{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 3693, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0008123476848090983, "grad_norm": 5.867218494415283, "learning_rate": 2.702702702702703e-08, "loss": 0.8628, "step": 1 }, { "epoch": 0.0016246953696181965, "grad_norm": 5.787890911102295, "learning_rate": 5.405405405405406e-08, "loss": 0.8838, "step": 2 }, { "epoch": 0.002437043054427295, "grad_norm": 6.13796329498291, "learning_rate": 8.108108108108109e-08, "loss": 0.9121, "step": 3 }, { "epoch": 0.003249390739236393, "grad_norm": 5.860419273376465, "learning_rate": 1.0810810810810812e-07, "loss": 0.8777, "step": 4 }, { "epoch": 0.004061738424045491, "grad_norm": 5.712265491485596, "learning_rate": 1.3513513513513515e-07, "loss": 0.8437, "step": 5 }, { "epoch": 0.00487408610885459, "grad_norm": 5.940003395080566, "learning_rate": 1.6216216216216218e-07, "loss": 0.8641, "step": 6 }, { "epoch": 0.005686433793663688, "grad_norm": 6.243408203125, "learning_rate": 1.8918918918918921e-07, "loss": 0.8939, "step": 7 }, { "epoch": 0.006498781478472786, "grad_norm": 5.8776726722717285, "learning_rate": 2.1621621621621625e-07, "loss": 0.8622, "step": 8 }, { "epoch": 0.007311129163281885, "grad_norm": 5.919240951538086, "learning_rate": 2.4324324324324326e-07, "loss": 0.8839, "step": 9 }, { "epoch": 0.008123476848090982, "grad_norm": 5.893919467926025, "learning_rate": 2.702702702702703e-07, "loss": 0.8916, "step": 10 }, { "epoch": 0.008935824532900082, "grad_norm": 5.7647318840026855, "learning_rate": 2.972972972972973e-07, "loss": 0.848, "step": 11 }, { "epoch": 0.00974817221770918, "grad_norm": 5.382693290710449, "learning_rate": 3.2432432432432436e-07, "loss": 0.8041, "step": 12 }, { "epoch": 0.010560519902518278, "grad_norm": 5.630814552307129, "learning_rate": 3.513513513513514e-07, "loss": 0.8585, "step": 13 }, { "epoch": 0.011372867587327376, "grad_norm": 5.537968635559082, "learning_rate": 3.7837837837837843e-07, "loss": 0.841, "step": 14 }, { "epoch": 0.012185215272136474, "grad_norm": 5.487429618835449, "learning_rate": 4.0540540540540546e-07, "loss": 0.8689, "step": 15 }, { "epoch": 0.012997562956945572, "grad_norm": 5.420842170715332, "learning_rate": 4.324324324324325e-07, "loss": 0.8562, "step": 16 }, { "epoch": 0.01380991064175467, "grad_norm": 5.586460113525391, "learning_rate": 4.5945945945945953e-07, "loss": 0.8716, "step": 17 }, { "epoch": 0.01462225832656377, "grad_norm": 5.419436454772949, "learning_rate": 4.864864864864865e-07, "loss": 0.8662, "step": 18 }, { "epoch": 0.015434606011372868, "grad_norm": 4.347222805023193, "learning_rate": 5.135135135135135e-07, "loss": 0.8066, "step": 19 }, { "epoch": 0.016246953696181964, "grad_norm": 4.490281581878662, "learning_rate": 5.405405405405406e-07, "loss": 0.8453, "step": 20 }, { "epoch": 0.017059301380991064, "grad_norm": 4.176577091217041, "learning_rate": 5.675675675675676e-07, "loss": 0.8095, "step": 21 }, { "epoch": 0.017871649065800164, "grad_norm": 4.286707401275635, "learning_rate": 5.945945945945947e-07, "loss": 0.839, "step": 22 }, { "epoch": 0.01868399675060926, "grad_norm": 4.250885486602783, "learning_rate": 6.216216216216217e-07, "loss": 0.8297, "step": 23 }, { "epoch": 0.01949634443541836, "grad_norm": 3.8648900985717773, "learning_rate": 6.486486486486487e-07, "loss": 0.8004, "step": 24 }, { "epoch": 0.020308692120227456, "grad_norm": 3.792182207107544, "learning_rate": 6.756756756756758e-07, "loss": 0.7713, "step": 25 }, { "epoch": 0.021121039805036556, "grad_norm": 2.3136017322540283, "learning_rate": 7.027027027027028e-07, "loss": 0.7494, "step": 26 }, { "epoch": 0.021933387489845652, "grad_norm": 2.258458137512207, "learning_rate": 7.297297297297298e-07, "loss": 0.7947, "step": 27 }, { "epoch": 0.022745735174654752, "grad_norm": 2.2972660064697266, "learning_rate": 7.567567567567569e-07, "loss": 0.8013, "step": 28 }, { "epoch": 0.023558082859463852, "grad_norm": 2.16971755027771, "learning_rate": 7.837837837837839e-07, "loss": 0.8025, "step": 29 }, { "epoch": 0.024370430544272948, "grad_norm": 2.128912925720215, "learning_rate": 8.108108108108109e-07, "loss": 0.8141, "step": 30 }, { "epoch": 0.025182778229082048, "grad_norm": 1.8542791604995728, "learning_rate": 8.37837837837838e-07, "loss": 0.7542, "step": 31 }, { "epoch": 0.025995125913891144, "grad_norm": 1.8852665424346924, "learning_rate": 8.64864864864865e-07, "loss": 0.7349, "step": 32 }, { "epoch": 0.026807473598700244, "grad_norm": 1.8394138813018799, "learning_rate": 8.91891891891892e-07, "loss": 0.7407, "step": 33 }, { "epoch": 0.02761982128350934, "grad_norm": 1.610342264175415, "learning_rate": 9.189189189189191e-07, "loss": 0.7365, "step": 34 }, { "epoch": 0.02843216896831844, "grad_norm": 1.4266325235366821, "learning_rate": 9.459459459459461e-07, "loss": 0.7566, "step": 35 }, { "epoch": 0.02924451665312754, "grad_norm": 1.8599340915679932, "learning_rate": 9.72972972972973e-07, "loss": 0.7626, "step": 36 }, { "epoch": 0.030056864337936636, "grad_norm": 2.0598795413970947, "learning_rate": 1.0000000000000002e-06, "loss": 0.7134, "step": 37 }, { "epoch": 0.030869212022745736, "grad_norm": 1.9298954010009766, "learning_rate": 1.027027027027027e-06, "loss": 0.6655, "step": 38 }, { "epoch": 0.03168155970755483, "grad_norm": 2.1654577255249023, "learning_rate": 1.0540540540540542e-06, "loss": 0.7339, "step": 39 }, { "epoch": 0.03249390739236393, "grad_norm": 2.0767595767974854, "learning_rate": 1.0810810810810812e-06, "loss": 0.7323, "step": 40 }, { "epoch": 0.03330625507717303, "grad_norm": 2.04421329498291, "learning_rate": 1.1081081081081083e-06, "loss": 0.7592, "step": 41 }, { "epoch": 0.03411860276198213, "grad_norm": 1.5938338041305542, "learning_rate": 1.1351351351351352e-06, "loss": 0.6743, "step": 42 }, { "epoch": 0.034930950446791224, "grad_norm": 1.3853787183761597, "learning_rate": 1.1621621621621624e-06, "loss": 0.6677, "step": 43 }, { "epoch": 0.03574329813160033, "grad_norm": 1.2020511627197266, "learning_rate": 1.1891891891891893e-06, "loss": 0.6686, "step": 44 }, { "epoch": 0.036555645816409424, "grad_norm": 1.1168713569641113, "learning_rate": 1.2162162162162164e-06, "loss": 0.7222, "step": 45 }, { "epoch": 0.03736799350121852, "grad_norm": 0.9262130856513977, "learning_rate": 1.2432432432432434e-06, "loss": 0.6936, "step": 46 }, { "epoch": 0.038180341186027617, "grad_norm": 0.9542868733406067, "learning_rate": 1.2702702702702705e-06, "loss": 0.7179, "step": 47 }, { "epoch": 0.03899268887083672, "grad_norm": 1.0097525119781494, "learning_rate": 1.2972972972972974e-06, "loss": 0.6715, "step": 48 }, { "epoch": 0.039805036555645816, "grad_norm": 1.116047739982605, "learning_rate": 1.3243243243243246e-06, "loss": 0.7075, "step": 49 }, { "epoch": 0.04061738424045491, "grad_norm": 1.0406266450881958, "learning_rate": 1.3513513513513515e-06, "loss": 0.6526, "step": 50 }, { "epoch": 0.041429731925264016, "grad_norm": 0.9339987635612488, "learning_rate": 1.3783783783783786e-06, "loss": 0.6424, "step": 51 }, { "epoch": 0.04224207961007311, "grad_norm": 0.9256858229637146, "learning_rate": 1.4054054054054056e-06, "loss": 0.6861, "step": 52 }, { "epoch": 0.04305442729488221, "grad_norm": 0.7636408805847168, "learning_rate": 1.4324324324324327e-06, "loss": 0.6613, "step": 53 }, { "epoch": 0.043866774979691305, "grad_norm": 0.8064551949501038, "learning_rate": 1.4594594594594596e-06, "loss": 0.6547, "step": 54 }, { "epoch": 0.04467912266450041, "grad_norm": 0.7482165098190308, "learning_rate": 1.4864864864864868e-06, "loss": 0.6766, "step": 55 }, { "epoch": 0.045491470349309504, "grad_norm": 0.7477511763572693, "learning_rate": 1.5135135135135137e-06, "loss": 0.6763, "step": 56 }, { "epoch": 0.0463038180341186, "grad_norm": 0.6924583911895752, "learning_rate": 1.5405405405405409e-06, "loss": 0.6503, "step": 57 }, { "epoch": 0.047116165718927704, "grad_norm": 0.6601155996322632, "learning_rate": 1.5675675675675678e-06, "loss": 0.6584, "step": 58 }, { "epoch": 0.0479285134037368, "grad_norm": 0.7005034685134888, "learning_rate": 1.5945945945945947e-06, "loss": 0.6486, "step": 59 }, { "epoch": 0.048740861088545896, "grad_norm": 0.7050193548202515, "learning_rate": 1.6216216216216219e-06, "loss": 0.6544, "step": 60 }, { "epoch": 0.04955320877335499, "grad_norm": 0.7136906385421753, "learning_rate": 1.6486486486486488e-06, "loss": 0.6371, "step": 61 }, { "epoch": 0.050365556458164096, "grad_norm": 0.5535628795623779, "learning_rate": 1.675675675675676e-06, "loss": 0.5961, "step": 62 }, { "epoch": 0.05117790414297319, "grad_norm": 0.5116686224937439, "learning_rate": 1.7027027027027028e-06, "loss": 0.6164, "step": 63 }, { "epoch": 0.05199025182778229, "grad_norm": 0.5073972940444946, "learning_rate": 1.72972972972973e-06, "loss": 0.6263, "step": 64 }, { "epoch": 0.05280259951259139, "grad_norm": 0.5199340581893921, "learning_rate": 1.756756756756757e-06, "loss": 0.6166, "step": 65 }, { "epoch": 0.05361494719740049, "grad_norm": 0.5627139210700989, "learning_rate": 1.783783783783784e-06, "loss": 0.6282, "step": 66 }, { "epoch": 0.054427294882209584, "grad_norm": 0.45891445875167847, "learning_rate": 1.810810810810811e-06, "loss": 0.6174, "step": 67 }, { "epoch": 0.05523964256701868, "grad_norm": 0.5445205569267273, "learning_rate": 1.8378378378378381e-06, "loss": 0.6451, "step": 68 }, { "epoch": 0.056051990251827784, "grad_norm": 0.5121186971664429, "learning_rate": 1.864864864864865e-06, "loss": 0.6647, "step": 69 }, { "epoch": 0.05686433793663688, "grad_norm": 0.479709267616272, "learning_rate": 1.8918918918918922e-06, "loss": 0.6136, "step": 70 }, { "epoch": 0.057676685621445976, "grad_norm": 0.4646925628185272, "learning_rate": 1.918918918918919e-06, "loss": 0.6147, "step": 71 }, { "epoch": 0.05848903330625508, "grad_norm": 0.48176467418670654, "learning_rate": 1.945945945945946e-06, "loss": 0.6171, "step": 72 }, { "epoch": 0.059301380991064176, "grad_norm": 0.4632847011089325, "learning_rate": 1.9729729729729734e-06, "loss": 0.609, "step": 73 }, { "epoch": 0.06011372867587327, "grad_norm": 0.45390772819519043, "learning_rate": 2.0000000000000003e-06, "loss": 0.5862, "step": 74 }, { "epoch": 0.060926076360682375, "grad_norm": 0.4966928958892822, "learning_rate": 2.0270270270270273e-06, "loss": 0.6049, "step": 75 }, { "epoch": 0.06173842404549147, "grad_norm": 0.4609528183937073, "learning_rate": 2.054054054054054e-06, "loss": 0.6285, "step": 76 }, { "epoch": 0.06255077173030058, "grad_norm": 0.48760563135147095, "learning_rate": 2.0810810810810815e-06, "loss": 0.6252, "step": 77 }, { "epoch": 0.06336311941510966, "grad_norm": 0.5056685209274292, "learning_rate": 2.1081081081081085e-06, "loss": 0.603, "step": 78 }, { "epoch": 0.06417546709991877, "grad_norm": 0.43392258882522583, "learning_rate": 2.1351351351351354e-06, "loss": 0.5914, "step": 79 }, { "epoch": 0.06498781478472786, "grad_norm": 0.47377288341522217, "learning_rate": 2.1621621621621623e-06, "loss": 0.6299, "step": 80 }, { "epoch": 0.06580016246953696, "grad_norm": 0.42371344566345215, "learning_rate": 2.1891891891891897e-06, "loss": 0.5846, "step": 81 }, { "epoch": 0.06661251015434606, "grad_norm": 0.4257068634033203, "learning_rate": 2.2162162162162166e-06, "loss": 0.5976, "step": 82 }, { "epoch": 0.06742485783915515, "grad_norm": 0.4191585183143616, "learning_rate": 2.2432432432432435e-06, "loss": 0.5979, "step": 83 }, { "epoch": 0.06823720552396426, "grad_norm": 0.5166892409324646, "learning_rate": 2.2702702702702705e-06, "loss": 0.6091, "step": 84 }, { "epoch": 0.06904955320877336, "grad_norm": 0.41576263308525085, "learning_rate": 2.297297297297298e-06, "loss": 0.5847, "step": 85 }, { "epoch": 0.06986190089358245, "grad_norm": 0.4071441888809204, "learning_rate": 2.3243243243243247e-06, "loss": 0.6053, "step": 86 }, { "epoch": 0.07067424857839155, "grad_norm": 0.45543980598449707, "learning_rate": 2.3513513513513517e-06, "loss": 0.6191, "step": 87 }, { "epoch": 0.07148659626320066, "grad_norm": 0.3929440975189209, "learning_rate": 2.3783783783783786e-06, "loss": 0.5987, "step": 88 }, { "epoch": 0.07229894394800974, "grad_norm": 0.42005449533462524, "learning_rate": 2.4054054054054055e-06, "loss": 0.557, "step": 89 }, { "epoch": 0.07311129163281885, "grad_norm": 0.40897229313850403, "learning_rate": 2.432432432432433e-06, "loss": 0.6215, "step": 90 }, { "epoch": 0.07392363931762795, "grad_norm": 0.4332367181777954, "learning_rate": 2.45945945945946e-06, "loss": 0.5999, "step": 91 }, { "epoch": 0.07473598700243704, "grad_norm": 0.40798962116241455, "learning_rate": 2.4864864864864867e-06, "loss": 0.5815, "step": 92 }, { "epoch": 0.07554833468724614, "grad_norm": 0.3801487386226654, "learning_rate": 2.5135135135135137e-06, "loss": 0.5443, "step": 93 }, { "epoch": 0.07636068237205523, "grad_norm": 0.45121127367019653, "learning_rate": 2.540540540540541e-06, "loss": 0.5544, "step": 94 }, { "epoch": 0.07717303005686434, "grad_norm": 0.4127246141433716, "learning_rate": 2.5675675675675675e-06, "loss": 0.582, "step": 95 }, { "epoch": 0.07798537774167344, "grad_norm": 0.42333000898361206, "learning_rate": 2.594594594594595e-06, "loss": 0.5627, "step": 96 }, { "epoch": 0.07879772542648253, "grad_norm": 0.3946426212787628, "learning_rate": 2.621621621621622e-06, "loss": 0.6107, "step": 97 }, { "epoch": 0.07961007311129163, "grad_norm": 0.3626299202442169, "learning_rate": 2.648648648648649e-06, "loss": 0.6022, "step": 98 }, { "epoch": 0.08042242079610074, "grad_norm": 0.4131406247615814, "learning_rate": 2.6756756756756757e-06, "loss": 0.5963, "step": 99 }, { "epoch": 0.08123476848090982, "grad_norm": 0.4487226903438568, "learning_rate": 2.702702702702703e-06, "loss": 0.5406, "step": 100 }, { "epoch": 0.08204711616571893, "grad_norm": 0.43455076217651367, "learning_rate": 2.72972972972973e-06, "loss": 0.5734, "step": 101 }, { "epoch": 0.08285946385052803, "grad_norm": 0.3843064308166504, "learning_rate": 2.7567567567567573e-06, "loss": 0.5854, "step": 102 }, { "epoch": 0.08367181153533712, "grad_norm": 0.3916749060153961, "learning_rate": 2.783783783783784e-06, "loss": 0.5693, "step": 103 }, { "epoch": 0.08448415922014622, "grad_norm": 0.38580673933029175, "learning_rate": 2.810810810810811e-06, "loss": 0.5653, "step": 104 }, { "epoch": 0.08529650690495533, "grad_norm": 0.41001734137535095, "learning_rate": 2.837837837837838e-06, "loss": 0.5571, "step": 105 }, { "epoch": 0.08610885458976442, "grad_norm": 0.4890478253364563, "learning_rate": 2.8648648648648654e-06, "loss": 0.5906, "step": 106 }, { "epoch": 0.08692120227457352, "grad_norm": 0.37780821323394775, "learning_rate": 2.891891891891892e-06, "loss": 0.5959, "step": 107 }, { "epoch": 0.08773354995938261, "grad_norm": 0.42310410737991333, "learning_rate": 2.9189189189189193e-06, "loss": 0.577, "step": 108 }, { "epoch": 0.08854589764419171, "grad_norm": 0.40478944778442383, "learning_rate": 2.9459459459459462e-06, "loss": 0.5985, "step": 109 }, { "epoch": 0.08935824532900082, "grad_norm": 0.42266377806663513, "learning_rate": 2.9729729729729736e-06, "loss": 0.5932, "step": 110 }, { "epoch": 0.0901705930138099, "grad_norm": 0.4191412627696991, "learning_rate": 3e-06, "loss": 0.5742, "step": 111 }, { "epoch": 0.09098294069861901, "grad_norm": 0.4237963557243347, "learning_rate": 3.0270270270270274e-06, "loss": 0.571, "step": 112 }, { "epoch": 0.09179528838342811, "grad_norm": 0.47107890248298645, "learning_rate": 3.0540540540540544e-06, "loss": 0.599, "step": 113 }, { "epoch": 0.0926076360682372, "grad_norm": 0.46631595492362976, "learning_rate": 3.0810810810810817e-06, "loss": 0.5853, "step": 114 }, { "epoch": 0.0934199837530463, "grad_norm": 0.4227235019207001, "learning_rate": 3.1081081081081082e-06, "loss": 0.5188, "step": 115 }, { "epoch": 0.09423233143785541, "grad_norm": 0.47727710008621216, "learning_rate": 3.1351351351351356e-06, "loss": 0.5775, "step": 116 }, { "epoch": 0.0950446791226645, "grad_norm": 0.49897849559783936, "learning_rate": 3.1621621621621625e-06, "loss": 0.5941, "step": 117 }, { "epoch": 0.0958570268074736, "grad_norm": 0.37962016463279724, "learning_rate": 3.1891891891891894e-06, "loss": 0.5693, "step": 118 }, { "epoch": 0.0966693744922827, "grad_norm": 0.44942930340766907, "learning_rate": 3.2162162162162164e-06, "loss": 0.5627, "step": 119 }, { "epoch": 0.09748172217709179, "grad_norm": 0.44786015152931213, "learning_rate": 3.2432432432432437e-06, "loss": 0.5683, "step": 120 }, { "epoch": 0.0982940698619009, "grad_norm": 0.4766939878463745, "learning_rate": 3.2702702702702706e-06, "loss": 0.5736, "step": 121 }, { "epoch": 0.09910641754670999, "grad_norm": 0.39454421401023865, "learning_rate": 3.2972972972972976e-06, "loss": 0.569, "step": 122 }, { "epoch": 0.09991876523151909, "grad_norm": 0.38677218556404114, "learning_rate": 3.3243243243243245e-06, "loss": 0.5604, "step": 123 }, { "epoch": 0.10073111291632819, "grad_norm": 0.42302370071411133, "learning_rate": 3.351351351351352e-06, "loss": 0.5698, "step": 124 }, { "epoch": 0.10154346060113728, "grad_norm": 0.43613728880882263, "learning_rate": 3.3783783783783788e-06, "loss": 0.5632, "step": 125 }, { "epoch": 0.10235580828594638, "grad_norm": 0.41828638315200806, "learning_rate": 3.4054054054054057e-06, "loss": 0.5853, "step": 126 }, { "epoch": 0.10316815597075549, "grad_norm": 0.4052099287509918, "learning_rate": 3.4324324324324326e-06, "loss": 0.5862, "step": 127 }, { "epoch": 0.10398050365556458, "grad_norm": 0.42845723032951355, "learning_rate": 3.45945945945946e-06, "loss": 0.5651, "step": 128 }, { "epoch": 0.10479285134037368, "grad_norm": 0.4249141216278076, "learning_rate": 3.4864864864864865e-06, "loss": 0.5746, "step": 129 }, { "epoch": 0.10560519902518278, "grad_norm": 0.41738975048065186, "learning_rate": 3.513513513513514e-06, "loss": 0.5356, "step": 130 }, { "epoch": 0.10641754670999187, "grad_norm": 0.5605790019035339, "learning_rate": 3.5405405405405408e-06, "loss": 0.5611, "step": 131 }, { "epoch": 0.10722989439480098, "grad_norm": 0.44710150361061096, "learning_rate": 3.567567567567568e-06, "loss": 0.5637, "step": 132 }, { "epoch": 0.10804224207961008, "grad_norm": 0.47365838289260864, "learning_rate": 3.5945945945945946e-06, "loss": 0.5996, "step": 133 }, { "epoch": 0.10885458976441917, "grad_norm": 0.4175429344177246, "learning_rate": 3.621621621621622e-06, "loss": 0.6035, "step": 134 }, { "epoch": 0.10966693744922827, "grad_norm": 0.43969622254371643, "learning_rate": 3.648648648648649e-06, "loss": 0.5491, "step": 135 }, { "epoch": 0.11047928513403736, "grad_norm": 0.4338403344154358, "learning_rate": 3.6756756756756763e-06, "loss": 0.5882, "step": 136 }, { "epoch": 0.11129163281884646, "grad_norm": 0.43711960315704346, "learning_rate": 3.7027027027027028e-06, "loss": 0.553, "step": 137 }, { "epoch": 0.11210398050365557, "grad_norm": 0.4391087591648102, "learning_rate": 3.72972972972973e-06, "loss": 0.5662, "step": 138 }, { "epoch": 0.11291632818846466, "grad_norm": 0.4613192081451416, "learning_rate": 3.756756756756757e-06, "loss": 0.5763, "step": 139 }, { "epoch": 0.11372867587327376, "grad_norm": 0.45040860772132874, "learning_rate": 3.7837837837837844e-06, "loss": 0.5379, "step": 140 }, { "epoch": 0.11454102355808286, "grad_norm": 0.4877156615257263, "learning_rate": 3.810810810810811e-06, "loss": 0.5796, "step": 141 }, { "epoch": 0.11535337124289195, "grad_norm": 0.43945345282554626, "learning_rate": 3.837837837837838e-06, "loss": 0.563, "step": 142 }, { "epoch": 0.11616571892770106, "grad_norm": 0.44278138875961304, "learning_rate": 3.864864864864865e-06, "loss": 0.535, "step": 143 }, { "epoch": 0.11697806661251016, "grad_norm": 0.4262688159942627, "learning_rate": 3.891891891891892e-06, "loss": 0.5812, "step": 144 }, { "epoch": 0.11779041429731925, "grad_norm": 0.4685794711112976, "learning_rate": 3.918918918918919e-06, "loss": 0.5475, "step": 145 }, { "epoch": 0.11860276198212835, "grad_norm": 0.4233499765396118, "learning_rate": 3.945945945945947e-06, "loss": 0.549, "step": 146 }, { "epoch": 0.11941510966693746, "grad_norm": 0.3941769599914551, "learning_rate": 3.972972972972973e-06, "loss": 0.5451, "step": 147 }, { "epoch": 0.12022745735174654, "grad_norm": 0.4458985924720764, "learning_rate": 4.000000000000001e-06, "loss": 0.5551, "step": 148 }, { "epoch": 0.12103980503655565, "grad_norm": 0.45144122838974, "learning_rate": 4.027027027027028e-06, "loss": 0.5712, "step": 149 }, { "epoch": 0.12185215272136475, "grad_norm": 0.4355672597885132, "learning_rate": 4.0540540540540545e-06, "loss": 0.5672, "step": 150 }, { "epoch": 0.12266450040617384, "grad_norm": 0.4658735990524292, "learning_rate": 4.0810810810810815e-06, "loss": 0.5494, "step": 151 }, { "epoch": 0.12347684809098294, "grad_norm": 0.3950033187866211, "learning_rate": 4.108108108108108e-06, "loss": 0.5015, "step": 152 }, { "epoch": 0.12428919577579203, "grad_norm": 0.4416932761669159, "learning_rate": 4.135135135135135e-06, "loss": 0.5385, "step": 153 }, { "epoch": 0.12510154346060115, "grad_norm": 0.406793475151062, "learning_rate": 4.162162162162163e-06, "loss": 0.5491, "step": 154 }, { "epoch": 0.12591389114541024, "grad_norm": 0.4446122944355011, "learning_rate": 4.189189189189189e-06, "loss": 0.5574, "step": 155 }, { "epoch": 0.12672623883021933, "grad_norm": 0.4093543291091919, "learning_rate": 4.216216216216217e-06, "loss": 0.5414, "step": 156 }, { "epoch": 0.12753858651502845, "grad_norm": 0.4124000370502472, "learning_rate": 4.243243243243244e-06, "loss": 0.5607, "step": 157 }, { "epoch": 0.12835093419983754, "grad_norm": 0.425896555185318, "learning_rate": 4.270270270270271e-06, "loss": 0.5647, "step": 158 }, { "epoch": 0.12916328188464662, "grad_norm": 0.4369313716888428, "learning_rate": 4.297297297297298e-06, "loss": 0.5322, "step": 159 }, { "epoch": 0.12997562956945571, "grad_norm": 0.4157240688800812, "learning_rate": 4.324324324324325e-06, "loss": 0.5556, "step": 160 }, { "epoch": 0.13078797725426483, "grad_norm": 0.415233850479126, "learning_rate": 4.351351351351352e-06, "loss": 0.5185, "step": 161 }, { "epoch": 0.13160032493907392, "grad_norm": 0.44863107800483704, "learning_rate": 4.378378378378379e-06, "loss": 0.5382, "step": 162 }, { "epoch": 0.132412672623883, "grad_norm": 0.4747501313686371, "learning_rate": 4.4054054054054054e-06, "loss": 0.5477, "step": 163 }, { "epoch": 0.13322502030869213, "grad_norm": 0.45311713218688965, "learning_rate": 4.432432432432433e-06, "loss": 0.5551, "step": 164 }, { "epoch": 0.13403736799350122, "grad_norm": 0.39696723222732544, "learning_rate": 4.45945945945946e-06, "loss": 0.5445, "step": 165 }, { "epoch": 0.1348497156783103, "grad_norm": 0.4417300224304199, "learning_rate": 4.486486486486487e-06, "loss": 0.5668, "step": 166 }, { "epoch": 0.13566206336311942, "grad_norm": 0.4647722542285919, "learning_rate": 4.513513513513514e-06, "loss": 0.5514, "step": 167 }, { "epoch": 0.1364744110479285, "grad_norm": 0.40805771946907043, "learning_rate": 4.540540540540541e-06, "loss": 0.5313, "step": 168 }, { "epoch": 0.1372867587327376, "grad_norm": 0.452298641204834, "learning_rate": 4.567567567567568e-06, "loss": 0.5164, "step": 169 }, { "epoch": 0.13809910641754672, "grad_norm": 0.45936301350593567, "learning_rate": 4.594594594594596e-06, "loss": 0.5456, "step": 170 }, { "epoch": 0.1389114541023558, "grad_norm": 0.42853599786758423, "learning_rate": 4.621621621621622e-06, "loss": 0.5451, "step": 171 }, { "epoch": 0.1397238017871649, "grad_norm": 0.4440080225467682, "learning_rate": 4.6486486486486495e-06, "loss": 0.5706, "step": 172 }, { "epoch": 0.14053614947197401, "grad_norm": 0.445077508687973, "learning_rate": 4.675675675675676e-06, "loss": 0.5392, "step": 173 }, { "epoch": 0.1413484971567831, "grad_norm": 0.4439601004123688, "learning_rate": 4.702702702702703e-06, "loss": 0.507, "step": 174 }, { "epoch": 0.1421608448415922, "grad_norm": 0.45404112339019775, "learning_rate": 4.72972972972973e-06, "loss": 0.5467, "step": 175 }, { "epoch": 0.1429731925264013, "grad_norm": 0.45123425126075745, "learning_rate": 4.756756756756757e-06, "loss": 0.5635, "step": 176 }, { "epoch": 0.1437855402112104, "grad_norm": 0.4142909646034241, "learning_rate": 4.783783783783784e-06, "loss": 0.5115, "step": 177 }, { "epoch": 0.1445978878960195, "grad_norm": 0.42664262652397156, "learning_rate": 4.810810810810811e-06, "loss": 0.5397, "step": 178 }, { "epoch": 0.1454102355808286, "grad_norm": 0.4265083968639374, "learning_rate": 4.837837837837838e-06, "loss": 0.5523, "step": 179 }, { "epoch": 0.1462225832656377, "grad_norm": 0.4110284745693207, "learning_rate": 4.864864864864866e-06, "loss": 0.5449, "step": 180 }, { "epoch": 0.14703493095044678, "grad_norm": 0.4318018853664398, "learning_rate": 4.891891891891893e-06, "loss": 0.5468, "step": 181 }, { "epoch": 0.1478472786352559, "grad_norm": 0.40929874777793884, "learning_rate": 4.91891891891892e-06, "loss": 0.5322, "step": 182 }, { "epoch": 0.148659626320065, "grad_norm": 0.4436834454536438, "learning_rate": 4.9459459459459466e-06, "loss": 0.5635, "step": 183 }, { "epoch": 0.14947197400487408, "grad_norm": 0.44753170013427734, "learning_rate": 4.9729729729729735e-06, "loss": 0.5219, "step": 184 }, { "epoch": 0.1502843216896832, "grad_norm": 0.4664373993873596, "learning_rate": 5e-06, "loss": 0.5567, "step": 185 }, { "epoch": 0.1510966693744923, "grad_norm": 0.39065948128700256, "learning_rate": 5.027027027027027e-06, "loss": 0.515, "step": 186 }, { "epoch": 0.15190901705930138, "grad_norm": 0.3841733932495117, "learning_rate": 5.054054054054054e-06, "loss": 0.5053, "step": 187 }, { "epoch": 0.15272136474411047, "grad_norm": 0.4314366579055786, "learning_rate": 5.081081081081082e-06, "loss": 0.5445, "step": 188 }, { "epoch": 0.15353371242891958, "grad_norm": 0.4173581302165985, "learning_rate": 5.108108108108108e-06, "loss": 0.5241, "step": 189 }, { "epoch": 0.15434606011372867, "grad_norm": 0.465297669172287, "learning_rate": 5.135135135135135e-06, "loss": 0.5366, "step": 190 }, { "epoch": 0.15515840779853776, "grad_norm": 0.4381401538848877, "learning_rate": 5.162162162162162e-06, "loss": 0.5174, "step": 191 }, { "epoch": 0.15597075548334688, "grad_norm": 0.460330992937088, "learning_rate": 5.18918918918919e-06, "loss": 0.5452, "step": 192 }, { "epoch": 0.15678310316815597, "grad_norm": 0.45626774430274963, "learning_rate": 5.216216216216217e-06, "loss": 0.5595, "step": 193 }, { "epoch": 0.15759545085296506, "grad_norm": 0.4322197735309601, "learning_rate": 5.243243243243244e-06, "loss": 0.5532, "step": 194 }, { "epoch": 0.15840779853777417, "grad_norm": 0.4883195757865906, "learning_rate": 5.2702702702702705e-06, "loss": 0.5393, "step": 195 }, { "epoch": 0.15922014622258326, "grad_norm": 0.415217787027359, "learning_rate": 5.297297297297298e-06, "loss": 0.5106, "step": 196 }, { "epoch": 0.16003249390739235, "grad_norm": 0.42783263325691223, "learning_rate": 5.324324324324324e-06, "loss": 0.5385, "step": 197 }, { "epoch": 0.16084484159220147, "grad_norm": 0.4532291889190674, "learning_rate": 5.351351351351351e-06, "loss": 0.5429, "step": 198 }, { "epoch": 0.16165718927701056, "grad_norm": 0.407629132270813, "learning_rate": 5.378378378378378e-06, "loss": 0.5101, "step": 199 }, { "epoch": 0.16246953696181965, "grad_norm": 0.40582939982414246, "learning_rate": 5.405405405405406e-06, "loss": 0.5362, "step": 200 }, { "epoch": 0.16328188464662877, "grad_norm": 0.44061997532844543, "learning_rate": 5.432432432432433e-06, "loss": 0.5373, "step": 201 }, { "epoch": 0.16409423233143786, "grad_norm": 0.4864961504936218, "learning_rate": 5.45945945945946e-06, "loss": 0.4986, "step": 202 }, { "epoch": 0.16490658001624695, "grad_norm": 0.40400636196136475, "learning_rate": 5.486486486486487e-06, "loss": 0.5369, "step": 203 }, { "epoch": 0.16571892770105606, "grad_norm": 0.41523101925849915, "learning_rate": 5.513513513513515e-06, "loss": 0.5249, "step": 204 }, { "epoch": 0.16653127538586515, "grad_norm": 0.4292742609977722, "learning_rate": 5.540540540540541e-06, "loss": 0.5031, "step": 205 }, { "epoch": 0.16734362307067424, "grad_norm": 0.4152248203754425, "learning_rate": 5.567567567567568e-06, "loss": 0.5132, "step": 206 }, { "epoch": 0.16815597075548336, "grad_norm": 0.45014795660972595, "learning_rate": 5.5945945945945945e-06, "loss": 0.523, "step": 207 }, { "epoch": 0.16896831844029245, "grad_norm": 0.4987364411354065, "learning_rate": 5.621621621621622e-06, "loss": 0.541, "step": 208 }, { "epoch": 0.16978066612510154, "grad_norm": 0.4945250451564789, "learning_rate": 5.648648648648649e-06, "loss": 0.5323, "step": 209 }, { "epoch": 0.17059301380991065, "grad_norm": 0.4806767404079437, "learning_rate": 5.675675675675676e-06, "loss": 0.5194, "step": 210 }, { "epoch": 0.17140536149471974, "grad_norm": 0.505066990852356, "learning_rate": 5.702702702702702e-06, "loss": 0.5307, "step": 211 }, { "epoch": 0.17221770917952883, "grad_norm": 0.41895273327827454, "learning_rate": 5.729729729729731e-06, "loss": 0.5142, "step": 212 }, { "epoch": 0.17303005686433795, "grad_norm": 0.5012500882148743, "learning_rate": 5.756756756756757e-06, "loss": 0.544, "step": 213 }, { "epoch": 0.17384240454914704, "grad_norm": 0.4409078061580658, "learning_rate": 5.783783783783784e-06, "loss": 0.4924, "step": 214 }, { "epoch": 0.17465475223395613, "grad_norm": 0.4610885977745056, "learning_rate": 5.810810810810811e-06, "loss": 0.523, "step": 215 }, { "epoch": 0.17546709991876522, "grad_norm": 0.4463043808937073, "learning_rate": 5.837837837837839e-06, "loss": 0.5052, "step": 216 }, { "epoch": 0.17627944760357434, "grad_norm": 0.47048816084861755, "learning_rate": 5.8648648648648655e-06, "loss": 0.5232, "step": 217 }, { "epoch": 0.17709179528838342, "grad_norm": 0.4656911790370941, "learning_rate": 5.8918918918918924e-06, "loss": 0.541, "step": 218 }, { "epoch": 0.17790414297319251, "grad_norm": 0.4964292049407959, "learning_rate": 5.9189189189189185e-06, "loss": 0.5378, "step": 219 }, { "epoch": 0.17871649065800163, "grad_norm": 0.45170658826828003, "learning_rate": 5.945945945945947e-06, "loss": 0.5148, "step": 220 }, { "epoch": 0.17952883834281072, "grad_norm": 0.45852646231651306, "learning_rate": 5.972972972972973e-06, "loss": 0.5222, "step": 221 }, { "epoch": 0.1803411860276198, "grad_norm": 0.4935731887817383, "learning_rate": 6e-06, "loss": 0.5266, "step": 222 }, { "epoch": 0.18115353371242893, "grad_norm": 0.46777376532554626, "learning_rate": 6.027027027027027e-06, "loss": 0.5467, "step": 223 }, { "epoch": 0.18196588139723802, "grad_norm": 0.41773721575737, "learning_rate": 6.054054054054055e-06, "loss": 0.525, "step": 224 }, { "epoch": 0.1827782290820471, "grad_norm": 0.43071162700653076, "learning_rate": 6.081081081081082e-06, "loss": 0.5313, "step": 225 }, { "epoch": 0.18359057676685622, "grad_norm": 0.4422866702079773, "learning_rate": 6.108108108108109e-06, "loss": 0.5069, "step": 226 }, { "epoch": 0.1844029244516653, "grad_norm": 0.39626345038414, "learning_rate": 6.135135135135135e-06, "loss": 0.5097, "step": 227 }, { "epoch": 0.1852152721364744, "grad_norm": 0.45402950048446655, "learning_rate": 6.162162162162163e-06, "loss": 0.5185, "step": 228 }, { "epoch": 0.18602761982128352, "grad_norm": 0.4395464360713959, "learning_rate": 6.1891891891891895e-06, "loss": 0.5518, "step": 229 }, { "epoch": 0.1868399675060926, "grad_norm": 0.4242137372493744, "learning_rate": 6.2162162162162164e-06, "loss": 0.5096, "step": 230 }, { "epoch": 0.1876523151909017, "grad_norm": 0.4502149522304535, "learning_rate": 6.243243243243243e-06, "loss": 0.5428, "step": 231 }, { "epoch": 0.18846466287571081, "grad_norm": 0.44240760803222656, "learning_rate": 6.270270270270271e-06, "loss": 0.4939, "step": 232 }, { "epoch": 0.1892770105605199, "grad_norm": 0.45318982005119324, "learning_rate": 6.297297297297298e-06, "loss": 0.5068, "step": 233 }, { "epoch": 0.190089358245329, "grad_norm": 0.4861313998699188, "learning_rate": 6.324324324324325e-06, "loss": 0.518, "step": 234 }, { "epoch": 0.1909017059301381, "grad_norm": 0.5122928023338318, "learning_rate": 6.351351351351351e-06, "loss": 0.5654, "step": 235 }, { "epoch": 0.1917140536149472, "grad_norm": 0.5766184329986572, "learning_rate": 6.378378378378379e-06, "loss": 0.4916, "step": 236 }, { "epoch": 0.1925264012997563, "grad_norm": 0.45587393641471863, "learning_rate": 6.405405405405406e-06, "loss": 0.5061, "step": 237 }, { "epoch": 0.1933387489845654, "grad_norm": 0.45259344577789307, "learning_rate": 6.432432432432433e-06, "loss": 0.5182, "step": 238 }, { "epoch": 0.1941510966693745, "grad_norm": 0.5314270257949829, "learning_rate": 6.45945945945946e-06, "loss": 0.5132, "step": 239 }, { "epoch": 0.19496344435418358, "grad_norm": 0.4795337915420532, "learning_rate": 6.486486486486487e-06, "loss": 0.5553, "step": 240 }, { "epoch": 0.1957757920389927, "grad_norm": 0.47671452164649963, "learning_rate": 6.513513513513514e-06, "loss": 0.5133, "step": 241 }, { "epoch": 0.1965881397238018, "grad_norm": 0.5184851884841919, "learning_rate": 6.540540540540541e-06, "loss": 0.5311, "step": 242 }, { "epoch": 0.19740048740861088, "grad_norm": 0.47618207335472107, "learning_rate": 6.567567567567567e-06, "loss": 0.5064, "step": 243 }, { "epoch": 0.19821283509341997, "grad_norm": 0.46554797887802124, "learning_rate": 6.594594594594595e-06, "loss": 0.5436, "step": 244 }, { "epoch": 0.1990251827782291, "grad_norm": 0.48086822032928467, "learning_rate": 6.621621621621622e-06, "loss": 0.5312, "step": 245 }, { "epoch": 0.19983753046303818, "grad_norm": 0.4239889085292816, "learning_rate": 6.648648648648649e-06, "loss": 0.4979, "step": 246 }, { "epoch": 0.20064987814784727, "grad_norm": 0.44131985306739807, "learning_rate": 6.675675675675676e-06, "loss": 0.5391, "step": 247 }, { "epoch": 0.20146222583265638, "grad_norm": 0.5699193477630615, "learning_rate": 6.702702702702704e-06, "loss": 0.5461, "step": 248 }, { "epoch": 0.20227457351746547, "grad_norm": 0.548466682434082, "learning_rate": 6.729729729729731e-06, "loss": 0.4953, "step": 249 }, { "epoch": 0.20308692120227456, "grad_norm": 0.49565795063972473, "learning_rate": 6.7567567567567575e-06, "loss": 0.5129, "step": 250 }, { "epoch": 0.20389926888708368, "grad_norm": 0.5184242725372314, "learning_rate": 6.783783783783784e-06, "loss": 0.51, "step": 251 }, { "epoch": 0.20471161657189277, "grad_norm": 0.4637296497821808, "learning_rate": 6.810810810810811e-06, "loss": 0.5139, "step": 252 }, { "epoch": 0.20552396425670186, "grad_norm": 0.6081213355064392, "learning_rate": 6.837837837837838e-06, "loss": 0.55, "step": 253 }, { "epoch": 0.20633631194151097, "grad_norm": 0.4519492983818054, "learning_rate": 6.864864864864865e-06, "loss": 0.5202, "step": 254 }, { "epoch": 0.20714865962632006, "grad_norm": 0.4584764838218689, "learning_rate": 6.891891891891892e-06, "loss": 0.5111, "step": 255 }, { "epoch": 0.20796100731112915, "grad_norm": 0.43585702776908875, "learning_rate": 6.91891891891892e-06, "loss": 0.5063, "step": 256 }, { "epoch": 0.20877335499593827, "grad_norm": 0.43584078550338745, "learning_rate": 6.945945945945947e-06, "loss": 0.5179, "step": 257 }, { "epoch": 0.20958570268074736, "grad_norm": 0.4439784586429596, "learning_rate": 6.972972972972973e-06, "loss": 0.544, "step": 258 }, { "epoch": 0.21039805036555645, "grad_norm": 0.5131964087486267, "learning_rate": 7e-06, "loss": 0.5093, "step": 259 }, { "epoch": 0.21121039805036557, "grad_norm": 0.4393364191055298, "learning_rate": 7.027027027027028e-06, "loss": 0.4839, "step": 260 }, { "epoch": 0.21202274573517466, "grad_norm": 0.4557783305644989, "learning_rate": 7.054054054054055e-06, "loss": 0.5038, "step": 261 }, { "epoch": 0.21283509341998375, "grad_norm": 0.47186943888664246, "learning_rate": 7.0810810810810815e-06, "loss": 0.5397, "step": 262 }, { "epoch": 0.21364744110479286, "grad_norm": 0.5184400677680969, "learning_rate": 7.1081081081081085e-06, "loss": 0.5285, "step": 263 }, { "epoch": 0.21445978878960195, "grad_norm": 0.480281800031662, "learning_rate": 7.135135135135136e-06, "loss": 0.5353, "step": 264 }, { "epoch": 0.21527213647441104, "grad_norm": 0.6046870946884155, "learning_rate": 7.162162162162163e-06, "loss": 0.531, "step": 265 }, { "epoch": 0.21608448415922016, "grad_norm": 0.5679700374603271, "learning_rate": 7.189189189189189e-06, "loss": 0.5187, "step": 266 }, { "epoch": 0.21689683184402925, "grad_norm": 0.41307082772254944, "learning_rate": 7.216216216216216e-06, "loss": 0.5186, "step": 267 }, { "epoch": 0.21770917952883834, "grad_norm": 0.5387875437736511, "learning_rate": 7.243243243243244e-06, "loss": 0.5371, "step": 268 }, { "epoch": 0.21852152721364745, "grad_norm": 0.5477129220962524, "learning_rate": 7.270270270270271e-06, "loss": 0.5251, "step": 269 }, { "epoch": 0.21933387489845654, "grad_norm": 0.496522456407547, "learning_rate": 7.297297297297298e-06, "loss": 0.5041, "step": 270 }, { "epoch": 0.22014622258326563, "grad_norm": 0.5140697360038757, "learning_rate": 7.324324324324325e-06, "loss": 0.4899, "step": 271 }, { "epoch": 0.22095857026807472, "grad_norm": 0.48648494482040405, "learning_rate": 7.3513513513513525e-06, "loss": 0.5148, "step": 272 }, { "epoch": 0.22177091795288384, "grad_norm": 0.49726030230522156, "learning_rate": 7.3783783783783794e-06, "loss": 0.5208, "step": 273 }, { "epoch": 0.22258326563769293, "grad_norm": 0.5484408736228943, "learning_rate": 7.4054054054054055e-06, "loss": 0.5195, "step": 274 }, { "epoch": 0.22339561332250202, "grad_norm": 0.44089898467063904, "learning_rate": 7.4324324324324324e-06, "loss": 0.488, "step": 275 }, { "epoch": 0.22420796100731114, "grad_norm": 0.47092267870903015, "learning_rate": 7.45945945945946e-06, "loss": 0.507, "step": 276 }, { "epoch": 0.22502030869212022, "grad_norm": 0.5094614028930664, "learning_rate": 7.486486486486487e-06, "loss": 0.5227, "step": 277 }, { "epoch": 0.2258326563769293, "grad_norm": 0.4911332130432129, "learning_rate": 7.513513513513514e-06, "loss": 0.5237, "step": 278 }, { "epoch": 0.22664500406173843, "grad_norm": 0.45606639981269836, "learning_rate": 7.540540540540541e-06, "loss": 0.5015, "step": 279 }, { "epoch": 0.22745735174654752, "grad_norm": 0.5105320811271667, "learning_rate": 7.567567567567569e-06, "loss": 0.532, "step": 280 }, { "epoch": 0.2282696994313566, "grad_norm": 0.4273783564567566, "learning_rate": 7.594594594594596e-06, "loss": 0.547, "step": 281 }, { "epoch": 0.22908204711616573, "grad_norm": 0.49930238723754883, "learning_rate": 7.621621621621622e-06, "loss": 0.5341, "step": 282 }, { "epoch": 0.22989439480097482, "grad_norm": 0.44750887155532837, "learning_rate": 7.648648648648649e-06, "loss": 0.4777, "step": 283 }, { "epoch": 0.2307067424857839, "grad_norm": 0.45860055088996887, "learning_rate": 7.675675675675676e-06, "loss": 0.5329, "step": 284 }, { "epoch": 0.23151909017059302, "grad_norm": 0.558803379535675, "learning_rate": 7.702702702702704e-06, "loss": 0.5081, "step": 285 }, { "epoch": 0.2323314378554021, "grad_norm": 0.4594920873641968, "learning_rate": 7.72972972972973e-06, "loss": 0.5215, "step": 286 }, { "epoch": 0.2331437855402112, "grad_norm": 0.43717220425605774, "learning_rate": 7.756756756756756e-06, "loss": 0.5127, "step": 287 }, { "epoch": 0.23395613322502032, "grad_norm": 0.45339101552963257, "learning_rate": 7.783783783783784e-06, "loss": 0.5066, "step": 288 }, { "epoch": 0.2347684809098294, "grad_norm": 0.4890469014644623, "learning_rate": 7.810810810810812e-06, "loss": 0.5342, "step": 289 }, { "epoch": 0.2355808285946385, "grad_norm": 0.4327148497104645, "learning_rate": 7.837837837837838e-06, "loss": 0.5093, "step": 290 }, { "epoch": 0.23639317627944761, "grad_norm": 0.44367021322250366, "learning_rate": 7.864864864864866e-06, "loss": 0.522, "step": 291 }, { "epoch": 0.2372055239642567, "grad_norm": 0.48489734530448914, "learning_rate": 7.891891891891894e-06, "loss": 0.489, "step": 292 }, { "epoch": 0.2380178716490658, "grad_norm": 0.4957413673400879, "learning_rate": 7.91891891891892e-06, "loss": 0.5282, "step": 293 }, { "epoch": 0.2388302193338749, "grad_norm": 0.5173468589782715, "learning_rate": 7.945945945945946e-06, "loss": 0.4961, "step": 294 }, { "epoch": 0.239642567018684, "grad_norm": 0.4977986216545105, "learning_rate": 7.972972972972974e-06, "loss": 0.4805, "step": 295 }, { "epoch": 0.2404549147034931, "grad_norm": 0.4972385764122009, "learning_rate": 8.000000000000001e-06, "loss": 0.4871, "step": 296 }, { "epoch": 0.2412672623883022, "grad_norm": 0.4595426023006439, "learning_rate": 8.027027027027027e-06, "loss": 0.5118, "step": 297 }, { "epoch": 0.2420796100731113, "grad_norm": 0.51301509141922, "learning_rate": 8.054054054054055e-06, "loss": 0.5301, "step": 298 }, { "epoch": 0.24289195775792038, "grad_norm": 0.43193161487579346, "learning_rate": 8.081081081081081e-06, "loss": 0.4991, "step": 299 }, { "epoch": 0.2437043054427295, "grad_norm": 0.5170255899429321, "learning_rate": 8.108108108108109e-06, "loss": 0.4897, "step": 300 }, { "epoch": 0.2445166531275386, "grad_norm": 0.48581281304359436, "learning_rate": 8.135135135135137e-06, "loss": 0.5072, "step": 301 }, { "epoch": 0.24532900081234768, "grad_norm": 0.544249415397644, "learning_rate": 8.162162162162163e-06, "loss": 0.5119, "step": 302 }, { "epoch": 0.24614134849715677, "grad_norm": 0.5191234350204468, "learning_rate": 8.189189189189189e-06, "loss": 0.4963, "step": 303 }, { "epoch": 0.2469536961819659, "grad_norm": 0.5369517207145691, "learning_rate": 8.216216216216217e-06, "loss": 0.4981, "step": 304 }, { "epoch": 0.24776604386677498, "grad_norm": 0.5271973013877869, "learning_rate": 8.243243243243245e-06, "loss": 0.511, "step": 305 }, { "epoch": 0.24857839155158407, "grad_norm": 0.5241037607192993, "learning_rate": 8.27027027027027e-06, "loss": 0.493, "step": 306 }, { "epoch": 0.24939073923639318, "grad_norm": 0.5022989511489868, "learning_rate": 8.297297297297298e-06, "loss": 0.4912, "step": 307 }, { "epoch": 0.2502030869212023, "grad_norm": 0.5528043508529663, "learning_rate": 8.324324324324326e-06, "loss": 0.551, "step": 308 }, { "epoch": 0.2510154346060114, "grad_norm": 0.5083795189857483, "learning_rate": 8.351351351351352e-06, "loss": 0.4961, "step": 309 }, { "epoch": 0.2518277822908205, "grad_norm": 0.5453813672065735, "learning_rate": 8.378378378378378e-06, "loss": 0.5321, "step": 310 }, { "epoch": 0.25264012997562957, "grad_norm": 0.47830072045326233, "learning_rate": 8.405405405405406e-06, "loss": 0.5063, "step": 311 }, { "epoch": 0.25345247766043866, "grad_norm": 0.5236061811447144, "learning_rate": 8.432432432432434e-06, "loss": 0.4915, "step": 312 }, { "epoch": 0.25426482534524775, "grad_norm": 0.6346319913864136, "learning_rate": 8.45945945945946e-06, "loss": 0.5184, "step": 313 }, { "epoch": 0.2550771730300569, "grad_norm": 0.5335103869438171, "learning_rate": 8.486486486486488e-06, "loss": 0.4769, "step": 314 }, { "epoch": 0.255889520714866, "grad_norm": 0.6964415311813354, "learning_rate": 8.513513513513514e-06, "loss": 0.5375, "step": 315 }, { "epoch": 0.25670186839967507, "grad_norm": 0.6465139985084534, "learning_rate": 8.540540540540542e-06, "loss": 0.5194, "step": 316 }, { "epoch": 0.25751421608448416, "grad_norm": 0.7245939373970032, "learning_rate": 8.567567567567568e-06, "loss": 0.5416, "step": 317 }, { "epoch": 0.25832656376929325, "grad_norm": 0.62569659948349, "learning_rate": 8.594594594594595e-06, "loss": 0.4882, "step": 318 }, { "epoch": 0.25913891145410234, "grad_norm": 0.5987505912780762, "learning_rate": 8.621621621621622e-06, "loss": 0.5005, "step": 319 }, { "epoch": 0.25995125913891143, "grad_norm": 0.7048532962799072, "learning_rate": 8.64864864864865e-06, "loss": 0.5245, "step": 320 }, { "epoch": 0.2607636068237206, "grad_norm": 0.55645352602005, "learning_rate": 8.675675675675677e-06, "loss": 0.5242, "step": 321 }, { "epoch": 0.26157595450852966, "grad_norm": 0.6503564715385437, "learning_rate": 8.702702702702703e-06, "loss": 0.5308, "step": 322 }, { "epoch": 0.26238830219333875, "grad_norm": 0.5956227779388428, "learning_rate": 8.72972972972973e-06, "loss": 0.5338, "step": 323 }, { "epoch": 0.26320064987814784, "grad_norm": 0.5507960319519043, "learning_rate": 8.756756756756759e-06, "loss": 0.5064, "step": 324 }, { "epoch": 0.26401299756295693, "grad_norm": 0.5373188257217407, "learning_rate": 8.783783783783785e-06, "loss": 0.4802, "step": 325 }, { "epoch": 0.264825345247766, "grad_norm": 0.545616626739502, "learning_rate": 8.810810810810811e-06, "loss": 0.5139, "step": 326 }, { "epoch": 0.26563769293257516, "grad_norm": 0.5328928232192993, "learning_rate": 8.837837837837839e-06, "loss": 0.5255, "step": 327 }, { "epoch": 0.26645004061738425, "grad_norm": 0.5485355854034424, "learning_rate": 8.864864864864866e-06, "loss": 0.5099, "step": 328 }, { "epoch": 0.26726238830219334, "grad_norm": 0.4735546112060547, "learning_rate": 8.891891891891893e-06, "loss": 0.5243, "step": 329 }, { "epoch": 0.26807473598700243, "grad_norm": 0.48040878772735596, "learning_rate": 8.91891891891892e-06, "loss": 0.5137, "step": 330 }, { "epoch": 0.2688870836718115, "grad_norm": 0.5022470355033875, "learning_rate": 8.945945945945946e-06, "loss": 0.5193, "step": 331 }, { "epoch": 0.2696994313566206, "grad_norm": 0.5143060088157654, "learning_rate": 8.972972972972974e-06, "loss": 0.5192, "step": 332 }, { "epoch": 0.27051177904142976, "grad_norm": 0.4853283166885376, "learning_rate": 9e-06, "loss": 0.499, "step": 333 }, { "epoch": 0.27132412672623885, "grad_norm": 0.49053311347961426, "learning_rate": 9.027027027027028e-06, "loss": 0.4912, "step": 334 }, { "epoch": 0.27213647441104794, "grad_norm": 0.5140088200569153, "learning_rate": 9.054054054054054e-06, "loss": 0.4754, "step": 335 }, { "epoch": 0.272948822095857, "grad_norm": 0.5208844542503357, "learning_rate": 9.081081081081082e-06, "loss": 0.5183, "step": 336 }, { "epoch": 0.2737611697806661, "grad_norm": 0.5226290225982666, "learning_rate": 9.10810810810811e-06, "loss": 0.5159, "step": 337 }, { "epoch": 0.2745735174654752, "grad_norm": 0.4992032051086426, "learning_rate": 9.135135135135136e-06, "loss": 0.5033, "step": 338 }, { "epoch": 0.27538586515028435, "grad_norm": 0.5840100049972534, "learning_rate": 9.162162162162162e-06, "loss": 0.5229, "step": 339 }, { "epoch": 0.27619821283509344, "grad_norm": 0.48474419116973877, "learning_rate": 9.189189189189191e-06, "loss": 0.5273, "step": 340 }, { "epoch": 0.2770105605199025, "grad_norm": 0.575573742389679, "learning_rate": 9.216216216216217e-06, "loss": 0.5146, "step": 341 }, { "epoch": 0.2778229082047116, "grad_norm": 0.5253869295120239, "learning_rate": 9.243243243243243e-06, "loss": 0.5438, "step": 342 }, { "epoch": 0.2786352558895207, "grad_norm": 0.5685439109802246, "learning_rate": 9.270270270270271e-06, "loss": 0.521, "step": 343 }, { "epoch": 0.2794476035743298, "grad_norm": 0.5274525284767151, "learning_rate": 9.297297297297299e-06, "loss": 0.5182, "step": 344 }, { "epoch": 0.2802599512591389, "grad_norm": 0.48353204131126404, "learning_rate": 9.324324324324325e-06, "loss": 0.4987, "step": 345 }, { "epoch": 0.28107229894394803, "grad_norm": 0.5360022187232971, "learning_rate": 9.351351351351353e-06, "loss": 0.5081, "step": 346 }, { "epoch": 0.2818846466287571, "grad_norm": 0.4673972725868225, "learning_rate": 9.378378378378379e-06, "loss": 0.5428, "step": 347 }, { "epoch": 0.2826969943135662, "grad_norm": 0.5286619067192078, "learning_rate": 9.405405405405407e-06, "loss": 0.5231, "step": 348 }, { "epoch": 0.2835093419983753, "grad_norm": 0.4709078073501587, "learning_rate": 9.432432432432433e-06, "loss": 0.4846, "step": 349 }, { "epoch": 0.2843216896831844, "grad_norm": 0.5183911919593811, "learning_rate": 9.45945945945946e-06, "loss": 0.5202, "step": 350 }, { "epoch": 0.2851340373679935, "grad_norm": 0.5216012597084045, "learning_rate": 9.486486486486487e-06, "loss": 0.5, "step": 351 }, { "epoch": 0.2859463850528026, "grad_norm": 0.4999358057975769, "learning_rate": 9.513513513513514e-06, "loss": 0.4956, "step": 352 }, { "epoch": 0.2867587327376117, "grad_norm": 0.5682715177536011, "learning_rate": 9.540540540540542e-06, "loss": 0.5013, "step": 353 }, { "epoch": 0.2875710804224208, "grad_norm": 0.5426211953163147, "learning_rate": 9.567567567567568e-06, "loss": 0.5104, "step": 354 }, { "epoch": 0.2883834281072299, "grad_norm": 0.5210642218589783, "learning_rate": 9.594594594594594e-06, "loss": 0.4759, "step": 355 }, { "epoch": 0.289195775792039, "grad_norm": 0.5944691896438599, "learning_rate": 9.621621621621622e-06, "loss": 0.526, "step": 356 }, { "epoch": 0.29000812347684807, "grad_norm": 0.5534200668334961, "learning_rate": 9.64864864864865e-06, "loss": 0.4956, "step": 357 }, { "epoch": 0.2908204711616572, "grad_norm": 0.5512939691543579, "learning_rate": 9.675675675675676e-06, "loss": 0.4808, "step": 358 }, { "epoch": 0.2916328188464663, "grad_norm": 0.5946492552757263, "learning_rate": 9.702702702702704e-06, "loss": 0.5054, "step": 359 }, { "epoch": 0.2924451665312754, "grad_norm": 0.536542534828186, "learning_rate": 9.729729729729732e-06, "loss": 0.4785, "step": 360 }, { "epoch": 0.2932575142160845, "grad_norm": 0.7262011170387268, "learning_rate": 9.756756756756758e-06, "loss": 0.5309, "step": 361 }, { "epoch": 0.29406986190089357, "grad_norm": 0.5309736132621765, "learning_rate": 9.783783783783785e-06, "loss": 0.4775, "step": 362 }, { "epoch": 0.29488220958570266, "grad_norm": 0.5744659304618835, "learning_rate": 9.810810810810811e-06, "loss": 0.4868, "step": 363 }, { "epoch": 0.2956945572705118, "grad_norm": 0.629359781742096, "learning_rate": 9.83783783783784e-06, "loss": 0.4783, "step": 364 }, { "epoch": 0.2965069049553209, "grad_norm": 0.5779407024383545, "learning_rate": 9.864864864864865e-06, "loss": 0.5131, "step": 365 }, { "epoch": 0.29731925264013, "grad_norm": 0.5458950400352478, "learning_rate": 9.891891891891893e-06, "loss": 0.529, "step": 366 }, { "epoch": 0.2981316003249391, "grad_norm": 0.6969084143638611, "learning_rate": 9.91891891891892e-06, "loss": 0.5064, "step": 367 }, { "epoch": 0.29894394800974816, "grad_norm": 0.5985047817230225, "learning_rate": 9.945945945945947e-06, "loss": 0.493, "step": 368 }, { "epoch": 0.29975629569455725, "grad_norm": 0.5648998022079468, "learning_rate": 9.972972972972975e-06, "loss": 0.4863, "step": 369 }, { "epoch": 0.3005686433793664, "grad_norm": 0.5135365128517151, "learning_rate": 1e-05, "loss": 0.5008, "step": 370 }, { "epoch": 0.3013809910641755, "grad_norm": 0.5521730780601501, "learning_rate": 9.999997765506791e-06, "loss": 0.5119, "step": 371 }, { "epoch": 0.3021933387489846, "grad_norm": 0.5166078209877014, "learning_rate": 9.99999106202916e-06, "loss": 0.5226, "step": 372 }, { "epoch": 0.30300568643379366, "grad_norm": 0.49592727422714233, "learning_rate": 9.9999798895731e-06, "loss": 0.5421, "step": 373 }, { "epoch": 0.30381803411860275, "grad_norm": 0.5873650312423706, "learning_rate": 9.999964248148597e-06, "loss": 0.4694, "step": 374 }, { "epoch": 0.30463038180341184, "grad_norm": 0.49393805861473083, "learning_rate": 9.999944137769629e-06, "loss": 0.5204, "step": 375 }, { "epoch": 0.30544272948822093, "grad_norm": 0.5189183354377747, "learning_rate": 9.999919558454171e-06, "loss": 0.4708, "step": 376 }, { "epoch": 0.3062550771730301, "grad_norm": 0.4769141674041748, "learning_rate": 9.999890510224195e-06, "loss": 0.4769, "step": 377 }, { "epoch": 0.30706742485783917, "grad_norm": 0.48614591360092163, "learning_rate": 9.999856993105661e-06, "loss": 0.491, "step": 378 }, { "epoch": 0.30787977254264826, "grad_norm": 0.5852887034416199, "learning_rate": 9.999819007128527e-06, "loss": 0.523, "step": 379 }, { "epoch": 0.30869212022745735, "grad_norm": 0.5015835762023926, "learning_rate": 9.999776552326747e-06, "loss": 0.5088, "step": 380 }, { "epoch": 0.30950446791226643, "grad_norm": 0.5359316468238831, "learning_rate": 9.999729628738263e-06, "loss": 0.5076, "step": 381 }, { "epoch": 0.3103168155970755, "grad_norm": 0.4959254264831543, "learning_rate": 9.99967823640502e-06, "loss": 0.5096, "step": 382 }, { "epoch": 0.31112916328188467, "grad_norm": 0.5693923234939575, "learning_rate": 9.99962237537295e-06, "loss": 0.5131, "step": 383 }, { "epoch": 0.31194151096669376, "grad_norm": 0.49897825717926025, "learning_rate": 9.999562045691979e-06, "loss": 0.5212, "step": 384 }, { "epoch": 0.31275385865150285, "grad_norm": 0.46268120408058167, "learning_rate": 9.999497247416035e-06, "loss": 0.5214, "step": 385 }, { "epoch": 0.31356620633631194, "grad_norm": 0.5341930389404297, "learning_rate": 9.99942798060303e-06, "loss": 0.5167, "step": 386 }, { "epoch": 0.314378554021121, "grad_norm": 0.49887049198150635, "learning_rate": 9.999354245314874e-06, "loss": 0.4917, "step": 387 }, { "epoch": 0.3151909017059301, "grad_norm": 0.45434051752090454, "learning_rate": 9.999276041617477e-06, "loss": 0.4837, "step": 388 }, { "epoch": 0.31600324939073926, "grad_norm": 0.4670315384864807, "learning_rate": 9.999193369580729e-06, "loss": 0.4866, "step": 389 }, { "epoch": 0.31681559707554835, "grad_norm": 0.5040281414985657, "learning_rate": 9.999106229278531e-06, "loss": 0.5217, "step": 390 }, { "epoch": 0.31762794476035744, "grad_norm": 0.45644405484199524, "learning_rate": 9.99901462078876e-06, "loss": 0.5072, "step": 391 }, { "epoch": 0.31844029244516653, "grad_norm": 0.5106453895568848, "learning_rate": 9.998918544193303e-06, "loss": 0.5121, "step": 392 }, { "epoch": 0.3192526401299756, "grad_norm": 0.47852852940559387, "learning_rate": 9.998817999578028e-06, "loss": 0.5053, "step": 393 }, { "epoch": 0.3200649878147847, "grad_norm": 0.5151554346084595, "learning_rate": 9.998712987032803e-06, "loss": 0.5146, "step": 394 }, { "epoch": 0.32087733549959385, "grad_norm": 0.5008130669593811, "learning_rate": 9.99860350665149e-06, "loss": 0.4942, "step": 395 }, { "epoch": 0.32168968318440294, "grad_norm": 0.5116183161735535, "learning_rate": 9.99848955853194e-06, "loss": 0.4726, "step": 396 }, { "epoch": 0.32250203086921203, "grad_norm": 0.48568010330200195, "learning_rate": 9.998371142776e-06, "loss": 0.5281, "step": 397 }, { "epoch": 0.3233143785540211, "grad_norm": 0.5491635799407959, "learning_rate": 9.998248259489509e-06, "loss": 0.5395, "step": 398 }, { "epoch": 0.3241267262388302, "grad_norm": 0.5306912064552307, "learning_rate": 9.998120908782301e-06, "loss": 0.4865, "step": 399 }, { "epoch": 0.3249390739236393, "grad_norm": 0.4622975289821625, "learning_rate": 9.9979890907682e-06, "loss": 0.4676, "step": 400 }, { "epoch": 0.3257514216084484, "grad_norm": 0.4424881339073181, "learning_rate": 9.997852805565026e-06, "loss": 0.5339, "step": 401 }, { "epoch": 0.32656376929325753, "grad_norm": 0.5120671987533569, "learning_rate": 9.99771205329459e-06, "loss": 0.5092, "step": 402 }, { "epoch": 0.3273761169780666, "grad_norm": 0.49095618724823, "learning_rate": 9.997566834082697e-06, "loss": 0.4705, "step": 403 }, { "epoch": 0.3281884646628757, "grad_norm": 0.4544297456741333, "learning_rate": 9.997417148059142e-06, "loss": 0.4814, "step": 404 }, { "epoch": 0.3290008123476848, "grad_norm": 0.5018048882484436, "learning_rate": 9.997262995357714e-06, "loss": 0.5072, "step": 405 }, { "epoch": 0.3298131600324939, "grad_norm": 0.5260384678840637, "learning_rate": 9.997104376116195e-06, "loss": 0.5237, "step": 406 }, { "epoch": 0.330625507717303, "grad_norm": 0.4336162805557251, "learning_rate": 9.996941290476359e-06, "loss": 0.4919, "step": 407 }, { "epoch": 0.3314378554021121, "grad_norm": 0.5632445812225342, "learning_rate": 9.99677373858397e-06, "loss": 0.4865, "step": 408 }, { "epoch": 0.3322502030869212, "grad_norm": 0.47161349654197693, "learning_rate": 9.996601720588787e-06, "loss": 0.4861, "step": 409 }, { "epoch": 0.3330625507717303, "grad_norm": 0.4954816699028015, "learning_rate": 9.996425236644558e-06, "loss": 0.499, "step": 410 }, { "epoch": 0.3338748984565394, "grad_norm": 0.5218966603279114, "learning_rate": 9.996244286909022e-06, "loss": 0.4934, "step": 411 }, { "epoch": 0.3346872461413485, "grad_norm": 0.6493699550628662, "learning_rate": 9.996058871543917e-06, "loss": 0.5176, "step": 412 }, { "epoch": 0.33549959382615757, "grad_norm": 0.5400435328483582, "learning_rate": 9.995868990714963e-06, "loss": 0.4863, "step": 413 }, { "epoch": 0.3363119415109667, "grad_norm": 0.6386740207672119, "learning_rate": 9.995674644591873e-06, "loss": 0.4976, "step": 414 }, { "epoch": 0.3371242891957758, "grad_norm": 0.48391029238700867, "learning_rate": 9.995475833348359e-06, "loss": 0.4967, "step": 415 }, { "epoch": 0.3379366368805849, "grad_norm": 0.5291131734848022, "learning_rate": 9.99527255716211e-06, "loss": 0.4949, "step": 416 }, { "epoch": 0.338748984565394, "grad_norm": 0.5030757188796997, "learning_rate": 9.99506481621482e-06, "loss": 0.4999, "step": 417 }, { "epoch": 0.3395613322502031, "grad_norm": 0.514165997505188, "learning_rate": 9.994852610692166e-06, "loss": 0.4944, "step": 418 }, { "epoch": 0.34037367993501216, "grad_norm": 0.5174204707145691, "learning_rate": 9.994635940783816e-06, "loss": 0.4805, "step": 419 }, { "epoch": 0.3411860276198213, "grad_norm": 0.5297481417655945, "learning_rate": 9.994414806683429e-06, "loss": 0.4921, "step": 420 }, { "epoch": 0.3419983753046304, "grad_norm": 0.5550931692123413, "learning_rate": 9.994189208588655e-06, "loss": 0.4705, "step": 421 }, { "epoch": 0.3428107229894395, "grad_norm": 0.5813543200492859, "learning_rate": 9.993959146701128e-06, "loss": 0.5151, "step": 422 }, { "epoch": 0.3436230706742486, "grad_norm": 0.581155002117157, "learning_rate": 9.993724621226484e-06, "loss": 0.5258, "step": 423 }, { "epoch": 0.34443541835905767, "grad_norm": 0.5383273363113403, "learning_rate": 9.993485632374338e-06, "loss": 0.513, "step": 424 }, { "epoch": 0.34524776604386676, "grad_norm": 0.4853130578994751, "learning_rate": 9.993242180358298e-06, "loss": 0.5066, "step": 425 }, { "epoch": 0.3460601137286759, "grad_norm": 0.5546746850013733, "learning_rate": 9.992994265395959e-06, "loss": 0.5198, "step": 426 }, { "epoch": 0.346872461413485, "grad_norm": 0.5029754638671875, "learning_rate": 9.992741887708908e-06, "loss": 0.5026, "step": 427 }, { "epoch": 0.3476848090982941, "grad_norm": 0.5960217714309692, "learning_rate": 9.99248504752272e-06, "loss": 0.4892, "step": 428 }, { "epoch": 0.34849715678310317, "grad_norm": 0.4362775683403015, "learning_rate": 9.992223745066959e-06, "loss": 0.4922, "step": 429 }, { "epoch": 0.34930950446791226, "grad_norm": 0.5453248023986816, "learning_rate": 9.991957980575172e-06, "loss": 0.4988, "step": 430 }, { "epoch": 0.35012185215272135, "grad_norm": 0.4515749514102936, "learning_rate": 9.991687754284904e-06, "loss": 0.4995, "step": 431 }, { "epoch": 0.35093419983753044, "grad_norm": 0.5404447913169861, "learning_rate": 9.99141306643768e-06, "loss": 0.5049, "step": 432 }, { "epoch": 0.3517465475223396, "grad_norm": 0.5164476633071899, "learning_rate": 9.991133917279013e-06, "loss": 0.4987, "step": 433 }, { "epoch": 0.35255889520714867, "grad_norm": 0.48054975271224976, "learning_rate": 9.990850307058411e-06, "loss": 0.4951, "step": 434 }, { "epoch": 0.35337124289195776, "grad_norm": 0.46658679842948914, "learning_rate": 9.99056223602936e-06, "loss": 0.51, "step": 435 }, { "epoch": 0.35418359057676685, "grad_norm": 0.5350185632705688, "learning_rate": 9.990269704449338e-06, "loss": 0.4972, "step": 436 }, { "epoch": 0.35499593826157594, "grad_norm": 0.4943006634712219, "learning_rate": 9.989972712579809e-06, "loss": 0.5131, "step": 437 }, { "epoch": 0.35580828594638503, "grad_norm": 0.5436580777168274, "learning_rate": 9.989671260686225e-06, "loss": 0.4959, "step": 438 }, { "epoch": 0.3566206336311942, "grad_norm": 0.4289551079273224, "learning_rate": 9.989365349038019e-06, "loss": 0.4881, "step": 439 }, { "epoch": 0.35743298131600326, "grad_norm": 0.49620842933654785, "learning_rate": 9.98905497790862e-06, "loss": 0.4859, "step": 440 }, { "epoch": 0.35824532900081235, "grad_norm": 0.5282759070396423, "learning_rate": 9.98874014757543e-06, "loss": 0.4918, "step": 441 }, { "epoch": 0.35905767668562144, "grad_norm": 0.41960445046424866, "learning_rate": 9.98842085831985e-06, "loss": 0.4782, "step": 442 }, { "epoch": 0.35987002437043053, "grad_norm": 0.4911501109600067, "learning_rate": 9.988097110427255e-06, "loss": 0.5329, "step": 443 }, { "epoch": 0.3606823720552396, "grad_norm": 0.514127254486084, "learning_rate": 9.987768904187014e-06, "loss": 0.4866, "step": 444 }, { "epoch": 0.36149471974004876, "grad_norm": 0.43918830156326294, "learning_rate": 9.987436239892472e-06, "loss": 0.4737, "step": 445 }, { "epoch": 0.36230706742485785, "grad_norm": 0.4967946410179138, "learning_rate": 9.987099117840969e-06, "loss": 0.4804, "step": 446 }, { "epoch": 0.36311941510966694, "grad_norm": 0.5516065955162048, "learning_rate": 9.986757538333817e-06, "loss": 0.4958, "step": 447 }, { "epoch": 0.36393176279447603, "grad_norm": 0.4820033311843872, "learning_rate": 9.986411501676327e-06, "loss": 0.5014, "step": 448 }, { "epoch": 0.3647441104792851, "grad_norm": 0.5718366503715515, "learning_rate": 9.986061008177779e-06, "loss": 0.5222, "step": 449 }, { "epoch": 0.3655564581640942, "grad_norm": 0.5417718887329102, "learning_rate": 9.985706058151446e-06, "loss": 0.5098, "step": 450 }, { "epoch": 0.36636880584890336, "grad_norm": 0.45447593927383423, "learning_rate": 9.985346651914581e-06, "loss": 0.4886, "step": 451 }, { "epoch": 0.36718115353371245, "grad_norm": 0.5571296215057373, "learning_rate": 9.98498278978842e-06, "loss": 0.4917, "step": 452 }, { "epoch": 0.36799350121852153, "grad_norm": 0.5669398307800293, "learning_rate": 9.984614472098181e-06, "loss": 0.5277, "step": 453 }, { "epoch": 0.3688058489033306, "grad_norm": 0.5737721920013428, "learning_rate": 9.984241699173069e-06, "loss": 0.5022, "step": 454 }, { "epoch": 0.3696181965881397, "grad_norm": 0.4828628897666931, "learning_rate": 9.983864471346263e-06, "loss": 0.5014, "step": 455 }, { "epoch": 0.3704305442729488, "grad_norm": 0.5983983278274536, "learning_rate": 9.983482788954931e-06, "loss": 0.4822, "step": 456 }, { "epoch": 0.37124289195775795, "grad_norm": 0.555889904499054, "learning_rate": 9.983096652340219e-06, "loss": 0.4897, "step": 457 }, { "epoch": 0.37205523964256704, "grad_norm": 0.5192045569419861, "learning_rate": 9.982706061847254e-06, "loss": 0.5043, "step": 458 }, { "epoch": 0.3728675873273761, "grad_norm": 0.6402643918991089, "learning_rate": 9.982311017825145e-06, "loss": 0.4694, "step": 459 }, { "epoch": 0.3736799350121852, "grad_norm": 0.5620437264442444, "learning_rate": 9.981911520626983e-06, "loss": 0.5058, "step": 460 }, { "epoch": 0.3744922826969943, "grad_norm": 0.5074028372764587, "learning_rate": 9.981507570609837e-06, "loss": 0.5001, "step": 461 }, { "epoch": 0.3753046303818034, "grad_norm": 0.5687691569328308, "learning_rate": 9.981099168134753e-06, "loss": 0.4963, "step": 462 }, { "epoch": 0.3761169780666125, "grad_norm": 0.5100858211517334, "learning_rate": 9.980686313566765e-06, "loss": 0.4913, "step": 463 }, { "epoch": 0.37692932575142163, "grad_norm": 0.4984515309333801, "learning_rate": 9.980269007274878e-06, "loss": 0.5059, "step": 464 }, { "epoch": 0.3777416734362307, "grad_norm": 0.49001407623291016, "learning_rate": 9.97984724963208e-06, "loss": 0.5053, "step": 465 }, { "epoch": 0.3785540211210398, "grad_norm": 0.5693202614784241, "learning_rate": 9.979421041015336e-06, "loss": 0.4776, "step": 466 }, { "epoch": 0.3793663688058489, "grad_norm": 0.5827487111091614, "learning_rate": 9.978990381805593e-06, "loss": 0.5039, "step": 467 }, { "epoch": 0.380178716490658, "grad_norm": 0.5298969149589539, "learning_rate": 9.978555272387771e-06, "loss": 0.5019, "step": 468 }, { "epoch": 0.3809910641754671, "grad_norm": 0.5810168385505676, "learning_rate": 9.978115713150768e-06, "loss": 0.4969, "step": 469 }, { "epoch": 0.3818034118602762, "grad_norm": 0.46066102385520935, "learning_rate": 9.977671704487465e-06, "loss": 0.5029, "step": 470 }, { "epoch": 0.3826157595450853, "grad_norm": 0.5538294911384583, "learning_rate": 9.977223246794712e-06, "loss": 0.4979, "step": 471 }, { "epoch": 0.3834281072298944, "grad_norm": 0.48471859097480774, "learning_rate": 9.97677034047334e-06, "loss": 0.5121, "step": 472 }, { "epoch": 0.3842404549147035, "grad_norm": 0.43323421478271484, "learning_rate": 9.976312985928159e-06, "loss": 0.4904, "step": 473 }, { "epoch": 0.3850528025995126, "grad_norm": 0.47097212076187134, "learning_rate": 9.975851183567945e-06, "loss": 0.4811, "step": 474 }, { "epoch": 0.38586515028432167, "grad_norm": 0.5560442209243774, "learning_rate": 9.975384933805461e-06, "loss": 0.5056, "step": 475 }, { "epoch": 0.3866774979691308, "grad_norm": 0.43062761425971985, "learning_rate": 9.974914237057435e-06, "loss": 0.5, "step": 476 }, { "epoch": 0.3874898456539399, "grad_norm": 0.5999177098274231, "learning_rate": 9.974439093744581e-06, "loss": 0.5237, "step": 477 }, { "epoch": 0.388302193338749, "grad_norm": 0.4601180851459503, "learning_rate": 9.973959504291574e-06, "loss": 0.4577, "step": 478 }, { "epoch": 0.3891145410235581, "grad_norm": 0.49394530057907104, "learning_rate": 9.973475469127074e-06, "loss": 0.4824, "step": 479 }, { "epoch": 0.38992688870836717, "grad_norm": 0.5397387146949768, "learning_rate": 9.972986988683711e-06, "loss": 0.5012, "step": 480 }, { "epoch": 0.39073923639317626, "grad_norm": 0.48552611470222473, "learning_rate": 9.972494063398083e-06, "loss": 0.4465, "step": 481 }, { "epoch": 0.3915515840779854, "grad_norm": 0.6346990466117859, "learning_rate": 9.971996693710768e-06, "loss": 0.4982, "step": 482 }, { "epoch": 0.3923639317627945, "grad_norm": 0.5586808919906616, "learning_rate": 9.971494880066315e-06, "loss": 0.5197, "step": 483 }, { "epoch": 0.3931762794476036, "grad_norm": 0.6436161398887634, "learning_rate": 9.97098862291324e-06, "loss": 0.4916, "step": 484 }, { "epoch": 0.39398862713241267, "grad_norm": 0.559062659740448, "learning_rate": 9.97047792270404e-06, "loss": 0.5095, "step": 485 }, { "epoch": 0.39480097481722176, "grad_norm": 0.5830823183059692, "learning_rate": 9.969962779895172e-06, "loss": 0.468, "step": 486 }, { "epoch": 0.39561332250203085, "grad_norm": 0.5965030789375305, "learning_rate": 9.96944319494707e-06, "loss": 0.4775, "step": 487 }, { "epoch": 0.39642567018683994, "grad_norm": 0.43599894642829895, "learning_rate": 9.96891916832414e-06, "loss": 0.487, "step": 488 }, { "epoch": 0.3972380178716491, "grad_norm": 0.6391028165817261, "learning_rate": 9.968390700494755e-06, "loss": 0.4762, "step": 489 }, { "epoch": 0.3980503655564582, "grad_norm": 0.5033403038978577, "learning_rate": 9.967857791931257e-06, "loss": 0.4693, "step": 490 }, { "epoch": 0.39886271324126726, "grad_norm": 0.49152690172195435, "learning_rate": 9.967320443109958e-06, "loss": 0.4839, "step": 491 }, { "epoch": 0.39967506092607635, "grad_norm": 0.6043562293052673, "learning_rate": 9.966778654511143e-06, "loss": 0.4826, "step": 492 }, { "epoch": 0.40048740861088544, "grad_norm": 0.4664207994937897, "learning_rate": 9.966232426619054e-06, "loss": 0.507, "step": 493 }, { "epoch": 0.40129975629569453, "grad_norm": 0.5082194209098816, "learning_rate": 9.965681759921912e-06, "loss": 0.4995, "step": 494 }, { "epoch": 0.4021121039805037, "grad_norm": 0.533515989780426, "learning_rate": 9.965126654911904e-06, "loss": 0.5112, "step": 495 }, { "epoch": 0.40292445166531277, "grad_norm": 0.5167580842971802, "learning_rate": 9.964567112085175e-06, "loss": 0.5051, "step": 496 }, { "epoch": 0.40373679935012186, "grad_norm": 0.5313193202018738, "learning_rate": 9.96400313194185e-06, "loss": 0.4613, "step": 497 }, { "epoch": 0.40454914703493094, "grad_norm": 0.5440222024917603, "learning_rate": 9.963434714986006e-06, "loss": 0.4901, "step": 498 }, { "epoch": 0.40536149471974003, "grad_norm": 0.4967605471611023, "learning_rate": 9.962861861725698e-06, "loss": 0.4649, "step": 499 }, { "epoch": 0.4061738424045491, "grad_norm": 0.5483538508415222, "learning_rate": 9.96228457267294e-06, "loss": 0.4956, "step": 500 }, { "epoch": 0.40698619008935827, "grad_norm": 0.5170942544937134, "learning_rate": 9.961702848343709e-06, "loss": 0.4722, "step": 501 }, { "epoch": 0.40779853777416736, "grad_norm": 0.510551393032074, "learning_rate": 9.961116689257949e-06, "loss": 0.4995, "step": 502 }, { "epoch": 0.40861088545897645, "grad_norm": 0.4607028663158417, "learning_rate": 9.960526095939566e-06, "loss": 0.505, "step": 503 }, { "epoch": 0.40942323314378554, "grad_norm": 0.5020343065261841, "learning_rate": 9.959931068916436e-06, "loss": 0.4858, "step": 504 }, { "epoch": 0.4102355808285946, "grad_norm": 0.451518177986145, "learning_rate": 9.959331608720386e-06, "loss": 0.5006, "step": 505 }, { "epoch": 0.4110479285134037, "grad_norm": 0.5158300399780273, "learning_rate": 9.958727715887218e-06, "loss": 0.4912, "step": 506 }, { "epoch": 0.41186027619821286, "grad_norm": 0.5170383453369141, "learning_rate": 9.958119390956685e-06, "loss": 0.4813, "step": 507 }, { "epoch": 0.41267262388302195, "grad_norm": 0.4723794758319855, "learning_rate": 9.95750663447251e-06, "loss": 0.5033, "step": 508 }, { "epoch": 0.41348497156783104, "grad_norm": 0.5372331142425537, "learning_rate": 9.956889446982369e-06, "loss": 0.4815, "step": 509 }, { "epoch": 0.41429731925264013, "grad_norm": 0.6045517325401306, "learning_rate": 9.956267829037906e-06, "loss": 0.491, "step": 510 }, { "epoch": 0.4151096669374492, "grad_norm": 0.505850613117218, "learning_rate": 9.95564178119472e-06, "loss": 0.5015, "step": 511 }, { "epoch": 0.4159220146222583, "grad_norm": 0.6147093772888184, "learning_rate": 9.95501130401237e-06, "loss": 0.4672, "step": 512 }, { "epoch": 0.41673436230706745, "grad_norm": 0.5220261216163635, "learning_rate": 9.954376398054377e-06, "loss": 0.4767, "step": 513 }, { "epoch": 0.41754670999187654, "grad_norm": 0.5472069978713989, "learning_rate": 9.953737063888216e-06, "loss": 0.4491, "step": 514 }, { "epoch": 0.41835905767668563, "grad_norm": 0.49514976143836975, "learning_rate": 9.953093302085324e-06, "loss": 0.483, "step": 515 }, { "epoch": 0.4191714053614947, "grad_norm": 0.565861701965332, "learning_rate": 9.952445113221093e-06, "loss": 0.5018, "step": 516 }, { "epoch": 0.4199837530463038, "grad_norm": 0.5246108174324036, "learning_rate": 9.95179249787487e-06, "loss": 0.5023, "step": 517 }, { "epoch": 0.4207961007311129, "grad_norm": 0.5922960042953491, "learning_rate": 9.951135456629966e-06, "loss": 0.4836, "step": 518 }, { "epoch": 0.421608448415922, "grad_norm": 0.5905640125274658, "learning_rate": 9.950473990073637e-06, "loss": 0.4899, "step": 519 }, { "epoch": 0.42242079610073113, "grad_norm": 0.6583929061889648, "learning_rate": 9.949808098797104e-06, "loss": 0.4987, "step": 520 }, { "epoch": 0.4232331437855402, "grad_norm": 0.508631706237793, "learning_rate": 9.949137783395537e-06, "loss": 0.5159, "step": 521 }, { "epoch": 0.4240454914703493, "grad_norm": 0.6548805236816406, "learning_rate": 9.948463044468063e-06, "loss": 0.4804, "step": 522 }, { "epoch": 0.4248578391551584, "grad_norm": 0.4542087912559509, "learning_rate": 9.94778388261776e-06, "loss": 0.4803, "step": 523 }, { "epoch": 0.4256701868399675, "grad_norm": 0.5409389734268188, "learning_rate": 9.947100298451663e-06, "loss": 0.4843, "step": 524 }, { "epoch": 0.4264825345247766, "grad_norm": 0.48888352513313293, "learning_rate": 9.946412292580757e-06, "loss": 0.4782, "step": 525 }, { "epoch": 0.4272948822095857, "grad_norm": 0.47382256388664246, "learning_rate": 9.94571986561998e-06, "loss": 0.4785, "step": 526 }, { "epoch": 0.4281072298943948, "grad_norm": 0.5466932058334351, "learning_rate": 9.945023018188222e-06, "loss": 0.4776, "step": 527 }, { "epoch": 0.4289195775792039, "grad_norm": 0.5342962145805359, "learning_rate": 9.944321750908321e-06, "loss": 0.5118, "step": 528 }, { "epoch": 0.429731925264013, "grad_norm": 0.488370805978775, "learning_rate": 9.94361606440707e-06, "loss": 0.526, "step": 529 }, { "epoch": 0.4305442729488221, "grad_norm": 0.47946813702583313, "learning_rate": 9.94290595931521e-06, "loss": 0.4782, "step": 530 }, { "epoch": 0.43135662063363117, "grad_norm": 0.45012009143829346, "learning_rate": 9.942191436267428e-06, "loss": 0.4913, "step": 531 }, { "epoch": 0.4321689683184403, "grad_norm": 0.5256201028823853, "learning_rate": 9.941472495902366e-06, "loss": 0.5257, "step": 532 }, { "epoch": 0.4329813160032494, "grad_norm": 0.48289409279823303, "learning_rate": 9.940749138862609e-06, "loss": 0.5141, "step": 533 }, { "epoch": 0.4337936636880585, "grad_norm": 0.5594226121902466, "learning_rate": 9.940021365794692e-06, "loss": 0.4803, "step": 534 }, { "epoch": 0.4346060113728676, "grad_norm": 0.4350510835647583, "learning_rate": 9.939289177349097e-06, "loss": 0.4807, "step": 535 }, { "epoch": 0.4354183590576767, "grad_norm": 0.45846202969551086, "learning_rate": 9.938552574180252e-06, "loss": 0.5102, "step": 536 }, { "epoch": 0.43623070674248576, "grad_norm": 0.48883935809135437, "learning_rate": 9.937811556946531e-06, "loss": 0.4995, "step": 537 }, { "epoch": 0.4370430544272949, "grad_norm": 0.4503771960735321, "learning_rate": 9.937066126310253e-06, "loss": 0.5122, "step": 538 }, { "epoch": 0.437855402112104, "grad_norm": 0.45419925451278687, "learning_rate": 9.936316282937682e-06, "loss": 0.4687, "step": 539 }, { "epoch": 0.4386677497969131, "grad_norm": 0.4319947361946106, "learning_rate": 9.935562027499026e-06, "loss": 0.4852, "step": 540 }, { "epoch": 0.4394800974817222, "grad_norm": 0.492436945438385, "learning_rate": 9.934803360668435e-06, "loss": 0.4971, "step": 541 }, { "epoch": 0.44029244516653127, "grad_norm": 0.5053468346595764, "learning_rate": 9.934040283124006e-06, "loss": 0.4551, "step": 542 }, { "epoch": 0.44110479285134035, "grad_norm": 0.5641597509384155, "learning_rate": 9.933272795547773e-06, "loss": 0.4665, "step": 543 }, { "epoch": 0.44191714053614944, "grad_norm": 0.5546810626983643, "learning_rate": 9.932500898625716e-06, "loss": 0.5135, "step": 544 }, { "epoch": 0.4427294882209586, "grad_norm": 0.6281055212020874, "learning_rate": 9.931724593047754e-06, "loss": 0.518, "step": 545 }, { "epoch": 0.4435418359057677, "grad_norm": 0.5220004320144653, "learning_rate": 9.930943879507748e-06, "loss": 0.4879, "step": 546 }, { "epoch": 0.44435418359057677, "grad_norm": 0.5242013335227966, "learning_rate": 9.930158758703495e-06, "loss": 0.4685, "step": 547 }, { "epoch": 0.44516653127538586, "grad_norm": 0.5622126460075378, "learning_rate": 9.929369231336735e-06, "loss": 0.4784, "step": 548 }, { "epoch": 0.44597887896019495, "grad_norm": 0.47500213980674744, "learning_rate": 9.928575298113146e-06, "loss": 0.4608, "step": 549 }, { "epoch": 0.44679122664500404, "grad_norm": 0.4946683943271637, "learning_rate": 9.927776959742344e-06, "loss": 0.4762, "step": 550 }, { "epoch": 0.4476035743298132, "grad_norm": 0.5434373617172241, "learning_rate": 9.92697421693788e-06, "loss": 0.4848, "step": 551 }, { "epoch": 0.44841592201462227, "grad_norm": 0.46261635422706604, "learning_rate": 9.926167070417243e-06, "loss": 0.4823, "step": 552 }, { "epoch": 0.44922826969943136, "grad_norm": 0.5457891821861267, "learning_rate": 9.925355520901861e-06, "loss": 0.4737, "step": 553 }, { "epoch": 0.45004061738424045, "grad_norm": 0.5264912247657776, "learning_rate": 9.924539569117092e-06, "loss": 0.4881, "step": 554 }, { "epoch": 0.45085296506904954, "grad_norm": 0.5739470720291138, "learning_rate": 9.923719215792233e-06, "loss": 0.4863, "step": 555 }, { "epoch": 0.4516653127538586, "grad_norm": 0.5150482058525085, "learning_rate": 9.922894461660513e-06, "loss": 0.4635, "step": 556 }, { "epoch": 0.4524776604386678, "grad_norm": 0.47487354278564453, "learning_rate": 9.922065307459096e-06, "loss": 0.4738, "step": 557 }, { "epoch": 0.45329000812347686, "grad_norm": 0.540755033493042, "learning_rate": 9.921231753929077e-06, "loss": 0.4934, "step": 558 }, { "epoch": 0.45410235580828595, "grad_norm": 0.5300335884094238, "learning_rate": 9.920393801815482e-06, "loss": 0.4936, "step": 559 }, { "epoch": 0.45491470349309504, "grad_norm": 0.5167898535728455, "learning_rate": 9.919551451867275e-06, "loss": 0.4782, "step": 560 }, { "epoch": 0.45572705117790413, "grad_norm": 0.49661946296691895, "learning_rate": 9.91870470483734e-06, "loss": 0.4855, "step": 561 }, { "epoch": 0.4565393988627132, "grad_norm": 0.5396080613136292, "learning_rate": 9.917853561482503e-06, "loss": 0.5012, "step": 562 }, { "epoch": 0.45735174654752236, "grad_norm": 0.4822172224521637, "learning_rate": 9.91699802256351e-06, "loss": 0.481, "step": 563 }, { "epoch": 0.45816409423233145, "grad_norm": 0.4852050542831421, "learning_rate": 9.916138088845042e-06, "loss": 0.4689, "step": 564 }, { "epoch": 0.45897644191714054, "grad_norm": 0.5295907855033875, "learning_rate": 9.915273761095701e-06, "loss": 0.4743, "step": 565 }, { "epoch": 0.45978878960194963, "grad_norm": 0.523033618927002, "learning_rate": 9.914405040088026e-06, "loss": 0.4816, "step": 566 }, { "epoch": 0.4606011372867587, "grad_norm": 0.5421017408370972, "learning_rate": 9.913531926598475e-06, "loss": 0.4883, "step": 567 }, { "epoch": 0.4614134849715678, "grad_norm": 0.4954166114330292, "learning_rate": 9.912654421407434e-06, "loss": 0.5007, "step": 568 }, { "epoch": 0.46222583265637696, "grad_norm": 0.5351802706718445, "learning_rate": 9.911772525299214e-06, "loss": 0.4997, "step": 569 }, { "epoch": 0.46303818034118605, "grad_norm": 0.581425666809082, "learning_rate": 9.910886239062054e-06, "loss": 0.5185, "step": 570 }, { "epoch": 0.46385052802599513, "grad_norm": 0.5007893443107605, "learning_rate": 9.909995563488116e-06, "loss": 0.505, "step": 571 }, { "epoch": 0.4646628757108042, "grad_norm": 0.49192190170288086, "learning_rate": 9.909100499373476e-06, "loss": 0.4626, "step": 572 }, { "epoch": 0.4654752233956133, "grad_norm": 0.5121611952781677, "learning_rate": 9.908201047518146e-06, "loss": 0.4804, "step": 573 }, { "epoch": 0.4662875710804224, "grad_norm": 0.4634596109390259, "learning_rate": 9.907297208726051e-06, "loss": 0.4977, "step": 574 }, { "epoch": 0.4670999187652315, "grad_norm": 0.5161271691322327, "learning_rate": 9.90638898380504e-06, "loss": 0.4866, "step": 575 }, { "epoch": 0.46791226645004064, "grad_norm": 0.56253981590271, "learning_rate": 9.905476373566884e-06, "loss": 0.5003, "step": 576 }, { "epoch": 0.4687246141348497, "grad_norm": 0.4513951241970062, "learning_rate": 9.90455937882727e-06, "loss": 0.4689, "step": 577 }, { "epoch": 0.4695369618196588, "grad_norm": 0.5390597581863403, "learning_rate": 9.903638000405805e-06, "loss": 0.5076, "step": 578 }, { "epoch": 0.4703493095044679, "grad_norm": 0.44172969460487366, "learning_rate": 9.902712239126016e-06, "loss": 0.4632, "step": 579 }, { "epoch": 0.471161657189277, "grad_norm": 0.5002825260162354, "learning_rate": 9.901782095815343e-06, "loss": 0.4785, "step": 580 }, { "epoch": 0.4719740048740861, "grad_norm": 0.5058497786521912, "learning_rate": 9.900847571305148e-06, "loss": 0.4649, "step": 581 }, { "epoch": 0.47278635255889523, "grad_norm": 0.5545551776885986, "learning_rate": 9.899908666430707e-06, "loss": 0.5074, "step": 582 }, { "epoch": 0.4735987002437043, "grad_norm": 0.46231019496917725, "learning_rate": 9.898965382031208e-06, "loss": 0.4606, "step": 583 }, { "epoch": 0.4744110479285134, "grad_norm": 0.4910362660884857, "learning_rate": 9.898017718949759e-06, "loss": 0.4667, "step": 584 }, { "epoch": 0.4752233956133225, "grad_norm": 0.4823734164237976, "learning_rate": 9.897065678033376e-06, "loss": 0.5033, "step": 585 }, { "epoch": 0.4760357432981316, "grad_norm": 0.48009324073791504, "learning_rate": 9.896109260132993e-06, "loss": 0.4629, "step": 586 }, { "epoch": 0.4768480909829407, "grad_norm": 0.5407451391220093, "learning_rate": 9.895148466103451e-06, "loss": 0.5136, "step": 587 }, { "epoch": 0.4776604386677498, "grad_norm": 0.45896950364112854, "learning_rate": 9.894183296803509e-06, "loss": 0.4758, "step": 588 }, { "epoch": 0.4784727863525589, "grad_norm": 0.4849286675453186, "learning_rate": 9.893213753095829e-06, "loss": 0.4784, "step": 589 }, { "epoch": 0.479285134037368, "grad_norm": 0.4668518304824829, "learning_rate": 9.892239835846988e-06, "loss": 0.4789, "step": 590 }, { "epoch": 0.4800974817221771, "grad_norm": 0.4598836898803711, "learning_rate": 9.891261545927469e-06, "loss": 0.4822, "step": 591 }, { "epoch": 0.4809098294069862, "grad_norm": 0.48795342445373535, "learning_rate": 9.890278884211668e-06, "loss": 0.5045, "step": 592 }, { "epoch": 0.48172217709179527, "grad_norm": 0.5036784410476685, "learning_rate": 9.889291851577885e-06, "loss": 0.4854, "step": 593 }, { "epoch": 0.4825345247766044, "grad_norm": 0.5608808398246765, "learning_rate": 9.888300448908324e-06, "loss": 0.4886, "step": 594 }, { "epoch": 0.4833468724614135, "grad_norm": 0.48936161398887634, "learning_rate": 9.8873046770891e-06, "loss": 0.496, "step": 595 }, { "epoch": 0.4841592201462226, "grad_norm": 0.4977943003177643, "learning_rate": 9.88630453701023e-06, "loss": 0.4643, "step": 596 }, { "epoch": 0.4849715678310317, "grad_norm": 0.4891336262226105, "learning_rate": 9.88530002956564e-06, "loss": 0.5098, "step": 597 }, { "epoch": 0.48578391551584077, "grad_norm": 0.4526248276233673, "learning_rate": 9.884291155653151e-06, "loss": 0.4456, "step": 598 }, { "epoch": 0.48659626320064986, "grad_norm": 0.49149397015571594, "learning_rate": 9.883277916174496e-06, "loss": 0.4719, "step": 599 }, { "epoch": 0.487408610885459, "grad_norm": 0.5347243547439575, "learning_rate": 9.882260312035304e-06, "loss": 0.4779, "step": 600 }, { "epoch": 0.4882209585702681, "grad_norm": 0.45201754570007324, "learning_rate": 9.881238344145106e-06, "loss": 0.4696, "step": 601 }, { "epoch": 0.4890333062550772, "grad_norm": 0.49777254462242126, "learning_rate": 9.880212013417334e-06, "loss": 0.4771, "step": 602 }, { "epoch": 0.48984565393988627, "grad_norm": 0.5125394463539124, "learning_rate": 9.879181320769322e-06, "loss": 0.4588, "step": 603 }, { "epoch": 0.49065800162469536, "grad_norm": 0.4388071596622467, "learning_rate": 9.878146267122298e-06, "loss": 0.4938, "step": 604 }, { "epoch": 0.49147034930950445, "grad_norm": 0.5140808820724487, "learning_rate": 9.877106853401392e-06, "loss": 0.4802, "step": 605 }, { "epoch": 0.49228269699431354, "grad_norm": 0.5832600593566895, "learning_rate": 9.876063080535627e-06, "loss": 0.4879, "step": 606 }, { "epoch": 0.4930950446791227, "grad_norm": 0.5455216765403748, "learning_rate": 9.875014949457926e-06, "loss": 0.5178, "step": 607 }, { "epoch": 0.4939073923639318, "grad_norm": 0.5727302432060242, "learning_rate": 9.873962461105104e-06, "loss": 0.4725, "step": 608 }, { "epoch": 0.49471974004874086, "grad_norm": 0.5598577857017517, "learning_rate": 9.872905616417875e-06, "loss": 0.4528, "step": 609 }, { "epoch": 0.49553208773354995, "grad_norm": 0.530923068523407, "learning_rate": 9.87184441634084e-06, "loss": 0.4808, "step": 610 }, { "epoch": 0.49634443541835904, "grad_norm": 0.6060804128646851, "learning_rate": 9.870778861822502e-06, "loss": 0.4711, "step": 611 }, { "epoch": 0.49715678310316813, "grad_norm": 0.5631340742111206, "learning_rate": 9.869708953815247e-06, "loss": 0.4793, "step": 612 }, { "epoch": 0.4979691307879773, "grad_norm": 0.45933523774147034, "learning_rate": 9.868634693275356e-06, "loss": 0.4766, "step": 613 }, { "epoch": 0.49878147847278637, "grad_norm": 0.6149038672447205, "learning_rate": 9.867556081163002e-06, "loss": 0.4552, "step": 614 }, { "epoch": 0.49959382615759546, "grad_norm": 0.5125203728675842, "learning_rate": 9.86647311844224e-06, "loss": 0.4561, "step": 615 }, { "epoch": 0.5004061738424046, "grad_norm": 0.5778510570526123, "learning_rate": 9.865385806081028e-06, "loss": 0.4641, "step": 616 }, { "epoch": 0.5012185215272137, "grad_norm": 0.5816516876220703, "learning_rate": 9.864294145051197e-06, "loss": 0.4829, "step": 617 }, { "epoch": 0.5020308692120228, "grad_norm": 0.5512207746505737, "learning_rate": 9.863198136328474e-06, "loss": 0.4899, "step": 618 }, { "epoch": 0.5028432168968319, "grad_norm": 0.5914095044136047, "learning_rate": 9.862097780892463e-06, "loss": 0.4957, "step": 619 }, { "epoch": 0.503655564581641, "grad_norm": 0.5292438864707947, "learning_rate": 9.860993079726665e-06, "loss": 0.4715, "step": 620 }, { "epoch": 0.50446791226645, "grad_norm": 0.4849179685115814, "learning_rate": 9.859884033818454e-06, "loss": 0.5114, "step": 621 }, { "epoch": 0.5052802599512591, "grad_norm": 0.606324315071106, "learning_rate": 9.858770644159098e-06, "loss": 0.4878, "step": 622 }, { "epoch": 0.5060926076360682, "grad_norm": 0.556729257106781, "learning_rate": 9.857652911743734e-06, "loss": 0.4821, "step": 623 }, { "epoch": 0.5069049553208773, "grad_norm": 0.46631667017936707, "learning_rate": 9.856530837571394e-06, "loss": 0.4851, "step": 624 }, { "epoch": 0.5077173030056864, "grad_norm": 0.5172593593597412, "learning_rate": 9.855404422644983e-06, "loss": 0.4602, "step": 625 }, { "epoch": 0.5085296506904955, "grad_norm": 0.5172812938690186, "learning_rate": 9.85427366797129e-06, "loss": 0.5199, "step": 626 }, { "epoch": 0.5093419983753046, "grad_norm": 0.44851288199424744, "learning_rate": 9.853138574560976e-06, "loss": 0.4775, "step": 627 }, { "epoch": 0.5101543460601138, "grad_norm": 0.4915592670440674, "learning_rate": 9.851999143428587e-06, "loss": 0.4633, "step": 628 }, { "epoch": 0.5109666937449229, "grad_norm": 0.565386176109314, "learning_rate": 9.850855375592543e-06, "loss": 0.4598, "step": 629 }, { "epoch": 0.511779041429732, "grad_norm": 0.5245419144630432, "learning_rate": 9.84970727207514e-06, "loss": 0.4599, "step": 630 }, { "epoch": 0.512591389114541, "grad_norm": 0.528085470199585, "learning_rate": 9.848554833902551e-06, "loss": 0.5036, "step": 631 }, { "epoch": 0.5134037367993501, "grad_norm": 0.5199056267738342, "learning_rate": 9.847398062104823e-06, "loss": 0.5008, "step": 632 }, { "epoch": 0.5142160844841592, "grad_norm": 0.549408495426178, "learning_rate": 9.846236957715872e-06, "loss": 0.4837, "step": 633 }, { "epoch": 0.5150284321689683, "grad_norm": 0.5063717365264893, "learning_rate": 9.845071521773492e-06, "loss": 0.487, "step": 634 }, { "epoch": 0.5158407798537774, "grad_norm": 0.49895793199539185, "learning_rate": 9.843901755319347e-06, "loss": 0.4898, "step": 635 }, { "epoch": 0.5166531275385865, "grad_norm": 0.5782054662704468, "learning_rate": 9.842727659398971e-06, "loss": 0.4842, "step": 636 }, { "epoch": 0.5174654752233956, "grad_norm": 0.5151228308677673, "learning_rate": 9.841549235061766e-06, "loss": 0.465, "step": 637 }, { "epoch": 0.5182778229082047, "grad_norm": 0.5027167797088623, "learning_rate": 9.840366483361008e-06, "loss": 0.4904, "step": 638 }, { "epoch": 0.5190901705930138, "grad_norm": 0.6593002676963806, "learning_rate": 9.839179405353834e-06, "loss": 0.5106, "step": 639 }, { "epoch": 0.5199025182778229, "grad_norm": 0.5689668655395508, "learning_rate": 9.837988002101253e-06, "loss": 0.489, "step": 640 }, { "epoch": 0.5207148659626321, "grad_norm": 0.47903433442115784, "learning_rate": 9.836792274668135e-06, "loss": 0.4495, "step": 641 }, { "epoch": 0.5215272136474411, "grad_norm": 0.5373712778091431, "learning_rate": 9.835592224123221e-06, "loss": 0.4957, "step": 642 }, { "epoch": 0.5223395613322502, "grad_norm": 0.5986288189888, "learning_rate": 9.834387851539113e-06, "loss": 0.4932, "step": 643 }, { "epoch": 0.5231519090170593, "grad_norm": 0.506561279296875, "learning_rate": 9.833179157992274e-06, "loss": 0.4941, "step": 644 }, { "epoch": 0.5239642567018684, "grad_norm": 0.5052922368049622, "learning_rate": 9.831966144563032e-06, "loss": 0.4756, "step": 645 }, { "epoch": 0.5247766043866775, "grad_norm": 0.5420305132865906, "learning_rate": 9.830748812335576e-06, "loss": 0.5138, "step": 646 }, { "epoch": 0.5255889520714866, "grad_norm": 0.47589266300201416, "learning_rate": 9.829527162397951e-06, "loss": 0.4789, "step": 647 }, { "epoch": 0.5264012997562957, "grad_norm": 0.5162862539291382, "learning_rate": 9.828301195842069e-06, "loss": 0.4985, "step": 648 }, { "epoch": 0.5272136474411048, "grad_norm": 0.5096306204795837, "learning_rate": 9.827070913763693e-06, "loss": 0.4969, "step": 649 }, { "epoch": 0.5280259951259139, "grad_norm": 0.499342143535614, "learning_rate": 9.825836317262445e-06, "loss": 0.4579, "step": 650 }, { "epoch": 0.528838342810723, "grad_norm": 0.5405427813529968, "learning_rate": 9.824597407441805e-06, "loss": 0.4864, "step": 651 }, { "epoch": 0.529650690495532, "grad_norm": 0.5131211876869202, "learning_rate": 9.823354185409108e-06, "loss": 0.4744, "step": 652 }, { "epoch": 0.5304630381803412, "grad_norm": 0.5098509788513184, "learning_rate": 9.822106652275541e-06, "loss": 0.5154, "step": 653 }, { "epoch": 0.5312753858651503, "grad_norm": 0.5229046940803528, "learning_rate": 9.820854809156146e-06, "loss": 0.4537, "step": 654 }, { "epoch": 0.5320877335499594, "grad_norm": 0.5336276888847351, "learning_rate": 9.819598657169819e-06, "loss": 0.5105, "step": 655 }, { "epoch": 0.5329000812347685, "grad_norm": 0.5914848446846008, "learning_rate": 9.818338197439304e-06, "loss": 0.4777, "step": 656 }, { "epoch": 0.5337124289195776, "grad_norm": 0.5238675475120544, "learning_rate": 9.817073431091194e-06, "loss": 0.505, "step": 657 }, { "epoch": 0.5345247766043867, "grad_norm": 0.4772823452949524, "learning_rate": 9.815804359255937e-06, "loss": 0.4745, "step": 658 }, { "epoch": 0.5353371242891958, "grad_norm": 0.5987365245819092, "learning_rate": 9.814530983067825e-06, "loss": 0.4878, "step": 659 }, { "epoch": 0.5361494719740049, "grad_norm": 0.4934236705303192, "learning_rate": 9.813253303664997e-06, "loss": 0.4597, "step": 660 }, { "epoch": 0.536961819658814, "grad_norm": 0.5166174173355103, "learning_rate": 9.811971322189442e-06, "loss": 0.4749, "step": 661 }, { "epoch": 0.537774167343623, "grad_norm": 0.5735315680503845, "learning_rate": 9.810685039786989e-06, "loss": 0.4802, "step": 662 }, { "epoch": 0.5385865150284321, "grad_norm": 0.5857974290847778, "learning_rate": 9.809394457607315e-06, "loss": 0.4896, "step": 663 }, { "epoch": 0.5393988627132412, "grad_norm": 0.5143821835517883, "learning_rate": 9.808099576803937e-06, "loss": 0.4738, "step": 664 }, { "epoch": 0.5402112103980503, "grad_norm": 0.5691672563552856, "learning_rate": 9.80680039853422e-06, "loss": 0.4645, "step": 665 }, { "epoch": 0.5410235580828595, "grad_norm": 0.6139249205589294, "learning_rate": 9.805496923959363e-06, "loss": 0.4683, "step": 666 }, { "epoch": 0.5418359057676686, "grad_norm": 0.5629609227180481, "learning_rate": 9.804189154244408e-06, "loss": 0.5007, "step": 667 }, { "epoch": 0.5426482534524777, "grad_norm": 0.5456512570381165, "learning_rate": 9.802877090558237e-06, "loss": 0.4663, "step": 668 }, { "epoch": 0.5434606011372868, "grad_norm": 0.6781876683235168, "learning_rate": 9.801560734073567e-06, "loss": 0.4915, "step": 669 }, { "epoch": 0.5442729488220959, "grad_norm": 0.5995696187019348, "learning_rate": 9.800240085966957e-06, "loss": 0.4921, "step": 670 }, { "epoch": 0.545085296506905, "grad_norm": 0.6287668943405151, "learning_rate": 9.798915147418795e-06, "loss": 0.481, "step": 671 }, { "epoch": 0.545897644191714, "grad_norm": 0.6584242582321167, "learning_rate": 9.797585919613311e-06, "loss": 0.5024, "step": 672 }, { "epoch": 0.5467099918765231, "grad_norm": 0.48127907514572144, "learning_rate": 9.796252403738564e-06, "loss": 0.4792, "step": 673 }, { "epoch": 0.5475223395613322, "grad_norm": 0.5172940492630005, "learning_rate": 9.794914600986446e-06, "loss": 0.4764, "step": 674 }, { "epoch": 0.5483346872461413, "grad_norm": 0.4964587986469269, "learning_rate": 9.793572512552683e-06, "loss": 0.4616, "step": 675 }, { "epoch": 0.5491470349309504, "grad_norm": 0.5081837177276611, "learning_rate": 9.792226139636827e-06, "loss": 0.5004, "step": 676 }, { "epoch": 0.5499593826157595, "grad_norm": 0.5880322456359863, "learning_rate": 9.790875483442265e-06, "loss": 0.4564, "step": 677 }, { "epoch": 0.5507717303005687, "grad_norm": 0.4805297255516052, "learning_rate": 9.78952054517621e-06, "loss": 0.5043, "step": 678 }, { "epoch": 0.5515840779853778, "grad_norm": 0.5655009746551514, "learning_rate": 9.788161326049702e-06, "loss": 0.4712, "step": 679 }, { "epoch": 0.5523964256701869, "grad_norm": 0.5046995878219604, "learning_rate": 9.786797827277607e-06, "loss": 0.475, "step": 680 }, { "epoch": 0.553208773354996, "grad_norm": 0.4422001838684082, "learning_rate": 9.785430050078616e-06, "loss": 0.4688, "step": 681 }, { "epoch": 0.554021121039805, "grad_norm": 0.5890532732009888, "learning_rate": 9.784057995675245e-06, "loss": 0.483, "step": 682 }, { "epoch": 0.5548334687246141, "grad_norm": 0.5244244337081909, "learning_rate": 9.782681665293831e-06, "loss": 0.5075, "step": 683 }, { "epoch": 0.5556458164094232, "grad_norm": 0.5230079293251038, "learning_rate": 9.781301060164537e-06, "loss": 0.4654, "step": 684 }, { "epoch": 0.5564581640942323, "grad_norm": 0.5398865938186646, "learning_rate": 9.779916181521342e-06, "loss": 0.4897, "step": 685 }, { "epoch": 0.5572705117790414, "grad_norm": 0.4909040033817291, "learning_rate": 9.778527030602049e-06, "loss": 0.4789, "step": 686 }, { "epoch": 0.5580828594638505, "grad_norm": 0.5490494966506958, "learning_rate": 9.777133608648276e-06, "loss": 0.4931, "step": 687 }, { "epoch": 0.5588952071486596, "grad_norm": 0.45627525448799133, "learning_rate": 9.775735916905458e-06, "loss": 0.4784, "step": 688 }, { "epoch": 0.5597075548334687, "grad_norm": 0.48990219831466675, "learning_rate": 9.77433395662285e-06, "loss": 0.4763, "step": 689 }, { "epoch": 0.5605199025182778, "grad_norm": 0.5045406818389893, "learning_rate": 9.77292772905352e-06, "loss": 0.4943, "step": 690 }, { "epoch": 0.561332250203087, "grad_norm": 0.4497394859790802, "learning_rate": 9.77151723545435e-06, "loss": 0.486, "step": 691 }, { "epoch": 0.5621445978878961, "grad_norm": 0.4579395651817322, "learning_rate": 9.77010247708604e-06, "loss": 0.4565, "step": 692 }, { "epoch": 0.5629569455727051, "grad_norm": 0.4906800091266632, "learning_rate": 9.768683455213089e-06, "loss": 0.475, "step": 693 }, { "epoch": 0.5637692932575142, "grad_norm": 0.4217583239078522, "learning_rate": 9.76726017110382e-06, "loss": 0.4665, "step": 694 }, { "epoch": 0.5645816409423233, "grad_norm": 0.45224449038505554, "learning_rate": 9.765832626030359e-06, "loss": 0.4681, "step": 695 }, { "epoch": 0.5653939886271324, "grad_norm": 0.42656925320625305, "learning_rate": 9.764400821268642e-06, "loss": 0.4569, "step": 696 }, { "epoch": 0.5662063363119415, "grad_norm": 0.43409469723701477, "learning_rate": 9.762964758098412e-06, "loss": 0.4987, "step": 697 }, { "epoch": 0.5670186839967506, "grad_norm": 0.5077341794967651, "learning_rate": 9.761524437803221e-06, "loss": 0.4587, "step": 698 }, { "epoch": 0.5678310316815597, "grad_norm": 0.4403238296508789, "learning_rate": 9.76007986167042e-06, "loss": 0.4775, "step": 699 }, { "epoch": 0.5686433793663688, "grad_norm": 0.533166229724884, "learning_rate": 9.758631030991168e-06, "loss": 0.4858, "step": 700 }, { "epoch": 0.5694557270511779, "grad_norm": 0.4663658142089844, "learning_rate": 9.757177947060426e-06, "loss": 0.4848, "step": 701 }, { "epoch": 0.570268074735987, "grad_norm": 0.45193350315093994, "learning_rate": 9.755720611176957e-06, "loss": 0.4688, "step": 702 }, { "epoch": 0.5710804224207962, "grad_norm": 0.608262300491333, "learning_rate": 9.754259024643324e-06, "loss": 0.4724, "step": 703 }, { "epoch": 0.5718927701056052, "grad_norm": 0.527118980884552, "learning_rate": 9.75279318876589e-06, "loss": 0.5159, "step": 704 }, { "epoch": 0.5727051177904143, "grad_norm": 0.5096942782402039, "learning_rate": 9.751323104854812e-06, "loss": 0.4663, "step": 705 }, { "epoch": 0.5735174654752234, "grad_norm": 0.5341753959655762, "learning_rate": 9.74984877422405e-06, "loss": 0.4845, "step": 706 }, { "epoch": 0.5743298131600325, "grad_norm": 0.6665489077568054, "learning_rate": 9.748370198191355e-06, "loss": 0.4564, "step": 707 }, { "epoch": 0.5751421608448416, "grad_norm": 0.42647042870521545, "learning_rate": 9.746887378078276e-06, "loss": 0.4888, "step": 708 }, { "epoch": 0.5759545085296507, "grad_norm": 0.6072022914886475, "learning_rate": 9.74540031521015e-06, "loss": 0.4829, "step": 709 }, { "epoch": 0.5767668562144598, "grad_norm": 0.6040878295898438, "learning_rate": 9.743909010916115e-06, "loss": 0.5014, "step": 710 }, { "epoch": 0.5775792038992689, "grad_norm": 0.5354973673820496, "learning_rate": 9.74241346652909e-06, "loss": 0.4721, "step": 711 }, { "epoch": 0.578391551584078, "grad_norm": 0.6246288418769836, "learning_rate": 9.740913683385791e-06, "loss": 0.498, "step": 712 }, { "epoch": 0.579203899268887, "grad_norm": 0.5609967708587646, "learning_rate": 9.73940966282672e-06, "loss": 0.5002, "step": 713 }, { "epoch": 0.5800162469536961, "grad_norm": 0.5037618279457092, "learning_rate": 9.737901406196164e-06, "loss": 0.4791, "step": 714 }, { "epoch": 0.5808285946385053, "grad_norm": 0.5292035937309265, "learning_rate": 9.736388914842203e-06, "loss": 0.466, "step": 715 }, { "epoch": 0.5816409423233144, "grad_norm": 0.526015043258667, "learning_rate": 9.734872190116695e-06, "loss": 0.4729, "step": 716 }, { "epoch": 0.5824532900081235, "grad_norm": 0.6039483547210693, "learning_rate": 9.733351233375283e-06, "loss": 0.4928, "step": 717 }, { "epoch": 0.5832656376929326, "grad_norm": 0.4864838123321533, "learning_rate": 9.731826045977397e-06, "loss": 0.4482, "step": 718 }, { "epoch": 0.5840779853777417, "grad_norm": 0.5396150350570679, "learning_rate": 9.730296629286245e-06, "loss": 0.4707, "step": 719 }, { "epoch": 0.5848903330625508, "grad_norm": 0.48373323678970337, "learning_rate": 9.728762984668813e-06, "loss": 0.5336, "step": 720 }, { "epoch": 0.5857026807473599, "grad_norm": 0.5778162479400635, "learning_rate": 9.727225113495871e-06, "loss": 0.4713, "step": 721 }, { "epoch": 0.586515028432169, "grad_norm": 0.5201114416122437, "learning_rate": 9.725683017141964e-06, "loss": 0.4858, "step": 722 }, { "epoch": 0.587327376116978, "grad_norm": 0.5454099178314209, "learning_rate": 9.724136696985412e-06, "loss": 0.4872, "step": 723 }, { "epoch": 0.5881397238017871, "grad_norm": 0.4987611174583435, "learning_rate": 9.722586154408312e-06, "loss": 0.4684, "step": 724 }, { "epoch": 0.5889520714865962, "grad_norm": 0.5262108445167542, "learning_rate": 9.721031390796535e-06, "loss": 0.5048, "step": 725 }, { "epoch": 0.5897644191714053, "grad_norm": 0.5154426097869873, "learning_rate": 9.719472407539725e-06, "loss": 0.4862, "step": 726 }, { "epoch": 0.5905767668562144, "grad_norm": 0.5164682865142822, "learning_rate": 9.717909206031295e-06, "loss": 0.4946, "step": 727 }, { "epoch": 0.5913891145410236, "grad_norm": 0.5091283321380615, "learning_rate": 9.716341787668434e-06, "loss": 0.4845, "step": 728 }, { "epoch": 0.5922014622258327, "grad_norm": 0.48636528849601746, "learning_rate": 9.714770153852093e-06, "loss": 0.4526, "step": 729 }, { "epoch": 0.5930138099106418, "grad_norm": 0.6054742336273193, "learning_rate": 9.713194305986995e-06, "loss": 0.4796, "step": 730 }, { "epoch": 0.5938261575954509, "grad_norm": 0.5157285332679749, "learning_rate": 9.71161424548163e-06, "loss": 0.509, "step": 731 }, { "epoch": 0.59463850528026, "grad_norm": 0.5601982474327087, "learning_rate": 9.710029973748249e-06, "loss": 0.4635, "step": 732 }, { "epoch": 0.595450852965069, "grad_norm": 0.4805345833301544, "learning_rate": 9.708441492202874e-06, "loss": 0.4524, "step": 733 }, { "epoch": 0.5962632006498781, "grad_norm": 0.5701055526733398, "learning_rate": 9.70684880226528e-06, "loss": 0.4593, "step": 734 }, { "epoch": 0.5970755483346872, "grad_norm": 0.5075626373291016, "learning_rate": 9.705251905359014e-06, "loss": 0.4746, "step": 735 }, { "epoch": 0.5978878960194963, "grad_norm": 0.5680609345436096, "learning_rate": 9.703650802911374e-06, "loss": 0.4969, "step": 736 }, { "epoch": 0.5987002437043054, "grad_norm": 0.45548075437545776, "learning_rate": 9.702045496353422e-06, "loss": 0.4722, "step": 737 }, { "epoch": 0.5995125913891145, "grad_norm": 0.5298168063163757, "learning_rate": 9.700435987119981e-06, "loss": 0.4985, "step": 738 }, { "epoch": 0.6003249390739236, "grad_norm": 0.43709418177604675, "learning_rate": 9.69882227664962e-06, "loss": 0.4489, "step": 739 }, { "epoch": 0.6011372867587328, "grad_norm": 0.5100873708724976, "learning_rate": 9.69720436638467e-06, "loss": 0.4577, "step": 740 }, { "epoch": 0.6019496344435419, "grad_norm": 0.5341557264328003, "learning_rate": 9.69558225777122e-06, "loss": 0.4638, "step": 741 }, { "epoch": 0.602761982128351, "grad_norm": 0.46457260847091675, "learning_rate": 9.693955952259099e-06, "loss": 0.461, "step": 742 }, { "epoch": 0.6035743298131601, "grad_norm": 0.44016724824905396, "learning_rate": 9.6923254513019e-06, "loss": 0.469, "step": 743 }, { "epoch": 0.6043866774979691, "grad_norm": 0.5114173293113708, "learning_rate": 9.69069075635696e-06, "loss": 0.4893, "step": 744 }, { "epoch": 0.6051990251827782, "grad_norm": 0.5072087645530701, "learning_rate": 9.689051868885362e-06, "loss": 0.4847, "step": 745 }, { "epoch": 0.6060113728675873, "grad_norm": 0.4003745913505554, "learning_rate": 9.68740879035194e-06, "loss": 0.5035, "step": 746 }, { "epoch": 0.6068237205523964, "grad_norm": 0.4884507954120636, "learning_rate": 9.685761522225271e-06, "loss": 0.4976, "step": 747 }, { "epoch": 0.6076360682372055, "grad_norm": 0.43283185362815857, "learning_rate": 9.684110065977685e-06, "loss": 0.4599, "step": 748 }, { "epoch": 0.6084484159220146, "grad_norm": 0.44943001866340637, "learning_rate": 9.682454423085244e-06, "loss": 0.4614, "step": 749 }, { "epoch": 0.6092607636068237, "grad_norm": 0.47455140948295593, "learning_rate": 9.680794595027761e-06, "loss": 0.473, "step": 750 }, { "epoch": 0.6100731112916328, "grad_norm": 0.4493069350719452, "learning_rate": 9.67913058328878e-06, "loss": 0.4658, "step": 751 }, { "epoch": 0.6108854589764419, "grad_norm": 0.5088927745819092, "learning_rate": 9.677462389355594e-06, "loss": 0.4935, "step": 752 }, { "epoch": 0.6116978066612511, "grad_norm": 0.495408296585083, "learning_rate": 9.675790014719231e-06, "loss": 0.4687, "step": 753 }, { "epoch": 0.6125101543460602, "grad_norm": 0.47937721014022827, "learning_rate": 9.674113460874453e-06, "loss": 0.4545, "step": 754 }, { "epoch": 0.6133225020308692, "grad_norm": 0.5113687515258789, "learning_rate": 9.67243272931976e-06, "loss": 0.4818, "step": 755 }, { "epoch": 0.6141348497156783, "grad_norm": 0.5597743391990662, "learning_rate": 9.670747821557387e-06, "loss": 0.4843, "step": 756 }, { "epoch": 0.6149471974004874, "grad_norm": 0.48122748732566833, "learning_rate": 9.669058739093296e-06, "loss": 0.4868, "step": 757 }, { "epoch": 0.6157595450852965, "grad_norm": 0.5828558206558228, "learning_rate": 9.667365483437188e-06, "loss": 0.4967, "step": 758 }, { "epoch": 0.6165718927701056, "grad_norm": 0.4902104139328003, "learning_rate": 9.665668056102487e-06, "loss": 0.4686, "step": 759 }, { "epoch": 0.6173842404549147, "grad_norm": 0.5212362408638, "learning_rate": 9.663966458606353e-06, "loss": 0.5234, "step": 760 }, { "epoch": 0.6181965881397238, "grad_norm": 0.4300674498081207, "learning_rate": 9.662260692469664e-06, "loss": 0.4581, "step": 761 }, { "epoch": 0.6190089358245329, "grad_norm": 0.4353412985801697, "learning_rate": 9.660550759217035e-06, "loss": 0.4522, "step": 762 }, { "epoch": 0.619821283509342, "grad_norm": 0.4798797369003296, "learning_rate": 9.658836660376796e-06, "loss": 0.4598, "step": 763 }, { "epoch": 0.620633631194151, "grad_norm": 0.45335501432418823, "learning_rate": 9.657118397481004e-06, "loss": 0.4696, "step": 764 }, { "epoch": 0.6214459788789602, "grad_norm": 0.5289939641952515, "learning_rate": 9.65539597206544e-06, "loss": 0.478, "step": 765 }, { "epoch": 0.6222583265637693, "grad_norm": 0.48557814955711365, "learning_rate": 9.6536693856696e-06, "loss": 0.4261, "step": 766 }, { "epoch": 0.6230706742485784, "grad_norm": 0.5315256714820862, "learning_rate": 9.651938639836705e-06, "loss": 0.4791, "step": 767 }, { "epoch": 0.6238830219333875, "grad_norm": 0.46642035245895386, "learning_rate": 9.650203736113689e-06, "loss": 0.4703, "step": 768 }, { "epoch": 0.6246953696181966, "grad_norm": 0.4311215281486511, "learning_rate": 9.648464676051206e-06, "loss": 0.4724, "step": 769 }, { "epoch": 0.6255077173030057, "grad_norm": 0.5052635669708252, "learning_rate": 9.64672146120362e-06, "loss": 0.4973, "step": 770 }, { "epoch": 0.6263200649878148, "grad_norm": 0.45287466049194336, "learning_rate": 9.644974093129017e-06, "loss": 0.471, "step": 771 }, { "epoch": 0.6271324126726239, "grad_norm": 0.47873014211654663, "learning_rate": 9.643222573389184e-06, "loss": 0.5173, "step": 772 }, { "epoch": 0.627944760357433, "grad_norm": 0.46916672587394714, "learning_rate": 9.641466903549628e-06, "loss": 0.4489, "step": 773 }, { "epoch": 0.628757108042242, "grad_norm": 0.4969717264175415, "learning_rate": 9.639707085179562e-06, "loss": 0.4852, "step": 774 }, { "epoch": 0.6295694557270511, "grad_norm": 0.47347816824913025, "learning_rate": 9.637943119851905e-06, "loss": 0.4954, "step": 775 }, { "epoch": 0.6303818034118602, "grad_norm": 0.44414928555488586, "learning_rate": 9.636175009143286e-06, "loss": 0.4495, "step": 776 }, { "epoch": 0.6311941510966693, "grad_norm": 0.5258604288101196, "learning_rate": 9.634402754634037e-06, "loss": 0.5053, "step": 777 }, { "epoch": 0.6320064987814785, "grad_norm": 0.42933136224746704, "learning_rate": 9.632626357908194e-06, "loss": 0.5004, "step": 778 }, { "epoch": 0.6328188464662876, "grad_norm": 0.4924398958683014, "learning_rate": 9.630845820553495e-06, "loss": 0.4607, "step": 779 }, { "epoch": 0.6336311941510967, "grad_norm": 0.51087486743927, "learning_rate": 9.629061144161383e-06, "loss": 0.4847, "step": 780 }, { "epoch": 0.6344435418359058, "grad_norm": 0.44060593843460083, "learning_rate": 9.627272330326993e-06, "loss": 0.4683, "step": 781 }, { "epoch": 0.6352558895207149, "grad_norm": 0.49551406502723694, "learning_rate": 9.625479380649163e-06, "loss": 0.4888, "step": 782 }, { "epoch": 0.636068237205524, "grad_norm": 0.47920048236846924, "learning_rate": 9.623682296730427e-06, "loss": 0.4824, "step": 783 }, { "epoch": 0.6368805848903331, "grad_norm": 0.5557704567909241, "learning_rate": 9.621881080177013e-06, "loss": 0.4689, "step": 784 }, { "epoch": 0.6376929325751421, "grad_norm": 0.505517840385437, "learning_rate": 9.620075732598843e-06, "loss": 0.4862, "step": 785 }, { "epoch": 0.6385052802599512, "grad_norm": 0.44242507219314575, "learning_rate": 9.618266255609533e-06, "loss": 0.4788, "step": 786 }, { "epoch": 0.6393176279447603, "grad_norm": 0.5345906615257263, "learning_rate": 9.616452650826392e-06, "loss": 0.4741, "step": 787 }, { "epoch": 0.6401299756295694, "grad_norm": 0.48230835795402527, "learning_rate": 9.614634919870407e-06, "loss": 0.4639, "step": 788 }, { "epoch": 0.6409423233143785, "grad_norm": 0.4664379358291626, "learning_rate": 9.612813064366267e-06, "loss": 0.463, "step": 789 }, { "epoch": 0.6417546709991877, "grad_norm": 0.44083043932914734, "learning_rate": 9.61098708594234e-06, "loss": 0.4766, "step": 790 }, { "epoch": 0.6425670186839968, "grad_norm": 0.4771829843521118, "learning_rate": 9.60915698623068e-06, "loss": 0.5132, "step": 791 }, { "epoch": 0.6433793663688059, "grad_norm": 0.44368380308151245, "learning_rate": 9.607322766867026e-06, "loss": 0.4686, "step": 792 }, { "epoch": 0.644191714053615, "grad_norm": 0.5316287279129028, "learning_rate": 9.605484429490796e-06, "loss": 0.4667, "step": 793 }, { "epoch": 0.6450040617384241, "grad_norm": 0.4671468734741211, "learning_rate": 9.603641975745095e-06, "loss": 0.4549, "step": 794 }, { "epoch": 0.6458164094232332, "grad_norm": 0.5129256844520569, "learning_rate": 9.601795407276699e-06, "loss": 0.4773, "step": 795 }, { "epoch": 0.6466287571080422, "grad_norm": 0.5458052754402161, "learning_rate": 9.59994472573607e-06, "loss": 0.4747, "step": 796 }, { "epoch": 0.6474411047928513, "grad_norm": 0.43545934557914734, "learning_rate": 9.598089932777339e-06, "loss": 0.4683, "step": 797 }, { "epoch": 0.6482534524776604, "grad_norm": 0.5019978284835815, "learning_rate": 9.596231030058315e-06, "loss": 0.4511, "step": 798 }, { "epoch": 0.6490658001624695, "grad_norm": 0.5115323066711426, "learning_rate": 9.594368019240484e-06, "loss": 0.5087, "step": 799 }, { "epoch": 0.6498781478472786, "grad_norm": 0.5263423919677734, "learning_rate": 9.592500901988994e-06, "loss": 0.4961, "step": 800 }, { "epoch": 0.6506904955320877, "grad_norm": 0.4772259294986725, "learning_rate": 9.590629679972673e-06, "loss": 0.4931, "step": 801 }, { "epoch": 0.6515028432168968, "grad_norm": 0.4585248529911041, "learning_rate": 9.588754354864014e-06, "loss": 0.4795, "step": 802 }, { "epoch": 0.652315190901706, "grad_norm": 0.4448578655719757, "learning_rate": 9.586874928339176e-06, "loss": 0.46, "step": 803 }, { "epoch": 0.6531275385865151, "grad_norm": 0.48814743757247925, "learning_rate": 9.584991402077988e-06, "loss": 0.4944, "step": 804 }, { "epoch": 0.6539398862713242, "grad_norm": 0.46126970648765564, "learning_rate": 9.583103777763938e-06, "loss": 0.4842, "step": 805 }, { "epoch": 0.6547522339561332, "grad_norm": 0.507763683795929, "learning_rate": 9.58121205708418e-06, "loss": 0.4579, "step": 806 }, { "epoch": 0.6555645816409423, "grad_norm": 0.5065272450447083, "learning_rate": 9.57931624172953e-06, "loss": 0.4959, "step": 807 }, { "epoch": 0.6563769293257514, "grad_norm": 0.5227736830711365, "learning_rate": 9.577416333394463e-06, "loss": 0.4893, "step": 808 }, { "epoch": 0.6571892770105605, "grad_norm": 0.4821391999721527, "learning_rate": 9.575512333777108e-06, "loss": 0.4629, "step": 809 }, { "epoch": 0.6580016246953696, "grad_norm": 0.5293917059898376, "learning_rate": 9.57360424457926e-06, "loss": 0.4785, "step": 810 }, { "epoch": 0.6588139723801787, "grad_norm": 0.5032228827476501, "learning_rate": 9.571692067506363e-06, "loss": 0.4613, "step": 811 }, { "epoch": 0.6596263200649878, "grad_norm": 0.47105273604393005, "learning_rate": 9.56977580426751e-06, "loss": 0.4741, "step": 812 }, { "epoch": 0.6604386677497969, "grad_norm": 0.5124861001968384, "learning_rate": 9.567855456575459e-06, "loss": 0.4701, "step": 813 }, { "epoch": 0.661251015434606, "grad_norm": 0.47613754868507385, "learning_rate": 9.565931026146607e-06, "loss": 0.4654, "step": 814 }, { "epoch": 0.6620633631194152, "grad_norm": 0.4638701379299164, "learning_rate": 9.564002514701006e-06, "loss": 0.4614, "step": 815 }, { "epoch": 0.6628757108042242, "grad_norm": 0.5349314212799072, "learning_rate": 9.562069923962355e-06, "loss": 0.4764, "step": 816 }, { "epoch": 0.6636880584890333, "grad_norm": 0.525813639163971, "learning_rate": 9.560133255657997e-06, "loss": 0.4897, "step": 817 }, { "epoch": 0.6645004061738424, "grad_norm": 0.47747093439102173, "learning_rate": 9.558192511518925e-06, "loss": 0.4728, "step": 818 }, { "epoch": 0.6653127538586515, "grad_norm": 0.5006821155548096, "learning_rate": 9.556247693279764e-06, "loss": 0.4641, "step": 819 }, { "epoch": 0.6661251015434606, "grad_norm": 0.515968382358551, "learning_rate": 9.554298802678793e-06, "loss": 0.4948, "step": 820 }, { "epoch": 0.6669374492282697, "grad_norm": 0.48413288593292236, "learning_rate": 9.552345841457922e-06, "loss": 0.5068, "step": 821 }, { "epoch": 0.6677497969130788, "grad_norm": 0.46980127692222595, "learning_rate": 9.550388811362704e-06, "loss": 0.4566, "step": 822 }, { "epoch": 0.6685621445978879, "grad_norm": 0.5134105682373047, "learning_rate": 9.548427714142326e-06, "loss": 0.462, "step": 823 }, { "epoch": 0.669374492282697, "grad_norm": 0.46410325169563293, "learning_rate": 9.546462551549612e-06, "loss": 0.4791, "step": 824 }, { "epoch": 0.670186839967506, "grad_norm": 0.4177771508693695, "learning_rate": 9.54449332534102e-06, "loss": 0.4911, "step": 825 }, { "epoch": 0.6709991876523151, "grad_norm": 0.5702652335166931, "learning_rate": 9.542520037276636e-06, "loss": 0.4894, "step": 826 }, { "epoch": 0.6718115353371243, "grad_norm": 0.481804758310318, "learning_rate": 9.540542689120184e-06, "loss": 0.4799, "step": 827 }, { "epoch": 0.6726238830219334, "grad_norm": 0.4519827663898468, "learning_rate": 9.538561282639008e-06, "loss": 0.4779, "step": 828 }, { "epoch": 0.6734362307067425, "grad_norm": 0.42120814323425293, "learning_rate": 9.536575819604087e-06, "loss": 0.4792, "step": 829 }, { "epoch": 0.6742485783915516, "grad_norm": 0.4413878917694092, "learning_rate": 9.534586301790021e-06, "loss": 0.4663, "step": 830 }, { "epoch": 0.6750609260763607, "grad_norm": 0.45427632331848145, "learning_rate": 9.532592730975035e-06, "loss": 0.5065, "step": 831 }, { "epoch": 0.6758732737611698, "grad_norm": 0.44048622250556946, "learning_rate": 9.530595108940978e-06, "loss": 0.4736, "step": 832 }, { "epoch": 0.6766856214459789, "grad_norm": 0.46253326535224915, "learning_rate": 9.52859343747332e-06, "loss": 0.4899, "step": 833 }, { "epoch": 0.677497969130788, "grad_norm": 0.4459482431411743, "learning_rate": 9.526587718361147e-06, "loss": 0.4746, "step": 834 }, { "epoch": 0.6783103168155971, "grad_norm": 0.43912217020988464, "learning_rate": 9.524577953397167e-06, "loss": 0.4715, "step": 835 }, { "epoch": 0.6791226645004061, "grad_norm": 0.4679292142391205, "learning_rate": 9.522564144377703e-06, "loss": 0.4626, "step": 836 }, { "epoch": 0.6799350121852152, "grad_norm": 0.48230791091918945, "learning_rate": 9.52054629310269e-06, "loss": 0.4665, "step": 837 }, { "epoch": 0.6807473598700243, "grad_norm": 0.4927789866924286, "learning_rate": 9.518524401375682e-06, "loss": 0.4618, "step": 838 }, { "epoch": 0.6815597075548334, "grad_norm": 0.4614965617656708, "learning_rate": 9.516498471003837e-06, "loss": 0.4592, "step": 839 }, { "epoch": 0.6823720552396426, "grad_norm": 0.5247054696083069, "learning_rate": 9.514468503797927e-06, "loss": 0.5022, "step": 840 }, { "epoch": 0.6831844029244517, "grad_norm": 0.49347978830337524, "learning_rate": 9.51243450157233e-06, "loss": 0.4839, "step": 841 }, { "epoch": 0.6839967506092608, "grad_norm": 0.5758035778999329, "learning_rate": 9.510396466145032e-06, "loss": 0.4515, "step": 842 }, { "epoch": 0.6848090982940699, "grad_norm": 0.4665607810020447, "learning_rate": 9.508354399337625e-06, "loss": 0.4906, "step": 843 }, { "epoch": 0.685621445978879, "grad_norm": 0.49946266412734985, "learning_rate": 9.5063083029753e-06, "loss": 0.46, "step": 844 }, { "epoch": 0.6864337936636881, "grad_norm": 0.5486260652542114, "learning_rate": 9.504258178886858e-06, "loss": 0.4537, "step": 845 }, { "epoch": 0.6872461413484972, "grad_norm": 0.43427589535713196, "learning_rate": 9.502204028904687e-06, "loss": 0.4716, "step": 846 }, { "epoch": 0.6880584890333062, "grad_norm": 0.4775094985961914, "learning_rate": 9.500145854864785e-06, "loss": 0.459, "step": 847 }, { "epoch": 0.6888708367181153, "grad_norm": 0.43652433156967163, "learning_rate": 9.498083658606744e-06, "loss": 0.4595, "step": 848 }, { "epoch": 0.6896831844029244, "grad_norm": 0.438930481672287, "learning_rate": 9.496017441973747e-06, "loss": 0.4562, "step": 849 }, { "epoch": 0.6904955320877335, "grad_norm": 0.510565996170044, "learning_rate": 9.49394720681257e-06, "loss": 0.4761, "step": 850 }, { "epoch": 0.6913078797725426, "grad_norm": 0.5602505803108215, "learning_rate": 9.49187295497359e-06, "loss": 0.4988, "step": 851 }, { "epoch": 0.6921202274573518, "grad_norm": 0.5348201990127563, "learning_rate": 9.48979468831076e-06, "loss": 0.4715, "step": 852 }, { "epoch": 0.6929325751421609, "grad_norm": 0.5365826487541199, "learning_rate": 9.487712408681635e-06, "loss": 0.4709, "step": 853 }, { "epoch": 0.69374492282697, "grad_norm": 0.4661272466182709, "learning_rate": 9.485626117947351e-06, "loss": 0.4742, "step": 854 }, { "epoch": 0.6945572705117791, "grad_norm": 0.476878821849823, "learning_rate": 9.483535817972625e-06, "loss": 0.4922, "step": 855 }, { "epoch": 0.6953696181965882, "grad_norm": 0.525993287563324, "learning_rate": 9.481441510625765e-06, "loss": 0.4918, "step": 856 }, { "epoch": 0.6961819658813972, "grad_norm": 0.48196908831596375, "learning_rate": 9.479343197778653e-06, "loss": 0.4718, "step": 857 }, { "epoch": 0.6969943135662063, "grad_norm": 0.5042597055435181, "learning_rate": 9.477240881306759e-06, "loss": 0.4362, "step": 858 }, { "epoch": 0.6978066612510154, "grad_norm": 0.5533737540245056, "learning_rate": 9.475134563089128e-06, "loss": 0.502, "step": 859 }, { "epoch": 0.6986190089358245, "grad_norm": 0.4748050570487976, "learning_rate": 9.47302424500838e-06, "loss": 0.4846, "step": 860 }, { "epoch": 0.6994313566206336, "grad_norm": 0.5643673539161682, "learning_rate": 9.47090992895071e-06, "loss": 0.4713, "step": 861 }, { "epoch": 0.7002437043054427, "grad_norm": 0.4783030152320862, "learning_rate": 9.468791616805893e-06, "loss": 0.4356, "step": 862 }, { "epoch": 0.7010560519902518, "grad_norm": 0.6046944260597229, "learning_rate": 9.466669310467266e-06, "loss": 0.5018, "step": 863 }, { "epoch": 0.7018683996750609, "grad_norm": 0.45827385783195496, "learning_rate": 9.464543011831742e-06, "loss": 0.4651, "step": 864 }, { "epoch": 0.7026807473598701, "grad_norm": 0.4919130504131317, "learning_rate": 9.462412722799801e-06, "loss": 0.4713, "step": 865 }, { "epoch": 0.7034930950446792, "grad_norm": 0.5300809144973755, "learning_rate": 9.46027844527549e-06, "loss": 0.4815, "step": 866 }, { "epoch": 0.7043054427294883, "grad_norm": 0.4729782044887543, "learning_rate": 9.45814018116642e-06, "loss": 0.4781, "step": 867 }, { "epoch": 0.7051177904142973, "grad_norm": 0.6003774404525757, "learning_rate": 9.455997932383766e-06, "loss": 0.4786, "step": 868 }, { "epoch": 0.7059301380991064, "grad_norm": 0.46799442172050476, "learning_rate": 9.453851700842262e-06, "loss": 0.4423, "step": 869 }, { "epoch": 0.7067424857839155, "grad_norm": 0.4956166446208954, "learning_rate": 9.451701488460207e-06, "loss": 0.4758, "step": 870 }, { "epoch": 0.7075548334687246, "grad_norm": 0.48664286732673645, "learning_rate": 9.449547297159453e-06, "loss": 0.5018, "step": 871 }, { "epoch": 0.7083671811535337, "grad_norm": 0.4308539032936096, "learning_rate": 9.447389128865413e-06, "loss": 0.4804, "step": 872 }, { "epoch": 0.7091795288383428, "grad_norm": 0.5070037841796875, "learning_rate": 9.445226985507047e-06, "loss": 0.478, "step": 873 }, { "epoch": 0.7099918765231519, "grad_norm": 0.47911974787712097, "learning_rate": 9.44306086901688e-06, "loss": 0.4659, "step": 874 }, { "epoch": 0.710804224207961, "grad_norm": 0.4171517789363861, "learning_rate": 9.440890781330974e-06, "loss": 0.4879, "step": 875 }, { "epoch": 0.7116165718927701, "grad_norm": 0.42292848229408264, "learning_rate": 9.43871672438895e-06, "loss": 0.454, "step": 876 }, { "epoch": 0.7124289195775793, "grad_norm": 0.4484780430793762, "learning_rate": 9.436538700133977e-06, "loss": 0.4644, "step": 877 }, { "epoch": 0.7132412672623883, "grad_norm": 0.4671514630317688, "learning_rate": 9.434356710512763e-06, "loss": 0.4908, "step": 878 }, { "epoch": 0.7140536149471974, "grad_norm": 0.44659608602523804, "learning_rate": 9.432170757475566e-06, "loss": 0.4421, "step": 879 }, { "epoch": 0.7148659626320065, "grad_norm": 0.4430672824382782, "learning_rate": 9.429980842976186e-06, "loss": 0.4479, "step": 880 }, { "epoch": 0.7156783103168156, "grad_norm": 0.4983255863189697, "learning_rate": 9.427786968971961e-06, "loss": 0.5018, "step": 881 }, { "epoch": 0.7164906580016247, "grad_norm": 0.42170456051826477, "learning_rate": 9.425589137423772e-06, "loss": 0.4745, "step": 882 }, { "epoch": 0.7173030056864338, "grad_norm": 0.4231012761592865, "learning_rate": 9.423387350296032e-06, "loss": 0.488, "step": 883 }, { "epoch": 0.7181153533712429, "grad_norm": 0.4280981421470642, "learning_rate": 9.421181609556693e-06, "loss": 0.4786, "step": 884 }, { "epoch": 0.718927701056052, "grad_norm": 0.4528018534183502, "learning_rate": 9.418971917177241e-06, "loss": 0.4628, "step": 885 }, { "epoch": 0.7197400487408611, "grad_norm": 0.48169681429862976, "learning_rate": 9.416758275132693e-06, "loss": 0.4934, "step": 886 }, { "epoch": 0.7205523964256701, "grad_norm": 0.45193174481391907, "learning_rate": 9.414540685401596e-06, "loss": 0.501, "step": 887 }, { "epoch": 0.7213647441104792, "grad_norm": 0.5240832567214966, "learning_rate": 9.412319149966025e-06, "loss": 0.4652, "step": 888 }, { "epoch": 0.7221770917952883, "grad_norm": 0.5229229927062988, "learning_rate": 9.410093670811582e-06, "loss": 0.4924, "step": 889 }, { "epoch": 0.7229894394800975, "grad_norm": 0.5595859885215759, "learning_rate": 9.407864249927396e-06, "loss": 0.5096, "step": 890 }, { "epoch": 0.7238017871649066, "grad_norm": 0.4172767102718353, "learning_rate": 9.405630889306116e-06, "loss": 0.4756, "step": 891 }, { "epoch": 0.7246141348497157, "grad_norm": 0.4398457705974579, "learning_rate": 9.403393590943916e-06, "loss": 0.4843, "step": 892 }, { "epoch": 0.7254264825345248, "grad_norm": 0.503291130065918, "learning_rate": 9.401152356840484e-06, "loss": 0.4821, "step": 893 }, { "epoch": 0.7262388302193339, "grad_norm": 0.426698237657547, "learning_rate": 9.39890718899903e-06, "loss": 0.4598, "step": 894 }, { "epoch": 0.727051177904143, "grad_norm": 0.43238916993141174, "learning_rate": 9.39665808942628e-06, "loss": 0.4927, "step": 895 }, { "epoch": 0.7278635255889521, "grad_norm": 0.4732600748538971, "learning_rate": 9.394405060132473e-06, "loss": 0.474, "step": 896 }, { "epoch": 0.7286758732737612, "grad_norm": 0.4712562561035156, "learning_rate": 9.392148103131358e-06, "loss": 0.4607, "step": 897 }, { "epoch": 0.7294882209585702, "grad_norm": 0.41925346851348877, "learning_rate": 9.3898872204402e-06, "loss": 0.4755, "step": 898 }, { "epoch": 0.7303005686433793, "grad_norm": 0.47284337878227234, "learning_rate": 9.38762241407977e-06, "loss": 0.477, "step": 899 }, { "epoch": 0.7311129163281884, "grad_norm": 0.4541894197463989, "learning_rate": 9.38535368607434e-06, "loss": 0.4774, "step": 900 }, { "epoch": 0.7319252640129975, "grad_norm": 0.4525924026966095, "learning_rate": 9.3830810384517e-06, "loss": 0.4643, "step": 901 }, { "epoch": 0.7327376116978067, "grad_norm": 0.4861850440502167, "learning_rate": 9.380804473243133e-06, "loss": 0.4724, "step": 902 }, { "epoch": 0.7335499593826158, "grad_norm": 0.4486435353755951, "learning_rate": 9.378523992483429e-06, "loss": 0.458, "step": 903 }, { "epoch": 0.7343623070674249, "grad_norm": 0.46833521127700806, "learning_rate": 9.376239598210873e-06, "loss": 0.4724, "step": 904 }, { "epoch": 0.735174654752234, "grad_norm": 0.40782931447029114, "learning_rate": 9.37395129246725e-06, "loss": 0.4602, "step": 905 }, { "epoch": 0.7359870024370431, "grad_norm": 0.5278148651123047, "learning_rate": 9.371659077297843e-06, "loss": 0.486, "step": 906 }, { "epoch": 0.7367993501218522, "grad_norm": 0.40506991744041443, "learning_rate": 9.369362954751428e-06, "loss": 0.4629, "step": 907 }, { "epoch": 0.7376116978066612, "grad_norm": 0.4837368130683899, "learning_rate": 9.367062926880273e-06, "loss": 0.457, "step": 908 }, { "epoch": 0.7384240454914703, "grad_norm": 0.5230798721313477, "learning_rate": 9.364758995740136e-06, "loss": 0.4658, "step": 909 }, { "epoch": 0.7392363931762794, "grad_norm": 0.42029985785484314, "learning_rate": 9.362451163390265e-06, "loss": 0.4536, "step": 910 }, { "epoch": 0.7400487408610885, "grad_norm": 0.496140718460083, "learning_rate": 9.360139431893394e-06, "loss": 0.4582, "step": 911 }, { "epoch": 0.7408610885458976, "grad_norm": 0.46102869510650635, "learning_rate": 9.35782380331574e-06, "loss": 0.4591, "step": 912 }, { "epoch": 0.7416734362307067, "grad_norm": 0.48244136571884155, "learning_rate": 9.35550427972701e-06, "loss": 0.4715, "step": 913 }, { "epoch": 0.7424857839155159, "grad_norm": 0.45504283905029297, "learning_rate": 9.353180863200385e-06, "loss": 0.4387, "step": 914 }, { "epoch": 0.743298131600325, "grad_norm": 0.5061055421829224, "learning_rate": 9.350853555812529e-06, "loss": 0.4508, "step": 915 }, { "epoch": 0.7441104792851341, "grad_norm": 0.41966158151626587, "learning_rate": 9.348522359643583e-06, "loss": 0.4909, "step": 916 }, { "epoch": 0.7449228269699432, "grad_norm": 0.463819682598114, "learning_rate": 9.346187276777163e-06, "loss": 0.4758, "step": 917 }, { "epoch": 0.7457351746547523, "grad_norm": 0.43058907985687256, "learning_rate": 9.34384830930036e-06, "loss": 0.4944, "step": 918 }, { "epoch": 0.7465475223395613, "grad_norm": 0.4928083121776581, "learning_rate": 9.34150545930374e-06, "loss": 0.4539, "step": 919 }, { "epoch": 0.7473598700243704, "grad_norm": 0.4363732933998108, "learning_rate": 9.339158728881332e-06, "loss": 0.4872, "step": 920 }, { "epoch": 0.7481722177091795, "grad_norm": 0.47651129961013794, "learning_rate": 9.336808120130638e-06, "loss": 0.4792, "step": 921 }, { "epoch": 0.7489845653939886, "grad_norm": 0.5103795528411865, "learning_rate": 9.334453635152628e-06, "loss": 0.4749, "step": 922 }, { "epoch": 0.7497969130787977, "grad_norm": 0.4488142728805542, "learning_rate": 9.332095276051729e-06, "loss": 0.476, "step": 923 }, { "epoch": 0.7506092607636068, "grad_norm": 0.5083579421043396, "learning_rate": 9.329733044935843e-06, "loss": 0.4661, "step": 924 }, { "epoch": 0.7514216084484159, "grad_norm": 0.41831204295158386, "learning_rate": 9.32736694391632e-06, "loss": 0.4782, "step": 925 }, { "epoch": 0.752233956133225, "grad_norm": 0.4422960579395294, "learning_rate": 9.324996975107978e-06, "loss": 0.4556, "step": 926 }, { "epoch": 0.7530463038180342, "grad_norm": 0.4336720407009125, "learning_rate": 9.322623140629088e-06, "loss": 0.4641, "step": 927 }, { "epoch": 0.7538586515028433, "grad_norm": 0.6038580536842346, "learning_rate": 9.320245442601377e-06, "loss": 0.4742, "step": 928 }, { "epoch": 0.7546709991876523, "grad_norm": 0.4041363000869751, "learning_rate": 9.317863883150022e-06, "loss": 0.4605, "step": 929 }, { "epoch": 0.7554833468724614, "grad_norm": 0.5121626257896423, "learning_rate": 9.31547846440366e-06, "loss": 0.4836, "step": 930 }, { "epoch": 0.7562956945572705, "grad_norm": 0.3995649814605713, "learning_rate": 9.313089188494366e-06, "loss": 0.4516, "step": 931 }, { "epoch": 0.7571080422420796, "grad_norm": 0.4669332504272461, "learning_rate": 9.31069605755767e-06, "loss": 0.4612, "step": 932 }, { "epoch": 0.7579203899268887, "grad_norm": 0.5015588998794556, "learning_rate": 9.30829907373255e-06, "loss": 0.4766, "step": 933 }, { "epoch": 0.7587327376116978, "grad_norm": 0.47153234481811523, "learning_rate": 9.30589823916142e-06, "loss": 0.4587, "step": 934 }, { "epoch": 0.7595450852965069, "grad_norm": 0.4716636836528778, "learning_rate": 9.303493555990139e-06, "loss": 0.4726, "step": 935 }, { "epoch": 0.760357432981316, "grad_norm": 0.4088805615901947, "learning_rate": 9.301085026368006e-06, "loss": 0.4327, "step": 936 }, { "epoch": 0.7611697806661251, "grad_norm": 0.5383111834526062, "learning_rate": 9.298672652447763e-06, "loss": 0.4595, "step": 937 }, { "epoch": 0.7619821283509342, "grad_norm": 0.4588027894496918, "learning_rate": 9.296256436385576e-06, "loss": 0.4859, "step": 938 }, { "epoch": 0.7627944760357434, "grad_norm": 0.5854985117912292, "learning_rate": 9.293836380341059e-06, "loss": 0.4543, "step": 939 }, { "epoch": 0.7636068237205524, "grad_norm": 0.4917309582233429, "learning_rate": 9.291412486477247e-06, "loss": 0.4731, "step": 940 }, { "epoch": 0.7644191714053615, "grad_norm": 0.6061102151870728, "learning_rate": 9.28898475696061e-06, "loss": 0.466, "step": 941 }, { "epoch": 0.7652315190901706, "grad_norm": 0.47461628913879395, "learning_rate": 9.28655319396105e-06, "loss": 0.4847, "step": 942 }, { "epoch": 0.7660438667749797, "grad_norm": 0.6091678142547607, "learning_rate": 9.284117799651887e-06, "loss": 0.4642, "step": 943 }, { "epoch": 0.7668562144597888, "grad_norm": 0.43186816573143005, "learning_rate": 9.281678576209873e-06, "loss": 0.4389, "step": 944 }, { "epoch": 0.7676685621445979, "grad_norm": 0.5412203073501587, "learning_rate": 9.279235525815177e-06, "loss": 0.4781, "step": 945 }, { "epoch": 0.768480909829407, "grad_norm": 0.5111507177352905, "learning_rate": 9.276788650651392e-06, "loss": 0.4739, "step": 946 }, { "epoch": 0.7692932575142161, "grad_norm": 0.5531342625617981, "learning_rate": 9.274337952905528e-06, "loss": 0.4635, "step": 947 }, { "epoch": 0.7701056051990252, "grad_norm": 0.4713832139968872, "learning_rate": 9.271883434768012e-06, "loss": 0.5138, "step": 948 }, { "epoch": 0.7709179528838342, "grad_norm": 0.5768186450004578, "learning_rate": 9.269425098432686e-06, "loss": 0.481, "step": 949 }, { "epoch": 0.7717303005686433, "grad_norm": 0.5014119148254395, "learning_rate": 9.266962946096802e-06, "loss": 0.4873, "step": 950 }, { "epoch": 0.7725426482534524, "grad_norm": 0.6314865350723267, "learning_rate": 9.264496979961031e-06, "loss": 0.4731, "step": 951 }, { "epoch": 0.7733549959382616, "grad_norm": 0.48227617144584656, "learning_rate": 9.26202720222944e-06, "loss": 0.4529, "step": 952 }, { "epoch": 0.7741673436230707, "grad_norm": 0.527912974357605, "learning_rate": 9.259553615109514e-06, "loss": 0.4869, "step": 953 }, { "epoch": 0.7749796913078798, "grad_norm": 0.5110397934913635, "learning_rate": 9.257076220812136e-06, "loss": 0.4968, "step": 954 }, { "epoch": 0.7757920389926889, "grad_norm": 0.46427398920059204, "learning_rate": 9.254595021551595e-06, "loss": 0.4468, "step": 955 }, { "epoch": 0.776604386677498, "grad_norm": 0.48961153626441956, "learning_rate": 9.252110019545581e-06, "loss": 0.4941, "step": 956 }, { "epoch": 0.7774167343623071, "grad_norm": 0.47760090231895447, "learning_rate": 9.249621217015182e-06, "loss": 0.4654, "step": 957 }, { "epoch": 0.7782290820471162, "grad_norm": 0.4788792133331299, "learning_rate": 9.247128616184881e-06, "loss": 0.4651, "step": 958 }, { "epoch": 0.7790414297319253, "grad_norm": 0.4785853922367096, "learning_rate": 9.244632219282561e-06, "loss": 0.4279, "step": 959 }, { "epoch": 0.7798537774167343, "grad_norm": 0.4722917377948761, "learning_rate": 9.242132028539493e-06, "loss": 0.4702, "step": 960 }, { "epoch": 0.7806661251015434, "grad_norm": 0.4376189410686493, "learning_rate": 9.239628046190342e-06, "loss": 0.4772, "step": 961 }, { "epoch": 0.7814784727863525, "grad_norm": 0.5341284275054932, "learning_rate": 9.237120274473157e-06, "loss": 0.4433, "step": 962 }, { "epoch": 0.7822908204711616, "grad_norm": 0.4427497684955597, "learning_rate": 9.23460871562938e-06, "loss": 0.4806, "step": 963 }, { "epoch": 0.7831031681559708, "grad_norm": 0.5416346788406372, "learning_rate": 9.232093371903836e-06, "loss": 0.4605, "step": 964 }, { "epoch": 0.7839155158407799, "grad_norm": 0.4424426257610321, "learning_rate": 9.229574245544732e-06, "loss": 0.4599, "step": 965 }, { "epoch": 0.784727863525589, "grad_norm": 0.49265486001968384, "learning_rate": 9.227051338803656e-06, "loss": 0.4661, "step": 966 }, { "epoch": 0.7855402112103981, "grad_norm": 0.4661208689212799, "learning_rate": 9.224524653935577e-06, "loss": 0.4685, "step": 967 }, { "epoch": 0.7863525588952072, "grad_norm": 0.5017091631889343, "learning_rate": 9.221994193198835e-06, "loss": 0.4856, "step": 968 }, { "epoch": 0.7871649065800163, "grad_norm": 0.42260468006134033, "learning_rate": 9.219459958855152e-06, "loss": 0.4637, "step": 969 }, { "epoch": 0.7879772542648253, "grad_norm": 0.4668169915676117, "learning_rate": 9.216921953169618e-06, "loss": 0.4522, "step": 970 }, { "epoch": 0.7887896019496344, "grad_norm": 0.4902791976928711, "learning_rate": 9.214380178410698e-06, "loss": 0.4952, "step": 971 }, { "epoch": 0.7896019496344435, "grad_norm": 0.4746381938457489, "learning_rate": 9.211834636850222e-06, "loss": 0.4903, "step": 972 }, { "epoch": 0.7904142973192526, "grad_norm": 0.4090979993343353, "learning_rate": 9.209285330763388e-06, "loss": 0.4461, "step": 973 }, { "epoch": 0.7912266450040617, "grad_norm": 0.49719369411468506, "learning_rate": 9.20673226242876e-06, "loss": 0.4644, "step": 974 }, { "epoch": 0.7920389926888708, "grad_norm": 0.44468948245048523, "learning_rate": 9.204175434128259e-06, "loss": 0.4873, "step": 975 }, { "epoch": 0.7928513403736799, "grad_norm": 0.5491762757301331, "learning_rate": 9.201614848147179e-06, "loss": 0.4855, "step": 976 }, { "epoch": 0.7936636880584891, "grad_norm": 0.4208453595638275, "learning_rate": 9.199050506774159e-06, "loss": 0.4889, "step": 977 }, { "epoch": 0.7944760357432982, "grad_norm": 0.5089733600616455, "learning_rate": 9.196482412301202e-06, "loss": 0.5147, "step": 978 }, { "epoch": 0.7952883834281073, "grad_norm": 0.481764018535614, "learning_rate": 9.193910567023663e-06, "loss": 0.481, "step": 979 }, { "epoch": 0.7961007311129163, "grad_norm": 0.4762742519378662, "learning_rate": 9.19133497324025e-06, "loss": 0.4394, "step": 980 }, { "epoch": 0.7969130787977254, "grad_norm": 0.5030269026756287, "learning_rate": 9.188755633253025e-06, "loss": 0.4802, "step": 981 }, { "epoch": 0.7977254264825345, "grad_norm": 0.48772624135017395, "learning_rate": 9.186172549367394e-06, "loss": 0.4739, "step": 982 }, { "epoch": 0.7985377741673436, "grad_norm": 0.46897926926612854, "learning_rate": 9.183585723892107e-06, "loss": 0.4616, "step": 983 }, { "epoch": 0.7993501218521527, "grad_norm": 0.399720698595047, "learning_rate": 9.180995159139265e-06, "loss": 0.4654, "step": 984 }, { "epoch": 0.8001624695369618, "grad_norm": 0.5188646912574768, "learning_rate": 9.178400857424305e-06, "loss": 0.4727, "step": 985 }, { "epoch": 0.8009748172217709, "grad_norm": 0.4100261330604553, "learning_rate": 9.175802821066009e-06, "loss": 0.4625, "step": 986 }, { "epoch": 0.80178716490658, "grad_norm": 0.45926401019096375, "learning_rate": 9.173201052386496e-06, "loss": 0.4729, "step": 987 }, { "epoch": 0.8025995125913891, "grad_norm": 0.4525834023952484, "learning_rate": 9.170595553711216e-06, "loss": 0.4773, "step": 988 }, { "epoch": 0.8034118602761983, "grad_norm": 0.4062269926071167, "learning_rate": 9.167986327368958e-06, "loss": 0.4554, "step": 989 }, { "epoch": 0.8042242079610074, "grad_norm": 0.4635365605354309, "learning_rate": 9.165373375691845e-06, "loss": 0.4797, "step": 990 }, { "epoch": 0.8050365556458164, "grad_norm": 0.4333793818950653, "learning_rate": 9.16275670101532e-06, "loss": 0.4565, "step": 991 }, { "epoch": 0.8058489033306255, "grad_norm": 0.4416877031326294, "learning_rate": 9.160136305678167e-06, "loss": 0.4856, "step": 992 }, { "epoch": 0.8066612510154346, "grad_norm": 0.43858057260513306, "learning_rate": 9.15751219202248e-06, "loss": 0.4562, "step": 993 }, { "epoch": 0.8074735987002437, "grad_norm": 0.4833579659461975, "learning_rate": 9.15488436239369e-06, "loss": 0.4744, "step": 994 }, { "epoch": 0.8082859463850528, "grad_norm": 0.4725137948989868, "learning_rate": 9.152252819140544e-06, "loss": 0.4757, "step": 995 }, { "epoch": 0.8090982940698619, "grad_norm": 0.3805605471134186, "learning_rate": 9.149617564615106e-06, "loss": 0.4352, "step": 996 }, { "epoch": 0.809910641754671, "grad_norm": 0.4492555260658264, "learning_rate": 9.146978601172761e-06, "loss": 0.4448, "step": 997 }, { "epoch": 0.8107229894394801, "grad_norm": 0.4411538243293762, "learning_rate": 9.144335931172205e-06, "loss": 0.509, "step": 998 }, { "epoch": 0.8115353371242892, "grad_norm": 0.4692782163619995, "learning_rate": 9.141689556975453e-06, "loss": 0.4838, "step": 999 }, { "epoch": 0.8123476848090982, "grad_norm": 0.49009406566619873, "learning_rate": 9.139039480947825e-06, "loss": 0.4653, "step": 1000 }, { "epoch": 0.8131600324939074, "grad_norm": 0.4031979739665985, "learning_rate": 9.136385705457951e-06, "loss": 0.4358, "step": 1001 }, { "epoch": 0.8139723801787165, "grad_norm": 0.5194141268730164, "learning_rate": 9.13372823287777e-06, "loss": 0.4813, "step": 1002 }, { "epoch": 0.8147847278635256, "grad_norm": 0.5044425129890442, "learning_rate": 9.131067065582521e-06, "loss": 0.4886, "step": 1003 }, { "epoch": 0.8155970755483347, "grad_norm": 0.4784109592437744, "learning_rate": 9.128402205950751e-06, "loss": 0.4583, "step": 1004 }, { "epoch": 0.8164094232331438, "grad_norm": 0.5194838643074036, "learning_rate": 9.125733656364304e-06, "loss": 0.4767, "step": 1005 }, { "epoch": 0.8172217709179529, "grad_norm": 0.4444587528705597, "learning_rate": 9.12306141920832e-06, "loss": 0.4876, "step": 1006 }, { "epoch": 0.818034118602762, "grad_norm": 0.449923574924469, "learning_rate": 9.12038549687124e-06, "loss": 0.4776, "step": 1007 }, { "epoch": 0.8188464662875711, "grad_norm": 0.42537087202072144, "learning_rate": 9.117705891744795e-06, "loss": 0.5032, "step": 1008 }, { "epoch": 0.8196588139723802, "grad_norm": 0.47723740339279175, "learning_rate": 9.115022606224008e-06, "loss": 0.4791, "step": 1009 }, { "epoch": 0.8204711616571893, "grad_norm": 0.43412908911705017, "learning_rate": 9.112335642707196e-06, "loss": 0.4724, "step": 1010 }, { "epoch": 0.8212835093419983, "grad_norm": 0.5175872445106506, "learning_rate": 9.109645003595954e-06, "loss": 0.482, "step": 1011 }, { "epoch": 0.8220958570268074, "grad_norm": 0.4394470453262329, "learning_rate": 9.106950691295172e-06, "loss": 0.4774, "step": 1012 }, { "epoch": 0.8229082047116165, "grad_norm": 0.43408331274986267, "learning_rate": 9.104252708213018e-06, "loss": 0.4179, "step": 1013 }, { "epoch": 0.8237205523964257, "grad_norm": 0.6224790811538696, "learning_rate": 9.101551056760942e-06, "loss": 0.473, "step": 1014 }, { "epoch": 0.8245329000812348, "grad_norm": 0.5133543014526367, "learning_rate": 9.098845739353672e-06, "loss": 0.4699, "step": 1015 }, { "epoch": 0.8253452477660439, "grad_norm": 0.5065946578979492, "learning_rate": 9.096136758409212e-06, "loss": 0.5049, "step": 1016 }, { "epoch": 0.826157595450853, "grad_norm": 0.5660102367401123, "learning_rate": 9.093424116348846e-06, "loss": 0.482, "step": 1017 }, { "epoch": 0.8269699431356621, "grad_norm": 0.6450397968292236, "learning_rate": 9.090707815597124e-06, "loss": 0.481, "step": 1018 }, { "epoch": 0.8277822908204712, "grad_norm": 0.5294673442840576, "learning_rate": 9.087987858581867e-06, "loss": 0.4995, "step": 1019 }, { "epoch": 0.8285946385052803, "grad_norm": 0.5429680943489075, "learning_rate": 9.085264247734165e-06, "loss": 0.4647, "step": 1020 }, { "epoch": 0.8294069861900893, "grad_norm": 0.46350201964378357, "learning_rate": 9.082536985488377e-06, "loss": 0.462, "step": 1021 }, { "epoch": 0.8302193338748984, "grad_norm": 0.5006203651428223, "learning_rate": 9.07980607428212e-06, "loss": 0.4625, "step": 1022 }, { "epoch": 0.8310316815597075, "grad_norm": 0.5279543995857239, "learning_rate": 9.077071516556276e-06, "loss": 0.4679, "step": 1023 }, { "epoch": 0.8318440292445166, "grad_norm": 0.4344598948955536, "learning_rate": 9.074333314754984e-06, "loss": 0.4532, "step": 1024 }, { "epoch": 0.8326563769293257, "grad_norm": 0.4854860305786133, "learning_rate": 9.071591471325644e-06, "loss": 0.4634, "step": 1025 }, { "epoch": 0.8334687246141349, "grad_norm": 0.565533459186554, "learning_rate": 9.068845988718906e-06, "loss": 0.4737, "step": 1026 }, { "epoch": 0.834281072298944, "grad_norm": 0.5328646898269653, "learning_rate": 9.066096869388674e-06, "loss": 0.448, "step": 1027 }, { "epoch": 0.8350934199837531, "grad_norm": 0.532748818397522, "learning_rate": 9.063344115792108e-06, "loss": 0.4532, "step": 1028 }, { "epoch": 0.8359057676685622, "grad_norm": 0.5555665493011475, "learning_rate": 9.060587730389606e-06, "loss": 0.481, "step": 1029 }, { "epoch": 0.8367181153533713, "grad_norm": 0.4458980858325958, "learning_rate": 9.05782771564482e-06, "loss": 0.48, "step": 1030 }, { "epoch": 0.8375304630381804, "grad_norm": 0.4995458126068115, "learning_rate": 9.055064074024646e-06, "loss": 0.5021, "step": 1031 }, { "epoch": 0.8383428107229894, "grad_norm": 0.5369895100593567, "learning_rate": 9.052296807999216e-06, "loss": 0.4567, "step": 1032 }, { "epoch": 0.8391551584077985, "grad_norm": 0.5668280124664307, "learning_rate": 9.049525920041909e-06, "loss": 0.4997, "step": 1033 }, { "epoch": 0.8399675060926076, "grad_norm": 0.5200647711753845, "learning_rate": 9.046751412629331e-06, "loss": 0.4664, "step": 1034 }, { "epoch": 0.8407798537774167, "grad_norm": 0.6480557918548584, "learning_rate": 9.043973288241334e-06, "loss": 0.4723, "step": 1035 }, { "epoch": 0.8415922014622258, "grad_norm": 0.5532544255256653, "learning_rate": 9.041191549360995e-06, "loss": 0.4714, "step": 1036 }, { "epoch": 0.8424045491470349, "grad_norm": 0.48528942465782166, "learning_rate": 9.038406198474627e-06, "loss": 0.4551, "step": 1037 }, { "epoch": 0.843216896831844, "grad_norm": 0.48096606135368347, "learning_rate": 9.03561723807177e-06, "loss": 0.4681, "step": 1038 }, { "epoch": 0.8440292445166532, "grad_norm": 0.5365792512893677, "learning_rate": 9.032824670645187e-06, "loss": 0.4586, "step": 1039 }, { "epoch": 0.8448415922014623, "grad_norm": 0.5250015258789062, "learning_rate": 9.030028498690866e-06, "loss": 0.5043, "step": 1040 }, { "epoch": 0.8456539398862714, "grad_norm": 0.46791160106658936, "learning_rate": 9.027228724708022e-06, "loss": 0.4653, "step": 1041 }, { "epoch": 0.8464662875710804, "grad_norm": 0.4984607994556427, "learning_rate": 9.02442535119908e-06, "loss": 0.4403, "step": 1042 }, { "epoch": 0.8472786352558895, "grad_norm": 0.48224472999572754, "learning_rate": 9.021618380669693e-06, "loss": 0.4393, "step": 1043 }, { "epoch": 0.8480909829406986, "grad_norm": 0.481342077255249, "learning_rate": 9.018807815628721e-06, "loss": 0.4512, "step": 1044 }, { "epoch": 0.8489033306255077, "grad_norm": 0.4520586431026459, "learning_rate": 9.01599365858824e-06, "loss": 0.4531, "step": 1045 }, { "epoch": 0.8497156783103168, "grad_norm": 0.47431260347366333, "learning_rate": 9.013175912063534e-06, "loss": 0.4537, "step": 1046 }, { "epoch": 0.8505280259951259, "grad_norm": 0.4946224093437195, "learning_rate": 9.0103545785731e-06, "loss": 0.4565, "step": 1047 }, { "epoch": 0.851340373679935, "grad_norm": 0.4639143645763397, "learning_rate": 9.007529660638637e-06, "loss": 0.4454, "step": 1048 }, { "epoch": 0.8521527213647441, "grad_norm": 0.5948116183280945, "learning_rate": 9.004701160785047e-06, "loss": 0.4702, "step": 1049 }, { "epoch": 0.8529650690495532, "grad_norm": 0.5105404257774353, "learning_rate": 9.00186908154044e-06, "loss": 0.4787, "step": 1050 }, { "epoch": 0.8537774167343624, "grad_norm": 0.4611700773239136, "learning_rate": 8.999033425436116e-06, "loss": 0.4757, "step": 1051 }, { "epoch": 0.8545897644191714, "grad_norm": 0.52210533618927, "learning_rate": 8.996194195006582e-06, "loss": 0.4614, "step": 1052 }, { "epoch": 0.8554021121039805, "grad_norm": 0.4310266673564911, "learning_rate": 8.993351392789529e-06, "loss": 0.4649, "step": 1053 }, { "epoch": 0.8562144597887896, "grad_norm": 0.5137235522270203, "learning_rate": 8.990505021325849e-06, "loss": 0.4712, "step": 1054 }, { "epoch": 0.8570268074735987, "grad_norm": 0.4465412497520447, "learning_rate": 8.987655083159618e-06, "loss": 0.4996, "step": 1055 }, { "epoch": 0.8578391551584078, "grad_norm": 0.49357977509498596, "learning_rate": 8.984801580838109e-06, "loss": 0.4832, "step": 1056 }, { "epoch": 0.8586515028432169, "grad_norm": 0.41144832968711853, "learning_rate": 8.98194451691177e-06, "loss": 0.4441, "step": 1057 }, { "epoch": 0.859463850528026, "grad_norm": 0.42749401926994324, "learning_rate": 8.979083893934236e-06, "loss": 0.4745, "step": 1058 }, { "epoch": 0.8602761982128351, "grad_norm": 0.5273240208625793, "learning_rate": 8.976219714462326e-06, "loss": 0.516, "step": 1059 }, { "epoch": 0.8610885458976442, "grad_norm": 0.43741098046302795, "learning_rate": 8.973351981056037e-06, "loss": 0.4689, "step": 1060 }, { "epoch": 0.8619008935824533, "grad_norm": 0.6278446912765503, "learning_rate": 8.97048069627854e-06, "loss": 0.4713, "step": 1061 }, { "epoch": 0.8627132412672623, "grad_norm": 0.4374808669090271, "learning_rate": 8.967605862696183e-06, "loss": 0.4891, "step": 1062 }, { "epoch": 0.8635255889520714, "grad_norm": 0.48301851749420166, "learning_rate": 8.964727482878482e-06, "loss": 0.4453, "step": 1063 }, { "epoch": 0.8643379366368806, "grad_norm": 0.45663905143737793, "learning_rate": 8.961845559398125e-06, "loss": 0.4582, "step": 1064 }, { "epoch": 0.8651502843216897, "grad_norm": 0.47542116045951843, "learning_rate": 8.95896009483097e-06, "loss": 0.467, "step": 1065 }, { "epoch": 0.8659626320064988, "grad_norm": 0.5842065215110779, "learning_rate": 8.956071091756036e-06, "loss": 0.4723, "step": 1066 }, { "epoch": 0.8667749796913079, "grad_norm": 0.4242716431617737, "learning_rate": 8.953178552755506e-06, "loss": 0.4837, "step": 1067 }, { "epoch": 0.867587327376117, "grad_norm": 0.49736928939819336, "learning_rate": 8.950282480414723e-06, "loss": 0.4955, "step": 1068 }, { "epoch": 0.8683996750609261, "grad_norm": 0.4428938031196594, "learning_rate": 8.947382877322189e-06, "loss": 0.4858, "step": 1069 }, { "epoch": 0.8692120227457352, "grad_norm": 0.46457770466804504, "learning_rate": 8.94447974606956e-06, "loss": 0.4677, "step": 1070 }, { "epoch": 0.8700243704305443, "grad_norm": 0.47255438566207886, "learning_rate": 8.941573089251652e-06, "loss": 0.4588, "step": 1071 }, { "epoch": 0.8708367181153533, "grad_norm": 0.4545036554336548, "learning_rate": 8.93866290946642e-06, "loss": 0.4763, "step": 1072 }, { "epoch": 0.8716490658001624, "grad_norm": 0.4806216061115265, "learning_rate": 8.935749209314981e-06, "loss": 0.4926, "step": 1073 }, { "epoch": 0.8724614134849715, "grad_norm": 0.5091702342033386, "learning_rate": 8.932831991401587e-06, "loss": 0.4978, "step": 1074 }, { "epoch": 0.8732737611697806, "grad_norm": 0.5218701958656311, "learning_rate": 8.929911258333644e-06, "loss": 0.4695, "step": 1075 }, { "epoch": 0.8740861088545898, "grad_norm": 0.5394851565361023, "learning_rate": 8.92698701272169e-06, "loss": 0.4783, "step": 1076 }, { "epoch": 0.8748984565393989, "grad_norm": 0.48874491453170776, "learning_rate": 8.924059257179414e-06, "loss": 0.469, "step": 1077 }, { "epoch": 0.875710804224208, "grad_norm": 0.4150269031524658, "learning_rate": 8.92112799432363e-06, "loss": 0.4946, "step": 1078 }, { "epoch": 0.8765231519090171, "grad_norm": 0.5228343605995178, "learning_rate": 8.9181932267743e-06, "loss": 0.4699, "step": 1079 }, { "epoch": 0.8773354995938262, "grad_norm": 0.5170260667800903, "learning_rate": 8.915254957154501e-06, "loss": 0.4841, "step": 1080 }, { "epoch": 0.8781478472786353, "grad_norm": 0.4113463759422302, "learning_rate": 8.91231318809046e-06, "loss": 0.4661, "step": 1081 }, { "epoch": 0.8789601949634444, "grad_norm": 0.4943483769893646, "learning_rate": 8.909367922211519e-06, "loss": 0.5024, "step": 1082 }, { "epoch": 0.8797725426482534, "grad_norm": 0.48948386311531067, "learning_rate": 8.906419162150147e-06, "loss": 0.4584, "step": 1083 }, { "epoch": 0.8805848903330625, "grad_norm": 0.4660533666610718, "learning_rate": 8.903466910541938e-06, "loss": 0.4873, "step": 1084 }, { "epoch": 0.8813972380178716, "grad_norm": 0.48683106899261475, "learning_rate": 8.900511170025609e-06, "loss": 0.4731, "step": 1085 }, { "epoch": 0.8822095857026807, "grad_norm": 0.45988956093788147, "learning_rate": 8.89755194324299e-06, "loss": 0.4584, "step": 1086 }, { "epoch": 0.8830219333874898, "grad_norm": 0.4323303997516632, "learning_rate": 8.89458923283903e-06, "loss": 0.4448, "step": 1087 }, { "epoch": 0.8838342810722989, "grad_norm": 0.4391065835952759, "learning_rate": 8.891623041461794e-06, "loss": 0.4634, "step": 1088 }, { "epoch": 0.8846466287571081, "grad_norm": 0.46511340141296387, "learning_rate": 8.888653371762454e-06, "loss": 0.485, "step": 1089 }, { "epoch": 0.8854589764419172, "grad_norm": 0.45805642008781433, "learning_rate": 8.885680226395295e-06, "loss": 0.4887, "step": 1090 }, { "epoch": 0.8862713241267263, "grad_norm": 0.44465896487236023, "learning_rate": 8.882703608017702e-06, "loss": 0.4613, "step": 1091 }, { "epoch": 0.8870836718115354, "grad_norm": 0.44617605209350586, "learning_rate": 8.87972351929017e-06, "loss": 0.4533, "step": 1092 }, { "epoch": 0.8878960194963444, "grad_norm": 0.44850581884384155, "learning_rate": 8.876739962876298e-06, "loss": 0.4901, "step": 1093 }, { "epoch": 0.8887083671811535, "grad_norm": 0.43714144825935364, "learning_rate": 8.873752941442775e-06, "loss": 0.4552, "step": 1094 }, { "epoch": 0.8895207148659626, "grad_norm": 0.5268635153770447, "learning_rate": 8.870762457659397e-06, "loss": 0.4833, "step": 1095 }, { "epoch": 0.8903330625507717, "grad_norm": 0.5036064982414246, "learning_rate": 8.867768514199046e-06, "loss": 0.4701, "step": 1096 }, { "epoch": 0.8911454102355808, "grad_norm": 0.5106784701347351, "learning_rate": 8.864771113737705e-06, "loss": 0.4671, "step": 1097 }, { "epoch": 0.8919577579203899, "grad_norm": 0.47654372453689575, "learning_rate": 8.86177025895444e-06, "loss": 0.4583, "step": 1098 }, { "epoch": 0.892770105605199, "grad_norm": 0.4870198667049408, "learning_rate": 8.858765952531407e-06, "loss": 0.4542, "step": 1099 }, { "epoch": 0.8935824532900081, "grad_norm": 0.4363579750061035, "learning_rate": 8.855758197153848e-06, "loss": 0.4549, "step": 1100 }, { "epoch": 0.8943948009748173, "grad_norm": 0.4570150673389435, "learning_rate": 8.852746995510084e-06, "loss": 0.4688, "step": 1101 }, { "epoch": 0.8952071486596264, "grad_norm": 0.4599553942680359, "learning_rate": 8.849732350291522e-06, "loss": 0.4765, "step": 1102 }, { "epoch": 0.8960194963444355, "grad_norm": 0.40944716334342957, "learning_rate": 8.846714264192642e-06, "loss": 0.4726, "step": 1103 }, { "epoch": 0.8968318440292445, "grad_norm": 0.4749741852283478, "learning_rate": 8.843692739911e-06, "loss": 0.4638, "step": 1104 }, { "epoch": 0.8976441917140536, "grad_norm": 0.4315512180328369, "learning_rate": 8.84066778014723e-06, "loss": 0.4563, "step": 1105 }, { "epoch": 0.8984565393988627, "grad_norm": 0.42155149579048157, "learning_rate": 8.837639387605031e-06, "loss": 0.43, "step": 1106 }, { "epoch": 0.8992688870836718, "grad_norm": 0.4843577444553375, "learning_rate": 8.834607564991168e-06, "loss": 0.4961, "step": 1107 }, { "epoch": 0.9000812347684809, "grad_norm": 0.4193378984928131, "learning_rate": 8.831572315015484e-06, "loss": 0.4351, "step": 1108 }, { "epoch": 0.90089358245329, "grad_norm": 0.43937820196151733, "learning_rate": 8.828533640390868e-06, "loss": 0.4657, "step": 1109 }, { "epoch": 0.9017059301380991, "grad_norm": 0.4227564334869385, "learning_rate": 8.825491543833286e-06, "loss": 0.4764, "step": 1110 }, { "epoch": 0.9025182778229082, "grad_norm": 0.47147610783576965, "learning_rate": 8.822446028061752e-06, "loss": 0.4755, "step": 1111 }, { "epoch": 0.9033306255077173, "grad_norm": 0.47162628173828125, "learning_rate": 8.819397095798343e-06, "loss": 0.4425, "step": 1112 }, { "epoch": 0.9041429731925265, "grad_norm": 0.40200889110565186, "learning_rate": 8.816344749768184e-06, "loss": 0.4576, "step": 1113 }, { "epoch": 0.9049553208773355, "grad_norm": 0.5315070748329163, "learning_rate": 8.813288992699454e-06, "loss": 0.4474, "step": 1114 }, { "epoch": 0.9057676685621446, "grad_norm": 0.513977587223053, "learning_rate": 8.810229827323378e-06, "loss": 0.466, "step": 1115 }, { "epoch": 0.9065800162469537, "grad_norm": 0.438178688287735, "learning_rate": 8.807167256374234e-06, "loss": 0.4184, "step": 1116 }, { "epoch": 0.9073923639317628, "grad_norm": 0.46907955408096313, "learning_rate": 8.804101282589338e-06, "loss": 0.4855, "step": 1117 }, { "epoch": 0.9082047116165719, "grad_norm": 0.50333172082901, "learning_rate": 8.801031908709047e-06, "loss": 0.4762, "step": 1118 }, { "epoch": 0.909017059301381, "grad_norm": 0.5059404969215393, "learning_rate": 8.797959137476763e-06, "loss": 0.458, "step": 1119 }, { "epoch": 0.9098294069861901, "grad_norm": 0.47010621428489685, "learning_rate": 8.794882971638916e-06, "loss": 0.4661, "step": 1120 }, { "epoch": 0.9106417546709992, "grad_norm": 0.5131338834762573, "learning_rate": 8.791803413944979e-06, "loss": 0.4717, "step": 1121 }, { "epoch": 0.9114541023558083, "grad_norm": 0.5202389359474182, "learning_rate": 8.78872046714745e-06, "loss": 0.5031, "step": 1122 }, { "epoch": 0.9122664500406173, "grad_norm": 0.4469550848007202, "learning_rate": 8.785634134001856e-06, "loss": 0.4455, "step": 1123 }, { "epoch": 0.9130787977254264, "grad_norm": 0.6121103167533875, "learning_rate": 8.782544417266758e-06, "loss": 0.449, "step": 1124 }, { "epoch": 0.9138911454102355, "grad_norm": 0.4597553014755249, "learning_rate": 8.779451319703735e-06, "loss": 0.4719, "step": 1125 }, { "epoch": 0.9147034930950447, "grad_norm": 0.4832404851913452, "learning_rate": 8.776354844077389e-06, "loss": 0.4398, "step": 1126 }, { "epoch": 0.9155158407798538, "grad_norm": 0.4932108521461487, "learning_rate": 8.77325499315534e-06, "loss": 0.5035, "step": 1127 }, { "epoch": 0.9163281884646629, "grad_norm": 0.5025442242622375, "learning_rate": 8.770151769708226e-06, "loss": 0.4627, "step": 1128 }, { "epoch": 0.917140536149472, "grad_norm": 0.4388456642627716, "learning_rate": 8.767045176509703e-06, "loss": 0.4818, "step": 1129 }, { "epoch": 0.9179528838342811, "grad_norm": 0.5696482062339783, "learning_rate": 8.763935216336433e-06, "loss": 0.4511, "step": 1130 }, { "epoch": 0.9187652315190902, "grad_norm": 0.4494622051715851, "learning_rate": 8.76082189196809e-06, "loss": 0.4584, "step": 1131 }, { "epoch": 0.9195775792038993, "grad_norm": 0.4613010585308075, "learning_rate": 8.757705206187357e-06, "loss": 0.4421, "step": 1132 }, { "epoch": 0.9203899268887084, "grad_norm": 0.5003172159194946, "learning_rate": 8.754585161779916e-06, "loss": 0.4897, "step": 1133 }, { "epoch": 0.9212022745735174, "grad_norm": 0.49380362033843994, "learning_rate": 8.751461761534457e-06, "loss": 0.4473, "step": 1134 }, { "epoch": 0.9220146222583265, "grad_norm": 0.5670063495635986, "learning_rate": 8.748335008242667e-06, "loss": 0.5035, "step": 1135 }, { "epoch": 0.9228269699431356, "grad_norm": 0.5831010937690735, "learning_rate": 8.745204904699226e-06, "loss": 0.4539, "step": 1136 }, { "epoch": 0.9236393176279447, "grad_norm": 0.5403640866279602, "learning_rate": 8.742071453701815e-06, "loss": 0.4789, "step": 1137 }, { "epoch": 0.9244516653127539, "grad_norm": 0.5296802520751953, "learning_rate": 8.738934658051103e-06, "loss": 0.5005, "step": 1138 }, { "epoch": 0.925264012997563, "grad_norm": 0.6219888925552368, "learning_rate": 8.73579452055075e-06, "loss": 0.4685, "step": 1139 }, { "epoch": 0.9260763606823721, "grad_norm": 0.44621542096138, "learning_rate": 8.732651044007402e-06, "loss": 0.4444, "step": 1140 }, { "epoch": 0.9268887083671812, "grad_norm": 0.6241264939308167, "learning_rate": 8.72950423123069e-06, "loss": 0.4538, "step": 1141 }, { "epoch": 0.9277010560519903, "grad_norm": 0.609604001045227, "learning_rate": 8.726354085033227e-06, "loss": 0.4951, "step": 1142 }, { "epoch": 0.9285134037367994, "grad_norm": 0.5452231168746948, "learning_rate": 8.723200608230605e-06, "loss": 0.4492, "step": 1143 }, { "epoch": 0.9293257514216084, "grad_norm": 0.6668962836265564, "learning_rate": 8.72004380364139e-06, "loss": 0.4582, "step": 1144 }, { "epoch": 0.9301380991064175, "grad_norm": 0.5767805576324463, "learning_rate": 8.71688367408713e-06, "loss": 0.4684, "step": 1145 }, { "epoch": 0.9309504467912266, "grad_norm": 0.5945210456848145, "learning_rate": 8.713720222392338e-06, "loss": 0.4777, "step": 1146 }, { "epoch": 0.9317627944760357, "grad_norm": 0.49358800053596497, "learning_rate": 8.710553451384499e-06, "loss": 0.449, "step": 1147 }, { "epoch": 0.9325751421608448, "grad_norm": 0.6295574307441711, "learning_rate": 8.707383363894064e-06, "loss": 0.4634, "step": 1148 }, { "epoch": 0.9333874898456539, "grad_norm": 0.5162637233734131, "learning_rate": 8.704209962754448e-06, "loss": 0.4476, "step": 1149 }, { "epoch": 0.934199837530463, "grad_norm": 0.6357398629188538, "learning_rate": 8.70103325080203e-06, "loss": 0.4563, "step": 1150 }, { "epoch": 0.9350121852152722, "grad_norm": 0.4998912215232849, "learning_rate": 8.697853230876145e-06, "loss": 0.4908, "step": 1151 }, { "epoch": 0.9358245329000813, "grad_norm": 0.5546163320541382, "learning_rate": 8.694669905819087e-06, "loss": 0.4496, "step": 1152 }, { "epoch": 0.9366368805848904, "grad_norm": 0.5610288381576538, "learning_rate": 8.691483278476103e-06, "loss": 0.4522, "step": 1153 }, { "epoch": 0.9374492282696995, "grad_norm": 0.5190321803092957, "learning_rate": 8.688293351695392e-06, "loss": 0.4547, "step": 1154 }, { "epoch": 0.9382615759545085, "grad_norm": 0.584850549697876, "learning_rate": 8.685100128328101e-06, "loss": 0.4406, "step": 1155 }, { "epoch": 0.9390739236393176, "grad_norm": 0.5948938131332397, "learning_rate": 8.681903611228327e-06, "loss": 0.4691, "step": 1156 }, { "epoch": 0.9398862713241267, "grad_norm": 0.49660372734069824, "learning_rate": 8.678703803253103e-06, "loss": 0.4622, "step": 1157 }, { "epoch": 0.9406986190089358, "grad_norm": 0.5179651379585266, "learning_rate": 8.675500707262415e-06, "loss": 0.4703, "step": 1158 }, { "epoch": 0.9415109666937449, "grad_norm": 0.4324522912502289, "learning_rate": 8.672294326119177e-06, "loss": 0.4568, "step": 1159 }, { "epoch": 0.942323314378554, "grad_norm": 0.4287954568862915, "learning_rate": 8.669084662689246e-06, "loss": 0.4523, "step": 1160 }, { "epoch": 0.9431356620633631, "grad_norm": 0.45402973890304565, "learning_rate": 8.66587171984141e-06, "loss": 0.4636, "step": 1161 }, { "epoch": 0.9439480097481722, "grad_norm": 0.4222104847431183, "learning_rate": 8.662655500447388e-06, "loss": 0.4803, "step": 1162 }, { "epoch": 0.9447603574329814, "grad_norm": 0.42634955048561096, "learning_rate": 8.65943600738183e-06, "loss": 0.4652, "step": 1163 }, { "epoch": 0.9455727051177905, "grad_norm": 0.4743734896183014, "learning_rate": 8.656213243522307e-06, "loss": 0.4749, "step": 1164 }, { "epoch": 0.9463850528025995, "grad_norm": 0.4085732698440552, "learning_rate": 8.65298721174932e-06, "loss": 0.4592, "step": 1165 }, { "epoch": 0.9471974004874086, "grad_norm": 0.4623430371284485, "learning_rate": 8.649757914946284e-06, "loss": 0.4304, "step": 1166 }, { "epoch": 0.9480097481722177, "grad_norm": 0.4583539664745331, "learning_rate": 8.646525355999538e-06, "loss": 0.4686, "step": 1167 }, { "epoch": 0.9488220958570268, "grad_norm": 0.42183730006217957, "learning_rate": 8.643289537798332e-06, "loss": 0.4649, "step": 1168 }, { "epoch": 0.9496344435418359, "grad_norm": 0.4787953495979309, "learning_rate": 8.640050463234837e-06, "loss": 0.4448, "step": 1169 }, { "epoch": 0.950446791226645, "grad_norm": 0.4666788876056671, "learning_rate": 8.636808135204123e-06, "loss": 0.513, "step": 1170 }, { "epoch": 0.9512591389114541, "grad_norm": 0.4887382984161377, "learning_rate": 8.633562556604177e-06, "loss": 0.486, "step": 1171 }, { "epoch": 0.9520714865962632, "grad_norm": 0.3983224630355835, "learning_rate": 8.630313730335888e-06, "loss": 0.4558, "step": 1172 }, { "epoch": 0.9528838342810723, "grad_norm": 0.5573809742927551, "learning_rate": 8.627061659303047e-06, "loss": 0.4378, "step": 1173 }, { "epoch": 0.9536961819658814, "grad_norm": 0.4736296534538269, "learning_rate": 8.623806346412349e-06, "loss": 0.4673, "step": 1174 }, { "epoch": 0.9545085296506904, "grad_norm": 0.4609343111515045, "learning_rate": 8.62054779457338e-06, "loss": 0.4607, "step": 1175 }, { "epoch": 0.9553208773354996, "grad_norm": 0.48708027601242065, "learning_rate": 8.617286006698628e-06, "loss": 0.4338, "step": 1176 }, { "epoch": 0.9561332250203087, "grad_norm": 0.5181538462638855, "learning_rate": 8.614020985703468e-06, "loss": 0.4693, "step": 1177 }, { "epoch": 0.9569455727051178, "grad_norm": 0.4115943908691406, "learning_rate": 8.610752734506166e-06, "loss": 0.4805, "step": 1178 }, { "epoch": 0.9577579203899269, "grad_norm": 0.5168994665145874, "learning_rate": 8.60748125602788e-06, "loss": 0.4549, "step": 1179 }, { "epoch": 0.958570268074736, "grad_norm": 0.43316850066185, "learning_rate": 8.604206553192647e-06, "loss": 0.4552, "step": 1180 }, { "epoch": 0.9593826157595451, "grad_norm": 0.4152318239212036, "learning_rate": 8.600928628927385e-06, "loss": 0.4611, "step": 1181 }, { "epoch": 0.9601949634443542, "grad_norm": 0.43464353680610657, "learning_rate": 8.597647486161898e-06, "loss": 0.4789, "step": 1182 }, { "epoch": 0.9610073111291633, "grad_norm": 0.4258009195327759, "learning_rate": 8.594363127828858e-06, "loss": 0.468, "step": 1183 }, { "epoch": 0.9618196588139724, "grad_norm": 0.45545142889022827, "learning_rate": 8.591075556863818e-06, "loss": 0.4931, "step": 1184 }, { "epoch": 0.9626320064987814, "grad_norm": 0.46154266595840454, "learning_rate": 8.5877847762052e-06, "loss": 0.4944, "step": 1185 }, { "epoch": 0.9634443541835905, "grad_norm": 0.4376266598701477, "learning_rate": 8.584490788794296e-06, "loss": 0.4481, "step": 1186 }, { "epoch": 0.9642567018683996, "grad_norm": 0.4428999423980713, "learning_rate": 8.581193597575259e-06, "loss": 0.4704, "step": 1187 }, { "epoch": 0.9650690495532088, "grad_norm": 0.548171877861023, "learning_rate": 8.577893205495113e-06, "loss": 0.4866, "step": 1188 }, { "epoch": 0.9658813972380179, "grad_norm": 0.4488280117511749, "learning_rate": 8.574589615503739e-06, "loss": 0.4651, "step": 1189 }, { "epoch": 0.966693744922827, "grad_norm": 0.5105303525924683, "learning_rate": 8.571282830553875e-06, "loss": 0.4571, "step": 1190 }, { "epoch": 0.9675060926076361, "grad_norm": 0.5015298128128052, "learning_rate": 8.567972853601119e-06, "loss": 0.4561, "step": 1191 }, { "epoch": 0.9683184402924452, "grad_norm": 0.4165172576904297, "learning_rate": 8.564659687603917e-06, "loss": 0.4413, "step": 1192 }, { "epoch": 0.9691307879772543, "grad_norm": 0.46881353855133057, "learning_rate": 8.56134333552357e-06, "loss": 0.4654, "step": 1193 }, { "epoch": 0.9699431356620634, "grad_norm": 0.49607816338539124, "learning_rate": 8.558023800324223e-06, "loss": 0.4588, "step": 1194 }, { "epoch": 0.9707554833468724, "grad_norm": 0.4980422854423523, "learning_rate": 8.554701084972869e-06, "loss": 0.4911, "step": 1195 }, { "epoch": 0.9715678310316815, "grad_norm": 0.4860203266143799, "learning_rate": 8.551375192439339e-06, "loss": 0.4769, "step": 1196 }, { "epoch": 0.9723801787164906, "grad_norm": 0.48375338315963745, "learning_rate": 8.54804612569631e-06, "loss": 0.4758, "step": 1197 }, { "epoch": 0.9731925264012997, "grad_norm": 0.49457746744155884, "learning_rate": 8.54471388771929e-06, "loss": 0.4697, "step": 1198 }, { "epoch": 0.9740048740861088, "grad_norm": 0.4287789463996887, "learning_rate": 8.541378481486627e-06, "loss": 0.449, "step": 1199 }, { "epoch": 0.974817221770918, "grad_norm": 0.48341500759124756, "learning_rate": 8.538039909979496e-06, "loss": 0.4525, "step": 1200 }, { "epoch": 0.9756295694557271, "grad_norm": 0.42321163415908813, "learning_rate": 8.534698176181904e-06, "loss": 0.4254, "step": 1201 }, { "epoch": 0.9764419171405362, "grad_norm": 0.4348788559436798, "learning_rate": 8.531353283080683e-06, "loss": 0.475, "step": 1202 }, { "epoch": 0.9772542648253453, "grad_norm": 0.5007128119468689, "learning_rate": 8.528005233665491e-06, "loss": 0.4671, "step": 1203 }, { "epoch": 0.9780666125101544, "grad_norm": 0.46965497732162476, "learning_rate": 8.524654030928802e-06, "loss": 0.4422, "step": 1204 }, { "epoch": 0.9788789601949635, "grad_norm": 0.5227495431900024, "learning_rate": 8.521299677865915e-06, "loss": 0.4676, "step": 1205 }, { "epoch": 0.9796913078797725, "grad_norm": 0.4029655158519745, "learning_rate": 8.517942177474943e-06, "loss": 0.4585, "step": 1206 }, { "epoch": 0.9805036555645816, "grad_norm": 0.4677031338214874, "learning_rate": 8.514581532756806e-06, "loss": 0.4694, "step": 1207 }, { "epoch": 0.9813160032493907, "grad_norm": 0.4862461984157562, "learning_rate": 8.511217746715243e-06, "loss": 0.4774, "step": 1208 }, { "epoch": 0.9821283509341998, "grad_norm": 0.406648188829422, "learning_rate": 8.507850822356796e-06, "loss": 0.4497, "step": 1209 }, { "epoch": 0.9829406986190089, "grad_norm": 0.4892263412475586, "learning_rate": 8.50448076269081e-06, "loss": 0.4572, "step": 1210 }, { "epoch": 0.983753046303818, "grad_norm": 0.39997056126594543, "learning_rate": 8.50110757072944e-06, "loss": 0.4423, "step": 1211 }, { "epoch": 0.9845653939886271, "grad_norm": 0.4816911220550537, "learning_rate": 8.497731249487632e-06, "loss": 0.4719, "step": 1212 }, { "epoch": 0.9853777416734363, "grad_norm": 0.4153689444065094, "learning_rate": 8.494351801983135e-06, "loss": 0.4374, "step": 1213 }, { "epoch": 0.9861900893582454, "grad_norm": 0.43972155451774597, "learning_rate": 8.490969231236487e-06, "loss": 0.449, "step": 1214 }, { "epoch": 0.9870024370430545, "grad_norm": 0.4446089267730713, "learning_rate": 8.487583540271024e-06, "loss": 0.4475, "step": 1215 }, { "epoch": 0.9878147847278635, "grad_norm": 0.4060399830341339, "learning_rate": 8.484194732112866e-06, "loss": 0.4485, "step": 1216 }, { "epoch": 0.9886271324126726, "grad_norm": 0.44879940152168274, "learning_rate": 8.480802809790922e-06, "loss": 0.4486, "step": 1217 }, { "epoch": 0.9894394800974817, "grad_norm": 0.4342731833457947, "learning_rate": 8.47740777633688e-06, "loss": 0.4699, "step": 1218 }, { "epoch": 0.9902518277822908, "grad_norm": 0.5003742575645447, "learning_rate": 8.474009634785214e-06, "loss": 0.4863, "step": 1219 }, { "epoch": 0.9910641754670999, "grad_norm": 0.4703499376773834, "learning_rate": 8.47060838817317e-06, "loss": 0.453, "step": 1220 }, { "epoch": 0.991876523151909, "grad_norm": 0.48684343695640564, "learning_rate": 8.467204039540779e-06, "loss": 0.4712, "step": 1221 }, { "epoch": 0.9926888708367181, "grad_norm": 0.435373455286026, "learning_rate": 8.463796591930833e-06, "loss": 0.4558, "step": 1222 }, { "epoch": 0.9935012185215272, "grad_norm": 0.40707623958587646, "learning_rate": 8.460386048388905e-06, "loss": 0.464, "step": 1223 }, { "epoch": 0.9943135662063363, "grad_norm": 0.46881210803985596, "learning_rate": 8.456972411963323e-06, "loss": 0.4841, "step": 1224 }, { "epoch": 0.9951259138911455, "grad_norm": 0.4050699770450592, "learning_rate": 8.453555685705192e-06, "loss": 0.4641, "step": 1225 }, { "epoch": 0.9959382615759546, "grad_norm": 0.4174027144908905, "learning_rate": 8.450135872668369e-06, "loss": 0.4453, "step": 1226 }, { "epoch": 0.9967506092607636, "grad_norm": 0.4392732083797455, "learning_rate": 8.446712975909474e-06, "loss": 0.4328, "step": 1227 }, { "epoch": 0.9975629569455727, "grad_norm": 0.4443771541118622, "learning_rate": 8.443286998487884e-06, "loss": 0.458, "step": 1228 }, { "epoch": 0.9983753046303818, "grad_norm": 0.41146790981292725, "learning_rate": 8.439857943465728e-06, "loss": 0.4828, "step": 1229 }, { "epoch": 0.9991876523151909, "grad_norm": 0.4455551207065582, "learning_rate": 8.436425813907885e-06, "loss": 0.4453, "step": 1230 }, { "epoch": 1.0, "grad_norm": 0.3927052319049835, "learning_rate": 8.432990612881985e-06, "loss": 0.474, "step": 1231 }, { "epoch": 1.0008123476848092, "grad_norm": 0.4331316649913788, "learning_rate": 8.4295523434584e-06, "loss": 0.4454, "step": 1232 }, { "epoch": 1.0016246953696182, "grad_norm": 0.41738376021385193, "learning_rate": 8.426111008710245e-06, "loss": 0.4314, "step": 1233 }, { "epoch": 1.0024370430544274, "grad_norm": 0.4252338111400604, "learning_rate": 8.422666611713378e-06, "loss": 0.4367, "step": 1234 }, { "epoch": 1.0032493907392364, "grad_norm": 0.4933944344520569, "learning_rate": 8.41921915554639e-06, "loss": 0.4711, "step": 1235 }, { "epoch": 1.0040617384240456, "grad_norm": 0.4027901887893677, "learning_rate": 8.415768643290608e-06, "loss": 0.4452, "step": 1236 }, { "epoch": 1.0048740861088545, "grad_norm": 0.45287829637527466, "learning_rate": 8.412315078030091e-06, "loss": 0.4539, "step": 1237 }, { "epoch": 1.0056864337936637, "grad_norm": 0.43848106265068054, "learning_rate": 8.408858462851625e-06, "loss": 0.4274, "step": 1238 }, { "epoch": 1.0064987814784727, "grad_norm": 0.42252078652381897, "learning_rate": 8.405398800844726e-06, "loss": 0.4333, "step": 1239 }, { "epoch": 1.007311129163282, "grad_norm": 0.4724125862121582, "learning_rate": 8.401936095101629e-06, "loss": 0.4193, "step": 1240 }, { "epoch": 1.008123476848091, "grad_norm": 0.4819019138813019, "learning_rate": 8.398470348717288e-06, "loss": 0.4229, "step": 1241 }, { "epoch": 1.0089358245329, "grad_norm": 0.5183465480804443, "learning_rate": 8.395001564789385e-06, "loss": 0.4688, "step": 1242 }, { "epoch": 1.009748172217709, "grad_norm": 0.4451766312122345, "learning_rate": 8.391529746418303e-06, "loss": 0.4484, "step": 1243 }, { "epoch": 1.0105605199025183, "grad_norm": 0.5225660800933838, "learning_rate": 8.388054896707146e-06, "loss": 0.4244, "step": 1244 }, { "epoch": 1.0113728675873275, "grad_norm": 0.47697004675865173, "learning_rate": 8.384577018761725e-06, "loss": 0.4325, "step": 1245 }, { "epoch": 1.0121852152721365, "grad_norm": 0.46439316868782043, "learning_rate": 8.38109611569056e-06, "loss": 0.4323, "step": 1246 }, { "epoch": 1.0129975629569457, "grad_norm": 0.49032357335090637, "learning_rate": 8.377612190604868e-06, "loss": 0.4581, "step": 1247 }, { "epoch": 1.0138099106417546, "grad_norm": 0.49629032611846924, "learning_rate": 8.374125246618576e-06, "loss": 0.4315, "step": 1248 }, { "epoch": 1.0146222583265638, "grad_norm": 0.46360233426094055, "learning_rate": 8.370635286848304e-06, "loss": 0.4383, "step": 1249 }, { "epoch": 1.0154346060113728, "grad_norm": 0.4170956015586853, "learning_rate": 8.367142314413368e-06, "loss": 0.4307, "step": 1250 }, { "epoch": 1.016246953696182, "grad_norm": 0.49460771679878235, "learning_rate": 8.363646332435778e-06, "loss": 0.4523, "step": 1251 }, { "epoch": 1.017059301380991, "grad_norm": 0.4730204641819, "learning_rate": 8.360147344040233e-06, "loss": 0.46, "step": 1252 }, { "epoch": 1.0178716490658002, "grad_norm": 0.5628587603569031, "learning_rate": 8.356645352354117e-06, "loss": 0.4305, "step": 1253 }, { "epoch": 1.0186839967506092, "grad_norm": 0.47420433163642883, "learning_rate": 8.353140360507506e-06, "loss": 0.4542, "step": 1254 }, { "epoch": 1.0194963444354184, "grad_norm": 0.4530784785747528, "learning_rate": 8.349632371633146e-06, "loss": 0.4572, "step": 1255 }, { "epoch": 1.0203086921202273, "grad_norm": 0.556462287902832, "learning_rate": 8.346121388866472e-06, "loss": 0.451, "step": 1256 }, { "epoch": 1.0211210398050365, "grad_norm": 0.49254006147384644, "learning_rate": 8.34260741534559e-06, "loss": 0.4212, "step": 1257 }, { "epoch": 1.0219333874898457, "grad_norm": 0.40716150403022766, "learning_rate": 8.339090454211278e-06, "loss": 0.423, "step": 1258 }, { "epoch": 1.0227457351746547, "grad_norm": 0.45480698347091675, "learning_rate": 8.33557050860699e-06, "loss": 0.4316, "step": 1259 }, { "epoch": 1.023558082859464, "grad_norm": 0.4937492907047272, "learning_rate": 8.332047581678839e-06, "loss": 0.4225, "step": 1260 }, { "epoch": 1.024370430544273, "grad_norm": 0.42261382937431335, "learning_rate": 8.328521676575612e-06, "loss": 0.4261, "step": 1261 }, { "epoch": 1.025182778229082, "grad_norm": 0.44886115193367004, "learning_rate": 8.324992796448752e-06, "loss": 0.4495, "step": 1262 }, { "epoch": 1.025995125913891, "grad_norm": 0.4230818748474121, "learning_rate": 8.321460944452364e-06, "loss": 0.4165, "step": 1263 }, { "epoch": 1.0268074735987003, "grad_norm": 0.41114264726638794, "learning_rate": 8.317926123743202e-06, "loss": 0.4386, "step": 1264 }, { "epoch": 1.0276198212835093, "grad_norm": 0.5249001979827881, "learning_rate": 8.314388337480686e-06, "loss": 0.4401, "step": 1265 }, { "epoch": 1.0284321689683185, "grad_norm": 0.5357246994972229, "learning_rate": 8.310847588826876e-06, "loss": 0.4491, "step": 1266 }, { "epoch": 1.0292445166531274, "grad_norm": 0.40458211302757263, "learning_rate": 8.307303880946485e-06, "loss": 0.4451, "step": 1267 }, { "epoch": 1.0300568643379366, "grad_norm": 0.47108572721481323, "learning_rate": 8.303757217006869e-06, "loss": 0.4226, "step": 1268 }, { "epoch": 1.0308692120227458, "grad_norm": 0.4525722861289978, "learning_rate": 8.300207600178024e-06, "loss": 0.434, "step": 1269 }, { "epoch": 1.0316815597075548, "grad_norm": 0.46809545159339905, "learning_rate": 8.296655033632592e-06, "loss": 0.4351, "step": 1270 }, { "epoch": 1.032493907392364, "grad_norm": 0.46970120072364807, "learning_rate": 8.293099520545845e-06, "loss": 0.4398, "step": 1271 }, { "epoch": 1.033306255077173, "grad_norm": 0.39223918318748474, "learning_rate": 8.289541064095693e-06, "loss": 0.4269, "step": 1272 }, { "epoch": 1.0341186027619822, "grad_norm": 0.4608410894870758, "learning_rate": 8.285979667462671e-06, "loss": 0.4359, "step": 1273 }, { "epoch": 1.0349309504467912, "grad_norm": 0.5276951789855957, "learning_rate": 8.28241533382995e-06, "loss": 0.4414, "step": 1274 }, { "epoch": 1.0357432981316004, "grad_norm": 0.37686455249786377, "learning_rate": 8.27884806638332e-06, "loss": 0.4182, "step": 1275 }, { "epoch": 1.0365556458164094, "grad_norm": 0.42418718338012695, "learning_rate": 8.275277868311193e-06, "loss": 0.433, "step": 1276 }, { "epoch": 1.0373679935012186, "grad_norm": 0.5256267189979553, "learning_rate": 8.271704742804605e-06, "loss": 0.4659, "step": 1277 }, { "epoch": 1.0381803411860275, "grad_norm": 0.4283134639263153, "learning_rate": 8.268128693057202e-06, "loss": 0.4352, "step": 1278 }, { "epoch": 1.0389926888708367, "grad_norm": 0.4393021762371063, "learning_rate": 8.264549722265254e-06, "loss": 0.4092, "step": 1279 }, { "epoch": 1.0398050365556457, "grad_norm": 0.46204668283462524, "learning_rate": 8.260967833627628e-06, "loss": 0.4216, "step": 1280 }, { "epoch": 1.040617384240455, "grad_norm": 0.3930058479309082, "learning_rate": 8.257383030345814e-06, "loss": 0.4469, "step": 1281 }, { "epoch": 1.0414297319252641, "grad_norm": 0.419111043214798, "learning_rate": 8.253795315623894e-06, "loss": 0.4417, "step": 1282 }, { "epoch": 1.042242079610073, "grad_norm": 0.4559578001499176, "learning_rate": 8.250204692668559e-06, "loss": 0.4332, "step": 1283 }, { "epoch": 1.0430544272948823, "grad_norm": 0.4236183762550354, "learning_rate": 8.246611164689097e-06, "loss": 0.4576, "step": 1284 }, { "epoch": 1.0438667749796913, "grad_norm": 0.3997756838798523, "learning_rate": 8.243014734897397e-06, "loss": 0.448, "step": 1285 }, { "epoch": 1.0446791226645005, "grad_norm": 0.44925230741500854, "learning_rate": 8.239415406507934e-06, "loss": 0.454, "step": 1286 }, { "epoch": 1.0454914703493094, "grad_norm": 0.4484761953353882, "learning_rate": 8.235813182737782e-06, "loss": 0.4205, "step": 1287 }, { "epoch": 1.0463038180341186, "grad_norm": 0.43182140588760376, "learning_rate": 8.232208066806596e-06, "loss": 0.4248, "step": 1288 }, { "epoch": 1.0471161657189276, "grad_norm": 0.4116652309894562, "learning_rate": 8.228600061936617e-06, "loss": 0.4085, "step": 1289 }, { "epoch": 1.0479285134037368, "grad_norm": 0.5102201700210571, "learning_rate": 8.224989171352675e-06, "loss": 0.4463, "step": 1290 }, { "epoch": 1.0487408610885458, "grad_norm": 0.4537793695926666, "learning_rate": 8.221375398282172e-06, "loss": 0.4561, "step": 1291 }, { "epoch": 1.049553208773355, "grad_norm": 0.4723230004310608, "learning_rate": 8.217758745955087e-06, "loss": 0.4576, "step": 1292 }, { "epoch": 1.0503655564581642, "grad_norm": 0.4999629855155945, "learning_rate": 8.214139217603975e-06, "loss": 0.4474, "step": 1293 }, { "epoch": 1.0511779041429732, "grad_norm": 0.5164727568626404, "learning_rate": 8.210516816463961e-06, "loss": 0.4397, "step": 1294 }, { "epoch": 1.0519902518277824, "grad_norm": 0.5361226797103882, "learning_rate": 8.206891545772738e-06, "loss": 0.4354, "step": 1295 }, { "epoch": 1.0528025995125914, "grad_norm": 0.5655002593994141, "learning_rate": 8.203263408770562e-06, "loss": 0.4347, "step": 1296 }, { "epoch": 1.0536149471974006, "grad_norm": 0.5547245740890503, "learning_rate": 8.19963240870025e-06, "loss": 0.462, "step": 1297 }, { "epoch": 1.0544272948822095, "grad_norm": 0.5319647192955017, "learning_rate": 8.195998548807185e-06, "loss": 0.442, "step": 1298 }, { "epoch": 1.0552396425670187, "grad_norm": 0.5249698758125305, "learning_rate": 8.192361832339299e-06, "loss": 0.4253, "step": 1299 }, { "epoch": 1.0560519902518277, "grad_norm": 0.4743811786174774, "learning_rate": 8.188722262547078e-06, "loss": 0.427, "step": 1300 }, { "epoch": 1.056864337936637, "grad_norm": 0.42148494720458984, "learning_rate": 8.185079842683558e-06, "loss": 0.4574, "step": 1301 }, { "epoch": 1.057676685621446, "grad_norm": 0.4681132137775421, "learning_rate": 8.181434576004328e-06, "loss": 0.4358, "step": 1302 }, { "epoch": 1.058489033306255, "grad_norm": 0.4373636841773987, "learning_rate": 8.177786465767517e-06, "loss": 0.412, "step": 1303 }, { "epoch": 1.059301380991064, "grad_norm": 0.36684393882751465, "learning_rate": 8.174135515233792e-06, "loss": 0.4344, "step": 1304 }, { "epoch": 1.0601137286758733, "grad_norm": 0.44498559832572937, "learning_rate": 8.170481727666366e-06, "loss": 0.4261, "step": 1305 }, { "epoch": 1.0609260763606825, "grad_norm": 0.40805113315582275, "learning_rate": 8.166825106330985e-06, "loss": 0.4419, "step": 1306 }, { "epoch": 1.0617384240454915, "grad_norm": 0.4648733139038086, "learning_rate": 8.163165654495925e-06, "loss": 0.4337, "step": 1307 }, { "epoch": 1.0625507717303007, "grad_norm": 0.3823528587818146, "learning_rate": 8.159503375431996e-06, "loss": 0.4264, "step": 1308 }, { "epoch": 1.0633631194151096, "grad_norm": 0.47934064269065857, "learning_rate": 8.155838272412531e-06, "loss": 0.4204, "step": 1309 }, { "epoch": 1.0641754670999188, "grad_norm": 0.49378839135169983, "learning_rate": 8.152170348713392e-06, "loss": 0.4404, "step": 1310 }, { "epoch": 1.0649878147847278, "grad_norm": 0.4586998224258423, "learning_rate": 8.148499607612956e-06, "loss": 0.4162, "step": 1311 }, { "epoch": 1.065800162469537, "grad_norm": 0.42415353655815125, "learning_rate": 8.144826052392123e-06, "loss": 0.4384, "step": 1312 }, { "epoch": 1.066612510154346, "grad_norm": 0.4287481904029846, "learning_rate": 8.141149686334307e-06, "loss": 0.4301, "step": 1313 }, { "epoch": 1.0674248578391552, "grad_norm": 0.618630588054657, "learning_rate": 8.137470512725434e-06, "loss": 0.4297, "step": 1314 }, { "epoch": 1.0682372055239642, "grad_norm": 0.4223957359790802, "learning_rate": 8.133788534853939e-06, "loss": 0.4334, "step": 1315 }, { "epoch": 1.0690495532087734, "grad_norm": 0.5580872893333435, "learning_rate": 8.130103756010763e-06, "loss": 0.4356, "step": 1316 }, { "epoch": 1.0698619008935824, "grad_norm": 0.46561360359191895, "learning_rate": 8.126416179489354e-06, "loss": 0.4379, "step": 1317 }, { "epoch": 1.0706742485783916, "grad_norm": 0.4603703022003174, "learning_rate": 8.122725808585654e-06, "loss": 0.4454, "step": 1318 }, { "epoch": 1.0714865962632008, "grad_norm": 0.5122115612030029, "learning_rate": 8.11903264659811e-06, "loss": 0.4618, "step": 1319 }, { "epoch": 1.0722989439480097, "grad_norm": 0.5117606520652771, "learning_rate": 8.115336696827658e-06, "loss": 0.4555, "step": 1320 }, { "epoch": 1.073111291632819, "grad_norm": 0.4663766920566559, "learning_rate": 8.111637962577728e-06, "loss": 0.4336, "step": 1321 }, { "epoch": 1.073923639317628, "grad_norm": 0.43328922986984253, "learning_rate": 8.10793644715424e-06, "loss": 0.4475, "step": 1322 }, { "epoch": 1.074735987002437, "grad_norm": 0.5659634470939636, "learning_rate": 8.104232153865597e-06, "loss": 0.4276, "step": 1323 }, { "epoch": 1.075548334687246, "grad_norm": 0.47168147563934326, "learning_rate": 8.100525086022687e-06, "loss": 0.4584, "step": 1324 }, { "epoch": 1.0763606823720553, "grad_norm": 0.4281969368457794, "learning_rate": 8.096815246938879e-06, "loss": 0.4434, "step": 1325 }, { "epoch": 1.0771730300568643, "grad_norm": 0.4717988073825836, "learning_rate": 8.093102639930013e-06, "loss": 0.4564, "step": 1326 }, { "epoch": 1.0779853777416735, "grad_norm": 0.46498191356658936, "learning_rate": 8.089387268314408e-06, "loss": 0.4577, "step": 1327 }, { "epoch": 1.0787977254264824, "grad_norm": 0.37887489795684814, "learning_rate": 8.085669135412856e-06, "loss": 0.4146, "step": 1328 }, { "epoch": 1.0796100731112916, "grad_norm": 0.48285171389579773, "learning_rate": 8.081948244548615e-06, "loss": 0.4245, "step": 1329 }, { "epoch": 1.0804224207961006, "grad_norm": 0.43382883071899414, "learning_rate": 8.0782245990474e-06, "loss": 0.4252, "step": 1330 }, { "epoch": 1.0812347684809098, "grad_norm": 0.4526747167110443, "learning_rate": 8.074498202237402e-06, "loss": 0.449, "step": 1331 }, { "epoch": 1.082047116165719, "grad_norm": 0.4427867531776428, "learning_rate": 8.070769057449262e-06, "loss": 0.4224, "step": 1332 }, { "epoch": 1.082859463850528, "grad_norm": 0.406307190656662, "learning_rate": 8.067037168016079e-06, "loss": 0.4241, "step": 1333 }, { "epoch": 1.0836718115353372, "grad_norm": 0.48406216502189636, "learning_rate": 8.063302537273407e-06, "loss": 0.4306, "step": 1334 }, { "epoch": 1.0844841592201462, "grad_norm": 0.4375550150871277, "learning_rate": 8.059565168559247e-06, "loss": 0.4717, "step": 1335 }, { "epoch": 1.0852965069049554, "grad_norm": 0.412991464138031, "learning_rate": 8.05582506521405e-06, "loss": 0.4266, "step": 1336 }, { "epoch": 1.0861088545897644, "grad_norm": 0.44577375054359436, "learning_rate": 8.052082230580711e-06, "loss": 0.4206, "step": 1337 }, { "epoch": 1.0869212022745736, "grad_norm": 0.43016287684440613, "learning_rate": 8.048336668004564e-06, "loss": 0.4426, "step": 1338 }, { "epoch": 1.0877335499593825, "grad_norm": 0.4562046527862549, "learning_rate": 8.044588380833384e-06, "loss": 0.4375, "step": 1339 }, { "epoch": 1.0885458976441917, "grad_norm": 0.39804595708847046, "learning_rate": 8.040837372417379e-06, "loss": 0.4272, "step": 1340 }, { "epoch": 1.0893582453290007, "grad_norm": 0.46848437190055847, "learning_rate": 8.03708364610919e-06, "loss": 0.4475, "step": 1341 }, { "epoch": 1.09017059301381, "grad_norm": 0.42733949422836304, "learning_rate": 8.033327205263888e-06, "loss": 0.4279, "step": 1342 }, { "epoch": 1.090982940698619, "grad_norm": 0.4692021310329437, "learning_rate": 8.02956805323897e-06, "loss": 0.4282, "step": 1343 }, { "epoch": 1.091795288383428, "grad_norm": 0.5373814105987549, "learning_rate": 8.025806193394354e-06, "loss": 0.4487, "step": 1344 }, { "epoch": 1.0926076360682373, "grad_norm": 0.4217304587364197, "learning_rate": 8.022041629092382e-06, "loss": 0.4169, "step": 1345 }, { "epoch": 1.0934199837530463, "grad_norm": 0.4654847979545593, "learning_rate": 8.01827436369781e-06, "loss": 0.4355, "step": 1346 }, { "epoch": 1.0942323314378555, "grad_norm": 0.5154445767402649, "learning_rate": 8.01450440057781e-06, "loss": 0.4459, "step": 1347 }, { "epoch": 1.0950446791226645, "grad_norm": 0.4609627425670624, "learning_rate": 8.010731743101967e-06, "loss": 0.425, "step": 1348 }, { "epoch": 1.0958570268074737, "grad_norm": 0.47743454575538635, "learning_rate": 8.006956394642268e-06, "loss": 0.4342, "step": 1349 }, { "epoch": 1.0966693744922826, "grad_norm": 0.514366090297699, "learning_rate": 8.003178358573112e-06, "loss": 0.4641, "step": 1350 }, { "epoch": 1.0974817221770918, "grad_norm": 0.4594586193561554, "learning_rate": 7.999397638271298e-06, "loss": 0.4387, "step": 1351 }, { "epoch": 1.0982940698619008, "grad_norm": 0.44047287106513977, "learning_rate": 7.99561423711602e-06, "loss": 0.4469, "step": 1352 }, { "epoch": 1.09910641754671, "grad_norm": 0.4455038607120514, "learning_rate": 7.991828158488875e-06, "loss": 0.4237, "step": 1353 }, { "epoch": 1.099918765231519, "grad_norm": 0.49357616901397705, "learning_rate": 7.98803940577385e-06, "loss": 0.4839, "step": 1354 }, { "epoch": 1.1007311129163282, "grad_norm": 0.40508750081062317, "learning_rate": 7.984247982357317e-06, "loss": 0.4362, "step": 1355 }, { "epoch": 1.1015434606011372, "grad_norm": 0.4900878965854645, "learning_rate": 7.980453891628047e-06, "loss": 0.4419, "step": 1356 }, { "epoch": 1.1023558082859464, "grad_norm": 0.5241803526878357, "learning_rate": 7.976657136977181e-06, "loss": 0.4253, "step": 1357 }, { "epoch": 1.1031681559707556, "grad_norm": 0.49766746163368225, "learning_rate": 7.972857721798254e-06, "loss": 0.4461, "step": 1358 }, { "epoch": 1.1039805036555645, "grad_norm": 0.4764297902584076, "learning_rate": 7.96905564948717e-06, "loss": 0.4606, "step": 1359 }, { "epoch": 1.1047928513403737, "grad_norm": 0.5020046830177307, "learning_rate": 7.96525092344221e-06, "loss": 0.4584, "step": 1360 }, { "epoch": 1.1056051990251827, "grad_norm": 0.3887828290462494, "learning_rate": 7.96144354706403e-06, "loss": 0.4112, "step": 1361 }, { "epoch": 1.106417546709992, "grad_norm": 0.4879290759563446, "learning_rate": 7.957633523755652e-06, "loss": 0.4433, "step": 1362 }, { "epoch": 1.107229894394801, "grad_norm": 0.4898350238800049, "learning_rate": 7.953820856922465e-06, "loss": 0.4268, "step": 1363 }, { "epoch": 1.10804224207961, "grad_norm": 0.46234890818595886, "learning_rate": 7.95000554997222e-06, "loss": 0.4477, "step": 1364 }, { "epoch": 1.108854589764419, "grad_norm": 0.5446366667747498, "learning_rate": 7.946187606315026e-06, "loss": 0.453, "step": 1365 }, { "epoch": 1.1096669374492283, "grad_norm": 0.5074025988578796, "learning_rate": 7.942367029363351e-06, "loss": 0.4397, "step": 1366 }, { "epoch": 1.1104792851340373, "grad_norm": 0.5596259236335754, "learning_rate": 7.938543822532019e-06, "loss": 0.4431, "step": 1367 }, { "epoch": 1.1112916328188465, "grad_norm": 0.5185883045196533, "learning_rate": 7.9347179892382e-06, "loss": 0.4314, "step": 1368 }, { "epoch": 1.1121039805036557, "grad_norm": 0.42015376687049866, "learning_rate": 7.930889532901416e-06, "loss": 0.4372, "step": 1369 }, { "epoch": 1.1129163281884646, "grad_norm": 0.46527034044265747, "learning_rate": 7.927058456943527e-06, "loss": 0.4443, "step": 1370 }, { "epoch": 1.1137286758732738, "grad_norm": 0.4787082076072693, "learning_rate": 7.92322476478874e-06, "loss": 0.4341, "step": 1371 }, { "epoch": 1.1145410235580828, "grad_norm": 0.5275425910949707, "learning_rate": 7.919388459863597e-06, "loss": 0.4341, "step": 1372 }, { "epoch": 1.115353371242892, "grad_norm": 0.4145677089691162, "learning_rate": 7.915549545596979e-06, "loss": 0.4327, "step": 1373 }, { "epoch": 1.116165718927701, "grad_norm": 0.5067287683486938, "learning_rate": 7.911708025420097e-06, "loss": 0.4491, "step": 1374 }, { "epoch": 1.1169780666125102, "grad_norm": 0.5521039962768555, "learning_rate": 7.907863902766494e-06, "loss": 0.4048, "step": 1375 }, { "epoch": 1.1177904142973192, "grad_norm": 0.45138511061668396, "learning_rate": 7.90401718107203e-06, "loss": 0.4559, "step": 1376 }, { "epoch": 1.1186027619821284, "grad_norm": 0.6131345629692078, "learning_rate": 7.900167863774896e-06, "loss": 0.4194, "step": 1377 }, { "epoch": 1.1194151096669374, "grad_norm": 0.522714376449585, "learning_rate": 7.896315954315608e-06, "loss": 0.4563, "step": 1378 }, { "epoch": 1.1202274573517466, "grad_norm": 0.48783406615257263, "learning_rate": 7.892461456136984e-06, "loss": 0.432, "step": 1379 }, { "epoch": 1.1210398050365558, "grad_norm": 0.5166853070259094, "learning_rate": 7.888604372684168e-06, "loss": 0.4593, "step": 1380 }, { "epoch": 1.1218521527213647, "grad_norm": 0.4254966080188751, "learning_rate": 7.88474470740461e-06, "loss": 0.4375, "step": 1381 }, { "epoch": 1.122664500406174, "grad_norm": 0.5695308446884155, "learning_rate": 7.880882463748067e-06, "loss": 0.4264, "step": 1382 }, { "epoch": 1.123476848090983, "grad_norm": 0.439438134431839, "learning_rate": 7.877017645166606e-06, "loss": 0.4337, "step": 1383 }, { "epoch": 1.1242891957757921, "grad_norm": 0.5017744302749634, "learning_rate": 7.873150255114588e-06, "loss": 0.4433, "step": 1384 }, { "epoch": 1.125101543460601, "grad_norm": 0.5389549732208252, "learning_rate": 7.869280297048674e-06, "loss": 0.4479, "step": 1385 }, { "epoch": 1.1259138911454103, "grad_norm": 0.4771658480167389, "learning_rate": 7.865407774427828e-06, "loss": 0.42, "step": 1386 }, { "epoch": 1.1267262388302193, "grad_norm": 0.5328062176704407, "learning_rate": 7.861532690713293e-06, "loss": 0.4251, "step": 1387 }, { "epoch": 1.1275385865150285, "grad_norm": 0.5170143842697144, "learning_rate": 7.857655049368614e-06, "loss": 0.428, "step": 1388 }, { "epoch": 1.1283509341998375, "grad_norm": 0.44764816761016846, "learning_rate": 7.853774853859612e-06, "loss": 0.4287, "step": 1389 }, { "epoch": 1.1291632818846467, "grad_norm": 0.4964679479598999, "learning_rate": 7.8498921076544e-06, "loss": 0.4361, "step": 1390 }, { "epoch": 1.1299756295694556, "grad_norm": 0.4497617781162262, "learning_rate": 7.846006814223362e-06, "loss": 0.4288, "step": 1391 }, { "epoch": 1.1307879772542648, "grad_norm": 0.42560264468193054, "learning_rate": 7.842118977039162e-06, "loss": 0.4438, "step": 1392 }, { "epoch": 1.131600324939074, "grad_norm": 0.4930296242237091, "learning_rate": 7.838228599576743e-06, "loss": 0.4181, "step": 1393 }, { "epoch": 1.132412672623883, "grad_norm": 0.4141998589038849, "learning_rate": 7.83433568531331e-06, "loss": 0.451, "step": 1394 }, { "epoch": 1.1332250203086922, "grad_norm": 0.38222232460975647, "learning_rate": 7.830440237728341e-06, "loss": 0.4361, "step": 1395 }, { "epoch": 1.1340373679935012, "grad_norm": 0.46584418416023254, "learning_rate": 7.826542260303576e-06, "loss": 0.4477, "step": 1396 }, { "epoch": 1.1348497156783104, "grad_norm": 0.4459529221057892, "learning_rate": 7.822641756523015e-06, "loss": 0.4547, "step": 1397 }, { "epoch": 1.1356620633631194, "grad_norm": 0.40063831210136414, "learning_rate": 7.818738729872921e-06, "loss": 0.434, "step": 1398 }, { "epoch": 1.1364744110479286, "grad_norm": 0.44158050417900085, "learning_rate": 7.814833183841805e-06, "loss": 0.4426, "step": 1399 }, { "epoch": 1.1372867587327375, "grad_norm": 0.40509259700775146, "learning_rate": 7.810925121920437e-06, "loss": 0.4464, "step": 1400 }, { "epoch": 1.1380991064175467, "grad_norm": 0.3798421323299408, "learning_rate": 7.80701454760183e-06, "loss": 0.4406, "step": 1401 }, { "epoch": 1.1389114541023557, "grad_norm": 0.4238473176956177, "learning_rate": 7.803101464381244e-06, "loss": 0.468, "step": 1402 }, { "epoch": 1.139723801787165, "grad_norm": 0.3825916051864624, "learning_rate": 7.799185875756184e-06, "loss": 0.4137, "step": 1403 }, { "epoch": 1.140536149471974, "grad_norm": 0.41741517186164856, "learning_rate": 7.795267785226392e-06, "loss": 0.4505, "step": 1404 }, { "epoch": 1.141348497156783, "grad_norm": 0.40400540828704834, "learning_rate": 7.791347196293846e-06, "loss": 0.4186, "step": 1405 }, { "epoch": 1.1421608448415923, "grad_norm": 0.43444114923477173, "learning_rate": 7.787424112462758e-06, "loss": 0.4226, "step": 1406 }, { "epoch": 1.1429731925264013, "grad_norm": 0.46810561418533325, "learning_rate": 7.783498537239568e-06, "loss": 0.4745, "step": 1407 }, { "epoch": 1.1437855402112105, "grad_norm": 0.46253445744514465, "learning_rate": 7.779570474132949e-06, "loss": 0.4334, "step": 1408 }, { "epoch": 1.1445978878960195, "grad_norm": 0.4158393144607544, "learning_rate": 7.775639926653789e-06, "loss": 0.4249, "step": 1409 }, { "epoch": 1.1454102355808287, "grad_norm": 0.4182446300983429, "learning_rate": 7.771706898315204e-06, "loss": 0.453, "step": 1410 }, { "epoch": 1.1462225832656376, "grad_norm": 0.43574872612953186, "learning_rate": 7.767771392632518e-06, "loss": 0.4216, "step": 1411 }, { "epoch": 1.1470349309504468, "grad_norm": 0.3994395136833191, "learning_rate": 7.763833413123281e-06, "loss": 0.4476, "step": 1412 }, { "epoch": 1.1478472786352558, "grad_norm": 0.4407590329647064, "learning_rate": 7.759892963307249e-06, "loss": 0.4582, "step": 1413 }, { "epoch": 1.148659626320065, "grad_norm": 0.3946763575077057, "learning_rate": 7.75595004670638e-06, "loss": 0.447, "step": 1414 }, { "epoch": 1.149471974004874, "grad_norm": 0.5031050443649292, "learning_rate": 7.752004666844849e-06, "loss": 0.4244, "step": 1415 }, { "epoch": 1.1502843216896832, "grad_norm": 0.3991170823574066, "learning_rate": 7.748056827249018e-06, "loss": 0.4286, "step": 1416 }, { "epoch": 1.1510966693744922, "grad_norm": 0.41863375902175903, "learning_rate": 7.744106531447462e-06, "loss": 0.4198, "step": 1417 }, { "epoch": 1.1519090170593014, "grad_norm": 0.47870102524757385, "learning_rate": 7.74015378297094e-06, "loss": 0.4425, "step": 1418 }, { "epoch": 1.1527213647441106, "grad_norm": 0.4682295024394989, "learning_rate": 7.736198585352411e-06, "loss": 0.4602, "step": 1419 }, { "epoch": 1.1535337124289196, "grad_norm": 0.4021531939506531, "learning_rate": 7.732240942127018e-06, "loss": 0.4374, "step": 1420 }, { "epoch": 1.1543460601137288, "grad_norm": 0.39077937602996826, "learning_rate": 7.728280856832094e-06, "loss": 0.433, "step": 1421 }, { "epoch": 1.1551584077985377, "grad_norm": 0.4612072706222534, "learning_rate": 7.72431833300715e-06, "loss": 0.4391, "step": 1422 }, { "epoch": 1.155970755483347, "grad_norm": 0.4032115936279297, "learning_rate": 7.72035337419388e-06, "loss": 0.4446, "step": 1423 }, { "epoch": 1.156783103168156, "grad_norm": 0.3930440843105316, "learning_rate": 7.716385983936154e-06, "loss": 0.4005, "step": 1424 }, { "epoch": 1.1575954508529651, "grad_norm": 0.48419514298439026, "learning_rate": 7.712416165780014e-06, "loss": 0.4509, "step": 1425 }, { "epoch": 1.158407798537774, "grad_norm": 0.4345088005065918, "learning_rate": 7.708443923273671e-06, "loss": 0.4587, "step": 1426 }, { "epoch": 1.1592201462225833, "grad_norm": 0.3952743113040924, "learning_rate": 7.704469259967509e-06, "loss": 0.4472, "step": 1427 }, { "epoch": 1.1600324939073923, "grad_norm": 0.4636023938655853, "learning_rate": 7.700492179414067e-06, "loss": 0.4318, "step": 1428 }, { "epoch": 1.1608448415922015, "grad_norm": 0.3977108895778656, "learning_rate": 7.69651268516805e-06, "loss": 0.4394, "step": 1429 }, { "epoch": 1.1616571892770104, "grad_norm": 0.393501341342926, "learning_rate": 7.692530780786321e-06, "loss": 0.4206, "step": 1430 }, { "epoch": 1.1624695369618196, "grad_norm": 0.46492549777030945, "learning_rate": 7.688546469827892e-06, "loss": 0.4283, "step": 1431 }, { "epoch": 1.1632818846466288, "grad_norm": 0.3878355622291565, "learning_rate": 7.684559755853932e-06, "loss": 0.4142, "step": 1432 }, { "epoch": 1.1640942323314378, "grad_norm": 0.4160679280757904, "learning_rate": 7.680570642427755e-06, "loss": 0.4203, "step": 1433 }, { "epoch": 1.164906580016247, "grad_norm": 0.4211764335632324, "learning_rate": 7.676579133114819e-06, "loss": 0.4576, "step": 1434 }, { "epoch": 1.165718927701056, "grad_norm": 0.3893836438655853, "learning_rate": 7.672585231482723e-06, "loss": 0.4411, "step": 1435 }, { "epoch": 1.1665312753858652, "grad_norm": 0.42600154876708984, "learning_rate": 7.668588941101208e-06, "loss": 0.456, "step": 1436 }, { "epoch": 1.1673436230706742, "grad_norm": 0.41839599609375, "learning_rate": 7.664590265542144e-06, "loss": 0.4428, "step": 1437 }, { "epoch": 1.1681559707554834, "grad_norm": 0.42473986744880676, "learning_rate": 7.66058920837954e-06, "loss": 0.4475, "step": 1438 }, { "epoch": 1.1689683184402924, "grad_norm": 0.370962530374527, "learning_rate": 7.65658577318953e-06, "loss": 0.4259, "step": 1439 }, { "epoch": 1.1697806661251016, "grad_norm": 0.4366114139556885, "learning_rate": 7.65257996355037e-06, "loss": 0.4177, "step": 1440 }, { "epoch": 1.1705930138099108, "grad_norm": 0.40339797735214233, "learning_rate": 7.648571783042445e-06, "loss": 0.4145, "step": 1441 }, { "epoch": 1.1714053614947197, "grad_norm": 0.37992992997169495, "learning_rate": 7.644561235248254e-06, "loss": 0.4282, "step": 1442 }, { "epoch": 1.1722177091795287, "grad_norm": 0.44465282559394836, "learning_rate": 7.640548323752412e-06, "loss": 0.4258, "step": 1443 }, { "epoch": 1.173030056864338, "grad_norm": 0.4468477964401245, "learning_rate": 7.636533052141653e-06, "loss": 0.4465, "step": 1444 }, { "epoch": 1.1738424045491471, "grad_norm": 0.4692933261394501, "learning_rate": 7.632515424004813e-06, "loss": 0.4381, "step": 1445 }, { "epoch": 1.174654752233956, "grad_norm": 0.3957800269126892, "learning_rate": 7.628495442932838e-06, "loss": 0.4279, "step": 1446 }, { "epoch": 1.1754670999187653, "grad_norm": 0.5090640187263489, "learning_rate": 7.624473112518777e-06, "loss": 0.432, "step": 1447 }, { "epoch": 1.1762794476035743, "grad_norm": 0.4562416672706604, "learning_rate": 7.6204484363577756e-06, "loss": 0.4417, "step": 1448 }, { "epoch": 1.1770917952883835, "grad_norm": 0.43862074613571167, "learning_rate": 7.61642141804708e-06, "loss": 0.4492, "step": 1449 }, { "epoch": 1.1779041429731925, "grad_norm": 0.5032182335853577, "learning_rate": 7.612392061186027e-06, "loss": 0.4297, "step": 1450 }, { "epoch": 1.1787164906580017, "grad_norm": 0.4196731746196747, "learning_rate": 7.608360369376047e-06, "loss": 0.4336, "step": 1451 }, { "epoch": 1.1795288383428106, "grad_norm": 0.48235753178596497, "learning_rate": 7.604326346220654e-06, "loss": 0.4341, "step": 1452 }, { "epoch": 1.1803411860276198, "grad_norm": 0.41192200779914856, "learning_rate": 7.600289995325446e-06, "loss": 0.446, "step": 1453 }, { "epoch": 1.181153533712429, "grad_norm": 0.4187053143978119, "learning_rate": 7.596251320298105e-06, "loss": 0.4116, "step": 1454 }, { "epoch": 1.181965881397238, "grad_norm": 0.4754772484302521, "learning_rate": 7.592210324748384e-06, "loss": 0.4519, "step": 1455 }, { "epoch": 1.182778229082047, "grad_norm": 0.3958410620689392, "learning_rate": 7.588167012288117e-06, "loss": 0.4041, "step": 1456 }, { "epoch": 1.1835905767668562, "grad_norm": 0.4172162413597107, "learning_rate": 7.584121386531205e-06, "loss": 0.4352, "step": 1457 }, { "epoch": 1.1844029244516654, "grad_norm": 0.4238002598285675, "learning_rate": 7.580073451093617e-06, "loss": 0.4425, "step": 1458 }, { "epoch": 1.1852152721364744, "grad_norm": 0.4225533902645111, "learning_rate": 7.576023209593386e-06, "loss": 0.4289, "step": 1459 }, { "epoch": 1.1860276198212836, "grad_norm": 0.41040223836898804, "learning_rate": 7.571970665650607e-06, "loss": 0.4543, "step": 1460 }, { "epoch": 1.1868399675060926, "grad_norm": 0.3907702565193176, "learning_rate": 7.567915822887434e-06, "loss": 0.4321, "step": 1461 }, { "epoch": 1.1876523151909018, "grad_norm": 0.39740344882011414, "learning_rate": 7.563858684928074e-06, "loss": 0.4181, "step": 1462 }, { "epoch": 1.1884646628757107, "grad_norm": 0.4074680805206299, "learning_rate": 7.559799255398785e-06, "loss": 0.4387, "step": 1463 }, { "epoch": 1.18927701056052, "grad_norm": 0.3857968747615814, "learning_rate": 7.555737537927875e-06, "loss": 0.4298, "step": 1464 }, { "epoch": 1.190089358245329, "grad_norm": 0.458016037940979, "learning_rate": 7.551673536145695e-06, "loss": 0.4367, "step": 1465 }, { "epoch": 1.190901705930138, "grad_norm": 0.4059580862522125, "learning_rate": 7.54760725368464e-06, "loss": 0.4503, "step": 1466 }, { "epoch": 1.1917140536149473, "grad_norm": 0.43018433451652527, "learning_rate": 7.543538694179142e-06, "loss": 0.419, "step": 1467 }, { "epoch": 1.1925264012997563, "grad_norm": 0.3979254364967346, "learning_rate": 7.539467861265668e-06, "loss": 0.4345, "step": 1468 }, { "epoch": 1.1933387489845655, "grad_norm": 0.43597379326820374, "learning_rate": 7.535394758582717e-06, "loss": 0.4381, "step": 1469 }, { "epoch": 1.1941510966693745, "grad_norm": 0.41497352719306946, "learning_rate": 7.531319389770818e-06, "loss": 0.458, "step": 1470 }, { "epoch": 1.1949634443541837, "grad_norm": 0.3971577286720276, "learning_rate": 7.527241758472525e-06, "loss": 0.4256, "step": 1471 }, { "epoch": 1.1957757920389926, "grad_norm": 0.44603872299194336, "learning_rate": 7.523161868332412e-06, "loss": 0.3994, "step": 1472 }, { "epoch": 1.1965881397238018, "grad_norm": 0.4628686308860779, "learning_rate": 7.519079722997076e-06, "loss": 0.4661, "step": 1473 }, { "epoch": 1.1974004874086108, "grad_norm": 0.4146297872066498, "learning_rate": 7.514995326115124e-06, "loss": 0.4388, "step": 1474 }, { "epoch": 1.19821283509342, "grad_norm": 0.41237735748291016, "learning_rate": 7.5109086813371835e-06, "loss": 0.4541, "step": 1475 }, { "epoch": 1.199025182778229, "grad_norm": 0.47962072491645813, "learning_rate": 7.506819792315883e-06, "loss": 0.4683, "step": 1476 }, { "epoch": 1.1998375304630382, "grad_norm": 0.4494830071926117, "learning_rate": 7.502728662705861e-06, "loss": 0.4259, "step": 1477 }, { "epoch": 1.2006498781478472, "grad_norm": 0.4391888976097107, "learning_rate": 7.498635296163758e-06, "loss": 0.4406, "step": 1478 }, { "epoch": 1.2014622258326564, "grad_norm": 0.507339596748352, "learning_rate": 7.494539696348215e-06, "loss": 0.4351, "step": 1479 }, { "epoch": 1.2022745735174656, "grad_norm": 0.4066133201122284, "learning_rate": 7.490441866919869e-06, "loss": 0.4397, "step": 1480 }, { "epoch": 1.2030869212022746, "grad_norm": 0.4261534810066223, "learning_rate": 7.486341811541345e-06, "loss": 0.4413, "step": 1481 }, { "epoch": 1.2038992688870838, "grad_norm": 0.45985305309295654, "learning_rate": 7.482239533877264e-06, "loss": 0.432, "step": 1482 }, { "epoch": 1.2047116165718927, "grad_norm": 0.45539411902427673, "learning_rate": 7.4781350375942296e-06, "loss": 0.4356, "step": 1483 }, { "epoch": 1.205523964256702, "grad_norm": 0.42351698875427246, "learning_rate": 7.4740283263608305e-06, "loss": 0.4325, "step": 1484 }, { "epoch": 1.206336311941511, "grad_norm": 0.4734072983264923, "learning_rate": 7.469919403847633e-06, "loss": 0.428, "step": 1485 }, { "epoch": 1.2071486596263201, "grad_norm": 0.4656287133693695, "learning_rate": 7.465808273727182e-06, "loss": 0.4448, "step": 1486 }, { "epoch": 1.207961007311129, "grad_norm": 0.36348557472229004, "learning_rate": 7.461694939673992e-06, "loss": 0.421, "step": 1487 }, { "epoch": 1.2087733549959383, "grad_norm": 0.4436640441417694, "learning_rate": 7.457579405364555e-06, "loss": 0.4165, "step": 1488 }, { "epoch": 1.2095857026807473, "grad_norm": 0.4206816554069519, "learning_rate": 7.453461674477318e-06, "loss": 0.4128, "step": 1489 }, { "epoch": 1.2103980503655565, "grad_norm": 0.4062630832195282, "learning_rate": 7.449341750692703e-06, "loss": 0.4489, "step": 1490 }, { "epoch": 1.2112103980503655, "grad_norm": 0.3898524343967438, "learning_rate": 7.445219637693082e-06, "loss": 0.4418, "step": 1491 }, { "epoch": 1.2120227457351747, "grad_norm": 0.4151443839073181, "learning_rate": 7.441095339162791e-06, "loss": 0.4493, "step": 1492 }, { "epoch": 1.2128350934199839, "grad_norm": 0.431013286113739, "learning_rate": 7.436968858788118e-06, "loss": 0.4246, "step": 1493 }, { "epoch": 1.2136474411047928, "grad_norm": 0.4703874886035919, "learning_rate": 7.4328402002572985e-06, "loss": 0.4594, "step": 1494 }, { "epoch": 1.214459788789602, "grad_norm": 0.39780861139297485, "learning_rate": 7.428709367260516e-06, "loss": 0.4357, "step": 1495 }, { "epoch": 1.215272136474411, "grad_norm": 0.47212347388267517, "learning_rate": 7.424576363489897e-06, "loss": 0.48, "step": 1496 }, { "epoch": 1.2160844841592202, "grad_norm": 0.4078375995159149, "learning_rate": 7.420441192639511e-06, "loss": 0.4436, "step": 1497 }, { "epoch": 1.2168968318440292, "grad_norm": 0.5259037613868713, "learning_rate": 7.416303858405363e-06, "loss": 0.4035, "step": 1498 }, { "epoch": 1.2177091795288384, "grad_norm": 0.5122765898704529, "learning_rate": 7.412164364485388e-06, "loss": 0.4526, "step": 1499 }, { "epoch": 1.2185215272136474, "grad_norm": 0.45020177960395813, "learning_rate": 7.408022714579457e-06, "loss": 0.4306, "step": 1500 }, { "epoch": 1.2193338748984566, "grad_norm": 0.5346155166625977, "learning_rate": 7.403878912389365e-06, "loss": 0.4527, "step": 1501 }, { "epoch": 1.2201462225832655, "grad_norm": 0.5493088960647583, "learning_rate": 7.3997329616188305e-06, "loss": 0.4779, "step": 1502 }, { "epoch": 1.2209585702680747, "grad_norm": 0.44448116421699524, "learning_rate": 7.395584865973493e-06, "loss": 0.4475, "step": 1503 }, { "epoch": 1.2217709179528837, "grad_norm": 0.47944480180740356, "learning_rate": 7.391434629160909e-06, "loss": 0.4375, "step": 1504 }, { "epoch": 1.222583265637693, "grad_norm": 0.4372415542602539, "learning_rate": 7.38728225489055e-06, "loss": 0.4376, "step": 1505 }, { "epoch": 1.2233956133225021, "grad_norm": 0.48701199889183044, "learning_rate": 7.383127746873796e-06, "loss": 0.4426, "step": 1506 }, { "epoch": 1.224207961007311, "grad_norm": 0.47440460324287415, "learning_rate": 7.378971108823933e-06, "loss": 0.4316, "step": 1507 }, { "epoch": 1.2250203086921203, "grad_norm": 0.4684794843196869, "learning_rate": 7.374812344456157e-06, "loss": 0.4458, "step": 1508 }, { "epoch": 1.2258326563769293, "grad_norm": 0.5145189166069031, "learning_rate": 7.370651457487559e-06, "loss": 0.4541, "step": 1509 }, { "epoch": 1.2266450040617385, "grad_norm": 0.42092692852020264, "learning_rate": 7.366488451637126e-06, "loss": 0.4284, "step": 1510 }, { "epoch": 1.2274573517465475, "grad_norm": 0.4637621343135834, "learning_rate": 7.362323330625744e-06, "loss": 0.4257, "step": 1511 }, { "epoch": 1.2282696994313567, "grad_norm": 0.47523191571235657, "learning_rate": 7.358156098176185e-06, "loss": 0.4433, "step": 1512 }, { "epoch": 1.2290820471161656, "grad_norm": 0.4811180830001831, "learning_rate": 7.353986758013112e-06, "loss": 0.4309, "step": 1513 }, { "epoch": 1.2298943948009748, "grad_norm": 0.49534186720848083, "learning_rate": 7.349815313863068e-06, "loss": 0.4497, "step": 1514 }, { "epoch": 1.2307067424857838, "grad_norm": 0.47659990191459656, "learning_rate": 7.3456417694544804e-06, "loss": 0.459, "step": 1515 }, { "epoch": 1.231519090170593, "grad_norm": 0.5005613565444946, "learning_rate": 7.341466128517649e-06, "loss": 0.4376, "step": 1516 }, { "epoch": 1.232331437855402, "grad_norm": 0.4576586186885834, "learning_rate": 7.337288394784754e-06, "loss": 0.423, "step": 1517 }, { "epoch": 1.2331437855402112, "grad_norm": 0.48960912227630615, "learning_rate": 7.3331085719898385e-06, "loss": 0.4358, "step": 1518 }, { "epoch": 1.2339561332250204, "grad_norm": 0.482206791639328, "learning_rate": 7.32892666386882e-06, "loss": 0.4044, "step": 1519 }, { "epoch": 1.2347684809098294, "grad_norm": 0.4415574371814728, "learning_rate": 7.324742674159475e-06, "loss": 0.4278, "step": 1520 }, { "epoch": 1.2355808285946386, "grad_norm": 0.5646147131919861, "learning_rate": 7.3205566066014436e-06, "loss": 0.433, "step": 1521 }, { "epoch": 1.2363931762794476, "grad_norm": 0.47684621810913086, "learning_rate": 7.316368464936219e-06, "loss": 0.4442, "step": 1522 }, { "epoch": 1.2372055239642568, "grad_norm": 0.5337287783622742, "learning_rate": 7.312178252907154e-06, "loss": 0.4039, "step": 1523 }, { "epoch": 1.2380178716490657, "grad_norm": 0.54087233543396, "learning_rate": 7.307985974259445e-06, "loss": 0.4583, "step": 1524 }, { "epoch": 1.238830219333875, "grad_norm": 0.45964881777763367, "learning_rate": 7.303791632740143e-06, "loss": 0.4456, "step": 1525 }, { "epoch": 1.239642567018684, "grad_norm": 0.4499828517436981, "learning_rate": 7.2995952320981356e-06, "loss": 0.4068, "step": 1526 }, { "epoch": 1.2404549147034931, "grad_norm": 0.5199623703956604, "learning_rate": 7.2953967760841585e-06, "loss": 0.4325, "step": 1527 }, { "epoch": 1.2412672623883023, "grad_norm": 0.4122292995452881, "learning_rate": 7.291196268450777e-06, "loss": 0.442, "step": 1528 }, { "epoch": 1.2420796100731113, "grad_norm": 0.4271869957447052, "learning_rate": 7.286993712952394e-06, "loss": 0.4162, "step": 1529 }, { "epoch": 1.2428919577579203, "grad_norm": 0.5271120667457581, "learning_rate": 7.282789113345242e-06, "loss": 0.4513, "step": 1530 }, { "epoch": 1.2437043054427295, "grad_norm": 0.4166598320007324, "learning_rate": 7.278582473387382e-06, "loss": 0.4481, "step": 1531 }, { "epoch": 1.2445166531275387, "grad_norm": 0.5159174799919128, "learning_rate": 7.274373796838696e-06, "loss": 0.4692, "step": 1532 }, { "epoch": 1.2453290008123477, "grad_norm": 0.44366762042045593, "learning_rate": 7.27016308746089e-06, "loss": 0.4398, "step": 1533 }, { "epoch": 1.2461413484971569, "grad_norm": 0.4072638154029846, "learning_rate": 7.265950349017481e-06, "loss": 0.4407, "step": 1534 }, { "epoch": 1.2469536961819658, "grad_norm": 0.39424508810043335, "learning_rate": 7.261735585273804e-06, "loss": 0.4333, "step": 1535 }, { "epoch": 1.247766043866775, "grad_norm": 0.5278186202049255, "learning_rate": 7.257518799997007e-06, "loss": 0.4563, "step": 1536 }, { "epoch": 1.248578391551584, "grad_norm": 0.4509982466697693, "learning_rate": 7.253299996956038e-06, "loss": 0.4108, "step": 1537 }, { "epoch": 1.2493907392363932, "grad_norm": 0.41750016808509827, "learning_rate": 7.249079179921652e-06, "loss": 0.418, "step": 1538 }, { "epoch": 1.2502030869212022, "grad_norm": 0.4403035342693329, "learning_rate": 7.244856352666404e-06, "loss": 0.4188, "step": 1539 }, { "epoch": 1.2510154346060114, "grad_norm": 0.4419596195220947, "learning_rate": 7.240631518964646e-06, "loss": 0.4431, "step": 1540 }, { "epoch": 1.2518277822908206, "grad_norm": 0.3949885964393616, "learning_rate": 7.236404682592523e-06, "loss": 0.4191, "step": 1541 }, { "epoch": 1.2526401299756296, "grad_norm": 0.4505539536476135, "learning_rate": 7.232175847327969e-06, "loss": 0.4276, "step": 1542 }, { "epoch": 1.2534524776604385, "grad_norm": 0.46323591470718384, "learning_rate": 7.227945016950706e-06, "loss": 0.4383, "step": 1543 }, { "epoch": 1.2542648253452477, "grad_norm": 0.4992050528526306, "learning_rate": 7.223712195242238e-06, "loss": 0.4413, "step": 1544 }, { "epoch": 1.255077173030057, "grad_norm": 0.41323262453079224, "learning_rate": 7.2194773859858495e-06, "loss": 0.4477, "step": 1545 }, { "epoch": 1.255889520714866, "grad_norm": 0.4885469079017639, "learning_rate": 7.215240592966603e-06, "loss": 0.4415, "step": 1546 }, { "epoch": 1.2567018683996751, "grad_norm": 0.5068345069885254, "learning_rate": 7.2110018199713325e-06, "loss": 0.462, "step": 1547 }, { "epoch": 1.257514216084484, "grad_norm": 0.4222823977470398, "learning_rate": 7.206761070788641e-06, "loss": 0.436, "step": 1548 }, { "epoch": 1.2583265637692933, "grad_norm": 0.42105600237846375, "learning_rate": 7.202518349208898e-06, "loss": 0.4343, "step": 1549 }, { "epoch": 1.2591389114541023, "grad_norm": 0.46136173605918884, "learning_rate": 7.1982736590242375e-06, "loss": 0.438, "step": 1550 }, { "epoch": 1.2599512591389115, "grad_norm": 0.4464051127433777, "learning_rate": 7.194027004028552e-06, "loss": 0.4243, "step": 1551 }, { "epoch": 1.2607636068237205, "grad_norm": 0.3976934254169464, "learning_rate": 7.189778388017489e-06, "loss": 0.4426, "step": 1552 }, { "epoch": 1.2615759545085297, "grad_norm": 0.4691995084285736, "learning_rate": 7.185527814788451e-06, "loss": 0.439, "step": 1553 }, { "epoch": 1.2623883021933389, "grad_norm": 0.439982533454895, "learning_rate": 7.18127528814059e-06, "loss": 0.4132, "step": 1554 }, { "epoch": 1.2632006498781478, "grad_norm": 0.46406736969947815, "learning_rate": 7.1770208118748e-06, "loss": 0.4113, "step": 1555 }, { "epoch": 1.2640129975629568, "grad_norm": 0.4355163872241974, "learning_rate": 7.172764389793723e-06, "loss": 0.4403, "step": 1556 }, { "epoch": 1.264825345247766, "grad_norm": 0.44774025678634644, "learning_rate": 7.168506025701734e-06, "loss": 0.4289, "step": 1557 }, { "epoch": 1.2656376929325752, "grad_norm": 0.4686010479927063, "learning_rate": 7.164245723404951e-06, "loss": 0.4323, "step": 1558 }, { "epoch": 1.2664500406173842, "grad_norm": 0.4414907991886139, "learning_rate": 7.159983486711219e-06, "loss": 0.432, "step": 1559 }, { "epoch": 1.2672623883021934, "grad_norm": 0.5167536735534668, "learning_rate": 7.155719319430114e-06, "loss": 0.4456, "step": 1560 }, { "epoch": 1.2680747359870024, "grad_norm": 0.43762633204460144, "learning_rate": 7.151453225372938e-06, "loss": 0.4471, "step": 1561 }, { "epoch": 1.2688870836718116, "grad_norm": 0.47158631682395935, "learning_rate": 7.147185208352712e-06, "loss": 0.4415, "step": 1562 }, { "epoch": 1.2696994313566206, "grad_norm": 0.4374321401119232, "learning_rate": 7.142915272184179e-06, "loss": 0.4281, "step": 1563 }, { "epoch": 1.2705117790414298, "grad_norm": 0.46433117985725403, "learning_rate": 7.138643420683795e-06, "loss": 0.4332, "step": 1564 }, { "epoch": 1.2713241267262387, "grad_norm": 0.3978220820426941, "learning_rate": 7.134369657669733e-06, "loss": 0.4477, "step": 1565 }, { "epoch": 1.272136474411048, "grad_norm": 0.45805713534355164, "learning_rate": 7.130093986961868e-06, "loss": 0.4332, "step": 1566 }, { "epoch": 1.2729488220958571, "grad_norm": 0.45946675539016724, "learning_rate": 7.1258164123817834e-06, "loss": 0.4233, "step": 1567 }, { "epoch": 1.2737611697806661, "grad_norm": 0.4473760426044464, "learning_rate": 7.121536937752764e-06, "loss": 0.4338, "step": 1568 }, { "epoch": 1.274573517465475, "grad_norm": 0.4813023507595062, "learning_rate": 7.1172555668997925e-06, "loss": 0.4317, "step": 1569 }, { "epoch": 1.2753858651502843, "grad_norm": 0.43594062328338623, "learning_rate": 7.112972303649546e-06, "loss": 0.4428, "step": 1570 }, { "epoch": 1.2761982128350935, "grad_norm": 0.45701050758361816, "learning_rate": 7.108687151830394e-06, "loss": 0.4271, "step": 1571 }, { "epoch": 1.2770105605199025, "grad_norm": 0.43664485216140747, "learning_rate": 7.104400115272393e-06, "loss": 0.4357, "step": 1572 }, { "epoch": 1.2778229082047117, "grad_norm": 0.41145092248916626, "learning_rate": 7.100111197807286e-06, "loss": 0.4418, "step": 1573 }, { "epoch": 1.2786352558895206, "grad_norm": 0.40245774388313293, "learning_rate": 7.095820403268494e-06, "loss": 0.46, "step": 1574 }, { "epoch": 1.2794476035743299, "grad_norm": 0.41749677062034607, "learning_rate": 7.091527735491119e-06, "loss": 0.4568, "step": 1575 }, { "epoch": 1.2802599512591388, "grad_norm": 0.4158957004547119, "learning_rate": 7.087233198311936e-06, "loss": 0.4519, "step": 1576 }, { "epoch": 1.281072298943948, "grad_norm": 0.41396236419677734, "learning_rate": 7.082936795569388e-06, "loss": 0.4667, "step": 1577 }, { "epoch": 1.281884646628757, "grad_norm": 0.41558006405830383, "learning_rate": 7.078638531103592e-06, "loss": 0.4625, "step": 1578 }, { "epoch": 1.2826969943135662, "grad_norm": 0.36386802792549133, "learning_rate": 7.074338408756321e-06, "loss": 0.442, "step": 1579 }, { "epoch": 1.2835093419983754, "grad_norm": 0.4059138000011444, "learning_rate": 7.070036432371017e-06, "loss": 0.4193, "step": 1580 }, { "epoch": 1.2843216896831844, "grad_norm": 0.43558838963508606, "learning_rate": 7.065732605792771e-06, "loss": 0.4383, "step": 1581 }, { "epoch": 1.2851340373679934, "grad_norm": 0.3820415437221527, "learning_rate": 7.061426932868334e-06, "loss": 0.427, "step": 1582 }, { "epoch": 1.2859463850528026, "grad_norm": 0.4230335056781769, "learning_rate": 7.057119417446102e-06, "loss": 0.4597, "step": 1583 }, { "epoch": 1.2867587327376118, "grad_norm": 0.3983877897262573, "learning_rate": 7.052810063376124e-06, "loss": 0.4251, "step": 1584 }, { "epoch": 1.2875710804224207, "grad_norm": 0.4683416783809662, "learning_rate": 7.048498874510088e-06, "loss": 0.4261, "step": 1585 }, { "epoch": 1.28838342810723, "grad_norm": 0.38484784960746765, "learning_rate": 7.044185854701321e-06, "loss": 0.4151, "step": 1586 }, { "epoch": 1.289195775792039, "grad_norm": 0.4502795338630676, "learning_rate": 7.03987100780479e-06, "loss": 0.4308, "step": 1587 }, { "epoch": 1.2900081234768481, "grad_norm": 0.4492860734462738, "learning_rate": 7.035554337677094e-06, "loss": 0.4432, "step": 1588 }, { "epoch": 1.2908204711616573, "grad_norm": 0.48498469591140747, "learning_rate": 7.03123584817646e-06, "loss": 0.4096, "step": 1589 }, { "epoch": 1.2916328188464663, "grad_norm": 0.50933438539505, "learning_rate": 7.026915543162741e-06, "loss": 0.4451, "step": 1590 }, { "epoch": 1.2924451665312753, "grad_norm": 0.450048565864563, "learning_rate": 7.0225934264974175e-06, "loss": 0.4494, "step": 1591 }, { "epoch": 1.2932575142160845, "grad_norm": 0.5877475142478943, "learning_rate": 7.018269502043581e-06, "loss": 0.4383, "step": 1592 }, { "epoch": 1.2940698619008937, "grad_norm": 0.49015185236930847, "learning_rate": 7.013943773665947e-06, "loss": 0.4545, "step": 1593 }, { "epoch": 1.2948822095857027, "grad_norm": 0.502072274684906, "learning_rate": 7.009616245230838e-06, "loss": 0.4703, "step": 1594 }, { "epoch": 1.2956945572705119, "grad_norm": 0.447099894285202, "learning_rate": 7.005286920606188e-06, "loss": 0.4297, "step": 1595 }, { "epoch": 1.2965069049553208, "grad_norm": 0.41516390442848206, "learning_rate": 7.0009558036615355e-06, "loss": 0.4233, "step": 1596 }, { "epoch": 1.29731925264013, "grad_norm": 0.43323948979377747, "learning_rate": 6.996622898268021e-06, "loss": 0.4499, "step": 1597 }, { "epoch": 1.298131600324939, "grad_norm": 0.38644006848335266, "learning_rate": 6.992288208298383e-06, "loss": 0.4357, "step": 1598 }, { "epoch": 1.2989439480097482, "grad_norm": 0.4490811824798584, "learning_rate": 6.987951737626956e-06, "loss": 0.4409, "step": 1599 }, { "epoch": 1.2997562956945572, "grad_norm": 0.5150893926620483, "learning_rate": 6.983613490129666e-06, "loss": 0.4366, "step": 1600 }, { "epoch": 1.3005686433793664, "grad_norm": 0.42927542328834534, "learning_rate": 6.979273469684026e-06, "loss": 0.4345, "step": 1601 }, { "epoch": 1.3013809910641756, "grad_norm": 0.4878041446208954, "learning_rate": 6.974931680169136e-06, "loss": 0.448, "step": 1602 }, { "epoch": 1.3021933387489846, "grad_norm": 0.44421038031578064, "learning_rate": 6.970588125465674e-06, "loss": 0.4176, "step": 1603 }, { "epoch": 1.3030056864337936, "grad_norm": 0.3810291290283203, "learning_rate": 6.966242809455899e-06, "loss": 0.4131, "step": 1604 }, { "epoch": 1.3038180341186028, "grad_norm": 0.3599824011325836, "learning_rate": 6.961895736023641e-06, "loss": 0.4211, "step": 1605 }, { "epoch": 1.304630381803412, "grad_norm": 0.4699738025665283, "learning_rate": 6.957546909054304e-06, "loss": 0.4344, "step": 1606 }, { "epoch": 1.305442729488221, "grad_norm": 0.4445282220840454, "learning_rate": 6.953196332434856e-06, "loss": 0.4423, "step": 1607 }, { "epoch": 1.3062550771730301, "grad_norm": 0.39019691944122314, "learning_rate": 6.948844010053832e-06, "loss": 0.4366, "step": 1608 }, { "epoch": 1.307067424857839, "grad_norm": 0.4367586076259613, "learning_rate": 6.944489945801326e-06, "loss": 0.422, "step": 1609 }, { "epoch": 1.3078797725426483, "grad_norm": 0.3910801410675049, "learning_rate": 6.940134143568987e-06, "loss": 0.4361, "step": 1610 }, { "epoch": 1.3086921202274573, "grad_norm": 0.47419318556785583, "learning_rate": 6.935776607250022e-06, "loss": 0.4476, "step": 1611 }, { "epoch": 1.3095044679122665, "grad_norm": 0.4078432321548462, "learning_rate": 6.931417340739183e-06, "loss": 0.4398, "step": 1612 }, { "epoch": 1.3103168155970755, "grad_norm": 0.4126666188240051, "learning_rate": 6.927056347932772e-06, "loss": 0.4381, "step": 1613 }, { "epoch": 1.3111291632818847, "grad_norm": 0.4156714379787445, "learning_rate": 6.92269363272863e-06, "loss": 0.4519, "step": 1614 }, { "epoch": 1.3119415109666939, "grad_norm": 0.43638691306114197, "learning_rate": 6.918329199026143e-06, "loss": 0.4489, "step": 1615 }, { "epoch": 1.3127538586515028, "grad_norm": 0.36737242341041565, "learning_rate": 6.913963050726226e-06, "loss": 0.4263, "step": 1616 }, { "epoch": 1.3135662063363118, "grad_norm": 0.43266183137893677, "learning_rate": 6.909595191731337e-06, "loss": 0.4238, "step": 1617 }, { "epoch": 1.314378554021121, "grad_norm": 0.40290603041648865, "learning_rate": 6.905225625945451e-06, "loss": 0.4253, "step": 1618 }, { "epoch": 1.3151909017059302, "grad_norm": 0.4044530391693115, "learning_rate": 6.900854357274075e-06, "loss": 0.4644, "step": 1619 }, { "epoch": 1.3160032493907392, "grad_norm": 0.40050724148750305, "learning_rate": 6.896481389624239e-06, "loss": 0.4405, "step": 1620 }, { "epoch": 1.3168155970755484, "grad_norm": 0.4185737669467926, "learning_rate": 6.892106726904486e-06, "loss": 0.4279, "step": 1621 }, { "epoch": 1.3176279447603574, "grad_norm": 0.3980472981929779, "learning_rate": 6.887730373024881e-06, "loss": 0.4457, "step": 1622 }, { "epoch": 1.3184402924451666, "grad_norm": 0.39266127347946167, "learning_rate": 6.883352331896998e-06, "loss": 0.4355, "step": 1623 }, { "epoch": 1.3192526401299756, "grad_norm": 0.3797450363636017, "learning_rate": 6.878972607433915e-06, "loss": 0.4115, "step": 1624 }, { "epoch": 1.3200649878147848, "grad_norm": 0.43725162744522095, "learning_rate": 6.8745912035502205e-06, "loss": 0.4234, "step": 1625 }, { "epoch": 1.3208773354995937, "grad_norm": 0.44563373923301697, "learning_rate": 6.870208124161998e-06, "loss": 0.4148, "step": 1626 }, { "epoch": 1.321689683184403, "grad_norm": 0.41500020027160645, "learning_rate": 6.8658233731868355e-06, "loss": 0.4183, "step": 1627 }, { "epoch": 1.3225020308692121, "grad_norm": 0.4586881399154663, "learning_rate": 6.86143695454381e-06, "loss": 0.4292, "step": 1628 }, { "epoch": 1.3233143785540211, "grad_norm": 0.44428837299346924, "learning_rate": 6.857048872153491e-06, "loss": 0.4267, "step": 1629 }, { "epoch": 1.32412672623883, "grad_norm": 0.36325597763061523, "learning_rate": 6.852659129937936e-06, "loss": 0.4374, "step": 1630 }, { "epoch": 1.3249390739236393, "grad_norm": 0.4576355516910553, "learning_rate": 6.84826773182068e-06, "loss": 0.4379, "step": 1631 }, { "epoch": 1.3257514216084485, "grad_norm": 0.4577658772468567, "learning_rate": 6.843874681726747e-06, "loss": 0.4371, "step": 1632 }, { "epoch": 1.3265637692932575, "grad_norm": 0.4804665744304657, "learning_rate": 6.839479983582632e-06, "loss": 0.4325, "step": 1633 }, { "epoch": 1.3273761169780667, "grad_norm": 0.421561598777771, "learning_rate": 6.835083641316304e-06, "loss": 0.4513, "step": 1634 }, { "epoch": 1.3281884646628757, "grad_norm": 0.4404643177986145, "learning_rate": 6.830685658857203e-06, "loss": 0.4031, "step": 1635 }, { "epoch": 1.3290008123476849, "grad_norm": 0.4036388397216797, "learning_rate": 6.826286040136232e-06, "loss": 0.4286, "step": 1636 }, { "epoch": 1.3298131600324938, "grad_norm": 0.4190489947795868, "learning_rate": 6.82188478908576e-06, "loss": 0.4047, "step": 1637 }, { "epoch": 1.330625507717303, "grad_norm": 0.4152657091617584, "learning_rate": 6.817481909639611e-06, "loss": 0.4472, "step": 1638 }, { "epoch": 1.331437855402112, "grad_norm": 0.4394361078739166, "learning_rate": 6.813077405733068e-06, "loss": 0.4315, "step": 1639 }, { "epoch": 1.3322502030869212, "grad_norm": 0.41955333948135376, "learning_rate": 6.808671281302865e-06, "loss": 0.4242, "step": 1640 }, { "epoch": 1.3330625507717304, "grad_norm": 0.3929923474788666, "learning_rate": 6.804263540287184e-06, "loss": 0.4399, "step": 1641 }, { "epoch": 1.3338748984565394, "grad_norm": 0.46274328231811523, "learning_rate": 6.799854186625651e-06, "loss": 0.4459, "step": 1642 }, { "epoch": 1.3346872461413484, "grad_norm": 0.44302669167518616, "learning_rate": 6.795443224259335e-06, "loss": 0.4466, "step": 1643 }, { "epoch": 1.3354995938261576, "grad_norm": 0.41105154156684875, "learning_rate": 6.7910306571307425e-06, "loss": 0.4448, "step": 1644 }, { "epoch": 1.3363119415109668, "grad_norm": 0.43146127462387085, "learning_rate": 6.786616489183814e-06, "loss": 0.4709, "step": 1645 }, { "epoch": 1.3371242891957758, "grad_norm": 0.4550401568412781, "learning_rate": 6.78220072436392e-06, "loss": 0.4622, "step": 1646 }, { "epoch": 1.337936636880585, "grad_norm": 0.45016199350357056, "learning_rate": 6.7777833666178594e-06, "loss": 0.4452, "step": 1647 }, { "epoch": 1.338748984565394, "grad_norm": 0.42926549911499023, "learning_rate": 6.773364419893854e-06, "loss": 0.4442, "step": 1648 }, { "epoch": 1.3395613322502031, "grad_norm": 0.36318108439445496, "learning_rate": 6.768943888141548e-06, "loss": 0.4095, "step": 1649 }, { "epoch": 1.340373679935012, "grad_norm": 0.4023272395133972, "learning_rate": 6.764521775312e-06, "loss": 0.4328, "step": 1650 }, { "epoch": 1.3411860276198213, "grad_norm": 0.42067837715148926, "learning_rate": 6.760098085357681e-06, "loss": 0.4285, "step": 1651 }, { "epoch": 1.3419983753046303, "grad_norm": 0.4386485517024994, "learning_rate": 6.755672822232475e-06, "loss": 0.413, "step": 1652 }, { "epoch": 1.3428107229894395, "grad_norm": 0.3671899735927582, "learning_rate": 6.7512459898916704e-06, "loss": 0.4521, "step": 1653 }, { "epoch": 1.3436230706742487, "grad_norm": 0.43575480580329895, "learning_rate": 6.746817592291957e-06, "loss": 0.4462, "step": 1654 }, { "epoch": 1.3444354183590577, "grad_norm": 0.41306522488594055, "learning_rate": 6.742387633391423e-06, "loss": 0.4451, "step": 1655 }, { "epoch": 1.3452477660438666, "grad_norm": 0.37121060490608215, "learning_rate": 6.737956117149555e-06, "loss": 0.422, "step": 1656 }, { "epoch": 1.3460601137286758, "grad_norm": 0.4351979196071625, "learning_rate": 6.733523047527231e-06, "loss": 0.4155, "step": 1657 }, { "epoch": 1.346872461413485, "grad_norm": 0.46888065338134766, "learning_rate": 6.729088428486717e-06, "loss": 0.4262, "step": 1658 }, { "epoch": 1.347684809098294, "grad_norm": 0.4376490116119385, "learning_rate": 6.72465226399166e-06, "loss": 0.4417, "step": 1659 }, { "epoch": 1.3484971567831032, "grad_norm": 0.38900598883628845, "learning_rate": 6.720214558007094e-06, "loss": 0.4209, "step": 1660 }, { "epoch": 1.3493095044679122, "grad_norm": 0.4824555516242981, "learning_rate": 6.715775314499431e-06, "loss": 0.4272, "step": 1661 }, { "epoch": 1.3501218521527214, "grad_norm": 0.44631344079971313, "learning_rate": 6.711334537436451e-06, "loss": 0.4656, "step": 1662 }, { "epoch": 1.3509341998375304, "grad_norm": 0.36818522214889526, "learning_rate": 6.7068922307873104e-06, "loss": 0.4165, "step": 1663 }, { "epoch": 1.3517465475223396, "grad_norm": 0.5114501118659973, "learning_rate": 6.7024483985225295e-06, "loss": 0.4353, "step": 1664 }, { "epoch": 1.3525588952071486, "grad_norm": 0.4142792224884033, "learning_rate": 6.698003044613997e-06, "loss": 0.4389, "step": 1665 }, { "epoch": 1.3533712428919578, "grad_norm": 0.3969666063785553, "learning_rate": 6.693556173034953e-06, "loss": 0.4351, "step": 1666 }, { "epoch": 1.354183590576767, "grad_norm": 0.4240264296531677, "learning_rate": 6.689107787760002e-06, "loss": 0.4277, "step": 1667 }, { "epoch": 1.354995938261576, "grad_norm": 0.47877803444862366, "learning_rate": 6.6846578927650985e-06, "loss": 0.4293, "step": 1668 }, { "epoch": 1.355808285946385, "grad_norm": 0.4045059084892273, "learning_rate": 6.6802064920275475e-06, "loss": 0.423, "step": 1669 }, { "epoch": 1.3566206336311941, "grad_norm": 0.4621323049068451, "learning_rate": 6.675753589525996e-06, "loss": 0.4318, "step": 1670 }, { "epoch": 1.3574329813160033, "grad_norm": 0.39301690459251404, "learning_rate": 6.671299189240439e-06, "loss": 0.4295, "step": 1671 }, { "epoch": 1.3582453290008123, "grad_norm": 0.40826693177223206, "learning_rate": 6.666843295152207e-06, "loss": 0.4454, "step": 1672 }, { "epoch": 1.3590576766856215, "grad_norm": 0.45076584815979004, "learning_rate": 6.662385911243965e-06, "loss": 0.4197, "step": 1673 }, { "epoch": 1.3598700243704305, "grad_norm": 0.38355278968811035, "learning_rate": 6.657927041499711e-06, "loss": 0.408, "step": 1674 }, { "epoch": 1.3606823720552397, "grad_norm": 0.42149946093559265, "learning_rate": 6.653466689904771e-06, "loss": 0.4533, "step": 1675 }, { "epoch": 1.3614947197400489, "grad_norm": 0.39136481285095215, "learning_rate": 6.649004860445795e-06, "loss": 0.4395, "step": 1676 }, { "epoch": 1.3623070674248579, "grad_norm": 0.4275790750980377, "learning_rate": 6.644541557110753e-06, "loss": 0.4585, "step": 1677 }, { "epoch": 1.3631194151096668, "grad_norm": 0.49846091866493225, "learning_rate": 6.640076783888934e-06, "loss": 0.4625, "step": 1678 }, { "epoch": 1.363931762794476, "grad_norm": 0.36266589164733887, "learning_rate": 6.63561054477094e-06, "loss": 0.412, "step": 1679 }, { "epoch": 1.3647441104792852, "grad_norm": 0.43483972549438477, "learning_rate": 6.631142843748686e-06, "loss": 0.4188, "step": 1680 }, { "epoch": 1.3655564581640942, "grad_norm": 0.37910351157188416, "learning_rate": 6.626673684815389e-06, "loss": 0.409, "step": 1681 }, { "epoch": 1.3663688058489034, "grad_norm": 0.365454763174057, "learning_rate": 6.62220307196557e-06, "loss": 0.4358, "step": 1682 }, { "epoch": 1.3671811535337124, "grad_norm": 0.4166135787963867, "learning_rate": 6.617731009195052e-06, "loss": 0.4335, "step": 1683 }, { "epoch": 1.3679935012185216, "grad_norm": 0.40873050689697266, "learning_rate": 6.613257500500952e-06, "loss": 0.4341, "step": 1684 }, { "epoch": 1.3688058489033306, "grad_norm": 0.4261147081851959, "learning_rate": 6.60878254988168e-06, "loss": 0.4445, "step": 1685 }, { "epoch": 1.3696181965881398, "grad_norm": 0.4532489776611328, "learning_rate": 6.6043061613369356e-06, "loss": 0.4575, "step": 1686 }, { "epoch": 1.3704305442729487, "grad_norm": 0.3907226622104645, "learning_rate": 6.5998283388677005e-06, "loss": 0.438, "step": 1687 }, { "epoch": 1.371242891957758, "grad_norm": 0.41578343510627747, "learning_rate": 6.595349086476244e-06, "loss": 0.4134, "step": 1688 }, { "epoch": 1.3720552396425671, "grad_norm": 0.42906829714775085, "learning_rate": 6.590868408166105e-06, "loss": 0.4408, "step": 1689 }, { "epoch": 1.3728675873273761, "grad_norm": 0.4461037516593933, "learning_rate": 6.586386307942105e-06, "loss": 0.415, "step": 1690 }, { "epoch": 1.373679935012185, "grad_norm": 0.3974899649620056, "learning_rate": 6.581902789810333e-06, "loss": 0.4204, "step": 1691 }, { "epoch": 1.3744922826969943, "grad_norm": 0.46569058299064636, "learning_rate": 6.577417857778142e-06, "loss": 0.4457, "step": 1692 }, { "epoch": 1.3753046303818035, "grad_norm": 0.43743613362312317, "learning_rate": 6.572931515854158e-06, "loss": 0.4074, "step": 1693 }, { "epoch": 1.3761169780666125, "grad_norm": 0.43243423104286194, "learning_rate": 6.568443768048254e-06, "loss": 0.4277, "step": 1694 }, { "epoch": 1.3769293257514217, "grad_norm": 0.475460410118103, "learning_rate": 6.563954618371573e-06, "loss": 0.4111, "step": 1695 }, { "epoch": 1.3777416734362307, "grad_norm": 0.37060701847076416, "learning_rate": 6.559464070836501e-06, "loss": 0.452, "step": 1696 }, { "epoch": 1.3785540211210399, "grad_norm": 0.4569481313228607, "learning_rate": 6.5549721294566794e-06, "loss": 0.423, "step": 1697 }, { "epoch": 1.3793663688058488, "grad_norm": 0.4465733766555786, "learning_rate": 6.550478798246991e-06, "loss": 0.4285, "step": 1698 }, { "epoch": 1.380178716490658, "grad_norm": 0.40262332558631897, "learning_rate": 6.545984081223564e-06, "loss": 0.446, "step": 1699 }, { "epoch": 1.380991064175467, "grad_norm": 0.42378148436546326, "learning_rate": 6.541487982403767e-06, "loss": 0.4369, "step": 1700 }, { "epoch": 1.3818034118602762, "grad_norm": 0.37862205505371094, "learning_rate": 6.536990505806197e-06, "loss": 0.4355, "step": 1701 }, { "epoch": 1.3826157595450854, "grad_norm": 0.4010162949562073, "learning_rate": 6.532491655450688e-06, "loss": 0.4192, "step": 1702 }, { "epoch": 1.3834281072298944, "grad_norm": 0.40317994356155396, "learning_rate": 6.5279914353582996e-06, "loss": 0.4091, "step": 1703 }, { "epoch": 1.3842404549147034, "grad_norm": 0.39625298976898193, "learning_rate": 6.523489849551316e-06, "loss": 0.3837, "step": 1704 }, { "epoch": 1.3850528025995126, "grad_norm": 0.3515082597732544, "learning_rate": 6.518986902053244e-06, "loss": 0.4157, "step": 1705 }, { "epoch": 1.3858651502843218, "grad_norm": 0.42249590158462524, "learning_rate": 6.514482596888807e-06, "loss": 0.4181, "step": 1706 }, { "epoch": 1.3866774979691308, "grad_norm": 0.3499023914337158, "learning_rate": 6.509976938083936e-06, "loss": 0.4265, "step": 1707 }, { "epoch": 1.38748984565394, "grad_norm": 0.37456512451171875, "learning_rate": 6.505469929665781e-06, "loss": 0.4187, "step": 1708 }, { "epoch": 1.388302193338749, "grad_norm": 0.43128740787506104, "learning_rate": 6.500961575662691e-06, "loss": 0.4467, "step": 1709 }, { "epoch": 1.3891145410235581, "grad_norm": 0.3839423954486847, "learning_rate": 6.496451880104222e-06, "loss": 0.4213, "step": 1710 }, { "epoch": 1.3899268887083671, "grad_norm": 0.3888901174068451, "learning_rate": 6.491940847021128e-06, "loss": 0.4065, "step": 1711 }, { "epoch": 1.3907392363931763, "grad_norm": 0.425380676984787, "learning_rate": 6.487428480445357e-06, "loss": 0.4264, "step": 1712 }, { "epoch": 1.3915515840779853, "grad_norm": 0.42465609312057495, "learning_rate": 6.48291478441005e-06, "loss": 0.4303, "step": 1713 }, { "epoch": 1.3923639317627945, "grad_norm": 0.3866753876209259, "learning_rate": 6.478399762949537e-06, "loss": 0.4294, "step": 1714 }, { "epoch": 1.3931762794476037, "grad_norm": 0.44158628582954407, "learning_rate": 6.47388342009933e-06, "loss": 0.4225, "step": 1715 }, { "epoch": 1.3939886271324127, "grad_norm": 0.43780654668807983, "learning_rate": 6.469365759896128e-06, "loss": 0.4324, "step": 1716 }, { "epoch": 1.3948009748172217, "grad_norm": 0.3996858298778534, "learning_rate": 6.464846786377801e-06, "loss": 0.4373, "step": 1717 }, { "epoch": 1.3956133225020309, "grad_norm": 0.4278011620044708, "learning_rate": 6.460326503583394e-06, "loss": 0.4077, "step": 1718 }, { "epoch": 1.39642567018684, "grad_norm": 0.3918190002441406, "learning_rate": 6.4558049155531255e-06, "loss": 0.3941, "step": 1719 }, { "epoch": 1.397238017871649, "grad_norm": 0.39194434881210327, "learning_rate": 6.451282026328377e-06, "loss": 0.4191, "step": 1720 }, { "epoch": 1.3980503655564582, "grad_norm": 0.4683765769004822, "learning_rate": 6.4467578399516965e-06, "loss": 0.4311, "step": 1721 }, { "epoch": 1.3988627132412672, "grad_norm": 0.4600546658039093, "learning_rate": 6.442232360466789e-06, "loss": 0.4347, "step": 1722 }, { "epoch": 1.3996750609260764, "grad_norm": 0.3983111083507538, "learning_rate": 6.437705591918514e-06, "loss": 0.4288, "step": 1723 }, { "epoch": 1.4004874086108854, "grad_norm": 0.3580684959888458, "learning_rate": 6.433177538352887e-06, "loss": 0.4402, "step": 1724 }, { "epoch": 1.4012997562956946, "grad_norm": 0.45322737097740173, "learning_rate": 6.428648203817069e-06, "loss": 0.4314, "step": 1725 }, { "epoch": 1.4021121039805036, "grad_norm": 0.42257973551750183, "learning_rate": 6.424117592359367e-06, "loss": 0.4206, "step": 1726 }, { "epoch": 1.4029244516653128, "grad_norm": 0.3810868561267853, "learning_rate": 6.419585708029229e-06, "loss": 0.4328, "step": 1727 }, { "epoch": 1.403736799350122, "grad_norm": 0.4666275382041931, "learning_rate": 6.415052554877241e-06, "loss": 0.4171, "step": 1728 }, { "epoch": 1.404549147034931, "grad_norm": 0.39973410964012146, "learning_rate": 6.410518136955124e-06, "loss": 0.4343, "step": 1729 }, { "epoch": 1.40536149471974, "grad_norm": 0.3998740017414093, "learning_rate": 6.405982458315727e-06, "loss": 0.4169, "step": 1730 }, { "epoch": 1.4061738424045491, "grad_norm": 0.3792087733745575, "learning_rate": 6.401445523013028e-06, "loss": 0.45, "step": 1731 }, { "epoch": 1.4069861900893583, "grad_norm": 0.4601117968559265, "learning_rate": 6.396907335102127e-06, "loss": 0.435, "step": 1732 }, { "epoch": 1.4077985377741673, "grad_norm": 0.3858169913291931, "learning_rate": 6.392367898639245e-06, "loss": 0.4334, "step": 1733 }, { "epoch": 1.4086108854589765, "grad_norm": 0.3918313682079315, "learning_rate": 6.387827217681717e-06, "loss": 0.4433, "step": 1734 }, { "epoch": 1.4094232331437855, "grad_norm": 0.4090438485145569, "learning_rate": 6.383285296287991e-06, "loss": 0.4151, "step": 1735 }, { "epoch": 1.4102355808285947, "grad_norm": 0.39105769991874695, "learning_rate": 6.378742138517624e-06, "loss": 0.4133, "step": 1736 }, { "epoch": 1.4110479285134037, "grad_norm": 0.4124877154827118, "learning_rate": 6.37419774843128e-06, "loss": 0.4275, "step": 1737 }, { "epoch": 1.4118602761982129, "grad_norm": 0.38867658376693726, "learning_rate": 6.3696521300907214e-06, "loss": 0.4273, "step": 1738 }, { "epoch": 1.4126726238830218, "grad_norm": 0.4474501311779022, "learning_rate": 6.3651052875588085e-06, "loss": 0.4019, "step": 1739 }, { "epoch": 1.413484971567831, "grad_norm": 0.4993278384208679, "learning_rate": 6.360557224899496e-06, "loss": 0.4083, "step": 1740 }, { "epoch": 1.4142973192526402, "grad_norm": 0.43720370531082153, "learning_rate": 6.356007946177833e-06, "loss": 0.4615, "step": 1741 }, { "epoch": 1.4151096669374492, "grad_norm": 0.4840840697288513, "learning_rate": 6.351457455459953e-06, "loss": 0.4228, "step": 1742 }, { "epoch": 1.4159220146222582, "grad_norm": 0.5005937218666077, "learning_rate": 6.346905756813069e-06, "loss": 0.4325, "step": 1743 }, { "epoch": 1.4167343623070674, "grad_norm": 0.46481356024742126, "learning_rate": 6.342352854305477e-06, "loss": 0.4401, "step": 1744 }, { "epoch": 1.4175467099918766, "grad_norm": 0.4409886300563812, "learning_rate": 6.337798752006551e-06, "loss": 0.4245, "step": 1745 }, { "epoch": 1.4183590576766856, "grad_norm": 0.4884374439716339, "learning_rate": 6.333243453986734e-06, "loss": 0.4394, "step": 1746 }, { "epoch": 1.4191714053614948, "grad_norm": 0.4498332440853119, "learning_rate": 6.32868696431754e-06, "loss": 0.4344, "step": 1747 }, { "epoch": 1.4199837530463038, "grad_norm": 0.4028746485710144, "learning_rate": 6.324129287071546e-06, "loss": 0.4205, "step": 1748 }, { "epoch": 1.420796100731113, "grad_norm": 0.44106325507164, "learning_rate": 6.3195704263223914e-06, "loss": 0.4203, "step": 1749 }, { "epoch": 1.421608448415922, "grad_norm": 0.4318805932998657, "learning_rate": 6.315010386144776e-06, "loss": 0.4639, "step": 1750 }, { "epoch": 1.4224207961007311, "grad_norm": 0.3923114836215973, "learning_rate": 6.310449170614446e-06, "loss": 0.4366, "step": 1751 }, { "epoch": 1.4232331437855401, "grad_norm": 0.45467549562454224, "learning_rate": 6.305886783808209e-06, "loss": 0.4402, "step": 1752 }, { "epoch": 1.4240454914703493, "grad_norm": 0.44564130902290344, "learning_rate": 6.301323229803913e-06, "loss": 0.4265, "step": 1753 }, { "epoch": 1.4248578391551585, "grad_norm": 0.44283732771873474, "learning_rate": 6.296758512680448e-06, "loss": 0.4201, "step": 1754 }, { "epoch": 1.4256701868399675, "grad_norm": 0.4538181722164154, "learning_rate": 6.292192636517744e-06, "loss": 0.4069, "step": 1755 }, { "epoch": 1.4264825345247765, "grad_norm": 0.4443502426147461, "learning_rate": 6.287625605396774e-06, "loss": 0.4296, "step": 1756 }, { "epoch": 1.4272948822095857, "grad_norm": 0.4522043466567993, "learning_rate": 6.2830574233995356e-06, "loss": 0.408, "step": 1757 }, { "epoch": 1.4281072298943949, "grad_norm": 0.4382094442844391, "learning_rate": 6.278488094609057e-06, "loss": 0.4545, "step": 1758 }, { "epoch": 1.4289195775792038, "grad_norm": 0.4091951549053192, "learning_rate": 6.273917623109392e-06, "loss": 0.4326, "step": 1759 }, { "epoch": 1.429731925264013, "grad_norm": 0.42136234045028687, "learning_rate": 6.269346012985617e-06, "loss": 0.4375, "step": 1760 }, { "epoch": 1.430544272948822, "grad_norm": 0.3803732991218567, "learning_rate": 6.264773268323821e-06, "loss": 0.4353, "step": 1761 }, { "epoch": 1.4313566206336312, "grad_norm": 0.43649783730506897, "learning_rate": 6.260199393211115e-06, "loss": 0.4128, "step": 1762 }, { "epoch": 1.4321689683184404, "grad_norm": 0.44440197944641113, "learning_rate": 6.255624391735614e-06, "loss": 0.4312, "step": 1763 }, { "epoch": 1.4329813160032494, "grad_norm": 0.425365686416626, "learning_rate": 6.2510482679864425e-06, "loss": 0.4243, "step": 1764 }, { "epoch": 1.4337936636880584, "grad_norm": 0.41090139746665955, "learning_rate": 6.246471026053727e-06, "loss": 0.4403, "step": 1765 }, { "epoch": 1.4346060113728676, "grad_norm": 0.4639219343662262, "learning_rate": 6.241892670028595e-06, "loss": 0.464, "step": 1766 }, { "epoch": 1.4354183590576768, "grad_norm": 0.47441554069519043, "learning_rate": 6.237313204003167e-06, "loss": 0.4261, "step": 1767 }, { "epoch": 1.4362307067424858, "grad_norm": 0.4320046305656433, "learning_rate": 6.23273263207056e-06, "loss": 0.4105, "step": 1768 }, { "epoch": 1.437043054427295, "grad_norm": 0.4138926565647125, "learning_rate": 6.228150958324872e-06, "loss": 0.3967, "step": 1769 }, { "epoch": 1.437855402112104, "grad_norm": 0.4537779688835144, "learning_rate": 6.223568186861195e-06, "loss": 0.4449, "step": 1770 }, { "epoch": 1.4386677497969131, "grad_norm": 0.4046548306941986, "learning_rate": 6.218984321775596e-06, "loss": 0.4367, "step": 1771 }, { "epoch": 1.4394800974817221, "grad_norm": 0.4062942862510681, "learning_rate": 6.214399367165122e-06, "loss": 0.4193, "step": 1772 }, { "epoch": 1.4402924451665313, "grad_norm": 0.3840895891189575, "learning_rate": 6.209813327127792e-06, "loss": 0.4081, "step": 1773 }, { "epoch": 1.4411047928513403, "grad_norm": 0.41630712151527405, "learning_rate": 6.205226205762595e-06, "loss": 0.4203, "step": 1774 }, { "epoch": 1.4419171405361495, "grad_norm": 0.45168817043304443, "learning_rate": 6.20063800716949e-06, "loss": 0.4305, "step": 1775 }, { "epoch": 1.4427294882209587, "grad_norm": 0.4464397132396698, "learning_rate": 6.196048735449396e-06, "loss": 0.4202, "step": 1776 }, { "epoch": 1.4435418359057677, "grad_norm": 0.46545231342315674, "learning_rate": 6.1914583947041906e-06, "loss": 0.4714, "step": 1777 }, { "epoch": 1.4443541835905767, "grad_norm": 0.43935084342956543, "learning_rate": 6.186866989036708e-06, "loss": 0.4402, "step": 1778 }, { "epoch": 1.4451665312753859, "grad_norm": 0.42537355422973633, "learning_rate": 6.1822745225507355e-06, "loss": 0.4442, "step": 1779 }, { "epoch": 1.445978878960195, "grad_norm": 0.3799642026424408, "learning_rate": 6.177680999351006e-06, "loss": 0.438, "step": 1780 }, { "epoch": 1.446791226645004, "grad_norm": 0.4475252330303192, "learning_rate": 6.173086423543198e-06, "loss": 0.4537, "step": 1781 }, { "epoch": 1.4476035743298132, "grad_norm": 0.4149285554885864, "learning_rate": 6.168490799233931e-06, "loss": 0.4252, "step": 1782 }, { "epoch": 1.4484159220146222, "grad_norm": 0.3624977171421051, "learning_rate": 6.163894130530761e-06, "loss": 0.4312, "step": 1783 }, { "epoch": 1.4492282696994314, "grad_norm": 0.47660931944847107, "learning_rate": 6.159296421542178e-06, "loss": 0.4458, "step": 1784 }, { "epoch": 1.4500406173842404, "grad_norm": 0.4471814036369324, "learning_rate": 6.154697676377605e-06, "loss": 0.4171, "step": 1785 }, { "epoch": 1.4508529650690496, "grad_norm": 0.3612496256828308, "learning_rate": 6.150097899147384e-06, "loss": 0.4236, "step": 1786 }, { "epoch": 1.4516653127538586, "grad_norm": 0.4935956597328186, "learning_rate": 6.1454970939627845e-06, "loss": 0.4671, "step": 1787 }, { "epoch": 1.4524776604386678, "grad_norm": 0.4217885136604309, "learning_rate": 6.140895264935993e-06, "loss": 0.4415, "step": 1788 }, { "epoch": 1.453290008123477, "grad_norm": 0.4079470634460449, "learning_rate": 6.136292416180114e-06, "loss": 0.4605, "step": 1789 }, { "epoch": 1.454102355808286, "grad_norm": 0.416427880525589, "learning_rate": 6.131688551809161e-06, "loss": 0.4185, "step": 1790 }, { "epoch": 1.454914703493095, "grad_norm": 0.42635297775268555, "learning_rate": 6.127083675938053e-06, "loss": 0.4634, "step": 1791 }, { "epoch": 1.4557270511779041, "grad_norm": 0.38852816820144653, "learning_rate": 6.122477792682616e-06, "loss": 0.431, "step": 1792 }, { "epoch": 1.4565393988627133, "grad_norm": 0.4585477113723755, "learning_rate": 6.117870906159577e-06, "loss": 0.4519, "step": 1793 }, { "epoch": 1.4573517465475223, "grad_norm": 0.4271377921104431, "learning_rate": 6.113263020486559e-06, "loss": 0.3997, "step": 1794 }, { "epoch": 1.4581640942323315, "grad_norm": 0.3462006747722626, "learning_rate": 6.1086541397820785e-06, "loss": 0.4336, "step": 1795 }, { "epoch": 1.4589764419171405, "grad_norm": 0.39981532096862793, "learning_rate": 6.104044268165539e-06, "loss": 0.4275, "step": 1796 }, { "epoch": 1.4597887896019497, "grad_norm": 0.4149695932865143, "learning_rate": 6.0994334097572305e-06, "loss": 0.4007, "step": 1797 }, { "epoch": 1.4606011372867587, "grad_norm": 0.45196834206581116, "learning_rate": 6.094821568678328e-06, "loss": 0.4295, "step": 1798 }, { "epoch": 1.4614134849715679, "grad_norm": 0.39436912536621094, "learning_rate": 6.090208749050881e-06, "loss": 0.4003, "step": 1799 }, { "epoch": 1.4622258326563768, "grad_norm": 0.423759400844574, "learning_rate": 6.085594954997814e-06, "loss": 0.4177, "step": 1800 }, { "epoch": 1.463038180341186, "grad_norm": 0.4793779253959656, "learning_rate": 6.080980190642927e-06, "loss": 0.4417, "step": 1801 }, { "epoch": 1.4638505280259952, "grad_norm": 0.4417078197002411, "learning_rate": 6.076364460110881e-06, "loss": 0.4376, "step": 1802 }, { "epoch": 1.4646628757108042, "grad_norm": 0.38632476329803467, "learning_rate": 6.071747767527204e-06, "loss": 0.4372, "step": 1803 }, { "epoch": 1.4654752233956132, "grad_norm": 0.3776713013648987, "learning_rate": 6.067130117018284e-06, "loss": 0.424, "step": 1804 }, { "epoch": 1.4662875710804224, "grad_norm": 0.37733107805252075, "learning_rate": 6.062511512711362e-06, "loss": 0.4226, "step": 1805 }, { "epoch": 1.4670999187652316, "grad_norm": 0.3816770613193512, "learning_rate": 6.057891958734538e-06, "loss": 0.409, "step": 1806 }, { "epoch": 1.4679122664500406, "grad_norm": 0.4150124788284302, "learning_rate": 6.053271459216755e-06, "loss": 0.4476, "step": 1807 }, { "epoch": 1.4687246141348498, "grad_norm": 0.34513142704963684, "learning_rate": 6.048650018287803e-06, "loss": 0.4163, "step": 1808 }, { "epoch": 1.4695369618196588, "grad_norm": 0.37753230333328247, "learning_rate": 6.044027640078312e-06, "loss": 0.4201, "step": 1809 }, { "epoch": 1.470349309504468, "grad_norm": 0.3681010901927948, "learning_rate": 6.0394043287197524e-06, "loss": 0.4123, "step": 1810 }, { "epoch": 1.471161657189277, "grad_norm": 0.3817704916000366, "learning_rate": 6.034780088344427e-06, "loss": 0.4116, "step": 1811 }, { "epoch": 1.4719740048740861, "grad_norm": 0.37053364515304565, "learning_rate": 6.030154923085469e-06, "loss": 0.4258, "step": 1812 }, { "epoch": 1.4727863525588951, "grad_norm": 0.3934813141822815, "learning_rate": 6.025528837076841e-06, "loss": 0.4461, "step": 1813 }, { "epoch": 1.4735987002437043, "grad_norm": 0.4011129140853882, "learning_rate": 6.020901834453322e-06, "loss": 0.4245, "step": 1814 }, { "epoch": 1.4744110479285135, "grad_norm": 0.4187600314617157, "learning_rate": 6.016273919350517e-06, "loss": 0.434, "step": 1815 }, { "epoch": 1.4752233956133225, "grad_norm": 0.4242841899394989, "learning_rate": 6.011645095904845e-06, "loss": 0.4271, "step": 1816 }, { "epoch": 1.4760357432981315, "grad_norm": 0.3730866312980652, "learning_rate": 6.0070153682535315e-06, "loss": 0.4339, "step": 1817 }, { "epoch": 1.4768480909829407, "grad_norm": 0.3687885105609894, "learning_rate": 6.002384740534619e-06, "loss": 0.4293, "step": 1818 }, { "epoch": 1.4776604386677499, "grad_norm": 0.4160657227039337, "learning_rate": 5.997753216886948e-06, "loss": 0.4695, "step": 1819 }, { "epoch": 1.4784727863525589, "grad_norm": 0.3806304931640625, "learning_rate": 5.993120801450162e-06, "loss": 0.4181, "step": 1820 }, { "epoch": 1.479285134037368, "grad_norm": 0.4010554850101471, "learning_rate": 5.9884874983647e-06, "loss": 0.4362, "step": 1821 }, { "epoch": 1.480097481722177, "grad_norm": 0.35981523990631104, "learning_rate": 5.983853311771799e-06, "loss": 0.4124, "step": 1822 }, { "epoch": 1.4809098294069862, "grad_norm": 0.369689017534256, "learning_rate": 5.979218245813478e-06, "loss": 0.4169, "step": 1823 }, { "epoch": 1.4817221770917952, "grad_norm": 0.41156864166259766, "learning_rate": 5.974582304632551e-06, "loss": 0.4341, "step": 1824 }, { "epoch": 1.4825345247766044, "grad_norm": 0.3808959722518921, "learning_rate": 5.969945492372606e-06, "loss": 0.422, "step": 1825 }, { "epoch": 1.4833468724614134, "grad_norm": 0.4518260061740875, "learning_rate": 5.965307813178015e-06, "loss": 0.4757, "step": 1826 }, { "epoch": 1.4841592201462226, "grad_norm": 0.4389316737651825, "learning_rate": 5.960669271193922e-06, "loss": 0.4722, "step": 1827 }, { "epoch": 1.4849715678310318, "grad_norm": 0.3936663269996643, "learning_rate": 5.9560298705662444e-06, "loss": 0.4303, "step": 1828 }, { "epoch": 1.4857839155158408, "grad_norm": 0.4025488793849945, "learning_rate": 5.951389615441666e-06, "loss": 0.435, "step": 1829 }, { "epoch": 1.4865962632006497, "grad_norm": 0.4273880124092102, "learning_rate": 5.9467485099676325e-06, "loss": 0.4105, "step": 1830 }, { "epoch": 1.487408610885459, "grad_norm": 0.37816867232322693, "learning_rate": 5.942106558292352e-06, "loss": 0.4468, "step": 1831 }, { "epoch": 1.4882209585702681, "grad_norm": 0.4034191966056824, "learning_rate": 5.93746376456479e-06, "loss": 0.4424, "step": 1832 }, { "epoch": 1.4890333062550771, "grad_norm": 0.35589340329170227, "learning_rate": 5.932820132934661e-06, "loss": 0.3949, "step": 1833 }, { "epoch": 1.4898456539398863, "grad_norm": 0.4092126488685608, "learning_rate": 5.928175667552432e-06, "loss": 0.4488, "step": 1834 }, { "epoch": 1.4906580016246953, "grad_norm": 0.3467622399330139, "learning_rate": 5.923530372569312e-06, "loss": 0.4237, "step": 1835 }, { "epoch": 1.4914703493095045, "grad_norm": 0.4262157380580902, "learning_rate": 5.9188842521372535e-06, "loss": 0.4445, "step": 1836 }, { "epoch": 1.4922826969943135, "grad_norm": 0.40147802233695984, "learning_rate": 5.914237310408947e-06, "loss": 0.4149, "step": 1837 }, { "epoch": 1.4930950446791227, "grad_norm": 0.4056001603603363, "learning_rate": 5.909589551537814e-06, "loss": 0.4372, "step": 1838 }, { "epoch": 1.4939073923639317, "grad_norm": 0.3613736927509308, "learning_rate": 5.904940979678012e-06, "loss": 0.4106, "step": 1839 }, { "epoch": 1.4947197400487409, "grad_norm": 0.3957826495170593, "learning_rate": 5.90029159898442e-06, "loss": 0.4002, "step": 1840 }, { "epoch": 1.49553208773355, "grad_norm": 0.40742602944374084, "learning_rate": 5.895641413612643e-06, "loss": 0.4537, "step": 1841 }, { "epoch": 1.496344435418359, "grad_norm": 0.38242772221565247, "learning_rate": 5.890990427719003e-06, "loss": 0.3988, "step": 1842 }, { "epoch": 1.497156783103168, "grad_norm": 0.37996768951416016, "learning_rate": 5.886338645460539e-06, "loss": 0.4388, "step": 1843 }, { "epoch": 1.4979691307879772, "grad_norm": 0.4365507960319519, "learning_rate": 5.881686070995001e-06, "loss": 0.4444, "step": 1844 }, { "epoch": 1.4987814784727864, "grad_norm": 0.4337053894996643, "learning_rate": 5.877032708480847e-06, "loss": 0.4339, "step": 1845 }, { "epoch": 1.4995938261575954, "grad_norm": 0.37773749232292175, "learning_rate": 5.872378562077241e-06, "loss": 0.4152, "step": 1846 }, { "epoch": 1.5004061738424046, "grad_norm": 0.35724160075187683, "learning_rate": 5.867723635944047e-06, "loss": 0.4142, "step": 1847 }, { "epoch": 1.5012185215272136, "grad_norm": 0.3920603394508362, "learning_rate": 5.863067934241823e-06, "loss": 0.4504, "step": 1848 }, { "epoch": 1.5020308692120228, "grad_norm": 0.3865125775337219, "learning_rate": 5.8584114611318225e-06, "loss": 0.4141, "step": 1849 }, { "epoch": 1.502843216896832, "grad_norm": 0.3598440885543823, "learning_rate": 5.853754220775992e-06, "loss": 0.418, "step": 1850 }, { "epoch": 1.503655564581641, "grad_norm": 0.39134517312049866, "learning_rate": 5.849096217336956e-06, "loss": 0.4171, "step": 1851 }, { "epoch": 1.50446791226645, "grad_norm": 0.4091382920742035, "learning_rate": 5.844437454978029e-06, "loss": 0.4152, "step": 1852 }, { "epoch": 1.5052802599512591, "grad_norm": 0.3923005163669586, "learning_rate": 5.8397779378631955e-06, "loss": 0.439, "step": 1853 }, { "epoch": 1.5060926076360683, "grad_norm": 0.38986265659332275, "learning_rate": 5.835117670157126e-06, "loss": 0.4383, "step": 1854 }, { "epoch": 1.5069049553208773, "grad_norm": 0.41697338223457336, "learning_rate": 5.83045665602515e-06, "loss": 0.4416, "step": 1855 }, { "epoch": 1.5077173030056863, "grad_norm": 0.3947635293006897, "learning_rate": 5.82579489963327e-06, "loss": 0.4299, "step": 1856 }, { "epoch": 1.5085296506904955, "grad_norm": 0.4014508128166199, "learning_rate": 5.8211324051481534e-06, "loss": 0.4454, "step": 1857 }, { "epoch": 1.5093419983753047, "grad_norm": 0.3723866939544678, "learning_rate": 5.816469176737122e-06, "loss": 0.4328, "step": 1858 }, { "epoch": 1.510154346060114, "grad_norm": 0.39850541949272156, "learning_rate": 5.8118052185681584e-06, "loss": 0.4271, "step": 1859 }, { "epoch": 1.5109666937449229, "grad_norm": 0.36388206481933594, "learning_rate": 5.807140534809896e-06, "loss": 0.4128, "step": 1860 }, { "epoch": 1.5117790414297319, "grad_norm": 0.4000585079193115, "learning_rate": 5.802475129631616e-06, "loss": 0.4496, "step": 1861 }, { "epoch": 1.512591389114541, "grad_norm": 0.4499792754650116, "learning_rate": 5.797809007203245e-06, "loss": 0.4181, "step": 1862 }, { "epoch": 1.5134037367993503, "grad_norm": 0.4053175151348114, "learning_rate": 5.79314217169535e-06, "loss": 0.4187, "step": 1863 }, { "epoch": 1.5142160844841592, "grad_norm": 0.38246864080429077, "learning_rate": 5.788474627279137e-06, "loss": 0.4363, "step": 1864 }, { "epoch": 1.5150284321689682, "grad_norm": 0.45316845178604126, "learning_rate": 5.783806378126443e-06, "loss": 0.4121, "step": 1865 }, { "epoch": 1.5158407798537774, "grad_norm": 0.455737829208374, "learning_rate": 5.779137428409738e-06, "loss": 0.4118, "step": 1866 }, { "epoch": 1.5166531275385866, "grad_norm": 0.4324571192264557, "learning_rate": 5.774467782302115e-06, "loss": 0.4068, "step": 1867 }, { "epoch": 1.5174654752233956, "grad_norm": 0.43511831760406494, "learning_rate": 5.7697974439772934e-06, "loss": 0.4403, "step": 1868 }, { "epoch": 1.5182778229082046, "grad_norm": 0.4236341118812561, "learning_rate": 5.765126417609606e-06, "loss": 0.4099, "step": 1869 }, { "epoch": 1.5190901705930138, "grad_norm": 0.3625388741493225, "learning_rate": 5.760454707374004e-06, "loss": 0.4224, "step": 1870 }, { "epoch": 1.519902518277823, "grad_norm": 0.4325239062309265, "learning_rate": 5.755782317446051e-06, "loss": 0.4351, "step": 1871 }, { "epoch": 1.5207148659626322, "grad_norm": 0.4011934995651245, "learning_rate": 5.751109252001916e-06, "loss": 0.4122, "step": 1872 }, { "epoch": 1.5215272136474411, "grad_norm": 0.37297022342681885, "learning_rate": 5.746435515218373e-06, "loss": 0.4371, "step": 1873 }, { "epoch": 1.5223395613322501, "grad_norm": 0.4015800654888153, "learning_rate": 5.741761111272792e-06, "loss": 0.4334, "step": 1874 }, { "epoch": 1.5231519090170593, "grad_norm": 0.42782503366470337, "learning_rate": 5.737086044343145e-06, "loss": 0.4373, "step": 1875 }, { "epoch": 1.5239642567018685, "grad_norm": 0.398815780878067, "learning_rate": 5.732410318607993e-06, "loss": 0.4397, "step": 1876 }, { "epoch": 1.5247766043866775, "grad_norm": 0.39515969157218933, "learning_rate": 5.72773393824649e-06, "loss": 0.4328, "step": 1877 }, { "epoch": 1.5255889520714865, "grad_norm": 0.4051194489002228, "learning_rate": 5.723056907438369e-06, "loss": 0.4448, "step": 1878 }, { "epoch": 1.5264012997562957, "grad_norm": 0.39927083253860474, "learning_rate": 5.718379230363948e-06, "loss": 0.457, "step": 1879 }, { "epoch": 1.5272136474411049, "grad_norm": 0.36176034808158875, "learning_rate": 5.713700911204121e-06, "loss": 0.4572, "step": 1880 }, { "epoch": 1.5280259951259139, "grad_norm": 0.41269826889038086, "learning_rate": 5.709021954140359e-06, "loss": 0.4197, "step": 1881 }, { "epoch": 1.5288383428107228, "grad_norm": 0.5117304921150208, "learning_rate": 5.704342363354701e-06, "loss": 0.4453, "step": 1882 }, { "epoch": 1.529650690495532, "grad_norm": 0.382647842168808, "learning_rate": 5.6996621430297514e-06, "loss": 0.4541, "step": 1883 }, { "epoch": 1.5304630381803412, "grad_norm": 0.4138002097606659, "learning_rate": 5.6949812973486785e-06, "loss": 0.4356, "step": 1884 }, { "epoch": 1.5312753858651504, "grad_norm": 0.4344785511493683, "learning_rate": 5.690299830495211e-06, "loss": 0.4543, "step": 1885 }, { "epoch": 1.5320877335499594, "grad_norm": 0.4192240536212921, "learning_rate": 5.685617746653629e-06, "loss": 0.4133, "step": 1886 }, { "epoch": 1.5329000812347684, "grad_norm": 0.3696334958076477, "learning_rate": 5.680935050008767e-06, "loss": 0.4405, "step": 1887 }, { "epoch": 1.5337124289195776, "grad_norm": 0.3675175607204437, "learning_rate": 5.676251744746008e-06, "loss": 0.4168, "step": 1888 }, { "epoch": 1.5345247766043868, "grad_norm": 0.41829437017440796, "learning_rate": 5.671567835051274e-06, "loss": 0.4297, "step": 1889 }, { "epoch": 1.5353371242891958, "grad_norm": 0.3887537717819214, "learning_rate": 5.666883325111035e-06, "loss": 0.4352, "step": 1890 }, { "epoch": 1.5361494719740048, "grad_norm": 0.3815261423587799, "learning_rate": 5.662198219112291e-06, "loss": 0.4592, "step": 1891 }, { "epoch": 1.536961819658814, "grad_norm": 0.41823646426200867, "learning_rate": 5.657512521242579e-06, "loss": 0.4413, "step": 1892 }, { "epoch": 1.5377741673436232, "grad_norm": 0.43080177903175354, "learning_rate": 5.652826235689961e-06, "loss": 0.4173, "step": 1893 }, { "epoch": 1.5385865150284321, "grad_norm": 0.3884052634239197, "learning_rate": 5.648139366643026e-06, "loss": 0.454, "step": 1894 }, { "epoch": 1.5393988627132411, "grad_norm": 0.42493098974227905, "learning_rate": 5.643451918290885e-06, "loss": 0.4259, "step": 1895 }, { "epoch": 1.5402112103980503, "grad_norm": 0.46387237310409546, "learning_rate": 5.638763894823169e-06, "loss": 0.4482, "step": 1896 }, { "epoch": 1.5410235580828595, "grad_norm": 0.3827616274356842, "learning_rate": 5.634075300430017e-06, "loss": 0.4113, "step": 1897 }, { "epoch": 1.5418359057676687, "grad_norm": 0.46530038118362427, "learning_rate": 5.6293861393020855e-06, "loss": 0.4024, "step": 1898 }, { "epoch": 1.5426482534524777, "grad_norm": 0.41372182965278625, "learning_rate": 5.624696415630532e-06, "loss": 0.4508, "step": 1899 }, { "epoch": 1.5434606011372867, "grad_norm": 0.4451163411140442, "learning_rate": 5.620006133607017e-06, "loss": 0.4269, "step": 1900 }, { "epoch": 1.5442729488220959, "grad_norm": 0.41111963987350464, "learning_rate": 5.615315297423707e-06, "loss": 0.4071, "step": 1901 }, { "epoch": 1.545085296506905, "grad_norm": 0.3793233036994934, "learning_rate": 5.610623911273254e-06, "loss": 0.4282, "step": 1902 }, { "epoch": 1.545897644191714, "grad_norm": 0.398641973733902, "learning_rate": 5.605931979348807e-06, "loss": 0.4131, "step": 1903 }, { "epoch": 1.546709991876523, "grad_norm": 0.4070829153060913, "learning_rate": 5.601239505844005e-06, "loss": 0.4196, "step": 1904 }, { "epoch": 1.5475223395613322, "grad_norm": 0.45268306136131287, "learning_rate": 5.596546494952965e-06, "loss": 0.4057, "step": 1905 }, { "epoch": 1.5483346872461414, "grad_norm": 0.41346582770347595, "learning_rate": 5.591852950870287e-06, "loss": 0.4306, "step": 1906 }, { "epoch": 1.5491470349309504, "grad_norm": 0.3893112540245056, "learning_rate": 5.58715887779105e-06, "loss": 0.4237, "step": 1907 }, { "epoch": 1.5499593826157594, "grad_norm": 0.5097365379333496, "learning_rate": 5.582464279910802e-06, "loss": 0.4385, "step": 1908 }, { "epoch": 1.5507717303005686, "grad_norm": 0.4035624861717224, "learning_rate": 5.577769161425563e-06, "loss": 0.4133, "step": 1909 }, { "epoch": 1.5515840779853778, "grad_norm": 0.44115886092185974, "learning_rate": 5.573073526531818e-06, "loss": 0.4614, "step": 1910 }, { "epoch": 1.552396425670187, "grad_norm": 0.4657628536224365, "learning_rate": 5.568377379426511e-06, "loss": 0.4302, "step": 1911 }, { "epoch": 1.553208773354996, "grad_norm": 0.375266969203949, "learning_rate": 5.563680724307046e-06, "loss": 0.4117, "step": 1912 }, { "epoch": 1.554021121039805, "grad_norm": 0.3807995617389679, "learning_rate": 5.558983565371281e-06, "loss": 0.4397, "step": 1913 }, { "epoch": 1.5548334687246141, "grad_norm": 0.4599345922470093, "learning_rate": 5.554285906817524e-06, "loss": 0.4398, "step": 1914 }, { "epoch": 1.5556458164094233, "grad_norm": 0.36266499757766724, "learning_rate": 5.549587752844529e-06, "loss": 0.4178, "step": 1915 }, { "epoch": 1.5564581640942323, "grad_norm": 0.389030784368515, "learning_rate": 5.5448891076514925e-06, "loss": 0.433, "step": 1916 }, { "epoch": 1.5572705117790413, "grad_norm": 0.45364755392074585, "learning_rate": 5.540189975438053e-06, "loss": 0.418, "step": 1917 }, { "epoch": 1.5580828594638505, "grad_norm": 0.39542555809020996, "learning_rate": 5.53549036040428e-06, "loss": 0.4586, "step": 1918 }, { "epoch": 1.5588952071486597, "grad_norm": 0.3738965690135956, "learning_rate": 5.530790266750675e-06, "loss": 0.4133, "step": 1919 }, { "epoch": 1.5597075548334687, "grad_norm": 0.4232056438922882, "learning_rate": 5.526089698678174e-06, "loss": 0.4634, "step": 1920 }, { "epoch": 1.5605199025182777, "grad_norm": 0.48626428842544556, "learning_rate": 5.521388660388129e-06, "loss": 0.4454, "step": 1921 }, { "epoch": 1.5613322502030869, "grad_norm": 0.3963759243488312, "learning_rate": 5.516687156082317e-06, "loss": 0.4242, "step": 1922 }, { "epoch": 1.562144597887896, "grad_norm": 0.39051321148872375, "learning_rate": 5.511985189962926e-06, "loss": 0.4376, "step": 1923 }, { "epoch": 1.5629569455727053, "grad_norm": 0.37828031182289124, "learning_rate": 5.507282766232565e-06, "loss": 0.4287, "step": 1924 }, { "epoch": 1.5637692932575142, "grad_norm": 0.39153003692626953, "learning_rate": 5.502579889094244e-06, "loss": 0.4443, "step": 1925 }, { "epoch": 1.5645816409423232, "grad_norm": 0.40584757924079895, "learning_rate": 5.497876562751384e-06, "loss": 0.4392, "step": 1926 }, { "epoch": 1.5653939886271324, "grad_norm": 0.4175497889518738, "learning_rate": 5.493172791407803e-06, "loss": 0.4229, "step": 1927 }, { "epoch": 1.5662063363119416, "grad_norm": 0.41563984751701355, "learning_rate": 5.488468579267721e-06, "loss": 0.4468, "step": 1928 }, { "epoch": 1.5670186839967506, "grad_norm": 0.42246225476264954, "learning_rate": 5.483763930535752e-06, "loss": 0.4329, "step": 1929 }, { "epoch": 1.5678310316815596, "grad_norm": 0.4455207884311676, "learning_rate": 5.479058849416894e-06, "loss": 0.4106, "step": 1930 }, { "epoch": 1.5686433793663688, "grad_norm": 0.3935076892375946, "learning_rate": 5.474353340116538e-06, "loss": 0.4161, "step": 1931 }, { "epoch": 1.569455727051178, "grad_norm": 0.44046279788017273, "learning_rate": 5.469647406840456e-06, "loss": 0.4378, "step": 1932 }, { "epoch": 1.570268074735987, "grad_norm": 0.37486255168914795, "learning_rate": 5.464941053794797e-06, "loss": 0.4277, "step": 1933 }, { "epoch": 1.5710804224207962, "grad_norm": 0.40192529559135437, "learning_rate": 5.460234285186087e-06, "loss": 0.4279, "step": 1934 }, { "epoch": 1.5718927701056051, "grad_norm": 0.4159044921398163, "learning_rate": 5.455527105221222e-06, "loss": 0.4591, "step": 1935 }, { "epoch": 1.5727051177904143, "grad_norm": 0.4004330039024353, "learning_rate": 5.450819518107468e-06, "loss": 0.4611, "step": 1936 }, { "epoch": 1.5735174654752235, "grad_norm": 0.40686550736427307, "learning_rate": 5.446111528052455e-06, "loss": 0.4224, "step": 1937 }, { "epoch": 1.5743298131600325, "grad_norm": 0.40855783224105835, "learning_rate": 5.441403139264169e-06, "loss": 0.402, "step": 1938 }, { "epoch": 1.5751421608448415, "grad_norm": 0.40024706721305847, "learning_rate": 5.436694355950955e-06, "loss": 0.4495, "step": 1939 }, { "epoch": 1.5759545085296507, "grad_norm": 0.3824950158596039, "learning_rate": 5.431985182321513e-06, "loss": 0.4313, "step": 1940 }, { "epoch": 1.5767668562144599, "grad_norm": 0.3968172073364258, "learning_rate": 5.427275622584888e-06, "loss": 0.4541, "step": 1941 }, { "epoch": 1.5775792038992689, "grad_norm": 0.35791343450546265, "learning_rate": 5.422565680950472e-06, "loss": 0.4236, "step": 1942 }, { "epoch": 1.5783915515840778, "grad_norm": 0.3950708210468292, "learning_rate": 5.417855361627998e-06, "loss": 0.4295, "step": 1943 }, { "epoch": 1.579203899268887, "grad_norm": 0.37179917097091675, "learning_rate": 5.413144668827538e-06, "loss": 0.4177, "step": 1944 }, { "epoch": 1.5800162469536962, "grad_norm": 0.3955080509185791, "learning_rate": 5.4084336067594945e-06, "loss": 0.4044, "step": 1945 }, { "epoch": 1.5808285946385054, "grad_norm": 0.39659351110458374, "learning_rate": 5.403722179634602e-06, "loss": 0.4258, "step": 1946 }, { "epoch": 1.5816409423233144, "grad_norm": 0.3683428168296814, "learning_rate": 5.399010391663923e-06, "loss": 0.4241, "step": 1947 }, { "epoch": 1.5824532900081234, "grad_norm": 0.3794843256473541, "learning_rate": 5.3942982470588386e-06, "loss": 0.4583, "step": 1948 }, { "epoch": 1.5832656376929326, "grad_norm": 0.3750239908695221, "learning_rate": 5.389585750031053e-06, "loss": 0.4351, "step": 1949 }, { "epoch": 1.5840779853777418, "grad_norm": 0.3974359631538391, "learning_rate": 5.384872904792583e-06, "loss": 0.4473, "step": 1950 }, { "epoch": 1.5848903330625508, "grad_norm": 0.3784196972846985, "learning_rate": 5.380159715555755e-06, "loss": 0.4399, "step": 1951 }, { "epoch": 1.5857026807473598, "grad_norm": 0.3932976722717285, "learning_rate": 5.375446186533206e-06, "loss": 0.4254, "step": 1952 }, { "epoch": 1.586515028432169, "grad_norm": 0.3713891804218292, "learning_rate": 5.370732321937876e-06, "loss": 0.4271, "step": 1953 }, { "epoch": 1.5873273761169782, "grad_norm": 0.4069165885448456, "learning_rate": 5.3660181259830024e-06, "loss": 0.4214, "step": 1954 }, { "epoch": 1.5881397238017871, "grad_norm": 0.3682200610637665, "learning_rate": 5.361303602882123e-06, "loss": 0.4165, "step": 1955 }, { "epoch": 1.5889520714865961, "grad_norm": 0.3878316283226013, "learning_rate": 5.356588756849065e-06, "loss": 0.434, "step": 1956 }, { "epoch": 1.5897644191714053, "grad_norm": 0.3716641962528229, "learning_rate": 5.3518735920979436e-06, "loss": 0.4185, "step": 1957 }, { "epoch": 1.5905767668562145, "grad_norm": 0.41925543546676636, "learning_rate": 5.347158112843163e-06, "loss": 0.4388, "step": 1958 }, { "epoch": 1.5913891145410237, "grad_norm": 0.36900174617767334, "learning_rate": 5.342442323299403e-06, "loss": 0.4064, "step": 1959 }, { "epoch": 1.5922014622258327, "grad_norm": 0.37261396646499634, "learning_rate": 5.3377262276816245e-06, "loss": 0.4232, "step": 1960 }, { "epoch": 1.5930138099106417, "grad_norm": 0.3895726501941681, "learning_rate": 5.3330098302050615e-06, "loss": 0.4382, "step": 1961 }, { "epoch": 1.5938261575954509, "grad_norm": 0.3930111527442932, "learning_rate": 5.328293135085215e-06, "loss": 0.4301, "step": 1962 }, { "epoch": 1.59463850528026, "grad_norm": 0.3927668631076813, "learning_rate": 5.323576146537858e-06, "loss": 0.4032, "step": 1963 }, { "epoch": 1.595450852965069, "grad_norm": 0.4086286425590515, "learning_rate": 5.31885886877902e-06, "loss": 0.4206, "step": 1964 }, { "epoch": 1.596263200649878, "grad_norm": 0.37573882937431335, "learning_rate": 5.314141306024991e-06, "loss": 0.4305, "step": 1965 }, { "epoch": 1.5970755483346872, "grad_norm": 0.39449894428253174, "learning_rate": 5.309423462492314e-06, "loss": 0.4243, "step": 1966 }, { "epoch": 1.5978878960194964, "grad_norm": 0.41781142354011536, "learning_rate": 5.30470534239779e-06, "loss": 0.4323, "step": 1967 }, { "epoch": 1.5987002437043054, "grad_norm": 0.35041487216949463, "learning_rate": 5.299986949958457e-06, "loss": 0.4086, "step": 1968 }, { "epoch": 1.5995125913891144, "grad_norm": 0.3616226017475128, "learning_rate": 5.295268289391603e-06, "loss": 0.4166, "step": 1969 }, { "epoch": 1.6003249390739236, "grad_norm": 0.42611250281333923, "learning_rate": 5.290549364914754e-06, "loss": 0.4338, "step": 1970 }, { "epoch": 1.6011372867587328, "grad_norm": 0.41363629698753357, "learning_rate": 5.285830180745673e-06, "loss": 0.465, "step": 1971 }, { "epoch": 1.601949634443542, "grad_norm": 0.38193294405937195, "learning_rate": 5.281110741102352e-06, "loss": 0.4381, "step": 1972 }, { "epoch": 1.602761982128351, "grad_norm": 0.399794340133667, "learning_rate": 5.276391050203015e-06, "loss": 0.4214, "step": 1973 }, { "epoch": 1.60357432981316, "grad_norm": 0.38528043031692505, "learning_rate": 5.2716711122661065e-06, "loss": 0.4198, "step": 1974 }, { "epoch": 1.6043866774979691, "grad_norm": 0.35136449337005615, "learning_rate": 5.266950931510296e-06, "loss": 0.413, "step": 1975 }, { "epoch": 1.6051990251827783, "grad_norm": 0.3463118076324463, "learning_rate": 5.262230512154469e-06, "loss": 0.4446, "step": 1976 }, { "epoch": 1.6060113728675873, "grad_norm": 0.39250296354293823, "learning_rate": 5.257509858417723e-06, "loss": 0.4194, "step": 1977 }, { "epoch": 1.6068237205523963, "grad_norm": 0.3944198787212372, "learning_rate": 5.252788974519364e-06, "loss": 0.4371, "step": 1978 }, { "epoch": 1.6076360682372055, "grad_norm": 0.3528541922569275, "learning_rate": 5.248067864678906e-06, "loss": 0.441, "step": 1979 }, { "epoch": 1.6084484159220147, "grad_norm": 0.38065677881240845, "learning_rate": 5.243346533116065e-06, "loss": 0.4232, "step": 1980 }, { "epoch": 1.6092607636068237, "grad_norm": 0.3996758460998535, "learning_rate": 5.238624984050754e-06, "loss": 0.4367, "step": 1981 }, { "epoch": 1.6100731112916327, "grad_norm": 0.36804139614105225, "learning_rate": 5.233903221703079e-06, "loss": 0.4225, "step": 1982 }, { "epoch": 1.6108854589764419, "grad_norm": 0.3769538700580597, "learning_rate": 5.229181250293341e-06, "loss": 0.4417, "step": 1983 }, { "epoch": 1.611697806661251, "grad_norm": 0.35993272066116333, "learning_rate": 5.224459074042023e-06, "loss": 0.4131, "step": 1984 }, { "epoch": 1.6125101543460603, "grad_norm": 0.4558311402797699, "learning_rate": 5.219736697169795e-06, "loss": 0.4477, "step": 1985 }, { "epoch": 1.6133225020308692, "grad_norm": 0.3750121295452118, "learning_rate": 5.215014123897504e-06, "loss": 0.4272, "step": 1986 }, { "epoch": 1.6141348497156782, "grad_norm": 0.3902049958705902, "learning_rate": 5.210291358446173e-06, "loss": 0.4222, "step": 1987 }, { "epoch": 1.6149471974004874, "grad_norm": 0.3493915796279907, "learning_rate": 5.205568405036996e-06, "loss": 0.4049, "step": 1988 }, { "epoch": 1.6157595450852966, "grad_norm": 0.37700775265693665, "learning_rate": 5.200845267891338e-06, "loss": 0.4304, "step": 1989 }, { "epoch": 1.6165718927701056, "grad_norm": 0.3725496232509613, "learning_rate": 5.196121951230726e-06, "loss": 0.4357, "step": 1990 }, { "epoch": 1.6173842404549146, "grad_norm": 0.3657264709472656, "learning_rate": 5.1913984592768455e-06, "loss": 0.408, "step": 1991 }, { "epoch": 1.6181965881397238, "grad_norm": 0.32991933822631836, "learning_rate": 5.186674796251543e-06, "loss": 0.4263, "step": 1992 }, { "epoch": 1.619008935824533, "grad_norm": 0.3903040587902069, "learning_rate": 5.181950966376814e-06, "loss": 0.4011, "step": 1993 }, { "epoch": 1.619821283509342, "grad_norm": 0.37245723605155945, "learning_rate": 5.177226973874805e-06, "loss": 0.4194, "step": 1994 }, { "epoch": 1.620633631194151, "grad_norm": 0.3556678593158722, "learning_rate": 5.172502822967808e-06, "loss": 0.4224, "step": 1995 }, { "epoch": 1.6214459788789601, "grad_norm": 0.3469019830226898, "learning_rate": 5.167778517878258e-06, "loss": 0.4026, "step": 1996 }, { "epoch": 1.6222583265637693, "grad_norm": 0.35179513692855835, "learning_rate": 5.163054062828724e-06, "loss": 0.4198, "step": 1997 }, { "epoch": 1.6230706742485785, "grad_norm": 0.3836049437522888, "learning_rate": 5.158329462041911e-06, "loss": 0.4429, "step": 1998 }, { "epoch": 1.6238830219333875, "grad_norm": 0.3886573910713196, "learning_rate": 5.153604719740655e-06, "loss": 0.4178, "step": 1999 }, { "epoch": 1.6246953696181965, "grad_norm": 0.3617821931838989, "learning_rate": 5.148879840147918e-06, "loss": 0.4486, "step": 2000 }, { "epoch": 1.6255077173030057, "grad_norm": 0.3990154266357422, "learning_rate": 5.1441548274867834e-06, "loss": 0.4385, "step": 2001 }, { "epoch": 1.626320064987815, "grad_norm": 0.32796114683151245, "learning_rate": 5.139429685980457e-06, "loss": 0.4132, "step": 2002 }, { "epoch": 1.6271324126726239, "grad_norm": 0.3868862986564636, "learning_rate": 5.134704419852253e-06, "loss": 0.4428, "step": 2003 }, { "epoch": 1.6279447603574329, "grad_norm": 0.384759783744812, "learning_rate": 5.129979033325608e-06, "loss": 0.4261, "step": 2004 }, { "epoch": 1.628757108042242, "grad_norm": 0.4078396260738373, "learning_rate": 5.125253530624054e-06, "loss": 0.4467, "step": 2005 }, { "epoch": 1.6295694557270513, "grad_norm": 0.4630512595176697, "learning_rate": 5.120527915971235e-06, "loss": 0.4419, "step": 2006 }, { "epoch": 1.6303818034118602, "grad_norm": 0.46145907044410706, "learning_rate": 5.115802193590893e-06, "loss": 0.4584, "step": 2007 }, { "epoch": 1.6311941510966692, "grad_norm": 0.3713025748729706, "learning_rate": 5.111076367706864e-06, "loss": 0.4409, "step": 2008 }, { "epoch": 1.6320064987814784, "grad_norm": 0.38065898418426514, "learning_rate": 5.10635044254308e-06, "loss": 0.414, "step": 2009 }, { "epoch": 1.6328188464662876, "grad_norm": 0.45707806944847107, "learning_rate": 5.1016244223235576e-06, "loss": 0.4297, "step": 2010 }, { "epoch": 1.6336311941510968, "grad_norm": 0.4434759318828583, "learning_rate": 5.096898311272404e-06, "loss": 0.4277, "step": 2011 }, { "epoch": 1.6344435418359058, "grad_norm": 0.4126327931880951, "learning_rate": 5.0921721136138025e-06, "loss": 0.457, "step": 2012 }, { "epoch": 1.6352558895207148, "grad_norm": 0.35178759694099426, "learning_rate": 5.087445833572017e-06, "loss": 0.4211, "step": 2013 }, { "epoch": 1.636068237205524, "grad_norm": 0.3697182238101959, "learning_rate": 5.082719475371382e-06, "loss": 0.4195, "step": 2014 }, { "epoch": 1.6368805848903332, "grad_norm": 0.35351112484931946, "learning_rate": 5.077993043236306e-06, "loss": 0.4135, "step": 2015 }, { "epoch": 1.6376929325751421, "grad_norm": 0.37736907601356506, "learning_rate": 5.073266541391259e-06, "loss": 0.3914, "step": 2016 }, { "epoch": 1.6385052802599511, "grad_norm": 0.4257614016532898, "learning_rate": 5.068539974060776e-06, "loss": 0.4659, "step": 2017 }, { "epoch": 1.6393176279447603, "grad_norm": 0.39920803904533386, "learning_rate": 5.06381334546945e-06, "loss": 0.4243, "step": 2018 }, { "epoch": 1.6401299756295695, "grad_norm": 0.39135390520095825, "learning_rate": 5.05908665984193e-06, "loss": 0.426, "step": 2019 }, { "epoch": 1.6409423233143785, "grad_norm": 0.38510334491729736, "learning_rate": 5.054359921402914e-06, "loss": 0.4025, "step": 2020 }, { "epoch": 1.6417546709991877, "grad_norm": 0.4193767011165619, "learning_rate": 5.049633134377147e-06, "loss": 0.4397, "step": 2021 }, { "epoch": 1.6425670186839967, "grad_norm": 0.41560494899749756, "learning_rate": 5.04490630298942e-06, "loss": 0.4144, "step": 2022 }, { "epoch": 1.6433793663688059, "grad_norm": 0.38020768761634827, "learning_rate": 5.04017943146456e-06, "loss": 0.4192, "step": 2023 }, { "epoch": 1.644191714053615, "grad_norm": 0.3637101948261261, "learning_rate": 5.035452524027436e-06, "loss": 0.4103, "step": 2024 }, { "epoch": 1.645004061738424, "grad_norm": 0.4004332423210144, "learning_rate": 5.03072558490294e-06, "loss": 0.4252, "step": 2025 }, { "epoch": 1.645816409423233, "grad_norm": 0.40274789929389954, "learning_rate": 5.0259986183160006e-06, "loss": 0.4487, "step": 2026 }, { "epoch": 1.6466287571080422, "grad_norm": 0.3644509017467499, "learning_rate": 5.021271628491566e-06, "loss": 0.4099, "step": 2027 }, { "epoch": 1.6474411047928514, "grad_norm": 0.3837379515171051, "learning_rate": 5.016544619654609e-06, "loss": 0.416, "step": 2028 }, { "epoch": 1.6482534524776604, "grad_norm": 0.4164125621318817, "learning_rate": 5.011817596030115e-06, "loss": 0.4328, "step": 2029 }, { "epoch": 1.6490658001624694, "grad_norm": 0.41627219319343567, "learning_rate": 5.0070905618430865e-06, "loss": 0.4441, "step": 2030 }, { "epoch": 1.6498781478472786, "grad_norm": 0.4131058156490326, "learning_rate": 5.0023635213185325e-06, "loss": 0.442, "step": 2031 }, { "epoch": 1.6506904955320878, "grad_norm": 0.3556484580039978, "learning_rate": 4.997636478681468e-06, "loss": 0.4457, "step": 2032 }, { "epoch": 1.6515028432168968, "grad_norm": 0.41273221373558044, "learning_rate": 4.992909438156915e-06, "loss": 0.4248, "step": 2033 }, { "epoch": 1.652315190901706, "grad_norm": 0.4287571609020233, "learning_rate": 4.988182403969886e-06, "loss": 0.4182, "step": 2034 }, { "epoch": 1.653127538586515, "grad_norm": 0.37732234597206116, "learning_rate": 4.983455380345392e-06, "loss": 0.4473, "step": 2035 }, { "epoch": 1.6539398862713242, "grad_norm": 0.36161917448043823, "learning_rate": 4.978728371508433e-06, "loss": 0.3954, "step": 2036 }, { "epoch": 1.6547522339561334, "grad_norm": 0.3491344749927521, "learning_rate": 4.974001381684e-06, "loss": 0.4276, "step": 2037 }, { "epoch": 1.6555645816409423, "grad_norm": 0.3846942186355591, "learning_rate": 4.96927441509706e-06, "loss": 0.428, "step": 2038 }, { "epoch": 1.6563769293257513, "grad_norm": 0.36397141218185425, "learning_rate": 4.964547475972565e-06, "loss": 0.4257, "step": 2039 }, { "epoch": 1.6571892770105605, "grad_norm": 0.3642084002494812, "learning_rate": 4.9598205685354404e-06, "loss": 0.4405, "step": 2040 }, { "epoch": 1.6580016246953697, "grad_norm": 0.40362438559532166, "learning_rate": 4.955093697010581e-06, "loss": 0.4449, "step": 2041 }, { "epoch": 1.6588139723801787, "grad_norm": 0.3561609387397766, "learning_rate": 4.950366865622854e-06, "loss": 0.4174, "step": 2042 }, { "epoch": 1.6596263200649877, "grad_norm": 0.40454813838005066, "learning_rate": 4.9456400785970875e-06, "loss": 0.4502, "step": 2043 }, { "epoch": 1.6604386677497969, "grad_norm": 0.36915281414985657, "learning_rate": 4.940913340158072e-06, "loss": 0.4487, "step": 2044 }, { "epoch": 1.661251015434606, "grad_norm": 0.39127498865127563, "learning_rate": 4.936186654530551e-06, "loss": 0.4436, "step": 2045 }, { "epoch": 1.6620633631194153, "grad_norm": 0.3916163444519043, "learning_rate": 4.931460025939226e-06, "loss": 0.423, "step": 2046 }, { "epoch": 1.6628757108042242, "grad_norm": 0.4523701071739197, "learning_rate": 4.926733458608744e-06, "loss": 0.4542, "step": 2047 }, { "epoch": 1.6636880584890332, "grad_norm": 0.351146399974823, "learning_rate": 4.922006956763697e-06, "loss": 0.4096, "step": 2048 }, { "epoch": 1.6645004061738424, "grad_norm": 0.3589871823787689, "learning_rate": 4.917280524628619e-06, "loss": 0.4322, "step": 2049 }, { "epoch": 1.6653127538586516, "grad_norm": 0.40618589520454407, "learning_rate": 4.912554166427985e-06, "loss": 0.4253, "step": 2050 }, { "epoch": 1.6661251015434606, "grad_norm": 0.43628552556037903, "learning_rate": 4.907827886386199e-06, "loss": 0.4292, "step": 2051 }, { "epoch": 1.6669374492282696, "grad_norm": 0.3647879362106323, "learning_rate": 4.903101688727598e-06, "loss": 0.3986, "step": 2052 }, { "epoch": 1.6677497969130788, "grad_norm": 0.38704797625541687, "learning_rate": 4.898375577676444e-06, "loss": 0.4523, "step": 2053 }, { "epoch": 1.668562144597888, "grad_norm": 0.40438055992126465, "learning_rate": 4.893649557456922e-06, "loss": 0.395, "step": 2054 }, { "epoch": 1.669374492282697, "grad_norm": 0.40994781255722046, "learning_rate": 4.888923632293137e-06, "loss": 0.3948, "step": 2055 }, { "epoch": 1.670186839967506, "grad_norm": 0.3617405891418457, "learning_rate": 4.884197806409109e-06, "loss": 0.406, "step": 2056 }, { "epoch": 1.6709991876523151, "grad_norm": 0.4087444841861725, "learning_rate": 4.8794720840287665e-06, "loss": 0.4302, "step": 2057 }, { "epoch": 1.6718115353371243, "grad_norm": 0.41078102588653564, "learning_rate": 4.874746469375947e-06, "loss": 0.4497, "step": 2058 }, { "epoch": 1.6726238830219335, "grad_norm": 0.39795446395874023, "learning_rate": 4.870020966674394e-06, "loss": 0.4232, "step": 2059 }, { "epoch": 1.6734362307067425, "grad_norm": 0.37849900126457214, "learning_rate": 4.865295580147747e-06, "loss": 0.4257, "step": 2060 }, { "epoch": 1.6742485783915515, "grad_norm": 0.3472154140472412, "learning_rate": 4.860570314019546e-06, "loss": 0.4168, "step": 2061 }, { "epoch": 1.6750609260763607, "grad_norm": 0.406711608171463, "learning_rate": 4.855845172513219e-06, "loss": 0.4211, "step": 2062 }, { "epoch": 1.67587327376117, "grad_norm": 0.3861650228500366, "learning_rate": 4.851120159852085e-06, "loss": 0.4087, "step": 2063 }, { "epoch": 1.6766856214459789, "grad_norm": 0.40679991245269775, "learning_rate": 4.846395280259348e-06, "loss": 0.4061, "step": 2064 }, { "epoch": 1.6774979691307879, "grad_norm": 0.39682650566101074, "learning_rate": 4.841670537958092e-06, "loss": 0.4258, "step": 2065 }, { "epoch": 1.678310316815597, "grad_norm": 0.3945506811141968, "learning_rate": 4.836945937171279e-06, "loss": 0.4505, "step": 2066 }, { "epoch": 1.6791226645004063, "grad_norm": 0.42214587330818176, "learning_rate": 4.832221482121745e-06, "loss": 0.4319, "step": 2067 }, { "epoch": 1.6799350121852152, "grad_norm": 0.42078980803489685, "learning_rate": 4.827497177032193e-06, "loss": 0.4397, "step": 2068 }, { "epoch": 1.6807473598700242, "grad_norm": 0.41046178340911865, "learning_rate": 4.822773026125197e-06, "loss": 0.4096, "step": 2069 }, { "epoch": 1.6815597075548334, "grad_norm": 0.3688640296459198, "learning_rate": 4.8180490336231865e-06, "loss": 0.4432, "step": 2070 }, { "epoch": 1.6823720552396426, "grad_norm": 0.4309726357460022, "learning_rate": 4.813325203748458e-06, "loss": 0.4085, "step": 2071 }, { "epoch": 1.6831844029244518, "grad_norm": 0.4271986484527588, "learning_rate": 4.8086015407231545e-06, "loss": 0.4354, "step": 2072 }, { "epoch": 1.6839967506092608, "grad_norm": 0.379728764295578, "learning_rate": 4.803878048769275e-06, "loss": 0.453, "step": 2073 }, { "epoch": 1.6848090982940698, "grad_norm": 0.3492249846458435, "learning_rate": 4.799154732108662e-06, "loss": 0.4265, "step": 2074 }, { "epoch": 1.685621445978879, "grad_norm": 0.39642855525016785, "learning_rate": 4.794431594963004e-06, "loss": 0.4071, "step": 2075 }, { "epoch": 1.6864337936636882, "grad_norm": 0.40923959016799927, "learning_rate": 4.789708641553828e-06, "loss": 0.4393, "step": 2076 }, { "epoch": 1.6872461413484972, "grad_norm": 0.39933842420578003, "learning_rate": 4.784985876102498e-06, "loss": 0.429, "step": 2077 }, { "epoch": 1.6880584890333061, "grad_norm": 0.3934655487537384, "learning_rate": 4.7802633028302065e-06, "loss": 0.4222, "step": 2078 }, { "epoch": 1.6888708367181153, "grad_norm": 0.38843926787376404, "learning_rate": 4.7755409259579786e-06, "loss": 0.3976, "step": 2079 }, { "epoch": 1.6896831844029245, "grad_norm": 0.3718976378440857, "learning_rate": 4.7708187497066595e-06, "loss": 0.417, "step": 2080 }, { "epoch": 1.6904955320877335, "grad_norm": 0.40562766790390015, "learning_rate": 4.766096778296922e-06, "loss": 0.4517, "step": 2081 }, { "epoch": 1.6913078797725425, "grad_norm": 0.43238475918769836, "learning_rate": 4.761375015949248e-06, "loss": 0.4228, "step": 2082 }, { "epoch": 1.6921202274573517, "grad_norm": 0.4059799015522003, "learning_rate": 4.756653466883937e-06, "loss": 0.4137, "step": 2083 }, { "epoch": 1.692932575142161, "grad_norm": 0.3899869918823242, "learning_rate": 4.751932135321095e-06, "loss": 0.3935, "step": 2084 }, { "epoch": 1.69374492282697, "grad_norm": 0.3797428011894226, "learning_rate": 4.7472110254806375e-06, "loss": 0.4424, "step": 2085 }, { "epoch": 1.694557270511779, "grad_norm": 0.36175915598869324, "learning_rate": 4.742490141582279e-06, "loss": 0.4381, "step": 2086 }, { "epoch": 1.695369618196588, "grad_norm": 0.3802557587623596, "learning_rate": 4.737769487845532e-06, "loss": 0.4707, "step": 2087 }, { "epoch": 1.6961819658813972, "grad_norm": 0.3841353952884674, "learning_rate": 4.733049068489704e-06, "loss": 0.4169, "step": 2088 }, { "epoch": 1.6969943135662064, "grad_norm": 0.37482720613479614, "learning_rate": 4.728328887733894e-06, "loss": 0.4392, "step": 2089 }, { "epoch": 1.6978066612510154, "grad_norm": 0.3448185920715332, "learning_rate": 4.723608949796987e-06, "loss": 0.4269, "step": 2090 }, { "epoch": 1.6986190089358244, "grad_norm": 0.34799543023109436, "learning_rate": 4.7188892588976496e-06, "loss": 0.4224, "step": 2091 }, { "epoch": 1.6994313566206336, "grad_norm": 0.3625708520412445, "learning_rate": 4.714169819254328e-06, "loss": 0.4686, "step": 2092 }, { "epoch": 1.7002437043054428, "grad_norm": 0.4206971526145935, "learning_rate": 4.709450635085246e-06, "loss": 0.4184, "step": 2093 }, { "epoch": 1.7010560519902518, "grad_norm": 0.40093880891799927, "learning_rate": 4.704731710608398e-06, "loss": 0.4341, "step": 2094 }, { "epoch": 1.7018683996750608, "grad_norm": 0.3436519503593445, "learning_rate": 4.700013050041544e-06, "loss": 0.4577, "step": 2095 }, { "epoch": 1.70268074735987, "grad_norm": 0.39040952920913696, "learning_rate": 4.695294657602212e-06, "loss": 0.4071, "step": 2096 }, { "epoch": 1.7034930950446792, "grad_norm": 0.3755565583705902, "learning_rate": 4.6905765375076865e-06, "loss": 0.4303, "step": 2097 }, { "epoch": 1.7043054427294884, "grad_norm": 0.41029202938079834, "learning_rate": 4.685858693975011e-06, "loss": 0.4376, "step": 2098 }, { "epoch": 1.7051177904142973, "grad_norm": 0.39143845438957214, "learning_rate": 4.681141131220982e-06, "loss": 0.4425, "step": 2099 }, { "epoch": 1.7059301380991063, "grad_norm": 0.3901256322860718, "learning_rate": 4.6764238534621436e-06, "loss": 0.4568, "step": 2100 }, { "epoch": 1.7067424857839155, "grad_norm": 0.4064679741859436, "learning_rate": 4.671706864914786e-06, "loss": 0.448, "step": 2101 }, { "epoch": 1.7075548334687247, "grad_norm": 0.35818979144096375, "learning_rate": 4.666990169794942e-06, "loss": 0.3967, "step": 2102 }, { "epoch": 1.7083671811535337, "grad_norm": 0.3777344226837158, "learning_rate": 4.662273772318378e-06, "loss": 0.4271, "step": 2103 }, { "epoch": 1.7091795288383427, "grad_norm": 0.39230087399482727, "learning_rate": 4.6575576767006e-06, "loss": 0.4291, "step": 2104 }, { "epoch": 1.7099918765231519, "grad_norm": 0.3484886884689331, "learning_rate": 4.652841887156841e-06, "loss": 0.4109, "step": 2105 }, { "epoch": 1.710804224207961, "grad_norm": 0.4115782082080841, "learning_rate": 4.648126407902058e-06, "loss": 0.4381, "step": 2106 }, { "epoch": 1.71161657189277, "grad_norm": 0.35441854596138, "learning_rate": 4.6434112431509376e-06, "loss": 0.4139, "step": 2107 }, { "epoch": 1.7124289195775793, "grad_norm": 0.33346840739250183, "learning_rate": 4.638696397117877e-06, "loss": 0.4586, "step": 2108 }, { "epoch": 1.7132412672623882, "grad_norm": 0.35904285311698914, "learning_rate": 4.6339818740169975e-06, "loss": 0.423, "step": 2109 }, { "epoch": 1.7140536149471974, "grad_norm": 0.3663124144077301, "learning_rate": 4.6292676780621245e-06, "loss": 0.4429, "step": 2110 }, { "epoch": 1.7148659626320066, "grad_norm": 0.39699602127075195, "learning_rate": 4.624553813466794e-06, "loss": 0.4262, "step": 2111 }, { "epoch": 1.7156783103168156, "grad_norm": 0.3997346758842468, "learning_rate": 4.619840284444245e-06, "loss": 0.4825, "step": 2112 }, { "epoch": 1.7164906580016246, "grad_norm": 0.38035809993743896, "learning_rate": 4.615127095207418e-06, "loss": 0.434, "step": 2113 }, { "epoch": 1.7173030056864338, "grad_norm": 0.4057188928127289, "learning_rate": 4.610414249968947e-06, "loss": 0.4182, "step": 2114 }, { "epoch": 1.718115353371243, "grad_norm": 0.4009607136249542, "learning_rate": 4.605701752941161e-06, "loss": 0.4508, "step": 2115 }, { "epoch": 1.718927701056052, "grad_norm": 0.3667864203453064, "learning_rate": 4.600989608336078e-06, "loss": 0.4107, "step": 2116 }, { "epoch": 1.719740048740861, "grad_norm": 0.3788108825683594, "learning_rate": 4.596277820365398e-06, "loss": 0.4254, "step": 2117 }, { "epoch": 1.7205523964256701, "grad_norm": 0.3890576958656311, "learning_rate": 4.591566393240507e-06, "loss": 0.4221, "step": 2118 }, { "epoch": 1.7213647441104794, "grad_norm": 0.38575151562690735, "learning_rate": 4.5868553311724635e-06, "loss": 0.4554, "step": 2119 }, { "epoch": 1.7221770917952883, "grad_norm": 0.4224950075149536, "learning_rate": 4.5821446383720025e-06, "loss": 0.4357, "step": 2120 }, { "epoch": 1.7229894394800975, "grad_norm": 0.38560542464256287, "learning_rate": 4.577434319049529e-06, "loss": 0.4403, "step": 2121 }, { "epoch": 1.7238017871649065, "grad_norm": 0.35568732023239136, "learning_rate": 4.572724377415114e-06, "loss": 0.4141, "step": 2122 }, { "epoch": 1.7246141348497157, "grad_norm": 0.3913602828979492, "learning_rate": 4.568014817678488e-06, "loss": 0.4452, "step": 2123 }, { "epoch": 1.725426482534525, "grad_norm": 0.38654786348342896, "learning_rate": 4.5633056440490455e-06, "loss": 0.4417, "step": 2124 }, { "epoch": 1.7262388302193339, "grad_norm": 0.3692072927951813, "learning_rate": 4.558596860735833e-06, "loss": 0.4616, "step": 2125 }, { "epoch": 1.7270511779041429, "grad_norm": 0.34955477714538574, "learning_rate": 4.553888471947546e-06, "loss": 0.3979, "step": 2126 }, { "epoch": 1.727863525588952, "grad_norm": 0.35733678936958313, "learning_rate": 4.549180481892532e-06, "loss": 0.4247, "step": 2127 }, { "epoch": 1.7286758732737613, "grad_norm": 0.3874279856681824, "learning_rate": 4.544472894778779e-06, "loss": 0.4195, "step": 2128 }, { "epoch": 1.7294882209585702, "grad_norm": 0.36150622367858887, "learning_rate": 4.539765714813915e-06, "loss": 0.4556, "step": 2129 }, { "epoch": 1.7303005686433792, "grad_norm": 0.3894542455673218, "learning_rate": 4.535058946205205e-06, "loss": 0.4194, "step": 2130 }, { "epoch": 1.7311129163281884, "grad_norm": 0.3600991368293762, "learning_rate": 4.530352593159546e-06, "loss": 0.3991, "step": 2131 }, { "epoch": 1.7319252640129976, "grad_norm": 0.41162586212158203, "learning_rate": 4.525646659883463e-06, "loss": 0.4717, "step": 2132 }, { "epoch": 1.7327376116978068, "grad_norm": 0.37818384170532227, "learning_rate": 4.5209411505831066e-06, "loss": 0.4116, "step": 2133 }, { "epoch": 1.7335499593826158, "grad_norm": 0.3957454562187195, "learning_rate": 4.51623606946425e-06, "loss": 0.4363, "step": 2134 }, { "epoch": 1.7343623070674248, "grad_norm": 0.39341485500335693, "learning_rate": 4.5115314207322794e-06, "loss": 0.4272, "step": 2135 }, { "epoch": 1.735174654752234, "grad_norm": 0.3467269241809845, "learning_rate": 4.506827208592198e-06, "loss": 0.4256, "step": 2136 }, { "epoch": 1.7359870024370432, "grad_norm": 0.37295210361480713, "learning_rate": 4.502123437248619e-06, "loss": 0.4293, "step": 2137 }, { "epoch": 1.7367993501218522, "grad_norm": 0.3725828230381012, "learning_rate": 4.497420110905758e-06, "loss": 0.4436, "step": 2138 }, { "epoch": 1.7376116978066611, "grad_norm": 0.4357910752296448, "learning_rate": 4.492717233767438e-06, "loss": 0.4299, "step": 2139 }, { "epoch": 1.7384240454914703, "grad_norm": 0.3973354697227478, "learning_rate": 4.4880148100370765e-06, "loss": 0.4685, "step": 2140 }, { "epoch": 1.7392363931762795, "grad_norm": 0.3896978497505188, "learning_rate": 4.483312843917685e-06, "loss": 0.4319, "step": 2141 }, { "epoch": 1.7400487408610885, "grad_norm": 0.38949155807495117, "learning_rate": 4.4786113396118715e-06, "loss": 0.414, "step": 2142 }, { "epoch": 1.7408610885458975, "grad_norm": 0.3961026966571808, "learning_rate": 4.473910301321827e-06, "loss": 0.4244, "step": 2143 }, { "epoch": 1.7416734362307067, "grad_norm": 0.374240517616272, "learning_rate": 4.469209733249326e-06, "loss": 0.4182, "step": 2144 }, { "epoch": 1.742485783915516, "grad_norm": 0.3563194274902344, "learning_rate": 4.4645096395957235e-06, "loss": 0.4197, "step": 2145 }, { "epoch": 1.743298131600325, "grad_norm": 0.3763846755027771, "learning_rate": 4.4598100245619505e-06, "loss": 0.4177, "step": 2146 }, { "epoch": 1.744110479285134, "grad_norm": 0.3694654107093811, "learning_rate": 4.455110892348508e-06, "loss": 0.4247, "step": 2147 }, { "epoch": 1.744922826969943, "grad_norm": 0.3405802547931671, "learning_rate": 4.450412247155473e-06, "loss": 0.4262, "step": 2148 }, { "epoch": 1.7457351746547523, "grad_norm": 0.3820209205150604, "learning_rate": 4.445714093182477e-06, "loss": 0.4314, "step": 2149 }, { "epoch": 1.7465475223395615, "grad_norm": 0.3941916823387146, "learning_rate": 4.441016434628719e-06, "loss": 0.4332, "step": 2150 }, { "epoch": 1.7473598700243704, "grad_norm": 0.3801328241825104, "learning_rate": 4.436319275692954e-06, "loss": 0.4456, "step": 2151 }, { "epoch": 1.7481722177091794, "grad_norm": 0.37573304772377014, "learning_rate": 4.43162262057349e-06, "loss": 0.4254, "step": 2152 }, { "epoch": 1.7489845653939886, "grad_norm": 0.37635961174964905, "learning_rate": 4.426926473468183e-06, "loss": 0.4268, "step": 2153 }, { "epoch": 1.7497969130787978, "grad_norm": 0.38333138823509216, "learning_rate": 4.4222308385744375e-06, "loss": 0.405, "step": 2154 }, { "epoch": 1.7506092607636068, "grad_norm": 0.3774360716342926, "learning_rate": 4.417535720089199e-06, "loss": 0.4086, "step": 2155 }, { "epoch": 1.7514216084484158, "grad_norm": 0.3827272653579712, "learning_rate": 4.412841122208951e-06, "loss": 0.4288, "step": 2156 }, { "epoch": 1.752233956133225, "grad_norm": 0.43773889541625977, "learning_rate": 4.408147049129715e-06, "loss": 0.4183, "step": 2157 }, { "epoch": 1.7530463038180342, "grad_norm": 0.42987060546875, "learning_rate": 4.403453505047037e-06, "loss": 0.4286, "step": 2158 }, { "epoch": 1.7538586515028434, "grad_norm": 0.4172975420951843, "learning_rate": 4.3987604941559966e-06, "loss": 0.4174, "step": 2159 }, { "epoch": 1.7546709991876523, "grad_norm": 0.36855989694595337, "learning_rate": 4.3940680206511935e-06, "loss": 0.439, "step": 2160 }, { "epoch": 1.7554833468724613, "grad_norm": 0.39440906047821045, "learning_rate": 4.389376088726747e-06, "loss": 0.4312, "step": 2161 }, { "epoch": 1.7562956945572705, "grad_norm": 0.4327711760997772, "learning_rate": 4.3846847025762944e-06, "loss": 0.4406, "step": 2162 }, { "epoch": 1.7571080422420797, "grad_norm": 0.45118004083633423, "learning_rate": 4.3799938663929835e-06, "loss": 0.4243, "step": 2163 }, { "epoch": 1.7579203899268887, "grad_norm": 0.36365261673927307, "learning_rate": 4.37530358436947e-06, "loss": 0.429, "step": 2164 }, { "epoch": 1.7587327376116977, "grad_norm": 0.37820473313331604, "learning_rate": 4.370613860697916e-06, "loss": 0.4339, "step": 2165 }, { "epoch": 1.7595450852965069, "grad_norm": 0.4791058897972107, "learning_rate": 4.3659246995699845e-06, "loss": 0.4225, "step": 2166 }, { "epoch": 1.760357432981316, "grad_norm": 0.3750168979167938, "learning_rate": 4.361236105176832e-06, "loss": 0.431, "step": 2167 }, { "epoch": 1.761169780666125, "grad_norm": 0.334994912147522, "learning_rate": 4.356548081709116e-06, "loss": 0.4143, "step": 2168 }, { "epoch": 1.761982128350934, "grad_norm": 0.3850284516811371, "learning_rate": 4.351860633356976e-06, "loss": 0.4154, "step": 2169 }, { "epoch": 1.7627944760357432, "grad_norm": 0.4072641134262085, "learning_rate": 4.347173764310041e-06, "loss": 0.4139, "step": 2170 }, { "epoch": 1.7636068237205524, "grad_norm": 0.39386770129203796, "learning_rate": 4.342487478757422e-06, "loss": 0.4262, "step": 2171 }, { "epoch": 1.7644191714053616, "grad_norm": 0.3814915418624878, "learning_rate": 4.3378017808877095e-06, "loss": 0.3985, "step": 2172 }, { "epoch": 1.7652315190901706, "grad_norm": 0.4008024036884308, "learning_rate": 4.333116674888966e-06, "loss": 0.4267, "step": 2173 }, { "epoch": 1.7660438667749796, "grad_norm": 0.3695429861545563, "learning_rate": 4.328432164948727e-06, "loss": 0.4057, "step": 2174 }, { "epoch": 1.7668562144597888, "grad_norm": 0.39434000849723816, "learning_rate": 4.323748255253995e-06, "loss": 0.4586, "step": 2175 }, { "epoch": 1.767668562144598, "grad_norm": 0.36962762475013733, "learning_rate": 4.319064949991236e-06, "loss": 0.4067, "step": 2176 }, { "epoch": 1.768480909829407, "grad_norm": 0.4006408154964447, "learning_rate": 4.314382253346374e-06, "loss": 0.445, "step": 2177 }, { "epoch": 1.769293257514216, "grad_norm": 0.35219964385032654, "learning_rate": 4.309700169504792e-06, "loss": 0.4638, "step": 2178 }, { "epoch": 1.7701056051990252, "grad_norm": 0.3404271602630615, "learning_rate": 4.305018702651323e-06, "loss": 0.3907, "step": 2179 }, { "epoch": 1.7709179528838344, "grad_norm": 0.4093247950077057, "learning_rate": 4.300337856970251e-06, "loss": 0.4374, "step": 2180 }, { "epoch": 1.7717303005686433, "grad_norm": 0.34633609652519226, "learning_rate": 4.295657636645302e-06, "loss": 0.4041, "step": 2181 }, { "epoch": 1.7725426482534523, "grad_norm": 0.32875415682792664, "learning_rate": 4.290978045859643e-06, "loss": 0.4246, "step": 2182 }, { "epoch": 1.7733549959382615, "grad_norm": 0.34525975584983826, "learning_rate": 4.2862990887958815e-06, "loss": 0.4145, "step": 2183 }, { "epoch": 1.7741673436230707, "grad_norm": 0.3990650475025177, "learning_rate": 4.281620769636055e-06, "loss": 0.4343, "step": 2184 }, { "epoch": 1.77497969130788, "grad_norm": 0.4007239043712616, "learning_rate": 4.276943092561632e-06, "loss": 0.4319, "step": 2185 }, { "epoch": 1.775792038992689, "grad_norm": 0.39259597659111023, "learning_rate": 4.2722660617535105e-06, "loss": 0.4282, "step": 2186 }, { "epoch": 1.7766043866774979, "grad_norm": 0.3571251928806305, "learning_rate": 4.267589681392006e-06, "loss": 0.428, "step": 2187 }, { "epoch": 1.777416734362307, "grad_norm": 0.4537633955478668, "learning_rate": 4.262913955656856e-06, "loss": 0.4531, "step": 2188 }, { "epoch": 1.7782290820471163, "grad_norm": 0.38622069358825684, "learning_rate": 4.258238888727209e-06, "loss": 0.4273, "step": 2189 }, { "epoch": 1.7790414297319253, "grad_norm": 0.39058756828308105, "learning_rate": 4.253564484781629e-06, "loss": 0.4285, "step": 2190 }, { "epoch": 1.7798537774167342, "grad_norm": 0.3843514919281006, "learning_rate": 4.248890747998085e-06, "loss": 0.4081, "step": 2191 }, { "epoch": 1.7806661251015434, "grad_norm": 0.41049879789352417, "learning_rate": 4.24421768255395e-06, "loss": 0.4246, "step": 2192 }, { "epoch": 1.7814784727863526, "grad_norm": 0.4386058747768402, "learning_rate": 4.2395452926259965e-06, "loss": 0.4112, "step": 2193 }, { "epoch": 1.7822908204711616, "grad_norm": 0.3687078058719635, "learning_rate": 4.234873582390395e-06, "loss": 0.4226, "step": 2194 }, { "epoch": 1.7831031681559708, "grad_norm": 0.3936542868614197, "learning_rate": 4.230202556022708e-06, "loss": 0.4079, "step": 2195 }, { "epoch": 1.7839155158407798, "grad_norm": 0.3892045021057129, "learning_rate": 4.225532217697886e-06, "loss": 0.4513, "step": 2196 }, { "epoch": 1.784727863525589, "grad_norm": 0.37418490648269653, "learning_rate": 4.220862571590264e-06, "loss": 0.4314, "step": 2197 }, { "epoch": 1.7855402112103982, "grad_norm": 0.3720824420452118, "learning_rate": 4.216193621873559e-06, "loss": 0.4031, "step": 2198 }, { "epoch": 1.7863525588952072, "grad_norm": 0.3943701982498169, "learning_rate": 4.211525372720865e-06, "loss": 0.4294, "step": 2199 }, { "epoch": 1.7871649065800161, "grad_norm": 0.393647164106369, "learning_rate": 4.206857828304651e-06, "loss": 0.4251, "step": 2200 }, { "epoch": 1.7879772542648253, "grad_norm": 0.3893287479877472, "learning_rate": 4.202190992796756e-06, "loss": 0.4482, "step": 2201 }, { "epoch": 1.7887896019496345, "grad_norm": 0.38627147674560547, "learning_rate": 4.1975248703683855e-06, "loss": 0.4055, "step": 2202 }, { "epoch": 1.7896019496344435, "grad_norm": 0.40966078639030457, "learning_rate": 4.1928594651901055e-06, "loss": 0.4285, "step": 2203 }, { "epoch": 1.7904142973192525, "grad_norm": 0.3683141767978668, "learning_rate": 4.188194781431843e-06, "loss": 0.4166, "step": 2204 }, { "epoch": 1.7912266450040617, "grad_norm": 0.4042739272117615, "learning_rate": 4.18353082326288e-06, "loss": 0.4128, "step": 2205 }, { "epoch": 1.792038992688871, "grad_norm": 0.3942421078681946, "learning_rate": 4.178867594851849e-06, "loss": 0.4156, "step": 2206 }, { "epoch": 1.7928513403736799, "grad_norm": 0.35137414932250977, "learning_rate": 4.1742051003667315e-06, "loss": 0.4284, "step": 2207 }, { "epoch": 1.793663688058489, "grad_norm": 0.4033411741256714, "learning_rate": 4.169543343974852e-06, "loss": 0.4133, "step": 2208 }, { "epoch": 1.794476035743298, "grad_norm": 0.37790918350219727, "learning_rate": 4.164882329842876e-06, "loss": 0.4108, "step": 2209 }, { "epoch": 1.7952883834281073, "grad_norm": 0.37882283329963684, "learning_rate": 4.160222062136805e-06, "loss": 0.4244, "step": 2210 }, { "epoch": 1.7961007311129165, "grad_norm": 0.4156208634376526, "learning_rate": 4.1555625450219725e-06, "loss": 0.4109, "step": 2211 }, { "epoch": 1.7969130787977254, "grad_norm": 0.390031635761261, "learning_rate": 4.150903782663045e-06, "loss": 0.4363, "step": 2212 }, { "epoch": 1.7977254264825344, "grad_norm": 0.35901251435279846, "learning_rate": 4.146245779224011e-06, "loss": 0.428, "step": 2213 }, { "epoch": 1.7985377741673436, "grad_norm": 0.3957388997077942, "learning_rate": 4.141588538868179e-06, "loss": 0.4238, "step": 2214 }, { "epoch": 1.7993501218521528, "grad_norm": 0.3760063946247101, "learning_rate": 4.13693206575818e-06, "loss": 0.4481, "step": 2215 }, { "epoch": 1.8001624695369618, "grad_norm": 0.3841516971588135, "learning_rate": 4.132276364055956e-06, "loss": 0.4268, "step": 2216 }, { "epoch": 1.8009748172217708, "grad_norm": 0.41985827684402466, "learning_rate": 4.127621437922761e-06, "loss": 0.4208, "step": 2217 }, { "epoch": 1.80178716490658, "grad_norm": 0.4195067286491394, "learning_rate": 4.122967291519155e-06, "loss": 0.4164, "step": 2218 }, { "epoch": 1.8025995125913892, "grad_norm": 0.430433988571167, "learning_rate": 4.118313929005002e-06, "loss": 0.43, "step": 2219 }, { "epoch": 1.8034118602761984, "grad_norm": 0.47560441493988037, "learning_rate": 4.113661354539464e-06, "loss": 0.435, "step": 2220 }, { "epoch": 1.8042242079610074, "grad_norm": 0.4253809154033661, "learning_rate": 4.109009572280999e-06, "loss": 0.4241, "step": 2221 }, { "epoch": 1.8050365556458163, "grad_norm": 0.4224478304386139, "learning_rate": 4.1043585863873595e-06, "loss": 0.4214, "step": 2222 }, { "epoch": 1.8058489033306255, "grad_norm": 0.5442357063293457, "learning_rate": 4.09970840101558e-06, "loss": 0.4389, "step": 2223 }, { "epoch": 1.8066612510154347, "grad_norm": 0.3897160291671753, "learning_rate": 4.095059020321988e-06, "loss": 0.4395, "step": 2224 }, { "epoch": 1.8074735987002437, "grad_norm": 0.40830859541893005, "learning_rate": 4.090410448462186e-06, "loss": 0.4209, "step": 2225 }, { "epoch": 1.8082859463850527, "grad_norm": 0.4319519102573395, "learning_rate": 4.085762689591054e-06, "loss": 0.4361, "step": 2226 }, { "epoch": 1.809098294069862, "grad_norm": 0.4178127944469452, "learning_rate": 4.081115747862747e-06, "loss": 0.4287, "step": 2227 }, { "epoch": 1.809910641754671, "grad_norm": 0.3760089874267578, "learning_rate": 4.076469627430689e-06, "loss": 0.4514, "step": 2228 }, { "epoch": 1.81072298943948, "grad_norm": 0.36359506845474243, "learning_rate": 4.071824332447568e-06, "loss": 0.44, "step": 2229 }, { "epoch": 1.811535337124289, "grad_norm": 0.42622217535972595, "learning_rate": 4.067179867065339e-06, "loss": 0.4136, "step": 2230 }, { "epoch": 1.8123476848090982, "grad_norm": 0.3691442906856537, "learning_rate": 4.0625362354352105e-06, "loss": 0.4366, "step": 2231 }, { "epoch": 1.8131600324939074, "grad_norm": 0.42135000228881836, "learning_rate": 4.0578934417076486e-06, "loss": 0.4577, "step": 2232 }, { "epoch": 1.8139723801787166, "grad_norm": 0.36476534605026245, "learning_rate": 4.053251490032368e-06, "loss": 0.4063, "step": 2233 }, { "epoch": 1.8147847278635256, "grad_norm": 0.37307024002075195, "learning_rate": 4.048610384558335e-06, "loss": 0.4379, "step": 2234 }, { "epoch": 1.8155970755483346, "grad_norm": 0.39189696311950684, "learning_rate": 4.043970129433757e-06, "loss": 0.4349, "step": 2235 }, { "epoch": 1.8164094232331438, "grad_norm": 0.44861528277397156, "learning_rate": 4.039330728806079e-06, "loss": 0.4451, "step": 2236 }, { "epoch": 1.817221770917953, "grad_norm": 0.39244669675827026, "learning_rate": 4.034692186821986e-06, "loss": 0.404, "step": 2237 }, { "epoch": 1.818034118602762, "grad_norm": 0.4264596700668335, "learning_rate": 4.030054507627395e-06, "loss": 0.4417, "step": 2238 }, { "epoch": 1.818846466287571, "grad_norm": 0.410950243473053, "learning_rate": 4.0254176953674505e-06, "loss": 0.4095, "step": 2239 }, { "epoch": 1.8196588139723802, "grad_norm": 0.40892279148101807, "learning_rate": 4.020781754186523e-06, "loss": 0.4328, "step": 2240 }, { "epoch": 1.8204711616571894, "grad_norm": 0.36550286412239075, "learning_rate": 4.016146688228203e-06, "loss": 0.4146, "step": 2241 }, { "epoch": 1.8212835093419983, "grad_norm": 0.4772724509239197, "learning_rate": 4.011512501635301e-06, "loss": 0.4364, "step": 2242 }, { "epoch": 1.8220958570268073, "grad_norm": 0.40436846017837524, "learning_rate": 4.00687919854984e-06, "loss": 0.4434, "step": 2243 }, { "epoch": 1.8229082047116165, "grad_norm": 0.3535594046115875, "learning_rate": 4.002246783113054e-06, "loss": 0.4382, "step": 2244 }, { "epoch": 1.8237205523964257, "grad_norm": 0.43400248885154724, "learning_rate": 3.997615259465382e-06, "loss": 0.3913, "step": 2245 }, { "epoch": 1.824532900081235, "grad_norm": 0.39666011929512024, "learning_rate": 3.992984631746469e-06, "loss": 0.4076, "step": 2246 }, { "epoch": 1.825345247766044, "grad_norm": 0.40733206272125244, "learning_rate": 3.988354904095157e-06, "loss": 0.4402, "step": 2247 }, { "epoch": 1.8261575954508529, "grad_norm": 0.3976712226867676, "learning_rate": 3.983726080649484e-06, "loss": 0.4116, "step": 2248 }, { "epoch": 1.826969943135662, "grad_norm": 0.3857514262199402, "learning_rate": 3.979098165546679e-06, "loss": 0.4153, "step": 2249 }, { "epoch": 1.8277822908204713, "grad_norm": 0.35972920060157776, "learning_rate": 3.974471162923162e-06, "loss": 0.4044, "step": 2250 }, { "epoch": 1.8285946385052803, "grad_norm": 0.3714330196380615, "learning_rate": 3.969845076914532e-06, "loss": 0.438, "step": 2251 }, { "epoch": 1.8294069861900892, "grad_norm": 0.431874543428421, "learning_rate": 3.965219911655575e-06, "loss": 0.4414, "step": 2252 }, { "epoch": 1.8302193338748984, "grad_norm": 0.3599884510040283, "learning_rate": 3.96059567128025e-06, "loss": 0.4279, "step": 2253 }, { "epoch": 1.8310316815597076, "grad_norm": 0.39495590329170227, "learning_rate": 3.9559723599216905e-06, "loss": 0.4438, "step": 2254 }, { "epoch": 1.8318440292445166, "grad_norm": 0.37928205728530884, "learning_rate": 3.951349981712199e-06, "loss": 0.4276, "step": 2255 }, { "epoch": 1.8326563769293256, "grad_norm": 0.38191428780555725, "learning_rate": 3.946728540783247e-06, "loss": 0.4272, "step": 2256 }, { "epoch": 1.8334687246141348, "grad_norm": 0.3588366210460663, "learning_rate": 3.942108041265463e-06, "loss": 0.4338, "step": 2257 }, { "epoch": 1.834281072298944, "grad_norm": 0.3449942469596863, "learning_rate": 3.9374884872886394e-06, "loss": 0.409, "step": 2258 }, { "epoch": 1.8350934199837532, "grad_norm": 0.38562554121017456, "learning_rate": 3.9328698829817195e-06, "loss": 0.4256, "step": 2259 }, { "epoch": 1.8359057676685622, "grad_norm": 0.3505711257457733, "learning_rate": 3.928252232472799e-06, "loss": 0.4158, "step": 2260 }, { "epoch": 1.8367181153533712, "grad_norm": 0.35499346256256104, "learning_rate": 3.923635539889121e-06, "loss": 0.4271, "step": 2261 }, { "epoch": 1.8375304630381804, "grad_norm": 0.358233243227005, "learning_rate": 3.919019809357075e-06, "loss": 0.4043, "step": 2262 }, { "epoch": 1.8383428107229896, "grad_norm": 0.359464168548584, "learning_rate": 3.914405045002186e-06, "loss": 0.4056, "step": 2263 }, { "epoch": 1.8391551584077985, "grad_norm": 0.4074437916278839, "learning_rate": 3.9097912509491195e-06, "loss": 0.4365, "step": 2264 }, { "epoch": 1.8399675060926075, "grad_norm": 0.3425873816013336, "learning_rate": 3.905178431321672e-06, "loss": 0.4344, "step": 2265 }, { "epoch": 1.8407798537774167, "grad_norm": 0.3793255686759949, "learning_rate": 3.9005665902427695e-06, "loss": 0.4082, "step": 2266 }, { "epoch": 1.841592201462226, "grad_norm": 0.3720904588699341, "learning_rate": 3.895955731834461e-06, "loss": 0.4335, "step": 2267 }, { "epoch": 1.8424045491470349, "grad_norm": 0.36797258257865906, "learning_rate": 3.8913458602179215e-06, "loss": 0.4197, "step": 2268 }, { "epoch": 1.8432168968318439, "grad_norm": 0.385223388671875, "learning_rate": 3.88673697951344e-06, "loss": 0.4144, "step": 2269 }, { "epoch": 1.844029244516653, "grad_norm": 0.37110820412635803, "learning_rate": 3.882129093840423e-06, "loss": 0.4222, "step": 2270 }, { "epoch": 1.8448415922014623, "grad_norm": 0.37669169902801514, "learning_rate": 3.8775222073173844e-06, "loss": 0.4026, "step": 2271 }, { "epoch": 1.8456539398862715, "grad_norm": 0.35052525997161865, "learning_rate": 3.872916324061949e-06, "loss": 0.4293, "step": 2272 }, { "epoch": 1.8464662875710804, "grad_norm": 0.38547343015670776, "learning_rate": 3.8683114481908415e-06, "loss": 0.3993, "step": 2273 }, { "epoch": 1.8472786352558894, "grad_norm": 0.43375420570373535, "learning_rate": 3.8637075838198864e-06, "loss": 0.4487, "step": 2274 }, { "epoch": 1.8480909829406986, "grad_norm": 0.37127524614334106, "learning_rate": 3.859104735064008e-06, "loss": 0.4155, "step": 2275 }, { "epoch": 1.8489033306255078, "grad_norm": 0.3490426540374756, "learning_rate": 3.854502906037217e-06, "loss": 0.4477, "step": 2276 }, { "epoch": 1.8497156783103168, "grad_norm": 0.41017231345176697, "learning_rate": 3.849902100852618e-06, "loss": 0.432, "step": 2277 }, { "epoch": 1.8505280259951258, "grad_norm": 0.42969265580177307, "learning_rate": 3.845302323622398e-06, "loss": 0.4492, "step": 2278 }, { "epoch": 1.851340373679935, "grad_norm": 0.3354117274284363, "learning_rate": 3.8407035784578226e-06, "loss": 0.4335, "step": 2279 }, { "epoch": 1.8521527213647442, "grad_norm": 0.40715301036834717, "learning_rate": 3.8361058694692414e-06, "loss": 0.4481, "step": 2280 }, { "epoch": 1.8529650690495532, "grad_norm": 0.3927744925022125, "learning_rate": 3.83150920076607e-06, "loss": 0.4287, "step": 2281 }, { "epoch": 1.8537774167343624, "grad_norm": 0.38053053617477417, "learning_rate": 3.826913576456803e-06, "loss": 0.4453, "step": 2282 }, { "epoch": 1.8545897644191713, "grad_norm": 0.3756331205368042, "learning_rate": 3.822319000648995e-06, "loss": 0.4375, "step": 2283 }, { "epoch": 1.8554021121039805, "grad_norm": 0.35971614718437195, "learning_rate": 3.817725477449265e-06, "loss": 0.4054, "step": 2284 }, { "epoch": 1.8562144597887897, "grad_norm": 0.3703836500644684, "learning_rate": 3.8131330109632925e-06, "loss": 0.4367, "step": 2285 }, { "epoch": 1.8570268074735987, "grad_norm": 0.3506587743759155, "learning_rate": 3.8085416052958107e-06, "loss": 0.4399, "step": 2286 }, { "epoch": 1.8578391551584077, "grad_norm": 0.3671494424343109, "learning_rate": 3.8039512645506055e-06, "loss": 0.4453, "step": 2287 }, { "epoch": 1.858651502843217, "grad_norm": 0.36680111289024353, "learning_rate": 3.7993619928305113e-06, "loss": 0.4388, "step": 2288 }, { "epoch": 1.859463850528026, "grad_norm": 0.3717835247516632, "learning_rate": 3.7947737942374065e-06, "loss": 0.4102, "step": 2289 }, { "epoch": 1.860276198212835, "grad_norm": 0.3481735289096832, "learning_rate": 3.7901866728722102e-06, "loss": 0.4172, "step": 2290 }, { "epoch": 1.861088545897644, "grad_norm": 0.3876075744628906, "learning_rate": 3.7856006328348805e-06, "loss": 0.4301, "step": 2291 }, { "epoch": 1.8619008935824533, "grad_norm": 0.36815333366394043, "learning_rate": 3.7810156782244055e-06, "loss": 0.4397, "step": 2292 }, { "epoch": 1.8627132412672625, "grad_norm": 0.3666851818561554, "learning_rate": 3.7764318131388067e-06, "loss": 0.4196, "step": 2293 }, { "epoch": 1.8635255889520714, "grad_norm": 0.35023778676986694, "learning_rate": 3.7718490416751297e-06, "loss": 0.4455, "step": 2294 }, { "epoch": 1.8643379366368806, "grad_norm": 0.35064491629600525, "learning_rate": 3.7672673679294436e-06, "loss": 0.4415, "step": 2295 }, { "epoch": 1.8651502843216896, "grad_norm": 0.34776070713996887, "learning_rate": 3.762686795996835e-06, "loss": 0.44, "step": 2296 }, { "epoch": 1.8659626320064988, "grad_norm": 0.34763410687446594, "learning_rate": 3.7581073299714077e-06, "loss": 0.4118, "step": 2297 }, { "epoch": 1.866774979691308, "grad_norm": 0.3268580138683319, "learning_rate": 3.7535289739462757e-06, "loss": 0.4227, "step": 2298 }, { "epoch": 1.867587327376117, "grad_norm": 0.36501774191856384, "learning_rate": 3.7489517320135583e-06, "loss": 0.4363, "step": 2299 }, { "epoch": 1.868399675060926, "grad_norm": 0.35482293367385864, "learning_rate": 3.7443756082643866e-06, "loss": 0.4454, "step": 2300 }, { "epoch": 1.8692120227457352, "grad_norm": 0.378528356552124, "learning_rate": 3.739800606788886e-06, "loss": 0.4515, "step": 2301 }, { "epoch": 1.8700243704305444, "grad_norm": 0.35682910680770874, "learning_rate": 3.735226731676179e-06, "loss": 0.4244, "step": 2302 }, { "epoch": 1.8708367181153533, "grad_norm": 0.38846689462661743, "learning_rate": 3.7306539870143853e-06, "loss": 0.4053, "step": 2303 }, { "epoch": 1.8716490658001623, "grad_norm": 0.3390444219112396, "learning_rate": 3.7260823768906096e-06, "loss": 0.4115, "step": 2304 }, { "epoch": 1.8724614134849715, "grad_norm": 0.34597837924957275, "learning_rate": 3.7215119053909444e-06, "loss": 0.4293, "step": 2305 }, { "epoch": 1.8732737611697807, "grad_norm": 0.35154134035110474, "learning_rate": 3.7169425766004653e-06, "loss": 0.4305, "step": 2306 }, { "epoch": 1.87408610885459, "grad_norm": 0.3287436366081238, "learning_rate": 3.7123743946032263e-06, "loss": 0.413, "step": 2307 }, { "epoch": 1.874898456539399, "grad_norm": 0.3321009874343872, "learning_rate": 3.7078073634822565e-06, "loss": 0.4211, "step": 2308 }, { "epoch": 1.8757108042242079, "grad_norm": 0.31308096647262573, "learning_rate": 3.7032414873195544e-06, "loss": 0.4221, "step": 2309 }, { "epoch": 1.876523151909017, "grad_norm": 0.35819676518440247, "learning_rate": 3.698676770196089e-06, "loss": 0.428, "step": 2310 }, { "epoch": 1.8773354995938263, "grad_norm": 0.3611733019351959, "learning_rate": 3.6941132161917916e-06, "loss": 0.4267, "step": 2311 }, { "epoch": 1.8781478472786353, "grad_norm": 0.3362630009651184, "learning_rate": 3.6895508293855546e-06, "loss": 0.4099, "step": 2312 }, { "epoch": 1.8789601949634442, "grad_norm": 0.3588707745075226, "learning_rate": 3.6849896138552265e-06, "loss": 0.4349, "step": 2313 }, { "epoch": 1.8797725426482534, "grad_norm": 0.3467176854610443, "learning_rate": 3.68042957367761e-06, "loss": 0.4422, "step": 2314 }, { "epoch": 1.8805848903330626, "grad_norm": 0.38330093026161194, "learning_rate": 3.6758707129284556e-06, "loss": 0.4118, "step": 2315 }, { "epoch": 1.8813972380178716, "grad_norm": 0.35750943422317505, "learning_rate": 3.6713130356824615e-06, "loss": 0.4221, "step": 2316 }, { "epoch": 1.8822095857026806, "grad_norm": 0.36062100529670715, "learning_rate": 3.666756546013267e-06, "loss": 0.4105, "step": 2317 }, { "epoch": 1.8830219333874898, "grad_norm": 0.36713045835494995, "learning_rate": 3.6622012479934505e-06, "loss": 0.4373, "step": 2318 }, { "epoch": 1.883834281072299, "grad_norm": 0.359055757522583, "learning_rate": 3.6576471456945244e-06, "loss": 0.4193, "step": 2319 }, { "epoch": 1.8846466287571082, "grad_norm": 0.3750867247581482, "learning_rate": 3.653094243186933e-06, "loss": 0.4159, "step": 2320 }, { "epoch": 1.8854589764419172, "grad_norm": 0.40929970145225525, "learning_rate": 3.648542544540049e-06, "loss": 0.4352, "step": 2321 }, { "epoch": 1.8862713241267262, "grad_norm": 0.33511918783187866, "learning_rate": 3.6439920538221675e-06, "loss": 0.4458, "step": 2322 }, { "epoch": 1.8870836718115354, "grad_norm": 0.3916158080101013, "learning_rate": 3.639442775100505e-06, "loss": 0.4513, "step": 2323 }, { "epoch": 1.8878960194963446, "grad_norm": 0.3896113336086273, "learning_rate": 3.634894712441193e-06, "loss": 0.4413, "step": 2324 }, { "epoch": 1.8887083671811535, "grad_norm": 0.36245447397232056, "learning_rate": 3.630347869909281e-06, "loss": 0.4139, "step": 2325 }, { "epoch": 1.8895207148659625, "grad_norm": 0.3438439965248108, "learning_rate": 3.6258022515687215e-06, "loss": 0.4545, "step": 2326 }, { "epoch": 1.8903330625507717, "grad_norm": 0.3938590884208679, "learning_rate": 3.6212578614823767e-06, "loss": 0.4112, "step": 2327 }, { "epoch": 1.891145410235581, "grad_norm": 0.3685142993927002, "learning_rate": 3.616714703712011e-06, "loss": 0.4262, "step": 2328 }, { "epoch": 1.89195775792039, "grad_norm": 0.37876415252685547, "learning_rate": 3.6121727823182854e-06, "loss": 0.4177, "step": 2329 }, { "epoch": 1.8927701056051989, "grad_norm": 0.37008458375930786, "learning_rate": 3.6076321013607575e-06, "loss": 0.4316, "step": 2330 }, { "epoch": 1.893582453290008, "grad_norm": 0.3596286475658417, "learning_rate": 3.6030926648978748e-06, "loss": 0.4144, "step": 2331 }, { "epoch": 1.8943948009748173, "grad_norm": 0.41880637407302856, "learning_rate": 3.598554476986974e-06, "loss": 0.4345, "step": 2332 }, { "epoch": 1.8952071486596265, "grad_norm": 0.37811264395713806, "learning_rate": 3.594017541684275e-06, "loss": 0.4377, "step": 2333 }, { "epoch": 1.8960194963444355, "grad_norm": 0.35049620270729065, "learning_rate": 3.589481863044878e-06, "loss": 0.4273, "step": 2334 }, { "epoch": 1.8968318440292444, "grad_norm": 0.3578229546546936, "learning_rate": 3.5849474451227607e-06, "loss": 0.4205, "step": 2335 }, { "epoch": 1.8976441917140536, "grad_norm": 0.4096352458000183, "learning_rate": 3.5804142919707734e-06, "loss": 0.4413, "step": 2336 }, { "epoch": 1.8984565393988628, "grad_norm": 0.3850961923599243, "learning_rate": 3.5758824076406358e-06, "loss": 0.4293, "step": 2337 }, { "epoch": 1.8992688870836718, "grad_norm": 0.3554864227771759, "learning_rate": 3.5713517961829314e-06, "loss": 0.4074, "step": 2338 }, { "epoch": 1.9000812347684808, "grad_norm": 0.40958690643310547, "learning_rate": 3.5668224616471136e-06, "loss": 0.4439, "step": 2339 }, { "epoch": 1.90089358245329, "grad_norm": 0.36483466625213623, "learning_rate": 3.562294408081486e-06, "loss": 0.4454, "step": 2340 }, { "epoch": 1.9017059301380992, "grad_norm": 0.3739314675331116, "learning_rate": 3.5577676395332116e-06, "loss": 0.4318, "step": 2341 }, { "epoch": 1.9025182778229082, "grad_norm": 0.3632747232913971, "learning_rate": 3.5532421600483035e-06, "loss": 0.4232, "step": 2342 }, { "epoch": 1.9033306255077171, "grad_norm": 0.33664384484291077, "learning_rate": 3.5487179736716233e-06, "loss": 0.4001, "step": 2343 }, { "epoch": 1.9041429731925263, "grad_norm": 0.357922226190567, "learning_rate": 3.5441950844468754e-06, "loss": 0.4009, "step": 2344 }, { "epoch": 1.9049553208773355, "grad_norm": 0.37992775440216064, "learning_rate": 3.539673496416607e-06, "loss": 0.4219, "step": 2345 }, { "epoch": 1.9057676685621447, "grad_norm": 0.3308415114879608, "learning_rate": 3.5351532136222012e-06, "loss": 0.4336, "step": 2346 }, { "epoch": 1.9065800162469537, "grad_norm": 0.31776073575019836, "learning_rate": 3.5306342401038727e-06, "loss": 0.4485, "step": 2347 }, { "epoch": 1.9073923639317627, "grad_norm": 0.3427734673023224, "learning_rate": 3.52611657990067e-06, "loss": 0.4086, "step": 2348 }, { "epoch": 1.908204711616572, "grad_norm": 0.36637574434280396, "learning_rate": 3.5216002370504656e-06, "loss": 0.4193, "step": 2349 }, { "epoch": 1.909017059301381, "grad_norm": 0.3516906201839447, "learning_rate": 3.5170852155899517e-06, "loss": 0.4208, "step": 2350 }, { "epoch": 1.90982940698619, "grad_norm": 0.32919102907180786, "learning_rate": 3.512571519554645e-06, "loss": 0.4459, "step": 2351 }, { "epoch": 1.910641754670999, "grad_norm": 0.3746897578239441, "learning_rate": 3.5080591529788736e-06, "loss": 0.4552, "step": 2352 }, { "epoch": 1.9114541023558083, "grad_norm": 0.3721083998680115, "learning_rate": 3.5035481198957794e-06, "loss": 0.4212, "step": 2353 }, { "epoch": 1.9122664500406175, "grad_norm": 0.40002909302711487, "learning_rate": 3.4990384243373105e-06, "loss": 0.4273, "step": 2354 }, { "epoch": 1.9130787977254264, "grad_norm": 0.3524540066719055, "learning_rate": 3.494530070334221e-06, "loss": 0.4205, "step": 2355 }, { "epoch": 1.9138911454102354, "grad_norm": 0.3941265642642975, "learning_rate": 3.490023061916065e-06, "loss": 0.4215, "step": 2356 }, { "epoch": 1.9147034930950446, "grad_norm": 0.4194328486919403, "learning_rate": 3.4855174031111947e-06, "loss": 0.4392, "step": 2357 }, { "epoch": 1.9155158407798538, "grad_norm": 0.41141635179519653, "learning_rate": 3.481013097946756e-06, "loss": 0.4514, "step": 2358 }, { "epoch": 1.916328188464663, "grad_norm": 0.3557471036911011, "learning_rate": 3.4765101504486842e-06, "loss": 0.4086, "step": 2359 }, { "epoch": 1.917140536149472, "grad_norm": 0.3776426315307617, "learning_rate": 3.4720085646417025e-06, "loss": 0.4562, "step": 2360 }, { "epoch": 1.917952883834281, "grad_norm": 0.39756423234939575, "learning_rate": 3.467508344549314e-06, "loss": 0.4495, "step": 2361 }, { "epoch": 1.9187652315190902, "grad_norm": 0.3712599277496338, "learning_rate": 3.4630094941938054e-06, "loss": 0.4232, "step": 2362 }, { "epoch": 1.9195775792038994, "grad_norm": 0.361512690782547, "learning_rate": 3.458512017596235e-06, "loss": 0.428, "step": 2363 }, { "epoch": 1.9203899268887084, "grad_norm": 0.3880564570426941, "learning_rate": 3.454015918776437e-06, "loss": 0.4267, "step": 2364 }, { "epoch": 1.9212022745735173, "grad_norm": 0.3553383946418762, "learning_rate": 3.449521201753011e-06, "loss": 0.4209, "step": 2365 }, { "epoch": 1.9220146222583265, "grad_norm": 0.3820456862449646, "learning_rate": 3.445027870543323e-06, "loss": 0.4349, "step": 2366 }, { "epoch": 1.9228269699431357, "grad_norm": 0.4123988449573517, "learning_rate": 3.4405359291635008e-06, "loss": 0.4115, "step": 2367 }, { "epoch": 1.9236393176279447, "grad_norm": 0.4753648638725281, "learning_rate": 3.436045381628429e-06, "loss": 0.4367, "step": 2368 }, { "epoch": 1.924451665312754, "grad_norm": 0.37252673506736755, "learning_rate": 3.431556231951747e-06, "loss": 0.4537, "step": 2369 }, { "epoch": 1.925264012997563, "grad_norm": 0.3678191006183624, "learning_rate": 3.4270684841458445e-06, "loss": 0.4107, "step": 2370 }, { "epoch": 1.926076360682372, "grad_norm": 0.4515796899795532, "learning_rate": 3.4225821422218585e-06, "loss": 0.4235, "step": 2371 }, { "epoch": 1.9268887083671813, "grad_norm": 0.36400699615478516, "learning_rate": 3.418097210189669e-06, "loss": 0.4398, "step": 2372 }, { "epoch": 1.9277010560519903, "grad_norm": 0.35223400592803955, "learning_rate": 3.4136136920578967e-06, "loss": 0.4604, "step": 2373 }, { "epoch": 1.9285134037367992, "grad_norm": 0.36282259225845337, "learning_rate": 3.4091315918338968e-06, "loss": 0.4369, "step": 2374 }, { "epoch": 1.9293257514216084, "grad_norm": 0.43420082330703735, "learning_rate": 3.4046509135237593e-06, "loss": 0.4361, "step": 2375 }, { "epoch": 1.9301380991064176, "grad_norm": 0.3349204957485199, "learning_rate": 3.400171661132299e-06, "loss": 0.4215, "step": 2376 }, { "epoch": 1.9309504467912266, "grad_norm": 0.38303491473197937, "learning_rate": 3.395693838663065e-06, "loss": 0.4063, "step": 2377 }, { "epoch": 1.9317627944760356, "grad_norm": 0.3924751877784729, "learning_rate": 3.3912174501183206e-06, "loss": 0.4434, "step": 2378 }, { "epoch": 1.9325751421608448, "grad_norm": 0.3639524579048157, "learning_rate": 3.3867424994990483e-06, "loss": 0.4144, "step": 2379 }, { "epoch": 1.933387489845654, "grad_norm": 0.31736093759536743, "learning_rate": 3.3822689908049487e-06, "loss": 0.4256, "step": 2380 }, { "epoch": 1.934199837530463, "grad_norm": 0.36893829703330994, "learning_rate": 3.37779692803443e-06, "loss": 0.4012, "step": 2381 }, { "epoch": 1.9350121852152722, "grad_norm": 0.3627434968948364, "learning_rate": 3.373326315184612e-06, "loss": 0.4363, "step": 2382 }, { "epoch": 1.9358245329000812, "grad_norm": 0.3698161840438843, "learning_rate": 3.368857156251314e-06, "loss": 0.4194, "step": 2383 }, { "epoch": 1.9366368805848904, "grad_norm": 0.4146488606929779, "learning_rate": 3.364389455229059e-06, "loss": 0.4312, "step": 2384 }, { "epoch": 1.9374492282696996, "grad_norm": 0.34686949849128723, "learning_rate": 3.359923216111067e-06, "loss": 0.4318, "step": 2385 }, { "epoch": 1.9382615759545085, "grad_norm": 0.3719654679298401, "learning_rate": 3.3554584428892488e-06, "loss": 0.4356, "step": 2386 }, { "epoch": 1.9390739236393175, "grad_norm": 0.3691517114639282, "learning_rate": 3.3509951395542073e-06, "loss": 0.4259, "step": 2387 }, { "epoch": 1.9398862713241267, "grad_norm": 0.4039909243583679, "learning_rate": 3.3465333100952306e-06, "loss": 0.4058, "step": 2388 }, { "epoch": 1.940698619008936, "grad_norm": 0.36259227991104126, "learning_rate": 3.342072958500291e-06, "loss": 0.4475, "step": 2389 }, { "epoch": 1.941510966693745, "grad_norm": 0.40982601046562195, "learning_rate": 3.337614088756037e-06, "loss": 0.45, "step": 2390 }, { "epoch": 1.9423233143785539, "grad_norm": 0.3331512212753296, "learning_rate": 3.3331567048477946e-06, "loss": 0.4145, "step": 2391 }, { "epoch": 1.943135662063363, "grad_norm": 0.41676145792007446, "learning_rate": 3.328700810759562e-06, "loss": 0.4495, "step": 2392 }, { "epoch": 1.9439480097481723, "grad_norm": 0.4020812213420868, "learning_rate": 3.324246410474006e-06, "loss": 0.4133, "step": 2393 }, { "epoch": 1.9447603574329815, "grad_norm": 0.3971535861492157, "learning_rate": 3.319793507972454e-06, "loss": 0.4347, "step": 2394 }, { "epoch": 1.9455727051177905, "grad_norm": 0.3705967664718628, "learning_rate": 3.3153421072349024e-06, "loss": 0.4066, "step": 2395 }, { "epoch": 1.9463850528025994, "grad_norm": 0.3734128177165985, "learning_rate": 3.310892212239999e-06, "loss": 0.4391, "step": 2396 }, { "epoch": 1.9471974004874086, "grad_norm": 0.38919153809547424, "learning_rate": 3.306443826965049e-06, "loss": 0.4147, "step": 2397 }, { "epoch": 1.9480097481722178, "grad_norm": 0.3683825433254242, "learning_rate": 3.3019969553860056e-06, "loss": 0.4079, "step": 2398 }, { "epoch": 1.9488220958570268, "grad_norm": 0.34082093834877014, "learning_rate": 3.2975516014774713e-06, "loss": 0.397, "step": 2399 }, { "epoch": 1.9496344435418358, "grad_norm": 0.36290448904037476, "learning_rate": 3.2931077692126912e-06, "loss": 0.4308, "step": 2400 }, { "epoch": 1.950446791226645, "grad_norm": 0.365864098072052, "learning_rate": 3.2886654625635507e-06, "loss": 0.4422, "step": 2401 }, { "epoch": 1.9512591389114542, "grad_norm": 0.3273039758205414, "learning_rate": 3.284224685500571e-06, "loss": 0.4198, "step": 2402 }, { "epoch": 1.9520714865962632, "grad_norm": 0.3724144399166107, "learning_rate": 3.2797854419929064e-06, "loss": 0.4257, "step": 2403 }, { "epoch": 1.9528838342810722, "grad_norm": 0.33202770352363586, "learning_rate": 3.2753477360083417e-06, "loss": 0.4296, "step": 2404 }, { "epoch": 1.9536961819658814, "grad_norm": 0.3399312496185303, "learning_rate": 3.2709115715132868e-06, "loss": 0.403, "step": 2405 }, { "epoch": 1.9545085296506906, "grad_norm": 0.37480056285858154, "learning_rate": 3.2664769524727712e-06, "loss": 0.3994, "step": 2406 }, { "epoch": 1.9553208773354998, "grad_norm": 0.3587169349193573, "learning_rate": 3.2620438828504473e-06, "loss": 0.4363, "step": 2407 }, { "epoch": 1.9561332250203087, "grad_norm": 0.38213640451431274, "learning_rate": 3.25761236660858e-06, "loss": 0.4437, "step": 2408 }, { "epoch": 1.9569455727051177, "grad_norm": 0.3766333758831024, "learning_rate": 3.2531824077080466e-06, "loss": 0.438, "step": 2409 }, { "epoch": 1.957757920389927, "grad_norm": 0.3457035720348358, "learning_rate": 3.2487540101083325e-06, "loss": 0.4207, "step": 2410 }, { "epoch": 1.958570268074736, "grad_norm": 0.36141955852508545, "learning_rate": 3.244327177767527e-06, "loss": 0.4355, "step": 2411 }, { "epoch": 1.959382615759545, "grad_norm": 0.36657431721687317, "learning_rate": 3.23990191464232e-06, "loss": 0.4325, "step": 2412 }, { "epoch": 1.960194963444354, "grad_norm": 0.3252984583377838, "learning_rate": 3.2354782246880016e-06, "loss": 0.4362, "step": 2413 }, { "epoch": 1.9610073111291633, "grad_norm": 0.36180341243743896, "learning_rate": 3.2310561118584526e-06, "loss": 0.4279, "step": 2414 }, { "epoch": 1.9618196588139725, "grad_norm": 0.40245968103408813, "learning_rate": 3.226635580106147e-06, "loss": 0.414, "step": 2415 }, { "epoch": 1.9626320064987814, "grad_norm": 0.41776153445243835, "learning_rate": 3.222216633382142e-06, "loss": 0.4078, "step": 2416 }, { "epoch": 1.9634443541835904, "grad_norm": 0.40520766377449036, "learning_rate": 3.2177992756360814e-06, "loss": 0.4335, "step": 2417 }, { "epoch": 1.9642567018683996, "grad_norm": 0.3651321232318878, "learning_rate": 3.2133835108161875e-06, "loss": 0.4274, "step": 2418 }, { "epoch": 1.9650690495532088, "grad_norm": 0.4189288020133972, "learning_rate": 3.2089693428692587e-06, "loss": 0.459, "step": 2419 }, { "epoch": 1.965881397238018, "grad_norm": 0.35893020033836365, "learning_rate": 3.204556775740665e-06, "loss": 0.4185, "step": 2420 }, { "epoch": 1.966693744922827, "grad_norm": 0.32167330384254456, "learning_rate": 3.200145813374349e-06, "loss": 0.4035, "step": 2421 }, { "epoch": 1.967506092607636, "grad_norm": 0.3753286302089691, "learning_rate": 3.195736459712817e-06, "loss": 0.3885, "step": 2422 }, { "epoch": 1.9683184402924452, "grad_norm": 0.3785870671272278, "learning_rate": 3.1913287186971353e-06, "loss": 0.4179, "step": 2423 }, { "epoch": 1.9691307879772544, "grad_norm": 0.38640910387039185, "learning_rate": 3.186922594266932e-06, "loss": 0.4336, "step": 2424 }, { "epoch": 1.9699431356620634, "grad_norm": 0.3653258681297302, "learning_rate": 3.1825180903603903e-06, "loss": 0.4444, "step": 2425 }, { "epoch": 1.9707554833468723, "grad_norm": 0.33298659324645996, "learning_rate": 3.178115210914242e-06, "loss": 0.4105, "step": 2426 }, { "epoch": 1.9715678310316815, "grad_norm": 0.37267443537712097, "learning_rate": 3.1737139598637688e-06, "loss": 0.421, "step": 2427 }, { "epoch": 1.9723801787164907, "grad_norm": 0.36115020513534546, "learning_rate": 3.1693143411427984e-06, "loss": 0.4458, "step": 2428 }, { "epoch": 1.9731925264012997, "grad_norm": 0.34956997632980347, "learning_rate": 3.164916358683697e-06, "loss": 0.4185, "step": 2429 }, { "epoch": 1.9740048740861087, "grad_norm": 0.3388860821723938, "learning_rate": 3.160520016417369e-06, "loss": 0.4112, "step": 2430 }, { "epoch": 1.974817221770918, "grad_norm": 0.3410119414329529, "learning_rate": 3.1561253182732544e-06, "loss": 0.3991, "step": 2431 }, { "epoch": 1.975629569455727, "grad_norm": 0.37571844458580017, "learning_rate": 3.1517322681793215e-06, "loss": 0.4194, "step": 2432 }, { "epoch": 1.9764419171405363, "grad_norm": 0.3356778025627136, "learning_rate": 3.1473408700620665e-06, "loss": 0.4178, "step": 2433 }, { "epoch": 1.9772542648253453, "grad_norm": 0.3451114892959595, "learning_rate": 3.1429511278465104e-06, "loss": 0.4179, "step": 2434 }, { "epoch": 1.9780666125101543, "grad_norm": 0.33805856108665466, "learning_rate": 3.138563045456191e-06, "loss": 0.4374, "step": 2435 }, { "epoch": 1.9788789601949635, "grad_norm": 0.3136672377586365, "learning_rate": 3.1341766268131657e-06, "loss": 0.4187, "step": 2436 }, { "epoch": 1.9796913078797727, "grad_norm": 0.3753918707370758, "learning_rate": 3.1297918758380035e-06, "loss": 0.4322, "step": 2437 }, { "epoch": 1.9805036555645816, "grad_norm": 0.35181042551994324, "learning_rate": 3.1254087964497816e-06, "loss": 0.4213, "step": 2438 }, { "epoch": 1.9813160032493906, "grad_norm": 0.33582577109336853, "learning_rate": 3.121027392566086e-06, "loss": 0.4142, "step": 2439 }, { "epoch": 1.9821283509341998, "grad_norm": 0.3428526222705841, "learning_rate": 3.1166476681030035e-06, "loss": 0.4312, "step": 2440 }, { "epoch": 1.982940698619009, "grad_norm": 0.41421839594841003, "learning_rate": 3.1122696269751196e-06, "loss": 0.43, "step": 2441 }, { "epoch": 1.983753046303818, "grad_norm": 0.3320109248161316, "learning_rate": 3.1078932730955152e-06, "loss": 0.4331, "step": 2442 }, { "epoch": 1.984565393988627, "grad_norm": 0.3382517695426941, "learning_rate": 3.103518610375764e-06, "loss": 0.4403, "step": 2443 }, { "epoch": 1.9853777416734362, "grad_norm": 0.35794687271118164, "learning_rate": 3.099145642725927e-06, "loss": 0.4108, "step": 2444 }, { "epoch": 1.9861900893582454, "grad_norm": 0.33276841044425964, "learning_rate": 3.0947743740545523e-06, "loss": 0.4375, "step": 2445 }, { "epoch": 1.9870024370430546, "grad_norm": 0.3354824185371399, "learning_rate": 3.0904048082686655e-06, "loss": 0.4006, "step": 2446 }, { "epoch": 1.9878147847278635, "grad_norm": 0.33709096908569336, "learning_rate": 3.0860369492737753e-06, "loss": 0.4462, "step": 2447 }, { "epoch": 1.9886271324126725, "grad_norm": 0.39353013038635254, "learning_rate": 3.0816708009738605e-06, "loss": 0.4155, "step": 2448 }, { "epoch": 1.9894394800974817, "grad_norm": 0.3257770836353302, "learning_rate": 3.0773063672713725e-06, "loss": 0.4332, "step": 2449 }, { "epoch": 1.990251827782291, "grad_norm": 0.3447599709033966, "learning_rate": 3.0729436520672318e-06, "loss": 0.4286, "step": 2450 }, { "epoch": 1.9910641754671, "grad_norm": 0.3332153856754303, "learning_rate": 3.0685826592608193e-06, "loss": 0.3994, "step": 2451 }, { "epoch": 1.9918765231519089, "grad_norm": 0.37756580114364624, "learning_rate": 3.0642233927499786e-06, "loss": 0.4347, "step": 2452 }, { "epoch": 1.992688870836718, "grad_norm": 0.34123897552490234, "learning_rate": 3.0598658564310122e-06, "loss": 0.423, "step": 2453 }, { "epoch": 1.9935012185215273, "grad_norm": 0.3321297764778137, "learning_rate": 3.055510054198675e-06, "loss": 0.4023, "step": 2454 }, { "epoch": 1.9943135662063363, "grad_norm": 0.321706086397171, "learning_rate": 3.0511559899461684e-06, "loss": 0.3883, "step": 2455 }, { "epoch": 1.9951259138911455, "grad_norm": 0.342475026845932, "learning_rate": 3.0468036675651447e-06, "loss": 0.4541, "step": 2456 }, { "epoch": 1.9959382615759544, "grad_norm": 0.35376742482185364, "learning_rate": 3.0424530909456974e-06, "loss": 0.4282, "step": 2457 }, { "epoch": 1.9967506092607636, "grad_norm": 0.3412845730781555, "learning_rate": 3.03810426397636e-06, "loss": 0.4284, "step": 2458 }, { "epoch": 1.9975629569455728, "grad_norm": 0.34521716833114624, "learning_rate": 3.033757190544102e-06, "loss": 0.4103, "step": 2459 }, { "epoch": 1.9983753046303818, "grad_norm": 0.33887413144111633, "learning_rate": 3.029411874534327e-06, "loss": 0.4115, "step": 2460 }, { "epoch": 1.9991876523151908, "grad_norm": 0.3351510167121887, "learning_rate": 3.0250683198308656e-06, "loss": 0.4242, "step": 2461 }, { "epoch": 2.0, "grad_norm": 0.3759629428386688, "learning_rate": 3.020726530315975e-06, "loss": 0.3929, "step": 2462 }, { "epoch": 2.000812347684809, "grad_norm": 0.37275058031082153, "learning_rate": 3.016386509870335e-06, "loss": 0.3906, "step": 2463 }, { "epoch": 2.0016246953696184, "grad_norm": 0.35512498021125793, "learning_rate": 3.0120482623730452e-06, "loss": 0.4208, "step": 2464 }, { "epoch": 2.002437043054427, "grad_norm": 0.37265199422836304, "learning_rate": 3.0077117917016187e-06, "loss": 0.3968, "step": 2465 }, { "epoch": 2.0032493907392364, "grad_norm": 0.36615651845932007, "learning_rate": 3.00337710173198e-06, "loss": 0.4127, "step": 2466 }, { "epoch": 2.0040617384240456, "grad_norm": 0.35998669266700745, "learning_rate": 2.9990441963384657e-06, "loss": 0.3923, "step": 2467 }, { "epoch": 2.0048740861088548, "grad_norm": 0.3655489683151245, "learning_rate": 2.9947130793938127e-06, "loss": 0.3979, "step": 2468 }, { "epoch": 2.0056864337936635, "grad_norm": 0.37442266941070557, "learning_rate": 2.9903837547691634e-06, "loss": 0.4253, "step": 2469 }, { "epoch": 2.0064987814784727, "grad_norm": 0.3869323432445526, "learning_rate": 2.986056226334054e-06, "loss": 0.4053, "step": 2470 }, { "epoch": 2.007311129163282, "grad_norm": 0.36181092262268066, "learning_rate": 2.9817304979564206e-06, "loss": 0.3967, "step": 2471 }, { "epoch": 2.008123476848091, "grad_norm": 0.3649786710739136, "learning_rate": 2.977406573502585e-06, "loss": 0.4441, "step": 2472 }, { "epoch": 2.0089358245329, "grad_norm": 0.3354089558124542, "learning_rate": 2.9730844568372598e-06, "loss": 0.3698, "step": 2473 }, { "epoch": 2.009748172217709, "grad_norm": 0.32851335406303406, "learning_rate": 2.968764151823542e-06, "loss": 0.4008, "step": 2474 }, { "epoch": 2.0105605199025183, "grad_norm": 0.3158995807170868, "learning_rate": 2.964445662322908e-06, "loss": 0.4085, "step": 2475 }, { "epoch": 2.0113728675873275, "grad_norm": 0.3282347023487091, "learning_rate": 2.960128992195211e-06, "loss": 0.3862, "step": 2476 }, { "epoch": 2.0121852152721367, "grad_norm": 0.3314456045627594, "learning_rate": 2.95581414529868e-06, "loss": 0.4031, "step": 2477 }, { "epoch": 2.0129975629569454, "grad_norm": 0.31788042187690735, "learning_rate": 2.951501125489914e-06, "loss": 0.4089, "step": 2478 }, { "epoch": 2.0138099106417546, "grad_norm": 0.3175959587097168, "learning_rate": 2.9471899366238777e-06, "loss": 0.3946, "step": 2479 }, { "epoch": 2.014622258326564, "grad_norm": 0.3431922495365143, "learning_rate": 2.9428805825538993e-06, "loss": 0.4162, "step": 2480 }, { "epoch": 2.015434606011373, "grad_norm": 0.3424859941005707, "learning_rate": 2.9385730671316687e-06, "loss": 0.381, "step": 2481 }, { "epoch": 2.016246953696182, "grad_norm": 0.360156774520874, "learning_rate": 2.934267394207231e-06, "loss": 0.3701, "step": 2482 }, { "epoch": 2.017059301380991, "grad_norm": 0.3426051735877991, "learning_rate": 2.929963567628985e-06, "loss": 0.4039, "step": 2483 }, { "epoch": 2.0178716490658, "grad_norm": 0.3281899392604828, "learning_rate": 2.9256615912436803e-06, "loss": 0.4044, "step": 2484 }, { "epoch": 2.0186839967506094, "grad_norm": 0.3540048599243164, "learning_rate": 2.9213614688964104e-06, "loss": 0.4303, "step": 2485 }, { "epoch": 2.019496344435418, "grad_norm": 0.36445194482803345, "learning_rate": 2.9170632044306137e-06, "loss": 0.4037, "step": 2486 }, { "epoch": 2.0203086921202273, "grad_norm": 0.3550053834915161, "learning_rate": 2.9127668016880674e-06, "loss": 0.3731, "step": 2487 }, { "epoch": 2.0211210398050365, "grad_norm": 0.4882868528366089, "learning_rate": 2.9084722645088825e-06, "loss": 0.4195, "step": 2488 }, { "epoch": 2.0219333874898457, "grad_norm": 0.33590832352638245, "learning_rate": 2.904179596731509e-06, "loss": 0.3859, "step": 2489 }, { "epoch": 2.022745735174655, "grad_norm": 0.342136412858963, "learning_rate": 2.8998888021927146e-06, "loss": 0.3811, "step": 2490 }, { "epoch": 2.0235580828594637, "grad_norm": 0.34826958179473877, "learning_rate": 2.895599884727608e-06, "loss": 0.3931, "step": 2491 }, { "epoch": 2.024370430544273, "grad_norm": 0.3417625427246094, "learning_rate": 2.8913128481696067e-06, "loss": 0.3932, "step": 2492 }, { "epoch": 2.025182778229082, "grad_norm": 0.3506479263305664, "learning_rate": 2.8870276963504562e-06, "loss": 0.3873, "step": 2493 }, { "epoch": 2.0259951259138913, "grad_norm": 0.35199618339538574, "learning_rate": 2.8827444331002087e-06, "loss": 0.4139, "step": 2494 }, { "epoch": 2.0268074735987, "grad_norm": 0.32923510670661926, "learning_rate": 2.8784630622472376e-06, "loss": 0.4102, "step": 2495 }, { "epoch": 2.0276198212835093, "grad_norm": 0.3562665283679962, "learning_rate": 2.8741835876182165e-06, "loss": 0.3878, "step": 2496 }, { "epoch": 2.0284321689683185, "grad_norm": 0.35663479566574097, "learning_rate": 2.8699060130381334e-06, "loss": 0.383, "step": 2497 }, { "epoch": 2.0292445166531277, "grad_norm": 0.3803117871284485, "learning_rate": 2.8656303423302678e-06, "loss": 0.424, "step": 2498 }, { "epoch": 2.0300568643379364, "grad_norm": 0.3355507552623749, "learning_rate": 2.861356579316204e-06, "loss": 0.3857, "step": 2499 }, { "epoch": 2.0308692120227456, "grad_norm": 0.35570433735847473, "learning_rate": 2.857084727815823e-06, "loss": 0.4119, "step": 2500 }, { "epoch": 2.031681559707555, "grad_norm": 0.3617430627346039, "learning_rate": 2.8528147916472895e-06, "loss": 0.3639, "step": 2501 }, { "epoch": 2.032493907392364, "grad_norm": 0.3566562831401825, "learning_rate": 2.848546774627064e-06, "loss": 0.3982, "step": 2502 }, { "epoch": 2.033306255077173, "grad_norm": 0.3602437674999237, "learning_rate": 2.8442806805698857e-06, "loss": 0.4049, "step": 2503 }, { "epoch": 2.034118602761982, "grad_norm": 0.36504390835762024, "learning_rate": 2.840016513288782e-06, "loss": 0.3957, "step": 2504 }, { "epoch": 2.034930950446791, "grad_norm": 0.36949270963668823, "learning_rate": 2.8357542765950493e-06, "loss": 0.3996, "step": 2505 }, { "epoch": 2.0357432981316004, "grad_norm": 0.389306902885437, "learning_rate": 2.8314939742982673e-06, "loss": 0.382, "step": 2506 }, { "epoch": 2.0365556458164096, "grad_norm": 0.38414281606674194, "learning_rate": 2.8272356102062792e-06, "loss": 0.4178, "step": 2507 }, { "epoch": 2.0373679935012183, "grad_norm": 0.41082218289375305, "learning_rate": 2.8229791881252023e-06, "loss": 0.396, "step": 2508 }, { "epoch": 2.0381803411860275, "grad_norm": 0.350232869386673, "learning_rate": 2.818724711859412e-06, "loss": 0.3792, "step": 2509 }, { "epoch": 2.0389926888708367, "grad_norm": 0.39827266335487366, "learning_rate": 2.814472185211551e-06, "loss": 0.4149, "step": 2510 }, { "epoch": 2.039805036555646, "grad_norm": 0.3866232931613922, "learning_rate": 2.810221611982512e-06, "loss": 0.4002, "step": 2511 }, { "epoch": 2.0406173842404547, "grad_norm": 0.4016122817993164, "learning_rate": 2.805972995971451e-06, "loss": 0.3982, "step": 2512 }, { "epoch": 2.041429731925264, "grad_norm": 0.318081259727478, "learning_rate": 2.8017263409757646e-06, "loss": 0.4326, "step": 2513 }, { "epoch": 2.042242079610073, "grad_norm": 0.3434903621673584, "learning_rate": 2.7974816507911047e-06, "loss": 0.3997, "step": 2514 }, { "epoch": 2.0430544272948823, "grad_norm": 0.36293721199035645, "learning_rate": 2.7932389292113604e-06, "loss": 0.4014, "step": 2515 }, { "epoch": 2.0438667749796915, "grad_norm": 0.3637715280056, "learning_rate": 2.788998180028668e-06, "loss": 0.4073, "step": 2516 }, { "epoch": 2.0446791226645002, "grad_norm": 0.32699674367904663, "learning_rate": 2.784759407033398e-06, "loss": 0.4187, "step": 2517 }, { "epoch": 2.0454914703493094, "grad_norm": 0.3277100920677185, "learning_rate": 2.7805226140141505e-06, "loss": 0.4284, "step": 2518 }, { "epoch": 2.0463038180341186, "grad_norm": 0.37753957509994507, "learning_rate": 2.776287804757765e-06, "loss": 0.4221, "step": 2519 }, { "epoch": 2.047116165718928, "grad_norm": 0.36457738280296326, "learning_rate": 2.772054983049296e-06, "loss": 0.4122, "step": 2520 }, { "epoch": 2.0479285134037366, "grad_norm": 0.359887033700943, "learning_rate": 2.7678241526720338e-06, "loss": 0.3914, "step": 2521 }, { "epoch": 2.048740861088546, "grad_norm": 0.3460420072078705, "learning_rate": 2.7635953174074787e-06, "loss": 0.4061, "step": 2522 }, { "epoch": 2.049553208773355, "grad_norm": 0.35709989070892334, "learning_rate": 2.759368481035356e-06, "loss": 0.4394, "step": 2523 }, { "epoch": 2.050365556458164, "grad_norm": 0.3503166735172272, "learning_rate": 2.755143647333597e-06, "loss": 0.398, "step": 2524 }, { "epoch": 2.0511779041429734, "grad_norm": 0.3550649881362915, "learning_rate": 2.7509208200783505e-06, "loss": 0.4104, "step": 2525 }, { "epoch": 2.051990251827782, "grad_norm": 0.36331310868263245, "learning_rate": 2.746700003043964e-06, "loss": 0.4108, "step": 2526 }, { "epoch": 2.0528025995125914, "grad_norm": 0.33469629287719727, "learning_rate": 2.7424812000029955e-06, "loss": 0.3966, "step": 2527 }, { "epoch": 2.0536149471974006, "grad_norm": 0.37598690390586853, "learning_rate": 2.738264414726196e-06, "loss": 0.4214, "step": 2528 }, { "epoch": 2.0544272948822098, "grad_norm": 0.3445940911769867, "learning_rate": 2.73404965098252e-06, "loss": 0.4105, "step": 2529 }, { "epoch": 2.0552396425670185, "grad_norm": 0.35055288672447205, "learning_rate": 2.7298369125391112e-06, "loss": 0.4134, "step": 2530 }, { "epoch": 2.0560519902518277, "grad_norm": 0.3420060873031616, "learning_rate": 2.7256262031613044e-06, "loss": 0.4014, "step": 2531 }, { "epoch": 2.056864337936637, "grad_norm": 0.35934704542160034, "learning_rate": 2.7214175266126184e-06, "loss": 0.4269, "step": 2532 }, { "epoch": 2.057676685621446, "grad_norm": 0.3491143584251404, "learning_rate": 2.7172108866547593e-06, "loss": 0.3852, "step": 2533 }, { "epoch": 2.058489033306255, "grad_norm": 0.3483504354953766, "learning_rate": 2.7130062870476072e-06, "loss": 0.3935, "step": 2534 }, { "epoch": 2.059301380991064, "grad_norm": 0.3194550573825836, "learning_rate": 2.7088037315492254e-06, "loss": 0.4123, "step": 2535 }, { "epoch": 2.0601137286758733, "grad_norm": 0.3074377477169037, "learning_rate": 2.7046032239158428e-06, "loss": 0.4043, "step": 2536 }, { "epoch": 2.0609260763606825, "grad_norm": 0.35129860043525696, "learning_rate": 2.7004047679018652e-06, "loss": 0.4046, "step": 2537 }, { "epoch": 2.0617384240454917, "grad_norm": 0.3560556173324585, "learning_rate": 2.6962083672598584e-06, "loss": 0.4036, "step": 2538 }, { "epoch": 2.0625507717303004, "grad_norm": 0.325373113155365, "learning_rate": 2.692014025740557e-06, "loss": 0.3886, "step": 2539 }, { "epoch": 2.0633631194151096, "grad_norm": 0.3447774648666382, "learning_rate": 2.687821747092848e-06, "loss": 0.4258, "step": 2540 }, { "epoch": 2.064175467099919, "grad_norm": 0.36525994539260864, "learning_rate": 2.683631535063783e-06, "loss": 0.4286, "step": 2541 }, { "epoch": 2.064987814784728, "grad_norm": 0.31490710377693176, "learning_rate": 2.679443393398558e-06, "loss": 0.3952, "step": 2542 }, { "epoch": 2.065800162469537, "grad_norm": 0.341650128364563, "learning_rate": 2.6752573258405245e-06, "loss": 0.4158, "step": 2543 }, { "epoch": 2.066612510154346, "grad_norm": 0.3392527103424072, "learning_rate": 2.6710733361311812e-06, "loss": 0.4078, "step": 2544 }, { "epoch": 2.067424857839155, "grad_norm": 0.32938235998153687, "learning_rate": 2.6668914280101615e-06, "loss": 0.3928, "step": 2545 }, { "epoch": 2.0682372055239644, "grad_norm": 0.3348221778869629, "learning_rate": 2.662711605215248e-06, "loss": 0.3892, "step": 2546 }, { "epoch": 2.069049553208773, "grad_norm": 0.33693021535873413, "learning_rate": 2.658533871482351e-06, "loss": 0.385, "step": 2547 }, { "epoch": 2.0698619008935824, "grad_norm": 0.3183956742286682, "learning_rate": 2.6543582305455217e-06, "loss": 0.3868, "step": 2548 }, { "epoch": 2.0706742485783916, "grad_norm": 0.3592449426651001, "learning_rate": 2.650184686136932e-06, "loss": 0.3918, "step": 2549 }, { "epoch": 2.0714865962632008, "grad_norm": 0.3549386262893677, "learning_rate": 2.6460132419868893e-06, "loss": 0.4251, "step": 2550 }, { "epoch": 2.07229894394801, "grad_norm": 0.35092395544052124, "learning_rate": 2.641843901823815e-06, "loss": 0.4007, "step": 2551 }, { "epoch": 2.0731112916328187, "grad_norm": 0.33407121896743774, "learning_rate": 2.637676669374258e-06, "loss": 0.4056, "step": 2552 }, { "epoch": 2.073923639317628, "grad_norm": 0.3169122338294983, "learning_rate": 2.6335115483628747e-06, "loss": 0.4161, "step": 2553 }, { "epoch": 2.074735987002437, "grad_norm": 0.32533520460128784, "learning_rate": 2.629348542512443e-06, "loss": 0.4329, "step": 2554 }, { "epoch": 2.0755483346872463, "grad_norm": 0.3764422535896301, "learning_rate": 2.625187655543844e-06, "loss": 0.3926, "step": 2555 }, { "epoch": 2.076360682372055, "grad_norm": 0.3179590702056885, "learning_rate": 2.6210288911760684e-06, "loss": 0.3888, "step": 2556 }, { "epoch": 2.0771730300568643, "grad_norm": 0.3587246239185333, "learning_rate": 2.616872253126207e-06, "loss": 0.4125, "step": 2557 }, { "epoch": 2.0779853777416735, "grad_norm": 0.3080962598323822, "learning_rate": 2.612717745109453e-06, "loss": 0.3866, "step": 2558 }, { "epoch": 2.0787977254264827, "grad_norm": 0.32065069675445557, "learning_rate": 2.6085653708390936e-06, "loss": 0.4128, "step": 2559 }, { "epoch": 2.0796100731112914, "grad_norm": 0.3520248532295227, "learning_rate": 2.604415134026509e-06, "loss": 0.4161, "step": 2560 }, { "epoch": 2.0804224207961006, "grad_norm": 0.3362104594707489, "learning_rate": 2.6002670383811724e-06, "loss": 0.3754, "step": 2561 }, { "epoch": 2.08123476848091, "grad_norm": 0.34961214661598206, "learning_rate": 2.596121087610636e-06, "loss": 0.3863, "step": 2562 }, { "epoch": 2.082047116165719, "grad_norm": 0.34757012128829956, "learning_rate": 2.591977285420545e-06, "loss": 0.397, "step": 2563 }, { "epoch": 2.0828594638505282, "grad_norm": 0.33569034934043884, "learning_rate": 2.587835635514613e-06, "loss": 0.4142, "step": 2564 }, { "epoch": 2.083671811535337, "grad_norm": 0.30626487731933594, "learning_rate": 2.5836961415946394e-06, "loss": 0.381, "step": 2565 }, { "epoch": 2.084484159220146, "grad_norm": 0.3378657102584839, "learning_rate": 2.579558807360489e-06, "loss": 0.4306, "step": 2566 }, { "epoch": 2.0852965069049554, "grad_norm": 0.34254515171051025, "learning_rate": 2.5754236365101024e-06, "loss": 0.4136, "step": 2567 }, { "epoch": 2.0861088545897646, "grad_norm": 0.3199291229248047, "learning_rate": 2.5712906327394836e-06, "loss": 0.4039, "step": 2568 }, { "epoch": 2.0869212022745733, "grad_norm": 0.3374512493610382, "learning_rate": 2.5671597997427023e-06, "loss": 0.3989, "step": 2569 }, { "epoch": 2.0877335499593825, "grad_norm": 0.34968093037605286, "learning_rate": 2.5630311412118815e-06, "loss": 0.3965, "step": 2570 }, { "epoch": 2.0885458976441917, "grad_norm": 0.3515475392341614, "learning_rate": 2.558904660837209e-06, "loss": 0.418, "step": 2571 }, { "epoch": 2.089358245329001, "grad_norm": 0.36579856276512146, "learning_rate": 2.554780362306918e-06, "loss": 0.3863, "step": 2572 }, { "epoch": 2.0901705930138097, "grad_norm": 0.32180267572402954, "learning_rate": 2.5506582493072994e-06, "loss": 0.3967, "step": 2573 }, { "epoch": 2.090982940698619, "grad_norm": 0.341371089220047, "learning_rate": 2.546538325522682e-06, "loss": 0.4066, "step": 2574 }, { "epoch": 2.091795288383428, "grad_norm": 0.35673025250434875, "learning_rate": 2.5424205946354475e-06, "loss": 0.4109, "step": 2575 }, { "epoch": 2.0926076360682373, "grad_norm": 0.3233785033226013, "learning_rate": 2.5383050603260074e-06, "loss": 0.4198, "step": 2576 }, { "epoch": 2.0934199837530465, "grad_norm": 0.3371371030807495, "learning_rate": 2.53419172627282e-06, "loss": 0.4039, "step": 2577 }, { "epoch": 2.0942323314378553, "grad_norm": 0.33923399448394775, "learning_rate": 2.5300805961523677e-06, "loss": 0.4435, "step": 2578 }, { "epoch": 2.0950446791226645, "grad_norm": 0.3154345452785492, "learning_rate": 2.525971673639172e-06, "loss": 0.4033, "step": 2579 }, { "epoch": 2.0958570268074737, "grad_norm": 0.36843085289001465, "learning_rate": 2.5218649624057717e-06, "loss": 0.3958, "step": 2580 }, { "epoch": 2.096669374492283, "grad_norm": 0.3461039662361145, "learning_rate": 2.517760466122739e-06, "loss": 0.3969, "step": 2581 }, { "epoch": 2.0974817221770916, "grad_norm": 0.3417900502681732, "learning_rate": 2.513658188458657e-06, "loss": 0.4303, "step": 2582 }, { "epoch": 2.098294069861901, "grad_norm": 0.3390774726867676, "learning_rate": 2.5095581330801345e-06, "loss": 0.3829, "step": 2583 }, { "epoch": 2.09910641754671, "grad_norm": 0.2999167740345001, "learning_rate": 2.5054603036517856e-06, "loss": 0.4308, "step": 2584 }, { "epoch": 2.099918765231519, "grad_norm": 0.3629439175128937, "learning_rate": 2.5013647038362437e-06, "loss": 0.3945, "step": 2585 }, { "epoch": 2.1007311129163284, "grad_norm": 0.3774712085723877, "learning_rate": 2.4972713372941406e-06, "loss": 0.3903, "step": 2586 }, { "epoch": 2.101543460601137, "grad_norm": 0.335875004529953, "learning_rate": 2.4931802076841176e-06, "loss": 0.3855, "step": 2587 }, { "epoch": 2.1023558082859464, "grad_norm": 0.31467002630233765, "learning_rate": 2.4890913186628178e-06, "loss": 0.3982, "step": 2588 }, { "epoch": 2.1031681559707556, "grad_norm": 0.3537750244140625, "learning_rate": 2.4850046738848752e-06, "loss": 0.3945, "step": 2589 }, { "epoch": 2.1039805036555648, "grad_norm": 0.32311877608299255, "learning_rate": 2.480920277002926e-06, "loss": 0.4034, "step": 2590 }, { "epoch": 2.1047928513403735, "grad_norm": 0.36299434304237366, "learning_rate": 2.4768381316675883e-06, "loss": 0.4245, "step": 2591 }, { "epoch": 2.1056051990251827, "grad_norm": 0.31940141320228577, "learning_rate": 2.472758241527477e-06, "loss": 0.3992, "step": 2592 }, { "epoch": 2.106417546709992, "grad_norm": 0.3453502058982849, "learning_rate": 2.468680610229182e-06, "loss": 0.4126, "step": 2593 }, { "epoch": 2.107229894394801, "grad_norm": 0.34232252836227417, "learning_rate": 2.4646052414172844e-06, "loss": 0.4117, "step": 2594 }, { "epoch": 2.10804224207961, "grad_norm": 0.3541296720504761, "learning_rate": 2.4605321387343334e-06, "loss": 0.3831, "step": 2595 }, { "epoch": 2.108854589764419, "grad_norm": 0.3366367220878601, "learning_rate": 2.4564613058208603e-06, "loss": 0.3583, "step": 2596 }, { "epoch": 2.1096669374492283, "grad_norm": 0.32279130816459656, "learning_rate": 2.452392746315361e-06, "loss": 0.3938, "step": 2597 }, { "epoch": 2.1104792851340375, "grad_norm": 0.3366466164588928, "learning_rate": 2.4483264638543076e-06, "loss": 0.4179, "step": 2598 }, { "epoch": 2.1112916328188467, "grad_norm": 0.3568834066390991, "learning_rate": 2.4442624620721274e-06, "loss": 0.3786, "step": 2599 }, { "epoch": 2.1121039805036554, "grad_norm": 0.3421010971069336, "learning_rate": 2.440200744601218e-06, "loss": 0.3889, "step": 2600 }, { "epoch": 2.1129163281884646, "grad_norm": 0.3540026545524597, "learning_rate": 2.4361413150719287e-06, "loss": 0.3964, "step": 2601 }, { "epoch": 2.113728675873274, "grad_norm": 0.3541000187397003, "learning_rate": 2.432084177112569e-06, "loss": 0.4068, "step": 2602 }, { "epoch": 2.114541023558083, "grad_norm": 0.32188883423805237, "learning_rate": 2.4280293343493954e-06, "loss": 0.3727, "step": 2603 }, { "epoch": 2.115353371242892, "grad_norm": 0.32976657152175903, "learning_rate": 2.423976790406616e-06, "loss": 0.3994, "step": 2604 }, { "epoch": 2.116165718927701, "grad_norm": 0.34797734022140503, "learning_rate": 2.4199265489063846e-06, "loss": 0.3936, "step": 2605 }, { "epoch": 2.11697806661251, "grad_norm": 0.31141889095306396, "learning_rate": 2.4158786134687966e-06, "loss": 0.381, "step": 2606 }, { "epoch": 2.1177904142973194, "grad_norm": 0.3162814974784851, "learning_rate": 2.4118329877118836e-06, "loss": 0.3794, "step": 2607 }, { "epoch": 2.118602761982128, "grad_norm": 0.3202272653579712, "learning_rate": 2.407789675251617e-06, "loss": 0.4062, "step": 2608 }, { "epoch": 2.1194151096669374, "grad_norm": 0.34154412150382996, "learning_rate": 2.403748679701896e-06, "loss": 0.3905, "step": 2609 }, { "epoch": 2.1202274573517466, "grad_norm": 0.33592063188552856, "learning_rate": 2.3997100046745554e-06, "loss": 0.3977, "step": 2610 }, { "epoch": 2.1210398050365558, "grad_norm": 0.3617228865623474, "learning_rate": 2.3956736537793468e-06, "loss": 0.3884, "step": 2611 }, { "epoch": 2.121852152721365, "grad_norm": 0.3250168263912201, "learning_rate": 2.3916396306239525e-06, "loss": 0.3723, "step": 2612 }, { "epoch": 2.1226645004061737, "grad_norm": 0.3317708969116211, "learning_rate": 2.3876079388139733e-06, "loss": 0.4174, "step": 2613 }, { "epoch": 2.123476848090983, "grad_norm": 0.3358134329319, "learning_rate": 2.38357858195292e-06, "loss": 0.3935, "step": 2614 }, { "epoch": 2.124289195775792, "grad_norm": 0.3573833703994751, "learning_rate": 2.3795515636422257e-06, "loss": 0.391, "step": 2615 }, { "epoch": 2.1251015434606013, "grad_norm": 0.40758025646209717, "learning_rate": 2.3755268874812237e-06, "loss": 0.4152, "step": 2616 }, { "epoch": 2.12591389114541, "grad_norm": 0.368748277425766, "learning_rate": 2.371504557067163e-06, "loss": 0.3944, "step": 2617 }, { "epoch": 2.1267262388302193, "grad_norm": 0.32097509503364563, "learning_rate": 2.367484575995187e-06, "loss": 0.4161, "step": 2618 }, { "epoch": 2.1275385865150285, "grad_norm": 0.33125802874565125, "learning_rate": 2.3634669478583484e-06, "loss": 0.4017, "step": 2619 }, { "epoch": 2.1283509341998377, "grad_norm": 0.353182852268219, "learning_rate": 2.3594516762475882e-06, "loss": 0.3903, "step": 2620 }, { "epoch": 2.1291632818846464, "grad_norm": 0.3532385230064392, "learning_rate": 2.3554387647517497e-06, "loss": 0.4184, "step": 2621 }, { "epoch": 2.1299756295694556, "grad_norm": 0.35239508748054504, "learning_rate": 2.3514282169575575e-06, "loss": 0.4142, "step": 2622 }, { "epoch": 2.130787977254265, "grad_norm": 0.32931041717529297, "learning_rate": 2.3474200364496326e-06, "loss": 0.3788, "step": 2623 }, { "epoch": 2.131600324939074, "grad_norm": 0.3142644166946411, "learning_rate": 2.343414226810472e-06, "loss": 0.4008, "step": 2624 }, { "epoch": 2.1324126726238832, "grad_norm": 0.31616324186325073, "learning_rate": 2.3394107916204617e-06, "loss": 0.3892, "step": 2625 }, { "epoch": 2.133225020308692, "grad_norm": 0.3371962904930115, "learning_rate": 2.3354097344578565e-06, "loss": 0.4061, "step": 2626 }, { "epoch": 2.134037367993501, "grad_norm": 0.3511751592159271, "learning_rate": 2.3314110588987946e-06, "loss": 0.3969, "step": 2627 }, { "epoch": 2.1348497156783104, "grad_norm": 0.3400145173072815, "learning_rate": 2.327414768517278e-06, "loss": 0.4184, "step": 2628 }, { "epoch": 2.1356620633631196, "grad_norm": 0.3161943554878235, "learning_rate": 2.3234208668851843e-06, "loss": 0.3861, "step": 2629 }, { "epoch": 2.1364744110479283, "grad_norm": 0.32528582215309143, "learning_rate": 2.3194293575722464e-06, "loss": 0.3825, "step": 2630 }, { "epoch": 2.1372867587327375, "grad_norm": 0.3669734001159668, "learning_rate": 2.3154402441460685e-06, "loss": 0.3874, "step": 2631 }, { "epoch": 2.1380991064175467, "grad_norm": 0.35076814889907837, "learning_rate": 2.311453530172109e-06, "loss": 0.3965, "step": 2632 }, { "epoch": 2.138911454102356, "grad_norm": 0.3505169153213501, "learning_rate": 2.3074692192136803e-06, "loss": 0.3897, "step": 2633 }, { "epoch": 2.1397238017871647, "grad_norm": 0.32077556848526, "learning_rate": 2.303487314831952e-06, "loss": 0.4072, "step": 2634 }, { "epoch": 2.140536149471974, "grad_norm": 0.34089285135269165, "learning_rate": 2.2995078205859344e-06, "loss": 0.3966, "step": 2635 }, { "epoch": 2.141348497156783, "grad_norm": 0.35310181975364685, "learning_rate": 2.295530740032494e-06, "loss": 0.3983, "step": 2636 }, { "epoch": 2.1421608448415923, "grad_norm": 0.33642950654029846, "learning_rate": 2.2915560767263296e-06, "loss": 0.3922, "step": 2637 }, { "epoch": 2.1429731925264015, "grad_norm": 0.3630782663822174, "learning_rate": 2.287583834219989e-06, "loss": 0.3899, "step": 2638 }, { "epoch": 2.1437855402112103, "grad_norm": 0.35471245646476746, "learning_rate": 2.2836140160638477e-06, "loss": 0.4193, "step": 2639 }, { "epoch": 2.1445978878960195, "grad_norm": 0.31771957874298096, "learning_rate": 2.2796466258061223e-06, "loss": 0.3901, "step": 2640 }, { "epoch": 2.1454102355808287, "grad_norm": 0.33326801657676697, "learning_rate": 2.2756816669928516e-06, "loss": 0.422, "step": 2641 }, { "epoch": 2.146222583265638, "grad_norm": 0.40002575516700745, "learning_rate": 2.2717191431679085e-06, "loss": 0.3901, "step": 2642 }, { "epoch": 2.1470349309504466, "grad_norm": 0.3248981833457947, "learning_rate": 2.2677590578729815e-06, "loss": 0.4017, "step": 2643 }, { "epoch": 2.147847278635256, "grad_norm": 0.3326977491378784, "learning_rate": 2.2638014146475894e-06, "loss": 0.3747, "step": 2644 }, { "epoch": 2.148659626320065, "grad_norm": 0.34488189220428467, "learning_rate": 2.2598462170290598e-06, "loss": 0.3993, "step": 2645 }, { "epoch": 2.149471974004874, "grad_norm": 0.3831392824649811, "learning_rate": 2.25589346855254e-06, "loss": 0.4341, "step": 2646 }, { "epoch": 2.150284321689683, "grad_norm": 0.39696526527404785, "learning_rate": 2.2519431727509823e-06, "loss": 0.4183, "step": 2647 }, { "epoch": 2.151096669374492, "grad_norm": 0.36571231484413147, "learning_rate": 2.2479953331551534e-06, "loss": 0.3673, "step": 2648 }, { "epoch": 2.1519090170593014, "grad_norm": 0.34326714277267456, "learning_rate": 2.2440499532936196e-06, "loss": 0.3693, "step": 2649 }, { "epoch": 2.1527213647441106, "grad_norm": 0.3607809841632843, "learning_rate": 2.240107036692753e-06, "loss": 0.3981, "step": 2650 }, { "epoch": 2.1535337124289198, "grad_norm": 0.34476950764656067, "learning_rate": 2.2361665868767185e-06, "loss": 0.4034, "step": 2651 }, { "epoch": 2.1543460601137285, "grad_norm": 0.3237300217151642, "learning_rate": 2.2322286073674832e-06, "loss": 0.3954, "step": 2652 }, { "epoch": 2.1551584077985377, "grad_norm": 0.335808664560318, "learning_rate": 2.228293101684799e-06, "loss": 0.3802, "step": 2653 }, { "epoch": 2.155970755483347, "grad_norm": 0.34860533475875854, "learning_rate": 2.224360073346213e-06, "loss": 0.3917, "step": 2654 }, { "epoch": 2.156783103168156, "grad_norm": 0.4348866045475006, "learning_rate": 2.2204295258670524e-06, "loss": 0.4205, "step": 2655 }, { "epoch": 2.157595450852965, "grad_norm": 0.3416139483451843, "learning_rate": 2.2165014627604313e-06, "loss": 0.4036, "step": 2656 }, { "epoch": 2.158407798537774, "grad_norm": 0.35119765996932983, "learning_rate": 2.212575887537244e-06, "loss": 0.3671, "step": 2657 }, { "epoch": 2.1592201462225833, "grad_norm": 0.3909545838832855, "learning_rate": 2.208652803706155e-06, "loss": 0.3933, "step": 2658 }, { "epoch": 2.1600324939073925, "grad_norm": 0.34906521439552307, "learning_rate": 2.2047322147736096e-06, "loss": 0.3889, "step": 2659 }, { "epoch": 2.1608448415922012, "grad_norm": 0.35982081294059753, "learning_rate": 2.2008141242438165e-06, "loss": 0.4106, "step": 2660 }, { "epoch": 2.1616571892770104, "grad_norm": 0.30278459191322327, "learning_rate": 2.196898535618757e-06, "loss": 0.4092, "step": 2661 }, { "epoch": 2.1624695369618196, "grad_norm": 0.3254021406173706, "learning_rate": 2.192985452398171e-06, "loss": 0.3915, "step": 2662 }, { "epoch": 2.163281884646629, "grad_norm": 0.35983291268348694, "learning_rate": 2.1890748780795644e-06, "loss": 0.4068, "step": 2663 }, { "epoch": 2.164094232331438, "grad_norm": 0.3626376688480377, "learning_rate": 2.1851668161581946e-06, "loss": 0.3907, "step": 2664 }, { "epoch": 2.164906580016247, "grad_norm": 0.3495860993862152, "learning_rate": 2.181261270127081e-06, "loss": 0.391, "step": 2665 }, { "epoch": 2.165718927701056, "grad_norm": 0.34304195642471313, "learning_rate": 2.1773582434769854e-06, "loss": 0.4035, "step": 2666 }, { "epoch": 2.166531275385865, "grad_norm": 0.3414693772792816, "learning_rate": 2.1734577396964264e-06, "loss": 0.4191, "step": 2667 }, { "epoch": 2.1673436230706744, "grad_norm": 0.3514431118965149, "learning_rate": 2.1695597622716603e-06, "loss": 0.3985, "step": 2668 }, { "epoch": 2.168155970755483, "grad_norm": 0.339547723531723, "learning_rate": 2.165664314686692e-06, "loss": 0.3891, "step": 2669 }, { "epoch": 2.1689683184402924, "grad_norm": 0.33227211236953735, "learning_rate": 2.1617714004232588e-06, "loss": 0.3969, "step": 2670 }, { "epoch": 2.1697806661251016, "grad_norm": 0.3789689242839813, "learning_rate": 2.15788102296084e-06, "loss": 0.4145, "step": 2671 }, { "epoch": 2.1705930138099108, "grad_norm": 0.3467925190925598, "learning_rate": 2.153993185776641e-06, "loss": 0.3805, "step": 2672 }, { "epoch": 2.1714053614947195, "grad_norm": 0.3369256854057312, "learning_rate": 2.150107892345602e-06, "loss": 0.4054, "step": 2673 }, { "epoch": 2.1722177091795287, "grad_norm": 0.33034881949424744, "learning_rate": 2.14622514614039e-06, "loss": 0.3999, "step": 2674 }, { "epoch": 2.173030056864338, "grad_norm": 0.3506893515586853, "learning_rate": 2.142344950631388e-06, "loss": 0.3834, "step": 2675 }, { "epoch": 2.173842404549147, "grad_norm": 0.34080591797828674, "learning_rate": 2.138467309286709e-06, "loss": 0.4198, "step": 2676 }, { "epoch": 2.1746547522339563, "grad_norm": 0.3784041702747345, "learning_rate": 2.1345922255721745e-06, "loss": 0.414, "step": 2677 }, { "epoch": 2.175467099918765, "grad_norm": 0.3215140402317047, "learning_rate": 2.1307197029513276e-06, "loss": 0.3929, "step": 2678 }, { "epoch": 2.1762794476035743, "grad_norm": 0.30066385865211487, "learning_rate": 2.1268497448854143e-06, "loss": 0.4175, "step": 2679 }, { "epoch": 2.1770917952883835, "grad_norm": 0.3157821595668793, "learning_rate": 2.1229823548333965e-06, "loss": 0.4042, "step": 2680 }, { "epoch": 2.1779041429731927, "grad_norm": 0.3348449170589447, "learning_rate": 2.1191175362519316e-06, "loss": 0.3719, "step": 2681 }, { "epoch": 2.1787164906580014, "grad_norm": 0.3133604824542999, "learning_rate": 2.115255292595391e-06, "loss": 0.4241, "step": 2682 }, { "epoch": 2.1795288383428106, "grad_norm": 0.3183619976043701, "learning_rate": 2.111395627315832e-06, "loss": 0.3697, "step": 2683 }, { "epoch": 2.18034118602762, "grad_norm": 0.31030163168907166, "learning_rate": 2.1075385438630168e-06, "loss": 0.4129, "step": 2684 }, { "epoch": 2.181153533712429, "grad_norm": 0.344056099653244, "learning_rate": 2.103684045684393e-06, "loss": 0.394, "step": 2685 }, { "epoch": 2.181965881397238, "grad_norm": 0.3242654800415039, "learning_rate": 2.0998321362251036e-06, "loss": 0.3939, "step": 2686 }, { "epoch": 2.182778229082047, "grad_norm": 0.3309515416622162, "learning_rate": 2.0959828189279712e-06, "loss": 0.3784, "step": 2687 }, { "epoch": 2.183590576766856, "grad_norm": 0.3284779191017151, "learning_rate": 2.0921360972335086e-06, "loss": 0.3937, "step": 2688 }, { "epoch": 2.1844029244516654, "grad_norm": 0.3392973840236664, "learning_rate": 2.0882919745799023e-06, "loss": 0.4115, "step": 2689 }, { "epoch": 2.1852152721364746, "grad_norm": 0.3263111412525177, "learning_rate": 2.084450454403022e-06, "loss": 0.3816, "step": 2690 }, { "epoch": 2.1860276198212834, "grad_norm": 0.36181673407554626, "learning_rate": 2.0806115401364036e-06, "loss": 0.4053, "step": 2691 }, { "epoch": 2.1868399675060926, "grad_norm": 0.3331276476383209, "learning_rate": 2.076775235211263e-06, "loss": 0.4015, "step": 2692 }, { "epoch": 2.1876523151909018, "grad_norm": 0.327770859003067, "learning_rate": 2.0729415430564748e-06, "loss": 0.3986, "step": 2693 }, { "epoch": 2.188464662875711, "grad_norm": 0.3493017852306366, "learning_rate": 2.0691104670985868e-06, "loss": 0.3903, "step": 2694 }, { "epoch": 2.1892770105605197, "grad_norm": 0.3758143484592438, "learning_rate": 2.0652820107618e-06, "loss": 0.4085, "step": 2695 }, { "epoch": 2.190089358245329, "grad_norm": 0.33815985918045044, "learning_rate": 2.0614561774679826e-06, "loss": 0.438, "step": 2696 }, { "epoch": 2.190901705930138, "grad_norm": 0.31782689690589905, "learning_rate": 2.0576329706366494e-06, "loss": 0.4182, "step": 2697 }, { "epoch": 2.1917140536149473, "grad_norm": 0.326326847076416, "learning_rate": 2.053812393684977e-06, "loss": 0.4039, "step": 2698 }, { "epoch": 2.192526401299756, "grad_norm": 0.3149928152561188, "learning_rate": 2.0499944500277825e-06, "loss": 0.3806, "step": 2699 }, { "epoch": 2.1933387489845653, "grad_norm": 0.3367812931537628, "learning_rate": 2.046179143077535e-06, "loss": 0.3987, "step": 2700 }, { "epoch": 2.1941510966693745, "grad_norm": 0.315292626619339, "learning_rate": 2.0423664762443483e-06, "loss": 0.3969, "step": 2701 }, { "epoch": 2.1949634443541837, "grad_norm": 0.33816057443618774, "learning_rate": 2.0385564529359696e-06, "loss": 0.3915, "step": 2702 }, { "epoch": 2.195775792038993, "grad_norm": 0.3316093683242798, "learning_rate": 2.034749076557791e-06, "loss": 0.4106, "step": 2703 }, { "epoch": 2.1965881397238016, "grad_norm": 0.3233179450035095, "learning_rate": 2.030944350512831e-06, "loss": 0.4077, "step": 2704 }, { "epoch": 2.197400487408611, "grad_norm": 0.30853042006492615, "learning_rate": 2.027142278201748e-06, "loss": 0.4229, "step": 2705 }, { "epoch": 2.19821283509342, "grad_norm": 0.3246832489967346, "learning_rate": 2.023342863022819e-06, "loss": 0.4126, "step": 2706 }, { "epoch": 2.1990251827782292, "grad_norm": 0.31879469752311707, "learning_rate": 2.0195461083719557e-06, "loss": 0.4005, "step": 2707 }, { "epoch": 2.199837530463038, "grad_norm": 0.3051415979862213, "learning_rate": 2.015752017642684e-06, "loss": 0.4122, "step": 2708 }, { "epoch": 2.200649878147847, "grad_norm": 0.32532262802124023, "learning_rate": 2.0119605942261535e-06, "loss": 0.4085, "step": 2709 }, { "epoch": 2.2014622258326564, "grad_norm": 0.34003743529319763, "learning_rate": 2.0081718415111263e-06, "loss": 0.4089, "step": 2710 }, { "epoch": 2.2022745735174656, "grad_norm": 0.3162637948989868, "learning_rate": 2.004385762883982e-06, "loss": 0.4134, "step": 2711 }, { "epoch": 2.2030869212022743, "grad_norm": 0.3363330662250519, "learning_rate": 2.000602361728704e-06, "loss": 0.4037, "step": 2712 }, { "epoch": 2.2038992688870835, "grad_norm": 0.327670693397522, "learning_rate": 1.99682164142689e-06, "loss": 0.392, "step": 2713 }, { "epoch": 2.2047116165718927, "grad_norm": 0.330199658870697, "learning_rate": 1.993043605357733e-06, "loss": 0.4252, "step": 2714 }, { "epoch": 2.205523964256702, "grad_norm": 0.33020225167274475, "learning_rate": 1.989268256898036e-06, "loss": 0.4121, "step": 2715 }, { "epoch": 2.206336311941511, "grad_norm": 0.36260074377059937, "learning_rate": 1.985495599422191e-06, "loss": 0.419, "step": 2716 }, { "epoch": 2.20714865962632, "grad_norm": 0.3468407690525055, "learning_rate": 1.9817256363021907e-06, "loss": 0.4028, "step": 2717 }, { "epoch": 2.207961007311129, "grad_norm": 0.3226333260536194, "learning_rate": 1.9779583709076207e-06, "loss": 0.4084, "step": 2718 }, { "epoch": 2.2087733549959383, "grad_norm": 0.3413955569267273, "learning_rate": 1.9741938066056475e-06, "loss": 0.4165, "step": 2719 }, { "epoch": 2.2095857026807475, "grad_norm": 0.3342687487602234, "learning_rate": 1.970431946761031e-06, "loss": 0.4064, "step": 2720 }, { "epoch": 2.2103980503655563, "grad_norm": 0.3254616856575012, "learning_rate": 1.9666727947361135e-06, "loss": 0.3845, "step": 2721 }, { "epoch": 2.2112103980503655, "grad_norm": 0.3489094376564026, "learning_rate": 1.962916353890811e-06, "loss": 0.4097, "step": 2722 }, { "epoch": 2.2120227457351747, "grad_norm": 0.3229878544807434, "learning_rate": 1.959162627582623e-06, "loss": 0.3916, "step": 2723 }, { "epoch": 2.212835093419984, "grad_norm": 0.3158610761165619, "learning_rate": 1.955411619166617e-06, "loss": 0.3944, "step": 2724 }, { "epoch": 2.213647441104793, "grad_norm": 0.31530219316482544, "learning_rate": 1.951663331995436e-06, "loss": 0.4251, "step": 2725 }, { "epoch": 2.214459788789602, "grad_norm": 0.36239251494407654, "learning_rate": 1.94791776941929e-06, "loss": 0.4044, "step": 2726 }, { "epoch": 2.215272136474411, "grad_norm": 0.3219406008720398, "learning_rate": 1.9441749347859503e-06, "loss": 0.3886, "step": 2727 }, { "epoch": 2.21608448415922, "grad_norm": 0.32228830456733704, "learning_rate": 1.9404348314407544e-06, "loss": 0.3986, "step": 2728 }, { "epoch": 2.2168968318440294, "grad_norm": 0.3324989676475525, "learning_rate": 1.9366974627265945e-06, "loss": 0.4015, "step": 2729 }, { "epoch": 2.217709179528838, "grad_norm": 0.3568439781665802, "learning_rate": 1.9329628319839227e-06, "loss": 0.3999, "step": 2730 }, { "epoch": 2.2185215272136474, "grad_norm": 0.3526439070701599, "learning_rate": 1.929230942550739e-06, "loss": 0.4068, "step": 2731 }, { "epoch": 2.2193338748984566, "grad_norm": 0.3537921905517578, "learning_rate": 1.9255017977625995e-06, "loss": 0.386, "step": 2732 }, { "epoch": 2.2201462225832658, "grad_norm": 0.2885175943374634, "learning_rate": 1.9217754009526e-06, "loss": 0.3863, "step": 2733 }, { "epoch": 2.2209585702680745, "grad_norm": 0.32086998224258423, "learning_rate": 1.9180517554513884e-06, "loss": 0.3698, "step": 2734 }, { "epoch": 2.2217709179528837, "grad_norm": 0.3574517071247101, "learning_rate": 1.914330864587144e-06, "loss": 0.4136, "step": 2735 }, { "epoch": 2.222583265637693, "grad_norm": 0.33149924874305725, "learning_rate": 1.910612731685593e-06, "loss": 0.3827, "step": 2736 }, { "epoch": 2.223395613322502, "grad_norm": 0.3127197027206421, "learning_rate": 1.9068973600699887e-06, "loss": 0.3974, "step": 2737 }, { "epoch": 2.2242079610073113, "grad_norm": 0.33789676427841187, "learning_rate": 1.9031847530611237e-06, "loss": 0.3953, "step": 2738 }, { "epoch": 2.22502030869212, "grad_norm": 0.34830495715141296, "learning_rate": 1.8994749139773133e-06, "loss": 0.3929, "step": 2739 }, { "epoch": 2.2258326563769293, "grad_norm": 0.36229902505874634, "learning_rate": 1.8957678461344048e-06, "loss": 0.3801, "step": 2740 }, { "epoch": 2.2266450040617385, "grad_norm": 0.36385151743888855, "learning_rate": 1.892063552845761e-06, "loss": 0.4184, "step": 2741 }, { "epoch": 2.2274573517465477, "grad_norm": 0.32817378640174866, "learning_rate": 1.888362037422274e-06, "loss": 0.4123, "step": 2742 }, { "epoch": 2.2282696994313564, "grad_norm": 0.3219839930534363, "learning_rate": 1.8846633031723438e-06, "loss": 0.3881, "step": 2743 }, { "epoch": 2.2290820471161656, "grad_norm": 0.3373190462589264, "learning_rate": 1.8809673534018908e-06, "loss": 0.3743, "step": 2744 }, { "epoch": 2.229894394800975, "grad_norm": 0.3381701707839966, "learning_rate": 1.8772741914143472e-06, "loss": 0.4229, "step": 2745 }, { "epoch": 2.230706742485784, "grad_norm": 0.3930266499519348, "learning_rate": 1.873583820510647e-06, "loss": 0.3994, "step": 2746 }, { "epoch": 2.2315190901705932, "grad_norm": 0.3690527677536011, "learning_rate": 1.8698962439892383e-06, "loss": 0.3875, "step": 2747 }, { "epoch": 2.232331437855402, "grad_norm": 0.3383232653141022, "learning_rate": 1.8662114651460617e-06, "loss": 0.4038, "step": 2748 }, { "epoch": 2.233143785540211, "grad_norm": 0.3454052805900574, "learning_rate": 1.8625294872745675e-06, "loss": 0.3975, "step": 2749 }, { "epoch": 2.2339561332250204, "grad_norm": 0.33133938908576965, "learning_rate": 1.8588503136656938e-06, "loss": 0.4223, "step": 2750 }, { "epoch": 2.2347684809098296, "grad_norm": 0.34136414527893066, "learning_rate": 1.8551739476078784e-06, "loss": 0.4091, "step": 2751 }, { "epoch": 2.2355808285946384, "grad_norm": 0.3554058372974396, "learning_rate": 1.851500392387045e-06, "loss": 0.3746, "step": 2752 }, { "epoch": 2.2363931762794476, "grad_norm": 0.3658202588558197, "learning_rate": 1.8478296512866105e-06, "loss": 0.414, "step": 2753 }, { "epoch": 2.2372055239642568, "grad_norm": 0.35714420676231384, "learning_rate": 1.8441617275874701e-06, "loss": 0.3827, "step": 2754 }, { "epoch": 2.238017871649066, "grad_norm": 0.3132324516773224, "learning_rate": 1.8404966245680068e-06, "loss": 0.4053, "step": 2755 }, { "epoch": 2.2388302193338747, "grad_norm": 0.3791865110397339, "learning_rate": 1.836834345504076e-06, "loss": 0.3912, "step": 2756 }, { "epoch": 2.239642567018684, "grad_norm": 0.32979816198349, "learning_rate": 1.8331748936690175e-06, "loss": 0.4005, "step": 2757 }, { "epoch": 2.240454914703493, "grad_norm": 0.3563513457775116, "learning_rate": 1.829518272333633e-06, "loss": 0.4006, "step": 2758 }, { "epoch": 2.2412672623883023, "grad_norm": 0.33698683977127075, "learning_rate": 1.8258644847662087e-06, "loss": 0.4019, "step": 2759 }, { "epoch": 2.2420796100731115, "grad_norm": 0.3503924012184143, "learning_rate": 1.8222135342324844e-06, "loss": 0.3719, "step": 2760 }, { "epoch": 2.2428919577579203, "grad_norm": 0.3702690899372101, "learning_rate": 1.8185654239956723e-06, "loss": 0.4036, "step": 2761 }, { "epoch": 2.2437043054427295, "grad_norm": 0.3358888328075409, "learning_rate": 1.8149201573164415e-06, "loss": 0.4183, "step": 2762 }, { "epoch": 2.2445166531275387, "grad_norm": 0.3720953166484833, "learning_rate": 1.811277737452924e-06, "loss": 0.4072, "step": 2763 }, { "epoch": 2.245329000812348, "grad_norm": 0.32277512550354004, "learning_rate": 1.807638167660702e-06, "loss": 0.3948, "step": 2764 }, { "epoch": 2.2461413484971566, "grad_norm": 0.3491523563861847, "learning_rate": 1.8040014511928155e-06, "loss": 0.4283, "step": 2765 }, { "epoch": 2.246953696181966, "grad_norm": 0.3734605610370636, "learning_rate": 1.8003675912997487e-06, "loss": 0.4295, "step": 2766 }, { "epoch": 2.247766043866775, "grad_norm": 0.31637808680534363, "learning_rate": 1.7967365912294398e-06, "loss": 0.379, "step": 2767 }, { "epoch": 2.2485783915515842, "grad_norm": 0.3423219621181488, "learning_rate": 1.7931084542272632e-06, "loss": 0.3984, "step": 2768 }, { "epoch": 2.249390739236393, "grad_norm": 0.3712254464626312, "learning_rate": 1.7894831835360387e-06, "loss": 0.4019, "step": 2769 }, { "epoch": 2.250203086921202, "grad_norm": 0.3479743003845215, "learning_rate": 1.7858607823960262e-06, "loss": 0.3961, "step": 2770 }, { "epoch": 2.2510154346060114, "grad_norm": 0.3553738594055176, "learning_rate": 1.7822412540449141e-06, "loss": 0.3663, "step": 2771 }, { "epoch": 2.2518277822908206, "grad_norm": 0.3650190830230713, "learning_rate": 1.77862460171783e-06, "loss": 0.4309, "step": 2772 }, { "epoch": 2.25264012997563, "grad_norm": 0.3229494094848633, "learning_rate": 1.7750108286473255e-06, "loss": 0.404, "step": 2773 }, { "epoch": 2.2534524776604385, "grad_norm": 0.3454046845436096, "learning_rate": 1.771399938063384e-06, "loss": 0.4047, "step": 2774 }, { "epoch": 2.2542648253452477, "grad_norm": 0.32091012597084045, "learning_rate": 1.7677919331934057e-06, "loss": 0.3934, "step": 2775 }, { "epoch": 2.255077173030057, "grad_norm": 0.33534422516822815, "learning_rate": 1.7641868172622201e-06, "loss": 0.3799, "step": 2776 }, { "epoch": 2.255889520714866, "grad_norm": 0.3297094702720642, "learning_rate": 1.7605845934920663e-06, "loss": 0.4042, "step": 2777 }, { "epoch": 2.256701868399675, "grad_norm": 0.3339848816394806, "learning_rate": 1.7569852651026048e-06, "loss": 0.3881, "step": 2778 }, { "epoch": 2.257514216084484, "grad_norm": 0.3145342171192169, "learning_rate": 1.7533888353109029e-06, "loss": 0.3939, "step": 2779 }, { "epoch": 2.2583265637692933, "grad_norm": 0.3555033504962921, "learning_rate": 1.7497953073314433e-06, "loss": 0.3984, "step": 2780 }, { "epoch": 2.2591389114541025, "grad_norm": 0.3382641077041626, "learning_rate": 1.7462046843761078e-06, "loss": 0.4181, "step": 2781 }, { "epoch": 2.2599512591389113, "grad_norm": 0.3483993113040924, "learning_rate": 1.7426169696541884e-06, "loss": 0.4087, "step": 2782 }, { "epoch": 2.2607636068237205, "grad_norm": 0.35030099749565125, "learning_rate": 1.739032166372372e-06, "loss": 0.4102, "step": 2783 }, { "epoch": 2.2615759545085297, "grad_norm": 0.3372749090194702, "learning_rate": 1.7354502777347487e-06, "loss": 0.3916, "step": 2784 }, { "epoch": 2.262388302193339, "grad_norm": 0.3328147828578949, "learning_rate": 1.731871306942799e-06, "loss": 0.3924, "step": 2785 }, { "epoch": 2.263200649878148, "grad_norm": 0.33908915519714355, "learning_rate": 1.7282952571953987e-06, "loss": 0.3914, "step": 2786 }, { "epoch": 2.264012997562957, "grad_norm": 0.3157394528388977, "learning_rate": 1.724722131688809e-06, "loss": 0.3937, "step": 2787 }, { "epoch": 2.264825345247766, "grad_norm": 0.33812418580055237, "learning_rate": 1.721151933616682e-06, "loss": 0.4208, "step": 2788 }, { "epoch": 2.265637692932575, "grad_norm": 0.32404249906539917, "learning_rate": 1.7175846661700518e-06, "loss": 0.3841, "step": 2789 }, { "epoch": 2.2664500406173844, "grad_norm": 0.3194350600242615, "learning_rate": 1.7140203325373294e-06, "loss": 0.4111, "step": 2790 }, { "epoch": 2.267262388302193, "grad_norm": 0.2978324592113495, "learning_rate": 1.7104589359043095e-06, "loss": 0.395, "step": 2791 }, { "epoch": 2.2680747359870024, "grad_norm": 0.34215494990348816, "learning_rate": 1.706900479454156e-06, "loss": 0.3882, "step": 2792 }, { "epoch": 2.2688870836718116, "grad_norm": 0.34362056851387024, "learning_rate": 1.7033449663674102e-06, "loss": 0.3845, "step": 2793 }, { "epoch": 2.2696994313566208, "grad_norm": 0.30932220816612244, "learning_rate": 1.6997923998219774e-06, "loss": 0.4087, "step": 2794 }, { "epoch": 2.2705117790414295, "grad_norm": 0.3131434917449951, "learning_rate": 1.6962427829931348e-06, "loss": 0.4092, "step": 2795 }, { "epoch": 2.2713241267262387, "grad_norm": 0.32000917196273804, "learning_rate": 1.6926961190535152e-06, "loss": 0.4096, "step": 2796 }, { "epoch": 2.272136474411048, "grad_norm": 0.30410441756248474, "learning_rate": 1.6891524111731246e-06, "loss": 0.4095, "step": 2797 }, { "epoch": 2.272948822095857, "grad_norm": 0.3193068206310272, "learning_rate": 1.6856116625193135e-06, "loss": 0.4101, "step": 2798 }, { "epoch": 2.2737611697806663, "grad_norm": 0.33229681849479675, "learning_rate": 1.6820738762567978e-06, "loss": 0.4086, "step": 2799 }, { "epoch": 2.274573517465475, "grad_norm": 0.33888503909111023, "learning_rate": 1.678539055547637e-06, "loss": 0.3926, "step": 2800 }, { "epoch": 2.2753858651502843, "grad_norm": 0.3132135272026062, "learning_rate": 1.6750072035512482e-06, "loss": 0.4233, "step": 2801 }, { "epoch": 2.2761982128350935, "grad_norm": 0.3082081079483032, "learning_rate": 1.6714783234243875e-06, "loss": 0.3752, "step": 2802 }, { "epoch": 2.2770105605199027, "grad_norm": 0.32040730118751526, "learning_rate": 1.6679524183211616e-06, "loss": 0.4148, "step": 2803 }, { "epoch": 2.2778229082047114, "grad_norm": 0.3278828263282776, "learning_rate": 1.6644294913930115e-06, "loss": 0.419, "step": 2804 }, { "epoch": 2.2786352558895206, "grad_norm": 0.36316508054733276, "learning_rate": 1.6609095457887237e-06, "loss": 0.4095, "step": 2805 }, { "epoch": 2.27944760357433, "grad_norm": 0.33542323112487793, "learning_rate": 1.657392584654412e-06, "loss": 0.4105, "step": 2806 }, { "epoch": 2.280259951259139, "grad_norm": 0.31213635206222534, "learning_rate": 1.6538786111335303e-06, "loss": 0.3911, "step": 2807 }, { "epoch": 2.281072298943948, "grad_norm": 0.36755165457725525, "learning_rate": 1.6503676283668553e-06, "loss": 0.3897, "step": 2808 }, { "epoch": 2.281884646628757, "grad_norm": 0.36153995990753174, "learning_rate": 1.6468596394924974e-06, "loss": 0.4075, "step": 2809 }, { "epoch": 2.282696994313566, "grad_norm": 0.32209739089012146, "learning_rate": 1.6433546476458833e-06, "loss": 0.3927, "step": 2810 }, { "epoch": 2.2835093419983754, "grad_norm": 0.3499915599822998, "learning_rate": 1.63985265595977e-06, "loss": 0.384, "step": 2811 }, { "epoch": 2.2843216896831846, "grad_norm": 0.3410784602165222, "learning_rate": 1.6363536675642233e-06, "loss": 0.371, "step": 2812 }, { "epoch": 2.2851340373679934, "grad_norm": 0.30357474088668823, "learning_rate": 1.632857685586632e-06, "loss": 0.4117, "step": 2813 }, { "epoch": 2.2859463850528026, "grad_norm": 0.345873087644577, "learning_rate": 1.629364713151697e-06, "loss": 0.4125, "step": 2814 }, { "epoch": 2.2867587327376118, "grad_norm": 0.3331557810306549, "learning_rate": 1.625874753381424e-06, "loss": 0.3942, "step": 2815 }, { "epoch": 2.287571080422421, "grad_norm": 0.3146262764930725, "learning_rate": 1.6223878093951329e-06, "loss": 0.3983, "step": 2816 }, { "epoch": 2.2883834281072297, "grad_norm": 0.3331737816333771, "learning_rate": 1.6189038843094419e-06, "loss": 0.3881, "step": 2817 }, { "epoch": 2.289195775792039, "grad_norm": 0.3366614580154419, "learning_rate": 1.615422981238276e-06, "loss": 0.3902, "step": 2818 }, { "epoch": 2.290008123476848, "grad_norm": 0.3490856885910034, "learning_rate": 1.6119451032928547e-06, "loss": 0.4049, "step": 2819 }, { "epoch": 2.2908204711616573, "grad_norm": 0.33703017234802246, "learning_rate": 1.6084702535816987e-06, "loss": 0.4165, "step": 2820 }, { "epoch": 2.291632818846466, "grad_norm": 0.31082579493522644, "learning_rate": 1.6049984352106163e-06, "loss": 0.3992, "step": 2821 }, { "epoch": 2.2924451665312753, "grad_norm": 0.33386117219924927, "learning_rate": 1.6015296512827122e-06, "loss": 0.3859, "step": 2822 }, { "epoch": 2.2932575142160845, "grad_norm": 0.3287014067173004, "learning_rate": 1.5980639048983726e-06, "loss": 0.4042, "step": 2823 }, { "epoch": 2.2940698619008937, "grad_norm": 0.36145922541618347, "learning_rate": 1.5946011991552757e-06, "loss": 0.4003, "step": 2824 }, { "epoch": 2.294882209585703, "grad_norm": 0.35328736901283264, "learning_rate": 1.5911415371483757e-06, "loss": 0.4066, "step": 2825 }, { "epoch": 2.2956945572705116, "grad_norm": 0.3109525442123413, "learning_rate": 1.587684921969912e-06, "loss": 0.3976, "step": 2826 }, { "epoch": 2.296506904955321, "grad_norm": 0.32561567425727844, "learning_rate": 1.5842313567093942e-06, "loss": 0.3837, "step": 2827 }, { "epoch": 2.29731925264013, "grad_norm": 0.30941569805145264, "learning_rate": 1.5807808444536132e-06, "loss": 0.3878, "step": 2828 }, { "epoch": 2.2981316003249392, "grad_norm": 0.3622291386127472, "learning_rate": 1.5773333882866242e-06, "loss": 0.4038, "step": 2829 }, { "epoch": 2.298943948009748, "grad_norm": 0.33826345205307007, "learning_rate": 1.5738889912897575e-06, "loss": 0.4043, "step": 2830 }, { "epoch": 2.299756295694557, "grad_norm": 0.3087054491043091, "learning_rate": 1.5704476565416027e-06, "loss": 0.3899, "step": 2831 }, { "epoch": 2.3005686433793664, "grad_norm": 0.32625290751457214, "learning_rate": 1.5670093871180165e-06, "loss": 0.395, "step": 2832 }, { "epoch": 2.3013809910641756, "grad_norm": 0.30848804116249084, "learning_rate": 1.5635741860921166e-06, "loss": 0.4153, "step": 2833 }, { "epoch": 2.3021933387489844, "grad_norm": 0.34557828307151794, "learning_rate": 1.5601420565342734e-06, "loss": 0.4303, "step": 2834 }, { "epoch": 2.3030056864337936, "grad_norm": 0.3346918821334839, "learning_rate": 1.5567130015121169e-06, "loss": 0.4077, "step": 2835 }, { "epoch": 2.3038180341186028, "grad_norm": 0.3229162395000458, "learning_rate": 1.5532870240905272e-06, "loss": 0.3972, "step": 2836 }, { "epoch": 2.304630381803412, "grad_norm": 0.3359091579914093, "learning_rate": 1.5498641273316322e-06, "loss": 0.3858, "step": 2837 }, { "epoch": 2.305442729488221, "grad_norm": 0.3416934907436371, "learning_rate": 1.5464443142948083e-06, "loss": 0.3815, "step": 2838 }, { "epoch": 2.30625507717303, "grad_norm": 0.3520924150943756, "learning_rate": 1.5430275880366774e-06, "loss": 0.4151, "step": 2839 }, { "epoch": 2.307067424857839, "grad_norm": 0.31596893072128296, "learning_rate": 1.5396139516110959e-06, "loss": 0.3943, "step": 2840 }, { "epoch": 2.3078797725426483, "grad_norm": 0.3388243019580841, "learning_rate": 1.5362034080691673e-06, "loss": 0.3881, "step": 2841 }, { "epoch": 2.3086921202274575, "grad_norm": 0.33511003851890564, "learning_rate": 1.5327959604592218e-06, "loss": 0.4036, "step": 2842 }, { "epoch": 2.3095044679122663, "grad_norm": 0.3075815439224243, "learning_rate": 1.5293916118268304e-06, "loss": 0.3914, "step": 2843 }, { "epoch": 2.3103168155970755, "grad_norm": 0.30811211466789246, "learning_rate": 1.5259903652147878e-06, "loss": 0.4252, "step": 2844 }, { "epoch": 2.3111291632818847, "grad_norm": 0.30946633219718933, "learning_rate": 1.5225922236631225e-06, "loss": 0.4049, "step": 2845 }, { "epoch": 2.311941510966694, "grad_norm": 0.35558176040649414, "learning_rate": 1.5191971902090797e-06, "loss": 0.4087, "step": 2846 }, { "epoch": 2.3127538586515026, "grad_norm": 0.3385161757469177, "learning_rate": 1.515805267887135e-06, "loss": 0.3912, "step": 2847 }, { "epoch": 2.313566206336312, "grad_norm": 0.3164585828781128, "learning_rate": 1.5124164597289765e-06, "loss": 0.3834, "step": 2848 }, { "epoch": 2.314378554021121, "grad_norm": 0.326754093170166, "learning_rate": 1.5090307687635141e-06, "loss": 0.4184, "step": 2849 }, { "epoch": 2.3151909017059302, "grad_norm": 0.30498093366622925, "learning_rate": 1.505648198016867e-06, "loss": 0.3851, "step": 2850 }, { "epoch": 2.3160032493907394, "grad_norm": 0.3096010088920593, "learning_rate": 1.5022687505123702e-06, "loss": 0.3851, "step": 2851 }, { "epoch": 2.316815597075548, "grad_norm": 0.3115338087081909, "learning_rate": 1.4988924292705615e-06, "loss": 0.4212, "step": 2852 }, { "epoch": 2.3176279447603574, "grad_norm": 0.3707998991012573, "learning_rate": 1.4955192373091915e-06, "loss": 0.4019, "step": 2853 }, { "epoch": 2.3184402924451666, "grad_norm": 0.37084218859672546, "learning_rate": 1.4921491776432062e-06, "loss": 0.4099, "step": 2854 }, { "epoch": 2.319252640129976, "grad_norm": 0.37690824270248413, "learning_rate": 1.488782253284759e-06, "loss": 0.4267, "step": 2855 }, { "epoch": 2.3200649878147845, "grad_norm": 0.3231479525566101, "learning_rate": 1.4854184672431948e-06, "loss": 0.3907, "step": 2856 }, { "epoch": 2.3208773354995937, "grad_norm": 0.31650710105895996, "learning_rate": 1.4820578225250582e-06, "loss": 0.4059, "step": 2857 }, { "epoch": 2.321689683184403, "grad_norm": 0.3414243161678314, "learning_rate": 1.4787003221340851e-06, "loss": 0.4046, "step": 2858 }, { "epoch": 2.322502030869212, "grad_norm": 0.34858524799346924, "learning_rate": 1.475345969071198e-06, "loss": 0.4119, "step": 2859 }, { "epoch": 2.323314378554021, "grad_norm": 0.3338465988636017, "learning_rate": 1.4719947663345112e-06, "loss": 0.3648, "step": 2860 }, { "epoch": 2.32412672623883, "grad_norm": 0.31978219747543335, "learning_rate": 1.468646716919318e-06, "loss": 0.377, "step": 2861 }, { "epoch": 2.3249390739236393, "grad_norm": 0.33066391944885254, "learning_rate": 1.4653018238180982e-06, "loss": 0.4025, "step": 2862 }, { "epoch": 2.3257514216084485, "grad_norm": 0.3189147710800171, "learning_rate": 1.461960090020505e-06, "loss": 0.4104, "step": 2863 }, { "epoch": 2.3265637692932577, "grad_norm": 0.30618587136268616, "learning_rate": 1.458621518513375e-06, "loss": 0.3828, "step": 2864 }, { "epoch": 2.3273761169780665, "grad_norm": 0.3717930316925049, "learning_rate": 1.455286112280711e-06, "loss": 0.4237, "step": 2865 }, { "epoch": 2.3281884646628757, "grad_norm": 0.3257165849208832, "learning_rate": 1.4519538743036927e-06, "loss": 0.4251, "step": 2866 }, { "epoch": 2.329000812347685, "grad_norm": 0.3218206465244293, "learning_rate": 1.4486248075606624e-06, "loss": 0.3902, "step": 2867 }, { "epoch": 2.329813160032494, "grad_norm": 0.3058813512325287, "learning_rate": 1.445298915027134e-06, "loss": 0.3941, "step": 2868 }, { "epoch": 2.330625507717303, "grad_norm": 0.3144093155860901, "learning_rate": 1.441976199675778e-06, "loss": 0.412, "step": 2869 }, { "epoch": 2.331437855402112, "grad_norm": 0.3257594406604767, "learning_rate": 1.4386566644764321e-06, "loss": 0.4309, "step": 2870 }, { "epoch": 2.332250203086921, "grad_norm": 0.3037477731704712, "learning_rate": 1.435340312396084e-06, "loss": 0.3828, "step": 2871 }, { "epoch": 2.3330625507717304, "grad_norm": 0.31920596957206726, "learning_rate": 1.4320271463988822e-06, "loss": 0.4043, "step": 2872 }, { "epoch": 2.333874898456539, "grad_norm": 0.3194264769554138, "learning_rate": 1.428717169446125e-06, "loss": 0.3826, "step": 2873 }, { "epoch": 2.3346872461413484, "grad_norm": 0.313929945230484, "learning_rate": 1.425410384496263e-06, "loss": 0.4219, "step": 2874 }, { "epoch": 2.3354995938261576, "grad_norm": 0.30695202946662903, "learning_rate": 1.4221067945048877e-06, "loss": 0.3973, "step": 2875 }, { "epoch": 2.3363119415109668, "grad_norm": 0.32783666253089905, "learning_rate": 1.4188064024247428e-06, "loss": 0.3771, "step": 2876 }, { "epoch": 2.337124289195776, "grad_norm": 0.3101159334182739, "learning_rate": 1.415509211205706e-06, "loss": 0.399, "step": 2877 }, { "epoch": 2.3379366368805847, "grad_norm": 0.371079683303833, "learning_rate": 1.4122152237948018e-06, "loss": 0.4026, "step": 2878 }, { "epoch": 2.338748984565394, "grad_norm": 0.3237926959991455, "learning_rate": 1.4089244431361826e-06, "loss": 0.4084, "step": 2879 }, { "epoch": 2.339561332250203, "grad_norm": 0.33235955238342285, "learning_rate": 1.405636872171144e-06, "loss": 0.376, "step": 2880 }, { "epoch": 2.3403736799350123, "grad_norm": 0.3051186203956604, "learning_rate": 1.4023525138381039e-06, "loss": 0.3756, "step": 2881 }, { "epoch": 2.3411860276198215, "grad_norm": 0.3069360852241516, "learning_rate": 1.3990713710726145e-06, "loss": 0.3731, "step": 2882 }, { "epoch": 2.3419983753046303, "grad_norm": 0.3547517657279968, "learning_rate": 1.395793446807354e-06, "loss": 0.4318, "step": 2883 }, { "epoch": 2.3428107229894395, "grad_norm": 0.3163175582885742, "learning_rate": 1.3925187439721193e-06, "loss": 0.3997, "step": 2884 }, { "epoch": 2.3436230706742487, "grad_norm": 0.3274424970149994, "learning_rate": 1.3892472654938338e-06, "loss": 0.4269, "step": 2885 }, { "epoch": 2.3444354183590574, "grad_norm": 0.325609415769577, "learning_rate": 1.385979014296533e-06, "loss": 0.3815, "step": 2886 }, { "epoch": 2.3452477660438666, "grad_norm": 0.31528210639953613, "learning_rate": 1.3827139933013739e-06, "loss": 0.3945, "step": 2887 }, { "epoch": 2.346060113728676, "grad_norm": 0.3128407299518585, "learning_rate": 1.3794522054266207e-06, "loss": 0.3924, "step": 2888 }, { "epoch": 2.346872461413485, "grad_norm": 0.3745444118976593, "learning_rate": 1.376193653587653e-06, "loss": 0.3902, "step": 2889 }, { "epoch": 2.3476848090982942, "grad_norm": 0.34880688786506653, "learning_rate": 1.3729383406969533e-06, "loss": 0.37, "step": 2890 }, { "epoch": 2.348497156783103, "grad_norm": 0.3270017206668854, "learning_rate": 1.3696862696641133e-06, "loss": 0.3838, "step": 2891 }, { "epoch": 2.349309504467912, "grad_norm": 0.305862158536911, "learning_rate": 1.366437443395824e-06, "loss": 0.3788, "step": 2892 }, { "epoch": 2.3501218521527214, "grad_norm": 0.3122928738594055, "learning_rate": 1.3631918647958786e-06, "loss": 0.4004, "step": 2893 }, { "epoch": 2.3509341998375306, "grad_norm": 0.30406659841537476, "learning_rate": 1.3599495367651645e-06, "loss": 0.3906, "step": 2894 }, { "epoch": 2.35174654752234, "grad_norm": 0.29217496514320374, "learning_rate": 1.3567104622016685e-06, "loss": 0.3809, "step": 2895 }, { "epoch": 2.3525588952071486, "grad_norm": 0.33505886793136597, "learning_rate": 1.3534746440004637e-06, "loss": 0.4022, "step": 2896 }, { "epoch": 2.3533712428919578, "grad_norm": 0.306234210729599, "learning_rate": 1.3502420850537185e-06, "loss": 0.4159, "step": 2897 }, { "epoch": 2.354183590576767, "grad_norm": 0.3248692750930786, "learning_rate": 1.3470127882506824e-06, "loss": 0.3957, "step": 2898 }, { "epoch": 2.3549959382615757, "grad_norm": 0.2986587584018707, "learning_rate": 1.3437867564776952e-06, "loss": 0.3939, "step": 2899 }, { "epoch": 2.355808285946385, "grad_norm": 0.2864691913127899, "learning_rate": 1.3405639926181718e-06, "loss": 0.3958, "step": 2900 }, { "epoch": 2.356620633631194, "grad_norm": 0.3094498813152313, "learning_rate": 1.337344499552612e-06, "loss": 0.4031, "step": 2901 }, { "epoch": 2.3574329813160033, "grad_norm": 0.3357416093349457, "learning_rate": 1.3341282801585914e-06, "loss": 0.39, "step": 2902 }, { "epoch": 2.3582453290008125, "grad_norm": 0.35451292991638184, "learning_rate": 1.3309153373107547e-06, "loss": 0.4029, "step": 2903 }, { "epoch": 2.3590576766856213, "grad_norm": 0.3414060175418854, "learning_rate": 1.3277056738808242e-06, "loss": 0.4079, "step": 2904 }, { "epoch": 2.3598700243704305, "grad_norm": 0.35274192690849304, "learning_rate": 1.3244992927375865e-06, "loss": 0.4172, "step": 2905 }, { "epoch": 2.3606823720552397, "grad_norm": 0.32465675473213196, "learning_rate": 1.3212961967468985e-06, "loss": 0.3984, "step": 2906 }, { "epoch": 2.361494719740049, "grad_norm": 0.29780328273773193, "learning_rate": 1.3180963887716758e-06, "loss": 0.4071, "step": 2907 }, { "epoch": 2.362307067424858, "grad_norm": 0.31804388761520386, "learning_rate": 1.3148998716719014e-06, "loss": 0.3707, "step": 2908 }, { "epoch": 2.363119415109667, "grad_norm": 0.3290041983127594, "learning_rate": 1.3117066483046098e-06, "loss": 0.3914, "step": 2909 }, { "epoch": 2.363931762794476, "grad_norm": 0.31818443536758423, "learning_rate": 1.3085167215238982e-06, "loss": 0.4013, "step": 2910 }, { "epoch": 2.3647441104792852, "grad_norm": 0.3272066116333008, "learning_rate": 1.3053300941809132e-06, "loss": 0.396, "step": 2911 }, { "epoch": 2.365556458164094, "grad_norm": 0.3467862010002136, "learning_rate": 1.302146769123856e-06, "loss": 0.4039, "step": 2912 }, { "epoch": 2.366368805848903, "grad_norm": 0.32177287340164185, "learning_rate": 1.2989667491979702e-06, "loss": 0.3966, "step": 2913 }, { "epoch": 2.3671811535337124, "grad_norm": 0.34209850430488586, "learning_rate": 1.2957900372455529e-06, "loss": 0.4009, "step": 2914 }, { "epoch": 2.3679935012185216, "grad_norm": 0.3264518678188324, "learning_rate": 1.2926166361059366e-06, "loss": 0.3958, "step": 2915 }, { "epoch": 2.368805848903331, "grad_norm": 0.35357701778411865, "learning_rate": 1.2894465486155028e-06, "loss": 0.4295, "step": 2916 }, { "epoch": 2.3696181965881395, "grad_norm": 0.3231707513332367, "learning_rate": 1.2862797776076624e-06, "loss": 0.3988, "step": 2917 }, { "epoch": 2.3704305442729487, "grad_norm": 0.36254674196243286, "learning_rate": 1.2831163259128714e-06, "loss": 0.4258, "step": 2918 }, { "epoch": 2.371242891957758, "grad_norm": 0.33546891808509827, "learning_rate": 1.2799561963586104e-06, "loss": 0.3925, "step": 2919 }, { "epoch": 2.372055239642567, "grad_norm": 0.33172982931137085, "learning_rate": 1.2767993917693984e-06, "loss": 0.4001, "step": 2920 }, { "epoch": 2.3728675873273763, "grad_norm": 0.32151344418525696, "learning_rate": 1.2736459149667747e-06, "loss": 0.389, "step": 2921 }, { "epoch": 2.373679935012185, "grad_norm": 0.316012442111969, "learning_rate": 1.2704957687693114e-06, "loss": 0.3965, "step": 2922 }, { "epoch": 2.3744922826969943, "grad_norm": 0.32510891556739807, "learning_rate": 1.2673489559925989e-06, "loss": 0.4008, "step": 2923 }, { "epoch": 2.3753046303818035, "grad_norm": 0.29814469814300537, "learning_rate": 1.2642054794492514e-06, "loss": 0.4009, "step": 2924 }, { "epoch": 2.3761169780666123, "grad_norm": 0.3240830600261688, "learning_rate": 1.2610653419488978e-06, "loss": 0.3924, "step": 2925 }, { "epoch": 2.3769293257514215, "grad_norm": 0.32139429450035095, "learning_rate": 1.2579285462981855e-06, "loss": 0.4026, "step": 2926 }, { "epoch": 2.3777416734362307, "grad_norm": 0.337580144405365, "learning_rate": 1.2547950953007754e-06, "loss": 0.3848, "step": 2927 }, { "epoch": 2.37855402112104, "grad_norm": 0.3653236925601959, "learning_rate": 1.2516649917573342e-06, "loss": 0.3888, "step": 2928 }, { "epoch": 2.379366368805849, "grad_norm": 0.4876585900783539, "learning_rate": 1.2485382384655437e-06, "loss": 0.4012, "step": 2929 }, { "epoch": 2.380178716490658, "grad_norm": 0.29925328493118286, "learning_rate": 1.2454148382200837e-06, "loss": 0.3909, "step": 2930 }, { "epoch": 2.380991064175467, "grad_norm": 0.32978683710098267, "learning_rate": 1.2422947938126445e-06, "loss": 0.3862, "step": 2931 }, { "epoch": 2.381803411860276, "grad_norm": 0.3155513405799866, "learning_rate": 1.2391781080319098e-06, "loss": 0.4011, "step": 2932 }, { "epoch": 2.3826157595450854, "grad_norm": 0.3448691964149475, "learning_rate": 1.236064783663568e-06, "loss": 0.3818, "step": 2933 }, { "epoch": 2.3834281072298946, "grad_norm": 0.35452166199684143, "learning_rate": 1.2329548234902976e-06, "loss": 0.3992, "step": 2934 }, { "epoch": 2.3842404549147034, "grad_norm": 0.3756706416606903, "learning_rate": 1.2298482302917752e-06, "loss": 0.3713, "step": 2935 }, { "epoch": 2.3850528025995126, "grad_norm": 0.3143619894981384, "learning_rate": 1.2267450068446617e-06, "loss": 0.4132, "step": 2936 }, { "epoch": 2.385865150284322, "grad_norm": 0.3295286297798157, "learning_rate": 1.2236451559226137e-06, "loss": 0.4018, "step": 2937 }, { "epoch": 2.386677497969131, "grad_norm": 0.32291191816329956, "learning_rate": 1.2205486802962656e-06, "loss": 0.3987, "step": 2938 }, { "epoch": 2.3874898456539397, "grad_norm": 0.3418336808681488, "learning_rate": 1.217455582733243e-06, "loss": 0.4035, "step": 2939 }, { "epoch": 2.388302193338749, "grad_norm": 0.301019549369812, "learning_rate": 1.2143658659981444e-06, "loss": 0.3841, "step": 2940 }, { "epoch": 2.389114541023558, "grad_norm": 0.30485448241233826, "learning_rate": 1.2112795328525529e-06, "loss": 0.4042, "step": 2941 }, { "epoch": 2.3899268887083673, "grad_norm": 0.3346501588821411, "learning_rate": 1.2081965860550227e-06, "loss": 0.4095, "step": 2942 }, { "epoch": 2.390739236393176, "grad_norm": 0.43871229887008667, "learning_rate": 1.2051170283610858e-06, "loss": 0.4053, "step": 2943 }, { "epoch": 2.3915515840779853, "grad_norm": 0.30541709065437317, "learning_rate": 1.2020408625232388e-06, "loss": 0.3674, "step": 2944 }, { "epoch": 2.3923639317627945, "grad_norm": 0.33992111682891846, "learning_rate": 1.198968091290953e-06, "loss": 0.3896, "step": 2945 }, { "epoch": 2.3931762794476037, "grad_norm": 0.3226431906223297, "learning_rate": 1.195898717410664e-06, "loss": 0.3945, "step": 2946 }, { "epoch": 2.393988627132413, "grad_norm": 0.3605234920978546, "learning_rate": 1.1928327436257669e-06, "loss": 0.3696, "step": 2947 }, { "epoch": 2.3948009748172217, "grad_norm": 0.323266863822937, "learning_rate": 1.189770172676623e-06, "loss": 0.4051, "step": 2948 }, { "epoch": 2.395613322502031, "grad_norm": 0.28726035356521606, "learning_rate": 1.1867110073005483e-06, "loss": 0.4063, "step": 2949 }, { "epoch": 2.39642567018684, "grad_norm": 0.3223848044872284, "learning_rate": 1.1836552502318171e-06, "loss": 0.4075, "step": 2950 }, { "epoch": 2.3972380178716493, "grad_norm": 0.31901854276657104, "learning_rate": 1.1806029042016581e-06, "loss": 0.3908, "step": 2951 }, { "epoch": 2.398050365556458, "grad_norm": 0.31621044874191284, "learning_rate": 1.1775539719382473e-06, "loss": 0.3863, "step": 2952 }, { "epoch": 2.398862713241267, "grad_norm": 0.2739246189594269, "learning_rate": 1.1745084561667136e-06, "loss": 0.4141, "step": 2953 }, { "epoch": 2.3996750609260764, "grad_norm": 0.33529770374298096, "learning_rate": 1.1714663596091319e-06, "loss": 0.403, "step": 2954 }, { "epoch": 2.4004874086108856, "grad_norm": 0.3150768280029297, "learning_rate": 1.1684276849845171e-06, "loss": 0.403, "step": 2955 }, { "epoch": 2.4012997562956944, "grad_norm": 0.3093283474445343, "learning_rate": 1.1653924350088319e-06, "loss": 0.3865, "step": 2956 }, { "epoch": 2.4021121039805036, "grad_norm": 0.31278371810913086, "learning_rate": 1.1623606123949705e-06, "loss": 0.4111, "step": 2957 }, { "epoch": 2.4029244516653128, "grad_norm": 0.36281001567840576, "learning_rate": 1.1593322198527707e-06, "loss": 0.4095, "step": 2958 }, { "epoch": 2.403736799350122, "grad_norm": 0.32101866602897644, "learning_rate": 1.1563072600889992e-06, "loss": 0.3955, "step": 2959 }, { "epoch": 2.404549147034931, "grad_norm": 0.3517989218235016, "learning_rate": 1.1532857358073597e-06, "loss": 0.3879, "step": 2960 }, { "epoch": 2.40536149471974, "grad_norm": 0.3430308699607849, "learning_rate": 1.1502676497084786e-06, "loss": 0.4078, "step": 2961 }, { "epoch": 2.406173842404549, "grad_norm": 0.3111429214477539, "learning_rate": 1.147253004489917e-06, "loss": 0.3979, "step": 2962 }, { "epoch": 2.4069861900893583, "grad_norm": 0.3368332087993622, "learning_rate": 1.1442418028461537e-06, "loss": 0.4397, "step": 2963 }, { "epoch": 2.4077985377741675, "grad_norm": 0.3263891339302063, "learning_rate": 1.1412340474685946e-06, "loss": 0.405, "step": 2964 }, { "epoch": 2.4086108854589763, "grad_norm": 0.3517023026943207, "learning_rate": 1.1382297410455612e-06, "loss": 0.3968, "step": 2965 }, { "epoch": 2.4094232331437855, "grad_norm": 0.2961708903312683, "learning_rate": 1.1352288862622968e-06, "loss": 0.3847, "step": 2966 }, { "epoch": 2.4102355808285947, "grad_norm": 0.31961700320243835, "learning_rate": 1.1322314858009547e-06, "loss": 0.4173, "step": 2967 }, { "epoch": 2.411047928513404, "grad_norm": 0.3213881254196167, "learning_rate": 1.129237542340606e-06, "loss": 0.3954, "step": 2968 }, { "epoch": 2.4118602761982126, "grad_norm": 0.31646502017974854, "learning_rate": 1.126247058557226e-06, "loss": 0.4002, "step": 2969 }, { "epoch": 2.412672623883022, "grad_norm": 0.3437632620334625, "learning_rate": 1.1232600371237035e-06, "loss": 0.3953, "step": 2970 }, { "epoch": 2.413484971567831, "grad_norm": 0.3372691571712494, "learning_rate": 1.1202764807098305e-06, "loss": 0.4261, "step": 2971 }, { "epoch": 2.4142973192526402, "grad_norm": 0.3224683403968811, "learning_rate": 1.1172963919822988e-06, "loss": 0.4051, "step": 2972 }, { "epoch": 2.4151096669374494, "grad_norm": 0.3201896548271179, "learning_rate": 1.1143197736047073e-06, "loss": 0.3983, "step": 2973 }, { "epoch": 2.415922014622258, "grad_norm": 0.3386845290660858, "learning_rate": 1.1113466282375457e-06, "loss": 0.4006, "step": 2974 }, { "epoch": 2.4167343623070674, "grad_norm": 0.3129492402076721, "learning_rate": 1.108376958538207e-06, "loss": 0.4077, "step": 2975 }, { "epoch": 2.4175467099918766, "grad_norm": 0.32023999094963074, "learning_rate": 1.1054107671609697e-06, "loss": 0.4023, "step": 2976 }, { "epoch": 2.418359057676686, "grad_norm": 0.3324236273765564, "learning_rate": 1.1024480567570123e-06, "loss": 0.4089, "step": 2977 }, { "epoch": 2.4191714053614946, "grad_norm": 0.3387928903102875, "learning_rate": 1.0994888299743929e-06, "loss": 0.3812, "step": 2978 }, { "epoch": 2.4199837530463038, "grad_norm": 0.3089993894100189, "learning_rate": 1.0965330894580638e-06, "loss": 0.4021, "step": 2979 }, { "epoch": 2.420796100731113, "grad_norm": 0.3302111029624939, "learning_rate": 1.0935808378498547e-06, "loss": 0.3971, "step": 2980 }, { "epoch": 2.421608448415922, "grad_norm": 0.32042378187179565, "learning_rate": 1.0906320777884833e-06, "loss": 0.421, "step": 2981 }, { "epoch": 2.422420796100731, "grad_norm": 0.32324162125587463, "learning_rate": 1.0876868119095407e-06, "loss": 0.3948, "step": 2982 }, { "epoch": 2.42323314378554, "grad_norm": 0.30912137031555176, "learning_rate": 1.0847450428454997e-06, "loss": 0.4006, "step": 2983 }, { "epoch": 2.4240454914703493, "grad_norm": 0.3188294768333435, "learning_rate": 1.0818067732257032e-06, "loss": 0.4116, "step": 2984 }, { "epoch": 2.4248578391551585, "grad_norm": 0.31744885444641113, "learning_rate": 1.0788720056763712e-06, "loss": 0.4045, "step": 2985 }, { "epoch": 2.4256701868399677, "grad_norm": 0.30359935760498047, "learning_rate": 1.075940742820588e-06, "loss": 0.3805, "step": 2986 }, { "epoch": 2.4264825345247765, "grad_norm": 0.3409248888492584, "learning_rate": 1.0730129872783102e-06, "loss": 0.4211, "step": 2987 }, { "epoch": 2.4272948822095857, "grad_norm": 0.3134811222553253, "learning_rate": 1.0700887416663575e-06, "loss": 0.3844, "step": 2988 }, { "epoch": 2.428107229894395, "grad_norm": 0.34443387389183044, "learning_rate": 1.0671680085984137e-06, "loss": 0.4239, "step": 2989 }, { "epoch": 2.428919577579204, "grad_norm": 0.34095945954322815, "learning_rate": 1.0642507906850202e-06, "loss": 0.4068, "step": 2990 }, { "epoch": 2.429731925264013, "grad_norm": 0.3262979984283447, "learning_rate": 1.0613370905335801e-06, "loss": 0.3872, "step": 2991 }, { "epoch": 2.430544272948822, "grad_norm": 0.3090975284576416, "learning_rate": 1.058426910748349e-06, "loss": 0.4057, "step": 2992 }, { "epoch": 2.4313566206336312, "grad_norm": 0.3218819499015808, "learning_rate": 1.0555202539304398e-06, "loss": 0.39, "step": 2993 }, { "epoch": 2.4321689683184404, "grad_norm": 0.32656538486480713, "learning_rate": 1.052617122677812e-06, "loss": 0.413, "step": 2994 }, { "epoch": 2.432981316003249, "grad_norm": 0.34132465720176697, "learning_rate": 1.0497175195852777e-06, "loss": 0.4281, "step": 2995 }, { "epoch": 2.4337936636880584, "grad_norm": 0.4224664568901062, "learning_rate": 1.0468214472444955e-06, "loss": 0.3982, "step": 2996 }, { "epoch": 2.4346060113728676, "grad_norm": 0.3518453538417816, "learning_rate": 1.0439289082439647e-06, "loss": 0.3782, "step": 2997 }, { "epoch": 2.435418359057677, "grad_norm": 0.3127150535583496, "learning_rate": 1.0410399051690306e-06, "loss": 0.3865, "step": 2998 }, { "epoch": 2.436230706742486, "grad_norm": 0.3268446624279022, "learning_rate": 1.0381544406018751e-06, "loss": 0.397, "step": 2999 }, { "epoch": 2.4370430544272947, "grad_norm": 0.32560864090919495, "learning_rate": 1.0352725171215194e-06, "loss": 0.4276, "step": 3000 }, { "epoch": 2.437855402112104, "grad_norm": 0.3014163076877594, "learning_rate": 1.0323941373038177e-06, "loss": 0.394, "step": 3001 }, { "epoch": 2.438667749796913, "grad_norm": 0.3396751284599304, "learning_rate": 1.0295193037214606e-06, "loss": 0.4119, "step": 3002 }, { "epoch": 2.4394800974817223, "grad_norm": 0.33938851952552795, "learning_rate": 1.0266480189439631e-06, "loss": 0.4121, "step": 3003 }, { "epoch": 2.440292445166531, "grad_norm": 0.3266737163066864, "learning_rate": 1.0237802855376745e-06, "loss": 0.3848, "step": 3004 }, { "epoch": 2.4411047928513403, "grad_norm": 0.3313482105731964, "learning_rate": 1.0209161060657651e-06, "loss": 0.4252, "step": 3005 }, { "epoch": 2.4419171405361495, "grad_norm": 0.30972322821617126, "learning_rate": 1.0180554830882333e-06, "loss": 0.3955, "step": 3006 }, { "epoch": 2.4427294882209587, "grad_norm": 0.31751471757888794, "learning_rate": 1.015198419161893e-06, "loss": 0.4147, "step": 3007 }, { "epoch": 2.4435418359057675, "grad_norm": 0.3034610450267792, "learning_rate": 1.0123449168403826e-06, "loss": 0.4159, "step": 3008 }, { "epoch": 2.4443541835905767, "grad_norm": 0.32123425602912903, "learning_rate": 1.0094949786741531e-06, "loss": 0.399, "step": 3009 }, { "epoch": 2.445166531275386, "grad_norm": 0.3134936988353729, "learning_rate": 1.0066486072104736e-06, "loss": 0.4048, "step": 3010 }, { "epoch": 2.445978878960195, "grad_norm": 0.32279106974601746, "learning_rate": 1.0038058049934203e-06, "loss": 0.4195, "step": 3011 }, { "epoch": 2.4467912266450043, "grad_norm": 0.33243227005004883, "learning_rate": 1.0009665745638852e-06, "loss": 0.3869, "step": 3012 }, { "epoch": 2.447603574329813, "grad_norm": 0.323783278465271, "learning_rate": 9.981309184595617e-07, "loss": 0.3758, "step": 3013 }, { "epoch": 2.448415922014622, "grad_norm": 0.3022133409976959, "learning_rate": 9.952988392149532e-07, "loss": 0.3791, "step": 3014 }, { "epoch": 2.4492282696994314, "grad_norm": 0.3224087357521057, "learning_rate": 9.924703393613655e-07, "loss": 0.3931, "step": 3015 }, { "epoch": 2.4500406173842406, "grad_norm": 0.31063491106033325, "learning_rate": 9.896454214269014e-07, "loss": 0.4091, "step": 3016 }, { "epoch": 2.4508529650690494, "grad_norm": 0.32100847363471985, "learning_rate": 9.868240879364676e-07, "loss": 0.4181, "step": 3017 }, { "epoch": 2.4516653127538586, "grad_norm": 0.3268452286720276, "learning_rate": 9.840063414117618e-07, "loss": 0.4121, "step": 3018 }, { "epoch": 2.4524776604386678, "grad_norm": 0.3526345193386078, "learning_rate": 9.811921843712813e-07, "loss": 0.3988, "step": 3019 }, { "epoch": 2.453290008123477, "grad_norm": 0.310891717672348, "learning_rate": 9.78381619330308e-07, "loss": 0.4001, "step": 3020 }, { "epoch": 2.4541023558082857, "grad_norm": 0.31552839279174805, "learning_rate": 9.75574648800921e-07, "loss": 0.4218, "step": 3021 }, { "epoch": 2.454914703493095, "grad_norm": 0.3576384484767914, "learning_rate": 9.727712752919804e-07, "loss": 0.4047, "step": 3022 }, { "epoch": 2.455727051177904, "grad_norm": 0.3197196424007416, "learning_rate": 9.699715013091354e-07, "loss": 0.4018, "step": 3023 }, { "epoch": 2.4565393988627133, "grad_norm": 0.34566831588745117, "learning_rate": 9.671753293548153e-07, "loss": 0.4037, "step": 3024 }, { "epoch": 2.4573517465475225, "grad_norm": 0.3130398392677307, "learning_rate": 9.643827619282309e-07, "loss": 0.3818, "step": 3025 }, { "epoch": 2.4581640942323313, "grad_norm": 0.32040145993232727, "learning_rate": 9.615938015253723e-07, "loss": 0.3938, "step": 3026 }, { "epoch": 2.4589764419171405, "grad_norm": 0.32835596799850464, "learning_rate": 9.588084506390055e-07, "loss": 0.3788, "step": 3027 }, { "epoch": 2.4597887896019497, "grad_norm": 0.3319845199584961, "learning_rate": 9.56026711758667e-07, "loss": 0.3942, "step": 3028 }, { "epoch": 2.460601137286759, "grad_norm": 0.3005688488483429, "learning_rate": 9.532485873706704e-07, "loss": 0.4051, "step": 3029 }, { "epoch": 2.4614134849715676, "grad_norm": 0.28133726119995117, "learning_rate": 9.504740799580931e-07, "loss": 0.3625, "step": 3030 }, { "epoch": 2.462225832656377, "grad_norm": 0.3373579680919647, "learning_rate": 9.477031920007845e-07, "loss": 0.4192, "step": 3031 }, { "epoch": 2.463038180341186, "grad_norm": 0.35395070910453796, "learning_rate": 9.449359259753543e-07, "loss": 0.3993, "step": 3032 }, { "epoch": 2.4638505280259952, "grad_norm": 0.32288557291030884, "learning_rate": 9.421722843551801e-07, "loss": 0.436, "step": 3033 }, { "epoch": 2.464662875710804, "grad_norm": 0.2946115732192993, "learning_rate": 9.394122696103947e-07, "loss": 0.3978, "step": 3034 }, { "epoch": 2.465475223395613, "grad_norm": 0.2839011549949646, "learning_rate": 9.366558842078943e-07, "loss": 0.3814, "step": 3035 }, { "epoch": 2.4662875710804224, "grad_norm": 0.30669400095939636, "learning_rate": 9.339031306113256e-07, "loss": 0.3919, "step": 3036 }, { "epoch": 2.4670999187652316, "grad_norm": 0.2918338179588318, "learning_rate": 9.311540112810957e-07, "loss": 0.3926, "step": 3037 }, { "epoch": 2.467912266450041, "grad_norm": 0.3361571133136749, "learning_rate": 9.284085286743571e-07, "loss": 0.3795, "step": 3038 }, { "epoch": 2.4687246141348496, "grad_norm": 0.30374228954315186, "learning_rate": 9.256666852450158e-07, "loss": 0.3815, "step": 3039 }, { "epoch": 2.4695369618196588, "grad_norm": 0.3204629123210907, "learning_rate": 9.229284834437252e-07, "loss": 0.3877, "step": 3040 }, { "epoch": 2.470349309504468, "grad_norm": 0.3358759880065918, "learning_rate": 9.20193925717881e-07, "loss": 0.4128, "step": 3041 }, { "epoch": 2.471161657189277, "grad_norm": 0.2967516779899597, "learning_rate": 9.174630145116248e-07, "loss": 0.3986, "step": 3042 }, { "epoch": 2.471974004874086, "grad_norm": 0.31051105260849, "learning_rate": 9.147357522658351e-07, "loss": 0.396, "step": 3043 }, { "epoch": 2.472786352558895, "grad_norm": 0.2986542880535126, "learning_rate": 9.120121414181354e-07, "loss": 0.3987, "step": 3044 }, { "epoch": 2.4735987002437043, "grad_norm": 0.32909372448921204, "learning_rate": 9.092921844028779e-07, "loss": 0.3919, "step": 3045 }, { "epoch": 2.4744110479285135, "grad_norm": 0.31840384006500244, "learning_rate": 9.065758836511556e-07, "loss": 0.397, "step": 3046 }, { "epoch": 2.4752233956133223, "grad_norm": 0.3519814610481262, "learning_rate": 9.038632415907883e-07, "loss": 0.3839, "step": 3047 }, { "epoch": 2.4760357432981315, "grad_norm": 0.34309276938438416, "learning_rate": 9.011542606463309e-07, "loss": 0.3927, "step": 3048 }, { "epoch": 2.4768480909829407, "grad_norm": 0.33575570583343506, "learning_rate": 8.984489432390597e-07, "loss": 0.3975, "step": 3049 }, { "epoch": 2.47766043866775, "grad_norm": 0.33665019273757935, "learning_rate": 8.957472917869836e-07, "loss": 0.4034, "step": 3050 }, { "epoch": 2.478472786352559, "grad_norm": 0.3149738013744354, "learning_rate": 8.930493087048286e-07, "loss": 0.3975, "step": 3051 }, { "epoch": 2.479285134037368, "grad_norm": 0.33006876707077026, "learning_rate": 8.90354996404047e-07, "loss": 0.3901, "step": 3052 }, { "epoch": 2.480097481722177, "grad_norm": 0.33566656708717346, "learning_rate": 8.876643572928056e-07, "loss": 0.41, "step": 3053 }, { "epoch": 2.4809098294069862, "grad_norm": 0.28900375962257385, "learning_rate": 8.84977393775992e-07, "loss": 0.3772, "step": 3054 }, { "epoch": 2.4817221770917954, "grad_norm": 0.2984250485897064, "learning_rate": 8.822941082552055e-07, "loss": 0.3628, "step": 3055 }, { "epoch": 2.4825345247766046, "grad_norm": 0.31372156739234924, "learning_rate": 8.796145031287612e-07, "loss": 0.4033, "step": 3056 }, { "epoch": 2.4833468724614134, "grad_norm": 0.28385254740715027, "learning_rate": 8.769385807916808e-07, "loss": 0.3876, "step": 3057 }, { "epoch": 2.4841592201462226, "grad_norm": 0.30986085534095764, "learning_rate": 8.742663436356969e-07, "loss": 0.4077, "step": 3058 }, { "epoch": 2.484971567831032, "grad_norm": 0.3043557107448578, "learning_rate": 8.715977940492504e-07, "loss": 0.4079, "step": 3059 }, { "epoch": 2.4857839155158405, "grad_norm": 0.30529287457466125, "learning_rate": 8.689329344174797e-07, "loss": 0.4187, "step": 3060 }, { "epoch": 2.4865962632006497, "grad_norm": 0.3195667862892151, "learning_rate": 8.662717671222326e-07, "loss": 0.4036, "step": 3061 }, { "epoch": 2.487408610885459, "grad_norm": 0.3265250325202942, "learning_rate": 8.6361429454205e-07, "loss": 0.3872, "step": 3062 }, { "epoch": 2.488220958570268, "grad_norm": 0.3277387320995331, "learning_rate": 8.609605190521752e-07, "loss": 0.4016, "step": 3063 }, { "epoch": 2.4890333062550773, "grad_norm": 0.3105236887931824, "learning_rate": 8.583104430245476e-07, "loss": 0.386, "step": 3064 }, { "epoch": 2.489845653939886, "grad_norm": 0.3036220371723175, "learning_rate": 8.55664068827794e-07, "loss": 0.376, "step": 3065 }, { "epoch": 2.4906580016246953, "grad_norm": 0.3134961724281311, "learning_rate": 8.53021398827239e-07, "loss": 0.3969, "step": 3066 }, { "epoch": 2.4914703493095045, "grad_norm": 0.31902387738227844, "learning_rate": 8.503824353848944e-07, "loss": 0.3789, "step": 3067 }, { "epoch": 2.4922826969943137, "grad_norm": 0.30067911744117737, "learning_rate": 8.477471808594556e-07, "loss": 0.4053, "step": 3068 }, { "epoch": 2.493095044679123, "grad_norm": 0.32964080572128296, "learning_rate": 8.451156376063096e-07, "loss": 0.4123, "step": 3069 }, { "epoch": 2.4939073923639317, "grad_norm": 0.3374527394771576, "learning_rate": 8.424878079775195e-07, "loss": 0.4003, "step": 3070 }, { "epoch": 2.494719740048741, "grad_norm": 0.3012656271457672, "learning_rate": 8.398636943218353e-07, "loss": 0.3883, "step": 3071 }, { "epoch": 2.49553208773355, "grad_norm": 0.3074401617050171, "learning_rate": 8.37243298984679e-07, "loss": 0.4055, "step": 3072 }, { "epoch": 2.496344435418359, "grad_norm": 0.3277352452278137, "learning_rate": 8.346266243081563e-07, "loss": 0.4446, "step": 3073 }, { "epoch": 2.497156783103168, "grad_norm": 0.3170638382434845, "learning_rate": 8.320136726310413e-07, "loss": 0.3809, "step": 3074 }, { "epoch": 2.497969130787977, "grad_norm": 0.31078654527664185, "learning_rate": 8.294044462887852e-07, "loss": 0.3845, "step": 3075 }, { "epoch": 2.4987814784727864, "grad_norm": 0.33386722207069397, "learning_rate": 8.267989476135052e-07, "loss": 0.39, "step": 3076 }, { "epoch": 2.4995938261575956, "grad_norm": 0.33588629961013794, "learning_rate": 8.24197178933992e-07, "loss": 0.3896, "step": 3077 }, { "epoch": 2.5004061738424044, "grad_norm": 0.3191992938518524, "learning_rate": 8.215991425756959e-07, "loss": 0.3865, "step": 3078 }, { "epoch": 2.5012185215272136, "grad_norm": 0.33356109261512756, "learning_rate": 8.190048408607376e-07, "loss": 0.3933, "step": 3079 }, { "epoch": 2.502030869212023, "grad_norm": 0.301811546087265, "learning_rate": 8.164142761078941e-07, "loss": 0.4275, "step": 3080 }, { "epoch": 2.502843216896832, "grad_norm": 0.3129429221153259, "learning_rate": 8.138274506326088e-07, "loss": 0.4046, "step": 3081 }, { "epoch": 2.503655564581641, "grad_norm": 0.30119991302490234, "learning_rate": 8.112443667469754e-07, "loss": 0.4069, "step": 3082 }, { "epoch": 2.50446791226645, "grad_norm": 0.3381614685058594, "learning_rate": 8.08665026759749e-07, "loss": 0.3866, "step": 3083 }, { "epoch": 2.505280259951259, "grad_norm": 0.34578147530555725, "learning_rate": 8.060894329763386e-07, "loss": 0.4152, "step": 3084 }, { "epoch": 2.5060926076360683, "grad_norm": 0.30772891640663147, "learning_rate": 8.035175876987994e-07, "loss": 0.3914, "step": 3085 }, { "epoch": 2.506904955320877, "grad_norm": 0.3040609657764435, "learning_rate": 8.009494932258427e-07, "loss": 0.4025, "step": 3086 }, { "epoch": 2.5077173030056863, "grad_norm": 0.32561513781547546, "learning_rate": 7.983851518528224e-07, "loss": 0.423, "step": 3087 }, { "epoch": 2.5085296506904955, "grad_norm": 0.3538568317890167, "learning_rate": 7.958245658717411e-07, "loss": 0.4009, "step": 3088 }, { "epoch": 2.5093419983753047, "grad_norm": 0.3462631106376648, "learning_rate": 7.932677375712423e-07, "loss": 0.4059, "step": 3089 }, { "epoch": 2.510154346060114, "grad_norm": 0.34786859154701233, "learning_rate": 7.907146692366135e-07, "loss": 0.4009, "step": 3090 }, { "epoch": 2.5109666937449227, "grad_norm": 0.323114812374115, "learning_rate": 7.881653631497793e-07, "loss": 0.3906, "step": 3091 }, { "epoch": 2.511779041429732, "grad_norm": 0.3402250409126282, "learning_rate": 7.856198215893035e-07, "loss": 0.3998, "step": 3092 }, { "epoch": 2.512591389114541, "grad_norm": 0.31516167521476746, "learning_rate": 7.830780468303828e-07, "loss": 0.3665, "step": 3093 }, { "epoch": 2.5134037367993503, "grad_norm": 0.33149728178977966, "learning_rate": 7.805400411448505e-07, "loss": 0.4046, "step": 3094 }, { "epoch": 2.5142160844841595, "grad_norm": 0.3011018931865692, "learning_rate": 7.78005806801167e-07, "loss": 0.4404, "step": 3095 }, { "epoch": 2.515028432168968, "grad_norm": 0.31024956703186035, "learning_rate": 7.754753460644266e-07, "loss": 0.4044, "step": 3096 }, { "epoch": 2.5158407798537774, "grad_norm": 0.33291688561439514, "learning_rate": 7.729486611963449e-07, "loss": 0.3878, "step": 3097 }, { "epoch": 2.5166531275385866, "grad_norm": 0.29554852843284607, "learning_rate": 7.704257544552696e-07, "loss": 0.3924, "step": 3098 }, { "epoch": 2.5174654752233954, "grad_norm": 0.34418779611587524, "learning_rate": 7.67906628096165e-07, "loss": 0.419, "step": 3099 }, { "epoch": 2.5182778229082046, "grad_norm": 0.3202332556247711, "learning_rate": 7.653912843706219e-07, "loss": 0.4269, "step": 3100 }, { "epoch": 2.5190901705930138, "grad_norm": 0.3088947832584381, "learning_rate": 7.628797255268439e-07, "loss": 0.4393, "step": 3101 }, { "epoch": 2.519902518277823, "grad_norm": 0.3221810758113861, "learning_rate": 7.603719538096604e-07, "loss": 0.4143, "step": 3102 }, { "epoch": 2.520714865962632, "grad_norm": 0.31965991854667664, "learning_rate": 7.57867971460507e-07, "loss": 0.3764, "step": 3103 }, { "epoch": 2.521527213647441, "grad_norm": 0.3272996246814728, "learning_rate": 7.553677807174392e-07, "loss": 0.3916, "step": 3104 }, { "epoch": 2.52233956133225, "grad_norm": 0.2941678464412689, "learning_rate": 7.52871383815118e-07, "loss": 0.384, "step": 3105 }, { "epoch": 2.5231519090170593, "grad_norm": 0.30207201838493347, "learning_rate": 7.503787829848191e-07, "loss": 0.3948, "step": 3106 }, { "epoch": 2.5239642567018685, "grad_norm": 0.32152262330055237, "learning_rate": 7.478899804544193e-07, "loss": 0.4141, "step": 3107 }, { "epoch": 2.5247766043866777, "grad_norm": 0.27521705627441406, "learning_rate": 7.454049784484063e-07, "loss": 0.3772, "step": 3108 }, { "epoch": 2.5255889520714865, "grad_norm": 0.29869428277015686, "learning_rate": 7.429237791878652e-07, "loss": 0.3877, "step": 3109 }, { "epoch": 2.5264012997562957, "grad_norm": 0.31091058254241943, "learning_rate": 7.40446384890487e-07, "loss": 0.4069, "step": 3110 }, { "epoch": 2.527213647441105, "grad_norm": 0.3161613941192627, "learning_rate": 7.379727977705614e-07, "loss": 0.3928, "step": 3111 }, { "epoch": 2.5280259951259136, "grad_norm": 0.31741422414779663, "learning_rate": 7.355030200389706e-07, "loss": 0.3965, "step": 3112 }, { "epoch": 2.528838342810723, "grad_norm": 0.31872835755348206, "learning_rate": 7.330370539031978e-07, "loss": 0.4133, "step": 3113 }, { "epoch": 2.529650690495532, "grad_norm": 0.29630106687545776, "learning_rate": 7.305749015673153e-07, "loss": 0.3944, "step": 3114 }, { "epoch": 2.5304630381803412, "grad_norm": 0.31562790274620056, "learning_rate": 7.281165652319893e-07, "loss": 0.3902, "step": 3115 }, { "epoch": 2.5312753858651504, "grad_norm": 0.3041161894798279, "learning_rate": 7.25662047094473e-07, "loss": 0.4394, "step": 3116 }, { "epoch": 2.532087733549959, "grad_norm": 0.2840907573699951, "learning_rate": 7.232113493486099e-07, "loss": 0.3778, "step": 3117 }, { "epoch": 2.5329000812347684, "grad_norm": 0.3163035213947296, "learning_rate": 7.207644741848241e-07, "loss": 0.3775, "step": 3118 }, { "epoch": 2.5337124289195776, "grad_norm": 0.32506245374679565, "learning_rate": 7.183214237901287e-07, "loss": 0.4272, "step": 3119 }, { "epoch": 2.534524776604387, "grad_norm": 0.3089554011821747, "learning_rate": 7.158822003481136e-07, "loss": 0.4111, "step": 3120 }, { "epoch": 2.535337124289196, "grad_norm": 0.3121393918991089, "learning_rate": 7.134468060389521e-07, "loss": 0.4037, "step": 3121 }, { "epoch": 2.5361494719740048, "grad_norm": 0.34509462118148804, "learning_rate": 7.110152430393902e-07, "loss": 0.3831, "step": 3122 }, { "epoch": 2.536961819658814, "grad_norm": 0.2968103289604187, "learning_rate": 7.085875135227555e-07, "loss": 0.404, "step": 3123 }, { "epoch": 2.537774167343623, "grad_norm": 0.29881343245506287, "learning_rate": 7.061636196589433e-07, "loss": 0.4062, "step": 3124 }, { "epoch": 2.538586515028432, "grad_norm": 0.31826266646385193, "learning_rate": 7.037435636144252e-07, "loss": 0.3973, "step": 3125 }, { "epoch": 2.539398862713241, "grad_norm": 0.2907699644565582, "learning_rate": 7.013273475522392e-07, "loss": 0.4059, "step": 3126 }, { "epoch": 2.5402112103980503, "grad_norm": 0.2929190397262573, "learning_rate": 6.989149736319939e-07, "loss": 0.4197, "step": 3127 }, { "epoch": 2.5410235580828595, "grad_norm": 0.294135719537735, "learning_rate": 6.965064440098629e-07, "loss": 0.3912, "step": 3128 }, { "epoch": 2.5418359057676687, "grad_norm": 0.34952181577682495, "learning_rate": 6.941017608385813e-07, "loss": 0.3833, "step": 3129 }, { "epoch": 2.5426482534524775, "grad_norm": 0.3249806761741638, "learning_rate": 6.917009262674512e-07, "loss": 0.406, "step": 3130 }, { "epoch": 2.5434606011372867, "grad_norm": 0.30944114923477173, "learning_rate": 6.893039424423298e-07, "loss": 0.405, "step": 3131 }, { "epoch": 2.544272948822096, "grad_norm": 0.3022415339946747, "learning_rate": 6.869108115056361e-07, "loss": 0.4198, "step": 3132 }, { "epoch": 2.545085296506905, "grad_norm": 0.3062586784362793, "learning_rate": 6.845215355963425e-07, "loss": 0.4015, "step": 3133 }, { "epoch": 2.5458976441917143, "grad_norm": 0.3151061534881592, "learning_rate": 6.821361168499796e-07, "loss": 0.4063, "step": 3134 }, { "epoch": 2.546709991876523, "grad_norm": 0.3095124065876007, "learning_rate": 6.797545573986253e-07, "loss": 0.3959, "step": 3135 }, { "epoch": 2.5475223395613322, "grad_norm": 0.31577879190444946, "learning_rate": 6.77376859370914e-07, "loss": 0.4374, "step": 3136 }, { "epoch": 2.5483346872461414, "grad_norm": 0.3263756334781647, "learning_rate": 6.750030248920231e-07, "loss": 0.3881, "step": 3137 }, { "epoch": 2.54914703493095, "grad_norm": 0.2995437979698181, "learning_rate": 6.726330560836813e-07, "loss": 0.4035, "step": 3138 }, { "epoch": 2.5499593826157594, "grad_norm": 0.2869778573513031, "learning_rate": 6.702669550641589e-07, "loss": 0.391, "step": 3139 }, { "epoch": 2.5507717303005686, "grad_norm": 0.30037182569503784, "learning_rate": 6.679047239482711e-07, "loss": 0.4073, "step": 3140 }, { "epoch": 2.551584077985378, "grad_norm": 0.2942163944244385, "learning_rate": 6.655463648473736e-07, "loss": 0.3945, "step": 3141 }, { "epoch": 2.552396425670187, "grad_norm": 0.3160702586174011, "learning_rate": 6.631918798693632e-07, "loss": 0.3917, "step": 3142 }, { "epoch": 2.5532087733549957, "grad_norm": 0.3064331114292145, "learning_rate": 6.60841271118669e-07, "loss": 0.3954, "step": 3143 }, { "epoch": 2.554021121039805, "grad_norm": 0.32766497135162354, "learning_rate": 6.584945406962617e-07, "loss": 0.4202, "step": 3144 }, { "epoch": 2.554833468724614, "grad_norm": 0.321922242641449, "learning_rate": 6.561516906996395e-07, "loss": 0.3931, "step": 3145 }, { "epoch": 2.5556458164094233, "grad_norm": 0.31717565655708313, "learning_rate": 6.53812723222838e-07, "loss": 0.4155, "step": 3146 }, { "epoch": 2.5564581640942325, "grad_norm": 0.3018951117992401, "learning_rate": 6.514776403564183e-07, "loss": 0.4059, "step": 3147 }, { "epoch": 2.5572705117790413, "grad_norm": 0.31250742077827454, "learning_rate": 6.491464441874728e-07, "loss": 0.393, "step": 3148 }, { "epoch": 2.5580828594638505, "grad_norm": 0.3197510242462158, "learning_rate": 6.468191367996163e-07, "loss": 0.3969, "step": 3149 }, { "epoch": 2.5588952071486597, "grad_norm": 0.3230994641780853, "learning_rate": 6.444957202729918e-07, "loss": 0.3784, "step": 3150 }, { "epoch": 2.5597075548334685, "grad_norm": 0.31886857748031616, "learning_rate": 6.421761966842604e-07, "loss": 0.4, "step": 3151 }, { "epoch": 2.5605199025182777, "grad_norm": 0.32426488399505615, "learning_rate": 6.398605681066083e-07, "loss": 0.4073, "step": 3152 }, { "epoch": 2.561332250203087, "grad_norm": 0.30014169216156006, "learning_rate": 6.37548836609736e-07, "loss": 0.3959, "step": 3153 }, { "epoch": 2.562144597887896, "grad_norm": 0.3184618055820465, "learning_rate": 6.352410042598639e-07, "loss": 0.4059, "step": 3154 }, { "epoch": 2.5629569455727053, "grad_norm": 0.2787325084209442, "learning_rate": 6.329370731197276e-07, "loss": 0.4166, "step": 3155 }, { "epoch": 2.563769293257514, "grad_norm": 0.3088829517364502, "learning_rate": 6.306370452485716e-07, "loss": 0.4086, "step": 3156 }, { "epoch": 2.564581640942323, "grad_norm": 0.29643672704696655, "learning_rate": 6.283409227021575e-07, "loss": 0.3962, "step": 3157 }, { "epoch": 2.5653939886271324, "grad_norm": 0.28540778160095215, "learning_rate": 6.260487075327504e-07, "loss": 0.4057, "step": 3158 }, { "epoch": 2.5662063363119416, "grad_norm": 0.30703288316726685, "learning_rate": 6.237604017891291e-07, "loss": 0.3804, "step": 3159 }, { "epoch": 2.567018683996751, "grad_norm": 0.31821268796920776, "learning_rate": 6.214760075165721e-07, "loss": 0.4091, "step": 3160 }, { "epoch": 2.5678310316815596, "grad_norm": 0.3305492699146271, "learning_rate": 6.191955267568678e-07, "loss": 0.4012, "step": 3161 }, { "epoch": 2.5686433793663688, "grad_norm": 0.28817540407180786, "learning_rate": 6.169189615483007e-07, "loss": 0.3798, "step": 3162 }, { "epoch": 2.569455727051178, "grad_norm": 0.33920520544052124, "learning_rate": 6.146463139256609e-07, "loss": 0.3696, "step": 3163 }, { "epoch": 2.5702680747359867, "grad_norm": 0.29678088426589966, "learning_rate": 6.12377585920233e-07, "loss": 0.3924, "step": 3164 }, { "epoch": 2.5710804224207964, "grad_norm": 0.31662172079086304, "learning_rate": 6.101127795598016e-07, "loss": 0.3917, "step": 3165 }, { "epoch": 2.571892770105605, "grad_norm": 0.29086583852767944, "learning_rate": 6.078518968686426e-07, "loss": 0.3856, "step": 3166 }, { "epoch": 2.5727051177904143, "grad_norm": 0.33175840973854065, "learning_rate": 6.055949398675293e-07, "loss": 0.4073, "step": 3167 }, { "epoch": 2.5735174654752235, "grad_norm": 0.34348851442337036, "learning_rate": 6.033419105737204e-07, "loss": 0.4311, "step": 3168 }, { "epoch": 2.5743298131600323, "grad_norm": 0.31555384397506714, "learning_rate": 6.010928110009711e-07, "loss": 0.376, "step": 3169 }, { "epoch": 2.5751421608448415, "grad_norm": 0.33284619450569153, "learning_rate": 5.98847643159517e-07, "loss": 0.3913, "step": 3170 }, { "epoch": 2.5759545085296507, "grad_norm": 0.3462708294391632, "learning_rate": 5.966064090560847e-07, "loss": 0.4132, "step": 3171 }, { "epoch": 2.57676685621446, "grad_norm": 0.30643415451049805, "learning_rate": 5.943691106938843e-07, "loss": 0.3897, "step": 3172 }, { "epoch": 2.577579203899269, "grad_norm": 0.31731587648391724, "learning_rate": 5.921357500726044e-07, "loss": 0.3838, "step": 3173 }, { "epoch": 2.578391551584078, "grad_norm": 0.3209911584854126, "learning_rate": 5.899063291884193e-07, "loss": 0.3848, "step": 3174 }, { "epoch": 2.579203899268887, "grad_norm": 0.3213132917881012, "learning_rate": 5.87680850033977e-07, "loss": 0.3947, "step": 3175 }, { "epoch": 2.5800162469536962, "grad_norm": 0.3072448670864105, "learning_rate": 5.854593145984067e-07, "loss": 0.4051, "step": 3176 }, { "epoch": 2.5808285946385054, "grad_norm": 0.30510398745536804, "learning_rate": 5.832417248673089e-07, "loss": 0.3872, "step": 3177 }, { "epoch": 2.5816409423233146, "grad_norm": 0.31009528040885925, "learning_rate": 5.810280828227599e-07, "loss": 0.4134, "step": 3178 }, { "epoch": 2.5824532900081234, "grad_norm": 0.2951056659221649, "learning_rate": 5.788183904433076e-07, "loss": 0.3888, "step": 3179 }, { "epoch": 2.5832656376929326, "grad_norm": 0.2965271770954132, "learning_rate": 5.766126497039698e-07, "loss": 0.3926, "step": 3180 }, { "epoch": 2.584077985377742, "grad_norm": 0.3197546601295471, "learning_rate": 5.744108625762291e-07, "loss": 0.4184, "step": 3181 }, { "epoch": 2.5848903330625506, "grad_norm": 0.3167704641819, "learning_rate": 5.722130310280394e-07, "loss": 0.3829, "step": 3182 }, { "epoch": 2.5857026807473598, "grad_norm": 0.28925126791000366, "learning_rate": 5.70019157023814e-07, "loss": 0.3965, "step": 3183 }, { "epoch": 2.586515028432169, "grad_norm": 0.3177677392959595, "learning_rate": 5.678292425244347e-07, "loss": 0.4197, "step": 3184 }, { "epoch": 2.587327376116978, "grad_norm": 0.30168697237968445, "learning_rate": 5.656432894872376e-07, "loss": 0.4054, "step": 3185 }, { "epoch": 2.5881397238017874, "grad_norm": 0.32591772079467773, "learning_rate": 5.634612998660249e-07, "loss": 0.3933, "step": 3186 }, { "epoch": 2.588952071486596, "grad_norm": 0.3192238211631775, "learning_rate": 5.612832756110498e-07, "loss": 0.3912, "step": 3187 }, { "epoch": 2.5897644191714053, "grad_norm": 0.28727754950523376, "learning_rate": 5.591092186690278e-07, "loss": 0.3912, "step": 3188 }, { "epoch": 2.5905767668562145, "grad_norm": 0.31550055742263794, "learning_rate": 5.569391309831218e-07, "loss": 0.4213, "step": 3189 }, { "epoch": 2.5913891145410237, "grad_norm": 0.3108750283718109, "learning_rate": 5.547730144929531e-07, "loss": 0.4164, "step": 3190 }, { "epoch": 2.592201462225833, "grad_norm": 0.30116260051727295, "learning_rate": 5.526108711345884e-07, "loss": 0.373, "step": 3191 }, { "epoch": 2.5930138099106417, "grad_norm": 0.30372902750968933, "learning_rate": 5.504527028405482e-07, "loss": 0.3825, "step": 3192 }, { "epoch": 2.593826157595451, "grad_norm": 0.334271103143692, "learning_rate": 5.482985115397943e-07, "loss": 0.4127, "step": 3193 }, { "epoch": 2.59463850528026, "grad_norm": 0.2973248064517975, "learning_rate": 5.461482991577399e-07, "loss": 0.391, "step": 3194 }, { "epoch": 2.595450852965069, "grad_norm": 0.30211353302001953, "learning_rate": 5.440020676162367e-07, "loss": 0.3817, "step": 3195 }, { "epoch": 2.596263200649878, "grad_norm": 0.30005940794944763, "learning_rate": 5.418598188335816e-07, "loss": 0.4088, "step": 3196 }, { "epoch": 2.5970755483346872, "grad_norm": 0.3088597059249878, "learning_rate": 5.397215547245116e-07, "loss": 0.3958, "step": 3197 }, { "epoch": 2.5978878960194964, "grad_norm": 0.29777616262435913, "learning_rate": 5.375872772002e-07, "loss": 0.3886, "step": 3198 }, { "epoch": 2.5987002437043056, "grad_norm": 0.307909220457077, "learning_rate": 5.354569881682592e-07, "loss": 0.3873, "step": 3199 }, { "epoch": 2.5995125913891144, "grad_norm": 0.30302149057388306, "learning_rate": 5.33330689532735e-07, "loss": 0.4009, "step": 3200 }, { "epoch": 2.6003249390739236, "grad_norm": 0.3177720606327057, "learning_rate": 5.312083831941084e-07, "loss": 0.3912, "step": 3201 }, { "epoch": 2.601137286758733, "grad_norm": 0.3247126638889313, "learning_rate": 5.290900710492891e-07, "loss": 0.4045, "step": 3202 }, { "epoch": 2.601949634443542, "grad_norm": 0.30737563967704773, "learning_rate": 5.269757549916216e-07, "loss": 0.4014, "step": 3203 }, { "epoch": 2.602761982128351, "grad_norm": 0.2867263853549957, "learning_rate": 5.24865436910873e-07, "loss": 0.4316, "step": 3204 }, { "epoch": 2.60357432981316, "grad_norm": 0.3193178176879883, "learning_rate": 5.227591186932418e-07, "loss": 0.3886, "step": 3205 }, { "epoch": 2.604386677497969, "grad_norm": 0.32047906517982483, "learning_rate": 5.206568022213482e-07, "loss": 0.3849, "step": 3206 }, { "epoch": 2.6051990251827783, "grad_norm": 0.3161729872226715, "learning_rate": 5.185584893742379e-07, "loss": 0.3739, "step": 3207 }, { "epoch": 2.606011372867587, "grad_norm": 0.3256489336490631, "learning_rate": 5.164641820273764e-07, "loss": 0.3843, "step": 3208 }, { "epoch": 2.6068237205523963, "grad_norm": 0.2820690870285034, "learning_rate": 5.143738820526517e-07, "loss": 0.3943, "step": 3209 }, { "epoch": 2.6076360682372055, "grad_norm": 0.3093954026699066, "learning_rate": 5.122875913183656e-07, "loss": 0.4092, "step": 3210 }, { "epoch": 2.6084484159220147, "grad_norm": 0.28831273317337036, "learning_rate": 5.102053116892414e-07, "loss": 0.3889, "step": 3211 }, { "epoch": 2.609260763606824, "grad_norm": 0.2972971200942993, "learning_rate": 5.081270450264125e-07, "loss": 0.4273, "step": 3212 }, { "epoch": 2.6100731112916327, "grad_norm": 0.29547664523124695, "learning_rate": 5.060527931874309e-07, "loss": 0.3597, "step": 3213 }, { "epoch": 2.610885458976442, "grad_norm": 0.31728243827819824, "learning_rate": 5.039825580262553e-07, "loss": 0.4216, "step": 3214 }, { "epoch": 2.611697806661251, "grad_norm": 0.31081485748291016, "learning_rate": 5.019163413932565e-07, "loss": 0.3993, "step": 3215 }, { "epoch": 2.6125101543460603, "grad_norm": 0.31164807081222534, "learning_rate": 4.998541451352134e-07, "loss": 0.4002, "step": 3216 }, { "epoch": 2.6133225020308695, "grad_norm": 0.3042896091938019, "learning_rate": 4.977959710953128e-07, "loss": 0.3888, "step": 3217 }, { "epoch": 2.614134849715678, "grad_norm": 0.3030947744846344, "learning_rate": 4.957418211131432e-07, "loss": 0.396, "step": 3218 }, { "epoch": 2.6149471974004874, "grad_norm": 0.31244611740112305, "learning_rate": 4.936916970246997e-07, "loss": 0.407, "step": 3219 }, { "epoch": 2.6157595450852966, "grad_norm": 0.31141746044158936, "learning_rate": 4.91645600662376e-07, "loss": 0.3927, "step": 3220 }, { "epoch": 2.6165718927701054, "grad_norm": 0.29062673449516296, "learning_rate": 4.896035338549693e-07, "loss": 0.3952, "step": 3221 }, { "epoch": 2.6173842404549146, "grad_norm": 0.315621554851532, "learning_rate": 4.875654984276718e-07, "loss": 0.3776, "step": 3222 }, { "epoch": 2.618196588139724, "grad_norm": 0.33212122321128845, "learning_rate": 4.855314962020746e-07, "loss": 0.4109, "step": 3223 }, { "epoch": 2.619008935824533, "grad_norm": 0.31794506311416626, "learning_rate": 4.835015289961642e-07, "loss": 0.4222, "step": 3224 }, { "epoch": 2.619821283509342, "grad_norm": 0.28830593824386597, "learning_rate": 4.81475598624318e-07, "loss": 0.4001, "step": 3225 }, { "epoch": 2.620633631194151, "grad_norm": 0.33178824186325073, "learning_rate": 4.794537068973093e-07, "loss": 0.4135, "step": 3226 }, { "epoch": 2.62144597887896, "grad_norm": 0.34291601181030273, "learning_rate": 4.774358556222974e-07, "loss": 0.395, "step": 3227 }, { "epoch": 2.6222583265637693, "grad_norm": 0.3098255395889282, "learning_rate": 4.7542204660283363e-07, "loss": 0.4076, "step": 3228 }, { "epoch": 2.6230706742485785, "grad_norm": 0.31660640239715576, "learning_rate": 4.7341228163885424e-07, "loss": 0.4128, "step": 3229 }, { "epoch": 2.6238830219333877, "grad_norm": 0.2963278591632843, "learning_rate": 4.7140656252668226e-07, "loss": 0.4019, "step": 3230 }, { "epoch": 2.6246953696181965, "grad_norm": 0.2995769679546356, "learning_rate": 4.6940489105902323e-07, "loss": 0.3959, "step": 3231 }, { "epoch": 2.6255077173030057, "grad_norm": 0.29976412653923035, "learning_rate": 4.67407269024967e-07, "loss": 0.4081, "step": 3232 }, { "epoch": 2.626320064987815, "grad_norm": 0.3118802607059479, "learning_rate": 4.654136982099805e-07, "loss": 0.4175, "step": 3233 }, { "epoch": 2.6271324126726237, "grad_norm": 0.29960232973098755, "learning_rate": 4.634241803959144e-07, "loss": 0.3983, "step": 3234 }, { "epoch": 2.627944760357433, "grad_norm": 0.30799466371536255, "learning_rate": 4.6143871736099246e-07, "loss": 0.4027, "step": 3235 }, { "epoch": 2.628757108042242, "grad_norm": 0.3332696855068207, "learning_rate": 4.5945731087981783e-07, "loss": 0.4084, "step": 3236 }, { "epoch": 2.6295694557270513, "grad_norm": 0.3433394432067871, "learning_rate": 4.574799627233639e-07, "loss": 0.4211, "step": 3237 }, { "epoch": 2.6303818034118605, "grad_norm": 0.2791535258293152, "learning_rate": 4.5550667465898204e-07, "loss": 0.4167, "step": 3238 }, { "epoch": 2.631194151096669, "grad_norm": 0.31284716725349426, "learning_rate": 4.535374484503885e-07, "loss": 0.4058, "step": 3239 }, { "epoch": 2.6320064987814784, "grad_norm": 0.31008487939834595, "learning_rate": 4.515722858576743e-07, "loss": 0.4076, "step": 3240 }, { "epoch": 2.6328188464662876, "grad_norm": 0.3172628581523895, "learning_rate": 4.4961118863729725e-07, "loss": 0.3932, "step": 3241 }, { "epoch": 2.633631194151097, "grad_norm": 0.29590314626693726, "learning_rate": 4.476541585420785e-07, "loss": 0.4123, "step": 3242 }, { "epoch": 2.634443541835906, "grad_norm": 0.3164519667625427, "learning_rate": 4.457011973212083e-07, "loss": 0.4146, "step": 3243 }, { "epoch": 2.6352558895207148, "grad_norm": 0.3092989921569824, "learning_rate": 4.437523067202365e-07, "loss": 0.3948, "step": 3244 }, { "epoch": 2.636068237205524, "grad_norm": 0.2959821820259094, "learning_rate": 4.4180748848107745e-07, "loss": 0.3817, "step": 3245 }, { "epoch": 2.636880584890333, "grad_norm": 0.31607359647750854, "learning_rate": 4.398667443420029e-07, "loss": 0.3983, "step": 3246 }, { "epoch": 2.637692932575142, "grad_norm": 0.28882747888565063, "learning_rate": 4.379300760376465e-07, "loss": 0.405, "step": 3247 }, { "epoch": 2.638505280259951, "grad_norm": 0.3079407513141632, "learning_rate": 4.359974852989951e-07, "loss": 0.3961, "step": 3248 }, { "epoch": 2.6393176279447603, "grad_norm": 0.3103860020637512, "learning_rate": 4.3406897385339544e-07, "loss": 0.3892, "step": 3249 }, { "epoch": 2.6401299756295695, "grad_norm": 0.30155372619628906, "learning_rate": 4.3214454342454357e-07, "loss": 0.3916, "step": 3250 }, { "epoch": 2.6409423233143787, "grad_norm": 0.3030792474746704, "learning_rate": 4.302241957324915e-07, "loss": 0.4452, "step": 3251 }, { "epoch": 2.6417546709991875, "grad_norm": 0.30642935633659363, "learning_rate": 4.283079324936401e-07, "loss": 0.3922, "step": 3252 }, { "epoch": 2.6425670186839967, "grad_norm": 0.3115255534648895, "learning_rate": 4.263957554207415e-07, "loss": 0.3762, "step": 3253 }, { "epoch": 2.643379366368806, "grad_norm": 0.3066229820251465, "learning_rate": 4.244876662228914e-07, "loss": 0.3925, "step": 3254 }, { "epoch": 2.644191714053615, "grad_norm": 0.314411997795105, "learning_rate": 4.2258366660553885e-07, "loss": 0.3868, "step": 3255 }, { "epoch": 2.6450040617384243, "grad_norm": 0.31996914744377136, "learning_rate": 4.206837582704704e-07, "loss": 0.3813, "step": 3256 }, { "epoch": 2.645816409423233, "grad_norm": 0.3181173503398895, "learning_rate": 4.187879429158204e-07, "loss": 0.4283, "step": 3257 }, { "epoch": 2.6466287571080422, "grad_norm": 0.3007601201534271, "learning_rate": 4.1689622223606283e-07, "loss": 0.3882, "step": 3258 }, { "epoch": 2.6474411047928514, "grad_norm": 0.3106346130371094, "learning_rate": 4.1500859792201363e-07, "loss": 0.3988, "step": 3259 }, { "epoch": 2.64825345247766, "grad_norm": 0.3245786726474762, "learning_rate": 4.1312507166082425e-07, "loss": 0.3971, "step": 3260 }, { "epoch": 2.6490658001624694, "grad_norm": 0.29172417521476746, "learning_rate": 4.1124564513598755e-07, "loss": 0.3963, "step": 3261 }, { "epoch": 2.6498781478472786, "grad_norm": 0.31069791316986084, "learning_rate": 4.09370320027328e-07, "loss": 0.3939, "step": 3262 }, { "epoch": 2.650690495532088, "grad_norm": 0.29607537388801575, "learning_rate": 4.0749909801100775e-07, "loss": 0.3924, "step": 3263 }, { "epoch": 2.651502843216897, "grad_norm": 0.33308812975883484, "learning_rate": 4.0563198075951824e-07, "loss": 0.3965, "step": 3264 }, { "epoch": 2.6523151909017058, "grad_norm": 0.30926012992858887, "learning_rate": 4.037689699416858e-07, "loss": 0.3996, "step": 3265 }, { "epoch": 2.653127538586515, "grad_norm": 0.31301313638687134, "learning_rate": 4.019100672226617e-07, "loss": 0.3952, "step": 3266 }, { "epoch": 2.653939886271324, "grad_norm": 0.32488787174224854, "learning_rate": 4.000552742639302e-07, "loss": 0.44, "step": 3267 }, { "epoch": 2.6547522339561334, "grad_norm": 0.30310705304145813, "learning_rate": 3.982045927233008e-07, "loss": 0.4203, "step": 3268 }, { "epoch": 2.6555645816409426, "grad_norm": 0.32868891954421997, "learning_rate": 3.963580242549053e-07, "loss": 0.3701, "step": 3269 }, { "epoch": 2.6563769293257513, "grad_norm": 0.3003014922142029, "learning_rate": 3.9451557050920406e-07, "loss": 0.3626, "step": 3270 }, { "epoch": 2.6571892770105605, "grad_norm": 0.3035599887371063, "learning_rate": 3.9267723313297535e-07, "loss": 0.4033, "step": 3271 }, { "epoch": 2.6580016246953697, "grad_norm": 0.3023796081542969, "learning_rate": 3.9084301376932143e-07, "loss": 0.4147, "step": 3272 }, { "epoch": 2.6588139723801785, "grad_norm": 0.3231407701969147, "learning_rate": 3.8901291405766097e-07, "loss": 0.4267, "step": 3273 }, { "epoch": 2.6596263200649877, "grad_norm": 0.29578468203544617, "learning_rate": 3.8718693563373376e-07, "loss": 0.4119, "step": 3274 }, { "epoch": 2.660438667749797, "grad_norm": 0.28106704354286194, "learning_rate": 3.853650801295933e-07, "loss": 0.3916, "step": 3275 }, { "epoch": 2.661251015434606, "grad_norm": 0.3064543306827545, "learning_rate": 3.8354734917361024e-07, "loss": 0.4239, "step": 3276 }, { "epoch": 2.6620633631194153, "grad_norm": 0.2946687936782837, "learning_rate": 3.817337443904662e-07, "loss": 0.4012, "step": 3277 }, { "epoch": 2.662875710804224, "grad_norm": 0.2835948169231415, "learning_rate": 3.799242674011572e-07, "loss": 0.3928, "step": 3278 }, { "epoch": 2.6636880584890332, "grad_norm": 0.31317368149757385, "learning_rate": 3.7811891982298843e-07, "loss": 0.3861, "step": 3279 }, { "epoch": 2.6645004061738424, "grad_norm": 0.31143370270729065, "learning_rate": 3.7631770326957475e-07, "loss": 0.3973, "step": 3280 }, { "epoch": 2.6653127538586516, "grad_norm": 0.31587716937065125, "learning_rate": 3.7452061935083826e-07, "loss": 0.4065, "step": 3281 }, { "epoch": 2.666125101543461, "grad_norm": 0.30511629581451416, "learning_rate": 3.7272766967300864e-07, "loss": 0.4011, "step": 3282 }, { "epoch": 2.6669374492282696, "grad_norm": 0.3353787362575531, "learning_rate": 3.709388558386179e-07, "loss": 0.4085, "step": 3283 }, { "epoch": 2.667749796913079, "grad_norm": 0.3356029689311981, "learning_rate": 3.691541794465042e-07, "loss": 0.3786, "step": 3284 }, { "epoch": 2.668562144597888, "grad_norm": 0.3079524636268616, "learning_rate": 3.673736420918067e-07, "loss": 0.3969, "step": 3285 }, { "epoch": 2.6693744922826967, "grad_norm": 0.2913385033607483, "learning_rate": 3.65597245365964e-07, "loss": 0.3841, "step": 3286 }, { "epoch": 2.670186839967506, "grad_norm": 0.30425670742988586, "learning_rate": 3.638249908567154e-07, "loss": 0.4118, "step": 3287 }, { "epoch": 2.670999187652315, "grad_norm": 0.35078132152557373, "learning_rate": 3.6205688014809615e-07, "loss": 0.4032, "step": 3288 }, { "epoch": 2.6718115353371243, "grad_norm": 0.2772643268108368, "learning_rate": 3.602929148204398e-07, "loss": 0.3828, "step": 3289 }, { "epoch": 2.6726238830219335, "grad_norm": 0.29602205753326416, "learning_rate": 3.5853309645037306e-07, "loss": 0.3901, "step": 3290 }, { "epoch": 2.6734362307067423, "grad_norm": 0.306713342666626, "learning_rate": 3.567774266108176e-07, "loss": 0.426, "step": 3291 }, { "epoch": 2.6742485783915515, "grad_norm": 0.2945929169654846, "learning_rate": 3.550259068709844e-07, "loss": 0.3957, "step": 3292 }, { "epoch": 2.6750609260763607, "grad_norm": 0.30694904923439026, "learning_rate": 3.5327853879637994e-07, "loss": 0.4015, "step": 3293 }, { "epoch": 2.67587327376117, "grad_norm": 0.2946479916572571, "learning_rate": 3.5153532394879507e-07, "loss": 0.4014, "step": 3294 }, { "epoch": 2.676685621445979, "grad_norm": 0.32288843393325806, "learning_rate": 3.4979626388631174e-07, "loss": 0.4173, "step": 3295 }, { "epoch": 2.677497969130788, "grad_norm": 0.3076927363872528, "learning_rate": 3.4806136016329537e-07, "loss": 0.399, "step": 3296 }, { "epoch": 2.678310316815597, "grad_norm": 0.2970230281352997, "learning_rate": 3.4633061433040014e-07, "loss": 0.4054, "step": 3297 }, { "epoch": 2.6791226645004063, "grad_norm": 0.30790555477142334, "learning_rate": 3.4460402793456083e-07, "loss": 0.4006, "step": 3298 }, { "epoch": 2.679935012185215, "grad_norm": 0.2994822859764099, "learning_rate": 3.4288160251899625e-07, "loss": 0.3891, "step": 3299 }, { "epoch": 2.680747359870024, "grad_norm": 0.2875164747238159, "learning_rate": 3.4116333962320445e-07, "loss": 0.3772, "step": 3300 }, { "epoch": 2.6815597075548334, "grad_norm": 0.320717453956604, "learning_rate": 3.394492407829658e-07, "loss": 0.4307, "step": 3301 }, { "epoch": 2.6823720552396426, "grad_norm": 0.345828115940094, "learning_rate": 3.3773930753033535e-07, "loss": 0.3859, "step": 3302 }, { "epoch": 2.683184402924452, "grad_norm": 0.2971956133842468, "learning_rate": 3.360335413936489e-07, "loss": 0.396, "step": 3303 }, { "epoch": 2.6839967506092606, "grad_norm": 0.31438204646110535, "learning_rate": 3.343319438975134e-07, "loss": 0.4044, "step": 3304 }, { "epoch": 2.6848090982940698, "grad_norm": 0.30116862058639526, "learning_rate": 3.3263451656281364e-07, "loss": 0.3778, "step": 3305 }, { "epoch": 2.685621445978879, "grad_norm": 0.3077612817287445, "learning_rate": 3.3094126090670477e-07, "loss": 0.3862, "step": 3306 }, { "epoch": 2.686433793663688, "grad_norm": 0.30740857124328613, "learning_rate": 3.2925217844261503e-07, "loss": 0.4032, "step": 3307 }, { "epoch": 2.6872461413484974, "grad_norm": 0.3080366551876068, "learning_rate": 3.275672706802402e-07, "loss": 0.4238, "step": 3308 }, { "epoch": 2.688058489033306, "grad_norm": 0.31856855750083923, "learning_rate": 3.258865391255478e-07, "loss": 0.4151, "step": 3309 }, { "epoch": 2.6888708367181153, "grad_norm": 0.32007700204849243, "learning_rate": 3.242099852807695e-07, "loss": 0.3968, "step": 3310 }, { "epoch": 2.6896831844029245, "grad_norm": 0.2984292805194855, "learning_rate": 3.225376106444056e-07, "loss": 0.4168, "step": 3311 }, { "epoch": 2.6904955320877333, "grad_norm": 0.3254469633102417, "learning_rate": 3.2086941671122084e-07, "loss": 0.4242, "step": 3312 }, { "epoch": 2.6913078797725425, "grad_norm": 0.30541539192199707, "learning_rate": 3.1920540497224095e-07, "loss": 0.3771, "step": 3313 }, { "epoch": 2.6921202274573517, "grad_norm": 0.2981858551502228, "learning_rate": 3.1754557691475617e-07, "loss": 0.413, "step": 3314 }, { "epoch": 2.692932575142161, "grad_norm": 0.31853339076042175, "learning_rate": 3.158899340223154e-07, "loss": 0.404, "step": 3315 }, { "epoch": 2.69374492282697, "grad_norm": 0.35278716683387756, "learning_rate": 3.142384777747287e-07, "loss": 0.3996, "step": 3316 }, { "epoch": 2.694557270511779, "grad_norm": 0.2994425296783447, "learning_rate": 3.125912096480621e-07, "loss": 0.3963, "step": 3317 }, { "epoch": 2.695369618196588, "grad_norm": 0.2902648150920868, "learning_rate": 3.1094813111464004e-07, "loss": 0.3707, "step": 3318 }, { "epoch": 2.6961819658813972, "grad_norm": 0.2849915623664856, "learning_rate": 3.0930924364304117e-07, "loss": 0.3801, "step": 3319 }, { "epoch": 2.6969943135662064, "grad_norm": 0.2985386550426483, "learning_rate": 3.076745486981003e-07, "loss": 0.3908, "step": 3320 }, { "epoch": 2.6978066612510156, "grad_norm": 0.32798177003860474, "learning_rate": 3.060440477409005e-07, "loss": 0.4035, "step": 3321 }, { "epoch": 2.6986190089358244, "grad_norm": 0.3014770746231079, "learning_rate": 3.0441774222878194e-07, "loss": 0.3938, "step": 3322 }, { "epoch": 2.6994313566206336, "grad_norm": 0.3308066725730896, "learning_rate": 3.027956336153298e-07, "loss": 0.408, "step": 3323 }, { "epoch": 2.700243704305443, "grad_norm": 0.31381842494010925, "learning_rate": 3.011777233503821e-07, "loss": 0.4048, "step": 3324 }, { "epoch": 2.7010560519902516, "grad_norm": 0.2945843040943146, "learning_rate": 2.995640128800209e-07, "loss": 0.386, "step": 3325 }, { "epoch": 2.7018683996750608, "grad_norm": 0.31614476442337036, "learning_rate": 2.9795450364657865e-07, "loss": 0.4274, "step": 3326 }, { "epoch": 2.70268074735987, "grad_norm": 0.29646214842796326, "learning_rate": 2.963491970886273e-07, "loss": 0.387, "step": 3327 }, { "epoch": 2.703493095044679, "grad_norm": 0.30427366495132446, "learning_rate": 2.947480946409875e-07, "loss": 0.3892, "step": 3328 }, { "epoch": 2.7043054427294884, "grad_norm": 0.2937681972980499, "learning_rate": 2.9315119773472067e-07, "loss": 0.4006, "step": 3329 }, { "epoch": 2.705117790414297, "grad_norm": 0.3158201277256012, "learning_rate": 2.9155850779712723e-07, "loss": 0.4281, "step": 3330 }, { "epoch": 2.7059301380991063, "grad_norm": 0.30112743377685547, "learning_rate": 2.899700262517502e-07, "loss": 0.3857, "step": 3331 }, { "epoch": 2.7067424857839155, "grad_norm": 0.31062421202659607, "learning_rate": 2.8838575451837046e-07, "loss": 0.4003, "step": 3332 }, { "epoch": 2.7075548334687247, "grad_norm": 0.28534382581710815, "learning_rate": 2.868056940130048e-07, "loss": 0.3835, "step": 3333 }, { "epoch": 2.708367181153534, "grad_norm": 0.29555192589759827, "learning_rate": 2.852298461479075e-07, "loss": 0.4127, "step": 3334 }, { "epoch": 2.7091795288383427, "grad_norm": 0.3164225220680237, "learning_rate": 2.8365821233156677e-07, "loss": 0.3823, "step": 3335 }, { "epoch": 2.709991876523152, "grad_norm": 0.357340931892395, "learning_rate": 2.8209079396870454e-07, "loss": 0.4322, "step": 3336 }, { "epoch": 2.710804224207961, "grad_norm": 0.3169068396091461, "learning_rate": 2.8052759246027626e-07, "loss": 0.4129, "step": 3337 }, { "epoch": 2.71161657189277, "grad_norm": 0.31965601444244385, "learning_rate": 2.7896860920346605e-07, "loss": 0.416, "step": 3338 }, { "epoch": 2.7124289195775795, "grad_norm": 0.31629157066345215, "learning_rate": 2.774138455916897e-07, "loss": 0.392, "step": 3339 }, { "epoch": 2.7132412672623882, "grad_norm": 0.3119846284389496, "learning_rate": 2.758633030145891e-07, "loss": 0.4079, "step": 3340 }, { "epoch": 2.7140536149471974, "grad_norm": 0.29546210169792175, "learning_rate": 2.743169828580372e-07, "loss": 0.4101, "step": 3341 }, { "epoch": 2.7148659626320066, "grad_norm": 0.3211430609226227, "learning_rate": 2.727748865041291e-07, "loss": 0.3913, "step": 3342 }, { "epoch": 2.7156783103168154, "grad_norm": 0.2952044904232025, "learning_rate": 2.712370153311872e-07, "loss": 0.3902, "step": 3343 }, { "epoch": 2.7164906580016246, "grad_norm": 0.3123466968536377, "learning_rate": 2.6970337071375584e-07, "loss": 0.3852, "step": 3344 }, { "epoch": 2.717303005686434, "grad_norm": 0.2983590364456177, "learning_rate": 2.681739540226036e-07, "loss": 0.4189, "step": 3345 }, { "epoch": 2.718115353371243, "grad_norm": 0.3091137707233429, "learning_rate": 2.6664876662471697e-07, "loss": 0.4046, "step": 3346 }, { "epoch": 2.718927701056052, "grad_norm": 0.29970207810401917, "learning_rate": 2.6512780988330654e-07, "loss": 0.3919, "step": 3347 }, { "epoch": 2.719740048740861, "grad_norm": 0.2968297600746155, "learning_rate": 2.6361108515779777e-07, "loss": 0.4131, "step": 3348 }, { "epoch": 2.72055239642567, "grad_norm": 0.2791298031806946, "learning_rate": 2.6209859380383673e-07, "loss": 0.3997, "step": 3349 }, { "epoch": 2.7213647441104794, "grad_norm": 0.29194870591163635, "learning_rate": 2.6059033717328154e-07, "loss": 0.4077, "step": 3350 }, { "epoch": 2.722177091795288, "grad_norm": 0.3261619508266449, "learning_rate": 2.590863166142105e-07, "loss": 0.387, "step": 3351 }, { "epoch": 2.7229894394800978, "grad_norm": 0.32189521193504333, "learning_rate": 2.5758653347091103e-07, "loss": 0.3958, "step": 3352 }, { "epoch": 2.7238017871649065, "grad_norm": 0.30409467220306396, "learning_rate": 2.560909890838864e-07, "loss": 0.3859, "step": 3353 }, { "epoch": 2.7246141348497157, "grad_norm": 0.32048603892326355, "learning_rate": 2.5459968478985074e-07, "loss": 0.4133, "step": 3354 }, { "epoch": 2.725426482534525, "grad_norm": 0.2803785800933838, "learning_rate": 2.531126219217256e-07, "loss": 0.4008, "step": 3355 }, { "epoch": 2.7262388302193337, "grad_norm": 0.2970365285873413, "learning_rate": 2.5162980180864605e-07, "loss": 0.4077, "step": 3356 }, { "epoch": 2.727051177904143, "grad_norm": 0.2962031662464142, "learning_rate": 2.5015122577595084e-07, "loss": 0.4067, "step": 3357 }, { "epoch": 2.727863525588952, "grad_norm": 0.2888890504837036, "learning_rate": 2.48676895145189e-07, "loss": 0.3986, "step": 3358 }, { "epoch": 2.7286758732737613, "grad_norm": 0.29607632756233215, "learning_rate": 2.472068112341114e-07, "loss": 0.4034, "step": 3359 }, { "epoch": 2.7294882209585705, "grad_norm": 0.317116916179657, "learning_rate": 2.4574097535667696e-07, "loss": 0.3951, "step": 3360 }, { "epoch": 2.7303005686433792, "grad_norm": 0.3157961070537567, "learning_rate": 2.4427938882304383e-07, "loss": 0.4275, "step": 3361 }, { "epoch": 2.7311129163281884, "grad_norm": 0.3267480134963989, "learning_rate": 2.428220529395753e-07, "loss": 0.4186, "step": 3362 }, { "epoch": 2.7319252640129976, "grad_norm": 0.3056676983833313, "learning_rate": 2.413689690088339e-07, "loss": 0.4052, "step": 3363 }, { "epoch": 2.732737611697807, "grad_norm": 0.321206659078598, "learning_rate": 2.399201383295824e-07, "loss": 0.4106, "step": 3364 }, { "epoch": 2.733549959382616, "grad_norm": 0.3254542648792267, "learning_rate": 2.3847556219678105e-07, "loss": 0.4207, "step": 3365 }, { "epoch": 2.734362307067425, "grad_norm": 0.3009594976902008, "learning_rate": 2.370352419015892e-07, "loss": 0.403, "step": 3366 }, { "epoch": 2.735174654752234, "grad_norm": 0.28849220275878906, "learning_rate": 2.3559917873135996e-07, "loss": 0.3918, "step": 3367 }, { "epoch": 2.735987002437043, "grad_norm": 0.31123340129852295, "learning_rate": 2.3416737396964373e-07, "loss": 0.3731, "step": 3368 }, { "epoch": 2.736799350121852, "grad_norm": 0.3478071987628937, "learning_rate": 2.3273982889618186e-07, "loss": 0.4247, "step": 3369 }, { "epoch": 2.737611697806661, "grad_norm": 0.3170619010925293, "learning_rate": 2.3131654478691258e-07, "loss": 0.4058, "step": 3370 }, { "epoch": 2.7384240454914703, "grad_norm": 0.29369670152664185, "learning_rate": 2.2989752291396218e-07, "loss": 0.3912, "step": 3371 }, { "epoch": 2.7392363931762795, "grad_norm": 0.30892622470855713, "learning_rate": 2.2848276454564943e-07, "loss": 0.3928, "step": 3372 }, { "epoch": 2.7400487408610887, "grad_norm": 0.3214389681816101, "learning_rate": 2.270722709464801e-07, "loss": 0.4096, "step": 3373 }, { "epoch": 2.7408610885458975, "grad_norm": 0.3106355369091034, "learning_rate": 2.2566604337715127e-07, "loss": 0.3824, "step": 3374 }, { "epoch": 2.7416734362307067, "grad_norm": 0.3384730815887451, "learning_rate": 2.242640830945436e-07, "loss": 0.4186, "step": 3375 }, { "epoch": 2.742485783915516, "grad_norm": 0.3003099858760834, "learning_rate": 2.228663913517265e-07, "loss": 0.4249, "step": 3376 }, { "epoch": 2.743298131600325, "grad_norm": 0.3158052861690521, "learning_rate": 2.2147296939795226e-07, "loss": 0.4014, "step": 3377 }, { "epoch": 2.7441104792851343, "grad_norm": 0.3097529113292694, "learning_rate": 2.2008381847865911e-07, "loss": 0.4122, "step": 3378 }, { "epoch": 2.744922826969943, "grad_norm": 0.30009838938713074, "learning_rate": 2.1869893983546442e-07, "loss": 0.4005, "step": 3379 }, { "epoch": 2.7457351746547523, "grad_norm": 0.32517653703689575, "learning_rate": 2.173183347061697e-07, "loss": 0.4041, "step": 3380 }, { "epoch": 2.7465475223395615, "grad_norm": 0.2956697344779968, "learning_rate": 2.1594200432475675e-07, "loss": 0.4026, "step": 3381 }, { "epoch": 2.74735987002437, "grad_norm": 0.31296494603157043, "learning_rate": 2.145699499213849e-07, "loss": 0.4039, "step": 3382 }, { "epoch": 2.7481722177091794, "grad_norm": 0.3011780381202698, "learning_rate": 2.132021727223943e-07, "loss": 0.3963, "step": 3383 }, { "epoch": 2.7489845653939886, "grad_norm": 0.2956334054470062, "learning_rate": 2.1183867395029812e-07, "loss": 0.4121, "step": 3384 }, { "epoch": 2.749796913078798, "grad_norm": 0.3177146017551422, "learning_rate": 2.1047945482378985e-07, "loss": 0.4019, "step": 3385 }, { "epoch": 2.750609260763607, "grad_norm": 0.2930009961128235, "learning_rate": 2.091245165577349e-07, "loss": 0.3995, "step": 3386 }, { "epoch": 2.7514216084484158, "grad_norm": 0.3092678189277649, "learning_rate": 2.0777386036317404e-07, "loss": 0.4036, "step": 3387 }, { "epoch": 2.752233956133225, "grad_norm": 0.2788461148738861, "learning_rate": 2.0642748744731878e-07, "loss": 0.3633, "step": 3388 }, { "epoch": 2.753046303818034, "grad_norm": 0.28698185086250305, "learning_rate": 2.050853990135554e-07, "loss": 0.3937, "step": 3389 }, { "epoch": 2.7538586515028434, "grad_norm": 0.3077094256877899, "learning_rate": 2.0374759626143714e-07, "loss": 0.4167, "step": 3390 }, { "epoch": 2.7546709991876526, "grad_norm": 0.2900007665157318, "learning_rate": 2.0241408038668976e-07, "loss": 0.3834, "step": 3391 }, { "epoch": 2.7554833468724613, "grad_norm": 0.3011857569217682, "learning_rate": 2.0108485258120536e-07, "loss": 0.4017, "step": 3392 }, { "epoch": 2.7562956945572705, "grad_norm": 0.3249492943286896, "learning_rate": 1.9975991403304529e-07, "loss": 0.3735, "step": 3393 }, { "epoch": 2.7571080422420797, "grad_norm": 0.3135182559490204, "learning_rate": 1.9843926592643392e-07, "loss": 0.4112, "step": 3394 }, { "epoch": 2.7579203899268885, "grad_norm": 0.30818063020706177, "learning_rate": 1.9712290944176537e-07, "loss": 0.4134, "step": 3395 }, { "epoch": 2.7587327376116977, "grad_norm": 0.2930564880371094, "learning_rate": 1.9581084575559352e-07, "loss": 0.3956, "step": 3396 }, { "epoch": 2.759545085296507, "grad_norm": 0.29711446166038513, "learning_rate": 1.94503076040638e-07, "loss": 0.4091, "step": 3397 }, { "epoch": 2.760357432981316, "grad_norm": 0.31607839465141296, "learning_rate": 1.9319960146578055e-07, "loss": 0.4116, "step": 3398 }, { "epoch": 2.7611697806661253, "grad_norm": 0.30882686376571655, "learning_rate": 1.919004231960625e-07, "loss": 0.3727, "step": 3399 }, { "epoch": 2.761982128350934, "grad_norm": 0.3330525755882263, "learning_rate": 1.9060554239268613e-07, "loss": 0.4009, "step": 3400 }, { "epoch": 2.7627944760357432, "grad_norm": 0.3090027868747711, "learning_rate": 1.8931496021301178e-07, "loss": 0.404, "step": 3401 }, { "epoch": 2.7636068237205524, "grad_norm": 0.2895274758338928, "learning_rate": 1.8802867781055943e-07, "loss": 0.3733, "step": 3402 }, { "epoch": 2.7644191714053616, "grad_norm": 0.31758832931518555, "learning_rate": 1.8674669633500342e-07, "loss": 0.4005, "step": 3403 }, { "epoch": 2.765231519090171, "grad_norm": 0.30862265825271606, "learning_rate": 1.8546901693217655e-07, "loss": 0.393, "step": 3404 }, { "epoch": 2.7660438667749796, "grad_norm": 0.31993189454078674, "learning_rate": 1.8419564074406427e-07, "loss": 0.4228, "step": 3405 }, { "epoch": 2.766856214459789, "grad_norm": 0.30134642124176025, "learning_rate": 1.8292656890880722e-07, "loss": 0.4058, "step": 3406 }, { "epoch": 2.767668562144598, "grad_norm": 0.30298781394958496, "learning_rate": 1.8166180256069753e-07, "loss": 0.3838, "step": 3407 }, { "epoch": 2.7684809098294068, "grad_norm": 0.31892693042755127, "learning_rate": 1.8040134283018207e-07, "loss": 0.4035, "step": 3408 }, { "epoch": 2.769293257514216, "grad_norm": 0.2683424949645996, "learning_rate": 1.7914519084385407e-07, "loss": 0.3971, "step": 3409 }, { "epoch": 2.770105605199025, "grad_norm": 0.2970157861709595, "learning_rate": 1.7789334772445987e-07, "loss": 0.4018, "step": 3410 }, { "epoch": 2.7709179528838344, "grad_norm": 0.3204529583454132, "learning_rate": 1.7664581459089335e-07, "loss": 0.4037, "step": 3411 }, { "epoch": 2.7717303005686436, "grad_norm": 0.2935613989830017, "learning_rate": 1.754025925581959e-07, "loss": 0.408, "step": 3412 }, { "epoch": 2.7725426482534523, "grad_norm": 0.31291264295578003, "learning_rate": 1.7416368273755646e-07, "loss": 0.4141, "step": 3413 }, { "epoch": 2.7733549959382615, "grad_norm": 0.33215782046318054, "learning_rate": 1.7292908623630866e-07, "loss": 0.3968, "step": 3414 }, { "epoch": 2.7741673436230707, "grad_norm": 0.32188722491264343, "learning_rate": 1.7169880415793206e-07, "loss": 0.4144, "step": 3415 }, { "epoch": 2.77497969130788, "grad_norm": 0.2978248596191406, "learning_rate": 1.7047283760204925e-07, "loss": 0.401, "step": 3416 }, { "epoch": 2.775792038992689, "grad_norm": 0.3360259234905243, "learning_rate": 1.6925118766442538e-07, "loss": 0.3971, "step": 3417 }, { "epoch": 2.776604386677498, "grad_norm": 0.3303055763244629, "learning_rate": 1.6803385543696872e-07, "loss": 0.4006, "step": 3418 }, { "epoch": 2.777416734362307, "grad_norm": 0.3225705623626709, "learning_rate": 1.6682084200772663e-07, "loss": 0.3888, "step": 3419 }, { "epoch": 2.7782290820471163, "grad_norm": 0.2911071479320526, "learning_rate": 1.65612148460888e-07, "loss": 0.3899, "step": 3420 }, { "epoch": 2.779041429731925, "grad_norm": 0.31420353055000305, "learning_rate": 1.6440777587677924e-07, "loss": 0.4233, "step": 3421 }, { "epoch": 2.7798537774167342, "grad_norm": 0.32015180587768555, "learning_rate": 1.6320772533186536e-07, "loss": 0.4061, "step": 3422 }, { "epoch": 2.7806661251015434, "grad_norm": 0.2907616198062897, "learning_rate": 1.6201199789874833e-07, "loss": 0.3942, "step": 3423 }, { "epoch": 2.7814784727863526, "grad_norm": 0.2958763539791107, "learning_rate": 1.6082059464616662e-07, "loss": 0.4241, "step": 3424 }, { "epoch": 2.782290820471162, "grad_norm": 0.2994837760925293, "learning_rate": 1.5963351663899284e-07, "loss": 0.3998, "step": 3425 }, { "epoch": 2.7831031681559706, "grad_norm": 0.2951434552669525, "learning_rate": 1.5845076493823331e-07, "loss": 0.3992, "step": 3426 }, { "epoch": 2.78391551584078, "grad_norm": 0.2950953543186188, "learning_rate": 1.572723406010296e-07, "loss": 0.393, "step": 3427 }, { "epoch": 2.784727863525589, "grad_norm": 0.31192269921302795, "learning_rate": 1.5609824468065306e-07, "loss": 0.3917, "step": 3428 }, { "epoch": 2.785540211210398, "grad_norm": 0.3042844831943512, "learning_rate": 1.549284782265087e-07, "loss": 0.4014, "step": 3429 }, { "epoch": 2.7863525588952074, "grad_norm": 0.28186631202697754, "learning_rate": 1.537630422841291e-07, "loss": 0.3879, "step": 3430 }, { "epoch": 2.787164906580016, "grad_norm": 0.3175770342350006, "learning_rate": 1.526019378951793e-07, "loss": 0.4137, "step": 3431 }, { "epoch": 2.7879772542648253, "grad_norm": 0.3156166672706604, "learning_rate": 1.5144516609744975e-07, "loss": 0.3985, "step": 3432 }, { "epoch": 2.7887896019496345, "grad_norm": 0.328734427690506, "learning_rate": 1.5029272792486116e-07, "loss": 0.4033, "step": 3433 }, { "epoch": 2.7896019496344433, "grad_norm": 0.32733696699142456, "learning_rate": 1.4914462440745847e-07, "loss": 0.409, "step": 3434 }, { "epoch": 2.7904142973192525, "grad_norm": 0.31674420833587646, "learning_rate": 1.480008565714147e-07, "loss": 0.4073, "step": 3435 }, { "epoch": 2.7912266450040617, "grad_norm": 0.32609647512435913, "learning_rate": 1.4686142543902548e-07, "loss": 0.3987, "step": 3436 }, { "epoch": 2.792038992688871, "grad_norm": 0.3194616138935089, "learning_rate": 1.4572633202871222e-07, "loss": 0.397, "step": 3437 }, { "epoch": 2.79285134037368, "grad_norm": 0.30249109864234924, "learning_rate": 1.4459557735501671e-07, "loss": 0.4122, "step": 3438 }, { "epoch": 2.793663688058489, "grad_norm": 0.30561989545822144, "learning_rate": 1.4346916242860664e-07, "loss": 0.3808, "step": 3439 }, { "epoch": 2.794476035743298, "grad_norm": 0.27477607131004333, "learning_rate": 1.4234708825626664e-07, "loss": 0.4025, "step": 3440 }, { "epoch": 2.7952883834281073, "grad_norm": 0.31942781805992126, "learning_rate": 1.4122935584090447e-07, "loss": 0.3887, "step": 3441 }, { "epoch": 2.7961007311129165, "grad_norm": 0.2941511273384094, "learning_rate": 1.4011596618154655e-07, "loss": 0.4124, "step": 3442 }, { "epoch": 2.7969130787977257, "grad_norm": 0.32168495655059814, "learning_rate": 1.3900692027333628e-07, "loss": 0.3842, "step": 3443 }, { "epoch": 2.7977254264825344, "grad_norm": 0.3252430558204651, "learning_rate": 1.37902219107538e-07, "loss": 0.4186, "step": 3444 }, { "epoch": 2.7985377741673436, "grad_norm": 0.2914152145385742, "learning_rate": 1.3680186367152848e-07, "loss": 0.4126, "step": 3445 }, { "epoch": 2.799350121852153, "grad_norm": 0.3054197430610657, "learning_rate": 1.3570585494880328e-07, "loss": 0.4045, "step": 3446 }, { "epoch": 2.8001624695369616, "grad_norm": 0.2976488769054413, "learning_rate": 1.3461419391897268e-07, "loss": 0.4072, "step": 3447 }, { "epoch": 2.8009748172217708, "grad_norm": 0.32117360830307007, "learning_rate": 1.3352688155775896e-07, "loss": 0.4058, "step": 3448 }, { "epoch": 2.80178716490658, "grad_norm": 0.3083198070526123, "learning_rate": 1.3244391883700025e-07, "loss": 0.4081, "step": 3449 }, { "epoch": 2.802599512591389, "grad_norm": 0.3051091134548187, "learning_rate": 1.3136530672464564e-07, "loss": 0.4319, "step": 3450 }, { "epoch": 2.8034118602761984, "grad_norm": 0.29489460587501526, "learning_rate": 1.3029104618475397e-07, "loss": 0.3997, "step": 3451 }, { "epoch": 2.804224207961007, "grad_norm": 0.31436192989349365, "learning_rate": 1.292211381774988e-07, "loss": 0.3863, "step": 3452 }, { "epoch": 2.8050365556458163, "grad_norm": 0.2925374507904053, "learning_rate": 1.2815558365915916e-07, "loss": 0.3804, "step": 3453 }, { "epoch": 2.8058489033306255, "grad_norm": 0.2933998107910156, "learning_rate": 1.2709438358212644e-07, "loss": 0.381, "step": 3454 }, { "epoch": 2.8066612510154347, "grad_norm": 0.3094556927680969, "learning_rate": 1.2603753889489645e-07, "loss": 0.425, "step": 3455 }, { "epoch": 2.807473598700244, "grad_norm": 0.30969417095184326, "learning_rate": 1.2498505054207578e-07, "loss": 0.3991, "step": 3456 }, { "epoch": 2.8082859463850527, "grad_norm": 0.3048311173915863, "learning_rate": 1.2393691946437414e-07, "loss": 0.3955, "step": 3457 }, { "epoch": 2.809098294069862, "grad_norm": 0.32327330112457275, "learning_rate": 1.228931465986094e-07, "loss": 0.4344, "step": 3458 }, { "epoch": 2.809910641754671, "grad_norm": 0.2948092222213745, "learning_rate": 1.2185373287770254e-07, "loss": 0.4103, "step": 3459 }, { "epoch": 2.81072298943948, "grad_norm": 0.2852489650249481, "learning_rate": 1.2081867923067925e-07, "loss": 0.4022, "step": 3460 }, { "epoch": 2.811535337124289, "grad_norm": 0.31764084100723267, "learning_rate": 1.1978798658266678e-07, "loss": 0.4153, "step": 3461 }, { "epoch": 2.8123476848090982, "grad_norm": 0.3021949529647827, "learning_rate": 1.1876165585489596e-07, "loss": 0.4347, "step": 3462 }, { "epoch": 2.8131600324939074, "grad_norm": 0.3054857552051544, "learning_rate": 1.1773968796469748e-07, "loss": 0.4171, "step": 3463 }, { "epoch": 2.8139723801787166, "grad_norm": 0.2940225899219513, "learning_rate": 1.1672208382550509e-07, "loss": 0.3975, "step": 3464 }, { "epoch": 2.8147847278635254, "grad_norm": 0.3131782114505768, "learning_rate": 1.1570884434684959e-07, "loss": 0.3762, "step": 3465 }, { "epoch": 2.8155970755483346, "grad_norm": 0.2872205078601837, "learning_rate": 1.1469997043436154e-07, "loss": 0.3841, "step": 3466 }, { "epoch": 2.816409423233144, "grad_norm": 0.3269827365875244, "learning_rate": 1.1369546298977018e-07, "loss": 0.3668, "step": 3467 }, { "epoch": 2.817221770917953, "grad_norm": 0.2918517589569092, "learning_rate": 1.1269532291090124e-07, "loss": 0.4235, "step": 3468 }, { "epoch": 2.818034118602762, "grad_norm": 0.30365756154060364, "learning_rate": 1.1169955109167741e-07, "loss": 0.3869, "step": 3469 }, { "epoch": 2.818846466287571, "grad_norm": 0.2992512583732605, "learning_rate": 1.1070814842211675e-07, "loss": 0.3984, "step": 3470 }, { "epoch": 2.81965881397238, "grad_norm": 0.30304834246635437, "learning_rate": 1.0972111578833266e-07, "loss": 0.4093, "step": 3471 }, { "epoch": 2.8204711616571894, "grad_norm": 0.31514546275138855, "learning_rate": 1.0873845407253114e-07, "loss": 0.3963, "step": 3472 }, { "epoch": 2.821283509341998, "grad_norm": 0.3160663843154907, "learning_rate": 1.0776016415301405e-07, "loss": 0.4017, "step": 3473 }, { "epoch": 2.8220958570268073, "grad_norm": 0.2795169949531555, "learning_rate": 1.067862469041725e-07, "loss": 0.3991, "step": 3474 }, { "epoch": 2.8229082047116165, "grad_norm": 0.29086241126060486, "learning_rate": 1.0581670319649295e-07, "loss": 0.3917, "step": 3475 }, { "epoch": 2.8237205523964257, "grad_norm": 0.3199886083602905, "learning_rate": 1.0485153389654945e-07, "loss": 0.4071, "step": 3476 }, { "epoch": 2.824532900081235, "grad_norm": 0.29486238956451416, "learning_rate": 1.038907398670086e-07, "loss": 0.421, "step": 3477 }, { "epoch": 2.8253452477660437, "grad_norm": 0.32456088066101074, "learning_rate": 1.0293432196662512e-07, "loss": 0.3963, "step": 3478 }, { "epoch": 2.826157595450853, "grad_norm": 0.2941395342350006, "learning_rate": 1.01982281050243e-07, "loss": 0.4066, "step": 3479 }, { "epoch": 2.826969943135662, "grad_norm": 0.30660057067871094, "learning_rate": 1.0103461796879265e-07, "loss": 0.3796, "step": 3480 }, { "epoch": 2.8277822908204713, "grad_norm": 0.30303823947906494, "learning_rate": 1.000913335692949e-07, "loss": 0.3938, "step": 3481 }, { "epoch": 2.8285946385052805, "grad_norm": 0.3147822618484497, "learning_rate": 9.915242869485309e-08, "loss": 0.4253, "step": 3482 }, { "epoch": 2.8294069861900892, "grad_norm": 0.306952565908432, "learning_rate": 9.821790418465815e-08, "loss": 0.4029, "step": 3483 }, { "epoch": 2.8302193338748984, "grad_norm": 0.30850479006767273, "learning_rate": 9.728776087398583e-08, "loss": 0.3955, "step": 3484 }, { "epoch": 2.8310316815597076, "grad_norm": 0.30633047223091125, "learning_rate": 9.636199959419557e-08, "loss": 0.397, "step": 3485 }, { "epoch": 2.8318440292445164, "grad_norm": 0.3147088289260864, "learning_rate": 9.544062117273045e-08, "loss": 0.3926, "step": 3486 }, { "epoch": 2.8326563769293256, "grad_norm": 0.30108803510665894, "learning_rate": 9.452362643311619e-08, "loss": 0.4033, "step": 3487 }, { "epoch": 2.833468724614135, "grad_norm": 0.299703985452652, "learning_rate": 9.361101619495994e-08, "loss": 0.3874, "step": 3488 }, { "epoch": 2.834281072298944, "grad_norm": 0.30293330550193787, "learning_rate": 9.270279127394977e-08, "loss": 0.386, "step": 3489 }, { "epoch": 2.835093419983753, "grad_norm": 0.2977336049079895, "learning_rate": 9.179895248185522e-08, "loss": 0.4152, "step": 3490 }, { "epoch": 2.835905767668562, "grad_norm": 0.28832748532295227, "learning_rate": 9.089950062652508e-08, "loss": 0.4115, "step": 3491 }, { "epoch": 2.836718115353371, "grad_norm": 0.3197004199028015, "learning_rate": 9.000443651188628e-08, "loss": 0.4165, "step": 3492 }, { "epoch": 2.8375304630381804, "grad_norm": 0.3217703402042389, "learning_rate": 8.911376093794555e-08, "loss": 0.4013, "step": 3493 }, { "epoch": 2.8383428107229896, "grad_norm": 0.30068984627723694, "learning_rate": 8.822747470078552e-08, "loss": 0.368, "step": 3494 }, { "epoch": 2.8391551584077988, "grad_norm": 0.3162768483161926, "learning_rate": 8.734557859256698e-08, "loss": 0.4155, "step": 3495 }, { "epoch": 2.8399675060926075, "grad_norm": 0.30976566672325134, "learning_rate": 8.646807340152607e-08, "loss": 0.3957, "step": 3496 }, { "epoch": 2.8407798537774167, "grad_norm": 0.28457963466644287, "learning_rate": 8.55949599119743e-08, "loss": 0.3864, "step": 3497 }, { "epoch": 2.841592201462226, "grad_norm": 0.3086813688278198, "learning_rate": 8.47262389042991e-08, "loss": 0.3844, "step": 3498 }, { "epoch": 2.8424045491470347, "grad_norm": 0.31234869360923767, "learning_rate": 8.386191115495934e-08, "loss": 0.4049, "step": 3499 }, { "epoch": 2.843216896831844, "grad_norm": 0.30212143063545227, "learning_rate": 8.300197743649041e-08, "loss": 0.4024, "step": 3500 }, { "epoch": 2.844029244516653, "grad_norm": 0.28561127185821533, "learning_rate": 8.214643851749748e-08, "loss": 0.4069, "step": 3501 }, { "epoch": 2.8448415922014623, "grad_norm": 0.2991864085197449, "learning_rate": 8.129529516266e-08, "loss": 0.4006, "step": 3502 }, { "epoch": 2.8456539398862715, "grad_norm": 0.3205910623073578, "learning_rate": 8.044854813272662e-08, "loss": 0.3892, "step": 3503 }, { "epoch": 2.8464662875710802, "grad_norm": 0.3079811930656433, "learning_rate": 7.960619818451865e-08, "loss": 0.4007, "step": 3504 }, { "epoch": 2.8472786352558894, "grad_norm": 0.3349230885505676, "learning_rate": 7.876824607092437e-08, "loss": 0.4081, "step": 3505 }, { "epoch": 2.8480909829406986, "grad_norm": 0.29307717084884644, "learning_rate": 7.793469254090524e-08, "loss": 0.3911, "step": 3506 }, { "epoch": 2.848903330625508, "grad_norm": 0.29644641280174255, "learning_rate": 7.710553833948753e-08, "loss": 0.3955, "step": 3507 }, { "epoch": 2.849715678310317, "grad_norm": 0.3100750744342804, "learning_rate": 7.628078420776786e-08, "loss": 0.4182, "step": 3508 }, { "epoch": 2.850528025995126, "grad_norm": 0.30939981341362, "learning_rate": 7.546043088290878e-08, "loss": 0.3678, "step": 3509 }, { "epoch": 2.851340373679935, "grad_norm": 0.3169429898262024, "learning_rate": 7.464447909814043e-08, "loss": 0.3745, "step": 3510 }, { "epoch": 2.852152721364744, "grad_norm": 0.302541047334671, "learning_rate": 7.383292958275778e-08, "loss": 0.3997, "step": 3511 }, { "epoch": 2.852965069049553, "grad_norm": 0.29039695858955383, "learning_rate": 7.302578306212116e-08, "loss": 0.3882, "step": 3512 }, { "epoch": 2.8537774167343626, "grad_norm": 0.31262582540512085, "learning_rate": 7.222304025765736e-08, "loss": 0.3879, "step": 3513 }, { "epoch": 2.8545897644191713, "grad_norm": 0.291723370552063, "learning_rate": 7.142470188685413e-08, "loss": 0.3843, "step": 3514 }, { "epoch": 2.8554021121039805, "grad_norm": 0.28224101662635803, "learning_rate": 7.063076866326568e-08, "loss": 0.3855, "step": 3515 }, { "epoch": 2.8562144597887897, "grad_norm": 0.28420501947402954, "learning_rate": 6.984124129650605e-08, "loss": 0.4178, "step": 3516 }, { "epoch": 2.8570268074735985, "grad_norm": 0.301545649766922, "learning_rate": 6.905612049225352e-08, "loss": 0.4122, "step": 3517 }, { "epoch": 2.8578391551584077, "grad_norm": 0.32055583596229553, "learning_rate": 6.827540695224622e-08, "loss": 0.3909, "step": 3518 }, { "epoch": 2.858651502843217, "grad_norm": 0.31663888692855835, "learning_rate": 6.749910137428484e-08, "loss": 0.416, "step": 3519 }, { "epoch": 2.859463850528026, "grad_norm": 0.2887202501296997, "learning_rate": 6.672720445222769e-08, "loss": 0.3823, "step": 3520 }, { "epoch": 2.8602761982128353, "grad_norm": 0.31757351756095886, "learning_rate": 6.595971687599567e-08, "loss": 0.4363, "step": 3521 }, { "epoch": 2.861088545897644, "grad_norm": 0.32522252202033997, "learning_rate": 6.519663933156562e-08, "loss": 0.3937, "step": 3522 }, { "epoch": 2.8619008935824533, "grad_norm": 0.29923468828201294, "learning_rate": 6.443797250097528e-08, "loss": 0.4011, "step": 3523 }, { "epoch": 2.8627132412672625, "grad_norm": 0.3006781339645386, "learning_rate": 6.368371706231891e-08, "loss": 0.4184, "step": 3524 }, { "epoch": 2.863525588952071, "grad_norm": 0.3106078505516052, "learning_rate": 6.29338736897478e-08, "loss": 0.3872, "step": 3525 }, { "epoch": 2.864337936636881, "grad_norm": 0.3211241364479065, "learning_rate": 6.218844305346916e-08, "loss": 0.4099, "step": 3526 }, { "epoch": 2.8651502843216896, "grad_norm": 0.28679415583610535, "learning_rate": 6.144742581974839e-08, "loss": 0.3815, "step": 3527 }, { "epoch": 2.865962632006499, "grad_norm": 0.2936805486679077, "learning_rate": 6.071082265090344e-08, "loss": 0.4129, "step": 3528 }, { "epoch": 2.866774979691308, "grad_norm": 0.3079078197479248, "learning_rate": 5.99786342053088e-08, "loss": 0.4318, "step": 3529 }, { "epoch": 2.8675873273761168, "grad_norm": 0.30371934175491333, "learning_rate": 5.925086113739209e-08, "loss": 0.3995, "step": 3530 }, { "epoch": 2.868399675060926, "grad_norm": 0.287179172039032, "learning_rate": 5.852750409763575e-08, "loss": 0.3882, "step": 3531 }, { "epoch": 2.869212022745735, "grad_norm": 0.31227585673332214, "learning_rate": 5.7808563732573155e-08, "loss": 0.421, "step": 3532 }, { "epoch": 2.8700243704305444, "grad_norm": 0.2952211797237396, "learning_rate": 5.709404068479196e-08, "loss": 0.4024, "step": 3533 }, { "epoch": 2.8708367181153536, "grad_norm": 0.28711700439453125, "learning_rate": 5.638393559293076e-08, "loss": 0.3891, "step": 3534 }, { "epoch": 2.8716490658001623, "grad_norm": 0.3000580668449402, "learning_rate": 5.5678249091679624e-08, "loss": 0.4199, "step": 3535 }, { "epoch": 2.8724614134849715, "grad_norm": 0.2873418629169464, "learning_rate": 5.4976981811779575e-08, "loss": 0.3931, "step": 3536 }, { "epoch": 2.8732737611697807, "grad_norm": 0.30064740777015686, "learning_rate": 5.4280134380020336e-08, "loss": 0.4138, "step": 3537 }, { "epoch": 2.87408610885459, "grad_norm": 0.295320987701416, "learning_rate": 5.358770741924368e-08, "loss": 0.3943, "step": 3538 }, { "epoch": 2.874898456539399, "grad_norm": 0.2832261323928833, "learning_rate": 5.28997015483379e-08, "loss": 0.3927, "step": 3539 }, { "epoch": 2.875710804224208, "grad_norm": 0.30327197909355164, "learning_rate": 5.2216117382241084e-08, "loss": 0.4152, "step": 3540 }, { "epoch": 2.876523151909017, "grad_norm": 0.30703404545783997, "learning_rate": 5.1536955531938937e-08, "loss": 0.4134, "step": 3541 }, { "epoch": 2.8773354995938263, "grad_norm": 0.2998286485671997, "learning_rate": 5.086221660446422e-08, "loss": 0.3929, "step": 3542 }, { "epoch": 2.878147847278635, "grad_norm": 0.33091166615486145, "learning_rate": 5.0191901202897295e-08, "loss": 0.398, "step": 3543 }, { "epoch": 2.8789601949634442, "grad_norm": 0.2909623980522156, "learning_rate": 4.952600992636392e-08, "loss": 0.4159, "step": 3544 }, { "epoch": 2.8797725426482534, "grad_norm": 0.32443371415138245, "learning_rate": 4.8864543370035764e-08, "loss": 0.4048, "step": 3545 }, { "epoch": 2.8805848903330626, "grad_norm": 0.28715866804122925, "learning_rate": 4.820750212513048e-08, "loss": 0.4128, "step": 3546 }, { "epoch": 2.881397238017872, "grad_norm": 0.3186452388763428, "learning_rate": 4.755488677890829e-08, "loss": 0.4061, "step": 3547 }, { "epoch": 2.8822095857026806, "grad_norm": 0.31681033968925476, "learning_rate": 4.6906697914677055e-08, "loss": 0.3888, "step": 3548 }, { "epoch": 2.88302193338749, "grad_norm": 0.2775630056858063, "learning_rate": 4.626293611178445e-08, "loss": 0.3959, "step": 3549 }, { "epoch": 2.883834281072299, "grad_norm": 0.3197926878929138, "learning_rate": 4.5623601945624096e-08, "loss": 0.4024, "step": 3550 }, { "epoch": 2.884646628757108, "grad_norm": 0.3088032901287079, "learning_rate": 4.498869598763056e-08, "loss": 0.4328, "step": 3551 }, { "epoch": 2.8854589764419174, "grad_norm": 0.33018359541893005, "learning_rate": 4.435821880528157e-08, "loss": 0.3716, "step": 3552 }, { "epoch": 2.886271324126726, "grad_norm": 0.3054960370063782, "learning_rate": 4.373217096209526e-08, "loss": 0.4018, "step": 3553 }, { "epoch": 2.8870836718115354, "grad_norm": 0.3194845914840698, "learning_rate": 4.3110553017631786e-08, "loss": 0.3917, "step": 3554 }, { "epoch": 2.8878960194963446, "grad_norm": 0.29551318287849426, "learning_rate": 4.249336552749172e-08, "loss": 0.3795, "step": 3555 }, { "epoch": 2.8887083671811533, "grad_norm": 0.295380562543869, "learning_rate": 4.188060904331548e-08, "loss": 0.3841, "step": 3556 }, { "epoch": 2.8895207148659625, "grad_norm": 0.286154180765152, "learning_rate": 4.127228411278383e-08, "loss": 0.4061, "step": 3557 }, { "epoch": 2.8903330625507717, "grad_norm": 0.2858513593673706, "learning_rate": 4.0668391279614085e-08, "loss": 0.4031, "step": 3558 }, { "epoch": 2.891145410235581, "grad_norm": 0.2881379723548889, "learning_rate": 4.0068931083566134e-08, "loss": 0.3866, "step": 3559 }, { "epoch": 2.89195775792039, "grad_norm": 0.3140018880367279, "learning_rate": 3.947390406043472e-08, "loss": 0.3937, "step": 3560 }, { "epoch": 2.892770105605199, "grad_norm": 0.29443052411079407, "learning_rate": 3.8883310742053316e-08, "loss": 0.3819, "step": 3561 }, { "epoch": 2.893582453290008, "grad_norm": 0.31405800580978394, "learning_rate": 3.8297151656293e-08, "loss": 0.3988, "step": 3562 }, { "epoch": 2.8943948009748173, "grad_norm": 0.30603235960006714, "learning_rate": 3.771542732706135e-08, "loss": 0.3803, "step": 3563 }, { "epoch": 2.8952071486596265, "grad_norm": 0.30026859045028687, "learning_rate": 3.713813827430135e-08, "loss": 0.3988, "step": 3564 }, { "epoch": 2.8960194963444357, "grad_norm": 0.2943122684955597, "learning_rate": 3.656528501399359e-08, "loss": 0.4243, "step": 3565 }, { "epoch": 2.8968318440292444, "grad_norm": 0.2953970730304718, "learning_rate": 3.599686805815128e-08, "loss": 0.3968, "step": 3566 }, { "epoch": 2.8976441917140536, "grad_norm": 0.3032015562057495, "learning_rate": 3.543288791482469e-08, "loss": 0.3704, "step": 3567 }, { "epoch": 2.898456539398863, "grad_norm": 0.3114636242389679, "learning_rate": 3.4873345088097255e-08, "loss": 0.3917, "step": 3568 }, { "epoch": 2.8992688870836716, "grad_norm": 0.3221556842327118, "learning_rate": 3.4318240078087816e-08, "loss": 0.3697, "step": 3569 }, { "epoch": 2.900081234768481, "grad_norm": 0.2832164466381073, "learning_rate": 3.3767573380946717e-08, "loss": 0.3858, "step": 3570 }, { "epoch": 2.90089358245329, "grad_norm": 0.32258501648902893, "learning_rate": 3.3221345488859134e-08, "loss": 0.3746, "step": 3571 }, { "epoch": 2.901705930138099, "grad_norm": 0.31264662742614746, "learning_rate": 3.267955689004176e-08, "loss": 0.384, "step": 3572 }, { "epoch": 2.9025182778229084, "grad_norm": 0.3008060157299042, "learning_rate": 3.214220806874335e-08, "loss": 0.3822, "step": 3573 }, { "epoch": 2.903330625507717, "grad_norm": 0.3327330946922302, "learning_rate": 3.1609299505245275e-08, "loss": 0.3916, "step": 3574 }, { "epoch": 2.9041429731925263, "grad_norm": 0.2949477434158325, "learning_rate": 3.1080831675859845e-08, "loss": 0.3986, "step": 3575 }, { "epoch": 2.9049553208773355, "grad_norm": 0.29786399006843567, "learning_rate": 3.055680505292979e-08, "loss": 0.4202, "step": 3576 }, { "epoch": 2.9057676685621447, "grad_norm": 0.30697497725486755, "learning_rate": 3.0037220104829344e-08, "loss": 0.4078, "step": 3577 }, { "epoch": 2.906580016246954, "grad_norm": 0.3340325653553009, "learning_rate": 2.952207729596146e-08, "loss": 0.3962, "step": 3578 }, { "epoch": 2.9073923639317627, "grad_norm": 0.3187118470668793, "learning_rate": 2.9011377086759496e-08, "loss": 0.3944, "step": 3579 }, { "epoch": 2.908204711616572, "grad_norm": 0.3035021722316742, "learning_rate": 2.85051199336861e-08, "loss": 0.3976, "step": 3580 }, { "epoch": 2.909017059301381, "grad_norm": 0.30085527896881104, "learning_rate": 2.800330628923209e-08, "loss": 0.3823, "step": 3581 }, { "epoch": 2.90982940698619, "grad_norm": 0.3071020841598511, "learning_rate": 2.7505936601918137e-08, "loss": 0.4002, "step": 3582 }, { "epoch": 2.910641754670999, "grad_norm": 0.3373470902442932, "learning_rate": 2.701301131629086e-08, "loss": 0.4095, "step": 3583 }, { "epoch": 2.9114541023558083, "grad_norm": 0.2999403774738312, "learning_rate": 2.652453087292617e-08, "loss": 0.4001, "step": 3584 }, { "epoch": 2.9122664500406175, "grad_norm": 0.29795733094215393, "learning_rate": 2.6040495708425928e-08, "loss": 0.3861, "step": 3585 }, { "epoch": 2.9130787977254267, "grad_norm": 0.30124548077583313, "learning_rate": 2.5560906255420737e-08, "loss": 0.4161, "step": 3586 }, { "epoch": 2.9138911454102354, "grad_norm": 0.31074258685112, "learning_rate": 2.5085762942564927e-08, "loss": 0.384, "step": 3587 }, { "epoch": 2.9147034930950446, "grad_norm": 0.27998292446136475, "learning_rate": 2.4615066194541015e-08, "loss": 0.3892, "step": 3588 }, { "epoch": 2.915515840779854, "grad_norm": 0.310269296169281, "learning_rate": 2.41488164320558e-08, "loss": 0.396, "step": 3589 }, { "epoch": 2.916328188464663, "grad_norm": 0.336788147687912, "learning_rate": 2.368701407184315e-08, "loss": 0.4027, "step": 3590 }, { "epoch": 2.917140536149472, "grad_norm": 0.3037167489528656, "learning_rate": 2.322965952666012e-08, "loss": 0.4262, "step": 3591 }, { "epoch": 2.917952883834281, "grad_norm": 0.32097792625427246, "learning_rate": 2.277675320528916e-08, "loss": 0.4271, "step": 3592 }, { "epoch": 2.91876523151909, "grad_norm": 0.2808135747909546, "learning_rate": 2.2328295512535902e-08, "loss": 0.3846, "step": 3593 }, { "epoch": 2.9195775792038994, "grad_norm": 0.30053290724754333, "learning_rate": 2.188428684923194e-08, "loss": 0.4063, "step": 3594 }, { "epoch": 2.920389926888708, "grad_norm": 0.29629120230674744, "learning_rate": 2.144472761222982e-08, "loss": 0.4012, "step": 3595 }, { "epoch": 2.9212022745735173, "grad_norm": 0.2765410840511322, "learning_rate": 2.10096181944075e-08, "loss": 0.4291, "step": 3596 }, { "epoch": 2.9220146222583265, "grad_norm": 0.31694161891937256, "learning_rate": 2.0578958984663888e-08, "loss": 0.3968, "step": 3597 }, { "epoch": 2.9228269699431357, "grad_norm": 0.30678892135620117, "learning_rate": 2.015275036792108e-08, "loss": 0.3941, "step": 3598 }, { "epoch": 2.923639317627945, "grad_norm": 0.29885348677635193, "learning_rate": 1.973099272512269e-08, "loss": 0.405, "step": 3599 }, { "epoch": 2.9244516653127537, "grad_norm": 0.2948175072669983, "learning_rate": 1.9313686433236057e-08, "loss": 0.3769, "step": 3600 }, { "epoch": 2.925264012997563, "grad_norm": 0.27978986501693726, "learning_rate": 1.8900831865246717e-08, "loss": 0.3934, "step": 3601 }, { "epoch": 2.926076360682372, "grad_norm": 0.30520179867744446, "learning_rate": 1.849242939016449e-08, "loss": 0.4087, "step": 3602 }, { "epoch": 2.9268887083671813, "grad_norm": 0.3178758919239044, "learning_rate": 1.8088479373016833e-08, "loss": 0.3833, "step": 3603 }, { "epoch": 2.9277010560519905, "grad_norm": 0.31621411442756653, "learning_rate": 1.768898217485493e-08, "loss": 0.4051, "step": 3604 }, { "epoch": 2.9285134037367992, "grad_norm": 0.30040243268013, "learning_rate": 1.7293938152746493e-08, "loss": 0.4002, "step": 3605 }, { "epoch": 2.9293257514216084, "grad_norm": 0.30989205837249756, "learning_rate": 1.6903347659781856e-08, "loss": 0.4014, "step": 3606 }, { "epoch": 2.9301380991064176, "grad_norm": 0.29258477687835693, "learning_rate": 1.6517211045069537e-08, "loss": 0.378, "step": 3607 }, { "epoch": 2.9309504467912264, "grad_norm": 0.2969825267791748, "learning_rate": 1.6135528653737353e-08, "loss": 0.4316, "step": 3608 }, { "epoch": 2.9317627944760356, "grad_norm": 0.3321622610092163, "learning_rate": 1.575830082693186e-08, "loss": 0.3983, "step": 3609 }, { "epoch": 2.932575142160845, "grad_norm": 0.32378455996513367, "learning_rate": 1.5385527901818907e-08, "loss": 0.4182, "step": 3610 }, { "epoch": 2.933387489845654, "grad_norm": 0.3056046962738037, "learning_rate": 1.5017210211580868e-08, "loss": 0.3998, "step": 3611 }, { "epoch": 2.934199837530463, "grad_norm": 0.29913386702537537, "learning_rate": 1.4653348085419961e-08, "loss": 0.418, "step": 3612 }, { "epoch": 2.935012185215272, "grad_norm": 0.2993413209915161, "learning_rate": 1.4293941848554926e-08, "loss": 0.3992, "step": 3613 }, { "epoch": 2.935824532900081, "grad_norm": 0.30811116099357605, "learning_rate": 1.3938991822222137e-08, "loss": 0.4015, "step": 3614 }, { "epoch": 2.9366368805848904, "grad_norm": 0.2973842918872833, "learning_rate": 1.3588498323674482e-08, "loss": 0.411, "step": 3615 }, { "epoch": 2.9374492282696996, "grad_norm": 0.2826515734195709, "learning_rate": 1.3242461666182483e-08, "loss": 0.3888, "step": 3616 }, { "epoch": 2.9382615759545088, "grad_norm": 0.30153322219848633, "learning_rate": 1.2900882159033178e-08, "loss": 0.393, "step": 3617 }, { "epoch": 2.9390739236393175, "grad_norm": 0.3052015006542206, "learning_rate": 1.2563760107528466e-08, "loss": 0.4121, "step": 3618 }, { "epoch": 2.9398862713241267, "grad_norm": 0.2943112850189209, "learning_rate": 1.223109581298787e-08, "loss": 0.4082, "step": 3619 }, { "epoch": 2.940698619008936, "grad_norm": 0.30807778239250183, "learning_rate": 1.1902889572745213e-08, "loss": 0.4095, "step": 3620 }, { "epoch": 2.9415109666937447, "grad_norm": 0.30686068534851074, "learning_rate": 1.1579141680150841e-08, "loss": 0.3904, "step": 3621 }, { "epoch": 2.942323314378554, "grad_norm": 0.31080901622772217, "learning_rate": 1.125985242456995e-08, "loss": 0.4029, "step": 3622 }, { "epoch": 2.943135662063363, "grad_norm": 0.3181033134460449, "learning_rate": 1.094502209138204e-08, "loss": 0.4007, "step": 3623 }, { "epoch": 2.9439480097481723, "grad_norm": 0.3211752474308014, "learning_rate": 1.0634650961981462e-08, "loss": 0.3923, "step": 3624 }, { "epoch": 2.9447603574329815, "grad_norm": 0.31242838501930237, "learning_rate": 1.0328739313776869e-08, "loss": 0.3992, "step": 3625 }, { "epoch": 2.9455727051177902, "grad_norm": 0.2889278829097748, "learning_rate": 1.0027287420192322e-08, "loss": 0.3894, "step": 3626 }, { "epoch": 2.9463850528025994, "grad_norm": 0.2933630645275116, "learning_rate": 9.730295550663959e-09, "loss": 0.3913, "step": 3627 }, { "epoch": 2.9471974004874086, "grad_norm": 0.32456400990486145, "learning_rate": 9.437763970642223e-09, "loss": 0.4313, "step": 3628 }, { "epoch": 2.948009748172218, "grad_norm": 0.30548420548439026, "learning_rate": 9.149692941590738e-09, "loss": 0.3935, "step": 3629 }, { "epoch": 2.948822095857027, "grad_norm": 0.3271596431732178, "learning_rate": 8.866082720987989e-09, "loss": 0.4036, "step": 3630 }, { "epoch": 2.949634443541836, "grad_norm": 0.2852364480495453, "learning_rate": 8.58693356232232e-09, "loss": 0.3943, "step": 3631 }, { "epoch": 2.950446791226645, "grad_norm": 0.30579474568367004, "learning_rate": 8.312245715097478e-09, "loss": 0.4014, "step": 3632 }, { "epoch": 2.951259138911454, "grad_norm": 0.30000039935112, "learning_rate": 8.042019424828739e-09, "loss": 0.3965, "step": 3633 }, { "epoch": 2.952071486596263, "grad_norm": 0.29591286182403564, "learning_rate": 7.776254933043458e-09, "loss": 0.3575, "step": 3634 }, { "epoch": 2.952883834281072, "grad_norm": 0.3090653717517853, "learning_rate": 7.514952477281067e-09, "loss": 0.3913, "step": 3635 }, { "epoch": 2.9536961819658814, "grad_norm": 0.36652871966362, "learning_rate": 7.258112291093078e-09, "loss": 0.4238, "step": 3636 }, { "epoch": 2.9545085296506906, "grad_norm": 0.33982953429222107, "learning_rate": 7.005734604041969e-09, "loss": 0.4158, "step": 3637 }, { "epoch": 2.9553208773354998, "grad_norm": 0.33272260427474976, "learning_rate": 6.757819641703966e-09, "loss": 0.4417, "step": 3638 }, { "epoch": 2.9561332250203085, "grad_norm": 0.31770890951156616, "learning_rate": 6.514367625662932e-09, "loss": 0.3922, "step": 3639 }, { "epoch": 2.9569455727051177, "grad_norm": 0.2908438444137573, "learning_rate": 6.275378773516472e-09, "loss": 0.3859, "step": 3640 }, { "epoch": 2.957757920389927, "grad_norm": 0.30604588985443115, "learning_rate": 6.040853298872051e-09, "loss": 0.4143, "step": 3641 }, { "epoch": 2.958570268074736, "grad_norm": 0.2890922427177429, "learning_rate": 5.8107914113475494e-09, "loss": 0.4012, "step": 3642 }, { "epoch": 2.9593826157595453, "grad_norm": 0.2838071286678314, "learning_rate": 5.585193316572368e-09, "loss": 0.3553, "step": 3643 }, { "epoch": 2.960194963444354, "grad_norm": 0.3178884983062744, "learning_rate": 5.364059216185213e-09, "loss": 0.4114, "step": 3644 }, { "epoch": 2.9610073111291633, "grad_norm": 0.2834209203720093, "learning_rate": 5.147389307834649e-09, "loss": 0.3938, "step": 3645 }, { "epoch": 2.9618196588139725, "grad_norm": 0.3206290900707245, "learning_rate": 4.935183785180209e-09, "loss": 0.409, "step": 3646 }, { "epoch": 2.9626320064987812, "grad_norm": 0.48397818207740784, "learning_rate": 4.727442837890173e-09, "loss": 0.4102, "step": 3647 }, { "epoch": 2.9634443541835904, "grad_norm": 0.36292949318885803, "learning_rate": 4.5241666516432355e-09, "loss": 0.3728, "step": 3648 }, { "epoch": 2.9642567018683996, "grad_norm": 0.32640111446380615, "learning_rate": 4.325355408127396e-09, "loss": 0.3963, "step": 3649 }, { "epoch": 2.965069049553209, "grad_norm": 0.3196983337402344, "learning_rate": 4.131009285038845e-09, "loss": 0.4004, "step": 3650 }, { "epoch": 2.965881397238018, "grad_norm": 0.31062671542167664, "learning_rate": 3.941128456083631e-09, "loss": 0.3996, "step": 3651 }, { "epoch": 2.966693744922827, "grad_norm": 0.3060777187347412, "learning_rate": 3.755713090977664e-09, "loss": 0.4421, "step": 3652 }, { "epoch": 2.967506092607636, "grad_norm": 0.2836207151412964, "learning_rate": 3.5747633554433782e-09, "loss": 0.3885, "step": 3653 }, { "epoch": 2.968318440292445, "grad_norm": 0.3095463216304779, "learning_rate": 3.3982794112141783e-09, "loss": 0.3994, "step": 3654 }, { "epoch": 2.9691307879772544, "grad_norm": 0.31077301502227783, "learning_rate": 3.226261416030552e-09, "loss": 0.409, "step": 3655 }, { "epoch": 2.9699431356620636, "grad_norm": 0.293522447347641, "learning_rate": 3.0587095236417342e-09, "loss": 0.391, "step": 3656 }, { "epoch": 2.9707554833468723, "grad_norm": 0.2877689003944397, "learning_rate": 2.8956238838051544e-09, "loss": 0.3777, "step": 3657 }, { "epoch": 2.9715678310316815, "grad_norm": 0.28567439317703247, "learning_rate": 2.737004642286434e-09, "loss": 0.3965, "step": 3658 }, { "epoch": 2.9723801787164907, "grad_norm": 0.3050907850265503, "learning_rate": 2.5828519408593877e-09, "loss": 0.3937, "step": 3659 }, { "epoch": 2.9731925264012995, "grad_norm": 0.3073970675468445, "learning_rate": 2.4331659173038038e-09, "loss": 0.3773, "step": 3660 }, { "epoch": 2.9740048740861087, "grad_norm": 0.31907224655151367, "learning_rate": 2.2879467054104378e-09, "loss": 0.3811, "step": 3661 }, { "epoch": 2.974817221770918, "grad_norm": 0.3014667332172394, "learning_rate": 2.1471944349749086e-09, "loss": 0.3984, "step": 3662 }, { "epoch": 2.975629569455727, "grad_norm": 0.28153878450393677, "learning_rate": 2.0109092318010283e-09, "loss": 0.3989, "step": 3663 }, { "epoch": 2.9764419171405363, "grad_norm": 0.2941541373729706, "learning_rate": 1.8790912177008015e-09, "loss": 0.4013, "step": 3664 }, { "epoch": 2.977254264825345, "grad_norm": 0.29759693145751953, "learning_rate": 1.7517405104922057e-09, "loss": 0.3942, "step": 3665 }, { "epoch": 2.9780666125101543, "grad_norm": 0.2826188802719116, "learning_rate": 1.6288572240014123e-09, "loss": 0.3883, "step": 3666 }, { "epoch": 2.9788789601949635, "grad_norm": 0.3148403465747833, "learning_rate": 1.510441468061119e-09, "loss": 0.409, "step": 3667 }, { "epoch": 2.9796913078797727, "grad_norm": 0.304959774017334, "learning_rate": 1.3964933485105526e-09, "loss": 0.4042, "step": 3668 }, { "epoch": 2.980503655564582, "grad_norm": 0.29764094948768616, "learning_rate": 1.287012967196577e-09, "loss": 0.3791, "step": 3669 }, { "epoch": 2.9813160032493906, "grad_norm": 0.3077965974807739, "learning_rate": 1.1820004219725845e-09, "loss": 0.393, "step": 3670 }, { "epoch": 2.9821283509342, "grad_norm": 0.3014600872993469, "learning_rate": 1.0814558066979397e-09, "loss": 0.4082, "step": 3671 }, { "epoch": 2.982940698619009, "grad_norm": 0.31328272819519043, "learning_rate": 9.853792112402005e-10, "loss": 0.4063, "step": 3672 }, { "epoch": 2.9837530463038178, "grad_norm": 0.29626762866973877, "learning_rate": 8.937707214712321e-10, "loss": 0.3851, "step": 3673 }, { "epoch": 2.984565393988627, "grad_norm": 0.2796812057495117, "learning_rate": 8.066304192710927e-10, "loss": 0.393, "step": 3674 }, { "epoch": 2.985377741673436, "grad_norm": 0.2912175953388214, "learning_rate": 7.239583825252583e-10, "loss": 0.3922, "step": 3675 }, { "epoch": 2.9861900893582454, "grad_norm": 0.2842952013015747, "learning_rate": 6.457546851262875e-10, "loss": 0.3892, "step": 3676 }, { "epoch": 2.9870024370430546, "grad_norm": 0.31342780590057373, "learning_rate": 5.720193969716015e-10, "loss": 0.388, "step": 3677 }, { "epoch": 2.9878147847278633, "grad_norm": 0.32050690054893494, "learning_rate": 5.027525839662594e-10, "loss": 0.4142, "step": 3678 }, { "epoch": 2.9886271324126725, "grad_norm": 0.30470016598701477, "learning_rate": 4.3795430802073825e-10, "loss": 0.4211, "step": 3679 }, { "epoch": 2.9894394800974817, "grad_norm": 0.31059303879737854, "learning_rate": 3.7762462705093204e-10, "loss": 0.4153, "step": 3680 }, { "epoch": 2.990251827782291, "grad_norm": 0.2829483151435852, "learning_rate": 3.2176359498037325e-10, "loss": 0.3742, "step": 3681 }, { "epoch": 2.9910641754671, "grad_norm": 0.32292118668556213, "learning_rate": 2.703712617363463e-10, "loss": 0.3999, "step": 3682 }, { "epoch": 2.991876523151909, "grad_norm": 0.290768563747406, "learning_rate": 2.234476732537738e-10, "loss": 0.398, "step": 3683 }, { "epoch": 2.992688870836718, "grad_norm": 0.30378708243370056, "learning_rate": 1.8099287147355095e-10, "loss": 0.3738, "step": 3684 }, { "epoch": 2.9935012185215273, "grad_norm": 0.3001213073730469, "learning_rate": 1.4300689434032512e-10, "loss": 0.4342, "step": 3685 }, { "epoch": 2.994313566206336, "grad_norm": 0.29953646659851074, "learning_rate": 1.0948977580638176e-10, "loss": 0.3979, "step": 3686 }, { "epoch": 2.9951259138911457, "grad_norm": 0.3066002130508423, "learning_rate": 8.044154582942388e-11, "loss": 0.3876, "step": 3687 }, { "epoch": 2.9959382615759544, "grad_norm": 0.30894726514816284, "learning_rate": 5.586223037257199e-11, "loss": 0.3872, "step": 3688 }, { "epoch": 2.9967506092607636, "grad_norm": 0.3294631540775299, "learning_rate": 3.575185140436421e-11, "loss": 0.3805, "step": 3689 }, { "epoch": 2.997562956945573, "grad_norm": 0.27827492356300354, "learning_rate": 2.0110426900421532e-11, "loss": 0.4142, "step": 3690 }, { "epoch": 2.9983753046303816, "grad_norm": 0.3080309331417084, "learning_rate": 8.937970840117161e-12, "loss": 0.4092, "step": 3691 }, { "epoch": 2.999187652315191, "grad_norm": 0.3040265738964081, "learning_rate": 2.2344932093520955e-12, "loss": 0.4135, "step": 3692 }, { "epoch": 3.0, "grad_norm": 0.2983072102069855, "learning_rate": 0.0, "loss": 0.4276, "step": 3693 }, { "epoch": 3.0, "step": 3693, "total_flos": 4922876784672768.0, "train_loss": 0.44633466947895384, "train_runtime": 126808.2431, "train_samples_per_second": 2.796, "train_steps_per_second": 0.029 } ], "logging_steps": 1.0, "max_steps": 3693, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 4922876784672768.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }