diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,12385 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 1765, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "grad_norm": 3.4841845237449234, + "learning_rate": 3.7735849056603774e-08, + "loss": 0.6308, + "step": 1 + }, + { + "epoch": 0.0, + "grad_norm": 4.42203786668807, + "learning_rate": 7.547169811320755e-08, + "loss": 0.6758, + "step": 2 + }, + { + "epoch": 0.0, + "grad_norm": 3.878145696353976, + "learning_rate": 1.1320754716981131e-07, + "loss": 0.6391, + "step": 3 + }, + { + "epoch": 0.0, + "grad_norm": 3.475344841495122, + "learning_rate": 1.509433962264151e-07, + "loss": 0.6518, + "step": 4 + }, + { + "epoch": 0.0, + "grad_norm": 3.4577169819897264, + "learning_rate": 1.8867924528301886e-07, + "loss": 0.6625, + "step": 5 + }, + { + "epoch": 0.0, + "grad_norm": 3.6314657760115416, + "learning_rate": 2.2641509433962263e-07, + "loss": 0.6405, + "step": 6 + }, + { + "epoch": 0.0, + "grad_norm": 3.2382705933247515, + "learning_rate": 2.641509433962264e-07, + "loss": 0.6715, + "step": 7 + }, + { + "epoch": 0.0, + "grad_norm": 5.678667419114082, + "learning_rate": 3.018867924528302e-07, + "loss": 0.6151, + "step": 8 + }, + { + "epoch": 0.01, + "grad_norm": 3.262379969483797, + "learning_rate": 3.396226415094339e-07, + "loss": 0.6107, + "step": 9 + }, + { + "epoch": 0.01, + "grad_norm": 4.489917095230589, + "learning_rate": 3.773584905660377e-07, + "loss": 0.626, + "step": 10 + }, + { + "epoch": 0.01, + "grad_norm": 3.4125483935497667, + "learning_rate": 4.150943396226415e-07, + "loss": 0.6132, + "step": 11 + }, + { + "epoch": 0.01, + "grad_norm": 3.79320406562782, + "learning_rate": 4.5283018867924526e-07, + "loss": 0.657, + "step": 12 + }, + { + "epoch": 0.01, + "grad_norm": 3.2675195669438626, + "learning_rate": 4.90566037735849e-07, + "loss": 0.6231, + "step": 13 + }, + { + "epoch": 0.01, + "grad_norm": 3.323174849476731, + "learning_rate": 5.283018867924528e-07, + "loss": 0.6052, + "step": 14 + }, + { + "epoch": 0.01, + "grad_norm": 2.924763211870166, + "learning_rate": 5.660377358490566e-07, + "loss": 0.5834, + "step": 15 + }, + { + "epoch": 0.01, + "grad_norm": 3.35179261011413, + "learning_rate": 6.037735849056604e-07, + "loss": 0.5923, + "step": 16 + }, + { + "epoch": 0.01, + "grad_norm": 3.334147599460606, + "learning_rate": 6.415094339622641e-07, + "loss": 0.5945, + "step": 17 + }, + { + "epoch": 0.01, + "grad_norm": 2.9466181761596992, + "learning_rate": 6.792452830188678e-07, + "loss": 0.5917, + "step": 18 + }, + { + "epoch": 0.01, + "grad_norm": 3.017575110561897, + "learning_rate": 7.169811320754716e-07, + "loss": 0.5754, + "step": 19 + }, + { + "epoch": 0.01, + "grad_norm": 3.229917335932235, + "learning_rate": 7.547169811320754e-07, + "loss": 0.5618, + "step": 20 + }, + { + "epoch": 0.01, + "grad_norm": 2.8131350538456132, + "learning_rate": 7.924528301886793e-07, + "loss": 0.572, + "step": 21 + }, + { + "epoch": 0.01, + "grad_norm": 2.72194929279167, + "learning_rate": 8.30188679245283e-07, + "loss": 0.513, + "step": 22 + }, + { + "epoch": 0.01, + "grad_norm": 2.3394944611913, + "learning_rate": 8.679245283018868e-07, + "loss": 0.457, + "step": 23 + }, + { + "epoch": 0.01, + "grad_norm": 2.931979495495408, + "learning_rate": 9.056603773584905e-07, + "loss": 0.4427, + "step": 24 + }, + { + "epoch": 0.01, + "grad_norm": 2.887347811966607, + "learning_rate": 9.433962264150943e-07, + "loss": 0.489, + "step": 25 + }, + { + "epoch": 0.01, + "grad_norm": 2.1514983983383695, + "learning_rate": 9.81132075471698e-07, + "loss": 0.4522, + "step": 26 + }, + { + "epoch": 0.02, + "grad_norm": 2.2995536440652637, + "learning_rate": 1.0188679245283019e-06, + "loss": 0.4751, + "step": 27 + }, + { + "epoch": 0.02, + "grad_norm": 2.1446285442921735, + "learning_rate": 1.0566037735849057e-06, + "loss": 0.44, + "step": 28 + }, + { + "epoch": 0.02, + "grad_norm": 2.2651023185345585, + "learning_rate": 1.0943396226415093e-06, + "loss": 0.4708, + "step": 29 + }, + { + "epoch": 0.02, + "grad_norm": 2.021080464753173, + "learning_rate": 1.1320754716981131e-06, + "loss": 0.4208, + "step": 30 + }, + { + "epoch": 0.02, + "grad_norm": 2.21584242208253, + "learning_rate": 1.169811320754717e-06, + "loss": 0.4305, + "step": 31 + }, + { + "epoch": 0.02, + "grad_norm": 2.3964757157004613, + "learning_rate": 1.2075471698113208e-06, + "loss": 0.4247, + "step": 32 + }, + { + "epoch": 0.02, + "grad_norm": 3.033542339756139, + "learning_rate": 1.2452830188679244e-06, + "loss": 0.3917, + "step": 33 + }, + { + "epoch": 0.02, + "grad_norm": 2.2140208027667616, + "learning_rate": 1.2830188679245282e-06, + "loss": 0.3896, + "step": 34 + }, + { + "epoch": 0.02, + "grad_norm": 2.2491772430285413, + "learning_rate": 1.320754716981132e-06, + "loss": 0.3995, + "step": 35 + }, + { + "epoch": 0.02, + "grad_norm": 2.4868984223931907, + "learning_rate": 1.3584905660377356e-06, + "loss": 0.3922, + "step": 36 + }, + { + "epoch": 0.02, + "grad_norm": 1.936398949468111, + "learning_rate": 1.3962264150943394e-06, + "loss": 0.4054, + "step": 37 + }, + { + "epoch": 0.02, + "grad_norm": 2.1541548439219653, + "learning_rate": 1.4339622641509432e-06, + "loss": 0.4242, + "step": 38 + }, + { + "epoch": 0.02, + "grad_norm": 1.9275413032125785, + "learning_rate": 1.471698113207547e-06, + "loss": 0.3636, + "step": 39 + }, + { + "epoch": 0.02, + "grad_norm": 1.8578963289661783, + "learning_rate": 1.5094339622641509e-06, + "loss": 0.4057, + "step": 40 + }, + { + "epoch": 0.02, + "grad_norm": 1.6999808011479187, + "learning_rate": 1.5471698113207547e-06, + "loss": 0.355, + "step": 41 + }, + { + "epoch": 0.02, + "grad_norm": 2.1149988645342224, + "learning_rate": 1.5849056603773585e-06, + "loss": 0.3504, + "step": 42 + }, + { + "epoch": 0.02, + "grad_norm": 1.791588524547875, + "learning_rate": 1.6226415094339623e-06, + "loss": 0.3673, + "step": 43 + }, + { + "epoch": 0.02, + "grad_norm": 1.8722057969802204, + "learning_rate": 1.660377358490566e-06, + "loss": 0.3671, + "step": 44 + }, + { + "epoch": 0.03, + "grad_norm": 2.2949583830173808, + "learning_rate": 1.6981132075471698e-06, + "loss": 0.3593, + "step": 45 + }, + { + "epoch": 0.03, + "grad_norm": 1.911131097602587, + "learning_rate": 1.7358490566037736e-06, + "loss": 0.3596, + "step": 46 + }, + { + "epoch": 0.03, + "grad_norm": 1.911161418733667, + "learning_rate": 1.7735849056603774e-06, + "loss": 0.3479, + "step": 47 + }, + { + "epoch": 0.03, + "grad_norm": 3.1118991017606383, + "learning_rate": 1.811320754716981e-06, + "loss": 0.3528, + "step": 48 + }, + { + "epoch": 0.03, + "grad_norm": 2.392602083325391, + "learning_rate": 1.8490566037735848e-06, + "loss": 0.3481, + "step": 49 + }, + { + "epoch": 0.03, + "grad_norm": 1.6918804428497827, + "learning_rate": 1.8867924528301887e-06, + "loss": 0.3295, + "step": 50 + }, + { + "epoch": 0.03, + "grad_norm": 1.6445974760767799, + "learning_rate": 1.9245283018867923e-06, + "loss": 0.3403, + "step": 51 + }, + { + "epoch": 0.03, + "grad_norm": 1.7998353066606536, + "learning_rate": 1.962264150943396e-06, + "loss": 0.3303, + "step": 52 + }, + { + "epoch": 0.03, + "grad_norm": 1.9466476774565185, + "learning_rate": 2e-06, + "loss": 0.3311, + "step": 53 + }, + { + "epoch": 0.03, + "grad_norm": 1.6838405510353802, + "learning_rate": 1.999998316310098e-06, + "loss": 0.3131, + "step": 54 + }, + { + "epoch": 0.03, + "grad_norm": 1.8674267622877876, + "learning_rate": 1.999993265246062e-06, + "loss": 0.3392, + "step": 55 + }, + { + "epoch": 0.03, + "grad_norm": 1.6346441469873199, + "learning_rate": 1.9999848468249006e-06, + "loss": 0.3276, + "step": 56 + }, + { + "epoch": 0.03, + "grad_norm": 1.4729957226872485, + "learning_rate": 1.999973061074962e-06, + "loss": 0.3436, + "step": 57 + }, + { + "epoch": 0.03, + "grad_norm": 3.078293012985477, + "learning_rate": 1.9999579080359334e-06, + "loss": 0.3347, + "step": 58 + }, + { + "epoch": 0.03, + "grad_norm": 2.297753681732222, + "learning_rate": 1.9999393877588407e-06, + "loss": 0.324, + "step": 59 + }, + { + "epoch": 0.03, + "grad_norm": 1.9264664474469204, + "learning_rate": 1.9999175003060483e-06, + "loss": 0.3362, + "step": 60 + }, + { + "epoch": 0.03, + "grad_norm": 2.0149148870938918, + "learning_rate": 1.9998922457512605e-06, + "loss": 0.3369, + "step": 61 + }, + { + "epoch": 0.04, + "grad_norm": 2.0618748458976106, + "learning_rate": 1.9998636241795183e-06, + "loss": 0.3308, + "step": 62 + }, + { + "epoch": 0.04, + "grad_norm": 1.8461171167550716, + "learning_rate": 1.999831635687201e-06, + "loss": 0.337, + "step": 63 + }, + { + "epoch": 0.04, + "grad_norm": 2.573323778446205, + "learning_rate": 1.999796280382027e-06, + "loss": 0.3035, + "step": 64 + }, + { + "epoch": 0.04, + "grad_norm": 1.9297388937697657, + "learning_rate": 1.9997575583830506e-06, + "loss": 0.3411, + "step": 65 + }, + { + "epoch": 0.04, + "grad_norm": 1.988741267914501, + "learning_rate": 1.9997154698206633e-06, + "loss": 0.3123, + "step": 66 + }, + { + "epoch": 0.04, + "grad_norm": 1.5698796744326022, + "learning_rate": 1.9996700148365936e-06, + "loss": 0.3086, + "step": 67 + }, + { + "epoch": 0.04, + "grad_norm": 1.4267305806521715, + "learning_rate": 1.999621193583905e-06, + "loss": 0.303, + "step": 68 + }, + { + "epoch": 0.04, + "grad_norm": 1.920967222581801, + "learning_rate": 1.9995690062269984e-06, + "loss": 0.3234, + "step": 69 + }, + { + "epoch": 0.04, + "grad_norm": 1.3938639600370097, + "learning_rate": 1.9995134529416073e-06, + "loss": 0.3083, + "step": 70 + }, + { + "epoch": 0.04, + "grad_norm": 1.819204014872305, + "learning_rate": 1.9994545339148013e-06, + "loss": 0.3146, + "step": 71 + }, + { + "epoch": 0.04, + "grad_norm": 2.360151055520758, + "learning_rate": 1.999392249344983e-06, + "loss": 0.3118, + "step": 72 + }, + { + "epoch": 0.04, + "grad_norm": 1.6417160504347441, + "learning_rate": 1.9993265994418883e-06, + "loss": 0.3251, + "step": 73 + }, + { + "epoch": 0.04, + "grad_norm": 1.7145584523487125, + "learning_rate": 1.9992575844265856e-06, + "loss": 0.3208, + "step": 74 + }, + { + "epoch": 0.04, + "grad_norm": 1.6072756512379518, + "learning_rate": 1.999185204531474e-06, + "loss": 0.2892, + "step": 75 + }, + { + "epoch": 0.04, + "grad_norm": 1.5028894089334375, + "learning_rate": 1.9991094600002847e-06, + "loss": 0.3396, + "step": 76 + }, + { + "epoch": 0.04, + "grad_norm": 1.526073412721908, + "learning_rate": 1.999030351088078e-06, + "loss": 0.3083, + "step": 77 + }, + { + "epoch": 0.04, + "grad_norm": 1.5290303153461566, + "learning_rate": 1.9989478780612433e-06, + "loss": 0.3026, + "step": 78 + }, + { + "epoch": 0.04, + "grad_norm": 1.6864641718579354, + "learning_rate": 1.998862041197499e-06, + "loss": 0.313, + "step": 79 + }, + { + "epoch": 0.05, + "grad_norm": 2.1460543565197776, + "learning_rate": 1.998772840785891e-06, + "loss": 0.2943, + "step": 80 + }, + { + "epoch": 0.05, + "grad_norm": 1.9579058234848254, + "learning_rate": 1.99868027712679e-06, + "loss": 0.3102, + "step": 81 + }, + { + "epoch": 0.05, + "grad_norm": 1.6024600368644362, + "learning_rate": 1.998584350531894e-06, + "loss": 0.3068, + "step": 82 + }, + { + "epoch": 0.05, + "grad_norm": 1.5500158132637425, + "learning_rate": 1.998485061324223e-06, + "loss": 0.3188, + "step": 83 + }, + { + "epoch": 0.05, + "grad_norm": 1.5721528341876851, + "learning_rate": 1.9983824098381224e-06, + "loss": 0.3115, + "step": 84 + }, + { + "epoch": 0.05, + "grad_norm": 1.4419242555613345, + "learning_rate": 1.9982763964192584e-06, + "loss": 0.3028, + "step": 85 + }, + { + "epoch": 0.05, + "grad_norm": 1.4989271888893556, + "learning_rate": 1.9981670214246184e-06, + "loss": 0.2916, + "step": 86 + }, + { + "epoch": 0.05, + "grad_norm": 2.999740923295262, + "learning_rate": 1.9980542852225097e-06, + "loss": 0.2996, + "step": 87 + }, + { + "epoch": 0.05, + "grad_norm": 1.6622698607673714, + "learning_rate": 1.997938188192558e-06, + "loss": 0.3124, + "step": 88 + }, + { + "epoch": 0.05, + "grad_norm": 1.6648545623356927, + "learning_rate": 1.997818730725706e-06, + "loss": 0.3117, + "step": 89 + }, + { + "epoch": 0.05, + "grad_norm": 1.4643519839399555, + "learning_rate": 1.9976959132242124e-06, + "loss": 0.3219, + "step": 90 + }, + { + "epoch": 0.05, + "grad_norm": 1.8554720628552361, + "learning_rate": 1.997569736101651e-06, + "loss": 0.2664, + "step": 91 + }, + { + "epoch": 0.05, + "grad_norm": 1.5102712470587052, + "learning_rate": 1.997440199782906e-06, + "loss": 0.2933, + "step": 92 + }, + { + "epoch": 0.05, + "grad_norm": 1.6419233696009894, + "learning_rate": 1.9973073047041776e-06, + "loss": 0.2904, + "step": 93 + }, + { + "epoch": 0.05, + "grad_norm": 1.763124553627459, + "learning_rate": 1.997171051312973e-06, + "loss": 0.2852, + "step": 94 + }, + { + "epoch": 0.05, + "grad_norm": 2.0291891591521485, + "learning_rate": 1.9970314400681097e-06, + "loss": 0.2702, + "step": 95 + }, + { + "epoch": 0.05, + "grad_norm": 1.9132613974926687, + "learning_rate": 1.996888471439711e-06, + "loss": 0.2993, + "step": 96 + }, + { + "epoch": 0.05, + "grad_norm": 1.6052018477653207, + "learning_rate": 1.9967421459092075e-06, + "loss": 0.2912, + "step": 97 + }, + { + "epoch": 0.06, + "grad_norm": 1.6611575462934254, + "learning_rate": 1.9965924639693317e-06, + "loss": 0.2803, + "step": 98 + }, + { + "epoch": 0.06, + "grad_norm": 1.6460993083091668, + "learning_rate": 1.9964394261241205e-06, + "loss": 0.2925, + "step": 99 + }, + { + "epoch": 0.06, + "grad_norm": 1.7929937658011097, + "learning_rate": 1.9962830328889104e-06, + "loss": 0.3104, + "step": 100 + }, + { + "epoch": 0.06, + "grad_norm": 2.7958338225406956, + "learning_rate": 1.996123284790336e-06, + "loss": 0.3095, + "step": 101 + }, + { + "epoch": 0.06, + "grad_norm": 1.7811495197980376, + "learning_rate": 1.99596018236633e-06, + "loss": 0.277, + "step": 102 + }, + { + "epoch": 0.06, + "grad_norm": 1.5261310813457682, + "learning_rate": 1.9957937261661213e-06, + "loss": 0.3078, + "step": 103 + }, + { + "epoch": 0.06, + "grad_norm": 2.0011507608468713, + "learning_rate": 1.99562391675023e-06, + "loss": 0.3086, + "step": 104 + }, + { + "epoch": 0.06, + "grad_norm": 1.6289150895375621, + "learning_rate": 1.9954507546904696e-06, + "loss": 0.3027, + "step": 105 + }, + { + "epoch": 0.06, + "grad_norm": 1.59315878335245, + "learning_rate": 1.995274240569942e-06, + "loss": 0.3002, + "step": 106 + }, + { + "epoch": 0.06, + "grad_norm": 1.493001623817555, + "learning_rate": 1.995094374983038e-06, + "loss": 0.2672, + "step": 107 + }, + { + "epoch": 0.06, + "grad_norm": 1.7264192215908158, + "learning_rate": 1.994911158535432e-06, + "loss": 0.2908, + "step": 108 + }, + { + "epoch": 0.06, + "grad_norm": 1.5788937616396206, + "learning_rate": 1.994724591844085e-06, + "loss": 0.295, + "step": 109 + }, + { + "epoch": 0.06, + "grad_norm": 1.658948012517768, + "learning_rate": 1.9945346755372364e-06, + "loss": 0.2807, + "step": 110 + }, + { + "epoch": 0.06, + "grad_norm": 2.1955561337720515, + "learning_rate": 1.9943414102544082e-06, + "loss": 0.3127, + "step": 111 + }, + { + "epoch": 0.06, + "grad_norm": 2.151960308631923, + "learning_rate": 1.9941447966463962e-06, + "loss": 0.3222, + "step": 112 + }, + { + "epoch": 0.06, + "grad_norm": 1.8981882303625344, + "learning_rate": 1.9939448353752743e-06, + "loss": 0.2892, + "step": 113 + }, + { + "epoch": 0.06, + "grad_norm": 1.9733942899011068, + "learning_rate": 1.9937415271143876e-06, + "loss": 0.3066, + "step": 114 + }, + { + "epoch": 0.07, + "grad_norm": 1.5952639087949956, + "learning_rate": 1.9935348725483525e-06, + "loss": 0.2694, + "step": 115 + }, + { + "epoch": 0.07, + "grad_norm": 1.6956577438970666, + "learning_rate": 1.9933248723730533e-06, + "loss": 0.2826, + "step": 116 + }, + { + "epoch": 0.07, + "grad_norm": 3.515417258009173, + "learning_rate": 1.9931115272956404e-06, + "loss": 0.2838, + "step": 117 + }, + { + "epoch": 0.07, + "grad_norm": 2.1533905590726836, + "learning_rate": 1.9928948380345273e-06, + "loss": 0.3044, + "step": 118 + }, + { + "epoch": 0.07, + "grad_norm": 2.414955521736467, + "learning_rate": 1.9926748053193894e-06, + "loss": 0.2866, + "step": 119 + }, + { + "epoch": 0.07, + "grad_norm": 1.8300861326639828, + "learning_rate": 1.9924514298911607e-06, + "loss": 0.2849, + "step": 120 + }, + { + "epoch": 0.07, + "grad_norm": 1.3507452847434371, + "learning_rate": 1.9922247125020305e-06, + "loss": 0.2736, + "step": 121 + }, + { + "epoch": 0.07, + "grad_norm": 1.7294482029952782, + "learning_rate": 1.9919946539154424e-06, + "loss": 0.2693, + "step": 122 + }, + { + "epoch": 0.07, + "grad_norm": 5.059252409832772, + "learning_rate": 1.9917612549060915e-06, + "loss": 0.2834, + "step": 123 + }, + { + "epoch": 0.07, + "grad_norm": 2.075710501140192, + "learning_rate": 1.991524516259921e-06, + "loss": 0.2755, + "step": 124 + }, + { + "epoch": 0.07, + "grad_norm": 1.8089419594122498, + "learning_rate": 1.991284438774119e-06, + "loss": 0.2929, + "step": 125 + }, + { + "epoch": 0.07, + "grad_norm": 1.6364760906705336, + "learning_rate": 1.9910410232571185e-06, + "loss": 0.266, + "step": 126 + }, + { + "epoch": 0.07, + "grad_norm": 1.6641452879186178, + "learning_rate": 1.9907942705285916e-06, + "loss": 0.2909, + "step": 127 + }, + { + "epoch": 0.07, + "grad_norm": 2.075940524759441, + "learning_rate": 1.990544181419448e-06, + "loss": 0.3055, + "step": 128 + }, + { + "epoch": 0.07, + "grad_norm": 1.6140752095755138, + "learning_rate": 1.9902907567718335e-06, + "loss": 0.2711, + "step": 129 + }, + { + "epoch": 0.07, + "grad_norm": 1.8458301967854023, + "learning_rate": 1.9900339974391247e-06, + "loss": 0.2915, + "step": 130 + }, + { + "epoch": 0.07, + "grad_norm": 2.5345258940635396, + "learning_rate": 1.9897739042859284e-06, + "loss": 0.285, + "step": 131 + }, + { + "epoch": 0.07, + "grad_norm": 2.313563333374535, + "learning_rate": 1.989510478188076e-06, + "loss": 0.2824, + "step": 132 + }, + { + "epoch": 0.08, + "grad_norm": 2.6660602786871563, + "learning_rate": 1.9892437200326236e-06, + "loss": 0.2885, + "step": 133 + }, + { + "epoch": 0.08, + "grad_norm": 1.757098250149227, + "learning_rate": 1.988973630717848e-06, + "loss": 0.2992, + "step": 134 + }, + { + "epoch": 0.08, + "grad_norm": 1.4219607783010701, + "learning_rate": 1.988700211153241e-06, + "loss": 0.2711, + "step": 135 + }, + { + "epoch": 0.08, + "grad_norm": 1.610322300712251, + "learning_rate": 1.9884234622595115e-06, + "loss": 0.3245, + "step": 136 + }, + { + "epoch": 0.08, + "grad_norm": 1.7034865974157456, + "learning_rate": 1.988143384968578e-06, + "loss": 0.2937, + "step": 137 + }, + { + "epoch": 0.08, + "grad_norm": 2.1472558512900806, + "learning_rate": 1.987859980223566e-06, + "loss": 0.2948, + "step": 138 + }, + { + "epoch": 0.08, + "grad_norm": 2.106917478213908, + "learning_rate": 1.9875732489788077e-06, + "loss": 0.2847, + "step": 139 + }, + { + "epoch": 0.08, + "grad_norm": 1.7261907563645862, + "learning_rate": 1.9872831921998367e-06, + "loss": 0.2714, + "step": 140 + }, + { + "epoch": 0.08, + "grad_norm": 3.227693070439494, + "learning_rate": 1.9869898108633835e-06, + "loss": 0.2761, + "step": 141 + }, + { + "epoch": 0.08, + "grad_norm": 1.3945652943122027, + "learning_rate": 1.9866931059573745e-06, + "loss": 0.2681, + "step": 142 + }, + { + "epoch": 0.08, + "grad_norm": 1.5682375981410632, + "learning_rate": 1.9863930784809278e-06, + "loss": 0.2869, + "step": 143 + }, + { + "epoch": 0.08, + "grad_norm": 1.6738297542661662, + "learning_rate": 1.9860897294443508e-06, + "loss": 0.3001, + "step": 144 + }, + { + "epoch": 0.08, + "grad_norm": 1.8476169282688792, + "learning_rate": 1.9857830598691335e-06, + "loss": 0.2943, + "step": 145 + }, + { + "epoch": 0.08, + "grad_norm": 1.585993793933756, + "learning_rate": 1.98547307078795e-06, + "loss": 0.2957, + "step": 146 + }, + { + "epoch": 0.08, + "grad_norm": 1.6324207123669265, + "learning_rate": 1.9851597632446505e-06, + "loss": 0.2728, + "step": 147 + }, + { + "epoch": 0.08, + "grad_norm": 1.3897614522732709, + "learning_rate": 1.9848431382942605e-06, + "loss": 0.2816, + "step": 148 + }, + { + "epoch": 0.08, + "grad_norm": 1.670002539945534, + "learning_rate": 1.9845231970029773e-06, + "loss": 0.2673, + "step": 149 + }, + { + "epoch": 0.08, + "grad_norm": 1.3998951955898702, + "learning_rate": 1.9841999404481636e-06, + "loss": 0.2835, + "step": 150 + }, + { + "epoch": 0.09, + "grad_norm": 1.8050324087254177, + "learning_rate": 1.983873369718348e-06, + "loss": 0.2955, + "step": 151 + }, + { + "epoch": 0.09, + "grad_norm": 1.6122607548590895, + "learning_rate": 1.983543485913218e-06, + "loss": 0.2651, + "step": 152 + }, + { + "epoch": 0.09, + "grad_norm": 5.636568772826462, + "learning_rate": 1.9832102901436167e-06, + "loss": 0.292, + "step": 153 + }, + { + "epoch": 0.09, + "grad_norm": 1.6032423566289615, + "learning_rate": 1.982873783531542e-06, + "loss": 0.2866, + "step": 154 + }, + { + "epoch": 0.09, + "grad_norm": 1.6140870587436835, + "learning_rate": 1.982533967210139e-06, + "loss": 0.2917, + "step": 155 + }, + { + "epoch": 0.09, + "grad_norm": 1.848769044209177, + "learning_rate": 1.9821908423236977e-06, + "loss": 0.2877, + "step": 156 + }, + { + "epoch": 0.09, + "grad_norm": 2.637509507675195, + "learning_rate": 1.9818444100276516e-06, + "loss": 0.2815, + "step": 157 + }, + { + "epoch": 0.09, + "grad_norm": 4.847148550709564, + "learning_rate": 1.981494671488568e-06, + "loss": 0.2647, + "step": 158 + }, + { + "epoch": 0.09, + "grad_norm": 1.457905784519377, + "learning_rate": 1.9811416278841502e-06, + "loss": 0.2788, + "step": 159 + }, + { + "epoch": 0.09, + "grad_norm": 1.4596375307905933, + "learning_rate": 1.98078528040323e-06, + "loss": 0.271, + "step": 160 + }, + { + "epoch": 0.09, + "grad_norm": 1.8185350940487643, + "learning_rate": 1.980425630245765e-06, + "loss": 0.2775, + "step": 161 + }, + { + "epoch": 0.09, + "grad_norm": 1.478348852209309, + "learning_rate": 1.9800626786228337e-06, + "loss": 0.2688, + "step": 162 + }, + { + "epoch": 0.09, + "grad_norm": 1.9825985671375521, + "learning_rate": 1.979696426756632e-06, + "loss": 0.2637, + "step": 163 + }, + { + "epoch": 0.09, + "grad_norm": 1.5013341808242229, + "learning_rate": 1.9793268758804687e-06, + "loss": 0.2679, + "step": 164 + }, + { + "epoch": 0.09, + "grad_norm": 2.5816863644661754, + "learning_rate": 1.9789540272387627e-06, + "loss": 0.268, + "step": 165 + }, + { + "epoch": 0.09, + "grad_norm": 1.8478316508698556, + "learning_rate": 1.978577882087036e-06, + "loss": 0.2837, + "step": 166 + }, + { + "epoch": 0.09, + "grad_norm": 2.6922064925766565, + "learning_rate": 1.9781984416919134e-06, + "loss": 0.2783, + "step": 167 + }, + { + "epoch": 0.1, + "grad_norm": 1.936362518787055, + "learning_rate": 1.977815707331114e-06, + "loss": 0.297, + "step": 168 + }, + { + "epoch": 0.1, + "grad_norm": 1.545874439153469, + "learning_rate": 1.97742968029345e-06, + "loss": 0.2907, + "step": 169 + }, + { + "epoch": 0.1, + "grad_norm": 1.8160738783954211, + "learning_rate": 1.977040361878821e-06, + "loss": 0.2721, + "step": 170 + }, + { + "epoch": 0.1, + "grad_norm": 2.807244593846061, + "learning_rate": 1.9766477533982094e-06, + "loss": 0.2821, + "step": 171 + }, + { + "epoch": 0.1, + "grad_norm": 1.7274195143294735, + "learning_rate": 1.976251856173678e-06, + "loss": 0.284, + "step": 172 + }, + { + "epoch": 0.1, + "grad_norm": 2.0244077785058754, + "learning_rate": 1.9758526715383627e-06, + "loss": 0.2846, + "step": 173 + }, + { + "epoch": 0.1, + "grad_norm": 2.024299787729565, + "learning_rate": 1.9754502008364698e-06, + "loss": 0.2926, + "step": 174 + }, + { + "epoch": 0.1, + "grad_norm": 1.7788767121304085, + "learning_rate": 1.975044445423271e-06, + "loss": 0.2941, + "step": 175 + }, + { + "epoch": 0.1, + "grad_norm": 1.4019859175935807, + "learning_rate": 1.974635406665099e-06, + "loss": 0.2761, + "step": 176 + }, + { + "epoch": 0.1, + "grad_norm": 2.2597958220548073, + "learning_rate": 1.974223085939343e-06, + "loss": 0.2623, + "step": 177 + }, + { + "epoch": 0.1, + "grad_norm": 1.5056594955124902, + "learning_rate": 1.9738074846344426e-06, + "loss": 0.2884, + "step": 178 + }, + { + "epoch": 0.1, + "grad_norm": 1.8849616223824353, + "learning_rate": 1.9733886041498856e-06, + "loss": 0.2744, + "step": 179 + }, + { + "epoch": 0.1, + "grad_norm": 2.2219466195904114, + "learning_rate": 1.972966445896202e-06, + "loss": 0.271, + "step": 180 + }, + { + "epoch": 0.1, + "grad_norm": 1.42849060120362, + "learning_rate": 1.972541011294959e-06, + "loss": 0.2735, + "step": 181 + }, + { + "epoch": 0.1, + "grad_norm": 2.2423472765799866, + "learning_rate": 1.972112301778756e-06, + "loss": 0.2977, + "step": 182 + }, + { + "epoch": 0.1, + "grad_norm": 1.6359840805861086, + "learning_rate": 1.9716803187912203e-06, + "loss": 0.2774, + "step": 183 + }, + { + "epoch": 0.1, + "grad_norm": 2.8153625109767515, + "learning_rate": 1.9712450637870045e-06, + "loss": 0.2919, + "step": 184 + }, + { + "epoch": 0.1, + "grad_norm": 2.0535893364579882, + "learning_rate": 1.970806538231776e-06, + "loss": 0.284, + "step": 185 + }, + { + "epoch": 0.11, + "grad_norm": 2.5494217943723756, + "learning_rate": 1.970364743602217e-06, + "loss": 0.2693, + "step": 186 + }, + { + "epoch": 0.11, + "grad_norm": 2.6549806664227336, + "learning_rate": 1.9699196813860184e-06, + "loss": 0.259, + "step": 187 + }, + { + "epoch": 0.11, + "grad_norm": 1.5769903194256722, + "learning_rate": 1.9694713530818735e-06, + "loss": 0.2611, + "step": 188 + }, + { + "epoch": 0.11, + "grad_norm": 1.8669250010493135, + "learning_rate": 1.969019760199474e-06, + "loss": 0.2828, + "step": 189 + }, + { + "epoch": 0.11, + "grad_norm": 1.795636962867612, + "learning_rate": 1.9685649042595045e-06, + "loss": 0.2832, + "step": 190 + }, + { + "epoch": 0.11, + "grad_norm": 1.645799112758709, + "learning_rate": 1.968106786793638e-06, + "loss": 0.2654, + "step": 191 + }, + { + "epoch": 0.11, + "grad_norm": 2.0536811264442205, + "learning_rate": 1.967645409344529e-06, + "loss": 0.2485, + "step": 192 + }, + { + "epoch": 0.11, + "grad_norm": 1.5843372589463587, + "learning_rate": 1.967180773465812e-06, + "loss": 0.2638, + "step": 193 + }, + { + "epoch": 0.11, + "grad_norm": 1.6189891086416026, + "learning_rate": 1.9667128807220916e-06, + "loss": 0.2764, + "step": 194 + }, + { + "epoch": 0.11, + "grad_norm": 1.618868426016012, + "learning_rate": 1.9662417326889407e-06, + "loss": 0.274, + "step": 195 + }, + { + "epoch": 0.11, + "grad_norm": 1.5563594433614312, + "learning_rate": 1.9657673309528936e-06, + "loss": 0.2832, + "step": 196 + }, + { + "epoch": 0.11, + "grad_norm": 1.903348639321686, + "learning_rate": 1.9652896771114414e-06, + "loss": 0.2674, + "step": 197 + }, + { + "epoch": 0.11, + "grad_norm": 4.299579541828807, + "learning_rate": 1.964808772773025e-06, + "loss": 0.2713, + "step": 198 + }, + { + "epoch": 0.11, + "grad_norm": 1.955556819058435, + "learning_rate": 1.9643246195570335e-06, + "loss": 0.2754, + "step": 199 + }, + { + "epoch": 0.11, + "grad_norm": 1.4246064872132271, + "learning_rate": 1.9638372190937934e-06, + "loss": 0.2804, + "step": 200 + }, + { + "epoch": 0.11, + "grad_norm": 1.7997151500172972, + "learning_rate": 1.963346573024568e-06, + "loss": 0.2711, + "step": 201 + }, + { + "epoch": 0.11, + "grad_norm": 1.8164292268565017, + "learning_rate": 1.962852683001548e-06, + "loss": 0.2841, + "step": 202 + }, + { + "epoch": 0.12, + "grad_norm": 1.8738622717571962, + "learning_rate": 1.9623555506878496e-06, + "loss": 0.2682, + "step": 203 + }, + { + "epoch": 0.12, + "grad_norm": 2.0588104641051057, + "learning_rate": 1.961855177757506e-06, + "loss": 0.2443, + "step": 204 + }, + { + "epoch": 0.12, + "grad_norm": 1.5447026778774853, + "learning_rate": 1.9613515658954624e-06, + "loss": 0.2676, + "step": 205 + }, + { + "epoch": 0.12, + "grad_norm": 58.17799785515925, + "learning_rate": 1.9608447167975712e-06, + "loss": 0.2719, + "step": 206 + }, + { + "epoch": 0.12, + "grad_norm": 1.4201318187648717, + "learning_rate": 1.960334632170587e-06, + "loss": 0.2711, + "step": 207 + }, + { + "epoch": 0.12, + "grad_norm": 1.6151061560904458, + "learning_rate": 1.9598213137321566e-06, + "loss": 0.2771, + "step": 208 + }, + { + "epoch": 0.12, + "grad_norm": 1.3624415422897582, + "learning_rate": 1.9593047632108198e-06, + "loss": 0.2795, + "step": 209 + }, + { + "epoch": 0.12, + "grad_norm": 1.7063809700167063, + "learning_rate": 1.9587849823459974e-06, + "loss": 0.262, + "step": 210 + }, + { + "epoch": 0.12, + "grad_norm": 1.704767470623922, + "learning_rate": 1.9582619728879896e-06, + "loss": 0.2671, + "step": 211 + }, + { + "epoch": 0.12, + "grad_norm": 1.9236883518597807, + "learning_rate": 1.957735736597967e-06, + "loss": 0.2587, + "step": 212 + }, + { + "epoch": 0.12, + "grad_norm": 1.4580317276355494, + "learning_rate": 1.957206275247968e-06, + "loss": 0.2626, + "step": 213 + }, + { + "epoch": 0.12, + "grad_norm": 1.609114235032971, + "learning_rate": 1.95667359062089e-06, + "loss": 0.2663, + "step": 214 + }, + { + "epoch": 0.12, + "grad_norm": 1.9122216272707904, + "learning_rate": 1.9561376845104836e-06, + "loss": 0.2569, + "step": 215 + }, + { + "epoch": 0.12, + "grad_norm": 1.8505334920885719, + "learning_rate": 1.9555985587213494e-06, + "loss": 0.2676, + "step": 216 + }, + { + "epoch": 0.12, + "grad_norm": 1.5798315592924423, + "learning_rate": 1.9550562150689275e-06, + "loss": 0.2811, + "step": 217 + }, + { + "epoch": 0.12, + "grad_norm": 1.624535997367186, + "learning_rate": 1.954510655379496e-06, + "loss": 0.2936, + "step": 218 + }, + { + "epoch": 0.12, + "grad_norm": 1.6653151163752762, + "learning_rate": 1.953961881490161e-06, + "loss": 0.2717, + "step": 219 + }, + { + "epoch": 0.12, + "grad_norm": 1.9766689422985182, + "learning_rate": 1.9534098952488526e-06, + "loss": 0.276, + "step": 220 + }, + { + "epoch": 0.13, + "grad_norm": 2.6101980482803224, + "learning_rate": 1.9528546985143177e-06, + "loss": 0.2644, + "step": 221 + }, + { + "epoch": 0.13, + "grad_norm": 5.352286483093476, + "learning_rate": 1.9522962931561157e-06, + "loss": 0.2642, + "step": 222 + }, + { + "epoch": 0.13, + "grad_norm": 8.357014330075064, + "learning_rate": 1.9517346810546093e-06, + "loss": 0.2621, + "step": 223 + }, + { + "epoch": 0.13, + "grad_norm": 1.8920999167934909, + "learning_rate": 1.9511698641009587e-06, + "loss": 0.2679, + "step": 224 + }, + { + "epoch": 0.13, + "grad_norm": 1.6327875286304598, + "learning_rate": 1.950601844197118e-06, + "loss": 0.2625, + "step": 225 + }, + { + "epoch": 0.13, + "grad_norm": 1.5348617277564598, + "learning_rate": 1.9500306232558257e-06, + "loss": 0.2721, + "step": 226 + }, + { + "epoch": 0.13, + "grad_norm": 1.3717764165183968, + "learning_rate": 1.9494562032006e-06, + "loss": 0.2683, + "step": 227 + }, + { + "epoch": 0.13, + "grad_norm": 2.2776086923388212, + "learning_rate": 1.948878585965731e-06, + "loss": 0.2713, + "step": 228 + }, + { + "epoch": 0.13, + "grad_norm": 1.6429171205503288, + "learning_rate": 1.9482977734962752e-06, + "loss": 0.2549, + "step": 229 + }, + { + "epoch": 0.13, + "grad_norm": 1.4645706920773651, + "learning_rate": 1.947713767748049e-06, + "loss": 0.2808, + "step": 230 + }, + { + "epoch": 0.13, + "grad_norm": 1.7805860800870479, + "learning_rate": 1.9471265706876218e-06, + "loss": 0.2845, + "step": 231 + }, + { + "epoch": 0.13, + "grad_norm": 2.2186326422427856, + "learning_rate": 1.9465361842923084e-06, + "loss": 0.2479, + "step": 232 + }, + { + "epoch": 0.13, + "grad_norm": 3.7364731511845473, + "learning_rate": 1.945942610550165e-06, + "loss": 0.2711, + "step": 233 + }, + { + "epoch": 0.13, + "grad_norm": 1.4438603297603503, + "learning_rate": 1.9453458514599794e-06, + "loss": 0.2766, + "step": 234 + }, + { + "epoch": 0.13, + "grad_norm": 1.8558861822807204, + "learning_rate": 1.9447459090312654e-06, + "loss": 0.2778, + "step": 235 + }, + { + "epoch": 0.13, + "grad_norm": 1.4958770312114351, + "learning_rate": 1.944142785284258e-06, + "loss": 0.2727, + "step": 236 + }, + { + "epoch": 0.13, + "grad_norm": 1.7408912789904278, + "learning_rate": 1.9435364822499036e-06, + "loss": 0.2896, + "step": 237 + }, + { + "epoch": 0.13, + "grad_norm": 2.2339047425800165, + "learning_rate": 1.942927001969855e-06, + "loss": 0.2743, + "step": 238 + }, + { + "epoch": 0.14, + "grad_norm": 1.8664554419736463, + "learning_rate": 1.942314346496463e-06, + "loss": 0.2951, + "step": 239 + }, + { + "epoch": 0.14, + "grad_norm": 1.754954964563153, + "learning_rate": 1.9416985178927723e-06, + "loss": 0.2607, + "step": 240 + }, + { + "epoch": 0.14, + "grad_norm": 1.4994427888165345, + "learning_rate": 1.941079518232511e-06, + "loss": 0.2546, + "step": 241 + }, + { + "epoch": 0.14, + "grad_norm": 5.861516380556998, + "learning_rate": 1.9404573496000865e-06, + "loss": 0.2603, + "step": 242 + }, + { + "epoch": 0.14, + "grad_norm": 1.5880495496350484, + "learning_rate": 1.939832014090576e-06, + "loss": 0.2641, + "step": 243 + }, + { + "epoch": 0.14, + "grad_norm": 1.5751997589373612, + "learning_rate": 1.939203513809723e-06, + "loss": 0.2701, + "step": 244 + }, + { + "epoch": 0.14, + "grad_norm": 1.495124724753987, + "learning_rate": 1.938571850873926e-06, + "loss": 0.251, + "step": 245 + }, + { + "epoch": 0.14, + "grad_norm": 1.427805485761381, + "learning_rate": 1.9379370274102336e-06, + "loss": 0.2657, + "step": 246 + }, + { + "epoch": 0.14, + "grad_norm": 1.4922932315972126, + "learning_rate": 1.9372990455563383e-06, + "loss": 0.2634, + "step": 247 + }, + { + "epoch": 0.14, + "grad_norm": 1.6420455833777385, + "learning_rate": 1.936657907460567e-06, + "loss": 0.2792, + "step": 248 + }, + { + "epoch": 0.14, + "grad_norm": 2.162890816351466, + "learning_rate": 1.9360136152818745e-06, + "loss": 0.2862, + "step": 249 + }, + { + "epoch": 0.14, + "grad_norm": 1.4162967948218883, + "learning_rate": 1.9353661711898382e-06, + "loss": 0.2475, + "step": 250 + }, + { + "epoch": 0.14, + "grad_norm": 1.7480275180658238, + "learning_rate": 1.9347155773646476e-06, + "loss": 0.2662, + "step": 251 + }, + { + "epoch": 0.14, + "grad_norm": 1.5194422222133057, + "learning_rate": 1.9340618359971e-06, + "loss": 0.2734, + "step": 252 + }, + { + "epoch": 0.14, + "grad_norm": 2.180744355718762, + "learning_rate": 1.93340494928859e-06, + "loss": 0.2641, + "step": 253 + }, + { + "epoch": 0.14, + "grad_norm": 1.817998622984753, + "learning_rate": 1.9327449194511057e-06, + "loss": 0.282, + "step": 254 + }, + { + "epoch": 0.14, + "grad_norm": 1.4470115577869191, + "learning_rate": 1.932081748707217e-06, + "loss": 0.2574, + "step": 255 + }, + { + "epoch": 0.15, + "grad_norm": 1.5333101714313249, + "learning_rate": 1.9314154392900724e-06, + "loss": 0.2587, + "step": 256 + }, + { + "epoch": 0.15, + "grad_norm": 1.6209363674828383, + "learning_rate": 1.9307459934433886e-06, + "loss": 0.2681, + "step": 257 + }, + { + "epoch": 0.15, + "grad_norm": 2.064766532745283, + "learning_rate": 1.930073413421444e-06, + "loss": 0.2691, + "step": 258 + }, + { + "epoch": 0.15, + "grad_norm": 1.425928225401545, + "learning_rate": 1.9293977014890715e-06, + "loss": 0.2661, + "step": 259 + }, + { + "epoch": 0.15, + "grad_norm": 1.8432398109184334, + "learning_rate": 1.9287188599216487e-06, + "loss": 0.2756, + "step": 260 + }, + { + "epoch": 0.15, + "grad_norm": 1.9419330252305926, + "learning_rate": 1.928036891005094e-06, + "loss": 0.26, + "step": 261 + }, + { + "epoch": 0.15, + "grad_norm": 1.9246952207139454, + "learning_rate": 1.9273517970358557e-06, + "loss": 0.2602, + "step": 262 + }, + { + "epoch": 0.15, + "grad_norm": 1.6182356210181768, + "learning_rate": 1.9266635803209045e-06, + "loss": 0.2683, + "step": 263 + }, + { + "epoch": 0.15, + "grad_norm": 1.6007631316186846, + "learning_rate": 1.9259722431777283e-06, + "loss": 0.2751, + "step": 264 + }, + { + "epoch": 0.15, + "grad_norm": 1.571410370519595, + "learning_rate": 1.9252777879343214e-06, + "loss": 0.2733, + "step": 265 + }, + { + "epoch": 0.15, + "grad_norm": 1.3967589720106062, + "learning_rate": 1.9245802169291786e-06, + "loss": 0.2658, + "step": 266 + }, + { + "epoch": 0.15, + "grad_norm": 2.0885804233758076, + "learning_rate": 1.9238795325112867e-06, + "loss": 0.2618, + "step": 267 + }, + { + "epoch": 0.15, + "grad_norm": 3.1804124273835255, + "learning_rate": 1.9231757370401155e-06, + "loss": 0.2482, + "step": 268 + }, + { + "epoch": 0.15, + "grad_norm": 1.4883966724737543, + "learning_rate": 1.9224688328856126e-06, + "loss": 0.2571, + "step": 269 + }, + { + "epoch": 0.15, + "grad_norm": 1.6229389625184016, + "learning_rate": 1.9217588224281916e-06, + "loss": 0.2805, + "step": 270 + }, + { + "epoch": 0.15, + "grad_norm": 1.3740620060528976, + "learning_rate": 1.9210457080587283e-06, + "loss": 0.2468, + "step": 271 + }, + { + "epoch": 0.15, + "grad_norm": 1.754068936990107, + "learning_rate": 1.9203294921785494e-06, + "loss": 0.2672, + "step": 272 + }, + { + "epoch": 0.15, + "grad_norm": 2.7039143913135644, + "learning_rate": 1.919610177199426e-06, + "loss": 0.2687, + "step": 273 + }, + { + "epoch": 0.16, + "grad_norm": 1.7295487903086968, + "learning_rate": 1.9188877655435643e-06, + "loss": 0.2567, + "step": 274 + }, + { + "epoch": 0.16, + "grad_norm": 2.7155931053445603, + "learning_rate": 1.918162259643599e-06, + "loss": 0.2461, + "step": 275 + }, + { + "epoch": 0.16, + "grad_norm": 1.7831424868639558, + "learning_rate": 1.917433661942584e-06, + "loss": 0.2669, + "step": 276 + }, + { + "epoch": 0.16, + "grad_norm": 2.049245904280425, + "learning_rate": 1.9167019748939844e-06, + "loss": 0.2497, + "step": 277 + }, + { + "epoch": 0.16, + "grad_norm": 1.5950707091096057, + "learning_rate": 1.9159672009616687e-06, + "loss": 0.2824, + "step": 278 + }, + { + "epoch": 0.16, + "grad_norm": 1.6817617181860023, + "learning_rate": 1.915229342619899e-06, + "loss": 0.2516, + "step": 279 + }, + { + "epoch": 0.16, + "grad_norm": 1.524769739672067, + "learning_rate": 1.914488402353326e-06, + "loss": 0.2646, + "step": 280 + }, + { + "epoch": 0.16, + "grad_norm": 1.5720744364995505, + "learning_rate": 1.9137443826569755e-06, + "loss": 0.2757, + "step": 281 + }, + { + "epoch": 0.16, + "grad_norm": 1.6625973270409442, + "learning_rate": 1.9129972860362456e-06, + "loss": 0.2463, + "step": 282 + }, + { + "epoch": 0.16, + "grad_norm": 1.5570685620443854, + "learning_rate": 1.9122471150068936e-06, + "loss": 0.2529, + "step": 283 + }, + { + "epoch": 0.16, + "grad_norm": 1.631556253373755, + "learning_rate": 1.9114938720950304e-06, + "loss": 0.2519, + "step": 284 + }, + { + "epoch": 0.16, + "grad_norm": 1.5618307115656873, + "learning_rate": 1.910737559837111e-06, + "loss": 0.2609, + "step": 285 + }, + { + "epoch": 0.16, + "grad_norm": 1.4955017462305111, + "learning_rate": 1.9099781807799263e-06, + "loss": 0.2857, + "step": 286 + }, + { + "epoch": 0.16, + "grad_norm": 1.3325718527553116, + "learning_rate": 1.9092157374805935e-06, + "loss": 0.257, + "step": 287 + }, + { + "epoch": 0.16, + "grad_norm": 1.4402867852956336, + "learning_rate": 1.9084502325065493e-06, + "loss": 0.2782, + "step": 288 + }, + { + "epoch": 0.16, + "grad_norm": 1.6336784855059647, + "learning_rate": 1.9076816684355396e-06, + "loss": 0.2581, + "step": 289 + }, + { + "epoch": 0.16, + "grad_norm": 1.648452115684247, + "learning_rate": 1.9069100478556109e-06, + "loss": 0.2641, + "step": 290 + }, + { + "epoch": 0.16, + "grad_norm": 4.29013379295967, + "learning_rate": 1.9061353733651035e-06, + "loss": 0.2608, + "step": 291 + }, + { + "epoch": 0.17, + "grad_norm": 4.427885638603707, + "learning_rate": 1.9053576475726402e-06, + "loss": 0.2576, + "step": 292 + }, + { + "epoch": 0.17, + "grad_norm": 1.4626100839525564, + "learning_rate": 1.9045768730971195e-06, + "loss": 0.2498, + "step": 293 + }, + { + "epoch": 0.17, + "grad_norm": 1.5571575693366928, + "learning_rate": 1.9037930525677055e-06, + "loss": 0.2532, + "step": 294 + }, + { + "epoch": 0.17, + "grad_norm": 2.041792290394832, + "learning_rate": 1.903006188623819e-06, + "loss": 0.2647, + "step": 295 + }, + { + "epoch": 0.17, + "grad_norm": 1.4571147687585155, + "learning_rate": 1.9022162839151309e-06, + "loss": 0.2732, + "step": 296 + }, + { + "epoch": 0.17, + "grad_norm": 1.57029661183803, + "learning_rate": 1.9014233411015495e-06, + "loss": 0.2678, + "step": 297 + }, + { + "epoch": 0.17, + "grad_norm": 2.095378117120327, + "learning_rate": 1.9006273628532144e-06, + "loss": 0.2571, + "step": 298 + }, + { + "epoch": 0.17, + "grad_norm": 1.9158555688964518, + "learning_rate": 1.8998283518504872e-06, + "loss": 0.25, + "step": 299 + }, + { + "epoch": 0.17, + "grad_norm": 1.4977815013315328, + "learning_rate": 1.8990263107839407e-06, + "loss": 0.2664, + "step": 300 + }, + { + "epoch": 0.17, + "grad_norm": 1.4524881944630132, + "learning_rate": 1.8982212423543527e-06, + "loss": 0.2618, + "step": 301 + }, + { + "epoch": 0.17, + "grad_norm": 1.453409301557865, + "learning_rate": 1.8974131492726938e-06, + "loss": 0.2512, + "step": 302 + }, + { + "epoch": 0.17, + "grad_norm": 3.484838904690488, + "learning_rate": 1.8966020342601202e-06, + "loss": 0.2599, + "step": 303 + }, + { + "epoch": 0.17, + "grad_norm": 1.548516570434266, + "learning_rate": 1.8957879000479646e-06, + "loss": 0.2759, + "step": 304 + }, + { + "epoch": 0.17, + "grad_norm": 8.857639079344162, + "learning_rate": 1.8949707493777258e-06, + "loss": 0.2714, + "step": 305 + }, + { + "epoch": 0.17, + "grad_norm": 2.649599781678979, + "learning_rate": 1.8941505850010611e-06, + "loss": 0.2768, + "step": 306 + }, + { + "epoch": 0.17, + "grad_norm": 2.907972020908407, + "learning_rate": 1.8933274096797742e-06, + "loss": 0.25, + "step": 307 + }, + { + "epoch": 0.17, + "grad_norm": 1.9997184476834842, + "learning_rate": 1.8925012261858101e-06, + "loss": 0.2596, + "step": 308 + }, + { + "epoch": 0.18, + "grad_norm": 1.688271694451874, + "learning_rate": 1.8916720373012423e-06, + "loss": 0.2637, + "step": 309 + }, + { + "epoch": 0.18, + "grad_norm": 1.5803539700538585, + "learning_rate": 1.8908398458182643e-06, + "loss": 0.2608, + "step": 310 + }, + { + "epoch": 0.18, + "grad_norm": 4.258078848126725, + "learning_rate": 1.8900046545391814e-06, + "loss": 0.2721, + "step": 311 + }, + { + "epoch": 0.18, + "grad_norm": 1.3656658504996113, + "learning_rate": 1.8891664662763995e-06, + "loss": 0.2516, + "step": 312 + }, + { + "epoch": 0.18, + "grad_norm": 1.3871598435416357, + "learning_rate": 1.8883252838524166e-06, + "loss": 0.255, + "step": 313 + }, + { + "epoch": 0.18, + "grad_norm": 1.6935400165801882, + "learning_rate": 1.8874811100998137e-06, + "loss": 0.2477, + "step": 314 + }, + { + "epoch": 0.18, + "grad_norm": 1.6335653329235087, + "learning_rate": 1.8866339478612448e-06, + "loss": 0.2425, + "step": 315 + }, + { + "epoch": 0.18, + "grad_norm": 1.858663218599638, + "learning_rate": 1.8857837999894264e-06, + "loss": 0.2773, + "step": 316 + }, + { + "epoch": 0.18, + "grad_norm": 1.3412242747183023, + "learning_rate": 1.8849306693471295e-06, + "loss": 0.2626, + "step": 317 + }, + { + "epoch": 0.18, + "grad_norm": 1.416778303285613, + "learning_rate": 1.8840745588071687e-06, + "loss": 0.2675, + "step": 318 + }, + { + "epoch": 0.18, + "grad_norm": 1.5314582573192885, + "learning_rate": 1.8832154712523935e-06, + "loss": 0.2505, + "step": 319 + }, + { + "epoch": 0.18, + "grad_norm": 2.3318619365410687, + "learning_rate": 1.8823534095756784e-06, + "loss": 0.2783, + "step": 320 + }, + { + "epoch": 0.18, + "grad_norm": 1.4583759244957066, + "learning_rate": 1.8814883766799117e-06, + "loss": 0.2578, + "step": 321 + }, + { + "epoch": 0.18, + "grad_norm": 1.497237501425145, + "learning_rate": 1.8806203754779885e-06, + "loss": 0.2418, + "step": 322 + }, + { + "epoch": 0.18, + "grad_norm": 1.5780661961703015, + "learning_rate": 1.8797494088927982e-06, + "loss": 0.2775, + "step": 323 + }, + { + "epoch": 0.18, + "grad_norm": 1.826755382387934, + "learning_rate": 1.8788754798572156e-06, + "loss": 0.2501, + "step": 324 + }, + { + "epoch": 0.18, + "grad_norm": 9.04923490348713, + "learning_rate": 1.8779985913140922e-06, + "loss": 0.2566, + "step": 325 + }, + { + "epoch": 0.18, + "grad_norm": 1.5771692851656443, + "learning_rate": 1.877118746216245e-06, + "loss": 0.2487, + "step": 326 + }, + { + "epoch": 0.19, + "grad_norm": 1.4053174744629435, + "learning_rate": 1.876235947526446e-06, + "loss": 0.2506, + "step": 327 + }, + { + "epoch": 0.19, + "grad_norm": 2.1378447373700036, + "learning_rate": 1.875350198217414e-06, + "loss": 0.261, + "step": 328 + }, + { + "epoch": 0.19, + "grad_norm": 1.5218176607440317, + "learning_rate": 1.8744615012718038e-06, + "loss": 0.2585, + "step": 329 + }, + { + "epoch": 0.19, + "grad_norm": 1.5103368426908612, + "learning_rate": 1.8735698596821946e-06, + "loss": 0.2754, + "step": 330 + }, + { + "epoch": 0.19, + "grad_norm": 1.5057746610256875, + "learning_rate": 1.8726752764510825e-06, + "loss": 0.2553, + "step": 331 + }, + { + "epoch": 0.19, + "grad_norm": 2.2904119881064395, + "learning_rate": 1.8717777545908694e-06, + "loss": 0.2696, + "step": 332 + }, + { + "epoch": 0.19, + "grad_norm": 1.5255556083995068, + "learning_rate": 1.8708772971238526e-06, + "loss": 0.2498, + "step": 333 + }, + { + "epoch": 0.19, + "grad_norm": 2.1437207390920414, + "learning_rate": 1.8699739070822136e-06, + "loss": 0.2508, + "step": 334 + }, + { + "epoch": 0.19, + "grad_norm": 3.2156588595090185, + "learning_rate": 1.86906758750801e-06, + "loss": 0.2584, + "step": 335 + }, + { + "epoch": 0.19, + "grad_norm": 1.453213205964556, + "learning_rate": 1.868158341453164e-06, + "loss": 0.265, + "step": 336 + }, + { + "epoch": 0.19, + "grad_norm": 1.5870385588706113, + "learning_rate": 1.867246171979453e-06, + "loss": 0.2771, + "step": 337 + }, + { + "epoch": 0.19, + "grad_norm": 1.6632905370336226, + "learning_rate": 1.8663310821584974e-06, + "loss": 0.2736, + "step": 338 + }, + { + "epoch": 0.19, + "grad_norm": 1.9573751783307998, + "learning_rate": 1.8654130750717522e-06, + "loss": 0.2542, + "step": 339 + }, + { + "epoch": 0.19, + "grad_norm": 2.014677579627816, + "learning_rate": 1.8644921538104963e-06, + "loss": 0.266, + "step": 340 + }, + { + "epoch": 0.19, + "grad_norm": 1.599367462960297, + "learning_rate": 1.8635683214758212e-06, + "loss": 0.2618, + "step": 341 + }, + { + "epoch": 0.19, + "grad_norm": 1.5394500305143561, + "learning_rate": 1.8626415811786212e-06, + "loss": 0.2739, + "step": 342 + }, + { + "epoch": 0.19, + "grad_norm": 1.7655995554131592, + "learning_rate": 1.8617119360395825e-06, + "loss": 0.2623, + "step": 343 + }, + { + "epoch": 0.19, + "grad_norm": 1.3443351289371104, + "learning_rate": 1.8607793891891741e-06, + "loss": 0.2453, + "step": 344 + }, + { + "epoch": 0.2, + "grad_norm": 2.706613653140644, + "learning_rate": 1.8598439437676349e-06, + "loss": 0.2544, + "step": 345 + }, + { + "epoch": 0.2, + "grad_norm": 1.4369277394486442, + "learning_rate": 1.858905602924965e-06, + "loss": 0.269, + "step": 346 + }, + { + "epoch": 0.2, + "grad_norm": 2.6458693329035445, + "learning_rate": 1.8579643698209145e-06, + "loss": 0.2442, + "step": 347 + }, + { + "epoch": 0.2, + "grad_norm": 1.6965174127684481, + "learning_rate": 1.8570202476249727e-06, + "loss": 0.267, + "step": 348 + }, + { + "epoch": 0.2, + "grad_norm": 1.9479394734771236, + "learning_rate": 1.856073239516358e-06, + "loss": 0.2502, + "step": 349 + }, + { + "epoch": 0.2, + "grad_norm": 1.8717497710304911, + "learning_rate": 1.8551233486840058e-06, + "loss": 0.2615, + "step": 350 + }, + { + "epoch": 0.2, + "grad_norm": 1.3961933051605424, + "learning_rate": 1.8541705783265595e-06, + "loss": 0.2479, + "step": 351 + }, + { + "epoch": 0.2, + "grad_norm": 1.548456709806826, + "learning_rate": 1.8532149316523587e-06, + "loss": 0.2458, + "step": 352 + }, + { + "epoch": 0.2, + "grad_norm": 1.4983113746510905, + "learning_rate": 1.8522564118794288e-06, + "loss": 0.2643, + "step": 353 + }, + { + "epoch": 0.2, + "grad_norm": 1.8441091849122442, + "learning_rate": 1.8512950222354702e-06, + "loss": 0.2429, + "step": 354 + }, + { + "epoch": 0.2, + "grad_norm": 3.3405305453121286, + "learning_rate": 1.8503307659578465e-06, + "loss": 0.2649, + "step": 355 + }, + { + "epoch": 0.2, + "grad_norm": 1.521549967718533, + "learning_rate": 1.849363646293575e-06, + "loss": 0.2719, + "step": 356 + }, + { + "epoch": 0.2, + "grad_norm": 1.9726008542682227, + "learning_rate": 1.848393666499315e-06, + "loss": 0.2529, + "step": 357 + }, + { + "epoch": 0.2, + "grad_norm": 3.1216441460870854, + "learning_rate": 1.8474208298413565e-06, + "loss": 0.2603, + "step": 358 + }, + { + "epoch": 0.2, + "grad_norm": 1.8168534441069981, + "learning_rate": 1.8464451395956107e-06, + "loss": 0.2732, + "step": 359 + }, + { + "epoch": 0.2, + "grad_norm": 1.336787937365217, + "learning_rate": 1.8454665990475963e-06, + "loss": 0.2409, + "step": 360 + }, + { + "epoch": 0.2, + "grad_norm": 2.016673189221436, + "learning_rate": 1.844485211492432e-06, + "loss": 0.2642, + "step": 361 + }, + { + "epoch": 0.21, + "grad_norm": 1.3679935456862138, + "learning_rate": 1.8435009802348219e-06, + "loss": 0.2528, + "step": 362 + }, + { + "epoch": 0.21, + "grad_norm": 1.6232171143940508, + "learning_rate": 1.8425139085890464e-06, + "loss": 0.2557, + "step": 363 + }, + { + "epoch": 0.21, + "grad_norm": 1.5680694316061323, + "learning_rate": 1.8415239998789508e-06, + "loss": 0.2641, + "step": 364 + }, + { + "epoch": 0.21, + "grad_norm": 1.5005299041051148, + "learning_rate": 1.840531257437934e-06, + "loss": 0.2542, + "step": 365 + }, + { + "epoch": 0.21, + "grad_norm": 2.2550744008937897, + "learning_rate": 1.8395356846089358e-06, + "loss": 0.2399, + "step": 366 + }, + { + "epoch": 0.21, + "grad_norm": 1.3923027217183879, + "learning_rate": 1.838537284744429e-06, + "loss": 0.2457, + "step": 367 + }, + { + "epoch": 0.21, + "grad_norm": 1.3855500063018884, + "learning_rate": 1.8375360612064048e-06, + "loss": 0.2285, + "step": 368 + }, + { + "epoch": 0.21, + "grad_norm": 1.3712056809511044, + "learning_rate": 1.8365320173663635e-06, + "loss": 0.2521, + "step": 369 + }, + { + "epoch": 0.21, + "grad_norm": 1.6662319800030796, + "learning_rate": 1.8355251566053014e-06, + "loss": 0.2663, + "step": 370 + }, + { + "epoch": 0.21, + "grad_norm": 1.7823378714354956, + "learning_rate": 1.8345154823137014e-06, + "loss": 0.25, + "step": 371 + }, + { + "epoch": 0.21, + "grad_norm": 1.316422764025108, + "learning_rate": 1.8335029978915207e-06, + "loss": 0.2439, + "step": 372 + }, + { + "epoch": 0.21, + "grad_norm": 2.182737096619672, + "learning_rate": 1.8324877067481782e-06, + "loss": 0.2613, + "step": 373 + }, + { + "epoch": 0.21, + "grad_norm": 1.5481659409963358, + "learning_rate": 1.8314696123025452e-06, + "loss": 0.2514, + "step": 374 + }, + { + "epoch": 0.21, + "grad_norm": 1.5793008682263645, + "learning_rate": 1.8304487179829323e-06, + "loss": 0.2578, + "step": 375 + }, + { + "epoch": 0.21, + "grad_norm": 1.3846024289902152, + "learning_rate": 1.8294250272270784e-06, + "loss": 0.2548, + "step": 376 + }, + { + "epoch": 0.21, + "grad_norm": 8.639190275028675, + "learning_rate": 1.8283985434821392e-06, + "loss": 0.2487, + "step": 377 + }, + { + "epoch": 0.21, + "grad_norm": 1.4960588909233208, + "learning_rate": 1.8273692702046751e-06, + "loss": 0.2605, + "step": 378 + }, + { + "epoch": 0.21, + "grad_norm": 1.8471412838734627, + "learning_rate": 1.82633721086064e-06, + "loss": 0.2558, + "step": 379 + }, + { + "epoch": 0.22, + "grad_norm": 2.7430897049166423, + "learning_rate": 1.8253023689253704e-06, + "loss": 0.2631, + "step": 380 + }, + { + "epoch": 0.22, + "grad_norm": 1.7067111615790653, + "learning_rate": 1.8242647478835716e-06, + "loss": 0.26, + "step": 381 + }, + { + "epoch": 0.22, + "grad_norm": 2.249256519894356, + "learning_rate": 1.8232243512293078e-06, + "loss": 0.2699, + "step": 382 + }, + { + "epoch": 0.22, + "grad_norm": 1.8911206849745297, + "learning_rate": 1.8221811824659897e-06, + "loss": 0.2434, + "step": 383 + }, + { + "epoch": 0.22, + "grad_norm": 2.2417249888846618, + "learning_rate": 1.8211352451063626e-06, + "loss": 0.2517, + "step": 384 + }, + { + "epoch": 0.22, + "grad_norm": 1.3609186236445172, + "learning_rate": 1.820086542672495e-06, + "loss": 0.2425, + "step": 385 + }, + { + "epoch": 0.22, + "grad_norm": 2.4401605272522, + "learning_rate": 1.8190350786957667e-06, + "loss": 0.2495, + "step": 386 + }, + { + "epoch": 0.22, + "grad_norm": 3.9867213064504, + "learning_rate": 1.8179808567168552e-06, + "loss": 0.2621, + "step": 387 + }, + { + "epoch": 0.22, + "grad_norm": 2.5269161987442152, + "learning_rate": 1.8169238802857274e-06, + "loss": 0.2326, + "step": 388 + }, + { + "epoch": 0.22, + "grad_norm": 1.515973959752345, + "learning_rate": 1.8158641529616236e-06, + "loss": 0.2659, + "step": 389 + }, + { + "epoch": 0.22, + "grad_norm": 1.7035163798577122, + "learning_rate": 1.814801678313049e-06, + "loss": 0.2525, + "step": 390 + }, + { + "epoch": 0.22, + "grad_norm": 1.7146882756588415, + "learning_rate": 1.8137364599177584e-06, + "loss": 0.2639, + "step": 391 + }, + { + "epoch": 0.22, + "grad_norm": 1.9168178698259475, + "learning_rate": 1.8126685013627472e-06, + "loss": 0.2787, + "step": 392 + }, + { + "epoch": 0.22, + "grad_norm": 1.4690690272480431, + "learning_rate": 1.8115978062442373e-06, + "loss": 0.2348, + "step": 393 + }, + { + "epoch": 0.22, + "grad_norm": 2.8100184186004884, + "learning_rate": 1.810524378167666e-06, + "loss": 0.2367, + "step": 394 + }, + { + "epoch": 0.22, + "grad_norm": 6.485932466140042, + "learning_rate": 1.8094482207476728e-06, + "loss": 0.2587, + "step": 395 + }, + { + "epoch": 0.22, + "grad_norm": 7.073727801578548, + "learning_rate": 1.8083693376080894e-06, + "loss": 0.2559, + "step": 396 + }, + { + "epoch": 0.22, + "grad_norm": 1.5409132585836345, + "learning_rate": 1.8072877323819244e-06, + "loss": 0.2471, + "step": 397 + }, + { + "epoch": 0.23, + "grad_norm": 1.4965005384663332, + "learning_rate": 1.8062034087113536e-06, + "loss": 0.2601, + "step": 398 + }, + { + "epoch": 0.23, + "grad_norm": 1.7113041600197383, + "learning_rate": 1.8051163702477063e-06, + "loss": 0.2732, + "step": 399 + }, + { + "epoch": 0.23, + "grad_norm": 1.406045617724908, + "learning_rate": 1.8040266206514545e-06, + "loss": 0.2394, + "step": 400 + }, + { + "epoch": 0.23, + "grad_norm": 1.508758015298593, + "learning_rate": 1.8029341635921985e-06, + "loss": 0.2451, + "step": 401 + }, + { + "epoch": 0.23, + "grad_norm": 1.632244989969479, + "learning_rate": 1.801839002748656e-06, + "loss": 0.2441, + "step": 402 + }, + { + "epoch": 0.23, + "grad_norm": 1.6557206877151516, + "learning_rate": 1.8007411418086503e-06, + "loss": 0.2612, + "step": 403 + }, + { + "epoch": 0.23, + "grad_norm": 1.5731162109957053, + "learning_rate": 1.7996405844690955e-06, + "loss": 0.232, + "step": 404 + }, + { + "epoch": 0.23, + "grad_norm": 1.5592785474295656, + "learning_rate": 1.798537334435986e-06, + "loss": 0.262, + "step": 405 + }, + { + "epoch": 0.23, + "grad_norm": 1.7766036403394192, + "learning_rate": 1.7974313954243843e-06, + "loss": 0.2386, + "step": 406 + }, + { + "epoch": 0.23, + "grad_norm": 1.5053094164582228, + "learning_rate": 1.7963227711584066e-06, + "loss": 0.2551, + "step": 407 + }, + { + "epoch": 0.23, + "grad_norm": 2.0419848727271144, + "learning_rate": 1.795211465371212e-06, + "loss": 0.2478, + "step": 408 + }, + { + "epoch": 0.23, + "grad_norm": 1.656239277547231, + "learning_rate": 1.7940974818049894e-06, + "loss": 0.2588, + "step": 409 + }, + { + "epoch": 0.23, + "grad_norm": 2.535248015867066, + "learning_rate": 1.7929808242109444e-06, + "loss": 0.2513, + "step": 410 + }, + { + "epoch": 0.23, + "grad_norm": 1.7593350582158802, + "learning_rate": 1.7918614963492868e-06, + "loss": 0.2641, + "step": 411 + }, + { + "epoch": 0.23, + "grad_norm": 1.6239493408782932, + "learning_rate": 1.7907395019892194e-06, + "loss": 0.26, + "step": 412 + }, + { + "epoch": 0.23, + "grad_norm": 1.8328916673718427, + "learning_rate": 1.7896148449089225e-06, + "loss": 0.2424, + "step": 413 + }, + { + "epoch": 0.23, + "grad_norm": 1.6488096268947292, + "learning_rate": 1.7884875288955446e-06, + "loss": 0.2426, + "step": 414 + }, + { + "epoch": 0.24, + "grad_norm": 1.4999408005310428, + "learning_rate": 1.787357557745186e-06, + "loss": 0.2512, + "step": 415 + }, + { + "epoch": 0.24, + "grad_norm": 1.719539454877777, + "learning_rate": 1.7862249352628895e-06, + "loss": 0.2691, + "step": 416 + }, + { + "epoch": 0.24, + "grad_norm": 2.476140799877467, + "learning_rate": 1.7850896652626246e-06, + "loss": 0.2688, + "step": 417 + }, + { + "epoch": 0.24, + "grad_norm": 1.379279565870243, + "learning_rate": 1.7839517515672768e-06, + "loss": 0.244, + "step": 418 + }, + { + "epoch": 0.24, + "grad_norm": 1.4101711514396826, + "learning_rate": 1.7828111980086335e-06, + "loss": 0.2417, + "step": 419 + }, + { + "epoch": 0.24, + "grad_norm": 1.769134440896085, + "learning_rate": 1.7816680084273724e-06, + "loss": 0.275, + "step": 420 + }, + { + "epoch": 0.24, + "grad_norm": 1.7447322283537994, + "learning_rate": 1.7805221866730458e-06, + "loss": 0.2462, + "step": 421 + }, + { + "epoch": 0.24, + "grad_norm": 1.3226792033914545, + "learning_rate": 1.7793737366040718e-06, + "loss": 0.2529, + "step": 422 + }, + { + "epoch": 0.24, + "grad_norm": 1.6647996204010023, + "learning_rate": 1.7782226620877175e-06, + "loss": 0.2536, + "step": 423 + }, + { + "epoch": 0.24, + "grad_norm": 1.4353109720906432, + "learning_rate": 1.777068967000088e-06, + "loss": 0.26, + "step": 424 + }, + { + "epoch": 0.24, + "grad_norm": 1.4435065706691035, + "learning_rate": 1.775912655226113e-06, + "loss": 0.2669, + "step": 425 + }, + { + "epoch": 0.24, + "grad_norm": 1.770720909184648, + "learning_rate": 1.7747537306595333e-06, + "loss": 0.2675, + "step": 426 + }, + { + "epoch": 0.24, + "grad_norm": 1.2747069276814509, + "learning_rate": 1.773592197202888e-06, + "loss": 0.2324, + "step": 427 + }, + { + "epoch": 0.24, + "grad_norm": 1.6447179982326599, + "learning_rate": 1.7724280587675015e-06, + "loss": 0.2469, + "step": 428 + }, + { + "epoch": 0.24, + "grad_norm": 1.6429984508102649, + "learning_rate": 1.7712613192734702e-06, + "loss": 0.2432, + "step": 429 + }, + { + "epoch": 0.24, + "grad_norm": 1.9367755129503237, + "learning_rate": 1.770091982649649e-06, + "loss": 0.2441, + "step": 430 + }, + { + "epoch": 0.24, + "grad_norm": 1.426679348855773, + "learning_rate": 1.7689200528336384e-06, + "loss": 0.2481, + "step": 431 + }, + { + "epoch": 0.24, + "grad_norm": 1.9894052680715406, + "learning_rate": 1.7677455337717708e-06, + "loss": 0.2361, + "step": 432 + }, + { + "epoch": 0.25, + "grad_norm": 1.5632096934948594, + "learning_rate": 1.7665684294190989e-06, + "loss": 0.2525, + "step": 433 + }, + { + "epoch": 0.25, + "grad_norm": 1.4333963564202992, + "learning_rate": 1.765388743739379e-06, + "loss": 0.2566, + "step": 434 + }, + { + "epoch": 0.25, + "grad_norm": 1.4007544482708665, + "learning_rate": 1.7642064807050615e-06, + "loss": 0.2411, + "step": 435 + }, + { + "epoch": 0.25, + "grad_norm": 1.5666245383536983, + "learning_rate": 1.7630216442972756e-06, + "loss": 0.2541, + "step": 436 + }, + { + "epoch": 0.25, + "grad_norm": 1.5252679112447733, + "learning_rate": 1.7618342385058145e-06, + "loss": 0.2488, + "step": 437 + }, + { + "epoch": 0.25, + "grad_norm": 1.7782030419957275, + "learning_rate": 1.7606442673291247e-06, + "loss": 0.2401, + "step": 438 + }, + { + "epoch": 0.25, + "grad_norm": 1.7810576859631657, + "learning_rate": 1.7594517347742913e-06, + "loss": 0.2663, + "step": 439 + }, + { + "epoch": 0.25, + "grad_norm": 2.349540538044517, + "learning_rate": 1.7582566448570244e-06, + "loss": 0.2623, + "step": 440 + }, + { + "epoch": 0.25, + "grad_norm": 1.8374885307017144, + "learning_rate": 1.7570590016016453e-06, + "loss": 0.2719, + "step": 441 + }, + { + "epoch": 0.25, + "grad_norm": 1.6591670657101392, + "learning_rate": 1.7558588090410745e-06, + "loss": 0.2708, + "step": 442 + }, + { + "epoch": 0.25, + "grad_norm": 1.3874442701643437, + "learning_rate": 1.7546560712168153e-06, + "loss": 0.2467, + "step": 443 + }, + { + "epoch": 0.25, + "grad_norm": 1.7881115535751977, + "learning_rate": 1.753450792178943e-06, + "loss": 0.2429, + "step": 444 + }, + { + "epoch": 0.25, + "grad_norm": 1.4700478447897465, + "learning_rate": 1.75224297598609e-06, + "loss": 0.2596, + "step": 445 + }, + { + "epoch": 0.25, + "grad_norm": 1.3756308145286567, + "learning_rate": 1.7510326267054324e-06, + "loss": 0.2443, + "step": 446 + }, + { + "epoch": 0.25, + "grad_norm": 1.4546591942630904, + "learning_rate": 1.7498197484126753e-06, + "loss": 0.251, + "step": 447 + }, + { + "epoch": 0.25, + "grad_norm": 1.4189605597831128, + "learning_rate": 1.7486043451920413e-06, + "loss": 0.2477, + "step": 448 + }, + { + "epoch": 0.25, + "grad_norm": 1.46931912434084, + "learning_rate": 1.7473864211362543e-06, + "loss": 0.2521, + "step": 449 + }, + { + "epoch": 0.25, + "grad_norm": 1.6788496933700936, + "learning_rate": 1.746165980346527e-06, + "loss": 0.2452, + "step": 450 + }, + { + "epoch": 0.26, + "grad_norm": 1.741628012364131, + "learning_rate": 1.744943026932547e-06, + "loss": 0.2415, + "step": 451 + }, + { + "epoch": 0.26, + "grad_norm": 1.4759944312767415, + "learning_rate": 1.7437175650124637e-06, + "loss": 0.2442, + "step": 452 + }, + { + "epoch": 0.26, + "grad_norm": 2.460598637476957, + "learning_rate": 1.742489598712872e-06, + "loss": 0.2345, + "step": 453 + }, + { + "epoch": 0.26, + "grad_norm": 1.6154575341587825, + "learning_rate": 1.7412591321688011e-06, + "loss": 0.2448, + "step": 454 + }, + { + "epoch": 0.26, + "grad_norm": 1.6350983593850692, + "learning_rate": 1.7400261695236993e-06, + "loss": 0.2524, + "step": 455 + }, + { + "epoch": 0.26, + "grad_norm": 1.343754316077936, + "learning_rate": 1.7387907149294198e-06, + "loss": 0.2537, + "step": 456 + }, + { + "epoch": 0.26, + "grad_norm": 1.9085229267662134, + "learning_rate": 1.7375527725462078e-06, + "loss": 0.2464, + "step": 457 + }, + { + "epoch": 0.26, + "grad_norm": 1.3834196533472318, + "learning_rate": 1.7363123465426852e-06, + "loss": 0.2552, + "step": 458 + }, + { + "epoch": 0.26, + "grad_norm": 1.4773476116537363, + "learning_rate": 1.7350694410958377e-06, + "loss": 0.2326, + "step": 459 + }, + { + "epoch": 0.26, + "grad_norm": 1.6942778350369863, + "learning_rate": 1.733824060391e-06, + "loss": 0.2471, + "step": 460 + }, + { + "epoch": 0.26, + "grad_norm": 2.295713670933045, + "learning_rate": 1.7325762086218414e-06, + "loss": 0.2429, + "step": 461 + }, + { + "epoch": 0.26, + "grad_norm": 2.7097616944136704, + "learning_rate": 1.7313258899903535e-06, + "loss": 0.2534, + "step": 462 + }, + { + "epoch": 0.26, + "grad_norm": 2.2580089656465554, + "learning_rate": 1.7300731087068336e-06, + "loss": 0.2319, + "step": 463 + }, + { + "epoch": 0.26, + "grad_norm": 1.5983957319863051, + "learning_rate": 1.7288178689898722e-06, + "loss": 0.2744, + "step": 464 + }, + { + "epoch": 0.26, + "grad_norm": 2.2575480793345224, + "learning_rate": 1.7275601750663384e-06, + "loss": 0.2401, + "step": 465 + }, + { + "epoch": 0.26, + "grad_norm": 1.384370331036085, + "learning_rate": 1.7263000311713648e-06, + "loss": 0.2437, + "step": 466 + }, + { + "epoch": 0.26, + "grad_norm": 1.7448252510746651, + "learning_rate": 1.7250374415483348e-06, + "loss": 0.2639, + "step": 467 + }, + { + "epoch": 0.27, + "grad_norm": 1.7024784295169932, + "learning_rate": 1.7237724104488673e-06, + "loss": 0.2595, + "step": 468 + }, + { + "epoch": 0.27, + "grad_norm": 1.6511619159157513, + "learning_rate": 1.7225049421328022e-06, + "loss": 0.2703, + "step": 469 + }, + { + "epoch": 0.27, + "grad_norm": 1.4470658970436698, + "learning_rate": 1.7212350408681867e-06, + "loss": 0.2351, + "step": 470 + }, + { + "epoch": 0.27, + "grad_norm": 1.3888372489540777, + "learning_rate": 1.7199627109312611e-06, + "loss": 0.2522, + "step": 471 + }, + { + "epoch": 0.27, + "grad_norm": 1.3972869351074375, + "learning_rate": 1.718687956606443e-06, + "loss": 0.2516, + "step": 472 + }, + { + "epoch": 0.27, + "grad_norm": 1.8218195729892743, + "learning_rate": 1.7174107821863149e-06, + "loss": 0.2513, + "step": 473 + }, + { + "epoch": 0.27, + "grad_norm": 1.4694234452554027, + "learning_rate": 1.7161311919716077e-06, + "loss": 0.2636, + "step": 474 + }, + { + "epoch": 0.27, + "grad_norm": 1.4820202577787849, + "learning_rate": 1.7148491902711882e-06, + "loss": 0.2567, + "step": 475 + }, + { + "epoch": 0.27, + "grad_norm": 1.4554857534150158, + "learning_rate": 1.713564781402042e-06, + "loss": 0.2449, + "step": 476 + }, + { + "epoch": 0.27, + "grad_norm": 1.4954620029036076, + "learning_rate": 1.7122779696892626e-06, + "loss": 0.228, + "step": 477 + }, + { + "epoch": 0.27, + "grad_norm": 1.73293954412979, + "learning_rate": 1.7109887594660335e-06, + "loss": 0.2514, + "step": 478 + }, + { + "epoch": 0.27, + "grad_norm": 1.3942916532008032, + "learning_rate": 1.709697155073615e-06, + "loss": 0.2486, + "step": 479 + }, + { + "epoch": 0.27, + "grad_norm": 1.431042232019366, + "learning_rate": 1.7084031608613296e-06, + "loss": 0.241, + "step": 480 + }, + { + "epoch": 0.27, + "grad_norm": 2.0896085623211715, + "learning_rate": 1.7071067811865474e-06, + "loss": 0.2373, + "step": 481 + }, + { + "epoch": 0.27, + "grad_norm": 1.497697061904849, + "learning_rate": 1.7058080204146715e-06, + "loss": 0.2529, + "step": 482 + }, + { + "epoch": 0.27, + "grad_norm": 1.4228466841677148, + "learning_rate": 1.7045068829191215e-06, + "loss": 0.2647, + "step": 483 + }, + { + "epoch": 0.27, + "grad_norm": 1.7833734496134566, + "learning_rate": 1.7032033730813233e-06, + "loss": 0.259, + "step": 484 + }, + { + "epoch": 0.27, + "grad_norm": 2.2616552646977355, + "learning_rate": 1.7018974952906882e-06, + "loss": 0.249, + "step": 485 + }, + { + "epoch": 0.28, + "grad_norm": 1.5650448508674943, + "learning_rate": 1.7005892539446033e-06, + "loss": 0.2363, + "step": 486 + }, + { + "epoch": 0.28, + "grad_norm": 1.9489724413907807, + "learning_rate": 1.6992786534484137e-06, + "loss": 0.2414, + "step": 487 + }, + { + "epoch": 0.28, + "grad_norm": 1.4293350033564403, + "learning_rate": 1.6979656982154093e-06, + "loss": 0.2282, + "step": 488 + }, + { + "epoch": 0.28, + "grad_norm": 1.4372108309514575, + "learning_rate": 1.6966503926668091e-06, + "loss": 0.2372, + "step": 489 + }, + { + "epoch": 0.28, + "grad_norm": 1.3027006415270381, + "learning_rate": 1.6953327412317467e-06, + "loss": 0.2304, + "step": 490 + }, + { + "epoch": 0.28, + "grad_norm": 1.6103229608755254, + "learning_rate": 1.6940127483472542e-06, + "loss": 0.2489, + "step": 491 + }, + { + "epoch": 0.28, + "grad_norm": 2.532446895620965, + "learning_rate": 1.6926904184582496e-06, + "loss": 0.241, + "step": 492 + }, + { + "epoch": 0.28, + "grad_norm": 1.4695172920292021, + "learning_rate": 1.6913657560175195e-06, + "loss": 0.2289, + "step": 493 + }, + { + "epoch": 0.28, + "grad_norm": 1.7748582684515035, + "learning_rate": 1.6900387654857052e-06, + "loss": 0.2391, + "step": 494 + }, + { + "epoch": 0.28, + "grad_norm": 1.5812444135508126, + "learning_rate": 1.6887094513312887e-06, + "loss": 0.2601, + "step": 495 + }, + { + "epoch": 0.28, + "grad_norm": 1.8602367294580844, + "learning_rate": 1.6873778180305749e-06, + "loss": 0.2496, + "step": 496 + }, + { + "epoch": 0.28, + "grad_norm": 1.8362834681849665, + "learning_rate": 1.6860438700676793e-06, + "loss": 0.2465, + "step": 497 + }, + { + "epoch": 0.28, + "grad_norm": 2.2795025182070305, + "learning_rate": 1.684707611934511e-06, + "loss": 0.2485, + "step": 498 + }, + { + "epoch": 0.28, + "grad_norm": 6.8233847922024555, + "learning_rate": 1.6833690481307588e-06, + "loss": 0.2517, + "step": 499 + }, + { + "epoch": 0.28, + "grad_norm": 1.499814068573968, + "learning_rate": 1.6820281831638757e-06, + "loss": 0.2421, + "step": 500 + }, + { + "epoch": 0.28, + "grad_norm": 2.504738024700644, + "learning_rate": 1.6806850215490629e-06, + "loss": 0.2593, + "step": 501 + }, + { + "epoch": 0.28, + "grad_norm": 1.5180706939565156, + "learning_rate": 1.6793395678092555e-06, + "loss": 0.2443, + "step": 502 + }, + { + "epoch": 0.28, + "grad_norm": 1.5701315323330418, + "learning_rate": 1.6779918264751078e-06, + "loss": 0.2347, + "step": 503 + }, + { + "epoch": 0.29, + "grad_norm": 1.5875402728952401, + "learning_rate": 1.6766418020849767e-06, + "loss": 0.254, + "step": 504 + }, + { + "epoch": 0.29, + "grad_norm": 1.5119259709416892, + "learning_rate": 1.6752894991849064e-06, + "loss": 0.2528, + "step": 505 + }, + { + "epoch": 0.29, + "grad_norm": 1.394478933649667, + "learning_rate": 1.6739349223286151e-06, + "loss": 0.2257, + "step": 506 + }, + { + "epoch": 0.29, + "grad_norm": 1.5507471034325682, + "learning_rate": 1.6725780760774772e-06, + "loss": 0.2468, + "step": 507 + }, + { + "epoch": 0.29, + "grad_norm": 1.3453249216548824, + "learning_rate": 1.67121896500051e-06, + "loss": 0.2394, + "step": 508 + }, + { + "epoch": 0.29, + "grad_norm": 1.511705626860942, + "learning_rate": 1.6698575936743556e-06, + "loss": 0.2638, + "step": 509 + }, + { + "epoch": 0.29, + "grad_norm": 1.6567154427999864, + "learning_rate": 1.6684939666832692e-06, + "loss": 0.261, + "step": 510 + }, + { + "epoch": 0.29, + "grad_norm": 1.8137535308533883, + "learning_rate": 1.6671280886191003e-06, + "loss": 0.2467, + "step": 511 + }, + { + "epoch": 0.29, + "grad_norm": 1.5582867916136136, + "learning_rate": 1.6657599640812792e-06, + "loss": 0.2539, + "step": 512 + }, + { + "epoch": 0.29, + "grad_norm": 3.984173244373959, + "learning_rate": 1.6643895976768009e-06, + "loss": 0.2545, + "step": 513 + }, + { + "epoch": 0.29, + "grad_norm": 1.5715533934321584, + "learning_rate": 1.6630169940202098e-06, + "loss": 0.2494, + "step": 514 + }, + { + "epoch": 0.29, + "grad_norm": 1.3096745928924098, + "learning_rate": 1.661642157733583e-06, + "loss": 0.2313, + "step": 515 + }, + { + "epoch": 0.29, + "grad_norm": 13.601183562179498, + "learning_rate": 1.6602650934465174e-06, + "loss": 0.265, + "step": 516 + }, + { + "epoch": 0.29, + "grad_norm": 1.7565832548198632, + "learning_rate": 1.658885805796111e-06, + "loss": 0.2547, + "step": 517 + }, + { + "epoch": 0.29, + "grad_norm": 2.5203761316567443, + "learning_rate": 1.6575042994269491e-06, + "loss": 0.2512, + "step": 518 + }, + { + "epoch": 0.29, + "grad_norm": 2.9480685309792034, + "learning_rate": 1.6561205789910883e-06, + "loss": 0.2408, + "step": 519 + }, + { + "epoch": 0.29, + "grad_norm": 1.6716362042729596, + "learning_rate": 1.6547346491480413e-06, + "loss": 0.2504, + "step": 520 + }, + { + "epoch": 0.3, + "grad_norm": 1.4594184859527592, + "learning_rate": 1.6533465145647595e-06, + "loss": 0.25, + "step": 521 + }, + { + "epoch": 0.3, + "grad_norm": 1.5346162471211768, + "learning_rate": 1.6519561799156202e-06, + "loss": 0.2521, + "step": 522 + }, + { + "epoch": 0.3, + "grad_norm": 1.571448825225742, + "learning_rate": 1.6505636498824072e-06, + "loss": 0.2357, + "step": 523 + }, + { + "epoch": 0.3, + "grad_norm": 1.404127365889556, + "learning_rate": 1.6491689291542988e-06, + "loss": 0.2399, + "step": 524 + }, + { + "epoch": 0.3, + "grad_norm": 1.3108564452782112, + "learning_rate": 1.6477720224278491e-06, + "loss": 0.2446, + "step": 525 + }, + { + "epoch": 0.3, + "grad_norm": 1.8644890179569715, + "learning_rate": 1.6463729344069734e-06, + "loss": 0.2378, + "step": 526 + }, + { + "epoch": 0.3, + "grad_norm": 2.0168442648773475, + "learning_rate": 1.6449716698029332e-06, + "loss": 0.2552, + "step": 527 + }, + { + "epoch": 0.3, + "grad_norm": 1.7567198705317597, + "learning_rate": 1.6435682333343176e-06, + "loss": 0.2461, + "step": 528 + }, + { + "epoch": 0.3, + "grad_norm": 1.3514730242494788, + "learning_rate": 1.6421626297270312e-06, + "loss": 0.2432, + "step": 529 + }, + { + "epoch": 0.3, + "grad_norm": 1.4791734006035535, + "learning_rate": 1.6407548637142746e-06, + "loss": 0.2496, + "step": 530 + }, + { + "epoch": 0.3, + "grad_norm": 1.4449855391993538, + "learning_rate": 1.6393449400365305e-06, + "loss": 0.26, + "step": 531 + }, + { + "epoch": 0.3, + "grad_norm": 1.3966386734231178, + "learning_rate": 1.6379328634415482e-06, + "loss": 0.2357, + "step": 532 + }, + { + "epoch": 0.3, + "grad_norm": 2.5075032731963334, + "learning_rate": 1.6365186386843249e-06, + "loss": 0.2448, + "step": 533 + }, + { + "epoch": 0.3, + "grad_norm": 2.220758069571181, + "learning_rate": 1.635102270527093e-06, + "loss": 0.2701, + "step": 534 + }, + { + "epoch": 0.3, + "grad_norm": 11.58565477454242, + "learning_rate": 1.6336837637393024e-06, + "loss": 0.2552, + "step": 535 + }, + { + "epoch": 0.3, + "grad_norm": 1.6818113484302808, + "learning_rate": 1.6322631230976034e-06, + "loss": 0.2455, + "step": 536 + }, + { + "epoch": 0.3, + "grad_norm": 1.5284239909672301, + "learning_rate": 1.6308403533858333e-06, + "loss": 0.2665, + "step": 537 + }, + { + "epoch": 0.3, + "grad_norm": 1.832182690361937, + "learning_rate": 1.6294154593949978e-06, + "loss": 0.2445, + "step": 538 + }, + { + "epoch": 0.31, + "grad_norm": 1.5102573779116684, + "learning_rate": 1.6279884459232558e-06, + "loss": 0.2546, + "step": 539 + }, + { + "epoch": 0.31, + "grad_norm": 1.656441969633844, + "learning_rate": 1.6265593177759046e-06, + "loss": 0.2536, + "step": 540 + }, + { + "epoch": 0.31, + "grad_norm": 1.276277241164629, + "learning_rate": 1.6251280797653604e-06, + "loss": 0.2398, + "step": 541 + }, + { + "epoch": 0.31, + "grad_norm": 2.2916378433516953, + "learning_rate": 1.6236947367111458e-06, + "loss": 0.2443, + "step": 542 + }, + { + "epoch": 0.31, + "grad_norm": 2.328588318706884, + "learning_rate": 1.6222592934398711e-06, + "loss": 0.2326, + "step": 543 + }, + { + "epoch": 0.31, + "grad_norm": 3.699419657474224, + "learning_rate": 1.620821754785219e-06, + "loss": 0.253, + "step": 544 + }, + { + "epoch": 0.31, + "grad_norm": 3.2591361809157973, + "learning_rate": 1.6193821255879279e-06, + "loss": 0.2492, + "step": 545 + }, + { + "epoch": 0.31, + "grad_norm": 3.4528952166263744, + "learning_rate": 1.6179404106957768e-06, + "loss": 0.2352, + "step": 546 + }, + { + "epoch": 0.31, + "grad_norm": 1.93037966555756, + "learning_rate": 1.6164966149635662e-06, + "loss": 0.2384, + "step": 547 + }, + { + "epoch": 0.31, + "grad_norm": 1.431970805701815, + "learning_rate": 1.6150507432531057e-06, + "loss": 0.2606, + "step": 548 + }, + { + "epoch": 0.31, + "grad_norm": 1.4417996810550193, + "learning_rate": 1.613602800433194e-06, + "loss": 0.2524, + "step": 549 + }, + { + "epoch": 0.31, + "grad_norm": 1.4688489486090681, + "learning_rate": 1.6121527913796043e-06, + "loss": 0.2431, + "step": 550 + }, + { + "epoch": 0.31, + "grad_norm": 1.3079982404963681, + "learning_rate": 1.6107007209750685e-06, + "loss": 0.2323, + "step": 551 + }, + { + "epoch": 0.31, + "grad_norm": 1.467565275670385, + "learning_rate": 1.6092465941092584e-06, + "loss": 0.2381, + "step": 552 + }, + { + "epoch": 0.31, + "grad_norm": 1.55919685015583, + "learning_rate": 1.6077904156787721e-06, + "loss": 0.2432, + "step": 553 + }, + { + "epoch": 0.31, + "grad_norm": 1.6213242392224096, + "learning_rate": 1.606332190587115e-06, + "loss": 0.2557, + "step": 554 + }, + { + "epoch": 0.31, + "grad_norm": 1.5469132712208518, + "learning_rate": 1.6048719237446849e-06, + "loss": 0.2425, + "step": 555 + }, + { + "epoch": 0.32, + "grad_norm": 1.5315139803248468, + "learning_rate": 1.603409620068755e-06, + "loss": 0.2282, + "step": 556 + }, + { + "epoch": 0.32, + "grad_norm": 1.57462813703486, + "learning_rate": 1.601945284483457e-06, + "loss": 0.2388, + "step": 557 + }, + { + "epoch": 0.32, + "grad_norm": 2.210990690597858, + "learning_rate": 1.6004789219197652e-06, + "loss": 0.2332, + "step": 558 + }, + { + "epoch": 0.32, + "grad_norm": 1.3945271541703466, + "learning_rate": 1.599010537315479e-06, + "loss": 0.2348, + "step": 559 + }, + { + "epoch": 0.32, + "grad_norm": 2.222087861663824, + "learning_rate": 1.5975401356152074e-06, + "loss": 0.2419, + "step": 560 + }, + { + "epoch": 0.32, + "grad_norm": 1.624595690795376, + "learning_rate": 1.596067721770351e-06, + "loss": 0.2431, + "step": 561 + }, + { + "epoch": 0.32, + "grad_norm": 1.7231652623202083, + "learning_rate": 1.5945933007390869e-06, + "loss": 0.2341, + "step": 562 + }, + { + "epoch": 0.32, + "grad_norm": 2.2704264718260427, + "learning_rate": 1.59311687748635e-06, + "loss": 0.2439, + "step": 563 + }, + { + "epoch": 0.32, + "grad_norm": 1.772855516533112, + "learning_rate": 1.5916384569838191e-06, + "loss": 0.2372, + "step": 564 + }, + { + "epoch": 0.32, + "grad_norm": 3.5245200150646485, + "learning_rate": 1.5901580442098968e-06, + "loss": 0.2478, + "step": 565 + }, + { + "epoch": 0.32, + "grad_norm": 1.6934029719701937, + "learning_rate": 1.5886756441496952e-06, + "loss": 0.2422, + "step": 566 + }, + { + "epoch": 0.32, + "grad_norm": 2.6670361413686003, + "learning_rate": 1.5871912617950188e-06, + "loss": 0.2512, + "step": 567 + }, + { + "epoch": 0.32, + "grad_norm": 1.9023601002273238, + "learning_rate": 1.5857049021443464e-06, + "loss": 0.2322, + "step": 568 + }, + { + "epoch": 0.32, + "grad_norm": 1.694549477666204, + "learning_rate": 1.5842165702028157e-06, + "loss": 0.2352, + "step": 569 + }, + { + "epoch": 0.32, + "grad_norm": 1.3979709252059453, + "learning_rate": 1.5827262709822053e-06, + "loss": 0.2454, + "step": 570 + }, + { + "epoch": 0.32, + "grad_norm": 1.893165295789768, + "learning_rate": 1.581234009500919e-06, + "loss": 0.2347, + "step": 571 + }, + { + "epoch": 0.32, + "grad_norm": 1.53281994906108, + "learning_rate": 1.5797397907839675e-06, + "loss": 0.2409, + "step": 572 + }, + { + "epoch": 0.32, + "grad_norm": 1.610407219402598, + "learning_rate": 1.5782436198629537e-06, + "loss": 0.2581, + "step": 573 + }, + { + "epoch": 0.33, + "grad_norm": 2.9006814764755284, + "learning_rate": 1.5767455017760523e-06, + "loss": 0.2451, + "step": 574 + }, + { + "epoch": 0.33, + "grad_norm": 1.4264945941320997, + "learning_rate": 1.5752454415679965e-06, + "loss": 0.2445, + "step": 575 + }, + { + "epoch": 0.33, + "grad_norm": 1.4978391958097854, + "learning_rate": 1.5737434442900583e-06, + "loss": 0.2382, + "step": 576 + }, + { + "epoch": 0.33, + "grad_norm": 2.1231475048486197, + "learning_rate": 1.5722395150000333e-06, + "loss": 0.2348, + "step": 577 + }, + { + "epoch": 0.33, + "grad_norm": 1.460810019376956, + "learning_rate": 1.570733658762223e-06, + "loss": 0.2429, + "step": 578 + }, + { + "epoch": 0.33, + "grad_norm": 1.993123316902666, + "learning_rate": 1.5692258806474165e-06, + "loss": 0.2417, + "step": 579 + }, + { + "epoch": 0.33, + "grad_norm": 1.453544059521118, + "learning_rate": 1.5677161857328758e-06, + "loss": 0.2534, + "step": 580 + }, + { + "epoch": 0.33, + "grad_norm": 1.507046324479192, + "learning_rate": 1.566204579102317e-06, + "loss": 0.24, + "step": 581 + }, + { + "epoch": 0.33, + "grad_norm": 1.4222461341313137, + "learning_rate": 1.5646910658458938e-06, + "loss": 0.2338, + "step": 582 + }, + { + "epoch": 0.33, + "grad_norm": 1.3258346979751718, + "learning_rate": 1.5631756510601802e-06, + "loss": 0.2425, + "step": 583 + }, + { + "epoch": 0.33, + "grad_norm": 1.7953242507064269, + "learning_rate": 1.5616583398481531e-06, + "loss": 0.2342, + "step": 584 + }, + { + "epoch": 0.33, + "grad_norm": 1.2481046588599674, + "learning_rate": 1.5601391373191757e-06, + "loss": 0.2442, + "step": 585 + }, + { + "epoch": 0.33, + "grad_norm": 1.5926068926765424, + "learning_rate": 1.5586180485889802e-06, + "loss": 0.2489, + "step": 586 + }, + { + "epoch": 0.33, + "grad_norm": 1.6072234068027944, + "learning_rate": 1.5570950787796498e-06, + "loss": 0.2444, + "step": 587 + }, + { + "epoch": 0.33, + "grad_norm": 1.3243454167963755, + "learning_rate": 1.5555702330196021e-06, + "loss": 0.2282, + "step": 588 + }, + { + "epoch": 0.33, + "grad_norm": 1.9066978756586013, + "learning_rate": 1.5540435164435723e-06, + "loss": 0.2425, + "step": 589 + }, + { + "epoch": 0.33, + "grad_norm": 1.3804190807524854, + "learning_rate": 1.552514934192595e-06, + "loss": 0.2498, + "step": 590 + }, + { + "epoch": 0.33, + "grad_norm": 1.745554122341435, + "learning_rate": 1.550984491413987e-06, + "loss": 0.2407, + "step": 591 + }, + { + "epoch": 0.34, + "grad_norm": 2.1080286514741533, + "learning_rate": 1.5494521932613305e-06, + "loss": 0.2346, + "step": 592 + }, + { + "epoch": 0.34, + "grad_norm": 1.6103571517460866, + "learning_rate": 1.5479180448944548e-06, + "loss": 0.2336, + "step": 593 + }, + { + "epoch": 0.34, + "grad_norm": 1.7179921691633244, + "learning_rate": 1.5463820514794212e-06, + "loss": 0.2343, + "step": 594 + }, + { + "epoch": 0.34, + "grad_norm": 1.4463733422442258, + "learning_rate": 1.5448442181885022e-06, + "loss": 0.2283, + "step": 595 + }, + { + "epoch": 0.34, + "grad_norm": 1.7018545113045958, + "learning_rate": 1.5433045502001665e-06, + "loss": 0.2376, + "step": 596 + }, + { + "epoch": 0.34, + "grad_norm": 1.42814722451918, + "learning_rate": 1.5417630526990612e-06, + "loss": 0.2348, + "step": 597 + }, + { + "epoch": 0.34, + "grad_norm": 1.6409457718379896, + "learning_rate": 1.540219730875994e-06, + "loss": 0.2557, + "step": 598 + }, + { + "epoch": 0.34, + "grad_norm": 2.52242951731036, + "learning_rate": 1.5386745899279153e-06, + "loss": 0.2435, + "step": 599 + }, + { + "epoch": 0.34, + "grad_norm": 1.632809515814534, + "learning_rate": 1.5371276350579017e-06, + "loss": 0.2433, + "step": 600 + }, + { + "epoch": 0.34, + "grad_norm": 1.5085269163324384, + "learning_rate": 1.5355788714751377e-06, + "loss": 0.2328, + "step": 601 + }, + { + "epoch": 0.34, + "grad_norm": 4.284329122719788, + "learning_rate": 1.5340283043948984e-06, + "loss": 0.2555, + "step": 602 + }, + { + "epoch": 0.34, + "grad_norm": 1.4186971896611364, + "learning_rate": 1.5324759390385324e-06, + "loss": 0.2258, + "step": 603 + }, + { + "epoch": 0.34, + "grad_norm": 1.4274452170413565, + "learning_rate": 1.5309217806334437e-06, + "loss": 0.226, + "step": 604 + }, + { + "epoch": 0.34, + "grad_norm": 1.4266812766580574, + "learning_rate": 1.5293658344130733e-06, + "loss": 0.2106, + "step": 605 + }, + { + "epoch": 0.34, + "grad_norm": 5.534877981606875, + "learning_rate": 1.527808105616883e-06, + "loss": 0.2457, + "step": 606 + }, + { + "epoch": 0.34, + "grad_norm": 1.3884602270041906, + "learning_rate": 1.5262485994903377e-06, + "loss": 0.232, + "step": 607 + }, + { + "epoch": 0.34, + "grad_norm": 1.3932552887625036, + "learning_rate": 1.524687321284887e-06, + "loss": 0.2308, + "step": 608 + }, + { + "epoch": 0.35, + "grad_norm": 4.337587362546552, + "learning_rate": 1.5231242762579472e-06, + "loss": 0.2396, + "step": 609 + }, + { + "epoch": 0.35, + "grad_norm": 1.508045670338887, + "learning_rate": 1.5215594696728847e-06, + "loss": 0.2505, + "step": 610 + }, + { + "epoch": 0.35, + "grad_norm": 1.479941382550365, + "learning_rate": 1.5199929067989974e-06, + "loss": 0.2351, + "step": 611 + }, + { + "epoch": 0.35, + "grad_norm": 1.3299254794638724, + "learning_rate": 1.5184245929114979e-06, + "loss": 0.2483, + "step": 612 + }, + { + "epoch": 0.35, + "grad_norm": 8.799198407832911, + "learning_rate": 1.5168545332914939e-06, + "loss": 0.2719, + "step": 613 + }, + { + "epoch": 0.35, + "grad_norm": 1.5428168526294368, + "learning_rate": 1.5152827332259736e-06, + "loss": 0.2378, + "step": 614 + }, + { + "epoch": 0.35, + "grad_norm": 1.5893678352580483, + "learning_rate": 1.5137091980077841e-06, + "loss": 0.2476, + "step": 615 + }, + { + "epoch": 0.35, + "grad_norm": 3.0175664109791964, + "learning_rate": 1.5121339329356162e-06, + "loss": 0.2524, + "step": 616 + }, + { + "epoch": 0.35, + "grad_norm": 1.471277714393818, + "learning_rate": 1.5105569433139858e-06, + "loss": 0.2378, + "step": 617 + }, + { + "epoch": 0.35, + "grad_norm": 1.4791048638564068, + "learning_rate": 1.5089782344532155e-06, + "loss": 0.232, + "step": 618 + }, + { + "epoch": 0.35, + "grad_norm": 1.9362092094555123, + "learning_rate": 1.507397811669418e-06, + "loss": 0.2504, + "step": 619 + }, + { + "epoch": 0.35, + "grad_norm": 1.2715850282844818, + "learning_rate": 1.5058156802844773e-06, + "loss": 0.2257, + "step": 620 + }, + { + "epoch": 0.35, + "grad_norm": 1.4521301345899846, + "learning_rate": 1.5042318456260303e-06, + "loss": 0.2365, + "step": 621 + }, + { + "epoch": 0.35, + "grad_norm": 1.7242427371506575, + "learning_rate": 1.5026463130274498e-06, + "loss": 0.2592, + "step": 622 + }, + { + "epoch": 0.35, + "grad_norm": 2.0353533791375185, + "learning_rate": 1.5010590878278265e-06, + "loss": 0.247, + "step": 623 + }, + { + "epoch": 0.35, + "grad_norm": 1.5663278785887618, + "learning_rate": 1.4994701753719503e-06, + "loss": 0.2674, + "step": 624 + }, + { + "epoch": 0.35, + "grad_norm": 1.452619304594392, + "learning_rate": 1.497879581010293e-06, + "loss": 0.2648, + "step": 625 + }, + { + "epoch": 0.35, + "grad_norm": 1.6958748760538278, + "learning_rate": 1.4962873100989899e-06, + "loss": 0.2568, + "step": 626 + }, + { + "epoch": 0.36, + "grad_norm": 1.5671504252150241, + "learning_rate": 1.4946933679998218e-06, + "loss": 0.2251, + "step": 627 + }, + { + "epoch": 0.36, + "grad_norm": 1.6043080892370902, + "learning_rate": 1.4930977600801976e-06, + "loss": 0.2612, + "step": 628 + }, + { + "epoch": 0.36, + "grad_norm": 1.687772844579193, + "learning_rate": 1.4915004917131344e-06, + "loss": 0.2614, + "step": 629 + }, + { + "epoch": 0.36, + "grad_norm": 1.891915737013904, + "learning_rate": 1.489901568277242e-06, + "loss": 0.2465, + "step": 630 + }, + { + "epoch": 0.36, + "grad_norm": 1.5303641552196703, + "learning_rate": 1.488300995156703e-06, + "loss": 0.2487, + "step": 631 + }, + { + "epoch": 0.36, + "grad_norm": 5.978683333749019, + "learning_rate": 1.4866987777412544e-06, + "loss": 0.244, + "step": 632 + }, + { + "epoch": 0.36, + "grad_norm": 2.3407629150290687, + "learning_rate": 1.4850949214261714e-06, + "loss": 0.2376, + "step": 633 + }, + { + "epoch": 0.36, + "grad_norm": 1.2700786130005546, + "learning_rate": 1.4834894316122468e-06, + "loss": 0.2298, + "step": 634 + }, + { + "epoch": 0.36, + "grad_norm": 1.559856046145795, + "learning_rate": 1.481882313705775e-06, + "loss": 0.2364, + "step": 635 + }, + { + "epoch": 0.36, + "grad_norm": 1.3746156504891356, + "learning_rate": 1.4802735731185325e-06, + "loss": 0.2419, + "step": 636 + }, + { + "epoch": 0.36, + "grad_norm": 1.6726222502356716, + "learning_rate": 1.4786632152677594e-06, + "loss": 0.2564, + "step": 637 + }, + { + "epoch": 0.36, + "grad_norm": 1.5285915969154347, + "learning_rate": 1.4770512455761424e-06, + "loss": 0.2556, + "step": 638 + }, + { + "epoch": 0.36, + "grad_norm": 1.806908243006475, + "learning_rate": 1.4754376694717955e-06, + "loss": 0.2258, + "step": 639 + }, + { + "epoch": 0.36, + "grad_norm": 1.3660760361469149, + "learning_rate": 1.4738224923882427e-06, + "loss": 0.2244, + "step": 640 + }, + { + "epoch": 0.36, + "grad_norm": 1.373135367525671, + "learning_rate": 1.4722057197643985e-06, + "loss": 0.2323, + "step": 641 + }, + { + "epoch": 0.36, + "grad_norm": 1.463014269196834, + "learning_rate": 1.4705873570445505e-06, + "loss": 0.2331, + "step": 642 + }, + { + "epoch": 0.36, + "grad_norm": 1.6399747398073876, + "learning_rate": 1.4689674096783403e-06, + "loss": 0.2443, + "step": 643 + }, + { + "epoch": 0.36, + "grad_norm": 1.7035037644428856, + "learning_rate": 1.467345883120746e-06, + "loss": 0.2213, + "step": 644 + }, + { + "epoch": 0.37, + "grad_norm": 1.4674625448492324, + "learning_rate": 1.4657227828320636e-06, + "loss": 0.2363, + "step": 645 + }, + { + "epoch": 0.37, + "grad_norm": 1.4251668467018777, + "learning_rate": 1.464098114277888e-06, + "loss": 0.2535, + "step": 646 + }, + { + "epoch": 0.37, + "grad_norm": 1.6312672801461443, + "learning_rate": 1.4624718829290959e-06, + "loss": 0.2547, + "step": 647 + }, + { + "epoch": 0.37, + "grad_norm": 1.6111424554559233, + "learning_rate": 1.460844094261825e-06, + "loss": 0.2437, + "step": 648 + }, + { + "epoch": 0.37, + "grad_norm": 1.4604307428917884, + "learning_rate": 1.4592147537574584e-06, + "loss": 0.2393, + "step": 649 + }, + { + "epoch": 0.37, + "grad_norm": 1.5971086562424008, + "learning_rate": 1.4575838669026044e-06, + "loss": 0.2517, + "step": 650 + }, + { + "epoch": 0.37, + "grad_norm": 44.99782717949104, + "learning_rate": 1.4559514391890785e-06, + "loss": 0.2359, + "step": 651 + }, + { + "epoch": 0.37, + "grad_norm": 1.6544382939149958, + "learning_rate": 1.454317476113885e-06, + "loss": 0.2529, + "step": 652 + }, + { + "epoch": 0.37, + "grad_norm": 2.2117339447654474, + "learning_rate": 1.4526819831791981e-06, + "loss": 0.2446, + "step": 653 + }, + { + "epoch": 0.37, + "grad_norm": 1.2939670086379962, + "learning_rate": 1.451044965892343e-06, + "loss": 0.2232, + "step": 654 + }, + { + "epoch": 0.37, + "grad_norm": 1.4230299949510998, + "learning_rate": 1.4494064297657794e-06, + "loss": 0.2484, + "step": 655 + }, + { + "epoch": 0.37, + "grad_norm": 1.3200605712840578, + "learning_rate": 1.4477663803170806e-06, + "loss": 0.2326, + "step": 656 + }, + { + "epoch": 0.37, + "grad_norm": 1.5386345957762018, + "learning_rate": 1.4461248230689158e-06, + "loss": 0.2666, + "step": 657 + }, + { + "epoch": 0.37, + "grad_norm": 1.7560315161844908, + "learning_rate": 1.444481763549032e-06, + "loss": 0.2558, + "step": 658 + }, + { + "epoch": 0.37, + "grad_norm": 1.4158456607635315, + "learning_rate": 1.442837207290234e-06, + "loss": 0.2472, + "step": 659 + }, + { + "epoch": 0.37, + "grad_norm": 1.7342914651432775, + "learning_rate": 1.4411911598303685e-06, + "loss": 0.232, + "step": 660 + }, + { + "epoch": 0.37, + "grad_norm": 1.4557994165516868, + "learning_rate": 1.4395436267123016e-06, + "loss": 0.218, + "step": 661 + }, + { + "epoch": 0.38, + "grad_norm": 1.4530783031187469, + "learning_rate": 1.437894613483903e-06, + "loss": 0.2507, + "step": 662 + }, + { + "epoch": 0.38, + "grad_norm": 2.65127120279145, + "learning_rate": 1.436244125698027e-06, + "loss": 0.242, + "step": 663 + }, + { + "epoch": 0.38, + "grad_norm": 2.059655918824336, + "learning_rate": 1.4345921689124927e-06, + "loss": 0.2435, + "step": 664 + }, + { + "epoch": 0.38, + "grad_norm": 4.285132910718842, + "learning_rate": 1.4329387486900657e-06, + "loss": 0.2393, + "step": 665 + }, + { + "epoch": 0.38, + "grad_norm": 1.3950307473975863, + "learning_rate": 1.43128387059844e-06, + "loss": 0.2273, + "step": 666 + }, + { + "epoch": 0.38, + "grad_norm": 1.507583565333135, + "learning_rate": 1.429627540210219e-06, + "loss": 0.238, + "step": 667 + }, + { + "epoch": 0.38, + "grad_norm": 1.6246448716329174, + "learning_rate": 1.427969763102896e-06, + "loss": 0.2319, + "step": 668 + }, + { + "epoch": 0.38, + "grad_norm": 1.3302902753083163, + "learning_rate": 1.4263105448588359e-06, + "loss": 0.2525, + "step": 669 + }, + { + "epoch": 0.38, + "grad_norm": 1.8903183848074376, + "learning_rate": 1.424649891065257e-06, + "loss": 0.2392, + "step": 670 + }, + { + "epoch": 0.38, + "grad_norm": 1.5206820011688553, + "learning_rate": 1.4229878073142112e-06, + "loss": 0.2517, + "step": 671 + }, + { + "epoch": 0.38, + "grad_norm": 1.3572466387943245, + "learning_rate": 1.421324299202566e-06, + "loss": 0.238, + "step": 672 + }, + { + "epoch": 0.38, + "grad_norm": 1.6910744091579193, + "learning_rate": 1.4196593723319848e-06, + "loss": 0.246, + "step": 673 + }, + { + "epoch": 0.38, + "grad_norm": 1.4361750576983177, + "learning_rate": 1.4179930323089086e-06, + "loss": 0.2328, + "step": 674 + }, + { + "epoch": 0.38, + "grad_norm": 1.3854353904309153, + "learning_rate": 1.4163252847445372e-06, + "loss": 0.2441, + "step": 675 + }, + { + "epoch": 0.38, + "grad_norm": 1.5610308695845496, + "learning_rate": 1.4146561352548108e-06, + "loss": 0.2292, + "step": 676 + }, + { + "epoch": 0.38, + "grad_norm": 8.181771111565979, + "learning_rate": 1.4129855894603884e-06, + "loss": 0.2442, + "step": 677 + }, + { + "epoch": 0.38, + "grad_norm": 1.4522836868752145, + "learning_rate": 1.411313652986633e-06, + "loss": 0.2389, + "step": 678 + }, + { + "epoch": 0.38, + "grad_norm": 1.4858942139409979, + "learning_rate": 1.4096403314635895e-06, + "loss": 0.2427, + "step": 679 + }, + { + "epoch": 0.39, + "grad_norm": 1.484250969842057, + "learning_rate": 1.407965630525967e-06, + "loss": 0.2514, + "step": 680 + }, + { + "epoch": 0.39, + "grad_norm": 1.5282402079255375, + "learning_rate": 1.4062895558131198e-06, + "loss": 0.2564, + "step": 681 + }, + { + "epoch": 0.39, + "grad_norm": 2.6052911158196297, + "learning_rate": 1.4046121129690277e-06, + "loss": 0.2417, + "step": 682 + }, + { + "epoch": 0.39, + "grad_norm": 3.0949683858356756, + "learning_rate": 1.402933307642278e-06, + "loss": 0.2533, + "step": 683 + }, + { + "epoch": 0.39, + "grad_norm": 6.205603720783408, + "learning_rate": 1.4012531454860461e-06, + "loss": 0.2329, + "step": 684 + }, + { + "epoch": 0.39, + "grad_norm": 1.7252669373083445, + "learning_rate": 1.399571632158076e-06, + "loss": 0.251, + "step": 685 + }, + { + "epoch": 0.39, + "grad_norm": 1.4293962307993813, + "learning_rate": 1.3978887733206613e-06, + "loss": 0.2502, + "step": 686 + }, + { + "epoch": 0.39, + "grad_norm": 1.633609347677397, + "learning_rate": 1.396204574640627e-06, + "loss": 0.2463, + "step": 687 + }, + { + "epoch": 0.39, + "grad_norm": 1.4704426867883098, + "learning_rate": 1.3945190417893105e-06, + "loss": 0.2537, + "step": 688 + }, + { + "epoch": 0.39, + "grad_norm": 1.3237709766489505, + "learning_rate": 1.3928321804425399e-06, + "loss": 0.2378, + "step": 689 + }, + { + "epoch": 0.39, + "grad_norm": 1.3983346305726356, + "learning_rate": 1.391143996280619e-06, + "loss": 0.2359, + "step": 690 + }, + { + "epoch": 0.39, + "grad_norm": 1.4163891827813617, + "learning_rate": 1.3894544949883044e-06, + "loss": 0.2313, + "step": 691 + }, + { + "epoch": 0.39, + "grad_norm": 1.8099217450230456, + "learning_rate": 1.387763682254789e-06, + "loss": 0.2396, + "step": 692 + }, + { + "epoch": 0.39, + "grad_norm": 1.7125173027138036, + "learning_rate": 1.3860715637736814e-06, + "loss": 0.2397, + "step": 693 + }, + { + "epoch": 0.39, + "grad_norm": 1.5034506987913445, + "learning_rate": 1.3843781452429873e-06, + "loss": 0.2365, + "step": 694 + }, + { + "epoch": 0.39, + "grad_norm": 1.5559763354515772, + "learning_rate": 1.3826834323650898e-06, + "loss": 0.2445, + "step": 695 + }, + { + "epoch": 0.39, + "grad_norm": 2.0427163147413303, + "learning_rate": 1.380987430846731e-06, + "loss": 0.2338, + "step": 696 + }, + { + "epoch": 0.39, + "grad_norm": 1.4987290587161617, + "learning_rate": 1.3792901463989916e-06, + "loss": 0.2304, + "step": 697 + }, + { + "epoch": 0.4, + "grad_norm": 1.6412575578042634, + "learning_rate": 1.377591584737274e-06, + "loss": 0.2296, + "step": 698 + }, + { + "epoch": 0.4, + "grad_norm": 1.5051938751091585, + "learning_rate": 1.3758917515812796e-06, + "loss": 0.2376, + "step": 699 + }, + { + "epoch": 0.4, + "grad_norm": 1.4623561245169414, + "learning_rate": 1.3741906526549927e-06, + "loss": 0.2443, + "step": 700 + }, + { + "epoch": 0.4, + "grad_norm": 1.911782647942742, + "learning_rate": 1.3724882936866593e-06, + "loss": 0.2407, + "step": 701 + }, + { + "epoch": 0.4, + "grad_norm": 1.4413271270555867, + "learning_rate": 1.3707846804087683e-06, + "loss": 0.2218, + "step": 702 + }, + { + "epoch": 0.4, + "grad_norm": 1.5528477592981256, + "learning_rate": 1.3690798185580334e-06, + "loss": 0.2493, + "step": 703 + }, + { + "epoch": 0.4, + "grad_norm": 1.4055580862022565, + "learning_rate": 1.3673737138753711e-06, + "loss": 0.2451, + "step": 704 + }, + { + "epoch": 0.4, + "grad_norm": 1.469249694354145, + "learning_rate": 1.3656663721058843e-06, + "loss": 0.2357, + "step": 705 + }, + { + "epoch": 0.4, + "grad_norm": 1.5045432076368959, + "learning_rate": 1.3639577989988416e-06, + "loss": 0.2448, + "step": 706 + }, + { + "epoch": 0.4, + "grad_norm": 2.1301487563589134, + "learning_rate": 1.3622480003076569e-06, + "loss": 0.2466, + "step": 707 + }, + { + "epoch": 0.4, + "grad_norm": 1.4189585132043432, + "learning_rate": 1.3605369817898717e-06, + "loss": 0.2359, + "step": 708 + }, + { + "epoch": 0.4, + "grad_norm": 1.4514728440393176, + "learning_rate": 1.3588247492071359e-06, + "loss": 0.226, + "step": 709 + }, + { + "epoch": 0.4, + "grad_norm": 1.3904903924117855, + "learning_rate": 1.3571113083251864e-06, + "loss": 0.2324, + "step": 710 + }, + { + "epoch": 0.4, + "grad_norm": 1.3953250336303555, + "learning_rate": 1.3553966649138295e-06, + "loss": 0.2383, + "step": 711 + }, + { + "epoch": 0.4, + "grad_norm": 3.7905850295064485, + "learning_rate": 1.3536808247469212e-06, + "loss": 0.2347, + "step": 712 + }, + { + "epoch": 0.4, + "grad_norm": 1.6624559881283447, + "learning_rate": 1.351963793602346e-06, + "loss": 0.2358, + "step": 713 + }, + { + "epoch": 0.4, + "grad_norm": 1.4082708258754444, + "learning_rate": 1.350245577262001e-06, + "loss": 0.238, + "step": 714 + }, + { + "epoch": 0.41, + "grad_norm": 2.169426868125172, + "learning_rate": 1.3485261815117727e-06, + "loss": 0.2366, + "step": 715 + }, + { + "epoch": 0.41, + "grad_norm": 1.5792019452823043, + "learning_rate": 1.34680561214152e-06, + "loss": 0.2313, + "step": 716 + }, + { + "epoch": 0.41, + "grad_norm": 1.4467397128946053, + "learning_rate": 1.3450838749450529e-06, + "loss": 0.2305, + "step": 717 + }, + { + "epoch": 0.41, + "grad_norm": 1.7921644987622443, + "learning_rate": 1.3433609757201145e-06, + "loss": 0.2334, + "step": 718 + }, + { + "epoch": 0.41, + "grad_norm": 1.5235185127916537, + "learning_rate": 1.3416369202683613e-06, + "loss": 0.2611, + "step": 719 + }, + { + "epoch": 0.41, + "grad_norm": 1.635762367890265, + "learning_rate": 1.3399117143953425e-06, + "loss": 0.2654, + "step": 720 + }, + { + "epoch": 0.41, + "grad_norm": 2.3777439887480383, + "learning_rate": 1.3381853639104813e-06, + "loss": 0.246, + "step": 721 + }, + { + "epoch": 0.41, + "grad_norm": 1.4145219668783935, + "learning_rate": 1.3364578746270563e-06, + "loss": 0.2158, + "step": 722 + }, + { + "epoch": 0.41, + "grad_norm": 6.061033422544863, + "learning_rate": 1.3347292523621794e-06, + "loss": 0.2443, + "step": 723 + }, + { + "epoch": 0.41, + "grad_norm": 1.7377303610175208, + "learning_rate": 1.332999502936778e-06, + "loss": 0.242, + "step": 724 + }, + { + "epoch": 0.41, + "grad_norm": 2.070947297246555, + "learning_rate": 1.331268632175576e-06, + "loss": 0.236, + "step": 725 + }, + { + "epoch": 0.41, + "grad_norm": 1.4463717824915143, + "learning_rate": 1.3295366459070723e-06, + "loss": 0.2307, + "step": 726 + }, + { + "epoch": 0.41, + "grad_norm": 1.7099924172725052, + "learning_rate": 1.3278035499635225e-06, + "loss": 0.237, + "step": 727 + }, + { + "epoch": 0.41, + "grad_norm": 2.245187652824694, + "learning_rate": 1.3260693501809193e-06, + "loss": 0.2421, + "step": 728 + }, + { + "epoch": 0.41, + "grad_norm": 1.4242925044873258, + "learning_rate": 1.3243340523989712e-06, + "loss": 0.2466, + "step": 729 + }, + { + "epoch": 0.41, + "grad_norm": 1.8076927166483903, + "learning_rate": 1.3225976624610856e-06, + "loss": 0.2493, + "step": 730 + }, + { + "epoch": 0.41, + "grad_norm": 1.8148339388198276, + "learning_rate": 1.3208601862143467e-06, + "loss": 0.2619, + "step": 731 + }, + { + "epoch": 0.41, + "grad_norm": 2.570532081333281, + "learning_rate": 1.3191216295094969e-06, + "loss": 0.2159, + "step": 732 + }, + { + "epoch": 0.42, + "grad_norm": 1.2785331783860376, + "learning_rate": 1.317381998200917e-06, + "loss": 0.2324, + "step": 733 + }, + { + "epoch": 0.42, + "grad_norm": 1.4804780251006124, + "learning_rate": 1.3156412981466062e-06, + "loss": 0.2433, + "step": 734 + }, + { + "epoch": 0.42, + "grad_norm": 1.7343326109787938, + "learning_rate": 1.3138995352081632e-06, + "loss": 0.2412, + "step": 735 + }, + { + "epoch": 0.42, + "grad_norm": 1.6656617108463214, + "learning_rate": 1.3121567152507644e-06, + "loss": 0.2244, + "step": 736 + }, + { + "epoch": 0.42, + "grad_norm": 2.568935035706889, + "learning_rate": 1.3104128441431473e-06, + "loss": 0.231, + "step": 737 + }, + { + "epoch": 0.42, + "grad_norm": 1.6914806965258493, + "learning_rate": 1.3086679277575878e-06, + "loss": 0.2292, + "step": 738 + }, + { + "epoch": 0.42, + "grad_norm": 1.4719232932654254, + "learning_rate": 1.3069219719698827e-06, + "loss": 0.2371, + "step": 739 + }, + { + "epoch": 0.42, + "grad_norm": 2.230955875734477, + "learning_rate": 1.3051749826593274e-06, + "loss": 0.2442, + "step": 740 + }, + { + "epoch": 0.42, + "grad_norm": 1.3232315136913047, + "learning_rate": 1.303426965708699e-06, + "loss": 0.2444, + "step": 741 + }, + { + "epoch": 0.42, + "grad_norm": 1.4246562123397692, + "learning_rate": 1.3016779270042344e-06, + "loss": 0.2462, + "step": 742 + }, + { + "epoch": 0.42, + "grad_norm": 7.97550610933391, + "learning_rate": 1.2999278724356114e-06, + "loss": 0.2321, + "step": 743 + }, + { + "epoch": 0.42, + "grad_norm": 2.336321289234588, + "learning_rate": 1.298176807895928e-06, + "loss": 0.2297, + "step": 744 + }, + { + "epoch": 0.42, + "grad_norm": 1.4229074241512198, + "learning_rate": 1.2964247392816838e-06, + "loss": 0.2401, + "step": 745 + }, + { + "epoch": 0.42, + "grad_norm": 1.4853670850295457, + "learning_rate": 1.2946716724927589e-06, + "loss": 0.2351, + "step": 746 + }, + { + "epoch": 0.42, + "grad_norm": 1.4895405566694082, + "learning_rate": 1.292917613432396e-06, + "loss": 0.2463, + "step": 747 + }, + { + "epoch": 0.42, + "grad_norm": 2.187389952293629, + "learning_rate": 1.2911625680071767e-06, + "loss": 0.2334, + "step": 748 + }, + { + "epoch": 0.42, + "grad_norm": 2.2031128372635234, + "learning_rate": 1.2894065421270068e-06, + "loss": 0.241, + "step": 749 + }, + { + "epoch": 0.42, + "grad_norm": 1.6084011710115733, + "learning_rate": 1.2876495417050914e-06, + "loss": 0.2412, + "step": 750 + }, + { + "epoch": 0.43, + "grad_norm": 1.793201570615233, + "learning_rate": 1.285891572657919e-06, + "loss": 0.247, + "step": 751 + }, + { + "epoch": 0.43, + "grad_norm": 1.5304091913442672, + "learning_rate": 1.2841326409052389e-06, + "loss": 0.2356, + "step": 752 + }, + { + "epoch": 0.43, + "grad_norm": 1.8788313424834322, + "learning_rate": 1.2823727523700418e-06, + "loss": 0.2334, + "step": 753 + }, + { + "epoch": 0.43, + "grad_norm": 1.4818960093012352, + "learning_rate": 1.2806119129785417e-06, + "loss": 0.2226, + "step": 754 + }, + { + "epoch": 0.43, + "grad_norm": 1.5440322928429044, + "learning_rate": 1.2788501286601532e-06, + "loss": 0.2493, + "step": 755 + }, + { + "epoch": 0.43, + "grad_norm": 1.3874101241721837, + "learning_rate": 1.2770874053474728e-06, + "loss": 0.2466, + "step": 756 + }, + { + "epoch": 0.43, + "grad_norm": 2.0055397623828224, + "learning_rate": 1.2753237489762597e-06, + "loss": 0.2332, + "step": 757 + }, + { + "epoch": 0.43, + "grad_norm": 1.396794085960728, + "learning_rate": 1.273559165485415e-06, + "loss": 0.242, + "step": 758 + }, + { + "epoch": 0.43, + "grad_norm": 1.4665641240013898, + "learning_rate": 1.2717936608169615e-06, + "loss": 0.2318, + "step": 759 + }, + { + "epoch": 0.43, + "grad_norm": 2.5635649259204056, + "learning_rate": 1.2700272409160238e-06, + "loss": 0.2407, + "step": 760 + }, + { + "epoch": 0.43, + "grad_norm": 1.4195120366636143, + "learning_rate": 1.2682599117308081e-06, + "loss": 0.2159, + "step": 761 + }, + { + "epoch": 0.43, + "grad_norm": 2.6135841247336473, + "learning_rate": 1.2664916792125837e-06, + "loss": 0.2373, + "step": 762 + }, + { + "epoch": 0.43, + "grad_norm": 1.3655061968704232, + "learning_rate": 1.2647225493156611e-06, + "loss": 0.2491, + "step": 763 + }, + { + "epoch": 0.43, + "grad_norm": 1.5763596670857873, + "learning_rate": 1.2629525279973718e-06, + "loss": 0.2423, + "step": 764 + }, + { + "epoch": 0.43, + "grad_norm": 1.4578250292551183, + "learning_rate": 1.2611816212180509e-06, + "loss": 0.2367, + "step": 765 + }, + { + "epoch": 0.43, + "grad_norm": 1.8802097748057356, + "learning_rate": 1.2594098349410134e-06, + "loss": 0.2332, + "step": 766 + }, + { + "epoch": 0.43, + "grad_norm": 1.2960015654555697, + "learning_rate": 1.257637175132537e-06, + "loss": 0.2341, + "step": 767 + }, + { + "epoch": 0.44, + "grad_norm": 2.2626472449586528, + "learning_rate": 1.2558636477618401e-06, + "loss": 0.2446, + "step": 768 + }, + { + "epoch": 0.44, + "grad_norm": 2.1635906088905563, + "learning_rate": 1.2540892588010632e-06, + "loss": 0.2199, + "step": 769 + }, + { + "epoch": 0.44, + "grad_norm": 1.5453098243782624, + "learning_rate": 1.252314014225248e-06, + "loss": 0.2476, + "step": 770 + }, + { + "epoch": 0.44, + "grad_norm": 2.552252317598482, + "learning_rate": 1.2505379200123172e-06, + "loss": 0.232, + "step": 771 + }, + { + "epoch": 0.44, + "grad_norm": 1.5988556908811609, + "learning_rate": 1.2487609821430543e-06, + "loss": 0.234, + "step": 772 + }, + { + "epoch": 0.44, + "grad_norm": 2.190749358007932, + "learning_rate": 1.2469832066010842e-06, + "loss": 0.2411, + "step": 773 + }, + { + "epoch": 0.44, + "grad_norm": 1.3747638121438797, + "learning_rate": 1.245204599372852e-06, + "loss": 0.2117, + "step": 774 + }, + { + "epoch": 0.44, + "grad_norm": 1.8828384176438921, + "learning_rate": 1.2434251664476047e-06, + "loss": 0.2485, + "step": 775 + }, + { + "epoch": 0.44, + "grad_norm": 1.5221148751431874, + "learning_rate": 1.2416449138173678e-06, + "loss": 0.2341, + "step": 776 + }, + { + "epoch": 0.44, + "grad_norm": 1.8229743137098027, + "learning_rate": 1.2398638474769287e-06, + "loss": 0.2393, + "step": 777 + }, + { + "epoch": 0.44, + "grad_norm": 1.9086756284202573, + "learning_rate": 1.2380819734238138e-06, + "loss": 0.261, + "step": 778 + }, + { + "epoch": 0.44, + "grad_norm": 1.5100134779139818, + "learning_rate": 1.2362992976582698e-06, + "loss": 0.2321, + "step": 779 + }, + { + "epoch": 0.44, + "grad_norm": 1.7137549352423147, + "learning_rate": 1.2345158261832433e-06, + "loss": 0.2375, + "step": 780 + }, + { + "epoch": 0.44, + "grad_norm": 2.11114580256908, + "learning_rate": 1.2327315650043604e-06, + "loss": 0.2382, + "step": 781 + }, + { + "epoch": 0.44, + "grad_norm": 1.6359697648719982, + "learning_rate": 1.2309465201299053e-06, + "loss": 0.2369, + "step": 782 + }, + { + "epoch": 0.44, + "grad_norm": 1.823882461016867, + "learning_rate": 1.2291606975708029e-06, + "loss": 0.2224, + "step": 783 + }, + { + "epoch": 0.44, + "grad_norm": 1.523948224892671, + "learning_rate": 1.2273741033405956e-06, + "loss": 0.2393, + "step": 784 + }, + { + "epoch": 0.44, + "grad_norm": 2.2285621538970166, + "learning_rate": 1.2255867434554245e-06, + "loss": 0.2502, + "step": 785 + }, + { + "epoch": 0.45, + "grad_norm": 1.5958187504880683, + "learning_rate": 1.2237986239340101e-06, + "loss": 0.2185, + "step": 786 + }, + { + "epoch": 0.45, + "grad_norm": 2.186458287348708, + "learning_rate": 1.222009750797629e-06, + "loss": 0.2394, + "step": 787 + }, + { + "epoch": 0.45, + "grad_norm": 3.403432871532403, + "learning_rate": 1.220220130070097e-06, + "loss": 0.2142, + "step": 788 + }, + { + "epoch": 0.45, + "grad_norm": 1.8832049654134484, + "learning_rate": 1.2184297677777462e-06, + "loss": 0.2491, + "step": 789 + }, + { + "epoch": 0.45, + "grad_norm": 1.839382832497527, + "learning_rate": 1.2166386699494072e-06, + "loss": 0.2377, + "step": 790 + }, + { + "epoch": 0.45, + "grad_norm": 3.3472688751471207, + "learning_rate": 1.2148468426163862e-06, + "loss": 0.2269, + "step": 791 + }, + { + "epoch": 0.45, + "grad_norm": 1.5478700836262334, + "learning_rate": 1.2130542918124467e-06, + "loss": 0.2559, + "step": 792 + }, + { + "epoch": 0.45, + "grad_norm": 7.722116112920835, + "learning_rate": 1.2112610235737876e-06, + "loss": 0.2508, + "step": 793 + }, + { + "epoch": 0.45, + "grad_norm": 1.8723490970359669, + "learning_rate": 1.2094670439390243e-06, + "loss": 0.2282, + "step": 794 + }, + { + "epoch": 0.45, + "grad_norm": 1.524774955720685, + "learning_rate": 1.207672358949168e-06, + "loss": 0.2385, + "step": 795 + }, + { + "epoch": 0.45, + "grad_norm": 1.8130800577279438, + "learning_rate": 1.205876974647604e-06, + "loss": 0.2397, + "step": 796 + }, + { + "epoch": 0.45, + "grad_norm": 1.7079887739199653, + "learning_rate": 1.204080897080074e-06, + "loss": 0.2561, + "step": 797 + }, + { + "epoch": 0.45, + "grad_norm": 1.4396323263531854, + "learning_rate": 1.2022841322946526e-06, + "loss": 0.2347, + "step": 798 + }, + { + "epoch": 0.45, + "grad_norm": 1.7839919562683673, + "learning_rate": 1.2004866863417295e-06, + "loss": 0.2403, + "step": 799 + }, + { + "epoch": 0.45, + "grad_norm": 1.739440758365247, + "learning_rate": 1.1986885652739878e-06, + "loss": 0.2457, + "step": 800 + }, + { + "epoch": 0.45, + "grad_norm": 1.3810375022654733, + "learning_rate": 1.196889775146384e-06, + "loss": 0.2342, + "step": 801 + }, + { + "epoch": 0.45, + "grad_norm": 1.6181809274658916, + "learning_rate": 1.1950903220161284e-06, + "loss": 0.2358, + "step": 802 + }, + { + "epoch": 0.45, + "grad_norm": 2.3804010150346278, + "learning_rate": 1.1932902119426622e-06, + "loss": 0.2306, + "step": 803 + }, + { + "epoch": 0.46, + "grad_norm": 1.500141766328542, + "learning_rate": 1.1914894509876398e-06, + "loss": 0.239, + "step": 804 + }, + { + "epoch": 0.46, + "grad_norm": 1.5193866579920117, + "learning_rate": 1.1896880452149077e-06, + "loss": 0.229, + "step": 805 + }, + { + "epoch": 0.46, + "grad_norm": 1.7622757436293444, + "learning_rate": 1.1878860006904832e-06, + "loss": 0.2552, + "step": 806 + }, + { + "epoch": 0.46, + "grad_norm": 1.4741533420013049, + "learning_rate": 1.1860833234825345e-06, + "loss": 0.2455, + "step": 807 + }, + { + "epoch": 0.46, + "grad_norm": 1.5089056184178953, + "learning_rate": 1.1842800196613607e-06, + "loss": 0.2426, + "step": 808 + }, + { + "epoch": 0.46, + "grad_norm": 1.5913601400643171, + "learning_rate": 1.18247609529937e-06, + "loss": 0.239, + "step": 809 + }, + { + "epoch": 0.46, + "grad_norm": 1.450866928480373, + "learning_rate": 1.1806715564710618e-06, + "loss": 0.2315, + "step": 810 + }, + { + "epoch": 0.46, + "grad_norm": 1.3997182563832737, + "learning_rate": 1.1788664092530029e-06, + "loss": 0.2316, + "step": 811 + }, + { + "epoch": 0.46, + "grad_norm": 4.069593617650151, + "learning_rate": 1.17706065972381e-06, + "loss": 0.2333, + "step": 812 + }, + { + "epoch": 0.46, + "grad_norm": 1.5387529950406302, + "learning_rate": 1.1752543139641276e-06, + "loss": 0.2374, + "step": 813 + }, + { + "epoch": 0.46, + "grad_norm": 1.5334244197823872, + "learning_rate": 1.1734473780566078e-06, + "loss": 0.2342, + "step": 814 + }, + { + "epoch": 0.46, + "grad_norm": 2.7971896669713443, + "learning_rate": 1.17163985808589e-06, + "loss": 0.24, + "step": 815 + }, + { + "epoch": 0.46, + "grad_norm": 1.2922849492342743, + "learning_rate": 1.1698317601385805e-06, + "loss": 0.2422, + "step": 816 + }, + { + "epoch": 0.46, + "grad_norm": 2.1812947550324324, + "learning_rate": 1.168023090303232e-06, + "loss": 0.2097, + "step": 817 + }, + { + "epoch": 0.46, + "grad_norm": 1.6151493728901585, + "learning_rate": 1.1662138546703228e-06, + "loss": 0.217, + "step": 818 + }, + { + "epoch": 0.46, + "grad_norm": 4.19418911561461, + "learning_rate": 1.1644040593322364e-06, + "loss": 0.2281, + "step": 819 + }, + { + "epoch": 0.46, + "grad_norm": 1.7171162552551054, + "learning_rate": 1.1625937103832406e-06, + "loss": 0.2269, + "step": 820 + }, + { + "epoch": 0.47, + "grad_norm": 1.7193487960955665, + "learning_rate": 1.1607828139194682e-06, + "loss": 0.2352, + "step": 821 + }, + { + "epoch": 0.47, + "grad_norm": 1.4784522512849319, + "learning_rate": 1.1589713760388957e-06, + "loss": 0.2343, + "step": 822 + }, + { + "epoch": 0.47, + "grad_norm": 2.2368261418131907, + "learning_rate": 1.1571594028413218e-06, + "loss": 0.2408, + "step": 823 + }, + { + "epoch": 0.47, + "grad_norm": 2.0486290073877953, + "learning_rate": 1.155346900428349e-06, + "loss": 0.2128, + "step": 824 + }, + { + "epoch": 0.47, + "grad_norm": 2.1420939607995306, + "learning_rate": 1.1535338749033612e-06, + "loss": 0.2356, + "step": 825 + }, + { + "epoch": 0.47, + "grad_norm": 1.4957737927607986, + "learning_rate": 1.151720332371504e-06, + "loss": 0.2288, + "step": 826 + }, + { + "epoch": 0.47, + "grad_norm": 1.6607090315800468, + "learning_rate": 1.1499062789396633e-06, + "loss": 0.2404, + "step": 827 + }, + { + "epoch": 0.47, + "grad_norm": 1.3671769211850193, + "learning_rate": 1.1480917207164466e-06, + "loss": 0.2509, + "step": 828 + }, + { + "epoch": 0.47, + "grad_norm": 1.9772503935616461, + "learning_rate": 1.146276663812161e-06, + "loss": 0.2283, + "step": 829 + }, + { + "epoch": 0.47, + "grad_norm": 1.4678916330892546, + "learning_rate": 1.1444611143387914e-06, + "loss": 0.244, + "step": 830 + }, + { + "epoch": 0.47, + "grad_norm": 1.5188005764910792, + "learning_rate": 1.1426450784099827e-06, + "loss": 0.2222, + "step": 831 + }, + { + "epoch": 0.47, + "grad_norm": 1.4437137069502337, + "learning_rate": 1.1408285621410183e-06, + "loss": 0.2201, + "step": 832 + }, + { + "epoch": 0.47, + "grad_norm": 3.5210224794297056, + "learning_rate": 1.1390115716487976e-06, + "loss": 0.241, + "step": 833 + }, + { + "epoch": 0.47, + "grad_norm": 1.3936924577704652, + "learning_rate": 1.1371941130518183e-06, + "loss": 0.2275, + "step": 834 + }, + { + "epoch": 0.47, + "grad_norm": 1.7996048398947955, + "learning_rate": 1.1353761924701537e-06, + "loss": 0.2398, + "step": 835 + }, + { + "epoch": 0.47, + "grad_norm": 1.3281395600591441, + "learning_rate": 1.1335578160254325e-06, + "loss": 0.2456, + "step": 836 + }, + { + "epoch": 0.47, + "grad_norm": 2.4363527171535297, + "learning_rate": 1.1317389898408187e-06, + "loss": 0.2163, + "step": 837 + }, + { + "epoch": 0.47, + "grad_norm": 1.431688622876347, + "learning_rate": 1.1299197200409917e-06, + "loss": 0.2366, + "step": 838 + }, + { + "epoch": 0.48, + "grad_norm": 1.9331732709637748, + "learning_rate": 1.1281000127521228e-06, + "loss": 0.2371, + "step": 839 + }, + { + "epoch": 0.48, + "grad_norm": 1.5632225542500495, + "learning_rate": 1.1262798741018582e-06, + "loss": 0.2395, + "step": 840 + }, + { + "epoch": 0.48, + "grad_norm": 1.5431152625799704, + "learning_rate": 1.1244593102192962e-06, + "loss": 0.2163, + "step": 841 + }, + { + "epoch": 0.48, + "grad_norm": 3.8374969939766235, + "learning_rate": 1.1226383272349664e-06, + "loss": 0.2304, + "step": 842 + }, + { + "epoch": 0.48, + "grad_norm": 1.626840201374071, + "learning_rate": 1.1208169312808103e-06, + "loss": 0.2206, + "step": 843 + }, + { + "epoch": 0.48, + "grad_norm": 2.7895304843064457, + "learning_rate": 1.1189951284901597e-06, + "loss": 0.2226, + "step": 844 + }, + { + "epoch": 0.48, + "grad_norm": 1.8593866565803943, + "learning_rate": 1.1171729249977168e-06, + "loss": 0.2391, + "step": 845 + }, + { + "epoch": 0.48, + "grad_norm": 1.6009221387417485, + "learning_rate": 1.115350326939533e-06, + "loss": 0.2144, + "step": 846 + }, + { + "epoch": 0.48, + "grad_norm": 1.9038368506897914, + "learning_rate": 1.1135273404529874e-06, + "loss": 0.2296, + "step": 847 + }, + { + "epoch": 0.48, + "grad_norm": 1.7814726561059937, + "learning_rate": 1.1117039716767687e-06, + "loss": 0.245, + "step": 848 + }, + { + "epoch": 0.48, + "grad_norm": 1.718656303156021, + "learning_rate": 1.1098802267508515e-06, + "loss": 0.2305, + "step": 849 + }, + { + "epoch": 0.48, + "grad_norm": 1.5180049463429541, + "learning_rate": 1.1080561118164783e-06, + "loss": 0.2368, + "step": 850 + }, + { + "epoch": 0.48, + "grad_norm": 1.6094012867037097, + "learning_rate": 1.1062316330161366e-06, + "loss": 0.2221, + "step": 851 + }, + { + "epoch": 0.48, + "grad_norm": 3.048125886891839, + "learning_rate": 1.1044067964935388e-06, + "loss": 0.253, + "step": 852 + }, + { + "epoch": 0.48, + "grad_norm": 1.5430318958857683, + "learning_rate": 1.1025816083936035e-06, + "loss": 0.2267, + "step": 853 + }, + { + "epoch": 0.48, + "grad_norm": 2.403321276134905, + "learning_rate": 1.1007560748624318e-06, + "loss": 0.2244, + "step": 854 + }, + { + "epoch": 0.48, + "grad_norm": 2.0921313212767934, + "learning_rate": 1.0989302020472884e-06, + "loss": 0.2363, + "step": 855 + }, + { + "epoch": 0.48, + "grad_norm": 1.4945977858368698, + "learning_rate": 1.097103996096581e-06, + "loss": 0.2283, + "step": 856 + }, + { + "epoch": 0.49, + "grad_norm": 1.7107548033495468, + "learning_rate": 1.0952774631598377e-06, + "loss": 0.2321, + "step": 857 + }, + { + "epoch": 0.49, + "grad_norm": 1.7269347706523859, + "learning_rate": 1.0934506093876896e-06, + "loss": 0.2459, + "step": 858 + }, + { + "epoch": 0.49, + "grad_norm": 1.6544378940242648, + "learning_rate": 1.0916234409318465e-06, + "loss": 0.234, + "step": 859 + }, + { + "epoch": 0.49, + "grad_norm": 1.5379374358875386, + "learning_rate": 1.0897959639450791e-06, + "loss": 0.2362, + "step": 860 + }, + { + "epoch": 0.49, + "grad_norm": 1.357387368149733, + "learning_rate": 1.0879681845811963e-06, + "loss": 0.2266, + "step": 861 + }, + { + "epoch": 0.49, + "grad_norm": 2.0858747713723886, + "learning_rate": 1.0861401089950254e-06, + "loss": 0.2145, + "step": 862 + }, + { + "epoch": 0.49, + "grad_norm": 1.6464408758752258, + "learning_rate": 1.084311743342391e-06, + "loss": 0.23, + "step": 863 + }, + { + "epoch": 0.49, + "grad_norm": 1.6376114408484301, + "learning_rate": 1.0824830937800947e-06, + "loss": 0.2299, + "step": 864 + }, + { + "epoch": 0.49, + "grad_norm": 1.3658402040192368, + "learning_rate": 1.0806541664658944e-06, + "loss": 0.2099, + "step": 865 + }, + { + "epoch": 0.49, + "grad_norm": 1.5379999961108974, + "learning_rate": 1.0788249675584831e-06, + "loss": 0.2323, + "step": 866 + }, + { + "epoch": 0.49, + "grad_norm": 2.125820269889813, + "learning_rate": 1.0769955032174678e-06, + "loss": 0.2222, + "step": 867 + }, + { + "epoch": 0.49, + "grad_norm": 2.014253739255314, + "learning_rate": 1.07516577960335e-06, + "loss": 0.2338, + "step": 868 + }, + { + "epoch": 0.49, + "grad_norm": 1.4689936945256907, + "learning_rate": 1.073335802877504e-06, + "loss": 0.2289, + "step": 869 + }, + { + "epoch": 0.49, + "grad_norm": 1.5416158261506738, + "learning_rate": 1.0715055792021565e-06, + "loss": 0.2343, + "step": 870 + }, + { + "epoch": 0.49, + "grad_norm": 1.4478998570865615, + "learning_rate": 1.0696751147403657e-06, + "loss": 0.2609, + "step": 871 + }, + { + "epoch": 0.49, + "grad_norm": 1.437933529501175, + "learning_rate": 1.067844415656001e-06, + "loss": 0.237, + "step": 872 + }, + { + "epoch": 0.49, + "grad_norm": 2.4930631242150403, + "learning_rate": 1.0660134881137208e-06, + "loss": 0.2475, + "step": 873 + }, + { + "epoch": 0.5, + "grad_norm": 1.5999432946340677, + "learning_rate": 1.0641823382789544e-06, + "loss": 0.2516, + "step": 874 + }, + { + "epoch": 0.5, + "grad_norm": 2.088772545721101, + "learning_rate": 1.062350972317878e-06, + "loss": 0.2332, + "step": 875 + }, + { + "epoch": 0.5, + "grad_norm": 1.4454856363661157, + "learning_rate": 1.0605193963973964e-06, + "loss": 0.2311, + "step": 876 + }, + { + "epoch": 0.5, + "grad_norm": 11.775387021821023, + "learning_rate": 1.0586876166851223e-06, + "loss": 0.2207, + "step": 877 + }, + { + "epoch": 0.5, + "grad_norm": 1.5147340938064329, + "learning_rate": 1.0568556393493527e-06, + "loss": 0.2212, + "step": 878 + }, + { + "epoch": 0.5, + "grad_norm": 1.68943517311353, + "learning_rate": 1.0550234705590513e-06, + "loss": 0.2342, + "step": 879 + }, + { + "epoch": 0.5, + "grad_norm": 1.6050576246953299, + "learning_rate": 1.0531911164838266e-06, + "loss": 0.244, + "step": 880 + }, + { + "epoch": 0.5, + "grad_norm": 1.5974690766421744, + "learning_rate": 1.0513585832939103e-06, + "loss": 0.2474, + "step": 881 + }, + { + "epoch": 0.5, + "grad_norm": 1.336901910698941, + "learning_rate": 1.0495258771601382e-06, + "loss": 0.2201, + "step": 882 + }, + { + "epoch": 0.5, + "grad_norm": 2.2263875875315953, + "learning_rate": 1.0476930042539273e-06, + "loss": 0.222, + "step": 883 + }, + { + "epoch": 0.5, + "grad_norm": 1.3146657739311325, + "learning_rate": 1.0458599707472572e-06, + "loss": 0.2296, + "step": 884 + }, + { + "epoch": 0.5, + "grad_norm": 1.7532706240665241, + "learning_rate": 1.0440267828126477e-06, + "loss": 0.2244, + "step": 885 + }, + { + "epoch": 0.5, + "grad_norm": 1.336115895913218, + "learning_rate": 1.0421934466231392e-06, + "loss": 0.2129, + "step": 886 + }, + { + "epoch": 0.5, + "grad_norm": 1.3547558095616752, + "learning_rate": 1.0403599683522703e-06, + "loss": 0.2315, + "step": 887 + }, + { + "epoch": 0.5, + "grad_norm": 1.6711691960415507, + "learning_rate": 1.0385263541740597e-06, + "loss": 0.2402, + "step": 888 + }, + { + "epoch": 0.5, + "grad_norm": 1.7748461016033732, + "learning_rate": 1.0366926102629815e-06, + "loss": 0.2241, + "step": 889 + }, + { + "epoch": 0.5, + "grad_norm": 1.4148145835950579, + "learning_rate": 1.034858742793949e-06, + "loss": 0.2255, + "step": 890 + }, + { + "epoch": 0.5, + "grad_norm": 1.512980409571602, + "learning_rate": 1.0330247579422897e-06, + "loss": 0.243, + "step": 891 + }, + { + "epoch": 0.51, + "grad_norm": 1.9993480248897966, + "learning_rate": 1.0311906618837276e-06, + "loss": 0.2214, + "step": 892 + }, + { + "epoch": 0.51, + "grad_norm": 1.6809629188621942, + "learning_rate": 1.0293564607943607e-06, + "loss": 0.2256, + "step": 893 + }, + { + "epoch": 0.51, + "grad_norm": 1.3579602300816607, + "learning_rate": 1.0275221608506407e-06, + "loss": 0.2368, + "step": 894 + }, + { + "epoch": 0.51, + "grad_norm": 1.303906517751258, + "learning_rate": 1.0256877682293518e-06, + "loss": 0.2333, + "step": 895 + }, + { + "epoch": 0.51, + "grad_norm": 1.6066375861873892, + "learning_rate": 1.023853289107591e-06, + "loss": 0.2483, + "step": 896 + }, + { + "epoch": 0.51, + "grad_norm": 1.622002113804314, + "learning_rate": 1.0220187296627463e-06, + "loss": 0.2273, + "step": 897 + }, + { + "epoch": 0.51, + "grad_norm": 1.997730551937032, + "learning_rate": 1.0201840960724763e-06, + "loss": 0.2461, + "step": 898 + }, + { + "epoch": 0.51, + "grad_norm": 2.282238724022589, + "learning_rate": 1.0183493945146886e-06, + "loss": 0.2228, + "step": 899 + }, + { + "epoch": 0.51, + "grad_norm": 2.551788022653091, + "learning_rate": 1.0165146311675204e-06, + "loss": 0.2257, + "step": 900 + }, + { + "epoch": 0.51, + "grad_norm": 1.3761203843599463, + "learning_rate": 1.0146798122093165e-06, + "loss": 0.2054, + "step": 901 + }, + { + "epoch": 0.51, + "grad_norm": 1.4267440867345338, + "learning_rate": 1.0128449438186096e-06, + "loss": 0.2444, + "step": 902 + }, + { + "epoch": 0.51, + "grad_norm": 1.4415973926165948, + "learning_rate": 1.0110100321740984e-06, + "loss": 0.2328, + "step": 903 + }, + { + "epoch": 0.51, + "grad_norm": 2.1117791795040217, + "learning_rate": 1.0091750834546273e-06, + "loss": 0.2211, + "step": 904 + }, + { + "epoch": 0.51, + "grad_norm": 1.9945951073594235, + "learning_rate": 1.0073401038391653e-06, + "loss": 0.2251, + "step": 905 + }, + { + "epoch": 0.51, + "grad_norm": 1.6837694129550345, + "learning_rate": 1.0055050995067863e-06, + "loss": 0.2373, + "step": 906 + }, + { + "epoch": 0.51, + "grad_norm": 1.4913237659032903, + "learning_rate": 1.0036700766366457e-06, + "loss": 0.2327, + "step": 907 + }, + { + "epoch": 0.51, + "grad_norm": 1.9681133457033684, + "learning_rate": 1.0018350414079635e-06, + "loss": 0.2368, + "step": 908 + }, + { + "epoch": 0.52, + "grad_norm": 1.40963052538597, + "learning_rate": 1e-06, + "loss": 0.2465, + "step": 909 + }, + { + "epoch": 0.52, + "grad_norm": 1.6336019559673207, + "learning_rate": 9.981649585920366e-07, + "loss": 0.229, + "step": 910 + }, + { + "epoch": 0.52, + "grad_norm": 2.047855323554034, + "learning_rate": 9.963299233633544e-07, + "loss": 0.2476, + "step": 911 + }, + { + "epoch": 0.52, + "grad_norm": 1.8238043849135097, + "learning_rate": 9.94494900493214e-07, + "loss": 0.2317, + "step": 912 + }, + { + "epoch": 0.52, + "grad_norm": 1.8984519837867802, + "learning_rate": 9.926598961608344e-07, + "loss": 0.2306, + "step": 913 + }, + { + "epoch": 0.52, + "grad_norm": 2.2051962807579297, + "learning_rate": 9.908249165453726e-07, + "loss": 0.223, + "step": 914 + }, + { + "epoch": 0.52, + "grad_norm": 2.0491183570022993, + "learning_rate": 9.889899678259015e-07, + "loss": 0.2138, + "step": 915 + }, + { + "epoch": 0.52, + "grad_norm": 1.8087699425891977, + "learning_rate": 9.871550561813905e-07, + "loss": 0.2383, + "step": 916 + }, + { + "epoch": 0.52, + "grad_norm": 1.9335801588158628, + "learning_rate": 9.853201877906834e-07, + "loss": 0.2248, + "step": 917 + }, + { + "epoch": 0.52, + "grad_norm": 1.460928493974189, + "learning_rate": 9.834853688324795e-07, + "loss": 0.2286, + "step": 918 + }, + { + "epoch": 0.52, + "grad_norm": 1.727351571517714, + "learning_rate": 9.816506054853115e-07, + "loss": 0.2196, + "step": 919 + }, + { + "epoch": 0.52, + "grad_norm": 1.4108583090360522, + "learning_rate": 9.798159039275239e-07, + "loss": 0.2268, + "step": 920 + }, + { + "epoch": 0.52, + "grad_norm": 1.4174505568920779, + "learning_rate": 9.779812703372536e-07, + "loss": 0.2425, + "step": 921 + }, + { + "epoch": 0.52, + "grad_norm": 1.2972251586360597, + "learning_rate": 9.761467108924088e-07, + "loss": 0.2364, + "step": 922 + }, + { + "epoch": 0.52, + "grad_norm": 3.4681627403764774, + "learning_rate": 9.74312231770648e-07, + "loss": 0.2345, + "step": 923 + }, + { + "epoch": 0.52, + "grad_norm": 1.436749559249778, + "learning_rate": 9.724778391493592e-07, + "loss": 0.2262, + "step": 924 + }, + { + "epoch": 0.52, + "grad_norm": 1.4913669267335037, + "learning_rate": 9.706435392056392e-07, + "loss": 0.228, + "step": 925 + }, + { + "epoch": 0.52, + "grad_norm": 1.5724775631725318, + "learning_rate": 9.688093381162724e-07, + "loss": 0.2387, + "step": 926 + }, + { + "epoch": 0.53, + "grad_norm": 1.7369559246126223, + "learning_rate": 9.669752420577103e-07, + "loss": 0.2074, + "step": 927 + }, + { + "epoch": 0.53, + "grad_norm": 1.507120098839554, + "learning_rate": 9.65141257206051e-07, + "loss": 0.2245, + "step": 928 + }, + { + "epoch": 0.53, + "grad_norm": 1.665596560584004, + "learning_rate": 9.633073897370182e-07, + "loss": 0.2031, + "step": 929 + }, + { + "epoch": 0.53, + "grad_norm": 1.3511207868052486, + "learning_rate": 9.614736458259404e-07, + "loss": 0.2195, + "step": 930 + }, + { + "epoch": 0.53, + "grad_norm": 1.5575188014305994, + "learning_rate": 9.596400316477296e-07, + "loss": 0.2313, + "step": 931 + }, + { + "epoch": 0.53, + "grad_norm": 1.4038769376886993, + "learning_rate": 9.57806553376861e-07, + "loss": 0.2262, + "step": 932 + }, + { + "epoch": 0.53, + "grad_norm": 1.4353978876044384, + "learning_rate": 9.559732171873522e-07, + "loss": 0.2268, + "step": 933 + }, + { + "epoch": 0.53, + "grad_norm": 3.9760204908546397, + "learning_rate": 9.541400292527427e-07, + "loss": 0.2312, + "step": 934 + }, + { + "epoch": 0.53, + "grad_norm": 1.6194187291766882, + "learning_rate": 9.523069957460727e-07, + "loss": 0.2277, + "step": 935 + }, + { + "epoch": 0.53, + "grad_norm": 1.4563484404944855, + "learning_rate": 9.50474122839862e-07, + "loss": 0.2342, + "step": 936 + }, + { + "epoch": 0.53, + "grad_norm": 1.438627980155567, + "learning_rate": 9.486414167060897e-07, + "loss": 0.2341, + "step": 937 + }, + { + "epoch": 0.53, + "grad_norm": 2.905277697371015, + "learning_rate": 9.468088835161736e-07, + "loss": 0.2401, + "step": 938 + }, + { + "epoch": 0.53, + "grad_norm": 2.2154129742831827, + "learning_rate": 9.449765294409488e-07, + "loss": 0.254, + "step": 939 + }, + { + "epoch": 0.53, + "grad_norm": 2.8111605023199746, + "learning_rate": 9.431443606506474e-07, + "loss": 0.2153, + "step": 940 + }, + { + "epoch": 0.53, + "grad_norm": 1.6680397678861656, + "learning_rate": 9.41312383314878e-07, + "loss": 0.2262, + "step": 941 + }, + { + "epoch": 0.53, + "grad_norm": 1.5565735646572942, + "learning_rate": 9.394806036026036e-07, + "loss": 0.2158, + "step": 942 + }, + { + "epoch": 0.53, + "grad_norm": 9.977166610258235, + "learning_rate": 9.376490276821223e-07, + "loss": 0.2485, + "step": 943 + }, + { + "epoch": 0.53, + "grad_norm": 1.7637091312467261, + "learning_rate": 9.358176617210458e-07, + "loss": 0.2258, + "step": 944 + }, + { + "epoch": 0.54, + "grad_norm": 1.9953490562691036, + "learning_rate": 9.339865118862791e-07, + "loss": 0.2233, + "step": 945 + }, + { + "epoch": 0.54, + "grad_norm": 1.325624511855323, + "learning_rate": 9.321555843439992e-07, + "loss": 0.2296, + "step": 946 + }, + { + "epoch": 0.54, + "grad_norm": 1.398606567284992, + "learning_rate": 9.303248852596344e-07, + "loss": 0.25, + "step": 947 + }, + { + "epoch": 0.54, + "grad_norm": 1.9371112269095365, + "learning_rate": 9.284944207978438e-07, + "loss": 0.221, + "step": 948 + }, + { + "epoch": 0.54, + "grad_norm": 1.4150928440448036, + "learning_rate": 9.266641971224962e-07, + "loss": 0.2269, + "step": 949 + }, + { + "epoch": 0.54, + "grad_norm": 1.3354287428752776, + "learning_rate": 9.248342203966499e-07, + "loss": 0.2246, + "step": 950 + }, + { + "epoch": 0.54, + "grad_norm": 1.5218673396522144, + "learning_rate": 9.230044967825322e-07, + "loss": 0.2294, + "step": 951 + }, + { + "epoch": 0.54, + "grad_norm": 1.447214742724845, + "learning_rate": 9.211750324415171e-07, + "loss": 0.233, + "step": 952 + }, + { + "epoch": 0.54, + "grad_norm": 1.632835870238796, + "learning_rate": 9.193458335341057e-07, + "loss": 0.2471, + "step": 953 + }, + { + "epoch": 0.54, + "grad_norm": 1.3440993073138094, + "learning_rate": 9.175169062199053e-07, + "loss": 0.2256, + "step": 954 + }, + { + "epoch": 0.54, + "grad_norm": 1.8859391124936533, + "learning_rate": 9.156882566576092e-07, + "loss": 0.229, + "step": 955 + }, + { + "epoch": 0.54, + "grad_norm": 1.5513154452283708, + "learning_rate": 9.138598910049747e-07, + "loss": 0.218, + "step": 956 + }, + { + "epoch": 0.54, + "grad_norm": 1.4966996201802323, + "learning_rate": 9.120318154188038e-07, + "loss": 0.2272, + "step": 957 + }, + { + "epoch": 0.54, + "grad_norm": 1.6145481695253427, + "learning_rate": 9.10204036054921e-07, + "loss": 0.2281, + "step": 958 + }, + { + "epoch": 0.54, + "grad_norm": 1.4910815484636624, + "learning_rate": 9.083765590681537e-07, + "loss": 0.2091, + "step": 959 + }, + { + "epoch": 0.54, + "grad_norm": 7.485583883885374, + "learning_rate": 9.065493906123104e-07, + "loss": 0.2226, + "step": 960 + }, + { + "epoch": 0.54, + "grad_norm": 1.912990247800358, + "learning_rate": 9.047225368401621e-07, + "loss": 0.2329, + "step": 961 + }, + { + "epoch": 0.55, + "grad_norm": 2.0491582437691136, + "learning_rate": 9.028960039034193e-07, + "loss": 0.2249, + "step": 962 + }, + { + "epoch": 0.55, + "grad_norm": 1.5422998027345127, + "learning_rate": 9.010697979527117e-07, + "loss": 0.221, + "step": 963 + }, + { + "epoch": 0.55, + "grad_norm": 1.470840447670189, + "learning_rate": 8.992439251375684e-07, + "loss": 0.2318, + "step": 964 + }, + { + "epoch": 0.55, + "grad_norm": 1.395836164203132, + "learning_rate": 8.974183916063967e-07, + "loss": 0.2416, + "step": 965 + }, + { + "epoch": 0.55, + "grad_norm": 1.4508381856859887, + "learning_rate": 8.95593203506461e-07, + "loss": 0.215, + "step": 966 + }, + { + "epoch": 0.55, + "grad_norm": 1.4840114476738164, + "learning_rate": 8.937683669838634e-07, + "loss": 0.202, + "step": 967 + }, + { + "epoch": 0.55, + "grad_norm": 1.4693327949000587, + "learning_rate": 8.919438881835216e-07, + "loss": 0.2636, + "step": 968 + }, + { + "epoch": 0.55, + "grad_norm": 1.633922747800585, + "learning_rate": 8.901197732491484e-07, + "loss": 0.2224, + "step": 969 + }, + { + "epoch": 0.55, + "grad_norm": 1.3697903692564926, + "learning_rate": 8.882960283232314e-07, + "loss": 0.2195, + "step": 970 + }, + { + "epoch": 0.55, + "grad_norm": 1.2856483439780895, + "learning_rate": 8.864726595470124e-07, + "loss": 0.2167, + "step": 971 + }, + { + "epoch": 0.55, + "grad_norm": 2.1394007640050443, + "learning_rate": 8.84649673060467e-07, + "loss": 0.23, + "step": 972 + }, + { + "epoch": 0.55, + "grad_norm": 1.449785208817842, + "learning_rate": 8.828270750022832e-07, + "loss": 0.2366, + "step": 973 + }, + { + "epoch": 0.55, + "grad_norm": 1.9284588726004648, + "learning_rate": 8.810048715098404e-07, + "loss": 0.2334, + "step": 974 + }, + { + "epoch": 0.55, + "grad_norm": 1.5975610644285747, + "learning_rate": 8.791830687191901e-07, + "loss": 0.2316, + "step": 975 + }, + { + "epoch": 0.55, + "grad_norm": 1.5083380445256873, + "learning_rate": 8.773616727650337e-07, + "loss": 0.2324, + "step": 976 + }, + { + "epoch": 0.55, + "grad_norm": 1.5932319930854235, + "learning_rate": 8.755406897807037e-07, + "loss": 0.2418, + "step": 977 + }, + { + "epoch": 0.55, + "grad_norm": 1.401709820682253, + "learning_rate": 8.737201258981417e-07, + "loss": 0.2362, + "step": 978 + }, + { + "epoch": 0.55, + "grad_norm": 1.547472494610388, + "learning_rate": 8.718999872478774e-07, + "loss": 0.2266, + "step": 979 + }, + { + "epoch": 0.56, + "grad_norm": 1.8623480105122379, + "learning_rate": 8.700802799590088e-07, + "loss": 0.2288, + "step": 980 + }, + { + "epoch": 0.56, + "grad_norm": 1.3986134423435082, + "learning_rate": 8.682610101591813e-07, + "loss": 0.2339, + "step": 981 + }, + { + "epoch": 0.56, + "grad_norm": 2.2426771313954332, + "learning_rate": 8.664421839745677e-07, + "loss": 0.2254, + "step": 982 + }, + { + "epoch": 0.56, + "grad_norm": 1.4153495454843168, + "learning_rate": 8.646238075298463e-07, + "loss": 0.2359, + "step": 983 + }, + { + "epoch": 0.56, + "grad_norm": 1.4758201419063472, + "learning_rate": 8.628058869481816e-07, + "loss": 0.2253, + "step": 984 + }, + { + "epoch": 0.56, + "grad_norm": 1.2952072671726105, + "learning_rate": 8.609884283512024e-07, + "loss": 0.2294, + "step": 985 + }, + { + "epoch": 0.56, + "grad_norm": 1.4157246382630677, + "learning_rate": 8.591714378589818e-07, + "loss": 0.2177, + "step": 986 + }, + { + "epoch": 0.56, + "grad_norm": 1.7710968037027621, + "learning_rate": 8.573549215900172e-07, + "loss": 0.2415, + "step": 987 + }, + { + "epoch": 0.56, + "grad_norm": 1.3984503689540595, + "learning_rate": 8.555388856612087e-07, + "loss": 0.2273, + "step": 988 + }, + { + "epoch": 0.56, + "grad_norm": 1.6649058673490156, + "learning_rate": 8.537233361878393e-07, + "loss": 0.2258, + "step": 989 + }, + { + "epoch": 0.56, + "grad_norm": 1.7195776370628344, + "learning_rate": 8.519082792835534e-07, + "loss": 0.2233, + "step": 990 + }, + { + "epoch": 0.56, + "grad_norm": 1.8240303874219226, + "learning_rate": 8.500937210603368e-07, + "loss": 0.22, + "step": 991 + }, + { + "epoch": 0.56, + "grad_norm": 6.218373254789966, + "learning_rate": 8.482796676284962e-07, + "loss": 0.2358, + "step": 992 + }, + { + "epoch": 0.56, + "grad_norm": 1.751351315178514, + "learning_rate": 8.464661250966386e-07, + "loss": 0.2275, + "step": 993 + }, + { + "epoch": 0.56, + "grad_norm": 1.9659299988921233, + "learning_rate": 8.446530995716508e-07, + "loss": 0.2323, + "step": 994 + }, + { + "epoch": 0.56, + "grad_norm": 1.607533231643826, + "learning_rate": 8.428405971586782e-07, + "loss": 0.2249, + "step": 995 + }, + { + "epoch": 0.56, + "grad_norm": 1.6727062358942648, + "learning_rate": 8.410286239611046e-07, + "loss": 0.2343, + "step": 996 + }, + { + "epoch": 0.56, + "grad_norm": 1.607867862013349, + "learning_rate": 8.392171860805318e-07, + "loss": 0.2168, + "step": 997 + }, + { + "epoch": 0.57, + "grad_norm": 1.3273732710853445, + "learning_rate": 8.374062896167595e-07, + "loss": 0.2267, + "step": 998 + }, + { + "epoch": 0.57, + "grad_norm": 1.684456608378398, + "learning_rate": 8.355959406677637e-07, + "loss": 0.2184, + "step": 999 + }, + { + "epoch": 0.57, + "grad_norm": 1.7359963320718517, + "learning_rate": 8.337861453296772e-07, + "loss": 0.2218, + "step": 1000 + }, + { + "epoch": 0.57, + "grad_norm": 1.621076478485808, + "learning_rate": 8.31976909696768e-07, + "loss": 0.224, + "step": 1001 + }, + { + "epoch": 0.57, + "grad_norm": 1.422820506951909, + "learning_rate": 8.301682398614194e-07, + "loss": 0.2271, + "step": 1002 + }, + { + "epoch": 0.57, + "grad_norm": 1.392580825145082, + "learning_rate": 8.283601419141099e-07, + "loss": 0.2122, + "step": 1003 + }, + { + "epoch": 0.57, + "grad_norm": 1.5493787221249964, + "learning_rate": 8.265526219433922e-07, + "loss": 0.2288, + "step": 1004 + }, + { + "epoch": 0.57, + "grad_norm": 1.9338607556424743, + "learning_rate": 8.247456860358724e-07, + "loss": 0.2138, + "step": 1005 + }, + { + "epoch": 0.57, + "grad_norm": 1.78869360751958, + "learning_rate": 8.229393402761901e-07, + "loss": 0.2288, + "step": 1006 + }, + { + "epoch": 0.57, + "grad_norm": 1.6896708412709238, + "learning_rate": 8.211335907469973e-07, + "loss": 0.2392, + "step": 1007 + }, + { + "epoch": 0.57, + "grad_norm": 2.885106014480173, + "learning_rate": 8.193284435289383e-07, + "loss": 0.2328, + "step": 1008 + }, + { + "epoch": 0.57, + "grad_norm": 1.7426360739276052, + "learning_rate": 8.175239047006297e-07, + "loss": 0.2182, + "step": 1009 + }, + { + "epoch": 0.57, + "grad_norm": 2.824243411756878, + "learning_rate": 8.157199803386395e-07, + "loss": 0.2127, + "step": 1010 + }, + { + "epoch": 0.57, + "grad_norm": 2.4290565263034054, + "learning_rate": 8.139166765174655e-07, + "loss": 0.2223, + "step": 1011 + }, + { + "epoch": 0.57, + "grad_norm": 1.5499150365876613, + "learning_rate": 8.121139993095169e-07, + "loss": 0.2256, + "step": 1012 + }, + { + "epoch": 0.57, + "grad_norm": 1.8290946634480763, + "learning_rate": 8.103119547850923e-07, + "loss": 0.2325, + "step": 1013 + }, + { + "epoch": 0.57, + "grad_norm": 2.154034309451538, + "learning_rate": 8.085105490123601e-07, + "loss": 0.2065, + "step": 1014 + }, + { + "epoch": 0.58, + "grad_norm": 1.4248992315551063, + "learning_rate": 8.06709788057338e-07, + "loss": 0.2332, + "step": 1015 + }, + { + "epoch": 0.58, + "grad_norm": 10.067959275634166, + "learning_rate": 8.049096779838717e-07, + "loss": 0.2323, + "step": 1016 + }, + { + "epoch": 0.58, + "grad_norm": 2.4003835102717757, + "learning_rate": 8.031102248536159e-07, + "loss": 0.2113, + "step": 1017 + }, + { + "epoch": 0.58, + "grad_norm": 1.7499042198611525, + "learning_rate": 8.013114347260123e-07, + "loss": 0.2376, + "step": 1018 + }, + { + "epoch": 0.58, + "grad_norm": 1.4633250103555142, + "learning_rate": 7.995133136582705e-07, + "loss": 0.2194, + "step": 1019 + }, + { + "epoch": 0.58, + "grad_norm": 1.4000338888931103, + "learning_rate": 7.977158677053473e-07, + "loss": 0.2172, + "step": 1020 + }, + { + "epoch": 0.58, + "grad_norm": 1.5223553913898549, + "learning_rate": 7.959191029199261e-07, + "loss": 0.202, + "step": 1021 + }, + { + "epoch": 0.58, + "grad_norm": 1.3659559207789822, + "learning_rate": 7.941230253523958e-07, + "loss": 0.2172, + "step": 1022 + }, + { + "epoch": 0.58, + "grad_norm": 1.6256711286079923, + "learning_rate": 7.923276410508322e-07, + "loss": 0.232, + "step": 1023 + }, + { + "epoch": 0.58, + "grad_norm": 1.4376527392166114, + "learning_rate": 7.905329560609757e-07, + "loss": 0.2311, + "step": 1024 + }, + { + "epoch": 0.58, + "grad_norm": 1.388324338534592, + "learning_rate": 7.887389764262123e-07, + "loss": 0.22, + "step": 1025 + }, + { + "epoch": 0.58, + "grad_norm": 1.7164963976083698, + "learning_rate": 7.869457081875532e-07, + "loss": 0.2347, + "step": 1026 + }, + { + "epoch": 0.58, + "grad_norm": 1.7922203565270243, + "learning_rate": 7.851531573836137e-07, + "loss": 0.2219, + "step": 1027 + }, + { + "epoch": 0.58, + "grad_norm": 1.3632989732266168, + "learning_rate": 7.833613300505929e-07, + "loss": 0.2274, + "step": 1028 + }, + { + "epoch": 0.58, + "grad_norm": 5.00537447094735, + "learning_rate": 7.815702322222537e-07, + "loss": 0.223, + "step": 1029 + }, + { + "epoch": 0.58, + "grad_norm": 1.56990340126771, + "learning_rate": 7.797798699299032e-07, + "loss": 0.2236, + "step": 1030 + }, + { + "epoch": 0.58, + "grad_norm": 1.4746601537517707, + "learning_rate": 7.779902492023709e-07, + "loss": 0.2146, + "step": 1031 + }, + { + "epoch": 0.58, + "grad_norm": 2.0748989265571502, + "learning_rate": 7.7620137606599e-07, + "loss": 0.2234, + "step": 1032 + }, + { + "epoch": 0.59, + "grad_norm": 2.336592795334383, + "learning_rate": 7.744132565445753e-07, + "loss": 0.2173, + "step": 1033 + }, + { + "epoch": 0.59, + "grad_norm": 1.6405872340384273, + "learning_rate": 7.726258966594045e-07, + "loss": 0.2363, + "step": 1034 + }, + { + "epoch": 0.59, + "grad_norm": 1.3869494717196675, + "learning_rate": 7.70839302429197e-07, + "loss": 0.2332, + "step": 1035 + }, + { + "epoch": 0.59, + "grad_norm": 1.5050492910113997, + "learning_rate": 7.690534798700945e-07, + "loss": 0.2276, + "step": 1036 + }, + { + "epoch": 0.59, + "grad_norm": 1.3830077513847308, + "learning_rate": 7.672684349956398e-07, + "loss": 0.2187, + "step": 1037 + }, + { + "epoch": 0.59, + "grad_norm": 1.9732498332999784, + "learning_rate": 7.654841738167567e-07, + "loss": 0.2169, + "step": 1038 + }, + { + "epoch": 0.59, + "grad_norm": 1.2474152123369173, + "learning_rate": 7.637007023417304e-07, + "loss": 0.2292, + "step": 1039 + }, + { + "epoch": 0.59, + "grad_norm": 2.0355442873811427, + "learning_rate": 7.619180265761865e-07, + "loss": 0.2438, + "step": 1040 + }, + { + "epoch": 0.59, + "grad_norm": 1.5015395722355682, + "learning_rate": 7.601361525230712e-07, + "loss": 0.2108, + "step": 1041 + }, + { + "epoch": 0.59, + "grad_norm": 1.4642010654442303, + "learning_rate": 7.58355086182632e-07, + "loss": 0.2452, + "step": 1042 + }, + { + "epoch": 0.59, + "grad_norm": 1.370419830966675, + "learning_rate": 7.565748335523954e-07, + "loss": 0.2245, + "step": 1043 + }, + { + "epoch": 0.59, + "grad_norm": 1.405544142080166, + "learning_rate": 7.54795400627148e-07, + "loss": 0.222, + "step": 1044 + }, + { + "epoch": 0.59, + "grad_norm": 1.5208098606914628, + "learning_rate": 7.53016793398916e-07, + "loss": 0.2304, + "step": 1045 + }, + { + "epoch": 0.59, + "grad_norm": 1.7705303114812307, + "learning_rate": 7.512390178569458e-07, + "loss": 0.1991, + "step": 1046 + }, + { + "epoch": 0.59, + "grad_norm": 1.4503254544004531, + "learning_rate": 7.494620799876828e-07, + "loss": 0.2215, + "step": 1047 + }, + { + "epoch": 0.59, + "grad_norm": 1.4303973835781347, + "learning_rate": 7.47685985774752e-07, + "loss": 0.2345, + "step": 1048 + }, + { + "epoch": 0.59, + "grad_norm": 1.8747660425070338, + "learning_rate": 7.459107411989369e-07, + "loss": 0.2111, + "step": 1049 + }, + { + "epoch": 0.59, + "grad_norm": 1.564262289262657, + "learning_rate": 7.441363522381601e-07, + "loss": 0.2367, + "step": 1050 + }, + { + "epoch": 0.6, + "grad_norm": 1.6484272263998976, + "learning_rate": 7.423628248674632e-07, + "loss": 0.2451, + "step": 1051 + }, + { + "epoch": 0.6, + "grad_norm": 1.4544302761164272, + "learning_rate": 7.405901650589865e-07, + "loss": 0.2035, + "step": 1052 + }, + { + "epoch": 0.6, + "grad_norm": 1.3812846171179063, + "learning_rate": 7.38818378781949e-07, + "loss": 0.2294, + "step": 1053 + }, + { + "epoch": 0.6, + "grad_norm": 1.664606875070604, + "learning_rate": 7.370474720026281e-07, + "loss": 0.2335, + "step": 1054 + }, + { + "epoch": 0.6, + "grad_norm": 2.0364849066672477, + "learning_rate": 7.352774506843393e-07, + "loss": 0.2177, + "step": 1055 + }, + { + "epoch": 0.6, + "grad_norm": 1.6105474255148198, + "learning_rate": 7.335083207874164e-07, + "loss": 0.2165, + "step": 1056 + }, + { + "epoch": 0.6, + "grad_norm": 1.3426561923500535, + "learning_rate": 7.317400882691918e-07, + "loss": 0.2344, + "step": 1057 + }, + { + "epoch": 0.6, + "grad_norm": 1.430775940616281, + "learning_rate": 7.299727590839764e-07, + "loss": 0.2309, + "step": 1058 + }, + { + "epoch": 0.6, + "grad_norm": 1.5889348874105849, + "learning_rate": 7.282063391830386e-07, + "loss": 0.2298, + "step": 1059 + }, + { + "epoch": 0.6, + "grad_norm": 1.5537874246029486, + "learning_rate": 7.264408345145851e-07, + "loss": 0.237, + "step": 1060 + }, + { + "epoch": 0.6, + "grad_norm": 1.3690981306499954, + "learning_rate": 7.246762510237403e-07, + "loss": 0.2294, + "step": 1061 + }, + { + "epoch": 0.6, + "grad_norm": 1.4474280925681746, + "learning_rate": 7.229125946525272e-07, + "loss": 0.2291, + "step": 1062 + }, + { + "epoch": 0.6, + "grad_norm": 1.618241985168829, + "learning_rate": 7.211498713398469e-07, + "loss": 0.2141, + "step": 1063 + }, + { + "epoch": 0.6, + "grad_norm": 1.4584229293581576, + "learning_rate": 7.193880870214582e-07, + "loss": 0.2265, + "step": 1064 + }, + { + "epoch": 0.6, + "grad_norm": 1.8349513763934644, + "learning_rate": 7.176272476299581e-07, + "loss": 0.2234, + "step": 1065 + }, + { + "epoch": 0.6, + "grad_norm": 1.7988168579615764, + "learning_rate": 7.158673590947614e-07, + "loss": 0.2263, + "step": 1066 + }, + { + "epoch": 0.6, + "grad_norm": 1.8708887748581664, + "learning_rate": 7.141084273420808e-07, + "loss": 0.2384, + "step": 1067 + }, + { + "epoch": 0.61, + "grad_norm": 1.707355501953993, + "learning_rate": 7.123504582949083e-07, + "loss": 0.2115, + "step": 1068 + }, + { + "epoch": 0.61, + "grad_norm": 1.3584161352614226, + "learning_rate": 7.105934578729933e-07, + "loss": 0.227, + "step": 1069 + }, + { + "epoch": 0.61, + "grad_norm": 1.3318995792607813, + "learning_rate": 7.088374319928234e-07, + "loss": 0.201, + "step": 1070 + }, + { + "epoch": 0.61, + "grad_norm": 1.6742261804527685, + "learning_rate": 7.070823865676045e-07, + "loss": 0.2331, + "step": 1071 + }, + { + "epoch": 0.61, + "grad_norm": 1.5879862425088984, + "learning_rate": 7.05328327507241e-07, + "loss": 0.2522, + "step": 1072 + }, + { + "epoch": 0.61, + "grad_norm": 1.5121161318469527, + "learning_rate": 7.035752607183161e-07, + "loss": 0.221, + "step": 1073 + }, + { + "epoch": 0.61, + "grad_norm": 1.652519439137743, + "learning_rate": 7.01823192104072e-07, + "loss": 0.2123, + "step": 1074 + }, + { + "epoch": 0.61, + "grad_norm": 1.6135109447341516, + "learning_rate": 7.000721275643886e-07, + "loss": 0.2273, + "step": 1075 + }, + { + "epoch": 0.61, + "grad_norm": 1.8711597053368574, + "learning_rate": 6.983220729957657e-07, + "loss": 0.2057, + "step": 1076 + }, + { + "epoch": 0.61, + "grad_norm": 1.476294404990509, + "learning_rate": 6.96573034291301e-07, + "loss": 0.2446, + "step": 1077 + }, + { + "epoch": 0.61, + "grad_norm": 1.5457467872108264, + "learning_rate": 6.948250173406725e-07, + "loss": 0.2457, + "step": 1078 + }, + { + "epoch": 0.61, + "grad_norm": 2.177811234005188, + "learning_rate": 6.930780280301174e-07, + "loss": 0.2354, + "step": 1079 + }, + { + "epoch": 0.61, + "grad_norm": 1.6128935667055158, + "learning_rate": 6.913320722424121e-07, + "loss": 0.2354, + "step": 1080 + }, + { + "epoch": 0.61, + "grad_norm": 1.457936318240009, + "learning_rate": 6.895871558568529e-07, + "loss": 0.2362, + "step": 1081 + }, + { + "epoch": 0.61, + "grad_norm": 1.3755696085436666, + "learning_rate": 6.878432847492359e-07, + "loss": 0.2084, + "step": 1082 + }, + { + "epoch": 0.61, + "grad_norm": 1.4866068452148193, + "learning_rate": 6.86100464791837e-07, + "loss": 0.2328, + "step": 1083 + }, + { + "epoch": 0.61, + "grad_norm": 1.585787474135499, + "learning_rate": 6.843587018533935e-07, + "loss": 0.2314, + "step": 1084 + }, + { + "epoch": 0.61, + "grad_norm": 1.5935623882176857, + "learning_rate": 6.826180017990828e-07, + "loss": 0.2349, + "step": 1085 + }, + { + "epoch": 0.62, + "grad_norm": 1.4064052304132924, + "learning_rate": 6.80878370490503e-07, + "loss": 0.2223, + "step": 1086 + }, + { + "epoch": 0.62, + "grad_norm": 2.468571397246631, + "learning_rate": 6.791398137856534e-07, + "loss": 0.2419, + "step": 1087 + }, + { + "epoch": 0.62, + "grad_norm": 1.4755183333658604, + "learning_rate": 6.774023375389144e-07, + "loss": 0.235, + "step": 1088 + }, + { + "epoch": 0.62, + "grad_norm": 1.35928939559579, + "learning_rate": 6.756659476010287e-07, + "loss": 0.2201, + "step": 1089 + }, + { + "epoch": 0.62, + "grad_norm": 1.6834149476099882, + "learning_rate": 6.739306498190809e-07, + "loss": 0.2412, + "step": 1090 + }, + { + "epoch": 0.62, + "grad_norm": 1.4310228636230502, + "learning_rate": 6.721964500364774e-07, + "loss": 0.2222, + "step": 1091 + }, + { + "epoch": 0.62, + "grad_norm": 1.5048672726114247, + "learning_rate": 6.704633540929279e-07, + "loss": 0.2271, + "step": 1092 + }, + { + "epoch": 0.62, + "grad_norm": 1.570323338066063, + "learning_rate": 6.687313678244242e-07, + "loss": 0.2109, + "step": 1093 + }, + { + "epoch": 0.62, + "grad_norm": 1.40312702947127, + "learning_rate": 6.670004970632219e-07, + "loss": 0.2293, + "step": 1094 + }, + { + "epoch": 0.62, + "grad_norm": 1.5121961714429724, + "learning_rate": 6.652707476378207e-07, + "loss": 0.2067, + "step": 1095 + }, + { + "epoch": 0.62, + "grad_norm": 2.3497453998784508, + "learning_rate": 6.635421253729436e-07, + "loss": 0.2166, + "step": 1096 + }, + { + "epoch": 0.62, + "grad_norm": 1.5757124943027156, + "learning_rate": 6.618146360895185e-07, + "loss": 0.2517, + "step": 1097 + }, + { + "epoch": 0.62, + "grad_norm": 1.5128125526041354, + "learning_rate": 6.600882856046578e-07, + "loss": 0.2227, + "step": 1098 + }, + { + "epoch": 0.62, + "grad_norm": 1.713335349133301, + "learning_rate": 6.583630797316386e-07, + "loss": 0.2299, + "step": 1099 + }, + { + "epoch": 0.62, + "grad_norm": 2.6952128503762367, + "learning_rate": 6.566390242798854e-07, + "loss": 0.2284, + "step": 1100 + }, + { + "epoch": 0.62, + "grad_norm": 1.759887957023974, + "learning_rate": 6.549161250549474e-07, + "loss": 0.2276, + "step": 1101 + }, + { + "epoch": 0.62, + "grad_norm": 1.5597579088808027, + "learning_rate": 6.531943878584802e-07, + "loss": 0.2319, + "step": 1102 + }, + { + "epoch": 0.62, + "grad_norm": 1.7147084406690423, + "learning_rate": 6.514738184882274e-07, + "loss": 0.2421, + "step": 1103 + }, + { + "epoch": 0.63, + "grad_norm": 1.420289740196203, + "learning_rate": 6.49754422737999e-07, + "loss": 0.2199, + "step": 1104 + }, + { + "epoch": 0.63, + "grad_norm": 2.5682641418236085, + "learning_rate": 6.480362063976539e-07, + "loss": 0.2239, + "step": 1105 + }, + { + "epoch": 0.63, + "grad_norm": 1.166355354214944, + "learning_rate": 6.46319175253079e-07, + "loss": 0.2036, + "step": 1106 + }, + { + "epoch": 0.63, + "grad_norm": 2.2922010368822745, + "learning_rate": 6.446033350861705e-07, + "loss": 0.2316, + "step": 1107 + }, + { + "epoch": 0.63, + "grad_norm": 1.367922481730153, + "learning_rate": 6.428886916748138e-07, + "loss": 0.2255, + "step": 1108 + }, + { + "epoch": 0.63, + "grad_norm": 1.5648994706332437, + "learning_rate": 6.411752507928641e-07, + "loss": 0.2224, + "step": 1109 + }, + { + "epoch": 0.63, + "grad_norm": 1.563867164772982, + "learning_rate": 6.394630182101282e-07, + "loss": 0.2178, + "step": 1110 + }, + { + "epoch": 0.63, + "grad_norm": 3.1466160585123317, + "learning_rate": 6.37751999692343e-07, + "loss": 0.2232, + "step": 1111 + }, + { + "epoch": 0.63, + "grad_norm": 1.4152489727322595, + "learning_rate": 6.360422010011584e-07, + "loss": 0.2154, + "step": 1112 + }, + { + "epoch": 0.63, + "grad_norm": 2.1598393496472807, + "learning_rate": 6.343336278941156e-07, + "loss": 0.2362, + "step": 1113 + }, + { + "epoch": 0.63, + "grad_norm": 1.4775419605349291, + "learning_rate": 6.32626286124629e-07, + "loss": 0.2235, + "step": 1114 + }, + { + "epoch": 0.63, + "grad_norm": 1.513547608872155, + "learning_rate": 6.309201814419667e-07, + "loss": 0.233, + "step": 1115 + }, + { + "epoch": 0.63, + "grad_norm": 1.6168266238771563, + "learning_rate": 6.292153195912315e-07, + "loss": 0.2178, + "step": 1116 + }, + { + "epoch": 0.63, + "grad_norm": 1.4181529704694595, + "learning_rate": 6.275117063133408e-07, + "loss": 0.2013, + "step": 1117 + }, + { + "epoch": 0.63, + "grad_norm": 2.2263079624619504, + "learning_rate": 6.258093473450075e-07, + "loss": 0.2239, + "step": 1118 + }, + { + "epoch": 0.63, + "grad_norm": 1.3411304082064495, + "learning_rate": 6.241082484187205e-07, + "loss": 0.2034, + "step": 1119 + }, + { + "epoch": 0.63, + "grad_norm": 1.5272384990449883, + "learning_rate": 6.224084152627261e-07, + "loss": 0.23, + "step": 1120 + }, + { + "epoch": 0.64, + "grad_norm": 1.8309033209745913, + "learning_rate": 6.207098536010082e-07, + "loss": 0.2604, + "step": 1121 + }, + { + "epoch": 0.64, + "grad_norm": 1.4313488494156839, + "learning_rate": 6.19012569153269e-07, + "loss": 0.2108, + "step": 1122 + }, + { + "epoch": 0.64, + "grad_norm": 1.4817364475701083, + "learning_rate": 6.173165676349102e-07, + "loss": 0.2178, + "step": 1123 + }, + { + "epoch": 0.64, + "grad_norm": 1.395517674787995, + "learning_rate": 6.156218547570129e-07, + "loss": 0.2366, + "step": 1124 + }, + { + "epoch": 0.64, + "grad_norm": 1.3641741417855122, + "learning_rate": 6.139284362263184e-07, + "loss": 0.2312, + "step": 1125 + }, + { + "epoch": 0.64, + "grad_norm": 1.3545941905756942, + "learning_rate": 6.122363177452108e-07, + "loss": 0.2181, + "step": 1126 + }, + { + "epoch": 0.64, + "grad_norm": 1.5648423077309812, + "learning_rate": 6.105455050116953e-07, + "loss": 0.2132, + "step": 1127 + }, + { + "epoch": 0.64, + "grad_norm": 1.3808975395729006, + "learning_rate": 6.08856003719381e-07, + "loss": 0.2169, + "step": 1128 + }, + { + "epoch": 0.64, + "grad_norm": 1.5941017671603206, + "learning_rate": 6.071678195574602e-07, + "loss": 0.2465, + "step": 1129 + }, + { + "epoch": 0.64, + "grad_norm": 1.4698342282853591, + "learning_rate": 6.054809582106899e-07, + "loss": 0.2319, + "step": 1130 + }, + { + "epoch": 0.64, + "grad_norm": 1.6866706358793475, + "learning_rate": 6.037954253593729e-07, + "loss": 0.2344, + "step": 1131 + }, + { + "epoch": 0.64, + "grad_norm": 1.740394826251841, + "learning_rate": 6.021112266793386e-07, + "loss": 0.2246, + "step": 1132 + }, + { + "epoch": 0.64, + "grad_norm": 2.0862530808853323, + "learning_rate": 6.004283678419242e-07, + "loss": 0.2319, + "step": 1133 + }, + { + "epoch": 0.64, + "grad_norm": 2.2283773791590544, + "learning_rate": 5.98746854513954e-07, + "loss": 0.2365, + "step": 1134 + }, + { + "epoch": 0.64, + "grad_norm": 1.6664026337890228, + "learning_rate": 5.970666923577221e-07, + "loss": 0.2197, + "step": 1135 + }, + { + "epoch": 0.64, + "grad_norm": 1.6352728608556466, + "learning_rate": 5.953878870309723e-07, + "loss": 0.2205, + "step": 1136 + }, + { + "epoch": 0.64, + "grad_norm": 1.626338805471819, + "learning_rate": 5.937104441868804e-07, + "loss": 0.2492, + "step": 1137 + }, + { + "epoch": 0.64, + "grad_norm": 1.424314940637421, + "learning_rate": 5.920343694740328e-07, + "loss": 0.2086, + "step": 1138 + }, + { + "epoch": 0.65, + "grad_norm": 1.33338948231071, + "learning_rate": 5.903596685364105e-07, + "loss": 0.2167, + "step": 1139 + }, + { + "epoch": 0.65, + "grad_norm": 1.6329612563054363, + "learning_rate": 5.886863470133671e-07, + "loss": 0.2022, + "step": 1140 + }, + { + "epoch": 0.65, + "grad_norm": 6.832540008845106, + "learning_rate": 5.870144105396118e-07, + "loss": 0.2087, + "step": 1141 + }, + { + "epoch": 0.65, + "grad_norm": 2.0760355709561913, + "learning_rate": 5.853438647451894e-07, + "loss": 0.2333, + "step": 1142 + }, + { + "epoch": 0.65, + "grad_norm": 1.3537998190536242, + "learning_rate": 5.836747152554624e-07, + "loss": 0.2179, + "step": 1143 + }, + { + "epoch": 0.65, + "grad_norm": 4.687083539016377, + "learning_rate": 5.820069676910915e-07, + "loss": 0.2338, + "step": 1144 + }, + { + "epoch": 0.65, + "grad_norm": 1.3068485147714688, + "learning_rate": 5.803406276680154e-07, + "loss": 0.2188, + "step": 1145 + }, + { + "epoch": 0.65, + "grad_norm": 1.6638389844293415, + "learning_rate": 5.78675700797434e-07, + "loss": 0.2313, + "step": 1146 + }, + { + "epoch": 0.65, + "grad_norm": 2.6441120604859694, + "learning_rate": 5.770121926857889e-07, + "loss": 0.239, + "step": 1147 + }, + { + "epoch": 0.65, + "grad_norm": 3.2782504767678478, + "learning_rate": 5.753501089347428e-07, + "loss": 0.2342, + "step": 1148 + }, + { + "epoch": 0.65, + "grad_norm": 1.2946495973598886, + "learning_rate": 5.736894551411642e-07, + "loss": 0.2014, + "step": 1149 + }, + { + "epoch": 0.65, + "grad_norm": 3.722993405932872, + "learning_rate": 5.720302368971042e-07, + "loss": 0.2221, + "step": 1150 + }, + { + "epoch": 0.65, + "grad_norm": 2.0257671160328474, + "learning_rate": 5.70372459789781e-07, + "loss": 0.2195, + "step": 1151 + }, + { + "epoch": 0.65, + "grad_norm": 1.24941644260534, + "learning_rate": 5.687161294015602e-07, + "loss": 0.2182, + "step": 1152 + }, + { + "epoch": 0.65, + "grad_norm": 1.3367264730276553, + "learning_rate": 5.670612513099342e-07, + "loss": 0.2069, + "step": 1153 + }, + { + "epoch": 0.65, + "grad_norm": 1.4026624466292554, + "learning_rate": 5.654078310875077e-07, + "loss": 0.2225, + "step": 1154 + }, + { + "epoch": 0.65, + "grad_norm": 1.396926491221855, + "learning_rate": 5.63755874301973e-07, + "loss": 0.2264, + "step": 1155 + }, + { + "epoch": 0.65, + "grad_norm": 1.9523486746262086, + "learning_rate": 5.62105386516097e-07, + "loss": 0.239, + "step": 1156 + }, + { + "epoch": 0.66, + "grad_norm": 1.6276013062192984, + "learning_rate": 5.604563732876988e-07, + "loss": 0.2372, + "step": 1157 + }, + { + "epoch": 0.66, + "grad_norm": 2.3682762607076273, + "learning_rate": 5.588088401696313e-07, + "loss": 0.2289, + "step": 1158 + }, + { + "epoch": 0.66, + "grad_norm": 1.4195496283970481, + "learning_rate": 5.571627927097657e-07, + "loss": 0.2262, + "step": 1159 + }, + { + "epoch": 0.66, + "grad_norm": 1.529248727085344, + "learning_rate": 5.555182364509681e-07, + "loss": 0.2265, + "step": 1160 + }, + { + "epoch": 0.66, + "grad_norm": 1.4439855754879787, + "learning_rate": 5.538751769310841e-07, + "loss": 0.2305, + "step": 1161 + }, + { + "epoch": 0.66, + "grad_norm": 1.4407617350581574, + "learning_rate": 5.522336196829196e-07, + "loss": 0.2367, + "step": 1162 + }, + { + "epoch": 0.66, + "grad_norm": 1.8661485794118338, + "learning_rate": 5.505935702342207e-07, + "loss": 0.2366, + "step": 1163 + }, + { + "epoch": 0.66, + "grad_norm": 3.2815367775934967, + "learning_rate": 5.489550341076569e-07, + "loss": 0.2256, + "step": 1164 + }, + { + "epoch": 0.66, + "grad_norm": 1.4638598260727298, + "learning_rate": 5.47318016820802e-07, + "loss": 0.2185, + "step": 1165 + }, + { + "epoch": 0.66, + "grad_norm": 1.673324071994042, + "learning_rate": 5.456825238861147e-07, + "loss": 0.2303, + "step": 1166 + }, + { + "epoch": 0.66, + "grad_norm": 1.7218049655523888, + "learning_rate": 5.440485608109215e-07, + "loss": 0.2136, + "step": 1167 + }, + { + "epoch": 0.66, + "grad_norm": 1.9650019081673185, + "learning_rate": 5.424161330973955e-07, + "loss": 0.228, + "step": 1168 + }, + { + "epoch": 0.66, + "grad_norm": 1.52377044516524, + "learning_rate": 5.407852462425416e-07, + "loss": 0.2248, + "step": 1169 + }, + { + "epoch": 0.66, + "grad_norm": 1.409218286992997, + "learning_rate": 5.391559057381749e-07, + "loss": 0.2089, + "step": 1170 + }, + { + "epoch": 0.66, + "grad_norm": 1.6950049017508326, + "learning_rate": 5.375281170709045e-07, + "loss": 0.2247, + "step": 1171 + }, + { + "epoch": 0.66, + "grad_norm": 1.359659268801696, + "learning_rate": 5.35901885722112e-07, + "loss": 0.209, + "step": 1172 + }, + { + "epoch": 0.66, + "grad_norm": 1.5666789243145594, + "learning_rate": 5.342772171679364e-07, + "loss": 0.2105, + "step": 1173 + }, + { + "epoch": 0.67, + "grad_norm": 1.3908105242227826, + "learning_rate": 5.32654116879254e-07, + "loss": 0.2273, + "step": 1174 + }, + { + "epoch": 0.67, + "grad_norm": 1.4908860706296103, + "learning_rate": 5.310325903216597e-07, + "loss": 0.2236, + "step": 1175 + }, + { + "epoch": 0.67, + "grad_norm": 1.7371547382682146, + "learning_rate": 5.294126429554498e-07, + "loss": 0.2316, + "step": 1176 + }, + { + "epoch": 0.67, + "grad_norm": 1.5950336486431171, + "learning_rate": 5.277942802356014e-07, + "loss": 0.219, + "step": 1177 + }, + { + "epoch": 0.67, + "grad_norm": 2.046357711381408, + "learning_rate": 5.261775076117571e-07, + "loss": 0.221, + "step": 1178 + }, + { + "epoch": 0.67, + "grad_norm": 1.9503517050956323, + "learning_rate": 5.245623305282046e-07, + "loss": 0.2217, + "step": 1179 + }, + { + "epoch": 0.67, + "grad_norm": 3.522985848810385, + "learning_rate": 5.229487544238574e-07, + "loss": 0.2178, + "step": 1180 + }, + { + "epoch": 0.67, + "grad_norm": 1.884544728995191, + "learning_rate": 5.213367847322407e-07, + "loss": 0.2213, + "step": 1181 + }, + { + "epoch": 0.67, + "grad_norm": 7.46301025022052, + "learning_rate": 5.197264268814676e-07, + "loss": 0.2194, + "step": 1182 + }, + { + "epoch": 0.67, + "grad_norm": 2.347259822924156, + "learning_rate": 5.181176862942247e-07, + "loss": 0.2061, + "step": 1183 + }, + { + "epoch": 0.67, + "grad_norm": 1.5683885147406846, + "learning_rate": 5.165105683877533e-07, + "loss": 0.2313, + "step": 1184 + }, + { + "epoch": 0.67, + "grad_norm": 1.3885376381655916, + "learning_rate": 5.149050785738284e-07, + "loss": 0.2221, + "step": 1185 + }, + { + "epoch": 0.67, + "grad_norm": 2.4816423921306807, + "learning_rate": 5.133012222587456e-07, + "loss": 0.2143, + "step": 1186 + }, + { + "epoch": 0.67, + "grad_norm": 1.8993455221606683, + "learning_rate": 5.116990048432971e-07, + "loss": 0.2317, + "step": 1187 + }, + { + "epoch": 0.67, + "grad_norm": 2.511055398241898, + "learning_rate": 5.100984317227579e-07, + "loss": 0.2117, + "step": 1188 + }, + { + "epoch": 0.67, + "grad_norm": 2.550305063894376, + "learning_rate": 5.084995082868657e-07, + "loss": 0.2245, + "step": 1189 + }, + { + "epoch": 0.67, + "grad_norm": 1.7236785847196716, + "learning_rate": 5.069022399198023e-07, + "loss": 0.2345, + "step": 1190 + }, + { + "epoch": 0.67, + "grad_norm": 1.3555666452285893, + "learning_rate": 5.053066320001782e-07, + "loss": 0.2271, + "step": 1191 + }, + { + "epoch": 0.68, + "grad_norm": 1.3577710477540295, + "learning_rate": 5.037126899010101e-07, + "loss": 0.2245, + "step": 1192 + }, + { + "epoch": 0.68, + "grad_norm": 1.546410690226597, + "learning_rate": 5.021204189897069e-07, + "loss": 0.21, + "step": 1193 + }, + { + "epoch": 0.68, + "grad_norm": 2.3982277509405088, + "learning_rate": 5.005298246280499e-07, + "loss": 0.2382, + "step": 1194 + }, + { + "epoch": 0.68, + "grad_norm": 6.281488171799301, + "learning_rate": 4.989409121721737e-07, + "loss": 0.2182, + "step": 1195 + }, + { + "epoch": 0.68, + "grad_norm": 2.036540896654236, + "learning_rate": 4.973536869725502e-07, + "loss": 0.2143, + "step": 1196 + }, + { + "epoch": 0.68, + "grad_norm": 1.4514477877644385, + "learning_rate": 4.957681543739697e-07, + "loss": 0.2163, + "step": 1197 + }, + { + "epoch": 0.68, + "grad_norm": 1.5410824657425053, + "learning_rate": 4.941843197155226e-07, + "loss": 0.2207, + "step": 1198 + }, + { + "epoch": 0.68, + "grad_norm": 4.389094217237689, + "learning_rate": 4.926021883305821e-07, + "loss": 0.2204, + "step": 1199 + }, + { + "epoch": 0.68, + "grad_norm": 1.4507992368856149, + "learning_rate": 4.910217655467847e-07, + "loss": 0.2302, + "step": 1200 + }, + { + "epoch": 0.68, + "grad_norm": 1.3697716147957977, + "learning_rate": 4.894430566860144e-07, + "loss": 0.2346, + "step": 1201 + }, + { + "epoch": 0.68, + "grad_norm": 1.9630217219063086, + "learning_rate": 4.878660670643838e-07, + "loss": 0.2315, + "step": 1202 + }, + { + "epoch": 0.68, + "grad_norm": 1.5843650771906488, + "learning_rate": 4.862908019922162e-07, + "loss": 0.2365, + "step": 1203 + }, + { + "epoch": 0.68, + "grad_norm": 2.537795064927633, + "learning_rate": 4.847172667740265e-07, + "loss": 0.2279, + "step": 1204 + }, + { + "epoch": 0.68, + "grad_norm": 1.6109790817380572, + "learning_rate": 4.831454667085058e-07, + "loss": 0.2079, + "step": 1205 + }, + { + "epoch": 0.68, + "grad_norm": 1.8990032857183288, + "learning_rate": 4.815754070885021e-07, + "loss": 0.2311, + "step": 1206 + }, + { + "epoch": 0.68, + "grad_norm": 2.0517199148101826, + "learning_rate": 4.800070932010026e-07, + "loss": 0.2134, + "step": 1207 + }, + { + "epoch": 0.68, + "grad_norm": 1.3100727721906615, + "learning_rate": 4.784405303271157e-07, + "loss": 0.2161, + "step": 1208 + }, + { + "epoch": 0.68, + "grad_norm": 3.067008789059033, + "learning_rate": 4.768757237420529e-07, + "loss": 0.2137, + "step": 1209 + }, + { + "epoch": 0.69, + "grad_norm": 1.3890297214797829, + "learning_rate": 4.753126787151129e-07, + "loss": 0.2239, + "step": 1210 + }, + { + "epoch": 0.69, + "grad_norm": 1.595204038268438, + "learning_rate": 4.7375140050966233e-07, + "loss": 0.2219, + "step": 1211 + }, + { + "epoch": 0.69, + "grad_norm": 1.7024497550602162, + "learning_rate": 4.721918943831167e-07, + "loss": 0.2363, + "step": 1212 + }, + { + "epoch": 0.69, + "grad_norm": 1.668263742777333, + "learning_rate": 4.706341655869269e-07, + "loss": 0.227, + "step": 1213 + }, + { + "epoch": 0.69, + "grad_norm": 2.570932078665602, + "learning_rate": 4.6907821936655633e-07, + "loss": 0.2155, + "step": 1214 + }, + { + "epoch": 0.69, + "grad_norm": 1.378026902775591, + "learning_rate": 4.6752406096146726e-07, + "loss": 0.2104, + "step": 1215 + }, + { + "epoch": 0.69, + "grad_norm": 1.5577785554961692, + "learning_rate": 4.6597169560510164e-07, + "loss": 0.2169, + "step": 1216 + }, + { + "epoch": 0.69, + "grad_norm": 2.3171743000445573, + "learning_rate": 4.6442112852486215e-07, + "loss": 0.2334, + "step": 1217 + }, + { + "epoch": 0.69, + "grad_norm": 1.3378837470851341, + "learning_rate": 4.628723649420985e-07, + "loss": 0.2339, + "step": 1218 + }, + { + "epoch": 0.69, + "grad_norm": 1.7773888353283567, + "learning_rate": 4.613254100720848e-07, + "loss": 0.2275, + "step": 1219 + }, + { + "epoch": 0.69, + "grad_norm": 15.874914022818503, + "learning_rate": 4.59780269124006e-07, + "loss": 0.206, + "step": 1220 + }, + { + "epoch": 0.69, + "grad_norm": 1.512934454573531, + "learning_rate": 4.58236947300939e-07, + "loss": 0.2443, + "step": 1221 + }, + { + "epoch": 0.69, + "grad_norm": 2.9837010437606826, + "learning_rate": 4.566954497998332e-07, + "loss": 0.2332, + "step": 1222 + }, + { + "epoch": 0.69, + "grad_norm": 1.3785349746856426, + "learning_rate": 4.5515578181149796e-07, + "loss": 0.2196, + "step": 1223 + }, + { + "epoch": 0.69, + "grad_norm": 1.314969917670223, + "learning_rate": 4.5361794852057875e-07, + "loss": 0.2254, + "step": 1224 + }, + { + "epoch": 0.69, + "grad_norm": 1.6708570767093758, + "learning_rate": 4.520819551055448e-07, + "loss": 0.213, + "step": 1225 + }, + { + "epoch": 0.69, + "grad_norm": 1.6232988842474056, + "learning_rate": 4.5054780673866975e-07, + "loss": 0.2152, + "step": 1226 + }, + { + "epoch": 0.7, + "grad_norm": 1.354337176615839, + "learning_rate": 4.49015508586013e-07, + "loss": 0.2175, + "step": 1227 + }, + { + "epoch": 0.7, + "grad_norm": 1.6175327187009665, + "learning_rate": 4.4748506580740487e-07, + "loss": 0.2099, + "step": 1228 + }, + { + "epoch": 0.7, + "grad_norm": 1.3491166944357529, + "learning_rate": 4.4595648355642744e-07, + "loss": 0.2159, + "step": 1229 + }, + { + "epoch": 0.7, + "grad_norm": 1.458814211080528, + "learning_rate": 4.4442976698039803e-07, + "loss": 0.2134, + "step": 1230 + }, + { + "epoch": 0.7, + "grad_norm": 1.736018860236866, + "learning_rate": 4.429049212203505e-07, + "loss": 0.2099, + "step": 1231 + }, + { + "epoch": 0.7, + "grad_norm": 1.689084747877659, + "learning_rate": 4.4138195141101997e-07, + "loss": 0.2332, + "step": 1232 + }, + { + "epoch": 0.7, + "grad_norm": 1.6286295946765985, + "learning_rate": 4.3986086268082425e-07, + "loss": 0.2229, + "step": 1233 + }, + { + "epoch": 0.7, + "grad_norm": 1.6117358741099328, + "learning_rate": 4.3834166015184694e-07, + "loss": 0.2188, + "step": 1234 + }, + { + "epoch": 0.7, + "grad_norm": 1.5077125897429717, + "learning_rate": 4.368243489398201e-07, + "loss": 0.237, + "step": 1235 + }, + { + "epoch": 0.7, + "grad_norm": 2.0315341030447036, + "learning_rate": 4.353089341541063e-07, + "loss": 0.2312, + "step": 1236 + }, + { + "epoch": 0.7, + "grad_norm": 1.2953521249304156, + "learning_rate": 4.3379542089768297e-07, + "loss": 0.2134, + "step": 1237 + }, + { + "epoch": 0.7, + "grad_norm": 1.3556970135757103, + "learning_rate": 4.3228381426712446e-07, + "loss": 0.2144, + "step": 1238 + }, + { + "epoch": 0.7, + "grad_norm": 1.6432642292254975, + "learning_rate": 4.3077411935258334e-07, + "loss": 0.2202, + "step": 1239 + }, + { + "epoch": 0.7, + "grad_norm": 1.9378193184288537, + "learning_rate": 4.292663412377772e-07, + "loss": 0.2303, + "step": 1240 + }, + { + "epoch": 0.7, + "grad_norm": 1.7000026387272886, + "learning_rate": 4.277604849999665e-07, + "loss": 0.2306, + "step": 1241 + }, + { + "epoch": 0.7, + "grad_norm": 1.7462537182904316, + "learning_rate": 4.2625655570994157e-07, + "loss": 0.2229, + "step": 1242 + }, + { + "epoch": 0.7, + "grad_norm": 1.9092802916569482, + "learning_rate": 4.247545584320038e-07, + "loss": 0.2155, + "step": 1243 + }, + { + "epoch": 0.7, + "grad_norm": 1.46405012692193, + "learning_rate": 4.2325449822394756e-07, + "loss": 0.2275, + "step": 1244 + }, + { + "epoch": 0.71, + "grad_norm": 1.302197037748767, + "learning_rate": 4.217563801370465e-07, + "loss": 0.2288, + "step": 1245 + }, + { + "epoch": 0.71, + "grad_norm": 1.7865081171511175, + "learning_rate": 4.2026020921603236e-07, + "loss": 0.2244, + "step": 1246 + }, + { + "epoch": 0.71, + "grad_norm": 2.3058085864911035, + "learning_rate": 4.18765990499081e-07, + "loss": 0.2301, + "step": 1247 + }, + { + "epoch": 0.71, + "grad_norm": 1.3570842705246915, + "learning_rate": 4.172737290177949e-07, + "loss": 0.2094, + "step": 1248 + }, + { + "epoch": 0.71, + "grad_norm": 1.7732459688578588, + "learning_rate": 4.157834297971842e-07, + "loss": 0.2183, + "step": 1249 + }, + { + "epoch": 0.71, + "grad_norm": 1.4540277412998732, + "learning_rate": 4.142950978556536e-07, + "loss": 0.2299, + "step": 1250 + }, + { + "epoch": 0.71, + "grad_norm": 2.162695728090071, + "learning_rate": 4.128087382049812e-07, + "loss": 0.2261, + "step": 1251 + }, + { + "epoch": 0.71, + "grad_norm": 1.976688314607081, + "learning_rate": 4.1132435585030453e-07, + "loss": 0.2309, + "step": 1252 + }, + { + "epoch": 0.71, + "grad_norm": 2.9366706626395893, + "learning_rate": 4.098419557901035e-07, + "loss": 0.2171, + "step": 1253 + }, + { + "epoch": 0.71, + "grad_norm": 1.3386529647337344, + "learning_rate": 4.083615430161811e-07, + "loss": 0.2108, + "step": 1254 + }, + { + "epoch": 0.71, + "grad_norm": 1.8890675113546014, + "learning_rate": 4.068831225136499e-07, + "loss": 0.2191, + "step": 1255 + }, + { + "epoch": 0.71, + "grad_norm": 4.961129311872351, + "learning_rate": 4.054066992609132e-07, + "loss": 0.2222, + "step": 1256 + }, + { + "epoch": 0.71, + "grad_norm": 1.3774748665810876, + "learning_rate": 4.0393227822964884e-07, + "loss": 0.2018, + "step": 1257 + }, + { + "epoch": 0.71, + "grad_norm": 2.2253829844331077, + "learning_rate": 4.024598643847927e-07, + "loss": 0.2428, + "step": 1258 + }, + { + "epoch": 0.71, + "grad_norm": 1.6894913004164593, + "learning_rate": 4.0098946268452105e-07, + "loss": 0.2178, + "step": 1259 + }, + { + "epoch": 0.71, + "grad_norm": 2.3241500149138434, + "learning_rate": 3.9952107808023484e-07, + "loss": 0.2251, + "step": 1260 + }, + { + "epoch": 0.71, + "grad_norm": 1.5739771020045248, + "learning_rate": 3.980547155165429e-07, + "loss": 0.2197, + "step": 1261 + }, + { + "epoch": 0.72, + "grad_norm": 1.2243318124162879, + "learning_rate": 3.965903799312452e-07, + "loss": 0.1949, + "step": 1262 + }, + { + "epoch": 0.72, + "grad_norm": 1.3046382940461914, + "learning_rate": 3.951280762553152e-07, + "loss": 0.2125, + "step": 1263 + }, + { + "epoch": 0.72, + "grad_norm": 1.685594992621803, + "learning_rate": 3.9366780941288515e-07, + "loss": 0.2225, + "step": 1264 + }, + { + "epoch": 0.72, + "grad_norm": 1.489390754103372, + "learning_rate": 3.9220958432122784e-07, + "loss": 0.2219, + "step": 1265 + }, + { + "epoch": 0.72, + "grad_norm": 1.5406503579629987, + "learning_rate": 3.907534058907414e-07, + "loss": 0.2098, + "step": 1266 + }, + { + "epoch": 0.72, + "grad_norm": 1.5834292214413508, + "learning_rate": 3.892992790249318e-07, + "loss": 0.2198, + "step": 1267 + }, + { + "epoch": 0.72, + "grad_norm": 1.2282394098910956, + "learning_rate": 3.878472086203957e-07, + "loss": 0.2069, + "step": 1268 + }, + { + "epoch": 0.72, + "grad_norm": 1.8149127926826298, + "learning_rate": 3.8639719956680615e-07, + "loss": 0.224, + "step": 1269 + }, + { + "epoch": 0.72, + "grad_norm": 2.0691258551332306, + "learning_rate": 3.849492567468947e-07, + "loss": 0.2335, + "step": 1270 + }, + { + "epoch": 0.72, + "grad_norm": 1.802155886357632, + "learning_rate": 3.8350338503643365e-07, + "loss": 0.2208, + "step": 1271 + }, + { + "epoch": 0.72, + "grad_norm": 1.3973794748914075, + "learning_rate": 3.8205958930422354e-07, + "loss": 0.2105, + "step": 1272 + }, + { + "epoch": 0.72, + "grad_norm": 1.5406436205361762, + "learning_rate": 3.8061787441207205e-07, + "loss": 0.2251, + "step": 1273 + }, + { + "epoch": 0.72, + "grad_norm": 1.610586845272582, + "learning_rate": 3.7917824521478106e-07, + "loss": 0.2148, + "step": 1274 + }, + { + "epoch": 0.72, + "grad_norm": 1.54345070390649, + "learning_rate": 3.7774070656012914e-07, + "loss": 0.2305, + "step": 1275 + }, + { + "epoch": 0.72, + "grad_norm": 1.631455569394719, + "learning_rate": 3.7630526328885405e-07, + "loss": 0.2213, + "step": 1276 + }, + { + "epoch": 0.72, + "grad_norm": 1.6886533745767076, + "learning_rate": 3.7487192023463966e-07, + "loss": 0.2493, + "step": 1277 + }, + { + "epoch": 0.72, + "grad_norm": 1.386118273870104, + "learning_rate": 3.734406822240955e-07, + "loss": 0.2012, + "step": 1278 + }, + { + "epoch": 0.72, + "grad_norm": 1.9432375355625244, + "learning_rate": 3.72011554076744e-07, + "loss": 0.2192, + "step": 1279 + }, + { + "epoch": 0.73, + "grad_norm": 24.323569770717064, + "learning_rate": 3.705845406050025e-07, + "loss": 0.2187, + "step": 1280 + }, + { + "epoch": 0.73, + "grad_norm": 1.359817486363084, + "learning_rate": 3.6915964661416653e-07, + "loss": 0.2335, + "step": 1281 + }, + { + "epoch": 0.73, + "grad_norm": 1.607039028098902, + "learning_rate": 3.6773687690239664e-07, + "loss": 0.2309, + "step": 1282 + }, + { + "epoch": 0.73, + "grad_norm": 1.3879806485178774, + "learning_rate": 3.6631623626069775e-07, + "loss": 0.2236, + "step": 1283 + }, + { + "epoch": 0.73, + "grad_norm": 1.6108210826221387, + "learning_rate": 3.6489772947290677e-07, + "loss": 0.2417, + "step": 1284 + }, + { + "epoch": 0.73, + "grad_norm": 1.61044372017858, + "learning_rate": 3.634813613156753e-07, + "loss": 0.2143, + "step": 1285 + }, + { + "epoch": 0.73, + "grad_norm": 1.5365326047391896, + "learning_rate": 3.620671365584521e-07, + "loss": 0.2104, + "step": 1286 + }, + { + "epoch": 0.73, + "grad_norm": 1.7150186417367956, + "learning_rate": 3.6065505996346946e-07, + "loss": 0.2385, + "step": 1287 + }, + { + "epoch": 0.73, + "grad_norm": 1.3115863367726706, + "learning_rate": 3.592451362857254e-07, + "loss": 0.2019, + "step": 1288 + }, + { + "epoch": 0.73, + "grad_norm": 1.9515643392293547, + "learning_rate": 3.5783737027296867e-07, + "loss": 0.2143, + "step": 1289 + }, + { + "epoch": 0.73, + "grad_norm": 1.9308591158303714, + "learning_rate": 3.564317666656823e-07, + "loss": 0.221, + "step": 1290 + }, + { + "epoch": 0.73, + "grad_norm": 1.402621361067043, + "learning_rate": 3.5502833019706693e-07, + "loss": 0.2325, + "step": 1291 + }, + { + "epoch": 0.73, + "grad_norm": 4.819937826549352, + "learning_rate": 3.5362706559302637e-07, + "loss": 0.2163, + "step": 1292 + }, + { + "epoch": 0.73, + "grad_norm": 1.4735014423494779, + "learning_rate": 3.5222797757215084e-07, + "loss": 0.2233, + "step": 1293 + }, + { + "epoch": 0.73, + "grad_norm": 3.367532274385393, + "learning_rate": 3.508310708457014e-07, + "loss": 0.2195, + "step": 1294 + }, + { + "epoch": 0.73, + "grad_norm": 2.3294914392790806, + "learning_rate": 3.494363501175929e-07, + "loss": 0.2268, + "step": 1295 + }, + { + "epoch": 0.73, + "grad_norm": 1.6103699509178353, + "learning_rate": 3.4804382008438007e-07, + "loss": 0.2249, + "step": 1296 + }, + { + "epoch": 0.73, + "grad_norm": 1.492243335760796, + "learning_rate": 3.466534854352403e-07, + "loss": 0.2118, + "step": 1297 + }, + { + "epoch": 0.74, + "grad_norm": 1.41292848776241, + "learning_rate": 3.4526535085195875e-07, + "loss": 0.2341, + "step": 1298 + }, + { + "epoch": 0.74, + "grad_norm": 7.457498337249308, + "learning_rate": 3.438794210089118e-07, + "loss": 0.2145, + "step": 1299 + }, + { + "epoch": 0.74, + "grad_norm": 1.7891132672143955, + "learning_rate": 3.42495700573051e-07, + "loss": 0.211, + "step": 1300 + }, + { + "epoch": 0.74, + "grad_norm": 2.4059450428501257, + "learning_rate": 3.4111419420388897e-07, + "loss": 0.2149, + "step": 1301 + }, + { + "epoch": 0.74, + "grad_norm": 1.3863411711735751, + "learning_rate": 3.397349065534827e-07, + "loss": 0.2058, + "step": 1302 + }, + { + "epoch": 0.74, + "grad_norm": 1.8369519446042217, + "learning_rate": 3.383578422664165e-07, + "loss": 0.2025, + "step": 1303 + }, + { + "epoch": 0.74, + "grad_norm": 2.045773560195026, + "learning_rate": 3.3698300597979037e-07, + "loss": 0.2185, + "step": 1304 + }, + { + "epoch": 0.74, + "grad_norm": 1.225262648313144, + "learning_rate": 3.3561040232319903e-07, + "loss": 0.1992, + "step": 1305 + }, + { + "epoch": 0.74, + "grad_norm": 1.3731561680860072, + "learning_rate": 3.3424003591872076e-07, + "loss": 0.2136, + "step": 1306 + }, + { + "epoch": 0.74, + "grad_norm": 1.521312662293047, + "learning_rate": 3.328719113809e-07, + "loss": 0.2215, + "step": 1307 + }, + { + "epoch": 0.74, + "grad_norm": 1.4656424845063694, + "learning_rate": 3.3150603331673065e-07, + "loss": 0.2181, + "step": 1308 + }, + { + "epoch": 0.74, + "grad_norm": 1.377716940544608, + "learning_rate": 3.3014240632564437e-07, + "loss": 0.217, + "step": 1309 + }, + { + "epoch": 0.74, + "grad_norm": 1.6089864415278838, + "learning_rate": 3.2878103499949004e-07, + "loss": 0.2298, + "step": 1310 + }, + { + "epoch": 0.74, + "grad_norm": 2.1562191422584784, + "learning_rate": 3.274219239225224e-07, + "loss": 0.2438, + "step": 1311 + }, + { + "epoch": 0.74, + "grad_norm": 2.3708410554376447, + "learning_rate": 3.2606507767138493e-07, + "loss": 0.223, + "step": 1312 + }, + { + "epoch": 0.74, + "grad_norm": 1.4240711643308386, + "learning_rate": 3.247105008150933e-07, + "loss": 0.2273, + "step": 1313 + }, + { + "epoch": 0.74, + "grad_norm": 1.6365621903401992, + "learning_rate": 3.233581979150235e-07, + "loss": 0.22, + "step": 1314 + }, + { + "epoch": 0.75, + "grad_norm": 1.5892334945254245, + "learning_rate": 3.220081735248922e-07, + "loss": 0.2169, + "step": 1315 + }, + { + "epoch": 0.75, + "grad_norm": 1.4984581770169263, + "learning_rate": 3.2066043219074435e-07, + "loss": 0.228, + "step": 1316 + }, + { + "epoch": 0.75, + "grad_norm": 1.3503949759265619, + "learning_rate": 3.1931497845093744e-07, + "loss": 0.2162, + "step": 1317 + }, + { + "epoch": 0.75, + "grad_norm": 11.644479303342717, + "learning_rate": 3.179718168361244e-07, + "loss": 0.2263, + "step": 1318 + }, + { + "epoch": 0.75, + "grad_norm": 1.4075713069004865, + "learning_rate": 3.1663095186924105e-07, + "loss": 0.216, + "step": 1319 + }, + { + "epoch": 0.75, + "grad_norm": 1.3834697556688493, + "learning_rate": 3.152923880654889e-07, + "loss": 0.2165, + "step": 1320 + }, + { + "epoch": 0.75, + "grad_norm": 1.5294091218866657, + "learning_rate": 3.1395612993232056e-07, + "loss": 0.2399, + "step": 1321 + }, + { + "epoch": 0.75, + "grad_norm": 1.5171049123519278, + "learning_rate": 3.1262218196942516e-07, + "loss": 0.2291, + "step": 1322 + }, + { + "epoch": 0.75, + "grad_norm": 2.094352144480436, + "learning_rate": 3.112905486687113e-07, + "loss": 0.2216, + "step": 1323 + }, + { + "epoch": 0.75, + "grad_norm": 1.4165249564285707, + "learning_rate": 3.0996123451429456e-07, + "loss": 0.2248, + "step": 1324 + }, + { + "epoch": 0.75, + "grad_norm": 2.2765359975748107, + "learning_rate": 3.0863424398248063e-07, + "loss": 0.2142, + "step": 1325 + }, + { + "epoch": 0.75, + "grad_norm": 1.33658333156668, + "learning_rate": 3.0730958154175066e-07, + "loss": 0.2087, + "step": 1326 + }, + { + "epoch": 0.75, + "grad_norm": 1.3622346436926145, + "learning_rate": 3.0598725165274597e-07, + "loss": 0.2179, + "step": 1327 + }, + { + "epoch": 0.75, + "grad_norm": 1.7357039066317481, + "learning_rate": 3.0466725876825337e-07, + "loss": 0.2036, + "step": 1328 + }, + { + "epoch": 0.75, + "grad_norm": 1.4566385045340606, + "learning_rate": 3.033496073331907e-07, + "loss": 0.2126, + "step": 1329 + }, + { + "epoch": 0.75, + "grad_norm": 1.5542517729429277, + "learning_rate": 3.020343017845904e-07, + "loss": 0.2444, + "step": 1330 + }, + { + "epoch": 0.75, + "grad_norm": 1.2664646588676471, + "learning_rate": 3.0072134655158645e-07, + "loss": 0.2173, + "step": 1331 + }, + { + "epoch": 0.75, + "grad_norm": 1.7150577332787311, + "learning_rate": 2.9941074605539687e-07, + "loss": 0.2179, + "step": 1332 + }, + { + "epoch": 0.76, + "grad_norm": 1.7337931127519945, + "learning_rate": 2.981025047093118e-07, + "loss": 0.2081, + "step": 1333 + }, + { + "epoch": 0.76, + "grad_norm": 2.1863626010385055, + "learning_rate": 2.96796626918677e-07, + "loss": 0.2133, + "step": 1334 + }, + { + "epoch": 0.76, + "grad_norm": 2.5940369827106675, + "learning_rate": 2.9549311708087797e-07, + "loss": 0.2264, + "step": 1335 + }, + { + "epoch": 0.76, + "grad_norm": 1.4939416035487627, + "learning_rate": 2.941919795853288e-07, + "loss": 0.2046, + "step": 1336 + }, + { + "epoch": 0.76, + "grad_norm": 1.740101885989287, + "learning_rate": 2.9289321881345254e-07, + "loss": 0.2221, + "step": 1337 + }, + { + "epoch": 0.76, + "grad_norm": 1.3124732724001025, + "learning_rate": 2.9159683913867026e-07, + "loss": 0.2237, + "step": 1338 + }, + { + "epoch": 0.76, + "grad_norm": 1.3636205470227236, + "learning_rate": 2.9030284492638526e-07, + "loss": 0.2115, + "step": 1339 + }, + { + "epoch": 0.76, + "grad_norm": 2.5856210819927727, + "learning_rate": 2.890112405339663e-07, + "loss": 0.2034, + "step": 1340 + }, + { + "epoch": 0.76, + "grad_norm": 1.6908836198750048, + "learning_rate": 2.8772203031073726e-07, + "loss": 0.2273, + "step": 1341 + }, + { + "epoch": 0.76, + "grad_norm": 1.5062993807299572, + "learning_rate": 2.864352185979578e-07, + "loss": 0.2183, + "step": 1342 + }, + { + "epoch": 0.76, + "grad_norm": 2.593853294463133, + "learning_rate": 2.851508097288119e-07, + "loss": 0.2499, + "step": 1343 + }, + { + "epoch": 0.76, + "grad_norm": 1.6276280910997505, + "learning_rate": 2.838688080283923e-07, + "loss": 0.2229, + "step": 1344 + }, + { + "epoch": 0.76, + "grad_norm": 1.6152432265276269, + "learning_rate": 2.8258921781368506e-07, + "loss": 0.2357, + "step": 1345 + }, + { + "epoch": 0.76, + "grad_norm": 1.417173802701744, + "learning_rate": 2.8131204339355673e-07, + "loss": 0.227, + "step": 1346 + }, + { + "epoch": 0.76, + "grad_norm": 1.6341536445289244, + "learning_rate": 2.800372890687388e-07, + "loss": 0.2243, + "step": 1347 + }, + { + "epoch": 0.76, + "grad_norm": 1.6343794151751727, + "learning_rate": 2.7876495913181297e-07, + "loss": 0.2311, + "step": 1348 + }, + { + "epoch": 0.76, + "grad_norm": 1.338481324304029, + "learning_rate": 2.7749505786719795e-07, + "loss": 0.2217, + "step": 1349 + }, + { + "epoch": 0.76, + "grad_norm": 1.416555808657952, + "learning_rate": 2.762275895511328e-07, + "loss": 0.2069, + "step": 1350 + }, + { + "epoch": 0.77, + "grad_norm": 1.5936125567441186, + "learning_rate": 2.7496255845166506e-07, + "loss": 0.221, + "step": 1351 + }, + { + "epoch": 0.77, + "grad_norm": 1.405637431706501, + "learning_rate": 2.736999688286351e-07, + "loss": 0.2033, + "step": 1352 + }, + { + "epoch": 0.77, + "grad_norm": 1.5387048546528532, + "learning_rate": 2.724398249336618e-07, + "loss": 0.2475, + "step": 1353 + }, + { + "epoch": 0.77, + "grad_norm": 1.383672015658528, + "learning_rate": 2.711821310101277e-07, + "loss": 0.2048, + "step": 1354 + }, + { + "epoch": 0.77, + "grad_norm": 1.454167406282567, + "learning_rate": 2.6992689129316624e-07, + "loss": 0.2264, + "step": 1355 + }, + { + "epoch": 0.77, + "grad_norm": 1.595225568693269, + "learning_rate": 2.6867411000964634e-07, + "loss": 0.2374, + "step": 1356 + }, + { + "epoch": 0.77, + "grad_norm": 1.725134382955814, + "learning_rate": 2.674237913781583e-07, + "loss": 0.2162, + "step": 1357 + }, + { + "epoch": 0.77, + "grad_norm": 1.4296092193612133, + "learning_rate": 2.6617593960900034e-07, + "loss": 0.21, + "step": 1358 + }, + { + "epoch": 0.77, + "grad_norm": 1.584780331169376, + "learning_rate": 2.649305589041624e-07, + "loss": 0.2394, + "step": 1359 + }, + { + "epoch": 0.77, + "grad_norm": 1.4759954009232976, + "learning_rate": 2.636876534573148e-07, + "loss": 0.194, + "step": 1360 + }, + { + "epoch": 0.77, + "grad_norm": 1.4181043191063678, + "learning_rate": 2.624472274537924e-07, + "loss": 0.2036, + "step": 1361 + }, + { + "epoch": 0.77, + "grad_norm": 1.3555138021307074, + "learning_rate": 2.6120928507058e-07, + "loss": 0.202, + "step": 1362 + }, + { + "epoch": 0.77, + "grad_norm": 2.1305070061333784, + "learning_rate": 2.5997383047630085e-07, + "loss": 0.2303, + "step": 1363 + }, + { + "epoch": 0.77, + "grad_norm": 1.577527323088917, + "learning_rate": 2.5874086783119896e-07, + "loss": 0.218, + "step": 1364 + }, + { + "epoch": 0.77, + "grad_norm": 1.5025459244893595, + "learning_rate": 2.57510401287128e-07, + "loss": 0.219, + "step": 1365 + }, + { + "epoch": 0.77, + "grad_norm": 2.6169502722849405, + "learning_rate": 2.5628243498753655e-07, + "loss": 0.2181, + "step": 1366 + }, + { + "epoch": 0.77, + "grad_norm": 1.8083706339088912, + "learning_rate": 2.5505697306745256e-07, + "loss": 0.209, + "step": 1367 + }, + { + "epoch": 0.78, + "grad_norm": 1.5961340940616482, + "learning_rate": 2.5383401965347307e-07, + "loss": 0.2247, + "step": 1368 + }, + { + "epoch": 0.78, + "grad_norm": 2.908178635650959, + "learning_rate": 2.526135788637458e-07, + "loss": 0.2059, + "step": 1369 + }, + { + "epoch": 0.78, + "grad_norm": 1.8111086918295838, + "learning_rate": 2.513956548079585e-07, + "loss": 0.2095, + "step": 1370 + }, + { + "epoch": 0.78, + "grad_norm": 1.6694986165534373, + "learning_rate": 2.501802515873247e-07, + "loss": 0.2176, + "step": 1371 + }, + { + "epoch": 0.78, + "grad_norm": 1.8383764533030202, + "learning_rate": 2.489673732945674e-07, + "loss": 0.2301, + "step": 1372 + }, + { + "epoch": 0.78, + "grad_norm": 1.5101003743018346, + "learning_rate": 2.4775702401390997e-07, + "loss": 0.2357, + "step": 1373 + }, + { + "epoch": 0.78, + "grad_norm": 2.284166190021383, + "learning_rate": 2.465492078210569e-07, + "loss": 0.2071, + "step": 1374 + }, + { + "epoch": 0.78, + "grad_norm": 6.420349501898126, + "learning_rate": 2.453439287831848e-07, + "loss": 0.2179, + "step": 1375 + }, + { + "epoch": 0.78, + "grad_norm": 1.3820586701320954, + "learning_rate": 2.441411909589257e-07, + "loss": 0.2429, + "step": 1376 + }, + { + "epoch": 0.78, + "grad_norm": 1.595073406583804, + "learning_rate": 2.4294099839835456e-07, + "loss": 0.234, + "step": 1377 + }, + { + "epoch": 0.78, + "grad_norm": 1.273128746045125, + "learning_rate": 2.417433551429756e-07, + "loss": 0.2148, + "step": 1378 + }, + { + "epoch": 0.78, + "grad_norm": 1.698788966978208, + "learning_rate": 2.405482652257086e-07, + "loss": 0.2214, + "step": 1379 + }, + { + "epoch": 0.78, + "grad_norm": 1.9997983198229585, + "learning_rate": 2.3935573267087516e-07, + "loss": 0.2188, + "step": 1380 + }, + { + "epoch": 0.78, + "grad_norm": 1.3806714877759103, + "learning_rate": 2.3816576149418578e-07, + "loss": 0.2182, + "step": 1381 + }, + { + "epoch": 0.78, + "grad_norm": 1.4960497394808543, + "learning_rate": 2.369783557027245e-07, + "loss": 0.224, + "step": 1382 + }, + { + "epoch": 0.78, + "grad_norm": 2.145391390070757, + "learning_rate": 2.3579351929493817e-07, + "loss": 0.2078, + "step": 1383 + }, + { + "epoch": 0.78, + "grad_norm": 2.5341691819921377, + "learning_rate": 2.3461125626062083e-07, + "loss": 0.2235, + "step": 1384 + }, + { + "epoch": 0.78, + "grad_norm": 2.388124706850279, + "learning_rate": 2.3343157058090147e-07, + "loss": 0.2209, + "step": 1385 + }, + { + "epoch": 0.79, + "grad_norm": 1.9649767591893454, + "learning_rate": 2.322544662282292e-07, + "loss": 0.2104, + "step": 1386 + }, + { + "epoch": 0.79, + "grad_norm": 1.914375887444368, + "learning_rate": 2.3107994716636182e-07, + "loss": 0.2133, + "step": 1387 + }, + { + "epoch": 0.79, + "grad_norm": 1.156564528037039, + "learning_rate": 2.2990801735035104e-07, + "loss": 0.2016, + "step": 1388 + }, + { + "epoch": 0.79, + "grad_norm": 1.566000949052137, + "learning_rate": 2.287386807265297e-07, + "loss": 0.2333, + "step": 1389 + }, + { + "epoch": 0.79, + "grad_norm": 1.28948641974671, + "learning_rate": 2.2757194123249847e-07, + "loss": 0.2164, + "step": 1390 + }, + { + "epoch": 0.79, + "grad_norm": 1.6901008782730746, + "learning_rate": 2.2640780279711202e-07, + "loss": 0.2204, + "step": 1391 + }, + { + "epoch": 0.79, + "grad_norm": 2.1623902195456437, + "learning_rate": 2.2524626934046664e-07, + "loss": 0.2222, + "step": 1392 + }, + { + "epoch": 0.79, + "grad_norm": 1.854888066509454, + "learning_rate": 2.2408734477388714e-07, + "loss": 0.2186, + "step": 1393 + }, + { + "epoch": 0.79, + "grad_norm": 1.6028515553049631, + "learning_rate": 2.229310329999118e-07, + "loss": 0.22, + "step": 1394 + }, + { + "epoch": 0.79, + "grad_norm": 1.3589399580865063, + "learning_rate": 2.217773379122826e-07, + "loss": 0.2227, + "step": 1395 + }, + { + "epoch": 0.79, + "grad_norm": 2.091830759360672, + "learning_rate": 2.2062626339592817e-07, + "loss": 0.2405, + "step": 1396 + }, + { + "epoch": 0.79, + "grad_norm": 1.5493381461728557, + "learning_rate": 2.1947781332695404e-07, + "loss": 0.2054, + "step": 1397 + }, + { + "epoch": 0.79, + "grad_norm": 1.3285246094589453, + "learning_rate": 2.1833199157262793e-07, + "loss": 0.2092, + "step": 1398 + }, + { + "epoch": 0.79, + "grad_norm": 1.5334976977250527, + "learning_rate": 2.1718880199136614e-07, + "loss": 0.2041, + "step": 1399 + }, + { + "epoch": 0.79, + "grad_norm": 2.3727345667201947, + "learning_rate": 2.1604824843272317e-07, + "loss": 0.2204, + "step": 1400 + }, + { + "epoch": 0.79, + "grad_norm": 1.4003133150084934, + "learning_rate": 2.149103347373753e-07, + "loss": 0.2151, + "step": 1401 + }, + { + "epoch": 0.79, + "grad_norm": 1.4954084415665423, + "learning_rate": 2.1377506473711037e-07, + "loss": 0.1953, + "step": 1402 + }, + { + "epoch": 0.79, + "grad_norm": 1.321318164532216, + "learning_rate": 2.1264244225481397e-07, + "loss": 0.2116, + "step": 1403 + }, + { + "epoch": 0.8, + "grad_norm": 1.4638325004418218, + "learning_rate": 2.1151247110445524e-07, + "loss": 0.2313, + "step": 1404 + }, + { + "epoch": 0.8, + "grad_norm": 1.6390035930382205, + "learning_rate": 2.1038515509107734e-07, + "loss": 0.2179, + "step": 1405 + }, + { + "epoch": 0.8, + "grad_norm": 1.5899837488223392, + "learning_rate": 2.092604980107807e-07, + "loss": 0.1991, + "step": 1406 + }, + { + "epoch": 0.8, + "grad_norm": 1.4731220414411554, + "learning_rate": 2.0813850365071306e-07, + "loss": 0.2132, + "step": 1407 + }, + { + "epoch": 0.8, + "grad_norm": 1.6341457852665624, + "learning_rate": 2.0701917578905593e-07, + "loss": 0.2281, + "step": 1408 + }, + { + "epoch": 0.8, + "grad_norm": 1.852974603582753, + "learning_rate": 2.059025181950107e-07, + "loss": 0.2099, + "step": 1409 + }, + { + "epoch": 0.8, + "grad_norm": 1.4613359041596845, + "learning_rate": 2.0478853462878797e-07, + "loss": 0.202, + "step": 1410 + }, + { + "epoch": 0.8, + "grad_norm": 1.5807084997236294, + "learning_rate": 2.0367722884159334e-07, + "loss": 0.2403, + "step": 1411 + }, + { + "epoch": 0.8, + "grad_norm": 1.5733803723236512, + "learning_rate": 2.0256860457561565e-07, + "loss": 0.2241, + "step": 1412 + }, + { + "epoch": 0.8, + "grad_norm": 1.458587557955409, + "learning_rate": 2.0146266556401403e-07, + "loss": 0.2237, + "step": 1413 + }, + { + "epoch": 0.8, + "grad_norm": 1.3969301733198365, + "learning_rate": 2.0035941553090463e-07, + "loss": 0.2211, + "step": 1414 + }, + { + "epoch": 0.8, + "grad_norm": 1.518107468406572, + "learning_rate": 1.9925885819134968e-07, + "loss": 0.2267, + "step": 1415 + }, + { + "epoch": 0.8, + "grad_norm": 1.5075259083068817, + "learning_rate": 1.9816099725134373e-07, + "loss": 0.213, + "step": 1416 + }, + { + "epoch": 0.8, + "grad_norm": 1.622192640052201, + "learning_rate": 1.9706583640780172e-07, + "loss": 0.2227, + "step": 1417 + }, + { + "epoch": 0.8, + "grad_norm": 1.4025989167170234, + "learning_rate": 1.9597337934854564e-07, + "loss": 0.1987, + "step": 1418 + }, + { + "epoch": 0.8, + "grad_norm": 1.3759385992313333, + "learning_rate": 1.9488362975229365e-07, + "loss": 0.2324, + "step": 1419 + }, + { + "epoch": 0.8, + "grad_norm": 1.5075169151808951, + "learning_rate": 1.9379659128864646e-07, + "loss": 0.2206, + "step": 1420 + }, + { + "epoch": 0.81, + "grad_norm": 1.2774983818513492, + "learning_rate": 1.9271226761807558e-07, + "loss": 0.2075, + "step": 1421 + }, + { + "epoch": 0.81, + "grad_norm": 1.4722434029404698, + "learning_rate": 1.9163066239191073e-07, + "loss": 0.2023, + "step": 1422 + }, + { + "epoch": 0.81, + "grad_norm": 1.6398943361574156, + "learning_rate": 1.9055177925232703e-07, + "loss": 0.2256, + "step": 1423 + }, + { + "epoch": 0.81, + "grad_norm": 2.0525606077723118, + "learning_rate": 1.8947562183233412e-07, + "loss": 0.2323, + "step": 1424 + }, + { + "epoch": 0.81, + "grad_norm": 1.5245276922945057, + "learning_rate": 1.8840219375576293e-07, + "loss": 0.2232, + "step": 1425 + }, + { + "epoch": 0.81, + "grad_norm": 2.4631199898561507, + "learning_rate": 1.8733149863725272e-07, + "loss": 0.2099, + "step": 1426 + }, + { + "epoch": 0.81, + "grad_norm": 1.6216924058052904, + "learning_rate": 1.8626354008224166e-07, + "loss": 0.2067, + "step": 1427 + }, + { + "epoch": 0.81, + "grad_norm": 1.4285023264814132, + "learning_rate": 1.8519832168695115e-07, + "loss": 0.2303, + "step": 1428 + }, + { + "epoch": 0.81, + "grad_norm": 2.079422020022985, + "learning_rate": 1.8413584703837615e-07, + "loss": 0.2243, + "step": 1429 + }, + { + "epoch": 0.81, + "grad_norm": 1.9915274983489897, + "learning_rate": 1.8307611971427273e-07, + "loss": 0.2146, + "step": 1430 + }, + { + "epoch": 0.81, + "grad_norm": 1.5536780420564804, + "learning_rate": 1.820191432831445e-07, + "loss": 0.2303, + "step": 1431 + }, + { + "epoch": 0.81, + "grad_norm": 1.4435851736782992, + "learning_rate": 1.8096492130423347e-07, + "loss": 0.2185, + "step": 1432 + }, + { + "epoch": 0.81, + "grad_norm": 1.5486834182273943, + "learning_rate": 1.7991345732750473e-07, + "loss": 0.2009, + "step": 1433 + }, + { + "epoch": 0.81, + "grad_norm": 2.8321943261182403, + "learning_rate": 1.7886475489363727e-07, + "loss": 0.1974, + "step": 1434 + }, + { + "epoch": 0.81, + "grad_norm": 1.748090603899352, + "learning_rate": 1.778188175340104e-07, + "loss": 0.2169, + "step": 1435 + }, + { + "epoch": 0.81, + "grad_norm": 1.4308200885604518, + "learning_rate": 1.7677564877069205e-07, + "loss": 0.206, + "step": 1436 + }, + { + "epoch": 0.81, + "grad_norm": 1.718387687971597, + "learning_rate": 1.7573525211642838e-07, + "loss": 0.2298, + "step": 1437 + }, + { + "epoch": 0.81, + "grad_norm": 3.6111442668759572, + "learning_rate": 1.7469763107462954e-07, + "loss": 0.225, + "step": 1438 + }, + { + "epoch": 0.82, + "grad_norm": 1.539649012251455, + "learning_rate": 1.7366278913935972e-07, + "loss": 0.2329, + "step": 1439 + }, + { + "epoch": 0.82, + "grad_norm": 1.6922532435508642, + "learning_rate": 1.726307297953251e-07, + "loss": 0.2184, + "step": 1440 + }, + { + "epoch": 0.82, + "grad_norm": 1.86131432698414, + "learning_rate": 1.7160145651786095e-07, + "loss": 0.2148, + "step": 1441 + }, + { + "epoch": 0.82, + "grad_norm": 1.455165212931987, + "learning_rate": 1.7057497277292154e-07, + "loss": 0.2233, + "step": 1442 + }, + { + "epoch": 0.82, + "grad_norm": 1.5145926827302396, + "learning_rate": 1.6955128201706758e-07, + "loss": 0.2303, + "step": 1443 + }, + { + "epoch": 0.82, + "grad_norm": 1.4344517820914653, + "learning_rate": 1.6853038769745465e-07, + "loss": 0.2215, + "step": 1444 + }, + { + "epoch": 0.82, + "grad_norm": 1.580862334032493, + "learning_rate": 1.6751229325182192e-07, + "loss": 0.2181, + "step": 1445 + }, + { + "epoch": 0.82, + "grad_norm": 1.9821739179695839, + "learning_rate": 1.664970021084794e-07, + "loss": 0.214, + "step": 1446 + }, + { + "epoch": 0.82, + "grad_norm": 1.2991392859101234, + "learning_rate": 1.6548451768629846e-07, + "loss": 0.2082, + "step": 1447 + }, + { + "epoch": 0.82, + "grad_norm": 1.6890038304575814, + "learning_rate": 1.644748433946984e-07, + "loss": 0.2171, + "step": 1448 + }, + { + "epoch": 0.82, + "grad_norm": 2.367217472313774, + "learning_rate": 1.6346798263363658e-07, + "loss": 0.2242, + "step": 1449 + }, + { + "epoch": 0.82, + "grad_norm": 1.3648812505589911, + "learning_rate": 1.624639387935951e-07, + "loss": 0.1901, + "step": 1450 + }, + { + "epoch": 0.82, + "grad_norm": 1.4952483754247154, + "learning_rate": 1.6146271525557088e-07, + "loss": 0.2276, + "step": 1451 + }, + { + "epoch": 0.82, + "grad_norm": 1.4178685575043481, + "learning_rate": 1.6046431539106432e-07, + "loss": 0.2012, + "step": 1452 + }, + { + "epoch": 0.82, + "grad_norm": 1.2949349025465102, + "learning_rate": 1.594687425620661e-07, + "loss": 0.1955, + "step": 1453 + }, + { + "epoch": 0.82, + "grad_norm": 1.856867449181767, + "learning_rate": 1.584760001210491e-07, + "loss": 0.2037, + "step": 1454 + }, + { + "epoch": 0.82, + "grad_norm": 1.4135145277557295, + "learning_rate": 1.574860914109535e-07, + "loss": 0.2145, + "step": 1455 + }, + { + "epoch": 0.82, + "grad_norm": 1.5557817511130325, + "learning_rate": 1.5649901976517798e-07, + "loss": 0.2043, + "step": 1456 + }, + { + "epoch": 0.83, + "grad_norm": 1.4223696662590923, + "learning_rate": 1.5551478850756795e-07, + "loss": 0.2175, + "step": 1457 + }, + { + "epoch": 0.83, + "grad_norm": 1.427155908043394, + "learning_rate": 1.545334009524033e-07, + "loss": 0.2245, + "step": 1458 + }, + { + "epoch": 0.83, + "grad_norm": 1.4890863782967008, + "learning_rate": 1.5355486040438949e-07, + "loss": 0.1928, + "step": 1459 + }, + { + "epoch": 0.83, + "grad_norm": 3.0513622448069078, + "learning_rate": 1.525791701586434e-07, + "loss": 0.209, + "step": 1460 + }, + { + "epoch": 0.83, + "grad_norm": 1.3840307569530503, + "learning_rate": 1.5160633350068507e-07, + "loss": 0.215, + "step": 1461 + }, + { + "epoch": 0.83, + "grad_norm": 1.3864285867462616, + "learning_rate": 1.5063635370642514e-07, + "loss": 0.2268, + "step": 1462 + }, + { + "epoch": 0.83, + "grad_norm": 1.4930927047416094, + "learning_rate": 1.4966923404215338e-07, + "loss": 0.2248, + "step": 1463 + }, + { + "epoch": 0.83, + "grad_norm": 1.4797517260115005, + "learning_rate": 1.4870497776452985e-07, + "loss": 0.2043, + "step": 1464 + }, + { + "epoch": 0.83, + "grad_norm": 1.4058790046703842, + "learning_rate": 1.4774358812057097e-07, + "loss": 0.2121, + "step": 1465 + }, + { + "epoch": 0.83, + "grad_norm": 1.6947643302132003, + "learning_rate": 1.4678506834764115e-07, + "loss": 0.231, + "step": 1466 + }, + { + "epoch": 0.83, + "grad_norm": 1.5023016331622012, + "learning_rate": 1.4582942167344058e-07, + "loss": 0.1959, + "step": 1467 + }, + { + "epoch": 0.83, + "grad_norm": 2.1116030454278163, + "learning_rate": 1.4487665131599426e-07, + "loss": 0.2319, + "step": 1468 + }, + { + "epoch": 0.83, + "grad_norm": 2.0786068390334056, + "learning_rate": 1.4392676048364193e-07, + "loss": 0.2113, + "step": 1469 + }, + { + "epoch": 0.83, + "grad_norm": 1.4212292413260825, + "learning_rate": 1.4297975237502701e-07, + "loss": 0.2194, + "step": 1470 + }, + { + "epoch": 0.83, + "grad_norm": 1.8743298150431202, + "learning_rate": 1.4203563017908526e-07, + "loss": 0.1955, + "step": 1471 + }, + { + "epoch": 0.83, + "grad_norm": 2.536748945615104, + "learning_rate": 1.41094397075035e-07, + "loss": 0.2265, + "step": 1472 + }, + { + "epoch": 0.83, + "grad_norm": 1.4422570910104409, + "learning_rate": 1.4015605623236514e-07, + "loss": 0.2172, + "step": 1473 + }, + { + "epoch": 0.84, + "grad_norm": 1.7953844653818516, + "learning_rate": 1.3922061081082593e-07, + "loss": 0.2203, + "step": 1474 + }, + { + "epoch": 0.84, + "grad_norm": 1.6873361576058994, + "learning_rate": 1.3828806396041738e-07, + "loss": 0.242, + "step": 1475 + }, + { + "epoch": 0.84, + "grad_norm": 1.529070658591192, + "learning_rate": 1.373584188213791e-07, + "loss": 0.2162, + "step": 1476 + }, + { + "epoch": 0.84, + "grad_norm": 1.2586192301592476, + "learning_rate": 1.364316785241789e-07, + "loss": 0.2097, + "step": 1477 + }, + { + "epoch": 0.84, + "grad_norm": 1.378659728650969, + "learning_rate": 1.355078461895036e-07, + "loss": 0.2271, + "step": 1478 + }, + { + "epoch": 0.84, + "grad_norm": 1.4864265260292164, + "learning_rate": 1.3458692492824763e-07, + "loss": 0.2066, + "step": 1479 + }, + { + "epoch": 0.84, + "grad_norm": 1.655328222051192, + "learning_rate": 1.3366891784150248e-07, + "loss": 0.2217, + "step": 1480 + }, + { + "epoch": 0.84, + "grad_norm": 2.6894621887133847, + "learning_rate": 1.3275382802054702e-07, + "loss": 0.2333, + "step": 1481 + }, + { + "epoch": 0.84, + "grad_norm": 1.9166050787129756, + "learning_rate": 1.318416585468357e-07, + "loss": 0.2086, + "step": 1482 + }, + { + "epoch": 0.84, + "grad_norm": 1.349943146212279, + "learning_rate": 1.3093241249198994e-07, + "loss": 0.2189, + "step": 1483 + }, + { + "epoch": 0.84, + "grad_norm": 1.626992565546153, + "learning_rate": 1.3002609291778655e-07, + "loss": 0.1957, + "step": 1484 + }, + { + "epoch": 0.84, + "grad_norm": 1.5852964974300052, + "learning_rate": 1.2912270287614734e-07, + "loss": 0.2211, + "step": 1485 + }, + { + "epoch": 0.84, + "grad_norm": 1.5817555093476803, + "learning_rate": 1.282222454091304e-07, + "loss": 0.2264, + "step": 1486 + }, + { + "epoch": 0.84, + "grad_norm": 1.507952778874696, + "learning_rate": 1.2732472354891733e-07, + "loss": 0.2122, + "step": 1487 + }, + { + "epoch": 0.84, + "grad_norm": 5.138776866929934, + "learning_rate": 1.2643014031780552e-07, + "loss": 0.2055, + "step": 1488 + }, + { + "epoch": 0.84, + "grad_norm": 1.4657619914701667, + "learning_rate": 1.2553849872819656e-07, + "loss": 0.2269, + "step": 1489 + }, + { + "epoch": 0.84, + "grad_norm": 1.607291502535302, + "learning_rate": 1.2464980178258567e-07, + "loss": 0.2106, + "step": 1490 + }, + { + "epoch": 0.84, + "grad_norm": 1.5548820815355142, + "learning_rate": 1.2376405247355394e-07, + "loss": 0.2198, + "step": 1491 + }, + { + "epoch": 0.85, + "grad_norm": 1.4594231840120386, + "learning_rate": 1.2288125378375502e-07, + "loss": 0.2293, + "step": 1492 + }, + { + "epoch": 0.85, + "grad_norm": 2.213619329997473, + "learning_rate": 1.220014086859076e-07, + "loss": 0.2362, + "step": 1493 + }, + { + "epoch": 0.85, + "grad_norm": 1.8278349949074923, + "learning_rate": 1.211245201427845e-07, + "loss": 0.2074, + "step": 1494 + }, + { + "epoch": 0.85, + "grad_norm": 1.4326511211468307, + "learning_rate": 1.2025059110720182e-07, + "loss": 0.2218, + "step": 1495 + }, + { + "epoch": 0.85, + "grad_norm": 1.6854068210313338, + "learning_rate": 1.1937962452201157e-07, + "loss": 0.1965, + "step": 1496 + }, + { + "epoch": 0.85, + "grad_norm": 6.419413343828208, + "learning_rate": 1.1851162332008813e-07, + "loss": 0.2239, + "step": 1497 + }, + { + "epoch": 0.85, + "grad_norm": 1.7322940990440954, + "learning_rate": 1.1764659042432157e-07, + "loss": 0.2074, + "step": 1498 + }, + { + "epoch": 0.85, + "grad_norm": 1.2884217915063167, + "learning_rate": 1.1678452874760646e-07, + "loss": 0.2164, + "step": 1499 + }, + { + "epoch": 0.85, + "grad_norm": 3.705299813388445, + "learning_rate": 1.1592544119283132e-07, + "loss": 0.2182, + "step": 1500 + }, + { + "epoch": 0.85, + "grad_norm": 5.584359657901772, + "learning_rate": 1.150693306528706e-07, + "loss": 0.2126, + "step": 1501 + }, + { + "epoch": 0.85, + "grad_norm": 1.3466422208879094, + "learning_rate": 1.1421620001057353e-07, + "loss": 0.2039, + "step": 1502 + }, + { + "epoch": 0.85, + "grad_norm": 1.2628932887910436, + "learning_rate": 1.1336605213875505e-07, + "loss": 0.2015, + "step": 1503 + }, + { + "epoch": 0.85, + "grad_norm": 1.269273740017695, + "learning_rate": 1.125188899001861e-07, + "loss": 0.2032, + "step": 1504 + }, + { + "epoch": 0.85, + "grad_norm": 1.5367900816241185, + "learning_rate": 1.1167471614758328e-07, + "loss": 0.2066, + "step": 1505 + }, + { + "epoch": 0.85, + "grad_norm": 1.485385802795097, + "learning_rate": 1.1083353372360049e-07, + "loss": 0.2208, + "step": 1506 + }, + { + "epoch": 0.85, + "grad_norm": 1.4548976676000265, + "learning_rate": 1.0999534546081846e-07, + "loss": 0.2194, + "step": 1507 + }, + { + "epoch": 0.85, + "grad_norm": 1.8774149353617495, + "learning_rate": 1.0916015418173552e-07, + "loss": 0.2214, + "step": 1508 + }, + { + "epoch": 0.85, + "grad_norm": 1.6497581971504462, + "learning_rate": 1.0832796269875755e-07, + "loss": 0.2281, + "step": 1509 + }, + { + "epoch": 0.86, + "grad_norm": 6.039211088755703, + "learning_rate": 1.0749877381418959e-07, + "loss": 0.2257, + "step": 1510 + }, + { + "epoch": 0.86, + "grad_norm": 1.4069529091058017, + "learning_rate": 1.0667259032022557e-07, + "loss": 0.2029, + "step": 1511 + }, + { + "epoch": 0.86, + "grad_norm": 1.3895974668708968, + "learning_rate": 1.0584941499893895e-07, + "loss": 0.21, + "step": 1512 + }, + { + "epoch": 0.86, + "grad_norm": 1.7871285649094508, + "learning_rate": 1.0502925062227408e-07, + "loss": 0.2117, + "step": 1513 + }, + { + "epoch": 0.86, + "grad_norm": 1.5037381905949265, + "learning_rate": 1.0421209995203528e-07, + "loss": 0.2257, + "step": 1514 + }, + { + "epoch": 0.86, + "grad_norm": 1.3692809132609176, + "learning_rate": 1.033979657398797e-07, + "loss": 0.2076, + "step": 1515 + }, + { + "epoch": 0.86, + "grad_norm": 2.0123278701859806, + "learning_rate": 1.0258685072730632e-07, + "loss": 0.2068, + "step": 1516 + }, + { + "epoch": 0.86, + "grad_norm": 1.6564062413010405, + "learning_rate": 1.0177875764564714e-07, + "loss": 0.2034, + "step": 1517 + }, + { + "epoch": 0.86, + "grad_norm": 1.25949945952813, + "learning_rate": 1.009736892160592e-07, + "loss": 0.2105, + "step": 1518 + }, + { + "epoch": 0.86, + "grad_norm": 2.1639413257146085, + "learning_rate": 1.0017164814951296e-07, + "loss": 0.2181, + "step": 1519 + }, + { + "epoch": 0.86, + "grad_norm": 1.6419162658352906, + "learning_rate": 9.937263714678556e-08, + "loss": 0.2339, + "step": 1520 + }, + { + "epoch": 0.86, + "grad_norm": 2.269329710263512, + "learning_rate": 9.85766588984508e-08, + "loss": 0.2199, + "step": 1521 + }, + { + "epoch": 0.86, + "grad_norm": 1.8557905832860966, + "learning_rate": 9.778371608486913e-08, + "loss": 0.2187, + "step": 1522 + }, + { + "epoch": 0.86, + "grad_norm": 1.4874218094611973, + "learning_rate": 9.69938113761808e-08, + "loss": 0.2129, + "step": 1523 + }, + { + "epoch": 0.86, + "grad_norm": 1.378058107097442, + "learning_rate": 9.620694743229462e-08, + "loss": 0.2057, + "step": 1524 + }, + { + "epoch": 0.86, + "grad_norm": 2.5935259943359603, + "learning_rate": 9.542312690288034e-08, + "loss": 0.2114, + "step": 1525 + }, + { + "epoch": 0.86, + "grad_norm": 1.7135607511606774, + "learning_rate": 9.464235242735974e-08, + "loss": 0.2246, + "step": 1526 + }, + { + "epoch": 0.87, + "grad_norm": 3.110210779535669, + "learning_rate": 9.386462663489636e-08, + "loss": 0.2148, + "step": 1527 + }, + { + "epoch": 0.87, + "grad_norm": 1.616985464905334, + "learning_rate": 9.308995214438908e-08, + "loss": 0.23, + "step": 1528 + }, + { + "epoch": 0.87, + "grad_norm": 1.3234503755083291, + "learning_rate": 9.231833156446056e-08, + "loss": 0.2136, + "step": 1529 + }, + { + "epoch": 0.87, + "grad_norm": 1.366973985422098, + "learning_rate": 9.154976749345056e-08, + "loss": 0.2143, + "step": 1530 + }, + { + "epoch": 0.87, + "grad_norm": 1.653512545268247, + "learning_rate": 9.078426251940641e-08, + "loss": 0.2333, + "step": 1531 + }, + { + "epoch": 0.87, + "grad_norm": 2.0296933082754105, + "learning_rate": 9.002181922007379e-08, + "loss": 0.2279, + "step": 1532 + }, + { + "epoch": 0.87, + "grad_norm": 2.320213778577294, + "learning_rate": 8.926244016288898e-08, + "loss": 0.2125, + "step": 1533 + }, + { + "epoch": 0.87, + "grad_norm": 1.5684473653552025, + "learning_rate": 8.850612790496959e-08, + "loss": 0.2219, + "step": 1534 + }, + { + "epoch": 0.87, + "grad_norm": 1.9502815681077792, + "learning_rate": 8.775288499310641e-08, + "loss": 0.2325, + "step": 1535 + }, + { + "epoch": 0.87, + "grad_norm": 2.2515644587913943, + "learning_rate": 8.700271396375446e-08, + "loss": 0.2228, + "step": 1536 + }, + { + "epoch": 0.87, + "grad_norm": 1.8401552034387312, + "learning_rate": 8.625561734302422e-08, + "loss": 0.2157, + "step": 1537 + }, + { + "epoch": 0.87, + "grad_norm": 2.072755257846267, + "learning_rate": 8.551159764667393e-08, + "loss": 0.2027, + "step": 1538 + }, + { + "epoch": 0.87, + "grad_norm": 1.7727705732292582, + "learning_rate": 8.477065738010058e-08, + "loss": 0.2251, + "step": 1539 + }, + { + "epoch": 0.87, + "grad_norm": 1.2717917638294784, + "learning_rate": 8.403279903833149e-08, + "loss": 0.2239, + "step": 1540 + }, + { + "epoch": 0.87, + "grad_norm": 1.566792344061754, + "learning_rate": 8.329802510601558e-08, + "loss": 0.2252, + "step": 1541 + }, + { + "epoch": 0.87, + "grad_norm": 2.299377149690567, + "learning_rate": 8.256633805741608e-08, + "loss": 0.2061, + "step": 1542 + }, + { + "epoch": 0.87, + "grad_norm": 3.81238619739377, + "learning_rate": 8.1837740356401e-08, + "loss": 0.209, + "step": 1543 + }, + { + "epoch": 0.87, + "grad_norm": 2.551457406437111, + "learning_rate": 8.111223445643568e-08, + "loss": 0.2117, + "step": 1544 + }, + { + "epoch": 0.88, + "grad_norm": 11.320553643271015, + "learning_rate": 8.038982280057415e-08, + "loss": 0.2225, + "step": 1545 + }, + { + "epoch": 0.88, + "grad_norm": 1.4050233940952197, + "learning_rate": 7.967050782145035e-08, + "loss": 0.2245, + "step": 1546 + }, + { + "epoch": 0.88, + "grad_norm": 1.3577032318459614, + "learning_rate": 7.895429194127145e-08, + "loss": 0.2054, + "step": 1547 + }, + { + "epoch": 0.88, + "grad_norm": 1.3193172384460126, + "learning_rate": 7.824117757180837e-08, + "loss": 0.2024, + "step": 1548 + }, + { + "epoch": 0.88, + "grad_norm": 1.3096352867462684, + "learning_rate": 7.753116711438745e-08, + "loss": 0.2157, + "step": 1549 + }, + { + "epoch": 0.88, + "grad_norm": 13.853697446027272, + "learning_rate": 7.682426295988431e-08, + "loss": 0.2047, + "step": 1550 + }, + { + "epoch": 0.88, + "grad_norm": 7.3063952827286265, + "learning_rate": 7.612046748871326e-08, + "loss": 0.211, + "step": 1551 + }, + { + "epoch": 0.88, + "grad_norm": 2.101446056695163, + "learning_rate": 7.541978307082109e-08, + "loss": 0.2187, + "step": 1552 + }, + { + "epoch": 0.88, + "grad_norm": 1.3470814479784097, + "learning_rate": 7.472221206567852e-08, + "loss": 0.2182, + "step": 1553 + }, + { + "epoch": 0.88, + "grad_norm": 1.3565615697673674, + "learning_rate": 7.402775682227157e-08, + "loss": 0.2128, + "step": 1554 + }, + { + "epoch": 0.88, + "grad_norm": 2.4485294758196794, + "learning_rate": 7.333641967909554e-08, + "loss": 0.2072, + "step": 1555 + }, + { + "epoch": 0.88, + "grad_norm": 1.6076167778655899, + "learning_rate": 7.264820296414442e-08, + "loss": 0.2293, + "step": 1556 + }, + { + "epoch": 0.88, + "grad_norm": 1.50390094731836, + "learning_rate": 7.196310899490576e-08, + "loss": 0.22, + "step": 1557 + }, + { + "epoch": 0.88, + "grad_norm": 1.5059929048196588, + "learning_rate": 7.128114007835118e-08, + "loss": 0.2298, + "step": 1558 + }, + { + "epoch": 0.88, + "grad_norm": 1.617337333010674, + "learning_rate": 7.060229851092859e-08, + "loss": 0.2176, + "step": 1559 + }, + { + "epoch": 0.88, + "grad_norm": 3.297107855112889, + "learning_rate": 6.992658657855566e-08, + "loss": 0.2211, + "step": 1560 + }, + { + "epoch": 0.88, + "grad_norm": 1.306340674816211, + "learning_rate": 6.925400655661117e-08, + "loss": 0.2014, + "step": 1561 + }, + { + "epoch": 0.88, + "grad_norm": 1.4010462642851482, + "learning_rate": 6.858456070992736e-08, + "loss": 0.1982, + "step": 1562 + }, + { + "epoch": 0.89, + "grad_norm": 1.6848447961721051, + "learning_rate": 6.791825129278294e-08, + "loss": 0.2197, + "step": 1563 + }, + { + "epoch": 0.89, + "grad_norm": 1.7033200241470479, + "learning_rate": 6.72550805488944e-08, + "loss": 0.2315, + "step": 1564 + }, + { + "epoch": 0.89, + "grad_norm": 1.4240213968688014, + "learning_rate": 6.659505071140958e-08, + "loss": 0.2164, + "step": 1565 + }, + { + "epoch": 0.89, + "grad_norm": 3.518752585986374, + "learning_rate": 6.593816400289986e-08, + "loss": 0.2139, + "step": 1566 + }, + { + "epoch": 0.89, + "grad_norm": 1.6739182297768318, + "learning_rate": 6.528442263535216e-08, + "loss": 0.2367, + "step": 1567 + }, + { + "epoch": 0.89, + "grad_norm": 2.186548787238609, + "learning_rate": 6.463382881016189e-08, + "loss": 0.2095, + "step": 1568 + }, + { + "epoch": 0.89, + "grad_norm": 1.5859753174396514, + "learning_rate": 6.398638471812546e-08, + "loss": 0.2274, + "step": 1569 + }, + { + "epoch": 0.89, + "grad_norm": 1.6009284253899283, + "learning_rate": 6.334209253943324e-08, + "loss": 0.2168, + "step": 1570 + }, + { + "epoch": 0.89, + "grad_norm": 1.5904440610041284, + "learning_rate": 6.270095444366164e-08, + "loss": 0.225, + "step": 1571 + }, + { + "epoch": 0.89, + "grad_norm": 1.4040342118985556, + "learning_rate": 6.206297258976634e-08, + "loss": 0.206, + "step": 1572 + }, + { + "epoch": 0.89, + "grad_norm": 1.3651449232732977, + "learning_rate": 6.142814912607408e-08, + "loss": 0.2122, + "step": 1573 + }, + { + "epoch": 0.89, + "grad_norm": 1.391074971878542, + "learning_rate": 6.079648619027678e-08, + "loss": 0.2224, + "step": 1574 + }, + { + "epoch": 0.89, + "grad_norm": 3.130118844364249, + "learning_rate": 6.016798590942374e-08, + "loss": 0.2268, + "step": 1575 + }, + { + "epoch": 0.89, + "grad_norm": 1.5101119683708426, + "learning_rate": 5.9542650399913576e-08, + "loss": 0.2147, + "step": 1576 + }, + { + "epoch": 0.89, + "grad_norm": 1.5680585813737002, + "learning_rate": 5.892048176748898e-08, + "loss": 0.2162, + "step": 1577 + }, + { + "epoch": 0.89, + "grad_norm": 1.4598061935813698, + "learning_rate": 5.830148210722774e-08, + "loss": 0.2189, + "step": 1578 + }, + { + "epoch": 0.89, + "grad_norm": 2.070201282277784, + "learning_rate": 5.7685653503536824e-08, + "loss": 0.2203, + "step": 1579 + }, + { + "epoch": 0.9, + "grad_norm": 9.594611540609964, + "learning_rate": 5.7072998030145205e-08, + "loss": 0.2255, + "step": 1580 + }, + { + "epoch": 0.9, + "grad_norm": 2.736973133246413, + "learning_rate": 5.646351775009617e-08, + "loss": 0.2146, + "step": 1581 + }, + { + "epoch": 0.9, + "grad_norm": 1.6497677300636338, + "learning_rate": 5.58572147157419e-08, + "loss": 0.1966, + "step": 1582 + }, + { + "epoch": 0.9, + "grad_norm": 1.4634227130137298, + "learning_rate": 5.525409096873446e-08, + "loss": 0.2195, + "step": 1583 + }, + { + "epoch": 0.9, + "grad_norm": 1.505947458006658, + "learning_rate": 5.465414854002071e-08, + "loss": 0.2154, + "step": 1584 + }, + { + "epoch": 0.9, + "grad_norm": 1.570411935526634, + "learning_rate": 5.405738944983496e-08, + "loss": 0.2186, + "step": 1585 + }, + { + "epoch": 0.9, + "grad_norm": 1.8010903311470088, + "learning_rate": 5.346381570769121e-08, + "loss": 0.2205, + "step": 1586 + }, + { + "epoch": 0.9, + "grad_norm": 4.825721403610928, + "learning_rate": 5.287342931237826e-08, + "loss": 0.2115, + "step": 1587 + }, + { + "epoch": 0.9, + "grad_norm": 1.5850289602846217, + "learning_rate": 5.2286232251950834e-08, + "loss": 0.2113, + "step": 1588 + }, + { + "epoch": 0.9, + "grad_norm": 1.3648614323848263, + "learning_rate": 5.170222650372469e-08, + "loss": 0.1983, + "step": 1589 + }, + { + "epoch": 0.9, + "grad_norm": 3.8680054409983224, + "learning_rate": 5.112141403426906e-08, + "loss": 0.2144, + "step": 1590 + }, + { + "epoch": 0.9, + "grad_norm": 2.268579713806137, + "learning_rate": 5.0543796799399904e-08, + "loss": 0.2309, + "step": 1591 + }, + { + "epoch": 0.9, + "grad_norm": 1.7419691975163984, + "learning_rate": 4.996937674417401e-08, + "loss": 0.2034, + "step": 1592 + }, + { + "epoch": 0.9, + "grad_norm": 3.6377344261105806, + "learning_rate": 4.9398155802881755e-08, + "loss": 0.228, + "step": 1593 + }, + { + "epoch": 0.9, + "grad_norm": 1.3434883962545732, + "learning_rate": 4.883013589904117e-08, + "loss": 0.2175, + "step": 1594 + }, + { + "epoch": 0.9, + "grad_norm": 2.02640517399693, + "learning_rate": 4.8265318945390874e-08, + "loss": 0.2079, + "step": 1595 + }, + { + "epoch": 0.9, + "grad_norm": 1.4936856398868752, + "learning_rate": 4.770370684388403e-08, + "loss": 0.2117, + "step": 1596 + }, + { + "epoch": 0.9, + "grad_norm": 20.274032433777478, + "learning_rate": 4.7145301485681965e-08, + "loss": 0.1969, + "step": 1597 + }, + { + "epoch": 0.91, + "grad_norm": 1.497128550590148, + "learning_rate": 4.659010475114755e-08, + "loss": 0.2225, + "step": 1598 + }, + { + "epoch": 0.91, + "grad_norm": 1.3871834547989423, + "learning_rate": 4.60381185098393e-08, + "loss": 0.2022, + "step": 1599 + }, + { + "epoch": 0.91, + "grad_norm": 1.87478263006469, + "learning_rate": 4.548934462050413e-08, + "loss": 0.1939, + "step": 1600 + }, + { + "epoch": 0.91, + "grad_norm": 1.767766326424203, + "learning_rate": 4.494378493107231e-08, + "loss": 0.2467, + "step": 1601 + }, + { + "epoch": 0.91, + "grad_norm": 1.5754705624250898, + "learning_rate": 4.440144127865064e-08, + "loss": 0.2028, + "step": 1602 + }, + { + "epoch": 0.91, + "grad_norm": 1.432396194607075, + "learning_rate": 4.386231548951613e-08, + "loss": 0.2123, + "step": 1603 + }, + { + "epoch": 0.91, + "grad_norm": 1.5347891574117098, + "learning_rate": 4.332640937911014e-08, + "loss": 0.2178, + "step": 1604 + }, + { + "epoch": 0.91, + "grad_norm": 1.5340164114516683, + "learning_rate": 4.27937247520318e-08, + "loss": 0.2248, + "step": 1605 + }, + { + "epoch": 0.91, + "grad_norm": 1.403051498466086, + "learning_rate": 4.226426340203282e-08, + "loss": 0.2158, + "step": 1606 + }, + { + "epoch": 0.91, + "grad_norm": 2.531605857189269, + "learning_rate": 4.173802711201069e-08, + "loss": 0.2312, + "step": 1607 + }, + { + "epoch": 0.91, + "grad_norm": 1.5801098954045238, + "learning_rate": 4.121501765400259e-08, + "loss": 0.2133, + "step": 1608 + }, + { + "epoch": 0.91, + "grad_norm": 9.937358497996158, + "learning_rate": 4.0695236789180296e-08, + "loss": 0.2253, + "step": 1609 + }, + { + "epoch": 0.91, + "grad_norm": 4.173942185576573, + "learning_rate": 4.0178686267843264e-08, + "loss": 0.231, + "step": 1610 + }, + { + "epoch": 0.91, + "grad_norm": 4.518971167916213, + "learning_rate": 3.966536782941332e-08, + "loss": 0.2239, + "step": 1611 + }, + { + "epoch": 0.91, + "grad_norm": 1.3956917843930003, + "learning_rate": 3.915528320242867e-08, + "loss": 0.2204, + "step": 1612 + }, + { + "epoch": 0.91, + "grad_norm": 1.8411508454573224, + "learning_rate": 3.8648434104537665e-08, + "loss": 0.2394, + "step": 1613 + }, + { + "epoch": 0.91, + "grad_norm": 1.4797409957087446, + "learning_rate": 3.8144822242494155e-08, + "loss": 0.2358, + "step": 1614 + }, + { + "epoch": 0.92, + "grad_norm": 1.6048927707177516, + "learning_rate": 3.764444931215027e-08, + "loss": 0.2337, + "step": 1615 + }, + { + "epoch": 0.92, + "grad_norm": 1.67415888713457, + "learning_rate": 3.714731699845186e-08, + "loss": 0.2269, + "step": 1616 + }, + { + "epoch": 0.92, + "grad_norm": 1.9908352914939207, + "learning_rate": 3.665342697543228e-08, + "loss": 0.2019, + "step": 1617 + }, + { + "epoch": 0.92, + "grad_norm": 2.096261371251174, + "learning_rate": 3.616278090620639e-08, + "loss": 0.2078, + "step": 1618 + }, + { + "epoch": 0.92, + "grad_norm": 1.49557750480535, + "learning_rate": 3.5675380442966476e-08, + "loss": 0.2348, + "step": 1619 + }, + { + "epoch": 0.92, + "grad_norm": 1.8244322568354607, + "learning_rate": 3.519122722697465e-08, + "loss": 0.2199, + "step": 1620 + }, + { + "epoch": 0.92, + "grad_norm": 7.9667062041397445, + "learning_rate": 3.4710322888558686e-08, + "loss": 0.2097, + "step": 1621 + }, + { + "epoch": 0.92, + "grad_norm": 1.349680728222318, + "learning_rate": 3.423266904710631e-08, + "loss": 0.2073, + "step": 1622 + }, + { + "epoch": 0.92, + "grad_norm": 1.7567408227096746, + "learning_rate": 3.3758267311059244e-08, + "loss": 0.218, + "step": 1623 + }, + { + "epoch": 0.92, + "grad_norm": 2.3798493322736607, + "learning_rate": 3.3287119277908394e-08, + "loss": 0.2236, + "step": 1624 + }, + { + "epoch": 0.92, + "grad_norm": 1.5980557158094102, + "learning_rate": 3.2819226534188e-08, + "loss": 0.2187, + "step": 1625 + }, + { + "epoch": 0.92, + "grad_norm": 1.6556403859671034, + "learning_rate": 3.235459065547086e-08, + "loss": 0.2272, + "step": 1626 + }, + { + "epoch": 0.92, + "grad_norm": 1.2704660633119456, + "learning_rate": 3.1893213206362404e-08, + "loss": 0.1996, + "step": 1627 + }, + { + "epoch": 0.92, + "grad_norm": 1.490056649033393, + "learning_rate": 3.143509574049552e-08, + "loss": 0.2141, + "step": 1628 + }, + { + "epoch": 0.92, + "grad_norm": 1.4551636235676808, + "learning_rate": 3.0980239800526e-08, + "loss": 0.2093, + "step": 1629 + }, + { + "epoch": 0.92, + "grad_norm": 2.593959272135738, + "learning_rate": 3.052864691812629e-08, + "loss": 0.223, + "step": 1630 + }, + { + "epoch": 0.92, + "grad_norm": 1.4857666116676196, + "learning_rate": 3.008031861398141e-08, + "loss": 0.2203, + "step": 1631 + }, + { + "epoch": 0.92, + "grad_norm": 1.3635366114841196, + "learning_rate": 2.9635256397782748e-08, + "loss": 0.2082, + "step": 1632 + }, + { + "epoch": 0.93, + "grad_norm": 1.476620834749662, + "learning_rate": 2.9193461768224016e-08, + "loss": 0.2174, + "step": 1633 + }, + { + "epoch": 0.93, + "grad_norm": 3.204260747580387, + "learning_rate": 2.8754936212995518e-08, + "loss": 0.2219, + "step": 1634 + }, + { + "epoch": 0.93, + "grad_norm": 1.4957137161142429, + "learning_rate": 2.8319681208779257e-08, + "loss": 0.2157, + "step": 1635 + }, + { + "epoch": 0.93, + "grad_norm": 2.156099477459119, + "learning_rate": 2.788769822124426e-08, + "loss": 0.207, + "step": 1636 + }, + { + "epoch": 0.93, + "grad_norm": 1.6328494750288434, + "learning_rate": 2.7458988705041152e-08, + "loss": 0.2247, + "step": 1637 + }, + { + "epoch": 0.93, + "grad_norm": 1.9436353830164899, + "learning_rate": 2.7033554103797928e-08, + "loss": 0.2226, + "step": 1638 + }, + { + "epoch": 0.93, + "grad_norm": 1.5099241671201995, + "learning_rate": 2.6611395850114294e-08, + "loss": 0.2296, + "step": 1639 + }, + { + "epoch": 0.93, + "grad_norm": 1.365804062564985, + "learning_rate": 2.6192515365557334e-08, + "loss": 0.2169, + "step": 1640 + }, + { + "epoch": 0.93, + "grad_norm": 1.56788695183528, + "learning_rate": 2.5776914060657074e-08, + "loss": 0.2146, + "step": 1641 + }, + { + "epoch": 0.93, + "grad_norm": 1.8794022535109987, + "learning_rate": 2.536459333490071e-08, + "loss": 0.2125, + "step": 1642 + }, + { + "epoch": 0.93, + "grad_norm": 1.6234388704582743, + "learning_rate": 2.4955554576728712e-08, + "loss": 0.2035, + "step": 1643 + }, + { + "epoch": 0.93, + "grad_norm": 1.3782615299906753, + "learning_rate": 2.454979916353006e-08, + "loss": 0.2226, + "step": 1644 + }, + { + "epoch": 0.93, + "grad_norm": 1.4275988733538552, + "learning_rate": 2.4147328461637138e-08, + "loss": 0.2097, + "step": 1645 + }, + { + "epoch": 0.93, + "grad_norm": 1.8311359865030032, + "learning_rate": 2.3748143826321954e-08, + "loss": 0.2128, + "step": 1646 + }, + { + "epoch": 0.93, + "grad_norm": 1.6775867813403897, + "learning_rate": 2.335224660179058e-08, + "loss": 0.2245, + "step": 1647 + }, + { + "epoch": 0.93, + "grad_norm": 2.2383467387738, + "learning_rate": 2.2959638121179292e-08, + "loss": 0.2107, + "step": 1648 + }, + { + "epoch": 0.93, + "grad_norm": 1.5239081303146866, + "learning_rate": 2.2570319706550212e-08, + "loss": 0.2289, + "step": 1649 + }, + { + "epoch": 0.93, + "grad_norm": 1.507463519802632, + "learning_rate": 2.2184292668885996e-08, + "loss": 0.2138, + "step": 1650 + }, + { + "epoch": 0.94, + "grad_norm": 1.4295100275646764, + "learning_rate": 2.18015583080865e-08, + "loss": 0.23, + "step": 1651 + }, + { + "epoch": 0.94, + "grad_norm": 1.786399786502824, + "learning_rate": 2.1422117912963556e-08, + "loss": 0.2121, + "step": 1652 + }, + { + "epoch": 0.94, + "grad_norm": 1.4173121039236776, + "learning_rate": 2.1045972761237208e-08, + "loss": 0.2249, + "step": 1653 + }, + { + "epoch": 0.94, + "grad_norm": 1.5543882499952295, + "learning_rate": 2.067312411953104e-08, + "loss": 0.225, + "step": 1654 + }, + { + "epoch": 0.94, + "grad_norm": 1.4923463631356784, + "learning_rate": 2.0303573243367955e-08, + "loss": 0.2072, + "step": 1655 + }, + { + "epoch": 0.94, + "grad_norm": 2.6430483694357796, + "learning_rate": 1.9937321377166085e-08, + "loss": 0.2231, + "step": 1656 + }, + { + "epoch": 0.94, + "grad_norm": 1.6521790224771524, + "learning_rate": 1.9574369754234654e-08, + "loss": 0.2265, + "step": 1657 + }, + { + "epoch": 0.94, + "grad_norm": 1.3453716734780017, + "learning_rate": 1.921471959676957e-08, + "loss": 0.2135, + "step": 1658 + }, + { + "epoch": 0.94, + "grad_norm": 1.9315979055113592, + "learning_rate": 1.8858372115849618e-08, + "loss": 0.2159, + "step": 1659 + }, + { + "epoch": 0.94, + "grad_norm": 1.5195138809846687, + "learning_rate": 1.850532851143194e-08, + "loss": 0.2301, + "step": 1660 + }, + { + "epoch": 0.94, + "grad_norm": 1.4173075177744054, + "learning_rate": 1.815558997234845e-08, + "loss": 0.2213, + "step": 1661 + }, + { + "epoch": 0.94, + "grad_norm": 1.473692107384903, + "learning_rate": 1.7809157676301755e-08, + "loss": 0.2183, + "step": 1662 + }, + { + "epoch": 0.94, + "grad_norm": 1.642739095041514, + "learning_rate": 1.746603278986114e-08, + "loss": 0.2229, + "step": 1663 + }, + { + "epoch": 0.94, + "grad_norm": 1.783393367447629, + "learning_rate": 1.712621646845802e-08, + "loss": 0.2088, + "step": 1664 + }, + { + "epoch": 0.94, + "grad_norm": 1.3490663075439162, + "learning_rate": 1.6789709856383172e-08, + "loss": 0.2123, + "step": 1665 + }, + { + "epoch": 0.94, + "grad_norm": 1.4871077552994478, + "learning_rate": 1.645651408678228e-08, + "loss": 0.1998, + "step": 1666 + }, + { + "epoch": 0.94, + "grad_norm": 1.492661640189132, + "learning_rate": 1.6126630281651843e-08, + "loss": 0.2181, + "step": 1667 + }, + { + "epoch": 0.95, + "grad_norm": 2.0898485545518417, + "learning_rate": 1.5800059551836276e-08, + "loss": 0.2329, + "step": 1668 + }, + { + "epoch": 0.95, + "grad_norm": 1.5865237819240359, + "learning_rate": 1.547680299702281e-08, + "loss": 0.2283, + "step": 1669 + }, + { + "epoch": 0.95, + "grad_norm": 1.5708981774765278, + "learning_rate": 1.5156861705739378e-08, + "loss": 0.2182, + "step": 1670 + }, + { + "epoch": 0.95, + "grad_norm": 1.4253287250210522, + "learning_rate": 1.4840236755349734e-08, + "loss": 0.2324, + "step": 1671 + }, + { + "epoch": 0.95, + "grad_norm": 1.5404023142116203, + "learning_rate": 1.4526929212050121e-08, + "loss": 0.2044, + "step": 1672 + }, + { + "epoch": 0.95, + "grad_norm": 1.6312167627054328, + "learning_rate": 1.4216940130866383e-08, + "loss": 0.2202, + "step": 1673 + }, + { + "epoch": 0.95, + "grad_norm": 1.3998689375748077, + "learning_rate": 1.3910270555649306e-08, + "loss": 0.2126, + "step": 1674 + }, + { + "epoch": 0.95, + "grad_norm": 1.9926382478476148, + "learning_rate": 1.3606921519071834e-08, + "loss": 0.2009, + "step": 1675 + }, + { + "epoch": 0.95, + "grad_norm": 1.5998954638513205, + "learning_rate": 1.3306894042625417e-08, + "loss": 0.2402, + "step": 1676 + }, + { + "epoch": 0.95, + "grad_norm": 1.483432459080721, + "learning_rate": 1.3010189136616445e-08, + "loss": 0.2264, + "step": 1677 + }, + { + "epoch": 0.95, + "grad_norm": 1.4478345606936291, + "learning_rate": 1.2716807800163154e-08, + "loss": 0.2253, + "step": 1678 + }, + { + "epoch": 0.95, + "grad_norm": 1.246370012322096, + "learning_rate": 1.2426751021191839e-08, + "loss": 0.1948, + "step": 1679 + }, + { + "epoch": 0.95, + "grad_norm": 1.8974726138329356, + "learning_rate": 1.2140019776433751e-08, + "loss": 0.2217, + "step": 1680 + }, + { + "epoch": 0.95, + "grad_norm": 1.3723676888800223, + "learning_rate": 1.1856615031422212e-08, + "loss": 0.2192, + "step": 1681 + }, + { + "epoch": 0.95, + "grad_norm": 1.8069982766502515, + "learning_rate": 1.1576537740488279e-08, + "loss": 0.2216, + "step": 1682 + }, + { + "epoch": 0.95, + "grad_norm": 1.4177935456943451, + "learning_rate": 1.129978884675875e-08, + "loss": 0.2146, + "step": 1683 + }, + { + "epoch": 0.95, + "grad_norm": 1.755283753533124, + "learning_rate": 1.1026369282152282e-08, + "loss": 0.1951, + "step": 1684 + }, + { + "epoch": 0.95, + "grad_norm": 1.5736974160620574, + "learning_rate": 1.075627996737627e-08, + "loss": 0.2173, + "step": 1685 + }, + { + "epoch": 0.96, + "grad_norm": 1.3627694353353759, + "learning_rate": 1.0489521811924085e-08, + "loss": 0.2095, + "step": 1686 + }, + { + "epoch": 0.96, + "grad_norm": 1.7844568756844557, + "learning_rate": 1.0226095714071736e-08, + "loss": 0.2159, + "step": 1687 + }, + { + "epoch": 0.96, + "grad_norm": 1.5122829790186212, + "learning_rate": 9.966002560874986e-09, + "loss": 0.2291, + "step": 1688 + }, + { + "epoch": 0.96, + "grad_norm": 1.6848929290314854, + "learning_rate": 9.709243228166242e-09, + "loss": 0.2205, + "step": 1689 + }, + { + "epoch": 0.96, + "grad_norm": 1.4297757700346223, + "learning_rate": 9.455818580551778e-09, + "loss": 0.2041, + "step": 1690 + }, + { + "epoch": 0.96, + "grad_norm": 2.566886180302486, + "learning_rate": 9.205729471408407e-09, + "loss": 0.2191, + "step": 1691 + }, + { + "epoch": 0.96, + "grad_norm": 1.5204098942937077, + "learning_rate": 8.958976742881374e-09, + "loss": 0.2268, + "step": 1692 + }, + { + "epoch": 0.96, + "grad_norm": 1.3736739493111134, + "learning_rate": 8.715561225880686e-09, + "loss": 0.2139, + "step": 1693 + }, + { + "epoch": 0.96, + "grad_norm": 1.4881190262164439, + "learning_rate": 8.475483740079003e-09, + "loss": 0.2281, + "step": 1694 + }, + { + "epoch": 0.96, + "grad_norm": 1.5636217330176294, + "learning_rate": 8.238745093908205e-09, + "loss": 0.2231, + "step": 1695 + }, + { + "epoch": 0.96, + "grad_norm": 1.5472527996577077, + "learning_rate": 8.005346084557386e-09, + "loss": 0.1908, + "step": 1696 + }, + { + "epoch": 0.96, + "grad_norm": 1.5004416569140946, + "learning_rate": 7.775287497969519e-09, + "loss": 0.2253, + "step": 1697 + }, + { + "epoch": 0.96, + "grad_norm": 1.8534499328257499, + "learning_rate": 7.54857010883947e-09, + "loss": 0.2182, + "step": 1698 + }, + { + "epoch": 0.96, + "grad_norm": 1.3382648382057611, + "learning_rate": 7.325194680610436e-09, + "loss": 0.1972, + "step": 1699 + }, + { + "epoch": 0.96, + "grad_norm": 1.5659600073117632, + "learning_rate": 7.105161965472617e-09, + "loss": 0.2142, + "step": 1700 + }, + { + "epoch": 0.96, + "grad_norm": 1.9514935130657527, + "learning_rate": 6.88847270435966e-09, + "loss": 0.2056, + "step": 1701 + }, + { + "epoch": 0.96, + "grad_norm": 1.605942494680415, + "learning_rate": 6.675127626946553e-09, + "loss": 0.2243, + "step": 1702 + }, + { + "epoch": 0.96, + "grad_norm": 1.7565254285536078, + "learning_rate": 6.46512745164729e-09, + "loss": 0.2289, + "step": 1703 + }, + { + "epoch": 0.97, + "grad_norm": 4.13556702752091, + "learning_rate": 6.258472885612098e-09, + "loss": 0.2161, + "step": 1704 + }, + { + "epoch": 0.97, + "grad_norm": 1.581102577206706, + "learning_rate": 6.055164624725551e-09, + "loss": 0.2352, + "step": 1705 + }, + { + "epoch": 0.97, + "grad_norm": 1.410106957250708, + "learning_rate": 5.855203353603677e-09, + "loss": 0.2262, + "step": 1706 + }, + { + "epoch": 0.97, + "grad_norm": 1.3592831661304072, + "learning_rate": 5.658589745591968e-09, + "loss": 0.2256, + "step": 1707 + }, + { + "epoch": 0.97, + "grad_norm": 1.5547871151239505, + "learning_rate": 5.465324462763266e-09, + "loss": 0.2398, + "step": 1708 + }, + { + "epoch": 0.97, + "grad_norm": 1.4198995333939302, + "learning_rate": 5.2754081559149844e-09, + "loss": 0.2014, + "step": 1709 + }, + { + "epoch": 0.97, + "grad_norm": 1.5400020307453715, + "learning_rate": 5.088841464567784e-09, + "loss": 0.2023, + "step": 1710 + }, + { + "epoch": 0.97, + "grad_norm": 1.3812754256037012, + "learning_rate": 4.905625016962234e-09, + "loss": 0.2181, + "step": 1711 + }, + { + "epoch": 0.97, + "grad_norm": 1.6152068825915296, + "learning_rate": 4.7257594300578184e-09, + "loss": 0.221, + "step": 1712 + }, + { + "epoch": 0.97, + "grad_norm": 1.5693585681503721, + "learning_rate": 4.54924530953027e-09, + "loss": 0.2129, + "step": 1713 + }, + { + "epoch": 0.97, + "grad_norm": 1.5542248440241997, + "learning_rate": 4.3760832497697905e-09, + "loss": 0.2407, + "step": 1714 + }, + { + "epoch": 0.97, + "grad_norm": 1.6045560815256188, + "learning_rate": 4.206273833878615e-09, + "loss": 0.2153, + "step": 1715 + }, + { + "epoch": 0.97, + "grad_norm": 1.591784114793385, + "learning_rate": 4.039817633669673e-09, + "loss": 0.2148, + "step": 1716 + }, + { + "epoch": 0.97, + "grad_norm": 1.3954936503921582, + "learning_rate": 3.8767152096641496e-09, + "loss": 0.2052, + "step": 1717 + }, + { + "epoch": 0.97, + "grad_norm": 1.3601376810220163, + "learning_rate": 3.7169671110899304e-09, + "loss": 0.2136, + "step": 1718 + }, + { + "epoch": 0.97, + "grad_norm": 1.502209290740096, + "learning_rate": 3.56057387587938e-09, + "loss": 0.2046, + "step": 1719 + }, + { + "epoch": 0.97, + "grad_norm": 1.5344847978424718, + "learning_rate": 3.407536030668123e-09, + "loss": 0.215, + "step": 1720 + }, + { + "epoch": 0.98, + "grad_norm": 1.451868504607553, + "learning_rate": 3.2578540907926e-09, + "loss": 0.2131, + "step": 1721 + }, + { + "epoch": 0.98, + "grad_norm": 1.4304961499600137, + "learning_rate": 3.111528560288734e-09, + "loss": 0.2191, + "step": 1722 + }, + { + "epoch": 0.98, + "grad_norm": 1.6344184430248494, + "learning_rate": 2.9685599318901576e-09, + "loss": 0.2097, + "step": 1723 + }, + { + "epoch": 0.98, + "grad_norm": 1.5043449810745821, + "learning_rate": 2.8289486870266554e-09, + "loss": 0.2311, + "step": 1724 + }, + { + "epoch": 0.98, + "grad_norm": 1.4986597599207385, + "learning_rate": 2.6926952958221673e-09, + "loss": 0.2144, + "step": 1725 + }, + { + "epoch": 0.98, + "grad_norm": 1.837441809837519, + "learning_rate": 2.5598002170936773e-09, + "loss": 0.2159, + "step": 1726 + }, + { + "epoch": 0.98, + "grad_norm": 1.2869602196355845, + "learning_rate": 2.4302638983494383e-09, + "loss": 0.1989, + "step": 1727 + }, + { + "epoch": 0.98, + "grad_norm": 2.80501369930481, + "learning_rate": 2.3040867757873062e-09, + "loss": 0.236, + "step": 1728 + }, + { + "epoch": 0.98, + "grad_norm": 2.026626697233838, + "learning_rate": 2.181269274293629e-09, + "loss": 0.2211, + "step": 1729 + }, + { + "epoch": 0.98, + "grad_norm": 4.51348700300453, + "learning_rate": 2.0618118074416937e-09, + "loss": 0.2205, + "step": 1730 + }, + { + "epoch": 0.98, + "grad_norm": 1.5596903400462347, + "learning_rate": 1.9457147774900595e-09, + "loss": 0.205, + "step": 1731 + }, + { + "epoch": 0.98, + "grad_norm": 1.882725397694564, + "learning_rate": 1.8329785753815608e-09, + "loss": 0.2317, + "step": 1732 + }, + { + "epoch": 0.98, + "grad_norm": 1.9512723895662012, + "learning_rate": 1.7236035807416393e-09, + "loss": 0.2201, + "step": 1733 + }, + { + "epoch": 0.98, + "grad_norm": 1.354336132015606, + "learning_rate": 1.61759016187768e-09, + "loss": 0.2037, + "step": 1734 + }, + { + "epoch": 0.98, + "grad_norm": 1.3803520556819424, + "learning_rate": 1.514938675777122e-09, + "loss": 0.203, + "step": 1735 + }, + { + "epoch": 0.98, + "grad_norm": 7.9565314526269555, + "learning_rate": 1.4156494681062391e-09, + "loss": 0.2131, + "step": 1736 + }, + { + "epoch": 0.98, + "grad_norm": 2.6119253614050866, + "learning_rate": 1.3197228732098053e-09, + "loss": 0.2166, + "step": 1737 + }, + { + "epoch": 0.98, + "grad_norm": 1.3965306376384607, + "learning_rate": 1.227159214108986e-09, + "loss": 0.2093, + "step": 1738 + }, + { + "epoch": 0.99, + "grad_norm": 2.1585813547747885, + "learning_rate": 1.1379588025006715e-09, + "loss": 0.2143, + "step": 1739 + }, + { + "epoch": 0.99, + "grad_norm": 2.381013444606734, + "learning_rate": 1.0521219387567004e-09, + "loss": 0.223, + "step": 1740 + }, + { + "epoch": 0.99, + "grad_norm": 2.098260845356692, + "learning_rate": 9.69648911922194e-10, + "loss": 0.2088, + "step": 1741 + }, + { + "epoch": 0.99, + "grad_norm": 1.8980378880567914, + "learning_rate": 8.90539999715445e-10, + "loss": 0.2189, + "step": 1742 + }, + { + "epoch": 0.99, + "grad_norm": 2.6418037592774217, + "learning_rate": 8.147954685259195e-10, + "loss": 0.2248, + "step": 1743 + }, + { + "epoch": 0.99, + "grad_norm": 2.626933319590434, + "learning_rate": 7.424155734143678e-10, + "loss": 0.2181, + "step": 1744 + }, + { + "epoch": 0.99, + "grad_norm": 1.5177164734687123, + "learning_rate": 6.734005581113811e-10, + "loss": 0.2207, + "step": 1745 + }, + { + "epoch": 0.99, + "grad_norm": 1.4065737839396852, + "learning_rate": 6.077506550167255e-10, + "loss": 0.2185, + "step": 1746 + }, + { + "epoch": 0.99, + "grad_norm": 1.5169283615668283, + "learning_rate": 5.454660851984539e-10, + "loss": 0.2139, + "step": 1747 + }, + { + "epoch": 0.99, + "grad_norm": 1.6397705969706347, + "learning_rate": 4.865470583925724e-10, + "loss": 0.2199, + "step": 1748 + }, + { + "epoch": 0.99, + "grad_norm": 1.3531140580530334, + "learning_rate": 4.309937730015978e-10, + "loss": 0.2181, + "step": 1749 + }, + { + "epoch": 0.99, + "grad_norm": 1.3849252208100733, + "learning_rate": 3.7880641609466804e-10, + "loss": 0.222, + "step": 1750 + }, + { + "epoch": 0.99, + "grad_norm": 1.4522196942042183, + "learning_rate": 3.2998516340654314e-10, + "loss": 0.211, + "step": 1751 + }, + { + "epoch": 0.99, + "grad_norm": 1.5915915167513823, + "learning_rate": 2.845301793366062e-10, + "loss": 0.2133, + "step": 1752 + }, + { + "epoch": 0.99, + "grad_norm": 1.8725093287117287, + "learning_rate": 2.424416169494181e-10, + "loss": 0.2217, + "step": 1753 + }, + { + "epoch": 0.99, + "grad_norm": 1.78230003998772, + "learning_rate": 2.0371961797283065e-10, + "loss": 0.2146, + "step": 1754 + }, + { + "epoch": 0.99, + "grad_norm": 1.3528306161369248, + "learning_rate": 1.683643127987633e-10, + "loss": 0.2171, + "step": 1755 + }, + { + "epoch": 0.99, + "grad_norm": 5.00774323400259, + "learning_rate": 1.36375820481871e-10, + "loss": 0.2199, + "step": 1756 + }, + { + "epoch": 1.0, + "grad_norm": 1.5365441696780864, + "learning_rate": 1.077542487394334e-10, + "loss": 0.2192, + "step": 1757 + }, + { + "epoch": 1.0, + "grad_norm": 1.8431032222811043, + "learning_rate": 8.249969395135448e-11, + "loss": 0.2233, + "step": 1758 + }, + { + "epoch": 1.0, + "grad_norm": 1.883075408661139, + "learning_rate": 6.061224115927465e-11, + "loss": 0.2305, + "step": 1759 + }, + { + "epoch": 1.0, + "grad_norm": 1.423053784552846, + "learning_rate": 4.209196406645965e-11, + "loss": 0.2155, + "step": 1760 + }, + { + "epoch": 1.0, + "grad_norm": 1.278276820338806, + "learning_rate": 2.6938925037689463e-11, + "loss": 0.2195, + "step": 1761 + }, + { + "epoch": 1.0, + "grad_norm": 1.3025894483812757, + "learning_rate": 1.5153175099147375e-11, + "loss": 0.2252, + "step": 1762 + }, + { + "epoch": 1.0, + "grad_norm": 1.6984971125125101, + "learning_rate": 6.7347539378648234e-12, + "loss": 0.2262, + "step": 1763 + }, + { + "epoch": 1.0, + "grad_norm": 1.3860308112942907, + "learning_rate": 1.6836899018324302e-12, + "loss": 0.2173, + "step": 1764 + }, + { + "epoch": 1.0, + "grad_norm": 1.4084210009732494, + "learning_rate": 0.0, + "loss": 0.2277, + "step": 1765 + }, + { + "epoch": 1.0, + "step": 1765, + "total_flos": 1353217200881664.0, + "train_loss": 0.2451857447877484, + "train_runtime": 48874.5213, + "train_samples_per_second": 2.311, + "train_steps_per_second": 0.036 + } + ], + "logging_steps": 1.0, + "max_steps": 1765, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 25000, + "total_flos": 1353217200881664.0, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}