06, "loss": 0.0414, "step": 2372 }, { "epoch": 2.1790633608815426, "grad_norm": 0.11813822388648987, "learning_rate": 4.223714887656049e-06, "loss": 0.0018, "step": 2373 }, { "epoch": 2.1799816345270893, "grad_norm": 0.04896951839327812, "learning_rate": 4.21499842513091e-06, "loss": 0.0012, "step": 2374 }, { "epoch": 2.1808999081726355, "grad_norm": 0.06720975041389465, "learning_rate": 4.206288563657506e-06, "loss": 0.0017, "step": 2375 }, { "epoch": 2.1818181818181817, "grad_norm": 1.7646563053131104, "learning_rate": 4.1975853131743395e-06, "loss": 0.0389, "step": 2376 }, { "epoch": 2.1827364554637283, "grad_norm": 1.5686593055725098, "learning_rate": 4.18888868361237e-06, "loss": 0.0553, "step": 2377 }, { "epoch": 2.1836547291092745, "grad_norm": 1.5425838232040405, "learning_rate": 4.180198684894995e-06, "loss": 0.0237, "step": 2378 }, { "epoch": 2.184573002754821, "grad_norm": 0.3182779848575592, "learning_rate": 4.171515326938058e-06, "loss": 0.0093, "step": 2379 }, { "epoch": 2.1854912764003673, "grad_norm": 1.2447888851165771, "learning_rate": 4.162838619649817e-06, "loss": 0.0285, "step": 2380 }, { "epoch": 2.1864095500459135, "grad_norm": 0.04312663525342941, "learning_rate": 4.154168572930941e-06, "loss": 0.001, "step": 2381 }, { "epoch": 2.18732782369146, "grad_norm": 1.5905214548110962, "learning_rate": 4.145505196674499e-06, "loss": 0.0796, "step": 2382 }, { "epoch": 2.1882460973370064, "grad_norm": 0.08280617743730545, "learning_rate": 4.136848500765948e-06, "loss": 0.0023, "step": 2383 }, { "epoch": 2.189164370982553, "grad_norm": 1.8333122730255127, "learning_rate": 4.128198495083131e-06, "loss": 0.0811, "step": 2384 }, { "epoch": 2.190082644628099, "grad_norm": 0.8395518064498901, "learning_rate": 4.119555189496246e-06, "loss": 0.0137, "step": 2385 }, { "epoch": 2.1910009182736454, "grad_norm": 1.873437762260437, "learning_rate": 4.110918593867851e-06, "loss": 0.1597, "step": 2386 }, { "epoch": 2.191919191919192, "grad_norm": 1.4029896259307861, "learning_rate": 4.102288718052845e-06, "loss": 0.0455, "step": 2387 }, { "epoch": 2.1928374655647382, "grad_norm": 0.531721293926239, "learning_rate": 4.093665571898457e-06, "loss": 0.0074, "step": 2388 }, { "epoch": 2.1937557392102844, "grad_norm": 0.22681565582752228, "learning_rate": 4.085049165244251e-06, "loss": 0.0067, "step": 2389 }, { "epoch": 2.194674012855831, "grad_norm": 1.4374407529830933, "learning_rate": 4.076439507922082e-06, "loss": 0.0782, "step": 2390 }, { "epoch": 2.1955922865013773, "grad_norm": 1.6869022846221924, "learning_rate": 4.067836609756116e-06, "loss": 0.0423, "step": 2391 }, { "epoch": 2.196510560146924, "grad_norm": 3.29586124420166, "learning_rate": 4.059240480562797e-06, "loss": 0.3392, "step": 2392 }, { "epoch": 2.19742883379247, "grad_norm": 1.4052507877349854, "learning_rate": 4.05065113015086e-06, "loss": 0.0441, "step": 2393 }, { "epoch": 2.1983471074380168, "grad_norm": 0.02917354181408882, "learning_rate": 4.04206856832129e-06, "loss": 0.0008, "step": 2394 }, { "epoch": 2.199265381083563, "grad_norm": 1.0327624082565308, "learning_rate": 4.033492804867331e-06, "loss": 0.0273, "step": 2395 }, { "epoch": 2.200183654729109, "grad_norm": 1.3676360845565796, "learning_rate": 4.024923849574474e-06, "loss": 0.108, "step": 2396 }, { "epoch": 2.201101928374656, "grad_norm": 2.5201096534729004, "learning_rate": 4.016361712220432e-06, "loss": 0.0987, "step": 2397 }, { "epoch": 2.202020202020202, "grad_norm": 0.41416481137275696, "learning_rate": 4.007806402575145e-06, "loss": 0.0072, "step": 2398 }, { "epoch": 2.202938475665748, "grad_norm": 0.10069648176431656, "learning_rate": 3.999257930400765e-06, "loss": 0.0021, "step": 2399 }, { "epoch": 2.203856749311295, "grad_norm": 0.6565502882003784, "learning_rate": 3.990716305451636e-06, "loss": 0.0103, "step": 2400 }, { "epoch": 2.204775022956841, "grad_norm": 0.05133352801203728, "learning_rate": 3.982181537474291e-06, "loss": 0.001, "step": 2401 }, { "epoch": 2.2056932966023877, "grad_norm": 1.3166080713272095, "learning_rate": 3.973653636207437e-06, "loss": 0.0749, "step": 2402 }, { "epoch": 2.206611570247934, "grad_norm": 0.09711894392967224, "learning_rate": 3.96513261138195e-06, "loss": 0.0012, "step": 2403 }, { "epoch": 2.20752984389348, "grad_norm": 2.089554786682129, "learning_rate": 3.956618472720851e-06, "loss": 0.0943, "step": 2404 }, { "epoch": 2.2084481175390267, "grad_norm": 0.7823900580406189, "learning_rate": 3.9481112299393175e-06, "loss": 0.0218, "step": 2405 }, { "epoch": 2.209366391184573, "grad_norm": 1.3393568992614746, "learning_rate": 3.939610892744647e-06, "loss": 0.0651, "step": 2406 }, { "epoch": 2.2102846648301195, "grad_norm": 0.8122115731239319, "learning_rate": 3.931117470836259e-06, "loss": 0.0188, "step": 2407 }, { "epoch": 2.2112029384756657, "grad_norm": 2.2816827297210693, "learning_rate": 3.922630973905685e-06, "loss": 0.1429, "step": 2408 }, { "epoch": 2.212121212121212, "grad_norm": 1.4124748706817627, "learning_rate": 3.9141514116365535e-06, "loss": 0.0382, "step": 2409 }, { "epoch": 2.2130394857667586, "grad_norm": 0.9464103579521179, "learning_rate": 3.905678793704575e-06, "loss": 0.0268, "step": 2410 }, { "epoch": 2.2139577594123048, "grad_norm": 1.1602070331573486, "learning_rate": 3.8972131297775485e-06, "loss": 0.1271, "step": 2411 }, { "epoch": 2.2148760330578514, "grad_norm": 0.3507850170135498, "learning_rate": 3.888754429515328e-06, "loss": 0.0071, "step": 2412 }, { "epoch": 2.2157943067033976, "grad_norm": 0.3688160479068756, "learning_rate": 3.880302702569823e-06, "loss": 0.0043, "step": 2413 }, { "epoch": 2.216712580348944, "grad_norm": 1.6793270111083984, "learning_rate": 3.87185795858499e-06, "loss": 0.0677, "step": 2414 }, { "epoch": 2.2176308539944904, "grad_norm": 0.31943562626838684, "learning_rate": 3.863420207196811e-06, "loss": 0.0066, "step": 2415 }, { "epoch": 2.2185491276400366, "grad_norm": 0.1738429069519043, "learning_rate": 3.8549894580332935e-06, "loss": 0.0027, "step": 2416 }, { "epoch": 2.2194674012855833, "grad_norm": 0.37680086493492126, "learning_rate": 3.846565720714451e-06, "loss": 0.0111, "step": 2417 }, { "epoch": 2.2203856749311295, "grad_norm": 2.7153451442718506, "learning_rate": 3.8381490048523075e-06, "loss": 0.1179, "step": 2418 }, { "epoch": 2.2213039485766757, "grad_norm": 0.08058220148086548, "learning_rate": 3.829739320050861e-06, "loss": 0.0019, "step": 2419 }, { "epoch": 2.2222222222222223, "grad_norm": 0.15834324061870575, "learning_rate": 3.821336675906094e-06, "loss": 0.0036, "step": 2420 }, { "epoch": 2.2231404958677685, "grad_norm": 0.864108145236969, "learning_rate": 3.8129410820059497e-06, "loss": 0.0185, "step": 2421 }, { "epoch": 2.224058769513315, "grad_norm": 0.5159684419631958, "learning_rate": 3.804552547930337e-06, "loss": 0.0087, "step": 2422 }, { "epoch": 2.2249770431588614, "grad_norm": 1.289405345916748, "learning_rate": 3.796171083251099e-06, "loss": 0.0369, "step": 2423 }, { "epoch": 2.2258953168044076, "grad_norm": 0.2897564768791199, "learning_rate": 3.7877966975320156e-06, "loss": 0.0053, "step": 2424 }, { "epoch": 2.226813590449954, "grad_norm": 0.23507395386695862, "learning_rate": 3.779429400328789e-06, "loss": 0.0043, "step": 2425 }, { "epoch": 2.2277318640955004, "grad_norm": 0.6375302076339722, "learning_rate": 3.7710692011890304e-06, "loss": 0.0671, "step": 2426 }, { "epoch": 2.228650137741047, "grad_norm": 1.1437077522277832, "learning_rate": 3.7627161096522614e-06, "loss": 0.0281, "step": 2427 }, { "epoch": 2.2295684113865932, "grad_norm": 0.5428879857063293, "learning_rate": 3.7543701352498817e-06, "loss": 0.0115, "step": 2428 }, { "epoch": 2.2304866850321394, "grad_norm": 0.2236398160457611, "learning_rate": 3.7460312875051763e-06, "loss": 0.0041, "step": 2429 }, { "epoch": 2.231404958677686, "grad_norm": 1.9673408269882202, "learning_rate": 3.7376995759332944e-06, "loss": 0.1308, "step": 2430 }, { "epoch": 2.2323232323232323, "grad_norm": 0.7397902011871338, "learning_rate": 3.729375010041243e-06, "loss": 0.0145, "step": 2431 }, { "epoch": 2.2332415059687785, "grad_norm": 0.4299418032169342, "learning_rate": 3.721057599327883e-06, "loss": 0.008, "step": 2432 }, { "epoch": 2.234159779614325, "grad_norm": 0.3570325970649719, "learning_rate": 3.712747353283901e-06, "loss": 0.0085, "step": 2433 }, { "epoch": 2.2350780532598713, "grad_norm": 0.7823039889335632, "learning_rate": 3.7044442813918125e-06, "loss": 0.0179, "step": 2434 }, { "epoch": 2.235996326905418, "grad_norm": 0.4119035601615906, "learning_rate": 3.6961483931259468e-06, "loss": 0.0059, "step": 2435 }, { "epoch": 2.236914600550964, "grad_norm": 0.8038538098335266, "learning_rate": 3.6878596979524306e-06, "loss": 0.0202, "step": 2436 }, { "epoch": 2.237832874196511, "grad_norm": 1.2781902551651, "learning_rate": 3.6795782053291963e-06, "loss": 0.0145, "step": 2437 }, { "epoch": 2.238751147842057, "grad_norm": 2.9761013984680176, "learning_rate": 3.671303924705946e-06, "loss": 0.0746, "step": 2438 }, { "epoch": 2.239669421487603, "grad_norm": 2.1986465454101562, "learning_rate": 3.6630368655241543e-06, "loss": 0.173, "step": 2439 }, { "epoch": 2.24058769513315, "grad_norm": 0.26577305793762207, "learning_rate": 3.6547770372170587e-06, "loss": 0.0031, "step": 2440 }, { "epoch": 2.241505968778696, "grad_norm": 0.27374911308288574, "learning_rate": 3.646524449209641e-06, "loss": 0.0052, "step": 2441 }, { "epoch": 2.242424242424242, "grad_norm": 0.36854618787765503, "learning_rate": 3.638279110918631e-06, "loss": 0.0074, "step": 2442 }, { "epoch": 2.243342516069789, "grad_norm": 1.3253982067108154, "learning_rate": 3.630041031752477e-06, "loss": 0.0845, "step": 2443 }, { "epoch": 2.244260789715335, "grad_norm": 1.2486592531204224, "learning_rate": 3.6218102211113458e-06, "loss": 0.0499, "step": 2444 }, { "epoch": 2.2451790633608817, "grad_norm": 1.1256142854690552, "learning_rate": 3.6135866883871097e-06, "loss": 0.0364, "step": 2445 }, { "epoch": 2.246097337006428, "grad_norm": 0.7972244024276733, "learning_rate": 3.605370442963344e-06, "loss": 0.0127, "step": 2446 }, { "epoch": 2.247015610651974, "grad_norm": 0.35453614592552185, "learning_rate": 3.5971614942153e-06, "loss": 0.0072, "step": 2447 }, { "epoch": 2.2479338842975207, "grad_norm": 0.1508754938840866, "learning_rate": 3.5889598515099066e-06, "loss": 0.0033, "step": 2448 }, { "epoch": 2.248852157943067, "grad_norm": 1.0750031471252441, "learning_rate": 3.580765524205755e-06, "loss": 0.0417, "step": 2449 }, { "epoch": 2.2497704315886136, "grad_norm": 1.7082961797714233, "learning_rate": 3.572578521653085e-06, "loss": 0.0736, "step": 2450 }, { "epoch": 2.2506887052341598, "grad_norm": 0.3136894106864929, "learning_rate": 3.5643988531937923e-06, "loss": 0.004, "step": 2451 }, { "epoch": 2.251606978879706, "grad_norm": 0.48219937086105347, "learning_rate": 3.5562265281613894e-06, "loss": 0.0081, "step": 2452 }, { "epoch": 2.2525252525252526, "grad_norm": 0.05858860909938812, "learning_rate": 3.548061555881015e-06, "loss": 0.0015, "step": 2453 }, { "epoch": 2.253443526170799, "grad_norm": 2.462329149246216, "learning_rate": 3.5399039456694186e-06, "loss": 0.121, "step": 2454 }, { "epoch": 2.2543617998163454, "grad_norm": 0.7764492630958557, "learning_rate": 3.5317537068349426e-06, "loss": 0.0125, "step": 2455 }, { "epoch": 2.2552800734618916, "grad_norm": 2.8270578384399414, "learning_rate": 3.5236108486775323e-06, "loss": 0.1855, "step": 2456 }, { "epoch": 2.256198347107438, "grad_norm": 1.729421854019165, "learning_rate": 3.5154753804886987e-06, "loss": 0.0878, "step": 2457 }, { "epoch": 2.2571166207529845, "grad_norm": 1.1429189443588257, "learning_rate": 3.507347311551523e-06, "loss": 0.0261, "step": 2458 }, { "epoch": 2.2580348943985307, "grad_norm": 0.19338180124759674, "learning_rate": 3.499226651140647e-06, "loss": 0.0037, "step": 2459 }, { "epoch": 2.2589531680440773, "grad_norm": 0.1777430772781372, "learning_rate": 3.4911134085222507e-06, "loss": 0.0046, "step": 2460 }, { "epoch": 2.2598714416896235, "grad_norm": 0.4482139050960541, "learning_rate": 3.483007592954064e-06, "loss": 0.0094, "step": 2461 }, { "epoch": 2.2607897153351697, "grad_norm": 0.11842993646860123, "learning_rate": 3.4749092136853304e-06, "loss": 0.0016, "step": 2462 }, { "epoch": 2.2617079889807163, "grad_norm": 0.13087882101535797, "learning_rate": 3.4668182799568104e-06, "loss": 0.0021, "step": 2463 }, { "epoch": 2.2626262626262625, "grad_norm": 0.6612945795059204, "learning_rate": 3.4587348010007703e-06, "loss": 0.0076, "step": 2464 }, { "epoch": 2.263544536271809, "grad_norm": 1.9466021060943604, "learning_rate": 3.450658786040967e-06, "loss": 0.0542, "step": 2465 }, { "epoch": 2.2644628099173554, "grad_norm": 0.0683688223361969, "learning_rate": 3.4425902442926495e-06, "loss": 0.0012, "step": 2466 }, { "epoch": 2.2653810835629016, "grad_norm": 0.30866166949272156, "learning_rate": 3.4345291849625283e-06, "loss": 0.01, "step": 2467 }, { "epoch": 2.266299357208448, "grad_norm": 0.3110257685184479, "learning_rate": 3.4264756172487813e-06, "loss": 0.0064, "step": 2468 }, { "epoch": 2.2672176308539944, "grad_norm": 0.36654701828956604, "learning_rate": 3.418429550341038e-06, "loss": 0.0095, "step": 2469 }, { "epoch": 2.268135904499541, "grad_norm": 1.5153489112854004, "learning_rate": 3.410390993420365e-06, "loss": 0.1149, "step": 2470 }, { "epoch": 2.2690541781450873, "grad_norm": 0.17080514132976532, "learning_rate": 3.402359955659269e-06, "loss": 0.003, "step": 2471 }, { "epoch": 2.2699724517906334, "grad_norm": 1.776566505432129, "learning_rate": 3.3943364462216665e-06, "loss": 0.1062, "step": 2472 }, { "epoch": 2.27089072543618, "grad_norm": 2.012523651123047, "learning_rate": 3.38632047426289e-06, "loss": 0.1008, "step": 2473 }, { "epoch": 2.2718089990817263, "grad_norm": 0.3849267065525055, "learning_rate": 3.3783120489296627e-06, "loss": 0.0072, "step": 2474 }, { "epoch": 2.2727272727272725, "grad_norm": 1.2448310852050781, "learning_rate": 3.370311179360114e-06, "loss": 0.0124, "step": 2475 }, { "epoch": 2.273645546372819, "grad_norm": 1.6750433444976807, "learning_rate": 3.3623178746837337e-06, "loss": 0.0745, "step": 2476 }, { "epoch": 2.2745638200183653, "grad_norm": 0.01864742301404476, "learning_rate": 3.3543321440213884e-06, "loss": 0.0005, "step": 2477 }, { "epoch": 2.275482093663912, "grad_norm": 2.979227066040039, "learning_rate": 3.346353996485301e-06, "loss": 0.0459, "step": 2478 }, { "epoch": 2.276400367309458, "grad_norm": 0.02859766036272049, "learning_rate": 3.338383441179036e-06, "loss": 0.0008, "step": 2479 }, { "epoch": 2.277318640955005, "grad_norm": 2.9358532428741455, "learning_rate": 3.330420487197509e-06, "loss": 0.3035, "step": 2480 }, { "epoch": 2.278236914600551, "grad_norm": 1.4725648164749146, "learning_rate": 3.3224651436269474e-06, "loss": 0.022, "step": 2481 }, { "epoch": 2.279155188246097, "grad_norm": 0.020239781588315964, "learning_rate": 3.3145174195449013e-06, "loss": 0.0006, "step": 2482 }, { "epoch": 2.280073461891644, "grad_norm": 0.18370984494686127, "learning_rate": 3.3065773240202235e-06, "loss": 0.0049, "step": 2483 }, { "epoch": 2.28099173553719, "grad_norm": 2.0112364292144775, "learning_rate": 3.298644866113063e-06, "loss": 0.0535, "step": 2484 }, { "epoch": 2.2819100091827362, "grad_norm": 0.13281796872615814, "learning_rate": 3.29072005487486e-06, "loss": 0.003, "step": 2485 }, { "epoch": 2.282828282828283, "grad_norm": 0.16511915624141693, "learning_rate": 3.282802899348321e-06, "loss": 0.0037, "step": 2486 }, { "epoch": 2.283746556473829, "grad_norm": 0.37696364521980286, "learning_rate": 3.274893408567419e-06, "loss": 0.0077, "step": 2487 }, { "epoch": 2.2846648301193757, "grad_norm": 1.7560406923294067, "learning_rate": 3.2669915915573836e-06, "loss": 0.0246, "step": 2488 }, { "epoch": 2.285583103764922, "grad_norm": 0.2240973860025406, "learning_rate": 3.259097457334681e-06, "loss": 0.0057, "step": 2489 }, { "epoch": 2.2865013774104685, "grad_norm": 0.12377997487783432, "learning_rate": 3.251211014907025e-06, "loss": 0.0025, "step": 2490 }, { "epoch": 2.2874196510560147, "grad_norm": 0.08734534680843353, "learning_rate": 3.2433322732733397e-06, "loss": 0.0019, "step": 2491 }, { "epoch": 2.288337924701561, "grad_norm": 0.5197944045066833, "learning_rate": 3.2354612414237673e-06, "loss": 0.017, "step": 2492 }, { "epoch": 2.2892561983471076, "grad_norm": 3.65740704536438, "learning_rate": 3.2275979283396495e-06, "loss": 0.0617, "step": 2493 }, { "epoch": 2.290174471992654, "grad_norm": 1.361937165260315, "learning_rate": 3.2197423429935217e-06, "loss": 0.0113, "step": 2494 }, { "epoch": 2.2910927456382, "grad_norm": 0.5386548638343811, "learning_rate": 3.211894494349107e-06, "loss": 0.0173, "step": 2495 }, { "epoch": 2.2920110192837466, "grad_norm": 1.692567229270935, "learning_rate": 3.2040543913612932e-06, "loss": 0.0532, "step": 2496 }, { "epoch": 2.292929292929293, "grad_norm": 0.6142813563346863, "learning_rate": 3.1962220429761336e-06, "loss": 0.0264, "step": 2497 }, { "epoch": 2.2938475665748395, "grad_norm": 1.1365841627120972, "learning_rate": 3.18839745813083e-06, "loss": 0.0296, "step": 2498 }, { "epoch": 2.2947658402203857, "grad_norm": 1.3411929607391357, "learning_rate": 3.1805806457537247e-06, "loss": 0.0646, "step": 2499 }, { "epoch": 2.2956841138659323, "grad_norm": 0.028581172227859497, "learning_rate": 3.1727716147643008e-06, "loss": 0.0007, "step": 2500 }, { "epoch": 2.2966023875114785, "grad_norm": 1.4188982248306274, "learning_rate": 3.1649703740731507e-06, "loss": 0.0496, "step": 2501 }, { "epoch": 2.2975206611570247, "grad_norm": 0.9670484662055969, "learning_rate": 3.1571769325819834e-06, "loss": 0.0181, "step": 2502 }, { "epoch": 2.2984389348025713, "grad_norm": 0.8114410042762756, "learning_rate": 3.1493912991836073e-06, "loss": 0.0207, "step": 2503 }, { "epoch": 2.2993572084481175, "grad_norm": 0.4215996563434601, "learning_rate": 3.1416134827619207e-06, "loss": 0.0107, "step": 2504 }, { "epoch": 2.3002754820936637, "grad_norm": 1.654029369354248, "learning_rate": 3.1338434921918993e-06, "loss": 0.0431, "step": 2505 }, { "epoch": 2.3011937557392104, "grad_norm": 2.572986364364624, "learning_rate": 3.1260813363396e-06, "loss": 0.1362, "step": 2506 }, { "epoch": 2.3021120293847566, "grad_norm": 0.3712492287158966, "learning_rate": 3.118327024062128e-06, "loss": 0.0054, "step": 2507 }, { "epoch": 2.303030303030303, "grad_norm": 1.2363102436065674, "learning_rate": 3.1105805642076435e-06, "loss": 0.0136, "step": 2508 }, { "epoch": 2.3039485766758494, "grad_norm": 3.174844980239868, "learning_rate": 3.102841965615345e-06, "loss": 0.1601, "step": 2509 }, { "epoch": 2.3048668503213956, "grad_norm": 0.33572930097579956, "learning_rate": 3.0951112371154635e-06, "loss": 0.007, "step": 2510 }, { "epoch": 2.3057851239669422, "grad_norm": 0.04543458670377731, "learning_rate": 3.0873883875292454e-06, "loss": 0.001, "step": 2511 }, { "epoch": 2.3067033976124884, "grad_norm": 0.031516265124082565, "learning_rate": 3.0796734256689476e-06, "loss": 0.0008, "step": 2512 }, { "epoch": 2.307621671258035, "grad_norm": 1.8570798635482788, "learning_rate": 3.0719663603378348e-06, "loss": 0.0876, "step": 2513 }, { "epoch": 2.3085399449035813, "grad_norm": 1.6901518106460571, "learning_rate": 3.06426720033015e-06, "loss": 0.0314, "step": 2514 }, { "epoch": 2.3094582185491275, "grad_norm": 0.1009795218706131, "learning_rate": 3.0565759544311223e-06, "loss": 0.0013, "step": 2515 }, { "epoch": 2.310376492194674, "grad_norm": 0.15918892621994019, "learning_rate": 3.0488926314169488e-06, "loss": 0.003, "step": 2516 }, { "epoch": 2.3112947658402203, "grad_norm": 0.21036164462566376, "learning_rate": 3.0412172400547836e-06, "loss": 0.0048, "step": 2517 }, { "epoch": 2.312213039485767, "grad_norm": 1.867573857307434, "learning_rate": 3.0335497891027298e-06, "loss": 0.1173, "step": 2518 }, { "epoch": 2.313131313131313, "grad_norm": 0.1875273436307907, "learning_rate": 3.0258902873098406e-06, "loss": 0.0036, "step": 2519 }, { "epoch": 2.3140495867768593, "grad_norm": 0.1415499746799469, "learning_rate": 3.0182387434160852e-06, "loss": 0.0024, "step": 2520 }, { "epoch": 2.314967860422406, "grad_norm": 0.26088228821754456, "learning_rate": 3.01059516615236e-06, "loss": 0.0039, "step": 2521 }, { "epoch": 2.315886134067952, "grad_norm": 0.12078636884689331, "learning_rate": 3.0029595642404686e-06, "loss": 0.0015, "step": 2522 }, { "epoch": 2.316804407713499, "grad_norm": 1.1396175622940063, "learning_rate": 2.995331946393111e-06, "loss": 0.0248, "step": 2523 }, { "epoch": 2.317722681359045, "grad_norm": 2.0162525177001953, "learning_rate": 2.987712321313885e-06, "loss": 0.0307, "step": 2524 }, { "epoch": 2.318640955004591, "grad_norm": 0.7487502694129944, "learning_rate": 2.9801006976972622e-06, "loss": 0.0337, "step": 2525 }, { "epoch": 2.319559228650138, "grad_norm": 1.7925126552581787, "learning_rate": 2.972497084228585e-06, "loss": 0.3401, "step": 2526 }, { "epoch": 2.320477502295684, "grad_norm": 0.2237427681684494, "learning_rate": 2.964901489584052e-06, "loss": 0.0053, "step": 2527 }, { "epoch": 2.3213957759412303, "grad_norm": 2.543327808380127, "learning_rate": 2.9573139224307234e-06, "loss": 0.039, "step": 2528 }, { "epoch": 2.322314049586777, "grad_norm": 0.2793721854686737, "learning_rate": 2.9497343914264885e-06, "loss": 0.0055, "step": 2529 }, { "epoch": 2.323232323232323, "grad_norm": 0.037090279161930084, "learning_rate": 2.9421629052200706e-06, "loss": 0.0008, "step": 2530 }, { "epoch": 2.3241505968778697, "grad_norm": 2.3366074562072754, "learning_rate": 2.934599472451012e-06, "loss": 0.0909, "step": 2531 }, { "epoch": 2.325068870523416, "grad_norm": 0.9602053761482239, "learning_rate": 2.927044101749663e-06, "loss": 0.0266, "step": 2532 }, { "epoch": 2.3259871441689626, "grad_norm": 0.08003444224596024, "learning_rate": 2.919496801737187e-06, "loss": 0.0013, "step": 2533 }, { "epoch": 2.3269054178145088, "grad_norm": 1.312590479850769, "learning_rate": 2.911957581025524e-06, "loss": 0.035, "step": 2534 }, { "epoch": 2.327823691460055, "grad_norm": 2.643156051635742, "learning_rate": 2.9044264482174023e-06, "loss": 0.0336, "step": 2535 }, { "epoch": 2.3287419651056016, "grad_norm": 0.7313743829727173, "learning_rate": 2.8969034119063176e-06, "loss": 0.0122, "step": 2536 }, { "epoch": 2.329660238751148, "grad_norm": 1.4735356569290161, "learning_rate": 2.8893884806765273e-06, "loss": 0.0496, "step": 2537 }, { "epoch": 2.330578512396694, "grad_norm": 1.7929165363311768, "learning_rate": 2.8818816631030466e-06, "loss": 0.0421, "step": 2538 }, { "epoch": 2.3314967860422406, "grad_norm": 4.305230617523193, "learning_rate": 2.8743829677516267e-06, "loss": 0.169, "step": 2539 }, { "epoch": 2.332415059687787, "grad_norm": 0.37915509939193726, "learning_rate": 2.8668924031787516e-06, "loss": 0.0083, "step": 2540 }, { "epoch": 2.3333333333333335, "grad_norm": 0.10107541084289551, "learning_rate": 2.8594099779316263e-06, "loss": 0.002, "step": 2541 }, { "epoch": 2.3342516069788797, "grad_norm": 0.28914961218833923, "learning_rate": 2.8519357005481685e-06, "loss": 0.0055, "step": 2542 }, { "epoch": 2.3351698806244263, "grad_norm": 0.20761729776859283, "learning_rate": 2.8444695795570067e-06, "loss": 0.0042, "step": 2543 }, { "epoch": 2.3360881542699725, "grad_norm": 0.1622348427772522, "learning_rate": 2.8370116234774513e-06, "loss": 0.0021, "step": 2544 }, { "epoch": 2.3370064279155187, "grad_norm": 1.319998860359192, "learning_rate": 2.8295618408195023e-06, "loss": 0.1116, "step": 2545 }, { "epoch": 2.3379247015610654, "grad_norm": 0.6892185807228088, "learning_rate": 2.822120240083831e-06, "loss": 0.0185, "step": 2546 }, { "epoch": 2.3388429752066116, "grad_norm": 0.3120400905609131, "learning_rate": 2.81468682976177e-06, "loss": 0.0063, "step": 2547 }, { "epoch": 2.3397612488521577, "grad_norm": 2.0055947303771973, "learning_rate": 2.8072616183353163e-06, "loss": 0.0812, "step": 2548 }, { "epoch": 2.3406795224977044, "grad_norm": 1.2935434579849243, "learning_rate": 2.7998446142771006e-06, "loss": 0.0259, "step": 2549 }, { "epoch": 2.3415977961432506, "grad_norm": 0.32454848289489746, "learning_rate": 2.7924358260503924e-06, "loss": 0.0062, "step": 2550 }, { "epoch": 2.3425160697887972, "grad_norm": 2.0418787002563477, "learning_rate": 2.785035262109087e-06, "loss": 0.0442, "step": 2551 }, { "epoch": 2.3434343434343434, "grad_norm": 0.046543002128601074, "learning_rate": 2.7776429308976914e-06, "loss": 0.0009, "step": 2552 }, { "epoch": 2.3443526170798896, "grad_norm": 0.1020079180598259, "learning_rate": 2.7702588408513276e-06, "loss": 0.0015, "step": 2553 }, { "epoch": 2.3452708907254363, "grad_norm": 0.8090242743492126, "learning_rate": 2.7628830003957063e-06, "loss": 0.0244, "step": 2554 }, { "epoch": 2.3461891643709825, "grad_norm": 0.22862492501735687, "learning_rate": 2.7555154179471256e-06, "loss": 0.0059, "step": 2555 }, { "epoch": 2.347107438016529, "grad_norm": 0.9009288549423218, "learning_rate": 2.748156101912459e-06, "loss": 0.0121, "step": 2556 }, { "epoch": 2.3480257116620753, "grad_norm": 0.04042191430926323, "learning_rate": 2.740805060689157e-06, "loss": 0.0011, "step": 2557 }, { "epoch": 2.3489439853076215, "grad_norm": 0.16544154286384583, "learning_rate": 2.7334623026652173e-06, "loss": 0.0031, "step": 2558 }, { "epoch": 2.349862258953168, "grad_norm": 1.748509168624878, "learning_rate": 2.726127836219191e-06, "loss": 0.0736, "step": 2559 }, { "epoch": 2.3507805325987143, "grad_norm": 1.4315698146820068, "learning_rate": 2.7188016697201658e-06, "loss": 0.0515, "step": 2560 }, { "epoch": 2.351698806244261, "grad_norm": 0.6745687127113342, "learning_rate": 2.7114838115277587e-06, "loss": 0.0161, "step": 2561 }, { "epoch": 2.352617079889807, "grad_norm": 2.490259885787964, "learning_rate": 2.7041742699921124e-06, "loss": 0.075, "step": 2562 }, { "epoch": 2.3535353535353534, "grad_norm": 1.2799923419952393, "learning_rate": 2.696873053453871e-06, "loss": 0.0522, "step": 2563 }, { "epoch": 2.3544536271809, "grad_norm": 1.3371598720550537, "learning_rate": 2.689580170244186e-06, "loss": 0.0649, "step": 2564 }, { "epoch": 2.355371900826446, "grad_norm": 0.288127601146698, "learning_rate": 2.6822956286846935e-06, "loss": 0.0033, "step": 2565 }, { "epoch": 2.356290174471993, "grad_norm": 0.4869992136955261, "learning_rate": 2.675019437087515e-06, "loss": 0.0097, "step": 2566 }, { "epoch": 2.357208448117539, "grad_norm": 0.774128794670105, "learning_rate": 2.667751603755249e-06, "loss": 0.0075, "step": 2567 }, { "epoch": 2.3581267217630852, "grad_norm": 1.4380367994308472, "learning_rate": 2.6604921369809476e-06, "loss": 0.0458, "step": 2568 }, { "epoch": 2.359044995408632, "grad_norm": 1.6380047798156738, "learning_rate": 2.653241045048123e-06, "loss": 0.0501, "step": 2569 }, { "epoch": 2.359963269054178, "grad_norm": 2.295801877975464, "learning_rate": 2.6459983362307263e-06, "loss": 0.1225, "step": 2570 }, { "epoch": 2.3608815426997243, "grad_norm": 0.29074129462242126, "learning_rate": 2.6387640187931428e-06, "loss": 0.0071, "step": 2571 }, { "epoch": 2.361799816345271, "grad_norm": 3.205228567123413, "learning_rate": 2.6315381009901907e-06, "loss": 0.1328, "step": 2572 }, { "epoch": 2.362718089990817, "grad_norm": 0.19601711630821228, "learning_rate": 2.6243205910670967e-06, "loss": 0.0034, "step": 2573 }, { "epoch": 2.3636363636363638, "grad_norm": 1.9930964708328247, "learning_rate": 2.6171114972594923e-06, "loss": 0.0204, "step": 2574 }, { "epoch": 2.36455463728191, "grad_norm": 1.957090139389038, "learning_rate": 2.6099108277934105e-06, "loss": 0.2454, "step": 2575 }, { "epoch": 2.3654729109274566, "grad_norm": 1.2435115575790405, "learning_rate": 2.602718590885266e-06, "loss": 0.0859, "step": 2576 }, { "epoch": 2.366391184573003, "grad_norm": 2.9373109340667725, "learning_rate": 2.5955347947418596e-06, "loss": 0.0977, "step": 2577 }, { "epoch": 2.367309458218549, "grad_norm": 0.8774164915084839, "learning_rate": 2.5883594475603547e-06, "loss": 0.015, "step": 2578 }, { "epoch": 2.3682277318640956, "grad_norm": 2.1985366344451904, "learning_rate": 2.5811925575282736e-06, "loss": 0.0428, "step": 2579 }, { "epoch": 2.369146005509642, "grad_norm": 0.9624273180961609, "learning_rate": 2.5740341328234875e-06, "loss": 0.0485, "step": 2580 }, { "epoch": 2.370064279155188, "grad_norm": 0.9280860424041748, "learning_rate": 2.5668841816142167e-06, "loss": 0.035, "step": 2581 }, { "epoch": 2.3709825528007347, "grad_norm": 1.0370209217071533, "learning_rate": 2.5597427120590036e-06, "loss": 0.0423, "step": 2582 }, { "epoch": 2.371900826446281, "grad_norm": 2.5337741374969482, "learning_rate": 2.5526097323067145e-06, "loss": 0.0463, "step": 2583 }, { "epoch": 2.3728191000918275, "grad_norm": 0.4135172367095947, "learning_rate": 2.54548525049653e-06, "loss": 0.0078, "step": 2584 }, { "epoch": 2.3737373737373737, "grad_norm": 1.5986052751541138, "learning_rate": 2.5383692747579313e-06, "loss": 0.0565, "step": 2585 }, { "epoch": 2.3746556473829203, "grad_norm": 0.915410578250885, "learning_rate": 2.5312618132107004e-06, "loss": 0.0204, "step": 2586 }, { "epoch": 2.3755739210284665, "grad_norm": 2.3877673149108887, "learning_rate": 2.524162873964896e-06, "loss": 0.1508, "step": 2587 }, { "epoch": 2.3764921946740127, "grad_norm": 0.41909176111221313, "learning_rate": 2.5170724651208567e-06, "loss": 0.004, "step": 2588 }, { "epoch": 2.3774104683195594, "grad_norm": 0.2534694969654083, "learning_rate": 2.509990594769186e-06, "loss": 0.0047, "step": 2589 }, { "epoch": 2.3783287419651056, "grad_norm": 0.36517879366874695, "learning_rate": 2.5029172709907414e-06, "loss": 0.0109, "step": 2590 }, { "epoch": 2.3792470156106518, "grad_norm": 2.232351064682007, "learning_rate": 2.495852501856637e-06, "loss": 0.0549, "step": 2591 }, { "epoch": 2.3801652892561984, "grad_norm": 0.09846024215221405, "learning_rate": 2.4887962954282175e-06, "loss": 0.0017, "step": 2592 }, { "epoch": 2.3810835629017446, "grad_norm": 0.0801422968506813, "learning_rate": 2.4817486597570606e-06, "loss": 0.0017, "step": 2593 }, { "epoch": 2.3820018365472913, "grad_norm": 0.548005998134613, "learning_rate": 2.4747096028849615e-06, "loss": 0.0198, "step": 2594 }, { "epoch": 2.3829201101928374, "grad_norm": 2.1759963035583496, "learning_rate": 2.4676791328439267e-06, "loss": 0.1102, "step": 2595 }, { "epoch": 2.3838383838383836, "grad_norm": 0.13843700289726257, "learning_rate": 2.4606572576561704e-06, "loss": 0.0022, "step": 2596 }, { "epoch": 2.3847566574839303, "grad_norm": 0.5533088445663452, "learning_rate": 2.4536439853340942e-06, "loss": 0.0162, "step": 2597 }, { "epoch": 2.3856749311294765, "grad_norm": 0.7968414425849915, "learning_rate": 2.446639323880282e-06, "loss": 0.0173, "step": 2598 }, { "epoch": 2.386593204775023, "grad_norm": 1.0637174844741821, "learning_rate": 2.4396432812874958e-06, "loss": 0.0179, "step": 2599 }, { "epoch": 2.3875114784205693, "grad_norm": 2.2582693099975586, "learning_rate": 2.432655865538659e-06, "loss": 0.1279, "step": 2600 }, { "epoch": 2.3884297520661155, "grad_norm": 0.5368929505348206, "learning_rate": 2.425677084606859e-06, "loss": 0.0138, "step": 2601 }, { "epoch": 2.389348025711662, "grad_norm": 1.8765724897384644, "learning_rate": 2.4187069464553215e-06, "loss": 0.0311, "step": 2602 }, { "epoch": 2.3902662993572084, "grad_norm": 0.16768869757652283, "learning_rate": 2.4117454590374134e-06, "loss": 0.0029, "step": 2603 }, { "epoch": 2.391184573002755, "grad_norm": 0.1828431934118271, "learning_rate": 2.404792630296633e-06, "loss": 0.0047, "step": 2604 }, { "epoch": 2.392102846648301, "grad_norm": 0.1970723271369934, "learning_rate": 2.397848468166596e-06, "loss": 0.0043, "step": 2605 }, { "epoch": 2.3930211202938474, "grad_norm": 0.2208438664674759, "learning_rate": 2.3909129805710284e-06, "loss": 0.0031, "step": 2606 }, { "epoch": 2.393939393939394, "grad_norm": 0.1597621589899063, "learning_rate": 2.383986175423757e-06, "loss": 0.0029, "step": 2607 }, { "epoch": 2.3948576675849402, "grad_norm": 0.42118754982948303, "learning_rate": 2.3770680606287088e-06, "loss": 0.004, "step": 2608 }, { "epoch": 2.395775941230487, "grad_norm": 2.7970027923583984, "learning_rate": 2.3701586440798863e-06, "loss": 0.0954, "step": 2609 }, { "epoch": 2.396694214876033, "grad_norm": 1.1904531717300415, "learning_rate": 2.3632579336613693e-06, "loss": 0.0315, "step": 2610 }, { "epoch": 2.3976124885215793, "grad_norm": 0.029392236843705177, "learning_rate": 2.3563659372473057e-06, "loss": 0.0006, "step": 2611 }, { "epoch": 2.398530762167126, "grad_norm": 1.426700472831726, "learning_rate": 2.3494826627018952e-06, "loss": 0.048, "step": 2612 }, { "epoch": 2.399449035812672, "grad_norm": 0.06780125945806503, "learning_rate": 2.342608117879388e-06, "loss": 0.0012, "step": 2613 }, { "epoch": 2.4003673094582183, "grad_norm": 1.6364099979400635, "learning_rate": 2.3357423106240772e-06, "loss": 0.1695, "step": 2614 }, { "epoch": 2.401285583103765, "grad_norm": 2.2678046226501465, "learning_rate": 2.328885248770281e-06, "loss": 0.1537, "step": 2615 }, { "epoch": 2.402203856749311, "grad_norm": 1.4187419414520264, "learning_rate": 2.322036940142338e-06, "loss": 0.1072, "step": 2616 }, { "epoch": 2.403122130394858, "grad_norm": 0.2581328749656677, "learning_rate": 2.3151973925546013e-06, "loss": 0.0053, "step": 2617 }, { "epoch": 2.404040404040404, "grad_norm": 0.31397542357444763, "learning_rate": 2.3083666138114246e-06, "loss": 0.0052, "step": 2618 }, { "epoch": 2.4049586776859506, "grad_norm": 1.7116750478744507, "learning_rate": 2.301544611707157e-06, "loss": 0.0118, "step": 2619 }, { "epoch": 2.405876951331497, "grad_norm": 0.06185396760702133, "learning_rate": 2.2947313940261373e-06, "loss": 0.0016, "step": 2620 }, { "epoch": 2.406795224977043, "grad_norm": 1.0356162786483765, "learning_rate": 2.2879269685426742e-06, "loss": 0.0155, "step": 2621 }, { "epoch": 2.4077134986225897, "grad_norm": 0.10553769767284393, "learning_rate": 2.2811313430210457e-06, "loss": 0.0014, "step": 2622 }, { "epoch": 2.408631772268136, "grad_norm": 2.0424914360046387, "learning_rate": 2.2743445252154907e-06, "loss": 0.1262, "step": 2623 }, { "epoch": 2.409550045913682, "grad_norm": 1.6168020963668823, "learning_rate": 2.2675665228701925e-06, "loss": 0.0659, "step": 2624 }, { "epoch": 2.4104683195592287, "grad_norm": 1.2901484966278076, "learning_rate": 2.2607973437192854e-06, "loss": 0.0242, "step": 2625 }, { "epoch": 2.411386593204775, "grad_norm": 0.3971695303916931, "learning_rate": 2.2540369954868267e-06, "loss": 0.0125, "step": 2626 }, { "epoch": 2.4123048668503215, "grad_norm": 2.6049318313598633, "learning_rate": 2.247285485886801e-06, "loss": 0.1614, "step": 2627 }, { "epoch": 2.4132231404958677, "grad_norm": 1.0835691690444946, "learning_rate": 2.2405428226231063e-06, "loss": 0.0177, "step": 2628 }, { "epoch": 2.4141414141414144, "grad_norm": 1.8864225149154663, "learning_rate": 2.233809013389543e-06, "loss": 0.149, "step": 2629 }, { "epoch": 2.4150596877869606, "grad_norm": 3.2133865356445312, "learning_rate": 2.227084065869819e-06, "loss": 0.2672, "step": 2630 }, { "epoch": 2.4159779614325068, "grad_norm": 1.3292202949523926, "learning_rate": 2.2203679877375205e-06, "loss": 0.0137, "step": 2631 }, { "epoch": 2.4168962350780534, "grad_norm": 0.6575788259506226, "learning_rate": 2.2136607866561155e-06, "loss": 0.0147, "step": 2632 }, { "epoch": 2.4178145087235996, "grad_norm": 0.7182263731956482, "learning_rate": 2.2069624702789426e-06, "loss": 0.009, "step": 2633 }, { "epoch": 2.418732782369146, "grad_norm": 1.3436055183410645, "learning_rate": 2.200273046249207e-06, "loss": 0.0336, "step": 2634 }, { "epoch": 2.4196510560146924, "grad_norm": 0.33913764357566833, "learning_rate": 2.1935925221999623e-06, "loss": 0.0087, "step": 2635 }, { "epoch": 2.4205693296602386, "grad_norm": 0.3970569670200348, "learning_rate": 2.1869209057541076e-06, "loss": 0.0084, "step": 2636 }, { "epoch": 2.4214876033057853, "grad_norm": 0.05815670266747475, "learning_rate": 2.1802582045243758e-06, "loss": 0.001, "step": 2637 }, { "epoch": 2.4224058769513315, "grad_norm": 0.22867566347122192, "learning_rate": 2.1736044261133305e-06, "loss": 0.0039, "step": 2638 }, { "epoch": 2.423324150596878, "grad_norm": 1.275484323501587, "learning_rate": 2.1669595781133545e-06, "loss": 0.0397, "step": 2639 }, { "epoch": 2.4242424242424243, "grad_norm": 0.9679417014122009, "learning_rate": 2.1603236681066374e-06, "loss": 0.0131, "step": 2640 }, { "epoch": 2.4251606978879705, "grad_norm": 0.2769470512866974, "learning_rate": 2.1536967036651722e-06, "loss": 0.0077, "step": 2641 }, { "epoch": 2.426078971533517, "grad_norm": 1.4241389036178589, "learning_rate": 2.1470786923507415e-06, "loss": 0.0431, "step": 2642 }, { "epoch": 2.4269972451790633, "grad_norm": 0.3719305396080017, "learning_rate": 2.140469641714912e-06, "loss": 0.0042, "step": 2643 }, { "epoch": 2.4279155188246095, "grad_norm": 0.8783397674560547, "learning_rate": 2.133869559299032e-06, "loss": 0.0181, "step": 2644 }, { "epoch": 2.428833792470156, "grad_norm": 1.3813852071762085, "learning_rate": 2.1272784526342105e-06, "loss": 0.0795, "step": 2645 }, { "epoch": 2.4297520661157024, "grad_norm": 1.422187328338623, "learning_rate": 2.120696329241315e-06, "loss": 0.0814, "step": 2646 }, { "epoch": 2.430670339761249, "grad_norm": 0.47917434573173523, "learning_rate": 2.114123196630964e-06, "loss": 0.0098, "step": 2647 }, { "epoch": 2.431588613406795, "grad_norm": 0.5204207301139832, "learning_rate": 2.1075590623035134e-06, "loss": 0.01, "step": 2648 }, { "epoch": 2.4325068870523414, "grad_norm": 0.06301619112491608, "learning_rate": 2.1010039337490597e-06, "loss": 0.0013, "step": 2649 }, { "epoch": 2.433425160697888, "grad_norm": 1.3351457118988037, "learning_rate": 2.094457818447416e-06, "loss": 0.0448, "step": 2650 }, { "epoch": 2.4343434343434343, "grad_norm": 0.7453467845916748, "learning_rate": 2.0879207238681133e-06, "loss": 0.0143, "step": 2651 }, { "epoch": 2.435261707988981, "grad_norm": 2.007072925567627, "learning_rate": 2.081392657470388e-06, "loss": 0.0788, "step": 2652 }, { "epoch": 2.436179981634527, "grad_norm": 1.33317232131958, "learning_rate": 2.074873626703172e-06, "loss": 0.0211, "step": 2653 }, { "epoch": 2.4370982552800733, "grad_norm": 0.11986751854419708, "learning_rate": 2.0683636390050986e-06, "loss": 0.0013, "step": 2654 }, { "epoch": 2.43801652892562, "grad_norm": 0.31331661343574524, "learning_rate": 2.06186270180447e-06, "loss": 0.0068, "step": 2655 }, { "epoch": 2.438934802571166, "grad_norm": 0.8929907083511353, "learning_rate": 2.055370822519266e-06, "loss": 0.012, "step": 2656 }, { "epoch": 2.4398530762167128, "grad_norm": 0.15286102890968323, "learning_rate": 2.0488880085571292e-06, "loss": 0.0043, "step": 2657 }, { "epoch": 2.440771349862259, "grad_norm": 0.5426670908927917, "learning_rate": 2.042414267315359e-06, "loss": 0.0123, "step": 2658 }, { "epoch": 2.441689623507805, "grad_norm": 1.734033226966858, "learning_rate": 2.0359496061809068e-06, "loss": 0.0982, "step": 2659 }, { "epoch": 2.442607897153352, "grad_norm": 3.085829734802246, "learning_rate": 2.029494032530356e-06, "loss": 0.2914, "step": 2660 }, { "epoch": 2.443526170798898, "grad_norm": 1.4394298791885376, "learning_rate": 2.0230475537299234e-06, "loss": 0.0662, "step": 2661 }, { "epoch": 2.4444444444444446, "grad_norm": 1.4141287803649902, "learning_rate": 2.016610177135445e-06, "loss": 0.0451, "step": 2662 }, { "epoch": 2.445362718089991, "grad_norm": 2.6319961547851562, "learning_rate": 2.010181910092379e-06, "loss": 0.0734, "step": 2663 }, { "epoch": 2.446280991735537, "grad_norm": 2.2960879802703857, "learning_rate": 2.0037627599357824e-06, "loss": 0.0701, "step": 2664 }, { "epoch": 2.4471992653810837, "grad_norm": 0.7884027361869812, "learning_rate": 1.9973527339903087e-06, "loss": 0.0506, "step": 2665 }, { "epoch": 2.44811753902663, "grad_norm": 0.7327689528465271, "learning_rate": 1.9909518395702022e-06, "loss": 0.0161, "step": 2666 }, { "epoch": 2.449035812672176, "grad_norm": 0.6578026413917542, "learning_rate": 1.9845600839792846e-06, "loss": 0.0103, "step": 2667 }, { "epoch": 2.4499540863177227, "grad_norm": 0.24825480580329895, "learning_rate": 1.9781774745109584e-06, "loss": 0.0037, "step": 2668 }, { "epoch": 2.450872359963269, "grad_norm": 0.20907500386238098, "learning_rate": 1.971804018448179e-06, "loss": 0.0028, "step": 2669 }, { "epoch": 2.4517906336088156, "grad_norm": 1.0168977975845337, "learning_rate": 1.965439723063465e-06, "loss": 0.0276, "step": 2670 }, { "epoch": 2.4527089072543617, "grad_norm": 1.0789552927017212, "learning_rate": 1.9590845956188754e-06, "loss": 0.0206, "step": 2671 }, { "epoch": 2.4536271808999084, "grad_norm": 0.6153807640075684, "learning_rate": 1.952738643366011e-06, "loss": 0.0106, "step": 2672 }, { "epoch": 2.4545454545454546, "grad_norm": 1.2309914827346802, "learning_rate": 1.946401873546009e-06, "loss": 0.0824, "step": 2673 }, { "epoch": 2.455463728191001, "grad_norm": 3.844825267791748, "learning_rate": 1.94007429338952e-06, "loss": 0.1099, "step": 2674 }, { "epoch": 2.4563820018365474, "grad_norm": 1.7821365594863892, "learning_rate": 1.9337559101167126e-06, "loss": 0.0519, "step": 2675 }, { "epoch": 2.4573002754820936, "grad_norm": 0.24311484396457672, "learning_rate": 1.9274467309372613e-06, "loss": 0.0031, "step": 2676 }, { "epoch": 2.45821854912764, "grad_norm": 0.8861656188964844, "learning_rate": 1.9211467630503333e-06, "loss": 0.016, "step": 2677 }, { "epoch": 2.4591368227731865, "grad_norm": 1.3018616437911987, "learning_rate": 1.914856013644596e-06, "loss": 0.0454, "step": 2678 }, { "epoch": 2.4600550964187327, "grad_norm": 1.6557914018630981, "learning_rate": 1.9085744898981886e-06, "loss": 0.0658, "step": 2679 }, { "epoch": 2.4609733700642793, "grad_norm": 0.7458274960517883, "learning_rate": 1.9023021989787249e-06, "loss": 0.0136, "step": 2680 }, { "epoch": 2.4618916437098255, "grad_norm": 0.1725546270608902, "learning_rate": 1.896039148043285e-06, "loss": 0.0044, "step": 2681 }, { "epoch": 2.462809917355372, "grad_norm": 3.438905954360962, "learning_rate": 1.8897853442384028e-06, "loss": 0.1762, "step": 2682 }, { "epoch": 2.4637281910009183, "grad_norm": 0.38326284289360046, "learning_rate": 1.8835407947000683e-06, "loss": 0.0062, "step": 2683 }, { "epoch": 2.4646464646464645, "grad_norm": 0.18040518462657928, "learning_rate": 1.8773055065537038e-06, "loss": 0.0049, "step": 2684 }, { "epoch": 2.465564738292011, "grad_norm": 0.20343388617038727, "learning_rate": 1.8710794869141668e-06, "loss": 0.0044, "step": 2685 }, { "epoch": 2.4664830119375574, "grad_norm": 2.018991470336914, "learning_rate": 1.864862742885737e-06, "loss": 0.0767, "step": 2686 }, { "epoch": 2.4674012855831036, "grad_norm": 1.170008897781372, "learning_rate": 1.8586552815621105e-06, "loss": 0.0751, "step": 2687 }, { "epoch": 2.46831955922865, "grad_norm": 1.198686957359314, "learning_rate": 1.8524571100263966e-06, "loss": 0.0475, "step": 2688 }, { "epoch": 2.4692378328741964, "grad_norm": 1.7048529386520386, "learning_rate": 1.8462682353510974e-06, "loss": 0.2304, "step": 2689 }, { "epoch": 2.470156106519743, "grad_norm": 0.33525997400283813, "learning_rate": 1.8400886645981098e-06, "loss": 0.0059, "step": 2690 }, { "epoch": 2.4710743801652892, "grad_norm": 2.000349283218384, "learning_rate": 1.833918404818711e-06, "loss": 0.0789, "step": 2691 }, { "epoch": 2.4719926538108354, "grad_norm": 0.5867419838905334, "learning_rate": 1.8277574630535611e-06, "loss": 0.0084, "step": 2692 }, { "epoch": 2.472910927456382, "grad_norm": 1.356623649597168, "learning_rate": 1.8216058463326824e-06, "loss": 0.2193, "step": 2693 }, { "epoch": 2.4738292011019283, "grad_norm": 2.9576778411865234, "learning_rate": 1.8154635616754568e-06, "loss": 0.1512, "step": 2694 }, { "epoch": 2.474747474747475, "grad_norm": 0.3592931032180786, "learning_rate": 1.8093306160906176e-06, "loss": 0.0061, "step": 2695 }, { "epoch": 2.475665748393021, "grad_norm": 1.5697753429412842, "learning_rate": 1.8032070165762416e-06, "loss": 0.0145, "step": 2696 }, { "epoch": 2.4765840220385673, "grad_norm": 2.7666683197021484, "learning_rate": 1.797092770119746e-06, "loss": 0.0833, "step": 2697 }, { "epoch": 2.477502295684114, "grad_norm": 0.11260843276977539, "learning_rate": 1.7909878836978712e-06, "loss": 0.0027, "step": 2698 }, { "epoch": 2.47842056932966, "grad_norm": 1.4134695529937744, "learning_rate": 1.7848923642766747e-06, "loss": 0.0349, "step": 2699 }, { "epoch": 2.479338842975207, "grad_norm": 1.3847624063491821, "learning_rate": 1.7788062188115307e-06, "loss": 0.0335, "step": 2700 }, { "epoch": 2.480257116620753, "grad_norm": 0.7041518688201904, "learning_rate": 1.7727294542471152e-06, "loss": 0.0328, "step": 2701 }, { "epoch": 2.481175390266299, "grad_norm": 0.5707157254219055, "learning_rate": 1.766662077517396e-06, "loss": 0.0096, "step": 2702 }, { "epoch": 2.482093663911846, "grad_norm": 0.8910525441169739, "learning_rate": 1.7606040955456394e-06, "loss": 0.014, "step": 2703 }, { "epoch": 2.483011937557392, "grad_norm": 0.617490828037262, "learning_rate": 1.7545555152443817e-06, "loss": 0.0101, "step": 2704 }, { "epoch": 2.4839302112029387, "grad_norm": 1.1317641735076904, "learning_rate": 1.748516343515434e-06, "loss": 0.0743, "step": 2705 }, { "epoch": 2.484848484848485, "grad_norm": 1.699375867843628, "learning_rate": 1.742486587249873e-06, "loss": 0.0502, "step": 2706 }, { "epoch": 2.485766758494031, "grad_norm": 3.5571696758270264, "learning_rate": 1.7364662533280308e-06, "loss": 0.0826, "step": 2707 }, { "epoch": 2.4866850321395777, "grad_norm": 0.32086634635925293, "learning_rate": 1.7304553486194852e-06, "loss": 0.0068, "step": 2708 }, { "epoch": 2.487603305785124, "grad_norm": 1.0324501991271973, "learning_rate": 1.7244538799830646e-06, "loss": 0.0228, "step": 2709 }, { "epoch": 2.48852157943067, "grad_norm": 0.3218711018562317, "learning_rate": 1.7184618542668173e-06, "loss": 0.0055, "step": 2710 }, { "epoch": 2.4894398530762167, "grad_norm": 0.15443678200244904, "learning_rate": 1.7124792783080258e-06, "loss": 0.0029, "step": 2711 }, { "epoch": 2.490358126721763, "grad_norm": 0.1050894483923912, "learning_rate": 1.7065061589331855e-06, "loss": 0.002, "step": 2712 }, { "epoch": 2.4912764003673096, "grad_norm": 0.11726363748311996, "learning_rate": 1.7005425029580024e-06, "loss": 0.0022, "step": 2713 }, { "epoch": 2.4921946740128558, "grad_norm": 0.7845532298088074, "learning_rate": 1.6945883171873822e-06, "loss": 0.038, "step": 2714 }, { "epoch": 2.4931129476584024, "grad_norm": 0.36780649423599243, "learning_rate": 1.68864360841543e-06, "loss": 0.0076, "step": 2715 }, { "epoch": 2.4940312213039486, "grad_norm": 0.09932395070791245, "learning_rate": 1.6827083834254331e-06, "loss": 0.0027, "step": 2716 }, { "epoch": 2.494949494949495, "grad_norm": 0.7114999294281006, "learning_rate": 1.6767826489898564e-06, "loss": 0.0145, "step": 2717 }, { "epoch": 2.4958677685950414, "grad_norm": 0.3183249831199646, "learning_rate": 1.6708664118703365e-06, "loss": 0.0101, "step": 2718 }, { "epoch": 2.4967860422405876, "grad_norm": 1.5758146047592163, "learning_rate": 1.6649596788176737e-06, "loss": 0.1342, "step": 2719 }, { "epoch": 2.497704315886134, "grad_norm": 0.39390358328819275, "learning_rate": 1.6590624565718193e-06, "loss": 0.003, "step": 2720 }, { "epoch": 2.4986225895316805, "grad_norm": 1.1878360509872437, "learning_rate": 1.6531747518618834e-06, "loss": 0.0137, "step": 2721 }, { "epoch": 2.4995408631772267, "grad_norm": 0.07133430987596512, "learning_rate": 1.647296571406104e-06, "loss": 0.0015, "step": 2722 }, { "epoch": 2.5004591368227733, "grad_norm": 0.8690692186355591, "learning_rate": 1.6414279219118568e-06, "loss": 0.0342, "step": 2723 }, { "epoch": 2.5013774104683195, "grad_norm": 0.08707112073898315, "learning_rate": 1.6355688100756407e-06, "loss": 0.0022, "step": 2724 }, { "epoch": 2.502295684113866, "grad_norm": 0.8081445693969727, "learning_rate": 1.6297192425830689e-06, "loss": 0.0149, "step": 2725 }, { "epoch": 2.5032139577594124, "grad_norm": 0.9214422106742859, "learning_rate": 1.623879226108872e-06, "loss": 0.0241, "step": 2726 }, { "epoch": 2.5041322314049586, "grad_norm": 2.4715449810028076, "learning_rate": 1.6180487673168754e-06, "loss": 0.1087, "step": 2727 }, { "epoch": 2.505050505050505, "grad_norm": 1.6869163513183594, "learning_rate": 1.6122278728600005e-06, "loss": 0.1234, "step": 2728 }, { "epoch": 2.5059687786960514, "grad_norm": 0.31028252840042114, "learning_rate": 1.6064165493802553e-06, "loss": 0.0059, "step": 2729 }, { "epoch": 2.5068870523415976, "grad_norm": 2.128185272216797, "learning_rate": 1.600614803508722e-06, "loss": 0.0364, "step": 2730 }, { "epoch": 2.5078053259871442, "grad_norm": 1.5447510480880737, "learning_rate": 1.5948226418655665e-06, "loss": 0.0192, "step": 2731 }, { "epoch": 2.5087235996326904, "grad_norm": 4.595907211303711, "learning_rate": 1.5890400710600063e-06, "loss": 0.2514, "step": 2732 }, { "epoch": 2.509641873278237, "grad_norm": 0.16256779432296753, "learning_rate": 1.5832670976903186e-06, "loss": 0.0042, "step": 2733 }, { "epoch": 2.5105601469237833, "grad_norm": 0.1701440066099167, "learning_rate": 1.5775037283438322e-06, "loss": 0.0024, "step": 2734 }, { "epoch": 2.51147842056933, "grad_norm": 0.5478557348251343, "learning_rate": 1.5717499695969118e-06, "loss": 0.0103, "step": 2735 }, { "epoch": 2.512396694214876, "grad_norm": 0.9022307991981506, "learning_rate": 1.5660058280149648e-06, "loss": 0.0205, "step": 2736 }, { "epoch": 2.5133149678604223, "grad_norm": 0.6886894702911377, "learning_rate": 1.5602713101524147e-06, "loss": 0.0118, "step": 2737 }, { "epoch": 2.514233241505969, "grad_norm": 2.526870012283325, "learning_rate": 1.5545464225527097e-06, "loss": 0.0811, "step": 2738 }, { "epoch": 2.515151515151515, "grad_norm": 0.4754156470298767, "learning_rate": 1.5488311717483085e-06, "loss": 0.0073, "step": 2739 }, { "epoch": 2.5160697887970613, "grad_norm": 0.25233811140060425, "learning_rate": 1.543125564260668e-06, "loss": 0.0054, "step": 2740 }, { "epoch": 2.516988062442608, "grad_norm": 1.2411692142486572, "learning_rate": 1.5374296066002526e-06, "loss": 0.0391, "step": 2741 }, { "epoch": 2.517906336088154, "grad_norm": 0.06376565992832184, "learning_rate": 1.5317433052665066e-06, "loss": 0.0016, "step": 2742 }, { "epoch": 2.5188246097337004, "grad_norm": 0.8136380910873413, "learning_rate": 1.5260666667478586e-06, "loss": 0.0135, "step": 2743 }, { "epoch": 2.519742883379247, "grad_norm": 1.289158821105957, "learning_rate": 1.5203996975217096e-06, "loss": 0.0284, "step": 2744 }, { "epoch": 2.5206611570247937, "grad_norm": 0.01640811562538147, "learning_rate": 1.5147424040544322e-06, "loss": 0.0004, "step": 2745 }, { "epoch": 2.52157943067034, "grad_norm": 4.026873588562012, "learning_rate": 1.5090947928013533e-06, "loss": 0.1924, "step": 2746 }, { "epoch": 2.522497704315886, "grad_norm": 0.43920791149139404, "learning_rate": 1.5034568702067543e-06, "loss": 0.0063, "step": 2747 }, { "epoch": 2.5234159779614327, "grad_norm": 0.6853976249694824, "learning_rate": 1.4978286427038602e-06, "loss": 0.0262, "step": 2748 }, { "epoch": 2.524334251606979, "grad_norm": 0.06293096393346786, "learning_rate": 1.4922101167148305e-06, "loss": 0.0015, "step": 2749 }, { "epoch": 2.525252525252525, "grad_norm": 0.5779849290847778, "learning_rate": 1.486601298650764e-06, "loss": 0.0138, "step": 2750 }, { "epoch": 2.5261707988980717, "grad_norm": 0.6501057744026184, "learning_rate": 1.4810021949116726e-06, "loss": 0.0166, "step": 2751 }, { "epoch": 2.527089072543618, "grad_norm": 0.7591861486434937, "learning_rate": 1.4754128118864874e-06, "loss": 0.0167, "step": 2752 }, { "epoch": 2.528007346189164, "grad_norm": 1.0648576021194458, "learning_rate": 1.4698331559530465e-06, "loss": 0.0173, "step": 2753 }, { "epoch": 2.5289256198347108, "grad_norm": 0.12788216769695282, "learning_rate": 1.464263233478086e-06, "loss": 0.0021, "step": 2754 }, { "epoch": 2.529843893480257, "grad_norm": 0.6297221183776855, "learning_rate": 1.4587030508172461e-06, "loss": 0.0106, "step": 2755 }, { "epoch": 2.5307621671258036, "grad_norm": 1.3269202709197998, "learning_rate": 1.453152614315041e-06, "loss": 0.0346, "step": 2756 }, { "epoch": 2.53168044077135, "grad_norm": 2.5461010932922363, "learning_rate": 1.4476119303048709e-06, "loss": 0.0663, "step": 2757 }, { "epoch": 2.5325987144168964, "grad_norm": 0.02888348326086998, "learning_rate": 1.442081005109005e-06, "loss": 0.0007, "step": 2758 }, { "epoch": 2.5335169880624426, "grad_norm": 0.12256601452827454, "learning_rate": 1.4365598450385742e-06, "loss": 0.0025, "step": 2759 }, { "epoch": 2.534435261707989, "grad_norm": 0.7276080250740051, "learning_rate": 1.4310484563935778e-06, "loss": 0.0146, "step": 2760 }, { "epoch": 2.5353535353535355, "grad_norm": 0.3559943735599518, "learning_rate": 1.4255468454628541e-06, "loss": 0.0033, "step": 2761 }, { "epoch": 2.5362718089990817, "grad_norm": 0.1259441375732422, "learning_rate": 1.420055018524088e-06, "loss": 0.0017, "step": 2762 }, { "epoch": 2.537190082644628, "grad_norm": 0.09513731300830841, "learning_rate": 1.4145729818438004e-06, "loss": 0.002, "step": 2763 }, { "epoch": 2.5381083562901745, "grad_norm": 0.1023062989115715, "learning_rate": 1.4091007416773395e-06, "loss": 0.0022, "step": 2764 }, { "epoch": 2.5390266299357207, "grad_norm": 4.23818302154541, "learning_rate": 1.4036383042688794e-06, "loss": 0.1492, "step": 2765 }, { "epoch": 2.5399449035812673, "grad_norm": 0.8712511658668518, "learning_rate": 1.3981856758514066e-06, "loss": 0.0096, "step": 2766 }, { "epoch": 2.5408631772268135, "grad_norm": 0.7189885973930359, "learning_rate": 1.392742862646711e-06, "loss": 0.0106, "step": 2767 }, { "epoch": 2.54178145087236, "grad_norm": 0.3436954915523529, "learning_rate": 1.3873098708653855e-06, "loss": 0.0075, "step": 2768 }, { "epoch": 2.5426997245179064, "grad_norm": 2.7826664447784424, "learning_rate": 1.3818867067068199e-06, "loss": 0.0874, "step": 2769 }, { "epoch": 2.5436179981634526, "grad_norm": 1.9858994483947754, "learning_rate": 1.3764733763591852e-06, "loss": 0.2644, "step": 2770 }, { "epoch": 2.544536271808999, "grad_norm": 1.466618537902832, "learning_rate": 1.371069885999432e-06, "loss": 0.06, "step": 2771 }, { "epoch": 2.5454545454545454, "grad_norm": 8.158011436462402, "learning_rate": 1.3656762417932845e-06, "loss": 0.5403, "step": 2772 }, { "epoch": 2.5463728191000916, "grad_norm": 3.076306104660034, "learning_rate": 1.3602924498952274e-06, "loss": 0.1969, "step": 2773 }, { "epoch": 2.5472910927456383, "grad_norm": 1.404654622077942, "learning_rate": 1.3549185164485135e-06, "loss": 0.0281, "step": 2774 }, { "epoch": 2.5482093663911844, "grad_norm": 0.7059173583984375, "learning_rate": 1.3495544475851352e-06, "loss": 0.0159, "step": 2775 }, { "epoch": 2.549127640036731, "grad_norm": 0.15187498927116394, "learning_rate": 1.3442002494258356e-06, "loss": 0.0029, "step": 2776 }, { "epoch": 2.5500459136822773, "grad_norm": 1.2744050025939941, "learning_rate": 1.3388559280800906e-06, "loss": 0.0118, "step": 2777 }, { "epoch": 2.550964187327824, "grad_norm": 1.4553592205047607, "learning_rate": 1.333521489646108e-06, "loss": 0.043, "step": 2778 }, { "epoch": 2.55188246097337, "grad_norm": 1.5437729358673096, "learning_rate": 1.3281969402108207e-06, "loss": 0.0304, "step": 2779 }, { "epoch": 2.5528007346189163, "grad_norm": 0.892447292804718, "learning_rate": 1.3228822858498746e-06, "loss": 0.0155, "step": 2780 }, { "epoch": 2.553719008264463, "grad_norm": 0.41516509652137756, "learning_rate": 1.3175775326276263e-06, "loss": 0.0086, "step": 2781 }, { "epoch": 2.554637281910009, "grad_norm": 0.05955673009157181, "learning_rate": 1.3122826865971328e-06, "loss": 0.0013, "step": 2782 }, { "epoch": 2.5555555555555554, "grad_norm": 2.9935965538024902, "learning_rate": 1.3069977538001443e-06, "loss": 0.1312, "step": 2783 }, { "epoch": 2.556473829201102, "grad_norm": 0.8288334608078003, "learning_rate": 1.3017227402671085e-06, "loss": 0.0192, "step": 2784 }, { "epoch": 2.557392102846648, "grad_norm": 0.8318669199943542, "learning_rate": 1.2964576520171458e-06, "loss": 0.0219, "step": 2785 }, { "epoch": 2.558310376492195, "grad_norm": 0.6842146515846252, "learning_rate": 1.2912024950580549e-06, "loss": 0.0216, "step": 2786 }, { "epoch": 2.559228650137741, "grad_norm": 0.05546356365084648, "learning_rate": 1.2859572753863004e-06, "loss": 0.0014, "step": 2787 }, { "epoch": 2.5601469237832877, "grad_norm": 0.11914683878421783, "learning_rate": 1.2807219989870056e-06, "loss": 0.0026, "step": 2788 }, { "epoch": 2.561065197428834, "grad_norm": 2.146909236907959, "learning_rate": 1.275496671833958e-06, "loss": 0.1278, "step": 2789 }, { "epoch": 2.56198347107438, "grad_norm": 1.6532269716262817, "learning_rate": 1.270281299889582e-06, "loss": 0.0621, "step": 2790 }, { "epoch": 2.5629017447199267, "grad_norm": 0.9215152859687805, "learning_rate": 1.2650758891049464e-06, "loss": 0.0242, "step": 2791 }, { "epoch": 2.563820018365473, "grad_norm": 0.09179025143384933, "learning_rate": 1.2598804454197522e-06, "loss": 0.0019, "step": 2792 }, { "epoch": 2.564738292011019, "grad_norm": 0.11939460784196854, "learning_rate": 1.2546949747623271e-06, "loss": 0.0033, "step": 2793 }, { "epoch": 2.5656565656565657, "grad_norm": 0.028229856863617897, "learning_rate": 1.249519483049626e-06, "loss": 0.0007, "step": 2794 }, { "epoch": 2.566574839302112, "grad_norm": 1.545397162437439, "learning_rate": 1.244353976187206e-06, "loss": 0.0498, "step": 2795 }, { "epoch": 2.567493112947658, "grad_norm": 1.9265278577804565, "learning_rate": 1.2391984600692386e-06, "loss": 0.0633, "step": 2796 }, { "epoch": 2.568411386593205, "grad_norm": 1.0827645063400269, "learning_rate": 1.2340529405784906e-06, "loss": 0.0078, "step": 2797 }, { "epoch": 2.5693296602387514, "grad_norm": 1.6443945169448853, "learning_rate": 1.2289174235863276e-06, "loss": 0.0509, "step": 2798 }, { "epoch": 2.5702479338842976, "grad_norm": 0.6539214253425598, "learning_rate": 1.2237919149526977e-06, "loss": 0.013, "step": 2799 }, { "epoch": 2.571166207529844, "grad_norm": 1.2936651706695557, "learning_rate": 1.218676420526128e-06, "loss": 0.0871, "step": 2800 }, { "epoch": 2.5720844811753905, "grad_norm": 2.310471296310425, "learning_rate": 1.2135709461437218e-06, "loss": 0.0524, "step": 2801 }, { "epoch": 2.5730027548209367, "grad_norm": 1.5138766765594482, "learning_rate": 1.2084754976311474e-06, "loss": 0.137, "step": 2802 }, { "epoch": 2.573921028466483, "grad_norm": 0.39970409870147705, "learning_rate": 1.20339008080263e-06, "loss": 0.0049, "step": 2803 }, { "epoch": 2.5748393021120295, "grad_norm": 0.4192346930503845, "learning_rate": 1.1983147014609552e-06, "loss": 0.0054, "step": 2804 }, { "epoch": 2.5757575757575757, "grad_norm": 0.12194506824016571, "learning_rate": 1.1932493653974487e-06, "loss": 0.0029, "step": 2805 }, { "epoch": 2.576675849403122, "grad_norm": 0.5168948769569397, "learning_rate": 1.1881940783919798e-06, "loss": 0.0075, "step": 2806 }, { "epoch": 2.5775941230486685, "grad_norm": 0.16746798157691956, "learning_rate": 1.1831488462129493e-06, "loss": 0.0024, "step": 2807 }, { "epoch": 2.5785123966942147, "grad_norm": 0.7471609711647034, "learning_rate": 1.178113674617285e-06, "loss": 0.0216, "step": 2808 }, { "epoch": 2.5794306703397614, "grad_norm": 0.0393698513507843, "learning_rate": 1.1730885693504323e-06, "loss": 0.0009, "step": 2809 }, { "epoch": 2.5803489439853076, "grad_norm": 2.1037158966064453, "learning_rate": 1.1680735361463601e-06, "loss": 0.0625, "step": 2810 }, { "epoch": 2.581267217630854, "grad_norm": 2.1559438705444336, "learning_rate": 1.1630685807275332e-06, "loss": 0.1354, "step": 2811 }, { "epoch": 2.5821854912764004, "grad_norm": 1.3067821264266968, "learning_rate": 1.1580737088049232e-06, "loss": 0.0407, "step": 2812 }, { "epoch": 2.5831037649219466, "grad_norm": 0.3614616394042969, "learning_rate": 1.1530889260779932e-06, "loss": 0.0078, "step": 2813 }, { "epoch": 2.5840220385674932, "grad_norm": 0.05607493594288826, "learning_rate": 1.1481142382346945e-06, "loss": 0.0014, "step": 2814 }, { "epoch": 2.5849403122130394, "grad_norm": 1.1692878007888794, "learning_rate": 1.1431496509514573e-06, "loss": 0.0147, "step": 2815 }, { "epoch": 2.5858585858585856, "grad_norm": 1.1829229593276978, "learning_rate": 1.1381951698931949e-06, "loss": 0.0601, "step": 2816 }, { "epoch": 2.5867768595041323, "grad_norm": 0.6711520552635193, "learning_rate": 1.133250800713278e-06, "loss": 0.0108, "step": 2817 }, { "epoch": 2.5876951331496785, "grad_norm": 0.2948712706565857, "learning_rate": 1.1283165490535463e-06, "loss": 0.0047, "step": 2818 }, { "epoch": 2.588613406795225, "grad_norm": 2.353947401046753, "learning_rate": 1.1233924205442915e-06, "loss": 0.1666, "step": 2819 }, { "epoch": 2.5895316804407713, "grad_norm": 0.8725816011428833, "learning_rate": 1.1184784208042532e-06, "loss": 0.014, "step": 2820 }, { "epoch": 2.590449954086318, "grad_norm": 0.3179554045200348, "learning_rate": 1.1135745554406153e-06, "loss": 0.0065, "step": 2821 }, { "epoch": 2.591368227731864, "grad_norm": 0.04472342133522034, "learning_rate": 1.1086808300489994e-06, "loss": 0.0009, "step": 2822 }, { "epoch": 2.5922865013774103, "grad_norm": 2.002721071243286, "learning_rate": 1.1037972502134554e-06, "loss": 0.0493, "step": 2823 }, { "epoch": 2.593204775022957, "grad_norm": 0.8469150066375732, "learning_rate": 1.0989238215064536e-06, "loss": 0.0266, "step": 2824 }, { "epoch": 2.594123048668503, "grad_norm": 0.2086629420518875, "learning_rate": 1.0940605494888856e-06, "loss": 0.0043, "step": 2825 }, { "epoch": 2.5950413223140494, "grad_norm": 1.3832448720932007, "learning_rate": 1.0892074397100473e-06, "loss": 0.0386, "step": 2826 }, { "epoch": 2.595959595959596, "grad_norm": 1.7735357284545898, "learning_rate": 1.0843644977076495e-06, "loss": 0.0581, "step": 2827 }, { "epoch": 2.596877869605142, "grad_norm": 2.4401235580444336, "learning_rate": 1.0795317290077912e-06, "loss": 0.1002, "step": 2828 }, { "epoch": 2.597796143250689, "grad_norm": 3.6091666221618652, "learning_rate": 1.0747091391249664e-06, "loss": 0.23, "step": 2829 }, { "epoch": 2.598714416896235, "grad_norm": 0.1981658935546875, "learning_rate": 1.0698967335620547e-06, "loss": 0.0031, "step": 2830 }, { "epoch": 2.5996326905417817, "grad_norm": 3.242581367492676, "learning_rate": 1.0650945178103123e-06, "loss": 0.381, "step": 2831 }, { "epoch": 2.600550964187328, "grad_norm": 0.31887778639793396, "learning_rate": 1.0603024973493747e-06, "loss": 0.0082, "step": 2832 }, { "epoch": 2.601469237832874, "grad_norm": 2.666067600250244, "learning_rate": 1.055520677647236e-06, "loss": 0.1277, "step": 2833 }, { "epoch": 2.6023875114784207, "grad_norm": 0.4910164475440979, "learning_rate": 1.050749064160257e-06, "loss": 0.0121, "step": 2834 }, { "epoch": 2.603305785123967, "grad_norm": 0.406756192445755, "learning_rate": 1.0459876623331467e-06, "loss": 0.008, "step": 2835 }, { "epoch": 2.604224058769513, "grad_norm": 0.044218119233846664, "learning_rate": 1.0412364775989647e-06, "loss": 0.0009, "step": 2836 }, { "epoch": 2.6051423324150598, "grad_norm": 0.10642480850219727, "learning_rate": 1.0364955153791157e-06, "loss": 0.0024, "step": 2837 }, { "epoch": 2.606060606060606, "grad_norm": 1.885723352432251, "learning_rate": 1.0317647810833354e-06, "loss": 0.0421, "step": 2838 }, { "epoch": 2.606978879706152, "grad_norm": 2.001011371612549, "learning_rate": 1.0270442801096903e-06, "loss": 0.1952, "step": 2839 }, { "epoch": 2.607897153351699, "grad_norm": 0.03794368356466293, "learning_rate": 1.0223340178445685e-06, "loss": 0.0012, "step": 2840 }, { "epoch": 2.6088154269972454, "grad_norm": 0.532000720500946, "learning_rate": 1.017633999662675e-06, "loss": 0.0088, "step": 2841 }, { "epoch": 2.6097337006427916, "grad_norm": 1.8204602003097534, "learning_rate": 1.012944230927031e-06, "loss": 0.0445, "step": 2842 }, { "epoch": 2.610651974288338, "grad_norm": 0.7875034213066101, "learning_rate": 1.008264716988957e-06, "loss": 0.0198, "step": 2843 }, { "epoch": 2.6115702479338845, "grad_norm": 1.0864038467407227, "learning_rate": 1.0035954631880728e-06, "loss": 0.021, "step": 2844 }, { "epoch": 2.6124885215794307, "grad_norm": 1.3388323783874512, "learning_rate": 9.989364748522922e-07, "loss": 0.0206, "step": 2845 }, { "epoch": 2.613406795224977, "grad_norm": 0.8711262941360474, "learning_rate": 9.942877572978127e-07, "loss": 0.024, "step": 2846 }, { "epoch": 2.6143250688705235, "grad_norm": 0.01607263833284378, "learning_rate": 9.896493158291176e-07, "loss": 0.0004, "step": 2847 }, { "epoch": 2.6152433425160697, "grad_norm": 0.11501925438642502, "learning_rate": 9.850211557389599e-07, "loss": 0.0027, "step": 2848 }, { "epoch": 2.616161616161616, "grad_norm": 0.0568469874560833, "learning_rate": 9.80403282308362e-07, "loss": 0.0015, "step": 2849 }, { "epoch": 2.6170798898071626, "grad_norm": 1.857568383216858, "learning_rate": 9.757957008066078e-07, "loss": 0.0608, "step": 2850 }, { "epoch": 2.6179981634527087, "grad_norm": 0.20476731657981873, "learning_rate": 9.711984164912402e-07, "loss": 0.0067, "step": 2851 }, { "epoch": 2.6189164370982554, "grad_norm": 0.8998540043830872, "learning_rate": 9.666114346080512e-07, "loss": 0.0277, "step": 2852 }, { "epoch": 2.6198347107438016, "grad_norm": 0.6825641989707947, "learning_rate": 9.62034760391074e-07, "loss": 0.0047, "step": 2853 }, { "epoch": 2.6207529843893482, "grad_norm": 0.06374778598546982, "learning_rate": 9.574683990625843e-07, "loss": 0.0011, "step": 2854 }, { "epoch": 2.6216712580348944, "grad_norm": 2.0198557376861572, "learning_rate": 9.529123558330866e-07, "loss": 0.1057, "step": 2855 }, { "epoch": 2.6225895316804406, "grad_norm": 0.9343361258506775, "learning_rate": 9.483666359013166e-07, "loss": 0.0219, "step": 2856 }, { "epoch": 2.6235078053259873, "grad_norm": 1.1411622762680054, "learning_rate": 9.438312444542264e-07, "loss": 0.0615, "step": 2857 }, { "epoch": 2.6244260789715335, "grad_norm": 0.16884204745292664, "learning_rate": 9.393061866669828e-07, "loss": 0.0036, "step": 2858 }, { "epoch": 2.6253443526170797, "grad_norm": 0.9165343642234802, "learning_rate": 9.347914677029624e-07, "loss": 0.0395, "step": 2859 }, { "epoch": 2.6262626262626263, "grad_norm": 1.5581132173538208, "learning_rate": 9.302870927137431e-07, "loss": 0.0258, "step": 2860 }, { "epoch": 2.6271808999081725, "grad_norm": 0.3434659540653229, "learning_rate": 9.257930668391035e-07, "loss": 0.0034, "step": 2861 }, { "epoch": 2.628099173553719, "grad_norm": 0.16584408283233643, "learning_rate": 9.213093952070107e-07, "loss": 0.0039, "step": 2862 }, { "epoch": 2.6290174471992653, "grad_norm": 0.6399905681610107, "learning_rate": 9.168360829336143e-07, "loss": 0.0051, "step": 2863 }, { "epoch": 2.629935720844812, "grad_norm": 0.15589985251426697, "learning_rate": 9.123731351232478e-07, "loss": 0.0037, "step": 2864 }, { "epoch": 2.630853994490358, "grad_norm": 0.13651470839977264, "learning_rate": 9.079205568684147e-07, "loss": 0.0026, "step": 2865 }, { "epoch": 2.6317722681359044, "grad_norm": 0.03408532217144966, "learning_rate": 9.034783532497893e-07, "loss": 0.0008, "step": 2866 }, { "epoch": 2.632690541781451, "grad_norm": 0.6617017388343811, "learning_rate": 8.99046529336206e-07, "loss": 0.0051, "step": 2867 }, { "epoch": 2.633608815426997, "grad_norm": 0.24223913252353668, "learning_rate": 8.946250901846553e-07, "loss": 0.0053, "step": 2868 }, { "epoch": 2.6345270890725434, "grad_norm": 0.6487122178077698, "learning_rate": 8.902140408402771e-07, "loss": 0.0098, "step": 2869 }, { "epoch": 2.63544536271809, "grad_norm": 0.199131578207016, "learning_rate": 8.858133863363561e-07, "loss": 0.0039, "step": 2870 }, { "epoch": 2.6363636363636362, "grad_norm": 0.957136332988739, "learning_rate": 8.8142313169432e-07, "loss": 0.0659, "step": 2871 }, { "epoch": 2.637281910009183, "grad_norm": 0.1790107786655426, "learning_rate": 8.77043281923724e-07, "loss": 0.0031, "step": 2872 }, { "epoch": 2.638200183654729, "grad_norm": 0.9351644515991211, "learning_rate": 8.726738420222547e-07, "loss": 0.0553, "step": 2873 }, { "epoch": 2.6391184573002757, "grad_norm": 1.39433753490448, "learning_rate": 8.683148169757161e-07, "loss": 0.021, "step": 2874 }, { "epoch": 2.640036730945822, "grad_norm": 0.10210525244474411, "learning_rate": 8.639662117580316e-07, "loss": 0.0019, "step": 2875 }, { "epoch": 2.640955004591368, "grad_norm": 0.3917114734649658, "learning_rate": 8.596280313312355e-07, "loss": 0.007, "step": 2876 }, { "epoch": 2.6418732782369148, "grad_norm": 0.9102626442909241, "learning_rate": 8.553002806454657e-07, "loss": 0.0273, "step": 2877 }, { "epoch": 2.642791551882461, "grad_norm": 0.7427564859390259, "learning_rate": 8.509829646389567e-07, "loss": 0.0176, "step": 2878 }, { "epoch": 2.643709825528007, "grad_norm": 3.7414515018463135, "learning_rate": 8.466760882380376e-07, "loss": 0.1141, "step": 2879 }, { "epoch": 2.644628099173554, "grad_norm": 1.70636785030365, "learning_rate": 8.423796563571296e-07, "loss": 0.0289, "step": 2880 }, { "epoch": 2.6455463728191, "grad_norm": 1.09941828250885, "learning_rate": 8.380936738987311e-07, "loss": 0.064, "step": 2881 }, { "epoch": 2.6464646464646466, "grad_norm": 2.2156598567962646, "learning_rate": 8.338181457534167e-07, "loss": 0.1052, "step": 2882 }, { "epoch": 2.647382920110193, "grad_norm": 1.224822998046875, "learning_rate": 8.295530767998361e-07, "loss": 0.0303, "step": 2883 }, { "epoch": 2.6483011937557395, "grad_norm": 0.5761744379997253, "learning_rate": 8.252984719046986e-07, "loss": 0.0109, "step": 2884 }, { "epoch": 2.6492194674012857, "grad_norm": 3.4248130321502686, "learning_rate": 8.210543359227818e-07, "loss": 0.0678, "step": 2885 }, { "epoch": 2.650137741046832, "grad_norm": 1.7152442932128906, "learning_rate": 8.168206736969086e-07, "loss": 0.0687, "step": 2886 }, { "epoch": 2.6510560146923785, "grad_norm": 2.5819900035858154, "learning_rate": 8.125974900579569e-07, "loss": 0.1083, "step": 2887 }, { "epoch": 2.6519742883379247, "grad_norm": 1.5239211320877075, "learning_rate": 8.083847898248442e-07, "loss": 0.0298, "step": 2888 }, { "epoch": 2.652892561983471, "grad_norm": 0.18692010641098022, "learning_rate": 8.041825778045265e-07, "loss": 0.003, "step": 2889 }, { "epoch": 2.6538108356290175, "grad_norm": 0.37729766964912415, "learning_rate": 7.999908587919958e-07, "loss": 0.0076, "step": 2890 }, { "epoch": 2.6547291092745637, "grad_norm": 0.6064942479133606, "learning_rate": 7.958096375702662e-07, "loss": 0.0131, "step": 2891 }, { "epoch": 2.65564738292011, "grad_norm": 0.28073135018348694, "learning_rate": 7.916389189103746e-07, "loss": 0.0064, "step": 2892 }, { "epoch": 2.6565656565656566, "grad_norm": 1.6347194910049438, "learning_rate": 7.874787075713742e-07, "loss": 0.0512, "step": 2893 }, { "epoch": 2.657483930211203, "grad_norm": 0.9469401240348816, "learning_rate": 7.833290083003276e-07, "loss": 0.013, "step": 2894 }, { "epoch": 2.6584022038567494, "grad_norm": 1.8440442085266113, "learning_rate": 7.79189825832306e-07, "loss": 0.0309, "step": 2895 }, { "epoch": 2.6593204775022956, "grad_norm": 1.0020002126693726, "learning_rate": 7.750611648903761e-07, "loss": 0.0332, "step": 2896 }, { "epoch": 2.6602387511478423, "grad_norm": 1.4172261953353882, "learning_rate": 7.709430301856014e-07, "loss": 0.0257, "step": 2897 }, { "epoch": 2.6611570247933884, "grad_norm": 0.07911404967308044, "learning_rate": 7.668354264170319e-07, "loss": 0.0015, "step": 2898 }, { "epoch": 2.6620752984389346, "grad_norm": 0.3122292757034302, "learning_rate": 7.627383582717018e-07, "loss": 0.0053, "step": 2899 }, { "epoch": 2.6629935720844813, "grad_norm": 1.116742491722107, "learning_rate": 7.586518304246282e-07, "loss": 0.0254, "step": 2900 }, { "epoch": 2.6639118457300275, "grad_norm": 0.24722304940223694, "learning_rate": 7.545758475387943e-07, "loss": 0.0069, "step": 2901 }, { "epoch": 2.6648301193755737, "grad_norm": 0.6742861866950989, "learning_rate": 7.505104142651543e-07, "loss": 0.0275, "step": 2902 }, { "epoch": 2.6657483930211203, "grad_norm": 0.8583983182907104, "learning_rate": 7.46455535242625e-07, "loss": 0.0264, "step": 2903 }, { "epoch": 2.6666666666666665, "grad_norm": 0.4618889391422272, "learning_rate": 7.424112150980756e-07, "loss": 0.0086, "step": 2904 }, { "epoch": 2.667584940312213, "grad_norm": 1.6976494789123535, "learning_rate": 7.383774584463343e-07, "loss": 0.0985, "step": 2905 }, { "epoch": 2.6685032139577594, "grad_norm": 0.17321307957172394, "learning_rate": 7.343542698901717e-07, "loss": 0.0047, "step": 2906 }, { "epoch": 2.669421487603306, "grad_norm": 0.8734964728355408, "learning_rate": 7.303416540202979e-07, "loss": 0.0337, "step": 2907 }, { "epoch": 2.670339761248852, "grad_norm": 0.825754702091217, "learning_rate": 7.263396154153624e-07, "loss": 0.0196, "step": 2908 }, { "epoch": 2.6712580348943984, "grad_norm": 0.8802013397216797, "learning_rate": 7.223481586419423e-07, "loss": 0.0237, "step": 2909 }, { "epoch": 2.672176308539945, "grad_norm": 1.0914276838302612, "learning_rate": 7.183672882545401e-07, "loss": 0.0212, "step": 2910 }, { "epoch": 2.6730945821854912, "grad_norm": 0.22917887568473816, "learning_rate": 7.143970087955843e-07, "loss": 0.0044, "step": 2911 }, { "epoch": 2.6740128558310374, "grad_norm": 0.1820296198129654, "learning_rate": 7.104373247954122e-07, "loss": 0.0042, "step": 2912 }, { "epoch": 2.674931129476584, "grad_norm": 3.3370842933654785, "learning_rate": 7.064882407722728e-07, "loss": 0.1212, "step": 2913 }, { "epoch": 2.6758494031221303, "grad_norm": 0.8496416211128235, "learning_rate": 7.025497612323206e-07, "loss": 0.027, "step": 2914 }, { "epoch": 2.676767676767677, "grad_norm": 0.17586636543273926, "learning_rate": 6.986218906696074e-07, "loss": 0.0035, "step": 2915 }, { "epoch": 2.677685950413223, "grad_norm": 0.03939807042479515, "learning_rate": 6.947046335660824e-07, "loss": 0.001, "step": 2916 }, { "epoch": 2.6786042240587697, "grad_norm": 0.38239291310310364, "learning_rate": 6.907979943915843e-07, "loss": 0.0057, "step": 2917 }, { "epoch": 2.679522497704316, "grad_norm": 1.7817931175231934, "learning_rate": 6.869019776038333e-07, "loss": 0.0343, "step": 2918 }, { "epoch": 2.680440771349862, "grad_norm": 0.09283678978681564, "learning_rate": 6.830165876484318e-07, "loss": 0.0022, "step": 2919 }, { "epoch": 2.681359044995409, "grad_norm": 0.1295529007911682, "learning_rate": 6.791418289588536e-07, "loss": 0.0027, "step": 2920 }, { "epoch": 2.682277318640955, "grad_norm": 1.1595604419708252, "learning_rate": 6.752777059564431e-07, "loss": 0.0358, "step": 2921 }, { "epoch": 2.683195592286501, "grad_norm": 0.5010924935340881, "learning_rate": 6.714242230504087e-07, "loss": 0.0091, "step": 2922 }, { "epoch": 2.684113865932048, "grad_norm": 2.733327865600586, "learning_rate": 6.675813846378154e-07, "loss": 0.1241, "step": 2923 }, { "epoch": 2.685032139577594, "grad_norm": 0.12037096917629242, "learning_rate": 6.637491951035879e-07, "loss": 0.0022, "step": 2924 }, { "epoch": 2.6859504132231407, "grad_norm": 0.24463437497615814, "learning_rate": 6.599276588204939e-07, "loss": 0.0037, "step": 2925 }, { "epoch": 2.686868686868687, "grad_norm": 0.030509361997246742, "learning_rate": 6.561167801491486e-07, "loss": 0.0007, "step": 2926 }, { "epoch": 2.6877869605142335, "grad_norm": 0.0935598760843277, "learning_rate": 6.523165634380047e-07, "loss": 0.0013, "step": 2927 }, { "epoch": 2.6887052341597797, "grad_norm": 0.08663608878850937, "learning_rate": 6.485270130233457e-07, "loss": 0.0018, "step": 2928 }, { "epoch": 2.689623507805326, "grad_norm": 0.08459111303091049, "learning_rate": 6.447481332292926e-07, "loss": 0.0019, "step": 2929 }, { "epoch": 2.6905417814508725, "grad_norm": 0.16754059493541718, "learning_rate": 6.409799283677843e-07, "loss": 0.0028, "step": 2930 }, { "epoch": 2.6914600550964187, "grad_norm": 0.7609684467315674, "learning_rate": 6.372224027385799e-07, "loss": 0.0158, "step": 2931 }, { "epoch": 2.692378328741965, "grad_norm": 1.711553692817688, "learning_rate": 6.334755606292508e-07, "loss": 0.044, "step": 2932 }, { "epoch": 2.6932966023875116, "grad_norm": 0.14905068278312683, "learning_rate": 6.29739406315184e-07, "loss": 0.0023, "step": 2933 }, { "epoch": 2.6942148760330578, "grad_norm": 1.7362515926361084, "learning_rate": 6.260139440595658e-07, "loss": 0.0186, "step": 2934 }, { "epoch": 2.695133149678604, "grad_norm": 4.0391411781311035, "learning_rate": 6.222991781133836e-07, "loss": 0.1286, "step": 2935 }, { "epoch": 2.6960514233241506, "grad_norm": 0.03592892363667488, "learning_rate": 6.1859511271542e-07, "loss": 0.0007, "step": 2936 }, { "epoch": 2.6969696969696972, "grad_norm": 0.39725461602211, "learning_rate": 6.149017520922451e-07, "loss": 0.005, "step": 2937 }, { "epoch": 2.6978879706152434, "grad_norm": 0.2982192635536194, "learning_rate": 6.112191004582191e-07, "loss": 0.0043, "step": 2938 }, { "epoch": 2.6988062442607896, "grad_norm": 0.5644283890724182, "learning_rate": 6.075471620154794e-07, "loss": 0.0163, "step": 2939 }, { "epoch": 2.6997245179063363, "grad_norm": 0.3217683434486389, "learning_rate": 6.038859409539389e-07, "loss": 0.0081, "step": 2940 }, { "epoch": 2.7006427915518825, "grad_norm": 0.23148080706596375, "learning_rate": 6.002354414512823e-07, "loss": 0.0036, "step": 2941 }, { "epoch": 2.7015610651974287, "grad_norm": 0.1715182363986969, "learning_rate": 5.965956676729568e-07, "loss": 0.0026, "step": 2942 }, { "epoch": 2.7024793388429753, "grad_norm": 0.4972032308578491, "learning_rate": 5.929666237721788e-07, "loss": 0.0095, "step": 2943 }, { "epoch": 2.7033976124885215, "grad_norm": 0.47591492533683777, "learning_rate": 5.893483138899125e-07, "loss": 0.0055, "step": 2944 }, { "epoch": 2.7043158861340677, "grad_norm": 2.038020372390747, "learning_rate": 5.8574074215488e-07, "loss": 0.0849, "step": 2945 }, { "epoch": 2.7052341597796143, "grad_norm": 0.8246241807937622, "learning_rate": 5.821439126835482e-07, "loss": 0.0183, "step": 2946 }, { "epoch": 2.7061524334251605, "grad_norm": 1.482092022895813, "learning_rate": 5.785578295801231e-07, "loss": 0.0436, "step": 2947 }, { "epoch": 2.707070707070707, "grad_norm": 0.08430731296539307, "learning_rate": 5.749824969365558e-07, "loss": 0.0017, "step": 2948 }, { "epoch": 2.7079889807162534, "grad_norm": 0.15671594440937042, "learning_rate": 5.714179188325253e-07, "loss": 0.0031, "step": 2949 }, { "epoch": 2.7089072543618, "grad_norm": 0.8623538613319397, "learning_rate": 5.678640993354389e-07, "loss": 0.0117, "step": 2950 }, { "epoch": 2.709825528007346, "grad_norm": 0.8815053105354309, "learning_rate": 5.643210425004297e-07, "loss": 0.0559, "step": 2951 }, { "epoch": 2.7107438016528924, "grad_norm": 0.09063015133142471, "learning_rate": 5.607887523703482e-07, "loss": 0.002, "step": 2952 }, { "epoch": 2.711662075298439, "grad_norm": 1.1986334323883057, "learning_rate": 5.572672329757622e-07, "loss": 0.0439, "step": 2953 }, { "epoch": 2.7125803489439853, "grad_norm": 0.408649742603302, "learning_rate": 5.537564883349455e-07, "loss": 0.0045, "step": 2954 }, { "epoch": 2.7134986225895315, "grad_norm": 2.5936026573181152, "learning_rate": 5.502565224538803e-07, "loss": 0.0814, "step": 2955 }, { "epoch": 2.714416896235078, "grad_norm": 2.5812549591064453, "learning_rate": 5.467673393262485e-07, "loss": 0.1311, "step": 2956 }, { "epoch": 2.7153351698806243, "grad_norm": 0.9354760050773621, "learning_rate": 5.432889429334265e-07, "loss": 0.0285, "step": 2957 }, { "epoch": 2.716253443526171, "grad_norm": 0.46559450030326843, "learning_rate": 5.39821337244486e-07, "loss": 0.0073, "step": 2958 }, { "epoch": 2.717171717171717, "grad_norm": 1.3008265495300293, "learning_rate": 5.363645262161843e-07, "loss": 0.0414, "step": 2959 }, { "epoch": 2.7180899908172638, "grad_norm": 1.4363903999328613, "learning_rate": 5.329185137929582e-07, "loss": 0.0688, "step": 2960 }, { "epoch": 2.71900826446281, "grad_norm": 0.3803216814994812, "learning_rate": 5.294833039069269e-07, "loss": 0.0052, "step": 2961 }, { "epoch": 2.719926538108356, "grad_norm": 1.6592466831207275, "learning_rate": 5.260589004778815e-07, "loss": 0.0833, "step": 2962 }, { "epoch": 2.720844811753903, "grad_norm": 0.09883728623390198, "learning_rate": 5.226453074132832e-07, "loss": 0.0014, "step": 2963 }, { "epoch": 2.721763085399449, "grad_norm": 1.76102876663208, "learning_rate": 5.192425286082559e-07, "loss": 0.0602, "step": 2964 }, { "epoch": 2.722681359044995, "grad_norm": 2.628971815109253, "learning_rate": 5.15850567945585e-07, "loss": 0.0739, "step": 2965 }, { "epoch": 2.723599632690542, "grad_norm": 0.2882360517978668, "learning_rate": 5.124694292957099e-07, "loss": 0.0039, "step": 2966 }, { "epoch": 2.724517906336088, "grad_norm": 1.5983134508132935, "learning_rate": 5.090991165167269e-07, "loss": 0.0356, "step": 2967 }, { "epoch": 2.7254361799816347, "grad_norm": 1.038012981414795, "learning_rate": 5.057396334543729e-07, "loss": 0.0285, "step": 2968 }, { "epoch": 2.726354453627181, "grad_norm": 0.3715822994709015, "learning_rate": 5.023909839420316e-07, "loss": 0.0054, "step": 2969 }, { "epoch": 2.7272727272727275, "grad_norm": 1.055738091468811, "learning_rate": 4.990531718007208e-07, "loss": 0.0346, "step": 2970 }, { "epoch": 2.7281910009182737, "grad_norm": 0.5632311701774597, "learning_rate": 4.957262008390962e-07, "loss": 0.0114, "step": 2971 }, { "epoch": 2.72910927456382, "grad_norm": 0.22483690083026886, "learning_rate": 4.924100748534422e-07, "loss": 0.007, "step": 2972 }, { "epoch": 2.7300275482093666, "grad_norm": 0.11211591213941574, "learning_rate": 4.891047976276674e-07, "loss": 0.0018, "step": 2973 }, { "epoch": 2.7309458218549127, "grad_norm": 1.141743779182434, "learning_rate": 4.858103729333008e-07, "loss": 0.0321, "step": 2974 }, { "epoch": 2.731864095500459, "grad_norm": 0.4293626844882965, "learning_rate": 4.825268045294895e-07, "loss": 0.005, "step": 2975 }, { "epoch": 2.7327823691460056, "grad_norm": 1.2076032161712646, "learning_rate": 4.79254096162991e-07, "loss": 0.0283, "step": 2976 }, { "epoch": 2.733700642791552, "grad_norm": 0.32515886425971985, "learning_rate": 4.7599225156817475e-07, "loss": 0.0078, "step": 2977 }, { "epoch": 2.734618916437098, "grad_norm": 1.7320493459701538, "learning_rate": 4.72741274467009e-07, "loss": 0.0248, "step": 2978 }, { "epoch": 2.7355371900826446, "grad_norm": 1.2904027700424194, "learning_rate": 4.6950116856906423e-07, "loss": 0.071, "step": 2979 }, { "epoch": 2.7364554637281913, "grad_norm": 1.2400047779083252, "learning_rate": 4.6627193757150524e-07, "loss": 0.0272, "step": 2980 }, { "epoch": 2.7373737373737375, "grad_norm": 2.9712631702423096, "learning_rate": 4.630535851590878e-07, "loss": 0.2376, "step": 2981 }, { "epoch": 2.7382920110192837, "grad_norm": 0.5598435997962952, "learning_rate": 4.598461150041578e-07, "loss": 0.0079, "step": 2982 }, { "epoch": 2.7392102846648303, "grad_norm": 0.33050432801246643, "learning_rate": 4.5664953076663855e-07, "loss": 0.005, "step": 2983 }, { "epoch": 2.7401285583103765, "grad_norm": 2.9320387840270996, "learning_rate": 4.5346383609403466e-07, "loss": 0.0662, "step": 2984 }, { "epoch": 2.7410468319559227, "grad_norm": 1.8376322984695435, "learning_rate": 4.502890346214239e-07, "loss": 0.1059, "step": 2985 }, { "epoch": 2.7419651056014693, "grad_norm": 0.9971739649772644, "learning_rate": 4.471251299714574e-07, "loss": 0.0127, "step": 2986 }, { "epoch": 2.7428833792470155, "grad_norm": 0.3624362051486969, "learning_rate": 4.4397212575434944e-07, "loss": 0.007, "step": 2987 }, { "epoch": 2.7438016528925617, "grad_norm": 1.0729924440383911, "learning_rate": 4.4083002556787434e-07, "loss": 0.0198, "step": 2988 }, { "epoch": 2.7447199265381084, "grad_norm": 1.2290228605270386, "learning_rate": 4.376988329973697e-07, "loss": 0.0216, "step": 2989 }, { "epoch": 2.7456382001836546, "grad_norm": 2.0748744010925293, "learning_rate": 4.345785516157208e-07, "loss": 0.0511, "step": 2990 }, { "epoch": 2.746556473829201, "grad_norm": 0.05007103830575943, "learning_rate": 4.3146918498336744e-07, "loss": 0.0011, "step": 2991 }, { "epoch": 2.7474747474747474, "grad_norm": 0.9901536107063293, "learning_rate": 4.2837073664829365e-07, "loss": 0.0336, "step": 2992 }, { "epoch": 2.748393021120294, "grad_norm": 0.6782660484313965, "learning_rate": 4.2528321014602356e-07, "loss": 0.0046, "step": 2993 }, { "epoch": 2.7493112947658402, "grad_norm": 0.3746846616268158, "learning_rate": 4.2220660899961905e-07, "loss": 0.0096, "step": 2994 }, { "epoch": 2.7502295684113864, "grad_norm": 1.3638575077056885, "learning_rate": 4.191409367196753e-07, "loss": 0.0235, "step": 2995 }, { "epoch": 2.751147842056933, "grad_norm": 0.5020732283592224, "learning_rate": 4.160861968043206e-07, "loss": 0.0157, "step": 2996 }, { "epoch": 2.7520661157024793, "grad_norm": 0.814428448677063, "learning_rate": 4.1304239273920464e-07, "loss": 0.013, "step": 2997 }, { "epoch": 2.7529843893480255, "grad_norm": 0.31162962317466736, "learning_rate": 4.100095279975003e-07, "loss": 0.0072, "step": 2998 }, { "epoch": 2.753902662993572, "grad_norm": 0.2786003649234772, "learning_rate": 4.0698760603989585e-07, "loss": 0.0059, "step": 2999 }, { "epoch": 2.7548209366391183, "grad_norm": 1.0214711427688599, "learning_rate": 4.0397663031459534e-07, "loss": 0.0254, "step": 3000 }, { "epoch": 2.755739210284665, "grad_norm": 0.5229830741882324, "learning_rate": 4.009766042573138e-07, "loss": 0.0104, "step": 3001 }, { "epoch": 2.756657483930211, "grad_norm": 2.5843100547790527, "learning_rate": 3.9798753129126956e-07, "loss": 0.1595, "step": 3002 }, { "epoch": 2.757575757575758, "grad_norm": 0.3334401547908783, "learning_rate": 3.9500941482718213e-07, "loss": 0.0033, "step": 3003 }, { "epoch": 2.758494031221304, "grad_norm": 1.2729233503341675, "learning_rate": 3.9204225826327213e-07, "loss": 0.0312, "step": 3004 }, { "epoch": 2.75941230486685, "grad_norm": 1.7363356351852417, "learning_rate": 3.8908606498524905e-07, "loss": 0.1268, "step": 3005 }, { "epoch": 2.760330578512397, "grad_norm": 0.1332523226737976, "learning_rate": 3.861408383663201e-07, "loss": 0.0026, "step": 3006 }, { "epoch": 2.761248852157943, "grad_norm": 0.4989207983016968, "learning_rate": 3.832065817671726e-07, "loss": 0.0052, "step": 3007 }, { "epoch": 2.762167125803489, "grad_norm": 0.23811666667461395, "learning_rate": 3.802832985359783e-07, "loss": 0.0029, "step": 3008 }, { "epoch": 2.763085399449036, "grad_norm": 0.18195079267024994, "learning_rate": 3.773709920083879e-07, "loss": 0.005, "step": 3009 }, { "epoch": 2.764003673094582, "grad_norm": 2.0789623260498047, "learning_rate": 3.744696655075275e-07, "loss": 0.0853, "step": 3010 }, { "epoch": 2.7649219467401287, "grad_norm": 0.9979847073554993, "learning_rate": 3.715793223439923e-07, "loss": 0.0076, "step": 3011 }, { "epoch": 2.765840220385675, "grad_norm": 1.945338487625122, "learning_rate": 3.6869996581584746e-07, "loss": 0.1562, "step": 3012 }, { "epoch": 2.7667584940312215, "grad_norm": 0.08289112895727158, "learning_rate": 3.658315992086203e-07, "loss": 0.0019, "step": 3013 }, { "epoch": 2.7676767676767677, "grad_norm": 1.3202627897262573, "learning_rate": 3.629742257952984e-07, "loss": 0.0154, "step": 3014 }, { "epoch": 2.768595041322314, "grad_norm": 0.35047104954719543, "learning_rate": 3.6012784883632355e-07, "loss": 0.0095, "step": 3015 }, { "epoch": 2.7695133149678606, "grad_norm": 0.15525496006011963, "learning_rate": 3.5729247157959337e-07, "loss": 0.0034, "step": 3016 }, { "epoch": 2.7704315886134068, "grad_norm": 3.3661739826202393, "learning_rate": 3.5446809726045083e-07, "loss": 0.0576, "step": 3017 }, { "epoch": 2.771349862258953, "grad_norm": 0.23415175080299377, "learning_rate": 3.516547291016836e-07, "loss": 0.0024, "step": 3018 }, { "epoch": 2.7722681359044996, "grad_norm": 1.1744816303253174, "learning_rate": 3.4885237031352604e-07, "loss": 0.0453, "step": 3019 }, { "epoch": 2.773186409550046, "grad_norm": 0.2955578565597534, "learning_rate": 3.460610240936446e-07, "loss": 0.0028, "step": 3020 }, { "epoch": 2.7741046831955924, "grad_norm": 1.9582228660583496, "learning_rate": 3.4328069362714157e-07, "loss": 0.0355, "step": 3021 }, { "epoch": 2.7750229568411386, "grad_norm": 0.3104550540447235, "learning_rate": 3.40511382086548e-07, "loss": 0.0069, "step": 3022 }, { "epoch": 2.7759412304866853, "grad_norm": 1.4517959356307983, "learning_rate": 3.3775309263182507e-07, "loss": 0.0227, "step": 3023 }, { "epoch": 2.7768595041322315, "grad_norm": 1.1276129484176636, "learning_rate": 3.350058284103519e-07, "loss": 0.0423, "step": 3024 }, { "epoch": 2.7777777777777777, "grad_norm": 0.47171011567115784, "learning_rate": 3.322695925569341e-07, "loss": 0.0091, "step": 3025 }, { "epoch": 2.7786960514233243, "grad_norm": 0.1047380343079567, "learning_rate": 3.2954438819378765e-07, "loss": 0.0014, "step": 3026 }, { "epoch": 2.7796143250688705, "grad_norm": 1.6442041397094727, "learning_rate": 3.2683021843054385e-07, "loss": 0.0433, "step": 3027 }, { "epoch": 2.7805325987144167, "grad_norm": 0.08780956268310547, "learning_rate": 3.2412708636423983e-07, "loss": 0.002, "step": 3028 }, { "epoch": 2.7814508723599634, "grad_norm": 0.2636961340904236, "learning_rate": 3.214349950793183e-07, "loss": 0.0066, "step": 3029 }, { "epoch": 2.7823691460055096, "grad_norm": 1.0960592031478882, "learning_rate": 3.187539476476298e-07, "loss": 0.0217, "step": 3030 }, { "epoch": 2.7832874196510557, "grad_norm": 2.208404302597046, "learning_rate": 3.1608394712841605e-07, "loss": 0.1092, "step": 3031 }, { "epoch": 2.7842056932966024, "grad_norm": 1.4103758335113525, "learning_rate": 3.1342499656831673e-07, "loss": 0.0719, "step": 3032 }, { "epoch": 2.785123966942149, "grad_norm": 1.2518872022628784, "learning_rate": 3.1077709900136033e-07, "loss": 0.0718, "step": 3033 }, { "epoch": 2.7860422405876952, "grad_norm": 0.039773307740688324, "learning_rate": 3.081402574489667e-07, "loss": 0.0008, "step": 3034 }, { "epoch": 2.7869605142332414, "grad_norm": 2.6027352809906006, "learning_rate": 3.055144749199379e-07, "loss": 0.0754, "step": 3035 }, { "epoch": 2.787878787878788, "grad_norm": 0.12104310840368271, "learning_rate": 3.028997544104584e-07, "loss": 0.0027, "step": 3036 }, { "epoch": 2.7887970615243343, "grad_norm": 1.0595043897628784, "learning_rate": 3.002960989040882e-07, "loss": 0.022, "step": 3037 }, { "epoch": 2.7897153351698805, "grad_norm": 0.08443102240562439, "learning_rate": 2.977035113717619e-07, "loss": 0.0015, "step": 3038 }, { "epoch": 2.790633608815427, "grad_norm": 0.09219575673341751, "learning_rate": 2.951219947717876e-07, "loss": 0.0022, "step": 3039 }, { "epoch": 2.7915518824609733, "grad_norm": 1.0649620294570923, "learning_rate": 2.9255155204983675e-07, "loss": 0.0796, "step": 3040 }, { "epoch": 2.7924701561065195, "grad_norm": 1.0873398780822754, "learning_rate": 2.8999218613894766e-07, "loss": 0.0312, "step": 3041 }, { "epoch": 2.793388429752066, "grad_norm": 0.662766695022583, "learning_rate": 2.874438999595175e-07, "loss": 0.0097, "step": 3042 }, { "epoch": 2.7943067033976123, "grad_norm": 1.8353675603866577, "learning_rate": 2.849066964193004e-07, "loss": 0.1581, "step": 3043 }, { "epoch": 2.795224977043159, "grad_norm": 1.5689135789871216, "learning_rate": 2.8238057841340814e-07, "loss": 0.0324, "step": 3044 }, { "epoch": 2.796143250688705, "grad_norm": 5.801629543304443, "learning_rate": 2.7986554882430073e-07, "loss": 0.2223, "step": 3045 }, { "epoch": 2.797061524334252, "grad_norm": 0.5001263618469238, "learning_rate": 2.773616105217836e-07, "loss": 0.0079, "step": 3046 }, { "epoch": 2.797979797979798, "grad_norm": 2.55608868598938, "learning_rate": 2.74868766363009e-07, "loss": 0.1126, "step": 3047 }, { "epoch": 2.798898071625344, "grad_norm": 0.4335533380508423, "learning_rate": 2.7238701919246934e-07, "loss": 0.0088, "step": 3048 }, { "epoch": 2.799816345270891, "grad_norm": 1.5154385566711426, "learning_rate": 2.6991637184199505e-07, "loss": 0.0087, "step": 3049 }, { "epoch": 2.800734618916437, "grad_norm": 1.4528425931930542, "learning_rate": 2.6745682713075205e-07, "loss": 0.0473, "step": 3050 }, { "epoch": 2.8016528925619832, "grad_norm": 2.39048433303833, "learning_rate": 2.650083878652332e-07, "loss": 0.0686, "step": 3051 }, { "epoch": 2.80257116620753, "grad_norm": 1.265404462814331, "learning_rate": 2.6257105683926366e-07, "loss": 0.0131, "step": 3052 }, { "epoch": 2.803489439853076, "grad_norm": 0.5127016305923462, "learning_rate": 2.6014483683399093e-07, "loss": 0.0081, "step": 3053 }, { "epoch": 2.8044077134986227, "grad_norm": 0.5520530343055725, "learning_rate": 2.577297306178872e-07, "loss": 0.0092, "step": 3054 }, { "epoch": 2.805325987144169, "grad_norm": 0.06618046760559082, "learning_rate": 2.553257409467391e-07, "loss": 0.0014, "step": 3055 }, { "epoch": 2.8062442607897156, "grad_norm": 0.5639001131057739, "learning_rate": 2.5293287056365136e-07, "loss": 0.0056, "step": 3056 }, { "epoch": 2.8071625344352618, "grad_norm": 0.04404551908373833, "learning_rate": 2.5055112219903974e-07, "loss": 0.001, "step": 3057 }, { "epoch": 2.808080808080808, "grad_norm": 1.5275850296020508, "learning_rate": 2.481804985706282e-07, "loss": 0.0303, "step": 3058 }, { "epoch": 2.8089990817263546, "grad_norm": 0.3461177945137024, "learning_rate": 2.4582100238345065e-07, "loss": 0.0056, "step": 3059 }, { "epoch": 2.809917355371901, "grad_norm": 1.5339081287384033, "learning_rate": 2.43472636329839e-07, "loss": 0.0431, "step": 3060 }, { "epoch": 2.810835629017447, "grad_norm": 0.054578959941864014, "learning_rate": 2.411354030894275e-07, "loss": 0.0013, "step": 3061 }, { "epoch": 2.8117539026629936, "grad_norm": 0.04946070536971092, "learning_rate": 2.388093053291463e-07, "loss": 0.0013, "step": 3062 }, { "epoch": 2.81267217630854, "grad_norm": 1.0762758255004883, "learning_rate": 2.3649434570321984e-07, "loss": 0.0325, "step": 3063 }, { "epoch": 2.8135904499540865, "grad_norm": 0.09016440063714981, "learning_rate": 2.3419052685316413e-07, "loss": 0.0017, "step": 3064 }, { "epoch": 2.8145087235996327, "grad_norm": 0.10233207792043686, "learning_rate": 2.3189785140778076e-07, "loss": 0.0021, "step": 3065 }, { "epoch": 2.8154269972451793, "grad_norm": 3.076625108718872, "learning_rate": 2.296163219831582e-07, "loss": 0.07, "step": 3066 }, { "epoch": 2.8163452708907255, "grad_norm": 0.23688285052776337, "learning_rate": 2.2734594118266284e-07, "loss": 0.0045, "step": 3067 }, { "epoch": 2.8172635445362717, "grad_norm": 1.152755856513977, "learning_rate": 2.2508671159694572e-07, "loss": 0.0527, "step": 3068 }, { "epoch": 2.8181818181818183, "grad_norm": 1.1403985023498535, "learning_rate": 2.2283863580392805e-07, "loss": 0.0569, "step": 3069 }, { "epoch": 2.8191000918273645, "grad_norm": 1.6253596544265747, "learning_rate": 2.2060171636880678e-07, "loss": 0.0358, "step": 3070 }, { "epoch": 2.8200183654729107, "grad_norm": 0.8324146866798401, "learning_rate": 2.1837595584404903e-07, "loss": 0.0117, "step": 3071 }, { "epoch": 2.8209366391184574, "grad_norm": 1.1190869808197021, "learning_rate": 2.1616135676938543e-07, "loss": 0.0479, "step": 3072 }, { "epoch": 2.8218549127640036, "grad_norm": 0.1217920184135437, "learning_rate": 2.1395792167181573e-07, "loss": 0.0021, "step": 3073 }, { "epoch": 2.8227731864095498, "grad_norm": 0.12983375787734985, "learning_rate": 2.117656530655976e-07, "loss": 0.0018, "step": 3074 }, { "epoch": 2.8236914600550964, "grad_norm": 2.6899046897888184, "learning_rate": 2.0958455345224783e-07, "loss": 0.1884, "step": 3075 }, { "epoch": 2.824609733700643, "grad_norm": 0.4124787151813507, "learning_rate": 2.0741462532053892e-07, "loss": 0.0049, "step": 3076 }, { "epoch": 2.8255280073461893, "grad_norm": 1.3023030757904053, "learning_rate": 2.052558711464947e-07, "loss": 0.0843, "step": 3077 }, { "epoch": 2.8264462809917354, "grad_norm": 0.14521706104278564, "learning_rate": 2.0310829339339034e-07, "loss": 0.0027, "step": 3078 }, { "epoch": 2.827364554637282, "grad_norm": 0.5954171419143677, "learning_rate": 2.0097189451174893e-07, "loss": 0.0123, "step": 3079 }, { "epoch": 2.8282828282828283, "grad_norm": 1.5349400043487549, "learning_rate": 1.988466769393349e-07, "loss": 0.0255, "step": 3080 }, { "epoch": 2.8292011019283745, "grad_norm": 1.1768803596496582, "learning_rate": 1.9673264310115513e-07, "loss": 0.0192, "step": 3081 }, { "epoch": 2.830119375573921, "grad_norm": 0.38872167468070984, "learning_rate": 1.946297954094567e-07, "loss": 0.0027, "step": 3082 }, { "epoch": 2.8310376492194673, "grad_norm": 0.9648703932762146, "learning_rate": 1.9253813626372132e-07, "loss": 0.0233, "step": 3083 }, { "epoch": 2.8319559228650135, "grad_norm": 1.0734480619430542, "learning_rate": 1.9045766805066312e-07, "loss": 0.049, "step": 3084 }, { "epoch": 2.83287419651056, "grad_norm": 0.5603837966918945, "learning_rate": 1.8838839314422986e-07, "loss": 0.014, "step": 3085 }, { "epoch": 2.8337924701561064, "grad_norm": 1.6676219701766968, "learning_rate": 1.8633031390559164e-07, "loss": 0.0321, "step": 3086 }, { "epoch": 2.834710743801653, "grad_norm": 0.5230815410614014, "learning_rate": 1.8428343268314884e-07, "loss": 0.0098, "step": 3087 }, { "epoch": 2.835629017447199, "grad_norm": 0.10117320716381073, "learning_rate": 1.8224775181252096e-07, "loss": 0.0016, "step": 3088 }, { "epoch": 2.836547291092746, "grad_norm": 4.292945861816406, "learning_rate": 1.802232736165499e-07, "loss": 0.2353, "step": 3089 }, { "epoch": 2.837465564738292, "grad_norm": 3.289274215698242, "learning_rate": 1.7821000040529114e-07, "loss": 0.1194, "step": 3090 }, { "epoch": 2.8383838383838382, "grad_norm": 0.18855465948581696, "learning_rate": 1.7620793447601702e-07, "loss": 0.0028, "step": 3091 }, { "epoch": 2.839302112029385, "grad_norm": 1.560431718826294, "learning_rate": 1.7421707811321132e-07, "loss": 0.0774, "step": 3092 }, { "epoch": 2.840220385674931, "grad_norm": 0.16909882426261902, "learning_rate": 1.7223743358856682e-07, "loss": 0.0039, "step": 3093 }, { "epoch": 2.8411386593204773, "grad_norm": 0.05686822906136513, "learning_rate": 1.7026900316098217e-07, "loss": 0.0009, "step": 3094 }, { "epoch": 2.842056932966024, "grad_norm": 0.03926891088485718, "learning_rate": 1.6831178907656175e-07, "loss": 0.0009, "step": 3095 }, { "epoch": 2.84297520661157, "grad_norm": 0.6271472573280334, "learning_rate": 1.6636579356860804e-07, "loss": 0.0107, "step": 3096 }, { "epoch": 2.8438934802571167, "grad_norm": 1.9849984645843506, "learning_rate": 1.6443101885762812e-07, "loss": 0.0642, "step": 3097 }, { "epoch": 2.844811753902663, "grad_norm": 0.4723215699195862, "learning_rate": 1.6250746715131938e-07, "loss": 0.0111, "step": 3098 }, { "epoch": 2.8457300275482096, "grad_norm": 1.0214556455612183, "learning_rate": 1.6059514064457717e-07, "loss": 0.0193, "step": 3099 }, { "epoch": 2.846648301193756, "grad_norm": 0.3185175359249115, "learning_rate": 1.5869404151948596e-07, "loss": 0.0035, "step": 3100 }, { "epoch": 2.847566574839302, "grad_norm": 1.2045021057128906, "learning_rate": 1.5680417194531838e-07, "loss": 0.0266, "step": 3101 }, { "epoch": 2.8484848484848486, "grad_norm": 0.2498399019241333, "learning_rate": 1.5492553407853716e-07, "loss": 0.0036, "step": 3102 }, { "epoch": 2.849403122130395, "grad_norm": 0.07353583723306656, "learning_rate": 1.5305813006278535e-07, "loss": 0.0019, "step": 3103 }, { "epoch": 2.850321395775941, "grad_norm": 0.10573957115411758, "learning_rate": 1.512019620288896e-07, "loss": 0.0013, "step": 3104 }, { "epoch": 2.8512396694214877, "grad_norm": 1.4359166622161865, "learning_rate": 1.493570320948534e-07, "loss": 0.0215, "step": 3105 }, { "epoch": 2.852157943067034, "grad_norm": 0.7380245327949524, "learning_rate": 1.475233423658595e-07, "loss": 0.0077, "step": 3106 }, { "epoch": 2.8530762167125805, "grad_norm": 1.3517699241638184, "learning_rate": 1.457008949342631e-07, "loss": 0.0836, "step": 3107 }, { "epoch": 2.8539944903581267, "grad_norm": 0.6036151647567749, "learning_rate": 1.438896918795907e-07, "loss": 0.0625, "step": 3108 }, { "epoch": 2.8549127640036733, "grad_norm": 3.339346408843994, "learning_rate": 1.420897352685413e-07, "loss": 0.0863, "step": 3109 }, { "epoch": 2.8558310376492195, "grad_norm": 0.11048352718353271, "learning_rate": 1.4030102715497762e-07, "loss": 0.0026, "step": 3110 }, { "epoch": 2.8567493112947657, "grad_norm": 0.5306409001350403, "learning_rate": 1.3852356957992807e-07, "loss": 0.0118, "step": 3111 }, { "epoch": 2.8576675849403124, "grad_norm": 0.2699042558670044, "learning_rate": 1.3675736457158584e-07, "loss": 0.0046, "step": 3112 }, { "epoch": 2.8585858585858586, "grad_norm": 0.9787731766700745, "learning_rate": 1.350024141452988e-07, "loss": 0.0105, "step": 3113 }, { "epoch": 2.8595041322314048, "grad_norm": 0.061481066048145294, "learning_rate": 1.3325872030357955e-07, "loss": 0.0011, "step": 3114 }, { "epoch": 2.8604224058769514, "grad_norm": 1.917507290840149, "learning_rate": 1.31526285036091e-07, "loss": 0.0863, "step": 3115 }, { "epoch": 2.8613406795224976, "grad_norm": 1.7572330236434937, "learning_rate": 1.2980511031965072e-07, "loss": 0.1114, "step": 3116 }, { "epoch": 2.8622589531680442, "grad_norm": 0.15686430037021637, "learning_rate": 1.280951981182299e-07, "loss": 0.0028, "step": 3117 }, { "epoch": 2.8631772268135904, "grad_norm": 0.9733131527900696, "learning_rate": 1.263965503829434e-07, "loss": 0.0168, "step": 3118 }, { "epoch": 2.864095500459137, "grad_norm": 0.0385260283946991, "learning_rate": 1.2470916905205633e-07, "loss": 0.0006, "step": 3119 }, { "epoch": 2.8650137741046833, "grad_norm": 0.7993068695068359, "learning_rate": 1.2303305605097958e-07, "loss": 0.0178, "step": 3120 }, { "epoch": 2.8659320477502295, "grad_norm": 0.5917356610298157, "learning_rate": 1.2136821329226e-07, "loss": 0.015, "step": 3121 }, { "epoch": 2.866850321395776, "grad_norm": 0.5922992825508118, "learning_rate": 1.197146426755913e-07, "loss": 0.0116, "step": 3122 }, { "epoch": 2.8677685950413223, "grad_norm": 0.11455473303794861, "learning_rate": 1.1807234608780083e-07, "loss": 0.0025, "step": 3123 }, { "epoch": 2.8686868686868685, "grad_norm": 0.6043168902397156, "learning_rate": 1.164413254028518e-07, "loss": 0.0108, "step": 3124 }, { "epoch": 2.869605142332415, "grad_norm": 0.3753260672092438, "learning_rate": 1.1482158248184106e-07, "loss": 0.0133, "step": 3125 }, { "epoch": 2.8705234159779613, "grad_norm": 3.5489041805267334, "learning_rate": 1.1321311917300016e-07, "loss": 0.0328, "step": 3126 }, { "epoch": 2.8714416896235075, "grad_norm": 0.15422485768795013, "learning_rate": 1.1161593731168652e-07, "loss": 0.0026, "step": 3127 }, { "epoch": 2.872359963269054, "grad_norm": 0.2785714864730835, "learning_rate": 1.1003003872038342e-07, "loss": 0.0057, "step": 3128 }, { "epoch": 2.873278236914601, "grad_norm": 0.1964033544063568, "learning_rate": 1.0845542520870223e-07, "loss": 0.0042, "step": 3129 }, { "epoch": 2.874196510560147, "grad_norm": 0.13215219974517822, "learning_rate": 1.0689209857337679e-07, "loss": 0.0024, "step": 3130 }, { "epoch": 2.875114784205693, "grad_norm": 1.4634959697723389, "learning_rate": 1.053400605982613e-07, "loss": 0.0266, "step": 3131 }, { "epoch": 2.87603305785124, "grad_norm": 2.30250883102417, "learning_rate": 1.0379931305432912e-07, "loss": 0.0352, "step": 3132 }, { "epoch": 2.876951331496786, "grad_norm": 0.4735690951347351, "learning_rate": 1.0226985769966946e-07, "loss": 0.008, "step": 3133 }, { "epoch": 2.8778696051423323, "grad_norm": 0.05447596311569214, "learning_rate": 1.0075169627948855e-07, "loss": 0.0014, "step": 3134 }, { "epoch": 2.878787878787879, "grad_norm": 1.0626271963119507, "learning_rate": 9.92448305261029e-08, "loss": 0.0315, "step": 3135 }, { "epoch": 2.879706152433425, "grad_norm": 1.567082166671753, "learning_rate": 9.774926215894376e-08, "loss": 0.0543, "step": 3136 }, { "epoch": 2.8806244260789713, "grad_norm": 2.278958320617676, "learning_rate": 9.626499288454604e-08, "loss": 0.0754, "step": 3137 }, { "epoch": 2.881542699724518, "grad_norm": 0.26005592942237854, "learning_rate": 9.479202439655722e-08, "loss": 0.0058, "step": 3138 }, { "epoch": 2.882460973370064, "grad_norm": 0.15916407108306885, "learning_rate": 9.333035837572612e-08, "loss": 0.0032, "step": 3139 }, { "epoch": 2.8833792470156108, "grad_norm": 0.7193595767021179, "learning_rate": 9.18799964899042e-08, "loss": 0.0117, "step": 3140 }, { "epoch": 2.884297520661157, "grad_norm": 0.6382821202278137, "learning_rate": 9.044094039404872e-08, "loss": 0.014, "step": 3141 }, { "epoch": 2.8852157943067036, "grad_norm": 0.5682835578918457, "learning_rate": 8.901319173021283e-08, "loss": 0.0101, "step": 3142 }, { "epoch": 2.88613406795225, "grad_norm": 1.5898805856704712, "learning_rate": 8.75967521275467e-08, "loss": 0.0587, "step": 3143 }, { "epoch": 2.887052341597796, "grad_norm": 0.3242902159690857, "learning_rate": 8.619162320229747e-08, "loss": 0.005, "step": 3144 }, { "epoch": 2.8879706152433426, "grad_norm": 0.2993893027305603, "learning_rate": 8.479780655780701e-08, "loss": 0.0056, "step": 3145 }, { "epoch": 2.888888888888889, "grad_norm": 2.2547664642333984, "learning_rate": 8.341530378450647e-08, "loss": 0.0453, "step": 3146 }, { "epoch": 2.889807162534435, "grad_norm": 0.3685355484485626, "learning_rate": 8.204411645992061e-08, "loss": 0.0028, "step": 3147 }, { "epoch": 2.8907254361799817, "grad_norm": 0.6575424075126648, "learning_rate": 8.0684246148659e-08, "loss": 0.0126, "step": 3148 }, { "epoch": 2.891643709825528, "grad_norm": 0.7583593130111694, "learning_rate": 7.933569440241928e-08, "loss": 0.0137, "step": 3149 }, { "epoch": 2.8925619834710745, "grad_norm": 0.10804170370101929, "learning_rate": 7.799846275998613e-08, "loss": 0.0025, "step": 3150 }, { "epoch": 2.8934802571166207, "grad_norm": 1.5065531730651855, "learning_rate": 7.66725527472234e-08, "loss": 0.0509, "step": 3151 }, { "epoch": 2.8943985307621674, "grad_norm": 0.5635542869567871, "learning_rate": 7.535796587707866e-08, "loss": 0.0127, "step": 3152 }, { "epoch": 2.8953168044077136, "grad_norm": 1.3557584285736084, "learning_rate": 7.405470364957868e-08, "loss": 0.0673, "step": 3153 }, { "epoch": 2.8962350780532597, "grad_norm": 0.23147182166576385, "learning_rate": 7.276276755182721e-08, "loss": 0.0029, "step": 3154 }, { "epoch": 2.8971533516988064, "grad_norm": 0.2670336961746216, "learning_rate": 7.148215905800504e-08, "loss": 0.0082, "step": 3155 }, { "epoch": 2.8980716253443526, "grad_norm": 1.1042649745941162, "learning_rate": 7.021287962936773e-08, "loss": 0.0465, "step": 3156 }, { "epoch": 2.898989898989899, "grad_norm": 0.19924212992191315, "learning_rate": 6.89549307142423e-08, "loss": 0.0026, "step": 3157 }, { "epoch": 2.8999081726354454, "grad_norm": 0.7726899981498718, "learning_rate": 6.770831374802944e-08, "loss": 0.0087, "step": 3158 }, { "epoch": 2.9008264462809916, "grad_norm": 1.5350456237792969, "learning_rate": 6.647303015319573e-08, "loss": 0.0623, "step": 3159 }, { "epoch": 2.9017447199265383, "grad_norm": 2.5826475620269775, "learning_rate": 6.524908133927921e-08, "loss": 0.0508, "step": 3160 }, { "epoch": 2.9026629935720845, "grad_norm": 0.2834948003292084, "learning_rate": 6.403646870288161e-08, "loss": 0.009, "step": 3161 }, { "epoch": 2.903581267217631, "grad_norm": 0.5433468818664551, "learning_rate": 6.283519362767166e-08, "loss": 0.0261, "step": 3162 }, { "epoch": 2.9044995408631773, "grad_norm": 0.33637240529060364, "learning_rate": 6.164525748437844e-08, "loss": 0.0061, "step": 3163 }, { "epoch": 2.9054178145087235, "grad_norm": 2.1217641830444336, "learning_rate": 6.046666163079585e-08, "loss": 0.0811, "step": 3164 }, { "epoch": 2.90633608815427, "grad_norm": 2.133303165435791, "learning_rate": 5.929940741177476e-08, "loss": 0.1138, "step": 3165 }, { "epoch": 2.9072543617998163, "grad_norm": 0.8302789926528931, "learning_rate": 5.814349615922643e-08, "loss": 0.0134, "step": 3166 }, { "epoch": 2.9081726354453625, "grad_norm": 1.9451009035110474, "learning_rate": 5.6998929192118025e-08, "loss": 0.03, "step": 3167 }, { "epoch": 2.909090909090909, "grad_norm": 0.16055302321910858, "learning_rate": 5.5865707816473714e-08, "loss": 0.003, "step": 3168 }, { "epoch": 2.9100091827364554, "grad_norm": 0.4948744773864746, "learning_rate": 5.4743833325370257e-08, "loss": 0.0084, "step": 3169 }, { "epoch": 2.9109274563820016, "grad_norm": 0.2971647381782532, "learning_rate": 5.363330699893587e-08, "loss": 0.0076, "step": 3170 }, { "epoch": 2.911845730027548, "grad_norm": 1.2985954284667969, "learning_rate": 5.25341301043536e-08, "loss": 0.0294, "step": 3171 }, { "epoch": 2.912764003673095, "grad_norm": 1.1135993003845215, "learning_rate": 5.144630389585126e-08, "loss": 0.0239, "step": 3172 }, { "epoch": 2.913682277318641, "grad_norm": 0.6765169501304626, "learning_rate": 5.036982961470816e-08, "loss": 0.0181, "step": 3173 }, { "epoch": 2.9146005509641872, "grad_norm": 0.8176794052124023, "learning_rate": 4.930470848924951e-08, "loss": 0.0158, "step": 3174 }, { "epoch": 2.915518824609734, "grad_norm": 1.5584675073623657, "learning_rate": 4.825094173484535e-08, "loss": 0.0829, "step": 3175 }, { "epoch": 2.91643709825528, "grad_norm": 1.810431957244873, "learning_rate": 4.7208530553910504e-08, "loss": 0.0509, "step": 3176 }, { "epoch": 2.9173553719008263, "grad_norm": 1.7732031345367432, "learning_rate": 4.617747613590129e-08, "loss": 0.0354, "step": 3177 }, { "epoch": 2.918273645546373, "grad_norm": 1.8618247509002686, "learning_rate": 4.5157779657316604e-08, "loss": 0.1003, "step": 3178 }, { "epoch": 2.919191919191919, "grad_norm": 0.8971810936927795, "learning_rate": 4.414944228169349e-08, "loss": 0.0181, "step": 3179 }, { "epoch": 2.9201101928374653, "grad_norm": 1.123865008354187, "learning_rate": 4.315246515960936e-08, "loss": 0.0084, "step": 3180 }, { "epoch": 2.921028466483012, "grad_norm": 1.6623347997665405, "learning_rate": 4.216684942867755e-08, "loss": 0.0973, "step": 3181 }, { "epoch": 2.921946740128558, "grad_norm": 0.47421637177467346, "learning_rate": 4.119259621354843e-08, "loss": 0.0046, "step": 3182 }, { "epoch": 2.922865013774105, "grad_norm": 1.293662428855896, "learning_rate": 4.0229706625904973e-08, "loss": 0.0229, "step": 3183 }, { "epoch": 2.923783287419651, "grad_norm": 0.2364083081483841, "learning_rate": 3.927818176446607e-08, "loss": 0.0065, "step": 3184 }, { "epoch": 2.9247015610651976, "grad_norm": 0.3507685959339142, "learning_rate": 3.8338022714980996e-08, "loss": 0.0059, "step": 3185 }, { "epoch": 2.925619834710744, "grad_norm": 0.14957334101200104, "learning_rate": 3.74092305502316e-08, "loss": 0.0038, "step": 3186 }, { "epoch": 2.92653810835629, "grad_norm": 1.5662870407104492, "learning_rate": 3.6491806330026805e-08, "loss": 0.0296, "step": 3187 }, { "epoch": 2.9274563820018367, "grad_norm": 0.5499269366264343, "learning_rate": 3.5585751101206986e-08, "loss": 0.0095, "step": 3188 }, { "epoch": 2.928374655647383, "grad_norm": 0.9087637066841125, "learning_rate": 3.4691065897637376e-08, "loss": 0.0177, "step": 3189 }, { "epoch": 2.929292929292929, "grad_norm": 1.9826748371124268, "learning_rate": 3.380775174021134e-08, "loss": 0.1436, "step": 3190 }, { "epoch": 2.9302112029384757, "grad_norm": 0.7138428688049316, "learning_rate": 3.293580963684484e-08, "loss": 0.0145, "step": 3191 }, { "epoch": 2.931129476584022, "grad_norm": 0.38268622756004333, "learning_rate": 3.207524058247868e-08, "loss": 0.0093, "step": 3192 }, { "epoch": 2.9320477502295685, "grad_norm": 1.5110130310058594, "learning_rate": 3.1226045559077376e-08, "loss": 0.0481, "step": 3193 }, { "epoch": 2.9329660238751147, "grad_norm": 0.027207711711525917, "learning_rate": 3.03882255356247e-08, "loss": 0.0007, "step": 3194 }, { "epoch": 2.9338842975206614, "grad_norm": 0.24859318137168884, "learning_rate": 2.956178146812816e-08, "loss": 0.0039, "step": 3195 }, { "epoch": 2.9348025711662076, "grad_norm": 2.73346209526062, "learning_rate": 2.8746714299611178e-08, "loss": 0.1703, "step": 3196 }, { "epoch": 2.9357208448117538, "grad_norm": 0.27133622765541077, "learning_rate": 2.794302496011758e-08, "loss": 0.0035, "step": 3197 }, { "epoch": 2.9366391184573004, "grad_norm": 0.5267176032066345, "learning_rate": 2.715071436670602e-08, "loss": 0.0077, "step": 3198 }, { "epoch": 2.9375573921028466, "grad_norm": 1.6207093000411987, "learning_rate": 2.636978342345553e-08, "loss": 0.0402, "step": 3199 }, { "epoch": 2.938475665748393, "grad_norm": 0.16108882427215576, "learning_rate": 2.5600233021455532e-08, "loss": 0.0037, "step": 3200 }, { "epoch": 2.9393939393939394, "grad_norm": 0.6857815980911255, "learning_rate": 2.4842064038811396e-08, "loss": 0.0179, "step": 3201 }, { "epoch": 2.9403122130394856, "grad_norm": 0.23014891147613525, "learning_rate": 2.4095277340642208e-08, "loss": 0.0047, "step": 3202 }, { "epoch": 2.9412304866850323, "grad_norm": 0.4055265188217163, "learning_rate": 2.3359873779079667e-08, "loss": 0.0072, "step": 3203 }, { "epoch": 2.9421487603305785, "grad_norm": 1.0405570268630981, "learning_rate": 2.263585419326475e-08, "loss": 0.053, "step": 3204 }, { "epoch": 2.943067033976125, "grad_norm": 2.178323745727539, "learning_rate": 2.1923219409349938e-08, "loss": 0.1611, "step": 3205 }, { "epoch": 2.9439853076216713, "grad_norm": 0.7953233122825623, "learning_rate": 2.122197024049588e-08, "loss": 0.0051, "step": 3206 }, { "epoch": 2.9449035812672175, "grad_norm": 0.9514625668525696, "learning_rate": 2.053210748687251e-08, "loss": 0.034, "step": 3207 }, { "epoch": 2.945821854912764, "grad_norm": 1.2754497528076172, "learning_rate": 1.9853631935656815e-08, "loss": 0.0265, "step": 3208 }, { "epoch": 2.9467401285583104, "grad_norm": 0.03885680064558983, "learning_rate": 1.9186544361031735e-08, "loss": 0.0011, "step": 3209 }, { "epoch": 2.9476584022038566, "grad_norm": 1.6792469024658203, "learning_rate": 1.853084552418616e-08, "loss": 0.053, "step": 3210 }, { "epoch": 2.948576675849403, "grad_norm": 0.5938102006912231, "learning_rate": 1.788653617331493e-08, "loss": 0.0142, "step": 3211 }, { "epoch": 2.9494949494949494, "grad_norm": 1.7057018280029297, "learning_rate": 1.72536170436155e-08, "loss": 0.0502, "step": 3212 }, { "epoch": 2.950413223140496, "grad_norm": 0.2296527922153473, "learning_rate": 1.6632088857286845e-08, "loss": 0.0043, "step": 3213 }, { "epoch": 2.9513314967860422, "grad_norm": 0.0996120274066925, "learning_rate": 1.6021952323532765e-08, "loss": 0.0018, "step": 3214 }, { "epoch": 2.952249770431589, "grad_norm": 0.3949110507965088, "learning_rate": 1.5423208138558578e-08, "loss": 0.0075, "step": 3215 }, { "epoch": 2.953168044077135, "grad_norm": 0.6914730668067932, "learning_rate": 1.4835856985568887e-08, "loss": 0.0137, "step": 3216 }, { "epoch": 2.9540863177226813, "grad_norm": 1.0207133293151855, "learning_rate": 1.4259899534766475e-08, "loss": 0.0123, "step": 3217 }, { "epoch": 2.955004591368228, "grad_norm": 0.13469620048999786, "learning_rate": 1.3695336443357854e-08, "loss": 0.0022, "step": 3218 }, { "epoch": 2.955922865013774, "grad_norm": 0.20656569302082062, "learning_rate": 1.3142168355542162e-08, "loss": 0.0026, "step": 3219 }, { "epoch": 2.9568411386593203, "grad_norm": 0.6795517206192017, "learning_rate": 1.2600395902522266e-08, "loss": 0.0179, "step": 3220 }, { "epoch": 2.957759412304867, "grad_norm": 0.7810941934585571, "learning_rate": 1.2070019702492553e-08, "loss": 0.0109, "step": 3221 }, { "epoch": 2.958677685950413, "grad_norm": 1.452011227607727, "learning_rate": 1.1551040360645582e-08, "loss": 0.0312, "step": 3222 }, { "epoch": 2.9595959595959593, "grad_norm": 2.39977765083313, "learning_rate": 1.1043458469170987e-08, "loss": 0.0791, "step": 3223 }, { "epoch": 2.960514233241506, "grad_norm": 3.314807653427124, "learning_rate": 1.0547274607251024e-08, "loss": 0.4784, "step": 3224 }, { "epoch": 2.9614325068870526, "grad_norm": 0.49781423807144165, "learning_rate": 1.0062489341062797e-08, "loss": 0.0082, "step": 3225 }, { "epoch": 2.962350780532599, "grad_norm": 2.183356761932373, "learning_rate": 9.589103223777152e-09, "loss": 0.0627, "step": 3226 }, { "epoch": 2.963269054178145, "grad_norm": 1.431169867515564, "learning_rate": 9.127116795556446e-09, "loss": 0.1082, "step": 3227 }, { "epoch": 2.9641873278236917, "grad_norm": 0.5601903200149536, "learning_rate": 8.676530583557884e-09, "loss": 0.0089, "step": 3228 }, { "epoch": 2.965105601469238, "grad_norm": 0.26850318908691406, "learning_rate": 8.237345101927973e-09, "loss": 0.0052, "step": 3229 }, { "epoch": 2.966023875114784, "grad_norm": 0.06995382159948349, "learning_rate": 7.809560851804731e-09, "loss": 0.0011, "step": 3230 }, { "epoch": 2.9669421487603307, "grad_norm": 0.4756215810775757, "learning_rate": 7.393178321316585e-09, "loss": 0.0088, "step": 3231 }, { "epoch": 2.967860422405877, "grad_norm": 0.09122544527053833, "learning_rate": 6.988197985583478e-09, "loss": 0.0015, "step": 3232 }, { "epoch": 2.968778696051423, "grad_norm": 0.15321795642375946, "learning_rate": 6.5946203067135395e-09, "loss": 0.003, "step": 3233 }, { "epoch": 2.9696969696969697, "grad_norm": 0.018853675574064255, "learning_rate": 6.212445733803085e-09, "loss": 0.0005, "step": 3234 }, { "epoch": 2.970615243342516, "grad_norm": 2.6415627002716064, "learning_rate": 5.841674702936617e-09, "loss": 0.0678, "step": 3235 }, { "epoch": 2.9715335169880626, "grad_norm": 1.5635310411453247, "learning_rate": 5.482307637189044e-09, "loss": 0.0393, "step": 3236 }, { "epoch": 2.9724517906336088, "grad_norm": 1.8492354154586792, "learning_rate": 5.134344946620129e-09, "loss": 0.0867, "step": 3237 }, { "epoch": 2.9733700642791554, "grad_norm": 0.04089601710438728, "learning_rate": 4.797787028276713e-09, "loss": 0.0011, "step": 3238 }, { "epoch": 2.9742883379247016, "grad_norm": 0.1822550892829895, "learning_rate": 4.4726342661927104e-09, "loss": 0.0037, "step": 3239 }, { "epoch": 2.975206611570248, "grad_norm": 1.0400606393814087, "learning_rate": 4.1588870313891135e-09, "loss": 0.0394, "step": 3240 }, { "epoch": 2.9761248852157944, "grad_norm": 3.1055660247802734, "learning_rate": 3.85654568186955e-09, "loss": 0.0817, "step": 3241 }, { "epoch": 2.9770431588613406, "grad_norm": 1.3689898252487183, "learning_rate": 3.565610562625832e-09, "loss": 0.0544, "step": 3242 }, { "epoch": 2.977961432506887, "grad_norm": 0.125425323843956, "learning_rate": 3.2860820056335174e-09, "loss": 0.0032, "step": 3243 }, { "epoch": 2.9788797061524335, "grad_norm": 0.07068412005901337, "learning_rate": 3.0179603298507996e-09, "loss": 0.0011, "step": 3244 }, { "epoch": 2.9797979797979797, "grad_norm": 3.6155176162719727, "learning_rate": 2.7612458412229483e-09, "loss": 0.0705, "step": 3245 }, { "epoch": 2.9807162534435263, "grad_norm": 1.338570237159729, "learning_rate": 2.5159388326767563e-09, "loss": 0.024, "step": 3246 }, { "epoch": 2.9816345270890725, "grad_norm": 0.11185647547245026, "learning_rate": 2.282039584122764e-09, "loss": 0.0017, "step": 3247 }, { "epoch": 2.982552800734619, "grad_norm": 1.5188997983932495, "learning_rate": 2.059548362455255e-09, "loss": 0.0585, "step": 3248 }, { "epoch": 2.9834710743801653, "grad_norm": 2.6831910610198975, "learning_rate": 1.8484654215500386e-09, "loss": 0.1462, "step": 3249 }, { "epoch": 2.9843893480257115, "grad_norm": 0.8003988862037659, "learning_rate": 1.6487910022666698e-09, "loss": 0.013, "step": 3250 }, { "epoch": 2.985307621671258, "grad_norm": 0.8887659907341003, "learning_rate": 1.4605253324462276e-09, "loss": 0.0167, "step": 3251 }, { "epoch": 2.9862258953168044, "grad_norm": 0.3433227241039276, "learning_rate": 1.2836686269124266e-09, "loss": 0.0051, "step": 3252 }, { "epoch": 2.9871441689623506, "grad_norm": 1.400734305381775, "learning_rate": 1.1182210874682853e-09, "loss": 0.0435, "step": 3253 }, { "epoch": 2.988062442607897, "grad_norm": 0.10417407751083374, "learning_rate": 9.641829029005677e-10, "loss": 0.0014, "step": 3254 }, { "epoch": 2.9889807162534434, "grad_norm": 1.2027952671051025, "learning_rate": 8.21554248977563e-10, "loss": 0.0384, "step": 3255 }, { "epoch": 2.98989898989899, "grad_norm": 0.3164428770542145, "learning_rate": 6.903352884457537e-10, "loss": 0.0058, "step": 3256 }, { "epoch": 2.9908172635445363, "grad_norm": 2.056800127029419, "learning_rate": 5.705261710364785e-10, "loss": 0.075, "step": 3257 }, { "epoch": 2.991735537190083, "grad_norm": 1.6685694456100464, "learning_rate": 4.6212703345704937e-10, "loss": 0.1229, "step": 3258 }, { "epoch": 2.992653810835629, "grad_norm": 0.1470831334590912, "learning_rate": 3.65137999399634e-10, "loss": 0.0027, "step": 3259 }, { "epoch": 2.9935720844811753, "grad_norm": 1.3659300804138184, "learning_rate": 2.7955917953348397e-10, "loss": 0.0478, "step": 3260 }, { "epoch": 2.994490358126722, "grad_norm": 0.08204853534698486, "learning_rate": 2.0539067151048586e-10, "loss": 0.0016, "step": 3261 }, { "epoch": 2.995408631772268, "grad_norm": 3.217071294784546, "learning_rate": 1.4263255996183056e-10, "loss": 0.1462, "step": 3262 }, { "epoch": 2.9963269054178143, "grad_norm": 0.10764630883932114, "learning_rate": 9.128491649690319e-11, "loss": 0.002, "step": 3263 }, { "epoch": 2.997245179063361, "grad_norm": 0.2585521638393402, "learning_rate": 5.134779970772385e-11, "loss": 0.0048, "step": 3264 }, { "epoch": 2.998163452708907, "grad_norm": 0.1300181746482849, "learning_rate": 2.282125516450684e-11, "loss": 0.0017, "step": 3265 }, { "epoch": 2.9990817263544534, "grad_norm": 0.19109271466732025, "learning_rate": 5.7053154189912194e-12, "loss": 0.0025, "step": 3266 }, { "epoch": 3.0, "grad_norm": 0.12415588647127151, "learning_rate": 0.0, "loss": 0.0028, "step": 3267 } ], "logging_steps": 1, "max_steps": 3267, "num_input_tokens_seen": 0, "num_train_epochs": 3, "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": 1.4232979906442035e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }