{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.9983314794215796, "eval_steps": 500, "global_step": 3594, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0008342602892102336, "grad_norm": 5.708919048309326, "learning_rate": 2.777777777777778e-08, "loss": 0.8173, "step": 1 }, { "epoch": 0.0016685205784204673, "grad_norm": 6.039193630218506, "learning_rate": 5.555555555555556e-08, "loss": 0.8758, "step": 2 }, { "epoch": 0.002502780867630701, "grad_norm": 6.0849385261535645, "learning_rate": 8.333333333333334e-08, "loss": 0.8937, "step": 3 }, { "epoch": 0.0033370411568409346, "grad_norm": 5.605709552764893, "learning_rate": 1.1111111111111112e-07, "loss": 0.8387, "step": 4 }, { "epoch": 0.004171301446051168, "grad_norm": 5.639842987060547, "learning_rate": 1.3888888888888888e-07, "loss": 0.8456, "step": 5 }, { "epoch": 0.005005561735261402, "grad_norm": 6.165993690490723, "learning_rate": 1.6666666666666668e-07, "loss": 0.9266, "step": 6 }, { "epoch": 0.005839822024471635, "grad_norm": 5.909365653991699, "learning_rate": 1.9444444444444447e-07, "loss": 0.89, "step": 7 }, { "epoch": 0.006674082313681869, "grad_norm": 5.679555416107178, "learning_rate": 2.2222222222222224e-07, "loss": 0.8404, "step": 8 }, { "epoch": 0.007508342602892102, "grad_norm": 5.7326812744140625, "learning_rate": 2.5000000000000004e-07, "loss": 0.8516, "step": 9 }, { "epoch": 0.008342602892102336, "grad_norm": 5.580862522125244, "learning_rate": 2.7777777777777776e-07, "loss": 0.8485, "step": 10 }, { "epoch": 0.00917686318131257, "grad_norm": 5.981043338775635, "learning_rate": 3.055555555555556e-07, "loss": 0.8714, "step": 11 }, { "epoch": 0.010011123470522803, "grad_norm": 5.619521141052246, "learning_rate": 3.3333333333333335e-07, "loss": 0.8719, "step": 12 }, { "epoch": 0.010845383759733037, "grad_norm": 5.510502815246582, "learning_rate": 3.611111111111111e-07, "loss": 0.8194, "step": 13 }, { "epoch": 0.01167964404894327, "grad_norm": 5.426331043243408, "learning_rate": 3.8888888888888895e-07, "loss": 0.8529, "step": 14 }, { "epoch": 0.012513904338153505, "grad_norm": 5.47323751449585, "learning_rate": 4.1666666666666667e-07, "loss": 0.8614, "step": 15 }, { "epoch": 0.013348164627363738, "grad_norm": 5.528432846069336, "learning_rate": 4.444444444444445e-07, "loss": 0.8676, "step": 16 }, { "epoch": 0.01418242491657397, "grad_norm": 5.31148099899292, "learning_rate": 4.7222222222222226e-07, "loss": 0.8316, "step": 17 }, { "epoch": 0.015016685205784204, "grad_norm": 5.178371906280518, "learning_rate": 5.000000000000001e-07, "loss": 0.837, "step": 18 }, { "epoch": 0.01585094549499444, "grad_norm": 4.678620338439941, "learning_rate": 5.277777777777779e-07, "loss": 0.8418, "step": 19 }, { "epoch": 0.01668520578420467, "grad_norm": 4.412558078765869, "learning_rate": 5.555555555555555e-07, "loss": 0.8096, "step": 20 }, { "epoch": 0.017519466073414907, "grad_norm": 4.304449558258057, "learning_rate": 5.833333333333334e-07, "loss": 0.8314, "step": 21 }, { "epoch": 0.01835372636262514, "grad_norm": 4.251956462860107, "learning_rate": 6.111111111111112e-07, "loss": 0.8036, "step": 22 }, { "epoch": 0.01918798665183537, "grad_norm": 4.043249607086182, "learning_rate": 6.388888888888889e-07, "loss": 0.8134, "step": 23 }, { "epoch": 0.020022246941045607, "grad_norm": 4.08378791809082, "learning_rate": 6.666666666666667e-07, "loss": 0.8028, "step": 24 }, { "epoch": 0.02085650723025584, "grad_norm": 3.1457738876342773, "learning_rate": 6.944444444444446e-07, "loss": 0.8339, "step": 25 }, { "epoch": 0.021690767519466074, "grad_norm": 2.344637155532837, "learning_rate": 7.222222222222222e-07, "loss": 0.768, "step": 26 }, { "epoch": 0.022525027808676306, "grad_norm": 2.2388687133789062, "learning_rate": 7.5e-07, "loss": 0.7745, "step": 27 }, { "epoch": 0.02335928809788654, "grad_norm": 2.2946319580078125, "learning_rate": 7.777777777777779e-07, "loss": 0.7469, "step": 28 }, { "epoch": 0.024193548387096774, "grad_norm": 2.161558151245117, "learning_rate": 8.055555555555557e-07, "loss": 0.7638, "step": 29 }, { "epoch": 0.02502780867630701, "grad_norm": 2.0922598838806152, "learning_rate": 8.333333333333333e-07, "loss": 0.7563, "step": 30 }, { "epoch": 0.02586206896551724, "grad_norm": 1.8901244401931763, "learning_rate": 8.611111111111112e-07, "loss": 0.7304, "step": 31 }, { "epoch": 0.026696329254727477, "grad_norm": 1.9145872592926025, "learning_rate": 8.88888888888889e-07, "loss": 0.7329, "step": 32 }, { "epoch": 0.02753058954393771, "grad_norm": 1.8416396379470825, "learning_rate": 9.166666666666666e-07, "loss": 0.7407, "step": 33 }, { "epoch": 0.02836484983314794, "grad_norm": 1.3365721702575684, "learning_rate": 9.444444444444445e-07, "loss": 0.6998, "step": 34 }, { "epoch": 0.029199110122358176, "grad_norm": 1.7359769344329834, "learning_rate": 9.722222222222224e-07, "loss": 0.7421, "step": 35 }, { "epoch": 0.030033370411568408, "grad_norm": 1.926587700843811, "learning_rate": 1.0000000000000002e-06, "loss": 0.7257, "step": 36 }, { "epoch": 0.030867630700778644, "grad_norm": 2.1734840869903564, "learning_rate": 1.0277777777777777e-06, "loss": 0.7432, "step": 37 }, { "epoch": 0.03170189098998888, "grad_norm": 2.1981430053710938, "learning_rate": 1.0555555555555557e-06, "loss": 0.7145, "step": 38 }, { "epoch": 0.03253615127919911, "grad_norm": 2.1273505687713623, "learning_rate": 1.0833333333333335e-06, "loss": 0.728, "step": 39 }, { "epoch": 0.03337041156840934, "grad_norm": 1.9261773824691772, "learning_rate": 1.111111111111111e-06, "loss": 0.6998, "step": 40 }, { "epoch": 0.03420467185761958, "grad_norm": 1.832566738128662, "learning_rate": 1.138888888888889e-06, "loss": 0.6724, "step": 41 }, { "epoch": 0.035038932146829814, "grad_norm": 1.6951864957809448, "learning_rate": 1.1666666666666668e-06, "loss": 0.7258, "step": 42 }, { "epoch": 0.03587319243604004, "grad_norm": 1.4489998817443848, "learning_rate": 1.1944444444444446e-06, "loss": 0.6928, "step": 43 }, { "epoch": 0.03670745272525028, "grad_norm": 1.1862679719924927, "learning_rate": 1.2222222222222223e-06, "loss": 0.6992, "step": 44 }, { "epoch": 0.037541713014460514, "grad_norm": 1.0113189220428467, "learning_rate": 1.25e-06, "loss": 0.6876, "step": 45 }, { "epoch": 0.03837597330367074, "grad_norm": 0.9587087035179138, "learning_rate": 1.2777777777777779e-06, "loss": 0.7014, "step": 46 }, { "epoch": 0.03921023359288098, "grad_norm": 0.9483241438865662, "learning_rate": 1.3055555555555556e-06, "loss": 0.6863, "step": 47 }, { "epoch": 0.04004449388209121, "grad_norm": 1.1822460889816284, "learning_rate": 1.3333333333333334e-06, "loss": 0.6594, "step": 48 }, { "epoch": 0.04087875417130145, "grad_norm": 1.1768932342529297, "learning_rate": 1.3611111111111112e-06, "loss": 0.6704, "step": 49 }, { "epoch": 0.04171301446051168, "grad_norm": 1.056631088256836, "learning_rate": 1.3888888888888892e-06, "loss": 0.6609, "step": 50 }, { "epoch": 0.04254727474972191, "grad_norm": 0.9872738718986511, "learning_rate": 1.4166666666666667e-06, "loss": 0.6539, "step": 51 }, { "epoch": 0.04338153503893215, "grad_norm": 0.8230380415916443, "learning_rate": 1.4444444444444445e-06, "loss": 0.7089, "step": 52 }, { "epoch": 0.044215795328142384, "grad_norm": 0.7862669229507446, "learning_rate": 1.4722222222222225e-06, "loss": 0.6435, "step": 53 }, { "epoch": 0.04505005561735261, "grad_norm": 0.8005119562149048, "learning_rate": 1.5e-06, "loss": 0.6508, "step": 54 }, { "epoch": 0.04588431590656285, "grad_norm": 0.8250868320465088, "learning_rate": 1.527777777777778e-06, "loss": 0.6477, "step": 55 }, { "epoch": 0.04671857619577308, "grad_norm": 0.7650861740112305, "learning_rate": 1.5555555555555558e-06, "loss": 0.6407, "step": 56 }, { "epoch": 0.04755283648498331, "grad_norm": 0.6561533212661743, "learning_rate": 1.5833333333333333e-06, "loss": 0.6363, "step": 57 }, { "epoch": 0.04838709677419355, "grad_norm": 0.5850383639335632, "learning_rate": 1.6111111111111113e-06, "loss": 0.6588, "step": 58 }, { "epoch": 0.04922135706340378, "grad_norm": 0.6280196905136108, "learning_rate": 1.638888888888889e-06, "loss": 0.6538, "step": 59 }, { "epoch": 0.05005561735261402, "grad_norm": 0.6843660473823547, "learning_rate": 1.6666666666666667e-06, "loss": 0.6485, "step": 60 }, { "epoch": 0.05088987764182425, "grad_norm": 0.6910691857337952, "learning_rate": 1.6944444444444446e-06, "loss": 0.6006, "step": 61 }, { "epoch": 0.05172413793103448, "grad_norm": 0.7332186698913574, "learning_rate": 1.7222222222222224e-06, "loss": 0.6513, "step": 62 }, { "epoch": 0.05255839822024472, "grad_norm": 0.5347396731376648, "learning_rate": 1.75e-06, "loss": 0.6066, "step": 63 }, { "epoch": 0.05339265850945495, "grad_norm": 0.5130331516265869, "learning_rate": 1.777777777777778e-06, "loss": 0.6062, "step": 64 }, { "epoch": 0.05422691879866518, "grad_norm": 0.5051006078720093, "learning_rate": 1.8055555555555557e-06, "loss": 0.6032, "step": 65 }, { "epoch": 0.05506117908787542, "grad_norm": 0.5098767876625061, "learning_rate": 1.8333333333333333e-06, "loss": 0.5959, "step": 66 }, { "epoch": 0.05589543937708565, "grad_norm": 0.5882676839828491, "learning_rate": 1.8611111111111113e-06, "loss": 0.6063, "step": 67 }, { "epoch": 0.05672969966629588, "grad_norm": 0.5112848281860352, "learning_rate": 1.888888888888889e-06, "loss": 0.5873, "step": 68 }, { "epoch": 0.05756395995550612, "grad_norm": 0.49226927757263184, "learning_rate": 1.916666666666667e-06, "loss": 0.6153, "step": 69 }, { "epoch": 0.05839822024471635, "grad_norm": 0.4759417772293091, "learning_rate": 1.944444444444445e-06, "loss": 0.6152, "step": 70 }, { "epoch": 0.05923248053392659, "grad_norm": 0.47083353996276855, "learning_rate": 1.9722222222222224e-06, "loss": 0.5974, "step": 71 }, { "epoch": 0.060066740823136816, "grad_norm": 0.44576266407966614, "learning_rate": 2.0000000000000003e-06, "loss": 0.5811, "step": 72 }, { "epoch": 0.06090100111234705, "grad_norm": 0.4522881507873535, "learning_rate": 2.027777777777778e-06, "loss": 0.611, "step": 73 }, { "epoch": 0.06173526140155729, "grad_norm": 0.43690255284309387, "learning_rate": 2.0555555555555555e-06, "loss": 0.6169, "step": 74 }, { "epoch": 0.06256952169076752, "grad_norm": 0.4156505763530731, "learning_rate": 2.0833333333333334e-06, "loss": 0.5975, "step": 75 }, { "epoch": 0.06340378197997776, "grad_norm": 0.4417053461074829, "learning_rate": 2.1111111111111114e-06, "loss": 0.6064, "step": 76 }, { "epoch": 0.06423804226918799, "grad_norm": 0.4280395209789276, "learning_rate": 2.138888888888889e-06, "loss": 0.591, "step": 77 }, { "epoch": 0.06507230255839822, "grad_norm": 0.3998802900314331, "learning_rate": 2.166666666666667e-06, "loss": 0.5804, "step": 78 }, { "epoch": 0.06590656284760846, "grad_norm": 0.4114524722099304, "learning_rate": 2.1944444444444445e-06, "loss": 0.5923, "step": 79 }, { "epoch": 0.06674082313681869, "grad_norm": 0.45238175988197327, "learning_rate": 2.222222222222222e-06, "loss": 0.5562, "step": 80 }, { "epoch": 0.06757508342602891, "grad_norm": 0.4305950701236725, "learning_rate": 2.25e-06, "loss": 0.5991, "step": 81 }, { "epoch": 0.06840934371523916, "grad_norm": 0.43620091676712036, "learning_rate": 2.277777777777778e-06, "loss": 0.588, "step": 82 }, { "epoch": 0.06924360400444939, "grad_norm": 0.4351707398891449, "learning_rate": 2.305555555555556e-06, "loss": 0.6127, "step": 83 }, { "epoch": 0.07007786429365963, "grad_norm": 0.4268796443939209, "learning_rate": 2.3333333333333336e-06, "loss": 0.5786, "step": 84 }, { "epoch": 0.07091212458286986, "grad_norm": 0.4452349543571472, "learning_rate": 2.361111111111111e-06, "loss": 0.592, "step": 85 }, { "epoch": 0.07174638487208009, "grad_norm": 0.4476031959056854, "learning_rate": 2.388888888888889e-06, "loss": 0.5628, "step": 86 }, { "epoch": 0.07258064516129033, "grad_norm": 0.4106993079185486, "learning_rate": 2.4166666666666667e-06, "loss": 0.5819, "step": 87 }, { "epoch": 0.07341490545050056, "grad_norm": 0.46695026755332947, "learning_rate": 2.4444444444444447e-06, "loss": 0.5796, "step": 88 }, { "epoch": 0.07424916573971078, "grad_norm": 0.4742920696735382, "learning_rate": 2.4722222222222226e-06, "loss": 0.5864, "step": 89 }, { "epoch": 0.07508342602892103, "grad_norm": 0.4124802350997925, "learning_rate": 2.5e-06, "loss": 0.5835, "step": 90 }, { "epoch": 0.07591768631813126, "grad_norm": 0.43789926171302795, "learning_rate": 2.5277777777777778e-06, "loss": 0.5787, "step": 91 }, { "epoch": 0.07675194660734148, "grad_norm": 0.46302932500839233, "learning_rate": 2.5555555555555557e-06, "loss": 0.5877, "step": 92 }, { "epoch": 0.07758620689655173, "grad_norm": 0.3888735771179199, "learning_rate": 2.5833333333333337e-06, "loss": 0.5779, "step": 93 }, { "epoch": 0.07842046718576196, "grad_norm": 0.40758630633354187, "learning_rate": 2.6111111111111113e-06, "loss": 0.5907, "step": 94 }, { "epoch": 0.0792547274749722, "grad_norm": 0.414559543132782, "learning_rate": 2.6388888888888893e-06, "loss": 0.5917, "step": 95 }, { "epoch": 0.08008898776418243, "grad_norm": 0.42353811860084534, "learning_rate": 2.666666666666667e-06, "loss": 0.5674, "step": 96 }, { "epoch": 0.08092324805339265, "grad_norm": 0.39419618248939514, "learning_rate": 2.6944444444444444e-06, "loss": 0.5801, "step": 97 }, { "epoch": 0.0817575083426029, "grad_norm": 0.4704502820968628, "learning_rate": 2.7222222222222224e-06, "loss": 0.5573, "step": 98 }, { "epoch": 0.08259176863181313, "grad_norm": 0.4100845754146576, "learning_rate": 2.7500000000000004e-06, "loss": 0.5576, "step": 99 }, { "epoch": 0.08342602892102335, "grad_norm": 0.4097414016723633, "learning_rate": 2.7777777777777783e-06, "loss": 0.5878, "step": 100 }, { "epoch": 0.0842602892102336, "grad_norm": 0.39492595195770264, "learning_rate": 2.805555555555556e-06, "loss": 0.5762, "step": 101 }, { "epoch": 0.08509454949944383, "grad_norm": 0.43790578842163086, "learning_rate": 2.8333333333333335e-06, "loss": 0.5597, "step": 102 }, { "epoch": 0.08592880978865405, "grad_norm": 0.39539381861686707, "learning_rate": 2.861111111111111e-06, "loss": 0.5668, "step": 103 }, { "epoch": 0.0867630700778643, "grad_norm": 0.41095635294914246, "learning_rate": 2.888888888888889e-06, "loss": 0.5404, "step": 104 }, { "epoch": 0.08759733036707452, "grad_norm": 0.4371747076511383, "learning_rate": 2.916666666666667e-06, "loss": 0.598, "step": 105 }, { "epoch": 0.08843159065628477, "grad_norm": 0.43250513076782227, "learning_rate": 2.944444444444445e-06, "loss": 0.5665, "step": 106 }, { "epoch": 0.089265850945495, "grad_norm": 0.4232328534126282, "learning_rate": 2.9722222222222225e-06, "loss": 0.5872, "step": 107 }, { "epoch": 0.09010011123470522, "grad_norm": 0.4156021177768707, "learning_rate": 3e-06, "loss": 0.5655, "step": 108 }, { "epoch": 0.09093437152391547, "grad_norm": 0.3746601939201355, "learning_rate": 3.0277777777777776e-06, "loss": 0.5645, "step": 109 }, { "epoch": 0.0917686318131257, "grad_norm": 0.3864150643348694, "learning_rate": 3.055555555555556e-06, "loss": 0.5563, "step": 110 }, { "epoch": 0.09260289210233592, "grad_norm": 0.38822275400161743, "learning_rate": 3.0833333333333336e-06, "loss": 0.6026, "step": 111 }, { "epoch": 0.09343715239154617, "grad_norm": 0.39356729388237, "learning_rate": 3.1111111111111116e-06, "loss": 0.5842, "step": 112 }, { "epoch": 0.0942714126807564, "grad_norm": 0.4090965986251831, "learning_rate": 3.138888888888889e-06, "loss": 0.5492, "step": 113 }, { "epoch": 0.09510567296996662, "grad_norm": 0.38893017172813416, "learning_rate": 3.1666666666666667e-06, "loss": 0.5647, "step": 114 }, { "epoch": 0.09593993325917687, "grad_norm": 0.40844276547431946, "learning_rate": 3.1944444444444443e-06, "loss": 0.5663, "step": 115 }, { "epoch": 0.0967741935483871, "grad_norm": 0.38572585582733154, "learning_rate": 3.2222222222222227e-06, "loss": 0.5306, "step": 116 }, { "epoch": 0.09760845383759734, "grad_norm": 0.4360794126987457, "learning_rate": 3.2500000000000002e-06, "loss": 0.5509, "step": 117 }, { "epoch": 0.09844271412680757, "grad_norm": 0.44063445925712585, "learning_rate": 3.277777777777778e-06, "loss": 0.5862, "step": 118 }, { "epoch": 0.0992769744160178, "grad_norm": 0.3998779356479645, "learning_rate": 3.3055555555555558e-06, "loss": 0.5673, "step": 119 }, { "epoch": 0.10011123470522804, "grad_norm": 0.4427625238895416, "learning_rate": 3.3333333333333333e-06, "loss": 0.5464, "step": 120 }, { "epoch": 0.10094549499443826, "grad_norm": 0.4133763909339905, "learning_rate": 3.3611111111111117e-06, "loss": 0.5805, "step": 121 }, { "epoch": 0.1017797552836485, "grad_norm": 0.4352451264858246, "learning_rate": 3.3888888888888893e-06, "loss": 0.5632, "step": 122 }, { "epoch": 0.10261401557285874, "grad_norm": 0.4551921486854553, "learning_rate": 3.416666666666667e-06, "loss": 0.5709, "step": 123 }, { "epoch": 0.10344827586206896, "grad_norm": 0.47087979316711426, "learning_rate": 3.444444444444445e-06, "loss": 0.5632, "step": 124 }, { "epoch": 0.10428253615127919, "grad_norm": 0.41226309537887573, "learning_rate": 3.4722222222222224e-06, "loss": 0.507, "step": 125 }, { "epoch": 0.10511679644048944, "grad_norm": 0.4352550506591797, "learning_rate": 3.5e-06, "loss": 0.5207, "step": 126 }, { "epoch": 0.10595105672969966, "grad_norm": 0.445148766040802, "learning_rate": 3.5277777777777784e-06, "loss": 0.5853, "step": 127 }, { "epoch": 0.1067853170189099, "grad_norm": 0.478374719619751, "learning_rate": 3.555555555555556e-06, "loss": 0.5232, "step": 128 }, { "epoch": 0.10761957730812013, "grad_norm": 0.4467496871948242, "learning_rate": 3.5833333333333335e-06, "loss": 0.5714, "step": 129 }, { "epoch": 0.10845383759733036, "grad_norm": 0.4308788478374481, "learning_rate": 3.6111111111111115e-06, "loss": 0.5416, "step": 130 }, { "epoch": 0.1092880978865406, "grad_norm": 0.4457818865776062, "learning_rate": 3.638888888888889e-06, "loss": 0.5261, "step": 131 }, { "epoch": 0.11012235817575083, "grad_norm": 0.4213798940181732, "learning_rate": 3.6666666666666666e-06, "loss": 0.5607, "step": 132 }, { "epoch": 0.11095661846496106, "grad_norm": 0.3845173120498657, "learning_rate": 3.694444444444445e-06, "loss": 0.5557, "step": 133 }, { "epoch": 0.1117908787541713, "grad_norm": 0.41657063364982605, "learning_rate": 3.7222222222222225e-06, "loss": 0.5553, "step": 134 }, { "epoch": 0.11262513904338153, "grad_norm": 0.4071827828884125, "learning_rate": 3.7500000000000005e-06, "loss": 0.5639, "step": 135 }, { "epoch": 0.11345939933259176, "grad_norm": 0.4514647126197815, "learning_rate": 3.777777777777778e-06, "loss": 0.558, "step": 136 }, { "epoch": 0.114293659621802, "grad_norm": 0.42188382148742676, "learning_rate": 3.8055555555555556e-06, "loss": 0.5353, "step": 137 }, { "epoch": 0.11512791991101223, "grad_norm": 0.4296925365924835, "learning_rate": 3.833333333333334e-06, "loss": 0.5688, "step": 138 }, { "epoch": 0.11596218020022248, "grad_norm": 0.4024839699268341, "learning_rate": 3.861111111111112e-06, "loss": 0.5467, "step": 139 }, { "epoch": 0.1167964404894327, "grad_norm": 0.4384309947490692, "learning_rate": 3.88888888888889e-06, "loss": 0.5561, "step": 140 }, { "epoch": 0.11763070077864293, "grad_norm": 0.5160971879959106, "learning_rate": 3.916666666666667e-06, "loss": 0.5385, "step": 141 }, { "epoch": 0.11846496106785318, "grad_norm": 0.4151046574115753, "learning_rate": 3.944444444444445e-06, "loss": 0.5505, "step": 142 }, { "epoch": 0.1192992213570634, "grad_norm": 0.39090967178344727, "learning_rate": 3.972222222222223e-06, "loss": 0.5128, "step": 143 }, { "epoch": 0.12013348164627363, "grad_norm": 0.4109446704387665, "learning_rate": 4.000000000000001e-06, "loss": 0.547, "step": 144 }, { "epoch": 0.12096774193548387, "grad_norm": 0.4755812883377075, "learning_rate": 4.027777777777779e-06, "loss": 0.546, "step": 145 }, { "epoch": 0.1218020022246941, "grad_norm": 0.4066700041294098, "learning_rate": 4.055555555555556e-06, "loss": 0.5333, "step": 146 }, { "epoch": 0.12263626251390433, "grad_norm": 0.4143904447555542, "learning_rate": 4.083333333333334e-06, "loss": 0.5359, "step": 147 }, { "epoch": 0.12347052280311457, "grad_norm": 0.40207016468048096, "learning_rate": 4.111111111111111e-06, "loss": 0.5305, "step": 148 }, { "epoch": 0.1243047830923248, "grad_norm": 0.5058565139770508, "learning_rate": 4.138888888888889e-06, "loss": 0.5754, "step": 149 }, { "epoch": 0.12513904338153503, "grad_norm": 0.41386109590530396, "learning_rate": 4.166666666666667e-06, "loss": 0.5274, "step": 150 }, { "epoch": 0.12597330367074527, "grad_norm": 0.4238750636577606, "learning_rate": 4.194444444444445e-06, "loss": 0.5162, "step": 151 }, { "epoch": 0.12680756395995552, "grad_norm": 0.4168976843357086, "learning_rate": 4.222222222222223e-06, "loss": 0.5799, "step": 152 }, { "epoch": 0.12764182424916573, "grad_norm": 0.43462175130844116, "learning_rate": 4.25e-06, "loss": 0.5328, "step": 153 }, { "epoch": 0.12847608453837597, "grad_norm": 0.45085975527763367, "learning_rate": 4.277777777777778e-06, "loss": 0.5413, "step": 154 }, { "epoch": 0.12931034482758622, "grad_norm": 0.4278835356235504, "learning_rate": 4.305555555555556e-06, "loss": 0.5145, "step": 155 }, { "epoch": 0.13014460511679643, "grad_norm": 0.43235158920288086, "learning_rate": 4.333333333333334e-06, "loss": 0.5679, "step": 156 }, { "epoch": 0.13097886540600667, "grad_norm": 0.39167067408561707, "learning_rate": 4.361111111111112e-06, "loss": 0.5167, "step": 157 }, { "epoch": 0.13181312569521692, "grad_norm": 0.4315093755722046, "learning_rate": 4.388888888888889e-06, "loss": 0.5242, "step": 158 }, { "epoch": 0.13264738598442713, "grad_norm": 0.46740204095840454, "learning_rate": 4.416666666666667e-06, "loss": 0.5659, "step": 159 }, { "epoch": 0.13348164627363737, "grad_norm": 0.4418814182281494, "learning_rate": 4.444444444444444e-06, "loss": 0.5396, "step": 160 }, { "epoch": 0.13431590656284761, "grad_norm": 0.4133063554763794, "learning_rate": 4.472222222222223e-06, "loss": 0.5503, "step": 161 }, { "epoch": 0.13515016685205783, "grad_norm": 0.45282894372940063, "learning_rate": 4.5e-06, "loss": 0.5347, "step": 162 }, { "epoch": 0.13598442714126807, "grad_norm": 0.41486236453056335, "learning_rate": 4.527777777777778e-06, "loss": 0.5604, "step": 163 }, { "epoch": 0.13681868743047831, "grad_norm": 0.41874799132347107, "learning_rate": 4.555555555555556e-06, "loss": 0.5365, "step": 164 }, { "epoch": 0.13765294771968856, "grad_norm": 0.4446343183517456, "learning_rate": 4.583333333333333e-06, "loss": 0.5494, "step": 165 }, { "epoch": 0.13848720800889877, "grad_norm": 0.441735178232193, "learning_rate": 4.611111111111112e-06, "loss": 0.5344, "step": 166 }, { "epoch": 0.139321468298109, "grad_norm": 0.4326004385948181, "learning_rate": 4.638888888888889e-06, "loss": 0.5154, "step": 167 }, { "epoch": 0.14015572858731926, "grad_norm": 0.50369793176651, "learning_rate": 4.666666666666667e-06, "loss": 0.5184, "step": 168 }, { "epoch": 0.14098998887652947, "grad_norm": 0.43184980750083923, "learning_rate": 4.694444444444445e-06, "loss": 0.5142, "step": 169 }, { "epoch": 0.1418242491657397, "grad_norm": 0.4378785192966461, "learning_rate": 4.722222222222222e-06, "loss": 0.5524, "step": 170 }, { "epoch": 0.14265850945494996, "grad_norm": 0.4451082646846771, "learning_rate": 4.75e-06, "loss": 0.5231, "step": 171 }, { "epoch": 0.14349276974416017, "grad_norm": 0.5083770155906677, "learning_rate": 4.777777777777778e-06, "loss": 0.59, "step": 172 }, { "epoch": 0.1443270300333704, "grad_norm": 0.4598393738269806, "learning_rate": 4.805555555555556e-06, "loss": 0.525, "step": 173 }, { "epoch": 0.14516129032258066, "grad_norm": 0.39461109042167664, "learning_rate": 4.833333333333333e-06, "loss": 0.534, "step": 174 }, { "epoch": 0.14599555061179087, "grad_norm": 0.4876367449760437, "learning_rate": 4.861111111111111e-06, "loss": 0.5495, "step": 175 }, { "epoch": 0.1468298109010011, "grad_norm": 0.4737469553947449, "learning_rate": 4.888888888888889e-06, "loss": 0.5457, "step": 176 }, { "epoch": 0.14766407119021135, "grad_norm": 0.46269896626472473, "learning_rate": 4.9166666666666665e-06, "loss": 0.5709, "step": 177 }, { "epoch": 0.14849833147942157, "grad_norm": 0.4733167290687561, "learning_rate": 4.944444444444445e-06, "loss": 0.5357, "step": 178 }, { "epoch": 0.1493325917686318, "grad_norm": 0.4146479666233063, "learning_rate": 4.9722222222222224e-06, "loss": 0.5251, "step": 179 }, { "epoch": 0.15016685205784205, "grad_norm": 0.46735939383506775, "learning_rate": 5e-06, "loss": 0.5164, "step": 180 }, { "epoch": 0.15100111234705227, "grad_norm": 0.452664315700531, "learning_rate": 5.027777777777778e-06, "loss": 0.539, "step": 181 }, { "epoch": 0.1518353726362625, "grad_norm": 0.423000693321228, "learning_rate": 5.0555555555555555e-06, "loss": 0.5292, "step": 182 }, { "epoch": 0.15266963292547275, "grad_norm": 0.4257478713989258, "learning_rate": 5.0833333333333335e-06, "loss": 0.5276, "step": 183 }, { "epoch": 0.15350389321468297, "grad_norm": 0.44887983798980713, "learning_rate": 5.1111111111111115e-06, "loss": 0.5087, "step": 184 }, { "epoch": 0.1543381535038932, "grad_norm": 0.46790847182273865, "learning_rate": 5.138888888888889e-06, "loss": 0.5367, "step": 185 }, { "epoch": 0.15517241379310345, "grad_norm": 0.4800509810447693, "learning_rate": 5.1666666666666675e-06, "loss": 0.5465, "step": 186 }, { "epoch": 0.1560066740823137, "grad_norm": 0.4490276575088501, "learning_rate": 5.1944444444444454e-06, "loss": 0.5301, "step": 187 }, { "epoch": 0.1568409343715239, "grad_norm": 0.4528900980949402, "learning_rate": 5.2222222222222226e-06, "loss": 0.5614, "step": 188 }, { "epoch": 0.15767519466073415, "grad_norm": 0.44372421503067017, "learning_rate": 5.2500000000000006e-06, "loss": 0.5258, "step": 189 }, { "epoch": 0.1585094549499444, "grad_norm": 0.4298260807991028, "learning_rate": 5.2777777777777785e-06, "loss": 0.5464, "step": 190 }, { "epoch": 0.1593437152391546, "grad_norm": 0.4085121750831604, "learning_rate": 5.305555555555556e-06, "loss": 0.5292, "step": 191 }, { "epoch": 0.16017797552836485, "grad_norm": 0.4379042387008667, "learning_rate": 5.333333333333334e-06, "loss": 0.5305, "step": 192 }, { "epoch": 0.1610122358175751, "grad_norm": 0.4323064982891083, "learning_rate": 5.361111111111112e-06, "loss": 0.5452, "step": 193 }, { "epoch": 0.1618464961067853, "grad_norm": 0.4173462986946106, "learning_rate": 5.388888888888889e-06, "loss": 0.5497, "step": 194 }, { "epoch": 0.16268075639599555, "grad_norm": 0.4222031235694885, "learning_rate": 5.416666666666667e-06, "loss": 0.5417, "step": 195 }, { "epoch": 0.1635150166852058, "grad_norm": 0.424387663602829, "learning_rate": 5.444444444444445e-06, "loss": 0.5016, "step": 196 }, { "epoch": 0.164349276974416, "grad_norm": 0.39744752645492554, "learning_rate": 5.4722222222222236e-06, "loss": 0.5192, "step": 197 }, { "epoch": 0.16518353726362625, "grad_norm": 0.4662140905857086, "learning_rate": 5.500000000000001e-06, "loss": 0.5316, "step": 198 }, { "epoch": 0.1660177975528365, "grad_norm": 0.42233771085739136, "learning_rate": 5.527777777777779e-06, "loss": 0.5204, "step": 199 }, { "epoch": 0.1668520578420467, "grad_norm": 0.39574185013771057, "learning_rate": 5.555555555555557e-06, "loss": 0.5349, "step": 200 }, { "epoch": 0.16768631813125695, "grad_norm": 0.4619927704334259, "learning_rate": 5.583333333333334e-06, "loss": 0.5376, "step": 201 }, { "epoch": 0.1685205784204672, "grad_norm": 0.39550814032554626, "learning_rate": 5.611111111111112e-06, "loss": 0.504, "step": 202 }, { "epoch": 0.1693548387096774, "grad_norm": 0.4330417513847351, "learning_rate": 5.638888888888889e-06, "loss": 0.5372, "step": 203 }, { "epoch": 0.17018909899888765, "grad_norm": 0.44979408383369446, "learning_rate": 5.666666666666667e-06, "loss": 0.5204, "step": 204 }, { "epoch": 0.1710233592880979, "grad_norm": 0.3971669673919678, "learning_rate": 5.694444444444445e-06, "loss": 0.4892, "step": 205 }, { "epoch": 0.1718576195773081, "grad_norm": 0.4077068567276001, "learning_rate": 5.722222222222222e-06, "loss": 0.5048, "step": 206 }, { "epoch": 0.17269187986651835, "grad_norm": 0.49847298860549927, "learning_rate": 5.75e-06, "loss": 0.5549, "step": 207 }, { "epoch": 0.1735261401557286, "grad_norm": 0.4292186498641968, "learning_rate": 5.777777777777778e-06, "loss": 0.5085, "step": 208 }, { "epoch": 0.17436040044493883, "grad_norm": 0.4549933075904846, "learning_rate": 5.805555555555557e-06, "loss": 0.5218, "step": 209 }, { "epoch": 0.17519466073414905, "grad_norm": 0.4953353703022003, "learning_rate": 5.833333333333334e-06, "loss": 0.5324, "step": 210 }, { "epoch": 0.1760289210233593, "grad_norm": 0.5215482115745544, "learning_rate": 5.861111111111112e-06, "loss": 0.5458, "step": 211 }, { "epoch": 0.17686318131256953, "grad_norm": 0.4282817244529724, "learning_rate": 5.88888888888889e-06, "loss": 0.5061, "step": 212 }, { "epoch": 0.17769744160177975, "grad_norm": 0.47634193301200867, "learning_rate": 5.916666666666667e-06, "loss": 0.5073, "step": 213 }, { "epoch": 0.17853170189099, "grad_norm": 0.47005677223205566, "learning_rate": 5.944444444444445e-06, "loss": 0.5479, "step": 214 }, { "epoch": 0.17936596218020023, "grad_norm": 0.41902974247932434, "learning_rate": 5.972222222222222e-06, "loss": 0.5189, "step": 215 }, { "epoch": 0.18020022246941045, "grad_norm": 0.49717599153518677, "learning_rate": 6e-06, "loss": 0.504, "step": 216 }, { "epoch": 0.1810344827586207, "grad_norm": 0.4664742946624756, "learning_rate": 6.027777777777778e-06, "loss": 0.5267, "step": 217 }, { "epoch": 0.18186874304783093, "grad_norm": 0.4234054982662201, "learning_rate": 6.055555555555555e-06, "loss": 0.5176, "step": 218 }, { "epoch": 0.18270300333704115, "grad_norm": 0.4274773895740509, "learning_rate": 6.083333333333333e-06, "loss": 0.546, "step": 219 }, { "epoch": 0.1835372636262514, "grad_norm": 0.40953537821769714, "learning_rate": 6.111111111111112e-06, "loss": 0.5206, "step": 220 }, { "epoch": 0.18437152391546163, "grad_norm": 0.4548538327217102, "learning_rate": 6.13888888888889e-06, "loss": 0.5202, "step": 221 }, { "epoch": 0.18520578420467185, "grad_norm": 0.4012257754802704, "learning_rate": 6.166666666666667e-06, "loss": 0.5645, "step": 222 }, { "epoch": 0.1860400444938821, "grad_norm": 0.48127222061157227, "learning_rate": 6.194444444444445e-06, "loss": 0.5063, "step": 223 }, { "epoch": 0.18687430478309233, "grad_norm": 0.3836106061935425, "learning_rate": 6.222222222222223e-06, "loss": 0.5113, "step": 224 }, { "epoch": 0.18770856507230255, "grad_norm": 0.4248840808868408, "learning_rate": 6.25e-06, "loss": 0.4992, "step": 225 }, { "epoch": 0.1885428253615128, "grad_norm": 0.4566095173358917, "learning_rate": 6.277777777777778e-06, "loss": 0.5286, "step": 226 }, { "epoch": 0.18937708565072303, "grad_norm": 0.47545602917671204, "learning_rate": 6.305555555555556e-06, "loss": 0.5271, "step": 227 }, { "epoch": 0.19021134593993325, "grad_norm": 0.4511077105998993, "learning_rate": 6.333333333333333e-06, "loss": 0.512, "step": 228 }, { "epoch": 0.1910456062291435, "grad_norm": 0.471103698015213, "learning_rate": 6.361111111111111e-06, "loss": 0.5276, "step": 229 }, { "epoch": 0.19187986651835373, "grad_norm": 0.5352763533592224, "learning_rate": 6.3888888888888885e-06, "loss": 0.5454, "step": 230 }, { "epoch": 0.19271412680756397, "grad_norm": 0.42069944739341736, "learning_rate": 6.416666666666667e-06, "loss": 0.5142, "step": 231 }, { "epoch": 0.1935483870967742, "grad_norm": 0.46758028864860535, "learning_rate": 6.444444444444445e-06, "loss": 0.5186, "step": 232 }, { "epoch": 0.19438264738598443, "grad_norm": 0.41542285680770874, "learning_rate": 6.472222222222223e-06, "loss": 0.5243, "step": 233 }, { "epoch": 0.19521690767519467, "grad_norm": 0.4628616273403168, "learning_rate": 6.5000000000000004e-06, "loss": 0.4936, "step": 234 }, { "epoch": 0.1960511679644049, "grad_norm": 0.45538580417633057, "learning_rate": 6.5277777777777784e-06, "loss": 0.5164, "step": 235 }, { "epoch": 0.19688542825361513, "grad_norm": 0.4266480803489685, "learning_rate": 6.555555555555556e-06, "loss": 0.5431, "step": 236 }, { "epoch": 0.19771968854282537, "grad_norm": 0.5166548490524292, "learning_rate": 6.5833333333333335e-06, "loss": 0.5358, "step": 237 }, { "epoch": 0.1985539488320356, "grad_norm": 0.4153410494327545, "learning_rate": 6.6111111111111115e-06, "loss": 0.5106, "step": 238 }, { "epoch": 0.19938820912124583, "grad_norm": 0.43755635619163513, "learning_rate": 6.6388888888888895e-06, "loss": 0.5217, "step": 239 }, { "epoch": 0.20022246941045607, "grad_norm": 0.536597728729248, "learning_rate": 6.666666666666667e-06, "loss": 0.5537, "step": 240 }, { "epoch": 0.2010567296996663, "grad_norm": 0.39942190051078796, "learning_rate": 6.694444444444445e-06, "loss": 0.521, "step": 241 }, { "epoch": 0.20189098998887653, "grad_norm": 0.46746277809143066, "learning_rate": 6.7222222222222235e-06, "loss": 0.527, "step": 242 }, { "epoch": 0.20272525027808677, "grad_norm": 0.4072192311286926, "learning_rate": 6.750000000000001e-06, "loss": 0.4949, "step": 243 }, { "epoch": 0.203559510567297, "grad_norm": 0.4744003415107727, "learning_rate": 6.777777777777779e-06, "loss": 0.5481, "step": 244 }, { "epoch": 0.20439377085650723, "grad_norm": 0.4428187906742096, "learning_rate": 6.8055555555555566e-06, "loss": 0.5067, "step": 245 }, { "epoch": 0.20522803114571747, "grad_norm": 0.4168057441711426, "learning_rate": 6.833333333333334e-06, "loss": 0.5326, "step": 246 }, { "epoch": 0.20606229143492769, "grad_norm": 0.4309629499912262, "learning_rate": 6.861111111111112e-06, "loss": 0.5113, "step": 247 }, { "epoch": 0.20689655172413793, "grad_norm": 0.5167315006256104, "learning_rate": 6.88888888888889e-06, "loss": 0.5326, "step": 248 }, { "epoch": 0.20773081201334817, "grad_norm": 0.4145900309085846, "learning_rate": 6.916666666666667e-06, "loss": 0.5131, "step": 249 }, { "epoch": 0.20856507230255839, "grad_norm": 0.5367866158485413, "learning_rate": 6.944444444444445e-06, "loss": 0.5474, "step": 250 }, { "epoch": 0.20939933259176863, "grad_norm": 0.46214938163757324, "learning_rate": 6.972222222222223e-06, "loss": 0.5536, "step": 251 }, { "epoch": 0.21023359288097887, "grad_norm": 0.40873825550079346, "learning_rate": 7e-06, "loss": 0.4923, "step": 252 }, { "epoch": 0.2110678531701891, "grad_norm": 0.4708506464958191, "learning_rate": 7.027777777777778e-06, "loss": 0.5083, "step": 253 }, { "epoch": 0.21190211345939933, "grad_norm": 0.4435998797416687, "learning_rate": 7.055555555555557e-06, "loss": 0.5068, "step": 254 }, { "epoch": 0.21273637374860957, "grad_norm": 0.4146523177623749, "learning_rate": 7.083333333333335e-06, "loss": 0.4828, "step": 255 }, { "epoch": 0.2135706340378198, "grad_norm": 0.5613356232643127, "learning_rate": 7.111111111111112e-06, "loss": 0.5204, "step": 256 }, { "epoch": 0.21440489432703003, "grad_norm": 0.47948330640792847, "learning_rate": 7.13888888888889e-06, "loss": 0.5299, "step": 257 }, { "epoch": 0.21523915461624027, "grad_norm": 0.4497644901275635, "learning_rate": 7.166666666666667e-06, "loss": 0.4939, "step": 258 }, { "epoch": 0.2160734149054505, "grad_norm": 0.520905077457428, "learning_rate": 7.194444444444445e-06, "loss": 0.5206, "step": 259 }, { "epoch": 0.21690767519466073, "grad_norm": 1.137113094329834, "learning_rate": 7.222222222222223e-06, "loss": 0.5313, "step": 260 }, { "epoch": 0.21774193548387097, "grad_norm": 0.5827332735061646, "learning_rate": 7.25e-06, "loss": 0.5523, "step": 261 }, { "epoch": 0.2185761957730812, "grad_norm": 0.5084248781204224, "learning_rate": 7.277777777777778e-06, "loss": 0.5309, "step": 262 }, { "epoch": 0.21941045606229143, "grad_norm": 0.42944765090942383, "learning_rate": 7.305555555555556e-06, "loss": 0.4928, "step": 263 }, { "epoch": 0.22024471635150167, "grad_norm": 0.5346848368644714, "learning_rate": 7.333333333333333e-06, "loss": 0.5227, "step": 264 }, { "epoch": 0.2210789766407119, "grad_norm": 0.41354894638061523, "learning_rate": 7.361111111111112e-06, "loss": 0.4994, "step": 265 }, { "epoch": 0.22191323692992213, "grad_norm": 0.41568654775619507, "learning_rate": 7.38888888888889e-06, "loss": 0.5281, "step": 266 }, { "epoch": 0.22274749721913237, "grad_norm": 0.43971559405326843, "learning_rate": 7.416666666666668e-06, "loss": 0.5286, "step": 267 }, { "epoch": 0.2235817575083426, "grad_norm": 0.47444212436676025, "learning_rate": 7.444444444444445e-06, "loss": 0.5269, "step": 268 }, { "epoch": 0.22441601779755282, "grad_norm": 0.442877858877182, "learning_rate": 7.472222222222223e-06, "loss": 0.5278, "step": 269 }, { "epoch": 0.22525027808676307, "grad_norm": 0.4597945213317871, "learning_rate": 7.500000000000001e-06, "loss": 0.4792, "step": 270 }, { "epoch": 0.2260845383759733, "grad_norm": 0.5012669563293457, "learning_rate": 7.527777777777778e-06, "loss": 0.5136, "step": 271 }, { "epoch": 0.22691879866518352, "grad_norm": 0.539006769657135, "learning_rate": 7.555555555555556e-06, "loss": 0.5285, "step": 272 }, { "epoch": 0.22775305895439377, "grad_norm": 0.48481157422065735, "learning_rate": 7.583333333333333e-06, "loss": 0.5124, "step": 273 }, { "epoch": 0.228587319243604, "grad_norm": 0.5181871056556702, "learning_rate": 7.611111111111111e-06, "loss": 0.513, "step": 274 }, { "epoch": 0.22942157953281425, "grad_norm": 0.552863359451294, "learning_rate": 7.638888888888888e-06, "loss": 0.5162, "step": 275 }, { "epoch": 0.23025583982202447, "grad_norm": 0.510677695274353, "learning_rate": 7.666666666666667e-06, "loss": 0.4953, "step": 276 }, { "epoch": 0.2310901001112347, "grad_norm": 0.5336751341819763, "learning_rate": 7.694444444444446e-06, "loss": 0.5418, "step": 277 }, { "epoch": 0.23192436040044495, "grad_norm": 0.5269660353660583, "learning_rate": 7.722222222222223e-06, "loss": 0.4944, "step": 278 }, { "epoch": 0.23275862068965517, "grad_norm": 0.5353904962539673, "learning_rate": 7.75e-06, "loss": 0.5214, "step": 279 }, { "epoch": 0.2335928809788654, "grad_norm": 0.4470595717430115, "learning_rate": 7.77777777777778e-06, "loss": 0.52, "step": 280 }, { "epoch": 0.23442714126807565, "grad_norm": 0.5386172533035278, "learning_rate": 7.805555555555556e-06, "loss": 0.493, "step": 281 }, { "epoch": 0.23526140155728587, "grad_norm": 0.4558522701263428, "learning_rate": 7.833333333333333e-06, "loss": 0.4978, "step": 282 }, { "epoch": 0.2360956618464961, "grad_norm": 0.5469487309455872, "learning_rate": 7.861111111111112e-06, "loss": 0.5251, "step": 283 }, { "epoch": 0.23692992213570635, "grad_norm": 0.422236829996109, "learning_rate": 7.88888888888889e-06, "loss": 0.4987, "step": 284 }, { "epoch": 0.23776418242491656, "grad_norm": 0.4462045729160309, "learning_rate": 7.916666666666667e-06, "loss": 0.5557, "step": 285 }, { "epoch": 0.2385984427141268, "grad_norm": 0.47547686100006104, "learning_rate": 7.944444444444445e-06, "loss": 0.5306, "step": 286 }, { "epoch": 0.23943270300333705, "grad_norm": 0.4801589250564575, "learning_rate": 7.972222222222224e-06, "loss": 0.4946, "step": 287 }, { "epoch": 0.24026696329254726, "grad_norm": 0.5033978223800659, "learning_rate": 8.000000000000001e-06, "loss": 0.505, "step": 288 }, { "epoch": 0.2411012235817575, "grad_norm": 0.48436596989631653, "learning_rate": 8.027777777777778e-06, "loss": 0.5325, "step": 289 }, { "epoch": 0.24193548387096775, "grad_norm": 0.5061163306236267, "learning_rate": 8.055555555555557e-06, "loss": 0.5067, "step": 290 }, { "epoch": 0.24276974416017796, "grad_norm": 0.4975638687610626, "learning_rate": 8.083333333333334e-06, "loss": 0.5198, "step": 291 }, { "epoch": 0.2436040044493882, "grad_norm": 0.45379534363746643, "learning_rate": 8.111111111111112e-06, "loss": 0.4993, "step": 292 }, { "epoch": 0.24443826473859845, "grad_norm": 0.543631374835968, "learning_rate": 8.138888888888889e-06, "loss": 0.5302, "step": 293 }, { "epoch": 0.24527252502780866, "grad_norm": 0.48415473103523254, "learning_rate": 8.166666666666668e-06, "loss": 0.5191, "step": 294 }, { "epoch": 0.2461067853170189, "grad_norm": 0.5045926570892334, "learning_rate": 8.194444444444445e-06, "loss": 0.5568, "step": 295 }, { "epoch": 0.24694104560622915, "grad_norm": 0.5577294230461121, "learning_rate": 8.222222222222222e-06, "loss": 0.5254, "step": 296 }, { "epoch": 0.2477753058954394, "grad_norm": 0.4676254987716675, "learning_rate": 8.25e-06, "loss": 0.514, "step": 297 }, { "epoch": 0.2486095661846496, "grad_norm": 0.48374131321907043, "learning_rate": 8.277777777777778e-06, "loss": 0.5002, "step": 298 }, { "epoch": 0.24944382647385985, "grad_norm": 0.5109012126922607, "learning_rate": 8.305555555555557e-06, "loss": 0.5017, "step": 299 }, { "epoch": 0.25027808676307006, "grad_norm": 0.47835978865623474, "learning_rate": 8.333333333333334e-06, "loss": 0.5301, "step": 300 }, { "epoch": 0.2511123470522803, "grad_norm": 0.4359108507633209, "learning_rate": 8.361111111111113e-06, "loss": 0.4859, "step": 301 }, { "epoch": 0.25194660734149055, "grad_norm": 0.45672059059143066, "learning_rate": 8.38888888888889e-06, "loss": 0.4904, "step": 302 }, { "epoch": 0.2527808676307008, "grad_norm": 0.46841976046562195, "learning_rate": 8.416666666666667e-06, "loss": 0.5073, "step": 303 }, { "epoch": 0.25361512791991103, "grad_norm": 0.42691048979759216, "learning_rate": 8.444444444444446e-06, "loss": 0.5158, "step": 304 }, { "epoch": 0.2544493882091212, "grad_norm": 0.49940377473831177, "learning_rate": 8.472222222222223e-06, "loss": 0.5123, "step": 305 }, { "epoch": 0.25528364849833146, "grad_norm": 0.5150557160377502, "learning_rate": 8.5e-06, "loss": 0.5061, "step": 306 }, { "epoch": 0.2561179087875417, "grad_norm": 0.4840739071369171, "learning_rate": 8.527777777777779e-06, "loss": 0.472, "step": 307 }, { "epoch": 0.25695216907675195, "grad_norm": 0.6029018759727478, "learning_rate": 8.555555555555556e-06, "loss": 0.512, "step": 308 }, { "epoch": 0.2577864293659622, "grad_norm": 0.44637009501457214, "learning_rate": 8.583333333333333e-06, "loss": 0.4886, "step": 309 }, { "epoch": 0.25862068965517243, "grad_norm": 0.5015267729759216, "learning_rate": 8.611111111111112e-06, "loss": 0.4738, "step": 310 }, { "epoch": 0.2594549499443827, "grad_norm": 0.48496803641319275, "learning_rate": 8.63888888888889e-06, "loss": 0.5085, "step": 311 }, { "epoch": 0.26028921023359286, "grad_norm": 0.48432299494743347, "learning_rate": 8.666666666666668e-06, "loss": 0.5098, "step": 312 }, { "epoch": 0.2611234705228031, "grad_norm": 0.4739378094673157, "learning_rate": 8.694444444444445e-06, "loss": 0.5096, "step": 313 }, { "epoch": 0.26195773081201335, "grad_norm": 0.4008074104785919, "learning_rate": 8.722222222222224e-06, "loss": 0.4934, "step": 314 }, { "epoch": 0.2627919911012236, "grad_norm": 0.4736694097518921, "learning_rate": 8.750000000000001e-06, "loss": 0.5102, "step": 315 }, { "epoch": 0.26362625139043383, "grad_norm": 0.4572477340698242, "learning_rate": 8.777777777777778e-06, "loss": 0.5422, "step": 316 }, { "epoch": 0.2644605116796441, "grad_norm": 0.404573917388916, "learning_rate": 8.805555555555557e-06, "loss": 0.4942, "step": 317 }, { "epoch": 0.26529477196885426, "grad_norm": 0.40781185030937195, "learning_rate": 8.833333333333334e-06, "loss": 0.4919, "step": 318 }, { "epoch": 0.2661290322580645, "grad_norm": 0.4529590606689453, "learning_rate": 8.861111111111111e-06, "loss": 0.5099, "step": 319 }, { "epoch": 0.26696329254727474, "grad_norm": 0.4221731424331665, "learning_rate": 8.888888888888888e-06, "loss": 0.5087, "step": 320 }, { "epoch": 0.267797552836485, "grad_norm": 0.411343514919281, "learning_rate": 8.916666666666667e-06, "loss": 0.4796, "step": 321 }, { "epoch": 0.26863181312569523, "grad_norm": 0.4464069902896881, "learning_rate": 8.944444444444446e-06, "loss": 0.5154, "step": 322 }, { "epoch": 0.26946607341490547, "grad_norm": 0.48135966062545776, "learning_rate": 8.972222222222223e-06, "loss": 0.512, "step": 323 }, { "epoch": 0.27030033370411566, "grad_norm": 0.48588791489601135, "learning_rate": 9e-06, "loss": 0.5124, "step": 324 }, { "epoch": 0.2711345939933259, "grad_norm": 0.4202023446559906, "learning_rate": 9.027777777777779e-06, "loss": 0.4919, "step": 325 }, { "epoch": 0.27196885428253614, "grad_norm": 0.4464338719844818, "learning_rate": 9.055555555555556e-06, "loss": 0.4956, "step": 326 }, { "epoch": 0.2728031145717464, "grad_norm": 0.4043933153152466, "learning_rate": 9.083333333333333e-06, "loss": 0.4658, "step": 327 }, { "epoch": 0.27363737486095663, "grad_norm": 0.48797979950904846, "learning_rate": 9.111111111111112e-06, "loss": 0.5179, "step": 328 }, { "epoch": 0.27447163515016687, "grad_norm": 0.43707919120788574, "learning_rate": 9.13888888888889e-06, "loss": 0.4841, "step": 329 }, { "epoch": 0.2753058954393771, "grad_norm": 0.479592502117157, "learning_rate": 9.166666666666666e-06, "loss": 0.5118, "step": 330 }, { "epoch": 0.2761401557285873, "grad_norm": 0.4346978962421417, "learning_rate": 9.194444444444445e-06, "loss": 0.5034, "step": 331 }, { "epoch": 0.27697441601779754, "grad_norm": 0.5436764359474182, "learning_rate": 9.222222222222224e-06, "loss": 0.5107, "step": 332 }, { "epoch": 0.2778086763070078, "grad_norm": 0.4696187674999237, "learning_rate": 9.250000000000001e-06, "loss": 0.4925, "step": 333 }, { "epoch": 0.278642936596218, "grad_norm": 0.5276966691017151, "learning_rate": 9.277777777777778e-06, "loss": 0.4961, "step": 334 }, { "epoch": 0.27947719688542827, "grad_norm": 0.5292407274246216, "learning_rate": 9.305555555555557e-06, "loss": 0.4976, "step": 335 }, { "epoch": 0.2803114571746385, "grad_norm": 0.4913964569568634, "learning_rate": 9.333333333333334e-06, "loss": 0.4844, "step": 336 }, { "epoch": 0.2811457174638487, "grad_norm": 0.5034012198448181, "learning_rate": 9.361111111111111e-06, "loss": 0.4851, "step": 337 }, { "epoch": 0.28197997775305894, "grad_norm": 0.5473969578742981, "learning_rate": 9.38888888888889e-06, "loss": 0.5206, "step": 338 }, { "epoch": 0.2828142380422692, "grad_norm": 0.47353893518447876, "learning_rate": 9.416666666666667e-06, "loss": 0.5048, "step": 339 }, { "epoch": 0.2836484983314794, "grad_norm": 0.5280002951622009, "learning_rate": 9.444444444444445e-06, "loss": 0.5122, "step": 340 }, { "epoch": 0.28448275862068967, "grad_norm": 0.5188562870025635, "learning_rate": 9.472222222222223e-06, "loss": 0.5204, "step": 341 }, { "epoch": 0.2853170189098999, "grad_norm": 0.5873152613639832, "learning_rate": 9.5e-06, "loss": 0.4939, "step": 342 }, { "epoch": 0.2861512791991101, "grad_norm": 0.47561976313591003, "learning_rate": 9.527777777777778e-06, "loss": 0.5012, "step": 343 }, { "epoch": 0.28698553948832034, "grad_norm": 0.6464734673500061, "learning_rate": 9.555555555555556e-06, "loss": 0.5049, "step": 344 }, { "epoch": 0.2878197997775306, "grad_norm": 0.42124444246292114, "learning_rate": 9.583333333333335e-06, "loss": 0.4564, "step": 345 }, { "epoch": 0.2886540600667408, "grad_norm": 0.4998268783092499, "learning_rate": 9.611111111111112e-06, "loss": 0.5007, "step": 346 }, { "epoch": 0.28948832035595107, "grad_norm": 0.4797537326812744, "learning_rate": 9.63888888888889e-06, "loss": 0.5312, "step": 347 }, { "epoch": 0.2903225806451613, "grad_norm": 0.4680449366569519, "learning_rate": 9.666666666666667e-06, "loss": 0.5133, "step": 348 }, { "epoch": 0.2911568409343715, "grad_norm": 0.49188801646232605, "learning_rate": 9.694444444444446e-06, "loss": 0.5028, "step": 349 }, { "epoch": 0.29199110122358174, "grad_norm": 0.43969807028770447, "learning_rate": 9.722222222222223e-06, "loss": 0.4539, "step": 350 }, { "epoch": 0.292825361512792, "grad_norm": 0.5416797995567322, "learning_rate": 9.75e-06, "loss": 0.4817, "step": 351 }, { "epoch": 0.2936596218020022, "grad_norm": 0.5135962963104248, "learning_rate": 9.777777777777779e-06, "loss": 0.5016, "step": 352 }, { "epoch": 0.29449388209121247, "grad_norm": 0.5412265062332153, "learning_rate": 9.805555555555556e-06, "loss": 0.5307, "step": 353 }, { "epoch": 0.2953281423804227, "grad_norm": 0.6637123227119446, "learning_rate": 9.833333333333333e-06, "loss": 0.5158, "step": 354 }, { "epoch": 0.29616240266963295, "grad_norm": 0.5118659734725952, "learning_rate": 9.861111111111112e-06, "loss": 0.4974, "step": 355 }, { "epoch": 0.29699666295884314, "grad_norm": 0.6933585405349731, "learning_rate": 9.88888888888889e-06, "loss": 0.4825, "step": 356 }, { "epoch": 0.2978309232480534, "grad_norm": 0.5294189453125, "learning_rate": 9.916666666666668e-06, "loss": 0.4746, "step": 357 }, { "epoch": 0.2986651835372636, "grad_norm": 0.5193678140640259, "learning_rate": 9.944444444444445e-06, "loss": 0.4908, "step": 358 }, { "epoch": 0.29949944382647387, "grad_norm": 0.5800821185112, "learning_rate": 9.972222222222224e-06, "loss": 0.5392, "step": 359 }, { "epoch": 0.3003337041156841, "grad_norm": 0.5917816758155823, "learning_rate": 1e-05, "loss": 0.5116, "step": 360 }, { "epoch": 0.30116796440489435, "grad_norm": 0.4675118029117584, "learning_rate": 9.999997640827546e-06, "loss": 0.4841, "step": 361 }, { "epoch": 0.30200222469410454, "grad_norm": 0.5768393874168396, "learning_rate": 9.99999056331241e-06, "loss": 0.5098, "step": 362 }, { "epoch": 0.3028364849833148, "grad_norm": 0.5361871123313904, "learning_rate": 9.999978767461268e-06, "loss": 0.4961, "step": 363 }, { "epoch": 0.303670745272525, "grad_norm": 0.48361095786094666, "learning_rate": 9.999962253285255e-06, "loss": 0.4837, "step": 364 }, { "epoch": 0.30450500556173526, "grad_norm": 0.620388925075531, "learning_rate": 9.999941020799956e-06, "loss": 0.5193, "step": 365 }, { "epoch": 0.3053392658509455, "grad_norm": 0.5575003623962402, "learning_rate": 9.999915070025401e-06, "loss": 0.5111, "step": 366 }, { "epoch": 0.30617352614015575, "grad_norm": 0.5322838425636292, "learning_rate": 9.999884400986087e-06, "loss": 0.508, "step": 367 }, { "epoch": 0.30700778642936594, "grad_norm": 0.4816237986087799, "learning_rate": 9.999849013710947e-06, "loss": 0.4901, "step": 368 }, { "epoch": 0.3078420467185762, "grad_norm": 0.5239679217338562, "learning_rate": 9.999808908233382e-06, "loss": 0.5064, "step": 369 }, { "epoch": 0.3086763070077864, "grad_norm": 0.4922145903110504, "learning_rate": 9.999764084591236e-06, "loss": 0.4851, "step": 370 }, { "epoch": 0.30951056729699666, "grad_norm": 0.4868338406085968, "learning_rate": 9.999714542826806e-06, "loss": 0.51, "step": 371 }, { "epoch": 0.3103448275862069, "grad_norm": 0.5109185576438904, "learning_rate": 9.999660282986846e-06, "loss": 0.5002, "step": 372 }, { "epoch": 0.31117908787541715, "grad_norm": 0.5069440603256226, "learning_rate": 9.999601305122555e-06, "loss": 0.5185, "step": 373 }, { "epoch": 0.3120133481646274, "grad_norm": 0.44173693656921387, "learning_rate": 9.999537609289592e-06, "loss": 0.4951, "step": 374 }, { "epoch": 0.3128476084538376, "grad_norm": 0.4982323944568634, "learning_rate": 9.999469195548066e-06, "loss": 0.467, "step": 375 }, { "epoch": 0.3136818687430478, "grad_norm": 0.49110421538352966, "learning_rate": 9.999396063962533e-06, "loss": 0.4924, "step": 376 }, { "epoch": 0.31451612903225806, "grad_norm": 0.48410937190055847, "learning_rate": 9.999318214602008e-06, "loss": 0.4776, "step": 377 }, { "epoch": 0.3153503893214683, "grad_norm": 0.5049404501914978, "learning_rate": 9.999235647539954e-06, "loss": 0.4994, "step": 378 }, { "epoch": 0.31618464961067855, "grad_norm": 0.5054913759231567, "learning_rate": 9.999148362854285e-06, "loss": 0.4761, "step": 379 }, { "epoch": 0.3170189098998888, "grad_norm": 0.49798208475112915, "learning_rate": 9.99905636062737e-06, "loss": 0.5148, "step": 380 }, { "epoch": 0.317853170189099, "grad_norm": 0.46227335929870605, "learning_rate": 9.998959640946033e-06, "loss": 0.4831, "step": 381 }, { "epoch": 0.3186874304783092, "grad_norm": 0.453815221786499, "learning_rate": 9.99885820390154e-06, "loss": 0.4727, "step": 382 }, { "epoch": 0.31952169076751946, "grad_norm": 0.4596490263938904, "learning_rate": 9.998752049589617e-06, "loss": 0.4606, "step": 383 }, { "epoch": 0.3203559510567297, "grad_norm": 0.44425299763679504, "learning_rate": 9.998641178110436e-06, "loss": 0.504, "step": 384 }, { "epoch": 0.32119021134593995, "grad_norm": 0.5349466800689697, "learning_rate": 9.998525589568624e-06, "loss": 0.4827, "step": 385 }, { "epoch": 0.3220244716351502, "grad_norm": 0.48766008019447327, "learning_rate": 9.99840528407326e-06, "loss": 0.482, "step": 386 }, { "epoch": 0.3228587319243604, "grad_norm": 0.48387590050697327, "learning_rate": 9.99828026173787e-06, "loss": 0.4946, "step": 387 }, { "epoch": 0.3236929922135706, "grad_norm": 0.47734877467155457, "learning_rate": 9.998150522680437e-06, "loss": 0.5092, "step": 388 }, { "epoch": 0.32452725250278086, "grad_norm": 0.435073584318161, "learning_rate": 9.998016067023386e-06, "loss": 0.4722, "step": 389 }, { "epoch": 0.3253615127919911, "grad_norm": 0.48233431577682495, "learning_rate": 9.997876894893606e-06, "loss": 0.4895, "step": 390 }, { "epoch": 0.32619577308120135, "grad_norm": 0.4786389470100403, "learning_rate": 9.997733006422423e-06, "loss": 0.5093, "step": 391 }, { "epoch": 0.3270300333704116, "grad_norm": 0.47974127531051636, "learning_rate": 9.997584401745623e-06, "loss": 0.4724, "step": 392 }, { "epoch": 0.3278642936596218, "grad_norm": 0.5294181108474731, "learning_rate": 9.99743108100344e-06, "loss": 0.4734, "step": 393 }, { "epoch": 0.328698553948832, "grad_norm": 0.4545994699001312, "learning_rate": 9.997273044340558e-06, "loss": 0.4937, "step": 394 }, { "epoch": 0.32953281423804226, "grad_norm": 0.5324988961219788, "learning_rate": 9.997110291906109e-06, "loss": 0.4742, "step": 395 }, { "epoch": 0.3303670745272525, "grad_norm": 0.5570563673973083, "learning_rate": 9.996942823853681e-06, "loss": 0.5109, "step": 396 }, { "epoch": 0.33120133481646274, "grad_norm": 0.5499030351638794, "learning_rate": 9.996770640341304e-06, "loss": 0.5021, "step": 397 }, { "epoch": 0.332035595105673, "grad_norm": 0.5759099125862122, "learning_rate": 9.996593741531468e-06, "loss": 0.5016, "step": 398 }, { "epoch": 0.33286985539488323, "grad_norm": 0.526728630065918, "learning_rate": 9.996412127591101e-06, "loss": 0.4708, "step": 399 }, { "epoch": 0.3337041156840934, "grad_norm": 0.51688551902771, "learning_rate": 9.99622579869159e-06, "loss": 0.5196, "step": 400 }, { "epoch": 0.33453837597330366, "grad_norm": 0.4946082532405853, "learning_rate": 9.996034755008767e-06, "loss": 0.4863, "step": 401 }, { "epoch": 0.3353726362625139, "grad_norm": 0.586512565612793, "learning_rate": 9.995838996722916e-06, "loss": 0.5189, "step": 402 }, { "epoch": 0.33620689655172414, "grad_norm": 0.5269147157669067, "learning_rate": 9.995638524018763e-06, "loss": 0.5194, "step": 403 }, { "epoch": 0.3370411568409344, "grad_norm": 0.6343860626220703, "learning_rate": 9.995433337085492e-06, "loss": 0.5005, "step": 404 }, { "epoch": 0.33787541713014463, "grad_norm": 0.512447714805603, "learning_rate": 9.99522343611673e-06, "loss": 0.52, "step": 405 }, { "epoch": 0.3387096774193548, "grad_norm": 0.6386441588401794, "learning_rate": 9.995008821310554e-06, "loss": 0.5065, "step": 406 }, { "epoch": 0.33954393770856506, "grad_norm": 0.4486370086669922, "learning_rate": 9.994789492869491e-06, "loss": 0.5054, "step": 407 }, { "epoch": 0.3403781979977753, "grad_norm": 0.603158175945282, "learning_rate": 9.994565451000513e-06, "loss": 0.5033, "step": 408 }, { "epoch": 0.34121245828698554, "grad_norm": 0.5226010084152222, "learning_rate": 9.994336695915041e-06, "loss": 0.4923, "step": 409 }, { "epoch": 0.3420467185761958, "grad_norm": 0.5318647027015686, "learning_rate": 9.994103227828943e-06, "loss": 0.5073, "step": 410 }, { "epoch": 0.342880978865406, "grad_norm": 0.5574752688407898, "learning_rate": 9.993865046962541e-06, "loss": 0.4769, "step": 411 }, { "epoch": 0.3437152391546162, "grad_norm": 0.44171372056007385, "learning_rate": 9.993622153540592e-06, "loss": 0.477, "step": 412 }, { "epoch": 0.34454949944382646, "grad_norm": 0.5763176083564758, "learning_rate": 9.993374547792311e-06, "loss": 0.5045, "step": 413 }, { "epoch": 0.3453837597330367, "grad_norm": 0.4335787892341614, "learning_rate": 9.993122229951355e-06, "loss": 0.5176, "step": 414 }, { "epoch": 0.34621802002224694, "grad_norm": 0.4284120798110962, "learning_rate": 9.992865200255829e-06, "loss": 0.4744, "step": 415 }, { "epoch": 0.3470522803114572, "grad_norm": 0.5243945121765137, "learning_rate": 9.992603458948282e-06, "loss": 0.4834, "step": 416 }, { "epoch": 0.3478865406006674, "grad_norm": 0.4780410826206207, "learning_rate": 9.992337006275715e-06, "loss": 0.4844, "step": 417 }, { "epoch": 0.34872080088987767, "grad_norm": 0.5417578816413879, "learning_rate": 9.992065842489567e-06, "loss": 0.5019, "step": 418 }, { "epoch": 0.34955506117908786, "grad_norm": 0.4520585834980011, "learning_rate": 9.991789967845729e-06, "loss": 0.4545, "step": 419 }, { "epoch": 0.3503893214682981, "grad_norm": 0.43547332286834717, "learning_rate": 9.991509382604534e-06, "loss": 0.4928, "step": 420 }, { "epoch": 0.35122358175750834, "grad_norm": 0.5224993228912354, "learning_rate": 9.991224087030763e-06, "loss": 0.4788, "step": 421 }, { "epoch": 0.3520578420467186, "grad_norm": 0.4621793329715729, "learning_rate": 9.990934081393641e-06, "loss": 0.4873, "step": 422 }, { "epoch": 0.3528921023359288, "grad_norm": 0.4531457722187042, "learning_rate": 9.990639365966835e-06, "loss": 0.4778, "step": 423 }, { "epoch": 0.35372636262513907, "grad_norm": 0.5211503505706787, "learning_rate": 9.990339941028463e-06, "loss": 0.5001, "step": 424 }, { "epoch": 0.35456062291434925, "grad_norm": 0.4248245358467102, "learning_rate": 9.990035806861076e-06, "loss": 0.4838, "step": 425 }, { "epoch": 0.3553948832035595, "grad_norm": 0.45071858167648315, "learning_rate": 9.989726963751683e-06, "loss": 0.4825, "step": 426 }, { "epoch": 0.35622914349276974, "grad_norm": 0.4202861785888672, "learning_rate": 9.989413411991727e-06, "loss": 0.489, "step": 427 }, { "epoch": 0.35706340378198, "grad_norm": 0.3932492136955261, "learning_rate": 9.989095151877096e-06, "loss": 0.4904, "step": 428 }, { "epoch": 0.3578976640711902, "grad_norm": 0.4064221978187561, "learning_rate": 9.988772183708122e-06, "loss": 0.4747, "step": 429 }, { "epoch": 0.35873192436040047, "grad_norm": 0.41099366545677185, "learning_rate": 9.988444507789584e-06, "loss": 0.4638, "step": 430 }, { "epoch": 0.35956618464961065, "grad_norm": 0.3796460032463074, "learning_rate": 9.988112124430693e-06, "loss": 0.4765, "step": 431 }, { "epoch": 0.3604004449388209, "grad_norm": 0.44086387753486633, "learning_rate": 9.987775033945116e-06, "loss": 0.486, "step": 432 }, { "epoch": 0.36123470522803114, "grad_norm": 0.4344269037246704, "learning_rate": 9.98743323665095e-06, "loss": 0.4945, "step": 433 }, { "epoch": 0.3620689655172414, "grad_norm": 0.4325614869594574, "learning_rate": 9.98708673287074e-06, "loss": 0.4839, "step": 434 }, { "epoch": 0.3629032258064516, "grad_norm": 0.4301924407482147, "learning_rate": 9.986735522931471e-06, "loss": 0.4858, "step": 435 }, { "epoch": 0.36373748609566187, "grad_norm": 0.4270024001598358, "learning_rate": 9.986379607164571e-06, "loss": 0.4844, "step": 436 }, { "epoch": 0.36457174638487205, "grad_norm": 0.4385048449039459, "learning_rate": 9.986018985905901e-06, "loss": 0.4798, "step": 437 }, { "epoch": 0.3654060066740823, "grad_norm": 0.4530482292175293, "learning_rate": 9.985653659495773e-06, "loss": 0.4969, "step": 438 }, { "epoch": 0.36624026696329254, "grad_norm": 0.42587095499038696, "learning_rate": 9.985283628278934e-06, "loss": 0.4707, "step": 439 }, { "epoch": 0.3670745272525028, "grad_norm": 0.45039263367652893, "learning_rate": 9.984908892604568e-06, "loss": 0.4849, "step": 440 }, { "epoch": 0.367908787541713, "grad_norm": 0.4711659252643585, "learning_rate": 9.984529452826304e-06, "loss": 0.5099, "step": 441 }, { "epoch": 0.36874304783092327, "grad_norm": 0.41641736030578613, "learning_rate": 9.984145309302207e-06, "loss": 0.474, "step": 442 }, { "epoch": 0.3695773081201335, "grad_norm": 0.5063542127609253, "learning_rate": 9.98375646239478e-06, "loss": 0.527, "step": 443 }, { "epoch": 0.3704115684093437, "grad_norm": 0.42919498682022095, "learning_rate": 9.983362912470967e-06, "loss": 0.4975, "step": 444 }, { "epoch": 0.37124582869855394, "grad_norm": 0.4322325885295868, "learning_rate": 9.982964659902148e-06, "loss": 0.4645, "step": 445 }, { "epoch": 0.3720800889877642, "grad_norm": 0.4132075607776642, "learning_rate": 9.982561705064144e-06, "loss": 0.4628, "step": 446 }, { "epoch": 0.3729143492769744, "grad_norm": 0.41358184814453125, "learning_rate": 9.982154048337209e-06, "loss": 0.4772, "step": 447 }, { "epoch": 0.37374860956618466, "grad_norm": 0.438781201839447, "learning_rate": 9.981741690106035e-06, "loss": 0.4858, "step": 448 }, { "epoch": 0.3745828698553949, "grad_norm": 0.43460530042648315, "learning_rate": 9.981324630759755e-06, "loss": 0.4727, "step": 449 }, { "epoch": 0.3754171301446051, "grad_norm": 0.463827908039093, "learning_rate": 9.980902870691931e-06, "loss": 0.4908, "step": 450 }, { "epoch": 0.37625139043381534, "grad_norm": 0.4200926423072815, "learning_rate": 9.980476410300567e-06, "loss": 0.4912, "step": 451 }, { "epoch": 0.3770856507230256, "grad_norm": 0.6460418105125427, "learning_rate": 9.9800452499881e-06, "loss": 0.4783, "step": 452 }, { "epoch": 0.3779199110122358, "grad_norm": 0.5319968461990356, "learning_rate": 9.979609390161405e-06, "loss": 0.4987, "step": 453 }, { "epoch": 0.37875417130144606, "grad_norm": 0.6477885842323303, "learning_rate": 9.979168831231785e-06, "loss": 0.4922, "step": 454 }, { "epoch": 0.3795884315906563, "grad_norm": 0.6888568997383118, "learning_rate": 9.978723573614984e-06, "loss": 0.5, "step": 455 }, { "epoch": 0.3804226918798665, "grad_norm": 0.5325235724449158, "learning_rate": 9.97827361773118e-06, "loss": 0.5245, "step": 456 }, { "epoch": 0.38125695216907673, "grad_norm": 0.5640732049942017, "learning_rate": 9.977818964004979e-06, "loss": 0.4947, "step": 457 }, { "epoch": 0.382091212458287, "grad_norm": 0.4858129620552063, "learning_rate": 9.977359612865424e-06, "loss": 0.4929, "step": 458 }, { "epoch": 0.3829254727474972, "grad_norm": 0.47492972016334534, "learning_rate": 9.976895564745993e-06, "loss": 0.4946, "step": 459 }, { "epoch": 0.38375973303670746, "grad_norm": 0.44178715348243713, "learning_rate": 9.97642682008459e-06, "loss": 0.4951, "step": 460 }, { "epoch": 0.3845939933259177, "grad_norm": 0.4377753436565399, "learning_rate": 9.975953379323557e-06, "loss": 0.4769, "step": 461 }, { "epoch": 0.38542825361512795, "grad_norm": 0.4591297209262848, "learning_rate": 9.975475242909667e-06, "loss": 0.504, "step": 462 }, { "epoch": 0.38626251390433813, "grad_norm": 0.43687117099761963, "learning_rate": 9.97499241129412e-06, "loss": 0.4962, "step": 463 }, { "epoch": 0.3870967741935484, "grad_norm": 0.4481161832809448, "learning_rate": 9.97450488493255e-06, "loss": 0.5073, "step": 464 }, { "epoch": 0.3879310344827586, "grad_norm": 0.4236924648284912, "learning_rate": 9.97401266428502e-06, "loss": 0.5243, "step": 465 }, { "epoch": 0.38876529477196886, "grad_norm": 0.4433425962924957, "learning_rate": 9.973515749816024e-06, "loss": 0.5051, "step": 466 }, { "epoch": 0.3895995550611791, "grad_norm": 0.4528418779373169, "learning_rate": 9.973014141994487e-06, "loss": 0.5372, "step": 467 }, { "epoch": 0.39043381535038935, "grad_norm": 0.4657782018184662, "learning_rate": 9.972507841293756e-06, "loss": 0.4727, "step": 468 }, { "epoch": 0.39126807563959953, "grad_norm": 0.4455312192440033, "learning_rate": 9.971996848191613e-06, "loss": 0.4879, "step": 469 }, { "epoch": 0.3921023359288098, "grad_norm": 0.4593108594417572, "learning_rate": 9.97148116317027e-06, "loss": 0.4913, "step": 470 }, { "epoch": 0.39293659621802, "grad_norm": 0.4544794261455536, "learning_rate": 9.970960786716357e-06, "loss": 0.4883, "step": 471 }, { "epoch": 0.39377085650723026, "grad_norm": 0.45691779255867004, "learning_rate": 9.97043571932094e-06, "loss": 0.5016, "step": 472 }, { "epoch": 0.3946051167964405, "grad_norm": 0.484771728515625, "learning_rate": 9.969905961479512e-06, "loss": 0.4857, "step": 473 }, { "epoch": 0.39543937708565075, "grad_norm": 0.4506840705871582, "learning_rate": 9.969371513691983e-06, "loss": 0.4768, "step": 474 }, { "epoch": 0.39627363737486093, "grad_norm": 0.48172155022621155, "learning_rate": 9.968832376462698e-06, "loss": 0.475, "step": 475 }, { "epoch": 0.3971078976640712, "grad_norm": 0.46071380376815796, "learning_rate": 9.968288550300424e-06, "loss": 0.4659, "step": 476 }, { "epoch": 0.3979421579532814, "grad_norm": 0.4384761452674866, "learning_rate": 9.967740035718353e-06, "loss": 0.4994, "step": 477 }, { "epoch": 0.39877641824249166, "grad_norm": 0.47761425375938416, "learning_rate": 9.967186833234101e-06, "loss": 0.4665, "step": 478 }, { "epoch": 0.3996106785317019, "grad_norm": 0.4359094202518463, "learning_rate": 9.966628943369708e-06, "loss": 0.4904, "step": 479 }, { "epoch": 0.40044493882091214, "grad_norm": 0.4435771703720093, "learning_rate": 9.966066366651636e-06, "loss": 0.4957, "step": 480 }, { "epoch": 0.40127919911012233, "grad_norm": 0.4432840943336487, "learning_rate": 9.965499103610775e-06, "loss": 0.4808, "step": 481 }, { "epoch": 0.4021134593993326, "grad_norm": 0.43742862343788147, "learning_rate": 9.964927154782429e-06, "loss": 0.5138, "step": 482 }, { "epoch": 0.4029477196885428, "grad_norm": 0.4387785792350769, "learning_rate": 9.964350520706329e-06, "loss": 0.4807, "step": 483 }, { "epoch": 0.40378197997775306, "grad_norm": 0.5197587609291077, "learning_rate": 9.96376920192663e-06, "loss": 0.4921, "step": 484 }, { "epoch": 0.4046162402669633, "grad_norm": 0.41846150159835815, "learning_rate": 9.963183198991904e-06, "loss": 0.4838, "step": 485 }, { "epoch": 0.40545050055617354, "grad_norm": 0.5465821027755737, "learning_rate": 9.96259251245514e-06, "loss": 0.4987, "step": 486 }, { "epoch": 0.4062847608453838, "grad_norm": 0.4456709325313568, "learning_rate": 9.961997142873754e-06, "loss": 0.4851, "step": 487 }, { "epoch": 0.407119021134594, "grad_norm": 0.4611912667751312, "learning_rate": 9.961397090809575e-06, "loss": 0.4838, "step": 488 }, { "epoch": 0.4079532814238042, "grad_norm": 0.4509412348270416, "learning_rate": 9.960792356828858e-06, "loss": 0.5275, "step": 489 }, { "epoch": 0.40878754171301446, "grad_norm": 0.4466070830821991, "learning_rate": 9.960182941502265e-06, "loss": 0.5086, "step": 490 }, { "epoch": 0.4096218020022247, "grad_norm": 0.45920494198799133, "learning_rate": 9.95956884540489e-06, "loss": 0.4815, "step": 491 }, { "epoch": 0.41045606229143494, "grad_norm": 0.42112889885902405, "learning_rate": 9.95895006911623e-06, "loss": 0.4608, "step": 492 }, { "epoch": 0.4112903225806452, "grad_norm": 0.5470235347747803, "learning_rate": 9.95832661322021e-06, "loss": 0.4973, "step": 493 }, { "epoch": 0.41212458286985537, "grad_norm": 0.4282953143119812, "learning_rate": 9.957698478305164e-06, "loss": 0.4745, "step": 494 }, { "epoch": 0.4129588431590656, "grad_norm": 0.46226781606674194, "learning_rate": 9.957065664963841e-06, "loss": 0.5027, "step": 495 }, { "epoch": 0.41379310344827586, "grad_norm": 0.47396498918533325, "learning_rate": 9.95642817379341e-06, "loss": 0.4879, "step": 496 }, { "epoch": 0.4146273637374861, "grad_norm": 0.4627336859703064, "learning_rate": 9.955786005395451e-06, "loss": 0.4954, "step": 497 }, { "epoch": 0.41546162402669634, "grad_norm": 0.4375038146972656, "learning_rate": 9.95513916037596e-06, "loss": 0.5054, "step": 498 }, { "epoch": 0.4162958843159066, "grad_norm": 0.4313870966434479, "learning_rate": 9.954487639345343e-06, "loss": 0.496, "step": 499 }, { "epoch": 0.41713014460511677, "grad_norm": 0.46265676617622375, "learning_rate": 9.953831442918418e-06, "loss": 0.4598, "step": 500 }, { "epoch": 0.417964404894327, "grad_norm": 0.48188352584838867, "learning_rate": 9.953170571714421e-06, "loss": 0.5118, "step": 501 }, { "epoch": 0.41879866518353726, "grad_norm": 0.4584149122238159, "learning_rate": 9.952505026356994e-06, "loss": 0.4755, "step": 502 }, { "epoch": 0.4196329254727475, "grad_norm": 0.5108729600906372, "learning_rate": 9.951834807474191e-06, "loss": 0.5054, "step": 503 }, { "epoch": 0.42046718576195774, "grad_norm": 0.4378891885280609, "learning_rate": 9.951159915698478e-06, "loss": 0.4774, "step": 504 }, { "epoch": 0.421301446051168, "grad_norm": 0.4527071714401245, "learning_rate": 9.950480351666729e-06, "loss": 0.4381, "step": 505 }, { "epoch": 0.4221357063403782, "grad_norm": 0.5536148548126221, "learning_rate": 9.949796116020227e-06, "loss": 0.5131, "step": 506 }, { "epoch": 0.4229699666295884, "grad_norm": 0.5109352469444275, "learning_rate": 9.949107209404664e-06, "loss": 0.4947, "step": 507 }, { "epoch": 0.42380422691879865, "grad_norm": 0.4513104259967804, "learning_rate": 9.94841363247014e-06, "loss": 0.4508, "step": 508 }, { "epoch": 0.4246384872080089, "grad_norm": 0.5248517394065857, "learning_rate": 9.947715385871164e-06, "loss": 0.4744, "step": 509 }, { "epoch": 0.42547274749721914, "grad_norm": 0.5093575119972229, "learning_rate": 9.947012470266645e-06, "loss": 0.5125, "step": 510 }, { "epoch": 0.4263070077864294, "grad_norm": 0.4292238652706146, "learning_rate": 9.946304886319908e-06, "loss": 0.4666, "step": 511 }, { "epoch": 0.4271412680756396, "grad_norm": 0.452281653881073, "learning_rate": 9.945592634698674e-06, "loss": 0.4772, "step": 512 }, { "epoch": 0.4279755283648498, "grad_norm": 0.4998204708099365, "learning_rate": 9.944875716075076e-06, "loss": 0.5022, "step": 513 }, { "epoch": 0.42880978865406005, "grad_norm": 0.4213614761829376, "learning_rate": 9.944154131125643e-06, "loss": 0.4919, "step": 514 }, { "epoch": 0.4296440489432703, "grad_norm": 0.4302680492401123, "learning_rate": 9.943427880531317e-06, "loss": 0.4931, "step": 515 }, { "epoch": 0.43047830923248054, "grad_norm": 0.5021787285804749, "learning_rate": 9.942696964977438e-06, "loss": 0.5154, "step": 516 }, { "epoch": 0.4313125695216908, "grad_norm": 0.4618604779243469, "learning_rate": 9.941961385153744e-06, "loss": 0.4936, "step": 517 }, { "epoch": 0.432146829810901, "grad_norm": 0.44708317518234253, "learning_rate": 9.941221141754385e-06, "loss": 0.4574, "step": 518 }, { "epoch": 0.4329810901001112, "grad_norm": 0.43967655301094055, "learning_rate": 9.9404762354779e-06, "loss": 0.4705, "step": 519 }, { "epoch": 0.43381535038932145, "grad_norm": 0.4660184681415558, "learning_rate": 9.939726667027238e-06, "loss": 0.4866, "step": 520 }, { "epoch": 0.4346496106785317, "grad_norm": 0.5399330258369446, "learning_rate": 9.938972437109742e-06, "loss": 0.5027, "step": 521 }, { "epoch": 0.43548387096774194, "grad_norm": 0.48371875286102295, "learning_rate": 9.938213546437154e-06, "loss": 0.5069, "step": 522 }, { "epoch": 0.4363181312569522, "grad_norm": 0.4957859218120575, "learning_rate": 9.937449995725617e-06, "loss": 0.4981, "step": 523 }, { "epoch": 0.4371523915461624, "grad_norm": 0.47849997878074646, "learning_rate": 9.936681785695672e-06, "loss": 0.4997, "step": 524 }, { "epoch": 0.4379866518353726, "grad_norm": 0.4498022794723511, "learning_rate": 9.935908917072253e-06, "loss": 0.498, "step": 525 }, { "epoch": 0.43882091212458285, "grad_norm": 0.44779595732688904, "learning_rate": 9.935131390584689e-06, "loss": 0.4813, "step": 526 }, { "epoch": 0.4396551724137931, "grad_norm": 0.4847566783428192, "learning_rate": 9.934349206966714e-06, "loss": 0.4734, "step": 527 }, { "epoch": 0.44048943270300334, "grad_norm": 0.4306754469871521, "learning_rate": 9.933562366956445e-06, "loss": 0.4786, "step": 528 }, { "epoch": 0.4413236929922136, "grad_norm": 0.466869592666626, "learning_rate": 9.932770871296403e-06, "loss": 0.4618, "step": 529 }, { "epoch": 0.4421579532814238, "grad_norm": 0.428120881319046, "learning_rate": 9.931974720733494e-06, "loss": 0.5106, "step": 530 }, { "epoch": 0.44299221357063406, "grad_norm": 0.4016896188259125, "learning_rate": 9.931173916019022e-06, "loss": 0.4861, "step": 531 }, { "epoch": 0.44382647385984425, "grad_norm": 0.46770769357681274, "learning_rate": 9.930368457908683e-06, "loss": 0.4593, "step": 532 }, { "epoch": 0.4446607341490545, "grad_norm": 0.4045330584049225, "learning_rate": 9.929558347162561e-06, "loss": 0.4815, "step": 533 }, { "epoch": 0.44549499443826474, "grad_norm": 0.5035044550895691, "learning_rate": 9.928743584545133e-06, "loss": 0.4862, "step": 534 }, { "epoch": 0.446329254727475, "grad_norm": 0.47986865043640137, "learning_rate": 9.927924170825266e-06, "loss": 0.4592, "step": 535 }, { "epoch": 0.4471635150166852, "grad_norm": 0.43214163184165955, "learning_rate": 9.927100106776213e-06, "loss": 0.506, "step": 536 }, { "epoch": 0.44799777530589546, "grad_norm": 0.4819169044494629, "learning_rate": 9.92627139317562e-06, "loss": 0.5073, "step": 537 }, { "epoch": 0.44883203559510565, "grad_norm": 0.4268386960029602, "learning_rate": 9.925438030805518e-06, "loss": 0.4525, "step": 538 }, { "epoch": 0.4496662958843159, "grad_norm": 0.46364253759384155, "learning_rate": 9.924600020452325e-06, "loss": 0.5019, "step": 539 }, { "epoch": 0.45050055617352613, "grad_norm": 0.43082213401794434, "learning_rate": 9.923757362906845e-06, "loss": 0.4815, "step": 540 }, { "epoch": 0.4513348164627364, "grad_norm": 0.452652245759964, "learning_rate": 9.922910058964268e-06, "loss": 0.4657, "step": 541 }, { "epoch": 0.4521690767519466, "grad_norm": 0.4956667721271515, "learning_rate": 9.922058109424168e-06, "loss": 0.4845, "step": 542 }, { "epoch": 0.45300333704115686, "grad_norm": 0.4981364905834198, "learning_rate": 9.921201515090503e-06, "loss": 0.4419, "step": 543 }, { "epoch": 0.45383759733036705, "grad_norm": 0.4571170508861542, "learning_rate": 9.920340276771615e-06, "loss": 0.5189, "step": 544 }, { "epoch": 0.4546718576195773, "grad_norm": 0.4796139597892761, "learning_rate": 9.91947439528023e-06, "loss": 0.4773, "step": 545 }, { "epoch": 0.45550611790878753, "grad_norm": 0.41199803352355957, "learning_rate": 9.91860387143345e-06, "loss": 0.4556, "step": 546 }, { "epoch": 0.4563403781979978, "grad_norm": 0.5168163776397705, "learning_rate": 9.917728706052765e-06, "loss": 0.4756, "step": 547 }, { "epoch": 0.457174638487208, "grad_norm": 0.47592148184776306, "learning_rate": 9.916848899964037e-06, "loss": 0.4964, "step": 548 }, { "epoch": 0.45800889877641826, "grad_norm": 0.4397949278354645, "learning_rate": 9.915964453997516e-06, "loss": 0.4643, "step": 549 }, { "epoch": 0.4588431590656285, "grad_norm": 0.5119925737380981, "learning_rate": 9.915075368987824e-06, "loss": 0.4772, "step": 550 }, { "epoch": 0.4596774193548387, "grad_norm": 0.3966693878173828, "learning_rate": 9.914181645773962e-06, "loss": 0.4384, "step": 551 }, { "epoch": 0.46051167964404893, "grad_norm": 0.5062504410743713, "learning_rate": 9.913283285199311e-06, "loss": 0.476, "step": 552 }, { "epoch": 0.4613459399332592, "grad_norm": 0.5168266892433167, "learning_rate": 9.912380288111625e-06, "loss": 0.4769, "step": 553 }, { "epoch": 0.4621802002224694, "grad_norm": 0.45059117674827576, "learning_rate": 9.911472655363035e-06, "loss": 0.4891, "step": 554 }, { "epoch": 0.46301446051167966, "grad_norm": 0.42274874448776245, "learning_rate": 9.910560387810046e-06, "loss": 0.4634, "step": 555 }, { "epoch": 0.4638487208008899, "grad_norm": 0.45587602257728577, "learning_rate": 9.909643486313533e-06, "loss": 0.4476, "step": 556 }, { "epoch": 0.4646829810901001, "grad_norm": 0.46408215165138245, "learning_rate": 9.908721951738754e-06, "loss": 0.4799, "step": 557 }, { "epoch": 0.46551724137931033, "grad_norm": 0.4534507393836975, "learning_rate": 9.907795784955327e-06, "loss": 0.4474, "step": 558 }, { "epoch": 0.4663515016685206, "grad_norm": 0.4155801236629486, "learning_rate": 9.90686498683725e-06, "loss": 0.4847, "step": 559 }, { "epoch": 0.4671857619577308, "grad_norm": 0.4850746989250183, "learning_rate": 9.905929558262887e-06, "loss": 0.4761, "step": 560 }, { "epoch": 0.46802002224694106, "grad_norm": 0.4280869960784912, "learning_rate": 9.904989500114973e-06, "loss": 0.4571, "step": 561 }, { "epoch": 0.4688542825361513, "grad_norm": 0.5070300102233887, "learning_rate": 9.904044813280613e-06, "loss": 0.4919, "step": 562 }, { "epoch": 0.4696885428253615, "grad_norm": 0.4651212692260742, "learning_rate": 9.903095498651276e-06, "loss": 0.4897, "step": 563 }, { "epoch": 0.47052280311457173, "grad_norm": 0.501735508441925, "learning_rate": 9.902141557122802e-06, "loss": 0.4958, "step": 564 }, { "epoch": 0.471357063403782, "grad_norm": 0.40163740515708923, "learning_rate": 9.901182989595397e-06, "loss": 0.459, "step": 565 }, { "epoch": 0.4721913236929922, "grad_norm": 0.38681790232658386, "learning_rate": 9.900219796973631e-06, "loss": 0.4696, "step": 566 }, { "epoch": 0.47302558398220246, "grad_norm": 0.42819729447364807, "learning_rate": 9.89925198016644e-06, "loss": 0.4889, "step": 567 }, { "epoch": 0.4738598442714127, "grad_norm": 0.400549978017807, "learning_rate": 9.89827954008712e-06, "loss": 0.4812, "step": 568 }, { "epoch": 0.4746941045606229, "grad_norm": 0.4154485762119293, "learning_rate": 9.897302477653334e-06, "loss": 0.5118, "step": 569 }, { "epoch": 0.47552836484983313, "grad_norm": 0.4281671941280365, "learning_rate": 9.896320793787106e-06, "loss": 0.4561, "step": 570 }, { "epoch": 0.47636262513904337, "grad_norm": 0.4837651252746582, "learning_rate": 9.89533448941482e-06, "loss": 0.4998, "step": 571 }, { "epoch": 0.4771968854282536, "grad_norm": 0.4793055057525635, "learning_rate": 9.894343565467222e-06, "loss": 0.4816, "step": 572 }, { "epoch": 0.47803114571746386, "grad_norm": 0.4413088858127594, "learning_rate": 9.893348022879414e-06, "loss": 0.4711, "step": 573 }, { "epoch": 0.4788654060066741, "grad_norm": 0.47595512866973877, "learning_rate": 9.892347862590861e-06, "loss": 0.4814, "step": 574 }, { "epoch": 0.47969966629588434, "grad_norm": 0.47067710757255554, "learning_rate": 9.89134308554538e-06, "loss": 0.4719, "step": 575 }, { "epoch": 0.48053392658509453, "grad_norm": 0.4779387414455414, "learning_rate": 9.890333692691152e-06, "loss": 0.4813, "step": 576 }, { "epoch": 0.48136818687430477, "grad_norm": 0.4446781873703003, "learning_rate": 9.889319684980707e-06, "loss": 0.485, "step": 577 }, { "epoch": 0.482202447163515, "grad_norm": 0.43165257573127747, "learning_rate": 9.888301063370934e-06, "loss": 0.4699, "step": 578 }, { "epoch": 0.48303670745272526, "grad_norm": 0.42691269516944885, "learning_rate": 9.887277828823074e-06, "loss": 0.4891, "step": 579 }, { "epoch": 0.4838709677419355, "grad_norm": 0.4564470946788788, "learning_rate": 9.88624998230272e-06, "loss": 0.4659, "step": 580 }, { "epoch": 0.48470522803114574, "grad_norm": 0.4319731593132019, "learning_rate": 9.885217524779823e-06, "loss": 0.4881, "step": 581 }, { "epoch": 0.4855394883203559, "grad_norm": 0.38786938786506653, "learning_rate": 9.884180457228678e-06, "loss": 0.4707, "step": 582 }, { "epoch": 0.48637374860956617, "grad_norm": 0.4109518826007843, "learning_rate": 9.883138780627933e-06, "loss": 0.4436, "step": 583 }, { "epoch": 0.4872080088987764, "grad_norm": 0.43407949805259705, "learning_rate": 9.882092495960589e-06, "loss": 0.4828, "step": 584 }, { "epoch": 0.48804226918798665, "grad_norm": 0.4262412488460541, "learning_rate": 9.881041604213988e-06, "loss": 0.4927, "step": 585 }, { "epoch": 0.4888765294771969, "grad_norm": 0.408741295337677, "learning_rate": 9.879986106379829e-06, "loss": 0.4907, "step": 586 }, { "epoch": 0.48971078976640714, "grad_norm": 0.4254915416240692, "learning_rate": 9.878926003454149e-06, "loss": 0.4604, "step": 587 }, { "epoch": 0.4905450500556173, "grad_norm": 0.4102683365345001, "learning_rate": 9.877861296437334e-06, "loss": 0.4955, "step": 588 }, { "epoch": 0.49137931034482757, "grad_norm": 0.4352142810821533, "learning_rate": 9.876791986334118e-06, "loss": 0.4559, "step": 589 }, { "epoch": 0.4922135706340378, "grad_norm": 0.42291560769081116, "learning_rate": 9.875718074153572e-06, "loss": 0.44, "step": 590 }, { "epoch": 0.49304783092324805, "grad_norm": 0.47134658694267273, "learning_rate": 9.874639560909118e-06, "loss": 0.4866, "step": 591 }, { "epoch": 0.4938820912124583, "grad_norm": 0.48203232884407043, "learning_rate": 9.873556447618513e-06, "loss": 0.4874, "step": 592 }, { "epoch": 0.49471635150166854, "grad_norm": 0.38159817457199097, "learning_rate": 9.872468735303857e-06, "loss": 0.4592, "step": 593 }, { "epoch": 0.4955506117908788, "grad_norm": 0.41757336258888245, "learning_rate": 9.87137642499159e-06, "loss": 0.4756, "step": 594 }, { "epoch": 0.49638487208008897, "grad_norm": 0.46518248319625854, "learning_rate": 9.870279517712494e-06, "loss": 0.468, "step": 595 }, { "epoch": 0.4972191323692992, "grad_norm": 0.43568670749664307, "learning_rate": 9.869178014501684e-06, "loss": 0.4858, "step": 596 }, { "epoch": 0.49805339265850945, "grad_norm": 0.41285020112991333, "learning_rate": 9.868071916398613e-06, "loss": 0.5006, "step": 597 }, { "epoch": 0.4988876529477197, "grad_norm": 0.4788348972797394, "learning_rate": 9.866961224447076e-06, "loss": 0.4889, "step": 598 }, { "epoch": 0.49972191323692994, "grad_norm": 0.44009971618652344, "learning_rate": 9.865845939695195e-06, "loss": 0.4553, "step": 599 }, { "epoch": 0.5005561735261401, "grad_norm": 0.46149957180023193, "learning_rate": 9.86472606319543e-06, "loss": 0.4921, "step": 600 }, { "epoch": 0.5013904338153504, "grad_norm": 0.46000486612319946, "learning_rate": 9.863601596004577e-06, "loss": 0.485, "step": 601 }, { "epoch": 0.5022246941045606, "grad_norm": 0.49990227818489075, "learning_rate": 9.862472539183757e-06, "loss": 0.4934, "step": 602 }, { "epoch": 0.5030589543937709, "grad_norm": 0.4986514449119568, "learning_rate": 9.861338893798426e-06, "loss": 0.4741, "step": 603 }, { "epoch": 0.5038932146829811, "grad_norm": 0.49131903052330017, "learning_rate": 9.860200660918371e-06, "loss": 0.4608, "step": 604 }, { "epoch": 0.5047274749721913, "grad_norm": 0.4317084550857544, "learning_rate": 9.859057841617709e-06, "loss": 0.4888, "step": 605 }, { "epoch": 0.5055617352614016, "grad_norm": 0.464290976524353, "learning_rate": 9.85791043697488e-06, "loss": 0.5016, "step": 606 }, { "epoch": 0.5063959955506118, "grad_norm": 0.4447185695171356, "learning_rate": 9.856758448072657e-06, "loss": 0.4784, "step": 607 }, { "epoch": 0.5072302558398221, "grad_norm": 0.39162972569465637, "learning_rate": 9.855601875998134e-06, "loss": 0.4837, "step": 608 }, { "epoch": 0.5080645161290323, "grad_norm": 0.4513096213340759, "learning_rate": 9.854440721842732e-06, "loss": 0.4759, "step": 609 }, { "epoch": 0.5088987764182424, "grad_norm": 0.45903152227401733, "learning_rate": 9.853274986702197e-06, "loss": 0.4849, "step": 610 }, { "epoch": 0.5097330367074527, "grad_norm": 0.36196327209472656, "learning_rate": 9.852104671676598e-06, "loss": 0.4745, "step": 611 }, { "epoch": 0.5105672969966629, "grad_norm": 0.44563594460487366, "learning_rate": 9.850929777870324e-06, "loss": 0.4698, "step": 612 }, { "epoch": 0.5114015572858732, "grad_norm": 0.4242342412471771, "learning_rate": 9.849750306392085e-06, "loss": 0.4825, "step": 613 }, { "epoch": 0.5122358175750834, "grad_norm": 0.47234001755714417, "learning_rate": 9.848566258354913e-06, "loss": 0.4841, "step": 614 }, { "epoch": 0.5130700778642937, "grad_norm": 0.42218708992004395, "learning_rate": 9.847377634876158e-06, "loss": 0.463, "step": 615 }, { "epoch": 0.5139043381535039, "grad_norm": 0.44886618852615356, "learning_rate": 9.846184437077483e-06, "loss": 0.4913, "step": 616 }, { "epoch": 0.5147385984427141, "grad_norm": 0.4369449019432068, "learning_rate": 9.84498666608488e-06, "loss": 0.4952, "step": 617 }, { "epoch": 0.5155728587319244, "grad_norm": 0.3921835124492645, "learning_rate": 9.843784323028639e-06, "loss": 0.4897, "step": 618 }, { "epoch": 0.5164071190211346, "grad_norm": 0.4219275414943695, "learning_rate": 9.84257740904338e-06, "loss": 0.5026, "step": 619 }, { "epoch": 0.5172413793103449, "grad_norm": 0.4749930202960968, "learning_rate": 9.841365925268027e-06, "loss": 0.4729, "step": 620 }, { "epoch": 0.518075639599555, "grad_norm": 0.4099540412425995, "learning_rate": 9.840149872845822e-06, "loss": 0.471, "step": 621 }, { "epoch": 0.5189098998887653, "grad_norm": 0.43642109632492065, "learning_rate": 9.838929252924313e-06, "loss": 0.4626, "step": 622 }, { "epoch": 0.5197441601779755, "grad_norm": 0.41526392102241516, "learning_rate": 9.837704066655367e-06, "loss": 0.5198, "step": 623 }, { "epoch": 0.5205784204671857, "grad_norm": 0.4362016022205353, "learning_rate": 9.836474315195148e-06, "loss": 0.469, "step": 624 }, { "epoch": 0.521412680756396, "grad_norm": 0.41387003660202026, "learning_rate": 9.835239999704136e-06, "loss": 0.4761, "step": 625 }, { "epoch": 0.5222469410456062, "grad_norm": 0.4288390576839447, "learning_rate": 9.83400112134712e-06, "loss": 0.4581, "step": 626 }, { "epoch": 0.5230812013348165, "grad_norm": 0.42333710193634033, "learning_rate": 9.832757681293185e-06, "loss": 0.4836, "step": 627 }, { "epoch": 0.5239154616240267, "grad_norm": 0.3941848576068878, "learning_rate": 9.83150968071573e-06, "loss": 0.4722, "step": 628 }, { "epoch": 0.5247497219132369, "grad_norm": 0.4287227392196655, "learning_rate": 9.830257120792455e-06, "loss": 0.4404, "step": 629 }, { "epoch": 0.5255839822024472, "grad_norm": 0.4020238220691681, "learning_rate": 9.82900000270536e-06, "loss": 0.4804, "step": 630 }, { "epoch": 0.5264182424916574, "grad_norm": 0.38638994097709656, "learning_rate": 9.827738327640753e-06, "loss": 0.4553, "step": 631 }, { "epoch": 0.5272525027808677, "grad_norm": 0.407690167427063, "learning_rate": 9.82647209678923e-06, "loss": 0.4781, "step": 632 }, { "epoch": 0.5280867630700778, "grad_norm": 0.3712944984436035, "learning_rate": 9.8252013113457e-06, "loss": 0.4867, "step": 633 }, { "epoch": 0.5289210233592881, "grad_norm": 0.40403303503990173, "learning_rate": 9.823925972509361e-06, "loss": 0.456, "step": 634 }, { "epoch": 0.5297552836484983, "grad_norm": 0.4010281562805176, "learning_rate": 9.822646081483713e-06, "loss": 0.4687, "step": 635 }, { "epoch": 0.5305895439377085, "grad_norm": 0.42941078543663025, "learning_rate": 9.821361639476548e-06, "loss": 0.5016, "step": 636 }, { "epoch": 0.5314238042269188, "grad_norm": 0.3991988003253937, "learning_rate": 9.820072647699951e-06, "loss": 0.5056, "step": 637 }, { "epoch": 0.532258064516129, "grad_norm": 0.4233654737472534, "learning_rate": 9.818779107370309e-06, "loss": 0.4702, "step": 638 }, { "epoch": 0.5330923248053393, "grad_norm": 0.4387751519680023, "learning_rate": 9.817481019708292e-06, "loss": 0.4508, "step": 639 }, { "epoch": 0.5339265850945495, "grad_norm": 0.38400793075561523, "learning_rate": 9.816178385938867e-06, "loss": 0.4809, "step": 640 }, { "epoch": 0.5347608453837598, "grad_norm": 0.42588305473327637, "learning_rate": 9.81487120729129e-06, "loss": 0.4624, "step": 641 }, { "epoch": 0.53559510567297, "grad_norm": 0.39758622646331787, "learning_rate": 9.813559484999102e-06, "loss": 0.4579, "step": 642 }, { "epoch": 0.5364293659621802, "grad_norm": 0.4103184640407562, "learning_rate": 9.812243220300137e-06, "loss": 0.4668, "step": 643 }, { "epoch": 0.5372636262513905, "grad_norm": 0.38489794731140137, "learning_rate": 9.810922414436512e-06, "loss": 0.4349, "step": 644 }, { "epoch": 0.5380978865406006, "grad_norm": 0.4336555600166321, "learning_rate": 9.809597068654631e-06, "loss": 0.4744, "step": 645 }, { "epoch": 0.5389321468298109, "grad_norm": 0.43442144989967346, "learning_rate": 9.808267184205182e-06, "loss": 0.4585, "step": 646 }, { "epoch": 0.5397664071190211, "grad_norm": 0.43225643038749695, "learning_rate": 9.806932762343136e-06, "loss": 0.4617, "step": 647 }, { "epoch": 0.5406006674082313, "grad_norm": 0.5174448490142822, "learning_rate": 9.805593804327742e-06, "loss": 0.4782, "step": 648 }, { "epoch": 0.5414349276974416, "grad_norm": 0.5096933841705322, "learning_rate": 9.80425031142254e-06, "loss": 0.5028, "step": 649 }, { "epoch": 0.5422691879866518, "grad_norm": 0.562730073928833, "learning_rate": 9.802902284895333e-06, "loss": 0.4572, "step": 650 }, { "epoch": 0.5431034482758621, "grad_norm": 0.5516816973686218, "learning_rate": 9.80154972601822e-06, "loss": 0.4887, "step": 651 }, { "epoch": 0.5439377085650723, "grad_norm": 0.48134392499923706, "learning_rate": 9.800192636067567e-06, "loss": 0.4485, "step": 652 }, { "epoch": 0.5447719688542826, "grad_norm": 0.521442711353302, "learning_rate": 9.798831016324017e-06, "loss": 0.4614, "step": 653 }, { "epoch": 0.5456062291434928, "grad_norm": 0.5239782333374023, "learning_rate": 9.797464868072489e-06, "loss": 0.4762, "step": 654 }, { "epoch": 0.546440489432703, "grad_norm": 0.5498670339584351, "learning_rate": 9.796094192602171e-06, "loss": 0.4981, "step": 655 }, { "epoch": 0.5472747497219133, "grad_norm": 0.4366642236709595, "learning_rate": 9.79471899120653e-06, "loss": 0.4835, "step": 656 }, { "epoch": 0.5481090100111234, "grad_norm": 0.4363177418708801, "learning_rate": 9.793339265183303e-06, "loss": 0.4765, "step": 657 }, { "epoch": 0.5489432703003337, "grad_norm": 0.4997605085372925, "learning_rate": 9.791955015834492e-06, "loss": 0.4774, "step": 658 }, { "epoch": 0.5497775305895439, "grad_norm": 0.3919442892074585, "learning_rate": 9.790566244466371e-06, "loss": 0.4763, "step": 659 }, { "epoch": 0.5506117908787542, "grad_norm": 0.4463404715061188, "learning_rate": 9.789172952389478e-06, "loss": 0.4579, "step": 660 }, { "epoch": 0.5514460511679644, "grad_norm": 0.43534791469573975, "learning_rate": 9.787775140918625e-06, "loss": 0.4699, "step": 661 }, { "epoch": 0.5522803114571746, "grad_norm": 0.4536166191101074, "learning_rate": 9.786372811372876e-06, "loss": 0.4982, "step": 662 }, { "epoch": 0.5531145717463849, "grad_norm": 0.4977605640888214, "learning_rate": 9.784965965075572e-06, "loss": 0.5016, "step": 663 }, { "epoch": 0.5539488320355951, "grad_norm": 0.410004585981369, "learning_rate": 9.783554603354308e-06, "loss": 0.4665, "step": 664 }, { "epoch": 0.5547830923248054, "grad_norm": 0.4869717061519623, "learning_rate": 9.782138727540942e-06, "loss": 0.4846, "step": 665 }, { "epoch": 0.5556173526140156, "grad_norm": 0.40355774760246277, "learning_rate": 9.780718338971591e-06, "loss": 0.5058, "step": 666 }, { "epoch": 0.5564516129032258, "grad_norm": 0.5235663056373596, "learning_rate": 9.779293438986635e-06, "loss": 0.4947, "step": 667 }, { "epoch": 0.557285873192436, "grad_norm": 0.3886733055114746, "learning_rate": 9.777864028930705e-06, "loss": 0.4837, "step": 668 }, { "epoch": 0.5581201334816462, "grad_norm": 0.42059266567230225, "learning_rate": 9.776430110152692e-06, "loss": 0.497, "step": 669 }, { "epoch": 0.5589543937708565, "grad_norm": 0.38878536224365234, "learning_rate": 9.77499168400574e-06, "loss": 0.4805, "step": 670 }, { "epoch": 0.5597886540600667, "grad_norm": 0.3996688425540924, "learning_rate": 9.773548751847249e-06, "loss": 0.4846, "step": 671 }, { "epoch": 0.560622914349277, "grad_norm": 0.4379034638404846, "learning_rate": 9.772101315038867e-06, "loss": 0.4715, "step": 672 }, { "epoch": 0.5614571746384872, "grad_norm": 0.4174176752567291, "learning_rate": 9.770649374946495e-06, "loss": 0.4859, "step": 673 }, { "epoch": 0.5622914349276974, "grad_norm": 0.4367312788963318, "learning_rate": 9.769192932940287e-06, "loss": 0.4843, "step": 674 }, { "epoch": 0.5631256952169077, "grad_norm": 0.4626831114292145, "learning_rate": 9.767731990394638e-06, "loss": 0.4829, "step": 675 }, { "epoch": 0.5639599555061179, "grad_norm": 0.38836783170700073, "learning_rate": 9.7662665486882e-06, "loss": 0.4683, "step": 676 }, { "epoch": 0.5647942157953282, "grad_norm": 0.4422423839569092, "learning_rate": 9.764796609203857e-06, "loss": 0.4811, "step": 677 }, { "epoch": 0.5656284760845384, "grad_norm": 0.3878195881843567, "learning_rate": 9.763322173328754e-06, "loss": 0.463, "step": 678 }, { "epoch": 0.5664627363737486, "grad_norm": 0.4068148136138916, "learning_rate": 9.761843242454261e-06, "loss": 0.4832, "step": 679 }, { "epoch": 0.5672969966629589, "grad_norm": 0.40299397706985474, "learning_rate": 9.760359817976007e-06, "loss": 0.4795, "step": 680 }, { "epoch": 0.568131256952169, "grad_norm": 0.48182982206344604, "learning_rate": 9.758871901293852e-06, "loss": 0.4697, "step": 681 }, { "epoch": 0.5689655172413793, "grad_norm": 0.40189141035079956, "learning_rate": 9.757379493811892e-06, "loss": 0.4829, "step": 682 }, { "epoch": 0.5697997775305895, "grad_norm": 0.4987938702106476, "learning_rate": 9.755882596938472e-06, "loss": 0.4848, "step": 683 }, { "epoch": 0.5706340378197998, "grad_norm": 0.4073486328125, "learning_rate": 9.754381212086163e-06, "loss": 0.4424, "step": 684 }, { "epoch": 0.57146829810901, "grad_norm": 0.3822236955165863, "learning_rate": 9.752875340671779e-06, "loss": 0.4571, "step": 685 }, { "epoch": 0.5723025583982202, "grad_norm": 0.471381813287735, "learning_rate": 9.751364984116359e-06, "loss": 0.4715, "step": 686 }, { "epoch": 0.5731368186874305, "grad_norm": 0.401653915643692, "learning_rate": 9.749850143845183e-06, "loss": 0.4972, "step": 687 }, { "epoch": 0.5739710789766407, "grad_norm": 0.43858474493026733, "learning_rate": 9.748330821287758e-06, "loss": 0.4938, "step": 688 }, { "epoch": 0.574805339265851, "grad_norm": 0.39525315165519714, "learning_rate": 9.746807017877823e-06, "loss": 0.4885, "step": 689 }, { "epoch": 0.5756395995550612, "grad_norm": 0.4266471564769745, "learning_rate": 9.745278735053345e-06, "loss": 0.4664, "step": 690 }, { "epoch": 0.5764738598442715, "grad_norm": 0.4280105531215668, "learning_rate": 9.743745974256511e-06, "loss": 0.4864, "step": 691 }, { "epoch": 0.5773081201334817, "grad_norm": 0.45957306027412415, "learning_rate": 9.742208736933746e-06, "loss": 0.4954, "step": 692 }, { "epoch": 0.5781423804226918, "grad_norm": 0.4806702435016632, "learning_rate": 9.740667024535691e-06, "loss": 0.4587, "step": 693 }, { "epoch": 0.5789766407119021, "grad_norm": 0.4538249671459198, "learning_rate": 9.739120838517211e-06, "loss": 0.4453, "step": 694 }, { "epoch": 0.5798109010011123, "grad_norm": 0.5237093567848206, "learning_rate": 9.737570180337395e-06, "loss": 0.4603, "step": 695 }, { "epoch": 0.5806451612903226, "grad_norm": 0.4676782786846161, "learning_rate": 9.736015051459551e-06, "loss": 0.4725, "step": 696 }, { "epoch": 0.5814794215795328, "grad_norm": 0.48610198497772217, "learning_rate": 9.734455453351205e-06, "loss": 0.4339, "step": 697 }, { "epoch": 0.582313681868743, "grad_norm": 0.4156615734100342, "learning_rate": 9.732891387484105e-06, "loss": 0.4759, "step": 698 }, { "epoch": 0.5831479421579533, "grad_norm": 0.4630693793296814, "learning_rate": 9.731322855334207e-06, "loss": 0.4922, "step": 699 }, { "epoch": 0.5839822024471635, "grad_norm": 0.40703749656677246, "learning_rate": 9.729749858381688e-06, "loss": 0.486, "step": 700 }, { "epoch": 0.5848164627363738, "grad_norm": 0.49949175119400024, "learning_rate": 9.728172398110935e-06, "loss": 0.5125, "step": 701 }, { "epoch": 0.585650723025584, "grad_norm": 0.42794883251190186, "learning_rate": 9.72659047601055e-06, "loss": 0.4634, "step": 702 }, { "epoch": 0.5864849833147943, "grad_norm": 0.44100436568260193, "learning_rate": 9.725004093573343e-06, "loss": 0.4686, "step": 703 }, { "epoch": 0.5873192436040044, "grad_norm": 0.46410587430000305, "learning_rate": 9.723413252296335e-06, "loss": 0.4663, "step": 704 }, { "epoch": 0.5881535038932146, "grad_norm": 0.4374300241470337, "learning_rate": 9.721817953680752e-06, "loss": 0.4854, "step": 705 }, { "epoch": 0.5889877641824249, "grad_norm": 0.4106270670890808, "learning_rate": 9.72021819923203e-06, "loss": 0.495, "step": 706 }, { "epoch": 0.5898220244716351, "grad_norm": 0.44270598888397217, "learning_rate": 9.718613990459806e-06, "loss": 0.5147, "step": 707 }, { "epoch": 0.5906562847608454, "grad_norm": 0.367276132106781, "learning_rate": 9.71700532887792e-06, "loss": 0.4508, "step": 708 }, { "epoch": 0.5914905450500556, "grad_norm": 0.47131553292274475, "learning_rate": 9.715392216004421e-06, "loss": 0.4932, "step": 709 }, { "epoch": 0.5923248053392659, "grad_norm": 0.4019921123981476, "learning_rate": 9.71377465336155e-06, "loss": 0.4648, "step": 710 }, { "epoch": 0.5931590656284761, "grad_norm": 0.42643240094184875, "learning_rate": 9.712152642475751e-06, "loss": 0.4638, "step": 711 }, { "epoch": 0.5939933259176863, "grad_norm": 0.4507838189601898, "learning_rate": 9.710526184877667e-06, "loss": 0.4827, "step": 712 }, { "epoch": 0.5948275862068966, "grad_norm": 0.46035051345825195, "learning_rate": 9.708895282102135e-06, "loss": 0.475, "step": 713 }, { "epoch": 0.5956618464961068, "grad_norm": 0.41170060634613037, "learning_rate": 9.707259935688187e-06, "loss": 0.4532, "step": 714 }, { "epoch": 0.5964961067853171, "grad_norm": 0.43891629576683044, "learning_rate": 9.705620147179046e-06, "loss": 0.4446, "step": 715 }, { "epoch": 0.5973303670745272, "grad_norm": 0.41178733110427856, "learning_rate": 9.703975918122134e-06, "loss": 0.4883, "step": 716 }, { "epoch": 0.5981646273637374, "grad_norm": 0.4536864757537842, "learning_rate": 9.702327250069058e-06, "loss": 0.4764, "step": 717 }, { "epoch": 0.5989988876529477, "grad_norm": 0.39851391315460205, "learning_rate": 9.700674144575614e-06, "loss": 0.467, "step": 718 }, { "epoch": 0.5998331479421579, "grad_norm": 0.46597108244895935, "learning_rate": 9.699016603201784e-06, "loss": 0.4741, "step": 719 }, { "epoch": 0.6006674082313682, "grad_norm": 0.4034475088119507, "learning_rate": 9.697354627511744e-06, "loss": 0.4351, "step": 720 }, { "epoch": 0.6015016685205784, "grad_norm": 0.40976834297180176, "learning_rate": 9.695688219073845e-06, "loss": 0.4929, "step": 721 }, { "epoch": 0.6023359288097887, "grad_norm": 0.42263010144233704, "learning_rate": 9.694017379460624e-06, "loss": 0.4792, "step": 722 }, { "epoch": 0.6031701890989989, "grad_norm": 0.44653940200805664, "learning_rate": 9.692342110248802e-06, "loss": 0.4565, "step": 723 }, { "epoch": 0.6040044493882091, "grad_norm": 0.3877444565296173, "learning_rate": 9.69066241301928e-06, "loss": 0.4629, "step": 724 }, { "epoch": 0.6048387096774194, "grad_norm": 0.3971206545829773, "learning_rate": 9.688978289357136e-06, "loss": 0.4599, "step": 725 }, { "epoch": 0.6056729699666296, "grad_norm": 0.4209384620189667, "learning_rate": 9.687289740851622e-06, "loss": 0.4733, "step": 726 }, { "epoch": 0.6065072302558399, "grad_norm": 0.41906723380088806, "learning_rate": 9.685596769096173e-06, "loss": 0.4525, "step": 727 }, { "epoch": 0.60734149054505, "grad_norm": 0.5063453316688538, "learning_rate": 9.68389937568839e-06, "loss": 0.4636, "step": 728 }, { "epoch": 0.6081757508342603, "grad_norm": 0.42767319083213806, "learning_rate": 9.682197562230054e-06, "loss": 0.481, "step": 729 }, { "epoch": 0.6090100111234705, "grad_norm": 0.49501386284828186, "learning_rate": 9.68049133032711e-06, "loss": 0.4739, "step": 730 }, { "epoch": 0.6098442714126807, "grad_norm": 0.45667776465415955, "learning_rate": 9.67878068158968e-06, "loss": 0.4614, "step": 731 }, { "epoch": 0.610678531701891, "grad_norm": 0.41828709840774536, "learning_rate": 9.67706561763205e-06, "loss": 0.4844, "step": 732 }, { "epoch": 0.6115127919911012, "grad_norm": 0.4679460823535919, "learning_rate": 9.675346140072667e-06, "loss": 0.4877, "step": 733 }, { "epoch": 0.6123470522803115, "grad_norm": 0.4925785958766937, "learning_rate": 9.673622250534155e-06, "loss": 0.4703, "step": 734 }, { "epoch": 0.6131813125695217, "grad_norm": 0.3913607895374298, "learning_rate": 9.671893950643294e-06, "loss": 0.4839, "step": 735 }, { "epoch": 0.6140155728587319, "grad_norm": 0.4483281075954437, "learning_rate": 9.670161242031024e-06, "loss": 0.473, "step": 736 }, { "epoch": 0.6148498331479422, "grad_norm": 0.4119316041469574, "learning_rate": 9.66842412633245e-06, "loss": 0.5065, "step": 737 }, { "epoch": 0.6156840934371524, "grad_norm": 0.4623095691204071, "learning_rate": 9.666682605186834e-06, "loss": 0.4813, "step": 738 }, { "epoch": 0.6165183537263627, "grad_norm": 0.352371484041214, "learning_rate": 9.664936680237597e-06, "loss": 0.4493, "step": 739 }, { "epoch": 0.6173526140155728, "grad_norm": 0.4493887424468994, "learning_rate": 9.663186353132312e-06, "loss": 0.4714, "step": 740 }, { "epoch": 0.6181868743047831, "grad_norm": 0.42149147391319275, "learning_rate": 9.66143162552271e-06, "loss": 0.467, "step": 741 }, { "epoch": 0.6190211345939933, "grad_norm": 0.45152875781059265, "learning_rate": 9.659672499064672e-06, "loss": 0.4625, "step": 742 }, { "epoch": 0.6198553948832035, "grad_norm": 0.3965662121772766, "learning_rate": 9.657908975418232e-06, "loss": 0.4684, "step": 743 }, { "epoch": 0.6206896551724138, "grad_norm": 0.4364754259586334, "learning_rate": 9.656141056247573e-06, "loss": 0.4857, "step": 744 }, { "epoch": 0.621523915461624, "grad_norm": 0.42279234528541565, "learning_rate": 9.654368743221022e-06, "loss": 0.4448, "step": 745 }, { "epoch": 0.6223581757508343, "grad_norm": 0.42457517981529236, "learning_rate": 9.652592038011059e-06, "loss": 0.489, "step": 746 }, { "epoch": 0.6231924360400445, "grad_norm": 0.42345842719078064, "learning_rate": 9.650810942294306e-06, "loss": 0.5197, "step": 747 }, { "epoch": 0.6240266963292548, "grad_norm": 0.415723592042923, "learning_rate": 9.649025457751526e-06, "loss": 0.4471, "step": 748 }, { "epoch": 0.624860956618465, "grad_norm": 0.3800181448459625, "learning_rate": 9.647235586067628e-06, "loss": 0.4624, "step": 749 }, { "epoch": 0.6256952169076752, "grad_norm": 0.3449118733406067, "learning_rate": 9.645441328931655e-06, "loss": 0.4679, "step": 750 }, { "epoch": 0.6265294771968855, "grad_norm": 0.3821072578430176, "learning_rate": 9.643642688036793e-06, "loss": 0.4918, "step": 751 }, { "epoch": 0.6273637374860956, "grad_norm": 0.39983755350112915, "learning_rate": 9.641839665080363e-06, "loss": 0.4622, "step": 752 }, { "epoch": 0.6281979977753059, "grad_norm": 0.3896723687648773, "learning_rate": 9.640032261763825e-06, "loss": 0.4972, "step": 753 }, { "epoch": 0.6290322580645161, "grad_norm": 0.3699634373188019, "learning_rate": 9.638220479792767e-06, "loss": 0.4565, "step": 754 }, { "epoch": 0.6298665183537263, "grad_norm": 0.3764488995075226, "learning_rate": 9.636404320876911e-06, "loss": 0.4742, "step": 755 }, { "epoch": 0.6307007786429366, "grad_norm": 0.43458330631256104, "learning_rate": 9.63458378673011e-06, "loss": 0.4423, "step": 756 }, { "epoch": 0.6315350389321468, "grad_norm": 0.41390514373779297, "learning_rate": 9.632758879070346e-06, "loss": 0.4551, "step": 757 }, { "epoch": 0.6323692992213571, "grad_norm": 0.44622522592544556, "learning_rate": 9.63092959961973e-06, "loss": 0.5088, "step": 758 }, { "epoch": 0.6332035595105673, "grad_norm": 0.41718927025794983, "learning_rate": 9.62909595010449e-06, "loss": 0.4823, "step": 759 }, { "epoch": 0.6340378197997776, "grad_norm": 0.41046199202537537, "learning_rate": 9.627257932254991e-06, "loss": 0.4816, "step": 760 }, { "epoch": 0.6348720800889878, "grad_norm": 0.43687310814857483, "learning_rate": 9.625415547805711e-06, "loss": 0.4612, "step": 761 }, { "epoch": 0.635706340378198, "grad_norm": 0.40793150663375854, "learning_rate": 9.62356879849525e-06, "loss": 0.47, "step": 762 }, { "epoch": 0.6365406006674083, "grad_norm": 0.38553401827812195, "learning_rate": 9.621717686066331e-06, "loss": 0.4573, "step": 763 }, { "epoch": 0.6373748609566184, "grad_norm": 0.4354100227355957, "learning_rate": 9.619862212265788e-06, "loss": 0.4706, "step": 764 }, { "epoch": 0.6382091212458287, "grad_norm": 0.45290783047676086, "learning_rate": 9.618002378844575e-06, "loss": 0.4448, "step": 765 }, { "epoch": 0.6390433815350389, "grad_norm": 0.4045596122741699, "learning_rate": 9.616138187557758e-06, "loss": 0.4649, "step": 766 }, { "epoch": 0.6398776418242491, "grad_norm": 0.4876764118671417, "learning_rate": 9.61426964016452e-06, "loss": 0.4605, "step": 767 }, { "epoch": 0.6407119021134594, "grad_norm": 0.43785810470581055, "learning_rate": 9.612396738428147e-06, "loss": 0.4858, "step": 768 }, { "epoch": 0.6415461624026696, "grad_norm": 0.39098039269447327, "learning_rate": 9.610519484116042e-06, "loss": 0.4648, "step": 769 }, { "epoch": 0.6423804226918799, "grad_norm": 0.5116364359855652, "learning_rate": 9.608637878999709e-06, "loss": 0.4346, "step": 770 }, { "epoch": 0.6432146829810901, "grad_norm": 0.4077700674533844, "learning_rate": 9.606751924854762e-06, "loss": 0.456, "step": 771 }, { "epoch": 0.6440489432703004, "grad_norm": 0.4522409737110138, "learning_rate": 9.604861623460918e-06, "loss": 0.4817, "step": 772 }, { "epoch": 0.6448832035595106, "grad_norm": 0.4127247929573059, "learning_rate": 9.602966976601995e-06, "loss": 0.4854, "step": 773 }, { "epoch": 0.6457174638487208, "grad_norm": 0.40627750754356384, "learning_rate": 9.60106798606591e-06, "loss": 0.4444, "step": 774 }, { "epoch": 0.646551724137931, "grad_norm": 0.41040539741516113, "learning_rate": 9.599164653644684e-06, "loss": 0.4779, "step": 775 }, { "epoch": 0.6473859844271412, "grad_norm": 0.41064515709877014, "learning_rate": 9.597256981134433e-06, "loss": 0.5026, "step": 776 }, { "epoch": 0.6482202447163515, "grad_norm": 0.4073631465435028, "learning_rate": 9.595344970335367e-06, "loss": 0.4714, "step": 777 }, { "epoch": 0.6490545050055617, "grad_norm": 0.41629865765571594, "learning_rate": 9.593428623051793e-06, "loss": 0.4732, "step": 778 }, { "epoch": 0.649888765294772, "grad_norm": 0.3929150104522705, "learning_rate": 9.591507941092106e-06, "loss": 0.4481, "step": 779 }, { "epoch": 0.6507230255839822, "grad_norm": 0.4257237911224365, "learning_rate": 9.589582926268798e-06, "loss": 0.4414, "step": 780 }, { "epoch": 0.6515572858731924, "grad_norm": 0.5401448011398315, "learning_rate": 9.587653580398442e-06, "loss": 0.4738, "step": 781 }, { "epoch": 0.6523915461624027, "grad_norm": 0.4575658142566681, "learning_rate": 9.585719905301703e-06, "loss": 0.4879, "step": 782 }, { "epoch": 0.6532258064516129, "grad_norm": 0.37722688913345337, "learning_rate": 9.583781902803329e-06, "loss": 0.463, "step": 783 }, { "epoch": 0.6540600667408232, "grad_norm": 0.4146757423877716, "learning_rate": 9.581839574732157e-06, "loss": 0.4451, "step": 784 }, { "epoch": 0.6548943270300334, "grad_norm": 0.4410708248615265, "learning_rate": 9.579892922921097e-06, "loss": 0.5062, "step": 785 }, { "epoch": 0.6557285873192435, "grad_norm": 0.4299607574939728, "learning_rate": 9.577941949207146e-06, "loss": 0.4896, "step": 786 }, { "epoch": 0.6565628476084538, "grad_norm": 0.37746211886405945, "learning_rate": 9.575986655431377e-06, "loss": 0.4865, "step": 787 }, { "epoch": 0.657397107897664, "grad_norm": 0.39992085099220276, "learning_rate": 9.574027043438942e-06, "loss": 0.4983, "step": 788 }, { "epoch": 0.6582313681868743, "grad_norm": 0.4093070328235626, "learning_rate": 9.572063115079063e-06, "loss": 0.4638, "step": 789 }, { "epoch": 0.6590656284760845, "grad_norm": 0.38917234539985657, "learning_rate": 9.57009487220504e-06, "loss": 0.4637, "step": 790 }, { "epoch": 0.6598998887652948, "grad_norm": 0.40783998370170593, "learning_rate": 9.568122316674242e-06, "loss": 0.4815, "step": 791 }, { "epoch": 0.660734149054505, "grad_norm": 0.43162935972213745, "learning_rate": 9.56614545034811e-06, "loss": 0.4865, "step": 792 }, { "epoch": 0.6615684093437152, "grad_norm": 0.37050920724868774, "learning_rate": 9.56416427509215e-06, "loss": 0.432, "step": 793 }, { "epoch": 0.6624026696329255, "grad_norm": 0.4352168142795563, "learning_rate": 9.562178792775936e-06, "loss": 0.4651, "step": 794 }, { "epoch": 0.6632369299221357, "grad_norm": 0.3818225860595703, "learning_rate": 9.560189005273106e-06, "loss": 0.4989, "step": 795 }, { "epoch": 0.664071190211346, "grad_norm": 0.4534338116645813, "learning_rate": 9.558194914461359e-06, "loss": 0.4726, "step": 796 }, { "epoch": 0.6649054505005562, "grad_norm": 0.41872018575668335, "learning_rate": 9.55619652222246e-06, "loss": 0.4987, "step": 797 }, { "epoch": 0.6657397107897665, "grad_norm": 0.445915162563324, "learning_rate": 9.554193830442228e-06, "loss": 0.5208, "step": 798 }, { "epoch": 0.6665739710789766, "grad_norm": 0.4388257563114166, "learning_rate": 9.552186841010543e-06, "loss": 0.451, "step": 799 }, { "epoch": 0.6674082313681868, "grad_norm": 0.41086870431900024, "learning_rate": 9.550175555821333e-06, "loss": 0.4715, "step": 800 }, { "epoch": 0.6682424916573971, "grad_norm": 0.5134502649307251, "learning_rate": 9.548159976772593e-06, "loss": 0.4475, "step": 801 }, { "epoch": 0.6690767519466073, "grad_norm": 0.4297252595424652, "learning_rate": 9.546140105766357e-06, "loss": 0.4704, "step": 802 }, { "epoch": 0.6699110122358176, "grad_norm": 0.4295264482498169, "learning_rate": 9.544115944708718e-06, "loss": 0.4412, "step": 803 }, { "epoch": 0.6707452725250278, "grad_norm": 0.4816243350505829, "learning_rate": 9.542087495509811e-06, "loss": 0.4555, "step": 804 }, { "epoch": 0.671579532814238, "grad_norm": 0.48643356561660767, "learning_rate": 9.540054760083822e-06, "loss": 0.477, "step": 805 }, { "epoch": 0.6724137931034483, "grad_norm": 0.4277946650981903, "learning_rate": 9.538017740348982e-06, "loss": 0.4639, "step": 806 }, { "epoch": 0.6732480533926585, "grad_norm": 0.41716429591178894, "learning_rate": 9.535976438227561e-06, "loss": 0.4634, "step": 807 }, { "epoch": 0.6740823136818688, "grad_norm": 0.4422438144683838, "learning_rate": 9.533930855645872e-06, "loss": 0.4905, "step": 808 }, { "epoch": 0.674916573971079, "grad_norm": 0.4607492983341217, "learning_rate": 9.53188099453427e-06, "loss": 0.4809, "step": 809 }, { "epoch": 0.6757508342602893, "grad_norm": 0.39637550711631775, "learning_rate": 9.529826856827144e-06, "loss": 0.4699, "step": 810 }, { "epoch": 0.6765850945494994, "grad_norm": 0.46183499693870544, "learning_rate": 9.527768444462922e-06, "loss": 0.4724, "step": 811 }, { "epoch": 0.6774193548387096, "grad_norm": 0.4133055508136749, "learning_rate": 9.525705759384063e-06, "loss": 0.4807, "step": 812 }, { "epoch": 0.6782536151279199, "grad_norm": 0.41410908102989197, "learning_rate": 9.523638803537057e-06, "loss": 0.4675, "step": 813 }, { "epoch": 0.6790878754171301, "grad_norm": 0.42624562978744507, "learning_rate": 9.521567578872427e-06, "loss": 0.4514, "step": 814 }, { "epoch": 0.6799221357063404, "grad_norm": 0.48528289794921875, "learning_rate": 9.519492087344724e-06, "loss": 0.4874, "step": 815 }, { "epoch": 0.6807563959955506, "grad_norm": 0.4477766156196594, "learning_rate": 9.517412330912524e-06, "loss": 0.4655, "step": 816 }, { "epoch": 0.6815906562847609, "grad_norm": 0.4273214638233185, "learning_rate": 9.51532831153843e-06, "loss": 0.4662, "step": 817 }, { "epoch": 0.6824249165739711, "grad_norm": 0.42267170548439026, "learning_rate": 9.513240031189068e-06, "loss": 0.4564, "step": 818 }, { "epoch": 0.6832591768631813, "grad_norm": 0.4416694939136505, "learning_rate": 9.51114749183508e-06, "loss": 0.4751, "step": 819 }, { "epoch": 0.6840934371523916, "grad_norm": 0.3952409327030182, "learning_rate": 9.509050695451131e-06, "loss": 0.4731, "step": 820 }, { "epoch": 0.6849276974416018, "grad_norm": 0.4402921795845032, "learning_rate": 9.506949644015905e-06, "loss": 0.4872, "step": 821 }, { "epoch": 0.685761957730812, "grad_norm": 0.4744360148906708, "learning_rate": 9.504844339512096e-06, "loss": 0.4964, "step": 822 }, { "epoch": 0.6865962180200222, "grad_norm": 0.4025762975215912, "learning_rate": 9.502734783926417e-06, "loss": 0.5127, "step": 823 }, { "epoch": 0.6874304783092324, "grad_norm": 0.45970141887664795, "learning_rate": 9.50062097924959e-06, "loss": 0.483, "step": 824 }, { "epoch": 0.6882647385984427, "grad_norm": 0.4756225347518921, "learning_rate": 9.498502927476345e-06, "loss": 0.4887, "step": 825 }, { "epoch": 0.6890989988876529, "grad_norm": 0.4579399526119232, "learning_rate": 9.496380630605424e-06, "loss": 0.4792, "step": 826 }, { "epoch": 0.6899332591768632, "grad_norm": 0.45419925451278687, "learning_rate": 9.494254090639572e-06, "loss": 0.4726, "step": 827 }, { "epoch": 0.6907675194660734, "grad_norm": 0.48215973377227783, "learning_rate": 9.492123309585538e-06, "loss": 0.4678, "step": 828 }, { "epoch": 0.6916017797552837, "grad_norm": 0.40905070304870605, "learning_rate": 9.489988289454073e-06, "loss": 0.4523, "step": 829 }, { "epoch": 0.6924360400444939, "grad_norm": 0.4414481222629547, "learning_rate": 9.487849032259932e-06, "loss": 0.5107, "step": 830 }, { "epoch": 0.6932703003337041, "grad_norm": 0.4442170262336731, "learning_rate": 9.485705540021866e-06, "loss": 0.4754, "step": 831 }, { "epoch": 0.6941045606229144, "grad_norm": 0.36192017793655396, "learning_rate": 9.483557814762619e-06, "loss": 0.4318, "step": 832 }, { "epoch": 0.6949388209121246, "grad_norm": 0.405210018157959, "learning_rate": 9.481405858508935e-06, "loss": 0.463, "step": 833 }, { "epoch": 0.6957730812013349, "grad_norm": 0.43775230646133423, "learning_rate": 9.479249673291546e-06, "loss": 0.4794, "step": 834 }, { "epoch": 0.696607341490545, "grad_norm": 0.38711220026016235, "learning_rate": 9.47708926114518e-06, "loss": 0.4569, "step": 835 }, { "epoch": 0.6974416017797553, "grad_norm": 0.46561262011528015, "learning_rate": 9.47492462410855e-06, "loss": 0.4587, "step": 836 }, { "epoch": 0.6982758620689655, "grad_norm": 0.41399070620536804, "learning_rate": 9.472755764224357e-06, "loss": 0.4536, "step": 837 }, { "epoch": 0.6991101223581757, "grad_norm": 0.42120206356048584, "learning_rate": 9.470582683539284e-06, "loss": 0.4613, "step": 838 }, { "epoch": 0.699944382647386, "grad_norm": 0.4780728816986084, "learning_rate": 9.468405384104005e-06, "loss": 0.4699, "step": 839 }, { "epoch": 0.7007786429365962, "grad_norm": 0.44980886578559875, "learning_rate": 9.466223867973166e-06, "loss": 0.4625, "step": 840 }, { "epoch": 0.7016129032258065, "grad_norm": 0.46513211727142334, "learning_rate": 9.464038137205395e-06, "loss": 0.4515, "step": 841 }, { "epoch": 0.7024471635150167, "grad_norm": 0.47072553634643555, "learning_rate": 9.461848193863305e-06, "loss": 0.4641, "step": 842 }, { "epoch": 0.7032814238042269, "grad_norm": 0.5269559621810913, "learning_rate": 9.45965404001347e-06, "loss": 0.5073, "step": 843 }, { "epoch": 0.7041156840934372, "grad_norm": 0.4510970711708069, "learning_rate": 9.457455677726447e-06, "loss": 0.4701, "step": 844 }, { "epoch": 0.7049499443826474, "grad_norm": 0.46201252937316895, "learning_rate": 9.455253109076767e-06, "loss": 0.4695, "step": 845 }, { "epoch": 0.7057842046718577, "grad_norm": 0.4317589998245239, "learning_rate": 9.453046336142921e-06, "loss": 0.4744, "step": 846 }, { "epoch": 0.7066184649610678, "grad_norm": 0.44674286246299744, "learning_rate": 9.450835361007372e-06, "loss": 0.4495, "step": 847 }, { "epoch": 0.7074527252502781, "grad_norm": 0.4899110794067383, "learning_rate": 9.448620185756549e-06, "loss": 0.4705, "step": 848 }, { "epoch": 0.7082869855394883, "grad_norm": 0.47531142830848694, "learning_rate": 9.446400812480847e-06, "loss": 0.4738, "step": 849 }, { "epoch": 0.7091212458286985, "grad_norm": 0.45990225672721863, "learning_rate": 9.444177243274619e-06, "loss": 0.4494, "step": 850 }, { "epoch": 0.7099555061179088, "grad_norm": 0.42081865668296814, "learning_rate": 9.441949480236174e-06, "loss": 0.4435, "step": 851 }, { "epoch": 0.710789766407119, "grad_norm": 0.45366358757019043, "learning_rate": 9.439717525467789e-06, "loss": 0.4687, "step": 852 }, { "epoch": 0.7116240266963293, "grad_norm": 0.4973547160625458, "learning_rate": 9.437481381075685e-06, "loss": 0.4791, "step": 853 }, { "epoch": 0.7124582869855395, "grad_norm": 0.4957771301269531, "learning_rate": 9.435241049170045e-06, "loss": 0.4934, "step": 854 }, { "epoch": 0.7132925472747497, "grad_norm": 0.43608108162879944, "learning_rate": 9.432996531865001e-06, "loss": 0.4747, "step": 855 }, { "epoch": 0.71412680756396, "grad_norm": 0.49690860509872437, "learning_rate": 9.430747831278632e-06, "loss": 0.489, "step": 856 }, { "epoch": 0.7149610678531702, "grad_norm": 0.41977497935295105, "learning_rate": 9.428494949532972e-06, "loss": 0.4607, "step": 857 }, { "epoch": 0.7157953281423804, "grad_norm": 0.4274574816226959, "learning_rate": 9.426237888753991e-06, "loss": 0.4718, "step": 858 }, { "epoch": 0.7166295884315906, "grad_norm": 0.4147649109363556, "learning_rate": 9.423976651071606e-06, "loss": 0.4813, "step": 859 }, { "epoch": 0.7174638487208009, "grad_norm": 0.4454214870929718, "learning_rate": 9.421711238619681e-06, "loss": 0.4589, "step": 860 }, { "epoch": 0.7182981090100111, "grad_norm": 0.4262605607509613, "learning_rate": 9.419441653536014e-06, "loss": 0.4344, "step": 861 }, { "epoch": 0.7191323692992213, "grad_norm": 0.48590415716171265, "learning_rate": 9.417167897962343e-06, "loss": 0.4684, "step": 862 }, { "epoch": 0.7199666295884316, "grad_norm": 0.48209530115127563, "learning_rate": 9.414889974044338e-06, "loss": 0.4817, "step": 863 }, { "epoch": 0.7208008898776418, "grad_norm": 0.4561712443828583, "learning_rate": 9.412607883931608e-06, "loss": 0.4571, "step": 864 }, { "epoch": 0.7216351501668521, "grad_norm": 0.45505762100219727, "learning_rate": 9.410321629777688e-06, "loss": 0.4472, "step": 865 }, { "epoch": 0.7224694104560623, "grad_norm": 0.4709140360355377, "learning_rate": 9.408031213740045e-06, "loss": 0.4691, "step": 866 }, { "epoch": 0.7233036707452726, "grad_norm": 0.5127201676368713, "learning_rate": 9.405736637980077e-06, "loss": 0.4504, "step": 867 }, { "epoch": 0.7241379310344828, "grad_norm": 0.3673347532749176, "learning_rate": 9.403437904663102e-06, "loss": 0.4331, "step": 868 }, { "epoch": 0.724972191323693, "grad_norm": 0.5430071353912354, "learning_rate": 9.401135015958363e-06, "loss": 0.458, "step": 869 }, { "epoch": 0.7258064516129032, "grad_norm": 0.4225972890853882, "learning_rate": 9.398827974039025e-06, "loss": 0.4308, "step": 870 }, { "epoch": 0.7266407119021134, "grad_norm": 0.4765642285346985, "learning_rate": 9.396516781082172e-06, "loss": 0.4521, "step": 871 }, { "epoch": 0.7274749721913237, "grad_norm": 0.46362507343292236, "learning_rate": 9.394201439268805e-06, "loss": 0.4724, "step": 872 }, { "epoch": 0.7283092324805339, "grad_norm": 0.4646399915218353, "learning_rate": 9.391881950783838e-06, "loss": 0.453, "step": 873 }, { "epoch": 0.7291434927697441, "grad_norm": 0.46840402483940125, "learning_rate": 9.389558317816105e-06, "loss": 0.4562, "step": 874 }, { "epoch": 0.7299777530589544, "grad_norm": 0.48050495982170105, "learning_rate": 9.387230542558343e-06, "loss": 0.4695, "step": 875 }, { "epoch": 0.7308120133481646, "grad_norm": 0.435786634683609, "learning_rate": 9.384898627207203e-06, "loss": 0.4697, "step": 876 }, { "epoch": 0.7316462736373749, "grad_norm": 0.41830170154571533, "learning_rate": 9.382562573963238e-06, "loss": 0.4309, "step": 877 }, { "epoch": 0.7324805339265851, "grad_norm": 0.43040770292282104, "learning_rate": 9.380222385030916e-06, "loss": 0.4834, "step": 878 }, { "epoch": 0.7333147942157954, "grad_norm": 0.4678661823272705, "learning_rate": 9.377878062618591e-06, "loss": 0.4957, "step": 879 }, { "epoch": 0.7341490545050056, "grad_norm": 0.42694494128227234, "learning_rate": 9.375529608938535e-06, "loss": 0.464, "step": 880 }, { "epoch": 0.7349833147942157, "grad_norm": 0.4166353642940521, "learning_rate": 9.373177026206907e-06, "loss": 0.4655, "step": 881 }, { "epoch": 0.735817575083426, "grad_norm": 0.4539768397808075, "learning_rate": 9.37082031664377e-06, "loss": 0.4654, "step": 882 }, { "epoch": 0.7366518353726362, "grad_norm": 0.42214322090148926, "learning_rate": 9.368459482473073e-06, "loss": 0.4677, "step": 883 }, { "epoch": 0.7374860956618465, "grad_norm": 0.48180848360061646, "learning_rate": 9.366094525922664e-06, "loss": 0.4873, "step": 884 }, { "epoch": 0.7383203559510567, "grad_norm": 0.42995765805244446, "learning_rate": 9.363725449224281e-06, "loss": 0.4736, "step": 885 }, { "epoch": 0.739154616240267, "grad_norm": 0.49140745401382446, "learning_rate": 9.361352254613547e-06, "loss": 0.4721, "step": 886 }, { "epoch": 0.7399888765294772, "grad_norm": 0.4049518406391144, "learning_rate": 9.35897494432997e-06, "loss": 0.4496, "step": 887 }, { "epoch": 0.7408231368186874, "grad_norm": 0.3842932879924774, "learning_rate": 9.356593520616948e-06, "loss": 0.4596, "step": 888 }, { "epoch": 0.7416573971078977, "grad_norm": 0.4287854731082916, "learning_rate": 9.354207985721753e-06, "loss": 0.4335, "step": 889 }, { "epoch": 0.7424916573971079, "grad_norm": 0.3899165391921997, "learning_rate": 9.351818341895542e-06, "loss": 0.4797, "step": 890 }, { "epoch": 0.7433259176863182, "grad_norm": 0.47397205233573914, "learning_rate": 9.349424591393348e-06, "loss": 0.4851, "step": 891 }, { "epoch": 0.7441601779755284, "grad_norm": 0.43816956877708435, "learning_rate": 9.347026736474077e-06, "loss": 0.423, "step": 892 }, { "epoch": 0.7449944382647385, "grad_norm": 0.4408695101737976, "learning_rate": 9.344624779400513e-06, "loss": 0.475, "step": 893 }, { "epoch": 0.7458286985539488, "grad_norm": 0.40708911418914795, "learning_rate": 9.34221872243931e-06, "loss": 0.4819, "step": 894 }, { "epoch": 0.746662958843159, "grad_norm": 0.43900659680366516, "learning_rate": 9.339808567860981e-06, "loss": 0.4693, "step": 895 }, { "epoch": 0.7474972191323693, "grad_norm": 0.4047435224056244, "learning_rate": 9.337394317939924e-06, "loss": 0.4488, "step": 896 }, { "epoch": 0.7483314794215795, "grad_norm": 0.3774779438972473, "learning_rate": 9.334975974954385e-06, "loss": 0.4873, "step": 897 }, { "epoch": 0.7491657397107898, "grad_norm": 0.40400227904319763, "learning_rate": 9.332553541186485e-06, "loss": 0.473, "step": 898 }, { "epoch": 0.75, "grad_norm": 0.35029685497283936, "learning_rate": 9.330127018922195e-06, "loss": 0.4414, "step": 899 }, { "epoch": 0.7508342602892102, "grad_norm": 0.4203338325023651, "learning_rate": 9.327696410451347e-06, "loss": 0.442, "step": 900 }, { "epoch": 0.7516685205784205, "grad_norm": 0.37832021713256836, "learning_rate": 9.325261718067638e-06, "loss": 0.4538, "step": 901 }, { "epoch": 0.7525027808676307, "grad_norm": 0.3839128315448761, "learning_rate": 9.322822944068606e-06, "loss": 0.4871, "step": 902 }, { "epoch": 0.753337041156841, "grad_norm": 0.4096129536628723, "learning_rate": 9.32038009075565e-06, "loss": 0.4667, "step": 903 }, { "epoch": 0.7541713014460512, "grad_norm": 0.36271363496780396, "learning_rate": 9.317933160434011e-06, "loss": 0.4859, "step": 904 }, { "epoch": 0.7550055617352615, "grad_norm": 0.3860838711261749, "learning_rate": 9.315482155412783e-06, "loss": 0.4396, "step": 905 }, { "epoch": 0.7558398220244716, "grad_norm": 0.3807103633880615, "learning_rate": 9.313027078004903e-06, "loss": 0.4826, "step": 906 }, { "epoch": 0.7566740823136818, "grad_norm": 0.392043799161911, "learning_rate": 9.310567930527155e-06, "loss": 0.4986, "step": 907 }, { "epoch": 0.7575083426028921, "grad_norm": 0.4067477881908417, "learning_rate": 9.308104715300155e-06, "loss": 0.4766, "step": 908 }, { "epoch": 0.7583426028921023, "grad_norm": 0.4207528531551361, "learning_rate": 9.305637434648364e-06, "loss": 0.4807, "step": 909 }, { "epoch": 0.7591768631813126, "grad_norm": 0.4014434814453125, "learning_rate": 9.303166090900082e-06, "loss": 0.4855, "step": 910 }, { "epoch": 0.7600111234705228, "grad_norm": 0.39711910486221313, "learning_rate": 9.300690686387435e-06, "loss": 0.5071, "step": 911 }, { "epoch": 0.760845383759733, "grad_norm": 0.3941214680671692, "learning_rate": 9.298211223446387e-06, "loss": 0.481, "step": 912 }, { "epoch": 0.7616796440489433, "grad_norm": 0.4359917938709259, "learning_rate": 9.295727704416731e-06, "loss": 0.4699, "step": 913 }, { "epoch": 0.7625139043381535, "grad_norm": 0.3672400414943695, "learning_rate": 9.293240131642087e-06, "loss": 0.4748, "step": 914 }, { "epoch": 0.7633481646273638, "grad_norm": 0.43665552139282227, "learning_rate": 9.290748507469897e-06, "loss": 0.4653, "step": 915 }, { "epoch": 0.764182424916574, "grad_norm": 0.3988327383995056, "learning_rate": 9.288252834251436e-06, "loss": 0.459, "step": 916 }, { "epoch": 0.7650166852057843, "grad_norm": 0.4199771285057068, "learning_rate": 9.285753114341787e-06, "loss": 0.468, "step": 917 }, { "epoch": 0.7658509454949944, "grad_norm": 0.4745626747608185, "learning_rate": 9.28324935009986e-06, "loss": 0.4537, "step": 918 }, { "epoch": 0.7666852057842046, "grad_norm": 0.3815878927707672, "learning_rate": 9.280741543888381e-06, "loss": 0.4533, "step": 919 }, { "epoch": 0.7675194660734149, "grad_norm": 0.4727926552295685, "learning_rate": 9.278229698073889e-06, "loss": 0.4519, "step": 920 }, { "epoch": 0.7683537263626251, "grad_norm": 0.4001463055610657, "learning_rate": 9.275713815026732e-06, "loss": 0.456, "step": 921 }, { "epoch": 0.7691879866518354, "grad_norm": 0.41719651222229004, "learning_rate": 9.273193897121074e-06, "loss": 0.4563, "step": 922 }, { "epoch": 0.7700222469410456, "grad_norm": 0.43005451560020447, "learning_rate": 9.27066994673488e-06, "loss": 0.4935, "step": 923 }, { "epoch": 0.7708565072302559, "grad_norm": 0.388235867023468, "learning_rate": 9.26814196624993e-06, "loss": 0.4694, "step": 924 }, { "epoch": 0.7716907675194661, "grad_norm": 0.42255866527557373, "learning_rate": 9.265609958051795e-06, "loss": 0.4712, "step": 925 }, { "epoch": 0.7725250278086763, "grad_norm": 0.4505786895751953, "learning_rate": 9.263073924529852e-06, "loss": 0.4753, "step": 926 }, { "epoch": 0.7733592880978866, "grad_norm": 0.4395304322242737, "learning_rate": 9.260533868077283e-06, "loss": 0.5047, "step": 927 }, { "epoch": 0.7741935483870968, "grad_norm": 0.42204567790031433, "learning_rate": 9.257989791091056e-06, "loss": 0.4671, "step": 928 }, { "epoch": 0.775027808676307, "grad_norm": 0.43199580907821655, "learning_rate": 9.255441695971938e-06, "loss": 0.4669, "step": 929 }, { "epoch": 0.7758620689655172, "grad_norm": 0.3820450007915497, "learning_rate": 9.252889585124489e-06, "loss": 0.434, "step": 930 }, { "epoch": 0.7766963292547274, "grad_norm": 0.47433188557624817, "learning_rate": 9.250333460957053e-06, "loss": 0.4558, "step": 931 }, { "epoch": 0.7775305895439377, "grad_norm": 0.46394258737564087, "learning_rate": 9.24777332588177e-06, "loss": 0.4254, "step": 932 }, { "epoch": 0.7783648498331479, "grad_norm": 0.3935208320617676, "learning_rate": 9.245209182314557e-06, "loss": 0.4619, "step": 933 }, { "epoch": 0.7791991101223582, "grad_norm": 0.43525421619415283, "learning_rate": 9.242641032675118e-06, "loss": 0.4686, "step": 934 }, { "epoch": 0.7800333704115684, "grad_norm": 0.3594815731048584, "learning_rate": 9.240068879386937e-06, "loss": 0.4589, "step": 935 }, { "epoch": 0.7808676307007787, "grad_norm": 0.407733678817749, "learning_rate": 9.237492724877272e-06, "loss": 0.4646, "step": 936 }, { "epoch": 0.7817018909899889, "grad_norm": 0.429595410823822, "learning_rate": 9.234912571577162e-06, "loss": 0.4816, "step": 937 }, { "epoch": 0.7825361512791991, "grad_norm": 0.4097098410129547, "learning_rate": 9.232328421921419e-06, "loss": 0.4526, "step": 938 }, { "epoch": 0.7833704115684094, "grad_norm": 0.4656343162059784, "learning_rate": 9.229740278348622e-06, "loss": 0.478, "step": 939 }, { "epoch": 0.7842046718576196, "grad_norm": 0.4694792330265045, "learning_rate": 9.227148143301125e-06, "loss": 0.4805, "step": 940 }, { "epoch": 0.7850389321468298, "grad_norm": 0.3811095654964447, "learning_rate": 9.224552019225044e-06, "loss": 0.4291, "step": 941 }, { "epoch": 0.78587319243604, "grad_norm": 0.464913934469223, "learning_rate": 9.221951908570259e-06, "loss": 0.4304, "step": 942 }, { "epoch": 0.7867074527252503, "grad_norm": 0.4236140549182892, "learning_rate": 9.219347813790416e-06, "loss": 0.4986, "step": 943 }, { "epoch": 0.7875417130144605, "grad_norm": 0.4009195864200592, "learning_rate": 9.216739737342919e-06, "loss": 0.4479, "step": 944 }, { "epoch": 0.7883759733036707, "grad_norm": 0.4322770833969116, "learning_rate": 9.214127681688927e-06, "loss": 0.4821, "step": 945 }, { "epoch": 0.789210233592881, "grad_norm": 0.48298022150993347, "learning_rate": 9.211511649293358e-06, "loss": 0.4562, "step": 946 }, { "epoch": 0.7900444938820912, "grad_norm": 0.3868677020072937, "learning_rate": 9.208891642624877e-06, "loss": 0.4694, "step": 947 }, { "epoch": 0.7908787541713015, "grad_norm": 0.5126819014549255, "learning_rate": 9.206267664155906e-06, "loss": 0.4866, "step": 948 }, { "epoch": 0.7917130144605117, "grad_norm": 0.4453577697277069, "learning_rate": 9.203639716362612e-06, "loss": 0.4796, "step": 949 }, { "epoch": 0.7925472747497219, "grad_norm": 0.42616450786590576, "learning_rate": 9.201007801724907e-06, "loss": 0.4706, "step": 950 }, { "epoch": 0.7933815350389322, "grad_norm": 0.44628676772117615, "learning_rate": 9.198371922726448e-06, "loss": 0.4534, "step": 951 }, { "epoch": 0.7942157953281423, "grad_norm": 0.4524332284927368, "learning_rate": 9.19573208185463e-06, "loss": 0.4846, "step": 952 }, { "epoch": 0.7950500556173526, "grad_norm": 0.397784024477005, "learning_rate": 9.193088281600592e-06, "loss": 0.4543, "step": 953 }, { "epoch": 0.7958843159065628, "grad_norm": 0.38583263754844666, "learning_rate": 9.190440524459203e-06, "loss": 0.4743, "step": 954 }, { "epoch": 0.7967185761957731, "grad_norm": 0.46938666701316833, "learning_rate": 9.187788812929074e-06, "loss": 0.4533, "step": 955 }, { "epoch": 0.7975528364849833, "grad_norm": 0.41229352355003357, "learning_rate": 9.185133149512538e-06, "loss": 0.4533, "step": 956 }, { "epoch": 0.7983870967741935, "grad_norm": 0.3625040352344513, "learning_rate": 9.182473536715666e-06, "loss": 0.4479, "step": 957 }, { "epoch": 0.7992213570634038, "grad_norm": 0.43269193172454834, "learning_rate": 9.179809977048249e-06, "loss": 0.4386, "step": 958 }, { "epoch": 0.800055617352614, "grad_norm": 0.36787861585617065, "learning_rate": 9.177142473023807e-06, "loss": 0.4676, "step": 959 }, { "epoch": 0.8008898776418243, "grad_norm": 0.389072448015213, "learning_rate": 9.174471027159582e-06, "loss": 0.4506, "step": 960 }, { "epoch": 0.8017241379310345, "grad_norm": 0.37853920459747314, "learning_rate": 9.171795641976532e-06, "loss": 0.449, "step": 961 }, { "epoch": 0.8025583982202447, "grad_norm": 0.4144928455352783, "learning_rate": 9.169116319999336e-06, "loss": 0.4873, "step": 962 }, { "epoch": 0.803392658509455, "grad_norm": 0.3869331181049347, "learning_rate": 9.16643306375639e-06, "loss": 0.4548, "step": 963 }, { "epoch": 0.8042269187986651, "grad_norm": 0.4542870819568634, "learning_rate": 9.163745875779795e-06, "loss": 0.473, "step": 964 }, { "epoch": 0.8050611790878754, "grad_norm": 0.4119902551174164, "learning_rate": 9.16105475860537e-06, "loss": 0.4683, "step": 965 }, { "epoch": 0.8058954393770856, "grad_norm": 0.4197347164154053, "learning_rate": 9.158359714772639e-06, "loss": 0.4338, "step": 966 }, { "epoch": 0.8067296996662959, "grad_norm": 0.3957701325416565, "learning_rate": 9.155660746824828e-06, "loss": 0.4649, "step": 967 }, { "epoch": 0.8075639599555061, "grad_norm": 0.4912014305591583, "learning_rate": 9.152957857308872e-06, "loss": 0.4385, "step": 968 }, { "epoch": 0.8083982202447163, "grad_norm": 0.42331212759017944, "learning_rate": 9.150251048775403e-06, "loss": 0.4711, "step": 969 }, { "epoch": 0.8092324805339266, "grad_norm": 0.43447136878967285, "learning_rate": 9.147540323778754e-06, "loss": 0.4293, "step": 970 }, { "epoch": 0.8100667408231368, "grad_norm": 0.4108222424983978, "learning_rate": 9.14482568487695e-06, "loss": 0.4811, "step": 971 }, { "epoch": 0.8109010011123471, "grad_norm": 0.38900354504585266, "learning_rate": 9.142107134631715e-06, "loss": 0.454, "step": 972 }, { "epoch": 0.8117352614015573, "grad_norm": 0.3810662031173706, "learning_rate": 9.139384675608455e-06, "loss": 0.4404, "step": 973 }, { "epoch": 0.8125695216907676, "grad_norm": 0.37715259194374084, "learning_rate": 9.136658310376274e-06, "loss": 0.4651, "step": 974 }, { "epoch": 0.8134037819799778, "grad_norm": 0.4148033559322357, "learning_rate": 9.133928041507957e-06, "loss": 0.4642, "step": 975 }, { "epoch": 0.814238042269188, "grad_norm": 0.3900693655014038, "learning_rate": 9.131193871579975e-06, "loss": 0.475, "step": 976 }, { "epoch": 0.8150723025583982, "grad_norm": 0.4385911822319031, "learning_rate": 9.128455803172478e-06, "loss": 0.4821, "step": 977 }, { "epoch": 0.8159065628476084, "grad_norm": 0.389020711183548, "learning_rate": 9.125713838869298e-06, "loss": 0.4615, "step": 978 }, { "epoch": 0.8167408231368187, "grad_norm": 0.42442530393600464, "learning_rate": 9.12296798125794e-06, "loss": 0.475, "step": 979 }, { "epoch": 0.8175750834260289, "grad_norm": 0.5191036462783813, "learning_rate": 9.120218232929585e-06, "loss": 0.4589, "step": 980 }, { "epoch": 0.8184093437152391, "grad_norm": 0.3842598497867584, "learning_rate": 9.117464596479087e-06, "loss": 0.5005, "step": 981 }, { "epoch": 0.8192436040044494, "grad_norm": 0.4663434326648712, "learning_rate": 9.114707074504966e-06, "loss": 0.4602, "step": 982 }, { "epoch": 0.8200778642936596, "grad_norm": 0.5281767845153809, "learning_rate": 9.111945669609408e-06, "loss": 0.4608, "step": 983 }, { "epoch": 0.8209121245828699, "grad_norm": 0.3892832100391388, "learning_rate": 9.109180384398267e-06, "loss": 0.4542, "step": 984 }, { "epoch": 0.8217463848720801, "grad_norm": 0.6043713092803955, "learning_rate": 9.106411221481057e-06, "loss": 0.475, "step": 985 }, { "epoch": 0.8225806451612904, "grad_norm": 0.49105918407440186, "learning_rate": 9.103638183470953e-06, "loss": 0.4607, "step": 986 }, { "epoch": 0.8234149054505006, "grad_norm": 0.4218956232070923, "learning_rate": 9.10086127298478e-06, "loss": 0.4728, "step": 987 }, { "epoch": 0.8242491657397107, "grad_norm": 0.4650801420211792, "learning_rate": 9.098080492643026e-06, "loss": 0.4617, "step": 988 }, { "epoch": 0.825083426028921, "grad_norm": 0.46243512630462646, "learning_rate": 9.095295845069827e-06, "loss": 0.4671, "step": 989 }, { "epoch": 0.8259176863181312, "grad_norm": 0.4062410593032837, "learning_rate": 9.092507332892968e-06, "loss": 0.5, "step": 990 }, { "epoch": 0.8267519466073415, "grad_norm": 0.4292340874671936, "learning_rate": 9.089714958743881e-06, "loss": 0.4844, "step": 991 }, { "epoch": 0.8275862068965517, "grad_norm": 0.42713284492492676, "learning_rate": 9.086918725257643e-06, "loss": 0.439, "step": 992 }, { "epoch": 0.828420467185762, "grad_norm": 0.36653932929039, "learning_rate": 9.084118635072974e-06, "loss": 0.4448, "step": 993 }, { "epoch": 0.8292547274749722, "grad_norm": 0.39162346720695496, "learning_rate": 9.08131469083223e-06, "loss": 0.4631, "step": 994 }, { "epoch": 0.8300889877641824, "grad_norm": 0.4857600927352905, "learning_rate": 9.078506895181408e-06, "loss": 0.4489, "step": 995 }, { "epoch": 0.8309232480533927, "grad_norm": 0.40378639101982117, "learning_rate": 9.075695250770138e-06, "loss": 0.461, "step": 996 }, { "epoch": 0.8317575083426029, "grad_norm": 0.4279227554798126, "learning_rate": 9.07287976025168e-06, "loss": 0.4739, "step": 997 }, { "epoch": 0.8325917686318132, "grad_norm": 0.4145830273628235, "learning_rate": 9.070060426282924e-06, "loss": 0.4647, "step": 998 }, { "epoch": 0.8334260289210234, "grad_norm": 0.458044171333313, "learning_rate": 9.067237251524392e-06, "loss": 0.4687, "step": 999 }, { "epoch": 0.8342602892102335, "grad_norm": 0.44754427671432495, "learning_rate": 9.064410238640225e-06, "loss": 0.4679, "step": 1000 }, { "epoch": 0.8350945494994438, "grad_norm": 0.4402177929878235, "learning_rate": 9.061579390298185e-06, "loss": 0.4697, "step": 1001 }, { "epoch": 0.835928809788654, "grad_norm": 0.4585135281085968, "learning_rate": 9.058744709169659e-06, "loss": 0.4761, "step": 1002 }, { "epoch": 0.8367630700778643, "grad_norm": 0.4092285633087158, "learning_rate": 9.055906197929645e-06, "loss": 0.45, "step": 1003 }, { "epoch": 0.8375973303670745, "grad_norm": 0.48842868208885193, "learning_rate": 9.05306385925676e-06, "loss": 0.4628, "step": 1004 }, { "epoch": 0.8384315906562848, "grad_norm": 0.4393143653869629, "learning_rate": 9.050217695833228e-06, "loss": 0.4801, "step": 1005 }, { "epoch": 0.839265850945495, "grad_norm": 0.3815311789512634, "learning_rate": 9.047367710344888e-06, "loss": 0.4604, "step": 1006 }, { "epoch": 0.8401001112347052, "grad_norm": 0.5236709117889404, "learning_rate": 9.044513905481182e-06, "loss": 0.4717, "step": 1007 }, { "epoch": 0.8409343715239155, "grad_norm": 0.39536532759666443, "learning_rate": 9.041656283935158e-06, "loss": 0.4357, "step": 1008 }, { "epoch": 0.8417686318131257, "grad_norm": 0.3960336148738861, "learning_rate": 9.038794848403463e-06, "loss": 0.4624, "step": 1009 }, { "epoch": 0.842602892102336, "grad_norm": 0.4288416802883148, "learning_rate": 9.035929601586347e-06, "loss": 0.4476, "step": 1010 }, { "epoch": 0.8434371523915462, "grad_norm": 0.43405285477638245, "learning_rate": 9.033060546187651e-06, "loss": 0.4598, "step": 1011 }, { "epoch": 0.8442714126807565, "grad_norm": 0.40863484144210815, "learning_rate": 9.030187684914817e-06, "loss": 0.4639, "step": 1012 }, { "epoch": 0.8451056729699666, "grad_norm": 0.4494283199310303, "learning_rate": 9.027311020478876e-06, "loss": 0.4791, "step": 1013 }, { "epoch": 0.8459399332591768, "grad_norm": 0.4609076976776123, "learning_rate": 9.024430555594443e-06, "loss": 0.4833, "step": 1014 }, { "epoch": 0.8467741935483871, "grad_norm": 0.39032191038131714, "learning_rate": 9.021546292979725e-06, "loss": 0.4489, "step": 1015 }, { "epoch": 0.8476084538375973, "grad_norm": 0.46165934205055237, "learning_rate": 9.018658235356514e-06, "loss": 0.4824, "step": 1016 }, { "epoch": 0.8484427141268076, "grad_norm": 0.43186482787132263, "learning_rate": 9.015766385450177e-06, "loss": 0.491, "step": 1017 }, { "epoch": 0.8492769744160178, "grad_norm": 0.45998620986938477, "learning_rate": 9.012870745989663e-06, "loss": 0.434, "step": 1018 }, { "epoch": 0.850111234705228, "grad_norm": 0.3752823770046234, "learning_rate": 9.0099713197075e-06, "loss": 0.4377, "step": 1019 }, { "epoch": 0.8509454949944383, "grad_norm": 0.46493247151374817, "learning_rate": 9.007068109339783e-06, "loss": 0.4732, "step": 1020 }, { "epoch": 0.8517797552836485, "grad_norm": 0.46487540006637573, "learning_rate": 9.004161117626186e-06, "loss": 0.4667, "step": 1021 }, { "epoch": 0.8526140155728588, "grad_norm": 0.36888015270233154, "learning_rate": 9.001250347309942e-06, "loss": 0.4532, "step": 1022 }, { "epoch": 0.853448275862069, "grad_norm": 0.501605749130249, "learning_rate": 8.998335801137858e-06, "loss": 0.4722, "step": 1023 }, { "epoch": 0.8542825361512792, "grad_norm": 0.3740626275539398, "learning_rate": 8.9954174818603e-06, "loss": 0.458, "step": 1024 }, { "epoch": 0.8551167964404894, "grad_norm": 0.37449079751968384, "learning_rate": 8.992495392231195e-06, "loss": 0.459, "step": 1025 }, { "epoch": 0.8559510567296996, "grad_norm": 0.44619497656822205, "learning_rate": 8.989569535008027e-06, "loss": 0.475, "step": 1026 }, { "epoch": 0.8567853170189099, "grad_norm": 0.37090879678726196, "learning_rate": 8.98663991295184e-06, "loss": 0.4438, "step": 1027 }, { "epoch": 0.8576195773081201, "grad_norm": 0.36330893635749817, "learning_rate": 8.983706528827226e-06, "loss": 0.4618, "step": 1028 }, { "epoch": 0.8584538375973304, "grad_norm": 0.4073713719844818, "learning_rate": 8.980769385402328e-06, "loss": 0.4597, "step": 1029 }, { "epoch": 0.8592880978865406, "grad_norm": 0.378522127866745, "learning_rate": 8.977828485448837e-06, "loss": 0.4406, "step": 1030 }, { "epoch": 0.8601223581757509, "grad_norm": 0.37273988127708435, "learning_rate": 8.97488383174199e-06, "loss": 0.4534, "step": 1031 }, { "epoch": 0.8609566184649611, "grad_norm": 0.40679633617401123, "learning_rate": 8.971935427060563e-06, "loss": 0.468, "step": 1032 }, { "epoch": 0.8617908787541713, "grad_norm": 0.3615160286426544, "learning_rate": 8.96898327418688e-06, "loss": 0.4553, "step": 1033 }, { "epoch": 0.8626251390433816, "grad_norm": 0.4568735957145691, "learning_rate": 8.966027375906787e-06, "loss": 0.4951, "step": 1034 }, { "epoch": 0.8634593993325917, "grad_norm": 0.4463333785533905, "learning_rate": 8.963067735009683e-06, "loss": 0.4894, "step": 1035 }, { "epoch": 0.864293659621802, "grad_norm": 0.3886243402957916, "learning_rate": 8.960104354288481e-06, "loss": 0.4563, "step": 1036 }, { "epoch": 0.8651279199110122, "grad_norm": 0.4114626348018646, "learning_rate": 8.957137236539638e-06, "loss": 0.4777, "step": 1037 }, { "epoch": 0.8659621802002224, "grad_norm": 0.42635855078697205, "learning_rate": 8.954166384563128e-06, "loss": 0.4474, "step": 1038 }, { "epoch": 0.8667964404894327, "grad_norm": 0.34951335191726685, "learning_rate": 8.951191801162453e-06, "loss": 0.4466, "step": 1039 }, { "epoch": 0.8676307007786429, "grad_norm": 0.4207518994808197, "learning_rate": 8.948213489144633e-06, "loss": 0.4478, "step": 1040 }, { "epoch": 0.8684649610678532, "grad_norm": 0.4197646677494049, "learning_rate": 8.945231451320208e-06, "loss": 0.5032, "step": 1041 }, { "epoch": 0.8692992213570634, "grad_norm": 0.35476863384246826, "learning_rate": 8.94224569050324e-06, "loss": 0.4252, "step": 1042 }, { "epoch": 0.8701334816462737, "grad_norm": 0.48652228713035583, "learning_rate": 8.939256209511293e-06, "loss": 0.4805, "step": 1043 }, { "epoch": 0.8709677419354839, "grad_norm": 0.4563238322734833, "learning_rate": 8.936263011165451e-06, "loss": 0.4975, "step": 1044 }, { "epoch": 0.8718020022246941, "grad_norm": 0.4107232391834259, "learning_rate": 8.933266098290301e-06, "loss": 0.4852, "step": 1045 }, { "epoch": 0.8726362625139044, "grad_norm": 0.4778875410556793, "learning_rate": 8.930265473713939e-06, "loss": 0.4942, "step": 1046 }, { "epoch": 0.8734705228031145, "grad_norm": 0.40038999915122986, "learning_rate": 8.927261140267956e-06, "loss": 0.4589, "step": 1047 }, { "epoch": 0.8743047830923248, "grad_norm": 0.41000422835350037, "learning_rate": 8.924253100787454e-06, "loss": 0.45, "step": 1048 }, { "epoch": 0.875139043381535, "grad_norm": 0.4242759943008423, "learning_rate": 8.921241358111023e-06, "loss": 0.4709, "step": 1049 }, { "epoch": 0.8759733036707452, "grad_norm": 0.45711013674736023, "learning_rate": 8.918225915080754e-06, "loss": 0.439, "step": 1050 }, { "epoch": 0.8768075639599555, "grad_norm": 0.46594780683517456, "learning_rate": 8.915206774542225e-06, "loss": 0.4547, "step": 1051 }, { "epoch": 0.8776418242491657, "grad_norm": 0.5010321736335754, "learning_rate": 8.912183939344505e-06, "loss": 0.4496, "step": 1052 }, { "epoch": 0.878476084538376, "grad_norm": 0.4597023129463196, "learning_rate": 8.90915741234015e-06, "loss": 0.439, "step": 1053 }, { "epoch": 0.8793103448275862, "grad_norm": 0.4273754954338074, "learning_rate": 8.906127196385202e-06, "loss": 0.4604, "step": 1054 }, { "epoch": 0.8801446051167965, "grad_norm": 0.4115888774394989, "learning_rate": 8.903093294339177e-06, "loss": 0.4656, "step": 1055 }, { "epoch": 0.8809788654060067, "grad_norm": 0.4938189387321472, "learning_rate": 8.900055709065078e-06, "loss": 0.4691, "step": 1056 }, { "epoch": 0.8818131256952169, "grad_norm": 0.36657044291496277, "learning_rate": 8.897014443429381e-06, "loss": 0.4503, "step": 1057 }, { "epoch": 0.8826473859844272, "grad_norm": 0.421212375164032, "learning_rate": 8.893969500302031e-06, "loss": 0.4651, "step": 1058 }, { "epoch": 0.8834816462736373, "grad_norm": 0.43546152114868164, "learning_rate": 8.89092088255645e-06, "loss": 0.4615, "step": 1059 }, { "epoch": 0.8843159065628476, "grad_norm": 0.3910651206970215, "learning_rate": 8.88786859306952e-06, "loss": 0.4441, "step": 1060 }, { "epoch": 0.8851501668520578, "grad_norm": 0.45837149024009705, "learning_rate": 8.884812634721592e-06, "loss": 0.4323, "step": 1061 }, { "epoch": 0.8859844271412681, "grad_norm": 0.4571623206138611, "learning_rate": 8.881753010396484e-06, "loss": 0.4551, "step": 1062 }, { "epoch": 0.8868186874304783, "grad_norm": 0.5300352573394775, "learning_rate": 8.878689722981462e-06, "loss": 0.4919, "step": 1063 }, { "epoch": 0.8876529477196885, "grad_norm": 0.3761040270328522, "learning_rate": 8.87562277536726e-06, "loss": 0.4541, "step": 1064 }, { "epoch": 0.8884872080088988, "grad_norm": 0.4909161627292633, "learning_rate": 8.87255217044806e-06, "loss": 0.4765, "step": 1065 }, { "epoch": 0.889321468298109, "grad_norm": 0.4710785448551178, "learning_rate": 8.869477911121495e-06, "loss": 0.4781, "step": 1066 }, { "epoch": 0.8901557285873193, "grad_norm": 0.38558924198150635, "learning_rate": 8.866400000288652e-06, "loss": 0.4586, "step": 1067 }, { "epoch": 0.8909899888765295, "grad_norm": 0.5103597640991211, "learning_rate": 8.863318440854057e-06, "loss": 0.4955, "step": 1068 }, { "epoch": 0.8918242491657397, "grad_norm": 0.41212478280067444, "learning_rate": 8.860233235725682e-06, "loss": 0.4671, "step": 1069 }, { "epoch": 0.89265850945495, "grad_norm": 0.40688809752464294, "learning_rate": 8.85714438781494e-06, "loss": 0.4609, "step": 1070 }, { "epoch": 0.8934927697441601, "grad_norm": 0.39361870288848877, "learning_rate": 8.854051900036683e-06, "loss": 0.4105, "step": 1071 }, { "epoch": 0.8943270300333704, "grad_norm": 0.4114530384540558, "learning_rate": 8.850955775309191e-06, "loss": 0.4819, "step": 1072 }, { "epoch": 0.8951612903225806, "grad_norm": 0.4132308065891266, "learning_rate": 8.847856016554187e-06, "loss": 0.4618, "step": 1073 }, { "epoch": 0.8959955506117909, "grad_norm": 0.41540011763572693, "learning_rate": 8.84475262669681e-06, "loss": 0.4493, "step": 1074 }, { "epoch": 0.8968298109010011, "grad_norm": 0.3894135653972626, "learning_rate": 8.84164560866564e-06, "loss": 0.4619, "step": 1075 }, { "epoch": 0.8976640711902113, "grad_norm": 0.4133654832839966, "learning_rate": 8.838534965392669e-06, "loss": 0.4181, "step": 1076 }, { "epoch": 0.8984983314794216, "grad_norm": 0.3832062780857086, "learning_rate": 8.835420699813315e-06, "loss": 0.4584, "step": 1077 }, { "epoch": 0.8993325917686318, "grad_norm": 0.3783755302429199, "learning_rate": 8.832302814866416e-06, "loss": 0.3918, "step": 1078 }, { "epoch": 0.9001668520578421, "grad_norm": 0.3818853199481964, "learning_rate": 8.829181313494222e-06, "loss": 0.4779, "step": 1079 }, { "epoch": 0.9010011123470523, "grad_norm": 0.394758939743042, "learning_rate": 8.826056198642396e-06, "loss": 0.4417, "step": 1080 }, { "epoch": 0.9018353726362626, "grad_norm": 0.4081030786037445, "learning_rate": 8.822927473260012e-06, "loss": 0.4933, "step": 1081 }, { "epoch": 0.9026696329254728, "grad_norm": 0.37765511870384216, "learning_rate": 8.819795140299554e-06, "loss": 0.4502, "step": 1082 }, { "epoch": 0.9035038932146829, "grad_norm": 0.39129066467285156, "learning_rate": 8.816659202716905e-06, "loss": 0.4504, "step": 1083 }, { "epoch": 0.9043381535038932, "grad_norm": 0.36765623092651367, "learning_rate": 8.813519663471351e-06, "loss": 0.4512, "step": 1084 }, { "epoch": 0.9051724137931034, "grad_norm": 0.4080364406108856, "learning_rate": 8.810376525525579e-06, "loss": 0.4943, "step": 1085 }, { "epoch": 0.9060066740823137, "grad_norm": 0.39574599266052246, "learning_rate": 8.807229791845673e-06, "loss": 0.4437, "step": 1086 }, { "epoch": 0.9068409343715239, "grad_norm": 0.3774428367614746, "learning_rate": 8.804079465401104e-06, "loss": 0.4388, "step": 1087 }, { "epoch": 0.9076751946607341, "grad_norm": 0.5327045917510986, "learning_rate": 8.800925549164742e-06, "loss": 0.4887, "step": 1088 }, { "epoch": 0.9085094549499444, "grad_norm": 0.3926681876182556, "learning_rate": 8.797768046112836e-06, "loss": 0.4895, "step": 1089 }, { "epoch": 0.9093437152391546, "grad_norm": 0.49882641434669495, "learning_rate": 8.794606959225024e-06, "loss": 0.4536, "step": 1090 }, { "epoch": 0.9101779755283649, "grad_norm": 0.3945174515247345, "learning_rate": 8.791442291484329e-06, "loss": 0.4647, "step": 1091 }, { "epoch": 0.9110122358175751, "grad_norm": 0.38362613320350647, "learning_rate": 8.788274045877145e-06, "loss": 0.4774, "step": 1092 }, { "epoch": 0.9118464961067854, "grad_norm": 0.4516408145427704, "learning_rate": 8.78510222539325e-06, "loss": 0.4852, "step": 1093 }, { "epoch": 0.9126807563959956, "grad_norm": 0.38279539346694946, "learning_rate": 8.781926833025793e-06, "loss": 0.4785, "step": 1094 }, { "epoch": 0.9135150166852057, "grad_norm": 0.4005299508571625, "learning_rate": 8.778747871771293e-06, "loss": 0.4618, "step": 1095 }, { "epoch": 0.914349276974416, "grad_norm": 0.38859423995018005, "learning_rate": 8.775565344629636e-06, "loss": 0.4329, "step": 1096 }, { "epoch": 0.9151835372636262, "grad_norm": 0.41498932242393494, "learning_rate": 8.772379254604074e-06, "loss": 0.462, "step": 1097 }, { "epoch": 0.9160177975528365, "grad_norm": 0.36600467562675476, "learning_rate": 8.769189604701221e-06, "loss": 0.4832, "step": 1098 }, { "epoch": 0.9168520578420467, "grad_norm": 0.39121365547180176, "learning_rate": 8.765996397931053e-06, "loss": 0.4654, "step": 1099 }, { "epoch": 0.917686318131257, "grad_norm": 0.4243551194667816, "learning_rate": 8.762799637306899e-06, "loss": 0.4566, "step": 1100 }, { "epoch": 0.9185205784204672, "grad_norm": 0.3581959009170532, "learning_rate": 8.759599325845442e-06, "loss": 0.466, "step": 1101 }, { "epoch": 0.9193548387096774, "grad_norm": 0.4194590747356415, "learning_rate": 8.756395466566718e-06, "loss": 0.4718, "step": 1102 }, { "epoch": 0.9201890989988877, "grad_norm": 0.384968101978302, "learning_rate": 8.75318806249411e-06, "loss": 0.4615, "step": 1103 }, { "epoch": 0.9210233592880979, "grad_norm": 0.37652865052223206, "learning_rate": 8.749977116654343e-06, "loss": 0.4484, "step": 1104 }, { "epoch": 0.9218576195773082, "grad_norm": 0.4349343180656433, "learning_rate": 8.74676263207749e-06, "loss": 0.4757, "step": 1105 }, { "epoch": 0.9226918798665183, "grad_norm": 0.37389683723449707, "learning_rate": 8.743544611796957e-06, "loss": 0.4607, "step": 1106 }, { "epoch": 0.9235261401557285, "grad_norm": 0.3857394754886627, "learning_rate": 8.740323058849494e-06, "loss": 0.4779, "step": 1107 }, { "epoch": 0.9243604004449388, "grad_norm": 0.38690078258514404, "learning_rate": 8.737097976275177e-06, "loss": 0.486, "step": 1108 }, { "epoch": 0.925194660734149, "grad_norm": 0.4097394049167633, "learning_rate": 8.73386936711742e-06, "loss": 0.4685, "step": 1109 }, { "epoch": 0.9260289210233593, "grad_norm": 0.3998280465602875, "learning_rate": 8.730637234422958e-06, "loss": 0.4745, "step": 1110 }, { "epoch": 0.9268631813125695, "grad_norm": 0.37257352471351624, "learning_rate": 8.727401581241857e-06, "loss": 0.4856, "step": 1111 }, { "epoch": 0.9276974416017798, "grad_norm": 0.4095642864704132, "learning_rate": 8.7241624106275e-06, "loss": 0.4446, "step": 1112 }, { "epoch": 0.92853170189099, "grad_norm": 0.413920134305954, "learning_rate": 8.720919725636592e-06, "loss": 0.4607, "step": 1113 }, { "epoch": 0.9293659621802002, "grad_norm": 0.40788841247558594, "learning_rate": 8.717673529329158e-06, "loss": 0.4698, "step": 1114 }, { "epoch": 0.9302002224694105, "grad_norm": 0.39772751927375793, "learning_rate": 8.714423824768527e-06, "loss": 0.4692, "step": 1115 }, { "epoch": 0.9310344827586207, "grad_norm": 0.4270651340484619, "learning_rate": 8.71117061502135e-06, "loss": 0.4785, "step": 1116 }, { "epoch": 0.931868743047831, "grad_norm": 0.39283356070518494, "learning_rate": 8.707913903157576e-06, "loss": 0.4544, "step": 1117 }, { "epoch": 0.9327030033370411, "grad_norm": 0.4233250319957733, "learning_rate": 8.704653692250466e-06, "loss": 0.4746, "step": 1118 }, { "epoch": 0.9335372636262514, "grad_norm": 0.437547504901886, "learning_rate": 8.701389985376578e-06, "loss": 0.4403, "step": 1119 }, { "epoch": 0.9343715239154616, "grad_norm": 0.3975749611854553, "learning_rate": 8.698122785615773e-06, "loss": 0.4584, "step": 1120 }, { "epoch": 0.9352057842046718, "grad_norm": 0.4367460608482361, "learning_rate": 8.694852096051203e-06, "loss": 0.4562, "step": 1121 }, { "epoch": 0.9360400444938821, "grad_norm": 0.4739750325679779, "learning_rate": 8.691577919769317e-06, "loss": 0.4715, "step": 1122 }, { "epoch": 0.9368743047830923, "grad_norm": 0.4404827654361725, "learning_rate": 8.688300259859855e-06, "loss": 0.4707, "step": 1123 }, { "epoch": 0.9377085650723026, "grad_norm": 0.3645053505897522, "learning_rate": 8.685019119415841e-06, "loss": 0.4542, "step": 1124 }, { "epoch": 0.9385428253615128, "grad_norm": 0.5085168480873108, "learning_rate": 8.681734501533588e-06, "loss": 0.487, "step": 1125 }, { "epoch": 0.939377085650723, "grad_norm": 0.41370585560798645, "learning_rate": 8.678446409312687e-06, "loss": 0.4723, "step": 1126 }, { "epoch": 0.9402113459399333, "grad_norm": 0.405166894197464, "learning_rate": 8.675154845856009e-06, "loss": 0.4411, "step": 1127 }, { "epoch": 0.9410456062291435, "grad_norm": 0.4386710524559021, "learning_rate": 8.671859814269698e-06, "loss": 0.4647, "step": 1128 }, { "epoch": 0.9418798665183538, "grad_norm": 0.45375555753707886, "learning_rate": 8.668561317663178e-06, "loss": 0.4457, "step": 1129 }, { "epoch": 0.942714126807564, "grad_norm": 0.47788673639297485, "learning_rate": 8.665259359149132e-06, "loss": 0.4422, "step": 1130 }, { "epoch": 0.9435483870967742, "grad_norm": 0.43363678455352783, "learning_rate": 8.661953941843521e-06, "loss": 0.5002, "step": 1131 }, { "epoch": 0.9443826473859844, "grad_norm": 0.44550037384033203, "learning_rate": 8.658645068865561e-06, "loss": 0.4291, "step": 1132 }, { "epoch": 0.9452169076751946, "grad_norm": 0.45737162232398987, "learning_rate": 8.655332743337735e-06, "loss": 0.4562, "step": 1133 }, { "epoch": 0.9460511679644049, "grad_norm": 0.42214739322662354, "learning_rate": 8.652016968385779e-06, "loss": 0.4864, "step": 1134 }, { "epoch": 0.9468854282536151, "grad_norm": 0.42924219369888306, "learning_rate": 8.64869774713869e-06, "loss": 0.4506, "step": 1135 }, { "epoch": 0.9477196885428254, "grad_norm": 0.4467397630214691, "learning_rate": 8.645375082728715e-06, "loss": 0.4624, "step": 1136 }, { "epoch": 0.9485539488320356, "grad_norm": 0.3702987730503082, "learning_rate": 8.642048978291347e-06, "loss": 0.451, "step": 1137 }, { "epoch": 0.9493882091212458, "grad_norm": 0.43321993947029114, "learning_rate": 8.638719436965327e-06, "loss": 0.4683, "step": 1138 }, { "epoch": 0.9502224694104561, "grad_norm": 0.4426807463169098, "learning_rate": 8.63538646189264e-06, "loss": 0.4547, "step": 1139 }, { "epoch": 0.9510567296996663, "grad_norm": 0.4010275900363922, "learning_rate": 8.632050056218515e-06, "loss": 0.4546, "step": 1140 }, { "epoch": 0.9518909899888766, "grad_norm": 0.40140464901924133, "learning_rate": 8.62871022309141e-06, "loss": 0.4741, "step": 1141 }, { "epoch": 0.9527252502780867, "grad_norm": 0.3634575307369232, "learning_rate": 8.625366965663024e-06, "loss": 0.4645, "step": 1142 }, { "epoch": 0.953559510567297, "grad_norm": 0.42506226897239685, "learning_rate": 8.622020287088286e-06, "loss": 0.4526, "step": 1143 }, { "epoch": 0.9543937708565072, "grad_norm": 0.3970424234867096, "learning_rate": 8.61867019052535e-06, "loss": 0.4339, "step": 1144 }, { "epoch": 0.9552280311457174, "grad_norm": 0.3777240216732025, "learning_rate": 8.615316679135602e-06, "loss": 0.4401, "step": 1145 }, { "epoch": 0.9560622914349277, "grad_norm": 0.3800307810306549, "learning_rate": 8.611959756083646e-06, "loss": 0.4565, "step": 1146 }, { "epoch": 0.9568965517241379, "grad_norm": 0.35066500306129456, "learning_rate": 8.6085994245373e-06, "loss": 0.4559, "step": 1147 }, { "epoch": 0.9577308120133482, "grad_norm": 0.3899294435977936, "learning_rate": 8.605235687667612e-06, "loss": 0.4492, "step": 1148 }, { "epoch": 0.9585650723025584, "grad_norm": 0.38308271765708923, "learning_rate": 8.601868548648834e-06, "loss": 0.4608, "step": 1149 }, { "epoch": 0.9593993325917687, "grad_norm": 0.41305652260780334, "learning_rate": 8.598498010658431e-06, "loss": 0.4655, "step": 1150 }, { "epoch": 0.9602335928809789, "grad_norm": 0.38035741448402405, "learning_rate": 8.595124076877074e-06, "loss": 0.4229, "step": 1151 }, { "epoch": 0.9610678531701891, "grad_norm": 0.38335081934928894, "learning_rate": 8.591746750488639e-06, "loss": 0.4647, "step": 1152 }, { "epoch": 0.9619021134593994, "grad_norm": 0.40261122584342957, "learning_rate": 8.588366034680205e-06, "loss": 0.4628, "step": 1153 }, { "epoch": 0.9627363737486095, "grad_norm": 0.418799489736557, "learning_rate": 8.58498193264205e-06, "loss": 0.4621, "step": 1154 }, { "epoch": 0.9635706340378198, "grad_norm": 0.37866294384002686, "learning_rate": 8.581594447567645e-06, "loss": 0.4695, "step": 1155 }, { "epoch": 0.96440489432703, "grad_norm": 0.4416775703430176, "learning_rate": 8.578203582653654e-06, "loss": 0.4813, "step": 1156 }, { "epoch": 0.9652391546162402, "grad_norm": 0.396340936422348, "learning_rate": 8.574809341099932e-06, "loss": 0.4812, "step": 1157 }, { "epoch": 0.9660734149054505, "grad_norm": 0.43418172001838684, "learning_rate": 8.571411726109518e-06, "loss": 0.4626, "step": 1158 }, { "epoch": 0.9669076751946607, "grad_norm": 0.44491395354270935, "learning_rate": 8.568010740888639e-06, "loss": 0.4573, "step": 1159 }, { "epoch": 0.967741935483871, "grad_norm": 0.4255155920982361, "learning_rate": 8.564606388646696e-06, "loss": 0.4397, "step": 1160 }, { "epoch": 0.9685761957730812, "grad_norm": 0.4496902525424957, "learning_rate": 8.561198672596271e-06, "loss": 0.4433, "step": 1161 }, { "epoch": 0.9694104560622915, "grad_norm": 0.3742341697216034, "learning_rate": 8.557787595953123e-06, "loss": 0.4464, "step": 1162 }, { "epoch": 0.9702447163515017, "grad_norm": 0.5138442516326904, "learning_rate": 8.554373161936176e-06, "loss": 0.4845, "step": 1163 }, { "epoch": 0.9710789766407119, "grad_norm": 0.3873581290245056, "learning_rate": 8.550955373767527e-06, "loss": 0.4467, "step": 1164 }, { "epoch": 0.9719132369299222, "grad_norm": 0.5042309761047363, "learning_rate": 8.547534234672435e-06, "loss": 0.4638, "step": 1165 }, { "epoch": 0.9727474972191323, "grad_norm": 0.41346895694732666, "learning_rate": 8.544109747879326e-06, "loss": 0.4532, "step": 1166 }, { "epoch": 0.9735817575083426, "grad_norm": 0.3949490785598755, "learning_rate": 8.540681916619778e-06, "loss": 0.464, "step": 1167 }, { "epoch": 0.9744160177975528, "grad_norm": 0.4092686176300049, "learning_rate": 8.537250744128531e-06, "loss": 0.4503, "step": 1168 }, { "epoch": 0.9752502780867631, "grad_norm": 0.39671772718429565, "learning_rate": 8.533816233643478e-06, "loss": 0.4789, "step": 1169 }, { "epoch": 0.9760845383759733, "grad_norm": 0.4615577459335327, "learning_rate": 8.530378388405656e-06, "loss": 0.4879, "step": 1170 }, { "epoch": 0.9769187986651835, "grad_norm": 0.4319697916507721, "learning_rate": 8.526937211659255e-06, "loss": 0.4727, "step": 1171 }, { "epoch": 0.9777530589543938, "grad_norm": 0.3359149992465973, "learning_rate": 8.523492706651607e-06, "loss": 0.4269, "step": 1172 }, { "epoch": 0.978587319243604, "grad_norm": 0.3656979501247406, "learning_rate": 8.520044876633186e-06, "loss": 0.4577, "step": 1173 }, { "epoch": 0.9794215795328143, "grad_norm": 0.42890122532844543, "learning_rate": 8.516593724857598e-06, "loss": 0.473, "step": 1174 }, { "epoch": 0.9802558398220245, "grad_norm": 0.3703756332397461, "learning_rate": 8.513139254581593e-06, "loss": 0.4785, "step": 1175 }, { "epoch": 0.9810901001112347, "grad_norm": 0.37994149327278137, "learning_rate": 8.509681469065042e-06, "loss": 0.4509, "step": 1176 }, { "epoch": 0.981924360400445, "grad_norm": 0.4303264319896698, "learning_rate": 8.506220371570956e-06, "loss": 0.4649, "step": 1177 }, { "epoch": 0.9827586206896551, "grad_norm": 0.43442729115486145, "learning_rate": 8.50275596536546e-06, "loss": 0.4535, "step": 1178 }, { "epoch": 0.9835928809788654, "grad_norm": 0.3773132264614105, "learning_rate": 8.49928825371781e-06, "loss": 0.4543, "step": 1179 }, { "epoch": 0.9844271412680756, "grad_norm": 0.4067005515098572, "learning_rate": 8.495817239900375e-06, "loss": 0.4926, "step": 1180 }, { "epoch": 0.9852614015572859, "grad_norm": 0.3736957311630249, "learning_rate": 8.492342927188647e-06, "loss": 0.476, "step": 1181 }, { "epoch": 0.9860956618464961, "grad_norm": 0.4138658344745636, "learning_rate": 8.488865318861223e-06, "loss": 0.4667, "step": 1182 }, { "epoch": 0.9869299221357063, "grad_norm": 0.38127124309539795, "learning_rate": 8.485384418199819e-06, "loss": 0.4524, "step": 1183 }, { "epoch": 0.9877641824249166, "grad_norm": 0.35918834805488586, "learning_rate": 8.481900228489248e-06, "loss": 0.4905, "step": 1184 }, { "epoch": 0.9885984427141268, "grad_norm": 0.4804007411003113, "learning_rate": 8.478412753017433e-06, "loss": 0.4768, "step": 1185 }, { "epoch": 0.9894327030033371, "grad_norm": 0.33645591139793396, "learning_rate": 8.474921995075399e-06, "loss": 0.4375, "step": 1186 }, { "epoch": 0.9902669632925473, "grad_norm": 0.34728148579597473, "learning_rate": 8.47142795795726e-06, "loss": 0.4437, "step": 1187 }, { "epoch": 0.9911012235817576, "grad_norm": 0.4384344518184662, "learning_rate": 8.46793064496024e-06, "loss": 0.5017, "step": 1188 }, { "epoch": 0.9919354838709677, "grad_norm": 0.3538776934146881, "learning_rate": 8.464430059384638e-06, "loss": 0.4407, "step": 1189 }, { "epoch": 0.9927697441601779, "grad_norm": 0.36891841888427734, "learning_rate": 8.460926204533849e-06, "loss": 0.4533, "step": 1190 }, { "epoch": 0.9936040044493882, "grad_norm": 0.43469569087028503, "learning_rate": 8.457419083714352e-06, "loss": 0.4408, "step": 1191 }, { "epoch": 0.9944382647385984, "grad_norm": 0.3407193720340729, "learning_rate": 8.453908700235711e-06, "loss": 0.4429, "step": 1192 }, { "epoch": 0.9952725250278087, "grad_norm": 0.3602113127708435, "learning_rate": 8.450395057410561e-06, "loss": 0.4579, "step": 1193 }, { "epoch": 0.9961067853170189, "grad_norm": 0.34148555994033813, "learning_rate": 8.44687815855462e-06, "loss": 0.4462, "step": 1194 }, { "epoch": 0.9969410456062291, "grad_norm": 0.3675920367240906, "learning_rate": 8.44335800698668e-06, "loss": 0.439, "step": 1195 }, { "epoch": 0.9977753058954394, "grad_norm": 0.39623668789863586, "learning_rate": 8.439834606028594e-06, "loss": 0.456, "step": 1196 }, { "epoch": 0.9986095661846496, "grad_norm": 0.4284643828868866, "learning_rate": 8.436307959005287e-06, "loss": 0.4606, "step": 1197 }, { "epoch": 0.9994438264738599, "grad_norm": 0.5012925267219543, "learning_rate": 8.432778069244748e-06, "loss": 0.4633, "step": 1198 }, { "epoch": 1.00027808676307, "grad_norm": 0.49264591932296753, "learning_rate": 8.429244940078025e-06, "loss": 0.6212, "step": 1199 }, { "epoch": 1.0011123470522802, "grad_norm": 0.40633392333984375, "learning_rate": 8.425708574839221e-06, "loss": 0.4423, "step": 1200 }, { "epoch": 1.0019466073414904, "grad_norm": 0.38383084535598755, "learning_rate": 8.422168976865495e-06, "loss": 0.4002, "step": 1201 }, { "epoch": 1.0027808676307008, "grad_norm": 0.3920266628265381, "learning_rate": 8.418626149497053e-06, "loss": 0.4385, "step": 1202 }, { "epoch": 1.003615127919911, "grad_norm": 0.38960161805152893, "learning_rate": 8.415080096077155e-06, "loss": 0.4123, "step": 1203 }, { "epoch": 1.0044493882091212, "grad_norm": 0.4278031587600708, "learning_rate": 8.4115308199521e-06, "loss": 0.4503, "step": 1204 }, { "epoch": 1.0052836484983314, "grad_norm": 0.34154215455055237, "learning_rate": 8.40797832447123e-06, "loss": 0.4144, "step": 1205 }, { "epoch": 1.0061179087875418, "grad_norm": 0.3662768304347992, "learning_rate": 8.404422612986924e-06, "loss": 0.4521, "step": 1206 }, { "epoch": 1.006952169076752, "grad_norm": 0.4136764705181122, "learning_rate": 8.400863688854598e-06, "loss": 0.4918, "step": 1207 }, { "epoch": 1.0077864293659622, "grad_norm": 0.3624766767024994, "learning_rate": 8.397301555432698e-06, "loss": 0.4204, "step": 1208 }, { "epoch": 1.0086206896551724, "grad_norm": 0.35145118832588196, "learning_rate": 8.393736216082697e-06, "loss": 0.3918, "step": 1209 }, { "epoch": 1.0094549499443826, "grad_norm": 0.3710238039493561, "learning_rate": 8.390167674169097e-06, "loss": 0.4586, "step": 1210 }, { "epoch": 1.010289210233593, "grad_norm": 0.37237706780433655, "learning_rate": 8.386595933059421e-06, "loss": 0.4291, "step": 1211 }, { "epoch": 1.0111234705228032, "grad_norm": 0.3923458158969879, "learning_rate": 8.383020996124208e-06, "loss": 0.4497, "step": 1212 }, { "epoch": 1.0119577308120133, "grad_norm": 0.35743463039398193, "learning_rate": 8.379442866737017e-06, "loss": 0.3981, "step": 1213 }, { "epoch": 1.0127919911012235, "grad_norm": 0.41525378823280334, "learning_rate": 8.375861548274417e-06, "loss": 0.5242, "step": 1214 }, { "epoch": 1.0136262513904337, "grad_norm": 0.35609331727027893, "learning_rate": 8.372277044115988e-06, "loss": 0.4096, "step": 1215 }, { "epoch": 1.0144605116796441, "grad_norm": 0.4297907054424286, "learning_rate": 8.368689357644314e-06, "loss": 0.463, "step": 1216 }, { "epoch": 1.0152947719688543, "grad_norm": 0.38932815194129944, "learning_rate": 8.365098492244983e-06, "loss": 0.4136, "step": 1217 }, { "epoch": 1.0161290322580645, "grad_norm": 0.4005068242549896, "learning_rate": 8.361504451306585e-06, "loss": 0.4369, "step": 1218 }, { "epoch": 1.0169632925472747, "grad_norm": 0.33839288353919983, "learning_rate": 8.357907238220704e-06, "loss": 0.3778, "step": 1219 }, { "epoch": 1.0177975528364849, "grad_norm": 0.40979301929473877, "learning_rate": 8.35430685638192e-06, "loss": 0.4356, "step": 1220 }, { "epoch": 1.0186318131256953, "grad_norm": 0.41428670287132263, "learning_rate": 8.3507033091878e-06, "loss": 0.4693, "step": 1221 }, { "epoch": 1.0194660734149055, "grad_norm": 0.38767558336257935, "learning_rate": 8.347096600038898e-06, "loss": 0.3957, "step": 1222 }, { "epoch": 1.0203003337041157, "grad_norm": 0.41299116611480713, "learning_rate": 8.343486732338758e-06, "loss": 0.4247, "step": 1223 }, { "epoch": 1.0211345939933258, "grad_norm": 0.41609710454940796, "learning_rate": 8.339873709493895e-06, "loss": 0.4582, "step": 1224 }, { "epoch": 1.0219688542825363, "grad_norm": 0.3619465231895447, "learning_rate": 8.336257534913813e-06, "loss": 0.3871, "step": 1225 }, { "epoch": 1.0228031145717464, "grad_norm": 0.39721059799194336, "learning_rate": 8.332638212010977e-06, "loss": 0.4497, "step": 1226 }, { "epoch": 1.0236373748609566, "grad_norm": 0.40791189670562744, "learning_rate": 8.329015744200834e-06, "loss": 0.4113, "step": 1227 }, { "epoch": 1.0244716351501668, "grad_norm": 0.4416361451148987, "learning_rate": 8.325390134901794e-06, "loss": 0.4454, "step": 1228 }, { "epoch": 1.025305895439377, "grad_norm": 0.43320390582084656, "learning_rate": 8.321761387535231e-06, "loss": 0.4156, "step": 1229 }, { "epoch": 1.0261401557285874, "grad_norm": 0.43886056542396545, "learning_rate": 8.31812950552548e-06, "loss": 0.4508, "step": 1230 }, { "epoch": 1.0269744160177976, "grad_norm": 0.40097135305404663, "learning_rate": 8.31449449229984e-06, "loss": 0.4652, "step": 1231 }, { "epoch": 1.0278086763070078, "grad_norm": 0.4846936762332916, "learning_rate": 8.310856351288558e-06, "loss": 0.3916, "step": 1232 }, { "epoch": 1.028642936596218, "grad_norm": 0.38358986377716064, "learning_rate": 8.307215085924833e-06, "loss": 0.4219, "step": 1233 }, { "epoch": 1.0294771968854282, "grad_norm": 0.48562484979629517, "learning_rate": 8.303570699644814e-06, "loss": 0.4063, "step": 1234 }, { "epoch": 1.0303114571746386, "grad_norm": 0.40883731842041016, "learning_rate": 8.299923195887599e-06, "loss": 0.399, "step": 1235 }, { "epoch": 1.0311457174638488, "grad_norm": 0.46416184306144714, "learning_rate": 8.296272578095218e-06, "loss": 0.4231, "step": 1236 }, { "epoch": 1.031979977753059, "grad_norm": 0.39179912209510803, "learning_rate": 8.292618849712655e-06, "loss": 0.3903, "step": 1237 }, { "epoch": 1.0328142380422691, "grad_norm": 0.48861297965049744, "learning_rate": 8.288962014187811e-06, "loss": 0.4088, "step": 1238 }, { "epoch": 1.0336484983314793, "grad_norm": 0.4720747768878937, "learning_rate": 8.285302074971532e-06, "loss": 0.419, "step": 1239 }, { "epoch": 1.0344827586206897, "grad_norm": 0.40325987339019775, "learning_rate": 8.281639035517591e-06, "loss": 0.4453, "step": 1240 }, { "epoch": 1.0353170189099, "grad_norm": 0.44201353192329407, "learning_rate": 8.277972899282684e-06, "loss": 0.3921, "step": 1241 }, { "epoch": 1.03615127919911, "grad_norm": 0.39162102341651917, "learning_rate": 8.274303669726427e-06, "loss": 0.4417, "step": 1242 }, { "epoch": 1.0369855394883203, "grad_norm": 0.3699743449687958, "learning_rate": 8.27063135031136e-06, "loss": 0.3955, "step": 1243 }, { "epoch": 1.0378197997775307, "grad_norm": 0.4512883126735687, "learning_rate": 8.266955944502937e-06, "loss": 0.4008, "step": 1244 }, { "epoch": 1.0386540600667409, "grad_norm": 0.3668423593044281, "learning_rate": 8.263277455769525e-06, "loss": 0.4264, "step": 1245 }, { "epoch": 1.039488320355951, "grad_norm": 0.3562275767326355, "learning_rate": 8.259595887582398e-06, "loss": 0.4133, "step": 1246 }, { "epoch": 1.0403225806451613, "grad_norm": 0.4208900034427643, "learning_rate": 8.25591124341574e-06, "loss": 0.4317, "step": 1247 }, { "epoch": 1.0411568409343714, "grad_norm": 0.35908329486846924, "learning_rate": 8.252223526746632e-06, "loss": 0.441, "step": 1248 }, { "epoch": 1.0419911012235819, "grad_norm": 0.41951847076416016, "learning_rate": 8.248532741055061e-06, "loss": 0.4027, "step": 1249 }, { "epoch": 1.042825361512792, "grad_norm": 0.44170820713043213, "learning_rate": 8.244838889823907e-06, "loss": 0.4535, "step": 1250 }, { "epoch": 1.0436596218020022, "grad_norm": 0.35065317153930664, "learning_rate": 8.241141976538944e-06, "loss": 0.4119, "step": 1251 }, { "epoch": 1.0444938820912124, "grad_norm": 0.46862509846687317, "learning_rate": 8.237442004688829e-06, "loss": 0.4797, "step": 1252 }, { "epoch": 1.0453281423804226, "grad_norm": 0.34548699855804443, "learning_rate": 8.233738977765116e-06, "loss": 0.3765, "step": 1253 }, { "epoch": 1.046162402669633, "grad_norm": 0.408844530582428, "learning_rate": 8.230032899262232e-06, "loss": 0.4609, "step": 1254 }, { "epoch": 1.0469966629588432, "grad_norm": 0.49015888571739197, "learning_rate": 8.226323772677492e-06, "loss": 0.462, "step": 1255 }, { "epoch": 1.0478309232480534, "grad_norm": 0.409572571516037, "learning_rate": 8.222611601511084e-06, "loss": 0.4396, "step": 1256 }, { "epoch": 1.0486651835372636, "grad_norm": 0.4696170687675476, "learning_rate": 8.218896389266065e-06, "loss": 0.4342, "step": 1257 }, { "epoch": 1.0494994438264738, "grad_norm": 0.4311964511871338, "learning_rate": 8.21517813944837e-06, "loss": 0.4421, "step": 1258 }, { "epoch": 1.0503337041156842, "grad_norm": 0.4286634624004364, "learning_rate": 8.211456855566793e-06, "loss": 0.4573, "step": 1259 }, { "epoch": 1.0511679644048944, "grad_norm": 0.422341912984848, "learning_rate": 8.207732541132996e-06, "loss": 0.4063, "step": 1260 }, { "epoch": 1.0520022246941045, "grad_norm": 0.38170957565307617, "learning_rate": 8.204005199661498e-06, "loss": 0.4224, "step": 1261 }, { "epoch": 1.0528364849833147, "grad_norm": 0.46123576164245605, "learning_rate": 8.200274834669675e-06, "loss": 0.417, "step": 1262 }, { "epoch": 1.0536707452725251, "grad_norm": 0.4075373411178589, "learning_rate": 8.196541449677758e-06, "loss": 0.4646, "step": 1263 }, { "epoch": 1.0545050055617353, "grad_norm": 0.41948607563972473, "learning_rate": 8.192805048208827e-06, "loss": 0.4084, "step": 1264 }, { "epoch": 1.0553392658509455, "grad_norm": 0.4773457646369934, "learning_rate": 8.189065633788805e-06, "loss": 0.4513, "step": 1265 }, { "epoch": 1.0561735261401557, "grad_norm": 0.3572321832180023, "learning_rate": 8.185323209946464e-06, "loss": 0.4048, "step": 1266 }, { "epoch": 1.0570077864293659, "grad_norm": 0.4075426459312439, "learning_rate": 8.181577780213415e-06, "loss": 0.4318, "step": 1267 }, { "epoch": 1.0578420467185763, "grad_norm": 0.4417472183704376, "learning_rate": 8.1778293481241e-06, "loss": 0.469, "step": 1268 }, { "epoch": 1.0586763070077865, "grad_norm": 0.37923336029052734, "learning_rate": 8.1740779172158e-06, "loss": 0.4057, "step": 1269 }, { "epoch": 1.0595105672969967, "grad_norm": 0.3904128968715668, "learning_rate": 8.170323491028625e-06, "loss": 0.4585, "step": 1270 }, { "epoch": 1.0603448275862069, "grad_norm": 0.4028118848800659, "learning_rate": 8.166566073105507e-06, "loss": 0.39, "step": 1271 }, { "epoch": 1.061179087875417, "grad_norm": 0.3972741663455963, "learning_rate": 8.16280566699221e-06, "loss": 0.4088, "step": 1272 }, { "epoch": 1.0620133481646274, "grad_norm": 0.4296902120113373, "learning_rate": 8.159042276237308e-06, "loss": 0.4687, "step": 1273 }, { "epoch": 1.0628476084538376, "grad_norm": 0.33471497893333435, "learning_rate": 8.155275904392197e-06, "loss": 0.3979, "step": 1274 }, { "epoch": 1.0636818687430478, "grad_norm": 0.39785945415496826, "learning_rate": 8.151506555011088e-06, "loss": 0.3928, "step": 1275 }, { "epoch": 1.064516129032258, "grad_norm": 0.45916348695755005, "learning_rate": 8.147734231650997e-06, "loss": 0.4437, "step": 1276 }, { "epoch": 1.0653503893214682, "grad_norm": 0.41307276487350464, "learning_rate": 8.143958937871748e-06, "loss": 0.4178, "step": 1277 }, { "epoch": 1.0661846496106786, "grad_norm": 0.43552249670028687, "learning_rate": 8.14018067723597e-06, "loss": 0.4457, "step": 1278 }, { "epoch": 1.0670189098998888, "grad_norm": 0.4023208022117615, "learning_rate": 8.13639945330909e-06, "loss": 0.4429, "step": 1279 }, { "epoch": 1.067853170189099, "grad_norm": 0.3628684878349304, "learning_rate": 8.132615269659329e-06, "loss": 0.4407, "step": 1280 }, { "epoch": 1.0686874304783092, "grad_norm": 0.46344175934791565, "learning_rate": 8.128828129857709e-06, "loss": 0.4768, "step": 1281 }, { "epoch": 1.0695216907675196, "grad_norm": 0.34299543499946594, "learning_rate": 8.125038037478033e-06, "loss": 0.4184, "step": 1282 }, { "epoch": 1.0703559510567298, "grad_norm": 0.4771042764186859, "learning_rate": 8.121244996096895e-06, "loss": 0.4913, "step": 1283 }, { "epoch": 1.07119021134594, "grad_norm": 0.3174145817756653, "learning_rate": 8.117449009293668e-06, "loss": 0.3952, "step": 1284 }, { "epoch": 1.0720244716351501, "grad_norm": 0.41328901052474976, "learning_rate": 8.11365008065051e-06, "loss": 0.4338, "step": 1285 }, { "epoch": 1.0728587319243603, "grad_norm": 0.45783644914627075, "learning_rate": 8.109848213752352e-06, "loss": 0.4763, "step": 1286 }, { "epoch": 1.0736929922135707, "grad_norm": 0.35163241624832153, "learning_rate": 8.106043412186894e-06, "loss": 0.4073, "step": 1287 }, { "epoch": 1.074527252502781, "grad_norm": 0.4496193528175354, "learning_rate": 8.102235679544612e-06, "loss": 0.4781, "step": 1288 }, { "epoch": 1.075361512791991, "grad_norm": 0.31854525208473206, "learning_rate": 8.098425019418746e-06, "loss": 0.4131, "step": 1289 }, { "epoch": 1.0761957730812013, "grad_norm": 0.4634130597114563, "learning_rate": 8.094611435405298e-06, "loss": 0.4857, "step": 1290 }, { "epoch": 1.0770300333704115, "grad_norm": 0.3165237605571747, "learning_rate": 8.090794931103026e-06, "loss": 0.3919, "step": 1291 }, { "epoch": 1.0778642936596219, "grad_norm": 0.38112467527389526, "learning_rate": 8.08697551011345e-06, "loss": 0.4677, "step": 1292 }, { "epoch": 1.078698553948832, "grad_norm": 0.4161386489868164, "learning_rate": 8.083153176040838e-06, "loss": 0.3982, "step": 1293 }, { "epoch": 1.0795328142380423, "grad_norm": 0.41562557220458984, "learning_rate": 8.079327932492207e-06, "loss": 0.4704, "step": 1294 }, { "epoch": 1.0803670745272524, "grad_norm": 0.4158051311969757, "learning_rate": 8.075499783077321e-06, "loss": 0.4541, "step": 1295 }, { "epoch": 1.0812013348164626, "grad_norm": 0.3613758087158203, "learning_rate": 8.071668731408686e-06, "loss": 0.4027, "step": 1296 }, { "epoch": 1.082035595105673, "grad_norm": 0.4970267117023468, "learning_rate": 8.067834781101548e-06, "loss": 0.4726, "step": 1297 }, { "epoch": 1.0828698553948832, "grad_norm": 0.3782969117164612, "learning_rate": 8.063997935773885e-06, "loss": 0.4046, "step": 1298 }, { "epoch": 1.0837041156840934, "grad_norm": 0.416192889213562, "learning_rate": 8.06015819904641e-06, "loss": 0.4569, "step": 1299 }, { "epoch": 1.0845383759733036, "grad_norm": 0.46574676036834717, "learning_rate": 8.056315574542564e-06, "loss": 0.4446, "step": 1300 }, { "epoch": 1.0853726362625138, "grad_norm": 0.42579254508018494, "learning_rate": 8.052470065888512e-06, "loss": 0.3926, "step": 1301 }, { "epoch": 1.0862068965517242, "grad_norm": 0.3901159465312958, "learning_rate": 8.04862167671314e-06, "loss": 0.4465, "step": 1302 }, { "epoch": 1.0870411568409344, "grad_norm": 0.3968401551246643, "learning_rate": 8.044770410648055e-06, "loss": 0.4577, "step": 1303 }, { "epoch": 1.0878754171301446, "grad_norm": 0.3455469608306885, "learning_rate": 8.040916271327577e-06, "loss": 0.405, "step": 1304 }, { "epoch": 1.0887096774193548, "grad_norm": 0.3721373975276947, "learning_rate": 8.03705926238874e-06, "loss": 0.4303, "step": 1305 }, { "epoch": 1.0895439377085652, "grad_norm": 0.39610087871551514, "learning_rate": 8.033199387471278e-06, "loss": 0.4209, "step": 1306 }, { "epoch": 1.0903781979977754, "grad_norm": 0.3774673342704773, "learning_rate": 8.029336650217638e-06, "loss": 0.476, "step": 1307 }, { "epoch": 1.0912124582869855, "grad_norm": 0.3908822536468506, "learning_rate": 8.025471054272969e-06, "loss": 0.4123, "step": 1308 }, { "epoch": 1.0920467185761957, "grad_norm": 0.3826358914375305, "learning_rate": 8.021602603285109e-06, "loss": 0.4093, "step": 1309 }, { "epoch": 1.092880978865406, "grad_norm": 0.3866932690143585, "learning_rate": 8.017731300904596e-06, "loss": 0.4293, "step": 1310 }, { "epoch": 1.0937152391546163, "grad_norm": 0.3523010313510895, "learning_rate": 8.01385715078466e-06, "loss": 0.4314, "step": 1311 }, { "epoch": 1.0945494994438265, "grad_norm": 0.3997926712036133, "learning_rate": 8.009980156581218e-06, "loss": 0.4493, "step": 1312 }, { "epoch": 1.0953837597330367, "grad_norm": 0.41041526198387146, "learning_rate": 8.006100321952864e-06, "loss": 0.4619, "step": 1313 }, { "epoch": 1.0962180200222469, "grad_norm": 0.341134250164032, "learning_rate": 8.002217650560881e-06, "loss": 0.3911, "step": 1314 }, { "epoch": 1.097052280311457, "grad_norm": 0.37675920128822327, "learning_rate": 7.998332146069224e-06, "loss": 0.4423, "step": 1315 }, { "epoch": 1.0978865406006675, "grad_norm": 0.44129353761672974, "learning_rate": 7.994443812144523e-06, "loss": 0.4812, "step": 1316 }, { "epoch": 1.0987208008898777, "grad_norm": 0.37104332447052, "learning_rate": 7.99055265245608e-06, "loss": 0.415, "step": 1317 }, { "epoch": 1.0995550611790879, "grad_norm": 0.40260326862335205, "learning_rate": 7.986658670675862e-06, "loss": 0.4038, "step": 1318 }, { "epoch": 1.100389321468298, "grad_norm": 0.3963145315647125, "learning_rate": 7.982761870478495e-06, "loss": 0.4771, "step": 1319 }, { "epoch": 1.1012235817575085, "grad_norm": 0.3831017017364502, "learning_rate": 7.97886225554127e-06, "loss": 0.4098, "step": 1320 }, { "epoch": 1.1020578420467186, "grad_norm": 0.33798912167549133, "learning_rate": 7.974959829544136e-06, "loss": 0.3894, "step": 1321 }, { "epoch": 1.1028921023359288, "grad_norm": 0.42702966928482056, "learning_rate": 7.971054596169688e-06, "loss": 0.4501, "step": 1322 }, { "epoch": 1.103726362625139, "grad_norm": 0.4006383419036865, "learning_rate": 7.967146559103177e-06, "loss": 0.4274, "step": 1323 }, { "epoch": 1.1045606229143492, "grad_norm": 0.3867168128490448, "learning_rate": 7.96323572203249e-06, "loss": 0.4162, "step": 1324 }, { "epoch": 1.1053948832035596, "grad_norm": 0.41627344489097595, "learning_rate": 7.959322088648166e-06, "loss": 0.4059, "step": 1325 }, { "epoch": 1.1062291434927698, "grad_norm": 0.4176735281944275, "learning_rate": 7.955405662643384e-06, "loss": 0.4457, "step": 1326 }, { "epoch": 1.10706340378198, "grad_norm": 0.3343712389469147, "learning_rate": 7.951486447713947e-06, "loss": 0.3982, "step": 1327 }, { "epoch": 1.1078976640711902, "grad_norm": 0.42032691836357117, "learning_rate": 7.9475644475583e-06, "loss": 0.4415, "step": 1328 }, { "epoch": 1.1087319243604004, "grad_norm": 0.39376381039619446, "learning_rate": 7.943639665877512e-06, "loss": 0.4554, "step": 1329 }, { "epoch": 1.1095661846496108, "grad_norm": 0.42125722765922546, "learning_rate": 7.939712106375277e-06, "loss": 0.4431, "step": 1330 }, { "epoch": 1.110400444938821, "grad_norm": 0.3674773871898651, "learning_rate": 7.935781772757913e-06, "loss": 0.3889, "step": 1331 }, { "epoch": 1.1112347052280311, "grad_norm": 0.36384931206703186, "learning_rate": 7.931848668734352e-06, "loss": 0.4689, "step": 1332 }, { "epoch": 1.1120689655172413, "grad_norm": 0.38162893056869507, "learning_rate": 7.927912798016144e-06, "loss": 0.4448, "step": 1333 }, { "epoch": 1.1129032258064515, "grad_norm": 0.41330212354660034, "learning_rate": 7.923974164317447e-06, "loss": 0.4221, "step": 1334 }, { "epoch": 1.113737486095662, "grad_norm": 0.46241480112075806, "learning_rate": 7.920032771355027e-06, "loss": 0.4505, "step": 1335 }, { "epoch": 1.114571746384872, "grad_norm": 0.4042946398258209, "learning_rate": 7.916088622848255e-06, "loss": 0.4172, "step": 1336 }, { "epoch": 1.1154060066740823, "grad_norm": 0.4135844111442566, "learning_rate": 7.912141722519103e-06, "loss": 0.4523, "step": 1337 }, { "epoch": 1.1162402669632925, "grad_norm": 0.44458505511283875, "learning_rate": 7.908192074092137e-06, "loss": 0.4266, "step": 1338 }, { "epoch": 1.1170745272525027, "grad_norm": 0.4653661847114563, "learning_rate": 7.904239681294515e-06, "loss": 0.4146, "step": 1339 }, { "epoch": 1.117908787541713, "grad_norm": 0.42476001381874084, "learning_rate": 7.900284547855992e-06, "loss": 0.4555, "step": 1340 }, { "epoch": 1.1187430478309233, "grad_norm": 0.4347173273563385, "learning_rate": 7.896326677508904e-06, "loss": 0.4315, "step": 1341 }, { "epoch": 1.1195773081201335, "grad_norm": 0.41023334860801697, "learning_rate": 7.892366073988167e-06, "loss": 0.4264, "step": 1342 }, { "epoch": 1.1204115684093436, "grad_norm": 0.42029741406440735, "learning_rate": 7.888402741031283e-06, "loss": 0.4395, "step": 1343 }, { "epoch": 1.121245828698554, "grad_norm": 0.4174759089946747, "learning_rate": 7.884436682378327e-06, "loss": 0.4017, "step": 1344 }, { "epoch": 1.1220800889877642, "grad_norm": 0.3997317850589752, "learning_rate": 7.880467901771941e-06, "loss": 0.4128, "step": 1345 }, { "epoch": 1.1229143492769744, "grad_norm": 0.3978201746940613, "learning_rate": 7.876496402957345e-06, "loss": 0.4477, "step": 1346 }, { "epoch": 1.1237486095661846, "grad_norm": 0.45930224657058716, "learning_rate": 7.872522189682318e-06, "loss": 0.4383, "step": 1347 }, { "epoch": 1.1245828698553948, "grad_norm": 0.41279423236846924, "learning_rate": 7.8685452656972e-06, "loss": 0.4383, "step": 1348 }, { "epoch": 1.1254171301446052, "grad_norm": 0.4058186113834381, "learning_rate": 7.864565634754893e-06, "loss": 0.4582, "step": 1349 }, { "epoch": 1.1262513904338154, "grad_norm": 0.423031747341156, "learning_rate": 7.860583300610849e-06, "loss": 0.4347, "step": 1350 }, { "epoch": 1.1270856507230256, "grad_norm": 0.37929490208625793, "learning_rate": 7.856598267023074e-06, "loss": 0.3817, "step": 1351 }, { "epoch": 1.1279199110122358, "grad_norm": 0.4737723171710968, "learning_rate": 7.852610537752123e-06, "loss": 0.503, "step": 1352 }, { "epoch": 1.1287541713014462, "grad_norm": 0.3358587622642517, "learning_rate": 7.84862011656109e-06, "loss": 0.3649, "step": 1353 }, { "epoch": 1.1295884315906564, "grad_norm": 0.43736356496810913, "learning_rate": 7.844627007215613e-06, "loss": 0.4565, "step": 1354 }, { "epoch": 1.1304226918798665, "grad_norm": 0.4603974223136902, "learning_rate": 7.840631213483862e-06, "loss": 0.448, "step": 1355 }, { "epoch": 1.1312569521690767, "grad_norm": 0.4198083281517029, "learning_rate": 7.83663273913655e-06, "loss": 0.4267, "step": 1356 }, { "epoch": 1.132091212458287, "grad_norm": 0.4083608388900757, "learning_rate": 7.832631587946905e-06, "loss": 0.4562, "step": 1357 }, { "epoch": 1.1329254727474973, "grad_norm": 0.4182080924510956, "learning_rate": 7.828627763690698e-06, "loss": 0.4214, "step": 1358 }, { "epoch": 1.1337597330367075, "grad_norm": 0.38977938890457153, "learning_rate": 7.824621270146207e-06, "loss": 0.4028, "step": 1359 }, { "epoch": 1.1345939933259177, "grad_norm": 0.44123324751853943, "learning_rate": 7.820612111094238e-06, "loss": 0.4968, "step": 1360 }, { "epoch": 1.135428253615128, "grad_norm": 0.3764444887638092, "learning_rate": 7.81660029031811e-06, "loss": 0.4149, "step": 1361 }, { "epoch": 1.136262513904338, "grad_norm": 0.4947524964809418, "learning_rate": 7.812585811603656e-06, "loss": 0.451, "step": 1362 }, { "epoch": 1.1370967741935485, "grad_norm": 0.40678539872169495, "learning_rate": 7.808568678739213e-06, "loss": 0.4364, "step": 1363 }, { "epoch": 1.1379310344827587, "grad_norm": 0.43148964643478394, "learning_rate": 7.804548895515622e-06, "loss": 0.4054, "step": 1364 }, { "epoch": 1.1387652947719689, "grad_norm": 0.48852941393852234, "learning_rate": 7.800526465726233e-06, "loss": 0.4378, "step": 1365 }, { "epoch": 1.139599555061179, "grad_norm": 0.3753541111946106, "learning_rate": 7.796501393166885e-06, "loss": 0.4425, "step": 1366 }, { "epoch": 1.1404338153503892, "grad_norm": 0.4130557179450989, "learning_rate": 7.792473681635913e-06, "loss": 0.4231, "step": 1367 }, { "epoch": 1.1412680756395996, "grad_norm": 0.37124696373939514, "learning_rate": 7.788443334934148e-06, "loss": 0.4056, "step": 1368 }, { "epoch": 1.1421023359288098, "grad_norm": 0.40169060230255127, "learning_rate": 7.7844103568649e-06, "loss": 0.4071, "step": 1369 }, { "epoch": 1.14293659621802, "grad_norm": 0.438406765460968, "learning_rate": 7.780374751233965e-06, "loss": 0.4922, "step": 1370 }, { "epoch": 1.1437708565072302, "grad_norm": 0.35231438279151917, "learning_rate": 7.776336521849618e-06, "loss": 0.4306, "step": 1371 }, { "epoch": 1.1446051167964404, "grad_norm": 0.39908567070961, "learning_rate": 7.772295672522615e-06, "loss": 0.3981, "step": 1372 }, { "epoch": 1.1454393770856508, "grad_norm": 0.3392985761165619, "learning_rate": 7.768252207066178e-06, "loss": 0.3842, "step": 1373 }, { "epoch": 1.146273637374861, "grad_norm": 0.35532334446907043, "learning_rate": 7.764206129295997e-06, "loss": 0.4212, "step": 1374 }, { "epoch": 1.1471078976640712, "grad_norm": 0.3943493366241455, "learning_rate": 7.760157443030234e-06, "loss": 0.4623, "step": 1375 }, { "epoch": 1.1479421579532814, "grad_norm": 0.37308013439178467, "learning_rate": 7.756106152089507e-06, "loss": 0.4619, "step": 1376 }, { "epoch": 1.1487764182424915, "grad_norm": 0.32562676072120667, "learning_rate": 7.752052260296894e-06, "loss": 0.3878, "step": 1377 }, { "epoch": 1.149610678531702, "grad_norm": 0.4006744623184204, "learning_rate": 7.747995771477928e-06, "loss": 0.4721, "step": 1378 }, { "epoch": 1.1504449388209121, "grad_norm": 0.37974750995635986, "learning_rate": 7.743936689460588e-06, "loss": 0.4314, "step": 1379 }, { "epoch": 1.1512791991101223, "grad_norm": 0.3907822370529175, "learning_rate": 7.739875018075307e-06, "loss": 0.4377, "step": 1380 }, { "epoch": 1.1521134593993325, "grad_norm": 0.338486909866333, "learning_rate": 7.735810761154959e-06, "loss": 0.4191, "step": 1381 }, { "epoch": 1.1529477196885427, "grad_norm": 0.3839567303657532, "learning_rate": 7.731743922534854e-06, "loss": 0.4343, "step": 1382 }, { "epoch": 1.1537819799777531, "grad_norm": 0.3401148319244385, "learning_rate": 7.727674506052744e-06, "loss": 0.4004, "step": 1383 }, { "epoch": 1.1546162402669633, "grad_norm": 0.36566048860549927, "learning_rate": 7.72360251554881e-06, "loss": 0.4507, "step": 1384 }, { "epoch": 1.1554505005561735, "grad_norm": 0.364854633808136, "learning_rate": 7.719527954865664e-06, "loss": 0.4051, "step": 1385 }, { "epoch": 1.1562847608453837, "grad_norm": 0.36764800548553467, "learning_rate": 7.715450827848341e-06, "loss": 0.422, "step": 1386 }, { "epoch": 1.157119021134594, "grad_norm": 0.3927608132362366, "learning_rate": 7.711371138344302e-06, "loss": 0.4237, "step": 1387 }, { "epoch": 1.1579532814238043, "grad_norm": 0.4566560983657837, "learning_rate": 7.70728889020342e-06, "loss": 0.4235, "step": 1388 }, { "epoch": 1.1587875417130145, "grad_norm": 0.36245205998420715, "learning_rate": 7.703204087277989e-06, "loss": 0.3851, "step": 1389 }, { "epoch": 1.1596218020022246, "grad_norm": 0.38626980781555176, "learning_rate": 7.69911673342271e-06, "loss": 0.4201, "step": 1390 }, { "epoch": 1.1604560622914348, "grad_norm": 0.36012300848960876, "learning_rate": 7.695026832494688e-06, "loss": 0.4413, "step": 1391 }, { "epoch": 1.1612903225806452, "grad_norm": 0.4370402693748474, "learning_rate": 7.690934388353442e-06, "loss": 0.4649, "step": 1392 }, { "epoch": 1.1621245828698554, "grad_norm": 0.4441777765750885, "learning_rate": 7.686839404860882e-06, "loss": 0.4541, "step": 1393 }, { "epoch": 1.1629588431590656, "grad_norm": 0.3619338572025299, "learning_rate": 7.682741885881314e-06, "loss": 0.4541, "step": 1394 }, { "epoch": 1.1637931034482758, "grad_norm": 0.39327967166900635, "learning_rate": 7.678641835281444e-06, "loss": 0.3942, "step": 1395 }, { "epoch": 1.1646273637374862, "grad_norm": 0.37201985716819763, "learning_rate": 7.674539256930364e-06, "loss": 0.4325, "step": 1396 }, { "epoch": 1.1654616240266964, "grad_norm": 0.3702593445777893, "learning_rate": 7.670434154699543e-06, "loss": 0.4122, "step": 1397 }, { "epoch": 1.1662958843159066, "grad_norm": 0.43530359864234924, "learning_rate": 7.666326532462842e-06, "loss": 0.4618, "step": 1398 }, { "epoch": 1.1671301446051168, "grad_norm": 0.382519006729126, "learning_rate": 7.662216394096498e-06, "loss": 0.4467, "step": 1399 }, { "epoch": 1.167964404894327, "grad_norm": 0.4178166687488556, "learning_rate": 7.658103743479121e-06, "loss": 0.423, "step": 1400 }, { "epoch": 1.1687986651835374, "grad_norm": 0.3858039379119873, "learning_rate": 7.653988584491692e-06, "loss": 0.4339, "step": 1401 }, { "epoch": 1.1696329254727476, "grad_norm": 0.3657691478729248, "learning_rate": 7.649870921017554e-06, "loss": 0.435, "step": 1402 }, { "epoch": 1.1704671857619577, "grad_norm": 0.3398030400276184, "learning_rate": 7.645750756942425e-06, "loss": 0.4418, "step": 1403 }, { "epoch": 1.171301446051168, "grad_norm": 0.4129176139831543, "learning_rate": 7.641628096154372e-06, "loss": 0.4218, "step": 1404 }, { "epoch": 1.1721357063403781, "grad_norm": 0.3641882836818695, "learning_rate": 7.637502942543825e-06, "loss": 0.4015, "step": 1405 }, { "epoch": 1.1729699666295885, "grad_norm": 0.3923196792602539, "learning_rate": 7.633375300003561e-06, "loss": 0.4807, "step": 1406 }, { "epoch": 1.1738042269187987, "grad_norm": 0.33696919679641724, "learning_rate": 7.6292451724287085e-06, "loss": 0.4038, "step": 1407 }, { "epoch": 1.174638487208009, "grad_norm": 0.3890690803527832, "learning_rate": 7.625112563716742e-06, "loss": 0.4353, "step": 1408 }, { "epoch": 1.175472747497219, "grad_norm": 0.3762977719306946, "learning_rate": 7.620977477767475e-06, "loss": 0.4662, "step": 1409 }, { "epoch": 1.1763070077864293, "grad_norm": 0.37485069036483765, "learning_rate": 7.616839918483061e-06, "loss": 0.4411, "step": 1410 }, { "epoch": 1.1771412680756397, "grad_norm": 0.36472249031066895, "learning_rate": 7.612699889767985e-06, "loss": 0.4386, "step": 1411 }, { "epoch": 1.1779755283648499, "grad_norm": 0.3695392906665802, "learning_rate": 7.608557395529064e-06, "loss": 0.4273, "step": 1412 }, { "epoch": 1.17880978865406, "grad_norm": 0.4389885365962982, "learning_rate": 7.604412439675444e-06, "loss": 0.4653, "step": 1413 }, { "epoch": 1.1796440489432702, "grad_norm": 0.3266173005104065, "learning_rate": 7.600265026118587e-06, "loss": 0.3963, "step": 1414 }, { "epoch": 1.1804783092324804, "grad_norm": 0.3468549847602844, "learning_rate": 7.596115158772281e-06, "loss": 0.3966, "step": 1415 }, { "epoch": 1.1813125695216908, "grad_norm": 0.3710923194885254, "learning_rate": 7.591962841552627e-06, "loss": 0.426, "step": 1416 }, { "epoch": 1.182146829810901, "grad_norm": 0.35736507177352905, "learning_rate": 7.587808078378036e-06, "loss": 0.3869, "step": 1417 }, { "epoch": 1.1829810901001112, "grad_norm": 0.39700913429260254, "learning_rate": 7.583650873169233e-06, "loss": 0.491, "step": 1418 }, { "epoch": 1.1838153503893214, "grad_norm": 0.32892197370529175, "learning_rate": 7.579491229849238e-06, "loss": 0.4023, "step": 1419 }, { "epoch": 1.1846496106785316, "grad_norm": 0.4308750629425049, "learning_rate": 7.575329152343383e-06, "loss": 0.4436, "step": 1420 }, { "epoch": 1.185483870967742, "grad_norm": 0.4100548028945923, "learning_rate": 7.57116464457929e-06, "loss": 0.4677, "step": 1421 }, { "epoch": 1.1863181312569522, "grad_norm": 0.368277907371521, "learning_rate": 7.566997710486873e-06, "loss": 0.4475, "step": 1422 }, { "epoch": 1.1871523915461624, "grad_norm": 0.38459235429763794, "learning_rate": 7.5628283539983415e-06, "loss": 0.4, "step": 1423 }, { "epoch": 1.1879866518353726, "grad_norm": 0.3957088887691498, "learning_rate": 7.5586565790481855e-06, "loss": 0.4367, "step": 1424 }, { "epoch": 1.188820912124583, "grad_norm": 0.3620186150074005, "learning_rate": 7.554482389573182e-06, "loss": 0.4286, "step": 1425 }, { "epoch": 1.1896551724137931, "grad_norm": 0.4163001477718353, "learning_rate": 7.550305789512385e-06, "loss": 0.5157, "step": 1426 }, { "epoch": 1.1904894327030033, "grad_norm": 0.3257404863834381, "learning_rate": 7.546126782807117e-06, "loss": 0.3881, "step": 1427 }, { "epoch": 1.1913236929922135, "grad_norm": 0.38778600096702576, "learning_rate": 7.541945373400981e-06, "loss": 0.4438, "step": 1428 }, { "epoch": 1.1921579532814237, "grad_norm": 0.3738137483596802, "learning_rate": 7.537761565239844e-06, "loss": 0.4474, "step": 1429 }, { "epoch": 1.1929922135706341, "grad_norm": 0.3332713544368744, "learning_rate": 7.533575362271834e-06, "loss": 0.4218, "step": 1430 }, { "epoch": 1.1938264738598443, "grad_norm": 0.36658287048339844, "learning_rate": 7.529386768447342e-06, "loss": 0.388, "step": 1431 }, { "epoch": 1.1946607341490545, "grad_norm": 0.47756433486938477, "learning_rate": 7.525195787719013e-06, "loss": 0.4596, "step": 1432 }, { "epoch": 1.1954949944382647, "grad_norm": 0.3945799171924591, "learning_rate": 7.521002424041748e-06, "loss": 0.4072, "step": 1433 }, { "epoch": 1.196329254727475, "grad_norm": 0.38227227330207825, "learning_rate": 7.51680668137269e-06, "loss": 0.4356, "step": 1434 }, { "epoch": 1.1971635150166853, "grad_norm": 0.4237688183784485, "learning_rate": 7.512608563671237e-06, "loss": 0.4026, "step": 1435 }, { "epoch": 1.1979977753058955, "grad_norm": 0.4947100281715393, "learning_rate": 7.5084080748990165e-06, "loss": 0.4483, "step": 1436 }, { "epoch": 1.1988320355951056, "grad_norm": 0.37258628010749817, "learning_rate": 7.504205219019904e-06, "loss": 0.4136, "step": 1437 }, { "epoch": 1.1996662958843158, "grad_norm": 0.33934515714645386, "learning_rate": 7.500000000000001e-06, "loss": 0.4116, "step": 1438 }, { "epoch": 1.2005005561735262, "grad_norm": 0.5108917355537415, "learning_rate": 7.495792421807644e-06, "loss": 0.4308, "step": 1439 }, { "epoch": 1.2013348164627364, "grad_norm": 0.4209623634815216, "learning_rate": 7.491582488413394e-06, "loss": 0.4015, "step": 1440 }, { "epoch": 1.2021690767519466, "grad_norm": 0.38059648871421814, "learning_rate": 7.487370203790033e-06, "loss": 0.4099, "step": 1441 }, { "epoch": 1.2030033370411568, "grad_norm": 0.4286749064922333, "learning_rate": 7.483155571912566e-06, "loss": 0.4277, "step": 1442 }, { "epoch": 1.203837597330367, "grad_norm": 0.41874897480010986, "learning_rate": 7.4789385967582075e-06, "loss": 0.4388, "step": 1443 }, { "epoch": 1.2046718576195774, "grad_norm": 0.4415771961212158, "learning_rate": 7.474719282306389e-06, "loss": 0.4306, "step": 1444 }, { "epoch": 1.2055061179087876, "grad_norm": 0.38208767771720886, "learning_rate": 7.470497632538743e-06, "loss": 0.4227, "step": 1445 }, { "epoch": 1.2063403781979978, "grad_norm": 0.35602617263793945, "learning_rate": 7.466273651439114e-06, "loss": 0.4181, "step": 1446 }, { "epoch": 1.207174638487208, "grad_norm": 0.38737812638282776, "learning_rate": 7.46204734299354e-06, "loss": 0.4522, "step": 1447 }, { "epoch": 1.2080088987764181, "grad_norm": 0.3424398899078369, "learning_rate": 7.457818711190255e-06, "loss": 0.43, "step": 1448 }, { "epoch": 1.2088431590656286, "grad_norm": 0.37164491415023804, "learning_rate": 7.453587760019691e-06, "loss": 0.4163, "step": 1449 }, { "epoch": 1.2096774193548387, "grad_norm": 0.35694190859794617, "learning_rate": 7.449354493474462e-06, "loss": 0.4035, "step": 1450 }, { "epoch": 1.210511679644049, "grad_norm": 0.38637179136276245, "learning_rate": 7.445118915549374e-06, "loss": 0.4629, "step": 1451 }, { "epoch": 1.2113459399332591, "grad_norm": 0.3525828719139099, "learning_rate": 7.440881030241407e-06, "loss": 0.4487, "step": 1452 }, { "epoch": 1.2121802002224693, "grad_norm": 0.3612954914569855, "learning_rate": 7.436640841549724e-06, "loss": 0.4593, "step": 1453 }, { "epoch": 1.2130144605116797, "grad_norm": 0.3518927991390228, "learning_rate": 7.43239835347566e-06, "loss": 0.4158, "step": 1454 }, { "epoch": 1.21384872080089, "grad_norm": 0.385416179895401, "learning_rate": 7.4281535700227185e-06, "loss": 0.4792, "step": 1455 }, { "epoch": 1.2146829810901, "grad_norm": 0.33973032236099243, "learning_rate": 7.423906495196569e-06, "loss": 0.3838, "step": 1456 }, { "epoch": 1.2155172413793103, "grad_norm": 0.32619166374206543, "learning_rate": 7.419657133005047e-06, "loss": 0.4536, "step": 1457 }, { "epoch": 1.2163515016685205, "grad_norm": 0.3492060601711273, "learning_rate": 7.41540548745814e-06, "loss": 0.4412, "step": 1458 }, { "epoch": 1.2171857619577309, "grad_norm": 0.3633149564266205, "learning_rate": 7.411151562567999e-06, "loss": 0.4379, "step": 1459 }, { "epoch": 1.218020022246941, "grad_norm": 0.3724212944507599, "learning_rate": 7.406895362348916e-06, "loss": 0.4602, "step": 1460 }, { "epoch": 1.2188542825361512, "grad_norm": 0.34466788172721863, "learning_rate": 7.4026368908173375e-06, "loss": 0.4141, "step": 1461 }, { "epoch": 1.2196885428253614, "grad_norm": 0.32960352301597595, "learning_rate": 7.39837615199185e-06, "loss": 0.442, "step": 1462 }, { "epoch": 1.2205228031145718, "grad_norm": 0.3426360487937927, "learning_rate": 7.394113149893182e-06, "loss": 0.4298, "step": 1463 }, { "epoch": 1.221357063403782, "grad_norm": 0.37420663237571716, "learning_rate": 7.389847888544196e-06, "loss": 0.4549, "step": 1464 }, { "epoch": 1.2221913236929922, "grad_norm": 0.36695924401283264, "learning_rate": 7.385580371969886e-06, "loss": 0.4326, "step": 1465 }, { "epoch": 1.2230255839822024, "grad_norm": 0.3566995859146118, "learning_rate": 7.381310604197375e-06, "loss": 0.4217, "step": 1466 }, { "epoch": 1.2238598442714126, "grad_norm": 0.364687979221344, "learning_rate": 7.3770385892559115e-06, "loss": 0.4207, "step": 1467 }, { "epoch": 1.224694104560623, "grad_norm": 0.4064037799835205, "learning_rate": 7.372764331176864e-06, "loss": 0.4492, "step": 1468 }, { "epoch": 1.2255283648498332, "grad_norm": 0.3697409927845001, "learning_rate": 7.3684878339937146e-06, "loss": 0.4425, "step": 1469 }, { "epoch": 1.2263626251390434, "grad_norm": 0.36389780044555664, "learning_rate": 7.364209101742062e-06, "loss": 0.4, "step": 1470 }, { "epoch": 1.2271968854282536, "grad_norm": 0.41046226024627686, "learning_rate": 7.359928138459615e-06, "loss": 0.4412, "step": 1471 }, { "epoch": 1.228031145717464, "grad_norm": 0.3425498604774475, "learning_rate": 7.355644948186185e-06, "loss": 0.3958, "step": 1472 }, { "epoch": 1.2288654060066742, "grad_norm": 0.36461883783340454, "learning_rate": 7.351359534963684e-06, "loss": 0.4335, "step": 1473 }, { "epoch": 1.2296996662958843, "grad_norm": 0.3706044852733612, "learning_rate": 7.347071902836127e-06, "loss": 0.4366, "step": 1474 }, { "epoch": 1.2305339265850945, "grad_norm": 0.38586944341659546, "learning_rate": 7.342782055849615e-06, "loss": 0.4361, "step": 1475 }, { "epoch": 1.2313681868743047, "grad_norm": 0.37868639826774597, "learning_rate": 7.338489998052347e-06, "loss": 0.4235, "step": 1476 }, { "epoch": 1.2322024471635151, "grad_norm": 0.39740055799484253, "learning_rate": 7.334195733494603e-06, "loss": 0.4506, "step": 1477 }, { "epoch": 1.2330367074527253, "grad_norm": 0.3991326689720154, "learning_rate": 7.3298992662287484e-06, "loss": 0.4452, "step": 1478 }, { "epoch": 1.2338709677419355, "grad_norm": 0.35111314058303833, "learning_rate": 7.325600600309226e-06, "loss": 0.4677, "step": 1479 }, { "epoch": 1.2347052280311457, "grad_norm": 0.3459174335002899, "learning_rate": 7.321299739792553e-06, "loss": 0.3901, "step": 1480 }, { "epoch": 1.2355394883203559, "grad_norm": 0.44568493962287903, "learning_rate": 7.316996688737319e-06, "loss": 0.425, "step": 1481 }, { "epoch": 1.2363737486095663, "grad_norm": 0.3591909110546112, "learning_rate": 7.312691451204178e-06, "loss": 0.4091, "step": 1482 }, { "epoch": 1.2372080088987765, "grad_norm": 0.3482624292373657, "learning_rate": 7.308384031255848e-06, "loss": 0.4547, "step": 1483 }, { "epoch": 1.2380422691879867, "grad_norm": 0.351936936378479, "learning_rate": 7.304074432957113e-06, "loss": 0.3762, "step": 1484 }, { "epoch": 1.2388765294771968, "grad_norm": 0.43568119406700134, "learning_rate": 7.299762660374803e-06, "loss": 0.5, "step": 1485 }, { "epoch": 1.239710789766407, "grad_norm": 0.33202552795410156, "learning_rate": 7.2954487175778044e-06, "loss": 0.4174, "step": 1486 }, { "epoch": 1.2405450500556174, "grad_norm": 0.396951287984848, "learning_rate": 7.291132608637053e-06, "loss": 0.4711, "step": 1487 }, { "epoch": 1.2413793103448276, "grad_norm": 0.38111114501953125, "learning_rate": 7.2868143376255254e-06, "loss": 0.4187, "step": 1488 }, { "epoch": 1.2422135706340378, "grad_norm": 0.3765943944454193, "learning_rate": 7.28249390861824e-06, "loss": 0.4389, "step": 1489 }, { "epoch": 1.243047830923248, "grad_norm": 0.37912416458129883, "learning_rate": 7.278171325692251e-06, "loss": 0.4554, "step": 1490 }, { "epoch": 1.2438820912124582, "grad_norm": 0.4324364960193634, "learning_rate": 7.2738465929266474e-06, "loss": 0.4196, "step": 1491 }, { "epoch": 1.2447163515016686, "grad_norm": 0.38898253440856934, "learning_rate": 7.269519714402545e-06, "loss": 0.4189, "step": 1492 }, { "epoch": 1.2455506117908788, "grad_norm": 0.41306301951408386, "learning_rate": 7.265190694203086e-06, "loss": 0.4616, "step": 1493 }, { "epoch": 1.246384872080089, "grad_norm": 0.375933438539505, "learning_rate": 7.260859536413429e-06, "loss": 0.435, "step": 1494 }, { "epoch": 1.2472191323692992, "grad_norm": 0.4032117426395416, "learning_rate": 7.2565262451207565e-06, "loss": 0.4384, "step": 1495 }, { "epoch": 1.2480533926585093, "grad_norm": 0.40048345923423767, "learning_rate": 7.252190824414259e-06, "loss": 0.3981, "step": 1496 }, { "epoch": 1.2488876529477198, "grad_norm": 0.5075517296791077, "learning_rate": 7.247853278385141e-06, "loss": 0.4611, "step": 1497 }, { "epoch": 1.24972191323693, "grad_norm": 0.36956506967544556, "learning_rate": 7.243513611126608e-06, "loss": 0.4064, "step": 1498 }, { "epoch": 1.2505561735261401, "grad_norm": 0.4639621376991272, "learning_rate": 7.23917182673387e-06, "loss": 0.4259, "step": 1499 }, { "epoch": 1.2513904338153503, "grad_norm": 0.46462148427963257, "learning_rate": 7.234827929304134e-06, "loss": 0.4724, "step": 1500 }, { "epoch": 1.2522246941045605, "grad_norm": 0.39841508865356445, "learning_rate": 7.2304819229366015e-06, "loss": 0.4124, "step": 1501 }, { "epoch": 1.253058954393771, "grad_norm": 0.4352608621120453, "learning_rate": 7.226133811732465e-06, "loss": 0.4066, "step": 1502 }, { "epoch": 1.253893214682981, "grad_norm": 0.44499877095222473, "learning_rate": 7.2217835997949e-06, "loss": 0.4121, "step": 1503 }, { "epoch": 1.2547274749721913, "grad_norm": 0.44167518615722656, "learning_rate": 7.217431291229068e-06, "loss": 0.4495, "step": 1504 }, { "epoch": 1.2555617352614017, "grad_norm": 0.3722616732120514, "learning_rate": 7.213076890142107e-06, "loss": 0.4332, "step": 1505 }, { "epoch": 1.2563959955506117, "grad_norm": 0.3937070965766907, "learning_rate": 7.208720400643131e-06, "loss": 0.4344, "step": 1506 }, { "epoch": 1.257230255839822, "grad_norm": 0.4235491156578064, "learning_rate": 7.204361826843223e-06, "loss": 0.427, "step": 1507 }, { "epoch": 1.2580645161290323, "grad_norm": 0.3435286283493042, "learning_rate": 7.200001172855436e-06, "loss": 0.4111, "step": 1508 }, { "epoch": 1.2588987764182424, "grad_norm": 0.3923328220844269, "learning_rate": 7.195638442794781e-06, "loss": 0.4148, "step": 1509 }, { "epoch": 1.2597330367074528, "grad_norm": 0.44932371377944946, "learning_rate": 7.191273640778232e-06, "loss": 0.4645, "step": 1510 }, { "epoch": 1.260567296996663, "grad_norm": 0.3333640396595001, "learning_rate": 7.186906770924719e-06, "loss": 0.3732, "step": 1511 }, { "epoch": 1.2614015572858732, "grad_norm": 0.3994513154029846, "learning_rate": 7.182537837355119e-06, "loss": 0.379, "step": 1512 }, { "epoch": 1.2622358175750834, "grad_norm": 0.4261477291584015, "learning_rate": 7.17816684419226e-06, "loss": 0.4729, "step": 1513 }, { "epoch": 1.2630700778642936, "grad_norm": 0.40953657031059265, "learning_rate": 7.173793795560915e-06, "loss": 0.4892, "step": 1514 }, { "epoch": 1.263904338153504, "grad_norm": 0.3802069127559662, "learning_rate": 7.169418695587791e-06, "loss": 0.3946, "step": 1515 }, { "epoch": 1.2647385984427142, "grad_norm": 0.4145275354385376, "learning_rate": 7.165041548401536e-06, "loss": 0.4772, "step": 1516 }, { "epoch": 1.2655728587319244, "grad_norm": 0.37408459186553955, "learning_rate": 7.1606623581327275e-06, "loss": 0.4058, "step": 1517 }, { "epoch": 1.2664071190211346, "grad_norm": 0.3672976493835449, "learning_rate": 7.156281128913871e-06, "loss": 0.4064, "step": 1518 }, { "epoch": 1.2672413793103448, "grad_norm": 0.3810864984989166, "learning_rate": 7.151897864879399e-06, "loss": 0.4583, "step": 1519 }, { "epoch": 1.2680756395995552, "grad_norm": 0.3849179446697235, "learning_rate": 7.147512570165659e-06, "loss": 0.4581, "step": 1520 }, { "epoch": 1.2689098998887653, "grad_norm": 0.3485821485519409, "learning_rate": 7.143125248910917e-06, "loss": 0.3994, "step": 1521 }, { "epoch": 1.2697441601779755, "grad_norm": 0.364894837141037, "learning_rate": 7.138735905255355e-06, "loss": 0.4555, "step": 1522 }, { "epoch": 1.2705784204671857, "grad_norm": 0.36580976843833923, "learning_rate": 7.1343445433410594e-06, "loss": 0.3883, "step": 1523 }, { "epoch": 1.271412680756396, "grad_norm": 0.38560718297958374, "learning_rate": 7.1299511673120214e-06, "loss": 0.4536, "step": 1524 }, { "epoch": 1.2722469410456063, "grad_norm": 0.35499194264411926, "learning_rate": 7.125555781314135e-06, "loss": 0.4195, "step": 1525 }, { "epoch": 1.2730812013348165, "grad_norm": 0.36469313502311707, "learning_rate": 7.121158389495187e-06, "loss": 0.4188, "step": 1526 }, { "epoch": 1.2739154616240267, "grad_norm": 0.33546513319015503, "learning_rate": 7.116758996004862e-06, "loss": 0.4175, "step": 1527 }, { "epoch": 1.2747497219132369, "grad_norm": 0.37647950649261475, "learning_rate": 7.112357604994732e-06, "loss": 0.419, "step": 1528 }, { "epoch": 1.275583982202447, "grad_norm": 0.3356795012950897, "learning_rate": 7.107954220618251e-06, "loss": 0.4454, "step": 1529 }, { "epoch": 1.2764182424916575, "grad_norm": 0.34670084714889526, "learning_rate": 7.103548847030757e-06, "loss": 0.3803, "step": 1530 }, { "epoch": 1.2772525027808677, "grad_norm": 0.36617186665534973, "learning_rate": 7.099141488389464e-06, "loss": 0.4218, "step": 1531 }, { "epoch": 1.2780867630700778, "grad_norm": 0.38970494270324707, "learning_rate": 7.09473214885346e-06, "loss": 0.4265, "step": 1532 }, { "epoch": 1.278921023359288, "grad_norm": 0.3900890648365021, "learning_rate": 7.090320832583704e-06, "loss": 0.4315, "step": 1533 }, { "epoch": 1.2797552836484982, "grad_norm": 0.4280264973640442, "learning_rate": 7.085907543743013e-06, "loss": 0.4489, "step": 1534 }, { "epoch": 1.2805895439377086, "grad_norm": 0.3273426294326782, "learning_rate": 7.081492286496077e-06, "loss": 0.4047, "step": 1535 }, { "epoch": 1.2814238042269188, "grad_norm": 0.38155993819236755, "learning_rate": 7.0770750650094335e-06, "loss": 0.3724, "step": 1536 }, { "epoch": 1.282258064516129, "grad_norm": 0.3943970799446106, "learning_rate": 7.072655883451478e-06, "loss": 0.4628, "step": 1537 }, { "epoch": 1.2830923248053392, "grad_norm": 0.35796424746513367, "learning_rate": 7.068234745992456e-06, "loss": 0.419, "step": 1538 }, { "epoch": 1.2839265850945494, "grad_norm": 0.35420385003089905, "learning_rate": 7.063811656804456e-06, "loss": 0.411, "step": 1539 }, { "epoch": 1.2847608453837598, "grad_norm": 0.38912665843963623, "learning_rate": 7.059386620061412e-06, "loss": 0.4411, "step": 1540 }, { "epoch": 1.28559510567297, "grad_norm": 0.4041212499141693, "learning_rate": 7.054959639939095e-06, "loss": 0.4435, "step": 1541 }, { "epoch": 1.2864293659621802, "grad_norm": 0.40881702303886414, "learning_rate": 7.0505307206151045e-06, "loss": 0.4576, "step": 1542 }, { "epoch": 1.2872636262513906, "grad_norm": 0.3394160270690918, "learning_rate": 7.046099866268878e-06, "loss": 0.3865, "step": 1543 }, { "epoch": 1.2880978865406005, "grad_norm": 0.41013771295547485, "learning_rate": 7.041667081081674e-06, "loss": 0.4833, "step": 1544 }, { "epoch": 1.288932146829811, "grad_norm": 0.33661574125289917, "learning_rate": 7.037232369236575e-06, "loss": 0.3415, "step": 1545 }, { "epoch": 1.2897664071190211, "grad_norm": 0.43068525195121765, "learning_rate": 7.032795734918479e-06, "loss": 0.4838, "step": 1546 }, { "epoch": 1.2906006674082313, "grad_norm": 0.4071597456932068, "learning_rate": 7.028357182314103e-06, "loss": 0.4087, "step": 1547 }, { "epoch": 1.2914349276974417, "grad_norm": 0.39967086911201477, "learning_rate": 7.023916715611969e-06, "loss": 0.4658, "step": 1548 }, { "epoch": 1.292269187986652, "grad_norm": 0.3541748821735382, "learning_rate": 7.01947433900241e-06, "loss": 0.4172, "step": 1549 }, { "epoch": 1.293103448275862, "grad_norm": 0.4110005795955658, "learning_rate": 7.015030056677559e-06, "loss": 0.4693, "step": 1550 }, { "epoch": 1.2939377085650723, "grad_norm": 0.38158583641052246, "learning_rate": 7.0105838728313436e-06, "loss": 0.4365, "step": 1551 }, { "epoch": 1.2947719688542825, "grad_norm": 0.3818630874156952, "learning_rate": 7.006135791659494e-06, "loss": 0.4685, "step": 1552 }, { "epoch": 1.2956062291434929, "grad_norm": 0.3660869002342224, "learning_rate": 7.001685817359524e-06, "loss": 0.4299, "step": 1553 }, { "epoch": 1.296440489432703, "grad_norm": 0.3399931490421295, "learning_rate": 6.997233954130737e-06, "loss": 0.4175, "step": 1554 }, { "epoch": 1.2972747497219133, "grad_norm": 0.36511334776878357, "learning_rate": 6.992780206174217e-06, "loss": 0.4517, "step": 1555 }, { "epoch": 1.2981090100111234, "grad_norm": 0.327958345413208, "learning_rate": 6.98832457769283e-06, "loss": 0.4109, "step": 1556 }, { "epoch": 1.2989432703003336, "grad_norm": 0.33503463864326477, "learning_rate": 6.983867072891213e-06, "loss": 0.3983, "step": 1557 }, { "epoch": 1.299777530589544, "grad_norm": 0.3728472888469696, "learning_rate": 6.979407695975776e-06, "loss": 0.414, "step": 1558 }, { "epoch": 1.3006117908787542, "grad_norm": 0.355268269777298, "learning_rate": 6.974946451154694e-06, "loss": 0.4472, "step": 1559 }, { "epoch": 1.3014460511679644, "grad_norm": 0.3459298014640808, "learning_rate": 6.970483342637906e-06, "loss": 0.3953, "step": 1560 }, { "epoch": 1.3022803114571746, "grad_norm": 0.3701285123825073, "learning_rate": 6.966018374637107e-06, "loss": 0.4502, "step": 1561 }, { "epoch": 1.3031145717463848, "grad_norm": 0.3382522463798523, "learning_rate": 6.961551551365752e-06, "loss": 0.447, "step": 1562 }, { "epoch": 1.3039488320355952, "grad_norm": 0.3431307375431061, "learning_rate": 6.957082877039041e-06, "loss": 0.4433, "step": 1563 }, { "epoch": 1.3047830923248054, "grad_norm": 0.3491676151752472, "learning_rate": 6.952612355873922e-06, "loss": 0.4041, "step": 1564 }, { "epoch": 1.3056173526140156, "grad_norm": 0.34047171473503113, "learning_rate": 6.9481399920890924e-06, "loss": 0.4367, "step": 1565 }, { "epoch": 1.3064516129032258, "grad_norm": 0.3346885144710541, "learning_rate": 6.943665789904979e-06, "loss": 0.4445, "step": 1566 }, { "epoch": 1.307285873192436, "grad_norm": 0.34209704399108887, "learning_rate": 6.939189753543749e-06, "loss": 0.4082, "step": 1567 }, { "epoch": 1.3081201334816464, "grad_norm": 0.32725247740745544, "learning_rate": 6.934711887229297e-06, "loss": 0.4197, "step": 1568 }, { "epoch": 1.3089543937708565, "grad_norm": 0.36859703063964844, "learning_rate": 6.930232195187251e-06, "loss": 0.4135, "step": 1569 }, { "epoch": 1.3097886540600667, "grad_norm": 0.33313897252082825, "learning_rate": 6.925750681644954e-06, "loss": 0.3923, "step": 1570 }, { "epoch": 1.310622914349277, "grad_norm": 0.3674825131893158, "learning_rate": 6.9212673508314734e-06, "loss": 0.4872, "step": 1571 }, { "epoch": 1.311457174638487, "grad_norm": 0.379022479057312, "learning_rate": 6.916782206977586e-06, "loss": 0.4473, "step": 1572 }, { "epoch": 1.3122914349276975, "grad_norm": 0.30735334753990173, "learning_rate": 6.912295254315786e-06, "loss": 0.3846, "step": 1573 }, { "epoch": 1.3131256952169077, "grad_norm": 0.3732188940048218, "learning_rate": 6.907806497080272e-06, "loss": 0.4203, "step": 1574 }, { "epoch": 1.3139599555061179, "grad_norm": 0.3908694386482239, "learning_rate": 6.9033159395069415e-06, "loss": 0.4183, "step": 1575 }, { "epoch": 1.314794215795328, "grad_norm": 0.34362950921058655, "learning_rate": 6.898823585833396e-06, "loss": 0.4442, "step": 1576 }, { "epoch": 1.3156284760845383, "grad_norm": 0.3639402389526367, "learning_rate": 6.894329440298932e-06, "loss": 0.423, "step": 1577 }, { "epoch": 1.3164627363737487, "grad_norm": 0.3810531198978424, "learning_rate": 6.889833507144534e-06, "loss": 0.4189, "step": 1578 }, { "epoch": 1.3172969966629589, "grad_norm": 0.4157160222530365, "learning_rate": 6.885335790612874e-06, "loss": 0.4496, "step": 1579 }, { "epoch": 1.318131256952169, "grad_norm": 0.3286227881908417, "learning_rate": 6.880836294948308e-06, "loss": 0.3813, "step": 1580 }, { "epoch": 1.3189655172413794, "grad_norm": 0.3627212941646576, "learning_rate": 6.876335024396872e-06, "loss": 0.4572, "step": 1581 }, { "epoch": 1.3197997775305894, "grad_norm": 0.3884158730506897, "learning_rate": 6.871831983206273e-06, "loss": 0.4451, "step": 1582 }, { "epoch": 1.3206340378197998, "grad_norm": 0.36515089869499207, "learning_rate": 6.867327175625893e-06, "loss": 0.3935, "step": 1583 }, { "epoch": 1.32146829810901, "grad_norm": 0.33927345275878906, "learning_rate": 6.862820605906778e-06, "loss": 0.4082, "step": 1584 }, { "epoch": 1.3223025583982202, "grad_norm": 0.3743673861026764, "learning_rate": 6.858312278301638e-06, "loss": 0.4077, "step": 1585 }, { "epoch": 1.3231368186874306, "grad_norm": 0.3939042389392853, "learning_rate": 6.8538021970648435e-06, "loss": 0.4705, "step": 1586 }, { "epoch": 1.3239710789766408, "grad_norm": 0.3493897616863251, "learning_rate": 6.849290366452417e-06, "loss": 0.3989, "step": 1587 }, { "epoch": 1.324805339265851, "grad_norm": 0.3618108332157135, "learning_rate": 6.8447767907220346e-06, "loss": 0.4181, "step": 1588 }, { "epoch": 1.3256395995550612, "grad_norm": 0.34700915217399597, "learning_rate": 6.840261474133014e-06, "loss": 0.3808, "step": 1589 }, { "epoch": 1.3264738598442714, "grad_norm": 0.37297502160072327, "learning_rate": 6.835744420946324e-06, "loss": 0.4493, "step": 1590 }, { "epoch": 1.3273081201334818, "grad_norm": 0.3632567524909973, "learning_rate": 6.831225635424567e-06, "loss": 0.4309, "step": 1591 }, { "epoch": 1.328142380422692, "grad_norm": 0.33368024230003357, "learning_rate": 6.8267051218319766e-06, "loss": 0.4217, "step": 1592 }, { "epoch": 1.3289766407119021, "grad_norm": 0.4136103093624115, "learning_rate": 6.822182884434424e-06, "loss": 0.4604, "step": 1593 }, { "epoch": 1.3298109010011123, "grad_norm": 0.3950430154800415, "learning_rate": 6.8176589274994044e-06, "loss": 0.4395, "step": 1594 }, { "epoch": 1.3306451612903225, "grad_norm": 0.33796441555023193, "learning_rate": 6.813133255296036e-06, "loss": 0.4118, "step": 1595 }, { "epoch": 1.331479421579533, "grad_norm": 0.39769691228866577, "learning_rate": 6.808605872095055e-06, "loss": 0.433, "step": 1596 }, { "epoch": 1.332313681868743, "grad_norm": 0.39284029603004456, "learning_rate": 6.8040767821688125e-06, "loss": 0.4203, "step": 1597 }, { "epoch": 1.3331479421579533, "grad_norm": 0.3829059302806854, "learning_rate": 6.799545989791268e-06, "loss": 0.4168, "step": 1598 }, { "epoch": 1.3339822024471635, "grad_norm": 0.36800655722618103, "learning_rate": 6.7950134992379935e-06, "loss": 0.4436, "step": 1599 }, { "epoch": 1.3348164627363737, "grad_norm": 0.4121303856372833, "learning_rate": 6.790479314786158e-06, "loss": 0.4901, "step": 1600 }, { "epoch": 1.335650723025584, "grad_norm": 0.36774975061416626, "learning_rate": 6.7859434407145305e-06, "loss": 0.4093, "step": 1601 }, { "epoch": 1.3364849833147943, "grad_norm": 0.39849355816841125, "learning_rate": 6.781405881303475e-06, "loss": 0.4337, "step": 1602 }, { "epoch": 1.3373192436040044, "grad_norm": 0.3597414493560791, "learning_rate": 6.7768666408349445e-06, "loss": 0.4116, "step": 1603 }, { "epoch": 1.3381535038932146, "grad_norm": 0.39071375131607056, "learning_rate": 6.772325723592481e-06, "loss": 0.4389, "step": 1604 }, { "epoch": 1.3389877641824248, "grad_norm": 0.3958578109741211, "learning_rate": 6.7677831338612076e-06, "loss": 0.461, "step": 1605 }, { "epoch": 1.3398220244716352, "grad_norm": 0.321527898311615, "learning_rate": 6.7632388759278225e-06, "loss": 0.3874, "step": 1606 }, { "epoch": 1.3406562847608454, "grad_norm": 0.40059611201286316, "learning_rate": 6.758692954080603e-06, "loss": 0.4586, "step": 1607 }, { "epoch": 1.3414905450500556, "grad_norm": 0.37041348218917847, "learning_rate": 6.7541453726093946e-06, "loss": 0.4574, "step": 1608 }, { "epoch": 1.3423248053392658, "grad_norm": 0.3925189673900604, "learning_rate": 6.749596135805609e-06, "loss": 0.4648, "step": 1609 }, { "epoch": 1.343159065628476, "grad_norm": 0.345735639333725, "learning_rate": 6.7450452479622186e-06, "loss": 0.4453, "step": 1610 }, { "epoch": 1.3439933259176864, "grad_norm": 0.35105717182159424, "learning_rate": 6.740492713373756e-06, "loss": 0.4227, "step": 1611 }, { "epoch": 1.3448275862068966, "grad_norm": 0.35010719299316406, "learning_rate": 6.735938536336309e-06, "loss": 0.4113, "step": 1612 }, { "epoch": 1.3456618464961068, "grad_norm": 0.39327704906463623, "learning_rate": 6.731382721147509e-06, "loss": 0.4144, "step": 1613 }, { "epoch": 1.346496106785317, "grad_norm": 0.374993234872818, "learning_rate": 6.726825272106539e-06, "loss": 0.4544, "step": 1614 }, { "epoch": 1.3473303670745271, "grad_norm": 0.3436034321784973, "learning_rate": 6.722266193514123e-06, "loss": 0.3865, "step": 1615 }, { "epoch": 1.3481646273637375, "grad_norm": 0.37302887439727783, "learning_rate": 6.717705489672524e-06, "loss": 0.4201, "step": 1616 }, { "epoch": 1.3489988876529477, "grad_norm": 0.3709654211997986, "learning_rate": 6.713143164885532e-06, "loss": 0.4414, "step": 1617 }, { "epoch": 1.349833147942158, "grad_norm": 0.35019373893737793, "learning_rate": 6.708579223458475e-06, "loss": 0.4041, "step": 1618 }, { "epoch": 1.350667408231368, "grad_norm": 0.34171605110168457, "learning_rate": 6.704013669698202e-06, "loss": 0.3922, "step": 1619 }, { "epoch": 1.3515016685205783, "grad_norm": 0.38731619715690613, "learning_rate": 6.699446507913083e-06, "loss": 0.4426, "step": 1620 }, { "epoch": 1.3523359288097887, "grad_norm": 0.37063726782798767, "learning_rate": 6.694877742413011e-06, "loss": 0.4439, "step": 1621 }, { "epoch": 1.3531701890989989, "grad_norm": 0.3400062620639801, "learning_rate": 6.690307377509385e-06, "loss": 0.4176, "step": 1622 }, { "epoch": 1.354004449388209, "grad_norm": 0.3863013684749603, "learning_rate": 6.685735417515116e-06, "loss": 0.3821, "step": 1623 }, { "epoch": 1.3548387096774195, "grad_norm": 0.3546510934829712, "learning_rate": 6.681161866744623e-06, "loss": 0.4127, "step": 1624 }, { "epoch": 1.3556729699666297, "grad_norm": 0.41468897461891174, "learning_rate": 6.676586729513823e-06, "loss": 0.4483, "step": 1625 }, { "epoch": 1.3565072302558399, "grad_norm": 0.3267693519592285, "learning_rate": 6.672010010140131e-06, "loss": 0.3922, "step": 1626 }, { "epoch": 1.35734149054505, "grad_norm": 0.34950146079063416, "learning_rate": 6.6674317129424535e-06, "loss": 0.4251, "step": 1627 }, { "epoch": 1.3581757508342602, "grad_norm": 0.3466767370700836, "learning_rate": 6.6628518422411916e-06, "loss": 0.4485, "step": 1628 }, { "epoch": 1.3590100111234706, "grad_norm": 0.3484038710594177, "learning_rate": 6.658270402358223e-06, "loss": 0.4399, "step": 1629 }, { "epoch": 1.3598442714126808, "grad_norm": 0.3672672212123871, "learning_rate": 6.653687397616915e-06, "loss": 0.4218, "step": 1630 }, { "epoch": 1.360678531701891, "grad_norm": 0.3670012950897217, "learning_rate": 6.649102832342103e-06, "loss": 0.438, "step": 1631 }, { "epoch": 1.3615127919911012, "grad_norm": 0.3717747926712036, "learning_rate": 6.6445167108601e-06, "loss": 0.4393, "step": 1632 }, { "epoch": 1.3623470522803114, "grad_norm": 0.3629494607448578, "learning_rate": 6.639929037498687e-06, "loss": 0.4127, "step": 1633 }, { "epoch": 1.3631813125695218, "grad_norm": 0.35044583678245544, "learning_rate": 6.635339816587109e-06, "loss": 0.483, "step": 1634 }, { "epoch": 1.364015572858732, "grad_norm": 0.3335733711719513, "learning_rate": 6.6307490524560726e-06, "loss": 0.4268, "step": 1635 }, { "epoch": 1.3648498331479422, "grad_norm": 0.34521371126174927, "learning_rate": 6.626156749437736e-06, "loss": 0.4029, "step": 1636 }, { "epoch": 1.3656840934371524, "grad_norm": 0.3520677387714386, "learning_rate": 6.621562911865718e-06, "loss": 0.4101, "step": 1637 }, { "epoch": 1.3665183537263625, "grad_norm": 0.3490772545337677, "learning_rate": 6.616967544075077e-06, "loss": 0.3904, "step": 1638 }, { "epoch": 1.367352614015573, "grad_norm": 0.318222314119339, "learning_rate": 6.61237065040232e-06, "loss": 0.3984, "step": 1639 }, { "epoch": 1.3681868743047831, "grad_norm": 0.35957300662994385, "learning_rate": 6.6077722351853925e-06, "loss": 0.471, "step": 1640 }, { "epoch": 1.3690211345939933, "grad_norm": 0.3508577346801758, "learning_rate": 6.6031723027636775e-06, "loss": 0.4573, "step": 1641 }, { "epoch": 1.3698553948832035, "grad_norm": 0.3350491225719452, "learning_rate": 6.598570857477989e-06, "loss": 0.4055, "step": 1642 }, { "epoch": 1.3706896551724137, "grad_norm": 0.34367936849594116, "learning_rate": 6.593967903670567e-06, "loss": 0.4354, "step": 1643 }, { "epoch": 1.371523915461624, "grad_norm": 0.35196825861930847, "learning_rate": 6.589363445685075e-06, "loss": 0.4384, "step": 1644 }, { "epoch": 1.3723581757508343, "grad_norm": 0.36445850133895874, "learning_rate": 6.5847574878666e-06, "loss": 0.441, "step": 1645 }, { "epoch": 1.3731924360400445, "grad_norm": 0.40169185400009155, "learning_rate": 6.580150034561639e-06, "loss": 0.4391, "step": 1646 }, { "epoch": 1.3740266963292547, "grad_norm": 0.3699248731136322, "learning_rate": 6.575541090118105e-06, "loss": 0.4629, "step": 1647 }, { "epoch": 1.3748609566184649, "grad_norm": 0.37944304943084717, "learning_rate": 6.570930658885314e-06, "loss": 0.4139, "step": 1648 }, { "epoch": 1.3756952169076753, "grad_norm": 0.40454912185668945, "learning_rate": 6.566318745213987e-06, "loss": 0.4466, "step": 1649 }, { "epoch": 1.3765294771968855, "grad_norm": 0.36432018876075745, "learning_rate": 6.561705353456245e-06, "loss": 0.403, "step": 1650 }, { "epoch": 1.3773637374860956, "grad_norm": 0.3831561207771301, "learning_rate": 6.5570904879656025e-06, "loss": 0.4249, "step": 1651 }, { "epoch": 1.3781979977753058, "grad_norm": 0.3690386712551117, "learning_rate": 6.552474153096966e-06, "loss": 0.483, "step": 1652 }, { "epoch": 1.379032258064516, "grad_norm": 0.3457645773887634, "learning_rate": 6.547856353206625e-06, "loss": 0.4255, "step": 1653 }, { "epoch": 1.3798665183537264, "grad_norm": 0.32802900671958923, "learning_rate": 6.543237092652255e-06, "loss": 0.4053, "step": 1654 }, { "epoch": 1.3807007786429366, "grad_norm": 0.3480626344680786, "learning_rate": 6.53861637579291e-06, "loss": 0.4212, "step": 1655 }, { "epoch": 1.3815350389321468, "grad_norm": 0.3589338958263397, "learning_rate": 6.533994206989016e-06, "loss": 0.4344, "step": 1656 }, { "epoch": 1.382369299221357, "grad_norm": 0.34494125843048096, "learning_rate": 6.52937059060237e-06, "loss": 0.452, "step": 1657 }, { "epoch": 1.3832035595105672, "grad_norm": 0.4095219671726227, "learning_rate": 6.524745530996137e-06, "loss": 0.4572, "step": 1658 }, { "epoch": 1.3840378197997776, "grad_norm": 0.3402360677719116, "learning_rate": 6.52011903253484e-06, "loss": 0.4232, "step": 1659 }, { "epoch": 1.3848720800889878, "grad_norm": 0.3665957748889923, "learning_rate": 6.515491099584366e-06, "loss": 0.4394, "step": 1660 }, { "epoch": 1.385706340378198, "grad_norm": 0.33454713225364685, "learning_rate": 6.5108617365119475e-06, "loss": 0.4072, "step": 1661 }, { "epoch": 1.3865406006674084, "grad_norm": 0.38357216119766235, "learning_rate": 6.5062309476861714e-06, "loss": 0.4314, "step": 1662 }, { "epoch": 1.3873748609566183, "grad_norm": 0.3535234332084656, "learning_rate": 6.501598737476973e-06, "loss": 0.3983, "step": 1663 }, { "epoch": 1.3882091212458287, "grad_norm": 0.34484124183654785, "learning_rate": 6.496965110255622e-06, "loss": 0.4263, "step": 1664 }, { "epoch": 1.389043381535039, "grad_norm": 0.3630712628364563, "learning_rate": 6.492330070394729e-06, "loss": 0.4152, "step": 1665 }, { "epoch": 1.389877641824249, "grad_norm": 0.38827309012413025, "learning_rate": 6.487693622268239e-06, "loss": 0.4361, "step": 1666 }, { "epoch": 1.3907119021134595, "grad_norm": 0.3070942163467407, "learning_rate": 6.4830557702514215e-06, "loss": 0.3971, "step": 1667 }, { "epoch": 1.3915461624026697, "grad_norm": 0.3782617747783661, "learning_rate": 6.478416518720878e-06, "loss": 0.4328, "step": 1668 }, { "epoch": 1.39238042269188, "grad_norm": 0.3913118839263916, "learning_rate": 6.473775872054522e-06, "loss": 0.4141, "step": 1669 }, { "epoch": 1.39321468298109, "grad_norm": 0.3455771803855896, "learning_rate": 6.4691338346315894e-06, "loss": 0.4178, "step": 1670 }, { "epoch": 1.3940489432703003, "grad_norm": 0.3841426968574524, "learning_rate": 6.464490410832628e-06, "loss": 0.454, "step": 1671 }, { "epoch": 1.3948832035595107, "grad_norm": 0.3589639961719513, "learning_rate": 6.459845605039491e-06, "loss": 0.4257, "step": 1672 }, { "epoch": 1.3957174638487209, "grad_norm": 0.37431642413139343, "learning_rate": 6.455199421635339e-06, "loss": 0.4189, "step": 1673 }, { "epoch": 1.396551724137931, "grad_norm": 0.41356030106544495, "learning_rate": 6.450551865004631e-06, "loss": 0.4024, "step": 1674 }, { "epoch": 1.3973859844271412, "grad_norm": 0.3427436053752899, "learning_rate": 6.445902939533119e-06, "loss": 0.4296, "step": 1675 }, { "epoch": 1.3982202447163514, "grad_norm": 0.3783082365989685, "learning_rate": 6.4412526496078555e-06, "loss": 0.4112, "step": 1676 }, { "epoch": 1.3990545050055618, "grad_norm": 0.4334007203578949, "learning_rate": 6.4366009996171705e-06, "loss": 0.406, "step": 1677 }, { "epoch": 1.399888765294772, "grad_norm": 0.35368117690086365, "learning_rate": 6.431947993950682e-06, "loss": 0.441, "step": 1678 }, { "epoch": 1.4007230255839822, "grad_norm": 0.42388230562210083, "learning_rate": 6.4272936369992885e-06, "loss": 0.4468, "step": 1679 }, { "epoch": 1.4015572858731924, "grad_norm": 0.36202552914619446, "learning_rate": 6.4226379331551625e-06, "loss": 0.3975, "step": 1680 }, { "epoch": 1.4023915461624026, "grad_norm": 0.3753626346588135, "learning_rate": 6.417980886811748e-06, "loss": 0.4884, "step": 1681 }, { "epoch": 1.403225806451613, "grad_norm": 0.34301289916038513, "learning_rate": 6.413322502363751e-06, "loss": 0.3983, "step": 1682 }, { "epoch": 1.4040600667408232, "grad_norm": 0.36388513445854187, "learning_rate": 6.408662784207149e-06, "loss": 0.446, "step": 1683 }, { "epoch": 1.4048943270300334, "grad_norm": 0.32237836718559265, "learning_rate": 6.404001736739171e-06, "loss": 0.4252, "step": 1684 }, { "epoch": 1.4057285873192435, "grad_norm": 0.35138484835624695, "learning_rate": 6.399339364358306e-06, "loss": 0.4412, "step": 1685 }, { "epoch": 1.4065628476084537, "grad_norm": 0.3409031629562378, "learning_rate": 6.394675671464285e-06, "loss": 0.4374, "step": 1686 }, { "epoch": 1.4073971078976641, "grad_norm": 0.3538261950016022, "learning_rate": 6.390010662458095e-06, "loss": 0.4333, "step": 1687 }, { "epoch": 1.4082313681868743, "grad_norm": 0.3358766436576843, "learning_rate": 6.385344341741957e-06, "loss": 0.446, "step": 1688 }, { "epoch": 1.4090656284760845, "grad_norm": 0.38416215777397156, "learning_rate": 6.380676713719336e-06, "loss": 0.4224, "step": 1689 }, { "epoch": 1.4098998887652947, "grad_norm": 0.3384018838405609, "learning_rate": 6.376007782794926e-06, "loss": 0.4147, "step": 1690 }, { "epoch": 1.410734149054505, "grad_norm": 0.3238848149776459, "learning_rate": 6.3713375533746525e-06, "loss": 0.4273, "step": 1691 }, { "epoch": 1.4115684093437153, "grad_norm": 0.3478248715400696, "learning_rate": 6.366666029865667e-06, "loss": 0.4352, "step": 1692 }, { "epoch": 1.4124026696329255, "grad_norm": 0.3329126536846161, "learning_rate": 6.36199321667634e-06, "loss": 0.432, "step": 1693 }, { "epoch": 1.4132369299221357, "grad_norm": 0.3433585464954376, "learning_rate": 6.357319118216263e-06, "loss": 0.3854, "step": 1694 }, { "epoch": 1.4140711902113459, "grad_norm": 0.32657405734062195, "learning_rate": 6.352643738896233e-06, "loss": 0.439, "step": 1695 }, { "epoch": 1.414905450500556, "grad_norm": 0.39330264925956726, "learning_rate": 6.347967083128264e-06, "loss": 0.4553, "step": 1696 }, { "epoch": 1.4157397107897665, "grad_norm": 0.35472023487091064, "learning_rate": 6.34328915532557e-06, "loss": 0.4164, "step": 1697 }, { "epoch": 1.4165739710789766, "grad_norm": 0.3473818600177765, "learning_rate": 6.338609959902569e-06, "loss": 0.4304, "step": 1698 }, { "epoch": 1.4174082313681868, "grad_norm": 0.35394665598869324, "learning_rate": 6.333929501274868e-06, "loss": 0.396, "step": 1699 }, { "epoch": 1.4182424916573972, "grad_norm": 0.33078086376190186, "learning_rate": 6.329247783859271e-06, "loss": 0.4133, "step": 1700 }, { "epoch": 1.4190767519466072, "grad_norm": 0.39534005522727966, "learning_rate": 6.324564812073774e-06, "loss": 0.4204, "step": 1701 }, { "epoch": 1.4199110122358176, "grad_norm": 0.4215700626373291, "learning_rate": 6.319880590337549e-06, "loss": 0.4328, "step": 1702 }, { "epoch": 1.4207452725250278, "grad_norm": 0.33816587924957275, "learning_rate": 6.315195123070952e-06, "loss": 0.4173, "step": 1703 }, { "epoch": 1.421579532814238, "grad_norm": 0.3583212196826935, "learning_rate": 6.310508414695511e-06, "loss": 0.4612, "step": 1704 }, { "epoch": 1.4224137931034484, "grad_norm": 0.3882175385951996, "learning_rate": 6.30582046963393e-06, "loss": 0.4481, "step": 1705 }, { "epoch": 1.4232480533926586, "grad_norm": 0.3504500389099121, "learning_rate": 6.301131292310077e-06, "loss": 0.4098, "step": 1706 }, { "epoch": 1.4240823136818688, "grad_norm": 0.33372482657432556, "learning_rate": 6.2964408871489806e-06, "loss": 0.4407, "step": 1707 }, { "epoch": 1.424916573971079, "grad_norm": 0.3384799361228943, "learning_rate": 6.291749258576831e-06, "loss": 0.4092, "step": 1708 }, { "epoch": 1.4257508342602891, "grad_norm": 0.3697969913482666, "learning_rate": 6.287056411020975e-06, "loss": 0.4321, "step": 1709 }, { "epoch": 1.4265850945494996, "grad_norm": 0.37322717905044556, "learning_rate": 6.282362348909907e-06, "loss": 0.4308, "step": 1710 }, { "epoch": 1.4274193548387097, "grad_norm": 0.34286564588546753, "learning_rate": 6.277667076673266e-06, "loss": 0.4616, "step": 1711 }, { "epoch": 1.42825361512792, "grad_norm": 0.3984396457672119, "learning_rate": 6.272970598741835e-06, "loss": 0.4559, "step": 1712 }, { "epoch": 1.4290878754171301, "grad_norm": 0.38473600149154663, "learning_rate": 6.268272919547537e-06, "loss": 0.4199, "step": 1713 }, { "epoch": 1.4299221357063403, "grad_norm": 0.4002467691898346, "learning_rate": 6.263574043523424e-06, "loss": 0.4704, "step": 1714 }, { "epoch": 1.4307563959955507, "grad_norm": 0.3591274321079254, "learning_rate": 6.258873975103681e-06, "loss": 0.4142, "step": 1715 }, { "epoch": 1.431590656284761, "grad_norm": 0.3514842092990875, "learning_rate": 6.254172718723613e-06, "loss": 0.4242, "step": 1716 }, { "epoch": 1.432424916573971, "grad_norm": 0.3902283310890198, "learning_rate": 6.249470278819655e-06, "loss": 0.405, "step": 1717 }, { "epoch": 1.4332591768631813, "grad_norm": 0.34500592947006226, "learning_rate": 6.244766659829351e-06, "loss": 0.4265, "step": 1718 }, { "epoch": 1.4340934371523915, "grad_norm": 0.32663965225219727, "learning_rate": 6.24006186619136e-06, "loss": 0.3973, "step": 1719 }, { "epoch": 1.4349276974416019, "grad_norm": 0.3638972043991089, "learning_rate": 6.2353559023454505e-06, "loss": 0.4413, "step": 1720 }, { "epoch": 1.435761957730812, "grad_norm": 0.3594224452972412, "learning_rate": 6.2306487727324925e-06, "loss": 0.4119, "step": 1721 }, { "epoch": 1.4365962180200222, "grad_norm": 0.3805740177631378, "learning_rate": 6.225940481794463e-06, "loss": 0.4492, "step": 1722 }, { "epoch": 1.4374304783092324, "grad_norm": 0.35666316747665405, "learning_rate": 6.221231033974427e-06, "loss": 0.4377, "step": 1723 }, { "epoch": 1.4382647385984426, "grad_norm": 0.35126715898513794, "learning_rate": 6.216520433716544e-06, "loss": 0.3878, "step": 1724 }, { "epoch": 1.439098998887653, "grad_norm": 0.404770165681839, "learning_rate": 6.211808685466063e-06, "loss": 0.4584, "step": 1725 }, { "epoch": 1.4399332591768632, "grad_norm": 0.35942354798316956, "learning_rate": 6.207095793669313e-06, "loss": 0.3868, "step": 1726 }, { "epoch": 1.4407675194660734, "grad_norm": 0.36309942603111267, "learning_rate": 6.202381762773703e-06, "loss": 0.46, "step": 1727 }, { "epoch": 1.4416017797552836, "grad_norm": 0.35771217942237854, "learning_rate": 6.197666597227722e-06, "loss": 0.3722, "step": 1728 }, { "epoch": 1.4424360400444938, "grad_norm": 0.34092283248901367, "learning_rate": 6.19295030148092e-06, "loss": 0.4278, "step": 1729 }, { "epoch": 1.4432703003337042, "grad_norm": 0.337833046913147, "learning_rate": 6.188232879983924e-06, "loss": 0.3865, "step": 1730 }, { "epoch": 1.4441045606229144, "grad_norm": 0.4635300934314728, "learning_rate": 6.183514337188416e-06, "loss": 0.5006, "step": 1731 }, { "epoch": 1.4449388209121246, "grad_norm": 0.3574368357658386, "learning_rate": 6.178794677547138e-06, "loss": 0.4358, "step": 1732 }, { "epoch": 1.4457730812013347, "grad_norm": 0.3945150077342987, "learning_rate": 6.174073905513885e-06, "loss": 0.4026, "step": 1733 }, { "epoch": 1.446607341490545, "grad_norm": 0.39540988206863403, "learning_rate": 6.169352025543507e-06, "loss": 0.39, "step": 1734 }, { "epoch": 1.4474416017797553, "grad_norm": 0.41786298155784607, "learning_rate": 6.164629042091894e-06, "loss": 0.462, "step": 1735 }, { "epoch": 1.4482758620689655, "grad_norm": 0.3581660985946655, "learning_rate": 6.1599049596159784e-06, "loss": 0.445, "step": 1736 }, { "epoch": 1.4491101223581757, "grad_norm": 0.3588237762451172, "learning_rate": 6.155179782573731e-06, "loss": 0.37, "step": 1737 }, { "epoch": 1.4499443826473861, "grad_norm": 0.42256447672843933, "learning_rate": 6.150453515424153e-06, "loss": 0.4633, "step": 1738 }, { "epoch": 1.450778642936596, "grad_norm": 0.344255268573761, "learning_rate": 6.145726162627278e-06, "loss": 0.4146, "step": 1739 }, { "epoch": 1.4516129032258065, "grad_norm": 0.34053128957748413, "learning_rate": 6.140997728644162e-06, "loss": 0.4287, "step": 1740 }, { "epoch": 1.4524471635150167, "grad_norm": 0.34585562348365784, "learning_rate": 6.136268217936881e-06, "loss": 0.4196, "step": 1741 }, { "epoch": 1.4532814238042269, "grad_norm": 0.38323643803596497, "learning_rate": 6.131537634968528e-06, "loss": 0.4088, "step": 1742 }, { "epoch": 1.4541156840934373, "grad_norm": 0.3722226321697235, "learning_rate": 6.126805984203208e-06, "loss": 0.3991, "step": 1743 }, { "epoch": 1.4549499443826475, "grad_norm": 0.39310893416404724, "learning_rate": 6.122073270106032e-06, "loss": 0.429, "step": 1744 }, { "epoch": 1.4557842046718577, "grad_norm": 0.3622687757015228, "learning_rate": 6.117339497143114e-06, "loss": 0.4647, "step": 1745 }, { "epoch": 1.4566184649610678, "grad_norm": 0.33306869864463806, "learning_rate": 6.112604669781572e-06, "loss": 0.4247, "step": 1746 }, { "epoch": 1.457452725250278, "grad_norm": 0.38984179496765137, "learning_rate": 6.1078687924895155e-06, "loss": 0.4542, "step": 1747 }, { "epoch": 1.4582869855394884, "grad_norm": 0.4093508720397949, "learning_rate": 6.103131869736042e-06, "loss": 0.4246, "step": 1748 }, { "epoch": 1.4591212458286986, "grad_norm": 0.36912301182746887, "learning_rate": 6.0983939059912405e-06, "loss": 0.3999, "step": 1749 }, { "epoch": 1.4599555061179088, "grad_norm": 0.3962744474411011, "learning_rate": 6.093654905726178e-06, "loss": 0.4294, "step": 1750 }, { "epoch": 1.460789766407119, "grad_norm": 0.3595927953720093, "learning_rate": 6.088914873412908e-06, "loss": 0.438, "step": 1751 }, { "epoch": 1.4616240266963292, "grad_norm": 0.3796848654747009, "learning_rate": 6.0841738135244454e-06, "loss": 0.4146, "step": 1752 }, { "epoch": 1.4624582869855396, "grad_norm": 0.35356003046035767, "learning_rate": 6.079431730534786e-06, "loss": 0.4183, "step": 1753 }, { "epoch": 1.4632925472747498, "grad_norm": 0.3418932855129242, "learning_rate": 6.074688628918884e-06, "loss": 0.4188, "step": 1754 }, { "epoch": 1.46412680756396, "grad_norm": 0.3649088740348816, "learning_rate": 6.069944513152656e-06, "loss": 0.4411, "step": 1755 }, { "epoch": 1.4649610678531702, "grad_norm": 0.35942530632019043, "learning_rate": 6.065199387712982e-06, "loss": 0.4675, "step": 1756 }, { "epoch": 1.4657953281423803, "grad_norm": 0.33811861276626587, "learning_rate": 6.060453257077686e-06, "loss": 0.4162, "step": 1757 }, { "epoch": 1.4666295884315907, "grad_norm": 0.37432366609573364, "learning_rate": 6.0557061257255426e-06, "loss": 0.4563, "step": 1758 }, { "epoch": 1.467463848720801, "grad_norm": 0.36456066370010376, "learning_rate": 6.0509579981362745e-06, "loss": 0.4168, "step": 1759 }, { "epoch": 1.4682981090100111, "grad_norm": 0.3582235872745514, "learning_rate": 6.046208878790543e-06, "loss": 0.4443, "step": 1760 }, { "epoch": 1.4691323692992213, "grad_norm": 0.37242966890335083, "learning_rate": 6.041458772169945e-06, "loss": 0.453, "step": 1761 }, { "epoch": 1.4699666295884315, "grad_norm": 0.3543757200241089, "learning_rate": 6.036707682757007e-06, "loss": 0.4207, "step": 1762 }, { "epoch": 1.470800889877642, "grad_norm": 0.37614893913269043, "learning_rate": 6.031955615035185e-06, "loss": 0.4394, "step": 1763 }, { "epoch": 1.471635150166852, "grad_norm": 0.32726022601127625, "learning_rate": 6.0272025734888595e-06, "loss": 0.4155, "step": 1764 }, { "epoch": 1.4724694104560623, "grad_norm": 0.33901551365852356, "learning_rate": 6.022448562603328e-06, "loss": 0.4461, "step": 1765 }, { "epoch": 1.4733036707452725, "grad_norm": 0.3379817605018616, "learning_rate": 6.017693586864803e-06, "loss": 0.3774, "step": 1766 }, { "epoch": 1.4741379310344827, "grad_norm": 0.4169909656047821, "learning_rate": 6.012937650760406e-06, "loss": 0.4854, "step": 1767 }, { "epoch": 1.474972191323693, "grad_norm": 0.33267632126808167, "learning_rate": 6.008180758778167e-06, "loss": 0.4166, "step": 1768 }, { "epoch": 1.4758064516129032, "grad_norm": 0.3553147315979004, "learning_rate": 6.003422915407019e-06, "loss": 0.4316, "step": 1769 }, { "epoch": 1.4766407119021134, "grad_norm": 0.386070191860199, "learning_rate": 5.998664125136791e-06, "loss": 0.4474, "step": 1770 }, { "epoch": 1.4774749721913236, "grad_norm": 0.37092092633247375, "learning_rate": 5.993904392458203e-06, "loss": 0.4519, "step": 1771 }, { "epoch": 1.4783092324805338, "grad_norm": 0.3352620601654053, "learning_rate": 5.989143721862869e-06, "loss": 0.3898, "step": 1772 }, { "epoch": 1.4791434927697442, "grad_norm": 0.34570157527923584, "learning_rate": 5.984382117843286e-06, "loss": 0.4025, "step": 1773 }, { "epoch": 1.4799777530589544, "grad_norm": 0.405752032995224, "learning_rate": 5.979619584892834e-06, "loss": 0.4599, "step": 1774 }, { "epoch": 1.4808120133481646, "grad_norm": 0.34562236070632935, "learning_rate": 5.974856127505765e-06, "loss": 0.4003, "step": 1775 }, { "epoch": 1.4816462736373748, "grad_norm": 0.3471370041370392, "learning_rate": 5.970091750177205e-06, "loss": 0.4426, "step": 1776 }, { "epoch": 1.482480533926585, "grad_norm": 0.38582542538642883, "learning_rate": 5.965326457403155e-06, "loss": 0.4414, "step": 1777 }, { "epoch": 1.4833147942157954, "grad_norm": 0.36598923802375793, "learning_rate": 5.960560253680467e-06, "loss": 0.4355, "step": 1778 }, { "epoch": 1.4841490545050056, "grad_norm": 0.35599008202552795, "learning_rate": 5.955793143506863e-06, "loss": 0.4135, "step": 1779 }, { "epoch": 1.4849833147942157, "grad_norm": 0.39755794405937195, "learning_rate": 5.951025131380915e-06, "loss": 0.4337, "step": 1780 }, { "epoch": 1.4858175750834262, "grad_norm": 0.351289838552475, "learning_rate": 5.946256221802052e-06, "loss": 0.4252, "step": 1781 }, { "epoch": 1.4866518353726363, "grad_norm": 0.34754008054733276, "learning_rate": 5.941486419270542e-06, "loss": 0.4141, "step": 1782 }, { "epoch": 1.4874860956618465, "grad_norm": 0.32769113779067993, "learning_rate": 5.936715728287503e-06, "loss": 0.3877, "step": 1783 }, { "epoch": 1.4883203559510567, "grad_norm": 0.414461225271225, "learning_rate": 5.931944153354886e-06, "loss": 0.473, "step": 1784 }, { "epoch": 1.489154616240267, "grad_norm": 0.3083798289299011, "learning_rate": 5.927171698975477e-06, "loss": 0.3736, "step": 1785 }, { "epoch": 1.4899888765294773, "grad_norm": 0.35010817646980286, "learning_rate": 5.922398369652896e-06, "loss": 0.4375, "step": 1786 }, { "epoch": 1.4908231368186875, "grad_norm": 0.3634519577026367, "learning_rate": 5.917624169891587e-06, "loss": 0.4414, "step": 1787 }, { "epoch": 1.4916573971078977, "grad_norm": 0.35225871205329895, "learning_rate": 5.91284910419681e-06, "loss": 0.3784, "step": 1788 }, { "epoch": 1.4924916573971079, "grad_norm": 0.3545885384082794, "learning_rate": 5.908073177074647e-06, "loss": 0.4663, "step": 1789 }, { "epoch": 1.493325917686318, "grad_norm": 0.3678167462348938, "learning_rate": 5.903296393031996e-06, "loss": 0.4115, "step": 1790 }, { "epoch": 1.4941601779755285, "grad_norm": 0.3385835886001587, "learning_rate": 5.898518756576557e-06, "loss": 0.423, "step": 1791 }, { "epoch": 1.4949944382647387, "grad_norm": 0.35776853561401367, "learning_rate": 5.893740272216837e-06, "loss": 0.4464, "step": 1792 }, { "epoch": 1.4958286985539488, "grad_norm": 0.34299585223197937, "learning_rate": 5.888960944462143e-06, "loss": 0.4755, "step": 1793 }, { "epoch": 1.496662958843159, "grad_norm": 0.3663299083709717, "learning_rate": 5.884180777822582e-06, "loss": 0.3842, "step": 1794 }, { "epoch": 1.4974972191323692, "grad_norm": 0.3727514147758484, "learning_rate": 5.879399776809047e-06, "loss": 0.4397, "step": 1795 }, { "epoch": 1.4983314794215796, "grad_norm": 0.437333345413208, "learning_rate": 5.874617945933218e-06, "loss": 0.4915, "step": 1796 }, { "epoch": 1.4991657397107898, "grad_norm": 0.3394482135772705, "learning_rate": 5.869835289707565e-06, "loss": 0.4289, "step": 1797 }, { "epoch": 1.5, "grad_norm": 0.33970800042152405, "learning_rate": 5.865051812645329e-06, "loss": 0.3928, "step": 1798 }, { "epoch": 1.5008342602892102, "grad_norm": 0.37352094054222107, "learning_rate": 5.86026751926053e-06, "loss": 0.4384, "step": 1799 }, { "epoch": 1.5016685205784204, "grad_norm": 0.375534325838089, "learning_rate": 5.855482414067957e-06, "loss": 0.4139, "step": 1800 }, { "epoch": 1.5025027808676308, "grad_norm": 0.3609679341316223, "learning_rate": 5.850696501583164e-06, "loss": 0.3878, "step": 1801 }, { "epoch": 1.503337041156841, "grad_norm": 0.3567720949649811, "learning_rate": 5.8459097863224705e-06, "loss": 0.4504, "step": 1802 }, { "epoch": 1.5041713014460512, "grad_norm": 0.38074713945388794, "learning_rate": 5.84112227280295e-06, "loss": 0.4524, "step": 1803 }, { "epoch": 1.5050055617352616, "grad_norm": 0.3865186274051666, "learning_rate": 5.836333965542431e-06, "loss": 0.4139, "step": 1804 }, { "epoch": 1.5058398220244715, "grad_norm": 0.3530215620994568, "learning_rate": 5.831544869059491e-06, "loss": 0.414, "step": 1805 }, { "epoch": 1.506674082313682, "grad_norm": 0.34515345096588135, "learning_rate": 5.8267549878734495e-06, "loss": 0.4326, "step": 1806 }, { "epoch": 1.5075083426028921, "grad_norm": 0.39897793531417847, "learning_rate": 5.821964326504372e-06, "loss": 0.4154, "step": 1807 }, { "epoch": 1.5083426028921023, "grad_norm": 0.38077181577682495, "learning_rate": 5.817172889473056e-06, "loss": 0.4233, "step": 1808 }, { "epoch": 1.5091768631813127, "grad_norm": 0.4097316265106201, "learning_rate": 5.812380681301031e-06, "loss": 0.4558, "step": 1809 }, { "epoch": 1.5100111234705227, "grad_norm": 0.3277016580104828, "learning_rate": 5.8075877065105535e-06, "loss": 0.3763, "step": 1810 }, { "epoch": 1.510845383759733, "grad_norm": 0.34882017970085144, "learning_rate": 5.802793969624609e-06, "loss": 0.3849, "step": 1811 }, { "epoch": 1.5116796440489433, "grad_norm": 0.43491142988204956, "learning_rate": 5.797999475166897e-06, "loss": 0.4799, "step": 1812 }, { "epoch": 1.5125139043381535, "grad_norm": 0.3510618209838867, "learning_rate": 5.793204227661833e-06, "loss": 0.435, "step": 1813 }, { "epoch": 1.5133481646273639, "grad_norm": 0.3862444758415222, "learning_rate": 5.788408231634541e-06, "loss": 0.4497, "step": 1814 }, { "epoch": 1.5141824249165738, "grad_norm": 0.36496201157569885, "learning_rate": 5.7836114916108576e-06, "loss": 0.4144, "step": 1815 }, { "epoch": 1.5150166852057843, "grad_norm": 0.36715638637542725, "learning_rate": 5.778814012117315e-06, "loss": 0.4409, "step": 1816 }, { "epoch": 1.5158509454949944, "grad_norm": 0.3462894856929779, "learning_rate": 5.774015797681148e-06, "loss": 0.3889, "step": 1817 }, { "epoch": 1.5166852057842046, "grad_norm": 0.3801722526550293, "learning_rate": 5.769216852830281e-06, "loss": 0.4188, "step": 1818 }, { "epoch": 1.517519466073415, "grad_norm": 0.36723944544792175, "learning_rate": 5.76441718209333e-06, "loss": 0.4467, "step": 1819 }, { "epoch": 1.518353726362625, "grad_norm": 0.3493143916130066, "learning_rate": 5.7596167899995934e-06, "loss": 0.402, "step": 1820 }, { "epoch": 1.5191879866518354, "grad_norm": 0.33326366543769836, "learning_rate": 5.754815681079055e-06, "loss": 0.3877, "step": 1821 }, { "epoch": 1.5200222469410456, "grad_norm": 0.34619376063346863, "learning_rate": 5.750013859862372e-06, "loss": 0.4534, "step": 1822 }, { "epoch": 1.5208565072302558, "grad_norm": 0.33616530895233154, "learning_rate": 5.745211330880872e-06, "loss": 0.4072, "step": 1823 }, { "epoch": 1.5216907675194662, "grad_norm": 0.36503005027770996, "learning_rate": 5.740408098666555e-06, "loss": 0.418, "step": 1824 }, { "epoch": 1.5225250278086762, "grad_norm": 0.3764933943748474, "learning_rate": 5.735604167752082e-06, "loss": 0.4239, "step": 1825 }, { "epoch": 1.5233592880978866, "grad_norm": 0.36556515097618103, "learning_rate": 5.730799542670774e-06, "loss": 0.4616, "step": 1826 }, { "epoch": 1.5241935483870968, "grad_norm": 0.3597460687160492, "learning_rate": 5.725994227956604e-06, "loss": 0.3992, "step": 1827 }, { "epoch": 1.525027808676307, "grad_norm": 0.364003986120224, "learning_rate": 5.721188228144201e-06, "loss": 0.4468, "step": 1828 }, { "epoch": 1.5258620689655173, "grad_norm": 0.3178982436656952, "learning_rate": 5.716381547768837e-06, "loss": 0.3988, "step": 1829 }, { "epoch": 1.5266963292547273, "grad_norm": 0.3744395673274994, "learning_rate": 5.711574191366427e-06, "loss": 0.5096, "step": 1830 }, { "epoch": 1.5275305895439377, "grad_norm": 0.36647582054138184, "learning_rate": 5.706766163473524e-06, "loss": 0.4228, "step": 1831 }, { "epoch": 1.528364849833148, "grad_norm": 0.3358461260795593, "learning_rate": 5.701957468627315e-06, "loss": 0.4124, "step": 1832 }, { "epoch": 1.529199110122358, "grad_norm": 0.3269394040107727, "learning_rate": 5.6971481113656165e-06, "loss": 0.4119, "step": 1833 }, { "epoch": 1.5300333704115685, "grad_norm": 0.35300952196121216, "learning_rate": 5.69233809622687e-06, "loss": 0.4294, "step": 1834 }, { "epoch": 1.5308676307007787, "grad_norm": 0.3265517055988312, "learning_rate": 5.687527427750138e-06, "loss": 0.3741, "step": 1835 }, { "epoch": 1.5317018909899889, "grad_norm": 0.3667660653591156, "learning_rate": 5.682716110475097e-06, "loss": 0.4663, "step": 1836 }, { "epoch": 1.532536151279199, "grad_norm": 0.3215627372264862, "learning_rate": 5.677904148942039e-06, "loss": 0.3891, "step": 1837 }, { "epoch": 1.5333704115684093, "grad_norm": 0.3344634175300598, "learning_rate": 5.673091547691866e-06, "loss": 0.4372, "step": 1838 }, { "epoch": 1.5342046718576197, "grad_norm": 0.34580373764038086, "learning_rate": 5.6682783112660744e-06, "loss": 0.4038, "step": 1839 }, { "epoch": 1.5350389321468298, "grad_norm": 0.34553655982017517, "learning_rate": 5.6634644442067695e-06, "loss": 0.4445, "step": 1840 }, { "epoch": 1.53587319243604, "grad_norm": 0.3413975238800049, "learning_rate": 5.6586499510566495e-06, "loss": 0.4345, "step": 1841 }, { "epoch": 1.5367074527252504, "grad_norm": 0.35421130061149597, "learning_rate": 5.653834836359e-06, "loss": 0.4044, "step": 1842 }, { "epoch": 1.5375417130144604, "grad_norm": 0.39411723613739014, "learning_rate": 5.649019104657697e-06, "loss": 0.4295, "step": 1843 }, { "epoch": 1.5383759733036708, "grad_norm": 0.40528348088264465, "learning_rate": 5.644202760497195e-06, "loss": 0.3852, "step": 1844 }, { "epoch": 1.539210233592881, "grad_norm": 0.3675856590270996, "learning_rate": 5.6393858084225305e-06, "loss": 0.4529, "step": 1845 }, { "epoch": 1.5400444938820912, "grad_norm": 0.3449460566043854, "learning_rate": 5.634568252979311e-06, "loss": 0.4006, "step": 1846 }, { "epoch": 1.5408787541713016, "grad_norm": 0.4057607054710388, "learning_rate": 5.629750098713716e-06, "loss": 0.4794, "step": 1847 }, { "epoch": 1.5417130144605116, "grad_norm": 0.34113919734954834, "learning_rate": 5.6249313501724844e-06, "loss": 0.3926, "step": 1848 }, { "epoch": 1.542547274749722, "grad_norm": 0.347791850566864, "learning_rate": 5.620112011902924e-06, "loss": 0.4588, "step": 1849 }, { "epoch": 1.5433815350389322, "grad_norm": 0.3991469144821167, "learning_rate": 5.615292088452892e-06, "loss": 0.4557, "step": 1850 }, { "epoch": 1.5442157953281423, "grad_norm": 0.3587254583835602, "learning_rate": 5.6104715843708e-06, "loss": 0.4168, "step": 1851 }, { "epoch": 1.5450500556173528, "grad_norm": 0.35422298312187195, "learning_rate": 5.605650504205611e-06, "loss": 0.4265, "step": 1852 }, { "epoch": 1.5458843159065627, "grad_norm": 0.369015634059906, "learning_rate": 5.600828852506826e-06, "loss": 0.4382, "step": 1853 }, { "epoch": 1.5467185761957731, "grad_norm": 0.32825449109077454, "learning_rate": 5.596006633824491e-06, "loss": 0.4201, "step": 1854 }, { "epoch": 1.5475528364849833, "grad_norm": 0.3557572364807129, "learning_rate": 5.591183852709184e-06, "loss": 0.4477, "step": 1855 }, { "epoch": 1.5483870967741935, "grad_norm": 0.360513836145401, "learning_rate": 5.586360513712011e-06, "loss": 0.4508, "step": 1856 }, { "epoch": 1.549221357063404, "grad_norm": 0.3667665719985962, "learning_rate": 5.581536621384607e-06, "loss": 0.45, "step": 1857 }, { "epoch": 1.5500556173526139, "grad_norm": 0.31304866075515747, "learning_rate": 5.576712180279134e-06, "loss": 0.388, "step": 1858 }, { "epoch": 1.5508898776418243, "grad_norm": 0.3820357322692871, "learning_rate": 5.571887194948266e-06, "loss": 0.4614, "step": 1859 }, { "epoch": 1.5517241379310345, "grad_norm": 0.3298532962799072, "learning_rate": 5.567061669945188e-06, "loss": 0.3387, "step": 1860 }, { "epoch": 1.5525583982202447, "grad_norm": 0.3124491274356842, "learning_rate": 5.562235609823604e-06, "loss": 0.432, "step": 1861 }, { "epoch": 1.553392658509455, "grad_norm": 0.3280312120914459, "learning_rate": 5.557409019137714e-06, "loss": 0.4406, "step": 1862 }, { "epoch": 1.554226918798665, "grad_norm": 0.35434046387672424, "learning_rate": 5.552581902442222e-06, "loss": 0.457, "step": 1863 }, { "epoch": 1.5550611790878754, "grad_norm": 0.33092817664146423, "learning_rate": 5.54775426429233e-06, "loss": 0.3932, "step": 1864 }, { "epoch": 1.5558954393770856, "grad_norm": 0.3245859146118164, "learning_rate": 5.542926109243727e-06, "loss": 0.4114, "step": 1865 }, { "epoch": 1.5567296996662958, "grad_norm": 0.42105039954185486, "learning_rate": 5.538097441852597e-06, "loss": 0.4255, "step": 1866 }, { "epoch": 1.5575639599555062, "grad_norm": 0.3745881915092468, "learning_rate": 5.533268266675601e-06, "loss": 0.477, "step": 1867 }, { "epoch": 1.5583982202447162, "grad_norm": 0.3418346643447876, "learning_rate": 5.528438588269883e-06, "loss": 0.3972, "step": 1868 }, { "epoch": 1.5592324805339266, "grad_norm": 0.3544463515281677, "learning_rate": 5.523608411193062e-06, "loss": 0.4473, "step": 1869 }, { "epoch": 1.5600667408231368, "grad_norm": 0.3591953217983246, "learning_rate": 5.518777740003223e-06, "loss": 0.4362, "step": 1870 }, { "epoch": 1.560901001112347, "grad_norm": 0.30156975984573364, "learning_rate": 5.513946579258923e-06, "loss": 0.3683, "step": 1871 }, { "epoch": 1.5617352614015574, "grad_norm": 0.3801504671573639, "learning_rate": 5.509114933519179e-06, "loss": 0.475, "step": 1872 }, { "epoch": 1.5625695216907676, "grad_norm": 0.3801041841506958, "learning_rate": 5.504282807343461e-06, "loss": 0.3934, "step": 1873 }, { "epoch": 1.5634037819799778, "grad_norm": 0.3276657462120056, "learning_rate": 5.499450205291702e-06, "loss": 0.4309, "step": 1874 }, { "epoch": 1.564238042269188, "grad_norm": 0.31744492053985596, "learning_rate": 5.494617131924276e-06, "loss": 0.4113, "step": 1875 }, { "epoch": 1.5650723025583981, "grad_norm": 0.4222368597984314, "learning_rate": 5.4897835918020045e-06, "loss": 0.4558, "step": 1876 }, { "epoch": 1.5659065628476085, "grad_norm": 0.39779284596443176, "learning_rate": 5.4849495894861495e-06, "loss": 0.4255, "step": 1877 }, { "epoch": 1.5667408231368187, "grad_norm": 0.35272958874702454, "learning_rate": 5.480115129538409e-06, "loss": 0.4489, "step": 1878 }, { "epoch": 1.567575083426029, "grad_norm": 0.31561389565467834, "learning_rate": 5.475280216520913e-06, "loss": 0.4104, "step": 1879 }, { "epoch": 1.568409343715239, "grad_norm": 0.391279011964798, "learning_rate": 5.4704448549962205e-06, "loss": 0.4352, "step": 1880 }, { "epoch": 1.5692436040044493, "grad_norm": 0.3497485816478729, "learning_rate": 5.465609049527311e-06, "loss": 0.4154, "step": 1881 }, { "epoch": 1.5700778642936597, "grad_norm": 0.34634244441986084, "learning_rate": 5.460772804677581e-06, "loss": 0.4197, "step": 1882 }, { "epoch": 1.5709121245828699, "grad_norm": 0.31684112548828125, "learning_rate": 5.45593612501085e-06, "loss": 0.4042, "step": 1883 }, { "epoch": 1.57174638487208, "grad_norm": 0.3875037729740143, "learning_rate": 5.451099015091339e-06, "loss": 0.4757, "step": 1884 }, { "epoch": 1.5725806451612905, "grad_norm": 0.32976391911506653, "learning_rate": 5.44626147948368e-06, "loss": 0.3931, "step": 1885 }, { "epoch": 1.5734149054505004, "grad_norm": 0.3115812838077545, "learning_rate": 5.441423522752904e-06, "loss": 0.3949, "step": 1886 }, { "epoch": 1.5742491657397109, "grad_norm": 0.36374416947364807, "learning_rate": 5.436585149464444e-06, "loss": 0.4314, "step": 1887 }, { "epoch": 1.575083426028921, "grad_norm": 0.3298008143901825, "learning_rate": 5.43174636418412e-06, "loss": 0.4143, "step": 1888 }, { "epoch": 1.5759176863181312, "grad_norm": 0.3393479585647583, "learning_rate": 5.426907171478143e-06, "loss": 0.4218, "step": 1889 }, { "epoch": 1.5767519466073416, "grad_norm": 0.37124401330947876, "learning_rate": 5.42206757591311e-06, "loss": 0.4508, "step": 1890 }, { "epoch": 1.5775862068965516, "grad_norm": 0.3406424820423126, "learning_rate": 5.417227582056e-06, "loss": 0.4134, "step": 1891 }, { "epoch": 1.578420467185762, "grad_norm": 0.3478071391582489, "learning_rate": 5.4123871944741605e-06, "loss": 0.4332, "step": 1892 }, { "epoch": 1.5792547274749722, "grad_norm": 0.3788225054740906, "learning_rate": 5.4075464177353165e-06, "loss": 0.4436, "step": 1893 }, { "epoch": 1.5800889877641824, "grad_norm": 0.3241221606731415, "learning_rate": 5.402705256407561e-06, "loss": 0.3782, "step": 1894 }, { "epoch": 1.5809232480533928, "grad_norm": 0.3935137093067169, "learning_rate": 5.397863715059344e-06, "loss": 0.412, "step": 1895 }, { "epoch": 1.5817575083426028, "grad_norm": 0.35025033354759216, "learning_rate": 5.393021798259482e-06, "loss": 0.395, "step": 1896 }, { "epoch": 1.5825917686318132, "grad_norm": 0.35507458448410034, "learning_rate": 5.388179510577139e-06, "loss": 0.4191, "step": 1897 }, { "epoch": 1.5834260289210234, "grad_norm": 0.3872746229171753, "learning_rate": 5.383336856581833e-06, "loss": 0.4344, "step": 1898 }, { "epoch": 1.5842602892102335, "grad_norm": 0.3771088123321533, "learning_rate": 5.378493840843426e-06, "loss": 0.4904, "step": 1899 }, { "epoch": 1.585094549499444, "grad_norm": 0.3232060372829437, "learning_rate": 5.373650467932122e-06, "loss": 0.4027, "step": 1900 }, { "epoch": 1.585928809788654, "grad_norm": 0.3607063889503479, "learning_rate": 5.368806742418461e-06, "loss": 0.3927, "step": 1901 }, { "epoch": 1.5867630700778643, "grad_norm": 0.36330488324165344, "learning_rate": 5.363962668873317e-06, "loss": 0.4358, "step": 1902 }, { "epoch": 1.5875973303670745, "grad_norm": 0.3343566954135895, "learning_rate": 5.359118251867891e-06, "loss": 0.4249, "step": 1903 }, { "epoch": 1.5884315906562847, "grad_norm": 0.4144919812679291, "learning_rate": 5.354273495973711e-06, "loss": 0.4829, "step": 1904 }, { "epoch": 1.589265850945495, "grad_norm": 0.33871355652809143, "learning_rate": 5.3494284057626215e-06, "loss": 0.3773, "step": 1905 }, { "epoch": 1.590100111234705, "grad_norm": 0.37109774351119995, "learning_rate": 5.344582985806783e-06, "loss": 0.4106, "step": 1906 }, { "epoch": 1.5909343715239155, "grad_norm": 0.3608848452568054, "learning_rate": 5.339737240678671e-06, "loss": 0.4331, "step": 1907 }, { "epoch": 1.5917686318131257, "grad_norm": 0.3356528878211975, "learning_rate": 5.334891174951061e-06, "loss": 0.4017, "step": 1908 }, { "epoch": 1.5926028921023359, "grad_norm": 0.3925946056842804, "learning_rate": 5.330044793197039e-06, "loss": 0.4548, "step": 1909 }, { "epoch": 1.5934371523915463, "grad_norm": 0.3699975609779358, "learning_rate": 5.325198099989982e-06, "loss": 0.4088, "step": 1910 }, { "epoch": 1.5942714126807565, "grad_norm": 0.38873329758644104, "learning_rate": 5.320351099903565e-06, "loss": 0.416, "step": 1911 }, { "epoch": 1.5951056729699666, "grad_norm": 0.35871514678001404, "learning_rate": 5.315503797511752e-06, "loss": 0.407, "step": 1912 }, { "epoch": 1.5959399332591768, "grad_norm": 0.4402942955493927, "learning_rate": 5.310656197388792e-06, "loss": 0.4275, "step": 1913 }, { "epoch": 1.596774193548387, "grad_norm": 0.33696022629737854, "learning_rate": 5.3058083041092145e-06, "loss": 0.4147, "step": 1914 }, { "epoch": 1.5976084538375974, "grad_norm": 0.4155772030353546, "learning_rate": 5.3009601222478254e-06, "loss": 0.4561, "step": 1915 }, { "epoch": 1.5984427141268076, "grad_norm": 0.40149742364883423, "learning_rate": 5.2961116563797045e-06, "loss": 0.4478, "step": 1916 }, { "epoch": 1.5992769744160178, "grad_norm": 0.3602691888809204, "learning_rate": 5.291262911080198e-06, "loss": 0.3836, "step": 1917 }, { "epoch": 1.600111234705228, "grad_norm": 0.43904805183410645, "learning_rate": 5.286413890924918e-06, "loss": 0.4183, "step": 1918 }, { "epoch": 1.6009454949944382, "grad_norm": 0.37550583481788635, "learning_rate": 5.281564600489732e-06, "loss": 0.4275, "step": 1919 }, { "epoch": 1.6017797552836486, "grad_norm": 0.3369218111038208, "learning_rate": 5.276715044350766e-06, "loss": 0.4383, "step": 1920 }, { "epoch": 1.6026140155728588, "grad_norm": 0.35058268904685974, "learning_rate": 5.271865227084397e-06, "loss": 0.4215, "step": 1921 }, { "epoch": 1.603448275862069, "grad_norm": 0.3335247039794922, "learning_rate": 5.267015153267246e-06, "loss": 0.4286, "step": 1922 }, { "epoch": 1.6042825361512794, "grad_norm": 0.37663722038269043, "learning_rate": 5.262164827476176e-06, "loss": 0.4698, "step": 1923 }, { "epoch": 1.6051167964404893, "grad_norm": 0.3418547511100769, "learning_rate": 5.25731425428829e-06, "loss": 0.4259, "step": 1924 }, { "epoch": 1.6059510567296997, "grad_norm": 0.3620426654815674, "learning_rate": 5.252463438280925e-06, "loss": 0.4088, "step": 1925 }, { "epoch": 1.60678531701891, "grad_norm": 0.3534836769104004, "learning_rate": 5.247612384031644e-06, "loss": 0.4414, "step": 1926 }, { "epoch": 1.60761957730812, "grad_norm": 0.32681530714035034, "learning_rate": 5.242761096118236e-06, "loss": 0.4045, "step": 1927 }, { "epoch": 1.6084538375973305, "grad_norm": 0.38962191343307495, "learning_rate": 5.237909579118713e-06, "loss": 0.4454, "step": 1928 }, { "epoch": 1.6092880978865405, "grad_norm": 0.3498905599117279, "learning_rate": 5.233057837611297e-06, "loss": 0.4075, "step": 1929 }, { "epoch": 1.610122358175751, "grad_norm": 0.3312284052371979, "learning_rate": 5.22820587617443e-06, "loss": 0.4193, "step": 1930 }, { "epoch": 1.610956618464961, "grad_norm": 0.37025687098503113, "learning_rate": 5.223353699386758e-06, "loss": 0.4337, "step": 1931 }, { "epoch": 1.6117908787541713, "grad_norm": 0.3225787878036499, "learning_rate": 5.218501311827128e-06, "loss": 0.3469, "step": 1932 }, { "epoch": 1.6126251390433817, "grad_norm": 0.34107398986816406, "learning_rate": 5.213648718074584e-06, "loss": 0.4694, "step": 1933 }, { "epoch": 1.6134593993325916, "grad_norm": 0.3138311207294464, "learning_rate": 5.208795922708374e-06, "loss": 0.4018, "step": 1934 }, { "epoch": 1.614293659621802, "grad_norm": 0.3693374693393707, "learning_rate": 5.2039429303079294e-06, "loss": 0.4238, "step": 1935 }, { "epoch": 1.6151279199110122, "grad_norm": 0.364181250333786, "learning_rate": 5.199089745452866e-06, "loss": 0.4684, "step": 1936 }, { "epoch": 1.6159621802002224, "grad_norm": 0.2857789695262909, "learning_rate": 5.194236372722984e-06, "loss": 0.3703, "step": 1937 }, { "epoch": 1.6167964404894328, "grad_norm": 0.3578060567378998, "learning_rate": 5.189382816698263e-06, "loss": 0.4255, "step": 1938 }, { "epoch": 1.6176307007786428, "grad_norm": 0.3604782521724701, "learning_rate": 5.184529081958853e-06, "loss": 0.4478, "step": 1939 }, { "epoch": 1.6184649610678532, "grad_norm": 0.31981563568115234, "learning_rate": 5.179675173085071e-06, "loss": 0.3837, "step": 1940 }, { "epoch": 1.6192992213570634, "grad_norm": 0.35473352670669556, "learning_rate": 5.174821094657403e-06, "loss": 0.4276, "step": 1941 }, { "epoch": 1.6201334816462736, "grad_norm": 0.34009769558906555, "learning_rate": 5.169966851256489e-06, "loss": 0.455, "step": 1942 }, { "epoch": 1.620967741935484, "grad_norm": 0.3425910770893097, "learning_rate": 5.165112447463131e-06, "loss": 0.4263, "step": 1943 }, { "epoch": 1.621802002224694, "grad_norm": 0.3475549817085266, "learning_rate": 5.160257887858278e-06, "loss": 0.4381, "step": 1944 }, { "epoch": 1.6226362625139044, "grad_norm": 0.3295859694480896, "learning_rate": 5.155403177023025e-06, "loss": 0.4257, "step": 1945 }, { "epoch": 1.6234705228031145, "grad_norm": 0.3199617266654968, "learning_rate": 5.150548319538614e-06, "loss": 0.4002, "step": 1946 }, { "epoch": 1.6243047830923247, "grad_norm": 0.3810797333717346, "learning_rate": 5.1456933199864255e-06, "loss": 0.4846, "step": 1947 }, { "epoch": 1.6251390433815351, "grad_norm": 0.3434957265853882, "learning_rate": 5.140838182947969e-06, "loss": 0.4185, "step": 1948 }, { "epoch": 1.625973303670745, "grad_norm": 0.33196794986724854, "learning_rate": 5.135982913004889e-06, "loss": 0.4404, "step": 1949 }, { "epoch": 1.6268075639599555, "grad_norm": 0.34848466515541077, "learning_rate": 5.13112751473895e-06, "loss": 0.4378, "step": 1950 }, { "epoch": 1.6276418242491657, "grad_norm": 0.3267745077610016, "learning_rate": 5.126271992732044e-06, "loss": 0.4293, "step": 1951 }, { "epoch": 1.628476084538376, "grad_norm": 0.3890758156776428, "learning_rate": 5.121416351566177e-06, "loss": 0.4592, "step": 1952 }, { "epoch": 1.6293103448275863, "grad_norm": 0.30736950039863586, "learning_rate": 5.116560595823463e-06, "loss": 0.3831, "step": 1953 }, { "epoch": 1.6301446051167965, "grad_norm": 0.3446367681026459, "learning_rate": 5.11170473008613e-06, "loss": 0.443, "step": 1954 }, { "epoch": 1.6309788654060067, "grad_norm": 0.3491804301738739, "learning_rate": 5.106848758936508e-06, "loss": 0.4168, "step": 1955 }, { "epoch": 1.6318131256952169, "grad_norm": 0.32931265234947205, "learning_rate": 5.101992686957028e-06, "loss": 0.4361, "step": 1956 }, { "epoch": 1.632647385984427, "grad_norm": 0.3387599587440491, "learning_rate": 5.097136518730211e-06, "loss": 0.4057, "step": 1957 }, { "epoch": 1.6334816462736375, "grad_norm": 0.3997807204723358, "learning_rate": 5.092280258838677e-06, "loss": 0.4823, "step": 1958 }, { "epoch": 1.6343159065628476, "grad_norm": 0.33180904388427734, "learning_rate": 5.087423911865121e-06, "loss": 0.4323, "step": 1959 }, { "epoch": 1.6351501668520578, "grad_norm": 0.3203374147415161, "learning_rate": 5.0825674823923345e-06, "loss": 0.4152, "step": 1960 }, { "epoch": 1.6359844271412682, "grad_norm": 0.33522137999534607, "learning_rate": 5.077710975003175e-06, "loss": 0.3803, "step": 1961 }, { "epoch": 1.6368186874304782, "grad_norm": 0.3476499915122986, "learning_rate": 5.0728543942805785e-06, "loss": 0.4878, "step": 1962 }, { "epoch": 1.6376529477196886, "grad_norm": 0.32842233777046204, "learning_rate": 5.06799774480755e-06, "loss": 0.3924, "step": 1963 }, { "epoch": 1.6384872080088988, "grad_norm": 0.39134475588798523, "learning_rate": 5.06314103116716e-06, "loss": 0.4214, "step": 1964 }, { "epoch": 1.639321468298109, "grad_norm": 0.3092092275619507, "learning_rate": 5.058284257942536e-06, "loss": 0.3941, "step": 1965 }, { "epoch": 1.6401557285873194, "grad_norm": 0.35948145389556885, "learning_rate": 5.053427429716867e-06, "loss": 0.4385, "step": 1966 }, { "epoch": 1.6409899888765294, "grad_norm": 0.3887205123901367, "learning_rate": 5.048570551073388e-06, "loss": 0.4781, "step": 1967 }, { "epoch": 1.6418242491657398, "grad_norm": 0.3425556719303131, "learning_rate": 5.043713626595388e-06, "loss": 0.3767, "step": 1968 }, { "epoch": 1.64265850945495, "grad_norm": 0.3756742477416992, "learning_rate": 5.038856660866194e-06, "loss": 0.4357, "step": 1969 }, { "epoch": 1.6434927697441601, "grad_norm": 0.3643359839916229, "learning_rate": 5.033999658469174e-06, "loss": 0.4161, "step": 1970 }, { "epoch": 1.6443270300333706, "grad_norm": 0.3236760199069977, "learning_rate": 5.029142623987732e-06, "loss": 0.4547, "step": 1971 }, { "epoch": 1.6451612903225805, "grad_norm": 0.3608415424823761, "learning_rate": 5.0242855620052985e-06, "loss": 0.448, "step": 1972 }, { "epoch": 1.645995550611791, "grad_norm": 0.30428338050842285, "learning_rate": 5.019428477105335e-06, "loss": 0.3883, "step": 1973 }, { "epoch": 1.6468298109010011, "grad_norm": 0.34165894985198975, "learning_rate": 5.014571373871319e-06, "loss": 0.4206, "step": 1974 }, { "epoch": 1.6476640711902113, "grad_norm": 0.31267908215522766, "learning_rate": 5.009714256886748e-06, "loss": 0.4533, "step": 1975 }, { "epoch": 1.6484983314794217, "grad_norm": 0.2867894172668457, "learning_rate": 5.004857130735137e-06, "loss": 0.3675, "step": 1976 }, { "epoch": 1.6493325917686317, "grad_norm": 0.3466445803642273, "learning_rate": 5e-06, "loss": 0.444, "step": 1977 }, { "epoch": 1.650166852057842, "grad_norm": 0.3509494662284851, "learning_rate": 4.995142869264867e-06, "loss": 0.4262, "step": 1978 }, { "epoch": 1.6510011123470523, "grad_norm": 0.32536476850509644, "learning_rate": 4.9902857431132525e-06, "loss": 0.4433, "step": 1979 }, { "epoch": 1.6518353726362625, "grad_norm": 0.33255723118782043, "learning_rate": 4.985428626128684e-06, "loss": 0.4303, "step": 1980 }, { "epoch": 1.6526696329254729, "grad_norm": 0.36425909399986267, "learning_rate": 4.980571522894668e-06, "loss": 0.4465, "step": 1981 }, { "epoch": 1.6535038932146828, "grad_norm": 0.3325182795524597, "learning_rate": 4.975714437994704e-06, "loss": 0.4281, "step": 1982 }, { "epoch": 1.6543381535038932, "grad_norm": 0.297776460647583, "learning_rate": 4.97085737601227e-06, "loss": 0.364, "step": 1983 }, { "epoch": 1.6551724137931034, "grad_norm": 0.35053402185440063, "learning_rate": 4.966000341530827e-06, "loss": 0.4582, "step": 1984 }, { "epoch": 1.6560066740823136, "grad_norm": 0.3339439630508423, "learning_rate": 4.961143339133806e-06, "loss": 0.42, "step": 1985 }, { "epoch": 1.656840934371524, "grad_norm": 0.36310577392578125, "learning_rate": 4.956286373404614e-06, "loss": 0.4403, "step": 1986 }, { "epoch": 1.657675194660734, "grad_norm": 0.32042670249938965, "learning_rate": 4.951429448926612e-06, "loss": 0.4223, "step": 1987 }, { "epoch": 1.6585094549499444, "grad_norm": 0.2969556152820587, "learning_rate": 4.946572570283135e-06, "loss": 0.3986, "step": 1988 }, { "epoch": 1.6593437152391546, "grad_norm": 0.32281747460365295, "learning_rate": 4.941715742057466e-06, "loss": 0.3988, "step": 1989 }, { "epoch": 1.6601779755283648, "grad_norm": 0.35498273372650146, "learning_rate": 4.936858968832841e-06, "loss": 0.453, "step": 1990 }, { "epoch": 1.6610122358175752, "grad_norm": 0.30598294734954834, "learning_rate": 4.932002255192452e-06, "loss": 0.4152, "step": 1991 }, { "epoch": 1.6618464961067854, "grad_norm": 0.3237619996070862, "learning_rate": 4.927145605719423e-06, "loss": 0.4213, "step": 1992 }, { "epoch": 1.6626807563959956, "grad_norm": 0.3265078663825989, "learning_rate": 4.922289024996827e-06, "loss": 0.418, "step": 1993 }, { "epoch": 1.6635150166852057, "grad_norm": 0.39053773880004883, "learning_rate": 4.917432517607666e-06, "loss": 0.4701, "step": 1994 }, { "epoch": 1.664349276974416, "grad_norm": 0.3603520095348358, "learning_rate": 4.91257608813488e-06, "loss": 0.3904, "step": 1995 }, { "epoch": 1.6651835372636263, "grad_norm": 0.3335421085357666, "learning_rate": 4.907719741161325e-06, "loss": 0.3913, "step": 1996 }, { "epoch": 1.6660177975528365, "grad_norm": 0.3502005338668823, "learning_rate": 4.902863481269789e-06, "loss": 0.4346, "step": 1997 }, { "epoch": 1.6668520578420467, "grad_norm": 0.346099317073822, "learning_rate": 4.898007313042975e-06, "loss": 0.4278, "step": 1998 }, { "epoch": 1.6676863181312571, "grad_norm": 0.3557526767253876, "learning_rate": 4.893151241063493e-06, "loss": 0.4284, "step": 1999 }, { "epoch": 1.668520578420467, "grad_norm": 0.3967556357383728, "learning_rate": 4.888295269913872e-06, "loss": 0.4541, "step": 2000 }, { "epoch": 1.6693548387096775, "grad_norm": 0.33592891693115234, "learning_rate": 4.8834394041765385e-06, "loss": 0.4171, "step": 2001 }, { "epoch": 1.6701890989988877, "grad_norm": 0.3323156237602234, "learning_rate": 4.878583648433826e-06, "loss": 0.4145, "step": 2002 }, { "epoch": 1.6710233592880979, "grad_norm": 0.35258743166923523, "learning_rate": 4.873728007267956e-06, "loss": 0.4367, "step": 2003 }, { "epoch": 1.6718576195773083, "grad_norm": 0.3728429973125458, "learning_rate": 4.868872485261051e-06, "loss": 0.4024, "step": 2004 }, { "epoch": 1.6726918798665182, "grad_norm": 0.3715766668319702, "learning_rate": 4.864017086995112e-06, "loss": 0.4157, "step": 2005 }, { "epoch": 1.6735261401557286, "grad_norm": 0.36981257796287537, "learning_rate": 4.859161817052031e-06, "loss": 0.4131, "step": 2006 }, { "epoch": 1.6743604004449388, "grad_norm": 0.3418399691581726, "learning_rate": 4.8543066800135745e-06, "loss": 0.4199, "step": 2007 }, { "epoch": 1.675194660734149, "grad_norm": 0.41819706559181213, "learning_rate": 4.8494516804613864e-06, "loss": 0.4461, "step": 2008 }, { "epoch": 1.6760289210233594, "grad_norm": 0.380597323179245, "learning_rate": 4.844596822976978e-06, "loss": 0.3931, "step": 2009 }, { "epoch": 1.6768631813125694, "grad_norm": 0.3272729516029358, "learning_rate": 4.839742112141725e-06, "loss": 0.3921, "step": 2010 }, { "epoch": 1.6776974416017798, "grad_norm": 0.3711085617542267, "learning_rate": 4.834887552536871e-06, "loss": 0.4333, "step": 2011 }, { "epoch": 1.67853170189099, "grad_norm": 0.3605871796607971, "learning_rate": 4.830033148743512e-06, "loss": 0.4114, "step": 2012 }, { "epoch": 1.6793659621802002, "grad_norm": 0.3562982678413391, "learning_rate": 4.825178905342599e-06, "loss": 0.3785, "step": 2013 }, { "epoch": 1.6802002224694106, "grad_norm": 0.3907734751701355, "learning_rate": 4.820324826914928e-06, "loss": 0.4691, "step": 2014 }, { "epoch": 1.6810344827586206, "grad_norm": 0.3734469711780548, "learning_rate": 4.8154709180411485e-06, "loss": 0.3895, "step": 2015 }, { "epoch": 1.681868743047831, "grad_norm": 0.374376505613327, "learning_rate": 4.810617183301737e-06, "loss": 0.4322, "step": 2016 }, { "epoch": 1.6827030033370411, "grad_norm": 0.3707486689090729, "learning_rate": 4.805763627277017e-06, "loss": 0.371, "step": 2017 }, { "epoch": 1.6835372636262513, "grad_norm": 0.3988339304924011, "learning_rate": 4.800910254547136e-06, "loss": 0.4237, "step": 2018 }, { "epoch": 1.6843715239154617, "grad_norm": 0.3413623571395874, "learning_rate": 4.796057069692073e-06, "loss": 0.3972, "step": 2019 }, { "epoch": 1.6852057842046717, "grad_norm": 0.3530906140804291, "learning_rate": 4.791204077291628e-06, "loss": 0.4246, "step": 2020 }, { "epoch": 1.6860400444938821, "grad_norm": 0.4002096951007843, "learning_rate": 4.786351281925417e-06, "loss": 0.4407, "step": 2021 }, { "epoch": 1.6868743047830923, "grad_norm": 0.3269517123699188, "learning_rate": 4.781498688172875e-06, "loss": 0.3806, "step": 2022 }, { "epoch": 1.6877085650723025, "grad_norm": 0.3389360308647156, "learning_rate": 4.776646300613243e-06, "loss": 0.4394, "step": 2023 }, { "epoch": 1.688542825361513, "grad_norm": 0.3714200556278229, "learning_rate": 4.7717941238255706e-06, "loss": 0.4362, "step": 2024 }, { "epoch": 1.6893770856507229, "grad_norm": 0.2980697751045227, "learning_rate": 4.766942162388703e-06, "loss": 0.3605, "step": 2025 }, { "epoch": 1.6902113459399333, "grad_norm": 0.36664795875549316, "learning_rate": 4.762090420881289e-06, "loss": 0.4335, "step": 2026 }, { "epoch": 1.6910456062291435, "grad_norm": 0.38145649433135986, "learning_rate": 4.757238903881764e-06, "loss": 0.4297, "step": 2027 }, { "epoch": 1.6918798665183536, "grad_norm": 0.3705378770828247, "learning_rate": 4.752387615968358e-06, "loss": 0.393, "step": 2028 }, { "epoch": 1.692714126807564, "grad_norm": 0.34017011523246765, "learning_rate": 4.7475365617190774e-06, "loss": 0.4108, "step": 2029 }, { "epoch": 1.6935483870967742, "grad_norm": 0.41382741928100586, "learning_rate": 4.742685745711711e-06, "loss": 0.4434, "step": 2030 }, { "epoch": 1.6943826473859844, "grad_norm": 0.3924911916255951, "learning_rate": 4.737835172523826e-06, "loss": 0.4577, "step": 2031 }, { "epoch": 1.6952169076751946, "grad_norm": 0.3370306193828583, "learning_rate": 4.732984846732755e-06, "loss": 0.4143, "step": 2032 }, { "epoch": 1.6960511679644048, "grad_norm": 0.4242655038833618, "learning_rate": 4.728134772915605e-06, "loss": 0.4382, "step": 2033 }, { "epoch": 1.6968854282536152, "grad_norm": 0.35472527146339417, "learning_rate": 4.7232849556492346e-06, "loss": 0.3892, "step": 2034 }, { "epoch": 1.6977196885428254, "grad_norm": 0.3224971294403076, "learning_rate": 4.718435399510269e-06, "loss": 0.3818, "step": 2035 }, { "epoch": 1.6985539488320356, "grad_norm": 0.4477153718471527, "learning_rate": 4.713586109075083e-06, "loss": 0.4545, "step": 2036 }, { "epoch": 1.6993882091212458, "grad_norm": 0.35747677087783813, "learning_rate": 4.7087370889198034e-06, "loss": 0.4206, "step": 2037 }, { "epoch": 1.700222469410456, "grad_norm": 0.3041766583919525, "learning_rate": 4.7038883436202955e-06, "loss": 0.4234, "step": 2038 }, { "epoch": 1.7010567296996664, "grad_norm": 0.40536001324653625, "learning_rate": 4.699039877752175e-06, "loss": 0.4244, "step": 2039 }, { "epoch": 1.7018909899888766, "grad_norm": 0.3881164789199829, "learning_rate": 4.694191695890788e-06, "loss": 0.4177, "step": 2040 }, { "epoch": 1.7027252502780867, "grad_norm": 0.3305839002132416, "learning_rate": 4.689343802611209e-06, "loss": 0.4003, "step": 2041 }, { "epoch": 1.7035595105672972, "grad_norm": 0.35763880610466003, "learning_rate": 4.68449620248825e-06, "loss": 0.459, "step": 2042 }, { "epoch": 1.7043937708565071, "grad_norm": 0.38510265946388245, "learning_rate": 4.679648900096436e-06, "loss": 0.4223, "step": 2043 }, { "epoch": 1.7052280311457175, "grad_norm": 0.3436644971370697, "learning_rate": 4.67480190001002e-06, "loss": 0.4332, "step": 2044 }, { "epoch": 1.7060622914349277, "grad_norm": 0.3052067458629608, "learning_rate": 4.6699552068029615e-06, "loss": 0.3823, "step": 2045 }, { "epoch": 1.706896551724138, "grad_norm": 0.3630422055721283, "learning_rate": 4.6651088250489405e-06, "loss": 0.4603, "step": 2046 }, { "epoch": 1.7077308120133483, "grad_norm": 0.3662045896053314, "learning_rate": 4.660262759321331e-06, "loss": 0.4637, "step": 2047 }, { "epoch": 1.7085650723025583, "grad_norm": 0.3272319436073303, "learning_rate": 4.655417014193218e-06, "loss": 0.3977, "step": 2048 }, { "epoch": 1.7093993325917687, "grad_norm": 0.3383512496948242, "learning_rate": 4.650571594237379e-06, "loss": 0.4222, "step": 2049 }, { "epoch": 1.7102335928809789, "grad_norm": 0.3223508298397064, "learning_rate": 4.645726504026291e-06, "loss": 0.399, "step": 2050 }, { "epoch": 1.711067853170189, "grad_norm": 0.3195055425167084, "learning_rate": 4.64088174813211e-06, "loss": 0.3992, "step": 2051 }, { "epoch": 1.7119021134593995, "grad_norm": 0.3342472314834595, "learning_rate": 4.636037331126684e-06, "loss": 0.4112, "step": 2052 }, { "epoch": 1.7127363737486094, "grad_norm": 0.3495992124080658, "learning_rate": 4.631193257581541e-06, "loss": 0.4578, "step": 2053 }, { "epoch": 1.7135706340378198, "grad_norm": 0.3018129765987396, "learning_rate": 4.626349532067879e-06, "loss": 0.4027, "step": 2054 }, { "epoch": 1.71440489432703, "grad_norm": 0.3246499001979828, "learning_rate": 4.6215061591565754e-06, "loss": 0.4603, "step": 2055 }, { "epoch": 1.7152391546162402, "grad_norm": 0.3835367262363434, "learning_rate": 4.6166631434181675e-06, "loss": 0.4324, "step": 2056 }, { "epoch": 1.7160734149054506, "grad_norm": 0.29748156666755676, "learning_rate": 4.611820489422862e-06, "loss": 0.401, "step": 2057 }, { "epoch": 1.7169076751946606, "grad_norm": 0.34028834104537964, "learning_rate": 4.606978201740519e-06, "loss": 0.4288, "step": 2058 }, { "epoch": 1.717741935483871, "grad_norm": 0.4010790288448334, "learning_rate": 4.602136284940657e-06, "loss": 0.4524, "step": 2059 }, { "epoch": 1.7185761957730812, "grad_norm": 0.33036044239997864, "learning_rate": 4.597294743592442e-06, "loss": 0.4531, "step": 2060 }, { "epoch": 1.7194104560622914, "grad_norm": 0.3538036346435547, "learning_rate": 4.592453582264684e-06, "loss": 0.4235, "step": 2061 }, { "epoch": 1.7202447163515018, "grad_norm": 0.34610629081726074, "learning_rate": 4.587612805525842e-06, "loss": 0.3721, "step": 2062 }, { "epoch": 1.7210789766407117, "grad_norm": 0.36061713099479675, "learning_rate": 4.582772417944003e-06, "loss": 0.4498, "step": 2063 }, { "epoch": 1.7219132369299222, "grad_norm": 0.34797388315200806, "learning_rate": 4.577932424086891e-06, "loss": 0.4056, "step": 2064 }, { "epoch": 1.7227474972191323, "grad_norm": 0.3233029246330261, "learning_rate": 4.573092828521857e-06, "loss": 0.3739, "step": 2065 }, { "epoch": 1.7235817575083425, "grad_norm": 0.372408926486969, "learning_rate": 4.568253635815882e-06, "loss": 0.464, "step": 2066 }, { "epoch": 1.724416017797553, "grad_norm": 0.3200142979621887, "learning_rate": 4.5634148505355566e-06, "loss": 0.4031, "step": 2067 }, { "epoch": 1.7252502780867631, "grad_norm": 0.37714990973472595, "learning_rate": 4.558576477247097e-06, "loss": 0.4684, "step": 2068 }, { "epoch": 1.7260845383759733, "grad_norm": 0.3718406856060028, "learning_rate": 4.553738520516321e-06, "loss": 0.4195, "step": 2069 }, { "epoch": 1.7269187986651835, "grad_norm": 0.38523221015930176, "learning_rate": 4.548900984908662e-06, "loss": 0.4523, "step": 2070 }, { "epoch": 1.7277530589543937, "grad_norm": 0.3816388249397278, "learning_rate": 4.544063874989153e-06, "loss": 0.3883, "step": 2071 }, { "epoch": 1.728587319243604, "grad_norm": 0.38949301838874817, "learning_rate": 4.539227195322421e-06, "loss": 0.4292, "step": 2072 }, { "epoch": 1.7294215795328143, "grad_norm": 0.3764198422431946, "learning_rate": 4.5343909504726915e-06, "loss": 0.4353, "step": 2073 }, { "epoch": 1.7302558398220245, "grad_norm": 0.3578626811504364, "learning_rate": 4.52955514500378e-06, "loss": 0.4603, "step": 2074 }, { "epoch": 1.7310901001112347, "grad_norm": 0.33561697602272034, "learning_rate": 4.524719783479088e-06, "loss": 0.3763, "step": 2075 }, { "epoch": 1.7319243604004448, "grad_norm": 0.4121699333190918, "learning_rate": 4.5198848704615915e-06, "loss": 0.4014, "step": 2076 }, { "epoch": 1.7327586206896552, "grad_norm": 0.3914092481136322, "learning_rate": 4.515050410513851e-06, "loss": 0.4539, "step": 2077 }, { "epoch": 1.7335928809788654, "grad_norm": 0.3629712760448456, "learning_rate": 4.510216408197996e-06, "loss": 0.4109, "step": 2078 }, { "epoch": 1.7344271412680756, "grad_norm": 0.353538453578949, "learning_rate": 4.505382868075726e-06, "loss": 0.4236, "step": 2079 }, { "epoch": 1.735261401557286, "grad_norm": 0.35204681754112244, "learning_rate": 4.500549794708301e-06, "loss": 0.39, "step": 2080 }, { "epoch": 1.736095661846496, "grad_norm": 0.3448371887207031, "learning_rate": 4.49571719265654e-06, "loss": 0.4217, "step": 2081 }, { "epoch": 1.7369299221357064, "grad_norm": 0.3754028379917145, "learning_rate": 4.4908850664808245e-06, "loss": 0.4155, "step": 2082 }, { "epoch": 1.7377641824249166, "grad_norm": 0.35702118277549744, "learning_rate": 4.486053420741079e-06, "loss": 0.3983, "step": 2083 }, { "epoch": 1.7385984427141268, "grad_norm": 0.40252161026000977, "learning_rate": 4.481222259996779e-06, "loss": 0.4839, "step": 2084 }, { "epoch": 1.7394327030033372, "grad_norm": 0.3217894732952118, "learning_rate": 4.47639158880694e-06, "loss": 0.4169, "step": 2085 }, { "epoch": 1.7402669632925472, "grad_norm": 0.3361777663230896, "learning_rate": 4.471561411730117e-06, "loss": 0.407, "step": 2086 }, { "epoch": 1.7411012235817576, "grad_norm": 0.41372883319854736, "learning_rate": 4.466731733324399e-06, "loss": 0.4292, "step": 2087 }, { "epoch": 1.7419354838709677, "grad_norm": 0.3565269410610199, "learning_rate": 4.461902558147405e-06, "loss": 0.4076, "step": 2088 }, { "epoch": 1.742769744160178, "grad_norm": 0.38486745953559875, "learning_rate": 4.457073890756273e-06, "loss": 0.4355, "step": 2089 }, { "epoch": 1.7436040044493883, "grad_norm": 0.3574977517127991, "learning_rate": 4.452245735707671e-06, "loss": 0.4495, "step": 2090 }, { "epoch": 1.7444382647385983, "grad_norm": 0.3536582887172699, "learning_rate": 4.44741809755778e-06, "loss": 0.3812, "step": 2091 }, { "epoch": 1.7452725250278087, "grad_norm": 0.3651394546031952, "learning_rate": 4.4425909808622875e-06, "loss": 0.4295, "step": 2092 }, { "epoch": 1.746106785317019, "grad_norm": 0.3784821927547455, "learning_rate": 4.437764390176399e-06, "loss": 0.4138, "step": 2093 }, { "epoch": 1.746941045606229, "grad_norm": 0.34921422600746155, "learning_rate": 4.432938330054813e-06, "loss": 0.4248, "step": 2094 }, { "epoch": 1.7477753058954395, "grad_norm": 0.3430158197879791, "learning_rate": 4.4281128050517365e-06, "loss": 0.3998, "step": 2095 }, { "epoch": 1.7486095661846495, "grad_norm": 0.4221331477165222, "learning_rate": 4.423287819720866e-06, "loss": 0.4477, "step": 2096 }, { "epoch": 1.7494438264738599, "grad_norm": 0.3847402334213257, "learning_rate": 4.418463378615395e-06, "loss": 0.406, "step": 2097 }, { "epoch": 1.75027808676307, "grad_norm": 0.3252056837081909, "learning_rate": 4.413639486287992e-06, "loss": 0.4272, "step": 2098 }, { "epoch": 1.7511123470522802, "grad_norm": 0.3303466737270355, "learning_rate": 4.408816147290819e-06, "loss": 0.4699, "step": 2099 }, { "epoch": 1.7519466073414907, "grad_norm": 0.36803579330444336, "learning_rate": 4.403993366175509e-06, "loss": 0.3976, "step": 2100 }, { "epoch": 1.7527808676307006, "grad_norm": 0.3701639771461487, "learning_rate": 4.3991711474931746e-06, "loss": 0.44, "step": 2101 }, { "epoch": 1.753615127919911, "grad_norm": 0.3463243842124939, "learning_rate": 4.394349495794391e-06, "loss": 0.473, "step": 2102 }, { "epoch": 1.7544493882091212, "grad_norm": 0.336681991815567, "learning_rate": 4.389528415629201e-06, "loss": 0.3784, "step": 2103 }, { "epoch": 1.7552836484983314, "grad_norm": 0.3713769018650055, "learning_rate": 4.384707911547111e-06, "loss": 0.4567, "step": 2104 }, { "epoch": 1.7561179087875418, "grad_norm": 0.2904699444770813, "learning_rate": 4.379887988097078e-06, "loss": 0.376, "step": 2105 }, { "epoch": 1.756952169076752, "grad_norm": 0.38665732741355896, "learning_rate": 4.375068649827517e-06, "loss": 0.454, "step": 2106 }, { "epoch": 1.7577864293659622, "grad_norm": 0.3559630811214447, "learning_rate": 4.370249901286286e-06, "loss": 0.4458, "step": 2107 }, { "epoch": 1.7586206896551724, "grad_norm": 0.3190843462944031, "learning_rate": 4.3654317470206894e-06, "loss": 0.3787, "step": 2108 }, { "epoch": 1.7594549499443826, "grad_norm": 0.39372408390045166, "learning_rate": 4.3606141915774695e-06, "loss": 0.4417, "step": 2109 }, { "epoch": 1.760289210233593, "grad_norm": 0.38130196928977966, "learning_rate": 4.355797239502807e-06, "loss": 0.3932, "step": 2110 }, { "epoch": 1.7611234705228032, "grad_norm": 0.328241765499115, "learning_rate": 4.350980895342306e-06, "loss": 0.4195, "step": 2111 }, { "epoch": 1.7619577308120133, "grad_norm": 0.36369842290878296, "learning_rate": 4.3461651636410014e-06, "loss": 0.4379, "step": 2112 }, { "epoch": 1.7627919911012235, "grad_norm": 0.36852437257766724, "learning_rate": 4.341350048943353e-06, "loss": 0.4056, "step": 2113 }, { "epoch": 1.7636262513904337, "grad_norm": 0.3463919758796692, "learning_rate": 4.336535555793231e-06, "loss": 0.4555, "step": 2114 }, { "epoch": 1.7644605116796441, "grad_norm": 0.36672672629356384, "learning_rate": 4.331721688733927e-06, "loss": 0.4143, "step": 2115 }, { "epoch": 1.7652947719688543, "grad_norm": 0.3515910506248474, "learning_rate": 4.326908452308136e-06, "loss": 0.387, "step": 2116 }, { "epoch": 1.7661290322580645, "grad_norm": 0.39994683861732483, "learning_rate": 4.322095851057962e-06, "loss": 0.4427, "step": 2117 }, { "epoch": 1.766963292547275, "grad_norm": 0.326637327671051, "learning_rate": 4.317283889524903e-06, "loss": 0.4283, "step": 2118 }, { "epoch": 1.7677975528364849, "grad_norm": 0.3344876170158386, "learning_rate": 4.3124725722498635e-06, "loss": 0.4052, "step": 2119 }, { "epoch": 1.7686318131256953, "grad_norm": 0.37111949920654297, "learning_rate": 4.307661903773129e-06, "loss": 0.4144, "step": 2120 }, { "epoch": 1.7694660734149055, "grad_norm": 0.352350652217865, "learning_rate": 4.302851888634384e-06, "loss": 0.4238, "step": 2121 }, { "epoch": 1.7703003337041157, "grad_norm": 0.3410108685493469, "learning_rate": 4.298042531372687e-06, "loss": 0.3947, "step": 2122 }, { "epoch": 1.771134593993326, "grad_norm": 0.41571545600891113, "learning_rate": 4.293233836526478e-06, "loss": 0.462, "step": 2123 }, { "epoch": 1.771968854282536, "grad_norm": 0.37703293561935425, "learning_rate": 4.2884258086335755e-06, "loss": 0.487, "step": 2124 }, { "epoch": 1.7728031145717464, "grad_norm": 0.32531753182411194, "learning_rate": 4.283618452231164e-06, "loss": 0.3976, "step": 2125 }, { "epoch": 1.7736373748609566, "grad_norm": 0.31765374541282654, "learning_rate": 4.2788117718558015e-06, "loss": 0.3809, "step": 2126 }, { "epoch": 1.7744716351501668, "grad_norm": 0.3829658031463623, "learning_rate": 4.274005772043397e-06, "loss": 0.4068, "step": 2127 }, { "epoch": 1.7753058954393772, "grad_norm": 0.3505929112434387, "learning_rate": 4.269200457329228e-06, "loss": 0.4089, "step": 2128 }, { "epoch": 1.7761401557285872, "grad_norm": 0.3437963128089905, "learning_rate": 4.2643958322479176e-06, "loss": 0.4218, "step": 2129 }, { "epoch": 1.7769744160177976, "grad_norm": 0.3291890025138855, "learning_rate": 4.2595919013334455e-06, "loss": 0.3993, "step": 2130 }, { "epoch": 1.7778086763070078, "grad_norm": 0.40227028727531433, "learning_rate": 4.254788669119127e-06, "loss": 0.4554, "step": 2131 }, { "epoch": 1.778642936596218, "grad_norm": 0.395007461309433, "learning_rate": 4.249986140137629e-06, "loss": 0.4076, "step": 2132 }, { "epoch": 1.7794771968854284, "grad_norm": 0.34915754199028015, "learning_rate": 4.245184318920947e-06, "loss": 0.4081, "step": 2133 }, { "epoch": 1.7803114571746383, "grad_norm": 0.3547031581401825, "learning_rate": 4.240383210000407e-06, "loss": 0.4272, "step": 2134 }, { "epoch": 1.7811457174638488, "grad_norm": 0.3240562081336975, "learning_rate": 4.235582817906673e-06, "loss": 0.3861, "step": 2135 }, { "epoch": 1.781979977753059, "grad_norm": 0.32610562443733215, "learning_rate": 4.23078314716972e-06, "loss": 0.3758, "step": 2136 }, { "epoch": 1.7828142380422691, "grad_norm": 0.40180131793022156, "learning_rate": 4.225984202318854e-06, "loss": 0.4564, "step": 2137 }, { "epoch": 1.7836484983314795, "grad_norm": 0.342831015586853, "learning_rate": 4.221185987882684e-06, "loss": 0.4044, "step": 2138 }, { "epoch": 1.7844827586206895, "grad_norm": 0.33297058939933777, "learning_rate": 4.216388508389144e-06, "loss": 0.4525, "step": 2139 }, { "epoch": 1.7853170189099, "grad_norm": 0.34853971004486084, "learning_rate": 4.211591768365459e-06, "loss": 0.3931, "step": 2140 }, { "epoch": 1.78615127919911, "grad_norm": 0.33653736114501953, "learning_rate": 4.206795772338169e-06, "loss": 0.4343, "step": 2141 }, { "epoch": 1.7869855394883203, "grad_norm": 0.35235658288002014, "learning_rate": 4.2020005248331056e-06, "loss": 0.4281, "step": 2142 }, { "epoch": 1.7878197997775307, "grad_norm": 0.32961341738700867, "learning_rate": 4.197206030375392e-06, "loss": 0.3756, "step": 2143 }, { "epoch": 1.7886540600667407, "grad_norm": 0.33251357078552246, "learning_rate": 4.192412293489448e-06, "loss": 0.4097, "step": 2144 }, { "epoch": 1.789488320355951, "grad_norm": 0.3196227550506592, "learning_rate": 4.187619318698971e-06, "loss": 0.4252, "step": 2145 }, { "epoch": 1.7903225806451613, "grad_norm": 0.3314020335674286, "learning_rate": 4.182827110526947e-06, "loss": 0.4035, "step": 2146 }, { "epoch": 1.7911568409343714, "grad_norm": 0.3491023778915405, "learning_rate": 4.178035673495628e-06, "loss": 0.4127, "step": 2147 }, { "epoch": 1.7919911012235819, "grad_norm": 0.32020241022109985, "learning_rate": 4.173245012126551e-06, "loss": 0.446, "step": 2148 }, { "epoch": 1.792825361512792, "grad_norm": 0.41506513953208923, "learning_rate": 4.16845513094051e-06, "loss": 0.4258, "step": 2149 }, { "epoch": 1.7936596218020022, "grad_norm": 0.3467378616333008, "learning_rate": 4.1636660344575705e-06, "loss": 0.4259, "step": 2150 }, { "epoch": 1.7944938820912124, "grad_norm": 0.3573753535747528, "learning_rate": 4.158877727197051e-06, "loss": 0.4445, "step": 2151 }, { "epoch": 1.7953281423804226, "grad_norm": 0.3284016251564026, "learning_rate": 4.154090213677531e-06, "loss": 0.3776, "step": 2152 }, { "epoch": 1.796162402669633, "grad_norm": 0.3396306335926056, "learning_rate": 4.149303498416838e-06, "loss": 0.4361, "step": 2153 }, { "epoch": 1.7969966629588432, "grad_norm": 0.3101923167705536, "learning_rate": 4.144517585932045e-06, "loss": 0.3905, "step": 2154 }, { "epoch": 1.7978309232480534, "grad_norm": 0.35775625705718994, "learning_rate": 4.139732480739473e-06, "loss": 0.4928, "step": 2155 }, { "epoch": 1.7986651835372638, "grad_norm": 0.35566189885139465, "learning_rate": 4.134948187354673e-06, "loss": 0.436, "step": 2156 }, { "epoch": 1.7994994438264738, "grad_norm": 0.3416772782802582, "learning_rate": 4.1301647102924355e-06, "loss": 0.4343, "step": 2157 }, { "epoch": 1.8003337041156842, "grad_norm": 0.35462114214897156, "learning_rate": 4.125382054066781e-06, "loss": 0.4409, "step": 2158 }, { "epoch": 1.8011679644048944, "grad_norm": 0.34269776940345764, "learning_rate": 4.120600223190955e-06, "loss": 0.4158, "step": 2159 }, { "epoch": 1.8020022246941045, "grad_norm": 0.34517303109169006, "learning_rate": 4.115819222177418e-06, "loss": 0.4418, "step": 2160 }, { "epoch": 1.802836484983315, "grad_norm": 0.34084394574165344, "learning_rate": 4.1110390555378575e-06, "loss": 0.3844, "step": 2161 }, { "epoch": 1.803670745272525, "grad_norm": 0.3680521845817566, "learning_rate": 4.106259727783164e-06, "loss": 0.4345, "step": 2162 }, { "epoch": 1.8045050055617353, "grad_norm": 0.34634819626808167, "learning_rate": 4.101481243423446e-06, "loss": 0.3916, "step": 2163 }, { "epoch": 1.8053392658509455, "grad_norm": 0.367437481880188, "learning_rate": 4.096703606968007e-06, "loss": 0.4226, "step": 2164 }, { "epoch": 1.8061735261401557, "grad_norm": 0.335455983877182, "learning_rate": 4.091926822925354e-06, "loss": 0.4419, "step": 2165 }, { "epoch": 1.807007786429366, "grad_norm": 0.3335912525653839, "learning_rate": 4.087150895803192e-06, "loss": 0.3959, "step": 2166 }, { "epoch": 1.807842046718576, "grad_norm": 0.3538050353527069, "learning_rate": 4.082375830108415e-06, "loss": 0.4406, "step": 2167 }, { "epoch": 1.8086763070077865, "grad_norm": 0.3584580421447754, "learning_rate": 4.077601630347105e-06, "loss": 0.4493, "step": 2168 }, { "epoch": 1.8095105672969967, "grad_norm": 0.32330915331840515, "learning_rate": 4.0728283010245236e-06, "loss": 0.4155, "step": 2169 }, { "epoch": 1.8103448275862069, "grad_norm": 0.31497472524642944, "learning_rate": 4.068055846645116e-06, "loss": 0.4395, "step": 2170 }, { "epoch": 1.8111790878754173, "grad_norm": 0.3434939980506897, "learning_rate": 4.063284271712498e-06, "loss": 0.422, "step": 2171 }, { "epoch": 1.8120133481646272, "grad_norm": 0.34802329540252686, "learning_rate": 4.058513580729459e-06, "loss": 0.4422, "step": 2172 }, { "epoch": 1.8128476084538376, "grad_norm": 0.325051873922348, "learning_rate": 4.053743778197951e-06, "loss": 0.4113, "step": 2173 }, { "epoch": 1.8136818687430478, "grad_norm": 0.3351990282535553, "learning_rate": 4.048974868619086e-06, "loss": 0.4358, "step": 2174 }, { "epoch": 1.814516129032258, "grad_norm": 0.30683091282844543, "learning_rate": 4.04420685649314e-06, "loss": 0.4102, "step": 2175 }, { "epoch": 1.8153503893214684, "grad_norm": 0.33208274841308594, "learning_rate": 4.0394397463195345e-06, "loss": 0.4367, "step": 2176 }, { "epoch": 1.8161846496106784, "grad_norm": 0.3181665539741516, "learning_rate": 4.034673542596849e-06, "loss": 0.4346, "step": 2177 }, { "epoch": 1.8170189098998888, "grad_norm": 0.3079817295074463, "learning_rate": 4.029908249822795e-06, "loss": 0.4065, "step": 2178 }, { "epoch": 1.817853170189099, "grad_norm": 0.3238179385662079, "learning_rate": 4.025143872494237e-06, "loss": 0.4236, "step": 2179 }, { "epoch": 1.8186874304783092, "grad_norm": 0.321132093667984, "learning_rate": 4.020380415107167e-06, "loss": 0.44, "step": 2180 }, { "epoch": 1.8195216907675196, "grad_norm": 0.3148971498012543, "learning_rate": 4.015617882156715e-06, "loss": 0.3995, "step": 2181 }, { "epoch": 1.8203559510567295, "grad_norm": 0.3377964496612549, "learning_rate": 4.010856278137132e-06, "loss": 0.4159, "step": 2182 }, { "epoch": 1.82119021134594, "grad_norm": 0.34815382957458496, "learning_rate": 4.006095607541799e-06, "loss": 0.4321, "step": 2183 }, { "epoch": 1.8220244716351501, "grad_norm": 0.3530846834182739, "learning_rate": 4.001335874863213e-06, "loss": 0.4385, "step": 2184 }, { "epoch": 1.8228587319243603, "grad_norm": 0.341614305973053, "learning_rate": 3.996577084592983e-06, "loss": 0.3991, "step": 2185 }, { "epoch": 1.8236929922135707, "grad_norm": 0.3433748483657837, "learning_rate": 3.991819241221836e-06, "loss": 0.4408, "step": 2186 }, { "epoch": 1.824527252502781, "grad_norm": 0.3895556330680847, "learning_rate": 3.987062349239596e-06, "loss": 0.4974, "step": 2187 }, { "epoch": 1.825361512791991, "grad_norm": 0.3340018093585968, "learning_rate": 3.982306413135199e-06, "loss": 0.367, "step": 2188 }, { "epoch": 1.8261957730812013, "grad_norm": 0.34353652596473694, "learning_rate": 3.977551437396672e-06, "loss": 0.433, "step": 2189 }, { "epoch": 1.8270300333704115, "grad_norm": 0.31687748432159424, "learning_rate": 3.972797426511141e-06, "loss": 0.3815, "step": 2190 }, { "epoch": 1.8278642936596219, "grad_norm": 0.3581458628177643, "learning_rate": 3.9680443849648154e-06, "loss": 0.4602, "step": 2191 }, { "epoch": 1.828698553948832, "grad_norm": 0.3098036050796509, "learning_rate": 3.963292317242994e-06, "loss": 0.3952, "step": 2192 }, { "epoch": 1.8295328142380423, "grad_norm": 0.35310354828834534, "learning_rate": 3.958541227830056e-06, "loss": 0.4037, "step": 2193 }, { "epoch": 1.8303670745272527, "grad_norm": 0.39254558086395264, "learning_rate": 3.953791121209458e-06, "loss": 0.473, "step": 2194 }, { "epoch": 1.8312013348164626, "grad_norm": 0.3474428057670593, "learning_rate": 3.949042001863727e-06, "loss": 0.3917, "step": 2195 }, { "epoch": 1.832035595105673, "grad_norm": 0.3494897782802582, "learning_rate": 3.944293874274459e-06, "loss": 0.4008, "step": 2196 }, { "epoch": 1.8328698553948832, "grad_norm": 0.3104712963104248, "learning_rate": 3.939546742922318e-06, "loss": 0.3988, "step": 2197 }, { "epoch": 1.8337041156840934, "grad_norm": 0.3722802400588989, "learning_rate": 3.93480061228702e-06, "loss": 0.4416, "step": 2198 }, { "epoch": 1.8345383759733038, "grad_norm": 0.37576010823249817, "learning_rate": 3.930055486847345e-06, "loss": 0.471, "step": 2199 }, { "epoch": 1.8353726362625138, "grad_norm": 0.3189527094364166, "learning_rate": 3.925311371081118e-06, "loss": 0.4136, "step": 2200 }, { "epoch": 1.8362068965517242, "grad_norm": 0.3355938196182251, "learning_rate": 3.920568269465216e-06, "loss": 0.4016, "step": 2201 }, { "epoch": 1.8370411568409344, "grad_norm": 0.35228610038757324, "learning_rate": 3.9158261864755545e-06, "loss": 0.4001, "step": 2202 }, { "epoch": 1.8378754171301446, "grad_norm": 0.3229486048221588, "learning_rate": 3.911085126587094e-06, "loss": 0.4419, "step": 2203 }, { "epoch": 1.838709677419355, "grad_norm": 0.32613855600357056, "learning_rate": 3.9063450942738225e-06, "loss": 0.4257, "step": 2204 }, { "epoch": 1.839543937708565, "grad_norm": 0.31504523754119873, "learning_rate": 3.901606094008761e-06, "loss": 0.3909, "step": 2205 }, { "epoch": 1.8403781979977754, "grad_norm": 0.37035560607910156, "learning_rate": 3.896868130263961e-06, "loss": 0.3862, "step": 2206 }, { "epoch": 1.8412124582869855, "grad_norm": 0.3331686854362488, "learning_rate": 3.892131207510487e-06, "loss": 0.4415, "step": 2207 }, { "epoch": 1.8420467185761957, "grad_norm": 0.336363822221756, "learning_rate": 3.887395330218429e-06, "loss": 0.4146, "step": 2208 }, { "epoch": 1.8428809788654061, "grad_norm": 0.34014326333999634, "learning_rate": 3.882660502856885e-06, "loss": 0.4075, "step": 2209 }, { "epoch": 1.843715239154616, "grad_norm": 0.340766966342926, "learning_rate": 3.87792672989397e-06, "loss": 0.421, "step": 2210 }, { "epoch": 1.8445494994438265, "grad_norm": 0.36345353722572327, "learning_rate": 3.8731940157967925e-06, "loss": 0.4412, "step": 2211 }, { "epoch": 1.8453837597330367, "grad_norm": 0.3718137741088867, "learning_rate": 3.868462365031473e-06, "loss": 0.4178, "step": 2212 }, { "epoch": 1.8462180200222469, "grad_norm": 0.33281439542770386, "learning_rate": 3.8637317820631195e-06, "loss": 0.4312, "step": 2213 }, { "epoch": 1.8470522803114573, "grad_norm": 0.35511812567710876, "learning_rate": 3.859002271355839e-06, "loss": 0.4447, "step": 2214 }, { "epoch": 1.8478865406006673, "grad_norm": 0.3564709722995758, "learning_rate": 3.854273837372724e-06, "loss": 0.433, "step": 2215 }, { "epoch": 1.8487208008898777, "grad_norm": 0.33320188522338867, "learning_rate": 3.849546484575849e-06, "loss": 0.4386, "step": 2216 }, { "epoch": 1.8495550611790879, "grad_norm": 0.3131037950515747, "learning_rate": 3.844820217426272e-06, "loss": 0.3991, "step": 2217 }, { "epoch": 1.850389321468298, "grad_norm": 0.3365841805934906, "learning_rate": 3.840095040384023e-06, "loss": 0.4127, "step": 2218 }, { "epoch": 1.8512235817575085, "grad_norm": 0.3213375508785248, "learning_rate": 3.835370957908108e-06, "loss": 0.4187, "step": 2219 }, { "epoch": 1.8520578420467184, "grad_norm": 0.33539897203445435, "learning_rate": 3.830647974456494e-06, "loss": 0.4186, "step": 2220 }, { "epoch": 1.8528921023359288, "grad_norm": 0.3356102705001831, "learning_rate": 3.825926094486116e-06, "loss": 0.4329, "step": 2221 }, { "epoch": 1.853726362625139, "grad_norm": 0.3109348714351654, "learning_rate": 3.821205322452863e-06, "loss": 0.3779, "step": 2222 }, { "epoch": 1.8545606229143492, "grad_norm": 0.4022214114665985, "learning_rate": 3.816485662811587e-06, "loss": 0.4598, "step": 2223 }, { "epoch": 1.8553948832035596, "grad_norm": 0.31630274653434753, "learning_rate": 3.8117671200160765e-06, "loss": 0.414, "step": 2224 }, { "epoch": 1.8562291434927698, "grad_norm": 0.32049354910850525, "learning_rate": 3.8070496985190806e-06, "loss": 0.4026, "step": 2225 }, { "epoch": 1.85706340378198, "grad_norm": 0.34562402963638306, "learning_rate": 3.8023334027722804e-06, "loss": 0.4652, "step": 2226 }, { "epoch": 1.8578976640711902, "grad_norm": 0.3275923728942871, "learning_rate": 3.797618237226297e-06, "loss": 0.4167, "step": 2227 }, { "epoch": 1.8587319243604004, "grad_norm": 0.3362849950790405, "learning_rate": 3.7929042063306896e-06, "loss": 0.3952, "step": 2228 }, { "epoch": 1.8595661846496108, "grad_norm": 0.3161216676235199, "learning_rate": 3.7881913145339387e-06, "loss": 0.3977, "step": 2229 }, { "epoch": 1.860400444938821, "grad_norm": 0.3419288396835327, "learning_rate": 3.783479566283457e-06, "loss": 0.4348, "step": 2230 }, { "epoch": 1.8612347052280311, "grad_norm": 0.38343125581741333, "learning_rate": 3.778768966025573e-06, "loss": 0.4199, "step": 2231 }, { "epoch": 1.8620689655172413, "grad_norm": 0.3117449879646301, "learning_rate": 3.7740595182055384e-06, "loss": 0.3788, "step": 2232 }, { "epoch": 1.8629032258064515, "grad_norm": 0.3373115062713623, "learning_rate": 3.769351227267507e-06, "loss": 0.405, "step": 2233 }, { "epoch": 1.863737486095662, "grad_norm": 0.33415982127189636, "learning_rate": 3.764644097654551e-06, "loss": 0.3692, "step": 2234 }, { "epoch": 1.864571746384872, "grad_norm": 0.33340150117874146, "learning_rate": 3.759938133808643e-06, "loss": 0.4418, "step": 2235 }, { "epoch": 1.8654060066740823, "grad_norm": 0.3263048529624939, "learning_rate": 3.7552333401706508e-06, "loss": 0.4408, "step": 2236 }, { "epoch": 1.8662402669632927, "grad_norm": 0.34095636010169983, "learning_rate": 3.7505297211803478e-06, "loss": 0.4547, "step": 2237 }, { "epoch": 1.8670745272525027, "grad_norm": 0.31763115525245667, "learning_rate": 3.7458272812763872e-06, "loss": 0.4327, "step": 2238 }, { "epoch": 1.867908787541713, "grad_norm": 0.308066725730896, "learning_rate": 3.741126024896321e-06, "loss": 0.4042, "step": 2239 }, { "epoch": 1.8687430478309233, "grad_norm": 0.3458216190338135, "learning_rate": 3.736425956476576e-06, "loss": 0.4272, "step": 2240 }, { "epoch": 1.8695773081201335, "grad_norm": 0.30907225608825684, "learning_rate": 3.731727080452464e-06, "loss": 0.3797, "step": 2241 }, { "epoch": 1.8704115684093439, "grad_norm": 0.32341885566711426, "learning_rate": 3.727029401258165e-06, "loss": 0.4149, "step": 2242 }, { "epoch": 1.8712458286985538, "grad_norm": 0.358316034078598, "learning_rate": 3.7223329233267354e-06, "loss": 0.445, "step": 2243 }, { "epoch": 1.8720800889877642, "grad_norm": 0.33664005994796753, "learning_rate": 3.7176376510900935e-06, "loss": 0.4302, "step": 2244 }, { "epoch": 1.8729143492769744, "grad_norm": 0.36642926931381226, "learning_rate": 3.7129435889790267e-06, "loss": 0.4583, "step": 2245 }, { "epoch": 1.8737486095661846, "grad_norm": 0.3228538930416107, "learning_rate": 3.7082507414231705e-06, "loss": 0.38, "step": 2246 }, { "epoch": 1.874582869855395, "grad_norm": 0.34694886207580566, "learning_rate": 3.703559112851022e-06, "loss": 0.4304, "step": 2247 }, { "epoch": 1.875417130144605, "grad_norm": 0.34376585483551025, "learning_rate": 3.6988687076899265e-06, "loss": 0.444, "step": 2248 }, { "epoch": 1.8762513904338154, "grad_norm": 0.30890923738479614, "learning_rate": 3.694179530366071e-06, "loss": 0.3943, "step": 2249 }, { "epoch": 1.8770856507230256, "grad_norm": 0.3383486866950989, "learning_rate": 3.689491585304491e-06, "loss": 0.4127, "step": 2250 }, { "epoch": 1.8779199110122358, "grad_norm": 0.3106797933578491, "learning_rate": 3.6848048769290488e-06, "loss": 0.407, "step": 2251 }, { "epoch": 1.8787541713014462, "grad_norm": 0.3436555564403534, "learning_rate": 3.6801194096624515e-06, "loss": 0.4087, "step": 2252 }, { "epoch": 1.8795884315906561, "grad_norm": 0.3124241828918457, "learning_rate": 3.675435187926226e-06, "loss": 0.4014, "step": 2253 }, { "epoch": 1.8804226918798665, "grad_norm": 0.3655535876750946, "learning_rate": 3.6707522161407294e-06, "loss": 0.4024, "step": 2254 }, { "epoch": 1.8812569521690767, "grad_norm": 0.36909767985343933, "learning_rate": 3.6660704987251337e-06, "loss": 0.4266, "step": 2255 }, { "epoch": 1.882091212458287, "grad_norm": 0.32996994256973267, "learning_rate": 3.6613900400974334e-06, "loss": 0.4234, "step": 2256 }, { "epoch": 1.8829254727474973, "grad_norm": 0.34742072224617004, "learning_rate": 3.6567108446744314e-06, "loss": 0.4297, "step": 2257 }, { "epoch": 1.8837597330367073, "grad_norm": 0.342094361782074, "learning_rate": 3.6520329168717374e-06, "loss": 0.426, "step": 2258 }, { "epoch": 1.8845939933259177, "grad_norm": 0.30557748675346375, "learning_rate": 3.647356261103769e-06, "loss": 0.4267, "step": 2259 }, { "epoch": 1.885428253615128, "grad_norm": 0.3127938210964203, "learning_rate": 3.6426808817837387e-06, "loss": 0.4156, "step": 2260 }, { "epoch": 1.886262513904338, "grad_norm": 0.34623152017593384, "learning_rate": 3.6380067833236613e-06, "loss": 0.3952, "step": 2261 }, { "epoch": 1.8870967741935485, "grad_norm": 0.336731880903244, "learning_rate": 3.633333970134334e-06, "loss": 0.4369, "step": 2262 }, { "epoch": 1.8879310344827587, "grad_norm": 0.321554958820343, "learning_rate": 3.6286624466253496e-06, "loss": 0.4427, "step": 2263 }, { "epoch": 1.8887652947719689, "grad_norm": 0.33362865447998047, "learning_rate": 3.623992217205075e-06, "loss": 0.4361, "step": 2264 }, { "epoch": 1.889599555061179, "grad_norm": 0.3425799310207367, "learning_rate": 3.619323286280665e-06, "loss": 0.4604, "step": 2265 }, { "epoch": 1.8904338153503892, "grad_norm": 0.34136834740638733, "learning_rate": 3.614655658258045e-06, "loss": 0.4302, "step": 2266 }, { "epoch": 1.8912680756395996, "grad_norm": 0.32766374945640564, "learning_rate": 3.609989337541907e-06, "loss": 0.3875, "step": 2267 }, { "epoch": 1.8921023359288098, "grad_norm": 0.3616121709346771, "learning_rate": 3.6053243285357163e-06, "loss": 0.4473, "step": 2268 }, { "epoch": 1.89293659621802, "grad_norm": 0.3073734939098358, "learning_rate": 3.6006606356416958e-06, "loss": 0.4112, "step": 2269 }, { "epoch": 1.8937708565072302, "grad_norm": 0.31997591257095337, "learning_rate": 3.5959982632608304e-06, "loss": 0.4667, "step": 2270 }, { "epoch": 1.8946051167964404, "grad_norm": 0.3109220862388611, "learning_rate": 3.5913372157928515e-06, "loss": 0.4038, "step": 2271 }, { "epoch": 1.8954393770856508, "grad_norm": 0.3522650897502899, "learning_rate": 3.5866774976362495e-06, "loss": 0.4577, "step": 2272 }, { "epoch": 1.896273637374861, "grad_norm": 0.36147645115852356, "learning_rate": 3.582019113188253e-06, "loss": 0.4494, "step": 2273 }, { "epoch": 1.8971078976640712, "grad_norm": 0.3251890242099762, "learning_rate": 3.5773620668448384e-06, "loss": 0.432, "step": 2274 }, { "epoch": 1.8979421579532816, "grad_norm": 0.2981336712837219, "learning_rate": 3.5727063630007115e-06, "loss": 0.4174, "step": 2275 }, { "epoch": 1.8987764182424915, "grad_norm": 0.33223384618759155, "learning_rate": 3.5680520060493197e-06, "loss": 0.399, "step": 2276 }, { "epoch": 1.899610678531702, "grad_norm": 0.31795069575309753, "learning_rate": 3.5633990003828324e-06, "loss": 0.4794, "step": 2277 }, { "epoch": 1.9004449388209121, "grad_norm": 0.29428133368492126, "learning_rate": 3.558747350392146e-06, "loss": 0.3999, "step": 2278 }, { "epoch": 1.9012791991101223, "grad_norm": 0.3348982334136963, "learning_rate": 3.5540970604668824e-06, "loss": 0.4346, "step": 2279 }, { "epoch": 1.9021134593993327, "grad_norm": 0.3025982975959778, "learning_rate": 3.5494481349953715e-06, "loss": 0.4247, "step": 2280 }, { "epoch": 1.9029477196885427, "grad_norm": 0.3185393512248993, "learning_rate": 3.5448005783646626e-06, "loss": 0.4242, "step": 2281 }, { "epoch": 1.9037819799777531, "grad_norm": 0.3097883462905884, "learning_rate": 3.5401543949605095e-06, "loss": 0.4213, "step": 2282 }, { "epoch": 1.9046162402669633, "grad_norm": 0.3358815014362335, "learning_rate": 3.535509589167374e-06, "loss": 0.4725, "step": 2283 }, { "epoch": 1.9054505005561735, "grad_norm": 0.30562302470207214, "learning_rate": 3.5308661653684114e-06, "loss": 0.4172, "step": 2284 }, { "epoch": 1.906284760845384, "grad_norm": 0.3131335377693176, "learning_rate": 3.526224127945479e-06, "loss": 0.3984, "step": 2285 }, { "epoch": 1.9071190211345939, "grad_norm": 0.34706947207450867, "learning_rate": 3.521583481279125e-06, "loss": 0.4631, "step": 2286 }, { "epoch": 1.9079532814238043, "grad_norm": 0.33318597078323364, "learning_rate": 3.516944229748579e-06, "loss": 0.4194, "step": 2287 }, { "epoch": 1.9087875417130145, "grad_norm": 0.3039299547672272, "learning_rate": 3.5123063777317644e-06, "loss": 0.4076, "step": 2288 }, { "epoch": 1.9096218020022246, "grad_norm": 0.32225292921066284, "learning_rate": 3.5076699296052728e-06, "loss": 0.4251, "step": 2289 }, { "epoch": 1.910456062291435, "grad_norm": 0.3512115478515625, "learning_rate": 3.5030348897443806e-06, "loss": 0.4467, "step": 2290 }, { "epoch": 1.911290322580645, "grad_norm": 0.30354228615760803, "learning_rate": 3.4984012625230284e-06, "loss": 0.3998, "step": 2291 }, { "epoch": 1.9121245828698554, "grad_norm": 0.3285408318042755, "learning_rate": 3.4937690523138302e-06, "loss": 0.3962, "step": 2292 }, { "epoch": 1.9129588431590656, "grad_norm": 0.3510547876358032, "learning_rate": 3.489138263488054e-06, "loss": 0.4636, "step": 2293 }, { "epoch": 1.9137931034482758, "grad_norm": 0.29534974694252014, "learning_rate": 3.4845089004156365e-06, "loss": 0.3463, "step": 2294 }, { "epoch": 1.9146273637374862, "grad_norm": 0.34923630952835083, "learning_rate": 3.4798809674651596e-06, "loss": 0.4199, "step": 2295 }, { "epoch": 1.9154616240266962, "grad_norm": 0.33547571301460266, "learning_rate": 3.475254469003865e-06, "loss": 0.4715, "step": 2296 }, { "epoch": 1.9162958843159066, "grad_norm": 0.30374789237976074, "learning_rate": 3.4706294093976313e-06, "loss": 0.4135, "step": 2297 }, { "epoch": 1.9171301446051168, "grad_norm": 0.30076679587364197, "learning_rate": 3.4660057930109853e-06, "loss": 0.3994, "step": 2298 }, { "epoch": 1.917964404894327, "grad_norm": 0.31907323002815247, "learning_rate": 3.461383624207092e-06, "loss": 0.4353, "step": 2299 }, { "epoch": 1.9187986651835374, "grad_norm": 0.31502845883369446, "learning_rate": 3.456762907347746e-06, "loss": 0.4523, "step": 2300 }, { "epoch": 1.9196329254727473, "grad_norm": 0.3288435935974121, "learning_rate": 3.4521436467933777e-06, "loss": 0.3931, "step": 2301 }, { "epoch": 1.9204671857619577, "grad_norm": 0.3266640305519104, "learning_rate": 3.4475258469030358e-06, "loss": 0.4305, "step": 2302 }, { "epoch": 1.921301446051168, "grad_norm": 0.2940015196800232, "learning_rate": 3.442909512034398e-06, "loss": 0.4058, "step": 2303 }, { "epoch": 1.9221357063403781, "grad_norm": 0.3194087743759155, "learning_rate": 3.438294646543755e-06, "loss": 0.4387, "step": 2304 }, { "epoch": 1.9229699666295885, "grad_norm": 0.3096925914287567, "learning_rate": 3.433681254786015e-06, "loss": 0.4063, "step": 2305 }, { "epoch": 1.9238042269187987, "grad_norm": 0.3414952754974365, "learning_rate": 3.4290693411146882e-06, "loss": 0.459, "step": 2306 }, { "epoch": 1.924638487208009, "grad_norm": 0.35306140780448914, "learning_rate": 3.424458909881897e-06, "loss": 0.4046, "step": 2307 }, { "epoch": 1.925472747497219, "grad_norm": 0.3284316658973694, "learning_rate": 3.4198499654383634e-06, "loss": 0.4528, "step": 2308 }, { "epoch": 1.9263070077864293, "grad_norm": 0.3222755491733551, "learning_rate": 3.415242512133402e-06, "loss": 0.4275, "step": 2309 }, { "epoch": 1.9271412680756397, "grad_norm": 0.36702752113342285, "learning_rate": 3.4106365543149267e-06, "loss": 0.4423, "step": 2310 }, { "epoch": 1.9279755283648499, "grad_norm": 0.37344807386398315, "learning_rate": 3.4060320963294345e-06, "loss": 0.4516, "step": 2311 }, { "epoch": 1.92880978865406, "grad_norm": 0.3229534924030304, "learning_rate": 3.4014291425220126e-06, "loss": 0.4357, "step": 2312 }, { "epoch": 1.9296440489432705, "grad_norm": 0.342750608921051, "learning_rate": 3.3968276972363224e-06, "loss": 0.4393, "step": 2313 }, { "epoch": 1.9304783092324804, "grad_norm": 0.34087005257606506, "learning_rate": 3.392227764814609e-06, "loss": 0.4314, "step": 2314 }, { "epoch": 1.9313125695216908, "grad_norm": 0.32580554485321045, "learning_rate": 3.3876293495976813e-06, "loss": 0.4062, "step": 2315 }, { "epoch": 1.932146829810901, "grad_norm": 0.33676525950431824, "learning_rate": 3.383032455924925e-06, "loss": 0.4369, "step": 2316 }, { "epoch": 1.9329810901001112, "grad_norm": 0.31994810700416565, "learning_rate": 3.378437088134285e-06, "loss": 0.4194, "step": 2317 }, { "epoch": 1.9338153503893216, "grad_norm": 0.3426682949066162, "learning_rate": 3.3738432505622653e-06, "loss": 0.3813, "step": 2318 }, { "epoch": 1.9346496106785316, "grad_norm": 0.33844465017318726, "learning_rate": 3.3692509475439304e-06, "loss": 0.4395, "step": 2319 }, { "epoch": 1.935483870967742, "grad_norm": 0.3103564381599426, "learning_rate": 3.3646601834128924e-06, "loss": 0.3928, "step": 2320 }, { "epoch": 1.9363181312569522, "grad_norm": 0.33597058057785034, "learning_rate": 3.360070962501315e-06, "loss": 0.4578, "step": 2321 }, { "epoch": 1.9371523915461624, "grad_norm": 0.3185846209526062, "learning_rate": 3.355483289139901e-06, "loss": 0.3718, "step": 2322 }, { "epoch": 1.9379866518353728, "grad_norm": 0.319553017616272, "learning_rate": 3.3508971676578993e-06, "loss": 0.435, "step": 2323 }, { "epoch": 1.9388209121245827, "grad_norm": 0.35344862937927246, "learning_rate": 3.346312602383086e-06, "loss": 0.446, "step": 2324 }, { "epoch": 1.9396551724137931, "grad_norm": 0.300991028547287, "learning_rate": 3.341729597641777e-06, "loss": 0.3895, "step": 2325 }, { "epoch": 1.9404894327030033, "grad_norm": 0.31422489881515503, "learning_rate": 3.3371481577588093e-06, "loss": 0.3861, "step": 2326 }, { "epoch": 1.9413236929922135, "grad_norm": 0.3401193618774414, "learning_rate": 3.3325682870575478e-06, "loss": 0.4074, "step": 2327 }, { "epoch": 1.942157953281424, "grad_norm": 0.3518194556236267, "learning_rate": 3.327989989859872e-06, "loss": 0.3704, "step": 2328 }, { "epoch": 1.942992213570634, "grad_norm": 0.33254581689834595, "learning_rate": 3.3234132704861786e-06, "loss": 0.4293, "step": 2329 }, { "epoch": 1.9438264738598443, "grad_norm": 0.293788880109787, "learning_rate": 3.31883813325538e-06, "loss": 0.3745, "step": 2330 }, { "epoch": 1.9446607341490545, "grad_norm": 0.35649773478507996, "learning_rate": 3.314264582484885e-06, "loss": 0.4474, "step": 2331 }, { "epoch": 1.9454949944382647, "grad_norm": 0.3288113474845886, "learning_rate": 3.3096926224906167e-06, "loss": 0.4408, "step": 2332 }, { "epoch": 1.946329254727475, "grad_norm": 0.3167007863521576, "learning_rate": 3.3051222575869897e-06, "loss": 0.4177, "step": 2333 }, { "epoch": 1.947163515016685, "grad_norm": 0.2876431345939636, "learning_rate": 3.3005534920869175e-06, "loss": 0.3758, "step": 2334 }, { "epoch": 1.9479977753058955, "grad_norm": 0.3686605393886566, "learning_rate": 3.295986330301799e-06, "loss": 0.4531, "step": 2335 }, { "epoch": 1.9488320355951056, "grad_norm": 0.33946412801742554, "learning_rate": 3.2914207765415263e-06, "loss": 0.4157, "step": 2336 }, { "epoch": 1.9496662958843158, "grad_norm": 0.325863778591156, "learning_rate": 3.286856835114468e-06, "loss": 0.3756, "step": 2337 }, { "epoch": 1.9505005561735262, "grad_norm": 0.3193451166152954, "learning_rate": 3.2822945103274783e-06, "loss": 0.4398, "step": 2338 }, { "epoch": 1.9513348164627362, "grad_norm": 0.31564050912857056, "learning_rate": 3.277733806485879e-06, "loss": 0.3987, "step": 2339 }, { "epoch": 1.9521690767519466, "grad_norm": 0.34202179312705994, "learning_rate": 3.273174727893463e-06, "loss": 0.457, "step": 2340 }, { "epoch": 1.9530033370411568, "grad_norm": 0.3142654597759247, "learning_rate": 3.268617278852494e-06, "loss": 0.388, "step": 2341 }, { "epoch": 1.953837597330367, "grad_norm": 0.329357773065567, "learning_rate": 3.264061463663693e-06, "loss": 0.4504, "step": 2342 }, { "epoch": 1.9546718576195774, "grad_norm": 0.30531638860702515, "learning_rate": 3.2595072866262456e-06, "loss": 0.4045, "step": 2343 }, { "epoch": 1.9555061179087876, "grad_norm": 0.3112248480319977, "learning_rate": 3.2549547520377823e-06, "loss": 0.4316, "step": 2344 }, { "epoch": 1.9563403781979978, "grad_norm": 0.2975502014160156, "learning_rate": 3.2504038641943927e-06, "loss": 0.3859, "step": 2345 }, { "epoch": 1.957174638487208, "grad_norm": 0.3432139456272125, "learning_rate": 3.2458546273906054e-06, "loss": 0.4235, "step": 2346 }, { "epoch": 1.9580088987764181, "grad_norm": 0.32590776681900024, "learning_rate": 3.2413070459193986e-06, "loss": 0.4163, "step": 2347 }, { "epoch": 1.9588431590656286, "grad_norm": 0.33472320437431335, "learning_rate": 3.2367611240721796e-06, "loss": 0.4569, "step": 2348 }, { "epoch": 1.9596774193548387, "grad_norm": 0.2986064553260803, "learning_rate": 3.232216866138794e-06, "loss": 0.3899, "step": 2349 }, { "epoch": 1.960511679644049, "grad_norm": 0.3079254925251007, "learning_rate": 3.227674276407521e-06, "loss": 0.4291, "step": 2350 }, { "epoch": 1.9613459399332593, "grad_norm": 0.34968146681785583, "learning_rate": 3.2231333591650567e-06, "loss": 0.4529, "step": 2351 }, { "epoch": 1.9621802002224693, "grad_norm": 0.3128223419189453, "learning_rate": 3.2185941186965275e-06, "loss": 0.4049, "step": 2352 }, { "epoch": 1.9630144605116797, "grad_norm": 0.3237152695655823, "learning_rate": 3.214056559285471e-06, "loss": 0.4502, "step": 2353 }, { "epoch": 1.96384872080089, "grad_norm": 0.31862348318099976, "learning_rate": 3.209520685213843e-06, "loss": 0.413, "step": 2354 }, { "epoch": 1.9646829810901, "grad_norm": 0.3319879174232483, "learning_rate": 3.204986500762006e-06, "loss": 0.3956, "step": 2355 }, { "epoch": 1.9655172413793105, "grad_norm": 0.3098748028278351, "learning_rate": 3.200454010208733e-06, "loss": 0.3999, "step": 2356 }, { "epoch": 1.9663515016685205, "grad_norm": 0.334347665309906, "learning_rate": 3.195923217831189e-06, "loss": 0.4305, "step": 2357 }, { "epoch": 1.9671857619577309, "grad_norm": 0.3212642967700958, "learning_rate": 3.1913941279049463e-06, "loss": 0.3965, "step": 2358 }, { "epoch": 1.968020022246941, "grad_norm": 0.32724618911743164, "learning_rate": 3.1868667447039657e-06, "loss": 0.475, "step": 2359 }, { "epoch": 1.9688542825361512, "grad_norm": 0.33089253306388855, "learning_rate": 3.1823410725005964e-06, "loss": 0.3942, "step": 2360 }, { "epoch": 1.9696885428253617, "grad_norm": 0.3488110303878784, "learning_rate": 3.177817115565578e-06, "loss": 0.4806, "step": 2361 }, { "epoch": 1.9705228031145716, "grad_norm": 0.30246812105178833, "learning_rate": 3.173294878168025e-06, "loss": 0.3889, "step": 2362 }, { "epoch": 1.971357063403782, "grad_norm": 0.2995411455631256, "learning_rate": 3.1687743645754367e-06, "loss": 0.4293, "step": 2363 }, { "epoch": 1.9721913236929922, "grad_norm": 0.3292909264564514, "learning_rate": 3.1642555790536767e-06, "loss": 0.4607, "step": 2364 }, { "epoch": 1.9730255839822024, "grad_norm": 0.3087635636329651, "learning_rate": 3.1597385258669872e-06, "loss": 0.4161, "step": 2365 }, { "epoch": 1.9738598442714128, "grad_norm": 0.3047802746295929, "learning_rate": 3.155223209277967e-06, "loss": 0.4117, "step": 2366 }, { "epoch": 1.9746941045606228, "grad_norm": 0.3234541416168213, "learning_rate": 3.150709633547584e-06, "loss": 0.4331, "step": 2367 }, { "epoch": 1.9755283648498332, "grad_norm": 0.3446621894836426, "learning_rate": 3.146197802935157e-06, "loss": 0.4359, "step": 2368 }, { "epoch": 1.9763626251390434, "grad_norm": 0.3143290579319, "learning_rate": 3.141687721698363e-06, "loss": 0.4056, "step": 2369 }, { "epoch": 1.9771968854282536, "grad_norm": 0.31421512365341187, "learning_rate": 3.1371793940932242e-06, "loss": 0.45, "step": 2370 }, { "epoch": 1.978031145717464, "grad_norm": 0.3237009346485138, "learning_rate": 3.132672824374109e-06, "loss": 0.4465, "step": 2371 }, { "epoch": 1.978865406006674, "grad_norm": 0.34867894649505615, "learning_rate": 3.1281680167937296e-06, "loss": 0.4437, "step": 2372 }, { "epoch": 1.9796996662958843, "grad_norm": 0.32679668068885803, "learning_rate": 3.12366497560313e-06, "loss": 0.4422, "step": 2373 }, { "epoch": 1.9805339265850945, "grad_norm": 0.3089198172092438, "learning_rate": 3.1191637050516932e-06, "loss": 0.4192, "step": 2374 }, { "epoch": 1.9813681868743047, "grad_norm": 0.28915393352508545, "learning_rate": 3.1146642093871263e-06, "loss": 0.3865, "step": 2375 }, { "epoch": 1.9822024471635151, "grad_norm": 0.30986833572387695, "learning_rate": 3.110166492855468e-06, "loss": 0.4705, "step": 2376 }, { "epoch": 1.983036707452725, "grad_norm": 0.28476881980895996, "learning_rate": 3.105670559701069e-06, "loss": 0.4044, "step": 2377 }, { "epoch": 1.9838709677419355, "grad_norm": 0.3143027424812317, "learning_rate": 3.1011764141666057e-06, "loss": 0.4144, "step": 2378 }, { "epoch": 1.9847052280311457, "grad_norm": 0.33094650506973267, "learning_rate": 3.0966840604930614e-06, "loss": 0.4037, "step": 2379 }, { "epoch": 1.9855394883203559, "grad_norm": 0.31206846237182617, "learning_rate": 3.09219350291973e-06, "loss": 0.4584, "step": 2380 }, { "epoch": 1.9863737486095663, "grad_norm": 0.27681300044059753, "learning_rate": 3.0877047456842157e-06, "loss": 0.3835, "step": 2381 }, { "epoch": 1.9872080088987765, "grad_norm": 0.3382240831851959, "learning_rate": 3.083217793022415e-06, "loss": 0.4432, "step": 2382 }, { "epoch": 1.9880422691879867, "grad_norm": 0.34051474928855896, "learning_rate": 3.0787326491685287e-06, "loss": 0.4143, "step": 2383 }, { "epoch": 1.9888765294771968, "grad_norm": 0.3428408205509186, "learning_rate": 3.074249318355046e-06, "loss": 0.4707, "step": 2384 }, { "epoch": 1.989710789766407, "grad_norm": 0.2981685996055603, "learning_rate": 3.06976780481275e-06, "loss": 0.3791, "step": 2385 }, { "epoch": 1.9905450500556174, "grad_norm": 0.32209986448287964, "learning_rate": 3.0652881127707023e-06, "loss": 0.3909, "step": 2386 }, { "epoch": 1.9913793103448276, "grad_norm": 0.33470839262008667, "learning_rate": 3.0608102464562527e-06, "loss": 0.4346, "step": 2387 }, { "epoch": 1.9922135706340378, "grad_norm": 0.31811508536338806, "learning_rate": 3.056334210095021e-06, "loss": 0.4234, "step": 2388 }, { "epoch": 1.993047830923248, "grad_norm": 0.3553306460380554, "learning_rate": 3.0518600079109096e-06, "loss": 0.4904, "step": 2389 }, { "epoch": 1.9938820912124582, "grad_norm": 0.3373257517814636, "learning_rate": 3.0473876441260786e-06, "loss": 0.3732, "step": 2390 }, { "epoch": 1.9947163515016686, "grad_norm": 0.34659111499786377, "learning_rate": 3.042917122960961e-06, "loss": 0.4244, "step": 2391 }, { "epoch": 1.9955506117908788, "grad_norm": 0.3290634751319885, "learning_rate": 3.0384484486342515e-06, "loss": 0.3982, "step": 2392 }, { "epoch": 1.996384872080089, "grad_norm": 0.3492671251296997, "learning_rate": 3.033981625362894e-06, "loss": 0.4013, "step": 2393 }, { "epoch": 1.9972191323692994, "grad_norm": 0.3552877902984619, "learning_rate": 3.029516657362097e-06, "loss": 0.3981, "step": 2394 }, { "epoch": 1.9980533926585093, "grad_norm": 0.33562472462654114, "learning_rate": 3.0250535488453077e-06, "loss": 0.4308, "step": 2395 }, { "epoch": 1.9988876529477198, "grad_norm": 0.34113645553588867, "learning_rate": 3.020592304024226e-06, "loss": 0.4568, "step": 2396 }, { "epoch": 1.99972191323693, "grad_norm": 0.4315108060836792, "learning_rate": 3.016132927108787e-06, "loss": 0.5246, "step": 2397 }, { "epoch": 2.00055617352614, "grad_norm": 0.39825209975242615, "learning_rate": 3.011675422307172e-06, "loss": 0.4446, "step": 2398 }, { "epoch": 2.0013904338153505, "grad_norm": 0.3343384861946106, "learning_rate": 3.007219793825784e-06, "loss": 0.375, "step": 2399 }, { "epoch": 2.0022246941045605, "grad_norm": 0.32050323486328125, "learning_rate": 3.0027660458692653e-06, "loss": 0.3833, "step": 2400 }, { "epoch": 2.003058954393771, "grad_norm": 0.3377557396888733, "learning_rate": 2.9983141826404786e-06, "loss": 0.3854, "step": 2401 }, { "epoch": 2.003893214682981, "grad_norm": 0.33381161093711853, "learning_rate": 2.9938642083405077e-06, "loss": 0.4179, "step": 2402 }, { "epoch": 2.0047274749721913, "grad_norm": 0.3337341248989105, "learning_rate": 2.9894161271686577e-06, "loss": 0.4141, "step": 2403 }, { "epoch": 2.0055617352614017, "grad_norm": 0.3213748037815094, "learning_rate": 2.9849699433224423e-06, "loss": 0.3834, "step": 2404 }, { "epoch": 2.0063959955506117, "grad_norm": 0.29469457268714905, "learning_rate": 2.980525660997591e-06, "loss": 0.3582, "step": 2405 }, { "epoch": 2.007230255839822, "grad_norm": 0.34047725796699524, "learning_rate": 2.976083284388031e-06, "loss": 0.4417, "step": 2406 }, { "epoch": 2.0080645161290325, "grad_norm": 0.34747496247291565, "learning_rate": 2.971642817685899e-06, "loss": 0.3665, "step": 2407 }, { "epoch": 2.0088987764182424, "grad_norm": 0.3109015226364136, "learning_rate": 2.9672042650815213e-06, "loss": 0.3593, "step": 2408 }, { "epoch": 2.009733036707453, "grad_norm": 0.31753817200660706, "learning_rate": 2.962767630763427e-06, "loss": 0.4034, "step": 2409 }, { "epoch": 2.010567296996663, "grad_norm": 0.30557411909103394, "learning_rate": 2.9583329189183284e-06, "loss": 0.4049, "step": 2410 }, { "epoch": 2.011401557285873, "grad_norm": 0.30013924837112427, "learning_rate": 2.9539001337311234e-06, "loss": 0.3979, "step": 2411 }, { "epoch": 2.0122358175750836, "grad_norm": 0.32679712772369385, "learning_rate": 2.9494692793848968e-06, "loss": 0.4098, "step": 2412 }, { "epoch": 2.0130700778642936, "grad_norm": 0.324676513671875, "learning_rate": 2.945040360060907e-06, "loss": 0.4078, "step": 2413 }, { "epoch": 2.013904338153504, "grad_norm": 0.33281078934669495, "learning_rate": 2.9406133799385894e-06, "loss": 0.4338, "step": 2414 }, { "epoch": 2.014738598442714, "grad_norm": 0.3054578900337219, "learning_rate": 2.9361883431955444e-06, "loss": 0.3823, "step": 2415 }, { "epoch": 2.0155728587319244, "grad_norm": 0.31917184591293335, "learning_rate": 2.9317652540075458e-06, "loss": 0.3898, "step": 2416 }, { "epoch": 2.016407119021135, "grad_norm": 0.3192262351512909, "learning_rate": 2.9273441165485227e-06, "loss": 0.3741, "step": 2417 }, { "epoch": 2.0172413793103448, "grad_norm": 0.3248962461948395, "learning_rate": 2.9229249349905686e-06, "loss": 0.4095, "step": 2418 }, { "epoch": 2.018075639599555, "grad_norm": 0.3395436704158783, "learning_rate": 2.918507713503924e-06, "loss": 0.4455, "step": 2419 }, { "epoch": 2.018909899888765, "grad_norm": 0.3182157278060913, "learning_rate": 2.914092456256987e-06, "loss": 0.3631, "step": 2420 }, { "epoch": 2.0197441601779755, "grad_norm": 0.32523906230926514, "learning_rate": 2.9096791674163004e-06, "loss": 0.3727, "step": 2421 }, { "epoch": 2.020578420467186, "grad_norm": 0.34601420164108276, "learning_rate": 2.9052678511465404e-06, "loss": 0.408, "step": 2422 }, { "epoch": 2.021412680756396, "grad_norm": 0.34236785769462585, "learning_rate": 2.900858511610537e-06, "loss": 0.3705, "step": 2423 }, { "epoch": 2.0222469410456063, "grad_norm": 0.3442624807357788, "learning_rate": 2.8964511529692454e-06, "loss": 0.442, "step": 2424 }, { "epoch": 2.0230812013348163, "grad_norm": 0.3221091032028198, "learning_rate": 2.8920457793817507e-06, "loss": 0.3841, "step": 2425 }, { "epoch": 2.0239154616240267, "grad_norm": 0.3166672885417938, "learning_rate": 2.88764239500527e-06, "loss": 0.3636, "step": 2426 }, { "epoch": 2.024749721913237, "grad_norm": 0.34055453538894653, "learning_rate": 2.8832410039951387e-06, "loss": 0.4223, "step": 2427 }, { "epoch": 2.025583982202447, "grad_norm": 0.32994741201400757, "learning_rate": 2.8788416105048124e-06, "loss": 0.4183, "step": 2428 }, { "epoch": 2.0264182424916575, "grad_norm": 0.3013457655906677, "learning_rate": 2.874444218685868e-06, "loss": 0.4106, "step": 2429 }, { "epoch": 2.0272525027808674, "grad_norm": 0.33392810821533203, "learning_rate": 2.870048832687978e-06, "loss": 0.3687, "step": 2430 }, { "epoch": 2.028086763070078, "grad_norm": 0.3370850086212158, "learning_rate": 2.8656554566589426e-06, "loss": 0.4216, "step": 2431 }, { "epoch": 2.0289210233592883, "grad_norm": 0.3084275424480438, "learning_rate": 2.861264094744647e-06, "loss": 0.412, "step": 2432 }, { "epoch": 2.029755283648498, "grad_norm": 0.2975429892539978, "learning_rate": 2.8568747510890826e-06, "loss": 0.3599, "step": 2433 }, { "epoch": 2.0305895439377086, "grad_norm": 0.3379700183868408, "learning_rate": 2.8524874298343442e-06, "loss": 0.4248, "step": 2434 }, { "epoch": 2.0314238042269186, "grad_norm": 0.28722667694091797, "learning_rate": 2.8481021351206024e-06, "loss": 0.3302, "step": 2435 }, { "epoch": 2.032258064516129, "grad_norm": 0.3404325842857361, "learning_rate": 2.843718871086129e-06, "loss": 0.4114, "step": 2436 }, { "epoch": 2.0330923248053394, "grad_norm": 0.302258163690567, "learning_rate": 2.8393376418672734e-06, "loss": 0.408, "step": 2437 }, { "epoch": 2.0339265850945494, "grad_norm": 0.2832605242729187, "learning_rate": 2.8349584515984653e-06, "loss": 0.3729, "step": 2438 }, { "epoch": 2.03476084538376, "grad_norm": 0.3298014998435974, "learning_rate": 2.83058130441221e-06, "loss": 0.4303, "step": 2439 }, { "epoch": 2.0355951056729698, "grad_norm": 0.30156224966049194, "learning_rate": 2.8262062044390864e-06, "loss": 0.3788, "step": 2440 }, { "epoch": 2.03642936596218, "grad_norm": 0.3153831362724304, "learning_rate": 2.8218331558077407e-06, "loss": 0.4154, "step": 2441 }, { "epoch": 2.0372636262513906, "grad_norm": 0.31013694405555725, "learning_rate": 2.817462162644883e-06, "loss": 0.3885, "step": 2442 }, { "epoch": 2.0380978865406005, "grad_norm": 0.32374078035354614, "learning_rate": 2.813093229075283e-06, "loss": 0.3794, "step": 2443 }, { "epoch": 2.038932146829811, "grad_norm": 0.37760987877845764, "learning_rate": 2.8087263592217696e-06, "loss": 0.4319, "step": 2444 }, { "epoch": 2.0397664071190214, "grad_norm": 0.3045150637626648, "learning_rate": 2.8043615572052217e-06, "loss": 0.3929, "step": 2445 }, { "epoch": 2.0406006674082313, "grad_norm": 0.32215702533721924, "learning_rate": 2.7999988271445643e-06, "loss": 0.3518, "step": 2446 }, { "epoch": 2.0414349276974417, "grad_norm": 0.32626840472221375, "learning_rate": 2.795638173156779e-06, "loss": 0.4248, "step": 2447 }, { "epoch": 2.0422691879866517, "grad_norm": 0.34439319372177124, "learning_rate": 2.79127959935687e-06, "loss": 0.3961, "step": 2448 }, { "epoch": 2.043103448275862, "grad_norm": 0.32189396023750305, "learning_rate": 2.786923109857893e-06, "loss": 0.3682, "step": 2449 }, { "epoch": 2.0439377085650725, "grad_norm": 0.33588990569114685, "learning_rate": 2.782568708770933e-06, "loss": 0.3845, "step": 2450 }, { "epoch": 2.0447719688542825, "grad_norm": 0.31993570923805237, "learning_rate": 2.778216400205101e-06, "loss": 0.4235, "step": 2451 }, { "epoch": 2.045606229143493, "grad_norm": 0.3332087993621826, "learning_rate": 2.7738661882675366e-06, "loss": 0.398, "step": 2452 }, { "epoch": 2.046440489432703, "grad_norm": 0.3746569752693176, "learning_rate": 2.7695180770633993e-06, "loss": 0.4263, "step": 2453 }, { "epoch": 2.0472747497219133, "grad_norm": 0.3593369722366333, "learning_rate": 2.7651720706958676e-06, "loss": 0.4041, "step": 2454 }, { "epoch": 2.0481090100111237, "grad_norm": 0.31307706236839294, "learning_rate": 2.7608281732661313e-06, "loss": 0.3653, "step": 2455 }, { "epoch": 2.0489432703003336, "grad_norm": 0.29822906851768494, "learning_rate": 2.756486388873394e-06, "loss": 0.3903, "step": 2456 }, { "epoch": 2.049777530589544, "grad_norm": 0.3027375042438507, "learning_rate": 2.7521467216148606e-06, "loss": 0.3908, "step": 2457 }, { "epoch": 2.050611790878754, "grad_norm": 0.33029675483703613, "learning_rate": 2.747809175585742e-06, "loss": 0.3848, "step": 2458 }, { "epoch": 2.0514460511679644, "grad_norm": 0.3474617898464203, "learning_rate": 2.7434737548792434e-06, "loss": 0.4246, "step": 2459 }, { "epoch": 2.052280311457175, "grad_norm": 0.31078991293907166, "learning_rate": 2.7391404635865725e-06, "loss": 0.3727, "step": 2460 }, { "epoch": 2.053114571746385, "grad_norm": 0.3351077735424042, "learning_rate": 2.734809305796915e-06, "loss": 0.4488, "step": 2461 }, { "epoch": 2.053948832035595, "grad_norm": 0.30430564284324646, "learning_rate": 2.730480285597455e-06, "loss": 0.379, "step": 2462 }, { "epoch": 2.054783092324805, "grad_norm": 0.3188287615776062, "learning_rate": 2.7261534070733547e-06, "loss": 0.4034, "step": 2463 }, { "epoch": 2.0556173526140156, "grad_norm": 0.3313559293746948, "learning_rate": 2.7218286743077495e-06, "loss": 0.3933, "step": 2464 }, { "epoch": 2.056451612903226, "grad_norm": 0.32156574726104736, "learning_rate": 2.717506091381762e-06, "loss": 0.4218, "step": 2465 }, { "epoch": 2.057285873192436, "grad_norm": 0.32254141569137573, "learning_rate": 2.7131856623744762e-06, "loss": 0.414, "step": 2466 }, { "epoch": 2.0581201334816464, "grad_norm": 0.2839086949825287, "learning_rate": 2.708867391362948e-06, "loss": 0.3475, "step": 2467 }, { "epoch": 2.0589543937708563, "grad_norm": 0.3326614499092102, "learning_rate": 2.7045512824221964e-06, "loss": 0.3987, "step": 2468 }, { "epoch": 2.0597886540600667, "grad_norm": 0.3299097716808319, "learning_rate": 2.7002373396251984e-06, "loss": 0.3757, "step": 2469 }, { "epoch": 2.060622914349277, "grad_norm": 0.3299124538898468, "learning_rate": 2.6959255670428887e-06, "loss": 0.4234, "step": 2470 }, { "epoch": 2.061457174638487, "grad_norm": 0.3157268464565277, "learning_rate": 2.6916159687441524e-06, "loss": 0.3921, "step": 2471 }, { "epoch": 2.0622914349276975, "grad_norm": 0.3450485169887543, "learning_rate": 2.687308548795825e-06, "loss": 0.3953, "step": 2472 }, { "epoch": 2.0631256952169075, "grad_norm": 0.298530638217926, "learning_rate": 2.683003311262684e-06, "loss": 0.3926, "step": 2473 }, { "epoch": 2.063959955506118, "grad_norm": 0.3111628293991089, "learning_rate": 2.678700260207449e-06, "loss": 0.3872, "step": 2474 }, { "epoch": 2.0647942157953283, "grad_norm": 0.34809643030166626, "learning_rate": 2.6743993996907742e-06, "loss": 0.3962, "step": 2475 }, { "epoch": 2.0656284760845383, "grad_norm": 0.33127614855766296, "learning_rate": 2.6701007337712536e-06, "loss": 0.371, "step": 2476 }, { "epoch": 2.0664627363737487, "grad_norm": 0.32816532254219055, "learning_rate": 2.6658042665053974e-06, "loss": 0.4424, "step": 2477 }, { "epoch": 2.0672969966629586, "grad_norm": 0.29663681983947754, "learning_rate": 2.6615100019476537e-06, "loss": 0.3527, "step": 2478 }, { "epoch": 2.068131256952169, "grad_norm": 0.3341652452945709, "learning_rate": 2.6572179441503852e-06, "loss": 0.4198, "step": 2479 }, { "epoch": 2.0689655172413794, "grad_norm": 0.3001740276813507, "learning_rate": 2.652928097163875e-06, "loss": 0.3563, "step": 2480 }, { "epoch": 2.0697997775305894, "grad_norm": 0.3564237952232361, "learning_rate": 2.648640465036316e-06, "loss": 0.4219, "step": 2481 }, { "epoch": 2.0706340378198, "grad_norm": 0.2996433973312378, "learning_rate": 2.6443550518138163e-06, "loss": 0.3661, "step": 2482 }, { "epoch": 2.07146829810901, "grad_norm": 0.33224916458129883, "learning_rate": 2.6400718615403852e-06, "loss": 0.4473, "step": 2483 }, { "epoch": 2.07230255839822, "grad_norm": 0.32246267795562744, "learning_rate": 2.6357908982579385e-06, "loss": 0.3875, "step": 2484 }, { "epoch": 2.0731368186874306, "grad_norm": 0.3157311975955963, "learning_rate": 2.631512166006287e-06, "loss": 0.407, "step": 2485 }, { "epoch": 2.0739710789766406, "grad_norm": 0.2806086540222168, "learning_rate": 2.6272356688231384e-06, "loss": 0.3349, "step": 2486 }, { "epoch": 2.074805339265851, "grad_norm": 0.3477040231227875, "learning_rate": 2.6229614107440898e-06, "loss": 0.4263, "step": 2487 }, { "epoch": 2.0756395995550614, "grad_norm": 0.32503464818000793, "learning_rate": 2.6186893958026245e-06, "loss": 0.4203, "step": 2488 }, { "epoch": 2.0764738598442714, "grad_norm": 0.2961250841617584, "learning_rate": 2.6144196280301158e-06, "loss": 0.3754, "step": 2489 }, { "epoch": 2.0773081201334818, "grad_norm": 0.29745084047317505, "learning_rate": 2.610152111455805e-06, "loss": 0.3718, "step": 2490 }, { "epoch": 2.0781423804226917, "grad_norm": 0.3539210259914398, "learning_rate": 2.605886850106818e-06, "loss": 0.4078, "step": 2491 }, { "epoch": 2.078976640711902, "grad_norm": 0.3128289580345154, "learning_rate": 2.601623848008152e-06, "loss": 0.3641, "step": 2492 }, { "epoch": 2.0798109010011125, "grad_norm": 0.2966078221797943, "learning_rate": 2.597363109182664e-06, "loss": 0.4002, "step": 2493 }, { "epoch": 2.0806451612903225, "grad_norm": 0.30546656250953674, "learning_rate": 2.5931046376510875e-06, "loss": 0.3952, "step": 2494 }, { "epoch": 2.081479421579533, "grad_norm": 0.38114428520202637, "learning_rate": 2.5888484374320033e-06, "loss": 0.4121, "step": 2495 }, { "epoch": 2.082313681868743, "grad_norm": 0.3055715262889862, "learning_rate": 2.5845945125418603e-06, "loss": 0.389, "step": 2496 }, { "epoch": 2.0831479421579533, "grad_norm": 0.3318061828613281, "learning_rate": 2.580342866994955e-06, "loss": 0.407, "step": 2497 }, { "epoch": 2.0839822024471637, "grad_norm": 0.31239941716194153, "learning_rate": 2.576093504803432e-06, "loss": 0.3664, "step": 2498 }, { "epoch": 2.0848164627363737, "grad_norm": 0.3259067237377167, "learning_rate": 2.5718464299772827e-06, "loss": 0.4013, "step": 2499 }, { "epoch": 2.085650723025584, "grad_norm": 0.32831817865371704, "learning_rate": 2.5676016465243414e-06, "loss": 0.399, "step": 2500 }, { "epoch": 2.086484983314794, "grad_norm": 0.30808135867118835, "learning_rate": 2.5633591584502753e-06, "loss": 0.3623, "step": 2501 }, { "epoch": 2.0873192436040044, "grad_norm": 0.36151373386383057, "learning_rate": 2.559118969758595e-06, "loss": 0.419, "step": 2502 }, { "epoch": 2.088153503893215, "grad_norm": 0.29532986879348755, "learning_rate": 2.554881084450629e-06, "loss": 0.3729, "step": 2503 }, { "epoch": 2.088987764182425, "grad_norm": 0.2804437577724457, "learning_rate": 2.550645506525538e-06, "loss": 0.3601, "step": 2504 }, { "epoch": 2.0898220244716352, "grad_norm": 0.2951928675174713, "learning_rate": 2.5464122399803126e-06, "loss": 0.4169, "step": 2505 }, { "epoch": 2.090656284760845, "grad_norm": 0.31487569212913513, "learning_rate": 2.542181288809745e-06, "loss": 0.4232, "step": 2506 }, { "epoch": 2.0914905450500556, "grad_norm": 0.27493560314178467, "learning_rate": 2.5379526570064635e-06, "loss": 0.3683, "step": 2507 }, { "epoch": 2.092324805339266, "grad_norm": 0.30253174901008606, "learning_rate": 2.5337263485608864e-06, "loss": 0.3855, "step": 2508 }, { "epoch": 2.093159065628476, "grad_norm": 0.3403005599975586, "learning_rate": 2.5295023674612568e-06, "loss": 0.4202, "step": 2509 }, { "epoch": 2.0939933259176864, "grad_norm": 0.28903496265411377, "learning_rate": 2.5252807176936124e-06, "loss": 0.339, "step": 2510 }, { "epoch": 2.0948275862068964, "grad_norm": 0.3059140741825104, "learning_rate": 2.5210614032417933e-06, "loss": 0.3941, "step": 2511 }, { "epoch": 2.0956618464961068, "grad_norm": 0.30906304717063904, "learning_rate": 2.516844428087436e-06, "loss": 0.4183, "step": 2512 }, { "epoch": 2.096496106785317, "grad_norm": 0.3132782280445099, "learning_rate": 2.5126297962099684e-06, "loss": 0.3889, "step": 2513 }, { "epoch": 2.097330367074527, "grad_norm": 0.31022217869758606, "learning_rate": 2.508417511586608e-06, "loss": 0.3937, "step": 2514 }, { "epoch": 2.0981646273637375, "grad_norm": 0.2862684428691864, "learning_rate": 2.504207578192358e-06, "loss": 0.3641, "step": 2515 }, { "epoch": 2.0989988876529475, "grad_norm": 0.3280887007713318, "learning_rate": 2.5000000000000015e-06, "loss": 0.444, "step": 2516 }, { "epoch": 2.099833147942158, "grad_norm": 0.3240327537059784, "learning_rate": 2.495794780980097e-06, "loss": 0.3747, "step": 2517 }, { "epoch": 2.1006674082313683, "grad_norm": 0.3226510286331177, "learning_rate": 2.491591925100985e-06, "loss": 0.3928, "step": 2518 }, { "epoch": 2.1015016685205783, "grad_norm": 0.3300107419490814, "learning_rate": 2.487391436328764e-06, "loss": 0.3938, "step": 2519 }, { "epoch": 2.1023359288097887, "grad_norm": 0.3120187222957611, "learning_rate": 2.4831933186273095e-06, "loss": 0.3848, "step": 2520 }, { "epoch": 2.103170189098999, "grad_norm": 0.33179065585136414, "learning_rate": 2.478997575958253e-06, "loss": 0.3931, "step": 2521 }, { "epoch": 2.104004449388209, "grad_norm": 0.2924414277076721, "learning_rate": 2.474804212280987e-06, "loss": 0.346, "step": 2522 }, { "epoch": 2.1048387096774195, "grad_norm": 0.3198830783367157, "learning_rate": 2.470613231552661e-06, "loss": 0.4335, "step": 2523 }, { "epoch": 2.1056729699666294, "grad_norm": 0.28610658645629883, "learning_rate": 2.466424637728167e-06, "loss": 0.3619, "step": 2524 }, { "epoch": 2.10650723025584, "grad_norm": 0.3129299581050873, "learning_rate": 2.4622384347601576e-06, "loss": 0.3695, "step": 2525 }, { "epoch": 2.1073414905450503, "grad_norm": 0.3356875777244568, "learning_rate": 2.45805462659902e-06, "loss": 0.3814, "step": 2526 }, { "epoch": 2.1081757508342602, "grad_norm": 0.31215232610702515, "learning_rate": 2.4538732171928847e-06, "loss": 0.4359, "step": 2527 }, { "epoch": 2.1090100111234706, "grad_norm": 0.2950201630592346, "learning_rate": 2.4496942104876182e-06, "loss": 0.3917, "step": 2528 }, { "epoch": 2.1098442714126806, "grad_norm": 0.30241620540618896, "learning_rate": 2.4455176104268193e-06, "loss": 0.3822, "step": 2529 }, { "epoch": 2.110678531701891, "grad_norm": 0.34169334173202515, "learning_rate": 2.4413434209518137e-06, "loss": 0.4553, "step": 2530 }, { "epoch": 2.1115127919911014, "grad_norm": 0.2908966839313507, "learning_rate": 2.437171646001661e-06, "loss": 0.3648, "step": 2531 }, { "epoch": 2.1123470522803114, "grad_norm": 0.2899493873119354, "learning_rate": 2.433002289513127e-06, "loss": 0.3608, "step": 2532 }, { "epoch": 2.113181312569522, "grad_norm": 0.33270958065986633, "learning_rate": 2.428835355420711e-06, "loss": 0.4032, "step": 2533 }, { "epoch": 2.1140155728587318, "grad_norm": 0.31102126836776733, "learning_rate": 2.4246708476566187e-06, "loss": 0.3584, "step": 2534 }, { "epoch": 2.114849833147942, "grad_norm": 0.3072962164878845, "learning_rate": 2.420508770150762e-06, "loss": 0.3596, "step": 2535 }, { "epoch": 2.1156840934371526, "grad_norm": 0.3226959705352783, "learning_rate": 2.416349126830771e-06, "loss": 0.4389, "step": 2536 }, { "epoch": 2.1165183537263625, "grad_norm": 0.3174997568130493, "learning_rate": 2.4121919216219646e-06, "loss": 0.4182, "step": 2537 }, { "epoch": 2.117352614015573, "grad_norm": 0.33749720454216003, "learning_rate": 2.408037158447375e-06, "loss": 0.4372, "step": 2538 }, { "epoch": 2.118186874304783, "grad_norm": 0.30083781480789185, "learning_rate": 2.4038848412277205e-06, "loss": 0.3656, "step": 2539 }, { "epoch": 2.1190211345939933, "grad_norm": 0.30593377351760864, "learning_rate": 2.3997349738814146e-06, "loss": 0.3941, "step": 2540 }, { "epoch": 2.1198553948832037, "grad_norm": 0.3079013526439667, "learning_rate": 2.3955875603245582e-06, "loss": 0.3914, "step": 2541 }, { "epoch": 2.1206896551724137, "grad_norm": 0.3344835937023163, "learning_rate": 2.3914426044709364e-06, "loss": 0.3751, "step": 2542 }, { "epoch": 2.121523915461624, "grad_norm": 0.3233895003795624, "learning_rate": 2.387300110232015e-06, "loss": 0.3758, "step": 2543 }, { "epoch": 2.122358175750834, "grad_norm": 0.3278207778930664, "learning_rate": 2.383160081516941e-06, "loss": 0.3955, "step": 2544 }, { "epoch": 2.1231924360400445, "grad_norm": 0.3323962688446045, "learning_rate": 2.379022522232527e-06, "loss": 0.4306, "step": 2545 }, { "epoch": 2.124026696329255, "grad_norm": 0.29836446046829224, "learning_rate": 2.374887436283258e-06, "loss": 0.3723, "step": 2546 }, { "epoch": 2.124860956618465, "grad_norm": 0.3010885715484619, "learning_rate": 2.370754827571293e-06, "loss": 0.3987, "step": 2547 }, { "epoch": 2.1256952169076753, "grad_norm": 0.33422359824180603, "learning_rate": 2.3666246999964392e-06, "loss": 0.4309, "step": 2548 }, { "epoch": 2.1265294771968852, "grad_norm": 0.29850563406944275, "learning_rate": 2.3624970574561773e-06, "loss": 0.3588, "step": 2549 }, { "epoch": 2.1273637374860956, "grad_norm": 0.3201923966407776, "learning_rate": 2.3583719038456277e-06, "loss": 0.3787, "step": 2550 }, { "epoch": 2.128197997775306, "grad_norm": 0.3185766041278839, "learning_rate": 2.3542492430575752e-06, "loss": 0.4116, "step": 2551 }, { "epoch": 2.129032258064516, "grad_norm": 0.3188939094543457, "learning_rate": 2.350129078982446e-06, "loss": 0.4148, "step": 2552 }, { "epoch": 2.1298665183537264, "grad_norm": 0.28966519236564636, "learning_rate": 2.34601141550831e-06, "loss": 0.3869, "step": 2553 }, { "epoch": 2.1307007786429364, "grad_norm": 0.3098064959049225, "learning_rate": 2.34189625652088e-06, "loss": 0.4347, "step": 2554 }, { "epoch": 2.131535038932147, "grad_norm": 0.28854382038116455, "learning_rate": 2.3377836059035025e-06, "loss": 0.3955, "step": 2555 }, { "epoch": 2.132369299221357, "grad_norm": 0.3675748407840729, "learning_rate": 2.3336734675371584e-06, "loss": 0.4749, "step": 2556 }, { "epoch": 2.133203559510567, "grad_norm": 0.30148911476135254, "learning_rate": 2.329565845300459e-06, "loss": 0.3571, "step": 2557 }, { "epoch": 2.1340378197997776, "grad_norm": 0.3329126536846161, "learning_rate": 2.3254607430696393e-06, "loss": 0.3848, "step": 2558 }, { "epoch": 2.1348720800889875, "grad_norm": 0.32700321078300476, "learning_rate": 2.321358164718554e-06, "loss": 0.447, "step": 2559 }, { "epoch": 2.135706340378198, "grad_norm": 0.3068147599697113, "learning_rate": 2.317258114118686e-06, "loss": 0.3878, "step": 2560 }, { "epoch": 2.1365406006674084, "grad_norm": 0.314846932888031, "learning_rate": 2.3131605951391184e-06, "loss": 0.3757, "step": 2561 }, { "epoch": 2.1373748609566183, "grad_norm": 0.3637123107910156, "learning_rate": 2.3090656116465597e-06, "loss": 0.4235, "step": 2562 }, { "epoch": 2.1382091212458287, "grad_norm": 0.2920368015766144, "learning_rate": 2.3049731675053116e-06, "loss": 0.3798, "step": 2563 }, { "epoch": 2.139043381535039, "grad_norm": 0.32462379336357117, "learning_rate": 2.3008832665772923e-06, "loss": 0.4419, "step": 2564 }, { "epoch": 2.139877641824249, "grad_norm": 0.2956613004207611, "learning_rate": 2.296795912722014e-06, "loss": 0.3714, "step": 2565 }, { "epoch": 2.1407119021134595, "grad_norm": 0.33574244379997253, "learning_rate": 2.2927111097965803e-06, "loss": 0.4048, "step": 2566 }, { "epoch": 2.1415461624026695, "grad_norm": 0.3212191164493561, "learning_rate": 2.2886288616556996e-06, "loss": 0.4178, "step": 2567 }, { "epoch": 2.14238042269188, "grad_norm": 0.31559816002845764, "learning_rate": 2.2845491721516595e-06, "loss": 0.3717, "step": 2568 }, { "epoch": 2.1432146829810903, "grad_norm": 0.32461607456207275, "learning_rate": 2.280472045134337e-06, "loss": 0.4146, "step": 2569 }, { "epoch": 2.1440489432703003, "grad_norm": 0.3373064398765564, "learning_rate": 2.276397484451191e-06, "loss": 0.4236, "step": 2570 }, { "epoch": 2.1448832035595107, "grad_norm": 0.30460992455482483, "learning_rate": 2.272325493947257e-06, "loss": 0.3932, "step": 2571 }, { "epoch": 2.1457174638487206, "grad_norm": 0.30839595198631287, "learning_rate": 2.2682560774651458e-06, "loss": 0.3455, "step": 2572 }, { "epoch": 2.146551724137931, "grad_norm": 0.33734768629074097, "learning_rate": 2.2641892388450433e-06, "loss": 0.4068, "step": 2573 }, { "epoch": 2.1473859844271415, "grad_norm": 0.33553802967071533, "learning_rate": 2.260124981924693e-06, "loss": 0.4087, "step": 2574 }, { "epoch": 2.1482202447163514, "grad_norm": 0.31311845779418945, "learning_rate": 2.2560633105394143e-06, "loss": 0.4025, "step": 2575 }, { "epoch": 2.149054505005562, "grad_norm": 0.30893412232398987, "learning_rate": 2.2520042285220754e-06, "loss": 0.3893, "step": 2576 }, { "epoch": 2.149888765294772, "grad_norm": 0.30590641498565674, "learning_rate": 2.247947739703106e-06, "loss": 0.3773, "step": 2577 }, { "epoch": 2.150723025583982, "grad_norm": 0.3323810398578644, "learning_rate": 2.2438938479104957e-06, "loss": 0.4019, "step": 2578 }, { "epoch": 2.1515572858731926, "grad_norm": 0.3288833200931549, "learning_rate": 2.2398425569697667e-06, "loss": 0.3968, "step": 2579 }, { "epoch": 2.1523915461624026, "grad_norm": 0.3134523928165436, "learning_rate": 2.235793870704004e-06, "loss": 0.4182, "step": 2580 }, { "epoch": 2.153225806451613, "grad_norm": 0.303055077791214, "learning_rate": 2.2317477929338236e-06, "loss": 0.3838, "step": 2581 }, { "epoch": 2.154060066740823, "grad_norm": 0.3088069260120392, "learning_rate": 2.2277043274773856e-06, "loss": 0.3836, "step": 2582 }, { "epoch": 2.1548943270300334, "grad_norm": 0.3468850553035736, "learning_rate": 2.2236634781503825e-06, "loss": 0.4334, "step": 2583 }, { "epoch": 2.1557285873192438, "grad_norm": 0.28516802191734314, "learning_rate": 2.219625248766037e-06, "loss": 0.3664, "step": 2584 }, { "epoch": 2.1565628476084537, "grad_norm": 0.313862144947052, "learning_rate": 2.215589643135102e-06, "loss": 0.4055, "step": 2585 }, { "epoch": 2.157397107897664, "grad_norm": 0.30492910742759705, "learning_rate": 2.211556665065854e-06, "loss": 0.382, "step": 2586 }, { "epoch": 2.158231368186874, "grad_norm": 0.3046795427799225, "learning_rate": 2.207526318364088e-06, "loss": 0.4386, "step": 2587 }, { "epoch": 2.1590656284760845, "grad_norm": 0.3131970465183258, "learning_rate": 2.203498606833118e-06, "loss": 0.3779, "step": 2588 }, { "epoch": 2.159899888765295, "grad_norm": 0.30875009298324585, "learning_rate": 2.1994735342737693e-06, "loss": 0.4105, "step": 2589 }, { "epoch": 2.160734149054505, "grad_norm": 0.30882230401039124, "learning_rate": 2.195451104484378e-06, "loss": 0.3712, "step": 2590 }, { "epoch": 2.1615684093437153, "grad_norm": 0.29419732093811035, "learning_rate": 2.1914313212607908e-06, "loss": 0.3863, "step": 2591 }, { "epoch": 2.1624026696329253, "grad_norm": 0.31885549426078796, "learning_rate": 2.1874141883963447e-06, "loss": 0.3742, "step": 2592 }, { "epoch": 2.1632369299221357, "grad_norm": 0.32298344373703003, "learning_rate": 2.1833997096818897e-06, "loss": 0.4163, "step": 2593 }, { "epoch": 2.164071190211346, "grad_norm": 0.2968355119228363, "learning_rate": 2.1793878889057628e-06, "loss": 0.4065, "step": 2594 }, { "epoch": 2.164905450500556, "grad_norm": 0.28976723551750183, "learning_rate": 2.175378729853794e-06, "loss": 0.3884, "step": 2595 }, { "epoch": 2.1657397107897665, "grad_norm": 0.3227960467338562, "learning_rate": 2.1713722363093044e-06, "loss": 0.387, "step": 2596 }, { "epoch": 2.166573971078977, "grad_norm": 0.3171670734882355, "learning_rate": 2.1673684120530954e-06, "loss": 0.418, "step": 2597 }, { "epoch": 2.167408231368187, "grad_norm": 0.2828860878944397, "learning_rate": 2.1633672608634524e-06, "loss": 0.387, "step": 2598 }, { "epoch": 2.1682424916573972, "grad_norm": 0.3218068778514862, "learning_rate": 2.1593687865161383e-06, "loss": 0.3753, "step": 2599 }, { "epoch": 2.169076751946607, "grad_norm": 0.34936580061912537, "learning_rate": 2.1553729927843894e-06, "loss": 0.4356, "step": 2600 }, { "epoch": 2.1699110122358176, "grad_norm": 0.3159432113170624, "learning_rate": 2.1513798834389115e-06, "loss": 0.3909, "step": 2601 }, { "epoch": 2.1707452725250276, "grad_norm": 0.3262786567211151, "learning_rate": 2.147389462247878e-06, "loss": 0.4059, "step": 2602 }, { "epoch": 2.171579532814238, "grad_norm": 0.3098350763320923, "learning_rate": 2.1434017329769246e-06, "loss": 0.392, "step": 2603 }, { "epoch": 2.1724137931034484, "grad_norm": 0.3304145336151123, "learning_rate": 2.139416699389153e-06, "loss": 0.3647, "step": 2604 }, { "epoch": 2.1732480533926584, "grad_norm": 0.33574002981185913, "learning_rate": 2.135434365245108e-06, "loss": 0.4037, "step": 2605 }, { "epoch": 2.1740823136818688, "grad_norm": 0.31210267543792725, "learning_rate": 2.131454734302801e-06, "loss": 0.3996, "step": 2606 }, { "epoch": 2.174916573971079, "grad_norm": 0.3199937045574188, "learning_rate": 2.1274778103176854e-06, "loss": 0.3791, "step": 2607 }, { "epoch": 2.175750834260289, "grad_norm": 0.3309960961341858, "learning_rate": 2.123503597042655e-06, "loss": 0.4145, "step": 2608 }, { "epoch": 2.1765850945494996, "grad_norm": 0.34267953038215637, "learning_rate": 2.1195320982280598e-06, "loss": 0.4372, "step": 2609 }, { "epoch": 2.1774193548387095, "grad_norm": 0.2989409565925598, "learning_rate": 2.1155633176216755e-06, "loss": 0.3939, "step": 2610 }, { "epoch": 2.17825361512792, "grad_norm": 0.3194637894630432, "learning_rate": 2.1115972589687184e-06, "loss": 0.3765, "step": 2611 }, { "epoch": 2.1790878754171303, "grad_norm": 0.3258320987224579, "learning_rate": 2.1076339260118343e-06, "loss": 0.4278, "step": 2612 }, { "epoch": 2.1799221357063403, "grad_norm": 0.29600444436073303, "learning_rate": 2.103673322491099e-06, "loss": 0.3841, "step": 2613 }, { "epoch": 2.1807563959955507, "grad_norm": 0.3132535517215729, "learning_rate": 2.09971545214401e-06, "loss": 0.3856, "step": 2614 }, { "epoch": 2.1815906562847607, "grad_norm": 0.3435145318508148, "learning_rate": 2.095760318705487e-06, "loss": 0.4751, "step": 2615 }, { "epoch": 2.182424916573971, "grad_norm": 0.31265854835510254, "learning_rate": 2.0918079259078667e-06, "loss": 0.373, "step": 2616 }, { "epoch": 2.1832591768631815, "grad_norm": 0.316650927066803, "learning_rate": 2.0878582774808997e-06, "loss": 0.4439, "step": 2617 }, { "epoch": 2.1840934371523915, "grad_norm": 0.3074955344200134, "learning_rate": 2.0839113771517465e-06, "loss": 0.3908, "step": 2618 }, { "epoch": 2.184927697441602, "grad_norm": 0.32587265968322754, "learning_rate": 2.0799672286449734e-06, "loss": 0.3946, "step": 2619 }, { "epoch": 2.185761957730812, "grad_norm": 0.3458298146724701, "learning_rate": 2.0760258356825556e-06, "loss": 0.3981, "step": 2620 }, { "epoch": 2.1865962180200222, "grad_norm": 0.32429951429367065, "learning_rate": 2.072087201983857e-06, "loss": 0.3959, "step": 2621 }, { "epoch": 2.1874304783092327, "grad_norm": 0.2964249849319458, "learning_rate": 2.0681513312656487e-06, "loss": 0.3982, "step": 2622 }, { "epoch": 2.1882647385984426, "grad_norm": 0.3039122223854065, "learning_rate": 2.064218227242088e-06, "loss": 0.3801, "step": 2623 }, { "epoch": 2.189098998887653, "grad_norm": 0.347340852022171, "learning_rate": 2.060287893624724e-06, "loss": 0.4251, "step": 2624 }, { "epoch": 2.189933259176863, "grad_norm": 0.3079792261123657, "learning_rate": 2.0563603341224897e-06, "loss": 0.3524, "step": 2625 }, { "epoch": 2.1907675194660734, "grad_norm": 0.3128872513771057, "learning_rate": 2.0524355524417017e-06, "loss": 0.3829, "step": 2626 }, { "epoch": 2.191601779755284, "grad_norm": 0.31552720069885254, "learning_rate": 2.0485135522860544e-06, "loss": 0.3376, "step": 2627 }, { "epoch": 2.1924360400444938, "grad_norm": 0.34576502442359924, "learning_rate": 2.0445943373566178e-06, "loss": 0.4066, "step": 2628 }, { "epoch": 2.193270300333704, "grad_norm": 0.32929980754852295, "learning_rate": 2.040677911351834e-06, "loss": 0.3668, "step": 2629 }, { "epoch": 2.194104560622914, "grad_norm": 0.3272038698196411, "learning_rate": 2.0367642779675116e-06, "loss": 0.3969, "step": 2630 }, { "epoch": 2.1949388209121246, "grad_norm": 0.3145748972892761, "learning_rate": 2.0328534408968263e-06, "loss": 0.4253, "step": 2631 }, { "epoch": 2.195773081201335, "grad_norm": 0.309112548828125, "learning_rate": 2.0289454038303115e-06, "loss": 0.4043, "step": 2632 }, { "epoch": 2.196607341490545, "grad_norm": 0.31137335300445557, "learning_rate": 2.0250401704558654e-06, "loss": 0.3948, "step": 2633 }, { "epoch": 2.1974416017797553, "grad_norm": 0.3313305377960205, "learning_rate": 2.0211377444587286e-06, "loss": 0.4033, "step": 2634 }, { "epoch": 2.1982758620689653, "grad_norm": 0.2846447825431824, "learning_rate": 2.017238129521506e-06, "loss": 0.3756, "step": 2635 }, { "epoch": 2.1991101223581757, "grad_norm": 0.3206008970737457, "learning_rate": 2.0133413293241393e-06, "loss": 0.3826, "step": 2636 }, { "epoch": 2.199944382647386, "grad_norm": 0.3218538165092468, "learning_rate": 2.00944734754392e-06, "loss": 0.4045, "step": 2637 }, { "epoch": 2.200778642936596, "grad_norm": 0.3313266336917877, "learning_rate": 2.005556187855479e-06, "loss": 0.4165, "step": 2638 }, { "epoch": 2.2016129032258065, "grad_norm": 0.2943177819252014, "learning_rate": 2.001667853930777e-06, "loss": 0.3688, "step": 2639 }, { "epoch": 2.202447163515017, "grad_norm": 0.2885214686393738, "learning_rate": 1.997782349439121e-06, "loss": 0.3806, "step": 2640 }, { "epoch": 2.203281423804227, "grad_norm": 0.32167738676071167, "learning_rate": 1.993899678047137e-06, "loss": 0.4176, "step": 2641 }, { "epoch": 2.2041156840934373, "grad_norm": 0.30459821224212646, "learning_rate": 1.9900198434187838e-06, "loss": 0.3952, "step": 2642 }, { "epoch": 2.2049499443826472, "grad_norm": 0.3536401689052582, "learning_rate": 1.9861428492153393e-06, "loss": 0.3952, "step": 2643 }, { "epoch": 2.2057842046718577, "grad_norm": 0.32320424914360046, "learning_rate": 1.9822686990954044e-06, "loss": 0.3831, "step": 2644 }, { "epoch": 2.206618464961068, "grad_norm": 0.291138619184494, "learning_rate": 1.978397396714891e-06, "loss": 0.386, "step": 2645 }, { "epoch": 2.207452725250278, "grad_norm": 0.33413878083229065, "learning_rate": 1.9745289457270335e-06, "loss": 0.4284, "step": 2646 }, { "epoch": 2.2082869855394884, "grad_norm": 0.30469003319740295, "learning_rate": 1.970663349782363e-06, "loss": 0.3987, "step": 2647 }, { "epoch": 2.2091212458286984, "grad_norm": 0.29255780577659607, "learning_rate": 1.966800612528723e-06, "loss": 0.4061, "step": 2648 }, { "epoch": 2.209955506117909, "grad_norm": 0.3220442831516266, "learning_rate": 1.962940737611264e-06, "loss": 0.3853, "step": 2649 }, { "epoch": 2.210789766407119, "grad_norm": 0.32286912202835083, "learning_rate": 1.959083728672423e-06, "loss": 0.3857, "step": 2650 }, { "epoch": 2.211624026696329, "grad_norm": 0.31698471307754517, "learning_rate": 1.955229589351947e-06, "loss": 0.3782, "step": 2651 }, { "epoch": 2.2124582869855396, "grad_norm": 0.2854025065898895, "learning_rate": 1.9513783232868603e-06, "loss": 0.3696, "step": 2652 }, { "epoch": 2.2132925472747496, "grad_norm": 0.2981434762477875, "learning_rate": 1.947529934111489e-06, "loss": 0.4092, "step": 2653 }, { "epoch": 2.21412680756396, "grad_norm": 0.29963335394859314, "learning_rate": 1.9436844254574366e-06, "loss": 0.3715, "step": 2654 }, { "epoch": 2.2149610678531704, "grad_norm": 0.3188994824886322, "learning_rate": 1.9398418009535906e-06, "loss": 0.3997, "step": 2655 }, { "epoch": 2.2157953281423803, "grad_norm": 0.30262845754623413, "learning_rate": 1.9360020642261155e-06, "loss": 0.3704, "step": 2656 }, { "epoch": 2.2166295884315907, "grad_norm": 0.3107425570487976, "learning_rate": 1.9321652188984535e-06, "loss": 0.4103, "step": 2657 }, { "epoch": 2.2174638487208007, "grad_norm": 0.295236200094223, "learning_rate": 1.928331268591315e-06, "loss": 0.3763, "step": 2658 }, { "epoch": 2.218298109010011, "grad_norm": 0.3207455575466156, "learning_rate": 1.9245002169226814e-06, "loss": 0.4587, "step": 2659 }, { "epoch": 2.2191323692992215, "grad_norm": 0.29384732246398926, "learning_rate": 1.9206720675077955e-06, "loss": 0.3677, "step": 2660 }, { "epoch": 2.2199666295884315, "grad_norm": 0.32689037919044495, "learning_rate": 1.9168468239591625e-06, "loss": 0.3709, "step": 2661 }, { "epoch": 2.220800889877642, "grad_norm": 0.327181875705719, "learning_rate": 1.913024489886552e-06, "loss": 0.4246, "step": 2662 }, { "epoch": 2.221635150166852, "grad_norm": 0.2876399755477905, "learning_rate": 1.9092050688969736e-06, "loss": 0.3719, "step": 2663 }, { "epoch": 2.2224694104560623, "grad_norm": 0.3239113390445709, "learning_rate": 1.9053885645947046e-06, "loss": 0.4307, "step": 2664 }, { "epoch": 2.2233036707452727, "grad_norm": 0.28594350814819336, "learning_rate": 1.901574980581254e-06, "loss": 0.3585, "step": 2665 }, { "epoch": 2.2241379310344827, "grad_norm": 0.2932165861129761, "learning_rate": 1.897764320455388e-06, "loss": 0.3625, "step": 2666 }, { "epoch": 2.224972191323693, "grad_norm": 0.32598814368247986, "learning_rate": 1.8939565878131071e-06, "loss": 0.3912, "step": 2667 }, { "epoch": 2.225806451612903, "grad_norm": 0.33622971177101135, "learning_rate": 1.8901517862476503e-06, "loss": 0.4035, "step": 2668 }, { "epoch": 2.2266407119021134, "grad_norm": 0.29595300555229187, "learning_rate": 1.886349919349491e-06, "loss": 0.3591, "step": 2669 }, { "epoch": 2.227474972191324, "grad_norm": 0.31503355503082275, "learning_rate": 1.8825509907063328e-06, "loss": 0.3721, "step": 2670 }, { "epoch": 2.228309232480534, "grad_norm": 0.3501476049423218, "learning_rate": 1.8787550039031066e-06, "loss": 0.4476, "step": 2671 }, { "epoch": 2.229143492769744, "grad_norm": 0.28197339177131653, "learning_rate": 1.8749619625219679e-06, "loss": 0.3567, "step": 2672 }, { "epoch": 2.2299777530589546, "grad_norm": 0.2993980050086975, "learning_rate": 1.871171870142292e-06, "loss": 0.4329, "step": 2673 }, { "epoch": 2.2308120133481646, "grad_norm": 0.28385934233665466, "learning_rate": 1.86738473034067e-06, "loss": 0.3501, "step": 2674 }, { "epoch": 2.231646273637375, "grad_norm": 0.3186591565608978, "learning_rate": 1.863600546690913e-06, "loss": 0.4047, "step": 2675 }, { "epoch": 2.232480533926585, "grad_norm": 0.3372349441051483, "learning_rate": 1.8598193227640305e-06, "loss": 0.4342, "step": 2676 }, { "epoch": 2.2333147942157954, "grad_norm": 0.32443442940711975, "learning_rate": 1.8560410621282543e-06, "loss": 0.3885, "step": 2677 }, { "epoch": 2.2341490545050053, "grad_norm": 0.2987229526042938, "learning_rate": 1.852265768349006e-06, "loss": 0.3935, "step": 2678 }, { "epoch": 2.2349833147942157, "grad_norm": 0.30233505368232727, "learning_rate": 1.8484934449889126e-06, "loss": 0.3894, "step": 2679 }, { "epoch": 2.235817575083426, "grad_norm": 0.3007335960865021, "learning_rate": 1.844724095607805e-06, "loss": 0.4191, "step": 2680 }, { "epoch": 2.236651835372636, "grad_norm": 0.3134855329990387, "learning_rate": 1.8409577237626935e-06, "loss": 0.3871, "step": 2681 }, { "epoch": 2.2374860956618465, "grad_norm": 0.3266913890838623, "learning_rate": 1.837194333007792e-06, "loss": 0.3949, "step": 2682 }, { "epoch": 2.238320355951057, "grad_norm": 0.3316470980644226, "learning_rate": 1.8334339268944934e-06, "loss": 0.3997, "step": 2683 }, { "epoch": 2.239154616240267, "grad_norm": 0.32572221755981445, "learning_rate": 1.829676508971377e-06, "loss": 0.3919, "step": 2684 }, { "epoch": 2.2399888765294773, "grad_norm": 0.3180612325668335, "learning_rate": 1.8259220827842011e-06, "loss": 0.4169, "step": 2685 }, { "epoch": 2.2408231368186873, "grad_norm": 0.3142438232898712, "learning_rate": 1.822170651875902e-06, "loss": 0.4013, "step": 2686 }, { "epoch": 2.2416573971078977, "grad_norm": 0.31687307357788086, "learning_rate": 1.8184222197865853e-06, "loss": 0.3963, "step": 2687 }, { "epoch": 2.242491657397108, "grad_norm": 0.31795868277549744, "learning_rate": 1.8146767900535368e-06, "loss": 0.4039, "step": 2688 }, { "epoch": 2.243325917686318, "grad_norm": 0.3098975419998169, "learning_rate": 1.8109343662111967e-06, "loss": 0.396, "step": 2689 }, { "epoch": 2.2441601779755285, "grad_norm": 0.3101857602596283, "learning_rate": 1.8071949517911758e-06, "loss": 0.3987, "step": 2690 }, { "epoch": 2.2449944382647384, "grad_norm": 0.3082296848297119, "learning_rate": 1.8034585503222441e-06, "loss": 0.4061, "step": 2691 }, { "epoch": 2.245828698553949, "grad_norm": 0.3301537334918976, "learning_rate": 1.7997251653303249e-06, "loss": 0.439, "step": 2692 }, { "epoch": 2.2466629588431593, "grad_norm": 0.30699047446250916, "learning_rate": 1.7959948003385042e-06, "loss": 0.4129, "step": 2693 }, { "epoch": 2.247497219132369, "grad_norm": 0.30936264991760254, "learning_rate": 1.7922674588670042e-06, "loss": 0.3989, "step": 2694 }, { "epoch": 2.2483314794215796, "grad_norm": 0.3112725615501404, "learning_rate": 1.7885431444332069e-06, "loss": 0.3983, "step": 2695 }, { "epoch": 2.2491657397107896, "grad_norm": 0.3039705753326416, "learning_rate": 1.7848218605516304e-06, "loss": 0.3791, "step": 2696 }, { "epoch": 2.25, "grad_norm": 0.3280637264251709, "learning_rate": 1.7811036107339351e-06, "loss": 0.3716, "step": 2697 }, { "epoch": 2.2508342602892104, "grad_norm": 0.3266432285308838, "learning_rate": 1.7773883984889178e-06, "loss": 0.4229, "step": 2698 }, { "epoch": 2.2516685205784204, "grad_norm": 0.31110578775405884, "learning_rate": 1.773676227322509e-06, "loss": 0.3734, "step": 2699 }, { "epoch": 2.252502780867631, "grad_norm": 0.3021202087402344, "learning_rate": 1.7699671007377694e-06, "loss": 0.3853, "step": 2700 }, { "epoch": 2.2533370411568407, "grad_norm": 0.2840973138809204, "learning_rate": 1.7662610222348869e-06, "loss": 0.4127, "step": 2701 }, { "epoch": 2.254171301446051, "grad_norm": 0.3143460154533386, "learning_rate": 1.7625579953111726e-06, "loss": 0.4169, "step": 2702 }, { "epoch": 2.2550055617352616, "grad_norm": 0.28967639803886414, "learning_rate": 1.7588580234610592e-06, "loss": 0.3977, "step": 2703 }, { "epoch": 2.2558398220244715, "grad_norm": 0.34158554673194885, "learning_rate": 1.755161110176094e-06, "loss": 0.4249, "step": 2704 }, { "epoch": 2.256674082313682, "grad_norm": 0.33009395003318787, "learning_rate": 1.7514672589449378e-06, "loss": 0.4345, "step": 2705 }, { "epoch": 2.2575083426028923, "grad_norm": 0.3080599009990692, "learning_rate": 1.7477764732533697e-06, "loss": 0.3745, "step": 2706 }, { "epoch": 2.2583426028921023, "grad_norm": 0.3056640625, "learning_rate": 1.7440887565842618e-06, "loss": 0.4031, "step": 2707 }, { "epoch": 2.2591768631813127, "grad_norm": 0.30819830298423767, "learning_rate": 1.7404041124176029e-06, "loss": 0.3779, "step": 2708 }, { "epoch": 2.2600111234705227, "grad_norm": 0.3141663372516632, "learning_rate": 1.7367225442304786e-06, "loss": 0.4008, "step": 2709 }, { "epoch": 2.260845383759733, "grad_norm": 0.29404333233833313, "learning_rate": 1.7330440554970646e-06, "loss": 0.3734, "step": 2710 }, { "epoch": 2.261679644048943, "grad_norm": 0.3162182867527008, "learning_rate": 1.729368649688642e-06, "loss": 0.4296, "step": 2711 }, { "epoch": 2.2625139043381535, "grad_norm": 0.3012341558933258, "learning_rate": 1.7256963302735752e-06, "loss": 0.3776, "step": 2712 }, { "epoch": 2.263348164627364, "grad_norm": 0.30861079692840576, "learning_rate": 1.7220271007173184e-06, "loss": 0.3922, "step": 2713 }, { "epoch": 2.264182424916574, "grad_norm": 0.335153192281723, "learning_rate": 1.7183609644824096e-06, "loss": 0.4311, "step": 2714 }, { "epoch": 2.2650166852057843, "grad_norm": 0.3252272307872772, "learning_rate": 1.7146979250284678e-06, "loss": 0.4093, "step": 2715 }, { "epoch": 2.2658509454949947, "grad_norm": 0.2915157973766327, "learning_rate": 1.71103798581219e-06, "loss": 0.3539, "step": 2716 }, { "epoch": 2.2666852057842046, "grad_norm": 0.3416847288608551, "learning_rate": 1.7073811502873476e-06, "loss": 0.4064, "step": 2717 }, { "epoch": 2.267519466073415, "grad_norm": 0.310714453458786, "learning_rate": 1.70372742190478e-06, "loss": 0.4124, "step": 2718 }, { "epoch": 2.268353726362625, "grad_norm": 0.31369876861572266, "learning_rate": 1.7000768041124038e-06, "loss": 0.4303, "step": 2719 }, { "epoch": 2.2691879866518354, "grad_norm": 0.2755182981491089, "learning_rate": 1.696429300355188e-06, "loss": 0.3704, "step": 2720 }, { "epoch": 2.2700222469410454, "grad_norm": 0.3183666467666626, "learning_rate": 1.6927849140751685e-06, "loss": 0.4039, "step": 2721 }, { "epoch": 2.270856507230256, "grad_norm": 0.3112582266330719, "learning_rate": 1.6891436487114448e-06, "loss": 0.3697, "step": 2722 }, { "epoch": 2.271690767519466, "grad_norm": 0.3100363314151764, "learning_rate": 1.6855055077001597e-06, "loss": 0.4355, "step": 2723 }, { "epoch": 2.272525027808676, "grad_norm": 0.33583199977874756, "learning_rate": 1.681870494474519e-06, "loss": 0.373, "step": 2724 }, { "epoch": 2.2733592880978866, "grad_norm": 0.31295427680015564, "learning_rate": 1.67823861246477e-06, "loss": 0.4106, "step": 2725 }, { "epoch": 2.274193548387097, "grad_norm": 0.30116719007492065, "learning_rate": 1.6746098650982072e-06, "loss": 0.3753, "step": 2726 }, { "epoch": 2.275027808676307, "grad_norm": 0.29683980345726013, "learning_rate": 1.6709842557991668e-06, "loss": 0.4027, "step": 2727 }, { "epoch": 2.2758620689655173, "grad_norm": 0.33427685499191284, "learning_rate": 1.6673617879890241e-06, "loss": 0.4218, "step": 2728 }, { "epoch": 2.2766963292547273, "grad_norm": 0.3259771764278412, "learning_rate": 1.6637424650861895e-06, "loss": 0.4126, "step": 2729 }, { "epoch": 2.2775305895439377, "grad_norm": 0.296382874250412, "learning_rate": 1.660126290506106e-06, "loss": 0.3545, "step": 2730 }, { "epoch": 2.278364849833148, "grad_norm": 0.3419109880924225, "learning_rate": 1.656513267661245e-06, "loss": 0.4331, "step": 2731 }, { "epoch": 2.279199110122358, "grad_norm": 0.33013615012168884, "learning_rate": 1.6529033999611038e-06, "loss": 0.4009, "step": 2732 }, { "epoch": 2.2800333704115685, "grad_norm": 0.2969238758087158, "learning_rate": 1.6492966908122033e-06, "loss": 0.3658, "step": 2733 }, { "epoch": 2.2808676307007785, "grad_norm": 0.3207879066467285, "learning_rate": 1.6456931436180812e-06, "loss": 0.4183, "step": 2734 }, { "epoch": 2.281701890989989, "grad_norm": 0.31107908487319946, "learning_rate": 1.642092761779298e-06, "loss": 0.4386, "step": 2735 }, { "epoch": 2.2825361512791993, "grad_norm": 0.3183040916919708, "learning_rate": 1.6384955486934157e-06, "loss": 0.3781, "step": 2736 }, { "epoch": 2.2833704115684093, "grad_norm": 0.3144783079624176, "learning_rate": 1.634901507755018e-06, "loss": 0.3982, "step": 2737 }, { "epoch": 2.2842046718576197, "grad_norm": 0.30719688534736633, "learning_rate": 1.631310642355688e-06, "loss": 0.3841, "step": 2738 }, { "epoch": 2.2850389321468296, "grad_norm": 0.28212541341781616, "learning_rate": 1.6277229558840136e-06, "loss": 0.4037, "step": 2739 }, { "epoch": 2.28587319243604, "grad_norm": 0.2954874336719513, "learning_rate": 1.6241384517255854e-06, "loss": 0.403, "step": 2740 }, { "epoch": 2.2867074527252504, "grad_norm": 0.2898341119289398, "learning_rate": 1.6205571332629839e-06, "loss": 0.389, "step": 2741 }, { "epoch": 2.2875417130144604, "grad_norm": 0.30213552713394165, "learning_rate": 1.6169790038757932e-06, "loss": 0.4063, "step": 2742 }, { "epoch": 2.288375973303671, "grad_norm": 0.3006368577480316, "learning_rate": 1.613404066940581e-06, "loss": 0.3706, "step": 2743 }, { "epoch": 2.289210233592881, "grad_norm": 0.3459871709346771, "learning_rate": 1.609832325830904e-06, "loss": 0.4002, "step": 2744 }, { "epoch": 2.290044493882091, "grad_norm": 0.35153865814208984, "learning_rate": 1.6062637839173045e-06, "loss": 0.4119, "step": 2745 }, { "epoch": 2.2908787541713016, "grad_norm": 0.30367162823677063, "learning_rate": 1.6026984445673043e-06, "loss": 0.3948, "step": 2746 }, { "epoch": 2.2917130144605116, "grad_norm": 0.29656633734703064, "learning_rate": 1.5991363111454023e-06, "loss": 0.4183, "step": 2747 }, { "epoch": 2.292547274749722, "grad_norm": 0.2764495015144348, "learning_rate": 1.595577387013078e-06, "loss": 0.358, "step": 2748 }, { "epoch": 2.2933815350389324, "grad_norm": 0.30261126160621643, "learning_rate": 1.5920216755287704e-06, "loss": 0.3775, "step": 2749 }, { "epoch": 2.2942157953281423, "grad_norm": 0.2842842638492584, "learning_rate": 1.5884691800479003e-06, "loss": 0.3781, "step": 2750 }, { "epoch": 2.2950500556173528, "grad_norm": 0.32256266474723816, "learning_rate": 1.5849199039228468e-06, "loss": 0.4043, "step": 2751 }, { "epoch": 2.2958843159065627, "grad_norm": 0.3366665840148926, "learning_rate": 1.5813738505029468e-06, "loss": 0.4377, "step": 2752 }, { "epoch": 2.296718576195773, "grad_norm": 0.3073413074016571, "learning_rate": 1.577831023134508e-06, "loss": 0.3779, "step": 2753 }, { "epoch": 2.297552836484983, "grad_norm": 0.31633391976356506, "learning_rate": 1.5742914251607794e-06, "loss": 0.4093, "step": 2754 }, { "epoch": 2.2983870967741935, "grad_norm": 0.3147065043449402, "learning_rate": 1.570755059921975e-06, "loss": 0.3882, "step": 2755 }, { "epoch": 2.299221357063404, "grad_norm": 0.2961532473564148, "learning_rate": 1.5672219307552511e-06, "loss": 0.3911, "step": 2756 }, { "epoch": 2.300055617352614, "grad_norm": 0.28965476155281067, "learning_rate": 1.5636920409947132e-06, "loss": 0.3969, "step": 2757 }, { "epoch": 2.3008898776418243, "grad_norm": 0.31148114800453186, "learning_rate": 1.5601653939714073e-06, "loss": 0.4369, "step": 2758 }, { "epoch": 2.3017241379310347, "grad_norm": 0.30000337958335876, "learning_rate": 1.5566419930133213e-06, "loss": 0.4112, "step": 2759 }, { "epoch": 2.3025583982202447, "grad_norm": 0.3028186559677124, "learning_rate": 1.5531218414453802e-06, "loss": 0.4002, "step": 2760 }, { "epoch": 2.303392658509455, "grad_norm": 0.311191201210022, "learning_rate": 1.549604942589441e-06, "loss": 0.3781, "step": 2761 }, { "epoch": 2.304226918798665, "grad_norm": 0.2823339104652405, "learning_rate": 1.5460912997642924e-06, "loss": 0.3936, "step": 2762 }, { "epoch": 2.3050611790878754, "grad_norm": 0.2850428521633148, "learning_rate": 1.5425809162856482e-06, "loss": 0.3629, "step": 2763 }, { "epoch": 2.3058954393770854, "grad_norm": 0.30343255400657654, "learning_rate": 1.5390737954661533e-06, "loss": 0.4243, "step": 2764 }, { "epoch": 2.306729699666296, "grad_norm": 0.30471670627593994, "learning_rate": 1.5355699406153628e-06, "loss": 0.4126, "step": 2765 }, { "epoch": 2.3075639599555062, "grad_norm": 0.3151152431964874, "learning_rate": 1.5320693550397619e-06, "loss": 0.4163, "step": 2766 }, { "epoch": 2.308398220244716, "grad_norm": 0.28400281071662903, "learning_rate": 1.5285720420427386e-06, "loss": 0.3705, "step": 2767 }, { "epoch": 2.3092324805339266, "grad_norm": 0.30862563848495483, "learning_rate": 1.5250780049246028e-06, "loss": 0.406, "step": 2768 }, { "epoch": 2.310066740823137, "grad_norm": 0.27777528762817383, "learning_rate": 1.5215872469825682e-06, "loss": 0.3606, "step": 2769 }, { "epoch": 2.310901001112347, "grad_norm": 0.3005344867706299, "learning_rate": 1.5180997715107543e-06, "loss": 0.4087, "step": 2770 }, { "epoch": 2.3117352614015574, "grad_norm": 0.2827872335910797, "learning_rate": 1.5146155818001835e-06, "loss": 0.3645, "step": 2771 }, { "epoch": 2.3125695216907673, "grad_norm": 0.3103359043598175, "learning_rate": 1.5111346811387772e-06, "loss": 0.3988, "step": 2772 }, { "epoch": 2.3134037819799778, "grad_norm": 0.3024842143058777, "learning_rate": 1.5076570728113543e-06, "loss": 0.4044, "step": 2773 }, { "epoch": 2.314238042269188, "grad_norm": 0.30052241683006287, "learning_rate": 1.5041827600996255e-06, "loss": 0.3955, "step": 2774 }, { "epoch": 2.315072302558398, "grad_norm": 0.2817320227622986, "learning_rate": 1.500711746282192e-06, "loss": 0.3841, "step": 2775 }, { "epoch": 2.3159065628476085, "grad_norm": 0.2977808713912964, "learning_rate": 1.4972440346345396e-06, "loss": 0.3984, "step": 2776 }, { "epoch": 2.3167408231368185, "grad_norm": 0.2998465597629547, "learning_rate": 1.4937796284290463e-06, "loss": 0.3875, "step": 2777 }, { "epoch": 2.317575083426029, "grad_norm": 0.29809197783470154, "learning_rate": 1.4903185309349571e-06, "loss": 0.3752, "step": 2778 }, { "epoch": 2.3184093437152393, "grad_norm": 0.3509664535522461, "learning_rate": 1.4868607454184092e-06, "loss": 0.4045, "step": 2779 }, { "epoch": 2.3192436040044493, "grad_norm": 0.30349424481391907, "learning_rate": 1.4834062751424018e-06, "loss": 0.4206, "step": 2780 }, { "epoch": 2.3200778642936597, "grad_norm": 0.3064546287059784, "learning_rate": 1.4799551233668152e-06, "loss": 0.3486, "step": 2781 }, { "epoch": 2.3209121245828697, "grad_norm": 0.3343088626861572, "learning_rate": 1.4765072933483949e-06, "loss": 0.3803, "step": 2782 }, { "epoch": 2.32174638487208, "grad_norm": 0.3040873110294342, "learning_rate": 1.4730627883407455e-06, "loss": 0.4329, "step": 2783 }, { "epoch": 2.3225806451612905, "grad_norm": 0.2858830690383911, "learning_rate": 1.4696216115943457e-06, "loss": 0.358, "step": 2784 }, { "epoch": 2.3234149054505004, "grad_norm": 0.29364022612571716, "learning_rate": 1.4661837663565249e-06, "loss": 0.3721, "step": 2785 }, { "epoch": 2.324249165739711, "grad_norm": 0.3050469160079956, "learning_rate": 1.4627492558714707e-06, "loss": 0.4014, "step": 2786 }, { "epoch": 2.325083426028921, "grad_norm": 0.31421467661857605, "learning_rate": 1.459318083380224e-06, "loss": 0.4237, "step": 2787 }, { "epoch": 2.3259176863181312, "grad_norm": 0.3104746341705322, "learning_rate": 1.455890252120677e-06, "loss": 0.4149, "step": 2788 }, { "epoch": 2.3267519466073416, "grad_norm": 0.29419419169425964, "learning_rate": 1.4524657653275653e-06, "loss": 0.3859, "step": 2789 }, { "epoch": 2.3275862068965516, "grad_norm": 0.285783052444458, "learning_rate": 1.4490446262324753e-06, "loss": 0.4296, "step": 2790 }, { "epoch": 2.328420467185762, "grad_norm": 0.27523231506347656, "learning_rate": 1.4456268380638262e-06, "loss": 0.3548, "step": 2791 }, { "epoch": 2.3292547274749724, "grad_norm": 0.3415546119213104, "learning_rate": 1.4422124040468772e-06, "loss": 0.47, "step": 2792 }, { "epoch": 2.3300889877641824, "grad_norm": 0.3007616698741913, "learning_rate": 1.43880132740373e-06, "loss": 0.3632, "step": 2793 }, { "epoch": 2.330923248053393, "grad_norm": 0.308025985956192, "learning_rate": 1.4353936113533045e-06, "loss": 0.4077, "step": 2794 }, { "epoch": 2.3317575083426028, "grad_norm": 0.29366764426231384, "learning_rate": 1.4319892591113631e-06, "loss": 0.3957, "step": 2795 }, { "epoch": 2.332591768631813, "grad_norm": 0.3047045171260834, "learning_rate": 1.4285882738904822e-06, "loss": 0.3637, "step": 2796 }, { "epoch": 2.333426028921023, "grad_norm": 0.3148644268512726, "learning_rate": 1.4251906589000692e-06, "loss": 0.4316, "step": 2797 }, { "epoch": 2.3342602892102335, "grad_norm": 0.3175455331802368, "learning_rate": 1.4217964173463472e-06, "loss": 0.3919, "step": 2798 }, { "epoch": 2.335094549499444, "grad_norm": 0.30425649881362915, "learning_rate": 1.4184055524323564e-06, "loss": 0.3888, "step": 2799 }, { "epoch": 2.335928809788654, "grad_norm": 0.33135339617729187, "learning_rate": 1.415018067357951e-06, "loss": 0.388, "step": 2800 }, { "epoch": 2.3367630700778643, "grad_norm": 0.30778247117996216, "learning_rate": 1.4116339653197958e-06, "loss": 0.3935, "step": 2801 }, { "epoch": 2.3375973303670747, "grad_norm": 0.29822850227355957, "learning_rate": 1.4082532495113627e-06, "loss": 0.3778, "step": 2802 }, { "epoch": 2.3384315906562847, "grad_norm": 0.33279499411582947, "learning_rate": 1.404875923122928e-06, "loss": 0.4151, "step": 2803 }, { "epoch": 2.339265850945495, "grad_norm": 0.2842143774032593, "learning_rate": 1.401501989341571e-06, "loss": 0.3939, "step": 2804 }, { "epoch": 2.340100111234705, "grad_norm": 0.3186204135417938, "learning_rate": 1.3981314513511657e-06, "loss": 0.4284, "step": 2805 }, { "epoch": 2.3409343715239155, "grad_norm": 0.2979459762573242, "learning_rate": 1.394764312332389e-06, "loss": 0.3832, "step": 2806 }, { "epoch": 2.341768631813126, "grad_norm": 0.3091980218887329, "learning_rate": 1.3914005754626996e-06, "loss": 0.4034, "step": 2807 }, { "epoch": 2.342602892102336, "grad_norm": 0.32874608039855957, "learning_rate": 1.3880402439163576e-06, "loss": 0.3955, "step": 2808 }, { "epoch": 2.3434371523915463, "grad_norm": 0.2876078188419342, "learning_rate": 1.3846833208643979e-06, "loss": 0.3513, "step": 2809 }, { "epoch": 2.3442714126807562, "grad_norm": 0.32261356711387634, "learning_rate": 1.3813298094746491e-06, "loss": 0.3873, "step": 2810 }, { "epoch": 2.3451056729699666, "grad_norm": 0.3095630705356598, "learning_rate": 1.3779797129117146e-06, "loss": 0.385, "step": 2811 }, { "epoch": 2.345939933259177, "grad_norm": 0.30513641238212585, "learning_rate": 1.3746330343369757e-06, "loss": 0.3904, "step": 2812 }, { "epoch": 2.346774193548387, "grad_norm": 0.302447646856308, "learning_rate": 1.3712897769085903e-06, "loss": 0.3728, "step": 2813 }, { "epoch": 2.3476084538375974, "grad_norm": 0.3309255838394165, "learning_rate": 1.3679499437814858e-06, "loss": 0.4123, "step": 2814 }, { "epoch": 2.3484427141268074, "grad_norm": 0.29397791624069214, "learning_rate": 1.36461353810736e-06, "loss": 0.4153, "step": 2815 }, { "epoch": 2.349276974416018, "grad_norm": 0.293869286775589, "learning_rate": 1.3612805630346742e-06, "loss": 0.3912, "step": 2816 }, { "epoch": 2.350111234705228, "grad_norm": 0.2878265976905823, "learning_rate": 1.357951021708655e-06, "loss": 0.349, "step": 2817 }, { "epoch": 2.350945494994438, "grad_norm": 0.29786062240600586, "learning_rate": 1.354624917271285e-06, "loss": 0.3949, "step": 2818 }, { "epoch": 2.3517797552836486, "grad_norm": 0.28689488768577576, "learning_rate": 1.3513022528613107e-06, "loss": 0.3782, "step": 2819 }, { "epoch": 2.3526140155728585, "grad_norm": 0.3022514879703522, "learning_rate": 1.347983031614221e-06, "loss": 0.4108, "step": 2820 }, { "epoch": 2.353448275862069, "grad_norm": 0.28900477290153503, "learning_rate": 1.3446672566622682e-06, "loss": 0.3942, "step": 2821 }, { "epoch": 2.3542825361512794, "grad_norm": 0.31759098172187805, "learning_rate": 1.3413549311344416e-06, "loss": 0.4163, "step": 2822 }, { "epoch": 2.3551167964404893, "grad_norm": 0.30652496218681335, "learning_rate": 1.3380460581564802e-06, "loss": 0.4057, "step": 2823 }, { "epoch": 2.3559510567296997, "grad_norm": 0.27905261516571045, "learning_rate": 1.3347406408508695e-06, "loss": 0.3451, "step": 2824 }, { "epoch": 2.35678531701891, "grad_norm": 0.27949586510658264, "learning_rate": 1.3314386823368237e-06, "loss": 0.3732, "step": 2825 }, { "epoch": 2.35761957730812, "grad_norm": 0.31075572967529297, "learning_rate": 1.3281401857303017e-06, "loss": 0.4072, "step": 2826 }, { "epoch": 2.3584538375973305, "grad_norm": 0.33084791898727417, "learning_rate": 1.3248451541439921e-06, "loss": 0.4083, "step": 2827 }, { "epoch": 2.3592880978865405, "grad_norm": 0.29269295930862427, "learning_rate": 1.3215535906873138e-06, "loss": 0.3854, "step": 2828 }, { "epoch": 2.360122358175751, "grad_norm": 0.3254573941230774, "learning_rate": 1.318265498466413e-06, "loss": 0.4054, "step": 2829 }, { "epoch": 2.360956618464961, "grad_norm": 0.29929226636886597, "learning_rate": 1.31498088058416e-06, "loss": 0.4101, "step": 2830 }, { "epoch": 2.3617908787541713, "grad_norm": 0.3091394305229187, "learning_rate": 1.311699740140146e-06, "loss": 0.3884, "step": 2831 }, { "epoch": 2.3626251390433817, "grad_norm": 0.30309727787971497, "learning_rate": 1.3084220802306853e-06, "loss": 0.4396, "step": 2832 }, { "epoch": 2.3634593993325916, "grad_norm": 0.2975016236305237, "learning_rate": 1.3051479039488002e-06, "loss": 0.3687, "step": 2833 }, { "epoch": 2.364293659621802, "grad_norm": 0.3315277695655823, "learning_rate": 1.3018772143842301e-06, "loss": 0.4115, "step": 2834 }, { "epoch": 2.3651279199110125, "grad_norm": 0.3222929537296295, "learning_rate": 1.298610014623423e-06, "loss": 0.3942, "step": 2835 }, { "epoch": 2.3659621802002224, "grad_norm": 0.28550970554351807, "learning_rate": 1.2953463077495338e-06, "loss": 0.399, "step": 2836 }, { "epoch": 2.366796440489433, "grad_norm": 0.3083926737308502, "learning_rate": 1.2920860968424247e-06, "loss": 0.4156, "step": 2837 }, { "epoch": 2.367630700778643, "grad_norm": 0.30489736795425415, "learning_rate": 1.2888293849786503e-06, "loss": 0.385, "step": 2838 }, { "epoch": 2.368464961067853, "grad_norm": 0.310162216424942, "learning_rate": 1.2855761752314726e-06, "loss": 0.3956, "step": 2839 }, { "epoch": 2.369299221357063, "grad_norm": 0.3120497465133667, "learning_rate": 1.2823264706708432e-06, "loss": 0.3742, "step": 2840 }, { "epoch": 2.3701334816462736, "grad_norm": 0.2774602472782135, "learning_rate": 1.2790802743634085e-06, "loss": 0.3638, "step": 2841 }, { "epoch": 2.370967741935484, "grad_norm": 0.3097188174724579, "learning_rate": 1.2758375893725016e-06, "loss": 0.4349, "step": 2842 }, { "epoch": 2.371802002224694, "grad_norm": 0.32209545373916626, "learning_rate": 1.2725984187581452e-06, "loss": 0.4194, "step": 2843 }, { "epoch": 2.3726362625139044, "grad_norm": 0.2814674973487854, "learning_rate": 1.2693627655770435e-06, "loss": 0.3773, "step": 2844 }, { "epoch": 2.3734705228031148, "grad_norm": 0.32393643260002136, "learning_rate": 1.2661306328825818e-06, "loss": 0.3975, "step": 2845 }, { "epoch": 2.3743047830923247, "grad_norm": 0.31731489300727844, "learning_rate": 1.2629020237248241e-06, "loss": 0.3796, "step": 2846 }, { "epoch": 2.375139043381535, "grad_norm": 0.32342132925987244, "learning_rate": 1.2596769411505077e-06, "loss": 0.4522, "step": 2847 }, { "epoch": 2.375973303670745, "grad_norm": 0.2736559808254242, "learning_rate": 1.2564553882030445e-06, "loss": 0.3618, "step": 2848 }, { "epoch": 2.3768075639599555, "grad_norm": 0.31809109449386597, "learning_rate": 1.2532373679225113e-06, "loss": 0.4175, "step": 2849 }, { "epoch": 2.377641824249166, "grad_norm": 0.2954404950141907, "learning_rate": 1.2500228833456595e-06, "loss": 0.3858, "step": 2850 }, { "epoch": 2.378476084538376, "grad_norm": 0.30347558856010437, "learning_rate": 1.2468119375058918e-06, "loss": 0.4098, "step": 2851 }, { "epoch": 2.3793103448275863, "grad_norm": 0.3123914897441864, "learning_rate": 1.2436045334332824e-06, "loss": 0.4125, "step": 2852 }, { "epoch": 2.3801446051167963, "grad_norm": 0.30376318097114563, "learning_rate": 1.24040067415456e-06, "loss": 0.391, "step": 2853 }, { "epoch": 2.3809788654060067, "grad_norm": 0.2839154899120331, "learning_rate": 1.2372003626931023e-06, "loss": 0.3693, "step": 2854 }, { "epoch": 2.381813125695217, "grad_norm": 0.3020245432853699, "learning_rate": 1.2340036020689477e-06, "loss": 0.4258, "step": 2855 }, { "epoch": 2.382647385984427, "grad_norm": 0.3141053020954132, "learning_rate": 1.23081039529878e-06, "loss": 0.4229, "step": 2856 }, { "epoch": 2.3834816462736375, "grad_norm": 0.29437947273254395, "learning_rate": 1.2276207453959283e-06, "loss": 0.3754, "step": 2857 }, { "epoch": 2.3843159065628474, "grad_norm": 0.27546951174736023, "learning_rate": 1.2244346553703668e-06, "loss": 0.4038, "step": 2858 }, { "epoch": 2.385150166852058, "grad_norm": 0.29869863390922546, "learning_rate": 1.2212521282287093e-06, "loss": 0.3695, "step": 2859 }, { "epoch": 2.3859844271412682, "grad_norm": 0.3130325675010681, "learning_rate": 1.218073166974208e-06, "loss": 0.4243, "step": 2860 }, { "epoch": 2.386818687430478, "grad_norm": 0.2968442142009735, "learning_rate": 1.2148977746067509e-06, "loss": 0.3477, "step": 2861 }, { "epoch": 2.3876529477196886, "grad_norm": 0.3211473822593689, "learning_rate": 1.2117259541228548e-06, "loss": 0.422, "step": 2862 }, { "epoch": 2.3884872080088986, "grad_norm": 0.3079310953617096, "learning_rate": 1.2085577085156735e-06, "loss": 0.4278, "step": 2863 }, { "epoch": 2.389321468298109, "grad_norm": 0.28990378975868225, "learning_rate": 1.205393040774977e-06, "loss": 0.3737, "step": 2864 }, { "epoch": 2.3901557285873194, "grad_norm": 0.3025454878807068, "learning_rate": 1.2022319538871646e-06, "loss": 0.4343, "step": 2865 }, { "epoch": 2.3909899888765294, "grad_norm": 0.30398377776145935, "learning_rate": 1.1990744508352604e-06, "loss": 0.4313, "step": 2866 }, { "epoch": 2.3918242491657398, "grad_norm": 0.3032330274581909, "learning_rate": 1.1959205345988955e-06, "loss": 0.4514, "step": 2867 }, { "epoch": 2.39265850945495, "grad_norm": 0.2900089919567108, "learning_rate": 1.1927702081543279e-06, "loss": 0.3405, "step": 2868 }, { "epoch": 2.39349276974416, "grad_norm": 0.31574609875679016, "learning_rate": 1.189623474474421e-06, "loss": 0.3827, "step": 2869 }, { "epoch": 2.3943270300333706, "grad_norm": 0.29861190915107727, "learning_rate": 1.1864803365286504e-06, "loss": 0.408, "step": 2870 }, { "epoch": 2.3951612903225805, "grad_norm": 0.3096163868904114, "learning_rate": 1.1833407972830973e-06, "loss": 0.4309, "step": 2871 }, { "epoch": 2.395995550611791, "grad_norm": 0.2866154909133911, "learning_rate": 1.1802048597004479e-06, "loss": 0.3695, "step": 2872 }, { "epoch": 2.396829810901001, "grad_norm": 0.2975037097930908, "learning_rate": 1.1770725267399892e-06, "loss": 0.3899, "step": 2873 }, { "epoch": 2.3976640711902113, "grad_norm": 0.29665178060531616, "learning_rate": 1.1739438013576066e-06, "loss": 0.3791, "step": 2874 }, { "epoch": 2.3984983314794217, "grad_norm": 0.3284822702407837, "learning_rate": 1.1708186865057808e-06, "loss": 0.4233, "step": 2875 }, { "epoch": 2.3993325917686317, "grad_norm": 0.3061011731624603, "learning_rate": 1.1676971851335861e-06, "loss": 0.4174, "step": 2876 }, { "epoch": 2.400166852057842, "grad_norm": 0.30806103348731995, "learning_rate": 1.1645793001866862e-06, "loss": 0.3817, "step": 2877 }, { "epoch": 2.4010011123470525, "grad_norm": 0.3035253882408142, "learning_rate": 1.161465034607332e-06, "loss": 0.4032, "step": 2878 }, { "epoch": 2.4018353726362625, "grad_norm": 0.31209564208984375, "learning_rate": 1.158354391334362e-06, "loss": 0.3962, "step": 2879 }, { "epoch": 2.402669632925473, "grad_norm": 0.3320656716823578, "learning_rate": 1.1552473733031893e-06, "loss": 0.4263, "step": 2880 }, { "epoch": 2.403503893214683, "grad_norm": 0.2853155732154846, "learning_rate": 1.1521439834458148e-06, "loss": 0.3756, "step": 2881 }, { "epoch": 2.4043381535038932, "grad_norm": 0.28644341230392456, "learning_rate": 1.149044224690809e-06, "loss": 0.3941, "step": 2882 }, { "epoch": 2.405172413793103, "grad_norm": 0.30181562900543213, "learning_rate": 1.1459480999633187e-06, "loss": 0.3951, "step": 2883 }, { "epoch": 2.4060066740823136, "grad_norm": 0.26482415199279785, "learning_rate": 1.1428556121850615e-06, "loss": 0.3318, "step": 2884 }, { "epoch": 2.406840934371524, "grad_norm": 0.298440545797348, "learning_rate": 1.1397667642743188e-06, "loss": 0.4276, "step": 2885 }, { "epoch": 2.407675194660734, "grad_norm": 0.3097051680088043, "learning_rate": 1.1366815591459446e-06, "loss": 0.4349, "step": 2886 }, { "epoch": 2.4085094549499444, "grad_norm": 0.32794445753097534, "learning_rate": 1.133599999711349e-06, "loss": 0.3973, "step": 2887 }, { "epoch": 2.409343715239155, "grad_norm": 0.290853351354599, "learning_rate": 1.1305220888785052e-06, "loss": 0.3844, "step": 2888 }, { "epoch": 2.4101779755283648, "grad_norm": 0.3135397434234619, "learning_rate": 1.1274478295519415e-06, "loss": 0.4639, "step": 2889 }, { "epoch": 2.411012235817575, "grad_norm": 0.3084127902984619, "learning_rate": 1.1243772246327416e-06, "loss": 0.3772, "step": 2890 }, { "epoch": 2.411846496106785, "grad_norm": 0.29788902401924133, "learning_rate": 1.121310277018538e-06, "loss": 0.3778, "step": 2891 }, { "epoch": 2.4126807563959956, "grad_norm": 0.2792569398880005, "learning_rate": 1.1182469896035193e-06, "loss": 0.3977, "step": 2892 }, { "epoch": 2.413515016685206, "grad_norm": 0.32008248567581177, "learning_rate": 1.1151873652784078e-06, "loss": 0.4342, "step": 2893 }, { "epoch": 2.414349276974416, "grad_norm": 0.30208131670951843, "learning_rate": 1.112131406930481e-06, "loss": 0.3892, "step": 2894 }, { "epoch": 2.4151835372636263, "grad_norm": 0.2919992506504059, "learning_rate": 1.1090791174435529e-06, "loss": 0.3686, "step": 2895 }, { "epoch": 2.4160177975528363, "grad_norm": 0.3303925395011902, "learning_rate": 1.1060304996979688e-06, "loss": 0.4216, "step": 2896 }, { "epoch": 2.4168520578420467, "grad_norm": 0.3172752559185028, "learning_rate": 1.1029855565706205e-06, "loss": 0.4161, "step": 2897 }, { "epoch": 2.417686318131257, "grad_norm": 0.2719566822052002, "learning_rate": 1.0999442909349217e-06, "loss": 0.3814, "step": 2898 }, { "epoch": 2.418520578420467, "grad_norm": 0.2920774817466736, "learning_rate": 1.0969067056608235e-06, "loss": 0.3596, "step": 2899 }, { "epoch": 2.4193548387096775, "grad_norm": 0.3034088909626007, "learning_rate": 1.0938728036148006e-06, "loss": 0.3906, "step": 2900 }, { "epoch": 2.420189098998888, "grad_norm": 0.324004590511322, "learning_rate": 1.0908425876598512e-06, "loss": 0.4308, "step": 2901 }, { "epoch": 2.421023359288098, "grad_norm": 0.30719316005706787, "learning_rate": 1.0878160606554972e-06, "loss": 0.384, "step": 2902 }, { "epoch": 2.4218576195773083, "grad_norm": 0.3040667772293091, "learning_rate": 1.084793225457777e-06, "loss": 0.407, "step": 2903 }, { "epoch": 2.4226918798665182, "grad_norm": 0.31594786047935486, "learning_rate": 1.0817740849192476e-06, "loss": 0.3721, "step": 2904 }, { "epoch": 2.4235261401557286, "grad_norm": 0.2929110527038574, "learning_rate": 1.0787586418889778e-06, "loss": 0.408, "step": 2905 }, { "epoch": 2.4243604004449386, "grad_norm": 0.28887131810188293, "learning_rate": 1.0757468992125476e-06, "loss": 0.3867, "step": 2906 }, { "epoch": 2.425194660734149, "grad_norm": 0.31083235144615173, "learning_rate": 1.0727388597320437e-06, "loss": 0.4285, "step": 2907 }, { "epoch": 2.4260289210233594, "grad_norm": 0.29951536655426025, "learning_rate": 1.0697345262860638e-06, "loss": 0.3696, "step": 2908 }, { "epoch": 2.4268631813125694, "grad_norm": 0.30940666794776917, "learning_rate": 1.066733901709699e-06, "loss": 0.4098, "step": 2909 }, { "epoch": 2.42769744160178, "grad_norm": 0.30643901228904724, "learning_rate": 1.0637369888345506e-06, "loss": 0.4039, "step": 2910 }, { "epoch": 2.42853170189099, "grad_norm": 0.2726360857486725, "learning_rate": 1.060743790488707e-06, "loss": 0.3194, "step": 2911 }, { "epoch": 2.4293659621802, "grad_norm": 0.29736897349357605, "learning_rate": 1.0577543094967613e-06, "loss": 0.4136, "step": 2912 }, { "epoch": 2.4302002224694106, "grad_norm": 0.3155801594257355, "learning_rate": 1.054768548679792e-06, "loss": 0.4103, "step": 2913 }, { "epoch": 2.4310344827586206, "grad_norm": 0.3199295401573181, "learning_rate": 1.0517865108553692e-06, "loss": 0.3945, "step": 2914 }, { "epoch": 2.431868743047831, "grad_norm": 0.31167230010032654, "learning_rate": 1.0488081988375493e-06, "loss": 0.3947, "step": 2915 }, { "epoch": 2.432703003337041, "grad_norm": 0.279599666595459, "learning_rate": 1.0458336154368732e-06, "loss": 0.3618, "step": 2916 }, { "epoch": 2.4335372636262513, "grad_norm": 0.3018079400062561, "learning_rate": 1.0428627634603628e-06, "loss": 0.4183, "step": 2917 }, { "epoch": 2.4343715239154617, "grad_norm": 0.2693529725074768, "learning_rate": 1.0398956457115195e-06, "loss": 0.3319, "step": 2918 }, { "epoch": 2.4352057842046717, "grad_norm": 0.31737491488456726, "learning_rate": 1.0369322649903197e-06, "loss": 0.461, "step": 2919 }, { "epoch": 2.436040044493882, "grad_norm": 0.27821725606918335, "learning_rate": 1.0339726240932124e-06, "loss": 0.3602, "step": 2920 }, { "epoch": 2.4368743047830925, "grad_norm": 0.3003993630409241, "learning_rate": 1.0310167258131232e-06, "loss": 0.4086, "step": 2921 }, { "epoch": 2.4377085650723025, "grad_norm": 0.29946818947792053, "learning_rate": 1.0280645729394368e-06, "loss": 0.4043, "step": 2922 }, { "epoch": 2.438542825361513, "grad_norm": 0.3006298243999481, "learning_rate": 1.0251161682580125e-06, "loss": 0.4309, "step": 2923 }, { "epoch": 2.439377085650723, "grad_norm": 0.2908306419849396, "learning_rate": 1.022171514551164e-06, "loss": 0.3571, "step": 2924 }, { "epoch": 2.4402113459399333, "grad_norm": 0.3080620765686035, "learning_rate": 1.0192306145976732e-06, "loss": 0.3945, "step": 2925 }, { "epoch": 2.4410456062291437, "grad_norm": 0.33930113911628723, "learning_rate": 1.0162934711727761e-06, "loss": 0.3563, "step": 2926 }, { "epoch": 2.4418798665183536, "grad_norm": 0.2985761761665344, "learning_rate": 1.0133600870481602e-06, "loss": 0.3924, "step": 2927 }, { "epoch": 2.442714126807564, "grad_norm": 0.29536521434783936, "learning_rate": 1.0104304649919733e-06, "loss": 0.4046, "step": 2928 }, { "epoch": 2.443548387096774, "grad_norm": 0.3101980686187744, "learning_rate": 1.0075046077688067e-06, "loss": 0.3688, "step": 2929 }, { "epoch": 2.4443826473859844, "grad_norm": 0.28825098276138306, "learning_rate": 1.0045825181397018e-06, "loss": 0.3967, "step": 2930 }, { "epoch": 2.445216907675195, "grad_norm": 0.3080148696899414, "learning_rate": 1.0016641988621434e-06, "loss": 0.4148, "step": 2931 }, { "epoch": 2.446051167964405, "grad_norm": 0.29866236448287964, "learning_rate": 9.987496526900593e-07, "loss": 0.3907, "step": 2932 }, { "epoch": 2.446885428253615, "grad_norm": 0.3033437430858612, "learning_rate": 9.958388823738146e-07, "loss": 0.4005, "step": 2933 }, { "epoch": 2.447719688542825, "grad_norm": 0.30878394842147827, "learning_rate": 9.929318906602176e-07, "loss": 0.3905, "step": 2934 }, { "epoch": 2.4485539488320356, "grad_norm": 0.2953241765499115, "learning_rate": 9.900286802925013e-07, "loss": 0.4159, "step": 2935 }, { "epoch": 2.449388209121246, "grad_norm": 0.30196720361709595, "learning_rate": 9.871292540103377e-07, "loss": 0.397, "step": 2936 }, { "epoch": 2.450222469410456, "grad_norm": 0.31449732184410095, "learning_rate": 9.84233614549825e-07, "loss": 0.4376, "step": 2937 }, { "epoch": 2.4510567296996664, "grad_norm": 0.2929903268814087, "learning_rate": 9.813417646434864e-07, "loss": 0.3942, "step": 2938 }, { "epoch": 2.4518909899888763, "grad_norm": 0.3063810169696808, "learning_rate": 9.784537070202754e-07, "loss": 0.398, "step": 2939 }, { "epoch": 2.4527252502780867, "grad_norm": 0.3011118173599243, "learning_rate": 9.755694444055574e-07, "loss": 0.4143, "step": 2940 }, { "epoch": 2.453559510567297, "grad_norm": 0.3112642168998718, "learning_rate": 9.726889795211253e-07, "loss": 0.4285, "step": 2941 }, { "epoch": 2.454393770856507, "grad_norm": 0.29086095094680786, "learning_rate": 9.698123150851835e-07, "loss": 0.3632, "step": 2942 }, { "epoch": 2.4552280311457175, "grad_norm": 0.3054104745388031, "learning_rate": 9.6693945381235e-07, "loss": 0.3912, "step": 2943 }, { "epoch": 2.456062291434928, "grad_norm": 0.2847020924091339, "learning_rate": 9.640703984136557e-07, "loss": 0.3432, "step": 2944 }, { "epoch": 2.456896551724138, "grad_norm": 0.314446359872818, "learning_rate": 9.612051515965388e-07, "loss": 0.3976, "step": 2945 }, { "epoch": 2.4577308120133483, "grad_norm": 0.30170273780822754, "learning_rate": 9.58343716064844e-07, "loss": 0.3758, "step": 2946 }, { "epoch": 2.4585650723025583, "grad_norm": 0.32040929794311523, "learning_rate": 9.554860945188193e-07, "loss": 0.3996, "step": 2947 }, { "epoch": 2.4593993325917687, "grad_norm": 0.2984631657600403, "learning_rate": 9.526322896551138e-07, "loss": 0.4057, "step": 2948 }, { "epoch": 2.4602335928809786, "grad_norm": 0.30202722549438477, "learning_rate": 9.497823041667742e-07, "loss": 0.3809, "step": 2949 }, { "epoch": 2.461067853170189, "grad_norm": 0.30153775215148926, "learning_rate": 9.469361407432431e-07, "loss": 0.4206, "step": 2950 }, { "epoch": 2.4619021134593995, "grad_norm": 0.30754852294921875, "learning_rate": 9.440938020703561e-07, "loss": 0.3669, "step": 2951 }, { "epoch": 2.4627363737486094, "grad_norm": 0.2698597013950348, "learning_rate": 9.412552908303436e-07, "loss": 0.3672, "step": 2952 }, { "epoch": 2.46357063403782, "grad_norm": 0.2915247082710266, "learning_rate": 9.384206097018156e-07, "loss": 0.3959, "step": 2953 }, { "epoch": 2.4644048943270302, "grad_norm": 0.30903884768486023, "learning_rate": 9.355897613597764e-07, "loss": 0.4097, "step": 2954 }, { "epoch": 2.46523915461624, "grad_norm": 0.2978751063346863, "learning_rate": 9.32762748475608e-07, "loss": 0.4205, "step": 2955 }, { "epoch": 2.4660734149054506, "grad_norm": 0.30277884006500244, "learning_rate": 9.299395737170758e-07, "loss": 0.3795, "step": 2956 }, { "epoch": 2.4669076751946606, "grad_norm": 0.3248867392539978, "learning_rate": 9.271202397483214e-07, "loss": 0.4167, "step": 2957 }, { "epoch": 2.467741935483871, "grad_norm": 0.31606051325798035, "learning_rate": 9.243047492298634e-07, "loss": 0.4157, "step": 2958 }, { "epoch": 2.468576195773081, "grad_norm": 0.2829505205154419, "learning_rate": 9.214931048185927e-07, "loss": 0.3603, "step": 2959 }, { "epoch": 2.4694104560622914, "grad_norm": 0.3315090239048004, "learning_rate": 9.186853091677711e-07, "loss": 0.3917, "step": 2960 }, { "epoch": 2.470244716351502, "grad_norm": 0.33292698860168457, "learning_rate": 9.158813649270276e-07, "loss": 0.3879, "step": 2961 }, { "epoch": 2.4710789766407117, "grad_norm": 0.28672370314598083, "learning_rate": 9.130812747423578e-07, "loss": 0.3661, "step": 2962 }, { "epoch": 2.471913236929922, "grad_norm": 0.2841660678386688, "learning_rate": 9.102850412561204e-07, "loss": 0.3616, "step": 2963 }, { "epoch": 2.4727474972191326, "grad_norm": 0.3068605065345764, "learning_rate": 9.074926671070322e-07, "loss": 0.3986, "step": 2964 }, { "epoch": 2.4735817575083425, "grad_norm": 0.29987263679504395, "learning_rate": 9.047041549301743e-07, "loss": 0.4044, "step": 2965 }, { "epoch": 2.474416017797553, "grad_norm": 0.3294799327850342, "learning_rate": 9.019195073569753e-07, "loss": 0.4093, "step": 2966 }, { "epoch": 2.475250278086763, "grad_norm": 0.2956344783306122, "learning_rate": 8.991387270152202e-07, "loss": 0.4045, "step": 2967 }, { "epoch": 2.4760845383759733, "grad_norm": 0.2836824655532837, "learning_rate": 8.9636181652905e-07, "loss": 0.369, "step": 2968 }, { "epoch": 2.4769187986651837, "grad_norm": 0.29442235827445984, "learning_rate": 8.935887785189435e-07, "loss": 0.3923, "step": 2969 }, { "epoch": 2.4777530589543937, "grad_norm": 0.3098742961883545, "learning_rate": 8.908196156017335e-07, "loss": 0.4045, "step": 2970 }, { "epoch": 2.478587319243604, "grad_norm": 0.3027343451976776, "learning_rate": 8.880543303905931e-07, "loss": 0.4158, "step": 2971 }, { "epoch": 2.479421579532814, "grad_norm": 0.2929569184780121, "learning_rate": 8.85292925495036e-07, "loss": 0.3776, "step": 2972 }, { "epoch": 2.4802558398220245, "grad_norm": 0.29937922954559326, "learning_rate": 8.825354035209138e-07, "loss": 0.3805, "step": 2973 }, { "epoch": 2.481090100111235, "grad_norm": 0.2864495813846588, "learning_rate": 8.797817670704151e-07, "loss": 0.3934, "step": 2974 }, { "epoch": 2.481924360400445, "grad_norm": 0.28260916471481323, "learning_rate": 8.770320187420606e-07, "loss": 0.422, "step": 2975 }, { "epoch": 2.4827586206896552, "grad_norm": 0.32750484347343445, "learning_rate": 8.742861611307029e-07, "loss": 0.4492, "step": 2976 }, { "epoch": 2.483592880978865, "grad_norm": 0.3051404356956482, "learning_rate": 8.715441968275229e-07, "loss": 0.372, "step": 2977 }, { "epoch": 2.4844271412680756, "grad_norm": 0.2971930205821991, "learning_rate": 8.688061284200266e-07, "loss": 0.3756, "step": 2978 }, { "epoch": 2.485261401557286, "grad_norm": 0.30458930134773254, "learning_rate": 8.660719584920441e-07, "loss": 0.384, "step": 2979 }, { "epoch": 2.486095661846496, "grad_norm": 0.3114638924598694, "learning_rate": 8.633416896237268e-07, "loss": 0.4362, "step": 2980 }, { "epoch": 2.4869299221357064, "grad_norm": 0.27140408754348755, "learning_rate": 8.606153243915466e-07, "loss": 0.3749, "step": 2981 }, { "epoch": 2.4877641824249164, "grad_norm": 0.2974949777126312, "learning_rate": 8.578928653682867e-07, "loss": 0.3922, "step": 2982 }, { "epoch": 2.488598442714127, "grad_norm": 0.3004611134529114, "learning_rate": 8.551743151230496e-07, "loss": 0.3994, "step": 2983 }, { "epoch": 2.489432703003337, "grad_norm": 0.280631959438324, "learning_rate": 8.524596762212461e-07, "loss": 0.3663, "step": 2984 }, { "epoch": 2.490266963292547, "grad_norm": 0.2980068624019623, "learning_rate": 8.497489512245971e-07, "loss": 0.4115, "step": 2985 }, { "epoch": 2.4911012235817576, "grad_norm": 0.2912701368331909, "learning_rate": 8.470421426911291e-07, "loss": 0.3923, "step": 2986 }, { "epoch": 2.491935483870968, "grad_norm": 0.2889801859855652, "learning_rate": 8.443392531751737e-07, "loss": 0.3729, "step": 2987 }, { "epoch": 2.492769744160178, "grad_norm": 0.2783116400241852, "learning_rate": 8.416402852273636e-07, "loss": 0.3869, "step": 2988 }, { "epoch": 2.4936040044493883, "grad_norm": 0.2838379144668579, "learning_rate": 8.389452413946314e-07, "loss": 0.4044, "step": 2989 }, { "epoch": 2.4944382647385983, "grad_norm": 0.3068615198135376, "learning_rate": 8.362541242202066e-07, "loss": 0.4498, "step": 2990 }, { "epoch": 2.4952725250278087, "grad_norm": 0.2967071235179901, "learning_rate": 8.335669362436121e-07, "loss": 0.3619, "step": 2991 }, { "epoch": 2.4961067853170187, "grad_norm": 0.3075029253959656, "learning_rate": 8.308836800006648e-07, "loss": 0.3977, "step": 2992 }, { "epoch": 2.496941045606229, "grad_norm": 0.2974574565887451, "learning_rate": 8.282043580234688e-07, "loss": 0.4391, "step": 2993 }, { "epoch": 2.4977753058954395, "grad_norm": 0.2833936810493469, "learning_rate": 8.255289728404208e-07, "loss": 0.4032, "step": 2994 }, { "epoch": 2.4986095661846495, "grad_norm": 0.2918832302093506, "learning_rate": 8.228575269761935e-07, "loss": 0.3685, "step": 2995 }, { "epoch": 2.49944382647386, "grad_norm": 0.29590904712677, "learning_rate": 8.201900229517523e-07, "loss": 0.3723, "step": 2996 }, { "epoch": 2.5002780867630703, "grad_norm": 0.3158923089504242, "learning_rate": 8.175264632843366e-07, "loss": 0.4125, "step": 2997 }, { "epoch": 2.5011123470522802, "grad_norm": 0.2983265221118927, "learning_rate": 8.148668504874624e-07, "loss": 0.4266, "step": 2998 }, { "epoch": 2.5019466073414907, "grad_norm": 0.2905604839324951, "learning_rate": 8.122111870709287e-07, "loss": 0.3594, "step": 2999 }, { "epoch": 2.5027808676307006, "grad_norm": 0.3031312823295593, "learning_rate": 8.095594755407971e-07, "loss": 0.393, "step": 3000 }, { "epoch": 2.503615127919911, "grad_norm": 0.31090012192726135, "learning_rate": 8.069117183994091e-07, "loss": 0.4171, "step": 3001 }, { "epoch": 2.504449388209121, "grad_norm": 0.30947375297546387, "learning_rate": 8.042679181453711e-07, "loss": 0.3684, "step": 3002 }, { "epoch": 2.5052836484983314, "grad_norm": 0.2908641993999481, "learning_rate": 8.01628077273554e-07, "loss": 0.3774, "step": 3003 }, { "epoch": 2.506117908787542, "grad_norm": 0.2930682599544525, "learning_rate": 7.989921982750942e-07, "loss": 0.3987, "step": 3004 }, { "epoch": 2.506952169076752, "grad_norm": 0.274057000875473, "learning_rate": 7.96360283637389e-07, "loss": 0.3918, "step": 3005 }, { "epoch": 2.507786429365962, "grad_norm": 0.29741156101226807, "learning_rate": 7.937323358440935e-07, "loss": 0.4212, "step": 3006 }, { "epoch": 2.5086206896551726, "grad_norm": 0.29102832078933716, "learning_rate": 7.911083573751244e-07, "loss": 0.4069, "step": 3007 }, { "epoch": 2.5094549499443826, "grad_norm": 0.28041404485702515, "learning_rate": 7.884883507066443e-07, "loss": 0.3705, "step": 3008 }, { "epoch": 2.510289210233593, "grad_norm": 0.2953092157840729, "learning_rate": 7.858723183110728e-07, "loss": 0.3412, "step": 3009 }, { "epoch": 2.5111234705228034, "grad_norm": 0.29672476649284363, "learning_rate": 7.832602626570824e-07, "loss": 0.4094, "step": 3010 }, { "epoch": 2.5119577308120133, "grad_norm": 0.3007173240184784, "learning_rate": 7.806521862095834e-07, "loss": 0.4108, "step": 3011 }, { "epoch": 2.5127919911012233, "grad_norm": 0.27530333399772644, "learning_rate": 7.780480914297422e-07, "loss": 0.3739, "step": 3012 }, { "epoch": 2.5136262513904337, "grad_norm": 0.294885516166687, "learning_rate": 7.754479807749571e-07, "loss": 0.4103, "step": 3013 }, { "epoch": 2.514460511679644, "grad_norm": 0.3044889569282532, "learning_rate": 7.728518566988758e-07, "loss": 0.3964, "step": 3014 }, { "epoch": 2.515294771968854, "grad_norm": 0.3085109293460846, "learning_rate": 7.702597216513786e-07, "loss": 0.4276, "step": 3015 }, { "epoch": 2.5161290322580645, "grad_norm": 0.3173978924751282, "learning_rate": 7.676715780785826e-07, "loss": 0.3753, "step": 3016 }, { "epoch": 2.516963292547275, "grad_norm": 0.3025672733783722, "learning_rate": 7.650874284228388e-07, "loss": 0.4069, "step": 3017 }, { "epoch": 2.517797552836485, "grad_norm": 0.28578469157218933, "learning_rate": 7.625072751227297e-07, "loss": 0.3768, "step": 3018 }, { "epoch": 2.5186318131256953, "grad_norm": 0.29843631386756897, "learning_rate": 7.599311206130655e-07, "loss": 0.3839, "step": 3019 }, { "epoch": 2.5194660734149057, "grad_norm": 0.29495760798454285, "learning_rate": 7.573589673248833e-07, "loss": 0.4132, "step": 3020 }, { "epoch": 2.5203003337041157, "grad_norm": 0.29186710715293884, "learning_rate": 7.54790817685444e-07, "loss": 0.3698, "step": 3021 }, { "epoch": 2.521134593993326, "grad_norm": 0.29619279503822327, "learning_rate": 7.522266741182305e-07, "loss": 0.3898, "step": 3022 }, { "epoch": 2.521968854282536, "grad_norm": 0.30557680130004883, "learning_rate": 7.49666539042948e-07, "loss": 0.4196, "step": 3023 }, { "epoch": 2.5228031145717464, "grad_norm": 0.28378981351852417, "learning_rate": 7.471104148755126e-07, "loss": 0.3934, "step": 3024 }, { "epoch": 2.5236373748609564, "grad_norm": 0.2965737283229828, "learning_rate": 7.445583040280636e-07, "loss": 0.4267, "step": 3025 }, { "epoch": 2.524471635150167, "grad_norm": 0.2945002615451813, "learning_rate": 7.420102089089448e-07, "loss": 0.3946, "step": 3026 }, { "epoch": 2.5253058954393772, "grad_norm": 0.3174886703491211, "learning_rate": 7.394661319227175e-07, "loss": 0.463, "step": 3027 }, { "epoch": 2.526140155728587, "grad_norm": 0.2920013666152954, "learning_rate": 7.369260754701485e-07, "loss": 0.3619, "step": 3028 }, { "epoch": 2.5269744160177976, "grad_norm": 0.286899596452713, "learning_rate": 7.343900419482064e-07, "loss": 0.3912, "step": 3029 }, { "epoch": 2.527808676307008, "grad_norm": 0.2929022014141083, "learning_rate": 7.318580337500713e-07, "loss": 0.3788, "step": 3030 }, { "epoch": 2.528642936596218, "grad_norm": 0.3051469326019287, "learning_rate": 7.293300532651193e-07, "loss": 0.384, "step": 3031 }, { "epoch": 2.5294771968854284, "grad_norm": 0.2993219494819641, "learning_rate": 7.268061028789275e-07, "loss": 0.3865, "step": 3032 }, { "epoch": 2.5303114571746383, "grad_norm": 0.3145145773887634, "learning_rate": 7.242861849732696e-07, "loss": 0.4207, "step": 3033 }, { "epoch": 2.5311457174638488, "grad_norm": 0.29131901264190674, "learning_rate": 7.217703019261135e-07, "loss": 0.3898, "step": 3034 }, { "epoch": 2.5319799777530587, "grad_norm": 0.2671528160572052, "learning_rate": 7.192584561116195e-07, "loss": 0.3533, "step": 3035 }, { "epoch": 2.532814238042269, "grad_norm": 0.30514389276504517, "learning_rate": 7.167506499001415e-07, "loss": 0.4137, "step": 3036 }, { "epoch": 2.5336484983314795, "grad_norm": 0.29500308632850647, "learning_rate": 7.142468856582141e-07, "loss": 0.3624, "step": 3037 }, { "epoch": 2.5344827586206895, "grad_norm": 0.29682791233062744, "learning_rate": 7.117471657485664e-07, "loss": 0.3733, "step": 3038 }, { "epoch": 2.5353170189099, "grad_norm": 0.302397757768631, "learning_rate": 7.09251492530103e-07, "loss": 0.4029, "step": 3039 }, { "epoch": 2.5361512791991103, "grad_norm": 0.31555670499801636, "learning_rate": 7.067598683579135e-07, "loss": 0.3906, "step": 3040 }, { "epoch": 2.5369855394883203, "grad_norm": 0.2922108769416809, "learning_rate": 7.042722955832703e-07, "loss": 0.4104, "step": 3041 }, { "epoch": 2.5378197997775307, "grad_norm": 0.2905091643333435, "learning_rate": 7.017887765536129e-07, "loss": 0.3973, "step": 3042 }, { "epoch": 2.5386540600667407, "grad_norm": 0.28624269366264343, "learning_rate": 6.993093136125656e-07, "loss": 0.3732, "step": 3043 }, { "epoch": 2.539488320355951, "grad_norm": 0.2894628942012787, "learning_rate": 6.968339090999188e-07, "loss": 0.3692, "step": 3044 }, { "epoch": 2.540322580645161, "grad_norm": 0.32212913036346436, "learning_rate": 6.943625653516356e-07, "loss": 0.4281, "step": 3045 }, { "epoch": 2.5411568409343714, "grad_norm": 0.3058313727378845, "learning_rate": 6.918952846998461e-07, "loss": 0.3859, "step": 3046 }, { "epoch": 2.541991101223582, "grad_norm": 0.29874539375305176, "learning_rate": 6.894320694728468e-07, "loss": 0.3822, "step": 3047 }, { "epoch": 2.542825361512792, "grad_norm": 0.3201921582221985, "learning_rate": 6.86972921995096e-07, "loss": 0.3982, "step": 3048 }, { "epoch": 2.5436596218020022, "grad_norm": 0.2768486440181732, "learning_rate": 6.845178445872181e-07, "loss": 0.3759, "step": 3049 }, { "epoch": 2.5444938820912126, "grad_norm": 0.33025282621383667, "learning_rate": 6.820668395659907e-07, "loss": 0.4758, "step": 3050 }, { "epoch": 2.5453281423804226, "grad_norm": 0.29065003991127014, "learning_rate": 6.796199092443523e-07, "loss": 0.38, "step": 3051 }, { "epoch": 2.546162402669633, "grad_norm": 0.28353971242904663, "learning_rate": 6.77177055931395e-07, "loss": 0.3775, "step": 3052 }, { "epoch": 2.5469966629588434, "grad_norm": 0.29327481985092163, "learning_rate": 6.747382819323622e-07, "loss": 0.4186, "step": 3053 }, { "epoch": 2.5478309232480534, "grad_norm": 0.30810412764549255, "learning_rate": 6.723035895486529e-07, "loss": 0.3781, "step": 3054 }, { "epoch": 2.548665183537264, "grad_norm": 0.2987021207809448, "learning_rate": 6.698729810778065e-07, "loss": 0.3441, "step": 3055 }, { "epoch": 2.5494994438264738, "grad_norm": 0.33564555644989014, "learning_rate": 6.674464588135165e-07, "loss": 0.4119, "step": 3056 }, { "epoch": 2.550333704115684, "grad_norm": 0.30483078956604004, "learning_rate": 6.650240250456147e-07, "loss": 0.369, "step": 3057 }, { "epoch": 2.551167964404894, "grad_norm": 0.3009977638721466, "learning_rate": 6.626056820600769e-07, "loss": 0.3956, "step": 3058 }, { "epoch": 2.5520022246941045, "grad_norm": 0.31366094946861267, "learning_rate": 6.601914321390185e-07, "loss": 0.4564, "step": 3059 }, { "epoch": 2.552836484983315, "grad_norm": 0.29573556780815125, "learning_rate": 6.577812775606929e-07, "loss": 0.3991, "step": 3060 }, { "epoch": 2.553670745272525, "grad_norm": 0.3137950301170349, "learning_rate": 6.553752205994874e-07, "loss": 0.4001, "step": 3061 }, { "epoch": 2.5545050055617353, "grad_norm": 0.2920612692832947, "learning_rate": 6.529732635259234e-07, "loss": 0.3568, "step": 3062 }, { "epoch": 2.5553392658509457, "grad_norm": 0.29269108176231384, "learning_rate": 6.50575408606654e-07, "loss": 0.4059, "step": 3063 }, { "epoch": 2.5561735261401557, "grad_norm": 0.3133447468280792, "learning_rate": 6.481816581044587e-07, "loss": 0.3911, "step": 3064 }, { "epoch": 2.557007786429366, "grad_norm": 0.3044225871562958, "learning_rate": 6.457920142782487e-07, "loss": 0.4169, "step": 3065 }, { "epoch": 2.557842046718576, "grad_norm": 0.29587242007255554, "learning_rate": 6.43406479383053e-07, "loss": 0.3656, "step": 3066 }, { "epoch": 2.5586763070077865, "grad_norm": 0.29437798261642456, "learning_rate": 6.410250556700309e-07, "loss": 0.4041, "step": 3067 }, { "epoch": 2.5595105672969964, "grad_norm": 0.29454711079597473, "learning_rate": 6.386477453864537e-07, "loss": 0.3975, "step": 3068 }, { "epoch": 2.560344827586207, "grad_norm": 0.28814536333084106, "learning_rate": 6.36274550775719e-07, "loss": 0.4091, "step": 3069 }, { "epoch": 2.5611790878754173, "grad_norm": 0.29001808166503906, "learning_rate": 6.33905474077337e-07, "loss": 0.4007, "step": 3070 }, { "epoch": 2.5620133481646272, "grad_norm": 0.31869572401046753, "learning_rate": 6.315405175269279e-07, "loss": 0.3992, "step": 3071 }, { "epoch": 2.5628476084538376, "grad_norm": 0.2786090672016144, "learning_rate": 6.291796833562319e-07, "loss": 0.3774, "step": 3072 }, { "epoch": 2.563681868743048, "grad_norm": 0.3102768361568451, "learning_rate": 6.268229737930931e-07, "loss": 0.4567, "step": 3073 }, { "epoch": 2.564516129032258, "grad_norm": 0.2771594524383545, "learning_rate": 6.244703910614663e-07, "loss": 0.3539, "step": 3074 }, { "epoch": 2.5653503893214684, "grad_norm": 0.30755615234375, "learning_rate": 6.221219373814097e-07, "loss": 0.4053, "step": 3075 }, { "epoch": 2.5661846496106784, "grad_norm": 0.2992633283138275, "learning_rate": 6.197776149690871e-07, "loss": 0.3537, "step": 3076 }, { "epoch": 2.567018909899889, "grad_norm": 0.3098567724227905, "learning_rate": 6.174374260367611e-07, "loss": 0.396, "step": 3077 }, { "epoch": 2.5678531701890988, "grad_norm": 0.28488320112228394, "learning_rate": 6.151013727927984e-07, "loss": 0.3572, "step": 3078 }, { "epoch": 2.568687430478309, "grad_norm": 0.2765231132507324, "learning_rate": 6.127694574416571e-07, "loss": 0.4005, "step": 3079 }, { "epoch": 2.5695216907675196, "grad_norm": 0.2999037504196167, "learning_rate": 6.10441682183896e-07, "loss": 0.4081, "step": 3080 }, { "epoch": 2.5703559510567295, "grad_norm": 0.2924880087375641, "learning_rate": 6.081180492161626e-07, "loss": 0.361, "step": 3081 }, { "epoch": 2.57119021134594, "grad_norm": 0.2962203919887543, "learning_rate": 6.057985607311962e-07, "loss": 0.4021, "step": 3082 }, { "epoch": 2.5720244716351504, "grad_norm": 0.2827346622943878, "learning_rate": 6.034832189178302e-07, "loss": 0.3732, "step": 3083 }, { "epoch": 2.5728587319243603, "grad_norm": 0.3268202543258667, "learning_rate": 6.011720259609754e-07, "loss": 0.4313, "step": 3084 }, { "epoch": 2.5736929922135707, "grad_norm": 0.286293625831604, "learning_rate": 5.988649840416377e-07, "loss": 0.3697, "step": 3085 }, { "epoch": 2.574527252502781, "grad_norm": 0.2841409146785736, "learning_rate": 5.965620953368983e-07, "loss": 0.3671, "step": 3086 }, { "epoch": 2.575361512791991, "grad_norm": 0.28253480792045593, "learning_rate": 5.942633620199234e-07, "loss": 0.402, "step": 3087 }, { "epoch": 2.576195773081201, "grad_norm": 0.2776295840740204, "learning_rate": 5.919687862599549e-07, "loss": 0.3632, "step": 3088 }, { "epoch": 2.5770300333704115, "grad_norm": 0.2871991991996765, "learning_rate": 5.896783702223136e-07, "loss": 0.4149, "step": 3089 }, { "epoch": 2.577864293659622, "grad_norm": 0.3033123314380646, "learning_rate": 5.873921160683943e-07, "loss": 0.4177, "step": 3090 }, { "epoch": 2.578698553948832, "grad_norm": 0.32237735390663147, "learning_rate": 5.85110025955663e-07, "loss": 0.3964, "step": 3091 }, { "epoch": 2.5795328142380423, "grad_norm": 0.29377758502960205, "learning_rate": 5.828321020376587e-07, "loss": 0.4079, "step": 3092 }, { "epoch": 2.5803670745272527, "grad_norm": 0.3066837191581726, "learning_rate": 5.805583464639863e-07, "loss": 0.3817, "step": 3093 }, { "epoch": 2.5812013348164626, "grad_norm": 0.2965717911720276, "learning_rate": 5.782887613803195e-07, "loss": 0.3603, "step": 3094 }, { "epoch": 2.582035595105673, "grad_norm": 0.2947966158390045, "learning_rate": 5.760233489283939e-07, "loss": 0.3981, "step": 3095 }, { "epoch": 2.5828698553948835, "grad_norm": 0.2998191714286804, "learning_rate": 5.737621112460117e-07, "loss": 0.4158, "step": 3096 }, { "epoch": 2.5837041156840934, "grad_norm": 0.2824717164039612, "learning_rate": 5.715050504670288e-07, "loss": 0.3856, "step": 3097 }, { "epoch": 2.584538375973304, "grad_norm": 0.279965341091156, "learning_rate": 5.692521687213671e-07, "loss": 0.3923, "step": 3098 }, { "epoch": 2.585372636262514, "grad_norm": 0.3081649839878082, "learning_rate": 5.670034681349995e-07, "loss": 0.4116, "step": 3099 }, { "epoch": 2.586206896551724, "grad_norm": 0.30319562554359436, "learning_rate": 5.647589508299555e-07, "loss": 0.4466, "step": 3100 }, { "epoch": 2.587041156840934, "grad_norm": 0.28980085253715515, "learning_rate": 5.625186189243159e-07, "loss": 0.3823, "step": 3101 }, { "epoch": 2.5878754171301446, "grad_norm": 0.29519158601760864, "learning_rate": 5.602824745322127e-07, "loss": 0.3723, "step": 3102 }, { "epoch": 2.588709677419355, "grad_norm": 0.3102172911167145, "learning_rate": 5.580505197638264e-07, "loss": 0.405, "step": 3103 }, { "epoch": 2.589543937708565, "grad_norm": 0.2959599494934082, "learning_rate": 5.558227567253832e-07, "loss": 0.3686, "step": 3104 }, { "epoch": 2.5903781979977754, "grad_norm": 0.2885022461414337, "learning_rate": 5.535991875191532e-07, "loss": 0.3996, "step": 3105 }, { "epoch": 2.5912124582869858, "grad_norm": 0.2985133230686188, "learning_rate": 5.513798142434512e-07, "loss": 0.4175, "step": 3106 }, { "epoch": 2.5920467185761957, "grad_norm": 0.30010277032852173, "learning_rate": 5.491646389926298e-07, "loss": 0.3998, "step": 3107 }, { "epoch": 2.592880978865406, "grad_norm": 0.3252725899219513, "learning_rate": 5.469536638570799e-07, "loss": 0.4158, "step": 3108 }, { "epoch": 2.593715239154616, "grad_norm": 0.2886245548725128, "learning_rate": 5.447468909232345e-07, "loss": 0.3806, "step": 3109 }, { "epoch": 2.5945494994438265, "grad_norm": 0.30152130126953125, "learning_rate": 5.425443222735527e-07, "loss": 0.396, "step": 3110 }, { "epoch": 2.5953837597330365, "grad_norm": 0.29885682463645935, "learning_rate": 5.403459599865307e-07, "loss": 0.3987, "step": 3111 }, { "epoch": 2.596218020022247, "grad_norm": 0.2793337106704712, "learning_rate": 5.381518061366975e-07, "loss": 0.3828, "step": 3112 }, { "epoch": 2.5970522803114573, "grad_norm": 0.2943042516708374, "learning_rate": 5.359618627946039e-07, "loss": 0.3965, "step": 3113 }, { "epoch": 2.5978865406006673, "grad_norm": 0.30580875277519226, "learning_rate": 5.337761320268353e-07, "loss": 0.381, "step": 3114 }, { "epoch": 2.5987208008898777, "grad_norm": 0.28757473826408386, "learning_rate": 5.315946158959962e-07, "loss": 0.3933, "step": 3115 }, { "epoch": 2.599555061179088, "grad_norm": 0.3094478249549866, "learning_rate": 5.294173164607158e-07, "loss": 0.4474, "step": 3116 }, { "epoch": 2.600389321468298, "grad_norm": 0.28737872838974, "learning_rate": 5.272442357756447e-07, "loss": 0.3647, "step": 3117 }, { "epoch": 2.6012235817575085, "grad_norm": 0.31689509749412537, "learning_rate": 5.250753758914506e-07, "loss": 0.4149, "step": 3118 }, { "epoch": 2.6020578420467184, "grad_norm": 0.3034818768501282, "learning_rate": 5.229107388548205e-07, "loss": 0.3823, "step": 3119 }, { "epoch": 2.602892102335929, "grad_norm": 0.2706875503063202, "learning_rate": 5.207503267084546e-07, "loss": 0.3612, "step": 3120 }, { "epoch": 2.603726362625139, "grad_norm": 0.30222293734550476, "learning_rate": 5.185941414910673e-07, "loss": 0.446, "step": 3121 }, { "epoch": 2.604560622914349, "grad_norm": 0.30778414011001587, "learning_rate": 5.164421852373824e-07, "loss": 0.4694, "step": 3122 }, { "epoch": 2.6053948832035596, "grad_norm": 0.2730533182621002, "learning_rate": 5.142944599781358e-07, "loss": 0.354, "step": 3123 }, { "epoch": 2.6062291434927696, "grad_norm": 0.2979160249233246, "learning_rate": 5.121509677400677e-07, "loss": 0.3957, "step": 3124 }, { "epoch": 2.60706340378198, "grad_norm": 0.31575047969818115, "learning_rate": 5.100117105459279e-07, "loss": 0.427, "step": 3125 }, { "epoch": 2.6078976640711904, "grad_norm": 0.29879340529441833, "learning_rate": 5.078766904144633e-07, "loss": 0.3818, "step": 3126 }, { "epoch": 2.6087319243604004, "grad_norm": 0.29934337735176086, "learning_rate": 5.057459093604294e-07, "loss": 0.4158, "step": 3127 }, { "epoch": 2.6095661846496108, "grad_norm": 0.3106815218925476, "learning_rate": 5.036193693945767e-07, "loss": 0.4012, "step": 3128 }, { "epoch": 2.610400444938821, "grad_norm": 0.30854007601737976, "learning_rate": 5.014970725236556e-07, "loss": 0.42, "step": 3129 }, { "epoch": 2.611234705228031, "grad_norm": 0.2966778874397278, "learning_rate": 4.993790207504112e-07, "loss": 0.3763, "step": 3130 }, { "epoch": 2.612068965517241, "grad_norm": 0.32756224274635315, "learning_rate": 4.972652160735836e-07, "loss": 0.4157, "step": 3131 }, { "epoch": 2.6129032258064515, "grad_norm": 0.28828343749046326, "learning_rate": 4.951556604879049e-07, "loss": 0.3716, "step": 3132 }, { "epoch": 2.613737486095662, "grad_norm": 0.29351601004600525, "learning_rate": 4.930503559840965e-07, "loss": 0.3944, "step": 3133 }, { "epoch": 2.614571746384872, "grad_norm": 0.304437518119812, "learning_rate": 4.9094930454887e-07, "loss": 0.4165, "step": 3134 }, { "epoch": 2.6154060066740823, "grad_norm": 0.2677256166934967, "learning_rate": 4.888525081649214e-07, "loss": 0.3541, "step": 3135 }, { "epoch": 2.6162402669632927, "grad_norm": 0.27725252509117126, "learning_rate": 4.867599688109337e-07, "loss": 0.3744, "step": 3136 }, { "epoch": 2.6170745272525027, "grad_norm": 0.3014366328716278, "learning_rate": 4.846716884615693e-07, "loss": 0.4242, "step": 3137 }, { "epoch": 2.617908787541713, "grad_norm": 0.27725306153297424, "learning_rate": 4.82587669087477e-07, "loss": 0.3515, "step": 3138 }, { "epoch": 2.6187430478309235, "grad_norm": 0.30495139956474304, "learning_rate": 4.80507912655277e-07, "loss": 0.4103, "step": 3139 }, { "epoch": 2.6195773081201335, "grad_norm": 0.31752339005470276, "learning_rate": 4.784324211275743e-07, "loss": 0.4061, "step": 3140 }, { "epoch": 2.620411568409344, "grad_norm": 0.2955370545387268, "learning_rate": 4.7636119646294555e-07, "loss": 0.392, "step": 3141 }, { "epoch": 2.621245828698554, "grad_norm": 0.302377313375473, "learning_rate": 4.7429424061593845e-07, "loss": 0.3902, "step": 3142 }, { "epoch": 2.6220800889877642, "grad_norm": 0.28408077359199524, "learning_rate": 4.722315555370793e-07, "loss": 0.3969, "step": 3143 }, { "epoch": 2.622914349276974, "grad_norm": 0.2815459668636322, "learning_rate": 4.7017314317285565e-07, "loss": 0.3647, "step": 3144 }, { "epoch": 2.6237486095661846, "grad_norm": 0.2965734899044037, "learning_rate": 4.681190054657303e-07, "loss": 0.4379, "step": 3145 }, { "epoch": 2.624582869855395, "grad_norm": 0.29109224677085876, "learning_rate": 4.660691443541282e-07, "loss": 0.3814, "step": 3146 }, { "epoch": 2.625417130144605, "grad_norm": 0.29665103554725647, "learning_rate": 4.640235617724409e-07, "loss": 0.4522, "step": 3147 }, { "epoch": 2.6262513904338154, "grad_norm": 0.303615003824234, "learning_rate": 4.619822596510193e-07, "loss": 0.3779, "step": 3148 }, { "epoch": 2.627085650723026, "grad_norm": 0.3119533658027649, "learning_rate": 4.5994523991617855e-07, "loss": 0.416, "step": 3149 }, { "epoch": 2.6279199110122358, "grad_norm": 0.2905789613723755, "learning_rate": 4.5791250449018943e-07, "loss": 0.3898, "step": 3150 }, { "epoch": 2.628754171301446, "grad_norm": 0.3045509457588196, "learning_rate": 4.558840552912841e-07, "loss": 0.3798, "step": 3151 }, { "epoch": 2.629588431590656, "grad_norm": 0.29854193329811096, "learning_rate": 4.538598942336442e-07, "loss": 0.3908, "step": 3152 }, { "epoch": 2.6304226918798665, "grad_norm": 0.2780359387397766, "learning_rate": 4.5184002322740784e-07, "loss": 0.3559, "step": 3153 }, { "epoch": 2.6312569521690765, "grad_norm": 0.3153265416622162, "learning_rate": 4.4982444417866753e-07, "loss": 0.4158, "step": 3154 }, { "epoch": 2.632091212458287, "grad_norm": 0.301856130361557, "learning_rate": 4.478131589894591e-07, "loss": 0.4152, "step": 3155 }, { "epoch": 2.6329254727474973, "grad_norm": 0.29711899161338806, "learning_rate": 4.4580616955777365e-07, "loss": 0.3871, "step": 3156 }, { "epoch": 2.6337597330367073, "grad_norm": 0.3019917607307434, "learning_rate": 4.4380347777754073e-07, "loss": 0.4222, "step": 3157 }, { "epoch": 2.6345939933259177, "grad_norm": 0.2791317403316498, "learning_rate": 4.418050855386413e-07, "loss": 0.3496, "step": 3158 }, { "epoch": 2.635428253615128, "grad_norm": 0.28393420577049255, "learning_rate": 4.398109947268958e-07, "loss": 0.4157, "step": 3159 }, { "epoch": 2.636262513904338, "grad_norm": 0.29753378033638, "learning_rate": 4.3782120722406565e-07, "loss": 0.4204, "step": 3160 }, { "epoch": 2.6370967741935485, "grad_norm": 0.2990167438983917, "learning_rate": 4.358357249078515e-07, "loss": 0.3741, "step": 3161 }, { "epoch": 2.637931034482759, "grad_norm": 0.32196131348609924, "learning_rate": 4.3385454965189087e-07, "loss": 0.4594, "step": 3162 }, { "epoch": 2.638765294771969, "grad_norm": 0.26464056968688965, "learning_rate": 4.3187768332575875e-07, "loss": 0.3351, "step": 3163 }, { "epoch": 2.639599555061179, "grad_norm": 0.2900596559047699, "learning_rate": 4.299051277949612e-07, "loss": 0.3575, "step": 3164 }, { "epoch": 2.6404338153503892, "grad_norm": 0.3220163881778717, "learning_rate": 4.279368849209381e-07, "loss": 0.4239, "step": 3165 }, { "epoch": 2.6412680756395996, "grad_norm": 0.291792094707489, "learning_rate": 4.2597295656105877e-07, "loss": 0.3748, "step": 3166 }, { "epoch": 2.6421023359288096, "grad_norm": 0.2969697415828705, "learning_rate": 4.2401334456862344e-07, "loss": 0.4162, "step": 3167 }, { "epoch": 2.64293659621802, "grad_norm": 0.2884451150894165, "learning_rate": 4.220580507928546e-07, "loss": 0.3842, "step": 3168 }, { "epoch": 2.6437708565072304, "grad_norm": 0.30128851532936096, "learning_rate": 4.201070770789045e-07, "loss": 0.4087, "step": 3169 }, { "epoch": 2.6446051167964404, "grad_norm": 0.3018232583999634, "learning_rate": 4.181604252678445e-07, "loss": 0.3865, "step": 3170 }, { "epoch": 2.645439377085651, "grad_norm": 0.2972775399684906, "learning_rate": 4.162180971966712e-07, "loss": 0.3957, "step": 3171 }, { "epoch": 2.646273637374861, "grad_norm": 0.2743283808231354, "learning_rate": 4.1428009469829947e-07, "loss": 0.4125, "step": 3172 }, { "epoch": 2.647107897664071, "grad_norm": 0.2790834605693817, "learning_rate": 4.123464196015592e-07, "loss": 0.4226, "step": 3173 }, { "epoch": 2.6479421579532816, "grad_norm": 0.28161856532096863, "learning_rate": 4.1041707373120354e-07, "loss": 0.3932, "step": 3174 }, { "epoch": 2.6487764182424915, "grad_norm": 0.28378602862358093, "learning_rate": 4.084920589078939e-07, "loss": 0.4071, "step": 3175 }, { "epoch": 2.649610678531702, "grad_norm": 0.2785118520259857, "learning_rate": 4.0657137694820826e-07, "loss": 0.4014, "step": 3176 }, { "epoch": 2.650444938820912, "grad_norm": 0.2589441239833832, "learning_rate": 4.04655029664634e-07, "loss": 0.3499, "step": 3177 }, { "epoch": 2.6512791991101223, "grad_norm": 0.31842419505119324, "learning_rate": 4.027430188655684e-07, "loss": 0.4293, "step": 3178 }, { "epoch": 2.6521134593993327, "grad_norm": 0.3101239502429962, "learning_rate": 4.0083534635531607e-07, "loss": 0.4449, "step": 3179 }, { "epoch": 2.6529477196885427, "grad_norm": 0.2996217906475067, "learning_rate": 3.989320139340919e-07, "loss": 0.3908, "step": 3180 }, { "epoch": 2.653781979977753, "grad_norm": 0.2939760386943817, "learning_rate": 3.9703302339800687e-07, "loss": 0.3796, "step": 3181 }, { "epoch": 2.6546162402669635, "grad_norm": 0.2922966182231903, "learning_rate": 3.951383765390832e-07, "loss": 0.3803, "step": 3182 }, { "epoch": 2.6554505005561735, "grad_norm": 0.2822796702384949, "learning_rate": 3.932480751452383e-07, "loss": 0.4098, "step": 3183 }, { "epoch": 2.656284760845384, "grad_norm": 0.29443418979644775, "learning_rate": 3.913621210002905e-07, "loss": 0.4245, "step": 3184 }, { "epoch": 2.657119021134594, "grad_norm": 0.2741539180278778, "learning_rate": 3.894805158839593e-07, "loss": 0.3509, "step": 3185 }, { "epoch": 2.6579532814238043, "grad_norm": 0.29160940647125244, "learning_rate": 3.8760326157185336e-07, "loss": 0.402, "step": 3186 }, { "epoch": 2.6587875417130142, "grad_norm": 0.29474547505378723, "learning_rate": 3.8573035983548167e-07, "loss": 0.4109, "step": 3187 }, { "epoch": 2.6596218020022246, "grad_norm": 0.3099055588245392, "learning_rate": 3.8386181244224274e-07, "loss": 0.4323, "step": 3188 }, { "epoch": 2.660456062291435, "grad_norm": 0.29274606704711914, "learning_rate": 3.8199762115542716e-07, "loss": 0.3866, "step": 3189 }, { "epoch": 2.661290322580645, "grad_norm": 0.31940582394599915, "learning_rate": 3.8013778773421386e-07, "loss": 0.436, "step": 3190 }, { "epoch": 2.6621245828698554, "grad_norm": 0.27543389797210693, "learning_rate": 3.7828231393367054e-07, "loss": 0.3315, "step": 3191 }, { "epoch": 2.662958843159066, "grad_norm": 0.28192925453186035, "learning_rate": 3.764312015047489e-07, "loss": 0.3479, "step": 3192 }, { "epoch": 2.663793103448276, "grad_norm": 0.305402010679245, "learning_rate": 3.745844521942893e-07, "loss": 0.4253, "step": 3193 }, { "epoch": 2.664627363737486, "grad_norm": 0.2898850739002228, "learning_rate": 3.7274206774500886e-07, "loss": 0.3933, "step": 3194 }, { "epoch": 2.665461624026696, "grad_norm": 0.30907320976257324, "learning_rate": 3.709040498955102e-07, "loss": 0.4254, "step": 3195 }, { "epoch": 2.6662958843159066, "grad_norm": 0.3104950487613678, "learning_rate": 3.690704003802725e-07, "loss": 0.4048, "step": 3196 }, { "epoch": 2.6671301446051165, "grad_norm": 0.2981571555137634, "learning_rate": 3.672411209296545e-07, "loss": 0.4106, "step": 3197 }, { "epoch": 2.667964404894327, "grad_norm": 0.3097781538963318, "learning_rate": 3.6541621326989183e-07, "loss": 0.3982, "step": 3198 }, { "epoch": 2.6687986651835374, "grad_norm": 0.3129294812679291, "learning_rate": 3.6359567912309045e-07, "loss": 0.4113, "step": 3199 }, { "epoch": 2.6696329254727473, "grad_norm": 0.29168057441711426, "learning_rate": 3.617795202072344e-07, "loss": 0.3962, "step": 3200 }, { "epoch": 2.6704671857619577, "grad_norm": 0.2945079505443573, "learning_rate": 3.5996773823617604e-07, "loss": 0.3869, "step": 3201 }, { "epoch": 2.671301446051168, "grad_norm": 0.3025236129760742, "learning_rate": 3.581603349196372e-07, "loss": 0.4107, "step": 3202 }, { "epoch": 2.672135706340378, "grad_norm": 0.2752647399902344, "learning_rate": 3.5635731196320853e-07, "loss": 0.3899, "step": 3203 }, { "epoch": 2.6729699666295885, "grad_norm": 0.260515034198761, "learning_rate": 3.5455867106834684e-07, "loss": 0.3405, "step": 3204 }, { "epoch": 2.673804226918799, "grad_norm": 0.2956203520298004, "learning_rate": 3.527644139323738e-07, "loss": 0.4491, "step": 3205 }, { "epoch": 2.674638487208009, "grad_norm": 0.3023238182067871, "learning_rate": 3.509745422484745e-07, "loss": 0.411, "step": 3206 }, { "epoch": 2.675472747497219, "grad_norm": 0.3026939630508423, "learning_rate": 3.49189057705695e-07, "loss": 0.4485, "step": 3207 }, { "epoch": 2.6763070077864293, "grad_norm": 0.2962052524089813, "learning_rate": 3.4740796198894143e-07, "loss": 0.373, "step": 3208 }, { "epoch": 2.6771412680756397, "grad_norm": 0.27983933687210083, "learning_rate": 3.4563125677897936e-07, "loss": 0.3373, "step": 3209 }, { "epoch": 2.6779755283648496, "grad_norm": 0.30185624957084656, "learning_rate": 3.4385894375242867e-07, "loss": 0.4108, "step": 3210 }, { "epoch": 2.67880978865406, "grad_norm": 0.2960197925567627, "learning_rate": 3.4209102458176936e-07, "loss": 0.401, "step": 3211 }, { "epoch": 2.6796440489432705, "grad_norm": 0.3187948763370514, "learning_rate": 3.403275009353285e-07, "loss": 0.4025, "step": 3212 }, { "epoch": 2.6804783092324804, "grad_norm": 0.2923695743083954, "learning_rate": 3.3856837447729054e-07, "loss": 0.3752, "step": 3213 }, { "epoch": 2.681312569521691, "grad_norm": 0.3134532868862152, "learning_rate": 3.368136468676891e-07, "loss": 0.3898, "step": 3214 }, { "epoch": 2.6821468298109012, "grad_norm": 0.2817475199699402, "learning_rate": 3.3506331976240367e-07, "loss": 0.3957, "step": 3215 }, { "epoch": 2.682981090100111, "grad_norm": 0.30992627143859863, "learning_rate": 3.333173948131663e-07, "loss": 0.4087, "step": 3216 }, { "epoch": 2.6838153503893216, "grad_norm": 0.2910294830799103, "learning_rate": 3.315758736675512e-07, "loss": 0.3503, "step": 3217 }, { "epoch": 2.6846496106785316, "grad_norm": 0.3198425769805908, "learning_rate": 3.2983875796897714e-07, "loss": 0.4422, "step": 3218 }, { "epoch": 2.685483870967742, "grad_norm": 0.27185824513435364, "learning_rate": 3.2810604935670734e-07, "loss": 0.3537, "step": 3219 }, { "epoch": 2.686318131256952, "grad_norm": 0.29563888907432556, "learning_rate": 3.263777494658449e-07, "loss": 0.4061, "step": 3220 }, { "epoch": 2.6871523915461624, "grad_norm": 0.29086634516716003, "learning_rate": 3.24653859927333e-07, "loss": 0.404, "step": 3221 }, { "epoch": 2.6879866518353728, "grad_norm": 0.305075466632843, "learning_rate": 3.2293438236795207e-07, "loss": 0.4375, "step": 3222 }, { "epoch": 2.6888209121245827, "grad_norm": 0.29730024933815, "learning_rate": 3.212193184103196e-07, "loss": 0.4077, "step": 3223 }, { "epoch": 2.689655172413793, "grad_norm": 0.28034645318984985, "learning_rate": 3.1950866967289e-07, "loss": 0.3488, "step": 3224 }, { "epoch": 2.6904894327030036, "grad_norm": 0.29451096057891846, "learning_rate": 3.1780243776994734e-07, "loss": 0.3689, "step": 3225 }, { "epoch": 2.6913236929922135, "grad_norm": 0.3121128976345062, "learning_rate": 3.1610062431161047e-07, "loss": 0.3906, "step": 3226 }, { "epoch": 2.692157953281424, "grad_norm": 0.2870229184627533, "learning_rate": 3.144032309038292e-07, "loss": 0.4129, "step": 3227 }, { "epoch": 2.692992213570634, "grad_norm": 0.2945619523525238, "learning_rate": 3.127102591483783e-07, "loss": 0.4487, "step": 3228 }, { "epoch": 2.6938264738598443, "grad_norm": 0.2702940106391907, "learning_rate": 3.1102171064286534e-07, "loss": 0.3723, "step": 3229 }, { "epoch": 2.6946607341490543, "grad_norm": 0.29362717270851135, "learning_rate": 3.0933758698072023e-07, "loss": 0.3799, "step": 3230 }, { "epoch": 2.6954949944382647, "grad_norm": 0.2827353775501251, "learning_rate": 3.076578897511978e-07, "loss": 0.3768, "step": 3231 }, { "epoch": 2.696329254727475, "grad_norm": 0.2771448791027069, "learning_rate": 3.059826205393773e-07, "loss": 0.3745, "step": 3232 }, { "epoch": 2.697163515016685, "grad_norm": 0.3077465891838074, "learning_rate": 3.04311780926157e-07, "loss": 0.4332, "step": 3233 }, { "epoch": 2.6979977753058955, "grad_norm": 0.28776803612709045, "learning_rate": 3.026453724882572e-07, "loss": 0.3887, "step": 3234 }, { "epoch": 2.698832035595106, "grad_norm": 0.3009152114391327, "learning_rate": 3.0098339679821566e-07, "loss": 0.4256, "step": 3235 }, { "epoch": 2.699666295884316, "grad_norm": 0.29475224018096924, "learning_rate": 2.993258554243877e-07, "loss": 0.4095, "step": 3236 }, { "epoch": 2.7005005561735262, "grad_norm": 0.284462571144104, "learning_rate": 2.9767274993094285e-07, "loss": 0.3702, "step": 3237 }, { "epoch": 2.701334816462736, "grad_norm": 0.2835181653499603, "learning_rate": 2.960240818778659e-07, "loss": 0.4035, "step": 3238 }, { "epoch": 2.7021690767519466, "grad_norm": 0.3016171455383301, "learning_rate": 2.943798528209535e-07, "loss": 0.3784, "step": 3239 }, { "epoch": 2.7030033370411566, "grad_norm": 0.31845951080322266, "learning_rate": 2.92740064311815e-07, "loss": 0.4201, "step": 3240 }, { "epoch": 2.703837597330367, "grad_norm": 0.29852157831192017, "learning_rate": 2.911047178978654e-07, "loss": 0.3823, "step": 3241 }, { "epoch": 2.7046718576195774, "grad_norm": 0.298768550157547, "learning_rate": 2.894738151223331e-07, "loss": 0.3693, "step": 3242 }, { "epoch": 2.7055061179087874, "grad_norm": 0.3083311915397644, "learning_rate": 2.878473575242485e-07, "loss": 0.4157, "step": 3243 }, { "epoch": 2.7063403781979978, "grad_norm": 0.2867836058139801, "learning_rate": 2.862253466384507e-07, "loss": 0.3823, "step": 3244 }, { "epoch": 2.707174638487208, "grad_norm": 0.3051856756210327, "learning_rate": 2.846077839955802e-07, "loss": 0.4089, "step": 3245 }, { "epoch": 2.708008898776418, "grad_norm": 0.2891470491886139, "learning_rate": 2.829946711220799e-07, "loss": 0.3426, "step": 3246 }, { "epoch": 2.7088431590656286, "grad_norm": 0.3278118968009949, "learning_rate": 2.8138600954019546e-07, "loss": 0.448, "step": 3247 }, { "epoch": 2.709677419354839, "grad_norm": 0.3022622764110565, "learning_rate": 2.797818007679715e-07, "loss": 0.3862, "step": 3248 }, { "epoch": 2.710511679644049, "grad_norm": 0.2982810139656067, "learning_rate": 2.781820463192486e-07, "loss": 0.4168, "step": 3249 }, { "epoch": 2.7113459399332593, "grad_norm": 0.29989078640937805, "learning_rate": 2.76586747703666e-07, "loss": 0.4131, "step": 3250 }, { "epoch": 2.7121802002224693, "grad_norm": 0.29547303915023804, "learning_rate": 2.7499590642665773e-07, "loss": 0.4151, "step": 3251 }, { "epoch": 2.7130144605116797, "grad_norm": 0.3001156747341156, "learning_rate": 2.7340952398945077e-07, "loss": 0.3571, "step": 3252 }, { "epoch": 2.7138487208008897, "grad_norm": 0.3134772479534149, "learning_rate": 2.71827601889067e-07, "loss": 0.4127, "step": 3253 }, { "epoch": 2.7146829810901, "grad_norm": 0.28235307335853577, "learning_rate": 2.70250141618314e-07, "loss": 0.3611, "step": 3254 }, { "epoch": 2.7155172413793105, "grad_norm": 0.3128192722797394, "learning_rate": 2.686771446657943e-07, "loss": 0.4666, "step": 3255 }, { "epoch": 2.7163515016685205, "grad_norm": 0.28323742747306824, "learning_rate": 2.671086125158973e-07, "loss": 0.3955, "step": 3256 }, { "epoch": 2.717185761957731, "grad_norm": 0.26181063055992126, "learning_rate": 2.6554454664879437e-07, "loss": 0.3396, "step": 3257 }, { "epoch": 2.7180200222469413, "grad_norm": 0.29119500517845154, "learning_rate": 2.6398494854045055e-07, "loss": 0.451, "step": 3258 }, { "epoch": 2.7188542825361512, "grad_norm": 0.29496321082115173, "learning_rate": 2.6242981966260574e-07, "loss": 0.4465, "step": 3259 }, { "epoch": 2.7196885428253617, "grad_norm": 0.2907100021839142, "learning_rate": 2.608791614827905e-07, "loss": 0.3935, "step": 3260 }, { "epoch": 2.7205228031145716, "grad_norm": 0.2587240934371948, "learning_rate": 2.5933297546431093e-07, "loss": 0.3236, "step": 3261 }, { "epoch": 2.721357063403782, "grad_norm": 0.30368977785110474, "learning_rate": 2.5779126306625493e-07, "loss": 0.4221, "step": 3262 }, { "epoch": 2.722191323692992, "grad_norm": 0.29711058735847473, "learning_rate": 2.5625402574348923e-07, "loss": 0.385, "step": 3263 }, { "epoch": 2.7230255839822024, "grad_norm": 0.2947615087032318, "learning_rate": 2.547212649466568e-07, "loss": 0.4249, "step": 3264 }, { "epoch": 2.723859844271413, "grad_norm": 0.30426573753356934, "learning_rate": 2.531929821221768e-07, "loss": 0.4205, "step": 3265 }, { "epoch": 2.7246941045606228, "grad_norm": 0.28079643845558167, "learning_rate": 2.516691787122416e-07, "loss": 0.3317, "step": 3266 }, { "epoch": 2.725528364849833, "grad_norm": 0.30024316906929016, "learning_rate": 2.5014985615481767e-07, "loss": 0.4375, "step": 3267 }, { "epoch": 2.7263626251390436, "grad_norm": 0.3106209635734558, "learning_rate": 2.4863501588364145e-07, "loss": 0.3923, "step": 3268 }, { "epoch": 2.7271968854282536, "grad_norm": 0.3049510717391968, "learning_rate": 2.471246593282234e-07, "loss": 0.4403, "step": 3269 }, { "epoch": 2.728031145717464, "grad_norm": 0.27871623635292053, "learning_rate": 2.456187879138372e-07, "loss": 0.3798, "step": 3270 }, { "epoch": 2.728865406006674, "grad_norm": 0.28311803936958313, "learning_rate": 2.441174030615295e-07, "loss": 0.3978, "step": 3271 }, { "epoch": 2.7296996662958843, "grad_norm": 0.2920793890953064, "learning_rate": 2.426205061881082e-07, "loss": 0.4003, "step": 3272 }, { "epoch": 2.7305339265850943, "grad_norm": 0.2904806435108185, "learning_rate": 2.4112809870614985e-07, "loss": 0.374, "step": 3273 }, { "epoch": 2.7313681868743047, "grad_norm": 0.2854829728603363, "learning_rate": 2.3964018202399355e-07, "loss": 0.3379, "step": 3274 }, { "epoch": 2.732202447163515, "grad_norm": 0.2889251112937927, "learning_rate": 2.3815675754573885e-07, "loss": 0.3879, "step": 3275 }, { "epoch": 2.733036707452725, "grad_norm": 0.28685539960861206, "learning_rate": 2.3667782667124884e-07, "loss": 0.4305, "step": 3276 }, { "epoch": 2.7338709677419355, "grad_norm": 0.271238774061203, "learning_rate": 2.3520339079614319e-07, "loss": 0.3771, "step": 3277 }, { "epoch": 2.734705228031146, "grad_norm": 0.3013826012611389, "learning_rate": 2.3373345131180224e-07, "loss": 0.4131, "step": 3278 }, { "epoch": 2.735539488320356, "grad_norm": 0.2905944585800171, "learning_rate": 2.32268009605362e-07, "loss": 0.4114, "step": 3279 }, { "epoch": 2.7363737486095663, "grad_norm": 0.2872878313064575, "learning_rate": 2.3080706705971457e-07, "loss": 0.3876, "step": 3280 }, { "epoch": 2.7372080088987767, "grad_norm": 0.3016399145126343, "learning_rate": 2.2935062505350547e-07, "loss": 0.4309, "step": 3281 }, { "epoch": 2.7380422691879867, "grad_norm": 0.2923175096511841, "learning_rate": 2.278986849611353e-07, "loss": 0.3711, "step": 3282 }, { "epoch": 2.7388765294771966, "grad_norm": 0.26824164390563965, "learning_rate": 2.2645124815275254e-07, "loss": 0.3597, "step": 3283 }, { "epoch": 2.739710789766407, "grad_norm": 0.3295307159423828, "learning_rate": 2.2500831599426176e-07, "loss": 0.4965, "step": 3284 }, { "epoch": 2.7405450500556174, "grad_norm": 0.27547112107276917, "learning_rate": 2.2356988984730932e-07, "loss": 0.3228, "step": 3285 }, { "epoch": 2.7413793103448274, "grad_norm": 0.295543909072876, "learning_rate": 2.2213597106929608e-07, "loss": 0.4682, "step": 3286 }, { "epoch": 2.742213570634038, "grad_norm": 0.2679787874221802, "learning_rate": 2.2070656101336686e-07, "loss": 0.3644, "step": 3287 }, { "epoch": 2.743047830923248, "grad_norm": 0.28594884276390076, "learning_rate": 2.192816610284093e-07, "loss": 0.3554, "step": 3288 }, { "epoch": 2.743882091212458, "grad_norm": 0.32329151034355164, "learning_rate": 2.178612724590595e-07, "loss": 0.445, "step": 3289 }, { "epoch": 2.7447163515016686, "grad_norm": 0.30242645740509033, "learning_rate": 2.1644539664569365e-07, "loss": 0.3886, "step": 3290 }, { "epoch": 2.745550611790879, "grad_norm": 0.33302366733551025, "learning_rate": 2.1503403492442898e-07, "loss": 0.42, "step": 3291 }, { "epoch": 2.746384872080089, "grad_norm": 0.27787497639656067, "learning_rate": 2.1362718862712462e-07, "loss": 0.3845, "step": 3292 }, { "epoch": 2.7472191323692994, "grad_norm": 0.2885017991065979, "learning_rate": 2.1222485908137747e-07, "loss": 0.3684, "step": 3293 }, { "epoch": 2.7480533926585093, "grad_norm": 0.2924964725971222, "learning_rate": 2.1082704761052176e-07, "loss": 0.4217, "step": 3294 }, { "epoch": 2.7488876529477198, "grad_norm": 0.30808934569358826, "learning_rate": 2.0943375553363066e-07, "loss": 0.382, "step": 3295 }, { "epoch": 2.7497219132369297, "grad_norm": 0.28866228461265564, "learning_rate": 2.0804498416550855e-07, "loss": 0.4147, "step": 3296 }, { "epoch": 2.75055617352614, "grad_norm": 0.2850414216518402, "learning_rate": 2.0666073481669714e-07, "loss": 0.3996, "step": 3297 }, { "epoch": 2.7513904338153505, "grad_norm": 0.294955313205719, "learning_rate": 2.0528100879346979e-07, "loss": 0.3936, "step": 3298 }, { "epoch": 2.7522246941045605, "grad_norm": 0.30047059059143066, "learning_rate": 2.039058073978295e-07, "loss": 0.3916, "step": 3299 }, { "epoch": 2.753058954393771, "grad_norm": 0.30105480551719666, "learning_rate": 2.0253513192751374e-07, "loss": 0.3924, "step": 3300 }, { "epoch": 2.7538932146829813, "grad_norm": 0.28604641556739807, "learning_rate": 2.0116898367598392e-07, "loss": 0.389, "step": 3301 }, { "epoch": 2.7547274749721913, "grad_norm": 0.2736835777759552, "learning_rate": 1.9980736393243327e-07, "loss": 0.3973, "step": 3302 }, { "epoch": 2.7555617352614017, "grad_norm": 0.30152949690818787, "learning_rate": 1.9845027398177952e-07, "loss": 0.3995, "step": 3303 }, { "epoch": 2.7563959955506117, "grad_norm": 0.30631908774375916, "learning_rate": 1.9709771510466713e-07, "loss": 0.4183, "step": 3304 }, { "epoch": 2.757230255839822, "grad_norm": 0.2899223566055298, "learning_rate": 1.957496885774629e-07, "loss": 0.3895, "step": 3305 }, { "epoch": 2.758064516129032, "grad_norm": 0.2760181128978729, "learning_rate": 1.944061956722587e-07, "loss": 0.3514, "step": 3306 }, { "epoch": 2.7588987764182424, "grad_norm": 0.269015371799469, "learning_rate": 1.9306723765686598e-07, "loss": 0.368, "step": 3307 }, { "epoch": 2.759733036707453, "grad_norm": 0.3026461601257324, "learning_rate": 1.9173281579481896e-07, "loss": 0.4297, "step": 3308 }, { "epoch": 2.760567296996663, "grad_norm": 0.323623389005661, "learning_rate": 1.9040293134536924e-07, "loss": 0.4247, "step": 3309 }, { "epoch": 2.761401557285873, "grad_norm": 0.28874653577804565, "learning_rate": 1.8907758556348853e-07, "loss": 0.3799, "step": 3310 }, { "epoch": 2.7622358175750836, "grad_norm": 0.2974967658519745, "learning_rate": 1.8775677969986362e-07, "loss": 0.3823, "step": 3311 }, { "epoch": 2.7630700778642936, "grad_norm": 0.30077195167541504, "learning_rate": 1.8644051500089744e-07, "loss": 0.4181, "step": 3312 }, { "epoch": 2.763904338153504, "grad_norm": 0.2814527153968811, "learning_rate": 1.8512879270871088e-07, "loss": 0.3793, "step": 3313 }, { "epoch": 2.764738598442714, "grad_norm": 0.32588040828704834, "learning_rate": 1.8382161406113208e-07, "loss": 0.4411, "step": 3314 }, { "epoch": 2.7655728587319244, "grad_norm": 0.3074783384799957, "learning_rate": 1.8251898029170767e-07, "loss": 0.3796, "step": 3315 }, { "epoch": 2.7664071190211343, "grad_norm": 0.28149497509002686, "learning_rate": 1.812208926296921e-07, "loss": 0.3924, "step": 3316 }, { "epoch": 2.7672413793103448, "grad_norm": 0.2700985074043274, "learning_rate": 1.7992735230004887e-07, "loss": 0.3697, "step": 3317 }, { "epoch": 2.768075639599555, "grad_norm": 0.3082144856452942, "learning_rate": 1.786383605234543e-07, "loss": 0.4352, "step": 3318 }, { "epoch": 2.768909899888765, "grad_norm": 0.31781986355781555, "learning_rate": 1.7735391851628814e-07, "loss": 0.4051, "step": 3319 }, { "epoch": 2.7697441601779755, "grad_norm": 0.2813461422920227, "learning_rate": 1.760740274906392e-07, "loss": 0.391, "step": 3320 }, { "epoch": 2.770578420467186, "grad_norm": 0.29099974036216736, "learning_rate": 1.7479868865430072e-07, "loss": 0.4359, "step": 3321 }, { "epoch": 2.771412680756396, "grad_norm": 0.29641345143318176, "learning_rate": 1.7352790321077063e-07, "loss": 0.383, "step": 3322 }, { "epoch": 2.7722469410456063, "grad_norm": 0.3072645962238312, "learning_rate": 1.7226167235924908e-07, "loss": 0.3652, "step": 3323 }, { "epoch": 2.7730812013348167, "grad_norm": 0.3004201352596283, "learning_rate": 1.7099999729464022e-07, "loss": 0.4318, "step": 3324 }, { "epoch": 2.7739154616240267, "grad_norm": 0.3168380558490753, "learning_rate": 1.6974287920754562e-07, "loss": 0.4208, "step": 3325 }, { "epoch": 2.7747497219132367, "grad_norm": 0.2729546129703522, "learning_rate": 1.6849031928427129e-07, "loss": 0.3946, "step": 3326 }, { "epoch": 2.775583982202447, "grad_norm": 0.2911393642425537, "learning_rate": 1.6724231870681673e-07, "loss": 0.3978, "step": 3327 }, { "epoch": 2.7764182424916575, "grad_norm": 0.27647945284843445, "learning_rate": 1.659988786528821e-07, "loss": 0.3799, "step": 3328 }, { "epoch": 2.7772525027808674, "grad_norm": 0.27860724925994873, "learning_rate": 1.6476000029586437e-07, "loss": 0.3621, "step": 3329 }, { "epoch": 2.778086763070078, "grad_norm": 0.30558112263679504, "learning_rate": 1.6352568480485277e-07, "loss": 0.4048, "step": 3330 }, { "epoch": 2.7789210233592883, "grad_norm": 0.28147411346435547, "learning_rate": 1.622959333446339e-07, "loss": 0.3935, "step": 3331 }, { "epoch": 2.779755283648498, "grad_norm": 0.3183319568634033, "learning_rate": 1.6107074707568616e-07, "loss": 0.4275, "step": 3332 }, { "epoch": 2.7805895439377086, "grad_norm": 0.27750369906425476, "learning_rate": 1.5985012715417858e-07, "loss": 0.3675, "step": 3333 }, { "epoch": 2.781423804226919, "grad_norm": 0.2739849388599396, "learning_rate": 1.5863407473197368e-07, "loss": 0.3861, "step": 3334 }, { "epoch": 2.782258064516129, "grad_norm": 0.2725929915904999, "learning_rate": 1.5742259095662126e-07, "loss": 0.3851, "step": 3335 }, { "epoch": 2.7830923248053394, "grad_norm": 0.2814646363258362, "learning_rate": 1.562156769713613e-07, "loss": 0.3731, "step": 3336 }, { "epoch": 2.7839265850945494, "grad_norm": 0.298306941986084, "learning_rate": 1.550133339151222e-07, "loss": 0.3808, "step": 3337 }, { "epoch": 2.78476084538376, "grad_norm": 0.28735390305519104, "learning_rate": 1.5381556292251632e-07, "loss": 0.397, "step": 3338 }, { "epoch": 2.7855951056729698, "grad_norm": 0.2901236414909363, "learning_rate": 1.5262236512384397e-07, "loss": 0.3917, "step": 3339 }, { "epoch": 2.78642936596218, "grad_norm": 0.29154109954833984, "learning_rate": 1.5143374164508784e-07, "loss": 0.3618, "step": 3340 }, { "epoch": 2.7872636262513906, "grad_norm": 0.2870076298713684, "learning_rate": 1.5024969360791564e-07, "loss": 0.3874, "step": 3341 }, { "epoch": 2.7880978865406005, "grad_norm": 0.285459965467453, "learning_rate": 1.4907022212967803e-07, "loss": 0.3639, "step": 3342 }, { "epoch": 2.788932146829811, "grad_norm": 0.30360954999923706, "learning_rate": 1.4789532832340304e-07, "loss": 0.3959, "step": 3343 }, { "epoch": 2.7897664071190214, "grad_norm": 0.26461079716682434, "learning_rate": 1.4672501329780374e-07, "loss": 0.3728, "step": 3344 }, { "epoch": 2.7906006674082313, "grad_norm": 0.2924419939517975, "learning_rate": 1.4555927815726955e-07, "loss": 0.4167, "step": 3345 }, { "epoch": 2.7914349276974417, "grad_norm": 0.28610262274742126, "learning_rate": 1.443981240018677e-07, "loss": 0.3706, "step": 3346 }, { "epoch": 2.7922691879866517, "grad_norm": 0.3064219653606415, "learning_rate": 1.4324155192734445e-07, "loss": 0.4262, "step": 3347 }, { "epoch": 2.793103448275862, "grad_norm": 0.29446282982826233, "learning_rate": 1.420895630251201e-07, "loss": 0.3618, "step": 3348 }, { "epoch": 2.793937708565072, "grad_norm": 0.29025495052337646, "learning_rate": 1.4094215838229176e-07, "loss": 0.4134, "step": 3349 }, { "epoch": 2.7947719688542825, "grad_norm": 0.27580779790878296, "learning_rate": 1.3979933908162878e-07, "loss": 0.3813, "step": 3350 }, { "epoch": 2.795606229143493, "grad_norm": 0.28983402252197266, "learning_rate": 1.386611062015747e-07, "loss": 0.4131, "step": 3351 }, { "epoch": 2.796440489432703, "grad_norm": 0.27870965003967285, "learning_rate": 1.375274608162447e-07, "loss": 0.4283, "step": 3352 }, { "epoch": 2.7972747497219133, "grad_norm": 0.295725554227829, "learning_rate": 1.3639840399542415e-07, "loss": 0.3939, "step": 3353 }, { "epoch": 2.7981090100111237, "grad_norm": 0.2669251263141632, "learning_rate": 1.3527393680456912e-07, "loss": 0.3921, "step": 3354 }, { "epoch": 2.7989432703003336, "grad_norm": 0.28645962476730347, "learning_rate": 1.3415406030480582e-07, "loss": 0.3843, "step": 3355 }, { "epoch": 2.799777530589544, "grad_norm": 0.29786163568496704, "learning_rate": 1.3303877555292443e-07, "loss": 0.4027, "step": 3356 }, { "epoch": 2.8006117908787544, "grad_norm": 0.2937493920326233, "learning_rate": 1.3192808360138699e-07, "loss": 0.3801, "step": 3357 }, { "epoch": 2.8014460511679644, "grad_norm": 0.29516565799713135, "learning_rate": 1.3082198549831838e-07, "loss": 0.4039, "step": 3358 }, { "epoch": 2.8022803114571744, "grad_norm": 0.2825239300727844, "learning_rate": 1.2972048228750755e-07, "loss": 0.3993, "step": 3359 }, { "epoch": 2.803114571746385, "grad_norm": 0.2994902729988098, "learning_rate": 1.2862357500841026e-07, "loss": 0.3742, "step": 3360 }, { "epoch": 2.803948832035595, "grad_norm": 0.2727040946483612, "learning_rate": 1.2753126469614407e-07, "loss": 0.3593, "step": 3361 }, { "epoch": 2.804783092324805, "grad_norm": 0.2898688316345215, "learning_rate": 1.2644355238148776e-07, "loss": 0.4387, "step": 3362 }, { "epoch": 2.8056173526140156, "grad_norm": 0.2956966757774353, "learning_rate": 1.253604390908819e-07, "loss": 0.3881, "step": 3363 }, { "epoch": 2.806451612903226, "grad_norm": 0.2990025281906128, "learning_rate": 1.2428192584642729e-07, "loss": 0.3838, "step": 3364 }, { "epoch": 2.807285873192436, "grad_norm": 0.31267911195755005, "learning_rate": 1.2320801366588253e-07, "loss": 0.4151, "step": 3365 }, { "epoch": 2.8081201334816464, "grad_norm": 0.2939951717853546, "learning_rate": 1.2213870356266645e-07, "loss": 0.4178, "step": 3366 }, { "epoch": 2.8089543937708568, "grad_norm": 0.27928388118743896, "learning_rate": 1.2107399654585183e-07, "loss": 0.3918, "step": 3367 }, { "epoch": 2.8097886540600667, "grad_norm": 0.3060562014579773, "learning_rate": 1.2001389362017223e-07, "loss": 0.4136, "step": 3368 }, { "epoch": 2.810622914349277, "grad_norm": 0.28037598729133606, "learning_rate": 1.1895839578601243e-07, "loss": 0.3759, "step": 3369 }, { "epoch": 2.811457174638487, "grad_norm": 0.28996023535728455, "learning_rate": 1.1790750403941231e-07, "loss": 0.4205, "step": 3370 }, { "epoch": 2.8122914349276975, "grad_norm": 0.287614107131958, "learning_rate": 1.1686121937206807e-07, "loss": 0.3788, "step": 3371 }, { "epoch": 2.8131256952169075, "grad_norm": 0.28932034969329834, "learning_rate": 1.1581954277132323e-07, "loss": 0.3782, "step": 3372 }, { "epoch": 2.813959955506118, "grad_norm": 0.30571091175079346, "learning_rate": 1.1478247522017816e-07, "loss": 0.4041, "step": 3373 }, { "epoch": 2.8147942157953283, "grad_norm": 0.2852344512939453, "learning_rate": 1.1375001769728e-07, "loss": 0.3877, "step": 3374 }, { "epoch": 2.8156284760845383, "grad_norm": 0.2832501232624054, "learning_rate": 1.1272217117692719e-07, "loss": 0.3896, "step": 3375 }, { "epoch": 2.8164627363737487, "grad_norm": 0.29212209582328796, "learning_rate": 1.1169893662906661e-07, "loss": 0.4286, "step": 3376 }, { "epoch": 2.817296996662959, "grad_norm": 0.2913340926170349, "learning_rate": 1.1068031501929366e-07, "loss": 0.4359, "step": 3377 }, { "epoch": 2.818131256952169, "grad_norm": 0.2783985435962677, "learning_rate": 1.0966630730884886e-07, "loss": 0.3417, "step": 3378 }, { "epoch": 2.8189655172413794, "grad_norm": 0.33768701553344727, "learning_rate": 1.0865691445462012e-07, "loss": 0.4537, "step": 3379 }, { "epoch": 2.8197997775305894, "grad_norm": 0.29121631383895874, "learning_rate": 1.076521374091405e-07, "loss": 0.3927, "step": 3380 }, { "epoch": 2.8206340378198, "grad_norm": 0.2912810444831848, "learning_rate": 1.0665197712058706e-07, "loss": 0.407, "step": 3381 }, { "epoch": 2.82146829810901, "grad_norm": 0.30418816208839417, "learning_rate": 1.0565643453277929e-07, "loss": 0.4143, "step": 3382 }, { "epoch": 2.82230255839822, "grad_norm": 0.3070503771305084, "learning_rate": 1.0466551058518015e-07, "loss": 0.3637, "step": 3383 }, { "epoch": 2.8231368186874306, "grad_norm": 0.2947136461734772, "learning_rate": 1.0367920621289496e-07, "loss": 0.3869, "step": 3384 }, { "epoch": 2.8239710789766406, "grad_norm": 0.2551459074020386, "learning_rate": 1.0269752234666642e-07, "loss": 0.3621, "step": 3385 }, { "epoch": 2.824805339265851, "grad_norm": 0.2721526324748993, "learning_rate": 1.017204599128807e-07, "loss": 0.4142, "step": 3386 }, { "epoch": 2.8256395995550614, "grad_norm": 0.28498178720474243, "learning_rate": 1.0074801983356086e-07, "loss": 0.3412, "step": 3387 }, { "epoch": 2.8264738598442714, "grad_norm": 0.30487504601478577, "learning_rate": 9.978020302636892e-08, "loss": 0.3965, "step": 3388 }, { "epoch": 2.8273081201334818, "grad_norm": 0.2758004665374756, "learning_rate": 9.881701040460379e-08, "loss": 0.3316, "step": 3389 }, { "epoch": 2.8281423804226917, "grad_norm": 0.3075368106365204, "learning_rate": 9.785844287719836e-08, "loss": 0.4385, "step": 3390 }, { "epoch": 2.828976640711902, "grad_norm": 0.29425230622291565, "learning_rate": 9.69045013487252e-08, "loss": 0.3946, "step": 3391 }, { "epoch": 2.829810901001112, "grad_norm": 0.30456775426864624, "learning_rate": 9.595518671938864e-08, "loss": 0.4131, "step": 3392 }, { "epoch": 2.8306451612903225, "grad_norm": 0.27334219217300415, "learning_rate": 9.501049988502764e-08, "loss": 0.342, "step": 3393 }, { "epoch": 2.831479421579533, "grad_norm": 0.28168052434921265, "learning_rate": 9.407044173711411e-08, "loss": 0.4107, "step": 3394 }, { "epoch": 2.832313681868743, "grad_norm": 0.2953903079032898, "learning_rate": 9.313501316275064e-08, "loss": 0.3714, "step": 3395 }, { "epoch": 2.8331479421579533, "grad_norm": 0.29103606939315796, "learning_rate": 9.22042150446728e-08, "loss": 0.4433, "step": 3396 }, { "epoch": 2.8339822024471637, "grad_norm": 0.2781805694103241, "learning_rate": 9.127804826124741e-08, "loss": 0.4117, "step": 3397 }, { "epoch": 2.8348164627363737, "grad_norm": 0.29511621594429016, "learning_rate": 9.035651368646647e-08, "loss": 0.3835, "step": 3398 }, { "epoch": 2.835650723025584, "grad_norm": 0.3077721893787384, "learning_rate": 8.943961218995544e-08, "loss": 0.4108, "step": 3399 }, { "epoch": 2.8364849833147945, "grad_norm": 0.2796066999435425, "learning_rate": 8.85273446369661e-08, "loss": 0.3748, "step": 3400 }, { "epoch": 2.8373192436040044, "grad_norm": 0.29605627059936523, "learning_rate": 8.761971188837537e-08, "loss": 0.3989, "step": 3401 }, { "epoch": 2.8381535038932144, "grad_norm": 0.2980910539627075, "learning_rate": 8.671671480069032e-08, "loss": 0.3572, "step": 3402 }, { "epoch": 2.838987764182425, "grad_norm": 0.28678300976753235, "learning_rate": 8.58183542260388e-08, "loss": 0.3768, "step": 3403 }, { "epoch": 2.8398220244716352, "grad_norm": 0.28513410687446594, "learning_rate": 8.492463101217763e-08, "loss": 0.4079, "step": 3404 }, { "epoch": 2.840656284760845, "grad_norm": 0.28315797448158264, "learning_rate": 8.403554600248498e-08, "loss": 0.3818, "step": 3405 }, { "epoch": 2.8414905450500556, "grad_norm": 0.3316526710987091, "learning_rate": 8.315110003596361e-08, "loss": 0.5016, "step": 3406 }, { "epoch": 2.842324805339266, "grad_norm": 0.2944430708885193, "learning_rate": 8.227129394723643e-08, "loss": 0.3619, "step": 3407 }, { "epoch": 2.843159065628476, "grad_norm": 0.2883685529232025, "learning_rate": 8.139612856655044e-08, "loss": 0.3993, "step": 3408 }, { "epoch": 2.8439933259176864, "grad_norm": 0.26984256505966187, "learning_rate": 8.052560471977111e-08, "loss": 0.3797, "step": 3409 }, { "epoch": 2.844827586206897, "grad_norm": 0.30006319284439087, "learning_rate": 7.965972322838522e-08, "loss": 0.418, "step": 3410 }, { "epoch": 2.8456618464961068, "grad_norm": 0.27099236845970154, "learning_rate": 7.879848490949804e-08, "loss": 0.3291, "step": 3411 }, { "epoch": 2.846496106785317, "grad_norm": 0.3038952946662903, "learning_rate": 7.794189057583335e-08, "loss": 0.4145, "step": 3412 }, { "epoch": 2.847330367074527, "grad_norm": 0.2966451644897461, "learning_rate": 7.708994103573397e-08, "loss": 0.3979, "step": 3413 }, { "epoch": 2.8481646273637375, "grad_norm": 0.2902075946331024, "learning_rate": 7.624263709315627e-08, "loss": 0.3814, "step": 3414 }, { "epoch": 2.8489988876529475, "grad_norm": 0.27492278814315796, "learning_rate": 7.53999795476762e-08, "loss": 0.3889, "step": 3415 }, { "epoch": 2.849833147942158, "grad_norm": 0.2608383893966675, "learning_rate": 7.456196919448211e-08, "loss": 0.3541, "step": 3416 }, { "epoch": 2.8506674082313683, "grad_norm": 0.3158596456050873, "learning_rate": 7.372860682437976e-08, "loss": 0.4288, "step": 3417 }, { "epoch": 2.8515016685205783, "grad_norm": 0.30074021220207214, "learning_rate": 7.289989322378732e-08, "loss": 0.4153, "step": 3418 }, { "epoch": 2.8523359288097887, "grad_norm": 0.2634953260421753, "learning_rate": 7.207582917473532e-08, "loss": 0.3624, "step": 3419 }, { "epoch": 2.853170189098999, "grad_norm": 0.2893223166465759, "learning_rate": 7.125641545486783e-08, "loss": 0.4299, "step": 3420 }, { "epoch": 2.854004449388209, "grad_norm": 0.276814341545105, "learning_rate": 7.044165283743965e-08, "loss": 0.393, "step": 3421 }, { "epoch": 2.8548387096774195, "grad_norm": 0.27974867820739746, "learning_rate": 6.963154209131795e-08, "loss": 0.4154, "step": 3422 }, { "epoch": 2.8556729699666294, "grad_norm": 0.299127459526062, "learning_rate": 6.882608398097846e-08, "loss": 0.4016, "step": 3423 }, { "epoch": 2.85650723025584, "grad_norm": 0.2782146632671356, "learning_rate": 6.80252792665076e-08, "loss": 0.3568, "step": 3424 }, { "epoch": 2.85734149054505, "grad_norm": 0.2920554280281067, "learning_rate": 6.722912870359865e-08, "loss": 0.4168, "step": 3425 }, { "epoch": 2.8581757508342602, "grad_norm": 0.30458396673202515, "learning_rate": 6.643763304355566e-08, "loss": 0.383, "step": 3426 }, { "epoch": 2.8590100111234706, "grad_norm": 0.29921793937683105, "learning_rate": 6.565079303328726e-08, "loss": 0.3942, "step": 3427 }, { "epoch": 2.8598442714126806, "grad_norm": 0.2839667499065399, "learning_rate": 6.48686094153117e-08, "loss": 0.3777, "step": 3428 }, { "epoch": 2.860678531701891, "grad_norm": 0.270255446434021, "learning_rate": 6.409108292774912e-08, "loss": 0.3573, "step": 3429 }, { "epoch": 2.8615127919911014, "grad_norm": 0.29771688580513, "learning_rate": 6.331821430432927e-08, "loss": 0.4077, "step": 3430 }, { "epoch": 2.8623470522803114, "grad_norm": 0.28371545672416687, "learning_rate": 6.255000427438318e-08, "loss": 0.3873, "step": 3431 }, { "epoch": 2.863181312569522, "grad_norm": 0.25725722312927246, "learning_rate": 6.178645356284652e-08, "loss": 0.3536, "step": 3432 }, { "epoch": 2.8640155728587318, "grad_norm": 0.29249241948127747, "learning_rate": 6.102756289025957e-08, "loss": 0.4233, "step": 3433 }, { "epoch": 2.864849833147942, "grad_norm": 0.2662562429904938, "learning_rate": 6.027333297276284e-08, "loss": 0.3932, "step": 3434 }, { "epoch": 2.865684093437152, "grad_norm": 0.28642210364341736, "learning_rate": 5.952376452210029e-08, "loss": 0.3624, "step": 3435 }, { "epoch": 2.8665183537263625, "grad_norm": 0.3193361759185791, "learning_rate": 5.8778858245616686e-08, "loss": 0.4262, "step": 3436 }, { "epoch": 2.867352614015573, "grad_norm": 0.3014802038669586, "learning_rate": 5.803861484625584e-08, "loss": 0.4185, "step": 3437 }, { "epoch": 2.868186874304783, "grad_norm": 0.3045835494995117, "learning_rate": 5.7303035022563404e-08, "loss": 0.3637, "step": 3438 }, { "epoch": 2.8690211345939933, "grad_norm": 0.32484540343284607, "learning_rate": 5.657211946868357e-08, "loss": 0.4095, "step": 3439 }, { "epoch": 2.8698553948832037, "grad_norm": 0.2702358365058899, "learning_rate": 5.584586887435739e-08, "loss": 0.353, "step": 3440 }, { "epoch": 2.8706896551724137, "grad_norm": 0.2933849096298218, "learning_rate": 5.512428392492608e-08, "loss": 0.3766, "step": 3441 }, { "epoch": 2.871523915461624, "grad_norm": 0.2865277826786041, "learning_rate": 5.4407365301326623e-08, "loss": 0.3857, "step": 3442 }, { "epoch": 2.8723581757508345, "grad_norm": 0.30091768503189087, "learning_rate": 5.369511368009284e-08, "loss": 0.4254, "step": 3443 }, { "epoch": 2.8731924360400445, "grad_norm": 0.2889598309993744, "learning_rate": 5.298752973335486e-08, "loss": 0.3607, "step": 3444 }, { "epoch": 2.874026696329255, "grad_norm": 0.2799936532974243, "learning_rate": 5.2284614128837454e-08, "loss": 0.3802, "step": 3445 }, { "epoch": 2.874860956618465, "grad_norm": 0.2854139506816864, "learning_rate": 5.1586367529860015e-08, "loss": 0.39, "step": 3446 }, { "epoch": 2.8756952169076753, "grad_norm": 0.2972990572452545, "learning_rate": 5.089279059533658e-08, "loss": 0.4151, "step": 3447 }, { "epoch": 2.8765294771968852, "grad_norm": 0.27479860186576843, "learning_rate": 5.02038839797736e-08, "loss": 0.3863, "step": 3448 }, { "epoch": 2.8773637374860956, "grad_norm": 0.3089061379432678, "learning_rate": 4.9519648333272164e-08, "loss": 0.4142, "step": 3449 }, { "epoch": 2.878197997775306, "grad_norm": 0.280931293964386, "learning_rate": 4.884008430152243e-08, "loss": 0.3841, "step": 3450 }, { "epoch": 2.879032258064516, "grad_norm": 0.30126655101776123, "learning_rate": 4.8165192525809754e-08, "loss": 0.4241, "step": 3451 }, { "epoch": 2.8798665183537264, "grad_norm": 0.29955974221229553, "learning_rate": 4.749497364300693e-08, "loss": 0.3947, "step": 3452 }, { "epoch": 2.880700778642937, "grad_norm": 0.29691264033317566, "learning_rate": 4.6829428285580235e-08, "loss": 0.4261, "step": 3453 }, { "epoch": 2.881535038932147, "grad_norm": 0.2838590145111084, "learning_rate": 4.6168557081582854e-08, "loss": 0.3855, "step": 3454 }, { "epoch": 2.882369299221357, "grad_norm": 0.2750702500343323, "learning_rate": 4.551236065465925e-08, "loss": 0.3954, "step": 3455 }, { "epoch": 2.883203559510567, "grad_norm": 0.27548539638519287, "learning_rate": 4.4860839624040756e-08, "loss": 0.4339, "step": 3456 }, { "epoch": 2.8840378197997776, "grad_norm": 0.28646010160446167, "learning_rate": 4.4213994604548894e-08, "loss": 0.3855, "step": 3457 }, { "epoch": 2.8848720800889875, "grad_norm": 0.2832009792327881, "learning_rate": 4.3571826206590395e-08, "loss": 0.3689, "step": 3458 }, { "epoch": 2.885706340378198, "grad_norm": 0.2908823788166046, "learning_rate": 4.29343350361594e-08, "loss": 0.4005, "step": 3459 }, { "epoch": 2.8865406006674084, "grad_norm": 0.29409629106521606, "learning_rate": 4.230152169483748e-08, "loss": 0.3978, "step": 3460 }, { "epoch": 2.8873748609566183, "grad_norm": 0.2751644253730774, "learning_rate": 4.167338677979027e-08, "loss": 0.3694, "step": 3461 }, { "epoch": 2.8882091212458287, "grad_norm": 0.2943800091743469, "learning_rate": 4.104993088376974e-08, "loss": 0.3498, "step": 3462 }, { "epoch": 2.889043381535039, "grad_norm": 0.3348623514175415, "learning_rate": 4.043115459511082e-08, "loss": 0.4026, "step": 3463 }, { "epoch": 2.889877641824249, "grad_norm": 0.28447192907333374, "learning_rate": 3.981705849773476e-08, "loss": 0.3648, "step": 3464 }, { "epoch": 2.8907119021134595, "grad_norm": 0.2703888714313507, "learning_rate": 3.920764317114412e-08, "loss": 0.3773, "step": 3465 }, { "epoch": 2.8915461624026695, "grad_norm": 0.2734789550304413, "learning_rate": 3.860290919042553e-08, "loss": 0.3595, "step": 3466 }, { "epoch": 2.89238042269188, "grad_norm": 0.2926599979400635, "learning_rate": 3.8002857126248074e-08, "loss": 0.4215, "step": 3467 }, { "epoch": 2.89321468298109, "grad_norm": 0.28259268403053284, "learning_rate": 3.7407487544861565e-08, "loss": 0.3763, "step": 3468 }, { "epoch": 2.8940489432703003, "grad_norm": 0.3167880177497864, "learning_rate": 3.6816801008097705e-08, "loss": 0.4246, "step": 3469 }, { "epoch": 2.8948832035595107, "grad_norm": 0.2713686525821686, "learning_rate": 3.62307980733706e-08, "loss": 0.3637, "step": 3470 }, { "epoch": 2.8957174638487206, "grad_norm": 0.27738142013549805, "learning_rate": 3.564947929367124e-08, "loss": 0.3915, "step": 3471 }, { "epoch": 2.896551724137931, "grad_norm": 0.30831918120384216, "learning_rate": 3.507284521757248e-08, "loss": 0.461, "step": 3472 }, { "epoch": 2.8973859844271415, "grad_norm": 0.2650093734264374, "learning_rate": 3.450089638922738e-08, "loss": 0.3471, "step": 3473 }, { "epoch": 2.8982202447163514, "grad_norm": 0.2824805676937103, "learning_rate": 3.393363334836475e-08, "loss": 0.3779, "step": 3474 }, { "epoch": 2.899054505005562, "grad_norm": 0.2820485234260559, "learning_rate": 3.337105663029361e-08, "loss": 0.3845, "step": 3475 }, { "epoch": 2.8998887652947722, "grad_norm": 0.2836378514766693, "learning_rate": 3.2813166765900403e-08, "loss": 0.384, "step": 3476 }, { "epoch": 2.900723025583982, "grad_norm": 0.3033655881881714, "learning_rate": 3.2259964281647905e-08, "loss": 0.4021, "step": 3477 }, { "epoch": 2.901557285873192, "grad_norm": 0.28571435809135437, "learning_rate": 3.1711449699576845e-08, "loss": 0.3973, "step": 3478 }, { "epoch": 2.9023915461624026, "grad_norm": 0.2829986810684204, "learning_rate": 3.116762353730318e-08, "loss": 0.3484, "step": 3479 }, { "epoch": 2.903225806451613, "grad_norm": 0.30450761318206787, "learning_rate": 3.0628486308018625e-08, "loss": 0.4043, "step": 3480 }, { "epoch": 2.904060066740823, "grad_norm": 0.2794939875602722, "learning_rate": 3.0094038520490646e-08, "loss": 0.3951, "step": 3481 }, { "epoch": 2.9048943270300334, "grad_norm": 0.2747998535633087, "learning_rate": 2.9564280679060255e-08, "loss": 0.361, "step": 3482 }, { "epoch": 2.9057285873192438, "grad_norm": 0.2993730306625366, "learning_rate": 2.903921328364423e-08, "loss": 0.403, "step": 3483 }, { "epoch": 2.9065628476084537, "grad_norm": 0.27738481760025024, "learning_rate": 2.8518836829732332e-08, "loss": 0.3638, "step": 3484 }, { "epoch": 2.907397107897664, "grad_norm": 0.2957957684993744, "learning_rate": 2.8003151808386753e-08, "loss": 0.3946, "step": 3485 }, { "epoch": 2.9082313681868746, "grad_norm": 0.28091463446617126, "learning_rate": 2.7492158706245443e-08, "loss": 0.3728, "step": 3486 }, { "epoch": 2.9090656284760845, "grad_norm": 0.29285937547683716, "learning_rate": 2.6985858005514342e-08, "loss": 0.3908, "step": 3487 }, { "epoch": 2.909899888765295, "grad_norm": 0.29561057686805725, "learning_rate": 2.6484250183975713e-08, "loss": 0.3859, "step": 3488 }, { "epoch": 2.910734149054505, "grad_norm": 0.28758639097213745, "learning_rate": 2.59873357149798e-08, "loss": 0.3942, "step": 3489 }, { "epoch": 2.9115684093437153, "grad_norm": 0.2638346552848816, "learning_rate": 2.5495115067450393e-08, "loss": 0.3438, "step": 3490 }, { "epoch": 2.9124026696329253, "grad_norm": 0.2851925790309906, "learning_rate": 2.500758870588038e-08, "loss": 0.4266, "step": 3491 }, { "epoch": 2.9132369299221357, "grad_norm": 0.2654818594455719, "learning_rate": 2.452475709033342e-08, "loss": 0.367, "step": 3492 }, { "epoch": 2.914071190211346, "grad_norm": 0.29980209469795227, "learning_rate": 2.4046620676442813e-08, "loss": 0.4313, "step": 3493 }, { "epoch": 2.914905450500556, "grad_norm": 0.3052956163883209, "learning_rate": 2.3573179915410972e-08, "loss": 0.3798, "step": 3494 }, { "epoch": 2.9157397107897665, "grad_norm": 0.30567771196365356, "learning_rate": 2.3104435254008852e-08, "loss": 0.3936, "step": 3495 }, { "epoch": 2.916573971078977, "grad_norm": 0.298231303691864, "learning_rate": 2.264038713457706e-08, "loss": 0.3947, "step": 3496 }, { "epoch": 2.917408231368187, "grad_norm": 0.28140056133270264, "learning_rate": 2.2181035995022527e-08, "loss": 0.3843, "step": 3497 }, { "epoch": 2.9182424916573972, "grad_norm": 0.29914212226867676, "learning_rate": 2.172638226882129e-08, "loss": 0.4304, "step": 3498 }, { "epoch": 2.919076751946607, "grad_norm": 0.2796788811683655, "learning_rate": 2.1276426385016257e-08, "loss": 0.3749, "step": 3499 }, { "epoch": 2.9199110122358176, "grad_norm": 0.3062983751296997, "learning_rate": 2.083116876821556e-08, "loss": 0.3995, "step": 3500 }, { "epoch": 2.9207452725250276, "grad_norm": 0.2826257646083832, "learning_rate": 2.039060983859642e-08, "loss": 0.3665, "step": 3501 }, { "epoch": 2.921579532814238, "grad_norm": 0.28888633847236633, "learning_rate": 1.995475001190017e-08, "loss": 0.406, "step": 3502 }, { "epoch": 2.9224137931034484, "grad_norm": 0.29964780807495117, "learning_rate": 1.9523589699433355e-08, "loss": 0.407, "step": 3503 }, { "epoch": 2.9232480533926584, "grad_norm": 0.2694574296474457, "learning_rate": 1.909712930806995e-08, "loss": 0.4073, "step": 3504 }, { "epoch": 2.9240823136818688, "grad_norm": 0.2828221917152405, "learning_rate": 1.8675369240246375e-08, "loss": 0.4442, "step": 3505 }, { "epoch": 2.924916573971079, "grad_norm": 0.2727067470550537, "learning_rate": 1.8258309893965375e-08, "loss": 0.41, "step": 3506 }, { "epoch": 2.925750834260289, "grad_norm": 0.2901587188243866, "learning_rate": 1.7845951662792128e-08, "loss": 0.3877, "step": 3507 }, { "epoch": 2.9265850945494996, "grad_norm": 0.29301509261131287, "learning_rate": 1.743829493585647e-08, "loss": 0.3801, "step": 3508 }, { "epoch": 2.9274193548387095, "grad_norm": 0.2955329120159149, "learning_rate": 1.7035340097851793e-08, "loss": 0.3788, "step": 3509 }, { "epoch": 2.92825361512792, "grad_norm": 0.28470882773399353, "learning_rate": 1.6637087529033925e-08, "loss": 0.3981, "step": 3510 }, { "epoch": 2.92908787541713, "grad_norm": 0.2758449912071228, "learning_rate": 1.6243537605221126e-08, "loss": 0.4005, "step": 3511 }, { "epoch": 2.9299221357063403, "grad_norm": 0.30064964294433594, "learning_rate": 1.585469069779466e-08, "loss": 0.4361, "step": 3512 }, { "epoch": 2.9307563959955507, "grad_norm": 0.28641098737716675, "learning_rate": 1.5470547173697114e-08, "loss": 0.4041, "step": 3513 }, { "epoch": 2.9315906562847607, "grad_norm": 0.2883082926273346, "learning_rate": 1.5091107395432958e-08, "loss": 0.3841, "step": 3514 }, { "epoch": 2.932424916573971, "grad_norm": 0.29947951436042786, "learning_rate": 1.4716371721067435e-08, "loss": 0.3744, "step": 3515 }, { "epoch": 2.9332591768631815, "grad_norm": 0.28807732462882996, "learning_rate": 1.4346340504227119e-08, "loss": 0.3963, "step": 3516 }, { "epoch": 2.9340934371523915, "grad_norm": 0.2691168785095215, "learning_rate": 1.3981014094099354e-08, "loss": 0.3797, "step": 3517 }, { "epoch": 2.934927697441602, "grad_norm": 0.2986096739768982, "learning_rate": 1.3620392835430596e-08, "loss": 0.4355, "step": 3518 }, { "epoch": 2.9357619577308123, "grad_norm": 0.2952001690864563, "learning_rate": 1.3264477068528625e-08, "loss": 0.385, "step": 3519 }, { "epoch": 2.9365962180200222, "grad_norm": 0.28507381677627563, "learning_rate": 1.2913267129260332e-08, "loss": 0.3817, "step": 3520 }, { "epoch": 2.937430478309232, "grad_norm": 0.28471943736076355, "learning_rate": 1.2566763349050604e-08, "loss": 0.3715, "step": 3521 }, { "epoch": 2.9382647385984426, "grad_norm": 0.2992861270904541, "learning_rate": 1.22249660548851e-08, "loss": 0.4274, "step": 3522 }, { "epoch": 2.939098998887653, "grad_norm": 0.29434365034103394, "learning_rate": 1.1887875569306928e-08, "loss": 0.4129, "step": 3523 }, { "epoch": 2.939933259176863, "grad_norm": 0.2795424461364746, "learning_rate": 1.1555492210418295e-08, "loss": 0.384, "step": 3524 }, { "epoch": 2.9407675194660734, "grad_norm": 0.2859519422054291, "learning_rate": 1.1227816291878857e-08, "loss": 0.3669, "step": 3525 }, { "epoch": 2.941601779755284, "grad_norm": 0.29043805599212646, "learning_rate": 1.0904848122906264e-08, "loss": 0.3964, "step": 3526 }, { "epoch": 2.9424360400444938, "grad_norm": 0.3057800233364105, "learning_rate": 1.0586588008275055e-08, "loss": 0.3659, "step": 3527 }, { "epoch": 2.943270300333704, "grad_norm": 0.30072230100631714, "learning_rate": 1.0273036248318325e-08, "loss": 0.392, "step": 3528 }, { "epoch": 2.9441045606229146, "grad_norm": 0.2981456518173218, "learning_rate": 9.964193138924383e-09, "loss": 0.4115, "step": 3529 }, { "epoch": 2.9449388209121246, "grad_norm": 0.3006112277507782, "learning_rate": 9.660058971540099e-09, "loss": 0.434, "step": 3530 }, { "epoch": 2.945773081201335, "grad_norm": 0.26545366644859314, "learning_rate": 9.36063403316534e-09, "loss": 0.3589, "step": 3531 }, { "epoch": 2.946607341490545, "grad_norm": 0.3039622902870178, "learning_rate": 9.06591860636019e-09, "loss": 0.4354, "step": 3532 }, { "epoch": 2.9474416017797553, "grad_norm": 0.28785961866378784, "learning_rate": 8.775912969237743e-09, "loss": 0.3637, "step": 3533 }, { "epoch": 2.9482758620689653, "grad_norm": 0.29674673080444336, "learning_rate": 8.49061739546686e-09, "loss": 0.4077, "step": 3534 }, { "epoch": 2.9491101223581757, "grad_norm": 0.2898532450199127, "learning_rate": 8.21003215427274e-09, "loss": 0.4087, "step": 3535 }, { "epoch": 2.949944382647386, "grad_norm": 0.28127920627593994, "learning_rate": 7.934157510434138e-09, "loss": 0.3764, "step": 3536 }, { "epoch": 2.950778642936596, "grad_norm": 0.28753623366355896, "learning_rate": 7.662993724286693e-09, "loss": 0.3937, "step": 3537 }, { "epoch": 2.9516129032258065, "grad_norm": 0.295627236366272, "learning_rate": 7.3965410517179426e-09, "loss": 0.4078, "step": 3538 }, { "epoch": 2.952447163515017, "grad_norm": 0.29336822032928467, "learning_rate": 7.13479974417175e-09, "loss": 0.4344, "step": 3539 }, { "epoch": 2.953281423804227, "grad_norm": 0.28384122252464294, "learning_rate": 6.877770048645538e-09, "loss": 0.3719, "step": 3540 }, { "epoch": 2.9541156840934373, "grad_norm": 0.2613103985786438, "learning_rate": 6.625452207689731e-09, "loss": 0.3884, "step": 3541 }, { "epoch": 2.9549499443826472, "grad_norm": 0.2776013910770416, "learning_rate": 6.37784645940831e-09, "loss": 0.3804, "step": 3542 }, { "epoch": 2.9557842046718577, "grad_norm": 0.3151768445968628, "learning_rate": 6.134953037460478e-09, "loss": 0.385, "step": 3543 }, { "epoch": 2.9566184649610676, "grad_norm": 0.27962344884872437, "learning_rate": 5.896772171056775e-09, "loss": 0.3957, "step": 3544 }, { "epoch": 2.957452725250278, "grad_norm": 0.29581794142723083, "learning_rate": 5.6633040849601865e-09, "loss": 0.3787, "step": 3545 }, { "epoch": 2.9582869855394884, "grad_norm": 0.28446370363235474, "learning_rate": 5.4345489994883645e-09, "loss": 0.3903, "step": 3546 }, { "epoch": 2.9591212458286984, "grad_norm": 0.2619404196739197, "learning_rate": 5.210507130509746e-09, "loss": 0.3397, "step": 3547 }, { "epoch": 2.959955506117909, "grad_norm": 0.2762039601802826, "learning_rate": 4.991178689445764e-09, "loss": 0.4103, "step": 3548 }, { "epoch": 2.960789766407119, "grad_norm": 0.30223143100738525, "learning_rate": 4.776563883270857e-09, "loss": 0.4092, "step": 3549 }, { "epoch": 2.961624026696329, "grad_norm": 0.2843014597892761, "learning_rate": 4.56666291450858e-09, "loss": 0.3507, "step": 3550 }, { "epoch": 2.9624582869855396, "grad_norm": 0.30934974551200867, "learning_rate": 4.361475981237706e-09, "loss": 0.4062, "step": 3551 }, { "epoch": 2.9632925472747496, "grad_norm": 0.27546247839927673, "learning_rate": 4.161003277085574e-09, "loss": 0.3504, "step": 3552 }, { "epoch": 2.96412680756396, "grad_norm": 0.29004156589508057, "learning_rate": 3.96524499123252e-09, "loss": 0.4385, "step": 3553 }, { "epoch": 2.96496106785317, "grad_norm": 0.2822257876396179, "learning_rate": 3.774201308409664e-09, "loss": 0.3917, "step": 3554 }, { "epoch": 2.9657953281423803, "grad_norm": 0.3003159463405609, "learning_rate": 3.587872408898907e-09, "loss": 0.4224, "step": 3555 }, { "epoch": 2.9666295884315907, "grad_norm": 0.29206693172454834, "learning_rate": 3.4062584685329304e-09, "loss": 0.4069, "step": 3556 }, { "epoch": 2.9674638487208007, "grad_norm": 0.3044872581958771, "learning_rate": 3.229359658695752e-09, "loss": 0.4014, "step": 3557 }, { "epoch": 2.968298109010011, "grad_norm": 0.3012820780277252, "learning_rate": 3.057176146319951e-09, "loss": 0.3758, "step": 3558 }, { "epoch": 2.9691323692992215, "grad_norm": 0.2815696597099304, "learning_rate": 2.8897080938916634e-09, "loss": 0.382, "step": 3559 }, { "epoch": 2.9699666295884315, "grad_norm": 0.2797418236732483, "learning_rate": 2.7269556594433645e-09, "loss": 0.4045, "step": 3560 }, { "epoch": 2.970800889877642, "grad_norm": 0.26507464051246643, "learning_rate": 2.568918996560532e-09, "loss": 0.352, "step": 3561 }, { "epoch": 2.9716351501668523, "grad_norm": 0.29521945118904114, "learning_rate": 2.41559825437776e-09, "loss": 0.4458, "step": 3562 }, { "epoch": 2.9724694104560623, "grad_norm": 0.27892816066741943, "learning_rate": 2.2669935775782026e-09, "loss": 0.3981, "step": 3563 }, { "epoch": 2.9733036707452727, "grad_norm": 0.2988332211971283, "learning_rate": 2.1231051063957955e-09, "loss": 0.415, "step": 3564 }, { "epoch": 2.9741379310344827, "grad_norm": 0.32047316431999207, "learning_rate": 1.9839329766141444e-09, "loss": 0.3763, "step": 3565 }, { "epoch": 2.974972191323693, "grad_norm": 0.3101384937763214, "learning_rate": 1.849477319564863e-09, "loss": 0.4227, "step": 3566 }, { "epoch": 2.975806451612903, "grad_norm": 0.2913734018802643, "learning_rate": 1.7197382621303437e-09, "loss": 0.3916, "step": 3567 }, { "epoch": 2.9766407119021134, "grad_norm": 0.27189216017723083, "learning_rate": 1.5947159267409863e-09, "loss": 0.3804, "step": 3568 }, { "epoch": 2.977474972191324, "grad_norm": 0.26789629459381104, "learning_rate": 1.4744104313763052e-09, "loss": 0.3754, "step": 3569 }, { "epoch": 2.978309232480534, "grad_norm": 0.3001950979232788, "learning_rate": 1.358821889564932e-09, "loss": 0.4324, "step": 3570 }, { "epoch": 2.979143492769744, "grad_norm": 0.2903038263320923, "learning_rate": 1.2479504103840579e-09, "loss": 0.4588, "step": 3571 }, { "epoch": 2.9799777530589546, "grad_norm": 0.28967899084091187, "learning_rate": 1.1417960984605459e-09, "loss": 0.349, "step": 3572 }, { "epoch": 2.9808120133481646, "grad_norm": 0.2984352707862854, "learning_rate": 1.040359053967599e-09, "loss": 0.3752, "step": 3573 }, { "epoch": 2.981646273637375, "grad_norm": 0.2965465188026428, "learning_rate": 9.436393726292014e-10, "loss": 0.3964, "step": 3574 }, { "epoch": 2.982480533926585, "grad_norm": 0.28276750445365906, "learning_rate": 8.516371457162331e-10, "loss": 0.3574, "step": 3575 }, { "epoch": 2.9833147942157954, "grad_norm": 0.29246285557746887, "learning_rate": 7.643524600481344e-10, "loss": 0.381, "step": 3576 }, { "epoch": 2.9841490545050053, "grad_norm": 0.29459959268569946, "learning_rate": 6.817853979934618e-10, "loss": 0.4241, "step": 3577 }, { "epoch": 2.9849833147942157, "grad_norm": 0.30874574184417725, "learning_rate": 6.039360374682224e-10, "loss": 0.4113, "step": 3578 }, { "epoch": 2.985817575083426, "grad_norm": 0.29302677512168884, "learning_rate": 5.308044519358735e-10, "loss": 0.357, "step": 3579 }, { "epoch": 2.986651835372636, "grad_norm": 0.29382452368736267, "learning_rate": 4.623907104084335e-10, "loss": 0.4329, "step": 3580 }, { "epoch": 2.9874860956618465, "grad_norm": 0.275650292634964, "learning_rate": 3.986948774459265e-10, "loss": 0.3881, "step": 3581 }, { "epoch": 2.988320355951057, "grad_norm": 0.29751113057136536, "learning_rate": 3.3971701315638205e-10, "loss": 0.3886, "step": 3582 }, { "epoch": 2.989154616240267, "grad_norm": 0.29810768365859985, "learning_rate": 2.854571731947253e-10, "loss": 0.4016, "step": 3583 }, { "epoch": 2.9899888765294773, "grad_norm": 0.30295199155807495, "learning_rate": 2.3591540876555244e-10, "loss": 0.4189, "step": 3584 }, { "epoch": 2.9908231368186873, "grad_norm": 0.2966918349266052, "learning_rate": 1.9109176661868957e-10, "loss": 0.3866, "step": 3585 }, { "epoch": 2.9916573971078977, "grad_norm": 0.2793084681034088, "learning_rate": 1.5098628905307887e-10, "loss": 0.3976, "step": 3586 }, { "epoch": 2.9924916573971077, "grad_norm": 0.2988084554672241, "learning_rate": 1.1559901391511308e-10, "loss": 0.3908, "step": 3587 }, { "epoch": 2.993325917686318, "grad_norm": 0.2782587707042694, "learning_rate": 8.49299745991905e-11, "loss": 0.3932, "step": 3588 }, { "epoch": 2.9941601779755285, "grad_norm": 0.29962342977523804, "learning_rate": 5.897920004604985e-11, "loss": 0.3996, "step": 3589 }, { "epoch": 2.9949944382647384, "grad_norm": 0.27788132429122925, "learning_rate": 3.7746714744435476e-11, "loss": 0.3844, "step": 3590 }, { "epoch": 2.995828698553949, "grad_norm": 0.27866125106811523, "learning_rate": 2.123253873165254e-11, "loss": 0.4018, "step": 3591 }, { "epoch": 2.9966629588431593, "grad_norm": 0.27451667189598083, "learning_rate": 9.436687591346527e-12, "loss": 0.3514, "step": 3592 }, { "epoch": 2.997497219132369, "grad_norm": 0.3089873194694519, "learning_rate": 2.359172454613479e-12, "loss": 0.4444, "step": 3593 }, { "epoch": 2.9983314794215796, "grad_norm": 0.3066548705101013, "learning_rate": 0.0, "loss": 0.4088, "step": 3594 }, { "epoch": 2.9983314794215796, "step": 3594, "total_flos": 4682415249588224.0, "train_loss": 0.44163876356485493, "train_runtime": 71857.4059, "train_samples_per_second": 4.804, "train_steps_per_second": 0.05 } ], "logging_steps": 1.0, "max_steps": 3594, "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": 4682415249588224.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }