{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 816, "global_step": 4077, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0, "eval_en_loss": 3.4737327098846436, "eval_en_runtime": 20.6086, "eval_en_samples_per_second": 47.456, "eval_en_steps_per_second": 0.097, "step": 0 }, { "epoch": 0, "eval_eu_loss": 2.365967035293579, "eval_eu_runtime": 14.9602, "eval_eu_samples_per_second": 65.841, "eval_eu_steps_per_second": 0.134, "step": 0 }, { "epoch": 0.0007358351729212656, "grad_norm": 34.71304655861751, "learning_rate": 0.0, "loss": 1.9089, "step": 1 }, { "epoch": 0.0014716703458425313, "grad_norm": 31.296394721296384, "learning_rate": 1.6260162601626018e-07, "loss": 1.8343, "step": 2 }, { "epoch": 0.002207505518763797, "grad_norm": 34.246858136040096, "learning_rate": 3.2520325203252037e-07, "loss": 1.9397, "step": 3 }, { "epoch": 0.0029433406916850625, "grad_norm": 33.23492141401048, "learning_rate": 4.878048780487805e-07, "loss": 1.913, "step": 4 }, { "epoch": 0.003679175864606328, "grad_norm": 36.80131334847028, "learning_rate": 6.504065040650407e-07, "loss": 1.9991, "step": 5 }, { "epoch": 0.004415011037527594, "grad_norm": 35.474660399737786, "learning_rate": 8.130081300813009e-07, "loss": 1.9507, "step": 6 }, { "epoch": 0.0051508462104488595, "grad_norm": 30.37409066098228, "learning_rate": 9.75609756097561e-07, "loss": 1.6268, "step": 7 }, { "epoch": 0.005886681383370125, "grad_norm": 41.78410416515968, "learning_rate": 1.1382113821138213e-06, "loss": 1.9673, "step": 8 }, { "epoch": 0.006622516556291391, "grad_norm": 20.051688072158004, "learning_rate": 1.3008130081300815e-06, "loss": 1.4953, "step": 9 }, { "epoch": 0.007358351729212656, "grad_norm": 14.53442846905539, "learning_rate": 1.4634146341463414e-06, "loss": 1.3256, "step": 10 }, { "epoch": 0.008094186902133923, "grad_norm": 15.21149459313074, "learning_rate": 1.6260162601626018e-06, "loss": 1.4715, "step": 11 }, { "epoch": 0.008830022075055188, "grad_norm": 12.304740511238986, "learning_rate": 1.788617886178862e-06, "loss": 1.2688, "step": 12 }, { "epoch": 0.009565857247976454, "grad_norm": 11.720170136803956, "learning_rate": 1.951219512195122e-06, "loss": 1.2851, "step": 13 }, { "epoch": 0.010301692420897719, "grad_norm": 12.096111053477248, "learning_rate": 2.1138211382113824e-06, "loss": 1.1905, "step": 14 }, { "epoch": 0.011037527593818985, "grad_norm": 11.941453012540283, "learning_rate": 2.2764227642276426e-06, "loss": 1.2377, "step": 15 }, { "epoch": 0.01177336276674025, "grad_norm": 10.180292692591019, "learning_rate": 2.4390243902439027e-06, "loss": 1.1824, "step": 16 }, { "epoch": 0.012509197939661517, "grad_norm": 13.299499627052464, "learning_rate": 2.601626016260163e-06, "loss": 1.1573, "step": 17 }, { "epoch": 0.013245033112582781, "grad_norm": 12.876296144673749, "learning_rate": 2.764227642276423e-06, "loss": 1.1384, "step": 18 }, { "epoch": 0.013980868285504048, "grad_norm": 12.670694815167291, "learning_rate": 2.926829268292683e-06, "loss": 1.1772, "step": 19 }, { "epoch": 0.014716703458425313, "grad_norm": 11.99487919272876, "learning_rate": 3.0894308943089435e-06, "loss": 1.1585, "step": 20 }, { "epoch": 0.01545253863134658, "grad_norm": 11.082441276606877, "learning_rate": 3.2520325203252037e-06, "loss": 1.1342, "step": 21 }, { "epoch": 0.016188373804267846, "grad_norm": 9.731399673976382, "learning_rate": 3.414634146341464e-06, "loss": 1.1688, "step": 22 }, { "epoch": 0.01692420897718911, "grad_norm": 9.733553571608118, "learning_rate": 3.577235772357724e-06, "loss": 1.0978, "step": 23 }, { "epoch": 0.017660044150110375, "grad_norm": 7.860245880651305, "learning_rate": 3.7398373983739838e-06, "loss": 1.1064, "step": 24 }, { "epoch": 0.01839587932303164, "grad_norm": 6.519335779739166, "learning_rate": 3.902439024390244e-06, "loss": 1.1054, "step": 25 }, { "epoch": 0.01913171449595291, "grad_norm": 6.968138839302284, "learning_rate": 4.0650406504065046e-06, "loss": 1.0969, "step": 26 }, { "epoch": 0.019867549668874173, "grad_norm": 5.368142521145898, "learning_rate": 4.227642276422765e-06, "loss": 1.0964, "step": 27 }, { "epoch": 0.020603384841795438, "grad_norm": 4.85014734674924, "learning_rate": 4.390243902439025e-06, "loss": 1.0502, "step": 28 }, { "epoch": 0.021339220014716703, "grad_norm": 4.941674101601081, "learning_rate": 4.552845528455285e-06, "loss": 1.0064, "step": 29 }, { "epoch": 0.02207505518763797, "grad_norm": 4.887071560193437, "learning_rate": 4.715447154471545e-06, "loss": 0.9632, "step": 30 }, { "epoch": 0.022810890360559236, "grad_norm": 4.756009631531132, "learning_rate": 4.8780487804878055e-06, "loss": 0.9603, "step": 31 }, { "epoch": 0.0235467255334805, "grad_norm": 4.838318970795568, "learning_rate": 5.040650406504065e-06, "loss": 1.0255, "step": 32 }, { "epoch": 0.024282560706401765, "grad_norm": 4.613111244302159, "learning_rate": 5.203252032520326e-06, "loss": 0.9496, "step": 33 }, { "epoch": 0.025018395879323033, "grad_norm": 4.62334793117491, "learning_rate": 5.365853658536586e-06, "loss": 1.0107, "step": 34 }, { "epoch": 0.025754231052244298, "grad_norm": 4.6525191208381536, "learning_rate": 5.528455284552846e-06, "loss": 0.9742, "step": 35 }, { "epoch": 0.026490066225165563, "grad_norm": 4.624116311195727, "learning_rate": 5.691056910569106e-06, "loss": 1.0077, "step": 36 }, { "epoch": 0.027225901398086828, "grad_norm": 4.846083994926852, "learning_rate": 5.853658536585366e-06, "loss": 1.0228, "step": 37 }, { "epoch": 0.027961736571008096, "grad_norm": 4.237776234793921, "learning_rate": 6.016260162601627e-06, "loss": 1.0367, "step": 38 }, { "epoch": 0.02869757174392936, "grad_norm": 4.166019970513924, "learning_rate": 6.178861788617887e-06, "loss": 1.0125, "step": 39 }, { "epoch": 0.029433406916850625, "grad_norm": 4.125846021034458, "learning_rate": 6.341463414634147e-06, "loss": 1.0095, "step": 40 }, { "epoch": 0.03016924208977189, "grad_norm": 3.6006519348376105, "learning_rate": 6.504065040650407e-06, "loss": 1.0015, "step": 41 }, { "epoch": 0.03090507726269316, "grad_norm": 2.992475560438197, "learning_rate": 6.666666666666667e-06, "loss": 0.959, "step": 42 }, { "epoch": 0.03164091243561442, "grad_norm": 2.824510707112987, "learning_rate": 6.829268292682928e-06, "loss": 0.9768, "step": 43 }, { "epoch": 0.03237674760853569, "grad_norm": 2.788412110807702, "learning_rate": 6.991869918699188e-06, "loss": 1.0384, "step": 44 }, { "epoch": 0.033112582781456956, "grad_norm": 2.6386315828646403, "learning_rate": 7.154471544715448e-06, "loss": 0.8869, "step": 45 }, { "epoch": 0.03384841795437822, "grad_norm": 2.419912559558062, "learning_rate": 7.317073170731707e-06, "loss": 0.9247, "step": 46 }, { "epoch": 0.034584253127299486, "grad_norm": 2.700007893000522, "learning_rate": 7.4796747967479676e-06, "loss": 0.8767, "step": 47 }, { "epoch": 0.03532008830022075, "grad_norm": 2.8672805351979203, "learning_rate": 7.64227642276423e-06, "loss": 0.9582, "step": 48 }, { "epoch": 0.036055923473142015, "grad_norm": 2.777490150063748, "learning_rate": 7.804878048780489e-06, "loss": 0.9809, "step": 49 }, { "epoch": 0.03679175864606328, "grad_norm": 2.670377910200324, "learning_rate": 7.967479674796748e-06, "loss": 0.991, "step": 50 }, { "epoch": 0.037527593818984545, "grad_norm": 2.6321587086569527, "learning_rate": 8.130081300813009e-06, "loss": 0.9956, "step": 51 }, { "epoch": 0.03826342899190582, "grad_norm": 2.7023743714755817, "learning_rate": 8.292682926829268e-06, "loss": 0.976, "step": 52 }, { "epoch": 0.03899926416482708, "grad_norm": 2.6819085080973246, "learning_rate": 8.45528455284553e-06, "loss": 0.954, "step": 53 }, { "epoch": 0.039735099337748346, "grad_norm": 2.6359838664131727, "learning_rate": 8.617886178861789e-06, "loss": 0.9555, "step": 54 }, { "epoch": 0.04047093451066961, "grad_norm": 2.8357053709541895, "learning_rate": 8.78048780487805e-06, "loss": 0.928, "step": 55 }, { "epoch": 0.041206769683590876, "grad_norm": 2.804712433485194, "learning_rate": 8.94308943089431e-06, "loss": 0.9368, "step": 56 }, { "epoch": 0.04194260485651214, "grad_norm": 2.507538654770431, "learning_rate": 9.10569105691057e-06, "loss": 0.9086, "step": 57 }, { "epoch": 0.042678440029433405, "grad_norm": 2.5861628311079983, "learning_rate": 9.268292682926831e-06, "loss": 0.9735, "step": 58 }, { "epoch": 0.04341427520235467, "grad_norm": 2.600943866451353, "learning_rate": 9.43089430894309e-06, "loss": 0.9352, "step": 59 }, { "epoch": 0.04415011037527594, "grad_norm": 2.7625822885243214, "learning_rate": 9.59349593495935e-06, "loss": 0.9553, "step": 60 }, { "epoch": 0.044885945548197206, "grad_norm": 2.7131422184252028, "learning_rate": 9.756097560975611e-06, "loss": 0.9223, "step": 61 }, { "epoch": 0.04562178072111847, "grad_norm": 2.653800777541093, "learning_rate": 9.91869918699187e-06, "loss": 0.9074, "step": 62 }, { "epoch": 0.046357615894039736, "grad_norm": 2.830905941412475, "learning_rate": 1.008130081300813e-05, "loss": 1.0239, "step": 63 }, { "epoch": 0.047093451066961, "grad_norm": 2.5470544391734564, "learning_rate": 1.024390243902439e-05, "loss": 0.9946, "step": 64 }, { "epoch": 0.047829286239882265, "grad_norm": 2.459966780166392, "learning_rate": 1.0406504065040652e-05, "loss": 0.9512, "step": 65 }, { "epoch": 0.04856512141280353, "grad_norm": 2.4670878582200757, "learning_rate": 1.0569105691056911e-05, "loss": 0.933, "step": 66 }, { "epoch": 0.049300956585724795, "grad_norm": 2.63320711234447, "learning_rate": 1.0731707317073172e-05, "loss": 0.9284, "step": 67 }, { "epoch": 0.05003679175864607, "grad_norm": 2.388494046201445, "learning_rate": 1.0894308943089431e-05, "loss": 0.9627, "step": 68 }, { "epoch": 0.05077262693156733, "grad_norm": 2.515277617774793, "learning_rate": 1.1056910569105692e-05, "loss": 0.9126, "step": 69 }, { "epoch": 0.051508462104488596, "grad_norm": 2.5903125441951027, "learning_rate": 1.1219512195121953e-05, "loss": 1.0658, "step": 70 }, { "epoch": 0.05224429727740986, "grad_norm": 2.6829768471708912, "learning_rate": 1.1382113821138213e-05, "loss": 0.9961, "step": 71 }, { "epoch": 0.052980132450331126, "grad_norm": 2.738457684197144, "learning_rate": 1.1544715447154474e-05, "loss": 0.8798, "step": 72 }, { "epoch": 0.05371596762325239, "grad_norm": 2.6204377805479484, "learning_rate": 1.1707317073170731e-05, "loss": 1.0242, "step": 73 }, { "epoch": 0.054451802796173655, "grad_norm": 2.6808470499495263, "learning_rate": 1.1869918699186992e-05, "loss": 0.9598, "step": 74 }, { "epoch": 0.05518763796909492, "grad_norm": 2.44948000201168, "learning_rate": 1.2032520325203254e-05, "loss": 0.9063, "step": 75 }, { "epoch": 0.05592347314201619, "grad_norm": 2.6909536482608, "learning_rate": 1.2195121951219513e-05, "loss": 0.9971, "step": 76 }, { "epoch": 0.05665930831493746, "grad_norm": 2.5013738990817145, "learning_rate": 1.2357723577235774e-05, "loss": 1.0317, "step": 77 }, { "epoch": 0.05739514348785872, "grad_norm": 2.657400012870715, "learning_rate": 1.2520325203252033e-05, "loss": 0.9938, "step": 78 }, { "epoch": 0.058130978660779986, "grad_norm": 2.459384350049531, "learning_rate": 1.2682926829268294e-05, "loss": 0.9756, "step": 79 }, { "epoch": 0.05886681383370125, "grad_norm": 2.5898617015022802, "learning_rate": 1.2845528455284555e-05, "loss": 0.9746, "step": 80 }, { "epoch": 0.059602649006622516, "grad_norm": 2.735883711631027, "learning_rate": 1.3008130081300815e-05, "loss": 0.9495, "step": 81 }, { "epoch": 0.06033848417954378, "grad_norm": 2.624030257323286, "learning_rate": 1.3170731707317076e-05, "loss": 0.9528, "step": 82 }, { "epoch": 0.061074319352465045, "grad_norm": 2.5046786213775465, "learning_rate": 1.3333333333333333e-05, "loss": 1.0233, "step": 83 }, { "epoch": 0.06181015452538632, "grad_norm": 2.6567444349123974, "learning_rate": 1.3495934959349594e-05, "loss": 0.9424, "step": 84 }, { "epoch": 0.06254598969830757, "grad_norm": 2.6885798542225126, "learning_rate": 1.3658536585365855e-05, "loss": 1.0064, "step": 85 }, { "epoch": 0.06328182487122884, "grad_norm": 2.4277308613040445, "learning_rate": 1.3821138211382115e-05, "loss": 0.9765, "step": 86 }, { "epoch": 0.0640176600441501, "grad_norm": 2.5651197700472124, "learning_rate": 1.3983739837398376e-05, "loss": 0.9821, "step": 87 }, { "epoch": 0.06475349521707138, "grad_norm": 2.8344240790843096, "learning_rate": 1.4146341463414635e-05, "loss": 0.9816, "step": 88 }, { "epoch": 0.06548933038999265, "grad_norm": 2.6667760008803976, "learning_rate": 1.4308943089430896e-05, "loss": 0.9439, "step": 89 }, { "epoch": 0.06622516556291391, "grad_norm": 2.7684861721552108, "learning_rate": 1.4471544715447157e-05, "loss": 0.9255, "step": 90 }, { "epoch": 0.06696100073583518, "grad_norm": 2.5134957152672306, "learning_rate": 1.4634146341463415e-05, "loss": 0.9356, "step": 91 }, { "epoch": 0.06769683590875644, "grad_norm": 2.5616482429585994, "learning_rate": 1.4796747967479676e-05, "loss": 0.9987, "step": 92 }, { "epoch": 0.0684326710816777, "grad_norm": 2.723451829017181, "learning_rate": 1.4959349593495935e-05, "loss": 0.9175, "step": 93 }, { "epoch": 0.06916850625459897, "grad_norm": 2.540719741548822, "learning_rate": 1.5121951219512196e-05, "loss": 0.9902, "step": 94 }, { "epoch": 0.06990434142752024, "grad_norm": 2.4444879964382507, "learning_rate": 1.528455284552846e-05, "loss": 0.9608, "step": 95 }, { "epoch": 0.0706401766004415, "grad_norm": 2.517708267354238, "learning_rate": 1.5447154471544717e-05, "loss": 1.003, "step": 96 }, { "epoch": 0.07137601177336277, "grad_norm": 2.4593767354977976, "learning_rate": 1.5609756097560978e-05, "loss": 0.969, "step": 97 }, { "epoch": 0.07211184694628403, "grad_norm": 2.7887583549101165, "learning_rate": 1.5772357723577235e-05, "loss": 1.0222, "step": 98 }, { "epoch": 0.0728476821192053, "grad_norm": 2.487289697139183, "learning_rate": 1.5934959349593496e-05, "loss": 0.9657, "step": 99 }, { "epoch": 0.07358351729212656, "grad_norm": 2.4905341325852217, "learning_rate": 1.6097560975609757e-05, "loss": 0.9734, "step": 100 }, { "epoch": 0.07431935246504782, "grad_norm": 2.3119442388708045, "learning_rate": 1.6260162601626018e-05, "loss": 0.9415, "step": 101 }, { "epoch": 0.07505518763796909, "grad_norm": 2.6881431850439492, "learning_rate": 1.642276422764228e-05, "loss": 1.0476, "step": 102 }, { "epoch": 0.07579102281089035, "grad_norm": 2.439731149975761, "learning_rate": 1.6585365853658537e-05, "loss": 1.0226, "step": 103 }, { "epoch": 0.07652685798381163, "grad_norm": 2.514518303557535, "learning_rate": 1.6747967479674798e-05, "loss": 1.0161, "step": 104 }, { "epoch": 0.0772626931567329, "grad_norm": 2.495705867448609, "learning_rate": 1.691056910569106e-05, "loss": 0.9757, "step": 105 }, { "epoch": 0.07799852832965416, "grad_norm": 2.5488745989148196, "learning_rate": 1.7073170731707317e-05, "loss": 1.0114, "step": 106 }, { "epoch": 0.07873436350257543, "grad_norm": 2.71708862115836, "learning_rate": 1.7235772357723578e-05, "loss": 1.0063, "step": 107 }, { "epoch": 0.07947019867549669, "grad_norm": 2.527320156595191, "learning_rate": 1.739837398373984e-05, "loss": 1.0297, "step": 108 }, { "epoch": 0.08020603384841796, "grad_norm": 2.611927765498986, "learning_rate": 1.75609756097561e-05, "loss": 1.032, "step": 109 }, { "epoch": 0.08094186902133922, "grad_norm": 2.4738626195489855, "learning_rate": 1.772357723577236e-05, "loss": 1.0009, "step": 110 }, { "epoch": 0.08167770419426049, "grad_norm": 2.532610281757583, "learning_rate": 1.788617886178862e-05, "loss": 1.0026, "step": 111 }, { "epoch": 0.08241353936718175, "grad_norm": 2.8196077995080358, "learning_rate": 1.804878048780488e-05, "loss": 1.0038, "step": 112 }, { "epoch": 0.08314937454010302, "grad_norm": 2.4889208650060284, "learning_rate": 1.821138211382114e-05, "loss": 0.9804, "step": 113 }, { "epoch": 0.08388520971302428, "grad_norm": 2.5906043911097116, "learning_rate": 1.83739837398374e-05, "loss": 1.0147, "step": 114 }, { "epoch": 0.08462104488594555, "grad_norm": 2.5965834309219322, "learning_rate": 1.8536585365853663e-05, "loss": 1.0038, "step": 115 }, { "epoch": 0.08535688005886681, "grad_norm": 2.622718628681369, "learning_rate": 1.869918699186992e-05, "loss": 1.046, "step": 116 }, { "epoch": 0.08609271523178808, "grad_norm": 2.558523465463052, "learning_rate": 1.886178861788618e-05, "loss": 0.9607, "step": 117 }, { "epoch": 0.08682855040470934, "grad_norm": 2.4735203422890804, "learning_rate": 1.902439024390244e-05, "loss": 1.0501, "step": 118 }, { "epoch": 0.0875643855776306, "grad_norm": 2.7638009907187957, "learning_rate": 1.91869918699187e-05, "loss": 1.0087, "step": 119 }, { "epoch": 0.08830022075055188, "grad_norm": 2.4674539283203076, "learning_rate": 1.934959349593496e-05, "loss": 0.9416, "step": 120 }, { "epoch": 0.08903605592347315, "grad_norm": 2.2846879552979154, "learning_rate": 1.9512195121951222e-05, "loss": 0.9157, "step": 121 }, { "epoch": 0.08977189109639441, "grad_norm": 2.3489286905446316, "learning_rate": 1.9674796747967483e-05, "loss": 0.9417, "step": 122 }, { "epoch": 0.09050772626931568, "grad_norm": 2.4774943821507263, "learning_rate": 1.983739837398374e-05, "loss": 1.0397, "step": 123 }, { "epoch": 0.09124356144223694, "grad_norm": 2.353350985388239, "learning_rate": 2e-05, "loss": 1.051, "step": 124 }, { "epoch": 0.09197939661515821, "grad_norm": 2.4291411275103, "learning_rate": 1.99999968435683e-05, "loss": 0.9729, "step": 125 }, { "epoch": 0.09271523178807947, "grad_norm": 2.148294551221167, "learning_rate": 1.9999987374275175e-05, "loss": 0.9068, "step": 126 }, { "epoch": 0.09345106696100074, "grad_norm": 2.4062786341340656, "learning_rate": 1.999997159212662e-05, "loss": 0.9409, "step": 127 }, { "epoch": 0.094186902133922, "grad_norm": 2.4234693749612415, "learning_rate": 1.9999949497132587e-05, "loss": 0.9274, "step": 128 }, { "epoch": 0.09492273730684327, "grad_norm": 2.5481893735183885, "learning_rate": 1.9999921089307026e-05, "loss": 0.9634, "step": 129 }, { "epoch": 0.09565857247976453, "grad_norm": 2.212567472234629, "learning_rate": 1.9999886368667875e-05, "loss": 0.9868, "step": 130 }, { "epoch": 0.0963944076526858, "grad_norm": 2.4126380032214327, "learning_rate": 1.999984533523705e-05, "loss": 0.9986, "step": 131 }, { "epoch": 0.09713024282560706, "grad_norm": 2.614237053409753, "learning_rate": 1.999979798904045e-05, "loss": 0.9805, "step": 132 }, { "epoch": 0.09786607799852833, "grad_norm": 2.3701208326378542, "learning_rate": 1.9999744330107972e-05, "loss": 1.0557, "step": 133 }, { "epoch": 0.09860191317144959, "grad_norm": 2.509839243244208, "learning_rate": 1.9999684358473485e-05, "loss": 0.9617, "step": 134 }, { "epoch": 0.09933774834437085, "grad_norm": 2.4608106052556677, "learning_rate": 1.9999618074174848e-05, "loss": 1.0271, "step": 135 }, { "epoch": 0.10007358351729213, "grad_norm": 2.3397374446469295, "learning_rate": 1.999954547725391e-05, "loss": 1.0299, "step": 136 }, { "epoch": 0.1008094186902134, "grad_norm": 2.3906292123453023, "learning_rate": 1.99994665677565e-05, "loss": 1.0032, "step": 137 }, { "epoch": 0.10154525386313466, "grad_norm": 2.372952953901993, "learning_rate": 1.9999381345732423e-05, "loss": 1.0233, "step": 138 }, { "epoch": 0.10228108903605593, "grad_norm": 2.457101621521935, "learning_rate": 1.9999289811235492e-05, "loss": 0.982, "step": 139 }, { "epoch": 0.10301692420897719, "grad_norm": 2.4820699243489077, "learning_rate": 1.9999191964323482e-05, "loss": 0.9499, "step": 140 }, { "epoch": 0.10375275938189846, "grad_norm": 2.386545541539904, "learning_rate": 1.9999087805058166e-05, "loss": 1.0004, "step": 141 }, { "epoch": 0.10448859455481972, "grad_norm": 2.2852550763592823, "learning_rate": 1.9998977333505298e-05, "loss": 0.9576, "step": 142 }, { "epoch": 0.10522442972774099, "grad_norm": 2.30716765583047, "learning_rate": 1.9998860549734618e-05, "loss": 0.9148, "step": 143 }, { "epoch": 0.10596026490066225, "grad_norm": 2.1760870684576665, "learning_rate": 1.9998737453819846e-05, "loss": 0.9574, "step": 144 }, { "epoch": 0.10669610007358352, "grad_norm": 2.134098389309033, "learning_rate": 1.9998608045838696e-05, "loss": 0.9739, "step": 145 }, { "epoch": 0.10743193524650478, "grad_norm": 2.3354752545241904, "learning_rate": 1.999847232587286e-05, "loss": 1.0072, "step": 146 }, { "epoch": 0.10816777041942605, "grad_norm": 2.3467868791976443, "learning_rate": 1.9998330294008016e-05, "loss": 1.0208, "step": 147 }, { "epoch": 0.10890360559234731, "grad_norm": 2.314643027702829, "learning_rate": 1.9998181950333825e-05, "loss": 0.9937, "step": 148 }, { "epoch": 0.10963944076526858, "grad_norm": 2.357817144369028, "learning_rate": 1.9998027294943934e-05, "loss": 0.9741, "step": 149 }, { "epoch": 0.11037527593818984, "grad_norm": 2.4433902186018313, "learning_rate": 1.999786632793598e-05, "loss": 1.0152, "step": 150 }, { "epoch": 0.1111111111111111, "grad_norm": 2.363422267834157, "learning_rate": 1.999769904941157e-05, "loss": 1.0448, "step": 151 }, { "epoch": 0.11184694628403238, "grad_norm": 2.4982958457756057, "learning_rate": 1.9997525459476313e-05, "loss": 1.0068, "step": 152 }, { "epoch": 0.11258278145695365, "grad_norm": 2.2925079318647805, "learning_rate": 1.9997345558239794e-05, "loss": 0.9858, "step": 153 }, { "epoch": 0.11331861662987491, "grad_norm": 2.2068096709501765, "learning_rate": 1.9997159345815577e-05, "loss": 0.9355, "step": 154 }, { "epoch": 0.11405445180279618, "grad_norm": 2.2125125959728957, "learning_rate": 1.9996966822321217e-05, "loss": 0.9633, "step": 155 }, { "epoch": 0.11479028697571744, "grad_norm": 2.143278756681051, "learning_rate": 1.9996767987878254e-05, "loss": 1.0215, "step": 156 }, { "epoch": 0.11552612214863871, "grad_norm": 2.126303450400665, "learning_rate": 1.9996562842612208e-05, "loss": 0.9888, "step": 157 }, { "epoch": 0.11626195732155997, "grad_norm": 2.2577264451313312, "learning_rate": 1.9996351386652583e-05, "loss": 0.9942, "step": 158 }, { "epoch": 0.11699779249448124, "grad_norm": 2.3521902598593325, "learning_rate": 1.999613362013287e-05, "loss": 0.9637, "step": 159 }, { "epoch": 0.1177336276674025, "grad_norm": 2.2187607276777164, "learning_rate": 1.999590954319054e-05, "loss": 0.9406, "step": 160 }, { "epoch": 0.11846946284032377, "grad_norm": 2.351517035093497, "learning_rate": 1.999567915596705e-05, "loss": 0.9795, "step": 161 }, { "epoch": 0.11920529801324503, "grad_norm": 2.4575754060545543, "learning_rate": 1.9995442458607846e-05, "loss": 0.9882, "step": 162 }, { "epoch": 0.1199411331861663, "grad_norm": 2.516389522094583, "learning_rate": 1.9995199451262348e-05, "loss": 0.9763, "step": 163 }, { "epoch": 0.12067696835908756, "grad_norm": 2.543380326040488, "learning_rate": 1.9994950134083954e-05, "loss": 0.9614, "step": 164 }, { "epoch": 0.12141280353200883, "grad_norm": 2.366129126486676, "learning_rate": 1.9994694507230072e-05, "loss": 0.9857, "step": 165 }, { "epoch": 0.12214863870493009, "grad_norm": 2.3954913369482234, "learning_rate": 1.999443257086206e-05, "loss": 0.9657, "step": 166 }, { "epoch": 0.12288447387785136, "grad_norm": 2.169137930273144, "learning_rate": 1.9994164325145283e-05, "loss": 0.9658, "step": 167 }, { "epoch": 0.12362030905077263, "grad_norm": 2.2993339852339165, "learning_rate": 1.9993889770249084e-05, "loss": 0.9801, "step": 168 }, { "epoch": 0.1243561442236939, "grad_norm": 2.27513539212771, "learning_rate": 1.9993608906346778e-05, "loss": 0.9986, "step": 169 }, { "epoch": 0.12509197939661515, "grad_norm": 2.4452951640509597, "learning_rate": 1.9993321733615674e-05, "loss": 1.0152, "step": 170 }, { "epoch": 0.12582781456953643, "grad_norm": 2.280299586978596, "learning_rate": 1.999302825223706e-05, "loss": 0.9701, "step": 171 }, { "epoch": 0.12656364974245768, "grad_norm": 2.2303350331339127, "learning_rate": 1.9992728462396207e-05, "loss": 0.9564, "step": 172 }, { "epoch": 0.12729948491537896, "grad_norm": 2.4274523907505516, "learning_rate": 1.9992422364282367e-05, "loss": 0.9492, "step": 173 }, { "epoch": 0.1280353200883002, "grad_norm": 2.319689377457269, "learning_rate": 1.9992109958088778e-05, "loss": 0.953, "step": 174 }, { "epoch": 0.1287711552612215, "grad_norm": 2.4703516265167638, "learning_rate": 1.9991791244012654e-05, "loss": 0.9192, "step": 175 }, { "epoch": 0.12950699043414277, "grad_norm": 2.4547683878822766, "learning_rate": 1.99914662222552e-05, "loss": 0.9523, "step": 176 }, { "epoch": 0.13024282560706402, "grad_norm": 2.157556676776675, "learning_rate": 1.9991134893021594e-05, "loss": 0.9672, "step": 177 }, { "epoch": 0.1309786607799853, "grad_norm": 2.1124489343675177, "learning_rate": 1.9990797256521e-05, "loss": 0.926, "step": 178 }, { "epoch": 0.13171449595290655, "grad_norm": 2.5430458881169877, "learning_rate": 1.9990453312966564e-05, "loss": 0.9337, "step": 179 }, { "epoch": 0.13245033112582782, "grad_norm": 2.2161776798457073, "learning_rate": 1.9990103062575415e-05, "loss": 0.9705, "step": 180 }, { "epoch": 0.13318616629874908, "grad_norm": 2.3859845750080813, "learning_rate": 1.9989746505568655e-05, "loss": 1.0498, "step": 181 }, { "epoch": 0.13392200147167035, "grad_norm": 2.54373525766505, "learning_rate": 1.9989383642171377e-05, "loss": 0.9731, "step": 182 }, { "epoch": 0.1346578366445916, "grad_norm": 2.143811029425878, "learning_rate": 1.9989014472612654e-05, "loss": 0.9483, "step": 183 }, { "epoch": 0.13539367181751288, "grad_norm": 2.2091137929129214, "learning_rate": 1.998863899712554e-05, "loss": 0.951, "step": 184 }, { "epoch": 0.13612950699043413, "grad_norm": 2.054877489936307, "learning_rate": 1.998825721594706e-05, "loss": 0.9411, "step": 185 }, { "epoch": 0.1368653421633554, "grad_norm": 2.3161293286292466, "learning_rate": 1.9987869129318232e-05, "loss": 1.0487, "step": 186 }, { "epoch": 0.13760117733627666, "grad_norm": 2.1109711377061857, "learning_rate": 1.998747473748405e-05, "loss": 0.986, "step": 187 }, { "epoch": 0.13833701250919794, "grad_norm": 2.2805794945772058, "learning_rate": 1.998707404069348e-05, "loss": 1.047, "step": 188 }, { "epoch": 0.1390728476821192, "grad_norm": 2.2165436461419854, "learning_rate": 1.9986667039199488e-05, "loss": 0.9002, "step": 189 }, { "epoch": 0.13980868285504047, "grad_norm": 2.3175729123375226, "learning_rate": 1.9986253733259004e-05, "loss": 1.0613, "step": 190 }, { "epoch": 0.14054451802796172, "grad_norm": 2.0975134027461446, "learning_rate": 1.9985834123132943e-05, "loss": 0.9873, "step": 191 }, { "epoch": 0.141280353200883, "grad_norm": 2.2109551478976366, "learning_rate": 1.9985408209086195e-05, "loss": 0.99, "step": 192 }, { "epoch": 0.14201618837380428, "grad_norm": 2.0645857247478983, "learning_rate": 1.998497599138764e-05, "loss": 0.9451, "step": 193 }, { "epoch": 0.14275202354672553, "grad_norm": 2.075277982246889, "learning_rate": 1.9984537470310127e-05, "loss": 0.9408, "step": 194 }, { "epoch": 0.1434878587196468, "grad_norm": 2.1879824255212883, "learning_rate": 1.9984092646130487e-05, "loss": 0.9512, "step": 195 }, { "epoch": 0.14422369389256806, "grad_norm": 2.185448713155068, "learning_rate": 1.9983641519129534e-05, "loss": 0.9645, "step": 196 }, { "epoch": 0.14495952906548934, "grad_norm": 2.4584233998002403, "learning_rate": 1.9983184089592062e-05, "loss": 0.9907, "step": 197 }, { "epoch": 0.1456953642384106, "grad_norm": 2.2904388279194485, "learning_rate": 1.9982720357806835e-05, "loss": 0.9278, "step": 198 }, { "epoch": 0.14643119941133187, "grad_norm": 2.2794667465379916, "learning_rate": 1.99822503240666e-05, "loss": 0.991, "step": 199 }, { "epoch": 0.14716703458425312, "grad_norm": 2.3638012102890755, "learning_rate": 1.9981773988668086e-05, "loss": 1.0589, "step": 200 }, { "epoch": 0.1479028697571744, "grad_norm": 2.1272577168734172, "learning_rate": 1.9981291351911995e-05, "loss": 0.9665, "step": 201 }, { "epoch": 0.14863870493009565, "grad_norm": 2.117827169058556, "learning_rate": 1.998080241410301e-05, "loss": 1.0189, "step": 202 }, { "epoch": 0.14937454010301693, "grad_norm": 2.015319902439755, "learning_rate": 1.998030717554979e-05, "loss": 0.96, "step": 203 }, { "epoch": 0.15011037527593818, "grad_norm": 2.4546846719895883, "learning_rate": 1.9979805636564974e-05, "loss": 1.0235, "step": 204 }, { "epoch": 0.15084621044885946, "grad_norm": 2.277124725913427, "learning_rate": 1.997929779746517e-05, "loss": 0.9624, "step": 205 }, { "epoch": 0.1515820456217807, "grad_norm": 2.279602285289561, "learning_rate": 1.9978783658570982e-05, "loss": 1.0016, "step": 206 }, { "epoch": 0.152317880794702, "grad_norm": 2.385758888257463, "learning_rate": 1.9978263220206967e-05, "loss": 0.945, "step": 207 }, { "epoch": 0.15305371596762327, "grad_norm": 2.0872220975236333, "learning_rate": 1.997773648270168e-05, "loss": 0.9747, "step": 208 }, { "epoch": 0.15378955114054452, "grad_norm": 2.1658009347634124, "learning_rate": 1.9977203446387638e-05, "loss": 0.9703, "step": 209 }, { "epoch": 0.1545253863134658, "grad_norm": 2.192934826014286, "learning_rate": 1.9976664111601337e-05, "loss": 0.9589, "step": 210 }, { "epoch": 0.15526122148638705, "grad_norm": 2.089134795552111, "learning_rate": 1.997611847868326e-05, "loss": 0.9311, "step": 211 }, { "epoch": 0.15599705665930833, "grad_norm": 2.248723518420459, "learning_rate": 1.997556654797785e-05, "loss": 0.9868, "step": 212 }, { "epoch": 0.15673289183222958, "grad_norm": 2.033763727854529, "learning_rate": 1.997500831983354e-05, "loss": 0.9625, "step": 213 }, { "epoch": 0.15746872700515085, "grad_norm": 2.0811451195277626, "learning_rate": 1.9974443794602723e-05, "loss": 0.9783, "step": 214 }, { "epoch": 0.1582045621780721, "grad_norm": 2.065627017799729, "learning_rate": 1.9973872972641783e-05, "loss": 0.9886, "step": 215 }, { "epoch": 0.15894039735099338, "grad_norm": 2.0376011279600035, "learning_rate": 1.997329585431107e-05, "loss": 0.9855, "step": 216 }, { "epoch": 0.15967623252391464, "grad_norm": 2.144343563462098, "learning_rate": 1.9972712439974912e-05, "loss": 0.9592, "step": 217 }, { "epoch": 0.1604120676968359, "grad_norm": 2.106842497691084, "learning_rate": 1.997212273000161e-05, "loss": 0.9735, "step": 218 }, { "epoch": 0.16114790286975716, "grad_norm": 2.1360979206432527, "learning_rate": 1.9971526724763438e-05, "loss": 0.9382, "step": 219 }, { "epoch": 0.16188373804267844, "grad_norm": 2.123321369845365, "learning_rate": 1.9970924424636645e-05, "loss": 1.0065, "step": 220 }, { "epoch": 0.1626195732155997, "grad_norm": 2.333703753668228, "learning_rate": 1.997031583000146e-05, "loss": 0.9777, "step": 221 }, { "epoch": 0.16335540838852097, "grad_norm": 2.068641957325135, "learning_rate": 1.996970094124208e-05, "loss": 0.9937, "step": 222 }, { "epoch": 0.16409124356144222, "grad_norm": 2.0196415265789724, "learning_rate": 1.996907975874667e-05, "loss": 0.9091, "step": 223 }, { "epoch": 0.1648270787343635, "grad_norm": 2.256838912853235, "learning_rate": 1.9968452282907377e-05, "loss": 0.9695, "step": 224 }, { "epoch": 0.16556291390728478, "grad_norm": 2.0713391814170103, "learning_rate": 1.996781851412032e-05, "loss": 0.9947, "step": 225 }, { "epoch": 0.16629874908020603, "grad_norm": 2.1225425720408757, "learning_rate": 1.9967178452785586e-05, "loss": 0.9734, "step": 226 }, { "epoch": 0.1670345842531273, "grad_norm": 2.2715384964019645, "learning_rate": 1.996653209930724e-05, "loss": 1.0133, "step": 227 }, { "epoch": 0.16777041942604856, "grad_norm": 2.272894771328333, "learning_rate": 1.9965879454093313e-05, "loss": 0.9902, "step": 228 }, { "epoch": 0.16850625459896984, "grad_norm": 2.2058529378861507, "learning_rate": 1.9965220517555814e-05, "loss": 0.9972, "step": 229 }, { "epoch": 0.1692420897718911, "grad_norm": 2.1672409499091847, "learning_rate": 1.9964555290110718e-05, "loss": 0.9314, "step": 230 }, { "epoch": 0.16997792494481237, "grad_norm": 2.3272620597467792, "learning_rate": 1.9963883772177973e-05, "loss": 1.0055, "step": 231 }, { "epoch": 0.17071376011773362, "grad_norm": 2.0496046058149116, "learning_rate": 1.9963205964181503e-05, "loss": 1.0462, "step": 232 }, { "epoch": 0.1714495952906549, "grad_norm": 2.1080311340115343, "learning_rate": 1.9962521866549194e-05, "loss": 0.9409, "step": 233 }, { "epoch": 0.17218543046357615, "grad_norm": 2.036982711009363, "learning_rate": 1.9961831479712916e-05, "loss": 0.9178, "step": 234 }, { "epoch": 0.17292126563649743, "grad_norm": 2.0922473775937807, "learning_rate": 1.996113480410849e-05, "loss": 1.0708, "step": 235 }, { "epoch": 0.17365710080941868, "grad_norm": 2.233410252176854, "learning_rate": 1.9960431840175722e-05, "loss": 1.0631, "step": 236 }, { "epoch": 0.17439293598233996, "grad_norm": 2.262266891411482, "learning_rate": 1.9959722588358387e-05, "loss": 0.9738, "step": 237 }, { "epoch": 0.1751287711552612, "grad_norm": 1.9301828692225271, "learning_rate": 1.9959007049104223e-05, "loss": 0.9261, "step": 238 }, { "epoch": 0.1758646063281825, "grad_norm": 2.0984663297838657, "learning_rate": 1.995828522286494e-05, "loss": 0.957, "step": 239 }, { "epoch": 0.17660044150110377, "grad_norm": 2.1108739555956895, "learning_rate": 1.995755711009622e-05, "loss": 0.9663, "step": 240 }, { "epoch": 0.17733627667402502, "grad_norm": 2.010676070954677, "learning_rate": 1.9956822711257708e-05, "loss": 0.9654, "step": 241 }, { "epoch": 0.1780721118469463, "grad_norm": 2.1904393857947952, "learning_rate": 1.995608202681302e-05, "loss": 0.9641, "step": 242 }, { "epoch": 0.17880794701986755, "grad_norm": 2.315086441989138, "learning_rate": 1.9955335057229738e-05, "loss": 1.0357, "step": 243 }, { "epoch": 0.17954378219278883, "grad_norm": 2.216672958849171, "learning_rate": 1.995458180297942e-05, "loss": 1.0359, "step": 244 }, { "epoch": 0.18027961736571008, "grad_norm": 2.0871996917752296, "learning_rate": 1.9953822264537573e-05, "loss": 0.9942, "step": 245 }, { "epoch": 0.18101545253863136, "grad_norm": 2.2827697026438427, "learning_rate": 1.99530564423837e-05, "loss": 0.951, "step": 246 }, { "epoch": 0.1817512877115526, "grad_norm": 2.259375644740999, "learning_rate": 1.9952284337001238e-05, "loss": 0.9964, "step": 247 }, { "epoch": 0.18248712288447388, "grad_norm": 2.090309259454565, "learning_rate": 1.995150594887762e-05, "loss": 0.9552, "step": 248 }, { "epoch": 0.18322295805739514, "grad_norm": 2.0956195864958436, "learning_rate": 1.9950721278504222e-05, "loss": 0.9796, "step": 249 }, { "epoch": 0.18395879323031641, "grad_norm": 2.2124002729988352, "learning_rate": 1.9949930326376403e-05, "loss": 1.0308, "step": 250 }, { "epoch": 0.18469462840323767, "grad_norm": 2.0689609292525257, "learning_rate": 1.9949133092993473e-05, "loss": 0.9069, "step": 251 }, { "epoch": 0.18543046357615894, "grad_norm": 2.1891621784606508, "learning_rate": 1.994832957885872e-05, "loss": 1.0065, "step": 252 }, { "epoch": 0.1861662987490802, "grad_norm": 2.115262669486497, "learning_rate": 1.994751978447939e-05, "loss": 0.9928, "step": 253 }, { "epoch": 0.18690213392200147, "grad_norm": 2.1203530156277526, "learning_rate": 1.9946703710366693e-05, "loss": 0.9285, "step": 254 }, { "epoch": 0.18763796909492272, "grad_norm": 2.2340023166739114, "learning_rate": 1.9945881357035808e-05, "loss": 1.0008, "step": 255 }, { "epoch": 0.188373804267844, "grad_norm": 2.0843381349806482, "learning_rate": 1.994505272500588e-05, "loss": 0.9891, "step": 256 }, { "epoch": 0.18910963944076528, "grad_norm": 1.9376532698148874, "learning_rate": 1.99442178148e-05, "loss": 0.9815, "step": 257 }, { "epoch": 0.18984547461368653, "grad_norm": 2.6146044819771324, "learning_rate": 1.9943376626945246e-05, "loss": 0.9565, "step": 258 }, { "epoch": 0.1905813097866078, "grad_norm": 2.277655724302375, "learning_rate": 1.9942529161972646e-05, "loss": 1.0076, "step": 259 }, { "epoch": 0.19131714495952906, "grad_norm": 1.974172014534775, "learning_rate": 1.994167542041719e-05, "loss": 0.9268, "step": 260 }, { "epoch": 0.19205298013245034, "grad_norm": 2.0840891072936247, "learning_rate": 1.994081540281784e-05, "loss": 0.9588, "step": 261 }, { "epoch": 0.1927888153053716, "grad_norm": 2.154328885959728, "learning_rate": 1.993994910971751e-05, "loss": 0.9818, "step": 262 }, { "epoch": 0.19352465047829287, "grad_norm": 2.1265382248737787, "learning_rate": 1.9939076541663077e-05, "loss": 1.0062, "step": 263 }, { "epoch": 0.19426048565121412, "grad_norm": 2.145988206864307, "learning_rate": 1.993819769920538e-05, "loss": 1.0185, "step": 264 }, { "epoch": 0.1949963208241354, "grad_norm": 2.1709193710832997, "learning_rate": 1.9937312582899224e-05, "loss": 0.9701, "step": 265 }, { "epoch": 0.19573215599705665, "grad_norm": 2.265515658421202, "learning_rate": 1.9936421193303373e-05, "loss": 0.9647, "step": 266 }, { "epoch": 0.19646799116997793, "grad_norm": 2.162441924194623, "learning_rate": 1.9935523530980542e-05, "loss": 0.952, "step": 267 }, { "epoch": 0.19720382634289918, "grad_norm": 2.0634503706883884, "learning_rate": 1.993461959649742e-05, "loss": 0.9244, "step": 268 }, { "epoch": 0.19793966151582046, "grad_norm": 2.205651043740088, "learning_rate": 1.993370939042464e-05, "loss": 0.9567, "step": 269 }, { "epoch": 0.1986754966887417, "grad_norm": 2.2763958710438477, "learning_rate": 1.9932792913336814e-05, "loss": 0.9885, "step": 270 }, { "epoch": 0.199411331861663, "grad_norm": 2.199247321581958, "learning_rate": 1.9931870165812492e-05, "loss": 0.9004, "step": 271 }, { "epoch": 0.20014716703458427, "grad_norm": 2.067748333070549, "learning_rate": 1.9930941148434197e-05, "loss": 0.9867, "step": 272 }, { "epoch": 0.20088300220750552, "grad_norm": 2.2291674017158702, "learning_rate": 1.99300058617884e-05, "loss": 1.0434, "step": 273 }, { "epoch": 0.2016188373804268, "grad_norm": 2.1713454748819885, "learning_rate": 1.9929064306465543e-05, "loss": 1.0298, "step": 274 }, { "epoch": 0.20235467255334805, "grad_norm": 2.079943470854805, "learning_rate": 1.992811648306001e-05, "loss": 0.9812, "step": 275 }, { "epoch": 0.20309050772626933, "grad_norm": 2.156390368017153, "learning_rate": 1.992716239217015e-05, "loss": 0.9937, "step": 276 }, { "epoch": 0.20382634289919058, "grad_norm": 2.005481724511299, "learning_rate": 1.992620203439827e-05, "loss": 0.9936, "step": 277 }, { "epoch": 0.20456217807211186, "grad_norm": 2.114580264046816, "learning_rate": 1.992523541035063e-05, "loss": 0.9895, "step": 278 }, { "epoch": 0.2052980132450331, "grad_norm": 2.0314066122732592, "learning_rate": 1.9924262520637442e-05, "loss": 0.9765, "step": 279 }, { "epoch": 0.20603384841795438, "grad_norm": 2.0383969926345498, "learning_rate": 1.9923283365872886e-05, "loss": 0.9664, "step": 280 }, { "epoch": 0.20676968359087564, "grad_norm": 2.173515487177821, "learning_rate": 1.9922297946675083e-05, "loss": 1.0026, "step": 281 }, { "epoch": 0.20750551876379691, "grad_norm": 2.003324149039148, "learning_rate": 1.9921306263666115e-05, "loss": 0.9017, "step": 282 }, { "epoch": 0.20824135393671817, "grad_norm": 2.1840290204841097, "learning_rate": 1.9920308317472023e-05, "loss": 0.9753, "step": 283 }, { "epoch": 0.20897718910963944, "grad_norm": 2.1733401922446056, "learning_rate": 1.991930410872279e-05, "loss": 0.9562, "step": 284 }, { "epoch": 0.2097130242825607, "grad_norm": 2.042850003931358, "learning_rate": 1.9918293638052365e-05, "loss": 0.9764, "step": 285 }, { "epoch": 0.21044885945548197, "grad_norm": 2.0450294815647267, "learning_rate": 1.9917276906098643e-05, "loss": 0.9204, "step": 286 }, { "epoch": 0.21118469462840322, "grad_norm": 2.060644112413295, "learning_rate": 1.991625391350347e-05, "loss": 1.0227, "step": 287 }, { "epoch": 0.2119205298013245, "grad_norm": 2.1564335032485515, "learning_rate": 1.9915224660912647e-05, "loss": 0.9059, "step": 288 }, { "epoch": 0.21265636497424578, "grad_norm": 2.08992526831175, "learning_rate": 1.991418914897593e-05, "loss": 1.0028, "step": 289 }, { "epoch": 0.21339220014716703, "grad_norm": 2.0959232566180046, "learning_rate": 1.9913147378347025e-05, "loss": 0.9436, "step": 290 }, { "epoch": 0.2141280353200883, "grad_norm": 2.023966058764531, "learning_rate": 1.9912099349683586e-05, "loss": 0.9487, "step": 291 }, { "epoch": 0.21486387049300956, "grad_norm": 2.1559697618182168, "learning_rate": 1.9911045063647214e-05, "loss": 1.0146, "step": 292 }, { "epoch": 0.21559970566593084, "grad_norm": 2.0765222246497297, "learning_rate": 1.9909984520903474e-05, "loss": 0.9493, "step": 293 }, { "epoch": 0.2163355408388521, "grad_norm": 2.065688227913734, "learning_rate": 1.9908917722121867e-05, "loss": 0.9589, "step": 294 }, { "epoch": 0.21707137601177337, "grad_norm": 1.9295777934190852, "learning_rate": 1.9907844667975847e-05, "loss": 0.9774, "step": 295 }, { "epoch": 0.21780721118469462, "grad_norm": 1.9153870319087736, "learning_rate": 1.9906765359142825e-05, "loss": 0.9921, "step": 296 }, { "epoch": 0.2185430463576159, "grad_norm": 1.9399627372849384, "learning_rate": 1.9905679796304148e-05, "loss": 0.9323, "step": 297 }, { "epoch": 0.21927888153053715, "grad_norm": 2.085212904086171, "learning_rate": 1.9904587980145117e-05, "loss": 0.8986, "step": 298 }, { "epoch": 0.22001471670345843, "grad_norm": 1.8791801248074498, "learning_rate": 1.9903489911354986e-05, "loss": 0.9401, "step": 299 }, { "epoch": 0.22075055187637968, "grad_norm": 1.9643773355796086, "learning_rate": 1.9902385590626947e-05, "loss": 0.9158, "step": 300 }, { "epoch": 0.22148638704930096, "grad_norm": 2.108968174424245, "learning_rate": 1.990127501865814e-05, "loss": 0.9561, "step": 301 }, { "epoch": 0.2222222222222222, "grad_norm": 1.9734430914124288, "learning_rate": 1.990015819614966e-05, "loss": 0.951, "step": 302 }, { "epoch": 0.2229580573951435, "grad_norm": 2.1006658420926994, "learning_rate": 1.9899035123806536e-05, "loss": 0.9842, "step": 303 }, { "epoch": 0.22369389256806477, "grad_norm": 1.9939056119168082, "learning_rate": 1.989790580233775e-05, "loss": 0.9503, "step": 304 }, { "epoch": 0.22442972774098602, "grad_norm": 1.9552519337078367, "learning_rate": 1.9896770232456233e-05, "loss": 0.8851, "step": 305 }, { "epoch": 0.2251655629139073, "grad_norm": 2.0635860121878244, "learning_rate": 1.9895628414878848e-05, "loss": 0.9886, "step": 306 }, { "epoch": 0.22590139808682855, "grad_norm": 2.079116924988777, "learning_rate": 1.989448035032641e-05, "loss": 0.968, "step": 307 }, { "epoch": 0.22663723325974983, "grad_norm": 2.1052033768366925, "learning_rate": 1.9893326039523675e-05, "loss": 0.9468, "step": 308 }, { "epoch": 0.22737306843267108, "grad_norm": 1.960624454345257, "learning_rate": 1.9892165483199347e-05, "loss": 0.9755, "step": 309 }, { "epoch": 0.22810890360559236, "grad_norm": 2.0288322489680977, "learning_rate": 1.989099868208607e-05, "loss": 0.983, "step": 310 }, { "epoch": 0.2288447387785136, "grad_norm": 1.915990839654117, "learning_rate": 1.9889825636920427e-05, "loss": 0.9606, "step": 311 }, { "epoch": 0.22958057395143489, "grad_norm": 2.011947058680569, "learning_rate": 1.9888646348442944e-05, "loss": 0.9248, "step": 312 }, { "epoch": 0.23031640912435614, "grad_norm": 1.8848728469254317, "learning_rate": 1.9887460817398093e-05, "loss": 0.9612, "step": 313 }, { "epoch": 0.23105224429727741, "grad_norm": 1.9882239090737062, "learning_rate": 1.9886269044534283e-05, "loss": 0.9435, "step": 314 }, { "epoch": 0.23178807947019867, "grad_norm": 1.93753869449512, "learning_rate": 1.9885071030603864e-05, "loss": 1.037, "step": 315 }, { "epoch": 0.23252391464311994, "grad_norm": 2.123476876452812, "learning_rate": 1.9883866776363123e-05, "loss": 0.9346, "step": 316 }, { "epoch": 0.2332597498160412, "grad_norm": 2.220372620809629, "learning_rate": 1.9882656282572293e-05, "loss": 1.0013, "step": 317 }, { "epoch": 0.23399558498896247, "grad_norm": 2.0226508902371303, "learning_rate": 1.988143954999554e-05, "loss": 0.9131, "step": 318 }, { "epoch": 0.23473142016188372, "grad_norm": 2.044651832743809, "learning_rate": 1.9880216579400972e-05, "loss": 0.9711, "step": 319 }, { "epoch": 0.235467255334805, "grad_norm": 2.0262366477930023, "learning_rate": 1.9878987371560633e-05, "loss": 0.9908, "step": 320 }, { "epoch": 0.23620309050772628, "grad_norm": 1.9280482255325142, "learning_rate": 1.9877751927250505e-05, "loss": 0.9309, "step": 321 }, { "epoch": 0.23693892568064753, "grad_norm": 2.0380383857973268, "learning_rate": 1.9876510247250506e-05, "loss": 0.9125, "step": 322 }, { "epoch": 0.2376747608535688, "grad_norm": 2.12600414868418, "learning_rate": 1.9875262332344495e-05, "loss": 0.9865, "step": 323 }, { "epoch": 0.23841059602649006, "grad_norm": 2.1441819288508808, "learning_rate": 1.987400818332026e-05, "loss": 0.9909, "step": 324 }, { "epoch": 0.23914643119941134, "grad_norm": 2.0258293222357455, "learning_rate": 1.9872747800969526e-05, "loss": 0.9448, "step": 325 }, { "epoch": 0.2398822663723326, "grad_norm": 2.065706198484501, "learning_rate": 1.9871481186087964e-05, "loss": 1.0199, "step": 326 }, { "epoch": 0.24061810154525387, "grad_norm": 2.0185933043691557, "learning_rate": 1.9870208339475162e-05, "loss": 0.9953, "step": 327 }, { "epoch": 0.24135393671817512, "grad_norm": 1.9381415550569754, "learning_rate": 1.9868929261934657e-05, "loss": 0.9309, "step": 328 }, { "epoch": 0.2420897718910964, "grad_norm": 1.9793169784319269, "learning_rate": 1.986764395427391e-05, "loss": 0.9646, "step": 329 }, { "epoch": 0.24282560706401765, "grad_norm": 2.0190062617871645, "learning_rate": 1.9866352417304316e-05, "loss": 1.0498, "step": 330 }, { "epoch": 0.24356144223693893, "grad_norm": 1.7594756245086014, "learning_rate": 1.986505465184121e-05, "loss": 0.9032, "step": 331 }, { "epoch": 0.24429727740986018, "grad_norm": 2.0526458234165905, "learning_rate": 1.9863750658703846e-05, "loss": 0.9934, "step": 332 }, { "epoch": 0.24503311258278146, "grad_norm": 2.2312584590668085, "learning_rate": 1.9862440438715427e-05, "loss": 0.975, "step": 333 }, { "epoch": 0.2457689477557027, "grad_norm": 2.0969021507021908, "learning_rate": 1.986112399270307e-05, "loss": 0.9552, "step": 334 }, { "epoch": 0.246504782928624, "grad_norm": 2.091094841724625, "learning_rate": 1.985980132149783e-05, "loss": 1.0132, "step": 335 }, { "epoch": 0.24724061810154527, "grad_norm": 1.9056737863345323, "learning_rate": 1.985847242593469e-05, "loss": 0.987, "step": 336 }, { "epoch": 0.24797645327446652, "grad_norm": 1.9832320512320325, "learning_rate": 1.985713730685257e-05, "loss": 0.9292, "step": 337 }, { "epoch": 0.2487122884473878, "grad_norm": 1.8605317210477095, "learning_rate": 1.9855795965094307e-05, "loss": 0.917, "step": 338 }, { "epoch": 0.24944812362030905, "grad_norm": 2.0182913104680704, "learning_rate": 1.9854448401506677e-05, "loss": 0.972, "step": 339 }, { "epoch": 0.2501839587932303, "grad_norm": 1.9692916849566924, "learning_rate": 1.985309461694037e-05, "loss": 0.9793, "step": 340 }, { "epoch": 0.2509197939661516, "grad_norm": 1.8850833862849983, "learning_rate": 1.9851734612250015e-05, "loss": 0.9429, "step": 341 }, { "epoch": 0.25165562913907286, "grad_norm": 1.9114610891912862, "learning_rate": 1.985036838829417e-05, "loss": 0.96, "step": 342 }, { "epoch": 0.2523914643119941, "grad_norm": 1.950180980158786, "learning_rate": 1.9848995945935305e-05, "loss": 1.0048, "step": 343 }, { "epoch": 0.25312729948491536, "grad_norm": 2.09961377765967, "learning_rate": 1.9847617286039834e-05, "loss": 0.9754, "step": 344 }, { "epoch": 0.25386313465783666, "grad_norm": 1.9746586475714387, "learning_rate": 1.9846232409478076e-05, "loss": 0.9749, "step": 345 }, { "epoch": 0.2545989698307579, "grad_norm": 2.074891844365653, "learning_rate": 1.984484131712429e-05, "loss": 0.9894, "step": 346 }, { "epoch": 0.25533480500367917, "grad_norm": 1.9251868371913734, "learning_rate": 1.9843444009856652e-05, "loss": 0.9289, "step": 347 }, { "epoch": 0.2560706401766004, "grad_norm": 2.1006618229314533, "learning_rate": 1.9842040488557267e-05, "loss": 0.9207, "step": 348 }, { "epoch": 0.2568064753495217, "grad_norm": 1.9804444607895395, "learning_rate": 1.9840630754112152e-05, "loss": 0.9226, "step": 349 }, { "epoch": 0.257542310522443, "grad_norm": 1.9030968881434533, "learning_rate": 1.983921480741126e-05, "loss": 0.9763, "step": 350 }, { "epoch": 0.2582781456953642, "grad_norm": 2.37497897821758, "learning_rate": 1.9837792649348454e-05, "loss": 0.9551, "step": 351 }, { "epoch": 0.25901398086828553, "grad_norm": 1.9998020953615339, "learning_rate": 1.9836364280821522e-05, "loss": 0.9861, "step": 352 }, { "epoch": 0.2597498160412068, "grad_norm": 2.008755702082269, "learning_rate": 1.983492970273218e-05, "loss": 1.0001, "step": 353 }, { "epoch": 0.26048565121412803, "grad_norm": 1.961889988324027, "learning_rate": 1.9833488915986054e-05, "loss": 0.9768, "step": 354 }, { "epoch": 0.2612214863870493, "grad_norm": 1.98577070219214, "learning_rate": 1.9832041921492688e-05, "loss": 0.9361, "step": 355 }, { "epoch": 0.2619573215599706, "grad_norm": 1.998276548022258, "learning_rate": 1.983058872016556e-05, "loss": 0.9741, "step": 356 }, { "epoch": 0.26269315673289184, "grad_norm": 2.4000543045757845, "learning_rate": 1.9829129312922044e-05, "loss": 0.9682, "step": 357 }, { "epoch": 0.2634289919058131, "grad_norm": 2.0304003590671216, "learning_rate": 1.9827663700683454e-05, "loss": 0.982, "step": 358 }, { "epoch": 0.26416482707873434, "grad_norm": 2.018993408440803, "learning_rate": 1.9826191884375007e-05, "loss": 0.9704, "step": 359 }, { "epoch": 0.26490066225165565, "grad_norm": 2.004732712309117, "learning_rate": 1.9824713864925843e-05, "loss": 0.9822, "step": 360 }, { "epoch": 0.2656364974245769, "grad_norm": 1.9762236957858388, "learning_rate": 1.982322964326901e-05, "loss": 0.998, "step": 361 }, { "epoch": 0.26637233259749815, "grad_norm": 1.9720249211529208, "learning_rate": 1.9821739220341483e-05, "loss": 0.9009, "step": 362 }, { "epoch": 0.2671081677704194, "grad_norm": 1.9006704648498403, "learning_rate": 1.982024259708414e-05, "loss": 0.9267, "step": 363 }, { "epoch": 0.2678440029433407, "grad_norm": 2.0730156536417885, "learning_rate": 1.9818739774441784e-05, "loss": 1.0246, "step": 364 }, { "epoch": 0.26857983811626196, "grad_norm": 2.048154153149425, "learning_rate": 1.9817230753363127e-05, "loss": 0.884, "step": 365 }, { "epoch": 0.2693156732891832, "grad_norm": 1.8400020216382713, "learning_rate": 1.9815715534800785e-05, "loss": 0.9252, "step": 366 }, { "epoch": 0.27005150846210446, "grad_norm": 2.0758226922913643, "learning_rate": 1.98141941197113e-05, "loss": 0.9758, "step": 367 }, { "epoch": 0.27078734363502577, "grad_norm": 1.8922960960111095, "learning_rate": 1.9812666509055124e-05, "loss": 0.927, "step": 368 }, { "epoch": 0.271523178807947, "grad_norm": 1.9689511820537138, "learning_rate": 1.9811132703796615e-05, "loss": 0.9552, "step": 369 }, { "epoch": 0.27225901398086827, "grad_norm": 1.946942633558561, "learning_rate": 1.980959270490404e-05, "loss": 0.9309, "step": 370 }, { "epoch": 0.2729948491537896, "grad_norm": 1.9486205080878791, "learning_rate": 1.9808046513349583e-05, "loss": 0.9275, "step": 371 }, { "epoch": 0.2737306843267108, "grad_norm": 1.9392462012821203, "learning_rate": 1.9806494130109327e-05, "loss": 0.9388, "step": 372 }, { "epoch": 0.2744665194996321, "grad_norm": 1.9248145571268096, "learning_rate": 1.980493555616328e-05, "loss": 0.9633, "step": 373 }, { "epoch": 0.27520235467255333, "grad_norm": 2.0210851342643528, "learning_rate": 1.980337079249534e-05, "loss": 0.9614, "step": 374 }, { "epoch": 0.27593818984547464, "grad_norm": 1.7715609820039433, "learning_rate": 1.9801799840093326e-05, "loss": 0.9157, "step": 375 }, { "epoch": 0.2766740250183959, "grad_norm": 2.088728474627495, "learning_rate": 1.980022269994896e-05, "loss": 1.0277, "step": 376 }, { "epoch": 0.27740986019131714, "grad_norm": 2.0112015896601205, "learning_rate": 1.9798639373057864e-05, "loss": 0.9942, "step": 377 }, { "epoch": 0.2781456953642384, "grad_norm": 1.9078231540533324, "learning_rate": 1.9797049860419572e-05, "loss": 0.9366, "step": 378 }, { "epoch": 0.2788815305371597, "grad_norm": 1.9482209318692174, "learning_rate": 1.9795454163037523e-05, "loss": 0.982, "step": 379 }, { "epoch": 0.27961736571008095, "grad_norm": 2.1008534439976962, "learning_rate": 1.9793852281919062e-05, "loss": 0.9571, "step": 380 }, { "epoch": 0.2803532008830022, "grad_norm": 1.938549580658199, "learning_rate": 1.9792244218075426e-05, "loss": 0.8706, "step": 381 }, { "epoch": 0.28108903605592345, "grad_norm": 2.0515152674573893, "learning_rate": 1.9790629972521772e-05, "loss": 0.9828, "step": 382 }, { "epoch": 0.28182487122884475, "grad_norm": 2.1498157610990374, "learning_rate": 1.9789009546277146e-05, "loss": 0.952, "step": 383 }, { "epoch": 0.282560706401766, "grad_norm": 2.073197378838484, "learning_rate": 1.9787382940364506e-05, "loss": 0.9911, "step": 384 }, { "epoch": 0.28329654157468726, "grad_norm": 2.002081519534446, "learning_rate": 1.97857501558107e-05, "loss": 0.9756, "step": 385 }, { "epoch": 0.28403237674760856, "grad_norm": 1.9516513036495493, "learning_rate": 1.978411119364649e-05, "loss": 0.8827, "step": 386 }, { "epoch": 0.2847682119205298, "grad_norm": 2.469935654137178, "learning_rate": 1.9782466054906525e-05, "loss": 1.0289, "step": 387 }, { "epoch": 0.28550404709345106, "grad_norm": 1.855039200361118, "learning_rate": 1.9780814740629357e-05, "loss": 0.9215, "step": 388 }, { "epoch": 0.2862398822663723, "grad_norm": 1.814701111142721, "learning_rate": 1.9779157251857445e-05, "loss": 0.9604, "step": 389 }, { "epoch": 0.2869757174392936, "grad_norm": 1.8608307307095602, "learning_rate": 1.977749358963713e-05, "loss": 0.9533, "step": 390 }, { "epoch": 0.28771155261221487, "grad_norm": 1.8376385918996372, "learning_rate": 1.9775823755018665e-05, "loss": 0.9568, "step": 391 }, { "epoch": 0.2884473877851361, "grad_norm": 1.890551011803974, "learning_rate": 1.97741477490562e-05, "loss": 0.9217, "step": 392 }, { "epoch": 0.2891832229580574, "grad_norm": 2.04447730452833, "learning_rate": 1.977246557280776e-05, "loss": 0.9765, "step": 393 }, { "epoch": 0.2899190581309787, "grad_norm": 1.987908576489268, "learning_rate": 1.9770777227335292e-05, "loss": 0.9692, "step": 394 }, { "epoch": 0.29065489330389993, "grad_norm": 2.097830958311067, "learning_rate": 1.9769082713704616e-05, "loss": 0.9464, "step": 395 }, { "epoch": 0.2913907284768212, "grad_norm": 1.9288362806405561, "learning_rate": 1.9767382032985466e-05, "loss": 0.9192, "step": 396 }, { "epoch": 0.29212656364974243, "grad_norm": 1.9771330296865681, "learning_rate": 1.976567518625145e-05, "loss": 0.9702, "step": 397 }, { "epoch": 0.29286239882266374, "grad_norm": 1.8953886822582893, "learning_rate": 1.976396217458008e-05, "loss": 0.9812, "step": 398 }, { "epoch": 0.293598233995585, "grad_norm": 1.9408677969145662, "learning_rate": 1.9762242999052756e-05, "loss": 0.9514, "step": 399 }, { "epoch": 0.29433406916850624, "grad_norm": 1.963752691549827, "learning_rate": 1.976051766075477e-05, "loss": 0.9667, "step": 400 }, { "epoch": 0.29506990434142755, "grad_norm": 1.9054044176818024, "learning_rate": 1.9758786160775306e-05, "loss": 0.919, "step": 401 }, { "epoch": 0.2958057395143488, "grad_norm": 1.8802508983592456, "learning_rate": 1.975704850020744e-05, "loss": 0.9429, "step": 402 }, { "epoch": 0.29654157468727005, "grad_norm": 1.8266176993856096, "learning_rate": 1.9755304680148125e-05, "loss": 0.9368, "step": 403 }, { "epoch": 0.2972774098601913, "grad_norm": 1.934665509900312, "learning_rate": 1.9753554701698215e-05, "loss": 0.954, "step": 404 }, { "epoch": 0.2980132450331126, "grad_norm": 1.8565427016700475, "learning_rate": 1.9751798565962446e-05, "loss": 0.9349, "step": 405 }, { "epoch": 0.29874908020603386, "grad_norm": 1.9486096281900174, "learning_rate": 1.9750036274049447e-05, "loss": 1.0342, "step": 406 }, { "epoch": 0.2994849153789551, "grad_norm": 1.9990787877101293, "learning_rate": 1.9748267827071723e-05, "loss": 0.9394, "step": 407 }, { "epoch": 0.30022075055187636, "grad_norm": 1.9407156818193239, "learning_rate": 1.9746493226145673e-05, "loss": 1.0028, "step": 408 }, { "epoch": 0.30095658572479767, "grad_norm": 1.9729693096526804, "learning_rate": 1.974471247239158e-05, "loss": 0.9401, "step": 409 }, { "epoch": 0.3016924208977189, "grad_norm": 2.0016410161220177, "learning_rate": 1.9742925566933607e-05, "loss": 0.9979, "step": 410 }, { "epoch": 0.30242825607064017, "grad_norm": 2.005108270665045, "learning_rate": 1.97411325108998e-05, "loss": 0.9333, "step": 411 }, { "epoch": 0.3031640912435614, "grad_norm": 2.050369352555863, "learning_rate": 1.97393333054221e-05, "loss": 0.9458, "step": 412 }, { "epoch": 0.3038999264164827, "grad_norm": 1.7552634466615784, "learning_rate": 1.973752795163631e-05, "loss": 0.9787, "step": 413 }, { "epoch": 0.304635761589404, "grad_norm": 1.8818555022898955, "learning_rate": 1.9735716450682137e-05, "loss": 0.9172, "step": 414 }, { "epoch": 0.3053715967623252, "grad_norm": 2.0232606696704076, "learning_rate": 1.9733898803703145e-05, "loss": 0.9481, "step": 415 }, { "epoch": 0.30610743193524653, "grad_norm": 1.9185297791293177, "learning_rate": 1.97320750118468e-05, "loss": 1.0035, "step": 416 }, { "epoch": 0.3068432671081678, "grad_norm": 1.9963383889626591, "learning_rate": 1.973024507626443e-05, "loss": 0.959, "step": 417 }, { "epoch": 0.30757910228108903, "grad_norm": 1.9284545776393582, "learning_rate": 1.972840899811125e-05, "loss": 0.9548, "step": 418 }, { "epoch": 0.3083149374540103, "grad_norm": 2.0832407645246978, "learning_rate": 1.972656677854635e-05, "loss": 1.0144, "step": 419 }, { "epoch": 0.3090507726269316, "grad_norm": 1.9327500795188939, "learning_rate": 1.97247184187327e-05, "loss": 0.9738, "step": 420 }, { "epoch": 0.30978660779985284, "grad_norm": 1.8799860878448367, "learning_rate": 1.9722863919837146e-05, "loss": 0.9867, "step": 421 }, { "epoch": 0.3105224429727741, "grad_norm": 2.012083707767368, "learning_rate": 1.9721003283030402e-05, "loss": 0.9425, "step": 422 }, { "epoch": 0.31125827814569534, "grad_norm": 1.9291435440451379, "learning_rate": 1.971913650948707e-05, "loss": 0.9092, "step": 423 }, { "epoch": 0.31199411331861665, "grad_norm": 1.8335897960527427, "learning_rate": 1.9717263600385614e-05, "loss": 0.9247, "step": 424 }, { "epoch": 0.3127299484915379, "grad_norm": 2.049072944592503, "learning_rate": 1.9715384556908375e-05, "loss": 0.9216, "step": 425 }, { "epoch": 0.31346578366445915, "grad_norm": 2.032206151642003, "learning_rate": 1.9713499380241568e-05, "loss": 0.9722, "step": 426 }, { "epoch": 0.3142016188373804, "grad_norm": 1.9299574740442451, "learning_rate": 1.9711608071575285e-05, "loss": 0.9222, "step": 427 }, { "epoch": 0.3149374540103017, "grad_norm": 1.867703147342899, "learning_rate": 1.9709710632103473e-05, "loss": 0.9605, "step": 428 }, { "epoch": 0.31567328918322296, "grad_norm": 1.9289979920292954, "learning_rate": 1.970780706302397e-05, "loss": 0.9578, "step": 429 }, { "epoch": 0.3164091243561442, "grad_norm": 1.68747361924899, "learning_rate": 1.970589736553847e-05, "loss": 0.8907, "step": 430 }, { "epoch": 0.31714495952906546, "grad_norm": 2.0156051212397386, "learning_rate": 1.9703981540852534e-05, "loss": 1.0074, "step": 431 }, { "epoch": 0.31788079470198677, "grad_norm": 1.8970737164453926, "learning_rate": 1.97020595901756e-05, "loss": 0.9393, "step": 432 }, { "epoch": 0.318616629874908, "grad_norm": 1.836791525061121, "learning_rate": 1.970013151472097e-05, "loss": 0.9433, "step": 433 }, { "epoch": 0.31935246504782927, "grad_norm": 1.9526207598526102, "learning_rate": 1.969819731570581e-05, "loss": 0.9556, "step": 434 }, { "epoch": 0.3200883002207506, "grad_norm": 1.891490801019088, "learning_rate": 1.969625699435115e-05, "loss": 0.9626, "step": 435 }, { "epoch": 0.3208241353936718, "grad_norm": 1.836787284780508, "learning_rate": 1.96943105518819e-05, "loss": 0.8895, "step": 436 }, { "epoch": 0.3215599705665931, "grad_norm": 2.01692944281003, "learning_rate": 1.969235798952681e-05, "loss": 1.0008, "step": 437 }, { "epoch": 0.32229580573951433, "grad_norm": 1.753186770441785, "learning_rate": 1.9690399308518508e-05, "loss": 0.909, "step": 438 }, { "epoch": 0.32303164091243564, "grad_norm": 1.797954227514674, "learning_rate": 1.968843451009349e-05, "loss": 0.8953, "step": 439 }, { "epoch": 0.3237674760853569, "grad_norm": 1.9485724431415787, "learning_rate": 1.96864635954921e-05, "loss": 0.9908, "step": 440 }, { "epoch": 0.32450331125827814, "grad_norm": 1.8265998704844313, "learning_rate": 1.968448656595855e-05, "loss": 0.9972, "step": 441 }, { "epoch": 0.3252391464311994, "grad_norm": 1.7688461464918745, "learning_rate": 1.9682503422740915e-05, "loss": 0.9137, "step": 442 }, { "epoch": 0.3259749816041207, "grad_norm": 1.9456273539216238, "learning_rate": 1.9680514167091124e-05, "loss": 0.9259, "step": 443 }, { "epoch": 0.32671081677704195, "grad_norm": 2.0411402763158044, "learning_rate": 1.9678518800264966e-05, "loss": 0.9239, "step": 444 }, { "epoch": 0.3274466519499632, "grad_norm": 1.9081880447683566, "learning_rate": 1.967651732352209e-05, "loss": 0.9162, "step": 445 }, { "epoch": 0.32818248712288445, "grad_norm": 1.8546758259852336, "learning_rate": 1.9674509738126e-05, "loss": 0.9583, "step": 446 }, { "epoch": 0.32891832229580575, "grad_norm": 2.0456204016597277, "learning_rate": 1.967249604534406e-05, "loss": 0.9333, "step": 447 }, { "epoch": 0.329654157468727, "grad_norm": 1.9255890346622127, "learning_rate": 1.9670476246447484e-05, "loss": 0.9643, "step": 448 }, { "epoch": 0.33038999264164826, "grad_norm": 1.9305732167332526, "learning_rate": 1.9668450342711343e-05, "loss": 0.8557, "step": 449 }, { "epoch": 0.33112582781456956, "grad_norm": 1.8937365434210784, "learning_rate": 1.9666418335414565e-05, "loss": 0.9204, "step": 450 }, { "epoch": 0.3318616629874908, "grad_norm": 1.8720145376470538, "learning_rate": 1.966438022583993e-05, "loss": 0.8842, "step": 451 }, { "epoch": 0.33259749816041206, "grad_norm": 1.8777577897810256, "learning_rate": 1.966233601527406e-05, "loss": 0.9648, "step": 452 }, { "epoch": 0.3333333333333333, "grad_norm": 1.8804390520380103, "learning_rate": 1.966028570500744e-05, "loss": 0.8687, "step": 453 }, { "epoch": 0.3340691685062546, "grad_norm": 1.7924742031640108, "learning_rate": 1.9658229296334416e-05, "loss": 0.9486, "step": 454 }, { "epoch": 0.33480500367917587, "grad_norm": 1.9576421113632076, "learning_rate": 1.9656166790553157e-05, "loss": 0.9441, "step": 455 }, { "epoch": 0.3355408388520971, "grad_norm": 1.9288360372366626, "learning_rate": 1.9654098188965695e-05, "loss": 0.9739, "step": 456 }, { "epoch": 0.3362766740250184, "grad_norm": 1.9786067042133868, "learning_rate": 1.9652023492877915e-05, "loss": 0.9585, "step": 457 }, { "epoch": 0.3370125091979397, "grad_norm": 1.9728285574839153, "learning_rate": 1.964994270359954e-05, "loss": 1.0002, "step": 458 }, { "epoch": 0.33774834437086093, "grad_norm": 1.8955978503452753, "learning_rate": 1.9647855822444152e-05, "loss": 0.9596, "step": 459 }, { "epoch": 0.3384841795437822, "grad_norm": 1.8801673820852893, "learning_rate": 1.964576285072916e-05, "loss": 0.941, "step": 460 }, { "epoch": 0.33922001471670343, "grad_norm": 1.873233827840871, "learning_rate": 1.9643663789775836e-05, "loss": 0.9203, "step": 461 }, { "epoch": 0.33995584988962474, "grad_norm": 1.9092362312401263, "learning_rate": 1.9641558640909286e-05, "loss": 0.9653, "step": 462 }, { "epoch": 0.340691685062546, "grad_norm": 1.9164184502700878, "learning_rate": 1.963944740545846e-05, "loss": 0.8861, "step": 463 }, { "epoch": 0.34142752023546724, "grad_norm": 1.9003573360930195, "learning_rate": 1.9637330084756154e-05, "loss": 0.9705, "step": 464 }, { "epoch": 0.34216335540838855, "grad_norm": 1.8592249316898481, "learning_rate": 1.9635206680139004e-05, "loss": 0.8892, "step": 465 }, { "epoch": 0.3428991905813098, "grad_norm": 1.9863717704437271, "learning_rate": 1.9633077192947486e-05, "loss": 0.9669, "step": 466 }, { "epoch": 0.34363502575423105, "grad_norm": 1.9177683171719417, "learning_rate": 1.9630941624525914e-05, "loss": 1.0019, "step": 467 }, { "epoch": 0.3443708609271523, "grad_norm": 1.8159502569861814, "learning_rate": 1.9628799976222444e-05, "loss": 0.9536, "step": 468 }, { "epoch": 0.3451066961000736, "grad_norm": 2.015019131429284, "learning_rate": 1.9626652249389076e-05, "loss": 0.9516, "step": 469 }, { "epoch": 0.34584253127299486, "grad_norm": 1.9011622050770476, "learning_rate": 1.962449844538163e-05, "loss": 0.9667, "step": 470 }, { "epoch": 0.3465783664459161, "grad_norm": 1.8712307983961374, "learning_rate": 1.9622338565559783e-05, "loss": 0.9086, "step": 471 }, { "epoch": 0.34731420161883736, "grad_norm": 1.9737488515762363, "learning_rate": 1.9620172611287028e-05, "loss": 0.9606, "step": 472 }, { "epoch": 0.34805003679175867, "grad_norm": 1.8405189548944296, "learning_rate": 1.9618000583930708e-05, "loss": 0.908, "step": 473 }, { "epoch": 0.3487858719646799, "grad_norm": 1.8438146342495425, "learning_rate": 1.9615822484861994e-05, "loss": 0.9519, "step": 474 }, { "epoch": 0.34952170713760117, "grad_norm": 1.9806524066227826, "learning_rate": 1.9613638315455888e-05, "loss": 0.9367, "step": 475 }, { "epoch": 0.3502575423105224, "grad_norm": 1.874488709367565, "learning_rate": 1.9611448077091228e-05, "loss": 0.9553, "step": 476 }, { "epoch": 0.3509933774834437, "grad_norm": 1.784937540761977, "learning_rate": 1.9609251771150682e-05, "loss": 0.8865, "step": 477 }, { "epoch": 0.351729212656365, "grad_norm": 1.9001233276529126, "learning_rate": 1.9607049399020746e-05, "loss": 0.9456, "step": 478 }, { "epoch": 0.3524650478292862, "grad_norm": 1.7842151859029445, "learning_rate": 1.960484096209175e-05, "loss": 0.9427, "step": 479 }, { "epoch": 0.35320088300220753, "grad_norm": 1.9153063153062702, "learning_rate": 1.960262646175785e-05, "loss": 0.9323, "step": 480 }, { "epoch": 0.3539367181751288, "grad_norm": 1.9661841162510894, "learning_rate": 1.9600405899417026e-05, "loss": 0.9577, "step": 481 }, { "epoch": 0.35467255334805003, "grad_norm": 1.9674230030039845, "learning_rate": 1.959817927647109e-05, "loss": 0.9233, "step": 482 }, { "epoch": 0.3554083885209713, "grad_norm": 1.8612874628157778, "learning_rate": 1.9595946594325684e-05, "loss": 0.9289, "step": 483 }, { "epoch": 0.3561442236938926, "grad_norm": 1.861694990620573, "learning_rate": 1.9593707854390263e-05, "loss": 0.9887, "step": 484 }, { "epoch": 0.35688005886681384, "grad_norm": 1.854895916875523, "learning_rate": 1.9591463058078118e-05, "loss": 0.9696, "step": 485 }, { "epoch": 0.3576158940397351, "grad_norm": 2.091779585257315, "learning_rate": 1.958921220680635e-05, "loss": 1.0057, "step": 486 }, { "epoch": 0.35835172921265634, "grad_norm": 1.9292872142343283, "learning_rate": 1.9586955301995903e-05, "loss": 0.9763, "step": 487 }, { "epoch": 0.35908756438557765, "grad_norm": 1.8353988510688606, "learning_rate": 1.958469234507152e-05, "loss": 0.9198, "step": 488 }, { "epoch": 0.3598233995584989, "grad_norm": 1.7780876711598792, "learning_rate": 1.958242333746178e-05, "loss": 0.8874, "step": 489 }, { "epoch": 0.36055923473142015, "grad_norm": 1.901732040188799, "learning_rate": 1.9580148280599075e-05, "loss": 0.8952, "step": 490 }, { "epoch": 0.3612950699043414, "grad_norm": 1.9864655503356903, "learning_rate": 1.9577867175919616e-05, "loss": 0.9439, "step": 491 }, { "epoch": 0.3620309050772627, "grad_norm": 2.0893196079258, "learning_rate": 1.9575580024863434e-05, "loss": 0.9841, "step": 492 }, { "epoch": 0.36276674025018396, "grad_norm": 1.8154764214499262, "learning_rate": 1.957328682887438e-05, "loss": 0.8781, "step": 493 }, { "epoch": 0.3635025754231052, "grad_norm": 1.9150305760340642, "learning_rate": 1.9570987589400108e-05, "loss": 0.9337, "step": 494 }, { "epoch": 0.36423841059602646, "grad_norm": 1.8042767959420936, "learning_rate": 1.9568682307892108e-05, "loss": 0.9146, "step": 495 }, { "epoch": 0.36497424576894777, "grad_norm": 1.7577518988858984, "learning_rate": 1.9566370985805666e-05, "loss": 0.9318, "step": 496 }, { "epoch": 0.365710080941869, "grad_norm": 1.832513077596428, "learning_rate": 1.9564053624599886e-05, "loss": 0.9182, "step": 497 }, { "epoch": 0.36644591611479027, "grad_norm": 1.8293816118268924, "learning_rate": 1.9561730225737694e-05, "loss": 0.9477, "step": 498 }, { "epoch": 0.3671817512877116, "grad_norm": 1.7842683049843435, "learning_rate": 1.9559400790685813e-05, "loss": 0.8826, "step": 499 }, { "epoch": 0.36791758646063283, "grad_norm": 1.8858839487613408, "learning_rate": 1.9557065320914787e-05, "loss": 0.9625, "step": 500 }, { "epoch": 0.3686534216335541, "grad_norm": 1.793654442213453, "learning_rate": 1.9554723817898965e-05, "loss": 0.9513, "step": 501 }, { "epoch": 0.36938925680647533, "grad_norm": 1.702464172505763, "learning_rate": 1.9552376283116508e-05, "loss": 0.8814, "step": 502 }, { "epoch": 0.37012509197939664, "grad_norm": 1.873537232198834, "learning_rate": 1.9550022718049383e-05, "loss": 0.9632, "step": 503 }, { "epoch": 0.3708609271523179, "grad_norm": 1.8512370897820394, "learning_rate": 1.9547663124183356e-05, "loss": 0.9371, "step": 504 }, { "epoch": 0.37159676232523914, "grad_norm": 1.7995742550131104, "learning_rate": 1.9545297503008014e-05, "loss": 0.9425, "step": 505 }, { "epoch": 0.3723325974981604, "grad_norm": 2.06740687996939, "learning_rate": 1.9542925856016742e-05, "loss": 0.9195, "step": 506 }, { "epoch": 0.3730684326710817, "grad_norm": 2.0238110393977222, "learning_rate": 1.954054818470672e-05, "loss": 0.9808, "step": 507 }, { "epoch": 0.37380426784400295, "grad_norm": 1.9472457883892635, "learning_rate": 1.953816449057895e-05, "loss": 0.9846, "step": 508 }, { "epoch": 0.3745401030169242, "grad_norm": 1.8845318396878805, "learning_rate": 1.9535774775138216e-05, "loss": 0.9962, "step": 509 }, { "epoch": 0.37527593818984545, "grad_norm": 1.7730098145249609, "learning_rate": 1.9533379039893118e-05, "loss": 0.9109, "step": 510 }, { "epoch": 0.37601177336276675, "grad_norm": 1.7662648461603967, "learning_rate": 1.9530977286356053e-05, "loss": 0.9158, "step": 511 }, { "epoch": 0.376747608535688, "grad_norm": 1.8745812537985784, "learning_rate": 1.952856951604321e-05, "loss": 0.9232, "step": 512 }, { "epoch": 0.37748344370860926, "grad_norm": 1.938292458823722, "learning_rate": 1.9526155730474586e-05, "loss": 0.9463, "step": 513 }, { "epoch": 0.37821927888153056, "grad_norm": 1.7606196767482192, "learning_rate": 1.9523735931173964e-05, "loss": 0.9527, "step": 514 }, { "epoch": 0.3789551140544518, "grad_norm": 1.8854033239827985, "learning_rate": 1.952131011966894e-05, "loss": 0.9359, "step": 515 }, { "epoch": 0.37969094922737306, "grad_norm": 1.9994380879657545, "learning_rate": 1.951887829749089e-05, "loss": 0.952, "step": 516 }, { "epoch": 0.3804267844002943, "grad_norm": 1.88687145708232, "learning_rate": 1.951644046617499e-05, "loss": 0.9024, "step": 517 }, { "epoch": 0.3811626195732156, "grad_norm": 1.9612912365521633, "learning_rate": 1.951399662726021e-05, "loss": 0.9379, "step": 518 }, { "epoch": 0.3818984547461369, "grad_norm": 1.7933780380157731, "learning_rate": 1.9511546782289307e-05, "loss": 0.9357, "step": 519 }, { "epoch": 0.3826342899190581, "grad_norm": 1.8739433914946975, "learning_rate": 1.950909093280885e-05, "loss": 0.9968, "step": 520 }, { "epoch": 0.3833701250919794, "grad_norm": 1.840831148940085, "learning_rate": 1.9506629080369163e-05, "loss": 0.9407, "step": 521 }, { "epoch": 0.3841059602649007, "grad_norm": 1.9078917290316066, "learning_rate": 1.9504161226524395e-05, "loss": 1.0213, "step": 522 }, { "epoch": 0.38484179543782193, "grad_norm": 1.8150236738025085, "learning_rate": 1.9501687372832466e-05, "loss": 0.8903, "step": 523 }, { "epoch": 0.3855776306107432, "grad_norm": 1.8484034199759047, "learning_rate": 1.9499207520855085e-05, "loss": 0.9549, "step": 524 }, { "epoch": 0.38631346578366443, "grad_norm": 1.9391423357798045, "learning_rate": 1.949672167215774e-05, "loss": 0.9277, "step": 525 }, { "epoch": 0.38704930095658574, "grad_norm": 1.9225924200348579, "learning_rate": 1.9494229828309724e-05, "loss": 0.8952, "step": 526 }, { "epoch": 0.387785136129507, "grad_norm": 1.843090991014795, "learning_rate": 1.9491731990884105e-05, "loss": 0.9306, "step": 527 }, { "epoch": 0.38852097130242824, "grad_norm": 1.9287330770404723, "learning_rate": 1.9489228161457724e-05, "loss": 0.9153, "step": 528 }, { "epoch": 0.38925680647534955, "grad_norm": 2.0046576103454274, "learning_rate": 1.948671834161122e-05, "loss": 0.9339, "step": 529 }, { "epoch": 0.3899926416482708, "grad_norm": 1.996460996792005, "learning_rate": 1.948420253292901e-05, "loss": 0.9381, "step": 530 }, { "epoch": 0.39072847682119205, "grad_norm": 1.8242008735727744, "learning_rate": 1.9481680736999283e-05, "loss": 0.9348, "step": 531 }, { "epoch": 0.3914643119941133, "grad_norm": 1.9404192212464668, "learning_rate": 1.947915295541402e-05, "loss": 0.9372, "step": 532 }, { "epoch": 0.3922001471670346, "grad_norm": 2.091749592810696, "learning_rate": 1.9476619189768978e-05, "loss": 0.9425, "step": 533 }, { "epoch": 0.39293598233995586, "grad_norm": 1.848410919904286, "learning_rate": 1.947407944166368e-05, "loss": 0.9546, "step": 534 }, { "epoch": 0.3936718175128771, "grad_norm": 1.9369718043701611, "learning_rate": 1.947153371270144e-05, "loss": 0.9977, "step": 535 }, { "epoch": 0.39440765268579836, "grad_norm": 1.803218933266155, "learning_rate": 1.946898200448934e-05, "loss": 0.9563, "step": 536 }, { "epoch": 0.39514348785871967, "grad_norm": 1.8670385664294848, "learning_rate": 1.9466424318638243e-05, "loss": 1.0314, "step": 537 }, { "epoch": 0.3958793230316409, "grad_norm": 1.866240192657055, "learning_rate": 1.946386065676277e-05, "loss": 0.9378, "step": 538 }, { "epoch": 0.39661515820456217, "grad_norm": 1.897408632183996, "learning_rate": 1.9461291020481336e-05, "loss": 0.9709, "step": 539 }, { "epoch": 0.3973509933774834, "grad_norm": 2.0270295582314786, "learning_rate": 1.9458715411416115e-05, "loss": 0.9693, "step": 540 }, { "epoch": 0.3980868285504047, "grad_norm": 2.0829822700902696, "learning_rate": 1.945613383119305e-05, "loss": 0.9934, "step": 541 }, { "epoch": 0.398822663723326, "grad_norm": 1.950376374311428, "learning_rate": 1.945354628144186e-05, "loss": 0.9126, "step": 542 }, { "epoch": 0.3995584988962472, "grad_norm": 1.8310428644167047, "learning_rate": 1.9450952763796035e-05, "loss": 0.9465, "step": 543 }, { "epoch": 0.40029433406916853, "grad_norm": 1.8777216360341267, "learning_rate": 1.944835327989282e-05, "loss": 0.9435, "step": 544 }, { "epoch": 0.4010301692420898, "grad_norm": 1.8347613659400537, "learning_rate": 1.9445747831373234e-05, "loss": 0.9169, "step": 545 }, { "epoch": 0.40176600441501104, "grad_norm": 1.8804860200195885, "learning_rate": 1.9443136419882064e-05, "loss": 0.9589, "step": 546 }, { "epoch": 0.4025018395879323, "grad_norm": 1.948690202103604, "learning_rate": 1.944051904706786e-05, "loss": 0.9793, "step": 547 }, { "epoch": 0.4032376747608536, "grad_norm": 1.8701440346073201, "learning_rate": 1.9437895714582926e-05, "loss": 0.9067, "step": 548 }, { "epoch": 0.40397350993377484, "grad_norm": 1.9549911415359942, "learning_rate": 1.9435266424083346e-05, "loss": 0.8891, "step": 549 }, { "epoch": 0.4047093451066961, "grad_norm": 1.8417660235559035, "learning_rate": 1.9432631177228948e-05, "loss": 0.9655, "step": 550 }, { "epoch": 0.40544518027961735, "grad_norm": 1.9949416356412781, "learning_rate": 1.942998997568333e-05, "loss": 0.906, "step": 551 }, { "epoch": 0.40618101545253865, "grad_norm": 1.9259627663946168, "learning_rate": 1.9427342821113845e-05, "loss": 0.9272, "step": 552 }, { "epoch": 0.4069168506254599, "grad_norm": 1.9825877357197181, "learning_rate": 1.942468971519161e-05, "loss": 0.9125, "step": 553 }, { "epoch": 0.40765268579838115, "grad_norm": 1.777872162696949, "learning_rate": 1.9422030659591485e-05, "loss": 0.9993, "step": 554 }, { "epoch": 0.4083885209713024, "grad_norm": 1.8405766699262434, "learning_rate": 1.941936565599211e-05, "loss": 0.9156, "step": 555 }, { "epoch": 0.4091243561442237, "grad_norm": 1.9390234080788202, "learning_rate": 1.941669470607585e-05, "loss": 0.9324, "step": 556 }, { "epoch": 0.40986019131714496, "grad_norm": 1.9064385322247352, "learning_rate": 1.941401781152885e-05, "loss": 0.9692, "step": 557 }, { "epoch": 0.4105960264900662, "grad_norm": 1.8339861772545685, "learning_rate": 1.9411334974040992e-05, "loss": 0.9553, "step": 558 }, { "epoch": 0.41133186166298746, "grad_norm": 1.8053677400324917, "learning_rate": 1.9408646195305914e-05, "loss": 0.8971, "step": 559 }, { "epoch": 0.41206769683590877, "grad_norm": 1.876785945039677, "learning_rate": 1.940595147702101e-05, "loss": 0.9027, "step": 560 }, { "epoch": 0.41280353200883, "grad_norm": 1.7266826486082822, "learning_rate": 1.9403250820887416e-05, "loss": 0.9399, "step": 561 }, { "epoch": 0.41353936718175127, "grad_norm": 1.8606739989070837, "learning_rate": 1.940054422861002e-05, "loss": 0.8776, "step": 562 }, { "epoch": 0.4142752023546726, "grad_norm": 1.8669654163054286, "learning_rate": 1.939783170189745e-05, "loss": 0.9261, "step": 563 }, { "epoch": 0.41501103752759383, "grad_norm": 1.733341107934013, "learning_rate": 1.93951132424621e-05, "loss": 0.9308, "step": 564 }, { "epoch": 0.4157468727005151, "grad_norm": 1.920363887509345, "learning_rate": 1.939238885202009e-05, "loss": 0.9543, "step": 565 }, { "epoch": 0.41648270787343633, "grad_norm": 2.0232463699994847, "learning_rate": 1.9389658532291286e-05, "loss": 0.9784, "step": 566 }, { "epoch": 0.41721854304635764, "grad_norm": 1.8726924342569566, "learning_rate": 1.9386922284999305e-05, "loss": 0.9293, "step": 567 }, { "epoch": 0.4179543782192789, "grad_norm": 1.8352152086185165, "learning_rate": 1.9384180111871502e-05, "loss": 0.8841, "step": 568 }, { "epoch": 0.41869021339220014, "grad_norm": 1.9093927258161947, "learning_rate": 1.9381432014638977e-05, "loss": 0.9898, "step": 569 }, { "epoch": 0.4194260485651214, "grad_norm": 1.7367191452751234, "learning_rate": 1.9378677995036564e-05, "loss": 0.9292, "step": 570 }, { "epoch": 0.4201618837380427, "grad_norm": 1.7815535628128507, "learning_rate": 1.9375918054802836e-05, "loss": 0.9603, "step": 571 }, { "epoch": 0.42089771891096395, "grad_norm": 1.9141058039417174, "learning_rate": 1.9373152195680106e-05, "loss": 0.8958, "step": 572 }, { "epoch": 0.4216335540838852, "grad_norm": 2.0159929575734967, "learning_rate": 1.9370380419414427e-05, "loss": 0.9362, "step": 573 }, { "epoch": 0.42236938925680645, "grad_norm": 1.814833717041072, "learning_rate": 1.936760272775558e-05, "loss": 0.8814, "step": 574 }, { "epoch": 0.42310522442972776, "grad_norm": 1.8249237454253078, "learning_rate": 1.9364819122457085e-05, "loss": 0.9031, "step": 575 }, { "epoch": 0.423841059602649, "grad_norm": 1.8896037666768113, "learning_rate": 1.9362029605276193e-05, "loss": 0.9437, "step": 576 }, { "epoch": 0.42457689477557026, "grad_norm": 1.7986353784404274, "learning_rate": 1.935923417797389e-05, "loss": 0.907, "step": 577 }, { "epoch": 0.42531272994849156, "grad_norm": 1.887989544973578, "learning_rate": 1.9356432842314886e-05, "loss": 0.9958, "step": 578 }, { "epoch": 0.4260485651214128, "grad_norm": 1.918543954583046, "learning_rate": 1.9353625600067632e-05, "loss": 0.9339, "step": 579 }, { "epoch": 0.42678440029433407, "grad_norm": 1.8892851634335837, "learning_rate": 1.9350812453004303e-05, "loss": 0.9731, "step": 580 }, { "epoch": 0.4275202354672553, "grad_norm": 1.7035889023509456, "learning_rate": 1.9347993402900792e-05, "loss": 0.9276, "step": 581 }, { "epoch": 0.4282560706401766, "grad_norm": 1.9510971147985623, "learning_rate": 1.934516845153673e-05, "loss": 0.964, "step": 582 }, { "epoch": 0.4289919058130979, "grad_norm": 1.7768907667183098, "learning_rate": 1.934233760069548e-05, "loss": 0.9216, "step": 583 }, { "epoch": 0.4297277409860191, "grad_norm": 1.7511256785295877, "learning_rate": 1.9339500852164104e-05, "loss": 0.9149, "step": 584 }, { "epoch": 0.4304635761589404, "grad_norm": 1.7976851728458334, "learning_rate": 1.933665820773341e-05, "loss": 0.8791, "step": 585 }, { "epoch": 0.4311994113318617, "grad_norm": 1.8039863942871988, "learning_rate": 1.933380966919792e-05, "loss": 0.9468, "step": 586 }, { "epoch": 0.43193524650478293, "grad_norm": 1.858563452202949, "learning_rate": 1.9330955238355883e-05, "loss": 0.9118, "step": 587 }, { "epoch": 0.4326710816777042, "grad_norm": 1.7072852036576611, "learning_rate": 1.9328094917009252e-05, "loss": 0.8642, "step": 588 }, { "epoch": 0.43340691685062543, "grad_norm": 1.840335724869349, "learning_rate": 1.9325228706963716e-05, "loss": 0.9381, "step": 589 }, { "epoch": 0.43414275202354674, "grad_norm": 1.8748116182276104, "learning_rate": 1.9322356610028673e-05, "loss": 0.9683, "step": 590 }, { "epoch": 0.434878587196468, "grad_norm": 1.8044832906265411, "learning_rate": 1.9319478628017237e-05, "loss": 0.9102, "step": 591 }, { "epoch": 0.43561442236938924, "grad_norm": 1.8114197072274565, "learning_rate": 1.9316594762746238e-05, "loss": 0.8785, "step": 592 }, { "epoch": 0.43635025754231055, "grad_norm": 1.800708982010269, "learning_rate": 1.9313705016036222e-05, "loss": 0.9054, "step": 593 }, { "epoch": 0.4370860927152318, "grad_norm": 1.8951038186460685, "learning_rate": 1.9310809389711446e-05, "loss": 0.922, "step": 594 }, { "epoch": 0.43782192788815305, "grad_norm": 1.8901583753270095, "learning_rate": 1.9307907885599883e-05, "loss": 0.8705, "step": 595 }, { "epoch": 0.4385577630610743, "grad_norm": 1.7717878973210721, "learning_rate": 1.930500050553321e-05, "loss": 0.9386, "step": 596 }, { "epoch": 0.4392935982339956, "grad_norm": 1.8245337981683667, "learning_rate": 1.9302087251346812e-05, "loss": 0.9535, "step": 597 }, { "epoch": 0.44002943340691686, "grad_norm": 1.6560653975815365, "learning_rate": 1.9299168124879798e-05, "loss": 0.9032, "step": 598 }, { "epoch": 0.4407652685798381, "grad_norm": 1.7022215068262343, "learning_rate": 1.929624312797496e-05, "loss": 0.9088, "step": 599 }, { "epoch": 0.44150110375275936, "grad_norm": 1.7723661504569475, "learning_rate": 1.9293312262478817e-05, "loss": 0.9062, "step": 600 }, { "epoch": 0.44223693892568067, "grad_norm": 1.8693085268415528, "learning_rate": 1.9290375530241577e-05, "loss": 0.9516, "step": 601 }, { "epoch": 0.4429727740986019, "grad_norm": 1.7398730245435485, "learning_rate": 1.928743293311717e-05, "loss": 0.8757, "step": 602 }, { "epoch": 0.44370860927152317, "grad_norm": 1.8911605678372523, "learning_rate": 1.928448447296321e-05, "loss": 0.9538, "step": 603 }, { "epoch": 0.4444444444444444, "grad_norm": 1.6890527072031125, "learning_rate": 1.9281530151641016e-05, "loss": 0.8719, "step": 604 }, { "epoch": 0.4451802796173657, "grad_norm": 1.9224261893388497, "learning_rate": 1.9278569971015617e-05, "loss": 0.9397, "step": 605 }, { "epoch": 0.445916114790287, "grad_norm": 1.87249619418816, "learning_rate": 1.9275603932955733e-05, "loss": 0.8599, "step": 606 }, { "epoch": 0.44665194996320823, "grad_norm": 1.8545458295045802, "learning_rate": 1.9272632039333784e-05, "loss": 0.8983, "step": 607 }, { "epoch": 0.44738778513612953, "grad_norm": 1.8887085725198027, "learning_rate": 1.9269654292025885e-05, "loss": 0.9179, "step": 608 }, { "epoch": 0.4481236203090508, "grad_norm": 1.9181020673786853, "learning_rate": 1.9266670692911848e-05, "loss": 0.934, "step": 609 }, { "epoch": 0.44885945548197204, "grad_norm": 1.909613347558481, "learning_rate": 1.9263681243875173e-05, "loss": 0.9197, "step": 610 }, { "epoch": 0.4495952906548933, "grad_norm": 1.8344810487260934, "learning_rate": 1.9260685946803067e-05, "loss": 0.9571, "step": 611 }, { "epoch": 0.4503311258278146, "grad_norm": 1.8170204593959955, "learning_rate": 1.9257684803586415e-05, "loss": 0.961, "step": 612 }, { "epoch": 0.45106696100073584, "grad_norm": 1.7760398825711987, "learning_rate": 1.92546778161198e-05, "loss": 0.8725, "step": 613 }, { "epoch": 0.4518027961736571, "grad_norm": 2.0073920177676547, "learning_rate": 1.9251664986301487e-05, "loss": 0.9287, "step": 614 }, { "epoch": 0.45253863134657835, "grad_norm": 1.8767913908454221, "learning_rate": 1.9248646316033444e-05, "loss": 0.9551, "step": 615 }, { "epoch": 0.45327446651949965, "grad_norm": 1.8275192809083611, "learning_rate": 1.9245621807221306e-05, "loss": 0.9437, "step": 616 }, { "epoch": 0.4540103016924209, "grad_norm": 1.8415022210913092, "learning_rate": 1.924259146177441e-05, "loss": 0.9679, "step": 617 }, { "epoch": 0.45474613686534215, "grad_norm": 1.8845379359302563, "learning_rate": 1.923955528160577e-05, "loss": 0.926, "step": 618 }, { "epoch": 0.4554819720382634, "grad_norm": 1.6954765992979466, "learning_rate": 1.9236513268632085e-05, "loss": 0.8604, "step": 619 }, { "epoch": 0.4562178072111847, "grad_norm": 1.7773485102869517, "learning_rate": 1.9233465424773737e-05, "loss": 0.9445, "step": 620 }, { "epoch": 0.45695364238410596, "grad_norm": 1.7806672386269224, "learning_rate": 1.9230411751954788e-05, "loss": 0.9132, "step": 621 }, { "epoch": 0.4576894775570272, "grad_norm": 1.6936402972863482, "learning_rate": 1.922735225210298e-05, "loss": 0.9801, "step": 622 }, { "epoch": 0.45842531272994846, "grad_norm": 1.718321280261051, "learning_rate": 1.922428692714973e-05, "loss": 0.9036, "step": 623 }, { "epoch": 0.45916114790286977, "grad_norm": 1.6760720282132584, "learning_rate": 1.9221215779030143e-05, "loss": 0.9236, "step": 624 }, { "epoch": 0.459896983075791, "grad_norm": 1.8867254079266473, "learning_rate": 1.9218138809682988e-05, "loss": 0.9553, "step": 625 }, { "epoch": 0.4606328182487123, "grad_norm": 1.8927969307621109, "learning_rate": 1.921505602105071e-05, "loss": 0.9268, "step": 626 }, { "epoch": 0.4613686534216336, "grad_norm": 1.768620364567468, "learning_rate": 1.921196741507944e-05, "loss": 0.9314, "step": 627 }, { "epoch": 0.46210448859455483, "grad_norm": 1.827822568494843, "learning_rate": 1.9208872993718967e-05, "loss": 0.931, "step": 628 }, { "epoch": 0.4628403237674761, "grad_norm": 1.9214924979615995, "learning_rate": 1.920577275892276e-05, "loss": 0.9564, "step": 629 }, { "epoch": 0.46357615894039733, "grad_norm": 1.8699433295154488, "learning_rate": 1.9202666712647953e-05, "loss": 0.9516, "step": 630 }, { "epoch": 0.46431199411331864, "grad_norm": 1.7685511691734337, "learning_rate": 1.919955485685535e-05, "loss": 0.9349, "step": 631 }, { "epoch": 0.4650478292862399, "grad_norm": 1.7580876571081694, "learning_rate": 1.9196437193509426e-05, "loss": 0.91, "step": 632 }, { "epoch": 0.46578366445916114, "grad_norm": 1.884928960772374, "learning_rate": 1.9193313724578316e-05, "loss": 0.9483, "step": 633 }, { "epoch": 0.4665194996320824, "grad_norm": 1.8603940149119056, "learning_rate": 1.9190184452033828e-05, "loss": 0.9909, "step": 634 }, { "epoch": 0.4672553348050037, "grad_norm": 1.8589357881568158, "learning_rate": 1.918704937785142e-05, "loss": 0.9359, "step": 635 }, { "epoch": 0.46799116997792495, "grad_norm": 1.7591447665104898, "learning_rate": 1.918390850401023e-05, "loss": 0.9656, "step": 636 }, { "epoch": 0.4687270051508462, "grad_norm": 2.0064405949609885, "learning_rate": 1.9180761832493045e-05, "loss": 0.9457, "step": 637 }, { "epoch": 0.46946284032376745, "grad_norm": 1.7665914456377785, "learning_rate": 1.9177609365286318e-05, "loss": 0.9043, "step": 638 }, { "epoch": 0.47019867549668876, "grad_norm": 1.758502940759409, "learning_rate": 1.9174451104380155e-05, "loss": 0.8839, "step": 639 }, { "epoch": 0.47093451066961, "grad_norm": 1.8273927835656736, "learning_rate": 1.917128705176833e-05, "loss": 0.947, "step": 640 }, { "epoch": 0.47167034584253126, "grad_norm": 1.737019092057898, "learning_rate": 1.916811720944826e-05, "loss": 0.8953, "step": 641 }, { "epoch": 0.47240618101545256, "grad_norm": 1.6973099440251174, "learning_rate": 1.9164941579421018e-05, "loss": 0.9127, "step": 642 }, { "epoch": 0.4731420161883738, "grad_norm": 1.7338072775122957, "learning_rate": 1.9161760163691347e-05, "loss": 0.9116, "step": 643 }, { "epoch": 0.47387785136129507, "grad_norm": 1.7790062298011433, "learning_rate": 1.9158572964267626e-05, "loss": 0.9272, "step": 644 }, { "epoch": 0.4746136865342163, "grad_norm": 1.9007692881330618, "learning_rate": 1.9155379983161895e-05, "loss": 0.9166, "step": 645 }, { "epoch": 0.4753495217071376, "grad_norm": 1.734145262249907, "learning_rate": 1.915218122238983e-05, "loss": 0.8589, "step": 646 }, { "epoch": 0.4760853568800589, "grad_norm": 1.8919518084497253, "learning_rate": 1.9148976683970774e-05, "loss": 0.9349, "step": 647 }, { "epoch": 0.4768211920529801, "grad_norm": 1.858858713237949, "learning_rate": 1.91457663699277e-05, "loss": 0.924, "step": 648 }, { "epoch": 0.4775570272259014, "grad_norm": 1.8247060894465648, "learning_rate": 1.9142550282287247e-05, "loss": 0.929, "step": 649 }, { "epoch": 0.4782928623988227, "grad_norm": 1.7518480503401153, "learning_rate": 1.9139328423079674e-05, "loss": 0.9196, "step": 650 }, { "epoch": 0.47902869757174393, "grad_norm": 1.8075998192390434, "learning_rate": 1.9136100794338907e-05, "loss": 0.9828, "step": 651 }, { "epoch": 0.4797645327446652, "grad_norm": 2.0774754392550197, "learning_rate": 1.9132867398102498e-05, "loss": 0.8761, "step": 652 }, { "epoch": 0.48050036791758644, "grad_norm": 1.837231777858835, "learning_rate": 1.912962823641165e-05, "loss": 0.8847, "step": 653 }, { "epoch": 0.48123620309050774, "grad_norm": 1.7755370889463498, "learning_rate": 1.91263833113112e-05, "loss": 0.9383, "step": 654 }, { "epoch": 0.481972038263429, "grad_norm": 1.7897218724402315, "learning_rate": 1.912313262484962e-05, "loss": 0.8756, "step": 655 }, { "epoch": 0.48270787343635024, "grad_norm": 1.8971674092531852, "learning_rate": 1.9119876179079032e-05, "loss": 0.9416, "step": 656 }, { "epoch": 0.48344370860927155, "grad_norm": 1.8097128589662772, "learning_rate": 1.911661397605518e-05, "loss": 0.9544, "step": 657 }, { "epoch": 0.4841795437821928, "grad_norm": 1.782447397592442, "learning_rate": 1.911334601783745e-05, "loss": 0.9288, "step": 658 }, { "epoch": 0.48491537895511405, "grad_norm": 1.7767210108022211, "learning_rate": 1.9110072306488862e-05, "loss": 0.9118, "step": 659 }, { "epoch": 0.4856512141280353, "grad_norm": 1.7907908509760848, "learning_rate": 1.910679284407606e-05, "loss": 0.9917, "step": 660 }, { "epoch": 0.4863870493009566, "grad_norm": 1.7324225695972038, "learning_rate": 1.910350763266933e-05, "loss": 0.9616, "step": 661 }, { "epoch": 0.48712288447387786, "grad_norm": 1.8460315936385159, "learning_rate": 1.9100216674342575e-05, "loss": 0.8992, "step": 662 }, { "epoch": 0.4878587196467991, "grad_norm": 1.7454861882311805, "learning_rate": 1.9096919971173336e-05, "loss": 0.8906, "step": 663 }, { "epoch": 0.48859455481972036, "grad_norm": 1.805711676292063, "learning_rate": 1.9093617525242772e-05, "loss": 0.9163, "step": 664 }, { "epoch": 0.48933038999264167, "grad_norm": 1.7315525899998578, "learning_rate": 1.909030933863568e-05, "loss": 0.9212, "step": 665 }, { "epoch": 0.4900662251655629, "grad_norm": 1.7257638628152343, "learning_rate": 1.908699541344047e-05, "loss": 0.9106, "step": 666 }, { "epoch": 0.49080206033848417, "grad_norm": 1.7292440347207856, "learning_rate": 1.9083675751749174e-05, "loss": 0.909, "step": 667 }, { "epoch": 0.4915378955114054, "grad_norm": 1.7974711560816725, "learning_rate": 1.908035035565745e-05, "loss": 0.9318, "step": 668 }, { "epoch": 0.4922737306843267, "grad_norm": 1.9487551762269577, "learning_rate": 1.9077019227264573e-05, "loss": 0.8635, "step": 669 }, { "epoch": 0.493009565857248, "grad_norm": 1.8058379252228474, "learning_rate": 1.907368236867345e-05, "loss": 0.8563, "step": 670 }, { "epoch": 0.49374540103016923, "grad_norm": 1.7602834128778306, "learning_rate": 1.907033978199058e-05, "loss": 0.9117, "step": 671 }, { "epoch": 0.49448123620309054, "grad_norm": 1.763022358328617, "learning_rate": 1.9066991469326103e-05, "loss": 0.9021, "step": 672 }, { "epoch": 0.4952170713760118, "grad_norm": 1.8455891477667328, "learning_rate": 1.9063637432793757e-05, "loss": 0.8893, "step": 673 }, { "epoch": 0.49595290654893304, "grad_norm": 1.7784026149319276, "learning_rate": 1.9060277674510904e-05, "loss": 0.9044, "step": 674 }, { "epoch": 0.4966887417218543, "grad_norm": 1.7024158778647673, "learning_rate": 1.905691219659851e-05, "loss": 0.9054, "step": 675 }, { "epoch": 0.4974245768947756, "grad_norm": 1.7985095107473885, "learning_rate": 1.9053541001181156e-05, "loss": 0.9229, "step": 676 }, { "epoch": 0.49816041206769685, "grad_norm": 1.7576158789893555, "learning_rate": 1.905016409038703e-05, "loss": 0.9258, "step": 677 }, { "epoch": 0.4988962472406181, "grad_norm": 1.7534793430088662, "learning_rate": 1.9046781466347935e-05, "loss": 0.9047, "step": 678 }, { "epoch": 0.49963208241353935, "grad_norm": 1.6130031233490505, "learning_rate": 1.9043393131199266e-05, "loss": 0.9096, "step": 679 }, { "epoch": 0.5003679175864606, "grad_norm": 1.817079244889132, "learning_rate": 1.9039999087080043e-05, "loss": 0.9018, "step": 680 }, { "epoch": 0.5011037527593819, "grad_norm": 1.7458259115415051, "learning_rate": 1.9036599336132876e-05, "loss": 0.898, "step": 681 }, { "epoch": 0.5018395879323032, "grad_norm": 1.6357182206630028, "learning_rate": 1.9033193880503976e-05, "loss": 0.9232, "step": 682 }, { "epoch": 0.5025754231052244, "grad_norm": 1.7950882889230875, "learning_rate": 1.9029782722343168e-05, "loss": 0.9824, "step": 683 }, { "epoch": 0.5033112582781457, "grad_norm": 1.6529413831872013, "learning_rate": 1.9026365863803863e-05, "loss": 0.8998, "step": 684 }, { "epoch": 0.5040470934510669, "grad_norm": 1.8408370470570072, "learning_rate": 1.9022943307043085e-05, "loss": 0.8638, "step": 685 }, { "epoch": 0.5047829286239882, "grad_norm": 1.7926674665741325, "learning_rate": 1.9019515054221443e-05, "loss": 0.9023, "step": 686 }, { "epoch": 0.5055187637969095, "grad_norm": 1.765173763114253, "learning_rate": 1.9016081107503145e-05, "loss": 0.9423, "step": 687 }, { "epoch": 0.5062545989698307, "grad_norm": 1.9260876382401408, "learning_rate": 1.9012641469055998e-05, "loss": 0.957, "step": 688 }, { "epoch": 0.506990434142752, "grad_norm": 1.8031956198907844, "learning_rate": 1.9009196141051392e-05, "loss": 0.9421, "step": 689 }, { "epoch": 0.5077262693156733, "grad_norm": 1.8570652118827466, "learning_rate": 1.9005745125664324e-05, "loss": 0.9295, "step": 690 }, { "epoch": 0.5084621044885945, "grad_norm": 1.7938463439794154, "learning_rate": 1.9002288425073367e-05, "loss": 0.9101, "step": 691 }, { "epoch": 0.5091979396615158, "grad_norm": 1.7719740299256987, "learning_rate": 1.899882604146069e-05, "loss": 0.8947, "step": 692 }, { "epoch": 0.5099337748344371, "grad_norm": 1.9707386295821665, "learning_rate": 1.8995357977012054e-05, "loss": 0.952, "step": 693 }, { "epoch": 0.5106696100073583, "grad_norm": 1.7200065029847882, "learning_rate": 1.8991884233916795e-05, "loss": 0.8846, "step": 694 }, { "epoch": 0.5114054451802796, "grad_norm": 1.7122197490294946, "learning_rate": 1.8988404814367837e-05, "loss": 0.9271, "step": 695 }, { "epoch": 0.5121412803532008, "grad_norm": 1.8204585814134358, "learning_rate": 1.89849197205617e-05, "loss": 0.9019, "step": 696 }, { "epoch": 0.5128771155261221, "grad_norm": 1.7248727489048938, "learning_rate": 1.8981428954698466e-05, "loss": 0.9168, "step": 697 }, { "epoch": 0.5136129506990434, "grad_norm": 1.7115533200547102, "learning_rate": 1.897793251898181e-05, "loss": 0.9063, "step": 698 }, { "epoch": 0.5143487858719646, "grad_norm": 1.7334606447648118, "learning_rate": 1.8974430415618987e-05, "loss": 0.9219, "step": 699 }, { "epoch": 0.515084621044886, "grad_norm": 1.9518678364879927, "learning_rate": 1.8970922646820825e-05, "loss": 0.9505, "step": 700 }, { "epoch": 0.5158204562178073, "grad_norm": 1.907140994082829, "learning_rate": 1.8967409214801737e-05, "loss": 0.8923, "step": 701 }, { "epoch": 0.5165562913907285, "grad_norm": 1.7477364594736575, "learning_rate": 1.896389012177969e-05, "loss": 0.8742, "step": 702 }, { "epoch": 0.5172921265636498, "grad_norm": 1.929462265731512, "learning_rate": 1.8960365369976256e-05, "loss": 0.8859, "step": 703 }, { "epoch": 0.5180279617365711, "grad_norm": 1.9035889626427007, "learning_rate": 1.895683496161655e-05, "loss": 0.8828, "step": 704 }, { "epoch": 0.5187637969094923, "grad_norm": 1.8137078479179929, "learning_rate": 1.895329889892928e-05, "loss": 0.8895, "step": 705 }, { "epoch": 0.5194996320824136, "grad_norm": 1.8318248677708597, "learning_rate": 1.8949757184146706e-05, "loss": 0.9079, "step": 706 }, { "epoch": 0.5202354672553348, "grad_norm": 1.9002746690907106, "learning_rate": 1.894620981950467e-05, "loss": 0.9615, "step": 707 }, { "epoch": 0.5209713024282561, "grad_norm": 1.730629165482707, "learning_rate": 1.894265680724257e-05, "loss": 0.9116, "step": 708 }, { "epoch": 0.5217071376011774, "grad_norm": 1.7516993257297155, "learning_rate": 1.893909814960338e-05, "loss": 0.9286, "step": 709 }, { "epoch": 0.5224429727740986, "grad_norm": 1.8221559650521686, "learning_rate": 1.8935533848833626e-05, "loss": 0.9116, "step": 710 }, { "epoch": 0.5231788079470199, "grad_norm": 1.7183723915190863, "learning_rate": 1.8931963907183406e-05, "loss": 0.9294, "step": 711 }, { "epoch": 0.5239146431199412, "grad_norm": 1.8026555440966985, "learning_rate": 1.8928388326906376e-05, "loss": 0.9052, "step": 712 }, { "epoch": 0.5246504782928624, "grad_norm": 1.7540008369514521, "learning_rate": 1.892480711025975e-05, "loss": 0.9164, "step": 713 }, { "epoch": 0.5253863134657837, "grad_norm": 1.8267437355732625, "learning_rate": 1.8921220259504297e-05, "loss": 0.9276, "step": 714 }, { "epoch": 0.5261221486387049, "grad_norm": 1.8426623637364203, "learning_rate": 1.8917627776904352e-05, "loss": 0.9909, "step": 715 }, { "epoch": 0.5268579838116262, "grad_norm": 1.6766366246826991, "learning_rate": 1.8914029664727803e-05, "loss": 0.8948, "step": 716 }, { "epoch": 0.5275938189845475, "grad_norm": 1.7185969095678189, "learning_rate": 1.891042592524608e-05, "loss": 0.9336, "step": 717 }, { "epoch": 0.5283296541574687, "grad_norm": 1.8201068441629522, "learning_rate": 1.8906816560734182e-05, "loss": 0.8983, "step": 718 }, { "epoch": 0.52906548933039, "grad_norm": 1.8633755063017456, "learning_rate": 1.8903201573470645e-05, "loss": 0.96, "step": 719 }, { "epoch": 0.5298013245033113, "grad_norm": 1.755837904102719, "learning_rate": 1.889958096573757e-05, "loss": 0.9281, "step": 720 }, { "epoch": 0.5305371596762325, "grad_norm": 1.753603736160547, "learning_rate": 1.889595473982059e-05, "loss": 0.907, "step": 721 }, { "epoch": 0.5312729948491538, "grad_norm": 1.8173745381864332, "learning_rate": 1.8892322898008893e-05, "loss": 0.9389, "step": 722 }, { "epoch": 0.5320088300220751, "grad_norm": 1.8408784873740462, "learning_rate": 1.8888685442595213e-05, "loss": 0.8842, "step": 723 }, { "epoch": 0.5327446651949963, "grad_norm": 1.8714418007993985, "learning_rate": 1.8885042375875825e-05, "loss": 0.9352, "step": 724 }, { "epoch": 0.5334805003679176, "grad_norm": 1.8878908864353927, "learning_rate": 1.888139370015055e-05, "loss": 0.8899, "step": 725 }, { "epoch": 0.5342163355408388, "grad_norm": 1.8612647867331715, "learning_rate": 1.887773941772274e-05, "loss": 0.895, "step": 726 }, { "epoch": 0.5349521707137601, "grad_norm": 1.7832960541723681, "learning_rate": 1.8874079530899298e-05, "loss": 0.9065, "step": 727 }, { "epoch": 0.5356880058866814, "grad_norm": 1.801067318350488, "learning_rate": 1.8870414041990662e-05, "loss": 0.8219, "step": 728 }, { "epoch": 0.5364238410596026, "grad_norm": 1.7745036849824285, "learning_rate": 1.8866742953310806e-05, "loss": 0.9139, "step": 729 }, { "epoch": 0.5371596762325239, "grad_norm": 1.7778134014956946, "learning_rate": 1.8863066267177234e-05, "loss": 0.9761, "step": 730 }, { "epoch": 0.5378955114054452, "grad_norm": 1.780182731547408, "learning_rate": 1.885938398591099e-05, "loss": 0.8887, "step": 731 }, { "epoch": 0.5386313465783664, "grad_norm": 1.7761202347328107, "learning_rate": 1.8855696111836645e-05, "loss": 0.9541, "step": 732 }, { "epoch": 0.5393671817512877, "grad_norm": 1.7290365777280763, "learning_rate": 1.885200264728231e-05, "loss": 0.8819, "step": 733 }, { "epoch": 0.5401030169242089, "grad_norm": 1.8049450941910286, "learning_rate": 1.8848303594579614e-05, "loss": 0.8967, "step": 734 }, { "epoch": 0.5408388520971302, "grad_norm": 1.70879733134883, "learning_rate": 1.8844598956063716e-05, "loss": 0.8486, "step": 735 }, { "epoch": 0.5415746872700515, "grad_norm": 2.0014124817412884, "learning_rate": 1.884088873407331e-05, "loss": 0.8662, "step": 736 }, { "epoch": 0.5423105224429727, "grad_norm": 1.8426425786670062, "learning_rate": 1.8837172930950602e-05, "loss": 0.9324, "step": 737 }, { "epoch": 0.543046357615894, "grad_norm": 1.819302369733713, "learning_rate": 1.8833451549041337e-05, "loss": 0.9324, "step": 738 }, { "epoch": 0.5437821927888153, "grad_norm": 1.8841808381782028, "learning_rate": 1.882972459069476e-05, "loss": 0.9823, "step": 739 }, { "epoch": 0.5445180279617365, "grad_norm": 1.7589045264160812, "learning_rate": 1.8825992058263656e-05, "loss": 0.8966, "step": 740 }, { "epoch": 0.5452538631346578, "grad_norm": 1.8043905384957963, "learning_rate": 1.8822253954104325e-05, "loss": 0.869, "step": 741 }, { "epoch": 0.5459896983075792, "grad_norm": 1.7587301739902648, "learning_rate": 1.8818510280576577e-05, "loss": 0.9137, "step": 742 }, { "epoch": 0.5467255334805003, "grad_norm": 1.6514800074063138, "learning_rate": 1.881476104004374e-05, "loss": 0.8792, "step": 743 }, { "epoch": 0.5474613686534217, "grad_norm": 1.7899338500452497, "learning_rate": 1.8811006234872665e-05, "loss": 0.9596, "step": 744 }, { "epoch": 0.5481972038263428, "grad_norm": 1.8009359321889098, "learning_rate": 1.88072458674337e-05, "loss": 0.9642, "step": 745 }, { "epoch": 0.5489330389992642, "grad_norm": 1.9174385439694683, "learning_rate": 1.8803479940100723e-05, "loss": 0.9031, "step": 746 }, { "epoch": 0.5496688741721855, "grad_norm": 1.7650008616778539, "learning_rate": 1.8799708455251103e-05, "loss": 0.96, "step": 747 }, { "epoch": 0.5504047093451067, "grad_norm": 1.708520624768988, "learning_rate": 1.8795931415265735e-05, "loss": 0.8784, "step": 748 }, { "epoch": 0.551140544518028, "grad_norm": 1.8498569203776378, "learning_rate": 1.879214882252901e-05, "loss": 0.891, "step": 749 }, { "epoch": 0.5518763796909493, "grad_norm": 1.7378902756964576, "learning_rate": 1.8788360679428822e-05, "loss": 0.9779, "step": 750 }, { "epoch": 0.5526122148638705, "grad_norm": 1.6673918447955127, "learning_rate": 1.8784566988356586e-05, "loss": 0.9407, "step": 751 }, { "epoch": 0.5533480500367918, "grad_norm": 1.6531132215960604, "learning_rate": 1.8780767751707193e-05, "loss": 0.9402, "step": 752 }, { "epoch": 0.5540838852097131, "grad_norm": 1.7520239468578516, "learning_rate": 1.8776962971879064e-05, "loss": 0.9108, "step": 753 }, { "epoch": 0.5548197203826343, "grad_norm": 1.7306163379171813, "learning_rate": 1.877315265127409e-05, "loss": 0.8904, "step": 754 }, { "epoch": 0.5555555555555556, "grad_norm": 1.697361119358812, "learning_rate": 1.8769336792297683e-05, "loss": 0.9112, "step": 755 }, { "epoch": 0.5562913907284768, "grad_norm": 1.70175181126646, "learning_rate": 1.8765515397358744e-05, "loss": 0.8653, "step": 756 }, { "epoch": 0.5570272259013981, "grad_norm": 1.7114001597544914, "learning_rate": 1.8761688468869658e-05, "loss": 0.9329, "step": 757 }, { "epoch": 0.5577630610743194, "grad_norm": 1.7175390803946604, "learning_rate": 1.8757856009246325e-05, "loss": 0.8255, "step": 758 }, { "epoch": 0.5584988962472406, "grad_norm": 1.7514122699790298, "learning_rate": 1.8754018020908113e-05, "loss": 0.9215, "step": 759 }, { "epoch": 0.5592347314201619, "grad_norm": 1.8686850399511217, "learning_rate": 1.8750174506277902e-05, "loss": 0.9402, "step": 760 }, { "epoch": 0.5599705665930832, "grad_norm": 1.8276931860538548, "learning_rate": 1.874632546778204e-05, "loss": 0.9309, "step": 761 }, { "epoch": 0.5607064017660044, "grad_norm": 1.929267963258629, "learning_rate": 1.8742470907850384e-05, "loss": 0.8404, "step": 762 }, { "epoch": 0.5614422369389257, "grad_norm": 1.7857734654885027, "learning_rate": 1.8738610828916255e-05, "loss": 0.954, "step": 763 }, { "epoch": 0.5621780721118469, "grad_norm": 1.6812611894081955, "learning_rate": 1.8734745233416475e-05, "loss": 0.9044, "step": 764 }, { "epoch": 0.5629139072847682, "grad_norm": 1.6832808344221268, "learning_rate": 1.8730874123791334e-05, "loss": 0.9026, "step": 765 }, { "epoch": 0.5636497424576895, "grad_norm": 1.6701047837059726, "learning_rate": 1.8726997502484617e-05, "loss": 0.8624, "step": 766 }, { "epoch": 0.5643855776306107, "grad_norm": 1.6318124963275, "learning_rate": 1.872311537194358e-05, "loss": 0.8734, "step": 767 }, { "epoch": 0.565121412803532, "grad_norm": 1.806108702331251, "learning_rate": 1.8719227734618955e-05, "loss": 0.9341, "step": 768 }, { "epoch": 0.5658572479764533, "grad_norm": 1.6726013004262954, "learning_rate": 1.8715334592964964e-05, "loss": 0.9051, "step": 769 }, { "epoch": 0.5665930831493745, "grad_norm": 1.7371875375220382, "learning_rate": 1.871143594943929e-05, "loss": 0.9216, "step": 770 }, { "epoch": 0.5673289183222958, "grad_norm": 1.9160271155716342, "learning_rate": 1.8707531806503086e-05, "loss": 0.8989, "step": 771 }, { "epoch": 0.5680647534952171, "grad_norm": 1.659713760472978, "learning_rate": 1.8703622166620995e-05, "loss": 0.8932, "step": 772 }, { "epoch": 0.5688005886681383, "grad_norm": 1.7377095542416774, "learning_rate": 1.8699707032261112e-05, "loss": 0.8978, "step": 773 }, { "epoch": 0.5695364238410596, "grad_norm": 1.72348913153184, "learning_rate": 1.8695786405895013e-05, "loss": 0.9306, "step": 774 }, { "epoch": 0.5702722590139808, "grad_norm": 1.6924952937505697, "learning_rate": 1.869186028999773e-05, "loss": 0.8513, "step": 775 }, { "epoch": 0.5710080941869021, "grad_norm": 1.98388698380918, "learning_rate": 1.868792868704777e-05, "loss": 0.9473, "step": 776 }, { "epoch": 0.5717439293598234, "grad_norm": 1.847833743098229, "learning_rate": 1.86839915995271e-05, "loss": 0.8879, "step": 777 }, { "epoch": 0.5724797645327446, "grad_norm": 1.694162709244805, "learning_rate": 1.868004902992115e-05, "loss": 0.8913, "step": 778 }, { "epoch": 0.5732155997056659, "grad_norm": 1.820893333808457, "learning_rate": 1.867610098071881e-05, "loss": 0.9312, "step": 779 }, { "epoch": 0.5739514348785872, "grad_norm": 1.732807216614163, "learning_rate": 1.867214745441243e-05, "loss": 0.9609, "step": 780 }, { "epoch": 0.5746872700515084, "grad_norm": 1.997654869967021, "learning_rate": 1.8668188453497814e-05, "loss": 0.925, "step": 781 }, { "epoch": 0.5754231052244297, "grad_norm": 1.7629325240895275, "learning_rate": 1.8664223980474226e-05, "loss": 0.9219, "step": 782 }, { "epoch": 0.5761589403973509, "grad_norm": 1.6684293989285983, "learning_rate": 1.866025403784439e-05, "loss": 0.9237, "step": 783 }, { "epoch": 0.5768947755702722, "grad_norm": 1.7368137723701447, "learning_rate": 1.865627862811447e-05, "loss": 0.8668, "step": 784 }, { "epoch": 0.5776306107431936, "grad_norm": 1.7494348843616063, "learning_rate": 1.8652297753794087e-05, "loss": 0.9001, "step": 785 }, { "epoch": 0.5783664459161147, "grad_norm": 1.7835632828910046, "learning_rate": 1.8648311417396314e-05, "loss": 0.9026, "step": 786 }, { "epoch": 0.579102281089036, "grad_norm": 1.6510261141492453, "learning_rate": 1.8644319621437677e-05, "loss": 0.8888, "step": 787 }, { "epoch": 0.5798381162619574, "grad_norm": 1.7690355105889088, "learning_rate": 1.8640322368438134e-05, "loss": 0.8648, "step": 788 }, { "epoch": 0.5805739514348786, "grad_norm": 1.8031156631347425, "learning_rate": 1.8636319660921098e-05, "loss": 0.9161, "step": 789 }, { "epoch": 0.5813097866077999, "grad_norm": 1.769311417425349, "learning_rate": 1.863231150141343e-05, "loss": 0.9004, "step": 790 }, { "epoch": 0.5820456217807212, "grad_norm": 1.6945605977074791, "learning_rate": 1.8628297892445417e-05, "loss": 0.9692, "step": 791 }, { "epoch": 0.5827814569536424, "grad_norm": 1.7369773143823, "learning_rate": 1.8624278836550803e-05, "loss": 0.9642, "step": 792 }, { "epoch": 0.5835172921265637, "grad_norm": 1.71683422344146, "learning_rate": 1.8620254336266757e-05, "loss": 0.9501, "step": 793 }, { "epoch": 0.5842531272994849, "grad_norm": 1.7962869255818152, "learning_rate": 1.8616224394133896e-05, "loss": 0.8742, "step": 794 }, { "epoch": 0.5849889624724062, "grad_norm": 1.8689197398600683, "learning_rate": 1.8612189012696263e-05, "loss": 0.981, "step": 795 }, { "epoch": 0.5857247976453275, "grad_norm": 1.7191755852163793, "learning_rate": 1.8608148194501343e-05, "loss": 0.9412, "step": 796 }, { "epoch": 0.5864606328182487, "grad_norm": 1.9942595543551032, "learning_rate": 1.860410194210005e-05, "loss": 0.8917, "step": 797 }, { "epoch": 0.58719646799117, "grad_norm": 1.7477232303806136, "learning_rate": 1.860005025804672e-05, "loss": 0.91, "step": 798 }, { "epoch": 0.5879323031640913, "grad_norm": 2.0321042149639594, "learning_rate": 1.8595993144899135e-05, "loss": 0.9822, "step": 799 }, { "epoch": 0.5886681383370125, "grad_norm": 1.772000024774547, "learning_rate": 1.8591930605218494e-05, "loss": 0.9763, "step": 800 }, { "epoch": 0.5894039735099338, "grad_norm": 1.7354266166182748, "learning_rate": 1.8587862641569417e-05, "loss": 0.8833, "step": 801 }, { "epoch": 0.5901398086828551, "grad_norm": 1.638334464927617, "learning_rate": 1.858378925651996e-05, "loss": 0.8987, "step": 802 }, { "epoch": 0.5908756438557763, "grad_norm": 1.749758656110327, "learning_rate": 1.857971045264159e-05, "loss": 0.8842, "step": 803 }, { "epoch": 0.5916114790286976, "grad_norm": 1.6351646964204207, "learning_rate": 1.8575626232509206e-05, "loss": 0.9136, "step": 804 }, { "epoch": 0.5923473142016188, "grad_norm": 1.7483210709552768, "learning_rate": 1.8571536598701114e-05, "loss": 0.9036, "step": 805 }, { "epoch": 0.5930831493745401, "grad_norm": 1.6837141580502386, "learning_rate": 1.856744155379905e-05, "loss": 0.9287, "step": 806 }, { "epoch": 0.5938189845474614, "grad_norm": 2.1805630545013406, "learning_rate": 1.8563341100388157e-05, "loss": 0.8319, "step": 807 }, { "epoch": 0.5945548197203826, "grad_norm": 1.8240101366808104, "learning_rate": 1.8559235241056994e-05, "loss": 0.9562, "step": 808 }, { "epoch": 0.5952906548933039, "grad_norm": 1.8420509580781002, "learning_rate": 1.8555123978397536e-05, "loss": 0.8682, "step": 809 }, { "epoch": 0.5960264900662252, "grad_norm": 1.7496539895971208, "learning_rate": 1.8551007315005165e-05, "loss": 0.883, "step": 810 }, { "epoch": 0.5967623252391464, "grad_norm": 1.895387398480887, "learning_rate": 1.8546885253478678e-05, "loss": 0.8986, "step": 811 }, { "epoch": 0.5974981604120677, "grad_norm": 1.718425980772248, "learning_rate": 1.8542757796420273e-05, "loss": 0.9586, "step": 812 }, { "epoch": 0.5982339955849889, "grad_norm": 1.6446148886677197, "learning_rate": 1.8538624946435555e-05, "loss": 0.8888, "step": 813 }, { "epoch": 0.5989698307579102, "grad_norm": 1.7144054503064052, "learning_rate": 1.853448670613354e-05, "loss": 0.9272, "step": 814 }, { "epoch": 0.5997056659308315, "grad_norm": 1.703630385648387, "learning_rate": 1.8530343078126645e-05, "loss": 0.9018, "step": 815 }, { "epoch": 0.6004415011037527, "grad_norm": 1.8034647023370793, "learning_rate": 1.852619406503068e-05, "loss": 0.9056, "step": 816 }, { "epoch": 0.6004415011037527, "eval_en_loss": 1.119572401046753, "eval_en_runtime": 14.2945, "eval_en_samples_per_second": 68.418, "eval_en_steps_per_second": 0.14, "step": 816 }, { "epoch": 0.6004415011037527, "eval_eu_loss": 0.9175521731376648, "eval_eu_runtime": 14.6261, "eval_eu_samples_per_second": 67.345, "eval_eu_steps_per_second": 0.137, "step": 816 }, { "epoch": 0.601177336276674, "grad_norm": 1.7694678707804954, "learning_rate": 1.8522039669464863e-05, "loss": 0.9152, "step": 817 }, { "epoch": 0.6019131714495953, "grad_norm": 1.7435007388746269, "learning_rate": 1.8517879894051804e-05, "loss": 0.9418, "step": 818 }, { "epoch": 0.6026490066225165, "grad_norm": 1.8981472261313714, "learning_rate": 1.8513714741417517e-05, "loss": 0.9249, "step": 819 }, { "epoch": 0.6033848417954378, "grad_norm": 1.8279687294485207, "learning_rate": 1.8509544214191403e-05, "loss": 0.8655, "step": 820 }, { "epoch": 0.6041206769683591, "grad_norm": 1.6951024445393148, "learning_rate": 1.8505368315006264e-05, "loss": 0.8698, "step": 821 }, { "epoch": 0.6048565121412803, "grad_norm": 1.7440793130492913, "learning_rate": 1.850118704649828e-05, "loss": 0.9291, "step": 822 }, { "epoch": 0.6055923473142016, "grad_norm": 1.746740195655706, "learning_rate": 1.8497000411307035e-05, "loss": 0.8761, "step": 823 }, { "epoch": 0.6063281824871228, "grad_norm": 1.715019647322021, "learning_rate": 1.8492808412075487e-05, "loss": 0.8724, "step": 824 }, { "epoch": 0.6070640176600441, "grad_norm": 1.6840939389144305, "learning_rate": 1.848861105145e-05, "loss": 0.835, "step": 825 }, { "epoch": 0.6077998528329654, "grad_norm": 1.6832386754720117, "learning_rate": 1.8484408332080298e-05, "loss": 0.8841, "step": 826 }, { "epoch": 0.6085356880058866, "grad_norm": 1.7718902776060905, "learning_rate": 1.8480200256619507e-05, "loss": 0.8948, "step": 827 }, { "epoch": 0.609271523178808, "grad_norm": 1.8072268174274644, "learning_rate": 1.8475986827724125e-05, "loss": 0.8929, "step": 828 }, { "epoch": 0.6100073583517293, "grad_norm": 1.6304629883115747, "learning_rate": 1.847176804805404e-05, "loss": 0.8913, "step": 829 }, { "epoch": 0.6107431935246505, "grad_norm": 1.7494968395111057, "learning_rate": 1.8467543920272498e-05, "loss": 0.9027, "step": 830 }, { "epoch": 0.6114790286975718, "grad_norm": 1.7535724107409771, "learning_rate": 1.846331444704614e-05, "loss": 0.8924, "step": 831 }, { "epoch": 0.6122148638704931, "grad_norm": 1.666750031415678, "learning_rate": 1.845907963104497e-05, "loss": 0.8806, "step": 832 }, { "epoch": 0.6129506990434143, "grad_norm": 1.7130363991130928, "learning_rate": 1.845483947494238e-05, "loss": 0.9253, "step": 833 }, { "epoch": 0.6136865342163356, "grad_norm": 1.7687800719303413, "learning_rate": 1.845059398141511e-05, "loss": 0.9349, "step": 834 }, { "epoch": 0.6144223693892568, "grad_norm": 1.7356703015205848, "learning_rate": 1.844634315314329e-05, "loss": 0.8707, "step": 835 }, { "epoch": 0.6151582045621781, "grad_norm": 1.7526367210055942, "learning_rate": 1.8442086992810407e-05, "loss": 0.9497, "step": 836 }, { "epoch": 0.6158940397350994, "grad_norm": 1.7181097201130866, "learning_rate": 1.8437825503103317e-05, "loss": 0.9277, "step": 837 }, { "epoch": 0.6166298749080206, "grad_norm": 1.9696926974240248, "learning_rate": 1.843355868671224e-05, "loss": 0.9041, "step": 838 }, { "epoch": 0.6173657100809419, "grad_norm": 1.8050681476856538, "learning_rate": 1.842928654633076e-05, "loss": 0.9067, "step": 839 }, { "epoch": 0.6181015452538632, "grad_norm": 1.7166663437448006, "learning_rate": 1.842500908465582e-05, "loss": 0.8297, "step": 840 }, { "epoch": 0.6188373804267844, "grad_norm": 1.8611662425182391, "learning_rate": 1.8420726304387723e-05, "loss": 0.8808, "step": 841 }, { "epoch": 0.6195732155997057, "grad_norm": 1.6850849064238824, "learning_rate": 1.841643820823013e-05, "loss": 0.8773, "step": 842 }, { "epoch": 0.6203090507726269, "grad_norm": 1.8492348271745245, "learning_rate": 1.841214479889006e-05, "loss": 0.8608, "step": 843 }, { "epoch": 0.6210448859455482, "grad_norm": 1.9405588458405014, "learning_rate": 1.840784607907788e-05, "loss": 0.9467, "step": 844 }, { "epoch": 0.6217807211184695, "grad_norm": 1.8151847314609755, "learning_rate": 1.8403542051507314e-05, "loss": 0.8866, "step": 845 }, { "epoch": 0.6225165562913907, "grad_norm": 1.7760452545953387, "learning_rate": 1.8399232718895435e-05, "loss": 0.9458, "step": 846 }, { "epoch": 0.623252391464312, "grad_norm": 1.7006584089494077, "learning_rate": 1.839491808396267e-05, "loss": 0.9029, "step": 847 }, { "epoch": 0.6239882266372333, "grad_norm": 1.7885509116934817, "learning_rate": 1.8390598149432784e-05, "loss": 0.9034, "step": 848 }, { "epoch": 0.6247240618101545, "grad_norm": 1.7309987186705038, "learning_rate": 1.8386272918032896e-05, "loss": 0.8896, "step": 849 }, { "epoch": 0.6254598969830758, "grad_norm": 1.7613103332773272, "learning_rate": 1.8381942392493464e-05, "loss": 0.9264, "step": 850 }, { "epoch": 0.6261957321559971, "grad_norm": 1.745185731810009, "learning_rate": 1.8377606575548288e-05, "loss": 0.9704, "step": 851 }, { "epoch": 0.6269315673289183, "grad_norm": 1.7021427401338243, "learning_rate": 1.8373265469934513e-05, "loss": 0.9339, "step": 852 }, { "epoch": 0.6276674025018396, "grad_norm": 1.6813710574755751, "learning_rate": 1.836891907839262e-05, "loss": 0.8861, "step": 853 }, { "epoch": 0.6284032376747608, "grad_norm": 1.7705052812094981, "learning_rate": 1.8364567403666426e-05, "loss": 0.9771, "step": 854 }, { "epoch": 0.6291390728476821, "grad_norm": 1.7602165719078846, "learning_rate": 1.836021044850308e-05, "loss": 0.9047, "step": 855 }, { "epoch": 0.6298749080206034, "grad_norm": 1.821816286570474, "learning_rate": 1.8355848215653073e-05, "loss": 0.9466, "step": 856 }, { "epoch": 0.6306107431935246, "grad_norm": 1.6949043179752452, "learning_rate": 1.835148070787022e-05, "loss": 0.8661, "step": 857 }, { "epoch": 0.6313465783664459, "grad_norm": 1.8075409244788812, "learning_rate": 1.8347107927911673e-05, "loss": 0.9189, "step": 858 }, { "epoch": 0.6320824135393672, "grad_norm": 1.7507023523585095, "learning_rate": 1.8342729878537903e-05, "loss": 0.8767, "step": 859 }, { "epoch": 0.6328182487122884, "grad_norm": 1.809824429108037, "learning_rate": 1.8338346562512716e-05, "loss": 0.9061, "step": 860 }, { "epoch": 0.6335540838852097, "grad_norm": 1.708677510082124, "learning_rate": 1.833395798260324e-05, "loss": 0.9135, "step": 861 }, { "epoch": 0.6342899190581309, "grad_norm": 1.697066768912668, "learning_rate": 1.8329564141579924e-05, "loss": 0.8548, "step": 862 }, { "epoch": 0.6350257542310522, "grad_norm": 1.7667874767670178, "learning_rate": 1.832516504221654e-05, "loss": 0.9156, "step": 863 }, { "epoch": 0.6357615894039735, "grad_norm": 1.6926252225049419, "learning_rate": 1.8320760687290177e-05, "loss": 0.8721, "step": 864 }, { "epoch": 0.6364974245768947, "grad_norm": 1.726748517207863, "learning_rate": 1.831635107958125e-05, "loss": 0.8524, "step": 865 }, { "epoch": 0.637233259749816, "grad_norm": 1.7386930746170506, "learning_rate": 1.8311936221873482e-05, "loss": 0.8697, "step": 866 }, { "epoch": 0.6379690949227373, "grad_norm": 2.0253964098194914, "learning_rate": 1.8307516116953906e-05, "loss": 0.9549, "step": 867 }, { "epoch": 0.6387049300956585, "grad_norm": 1.774307632696965, "learning_rate": 1.8303090767612882e-05, "loss": 0.8958, "step": 868 }, { "epoch": 0.6394407652685798, "grad_norm": 1.6813763578091538, "learning_rate": 1.8298660176644075e-05, "loss": 0.8931, "step": 869 }, { "epoch": 0.6401766004415012, "grad_norm": 1.6837817782257916, "learning_rate": 1.8294224346844445e-05, "loss": 0.8491, "step": 870 }, { "epoch": 0.6409124356144223, "grad_norm": 1.758627756453986, "learning_rate": 1.828978328101428e-05, "loss": 0.9063, "step": 871 }, { "epoch": 0.6416482707873437, "grad_norm": 1.742109786438671, "learning_rate": 1.828533698195716e-05, "loss": 0.8832, "step": 872 }, { "epoch": 0.6423841059602649, "grad_norm": 1.8075858543941092, "learning_rate": 1.828088545247997e-05, "loss": 0.923, "step": 873 }, { "epoch": 0.6431199411331862, "grad_norm": 1.6032483606331305, "learning_rate": 1.8276428695392908e-05, "loss": 0.8655, "step": 874 }, { "epoch": 0.6438557763061075, "grad_norm": 1.7567804469127666, "learning_rate": 1.827196671350946e-05, "loss": 0.9348, "step": 875 }, { "epoch": 0.6445916114790287, "grad_norm": 1.7327731742248051, "learning_rate": 1.8267499509646414e-05, "loss": 0.946, "step": 876 }, { "epoch": 0.64532744665195, "grad_norm": 1.78869601194571, "learning_rate": 1.8263027086623852e-05, "loss": 0.9124, "step": 877 }, { "epoch": 0.6460632818248713, "grad_norm": 1.9664065964549209, "learning_rate": 1.825854944726516e-05, "loss": 0.926, "step": 878 }, { "epoch": 0.6467991169977925, "grad_norm": 1.7193164977637796, "learning_rate": 1.8254066594397004e-05, "loss": 0.8598, "step": 879 }, { "epoch": 0.6475349521707138, "grad_norm": 1.8004704404143086, "learning_rate": 1.824957853084935e-05, "loss": 0.9192, "step": 880 }, { "epoch": 0.6482707873436351, "grad_norm": 1.5850855104913149, "learning_rate": 1.8245085259455453e-05, "loss": 0.8495, "step": 881 }, { "epoch": 0.6490066225165563, "grad_norm": 1.6812405500995857, "learning_rate": 1.8240586783051854e-05, "loss": 0.877, "step": 882 }, { "epoch": 0.6497424576894776, "grad_norm": 1.6121343365643663, "learning_rate": 1.8236083104478373e-05, "loss": 0.8948, "step": 883 }, { "epoch": 0.6504782928623988, "grad_norm": 1.8820654126864933, "learning_rate": 1.8231574226578134e-05, "loss": 0.9078, "step": 884 }, { "epoch": 0.6512141280353201, "grad_norm": 1.8851234189402708, "learning_rate": 1.822706015219752e-05, "loss": 0.8958, "step": 885 }, { "epoch": 0.6519499632082414, "grad_norm": 1.6863811026201005, "learning_rate": 1.82225408841862e-05, "loss": 0.8929, "step": 886 }, { "epoch": 0.6526857983811626, "grad_norm": 1.791821225503059, "learning_rate": 1.8218016425397142e-05, "loss": 0.9141, "step": 887 }, { "epoch": 0.6534216335540839, "grad_norm": 1.692348655201856, "learning_rate": 1.821348677868656e-05, "loss": 0.9526, "step": 888 }, { "epoch": 0.6541574687270052, "grad_norm": 1.7751897385551239, "learning_rate": 1.8208951946913965e-05, "loss": 0.9293, "step": 889 }, { "epoch": 0.6548933038999264, "grad_norm": 1.6575173426230936, "learning_rate": 1.8204411932942137e-05, "loss": 0.8837, "step": 890 }, { "epoch": 0.6556291390728477, "grad_norm": 1.6478778490364037, "learning_rate": 1.8199866739637117e-05, "loss": 0.9025, "step": 891 }, { "epoch": 0.6563649742457689, "grad_norm": 1.6404059824751942, "learning_rate": 1.819531636986823e-05, "loss": 0.8947, "step": 892 }, { "epoch": 0.6571008094186902, "grad_norm": 1.782971711524768, "learning_rate": 1.8190760826508058e-05, "loss": 0.8546, "step": 893 }, { "epoch": 0.6578366445916115, "grad_norm": 1.8760942410317862, "learning_rate": 1.818620011243246e-05, "loss": 0.909, "step": 894 }, { "epoch": 0.6585724797645327, "grad_norm": 1.667554215287101, "learning_rate": 1.8181634230520537e-05, "loss": 0.9186, "step": 895 }, { "epoch": 0.659308314937454, "grad_norm": 1.6456027113026317, "learning_rate": 1.8177063183654688e-05, "loss": 0.8747, "step": 896 }, { "epoch": 0.6600441501103753, "grad_norm": 1.6486389713183311, "learning_rate": 1.817248697472054e-05, "loss": 0.912, "step": 897 }, { "epoch": 0.6607799852832965, "grad_norm": 1.5924163421086241, "learning_rate": 1.8167905606606995e-05, "loss": 0.8249, "step": 898 }, { "epoch": 0.6615158204562178, "grad_norm": 1.6068193794019177, "learning_rate": 1.8163319082206206e-05, "loss": 0.8377, "step": 899 }, { "epoch": 0.6622516556291391, "grad_norm": 1.7248996706571482, "learning_rate": 1.8158727404413584e-05, "loss": 0.8769, "step": 900 }, { "epoch": 0.6629874908020603, "grad_norm": 1.5779295322528013, "learning_rate": 1.8154130576127794e-05, "loss": 0.8435, "step": 901 }, { "epoch": 0.6637233259749816, "grad_norm": 1.7476964042948075, "learning_rate": 1.814952860025075e-05, "loss": 0.9438, "step": 902 }, { "epoch": 0.6644591611479028, "grad_norm": 1.8093028936934659, "learning_rate": 1.8144921479687613e-05, "loss": 0.8299, "step": 903 }, { "epoch": 0.6651949963208241, "grad_norm": 1.7472357997397536, "learning_rate": 1.8140309217346805e-05, "loss": 0.882, "step": 904 }, { "epoch": 0.6659308314937454, "grad_norm": 1.7905477200220201, "learning_rate": 1.8135691816139974e-05, "loss": 0.9454, "step": 905 }, { "epoch": 0.6666666666666666, "grad_norm": 1.7725162183692944, "learning_rate": 1.8131069278982026e-05, "loss": 0.9313, "step": 906 }, { "epoch": 0.6674025018395879, "grad_norm": 1.579034379156971, "learning_rate": 1.812644160879111e-05, "loss": 0.8933, "step": 907 }, { "epoch": 0.6681383370125092, "grad_norm": 1.7626946208567593, "learning_rate": 1.81218088084886e-05, "loss": 0.9348, "step": 908 }, { "epoch": 0.6688741721854304, "grad_norm": 1.8539726268675878, "learning_rate": 1.8117170880999132e-05, "loss": 0.8971, "step": 909 }, { "epoch": 0.6696100073583517, "grad_norm": 1.8243460442984096, "learning_rate": 1.8112527829250558e-05, "loss": 1.0195, "step": 910 }, { "epoch": 0.6703458425312729, "grad_norm": 1.6232410747012707, "learning_rate": 1.8107879656173977e-05, "loss": 0.881, "step": 911 }, { "epoch": 0.6710816777041942, "grad_norm": 1.7473877360372896, "learning_rate": 1.8103226364703714e-05, "loss": 0.8758, "step": 912 }, { "epoch": 0.6718175128771156, "grad_norm": 1.8513546933301925, "learning_rate": 1.809856795777733e-05, "loss": 0.9516, "step": 913 }, { "epoch": 0.6725533480500367, "grad_norm": 1.8157012017857281, "learning_rate": 1.8093904438335617e-05, "loss": 0.9212, "step": 914 }, { "epoch": 0.673289183222958, "grad_norm": 1.9435291789116511, "learning_rate": 1.8089235809322585e-05, "loss": 0.8629, "step": 915 }, { "epoch": 0.6740250183958794, "grad_norm": 1.928221535166243, "learning_rate": 1.8084562073685482e-05, "loss": 0.9452, "step": 916 }, { "epoch": 0.6747608535688006, "grad_norm": 1.6981223625397994, "learning_rate": 1.8079883234374765e-05, "loss": 0.9175, "step": 917 }, { "epoch": 0.6754966887417219, "grad_norm": 1.7518387855721922, "learning_rate": 1.8075199294344132e-05, "loss": 0.917, "step": 918 }, { "epoch": 0.6762325239146432, "grad_norm": 1.75131694592391, "learning_rate": 1.807051025655048e-05, "loss": 0.8946, "step": 919 }, { "epoch": 0.6769683590875644, "grad_norm": 1.7122999705950404, "learning_rate": 1.8065816123953944e-05, "loss": 0.9322, "step": 920 }, { "epoch": 0.6777041942604857, "grad_norm": 1.6328412374873396, "learning_rate": 1.8061116899517863e-05, "loss": 0.9223, "step": 921 }, { "epoch": 0.6784400294334069, "grad_norm": 1.6430916248377212, "learning_rate": 1.8056412586208784e-05, "loss": 0.9074, "step": 922 }, { "epoch": 0.6791758646063282, "grad_norm": 1.6466536575741342, "learning_rate": 1.805170318699649e-05, "loss": 0.8811, "step": 923 }, { "epoch": 0.6799116997792495, "grad_norm": 1.6933089602955034, "learning_rate": 1.8046988704853946e-05, "loss": 0.9611, "step": 924 }, { "epoch": 0.6806475349521707, "grad_norm": 1.698774048346321, "learning_rate": 1.8042269142757354e-05, "loss": 0.9183, "step": 925 }, { "epoch": 0.681383370125092, "grad_norm": 1.6953753994958416, "learning_rate": 1.80375445036861e-05, "loss": 0.8625, "step": 926 }, { "epoch": 0.6821192052980133, "grad_norm": 1.81952311519756, "learning_rate": 1.803281479062279e-05, "loss": 0.9331, "step": 927 }, { "epoch": 0.6828550404709345, "grad_norm": 1.6845783588375605, "learning_rate": 1.8028080006553223e-05, "loss": 0.8991, "step": 928 }, { "epoch": 0.6835908756438558, "grad_norm": 1.6718647961304327, "learning_rate": 1.8023340154466402e-05, "loss": 0.8933, "step": 929 }, { "epoch": 0.6843267108167771, "grad_norm": 1.6362867895738917, "learning_rate": 1.8018595237354536e-05, "loss": 0.8941, "step": 930 }, { "epoch": 0.6850625459896983, "grad_norm": 1.826713297641183, "learning_rate": 1.8013845258213024e-05, "loss": 0.9094, "step": 931 }, { "epoch": 0.6857983811626196, "grad_norm": 1.694122152453715, "learning_rate": 1.800909022004046e-05, "loss": 0.8808, "step": 932 }, { "epoch": 0.6865342163355408, "grad_norm": 1.7214137026363854, "learning_rate": 1.8004330125838642e-05, "loss": 0.9455, "step": 933 }, { "epoch": 0.6872700515084621, "grad_norm": 1.6812997312860778, "learning_rate": 1.7999564978612544e-05, "loss": 0.8952, "step": 934 }, { "epoch": 0.6880058866813834, "grad_norm": 1.667609992534637, "learning_rate": 1.799479478137035e-05, "loss": 0.9091, "step": 935 }, { "epoch": 0.6887417218543046, "grad_norm": 1.7060439788401367, "learning_rate": 1.7990019537123403e-05, "loss": 0.8719, "step": 936 }, { "epoch": 0.6894775570272259, "grad_norm": 1.7624113971321436, "learning_rate": 1.7985239248886264e-05, "loss": 0.9371, "step": 937 }, { "epoch": 0.6902133922001472, "grad_norm": 1.705172001005095, "learning_rate": 1.7980453919676658e-05, "loss": 0.8986, "step": 938 }, { "epoch": 0.6909492273730684, "grad_norm": 1.7773321415933379, "learning_rate": 1.79756635525155e-05, "loss": 0.8974, "step": 939 }, { "epoch": 0.6916850625459897, "grad_norm": 1.6522655703241387, "learning_rate": 1.797086815042688e-05, "loss": 0.9165, "step": 940 }, { "epoch": 0.6924208977189109, "grad_norm": 1.5744038358014787, "learning_rate": 1.7966067716438075e-05, "loss": 0.8713, "step": 941 }, { "epoch": 0.6931567328918322, "grad_norm": 1.651733681232309, "learning_rate": 1.7961262253579526e-05, "loss": 0.8783, "step": 942 }, { "epoch": 0.6938925680647535, "grad_norm": 1.7014541890273394, "learning_rate": 1.7956451764884862e-05, "loss": 0.8717, "step": 943 }, { "epoch": 0.6946284032376747, "grad_norm": 1.6126638850942638, "learning_rate": 1.795163625339088e-05, "loss": 0.9015, "step": 944 }, { "epoch": 0.695364238410596, "grad_norm": 1.702056385005647, "learning_rate": 1.7946815722137543e-05, "loss": 0.9235, "step": 945 }, { "epoch": 0.6961000735835173, "grad_norm": 1.7284635730362221, "learning_rate": 1.7941990174167987e-05, "loss": 0.9184, "step": 946 }, { "epoch": 0.6968359087564385, "grad_norm": 1.7498698852359829, "learning_rate": 1.7937159612528514e-05, "loss": 0.9504, "step": 947 }, { "epoch": 0.6975717439293598, "grad_norm": 1.757413054612637, "learning_rate": 1.7932324040268594e-05, "loss": 0.8954, "step": 948 }, { "epoch": 0.6983075791022811, "grad_norm": 1.9380371146612057, "learning_rate": 1.7927483460440857e-05, "loss": 0.9544, "step": 949 }, { "epoch": 0.6990434142752023, "grad_norm": 1.7866892362774434, "learning_rate": 1.792263787610109e-05, "loss": 0.8747, "step": 950 }, { "epoch": 0.6997792494481236, "grad_norm": 1.716728117866327, "learning_rate": 1.791778729030825e-05, "loss": 0.8863, "step": 951 }, { "epoch": 0.7005150846210448, "grad_norm": 1.7095048569976103, "learning_rate": 1.7912931706124447e-05, "loss": 0.8586, "step": 952 }, { "epoch": 0.7012509197939661, "grad_norm": 1.7607191238292803, "learning_rate": 1.790807112661494e-05, "loss": 0.952, "step": 953 }, { "epoch": 0.7019867549668874, "grad_norm": 1.806719390437014, "learning_rate": 1.7903205554848148e-05, "loss": 0.929, "step": 954 }, { "epoch": 0.7027225901398086, "grad_norm": 1.638509755904781, "learning_rate": 1.789833499389564e-05, "loss": 0.8696, "step": 955 }, { "epoch": 0.70345842531273, "grad_norm": 1.681582865538134, "learning_rate": 1.7893459446832135e-05, "loss": 0.8662, "step": 956 }, { "epoch": 0.7041942604856513, "grad_norm": 1.7755801938149824, "learning_rate": 1.7888578916735502e-05, "loss": 0.8987, "step": 957 }, { "epoch": 0.7049300956585725, "grad_norm": 1.73403172221321, "learning_rate": 1.7883693406686746e-05, "loss": 0.8761, "step": 958 }, { "epoch": 0.7056659308314938, "grad_norm": 1.6576610953169342, "learning_rate": 1.787880291977003e-05, "loss": 0.8572, "step": 959 }, { "epoch": 0.7064017660044151, "grad_norm": 1.6503140413697055, "learning_rate": 1.7873907459072648e-05, "loss": 0.8426, "step": 960 }, { "epoch": 0.7071376011773363, "grad_norm": 1.6942505869289013, "learning_rate": 1.786900702768504e-05, "loss": 0.9118, "step": 961 }, { "epoch": 0.7078734363502576, "grad_norm": 1.6735250606275054, "learning_rate": 1.7864101628700776e-05, "loss": 0.8487, "step": 962 }, { "epoch": 0.7086092715231788, "grad_norm": 1.7549219787295118, "learning_rate": 1.7859191265216572e-05, "loss": 0.9086, "step": 963 }, { "epoch": 0.7093451066961001, "grad_norm": 1.7215752046371642, "learning_rate": 1.7854275940332272e-05, "loss": 0.9001, "step": 964 }, { "epoch": 0.7100809418690214, "grad_norm": 1.7987183798278892, "learning_rate": 1.7849355657150857e-05, "loss": 0.9075, "step": 965 }, { "epoch": 0.7108167770419426, "grad_norm": 1.686377758897606, "learning_rate": 1.7844430418778426e-05, "loss": 0.8887, "step": 966 }, { "epoch": 0.7115526122148639, "grad_norm": 1.6569103091368755, "learning_rate": 1.7839500228324223e-05, "loss": 0.9255, "step": 967 }, { "epoch": 0.7122884473877852, "grad_norm": 1.6714410041998493, "learning_rate": 1.7834565088900605e-05, "loss": 0.9395, "step": 968 }, { "epoch": 0.7130242825607064, "grad_norm": 1.6963489834413195, "learning_rate": 1.7829625003623063e-05, "loss": 0.8526, "step": 969 }, { "epoch": 0.7137601177336277, "grad_norm": 1.6956640854686773, "learning_rate": 1.78246799756102e-05, "loss": 0.8726, "step": 970 }, { "epoch": 0.7144959529065489, "grad_norm": 1.6816352549130953, "learning_rate": 1.7819730007983747e-05, "loss": 0.9106, "step": 971 }, { "epoch": 0.7152317880794702, "grad_norm": 1.7580351860771737, "learning_rate": 1.7814775103868553e-05, "loss": 0.9133, "step": 972 }, { "epoch": 0.7159676232523915, "grad_norm": 1.8352119330961616, "learning_rate": 1.7809815266392575e-05, "loss": 0.878, "step": 973 }, { "epoch": 0.7167034584253127, "grad_norm": 1.725580703862397, "learning_rate": 1.78048504986869e-05, "loss": 0.8793, "step": 974 }, { "epoch": 0.717439293598234, "grad_norm": 1.6808299579360038, "learning_rate": 1.779988080388571e-05, "loss": 0.8643, "step": 975 }, { "epoch": 0.7181751287711553, "grad_norm": 1.7248499204061802, "learning_rate": 1.779490618512631e-05, "loss": 0.919, "step": 976 }, { "epoch": 0.7189109639440765, "grad_norm": 1.7226119184421844, "learning_rate": 1.7789926645549103e-05, "loss": 0.9246, "step": 977 }, { "epoch": 0.7196467991169978, "grad_norm": 1.7329662031062194, "learning_rate": 1.7784942188297614e-05, "loss": 0.929, "step": 978 }, { "epoch": 0.7203826342899191, "grad_norm": 1.745386414974443, "learning_rate": 1.7779952816518454e-05, "loss": 0.9224, "step": 979 }, { "epoch": 0.7211184694628403, "grad_norm": 1.7526868710727221, "learning_rate": 1.777495853336135e-05, "loss": 0.9007, "step": 980 }, { "epoch": 0.7218543046357616, "grad_norm": 1.6509791154916162, "learning_rate": 1.7769959341979123e-05, "loss": 0.8921, "step": 981 }, { "epoch": 0.7225901398086828, "grad_norm": 1.5612527359365969, "learning_rate": 1.7764955245527693e-05, "loss": 0.9084, "step": 982 }, { "epoch": 0.7233259749816041, "grad_norm": 1.67218362154563, "learning_rate": 1.7759946247166083e-05, "loss": 0.889, "step": 983 }, { "epoch": 0.7240618101545254, "grad_norm": 1.6560543746650964, "learning_rate": 1.7754932350056395e-05, "loss": 0.8831, "step": 984 }, { "epoch": 0.7247976453274466, "grad_norm": 1.8485117379598066, "learning_rate": 1.7749913557363844e-05, "loss": 0.9212, "step": 985 }, { "epoch": 0.7255334805003679, "grad_norm": 1.6271574110106015, "learning_rate": 1.7744889872256724e-05, "loss": 0.9064, "step": 986 }, { "epoch": 0.7262693156732892, "grad_norm": 1.7287499860734992, "learning_rate": 1.7739861297906413e-05, "loss": 0.884, "step": 987 }, { "epoch": 0.7270051508462104, "grad_norm": 1.7927229982040551, "learning_rate": 1.7734827837487386e-05, "loss": 0.949, "step": 988 }, { "epoch": 0.7277409860191317, "grad_norm": 1.7654777731245228, "learning_rate": 1.77297894941772e-05, "loss": 0.8842, "step": 989 }, { "epoch": 0.7284768211920529, "grad_norm": 1.6845305901819827, "learning_rate": 1.7724746271156482e-05, "loss": 0.8903, "step": 990 }, { "epoch": 0.7292126563649742, "grad_norm": 1.7962609722485419, "learning_rate": 1.771969817160896e-05, "loss": 0.9198, "step": 991 }, { "epoch": 0.7299484915378955, "grad_norm": 1.7329976780449403, "learning_rate": 1.771464519872143e-05, "loss": 0.9281, "step": 992 }, { "epoch": 0.7306843267108167, "grad_norm": 1.6318469599505676, "learning_rate": 1.7709587355683756e-05, "loss": 0.8783, "step": 993 }, { "epoch": 0.731420161883738, "grad_norm": 1.7025774764829062, "learning_rate": 1.770452464568889e-05, "loss": 0.8508, "step": 994 }, { "epoch": 0.7321559970566593, "grad_norm": 1.7361642533610608, "learning_rate": 1.7699457071932858e-05, "loss": 0.8911, "step": 995 }, { "epoch": 0.7328918322295805, "grad_norm": 1.7658009895161806, "learning_rate": 1.7694384637614745e-05, "loss": 0.9251, "step": 996 }, { "epoch": 0.7336276674025018, "grad_norm": 1.7202050689837307, "learning_rate": 1.7689307345936705e-05, "loss": 0.8979, "step": 997 }, { "epoch": 0.7343635025754232, "grad_norm": 1.6931068723782594, "learning_rate": 1.768422520010397e-05, "loss": 0.9099, "step": 998 }, { "epoch": 0.7350993377483444, "grad_norm": 1.593415443042027, "learning_rate": 1.7679138203324823e-05, "loss": 0.8464, "step": 999 }, { "epoch": 0.7358351729212657, "grad_norm": 1.587267044476069, "learning_rate": 1.767404635881062e-05, "loss": 0.8554, "step": 1000 }, { "epoch": 0.7365710080941869, "grad_norm": 1.5809262925188297, "learning_rate": 1.766894966977577e-05, "loss": 0.8707, "step": 1001 }, { "epoch": 0.7373068432671082, "grad_norm": 1.7481725206393877, "learning_rate": 1.7663848139437747e-05, "loss": 0.8916, "step": 1002 }, { "epoch": 0.7380426784400295, "grad_norm": 1.883512177033637, "learning_rate": 1.7658741771017076e-05, "loss": 0.9429, "step": 1003 }, { "epoch": 0.7387785136129507, "grad_norm": 1.7392422186013168, "learning_rate": 1.7653630567737337e-05, "loss": 0.8828, "step": 1004 }, { "epoch": 0.739514348785872, "grad_norm": 1.7674569230695805, "learning_rate": 1.764851453282516e-05, "loss": 0.8813, "step": 1005 }, { "epoch": 0.7402501839587933, "grad_norm": 1.6196053271068664, "learning_rate": 1.7643393669510236e-05, "loss": 0.8366, "step": 1006 }, { "epoch": 0.7409860191317145, "grad_norm": 1.5752540742710246, "learning_rate": 1.7638267981025286e-05, "loss": 0.857, "step": 1007 }, { "epoch": 0.7417218543046358, "grad_norm": 1.6866343967905144, "learning_rate": 1.7633137470606095e-05, "loss": 0.9057, "step": 1008 }, { "epoch": 0.7424576894775571, "grad_norm": 1.6406346644779626, "learning_rate": 1.7628002141491477e-05, "loss": 0.8959, "step": 1009 }, { "epoch": 0.7431935246504783, "grad_norm": 1.6909854729348062, "learning_rate": 1.7622861996923305e-05, "loss": 0.856, "step": 1010 }, { "epoch": 0.7439293598233996, "grad_norm": 1.7015338840302223, "learning_rate": 1.761771704014647e-05, "loss": 0.9054, "step": 1011 }, { "epoch": 0.7446651949963208, "grad_norm": 1.6913350938778275, "learning_rate": 1.7612567274408925e-05, "loss": 0.877, "step": 1012 }, { "epoch": 0.7454010301692421, "grad_norm": 1.6782898861728612, "learning_rate": 1.7607412702961638e-05, "loss": 0.87, "step": 1013 }, { "epoch": 0.7461368653421634, "grad_norm": 1.6333663534510834, "learning_rate": 1.760225332905862e-05, "loss": 0.9007, "step": 1014 }, { "epoch": 0.7468727005150846, "grad_norm": 1.6990126921499376, "learning_rate": 1.759708915595692e-05, "loss": 0.8599, "step": 1015 }, { "epoch": 0.7476085356880059, "grad_norm": 1.6528905682971202, "learning_rate": 1.7591920186916605e-05, "loss": 0.8189, "step": 1016 }, { "epoch": 0.7483443708609272, "grad_norm": 1.8251624747316608, "learning_rate": 1.7586746425200773e-05, "loss": 0.8779, "step": 1017 }, { "epoch": 0.7490802060338484, "grad_norm": 1.7212940497575995, "learning_rate": 1.7581567874075552e-05, "loss": 0.9703, "step": 1018 }, { "epoch": 0.7498160412067697, "grad_norm": 1.8257756040729594, "learning_rate": 1.757638453681009e-05, "loss": 0.9283, "step": 1019 }, { "epoch": 0.7505518763796909, "grad_norm": 1.748739488286786, "learning_rate": 1.757119641667656e-05, "loss": 0.8851, "step": 1020 }, { "epoch": 0.7512877115526122, "grad_norm": 1.6489910239232946, "learning_rate": 1.7566003516950146e-05, "loss": 0.9179, "step": 1021 }, { "epoch": 0.7520235467255335, "grad_norm": 1.787728701376888, "learning_rate": 1.7560805840909058e-05, "loss": 0.9349, "step": 1022 }, { "epoch": 0.7527593818984547, "grad_norm": 1.6217312797735408, "learning_rate": 1.7555603391834514e-05, "loss": 0.8456, "step": 1023 }, { "epoch": 0.753495217071376, "grad_norm": 1.6630175027504652, "learning_rate": 1.755039617301075e-05, "loss": 0.8528, "step": 1024 }, { "epoch": 0.7542310522442973, "grad_norm": 1.568874941088452, "learning_rate": 1.754518418772502e-05, "loss": 0.8518, "step": 1025 }, { "epoch": 0.7549668874172185, "grad_norm": 1.694938445537615, "learning_rate": 1.7539967439267574e-05, "loss": 0.9085, "step": 1026 }, { "epoch": 0.7557027225901398, "grad_norm": 1.7098006726041466, "learning_rate": 1.753474593093167e-05, "loss": 0.9245, "step": 1027 }, { "epoch": 0.7564385577630611, "grad_norm": 1.734633064692764, "learning_rate": 1.7529519666013578e-05, "loss": 0.8493, "step": 1028 }, { "epoch": 0.7571743929359823, "grad_norm": 1.736056457074215, "learning_rate": 1.752428864781257e-05, "loss": 0.8363, "step": 1029 }, { "epoch": 0.7579102281089036, "grad_norm": 1.6944498981497336, "learning_rate": 1.751905287963091e-05, "loss": 0.9095, "step": 1030 }, { "epoch": 0.7586460632818248, "grad_norm": 1.7738905000974685, "learning_rate": 1.7513812364773876e-05, "loss": 0.8782, "step": 1031 }, { "epoch": 0.7593818984547461, "grad_norm": 1.6751712076715068, "learning_rate": 1.7508567106549725e-05, "loss": 0.8795, "step": 1032 }, { "epoch": 0.7601177336276674, "grad_norm": 1.8477432766301547, "learning_rate": 1.7503317108269722e-05, "loss": 0.8993, "step": 1033 }, { "epoch": 0.7608535688005886, "grad_norm": 1.751786907114036, "learning_rate": 1.7498062373248118e-05, "loss": 0.9444, "step": 1034 }, { "epoch": 0.7615894039735099, "grad_norm": 1.7423014887005308, "learning_rate": 1.7492802904802155e-05, "loss": 0.873, "step": 1035 }, { "epoch": 0.7623252391464312, "grad_norm": 1.5935790770419576, "learning_rate": 1.7487538706252062e-05, "loss": 0.8562, "step": 1036 }, { "epoch": 0.7630610743193524, "grad_norm": 1.5295017019675006, "learning_rate": 1.7482269780921056e-05, "loss": 0.8703, "step": 1037 }, { "epoch": 0.7637969094922737, "grad_norm": 1.5749604327197473, "learning_rate": 1.7476996132135342e-05, "loss": 0.9107, "step": 1038 }, { "epoch": 0.7645327446651949, "grad_norm": 1.7014820992965765, "learning_rate": 1.7471717763224096e-05, "loss": 0.8496, "step": 1039 }, { "epoch": 0.7652685798381162, "grad_norm": 1.6597765029955829, "learning_rate": 1.7466434677519485e-05, "loss": 0.8771, "step": 1040 }, { "epoch": 0.7660044150110376, "grad_norm": 1.7041948620419305, "learning_rate": 1.746114687835665e-05, "loss": 0.9351, "step": 1041 }, { "epoch": 0.7667402501839587, "grad_norm": 1.8040072988679428, "learning_rate": 1.7455854369073703e-05, "loss": 0.9225, "step": 1042 }, { "epoch": 0.7674760853568801, "grad_norm": 1.7224457028896747, "learning_rate": 1.745055715301173e-05, "loss": 0.9376, "step": 1043 }, { "epoch": 0.7682119205298014, "grad_norm": 1.6780584475153935, "learning_rate": 1.7445255233514794e-05, "loss": 0.8468, "step": 1044 }, { "epoch": 0.7689477557027226, "grad_norm": 1.6142850849363055, "learning_rate": 1.7439948613929928e-05, "loss": 0.8897, "step": 1045 }, { "epoch": 0.7696835908756439, "grad_norm": 1.719451088023112, "learning_rate": 1.7434637297607125e-05, "loss": 0.9275, "step": 1046 }, { "epoch": 0.7704194260485652, "grad_norm": 1.6510986546566475, "learning_rate": 1.742932128789935e-05, "loss": 0.8717, "step": 1047 }, { "epoch": 0.7711552612214864, "grad_norm": 1.9456984417645686, "learning_rate": 1.742400058816252e-05, "loss": 0.8817, "step": 1048 }, { "epoch": 0.7718910963944077, "grad_norm": 1.5663808120956695, "learning_rate": 1.7418675201755525e-05, "loss": 0.8479, "step": 1049 }, { "epoch": 0.7726269315673289, "grad_norm": 1.6068194077425801, "learning_rate": 1.741334513204021e-05, "loss": 0.8674, "step": 1050 }, { "epoch": 0.7733627667402502, "grad_norm": 1.7311901949913588, "learning_rate": 1.740801038238137e-05, "loss": 0.8719, "step": 1051 }, { "epoch": 0.7740986019131715, "grad_norm": 1.7202977237586272, "learning_rate": 1.7402670956146766e-05, "loss": 0.9192, "step": 1052 }, { "epoch": 0.7748344370860927, "grad_norm": 1.777253786119638, "learning_rate": 1.7397326856707102e-05, "loss": 0.8966, "step": 1053 }, { "epoch": 0.775570272259014, "grad_norm": 1.608643536296025, "learning_rate": 1.7391978087436032e-05, "loss": 0.8942, "step": 1054 }, { "epoch": 0.7763061074319353, "grad_norm": 1.5897881513138783, "learning_rate": 1.7386624651710165e-05, "loss": 0.8294, "step": 1055 }, { "epoch": 0.7770419426048565, "grad_norm": 1.7609020926304395, "learning_rate": 1.738126655290905e-05, "loss": 0.9192, "step": 1056 }, { "epoch": 0.7777777777777778, "grad_norm": 1.666958985843096, "learning_rate": 1.737590379441518e-05, "loss": 0.9058, "step": 1057 }, { "epoch": 0.7785136129506991, "grad_norm": 1.7128350340595238, "learning_rate": 1.7370536379613998e-05, "loss": 0.8636, "step": 1058 }, { "epoch": 0.7792494481236203, "grad_norm": 1.6848298776114112, "learning_rate": 1.7365164311893868e-05, "loss": 0.8535, "step": 1059 }, { "epoch": 0.7799852832965416, "grad_norm": 1.6781959818499612, "learning_rate": 1.735978759464612e-05, "loss": 0.8814, "step": 1060 }, { "epoch": 0.7807211184694628, "grad_norm": 1.6726838507290587, "learning_rate": 1.7354406231264983e-05, "loss": 0.8477, "step": 1061 }, { "epoch": 0.7814569536423841, "grad_norm": 1.7808032334364101, "learning_rate": 1.7349020225147653e-05, "loss": 0.8623, "step": 1062 }, { "epoch": 0.7821927888153054, "grad_norm": 1.7540144354508909, "learning_rate": 1.734362957969423e-05, "loss": 0.9066, "step": 1063 }, { "epoch": 0.7829286239882266, "grad_norm": 1.6376905470414378, "learning_rate": 1.7338234298307764e-05, "loss": 0.9069, "step": 1064 }, { "epoch": 0.7836644591611479, "grad_norm": 1.788310034625574, "learning_rate": 1.7332834384394225e-05, "loss": 0.9449, "step": 1065 }, { "epoch": 0.7844002943340692, "grad_norm": 1.692642770046459, "learning_rate": 1.7327429841362494e-05, "loss": 0.9244, "step": 1066 }, { "epoch": 0.7851361295069904, "grad_norm": 1.582186543179142, "learning_rate": 1.732202067262439e-05, "loss": 0.8561, "step": 1067 }, { "epoch": 0.7858719646799117, "grad_norm": 1.6594537708198056, "learning_rate": 1.7316606881594652e-05, "loss": 0.9456, "step": 1068 }, { "epoch": 0.7866077998528329, "grad_norm": 1.7401806751262032, "learning_rate": 1.7311188471690925e-05, "loss": 0.8596, "step": 1069 }, { "epoch": 0.7873436350257542, "grad_norm": 1.7424209969054798, "learning_rate": 1.7305765446333783e-05, "loss": 0.8681, "step": 1070 }, { "epoch": 0.7880794701986755, "grad_norm": 1.840353121846038, "learning_rate": 1.7300337808946703e-05, "loss": 0.907, "step": 1071 }, { "epoch": 0.7888153053715967, "grad_norm": 1.7360232098766533, "learning_rate": 1.729490556295608e-05, "loss": 0.9603, "step": 1072 }, { "epoch": 0.789551140544518, "grad_norm": 1.6978928596665108, "learning_rate": 1.728946871179122e-05, "loss": 0.9176, "step": 1073 }, { "epoch": 0.7902869757174393, "grad_norm": 1.793958587547566, "learning_rate": 1.728402725888433e-05, "loss": 0.8967, "step": 1074 }, { "epoch": 0.7910228108903605, "grad_norm": 1.6486567737934503, "learning_rate": 1.7278581207670522e-05, "loss": 0.8742, "step": 1075 }, { "epoch": 0.7917586460632818, "grad_norm": 1.595935565637869, "learning_rate": 1.7273130561587822e-05, "loss": 0.8689, "step": 1076 }, { "epoch": 0.7924944812362031, "grad_norm": 1.7111177623981513, "learning_rate": 1.7267675324077142e-05, "loss": 0.9064, "step": 1077 }, { "epoch": 0.7932303164091243, "grad_norm": 1.812932379729852, "learning_rate": 1.72622154985823e-05, "loss": 0.9059, "step": 1078 }, { "epoch": 0.7939661515820456, "grad_norm": 1.6534077109272434, "learning_rate": 1.7256751088550007e-05, "loss": 0.9022, "step": 1079 }, { "epoch": 0.7947019867549668, "grad_norm": 1.7849569753686683, "learning_rate": 1.7251282097429876e-05, "loss": 0.8942, "step": 1080 }, { "epoch": 0.7954378219278881, "grad_norm": 2.0705364391658745, "learning_rate": 1.7245808528674403e-05, "loss": 0.905, "step": 1081 }, { "epoch": 0.7961736571008095, "grad_norm": 1.8012740629931383, "learning_rate": 1.7240330385738974e-05, "loss": 0.8765, "step": 1082 }, { "epoch": 0.7969094922737306, "grad_norm": 1.5716514721663128, "learning_rate": 1.7234847672081875e-05, "loss": 0.8626, "step": 1083 }, { "epoch": 0.797645327446652, "grad_norm": 1.6990960705932292, "learning_rate": 1.7229360391164256e-05, "loss": 0.8437, "step": 1084 }, { "epoch": 0.7983811626195733, "grad_norm": 1.6318349681502868, "learning_rate": 1.7223868546450172e-05, "loss": 0.8773, "step": 1085 }, { "epoch": 0.7991169977924945, "grad_norm": 1.7289064052650287, "learning_rate": 1.721837214140655e-05, "loss": 0.8945, "step": 1086 }, { "epoch": 0.7998528329654158, "grad_norm": 1.6852159766390455, "learning_rate": 1.7212871179503188e-05, "loss": 0.8626, "step": 1087 }, { "epoch": 0.8005886681383371, "grad_norm": 1.6634107319622, "learning_rate": 1.7207365664212773e-05, "loss": 0.9212, "step": 1088 }, { "epoch": 0.8013245033112583, "grad_norm": 1.6530804658184248, "learning_rate": 1.720185559901086e-05, "loss": 0.8489, "step": 1089 }, { "epoch": 0.8020603384841796, "grad_norm": 1.732736325429605, "learning_rate": 1.719634098737588e-05, "loss": 0.8636, "step": 1090 }, { "epoch": 0.8027961736571008, "grad_norm": 1.8105815602142552, "learning_rate": 1.719082183278913e-05, "loss": 0.8515, "step": 1091 }, { "epoch": 0.8035320088300221, "grad_norm": 1.6998000547887002, "learning_rate": 1.718529813873478e-05, "loss": 0.9133, "step": 1092 }, { "epoch": 0.8042678440029434, "grad_norm": 1.7743858257210108, "learning_rate": 1.7179769908699856e-05, "loss": 0.8316, "step": 1093 }, { "epoch": 0.8050036791758646, "grad_norm": 1.7394802638295903, "learning_rate": 1.717423714617426e-05, "loss": 0.8845, "step": 1094 }, { "epoch": 0.8057395143487859, "grad_norm": 1.898771873295927, "learning_rate": 1.716869985465075e-05, "loss": 0.8813, "step": 1095 }, { "epoch": 0.8064753495217072, "grad_norm": 1.625729436491641, "learning_rate": 1.716315803762494e-05, "loss": 0.8631, "step": 1096 }, { "epoch": 0.8072111846946284, "grad_norm": 1.5998110993133852, "learning_rate": 1.71576116985953e-05, "loss": 0.8207, "step": 1097 }, { "epoch": 0.8079470198675497, "grad_norm": 1.6659216364274672, "learning_rate": 1.7152060841063163e-05, "loss": 0.91, "step": 1098 }, { "epoch": 0.8086828550404709, "grad_norm": 1.5624315664740804, "learning_rate": 1.7146505468532707e-05, "loss": 0.925, "step": 1099 }, { "epoch": 0.8094186902133922, "grad_norm": 1.5164289923925567, "learning_rate": 1.714094558451097e-05, "loss": 0.8846, "step": 1100 }, { "epoch": 0.8101545253863135, "grad_norm": 1.7044509998898014, "learning_rate": 1.7135381192507817e-05, "loss": 0.8734, "step": 1101 }, { "epoch": 0.8108903605592347, "grad_norm": 1.6824724704238292, "learning_rate": 1.7129812296035985e-05, "loss": 0.9615, "step": 1102 }, { "epoch": 0.811626195732156, "grad_norm": 1.656972572538243, "learning_rate": 1.7124238898611035e-05, "loss": 0.9019, "step": 1103 }, { "epoch": 0.8123620309050773, "grad_norm": 1.9060698058569623, "learning_rate": 1.7118661003751383e-05, "loss": 0.905, "step": 1104 }, { "epoch": 0.8130978660779985, "grad_norm": 1.7745151557463446, "learning_rate": 1.711307861497827e-05, "loss": 0.898, "step": 1105 }, { "epoch": 0.8138337012509198, "grad_norm": 1.5386979975776054, "learning_rate": 1.710749173581579e-05, "loss": 0.8514, "step": 1106 }, { "epoch": 0.8145695364238411, "grad_norm": 1.5535424819510375, "learning_rate": 1.7101900369790857e-05, "loss": 0.8951, "step": 1107 }, { "epoch": 0.8153053715967623, "grad_norm": 1.696140422426633, "learning_rate": 1.709630452043323e-05, "loss": 0.8744, "step": 1108 }, { "epoch": 0.8160412067696836, "grad_norm": 1.7095263873266016, "learning_rate": 1.7090704191275486e-05, "loss": 0.9238, "step": 1109 }, { "epoch": 0.8167770419426048, "grad_norm": 1.6756415862439744, "learning_rate": 1.7085099385853036e-05, "loss": 0.8666, "step": 1110 }, { "epoch": 0.8175128771155261, "grad_norm": 1.9279733754511357, "learning_rate": 1.707949010770413e-05, "loss": 0.901, "step": 1111 }, { "epoch": 0.8182487122884474, "grad_norm": 1.7843755415025104, "learning_rate": 1.707387636036981e-05, "loss": 0.8775, "step": 1112 }, { "epoch": 0.8189845474613686, "grad_norm": 1.6576723533586375, "learning_rate": 1.7068258147393967e-05, "loss": 0.9045, "step": 1113 }, { "epoch": 0.8197203826342899, "grad_norm": 1.8341227234482658, "learning_rate": 1.7062635472323306e-05, "loss": 0.9222, "step": 1114 }, { "epoch": 0.8204562178072112, "grad_norm": 1.6593984538959745, "learning_rate": 1.7057008338707342e-05, "loss": 0.8639, "step": 1115 }, { "epoch": 0.8211920529801324, "grad_norm": 1.6828687358234797, "learning_rate": 1.7051376750098404e-05, "loss": 0.8835, "step": 1116 }, { "epoch": 0.8219278881530537, "grad_norm": 1.5829433587578319, "learning_rate": 1.7045740710051637e-05, "loss": 0.872, "step": 1117 }, { "epoch": 0.8226637233259749, "grad_norm": 1.7315416091450515, "learning_rate": 1.7040100222125e-05, "loss": 0.8535, "step": 1118 }, { "epoch": 0.8233995584988962, "grad_norm": 1.6592406607044399, "learning_rate": 1.703445528987926e-05, "loss": 0.8585, "step": 1119 }, { "epoch": 0.8241353936718175, "grad_norm": 1.689273280821189, "learning_rate": 1.7028805916877975e-05, "loss": 0.9136, "step": 1120 }, { "epoch": 0.8248712288447387, "grad_norm": 1.6011165133040948, "learning_rate": 1.7023152106687525e-05, "loss": 0.8624, "step": 1121 }, { "epoch": 0.82560706401766, "grad_norm": 1.7511161879496535, "learning_rate": 1.701749386287708e-05, "loss": 0.868, "step": 1122 }, { "epoch": 0.8263428991905813, "grad_norm": 1.6149856081557823, "learning_rate": 1.7011831189018607e-05, "loss": 0.8272, "step": 1123 }, { "epoch": 0.8270787343635025, "grad_norm": 1.6935459916004387, "learning_rate": 1.700616408868689e-05, "loss": 0.8666, "step": 1124 }, { "epoch": 0.8278145695364238, "grad_norm": 1.6858832761442941, "learning_rate": 1.7000492565459476e-05, "loss": 0.9452, "step": 1125 }, { "epoch": 0.8285504047093452, "grad_norm": 1.6653190253290648, "learning_rate": 1.6994816622916726e-05, "loss": 0.8924, "step": 1126 }, { "epoch": 0.8292862398822664, "grad_norm": 1.6695844800740303, "learning_rate": 1.6989136264641785e-05, "loss": 0.8832, "step": 1127 }, { "epoch": 0.8300220750551877, "grad_norm": 1.7203400373130469, "learning_rate": 1.6983451494220587e-05, "loss": 0.8721, "step": 1128 }, { "epoch": 0.8307579102281089, "grad_norm": 1.5995006198597135, "learning_rate": 1.697776231524185e-05, "loss": 0.8804, "step": 1129 }, { "epoch": 0.8314937454010302, "grad_norm": 1.6746299226653387, "learning_rate": 1.6972068731297075e-05, "loss": 0.8679, "step": 1130 }, { "epoch": 0.8322295805739515, "grad_norm": 1.6733475077914468, "learning_rate": 1.696637074598054e-05, "loss": 0.8921, "step": 1131 }, { "epoch": 0.8329654157468727, "grad_norm": 1.6654781488788535, "learning_rate": 1.696066836288931e-05, "loss": 0.8872, "step": 1132 }, { "epoch": 0.833701250919794, "grad_norm": 1.6346955055827317, "learning_rate": 1.6954961585623214e-05, "loss": 0.8767, "step": 1133 }, { "epoch": 0.8344370860927153, "grad_norm": 1.6440525794979899, "learning_rate": 1.6949250417784873e-05, "loss": 0.8595, "step": 1134 }, { "epoch": 0.8351729212656365, "grad_norm": 1.7120600561412382, "learning_rate": 1.694353486297966e-05, "loss": 0.9088, "step": 1135 }, { "epoch": 0.8359087564385578, "grad_norm": 1.718473177325096, "learning_rate": 1.6937814924815733e-05, "loss": 0.8437, "step": 1136 }, { "epoch": 0.8366445916114791, "grad_norm": 1.6745860107970816, "learning_rate": 1.6932090606904012e-05, "loss": 0.8831, "step": 1137 }, { "epoch": 0.8373804267844003, "grad_norm": 1.7911109591694705, "learning_rate": 1.6926361912858172e-05, "loss": 0.8691, "step": 1138 }, { "epoch": 0.8381162619573216, "grad_norm": 1.697555247325954, "learning_rate": 1.692062884629467e-05, "loss": 0.8768, "step": 1139 }, { "epoch": 0.8388520971302428, "grad_norm": 1.5785420646198036, "learning_rate": 1.6914891410832702e-05, "loss": 0.8739, "step": 1140 }, { "epoch": 0.8395879323031641, "grad_norm": 1.7545893811663347, "learning_rate": 1.6909149610094245e-05, "loss": 0.8931, "step": 1141 }, { "epoch": 0.8403237674760854, "grad_norm": 1.770506859254838, "learning_rate": 1.6903403447704004e-05, "loss": 0.879, "step": 1142 }, { "epoch": 0.8410596026490066, "grad_norm": 1.7444278975164118, "learning_rate": 1.6897652927289466e-05, "loss": 0.883, "step": 1143 }, { "epoch": 0.8417954378219279, "grad_norm": 1.7067821989973064, "learning_rate": 1.689189805248085e-05, "loss": 0.8346, "step": 1144 }, { "epoch": 0.8425312729948492, "grad_norm": 1.7039886223387302, "learning_rate": 1.6886138826911132e-05, "loss": 0.8468, "step": 1145 }, { "epoch": 0.8432671081677704, "grad_norm": 1.5462077550249482, "learning_rate": 1.688037525421603e-05, "loss": 0.9137, "step": 1146 }, { "epoch": 0.8440029433406917, "grad_norm": 1.6489514650645816, "learning_rate": 1.6874607338034015e-05, "loss": 0.8889, "step": 1147 }, { "epoch": 0.8447387785136129, "grad_norm": 1.6245729897937689, "learning_rate": 1.6868835082006283e-05, "loss": 0.8624, "step": 1148 }, { "epoch": 0.8454746136865342, "grad_norm": 1.5888154818298126, "learning_rate": 1.6863058489776788e-05, "loss": 0.8604, "step": 1149 }, { "epoch": 0.8462104488594555, "grad_norm": 1.7528797048859943, "learning_rate": 1.6857277564992212e-05, "loss": 0.8346, "step": 1150 }, { "epoch": 0.8469462840323767, "grad_norm": 1.781204082062478, "learning_rate": 1.6851492311301973e-05, "loss": 0.8122, "step": 1151 }, { "epoch": 0.847682119205298, "grad_norm": 1.613556494911429, "learning_rate": 1.6845702732358224e-05, "loss": 0.888, "step": 1152 }, { "epoch": 0.8484179543782193, "grad_norm": 1.6473072745851207, "learning_rate": 1.683990883181585e-05, "loss": 0.8078, "step": 1153 }, { "epoch": 0.8491537895511405, "grad_norm": 1.7002654965083253, "learning_rate": 1.6834110613332453e-05, "loss": 0.894, "step": 1154 }, { "epoch": 0.8498896247240618, "grad_norm": 1.6462593850723826, "learning_rate": 1.6828308080568376e-05, "loss": 0.8881, "step": 1155 }, { "epoch": 0.8506254598969831, "grad_norm": 1.638352266967807, "learning_rate": 1.6822501237186677e-05, "loss": 0.8215, "step": 1156 }, { "epoch": 0.8513612950699043, "grad_norm": 1.7365826484194085, "learning_rate": 1.681669008685314e-05, "loss": 0.9005, "step": 1157 }, { "epoch": 0.8520971302428256, "grad_norm": 1.6524690847066916, "learning_rate": 1.6810874633236255e-05, "loss": 0.8642, "step": 1158 }, { "epoch": 0.8528329654157468, "grad_norm": 1.6822414515300266, "learning_rate": 1.680505488000725e-05, "loss": 0.8497, "step": 1159 }, { "epoch": 0.8535688005886681, "grad_norm": 1.8918047557173674, "learning_rate": 1.6799230830840048e-05, "loss": 0.863, "step": 1160 }, { "epoch": 0.8543046357615894, "grad_norm": 1.648102145625533, "learning_rate": 1.6793402489411293e-05, "loss": 0.8271, "step": 1161 }, { "epoch": 0.8550404709345106, "grad_norm": 1.6399071281031896, "learning_rate": 1.678756985940034e-05, "loss": 0.8663, "step": 1162 }, { "epoch": 0.8557763061074319, "grad_norm": 1.6748691177212722, "learning_rate": 1.6781732944489248e-05, "loss": 0.8801, "step": 1163 }, { "epoch": 0.8565121412803532, "grad_norm": 1.677849035823229, "learning_rate": 1.677589174836278e-05, "loss": 0.8733, "step": 1164 }, { "epoch": 0.8572479764532744, "grad_norm": 1.6218082809420284, "learning_rate": 1.6770046274708404e-05, "loss": 0.856, "step": 1165 }, { "epoch": 0.8579838116261957, "grad_norm": 1.6312752413809628, "learning_rate": 1.676419652721629e-05, "loss": 0.8565, "step": 1166 }, { "epoch": 0.8587196467991169, "grad_norm": 1.5993381171032048, "learning_rate": 1.67583425095793e-05, "loss": 0.8863, "step": 1167 }, { "epoch": 0.8594554819720382, "grad_norm": 1.6931121900383692, "learning_rate": 1.6752484225493e-05, "loss": 0.9411, "step": 1168 }, { "epoch": 0.8601913171449596, "grad_norm": 1.5832405329893218, "learning_rate": 1.6746621678655636e-05, "loss": 0.9219, "step": 1169 }, { "epoch": 0.8609271523178808, "grad_norm": 1.6257220868616764, "learning_rate": 1.6740754872768165e-05, "loss": 0.9243, "step": 1170 }, { "epoch": 0.8616629874908021, "grad_norm": 1.6079239230062925, "learning_rate": 1.673488381153421e-05, "loss": 0.9001, "step": 1171 }, { "epoch": 0.8623988226637234, "grad_norm": 1.6845694693563817, "learning_rate": 1.6729008498660104e-05, "loss": 0.8283, "step": 1172 }, { "epoch": 0.8631346578366446, "grad_norm": 1.526437653793957, "learning_rate": 1.672312893785484e-05, "loss": 0.8457, "step": 1173 }, { "epoch": 0.8638704930095659, "grad_norm": 1.6648427275772735, "learning_rate": 1.6717245132830114e-05, "loss": 0.8757, "step": 1174 }, { "epoch": 0.8646063281824872, "grad_norm": 1.691314475252365, "learning_rate": 1.6711357087300287e-05, "loss": 0.8891, "step": 1175 }, { "epoch": 0.8653421633554084, "grad_norm": 1.7726153211412188, "learning_rate": 1.67054648049824e-05, "loss": 0.8883, "step": 1176 }, { "epoch": 0.8660779985283297, "grad_norm": 1.6978851962258021, "learning_rate": 1.6699568289596175e-05, "loss": 0.9039, "step": 1177 }, { "epoch": 0.8668138337012509, "grad_norm": 1.7044806316112322, "learning_rate": 1.6693667544863998e-05, "loss": 0.8898, "step": 1178 }, { "epoch": 0.8675496688741722, "grad_norm": 1.6822395654708309, "learning_rate": 1.6687762574510927e-05, "loss": 0.8911, "step": 1179 }, { "epoch": 0.8682855040470935, "grad_norm": 1.6136477174108022, "learning_rate": 1.66818533822647e-05, "loss": 0.8392, "step": 1180 }, { "epoch": 0.8690213392200147, "grad_norm": 1.6184890637625937, "learning_rate": 1.6675939971855694e-05, "loss": 0.8603, "step": 1181 }, { "epoch": 0.869757174392936, "grad_norm": 1.9897279716875635, "learning_rate": 1.6670022347016972e-05, "loss": 0.8784, "step": 1182 }, { "epoch": 0.8704930095658573, "grad_norm": 1.74703262535965, "learning_rate": 1.6664100511484252e-05, "loss": 0.8375, "step": 1183 }, { "epoch": 0.8712288447387785, "grad_norm": 1.5941584815980894, "learning_rate": 1.6658174468995903e-05, "loss": 0.8411, "step": 1184 }, { "epoch": 0.8719646799116998, "grad_norm": 1.5807936584898787, "learning_rate": 1.665224422329296e-05, "loss": 0.8515, "step": 1185 }, { "epoch": 0.8727005150846211, "grad_norm": 1.6558569882739427, "learning_rate": 1.66463097781191e-05, "loss": 0.8814, "step": 1186 }, { "epoch": 0.8734363502575423, "grad_norm": 1.6998155215371888, "learning_rate": 1.664037113722066e-05, "loss": 0.8812, "step": 1187 }, { "epoch": 0.8741721854304636, "grad_norm": 1.5611138474274386, "learning_rate": 1.6634428304346623e-05, "loss": 0.9043, "step": 1188 }, { "epoch": 0.8749080206033848, "grad_norm": 1.6283179455146963, "learning_rate": 1.662848128324862e-05, "loss": 0.8477, "step": 1189 }, { "epoch": 0.8756438557763061, "grad_norm": 1.7427604928389562, "learning_rate": 1.662253007768092e-05, "loss": 0.921, "step": 1190 }, { "epoch": 0.8763796909492274, "grad_norm": 1.6129376914608529, "learning_rate": 1.6616574691400443e-05, "loss": 0.831, "step": 1191 }, { "epoch": 0.8771155261221486, "grad_norm": 1.718926699219354, "learning_rate": 1.6610615128166738e-05, "loss": 0.885, "step": 1192 }, { "epoch": 0.8778513612950699, "grad_norm": 1.7149420278779652, "learning_rate": 1.6604651391742e-05, "loss": 0.885, "step": 1193 }, { "epoch": 0.8785871964679912, "grad_norm": 1.7299632943474177, "learning_rate": 1.6598683485891053e-05, "loss": 0.9121, "step": 1194 }, { "epoch": 0.8793230316409124, "grad_norm": 1.6893158474618697, "learning_rate": 1.659271141438135e-05, "loss": 0.8628, "step": 1195 }, { "epoch": 0.8800588668138337, "grad_norm": 1.9503505903693472, "learning_rate": 1.6586735180982986e-05, "loss": 0.9085, "step": 1196 }, { "epoch": 0.8807947019867549, "grad_norm": 1.6792579861055341, "learning_rate": 1.6580754789468666e-05, "loss": 0.8167, "step": 1197 }, { "epoch": 0.8815305371596762, "grad_norm": 1.5858881990458615, "learning_rate": 1.657477024361374e-05, "loss": 0.84, "step": 1198 }, { "epoch": 0.8822663723325975, "grad_norm": 1.6970511879826335, "learning_rate": 1.6568781547196163e-05, "loss": 0.8632, "step": 1199 }, { "epoch": 0.8830022075055187, "grad_norm": 1.70018529729212, "learning_rate": 1.656278870399652e-05, "loss": 0.825, "step": 1200 }, { "epoch": 0.88373804267844, "grad_norm": 1.6914326632069463, "learning_rate": 1.6556791717798013e-05, "loss": 0.9378, "step": 1201 }, { "epoch": 0.8844738778513613, "grad_norm": 1.7776013572859206, "learning_rate": 1.655079059238645e-05, "loss": 0.8513, "step": 1202 }, { "epoch": 0.8852097130242825, "grad_norm": 1.6890346892642896, "learning_rate": 1.6544785331550265e-05, "loss": 0.889, "step": 1203 }, { "epoch": 0.8859455481972038, "grad_norm": 1.6917639146380246, "learning_rate": 1.65387759390805e-05, "loss": 0.8449, "step": 1204 }, { "epoch": 0.8866813833701251, "grad_norm": 1.6895442408989214, "learning_rate": 1.6532762418770797e-05, "loss": 0.8283, "step": 1205 }, { "epoch": 0.8874172185430463, "grad_norm": 1.6756404944735936, "learning_rate": 1.652674477441741e-05, "loss": 0.8542, "step": 1206 }, { "epoch": 0.8881530537159676, "grad_norm": 1.7385620471465553, "learning_rate": 1.65207230098192e-05, "loss": 0.8456, "step": 1207 }, { "epoch": 0.8888888888888888, "grad_norm": 1.6665290236773058, "learning_rate": 1.6514697128777615e-05, "loss": 0.8933, "step": 1208 }, { "epoch": 0.8896247240618101, "grad_norm": 1.72955582747036, "learning_rate": 1.6508667135096727e-05, "loss": 0.8104, "step": 1209 }, { "epoch": 0.8903605592347315, "grad_norm": 2.0062730119411945, "learning_rate": 1.6502633032583173e-05, "loss": 0.886, "step": 1210 }, { "epoch": 0.8910963944076526, "grad_norm": 1.6373961494588334, "learning_rate": 1.6496594825046206e-05, "loss": 0.8591, "step": 1211 }, { "epoch": 0.891832229580574, "grad_norm": 1.6546510132788108, "learning_rate": 1.6490552516297667e-05, "loss": 0.8656, "step": 1212 }, { "epoch": 0.8925680647534953, "grad_norm": 1.6760475022787409, "learning_rate": 1.6484506110151977e-05, "loss": 0.9081, "step": 1213 }, { "epoch": 0.8933038999264165, "grad_norm": 1.7644305814462375, "learning_rate": 1.6478455610426157e-05, "loss": 0.8537, "step": 1214 }, { "epoch": 0.8940397350993378, "grad_norm": 1.7252321113624203, "learning_rate": 1.6472401020939797e-05, "loss": 0.8403, "step": 1215 }, { "epoch": 0.8947755702722591, "grad_norm": 1.637141278507686, "learning_rate": 1.646634234551508e-05, "loss": 0.8571, "step": 1216 }, { "epoch": 0.8955114054451803, "grad_norm": 1.6073068864075206, "learning_rate": 1.6460279587976765e-05, "loss": 0.8467, "step": 1217 }, { "epoch": 0.8962472406181016, "grad_norm": 1.6663934828043208, "learning_rate": 1.645421275215219e-05, "loss": 0.8476, "step": 1218 }, { "epoch": 0.8969830757910228, "grad_norm": 1.692340646576524, "learning_rate": 1.6448141841871262e-05, "loss": 0.8962, "step": 1219 }, { "epoch": 0.8977189109639441, "grad_norm": 1.7630715496964604, "learning_rate": 1.6442066860966466e-05, "loss": 0.8915, "step": 1220 }, { "epoch": 0.8984547461368654, "grad_norm": 1.8451269973562627, "learning_rate": 1.6435987813272854e-05, "loss": 0.922, "step": 1221 }, { "epoch": 0.8991905813097866, "grad_norm": 1.7106557654966517, "learning_rate": 1.6429904702628044e-05, "loss": 0.8755, "step": 1222 }, { "epoch": 0.8999264164827079, "grad_norm": 1.663563319478606, "learning_rate": 1.642381753287223e-05, "loss": 0.8933, "step": 1223 }, { "epoch": 0.9006622516556292, "grad_norm": 1.6098266566452857, "learning_rate": 1.6417726307848143e-05, "loss": 0.8371, "step": 1224 }, { "epoch": 0.9013980868285504, "grad_norm": 1.8085298904410378, "learning_rate": 1.64116310314011e-05, "loss": 0.868, "step": 1225 }, { "epoch": 0.9021339220014717, "grad_norm": 1.6785380601913962, "learning_rate": 1.6405531707378965e-05, "loss": 0.9028, "step": 1226 }, { "epoch": 0.9028697571743929, "grad_norm": 1.666236942421899, "learning_rate": 1.639942833963216e-05, "loss": 0.8791, "step": 1227 }, { "epoch": 0.9036055923473142, "grad_norm": 1.6317077892545375, "learning_rate": 1.639332093201365e-05, "loss": 0.8698, "step": 1228 }, { "epoch": 0.9043414275202355, "grad_norm": 1.6611013887304678, "learning_rate": 1.6387209488378966e-05, "loss": 0.9568, "step": 1229 }, { "epoch": 0.9050772626931567, "grad_norm": 1.6001101302659493, "learning_rate": 1.6381094012586176e-05, "loss": 0.8231, "step": 1230 }, { "epoch": 0.905813097866078, "grad_norm": 1.703943383647582, "learning_rate": 1.6374974508495895e-05, "loss": 0.8794, "step": 1231 }, { "epoch": 0.9065489330389993, "grad_norm": 1.7572312322246313, "learning_rate": 1.6368850979971287e-05, "loss": 0.8896, "step": 1232 }, { "epoch": 0.9072847682119205, "grad_norm": 1.6040778974244314, "learning_rate": 1.6362723430878045e-05, "loss": 0.8195, "step": 1233 }, { "epoch": 0.9080206033848418, "grad_norm": 1.5361068722609248, "learning_rate": 1.6356591865084413e-05, "loss": 0.8669, "step": 1234 }, { "epoch": 0.9087564385577631, "grad_norm": 1.6494381998799204, "learning_rate": 1.635045628646116e-05, "loss": 0.8781, "step": 1235 }, { "epoch": 0.9094922737306843, "grad_norm": 1.575525596016636, "learning_rate": 1.6344316698881595e-05, "loss": 0.8407, "step": 1236 }, { "epoch": 0.9102281089036056, "grad_norm": 1.6755394357231512, "learning_rate": 1.633817310622156e-05, "loss": 0.861, "step": 1237 }, { "epoch": 0.9109639440765268, "grad_norm": 1.658150925089117, "learning_rate": 1.6332025512359412e-05, "loss": 0.8325, "step": 1238 }, { "epoch": 0.9116997792494481, "grad_norm": 1.7246494054420414, "learning_rate": 1.632587392117605e-05, "loss": 0.8793, "step": 1239 }, { "epoch": 0.9124356144223694, "grad_norm": 1.5853902338497228, "learning_rate": 1.631971833655489e-05, "loss": 0.8772, "step": 1240 }, { "epoch": 0.9131714495952906, "grad_norm": 1.63522313444543, "learning_rate": 1.6313558762381865e-05, "loss": 0.8936, "step": 1241 }, { "epoch": 0.9139072847682119, "grad_norm": 1.6087724388000149, "learning_rate": 1.6307395202545427e-05, "loss": 0.8742, "step": 1242 }, { "epoch": 0.9146431199411332, "grad_norm": 2.311764385894833, "learning_rate": 1.630122766093656e-05, "loss": 0.8741, "step": 1243 }, { "epoch": 0.9153789551140544, "grad_norm": 1.9274776359460701, "learning_rate": 1.6295056141448734e-05, "loss": 0.9067, "step": 1244 }, { "epoch": 0.9161147902869757, "grad_norm": 1.7044887455898874, "learning_rate": 1.6288880647977953e-05, "loss": 0.8536, "step": 1245 }, { "epoch": 0.9168506254598969, "grad_norm": 1.609894710470609, "learning_rate": 1.6282701184422717e-05, "loss": 0.8438, "step": 1246 }, { "epoch": 0.9175864606328182, "grad_norm": 1.7207960947572836, "learning_rate": 1.6276517754684044e-05, "loss": 0.8323, "step": 1247 }, { "epoch": 0.9183222958057395, "grad_norm": 1.5703486891287204, "learning_rate": 1.627033036266544e-05, "loss": 0.829, "step": 1248 }, { "epoch": 0.9190581309786607, "grad_norm": 1.7493941145194132, "learning_rate": 1.6264139012272927e-05, "loss": 0.9007, "step": 1249 }, { "epoch": 0.919793966151582, "grad_norm": 1.6485329929053676, "learning_rate": 1.625794370741502e-05, "loss": 0.8566, "step": 1250 }, { "epoch": 0.9205298013245033, "grad_norm": 1.6342838676541451, "learning_rate": 1.625174445200273e-05, "loss": 0.884, "step": 1251 }, { "epoch": 0.9212656364974245, "grad_norm": 1.5221471483793232, "learning_rate": 1.6245541249949558e-05, "loss": 0.8641, "step": 1252 }, { "epoch": 0.9220014716703459, "grad_norm": 1.5834614164262069, "learning_rate": 1.6239334105171506e-05, "loss": 0.8497, "step": 1253 }, { "epoch": 0.9227373068432672, "grad_norm": 1.8304292527355235, "learning_rate": 1.6233123021587057e-05, "loss": 0.8936, "step": 1254 }, { "epoch": 0.9234731420161884, "grad_norm": 1.5703630788197824, "learning_rate": 1.622690800311718e-05, "loss": 0.8641, "step": 1255 }, { "epoch": 0.9242089771891097, "grad_norm": 1.6243127340822647, "learning_rate": 1.622068905368534e-05, "loss": 0.8389, "step": 1256 }, { "epoch": 0.9249448123620309, "grad_norm": 1.6306055640525847, "learning_rate": 1.621446617721747e-05, "loss": 0.8536, "step": 1257 }, { "epoch": 0.9256806475349522, "grad_norm": 1.5102031041823758, "learning_rate": 1.620823937764198e-05, "loss": 0.8829, "step": 1258 }, { "epoch": 0.9264164827078735, "grad_norm": 1.6988717298160818, "learning_rate": 1.6202008658889776e-05, "loss": 0.8928, "step": 1259 }, { "epoch": 0.9271523178807947, "grad_norm": 1.6303144983867193, "learning_rate": 1.619577402489422e-05, "loss": 0.8542, "step": 1260 }, { "epoch": 0.927888153053716, "grad_norm": 1.6125876077647925, "learning_rate": 1.618953547959115e-05, "loss": 0.8724, "step": 1261 }, { "epoch": 0.9286239882266373, "grad_norm": 1.7049261039615058, "learning_rate": 1.6183293026918875e-05, "loss": 0.8965, "step": 1262 }, { "epoch": 0.9293598233995585, "grad_norm": 1.664825445122122, "learning_rate": 1.617704667081817e-05, "loss": 0.8434, "step": 1263 }, { "epoch": 0.9300956585724798, "grad_norm": 1.582617038769996, "learning_rate": 1.6170796415232278e-05, "loss": 0.862, "step": 1264 }, { "epoch": 0.9308314937454011, "grad_norm": 1.8111074745037288, "learning_rate": 1.6164542264106895e-05, "loss": 0.8835, "step": 1265 }, { "epoch": 0.9315673289183223, "grad_norm": 1.6888370913060013, "learning_rate": 1.615828422139018e-05, "loss": 0.874, "step": 1266 }, { "epoch": 0.9323031640912436, "grad_norm": 1.7015885965001105, "learning_rate": 1.615202229103276e-05, "loss": 0.8904, "step": 1267 }, { "epoch": 0.9330389992641648, "grad_norm": 1.6838509775799142, "learning_rate": 1.6145756476987692e-05, "loss": 0.9168, "step": 1268 }, { "epoch": 0.9337748344370861, "grad_norm": 1.661998085657079, "learning_rate": 1.613948678321051e-05, "loss": 0.8366, "step": 1269 }, { "epoch": 0.9345106696100074, "grad_norm": 1.8157596833499021, "learning_rate": 1.613321321365918e-05, "loss": 0.9032, "step": 1270 }, { "epoch": 0.9352465047829286, "grad_norm": 1.5642086589678073, "learning_rate": 1.6126935772294127e-05, "loss": 0.8436, "step": 1271 }, { "epoch": 0.9359823399558499, "grad_norm": 1.6900261294841192, "learning_rate": 1.612065446307821e-05, "loss": 0.9066, "step": 1272 }, { "epoch": 0.9367181751287712, "grad_norm": 1.7005999586652023, "learning_rate": 1.6114369289976727e-05, "loss": 0.9077, "step": 1273 }, { "epoch": 0.9374540103016924, "grad_norm": 1.6233076055400244, "learning_rate": 1.6108080256957434e-05, "loss": 0.8923, "step": 1274 }, { "epoch": 0.9381898454746137, "grad_norm": 1.5894642425770142, "learning_rate": 1.6101787367990503e-05, "loss": 0.8732, "step": 1275 }, { "epoch": 0.9389256806475349, "grad_norm": 1.680277310521377, "learning_rate": 1.609549062704855e-05, "loss": 0.827, "step": 1276 }, { "epoch": 0.9396615158204562, "grad_norm": 1.7499567528972784, "learning_rate": 1.6089190038106628e-05, "loss": 0.8636, "step": 1277 }, { "epoch": 0.9403973509933775, "grad_norm": 1.6885177683802415, "learning_rate": 1.608288560514221e-05, "loss": 0.8668, "step": 1278 }, { "epoch": 0.9411331861662987, "grad_norm": 1.672896623182125, "learning_rate": 1.607657733213519e-05, "loss": 0.8889, "step": 1279 }, { "epoch": 0.94186902133922, "grad_norm": 1.6281274720098766, "learning_rate": 1.6070265223067904e-05, "loss": 0.8755, "step": 1280 }, { "epoch": 0.9426048565121413, "grad_norm": 1.6679572677108152, "learning_rate": 1.6063949281925095e-05, "loss": 0.8509, "step": 1281 }, { "epoch": 0.9433406916850625, "grad_norm": 1.7695484020504688, "learning_rate": 1.6057629512693938e-05, "loss": 0.9189, "step": 1282 }, { "epoch": 0.9440765268579838, "grad_norm": 1.6094859614012644, "learning_rate": 1.6051305919364008e-05, "loss": 0.8794, "step": 1283 }, { "epoch": 0.9448123620309051, "grad_norm": 1.6903468156428685, "learning_rate": 1.604497850592731e-05, "loss": 0.8928, "step": 1284 }, { "epoch": 0.9455481972038263, "grad_norm": 1.5614472250024436, "learning_rate": 1.6038647276378246e-05, "loss": 0.8647, "step": 1285 }, { "epoch": 0.9462840323767476, "grad_norm": 1.52051434450752, "learning_rate": 1.603231223471364e-05, "loss": 0.8461, "step": 1286 }, { "epoch": 0.9470198675496688, "grad_norm": 1.607535431925128, "learning_rate": 1.6025973384932722e-05, "loss": 0.8594, "step": 1287 }, { "epoch": 0.9477557027225901, "grad_norm": 1.641622512401732, "learning_rate": 1.601963073103711e-05, "loss": 0.8147, "step": 1288 }, { "epoch": 0.9484915378955114, "grad_norm": 1.6998158087302542, "learning_rate": 1.6013284277030844e-05, "loss": 0.8268, "step": 1289 }, { "epoch": 0.9492273730684326, "grad_norm": 1.7441194033702334, "learning_rate": 1.6006934026920347e-05, "loss": 0.7887, "step": 1290 }, { "epoch": 0.9499632082413539, "grad_norm": 1.666736639134981, "learning_rate": 1.6000579984714453e-05, "loss": 0.8601, "step": 1291 }, { "epoch": 0.9506990434142752, "grad_norm": 1.593701329215587, "learning_rate": 1.5994222154424376e-05, "loss": 0.8596, "step": 1292 }, { "epoch": 0.9514348785871964, "grad_norm": 1.6474527489110804, "learning_rate": 1.5987860540063725e-05, "loss": 0.8691, "step": 1293 }, { "epoch": 0.9521707137601177, "grad_norm": 1.676381664117038, "learning_rate": 1.5981495145648507e-05, "loss": 0.8375, "step": 1294 }, { "epoch": 0.9529065489330389, "grad_norm": 1.6651640849791232, "learning_rate": 1.5975125975197107e-05, "loss": 0.8428, "step": 1295 }, { "epoch": 0.9536423841059603, "grad_norm": 1.6230767505687993, "learning_rate": 1.596875303273029e-05, "loss": 0.8453, "step": 1296 }, { "epoch": 0.9543782192788816, "grad_norm": 1.6341139368181008, "learning_rate": 1.5962376322271218e-05, "loss": 0.8939, "step": 1297 }, { "epoch": 0.9551140544518028, "grad_norm": 1.580912577416275, "learning_rate": 1.5955995847845415e-05, "loss": 0.8702, "step": 1298 }, { "epoch": 0.9558498896247241, "grad_norm": 1.6845602047275792, "learning_rate": 1.5949611613480784e-05, "loss": 0.8818, "step": 1299 }, { "epoch": 0.9565857247976454, "grad_norm": 1.6170527562041035, "learning_rate": 1.5943223623207608e-05, "loss": 0.8836, "step": 1300 }, { "epoch": 0.9573215599705666, "grad_norm": 1.5802418331986987, "learning_rate": 1.5936831881058533e-05, "loss": 0.8775, "step": 1301 }, { "epoch": 0.9580573951434879, "grad_norm": 1.7797669857682274, "learning_rate": 1.593043639106859e-05, "loss": 0.9234, "step": 1302 }, { "epoch": 0.9587932303164092, "grad_norm": 1.6394385386284716, "learning_rate": 1.5924037157275156e-05, "loss": 0.8783, "step": 1303 }, { "epoch": 0.9595290654893304, "grad_norm": 1.833387428218918, "learning_rate": 1.5917634183717985e-05, "loss": 0.8741, "step": 1304 }, { "epoch": 0.9602649006622517, "grad_norm": 1.7852045065171982, "learning_rate": 1.591122747443918e-05, "loss": 0.9114, "step": 1305 }, { "epoch": 0.9610007358351729, "grad_norm": 1.721123718917073, "learning_rate": 1.5904817033483216e-05, "loss": 0.8187, "step": 1306 }, { "epoch": 0.9617365710080942, "grad_norm": 1.6463829855498104, "learning_rate": 1.5898402864896908e-05, "loss": 0.851, "step": 1307 }, { "epoch": 0.9624724061810155, "grad_norm": 1.707400093466762, "learning_rate": 1.5891984972729448e-05, "loss": 0.9106, "step": 1308 }, { "epoch": 0.9632082413539367, "grad_norm": 1.6469327198901387, "learning_rate": 1.588556336103235e-05, "loss": 0.8762, "step": 1309 }, { "epoch": 0.963944076526858, "grad_norm": 1.5691236352707765, "learning_rate": 1.587913803385949e-05, "loss": 0.7902, "step": 1310 }, { "epoch": 0.9646799116997793, "grad_norm": 1.6712245851305203, "learning_rate": 1.5872708995267094e-05, "loss": 0.9158, "step": 1311 }, { "epoch": 0.9654157468727005, "grad_norm": 1.6049407730498289, "learning_rate": 1.586627624931373e-05, "loss": 0.865, "step": 1312 }, { "epoch": 0.9661515820456218, "grad_norm": 1.680470651582958, "learning_rate": 1.5859839800060295e-05, "loss": 0.858, "step": 1313 }, { "epoch": 0.9668874172185431, "grad_norm": 1.701944942526223, "learning_rate": 1.5853399651570036e-05, "loss": 0.9006, "step": 1314 }, { "epoch": 0.9676232523914643, "grad_norm": 1.6662591444272359, "learning_rate": 1.584695580790853e-05, "loss": 0.8577, "step": 1315 }, { "epoch": 0.9683590875643856, "grad_norm": 1.732795235066981, "learning_rate": 1.5840508273143687e-05, "loss": 0.8853, "step": 1316 }, { "epoch": 0.9690949227373068, "grad_norm": 1.5645619619625495, "learning_rate": 1.5834057051345745e-05, "loss": 0.8749, "step": 1317 }, { "epoch": 0.9698307579102281, "grad_norm": 1.6409321717815217, "learning_rate": 1.5827602146587277e-05, "loss": 0.8583, "step": 1318 }, { "epoch": 0.9705665930831494, "grad_norm": 1.5428926009857327, "learning_rate": 1.582114356294317e-05, "loss": 0.8608, "step": 1319 }, { "epoch": 0.9713024282560706, "grad_norm": 1.659315765939285, "learning_rate": 1.5814681304490653e-05, "loss": 0.8786, "step": 1320 }, { "epoch": 0.9720382634289919, "grad_norm": 1.6148126020978948, "learning_rate": 1.5808215375309243e-05, "loss": 0.8767, "step": 1321 }, { "epoch": 0.9727740986019132, "grad_norm": 1.7070694843745031, "learning_rate": 1.5801745779480807e-05, "loss": 0.8931, "step": 1322 }, { "epoch": 0.9735099337748344, "grad_norm": 1.7705083455043966, "learning_rate": 1.5795272521089503e-05, "loss": 0.897, "step": 1323 }, { "epoch": 0.9742457689477557, "grad_norm": 1.5971326824880283, "learning_rate": 1.578879560422182e-05, "loss": 0.8573, "step": 1324 }, { "epoch": 0.9749816041206769, "grad_norm": 1.8019419493088522, "learning_rate": 1.578231503296654e-05, "loss": 0.8913, "step": 1325 }, { "epoch": 0.9757174392935982, "grad_norm": 1.7456412220535933, "learning_rate": 1.5775830811414762e-05, "loss": 0.8858, "step": 1326 }, { "epoch": 0.9764532744665195, "grad_norm": 1.6337227360079416, "learning_rate": 1.576934294365988e-05, "loss": 0.8704, "step": 1327 }, { "epoch": 0.9771891096394407, "grad_norm": 1.5180483403633172, "learning_rate": 1.576285143379761e-05, "loss": 0.876, "step": 1328 }, { "epoch": 0.977924944812362, "grad_norm": 1.6024412672155344, "learning_rate": 1.5756356285925943e-05, "loss": 0.8111, "step": 1329 }, { "epoch": 0.9786607799852833, "grad_norm": 1.6154350063860536, "learning_rate": 1.574985750414518e-05, "loss": 0.8695, "step": 1330 }, { "epoch": 0.9793966151582045, "grad_norm": 1.5075980649703777, "learning_rate": 1.5743355092557912e-05, "loss": 0.8384, "step": 1331 }, { "epoch": 0.9801324503311258, "grad_norm": 1.6032874772000885, "learning_rate": 1.5736849055269024e-05, "loss": 0.8963, "step": 1332 }, { "epoch": 0.9808682855040471, "grad_norm": 1.846795242544621, "learning_rate": 1.5730339396385684e-05, "loss": 0.8577, "step": 1333 }, { "epoch": 0.9816041206769683, "grad_norm": 1.7919915666918487, "learning_rate": 1.5723826120017363e-05, "loss": 0.8794, "step": 1334 }, { "epoch": 0.9823399558498896, "grad_norm": 1.6525546436710281, "learning_rate": 1.571730923027579e-05, "loss": 0.8259, "step": 1335 }, { "epoch": 0.9830757910228108, "grad_norm": 1.6561300591617003, "learning_rate": 1.571078873127499e-05, "loss": 0.8925, "step": 1336 }, { "epoch": 0.9838116261957321, "grad_norm": 1.8063479298692757, "learning_rate": 1.5704264627131274e-05, "loss": 0.8526, "step": 1337 }, { "epoch": 0.9845474613686535, "grad_norm": 1.5161949045524086, "learning_rate": 1.5697736921963212e-05, "loss": 0.8486, "step": 1338 }, { "epoch": 0.9852832965415746, "grad_norm": 1.6620034756834448, "learning_rate": 1.569120561989166e-05, "loss": 0.8664, "step": 1339 }, { "epoch": 0.986019131714496, "grad_norm": 1.7497658623310905, "learning_rate": 1.5684670725039736e-05, "loss": 0.849, "step": 1340 }, { "epoch": 0.9867549668874173, "grad_norm": 1.741350375764251, "learning_rate": 1.567813224153283e-05, "loss": 0.8416, "step": 1341 }, { "epoch": 0.9874908020603385, "grad_norm": 1.5691725054228542, "learning_rate": 1.5671590173498602e-05, "loss": 0.8795, "step": 1342 }, { "epoch": 0.9882266372332598, "grad_norm": 1.7529060717257474, "learning_rate": 1.5665044525066963e-05, "loss": 0.9061, "step": 1343 }, { "epoch": 0.9889624724061811, "grad_norm": 1.636105093613301, "learning_rate": 1.5658495300370093e-05, "loss": 0.895, "step": 1344 }, { "epoch": 0.9896983075791023, "grad_norm": 1.6955325554768137, "learning_rate": 1.5651942503542435e-05, "loss": 0.862, "step": 1345 }, { "epoch": 0.9904341427520236, "grad_norm": 1.5243116753356085, "learning_rate": 1.5645386138720672e-05, "loss": 0.808, "step": 1346 }, { "epoch": 0.9911699779249448, "grad_norm": 1.5798730797155465, "learning_rate": 1.5638826210043752e-05, "loss": 0.8388, "step": 1347 }, { "epoch": 0.9919058130978661, "grad_norm": 1.5234690739539196, "learning_rate": 1.563226272165287e-05, "loss": 0.8441, "step": 1348 }, { "epoch": 0.9926416482707874, "grad_norm": 1.6326736707194998, "learning_rate": 1.562569567769146e-05, "loss": 0.8439, "step": 1349 }, { "epoch": 0.9933774834437086, "grad_norm": 1.8130786307215085, "learning_rate": 1.561912508230521e-05, "loss": 0.8826, "step": 1350 }, { "epoch": 0.9941133186166299, "grad_norm": 1.553947357143518, "learning_rate": 1.561255093964205e-05, "loss": 0.8278, "step": 1351 }, { "epoch": 0.9948491537895512, "grad_norm": 1.7147704962117294, "learning_rate": 1.5605973253852148e-05, "loss": 0.873, "step": 1352 }, { "epoch": 0.9955849889624724, "grad_norm": 1.586479071109721, "learning_rate": 1.5599392029087897e-05, "loss": 0.843, "step": 1353 }, { "epoch": 0.9963208241353937, "grad_norm": 1.6843666372385968, "learning_rate": 1.559280726950395e-05, "loss": 0.8292, "step": 1354 }, { "epoch": 0.9970566593083149, "grad_norm": 1.5559448319593558, "learning_rate": 1.5586218979257158e-05, "loss": 0.8517, "step": 1355 }, { "epoch": 0.9977924944812362, "grad_norm": 1.6081185108126337, "learning_rate": 1.557962716250663e-05, "loss": 0.8437, "step": 1356 }, { "epoch": 0.9985283296541575, "grad_norm": 1.6161904850867905, "learning_rate": 1.557303182341369e-05, "loss": 0.85, "step": 1357 }, { "epoch": 0.9992641648270787, "grad_norm": 1.6955881467843927, "learning_rate": 1.556643296614188e-05, "loss": 0.8614, "step": 1358 }, { "epoch": 1.0, "grad_norm": 1.6515905679635714, "learning_rate": 1.5559830594856973e-05, "loss": 0.8284, "step": 1359 }, { "epoch": 1.0007358351729212, "grad_norm": 1.782079387006564, "learning_rate": 1.5553224713726954e-05, "loss": 0.5822, "step": 1360 }, { "epoch": 1.0014716703458426, "grad_norm": 1.6887235430913803, "learning_rate": 1.5546615326922024e-05, "loss": 0.6441, "step": 1361 }, { "epoch": 1.0022075055187638, "grad_norm": 1.6233598610323619, "learning_rate": 1.5540002438614605e-05, "loss": 0.6119, "step": 1362 }, { "epoch": 1.002943340691685, "grad_norm": 1.5704301863890955, "learning_rate": 1.553338605297931e-05, "loss": 0.6174, "step": 1363 }, { "epoch": 1.0036791758646064, "grad_norm": 1.5383420625044097, "learning_rate": 1.552676617419299e-05, "loss": 0.5497, "step": 1364 }, { "epoch": 1.0044150110375276, "grad_norm": 1.8738474635167315, "learning_rate": 1.5520142806434667e-05, "loss": 0.6109, "step": 1365 }, { "epoch": 1.0051508462104488, "grad_norm": 1.8662583772828845, "learning_rate": 1.55135159538856e-05, "loss": 0.5732, "step": 1366 }, { "epoch": 1.0058866813833702, "grad_norm": 1.960140064274908, "learning_rate": 1.5506885620729215e-05, "loss": 0.5649, "step": 1367 }, { "epoch": 1.0066225165562914, "grad_norm": 2.005060599719176, "learning_rate": 1.550025181115116e-05, "loss": 0.613, "step": 1368 }, { "epoch": 1.0073583517292126, "grad_norm": 1.8099834396788288, "learning_rate": 1.549361452933926e-05, "loss": 0.5692, "step": 1369 }, { "epoch": 1.0080941869021338, "grad_norm": 1.6824608657748283, "learning_rate": 1.548697377948355e-05, "loss": 0.543, "step": 1370 }, { "epoch": 1.0088300220750552, "grad_norm": 1.6922837082813313, "learning_rate": 1.548032956577624e-05, "loss": 0.581, "step": 1371 }, { "epoch": 1.0095658572479764, "grad_norm": 1.6450883958239972, "learning_rate": 1.5473681892411733e-05, "loss": 0.573, "step": 1372 }, { "epoch": 1.0103016924208976, "grad_norm": 1.7004120441960857, "learning_rate": 1.546703076358661e-05, "loss": 0.5747, "step": 1373 }, { "epoch": 1.011037527593819, "grad_norm": 1.5924659516496262, "learning_rate": 1.5460376183499645e-05, "loss": 0.6099, "step": 1374 }, { "epoch": 1.0117733627667402, "grad_norm": 1.6487549231617062, "learning_rate": 1.5453718156351775e-05, "loss": 0.5769, "step": 1375 }, { "epoch": 1.0125091979396614, "grad_norm": 1.5823831935247734, "learning_rate": 1.544705668634613e-05, "loss": 0.5319, "step": 1376 }, { "epoch": 1.0132450331125828, "grad_norm": 1.524542507630712, "learning_rate": 1.5440391777688e-05, "loss": 0.5301, "step": 1377 }, { "epoch": 1.013980868285504, "grad_norm": 1.7114794160709876, "learning_rate": 1.543372343458485e-05, "loss": 0.5803, "step": 1378 }, { "epoch": 1.0147167034584252, "grad_norm": 1.7529955422262211, "learning_rate": 1.542705166124632e-05, "loss": 0.5764, "step": 1379 }, { "epoch": 1.0154525386313467, "grad_norm": 1.7926566996002815, "learning_rate": 1.5420376461884203e-05, "loss": 0.5987, "step": 1380 }, { "epoch": 1.0161883738042679, "grad_norm": 1.6883168760944591, "learning_rate": 1.541369784071246e-05, "loss": 0.5631, "step": 1381 }, { "epoch": 1.016924208977189, "grad_norm": 1.6880484692792852, "learning_rate": 1.5407015801947224e-05, "loss": 0.5802, "step": 1382 }, { "epoch": 1.0176600441501105, "grad_norm": 1.8197674885899708, "learning_rate": 1.540033034980676e-05, "loss": 0.5698, "step": 1383 }, { "epoch": 1.0183958793230317, "grad_norm": 1.6042665320259755, "learning_rate": 1.5393641488511514e-05, "loss": 0.5545, "step": 1384 }, { "epoch": 1.0191317144959529, "grad_norm": 1.7333612911154268, "learning_rate": 1.5386949222284072e-05, "loss": 0.5708, "step": 1385 }, { "epoch": 1.0198675496688743, "grad_norm": 1.7285364244296912, "learning_rate": 1.5380253555349163e-05, "loss": 0.5878, "step": 1386 }, { "epoch": 1.0206033848417955, "grad_norm": 1.742723813375258, "learning_rate": 1.537355449193367e-05, "loss": 0.6019, "step": 1387 }, { "epoch": 1.0213392200147167, "grad_norm": 1.7256907153541874, "learning_rate": 1.5366852036266636e-05, "loss": 0.5968, "step": 1388 }, { "epoch": 1.022075055187638, "grad_norm": 1.7757494246067083, "learning_rate": 1.536014619257921e-05, "loss": 0.6127, "step": 1389 }, { "epoch": 1.0228108903605593, "grad_norm": 1.6204138052183168, "learning_rate": 1.5353436965104708e-05, "loss": 0.5678, "step": 1390 }, { "epoch": 1.0235467255334805, "grad_norm": 1.5657742082584674, "learning_rate": 1.5346724358078576e-05, "loss": 0.5709, "step": 1391 }, { "epoch": 1.0242825607064017, "grad_norm": 1.6576821976972977, "learning_rate": 1.5340008375738385e-05, "loss": 0.5793, "step": 1392 }, { "epoch": 1.025018395879323, "grad_norm": 1.758008671060223, "learning_rate": 1.533328902232385e-05, "loss": 0.5904, "step": 1393 }, { "epoch": 1.0257542310522443, "grad_norm": 1.6061760501231, "learning_rate": 1.53265663020768e-05, "loss": 0.5966, "step": 1394 }, { "epoch": 1.0264900662251655, "grad_norm": 1.602857073116805, "learning_rate": 1.5319840219241205e-05, "loss": 0.5928, "step": 1395 }, { "epoch": 1.027225901398087, "grad_norm": 1.6702467959142109, "learning_rate": 1.5313110778063142e-05, "loss": 0.6051, "step": 1396 }, { "epoch": 1.027961736571008, "grad_norm": 1.6892664878480965, "learning_rate": 1.5306377982790818e-05, "loss": 0.5842, "step": 1397 }, { "epoch": 1.0286975717439293, "grad_norm": 1.7014488397602208, "learning_rate": 1.529964183767455e-05, "loss": 0.5757, "step": 1398 }, { "epoch": 1.0294334069168507, "grad_norm": 1.5223641591232853, "learning_rate": 1.5292902346966782e-05, "loss": 0.56, "step": 1399 }, { "epoch": 1.030169242089772, "grad_norm": 1.5835810290615249, "learning_rate": 1.528615951492206e-05, "loss": 0.5782, "step": 1400 }, { "epoch": 1.030905077262693, "grad_norm": 1.6007806240067257, "learning_rate": 1.5279413345797037e-05, "loss": 0.5918, "step": 1401 }, { "epoch": 1.0316409124356145, "grad_norm": 1.706083245597314, "learning_rate": 1.5272663843850484e-05, "loss": 0.5916, "step": 1402 }, { "epoch": 1.0323767476085357, "grad_norm": 1.6154185026024042, "learning_rate": 1.5265911013343265e-05, "loss": 0.5731, "step": 1403 }, { "epoch": 1.033112582781457, "grad_norm": 1.6979515609089855, "learning_rate": 1.5259154858538354e-05, "loss": 0.6037, "step": 1404 }, { "epoch": 1.0338484179543783, "grad_norm": 1.723079066677853, "learning_rate": 1.5252395383700815e-05, "loss": 0.5323, "step": 1405 }, { "epoch": 1.0345842531272995, "grad_norm": 1.7238411567174703, "learning_rate": 1.5245632593097814e-05, "loss": 0.598, "step": 1406 }, { "epoch": 1.0353200883002207, "grad_norm": 1.6490576329298072, "learning_rate": 1.5238866490998607e-05, "loss": 0.5838, "step": 1407 }, { "epoch": 1.0360559234731421, "grad_norm": 1.6882993369758694, "learning_rate": 1.5232097081674542e-05, "loss": 0.5656, "step": 1408 }, { "epoch": 1.0367917586460633, "grad_norm": 1.6257161886084623, "learning_rate": 1.5225324369399058e-05, "loss": 0.5785, "step": 1409 }, { "epoch": 1.0375275938189845, "grad_norm": 1.6309056613116781, "learning_rate": 1.5218548358447668e-05, "loss": 0.5586, "step": 1410 }, { "epoch": 1.0382634289919057, "grad_norm": 1.7396787733974617, "learning_rate": 1.5211769053097984e-05, "loss": 0.585, "step": 1411 }, { "epoch": 1.0389992641648271, "grad_norm": 1.6962561698247274, "learning_rate": 1.5204986457629685e-05, "loss": 0.6273, "step": 1412 }, { "epoch": 1.0397350993377483, "grad_norm": 1.5618875482206627, "learning_rate": 1.5198200576324534e-05, "loss": 0.5622, "step": 1413 }, { "epoch": 1.0404709345106695, "grad_norm": 1.6069903648800956, "learning_rate": 1.5191411413466357e-05, "loss": 0.5868, "step": 1414 }, { "epoch": 1.041206769683591, "grad_norm": 1.5728078826754839, "learning_rate": 1.5184618973341068e-05, "loss": 0.5877, "step": 1415 }, { "epoch": 1.0419426048565121, "grad_norm": 1.6383176406942026, "learning_rate": 1.5177823260236636e-05, "loss": 0.5694, "step": 1416 }, { "epoch": 1.0426784400294333, "grad_norm": 1.6297227930335052, "learning_rate": 1.5171024278443108e-05, "loss": 0.5796, "step": 1417 }, { "epoch": 1.0434142752023547, "grad_norm": 1.6748597926021185, "learning_rate": 1.5164222032252585e-05, "loss": 0.5559, "step": 1418 }, { "epoch": 1.044150110375276, "grad_norm": 1.5158863555291358, "learning_rate": 1.5157416525959227e-05, "loss": 0.5927, "step": 1419 }, { "epoch": 1.0448859455481971, "grad_norm": 1.5763563121707245, "learning_rate": 1.5150607763859266e-05, "loss": 0.5257, "step": 1420 }, { "epoch": 1.0456217807211186, "grad_norm": 1.63524824180237, "learning_rate": 1.5143795750250976e-05, "loss": 0.5929, "step": 1421 }, { "epoch": 1.0463576158940397, "grad_norm": 1.7529516417524882, "learning_rate": 1.5136980489434691e-05, "loss": 0.6105, "step": 1422 }, { "epoch": 1.047093451066961, "grad_norm": 1.7563447589946959, "learning_rate": 1.5130161985712786e-05, "loss": 0.6112, "step": 1423 }, { "epoch": 1.0478292862398824, "grad_norm": 1.7237680383831018, "learning_rate": 1.5123340243389697e-05, "loss": 0.5675, "step": 1424 }, { "epoch": 1.0485651214128036, "grad_norm": 1.6230600795688306, "learning_rate": 1.5116515266771888e-05, "loss": 0.5726, "step": 1425 }, { "epoch": 1.0493009565857248, "grad_norm": 1.6525898581140253, "learning_rate": 1.510968706016788e-05, "loss": 0.6253, "step": 1426 }, { "epoch": 1.0500367917586462, "grad_norm": 1.6315661324131145, "learning_rate": 1.5102855627888228e-05, "loss": 0.5655, "step": 1427 }, { "epoch": 1.0507726269315674, "grad_norm": 1.5481636196017934, "learning_rate": 1.5096020974245518e-05, "loss": 0.5783, "step": 1428 }, { "epoch": 1.0515084621044886, "grad_norm": 1.628197425947137, "learning_rate": 1.5089183103554372e-05, "loss": 0.5743, "step": 1429 }, { "epoch": 1.0522442972774098, "grad_norm": 1.7575925371088177, "learning_rate": 1.5082342020131448e-05, "loss": 0.5948, "step": 1430 }, { "epoch": 1.0529801324503312, "grad_norm": 1.716659217582222, "learning_rate": 1.5075497728295427e-05, "loss": 0.5763, "step": 1431 }, { "epoch": 1.0537159676232524, "grad_norm": 1.7971200513415775, "learning_rate": 1.506865023236702e-05, "loss": 0.5627, "step": 1432 }, { "epoch": 1.0544518027961736, "grad_norm": 1.7167561129667523, "learning_rate": 1.5061799536668951e-05, "loss": 0.5931, "step": 1433 }, { "epoch": 1.055187637969095, "grad_norm": 1.7717743369074668, "learning_rate": 1.5054945645525973e-05, "loss": 0.5645, "step": 1434 }, { "epoch": 1.0559234731420162, "grad_norm": 1.6845420083537175, "learning_rate": 1.504808856326486e-05, "loss": 0.5805, "step": 1435 }, { "epoch": 1.0566593083149374, "grad_norm": 1.6461431367235213, "learning_rate": 1.5041228294214388e-05, "loss": 0.5657, "step": 1436 }, { "epoch": 1.0573951434878588, "grad_norm": 1.7663370638242508, "learning_rate": 1.5034364842705352e-05, "loss": 0.5572, "step": 1437 }, { "epoch": 1.05813097866078, "grad_norm": 1.6845808383539054, "learning_rate": 1.5027498213070558e-05, "loss": 0.5903, "step": 1438 }, { "epoch": 1.0588668138337012, "grad_norm": 1.5644535537658235, "learning_rate": 1.5020628409644813e-05, "loss": 0.549, "step": 1439 }, { "epoch": 1.0596026490066226, "grad_norm": 1.5752901700825785, "learning_rate": 1.5013755436764928e-05, "loss": 0.5901, "step": 1440 }, { "epoch": 1.0603384841795438, "grad_norm": 1.6913022108296392, "learning_rate": 1.5006879298769721e-05, "loss": 0.5766, "step": 1441 }, { "epoch": 1.061074319352465, "grad_norm": 1.6776978475756528, "learning_rate": 1.5000000000000002e-05, "loss": 0.5875, "step": 1442 }, { "epoch": 1.0618101545253864, "grad_norm": 1.760459177090869, "learning_rate": 1.4993117544798578e-05, "loss": 0.5793, "step": 1443 }, { "epoch": 1.0625459896983076, "grad_norm": 1.5803225455115626, "learning_rate": 1.498623193751025e-05, "loss": 0.5303, "step": 1444 }, { "epoch": 1.0632818248712288, "grad_norm": 1.7201610052611787, "learning_rate": 1.4979343182481812e-05, "loss": 0.592, "step": 1445 }, { "epoch": 1.0640176600441502, "grad_norm": 1.6890139877290564, "learning_rate": 1.497245128406203e-05, "loss": 0.6195, "step": 1446 }, { "epoch": 1.0647534952170714, "grad_norm": 1.655042286148294, "learning_rate": 1.4965556246601677e-05, "loss": 0.5568, "step": 1447 }, { "epoch": 1.0654893303899926, "grad_norm": 1.7308539340773277, "learning_rate": 1.495865807445349e-05, "loss": 0.5738, "step": 1448 }, { "epoch": 1.0662251655629138, "grad_norm": 1.671041422429769, "learning_rate": 1.495175677197219e-05, "loss": 0.5818, "step": 1449 }, { "epoch": 1.0669610007358352, "grad_norm": 1.710498157765487, "learning_rate": 1.4944852343514478e-05, "loss": 0.5911, "step": 1450 }, { "epoch": 1.0676968359087564, "grad_norm": 1.6792274242082414, "learning_rate": 1.4937944793439023e-05, "loss": 0.5712, "step": 1451 }, { "epoch": 1.0684326710816776, "grad_norm": 1.5327797191921197, "learning_rate": 1.4931034126106468e-05, "loss": 0.5592, "step": 1452 }, { "epoch": 1.069168506254599, "grad_norm": 1.64901615849467, "learning_rate": 1.4924120345879422e-05, "loss": 0.5755, "step": 1453 }, { "epoch": 1.0699043414275202, "grad_norm": 1.5434711805628378, "learning_rate": 1.4917203457122465e-05, "loss": 0.5827, "step": 1454 }, { "epoch": 1.0706401766004414, "grad_norm": 1.6329050403298664, "learning_rate": 1.4910283464202126e-05, "loss": 0.5899, "step": 1455 }, { "epoch": 1.0713760117733628, "grad_norm": 1.6261713360102217, "learning_rate": 1.490336037148691e-05, "loss": 0.5762, "step": 1456 }, { "epoch": 1.072111846946284, "grad_norm": 1.637300761395057, "learning_rate": 1.4896434183347266e-05, "loss": 0.5254, "step": 1457 }, { "epoch": 1.0728476821192052, "grad_norm": 1.7173751092330352, "learning_rate": 1.4889504904155602e-05, "loss": 0.5582, "step": 1458 }, { "epoch": 1.0735835172921266, "grad_norm": 1.6623116535658022, "learning_rate": 1.4882572538286279e-05, "loss": 0.5855, "step": 1459 }, { "epoch": 1.0743193524650478, "grad_norm": 1.7871525660825895, "learning_rate": 1.4875637090115605e-05, "loss": 0.5809, "step": 1460 }, { "epoch": 1.075055187637969, "grad_norm": 1.782027252455113, "learning_rate": 1.4868698564021833e-05, "loss": 0.6237, "step": 1461 }, { "epoch": 1.0757910228108905, "grad_norm": 1.7550346071806242, "learning_rate": 1.486175696438516e-05, "loss": 0.5699, "step": 1462 }, { "epoch": 1.0765268579838116, "grad_norm": 1.583003226237981, "learning_rate": 1.4854812295587722e-05, "loss": 0.5898, "step": 1463 }, { "epoch": 1.0772626931567328, "grad_norm": 1.674278699858556, "learning_rate": 1.4847864562013596e-05, "loss": 0.5688, "step": 1464 }, { "epoch": 1.0779985283296543, "grad_norm": 1.609663205752749, "learning_rate": 1.4840913768048788e-05, "loss": 0.5661, "step": 1465 }, { "epoch": 1.0787343635025755, "grad_norm": 1.6351739172816648, "learning_rate": 1.4833959918081244e-05, "loss": 0.6082, "step": 1466 }, { "epoch": 1.0794701986754967, "grad_norm": 1.6319884679811134, "learning_rate": 1.4827003016500827e-05, "loss": 0.598, "step": 1467 }, { "epoch": 1.0802060338484178, "grad_norm": 1.6821159191909472, "learning_rate": 1.4820043067699342e-05, "loss": 0.6088, "step": 1468 }, { "epoch": 1.0809418690213393, "grad_norm": 1.7785488842351798, "learning_rate": 1.4813080076070505e-05, "loss": 0.581, "step": 1469 }, { "epoch": 1.0816777041942605, "grad_norm": 1.7034808998247117, "learning_rate": 1.4806114046009958e-05, "loss": 0.6089, "step": 1470 }, { "epoch": 1.0824135393671817, "grad_norm": 1.6922388603607321, "learning_rate": 1.479914498191526e-05, "loss": 0.5898, "step": 1471 }, { "epoch": 1.083149374540103, "grad_norm": 1.6453811446185473, "learning_rate": 1.4792172888185891e-05, "loss": 0.5845, "step": 1472 }, { "epoch": 1.0838852097130243, "grad_norm": 1.6943380207465062, "learning_rate": 1.4785197769223234e-05, "loss": 0.6145, "step": 1473 }, { "epoch": 1.0846210448859455, "grad_norm": 1.731184482119423, "learning_rate": 1.4778219629430585e-05, "loss": 0.5986, "step": 1474 }, { "epoch": 1.0853568800588669, "grad_norm": 1.6999622112292205, "learning_rate": 1.4771238473213153e-05, "loss": 0.6108, "step": 1475 }, { "epoch": 1.086092715231788, "grad_norm": 1.652919538915065, "learning_rate": 1.4764254304978042e-05, "loss": 0.5996, "step": 1476 }, { "epoch": 1.0868285504047093, "grad_norm": 1.6050634003766855, "learning_rate": 1.4757267129134266e-05, "loss": 0.5678, "step": 1477 }, { "epoch": 1.0875643855776307, "grad_norm": 1.578419437273882, "learning_rate": 1.4750276950092728e-05, "loss": 0.581, "step": 1478 }, { "epoch": 1.0883002207505519, "grad_norm": 1.6858773863097256, "learning_rate": 1.4743283772266238e-05, "loss": 0.5518, "step": 1479 }, { "epoch": 1.089036055923473, "grad_norm": 1.7106942290416567, "learning_rate": 1.4736287600069493e-05, "loss": 0.603, "step": 1480 }, { "epoch": 1.0897718910963945, "grad_norm": 1.6900890102712007, "learning_rate": 1.4729288437919078e-05, "loss": 0.614, "step": 1481 }, { "epoch": 1.0905077262693157, "grad_norm": 1.629569628389968, "learning_rate": 1.4722286290233474e-05, "loss": 0.5923, "step": 1482 }, { "epoch": 1.0912435614422369, "grad_norm": 1.7099822135400438, "learning_rate": 1.4715281161433032e-05, "loss": 0.5808, "step": 1483 }, { "epoch": 1.0919793966151583, "grad_norm": 1.7199280475908123, "learning_rate": 1.4708273055940003e-05, "loss": 0.613, "step": 1484 }, { "epoch": 1.0927152317880795, "grad_norm": 1.747739852450082, "learning_rate": 1.4701261978178506e-05, "loss": 0.5487, "step": 1485 }, { "epoch": 1.0934510669610007, "grad_norm": 1.664887337282694, "learning_rate": 1.4694247932574533e-05, "loss": 0.5694, "step": 1486 }, { "epoch": 1.094186902133922, "grad_norm": 1.7846635806285405, "learning_rate": 1.4687230923555961e-05, "loss": 0.5908, "step": 1487 }, { "epoch": 1.0949227373068433, "grad_norm": 1.567487754180324, "learning_rate": 1.4680210955552535e-05, "loss": 0.569, "step": 1488 }, { "epoch": 1.0956585724797645, "grad_norm": 1.7656700228406268, "learning_rate": 1.4673188032995858e-05, "loss": 0.5697, "step": 1489 }, { "epoch": 1.0963944076526857, "grad_norm": 1.5544584635537868, "learning_rate": 1.4666162160319406e-05, "loss": 0.5336, "step": 1490 }, { "epoch": 1.0971302428256071, "grad_norm": 1.6987069978868863, "learning_rate": 1.465913334195852e-05, "loss": 0.5738, "step": 1491 }, { "epoch": 1.0978660779985283, "grad_norm": 1.6716984890366846, "learning_rate": 1.4652101582350394e-05, "loss": 0.5789, "step": 1492 }, { "epoch": 1.0986019131714495, "grad_norm": 1.7205014103070821, "learning_rate": 1.464506688593408e-05, "loss": 0.5885, "step": 1493 }, { "epoch": 1.099337748344371, "grad_norm": 1.7165007065426567, "learning_rate": 1.4638029257150494e-05, "loss": 0.5733, "step": 1494 }, { "epoch": 1.1000735835172921, "grad_norm": 1.6705420284021413, "learning_rate": 1.4630988700442386e-05, "loss": 0.5619, "step": 1495 }, { "epoch": 1.1008094186902133, "grad_norm": 1.7354543076693514, "learning_rate": 1.4623945220254367e-05, "loss": 0.551, "step": 1496 }, { "epoch": 1.1015452538631347, "grad_norm": 1.7145396477135693, "learning_rate": 1.4616898821032892e-05, "loss": 0.5646, "step": 1497 }, { "epoch": 1.102281089036056, "grad_norm": 1.6687535338767427, "learning_rate": 1.4609849507226254e-05, "loss": 0.5331, "step": 1498 }, { "epoch": 1.1030169242089771, "grad_norm": 1.659482380731826, "learning_rate": 1.4602797283284586e-05, "loss": 0.547, "step": 1499 }, { "epoch": 1.1037527593818985, "grad_norm": 1.660927973615478, "learning_rate": 1.4595742153659867e-05, "loss": 0.5891, "step": 1500 }, { "epoch": 1.1044885945548197, "grad_norm": 1.6485195346627006, "learning_rate": 1.4588684122805895e-05, "loss": 0.584, "step": 1501 }, { "epoch": 1.105224429727741, "grad_norm": 1.5703893795634107, "learning_rate": 1.4581623195178318e-05, "loss": 0.5562, "step": 1502 }, { "epoch": 1.1059602649006623, "grad_norm": 1.659864497873078, "learning_rate": 1.45745593752346e-05, "loss": 0.5815, "step": 1503 }, { "epoch": 1.1066961000735835, "grad_norm": 1.5655492599997627, "learning_rate": 1.4567492667434031e-05, "loss": 0.5535, "step": 1504 }, { "epoch": 1.1074319352465047, "grad_norm": 1.723217454063725, "learning_rate": 1.4560423076237729e-05, "loss": 0.5686, "step": 1505 }, { "epoch": 1.108167770419426, "grad_norm": 1.7209848980652787, "learning_rate": 1.4553350606108632e-05, "loss": 0.5822, "step": 1506 }, { "epoch": 1.1089036055923474, "grad_norm": 1.711695633468652, "learning_rate": 1.4546275261511493e-05, "loss": 0.5425, "step": 1507 }, { "epoch": 1.1096394407652685, "grad_norm": 1.7883056394808705, "learning_rate": 1.4539197046912876e-05, "loss": 0.5904, "step": 1508 }, { "epoch": 1.1103752759381897, "grad_norm": 1.624480378480914, "learning_rate": 1.4532115966781168e-05, "loss": 0.5574, "step": 1509 }, { "epoch": 1.1111111111111112, "grad_norm": 1.7501186485980251, "learning_rate": 1.4525032025586555e-05, "loss": 0.5866, "step": 1510 }, { "epoch": 1.1118469462840324, "grad_norm": 1.6360489990309033, "learning_rate": 1.4517945227801029e-05, "loss": 0.5778, "step": 1511 }, { "epoch": 1.1125827814569536, "grad_norm": 1.6620875222370601, "learning_rate": 1.4510855577898397e-05, "loss": 0.577, "step": 1512 }, { "epoch": 1.113318616629875, "grad_norm": 1.6804572665052948, "learning_rate": 1.450376308035425e-05, "loss": 0.5469, "step": 1513 }, { "epoch": 1.1140544518027962, "grad_norm": 1.6431752080297655, "learning_rate": 1.4496667739645986e-05, "loss": 0.5691, "step": 1514 }, { "epoch": 1.1147902869757174, "grad_norm": 1.6589063655289644, "learning_rate": 1.4489569560252804e-05, "loss": 0.538, "step": 1515 }, { "epoch": 1.1155261221486388, "grad_norm": 1.6324948830534771, "learning_rate": 1.4482468546655679e-05, "loss": 0.5746, "step": 1516 }, { "epoch": 1.11626195732156, "grad_norm": 1.6280883114166111, "learning_rate": 1.447536470333739e-05, "loss": 0.581, "step": 1517 }, { "epoch": 1.1169977924944812, "grad_norm": 1.6045441998375134, "learning_rate": 1.446825803478249e-05, "loss": 0.5891, "step": 1518 }, { "epoch": 1.1177336276674026, "grad_norm": 1.797232287376697, "learning_rate": 1.4461148545477328e-05, "loss": 0.618, "step": 1519 }, { "epoch": 1.1184694628403238, "grad_norm": 1.6041387812362757, "learning_rate": 1.4454036239910022e-05, "loss": 0.6177, "step": 1520 }, { "epoch": 1.119205298013245, "grad_norm": 1.5855854495509063, "learning_rate": 1.4446921122570478e-05, "loss": 0.5447, "step": 1521 }, { "epoch": 1.1199411331861664, "grad_norm": 1.6382972876805815, "learning_rate": 1.443980319795037e-05, "loss": 0.5647, "step": 1522 }, { "epoch": 1.1206769683590876, "grad_norm": 1.6701457389613594, "learning_rate": 1.4432682470543149e-05, "loss": 0.619, "step": 1523 }, { "epoch": 1.1214128035320088, "grad_norm": 1.6022644977608829, "learning_rate": 1.4425558944844027e-05, "loss": 0.5436, "step": 1524 }, { "epoch": 1.12214863870493, "grad_norm": 1.658821270048333, "learning_rate": 1.4418432625349997e-05, "loss": 0.5934, "step": 1525 }, { "epoch": 1.1228844738778514, "grad_norm": 1.6587585106725538, "learning_rate": 1.44113035165598e-05, "loss": 0.5783, "step": 1526 }, { "epoch": 1.1236203090507726, "grad_norm": 1.7279704267146356, "learning_rate": 1.440417162297395e-05, "loss": 0.5674, "step": 1527 }, { "epoch": 1.124356144223694, "grad_norm": 2.0869149854829785, "learning_rate": 1.439703694909471e-05, "loss": 0.6184, "step": 1528 }, { "epoch": 1.1250919793966152, "grad_norm": 1.6243091856714826, "learning_rate": 1.4389899499426103e-05, "loss": 0.5518, "step": 1529 }, { "epoch": 1.1258278145695364, "grad_norm": 1.6313619588044435, "learning_rate": 1.4382759278473906e-05, "loss": 0.5597, "step": 1530 }, { "epoch": 1.1265636497424576, "grad_norm": 1.9554796507372152, "learning_rate": 1.437561629074564e-05, "loss": 0.591, "step": 1531 }, { "epoch": 1.127299484915379, "grad_norm": 1.6693713036397673, "learning_rate": 1.4368470540750578e-05, "loss": 0.6124, "step": 1532 }, { "epoch": 1.1280353200883002, "grad_norm": 1.600385117843736, "learning_rate": 1.4361322032999736e-05, "loss": 0.5457, "step": 1533 }, { "epoch": 1.1287711552612214, "grad_norm": 1.6395111744402493, "learning_rate": 1.4354170772005862e-05, "loss": 0.5778, "step": 1534 }, { "epoch": 1.1295069904341428, "grad_norm": 1.6717455913692627, "learning_rate": 1.4347016762283454e-05, "loss": 0.5188, "step": 1535 }, { "epoch": 1.130242825607064, "grad_norm": 1.6340488274131453, "learning_rate": 1.4339860008348743e-05, "loss": 0.5981, "step": 1536 }, { "epoch": 1.1309786607799852, "grad_norm": 1.7890508288582891, "learning_rate": 1.4332700514719687e-05, "loss": 0.6292, "step": 1537 }, { "epoch": 1.1317144959529066, "grad_norm": 1.696844946387311, "learning_rate": 1.4325538285915974e-05, "loss": 0.5795, "step": 1538 }, { "epoch": 1.1324503311258278, "grad_norm": 1.6277198280485496, "learning_rate": 1.4318373326459027e-05, "loss": 0.5057, "step": 1539 }, { "epoch": 1.133186166298749, "grad_norm": 1.748603148522373, "learning_rate": 1.4311205640871985e-05, "loss": 0.5919, "step": 1540 }, { "epoch": 1.1339220014716704, "grad_norm": 1.607139796524255, "learning_rate": 1.430403523367971e-05, "loss": 0.5366, "step": 1541 }, { "epoch": 1.1346578366445916, "grad_norm": 1.7285451292614846, "learning_rate": 1.429686210940878e-05, "loss": 0.5663, "step": 1542 }, { "epoch": 1.1353936718175128, "grad_norm": 1.5823478919138387, "learning_rate": 1.4289686272587493e-05, "loss": 0.5549, "step": 1543 }, { "epoch": 1.136129506990434, "grad_norm": 1.6428598929364049, "learning_rate": 1.4282507727745853e-05, "loss": 0.5588, "step": 1544 }, { "epoch": 1.1368653421633554, "grad_norm": 1.6794509652346086, "learning_rate": 1.4275326479415581e-05, "loss": 0.5761, "step": 1545 }, { "epoch": 1.1376011773362766, "grad_norm": 1.6022456162612353, "learning_rate": 1.4268142532130102e-05, "loss": 0.5924, "step": 1546 }, { "epoch": 1.138337012509198, "grad_norm": 1.6154237480724383, "learning_rate": 1.4260955890424539e-05, "loss": 0.5316, "step": 1547 }, { "epoch": 1.1390728476821192, "grad_norm": 1.6358623409529394, "learning_rate": 1.4253766558835724e-05, "loss": 0.552, "step": 1548 }, { "epoch": 1.1398086828550404, "grad_norm": 1.5950657235285242, "learning_rate": 1.4246574541902187e-05, "loss": 0.5882, "step": 1549 }, { "epoch": 1.1405445180279616, "grad_norm": 1.7487361750254857, "learning_rate": 1.4239379844164144e-05, "loss": 0.6143, "step": 1550 }, { "epoch": 1.141280353200883, "grad_norm": 1.5823086697060833, "learning_rate": 1.423218247016351e-05, "loss": 0.5676, "step": 1551 }, { "epoch": 1.1420161883738043, "grad_norm": 1.7406317405202547, "learning_rate": 1.4224982424443894e-05, "loss": 0.5992, "step": 1552 }, { "epoch": 1.1427520235467254, "grad_norm": 1.656077761498962, "learning_rate": 1.4217779711550578e-05, "loss": 0.5871, "step": 1553 }, { "epoch": 1.1434878587196469, "grad_norm": 1.7037863230733818, "learning_rate": 1.4210574336030545e-05, "loss": 0.585, "step": 1554 }, { "epoch": 1.144223693892568, "grad_norm": 1.5834560086322158, "learning_rate": 1.4203366302432447e-05, "loss": 0.59, "step": 1555 }, { "epoch": 1.1449595290654893, "grad_norm": 1.7334688901167137, "learning_rate": 1.4196155615306615e-05, "loss": 0.5959, "step": 1556 }, { "epoch": 1.1456953642384107, "grad_norm": 1.6758767856411003, "learning_rate": 1.418894227920506e-05, "loss": 0.561, "step": 1557 }, { "epoch": 1.1464311994113319, "grad_norm": 1.6558018100970515, "learning_rate": 1.4181726298681462e-05, "loss": 0.5379, "step": 1558 }, { "epoch": 1.147167034584253, "grad_norm": 1.7455715571513901, "learning_rate": 1.4174507678291169e-05, "loss": 0.5779, "step": 1559 }, { "epoch": 1.1479028697571745, "grad_norm": 1.6213067112442727, "learning_rate": 1.41672864225912e-05, "loss": 0.581, "step": 1560 }, { "epoch": 1.1486387049300957, "grad_norm": 1.702070096793928, "learning_rate": 1.4160062536140235e-05, "loss": 0.5918, "step": 1561 }, { "epoch": 1.1493745401030169, "grad_norm": 1.5979378712055694, "learning_rate": 1.4152836023498613e-05, "loss": 0.5461, "step": 1562 }, { "epoch": 1.150110375275938, "grad_norm": 1.8743854178384853, "learning_rate": 1.4145606889228331e-05, "loss": 0.5863, "step": 1563 }, { "epoch": 1.1508462104488595, "grad_norm": 1.5290462177667685, "learning_rate": 1.413837513789305e-05, "loss": 0.5615, "step": 1564 }, { "epoch": 1.1515820456217807, "grad_norm": 1.7500877269467336, "learning_rate": 1.4131140774058067e-05, "loss": 0.6364, "step": 1565 }, { "epoch": 1.152317880794702, "grad_norm": 1.6770083722631624, "learning_rate": 1.4123903802290346e-05, "loss": 0.55, "step": 1566 }, { "epoch": 1.1530537159676233, "grad_norm": 1.6373452873817556, "learning_rate": 1.4116664227158481e-05, "loss": 0.5556, "step": 1567 }, { "epoch": 1.1537895511405445, "grad_norm": 1.5908972113545479, "learning_rate": 1.410942205323272e-05, "loss": 0.5809, "step": 1568 }, { "epoch": 1.1545253863134657, "grad_norm": 1.5376796010167992, "learning_rate": 1.4102177285084946e-05, "loss": 0.5756, "step": 1569 }, { "epoch": 1.155261221486387, "grad_norm": 1.6470887459767158, "learning_rate": 1.4094929927288688e-05, "loss": 0.5758, "step": 1570 }, { "epoch": 1.1559970566593083, "grad_norm": 1.6305572864260847, "learning_rate": 1.4087679984419096e-05, "loss": 0.5862, "step": 1571 }, { "epoch": 1.1567328918322295, "grad_norm": 1.6906244852379722, "learning_rate": 1.4080427461052966e-05, "loss": 0.5666, "step": 1572 }, { "epoch": 1.157468727005151, "grad_norm": 1.6509406969819842, "learning_rate": 1.4073172361768715e-05, "loss": 0.5564, "step": 1573 }, { "epoch": 1.158204562178072, "grad_norm": 1.7222849247361465, "learning_rate": 1.4065914691146388e-05, "loss": 0.5714, "step": 1574 }, { "epoch": 1.1589403973509933, "grad_norm": 1.7007571047908197, "learning_rate": 1.4058654453767655e-05, "loss": 0.5442, "step": 1575 }, { "epoch": 1.1596762325239147, "grad_norm": 1.5869261717617482, "learning_rate": 1.4051391654215803e-05, "loss": 0.5739, "step": 1576 }, { "epoch": 1.160412067696836, "grad_norm": 1.5538401172935432, "learning_rate": 1.4044126297075735e-05, "loss": 0.578, "step": 1577 }, { "epoch": 1.161147902869757, "grad_norm": 1.7738895147645604, "learning_rate": 1.4036858386933977e-05, "loss": 0.5991, "step": 1578 }, { "epoch": 1.1618837380426785, "grad_norm": 1.7428981763616738, "learning_rate": 1.402958792837866e-05, "loss": 0.6042, "step": 1579 }, { "epoch": 1.1626195732155997, "grad_norm": 1.7940449087850439, "learning_rate": 1.4022314925999525e-05, "loss": 0.5668, "step": 1580 }, { "epoch": 1.163355408388521, "grad_norm": 1.647028410264198, "learning_rate": 1.4015039384387916e-05, "loss": 0.6094, "step": 1581 }, { "epoch": 1.1640912435614421, "grad_norm": 1.6413741009463292, "learning_rate": 1.4007761308136791e-05, "loss": 0.5989, "step": 1582 }, { "epoch": 1.1648270787343635, "grad_norm": 1.6742369033700673, "learning_rate": 1.4000480701840688e-05, "loss": 0.5679, "step": 1583 }, { "epoch": 1.1655629139072847, "grad_norm": 1.6484776675804262, "learning_rate": 1.3993197570095766e-05, "loss": 0.5584, "step": 1584 }, { "epoch": 1.1662987490802061, "grad_norm": 1.66416505782721, "learning_rate": 1.3985911917499764e-05, "loss": 0.6235, "step": 1585 }, { "epoch": 1.1670345842531273, "grad_norm": 1.7378279190943635, "learning_rate": 1.3978623748652007e-05, "loss": 0.5881, "step": 1586 }, { "epoch": 1.1677704194260485, "grad_norm": 1.7718540380219283, "learning_rate": 1.3971333068153423e-05, "loss": 0.6266, "step": 1587 }, { "epoch": 1.1685062545989697, "grad_norm": 1.6959831430840786, "learning_rate": 1.3964039880606522e-05, "loss": 0.5727, "step": 1588 }, { "epoch": 1.1692420897718911, "grad_norm": 1.636435304737457, "learning_rate": 1.3956744190615385e-05, "loss": 0.6004, "step": 1589 }, { "epoch": 1.1699779249448123, "grad_norm": 1.5763036262034176, "learning_rate": 1.394944600278569e-05, "loss": 0.5675, "step": 1590 }, { "epoch": 1.1707137601177335, "grad_norm": 1.7216596539386162, "learning_rate": 1.3942145321724678e-05, "loss": 0.5685, "step": 1591 }, { "epoch": 1.171449595290655, "grad_norm": 1.5411792748371422, "learning_rate": 1.3934842152041173e-05, "loss": 0.5854, "step": 1592 }, { "epoch": 1.1721854304635762, "grad_norm": 1.643367507121917, "learning_rate": 1.392753649834556e-05, "loss": 0.5786, "step": 1593 }, { "epoch": 1.1729212656364973, "grad_norm": 1.650868834375684, "learning_rate": 1.3920228365249807e-05, "loss": 0.5866, "step": 1594 }, { "epoch": 1.1736571008094188, "grad_norm": 1.701797932427629, "learning_rate": 1.3912917757367429e-05, "loss": 0.5976, "step": 1595 }, { "epoch": 1.17439293598234, "grad_norm": 1.7071478894154375, "learning_rate": 1.390560467931352e-05, "loss": 0.6005, "step": 1596 }, { "epoch": 1.1751287711552612, "grad_norm": 1.6572537589264553, "learning_rate": 1.3898289135704726e-05, "loss": 0.5854, "step": 1597 }, { "epoch": 1.1758646063281826, "grad_norm": 1.6313289515684164, "learning_rate": 1.3890971131159246e-05, "loss": 0.5893, "step": 1598 }, { "epoch": 1.1766004415011038, "grad_norm": 1.7528431664755124, "learning_rate": 1.3883650670296838e-05, "loss": 0.6031, "step": 1599 }, { "epoch": 1.177336276674025, "grad_norm": 1.6602624828494168, "learning_rate": 1.387632775773881e-05, "loss": 0.5311, "step": 1600 }, { "epoch": 1.1780721118469464, "grad_norm": 1.6457442150228296, "learning_rate": 1.3869002398108017e-05, "loss": 0.5896, "step": 1601 }, { "epoch": 1.1788079470198676, "grad_norm": 1.676888720761016, "learning_rate": 1.3861674596028855e-05, "loss": 0.5205, "step": 1602 }, { "epoch": 1.1795437821927888, "grad_norm": 1.687988341741529, "learning_rate": 1.3854344356127272e-05, "loss": 0.6005, "step": 1603 }, { "epoch": 1.1802796173657102, "grad_norm": 1.6642243577544955, "learning_rate": 1.384701168303074e-05, "loss": 0.5529, "step": 1604 }, { "epoch": 1.1810154525386314, "grad_norm": 1.6730757553990256, "learning_rate": 1.383967658136828e-05, "loss": 0.6072, "step": 1605 }, { "epoch": 1.1817512877115526, "grad_norm": 1.743584394740668, "learning_rate": 1.3832339055770443e-05, "loss": 0.5677, "step": 1606 }, { "epoch": 1.1824871228844738, "grad_norm": 1.6925666679275104, "learning_rate": 1.3824999110869306e-05, "loss": 0.5687, "step": 1607 }, { "epoch": 1.1832229580573952, "grad_norm": 1.7000070387544461, "learning_rate": 1.3817656751298474e-05, "loss": 0.5721, "step": 1608 }, { "epoch": 1.1839587932303164, "grad_norm": 1.5734835477683098, "learning_rate": 1.3810311981693084e-05, "loss": 0.5888, "step": 1609 }, { "epoch": 1.1846946284032376, "grad_norm": 1.6692704330619736, "learning_rate": 1.3802964806689784e-05, "loss": 0.5852, "step": 1610 }, { "epoch": 1.185430463576159, "grad_norm": 1.6813900761338572, "learning_rate": 1.3795615230926748e-05, "loss": 0.6222, "step": 1611 }, { "epoch": 1.1861662987490802, "grad_norm": 1.5289991180913018, "learning_rate": 1.378826325904366e-05, "loss": 0.537, "step": 1612 }, { "epoch": 1.1869021339220014, "grad_norm": 1.7300043569088797, "learning_rate": 1.3780908895681725e-05, "loss": 0.5944, "step": 1613 }, { "epoch": 1.1876379690949228, "grad_norm": 1.6455125918312654, "learning_rate": 1.3773552145483642e-05, "loss": 0.5643, "step": 1614 }, { "epoch": 1.188373804267844, "grad_norm": 1.6727308219501065, "learning_rate": 1.3766193013093637e-05, "loss": 0.5722, "step": 1615 }, { "epoch": 1.1891096394407652, "grad_norm": 1.743811490164519, "learning_rate": 1.3758831503157426e-05, "loss": 0.5702, "step": 1616 }, { "epoch": 1.1898454746136866, "grad_norm": 1.8167212245325182, "learning_rate": 1.3751467620322228e-05, "loss": 0.5784, "step": 1617 }, { "epoch": 1.1905813097866078, "grad_norm": 1.7791567433352524, "learning_rate": 1.3744101369236765e-05, "loss": 0.5774, "step": 1618 }, { "epoch": 1.191317144959529, "grad_norm": 1.7284023777763204, "learning_rate": 1.373673275455125e-05, "loss": 0.5596, "step": 1619 }, { "epoch": 1.1920529801324504, "grad_norm": 1.7882047717011604, "learning_rate": 1.3729361780917383e-05, "loss": 0.6169, "step": 1620 }, { "epoch": 1.1927888153053716, "grad_norm": 1.9563761629649519, "learning_rate": 1.3721988452988366e-05, "loss": 0.6443, "step": 1621 }, { "epoch": 1.1935246504782928, "grad_norm": 1.6353304210098594, "learning_rate": 1.3714612775418878e-05, "loss": 0.5539, "step": 1622 }, { "epoch": 1.1942604856512142, "grad_norm": 1.5825641170729703, "learning_rate": 1.3707234752865084e-05, "loss": 0.5609, "step": 1623 }, { "epoch": 1.1949963208241354, "grad_norm": 1.903390621485769, "learning_rate": 1.3699854389984626e-05, "loss": 0.6167, "step": 1624 }, { "epoch": 1.1957321559970566, "grad_norm": 1.6772826191646968, "learning_rate": 1.3692471691436632e-05, "loss": 0.5392, "step": 1625 }, { "epoch": 1.1964679911699778, "grad_norm": 1.6282351321690154, "learning_rate": 1.3685086661881695e-05, "loss": 0.5597, "step": 1626 }, { "epoch": 1.1972038263428992, "grad_norm": 1.659578913466149, "learning_rate": 1.367769930598188e-05, "loss": 0.5696, "step": 1627 }, { "epoch": 1.1979396615158204, "grad_norm": 1.801051652692543, "learning_rate": 1.3670309628400733e-05, "loss": 0.5963, "step": 1628 }, { "epoch": 1.1986754966887416, "grad_norm": 1.789580404770226, "learning_rate": 1.3662917633803246e-05, "loss": 0.5792, "step": 1629 }, { "epoch": 1.199411331861663, "grad_norm": 1.8401904115377323, "learning_rate": 1.3655523326855889e-05, "loss": 0.5751, "step": 1630 }, { "epoch": 1.2001471670345842, "grad_norm": 1.7015509173475396, "learning_rate": 1.364812671222659e-05, "loss": 0.5768, "step": 1631 }, { "epoch": 1.2008830022075054, "grad_norm": 1.7033015211894509, "learning_rate": 1.3640727794584725e-05, "loss": 0.5884, "step": 1632 }, { "epoch": 1.2008830022075054, "eval_en_loss": 1.0250611305236816, "eval_en_runtime": 16.5658, "eval_en_samples_per_second": 59.037, "eval_en_steps_per_second": 0.121, "step": 1632 }, { "epoch": 1.2008830022075054, "eval_eu_loss": 0.8758312463760376, "eval_eu_runtime": 14.9439, "eval_eu_samples_per_second": 65.913, "eval_eu_steps_per_second": 0.134, "step": 1632 }, { "epoch": 1.2016188373804269, "grad_norm": 1.6330532152335033, "learning_rate": 1.3633326578601133e-05, "loss": 0.6043, "step": 1633 }, { "epoch": 1.202354672553348, "grad_norm": 1.6688800854261605, "learning_rate": 1.3625923068948103e-05, "loss": 0.5345, "step": 1634 }, { "epoch": 1.2030905077262692, "grad_norm": 1.555622384458301, "learning_rate": 1.3618517270299367e-05, "loss": 0.5352, "step": 1635 }, { "epoch": 1.2038263428991907, "grad_norm": 1.6553138110213463, "learning_rate": 1.36111091873301e-05, "loss": 0.5464, "step": 1636 }, { "epoch": 1.2045621780721119, "grad_norm": 1.66654454774083, "learning_rate": 1.360369882471693e-05, "loss": 0.578, "step": 1637 }, { "epoch": 1.205298013245033, "grad_norm": 1.7180265984543581, "learning_rate": 1.3596286187137917e-05, "loss": 0.5546, "step": 1638 }, { "epoch": 1.2060338484179545, "grad_norm": 1.658234305227263, "learning_rate": 1.3588871279272553e-05, "loss": 0.6262, "step": 1639 }, { "epoch": 1.2067696835908757, "grad_norm": 1.7570927090376205, "learning_rate": 1.3581454105801775e-05, "loss": 0.5912, "step": 1640 }, { "epoch": 1.2075055187637969, "grad_norm": 1.6993635184965636, "learning_rate": 1.3574034671407937e-05, "loss": 0.5694, "step": 1641 }, { "epoch": 1.2082413539367183, "grad_norm": 1.704409588815701, "learning_rate": 1.356661298077483e-05, "loss": 0.565, "step": 1642 }, { "epoch": 1.2089771891096395, "grad_norm": 1.6889435734827263, "learning_rate": 1.3559189038587665e-05, "loss": 0.5431, "step": 1643 }, { "epoch": 1.2097130242825607, "grad_norm": 1.727791735382705, "learning_rate": 1.3551762849533077e-05, "loss": 0.5544, "step": 1644 }, { "epoch": 1.2104488594554819, "grad_norm": 1.6009063824542131, "learning_rate": 1.3544334418299115e-05, "loss": 0.5849, "step": 1645 }, { "epoch": 1.2111846946284033, "grad_norm": 1.7219984150226733, "learning_rate": 1.3536903749575248e-05, "loss": 0.5644, "step": 1646 }, { "epoch": 1.2119205298013245, "grad_norm": 1.5540583192397361, "learning_rate": 1.3529470848052353e-05, "loss": 0.5685, "step": 1647 }, { "epoch": 1.212656364974246, "grad_norm": 1.683219545313334, "learning_rate": 1.3522035718422722e-05, "loss": 0.5948, "step": 1648 }, { "epoch": 1.213392200147167, "grad_norm": 1.5898381007315152, "learning_rate": 1.3514598365380051e-05, "loss": 0.5558, "step": 1649 }, { "epoch": 1.2141280353200883, "grad_norm": 1.7742680939357531, "learning_rate": 1.3507158793619438e-05, "loss": 0.5587, "step": 1650 }, { "epoch": 1.2148638704930095, "grad_norm": 1.692234369212239, "learning_rate": 1.3499717007837381e-05, "loss": 0.5609, "step": 1651 }, { "epoch": 1.215599705665931, "grad_norm": 1.6322853948214673, "learning_rate": 1.3492273012731785e-05, "loss": 0.5915, "step": 1652 }, { "epoch": 1.216335540838852, "grad_norm": 1.712272407528033, "learning_rate": 1.3484826813001934e-05, "loss": 0.5812, "step": 1653 }, { "epoch": 1.2170713760117733, "grad_norm": 1.780518298565973, "learning_rate": 1.3477378413348516e-05, "loss": 0.5726, "step": 1654 }, { "epoch": 1.2178072111846947, "grad_norm": 1.8165982051431835, "learning_rate": 1.3469927818473607e-05, "loss": 0.5843, "step": 1655 }, { "epoch": 1.218543046357616, "grad_norm": 1.5850178299828417, "learning_rate": 1.346247503308066e-05, "loss": 0.5617, "step": 1656 }, { "epoch": 1.219278881530537, "grad_norm": 1.6008771815494183, "learning_rate": 1.3455020061874517e-05, "loss": 0.5842, "step": 1657 }, { "epoch": 1.2200147167034585, "grad_norm": 1.7083576738394919, "learning_rate": 1.3447562909561405e-05, "loss": 0.5539, "step": 1658 }, { "epoch": 1.2207505518763797, "grad_norm": 1.6794509054230666, "learning_rate": 1.3440103580848919e-05, "loss": 0.6087, "step": 1659 }, { "epoch": 1.221486387049301, "grad_norm": 1.6882314970079615, "learning_rate": 1.343264208044603e-05, "loss": 0.5868, "step": 1660 }, { "epoch": 1.2222222222222223, "grad_norm": 1.7456570105060547, "learning_rate": 1.3425178413063082e-05, "loss": 0.6058, "step": 1661 }, { "epoch": 1.2229580573951435, "grad_norm": 1.6379752977829967, "learning_rate": 1.3417712583411787e-05, "loss": 0.5692, "step": 1662 }, { "epoch": 1.2236938925680647, "grad_norm": 1.6983491309305794, "learning_rate": 1.3410244596205222e-05, "loss": 0.5778, "step": 1663 }, { "epoch": 1.224429727740986, "grad_norm": 1.5870457571437877, "learning_rate": 1.3402774456157825e-05, "loss": 0.5487, "step": 1664 }, { "epoch": 1.2251655629139073, "grad_norm": 1.6109719183125744, "learning_rate": 1.3395302167985393e-05, "loss": 0.5354, "step": 1665 }, { "epoch": 1.2259013980868285, "grad_norm": 1.7714999003241316, "learning_rate": 1.3387827736405079e-05, "loss": 0.6407, "step": 1666 }, { "epoch": 1.22663723325975, "grad_norm": 1.6420074705649013, "learning_rate": 1.338035116613539e-05, "loss": 0.5966, "step": 1667 }, { "epoch": 1.2273730684326711, "grad_norm": 1.6714357290954576, "learning_rate": 1.337287246189618e-05, "loss": 0.5897, "step": 1668 }, { "epoch": 1.2281089036055923, "grad_norm": 1.5826469758645965, "learning_rate": 1.336539162840866e-05, "loss": 0.5867, "step": 1669 }, { "epoch": 1.2288447387785135, "grad_norm": 1.7305895351493574, "learning_rate": 1.3357908670395371e-05, "loss": 0.5944, "step": 1670 }, { "epoch": 1.229580573951435, "grad_norm": 1.669594352638318, "learning_rate": 1.3350423592580204e-05, "loss": 0.5921, "step": 1671 }, { "epoch": 1.2303164091243561, "grad_norm": 1.8265369238990792, "learning_rate": 1.3342936399688387e-05, "loss": 0.5642, "step": 1672 }, { "epoch": 1.2310522442972773, "grad_norm": 1.7033280315292434, "learning_rate": 1.3335447096446481e-05, "loss": 0.5865, "step": 1673 }, { "epoch": 1.2317880794701987, "grad_norm": 1.792590892805456, "learning_rate": 1.3327955687582389e-05, "loss": 0.6019, "step": 1674 }, { "epoch": 1.23252391464312, "grad_norm": 1.7191931591905705, "learning_rate": 1.3320462177825321e-05, "loss": 0.5699, "step": 1675 }, { "epoch": 1.2332597498160411, "grad_norm": 1.6711981572435757, "learning_rate": 1.3312966571905836e-05, "loss": 0.5624, "step": 1676 }, { "epoch": 1.2339955849889626, "grad_norm": 1.7349058593984188, "learning_rate": 1.3305468874555811e-05, "loss": 0.56, "step": 1677 }, { "epoch": 1.2347314201618838, "grad_norm": 1.8244916291982227, "learning_rate": 1.3297969090508434e-05, "loss": 0.5443, "step": 1678 }, { "epoch": 1.235467255334805, "grad_norm": 1.8519277420883375, "learning_rate": 1.3290467224498214e-05, "loss": 0.584, "step": 1679 }, { "epoch": 1.2362030905077264, "grad_norm": 1.6188463570874905, "learning_rate": 1.3282963281260985e-05, "loss": 0.5631, "step": 1680 }, { "epoch": 1.2369389256806476, "grad_norm": 1.7158127556501868, "learning_rate": 1.3275457265533876e-05, "loss": 0.5924, "step": 1681 }, { "epoch": 1.2376747608535688, "grad_norm": 1.630579026047676, "learning_rate": 1.3267949182055335e-05, "loss": 0.599, "step": 1682 }, { "epoch": 1.23841059602649, "grad_norm": 1.7145666893951161, "learning_rate": 1.3260439035565114e-05, "loss": 0.5428, "step": 1683 }, { "epoch": 1.2391464311994114, "grad_norm": 1.6180328278625875, "learning_rate": 1.3252926830804264e-05, "loss": 0.5489, "step": 1684 }, { "epoch": 1.2398822663723326, "grad_norm": 1.5908935407880533, "learning_rate": 1.3245412572515137e-05, "loss": 0.5739, "step": 1685 }, { "epoch": 1.240618101545254, "grad_norm": 1.6536757390111427, "learning_rate": 1.3237896265441386e-05, "loss": 0.5389, "step": 1686 }, { "epoch": 1.2413539367181752, "grad_norm": 1.7206738142425848, "learning_rate": 1.323037791432795e-05, "loss": 0.5628, "step": 1687 }, { "epoch": 1.2420897718910964, "grad_norm": 1.6795923897985823, "learning_rate": 1.3222857523921058e-05, "loss": 0.6199, "step": 1688 }, { "epoch": 1.2428256070640176, "grad_norm": 1.580804696386432, "learning_rate": 1.3215335098968234e-05, "loss": 0.6147, "step": 1689 }, { "epoch": 1.243561442236939, "grad_norm": 1.7751074936288331, "learning_rate": 1.3207810644218277e-05, "loss": 0.5743, "step": 1690 }, { "epoch": 1.2442972774098602, "grad_norm": 1.6274351062919536, "learning_rate": 1.320028416442128e-05, "loss": 0.5413, "step": 1691 }, { "epoch": 1.2450331125827814, "grad_norm": 1.6266512157849864, "learning_rate": 1.3192755664328605e-05, "loss": 0.5519, "step": 1692 }, { "epoch": 1.2457689477557028, "grad_norm": 1.4993791411725446, "learning_rate": 1.3185225148692884e-05, "loss": 0.5831, "step": 1693 }, { "epoch": 1.246504782928624, "grad_norm": 1.6180187772483912, "learning_rate": 1.3177692622268039e-05, "loss": 0.5811, "step": 1694 }, { "epoch": 1.2472406181015452, "grad_norm": 1.5789400570271521, "learning_rate": 1.3170158089809246e-05, "loss": 0.5796, "step": 1695 }, { "epoch": 1.2479764532744666, "grad_norm": 1.571531853980459, "learning_rate": 1.316262155607295e-05, "loss": 0.5806, "step": 1696 }, { "epoch": 1.2487122884473878, "grad_norm": 1.503546640150044, "learning_rate": 1.3155083025816864e-05, "loss": 0.5312, "step": 1697 }, { "epoch": 1.249448123620309, "grad_norm": 1.62367085635759, "learning_rate": 1.3147542503799963e-05, "loss": 0.5578, "step": 1698 }, { "epoch": 1.2501839587932304, "grad_norm": 1.5964710051743565, "learning_rate": 1.3139999994782468e-05, "loss": 0.5605, "step": 1699 }, { "epoch": 1.2509197939661516, "grad_norm": 1.6387861545065932, "learning_rate": 1.313245550352587e-05, "loss": 0.5062, "step": 1700 }, { "epoch": 1.2516556291390728, "grad_norm": 1.5871717874340798, "learning_rate": 1.3124909034792896e-05, "loss": 0.5644, "step": 1701 }, { "epoch": 1.252391464311994, "grad_norm": 1.6747988679671229, "learning_rate": 1.3117360593347535e-05, "loss": 0.5885, "step": 1702 }, { "epoch": 1.2531272994849154, "grad_norm": 1.692072227990389, "learning_rate": 1.3109810183955012e-05, "loss": 0.6006, "step": 1703 }, { "epoch": 1.2538631346578366, "grad_norm": 1.6549603823183472, "learning_rate": 1.3102257811381795e-05, "loss": 0.5698, "step": 1704 }, { "epoch": 1.254598969830758, "grad_norm": 1.5938180496151089, "learning_rate": 1.3094703480395597e-05, "loss": 0.5334, "step": 1705 }, { "epoch": 1.2553348050036792, "grad_norm": 1.6392226213805188, "learning_rate": 1.3087147195765361e-05, "loss": 0.5543, "step": 1706 }, { "epoch": 1.2560706401766004, "grad_norm": 1.5826881485257616, "learning_rate": 1.3079588962261271e-05, "loss": 0.5587, "step": 1707 }, { "epoch": 1.2568064753495216, "grad_norm": 1.5573496046642847, "learning_rate": 1.3072028784654732e-05, "loss": 0.5985, "step": 1708 }, { "epoch": 1.257542310522443, "grad_norm": 1.6741571689112613, "learning_rate": 1.306446666771838e-05, "loss": 0.5689, "step": 1709 }, { "epoch": 1.2582781456953642, "grad_norm": 1.7298734419408972, "learning_rate": 1.3056902616226084e-05, "loss": 0.6022, "step": 1710 }, { "epoch": 1.2590139808682856, "grad_norm": 1.5872106350412893, "learning_rate": 1.3049336634952918e-05, "loss": 0.5692, "step": 1711 }, { "epoch": 1.2597498160412068, "grad_norm": 1.7741657049289108, "learning_rate": 1.3041768728675189e-05, "loss": 0.5845, "step": 1712 }, { "epoch": 1.260485651214128, "grad_norm": 1.806168575584835, "learning_rate": 1.3034198902170407e-05, "loss": 0.5801, "step": 1713 }, { "epoch": 1.2612214863870492, "grad_norm": 1.6960269222749766, "learning_rate": 1.3026627160217302e-05, "loss": 0.6014, "step": 1714 }, { "epoch": 1.2619573215599706, "grad_norm": 1.705863126349117, "learning_rate": 1.3019053507595814e-05, "loss": 0.5402, "step": 1715 }, { "epoch": 1.2626931567328918, "grad_norm": 1.741561663782613, "learning_rate": 1.3011477949087083e-05, "loss": 0.589, "step": 1716 }, { "epoch": 1.263428991905813, "grad_norm": 1.6310345137539242, "learning_rate": 1.3003900489473455e-05, "loss": 0.5973, "step": 1717 }, { "epoch": 1.2641648270787345, "grad_norm": 1.7719825016076658, "learning_rate": 1.2996321133538482e-05, "loss": 0.5819, "step": 1718 }, { "epoch": 1.2649006622516556, "grad_norm": 1.6559772911720225, "learning_rate": 1.2988739886066904e-05, "loss": 0.5981, "step": 1719 }, { "epoch": 1.2656364974245768, "grad_norm": 1.6733621323558092, "learning_rate": 1.2981156751844659e-05, "loss": 0.5553, "step": 1720 }, { "epoch": 1.266372332597498, "grad_norm": 1.6783813836881418, "learning_rate": 1.2973571735658878e-05, "loss": 0.5599, "step": 1721 }, { "epoch": 1.2671081677704195, "grad_norm": 1.7105796895326966, "learning_rate": 1.2965984842297876e-05, "loss": 0.6156, "step": 1722 }, { "epoch": 1.2678440029433407, "grad_norm": 1.654341775866032, "learning_rate": 1.2958396076551157e-05, "loss": 0.5593, "step": 1723 }, { "epoch": 1.268579838116262, "grad_norm": 1.618723177392311, "learning_rate": 1.29508054432094e-05, "loss": 0.5629, "step": 1724 }, { "epoch": 1.2693156732891833, "grad_norm": 1.7096428540058304, "learning_rate": 1.2943212947064477e-05, "loss": 0.6188, "step": 1725 }, { "epoch": 1.2700515084621045, "grad_norm": 1.5913704730907832, "learning_rate": 1.2935618592909419e-05, "loss": 0.6111, "step": 1726 }, { "epoch": 1.2707873436350257, "grad_norm": 1.6538891970075458, "learning_rate": 1.2928022385538441e-05, "loss": 0.6131, "step": 1727 }, { "epoch": 1.271523178807947, "grad_norm": 1.696791709142469, "learning_rate": 1.2920424329746928e-05, "loss": 0.5861, "step": 1728 }, { "epoch": 1.2722590139808683, "grad_norm": 1.7089623145281458, "learning_rate": 1.2912824430331425e-05, "loss": 0.6208, "step": 1729 }, { "epoch": 1.2729948491537897, "grad_norm": 1.5615918110924438, "learning_rate": 1.2905222692089647e-05, "loss": 0.5508, "step": 1730 }, { "epoch": 1.2737306843267109, "grad_norm": 1.7026099241413428, "learning_rate": 1.2897619119820466e-05, "loss": 0.5743, "step": 1731 }, { "epoch": 1.274466519499632, "grad_norm": 1.5780914174208969, "learning_rate": 1.2890013718323913e-05, "loss": 0.5608, "step": 1732 }, { "epoch": 1.2752023546725533, "grad_norm": 1.7869012200404912, "learning_rate": 1.2882406492401174e-05, "loss": 0.5736, "step": 1733 }, { "epoch": 1.2759381898454747, "grad_norm": 1.6728127374085477, "learning_rate": 1.287479744685459e-05, "loss": 0.5803, "step": 1734 }, { "epoch": 1.2766740250183959, "grad_norm": 1.6618979126091928, "learning_rate": 1.2867186586487642e-05, "loss": 0.5842, "step": 1735 }, { "epoch": 1.277409860191317, "grad_norm": 1.6807765940379928, "learning_rate": 1.2859573916104966e-05, "loss": 0.5998, "step": 1736 }, { "epoch": 1.2781456953642385, "grad_norm": 1.7103796889262586, "learning_rate": 1.2851959440512338e-05, "loss": 0.5557, "step": 1737 }, { "epoch": 1.2788815305371597, "grad_norm": 1.7778433115668713, "learning_rate": 1.2844343164516671e-05, "loss": 0.5963, "step": 1738 }, { "epoch": 1.279617365710081, "grad_norm": 1.8413011404586914, "learning_rate": 1.2836725092926013e-05, "loss": 0.572, "step": 1739 }, { "epoch": 1.280353200883002, "grad_norm": 1.749167299505893, "learning_rate": 1.2829105230549552e-05, "loss": 0.5748, "step": 1740 }, { "epoch": 1.2810890360559235, "grad_norm": 1.5639732253718703, "learning_rate": 1.2821483582197604e-05, "loss": 0.5543, "step": 1741 }, { "epoch": 1.2818248712288447, "grad_norm": 1.621413538693543, "learning_rate": 1.2813860152681607e-05, "loss": 0.526, "step": 1742 }, { "epoch": 1.2825607064017661, "grad_norm": 1.6437055948381856, "learning_rate": 1.280623494681413e-05, "loss": 0.5671, "step": 1743 }, { "epoch": 1.2832965415746873, "grad_norm": 1.584492946243981, "learning_rate": 1.2798607969408865e-05, "loss": 0.5605, "step": 1744 }, { "epoch": 1.2840323767476085, "grad_norm": 1.6621133912143302, "learning_rate": 1.2790979225280614e-05, "loss": 0.5738, "step": 1745 }, { "epoch": 1.2847682119205297, "grad_norm": 1.6762339400548576, "learning_rate": 1.2783348719245301e-05, "loss": 0.5966, "step": 1746 }, { "epoch": 1.2855040470934511, "grad_norm": 1.644351086871114, "learning_rate": 1.2775716456119962e-05, "loss": 0.5645, "step": 1747 }, { "epoch": 1.2862398822663723, "grad_norm": 1.6898388926720624, "learning_rate": 1.2768082440722735e-05, "loss": 0.6134, "step": 1748 }, { "epoch": 1.2869757174392937, "grad_norm": 1.5811074270924006, "learning_rate": 1.2760446677872873e-05, "loss": 0.6092, "step": 1749 }, { "epoch": 1.287711552612215, "grad_norm": 1.6648369832318701, "learning_rate": 1.275280917239073e-05, "loss": 0.6139, "step": 1750 }, { "epoch": 1.2884473877851361, "grad_norm": 1.7023994948275034, "learning_rate": 1.2745169929097756e-05, "loss": 0.6075, "step": 1751 }, { "epoch": 1.2891832229580573, "grad_norm": 1.6511192139327608, "learning_rate": 1.2737528952816503e-05, "loss": 0.6089, "step": 1752 }, { "epoch": 1.2899190581309787, "grad_norm": 1.66395176504933, "learning_rate": 1.2729886248370616e-05, "loss": 0.579, "step": 1753 }, { "epoch": 1.2906548933039, "grad_norm": 1.747241921423826, "learning_rate": 1.2722241820584827e-05, "loss": 0.5697, "step": 1754 }, { "epoch": 1.2913907284768211, "grad_norm": 1.6579344729624301, "learning_rate": 1.271459567428496e-05, "loss": 0.5913, "step": 1755 }, { "epoch": 1.2921265636497425, "grad_norm": 1.7182601514835805, "learning_rate": 1.2706947814297923e-05, "loss": 0.6293, "step": 1756 }, { "epoch": 1.2928623988226637, "grad_norm": 1.6139943249219948, "learning_rate": 1.2699298245451705e-05, "loss": 0.6049, "step": 1757 }, { "epoch": 1.293598233995585, "grad_norm": 1.7999094246079648, "learning_rate": 1.2691646972575376e-05, "loss": 0.6002, "step": 1758 }, { "epoch": 1.2943340691685061, "grad_norm": 1.6265066932174141, "learning_rate": 1.2683994000499078e-05, "loss": 0.571, "step": 1759 }, { "epoch": 1.2950699043414275, "grad_norm": 1.6273064163832915, "learning_rate": 1.2676339334054029e-05, "loss": 0.5706, "step": 1760 }, { "epoch": 1.2958057395143487, "grad_norm": 1.5378678985252683, "learning_rate": 1.2668682978072512e-05, "loss": 0.5423, "step": 1761 }, { "epoch": 1.2965415746872702, "grad_norm": 1.6075135404557672, "learning_rate": 1.2661024937387888e-05, "loss": 0.5546, "step": 1762 }, { "epoch": 1.2972774098601914, "grad_norm": 1.656212746934162, "learning_rate": 1.2653365216834567e-05, "loss": 0.5506, "step": 1763 }, { "epoch": 1.2980132450331126, "grad_norm": 1.641102707618753, "learning_rate": 1.2645703821248026e-05, "loss": 0.5655, "step": 1764 }, { "epoch": 1.2987490802060337, "grad_norm": 1.6590691151604982, "learning_rate": 1.2638040755464802e-05, "loss": 0.576, "step": 1765 }, { "epoch": 1.2994849153789552, "grad_norm": 1.6175751978619162, "learning_rate": 1.2630376024322481e-05, "loss": 0.5633, "step": 1766 }, { "epoch": 1.3002207505518764, "grad_norm": 1.6023602303562001, "learning_rate": 1.2622709632659706e-05, "loss": 0.5829, "step": 1767 }, { "epoch": 1.3009565857247978, "grad_norm": 1.6889810358630841, "learning_rate": 1.2615041585316163e-05, "loss": 0.5918, "step": 1768 }, { "epoch": 1.301692420897719, "grad_norm": 1.7752225183254047, "learning_rate": 1.2607371887132586e-05, "loss": 0.584, "step": 1769 }, { "epoch": 1.3024282560706402, "grad_norm": 1.69350433917708, "learning_rate": 1.2599700542950752e-05, "loss": 0.5915, "step": 1770 }, { "epoch": 1.3031640912435614, "grad_norm": 1.5657938537795593, "learning_rate": 1.2592027557613476e-05, "loss": 0.5524, "step": 1771 }, { "epoch": 1.3038999264164828, "grad_norm": 1.5751618566596928, "learning_rate": 1.2584352935964608e-05, "loss": 0.5174, "step": 1772 }, { "epoch": 1.304635761589404, "grad_norm": 1.769509494506034, "learning_rate": 1.257667668284903e-05, "loss": 0.589, "step": 1773 }, { "epoch": 1.3053715967623252, "grad_norm": 1.6624542128567834, "learning_rate": 1.2568998803112658e-05, "loss": 0.5819, "step": 1774 }, { "epoch": 1.3061074319352466, "grad_norm": 1.7978224046179736, "learning_rate": 1.256131930160243e-05, "loss": 0.6106, "step": 1775 }, { "epoch": 1.3068432671081678, "grad_norm": 1.7127366948887959, "learning_rate": 1.2553638183166315e-05, "loss": 0.5585, "step": 1776 }, { "epoch": 1.307579102281089, "grad_norm": 1.6956620969315237, "learning_rate": 1.2545955452653294e-05, "loss": 0.5821, "step": 1777 }, { "epoch": 1.3083149374540102, "grad_norm": 1.716989989513977, "learning_rate": 1.2538271114913372e-05, "loss": 0.5982, "step": 1778 }, { "epoch": 1.3090507726269316, "grad_norm": 1.604760090072921, "learning_rate": 1.2530585174797563e-05, "loss": 0.5808, "step": 1779 }, { "epoch": 1.3097866077998528, "grad_norm": 1.5841500263615986, "learning_rate": 1.2522897637157905e-05, "loss": 0.5535, "step": 1780 }, { "epoch": 1.3105224429727742, "grad_norm": 1.5779068275282342, "learning_rate": 1.2515208506847424e-05, "loss": 0.5996, "step": 1781 }, { "epoch": 1.3112582781456954, "grad_norm": 1.6520704217755668, "learning_rate": 1.2507517788720172e-05, "loss": 0.5661, "step": 1782 }, { "epoch": 1.3119941133186166, "grad_norm": 1.5924771037287082, "learning_rate": 1.249982548763119e-05, "loss": 0.5597, "step": 1783 }, { "epoch": 1.3127299484915378, "grad_norm": 1.6890827045258607, "learning_rate": 1.2492131608436519e-05, "loss": 0.5452, "step": 1784 }, { "epoch": 1.3134657836644592, "grad_norm": 1.7216563277265429, "learning_rate": 1.2484436155993207e-05, "loss": 0.5996, "step": 1785 }, { "epoch": 1.3142016188373804, "grad_norm": 1.7247451327854124, "learning_rate": 1.2476739135159286e-05, "loss": 0.5512, "step": 1786 }, { "epoch": 1.3149374540103018, "grad_norm": 1.7355103973971557, "learning_rate": 1.2469040550793777e-05, "loss": 0.5431, "step": 1787 }, { "epoch": 1.315673289183223, "grad_norm": 1.6422607316406208, "learning_rate": 1.2461340407756692e-05, "loss": 0.5875, "step": 1788 }, { "epoch": 1.3164091243561442, "grad_norm": 1.867081901480372, "learning_rate": 1.2453638710909033e-05, "loss": 0.6107, "step": 1789 }, { "epoch": 1.3171449595290654, "grad_norm": 1.6506631898805941, "learning_rate": 1.2445935465112769e-05, "loss": 0.546, "step": 1790 }, { "epoch": 1.3178807947019868, "grad_norm": 1.6833951303019015, "learning_rate": 1.2438230675230851e-05, "loss": 0.6185, "step": 1791 }, { "epoch": 1.318616629874908, "grad_norm": 1.6313647930162583, "learning_rate": 1.2430524346127215e-05, "loss": 0.5495, "step": 1792 }, { "epoch": 1.3193524650478292, "grad_norm": 1.7728323693188524, "learning_rate": 1.2422816482666757e-05, "loss": 0.5837, "step": 1793 }, { "epoch": 1.3200883002207506, "grad_norm": 1.6349436874241552, "learning_rate": 1.2415107089715348e-05, "loss": 0.5795, "step": 1794 }, { "epoch": 1.3208241353936718, "grad_norm": 1.4809571961424624, "learning_rate": 1.2407396172139822e-05, "loss": 0.5241, "step": 1795 }, { "epoch": 1.321559970566593, "grad_norm": 1.6501080078482473, "learning_rate": 1.2399683734807974e-05, "loss": 0.5875, "step": 1796 }, { "epoch": 1.3222958057395142, "grad_norm": 1.5728940182859035, "learning_rate": 1.2391969782588564e-05, "loss": 0.54, "step": 1797 }, { "epoch": 1.3230316409124356, "grad_norm": 1.7797381385702395, "learning_rate": 1.2384254320351301e-05, "loss": 0.5752, "step": 1798 }, { "epoch": 1.3237674760853568, "grad_norm": 2.6579092280597325, "learning_rate": 1.2376537352966856e-05, "loss": 0.6352, "step": 1799 }, { "epoch": 1.3245033112582782, "grad_norm": 1.515032906319599, "learning_rate": 1.2368818885306838e-05, "loss": 0.5384, "step": 1800 }, { "epoch": 1.3252391464311994, "grad_norm": 1.5367878372831336, "learning_rate": 1.2361098922243812e-05, "loss": 0.5334, "step": 1801 }, { "epoch": 1.3259749816041206, "grad_norm": 1.5906433332976568, "learning_rate": 1.2353377468651288e-05, "loss": 0.5597, "step": 1802 }, { "epoch": 1.3267108167770418, "grad_norm": 1.482919759434943, "learning_rate": 1.2345654529403714e-05, "loss": 0.5618, "step": 1803 }, { "epoch": 1.3274466519499633, "grad_norm": 1.656716948142424, "learning_rate": 1.233793010937648e-05, "loss": 0.5708, "step": 1804 }, { "epoch": 1.3281824871228844, "grad_norm": 1.6959818062290408, "learning_rate": 1.2330204213445896e-05, "loss": 0.5929, "step": 1805 }, { "epoch": 1.3289183222958059, "grad_norm": 1.5877157989914679, "learning_rate": 1.2322476846489222e-05, "loss": 0.5889, "step": 1806 }, { "epoch": 1.329654157468727, "grad_norm": 1.5763021034257718, "learning_rate": 1.2314748013384639e-05, "loss": 0.5735, "step": 1807 }, { "epoch": 1.3303899926416483, "grad_norm": 1.6707131990363375, "learning_rate": 1.2307017719011253e-05, "loss": 0.5792, "step": 1808 }, { "epoch": 1.3311258278145695, "grad_norm": 1.7451251236620564, "learning_rate": 1.2299285968249091e-05, "loss": 0.6163, "step": 1809 }, { "epoch": 1.3318616629874909, "grad_norm": 1.586491550989801, "learning_rate": 1.2291552765979104e-05, "loss": 0.5416, "step": 1810 }, { "epoch": 1.332597498160412, "grad_norm": 1.6718360301710564, "learning_rate": 1.2283818117083157e-05, "loss": 0.5741, "step": 1811 }, { "epoch": 1.3333333333333333, "grad_norm": 1.6760472520563194, "learning_rate": 1.2276082026444026e-05, "loss": 0.5809, "step": 1812 }, { "epoch": 1.3340691685062547, "grad_norm": 1.606633303059445, "learning_rate": 1.2268344498945404e-05, "loss": 0.5884, "step": 1813 }, { "epoch": 1.3348050036791759, "grad_norm": 1.6313880395281861, "learning_rate": 1.2260605539471884e-05, "loss": 0.5423, "step": 1814 }, { "epoch": 1.335540838852097, "grad_norm": 1.7960691482278495, "learning_rate": 1.2252865152908964e-05, "loss": 0.5952, "step": 1815 }, { "epoch": 1.3362766740250183, "grad_norm": 1.737692707279411, "learning_rate": 1.2245123344143044e-05, "loss": 0.5854, "step": 1816 }, { "epoch": 1.3370125091979397, "grad_norm": 1.6629289899430886, "learning_rate": 1.2237380118061423e-05, "loss": 0.5656, "step": 1817 }, { "epoch": 1.3377483443708609, "grad_norm": 1.6892505452061881, "learning_rate": 1.2229635479552295e-05, "loss": 0.6144, "step": 1818 }, { "epoch": 1.3384841795437823, "grad_norm": 1.6828565454235789, "learning_rate": 1.2221889433504743e-05, "loss": 0.5744, "step": 1819 }, { "epoch": 1.3392200147167035, "grad_norm": 1.939763812838378, "learning_rate": 1.2214141984808742e-05, "loss": 0.5246, "step": 1820 }, { "epoch": 1.3399558498896247, "grad_norm": 1.6079817042069515, "learning_rate": 1.220639313835515e-05, "loss": 0.5788, "step": 1821 }, { "epoch": 1.3406916850625459, "grad_norm": 1.559707428737149, "learning_rate": 1.2198642899035704e-05, "loss": 0.5948, "step": 1822 }, { "epoch": 1.3414275202354673, "grad_norm": 1.5826483310634691, "learning_rate": 1.2190891271743032e-05, "loss": 0.5656, "step": 1823 }, { "epoch": 1.3421633554083885, "grad_norm": 1.7266761764611274, "learning_rate": 1.2183138261370624e-05, "loss": 0.5781, "step": 1824 }, { "epoch": 1.34289919058131, "grad_norm": 1.6361928497445453, "learning_rate": 1.2175383872812851e-05, "loss": 0.5598, "step": 1825 }, { "epoch": 1.343635025754231, "grad_norm": 1.6378684827569578, "learning_rate": 1.2167628110964953e-05, "loss": 0.5575, "step": 1826 }, { "epoch": 1.3443708609271523, "grad_norm": 1.6204273709980654, "learning_rate": 1.2159870980723037e-05, "loss": 0.5826, "step": 1827 }, { "epoch": 1.3451066961000735, "grad_norm": 1.6589119029977324, "learning_rate": 1.2152112486984071e-05, "loss": 0.562, "step": 1828 }, { "epoch": 1.345842531272995, "grad_norm": 1.6621645715323754, "learning_rate": 1.214435263464589e-05, "loss": 0.526, "step": 1829 }, { "epoch": 1.346578366445916, "grad_norm": 1.6007229825258555, "learning_rate": 1.2136591428607181e-05, "loss": 0.5652, "step": 1830 }, { "epoch": 1.3473142016188373, "grad_norm": 1.6027512548703706, "learning_rate": 1.2128828873767487e-05, "loss": 0.5163, "step": 1831 }, { "epoch": 1.3480500367917587, "grad_norm": 1.6368175208977236, "learning_rate": 1.2121064975027203e-05, "loss": 0.547, "step": 1832 }, { "epoch": 1.34878587196468, "grad_norm": 1.684754525571462, "learning_rate": 1.2113299737287571e-05, "loss": 0.5363, "step": 1833 }, { "epoch": 1.3495217071376011, "grad_norm": 1.7070924386790034, "learning_rate": 1.210553316545068e-05, "loss": 0.555, "step": 1834 }, { "epoch": 1.3502575423105223, "grad_norm": 1.680724563410467, "learning_rate": 1.2097765264419463e-05, "loss": 0.606, "step": 1835 }, { "epoch": 1.3509933774834437, "grad_norm": 1.679104129997455, "learning_rate": 1.208999603909769e-05, "loss": 0.5959, "step": 1836 }, { "epoch": 1.351729212656365, "grad_norm": 1.5879384952252842, "learning_rate": 1.2082225494389961e-05, "loss": 0.5804, "step": 1837 }, { "epoch": 1.3524650478292863, "grad_norm": 1.6032655989504605, "learning_rate": 1.2074453635201722e-05, "loss": 0.5697, "step": 1838 }, { "epoch": 1.3532008830022075, "grad_norm": 1.5859119587952908, "learning_rate": 1.2066680466439236e-05, "loss": 0.5276, "step": 1839 }, { "epoch": 1.3539367181751287, "grad_norm": 1.606257052882603, "learning_rate": 1.2058905993009604e-05, "loss": 0.5505, "step": 1840 }, { "epoch": 1.35467255334805, "grad_norm": 1.54776216431662, "learning_rate": 1.2051130219820741e-05, "loss": 0.5397, "step": 1841 }, { "epoch": 1.3554083885209713, "grad_norm": 1.6313624175591792, "learning_rate": 1.2043353151781384e-05, "loss": 0.582, "step": 1842 }, { "epoch": 1.3561442236938925, "grad_norm": 1.662326298563422, "learning_rate": 1.2035574793801095e-05, "loss": 0.5703, "step": 1843 }, { "epoch": 1.356880058866814, "grad_norm": 1.5759336058941382, "learning_rate": 1.2027795150790242e-05, "loss": 0.5452, "step": 1844 }, { "epoch": 1.3576158940397351, "grad_norm": 1.7171116938490618, "learning_rate": 1.2020014227660008e-05, "loss": 0.5644, "step": 1845 }, { "epoch": 1.3583517292126563, "grad_norm": 1.6311070512711563, "learning_rate": 1.2012232029322384e-05, "loss": 0.571, "step": 1846 }, { "epoch": 1.3590875643855775, "grad_norm": 1.6553463829756712, "learning_rate": 1.2004448560690167e-05, "loss": 0.5488, "step": 1847 }, { "epoch": 1.359823399558499, "grad_norm": 1.5239058578728644, "learning_rate": 1.199666382667695e-05, "loss": 0.5419, "step": 1848 }, { "epoch": 1.3605592347314202, "grad_norm": 1.5780045401166218, "learning_rate": 1.1988877832197135e-05, "loss": 0.5443, "step": 1849 }, { "epoch": 1.3612950699043413, "grad_norm": 1.625041806359, "learning_rate": 1.1981090582165911e-05, "loss": 0.5865, "step": 1850 }, { "epoch": 1.3620309050772628, "grad_norm": 1.7218196313070446, "learning_rate": 1.1973302081499259e-05, "loss": 0.6026, "step": 1851 }, { "epoch": 1.362766740250184, "grad_norm": 1.599094593723593, "learning_rate": 1.1965512335113958e-05, "loss": 0.5972, "step": 1852 }, { "epoch": 1.3635025754231052, "grad_norm": 1.7682761975488557, "learning_rate": 1.1957721347927568e-05, "loss": 0.5592, "step": 1853 }, { "epoch": 1.3642384105960264, "grad_norm": 1.6450817625881806, "learning_rate": 1.194992912485843e-05, "loss": 0.5788, "step": 1854 }, { "epoch": 1.3649742457689478, "grad_norm": 1.720870442123089, "learning_rate": 1.1942135670825672e-05, "loss": 0.5625, "step": 1855 }, { "epoch": 1.365710080941869, "grad_norm": 1.6311889919584903, "learning_rate": 1.1934340990749193e-05, "loss": 0.5567, "step": 1856 }, { "epoch": 1.3664459161147904, "grad_norm": 1.7272209945424617, "learning_rate": 1.1926545089549666e-05, "loss": 0.5755, "step": 1857 }, { "epoch": 1.3671817512877116, "grad_norm": 1.6888796854231565, "learning_rate": 1.1918747972148541e-05, "loss": 0.6166, "step": 1858 }, { "epoch": 1.3679175864606328, "grad_norm": 1.608232387552836, "learning_rate": 1.191094964346803e-05, "loss": 0.5479, "step": 1859 }, { "epoch": 1.368653421633554, "grad_norm": 1.629847295524695, "learning_rate": 1.190315010843111e-05, "loss": 0.6069, "step": 1860 }, { "epoch": 1.3693892568064754, "grad_norm": 1.6994660769006389, "learning_rate": 1.189534937196152e-05, "loss": 0.5889, "step": 1861 }, { "epoch": 1.3701250919793966, "grad_norm": 1.741199292426383, "learning_rate": 1.1887547438983764e-05, "loss": 0.5683, "step": 1862 }, { "epoch": 1.370860927152318, "grad_norm": 1.5866605492583288, "learning_rate": 1.187974431442309e-05, "loss": 0.5802, "step": 1863 }, { "epoch": 1.3715967623252392, "grad_norm": 1.5851386019419855, "learning_rate": 1.1871940003205505e-05, "loss": 0.5647, "step": 1864 }, { "epoch": 1.3723325974981604, "grad_norm": 1.6535506927239185, "learning_rate": 1.1864134510257767e-05, "loss": 0.5591, "step": 1865 }, { "epoch": 1.3730684326710816, "grad_norm": 1.713060367066935, "learning_rate": 1.1856327840507377e-05, "loss": 0.5953, "step": 1866 }, { "epoch": 1.373804267844003, "grad_norm": 1.6524981608639844, "learning_rate": 1.1848519998882572e-05, "loss": 0.5575, "step": 1867 }, { "epoch": 1.3745401030169242, "grad_norm": 1.6642472449594576, "learning_rate": 1.1840710990312347e-05, "loss": 0.5518, "step": 1868 }, { "epoch": 1.3752759381898454, "grad_norm": 1.6783776706818672, "learning_rate": 1.1832900819726412e-05, "loss": 0.5328, "step": 1869 }, { "epoch": 1.3760117733627668, "grad_norm": 1.6402130555115022, "learning_rate": 1.1825089492055227e-05, "loss": 0.5987, "step": 1870 }, { "epoch": 1.376747608535688, "grad_norm": 1.5682795294780574, "learning_rate": 1.1817277012229973e-05, "loss": 0.5647, "step": 1871 }, { "epoch": 1.3774834437086092, "grad_norm": 1.6648720300638573, "learning_rate": 1.1809463385182568e-05, "loss": 0.5927, "step": 1872 }, { "epoch": 1.3782192788815306, "grad_norm": 1.607275969593695, "learning_rate": 1.1801648615845638e-05, "loss": 0.5577, "step": 1873 }, { "epoch": 1.3789551140544518, "grad_norm": 1.707664405966891, "learning_rate": 1.1793832709152548e-05, "loss": 0.582, "step": 1874 }, { "epoch": 1.379690949227373, "grad_norm": 1.6560591375667133, "learning_rate": 1.1786015670037374e-05, "loss": 0.6133, "step": 1875 }, { "epoch": 1.3804267844002944, "grad_norm": 1.635787839015738, "learning_rate": 1.17781975034349e-05, "loss": 0.598, "step": 1876 }, { "epoch": 1.3811626195732156, "grad_norm": 1.749951995242025, "learning_rate": 1.1770378214280631e-05, "loss": 0.5551, "step": 1877 }, { "epoch": 1.3818984547461368, "grad_norm": 1.6537613371875242, "learning_rate": 1.176255780751078e-05, "loss": 0.5643, "step": 1878 }, { "epoch": 1.382634289919058, "grad_norm": 1.7074724786685735, "learning_rate": 1.1754736288062256e-05, "loss": 0.5483, "step": 1879 }, { "epoch": 1.3833701250919794, "grad_norm": 1.612985773101468, "learning_rate": 1.1746913660872681e-05, "loss": 0.5369, "step": 1880 }, { "epoch": 1.3841059602649006, "grad_norm": 1.7295379706423881, "learning_rate": 1.1739089930880377e-05, "loss": 0.5363, "step": 1881 }, { "epoch": 1.384841795437822, "grad_norm": 1.6962925713329144, "learning_rate": 1.1731265103024351e-05, "loss": 0.5366, "step": 1882 }, { "epoch": 1.3855776306107432, "grad_norm": 1.6240826468506437, "learning_rate": 1.1723439182244316e-05, "loss": 0.6121, "step": 1883 }, { "epoch": 1.3863134657836644, "grad_norm": 1.5642100674204396, "learning_rate": 1.1715612173480668e-05, "loss": 0.5542, "step": 1884 }, { "epoch": 1.3870493009565856, "grad_norm": 1.6749740433524571, "learning_rate": 1.1707784081674482e-05, "loss": 0.5542, "step": 1885 }, { "epoch": 1.387785136129507, "grad_norm": 1.7551898747073287, "learning_rate": 1.1699954911767537e-05, "loss": 0.5865, "step": 1886 }, { "epoch": 1.3885209713024282, "grad_norm": 1.6864634995344123, "learning_rate": 1.1692124668702276e-05, "loss": 0.5305, "step": 1887 }, { "epoch": 1.3892568064753497, "grad_norm": 1.649576777337129, "learning_rate": 1.1684293357421824e-05, "loss": 0.5449, "step": 1888 }, { "epoch": 1.3899926416482709, "grad_norm": 1.6172605238846391, "learning_rate": 1.1676460982869977e-05, "loss": 0.5984, "step": 1889 }, { "epoch": 1.390728476821192, "grad_norm": 1.7127923091970172, "learning_rate": 1.1668627549991216e-05, "loss": 0.6109, "step": 1890 }, { "epoch": 1.3914643119941132, "grad_norm": 1.6623582784696604, "learning_rate": 1.1660793063730674e-05, "loss": 0.5724, "step": 1891 }, { "epoch": 1.3922001471670347, "grad_norm": 1.6156929256506605, "learning_rate": 1.1652957529034155e-05, "loss": 0.5244, "step": 1892 }, { "epoch": 1.3929359823399559, "grad_norm": 1.6138343897545238, "learning_rate": 1.1645120950848127e-05, "loss": 0.5285, "step": 1893 }, { "epoch": 1.393671817512877, "grad_norm": 1.5830362578660204, "learning_rate": 1.1637283334119713e-05, "loss": 0.5432, "step": 1894 }, { "epoch": 1.3944076526857985, "grad_norm": 1.8084531409745215, "learning_rate": 1.1629444683796694e-05, "loss": 0.561, "step": 1895 }, { "epoch": 1.3951434878587197, "grad_norm": 1.7086592352223076, "learning_rate": 1.1621605004827503e-05, "loss": 0.5862, "step": 1896 }, { "epoch": 1.3958793230316409, "grad_norm": 1.6183546818666004, "learning_rate": 1.1613764302161222e-05, "loss": 0.6038, "step": 1897 }, { "epoch": 1.396615158204562, "grad_norm": 1.6387191689133846, "learning_rate": 1.1605922580747578e-05, "loss": 0.5377, "step": 1898 }, { "epoch": 1.3973509933774835, "grad_norm": 1.5998262732748012, "learning_rate": 1.1598079845536947e-05, "loss": 0.6078, "step": 1899 }, { "epoch": 1.3980868285504047, "grad_norm": 1.478937280918058, "learning_rate": 1.1590236101480339e-05, "loss": 0.5507, "step": 1900 }, { "epoch": 1.398822663723326, "grad_norm": 1.6875290278767006, "learning_rate": 1.1582391353529399e-05, "loss": 0.5891, "step": 1901 }, { "epoch": 1.3995584988962473, "grad_norm": 1.5930524312373293, "learning_rate": 1.1574545606636413e-05, "loss": 0.5457, "step": 1902 }, { "epoch": 1.4002943340691685, "grad_norm": 1.7804187067921255, "learning_rate": 1.1566698865754291e-05, "loss": 0.6034, "step": 1903 }, { "epoch": 1.4010301692420897, "grad_norm": 1.6427767506717992, "learning_rate": 1.1558851135836573e-05, "loss": 0.578, "step": 1904 }, { "epoch": 1.401766004415011, "grad_norm": 1.6277125696160584, "learning_rate": 1.155100242183743e-05, "loss": 0.5442, "step": 1905 }, { "epoch": 1.4025018395879323, "grad_norm": 1.680868469331632, "learning_rate": 1.154315272871164e-05, "loss": 0.5958, "step": 1906 }, { "epoch": 1.4032376747608537, "grad_norm": 1.5752142944230687, "learning_rate": 1.1535302061414613e-05, "loss": 0.5645, "step": 1907 }, { "epoch": 1.403973509933775, "grad_norm": 1.633720467870632, "learning_rate": 1.1527450424902364e-05, "loss": 0.5337, "step": 1908 }, { "epoch": 1.404709345106696, "grad_norm": 1.6486558472981163, "learning_rate": 1.1519597824131527e-05, "loss": 0.6105, "step": 1909 }, { "epoch": 1.4054451802796173, "grad_norm": 1.6084331303014758, "learning_rate": 1.1511744264059339e-05, "loss": 0.5749, "step": 1910 }, { "epoch": 1.4061810154525387, "grad_norm": 1.7063800582542832, "learning_rate": 1.1503889749643649e-05, "loss": 0.5514, "step": 1911 }, { "epoch": 1.40691685062546, "grad_norm": 1.6199664297379, "learning_rate": 1.1496034285842897e-05, "loss": 0.5962, "step": 1912 }, { "epoch": 1.407652685798381, "grad_norm": 1.5373712139493687, "learning_rate": 1.1488177877616136e-05, "loss": 0.5781, "step": 1913 }, { "epoch": 1.4083885209713025, "grad_norm": 1.7098788559534626, "learning_rate": 1.148032052992301e-05, "loss": 0.5421, "step": 1914 }, { "epoch": 1.4091243561442237, "grad_norm": 1.6143801248903742, "learning_rate": 1.1472462247723752e-05, "loss": 0.5333, "step": 1915 }, { "epoch": 1.409860191317145, "grad_norm": 1.6975960857075199, "learning_rate": 1.1464603035979192e-05, "loss": 0.5714, "step": 1916 }, { "epoch": 1.410596026490066, "grad_norm": 1.6109764048589759, "learning_rate": 1.1456742899650739e-05, "loss": 0.5532, "step": 1917 }, { "epoch": 1.4113318616629875, "grad_norm": 1.7042474870872868, "learning_rate": 1.1448881843700392e-05, "loss": 0.6076, "step": 1918 }, { "epoch": 1.4120676968359087, "grad_norm": 1.6720477916068792, "learning_rate": 1.1441019873090726e-05, "loss": 0.605, "step": 1919 }, { "epoch": 1.4128035320088301, "grad_norm": 1.6712023263758147, "learning_rate": 1.1433156992784899e-05, "loss": 0.54, "step": 1920 }, { "epoch": 1.4135393671817513, "grad_norm": 1.8391068511348356, "learning_rate": 1.1425293207746638e-05, "loss": 0.5572, "step": 1921 }, { "epoch": 1.4142752023546725, "grad_norm": 1.6703928914131883, "learning_rate": 1.1417428522940242e-05, "loss": 0.5825, "step": 1922 }, { "epoch": 1.4150110375275937, "grad_norm": 1.6163998851004484, "learning_rate": 1.140956294333058e-05, "loss": 0.5662, "step": 1923 }, { "epoch": 1.4157468727005151, "grad_norm": 1.6750188337501677, "learning_rate": 1.1401696473883086e-05, "loss": 0.5828, "step": 1924 }, { "epoch": 1.4164827078734363, "grad_norm": 1.763476082524852, "learning_rate": 1.1393829119563752e-05, "loss": 0.5992, "step": 1925 }, { "epoch": 1.4172185430463577, "grad_norm": 1.7207737362621254, "learning_rate": 1.1385960885339138e-05, "loss": 0.5378, "step": 1926 }, { "epoch": 1.417954378219279, "grad_norm": 1.6529237571967215, "learning_rate": 1.1378091776176348e-05, "loss": 0.5755, "step": 1927 }, { "epoch": 1.4186902133922001, "grad_norm": 1.5681337699980944, "learning_rate": 1.1370221797043039e-05, "loss": 0.5884, "step": 1928 }, { "epoch": 1.4194260485651213, "grad_norm": 1.6198920514304507, "learning_rate": 1.136235095290743e-05, "loss": 0.4905, "step": 1929 }, { "epoch": 1.4201618837380428, "grad_norm": 1.6600101419772877, "learning_rate": 1.1354479248738271e-05, "loss": 0.5553, "step": 1930 }, { "epoch": 1.420897718910964, "grad_norm": 1.6693008551124882, "learning_rate": 1.1346606689504862e-05, "loss": 0.5713, "step": 1931 }, { "epoch": 1.4216335540838851, "grad_norm": 1.5646647061251102, "learning_rate": 1.1338733280177046e-05, "loss": 0.5585, "step": 1932 }, { "epoch": 1.4223693892568066, "grad_norm": 1.731601682965176, "learning_rate": 1.1330859025725193e-05, "loss": 0.6056, "step": 1933 }, { "epoch": 1.4231052244297278, "grad_norm": 1.728981233246729, "learning_rate": 1.1322983931120217e-05, "loss": 0.5618, "step": 1934 }, { "epoch": 1.423841059602649, "grad_norm": 1.6070372309302887, "learning_rate": 1.1315108001333556e-05, "loss": 0.6002, "step": 1935 }, { "epoch": 1.4245768947755701, "grad_norm": 1.667125580563415, "learning_rate": 1.130723124133718e-05, "loss": 0.5765, "step": 1936 }, { "epoch": 1.4253127299484916, "grad_norm": 1.599844733245482, "learning_rate": 1.1299353656103573e-05, "loss": 0.5692, "step": 1937 }, { "epoch": 1.4260485651214128, "grad_norm": 1.614528022530162, "learning_rate": 1.129147525060575e-05, "loss": 0.5715, "step": 1938 }, { "epoch": 1.4267844002943342, "grad_norm": 1.581977112392787, "learning_rate": 1.1283596029817248e-05, "loss": 0.568, "step": 1939 }, { "epoch": 1.4275202354672554, "grad_norm": 1.5861004346969427, "learning_rate": 1.1275715998712098e-05, "loss": 0.5441, "step": 1940 }, { "epoch": 1.4282560706401766, "grad_norm": 1.6654359341870644, "learning_rate": 1.1267835162264867e-05, "loss": 0.5021, "step": 1941 }, { "epoch": 1.4289919058130978, "grad_norm": 1.690467463222397, "learning_rate": 1.1259953525450616e-05, "loss": 0.5918, "step": 1942 }, { "epoch": 1.4297277409860192, "grad_norm": 1.5967295822478393, "learning_rate": 1.1252071093244915e-05, "loss": 0.5462, "step": 1943 }, { "epoch": 1.4304635761589404, "grad_norm": 1.6212385380172984, "learning_rate": 1.1244187870623832e-05, "loss": 0.5687, "step": 1944 }, { "epoch": 1.4311994113318618, "grad_norm": 1.62678090762976, "learning_rate": 1.1236303862563945e-05, "loss": 0.576, "step": 1945 }, { "epoch": 1.431935246504783, "grad_norm": 1.6036764144180038, "learning_rate": 1.1228419074042312e-05, "loss": 0.5434, "step": 1946 }, { "epoch": 1.4326710816777042, "grad_norm": 1.6366891748147494, "learning_rate": 1.1220533510036497e-05, "loss": 0.5865, "step": 1947 }, { "epoch": 1.4334069168506254, "grad_norm": 1.7443459671694674, "learning_rate": 1.1212647175524551e-05, "loss": 0.5733, "step": 1948 }, { "epoch": 1.4341427520235468, "grad_norm": 1.6571645516022788, "learning_rate": 1.1204760075485007e-05, "loss": 0.5491, "step": 1949 }, { "epoch": 1.434878587196468, "grad_norm": 1.673632053689673, "learning_rate": 1.1196872214896881e-05, "loss": 0.5885, "step": 1950 }, { "epoch": 1.4356144223693892, "grad_norm": 1.6771799352154384, "learning_rate": 1.1188983598739675e-05, "loss": 0.6214, "step": 1951 }, { "epoch": 1.4363502575423106, "grad_norm": 1.6937333457984918, "learning_rate": 1.1181094231993365e-05, "loss": 0.5935, "step": 1952 }, { "epoch": 1.4370860927152318, "grad_norm": 1.6708321791000225, "learning_rate": 1.1173204119638399e-05, "loss": 0.5678, "step": 1953 }, { "epoch": 1.437821927888153, "grad_norm": 1.59797790843739, "learning_rate": 1.1165313266655698e-05, "loss": 0.5506, "step": 1954 }, { "epoch": 1.4385577630610742, "grad_norm": 1.7082930393220228, "learning_rate": 1.1157421678026645e-05, "loss": 0.5345, "step": 1955 }, { "epoch": 1.4392935982339956, "grad_norm": 1.7444808853044957, "learning_rate": 1.1149529358733098e-05, "loss": 0.5813, "step": 1956 }, { "epoch": 1.4400294334069168, "grad_norm": 1.6256066426254665, "learning_rate": 1.1141636313757369e-05, "loss": 0.5516, "step": 1957 }, { "epoch": 1.4407652685798382, "grad_norm": 1.5806772541891903, "learning_rate": 1.113374254808223e-05, "loss": 0.5504, "step": 1958 }, { "epoch": 1.4415011037527594, "grad_norm": 1.6285248541606476, "learning_rate": 1.1125848066690903e-05, "loss": 0.6092, "step": 1959 }, { "epoch": 1.4422369389256806, "grad_norm": 1.7339105980649374, "learning_rate": 1.1117952874567073e-05, "loss": 0.5557, "step": 1960 }, { "epoch": 1.4429727740986018, "grad_norm": 1.8418131684623091, "learning_rate": 1.1110056976694862e-05, "loss": 0.5667, "step": 1961 }, { "epoch": 1.4437086092715232, "grad_norm": 1.6627238460168583, "learning_rate": 1.1102160378058844e-05, "loss": 0.5326, "step": 1962 }, { "epoch": 1.4444444444444444, "grad_norm": 1.6510043239636674, "learning_rate": 1.1094263083644036e-05, "loss": 0.5692, "step": 1963 }, { "epoch": 1.4451802796173658, "grad_norm": 1.594389765879526, "learning_rate": 1.1086365098435888e-05, "loss": 0.5851, "step": 1964 }, { "epoch": 1.445916114790287, "grad_norm": 1.5674802726240622, "learning_rate": 1.1078466427420292e-05, "loss": 0.5472, "step": 1965 }, { "epoch": 1.4466519499632082, "grad_norm": 1.6339762561138564, "learning_rate": 1.1070567075583572e-05, "loss": 0.5414, "step": 1966 }, { "epoch": 1.4473877851361294, "grad_norm": 1.7053713467886087, "learning_rate": 1.106266704791248e-05, "loss": 0.586, "step": 1967 }, { "epoch": 1.4481236203090508, "grad_norm": 1.7066289519501057, "learning_rate": 1.1054766349394198e-05, "loss": 0.6009, "step": 1968 }, { "epoch": 1.448859455481972, "grad_norm": 1.6859953563191974, "learning_rate": 1.1046864985016326e-05, "loss": 0.5461, "step": 1969 }, { "epoch": 1.4495952906548932, "grad_norm": 1.6049439444812885, "learning_rate": 1.1038962959766883e-05, "loss": 0.4992, "step": 1970 }, { "epoch": 1.4503311258278146, "grad_norm": 1.6241353541161443, "learning_rate": 1.103106027863432e-05, "loss": 0.5344, "step": 1971 }, { "epoch": 1.4510669610007358, "grad_norm": 1.679905047176464, "learning_rate": 1.1023156946607485e-05, "loss": 0.5635, "step": 1972 }, { "epoch": 1.451802796173657, "grad_norm": 1.724069193282252, "learning_rate": 1.1015252968675645e-05, "loss": 0.5502, "step": 1973 }, { "epoch": 1.4525386313465782, "grad_norm": 1.7231251300689372, "learning_rate": 1.1007348349828473e-05, "loss": 0.6257, "step": 1974 }, { "epoch": 1.4532744665194997, "grad_norm": 1.7635351837888371, "learning_rate": 1.0999443095056051e-05, "loss": 0.5968, "step": 1975 }, { "epoch": 1.4540103016924208, "grad_norm": 1.6688543283489172, "learning_rate": 1.099153720934885e-05, "loss": 0.5428, "step": 1976 }, { "epoch": 1.4547461368653423, "grad_norm": 1.6343255543122703, "learning_rate": 1.0983630697697755e-05, "loss": 0.5729, "step": 1977 }, { "epoch": 1.4554819720382635, "grad_norm": 1.7110556710172071, "learning_rate": 1.0975723565094036e-05, "loss": 0.5444, "step": 1978 }, { "epoch": 1.4562178072111847, "grad_norm": 1.6941852315947088, "learning_rate": 1.0967815816529358e-05, "loss": 0.5882, "step": 1979 }, { "epoch": 1.4569536423841059, "grad_norm": 1.6818970254123977, "learning_rate": 1.095990745699577e-05, "loss": 0.5813, "step": 1980 }, { "epoch": 1.4576894775570273, "grad_norm": 1.6642388699269959, "learning_rate": 1.0951998491485722e-05, "loss": 0.558, "step": 1981 }, { "epoch": 1.4584253127299485, "grad_norm": 1.6890422402494658, "learning_rate": 1.0944088924992026e-05, "loss": 0.6008, "step": 1982 }, { "epoch": 1.4591611479028699, "grad_norm": 1.65161857108531, "learning_rate": 1.0936178762507887e-05, "loss": 0.5929, "step": 1983 }, { "epoch": 1.459896983075791, "grad_norm": 1.6739425087378619, "learning_rate": 1.0928268009026885e-05, "loss": 0.6036, "step": 1984 }, { "epoch": 1.4606328182487123, "grad_norm": 1.584961269388695, "learning_rate": 1.092035666954297e-05, "loss": 0.5206, "step": 1985 }, { "epoch": 1.4613686534216335, "grad_norm": 1.637499083673002, "learning_rate": 1.0912444749050457e-05, "loss": 0.5364, "step": 1986 }, { "epoch": 1.4621044885945549, "grad_norm": 1.627823209892429, "learning_rate": 1.090453225254404e-05, "loss": 0.5647, "step": 1987 }, { "epoch": 1.462840323767476, "grad_norm": 1.8464593175471395, "learning_rate": 1.0896619185018764e-05, "loss": 0.6042, "step": 1988 }, { "epoch": 1.4635761589403973, "grad_norm": 1.6603853596453817, "learning_rate": 1.0888705551470045e-05, "loss": 0.55, "step": 1989 }, { "epoch": 1.4643119941133187, "grad_norm": 1.6285994676265145, "learning_rate": 1.0880791356893652e-05, "loss": 0.55, "step": 1990 }, { "epoch": 1.4650478292862399, "grad_norm": 1.6779663565709444, "learning_rate": 1.0872876606285702e-05, "loss": 0.5535, "step": 1991 }, { "epoch": 1.465783664459161, "grad_norm": 1.6402788477421373, "learning_rate": 1.0864961304642676e-05, "loss": 0.5429, "step": 1992 }, { "epoch": 1.4665194996320823, "grad_norm": 1.66842324833599, "learning_rate": 1.0857045456961394e-05, "loss": 0.6095, "step": 1993 }, { "epoch": 1.4672553348050037, "grad_norm": 1.58388538674602, "learning_rate": 1.084912906823902e-05, "loss": 0.5675, "step": 1994 }, { "epoch": 1.467991169977925, "grad_norm": 1.7555552993643715, "learning_rate": 1.0841212143473065e-05, "loss": 0.5898, "step": 1995 }, { "epoch": 1.4687270051508463, "grad_norm": 1.7547462231900763, "learning_rate": 1.0833294687661376e-05, "loss": 0.5589, "step": 1996 }, { "epoch": 1.4694628403237675, "grad_norm": 1.6781087504906314, "learning_rate": 1.0825376705802129e-05, "loss": 0.5432, "step": 1997 }, { "epoch": 1.4701986754966887, "grad_norm": 1.6384843699694462, "learning_rate": 1.081745820289384e-05, "loss": 0.523, "step": 1998 }, { "epoch": 1.47093451066961, "grad_norm": 1.6928713650554745, "learning_rate": 1.0809539183935358e-05, "loss": 0.5785, "step": 1999 }, { "epoch": 1.4716703458425313, "grad_norm": 1.8569552735031456, "learning_rate": 1.0801619653925843e-05, "loss": 0.6188, "step": 2000 }, { "epoch": 1.4724061810154525, "grad_norm": 1.6691809008052516, "learning_rate": 1.079369961786479e-05, "loss": 0.6226, "step": 2001 }, { "epoch": 1.473142016188374, "grad_norm": 1.6421720516596607, "learning_rate": 1.0785779080752012e-05, "loss": 0.5579, "step": 2002 }, { "epoch": 1.4738778513612951, "grad_norm": 1.6840925382575136, "learning_rate": 1.0777858047587635e-05, "loss": 0.5874, "step": 2003 }, { "epoch": 1.4746136865342163, "grad_norm": 1.650972043952294, "learning_rate": 1.0769936523372094e-05, "loss": 0.5441, "step": 2004 }, { "epoch": 1.4753495217071375, "grad_norm": 1.6626786736293468, "learning_rate": 1.0762014513106143e-05, "loss": 0.5749, "step": 2005 }, { "epoch": 1.476085356880059, "grad_norm": 1.6610215666487698, "learning_rate": 1.0754092021790838e-05, "loss": 0.6119, "step": 2006 }, { "epoch": 1.4768211920529801, "grad_norm": 1.7220757421666644, "learning_rate": 1.0746169054427543e-05, "loss": 0.5659, "step": 2007 }, { "epoch": 1.4775570272259013, "grad_norm": 1.6714974053973717, "learning_rate": 1.073824561601791e-05, "loss": 0.595, "step": 2008 }, { "epoch": 1.4782928623988227, "grad_norm": 1.640814200806081, "learning_rate": 1.0730321711563909e-05, "loss": 0.5374, "step": 2009 }, { "epoch": 1.479028697571744, "grad_norm": 1.613606630384996, "learning_rate": 1.0722397346067785e-05, "loss": 0.5834, "step": 2010 }, { "epoch": 1.4797645327446651, "grad_norm": 1.7244982676027247, "learning_rate": 1.0714472524532085e-05, "loss": 0.5567, "step": 2011 }, { "epoch": 1.4805003679175863, "grad_norm": 1.9345436926248396, "learning_rate": 1.070654725195964e-05, "loss": 0.5517, "step": 2012 }, { "epoch": 1.4812362030905077, "grad_norm": 1.7256806040276527, "learning_rate": 1.0698621533353562e-05, "loss": 0.5848, "step": 2013 }, { "epoch": 1.481972038263429, "grad_norm": 1.627532414983298, "learning_rate": 1.0690695373717254e-05, "loss": 0.5241, "step": 2014 }, { "epoch": 1.4827078734363504, "grad_norm": 1.6827098348208358, "learning_rate": 1.0682768778054394e-05, "loss": 0.5653, "step": 2015 }, { "epoch": 1.4834437086092715, "grad_norm": 1.686153194241136, "learning_rate": 1.0674841751368925e-05, "loss": 0.5428, "step": 2016 }, { "epoch": 1.4841795437821927, "grad_norm": 1.5658719996749786, "learning_rate": 1.0666914298665079e-05, "loss": 0.5294, "step": 2017 }, { "epoch": 1.484915378955114, "grad_norm": 1.6433072417452517, "learning_rate": 1.0658986424947347e-05, "loss": 0.5717, "step": 2018 }, { "epoch": 1.4856512141280354, "grad_norm": 1.60049988751329, "learning_rate": 1.0651058135220484e-05, "loss": 0.544, "step": 2019 }, { "epoch": 1.4863870493009566, "grad_norm": 1.7383971559265776, "learning_rate": 1.0643129434489514e-05, "loss": 0.6787, "step": 2020 }, { "epoch": 1.487122884473878, "grad_norm": 1.6328412054965977, "learning_rate": 1.0635200327759719e-05, "loss": 0.5742, "step": 2021 }, { "epoch": 1.4878587196467992, "grad_norm": 1.5916157494415346, "learning_rate": 1.062727082003663e-05, "loss": 0.5687, "step": 2022 }, { "epoch": 1.4885945548197204, "grad_norm": 1.6026132691364001, "learning_rate": 1.0619340916326039e-05, "loss": 0.5549, "step": 2023 }, { "epoch": 1.4893303899926416, "grad_norm": 1.7956380807260615, "learning_rate": 1.0611410621633992e-05, "loss": 0.6146, "step": 2024 }, { "epoch": 1.490066225165563, "grad_norm": 1.7294072432119165, "learning_rate": 1.0603479940966769e-05, "loss": 0.552, "step": 2025 }, { "epoch": 1.4908020603384842, "grad_norm": 1.715478782830011, "learning_rate": 1.05955488793309e-05, "loss": 0.5377, "step": 2026 }, { "epoch": 1.4915378955114054, "grad_norm": 1.5777164882383876, "learning_rate": 1.058761744173316e-05, "loss": 0.585, "step": 2027 }, { "epoch": 1.4922737306843268, "grad_norm": 1.6460329612493232, "learning_rate": 1.0579685633180556e-05, "loss": 0.5385, "step": 2028 }, { "epoch": 1.493009565857248, "grad_norm": 1.7779789258350518, "learning_rate": 1.0571753458680329e-05, "loss": 0.5827, "step": 2029 }, { "epoch": 1.4937454010301692, "grad_norm": 1.6575207581218216, "learning_rate": 1.0563820923239956e-05, "loss": 0.5867, "step": 2030 }, { "epoch": 1.4944812362030906, "grad_norm": 1.6617978652193899, "learning_rate": 1.0555888031867132e-05, "loss": 0.6062, "step": 2031 }, { "epoch": 1.4952170713760118, "grad_norm": 1.7707039329787393, "learning_rate": 1.0547954789569785e-05, "loss": 0.5444, "step": 2032 }, { "epoch": 1.495952906548933, "grad_norm": 1.685860062072796, "learning_rate": 1.0540021201356067e-05, "loss": 0.5789, "step": 2033 }, { "epoch": 1.4966887417218544, "grad_norm": 1.6767388793568552, "learning_rate": 1.0532087272234341e-05, "loss": 0.548, "step": 2034 }, { "epoch": 1.4974245768947756, "grad_norm": 1.5866764072711037, "learning_rate": 1.0524153007213185e-05, "loss": 0.5723, "step": 2035 }, { "epoch": 1.4981604120676968, "grad_norm": 1.7674611806084748, "learning_rate": 1.05162184113014e-05, "loss": 0.6195, "step": 2036 }, { "epoch": 1.498896247240618, "grad_norm": 1.6366291083124846, "learning_rate": 1.050828348950798e-05, "loss": 0.5872, "step": 2037 }, { "epoch": 1.4996320824135394, "grad_norm": 1.716193573699539, "learning_rate": 1.0500348246842136e-05, "loss": 0.5532, "step": 2038 }, { "epoch": 1.5003679175864606, "grad_norm": 1.5465610762828967, "learning_rate": 1.0492412688313279e-05, "loss": 0.5515, "step": 2039 }, { "epoch": 1.501103752759382, "grad_norm": 1.6639093972176955, "learning_rate": 1.0484476818931015e-05, "loss": 0.5572, "step": 2040 }, { "epoch": 1.5018395879323032, "grad_norm": 1.5602221379341157, "learning_rate": 1.0476540643705153e-05, "loss": 0.5594, "step": 2041 }, { "epoch": 1.5025754231052244, "grad_norm": 1.799454763927708, "learning_rate": 1.0468604167645695e-05, "loss": 0.5696, "step": 2042 }, { "epoch": 1.5033112582781456, "grad_norm": 1.790071142028682, "learning_rate": 1.0460667395762825e-05, "loss": 0.5843, "step": 2043 }, { "epoch": 1.5040470934510668, "grad_norm": 1.61839206222413, "learning_rate": 1.0452730333066919e-05, "loss": 0.5771, "step": 2044 }, { "epoch": 1.5047829286239882, "grad_norm": 1.642017577726343, "learning_rate": 1.044479298456854e-05, "loss": 0.5569, "step": 2045 }, { "epoch": 1.5055187637969096, "grad_norm": 1.658192775106458, "learning_rate": 1.0436855355278426e-05, "loss": 0.5759, "step": 2046 }, { "epoch": 1.5062545989698308, "grad_norm": 1.6951597803312521, "learning_rate": 1.0428917450207489e-05, "loss": 0.5836, "step": 2047 }, { "epoch": 1.506990434142752, "grad_norm": 1.6518575500035901, "learning_rate": 1.042097927436683e-05, "loss": 0.5423, "step": 2048 }, { "epoch": 1.5077262693156732, "grad_norm": 1.6097028499714123, "learning_rate": 1.0413040832767702e-05, "loss": 0.5886, "step": 2049 }, { "epoch": 1.5084621044885944, "grad_norm": 1.6964167521822195, "learning_rate": 1.0405102130421536e-05, "loss": 0.5678, "step": 2050 }, { "epoch": 1.5091979396615158, "grad_norm": 1.7094707645885585, "learning_rate": 1.0397163172339932e-05, "loss": 0.5494, "step": 2051 }, { "epoch": 1.5099337748344372, "grad_norm": 1.75317922921274, "learning_rate": 1.038922396353464e-05, "loss": 0.5574, "step": 2052 }, { "epoch": 1.5106696100073584, "grad_norm": 1.9011057496624109, "learning_rate": 1.0381284509017578e-05, "loss": 0.5965, "step": 2053 }, { "epoch": 1.5114054451802796, "grad_norm": 1.6135075692657024, "learning_rate": 1.0373344813800814e-05, "loss": 0.5092, "step": 2054 }, { "epoch": 1.5121412803532008, "grad_norm": 1.6007244240944427, "learning_rate": 1.0365404882896568e-05, "loss": 0.5456, "step": 2055 }, { "epoch": 1.512877115526122, "grad_norm": 1.7717315411449996, "learning_rate": 1.035746472131721e-05, "loss": 0.6043, "step": 2056 }, { "epoch": 1.5136129506990434, "grad_norm": 1.6282654839297135, "learning_rate": 1.0349524334075256e-05, "loss": 0.5688, "step": 2057 }, { "epoch": 1.5143487858719646, "grad_norm": 1.6408549324394774, "learning_rate": 1.0341583726183365e-05, "loss": 0.5304, "step": 2058 }, { "epoch": 1.515084621044886, "grad_norm": 1.5638135216422413, "learning_rate": 1.033364290265433e-05, "loss": 0.5667, "step": 2059 }, { "epoch": 1.5158204562178073, "grad_norm": 1.519972430920829, "learning_rate": 1.0325701868501092e-05, "loss": 0.565, "step": 2060 }, { "epoch": 1.5165562913907285, "grad_norm": 1.6556634745940395, "learning_rate": 1.031776062873671e-05, "loss": 0.569, "step": 2061 }, { "epoch": 1.5172921265636496, "grad_norm": 1.682816278525133, "learning_rate": 1.0309819188374386e-05, "loss": 0.571, "step": 2062 }, { "epoch": 1.518027961736571, "grad_norm": 1.6745445862497832, "learning_rate": 1.030187755242744e-05, "loss": 0.5679, "step": 2063 }, { "epoch": 1.5187637969094923, "grad_norm": 1.5108136080007393, "learning_rate": 1.0293935725909318e-05, "loss": 0.5248, "step": 2064 }, { "epoch": 1.5194996320824137, "grad_norm": 1.7862014160610131, "learning_rate": 1.0285993713833586e-05, "loss": 0.5916, "step": 2065 }, { "epoch": 1.5202354672553349, "grad_norm": 1.6177638593372496, "learning_rate": 1.0278051521213931e-05, "loss": 0.5332, "step": 2066 }, { "epoch": 1.520971302428256, "grad_norm": 1.7027271067708998, "learning_rate": 1.0270109153064147e-05, "loss": 0.5695, "step": 2067 }, { "epoch": 1.5217071376011773, "grad_norm": 1.6239153232069632, "learning_rate": 1.0262166614398144e-05, "loss": 0.549, "step": 2068 }, { "epoch": 1.5224429727740985, "grad_norm": 1.6377133995099697, "learning_rate": 1.0254223910229939e-05, "loss": 0.5741, "step": 2069 }, { "epoch": 1.5231788079470199, "grad_norm": 1.734432170438945, "learning_rate": 1.0246281045573654e-05, "loss": 0.5842, "step": 2070 }, { "epoch": 1.5239146431199413, "grad_norm": 1.5710762274284589, "learning_rate": 1.0238338025443507e-05, "loss": 0.5452, "step": 2071 }, { "epoch": 1.5246504782928625, "grad_norm": 1.7181406033516133, "learning_rate": 1.023039485485382e-05, "loss": 0.5658, "step": 2072 }, { "epoch": 1.5253863134657837, "grad_norm": 1.6321429881434064, "learning_rate": 1.022245153881901e-05, "loss": 0.5698, "step": 2073 }, { "epoch": 1.5261221486387049, "grad_norm": 1.7795839797973398, "learning_rate": 1.0214508082353578e-05, "loss": 0.5955, "step": 2074 }, { "epoch": 1.526857983811626, "grad_norm": 1.555187709325409, "learning_rate": 1.0206564490472124e-05, "loss": 0.5475, "step": 2075 }, { "epoch": 1.5275938189845475, "grad_norm": 1.641074014317317, "learning_rate": 1.019862076818933e-05, "loss": 0.5679, "step": 2076 }, { "epoch": 1.5283296541574687, "grad_norm": 1.6024623386492252, "learning_rate": 1.019067692051996e-05, "loss": 0.6049, "step": 2077 }, { "epoch": 1.52906548933039, "grad_norm": 1.8011231457284704, "learning_rate": 1.0182732952478851e-05, "loss": 0.6024, "step": 2078 }, { "epoch": 1.5298013245033113, "grad_norm": 1.6271770379886934, "learning_rate": 1.0174788869080929e-05, "loss": 0.5788, "step": 2079 }, { "epoch": 1.5305371596762325, "grad_norm": 1.5836942002744068, "learning_rate": 1.016684467534118e-05, "loss": 0.5172, "step": 2080 }, { "epoch": 1.5312729948491537, "grad_norm": 1.5931852358226284, "learning_rate": 1.0158900376274665e-05, "loss": 0.533, "step": 2081 }, { "epoch": 1.532008830022075, "grad_norm": 1.5618797270517855, "learning_rate": 1.0150955976896515e-05, "loss": 0.5648, "step": 2082 }, { "epoch": 1.5327446651949963, "grad_norm": 1.8388198012447665, "learning_rate": 1.0143011482221916e-05, "loss": 0.6067, "step": 2083 }, { "epoch": 1.5334805003679177, "grad_norm": 1.5965411804501952, "learning_rate": 1.0135066897266124e-05, "loss": 0.5807, "step": 2084 }, { "epoch": 1.534216335540839, "grad_norm": 1.6968827192129694, "learning_rate": 1.0127122227044444e-05, "loss": 0.5981, "step": 2085 }, { "epoch": 1.5349521707137601, "grad_norm": 1.5736572666214013, "learning_rate": 1.0119177476572237e-05, "loss": 0.5509, "step": 2086 }, { "epoch": 1.5356880058866813, "grad_norm": 1.62823545858451, "learning_rate": 1.0111232650864918e-05, "loss": 0.5688, "step": 2087 }, { "epoch": 1.5364238410596025, "grad_norm": 1.598960755481244, "learning_rate": 1.0103287754937946e-05, "loss": 0.6237, "step": 2088 }, { "epoch": 1.537159676232524, "grad_norm": 1.5452866584513607, "learning_rate": 1.0095342793806828e-05, "loss": 0.5408, "step": 2089 }, { "epoch": 1.5378955114054453, "grad_norm": 1.6408919104546065, "learning_rate": 1.0087397772487102e-05, "loss": 0.5463, "step": 2090 }, { "epoch": 1.5386313465783665, "grad_norm": 1.706523763374258, "learning_rate": 1.0079452695994358e-05, "loss": 0.5553, "step": 2091 }, { "epoch": 1.5393671817512877, "grad_norm": 1.6445033177842225, "learning_rate": 1.0071507569344213e-05, "loss": 0.5698, "step": 2092 }, { "epoch": 1.540103016924209, "grad_norm": 1.6426622909601079, "learning_rate": 1.0063562397552314e-05, "loss": 0.5722, "step": 2093 }, { "epoch": 1.5408388520971301, "grad_norm": 1.6622740231367648, "learning_rate": 1.0055617185634344e-05, "loss": 0.5834, "step": 2094 }, { "epoch": 1.5415746872700515, "grad_norm": 1.6057561098527635, "learning_rate": 1.0047671938606002e-05, "loss": 0.5, "step": 2095 }, { "epoch": 1.5423105224429727, "grad_norm": 1.6895075453711903, "learning_rate": 1.0039726661483019e-05, "loss": 0.5808, "step": 2096 }, { "epoch": 1.5430463576158941, "grad_norm": 1.6630627250217922, "learning_rate": 1.0031781359281135e-05, "loss": 0.5556, "step": 2097 }, { "epoch": 1.5437821927888153, "grad_norm": 1.6090736300239246, "learning_rate": 1.0023836037016115e-05, "loss": 0.541, "step": 2098 }, { "epoch": 1.5445180279617365, "grad_norm": 1.5826076902154915, "learning_rate": 1.0015890699703727e-05, "loss": 0.5369, "step": 2099 }, { "epoch": 1.5452538631346577, "grad_norm": 1.688271952329565, "learning_rate": 1.0007945352359763e-05, "loss": 0.5762, "step": 2100 }, { "epoch": 1.5459896983075792, "grad_norm": 1.7944503978382123, "learning_rate": 1e-05, "loss": 0.557, "step": 2101 }, { "epoch": 1.5467255334805003, "grad_norm": 1.5867695371927264, "learning_rate": 9.992054647640242e-06, "loss": 0.5678, "step": 2102 }, { "epoch": 1.5474613686534218, "grad_norm": 1.624555936034828, "learning_rate": 9.984109300296271e-06, "loss": 0.5906, "step": 2103 }, { "epoch": 1.548197203826343, "grad_norm": 1.624040178368215, "learning_rate": 9.976163962983889e-06, "loss": 0.5899, "step": 2104 }, { "epoch": 1.5489330389992642, "grad_norm": 1.5437497421835324, "learning_rate": 9.968218640718867e-06, "loss": 0.598, "step": 2105 }, { "epoch": 1.5496688741721854, "grad_norm": 1.5668542810542732, "learning_rate": 9.960273338516983e-06, "loss": 0.6012, "step": 2106 }, { "epoch": 1.5504047093451065, "grad_norm": 1.6635954319733146, "learning_rate": 9.952328061394001e-06, "loss": 0.5545, "step": 2107 }, { "epoch": 1.551140544518028, "grad_norm": 1.5825990368421707, "learning_rate": 9.944382814365661e-06, "loss": 0.5907, "step": 2108 }, { "epoch": 1.5518763796909494, "grad_norm": 1.6436585945994533, "learning_rate": 9.936437602447687e-06, "loss": 0.5535, "step": 2109 }, { "epoch": 1.5526122148638706, "grad_norm": 1.6836989476521598, "learning_rate": 9.92849243065579e-06, "loss": 0.564, "step": 2110 }, { "epoch": 1.5533480500367918, "grad_norm": 1.591594799684637, "learning_rate": 9.920547304005645e-06, "loss": 0.5836, "step": 2111 }, { "epoch": 1.554083885209713, "grad_norm": 1.5472420964422224, "learning_rate": 9.912602227512898e-06, "loss": 0.5244, "step": 2112 }, { "epoch": 1.5548197203826342, "grad_norm": 1.6534774529655973, "learning_rate": 9.904657206193175e-06, "loss": 0.5374, "step": 2113 }, { "epoch": 1.5555555555555556, "grad_norm": 1.701675755100576, "learning_rate": 9.896712245062055e-06, "loss": 0.5697, "step": 2114 }, { "epoch": 1.5562913907284768, "grad_norm": 1.6945691209982439, "learning_rate": 9.888767349135084e-06, "loss": 0.5959, "step": 2115 }, { "epoch": 1.5570272259013982, "grad_norm": 1.763009683734181, "learning_rate": 9.880822523427766e-06, "loss": 0.5849, "step": 2116 }, { "epoch": 1.5577630610743194, "grad_norm": 1.7556564694764039, "learning_rate": 9.872877772955561e-06, "loss": 0.5773, "step": 2117 }, { "epoch": 1.5584988962472406, "grad_norm": 1.723101409681146, "learning_rate": 9.864933102733878e-06, "loss": 0.5962, "step": 2118 }, { "epoch": 1.5592347314201618, "grad_norm": 1.6577333961097547, "learning_rate": 9.856988517778086e-06, "loss": 0.5783, "step": 2119 }, { "epoch": 1.5599705665930832, "grad_norm": 1.7036541897640411, "learning_rate": 9.84904402310349e-06, "loss": 0.5971, "step": 2120 }, { "epoch": 1.5607064017660044, "grad_norm": 1.7102249834339234, "learning_rate": 9.841099623725335e-06, "loss": 0.5608, "step": 2121 }, { "epoch": 1.5614422369389258, "grad_norm": 1.628764327564998, "learning_rate": 9.833155324658823e-06, "loss": 0.5933, "step": 2122 }, { "epoch": 1.562178072111847, "grad_norm": 1.684117488601007, "learning_rate": 9.825211130919075e-06, "loss": 0.5502, "step": 2123 }, { "epoch": 1.5629139072847682, "grad_norm": 1.6082524958645772, "learning_rate": 9.81726704752115e-06, "loss": 0.5366, "step": 2124 }, { "epoch": 1.5636497424576894, "grad_norm": 1.6828837548716284, "learning_rate": 9.809323079480043e-06, "loss": 0.5671, "step": 2125 }, { "epoch": 1.5643855776306106, "grad_norm": 1.7650544016620995, "learning_rate": 9.801379231810673e-06, "loss": 0.5568, "step": 2126 }, { "epoch": 1.565121412803532, "grad_norm": 1.6010684658090801, "learning_rate": 9.793435509527876e-06, "loss": 0.5636, "step": 2127 }, { "epoch": 1.5658572479764534, "grad_norm": 1.6962730472696914, "learning_rate": 9.785491917646425e-06, "loss": 0.6098, "step": 2128 }, { "epoch": 1.5665930831493746, "grad_norm": 1.6303890964386876, "learning_rate": 9.777548461180997e-06, "loss": 0.5683, "step": 2129 }, { "epoch": 1.5673289183222958, "grad_norm": 1.6177549465904468, "learning_rate": 9.769605145146184e-06, "loss": 0.5472, "step": 2130 }, { "epoch": 1.568064753495217, "grad_norm": 1.7938646460200085, "learning_rate": 9.761661974556495e-06, "loss": 0.5802, "step": 2131 }, { "epoch": 1.5688005886681382, "grad_norm": 1.649422920259423, "learning_rate": 9.75371895442635e-06, "loss": 0.5795, "step": 2132 }, { "epoch": 1.5695364238410596, "grad_norm": 1.6031704869927177, "learning_rate": 9.745776089770061e-06, "loss": 0.5588, "step": 2133 }, { "epoch": 1.5702722590139808, "grad_norm": 1.6655299038269809, "learning_rate": 9.737833385601858e-06, "loss": 0.5751, "step": 2134 }, { "epoch": 1.5710080941869022, "grad_norm": 1.677933816223336, "learning_rate": 9.729890846935856e-06, "loss": 0.5541, "step": 2135 }, { "epoch": 1.5717439293598234, "grad_norm": 1.6039066238098938, "learning_rate": 9.72194847878607e-06, "loss": 0.5766, "step": 2136 }, { "epoch": 1.5724797645327446, "grad_norm": 1.6361193506549443, "learning_rate": 9.714006286166416e-06, "loss": 0.5881, "step": 2137 }, { "epoch": 1.5732155997056658, "grad_norm": 1.6155971628128232, "learning_rate": 9.706064274090687e-06, "loss": 0.5908, "step": 2138 }, { "epoch": 1.5739514348785872, "grad_norm": 1.5858105741764166, "learning_rate": 9.698122447572563e-06, "loss": 0.4995, "step": 2139 }, { "epoch": 1.5746872700515084, "grad_norm": 1.6721308792694551, "learning_rate": 9.690180811625618e-06, "loss": 0.531, "step": 2140 }, { "epoch": 1.5754231052244299, "grad_norm": 1.6811258989389235, "learning_rate": 9.682239371263293e-06, "loss": 0.5745, "step": 2141 }, { "epoch": 1.576158940397351, "grad_norm": 1.8065420828867766, "learning_rate": 9.67429813149891e-06, "loss": 0.5922, "step": 2142 }, { "epoch": 1.5768947755702722, "grad_norm": 1.642511793989236, "learning_rate": 9.666357097345672e-06, "loss": 0.5481, "step": 2143 }, { "epoch": 1.5776306107431934, "grad_norm": 1.7041838395045736, "learning_rate": 9.65841627381664e-06, "loss": 0.565, "step": 2144 }, { "epoch": 1.5783664459161146, "grad_norm": 1.6529994220419666, "learning_rate": 9.650475665924747e-06, "loss": 0.6069, "step": 2145 }, { "epoch": 1.579102281089036, "grad_norm": 1.6512487159930365, "learning_rate": 9.642535278682795e-06, "loss": 0.5794, "step": 2146 }, { "epoch": 1.5798381162619575, "grad_norm": 1.5409595427783993, "learning_rate": 9.634595117103437e-06, "loss": 0.518, "step": 2147 }, { "epoch": 1.5805739514348787, "grad_norm": 1.7050307836573118, "learning_rate": 9.62665518619919e-06, "loss": 0.5342, "step": 2148 }, { "epoch": 1.5813097866077999, "grad_norm": 1.5566318019721967, "learning_rate": 9.618715490982425e-06, "loss": 0.5353, "step": 2149 }, { "epoch": 1.582045621780721, "grad_norm": 1.704589848431604, "learning_rate": 9.610776036465364e-06, "loss": 0.5491, "step": 2150 }, { "epoch": 1.5827814569536423, "grad_norm": 1.7159051522823725, "learning_rate": 9.60283682766007e-06, "loss": 0.626, "step": 2151 }, { "epoch": 1.5835172921265637, "grad_norm": 1.579002363781222, "learning_rate": 9.594897869578466e-06, "loss": 0.5448, "step": 2152 }, { "epoch": 1.5842531272994849, "grad_norm": 1.657222754412775, "learning_rate": 9.586959167232304e-06, "loss": 0.5686, "step": 2153 }, { "epoch": 1.5849889624724063, "grad_norm": 1.5831061195439633, "learning_rate": 9.579020725633172e-06, "loss": 0.5946, "step": 2154 }, { "epoch": 1.5857247976453275, "grad_norm": 1.7722502015716135, "learning_rate": 9.571082549792513e-06, "loss": 0.5766, "step": 2155 }, { "epoch": 1.5864606328182487, "grad_norm": 1.5254012169183675, "learning_rate": 9.56314464472158e-06, "loss": 0.5317, "step": 2156 }, { "epoch": 1.5871964679911699, "grad_norm": 1.5914122409745899, "learning_rate": 9.555207015431462e-06, "loss": 0.5436, "step": 2157 }, { "epoch": 1.5879323031640913, "grad_norm": 1.6586727125163816, "learning_rate": 9.547269666933085e-06, "loss": 0.5331, "step": 2158 }, { "epoch": 1.5886681383370125, "grad_norm": 1.640506058266485, "learning_rate": 9.53933260423718e-06, "loss": 0.5822, "step": 2159 }, { "epoch": 1.589403973509934, "grad_norm": 1.7253995262440898, "learning_rate": 9.531395832354308e-06, "loss": 0.5762, "step": 2160 }, { "epoch": 1.590139808682855, "grad_norm": 1.9148297242372818, "learning_rate": 9.523459356294849e-06, "loss": 0.5948, "step": 2161 }, { "epoch": 1.5908756438557763, "grad_norm": 1.7291790035344345, "learning_rate": 9.51552318106899e-06, "loss": 0.5934, "step": 2162 }, { "epoch": 1.5916114790286975, "grad_norm": 1.7466892638865619, "learning_rate": 9.507587311686725e-06, "loss": 0.594, "step": 2163 }, { "epoch": 1.5923473142016187, "grad_norm": 1.6238532741698948, "learning_rate": 9.499651753157869e-06, "loss": 0.5354, "step": 2164 }, { "epoch": 1.59308314937454, "grad_norm": 1.9279882649847617, "learning_rate": 9.491716510492025e-06, "loss": 0.5527, "step": 2165 }, { "epoch": 1.5938189845474615, "grad_norm": 1.6360138263698285, "learning_rate": 9.483781588698603e-06, "loss": 0.5291, "step": 2166 }, { "epoch": 1.5945548197203827, "grad_norm": 1.793065661050235, "learning_rate": 9.475846992786817e-06, "loss": 0.5932, "step": 2167 }, { "epoch": 1.595290654893304, "grad_norm": 1.7123736936856366, "learning_rate": 9.46791272776566e-06, "loss": 0.5217, "step": 2168 }, { "epoch": 1.596026490066225, "grad_norm": 1.6967080511504415, "learning_rate": 9.459978798643934e-06, "loss": 0.5557, "step": 2169 }, { "epoch": 1.5967623252391463, "grad_norm": 1.701916203826406, "learning_rate": 9.452045210430218e-06, "loss": 0.5725, "step": 2170 }, { "epoch": 1.5974981604120677, "grad_norm": 1.7050888651819072, "learning_rate": 9.444111968132868e-06, "loss": 0.5695, "step": 2171 }, { "epoch": 1.598233995584989, "grad_norm": 1.687452380252958, "learning_rate": 9.436179076760047e-06, "loss": 0.5615, "step": 2172 }, { "epoch": 1.5989698307579103, "grad_norm": 1.657674422444159, "learning_rate": 9.428246541319673e-06, "loss": 0.5551, "step": 2173 }, { "epoch": 1.5997056659308315, "grad_norm": 1.576576335443867, "learning_rate": 9.420314366819446e-06, "loss": 0.5666, "step": 2174 }, { "epoch": 1.6004415011037527, "grad_norm": 1.6825804304086946, "learning_rate": 9.412382558266841e-06, "loss": 0.5415, "step": 2175 }, { "epoch": 1.601177336276674, "grad_norm": 1.7354327582894893, "learning_rate": 9.404451120669102e-06, "loss": 0.576, "step": 2176 }, { "epoch": 1.6019131714495953, "grad_norm": 1.6027823702465192, "learning_rate": 9.396520059033233e-06, "loss": 0.5371, "step": 2177 }, { "epoch": 1.6026490066225165, "grad_norm": 1.6633192299453694, "learning_rate": 9.388589378366011e-06, "loss": 0.5988, "step": 2178 }, { "epoch": 1.603384841795438, "grad_norm": 1.592799484919593, "learning_rate": 9.380659083673963e-06, "loss": 0.5194, "step": 2179 }, { "epoch": 1.6041206769683591, "grad_norm": 1.5511989601824707, "learning_rate": 9.37272917996337e-06, "loss": 0.5391, "step": 2180 }, { "epoch": 1.6048565121412803, "grad_norm": 1.6578304665549137, "learning_rate": 9.364799672240283e-06, "loss": 0.5632, "step": 2181 }, { "epoch": 1.6055923473142015, "grad_norm": 1.705590202546066, "learning_rate": 9.356870565510488e-06, "loss": 0.5612, "step": 2182 }, { "epoch": 1.6063281824871227, "grad_norm": 1.6497087445883332, "learning_rate": 9.348941864779517e-06, "loss": 0.5538, "step": 2183 }, { "epoch": 1.6070640176600441, "grad_norm": 1.6246493215564124, "learning_rate": 9.341013575052656e-06, "loss": 0.553, "step": 2184 }, { "epoch": 1.6077998528329656, "grad_norm": 1.7221388733092957, "learning_rate": 9.333085701334925e-06, "loss": 0.5735, "step": 2185 }, { "epoch": 1.6085356880058868, "grad_norm": 1.6806298367156012, "learning_rate": 9.325158248631075e-06, "loss": 0.5584, "step": 2186 }, { "epoch": 1.609271523178808, "grad_norm": 1.6237418732164348, "learning_rate": 9.317231221945611e-06, "loss": 0.5947, "step": 2187 }, { "epoch": 1.6100073583517291, "grad_norm": 1.622892714402971, "learning_rate": 9.30930462628275e-06, "loss": 0.5614, "step": 2188 }, { "epoch": 1.6107431935246503, "grad_norm": 1.6294551237481694, "learning_rate": 9.301378466646438e-06, "loss": 0.5209, "step": 2189 }, { "epoch": 1.6114790286975718, "grad_norm": 1.6210482809638986, "learning_rate": 9.293452748040363e-06, "loss": 0.5712, "step": 2190 }, { "epoch": 1.6122148638704932, "grad_norm": 1.6495636836365193, "learning_rate": 9.285527475467918e-06, "loss": 0.5547, "step": 2191 }, { "epoch": 1.6129506990434144, "grad_norm": 1.7018798275356282, "learning_rate": 9.277602653932217e-06, "loss": 0.5914, "step": 2192 }, { "epoch": 1.6136865342163356, "grad_norm": 1.7341373941508667, "learning_rate": 9.269678288436094e-06, "loss": 0.5436, "step": 2193 }, { "epoch": 1.6144223693892568, "grad_norm": 1.7872375309972144, "learning_rate": 9.261754383982093e-06, "loss": 0.6134, "step": 2194 }, { "epoch": 1.615158204562178, "grad_norm": 1.6588667375753743, "learning_rate": 9.253830945572459e-06, "loss": 0.5682, "step": 2195 }, { "epoch": 1.6158940397350994, "grad_norm": 1.8540758829211172, "learning_rate": 9.245907978209164e-06, "loss": 0.5339, "step": 2196 }, { "epoch": 1.6166298749080206, "grad_norm": 1.7315724143629152, "learning_rate": 9.23798548689386e-06, "loss": 0.5567, "step": 2197 }, { "epoch": 1.617365710080942, "grad_norm": 1.7491849501118528, "learning_rate": 9.230063476627908e-06, "loss": 0.5282, "step": 2198 }, { "epoch": 1.6181015452538632, "grad_norm": 1.6659750530777746, "learning_rate": 9.222141952412369e-06, "loss": 0.57, "step": 2199 }, { "epoch": 1.6188373804267844, "grad_norm": 1.628932234884798, "learning_rate": 9.21422091924799e-06, "loss": 0.5873, "step": 2200 }, { "epoch": 1.6195732155997056, "grad_norm": 1.6276910207901034, "learning_rate": 9.20630038213521e-06, "loss": 0.5519, "step": 2201 }, { "epoch": 1.6203090507726268, "grad_norm": 1.5538218307416956, "learning_rate": 9.198380346074159e-06, "loss": 0.5842, "step": 2202 }, { "epoch": 1.6210448859455482, "grad_norm": 1.736188536068997, "learning_rate": 9.190460816064649e-06, "loss": 0.5811, "step": 2203 }, { "epoch": 1.6217807211184696, "grad_norm": 1.5767947912447513, "learning_rate": 9.18254179710616e-06, "loss": 0.5733, "step": 2204 }, { "epoch": 1.6225165562913908, "grad_norm": 1.6353225874547221, "learning_rate": 9.174623294197875e-06, "loss": 0.5573, "step": 2205 }, { "epoch": 1.623252391464312, "grad_norm": 1.6786318819946457, "learning_rate": 9.16670531233863e-06, "loss": 0.5576, "step": 2206 }, { "epoch": 1.6239882266372332, "grad_norm": 1.7706063440690827, "learning_rate": 9.158787856526937e-06, "loss": 0.604, "step": 2207 }, { "epoch": 1.6247240618101544, "grad_norm": 1.533680876600845, "learning_rate": 9.150870931760982e-06, "loss": 0.5512, "step": 2208 }, { "epoch": 1.6254598969830758, "grad_norm": 1.6570511029666237, "learning_rate": 9.14295454303861e-06, "loss": 0.5948, "step": 2209 }, { "epoch": 1.6261957321559972, "grad_norm": 1.6351336457367263, "learning_rate": 9.135038695357324e-06, "loss": 0.5892, "step": 2210 }, { "epoch": 1.6269315673289184, "grad_norm": 1.6529320797216236, "learning_rate": 9.1271233937143e-06, "loss": 0.5061, "step": 2211 }, { "epoch": 1.6276674025018396, "grad_norm": 1.6018770155409041, "learning_rate": 9.119208643106353e-06, "loss": 0.5642, "step": 2212 }, { "epoch": 1.6284032376747608, "grad_norm": 1.6346781818779696, "learning_rate": 9.111294448529955e-06, "loss": 0.5311, "step": 2213 }, { "epoch": 1.629139072847682, "grad_norm": 1.7530388383569955, "learning_rate": 9.10338081498124e-06, "loss": 0.5437, "step": 2214 }, { "epoch": 1.6298749080206034, "grad_norm": 1.5245480943069292, "learning_rate": 9.095467747455965e-06, "loss": 0.524, "step": 2215 }, { "epoch": 1.6306107431935246, "grad_norm": 1.7324285265809922, "learning_rate": 9.087555250949546e-06, "loss": 0.5668, "step": 2216 }, { "epoch": 1.631346578366446, "grad_norm": 1.6252799982179105, "learning_rate": 9.079643330457036e-06, "loss": 0.5518, "step": 2217 }, { "epoch": 1.6320824135393672, "grad_norm": 1.71022899602244, "learning_rate": 9.071731990973118e-06, "loss": 0.6237, "step": 2218 }, { "epoch": 1.6328182487122884, "grad_norm": 1.7123742925225551, "learning_rate": 9.063821237492113e-06, "loss": 0.5857, "step": 2219 }, { "epoch": 1.6335540838852096, "grad_norm": 1.714417150350372, "learning_rate": 9.055911075007977e-06, "loss": 0.59, "step": 2220 }, { "epoch": 1.6342899190581308, "grad_norm": 1.5478913585694138, "learning_rate": 9.048001508514283e-06, "loss": 0.5419, "step": 2221 }, { "epoch": 1.6350257542310522, "grad_norm": 1.7191387682602886, "learning_rate": 9.04009254300423e-06, "loss": 0.5541, "step": 2222 }, { "epoch": 1.6357615894039736, "grad_norm": 1.5816129458619532, "learning_rate": 9.032184183470648e-06, "loss": 0.5609, "step": 2223 }, { "epoch": 1.6364974245768948, "grad_norm": 1.560322840742248, "learning_rate": 9.024276434905969e-06, "loss": 0.5792, "step": 2224 }, { "epoch": 1.637233259749816, "grad_norm": 1.6176043635943607, "learning_rate": 9.016369302302248e-06, "loss": 0.6074, "step": 2225 }, { "epoch": 1.6379690949227372, "grad_norm": 1.7258903182029444, "learning_rate": 9.008462790651154e-06, "loss": 0.5853, "step": 2226 }, { "epoch": 1.6387049300956584, "grad_norm": 1.7345197602014422, "learning_rate": 9.000556904943956e-06, "loss": 0.5973, "step": 2227 }, { "epoch": 1.6394407652685798, "grad_norm": 1.573678368440037, "learning_rate": 8.992651650171528e-06, "loss": 0.5509, "step": 2228 }, { "epoch": 1.6401766004415013, "grad_norm": 1.6482978990993622, "learning_rate": 8.984747031324357e-06, "loss": 0.5875, "step": 2229 }, { "epoch": 1.6409124356144225, "grad_norm": 1.664215281286393, "learning_rate": 8.976843053392518e-06, "loss": 0.5777, "step": 2230 }, { "epoch": 1.6416482707873437, "grad_norm": 1.8125556645917122, "learning_rate": 8.968939721365682e-06, "loss": 0.5785, "step": 2231 }, { "epoch": 1.6423841059602649, "grad_norm": 1.5919435865526799, "learning_rate": 8.961037040233119e-06, "loss": 0.6012, "step": 2232 }, { "epoch": 1.643119941133186, "grad_norm": 1.695356064592736, "learning_rate": 8.95313501498368e-06, "loss": 0.5462, "step": 2233 }, { "epoch": 1.6438557763061075, "grad_norm": 1.7541172255615503, "learning_rate": 8.945233650605808e-06, "loss": 0.5485, "step": 2234 }, { "epoch": 1.6445916114790287, "grad_norm": 1.6634805378453645, "learning_rate": 8.937332952087522e-06, "loss": 0.5731, "step": 2235 }, { "epoch": 1.64532744665195, "grad_norm": 1.6518643143116662, "learning_rate": 8.929432924416433e-06, "loss": 0.5364, "step": 2236 }, { "epoch": 1.6460632818248713, "grad_norm": 1.749948514554574, "learning_rate": 8.92153357257971e-06, "loss": 0.5239, "step": 2237 }, { "epoch": 1.6467991169977925, "grad_norm": 1.7286038375774255, "learning_rate": 8.913634901564115e-06, "loss": 0.568, "step": 2238 }, { "epoch": 1.6475349521707137, "grad_norm": 1.7228506670486763, "learning_rate": 8.905736916355969e-06, "loss": 0.606, "step": 2239 }, { "epoch": 1.648270787343635, "grad_norm": 1.613333466904268, "learning_rate": 8.897839621941156e-06, "loss": 0.5285, "step": 2240 }, { "epoch": 1.6490066225165563, "grad_norm": 1.7024545879249744, "learning_rate": 8.88994302330514e-06, "loss": 0.5927, "step": 2241 }, { "epoch": 1.6497424576894777, "grad_norm": 1.634678526516545, "learning_rate": 8.882047125432929e-06, "loss": 0.5453, "step": 2242 }, { "epoch": 1.6504782928623989, "grad_norm": 1.6221948629359602, "learning_rate": 8.874151933309099e-06, "loss": 0.5253, "step": 2243 }, { "epoch": 1.65121412803532, "grad_norm": 1.6261910765695433, "learning_rate": 8.866257451917774e-06, "loss": 0.6038, "step": 2244 }, { "epoch": 1.6519499632082413, "grad_norm": 1.5566487211177418, "learning_rate": 8.858363686242635e-06, "loss": 0.5097, "step": 2245 }, { "epoch": 1.6526857983811625, "grad_norm": 1.6493734687303732, "learning_rate": 8.850470641266903e-06, "loss": 0.5638, "step": 2246 }, { "epoch": 1.653421633554084, "grad_norm": 1.5699426860809944, "learning_rate": 8.842578321973358e-06, "loss": 0.5856, "step": 2247 }, { "epoch": 1.6541574687270053, "grad_norm": 1.5930860642710145, "learning_rate": 8.834686733344309e-06, "loss": 0.5418, "step": 2248 }, { "epoch": 1.6548933038999265, "grad_norm": 1.7721255382862384, "learning_rate": 8.826795880361603e-06, "loss": 0.5701, "step": 2249 }, { "epoch": 1.6556291390728477, "grad_norm": 1.7429374555236783, "learning_rate": 8.818905768006637e-06, "loss": 0.5465, "step": 2250 }, { "epoch": 1.656364974245769, "grad_norm": 1.6107035700687782, "learning_rate": 8.811016401260327e-06, "loss": 0.5508, "step": 2251 }, { "epoch": 1.65710080941869, "grad_norm": 1.603978849668595, "learning_rate": 8.80312778510312e-06, "loss": 0.5778, "step": 2252 }, { "epoch": 1.6578366445916115, "grad_norm": 1.538191738470185, "learning_rate": 8.795239924514997e-06, "loss": 0.5266, "step": 2253 }, { "epoch": 1.6585724797645327, "grad_norm": 1.710752762097201, "learning_rate": 8.787352824475454e-06, "loss": 0.5523, "step": 2254 }, { "epoch": 1.6593083149374541, "grad_norm": 1.7749029279045965, "learning_rate": 8.779466489963503e-06, "loss": 0.5521, "step": 2255 }, { "epoch": 1.6600441501103753, "grad_norm": 1.6662001059676703, "learning_rate": 8.771580925957691e-06, "loss": 0.5554, "step": 2256 }, { "epoch": 1.6607799852832965, "grad_norm": 1.5787138291569858, "learning_rate": 8.76369613743606e-06, "loss": 0.5372, "step": 2257 }, { "epoch": 1.6615158204562177, "grad_norm": 1.6151820314628083, "learning_rate": 8.75581212937617e-06, "loss": 0.5706, "step": 2258 }, { "epoch": 1.6622516556291391, "grad_norm": 1.5960295136163756, "learning_rate": 8.747928906755087e-06, "loss": 0.5957, "step": 2259 }, { "epoch": 1.6629874908020603, "grad_norm": 1.6072879386256906, "learning_rate": 8.740046474549387e-06, "loss": 0.5185, "step": 2260 }, { "epoch": 1.6637233259749817, "grad_norm": 1.5629521641563562, "learning_rate": 8.732164837735135e-06, "loss": 0.5517, "step": 2261 }, { "epoch": 1.664459161147903, "grad_norm": 1.5775284916102843, "learning_rate": 8.724284001287903e-06, "loss": 0.5296, "step": 2262 }, { "epoch": 1.6651949963208241, "grad_norm": 1.7208078469790558, "learning_rate": 8.716403970182759e-06, "loss": 0.5544, "step": 2263 }, { "epoch": 1.6659308314937453, "grad_norm": 1.7993894558719816, "learning_rate": 8.708524749394249e-06, "loss": 0.5795, "step": 2264 }, { "epoch": 1.6666666666666665, "grad_norm": 1.5867212658614211, "learning_rate": 8.70064634389643e-06, "loss": 0.5256, "step": 2265 }, { "epoch": 1.667402501839588, "grad_norm": 1.706606466157513, "learning_rate": 8.692768758662827e-06, "loss": 0.547, "step": 2266 }, { "epoch": 1.6681383370125094, "grad_norm": 1.6522592456157532, "learning_rate": 8.684891998666445e-06, "loss": 0.5305, "step": 2267 }, { "epoch": 1.6688741721854305, "grad_norm": 1.6345587430932267, "learning_rate": 8.677016068879785e-06, "loss": 0.4987, "step": 2268 }, { "epoch": 1.6696100073583517, "grad_norm": 1.6439617284711228, "learning_rate": 8.66914097427481e-06, "loss": 0.5687, "step": 2269 }, { "epoch": 1.670345842531273, "grad_norm": 1.5877738351915347, "learning_rate": 8.661266719822959e-06, "loss": 0.5325, "step": 2270 }, { "epoch": 1.6710816777041941, "grad_norm": 1.6640079885090093, "learning_rate": 8.653393310495141e-06, "loss": 0.5852, "step": 2271 }, { "epoch": 1.6718175128771156, "grad_norm": 1.672159298027328, "learning_rate": 8.645520751261736e-06, "loss": 0.6046, "step": 2272 }, { "epoch": 1.6725533480500367, "grad_norm": 1.6080914326098872, "learning_rate": 8.637649047092573e-06, "loss": 0.5487, "step": 2273 }, { "epoch": 1.6732891832229582, "grad_norm": 1.5720337674368055, "learning_rate": 8.629778202956963e-06, "loss": 0.5017, "step": 2274 }, { "epoch": 1.6740250183958794, "grad_norm": 1.6867455318463775, "learning_rate": 8.621908223823657e-06, "loss": 0.6039, "step": 2275 }, { "epoch": 1.6747608535688006, "grad_norm": 1.6062406724987193, "learning_rate": 8.614039114660864e-06, "loss": 0.557, "step": 2276 }, { "epoch": 1.6754966887417218, "grad_norm": 1.6673307668031354, "learning_rate": 8.60617088043625e-06, "loss": 0.5652, "step": 2277 }, { "epoch": 1.6762325239146432, "grad_norm": 1.6169228564751406, "learning_rate": 8.598303526116916e-06, "loss": 0.5192, "step": 2278 }, { "epoch": 1.6769683590875644, "grad_norm": 1.671756544536886, "learning_rate": 8.590437056669423e-06, "loss": 0.5682, "step": 2279 }, { "epoch": 1.6777041942604858, "grad_norm": 1.8075996267726218, "learning_rate": 8.582571477059763e-06, "loss": 0.5906, "step": 2280 }, { "epoch": 1.678440029433407, "grad_norm": 1.645710944689608, "learning_rate": 8.574706792253367e-06, "loss": 0.5827, "step": 2281 }, { "epoch": 1.6791758646063282, "grad_norm": 1.613971036156254, "learning_rate": 8.566843007215103e-06, "loss": 0.5675, "step": 2282 }, { "epoch": 1.6799116997792494, "grad_norm": 1.6117077557520547, "learning_rate": 8.558980126909277e-06, "loss": 0.5524, "step": 2283 }, { "epoch": 1.6806475349521706, "grad_norm": 1.7245589921635787, "learning_rate": 8.551118156299613e-06, "loss": 0.5507, "step": 2284 }, { "epoch": 1.681383370125092, "grad_norm": 1.7158836936532758, "learning_rate": 8.543257100349265e-06, "loss": 0.6029, "step": 2285 }, { "epoch": 1.6821192052980134, "grad_norm": 1.7018494176696062, "learning_rate": 8.535396964020812e-06, "loss": 0.5294, "step": 2286 }, { "epoch": 1.6828550404709346, "grad_norm": 1.6696006096001264, "learning_rate": 8.527537752276251e-06, "loss": 0.6084, "step": 2287 }, { "epoch": 1.6835908756438558, "grad_norm": 1.7019338610022428, "learning_rate": 8.519679470076992e-06, "loss": 0.538, "step": 2288 }, { "epoch": 1.684326710816777, "grad_norm": 1.7153644045731073, "learning_rate": 8.511822122383866e-06, "loss": 0.5712, "step": 2289 }, { "epoch": 1.6850625459896982, "grad_norm": 1.7203027484483036, "learning_rate": 8.503965714157108e-06, "loss": 0.5814, "step": 2290 }, { "epoch": 1.6857983811626196, "grad_norm": 1.5722788503914524, "learning_rate": 8.496110250356355e-06, "loss": 0.589, "step": 2291 }, { "epoch": 1.6865342163355408, "grad_norm": 1.6236246129774747, "learning_rate": 8.488255735940663e-06, "loss": 0.5815, "step": 2292 }, { "epoch": 1.6872700515084622, "grad_norm": 1.687828584582342, "learning_rate": 8.480402175868477e-06, "loss": 0.5986, "step": 2293 }, { "epoch": 1.6880058866813834, "grad_norm": 1.6265768446397555, "learning_rate": 8.472549575097639e-06, "loss": 0.5642, "step": 2294 }, { "epoch": 1.6887417218543046, "grad_norm": 1.6727493923718564, "learning_rate": 8.464697938585392e-06, "loss": 0.5633, "step": 2295 }, { "epoch": 1.6894775570272258, "grad_norm": 1.5032135830218314, "learning_rate": 8.456847271288365e-06, "loss": 0.5545, "step": 2296 }, { "epoch": 1.6902133922001472, "grad_norm": 1.5122835212722185, "learning_rate": 8.448997578162572e-06, "loss": 0.5431, "step": 2297 }, { "epoch": 1.6909492273730684, "grad_norm": 1.6398408032847036, "learning_rate": 8.441148864163429e-06, "loss": 0.5217, "step": 2298 }, { "epoch": 1.6916850625459898, "grad_norm": 1.607285522549659, "learning_rate": 8.43330113424571e-06, "loss": 0.5538, "step": 2299 }, { "epoch": 1.692420897718911, "grad_norm": 1.6306080378311107, "learning_rate": 8.42545439336359e-06, "loss": 0.5541, "step": 2300 }, { "epoch": 1.6931567328918322, "grad_norm": 1.6392445232692874, "learning_rate": 8.417608646470605e-06, "loss": 0.5595, "step": 2301 }, { "epoch": 1.6938925680647534, "grad_norm": 1.693342297824956, "learning_rate": 8.409763898519665e-06, "loss": 0.5595, "step": 2302 }, { "epoch": 1.6946284032376746, "grad_norm": 1.6223185269174518, "learning_rate": 8.401920154463056e-06, "loss": 0.5576, "step": 2303 }, { "epoch": 1.695364238410596, "grad_norm": 1.7329713602767167, "learning_rate": 8.394077419252425e-06, "loss": 0.6099, "step": 2304 }, { "epoch": 1.6961000735835174, "grad_norm": 1.5827740231326806, "learning_rate": 8.38623569783878e-06, "loss": 0.5579, "step": 2305 }, { "epoch": 1.6968359087564386, "grad_norm": 1.5674404049272082, "learning_rate": 8.3783949951725e-06, "loss": 0.546, "step": 2306 }, { "epoch": 1.6975717439293598, "grad_norm": 1.5867146591594987, "learning_rate": 8.37055531620331e-06, "loss": 0.542, "step": 2307 }, { "epoch": 1.698307579102281, "grad_norm": 1.6027263056165313, "learning_rate": 8.362716665880289e-06, "loss": 0.5428, "step": 2308 }, { "epoch": 1.6990434142752022, "grad_norm": 1.6551280318358603, "learning_rate": 8.354879049151875e-06, "loss": 0.5106, "step": 2309 }, { "epoch": 1.6997792494481236, "grad_norm": 1.5740822714015474, "learning_rate": 8.347042470965846e-06, "loss": 0.5846, "step": 2310 }, { "epoch": 1.7005150846210448, "grad_norm": 1.7613652519122764, "learning_rate": 8.339206936269328e-06, "loss": 0.5847, "step": 2311 }, { "epoch": 1.7012509197939663, "grad_norm": 1.6239702184496714, "learning_rate": 8.331372450008786e-06, "loss": 0.5445, "step": 2312 }, { "epoch": 1.7019867549668874, "grad_norm": 1.6708240543330075, "learning_rate": 8.323539017130024e-06, "loss": 0.5923, "step": 2313 }, { "epoch": 1.7027225901398086, "grad_norm": 1.6652296135676659, "learning_rate": 8.315706642578178e-06, "loss": 0.5521, "step": 2314 }, { "epoch": 1.7034584253127298, "grad_norm": 1.6203638241589624, "learning_rate": 8.307875331297727e-06, "loss": 0.5562, "step": 2315 }, { "epoch": 1.7041942604856513, "grad_norm": 1.606331685114908, "learning_rate": 8.300045088232468e-06, "loss": 0.595, "step": 2316 }, { "epoch": 1.7049300956585725, "grad_norm": 1.6977046800839937, "learning_rate": 8.292215918325518e-06, "loss": 0.5614, "step": 2317 }, { "epoch": 1.7056659308314939, "grad_norm": 1.7366720263620359, "learning_rate": 8.284387826519335e-06, "loss": 0.5993, "step": 2318 }, { "epoch": 1.706401766004415, "grad_norm": 1.601257158425975, "learning_rate": 8.276560817755686e-06, "loss": 0.5627, "step": 2319 }, { "epoch": 1.7071376011773363, "grad_norm": 1.6566528048071607, "learning_rate": 8.268734896975649e-06, "loss": 0.539, "step": 2320 }, { "epoch": 1.7078734363502575, "grad_norm": 1.4473446429574186, "learning_rate": 8.260910069119625e-06, "loss": 0.5121, "step": 2321 }, { "epoch": 1.7086092715231787, "grad_norm": 1.6119999464243622, "learning_rate": 8.25308633912732e-06, "loss": 0.551, "step": 2322 }, { "epoch": 1.7093451066961, "grad_norm": 1.577112275054741, "learning_rate": 8.245263711937746e-06, "loss": 0.5492, "step": 2323 }, { "epoch": 1.7100809418690215, "grad_norm": 1.7200820841351514, "learning_rate": 8.237442192489225e-06, "loss": 0.5389, "step": 2324 }, { "epoch": 1.7108167770419427, "grad_norm": 1.6614111204444406, "learning_rate": 8.229621785719372e-06, "loss": 0.5498, "step": 2325 }, { "epoch": 1.7115526122148639, "grad_norm": 1.7028884354392748, "learning_rate": 8.221802496565102e-06, "loss": 0.5564, "step": 2326 }, { "epoch": 1.712288447387785, "grad_norm": 1.7207361429957186, "learning_rate": 8.213984329962629e-06, "loss": 0.5603, "step": 2327 }, { "epoch": 1.7130242825607063, "grad_norm": 1.5800552731620217, "learning_rate": 8.206167290847454e-06, "loss": 0.554, "step": 2328 }, { "epoch": 1.7137601177336277, "grad_norm": 1.5652064164750294, "learning_rate": 8.198351384154363e-06, "loss": 0.5455, "step": 2329 }, { "epoch": 1.7144959529065489, "grad_norm": 1.6044725778327948, "learning_rate": 8.190536614817437e-06, "loss": 0.5679, "step": 2330 }, { "epoch": 1.7152317880794703, "grad_norm": 1.5734555570538142, "learning_rate": 8.18272298777003e-06, "loss": 0.5722, "step": 2331 }, { "epoch": 1.7159676232523915, "grad_norm": 1.6698821282868337, "learning_rate": 8.174910507944775e-06, "loss": 0.5989, "step": 2332 }, { "epoch": 1.7167034584253127, "grad_norm": 1.702871954466965, "learning_rate": 8.167099180273591e-06, "loss": 0.5739, "step": 2333 }, { "epoch": 1.7174392935982339, "grad_norm": 1.6324673015465838, "learning_rate": 8.159289009687658e-06, "loss": 0.5219, "step": 2334 }, { "epoch": 1.7181751287711553, "grad_norm": 1.7562046572960892, "learning_rate": 8.15148000111743e-06, "loss": 0.5727, "step": 2335 }, { "epoch": 1.7189109639440765, "grad_norm": 1.7456403349641283, "learning_rate": 8.143672159492626e-06, "loss": 0.5743, "step": 2336 }, { "epoch": 1.719646799116998, "grad_norm": 1.6204688888541847, "learning_rate": 8.135865489742235e-06, "loss": 0.5443, "step": 2337 }, { "epoch": 1.720382634289919, "grad_norm": 1.631743958812732, "learning_rate": 8.128059996794495e-06, "loss": 0.5245, "step": 2338 }, { "epoch": 1.7211184694628403, "grad_norm": 1.6152133250670444, "learning_rate": 8.120255685576913e-06, "loss": 0.5152, "step": 2339 }, { "epoch": 1.7218543046357615, "grad_norm": 1.7391672920702943, "learning_rate": 8.112452561016241e-06, "loss": 0.5433, "step": 2340 }, { "epoch": 1.7225901398086827, "grad_norm": 1.6753847584500896, "learning_rate": 8.10465062803848e-06, "loss": 0.6017, "step": 2341 }, { "epoch": 1.7233259749816041, "grad_norm": 1.888313842415996, "learning_rate": 8.096849891568894e-06, "loss": 0.5447, "step": 2342 }, { "epoch": 1.7240618101545255, "grad_norm": 1.6639416846360875, "learning_rate": 8.089050356531975e-06, "loss": 0.585, "step": 2343 }, { "epoch": 1.7247976453274467, "grad_norm": 1.7184865612727462, "learning_rate": 8.08125202785146e-06, "loss": 0.5431, "step": 2344 }, { "epoch": 1.725533480500368, "grad_norm": 1.5540325800232306, "learning_rate": 8.073454910450337e-06, "loss": 0.5389, "step": 2345 }, { "epoch": 1.7262693156732891, "grad_norm": 1.5954474584518752, "learning_rate": 8.065659009250812e-06, "loss": 0.5245, "step": 2346 }, { "epoch": 1.7270051508462103, "grad_norm": 1.5471955560488546, "learning_rate": 8.05786432917433e-06, "loss": 0.5214, "step": 2347 }, { "epoch": 1.7277409860191317, "grad_norm": 1.7239617699306686, "learning_rate": 8.050070875141572e-06, "loss": 0.5903, "step": 2348 }, { "epoch": 1.728476821192053, "grad_norm": 1.677923503270469, "learning_rate": 8.042278652072437e-06, "loss": 0.5848, "step": 2349 }, { "epoch": 1.7292126563649743, "grad_norm": 1.6611223285601262, "learning_rate": 8.034487664886042e-06, "loss": 0.5498, "step": 2350 }, { "epoch": 1.7299484915378955, "grad_norm": 1.5823821015989217, "learning_rate": 8.026697918500743e-06, "loss": 0.5977, "step": 2351 }, { "epoch": 1.7306843267108167, "grad_norm": 1.664323916817365, "learning_rate": 8.018909417834096e-06, "loss": 0.5501, "step": 2352 }, { "epoch": 1.731420161883738, "grad_norm": 1.6839130853207227, "learning_rate": 8.011122167802869e-06, "loss": 0.5836, "step": 2353 }, { "epoch": 1.7321559970566593, "grad_norm": 1.6580549901544421, "learning_rate": 8.003336173323053e-06, "loss": 0.5566, "step": 2354 }, { "epoch": 1.7328918322295805, "grad_norm": 1.680937274862776, "learning_rate": 7.99555143930984e-06, "loss": 0.5816, "step": 2355 }, { "epoch": 1.733627667402502, "grad_norm": 1.7373532179210185, "learning_rate": 7.987767970677618e-06, "loss": 0.5648, "step": 2356 }, { "epoch": 1.7343635025754232, "grad_norm": 1.5713671346341733, "learning_rate": 7.979985772339995e-06, "loss": 0.4989, "step": 2357 }, { "epoch": 1.7350993377483444, "grad_norm": 1.653914139549647, "learning_rate": 7.972204849209763e-06, "loss": 0.5702, "step": 2358 }, { "epoch": 1.7358351729212655, "grad_norm": 1.6590101925081517, "learning_rate": 7.964425206198907e-06, "loss": 0.563, "step": 2359 }, { "epoch": 1.7365710080941867, "grad_norm": 1.5976975067969488, "learning_rate": 7.956646848218619e-06, "loss": 0.5784, "step": 2360 }, { "epoch": 1.7373068432671082, "grad_norm": 1.6267416036666313, "learning_rate": 7.948869780179266e-06, "loss": 0.5609, "step": 2361 }, { "epoch": 1.7380426784400296, "grad_norm": 1.642152657879642, "learning_rate": 7.941094006990398e-06, "loss": 0.5265, "step": 2362 }, { "epoch": 1.7387785136129508, "grad_norm": 1.5849540518433114, "learning_rate": 7.933319533560766e-06, "loss": 0.5533, "step": 2363 }, { "epoch": 1.739514348785872, "grad_norm": 1.640023987666978, "learning_rate": 7.925546364798283e-06, "loss": 0.5461, "step": 2364 }, { "epoch": 1.7402501839587932, "grad_norm": 1.636263706748099, "learning_rate": 7.917774505610039e-06, "loss": 0.5118, "step": 2365 }, { "epoch": 1.7409860191317144, "grad_norm": 1.7078022051880326, "learning_rate": 7.910003960902314e-06, "loss": 0.5518, "step": 2366 }, { "epoch": 1.7417218543046358, "grad_norm": 1.6661232133466564, "learning_rate": 7.90223473558054e-06, "loss": 0.5606, "step": 2367 }, { "epoch": 1.7424576894775572, "grad_norm": 1.7486966915357776, "learning_rate": 7.89446683454932e-06, "loss": 0.5485, "step": 2368 }, { "epoch": 1.7431935246504784, "grad_norm": 1.6114082468974358, "learning_rate": 7.886700262712432e-06, "loss": 0.5371, "step": 2369 }, { "epoch": 1.7439293598233996, "grad_norm": 1.75311522577699, "learning_rate": 7.878935024972802e-06, "loss": 0.5717, "step": 2370 }, { "epoch": 1.7446651949963208, "grad_norm": 1.5873927446943776, "learning_rate": 7.871171126232516e-06, "loss": 0.5339, "step": 2371 }, { "epoch": 1.745401030169242, "grad_norm": 1.6446649625403258, "learning_rate": 7.863408571392824e-06, "loss": 0.5465, "step": 2372 }, { "epoch": 1.7461368653421634, "grad_norm": 1.708806798204539, "learning_rate": 7.855647365354114e-06, "loss": 0.5811, "step": 2373 }, { "epoch": 1.7468727005150846, "grad_norm": 1.766009653544323, "learning_rate": 7.84788751301593e-06, "loss": 0.5477, "step": 2374 }, { "epoch": 1.747608535688006, "grad_norm": 1.6028590631773518, "learning_rate": 7.840129019276966e-06, "loss": 0.5655, "step": 2375 }, { "epoch": 1.7483443708609272, "grad_norm": 1.5836286375032533, "learning_rate": 7.832371889035052e-06, "loss": 0.5707, "step": 2376 }, { "epoch": 1.7490802060338484, "grad_norm": 1.6369240092345068, "learning_rate": 7.82461612718715e-06, "loss": 0.5615, "step": 2377 }, { "epoch": 1.7498160412067696, "grad_norm": 1.581660114053097, "learning_rate": 7.816861738629377e-06, "loss": 0.5213, "step": 2378 }, { "epoch": 1.7505518763796908, "grad_norm": 1.6334029033422497, "learning_rate": 7.809108728256971e-06, "loss": 0.568, "step": 2379 }, { "epoch": 1.7512877115526122, "grad_norm": 1.6324893540240375, "learning_rate": 7.801357100964295e-06, "loss": 0.5486, "step": 2380 }, { "epoch": 1.7520235467255336, "grad_norm": 1.6426395509960734, "learning_rate": 7.793606861644854e-06, "loss": 0.589, "step": 2381 }, { "epoch": 1.7527593818984548, "grad_norm": 1.7413401945943427, "learning_rate": 7.785858015191261e-06, "loss": 0.5676, "step": 2382 }, { "epoch": 1.753495217071376, "grad_norm": 1.636120809713823, "learning_rate": 7.778110566495256e-06, "loss": 0.5386, "step": 2383 }, { "epoch": 1.7542310522442972, "grad_norm": 1.6842391723874712, "learning_rate": 7.770364520447706e-06, "loss": 0.5633, "step": 2384 }, { "epoch": 1.7549668874172184, "grad_norm": 1.5872938069971727, "learning_rate": 7.76261988193858e-06, "loss": 0.5719, "step": 2385 }, { "epoch": 1.7557027225901398, "grad_norm": 1.5970246873108196, "learning_rate": 7.754876655856957e-06, "loss": 0.5508, "step": 2386 }, { "epoch": 1.7564385577630612, "grad_norm": 1.5733696120181273, "learning_rate": 7.747134847091039e-06, "loss": 0.5462, "step": 2387 }, { "epoch": 1.7571743929359824, "grad_norm": 1.595190052372625, "learning_rate": 7.73939446052812e-06, "loss": 0.5816, "step": 2388 }, { "epoch": 1.7579102281089036, "grad_norm": 1.6509661820675576, "learning_rate": 7.731655501054597e-06, "loss": 0.53, "step": 2389 }, { "epoch": 1.7586460632818248, "grad_norm": 1.6833052410045835, "learning_rate": 7.723917973555975e-06, "loss": 0.5756, "step": 2390 }, { "epoch": 1.759381898454746, "grad_norm": 1.5735677702240343, "learning_rate": 7.716181882916846e-06, "loss": 0.5526, "step": 2391 }, { "epoch": 1.7601177336276674, "grad_norm": 1.6445821530968614, "learning_rate": 7.708447234020898e-06, "loss": 0.5353, "step": 2392 }, { "epoch": 1.7608535688005886, "grad_norm": 1.595109570651103, "learning_rate": 7.700714031750912e-06, "loss": 0.4981, "step": 2393 }, { "epoch": 1.76158940397351, "grad_norm": 1.7169635813315205, "learning_rate": 7.692982280988752e-06, "loss": 0.5803, "step": 2394 }, { "epoch": 1.7623252391464312, "grad_norm": 1.6875808496800762, "learning_rate": 7.685251986615363e-06, "loss": 0.5385, "step": 2395 }, { "epoch": 1.7630610743193524, "grad_norm": 1.7006490613520282, "learning_rate": 7.67752315351078e-06, "loss": 0.5458, "step": 2396 }, { "epoch": 1.7637969094922736, "grad_norm": 1.644081569658929, "learning_rate": 7.669795786554107e-06, "loss": 0.5457, "step": 2397 }, { "epoch": 1.7645327446651948, "grad_norm": 1.61430603407318, "learning_rate": 7.662069890623525e-06, "loss": 0.5303, "step": 2398 }, { "epoch": 1.7652685798381162, "grad_norm": 1.5855191869860212, "learning_rate": 7.654345470596288e-06, "loss": 0.5496, "step": 2399 }, { "epoch": 1.7660044150110377, "grad_norm": 1.7045765539262054, "learning_rate": 7.646622531348715e-06, "loss": 0.5602, "step": 2400 }, { "epoch": 1.7667402501839589, "grad_norm": 1.59237309682843, "learning_rate": 7.63890107775619e-06, "loss": 0.5431, "step": 2401 }, { "epoch": 1.76747608535688, "grad_norm": 1.677265497991892, "learning_rate": 7.631181114693167e-06, "loss": 0.5477, "step": 2402 }, { "epoch": 1.7682119205298013, "grad_norm": 1.7436912310720725, "learning_rate": 7.623462647033151e-06, "loss": 0.5947, "step": 2403 }, { "epoch": 1.7689477557027224, "grad_norm": 1.6579926802709544, "learning_rate": 7.615745679648702e-06, "loss": 0.5641, "step": 2404 }, { "epoch": 1.7696835908756439, "grad_norm": 1.6176652502500397, "learning_rate": 7.608030217411438e-06, "loss": 0.5646, "step": 2405 }, { "epoch": 1.7704194260485653, "grad_norm": 1.6121216892370052, "learning_rate": 7.600316265192028e-06, "loss": 0.5317, "step": 2406 }, { "epoch": 1.7711552612214865, "grad_norm": 1.5488880370037712, "learning_rate": 7.59260382786018e-06, "loss": 0.527, "step": 2407 }, { "epoch": 1.7718910963944077, "grad_norm": 1.7180131106737515, "learning_rate": 7.584892910284655e-06, "loss": 0.5897, "step": 2408 }, { "epoch": 1.7726269315673289, "grad_norm": 1.761358583186697, "learning_rate": 7.577183517333247e-06, "loss": 0.584, "step": 2409 }, { "epoch": 1.77336276674025, "grad_norm": 1.6450252220106238, "learning_rate": 7.569475653872787e-06, "loss": 0.5416, "step": 2410 }, { "epoch": 1.7740986019131715, "grad_norm": 1.578261287422815, "learning_rate": 7.561769324769151e-06, "loss": 0.5267, "step": 2411 }, { "epoch": 1.7748344370860927, "grad_norm": 1.5683703979637185, "learning_rate": 7.554064534887238e-06, "loss": 0.4827, "step": 2412 }, { "epoch": 1.775570272259014, "grad_norm": 1.6908131383052027, "learning_rate": 7.546361289090971e-06, "loss": 0.5504, "step": 2413 }, { "epoch": 1.7763061074319353, "grad_norm": 1.6583149500726926, "learning_rate": 7.538659592243309e-06, "loss": 0.5439, "step": 2414 }, { "epoch": 1.7770419426048565, "grad_norm": 1.5909351946379584, "learning_rate": 7.5309594492062256e-06, "loss": 0.5425, "step": 2415 }, { "epoch": 1.7777777777777777, "grad_norm": 1.615174022397961, "learning_rate": 7.5232608648407166e-06, "loss": 0.5334, "step": 2416 }, { "epoch": 1.778513612950699, "grad_norm": 1.6704061895171465, "learning_rate": 7.515563844006796e-06, "loss": 0.5715, "step": 2417 }, { "epoch": 1.7792494481236203, "grad_norm": 1.648762184059532, "learning_rate": 7.507868391563485e-06, "loss": 0.5273, "step": 2418 }, { "epoch": 1.7799852832965417, "grad_norm": 1.622958340523172, "learning_rate": 7.500174512368814e-06, "loss": 0.5574, "step": 2419 }, { "epoch": 1.780721118469463, "grad_norm": 1.5286454291112226, "learning_rate": 7.492482211279832e-06, "loss": 0.4878, "step": 2420 }, { "epoch": 1.781456953642384, "grad_norm": 1.6335074557629845, "learning_rate": 7.484791493152579e-06, "loss": 0.5564, "step": 2421 }, { "epoch": 1.7821927888153053, "grad_norm": 1.6875647997436443, "learning_rate": 7.477102362842099e-06, "loss": 0.5527, "step": 2422 }, { "epoch": 1.7829286239882265, "grad_norm": 1.6787247706507753, "learning_rate": 7.469414825202439e-06, "loss": 0.5593, "step": 2423 }, { "epoch": 1.783664459161148, "grad_norm": 1.7499793856814168, "learning_rate": 7.461728885086634e-06, "loss": 0.5416, "step": 2424 }, { "epoch": 1.7844002943340693, "grad_norm": 1.6083983526946115, "learning_rate": 7.454044547346708e-06, "loss": 0.5293, "step": 2425 }, { "epoch": 1.7851361295069905, "grad_norm": 1.538599666228674, "learning_rate": 7.44636181683369e-06, "loss": 0.5242, "step": 2426 }, { "epoch": 1.7858719646799117, "grad_norm": 1.6452821459588376, "learning_rate": 7.438680698397575e-06, "loss": 0.5257, "step": 2427 }, { "epoch": 1.786607799852833, "grad_norm": 1.5244775563866868, "learning_rate": 7.431001196887345e-06, "loss": 0.5207, "step": 2428 }, { "epoch": 1.787343635025754, "grad_norm": 1.723396919319546, "learning_rate": 7.423323317150975e-06, "loss": 0.5336, "step": 2429 }, { "epoch": 1.7880794701986755, "grad_norm": 1.6633535107955746, "learning_rate": 7.415647064035397e-06, "loss": 0.5508, "step": 2430 }, { "epoch": 1.7888153053715967, "grad_norm": 1.5926187590521532, "learning_rate": 7.407972442386527e-06, "loss": 0.5317, "step": 2431 }, { "epoch": 1.7895511405445181, "grad_norm": 1.6518002532817206, "learning_rate": 7.400299457049252e-06, "loss": 0.5475, "step": 2432 }, { "epoch": 1.7902869757174393, "grad_norm": 1.6647768505121943, "learning_rate": 7.392628112867415e-06, "loss": 0.5023, "step": 2433 }, { "epoch": 1.7910228108903605, "grad_norm": 1.6972679340148449, "learning_rate": 7.384958414683839e-06, "loss": 0.5849, "step": 2434 }, { "epoch": 1.7917586460632817, "grad_norm": 1.6816549529078637, "learning_rate": 7.377290367340297e-06, "loss": 0.5107, "step": 2435 }, { "epoch": 1.7924944812362031, "grad_norm": 1.748379882461539, "learning_rate": 7.369623975677519e-06, "loss": 0.563, "step": 2436 }, { "epoch": 1.7932303164091243, "grad_norm": 1.7484403157343549, "learning_rate": 7.361959244535199e-06, "loss": 0.5752, "step": 2437 }, { "epoch": 1.7939661515820458, "grad_norm": 1.6084929021138032, "learning_rate": 7.354296178751976e-06, "loss": 0.5548, "step": 2438 }, { "epoch": 1.794701986754967, "grad_norm": 1.6839945339805285, "learning_rate": 7.346634783165435e-06, "loss": 0.5861, "step": 2439 }, { "epoch": 1.7954378219278881, "grad_norm": 1.6382708142678741, "learning_rate": 7.338975062612115e-06, "loss": 0.5218, "step": 2440 }, { "epoch": 1.7961736571008093, "grad_norm": 1.5389504740216644, "learning_rate": 7.331317021927489e-06, "loss": 0.5312, "step": 2441 }, { "epoch": 1.7969094922737305, "grad_norm": 1.7185335450944204, "learning_rate": 7.323660665945972e-06, "loss": 0.5346, "step": 2442 }, { "epoch": 1.797645327446652, "grad_norm": 1.766984177268722, "learning_rate": 7.316005999500924e-06, "loss": 0.567, "step": 2443 }, { "epoch": 1.7983811626195734, "grad_norm": 1.6729664847955887, "learning_rate": 7.308353027424629e-06, "loss": 0.4978, "step": 2444 }, { "epoch": 1.7991169977924946, "grad_norm": 1.6432524315786419, "learning_rate": 7.300701754548295e-06, "loss": 0.5478, "step": 2445 }, { "epoch": 1.7998528329654158, "grad_norm": 1.6782276202259596, "learning_rate": 7.293052185702079e-06, "loss": 0.5109, "step": 2446 }, { "epoch": 1.800588668138337, "grad_norm": 1.656255024257162, "learning_rate": 7.285404325715044e-06, "loss": 0.5475, "step": 2447 }, { "epoch": 1.8013245033112582, "grad_norm": 1.7566996931423353, "learning_rate": 7.277758179415175e-06, "loss": 0.5416, "step": 2448 }, { "epoch": 1.8013245033112582, "eval_en_loss": 0.9356659054756165, "eval_en_runtime": 14.6694, "eval_en_samples_per_second": 66.669, "eval_en_steps_per_second": 0.136, "step": 2448 }, { "epoch": 1.8013245033112582, "eval_eu_loss": 0.8410176634788513, "eval_eu_runtime": 14.9659, "eval_eu_samples_per_second": 65.816, "eval_eu_steps_per_second": 0.134, "step": 2448 }, { "epoch": 1.8020603384841796, "grad_norm": 1.7449639116331777, "learning_rate": 7.270113751629388e-06, "loss": 0.5646, "step": 2449 }, { "epoch": 1.8027961736571008, "grad_norm": 1.617486670341438, "learning_rate": 7.262471047183501e-06, "loss": 0.5478, "step": 2450 }, { "epoch": 1.8035320088300222, "grad_norm": 1.734893119470817, "learning_rate": 7.2548300709022455e-06, "loss": 0.5785, "step": 2451 }, { "epoch": 1.8042678440029434, "grad_norm": 1.575752665100807, "learning_rate": 7.247190827609273e-06, "loss": 0.5198, "step": 2452 }, { "epoch": 1.8050036791758646, "grad_norm": 1.5889943063916812, "learning_rate": 7.239553322127131e-06, "loss": 0.5513, "step": 2453 }, { "epoch": 1.8057395143487858, "grad_norm": 1.6079208283937405, "learning_rate": 7.231917559277266e-06, "loss": 0.5171, "step": 2454 }, { "epoch": 1.8064753495217072, "grad_norm": 1.6559644835918488, "learning_rate": 7.224283543880041e-06, "loss": 0.5705, "step": 2455 }, { "epoch": 1.8072111846946284, "grad_norm": 1.6437516429969772, "learning_rate": 7.216651280754701e-06, "loss": 0.5468, "step": 2456 }, { "epoch": 1.8079470198675498, "grad_norm": 1.7573204001279938, "learning_rate": 7.209020774719386e-06, "loss": 0.545, "step": 2457 }, { "epoch": 1.808682855040471, "grad_norm": 2.267932884670688, "learning_rate": 7.201392030591137e-06, "loss": 0.5645, "step": 2458 }, { "epoch": 1.8094186902133922, "grad_norm": 1.693189869779365, "learning_rate": 7.193765053185872e-06, "loss": 0.5913, "step": 2459 }, { "epoch": 1.8101545253863134, "grad_norm": 1.61088982513275, "learning_rate": 7.186139847318394e-06, "loss": 0.566, "step": 2460 }, { "epoch": 1.8108903605592346, "grad_norm": 1.599311415818047, "learning_rate": 7.178516417802399e-06, "loss": 0.5122, "step": 2461 }, { "epoch": 1.811626195732156, "grad_norm": 1.7264345591286077, "learning_rate": 7.170894769450451e-06, "loss": 0.5848, "step": 2462 }, { "epoch": 1.8123620309050774, "grad_norm": 1.6194197434537798, "learning_rate": 7.163274907073987e-06, "loss": 0.5311, "step": 2463 }, { "epoch": 1.8130978660779986, "grad_norm": 1.6032688413640344, "learning_rate": 7.155656835483331e-06, "loss": 0.5199, "step": 2464 }, { "epoch": 1.8138337012509198, "grad_norm": 1.6134420943726775, "learning_rate": 7.148040559487664e-06, "loss": 0.5477, "step": 2465 }, { "epoch": 1.814569536423841, "grad_norm": 1.6064967398540455, "learning_rate": 7.1404260838950344e-06, "loss": 0.5551, "step": 2466 }, { "epoch": 1.8153053715967622, "grad_norm": 1.707042434619857, "learning_rate": 7.132813413512361e-06, "loss": 0.5349, "step": 2467 }, { "epoch": 1.8160412067696836, "grad_norm": 1.6149727032631507, "learning_rate": 7.125202553145416e-06, "loss": 0.5748, "step": 2468 }, { "epoch": 1.8167770419426048, "grad_norm": 1.6370941180596321, "learning_rate": 7.117593507598827e-06, "loss": 0.538, "step": 2469 }, { "epoch": 1.8175128771155262, "grad_norm": 1.6144246168661425, "learning_rate": 7.10998628167609e-06, "loss": 0.6074, "step": 2470 }, { "epoch": 1.8182487122884474, "grad_norm": 1.6120246862615482, "learning_rate": 7.102380880179538e-06, "loss": 0.5621, "step": 2471 }, { "epoch": 1.8189845474613686, "grad_norm": 1.5885970711491622, "learning_rate": 7.0947773079103565e-06, "loss": 0.5286, "step": 2472 }, { "epoch": 1.8197203826342898, "grad_norm": 1.508841002587159, "learning_rate": 7.087175569668576e-06, "loss": 0.512, "step": 2473 }, { "epoch": 1.8204562178072112, "grad_norm": 1.6350653023073896, "learning_rate": 7.079575670253074e-06, "loss": 0.5229, "step": 2474 }, { "epoch": 1.8211920529801324, "grad_norm": 1.6833101935089136, "learning_rate": 7.071977614461559e-06, "loss": 0.5937, "step": 2475 }, { "epoch": 1.8219278881530538, "grad_norm": 1.6231238642309747, "learning_rate": 7.064381407090584e-06, "loss": 0.5882, "step": 2476 }, { "epoch": 1.822663723325975, "grad_norm": 1.6313035111072751, "learning_rate": 7.056787052935528e-06, "loss": 0.532, "step": 2477 }, { "epoch": 1.8233995584988962, "grad_norm": 1.7063820756730512, "learning_rate": 7.049194556790601e-06, "loss": 0.5906, "step": 2478 }, { "epoch": 1.8241353936718174, "grad_norm": 1.7087975826419117, "learning_rate": 7.041603923448847e-06, "loss": 0.5513, "step": 2479 }, { "epoch": 1.8248712288447386, "grad_norm": 1.699222207881662, "learning_rate": 7.034015157702127e-06, "loss": 0.5482, "step": 2480 }, { "epoch": 1.82560706401766, "grad_norm": 1.7140358834668825, "learning_rate": 7.026428264341125e-06, "loss": 0.581, "step": 2481 }, { "epoch": 1.8263428991905815, "grad_norm": 1.6260044032241605, "learning_rate": 7.018843248155345e-06, "loss": 0.5566, "step": 2482 }, { "epoch": 1.8270787343635027, "grad_norm": 1.627300953709885, "learning_rate": 7.0112601139331005e-06, "loss": 0.511, "step": 2483 }, { "epoch": 1.8278145695364238, "grad_norm": 1.6140762825654333, "learning_rate": 7.003678866461518e-06, "loss": 0.5523, "step": 2484 }, { "epoch": 1.828550404709345, "grad_norm": 1.6589114230298583, "learning_rate": 6.996099510526546e-06, "loss": 0.5434, "step": 2485 }, { "epoch": 1.8292862398822662, "grad_norm": 1.6649438716151623, "learning_rate": 6.988522050912921e-06, "loss": 0.58, "step": 2486 }, { "epoch": 1.8300220750551877, "grad_norm": 1.5927153376244017, "learning_rate": 6.980946492404188e-06, "loss": 0.5376, "step": 2487 }, { "epoch": 1.8307579102281089, "grad_norm": 1.533368327120076, "learning_rate": 6.973372839782699e-06, "loss": 0.5518, "step": 2488 }, { "epoch": 1.8314937454010303, "grad_norm": 1.6962076199099696, "learning_rate": 6.965801097829596e-06, "loss": 0.582, "step": 2489 }, { "epoch": 1.8322295805739515, "grad_norm": 1.6004113158921032, "learning_rate": 6.9582312713248144e-06, "loss": 0.5496, "step": 2490 }, { "epoch": 1.8329654157468727, "grad_norm": 1.6746890331270168, "learning_rate": 6.950663365047083e-06, "loss": 0.5587, "step": 2491 }, { "epoch": 1.8337012509197939, "grad_norm": 1.788091682787091, "learning_rate": 6.9430973837739204e-06, "loss": 0.5365, "step": 2492 }, { "epoch": 1.8344370860927153, "grad_norm": 1.6531657042415953, "learning_rate": 6.935533332281619e-06, "loss": 0.5967, "step": 2493 }, { "epoch": 1.8351729212656365, "grad_norm": 1.6398564449762503, "learning_rate": 6.927971215345271e-06, "loss": 0.5696, "step": 2494 }, { "epoch": 1.8359087564385579, "grad_norm": 1.6616860109979101, "learning_rate": 6.920411037738735e-06, "loss": 0.5595, "step": 2495 }, { "epoch": 1.836644591611479, "grad_norm": 1.582559586372339, "learning_rate": 6.91285280423464e-06, "loss": 0.5141, "step": 2496 }, { "epoch": 1.8373804267844003, "grad_norm": 1.556844945621914, "learning_rate": 6.905296519604407e-06, "loss": 0.5244, "step": 2497 }, { "epoch": 1.8381162619573215, "grad_norm": 1.592103085787235, "learning_rate": 6.89774218861821e-06, "loss": 0.563, "step": 2498 }, { "epoch": 1.8388520971302427, "grad_norm": 1.5903471928527984, "learning_rate": 6.890189816044992e-06, "loss": 0.5533, "step": 2499 }, { "epoch": 1.839587932303164, "grad_norm": 1.701284497764011, "learning_rate": 6.8826394066524695e-06, "loss": 0.5404, "step": 2500 }, { "epoch": 1.8403237674760855, "grad_norm": 1.6392798716558004, "learning_rate": 6.875090965207108e-06, "loss": 0.5376, "step": 2501 }, { "epoch": 1.8410596026490067, "grad_norm": 1.6863212556939284, "learning_rate": 6.867544496474132e-06, "loss": 0.5472, "step": 2502 }, { "epoch": 1.841795437821928, "grad_norm": 1.764096288998047, "learning_rate": 6.860000005217533e-06, "loss": 0.5618, "step": 2503 }, { "epoch": 1.842531272994849, "grad_norm": 1.7151585556247095, "learning_rate": 6.852457496200042e-06, "loss": 0.5206, "step": 2504 }, { "epoch": 1.8432671081677703, "grad_norm": 1.5937678842578054, "learning_rate": 6.8449169741831356e-06, "loss": 0.5073, "step": 2505 }, { "epoch": 1.8440029433406917, "grad_norm": 1.6004373924087856, "learning_rate": 6.837378443927052e-06, "loss": 0.5671, "step": 2506 }, { "epoch": 1.844738778513613, "grad_norm": 1.710818085334355, "learning_rate": 6.829841910190759e-06, "loss": 0.5585, "step": 2507 }, { "epoch": 1.8454746136865343, "grad_norm": 1.6894369518997978, "learning_rate": 6.8223073777319625e-06, "loss": 0.5585, "step": 2508 }, { "epoch": 1.8462104488594555, "grad_norm": 1.6150460851374673, "learning_rate": 6.814774851307118e-06, "loss": 0.5361, "step": 2509 }, { "epoch": 1.8469462840323767, "grad_norm": 1.609935126827577, "learning_rate": 6.807244335671401e-06, "loss": 0.5416, "step": 2510 }, { "epoch": 1.847682119205298, "grad_norm": 1.642719241641002, "learning_rate": 6.799715835578721e-06, "loss": 0.5814, "step": 2511 }, { "epoch": 1.8484179543782193, "grad_norm": 1.6401808764875245, "learning_rate": 6.7921893557817246e-06, "loss": 0.593, "step": 2512 }, { "epoch": 1.8491537895511405, "grad_norm": 1.6603142455158606, "learning_rate": 6.784664901031772e-06, "loss": 0.5346, "step": 2513 }, { "epoch": 1.849889624724062, "grad_norm": 1.5731073507396458, "learning_rate": 6.777142476078944e-06, "loss": 0.4906, "step": 2514 }, { "epoch": 1.8506254598969831, "grad_norm": 1.6737541482274731, "learning_rate": 6.769622085672054e-06, "loss": 0.5699, "step": 2515 }, { "epoch": 1.8513612950699043, "grad_norm": 1.5803858193880294, "learning_rate": 6.762103734558616e-06, "loss": 0.5063, "step": 2516 }, { "epoch": 1.8520971302428255, "grad_norm": 1.6048100720079939, "learning_rate": 6.754587427484862e-06, "loss": 0.5533, "step": 2517 }, { "epoch": 1.8528329654157467, "grad_norm": 1.6900092787571757, "learning_rate": 6.747073169195739e-06, "loss": 0.5702, "step": 2518 }, { "epoch": 1.8535688005886681, "grad_norm": 1.5654229811784632, "learning_rate": 6.7395609644348905e-06, "loss": 0.519, "step": 2519 }, { "epoch": 1.8543046357615895, "grad_norm": 1.61269058567931, "learning_rate": 6.732050817944666e-06, "loss": 0.5217, "step": 2520 }, { "epoch": 1.8550404709345107, "grad_norm": 1.598716329743238, "learning_rate": 6.724542734466127e-06, "loss": 0.5347, "step": 2521 }, { "epoch": 1.855776306107432, "grad_norm": 1.645617284999151, "learning_rate": 6.717036718739021e-06, "loss": 0.5829, "step": 2522 }, { "epoch": 1.8565121412803531, "grad_norm": 1.6528100011498184, "learning_rate": 6.7095327755017856e-06, "loss": 0.5513, "step": 2523 }, { "epoch": 1.8572479764532743, "grad_norm": 1.6484316627264894, "learning_rate": 6.70203090949157e-06, "loss": 0.5409, "step": 2524 }, { "epoch": 1.8579838116261957, "grad_norm": 1.7373125869873673, "learning_rate": 6.694531125444193e-06, "loss": 0.5739, "step": 2525 }, { "epoch": 1.858719646799117, "grad_norm": 1.6237535512334966, "learning_rate": 6.687033428094164e-06, "loss": 0.5764, "step": 2526 }, { "epoch": 1.8594554819720384, "grad_norm": 1.533249219433469, "learning_rate": 6.679537822174682e-06, "loss": 0.5372, "step": 2527 }, { "epoch": 1.8601913171449596, "grad_norm": 1.6727696557298795, "learning_rate": 6.672044312417618e-06, "loss": 0.554, "step": 2528 }, { "epoch": 1.8609271523178808, "grad_norm": 1.6733929387010986, "learning_rate": 6.664552903553519e-06, "loss": 0.5681, "step": 2529 }, { "epoch": 1.861662987490802, "grad_norm": 1.6027874413840584, "learning_rate": 6.657063600311616e-06, "loss": 0.5573, "step": 2530 }, { "epoch": 1.8623988226637234, "grad_norm": 1.634447811935984, "learning_rate": 6.649576407419801e-06, "loss": 0.5438, "step": 2531 }, { "epoch": 1.8631346578366446, "grad_norm": 1.6099005665041244, "learning_rate": 6.642091329604632e-06, "loss": 0.519, "step": 2532 }, { "epoch": 1.863870493009566, "grad_norm": 1.6322859624923425, "learning_rate": 6.634608371591343e-06, "loss": 0.5365, "step": 2533 }, { "epoch": 1.8646063281824872, "grad_norm": 1.6804092639271562, "learning_rate": 6.627127538103821e-06, "loss": 0.5482, "step": 2534 }, { "epoch": 1.8653421633554084, "grad_norm": 1.6552416628264206, "learning_rate": 6.619648833864614e-06, "loss": 0.5591, "step": 2535 }, { "epoch": 1.8660779985283296, "grad_norm": 1.6117413146056658, "learning_rate": 6.6121722635949244e-06, "loss": 0.5419, "step": 2536 }, { "epoch": 1.8668138337012508, "grad_norm": 1.7919723110568364, "learning_rate": 6.604697832014612e-06, "loss": 0.5156, "step": 2537 }, { "epoch": 1.8675496688741722, "grad_norm": 1.5443322875358525, "learning_rate": 6.597225543842176e-06, "loss": 0.5244, "step": 2538 }, { "epoch": 1.8682855040470936, "grad_norm": 1.7595462541095048, "learning_rate": 6.58975540379478e-06, "loss": 0.5021, "step": 2539 }, { "epoch": 1.8690213392200148, "grad_norm": 1.659490253597181, "learning_rate": 6.582287416588216e-06, "loss": 0.5845, "step": 2540 }, { "epoch": 1.869757174392936, "grad_norm": 1.8057211058164824, "learning_rate": 6.574821586936921e-06, "loss": 0.5623, "step": 2541 }, { "epoch": 1.8704930095658572, "grad_norm": 1.5792076347031863, "learning_rate": 6.567357919553973e-06, "loss": 0.5539, "step": 2542 }, { "epoch": 1.8712288447387784, "grad_norm": 1.5672177511821372, "learning_rate": 6.559896419151085e-06, "loss": 0.5652, "step": 2543 }, { "epoch": 1.8719646799116998, "grad_norm": 1.656213193693715, "learning_rate": 6.552437090438597e-06, "loss": 0.5574, "step": 2544 }, { "epoch": 1.8727005150846212, "grad_norm": 1.6434800664945937, "learning_rate": 6.544979938125485e-06, "loss": 0.5405, "step": 2545 }, { "epoch": 1.8734363502575424, "grad_norm": 1.5707822332704562, "learning_rate": 6.537524966919346e-06, "loss": 0.5483, "step": 2546 }, { "epoch": 1.8741721854304636, "grad_norm": 1.5890171147655536, "learning_rate": 6.5300721815263956e-06, "loss": 0.5531, "step": 2547 }, { "epoch": 1.8749080206033848, "grad_norm": 1.5866948525561457, "learning_rate": 6.522621586651485e-06, "loss": 0.5818, "step": 2548 }, { "epoch": 1.875643855776306, "grad_norm": 1.566322621886349, "learning_rate": 6.51517318699807e-06, "loss": 0.5545, "step": 2549 }, { "epoch": 1.8763796909492274, "grad_norm": 1.645678798532333, "learning_rate": 6.507726987268219e-06, "loss": 0.5368, "step": 2550 }, { "epoch": 1.8771155261221486, "grad_norm": 1.693955891727557, "learning_rate": 6.5002829921626206e-06, "loss": 0.5503, "step": 2551 }, { "epoch": 1.87785136129507, "grad_norm": 1.649698106413513, "learning_rate": 6.492841206380566e-06, "loss": 0.5474, "step": 2552 }, { "epoch": 1.8785871964679912, "grad_norm": 1.6391957930610064, "learning_rate": 6.485401634619952e-06, "loss": 0.497, "step": 2553 }, { "epoch": 1.8793230316409124, "grad_norm": 1.5903793241195672, "learning_rate": 6.477964281577282e-06, "loss": 0.5474, "step": 2554 }, { "epoch": 1.8800588668138336, "grad_norm": 1.7354805371091313, "learning_rate": 6.470529151947652e-06, "loss": 0.5668, "step": 2555 }, { "epoch": 1.8807947019867548, "grad_norm": 1.592599614475279, "learning_rate": 6.463096250424756e-06, "loss": 0.5638, "step": 2556 }, { "epoch": 1.8815305371596762, "grad_norm": 1.5688830403892253, "learning_rate": 6.4556655817008895e-06, "loss": 0.5318, "step": 2557 }, { "epoch": 1.8822663723325976, "grad_norm": 1.5345508967479415, "learning_rate": 6.4482371504669276e-06, "loss": 0.5222, "step": 2558 }, { "epoch": 1.8830022075055188, "grad_norm": 1.6723727645098303, "learning_rate": 6.440810961412338e-06, "loss": 0.5301, "step": 2559 }, { "epoch": 1.88373804267844, "grad_norm": 1.5520905262187248, "learning_rate": 6.433387019225175e-06, "loss": 0.5471, "step": 2560 }, { "epoch": 1.8844738778513612, "grad_norm": 1.68273223493968, "learning_rate": 6.425965328592068e-06, "loss": 0.5607, "step": 2561 }, { "epoch": 1.8852097130242824, "grad_norm": 1.624531205363433, "learning_rate": 6.418545894198228e-06, "loss": 0.5607, "step": 2562 }, { "epoch": 1.8859455481972038, "grad_norm": 1.8456497133278564, "learning_rate": 6.411128720727448e-06, "loss": 0.5697, "step": 2563 }, { "epoch": 1.8866813833701253, "grad_norm": 1.699474489459609, "learning_rate": 6.403713812862085e-06, "loss": 0.5395, "step": 2564 }, { "epoch": 1.8874172185430464, "grad_norm": 1.6844397121028625, "learning_rate": 6.3963011752830706e-06, "loss": 0.5567, "step": 2565 }, { "epoch": 1.8881530537159676, "grad_norm": 1.6594475874052153, "learning_rate": 6.3888908126699015e-06, "loss": 0.5201, "step": 2566 }, { "epoch": 1.8888888888888888, "grad_norm": 1.5723471705275496, "learning_rate": 6.381482729700636e-06, "loss": 0.5963, "step": 2567 }, { "epoch": 1.88962472406181, "grad_norm": 1.583036762508335, "learning_rate": 6.374076931051899e-06, "loss": 0.4879, "step": 2568 }, { "epoch": 1.8903605592347315, "grad_norm": 1.5947244371296856, "learning_rate": 6.366673421398869e-06, "loss": 0.4865, "step": 2569 }, { "epoch": 1.8910963944076526, "grad_norm": 1.635540688315594, "learning_rate": 6.359272205415274e-06, "loss": 0.5436, "step": 2570 }, { "epoch": 1.891832229580574, "grad_norm": 1.5714791585374204, "learning_rate": 6.351873287773413e-06, "loss": 0.5428, "step": 2571 }, { "epoch": 1.8925680647534953, "grad_norm": 1.5610084180596018, "learning_rate": 6.344476673144113e-06, "loss": 0.4919, "step": 2572 }, { "epoch": 1.8933038999264165, "grad_norm": 1.5818852211953949, "learning_rate": 6.337082366196756e-06, "loss": 0.5454, "step": 2573 }, { "epoch": 1.8940397350993377, "grad_norm": 1.5762776709372677, "learning_rate": 6.329690371599271e-06, "loss": 0.5326, "step": 2574 }, { "epoch": 1.894775570272259, "grad_norm": 1.5202112708620692, "learning_rate": 6.322300694018122e-06, "loss": 0.5514, "step": 2575 }, { "epoch": 1.8955114054451803, "grad_norm": 1.5636981714913794, "learning_rate": 6.314913338118308e-06, "loss": 0.5457, "step": 2576 }, { "epoch": 1.8962472406181017, "grad_norm": 1.6257314993463274, "learning_rate": 6.307528308563371e-06, "loss": 0.5105, "step": 2577 }, { "epoch": 1.8969830757910229, "grad_norm": 1.630997988725555, "learning_rate": 6.3001456100153754e-06, "loss": 0.535, "step": 2578 }, { "epoch": 1.897718910963944, "grad_norm": 1.5787127860852082, "learning_rate": 6.292765247134916e-06, "loss": 0.5297, "step": 2579 }, { "epoch": 1.8984547461368653, "grad_norm": 1.6436782038906417, "learning_rate": 6.285387224581123e-06, "loss": 0.5326, "step": 2580 }, { "epoch": 1.8991905813097865, "grad_norm": 1.587312352690787, "learning_rate": 6.278011547011638e-06, "loss": 0.5375, "step": 2581 }, { "epoch": 1.8999264164827079, "grad_norm": 1.6591057984804471, "learning_rate": 6.270638219082617e-06, "loss": 0.5641, "step": 2582 }, { "epoch": 1.9006622516556293, "grad_norm": 1.8031082666942515, "learning_rate": 6.263267245448754e-06, "loss": 0.5809, "step": 2583 }, { "epoch": 1.9013980868285505, "grad_norm": 1.5460324154254546, "learning_rate": 6.255898630763238e-06, "loss": 0.5297, "step": 2584 }, { "epoch": 1.9021339220014717, "grad_norm": 1.6329103998149286, "learning_rate": 6.248532379677773e-06, "loss": 0.5187, "step": 2585 }, { "epoch": 1.9028697571743929, "grad_norm": 1.6645090978447852, "learning_rate": 6.241168496842578e-06, "loss": 0.5512, "step": 2586 }, { "epoch": 1.903605592347314, "grad_norm": 1.5595548772765422, "learning_rate": 6.233806986906367e-06, "loss": 0.4824, "step": 2587 }, { "epoch": 1.9043414275202355, "grad_norm": 1.6855967657348787, "learning_rate": 6.226447854516358e-06, "loss": 0.5357, "step": 2588 }, { "epoch": 1.9050772626931567, "grad_norm": 1.6064206636391902, "learning_rate": 6.21909110431828e-06, "loss": 0.5275, "step": 2589 }, { "epoch": 1.905813097866078, "grad_norm": 1.6980584686786258, "learning_rate": 6.211736740956343e-06, "loss": 0.557, "step": 2590 }, { "epoch": 1.9065489330389993, "grad_norm": 1.669551147401288, "learning_rate": 6.204384769073253e-06, "loss": 0.5456, "step": 2591 }, { "epoch": 1.9072847682119205, "grad_norm": 1.6447158103575996, "learning_rate": 6.197035193310217e-06, "loss": 0.5099, "step": 2592 }, { "epoch": 1.9080206033848417, "grad_norm": 1.765159544015241, "learning_rate": 6.189688018306919e-06, "loss": 0.5595, "step": 2593 }, { "epoch": 1.9087564385577631, "grad_norm": 1.6215430195833724, "learning_rate": 6.1823432487015265e-06, "loss": 0.5279, "step": 2594 }, { "epoch": 1.9094922737306843, "grad_norm": 1.6858106401422044, "learning_rate": 6.1750008891306975e-06, "loss": 0.528, "step": 2595 }, { "epoch": 1.9102281089036057, "grad_norm": 1.724121876167167, "learning_rate": 6.167660944229561e-06, "loss": 0.5302, "step": 2596 }, { "epoch": 1.910963944076527, "grad_norm": 1.6751726052186093, "learning_rate": 6.160323418631721e-06, "loss": 0.5659, "step": 2597 }, { "epoch": 1.9116997792494481, "grad_norm": 1.7169327292665502, "learning_rate": 6.152988316969263e-06, "loss": 0.5262, "step": 2598 }, { "epoch": 1.9124356144223693, "grad_norm": 1.5950622592907109, "learning_rate": 6.145655643872733e-06, "loss": 0.535, "step": 2599 }, { "epoch": 1.9131714495952905, "grad_norm": 1.6349425902705315, "learning_rate": 6.138325403971144e-06, "loss": 0.5483, "step": 2600 }, { "epoch": 1.913907284768212, "grad_norm": 1.617142832720518, "learning_rate": 6.130997601891986e-06, "loss": 0.5524, "step": 2601 }, { "epoch": 1.9146431199411333, "grad_norm": 1.5279990335462168, "learning_rate": 6.123672242261191e-06, "loss": 0.4969, "step": 2602 }, { "epoch": 1.9153789551140545, "grad_norm": 1.667627495984815, "learning_rate": 6.116349329703164e-06, "loss": 0.55, "step": 2603 }, { "epoch": 1.9161147902869757, "grad_norm": 1.56166516402945, "learning_rate": 6.1090288688407585e-06, "loss": 0.5297, "step": 2604 }, { "epoch": 1.916850625459897, "grad_norm": 1.539480581539991, "learning_rate": 6.101710864295279e-06, "loss": 0.4861, "step": 2605 }, { "epoch": 1.9175864606328181, "grad_norm": 1.6790185374482078, "learning_rate": 6.094395320686482e-06, "loss": 0.5614, "step": 2606 }, { "epoch": 1.9183222958057395, "grad_norm": 1.6724396926812855, "learning_rate": 6.087082242632574e-06, "loss": 0.577, "step": 2607 }, { "epoch": 1.9190581309786607, "grad_norm": 1.676466938668875, "learning_rate": 6.0797716347502e-06, "loss": 0.5598, "step": 2608 }, { "epoch": 1.9197939661515822, "grad_norm": 1.6765882239371444, "learning_rate": 6.0724635016544434e-06, "loss": 0.5304, "step": 2609 }, { "epoch": 1.9205298013245033, "grad_norm": 1.6123697950169587, "learning_rate": 6.065157847958831e-06, "loss": 0.5298, "step": 2610 }, { "epoch": 1.9212656364974245, "grad_norm": 1.8851748911685364, "learning_rate": 6.057854678275326e-06, "loss": 0.5491, "step": 2611 }, { "epoch": 1.9220014716703457, "grad_norm": 1.7168627073570941, "learning_rate": 6.050553997214312e-06, "loss": 0.5926, "step": 2612 }, { "epoch": 1.9227373068432672, "grad_norm": 1.659493470069045, "learning_rate": 6.043255809384617e-06, "loss": 0.5546, "step": 2613 }, { "epoch": 1.9234731420161884, "grad_norm": 1.6209844641354971, "learning_rate": 6.035960119393483e-06, "loss": 0.5335, "step": 2614 }, { "epoch": 1.9242089771891098, "grad_norm": 1.7100690892383974, "learning_rate": 6.028666931846576e-06, "loss": 0.5755, "step": 2615 }, { "epoch": 1.924944812362031, "grad_norm": 1.6540585123850629, "learning_rate": 6.021376251347994e-06, "loss": 0.555, "step": 2616 }, { "epoch": 1.9256806475349522, "grad_norm": 1.605903094292241, "learning_rate": 6.014088082500241e-06, "loss": 0.5384, "step": 2617 }, { "epoch": 1.9264164827078734, "grad_norm": 1.6123723716876128, "learning_rate": 6.006802429904235e-06, "loss": 0.5687, "step": 2618 }, { "epoch": 1.9271523178807946, "grad_norm": 1.6352638594436353, "learning_rate": 5.999519298159313e-06, "loss": 0.5335, "step": 2619 }, { "epoch": 1.927888153053716, "grad_norm": 1.5173717621328167, "learning_rate": 5.9922386918632145e-06, "loss": 0.4941, "step": 2620 }, { "epoch": 1.9286239882266374, "grad_norm": 1.613806034867894, "learning_rate": 5.984960615612083e-06, "loss": 0.5737, "step": 2621 }, { "epoch": 1.9293598233995586, "grad_norm": 1.6312663432312176, "learning_rate": 5.977685074000478e-06, "loss": 0.5686, "step": 2622 }, { "epoch": 1.9300956585724798, "grad_norm": 1.6406519118170466, "learning_rate": 5.9704120716213435e-06, "loss": 0.5608, "step": 2623 }, { "epoch": 1.930831493745401, "grad_norm": 1.6123048431693174, "learning_rate": 5.9631416130660235e-06, "loss": 0.5521, "step": 2624 }, { "epoch": 1.9315673289183222, "grad_norm": 1.630176381086247, "learning_rate": 5.955873702924266e-06, "loss": 0.5668, "step": 2625 }, { "epoch": 1.9323031640912436, "grad_norm": 1.551831959372265, "learning_rate": 5.948608345784201e-06, "loss": 0.5548, "step": 2626 }, { "epoch": 1.9330389992641648, "grad_norm": 1.6540768743110155, "learning_rate": 5.941345546232348e-06, "loss": 0.5551, "step": 2627 }, { "epoch": 1.9337748344370862, "grad_norm": 1.7060219196509332, "learning_rate": 5.9340853088536145e-06, "loss": 0.5556, "step": 2628 }, { "epoch": 1.9345106696100074, "grad_norm": 1.7334025542671034, "learning_rate": 5.926827638231289e-06, "loss": 0.5765, "step": 2629 }, { "epoch": 1.9352465047829286, "grad_norm": 1.7466160665369697, "learning_rate": 5.919572538947036e-06, "loss": 0.5676, "step": 2630 }, { "epoch": 1.9359823399558498, "grad_norm": 1.6673975314878668, "learning_rate": 5.912320015580907e-06, "loss": 0.5437, "step": 2631 }, { "epoch": 1.9367181751287712, "grad_norm": 1.587606250150098, "learning_rate": 5.905070072711318e-06, "loss": 0.5349, "step": 2632 }, { "epoch": 1.9374540103016924, "grad_norm": 1.5620723057509551, "learning_rate": 5.8978227149150545e-06, "loss": 0.5469, "step": 2633 }, { "epoch": 1.9381898454746138, "grad_norm": 1.6800692758266484, "learning_rate": 5.890577946767284e-06, "loss": 0.55, "step": 2634 }, { "epoch": 1.938925680647535, "grad_norm": 1.6798500957910225, "learning_rate": 5.883335772841523e-06, "loss": 0.5707, "step": 2635 }, { "epoch": 1.9396615158204562, "grad_norm": 1.5496690059991356, "learning_rate": 5.876096197709658e-06, "loss": 0.5411, "step": 2636 }, { "epoch": 1.9403973509933774, "grad_norm": 1.6515308153200625, "learning_rate": 5.868859225941934e-06, "loss": 0.5822, "step": 2637 }, { "epoch": 1.9411331861662986, "grad_norm": 1.726630993426133, "learning_rate": 5.8616248621069545e-06, "loss": 0.6071, "step": 2638 }, { "epoch": 1.94186902133922, "grad_norm": 1.6607874018143791, "learning_rate": 5.854393110771668e-06, "loss": 0.5627, "step": 2639 }, { "epoch": 1.9426048565121414, "grad_norm": 1.5762980808230012, "learning_rate": 5.847163976501391e-06, "loss": 0.5646, "step": 2640 }, { "epoch": 1.9433406916850626, "grad_norm": 1.5685222471121278, "learning_rate": 5.83993746385977e-06, "loss": 0.5095, "step": 2641 }, { "epoch": 1.9440765268579838, "grad_norm": 1.6250536970002132, "learning_rate": 5.8327135774088e-06, "loss": 0.5255, "step": 2642 }, { "epoch": 1.944812362030905, "grad_norm": 1.681788991135852, "learning_rate": 5.825492321708834e-06, "loss": 0.5487, "step": 2643 }, { "epoch": 1.9455481972038262, "grad_norm": 1.705787089937453, "learning_rate": 5.818273701318542e-06, "loss": 0.5426, "step": 2644 }, { "epoch": 1.9462840323767476, "grad_norm": 1.7465452329996107, "learning_rate": 5.811057720794941e-06, "loss": 0.5164, "step": 2645 }, { "epoch": 1.9470198675496688, "grad_norm": 1.6519703057100152, "learning_rate": 5.803844384693386e-06, "loss": 0.5595, "step": 2646 }, { "epoch": 1.9477557027225902, "grad_norm": 1.6366980049269149, "learning_rate": 5.796633697567557e-06, "loss": 0.5365, "step": 2647 }, { "epoch": 1.9484915378955114, "grad_norm": 1.5885974009106256, "learning_rate": 5.789425663969458e-06, "loss": 0.5621, "step": 2648 }, { "epoch": 1.9492273730684326, "grad_norm": 1.614670678693086, "learning_rate": 5.782220288449422e-06, "loss": 0.5389, "step": 2649 }, { "epoch": 1.9499632082413538, "grad_norm": 1.6545119627747331, "learning_rate": 5.77501757555611e-06, "loss": 0.5704, "step": 2650 }, { "epoch": 1.9506990434142752, "grad_norm": 1.5894104512483067, "learning_rate": 5.76781752983649e-06, "loss": 0.5011, "step": 2651 }, { "epoch": 1.9514348785871964, "grad_norm": 1.4921377080256133, "learning_rate": 5.760620155835861e-06, "loss": 0.5108, "step": 2652 }, { "epoch": 1.9521707137601179, "grad_norm": 1.6613856013654442, "learning_rate": 5.753425458097817e-06, "loss": 0.5156, "step": 2653 }, { "epoch": 1.952906548933039, "grad_norm": 1.8017650930428608, "learning_rate": 5.746233441164275e-06, "loss": 0.4953, "step": 2654 }, { "epoch": 1.9536423841059603, "grad_norm": 1.516173967971116, "learning_rate": 5.73904410957546e-06, "loss": 0.5061, "step": 2655 }, { "epoch": 1.9543782192788814, "grad_norm": 1.7115581170330667, "learning_rate": 5.731857467869902e-06, "loss": 0.5621, "step": 2656 }, { "epoch": 1.9551140544518026, "grad_norm": 1.6073023157565043, "learning_rate": 5.724673520584421e-06, "loss": 0.5349, "step": 2657 }, { "epoch": 1.955849889624724, "grad_norm": 1.7032915815022085, "learning_rate": 5.717492272254148e-06, "loss": 0.587, "step": 2658 }, { "epoch": 1.9565857247976455, "grad_norm": 1.6699912916149007, "learning_rate": 5.710313727412513e-06, "loss": 0.5208, "step": 2659 }, { "epoch": 1.9573215599705667, "grad_norm": 1.6306379131913247, "learning_rate": 5.703137890591222e-06, "loss": 0.5163, "step": 2660 }, { "epoch": 1.9580573951434879, "grad_norm": 1.6508029940053606, "learning_rate": 5.695964766320294e-06, "loss": 0.5362, "step": 2661 }, { "epoch": 1.958793230316409, "grad_norm": 1.5469372459082686, "learning_rate": 5.688794359128018e-06, "loss": 0.5253, "step": 2662 }, { "epoch": 1.9595290654893303, "grad_norm": 1.6518939809539501, "learning_rate": 5.681626673540974e-06, "loss": 0.5442, "step": 2663 }, { "epoch": 1.9602649006622517, "grad_norm": 1.7344625791667536, "learning_rate": 5.674461714084026e-06, "loss": 0.5273, "step": 2664 }, { "epoch": 1.9610007358351729, "grad_norm": 1.683927728553634, "learning_rate": 5.6672994852803184e-06, "loss": 0.5132, "step": 2665 }, { "epoch": 1.9617365710080943, "grad_norm": 1.6710182019961588, "learning_rate": 5.660139991651261e-06, "loss": 0.519, "step": 2666 }, { "epoch": 1.9624724061810155, "grad_norm": 2.60331211113603, "learning_rate": 5.652983237716549e-06, "loss": 0.5699, "step": 2667 }, { "epoch": 1.9632082413539367, "grad_norm": 1.5905411047904245, "learning_rate": 5.645829227994146e-06, "loss": 0.521, "step": 2668 }, { "epoch": 1.9639440765268579, "grad_norm": 1.6526535015849972, "learning_rate": 5.638677967000268e-06, "loss": 0.5528, "step": 2669 }, { "epoch": 1.9646799116997793, "grad_norm": 1.786945165726115, "learning_rate": 5.631529459249426e-06, "loss": 0.5669, "step": 2670 }, { "epoch": 1.9654157468727005, "grad_norm": 1.6909605209474077, "learning_rate": 5.624383709254363e-06, "loss": 0.5318, "step": 2671 }, { "epoch": 1.966151582045622, "grad_norm": 1.70866158348278, "learning_rate": 5.617240721526096e-06, "loss": 0.5111, "step": 2672 }, { "epoch": 1.966887417218543, "grad_norm": 1.517037703618304, "learning_rate": 5.610100500573898e-06, "loss": 0.5082, "step": 2673 }, { "epoch": 1.9676232523914643, "grad_norm": 1.590975644442153, "learning_rate": 5.602963050905296e-06, "loss": 0.5708, "step": 2674 }, { "epoch": 1.9683590875643855, "grad_norm": 1.6848583296141102, "learning_rate": 5.5958283770260545e-06, "loss": 0.5135, "step": 2675 }, { "epoch": 1.9690949227373067, "grad_norm": 1.64594285383439, "learning_rate": 5.588696483440203e-06, "loss": 0.5264, "step": 2676 }, { "epoch": 1.969830757910228, "grad_norm": 1.6118736979344819, "learning_rate": 5.58156737465001e-06, "loss": 0.5473, "step": 2677 }, { "epoch": 1.9705665930831495, "grad_norm": 1.6094214977055437, "learning_rate": 5.5744410551559725e-06, "loss": 0.571, "step": 2678 }, { "epoch": 1.9713024282560707, "grad_norm": 1.7149360842409607, "learning_rate": 5.567317529456856e-06, "loss": 0.5309, "step": 2679 }, { "epoch": 1.972038263428992, "grad_norm": 1.6292807757127579, "learning_rate": 5.560196802049633e-06, "loss": 0.5529, "step": 2680 }, { "epoch": 1.972774098601913, "grad_norm": 1.6441489976844779, "learning_rate": 5.553078877429524e-06, "loss": 0.5591, "step": 2681 }, { "epoch": 1.9735099337748343, "grad_norm": 1.7615867348401815, "learning_rate": 5.545963760089978e-06, "loss": 0.5444, "step": 2682 }, { "epoch": 1.9742457689477557, "grad_norm": 1.6904624498567697, "learning_rate": 5.538851454522678e-06, "loss": 0.5442, "step": 2683 }, { "epoch": 1.974981604120677, "grad_norm": 1.6325830075672203, "learning_rate": 5.531741965217514e-06, "loss": 0.585, "step": 2684 }, { "epoch": 1.9757174392935983, "grad_norm": 1.5721824447901875, "learning_rate": 5.524635296662615e-06, "loss": 0.539, "step": 2685 }, { "epoch": 1.9764532744665195, "grad_norm": 1.670977546142154, "learning_rate": 5.517531453344327e-06, "loss": 0.5625, "step": 2686 }, { "epoch": 1.9771891096394407, "grad_norm": 1.7351410081250698, "learning_rate": 5.510430439747197e-06, "loss": 0.5632, "step": 2687 }, { "epoch": 1.977924944812362, "grad_norm": 1.685821692080219, "learning_rate": 5.503332260354017e-06, "loss": 0.5711, "step": 2688 }, { "epoch": 1.9786607799852833, "grad_norm": 1.6106661916226772, "learning_rate": 5.496236919645754e-06, "loss": 0.5469, "step": 2689 }, { "epoch": 1.9793966151582045, "grad_norm": 1.4689792225324234, "learning_rate": 5.489144422101607e-06, "loss": 0.524, "step": 2690 }, { "epoch": 1.980132450331126, "grad_norm": 1.5768532983576036, "learning_rate": 5.482054772198972e-06, "loss": 0.5237, "step": 2691 }, { "epoch": 1.9808682855040471, "grad_norm": 1.5937938746513476, "learning_rate": 5.474967974413451e-06, "loss": 0.51, "step": 2692 }, { "epoch": 1.9816041206769683, "grad_norm": 1.633906150055981, "learning_rate": 5.467884033218837e-06, "loss": 0.5435, "step": 2693 }, { "epoch": 1.9823399558498895, "grad_norm": 1.5627463336737129, "learning_rate": 5.460802953087126e-06, "loss": 0.5458, "step": 2694 }, { "epoch": 1.9830757910228107, "grad_norm": 1.57398646304602, "learning_rate": 5.453724738488511e-06, "loss": 0.542, "step": 2695 }, { "epoch": 1.9838116261957321, "grad_norm": 1.7467419681788743, "learning_rate": 5.4466493938913705e-06, "loss": 0.5397, "step": 2696 }, { "epoch": 1.9845474613686536, "grad_norm": 1.6084674923050672, "learning_rate": 5.439576923762275e-06, "loss": 0.5563, "step": 2697 }, { "epoch": 1.9852832965415748, "grad_norm": 1.5634811912292594, "learning_rate": 5.432507332565968e-06, "loss": 0.5472, "step": 2698 }, { "epoch": 1.986019131714496, "grad_norm": 1.5594834776134685, "learning_rate": 5.425440624765403e-06, "loss": 0.5522, "step": 2699 }, { "epoch": 1.9867549668874172, "grad_norm": 1.673017701707529, "learning_rate": 5.418376804821682e-06, "loss": 0.5391, "step": 2700 }, { "epoch": 1.9874908020603383, "grad_norm": 1.5651970939384492, "learning_rate": 5.411315877194104e-06, "loss": 0.5177, "step": 2701 }, { "epoch": 1.9882266372332598, "grad_norm": 1.5405642710953777, "learning_rate": 5.404257846340138e-06, "loss": 0.4942, "step": 2702 }, { "epoch": 1.9889624724061812, "grad_norm": 1.6317828592493615, "learning_rate": 5.397202716715417e-06, "loss": 0.5393, "step": 2703 }, { "epoch": 1.9896983075791024, "grad_norm": 1.599426194668401, "learning_rate": 5.390150492773749e-06, "loss": 0.5214, "step": 2704 }, { "epoch": 1.9904341427520236, "grad_norm": 1.5325230052827423, "learning_rate": 5.383101178967109e-06, "loss": 0.5098, "step": 2705 }, { "epoch": 1.9911699779249448, "grad_norm": 1.6991370969409885, "learning_rate": 5.376054779745637e-06, "loss": 0.5162, "step": 2706 }, { "epoch": 1.991905813097866, "grad_norm": 1.7540800488807364, "learning_rate": 5.369011299557617e-06, "loss": 0.552, "step": 2707 }, { "epoch": 1.9926416482707874, "grad_norm": 1.6410147735727059, "learning_rate": 5.361970742849509e-06, "loss": 0.5327, "step": 2708 }, { "epoch": 1.9933774834437086, "grad_norm": 1.5981214491565192, "learning_rate": 5.35493311406592e-06, "loss": 0.5142, "step": 2709 }, { "epoch": 1.99411331861663, "grad_norm": 1.5924759677452518, "learning_rate": 5.347898417649609e-06, "loss": 0.5316, "step": 2710 }, { "epoch": 1.9948491537895512, "grad_norm": 1.5474903407127287, "learning_rate": 5.340866658041486e-06, "loss": 0.5267, "step": 2711 }, { "epoch": 1.9955849889624724, "grad_norm": 1.6541290105824582, "learning_rate": 5.333837839680598e-06, "loss": 0.5243, "step": 2712 }, { "epoch": 1.9963208241353936, "grad_norm": 1.643213009520828, "learning_rate": 5.3268119670041465e-06, "loss": 0.5443, "step": 2713 }, { "epoch": 1.9970566593083148, "grad_norm": 1.6008171708850065, "learning_rate": 5.319789044447468e-06, "loss": 0.5364, "step": 2714 }, { "epoch": 1.9977924944812362, "grad_norm": 1.7365861596097103, "learning_rate": 5.312769076444041e-06, "loss": 0.5678, "step": 2715 }, { "epoch": 1.9985283296541576, "grad_norm": 1.5730084449277337, "learning_rate": 5.305752067425469e-06, "loss": 0.5112, "step": 2716 }, { "epoch": 1.9992641648270788, "grad_norm": 1.6005773436268447, "learning_rate": 5.298738021821498e-06, "loss": 0.5629, "step": 2717 }, { "epoch": 2.0, "grad_norm": 1.7511316630968148, "learning_rate": 5.291726944059999e-06, "loss": 0.5467, "step": 2718 }, { "epoch": 2.000735835172921, "grad_norm": 2.0126826891506493, "learning_rate": 5.284718838566968e-06, "loss": 0.3027, "step": 2719 }, { "epoch": 2.0014716703458424, "grad_norm": 1.94220863291904, "learning_rate": 5.277713709766532e-06, "loss": 0.2917, "step": 2720 }, { "epoch": 2.0022075055187636, "grad_norm": 1.9453165309121063, "learning_rate": 5.270711562080923e-06, "loss": 0.2743, "step": 2721 }, { "epoch": 2.0029433406916852, "grad_norm": 1.7107764648621309, "learning_rate": 5.26371239993051e-06, "loss": 0.2507, "step": 2722 }, { "epoch": 2.0036791758646064, "grad_norm": 1.7292708944113584, "learning_rate": 5.256716227733762e-06, "loss": 0.2759, "step": 2723 }, { "epoch": 2.0044150110375276, "grad_norm": 1.6172037774650876, "learning_rate": 5.249723049907276e-06, "loss": 0.2763, "step": 2724 }, { "epoch": 2.005150846210449, "grad_norm": 1.6218922096022093, "learning_rate": 5.242732870865739e-06, "loss": 0.2517, "step": 2725 }, { "epoch": 2.00588668138337, "grad_norm": 1.8430953953107103, "learning_rate": 5.235745695021961e-06, "loss": 0.2539, "step": 2726 }, { "epoch": 2.006622516556291, "grad_norm": 2.019607883560947, "learning_rate": 5.228761526786853e-06, "loss": 0.2811, "step": 2727 }, { "epoch": 2.007358351729213, "grad_norm": 2.0911294800961877, "learning_rate": 5.221780370569415e-06, "loss": 0.2736, "step": 2728 }, { "epoch": 2.008094186902134, "grad_norm": 1.935454424340637, "learning_rate": 5.21480223077677e-06, "loss": 0.2624, "step": 2729 }, { "epoch": 2.0088300220750552, "grad_norm": 1.8372238594059738, "learning_rate": 5.2078271118141114e-06, "loss": 0.2469, "step": 2730 }, { "epoch": 2.0095658572479764, "grad_norm": 1.7631885279367054, "learning_rate": 5.2008550180847394e-06, "loss": 0.2752, "step": 2731 }, { "epoch": 2.0103016924208976, "grad_norm": 1.7571225166448643, "learning_rate": 5.193885953990043e-06, "loss": 0.2896, "step": 2732 }, { "epoch": 2.011037527593819, "grad_norm": 1.850762052562353, "learning_rate": 5.186919923929499e-06, "loss": 0.2419, "step": 2733 }, { "epoch": 2.0117733627667405, "grad_norm": 1.7138379581904337, "learning_rate": 5.1799569323006615e-06, "loss": 0.2598, "step": 2734 }, { "epoch": 2.0125091979396617, "grad_norm": 1.7406064681912203, "learning_rate": 5.172996983499174e-06, "loss": 0.252, "step": 2735 }, { "epoch": 2.013245033112583, "grad_norm": 1.773451509885734, "learning_rate": 5.166040081918764e-06, "loss": 0.2878, "step": 2736 }, { "epoch": 2.013980868285504, "grad_norm": 1.812368157826744, "learning_rate": 5.159086231951213e-06, "loss": 0.26, "step": 2737 }, { "epoch": 2.0147167034584252, "grad_norm": 1.734052879337422, "learning_rate": 5.152135437986409e-06, "loss": 0.2711, "step": 2738 }, { "epoch": 2.0154525386313464, "grad_norm": 1.765354667818382, "learning_rate": 5.14518770441228e-06, "loss": 0.2722, "step": 2739 }, { "epoch": 2.0161883738042676, "grad_norm": 1.6798781472592301, "learning_rate": 5.138243035614842e-06, "loss": 0.2551, "step": 2740 }, { "epoch": 2.0169242089771893, "grad_norm": 1.7243845629666414, "learning_rate": 5.131301435978169e-06, "loss": 0.2842, "step": 2741 }, { "epoch": 2.0176600441501105, "grad_norm": 1.7013171487791972, "learning_rate": 5.124362909884399e-06, "loss": 0.2587, "step": 2742 }, { "epoch": 2.0183958793230317, "grad_norm": 1.66308764413022, "learning_rate": 5.117427461713724e-06, "loss": 0.2962, "step": 2743 }, { "epoch": 2.019131714495953, "grad_norm": 1.6818823780115038, "learning_rate": 5.110495095844401e-06, "loss": 0.2635, "step": 2744 }, { "epoch": 2.019867549668874, "grad_norm": 1.686108029412894, "learning_rate": 5.10356581665274e-06, "loss": 0.269, "step": 2745 }, { "epoch": 2.0206033848417952, "grad_norm": 1.6025034015950004, "learning_rate": 5.096639628513092e-06, "loss": 0.2747, "step": 2746 }, { "epoch": 2.021339220014717, "grad_norm": 1.7419883153601194, "learning_rate": 5.089716535797877e-06, "loss": 0.2662, "step": 2747 }, { "epoch": 2.022075055187638, "grad_norm": 1.7358726267423308, "learning_rate": 5.082796542877539e-06, "loss": 0.2791, "step": 2748 }, { "epoch": 2.0228108903605593, "grad_norm": 1.7651735244373148, "learning_rate": 5.0758796541205794e-06, "loss": 0.26, "step": 2749 }, { "epoch": 2.0235467255334805, "grad_norm": 1.6865433198950779, "learning_rate": 5.068965873893534e-06, "loss": 0.2609, "step": 2750 }, { "epoch": 2.0242825607064017, "grad_norm": 1.7467647245185198, "learning_rate": 5.062055206560982e-06, "loss": 0.2837, "step": 2751 }, { "epoch": 2.025018395879323, "grad_norm": 1.624724380396991, "learning_rate": 5.055147656485526e-06, "loss": 0.2513, "step": 2752 }, { "epoch": 2.0257542310522445, "grad_norm": 1.5466840411333151, "learning_rate": 5.048243228027812e-06, "loss": 0.2523, "step": 2753 }, { "epoch": 2.0264900662251657, "grad_norm": 1.7243918791684267, "learning_rate": 5.0413419255465165e-06, "loss": 0.231, "step": 2754 }, { "epoch": 2.027225901398087, "grad_norm": 1.6634245354119095, "learning_rate": 5.034443753398323e-06, "loss": 0.2755, "step": 2755 }, { "epoch": 2.027961736571008, "grad_norm": 1.6721754396099204, "learning_rate": 5.0275487159379725e-06, "loss": 0.2616, "step": 2756 }, { "epoch": 2.0286975717439293, "grad_norm": 1.6368898111897827, "learning_rate": 5.020656817518192e-06, "loss": 0.2833, "step": 2757 }, { "epoch": 2.0294334069168505, "grad_norm": 1.6163726469706907, "learning_rate": 5.01376806248975e-06, "loss": 0.2777, "step": 2758 }, { "epoch": 2.0301692420897717, "grad_norm": 1.7232612264424614, "learning_rate": 5.006882455201423e-06, "loss": 0.2769, "step": 2759 }, { "epoch": 2.0309050772626933, "grad_norm": 1.6072354381177674, "learning_rate": 5.000000000000003e-06, "loss": 0.2577, "step": 2760 }, { "epoch": 2.0316409124356145, "grad_norm": 1.7435134223416295, "learning_rate": 4.993120701230283e-06, "loss": 0.2679, "step": 2761 }, { "epoch": 2.0323767476085357, "grad_norm": 1.7076428974719375, "learning_rate": 4.986244563235075e-06, "loss": 0.2728, "step": 2762 }, { "epoch": 2.033112582781457, "grad_norm": 1.7188782342892146, "learning_rate": 4.9793715903551946e-06, "loss": 0.2602, "step": 2763 }, { "epoch": 2.033848417954378, "grad_norm": 1.715196083727058, "learning_rate": 4.972501786929443e-06, "loss": 0.2804, "step": 2764 }, { "epoch": 2.0345842531272993, "grad_norm": 1.7128258928273374, "learning_rate": 4.965635157294651e-06, "loss": 0.248, "step": 2765 }, { "epoch": 2.035320088300221, "grad_norm": 1.5012796119688065, "learning_rate": 4.9587717057856155e-06, "loss": 0.2269, "step": 2766 }, { "epoch": 2.036055923473142, "grad_norm": 1.7757963012915579, "learning_rate": 4.951911436735142e-06, "loss": 0.2696, "step": 2767 }, { "epoch": 2.0367917586460633, "grad_norm": 1.691899572917995, "learning_rate": 4.945054354474026e-06, "loss": 0.2844, "step": 2768 }, { "epoch": 2.0375275938189845, "grad_norm": 1.6333645639495764, "learning_rate": 4.938200463331054e-06, "loss": 0.2778, "step": 2769 }, { "epoch": 2.0382634289919057, "grad_norm": 1.7754779539390873, "learning_rate": 4.931349767632985e-06, "loss": 0.2921, "step": 2770 }, { "epoch": 2.038999264164827, "grad_norm": 1.6733950660290315, "learning_rate": 4.924502271704574e-06, "loss": 0.2417, "step": 2771 }, { "epoch": 2.0397350993377485, "grad_norm": 1.6431476978588555, "learning_rate": 4.9176579798685566e-06, "loss": 0.2692, "step": 2772 }, { "epoch": 2.0404709345106697, "grad_norm": 1.7735522275799547, "learning_rate": 4.910816896445628e-06, "loss": 0.2499, "step": 2773 }, { "epoch": 2.041206769683591, "grad_norm": 1.6751955223816286, "learning_rate": 4.9039790257544866e-06, "loss": 0.2809, "step": 2774 }, { "epoch": 2.041942604856512, "grad_norm": 1.6364095357634862, "learning_rate": 4.897144372111774e-06, "loss": 0.2546, "step": 2775 }, { "epoch": 2.0426784400294333, "grad_norm": 1.579037329869153, "learning_rate": 4.890312939832119e-06, "loss": 0.2453, "step": 2776 }, { "epoch": 2.0434142752023545, "grad_norm": 1.5871335885946565, "learning_rate": 4.883484733228112e-06, "loss": 0.2298, "step": 2777 }, { "epoch": 2.044150110375276, "grad_norm": 1.7680821459027196, "learning_rate": 4.876659756610309e-06, "loss": 0.2983, "step": 2778 }, { "epoch": 2.0448859455481974, "grad_norm": 1.765588844774007, "learning_rate": 4.869838014287217e-06, "loss": 0.2542, "step": 2779 }, { "epoch": 2.0456217807211186, "grad_norm": 1.6623558917131338, "learning_rate": 4.863019510565312e-06, "loss": 0.2668, "step": 2780 }, { "epoch": 2.0463576158940397, "grad_norm": 1.6980197098902883, "learning_rate": 4.856204249749028e-06, "loss": 0.2739, "step": 2781 }, { "epoch": 2.047093451066961, "grad_norm": 1.8176043438492988, "learning_rate": 4.849392236140734e-06, "loss": 0.2698, "step": 2782 }, { "epoch": 2.047829286239882, "grad_norm": 1.7548017171614259, "learning_rate": 4.842583474040775e-06, "loss": 0.2604, "step": 2783 }, { "epoch": 2.0485651214128033, "grad_norm": 1.714657681241402, "learning_rate": 4.835777967747419e-06, "loss": 0.2726, "step": 2784 }, { "epoch": 2.049300956585725, "grad_norm": 1.6272123706327266, "learning_rate": 4.828975721556895e-06, "loss": 0.2088, "step": 2785 }, { "epoch": 2.050036791758646, "grad_norm": 1.70879800844672, "learning_rate": 4.822176739763368e-06, "loss": 0.2618, "step": 2786 }, { "epoch": 2.0507726269315674, "grad_norm": 1.6227249867598725, "learning_rate": 4.815381026658937e-06, "loss": 0.2642, "step": 2787 }, { "epoch": 2.0515084621044886, "grad_norm": 1.5863545838857975, "learning_rate": 4.808588586533646e-06, "loss": 0.2424, "step": 2788 }, { "epoch": 2.0522442972774098, "grad_norm": 1.7078485366422074, "learning_rate": 4.80179942367547e-06, "loss": 0.2691, "step": 2789 }, { "epoch": 2.052980132450331, "grad_norm": 1.8812549861440595, "learning_rate": 4.7950135423703195e-06, "loss": 0.2892, "step": 2790 }, { "epoch": 2.0537159676232526, "grad_norm": 1.606701896933876, "learning_rate": 4.788230946902015e-06, "loss": 0.2767, "step": 2791 }, { "epoch": 2.054451802796174, "grad_norm": 1.7059092413623531, "learning_rate": 4.781451641552333e-06, "loss": 0.2405, "step": 2792 }, { "epoch": 2.055187637969095, "grad_norm": 1.668541989834053, "learning_rate": 4.774675630600947e-06, "loss": 0.2602, "step": 2793 }, { "epoch": 2.055923473142016, "grad_norm": 1.6588262473973225, "learning_rate": 4.76790291832546e-06, "loss": 0.262, "step": 2794 }, { "epoch": 2.0566593083149374, "grad_norm": 1.6095841457711229, "learning_rate": 4.761133509001399e-06, "loss": 0.2252, "step": 2795 }, { "epoch": 2.0573951434878586, "grad_norm": 1.663527918349439, "learning_rate": 4.75436740690219e-06, "loss": 0.2471, "step": 2796 }, { "epoch": 2.05813097866078, "grad_norm": 1.736161164610128, "learning_rate": 4.747604616299189e-06, "loss": 0.2836, "step": 2797 }, { "epoch": 2.0588668138337014, "grad_norm": 1.7848392107895235, "learning_rate": 4.740845141461649e-06, "loss": 0.2854, "step": 2798 }, { "epoch": 2.0596026490066226, "grad_norm": 1.6798913717158641, "learning_rate": 4.734088986656738e-06, "loss": 0.2514, "step": 2799 }, { "epoch": 2.060338484179544, "grad_norm": 1.7965016484838376, "learning_rate": 4.727336156149516e-06, "loss": 0.2853, "step": 2800 }, { "epoch": 2.061074319352465, "grad_norm": 1.6327285114029553, "learning_rate": 4.720586654202965e-06, "loss": 0.2513, "step": 2801 }, { "epoch": 2.061810154525386, "grad_norm": 1.6114300536745494, "learning_rate": 4.713840485077943e-06, "loss": 0.2714, "step": 2802 }, { "epoch": 2.0625459896983074, "grad_norm": 1.791958308221568, "learning_rate": 4.707097653033219e-06, "loss": 0.2661, "step": 2803 }, { "epoch": 2.063281824871229, "grad_norm": 1.7423658650690017, "learning_rate": 4.700358162325453e-06, "loss": 0.2495, "step": 2804 }, { "epoch": 2.06401766004415, "grad_norm": 1.6655384237883362, "learning_rate": 4.6936220172091875e-06, "loss": 0.246, "step": 2805 }, { "epoch": 2.0647534952170714, "grad_norm": 1.6148803828653253, "learning_rate": 4.686889221936861e-06, "loss": 0.2182, "step": 2806 }, { "epoch": 2.0654893303899926, "grad_norm": 1.6547523601665712, "learning_rate": 4.680159780758797e-06, "loss": 0.2543, "step": 2807 }, { "epoch": 2.066225165562914, "grad_norm": 1.7101533607423574, "learning_rate": 4.673433697923202e-06, "loss": 0.244, "step": 2808 }, { "epoch": 2.066961000735835, "grad_norm": 1.6842760921005742, "learning_rate": 4.66671097767615e-06, "loss": 0.2638, "step": 2809 }, { "epoch": 2.0676968359087566, "grad_norm": 1.8390592608545313, "learning_rate": 4.659991624261616e-06, "loss": 0.2675, "step": 2810 }, { "epoch": 2.068432671081678, "grad_norm": 1.751092004664727, "learning_rate": 4.653275641921428e-06, "loss": 0.2681, "step": 2811 }, { "epoch": 2.069168506254599, "grad_norm": 1.7650448868143285, "learning_rate": 4.646563034895293e-06, "loss": 0.2577, "step": 2812 }, { "epoch": 2.0699043414275202, "grad_norm": 1.789277628967715, "learning_rate": 4.639853807420796e-06, "loss": 0.2475, "step": 2813 }, { "epoch": 2.0706401766004414, "grad_norm": 1.6565294775695083, "learning_rate": 4.633147963733371e-06, "loss": 0.2514, "step": 2814 }, { "epoch": 2.0713760117733626, "grad_norm": 1.7306793234901154, "learning_rate": 4.626445508066329e-06, "loss": 0.2692, "step": 2815 }, { "epoch": 2.0721118469462843, "grad_norm": 1.5990991114197384, "learning_rate": 4.61974644465084e-06, "loss": 0.2504, "step": 2816 }, { "epoch": 2.0728476821192054, "grad_norm": 1.6803251153482832, "learning_rate": 4.613050777715935e-06, "loss": 0.2266, "step": 2817 }, { "epoch": 2.0735835172921266, "grad_norm": 1.72504843033123, "learning_rate": 4.606358511488486e-06, "loss": 0.2876, "step": 2818 }, { "epoch": 2.074319352465048, "grad_norm": 1.7182090463153759, "learning_rate": 4.599669650193241e-06, "loss": 0.2385, "step": 2819 }, { "epoch": 2.075055187637969, "grad_norm": 1.7480485867064028, "learning_rate": 4.59298419805278e-06, "loss": 0.2416, "step": 2820 }, { "epoch": 2.0757910228108902, "grad_norm": 1.6123153714897556, "learning_rate": 4.58630215928754e-06, "loss": 0.2469, "step": 2821 }, { "epoch": 2.0765268579838114, "grad_norm": 1.6672871190268874, "learning_rate": 4.579623538115803e-06, "loss": 0.2729, "step": 2822 }, { "epoch": 2.077262693156733, "grad_norm": 1.805403265879931, "learning_rate": 4.5729483387536845e-06, "loss": 0.2534, "step": 2823 }, { "epoch": 2.0779985283296543, "grad_norm": 1.7214565735333842, "learning_rate": 4.566276565415152e-06, "loss": 0.2717, "step": 2824 }, { "epoch": 2.0787343635025755, "grad_norm": 1.6110684679367981, "learning_rate": 4.559608222312002e-06, "loss": 0.2325, "step": 2825 }, { "epoch": 2.0794701986754967, "grad_norm": 1.7198476272139676, "learning_rate": 4.552943313653871e-06, "loss": 0.2448, "step": 2826 }, { "epoch": 2.080206033848418, "grad_norm": 1.7490142959401835, "learning_rate": 4.5462818436482245e-06, "loss": 0.2552, "step": 2827 }, { "epoch": 2.080941869021339, "grad_norm": 1.694967072782453, "learning_rate": 4.539623816500361e-06, "loss": 0.2416, "step": 2828 }, { "epoch": 2.0816777041942607, "grad_norm": 1.8303285916344003, "learning_rate": 4.532969236413391e-06, "loss": 0.256, "step": 2829 }, { "epoch": 2.082413539367182, "grad_norm": 1.7478790990599542, "learning_rate": 4.52631810758827e-06, "loss": 0.245, "step": 2830 }, { "epoch": 2.083149374540103, "grad_norm": 1.720993925063251, "learning_rate": 4.519670434223765e-06, "loss": 0.2681, "step": 2831 }, { "epoch": 2.0838852097130243, "grad_norm": 1.7996025571507865, "learning_rate": 4.51302622051645e-06, "loss": 0.2945, "step": 2832 }, { "epoch": 2.0846210448859455, "grad_norm": 1.9160351957967066, "learning_rate": 4.506385470660742e-06, "loss": 0.2741, "step": 2833 }, { "epoch": 2.0853568800588667, "grad_norm": 1.6665766716653703, "learning_rate": 4.499748188848846e-06, "loss": 0.2591, "step": 2834 }, { "epoch": 2.0860927152317883, "grad_norm": 1.6541150129833835, "learning_rate": 4.493114379270788e-06, "loss": 0.214, "step": 2835 }, { "epoch": 2.0868285504047095, "grad_norm": 1.5734256313492068, "learning_rate": 4.486484046114403e-06, "loss": 0.2446, "step": 2836 }, { "epoch": 2.0875643855776307, "grad_norm": 1.6752994721385326, "learning_rate": 4.4798571935653345e-06, "loss": 0.263, "step": 2837 }, { "epoch": 2.088300220750552, "grad_norm": 1.6594906505996463, "learning_rate": 4.473233825807014e-06, "loss": 0.2739, "step": 2838 }, { "epoch": 2.089036055923473, "grad_norm": 1.6285460120416235, "learning_rate": 4.466613947020689e-06, "loss": 0.2516, "step": 2839 }, { "epoch": 2.0897718910963943, "grad_norm": 1.6789012646907804, "learning_rate": 4.459997561385402e-06, "loss": 0.2508, "step": 2840 }, { "epoch": 2.0905077262693155, "grad_norm": 1.6555276489006987, "learning_rate": 4.453384673077975e-06, "loss": 0.2721, "step": 2841 }, { "epoch": 2.091243561442237, "grad_norm": 1.810748193674113, "learning_rate": 4.4467752862730485e-06, "loss": 0.289, "step": 2842 }, { "epoch": 2.0919793966151583, "grad_norm": 1.7229144327112533, "learning_rate": 4.4401694051430284e-06, "loss": 0.2544, "step": 2843 }, { "epoch": 2.0927152317880795, "grad_norm": 1.733518274251189, "learning_rate": 4.43356703385812e-06, "loss": 0.2576, "step": 2844 }, { "epoch": 2.0934510669610007, "grad_norm": 1.6411739656354074, "learning_rate": 4.42696817658631e-06, "loss": 0.2803, "step": 2845 }, { "epoch": 2.094186902133922, "grad_norm": 1.668509957335598, "learning_rate": 4.420372837493372e-06, "loss": 0.2174, "step": 2846 }, { "epoch": 2.094922737306843, "grad_norm": 1.7480906453624934, "learning_rate": 4.413781020742843e-06, "loss": 0.279, "step": 2847 }, { "epoch": 2.0956585724797647, "grad_norm": 1.787640078951225, "learning_rate": 4.4071927304960534e-06, "loss": 0.2589, "step": 2848 }, { "epoch": 2.096394407652686, "grad_norm": 1.9407000589315946, "learning_rate": 4.400607970912105e-06, "loss": 0.2698, "step": 2849 }, { "epoch": 2.097130242825607, "grad_norm": 1.7366696437119893, "learning_rate": 4.394026746147852e-06, "loss": 0.2641, "step": 2850 }, { "epoch": 2.0978660779985283, "grad_norm": 1.6861269615256846, "learning_rate": 4.38744906035795e-06, "loss": 0.2679, "step": 2851 }, { "epoch": 2.0986019131714495, "grad_norm": 1.8969805663586858, "learning_rate": 4.380874917694789e-06, "loss": 0.2565, "step": 2852 }, { "epoch": 2.0993377483443707, "grad_norm": 1.7194499085425645, "learning_rate": 4.374304322308541e-06, "loss": 0.2374, "step": 2853 }, { "epoch": 2.1000735835172923, "grad_norm": 1.7230559451408698, "learning_rate": 4.367737278347136e-06, "loss": 0.2677, "step": 2854 }, { "epoch": 2.1008094186902135, "grad_norm": 1.6658122762704137, "learning_rate": 4.36117378995625e-06, "loss": 0.2608, "step": 2855 }, { "epoch": 2.1015452538631347, "grad_norm": 1.7500353717197892, "learning_rate": 4.354613861279329e-06, "loss": 0.2527, "step": 2856 }, { "epoch": 2.102281089036056, "grad_norm": 1.7970296981722667, "learning_rate": 4.348057496457567e-06, "loss": 0.2659, "step": 2857 }, { "epoch": 2.103016924208977, "grad_norm": 1.6431609113101793, "learning_rate": 4.3415046996299105e-06, "loss": 0.2745, "step": 2858 }, { "epoch": 2.1037527593818983, "grad_norm": 1.7830923952105473, "learning_rate": 4.334955474933039e-06, "loss": 0.264, "step": 2859 }, { "epoch": 2.1044885945548195, "grad_norm": 1.699657072812776, "learning_rate": 4.328409826501403e-06, "loss": 0.2675, "step": 2860 }, { "epoch": 2.105224429727741, "grad_norm": 1.6535477884786667, "learning_rate": 4.321867758467172e-06, "loss": 0.2601, "step": 2861 }, { "epoch": 2.1059602649006623, "grad_norm": 1.744860413266754, "learning_rate": 4.315329274960266e-06, "loss": 0.2727, "step": 2862 }, { "epoch": 2.1066961000735835, "grad_norm": 1.7143379605041997, "learning_rate": 4.3087943801083445e-06, "loss": 0.2512, "step": 2863 }, { "epoch": 2.1074319352465047, "grad_norm": 1.6470898342027513, "learning_rate": 4.302263078036789e-06, "loss": 0.2421, "step": 2864 }, { "epoch": 2.108167770419426, "grad_norm": 1.6642090200743618, "learning_rate": 4.295735372868728e-06, "loss": 0.2303, "step": 2865 }, { "epoch": 2.108903605592347, "grad_norm": 1.8120947303513626, "learning_rate": 4.289211268725009e-06, "loss": 0.3068, "step": 2866 }, { "epoch": 2.1096394407652688, "grad_norm": 1.6932947812287773, "learning_rate": 4.282690769724216e-06, "loss": 0.2396, "step": 2867 }, { "epoch": 2.11037527593819, "grad_norm": 1.7270987455295186, "learning_rate": 4.276173879982639e-06, "loss": 0.2614, "step": 2868 }, { "epoch": 2.111111111111111, "grad_norm": 1.7739445891598036, "learning_rate": 4.269660603614316e-06, "loss": 0.2652, "step": 2869 }, { "epoch": 2.1118469462840324, "grad_norm": 1.6687970383618138, "learning_rate": 4.2631509447309785e-06, "loss": 0.2794, "step": 2870 }, { "epoch": 2.1125827814569536, "grad_norm": 1.6043069754636727, "learning_rate": 4.25664490744209e-06, "loss": 0.2338, "step": 2871 }, { "epoch": 2.1133186166298747, "grad_norm": 1.6392291257281946, "learning_rate": 4.250142495854825e-06, "loss": 0.2688, "step": 2872 }, { "epoch": 2.1140544518027964, "grad_norm": 1.6470997492495298, "learning_rate": 4.243643714074061e-06, "loss": 0.2497, "step": 2873 }, { "epoch": 2.1147902869757176, "grad_norm": 1.7946933648762862, "learning_rate": 4.237148566202392e-06, "loss": 0.2505, "step": 2874 }, { "epoch": 2.1155261221486388, "grad_norm": 1.7142882585306836, "learning_rate": 4.2306570563401185e-06, "loss": 0.27, "step": 2875 }, { "epoch": 2.11626195732156, "grad_norm": 1.6052584684390891, "learning_rate": 4.224169188585243e-06, "loss": 0.2492, "step": 2876 }, { "epoch": 2.116997792494481, "grad_norm": 1.8038118751107217, "learning_rate": 4.217684967033461e-06, "loss": 0.2592, "step": 2877 }, { "epoch": 2.1177336276674024, "grad_norm": 1.737454708095793, "learning_rate": 4.211204395778183e-06, "loss": 0.2493, "step": 2878 }, { "epoch": 2.1184694628403236, "grad_norm": 1.6325458298691962, "learning_rate": 4.204727478910497e-06, "loss": 0.2632, "step": 2879 }, { "epoch": 2.119205298013245, "grad_norm": 1.694918260961312, "learning_rate": 4.198254220519195e-06, "loss": 0.2516, "step": 2880 }, { "epoch": 2.1199411331861664, "grad_norm": 1.7241857431103218, "learning_rate": 4.19178462469076e-06, "loss": 0.2738, "step": 2881 }, { "epoch": 2.1206769683590876, "grad_norm": 1.7605287641539007, "learning_rate": 4.185318695509352e-06, "loss": 0.2698, "step": 2882 }, { "epoch": 2.121412803532009, "grad_norm": 1.6990155755041225, "learning_rate": 4.178856437056829e-06, "loss": 0.261, "step": 2883 }, { "epoch": 2.12214863870493, "grad_norm": 1.842881814045483, "learning_rate": 4.172397853412725e-06, "loss": 0.2323, "step": 2884 }, { "epoch": 2.122884473877851, "grad_norm": 1.6418650891941826, "learning_rate": 4.16594294865426e-06, "loss": 0.2444, "step": 2885 }, { "epoch": 2.123620309050773, "grad_norm": 1.7948519632659936, "learning_rate": 4.159491726856315e-06, "loss": 0.2653, "step": 2886 }, { "epoch": 2.124356144223694, "grad_norm": 1.6735194650491942, "learning_rate": 4.1530441920914746e-06, "loss": 0.2429, "step": 2887 }, { "epoch": 2.125091979396615, "grad_norm": 1.685182381775674, "learning_rate": 4.146600348429967e-06, "loss": 0.249, "step": 2888 }, { "epoch": 2.1258278145695364, "grad_norm": 1.8640194093258522, "learning_rate": 4.140160199939707e-06, "loss": 0.2768, "step": 2889 }, { "epoch": 2.1265636497424576, "grad_norm": 1.689221416230374, "learning_rate": 4.1337237506862744e-06, "loss": 0.2498, "step": 2890 }, { "epoch": 2.127299484915379, "grad_norm": 1.7126020652136995, "learning_rate": 4.127291004732908e-06, "loss": 0.248, "step": 2891 }, { "epoch": 2.1280353200883004, "grad_norm": 1.7249843237755398, "learning_rate": 4.1208619661405136e-06, "loss": 0.26, "step": 2892 }, { "epoch": 2.1287711552612216, "grad_norm": 1.5822993100723823, "learning_rate": 4.114436638967656e-06, "loss": 0.2621, "step": 2893 }, { "epoch": 2.129506990434143, "grad_norm": 1.6684348040357797, "learning_rate": 4.108015027270559e-06, "loss": 0.2659, "step": 2894 }, { "epoch": 2.130242825607064, "grad_norm": 1.6882321759227472, "learning_rate": 4.101597135103089e-06, "loss": 0.2527, "step": 2895 }, { "epoch": 2.130978660779985, "grad_norm": 1.6601082265747429, "learning_rate": 4.095182966516787e-06, "loss": 0.2614, "step": 2896 }, { "epoch": 2.1317144959529064, "grad_norm": 1.7716782564266036, "learning_rate": 4.088772525560821e-06, "loss": 0.2241, "step": 2897 }, { "epoch": 2.1324503311258276, "grad_norm": 1.7413750908433892, "learning_rate": 4.082365816282017e-06, "loss": 0.2841, "step": 2898 }, { "epoch": 2.1331861662987492, "grad_norm": 1.8396463919605357, "learning_rate": 4.075962842724847e-06, "loss": 0.2525, "step": 2899 }, { "epoch": 2.1339220014716704, "grad_norm": 1.7165167526735643, "learning_rate": 4.069563608931413e-06, "loss": 0.2462, "step": 2900 }, { "epoch": 2.1346578366445916, "grad_norm": 1.7000203886921554, "learning_rate": 4.063168118941467e-06, "loss": 0.2425, "step": 2901 }, { "epoch": 2.135393671817513, "grad_norm": 1.7445913893850693, "learning_rate": 4.0567763767923965e-06, "loss": 0.2761, "step": 2902 }, { "epoch": 2.136129506990434, "grad_norm": 1.6191259065311934, "learning_rate": 4.050388386519223e-06, "loss": 0.2702, "step": 2903 }, { "epoch": 2.136865342163355, "grad_norm": 1.7548650960712424, "learning_rate": 4.044004152154588e-06, "loss": 0.2795, "step": 2904 }, { "epoch": 2.137601177336277, "grad_norm": 1.679635217784513, "learning_rate": 4.037623677728783e-06, "loss": 0.2453, "step": 2905 }, { "epoch": 2.138337012509198, "grad_norm": 1.795932673139973, "learning_rate": 4.031246967269707e-06, "loss": 0.2848, "step": 2906 }, { "epoch": 2.1390728476821192, "grad_norm": 1.6608231016584178, "learning_rate": 4.024874024802893e-06, "loss": 0.2449, "step": 2907 }, { "epoch": 2.1398086828550404, "grad_norm": 1.7228457576363554, "learning_rate": 4.018504854351495e-06, "loss": 0.2641, "step": 2908 }, { "epoch": 2.1405445180279616, "grad_norm": 1.689657168489886, "learning_rate": 4.0121394599362765e-06, "loss": 0.279, "step": 2909 }, { "epoch": 2.141280353200883, "grad_norm": 1.620492291256972, "learning_rate": 4.005777845575628e-06, "loss": 0.24, "step": 2910 }, { "epoch": 2.1420161883738045, "grad_norm": 1.7017218564142662, "learning_rate": 3.999420015285549e-06, "loss": 0.2587, "step": 2911 }, { "epoch": 2.1427520235467257, "grad_norm": 1.5828005626341604, "learning_rate": 3.993065973079656e-06, "loss": 0.2443, "step": 2912 }, { "epoch": 2.143487858719647, "grad_norm": 1.7729966259345669, "learning_rate": 3.986715722969157e-06, "loss": 0.254, "step": 2913 }, { "epoch": 2.144223693892568, "grad_norm": 1.9393626636979586, "learning_rate": 3.980369268962893e-06, "loss": 0.2775, "step": 2914 }, { "epoch": 2.1449595290654893, "grad_norm": 1.693959885870409, "learning_rate": 3.974026615067282e-06, "loss": 0.2645, "step": 2915 }, { "epoch": 2.1456953642384105, "grad_norm": 1.6819533688499209, "learning_rate": 3.967687765286359e-06, "loss": 0.2619, "step": 2916 }, { "epoch": 2.1464311994113316, "grad_norm": 1.8773974307094918, "learning_rate": 3.961352723621757e-06, "loss": 0.2487, "step": 2917 }, { "epoch": 2.1471670345842533, "grad_norm": 1.8217857834106685, "learning_rate": 3.9550214940726946e-06, "loss": 0.2588, "step": 2918 }, { "epoch": 2.1479028697571745, "grad_norm": 1.6654653437376334, "learning_rate": 3.948694080635994e-06, "loss": 0.2548, "step": 2919 }, { "epoch": 2.1486387049300957, "grad_norm": 1.7128036720282651, "learning_rate": 3.942370487306064e-06, "loss": 0.2554, "step": 2920 }, { "epoch": 2.149374540103017, "grad_norm": 1.8740614033697067, "learning_rate": 3.936050718074908e-06, "loss": 0.2668, "step": 2921 }, { "epoch": 2.150110375275938, "grad_norm": 1.6291197330849263, "learning_rate": 3.9297347769321e-06, "loss": 0.2551, "step": 2922 }, { "epoch": 2.1508462104488593, "grad_norm": 1.794759133728299, "learning_rate": 3.923422667864814e-06, "loss": 0.2989, "step": 2923 }, { "epoch": 2.151582045621781, "grad_norm": 1.739551394382465, "learning_rate": 3.917114394857796e-06, "loss": 0.2499, "step": 2924 }, { "epoch": 2.152317880794702, "grad_norm": 1.7742742825584192, "learning_rate": 3.910809961893374e-06, "loss": 0.2688, "step": 2925 }, { "epoch": 2.1530537159676233, "grad_norm": 1.668879745561898, "learning_rate": 3.904509372951453e-06, "loss": 0.261, "step": 2926 }, { "epoch": 2.1537895511405445, "grad_norm": 1.6542487905391288, "learning_rate": 3.8982126320095015e-06, "loss": 0.2594, "step": 2927 }, { "epoch": 2.1545253863134657, "grad_norm": 1.7120251206605475, "learning_rate": 3.89191974304257e-06, "loss": 0.2737, "step": 2928 }, { "epoch": 2.155261221486387, "grad_norm": 1.6358982806793592, "learning_rate": 3.885630710023275e-06, "loss": 0.2386, "step": 2929 }, { "epoch": 2.1559970566593085, "grad_norm": 1.5775534398278963, "learning_rate": 3.879345536921798e-06, "loss": 0.2809, "step": 2930 }, { "epoch": 2.1567328918322297, "grad_norm": 1.7532982682164586, "learning_rate": 3.873064227705876e-06, "loss": 0.2667, "step": 2931 }, { "epoch": 2.157468727005151, "grad_norm": 1.7265387572489392, "learning_rate": 3.866786786340821e-06, "loss": 0.2405, "step": 2932 }, { "epoch": 2.158204562178072, "grad_norm": 1.6676945799858067, "learning_rate": 3.860513216789494e-06, "loss": 0.2684, "step": 2933 }, { "epoch": 2.1589403973509933, "grad_norm": 1.6694407603957206, "learning_rate": 3.854243523012307e-06, "loss": 0.2532, "step": 2934 }, { "epoch": 2.1596762325239145, "grad_norm": 1.7280148350690785, "learning_rate": 3.847977708967246e-06, "loss": 0.2894, "step": 2935 }, { "epoch": 2.1604120676968357, "grad_norm": 1.6026138714924427, "learning_rate": 3.841715778609822e-06, "loss": 0.2512, "step": 2936 }, { "epoch": 2.1611479028697573, "grad_norm": 1.6824913791307485, "learning_rate": 3.835457735893109e-06, "loss": 0.2655, "step": 2937 }, { "epoch": 2.1618837380426785, "grad_norm": 1.6705958814539175, "learning_rate": 3.829203584767724e-06, "loss": 0.2687, "step": 2938 }, { "epoch": 2.1626195732155997, "grad_norm": 1.7676878803362104, "learning_rate": 3.8229533291818335e-06, "loss": 0.2711, "step": 2939 }, { "epoch": 2.163355408388521, "grad_norm": 1.6999017611719713, "learning_rate": 3.816706973081129e-06, "loss": 0.2562, "step": 2940 }, { "epoch": 2.164091243561442, "grad_norm": 1.6225974538791643, "learning_rate": 3.810464520408853e-06, "loss": 0.2603, "step": 2941 }, { "epoch": 2.1648270787343633, "grad_norm": 1.679973292164748, "learning_rate": 3.804225975105786e-06, "loss": 0.2569, "step": 2942 }, { "epoch": 2.165562913907285, "grad_norm": 1.6996309039883775, "learning_rate": 3.797991341110224e-06, "loss": 0.2656, "step": 2943 }, { "epoch": 2.166298749080206, "grad_norm": 1.756416119058585, "learning_rate": 3.7917606223580217e-06, "loss": 0.237, "step": 2944 }, { "epoch": 2.1670345842531273, "grad_norm": 1.7043245802640696, "learning_rate": 3.7855338227825357e-06, "loss": 0.2522, "step": 2945 }, { "epoch": 2.1677704194260485, "grad_norm": 1.676285699980166, "learning_rate": 3.7793109463146627e-06, "loss": 0.2672, "step": 2946 }, { "epoch": 2.1685062545989697, "grad_norm": 1.7590269804614833, "learning_rate": 3.7730919968828194e-06, "loss": 0.2736, "step": 2947 }, { "epoch": 2.169242089771891, "grad_norm": 1.6940259130351618, "learning_rate": 3.766876978412949e-06, "loss": 0.2621, "step": 2948 }, { "epoch": 2.1699779249448126, "grad_norm": 1.8075268150123538, "learning_rate": 3.760665894828498e-06, "loss": 0.2414, "step": 2949 }, { "epoch": 2.1707137601177338, "grad_norm": 1.836574324785131, "learning_rate": 3.754458750050445e-06, "loss": 0.2756, "step": 2950 }, { "epoch": 2.171449595290655, "grad_norm": 1.6477436696617922, "learning_rate": 3.748255547997276e-06, "loss": 0.2626, "step": 2951 }, { "epoch": 2.172185430463576, "grad_norm": 1.6664203525762562, "learning_rate": 3.74205629258498e-06, "loss": 0.2399, "step": 2952 }, { "epoch": 2.1729212656364973, "grad_norm": 1.6909134293946713, "learning_rate": 3.7358609877270746e-06, "loss": 0.2513, "step": 2953 }, { "epoch": 2.1736571008094185, "grad_norm": 1.7054357885493687, "learning_rate": 3.729669637334562e-06, "loss": 0.2205, "step": 2954 }, { "epoch": 2.1743929359823397, "grad_norm": 1.7996885987201272, "learning_rate": 3.7234822453159593e-06, "loss": 0.2241, "step": 2955 }, { "epoch": 2.1751287711552614, "grad_norm": 1.6279008570704054, "learning_rate": 3.717298815577284e-06, "loss": 0.2656, "step": 2956 }, { "epoch": 2.1758646063281826, "grad_norm": 1.6807051671295137, "learning_rate": 3.7111193520220523e-06, "loss": 0.2671, "step": 2957 }, { "epoch": 2.1766004415011038, "grad_norm": 1.7742737673214088, "learning_rate": 3.7049438585512697e-06, "loss": 0.2555, "step": 2958 }, { "epoch": 2.177336276674025, "grad_norm": 1.9088557569641422, "learning_rate": 3.6987723390634447e-06, "loss": 0.276, "step": 2959 }, { "epoch": 2.178072111846946, "grad_norm": 1.6853722661976052, "learning_rate": 3.692604797454572e-06, "loss": 0.2636, "step": 2960 }, { "epoch": 2.1788079470198674, "grad_norm": 1.7355400670694985, "learning_rate": 3.686441237618137e-06, "loss": 0.2462, "step": 2961 }, { "epoch": 2.179543782192789, "grad_norm": 1.810949934051643, "learning_rate": 3.6802816634451144e-06, "loss": 0.2775, "step": 2962 }, { "epoch": 2.18027961736571, "grad_norm": 1.6339399563747659, "learning_rate": 3.6741260788239484e-06, "loss": 0.2658, "step": 2963 }, { "epoch": 2.1810154525386314, "grad_norm": 1.6072054894481669, "learning_rate": 3.66797448764059e-06, "loss": 0.2458, "step": 2964 }, { "epoch": 2.1817512877115526, "grad_norm": 1.6054716157510847, "learning_rate": 3.661826893778443e-06, "loss": 0.251, "step": 2965 }, { "epoch": 2.1824871228844738, "grad_norm": 1.6488444655977466, "learning_rate": 3.6556833011184045e-06, "loss": 0.2502, "step": 2966 }, { "epoch": 2.183222958057395, "grad_norm": 1.6197848546183027, "learning_rate": 3.6495437135388434e-06, "loss": 0.2414, "step": 2967 }, { "epoch": 2.1839587932303166, "grad_norm": 1.787408330241592, "learning_rate": 3.6434081349155903e-06, "loss": 0.2494, "step": 2968 }, { "epoch": 2.184694628403238, "grad_norm": 1.7133212887456388, "learning_rate": 3.637276569121957e-06, "loss": 0.2704, "step": 2969 }, { "epoch": 2.185430463576159, "grad_norm": 1.7088689038493476, "learning_rate": 3.6311490200287146e-06, "loss": 0.2502, "step": 2970 }, { "epoch": 2.18616629874908, "grad_norm": 1.856641480524577, "learning_rate": 3.6250254915041073e-06, "loss": 0.2831, "step": 2971 }, { "epoch": 2.1869021339220014, "grad_norm": 1.7448766587000537, "learning_rate": 3.6189059874138235e-06, "loss": 0.2561, "step": 2972 }, { "epoch": 2.1876379690949226, "grad_norm": 1.8807520130458797, "learning_rate": 3.612790511621036e-06, "loss": 0.2666, "step": 2973 }, { "epoch": 2.188373804267844, "grad_norm": 1.881066926446072, "learning_rate": 3.6066790679863505e-06, "loss": 0.2464, "step": 2974 }, { "epoch": 2.1891096394407654, "grad_norm": 1.6493010901611862, "learning_rate": 3.600571660367843e-06, "loss": 0.2527, "step": 2975 }, { "epoch": 2.1898454746136866, "grad_norm": 1.7194364354522513, "learning_rate": 3.5944682926210384e-06, "loss": 0.2397, "step": 2976 }, { "epoch": 2.190581309786608, "grad_norm": 1.796438081094231, "learning_rate": 3.588368968598903e-06, "loss": 0.2882, "step": 2977 }, { "epoch": 2.191317144959529, "grad_norm": 1.8256191286328383, "learning_rate": 3.582273692151861e-06, "loss": 0.2407, "step": 2978 }, { "epoch": 2.19205298013245, "grad_norm": 1.7658944054681898, "learning_rate": 3.5761824671277747e-06, "loss": 0.2702, "step": 2979 }, { "epoch": 2.1927888153053714, "grad_norm": 1.7703887142613088, "learning_rate": 3.5700952973719573e-06, "loss": 0.2513, "step": 2980 }, { "epoch": 2.193524650478293, "grad_norm": 1.6903630416310593, "learning_rate": 3.564012186727148e-06, "loss": 0.2512, "step": 2981 }, { "epoch": 2.1942604856512142, "grad_norm": 1.7673045842335058, "learning_rate": 3.557933139033536e-06, "loss": 0.26, "step": 2982 }, { "epoch": 2.1949963208241354, "grad_norm": 1.7340542794894447, "learning_rate": 3.551858158128739e-06, "loss": 0.2499, "step": 2983 }, { "epoch": 2.1957321559970566, "grad_norm": 1.866526465930817, "learning_rate": 3.5457872478478105e-06, "loss": 0.2616, "step": 2984 }, { "epoch": 2.196467991169978, "grad_norm": 1.6669653441960506, "learning_rate": 3.5397204120232376e-06, "loss": 0.2549, "step": 2985 }, { "epoch": 2.197203826342899, "grad_norm": 1.680901048524847, "learning_rate": 3.533657654484922e-06, "loss": 0.2766, "step": 2986 }, { "epoch": 2.1979396615158207, "grad_norm": 1.6726298109217725, "learning_rate": 3.527598979060206e-06, "loss": 0.2477, "step": 2987 }, { "epoch": 2.198675496688742, "grad_norm": 1.6685481221236083, "learning_rate": 3.5215443895738454e-06, "loss": 0.2458, "step": 2988 }, { "epoch": 2.199411331861663, "grad_norm": 1.6724493190763365, "learning_rate": 3.515493889848025e-06, "loss": 0.249, "step": 2989 }, { "epoch": 2.2001471670345842, "grad_norm": 1.7855810312439104, "learning_rate": 3.5094474837023362e-06, "loss": 0.2424, "step": 2990 }, { "epoch": 2.2008830022075054, "grad_norm": 1.7293283731430742, "learning_rate": 3.5034051749537956e-06, "loss": 0.2408, "step": 2991 }, { "epoch": 2.2016188373804266, "grad_norm": 1.7298592602837097, "learning_rate": 3.49736696741683e-06, "loss": 0.2538, "step": 2992 }, { "epoch": 2.202354672553348, "grad_norm": 1.7928564633277904, "learning_rate": 3.491332864903276e-06, "loss": 0.2664, "step": 2993 }, { "epoch": 2.2030905077262695, "grad_norm": 1.9377382679832291, "learning_rate": 3.485302871222386e-06, "loss": 0.2765, "step": 2994 }, { "epoch": 2.2038263428991907, "grad_norm": 1.706210921099338, "learning_rate": 3.4792769901808043e-06, "loss": 0.2447, "step": 2995 }, { "epoch": 2.204562178072112, "grad_norm": 1.858840214758188, "learning_rate": 3.4732552255825913e-06, "loss": 0.2579, "step": 2996 }, { "epoch": 2.205298013245033, "grad_norm": 1.7415768651340748, "learning_rate": 3.467237581229205e-06, "loss": 0.2785, "step": 2997 }, { "epoch": 2.2060338484179542, "grad_norm": 1.6980851898510514, "learning_rate": 3.4612240609195034e-06, "loss": 0.2323, "step": 2998 }, { "epoch": 2.2067696835908754, "grad_norm": 1.5994702517845278, "learning_rate": 3.4552146684497366e-06, "loss": 0.228, "step": 2999 }, { "epoch": 2.207505518763797, "grad_norm": 1.6656995963197228, "learning_rate": 3.449209407613552e-06, "loss": 0.2909, "step": 3000 }, { "epoch": 2.2082413539367183, "grad_norm": 1.607060393701296, "learning_rate": 3.443208282201994e-06, "loss": 0.2596, "step": 3001 }, { "epoch": 2.2089771891096395, "grad_norm": 1.7054406006766685, "learning_rate": 3.43721129600348e-06, "loss": 0.2552, "step": 3002 }, { "epoch": 2.2097130242825607, "grad_norm": 1.7428154026262583, "learning_rate": 3.4312184528038397e-06, "loss": 0.2559, "step": 3003 }, { "epoch": 2.210448859455482, "grad_norm": 1.678322036581353, "learning_rate": 3.4252297563862625e-06, "loss": 0.2331, "step": 3004 }, { "epoch": 2.211184694628403, "grad_norm": 1.7145112154259547, "learning_rate": 3.4192452105313346e-06, "loss": 0.2766, "step": 3005 }, { "epoch": 2.2119205298013247, "grad_norm": 1.661552181132433, "learning_rate": 3.413264819017017e-06, "loss": 0.2426, "step": 3006 }, { "epoch": 2.212656364974246, "grad_norm": 1.7750298631078716, "learning_rate": 3.407288585618654e-06, "loss": 0.2713, "step": 3007 }, { "epoch": 2.213392200147167, "grad_norm": 1.7689454153200712, "learning_rate": 3.401316514108952e-06, "loss": 0.2749, "step": 3008 }, { "epoch": 2.2141280353200883, "grad_norm": 1.6410770527072946, "learning_rate": 3.3953486082580025e-06, "loss": 0.2394, "step": 3009 }, { "epoch": 2.2148638704930095, "grad_norm": 1.648798856426521, "learning_rate": 3.3893848718332665e-06, "loss": 0.2413, "step": 3010 }, { "epoch": 2.2155997056659307, "grad_norm": 1.9035057905092567, "learning_rate": 3.3834253085995583e-06, "loss": 0.2894, "step": 3011 }, { "epoch": 2.216335540838852, "grad_norm": 1.6975379499444323, "learning_rate": 3.377469922319082e-06, "loss": 0.2458, "step": 3012 }, { "epoch": 2.2170713760117735, "grad_norm": 1.7049004894449327, "learning_rate": 3.371518716751383e-06, "loss": 0.2844, "step": 3013 }, { "epoch": 2.2178072111846947, "grad_norm": 1.7974727305772311, "learning_rate": 3.365571695653378e-06, "loss": 0.2565, "step": 3014 }, { "epoch": 2.218543046357616, "grad_norm": 1.6111941961789387, "learning_rate": 3.3596288627793417e-06, "loss": 0.2421, "step": 3015 }, { "epoch": 2.219278881530537, "grad_norm": 1.788207516079043, "learning_rate": 3.3536902218809043e-06, "loss": 0.2614, "step": 3016 }, { "epoch": 2.2200147167034583, "grad_norm": 1.622300218560251, "learning_rate": 3.3477557767070447e-06, "loss": 0.2461, "step": 3017 }, { "epoch": 2.2207505518763795, "grad_norm": 1.6817912334519671, "learning_rate": 3.341825531004099e-06, "loss": 0.2508, "step": 3018 }, { "epoch": 2.221486387049301, "grad_norm": 1.647004529751363, "learning_rate": 3.3358994885157537e-06, "loss": 0.2506, "step": 3019 }, { "epoch": 2.2222222222222223, "grad_norm": 1.7252379309068373, "learning_rate": 3.3299776529830285e-06, "loss": 0.2629, "step": 3020 }, { "epoch": 2.2229580573951435, "grad_norm": 1.660366854465252, "learning_rate": 3.32406002814431e-06, "loss": 0.2781, "step": 3021 }, { "epoch": 2.2236938925680647, "grad_norm": 1.642551379145196, "learning_rate": 3.318146617735306e-06, "loss": 0.2594, "step": 3022 }, { "epoch": 2.224429727740986, "grad_norm": 1.834930187277821, "learning_rate": 3.3122374254890734e-06, "loss": 0.2421, "step": 3023 }, { "epoch": 2.225165562913907, "grad_norm": 1.722369264444769, "learning_rate": 3.3063324551360044e-06, "loss": 0.2637, "step": 3024 }, { "epoch": 2.2259013980868287, "grad_norm": 1.6518785950458212, "learning_rate": 3.3004317104038296e-06, "loss": 0.2501, "step": 3025 }, { "epoch": 2.22663723325975, "grad_norm": 1.8344447219505913, "learning_rate": 3.294535195017603e-06, "loss": 0.2306, "step": 3026 }, { "epoch": 2.227373068432671, "grad_norm": 1.789780256916839, "learning_rate": 3.2886429126997165e-06, "loss": 0.2673, "step": 3027 }, { "epoch": 2.2281089036055923, "grad_norm": 1.8053740215536263, "learning_rate": 3.2827548671698907e-06, "loss": 0.2833, "step": 3028 }, { "epoch": 2.2288447387785135, "grad_norm": 1.6918145124508581, "learning_rate": 3.276871062145159e-06, "loss": 0.2242, "step": 3029 }, { "epoch": 2.2295805739514347, "grad_norm": 1.7177321857740202, "learning_rate": 3.2709915013398996e-06, "loss": 0.2173, "step": 3030 }, { "epoch": 2.230316409124356, "grad_norm": 1.8201902834394679, "learning_rate": 3.26511618846579e-06, "loss": 0.2807, "step": 3031 }, { "epoch": 2.2310522442972776, "grad_norm": 1.7296835224414993, "learning_rate": 3.259245127231838e-06, "loss": 0.2526, "step": 3032 }, { "epoch": 2.2317880794701987, "grad_norm": 1.5926534431699046, "learning_rate": 3.253378321344364e-06, "loss": 0.2409, "step": 3033 }, { "epoch": 2.23252391464312, "grad_norm": 1.6753481039714735, "learning_rate": 3.247515774507005e-06, "loss": 0.2375, "step": 3034 }, { "epoch": 2.233259749816041, "grad_norm": 1.8381991424724584, "learning_rate": 3.241657490420702e-06, "loss": 0.2589, "step": 3035 }, { "epoch": 2.2339955849889623, "grad_norm": 1.6574488511282315, "learning_rate": 3.2358034727837116e-06, "loss": 0.238, "step": 3036 }, { "epoch": 2.2347314201618835, "grad_norm": 1.6761051556967888, "learning_rate": 3.2299537252915993e-06, "loss": 0.2566, "step": 3037 }, { "epoch": 2.235467255334805, "grad_norm": 1.6126217939902072, "learning_rate": 3.22410825163722e-06, "loss": 0.2766, "step": 3038 }, { "epoch": 2.2362030905077264, "grad_norm": 1.6244631593420706, "learning_rate": 3.2182670555107543e-06, "loss": 0.2337, "step": 3039 }, { "epoch": 2.2369389256806476, "grad_norm": 1.81028995614845, "learning_rate": 3.2124301405996616e-06, "loss": 0.2389, "step": 3040 }, { "epoch": 2.2376747608535688, "grad_norm": 1.7077931849981272, "learning_rate": 3.2065975105887082e-06, "loss": 0.2323, "step": 3041 }, { "epoch": 2.23841059602649, "grad_norm": 1.7437563291959581, "learning_rate": 3.200769169159954e-06, "loss": 0.2559, "step": 3042 }, { "epoch": 2.239146431199411, "grad_norm": 1.589329767939776, "learning_rate": 3.194945119992755e-06, "loss": 0.2611, "step": 3043 }, { "epoch": 2.239882266372333, "grad_norm": 1.6893303906252255, "learning_rate": 3.1891253667637467e-06, "loss": 0.237, "step": 3044 }, { "epoch": 2.240618101545254, "grad_norm": 1.6667396965054875, "learning_rate": 3.183309913146865e-06, "loss": 0.2524, "step": 3045 }, { "epoch": 2.241353936718175, "grad_norm": 1.8150345209054146, "learning_rate": 3.177498762813327e-06, "loss": 0.2445, "step": 3046 }, { "epoch": 2.2420897718910964, "grad_norm": 1.669927753852559, "learning_rate": 3.171691919431623e-06, "loss": 0.2332, "step": 3047 }, { "epoch": 2.2428256070640176, "grad_norm": 1.794384655172415, "learning_rate": 3.1658893866675487e-06, "loss": 0.2492, "step": 3048 }, { "epoch": 2.2435614422369388, "grad_norm": 1.6853452748151363, "learning_rate": 3.160091168184154e-06, "loss": 0.2519, "step": 3049 }, { "epoch": 2.24429727740986, "grad_norm": 1.7252105382031155, "learning_rate": 3.154297267641776e-06, "loss": 0.2455, "step": 3050 }, { "epoch": 2.2450331125827816, "grad_norm": 1.722679038162274, "learning_rate": 3.148507688698027e-06, "loss": 0.2483, "step": 3051 }, { "epoch": 2.245768947755703, "grad_norm": 1.69455472010349, "learning_rate": 3.142722435007791e-06, "loss": 0.2829, "step": 3052 }, { "epoch": 2.246504782928624, "grad_norm": 1.7448389255962133, "learning_rate": 3.1369415102232148e-06, "loss": 0.2363, "step": 3053 }, { "epoch": 2.247240618101545, "grad_norm": 1.6546206669388588, "learning_rate": 3.13116491799372e-06, "loss": 0.2386, "step": 3054 }, { "epoch": 2.2479764532744664, "grad_norm": 1.6449584932819352, "learning_rate": 3.1253926619659912e-06, "loss": 0.243, "step": 3055 }, { "epoch": 2.248712288447388, "grad_norm": 1.7987099828824784, "learning_rate": 3.119624745783969e-06, "loss": 0.2715, "step": 3056 }, { "epoch": 2.249448123620309, "grad_norm": 1.831428485711199, "learning_rate": 3.1138611730888703e-06, "loss": 0.2562, "step": 3057 }, { "epoch": 2.2501839587932304, "grad_norm": 1.6686089462341602, "learning_rate": 3.108101947519151e-06, "loss": 0.2533, "step": 3058 }, { "epoch": 2.2509197939661516, "grad_norm": 1.6316905580616181, "learning_rate": 3.1023470727105344e-06, "loss": 0.2438, "step": 3059 }, { "epoch": 2.251655629139073, "grad_norm": 1.7386800793035655, "learning_rate": 3.0965965522959985e-06, "loss": 0.2632, "step": 3060 }, { "epoch": 2.252391464311994, "grad_norm": 1.645460864313617, "learning_rate": 3.0908503899057605e-06, "loss": 0.2419, "step": 3061 }, { "epoch": 2.253127299484915, "grad_norm": 1.65637239127045, "learning_rate": 3.0851085891672984e-06, "loss": 0.2241, "step": 3062 }, { "epoch": 2.253863134657837, "grad_norm": 1.8100325432125162, "learning_rate": 3.0793711537053328e-06, "loss": 0.2558, "step": 3063 }, { "epoch": 2.254598969830758, "grad_norm": 1.6786903419185508, "learning_rate": 3.0736380871418305e-06, "loss": 0.2719, "step": 3064 }, { "epoch": 2.255334805003679, "grad_norm": 1.676187646492454, "learning_rate": 3.067909393095989e-06, "loss": 0.2434, "step": 3065 }, { "epoch": 2.2560706401766004, "grad_norm": 1.6984282659017687, "learning_rate": 3.0621850751842685e-06, "loss": 0.2554, "step": 3066 }, { "epoch": 2.2568064753495216, "grad_norm": 1.8490036730328823, "learning_rate": 3.0564651370203414e-06, "loss": 0.2431, "step": 3067 }, { "epoch": 2.257542310522443, "grad_norm": 1.6030065620150227, "learning_rate": 3.0507495822151302e-06, "loss": 0.2569, "step": 3068 }, { "epoch": 2.258278145695364, "grad_norm": 1.728948842474602, "learning_rate": 3.0450384143767906e-06, "loss": 0.2585, "step": 3069 }, { "epoch": 2.2590139808682856, "grad_norm": 1.704389014071706, "learning_rate": 3.039331637110697e-06, "loss": 0.2301, "step": 3070 }, { "epoch": 2.259749816041207, "grad_norm": 1.7123356355869208, "learning_rate": 3.0336292540194645e-06, "loss": 0.2535, "step": 3071 }, { "epoch": 2.260485651214128, "grad_norm": 1.6808696018154081, "learning_rate": 3.0279312687029295e-06, "loss": 0.2572, "step": 3072 }, { "epoch": 2.2612214863870492, "grad_norm": 1.7066520311861184, "learning_rate": 3.0222376847581546e-06, "loss": 0.2533, "step": 3073 }, { "epoch": 2.2619573215599704, "grad_norm": 1.6795571083299847, "learning_rate": 3.016548505779413e-06, "loss": 0.2809, "step": 3074 }, { "epoch": 2.262693156732892, "grad_norm": 1.5865656584375918, "learning_rate": 3.010863735358217e-06, "loss": 0.2709, "step": 3075 }, { "epoch": 2.2634289919058133, "grad_norm": 1.741145094563952, "learning_rate": 3.005183377083277e-06, "loss": 0.2764, "step": 3076 }, { "epoch": 2.2641648270787345, "grad_norm": 1.722627743361094, "learning_rate": 2.999507434540527e-06, "loss": 0.247, "step": 3077 }, { "epoch": 2.2649006622516556, "grad_norm": 1.7014989398600358, "learning_rate": 2.9938359113131167e-06, "loss": 0.2745, "step": 3078 }, { "epoch": 2.265636497424577, "grad_norm": 1.6052246110208996, "learning_rate": 2.9881688109813933e-06, "loss": 0.2532, "step": 3079 }, { "epoch": 2.266372332597498, "grad_norm": 1.646411656397032, "learning_rate": 2.9825061371229248e-06, "loss": 0.2372, "step": 3080 }, { "epoch": 2.2671081677704192, "grad_norm": 1.6650832982578632, "learning_rate": 2.9768478933124777e-06, "loss": 0.2484, "step": 3081 }, { "epoch": 2.267844002943341, "grad_norm": 1.6369296519473717, "learning_rate": 2.971194083122029e-06, "loss": 0.2605, "step": 3082 }, { "epoch": 2.268579838116262, "grad_norm": 1.7579773965430159, "learning_rate": 2.965544710120741e-06, "loss": 0.222, "step": 3083 }, { "epoch": 2.2693156732891833, "grad_norm": 1.6430295580302254, "learning_rate": 2.9598997778750005e-06, "loss": 0.2596, "step": 3084 }, { "epoch": 2.2700515084621045, "grad_norm": 1.765882210839475, "learning_rate": 2.9542592899483633e-06, "loss": 0.2666, "step": 3085 }, { "epoch": 2.2707873436350257, "grad_norm": 1.7527665689101135, "learning_rate": 2.9486232499015988e-06, "loss": 0.2542, "step": 3086 }, { "epoch": 2.271523178807947, "grad_norm": 1.7960024089250144, "learning_rate": 2.9429916612926636e-06, "loss": 0.2562, "step": 3087 }, { "epoch": 2.272259013980868, "grad_norm": 1.709402754977815, "learning_rate": 2.937364527676697e-06, "loss": 0.2781, "step": 3088 }, { "epoch": 2.2729948491537897, "grad_norm": 1.7311102091343018, "learning_rate": 2.931741852606034e-06, "loss": 0.2335, "step": 3089 }, { "epoch": 2.273730684326711, "grad_norm": 1.8007663342179443, "learning_rate": 2.926123639630193e-06, "loss": 0.2662, "step": 3090 }, { "epoch": 2.274466519499632, "grad_norm": 1.6464159366020426, "learning_rate": 2.920509892295875e-06, "loss": 0.2497, "step": 3091 }, { "epoch": 2.2752023546725533, "grad_norm": 1.8185533942972576, "learning_rate": 2.9149006141469616e-06, "loss": 0.2648, "step": 3092 }, { "epoch": 2.2759381898454745, "grad_norm": 1.6538733467276947, "learning_rate": 2.9092958087245184e-06, "loss": 0.233, "step": 3093 }, { "epoch": 2.276674025018396, "grad_norm": 1.7053476238924323, "learning_rate": 2.903695479566774e-06, "loss": 0.2555, "step": 3094 }, { "epoch": 2.2774098601913173, "grad_norm": 1.6510596862565527, "learning_rate": 2.8980996302091434e-06, "loss": 0.2777, "step": 3095 }, { "epoch": 2.2781456953642385, "grad_norm": 1.6753741210205064, "learning_rate": 2.8925082641842137e-06, "loss": 0.2713, "step": 3096 }, { "epoch": 2.2788815305371597, "grad_norm": 1.5441894855336937, "learning_rate": 2.886921385021729e-06, "loss": 0.2304, "step": 3097 }, { "epoch": 2.279617365710081, "grad_norm": 1.723084045133016, "learning_rate": 2.8813389962486205e-06, "loss": 0.2275, "step": 3098 }, { "epoch": 2.280353200883002, "grad_norm": 1.5980311281497175, "learning_rate": 2.8757611013889663e-06, "loss": 0.2146, "step": 3099 }, { "epoch": 2.2810890360559233, "grad_norm": 1.8127826828542457, "learning_rate": 2.870187703964017e-06, "loss": 0.2566, "step": 3100 }, { "epoch": 2.281824871228845, "grad_norm": 1.615441371398955, "learning_rate": 2.864618807492183e-06, "loss": 0.2352, "step": 3101 }, { "epoch": 2.282560706401766, "grad_norm": 1.7936263487357795, "learning_rate": 2.8590544154890364e-06, "loss": 0.2753, "step": 3102 }, { "epoch": 2.2832965415746873, "grad_norm": 1.8052894638857888, "learning_rate": 2.8534945314672946e-06, "loss": 0.2642, "step": 3103 }, { "epoch": 2.2840323767476085, "grad_norm": 1.7163785204534308, "learning_rate": 2.847939158936839e-06, "loss": 0.2238, "step": 3104 }, { "epoch": 2.2847682119205297, "grad_norm": 1.856247948390034, "learning_rate": 2.842388301404705e-06, "loss": 0.275, "step": 3105 }, { "epoch": 2.285504047093451, "grad_norm": 1.7877785367034391, "learning_rate": 2.8368419623750633e-06, "loss": 0.2451, "step": 3106 }, { "epoch": 2.286239882266372, "grad_norm": 1.8099373760914674, "learning_rate": 2.831300145349254e-06, "loss": 0.2471, "step": 3107 }, { "epoch": 2.2869757174392937, "grad_norm": 1.6548167553014372, "learning_rate": 2.825762853825741e-06, "loss": 0.2541, "step": 3108 }, { "epoch": 2.287711552612215, "grad_norm": 1.7297742752005965, "learning_rate": 2.8202300913001445e-06, "loss": 0.2276, "step": 3109 }, { "epoch": 2.288447387785136, "grad_norm": 1.749490420144084, "learning_rate": 2.8147018612652223e-06, "loss": 0.2554, "step": 3110 }, { "epoch": 2.2891832229580573, "grad_norm": 1.7592059090142704, "learning_rate": 2.809178167210872e-06, "loss": 0.2633, "step": 3111 }, { "epoch": 2.2899190581309785, "grad_norm": 1.736145780111217, "learning_rate": 2.8036590126241226e-06, "loss": 0.2627, "step": 3112 }, { "epoch": 2.2906548933039, "grad_norm": 1.8448152942354743, "learning_rate": 2.798144400989141e-06, "loss": 0.2526, "step": 3113 }, { "epoch": 2.2913907284768213, "grad_norm": 1.6071700092674437, "learning_rate": 2.792634335787231e-06, "loss": 0.2469, "step": 3114 }, { "epoch": 2.2921265636497425, "grad_norm": 1.8432304223474651, "learning_rate": 2.7871288204968127e-06, "loss": 0.2537, "step": 3115 }, { "epoch": 2.2928623988226637, "grad_norm": 1.7979321118626523, "learning_rate": 2.7816278585934544e-06, "loss": 0.2852, "step": 3116 }, { "epoch": 2.293598233995585, "grad_norm": 1.6251714534415471, "learning_rate": 2.7761314535498286e-06, "loss": 0.2625, "step": 3117 }, { "epoch": 2.294334069168506, "grad_norm": 1.7682654355212204, "learning_rate": 2.7706396088357444e-06, "loss": 0.2514, "step": 3118 }, { "epoch": 2.2950699043414273, "grad_norm": 1.6801328357227503, "learning_rate": 2.765152327918128e-06, "loss": 0.2135, "step": 3119 }, { "epoch": 2.295805739514349, "grad_norm": 1.7932835871211037, "learning_rate": 2.759669614261028e-06, "loss": 0.2607, "step": 3120 }, { "epoch": 2.29654157468727, "grad_norm": 1.7230611694730817, "learning_rate": 2.754191471325601e-06, "loss": 0.2326, "step": 3121 }, { "epoch": 2.2972774098601914, "grad_norm": 1.808331804021583, "learning_rate": 2.748717902570126e-06, "loss": 0.2428, "step": 3122 }, { "epoch": 2.2980132450331126, "grad_norm": 1.6243952036559184, "learning_rate": 2.743248911449996e-06, "loss": 0.2461, "step": 3123 }, { "epoch": 2.2987490802060337, "grad_norm": 1.7084168474428696, "learning_rate": 2.737784501417702e-06, "loss": 0.264, "step": 3124 }, { "epoch": 2.299484915378955, "grad_norm": 1.8123407323546499, "learning_rate": 2.7323246759228605e-06, "loss": 0.2412, "step": 3125 }, { "epoch": 2.300220750551876, "grad_norm": 1.8643404190429422, "learning_rate": 2.7268694384121798e-06, "loss": 0.251, "step": 3126 }, { "epoch": 2.3009565857247978, "grad_norm": 1.7738659561820855, "learning_rate": 2.7214187923294766e-06, "loss": 0.2799, "step": 3127 }, { "epoch": 2.301692420897719, "grad_norm": 1.648417358299595, "learning_rate": 2.7159727411156734e-06, "loss": 0.2369, "step": 3128 }, { "epoch": 2.30242825607064, "grad_norm": 1.8594249783298513, "learning_rate": 2.710531288208782e-06, "loss": 0.2413, "step": 3129 }, { "epoch": 2.3031640912435614, "grad_norm": 1.7531071675229246, "learning_rate": 2.70509443704392e-06, "loss": 0.2334, "step": 3130 }, { "epoch": 2.3038999264164826, "grad_norm": 1.6730748091771641, "learning_rate": 2.699662191053298e-06, "loss": 0.2593, "step": 3131 }, { "epoch": 2.304635761589404, "grad_norm": 1.6885920117897, "learning_rate": 2.6942345536662218e-06, "loss": 0.2386, "step": 3132 }, { "epoch": 2.3053715967623254, "grad_norm": 1.7113485771644745, "learning_rate": 2.6888115283090754e-06, "loss": 0.2586, "step": 3133 }, { "epoch": 2.3061074319352466, "grad_norm": 1.9160098300003527, "learning_rate": 2.683393118405352e-06, "loss": 0.2572, "step": 3134 }, { "epoch": 2.306843267108168, "grad_norm": 1.7273030277485963, "learning_rate": 2.6779793273756105e-06, "loss": 0.2344, "step": 3135 }, { "epoch": 2.307579102281089, "grad_norm": 1.773722604501024, "learning_rate": 2.6725701586375075e-06, "loss": 0.2851, "step": 3136 }, { "epoch": 2.30831493745401, "grad_norm": 1.6860806985921892, "learning_rate": 2.66716561560578e-06, "loss": 0.2519, "step": 3137 }, { "epoch": 2.3090507726269314, "grad_norm": 1.6120045556343952, "learning_rate": 2.661765701692236e-06, "loss": 0.2759, "step": 3138 }, { "epoch": 2.309786607799853, "grad_norm": 1.6700091692316805, "learning_rate": 2.6563704203057704e-06, "loss": 0.2311, "step": 3139 }, { "epoch": 2.310522442972774, "grad_norm": 1.6987471239495073, "learning_rate": 2.6509797748523504e-06, "loss": 0.2545, "step": 3140 }, { "epoch": 2.3112582781456954, "grad_norm": 1.805670688306748, "learning_rate": 2.645593768735021e-06, "loss": 0.2419, "step": 3141 }, { "epoch": 2.3119941133186166, "grad_norm": 1.7852268585779296, "learning_rate": 2.6402124053538837e-06, "loss": 0.2364, "step": 3142 }, { "epoch": 2.312729948491538, "grad_norm": 1.8154583000045406, "learning_rate": 2.634835688106132e-06, "loss": 0.2619, "step": 3143 }, { "epoch": 2.313465783664459, "grad_norm": 1.7181694104645389, "learning_rate": 2.6294636203860058e-06, "loss": 0.2402, "step": 3144 }, { "epoch": 2.31420161883738, "grad_norm": 1.751040944590594, "learning_rate": 2.6240962055848196e-06, "loss": 0.2522, "step": 3145 }, { "epoch": 2.314937454010302, "grad_norm": 1.8195932546318754, "learning_rate": 2.6187334470909543e-06, "loss": 0.2717, "step": 3146 }, { "epoch": 2.315673289183223, "grad_norm": 1.7607905234721462, "learning_rate": 2.6133753482898396e-06, "loss": 0.27, "step": 3147 }, { "epoch": 2.316409124356144, "grad_norm": 1.7585589655084422, "learning_rate": 2.6080219125639703e-06, "loss": 0.2591, "step": 3148 }, { "epoch": 2.3171449595290654, "grad_norm": 1.779331736059742, "learning_rate": 2.6026731432929008e-06, "loss": 0.2725, "step": 3149 }, { "epoch": 2.3178807947019866, "grad_norm": 1.6816406406171467, "learning_rate": 2.5973290438532384e-06, "loss": 0.238, "step": 3150 }, { "epoch": 2.3186166298749082, "grad_norm": 1.7625996219096118, "learning_rate": 2.5919896176186287e-06, "loss": 0.2219, "step": 3151 }, { "epoch": 2.3193524650478294, "grad_norm": 1.5848648130622267, "learning_rate": 2.586654867959794e-06, "loss": 0.2466, "step": 3152 }, { "epoch": 2.3200883002207506, "grad_norm": 1.7396530710346716, "learning_rate": 2.5813247982444767e-06, "loss": 0.274, "step": 3153 }, { "epoch": 2.320824135393672, "grad_norm": 1.6815088406259502, "learning_rate": 2.575999411837481e-06, "loss": 0.2634, "step": 3154 }, { "epoch": 2.321559970566593, "grad_norm": 1.6926450996562248, "learning_rate": 2.5706787121006547e-06, "loss": 0.2509, "step": 3155 }, { "epoch": 2.322295805739514, "grad_norm": 1.8086302099797613, "learning_rate": 2.5653627023928764e-06, "loss": 0.2774, "step": 3156 }, { "epoch": 2.3230316409124354, "grad_norm": 1.6738590449783146, "learning_rate": 2.560051386070073e-06, "loss": 0.2391, "step": 3157 }, { "epoch": 2.323767476085357, "grad_norm": 1.7207221214222048, "learning_rate": 2.5547447664852066e-06, "loss": 0.2454, "step": 3158 }, { "epoch": 2.3245033112582782, "grad_norm": 1.813442543672861, "learning_rate": 2.5494428469882747e-06, "loss": 0.2481, "step": 3159 }, { "epoch": 2.3252391464311994, "grad_norm": 1.7691482219625898, "learning_rate": 2.5441456309263e-06, "loss": 0.2789, "step": 3160 }, { "epoch": 2.3259749816041206, "grad_norm": 1.6759464460858196, "learning_rate": 2.538853121643354e-06, "loss": 0.2404, "step": 3161 }, { "epoch": 2.326710816777042, "grad_norm": 1.646584010131482, "learning_rate": 2.533565322480516e-06, "loss": 0.2403, "step": 3162 }, { "epoch": 2.327446651949963, "grad_norm": 1.762789607152588, "learning_rate": 2.5282822367759054e-06, "loss": 0.2567, "step": 3163 }, { "epoch": 2.3281824871228842, "grad_norm": 1.784862220530603, "learning_rate": 2.523003867864663e-06, "loss": 0.249, "step": 3164 }, { "epoch": 2.328918322295806, "grad_norm": 1.6839521498968186, "learning_rate": 2.5177302190789465e-06, "loss": 0.253, "step": 3165 }, { "epoch": 2.329654157468727, "grad_norm": 1.682306369587713, "learning_rate": 2.512461293747942e-06, "loss": 0.2498, "step": 3166 }, { "epoch": 2.3303899926416483, "grad_norm": 1.889800297622584, "learning_rate": 2.5071970951978487e-06, "loss": 0.2587, "step": 3167 }, { "epoch": 2.3311258278145695, "grad_norm": 1.77097444761681, "learning_rate": 2.501937626751887e-06, "loss": 0.2733, "step": 3168 }, { "epoch": 2.3318616629874906, "grad_norm": 1.8586974823282683, "learning_rate": 2.496682891730279e-06, "loss": 0.2474, "step": 3169 }, { "epoch": 2.3325974981604123, "grad_norm": 1.6854400705431196, "learning_rate": 2.4914328934502774e-06, "loss": 0.233, "step": 3170 }, { "epoch": 2.3333333333333335, "grad_norm": 1.6884699856366032, "learning_rate": 2.486187635226127e-06, "loss": 0.2793, "step": 3171 }, { "epoch": 2.3340691685062547, "grad_norm": 1.8588550824113401, "learning_rate": 2.480947120369089e-06, "loss": 0.2474, "step": 3172 }, { "epoch": 2.334805003679176, "grad_norm": 1.7295940299471173, "learning_rate": 2.4757113521874344e-06, "loss": 0.241, "step": 3173 }, { "epoch": 2.335540838852097, "grad_norm": 1.6226550170485745, "learning_rate": 2.4704803339864237e-06, "loss": 0.2612, "step": 3174 }, { "epoch": 2.3362766740250183, "grad_norm": 1.8818996743415843, "learning_rate": 2.4652540690683315e-06, "loss": 0.2482, "step": 3175 }, { "epoch": 2.3370125091979395, "grad_norm": 1.8360943589851486, "learning_rate": 2.460032560732427e-06, "loss": 0.225, "step": 3176 }, { "epoch": 2.337748344370861, "grad_norm": 1.7738289702829784, "learning_rate": 2.454815812274981e-06, "loss": 0.2487, "step": 3177 }, { "epoch": 2.3384841795437823, "grad_norm": 1.7304346380570972, "learning_rate": 2.4496038269892455e-06, "loss": 0.2456, "step": 3178 }, { "epoch": 2.3392200147167035, "grad_norm": 1.8703448487171839, "learning_rate": 2.444396608165488e-06, "loss": 0.2673, "step": 3179 }, { "epoch": 2.3399558498896247, "grad_norm": 1.6890271758036735, "learning_rate": 2.439194159090945e-06, "loss": 0.2832, "step": 3180 }, { "epoch": 2.340691685062546, "grad_norm": 1.6716779413532699, "learning_rate": 2.433996483049855e-06, "loss": 0.2351, "step": 3181 }, { "epoch": 2.341427520235467, "grad_norm": 1.8337646456959118, "learning_rate": 2.4288035833234437e-06, "loss": 0.2518, "step": 3182 }, { "epoch": 2.3421633554083887, "grad_norm": 1.7533141947215194, "learning_rate": 2.423615463189911e-06, "loss": 0.2343, "step": 3183 }, { "epoch": 2.34289919058131, "grad_norm": 1.6484029053288234, "learning_rate": 2.418432125924449e-06, "loss": 0.2389, "step": 3184 }, { "epoch": 2.343635025754231, "grad_norm": 1.8060211968963762, "learning_rate": 2.413253574799228e-06, "loss": 0.2648, "step": 3185 }, { "epoch": 2.3443708609271523, "grad_norm": 1.8057495000967085, "learning_rate": 2.4080798130834004e-06, "loss": 0.311, "step": 3186 }, { "epoch": 2.3451066961000735, "grad_norm": 1.7108498522077873, "learning_rate": 2.4029108440430838e-06, "loss": 0.2308, "step": 3187 }, { "epoch": 2.3458425312729947, "grad_norm": 1.7386920064150027, "learning_rate": 2.397746670941381e-06, "loss": 0.2344, "step": 3188 }, { "epoch": 2.3465783664459163, "grad_norm": 1.8028763381576836, "learning_rate": 2.3925872970383657e-06, "loss": 0.2437, "step": 3189 }, { "epoch": 2.3473142016188375, "grad_norm": 1.8058972318769722, "learning_rate": 2.387432725591078e-06, "loss": 0.2488, "step": 3190 }, { "epoch": 2.3480500367917587, "grad_norm": 1.7197769359220054, "learning_rate": 2.3822829598535324e-06, "loss": 0.2669, "step": 3191 }, { "epoch": 2.34878587196468, "grad_norm": 1.6987180214464481, "learning_rate": 2.3771380030767e-06, "loss": 0.22, "step": 3192 }, { "epoch": 2.349521707137601, "grad_norm": 1.7987883381675105, "learning_rate": 2.3719978585085234e-06, "loss": 0.2786, "step": 3193 }, { "epoch": 2.3502575423105223, "grad_norm": 1.705334231476841, "learning_rate": 2.3668625293939085e-06, "loss": 0.2597, "step": 3194 }, { "epoch": 2.3509933774834435, "grad_norm": 1.688366042513513, "learning_rate": 2.3617320189747173e-06, "loss": 0.2318, "step": 3195 }, { "epoch": 2.351729212656365, "grad_norm": 1.652186669360767, "learning_rate": 2.356606330489769e-06, "loss": 0.2272, "step": 3196 }, { "epoch": 2.3524650478292863, "grad_norm": 1.7607215475982534, "learning_rate": 2.351485467174841e-06, "loss": 0.2713, "step": 3197 }, { "epoch": 2.3532008830022075, "grad_norm": 1.6694609409760766, "learning_rate": 2.346369432262665e-06, "loss": 0.2673, "step": 3198 }, { "epoch": 2.3539367181751287, "grad_norm": 1.6876852463560286, "learning_rate": 2.3412582289829254e-06, "loss": 0.2525, "step": 3199 }, { "epoch": 2.35467255334805, "grad_norm": 1.595270499737159, "learning_rate": 2.336151860562256e-06, "loss": 0.206, "step": 3200 }, { "epoch": 2.355408388520971, "grad_norm": 1.7494754596756468, "learning_rate": 2.3310503302242317e-06, "loss": 0.2481, "step": 3201 }, { "epoch": 2.3561442236938928, "grad_norm": 1.7886681231581956, "learning_rate": 2.3259536411893836e-06, "loss": 0.2711, "step": 3202 }, { "epoch": 2.356880058866814, "grad_norm": 1.71188281169439, "learning_rate": 2.32086179667518e-06, "loss": 0.2478, "step": 3203 }, { "epoch": 2.357615894039735, "grad_norm": 1.7331239366384568, "learning_rate": 2.315774799896037e-06, "loss": 0.2696, "step": 3204 }, { "epoch": 2.3583517292126563, "grad_norm": 1.6227824239151234, "learning_rate": 2.3106926540633e-06, "loss": 0.2376, "step": 3205 }, { "epoch": 2.3590875643855775, "grad_norm": 1.7526278104083681, "learning_rate": 2.3056153623852607e-06, "loss": 0.2466, "step": 3206 }, { "epoch": 2.3598233995584987, "grad_norm": 1.813226410056469, "learning_rate": 2.300542928067148e-06, "loss": 0.2726, "step": 3207 }, { "epoch": 2.3605592347314204, "grad_norm": 1.61741600956033, "learning_rate": 2.2954753543111097e-06, "loss": 0.2379, "step": 3208 }, { "epoch": 2.3612950699043416, "grad_norm": 1.7720220493639876, "learning_rate": 2.290412644316249e-06, "loss": 0.2593, "step": 3209 }, { "epoch": 2.3620309050772628, "grad_norm": 1.715687225443807, "learning_rate": 2.2853548012785764e-06, "loss": 0.2814, "step": 3210 }, { "epoch": 2.362766740250184, "grad_norm": 1.719236133007251, "learning_rate": 2.2803018283910415e-06, "loss": 0.2664, "step": 3211 }, { "epoch": 2.363502575423105, "grad_norm": 1.6909130567292407, "learning_rate": 2.2752537288435194e-06, "loss": 0.2716, "step": 3212 }, { "epoch": 2.3642384105960264, "grad_norm": 1.6998884807021137, "learning_rate": 2.2702105058228063e-06, "loss": 0.231, "step": 3213 }, { "epoch": 2.3649742457689475, "grad_norm": 1.7203724878233522, "learning_rate": 2.2651721625126167e-06, "loss": 0.2286, "step": 3214 }, { "epoch": 2.365710080941869, "grad_norm": 1.5828691337622731, "learning_rate": 2.2601387020935884e-06, "loss": 0.2292, "step": 3215 }, { "epoch": 2.3664459161147904, "grad_norm": 1.8526175346680884, "learning_rate": 2.255110127743281e-06, "loss": 0.2664, "step": 3216 }, { "epoch": 2.3671817512877116, "grad_norm": 1.7927824258438587, "learning_rate": 2.2500864426361556e-06, "loss": 0.2525, "step": 3217 }, { "epoch": 2.3679175864606328, "grad_norm": 1.6703405260980453, "learning_rate": 2.2450676499436063e-06, "loss": 0.2331, "step": 3218 }, { "epoch": 2.368653421633554, "grad_norm": 1.7284253263313913, "learning_rate": 2.240053752833923e-06, "loss": 0.2261, "step": 3219 }, { "epoch": 2.369389256806475, "grad_norm": 1.7406757114379334, "learning_rate": 2.23504475447231e-06, "loss": 0.2633, "step": 3220 }, { "epoch": 2.370125091979397, "grad_norm": 1.7599671181306462, "learning_rate": 2.23004065802088e-06, "loss": 0.2548, "step": 3221 }, { "epoch": 2.370860927152318, "grad_norm": 1.6972909501193318, "learning_rate": 2.225041466638652e-06, "loss": 0.2524, "step": 3222 }, { "epoch": 2.371596762325239, "grad_norm": 1.7668484320838462, "learning_rate": 2.2200471834815497e-06, "loss": 0.2548, "step": 3223 }, { "epoch": 2.3723325974981604, "grad_norm": 1.7137242698608945, "learning_rate": 2.2150578117023904e-06, "loss": 0.227, "step": 3224 }, { "epoch": 2.3730684326710816, "grad_norm": 1.8075157152368038, "learning_rate": 2.210073354450898e-06, "loss": 0.2469, "step": 3225 }, { "epoch": 2.373804267844003, "grad_norm": 1.7879335377940944, "learning_rate": 2.2050938148736934e-06, "loss": 0.261, "step": 3226 }, { "epoch": 2.3745401030169244, "grad_norm": 1.6277624226164618, "learning_rate": 2.200119196114294e-06, "loss": 0.2798, "step": 3227 }, { "epoch": 2.3752759381898456, "grad_norm": 1.6163460114420654, "learning_rate": 2.195149501313101e-06, "loss": 0.2171, "step": 3228 }, { "epoch": 2.376011773362767, "grad_norm": 1.674602674757152, "learning_rate": 2.1901847336074258e-06, "loss": 0.2396, "step": 3229 }, { "epoch": 2.376747608535688, "grad_norm": 1.73038090238588, "learning_rate": 2.1852248961314493e-06, "loss": 0.2483, "step": 3230 }, { "epoch": 2.377483443708609, "grad_norm": 1.6805222562469322, "learning_rate": 2.180269992016254e-06, "loss": 0.2463, "step": 3231 }, { "epoch": 2.3782192788815304, "grad_norm": 1.8075253897015147, "learning_rate": 2.1753200243898032e-06, "loss": 0.2129, "step": 3232 }, { "epoch": 2.3789551140544516, "grad_norm": 1.7761542150856577, "learning_rate": 2.1703749963769406e-06, "loss": 0.2824, "step": 3233 }, { "epoch": 2.3796909492273732, "grad_norm": 1.7314142369449934, "learning_rate": 2.1654349110993957e-06, "loss": 0.2787, "step": 3234 }, { "epoch": 2.3804267844002944, "grad_norm": 1.7738230049271393, "learning_rate": 2.160499771675778e-06, "loss": 0.2285, "step": 3235 }, { "epoch": 2.3811626195732156, "grad_norm": 1.7464603219856476, "learning_rate": 2.1555695812215773e-06, "loss": 0.2313, "step": 3236 }, { "epoch": 2.381898454746137, "grad_norm": 1.6712272589550297, "learning_rate": 2.1506443428491454e-06, "loss": 0.2527, "step": 3237 }, { "epoch": 2.382634289919058, "grad_norm": 1.7094123032983815, "learning_rate": 2.14572405966773e-06, "loss": 0.2382, "step": 3238 }, { "epoch": 2.383370125091979, "grad_norm": 1.6600570028435553, "learning_rate": 2.1408087347834293e-06, "loss": 0.2657, "step": 3239 }, { "epoch": 2.384105960264901, "grad_norm": 1.7078291858383108, "learning_rate": 2.1358983712992253e-06, "loss": 0.2451, "step": 3240 }, { "epoch": 2.384841795437822, "grad_norm": 1.7653303622113787, "learning_rate": 2.130992972314965e-06, "loss": 0.2631, "step": 3241 }, { "epoch": 2.3855776306107432, "grad_norm": 1.5696093347681386, "learning_rate": 2.1260925409273536e-06, "loss": 0.2256, "step": 3242 }, { "epoch": 2.3863134657836644, "grad_norm": 1.7700116262724819, "learning_rate": 2.121197080229971e-06, "loss": 0.2639, "step": 3243 }, { "epoch": 2.3870493009565856, "grad_norm": 1.7534476098944274, "learning_rate": 2.1163065933132544e-06, "loss": 0.2335, "step": 3244 }, { "epoch": 2.387785136129507, "grad_norm": 1.6575978912767368, "learning_rate": 2.1114210832645023e-06, "loss": 0.2375, "step": 3245 }, { "epoch": 2.3885209713024285, "grad_norm": 1.7084368391579063, "learning_rate": 2.106540553167864e-06, "loss": 0.2299, "step": 3246 }, { "epoch": 2.3892568064753497, "grad_norm": 1.8703478820831794, "learning_rate": 2.101665006104362e-06, "loss": 0.2581, "step": 3247 }, { "epoch": 2.389992641648271, "grad_norm": 1.7408269422810914, "learning_rate": 2.096794445151854e-06, "loss": 0.2573, "step": 3248 }, { "epoch": 2.390728476821192, "grad_norm": 1.6637502345480468, "learning_rate": 2.091928873385062e-06, "loss": 0.2515, "step": 3249 }, { "epoch": 2.3914643119941132, "grad_norm": 1.9812563510977517, "learning_rate": 2.087068293875557e-06, "loss": 0.246, "step": 3250 }, { "epoch": 2.3922001471670344, "grad_norm": 1.7560972327538718, "learning_rate": 2.082212709691751e-06, "loss": 0.2607, "step": 3251 }, { "epoch": 2.3929359823399556, "grad_norm": 1.7476337300952962, "learning_rate": 2.0773621238989117e-06, "loss": 0.2448, "step": 3252 }, { "epoch": 2.3936718175128773, "grad_norm": 1.8094879607575942, "learning_rate": 2.0725165395591472e-06, "loss": 0.2475, "step": 3253 }, { "epoch": 2.3944076526857985, "grad_norm": 1.7306422686988285, "learning_rate": 2.067675959731411e-06, "loss": 0.2595, "step": 3254 }, { "epoch": 2.3951434878587197, "grad_norm": 1.758675622118065, "learning_rate": 2.06284038747149e-06, "loss": 0.2215, "step": 3255 }, { "epoch": 2.395879323031641, "grad_norm": 1.7455564927688045, "learning_rate": 2.0580098258320167e-06, "loss": 0.2618, "step": 3256 }, { "epoch": 2.396615158204562, "grad_norm": 1.6811815967083927, "learning_rate": 2.05318427786246e-06, "loss": 0.2678, "step": 3257 }, { "epoch": 2.3973509933774833, "grad_norm": 1.8232413117199282, "learning_rate": 2.048363746609121e-06, "loss": 0.2785, "step": 3258 }, { "epoch": 2.398086828550405, "grad_norm": 1.874369636524774, "learning_rate": 2.043548235115139e-06, "loss": 0.2442, "step": 3259 }, { "epoch": 2.398822663723326, "grad_norm": 1.72168503029065, "learning_rate": 2.038737746420476e-06, "loss": 0.2702, "step": 3260 }, { "epoch": 2.3995584988962473, "grad_norm": 1.692517415791343, "learning_rate": 2.033932283561929e-06, "loss": 0.2575, "step": 3261 }, { "epoch": 2.4002943340691685, "grad_norm": 1.6979464705003244, "learning_rate": 2.0291318495731215e-06, "loss": 0.243, "step": 3262 }, { "epoch": 2.4010301692420897, "grad_norm": 1.7705359018948497, "learning_rate": 2.024336447484504e-06, "loss": 0.2608, "step": 3263 }, { "epoch": 2.401766004415011, "grad_norm": 1.7285618164725964, "learning_rate": 2.019546080323345e-06, "loss": 0.2514, "step": 3264 }, { "epoch": 2.401766004415011, "eval_en_loss": 1.0151522159576416, "eval_en_runtime": 14.9015, "eval_en_samples_per_second": 65.631, "eval_en_steps_per_second": 0.134, "step": 3264 }, { "epoch": 2.401766004415011, "eval_eu_loss": 0.9455450177192688, "eval_eu_runtime": 14.9469, "eval_eu_samples_per_second": 65.9, "eval_eu_steps_per_second": 0.134, "step": 3264 }, { "epoch": 2.4025018395879325, "grad_norm": 1.6366934401941358, "learning_rate": 2.014760751113738e-06, "loss": 0.2185, "step": 3265 }, { "epoch": 2.4032376747608537, "grad_norm": 1.6558153879936952, "learning_rate": 2.0099804628766006e-06, "loss": 0.2457, "step": 3266 }, { "epoch": 2.403973509933775, "grad_norm": 1.5385993003099847, "learning_rate": 2.0052052186296546e-06, "loss": 0.239, "step": 3267 }, { "epoch": 2.404709345106696, "grad_norm": 1.6455680812676081, "learning_rate": 2.000435021387457e-06, "loss": 0.259, "step": 3268 }, { "epoch": 2.4054451802796173, "grad_norm": 1.6894758200486009, "learning_rate": 1.99566987416136e-06, "loss": 0.2495, "step": 3269 }, { "epoch": 2.4061810154525385, "grad_norm": 1.830035704958596, "learning_rate": 1.99090977995954e-06, "loss": 0.2774, "step": 3270 }, { "epoch": 2.4069168506254597, "grad_norm": 1.771882660019693, "learning_rate": 1.9861547417869776e-06, "loss": 0.2401, "step": 3271 }, { "epoch": 2.4076526857983813, "grad_norm": 1.6788794234900337, "learning_rate": 1.981404762645467e-06, "loss": 0.2467, "step": 3272 }, { "epoch": 2.4083885209713025, "grad_norm": 1.6402621527837005, "learning_rate": 1.9766598455335994e-06, "loss": 0.2053, "step": 3273 }, { "epoch": 2.4091243561442237, "grad_norm": 1.5739696841753201, "learning_rate": 1.9719199934467804e-06, "loss": 0.2566, "step": 3274 }, { "epoch": 2.409860191317145, "grad_norm": 1.6530098393342911, "learning_rate": 1.9671852093772147e-06, "loss": 0.2703, "step": 3275 }, { "epoch": 2.410596026490066, "grad_norm": 1.7717601283510982, "learning_rate": 1.9624554963138996e-06, "loss": 0.2527, "step": 3276 }, { "epoch": 2.4113318616629873, "grad_norm": 1.7614861201233665, "learning_rate": 1.957730857242649e-06, "loss": 0.2598, "step": 3277 }, { "epoch": 2.412067696835909, "grad_norm": 1.6715689961867617, "learning_rate": 1.9530112951460545e-06, "loss": 0.2512, "step": 3278 }, { "epoch": 2.41280353200883, "grad_norm": 1.7387368709315854, "learning_rate": 1.948296813003514e-06, "loss": 0.2416, "step": 3279 }, { "epoch": 2.4135393671817513, "grad_norm": 1.810654416412113, "learning_rate": 1.943587413791217e-06, "loss": 0.2862, "step": 3280 }, { "epoch": 2.4142752023546725, "grad_norm": 1.8128083098209944, "learning_rate": 1.938883100482144e-06, "loss": 0.2666, "step": 3281 }, { "epoch": 2.4150110375275937, "grad_norm": 1.7801829647915604, "learning_rate": 1.9341838760460587e-06, "loss": 0.2388, "step": 3282 }, { "epoch": 2.415746872700515, "grad_norm": 1.7450589823748002, "learning_rate": 1.9294897434495196e-06, "loss": 0.2568, "step": 3283 }, { "epoch": 2.4164827078734366, "grad_norm": 1.781881934754957, "learning_rate": 1.9248007056558725e-06, "loss": 0.2214, "step": 3284 }, { "epoch": 2.4172185430463577, "grad_norm": 1.8262039594334878, "learning_rate": 1.920116765625234e-06, "loss": 0.2491, "step": 3285 }, { "epoch": 2.417954378219279, "grad_norm": 1.8009774753215544, "learning_rate": 1.915437926314523e-06, "loss": 0.2953, "step": 3286 }, { "epoch": 2.4186902133922, "grad_norm": 1.6595949325338821, "learning_rate": 1.9107641906774166e-06, "loss": 0.2371, "step": 3287 }, { "epoch": 2.4194260485651213, "grad_norm": 1.7267553716949737, "learning_rate": 1.906095561664385e-06, "loss": 0.2158, "step": 3288 }, { "epoch": 2.4201618837380425, "grad_norm": 1.8527888215648394, "learning_rate": 1.9014320422226707e-06, "loss": 0.2688, "step": 3289 }, { "epoch": 2.4208977189109637, "grad_norm": 1.7378926493991702, "learning_rate": 1.8967736352962895e-06, "loss": 0.2388, "step": 3290 }, { "epoch": 2.4216335540838854, "grad_norm": 1.7833033444606217, "learning_rate": 1.8921203438260272e-06, "loss": 0.2447, "step": 3291 }, { "epoch": 2.4223693892568066, "grad_norm": 1.6836683101627505, "learning_rate": 1.8874721707494448e-06, "loss": 0.2503, "step": 3292 }, { "epoch": 2.4231052244297278, "grad_norm": 1.7138451678399624, "learning_rate": 1.882829119000873e-06, "loss": 0.262, "step": 3293 }, { "epoch": 2.423841059602649, "grad_norm": 1.7311960068334658, "learning_rate": 1.8781911915113993e-06, "loss": 0.2617, "step": 3294 }, { "epoch": 2.42457689477557, "grad_norm": 1.7369160048526153, "learning_rate": 1.8735583912088951e-06, "loss": 0.2741, "step": 3295 }, { "epoch": 2.425312729948492, "grad_norm": 1.8523235221539296, "learning_rate": 1.8689307210179753e-06, "loss": 0.235, "step": 3296 }, { "epoch": 2.426048565121413, "grad_norm": 1.6843343512719908, "learning_rate": 1.8643081838600275e-06, "loss": 0.249, "step": 3297 }, { "epoch": 2.426784400294334, "grad_norm": 1.6804029964259308, "learning_rate": 1.8596907826531962e-06, "loss": 0.2233, "step": 3298 }, { "epoch": 2.4275202354672554, "grad_norm": 1.7317925061179722, "learning_rate": 1.8550785203123877e-06, "loss": 0.2376, "step": 3299 }, { "epoch": 2.4282560706401766, "grad_norm": 1.62851192129948, "learning_rate": 1.850471399749253e-06, "loss": 0.2331, "step": 3300 }, { "epoch": 2.4289919058130978, "grad_norm": 1.789083819840286, "learning_rate": 1.8458694238722086e-06, "loss": 0.2489, "step": 3301 }, { "epoch": 2.429727740986019, "grad_norm": 1.693119677937614, "learning_rate": 1.8412725955864209e-06, "loss": 0.2223, "step": 3302 }, { "epoch": 2.4304635761589406, "grad_norm": 1.6666830658714717, "learning_rate": 1.8366809177937961e-06, "loss": 0.2133, "step": 3303 }, { "epoch": 2.431199411331862, "grad_norm": 1.7354625006043276, "learning_rate": 1.8320943933930103e-06, "loss": 0.2646, "step": 3304 }, { "epoch": 2.431935246504783, "grad_norm": 1.7113698234790098, "learning_rate": 1.8275130252794637e-06, "loss": 0.2764, "step": 3305 }, { "epoch": 2.432671081677704, "grad_norm": 1.7938843400202737, "learning_rate": 1.8229368163453153e-06, "loss": 0.2479, "step": 3306 }, { "epoch": 2.4334069168506254, "grad_norm": 1.8636319832238406, "learning_rate": 1.818365769479462e-06, "loss": 0.2557, "step": 3307 }, { "epoch": 2.4341427520235466, "grad_norm": 1.5970696687393966, "learning_rate": 1.8137998875675477e-06, "loss": 0.2314, "step": 3308 }, { "epoch": 2.4348785871964678, "grad_norm": 1.6729359693367711, "learning_rate": 1.8092391734919445e-06, "loss": 0.2238, "step": 3309 }, { "epoch": 2.4356144223693894, "grad_norm": 1.6617606074044837, "learning_rate": 1.8046836301317727e-06, "loss": 0.287, "step": 3310 }, { "epoch": 2.4363502575423106, "grad_norm": 1.7283352028794194, "learning_rate": 1.8001332603628874e-06, "loss": 0.2175, "step": 3311 }, { "epoch": 2.437086092715232, "grad_norm": 1.5639557556609274, "learning_rate": 1.795588067057865e-06, "loss": 0.232, "step": 3312 }, { "epoch": 2.437821927888153, "grad_norm": 1.7843655835109524, "learning_rate": 1.7910480530860363e-06, "loss": 0.264, "step": 3313 }, { "epoch": 2.438557763061074, "grad_norm": 1.6792411836986212, "learning_rate": 1.7865132213134417e-06, "loss": 0.2395, "step": 3314 }, { "epoch": 2.439293598233996, "grad_norm": 1.770699594199088, "learning_rate": 1.7819835746028602e-06, "loss": 0.2593, "step": 3315 }, { "epoch": 2.440029433406917, "grad_norm": 1.7505008066228784, "learning_rate": 1.7774591158137977e-06, "loss": 0.2583, "step": 3316 }, { "epoch": 2.440765268579838, "grad_norm": 1.5821536942121672, "learning_rate": 1.7729398478024852e-06, "loss": 0.279, "step": 3317 }, { "epoch": 2.4415011037527594, "grad_norm": 1.7827035050184032, "learning_rate": 1.768425773421869e-06, "loss": 0.2437, "step": 3318 }, { "epoch": 2.4422369389256806, "grad_norm": 1.9084830924446659, "learning_rate": 1.7639168955216257e-06, "loss": 0.2807, "step": 3319 }, { "epoch": 2.442972774098602, "grad_norm": 1.7372146532088621, "learning_rate": 1.7594132169481504e-06, "loss": 0.2568, "step": 3320 }, { "epoch": 2.443708609271523, "grad_norm": 1.705685796323443, "learning_rate": 1.7549147405445478e-06, "loss": 0.2689, "step": 3321 }, { "epoch": 2.4444444444444446, "grad_norm": 1.70487549962687, "learning_rate": 1.7504214691506527e-06, "loss": 0.2552, "step": 3322 }, { "epoch": 2.445180279617366, "grad_norm": 1.7087364220120773, "learning_rate": 1.7459334056029998e-06, "loss": 0.2389, "step": 3323 }, { "epoch": 2.445916114790287, "grad_norm": 1.8010820330052006, "learning_rate": 1.7414505527348436e-06, "loss": 0.2613, "step": 3324 }, { "epoch": 2.4466519499632082, "grad_norm": 1.7142614541819146, "learning_rate": 1.7369729133761493e-06, "loss": 0.2657, "step": 3325 }, { "epoch": 2.4473877851361294, "grad_norm": 1.7917313058584114, "learning_rate": 1.7325004903535903e-06, "loss": 0.2411, "step": 3326 }, { "epoch": 2.4481236203090506, "grad_norm": 1.7203020522493324, "learning_rate": 1.7280332864905435e-06, "loss": 0.2836, "step": 3327 }, { "epoch": 2.448859455481972, "grad_norm": 1.816297013762729, "learning_rate": 1.7235713046070935e-06, "loss": 0.2729, "step": 3328 }, { "epoch": 2.4495952906548935, "grad_norm": 1.7328982952414567, "learning_rate": 1.7191145475200333e-06, "loss": 0.2323, "step": 3329 }, { "epoch": 2.4503311258278146, "grad_norm": 1.7127940065736864, "learning_rate": 1.7146630180428435e-06, "loss": 0.2382, "step": 3330 }, { "epoch": 2.451066961000736, "grad_norm": 1.674202283145438, "learning_rate": 1.7102167189857255e-06, "loss": 0.2238, "step": 3331 }, { "epoch": 2.451802796173657, "grad_norm": 1.6505731385236384, "learning_rate": 1.705775653155558e-06, "loss": 0.2128, "step": 3332 }, { "epoch": 2.4525386313465782, "grad_norm": 1.8406698624525506, "learning_rate": 1.7013398233559286e-06, "loss": 0.2664, "step": 3333 }, { "epoch": 2.4532744665195, "grad_norm": 1.6129360025080224, "learning_rate": 1.6969092323871195e-06, "loss": 0.2384, "step": 3334 }, { "epoch": 2.454010301692421, "grad_norm": 1.6257494095110958, "learning_rate": 1.6924838830460955e-06, "loss": 0.2531, "step": 3335 }, { "epoch": 2.4547461368653423, "grad_norm": 1.6497536286720702, "learning_rate": 1.6880637781265218e-06, "loss": 0.2615, "step": 3336 }, { "epoch": 2.4554819720382635, "grad_norm": 1.7297938405108186, "learning_rate": 1.6836489204187511e-06, "loss": 0.2518, "step": 3337 }, { "epoch": 2.4562178072111847, "grad_norm": 1.779267161362983, "learning_rate": 1.6792393127098261e-06, "loss": 0.2492, "step": 3338 }, { "epoch": 2.456953642384106, "grad_norm": 1.7678544518162518, "learning_rate": 1.6748349577834622e-06, "loss": 0.2432, "step": 3339 }, { "epoch": 2.457689477557027, "grad_norm": 1.7700587676647996, "learning_rate": 1.6704358584200809e-06, "loss": 0.2577, "step": 3340 }, { "epoch": 2.4584253127299487, "grad_norm": 1.76197696395195, "learning_rate": 1.666042017396764e-06, "loss": 0.2692, "step": 3341 }, { "epoch": 2.45916114790287, "grad_norm": 1.8315188614546565, "learning_rate": 1.661653437487286e-06, "loss": 0.2626, "step": 3342 }, { "epoch": 2.459896983075791, "grad_norm": 1.7625848683654388, "learning_rate": 1.6572701214621013e-06, "loss": 0.2269, "step": 3343 }, { "epoch": 2.4606328182487123, "grad_norm": 1.7829764784952207, "learning_rate": 1.6528920720883323e-06, "loss": 0.2438, "step": 3344 }, { "epoch": 2.4613686534216335, "grad_norm": 1.695963714504497, "learning_rate": 1.6485192921297833e-06, "loss": 0.245, "step": 3345 }, { "epoch": 2.4621044885945547, "grad_norm": 1.7407759976841757, "learning_rate": 1.6441517843469302e-06, "loss": 0.254, "step": 3346 }, { "epoch": 2.462840323767476, "grad_norm": 1.6359952466304646, "learning_rate": 1.6397895514969253e-06, "loss": 0.2561, "step": 3347 }, { "epoch": 2.4635761589403975, "grad_norm": 1.6305314404463196, "learning_rate": 1.6354325963335772e-06, "loss": 0.2435, "step": 3348 }, { "epoch": 2.4643119941133187, "grad_norm": 1.8727688388919523, "learning_rate": 1.631080921607383e-06, "loss": 0.2673, "step": 3349 }, { "epoch": 2.46504782928624, "grad_norm": 1.6559582254891654, "learning_rate": 1.6267345300654891e-06, "loss": 0.2605, "step": 3350 }, { "epoch": 2.465783664459161, "grad_norm": 1.6401300635429674, "learning_rate": 1.6223934244517136e-06, "loss": 0.2617, "step": 3351 }, { "epoch": 2.4665194996320823, "grad_norm": 1.7156404962461203, "learning_rate": 1.6180576075065412e-06, "loss": 0.2715, "step": 3352 }, { "epoch": 2.467255334805004, "grad_norm": 1.659264720304575, "learning_rate": 1.6137270819671081e-06, "loss": 0.2531, "step": 3353 }, { "epoch": 2.467991169977925, "grad_norm": 1.9264956675087563, "learning_rate": 1.6094018505672194e-06, "loss": 0.286, "step": 3354 }, { "epoch": 2.4687270051508463, "grad_norm": 1.7338262570343541, "learning_rate": 1.6050819160373331e-06, "loss": 0.2426, "step": 3355 }, { "epoch": 2.4694628403237675, "grad_norm": 1.6876160410360104, "learning_rate": 1.600767281104566e-06, "loss": 0.2401, "step": 3356 }, { "epoch": 2.4701986754966887, "grad_norm": 1.6607983900961285, "learning_rate": 1.596457948492688e-06, "loss": 0.2365, "step": 3357 }, { "epoch": 2.47093451066961, "grad_norm": 1.643242781996778, "learning_rate": 1.5921539209221238e-06, "loss": 0.2385, "step": 3358 }, { "epoch": 2.471670345842531, "grad_norm": 1.677361036481059, "learning_rate": 1.5878552011099424e-06, "loss": 0.2301, "step": 3359 }, { "epoch": 2.4724061810154527, "grad_norm": 1.665257452074283, "learning_rate": 1.583561791769871e-06, "loss": 0.2624, "step": 3360 }, { "epoch": 2.473142016188374, "grad_norm": 1.6830302837815145, "learning_rate": 1.5792736956122801e-06, "loss": 0.2201, "step": 3361 }, { "epoch": 2.473877851361295, "grad_norm": 1.7150926565787838, "learning_rate": 1.5749909153441812e-06, "loss": 0.2452, "step": 3362 }, { "epoch": 2.4746136865342163, "grad_norm": 1.4995312444352122, "learning_rate": 1.5707134536692425e-06, "loss": 0.2147, "step": 3363 }, { "epoch": 2.4753495217071375, "grad_norm": 1.7743240939070206, "learning_rate": 1.566441313287762e-06, "loss": 0.2345, "step": 3364 }, { "epoch": 2.4760853568800587, "grad_norm": 1.7329929865516152, "learning_rate": 1.5621744968966846e-06, "loss": 0.2316, "step": 3365 }, { "epoch": 2.47682119205298, "grad_norm": 1.751778127137392, "learning_rate": 1.5579130071895944e-06, "loss": 0.2562, "step": 3366 }, { "epoch": 2.4775570272259015, "grad_norm": 1.7067541332891576, "learning_rate": 1.5536568468567126e-06, "loss": 0.252, "step": 3367 }, { "epoch": 2.4782928623988227, "grad_norm": 1.794575471623816, "learning_rate": 1.5494060185848915e-06, "loss": 0.2557, "step": 3368 }, { "epoch": 2.479028697571744, "grad_norm": 1.8653194425755744, "learning_rate": 1.545160525057623e-06, "loss": 0.2539, "step": 3369 }, { "epoch": 2.479764532744665, "grad_norm": 1.642701227572327, "learning_rate": 1.5409203689550313e-06, "loss": 0.2326, "step": 3370 }, { "epoch": 2.4805003679175863, "grad_norm": 1.6289429854191715, "learning_rate": 1.5366855529538615e-06, "loss": 0.229, "step": 3371 }, { "epoch": 2.481236203090508, "grad_norm": 1.773658178710421, "learning_rate": 1.5324560797275056e-06, "loss": 0.2595, "step": 3372 }, { "epoch": 2.481972038263429, "grad_norm": 1.7655703551050013, "learning_rate": 1.5282319519459643e-06, "loss": 0.2346, "step": 3373 }, { "epoch": 2.4827078734363504, "grad_norm": 1.7439779236930453, "learning_rate": 1.5240131722758745e-06, "loss": 0.2552, "step": 3374 }, { "epoch": 2.4834437086092715, "grad_norm": 1.6841634921448754, "learning_rate": 1.5197997433804945e-06, "loss": 0.261, "step": 3375 }, { "epoch": 2.4841795437821927, "grad_norm": 1.7193514872646978, "learning_rate": 1.5155916679197057e-06, "loss": 0.2489, "step": 3376 }, { "epoch": 2.484915378955114, "grad_norm": 1.7217411787016261, "learning_rate": 1.5113889485500056e-06, "loss": 0.2516, "step": 3377 }, { "epoch": 2.485651214128035, "grad_norm": 1.6028619845243868, "learning_rate": 1.5071915879245147e-06, "loss": 0.2291, "step": 3378 }, { "epoch": 2.4863870493009568, "grad_norm": 1.6764016275814222, "learning_rate": 1.5029995886929717e-06, "loss": 0.2267, "step": 3379 }, { "epoch": 2.487122884473878, "grad_norm": 1.6351027195235452, "learning_rate": 1.4988129535017216e-06, "loss": 0.2323, "step": 3380 }, { "epoch": 2.487858719646799, "grad_norm": 1.8198871417574258, "learning_rate": 1.4946316849937403e-06, "loss": 0.2393, "step": 3381 }, { "epoch": 2.4885945548197204, "grad_norm": 1.738147092622527, "learning_rate": 1.4904557858085967e-06, "loss": 0.252, "step": 3382 }, { "epoch": 2.4893303899926416, "grad_norm": 1.8356603931613176, "learning_rate": 1.4862852585824827e-06, "loss": 0.2483, "step": 3383 }, { "epoch": 2.4900662251655628, "grad_norm": 1.8607149975431514, "learning_rate": 1.482120105948196e-06, "loss": 0.2859, "step": 3384 }, { "epoch": 2.490802060338484, "grad_norm": 1.759207757339279, "learning_rate": 1.4779603305351397e-06, "loss": 0.2433, "step": 3385 }, { "epoch": 2.4915378955114056, "grad_norm": 1.6686086922577186, "learning_rate": 1.4738059349693213e-06, "loss": 0.226, "step": 3386 }, { "epoch": 2.492273730684327, "grad_norm": 1.7085891207406119, "learning_rate": 1.4696569218733558e-06, "loss": 0.2355, "step": 3387 }, { "epoch": 2.493009565857248, "grad_norm": 1.746918460592173, "learning_rate": 1.4655132938664607e-06, "loss": 0.2358, "step": 3388 }, { "epoch": 2.493745401030169, "grad_norm": 1.8101650383372112, "learning_rate": 1.4613750535644445e-06, "loss": 0.2875, "step": 3389 }, { "epoch": 2.4944812362030904, "grad_norm": 1.790877126895719, "learning_rate": 1.4572422035797306e-06, "loss": 0.2487, "step": 3390 }, { "epoch": 2.495217071376012, "grad_norm": 1.6391345849467351, "learning_rate": 1.4531147465213247e-06, "loss": 0.217, "step": 3391 }, { "epoch": 2.495952906548933, "grad_norm": 1.7828338000028152, "learning_rate": 1.4489926849948355e-06, "loss": 0.2463, "step": 3392 }, { "epoch": 2.4966887417218544, "grad_norm": 1.7528228116480327, "learning_rate": 1.444876021602465e-06, "loss": 0.2475, "step": 3393 }, { "epoch": 2.4974245768947756, "grad_norm": 1.4750278537652857, "learning_rate": 1.4407647589430084e-06, "loss": 0.2225, "step": 3394 }, { "epoch": 2.498160412067697, "grad_norm": 1.618005196853689, "learning_rate": 1.4366588996118458e-06, "loss": 0.2333, "step": 3395 }, { "epoch": 2.498896247240618, "grad_norm": 1.7336839732269556, "learning_rate": 1.4325584462009512e-06, "loss": 0.2508, "step": 3396 }, { "epoch": 2.499632082413539, "grad_norm": 1.7112504921051637, "learning_rate": 1.4284634012988886e-06, "loss": 0.2425, "step": 3397 }, { "epoch": 2.500367917586461, "grad_norm": 1.6662794569858832, "learning_rate": 1.4243737674907953e-06, "loss": 0.2412, "step": 3398 }, { "epoch": 2.501103752759382, "grad_norm": 1.8231450931747675, "learning_rate": 1.420289547358411e-06, "loss": 0.2531, "step": 3399 }, { "epoch": 2.501839587932303, "grad_norm": 1.7156471336063879, "learning_rate": 1.4162107434800422e-06, "loss": 0.2614, "step": 3400 }, { "epoch": 2.5025754231052244, "grad_norm": 1.7730356736266544, "learning_rate": 1.412137358430583e-06, "loss": 0.2436, "step": 3401 }, { "epoch": 2.5033112582781456, "grad_norm": 1.7654702099426187, "learning_rate": 1.4080693947815093e-06, "loss": 0.2468, "step": 3402 }, { "epoch": 2.504047093451067, "grad_norm": 1.6731649815957215, "learning_rate": 1.4040068551008658e-06, "loss": 0.2513, "step": 3403 }, { "epoch": 2.504782928623988, "grad_norm": 1.7420473483165488, "learning_rate": 1.3999497419532804e-06, "loss": 0.2414, "step": 3404 }, { "epoch": 2.5055187637969096, "grad_norm": 1.6656373448059687, "learning_rate": 1.3958980578999538e-06, "loss": 0.2506, "step": 3405 }, { "epoch": 2.506254598969831, "grad_norm": 1.6400899853119928, "learning_rate": 1.3918518054986607e-06, "loss": 0.2388, "step": 3406 }, { "epoch": 2.506990434142752, "grad_norm": 1.582944315378475, "learning_rate": 1.3878109873037382e-06, "loss": 0.2229, "step": 3407 }, { "epoch": 2.507726269315673, "grad_norm": 1.7961068890394427, "learning_rate": 1.3837756058661078e-06, "loss": 0.2436, "step": 3408 }, { "epoch": 2.5084621044885944, "grad_norm": 1.711412338802369, "learning_rate": 1.3797456637332451e-06, "loss": 0.2471, "step": 3409 }, { "epoch": 2.509197939661516, "grad_norm": 1.7446408820211619, "learning_rate": 1.3757211634492007e-06, "loss": 0.2322, "step": 3410 }, { "epoch": 2.5099337748344372, "grad_norm": 1.8107424947861952, "learning_rate": 1.3717021075545866e-06, "loss": 0.2595, "step": 3411 }, { "epoch": 2.5106696100073584, "grad_norm": 1.7023893701482797, "learning_rate": 1.3676884985865735e-06, "loss": 0.2483, "step": 3412 }, { "epoch": 2.5114054451802796, "grad_norm": 1.7883963555287943, "learning_rate": 1.363680339078902e-06, "loss": 0.2608, "step": 3413 }, { "epoch": 2.512141280353201, "grad_norm": 1.7434221281558833, "learning_rate": 1.3596776315618676e-06, "loss": 0.2507, "step": 3414 }, { "epoch": 2.512877115526122, "grad_norm": 1.6693348323792023, "learning_rate": 1.3556803785623274e-06, "loss": 0.235, "step": 3415 }, { "epoch": 2.5136129506990432, "grad_norm": 1.827378249004401, "learning_rate": 1.3516885826036852e-06, "loss": 0.2494, "step": 3416 }, { "epoch": 2.514348785871965, "grad_norm": 1.8340407764559044, "learning_rate": 1.3477022462059164e-06, "loss": 0.2215, "step": 3417 }, { "epoch": 2.515084621044886, "grad_norm": 1.5866914097953888, "learning_rate": 1.3437213718855347e-06, "loss": 0.218, "step": 3418 }, { "epoch": 2.5158204562178073, "grad_norm": 1.702391103363591, "learning_rate": 1.339745962155613e-06, "loss": 0.2398, "step": 3419 }, { "epoch": 2.5165562913907285, "grad_norm": 1.8377242553221806, "learning_rate": 1.3357760195257764e-06, "loss": 0.2534, "step": 3420 }, { "epoch": 2.5172921265636496, "grad_norm": 1.6865820524963764, "learning_rate": 1.3318115465021896e-06, "loss": 0.236, "step": 3421 }, { "epoch": 2.5180279617365713, "grad_norm": 1.7943735597145694, "learning_rate": 1.3278525455875736e-06, "loss": 0.2566, "step": 3422 }, { "epoch": 2.518763796909492, "grad_norm": 1.8269555614249309, "learning_rate": 1.3238990192811917e-06, "loss": 0.2388, "step": 3423 }, { "epoch": 2.5194996320824137, "grad_norm": 1.7891109342889397, "learning_rate": 1.3199509700788527e-06, "loss": 0.2495, "step": 3424 }, { "epoch": 2.520235467255335, "grad_norm": 1.6977763137306137, "learning_rate": 1.316008400472899e-06, "loss": 0.2229, "step": 3425 }, { "epoch": 2.520971302428256, "grad_norm": 1.9994579539040906, "learning_rate": 1.312071312952231e-06, "loss": 0.2559, "step": 3426 }, { "epoch": 2.5217071376011773, "grad_norm": 1.6976166220924989, "learning_rate": 1.3081397100022718e-06, "loss": 0.2365, "step": 3427 }, { "epoch": 2.5224429727740985, "grad_norm": 1.9035634557320873, "learning_rate": 1.3042135941049894e-06, "loss": 0.2678, "step": 3428 }, { "epoch": 2.52317880794702, "grad_norm": 1.8151203294040503, "learning_rate": 1.3002929677388898e-06, "loss": 0.254, "step": 3429 }, { "epoch": 2.5239146431199413, "grad_norm": 1.801281064708721, "learning_rate": 1.2963778333790067e-06, "loss": 0.2851, "step": 3430 }, { "epoch": 2.5246504782928625, "grad_norm": 1.8176148029595072, "learning_rate": 1.292468193496914e-06, "loss": 0.2425, "step": 3431 }, { "epoch": 2.5253863134657837, "grad_norm": 1.6655436469392781, "learning_rate": 1.2885640505607122e-06, "loss": 0.2462, "step": 3432 }, { "epoch": 2.526122148638705, "grad_norm": 1.6419291803712979, "learning_rate": 1.2846654070350372e-06, "loss": 0.2229, "step": 3433 }, { "epoch": 2.526857983811626, "grad_norm": 1.6886899319983077, "learning_rate": 1.2807722653810428e-06, "loss": 0.2415, "step": 3434 }, { "epoch": 2.5275938189845473, "grad_norm": 1.8228434795182042, "learning_rate": 1.276884628056423e-06, "loss": 0.2539, "step": 3435 }, { "epoch": 2.528329654157469, "grad_norm": 1.6439433142139264, "learning_rate": 1.2730024975153854e-06, "loss": 0.1988, "step": 3436 }, { "epoch": 2.52906548933039, "grad_norm": 1.716548855333449, "learning_rate": 1.2691258762086677e-06, "loss": 0.2848, "step": 3437 }, { "epoch": 2.5298013245033113, "grad_norm": 1.7727817774463785, "learning_rate": 1.2652547665835301e-06, "loss": 0.2407, "step": 3438 }, { "epoch": 2.5305371596762325, "grad_norm": 1.7698048103923991, "learning_rate": 1.2613891710837467e-06, "loss": 0.2595, "step": 3439 }, { "epoch": 2.5312729948491537, "grad_norm": 1.629096781110587, "learning_rate": 1.2575290921496185e-06, "loss": 0.2601, "step": 3440 }, { "epoch": 2.5320088300220753, "grad_norm": 1.7275205841702712, "learning_rate": 1.2536745322179588e-06, "loss": 0.2364, "step": 3441 }, { "epoch": 2.532744665194996, "grad_norm": 1.9111824822922927, "learning_rate": 1.249825493722101e-06, "loss": 0.2667, "step": 3442 }, { "epoch": 2.5334805003679177, "grad_norm": 1.704777768085588, "learning_rate": 1.2459819790918859e-06, "loss": 0.2149, "step": 3443 }, { "epoch": 2.534216335540839, "grad_norm": 1.742959344675288, "learning_rate": 1.2421439907536781e-06, "loss": 0.2482, "step": 3444 }, { "epoch": 2.53495217071376, "grad_norm": 1.6346644506805557, "learning_rate": 1.2383115311303417e-06, "loss": 0.2311, "step": 3445 }, { "epoch": 2.5356880058866813, "grad_norm": 1.7546368701777955, "learning_rate": 1.2344846026412593e-06, "loss": 0.2617, "step": 3446 }, { "epoch": 2.5364238410596025, "grad_norm": 1.7659462690205396, "learning_rate": 1.2306632077023195e-06, "loss": 0.2555, "step": 3447 }, { "epoch": 2.537159676232524, "grad_norm": 1.8275505358940825, "learning_rate": 1.2268473487259124e-06, "loss": 0.25, "step": 3448 }, { "epoch": 2.5378955114054453, "grad_norm": 1.7832522179029444, "learning_rate": 1.2230370281209403e-06, "loss": 0.2586, "step": 3449 }, { "epoch": 2.5386313465783665, "grad_norm": 1.710486780829863, "learning_rate": 1.2192322482928066e-06, "loss": 0.2474, "step": 3450 }, { "epoch": 2.5393671817512877, "grad_norm": 1.6510735253886435, "learning_rate": 1.2154330116434188e-06, "loss": 0.2279, "step": 3451 }, { "epoch": 2.540103016924209, "grad_norm": 1.7971277252638231, "learning_rate": 1.2116393205711762e-06, "loss": 0.2424, "step": 3452 }, { "epoch": 2.54083885209713, "grad_norm": 1.6457037075334149, "learning_rate": 1.2078511774709932e-06, "loss": 0.2371, "step": 3453 }, { "epoch": 2.5415746872700513, "grad_norm": 1.668364583335112, "learning_rate": 1.204068584734267e-06, "loss": 0.248, "step": 3454 }, { "epoch": 2.542310522442973, "grad_norm": 1.5461779034480625, "learning_rate": 1.2002915447488972e-06, "loss": 0.2354, "step": 3455 }, { "epoch": 2.543046357615894, "grad_norm": 1.6883054967470128, "learning_rate": 1.196520059899282e-06, "loss": 0.2362, "step": 3456 }, { "epoch": 2.5437821927888153, "grad_norm": 1.744050323082429, "learning_rate": 1.1927541325663018e-06, "loss": 0.2502, "step": 3457 }, { "epoch": 2.5445180279617365, "grad_norm": 1.7117640292150291, "learning_rate": 1.1889937651273375e-06, "loss": 0.251, "step": 3458 }, { "epoch": 2.5452538631346577, "grad_norm": 1.752911427595205, "learning_rate": 1.1852389599562596e-06, "loss": 0.2455, "step": 3459 }, { "epoch": 2.5459896983075794, "grad_norm": 2.022543785102748, "learning_rate": 1.1814897194234253e-06, "loss": 0.2304, "step": 3460 }, { "epoch": 2.5467255334805, "grad_norm": 1.8380649786538275, "learning_rate": 1.1777460458956757e-06, "loss": 0.2611, "step": 3461 }, { "epoch": 2.5474613686534218, "grad_norm": 1.732633616221098, "learning_rate": 1.1740079417363438e-06, "loss": 0.2653, "step": 3462 }, { "epoch": 2.548197203826343, "grad_norm": 1.7110837403000378, "learning_rate": 1.1702754093052415e-06, "loss": 0.249, "step": 3463 }, { "epoch": 2.548933038999264, "grad_norm": 1.5864044915275304, "learning_rate": 1.1665484509586666e-06, "loss": 0.2296, "step": 3464 }, { "epoch": 2.5496688741721854, "grad_norm": 1.6513235780166196, "learning_rate": 1.1628270690493993e-06, "loss": 0.2368, "step": 3465 }, { "epoch": 2.5504047093451065, "grad_norm": 1.8085610582375373, "learning_rate": 1.1591112659266934e-06, "loss": 0.2354, "step": 3466 }, { "epoch": 2.551140544518028, "grad_norm": 1.621861597853751, "learning_rate": 1.1554010439362861e-06, "loss": 0.2328, "step": 3467 }, { "epoch": 2.5518763796909494, "grad_norm": 1.7090521755523793, "learning_rate": 1.1516964054203893e-06, "loss": 0.2719, "step": 3468 }, { "epoch": 2.5526122148638706, "grad_norm": 1.7544700270998157, "learning_rate": 1.1479973527176935e-06, "loss": 0.2439, "step": 3469 }, { "epoch": 2.5533480500367918, "grad_norm": 1.757497705901203, "learning_rate": 1.1443038881633561e-06, "loss": 0.2472, "step": 3470 }, { "epoch": 2.554083885209713, "grad_norm": 1.647078643187675, "learning_rate": 1.1406160140890132e-06, "loss": 0.2221, "step": 3471 }, { "epoch": 2.554819720382634, "grad_norm": 1.635437912906774, "learning_rate": 1.136933732822768e-06, "loss": 0.2654, "step": 3472 }, { "epoch": 2.5555555555555554, "grad_norm": 1.7250360481152613, "learning_rate": 1.1332570466891956e-06, "loss": 0.2413, "step": 3473 }, { "epoch": 2.556291390728477, "grad_norm": 1.7335133459110221, "learning_rate": 1.1295859580093394e-06, "loss": 0.2669, "step": 3474 }, { "epoch": 2.557027225901398, "grad_norm": 1.7758638529652733, "learning_rate": 1.125920469100704e-06, "loss": 0.2773, "step": 3475 }, { "epoch": 2.5577630610743194, "grad_norm": 1.7955485749449285, "learning_rate": 1.1222605822772637e-06, "loss": 0.2497, "step": 3476 }, { "epoch": 2.5584988962472406, "grad_norm": 1.70845650204259, "learning_rate": 1.118606299849455e-06, "loss": 0.2489, "step": 3477 }, { "epoch": 2.559234731420162, "grad_norm": 1.7623239395180874, "learning_rate": 1.1149576241241788e-06, "loss": 0.2533, "step": 3478 }, { "epoch": 2.5599705665930834, "grad_norm": 1.6791752940682136, "learning_rate": 1.11131455740479e-06, "loss": 0.2261, "step": 3479 }, { "epoch": 2.560706401766004, "grad_norm": 1.6860500568371508, "learning_rate": 1.1076771019911093e-06, "loss": 0.2422, "step": 3480 }, { "epoch": 2.561442236938926, "grad_norm": 1.6631570777266276, "learning_rate": 1.104045260179415e-06, "loss": 0.2392, "step": 3481 }, { "epoch": 2.562178072111847, "grad_norm": 1.7066016353660567, "learning_rate": 1.100419034262432e-06, "loss": 0.2576, "step": 3482 }, { "epoch": 2.562913907284768, "grad_norm": 1.6276368443356057, "learning_rate": 1.0967984265293562e-06, "loss": 0.2568, "step": 3483 }, { "epoch": 2.5636497424576894, "grad_norm": 1.688299862894194, "learning_rate": 1.0931834392658213e-06, "loss": 0.2718, "step": 3484 }, { "epoch": 2.5643855776306106, "grad_norm": 1.7174822540169516, "learning_rate": 1.0895740747539219e-06, "loss": 0.2475, "step": 3485 }, { "epoch": 2.5651214128035322, "grad_norm": 1.7161550571835344, "learning_rate": 1.0859703352721995e-06, "loss": 0.2342, "step": 3486 }, { "epoch": 2.5658572479764534, "grad_norm": 1.7590758130420316, "learning_rate": 1.082372223095647e-06, "loss": 0.2379, "step": 3487 }, { "epoch": 2.5665930831493746, "grad_norm": 1.740692594815057, "learning_rate": 1.0787797404957046e-06, "loss": 0.2445, "step": 3488 }, { "epoch": 2.567328918322296, "grad_norm": 1.7797454202727634, "learning_rate": 1.075192889740253e-06, "loss": 0.2587, "step": 3489 }, { "epoch": 2.568064753495217, "grad_norm": 1.7157830043499667, "learning_rate": 1.0716116730936254e-06, "loss": 0.2382, "step": 3490 }, { "epoch": 2.568800588668138, "grad_norm": 1.6949239310539568, "learning_rate": 1.0680360928165944e-06, "loss": 0.2278, "step": 3491 }, { "epoch": 2.5695364238410594, "grad_norm": 1.7165453252534164, "learning_rate": 1.0644661511663756e-06, "loss": 0.2758, "step": 3492 }, { "epoch": 2.570272259013981, "grad_norm": 1.6534229919410828, "learning_rate": 1.0609018503966207e-06, "loss": 0.2272, "step": 3493 }, { "epoch": 2.5710080941869022, "grad_norm": 1.7911636202957402, "learning_rate": 1.0573431927574306e-06, "loss": 0.2536, "step": 3494 }, { "epoch": 2.5717439293598234, "grad_norm": 1.7778025221471867, "learning_rate": 1.053790180495332e-06, "loss": 0.2429, "step": 3495 }, { "epoch": 2.5724797645327446, "grad_norm": 1.7357916518177678, "learning_rate": 1.0502428158532952e-06, "loss": 0.2389, "step": 3496 }, { "epoch": 2.573215599705666, "grad_norm": 1.7581564498632236, "learning_rate": 1.046701101070724e-06, "loss": 0.2342, "step": 3497 }, { "epoch": 2.5739514348785875, "grad_norm": 1.7153012571003765, "learning_rate": 1.0431650383834512e-06, "loss": 0.2327, "step": 3498 }, { "epoch": 2.574687270051508, "grad_norm": 1.6385513834749612, "learning_rate": 1.039634630023747e-06, "loss": 0.2262, "step": 3499 }, { "epoch": 2.57542310522443, "grad_norm": 1.8583369322114556, "learning_rate": 1.0361098782203094e-06, "loss": 0.2304, "step": 3500 }, { "epoch": 2.576158940397351, "grad_norm": 1.669562172982025, "learning_rate": 1.0325907851982686e-06, "loss": 0.2361, "step": 3501 }, { "epoch": 2.5768947755702722, "grad_norm": 1.6585497686031854, "learning_rate": 1.0290773531791743e-06, "loss": 0.2363, "step": 3502 }, { "epoch": 2.5776306107431934, "grad_norm": 1.6552497639752566, "learning_rate": 1.0255695843810153e-06, "loss": 0.2232, "step": 3503 }, { "epoch": 2.5783664459161146, "grad_norm": 1.8358599266715137, "learning_rate": 1.022067481018192e-06, "loss": 0.2574, "step": 3504 }, { "epoch": 2.5791022810890363, "grad_norm": 1.5965542799946373, "learning_rate": 1.0185710453015374e-06, "loss": 0.2388, "step": 3505 }, { "epoch": 2.5798381162619575, "grad_norm": 1.7968617262211553, "learning_rate": 1.0150802794383052e-06, "loss": 0.2616, "step": 3506 }, { "epoch": 2.5805739514348787, "grad_norm": 1.5819761412045519, "learning_rate": 1.0115951856321638e-06, "loss": 0.2241, "step": 3507 }, { "epoch": 2.5813097866078, "grad_norm": 1.7780470172050378, "learning_rate": 1.0081157660832086e-06, "loss": 0.2148, "step": 3508 }, { "epoch": 2.582045621780721, "grad_norm": 1.7528798731989228, "learning_rate": 1.0046420229879473e-06, "loss": 0.2349, "step": 3509 }, { "epoch": 2.5827814569536423, "grad_norm": 1.6420348521368329, "learning_rate": 1.00117395853931e-06, "loss": 0.2327, "step": 3510 }, { "epoch": 2.5835172921265634, "grad_norm": 1.6989068676051837, "learning_rate": 9.977115749266331e-07, "loss": 0.2268, "step": 3511 }, { "epoch": 2.584253127299485, "grad_norm": 1.7958992008931811, "learning_rate": 9.942548743356783e-07, "loss": 0.2361, "step": 3512 }, { "epoch": 2.5849889624724063, "grad_norm": 1.721437652138349, "learning_rate": 9.90803858948609e-07, "loss": 0.2707, "step": 3513 }, { "epoch": 2.5857247976453275, "grad_norm": 1.7621584253817384, "learning_rate": 9.87358530944006e-07, "loss": 0.2454, "step": 3514 }, { "epoch": 2.5864606328182487, "grad_norm": 1.7160193206347139, "learning_rate": 9.839188924968578e-07, "loss": 0.273, "step": 3515 }, { "epoch": 2.58719646799117, "grad_norm": 1.729438092017295, "learning_rate": 9.804849457785604e-07, "loss": 0.2629, "step": 3516 }, { "epoch": 2.5879323031640915, "grad_norm": 1.7081467485121768, "learning_rate": 9.77056692956916e-07, "loss": 0.2642, "step": 3517 }, { "epoch": 2.5886681383370123, "grad_norm": 1.761696963835157, "learning_rate": 9.736341361961367e-07, "loss": 0.2481, "step": 3518 }, { "epoch": 2.589403973509934, "grad_norm": 1.7871382160655724, "learning_rate": 9.702172776568363e-07, "loss": 0.2479, "step": 3519 }, { "epoch": 2.590139808682855, "grad_norm": 2.2665582734277185, "learning_rate": 9.668061194960255e-07, "loss": 0.2224, "step": 3520 }, { "epoch": 2.5908756438557763, "grad_norm": 1.6413708996229432, "learning_rate": 9.63400663867129e-07, "loss": 0.2354, "step": 3521 }, { "epoch": 2.5916114790286975, "grad_norm": 1.697872807858783, "learning_rate": 9.600009129199594e-07, "loss": 0.2512, "step": 3522 }, { "epoch": 2.5923473142016187, "grad_norm": 1.7359899914043861, "learning_rate": 9.566068688007346e-07, "loss": 0.2607, "step": 3523 }, { "epoch": 2.5930831493745403, "grad_norm": 1.8108434265075128, "learning_rate": 9.532185336520706e-07, "loss": 0.2656, "step": 3524 }, { "epoch": 2.5938189845474615, "grad_norm": 1.6561951874680505, "learning_rate": 9.498359096129717e-07, "loss": 0.2251, "step": 3525 }, { "epoch": 2.5945548197203827, "grad_norm": 1.774234350706018, "learning_rate": 9.464589988188466e-07, "loss": 0.2326, "step": 3526 }, { "epoch": 2.595290654893304, "grad_norm": 1.7285407285015244, "learning_rate": 9.430878034014912e-07, "loss": 0.2469, "step": 3527 }, { "epoch": 2.596026490066225, "grad_norm": 1.6505141927247093, "learning_rate": 9.397223254890986e-07, "loss": 0.2291, "step": 3528 }, { "epoch": 2.5967623252391463, "grad_norm": 1.6340620205095813, "learning_rate": 9.363625672062427e-07, "loss": 0.2699, "step": 3529 }, { "epoch": 2.5974981604120675, "grad_norm": 1.740620908692019, "learning_rate": 9.330085306738979e-07, "loss": 0.2253, "step": 3530 }, { "epoch": 2.598233995584989, "grad_norm": 1.8602958726346162, "learning_rate": 9.296602180094194e-07, "loss": 0.2314, "step": 3531 }, { "epoch": 2.5989698307579103, "grad_norm": 1.741894575062516, "learning_rate": 9.263176313265521e-07, "loss": 0.2607, "step": 3532 }, { "epoch": 2.5997056659308315, "grad_norm": 1.6731973332881023, "learning_rate": 9.229807727354267e-07, "loss": 0.2408, "step": 3533 }, { "epoch": 2.6004415011037527, "grad_norm": 1.7481410169797778, "learning_rate": 9.19649644342554e-07, "loss": 0.2288, "step": 3534 }, { "epoch": 2.601177336276674, "grad_norm": 1.6841720812965884, "learning_rate": 9.163242482508306e-07, "loss": 0.2393, "step": 3535 }, { "epoch": 2.6019131714495956, "grad_norm": 1.8462062958966798, "learning_rate": 9.130045865595327e-07, "loss": 0.2345, "step": 3536 }, { "epoch": 2.6026490066225163, "grad_norm": 1.6742490240333734, "learning_rate": 9.09690661364322e-07, "loss": 0.2486, "step": 3537 }, { "epoch": 2.603384841795438, "grad_norm": 1.823600078003075, "learning_rate": 9.06382474757228e-07, "loss": 0.2722, "step": 3538 }, { "epoch": 2.604120676968359, "grad_norm": 1.781430731498435, "learning_rate": 9.030800288266673e-07, "loss": 0.2639, "step": 3539 }, { "epoch": 2.6048565121412803, "grad_norm": 1.7633747876485586, "learning_rate": 8.997833256574295e-07, "loss": 0.2453, "step": 3540 }, { "epoch": 2.6055923473142015, "grad_norm": 1.6809968907688126, "learning_rate": 8.964923673306725e-07, "loss": 0.2225, "step": 3541 }, { "epoch": 2.6063281824871227, "grad_norm": 1.7422806453953557, "learning_rate": 8.932071559239419e-07, "loss": 0.2345, "step": 3542 }, { "epoch": 2.6070640176600444, "grad_norm": 1.633567880850147, "learning_rate": 8.899276935111412e-07, "loss": 0.2624, "step": 3543 }, { "epoch": 2.6077998528329656, "grad_norm": 1.7665517018258974, "learning_rate": 8.866539821625519e-07, "loss": 0.2495, "step": 3544 }, { "epoch": 2.6085356880058868, "grad_norm": 1.8073969810792523, "learning_rate": 8.833860239448222e-07, "loss": 0.2779, "step": 3545 }, { "epoch": 2.609271523178808, "grad_norm": 1.7636908994572267, "learning_rate": 8.801238209209717e-07, "loss": 0.2446, "step": 3546 }, { "epoch": 2.610007358351729, "grad_norm": 1.6319600728447423, "learning_rate": 8.76867375150382e-07, "loss": 0.2338, "step": 3547 }, { "epoch": 2.6107431935246503, "grad_norm": 1.7846730809812417, "learning_rate": 8.736166886888042e-07, "loss": 0.2587, "step": 3548 }, { "epoch": 2.6114790286975715, "grad_norm": 1.6156414132711545, "learning_rate": 8.703717635883536e-07, "loss": 0.2274, "step": 3549 }, { "epoch": 2.612214863870493, "grad_norm": 1.6260399028990873, "learning_rate": 8.671326018975024e-07, "loss": 0.2237, "step": 3550 }, { "epoch": 2.6129506990434144, "grad_norm": 1.6980909311067829, "learning_rate": 8.638992056610962e-07, "loss": 0.2435, "step": 3551 }, { "epoch": 2.6136865342163356, "grad_norm": 1.7604440286588718, "learning_rate": 8.606715769203267e-07, "loss": 0.2282, "step": 3552 }, { "epoch": 2.6144223693892568, "grad_norm": 1.652772436041799, "learning_rate": 8.574497177127561e-07, "loss": 0.2426, "step": 3553 }, { "epoch": 2.615158204562178, "grad_norm": 1.6815045001310194, "learning_rate": 8.542336300722998e-07, "loss": 0.2712, "step": 3554 }, { "epoch": 2.6158940397350996, "grad_norm": 1.7837158571944622, "learning_rate": 8.510233160292302e-07, "loss": 0.2575, "step": 3555 }, { "epoch": 2.6166298749080203, "grad_norm": 1.6864460779862458, "learning_rate": 8.47818777610172e-07, "loss": 0.2253, "step": 3556 }, { "epoch": 2.617365710080942, "grad_norm": 1.667162762881755, "learning_rate": 8.44620016838108e-07, "loss": 0.2198, "step": 3557 }, { "epoch": 2.618101545253863, "grad_norm": 1.7781867042074555, "learning_rate": 8.414270357323751e-07, "loss": 0.232, "step": 3558 }, { "epoch": 2.6188373804267844, "grad_norm": 1.6934778257106227, "learning_rate": 8.38239836308653e-07, "loss": 0.2566, "step": 3559 }, { "epoch": 2.6195732155997056, "grad_norm": 1.7828237780587954, "learning_rate": 8.350584205789825e-07, "loss": 0.2629, "step": 3560 }, { "epoch": 2.6203090507726268, "grad_norm": 1.6285513197395665, "learning_rate": 8.318827905517446e-07, "loss": 0.2523, "step": 3561 }, { "epoch": 2.6210448859455484, "grad_norm": 1.7148472929633534, "learning_rate": 8.287129482316725e-07, "loss": 0.2351, "step": 3562 }, { "epoch": 2.6217807211184696, "grad_norm": 1.8248595346377154, "learning_rate": 8.255488956198443e-07, "loss": 0.2539, "step": 3563 }, { "epoch": 2.622516556291391, "grad_norm": 1.7959588618134783, "learning_rate": 8.223906347136835e-07, "loss": 0.2399, "step": 3564 }, { "epoch": 2.623252391464312, "grad_norm": 1.6677944468996406, "learning_rate": 8.192381675069561e-07, "loss": 0.2656, "step": 3565 }, { "epoch": 2.623988226637233, "grad_norm": 1.6915831758985198, "learning_rate": 8.160914959897714e-07, "loss": 0.2692, "step": 3566 }, { "epoch": 2.6247240618101544, "grad_norm": 1.6938606612308869, "learning_rate": 8.129506221485828e-07, "loss": 0.2586, "step": 3567 }, { "epoch": 2.6254598969830756, "grad_norm": 1.6378173016296875, "learning_rate": 8.098155479661751e-07, "loss": 0.2512, "step": 3568 }, { "epoch": 2.626195732155997, "grad_norm": 1.7903417304874525, "learning_rate": 8.066862754216853e-07, "loss": 0.2286, "step": 3569 }, { "epoch": 2.6269315673289184, "grad_norm": 1.6270250899449246, "learning_rate": 8.035628064905754e-07, "loss": 0.2526, "step": 3570 }, { "epoch": 2.6276674025018396, "grad_norm": 1.760850204762114, "learning_rate": 8.004451431446503e-07, "loss": 0.2468, "step": 3571 }, { "epoch": 2.628403237674761, "grad_norm": 1.6345146143253098, "learning_rate": 7.973332873520479e-07, "loss": 0.2424, "step": 3572 }, { "epoch": 2.629139072847682, "grad_norm": 1.7152667976425173, "learning_rate": 7.942272410772423e-07, "loss": 0.2453, "step": 3573 }, { "epoch": 2.6298749080206036, "grad_norm": 1.7757938963338793, "learning_rate": 7.911270062810338e-07, "loss": 0.28, "step": 3574 }, { "epoch": 2.6306107431935244, "grad_norm": 1.7155923335447645, "learning_rate": 7.880325849205617e-07, "loss": 0.2381, "step": 3575 }, { "epoch": 2.631346578366446, "grad_norm": 1.7231540953307327, "learning_rate": 7.849439789492918e-07, "loss": 0.2442, "step": 3576 }, { "epoch": 2.6320824135393672, "grad_norm": 1.7417300346505762, "learning_rate": 7.818611903170159e-07, "loss": 0.2503, "step": 3577 }, { "epoch": 2.6328182487122884, "grad_norm": 1.6477016665310134, "learning_rate": 7.787842209698615e-07, "loss": 0.2664, "step": 3578 }, { "epoch": 2.6335540838852096, "grad_norm": 1.7339223263940868, "learning_rate": 7.757130728502715e-07, "loss": 0.2313, "step": 3579 }, { "epoch": 2.634289919058131, "grad_norm": 1.7433588725869087, "learning_rate": 7.72647747897024e-07, "loss": 0.2372, "step": 3580 }, { "epoch": 2.6350257542310525, "grad_norm": 1.6704585803897807, "learning_rate": 7.695882480452143e-07, "loss": 0.2726, "step": 3581 }, { "epoch": 2.6357615894039736, "grad_norm": 1.875096624648085, "learning_rate": 7.665345752262665e-07, "loss": 0.2634, "step": 3582 }, { "epoch": 2.636497424576895, "grad_norm": 1.675165860936199, "learning_rate": 7.634867313679172e-07, "loss": 0.2395, "step": 3583 }, { "epoch": 2.637233259749816, "grad_norm": 1.8483410261736228, "learning_rate": 7.604447183942321e-07, "loss": 0.2923, "step": 3584 }, { "epoch": 2.6379690949227372, "grad_norm": 1.9060034541429833, "learning_rate": 7.574085382255936e-07, "loss": 0.2458, "step": 3585 }, { "epoch": 2.6387049300956584, "grad_norm": 1.563536047748598, "learning_rate": 7.543781927786953e-07, "loss": 0.1886, "step": 3586 }, { "epoch": 2.6394407652685796, "grad_norm": 1.684916289202401, "learning_rate": 7.513536839665602e-07, "loss": 0.2565, "step": 3587 }, { "epoch": 2.6401766004415013, "grad_norm": 1.7660008283862516, "learning_rate": 7.483350136985135e-07, "loss": 0.2617, "step": 3588 }, { "epoch": 2.6409124356144225, "grad_norm": 1.636262102646573, "learning_rate": 7.453221838802027e-07, "loss": 0.233, "step": 3589 }, { "epoch": 2.6416482707873437, "grad_norm": 1.5629948984752182, "learning_rate": 7.423151964135855e-07, "loss": 0.2475, "step": 3590 }, { "epoch": 2.642384105960265, "grad_norm": 1.7148037057003085, "learning_rate": 7.393140531969345e-07, "loss": 0.2292, "step": 3591 }, { "epoch": 2.643119941133186, "grad_norm": 1.8107938751920416, "learning_rate": 7.363187561248275e-07, "loss": 0.2689, "step": 3592 }, { "epoch": 2.6438557763061077, "grad_norm": 1.608019278755651, "learning_rate": 7.333293070881553e-07, "loss": 0.2225, "step": 3593 }, { "epoch": 2.6445916114790284, "grad_norm": 1.672692022606064, "learning_rate": 7.303457079741183e-07, "loss": 0.2234, "step": 3594 }, { "epoch": 2.64532744665195, "grad_norm": 1.735731340586173, "learning_rate": 7.273679606662166e-07, "loss": 0.2674, "step": 3595 }, { "epoch": 2.6460632818248713, "grad_norm": 1.9162806198401496, "learning_rate": 7.243960670442685e-07, "loss": 0.2648, "step": 3596 }, { "epoch": 2.6467991169977925, "grad_norm": 1.7407594558054649, "learning_rate": 7.214300289843845e-07, "loss": 0.2745, "step": 3597 }, { "epoch": 2.6475349521707137, "grad_norm": 1.633274610238279, "learning_rate": 7.184698483589858e-07, "loss": 0.2267, "step": 3598 }, { "epoch": 2.648270787343635, "grad_norm": 1.6901819277951884, "learning_rate": 7.155155270367931e-07, "loss": 0.2392, "step": 3599 }, { "epoch": 2.6490066225165565, "grad_norm": 1.7225903099096995, "learning_rate": 7.125670668828332e-07, "loss": 0.2288, "step": 3600 }, { "epoch": 2.6497424576894777, "grad_norm": 1.6802752760091644, "learning_rate": 7.096244697584221e-07, "loss": 0.2443, "step": 3601 }, { "epoch": 2.650478292862399, "grad_norm": 1.7432566369794602, "learning_rate": 7.066877375211855e-07, "loss": 0.2347, "step": 3602 }, { "epoch": 2.65121412803532, "grad_norm": 1.6497560262480746, "learning_rate": 7.037568720250431e-07, "loss": 0.2584, "step": 3603 }, { "epoch": 2.6519499632082413, "grad_norm": 1.8086860165335343, "learning_rate": 7.008318751202048e-07, "loss": 0.2324, "step": 3604 }, { "epoch": 2.6526857983811625, "grad_norm": 1.7556735585651093, "learning_rate": 6.979127486531878e-07, "loss": 0.2523, "step": 3605 }, { "epoch": 2.6534216335540837, "grad_norm": 1.6661798128603398, "learning_rate": 6.949994944667926e-07, "loss": 0.2355, "step": 3606 }, { "epoch": 2.6541574687270053, "grad_norm": 1.746572698501227, "learning_rate": 6.92092114400118e-07, "loss": 0.2779, "step": 3607 }, { "epoch": 2.6548933038999265, "grad_norm": 1.7700421179440646, "learning_rate": 6.891906102885559e-07, "loss": 0.2479, "step": 3608 }, { "epoch": 2.6556291390728477, "grad_norm": 1.71384059505109, "learning_rate": 6.862949839637811e-07, "loss": 0.2509, "step": 3609 }, { "epoch": 2.656364974245769, "grad_norm": 1.7211717187607545, "learning_rate": 6.834052372537658e-07, "loss": 0.2272, "step": 3610 }, { "epoch": 2.65710080941869, "grad_norm": 1.5626432885654995, "learning_rate": 6.805213719827664e-07, "loss": 0.2313, "step": 3611 }, { "epoch": 2.6578366445916117, "grad_norm": 1.7658238557985708, "learning_rate": 6.776433899713319e-07, "loss": 0.2399, "step": 3612 }, { "epoch": 2.6585724797645325, "grad_norm": 1.6808474441723875, "learning_rate": 6.747712930362848e-07, "loss": 0.2318, "step": 3613 }, { "epoch": 2.659308314937454, "grad_norm": 1.7104013282410016, "learning_rate": 6.719050829907503e-07, "loss": 0.2351, "step": 3614 }, { "epoch": 2.6600441501103753, "grad_norm": 1.7443940022531135, "learning_rate": 6.690447616441209e-07, "loss": 0.2429, "step": 3615 }, { "epoch": 2.6607799852832965, "grad_norm": 1.7085463783220276, "learning_rate": 6.661903308020801e-07, "loss": 0.2098, "step": 3616 }, { "epoch": 2.6615158204562177, "grad_norm": 1.9058197736526132, "learning_rate": 6.63341792266593e-07, "loss": 0.2326, "step": 3617 }, { "epoch": 2.662251655629139, "grad_norm": 1.6792689585671727, "learning_rate": 6.60499147835898e-07, "loss": 0.2596, "step": 3618 }, { "epoch": 2.6629874908020605, "grad_norm": 1.7537737251477834, "learning_rate": 6.57662399304525e-07, "loss": 0.2404, "step": 3619 }, { "epoch": 2.6637233259749817, "grad_norm": 1.7660822325633982, "learning_rate": 6.548315484632695e-07, "loss": 0.2515, "step": 3620 }, { "epoch": 2.664459161147903, "grad_norm": 1.7432804430292619, "learning_rate": 6.520065970992095e-07, "loss": 0.2551, "step": 3621 }, { "epoch": 2.665194996320824, "grad_norm": 1.6615462600872102, "learning_rate": 6.491875469956998e-07, "loss": 0.2482, "step": 3622 }, { "epoch": 2.6659308314937453, "grad_norm": 1.8702801022619462, "learning_rate": 6.46374399932369e-07, "loss": 0.2719, "step": 3623 }, { "epoch": 2.6666666666666665, "grad_norm": 1.6042668328554512, "learning_rate": 6.435671576851144e-07, "loss": 0.2385, "step": 3624 }, { "epoch": 2.6674025018395877, "grad_norm": 1.69341865091164, "learning_rate": 6.407658220261126e-07, "loss": 0.2535, "step": 3625 }, { "epoch": 2.6681383370125094, "grad_norm": 1.7893440790345283, "learning_rate": 6.379703947238092e-07, "loss": 0.2523, "step": 3626 }, { "epoch": 2.6688741721854305, "grad_norm": 1.5732369239206023, "learning_rate": 6.351808775429158e-07, "loss": 0.2464, "step": 3627 }, { "epoch": 2.6696100073583517, "grad_norm": 1.664058116902164, "learning_rate": 6.323972722444215e-07, "loss": 0.2369, "step": 3628 }, { "epoch": 2.670345842531273, "grad_norm": 1.6623564681572283, "learning_rate": 6.296195805855743e-07, "loss": 0.23, "step": 3629 }, { "epoch": 2.671081677704194, "grad_norm": 1.7129560372163681, "learning_rate": 6.268478043198944e-07, "loss": 0.2596, "step": 3630 }, { "epoch": 2.6718175128771158, "grad_norm": 1.6640922810555203, "learning_rate": 6.240819451971658e-07, "loss": 0.2287, "step": 3631 }, { "epoch": 2.6725533480500365, "grad_norm": 1.692991177830461, "learning_rate": 6.213220049634394e-07, "loss": 0.2208, "step": 3632 }, { "epoch": 2.673289183222958, "grad_norm": 1.6060975434576932, "learning_rate": 6.185679853610249e-07, "loss": 0.2115, "step": 3633 }, { "epoch": 2.6740250183958794, "grad_norm": 1.756479951784886, "learning_rate": 6.158198881284994e-07, "loss": 0.2222, "step": 3634 }, { "epoch": 2.6747608535688006, "grad_norm": 1.728600783017709, "learning_rate": 6.130777150006995e-07, "loss": 0.248, "step": 3635 }, { "epoch": 2.6754966887417218, "grad_norm": 1.6448651445096047, "learning_rate": 6.103414677087183e-07, "loss": 0.242, "step": 3636 }, { "epoch": 2.676232523914643, "grad_norm": 1.7204014912605838, "learning_rate": 6.076111479799162e-07, "loss": 0.2533, "step": 3637 }, { "epoch": 2.6769683590875646, "grad_norm": 1.7393642652790697, "learning_rate": 6.048867575379025e-07, "loss": 0.2596, "step": 3638 }, { "epoch": 2.677704194260486, "grad_norm": 1.7533394658909294, "learning_rate": 6.021682981025489e-07, "loss": 0.2872, "step": 3639 }, { "epoch": 2.678440029433407, "grad_norm": 1.6603196409225496, "learning_rate": 5.994557713899829e-07, "loss": 0.2276, "step": 3640 }, { "epoch": 2.679175864606328, "grad_norm": 1.7451788285633048, "learning_rate": 5.967491791125868e-07, "loss": 0.2732, "step": 3641 }, { "epoch": 2.6799116997792494, "grad_norm": 1.699755999271252, "learning_rate": 5.940485229789917e-07, "loss": 0.2591, "step": 3642 }, { "epoch": 2.6806475349521706, "grad_norm": 1.7850447918714751, "learning_rate": 5.913538046940859e-07, "loss": 0.2758, "step": 3643 }, { "epoch": 2.6813833701250918, "grad_norm": 1.7777883822302472, "learning_rate": 5.88665025959011e-07, "loss": 0.2699, "step": 3644 }, { "epoch": 2.6821192052980134, "grad_norm": 1.7015795328606504, "learning_rate": 5.859821884711514e-07, "loss": 0.2165, "step": 3645 }, { "epoch": 2.6828550404709346, "grad_norm": 1.8614859390100824, "learning_rate": 5.833052939241513e-07, "loss": 0.2377, "step": 3646 }, { "epoch": 2.683590875643856, "grad_norm": 1.7233668228555226, "learning_rate": 5.806343440078932e-07, "loss": 0.2101, "step": 3647 }, { "epoch": 2.684326710816777, "grad_norm": 1.7357894251768138, "learning_rate": 5.77969340408514e-07, "loss": 0.2387, "step": 3648 }, { "epoch": 2.685062545989698, "grad_norm": 1.6185190209840916, "learning_rate": 5.753102848083924e-07, "loss": 0.2149, "step": 3649 }, { "epoch": 2.68579838116262, "grad_norm": 1.8156651825621444, "learning_rate": 5.726571788861568e-07, "loss": 0.28, "step": 3650 }, { "epoch": 2.6865342163355406, "grad_norm": 1.641521493675239, "learning_rate": 5.700100243166729e-07, "loss": 0.265, "step": 3651 }, { "epoch": 2.687270051508462, "grad_norm": 1.727924902110189, "learning_rate": 5.673688227710539e-07, "loss": 0.2336, "step": 3652 }, { "epoch": 2.6880058866813834, "grad_norm": 1.6921563711320895, "learning_rate": 5.647335759166573e-07, "loss": 0.2423, "step": 3653 }, { "epoch": 2.6887417218543046, "grad_norm": 1.7220010823845595, "learning_rate": 5.621042854170744e-07, "loss": 0.2464, "step": 3654 }, { "epoch": 2.689477557027226, "grad_norm": 1.7956751957689434, "learning_rate": 5.594809529321443e-07, "loss": 0.23, "step": 3655 }, { "epoch": 2.690213392200147, "grad_norm": 1.6281529777180455, "learning_rate": 5.568635801179378e-07, "loss": 0.2339, "step": 3656 }, { "epoch": 2.6909492273730686, "grad_norm": 1.7983049466370802, "learning_rate": 5.542521686267677e-07, "loss": 0.2596, "step": 3657 }, { "epoch": 2.69168506254599, "grad_norm": 1.7113824698698306, "learning_rate": 5.516467201071829e-07, "loss": 0.23, "step": 3658 }, { "epoch": 2.692420897718911, "grad_norm": 1.6607457860220147, "learning_rate": 5.490472362039679e-07, "loss": 0.2451, "step": 3659 }, { "epoch": 2.693156732891832, "grad_norm": 1.8711570744522312, "learning_rate": 5.464537185581398e-07, "loss": 0.2565, "step": 3660 }, { "epoch": 2.6938925680647534, "grad_norm": 1.8271533524412866, "learning_rate": 5.438661688069513e-07, "loss": 0.2492, "step": 3661 }, { "epoch": 2.6946284032376746, "grad_norm": 1.6660047518210401, "learning_rate": 5.412845885838891e-07, "loss": 0.2632, "step": 3662 }, { "epoch": 2.695364238410596, "grad_norm": 1.6821851024228875, "learning_rate": 5.38708979518665e-07, "loss": 0.252, "step": 3663 }, { "epoch": 2.6961000735835174, "grad_norm": 1.6447309820882021, "learning_rate": 5.361393432372319e-07, "loss": 0.2476, "step": 3664 }, { "epoch": 2.6968359087564386, "grad_norm": 1.7744165180311415, "learning_rate": 5.335756813617621e-07, "loss": 0.2392, "step": 3665 }, { "epoch": 2.69757174392936, "grad_norm": 1.8574181263503295, "learning_rate": 5.310179955106598e-07, "loss": 0.26, "step": 3666 }, { "epoch": 2.698307579102281, "grad_norm": 1.7028625447717498, "learning_rate": 5.284662872985602e-07, "loss": 0.2157, "step": 3667 }, { "epoch": 2.6990434142752022, "grad_norm": 1.7536896773632082, "learning_rate": 5.259205583363203e-07, "loss": 0.2478, "step": 3668 }, { "epoch": 2.699779249448124, "grad_norm": 1.6987492324581277, "learning_rate": 5.233808102310234e-07, "loss": 0.2583, "step": 3669 }, { "epoch": 2.7005150846210446, "grad_norm": 1.7215407092159842, "learning_rate": 5.208470445859782e-07, "loss": 0.2562, "step": 3670 }, { "epoch": 2.7012509197939663, "grad_norm": 1.6693060548761272, "learning_rate": 5.183192630007172e-07, "loss": 0.254, "step": 3671 }, { "epoch": 2.7019867549668874, "grad_norm": 1.7698609065468067, "learning_rate": 5.15797467070992e-07, "loss": 0.2739, "step": 3672 }, { "epoch": 2.7027225901398086, "grad_norm": 1.757347738354259, "learning_rate": 5.132816583887812e-07, "loss": 0.2582, "step": 3673 }, { "epoch": 2.70345842531273, "grad_norm": 1.6596860781432068, "learning_rate": 5.107718385422778e-07, "loss": 0.2397, "step": 3674 }, { "epoch": 2.704194260485651, "grad_norm": 1.6813230142360205, "learning_rate": 5.082680091158987e-07, "loss": 0.2517, "step": 3675 }, { "epoch": 2.7049300956585727, "grad_norm": 1.712051535881859, "learning_rate": 5.057701716902764e-07, "loss": 0.2588, "step": 3676 }, { "epoch": 2.705665930831494, "grad_norm": 1.6423474893907533, "learning_rate": 5.032783278422604e-07, "loss": 0.228, "step": 3677 }, { "epoch": 2.706401766004415, "grad_norm": 1.6342266877643448, "learning_rate": 5.007924791449193e-07, "loss": 0.2505, "step": 3678 }, { "epoch": 2.7071376011773363, "grad_norm": 1.7986514295512916, "learning_rate": 4.983126271675354e-07, "loss": 0.2527, "step": 3679 }, { "epoch": 2.7078734363502575, "grad_norm": 1.6965838104070459, "learning_rate": 4.958387734756054e-07, "loss": 0.2469, "step": 3680 }, { "epoch": 2.7086092715231787, "grad_norm": 1.7404297701319238, "learning_rate": 4.933709196308356e-07, "loss": 0.2329, "step": 3681 }, { "epoch": 2.7093451066961, "grad_norm": 1.6394419743825923, "learning_rate": 4.909090671911554e-07, "loss": 0.2416, "step": 3682 }, { "epoch": 2.7100809418690215, "grad_norm": 1.8164884991653758, "learning_rate": 4.884532177106927e-07, "loss": 0.2448, "step": 3683 }, { "epoch": 2.7108167770419427, "grad_norm": 1.729026902117642, "learning_rate": 4.86003372739795e-07, "loss": 0.2515, "step": 3684 }, { "epoch": 2.711552612214864, "grad_norm": 1.7275074750049675, "learning_rate": 4.835595338250155e-07, "loss": 0.228, "step": 3685 }, { "epoch": 2.712288447387785, "grad_norm": 1.8210947066566487, "learning_rate": 4.811217025091141e-07, "loss": 0.2677, "step": 3686 }, { "epoch": 2.7130242825607063, "grad_norm": 1.8610044102290322, "learning_rate": 4.786898803310624e-07, "loss": 0.272, "step": 3687 }, { "epoch": 2.713760117733628, "grad_norm": 1.747349543212748, "learning_rate": 4.762640688260356e-07, "loss": 0.2549, "step": 3688 }, { "epoch": 2.7144959529065487, "grad_norm": 1.7481944169892236, "learning_rate": 4.738442695254175e-07, "loss": 0.2513, "step": 3689 }, { "epoch": 2.7152317880794703, "grad_norm": 1.6777000036963645, "learning_rate": 4.714304839567907e-07, "loss": 0.233, "step": 3690 }, { "epoch": 2.7159676232523915, "grad_norm": 1.6779488948917345, "learning_rate": 4.690227136439496e-07, "loss": 0.2752, "step": 3691 }, { "epoch": 2.7167034584253127, "grad_norm": 1.8204323656082873, "learning_rate": 4.666209601068827e-07, "loss": 0.2539, "step": 3692 }, { "epoch": 2.717439293598234, "grad_norm": 1.826138790060875, "learning_rate": 4.642252248617851e-07, "loss": 0.256, "step": 3693 }, { "epoch": 2.718175128771155, "grad_norm": 1.7654919792573203, "learning_rate": 4.618355094210547e-07, "loss": 0.2361, "step": 3694 }, { "epoch": 2.7189109639440767, "grad_norm": 1.8295538324894542, "learning_rate": 4.594518152932814e-07, "loss": 0.2505, "step": 3695 }, { "epoch": 2.719646799116998, "grad_norm": 1.7959884057125162, "learning_rate": 4.570741439832627e-07, "loss": 0.256, "step": 3696 }, { "epoch": 2.720382634289919, "grad_norm": 1.803395833601212, "learning_rate": 4.5470249699198667e-07, "loss": 0.2554, "step": 3697 }, { "epoch": 2.7211184694628403, "grad_norm": 1.743520353166018, "learning_rate": 4.523368758166458e-07, "loss": 0.2347, "step": 3698 }, { "epoch": 2.7218543046357615, "grad_norm": 1.792788839009114, "learning_rate": 4.4997728195061983e-07, "loss": 0.2534, "step": 3699 }, { "epoch": 2.7225901398086827, "grad_norm": 1.792657168217236, "learning_rate": 4.476237168834929e-07, "loss": 0.2659, "step": 3700 }, { "epoch": 2.723325974981604, "grad_norm": 1.5992640311995765, "learning_rate": 4.452761821010343e-07, "loss": 0.2349, "step": 3701 }, { "epoch": 2.7240618101545255, "grad_norm": 1.881021978950351, "learning_rate": 4.4293467908521293e-07, "loss": 0.2691, "step": 3702 }, { "epoch": 2.7247976453274467, "grad_norm": 1.706138284954923, "learning_rate": 4.4059920931418866e-07, "loss": 0.2685, "step": 3703 }, { "epoch": 2.725533480500368, "grad_norm": 1.6411484442315605, "learning_rate": 4.382697742623088e-07, "loss": 0.2326, "step": 3704 }, { "epoch": 2.726269315673289, "grad_norm": 1.6646086066970653, "learning_rate": 4.3594637540011477e-07, "loss": 0.2436, "step": 3705 }, { "epoch": 2.7270051508462103, "grad_norm": 1.7052806824599394, "learning_rate": 4.336290141943367e-07, "loss": 0.2423, "step": 3706 }, { "epoch": 2.727740986019132, "grad_norm": 1.6321292124201308, "learning_rate": 4.3131769210789545e-07, "loss": 0.2306, "step": 3707 }, { "epoch": 2.7284768211920527, "grad_norm": 1.5767463580116345, "learning_rate": 4.290124105998916e-07, "loss": 0.2302, "step": 3708 }, { "epoch": 2.7292126563649743, "grad_norm": 1.6655928622345058, "learning_rate": 4.267131711256245e-07, "loss": 0.2391, "step": 3709 }, { "epoch": 2.7299484915378955, "grad_norm": 1.744343153833518, "learning_rate": 4.244199751365685e-07, "loss": 0.2426, "step": 3710 }, { "epoch": 2.7306843267108167, "grad_norm": 1.7623861661887792, "learning_rate": 4.221328240803868e-07, "loss": 0.2381, "step": 3711 }, { "epoch": 2.731420161883738, "grad_norm": 1.7062503033345042, "learning_rate": 4.1985171940092884e-07, "loss": 0.2514, "step": 3712 }, { "epoch": 2.732155997056659, "grad_norm": 1.672376550016203, "learning_rate": 4.175766625382227e-07, "loss": 0.2327, "step": 3713 }, { "epoch": 2.7328918322295808, "grad_norm": 1.6920938347178687, "learning_rate": 4.153076549284818e-07, "loss": 0.2518, "step": 3714 }, { "epoch": 2.733627667402502, "grad_norm": 1.7751234236375826, "learning_rate": 4.130446980041003e-07, "loss": 0.2593, "step": 3715 }, { "epoch": 2.734363502575423, "grad_norm": 1.5593651943573448, "learning_rate": 4.1078779319365103e-07, "loss": 0.2328, "step": 3716 }, { "epoch": 2.7350993377483444, "grad_norm": 1.768691085370638, "learning_rate": 4.085369419218854e-07, "loss": 0.236, "step": 3717 }, { "epoch": 2.7358351729212655, "grad_norm": 1.8294261423244818, "learning_rate": 4.0629214560973907e-07, "loss": 0.2747, "step": 3718 }, { "epoch": 2.7365710080941867, "grad_norm": 1.9432579954660227, "learning_rate": 4.040534056743184e-07, "loss": 0.2402, "step": 3719 }, { "epoch": 2.737306843267108, "grad_norm": 1.7503692724288618, "learning_rate": 4.018207235289107e-07, "loss": 0.241, "step": 3720 }, { "epoch": 2.7380426784400296, "grad_norm": 1.8138802180242528, "learning_rate": 3.995941005829773e-07, "loss": 0.2603, "step": 3721 }, { "epoch": 2.7387785136129508, "grad_norm": 1.6795467120054093, "learning_rate": 3.9737353824215395e-07, "loss": 0.2533, "step": 3722 }, { "epoch": 2.739514348785872, "grad_norm": 1.6156832694062135, "learning_rate": 3.951590379082515e-07, "loss": 0.2622, "step": 3723 }, { "epoch": 2.740250183958793, "grad_norm": 1.7552383077285774, "learning_rate": 3.92950600979255e-07, "loss": 0.2752, "step": 3724 }, { "epoch": 2.7409860191317144, "grad_norm": 1.820045324381553, "learning_rate": 3.907482288493214e-07, "loss": 0.2345, "step": 3725 }, { "epoch": 2.741721854304636, "grad_norm": 1.6661055419765265, "learning_rate": 3.885519229087731e-07, "loss": 0.2318, "step": 3726 }, { "epoch": 2.742457689477557, "grad_norm": 1.566975754651405, "learning_rate": 3.863616845441154e-07, "loss": 0.2313, "step": 3727 }, { "epoch": 2.7431935246504784, "grad_norm": 1.7183789562541831, "learning_rate": 3.8417751513801007e-07, "loss": 0.243, "step": 3728 }, { "epoch": 2.7439293598233996, "grad_norm": 2.0474033775625924, "learning_rate": 3.8199941606929524e-07, "loss": 0.2494, "step": 3729 }, { "epoch": 2.744665194996321, "grad_norm": 1.534108636184764, "learning_rate": 3.798273887129755e-07, "loss": 0.2264, "step": 3730 }, { "epoch": 2.745401030169242, "grad_norm": 1.6313195071268691, "learning_rate": 3.7766143444022184e-07, "loss": 0.2558, "step": 3731 }, { "epoch": 2.746136865342163, "grad_norm": 1.7851899440202184, "learning_rate": 3.7550155461837155e-07, "loss": 0.2606, "step": 3732 }, { "epoch": 2.746872700515085, "grad_norm": 1.91306598529346, "learning_rate": 3.733477506109262e-07, "loss": 0.2443, "step": 3733 }, { "epoch": 2.747608535688006, "grad_norm": 1.7036401206816916, "learning_rate": 3.7120002377755594e-07, "loss": 0.2499, "step": 3734 }, { "epoch": 2.748344370860927, "grad_norm": 1.7661407972507936, "learning_rate": 3.690583754740884e-07, "loss": 0.2535, "step": 3735 }, { "epoch": 2.7490802060338484, "grad_norm": 1.7688962888683195, "learning_rate": 3.669228070525177e-07, "loss": 0.2462, "step": 3736 }, { "epoch": 2.7498160412067696, "grad_norm": 1.8550884936713732, "learning_rate": 3.6479331986099877e-07, "loss": 0.2719, "step": 3737 }, { "epoch": 2.750551876379691, "grad_norm": 1.794699618310612, "learning_rate": 3.626699152438484e-07, "loss": 0.2338, "step": 3738 }, { "epoch": 2.751287711552612, "grad_norm": 1.7408553722781128, "learning_rate": 3.6055259454154334e-07, "loss": 0.2489, "step": 3739 }, { "epoch": 2.7520235467255336, "grad_norm": 1.653685677987556, "learning_rate": 3.5844135909071763e-07, "loss": 0.2335, "step": 3740 }, { "epoch": 2.752759381898455, "grad_norm": 1.6401196008369283, "learning_rate": 3.563362102241663e-07, "loss": 0.2297, "step": 3741 }, { "epoch": 2.753495217071376, "grad_norm": 1.650114086179249, "learning_rate": 3.5423714927084186e-07, "loss": 0.222, "step": 3742 }, { "epoch": 2.754231052244297, "grad_norm": 1.7529624656958585, "learning_rate": 3.5214417755585207e-07, "loss": 0.2352, "step": 3743 }, { "epoch": 2.7549668874172184, "grad_norm": 1.7476141744148208, "learning_rate": 3.500572964004612e-07, "loss": 0.2515, "step": 3744 }, { "epoch": 2.75570272259014, "grad_norm": 1.684113881599882, "learning_rate": 3.4797650712208863e-07, "loss": 0.2382, "step": 3745 }, { "epoch": 2.7564385577630612, "grad_norm": 1.7226077801440611, "learning_rate": 3.459018110343082e-07, "loss": 0.2572, "step": 3746 }, { "epoch": 2.7571743929359824, "grad_norm": 1.7000614640772744, "learning_rate": 3.438332094468477e-07, "loss": 0.2667, "step": 3747 }, { "epoch": 2.7579102281089036, "grad_norm": 1.6811733152400394, "learning_rate": 3.417707036655882e-07, "loss": 0.2426, "step": 3748 }, { "epoch": 2.758646063281825, "grad_norm": 1.8433096112063967, "learning_rate": 3.3971429499255805e-07, "loss": 0.2361, "step": 3749 }, { "epoch": 2.759381898454746, "grad_norm": 1.7589599461935912, "learning_rate": 3.376639847259433e-07, "loss": 0.2627, "step": 3750 }, { "epoch": 2.760117733627667, "grad_norm": 1.7222518655753956, "learning_rate": 3.356197741600753e-07, "loss": 0.2411, "step": 3751 }, { "epoch": 2.760853568800589, "grad_norm": 1.7306039299830858, "learning_rate": 3.335816645854373e-07, "loss": 0.2439, "step": 3752 }, { "epoch": 2.76158940397351, "grad_norm": 1.7204833040722294, "learning_rate": 3.3154965728865893e-07, "loss": 0.2311, "step": 3753 }, { "epoch": 2.7623252391464312, "grad_norm": 1.6492223860789363, "learning_rate": 3.2952375355251865e-07, "loss": 0.2349, "step": 3754 }, { "epoch": 2.7630610743193524, "grad_norm": 1.8847412389388334, "learning_rate": 3.275039546559422e-07, "loss": 0.2134, "step": 3755 }, { "epoch": 2.7637969094922736, "grad_norm": 1.9900757823446324, "learning_rate": 3.254902618740008e-07, "loss": 0.2387, "step": 3756 }, { "epoch": 2.764532744665195, "grad_norm": 1.6450134683675928, "learning_rate": 3.234826764779131e-07, "loss": 0.2479, "step": 3757 }, { "epoch": 2.765268579838116, "grad_norm": 1.7328377798045203, "learning_rate": 3.2148119973503534e-07, "loss": 0.2184, "step": 3758 }, { "epoch": 2.7660044150110377, "grad_norm": 1.7715747861024864, "learning_rate": 3.1948583290887904e-07, "loss": 0.2405, "step": 3759 }, { "epoch": 2.766740250183959, "grad_norm": 1.6941783683341627, "learning_rate": 3.174965772590866e-07, "loss": 0.2334, "step": 3760 }, { "epoch": 2.76747608535688, "grad_norm": 1.7110854805425397, "learning_rate": 3.1551343404145005e-07, "loss": 0.2478, "step": 3761 }, { "epoch": 2.7682119205298013, "grad_norm": 1.6485770451255919, "learning_rate": 3.1353640450790257e-07, "loss": 0.241, "step": 3762 }, { "epoch": 2.7689477557027224, "grad_norm": 1.819691858786085, "learning_rate": 3.1156548990651237e-07, "loss": 0.2453, "step": 3763 }, { "epoch": 2.769683590875644, "grad_norm": 1.8301955155987677, "learning_rate": 3.0960069148149306e-07, "loss": 0.2665, "step": 3764 }, { "epoch": 2.7704194260485653, "grad_norm": 1.7710103687449257, "learning_rate": 3.0764201047319364e-07, "loss": 0.243, "step": 3765 }, { "epoch": 2.7711552612214865, "grad_norm": 1.7408209743442606, "learning_rate": 3.0568944811810497e-07, "loss": 0.2242, "step": 3766 }, { "epoch": 2.7718910963944077, "grad_norm": 1.7828223897212827, "learning_rate": 3.037430056488488e-07, "loss": 0.2289, "step": 3767 }, { "epoch": 2.772626931567329, "grad_norm": 1.6460926961312006, "learning_rate": 3.0180268429419345e-07, "loss": 0.2591, "step": 3768 }, { "epoch": 2.77336276674025, "grad_norm": 1.7383097399314889, "learning_rate": 2.9986848527903347e-07, "loss": 0.2386, "step": 3769 }, { "epoch": 2.7740986019131713, "grad_norm": 1.6609537813471469, "learning_rate": 2.979404098244021e-07, "loss": 0.2356, "step": 3770 }, { "epoch": 2.774834437086093, "grad_norm": 1.7039937821432207, "learning_rate": 2.9601845914746907e-07, "loss": 0.2283, "step": 3771 }, { "epoch": 2.775570272259014, "grad_norm": 1.6048995596575188, "learning_rate": 2.9410263446153385e-07, "loss": 0.2445, "step": 3772 }, { "epoch": 2.7763061074319353, "grad_norm": 1.7386979238894014, "learning_rate": 2.9219293697603123e-07, "loss": 0.2445, "step": 3773 }, { "epoch": 2.7770419426048565, "grad_norm": 1.6188822554534645, "learning_rate": 2.9028936789652685e-07, "loss": 0.2147, "step": 3774 }, { "epoch": 2.7777777777777777, "grad_norm": 1.6437033861436134, "learning_rate": 2.8839192842471943e-07, "loss": 0.2428, "step": 3775 }, { "epoch": 2.7785136129506993, "grad_norm": 1.86125843765982, "learning_rate": 2.865006197584319e-07, "loss": 0.2503, "step": 3776 }, { "epoch": 2.77924944812362, "grad_norm": 1.9847081068617871, "learning_rate": 2.8461544309162816e-07, "loss": 0.2469, "step": 3777 }, { "epoch": 2.7799852832965417, "grad_norm": 1.7492899327715994, "learning_rate": 2.827363996143895e-07, "loss": 0.2347, "step": 3778 }, { "epoch": 2.780721118469463, "grad_norm": 1.663247567913572, "learning_rate": 2.808634905129326e-07, "loss": 0.2458, "step": 3779 }, { "epoch": 2.781456953642384, "grad_norm": 1.7366927205325497, "learning_rate": 2.789967169695995e-07, "loss": 0.2329, "step": 3780 }, { "epoch": 2.7821927888153053, "grad_norm": 1.7359097658973703, "learning_rate": 2.771360801628575e-07, "loss": 0.2431, "step": 3781 }, { "epoch": 2.7829286239882265, "grad_norm": 1.8451216094042817, "learning_rate": 2.7528158126730155e-07, "loss": 0.2503, "step": 3782 }, { "epoch": 2.783664459161148, "grad_norm": 1.7150824130698195, "learning_rate": 2.734332214536517e-07, "loss": 0.2315, "step": 3783 }, { "epoch": 2.7844002943340693, "grad_norm": 1.7434566435094143, "learning_rate": 2.7159100188875355e-07, "loss": 0.2515, "step": 3784 }, { "epoch": 2.7851361295069905, "grad_norm": 1.611414517244581, "learning_rate": 2.697549237355723e-07, "loss": 0.2187, "step": 3785 }, { "epoch": 2.7858719646799117, "grad_norm": 1.6858545161068759, "learning_rate": 2.6792498815320314e-07, "loss": 0.2236, "step": 3786 }, { "epoch": 2.786607799852833, "grad_norm": 1.6825226212960303, "learning_rate": 2.6610119629685517e-07, "loss": 0.2573, "step": 3787 }, { "epoch": 2.787343635025754, "grad_norm": 1.6892643290413476, "learning_rate": 2.6428354931786525e-07, "loss": 0.2401, "step": 3788 }, { "epoch": 2.7880794701986753, "grad_norm": 1.7156295264269519, "learning_rate": 2.6247204836369e-07, "loss": 0.2595, "step": 3789 }, { "epoch": 2.788815305371597, "grad_norm": 1.6922051934526812, "learning_rate": 2.606666945779024e-07, "loss": 0.2513, "step": 3790 }, { "epoch": 2.789551140544518, "grad_norm": 1.7448199978073464, "learning_rate": 2.5886748910020075e-07, "loss": 0.2408, "step": 3791 }, { "epoch": 2.7902869757174393, "grad_norm": 1.7829589327199624, "learning_rate": 2.570744330663966e-07, "loss": 0.2674, "step": 3792 }, { "epoch": 2.7910228108903605, "grad_norm": 1.6783520778816432, "learning_rate": 2.552875276084232e-07, "loss": 0.2494, "step": 3793 }, { "epoch": 2.7917586460632817, "grad_norm": 1.809794909271292, "learning_rate": 2.5350677385432843e-07, "loss": 0.2406, "step": 3794 }, { "epoch": 2.7924944812362034, "grad_norm": 1.7836483063918014, "learning_rate": 2.5173217292827954e-07, "loss": 0.2641, "step": 3795 }, { "epoch": 2.793230316409124, "grad_norm": 1.6050419888213217, "learning_rate": 2.4996372595055605e-07, "loss": 0.221, "step": 3796 }, { "epoch": 2.7939661515820458, "grad_norm": 1.69618502636022, "learning_rate": 2.482014340375549e-07, "loss": 0.255, "step": 3797 }, { "epoch": 2.794701986754967, "grad_norm": 1.7161160945449287, "learning_rate": 2.4644529830178733e-07, "loss": 0.2511, "step": 3798 }, { "epoch": 2.795437821927888, "grad_norm": 1.586525383058371, "learning_rate": 2.446953198518776e-07, "loss": 0.231, "step": 3799 }, { "epoch": 2.7961736571008093, "grad_norm": 1.6679864226039085, "learning_rate": 2.4295149979256304e-07, "loss": 0.2462, "step": 3800 }, { "epoch": 2.7969094922737305, "grad_norm": 1.7130687495159882, "learning_rate": 2.412138392246932e-07, "loss": 0.2435, "step": 3801 }, { "epoch": 2.797645327446652, "grad_norm": 1.7055731925857127, "learning_rate": 2.394823392452306e-07, "loss": 0.2227, "step": 3802 }, { "epoch": 2.7983811626195734, "grad_norm": 1.7968640320759925, "learning_rate": 2.377570009472463e-07, "loss": 0.2425, "step": 3803 }, { "epoch": 2.7991169977924946, "grad_norm": 1.7008692340704223, "learning_rate": 2.3603782541992247e-07, "loss": 0.2405, "step": 3804 }, { "epoch": 2.7998528329654158, "grad_norm": 1.622108295457621, "learning_rate": 2.343248137485532e-07, "loss": 0.2639, "step": 3805 }, { "epoch": 2.800588668138337, "grad_norm": 1.6999208700508, "learning_rate": 2.3261796701453676e-07, "loss": 0.26, "step": 3806 }, { "epoch": 2.801324503311258, "grad_norm": 1.7051188122972851, "learning_rate": 2.3091728629538456e-07, "loss": 0.2351, "step": 3807 }, { "epoch": 2.8020603384841793, "grad_norm": 1.808546627779052, "learning_rate": 2.2922277266471226e-07, "loss": 0.2308, "step": 3808 }, { "epoch": 2.802796173657101, "grad_norm": 1.7450576217667282, "learning_rate": 2.2753442719224195e-07, "loss": 0.2591, "step": 3809 }, { "epoch": 2.803532008830022, "grad_norm": 1.6812388920597305, "learning_rate": 2.2585225094380435e-07, "loss": 0.218, "step": 3810 }, { "epoch": 2.8042678440029434, "grad_norm": 1.7290820435981877, "learning_rate": 2.241762449813345e-07, "loss": 0.2222, "step": 3811 }, { "epoch": 2.8050036791758646, "grad_norm": 1.7483591107645768, "learning_rate": 2.2250641036287156e-07, "loss": 0.2565, "step": 3812 }, { "epoch": 2.8057395143487858, "grad_norm": 1.6431266250107661, "learning_rate": 2.20842748142559e-07, "loss": 0.2327, "step": 3813 }, { "epoch": 2.8064753495217074, "grad_norm": 1.779611723961918, "learning_rate": 2.191852593706456e-07, "loss": 0.2576, "step": 3814 }, { "epoch": 2.807211184694628, "grad_norm": 1.7626085478305051, "learning_rate": 2.175339450934777e-07, "loss": 0.2608, "step": 3815 }, { "epoch": 2.80794701986755, "grad_norm": 1.748439382047381, "learning_rate": 2.158888063535125e-07, "loss": 0.2509, "step": 3816 }, { "epoch": 2.808682855040471, "grad_norm": 1.7013431812034925, "learning_rate": 2.142498441893004e-07, "loss": 0.2378, "step": 3817 }, { "epoch": 2.809418690213392, "grad_norm": 1.6191607823494434, "learning_rate": 2.1261705963549595e-07, "loss": 0.2255, "step": 3818 }, { "epoch": 2.8101545253863134, "grad_norm": 1.6835496697102084, "learning_rate": 2.1099045372285465e-07, "loss": 0.2492, "step": 3819 }, { "epoch": 2.8108903605592346, "grad_norm": 1.7305485123754214, "learning_rate": 2.0937002747823067e-07, "loss": 0.2578, "step": 3820 }, { "epoch": 2.811626195732156, "grad_norm": 1.7335360204400307, "learning_rate": 2.077557819245757e-07, "loss": 0.2691, "step": 3821 }, { "epoch": 2.8123620309050774, "grad_norm": 1.697816047503833, "learning_rate": 2.0614771808094237e-07, "loss": 0.2556, "step": 3822 }, { "epoch": 2.8130978660779986, "grad_norm": 1.7041439688070037, "learning_rate": 2.0454583696247864e-07, "loss": 0.227, "step": 3823 }, { "epoch": 2.81383370125092, "grad_norm": 1.691147827037893, "learning_rate": 2.0295013958042899e-07, "loss": 0.2615, "step": 3824 }, { "epoch": 2.814569536423841, "grad_norm": 1.7361615374444903, "learning_rate": 2.0136062694213865e-07, "loss": 0.255, "step": 3825 }, { "epoch": 2.815305371596762, "grad_norm": 1.679091764519708, "learning_rate": 1.997773000510428e-07, "loss": 0.2421, "step": 3826 }, { "epoch": 2.8160412067696834, "grad_norm": 1.6994974649782233, "learning_rate": 1.9820015990667408e-07, "loss": 0.2505, "step": 3827 }, { "epoch": 2.816777041942605, "grad_norm": 1.8361834940641757, "learning_rate": 1.9662920750466053e-07, "loss": 0.2276, "step": 3828 }, { "epoch": 2.8175128771155262, "grad_norm": 1.672612152711852, "learning_rate": 1.9506444383672328e-07, "loss": 0.2338, "step": 3829 }, { "epoch": 2.8182487122884474, "grad_norm": 1.7272120529553066, "learning_rate": 1.9350586989067443e-07, "loss": 0.2305, "step": 3830 }, { "epoch": 2.8189845474613686, "grad_norm": 1.7379486698667923, "learning_rate": 1.9195348665042134e-07, "loss": 0.2605, "step": 3831 }, { "epoch": 2.81972038263429, "grad_norm": 1.7642422455018014, "learning_rate": 1.9040729509596235e-07, "loss": 0.2581, "step": 3832 }, { "epoch": 2.8204562178072115, "grad_norm": 1.7667279173122372, "learning_rate": 1.8886729620338662e-07, "loss": 0.2062, "step": 3833 }, { "epoch": 2.821192052980132, "grad_norm": 1.7446347653386507, "learning_rate": 1.873334909448765e-07, "loss": 0.2391, "step": 3834 }, { "epoch": 2.821927888153054, "grad_norm": 1.5362138635167109, "learning_rate": 1.8580588028869972e-07, "loss": 0.2084, "step": 3835 }, { "epoch": 2.822663723325975, "grad_norm": 1.7056482999401628, "learning_rate": 1.8428446519921704e-07, "loss": 0.2288, "step": 3836 }, { "epoch": 2.8233995584988962, "grad_norm": 1.71765510442023, "learning_rate": 1.827692466368769e-07, "loss": 0.2757, "step": 3837 }, { "epoch": 2.8241353936718174, "grad_norm": 1.7987636298316332, "learning_rate": 1.8126022555821742e-07, "loss": 0.251, "step": 3838 }, { "epoch": 2.8248712288447386, "grad_norm": 1.72722017655643, "learning_rate": 1.7975740291586109e-07, "loss": 0.2224, "step": 3839 }, { "epoch": 2.8256070640176603, "grad_norm": 1.8166448020854822, "learning_rate": 1.7826077965852008e-07, "loss": 0.264, "step": 3840 }, { "epoch": 2.8263428991905815, "grad_norm": 1.7068395150448654, "learning_rate": 1.7677035673099196e-07, "loss": 0.2601, "step": 3841 }, { "epoch": 2.8270787343635027, "grad_norm": 1.6752833324191418, "learning_rate": 1.752861350741597e-07, "loss": 0.2211, "step": 3842 }, { "epoch": 2.827814569536424, "grad_norm": 1.8490736586032097, "learning_rate": 1.738081156249949e-07, "loss": 0.2599, "step": 3843 }, { "epoch": 2.828550404709345, "grad_norm": 1.849256559249951, "learning_rate": 1.7233629931654782e-07, "loss": 0.266, "step": 3844 }, { "epoch": 2.8292862398822662, "grad_norm": 1.6879020483180327, "learning_rate": 1.708706870779575e-07, "loss": 0.2634, "step": 3845 }, { "epoch": 2.8300220750551874, "grad_norm": 1.6469948320387187, "learning_rate": 1.6941127983444494e-07, "loss": 0.2325, "step": 3846 }, { "epoch": 2.830757910228109, "grad_norm": 1.682773527355791, "learning_rate": 1.6795807850731428e-07, "loss": 0.219, "step": 3847 }, { "epoch": 2.8314937454010303, "grad_norm": 1.6260516833538026, "learning_rate": 1.6651108401395054e-07, "loss": 0.258, "step": 3848 }, { "epoch": 2.8322295805739515, "grad_norm": 1.7214682765400753, "learning_rate": 1.6507029726782197e-07, "loss": 0.2199, "step": 3849 }, { "epoch": 2.8329654157468727, "grad_norm": 1.7480837771183404, "learning_rate": 1.6363571917847875e-07, "loss": 0.2449, "step": 3850 }, { "epoch": 2.833701250919794, "grad_norm": 1.681842245536936, "learning_rate": 1.6220735065154758e-07, "loss": 0.2455, "step": 3851 }, { "epoch": 2.8344370860927155, "grad_norm": 1.738715662382036, "learning_rate": 1.607851925887427e-07, "loss": 0.2604, "step": 3852 }, { "epoch": 2.8351729212656362, "grad_norm": 1.9047082165632614, "learning_rate": 1.593692458878482e-07, "loss": 0.2481, "step": 3853 }, { "epoch": 2.835908756438558, "grad_norm": 1.5903641727958264, "learning_rate": 1.5795951144273459e-07, "loss": 0.2246, "step": 3854 }, { "epoch": 2.836644591611479, "grad_norm": 1.6712823649223019, "learning_rate": 1.5655599014334776e-07, "loss": 0.2217, "step": 3855 }, { "epoch": 2.8373804267844003, "grad_norm": 1.5823660561611348, "learning_rate": 1.5515868287571124e-07, "loss": 0.2189, "step": 3856 }, { "epoch": 2.8381162619573215, "grad_norm": 1.766854948021502, "learning_rate": 1.5376759052192603e-07, "loss": 0.2826, "step": 3857 }, { "epoch": 2.8388520971302427, "grad_norm": 1.7783586681881205, "learning_rate": 1.5238271396016858e-07, "loss": 0.2575, "step": 3858 }, { "epoch": 2.8395879323031643, "grad_norm": 1.7535192234977668, "learning_rate": 1.5100405406469508e-07, "loss": 0.2273, "step": 3859 }, { "epoch": 2.8403237674760855, "grad_norm": 1.8553262616949613, "learning_rate": 1.4963161170583162e-07, "loss": 0.2623, "step": 3860 }, { "epoch": 2.8410596026490067, "grad_norm": 1.7334361248509211, "learning_rate": 1.482653877499851e-07, "loss": 0.2351, "step": 3861 }, { "epoch": 2.841795437821928, "grad_norm": 1.625099183941823, "learning_rate": 1.469053830596323e-07, "loss": 0.242, "step": 3862 }, { "epoch": 2.842531272994849, "grad_norm": 1.7375361092336434, "learning_rate": 1.4555159849332644e-07, "loss": 0.2317, "step": 3863 }, { "epoch": 2.8432671081677703, "grad_norm": 1.8159998513350786, "learning_rate": 1.442040349056928e-07, "loss": 0.2741, "step": 3864 }, { "epoch": 2.8440029433406915, "grad_norm": 1.700104235352774, "learning_rate": 1.4286269314743085e-07, "loss": 0.2207, "step": 3865 }, { "epoch": 2.844738778513613, "grad_norm": 1.7266079925122733, "learning_rate": 1.4152757406530993e-07, "loss": 0.2398, "step": 3866 }, { "epoch": 2.8454746136865343, "grad_norm": 1.796129301468646, "learning_rate": 1.401986785021725e-07, "loss": 0.2503, "step": 3867 }, { "epoch": 2.8462104488594555, "grad_norm": 1.8288480000370126, "learning_rate": 1.3887600729693307e-07, "loss": 0.2575, "step": 3868 }, { "epoch": 2.8469462840323767, "grad_norm": 1.6155849653803234, "learning_rate": 1.3755956128457482e-07, "loss": 0.2287, "step": 3869 }, { "epoch": 2.847682119205298, "grad_norm": 1.6745176282831853, "learning_rate": 1.3624934129615407e-07, "loss": 0.2347, "step": 3870 }, { "epoch": 2.8484179543782195, "grad_norm": 1.5375708927140899, "learning_rate": 1.3494534815879257e-07, "loss": 0.2311, "step": 3871 }, { "epoch": 2.8491537895511403, "grad_norm": 1.7427531524604531, "learning_rate": 1.3364758269568513e-07, "loss": 0.261, "step": 3872 }, { "epoch": 2.849889624724062, "grad_norm": 1.5947660433671558, "learning_rate": 1.3235604572609194e-07, "loss": 0.2245, "step": 3873 }, { "epoch": 2.850625459896983, "grad_norm": 1.681111269451823, "learning_rate": 1.310707380653442e-07, "loss": 0.2413, "step": 3874 }, { "epoch": 2.8513612950699043, "grad_norm": 1.660500488982163, "learning_rate": 1.2979166052483838e-07, "loss": 0.2398, "step": 3875 }, { "epoch": 2.8520971302428255, "grad_norm": 1.805807241744587, "learning_rate": 1.2851881391203746e-07, "loss": 0.2491, "step": 3876 }, { "epoch": 2.8528329654157467, "grad_norm": 1.6632693241905476, "learning_rate": 1.2725219903047425e-07, "loss": 0.2319, "step": 3877 }, { "epoch": 2.8535688005886684, "grad_norm": 1.7410399160438021, "learning_rate": 1.2599181667974249e-07, "loss": 0.2493, "step": 3878 }, { "epoch": 2.8543046357615895, "grad_norm": 1.7105156127289067, "learning_rate": 1.2473766765550786e-07, "loss": 0.2527, "step": 3879 }, { "epoch": 2.8550404709345107, "grad_norm": 1.6541770110670573, "learning_rate": 1.2348975274949605e-07, "loss": 0.2356, "step": 3880 }, { "epoch": 2.855776306107432, "grad_norm": 1.774632855839168, "learning_rate": 1.222480727494968e-07, "loss": 0.2467, "step": 3881 }, { "epoch": 2.856512141280353, "grad_norm": 1.5760161413871805, "learning_rate": 1.2101262843936867e-07, "loss": 0.2423, "step": 3882 }, { "epoch": 2.8572479764532743, "grad_norm": 1.6960402318525245, "learning_rate": 1.1978342059902892e-07, "loss": 0.276, "step": 3883 }, { "epoch": 2.8579838116261955, "grad_norm": 1.663118558932001, "learning_rate": 1.1856045000446126e-07, "loss": 0.2622, "step": 3884 }, { "epoch": 2.858719646799117, "grad_norm": 1.6850516527069948, "learning_rate": 1.1734371742770811e-07, "loss": 0.2424, "step": 3885 }, { "epoch": 2.8594554819720384, "grad_norm": 1.678672565355667, "learning_rate": 1.161332236368784e-07, "loss": 0.2224, "step": 3886 }, { "epoch": 2.8601913171449596, "grad_norm": 1.7041939373481028, "learning_rate": 1.1492896939613862e-07, "loss": 0.2534, "step": 3887 }, { "epoch": 2.8609271523178808, "grad_norm": 1.7764647502009725, "learning_rate": 1.1373095546571844e-07, "loss": 0.2544, "step": 3888 }, { "epoch": 2.861662987490802, "grad_norm": 1.60306812298855, "learning_rate": 1.1253918260190844e-07, "loss": 0.2124, "step": 3889 }, { "epoch": 2.8623988226637236, "grad_norm": 1.7697527589120665, "learning_rate": 1.1135365155705791e-07, "loss": 0.239, "step": 3890 }, { "epoch": 2.8631346578366443, "grad_norm": 1.7553991062647485, "learning_rate": 1.1017436307957708e-07, "loss": 0.2692, "step": 3891 }, { "epoch": 2.863870493009566, "grad_norm": 1.7099667448181222, "learning_rate": 1.0900131791393265e-07, "loss": 0.204, "step": 3892 }, { "epoch": 2.864606328182487, "grad_norm": 1.7267175690766114, "learning_rate": 1.0783451680065448e-07, "loss": 0.2632, "step": 3893 }, { "epoch": 2.8653421633554084, "grad_norm": 1.6990397168793294, "learning_rate": 1.0667396047632783e-07, "loss": 0.2612, "step": 3894 }, { "epoch": 2.8660779985283296, "grad_norm": 1.654645416275204, "learning_rate": 1.0551964967359441e-07, "loss": 0.2236, "step": 3895 }, { "epoch": 2.8668138337012508, "grad_norm": 1.7300435988606697, "learning_rate": 1.0437158512115575e-07, "loss": 0.2102, "step": 3896 }, { "epoch": 2.8675496688741724, "grad_norm": 1.8858286272858193, "learning_rate": 1.032297675437699e-07, "loss": 0.2693, "step": 3897 }, { "epoch": 2.8682855040470936, "grad_norm": 1.7711442445045276, "learning_rate": 1.0209419766225026e-07, "loss": 0.226, "step": 3898 }, { "epoch": 2.869021339220015, "grad_norm": 1.6997018545775766, "learning_rate": 1.0096487619346563e-07, "loss": 0.2536, "step": 3899 }, { "epoch": 2.869757174392936, "grad_norm": 1.792419731788264, "learning_rate": 9.984180385034347e-08, "loss": 0.2713, "step": 3900 }, { "epoch": 2.870493009565857, "grad_norm": 1.698161778290663, "learning_rate": 9.872498134186115e-08, "loss": 0.2419, "step": 3901 }, { "epoch": 2.8712288447387784, "grad_norm": 1.6834163653122964, "learning_rate": 9.761440937305577e-08, "loss": 0.2234, "step": 3902 }, { "epoch": 2.8719646799116996, "grad_norm": 1.7911694623262744, "learning_rate": 9.651008864501543e-08, "loss": 0.2642, "step": 3903 }, { "epoch": 2.872700515084621, "grad_norm": 1.7200443127302365, "learning_rate": 9.541201985488358e-08, "loss": 0.2043, "step": 3904 }, { "epoch": 2.8734363502575424, "grad_norm": 1.7521098261829209, "learning_rate": 9.432020369585348e-08, "loss": 0.2458, "step": 3905 }, { "epoch": 2.8741721854304636, "grad_norm": 1.8503478790003667, "learning_rate": 9.323464085717715e-08, "loss": 0.2716, "step": 3906 }, { "epoch": 2.874908020603385, "grad_norm": 1.7415291633190446, "learning_rate": 9.215533202415306e-08, "loss": 0.2319, "step": 3907 }, { "epoch": 2.875643855776306, "grad_norm": 1.7107504735755665, "learning_rate": 9.108227787813506e-08, "loss": 0.245, "step": 3908 }, { "epoch": 2.8763796909492276, "grad_norm": 1.698851471420804, "learning_rate": 9.001547909652909e-08, "loss": 0.2299, "step": 3909 }, { "epoch": 2.8771155261221484, "grad_norm": 1.7649974101369026, "learning_rate": 8.89549363527864e-08, "loss": 0.265, "step": 3910 }, { "epoch": 2.87785136129507, "grad_norm": 1.7013650718696174, "learning_rate": 8.790065031641815e-08, "loss": 0.2489, "step": 3911 }, { "epoch": 2.878587196467991, "grad_norm": 1.7854832822149096, "learning_rate": 8.685262165297637e-08, "loss": 0.2634, "step": 3912 }, { "epoch": 2.8793230316409124, "grad_norm": 1.743580186114551, "learning_rate": 8.581085102407072e-08, "loss": 0.2266, "step": 3913 }, { "epoch": 2.8800588668138336, "grad_norm": 1.8084258365221055, "learning_rate": 8.477533908735402e-08, "loss": 0.2267, "step": 3914 }, { "epoch": 2.880794701986755, "grad_norm": 1.7189138185116613, "learning_rate": 8.374608649653337e-08, "loss": 0.2402, "step": 3915 }, { "epoch": 2.8815305371596764, "grad_norm": 1.8841224253447875, "learning_rate": 8.272309390136013e-08, "loss": 0.2474, "step": 3916 }, { "epoch": 2.8822663723325976, "grad_norm": 1.6172860346966342, "learning_rate": 8.170636194763659e-08, "loss": 0.2657, "step": 3917 }, { "epoch": 2.883002207505519, "grad_norm": 1.7778769648832287, "learning_rate": 8.069589127721044e-08, "loss": 0.2548, "step": 3918 }, { "epoch": 2.88373804267844, "grad_norm": 1.7708081358154095, "learning_rate": 7.96916825279781e-08, "loss": 0.2494, "step": 3919 }, { "epoch": 2.8844738778513612, "grad_norm": 1.8043890203756383, "learning_rate": 7.869373633388466e-08, "loss": 0.239, "step": 3920 }, { "epoch": 2.8852097130242824, "grad_norm": 1.78032293394635, "learning_rate": 7.770205332491842e-08, "loss": 0.2471, "step": 3921 }, { "epoch": 2.8859455481972036, "grad_norm": 1.7040158091106619, "learning_rate": 7.671663412711527e-08, "loss": 0.2769, "step": 3922 }, { "epoch": 2.8866813833701253, "grad_norm": 1.7862041174426364, "learning_rate": 7.573747936255759e-08, "loss": 0.2577, "step": 3923 }, { "epoch": 2.8874172185430464, "grad_norm": 1.7753709768672823, "learning_rate": 7.476458964937316e-08, "loss": 0.2583, "step": 3924 }, { "epoch": 2.8881530537159676, "grad_norm": 1.816886631085929, "learning_rate": 7.37979656017318e-08, "loss": 0.2791, "step": 3925 }, { "epoch": 2.888888888888889, "grad_norm": 1.7695410883723113, "learning_rate": 7.283760782985095e-08, "loss": 0.235, "step": 3926 }, { "epoch": 2.88962472406181, "grad_norm": 1.64846769750072, "learning_rate": 7.188351693999229e-08, "loss": 0.2156, "step": 3927 }, { "epoch": 2.8903605592347317, "grad_norm": 1.7353518182591017, "learning_rate": 7.09356935344585e-08, "loss": 0.2517, "step": 3928 }, { "epoch": 2.8910963944076524, "grad_norm": 1.7645881210634784, "learning_rate": 6.999413821159873e-08, "loss": 0.2371, "step": 3929 }, { "epoch": 2.891832229580574, "grad_norm": 1.7469616305671387, "learning_rate": 6.905885156580528e-08, "loss": 0.2564, "step": 3930 }, { "epoch": 2.8925680647534953, "grad_norm": 1.8049932633744228, "learning_rate": 6.812983418750917e-08, "loss": 0.249, "step": 3931 }, { "epoch": 2.8933038999264165, "grad_norm": 1.7324836601006772, "learning_rate": 6.720708666318797e-08, "loss": 0.2475, "step": 3932 }, { "epoch": 2.8940397350993377, "grad_norm": 1.7649819957764241, "learning_rate": 6.629060957536016e-08, "loss": 0.2568, "step": 3933 }, { "epoch": 2.894775570272259, "grad_norm": 1.834587845110287, "learning_rate": 6.538040350258401e-08, "loss": 0.2515, "step": 3934 }, { "epoch": 2.8955114054451805, "grad_norm": 1.603788825039022, "learning_rate": 6.447646901945992e-08, "loss": 0.2166, "step": 3935 }, { "epoch": 2.8962472406181017, "grad_norm": 1.7256736357967237, "learning_rate": 6.357880669663141e-08, "loss": 0.2163, "step": 3936 }, { "epoch": 2.896983075791023, "grad_norm": 1.7692792045771981, "learning_rate": 6.268741710077741e-08, "loss": 0.2207, "step": 3937 }, { "epoch": 2.897718910963944, "grad_norm": 1.680337233513465, "learning_rate": 6.180230079462224e-08, "loss": 0.2606, "step": 3938 }, { "epoch": 2.8984547461368653, "grad_norm": 1.7087576597713394, "learning_rate": 6.09234583369278e-08, "loss": 0.2691, "step": 3939 }, { "epoch": 2.8991905813097865, "grad_norm": 1.7563729182412025, "learning_rate": 6.005089028249366e-08, "loss": 0.242, "step": 3940 }, { "epoch": 2.8999264164827077, "grad_norm": 1.6879711436989635, "learning_rate": 5.918459718216141e-08, "loss": 0.2271, "step": 3941 }, { "epoch": 2.9006622516556293, "grad_norm": 1.66766017527213, "learning_rate": 5.832457958281024e-08, "loss": 0.2213, "step": 3942 }, { "epoch": 2.9013980868285505, "grad_norm": 1.6302315374162872, "learning_rate": 5.747083802735587e-08, "loss": 0.2455, "step": 3943 }, { "epoch": 2.9021339220014717, "grad_norm": 1.6992963017607876, "learning_rate": 5.662337305475607e-08, "loss": 0.2355, "step": 3944 }, { "epoch": 2.902869757174393, "grad_norm": 1.6626249391619667, "learning_rate": 5.5782185200002894e-08, "loss": 0.2308, "step": 3945 }, { "epoch": 2.903605592347314, "grad_norm": 1.7493284052263007, "learning_rate": 5.494727499412489e-08, "loss": 0.2229, "step": 3946 }, { "epoch": 2.9043414275202357, "grad_norm": 1.7744030940731843, "learning_rate": 5.4118642964192667e-08, "loss": 0.2587, "step": 3947 }, { "epoch": 2.9050772626931565, "grad_norm": 1.6015518279012513, "learning_rate": 5.329628963330779e-08, "loss": 0.2173, "step": 3948 }, { "epoch": 2.905813097866078, "grad_norm": 1.7237465312306102, "learning_rate": 5.2480215520611665e-08, "loss": 0.2558, "step": 3949 }, { "epoch": 2.9065489330389993, "grad_norm": 1.7660219818119478, "learning_rate": 5.167042114128218e-08, "loss": 0.2404, "step": 3950 }, { "epoch": 2.9072847682119205, "grad_norm": 1.8650506141659138, "learning_rate": 5.086690700652819e-08, "loss": 0.2465, "step": 3951 }, { "epoch": 2.9080206033848417, "grad_norm": 1.792688462332254, "learning_rate": 5.006967362359949e-08, "loss": 0.2404, "step": 3952 }, { "epoch": 2.908756438557763, "grad_norm": 1.6495879196116516, "learning_rate": 4.927872149577795e-08, "loss": 0.2291, "step": 3953 }, { "epoch": 2.9094922737306845, "grad_norm": 1.6484836057722034, "learning_rate": 4.8494051122381925e-08, "loss": 0.2485, "step": 3954 }, { "epoch": 2.9102281089036057, "grad_norm": 1.7260817053464583, "learning_rate": 4.7715662998760735e-08, "loss": 0.2642, "step": 3955 }, { "epoch": 2.910963944076527, "grad_norm": 1.7453178597886456, "learning_rate": 4.6943557616302424e-08, "loss": 0.2764, "step": 3956 }, { "epoch": 2.911699779249448, "grad_norm": 1.6141114357782218, "learning_rate": 4.617773546242599e-08, "loss": 0.2265, "step": 3957 }, { "epoch": 2.9124356144223693, "grad_norm": 1.695991565725627, "learning_rate": 4.54181970205847e-08, "loss": 0.2543, "step": 3958 }, { "epoch": 2.9131714495952905, "grad_norm": 1.7647184170576748, "learning_rate": 4.4664942770263894e-08, "loss": 0.2467, "step": 3959 }, { "epoch": 2.9139072847682117, "grad_norm": 1.8503596768226047, "learning_rate": 4.391797318698321e-08, "loss": 0.2139, "step": 3960 }, { "epoch": 2.9146431199411333, "grad_norm": 1.6503332381647209, "learning_rate": 4.317728874229321e-08, "loss": 0.2196, "step": 3961 }, { "epoch": 2.9153789551140545, "grad_norm": 1.7300859051374122, "learning_rate": 4.244288990377987e-08, "loss": 0.2514, "step": 3962 }, { "epoch": 2.9161147902869757, "grad_norm": 1.7115242978562861, "learning_rate": 4.1714777135059005e-08, "loss": 0.2369, "step": 3963 }, { "epoch": 2.916850625459897, "grad_norm": 1.6210946236359676, "learning_rate": 4.0992950895776265e-08, "loss": 0.2577, "step": 3964 }, { "epoch": 2.917586460632818, "grad_norm": 1.6769143074304844, "learning_rate": 4.027741164161269e-08, "loss": 0.2337, "step": 3965 }, { "epoch": 2.9183222958057398, "grad_norm": 1.718561733469092, "learning_rate": 3.9568159824278084e-08, "loss": 0.2336, "step": 3966 }, { "epoch": 2.9190581309786605, "grad_norm": 1.725334836518469, "learning_rate": 3.8865195891512054e-08, "loss": 0.2407, "step": 3967 }, { "epoch": 2.919793966151582, "grad_norm": 1.786565909794773, "learning_rate": 3.816852028708739e-08, "loss": 0.2499, "step": 3968 }, { "epoch": 2.9205298013245033, "grad_norm": 1.8483599999901057, "learning_rate": 3.7478133450805644e-08, "loss": 0.2688, "step": 3969 }, { "epoch": 2.9212656364974245, "grad_norm": 1.6866063862736622, "learning_rate": 3.679403581849927e-08, "loss": 0.2455, "step": 3970 }, { "epoch": 2.9220014716703457, "grad_norm": 1.6112338805691302, "learning_rate": 3.6116227822028396e-08, "loss": 0.2364, "step": 3971 }, { "epoch": 2.922737306843267, "grad_norm": 1.6304663417492535, "learning_rate": 3.5444709889284056e-08, "loss": 0.2434, "step": 3972 }, { "epoch": 2.9234731420161886, "grad_norm": 1.8452578767862138, "learning_rate": 3.477948244418716e-08, "loss": 0.2328, "step": 3973 }, { "epoch": 2.9242089771891098, "grad_norm": 1.687145343627579, "learning_rate": 3.412054590668734e-08, "loss": 0.2536, "step": 3974 }, { "epoch": 2.924944812362031, "grad_norm": 1.6909918448499368, "learning_rate": 3.346790069276073e-08, "loss": 0.2471, "step": 3975 }, { "epoch": 2.925680647534952, "grad_norm": 1.7582689611828515, "learning_rate": 3.2821547214413327e-08, "loss": 0.2403, "step": 3976 }, { "epoch": 2.9264164827078734, "grad_norm": 1.6923429920291118, "learning_rate": 3.218148587968095e-08, "loss": 0.2365, "step": 3977 }, { "epoch": 2.9271523178807946, "grad_norm": 1.697567634867341, "learning_rate": 3.154771709262372e-08, "loss": 0.2438, "step": 3978 }, { "epoch": 2.9278881530537157, "grad_norm": 1.9911689877104657, "learning_rate": 3.0920241253331596e-08, "loss": 0.2565, "step": 3979 }, { "epoch": 2.9286239882266374, "grad_norm": 1.66746756495546, "learning_rate": 3.029905875792327e-08, "loss": 0.2748, "step": 3980 }, { "epoch": 2.9293598233995586, "grad_norm": 1.747655642344477, "learning_rate": 2.9684169998540623e-08, "loss": 0.2532, "step": 3981 }, { "epoch": 2.9300956585724798, "grad_norm": 1.7025002737649482, "learning_rate": 2.9075575363355368e-08, "loss": 0.2256, "step": 3982 }, { "epoch": 2.930831493745401, "grad_norm": 1.7299149792082378, "learning_rate": 2.8473275236564624e-08, "loss": 0.2599, "step": 3983 }, { "epoch": 2.931567328918322, "grad_norm": 1.6592356499933907, "learning_rate": 2.787726999839313e-08, "loss": 0.2607, "step": 3984 }, { "epoch": 2.932303164091244, "grad_norm": 1.7859606967223403, "learning_rate": 2.728756002508881e-08, "loss": 0.2905, "step": 3985 }, { "epoch": 2.9330389992641646, "grad_norm": 1.6945417389831066, "learning_rate": 2.6704145688930537e-08, "loss": 0.2356, "step": 3986 }, { "epoch": 2.933774834437086, "grad_norm": 1.767884119811445, "learning_rate": 2.612702735821815e-08, "loss": 0.2701, "step": 3987 }, { "epoch": 2.9345106696100074, "grad_norm": 1.7685506153275976, "learning_rate": 2.555620539727799e-08, "loss": 0.2455, "step": 3988 }, { "epoch": 2.9352465047829286, "grad_norm": 1.7125831788915529, "learning_rate": 2.499168016646292e-08, "loss": 0.2224, "step": 3989 }, { "epoch": 2.93598233995585, "grad_norm": 1.6180602942617421, "learning_rate": 2.4433452022150083e-08, "loss": 0.2393, "step": 3990 }, { "epoch": 2.936718175128771, "grad_norm": 1.712067063653246, "learning_rate": 2.388152131674093e-08, "loss": 0.2344, "step": 3991 }, { "epoch": 2.9374540103016926, "grad_norm": 1.6433071773205776, "learning_rate": 2.33358883986623e-08, "loss": 0.2126, "step": 3992 }, { "epoch": 2.938189845474614, "grad_norm": 1.7435907891861424, "learning_rate": 2.2796553612364215e-08, "loss": 0.2454, "step": 3993 }, { "epoch": 2.938925680647535, "grad_norm": 1.6790160447393114, "learning_rate": 2.2263517298320992e-08, "loss": 0.2493, "step": 3994 }, { "epoch": 2.939661515820456, "grad_norm": 1.712606209355147, "learning_rate": 2.173677979303346e-08, "loss": 0.2536, "step": 3995 }, { "epoch": 2.9403973509933774, "grad_norm": 1.6878574510776758, "learning_rate": 2.121634142902007e-08, "loss": 0.2629, "step": 3996 }, { "epoch": 2.9411331861662986, "grad_norm": 1.6955807491233308, "learning_rate": 2.070220253483024e-08, "loss": 0.2344, "step": 3997 }, { "epoch": 2.94186902133922, "grad_norm": 1.5580717322113489, "learning_rate": 2.0194363435029897e-08, "loss": 0.2265, "step": 3998 }, { "epoch": 2.9426048565121414, "grad_norm": 1.6162039861501551, "learning_rate": 1.969282445021259e-08, "loss": 0.2347, "step": 3999 }, { "epoch": 2.9433406916850626, "grad_norm": 1.764457028718505, "learning_rate": 1.919758589699283e-08, "loss": 0.2508, "step": 4000 }, { "epoch": 2.944076526857984, "grad_norm": 1.782712003795116, "learning_rate": 1.870864808800721e-08, "loss": 0.2338, "step": 4001 }, { "epoch": 2.944812362030905, "grad_norm": 1.7147194582544212, "learning_rate": 1.8226011331916594e-08, "loss": 0.2571, "step": 4002 }, { "epoch": 2.945548197203826, "grad_norm": 1.5955996500730532, "learning_rate": 1.774967593340171e-08, "loss": 0.2184, "step": 4003 }, { "epoch": 2.946284032376748, "grad_norm": 1.6890582331058435, "learning_rate": 1.727964219316758e-08, "loss": 0.2605, "step": 4004 }, { "epoch": 2.9470198675496686, "grad_norm": 1.6868944870334195, "learning_rate": 1.6815910407939062e-08, "loss": 0.2322, "step": 4005 }, { "epoch": 2.9477557027225902, "grad_norm": 1.7036969855230735, "learning_rate": 1.635848087046532e-08, "loss": 0.2365, "step": 4006 }, { "epoch": 2.9484915378955114, "grad_norm": 1.6331810605869743, "learning_rate": 1.5907353869514252e-08, "loss": 0.2168, "step": 4007 }, { "epoch": 2.9492273730684326, "grad_norm": 1.7813894939481385, "learning_rate": 1.5462529689875827e-08, "loss": 0.2562, "step": 4008 }, { "epoch": 2.949963208241354, "grad_norm": 1.6422978487109332, "learning_rate": 1.5024008612363196e-08, "loss": 0.2366, "step": 4009 }, { "epoch": 2.950699043414275, "grad_norm": 1.8022417872967114, "learning_rate": 1.4591790913806025e-08, "loss": 0.2631, "step": 4010 }, { "epoch": 2.9514348785871967, "grad_norm": 1.7424687428029613, "learning_rate": 1.4165876867059391e-08, "loss": 0.2496, "step": 4011 }, { "epoch": 2.952170713760118, "grad_norm": 1.7799284371788084, "learning_rate": 1.3746266740997104e-08, "loss": 0.245, "step": 4012 }, { "epoch": 2.952906548933039, "grad_norm": 1.8080204637884234, "learning_rate": 1.3332960800512829e-08, "loss": 0.24, "step": 4013 }, { "epoch": 2.9536423841059603, "grad_norm": 1.6437354156481234, "learning_rate": 1.2925959306520075e-08, "loss": 0.2513, "step": 4014 }, { "epoch": 2.9543782192788814, "grad_norm": 1.6491082097660745, "learning_rate": 1.2525262515954429e-08, "loss": 0.2031, "step": 4015 }, { "epoch": 2.9551140544518026, "grad_norm": 1.759923806172501, "learning_rate": 1.2130870681770213e-08, "loss": 0.2393, "step": 4016 }, { "epoch": 2.955849889624724, "grad_norm": 1.725005918603718, "learning_rate": 1.1742784052941603e-08, "loss": 0.2431, "step": 4017 }, { "epoch": 2.9565857247976455, "grad_norm": 1.8219965327986294, "learning_rate": 1.1361002874461512e-08, "loss": 0.23, "step": 4018 }, { "epoch": 2.9573215599705667, "grad_norm": 1.8510207833705905, "learning_rate": 1.098552738734493e-08, "loss": 0.266, "step": 4019 }, { "epoch": 2.958057395143488, "grad_norm": 1.6304325326519737, "learning_rate": 1.0616357828622248e-08, "loss": 0.2548, "step": 4020 }, { "epoch": 2.958793230316409, "grad_norm": 1.6491206322228935, "learning_rate": 1.0253494431347045e-08, "loss": 0.2401, "step": 4021 }, { "epoch": 2.9595290654893303, "grad_norm": 1.6666661759027839, "learning_rate": 9.896937424589415e-09, "loss": 0.2076, "step": 4022 }, { "epoch": 2.960264900662252, "grad_norm": 1.549614123934888, "learning_rate": 9.546687033437086e-09, "loss": 0.2527, "step": 4023 }, { "epoch": 2.9610007358351726, "grad_norm": 1.9003187259923657, "learning_rate": 9.202743479002074e-09, "loss": 0.2647, "step": 4024 }, { "epoch": 2.9617365710080943, "grad_norm": 1.6377511312762252, "learning_rate": 8.865106978408478e-09, "loss": 0.2368, "step": 4025 }, { "epoch": 2.9624724061810155, "grad_norm": 1.713942294078811, "learning_rate": 8.533777744801352e-09, "loss": 0.2237, "step": 4026 }, { "epoch": 2.9632082413539367, "grad_norm": 1.8302715835481478, "learning_rate": 8.208755987346717e-09, "loss": 0.2254, "step": 4027 }, { "epoch": 2.963944076526858, "grad_norm": 1.6993683519483207, "learning_rate": 7.890041911223779e-09, "loss": 0.2666, "step": 4028 }, { "epoch": 2.964679911699779, "grad_norm": 1.6466132157017925, "learning_rate": 7.577635717633813e-09, "loss": 0.237, "step": 4029 }, { "epoch": 2.9654157468727007, "grad_norm": 1.7189614162663858, "learning_rate": 7.27153760379462e-09, "loss": 0.2264, "step": 4030 }, { "epoch": 2.966151582045622, "grad_norm": 1.7790381524509684, "learning_rate": 6.971747762941627e-09, "loss": 0.2643, "step": 4031 }, { "epoch": 2.966887417218543, "grad_norm": 1.6363940518886981, "learning_rate": 6.678266384327892e-09, "loss": 0.2548, "step": 4032 }, { "epoch": 2.9676232523914643, "grad_norm": 1.687120373733201, "learning_rate": 6.391093653224101e-09, "loss": 0.2508, "step": 4033 }, { "epoch": 2.9683590875643855, "grad_norm": 1.7114690617673218, "learning_rate": 6.110229750917463e-09, "loss": 0.2461, "step": 4034 }, { "epoch": 2.9690949227373067, "grad_norm": 1.6841098619772832, "learning_rate": 5.8356748547161445e-09, "loss": 0.2533, "step": 4035 }, { "epoch": 2.969830757910228, "grad_norm": 1.6961325930604967, "learning_rate": 5.567429137940395e-09, "loss": 0.2555, "step": 4036 }, { "epoch": 2.9705665930831495, "grad_norm": 1.778204538934733, "learning_rate": 5.30549276993142e-09, "loss": 0.2441, "step": 4037 }, { "epoch": 2.9713024282560707, "grad_norm": 1.7294888387637952, "learning_rate": 5.049865916045837e-09, "loss": 0.2293, "step": 4038 }, { "epoch": 2.972038263428992, "grad_norm": 1.6666115385959202, "learning_rate": 4.800548737656785e-09, "loss": 0.2296, "step": 4039 }, { "epoch": 2.972774098601913, "grad_norm": 1.68158053923391, "learning_rate": 4.5575413921561395e-09, "loss": 0.2429, "step": 4040 }, { "epoch": 2.9735099337748343, "grad_norm": 1.7304103386381005, "learning_rate": 4.320844032950078e-09, "loss": 0.2291, "step": 4041 }, { "epoch": 2.974245768947756, "grad_norm": 1.6755097655387852, "learning_rate": 4.090456809462407e-09, "loss": 0.2457, "step": 4042 }, { "epoch": 2.9749816041206767, "grad_norm": 1.7532414372386098, "learning_rate": 3.8663798671323415e-09, "loss": 0.2297, "step": 4043 }, { "epoch": 2.9757174392935983, "grad_norm": 1.5845404985726375, "learning_rate": 3.64861334741895e-09, "loss": 0.2255, "step": 4044 }, { "epoch": 2.9764532744665195, "grad_norm": 1.7739240720481877, "learning_rate": 3.4371573877944874e-09, "loss": 0.2279, "step": 4045 }, { "epoch": 2.9771891096394407, "grad_norm": 1.7688224764261355, "learning_rate": 3.2320121217477295e-09, "loss": 0.2436, "step": 4046 }, { "epoch": 2.977924944812362, "grad_norm": 1.7165583472458874, "learning_rate": 3.0331776787839718e-09, "loss": 0.2415, "step": 4047 }, { "epoch": 2.978660779985283, "grad_norm": 1.6624735466100926, "learning_rate": 2.840654184425029e-09, "loss": 0.2286, "step": 4048 }, { "epoch": 2.9793966151582048, "grad_norm": 1.7207393085565563, "learning_rate": 2.6544417602081263e-09, "loss": 0.2504, "step": 4049 }, { "epoch": 2.980132450331126, "grad_norm": 1.75649030639798, "learning_rate": 2.474540523687008e-09, "loss": 0.2421, "step": 4050 }, { "epoch": 2.980868285504047, "grad_norm": 1.7150199284345495, "learning_rate": 2.300950588430828e-09, "loss": 0.2147, "step": 4051 }, { "epoch": 2.9816041206769683, "grad_norm": 1.759292414404851, "learning_rate": 2.1336720640241504e-09, "loss": 0.2233, "step": 4052 }, { "epoch": 2.9823399558498895, "grad_norm": 1.7468789156851088, "learning_rate": 1.9727050560680584e-09, "loss": 0.2433, "step": 4053 }, { "epoch": 2.9830757910228107, "grad_norm": 1.7562236905334667, "learning_rate": 1.8180496661779346e-09, "loss": 0.2526, "step": 4054 }, { "epoch": 2.983811626195732, "grad_norm": 1.8331788126296142, "learning_rate": 1.669705991986792e-09, "loss": 0.2688, "step": 4055 }, { "epoch": 2.9845474613686536, "grad_norm": 1.7094097094096863, "learning_rate": 1.5276741271419427e-09, "loss": 0.2507, "step": 4056 }, { "epoch": 2.9852832965415748, "grad_norm": 1.620215153784787, "learning_rate": 1.391954161304998e-09, "loss": 0.2528, "step": 4057 }, { "epoch": 2.986019131714496, "grad_norm": 1.7346575427883826, "learning_rate": 1.2625461801540896e-09, "loss": 0.2703, "step": 4058 }, { "epoch": 2.986754966887417, "grad_norm": 1.772650656154652, "learning_rate": 1.139450265383868e-09, "loss": 0.2461, "step": 4059 }, { "epoch": 2.9874908020603383, "grad_norm": 1.6718380115946625, "learning_rate": 1.0226664947032838e-09, "loss": 0.2535, "step": 4060 }, { "epoch": 2.98822663723326, "grad_norm": 1.6889537536858779, "learning_rate": 9.121949418355869e-10, "loss": 0.228, "step": 4061 }, { "epoch": 2.988962472406181, "grad_norm": 1.6319492348321627, "learning_rate": 8.080356765194364e-10, "loss": 0.2243, "step": 4062 }, { "epoch": 2.9896983075791024, "grad_norm": 1.6772605492450758, "learning_rate": 7.101887645100114e-10, "loss": 0.2467, "step": 4063 }, { "epoch": 2.9904341427520236, "grad_norm": 1.6977583048023055, "learning_rate": 6.186542675767903e-10, "loss": 0.2515, "step": 4064 }, { "epoch": 2.9911699779249448, "grad_norm": 1.6878007777937696, "learning_rate": 5.33432243503551e-10, "loss": 0.2265, "step": 4065 }, { "epoch": 2.991905813097866, "grad_norm": 1.8537993807280682, "learning_rate": 4.54522746090591e-10, "loss": 0.2362, "step": 4066 }, { "epoch": 2.992641648270787, "grad_norm": 1.8209215118989313, "learning_rate": 3.819258251525071e-10, "loss": 0.2297, "step": 4067 }, { "epoch": 2.993377483443709, "grad_norm": 1.6575686230345335, "learning_rate": 3.1564152651708536e-10, "loss": 0.2527, "step": 4068 }, { "epoch": 2.99411331861663, "grad_norm": 1.7567506336143794, "learning_rate": 2.55669892030852e-10, "loss": 0.2667, "step": 4069 }, { "epoch": 2.994849153789551, "grad_norm": 1.7438553657569038, "learning_rate": 2.0201095955130202e-10, "loss": 0.2398, "step": 4070 }, { "epoch": 2.9955849889624724, "grad_norm": 1.6003878280216197, "learning_rate": 1.5466476295356025e-10, "loss": 0.2342, "step": 4071 }, { "epoch": 2.9963208241353936, "grad_norm": 1.6851923895112628, "learning_rate": 1.1363133212705102e-10, "loss": 0.2462, "step": 4072 }, { "epoch": 2.9970566593083148, "grad_norm": 1.7021071978643727, "learning_rate": 7.891069297438769e-11, "loss": 0.2317, "step": 4073 }, { "epoch": 2.997792494481236, "grad_norm": 1.6733195777196905, "learning_rate": 5.0502867414703447e-11, "loss": 0.2336, "step": 4074 }, { "epoch": 2.9985283296541576, "grad_norm": 1.786778561090688, "learning_rate": 2.840787338254103e-11, "loss": 0.2406, "step": 4075 }, { "epoch": 2.999264164827079, "grad_norm": 1.6387755046477843, "learning_rate": 1.2625724824522068e-11, "loss": 0.2418, "step": 4076 }, { "epoch": 3.0, "grad_norm": 1.5633601149576242, "learning_rate": 3.156431703787988e-12, "loss": 0.2452, "step": 4077 }, { "epoch": 3.0, "step": 4077, "total_flos": 3502095876489216.0, "train_loss": 0.5817589285087433, "train_runtime": 15713.503, "train_samples_per_second": 33.219, "train_steps_per_second": 0.259 } ], "logging_steps": 1.0, "max_steps": 4077, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 816, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3502095876489216.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }