diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,54714 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.999936, + "eval_steps": 500, + "global_step": 7812, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.000128, + "grad_norm": 137.0978546142578, + "learning_rate": 4.255319148936171e-06, + "loss": 8.1484, + "step": 1 + }, + { + "epoch": 0.000256, + "grad_norm": 92.25102996826172, + "learning_rate": 8.510638297872341e-06, + "loss": 8.011, + "step": 2 + }, + { + "epoch": 0.000384, + "grad_norm": 148.20196533203125, + "learning_rate": 1.2765957446808511e-05, + "loss": 8.0973, + "step": 3 + }, + { + "epoch": 0.000512, + "grad_norm": 77.27680206298828, + "learning_rate": 1.7021276595744682e-05, + "loss": 7.5874, + "step": 4 + }, + { + "epoch": 0.00064, + "grad_norm": 83.11038970947266, + "learning_rate": 2.1276595744680852e-05, + "loss": 6.9114, + "step": 5 + }, + { + "epoch": 0.000768, + "grad_norm": 31.194805145263672, + "learning_rate": 2.5531914893617022e-05, + "loss": 6.1825, + "step": 6 + }, + { + "epoch": 0.000896, + "grad_norm": 34.35151290893555, + "learning_rate": 2.9787234042553195e-05, + "loss": 5.7855, + "step": 7 + }, + { + "epoch": 0.001024, + "grad_norm": 23.203657150268555, + "learning_rate": 3.4042553191489365e-05, + "loss": 5.7993, + "step": 8 + }, + { + "epoch": 0.001152, + "grad_norm": 18.102510452270508, + "learning_rate": 3.829787234042553e-05, + "loss": 5.3982, + "step": 9 + }, + { + "epoch": 0.00128, + "grad_norm": 18.51192283630371, + "learning_rate": 4.2553191489361704e-05, + "loss": 5.0888, + "step": 10 + }, + { + "epoch": 0.001408, + "grad_norm": 14.597257614135742, + "learning_rate": 4.680851063829787e-05, + "loss": 4.9272, + "step": 11 + }, + { + "epoch": 0.001536, + "grad_norm": 12.254507064819336, + "learning_rate": 5.1063829787234044e-05, + "loss": 4.7903, + "step": 12 + }, + { + "epoch": 0.001664, + "grad_norm": 11.385390281677246, + "learning_rate": 5.531914893617021e-05, + "loss": 4.6265, + "step": 13 + }, + { + "epoch": 0.001792, + "grad_norm": 9.810650825500488, + "learning_rate": 5.957446808510639e-05, + "loss": 4.5584, + "step": 14 + }, + { + "epoch": 0.00192, + "grad_norm": 8.439884185791016, + "learning_rate": 6.382978723404255e-05, + "loss": 4.3065, + "step": 15 + }, + { + "epoch": 0.002048, + "grad_norm": 7.719231605529785, + "learning_rate": 6.808510638297873e-05, + "loss": 4.199, + "step": 16 + }, + { + "epoch": 0.002176, + "grad_norm": 4.830698013305664, + "learning_rate": 7.234042553191488e-05, + "loss": 4.1823, + "step": 17 + }, + { + "epoch": 0.002304, + "grad_norm": 4.571081161499023, + "learning_rate": 7.659574468085106e-05, + "loss": 4.1156, + "step": 18 + }, + { + "epoch": 0.002432, + "grad_norm": 4.783127784729004, + "learning_rate": 8.085106382978724e-05, + "loss": 3.995, + "step": 19 + }, + { + "epoch": 0.00256, + "grad_norm": 3.8924221992492676, + "learning_rate": 8.510638297872341e-05, + "loss": 4.0149, + "step": 20 + }, + { + "epoch": 0.002688, + "grad_norm": 4.823347091674805, + "learning_rate": 8.936170212765958e-05, + "loss": 4.0324, + "step": 21 + }, + { + "epoch": 0.002816, + "grad_norm": 3.839674711227417, + "learning_rate": 9.361702127659574e-05, + "loss": 4.0103, + "step": 22 + }, + { + "epoch": 0.002944, + "grad_norm": 2.3170769214630127, + "learning_rate": 9.787234042553192e-05, + "loss": 3.9063, + "step": 23 + }, + { + "epoch": 0.003072, + "grad_norm": 2.108315944671631, + "learning_rate": 0.00010212765957446809, + "loss": 3.7813, + "step": 24 + }, + { + "epoch": 0.0032, + "grad_norm": 2.423905611038208, + "learning_rate": 0.00010638297872340425, + "loss": 3.9902, + "step": 25 + }, + { + "epoch": 0.003328, + "grad_norm": 2.6972157955169678, + "learning_rate": 0.00011063829787234042, + "loss": 4.0109, + "step": 26 + }, + { + "epoch": 0.003456, + "grad_norm": 2.43788743019104, + "learning_rate": 0.0001148936170212766, + "loss": 3.7552, + "step": 27 + }, + { + "epoch": 0.003584, + "grad_norm": 1.981242060661316, + "learning_rate": 0.00011914893617021278, + "loss": 3.8227, + "step": 28 + }, + { + "epoch": 0.003712, + "grad_norm": 1.8447303771972656, + "learning_rate": 0.00012340425531914893, + "loss": 3.7902, + "step": 29 + }, + { + "epoch": 0.00384, + "grad_norm": 1.7584038972854614, + "learning_rate": 0.0001276595744680851, + "loss": 3.8017, + "step": 30 + }, + { + "epoch": 0.003968, + "grad_norm": 1.6604690551757812, + "learning_rate": 0.00013191489361702127, + "loss": 3.8535, + "step": 31 + }, + { + "epoch": 0.004096, + "grad_norm": 1.6179003715515137, + "learning_rate": 0.00013617021276595746, + "loss": 3.8878, + "step": 32 + }, + { + "epoch": 0.004224, + "grad_norm": 1.416571021080017, + "learning_rate": 0.00014042553191489363, + "loss": 3.7505, + "step": 33 + }, + { + "epoch": 0.004352, + "grad_norm": 1.402642011642456, + "learning_rate": 0.00014468085106382977, + "loss": 3.7275, + "step": 34 + }, + { + "epoch": 0.00448, + "grad_norm": 1.2541730403900146, + "learning_rate": 0.00014893617021276596, + "loss": 3.6886, + "step": 35 + }, + { + "epoch": 0.004608, + "grad_norm": 1.3865234851837158, + "learning_rate": 0.00015319148936170213, + "loss": 3.6909, + "step": 36 + }, + { + "epoch": 0.004736, + "grad_norm": 1.1091139316558838, + "learning_rate": 0.00015744680851063832, + "loss": 3.6254, + "step": 37 + }, + { + "epoch": 0.004864, + "grad_norm": 1.0914801359176636, + "learning_rate": 0.00016170212765957449, + "loss": 3.7275, + "step": 38 + }, + { + "epoch": 0.004992, + "grad_norm": 1.1153056621551514, + "learning_rate": 0.00016595744680851062, + "loss": 3.7522, + "step": 39 + }, + { + "epoch": 0.00512, + "grad_norm": 0.9447570443153381, + "learning_rate": 0.00017021276595744682, + "loss": 3.6581, + "step": 40 + }, + { + "epoch": 0.005248, + "grad_norm": 0.9334722757339478, + "learning_rate": 0.00017446808510638298, + "loss": 3.5512, + "step": 41 + }, + { + "epoch": 0.005376, + "grad_norm": 0.9279382228851318, + "learning_rate": 0.00017872340425531915, + "loss": 3.6989, + "step": 42 + }, + { + "epoch": 0.005504, + "grad_norm": 1.0880032777786255, + "learning_rate": 0.00018297872340425532, + "loss": 3.7452, + "step": 43 + }, + { + "epoch": 0.005632, + "grad_norm": 0.8086416721343994, + "learning_rate": 0.00018723404255319148, + "loss": 3.7177, + "step": 44 + }, + { + "epoch": 0.00576, + "grad_norm": 1.4188841581344604, + "learning_rate": 0.00019148936170212765, + "loss": 3.7037, + "step": 45 + }, + { + "epoch": 0.005888, + "grad_norm": 0.8778223395347595, + "learning_rate": 0.00019574468085106384, + "loss": 3.6689, + "step": 46 + }, + { + "epoch": 0.006016, + "grad_norm": 0.7749184370040894, + "learning_rate": 0.0002, + "loss": 3.6657, + "step": 47 + }, + { + "epoch": 0.006144, + "grad_norm": 0.7862442135810852, + "learning_rate": 0.00020425531914893618, + "loss": 3.67, + "step": 48 + }, + { + "epoch": 0.006272, + "grad_norm": 0.8049819469451904, + "learning_rate": 0.00020851063829787234, + "loss": 3.6149, + "step": 49 + }, + { + "epoch": 0.0064, + "grad_norm": 0.734851062297821, + "learning_rate": 0.0002127659574468085, + "loss": 3.516, + "step": 50 + }, + { + "epoch": 0.006528, + "grad_norm": 0.6626814007759094, + "learning_rate": 0.0002170212765957447, + "loss": 3.695, + "step": 51 + }, + { + "epoch": 0.006656, + "grad_norm": 0.7173753380775452, + "learning_rate": 0.00022127659574468084, + "loss": 3.6227, + "step": 52 + }, + { + "epoch": 0.006784, + "grad_norm": 0.806590735912323, + "learning_rate": 0.000225531914893617, + "loss": 3.7354, + "step": 53 + }, + { + "epoch": 0.006912, + "grad_norm": 0.7284512519836426, + "learning_rate": 0.0002297872340425532, + "loss": 3.746, + "step": 54 + }, + { + "epoch": 0.00704, + "grad_norm": 0.6704588532447815, + "learning_rate": 0.00023404255319148937, + "loss": 3.6246, + "step": 55 + }, + { + "epoch": 0.007168, + "grad_norm": 0.735169529914856, + "learning_rate": 0.00023829787234042556, + "loss": 3.7187, + "step": 56 + }, + { + "epoch": 0.007296, + "grad_norm": 0.7549604177474976, + "learning_rate": 0.0002425531914893617, + "loss": 3.6157, + "step": 57 + }, + { + "epoch": 0.007424, + "grad_norm": 0.7055206298828125, + "learning_rate": 0.00024680851063829787, + "loss": 3.6581, + "step": 58 + }, + { + "epoch": 0.007552, + "grad_norm": 0.6264505386352539, + "learning_rate": 0.00025106382978723403, + "loss": 3.5261, + "step": 59 + }, + { + "epoch": 0.00768, + "grad_norm": 0.7330943942070007, + "learning_rate": 0.0002553191489361702, + "loss": 3.5507, + "step": 60 + }, + { + "epoch": 0.007808, + "grad_norm": 0.6582585573196411, + "learning_rate": 0.00025957446808510637, + "loss": 3.6256, + "step": 61 + }, + { + "epoch": 0.007936, + "grad_norm": 0.5932968258857727, + "learning_rate": 0.00026382978723404253, + "loss": 3.6048, + "step": 62 + }, + { + "epoch": 0.008064, + "grad_norm": 0.8019038438796997, + "learning_rate": 0.00026808510638297875, + "loss": 3.7236, + "step": 63 + }, + { + "epoch": 0.008192, + "grad_norm": 0.7750895023345947, + "learning_rate": 0.0002723404255319149, + "loss": 3.6348, + "step": 64 + }, + { + "epoch": 0.00832, + "grad_norm": 0.7497636675834656, + "learning_rate": 0.0002765957446808511, + "loss": 3.6499, + "step": 65 + }, + { + "epoch": 0.008448, + "grad_norm": 0.5822744965553284, + "learning_rate": 0.00028085106382978725, + "loss": 3.4884, + "step": 66 + }, + { + "epoch": 0.008576, + "grad_norm": 0.6551232933998108, + "learning_rate": 0.0002851063829787234, + "loss": 3.6917, + "step": 67 + }, + { + "epoch": 0.008704, + "grad_norm": 0.6105189919471741, + "learning_rate": 0.00028936170212765953, + "loss": 3.5637, + "step": 68 + }, + { + "epoch": 0.008832, + "grad_norm": 0.5864344239234924, + "learning_rate": 0.00029361702127659575, + "loss": 3.5028, + "step": 69 + }, + { + "epoch": 0.00896, + "grad_norm": 0.6039401292800903, + "learning_rate": 0.0002978723404255319, + "loss": 3.6666, + "step": 70 + }, + { + "epoch": 0.009088, + "grad_norm": 0.6590763926506042, + "learning_rate": 0.0003021276595744681, + "loss": 3.692, + "step": 71 + }, + { + "epoch": 0.009216, + "grad_norm": 0.6972270011901855, + "learning_rate": 0.00030638297872340425, + "loss": 3.5489, + "step": 72 + }, + { + "epoch": 0.009344, + "grad_norm": 0.7538952231407166, + "learning_rate": 0.0003106382978723404, + "loss": 3.6564, + "step": 73 + }, + { + "epoch": 0.009472, + "grad_norm": 0.7196582555770874, + "learning_rate": 0.00031489361702127664, + "loss": 3.5602, + "step": 74 + }, + { + "epoch": 0.0096, + "grad_norm": 0.5942358374595642, + "learning_rate": 0.0003191489361702128, + "loss": 3.6068, + "step": 75 + }, + { + "epoch": 0.009728, + "grad_norm": 0.7040883302688599, + "learning_rate": 0.00032340425531914897, + "loss": 3.6537, + "step": 76 + }, + { + "epoch": 0.009856, + "grad_norm": 0.8344489932060242, + "learning_rate": 0.0003276595744680851, + "loss": 3.608, + "step": 77 + }, + { + "epoch": 0.009984, + "grad_norm": 1.0256277322769165, + "learning_rate": 0.00033191489361702125, + "loss": 3.5692, + "step": 78 + }, + { + "epoch": 0.010112, + "grad_norm": 0.8147671222686768, + "learning_rate": 0.0003361702127659574, + "loss": 3.5866, + "step": 79 + }, + { + "epoch": 0.01024, + "grad_norm": 0.7291015386581421, + "learning_rate": 0.00034042553191489364, + "loss": 3.5938, + "step": 80 + }, + { + "epoch": 0.010368, + "grad_norm": 1.0605134963989258, + "learning_rate": 0.0003446808510638298, + "loss": 3.5379, + "step": 81 + }, + { + "epoch": 0.010496, + "grad_norm": 0.8678496479988098, + "learning_rate": 0.00034893617021276597, + "loss": 3.6474, + "step": 82 + }, + { + "epoch": 0.010624, + "grad_norm": 0.8527273535728455, + "learning_rate": 0.00035319148936170213, + "loss": 3.6066, + "step": 83 + }, + { + "epoch": 0.010752, + "grad_norm": 0.9334797859191895, + "learning_rate": 0.0003574468085106383, + "loss": 3.5172, + "step": 84 + }, + { + "epoch": 0.01088, + "grad_norm": 0.7257692813873291, + "learning_rate": 0.0003617021276595745, + "loss": 3.6018, + "step": 85 + }, + { + "epoch": 0.011008, + "grad_norm": 1.1335358619689941, + "learning_rate": 0.00036595744680851063, + "loss": 3.5871, + "step": 86 + }, + { + "epoch": 0.011136, + "grad_norm": 0.9640700817108154, + "learning_rate": 0.0003702127659574468, + "loss": 3.6472, + "step": 87 + }, + { + "epoch": 0.011264, + "grad_norm": 0.8622243404388428, + "learning_rate": 0.00037446808510638297, + "loss": 3.6049, + "step": 88 + }, + { + "epoch": 0.011392, + "grad_norm": 0.9182441830635071, + "learning_rate": 0.00037872340425531913, + "loss": 3.7062, + "step": 89 + }, + { + "epoch": 0.01152, + "grad_norm": 1.099652886390686, + "learning_rate": 0.0003829787234042553, + "loss": 3.6639, + "step": 90 + }, + { + "epoch": 0.011648, + "grad_norm": 0.97944176197052, + "learning_rate": 0.0003872340425531915, + "loss": 3.6867, + "step": 91 + }, + { + "epoch": 0.011776, + "grad_norm": 0.9478636384010315, + "learning_rate": 0.0003914893617021277, + "loss": 3.4808, + "step": 92 + }, + { + "epoch": 0.011904, + "grad_norm": 0.9662417769432068, + "learning_rate": 0.00039574468085106385, + "loss": 3.4471, + "step": 93 + }, + { + "epoch": 0.012032, + "grad_norm": 0.9023204445838928, + "learning_rate": 0.0004, + "loss": 3.5289, + "step": 94 + }, + { + "epoch": 0.01216, + "grad_norm": 1.3472074270248413, + "learning_rate": 0.00040425531914893613, + "loss": 3.4973, + "step": 95 + }, + { + "epoch": 0.012288, + "grad_norm": 0.9390718340873718, + "learning_rate": 0.00040851063829787235, + "loss": 3.4974, + "step": 96 + }, + { + "epoch": 0.012416, + "grad_norm": 1.3564459085464478, + "learning_rate": 0.0004127659574468085, + "loss": 3.48, + "step": 97 + }, + { + "epoch": 0.012544, + "grad_norm": 1.183982491493225, + "learning_rate": 0.0004170212765957447, + "loss": 3.6251, + "step": 98 + }, + { + "epoch": 0.012672, + "grad_norm": 1.2942726612091064, + "learning_rate": 0.00042127659574468085, + "loss": 3.5328, + "step": 99 + }, + { + "epoch": 0.0128, + "grad_norm": 0.8319986462593079, + "learning_rate": 0.000425531914893617, + "loss": 3.5118, + "step": 100 + }, + { + "epoch": 0.012928, + "grad_norm": 1.131800651550293, + "learning_rate": 0.0004297872340425532, + "loss": 3.566, + "step": 101 + }, + { + "epoch": 0.013056, + "grad_norm": 0.8964243531227112, + "learning_rate": 0.0004340425531914894, + "loss": 3.412, + "step": 102 + }, + { + "epoch": 0.013184, + "grad_norm": 0.8629341721534729, + "learning_rate": 0.00043829787234042557, + "loss": 3.4855, + "step": 103 + }, + { + "epoch": 0.013312, + "grad_norm": 1.2416704893112183, + "learning_rate": 0.0004425531914893617, + "loss": 3.341, + "step": 104 + }, + { + "epoch": 0.01344, + "grad_norm": 27.238094329833984, + "learning_rate": 0.00044680851063829785, + "loss": 3.495, + "step": 105 + }, + { + "epoch": 0.013568, + "grad_norm": 4.09493350982666, + "learning_rate": 0.000451063829787234, + "loss": 3.694, + "step": 106 + }, + { + "epoch": 0.013696, + "grad_norm": 2.982325553894043, + "learning_rate": 0.00045531914893617024, + "loss": 3.6086, + "step": 107 + }, + { + "epoch": 0.013824, + "grad_norm": 3.10099720954895, + "learning_rate": 0.0004595744680851064, + "loss": 3.7353, + "step": 108 + }, + { + "epoch": 0.013952, + "grad_norm": 2.6089165210723877, + "learning_rate": 0.00046382978723404257, + "loss": 3.6823, + "step": 109 + }, + { + "epoch": 0.01408, + "grad_norm": 1.9390264749526978, + "learning_rate": 0.00046808510638297874, + "loss": 3.8355, + "step": 110 + }, + { + "epoch": 0.014208, + "grad_norm": 1.4381650686264038, + "learning_rate": 0.0004723404255319149, + "loss": 3.6121, + "step": 111 + }, + { + "epoch": 0.014336, + "grad_norm": 1.036955714225769, + "learning_rate": 0.0004765957446808511, + "loss": 3.5592, + "step": 112 + }, + { + "epoch": 0.014464, + "grad_norm": 1.0634974241256714, + "learning_rate": 0.00048085106382978723, + "loss": 3.8075, + "step": 113 + }, + { + "epoch": 0.014592, + "grad_norm": 1.0217565298080444, + "learning_rate": 0.0004851063829787234, + "loss": 3.6921, + "step": 114 + }, + { + "epoch": 0.01472, + "grad_norm": 1.0535690784454346, + "learning_rate": 0.0004893617021276596, + "loss": 3.7057, + "step": 115 + }, + { + "epoch": 0.014848, + "grad_norm": 1.2743030786514282, + "learning_rate": 0.0004936170212765957, + "loss": 3.6748, + "step": 116 + }, + { + "epoch": 0.014976, + "grad_norm": 0.728239893913269, + "learning_rate": 0.000497872340425532, + "loss": 3.7869, + "step": 117 + }, + { + "epoch": 0.015104, + "grad_norm": 0.8120728135108948, + "learning_rate": 0.0005021276595744681, + "loss": 3.78, + "step": 118 + }, + { + "epoch": 0.015232, + "grad_norm": 0.707918107509613, + "learning_rate": 0.0005063829787234042, + "loss": 3.7082, + "step": 119 + }, + { + "epoch": 0.01536, + "grad_norm": 0.7319228649139404, + "learning_rate": 0.0005106382978723404, + "loss": 3.5435, + "step": 120 + }, + { + "epoch": 0.015488, + "grad_norm": 0.66538405418396, + "learning_rate": 0.0005148936170212766, + "loss": 3.613, + "step": 121 + }, + { + "epoch": 0.015616, + "grad_norm": 2.0641486644744873, + "learning_rate": 0.0005191489361702127, + "loss": 3.6249, + "step": 122 + }, + { + "epoch": 0.015744, + "grad_norm": 1.0651967525482178, + "learning_rate": 0.000523404255319149, + "loss": 3.6545, + "step": 123 + }, + { + "epoch": 0.015872, + "grad_norm": 0.5395228862762451, + "learning_rate": 0.0005276595744680851, + "loss": 3.7132, + "step": 124 + }, + { + "epoch": 0.016, + "grad_norm": 0.6075120568275452, + "learning_rate": 0.0005319148936170213, + "loss": 3.5261, + "step": 125 + }, + { + "epoch": 0.016128, + "grad_norm": 0.6802886724472046, + "learning_rate": 0.0005361702127659575, + "loss": 3.5731, + "step": 126 + }, + { + "epoch": 0.016256, + "grad_norm": 0.6287043690681458, + "learning_rate": 0.0005404255319148936, + "loss": 3.6603, + "step": 127 + }, + { + "epoch": 0.016384, + "grad_norm": 0.5752686262130737, + "learning_rate": 0.0005446808510638298, + "loss": 3.7146, + "step": 128 + }, + { + "epoch": 0.016512, + "grad_norm": 0.6011753082275391, + "learning_rate": 0.000548936170212766, + "loss": 3.6363, + "step": 129 + }, + { + "epoch": 0.01664, + "grad_norm": 0.6716907620429993, + "learning_rate": 0.0005531914893617022, + "loss": 3.6015, + "step": 130 + }, + { + "epoch": 0.016768, + "grad_norm": 0.5629003643989563, + "learning_rate": 0.0005574468085106383, + "loss": 3.5962, + "step": 131 + }, + { + "epoch": 0.016896, + "grad_norm": 0.7295985817909241, + "learning_rate": 0.0005617021276595745, + "loss": 3.599, + "step": 132 + }, + { + "epoch": 0.017024, + "grad_norm": 0.8673283457756042, + "learning_rate": 0.0005659574468085107, + "loss": 3.7163, + "step": 133 + }, + { + "epoch": 0.017152, + "grad_norm": 0.5329201817512512, + "learning_rate": 0.0005702127659574468, + "loss": 3.6449, + "step": 134 + }, + { + "epoch": 0.01728, + "grad_norm": 0.9048171639442444, + "learning_rate": 0.0005744680851063831, + "loss": 3.5008, + "step": 135 + }, + { + "epoch": 0.017408, + "grad_norm": 0.5564020276069641, + "learning_rate": 0.0005787234042553191, + "loss": 3.5677, + "step": 136 + }, + { + "epoch": 0.017536, + "grad_norm": 0.6164737343788147, + "learning_rate": 0.0005829787234042553, + "loss": 3.5627, + "step": 137 + }, + { + "epoch": 0.017664, + "grad_norm": 1.9016156196594238, + "learning_rate": 0.0005872340425531915, + "loss": 3.4632, + "step": 138 + }, + { + "epoch": 0.017792, + "grad_norm": 1.971512794494629, + "learning_rate": 0.0005914893617021276, + "loss": 3.6326, + "step": 139 + }, + { + "epoch": 0.01792, + "grad_norm": 2.0172057151794434, + "learning_rate": 0.0005957446808510638, + "loss": 3.5098, + "step": 140 + }, + { + "epoch": 0.018048, + "grad_norm": 1.1590466499328613, + "learning_rate": 0.0006, + "loss": 3.6471, + "step": 141 + }, + { + "epoch": 0.018176, + "grad_norm": 1.563492774963379, + "learning_rate": 0.0006042553191489362, + "loss": 3.6579, + "step": 142 + }, + { + "epoch": 0.018304, + "grad_norm": 0.8897820711135864, + "learning_rate": 0.0006085106382978724, + "loss": 3.5453, + "step": 143 + }, + { + "epoch": 0.018432, + "grad_norm": 0.8937811255455017, + "learning_rate": 0.0006127659574468085, + "loss": 3.5618, + "step": 144 + }, + { + "epoch": 0.01856, + "grad_norm": 0.9533325433731079, + "learning_rate": 0.0006170212765957447, + "loss": 3.3462, + "step": 145 + }, + { + "epoch": 0.018688, + "grad_norm": 0.9393768310546875, + "learning_rate": 0.0006212765957446808, + "loss": 3.5499, + "step": 146 + }, + { + "epoch": 0.018816, + "grad_norm": 0.8685979843139648, + "learning_rate": 0.000625531914893617, + "loss": 3.3623, + "step": 147 + }, + { + "epoch": 0.018944, + "grad_norm": 0.8216493129730225, + "learning_rate": 0.0006297872340425533, + "loss": 3.507, + "step": 148 + }, + { + "epoch": 0.019072, + "grad_norm": 0.7729883193969727, + "learning_rate": 0.0006340425531914894, + "loss": 3.476, + "step": 149 + }, + { + "epoch": 0.0192, + "grad_norm": 0.7121238708496094, + "learning_rate": 0.0006382978723404256, + "loss": 3.4699, + "step": 150 + }, + { + "epoch": 0.019328, + "grad_norm": 0.7459253072738647, + "learning_rate": 0.0006425531914893617, + "loss": 3.4394, + "step": 151 + }, + { + "epoch": 0.019456, + "grad_norm": 0.7433508634567261, + "learning_rate": 0.0006468085106382979, + "loss": 3.4839, + "step": 152 + }, + { + "epoch": 0.019584, + "grad_norm": 0.9679008722305298, + "learning_rate": 0.0006510638297872342, + "loss": 3.3985, + "step": 153 + }, + { + "epoch": 0.019712, + "grad_norm": 0.9967141151428223, + "learning_rate": 0.0006553191489361702, + "loss": 3.4481, + "step": 154 + }, + { + "epoch": 0.01984, + "grad_norm": 0.9571322798728943, + "learning_rate": 0.0006595744680851064, + "loss": 3.4819, + "step": 155 + }, + { + "epoch": 0.019968, + "grad_norm": 0.9345116019248962, + "learning_rate": 0.0006638297872340425, + "loss": 3.4828, + "step": 156 + }, + { + "epoch": 0.020096, + "grad_norm": 0.7965233325958252, + "learning_rate": 0.0006680851063829787, + "loss": 3.3709, + "step": 157 + }, + { + "epoch": 0.020224, + "grad_norm": 0.8162989020347595, + "learning_rate": 0.0006723404255319148, + "loss": 3.4867, + "step": 158 + }, + { + "epoch": 0.020352, + "grad_norm": 1.1211918592453003, + "learning_rate": 0.000676595744680851, + "loss": 3.5553, + "step": 159 + }, + { + "epoch": 0.02048, + "grad_norm": 0.9674050211906433, + "learning_rate": 0.0006808510638297873, + "loss": 3.4788, + "step": 160 + }, + { + "epoch": 0.020608, + "grad_norm": 0.9817243218421936, + "learning_rate": 0.0006851063829787234, + "loss": 3.3629, + "step": 161 + }, + { + "epoch": 0.020736, + "grad_norm": 0.8595209717750549, + "learning_rate": 0.0006893617021276596, + "loss": 3.4147, + "step": 162 + }, + { + "epoch": 0.020864, + "grad_norm": 1.0436092615127563, + "learning_rate": 0.0006936170212765957, + "loss": 3.4755, + "step": 163 + }, + { + "epoch": 0.020992, + "grad_norm": 0.771826982498169, + "learning_rate": 0.0006978723404255319, + "loss": 3.3852, + "step": 164 + }, + { + "epoch": 0.02112, + "grad_norm": 0.9518557190895081, + "learning_rate": 0.0007021276595744682, + "loss": 3.4391, + "step": 165 + }, + { + "epoch": 0.021248, + "grad_norm": 0.8252353072166443, + "learning_rate": 0.0007063829787234043, + "loss": 3.3634, + "step": 166 + }, + { + "epoch": 0.021376, + "grad_norm": 0.9902182817459106, + "learning_rate": 0.0007106382978723405, + "loss": 3.445, + "step": 167 + }, + { + "epoch": 0.021504, + "grad_norm": 0.8893678188323975, + "learning_rate": 0.0007148936170212766, + "loss": 3.3922, + "step": 168 + }, + { + "epoch": 0.021632, + "grad_norm": 0.8823418617248535, + "learning_rate": 0.0007191489361702128, + "loss": 3.5372, + "step": 169 + }, + { + "epoch": 0.02176, + "grad_norm": 1.2956855297088623, + "learning_rate": 0.000723404255319149, + "loss": 3.3809, + "step": 170 + }, + { + "epoch": 0.021888, + "grad_norm": 0.9388294816017151, + "learning_rate": 0.0007276595744680852, + "loss": 3.4143, + "step": 171 + }, + { + "epoch": 0.022016, + "grad_norm": 1.1023266315460205, + "learning_rate": 0.0007319148936170213, + "loss": 3.3745, + "step": 172 + }, + { + "epoch": 0.022144, + "grad_norm": 1.0422903299331665, + "learning_rate": 0.0007361702127659574, + "loss": 3.3941, + "step": 173 + }, + { + "epoch": 0.022272, + "grad_norm": 0.823570966720581, + "learning_rate": 0.0007404255319148936, + "loss": 3.4807, + "step": 174 + }, + { + "epoch": 0.0224, + "grad_norm": 1.1146208047866821, + "learning_rate": 0.0007446808510638298, + "loss": 3.4888, + "step": 175 + }, + { + "epoch": 0.022528, + "grad_norm": 1.0240081548690796, + "learning_rate": 0.0007489361702127659, + "loss": 3.2523, + "step": 176 + }, + { + "epoch": 0.022656, + "grad_norm": 1.0671203136444092, + "learning_rate": 0.0007531914893617022, + "loss": 3.5353, + "step": 177 + }, + { + "epoch": 0.022784, + "grad_norm": 1.0950299501419067, + "learning_rate": 0.0007574468085106383, + "loss": 3.2867, + "step": 178 + }, + { + "epoch": 0.022912, + "grad_norm": 0.9805471301078796, + "learning_rate": 0.0007617021276595745, + "loss": 3.3335, + "step": 179 + }, + { + "epoch": 0.02304, + "grad_norm": 1.1763050556182861, + "learning_rate": 0.0007659574468085106, + "loss": 3.3488, + "step": 180 + }, + { + "epoch": 0.023168, + "grad_norm": 1.0312926769256592, + "learning_rate": 0.0007702127659574468, + "loss": 3.381, + "step": 181 + }, + { + "epoch": 0.023296, + "grad_norm": 1.1540509462356567, + "learning_rate": 0.000774468085106383, + "loss": 3.3357, + "step": 182 + }, + { + "epoch": 0.023424, + "grad_norm": 0.7416232824325562, + "learning_rate": 0.0007787234042553192, + "loss": 3.3159, + "step": 183 + }, + { + "epoch": 0.023552, + "grad_norm": 0.9957670569419861, + "learning_rate": 0.0007829787234042554, + "loss": 3.3222, + "step": 184 + }, + { + "epoch": 0.02368, + "grad_norm": 1.089540719985962, + "learning_rate": 0.0007872340425531915, + "loss": 3.4317, + "step": 185 + }, + { + "epoch": 0.023808, + "grad_norm": 0.9947872757911682, + "learning_rate": 0.0007914893617021277, + "loss": 3.3757, + "step": 186 + }, + { + "epoch": 0.023936, + "grad_norm": 1.0085082054138184, + "learning_rate": 0.0007957446808510639, + "loss": 3.2999, + "step": 187 + }, + { + "epoch": 0.024064, + "grad_norm": 1.2522938251495361, + "learning_rate": 0.0008, + "loss": 3.3293, + "step": 188 + }, + { + "epoch": 0.024192, + "grad_norm": 0.9051831364631653, + "learning_rate": 0.0008042553191489363, + "loss": 3.3556, + "step": 189 + }, + { + "epoch": 0.02432, + "grad_norm": 1.09893000125885, + "learning_rate": 0.0008085106382978723, + "loss": 3.3504, + "step": 190 + }, + { + "epoch": 0.024448, + "grad_norm": 1.5150002241134644, + "learning_rate": 0.0008127659574468085, + "loss": 3.251, + "step": 191 + }, + { + "epoch": 0.024576, + "grad_norm": 1.3438389301300049, + "learning_rate": 0.0008170212765957447, + "loss": 3.3291, + "step": 192 + }, + { + "epoch": 0.024704, + "grad_norm": 1.4488674402236938, + "learning_rate": 0.0008212765957446808, + "loss": 3.4458, + "step": 193 + }, + { + "epoch": 0.024832, + "grad_norm": 1.214733362197876, + "learning_rate": 0.000825531914893617, + "loss": 3.2856, + "step": 194 + }, + { + "epoch": 0.02496, + "grad_norm": 1.0625321865081787, + "learning_rate": 0.0008297872340425531, + "loss": 3.4854, + "step": 195 + }, + { + "epoch": 0.025088, + "grad_norm": 0.9422627091407776, + "learning_rate": 0.0008340425531914894, + "loss": 3.3307, + "step": 196 + }, + { + "epoch": 0.025216, + "grad_norm": 1.089637279510498, + "learning_rate": 0.0008382978723404256, + "loss": 3.2985, + "step": 197 + }, + { + "epoch": 0.025344, + "grad_norm": 0.8540605306625366, + "learning_rate": 0.0008425531914893617, + "loss": 3.2847, + "step": 198 + }, + { + "epoch": 0.025472, + "grad_norm": 0.9661247730255127, + "learning_rate": 0.0008468085106382979, + "loss": 3.3252, + "step": 199 + }, + { + "epoch": 0.0256, + "grad_norm": 1.1259022951126099, + "learning_rate": 0.000851063829787234, + "loss": 3.3286, + "step": 200 + }, + { + "epoch": 0.025728, + "grad_norm": 1.0061057806015015, + "learning_rate": 0.0008553191489361703, + "loss": 3.2581, + "step": 201 + }, + { + "epoch": 0.025856, + "grad_norm": 1.0879464149475098, + "learning_rate": 0.0008595744680851064, + "loss": 3.3087, + "step": 202 + }, + { + "epoch": 0.025984, + "grad_norm": 1.061371088027954, + "learning_rate": 0.0008638297872340426, + "loss": 3.3182, + "step": 203 + }, + { + "epoch": 0.026112, + "grad_norm": 0.8742418885231018, + "learning_rate": 0.0008680851063829788, + "loss": 3.3001, + "step": 204 + }, + { + "epoch": 0.02624, + "grad_norm": 0.9869832992553711, + "learning_rate": 0.0008723404255319149, + "loss": 3.3445, + "step": 205 + }, + { + "epoch": 0.026368, + "grad_norm": 0.9327910542488098, + "learning_rate": 0.0008765957446808511, + "loss": 3.2075, + "step": 206 + }, + { + "epoch": 0.026496, + "grad_norm": 0.8447566628456116, + "learning_rate": 0.0008808510638297873, + "loss": 3.2423, + "step": 207 + }, + { + "epoch": 0.026624, + "grad_norm": 0.8698275089263916, + "learning_rate": 0.0008851063829787234, + "loss": 3.2062, + "step": 208 + }, + { + "epoch": 0.026752, + "grad_norm": 0.9042931199073792, + "learning_rate": 0.0008893617021276596, + "loss": 3.296, + "step": 209 + }, + { + "epoch": 0.02688, + "grad_norm": 0.9367974400520325, + "learning_rate": 0.0008936170212765957, + "loss": 3.3047, + "step": 210 + }, + { + "epoch": 0.027008, + "grad_norm": 0.8877196907997131, + "learning_rate": 0.0008978723404255319, + "loss": 3.1979, + "step": 211 + }, + { + "epoch": 0.027136, + "grad_norm": 1.025023102760315, + "learning_rate": 0.000902127659574468, + "loss": 3.219, + "step": 212 + }, + { + "epoch": 0.027264, + "grad_norm": 1.084191918373108, + "learning_rate": 0.0009063829787234043, + "loss": 3.2532, + "step": 213 + }, + { + "epoch": 0.027392, + "grad_norm": 1.0451518297195435, + "learning_rate": 0.0009106382978723405, + "loss": 3.1949, + "step": 214 + }, + { + "epoch": 0.02752, + "grad_norm": 1.00088632106781, + "learning_rate": 0.0009148936170212766, + "loss": 3.1486, + "step": 215 + }, + { + "epoch": 0.027648, + "grad_norm": 0.9949928522109985, + "learning_rate": 0.0009191489361702128, + "loss": 3.1372, + "step": 216 + }, + { + "epoch": 0.027776, + "grad_norm": 0.9554286599159241, + "learning_rate": 0.0009234042553191489, + "loss": 3.1717, + "step": 217 + }, + { + "epoch": 0.027904, + "grad_norm": 0.857183039188385, + "learning_rate": 0.0009276595744680851, + "loss": 3.1946, + "step": 218 + }, + { + "epoch": 0.028032, + "grad_norm": 1.1246652603149414, + "learning_rate": 0.0009319148936170214, + "loss": 3.1488, + "step": 219 + }, + { + "epoch": 0.02816, + "grad_norm": 1.0726946592330933, + "learning_rate": 0.0009361702127659575, + "loss": 3.3005, + "step": 220 + }, + { + "epoch": 0.028288, + "grad_norm": 0.9822759032249451, + "learning_rate": 0.0009404255319148937, + "loss": 3.24, + "step": 221 + }, + { + "epoch": 0.028416, + "grad_norm": 1.0777263641357422, + "learning_rate": 0.0009446808510638298, + "loss": 3.2488, + "step": 222 + }, + { + "epoch": 0.028544, + "grad_norm": 0.9558141231536865, + "learning_rate": 0.000948936170212766, + "loss": 3.1317, + "step": 223 + }, + { + "epoch": 0.028672, + "grad_norm": 1.1451855897903442, + "learning_rate": 0.0009531914893617022, + "loss": 3.221, + "step": 224 + }, + { + "epoch": 0.0288, + "grad_norm": 0.9488675594329834, + "learning_rate": 0.0009574468085106384, + "loss": 3.1229, + "step": 225 + }, + { + "epoch": 0.028928, + "grad_norm": 1.1733403205871582, + "learning_rate": 0.0009617021276595745, + "loss": 3.2061, + "step": 226 + }, + { + "epoch": 0.029056, + "grad_norm": 0.9430031180381775, + "learning_rate": 0.0009659574468085106, + "loss": 3.142, + "step": 227 + }, + { + "epoch": 0.029184, + "grad_norm": 1.0563158988952637, + "learning_rate": 0.0009702127659574468, + "loss": 3.2456, + "step": 228 + }, + { + "epoch": 0.029312, + "grad_norm": 0.9885168671607971, + "learning_rate": 0.0009744680851063829, + "loss": 3.2132, + "step": 229 + }, + { + "epoch": 0.02944, + "grad_norm": 0.8902058601379395, + "learning_rate": 0.0009787234042553192, + "loss": 3.0391, + "step": 230 + }, + { + "epoch": 0.029568, + "grad_norm": 0.9703031778335571, + "learning_rate": 0.0009829787234042554, + "loss": 3.243, + "step": 231 + }, + { + "epoch": 0.029696, + "grad_norm": 1.0198436975479126, + "learning_rate": 0.0009872340425531915, + "loss": 3.1992, + "step": 232 + }, + { + "epoch": 0.029824, + "grad_norm": 0.9725392460823059, + "learning_rate": 0.0009914893617021276, + "loss": 3.1922, + "step": 233 + }, + { + "epoch": 0.029952, + "grad_norm": 0.8880071640014648, + "learning_rate": 0.000995744680851064, + "loss": 3.0622, + "step": 234 + }, + { + "epoch": 0.03008, + "grad_norm": 0.9277142286300659, + "learning_rate": 0.001, + "loss": 3.1859, + "step": 235 + }, + { + "epoch": 0.030208, + "grad_norm": 0.9669767022132874, + "learning_rate": 0.0009999999570221019, + "loss": 3.075, + "step": 236 + }, + { + "epoch": 0.030336, + "grad_norm": 0.974901556968689, + "learning_rate": 0.0009999998280884144, + "loss": 3.0961, + "step": 237 + }, + { + "epoch": 0.030464, + "grad_norm": 0.9276684522628784, + "learning_rate": 0.0009999996131989603, + "loss": 3.2345, + "step": 238 + }, + { + "epoch": 0.030592, + "grad_norm": 1.280395746231079, + "learning_rate": 0.0009999993123537758, + "loss": 3.2413, + "step": 239 + }, + { + "epoch": 0.03072, + "grad_norm": 0.8946695327758789, + "learning_rate": 0.0009999989255529132, + "loss": 3.1446, + "step": 240 + }, + { + "epoch": 0.030848, + "grad_norm": 0.9685547351837158, + "learning_rate": 0.0009999984527964388, + "loss": 3.1467, + "step": 241 + }, + { + "epoch": 0.030976, + "grad_norm": 0.8855364918708801, + "learning_rate": 0.000999997894084434, + "loss": 3.0503, + "step": 242 + }, + { + "epoch": 0.031104, + "grad_norm": 0.8917241096496582, + "learning_rate": 0.0009999972494169947, + "loss": 3.1722, + "step": 243 + }, + { + "epoch": 0.031232, + "grad_norm": 0.8910233378410339, + "learning_rate": 0.0009999965187942315, + "loss": 3.2075, + "step": 244 + }, + { + "epoch": 0.03136, + "grad_norm": 1.0061917304992676, + "learning_rate": 0.0009999957022162706, + "loss": 3.0826, + "step": 245 + }, + { + "epoch": 0.031488, + "grad_norm": 1.034176230430603, + "learning_rate": 0.0009999947996832518, + "loss": 3.1446, + "step": 246 + }, + { + "epoch": 0.031616, + "grad_norm": 0.943690836429596, + "learning_rate": 0.0009999938111953305, + "loss": 3.1511, + "step": 247 + }, + { + "epoch": 0.031744, + "grad_norm": 1.0145138502120972, + "learning_rate": 0.0009999927367526768, + "loss": 3.04, + "step": 248 + }, + { + "epoch": 0.031872, + "grad_norm": 0.927814245223999, + "learning_rate": 0.0009999915763554753, + "loss": 3.0253, + "step": 249 + }, + { + "epoch": 0.032, + "grad_norm": 1.1053346395492554, + "learning_rate": 0.0009999903300039253, + "loss": 3.2994, + "step": 250 + }, + { + "epoch": 0.032128, + "grad_norm": 0.9806634187698364, + "learning_rate": 0.0009999889976982413, + "loss": 3.0757, + "step": 251 + }, + { + "epoch": 0.032256, + "grad_norm": 0.9248946905136108, + "learning_rate": 0.000999987579438652, + "loss": 3.2575, + "step": 252 + }, + { + "epoch": 0.032384, + "grad_norm": 0.9743854403495789, + "learning_rate": 0.0009999860752254014, + "loss": 3.1359, + "step": 253 + }, + { + "epoch": 0.032512, + "grad_norm": 0.8590320944786072, + "learning_rate": 0.0009999844850587486, + "loss": 3.0491, + "step": 254 + }, + { + "epoch": 0.03264, + "grad_norm": 0.9440402984619141, + "learning_rate": 0.000999982808938966, + "loss": 3.1793, + "step": 255 + }, + { + "epoch": 0.032768, + "grad_norm": 0.9098583459854126, + "learning_rate": 0.0009999810468663425, + "loss": 3.11, + "step": 256 + }, + { + "epoch": 0.032896, + "grad_norm": 0.9364376664161682, + "learning_rate": 0.0009999791988411805, + "loss": 3.149, + "step": 257 + }, + { + "epoch": 0.033024, + "grad_norm": 0.9506096839904785, + "learning_rate": 0.0009999772648637984, + "loss": 3.0807, + "step": 258 + }, + { + "epoch": 0.033152, + "grad_norm": 2.148585081100464, + "learning_rate": 0.000999975244934528, + "loss": 3.1175, + "step": 259 + }, + { + "epoch": 0.03328, + "grad_norm": 0.9461807608604431, + "learning_rate": 0.0009999731390537168, + "loss": 3.0917, + "step": 260 + }, + { + "epoch": 0.033408, + "grad_norm": 0.9068469405174255, + "learning_rate": 0.0009999709472217266, + "loss": 3.0895, + "step": 261 + }, + { + "epoch": 0.033536, + "grad_norm": 0.9892480969429016, + "learning_rate": 0.0009999686694389348, + "loss": 3.1054, + "step": 262 + }, + { + "epoch": 0.033664, + "grad_norm": 0.8607271313667297, + "learning_rate": 0.0009999663057057323, + "loss": 3.0972, + "step": 263 + }, + { + "epoch": 0.033792, + "grad_norm": 0.970432460308075, + "learning_rate": 0.0009999638560225258, + "loss": 3.1763, + "step": 264 + }, + { + "epoch": 0.03392, + "grad_norm": 0.9157929420471191, + "learning_rate": 0.0009999613203897364, + "loss": 3.0811, + "step": 265 + }, + { + "epoch": 0.034048, + "grad_norm": 0.9974496960639954, + "learning_rate": 0.0009999586988078, + "loss": 3.1489, + "step": 266 + }, + { + "epoch": 0.034176, + "grad_norm": 0.9609981775283813, + "learning_rate": 0.0009999559912771672, + "loss": 3.159, + "step": 267 + }, + { + "epoch": 0.034304, + "grad_norm": 0.847595751285553, + "learning_rate": 0.0009999531977983036, + "loss": 3.0848, + "step": 268 + }, + { + "epoch": 0.034432, + "grad_norm": 0.8561301827430725, + "learning_rate": 0.0009999503183716894, + "loss": 2.9538, + "step": 269 + }, + { + "epoch": 0.03456, + "grad_norm": 0.9255397915840149, + "learning_rate": 0.0009999473529978193, + "loss": 2.9977, + "step": 270 + }, + { + "epoch": 0.034688, + "grad_norm": 0.9491379261016846, + "learning_rate": 0.0009999443016772036, + "loss": 3.1228, + "step": 271 + }, + { + "epoch": 0.034816, + "grad_norm": 1.189246654510498, + "learning_rate": 0.0009999411644103665, + "loss": 3.1584, + "step": 272 + }, + { + "epoch": 0.034944, + "grad_norm": 1.116546392440796, + "learning_rate": 0.0009999379411978473, + "loss": 3.1514, + "step": 273 + }, + { + "epoch": 0.035072, + "grad_norm": 0.9567472338676453, + "learning_rate": 0.0009999346320402002, + "loss": 3.0595, + "step": 274 + }, + { + "epoch": 0.0352, + "grad_norm": 0.9525822997093201, + "learning_rate": 0.0009999312369379943, + "loss": 3.1966, + "step": 275 + }, + { + "epoch": 0.035328, + "grad_norm": 0.8483976125717163, + "learning_rate": 0.000999927755891813, + "loss": 3.0836, + "step": 276 + }, + { + "epoch": 0.035456, + "grad_norm": 0.853813886642456, + "learning_rate": 0.0009999241889022549, + "loss": 3.0991, + "step": 277 + }, + { + "epoch": 0.035584, + "grad_norm": 0.8766281008720398, + "learning_rate": 0.000999920535969933, + "loss": 3.0817, + "step": 278 + }, + { + "epoch": 0.035712, + "grad_norm": 0.7899410128593445, + "learning_rate": 0.0009999167970954753, + "loss": 3.0786, + "step": 279 + }, + { + "epoch": 0.03584, + "grad_norm": 0.959100604057312, + "learning_rate": 0.0009999129722795248, + "loss": 2.9862, + "step": 280 + }, + { + "epoch": 0.035968, + "grad_norm": 0.9234681725502014, + "learning_rate": 0.000999909061522739, + "loss": 3.1799, + "step": 281 + }, + { + "epoch": 0.036096, + "grad_norm": 0.8719414472579956, + "learning_rate": 0.0009999050648257899, + "loss": 2.9985, + "step": 282 + }, + { + "epoch": 0.036224, + "grad_norm": 0.9887234568595886, + "learning_rate": 0.0009999009821893647, + "loss": 3.1257, + "step": 283 + }, + { + "epoch": 0.036352, + "grad_norm": 0.9738095998764038, + "learning_rate": 0.0009998968136141653, + "loss": 3.0401, + "step": 284 + }, + { + "epoch": 0.03648, + "grad_norm": 0.85678631067276, + "learning_rate": 0.0009998925591009084, + "loss": 3.0317, + "step": 285 + }, + { + "epoch": 0.036608, + "grad_norm": 0.8604729175567627, + "learning_rate": 0.0009998882186503255, + "loss": 3.142, + "step": 286 + }, + { + "epoch": 0.036736, + "grad_norm": 1.0266964435577393, + "learning_rate": 0.0009998837922631624, + "loss": 3.0753, + "step": 287 + }, + { + "epoch": 0.036864, + "grad_norm": 0.8378276228904724, + "learning_rate": 0.0009998792799401804, + "loss": 3.0958, + "step": 288 + }, + { + "epoch": 0.036992, + "grad_norm": 0.8743003606796265, + "learning_rate": 0.000999874681682155, + "loss": 3.1208, + "step": 289 + }, + { + "epoch": 0.03712, + "grad_norm": 0.9393128156661987, + "learning_rate": 0.0009998699974898767, + "loss": 3.1858, + "step": 290 + }, + { + "epoch": 0.037248, + "grad_norm": 1.354142189025879, + "learning_rate": 0.000999865227364151, + "loss": 3.1365, + "step": 291 + }, + { + "epoch": 0.037376, + "grad_norm": 0.9946127533912659, + "learning_rate": 0.0009998603713057977, + "loss": 3.1273, + "step": 292 + }, + { + "epoch": 0.037504, + "grad_norm": 0.9451419115066528, + "learning_rate": 0.0009998554293156516, + "loss": 2.9788, + "step": 293 + }, + { + "epoch": 0.037632, + "grad_norm": 0.861698567867279, + "learning_rate": 0.0009998504013945626, + "loss": 3.0153, + "step": 294 + }, + { + "epoch": 0.03776, + "grad_norm": 0.9182236790657043, + "learning_rate": 0.0009998452875433948, + "loss": 3.0531, + "step": 295 + }, + { + "epoch": 0.037888, + "grad_norm": 0.8742627501487732, + "learning_rate": 0.0009998400877630272, + "loss": 2.9862, + "step": 296 + }, + { + "epoch": 0.038016, + "grad_norm": 0.84471595287323, + "learning_rate": 0.000999834802054354, + "loss": 2.9123, + "step": 297 + }, + { + "epoch": 0.038144, + "grad_norm": 0.888887345790863, + "learning_rate": 0.0009998294304182837, + "loss": 3.0967, + "step": 298 + }, + { + "epoch": 0.038272, + "grad_norm": 0.996583878993988, + "learning_rate": 0.0009998239728557397, + "loss": 2.9616, + "step": 299 + }, + { + "epoch": 0.0384, + "grad_norm": 1.152941107749939, + "learning_rate": 0.0009998184293676603, + "loss": 3.0105, + "step": 300 + }, + { + "epoch": 0.038528, + "grad_norm": 0.9084758758544922, + "learning_rate": 0.0009998127999549988, + "loss": 3.1371, + "step": 301 + }, + { + "epoch": 0.038656, + "grad_norm": 0.9090200662612915, + "learning_rate": 0.0009998070846187223, + "loss": 3.0323, + "step": 302 + }, + { + "epoch": 0.038784, + "grad_norm": 0.8822016716003418, + "learning_rate": 0.000999801283359814, + "loss": 3.0405, + "step": 303 + }, + { + "epoch": 0.038912, + "grad_norm": 0.8099454641342163, + "learning_rate": 0.0009997953961792708, + "loss": 3.1324, + "step": 304 + }, + { + "epoch": 0.03904, + "grad_norm": 0.84024977684021, + "learning_rate": 0.0009997894230781047, + "loss": 3.1308, + "step": 305 + }, + { + "epoch": 0.039168, + "grad_norm": 0.8685364723205566, + "learning_rate": 0.0009997833640573428, + "loss": 3.0935, + "step": 306 + }, + { + "epoch": 0.039296, + "grad_norm": 0.9617405533790588, + "learning_rate": 0.0009997772191180268, + "loss": 3.1709, + "step": 307 + }, + { + "epoch": 0.039424, + "grad_norm": 0.863365113735199, + "learning_rate": 0.0009997709882612128, + "loss": 3.0435, + "step": 308 + }, + { + "epoch": 0.039552, + "grad_norm": 0.8060149550437927, + "learning_rate": 0.000999764671487972, + "loss": 3.065, + "step": 309 + }, + { + "epoch": 0.03968, + "grad_norm": 0.8881270289421082, + "learning_rate": 0.0009997582687993904, + "loss": 2.9841, + "step": 310 + }, + { + "epoch": 0.039808, + "grad_norm": 0.8435242176055908, + "learning_rate": 0.000999751780196569, + "loss": 2.9915, + "step": 311 + }, + { + "epoch": 0.039936, + "grad_norm": 0.8635302186012268, + "learning_rate": 0.0009997452056806227, + "loss": 3.0157, + "step": 312 + }, + { + "epoch": 0.040064, + "grad_norm": 0.891107439994812, + "learning_rate": 0.000999738545252682, + "loss": 3.0515, + "step": 313 + }, + { + "epoch": 0.040192, + "grad_norm": 0.8483836650848389, + "learning_rate": 0.0009997317989138919, + "loss": 3.1188, + "step": 314 + }, + { + "epoch": 0.04032, + "grad_norm": 0.868516206741333, + "learning_rate": 0.0009997249666654122, + "loss": 3.0692, + "step": 315 + }, + { + "epoch": 0.040448, + "grad_norm": 0.9362041354179382, + "learning_rate": 0.0009997180485084175, + "loss": 3.0298, + "step": 316 + }, + { + "epoch": 0.040576, + "grad_norm": 0.9011420011520386, + "learning_rate": 0.000999711044444097, + "loss": 3.0949, + "step": 317 + }, + { + "epoch": 0.040704, + "grad_norm": 0.8734798431396484, + "learning_rate": 0.0009997039544736546, + "loss": 3.0913, + "step": 318 + }, + { + "epoch": 0.040832, + "grad_norm": 0.8192832469940186, + "learning_rate": 0.0009996967785983097, + "loss": 2.9319, + "step": 319 + }, + { + "epoch": 0.04096, + "grad_norm": 0.907325029373169, + "learning_rate": 0.0009996895168192955, + "loss": 3.0661, + "step": 320 + }, + { + "epoch": 0.041088, + "grad_norm": 0.8035445809364319, + "learning_rate": 0.0009996821691378602, + "loss": 3.0069, + "step": 321 + }, + { + "epoch": 0.041216, + "grad_norm": 0.7744388580322266, + "learning_rate": 0.0009996747355552674, + "loss": 2.9989, + "step": 322 + }, + { + "epoch": 0.041344, + "grad_norm": 0.8637117147445679, + "learning_rate": 0.000999667216072795, + "loss": 3.0228, + "step": 323 + }, + { + "epoch": 0.041472, + "grad_norm": 0.914459228515625, + "learning_rate": 0.0009996596106917352, + "loss": 3.026, + "step": 324 + }, + { + "epoch": 0.0416, + "grad_norm": 0.8508589267730713, + "learning_rate": 0.000999651919413396, + "loss": 2.9119, + "step": 325 + }, + { + "epoch": 0.041728, + "grad_norm": 0.8922731876373291, + "learning_rate": 0.0009996441422390992, + "loss": 3.0642, + "step": 326 + }, + { + "epoch": 0.041856, + "grad_norm": 0.9900521636009216, + "learning_rate": 0.0009996362791701821, + "loss": 2.9388, + "step": 327 + }, + { + "epoch": 0.041984, + "grad_norm": 0.8944075107574463, + "learning_rate": 0.0009996283302079964, + "loss": 3.042, + "step": 328 + }, + { + "epoch": 0.042112, + "grad_norm": 0.8854784965515137, + "learning_rate": 0.0009996202953539083, + "loss": 3.0592, + "step": 329 + }, + { + "epoch": 0.04224, + "grad_norm": 0.9966174960136414, + "learning_rate": 0.0009996121746092997, + "loss": 3.0979, + "step": 330 + }, + { + "epoch": 0.042368, + "grad_norm": 1.0372589826583862, + "learning_rate": 0.000999603967975566, + "loss": 2.9877, + "step": 331 + }, + { + "epoch": 0.042496, + "grad_norm": 0.9314953684806824, + "learning_rate": 0.0009995956754541182, + "loss": 2.9571, + "step": 332 + }, + { + "epoch": 0.042624, + "grad_norm": 0.8589298129081726, + "learning_rate": 0.0009995872970463822, + "loss": 3.06, + "step": 333 + }, + { + "epoch": 0.042752, + "grad_norm": 0.9840624928474426, + "learning_rate": 0.0009995788327537983, + "loss": 2.9912, + "step": 334 + }, + { + "epoch": 0.04288, + "grad_norm": 1.079399585723877, + "learning_rate": 0.0009995702825778211, + "loss": 2.9613, + "step": 335 + }, + { + "epoch": 0.043008, + "grad_norm": 0.926767885684967, + "learning_rate": 0.0009995616465199209, + "loss": 2.9507, + "step": 336 + }, + { + "epoch": 0.043136, + "grad_norm": 0.7876798510551453, + "learning_rate": 0.0009995529245815822, + "loss": 2.9539, + "step": 337 + }, + { + "epoch": 0.043264, + "grad_norm": 0.8391332030296326, + "learning_rate": 0.0009995441167643046, + "loss": 2.9179, + "step": 338 + }, + { + "epoch": 0.043392, + "grad_norm": 0.8193027377128601, + "learning_rate": 0.000999535223069602, + "loss": 3.0505, + "step": 339 + }, + { + "epoch": 0.04352, + "grad_norm": 0.8525395393371582, + "learning_rate": 0.0009995262434990037, + "loss": 2.9848, + "step": 340 + }, + { + "epoch": 0.043648, + "grad_norm": 0.7669763565063477, + "learning_rate": 0.000999517178054053, + "loss": 3.0447, + "step": 341 + }, + { + "epoch": 0.043776, + "grad_norm": 1.0858415365219116, + "learning_rate": 0.0009995080267363081, + "loss": 2.9945, + "step": 342 + }, + { + "epoch": 0.043904, + "grad_norm": 0.9280194044113159, + "learning_rate": 0.000999498789547343, + "loss": 2.9846, + "step": 343 + }, + { + "epoch": 0.044032, + "grad_norm": 0.8060023784637451, + "learning_rate": 0.0009994894664887453, + "loss": 2.9991, + "step": 344 + }, + { + "epoch": 0.04416, + "grad_norm": 0.814110279083252, + "learning_rate": 0.0009994800575621175, + "loss": 3.0132, + "step": 345 + }, + { + "epoch": 0.044288, + "grad_norm": 0.9309835433959961, + "learning_rate": 0.0009994705627690776, + "loss": 2.9816, + "step": 346 + }, + { + "epoch": 0.044416, + "grad_norm": 0.8402641415596008, + "learning_rate": 0.0009994609821112575, + "loss": 2.9936, + "step": 347 + }, + { + "epoch": 0.044544, + "grad_norm": 0.9190765619277954, + "learning_rate": 0.0009994513155903043, + "loss": 3.027, + "step": 348 + }, + { + "epoch": 0.044672, + "grad_norm": 0.8800088167190552, + "learning_rate": 0.0009994415632078796, + "loss": 2.966, + "step": 349 + }, + { + "epoch": 0.0448, + "grad_norm": 0.9064901471138, + "learning_rate": 0.0009994317249656607, + "loss": 2.9166, + "step": 350 + }, + { + "epoch": 0.044928, + "grad_norm": 0.847995936870575, + "learning_rate": 0.000999421800865338, + "loss": 2.9772, + "step": 351 + }, + { + "epoch": 0.045056, + "grad_norm": 0.8826605081558228, + "learning_rate": 0.0009994117909086179, + "loss": 3.0622, + "step": 352 + }, + { + "epoch": 0.045184, + "grad_norm": 0.8990454077720642, + "learning_rate": 0.0009994016950972213, + "loss": 2.9984, + "step": 353 + }, + { + "epoch": 0.045312, + "grad_norm": 0.8726716637611389, + "learning_rate": 0.000999391513432884, + "loss": 3.0057, + "step": 354 + }, + { + "epoch": 0.04544, + "grad_norm": 0.8427373170852661, + "learning_rate": 0.0009993812459173556, + "loss": 3.0017, + "step": 355 + }, + { + "epoch": 0.045568, + "grad_norm": 0.9117828011512756, + "learning_rate": 0.000999370892552402, + "loss": 2.9501, + "step": 356 + }, + { + "epoch": 0.045696, + "grad_norm": 0.8523821830749512, + "learning_rate": 0.000999360453339803, + "loss": 2.9891, + "step": 357 + }, + { + "epoch": 0.045824, + "grad_norm": 0.8968411684036255, + "learning_rate": 0.0009993499282813527, + "loss": 2.962, + "step": 358 + }, + { + "epoch": 0.045952, + "grad_norm": 1.271944522857666, + "learning_rate": 0.0009993393173788607, + "loss": 3.0918, + "step": 359 + }, + { + "epoch": 0.04608, + "grad_norm": 0.9753352403640747, + "learning_rate": 0.0009993286206341513, + "loss": 3.05, + "step": 360 + }, + { + "epoch": 0.046208, + "grad_norm": 0.9326463937759399, + "learning_rate": 0.0009993178380490635, + "loss": 3.1294, + "step": 361 + }, + { + "epoch": 0.046336, + "grad_norm": 0.8117707967758179, + "learning_rate": 0.0009993069696254506, + "loss": 2.9282, + "step": 362 + }, + { + "epoch": 0.046464, + "grad_norm": 1.3575208187103271, + "learning_rate": 0.0009992960153651811, + "loss": 3.0622, + "step": 363 + }, + { + "epoch": 0.046592, + "grad_norm": 1.1114412546157837, + "learning_rate": 0.0009992849752701383, + "loss": 3.0044, + "step": 364 + }, + { + "epoch": 0.04672, + "grad_norm": 0.8127191066741943, + "learning_rate": 0.00099927384934222, + "loss": 2.9507, + "step": 365 + }, + { + "epoch": 0.046848, + "grad_norm": 0.8835131525993347, + "learning_rate": 0.0009992626375833389, + "loss": 2.9597, + "step": 366 + }, + { + "epoch": 0.046976, + "grad_norm": 0.8865495920181274, + "learning_rate": 0.0009992513399954225, + "loss": 2.998, + "step": 367 + }, + { + "epoch": 0.047104, + "grad_norm": 0.9364601373672485, + "learning_rate": 0.0009992399565804128, + "loss": 2.8971, + "step": 368 + }, + { + "epoch": 0.047232, + "grad_norm": 0.87732994556427, + "learning_rate": 0.000999228487340267, + "loss": 2.9034, + "step": 369 + }, + { + "epoch": 0.04736, + "grad_norm": 0.9213501214981079, + "learning_rate": 0.0009992169322769566, + "loss": 3.0737, + "step": 370 + }, + { + "epoch": 0.047488, + "grad_norm": 0.9006425738334656, + "learning_rate": 0.0009992052913924683, + "loss": 3.0688, + "step": 371 + }, + { + "epoch": 0.047616, + "grad_norm": 0.8741381764411926, + "learning_rate": 0.000999193564688803, + "loss": 2.9938, + "step": 372 + }, + { + "epoch": 0.047744, + "grad_norm": 0.8652103543281555, + "learning_rate": 0.0009991817521679768, + "loss": 3.0678, + "step": 373 + }, + { + "epoch": 0.047872, + "grad_norm": 0.9090451598167419, + "learning_rate": 0.0009991698538320203, + "loss": 3.0685, + "step": 374 + }, + { + "epoch": 0.048, + "grad_norm": 0.8936727643013, + "learning_rate": 0.0009991578696829794, + "loss": 2.91, + "step": 375 + }, + { + "epoch": 0.048128, + "grad_norm": 0.8842224478721619, + "learning_rate": 0.0009991457997229134, + "loss": 3.0261, + "step": 376 + }, + { + "epoch": 0.048256, + "grad_norm": 0.8420811295509338, + "learning_rate": 0.0009991336439538984, + "loss": 2.9911, + "step": 377 + }, + { + "epoch": 0.048384, + "grad_norm": 0.8568455576896667, + "learning_rate": 0.000999121402378023, + "loss": 2.9518, + "step": 378 + }, + { + "epoch": 0.048512, + "grad_norm": 0.8723102807998657, + "learning_rate": 0.0009991090749973926, + "loss": 2.9701, + "step": 379 + }, + { + "epoch": 0.04864, + "grad_norm": 0.8671232461929321, + "learning_rate": 0.000999096661814126, + "loss": 3.0896, + "step": 380 + }, + { + "epoch": 0.048768, + "grad_norm": 0.8732597827911377, + "learning_rate": 0.0009990841628303572, + "loss": 2.9999, + "step": 381 + }, + { + "epoch": 0.048896, + "grad_norm": 0.7922152280807495, + "learning_rate": 0.0009990715780482347, + "loss": 2.9183, + "step": 382 + }, + { + "epoch": 0.049024, + "grad_norm": 0.9291329383850098, + "learning_rate": 0.0009990589074699225, + "loss": 2.9828, + "step": 383 + }, + { + "epoch": 0.049152, + "grad_norm": 0.8204339742660522, + "learning_rate": 0.0009990461510975984, + "loss": 3.0078, + "step": 384 + }, + { + "epoch": 0.04928, + "grad_norm": 0.8787654638290405, + "learning_rate": 0.0009990333089334555, + "loss": 2.9734, + "step": 385 + }, + { + "epoch": 0.049408, + "grad_norm": 0.8029327392578125, + "learning_rate": 0.0009990203809797013, + "loss": 2.9345, + "step": 386 + }, + { + "epoch": 0.049536, + "grad_norm": 0.8262515664100647, + "learning_rate": 0.0009990073672385587, + "loss": 2.9434, + "step": 387 + }, + { + "epoch": 0.049664, + "grad_norm": 0.991312563419342, + "learning_rate": 0.0009989942677122647, + "loss": 2.9268, + "step": 388 + }, + { + "epoch": 0.049792, + "grad_norm": 0.8821982145309448, + "learning_rate": 0.0009989810824030713, + "loss": 2.9826, + "step": 389 + }, + { + "epoch": 0.04992, + "grad_norm": 0.8759288191795349, + "learning_rate": 0.000998967811313245, + "loss": 3.0323, + "step": 390 + }, + { + "epoch": 0.050048, + "grad_norm": 0.8573916554450989, + "learning_rate": 0.0009989544544450673, + "loss": 3.0054, + "step": 391 + }, + { + "epoch": 0.050176, + "grad_norm": 0.857089638710022, + "learning_rate": 0.0009989410118008348, + "loss": 2.8532, + "step": 392 + }, + { + "epoch": 0.050304, + "grad_norm": 0.8681499361991882, + "learning_rate": 0.000998927483382858, + "loss": 2.9465, + "step": 393 + }, + { + "epoch": 0.050432, + "grad_norm": 0.9481124877929688, + "learning_rate": 0.0009989138691934628, + "loss": 2.9855, + "step": 394 + }, + { + "epoch": 0.05056, + "grad_norm": 0.9023472666740417, + "learning_rate": 0.0009989001692349894, + "loss": 3.0722, + "step": 395 + }, + { + "epoch": 0.050688, + "grad_norm": 0.8562093377113342, + "learning_rate": 0.0009988863835097932, + "loss": 2.8906, + "step": 396 + }, + { + "epoch": 0.050816, + "grad_norm": 0.8540483117103577, + "learning_rate": 0.0009988725120202442, + "loss": 2.9284, + "step": 397 + }, + { + "epoch": 0.050944, + "grad_norm": 0.9885972142219543, + "learning_rate": 0.000998858554768727, + "loss": 2.8781, + "step": 398 + }, + { + "epoch": 0.051072, + "grad_norm": 0.8391234278678894, + "learning_rate": 0.0009988445117576409, + "loss": 3.0074, + "step": 399 + }, + { + "epoch": 0.0512, + "grad_norm": 0.8503037095069885, + "learning_rate": 0.0009988303829894002, + "loss": 2.8947, + "step": 400 + }, + { + "epoch": 0.051328, + "grad_norm": 0.9345840811729431, + "learning_rate": 0.0009988161684664335, + "loss": 3.0029, + "step": 401 + }, + { + "epoch": 0.051456, + "grad_norm": 0.8529277443885803, + "learning_rate": 0.0009988018681911849, + "loss": 3.012, + "step": 402 + }, + { + "epoch": 0.051584, + "grad_norm": 0.7660846710205078, + "learning_rate": 0.0009987874821661123, + "loss": 2.9069, + "step": 403 + }, + { + "epoch": 0.051712, + "grad_norm": 0.8276880383491516, + "learning_rate": 0.0009987730103936895, + "loss": 2.9685, + "step": 404 + }, + { + "epoch": 0.05184, + "grad_norm": 0.8381699323654175, + "learning_rate": 0.0009987584528764035, + "loss": 2.9186, + "step": 405 + }, + { + "epoch": 0.051968, + "grad_norm": 0.9605251550674438, + "learning_rate": 0.0009987438096167576, + "loss": 2.9357, + "step": 406 + }, + { + "epoch": 0.052096, + "grad_norm": 0.9230712056159973, + "learning_rate": 0.0009987290806172688, + "loss": 2.9855, + "step": 407 + }, + { + "epoch": 0.052224, + "grad_norm": 0.8335307836532593, + "learning_rate": 0.0009987142658804693, + "loss": 3.0306, + "step": 408 + }, + { + "epoch": 0.052352, + "grad_norm": 0.9176110625267029, + "learning_rate": 0.0009986993654089059, + "loss": 2.9599, + "step": 409 + }, + { + "epoch": 0.05248, + "grad_norm": 0.7880818247795105, + "learning_rate": 0.00099868437920514, + "loss": 3.0114, + "step": 410 + }, + { + "epoch": 0.052608, + "grad_norm": 0.8253980875015259, + "learning_rate": 0.0009986693072717484, + "loss": 2.9577, + "step": 411 + }, + { + "epoch": 0.052736, + "grad_norm": 0.7556925415992737, + "learning_rate": 0.0009986541496113215, + "loss": 2.994, + "step": 412 + }, + { + "epoch": 0.052864, + "grad_norm": 0.8475455045700073, + "learning_rate": 0.0009986389062264657, + "loss": 2.9708, + "step": 413 + }, + { + "epoch": 0.052992, + "grad_norm": 0.8585573434829712, + "learning_rate": 0.0009986235771198009, + "loss": 2.9039, + "step": 414 + }, + { + "epoch": 0.05312, + "grad_norm": 0.8316100239753723, + "learning_rate": 0.000998608162293963, + "loss": 2.9046, + "step": 415 + }, + { + "epoch": 0.053248, + "grad_norm": 0.9127733111381531, + "learning_rate": 0.0009985926617516012, + "loss": 3.0002, + "step": 416 + }, + { + "epoch": 0.053376, + "grad_norm": 0.9314609169960022, + "learning_rate": 0.0009985770754953808, + "loss": 2.9791, + "step": 417 + }, + { + "epoch": 0.053504, + "grad_norm": 0.9396356344223022, + "learning_rate": 0.0009985614035279812, + "loss": 2.8337, + "step": 418 + }, + { + "epoch": 0.053632, + "grad_norm": 0.86746746301651, + "learning_rate": 0.0009985456458520968, + "loss": 2.9744, + "step": 419 + }, + { + "epoch": 0.05376, + "grad_norm": 0.8200275897979736, + "learning_rate": 0.000998529802470436, + "loss": 2.9296, + "step": 420 + }, + { + "epoch": 0.053888, + "grad_norm": 1.2117091417312622, + "learning_rate": 0.0009985138733857224, + "loss": 2.8779, + "step": 421 + }, + { + "epoch": 0.054016, + "grad_norm": 0.8219243288040161, + "learning_rate": 0.000998497858600695, + "loss": 2.9163, + "step": 422 + }, + { + "epoch": 0.054144, + "grad_norm": 1.014870285987854, + "learning_rate": 0.0009984817581181068, + "loss": 2.9808, + "step": 423 + }, + { + "epoch": 0.054272, + "grad_norm": 0.927848756313324, + "learning_rate": 0.0009984655719407253, + "loss": 2.9895, + "step": 424 + }, + { + "epoch": 0.0544, + "grad_norm": 0.8441425561904907, + "learning_rate": 0.000998449300071333, + "loss": 2.8933, + "step": 425 + }, + { + "epoch": 0.054528, + "grad_norm": 0.7969479560852051, + "learning_rate": 0.0009984329425127279, + "loss": 2.9476, + "step": 426 + }, + { + "epoch": 0.054656, + "grad_norm": 0.8277332186698914, + "learning_rate": 0.0009984164992677214, + "loss": 2.9638, + "step": 427 + }, + { + "epoch": 0.054784, + "grad_norm": 0.8566104769706726, + "learning_rate": 0.0009983999703391407, + "loss": 2.8862, + "step": 428 + }, + { + "epoch": 0.054912, + "grad_norm": 0.8398438096046448, + "learning_rate": 0.000998383355729827, + "loss": 2.9274, + "step": 429 + }, + { + "epoch": 0.05504, + "grad_norm": 0.9653627872467041, + "learning_rate": 0.0009983666554426367, + "loss": 3.0354, + "step": 430 + }, + { + "epoch": 0.055168, + "grad_norm": 0.8621662259101868, + "learning_rate": 0.000998349869480441, + "loss": 2.9898, + "step": 431 + }, + { + "epoch": 0.055296, + "grad_norm": 0.932272732257843, + "learning_rate": 0.0009983329978461252, + "loss": 3.0216, + "step": 432 + }, + { + "epoch": 0.055424, + "grad_norm": 0.8948084712028503, + "learning_rate": 0.0009983160405425899, + "loss": 2.9169, + "step": 433 + }, + { + "epoch": 0.055552, + "grad_norm": 0.8620967268943787, + "learning_rate": 0.0009982989975727502, + "loss": 2.9682, + "step": 434 + }, + { + "epoch": 0.05568, + "grad_norm": 0.8760973811149597, + "learning_rate": 0.0009982818689395362, + "loss": 2.9727, + "step": 435 + }, + { + "epoch": 0.055808, + "grad_norm": 0.863288402557373, + "learning_rate": 0.0009982646546458922, + "loss": 2.7776, + "step": 436 + }, + { + "epoch": 0.055936, + "grad_norm": 0.8769680261611938, + "learning_rate": 0.0009982473546947777, + "loss": 2.91, + "step": 437 + }, + { + "epoch": 0.056064, + "grad_norm": 0.8858181834220886, + "learning_rate": 0.0009982299690891668, + "loss": 2.8306, + "step": 438 + }, + { + "epoch": 0.056192, + "grad_norm": 1.0747575759887695, + "learning_rate": 0.0009982124978320482, + "loss": 2.961, + "step": 439 + }, + { + "epoch": 0.05632, + "grad_norm": 0.8561440706253052, + "learning_rate": 0.0009981949409264255, + "loss": 2.9801, + "step": 440 + }, + { + "epoch": 0.056448, + "grad_norm": 0.8666918873786926, + "learning_rate": 0.0009981772983753167, + "loss": 2.9768, + "step": 441 + }, + { + "epoch": 0.056576, + "grad_norm": 0.8439273238182068, + "learning_rate": 0.000998159570181755, + "loss": 2.9625, + "step": 442 + }, + { + "epoch": 0.056704, + "grad_norm": 1.2183685302734375, + "learning_rate": 0.0009981417563487884, + "loss": 2.923, + "step": 443 + }, + { + "epoch": 0.056832, + "grad_norm": 0.7799296379089355, + "learning_rate": 0.0009981238568794787, + "loss": 2.9746, + "step": 444 + }, + { + "epoch": 0.05696, + "grad_norm": 0.8253427743911743, + "learning_rate": 0.000998105871776903, + "loss": 2.9432, + "step": 445 + }, + { + "epoch": 0.057088, + "grad_norm": 0.8542295694351196, + "learning_rate": 0.0009980878010441537, + "loss": 2.9697, + "step": 446 + }, + { + "epoch": 0.057216, + "grad_norm": 0.8216561675071716, + "learning_rate": 0.000998069644684337, + "loss": 2.8627, + "step": 447 + }, + { + "epoch": 0.057344, + "grad_norm": 0.8315777778625488, + "learning_rate": 0.0009980514027005743, + "loss": 2.9353, + "step": 448 + }, + { + "epoch": 0.057472, + "grad_norm": 0.7814035415649414, + "learning_rate": 0.0009980330750960014, + "loss": 2.91, + "step": 449 + }, + { + "epoch": 0.0576, + "grad_norm": 0.80046546459198, + "learning_rate": 0.0009980146618737693, + "loss": 2.942, + "step": 450 + }, + { + "epoch": 0.057728, + "grad_norm": 0.808443009853363, + "learning_rate": 0.0009979961630370434, + "loss": 2.9916, + "step": 451 + }, + { + "epoch": 0.057856, + "grad_norm": 0.8269020915031433, + "learning_rate": 0.0009979775785890037, + "loss": 2.9476, + "step": 452 + }, + { + "epoch": 0.057984, + "grad_norm": 0.8049909472465515, + "learning_rate": 0.0009979589085328453, + "loss": 2.9747, + "step": 453 + }, + { + "epoch": 0.058112, + "grad_norm": 0.8847025036811829, + "learning_rate": 0.0009979401528717776, + "loss": 2.9406, + "step": 454 + }, + { + "epoch": 0.05824, + "grad_norm": 0.9976103901863098, + "learning_rate": 0.0009979213116090252, + "loss": 2.8871, + "step": 455 + }, + { + "epoch": 0.058368, + "grad_norm": 0.8739810585975647, + "learning_rate": 0.0009979023847478266, + "loss": 2.8757, + "step": 456 + }, + { + "epoch": 0.058496, + "grad_norm": 0.8689247965812683, + "learning_rate": 0.0009978833722914362, + "loss": 3.0059, + "step": 457 + }, + { + "epoch": 0.058624, + "grad_norm": 0.8380698561668396, + "learning_rate": 0.000997864274243122, + "loss": 2.9699, + "step": 458 + }, + { + "epoch": 0.058752, + "grad_norm": 0.837498128414154, + "learning_rate": 0.0009978450906061673, + "loss": 2.8966, + "step": 459 + }, + { + "epoch": 0.05888, + "grad_norm": 0.8527972102165222, + "learning_rate": 0.00099782582138387, + "loss": 2.9648, + "step": 460 + }, + { + "epoch": 0.059008, + "grad_norm": 0.9263566136360168, + "learning_rate": 0.0009978064665795428, + "loss": 2.9747, + "step": 461 + }, + { + "epoch": 0.059136, + "grad_norm": 0.790797233581543, + "learning_rate": 0.0009977870261965128, + "loss": 2.9059, + "step": 462 + }, + { + "epoch": 0.059264, + "grad_norm": 0.9020636081695557, + "learning_rate": 0.000997767500238122, + "loss": 2.8904, + "step": 463 + }, + { + "epoch": 0.059392, + "grad_norm": 0.8688124418258667, + "learning_rate": 0.0009977478887077276, + "loss": 2.9342, + "step": 464 + }, + { + "epoch": 0.05952, + "grad_norm": 0.9296693801879883, + "learning_rate": 0.0009977281916087007, + "loss": 2.8909, + "step": 465 + }, + { + "epoch": 0.059648, + "grad_norm": 0.881331741809845, + "learning_rate": 0.0009977084089444275, + "loss": 2.9117, + "step": 466 + }, + { + "epoch": 0.059776, + "grad_norm": 0.8821720480918884, + "learning_rate": 0.000997688540718309, + "loss": 2.8713, + "step": 467 + }, + { + "epoch": 0.059904, + "grad_norm": 0.8100451827049255, + "learning_rate": 0.0009976685869337602, + "loss": 2.9473, + "step": 468 + }, + { + "epoch": 0.060032, + "grad_norm": 0.7819488644599915, + "learning_rate": 0.0009976485475942122, + "loss": 2.9267, + "step": 469 + }, + { + "epoch": 0.06016, + "grad_norm": 0.7937043905258179, + "learning_rate": 0.0009976284227031097, + "loss": 2.9057, + "step": 470 + }, + { + "epoch": 0.060288, + "grad_norm": 0.7939175963401794, + "learning_rate": 0.000997608212263912, + "loss": 2.8957, + "step": 471 + }, + { + "epoch": 0.060416, + "grad_norm": 0.8964297771453857, + "learning_rate": 0.000997587916280094, + "loss": 2.947, + "step": 472 + }, + { + "epoch": 0.060544, + "grad_norm": 0.7535595893859863, + "learning_rate": 0.0009975675347551447, + "loss": 2.9387, + "step": 473 + }, + { + "epoch": 0.060672, + "grad_norm": 0.8702789545059204, + "learning_rate": 0.000997547067692568, + "loss": 2.9145, + "step": 474 + }, + { + "epoch": 0.0608, + "grad_norm": 0.8363080620765686, + "learning_rate": 0.000997526515095882, + "loss": 2.9209, + "step": 475 + }, + { + "epoch": 0.060928, + "grad_norm": 0.771239161491394, + "learning_rate": 0.0009975058769686204, + "loss": 2.8419, + "step": 476 + }, + { + "epoch": 0.061056, + "grad_norm": 0.8436124920845032, + "learning_rate": 0.000997485153314331, + "loss": 2.9084, + "step": 477 + }, + { + "epoch": 0.061184, + "grad_norm": 0.8249334096908569, + "learning_rate": 0.0009974643441365764, + "loss": 2.893, + "step": 478 + }, + { + "epoch": 0.061312, + "grad_norm": 0.8912394046783447, + "learning_rate": 0.000997443449438934, + "loss": 2.9943, + "step": 479 + }, + { + "epoch": 0.06144, + "grad_norm": 0.9046705961227417, + "learning_rate": 0.0009974224692249956, + "loss": 2.8861, + "step": 480 + }, + { + "epoch": 0.061568, + "grad_norm": 0.7414267659187317, + "learning_rate": 0.0009974014034983684, + "loss": 2.8076, + "step": 481 + }, + { + "epoch": 0.061696, + "grad_norm": 0.8210133910179138, + "learning_rate": 0.0009973802522626732, + "loss": 2.9479, + "step": 482 + }, + { + "epoch": 0.061824, + "grad_norm": 0.7954651117324829, + "learning_rate": 0.0009973590155215467, + "loss": 2.9551, + "step": 483 + }, + { + "epoch": 0.061952, + "grad_norm": 0.7361330389976501, + "learning_rate": 0.0009973376932786396, + "loss": 2.7989, + "step": 484 + }, + { + "epoch": 0.06208, + "grad_norm": 0.8215603828430176, + "learning_rate": 0.0009973162855376172, + "loss": 2.9658, + "step": 485 + }, + { + "epoch": 0.062208, + "grad_norm": 0.9567349553108215, + "learning_rate": 0.0009972947923021602, + "loss": 2.9256, + "step": 486 + }, + { + "epoch": 0.062336, + "grad_norm": 0.7475981116294861, + "learning_rate": 0.000997273213575963, + "loss": 3.0095, + "step": 487 + }, + { + "epoch": 0.062464, + "grad_norm": 0.8618100881576538, + "learning_rate": 0.0009972515493627356, + "loss": 2.9352, + "step": 488 + }, + { + "epoch": 0.062592, + "grad_norm": 0.7602435350418091, + "learning_rate": 0.0009972297996662023, + "loss": 2.8602, + "step": 489 + }, + { + "epoch": 0.06272, + "grad_norm": 0.8116627335548401, + "learning_rate": 0.0009972079644901019, + "loss": 2.9571, + "step": 490 + }, + { + "epoch": 0.062848, + "grad_norm": 0.9534598588943481, + "learning_rate": 0.0009971860438381885, + "loss": 2.9747, + "step": 491 + }, + { + "epoch": 0.062976, + "grad_norm": 0.8760648965835571, + "learning_rate": 0.0009971640377142302, + "loss": 2.837, + "step": 492 + }, + { + "epoch": 0.063104, + "grad_norm": 0.8705786466598511, + "learning_rate": 0.0009971419461220102, + "loss": 2.954, + "step": 493 + }, + { + "epoch": 0.063232, + "grad_norm": 0.8618326783180237, + "learning_rate": 0.0009971197690653263, + "loss": 2.8854, + "step": 494 + }, + { + "epoch": 0.06336, + "grad_norm": 0.8591819405555725, + "learning_rate": 0.000997097506547991, + "loss": 2.9361, + "step": 495 + }, + { + "epoch": 0.063488, + "grad_norm": 0.8599340319633484, + "learning_rate": 0.0009970751585738314, + "loss": 2.9133, + "step": 496 + }, + { + "epoch": 0.063616, + "grad_norm": 0.8033486008644104, + "learning_rate": 0.0009970527251466894, + "loss": 2.8282, + "step": 497 + }, + { + "epoch": 0.063744, + "grad_norm": 0.8381534814834595, + "learning_rate": 0.0009970302062704216, + "loss": 2.958, + "step": 498 + }, + { + "epoch": 0.063872, + "grad_norm": 0.8161300420761108, + "learning_rate": 0.0009970076019488996, + "loss": 2.949, + "step": 499 + }, + { + "epoch": 0.064, + "grad_norm": 0.8745964169502258, + "learning_rate": 0.0009969849121860088, + "loss": 2.9166, + "step": 500 + }, + { + "epoch": 0.064128, + "grad_norm": 1.519952416419983, + "learning_rate": 0.0009969621369856501, + "loss": 2.9837, + "step": 501 + }, + { + "epoch": 0.064256, + "grad_norm": 0.8925809264183044, + "learning_rate": 0.0009969392763517387, + "loss": 2.9434, + "step": 502 + }, + { + "epoch": 0.064384, + "grad_norm": 0.8160974979400635, + "learning_rate": 0.0009969163302882048, + "loss": 2.9056, + "step": 503 + }, + { + "epoch": 0.064512, + "grad_norm": 0.8990865349769592, + "learning_rate": 0.0009968932987989927, + "loss": 2.9486, + "step": 504 + }, + { + "epoch": 0.06464, + "grad_norm": 0.8527535796165466, + "learning_rate": 0.0009968701818880623, + "loss": 2.7934, + "step": 505 + }, + { + "epoch": 0.064768, + "grad_norm": 0.9045835733413696, + "learning_rate": 0.0009968469795593874, + "loss": 3.002, + "step": 506 + }, + { + "epoch": 0.064896, + "grad_norm": 0.7804040312767029, + "learning_rate": 0.0009968236918169566, + "loss": 2.9604, + "step": 507 + }, + { + "epoch": 0.065024, + "grad_norm": 0.9197160601615906, + "learning_rate": 0.0009968003186647735, + "loss": 2.8559, + "step": 508 + }, + { + "epoch": 0.065152, + "grad_norm": 0.8895035982131958, + "learning_rate": 0.0009967768601068566, + "loss": 2.8417, + "step": 509 + }, + { + "epoch": 0.06528, + "grad_norm": 0.7669428586959839, + "learning_rate": 0.000996753316147238, + "loss": 2.7728, + "step": 510 + }, + { + "epoch": 0.065408, + "grad_norm": 0.7975584864616394, + "learning_rate": 0.0009967296867899656, + "loss": 2.8815, + "step": 511 + }, + { + "epoch": 0.065536, + "grad_norm": 0.8407531380653381, + "learning_rate": 0.0009967059720391014, + "loss": 2.9233, + "step": 512 + }, + { + "epoch": 0.065664, + "grad_norm": 0.6987216472625732, + "learning_rate": 0.0009966821718987223, + "loss": 2.8908, + "step": 513 + }, + { + "epoch": 0.065792, + "grad_norm": 0.835658848285675, + "learning_rate": 0.0009966582863729198, + "loss": 2.8729, + "step": 514 + }, + { + "epoch": 0.06592, + "grad_norm": 0.8763951063156128, + "learning_rate": 0.0009966343154658001, + "loss": 2.8365, + "step": 515 + }, + { + "epoch": 0.066048, + "grad_norm": 0.8309316635131836, + "learning_rate": 0.0009966102591814842, + "loss": 3.0225, + "step": 516 + }, + { + "epoch": 0.066176, + "grad_norm": 0.7884944081306458, + "learning_rate": 0.0009965861175241075, + "loss": 2.8817, + "step": 517 + }, + { + "epoch": 0.066304, + "grad_norm": 0.8770759701728821, + "learning_rate": 0.0009965618904978202, + "loss": 2.9296, + "step": 518 + }, + { + "epoch": 0.066432, + "grad_norm": 0.7792795300483704, + "learning_rate": 0.0009965375781067873, + "loss": 2.8819, + "step": 519 + }, + { + "epoch": 0.06656, + "grad_norm": 0.9122428297996521, + "learning_rate": 0.0009965131803551886, + "loss": 3.0358, + "step": 520 + }, + { + "epoch": 0.066688, + "grad_norm": 0.8287910223007202, + "learning_rate": 0.0009964886972472179, + "loss": 2.893, + "step": 521 + }, + { + "epoch": 0.066816, + "grad_norm": 0.8577980399131775, + "learning_rate": 0.0009964641287870842, + "loss": 2.9793, + "step": 522 + }, + { + "epoch": 0.066944, + "grad_norm": 0.797559380531311, + "learning_rate": 0.0009964394749790116, + "loss": 2.8914, + "step": 523 + }, + { + "epoch": 0.067072, + "grad_norm": 0.7894203662872314, + "learning_rate": 0.0009964147358272378, + "loss": 2.907, + "step": 524 + }, + { + "epoch": 0.0672, + "grad_norm": 0.9896024465560913, + "learning_rate": 0.000996389911336016, + "loss": 2.8875, + "step": 525 + }, + { + "epoch": 0.067328, + "grad_norm": 0.7524051070213318, + "learning_rate": 0.000996365001509614, + "loss": 2.9047, + "step": 526 + }, + { + "epoch": 0.067456, + "grad_norm": 0.8152078986167908, + "learning_rate": 0.0009963400063523135, + "loss": 2.8397, + "step": 527 + }, + { + "epoch": 0.067584, + "grad_norm": 0.822388231754303, + "learning_rate": 0.0009963149258684122, + "loss": 2.9451, + "step": 528 + }, + { + "epoch": 0.067712, + "grad_norm": 0.8943387866020203, + "learning_rate": 0.000996289760062221, + "loss": 2.9311, + "step": 529 + }, + { + "epoch": 0.06784, + "grad_norm": 0.859833836555481, + "learning_rate": 0.000996264508938067, + "loss": 2.8859, + "step": 530 + }, + { + "epoch": 0.067968, + "grad_norm": 0.8182249665260315, + "learning_rate": 0.0009962391725002904, + "loss": 2.8287, + "step": 531 + }, + { + "epoch": 0.068096, + "grad_norm": 0.7720870971679688, + "learning_rate": 0.0009962137507532475, + "loss": 2.9438, + "step": 532 + }, + { + "epoch": 0.068224, + "grad_norm": 0.8875404000282288, + "learning_rate": 0.000996188243701308, + "loss": 2.873, + "step": 533 + }, + { + "epoch": 0.068352, + "grad_norm": 0.8407955169677734, + "learning_rate": 0.000996162651348857, + "loss": 2.8709, + "step": 534 + }, + { + "epoch": 0.06848, + "grad_norm": 0.8966158032417297, + "learning_rate": 0.0009961369737002943, + "loss": 2.843, + "step": 535 + }, + { + "epoch": 0.068608, + "grad_norm": 0.8447369933128357, + "learning_rate": 0.0009961112107600343, + "loss": 2.9387, + "step": 536 + }, + { + "epoch": 0.068736, + "grad_norm": 0.7806838750839233, + "learning_rate": 0.0009960853625325054, + "loss": 2.7775, + "step": 537 + }, + { + "epoch": 0.068864, + "grad_norm": 0.806995689868927, + "learning_rate": 0.0009960594290221519, + "loss": 2.8192, + "step": 538 + }, + { + "epoch": 0.068992, + "grad_norm": 0.776648223400116, + "learning_rate": 0.0009960334102334315, + "loss": 2.99, + "step": 539 + }, + { + "epoch": 0.06912, + "grad_norm": 0.8101557493209839, + "learning_rate": 0.0009960073061708176, + "loss": 2.8417, + "step": 540 + }, + { + "epoch": 0.069248, + "grad_norm": 0.7877004742622375, + "learning_rate": 0.0009959811168387973, + "loss": 2.9435, + "step": 541 + }, + { + "epoch": 0.069376, + "grad_norm": 0.8213893175125122, + "learning_rate": 0.0009959548422418733, + "loss": 2.8216, + "step": 542 + }, + { + "epoch": 0.069504, + "grad_norm": 0.8242461681365967, + "learning_rate": 0.0009959284823845623, + "loss": 2.8313, + "step": 543 + }, + { + "epoch": 0.069632, + "grad_norm": 0.8038188219070435, + "learning_rate": 0.0009959020372713957, + "loss": 2.897, + "step": 544 + }, + { + "epoch": 0.06976, + "grad_norm": 0.7908174395561218, + "learning_rate": 0.00099587550690692, + "loss": 2.8991, + "step": 545 + }, + { + "epoch": 0.069888, + "grad_norm": 0.8224116563796997, + "learning_rate": 0.000995848891295696, + "loss": 2.9543, + "step": 546 + }, + { + "epoch": 0.070016, + "grad_norm": 0.7777135372161865, + "learning_rate": 0.0009958221904422992, + "loss": 2.8846, + "step": 547 + }, + { + "epoch": 0.070144, + "grad_norm": 0.7755223512649536, + "learning_rate": 0.00099579540435132, + "loss": 2.7546, + "step": 548 + }, + { + "epoch": 0.070272, + "grad_norm": 1.1651290655136108, + "learning_rate": 0.0009957685330273629, + "loss": 2.8372, + "step": 549 + }, + { + "epoch": 0.0704, + "grad_norm": 0.7909471392631531, + "learning_rate": 0.0009957415764750475, + "loss": 2.9303, + "step": 550 + }, + { + "epoch": 0.070528, + "grad_norm": 0.7924452424049377, + "learning_rate": 0.000995714534699008, + "loss": 2.9104, + "step": 551 + }, + { + "epoch": 0.070656, + "grad_norm": 0.856462836265564, + "learning_rate": 0.000995687407703893, + "loss": 2.8913, + "step": 552 + }, + { + "epoch": 0.070784, + "grad_norm": 0.8432139158248901, + "learning_rate": 0.0009956601954943663, + "loss": 3.0183, + "step": 553 + }, + { + "epoch": 0.070912, + "grad_norm": 0.8769915103912354, + "learning_rate": 0.0009956328980751058, + "loss": 2.9622, + "step": 554 + }, + { + "epoch": 0.07104, + "grad_norm": 0.7732940912246704, + "learning_rate": 0.000995605515450804, + "loss": 2.89, + "step": 555 + }, + { + "epoch": 0.071168, + "grad_norm": 0.7671625018119812, + "learning_rate": 0.0009955780476261689, + "loss": 2.9004, + "step": 556 + }, + { + "epoch": 0.071296, + "grad_norm": 0.7472454309463501, + "learning_rate": 0.000995550494605922, + "loss": 2.8877, + "step": 557 + }, + { + "epoch": 0.071424, + "grad_norm": 0.7935909032821655, + "learning_rate": 0.0009955228563948003, + "loss": 2.9286, + "step": 558 + }, + { + "epoch": 0.071552, + "grad_norm": 0.8747076392173767, + "learning_rate": 0.0009954951329975547, + "loss": 2.8559, + "step": 559 + }, + { + "epoch": 0.07168, + "grad_norm": 0.7402902841567993, + "learning_rate": 0.0009954673244189519, + "loss": 2.8588, + "step": 560 + }, + { + "epoch": 0.071808, + "grad_norm": 0.7989487648010254, + "learning_rate": 0.0009954394306637717, + "loss": 2.8495, + "step": 561 + }, + { + "epoch": 0.071936, + "grad_norm": 0.8087185621261597, + "learning_rate": 0.00099541145173681, + "loss": 2.8722, + "step": 562 + }, + { + "epoch": 0.072064, + "grad_norm": 0.7866984009742737, + "learning_rate": 0.0009953833876428762, + "loss": 2.8857, + "step": 563 + }, + { + "epoch": 0.072192, + "grad_norm": 0.7609816789627075, + "learning_rate": 0.0009953552383867952, + "loss": 2.8663, + "step": 564 + }, + { + "epoch": 0.07232, + "grad_norm": 0.811732828617096, + "learning_rate": 0.000995327003973406, + "loss": 2.8569, + "step": 565 + }, + { + "epoch": 0.072448, + "grad_norm": 0.7933493852615356, + "learning_rate": 0.0009952986844075627, + "loss": 2.8341, + "step": 566 + }, + { + "epoch": 0.072576, + "grad_norm": 0.8152285218238831, + "learning_rate": 0.0009952702796941334, + "loss": 2.8615, + "step": 567 + }, + { + "epoch": 0.072704, + "grad_norm": 0.8959524631500244, + "learning_rate": 0.0009952417898380014, + "loss": 2.8771, + "step": 568 + }, + { + "epoch": 0.072832, + "grad_norm": 0.856249988079071, + "learning_rate": 0.0009952132148440644, + "loss": 3.0105, + "step": 569 + }, + { + "epoch": 0.07296, + "grad_norm": 0.8343120813369751, + "learning_rate": 0.0009951845547172347, + "loss": 2.8764, + "step": 570 + }, + { + "epoch": 0.073088, + "grad_norm": 0.7916964888572693, + "learning_rate": 0.0009951558094624395, + "loss": 2.8414, + "step": 571 + }, + { + "epoch": 0.073216, + "grad_norm": 0.9378660321235657, + "learning_rate": 0.0009951269790846203, + "loss": 2.908, + "step": 572 + }, + { + "epoch": 0.073344, + "grad_norm": 0.7986360788345337, + "learning_rate": 0.0009950980635887333, + "loss": 2.9188, + "step": 573 + }, + { + "epoch": 0.073472, + "grad_norm": 0.8264856338500977, + "learning_rate": 0.0009950690629797494, + "loss": 2.8016, + "step": 574 + }, + { + "epoch": 0.0736, + "grad_norm": 0.8270865082740784, + "learning_rate": 0.0009950399772626543, + "loss": 2.8371, + "step": 575 + }, + { + "epoch": 0.073728, + "grad_norm": 0.9141223430633545, + "learning_rate": 0.0009950108064424481, + "loss": 2.8804, + "step": 576 + }, + { + "epoch": 0.073856, + "grad_norm": 0.7859342694282532, + "learning_rate": 0.0009949815505241458, + "loss": 2.8972, + "step": 577 + }, + { + "epoch": 0.073984, + "grad_norm": 0.7929102182388306, + "learning_rate": 0.0009949522095127766, + "loss": 2.7401, + "step": 578 + }, + { + "epoch": 0.074112, + "grad_norm": 0.7637389302253723, + "learning_rate": 0.0009949227834133846, + "loss": 2.9515, + "step": 579 + }, + { + "epoch": 0.07424, + "grad_norm": 0.7954878211021423, + "learning_rate": 0.0009948932722310283, + "loss": 2.9101, + "step": 580 + }, + { + "epoch": 0.074368, + "grad_norm": 0.7722182273864746, + "learning_rate": 0.0009948636759707814, + "loss": 2.9565, + "step": 581 + }, + { + "epoch": 0.074496, + "grad_norm": 0.8248521685600281, + "learning_rate": 0.0009948339946377316, + "loss": 2.8683, + "step": 582 + }, + { + "epoch": 0.074624, + "grad_norm": 0.7313053607940674, + "learning_rate": 0.0009948042282369816, + "loss": 2.9094, + "step": 583 + }, + { + "epoch": 0.074752, + "grad_norm": 0.8378648161888123, + "learning_rate": 0.0009947743767736484, + "loss": 2.8072, + "step": 584 + }, + { + "epoch": 0.07488, + "grad_norm": 0.7735212445259094, + "learning_rate": 0.0009947444402528638, + "loss": 2.9109, + "step": 585 + }, + { + "epoch": 0.075008, + "grad_norm": 0.7970257997512817, + "learning_rate": 0.0009947144186797745, + "loss": 2.8471, + "step": 586 + }, + { + "epoch": 0.075136, + "grad_norm": 0.8226315379142761, + "learning_rate": 0.0009946843120595414, + "loss": 2.793, + "step": 587 + }, + { + "epoch": 0.075264, + "grad_norm": 0.8299954533576965, + "learning_rate": 0.00099465412039734, + "loss": 2.7698, + "step": 588 + }, + { + "epoch": 0.075392, + "grad_norm": 0.7330420017242432, + "learning_rate": 0.000994623843698361, + "loss": 2.8638, + "step": 589 + }, + { + "epoch": 0.07552, + "grad_norm": 0.8606112599372864, + "learning_rate": 0.000994593481967809, + "loss": 2.8662, + "step": 590 + }, + { + "epoch": 0.075648, + "grad_norm": 0.7900723814964294, + "learning_rate": 0.0009945630352109038, + "loss": 2.8116, + "step": 591 + }, + { + "epoch": 0.075776, + "grad_norm": 0.7945204973220825, + "learning_rate": 0.0009945325034328792, + "loss": 2.7642, + "step": 592 + }, + { + "epoch": 0.075904, + "grad_norm": 0.8132157921791077, + "learning_rate": 0.0009945018866389843, + "loss": 2.8659, + "step": 593 + }, + { + "epoch": 0.076032, + "grad_norm": 0.805907130241394, + "learning_rate": 0.0009944711848344822, + "loss": 2.8246, + "step": 594 + }, + { + "epoch": 0.07616, + "grad_norm": 0.9272283911705017, + "learning_rate": 0.000994440398024651, + "loss": 2.8498, + "step": 595 + }, + { + "epoch": 0.076288, + "grad_norm": 0.7557371258735657, + "learning_rate": 0.0009944095262147835, + "loss": 2.9491, + "step": 596 + }, + { + "epoch": 0.076416, + "grad_norm": 0.7209718823432922, + "learning_rate": 0.0009943785694101865, + "loss": 2.8825, + "step": 597 + }, + { + "epoch": 0.076544, + "grad_norm": 0.7606251835823059, + "learning_rate": 0.0009943475276161823, + "loss": 2.8307, + "step": 598 + }, + { + "epoch": 0.076672, + "grad_norm": 0.7491905093193054, + "learning_rate": 0.0009943164008381071, + "loss": 2.8562, + "step": 599 + }, + { + "epoch": 0.0768, + "grad_norm": 0.8205605149269104, + "learning_rate": 0.000994285189081312, + "loss": 2.8982, + "step": 600 + }, + { + "epoch": 0.076928, + "grad_norm": 0.8395471572875977, + "learning_rate": 0.0009942538923511626, + "loss": 2.8972, + "step": 601 + }, + { + "epoch": 0.077056, + "grad_norm": 0.8139408826828003, + "learning_rate": 0.000994222510653039, + "loss": 2.8718, + "step": 602 + }, + { + "epoch": 0.077184, + "grad_norm": 0.7475836277008057, + "learning_rate": 0.0009941910439923366, + "loss": 2.9444, + "step": 603 + }, + { + "epoch": 0.077312, + "grad_norm": 0.8095919489860535, + "learning_rate": 0.0009941594923744647, + "loss": 2.9417, + "step": 604 + }, + { + "epoch": 0.07744, + "grad_norm": 0.8157938122749329, + "learning_rate": 0.000994127855804847, + "loss": 2.8589, + "step": 605 + }, + { + "epoch": 0.077568, + "grad_norm": 0.7190120220184326, + "learning_rate": 0.0009940961342889225, + "loss": 2.9675, + "step": 606 + }, + { + "epoch": 0.077696, + "grad_norm": 0.8949825167655945, + "learning_rate": 0.0009940643278321447, + "loss": 2.8728, + "step": 607 + }, + { + "epoch": 0.077824, + "grad_norm": 0.7941190600395203, + "learning_rate": 0.0009940324364399809, + "loss": 2.76, + "step": 608 + }, + { + "epoch": 0.077952, + "grad_norm": 0.7857587933540344, + "learning_rate": 0.0009940004601179141, + "loss": 2.875, + "step": 609 + }, + { + "epoch": 0.07808, + "grad_norm": 1.2257188558578491, + "learning_rate": 0.0009939683988714413, + "loss": 2.8812, + "step": 610 + }, + { + "epoch": 0.078208, + "grad_norm": 0.8046669363975525, + "learning_rate": 0.0009939362527060743, + "loss": 2.8239, + "step": 611 + }, + { + "epoch": 0.078336, + "grad_norm": 0.7974241971969604, + "learning_rate": 0.0009939040216273391, + "loss": 2.9013, + "step": 612 + }, + { + "epoch": 0.078464, + "grad_norm": 0.8709222674369812, + "learning_rate": 0.0009938717056407768, + "loss": 2.8707, + "step": 613 + }, + { + "epoch": 0.078592, + "grad_norm": 0.80191570520401, + "learning_rate": 0.0009938393047519427, + "loss": 2.8831, + "step": 614 + }, + { + "epoch": 0.07872, + "grad_norm": 1.3062794208526611, + "learning_rate": 0.0009938068189664071, + "loss": 2.8474, + "step": 615 + }, + { + "epoch": 0.078848, + "grad_norm": 0.9087037444114685, + "learning_rate": 0.0009937742482897549, + "loss": 2.8886, + "step": 616 + }, + { + "epoch": 0.078976, + "grad_norm": 0.7073392868041992, + "learning_rate": 0.0009937415927275847, + "loss": 2.8203, + "step": 617 + }, + { + "epoch": 0.079104, + "grad_norm": 0.8253702521324158, + "learning_rate": 0.000993708852285511, + "loss": 2.914, + "step": 618 + }, + { + "epoch": 0.079232, + "grad_norm": 0.8949381709098816, + "learning_rate": 0.0009936760269691621, + "loss": 2.9708, + "step": 619 + }, + { + "epoch": 0.07936, + "grad_norm": 0.804724395275116, + "learning_rate": 0.0009936431167841806, + "loss": 2.8114, + "step": 620 + }, + { + "epoch": 0.079488, + "grad_norm": 0.8124799728393555, + "learning_rate": 0.000993610121736225, + "loss": 2.7669, + "step": 621 + }, + { + "epoch": 0.079616, + "grad_norm": 0.9133097529411316, + "learning_rate": 0.0009935770418309667, + "loss": 2.9434, + "step": 622 + }, + { + "epoch": 0.079744, + "grad_norm": 0.7750399112701416, + "learning_rate": 0.000993543877074093, + "loss": 2.812, + "step": 623 + }, + { + "epoch": 0.079872, + "grad_norm": 0.8279737830162048, + "learning_rate": 0.0009935106274713053, + "loss": 2.9426, + "step": 624 + }, + { + "epoch": 0.08, + "grad_norm": 0.7831854224205017, + "learning_rate": 0.0009934772930283193, + "loss": 2.8041, + "step": 625 + }, + { + "epoch": 0.080128, + "grad_norm": 0.7832692861557007, + "learning_rate": 0.0009934438737508658, + "loss": 2.7943, + "step": 626 + }, + { + "epoch": 0.080256, + "grad_norm": 0.7143595814704895, + "learning_rate": 0.00099341036964469, + "loss": 2.8601, + "step": 627 + }, + { + "epoch": 0.080384, + "grad_norm": 0.8304581642150879, + "learning_rate": 0.0009933767807155515, + "loss": 2.8257, + "step": 628 + }, + { + "epoch": 0.080512, + "grad_norm": 0.6824816465377808, + "learning_rate": 0.0009933431069692245, + "loss": 2.8688, + "step": 629 + }, + { + "epoch": 0.08064, + "grad_norm": 0.7557610273361206, + "learning_rate": 0.0009933093484114982, + "loss": 2.9305, + "step": 630 + }, + { + "epoch": 0.080768, + "grad_norm": 0.8651634454727173, + "learning_rate": 0.000993275505048176, + "loss": 2.8681, + "step": 631 + }, + { + "epoch": 0.080896, + "grad_norm": 0.7145448327064514, + "learning_rate": 0.000993241576885076, + "loss": 2.7643, + "step": 632 + }, + { + "epoch": 0.081024, + "grad_norm": 0.7377251982688904, + "learning_rate": 0.0009932075639280306, + "loss": 2.8866, + "step": 633 + }, + { + "epoch": 0.081152, + "grad_norm": 0.7124603390693665, + "learning_rate": 0.0009931734661828874, + "loss": 2.8905, + "step": 634 + }, + { + "epoch": 0.08128, + "grad_norm": 0.77359938621521, + "learning_rate": 0.000993139283655508, + "loss": 2.9134, + "step": 635 + }, + { + "epoch": 0.081408, + "grad_norm": 0.7477929592132568, + "learning_rate": 0.0009931050163517687, + "loss": 2.9645, + "step": 636 + }, + { + "epoch": 0.081536, + "grad_norm": 0.7312575578689575, + "learning_rate": 0.0009930706642775607, + "loss": 2.8397, + "step": 637 + }, + { + "epoch": 0.081664, + "grad_norm": 0.796295166015625, + "learning_rate": 0.0009930362274387892, + "loss": 2.831, + "step": 638 + }, + { + "epoch": 0.081792, + "grad_norm": 0.760362982749939, + "learning_rate": 0.0009930017058413744, + "loss": 2.7266, + "step": 639 + }, + { + "epoch": 0.08192, + "grad_norm": 0.7149213552474976, + "learning_rate": 0.000992967099491251, + "loss": 2.8616, + "step": 640 + }, + { + "epoch": 0.082048, + "grad_norm": 0.7547613382339478, + "learning_rate": 0.0009929324083943682, + "loss": 2.7998, + "step": 641 + }, + { + "epoch": 0.082176, + "grad_norm": 0.770836591720581, + "learning_rate": 0.00099289763255669, + "loss": 2.8115, + "step": 642 + }, + { + "epoch": 0.082304, + "grad_norm": 0.7519711852073669, + "learning_rate": 0.0009928627719841944, + "loss": 2.868, + "step": 643 + }, + { + "epoch": 0.082432, + "grad_norm": 0.8282558917999268, + "learning_rate": 0.0009928278266828747, + "loss": 2.9219, + "step": 644 + }, + { + "epoch": 0.08256, + "grad_norm": 0.7945367097854614, + "learning_rate": 0.0009927927966587383, + "loss": 2.908, + "step": 645 + }, + { + "epoch": 0.082688, + "grad_norm": 0.8094933032989502, + "learning_rate": 0.000992757681917807, + "loss": 2.8664, + "step": 646 + }, + { + "epoch": 0.082816, + "grad_norm": 0.9463929533958435, + "learning_rate": 0.0009927224824661178, + "loss": 2.8313, + "step": 647 + }, + { + "epoch": 0.082944, + "grad_norm": 0.8298996686935425, + "learning_rate": 0.0009926871983097216, + "loss": 2.8777, + "step": 648 + }, + { + "epoch": 0.083072, + "grad_norm": 0.79859858751297, + "learning_rate": 0.0009926518294546844, + "loss": 2.8982, + "step": 649 + }, + { + "epoch": 0.0832, + "grad_norm": 0.6972382068634033, + "learning_rate": 0.0009926163759070863, + "loss": 2.7967, + "step": 650 + }, + { + "epoch": 0.083328, + "grad_norm": 0.7715002298355103, + "learning_rate": 0.0009925808376730224, + "loss": 2.9291, + "step": 651 + }, + { + "epoch": 0.083456, + "grad_norm": 0.7525192499160767, + "learning_rate": 0.000992545214758602, + "loss": 2.8242, + "step": 652 + }, + { + "epoch": 0.083584, + "grad_norm": 0.932239294052124, + "learning_rate": 0.000992509507169949, + "loss": 2.8458, + "step": 653 + }, + { + "epoch": 0.083712, + "grad_norm": 0.7639808654785156, + "learning_rate": 0.0009924737149132022, + "loss": 2.8888, + "step": 654 + }, + { + "epoch": 0.08384, + "grad_norm": 0.7018510103225708, + "learning_rate": 0.0009924378379945146, + "loss": 2.8581, + "step": 655 + }, + { + "epoch": 0.083968, + "grad_norm": 1.0078685283660889, + "learning_rate": 0.0009924018764200537, + "loss": 2.9095, + "step": 656 + }, + { + "epoch": 0.084096, + "grad_norm": 0.814847469329834, + "learning_rate": 0.0009923658301960018, + "loss": 2.9192, + "step": 657 + }, + { + "epoch": 0.084224, + "grad_norm": 0.8424510955810547, + "learning_rate": 0.0009923296993285557, + "loss": 2.8605, + "step": 658 + }, + { + "epoch": 0.084352, + "grad_norm": 0.8350453972816467, + "learning_rate": 0.0009922934838239269, + "loss": 2.9076, + "step": 659 + }, + { + "epoch": 0.08448, + "grad_norm": 0.7243382334709167, + "learning_rate": 0.0009922571836883409, + "loss": 2.9039, + "step": 660 + }, + { + "epoch": 0.084608, + "grad_norm": 0.7178320288658142, + "learning_rate": 0.0009922207989280383, + "loss": 2.7637, + "step": 661 + }, + { + "epoch": 0.084736, + "grad_norm": 0.7874022722244263, + "learning_rate": 0.000992184329549274, + "loss": 2.795, + "step": 662 + }, + { + "epoch": 0.084864, + "grad_norm": 0.8321793079376221, + "learning_rate": 0.0009921477755583177, + "loss": 2.7921, + "step": 663 + }, + { + "epoch": 0.084992, + "grad_norm": 0.8057078123092651, + "learning_rate": 0.0009921111369614532, + "loss": 2.8738, + "step": 664 + }, + { + "epoch": 0.08512, + "grad_norm": 0.7248581051826477, + "learning_rate": 0.0009920744137649792, + "loss": 2.8604, + "step": 665 + }, + { + "epoch": 0.085248, + "grad_norm": 0.6911950707435608, + "learning_rate": 0.000992037605975209, + "loss": 2.8755, + "step": 666 + }, + { + "epoch": 0.085376, + "grad_norm": 0.7698063850402832, + "learning_rate": 0.00099200071359847, + "loss": 2.8539, + "step": 667 + }, + { + "epoch": 0.085504, + "grad_norm": 0.8337739706039429, + "learning_rate": 0.0009919637366411047, + "loss": 3.0345, + "step": 668 + }, + { + "epoch": 0.085632, + "grad_norm": 0.7525111436843872, + "learning_rate": 0.0009919266751094696, + "loss": 2.9439, + "step": 669 + }, + { + "epoch": 0.08576, + "grad_norm": 0.760071873664856, + "learning_rate": 0.0009918895290099363, + "loss": 2.9046, + "step": 670 + }, + { + "epoch": 0.085888, + "grad_norm": 0.8926440477371216, + "learning_rate": 0.0009918522983488904, + "loss": 2.8348, + "step": 671 + }, + { + "epoch": 0.086016, + "grad_norm": 0.7180164456367493, + "learning_rate": 0.0009918149831327323, + "loss": 2.7924, + "step": 672 + }, + { + "epoch": 0.086144, + "grad_norm": 0.8653330206871033, + "learning_rate": 0.0009917775833678771, + "loss": 2.943, + "step": 673 + }, + { + "epoch": 0.086272, + "grad_norm": 1.3636759519577026, + "learning_rate": 0.000991740099060754, + "loss": 3.0195, + "step": 674 + }, + { + "epoch": 0.0864, + "grad_norm": 3.0478503704071045, + "learning_rate": 0.0009917025302178073, + "loss": 2.9604, + "step": 675 + }, + { + "epoch": 0.086528, + "grad_norm": 0.7441273927688599, + "learning_rate": 0.0009916648768454955, + "loss": 2.8175, + "step": 676 + }, + { + "epoch": 0.086656, + "grad_norm": 0.7335507869720459, + "learning_rate": 0.000991627138950291, + "loss": 2.8641, + "step": 677 + }, + { + "epoch": 0.086784, + "grad_norm": 0.7869941592216492, + "learning_rate": 0.0009915893165386824, + "loss": 2.899, + "step": 678 + }, + { + "epoch": 0.086912, + "grad_norm": 0.7374983429908752, + "learning_rate": 0.0009915514096171711, + "loss": 2.9255, + "step": 679 + }, + { + "epoch": 0.08704, + "grad_norm": 0.7504434585571289, + "learning_rate": 0.0009915134181922739, + "loss": 2.9115, + "step": 680 + }, + { + "epoch": 0.087168, + "grad_norm": 0.723962664604187, + "learning_rate": 0.000991475342270522, + "loss": 2.9259, + "step": 681 + }, + { + "epoch": 0.087296, + "grad_norm": 0.8230690360069275, + "learning_rate": 0.0009914371818584612, + "loss": 2.8863, + "step": 682 + }, + { + "epoch": 0.087424, + "grad_norm": 0.7582704424858093, + "learning_rate": 0.0009913989369626515, + "loss": 2.8455, + "step": 683 + }, + { + "epoch": 0.087552, + "grad_norm": 0.7423965930938721, + "learning_rate": 0.0009913606075896675, + "loss": 2.8745, + "step": 684 + }, + { + "epoch": 0.08768, + "grad_norm": 0.8798903822898865, + "learning_rate": 0.000991322193746099, + "loss": 2.9666, + "step": 685 + }, + { + "epoch": 0.087808, + "grad_norm": 0.9407333135604858, + "learning_rate": 0.0009912836954385496, + "loss": 2.8564, + "step": 686 + }, + { + "epoch": 0.087936, + "grad_norm": 0.8778307437896729, + "learning_rate": 0.0009912451126736373, + "loss": 2.8411, + "step": 687 + }, + { + "epoch": 0.088064, + "grad_norm": 0.9730737209320068, + "learning_rate": 0.0009912064454579953, + "loss": 2.9251, + "step": 688 + }, + { + "epoch": 0.088192, + "grad_norm": 0.9133803248405457, + "learning_rate": 0.0009911676937982706, + "loss": 2.8953, + "step": 689 + }, + { + "epoch": 0.08832, + "grad_norm": 0.7632529139518738, + "learning_rate": 0.0009911288577011254, + "loss": 2.903, + "step": 690 + }, + { + "epoch": 0.088448, + "grad_norm": 0.7411234378814697, + "learning_rate": 0.0009910899371732358, + "loss": 2.8768, + "step": 691 + }, + { + "epoch": 0.088576, + "grad_norm": 0.7168850302696228, + "learning_rate": 0.0009910509322212928, + "loss": 2.8769, + "step": 692 + }, + { + "epoch": 0.088704, + "grad_norm": 0.7761476635932922, + "learning_rate": 0.000991011842852002, + "loss": 2.7561, + "step": 693 + }, + { + "epoch": 0.088832, + "grad_norm": 0.8862825036048889, + "learning_rate": 0.0009909726690720828, + "loss": 2.7818, + "step": 694 + }, + { + "epoch": 0.08896, + "grad_norm": 0.7955124378204346, + "learning_rate": 0.00099093341088827, + "loss": 2.7901, + "step": 695 + }, + { + "epoch": 0.089088, + "grad_norm": 0.7089766263961792, + "learning_rate": 0.0009908940683073127, + "loss": 2.8501, + "step": 696 + }, + { + "epoch": 0.089216, + "grad_norm": 0.7886579632759094, + "learning_rate": 0.000990854641335974, + "loss": 2.8351, + "step": 697 + }, + { + "epoch": 0.089344, + "grad_norm": 0.6849287152290344, + "learning_rate": 0.000990815129981032, + "loss": 2.8203, + "step": 698 + }, + { + "epoch": 0.089472, + "grad_norm": 0.6867862343788147, + "learning_rate": 0.000990775534249279, + "loss": 2.8234, + "step": 699 + }, + { + "epoch": 0.0896, + "grad_norm": 0.7529506087303162, + "learning_rate": 0.0009907358541475222, + "loss": 2.9818, + "step": 700 + }, + { + "epoch": 0.089728, + "grad_norm": 0.8373282551765442, + "learning_rate": 0.0009906960896825829, + "loss": 2.8342, + "step": 701 + }, + { + "epoch": 0.089856, + "grad_norm": 0.6819619536399841, + "learning_rate": 0.0009906562408612974, + "loss": 2.79, + "step": 702 + }, + { + "epoch": 0.089984, + "grad_norm": 0.6994437575340271, + "learning_rate": 0.0009906163076905158, + "loss": 2.6912, + "step": 703 + }, + { + "epoch": 0.090112, + "grad_norm": 0.7541998028755188, + "learning_rate": 0.000990576290177103, + "loss": 2.9513, + "step": 704 + }, + { + "epoch": 0.09024, + "grad_norm": 0.8213698267936707, + "learning_rate": 0.0009905361883279387, + "loss": 2.9041, + "step": 705 + }, + { + "epoch": 0.090368, + "grad_norm": 0.7771000266075134, + "learning_rate": 0.000990496002149917, + "loss": 2.8304, + "step": 706 + }, + { + "epoch": 0.090496, + "grad_norm": 0.7140235304832458, + "learning_rate": 0.0009904557316499463, + "loss": 2.7972, + "step": 707 + }, + { + "epoch": 0.090624, + "grad_norm": 0.7459695339202881, + "learning_rate": 0.0009904153768349492, + "loss": 2.8207, + "step": 708 + }, + { + "epoch": 0.090752, + "grad_norm": 0.7276132702827454, + "learning_rate": 0.0009903749377118637, + "loss": 2.8382, + "step": 709 + }, + { + "epoch": 0.09088, + "grad_norm": 0.747065544128418, + "learning_rate": 0.0009903344142876414, + "loss": 2.8572, + "step": 710 + }, + { + "epoch": 0.091008, + "grad_norm": 0.7954176664352417, + "learning_rate": 0.0009902938065692486, + "loss": 2.95, + "step": 711 + }, + { + "epoch": 0.091136, + "grad_norm": 0.9190263152122498, + "learning_rate": 0.000990253114563667, + "loss": 2.854, + "step": 712 + }, + { + "epoch": 0.091264, + "grad_norm": 0.7369100451469421, + "learning_rate": 0.000990212338277891, + "loss": 2.806, + "step": 713 + }, + { + "epoch": 0.091392, + "grad_norm": 0.7376288175582886, + "learning_rate": 0.0009901714777189312, + "loss": 2.8353, + "step": 714 + }, + { + "epoch": 0.09152, + "grad_norm": 0.756733775138855, + "learning_rate": 0.0009901305328938118, + "loss": 2.8409, + "step": 715 + }, + { + "epoch": 0.091648, + "grad_norm": 0.7417981624603271, + "learning_rate": 0.000990089503809572, + "loss": 2.8773, + "step": 716 + }, + { + "epoch": 0.091776, + "grad_norm": 0.7322847843170166, + "learning_rate": 0.0009900483904732645, + "loss": 2.8745, + "step": 717 + }, + { + "epoch": 0.091904, + "grad_norm": 0.6920481324195862, + "learning_rate": 0.0009900071928919578, + "loss": 2.8209, + "step": 718 + }, + { + "epoch": 0.092032, + "grad_norm": 0.7069603204727173, + "learning_rate": 0.0009899659110727337, + "loss": 2.8178, + "step": 719 + }, + { + "epoch": 0.09216, + "grad_norm": 0.7882048487663269, + "learning_rate": 0.0009899245450226896, + "loss": 2.8061, + "step": 720 + }, + { + "epoch": 0.092288, + "grad_norm": 0.722066342830658, + "learning_rate": 0.0009898830947489366, + "loss": 2.8207, + "step": 721 + }, + { + "epoch": 0.092416, + "grad_norm": 0.7623542547225952, + "learning_rate": 0.0009898415602586001, + "loss": 2.7644, + "step": 722 + }, + { + "epoch": 0.092544, + "grad_norm": 0.7430248260498047, + "learning_rate": 0.000989799941558821, + "loss": 2.8369, + "step": 723 + }, + { + "epoch": 0.092672, + "grad_norm": 0.6972943544387817, + "learning_rate": 0.000989758238656753, + "loss": 2.8616, + "step": 724 + }, + { + "epoch": 0.0928, + "grad_norm": 0.8298338651657104, + "learning_rate": 0.0009897164515595667, + "loss": 2.8489, + "step": 725 + }, + { + "epoch": 0.092928, + "grad_norm": 0.7238650918006897, + "learning_rate": 0.0009896745802744449, + "loss": 2.7586, + "step": 726 + }, + { + "epoch": 0.093056, + "grad_norm": 0.6935379505157471, + "learning_rate": 0.000989632624808586, + "loss": 2.8891, + "step": 727 + }, + { + "epoch": 0.093184, + "grad_norm": 0.7249261736869812, + "learning_rate": 0.0009895905851692026, + "loss": 2.7736, + "step": 728 + }, + { + "epoch": 0.093312, + "grad_norm": 0.7214240431785583, + "learning_rate": 0.0009895484613635216, + "loss": 2.8321, + "step": 729 + }, + { + "epoch": 0.09344, + "grad_norm": 0.6799103021621704, + "learning_rate": 0.0009895062533987848, + "loss": 2.8393, + "step": 730 + }, + { + "epoch": 0.093568, + "grad_norm": 0.7305128574371338, + "learning_rate": 0.0009894639612822486, + "loss": 2.8625, + "step": 731 + }, + { + "epoch": 0.093696, + "grad_norm": 0.7512807846069336, + "learning_rate": 0.0009894215850211824, + "loss": 2.8335, + "step": 732 + }, + { + "epoch": 0.093824, + "grad_norm": 0.7089670300483704, + "learning_rate": 0.0009893791246228724, + "loss": 2.8225, + "step": 733 + }, + { + "epoch": 0.093952, + "grad_norm": 0.7406983375549316, + "learning_rate": 0.0009893365800946175, + "loss": 2.8781, + "step": 734 + }, + { + "epoch": 0.09408, + "grad_norm": 0.7323946356773376, + "learning_rate": 0.0009892939514437313, + "loss": 2.8625, + "step": 735 + }, + { + "epoch": 0.094208, + "grad_norm": 0.7321927547454834, + "learning_rate": 0.0009892512386775428, + "loss": 2.8422, + "step": 736 + }, + { + "epoch": 0.094336, + "grad_norm": 0.6861856579780579, + "learning_rate": 0.000989208441803394, + "loss": 2.7935, + "step": 737 + }, + { + "epoch": 0.094464, + "grad_norm": 0.7849147915840149, + "learning_rate": 0.0009891655608286433, + "loss": 2.9138, + "step": 738 + }, + { + "epoch": 0.094592, + "grad_norm": 0.7580278515815735, + "learning_rate": 0.0009891225957606613, + "loss": 2.7232, + "step": 739 + }, + { + "epoch": 0.09472, + "grad_norm": 0.826312243938446, + "learning_rate": 0.000989079546606835, + "loss": 2.8195, + "step": 740 + }, + { + "epoch": 0.094848, + "grad_norm": 0.7311555743217468, + "learning_rate": 0.0009890364133745646, + "loss": 2.7614, + "step": 741 + }, + { + "epoch": 0.094976, + "grad_norm": 0.763739287853241, + "learning_rate": 0.0009889931960712652, + "loss": 2.8756, + "step": 742 + }, + { + "epoch": 0.095104, + "grad_norm": 0.6614788770675659, + "learning_rate": 0.0009889498947043668, + "loss": 2.8389, + "step": 743 + }, + { + "epoch": 0.095232, + "grad_norm": 0.6671706438064575, + "learning_rate": 0.000988906509281313, + "loss": 2.8457, + "step": 744 + }, + { + "epoch": 0.09536, + "grad_norm": 0.7160948514938354, + "learning_rate": 0.0009888630398095623, + "loss": 2.8916, + "step": 745 + }, + { + "epoch": 0.095488, + "grad_norm": 0.7248741388320923, + "learning_rate": 0.0009888194862965878, + "loss": 2.848, + "step": 746 + }, + { + "epoch": 0.095616, + "grad_norm": 0.7239066362380981, + "learning_rate": 0.0009887758487498765, + "loss": 2.8461, + "step": 747 + }, + { + "epoch": 0.095744, + "grad_norm": 0.7101806402206421, + "learning_rate": 0.0009887321271769304, + "loss": 2.8173, + "step": 748 + }, + { + "epoch": 0.095872, + "grad_norm": 0.7313535809516907, + "learning_rate": 0.000988688321585266, + "loss": 2.8384, + "step": 749 + }, + { + "epoch": 0.096, + "grad_norm": 0.7425237894058228, + "learning_rate": 0.0009886444319824137, + "loss": 2.8071, + "step": 750 + }, + { + "epoch": 0.096128, + "grad_norm": 0.7459864616394043, + "learning_rate": 0.0009886004583759187, + "loss": 2.8968, + "step": 751 + }, + { + "epoch": 0.096256, + "grad_norm": 0.7610161900520325, + "learning_rate": 0.0009885564007733405, + "loss": 2.8675, + "step": 752 + }, + { + "epoch": 0.096384, + "grad_norm": 0.7036921977996826, + "learning_rate": 0.0009885122591822533, + "loss": 2.8448, + "step": 753 + }, + { + "epoch": 0.096512, + "grad_norm": 0.769883394241333, + "learning_rate": 0.0009884680336102452, + "loss": 2.938, + "step": 754 + }, + { + "epoch": 0.09664, + "grad_norm": 0.7477030754089355, + "learning_rate": 0.0009884237240649193, + "loss": 2.8023, + "step": 755 + }, + { + "epoch": 0.096768, + "grad_norm": 0.7171107530593872, + "learning_rate": 0.0009883793305538932, + "loss": 2.8496, + "step": 756 + }, + { + "epoch": 0.096896, + "grad_norm": 0.7267308831214905, + "learning_rate": 0.0009883348530847983, + "loss": 2.7736, + "step": 757 + }, + { + "epoch": 0.097024, + "grad_norm": 0.6505060195922852, + "learning_rate": 0.0009882902916652809, + "loss": 2.9694, + "step": 758 + }, + { + "epoch": 0.097152, + "grad_norm": 0.6769353151321411, + "learning_rate": 0.0009882456463030016, + "loss": 2.6836, + "step": 759 + }, + { + "epoch": 0.09728, + "grad_norm": 0.6956661343574524, + "learning_rate": 0.0009882009170056353, + "loss": 2.7952, + "step": 760 + }, + { + "epoch": 0.097408, + "grad_norm": 0.7350356578826904, + "learning_rate": 0.000988156103780872, + "loss": 2.797, + "step": 761 + }, + { + "epoch": 0.097536, + "grad_norm": 0.6989827156066895, + "learning_rate": 0.000988111206636415, + "loss": 2.809, + "step": 762 + }, + { + "epoch": 0.097664, + "grad_norm": 0.7215303182601929, + "learning_rate": 0.000988066225579983, + "loss": 2.9055, + "step": 763 + }, + { + "epoch": 0.097792, + "grad_norm": 0.7074881792068481, + "learning_rate": 0.0009880211606193088, + "loss": 2.7201, + "step": 764 + }, + { + "epoch": 0.09792, + "grad_norm": 0.7426018118858337, + "learning_rate": 0.0009879760117621392, + "loss": 2.761, + "step": 765 + }, + { + "epoch": 0.098048, + "grad_norm": 0.8300538659095764, + "learning_rate": 0.0009879307790162362, + "loss": 2.7843, + "step": 766 + }, + { + "epoch": 0.098176, + "grad_norm": 0.691016674041748, + "learning_rate": 0.000987885462389376, + "loss": 2.7799, + "step": 767 + }, + { + "epoch": 0.098304, + "grad_norm": 0.7044005393981934, + "learning_rate": 0.0009878400618893483, + "loss": 2.8414, + "step": 768 + }, + { + "epoch": 0.098432, + "grad_norm": 0.6454651355743408, + "learning_rate": 0.0009877945775239585, + "loss": 2.8454, + "step": 769 + }, + { + "epoch": 0.09856, + "grad_norm": 0.6972115635871887, + "learning_rate": 0.0009877490093010259, + "loss": 2.8577, + "step": 770 + }, + { + "epoch": 0.098688, + "grad_norm": 0.7290935516357422, + "learning_rate": 0.0009877033572283842, + "loss": 2.9073, + "step": 771 + }, + { + "epoch": 0.098816, + "grad_norm": 0.7479534149169922, + "learning_rate": 0.0009876576213138815, + "loss": 2.9625, + "step": 772 + }, + { + "epoch": 0.098944, + "grad_norm": 0.7013347744941711, + "learning_rate": 0.0009876118015653803, + "loss": 2.8044, + "step": 773 + }, + { + "epoch": 0.099072, + "grad_norm": 0.6676444411277771, + "learning_rate": 0.0009875658979907574, + "loss": 2.9243, + "step": 774 + }, + { + "epoch": 0.0992, + "grad_norm": 0.6584591865539551, + "learning_rate": 0.0009875199105979042, + "loss": 2.8196, + "step": 775 + }, + { + "epoch": 0.099328, + "grad_norm": 0.6920338273048401, + "learning_rate": 0.0009874738393947267, + "loss": 2.8383, + "step": 776 + }, + { + "epoch": 0.099456, + "grad_norm": 0.6311237215995789, + "learning_rate": 0.0009874276843891449, + "loss": 2.8748, + "step": 777 + }, + { + "epoch": 0.099584, + "grad_norm": 1.332052230834961, + "learning_rate": 0.0009873814455890935, + "loss": 2.8264, + "step": 778 + }, + { + "epoch": 0.099712, + "grad_norm": 0.7108719348907471, + "learning_rate": 0.0009873351230025211, + "loss": 2.8503, + "step": 779 + }, + { + "epoch": 0.09984, + "grad_norm": 0.6798509359359741, + "learning_rate": 0.0009872887166373916, + "loss": 2.8353, + "step": 780 + }, + { + "epoch": 0.099968, + "grad_norm": 0.681185245513916, + "learning_rate": 0.0009872422265016826, + "loss": 2.8279, + "step": 781 + }, + { + "epoch": 0.100096, + "grad_norm": 0.7570596933364868, + "learning_rate": 0.0009871956526033862, + "loss": 2.7364, + "step": 782 + }, + { + "epoch": 0.100224, + "grad_norm": 0.72552490234375, + "learning_rate": 0.0009871489949505091, + "loss": 2.8333, + "step": 783 + }, + { + "epoch": 0.100352, + "grad_norm": 0.7157814502716064, + "learning_rate": 0.0009871022535510722, + "loss": 2.7727, + "step": 784 + }, + { + "epoch": 0.10048, + "grad_norm": 0.6475344896316528, + "learning_rate": 0.0009870554284131111, + "loss": 2.7494, + "step": 785 + }, + { + "epoch": 0.100608, + "grad_norm": 0.6657574772834778, + "learning_rate": 0.0009870085195446752, + "loss": 2.8317, + "step": 786 + }, + { + "epoch": 0.100736, + "grad_norm": 0.9935200214385986, + "learning_rate": 0.000986961526953829, + "loss": 2.8641, + "step": 787 + }, + { + "epoch": 0.100864, + "grad_norm": 0.751855731010437, + "learning_rate": 0.000986914450648651, + "loss": 2.7417, + "step": 788 + }, + { + "epoch": 0.100992, + "grad_norm": 0.7009740471839905, + "learning_rate": 0.0009868672906372341, + "loss": 2.8265, + "step": 789 + }, + { + "epoch": 0.10112, + "grad_norm": 0.6803882718086243, + "learning_rate": 0.0009868200469276857, + "loss": 2.9041, + "step": 790 + }, + { + "epoch": 0.101248, + "grad_norm": 0.6992521286010742, + "learning_rate": 0.0009867727195281275, + "loss": 2.8376, + "step": 791 + }, + { + "epoch": 0.101376, + "grad_norm": 0.6924471259117126, + "learning_rate": 0.0009867253084466956, + "loss": 2.9402, + "step": 792 + }, + { + "epoch": 0.101504, + "grad_norm": 0.6720661520957947, + "learning_rate": 0.0009866778136915408, + "loss": 2.7875, + "step": 793 + }, + { + "epoch": 0.101632, + "grad_norm": 0.7311491370201111, + "learning_rate": 0.0009866302352708276, + "loss": 2.8347, + "step": 794 + }, + { + "epoch": 0.10176, + "grad_norm": 0.8350388407707214, + "learning_rate": 0.0009865825731927354, + "loss": 2.8592, + "step": 795 + }, + { + "epoch": 0.101888, + "grad_norm": 0.7412349581718445, + "learning_rate": 0.000986534827465458, + "loss": 2.8806, + "step": 796 + }, + { + "epoch": 0.102016, + "grad_norm": 0.7111446261405945, + "learning_rate": 0.0009864869980972034, + "loss": 2.7907, + "step": 797 + }, + { + "epoch": 0.102144, + "grad_norm": 0.7535229325294495, + "learning_rate": 0.000986439085096194, + "loss": 2.6987, + "step": 798 + }, + { + "epoch": 0.102272, + "grad_norm": 0.7558249235153198, + "learning_rate": 0.0009863910884706666, + "loss": 2.9187, + "step": 799 + }, + { + "epoch": 0.1024, + "grad_norm": 0.7505187392234802, + "learning_rate": 0.0009863430082288722, + "loss": 2.8993, + "step": 800 + }, + { + "epoch": 0.102528, + "grad_norm": 0.7182137370109558, + "learning_rate": 0.0009862948443790766, + "loss": 2.7663, + "step": 801 + }, + { + "epoch": 0.102656, + "grad_norm": 0.6924146413803101, + "learning_rate": 0.0009862465969295597, + "loss": 2.8313, + "step": 802 + }, + { + "epoch": 0.102784, + "grad_norm": 0.6943239569664001, + "learning_rate": 0.0009861982658886158, + "loss": 2.8712, + "step": 803 + }, + { + "epoch": 0.102912, + "grad_norm": 0.7231345772743225, + "learning_rate": 0.0009861498512645532, + "loss": 2.9308, + "step": 804 + }, + { + "epoch": 0.10304, + "grad_norm": 0.7031784057617188, + "learning_rate": 0.0009861013530656955, + "loss": 2.8459, + "step": 805 + }, + { + "epoch": 0.103168, + "grad_norm": 0.6744710206985474, + "learning_rate": 0.0009860527713003798, + "loss": 2.7286, + "step": 806 + }, + { + "epoch": 0.103296, + "grad_norm": 0.7508935928344727, + "learning_rate": 0.0009860041059769579, + "loss": 2.8955, + "step": 807 + }, + { + "epoch": 0.103424, + "grad_norm": 0.7038334608078003, + "learning_rate": 0.000985955357103796, + "loss": 2.9181, + "step": 808 + }, + { + "epoch": 0.103552, + "grad_norm": 0.7346904277801514, + "learning_rate": 0.0009859065246892744, + "loss": 2.7801, + "step": 809 + }, + { + "epoch": 0.10368, + "grad_norm": 0.6825481057167053, + "learning_rate": 0.000985857608741788, + "loss": 2.7762, + "step": 810 + }, + { + "epoch": 0.103808, + "grad_norm": 0.7063619494438171, + "learning_rate": 0.0009858086092697463, + "loss": 2.7687, + "step": 811 + }, + { + "epoch": 0.103936, + "grad_norm": 0.6591432690620422, + "learning_rate": 0.0009857595262815727, + "loss": 2.872, + "step": 812 + }, + { + "epoch": 0.104064, + "grad_norm": 0.6454761624336243, + "learning_rate": 0.000985710359785705, + "loss": 2.8229, + "step": 813 + }, + { + "epoch": 0.104192, + "grad_norm": 0.6330951452255249, + "learning_rate": 0.0009856611097905955, + "loss": 2.7974, + "step": 814 + }, + { + "epoch": 0.10432, + "grad_norm": 0.6317194700241089, + "learning_rate": 0.0009856117763047114, + "loss": 2.9071, + "step": 815 + }, + { + "epoch": 0.104448, + "grad_norm": 0.6900577545166016, + "learning_rate": 0.0009855623593365331, + "loss": 2.9042, + "step": 816 + }, + { + "epoch": 0.104576, + "grad_norm": 0.7121096849441528, + "learning_rate": 0.000985512858894556, + "loss": 2.7124, + "step": 817 + }, + { + "epoch": 0.104704, + "grad_norm": 0.8220340609550476, + "learning_rate": 0.0009854632749872899, + "loss": 2.8511, + "step": 818 + }, + { + "epoch": 0.104832, + "grad_norm": 0.6916947960853577, + "learning_rate": 0.0009854136076232587, + "loss": 2.8563, + "step": 819 + }, + { + "epoch": 0.10496, + "grad_norm": 0.709361732006073, + "learning_rate": 0.0009853638568110012, + "loss": 2.8572, + "step": 820 + }, + { + "epoch": 0.105088, + "grad_norm": 0.6753686666488647, + "learning_rate": 0.0009853140225590698, + "loss": 2.8661, + "step": 821 + }, + { + "epoch": 0.105216, + "grad_norm": 0.7245519757270813, + "learning_rate": 0.0009852641048760317, + "loss": 2.8319, + "step": 822 + }, + { + "epoch": 0.105344, + "grad_norm": 0.6777408719062805, + "learning_rate": 0.0009852141037704682, + "loss": 2.8021, + "step": 823 + }, + { + "epoch": 0.105472, + "grad_norm": 0.641461968421936, + "learning_rate": 0.0009851640192509753, + "loss": 2.7147, + "step": 824 + }, + { + "epoch": 0.1056, + "grad_norm": 0.6775611042976379, + "learning_rate": 0.0009851138513261631, + "loss": 2.8585, + "step": 825 + }, + { + "epoch": 0.105728, + "grad_norm": 0.6586411595344543, + "learning_rate": 0.0009850636000046558, + "loss": 2.8377, + "step": 826 + }, + { + "epoch": 0.105856, + "grad_norm": 0.6785038113594055, + "learning_rate": 0.0009850132652950922, + "loss": 2.7795, + "step": 827 + }, + { + "epoch": 0.105984, + "grad_norm": 0.7746644020080566, + "learning_rate": 0.0009849628472061258, + "loss": 2.8131, + "step": 828 + }, + { + "epoch": 0.106112, + "grad_norm": 0.741019070148468, + "learning_rate": 0.0009849123457464236, + "loss": 2.7746, + "step": 829 + }, + { + "epoch": 0.10624, + "grad_norm": 0.6637568473815918, + "learning_rate": 0.0009848617609246677, + "loss": 2.7743, + "step": 830 + }, + { + "epoch": 0.106368, + "grad_norm": 0.7340434789657593, + "learning_rate": 0.0009848110927495538, + "loss": 2.7682, + "step": 831 + }, + { + "epoch": 0.106496, + "grad_norm": 0.7781968712806702, + "learning_rate": 0.000984760341229793, + "loss": 2.755, + "step": 832 + }, + { + "epoch": 0.106624, + "grad_norm": 0.7643356919288635, + "learning_rate": 0.0009847095063741096, + "loss": 2.8485, + "step": 833 + }, + { + "epoch": 0.106752, + "grad_norm": 0.6973885297775269, + "learning_rate": 0.000984658588191243, + "loss": 2.7982, + "step": 834 + }, + { + "epoch": 0.10688, + "grad_norm": 0.8116495609283447, + "learning_rate": 0.0009846075866899462, + "loss": 2.8446, + "step": 835 + }, + { + "epoch": 0.107008, + "grad_norm": 0.7262203097343445, + "learning_rate": 0.0009845565018789873, + "loss": 2.818, + "step": 836 + }, + { + "epoch": 0.107136, + "grad_norm": 0.6293861269950867, + "learning_rate": 0.0009845053337671481, + "loss": 2.8237, + "step": 837 + }, + { + "epoch": 0.107264, + "grad_norm": 0.652298629283905, + "learning_rate": 0.0009844540823632252, + "loss": 2.8144, + "step": 838 + }, + { + "epoch": 0.107392, + "grad_norm": 0.7364869713783264, + "learning_rate": 0.0009844027476760295, + "loss": 2.8503, + "step": 839 + }, + { + "epoch": 0.10752, + "grad_norm": 0.7778275012969971, + "learning_rate": 0.0009843513297143856, + "loss": 2.8036, + "step": 840 + }, + { + "epoch": 0.107648, + "grad_norm": 0.7322375774383545, + "learning_rate": 0.0009842998284871332, + "loss": 2.8393, + "step": 841 + }, + { + "epoch": 0.107776, + "grad_norm": 0.7101910710334778, + "learning_rate": 0.0009842482440031255, + "loss": 2.8586, + "step": 842 + }, + { + "epoch": 0.107904, + "grad_norm": 0.7225572466850281, + "learning_rate": 0.0009841965762712312, + "loss": 2.7813, + "step": 843 + }, + { + "epoch": 0.108032, + "grad_norm": 0.7090955376625061, + "learning_rate": 0.0009841448253003317, + "loss": 2.8631, + "step": 844 + }, + { + "epoch": 0.10816, + "grad_norm": 0.6929117441177368, + "learning_rate": 0.0009840929910993244, + "loss": 2.8246, + "step": 845 + }, + { + "epoch": 0.108288, + "grad_norm": 0.6790797710418701, + "learning_rate": 0.0009840410736771197, + "loss": 2.6924, + "step": 846 + }, + { + "epoch": 0.108416, + "grad_norm": 0.6415429711341858, + "learning_rate": 0.000983989073042643, + "loss": 2.8075, + "step": 847 + }, + { + "epoch": 0.108544, + "grad_norm": 0.7478675842285156, + "learning_rate": 0.0009839369892048336, + "loss": 2.9357, + "step": 848 + }, + { + "epoch": 0.108672, + "grad_norm": 0.6312830448150635, + "learning_rate": 0.0009838848221726455, + "loss": 2.775, + "step": 849 + }, + { + "epoch": 0.1088, + "grad_norm": 0.8103664517402649, + "learning_rate": 0.000983832571955047, + "loss": 2.9053, + "step": 850 + }, + { + "epoch": 0.108928, + "grad_norm": 0.711162805557251, + "learning_rate": 0.00098378023856102, + "loss": 2.8498, + "step": 851 + }, + { + "epoch": 0.109056, + "grad_norm": 0.6967734694480896, + "learning_rate": 0.0009837278219995617, + "loss": 2.8671, + "step": 852 + }, + { + "epoch": 0.109184, + "grad_norm": 0.7177896499633789, + "learning_rate": 0.000983675322279683, + "loss": 2.8118, + "step": 853 + }, + { + "epoch": 0.109312, + "grad_norm": 0.6779946088790894, + "learning_rate": 0.0009836227394104092, + "loss": 2.8135, + "step": 854 + }, + { + "epoch": 0.10944, + "grad_norm": 0.6368451118469238, + "learning_rate": 0.0009835700734007796, + "loss": 2.8067, + "step": 855 + }, + { + "epoch": 0.109568, + "grad_norm": 0.6852014660835266, + "learning_rate": 0.0009835173242598485, + "loss": 2.8982, + "step": 856 + }, + { + "epoch": 0.109696, + "grad_norm": 0.634838342666626, + "learning_rate": 0.0009834644919966841, + "loss": 2.7611, + "step": 857 + }, + { + "epoch": 0.109824, + "grad_norm": 0.6868596076965332, + "learning_rate": 0.0009834115766203686, + "loss": 2.7954, + "step": 858 + }, + { + "epoch": 0.109952, + "grad_norm": 0.7114148736000061, + "learning_rate": 0.000983358578139999, + "loss": 2.7957, + "step": 859 + }, + { + "epoch": 0.11008, + "grad_norm": 0.710063636302948, + "learning_rate": 0.000983305496564686, + "loss": 2.9228, + "step": 860 + }, + { + "epoch": 0.110208, + "grad_norm": 0.6717804074287415, + "learning_rate": 0.0009832523319035554, + "loss": 2.8622, + "step": 861 + }, + { + "epoch": 0.110336, + "grad_norm": 0.6814509034156799, + "learning_rate": 0.0009831990841657463, + "loss": 2.682, + "step": 862 + }, + { + "epoch": 0.110464, + "grad_norm": 0.7521986365318298, + "learning_rate": 0.0009831457533604133, + "loss": 2.8675, + "step": 863 + }, + { + "epoch": 0.110592, + "grad_norm": 0.6606860756874084, + "learning_rate": 0.000983092339496724, + "loss": 2.7467, + "step": 864 + }, + { + "epoch": 0.11072, + "grad_norm": 0.7901679277420044, + "learning_rate": 0.000983038842583861, + "loss": 2.8666, + "step": 865 + }, + { + "epoch": 0.110848, + "grad_norm": 0.6971216201782227, + "learning_rate": 0.0009829852626310211, + "loss": 2.6926, + "step": 866 + }, + { + "epoch": 0.110976, + "grad_norm": 0.6423192024230957, + "learning_rate": 0.0009829315996474155, + "loss": 2.7941, + "step": 867 + }, + { + "epoch": 0.111104, + "grad_norm": 0.6556682586669922, + "learning_rate": 0.0009828778536422691, + "loss": 2.8821, + "step": 868 + }, + { + "epoch": 0.111232, + "grad_norm": 0.641143798828125, + "learning_rate": 0.0009828240246248217, + "loss": 2.8299, + "step": 869 + }, + { + "epoch": 0.11136, + "grad_norm": 0.7321833968162537, + "learning_rate": 0.0009827701126043271, + "loss": 2.8916, + "step": 870 + }, + { + "epoch": 0.111488, + "grad_norm": 0.7276745438575745, + "learning_rate": 0.0009827161175900534, + "loss": 2.7032, + "step": 871 + }, + { + "epoch": 0.111616, + "grad_norm": 0.6712154150009155, + "learning_rate": 0.0009826620395912832, + "loss": 2.8384, + "step": 872 + }, + { + "epoch": 0.111744, + "grad_norm": 0.6931195259094238, + "learning_rate": 0.0009826078786173125, + "loss": 2.8016, + "step": 873 + }, + { + "epoch": 0.111872, + "grad_norm": 0.9606790542602539, + "learning_rate": 0.000982553634677453, + "loss": 2.767, + "step": 874 + }, + { + "epoch": 0.112, + "grad_norm": 0.749415397644043, + "learning_rate": 0.000982499307781029, + "loss": 2.8551, + "step": 875 + }, + { + "epoch": 0.112128, + "grad_norm": 0.6956034302711487, + "learning_rate": 0.0009824448979373806, + "loss": 2.6836, + "step": 876 + }, + { + "epoch": 0.112256, + "grad_norm": 0.718097984790802, + "learning_rate": 0.0009823904051558612, + "loss": 2.8964, + "step": 877 + }, + { + "epoch": 0.112384, + "grad_norm": 0.7438269853591919, + "learning_rate": 0.0009823358294458388, + "loss": 2.9709, + "step": 878 + }, + { + "epoch": 0.112512, + "grad_norm": 0.7444434762001038, + "learning_rate": 0.0009822811708166959, + "loss": 2.8659, + "step": 879 + }, + { + "epoch": 0.11264, + "grad_norm": 0.655365526676178, + "learning_rate": 0.0009822264292778282, + "loss": 2.7954, + "step": 880 + }, + { + "epoch": 0.112768, + "grad_norm": 0.653344452381134, + "learning_rate": 0.0009821716048386473, + "loss": 2.763, + "step": 881 + }, + { + "epoch": 0.112896, + "grad_norm": 0.7563408017158508, + "learning_rate": 0.0009821166975085773, + "loss": 2.9076, + "step": 882 + }, + { + "epoch": 0.113024, + "grad_norm": 0.662342369556427, + "learning_rate": 0.0009820617072970582, + "loss": 2.7371, + "step": 883 + }, + { + "epoch": 0.113152, + "grad_norm": 0.6423828601837158, + "learning_rate": 0.0009820066342135427, + "loss": 2.8056, + "step": 884 + }, + { + "epoch": 0.11328, + "grad_norm": 0.6426447033882141, + "learning_rate": 0.0009819514782674992, + "loss": 2.8094, + "step": 885 + }, + { + "epoch": 0.113408, + "grad_norm": 0.6390810012817383, + "learning_rate": 0.0009818962394684092, + "loss": 2.8013, + "step": 886 + }, + { + "epoch": 0.113536, + "grad_norm": 0.652826726436615, + "learning_rate": 0.000981840917825769, + "loss": 2.8369, + "step": 887 + }, + { + "epoch": 0.113664, + "grad_norm": 0.71772301197052, + "learning_rate": 0.0009817855133490891, + "loss": 2.8958, + "step": 888 + }, + { + "epoch": 0.113792, + "grad_norm": 0.664911687374115, + "learning_rate": 0.000981730026047894, + "loss": 2.8067, + "step": 889 + }, + { + "epoch": 0.11392, + "grad_norm": 0.840049684047699, + "learning_rate": 0.0009816744559317228, + "loss": 2.7701, + "step": 890 + }, + { + "epoch": 0.114048, + "grad_norm": 0.6103021502494812, + "learning_rate": 0.0009816188030101285, + "loss": 2.7668, + "step": 891 + }, + { + "epoch": 0.114176, + "grad_norm": 0.6364806890487671, + "learning_rate": 0.0009815630672926788, + "loss": 2.8146, + "step": 892 + }, + { + "epoch": 0.114304, + "grad_norm": 0.6550793051719666, + "learning_rate": 0.0009815072487889549, + "loss": 2.9055, + "step": 893 + }, + { + "epoch": 0.114432, + "grad_norm": 0.6753566861152649, + "learning_rate": 0.000981451347508553, + "loss": 2.8172, + "step": 894 + }, + { + "epoch": 0.11456, + "grad_norm": 0.6634787917137146, + "learning_rate": 0.0009813953634610826, + "loss": 2.7864, + "step": 895 + }, + { + "epoch": 0.114688, + "grad_norm": 0.7429337501525879, + "learning_rate": 0.0009813392966561687, + "loss": 2.7738, + "step": 896 + }, + { + "epoch": 0.114816, + "grad_norm": 0.6683288216590881, + "learning_rate": 0.0009812831471034495, + "loss": 2.839, + "step": 897 + }, + { + "epoch": 0.114944, + "grad_norm": 0.674822986125946, + "learning_rate": 0.0009812269148125777, + "loss": 2.7721, + "step": 898 + }, + { + "epoch": 0.115072, + "grad_norm": 0.7153884172439575, + "learning_rate": 0.0009811705997932205, + "loss": 2.7594, + "step": 899 + }, + { + "epoch": 0.1152, + "grad_norm": 0.6432443857192993, + "learning_rate": 0.000981114202055059, + "loss": 2.7413, + "step": 900 + }, + { + "epoch": 0.115328, + "grad_norm": 0.6570873260498047, + "learning_rate": 0.0009810577216077885, + "loss": 2.8455, + "step": 901 + }, + { + "epoch": 0.115456, + "grad_norm": 0.64909428358078, + "learning_rate": 0.000981001158461119, + "loss": 2.8004, + "step": 902 + }, + { + "epoch": 0.115584, + "grad_norm": 0.6600486636161804, + "learning_rate": 0.0009809445126247739, + "loss": 2.7458, + "step": 903 + }, + { + "epoch": 0.115712, + "grad_norm": 0.6748130321502686, + "learning_rate": 0.0009808877841084914, + "loss": 2.8012, + "step": 904 + }, + { + "epoch": 0.11584, + "grad_norm": 0.5952267646789551, + "learning_rate": 0.000980830972922024, + "loss": 2.7863, + "step": 905 + }, + { + "epoch": 0.115968, + "grad_norm": 0.645725429058075, + "learning_rate": 0.0009807740790751382, + "loss": 2.7333, + "step": 906 + }, + { + "epoch": 0.116096, + "grad_norm": 0.7137587070465088, + "learning_rate": 0.0009807171025776145, + "loss": 2.8128, + "step": 907 + }, + { + "epoch": 0.116224, + "grad_norm": 0.647759735584259, + "learning_rate": 0.0009806600434392478, + "loss": 2.8675, + "step": 908 + }, + { + "epoch": 0.116352, + "grad_norm": 0.659572422504425, + "learning_rate": 0.0009806029016698474, + "loss": 2.7805, + "step": 909 + }, + { + "epoch": 0.11648, + "grad_norm": 0.6541268229484558, + "learning_rate": 0.0009805456772792368, + "loss": 2.8172, + "step": 910 + }, + { + "epoch": 0.116608, + "grad_norm": 0.6538609266281128, + "learning_rate": 0.0009804883702772531, + "loss": 2.7849, + "step": 911 + }, + { + "epoch": 0.116736, + "grad_norm": 0.5993735790252686, + "learning_rate": 0.0009804309806737482, + "loss": 2.778, + "step": 912 + }, + { + "epoch": 0.116864, + "grad_norm": 0.6493178606033325, + "learning_rate": 0.0009803735084785883, + "loss": 2.8267, + "step": 913 + }, + { + "epoch": 0.116992, + "grad_norm": 0.6798108816146851, + "learning_rate": 0.000980315953701653, + "loss": 2.8975, + "step": 914 + }, + { + "epoch": 0.11712, + "grad_norm": 0.7133318185806274, + "learning_rate": 0.0009802583163528373, + "loss": 2.8243, + "step": 915 + }, + { + "epoch": 0.117248, + "grad_norm": 0.6855413317680359, + "learning_rate": 0.0009802005964420493, + "loss": 2.8432, + "step": 916 + }, + { + "epoch": 0.117376, + "grad_norm": 0.6447359323501587, + "learning_rate": 0.0009801427939792118, + "loss": 2.793, + "step": 917 + }, + { + "epoch": 0.117504, + "grad_norm": 0.6987738013267517, + "learning_rate": 0.0009800849089742615, + "loss": 2.7956, + "step": 918 + }, + { + "epoch": 0.117632, + "grad_norm": 0.6707319021224976, + "learning_rate": 0.0009800269414371502, + "loss": 2.8116, + "step": 919 + }, + { + "epoch": 0.11776, + "grad_norm": 0.589859664440155, + "learning_rate": 0.000979968891377842, + "loss": 2.8821, + "step": 920 + }, + { + "epoch": 0.117888, + "grad_norm": 0.6681221127510071, + "learning_rate": 0.0009799107588063178, + "loss": 2.763, + "step": 921 + }, + { + "epoch": 0.118016, + "grad_norm": 0.6464031338691711, + "learning_rate": 0.0009798525437325703, + "loss": 2.9777, + "step": 922 + }, + { + "epoch": 0.118144, + "grad_norm": 0.6249955892562866, + "learning_rate": 0.0009797942461666075, + "loss": 2.8404, + "step": 923 + }, + { + "epoch": 0.118272, + "grad_norm": 0.7493773698806763, + "learning_rate": 0.0009797358661184517, + "loss": 2.8528, + "step": 924 + }, + { + "epoch": 0.1184, + "grad_norm": 0.651437520980835, + "learning_rate": 0.0009796774035981388, + "loss": 2.7262, + "step": 925 + }, + { + "epoch": 0.118528, + "grad_norm": 0.7157898545265198, + "learning_rate": 0.0009796188586157193, + "loss": 2.868, + "step": 926 + }, + { + "epoch": 0.118656, + "grad_norm": 0.6863674521446228, + "learning_rate": 0.000979560231181258, + "loss": 2.6468, + "step": 927 + }, + { + "epoch": 0.118784, + "grad_norm": 0.6503874063491821, + "learning_rate": 0.0009795015213048333, + "loss": 2.8357, + "step": 928 + }, + { + "epoch": 0.118912, + "grad_norm": 0.6674947738647461, + "learning_rate": 0.0009794427289965381, + "loss": 2.7295, + "step": 929 + }, + { + "epoch": 0.11904, + "grad_norm": 0.6343032121658325, + "learning_rate": 0.0009793838542664797, + "loss": 2.7392, + "step": 930 + }, + { + "epoch": 0.119168, + "grad_norm": 0.6681572794914246, + "learning_rate": 0.0009793248971247791, + "loss": 2.7561, + "step": 931 + }, + { + "epoch": 0.119296, + "grad_norm": 0.66341233253479, + "learning_rate": 0.0009792658575815724, + "loss": 2.6855, + "step": 932 + }, + { + "epoch": 0.119424, + "grad_norm": 0.6416642069816589, + "learning_rate": 0.0009792067356470082, + "loss": 2.8765, + "step": 933 + }, + { + "epoch": 0.119552, + "grad_norm": 0.6245805025100708, + "learning_rate": 0.0009791475313312508, + "loss": 2.8072, + "step": 934 + }, + { + "epoch": 0.11968, + "grad_norm": 0.6515790820121765, + "learning_rate": 0.000979088244644478, + "loss": 2.8197, + "step": 935 + }, + { + "epoch": 0.119808, + "grad_norm": 0.6365097761154175, + "learning_rate": 0.000979028875596882, + "loss": 2.8267, + "step": 936 + }, + { + "epoch": 0.119936, + "grad_norm": 0.6251488327980042, + "learning_rate": 0.0009789694241986687, + "loss": 2.7424, + "step": 937 + }, + { + "epoch": 0.120064, + "grad_norm": 0.6412323713302612, + "learning_rate": 0.0009789098904600588, + "loss": 2.767, + "step": 938 + }, + { + "epoch": 0.120192, + "grad_norm": 0.6778679490089417, + "learning_rate": 0.0009788502743912867, + "loss": 2.8081, + "step": 939 + }, + { + "epoch": 0.12032, + "grad_norm": 0.6563493609428406, + "learning_rate": 0.000978790576002601, + "loss": 2.7526, + "step": 940 + }, + { + "epoch": 0.120448, + "grad_norm": 0.644055187702179, + "learning_rate": 0.000978730795304265, + "loss": 2.7783, + "step": 941 + }, + { + "epoch": 0.120576, + "grad_norm": 0.6656532883644104, + "learning_rate": 0.0009786709323065552, + "loss": 2.8426, + "step": 942 + }, + { + "epoch": 0.120704, + "grad_norm": 0.6300017237663269, + "learning_rate": 0.0009786109870197628, + "loss": 2.8214, + "step": 943 + }, + { + "epoch": 0.120832, + "grad_norm": 0.6997824311256409, + "learning_rate": 0.0009785509594541933, + "loss": 2.78, + "step": 944 + }, + { + "epoch": 0.12096, + "grad_norm": 0.6612046957015991, + "learning_rate": 0.000978490849620166, + "loss": 2.8445, + "step": 945 + }, + { + "epoch": 0.121088, + "grad_norm": 0.6877278685569763, + "learning_rate": 0.0009784306575280144, + "loss": 2.7874, + "step": 946 + }, + { + "epoch": 0.121216, + "grad_norm": 0.6226553320884705, + "learning_rate": 0.0009783703831880866, + "loss": 2.7485, + "step": 947 + }, + { + "epoch": 0.121344, + "grad_norm": 0.6674754619598389, + "learning_rate": 0.000978310026610744, + "loss": 2.7538, + "step": 948 + }, + { + "epoch": 0.121472, + "grad_norm": 0.6339161992073059, + "learning_rate": 0.0009782495878063629, + "loss": 2.7222, + "step": 949 + }, + { + "epoch": 0.1216, + "grad_norm": 0.6263079643249512, + "learning_rate": 0.000978189066785333, + "loss": 2.8152, + "step": 950 + }, + { + "epoch": 0.121728, + "grad_norm": 0.6970793604850769, + "learning_rate": 0.0009781284635580591, + "loss": 2.7695, + "step": 951 + }, + { + "epoch": 0.121856, + "grad_norm": 0.712485134601593, + "learning_rate": 0.0009780677781349595, + "loss": 2.8133, + "step": 952 + }, + { + "epoch": 0.121984, + "grad_norm": 0.7340429425239563, + "learning_rate": 0.0009780070105264666, + "loss": 2.7361, + "step": 953 + }, + { + "epoch": 0.122112, + "grad_norm": 0.6385213732719421, + "learning_rate": 0.0009779461607430267, + "loss": 2.9074, + "step": 954 + }, + { + "epoch": 0.12224, + "grad_norm": 0.7070710062980652, + "learning_rate": 0.0009778852287951013, + "loss": 2.8242, + "step": 955 + }, + { + "epoch": 0.122368, + "grad_norm": 0.6693868041038513, + "learning_rate": 0.000977824214693165, + "loss": 2.8353, + "step": 956 + }, + { + "epoch": 0.122496, + "grad_norm": 0.6536783576011658, + "learning_rate": 0.0009777631184477065, + "loss": 2.7421, + "step": 957 + }, + { + "epoch": 0.122624, + "grad_norm": 0.6139789819717407, + "learning_rate": 0.0009777019400692291, + "loss": 2.7547, + "step": 958 + }, + { + "epoch": 0.122752, + "grad_norm": 0.6508567929267883, + "learning_rate": 0.0009776406795682505, + "loss": 2.7931, + "step": 959 + }, + { + "epoch": 0.12288, + "grad_norm": 0.6478503942489624, + "learning_rate": 0.0009775793369553017, + "loss": 2.7575, + "step": 960 + }, + { + "epoch": 0.123008, + "grad_norm": 0.6180828213691711, + "learning_rate": 0.0009775179122409284, + "loss": 2.7397, + "step": 961 + }, + { + "epoch": 0.123136, + "grad_norm": 0.6681183576583862, + "learning_rate": 0.00097745640543569, + "loss": 2.7475, + "step": 962 + }, + { + "epoch": 0.123264, + "grad_norm": 0.6023115515708923, + "learning_rate": 0.0009773948165501605, + "loss": 2.7286, + "step": 963 + }, + { + "epoch": 0.123392, + "grad_norm": 0.6685054302215576, + "learning_rate": 0.0009773331455949275, + "loss": 2.7445, + "step": 964 + }, + { + "epoch": 0.12352, + "grad_norm": 0.717911958694458, + "learning_rate": 0.000977271392580593, + "loss": 2.8414, + "step": 965 + }, + { + "epoch": 0.123648, + "grad_norm": 0.6953067779541016, + "learning_rate": 0.0009772095575177732, + "loss": 2.8082, + "step": 966 + }, + { + "epoch": 0.123776, + "grad_norm": 0.7320588827133179, + "learning_rate": 0.0009771476404170983, + "loss": 2.8264, + "step": 967 + }, + { + "epoch": 0.123904, + "grad_norm": 0.7053120136260986, + "learning_rate": 0.0009770856412892122, + "loss": 2.8206, + "step": 968 + }, + { + "epoch": 0.124032, + "grad_norm": 0.6515536904335022, + "learning_rate": 0.0009770235601447737, + "loss": 2.8874, + "step": 969 + }, + { + "epoch": 0.12416, + "grad_norm": 0.6852850914001465, + "learning_rate": 0.0009769613969944548, + "loss": 2.8051, + "step": 970 + }, + { + "epoch": 0.124288, + "grad_norm": 0.674262523651123, + "learning_rate": 0.0009768991518489425, + "loss": 2.762, + "step": 971 + }, + { + "epoch": 0.124416, + "grad_norm": 0.6374111175537109, + "learning_rate": 0.0009768368247189375, + "loss": 2.7824, + "step": 972 + }, + { + "epoch": 0.124544, + "grad_norm": 0.6575995683670044, + "learning_rate": 0.0009767744156151542, + "loss": 2.7471, + "step": 973 + }, + { + "epoch": 0.124672, + "grad_norm": 0.6892423629760742, + "learning_rate": 0.0009767119245483215, + "loss": 2.8076, + "step": 974 + }, + { + "epoch": 0.1248, + "grad_norm": 0.7398384809494019, + "learning_rate": 0.0009766493515291827, + "loss": 2.9103, + "step": 975 + }, + { + "epoch": 0.124928, + "grad_norm": 0.7400169372558594, + "learning_rate": 0.0009765866965684945, + "loss": 2.7363, + "step": 976 + }, + { + "epoch": 0.125056, + "grad_norm": 0.5914911031723022, + "learning_rate": 0.0009765239596770279, + "loss": 2.8261, + "step": 977 + }, + { + "epoch": 0.125184, + "grad_norm": 0.6973153352737427, + "learning_rate": 0.0009764611408655685, + "loss": 2.8456, + "step": 978 + }, + { + "epoch": 0.125312, + "grad_norm": 0.703074038028717, + "learning_rate": 0.0009763982401449153, + "loss": 2.841, + "step": 979 + }, + { + "epoch": 0.12544, + "grad_norm": 0.6340558528900146, + "learning_rate": 0.0009763352575258817, + "loss": 2.7828, + "step": 980 + }, + { + "epoch": 0.125568, + "grad_norm": 0.659621000289917, + "learning_rate": 0.0009762721930192952, + "loss": 2.7667, + "step": 981 + }, + { + "epoch": 0.125696, + "grad_norm": 0.6087130904197693, + "learning_rate": 0.0009762090466359974, + "loss": 2.7526, + "step": 982 + }, + { + "epoch": 0.125824, + "grad_norm": 0.737118661403656, + "learning_rate": 0.0009761458183868437, + "loss": 2.8201, + "step": 983 + }, + { + "epoch": 0.125952, + "grad_norm": 0.7633397579193115, + "learning_rate": 0.0009760825082827039, + "loss": 2.7909, + "step": 984 + }, + { + "epoch": 0.12608, + "grad_norm": 0.6944820284843445, + "learning_rate": 0.0009760191163344617, + "loss": 2.7003, + "step": 985 + }, + { + "epoch": 0.126208, + "grad_norm": 0.7229971289634705, + "learning_rate": 0.000975955642553015, + "loss": 2.7288, + "step": 986 + }, + { + "epoch": 0.126336, + "grad_norm": 0.7411198019981384, + "learning_rate": 0.0009758920869492754, + "loss": 2.7737, + "step": 987 + }, + { + "epoch": 0.126464, + "grad_norm": 0.6268801093101501, + "learning_rate": 0.0009758284495341692, + "loss": 2.7942, + "step": 988 + }, + { + "epoch": 0.126592, + "grad_norm": 0.6473471522331238, + "learning_rate": 0.0009757647303186361, + "loss": 2.82, + "step": 989 + }, + { + "epoch": 0.12672, + "grad_norm": 0.6369385123252869, + "learning_rate": 0.0009757009293136303, + "loss": 2.7601, + "step": 990 + }, + { + "epoch": 0.126848, + "grad_norm": 0.6501265168190002, + "learning_rate": 0.00097563704653012, + "loss": 2.7406, + "step": 991 + }, + { + "epoch": 0.126976, + "grad_norm": 0.6793976426124573, + "learning_rate": 0.0009755730819790876, + "loss": 2.8869, + "step": 992 + }, + { + "epoch": 0.127104, + "grad_norm": 0.6049587726593018, + "learning_rate": 0.0009755090356715288, + "loss": 2.8207, + "step": 993 + }, + { + "epoch": 0.127232, + "grad_norm": 0.6973533034324646, + "learning_rate": 0.0009754449076184543, + "loss": 2.7578, + "step": 994 + }, + { + "epoch": 0.12736, + "grad_norm": 0.6367203593254089, + "learning_rate": 0.0009753806978308883, + "loss": 2.8889, + "step": 995 + }, + { + "epoch": 0.127488, + "grad_norm": 0.6487019658088684, + "learning_rate": 0.0009753164063198693, + "loss": 2.7837, + "step": 996 + }, + { + "epoch": 0.127616, + "grad_norm": 0.6247488260269165, + "learning_rate": 0.0009752520330964496, + "loss": 2.7254, + "step": 997 + }, + { + "epoch": 0.127744, + "grad_norm": 0.597643256187439, + "learning_rate": 0.000975187578171696, + "loss": 2.7174, + "step": 998 + }, + { + "epoch": 0.127872, + "grad_norm": 0.6663332581520081, + "learning_rate": 0.0009751230415566887, + "loss": 2.9881, + "step": 999 + }, + { + "epoch": 0.128, + "grad_norm": 0.6311107873916626, + "learning_rate": 0.0009750584232625226, + "loss": 2.8583, + "step": 1000 + }, + { + "epoch": 0.128128, + "grad_norm": 0.6396372318267822, + "learning_rate": 0.0009749937233003061, + "loss": 2.6397, + "step": 1001 + }, + { + "epoch": 0.128256, + "grad_norm": 0.6171582937240601, + "learning_rate": 0.0009749289416811621, + "loss": 2.7709, + "step": 1002 + }, + { + "epoch": 0.128384, + "grad_norm": 0.6234565377235413, + "learning_rate": 0.000974864078416227, + "loss": 2.819, + "step": 1003 + }, + { + "epoch": 0.128512, + "grad_norm": 0.6170878410339355, + "learning_rate": 0.0009747991335166519, + "loss": 2.7638, + "step": 1004 + }, + { + "epoch": 0.12864, + "grad_norm": 0.6904697418212891, + "learning_rate": 0.0009747341069936013, + "loss": 2.7552, + "step": 1005 + }, + { + "epoch": 0.128768, + "grad_norm": 0.6638295650482178, + "learning_rate": 0.000974668998858254, + "loss": 2.7845, + "step": 1006 + }, + { + "epoch": 0.128896, + "grad_norm": 0.6663033366203308, + "learning_rate": 0.0009746038091218031, + "loss": 2.833, + "step": 1007 + }, + { + "epoch": 0.129024, + "grad_norm": 0.6373790502548218, + "learning_rate": 0.0009745385377954555, + "loss": 2.8155, + "step": 1008 + }, + { + "epoch": 0.129152, + "grad_norm": 0.6730905771255493, + "learning_rate": 0.0009744731848904315, + "loss": 2.772, + "step": 1009 + }, + { + "epoch": 0.12928, + "grad_norm": 0.6039687395095825, + "learning_rate": 0.0009744077504179669, + "loss": 2.7551, + "step": 1010 + }, + { + "epoch": 0.129408, + "grad_norm": 0.6682738661766052, + "learning_rate": 0.0009743422343893098, + "loss": 2.8094, + "step": 1011 + }, + { + "epoch": 0.129536, + "grad_norm": 0.6548439264297485, + "learning_rate": 0.0009742766368157239, + "loss": 2.762, + "step": 1012 + }, + { + "epoch": 0.129664, + "grad_norm": 0.6597762703895569, + "learning_rate": 0.0009742109577084856, + "loss": 2.847, + "step": 1013 + }, + { + "epoch": 0.129792, + "grad_norm": 0.6590989828109741, + "learning_rate": 0.0009741451970788862, + "loss": 2.7083, + "step": 1014 + }, + { + "epoch": 0.12992, + "grad_norm": 0.6737598776817322, + "learning_rate": 0.0009740793549382307, + "loss": 2.9241, + "step": 1015 + }, + { + "epoch": 0.130048, + "grad_norm": 0.6716289520263672, + "learning_rate": 0.0009740134312978382, + "loss": 2.811, + "step": 1016 + }, + { + "epoch": 0.130176, + "grad_norm": 0.6303693056106567, + "learning_rate": 0.0009739474261690415, + "loss": 2.8129, + "step": 1017 + }, + { + "epoch": 0.130304, + "grad_norm": 0.7026432156562805, + "learning_rate": 0.0009738813395631878, + "loss": 2.8535, + "step": 1018 + }, + { + "epoch": 0.130432, + "grad_norm": 0.7056198716163635, + "learning_rate": 0.0009738151714916381, + "loss": 2.6313, + "step": 1019 + }, + { + "epoch": 0.13056, + "grad_norm": 0.6290691494941711, + "learning_rate": 0.0009737489219657675, + "loss": 2.7692, + "step": 1020 + }, + { + "epoch": 0.130688, + "grad_norm": 0.6588796377182007, + "learning_rate": 0.000973682590996965, + "loss": 2.7381, + "step": 1021 + }, + { + "epoch": 0.130816, + "grad_norm": 0.6618863344192505, + "learning_rate": 0.0009736161785966337, + "loss": 2.8805, + "step": 1022 + }, + { + "epoch": 0.130944, + "grad_norm": 0.677842915058136, + "learning_rate": 0.0009735496847761908, + "loss": 2.772, + "step": 1023 + }, + { + "epoch": 0.131072, + "grad_norm": 0.8161168098449707, + "learning_rate": 0.0009734831095470671, + "loss": 2.7254, + "step": 1024 + }, + { + "epoch": 0.1312, + "grad_norm": 0.6861282587051392, + "learning_rate": 0.0009734164529207079, + "loss": 2.7473, + "step": 1025 + }, + { + "epoch": 0.131328, + "grad_norm": 0.8163470029830933, + "learning_rate": 0.0009733497149085719, + "loss": 2.7431, + "step": 1026 + }, + { + "epoch": 0.131456, + "grad_norm": 0.6742125749588013, + "learning_rate": 0.0009732828955221327, + "loss": 2.7688, + "step": 1027 + }, + { + "epoch": 0.131584, + "grad_norm": 0.7708075642585754, + "learning_rate": 0.0009732159947728767, + "loss": 2.7619, + "step": 1028 + }, + { + "epoch": 0.131712, + "grad_norm": 0.6749203205108643, + "learning_rate": 0.0009731490126723052, + "loss": 2.7254, + "step": 1029 + }, + { + "epoch": 0.13184, + "grad_norm": 0.6552834510803223, + "learning_rate": 0.0009730819492319335, + "loss": 2.8324, + "step": 1030 + }, + { + "epoch": 0.131968, + "grad_norm": 0.6241470575332642, + "learning_rate": 0.0009730148044632901, + "loss": 2.8069, + "step": 1031 + }, + { + "epoch": 0.132096, + "grad_norm": 0.6623492240905762, + "learning_rate": 0.0009729475783779181, + "loss": 2.8324, + "step": 1032 + }, + { + "epoch": 0.132224, + "grad_norm": 0.6525461673736572, + "learning_rate": 0.0009728802709873747, + "loss": 2.6984, + "step": 1033 + }, + { + "epoch": 0.132352, + "grad_norm": 0.6781845688819885, + "learning_rate": 0.0009728128823032305, + "loss": 2.8353, + "step": 1034 + }, + { + "epoch": 0.13248, + "grad_norm": 0.6478338837623596, + "learning_rate": 0.0009727454123370705, + "loss": 2.7882, + "step": 1035 + }, + { + "epoch": 0.132608, + "grad_norm": 0.6432681679725647, + "learning_rate": 0.0009726778611004935, + "loss": 2.837, + "step": 1036 + }, + { + "epoch": 0.132736, + "grad_norm": 0.631017804145813, + "learning_rate": 0.0009726102286051127, + "loss": 2.7962, + "step": 1037 + }, + { + "epoch": 0.132864, + "grad_norm": 0.6605352759361267, + "learning_rate": 0.0009725425148625542, + "loss": 2.7579, + "step": 1038 + }, + { + "epoch": 0.132992, + "grad_norm": 0.6587074398994446, + "learning_rate": 0.0009724747198844595, + "loss": 2.8714, + "step": 1039 + }, + { + "epoch": 0.13312, + "grad_norm": 0.6540379524230957, + "learning_rate": 0.0009724068436824831, + "loss": 2.8074, + "step": 1040 + }, + { + "epoch": 0.133248, + "grad_norm": 0.6477424502372742, + "learning_rate": 0.0009723388862682936, + "loss": 2.6658, + "step": 1041 + }, + { + "epoch": 0.133376, + "grad_norm": 0.7079359292984009, + "learning_rate": 0.0009722708476535736, + "loss": 2.8796, + "step": 1042 + }, + { + "epoch": 0.133504, + "grad_norm": 0.7030080556869507, + "learning_rate": 0.0009722027278500201, + "loss": 2.8629, + "step": 1043 + }, + { + "epoch": 0.133632, + "grad_norm": 0.639582633972168, + "learning_rate": 0.0009721345268693432, + "loss": 2.7928, + "step": 1044 + }, + { + "epoch": 0.13376, + "grad_norm": 0.647226095199585, + "learning_rate": 0.0009720662447232677, + "loss": 2.8229, + "step": 1045 + }, + { + "epoch": 0.133888, + "grad_norm": 0.6455651521682739, + "learning_rate": 0.0009719978814235322, + "loss": 2.848, + "step": 1046 + }, + { + "epoch": 0.134016, + "grad_norm": 0.6383050084114075, + "learning_rate": 0.000971929436981889, + "loss": 2.8001, + "step": 1047 + }, + { + "epoch": 0.134144, + "grad_norm": 0.6407660841941833, + "learning_rate": 0.0009718609114101045, + "loss": 2.7806, + "step": 1048 + }, + { + "epoch": 0.134272, + "grad_norm": 0.6269516944885254, + "learning_rate": 0.0009717923047199591, + "loss": 2.6748, + "step": 1049 + }, + { + "epoch": 0.1344, + "grad_norm": 0.6194528341293335, + "learning_rate": 0.0009717236169232468, + "loss": 2.854, + "step": 1050 + }, + { + "epoch": 0.134528, + "grad_norm": 0.61770099401474, + "learning_rate": 0.0009716548480317763, + "loss": 2.718, + "step": 1051 + }, + { + "epoch": 0.134656, + "grad_norm": 0.6993018984794617, + "learning_rate": 0.0009715859980573694, + "loss": 2.7862, + "step": 1052 + }, + { + "epoch": 0.134784, + "grad_norm": 0.6865027546882629, + "learning_rate": 0.0009715170670118624, + "loss": 2.6649, + "step": 1053 + }, + { + "epoch": 0.134912, + "grad_norm": 0.6332573890686035, + "learning_rate": 0.0009714480549071052, + "loss": 2.8242, + "step": 1054 + }, + { + "epoch": 0.13504, + "grad_norm": 0.6089667677879333, + "learning_rate": 0.000971378961754962, + "loss": 2.7674, + "step": 1055 + }, + { + "epoch": 0.135168, + "grad_norm": 0.6154025197029114, + "learning_rate": 0.0009713097875673105, + "loss": 2.8201, + "step": 1056 + }, + { + "epoch": 0.135296, + "grad_norm": 0.6062637567520142, + "learning_rate": 0.0009712405323560427, + "loss": 2.7369, + "step": 1057 + }, + { + "epoch": 0.135424, + "grad_norm": 0.6019504070281982, + "learning_rate": 0.0009711711961330643, + "loss": 2.7938, + "step": 1058 + }, + { + "epoch": 0.135552, + "grad_norm": 0.7028970718383789, + "learning_rate": 0.0009711017789102948, + "loss": 2.8732, + "step": 1059 + }, + { + "epoch": 0.13568, + "grad_norm": 0.6612076163291931, + "learning_rate": 0.0009710322806996682, + "loss": 2.8437, + "step": 1060 + }, + { + "epoch": 0.135808, + "grad_norm": 0.6898564100265503, + "learning_rate": 0.0009709627015131319, + "loss": 2.8254, + "step": 1061 + }, + { + "epoch": 0.135936, + "grad_norm": 0.663036048412323, + "learning_rate": 0.0009708930413626473, + "loss": 2.8254, + "step": 1062 + }, + { + "epoch": 0.136064, + "grad_norm": 0.6595954298973083, + "learning_rate": 0.0009708233002601898, + "loss": 2.7632, + "step": 1063 + }, + { + "epoch": 0.136192, + "grad_norm": 0.6536065936088562, + "learning_rate": 0.0009707534782177487, + "loss": 2.7966, + "step": 1064 + }, + { + "epoch": 0.13632, + "grad_norm": 0.6059527397155762, + "learning_rate": 0.0009706835752473273, + "loss": 2.8212, + "step": 1065 + }, + { + "epoch": 0.136448, + "grad_norm": 0.7033224701881409, + "learning_rate": 0.0009706135913609426, + "loss": 2.7958, + "step": 1066 + }, + { + "epoch": 0.136576, + "grad_norm": 0.6728179454803467, + "learning_rate": 0.0009705435265706258, + "loss": 2.7559, + "step": 1067 + }, + { + "epoch": 0.136704, + "grad_norm": 0.6774322390556335, + "learning_rate": 0.0009704733808884218, + "loss": 2.8125, + "step": 1068 + }, + { + "epoch": 0.136832, + "grad_norm": 0.6221113204956055, + "learning_rate": 0.0009704031543263893, + "loss": 2.728, + "step": 1069 + }, + { + "epoch": 0.13696, + "grad_norm": 0.6364342570304871, + "learning_rate": 0.0009703328468966015, + "loss": 2.8536, + "step": 1070 + }, + { + "epoch": 0.137088, + "grad_norm": 0.6841943264007568, + "learning_rate": 0.0009702624586111445, + "loss": 2.7347, + "step": 1071 + }, + { + "epoch": 0.137216, + "grad_norm": 0.6921083331108093, + "learning_rate": 0.0009701919894821192, + "loss": 2.8149, + "step": 1072 + }, + { + "epoch": 0.137344, + "grad_norm": 0.626365065574646, + "learning_rate": 0.0009701214395216399, + "loss": 2.768, + "step": 1073 + }, + { + "epoch": 0.137472, + "grad_norm": 0.6088806390762329, + "learning_rate": 0.0009700508087418351, + "loss": 2.772, + "step": 1074 + }, + { + "epoch": 0.1376, + "grad_norm": 0.6759106516838074, + "learning_rate": 0.0009699800971548469, + "loss": 2.7984, + "step": 1075 + }, + { + "epoch": 0.137728, + "grad_norm": 0.6283963322639465, + "learning_rate": 0.0009699093047728316, + "loss": 2.8533, + "step": 1076 + }, + { + "epoch": 0.137856, + "grad_norm": 0.633419394493103, + "learning_rate": 0.0009698384316079591, + "loss": 2.8434, + "step": 1077 + }, + { + "epoch": 0.137984, + "grad_norm": 0.5863417983055115, + "learning_rate": 0.0009697674776724135, + "loss": 2.7536, + "step": 1078 + }, + { + "epoch": 0.138112, + "grad_norm": 0.6459261178970337, + "learning_rate": 0.0009696964429783923, + "loss": 2.7936, + "step": 1079 + }, + { + "epoch": 0.13824, + "grad_norm": 0.7249024510383606, + "learning_rate": 0.0009696253275381074, + "loss": 2.8038, + "step": 1080 + }, + { + "epoch": 0.138368, + "grad_norm": 0.6486040949821472, + "learning_rate": 0.0009695541313637843, + "loss": 2.7557, + "step": 1081 + }, + { + "epoch": 0.138496, + "grad_norm": 0.6441357731819153, + "learning_rate": 0.0009694828544676626, + "loss": 2.8585, + "step": 1082 + }, + { + "epoch": 0.138624, + "grad_norm": 0.6527015566825867, + "learning_rate": 0.0009694114968619954, + "loss": 2.7669, + "step": 1083 + }, + { + "epoch": 0.138752, + "grad_norm": 0.6376598477363586, + "learning_rate": 0.0009693400585590501, + "loss": 2.6872, + "step": 1084 + }, + { + "epoch": 0.13888, + "grad_norm": 0.5876955389976501, + "learning_rate": 0.0009692685395711076, + "loss": 2.6878, + "step": 1085 + }, + { + "epoch": 0.139008, + "grad_norm": 0.59808349609375, + "learning_rate": 0.000969196939910463, + "loss": 2.7934, + "step": 1086 + }, + { + "epoch": 0.139136, + "grad_norm": 0.6211817264556885, + "learning_rate": 0.000969125259589425, + "loss": 2.6362, + "step": 1087 + }, + { + "epoch": 0.139264, + "grad_norm": 0.7010995149612427, + "learning_rate": 0.0009690534986203164, + "loss": 2.7979, + "step": 1088 + }, + { + "epoch": 0.139392, + "grad_norm": 0.6464605331420898, + "learning_rate": 0.0009689816570154734, + "loss": 2.7518, + "step": 1089 + }, + { + "epoch": 0.13952, + "grad_norm": 0.6146900057792664, + "learning_rate": 0.0009689097347872469, + "loss": 2.6993, + "step": 1090 + }, + { + "epoch": 0.139648, + "grad_norm": 0.6473577618598938, + "learning_rate": 0.0009688377319480009, + "loss": 2.6823, + "step": 1091 + }, + { + "epoch": 0.139776, + "grad_norm": 0.7197511792182922, + "learning_rate": 0.0009687656485101132, + "loss": 2.7708, + "step": 1092 + }, + { + "epoch": 0.139904, + "grad_norm": 0.6540927290916443, + "learning_rate": 0.0009686934844859765, + "loss": 2.77, + "step": 1093 + }, + { + "epoch": 0.140032, + "grad_norm": 0.6862616539001465, + "learning_rate": 0.0009686212398879962, + "loss": 2.9517, + "step": 1094 + }, + { + "epoch": 0.14016, + "grad_norm": 0.6811481714248657, + "learning_rate": 0.000968548914728592, + "loss": 2.8188, + "step": 1095 + }, + { + "epoch": 0.140288, + "grad_norm": 0.6409133672714233, + "learning_rate": 0.0009684765090201974, + "loss": 2.7447, + "step": 1096 + }, + { + "epoch": 0.140416, + "grad_norm": 0.6580144762992859, + "learning_rate": 0.0009684040227752599, + "loss": 2.848, + "step": 1097 + }, + { + "epoch": 0.140544, + "grad_norm": 0.6062847375869751, + "learning_rate": 0.0009683314560062408, + "loss": 2.6222, + "step": 1098 + }, + { + "epoch": 0.140672, + "grad_norm": 0.6041891574859619, + "learning_rate": 0.0009682588087256148, + "loss": 2.8284, + "step": 1099 + }, + { + "epoch": 0.1408, + "grad_norm": 0.6145558953285217, + "learning_rate": 0.0009681860809458713, + "loss": 2.7529, + "step": 1100 + }, + { + "epoch": 0.140928, + "grad_norm": 0.6040167808532715, + "learning_rate": 0.0009681132726795128, + "loss": 2.7959, + "step": 1101 + }, + { + "epoch": 0.141056, + "grad_norm": 0.6365892291069031, + "learning_rate": 0.0009680403839390557, + "loss": 2.7487, + "step": 1102 + }, + { + "epoch": 0.141184, + "grad_norm": 0.57224440574646, + "learning_rate": 0.0009679674147370308, + "loss": 2.801, + "step": 1103 + }, + { + "epoch": 0.141312, + "grad_norm": 0.6292387843132019, + "learning_rate": 0.0009678943650859819, + "loss": 2.7962, + "step": 1104 + }, + { + "epoch": 0.14144, + "grad_norm": 0.622048556804657, + "learning_rate": 0.0009678212349984676, + "loss": 2.6481, + "step": 1105 + }, + { + "epoch": 0.141568, + "grad_norm": 0.586182177066803, + "learning_rate": 0.0009677480244870596, + "loss": 2.7329, + "step": 1106 + }, + { + "epoch": 0.141696, + "grad_norm": 0.7933721542358398, + "learning_rate": 0.0009676747335643435, + "loss": 2.7707, + "step": 1107 + }, + { + "epoch": 0.141824, + "grad_norm": 0.613935112953186, + "learning_rate": 0.0009676013622429187, + "loss": 2.7915, + "step": 1108 + }, + { + "epoch": 0.141952, + "grad_norm": 0.6614904999732971, + "learning_rate": 0.0009675279105353992, + "loss": 2.7757, + "step": 1109 + }, + { + "epoch": 0.14208, + "grad_norm": 0.6214826703071594, + "learning_rate": 0.0009674543784544117, + "loss": 2.8229, + "step": 1110 + }, + { + "epoch": 0.142208, + "grad_norm": 0.7020379304885864, + "learning_rate": 0.0009673807660125973, + "loss": 2.7546, + "step": 1111 + }, + { + "epoch": 0.142336, + "grad_norm": 0.6840025782585144, + "learning_rate": 0.0009673070732226108, + "loss": 2.6851, + "step": 1112 + }, + { + "epoch": 0.142464, + "grad_norm": 0.6607086062431335, + "learning_rate": 0.000967233300097121, + "loss": 2.7353, + "step": 1113 + }, + { + "epoch": 0.142592, + "grad_norm": 0.7151747345924377, + "learning_rate": 0.0009671594466488104, + "loss": 2.7046, + "step": 1114 + }, + { + "epoch": 0.14272, + "grad_norm": 0.6117125153541565, + "learning_rate": 0.000967085512890375, + "loss": 2.7835, + "step": 1115 + }, + { + "epoch": 0.142848, + "grad_norm": 0.6023160815238953, + "learning_rate": 0.000967011498834525, + "loss": 2.6878, + "step": 1116 + }, + { + "epoch": 0.142976, + "grad_norm": 0.6244572997093201, + "learning_rate": 0.0009669374044939845, + "loss": 2.7405, + "step": 1117 + }, + { + "epoch": 0.143104, + "grad_norm": 0.5718654990196228, + "learning_rate": 0.0009668632298814908, + "loss": 2.7043, + "step": 1118 + }, + { + "epoch": 0.143232, + "grad_norm": 0.5859500765800476, + "learning_rate": 0.0009667889750097956, + "loss": 2.7149, + "step": 1119 + }, + { + "epoch": 0.14336, + "grad_norm": 0.5724707245826721, + "learning_rate": 0.000966714639891664, + "loss": 2.8116, + "step": 1120 + }, + { + "epoch": 0.143488, + "grad_norm": 0.6986823081970215, + "learning_rate": 0.0009666402245398754, + "loss": 2.8539, + "step": 1121 + }, + { + "epoch": 0.143616, + "grad_norm": 0.6466898918151855, + "learning_rate": 0.0009665657289672222, + "loss": 2.7097, + "step": 1122 + }, + { + "epoch": 0.143744, + "grad_norm": 0.6617648601531982, + "learning_rate": 0.0009664911531865114, + "loss": 2.8067, + "step": 1123 + }, + { + "epoch": 0.143872, + "grad_norm": 0.6437375545501709, + "learning_rate": 0.0009664164972105633, + "loss": 2.8054, + "step": 1124 + }, + { + "epoch": 0.144, + "grad_norm": 0.6362265944480896, + "learning_rate": 0.0009663417610522123, + "loss": 2.8661, + "step": 1125 + }, + { + "epoch": 0.144128, + "grad_norm": 0.5870741605758667, + "learning_rate": 0.0009662669447243061, + "loss": 2.764, + "step": 1126 + }, + { + "epoch": 0.144256, + "grad_norm": 0.6248798370361328, + "learning_rate": 0.0009661920482397068, + "loss": 2.7285, + "step": 1127 + }, + { + "epoch": 0.144384, + "grad_norm": 0.800068199634552, + "learning_rate": 0.0009661170716112897, + "loss": 2.7794, + "step": 1128 + }, + { + "epoch": 0.144512, + "grad_norm": 0.5620113015174866, + "learning_rate": 0.0009660420148519444, + "loss": 2.8058, + "step": 1129 + }, + { + "epoch": 0.14464, + "grad_norm": 0.655349612236023, + "learning_rate": 0.000965966877974574, + "loss": 2.7739, + "step": 1130 + }, + { + "epoch": 0.144768, + "grad_norm": 0.626011848449707, + "learning_rate": 0.0009658916609920951, + "loss": 2.8322, + "step": 1131 + }, + { + "epoch": 0.144896, + "grad_norm": 0.6717645525932312, + "learning_rate": 0.0009658163639174389, + "loss": 2.754, + "step": 1132 + }, + { + "epoch": 0.145024, + "grad_norm": 0.7158605456352234, + "learning_rate": 0.0009657409867635492, + "loss": 2.7654, + "step": 1133 + }, + { + "epoch": 0.145152, + "grad_norm": 0.5941844582557678, + "learning_rate": 0.0009656655295433848, + "loss": 2.8594, + "step": 1134 + }, + { + "epoch": 0.14528, + "grad_norm": 0.6939860582351685, + "learning_rate": 0.0009655899922699174, + "loss": 2.8006, + "step": 1135 + }, + { + "epoch": 0.145408, + "grad_norm": 0.8847682476043701, + "learning_rate": 0.0009655143749561325, + "loss": 2.7621, + "step": 1136 + }, + { + "epoch": 0.145536, + "grad_norm": 0.6102097630500793, + "learning_rate": 0.0009654386776150302, + "loss": 2.7634, + "step": 1137 + }, + { + "epoch": 0.145664, + "grad_norm": 0.599346935749054, + "learning_rate": 0.0009653629002596231, + "loss": 2.7016, + "step": 1138 + }, + { + "epoch": 0.145792, + "grad_norm": 0.7723658680915833, + "learning_rate": 0.0009652870429029386, + "loss": 2.6794, + "step": 1139 + }, + { + "epoch": 0.14592, + "grad_norm": 0.6448710560798645, + "learning_rate": 0.0009652111055580174, + "loss": 2.8012, + "step": 1140 + }, + { + "epoch": 0.146048, + "grad_norm": 0.6397614479064941, + "learning_rate": 0.0009651350882379139, + "loss": 2.7109, + "step": 1141 + }, + { + "epoch": 0.146176, + "grad_norm": 0.6409506797790527, + "learning_rate": 0.0009650589909556964, + "loss": 2.9143, + "step": 1142 + }, + { + "epoch": 0.146304, + "grad_norm": 0.6069795489311218, + "learning_rate": 0.000964982813724447, + "loss": 2.785, + "step": 1143 + }, + { + "epoch": 0.146432, + "grad_norm": 0.7140107154846191, + "learning_rate": 0.0009649065565572614, + "loss": 2.8721, + "step": 1144 + }, + { + "epoch": 0.14656, + "grad_norm": 0.6629038453102112, + "learning_rate": 0.0009648302194672489, + "loss": 2.8241, + "step": 1145 + }, + { + "epoch": 0.146688, + "grad_norm": 0.6241483688354492, + "learning_rate": 0.0009647538024675331, + "loss": 2.8294, + "step": 1146 + }, + { + "epoch": 0.146816, + "grad_norm": 0.6423130631446838, + "learning_rate": 0.0009646773055712506, + "loss": 2.7404, + "step": 1147 + }, + { + "epoch": 0.146944, + "grad_norm": 0.5670340657234192, + "learning_rate": 0.0009646007287915523, + "loss": 2.7156, + "step": 1148 + }, + { + "epoch": 0.147072, + "grad_norm": 0.6382002830505371, + "learning_rate": 0.0009645240721416028, + "loss": 2.7945, + "step": 1149 + }, + { + "epoch": 0.1472, + "grad_norm": 0.6566773653030396, + "learning_rate": 0.00096444733563458, + "loss": 2.9014, + "step": 1150 + }, + { + "epoch": 0.147328, + "grad_norm": 0.6417313814163208, + "learning_rate": 0.0009643705192836758, + "loss": 2.7864, + "step": 1151 + }, + { + "epoch": 0.147456, + "grad_norm": 0.5897763967514038, + "learning_rate": 0.0009642936231020958, + "loss": 2.8688, + "step": 1152 + }, + { + "epoch": 0.147584, + "grad_norm": 0.5739684104919434, + "learning_rate": 0.0009642166471030596, + "loss": 2.7414, + "step": 1153 + }, + { + "epoch": 0.147712, + "grad_norm": 0.6190776824951172, + "learning_rate": 0.0009641395912998, + "loss": 2.8119, + "step": 1154 + }, + { + "epoch": 0.14784, + "grad_norm": 0.6303085684776306, + "learning_rate": 0.0009640624557055639, + "loss": 2.8086, + "step": 1155 + }, + { + "epoch": 0.147968, + "grad_norm": 0.5856841802597046, + "learning_rate": 0.0009639852403336118, + "loss": 2.8147, + "step": 1156 + }, + { + "epoch": 0.148096, + "grad_norm": 0.6192725896835327, + "learning_rate": 0.0009639079451972179, + "loss": 2.7368, + "step": 1157 + }, + { + "epoch": 0.148224, + "grad_norm": 0.6907121539115906, + "learning_rate": 0.0009638305703096701, + "loss": 2.8598, + "step": 1158 + }, + { + "epoch": 0.148352, + "grad_norm": 0.6590476632118225, + "learning_rate": 0.0009637531156842702, + "loss": 2.7657, + "step": 1159 + }, + { + "epoch": 0.14848, + "grad_norm": 0.5847887992858887, + "learning_rate": 0.0009636755813343334, + "loss": 2.8395, + "step": 1160 + }, + { + "epoch": 0.148608, + "grad_norm": 0.6158317923545837, + "learning_rate": 0.0009635979672731887, + "loss": 2.7726, + "step": 1161 + }, + { + "epoch": 0.148736, + "grad_norm": 0.6245673894882202, + "learning_rate": 0.000963520273514179, + "loss": 2.7752, + "step": 1162 + }, + { + "epoch": 0.148864, + "grad_norm": 0.6345449090003967, + "learning_rate": 0.0009634425000706607, + "loss": 2.8313, + "step": 1163 + }, + { + "epoch": 0.148992, + "grad_norm": 0.5736799240112305, + "learning_rate": 0.0009633646469560039, + "loss": 2.6588, + "step": 1164 + }, + { + "epoch": 0.14912, + "grad_norm": 0.590170681476593, + "learning_rate": 0.0009632867141835926, + "loss": 2.7298, + "step": 1165 + }, + { + "epoch": 0.149248, + "grad_norm": 0.6257349252700806, + "learning_rate": 0.000963208701766824, + "loss": 2.76, + "step": 1166 + }, + { + "epoch": 0.149376, + "grad_norm": 0.6199800968170166, + "learning_rate": 0.0009631306097191099, + "loss": 2.7311, + "step": 1167 + }, + { + "epoch": 0.149504, + "grad_norm": 0.6486892700195312, + "learning_rate": 0.0009630524380538748, + "loss": 2.6942, + "step": 1168 + }, + { + "epoch": 0.149632, + "grad_norm": 0.6119816899299622, + "learning_rate": 0.0009629741867845573, + "loss": 2.8008, + "step": 1169 + }, + { + "epoch": 0.14976, + "grad_norm": 0.6299313902854919, + "learning_rate": 0.00096289585592461, + "loss": 2.747, + "step": 1170 + }, + { + "epoch": 0.149888, + "grad_norm": 0.6144882440567017, + "learning_rate": 0.0009628174454874987, + "loss": 2.8278, + "step": 1171 + }, + { + "epoch": 0.150016, + "grad_norm": 0.6176167726516724, + "learning_rate": 0.0009627389554867031, + "loss": 2.8126, + "step": 1172 + }, + { + "epoch": 0.150144, + "grad_norm": 0.6355456113815308, + "learning_rate": 0.0009626603859357165, + "loss": 2.8809, + "step": 1173 + }, + { + "epoch": 0.150272, + "grad_norm": 0.6661367416381836, + "learning_rate": 0.0009625817368480458, + "loss": 2.7038, + "step": 1174 + }, + { + "epoch": 0.1504, + "grad_norm": 0.5979088544845581, + "learning_rate": 0.000962503008237212, + "loss": 2.7906, + "step": 1175 + }, + { + "epoch": 0.150528, + "grad_norm": 0.6040823459625244, + "learning_rate": 0.0009624242001167493, + "loss": 2.7942, + "step": 1176 + }, + { + "epoch": 0.150656, + "grad_norm": 0.6401028633117676, + "learning_rate": 0.0009623453125002056, + "loss": 2.7368, + "step": 1177 + }, + { + "epoch": 0.150784, + "grad_norm": 0.6093289256095886, + "learning_rate": 0.0009622663454011427, + "loss": 2.7956, + "step": 1178 + }, + { + "epoch": 0.150912, + "grad_norm": 0.5955445170402527, + "learning_rate": 0.0009621872988331362, + "loss": 2.788, + "step": 1179 + }, + { + "epoch": 0.15104, + "grad_norm": 0.6452881097793579, + "learning_rate": 0.0009621081728097747, + "loss": 2.6621, + "step": 1180 + }, + { + "epoch": 0.151168, + "grad_norm": 0.6016805768013, + "learning_rate": 0.000962028967344661, + "loss": 2.7408, + "step": 1181 + }, + { + "epoch": 0.151296, + "grad_norm": 0.620296835899353, + "learning_rate": 0.0009619496824514117, + "loss": 2.6526, + "step": 1182 + }, + { + "epoch": 0.151424, + "grad_norm": 0.6367105841636658, + "learning_rate": 0.0009618703181436566, + "loss": 2.8862, + "step": 1183 + }, + { + "epoch": 0.151552, + "grad_norm": 0.6269406676292419, + "learning_rate": 0.0009617908744350393, + "loss": 2.6917, + "step": 1184 + }, + { + "epoch": 0.15168, + "grad_norm": 0.6970648169517517, + "learning_rate": 0.000961711351339217, + "loss": 2.8413, + "step": 1185 + }, + { + "epoch": 0.151808, + "grad_norm": 0.6343854069709778, + "learning_rate": 0.0009616317488698609, + "loss": 2.7515, + "step": 1186 + }, + { + "epoch": 0.151936, + "grad_norm": 0.7058969736099243, + "learning_rate": 0.0009615520670406554, + "loss": 2.72, + "step": 1187 + }, + { + "epoch": 0.152064, + "grad_norm": 0.5797877311706543, + "learning_rate": 0.000961472305865299, + "loss": 2.7289, + "step": 1188 + }, + { + "epoch": 0.152192, + "grad_norm": 0.60205078125, + "learning_rate": 0.0009613924653575032, + "loss": 2.7016, + "step": 1189 + }, + { + "epoch": 0.15232, + "grad_norm": 0.5990620851516724, + "learning_rate": 0.0009613125455309939, + "loss": 2.7174, + "step": 1190 + }, + { + "epoch": 0.152448, + "grad_norm": 0.6358687877655029, + "learning_rate": 0.0009612325463995099, + "loss": 2.806, + "step": 1191 + }, + { + "epoch": 0.152576, + "grad_norm": 0.6182118058204651, + "learning_rate": 0.000961152467976804, + "loss": 2.7658, + "step": 1192 + }, + { + "epoch": 0.152704, + "grad_norm": 1.1248902082443237, + "learning_rate": 0.0009610723102766428, + "loss": 2.7065, + "step": 1193 + }, + { + "epoch": 0.152832, + "grad_norm": 0.6031314134597778, + "learning_rate": 0.0009609920733128064, + "loss": 2.7815, + "step": 1194 + }, + { + "epoch": 0.15296, + "grad_norm": 0.6804521679878235, + "learning_rate": 0.000960911757099088, + "loss": 2.7341, + "step": 1195 + }, + { + "epoch": 0.153088, + "grad_norm": 0.6382904052734375, + "learning_rate": 0.0009608313616492955, + "loss": 2.7273, + "step": 1196 + }, + { + "epoch": 0.153216, + "grad_norm": 0.626151442527771, + "learning_rate": 0.0009607508869772495, + "loss": 2.6974, + "step": 1197 + }, + { + "epoch": 0.153344, + "grad_norm": 0.6176477670669556, + "learning_rate": 0.0009606703330967843, + "loss": 2.8033, + "step": 1198 + }, + { + "epoch": 0.153472, + "grad_norm": 0.583111584186554, + "learning_rate": 0.0009605897000217485, + "loss": 2.7218, + "step": 1199 + }, + { + "epoch": 0.1536, + "grad_norm": 0.6251111626625061, + "learning_rate": 0.0009605089877660036, + "loss": 2.8977, + "step": 1200 + }, + { + "epoch": 0.153728, + "grad_norm": 0.5900866389274597, + "learning_rate": 0.000960428196343425, + "loss": 2.8832, + "step": 1201 + }, + { + "epoch": 0.153856, + "grad_norm": 0.590005099773407, + "learning_rate": 0.0009603473257679018, + "loss": 2.7922, + "step": 1202 + }, + { + "epoch": 0.153984, + "grad_norm": 0.6283178329467773, + "learning_rate": 0.0009602663760533363, + "loss": 2.7483, + "step": 1203 + }, + { + "epoch": 0.154112, + "grad_norm": 0.6132609844207764, + "learning_rate": 0.000960185347213645, + "loss": 2.8257, + "step": 1204 + }, + { + "epoch": 0.15424, + "grad_norm": 0.6836451888084412, + "learning_rate": 0.0009601042392627576, + "loss": 2.7227, + "step": 1205 + }, + { + "epoch": 0.154368, + "grad_norm": 0.6225904822349548, + "learning_rate": 0.0009600230522146176, + "loss": 2.6897, + "step": 1206 + }, + { + "epoch": 0.154496, + "grad_norm": 0.6171424388885498, + "learning_rate": 0.0009599417860831817, + "loss": 2.7832, + "step": 1207 + }, + { + "epoch": 0.154624, + "grad_norm": 0.6191909909248352, + "learning_rate": 0.0009598604408824207, + "loss": 2.813, + "step": 1208 + }, + { + "epoch": 0.154752, + "grad_norm": 0.5952314734458923, + "learning_rate": 0.0009597790166263189, + "loss": 2.7753, + "step": 1209 + }, + { + "epoch": 0.15488, + "grad_norm": 0.616737961769104, + "learning_rate": 0.0009596975133288738, + "loss": 2.8577, + "step": 1210 + }, + { + "epoch": 0.155008, + "grad_norm": 0.5456274151802063, + "learning_rate": 0.0009596159310040969, + "loss": 2.6741, + "step": 1211 + }, + { + "epoch": 0.155136, + "grad_norm": 0.6556591391563416, + "learning_rate": 0.0009595342696660132, + "loss": 2.8209, + "step": 1212 + }, + { + "epoch": 0.155264, + "grad_norm": 0.5862638354301453, + "learning_rate": 0.000959452529328661, + "loss": 2.8454, + "step": 1213 + }, + { + "epoch": 0.155392, + "grad_norm": 0.5989412069320679, + "learning_rate": 0.0009593707100060926, + "loss": 2.758, + "step": 1214 + }, + { + "epoch": 0.15552, + "grad_norm": 0.5916311740875244, + "learning_rate": 0.000959288811712374, + "loss": 2.7898, + "step": 1215 + }, + { + "epoch": 0.155648, + "grad_norm": 0.6440004706382751, + "learning_rate": 0.0009592068344615837, + "loss": 2.7254, + "step": 1216 + }, + { + "epoch": 0.155776, + "grad_norm": 0.6754725575447083, + "learning_rate": 0.0009591247782678152, + "loss": 2.691, + "step": 1217 + }, + { + "epoch": 0.155904, + "grad_norm": 0.63521808385849, + "learning_rate": 0.0009590426431451746, + "loss": 2.8096, + "step": 1218 + }, + { + "epoch": 0.156032, + "grad_norm": 0.5946747660636902, + "learning_rate": 0.000958960429107782, + "loss": 2.8005, + "step": 1219 + }, + { + "epoch": 0.15616, + "grad_norm": 0.5706272721290588, + "learning_rate": 0.0009588781361697709, + "loss": 2.6387, + "step": 1220 + }, + { + "epoch": 0.156288, + "grad_norm": 0.6216918230056763, + "learning_rate": 0.0009587957643452885, + "loss": 2.7859, + "step": 1221 + }, + { + "epoch": 0.156416, + "grad_norm": 0.5551254749298096, + "learning_rate": 0.0009587133136484953, + "loss": 2.8187, + "step": 1222 + }, + { + "epoch": 0.156544, + "grad_norm": 0.576370358467102, + "learning_rate": 0.0009586307840935657, + "loss": 2.6907, + "step": 1223 + }, + { + "epoch": 0.156672, + "grad_norm": 0.5974119901657104, + "learning_rate": 0.0009585481756946874, + "loss": 2.8484, + "step": 1224 + }, + { + "epoch": 0.1568, + "grad_norm": 0.5724311470985413, + "learning_rate": 0.0009584654884660618, + "loss": 2.6906, + "step": 1225 + }, + { + "epoch": 0.156928, + "grad_norm": 0.6312941908836365, + "learning_rate": 0.0009583827224219035, + "loss": 2.7417, + "step": 1226 + }, + { + "epoch": 0.157056, + "grad_norm": 0.6085939407348633, + "learning_rate": 0.0009582998775764413, + "loss": 2.8088, + "step": 1227 + }, + { + "epoch": 0.157184, + "grad_norm": 0.6661676168441772, + "learning_rate": 0.0009582169539439171, + "loss": 2.7252, + "step": 1228 + }, + { + "epoch": 0.157312, + "grad_norm": 0.6060323715209961, + "learning_rate": 0.0009581339515385865, + "loss": 2.7819, + "step": 1229 + }, + { + "epoch": 0.15744, + "grad_norm": 0.6148831844329834, + "learning_rate": 0.0009580508703747184, + "loss": 2.7751, + "step": 1230 + }, + { + "epoch": 0.157568, + "grad_norm": 0.729153573513031, + "learning_rate": 0.0009579677104665956, + "loss": 2.7651, + "step": 1231 + }, + { + "epoch": 0.157696, + "grad_norm": 0.5628499388694763, + "learning_rate": 0.000957884471828514, + "loss": 2.7495, + "step": 1232 + }, + { + "epoch": 0.157824, + "grad_norm": 0.7604974508285522, + "learning_rate": 0.0009578011544747835, + "loss": 2.7808, + "step": 1233 + }, + { + "epoch": 0.157952, + "grad_norm": 0.6171746850013733, + "learning_rate": 0.0009577177584197274, + "loss": 2.8084, + "step": 1234 + }, + { + "epoch": 0.15808, + "grad_norm": 0.6707436442375183, + "learning_rate": 0.0009576342836776821, + "loss": 2.8787, + "step": 1235 + }, + { + "epoch": 0.158208, + "grad_norm": 0.6785359978675842, + "learning_rate": 0.0009575507302629982, + "loss": 2.8919, + "step": 1236 + }, + { + "epoch": 0.158336, + "grad_norm": 0.6822549104690552, + "learning_rate": 0.0009574670981900393, + "loss": 2.7134, + "step": 1237 + }, + { + "epoch": 0.158464, + "grad_norm": 0.665725827217102, + "learning_rate": 0.0009573833874731829, + "loss": 2.6667, + "step": 1238 + }, + { + "epoch": 0.158592, + "grad_norm": 0.6216881275177002, + "learning_rate": 0.0009572995981268198, + "loss": 2.8436, + "step": 1239 + }, + { + "epoch": 0.15872, + "grad_norm": 0.6793299913406372, + "learning_rate": 0.0009572157301653542, + "loss": 2.8164, + "step": 1240 + }, + { + "epoch": 0.158848, + "grad_norm": 0.6019121408462524, + "learning_rate": 0.0009571317836032041, + "loss": 2.7171, + "step": 1241 + }, + { + "epoch": 0.158976, + "grad_norm": 0.6011812686920166, + "learning_rate": 0.0009570477584548007, + "loss": 2.699, + "step": 1242 + }, + { + "epoch": 0.159104, + "grad_norm": 0.6519502401351929, + "learning_rate": 0.0009569636547345894, + "loss": 2.6971, + "step": 1243 + }, + { + "epoch": 0.159232, + "grad_norm": 0.7981594204902649, + "learning_rate": 0.0009568794724570281, + "loss": 2.7606, + "step": 1244 + }, + { + "epoch": 0.15936, + "grad_norm": 0.7257503867149353, + "learning_rate": 0.0009567952116365889, + "loss": 2.7178, + "step": 1245 + }, + { + "epoch": 0.159488, + "grad_norm": 0.6252155900001526, + "learning_rate": 0.0009567108722877572, + "loss": 2.8431, + "step": 1246 + }, + { + "epoch": 0.159616, + "grad_norm": 0.6213288903236389, + "learning_rate": 0.0009566264544250318, + "loss": 2.793, + "step": 1247 + }, + { + "epoch": 0.159744, + "grad_norm": 0.5798444747924805, + "learning_rate": 0.0009565419580629253, + "loss": 2.7343, + "step": 1248 + }, + { + "epoch": 0.159872, + "grad_norm": 1.9296504259109497, + "learning_rate": 0.0009564573832159637, + "loss": 2.7724, + "step": 1249 + }, + { + "epoch": 0.16, + "grad_norm": 0.6001925468444824, + "learning_rate": 0.0009563727298986859, + "loss": 2.8066, + "step": 1250 + }, + { + "epoch": 0.160128, + "grad_norm": 1.981367588043213, + "learning_rate": 0.0009562879981256454, + "loss": 2.8143, + "step": 1251 + }, + { + "epoch": 0.160256, + "grad_norm": 0.7230203151702881, + "learning_rate": 0.0009562031879114081, + "loss": 2.8689, + "step": 1252 + }, + { + "epoch": 0.160384, + "grad_norm": 0.9239676594734192, + "learning_rate": 0.0009561182992705541, + "loss": 2.753, + "step": 1253 + }, + { + "epoch": 0.160512, + "grad_norm": 0.681476891040802, + "learning_rate": 0.0009560333322176767, + "loss": 2.6419, + "step": 1254 + }, + { + "epoch": 0.16064, + "grad_norm": 0.7136267423629761, + "learning_rate": 0.0009559482867673825, + "loss": 2.7773, + "step": 1255 + }, + { + "epoch": 0.160768, + "grad_norm": 0.6641641855239868, + "learning_rate": 0.0009558631629342922, + "loss": 2.8487, + "step": 1256 + }, + { + "epoch": 0.160896, + "grad_norm": 0.7043903470039368, + "learning_rate": 0.0009557779607330393, + "loss": 2.901, + "step": 1257 + }, + { + "epoch": 0.161024, + "grad_norm": 0.680356502532959, + "learning_rate": 0.0009556926801782713, + "loss": 2.8636, + "step": 1258 + }, + { + "epoch": 0.161152, + "grad_norm": 1.099178433418274, + "learning_rate": 0.0009556073212846485, + "loss": 2.6953, + "step": 1259 + }, + { + "epoch": 0.16128, + "grad_norm": 0.7003170251846313, + "learning_rate": 0.0009555218840668454, + "loss": 2.7585, + "step": 1260 + }, + { + "epoch": 0.161408, + "grad_norm": 0.6663305759429932, + "learning_rate": 0.0009554363685395495, + "loss": 2.7458, + "step": 1261 + }, + { + "epoch": 0.161536, + "grad_norm": 0.6601340770721436, + "learning_rate": 0.0009553507747174622, + "loss": 2.8318, + "step": 1262 + }, + { + "epoch": 0.161664, + "grad_norm": 0.6770673990249634, + "learning_rate": 0.0009552651026152976, + "loss": 2.7154, + "step": 1263 + }, + { + "epoch": 0.161792, + "grad_norm": 0.9603923559188843, + "learning_rate": 0.0009551793522477841, + "loss": 2.7669, + "step": 1264 + }, + { + "epoch": 0.16192, + "grad_norm": 0.6503064632415771, + "learning_rate": 0.0009550935236296629, + "loss": 2.8893, + "step": 1265 + }, + { + "epoch": 0.162048, + "grad_norm": 0.6399716138839722, + "learning_rate": 0.0009550076167756892, + "loss": 2.8002, + "step": 1266 + }, + { + "epoch": 0.162176, + "grad_norm": 1.0242973566055298, + "learning_rate": 0.0009549216317006313, + "loss": 2.8138, + "step": 1267 + }, + { + "epoch": 0.162304, + "grad_norm": 0.6370578408241272, + "learning_rate": 0.0009548355684192711, + "loss": 2.7651, + "step": 1268 + }, + { + "epoch": 0.162432, + "grad_norm": 1.0837109088897705, + "learning_rate": 0.0009547494269464036, + "loss": 2.9206, + "step": 1269 + }, + { + "epoch": 0.16256, + "grad_norm": 0.6916212439537048, + "learning_rate": 0.0009546632072968378, + "loss": 2.8592, + "step": 1270 + }, + { + "epoch": 0.162688, + "grad_norm": 0.7830777168273926, + "learning_rate": 0.0009545769094853958, + "loss": 2.8092, + "step": 1271 + }, + { + "epoch": 0.162816, + "grad_norm": 0.6350697875022888, + "learning_rate": 0.000954490533526913, + "loss": 2.715, + "step": 1272 + }, + { + "epoch": 0.162944, + "grad_norm": 0.6429615020751953, + "learning_rate": 0.0009544040794362388, + "loss": 2.808, + "step": 1273 + }, + { + "epoch": 0.163072, + "grad_norm": 0.6584208607673645, + "learning_rate": 0.0009543175472282352, + "loss": 2.7914, + "step": 1274 + }, + { + "epoch": 0.1632, + "grad_norm": 0.6702423095703125, + "learning_rate": 0.0009542309369177784, + "loss": 2.772, + "step": 1275 + }, + { + "epoch": 0.163328, + "grad_norm": 0.8203039765357971, + "learning_rate": 0.0009541442485197577, + "loss": 2.742, + "step": 1276 + }, + { + "epoch": 0.163456, + "grad_norm": 0.7193248271942139, + "learning_rate": 0.0009540574820490757, + "loss": 2.6707, + "step": 1277 + }, + { + "epoch": 0.163584, + "grad_norm": 0.6296414136886597, + "learning_rate": 0.0009539706375206487, + "loss": 2.7846, + "step": 1278 + }, + { + "epoch": 0.163712, + "grad_norm": 0.6784297227859497, + "learning_rate": 0.0009538837149494064, + "loss": 2.7578, + "step": 1279 + }, + { + "epoch": 0.16384, + "grad_norm": 2.5007293224334717, + "learning_rate": 0.0009537967143502915, + "loss": 2.6891, + "step": 1280 + }, + { + "epoch": 0.163968, + "grad_norm": 1.2267050743103027, + "learning_rate": 0.0009537096357382604, + "loss": 2.8383, + "step": 1281 + }, + { + "epoch": 0.164096, + "grad_norm": 0.7336238026618958, + "learning_rate": 0.0009536224791282833, + "loss": 2.8333, + "step": 1282 + }, + { + "epoch": 0.164224, + "grad_norm": 0.762978732585907, + "learning_rate": 0.000953535244535343, + "loss": 2.9385, + "step": 1283 + }, + { + "epoch": 0.164352, + "grad_norm": 0.6125741004943848, + "learning_rate": 0.0009534479319744365, + "loss": 2.8607, + "step": 1284 + }, + { + "epoch": 0.16448, + "grad_norm": 0.6245619058609009, + "learning_rate": 0.0009533605414605735, + "loss": 2.854, + "step": 1285 + }, + { + "epoch": 0.164608, + "grad_norm": 0.6322377324104309, + "learning_rate": 0.0009532730730087778, + "loss": 2.7703, + "step": 1286 + }, + { + "epoch": 0.164736, + "grad_norm": 0.5940580368041992, + "learning_rate": 0.0009531855266340859, + "loss": 2.6507, + "step": 1287 + }, + { + "epoch": 0.164864, + "grad_norm": 0.6305710673332214, + "learning_rate": 0.0009530979023515483, + "loss": 2.7488, + "step": 1288 + }, + { + "epoch": 0.164992, + "grad_norm": 0.6721678972244263, + "learning_rate": 0.0009530102001762284, + "loss": 2.9247, + "step": 1289 + }, + { + "epoch": 0.16512, + "grad_norm": 0.6502331495285034, + "learning_rate": 0.0009529224201232034, + "loss": 2.8565, + "step": 1290 + }, + { + "epoch": 0.165248, + "grad_norm": 0.5964451432228088, + "learning_rate": 0.0009528345622075635, + "loss": 2.8258, + "step": 1291 + }, + { + "epoch": 0.165376, + "grad_norm": 0.5776752829551697, + "learning_rate": 0.000952746626444413, + "loss": 2.6559, + "step": 1292 + }, + { + "epoch": 0.165504, + "grad_norm": 0.6667888760566711, + "learning_rate": 0.0009526586128488685, + "loss": 2.848, + "step": 1293 + }, + { + "epoch": 0.165632, + "grad_norm": 0.660143256187439, + "learning_rate": 0.0009525705214360609, + "loss": 2.8078, + "step": 1294 + }, + { + "epoch": 0.16576, + "grad_norm": 0.7522936463356018, + "learning_rate": 0.0009524823522211339, + "loss": 2.7187, + "step": 1295 + }, + { + "epoch": 0.165888, + "grad_norm": 0.9888327717781067, + "learning_rate": 0.000952394105219245, + "loss": 2.8511, + "step": 1296 + }, + { + "epoch": 0.166016, + "grad_norm": 0.6380833387374878, + "learning_rate": 0.0009523057804455648, + "loss": 2.8179, + "step": 1297 + }, + { + "epoch": 0.166144, + "grad_norm": 0.6702451109886169, + "learning_rate": 0.0009522173779152773, + "loss": 2.7516, + "step": 1298 + }, + { + "epoch": 0.166272, + "grad_norm": 0.6157838702201843, + "learning_rate": 0.00095212889764358, + "loss": 2.8981, + "step": 1299 + }, + { + "epoch": 0.1664, + "grad_norm": 0.6978871822357178, + "learning_rate": 0.0009520403396456838, + "loss": 2.6791, + "step": 1300 + }, + { + "epoch": 0.166528, + "grad_norm": 0.643008291721344, + "learning_rate": 0.0009519517039368126, + "loss": 2.8744, + "step": 1301 + }, + { + "epoch": 0.166656, + "grad_norm": 0.6399442553520203, + "learning_rate": 0.000951862990532204, + "loss": 2.7789, + "step": 1302 + }, + { + "epoch": 0.166784, + "grad_norm": 1.1904098987579346, + "learning_rate": 0.000951774199447109, + "loss": 2.7787, + "step": 1303 + }, + { + "epoch": 0.166912, + "grad_norm": 0.7599002122879028, + "learning_rate": 0.0009516853306967915, + "loss": 2.8014, + "step": 1304 + }, + { + "epoch": 0.16704, + "grad_norm": 1.0064979791641235, + "learning_rate": 0.0009515963842965294, + "loss": 2.7574, + "step": 1305 + }, + { + "epoch": 0.167168, + "grad_norm": 0.9409542083740234, + "learning_rate": 0.0009515073602616135, + "loss": 2.696, + "step": 1306 + }, + { + "epoch": 0.167296, + "grad_norm": 0.6813908815383911, + "learning_rate": 0.000951418258607348, + "loss": 2.8158, + "step": 1307 + }, + { + "epoch": 0.167424, + "grad_norm": 0.6212358474731445, + "learning_rate": 0.0009513290793490505, + "loss": 2.7905, + "step": 1308 + }, + { + "epoch": 0.167552, + "grad_norm": 0.7596809267997742, + "learning_rate": 0.0009512398225020522, + "loss": 2.7433, + "step": 1309 + }, + { + "epoch": 0.16768, + "grad_norm": 0.5821667313575745, + "learning_rate": 0.000951150488081697, + "loss": 2.8046, + "step": 1310 + }, + { + "epoch": 0.167808, + "grad_norm": 0.606890082359314, + "learning_rate": 0.0009510610761033427, + "loss": 2.7411, + "step": 1311 + }, + { + "epoch": 0.167936, + "grad_norm": 0.5842027068138123, + "learning_rate": 0.0009509715865823604, + "loss": 2.7722, + "step": 1312 + }, + { + "epoch": 0.168064, + "grad_norm": 0.625555694103241, + "learning_rate": 0.0009508820195341342, + "loss": 2.7176, + "step": 1313 + }, + { + "epoch": 0.168192, + "grad_norm": 0.6106181740760803, + "learning_rate": 0.0009507923749740618, + "loss": 2.6703, + "step": 1314 + }, + { + "epoch": 0.16832, + "grad_norm": 0.5952621698379517, + "learning_rate": 0.000950702652917554, + "loss": 2.7373, + "step": 1315 + }, + { + "epoch": 0.168448, + "grad_norm": 0.6135830283164978, + "learning_rate": 0.0009506128533800354, + "loss": 2.7406, + "step": 1316 + }, + { + "epoch": 0.168576, + "grad_norm": 0.764107882976532, + "learning_rate": 0.0009505229763769431, + "loss": 2.7293, + "step": 1317 + }, + { + "epoch": 0.168704, + "grad_norm": 0.7059986591339111, + "learning_rate": 0.0009504330219237283, + "loss": 2.8393, + "step": 1318 + }, + { + "epoch": 0.168832, + "grad_norm": 0.7599570155143738, + "learning_rate": 0.0009503429900358553, + "loss": 2.8424, + "step": 1319 + }, + { + "epoch": 0.16896, + "grad_norm": 0.6840741038322449, + "learning_rate": 0.0009502528807288014, + "loss": 2.7594, + "step": 1320 + }, + { + "epoch": 0.169088, + "grad_norm": 0.6327698230743408, + "learning_rate": 0.0009501626940180575, + "loss": 2.7982, + "step": 1321 + }, + { + "epoch": 0.169216, + "grad_norm": 0.6820536851882935, + "learning_rate": 0.0009500724299191278, + "loss": 2.7634, + "step": 1322 + }, + { + "epoch": 0.169344, + "grad_norm": 0.6757768392562866, + "learning_rate": 0.0009499820884475296, + "loss": 2.6732, + "step": 1323 + }, + { + "epoch": 0.169472, + "grad_norm": 0.646988570690155, + "learning_rate": 0.000949891669618794, + "loss": 2.7501, + "step": 1324 + }, + { + "epoch": 0.1696, + "grad_norm": 0.610481858253479, + "learning_rate": 0.0009498011734484646, + "loss": 2.7314, + "step": 1325 + }, + { + "epoch": 0.169728, + "grad_norm": 0.6370837688446045, + "learning_rate": 0.0009497105999520989, + "loss": 2.8036, + "step": 1326 + }, + { + "epoch": 0.169856, + "grad_norm": 0.5803393721580505, + "learning_rate": 0.0009496199491452679, + "loss": 2.751, + "step": 1327 + }, + { + "epoch": 0.169984, + "grad_norm": 1.1647971868515015, + "learning_rate": 0.0009495292210435548, + "loss": 2.7449, + "step": 1328 + }, + { + "epoch": 0.170112, + "grad_norm": 0.6656283736228943, + "learning_rate": 0.0009494384156625575, + "loss": 2.801, + "step": 1329 + }, + { + "epoch": 0.17024, + "grad_norm": 0.6291475296020508, + "learning_rate": 0.000949347533017886, + "loss": 2.8354, + "step": 1330 + }, + { + "epoch": 0.170368, + "grad_norm": 0.7014458179473877, + "learning_rate": 0.0009492565731251644, + "loss": 2.9383, + "step": 1331 + }, + { + "epoch": 0.170496, + "grad_norm": 0.6213318705558777, + "learning_rate": 0.0009491655360000297, + "loss": 2.7622, + "step": 1332 + }, + { + "epoch": 0.170624, + "grad_norm": 0.7062961459159851, + "learning_rate": 0.0009490744216581322, + "loss": 2.7611, + "step": 1333 + }, + { + "epoch": 0.170752, + "grad_norm": 0.6952922940254211, + "learning_rate": 0.0009489832301151353, + "loss": 2.7724, + "step": 1334 + }, + { + "epoch": 0.17088, + "grad_norm": 0.6644594669342041, + "learning_rate": 0.0009488919613867162, + "loss": 2.7065, + "step": 1335 + }, + { + "epoch": 0.171008, + "grad_norm": 0.6368440389633179, + "learning_rate": 0.000948800615488565, + "loss": 2.7706, + "step": 1336 + }, + { + "epoch": 0.171136, + "grad_norm": 0.6296830177307129, + "learning_rate": 0.0009487091924363851, + "loss": 2.8363, + "step": 1337 + }, + { + "epoch": 0.171264, + "grad_norm": 0.6764806509017944, + "learning_rate": 0.0009486176922458929, + "loss": 2.7548, + "step": 1338 + }, + { + "epoch": 0.171392, + "grad_norm": 0.6404110193252563, + "learning_rate": 0.0009485261149328188, + "loss": 2.6842, + "step": 1339 + }, + { + "epoch": 0.17152, + "grad_norm": 0.6751078367233276, + "learning_rate": 0.0009484344605129056, + "loss": 2.8062, + "step": 1340 + }, + { + "epoch": 0.171648, + "grad_norm": 0.6456961631774902, + "learning_rate": 0.0009483427290019101, + "loss": 2.8425, + "step": 1341 + }, + { + "epoch": 0.171776, + "grad_norm": 0.7031457424163818, + "learning_rate": 0.0009482509204156018, + "loss": 2.7528, + "step": 1342 + }, + { + "epoch": 0.171904, + "grad_norm": 0.5889406204223633, + "learning_rate": 0.0009481590347697637, + "loss": 2.7855, + "step": 1343 + }, + { + "epoch": 0.172032, + "grad_norm": 0.6444111466407776, + "learning_rate": 0.0009480670720801921, + "loss": 2.698, + "step": 1344 + }, + { + "epoch": 0.17216, + "grad_norm": 0.6742934584617615, + "learning_rate": 0.0009479750323626962, + "loss": 2.8078, + "step": 1345 + }, + { + "epoch": 0.172288, + "grad_norm": 0.6685083508491516, + "learning_rate": 0.000947882915633099, + "loss": 2.8213, + "step": 1346 + }, + { + "epoch": 0.172416, + "grad_norm": 0.6732606887817383, + "learning_rate": 0.0009477907219072361, + "loss": 2.9879, + "step": 1347 + }, + { + "epoch": 0.172544, + "grad_norm": 0.6535918712615967, + "learning_rate": 0.0009476984512009571, + "loss": 2.736, + "step": 1348 + }, + { + "epoch": 0.172672, + "grad_norm": 0.628352701663971, + "learning_rate": 0.0009476061035301241, + "loss": 2.7522, + "step": 1349 + }, + { + "epoch": 0.1728, + "grad_norm": 0.6275818943977356, + "learning_rate": 0.0009475136789106128, + "loss": 2.8036, + "step": 1350 + }, + { + "epoch": 0.172928, + "grad_norm": 0.6174351572990417, + "learning_rate": 0.0009474211773583121, + "loss": 2.6708, + "step": 1351 + }, + { + "epoch": 0.173056, + "grad_norm": 0.6139631867408752, + "learning_rate": 0.000947328598889124, + "loss": 2.891, + "step": 1352 + }, + { + "epoch": 0.173184, + "grad_norm": 0.7011117935180664, + "learning_rate": 0.000947235943518964, + "loss": 2.7866, + "step": 1353 + }, + { + "epoch": 0.173312, + "grad_norm": 0.631915271282196, + "learning_rate": 0.0009471432112637604, + "loss": 2.6762, + "step": 1354 + }, + { + "epoch": 0.17344, + "grad_norm": 0.6609591245651245, + "learning_rate": 0.000947050402139455, + "loss": 2.8273, + "step": 1355 + }, + { + "epoch": 0.173568, + "grad_norm": 0.6784368753433228, + "learning_rate": 0.0009469575161620029, + "loss": 2.7808, + "step": 1356 + }, + { + "epoch": 0.173696, + "grad_norm": 0.6232576966285706, + "learning_rate": 0.0009468645533473721, + "loss": 2.7978, + "step": 1357 + }, + { + "epoch": 0.173824, + "grad_norm": 0.6302456259727478, + "learning_rate": 0.0009467715137115442, + "loss": 2.7678, + "step": 1358 + }, + { + "epoch": 0.173952, + "grad_norm": 0.7115084528923035, + "learning_rate": 0.0009466783972705136, + "loss": 2.7571, + "step": 1359 + }, + { + "epoch": 0.17408, + "grad_norm": 0.6183989644050598, + "learning_rate": 0.0009465852040402883, + "loss": 2.8136, + "step": 1360 + }, + { + "epoch": 0.174208, + "grad_norm": 0.6962798833847046, + "learning_rate": 0.000946491934036889, + "loss": 2.7847, + "step": 1361 + }, + { + "epoch": 0.174336, + "grad_norm": 0.8298761248588562, + "learning_rate": 0.00094639858727635, + "loss": 2.807, + "step": 1362 + }, + { + "epoch": 0.174464, + "grad_norm": 0.6634453535079956, + "learning_rate": 0.0009463051637747191, + "loss": 2.7381, + "step": 1363 + }, + { + "epoch": 0.174592, + "grad_norm": 0.7687205672264099, + "learning_rate": 0.0009462116635480562, + "loss": 2.6183, + "step": 1364 + }, + { + "epoch": 0.17472, + "grad_norm": 0.5850154757499695, + "learning_rate": 0.0009461180866124356, + "loss": 2.7144, + "step": 1365 + }, + { + "epoch": 0.174848, + "grad_norm": 1.8635470867156982, + "learning_rate": 0.000946024432983944, + "loss": 2.9338, + "step": 1366 + }, + { + "epoch": 0.174976, + "grad_norm": 0.5905047655105591, + "learning_rate": 0.0009459307026786816, + "loss": 2.7844, + "step": 1367 + }, + { + "epoch": 0.175104, + "grad_norm": 0.8280892372131348, + "learning_rate": 0.0009458368957127617, + "loss": 2.8692, + "step": 1368 + }, + { + "epoch": 0.175232, + "grad_norm": 0.7254469990730286, + "learning_rate": 0.000945743012102311, + "loss": 2.7803, + "step": 1369 + }, + { + "epoch": 0.17536, + "grad_norm": 0.6624941825866699, + "learning_rate": 0.0009456490518634688, + "loss": 2.6803, + "step": 1370 + }, + { + "epoch": 0.175488, + "grad_norm": 0.6648242473602295, + "learning_rate": 0.0009455550150123883, + "loss": 2.7469, + "step": 1371 + }, + { + "epoch": 0.175616, + "grad_norm": 0.7350017428398132, + "learning_rate": 0.0009454609015652353, + "loss": 2.8066, + "step": 1372 + }, + { + "epoch": 0.175744, + "grad_norm": 0.7223126888275146, + "learning_rate": 0.0009453667115381892, + "loss": 2.8088, + "step": 1373 + }, + { + "epoch": 0.175872, + "grad_norm": 0.724946916103363, + "learning_rate": 0.0009452724449474422, + "loss": 2.7509, + "step": 1374 + }, + { + "epoch": 0.176, + "grad_norm": 0.6045209169387817, + "learning_rate": 0.0009451781018091999, + "loss": 2.8458, + "step": 1375 + }, + { + "epoch": 0.176128, + "grad_norm": 0.6119378209114075, + "learning_rate": 0.000945083682139681, + "loss": 2.7519, + "step": 1376 + }, + { + "epoch": 0.176256, + "grad_norm": 0.6600387096405029, + "learning_rate": 0.0009449891859551172, + "loss": 2.8469, + "step": 1377 + }, + { + "epoch": 0.176384, + "grad_norm": 0.5878028869628906, + "learning_rate": 0.0009448946132717536, + "loss": 2.7051, + "step": 1378 + }, + { + "epoch": 0.176512, + "grad_norm": 0.9388783574104309, + "learning_rate": 0.0009447999641058485, + "loss": 2.7926, + "step": 1379 + }, + { + "epoch": 0.17664, + "grad_norm": 0.7119961380958557, + "learning_rate": 0.0009447052384736728, + "loss": 2.7233, + "step": 1380 + }, + { + "epoch": 0.176768, + "grad_norm": 0.6239910125732422, + "learning_rate": 0.0009446104363915112, + "loss": 2.7465, + "step": 1381 + }, + { + "epoch": 0.176896, + "grad_norm": 5.004570484161377, + "learning_rate": 0.0009445155578756611, + "loss": 2.8036, + "step": 1382 + }, + { + "epoch": 0.177024, + "grad_norm": 0.7843655347824097, + "learning_rate": 0.0009444206029424334, + "loss": 2.7545, + "step": 1383 + }, + { + "epoch": 0.177152, + "grad_norm": 0.7828148007392883, + "learning_rate": 0.0009443255716081521, + "loss": 2.7922, + "step": 1384 + }, + { + "epoch": 0.17728, + "grad_norm": 0.7413930296897888, + "learning_rate": 0.0009442304638891537, + "loss": 2.8447, + "step": 1385 + }, + { + "epoch": 0.177408, + "grad_norm": 0.7736758589744568, + "learning_rate": 0.0009441352798017886, + "loss": 2.8227, + "step": 1386 + }, + { + "epoch": 0.177536, + "grad_norm": 0.7102529406547546, + "learning_rate": 0.0009440400193624201, + "loss": 2.8405, + "step": 1387 + }, + { + "epoch": 0.177664, + "grad_norm": 0.624217689037323, + "learning_rate": 0.0009439446825874247, + "loss": 2.8245, + "step": 1388 + }, + { + "epoch": 0.177792, + "grad_norm": 0.6483398675918579, + "learning_rate": 0.0009438492694931915, + "loss": 2.855, + "step": 1389 + }, + { + "epoch": 0.17792, + "grad_norm": 0.7867991924285889, + "learning_rate": 0.0009437537800961234, + "loss": 2.8044, + "step": 1390 + }, + { + "epoch": 0.178048, + "grad_norm": 0.7120444774627686, + "learning_rate": 0.0009436582144126361, + "loss": 2.8488, + "step": 1391 + }, + { + "epoch": 0.178176, + "grad_norm": 0.7026340961456299, + "learning_rate": 0.0009435625724591583, + "loss": 2.8034, + "step": 1392 + }, + { + "epoch": 0.178304, + "grad_norm": 0.7869840860366821, + "learning_rate": 0.0009434668542521323, + "loss": 2.859, + "step": 1393 + }, + { + "epoch": 0.178432, + "grad_norm": 0.6876501441001892, + "learning_rate": 0.0009433710598080127, + "loss": 2.8002, + "step": 1394 + }, + { + "epoch": 0.17856, + "grad_norm": 1.3003562688827515, + "learning_rate": 0.000943275189143268, + "loss": 2.7837, + "step": 1395 + }, + { + "epoch": 0.178688, + "grad_norm": 0.7743299603462219, + "learning_rate": 0.0009431792422743794, + "loss": 2.7246, + "step": 1396 + }, + { + "epoch": 0.178816, + "grad_norm": 0.70578533411026, + "learning_rate": 0.0009430832192178413, + "loss": 2.7727, + "step": 1397 + }, + { + "epoch": 0.178944, + "grad_norm": 0.6332562565803528, + "learning_rate": 0.0009429871199901613, + "loss": 2.8604, + "step": 1398 + }, + { + "epoch": 0.179072, + "grad_norm": 0.6262220144271851, + "learning_rate": 0.0009428909446078598, + "loss": 2.9018, + "step": 1399 + }, + { + "epoch": 0.1792, + "grad_norm": 0.6338145732879639, + "learning_rate": 0.0009427946930874703, + "loss": 2.7546, + "step": 1400 + }, + { + "epoch": 0.179328, + "grad_norm": 0.6016040444374084, + "learning_rate": 0.0009426983654455398, + "loss": 2.8303, + "step": 1401 + }, + { + "epoch": 0.179456, + "grad_norm": 0.6110451221466064, + "learning_rate": 0.0009426019616986281, + "loss": 2.8508, + "step": 1402 + }, + { + "epoch": 0.179584, + "grad_norm": 0.6073960065841675, + "learning_rate": 0.0009425054818633081, + "loss": 2.8901, + "step": 1403 + }, + { + "epoch": 0.179712, + "grad_norm": 0.6042421460151672, + "learning_rate": 0.0009424089259561657, + "loss": 2.7677, + "step": 1404 + }, + { + "epoch": 0.17984, + "grad_norm": 0.621364176273346, + "learning_rate": 0.0009423122939938003, + "loss": 2.7622, + "step": 1405 + }, + { + "epoch": 0.179968, + "grad_norm": 0.6646581888198853, + "learning_rate": 0.0009422155859928236, + "loss": 2.8909, + "step": 1406 + }, + { + "epoch": 0.180096, + "grad_norm": 0.6159442663192749, + "learning_rate": 0.0009421188019698612, + "loss": 2.8212, + "step": 1407 + }, + { + "epoch": 0.180224, + "grad_norm": 0.636151134967804, + "learning_rate": 0.0009420219419415514, + "loss": 2.8667, + "step": 1408 + }, + { + "epoch": 0.180352, + "grad_norm": 0.6391863226890564, + "learning_rate": 0.0009419250059245452, + "loss": 2.6823, + "step": 1409 + }, + { + "epoch": 0.18048, + "grad_norm": 0.7106063365936279, + "learning_rate": 0.0009418279939355074, + "loss": 2.6509, + "step": 1410 + }, + { + "epoch": 0.180608, + "grad_norm": 0.6507261395454407, + "learning_rate": 0.0009417309059911152, + "loss": 2.8751, + "step": 1411 + }, + { + "epoch": 0.180736, + "grad_norm": 0.6292952299118042, + "learning_rate": 0.0009416337421080593, + "loss": 2.7449, + "step": 1412 + }, + { + "epoch": 0.180864, + "grad_norm": 0.5959364771842957, + "learning_rate": 0.0009415365023030433, + "loss": 2.7854, + "step": 1413 + }, + { + "epoch": 0.180992, + "grad_norm": 0.5956499576568604, + "learning_rate": 0.0009414391865927838, + "loss": 2.6975, + "step": 1414 + }, + { + "epoch": 0.18112, + "grad_norm": 0.6616135835647583, + "learning_rate": 0.0009413417949940107, + "loss": 2.7243, + "step": 1415 + }, + { + "epoch": 0.181248, + "grad_norm": 0.6522478461265564, + "learning_rate": 0.0009412443275234663, + "loss": 2.7649, + "step": 1416 + }, + { + "epoch": 0.181376, + "grad_norm": 0.6037887334823608, + "learning_rate": 0.0009411467841979068, + "loss": 2.6776, + "step": 1417 + }, + { + "epoch": 0.181504, + "grad_norm": 0.6460844874382019, + "learning_rate": 0.0009410491650341009, + "loss": 2.7919, + "step": 1418 + }, + { + "epoch": 0.181632, + "grad_norm": 0.5915414094924927, + "learning_rate": 0.0009409514700488304, + "loss": 2.8455, + "step": 1419 + }, + { + "epoch": 0.18176, + "grad_norm": 0.6247175931930542, + "learning_rate": 0.0009408536992588902, + "loss": 2.8183, + "step": 1420 + }, + { + "epoch": 0.181888, + "grad_norm": 0.6534907817840576, + "learning_rate": 0.0009407558526810883, + "loss": 2.7517, + "step": 1421 + }, + { + "epoch": 0.182016, + "grad_norm": 0.5965962409973145, + "learning_rate": 0.0009406579303322458, + "loss": 2.7719, + "step": 1422 + }, + { + "epoch": 0.182144, + "grad_norm": 0.7130102515220642, + "learning_rate": 0.0009405599322291962, + "loss": 2.7813, + "step": 1423 + }, + { + "epoch": 0.182272, + "grad_norm": 0.6599665880203247, + "learning_rate": 0.0009404618583887872, + "loss": 2.8716, + "step": 1424 + }, + { + "epoch": 0.1824, + "grad_norm": 0.577401340007782, + "learning_rate": 0.0009403637088278783, + "loss": 2.7756, + "step": 1425 + }, + { + "epoch": 0.182528, + "grad_norm": 0.7568121552467346, + "learning_rate": 0.0009402654835633427, + "loss": 2.8136, + "step": 1426 + }, + { + "epoch": 0.182656, + "grad_norm": 0.6775643825531006, + "learning_rate": 0.0009401671826120665, + "loss": 2.8308, + "step": 1427 + }, + { + "epoch": 0.182784, + "grad_norm": 0.6787882447242737, + "learning_rate": 0.0009400688059909487, + "loss": 2.909, + "step": 1428 + }, + { + "epoch": 0.182912, + "grad_norm": 0.6377055644989014, + "learning_rate": 0.0009399703537169016, + "loss": 2.7228, + "step": 1429 + }, + { + "epoch": 0.18304, + "grad_norm": 0.6159709095954895, + "learning_rate": 0.0009398718258068502, + "loss": 2.721, + "step": 1430 + }, + { + "epoch": 0.183168, + "grad_norm": 0.5609552264213562, + "learning_rate": 0.0009397732222777322, + "loss": 2.7963, + "step": 1431 + }, + { + "epoch": 0.183296, + "grad_norm": 0.5538759231567383, + "learning_rate": 0.0009396745431464993, + "loss": 2.8008, + "step": 1432 + }, + { + "epoch": 0.183424, + "grad_norm": 0.6544602513313293, + "learning_rate": 0.000939575788430115, + "loss": 2.701, + "step": 1433 + }, + { + "epoch": 0.183552, + "grad_norm": 0.6217986941337585, + "learning_rate": 0.0009394769581455569, + "loss": 2.7019, + "step": 1434 + }, + { + "epoch": 0.18368, + "grad_norm": 0.5695937275886536, + "learning_rate": 0.0009393780523098146, + "loss": 2.7395, + "step": 1435 + }, + { + "epoch": 0.183808, + "grad_norm": 0.5902389287948608, + "learning_rate": 0.0009392790709398917, + "loss": 2.693, + "step": 1436 + }, + { + "epoch": 0.183936, + "grad_norm": 0.5968219637870789, + "learning_rate": 0.0009391800140528037, + "loss": 2.765, + "step": 1437 + }, + { + "epoch": 0.184064, + "grad_norm": 0.6506455540657043, + "learning_rate": 0.0009390808816655801, + "loss": 2.7197, + "step": 1438 + }, + { + "epoch": 0.184192, + "grad_norm": 0.6647669076919556, + "learning_rate": 0.0009389816737952623, + "loss": 2.8478, + "step": 1439 + }, + { + "epoch": 0.18432, + "grad_norm": 0.6308488845825195, + "learning_rate": 0.0009388823904589061, + "loss": 2.6609, + "step": 1440 + }, + { + "epoch": 0.184448, + "grad_norm": 0.6091712713241577, + "learning_rate": 0.0009387830316735788, + "loss": 2.7415, + "step": 1441 + }, + { + "epoch": 0.184576, + "grad_norm": 0.602868914604187, + "learning_rate": 0.0009386835974563615, + "loss": 2.7626, + "step": 1442 + }, + { + "epoch": 0.184704, + "grad_norm": 0.6455078125, + "learning_rate": 0.0009385840878243483, + "loss": 2.7107, + "step": 1443 + }, + { + "epoch": 0.184832, + "grad_norm": 0.6361019015312195, + "learning_rate": 0.0009384845027946458, + "loss": 2.7574, + "step": 1444 + }, + { + "epoch": 0.18496, + "grad_norm": 0.5770986080169678, + "learning_rate": 0.000938384842384374, + "loss": 2.6159, + "step": 1445 + }, + { + "epoch": 0.185088, + "grad_norm": 0.6655343770980835, + "learning_rate": 0.0009382851066106655, + "loss": 2.865, + "step": 1446 + }, + { + "epoch": 0.185216, + "grad_norm": 0.6070794463157654, + "learning_rate": 0.0009381852954906661, + "loss": 2.6948, + "step": 1447 + }, + { + "epoch": 0.185344, + "grad_norm": 0.6329118609428406, + "learning_rate": 0.0009380854090415347, + "loss": 2.6673, + "step": 1448 + }, + { + "epoch": 0.185472, + "grad_norm": 0.5782359838485718, + "learning_rate": 0.0009379854472804425, + "loss": 2.6902, + "step": 1449 + }, + { + "epoch": 0.1856, + "grad_norm": 0.6422058939933777, + "learning_rate": 0.0009378854102245746, + "loss": 2.75, + "step": 1450 + }, + { + "epoch": 0.185728, + "grad_norm": 0.6068655252456665, + "learning_rate": 0.0009377852978911281, + "loss": 2.814, + "step": 1451 + }, + { + "epoch": 0.185856, + "grad_norm": 0.5889601111412048, + "learning_rate": 0.0009376851102973139, + "loss": 2.7142, + "step": 1452 + }, + { + "epoch": 0.185984, + "grad_norm": 0.6073530912399292, + "learning_rate": 0.0009375848474603549, + "loss": 2.8483, + "step": 1453 + }, + { + "epoch": 0.186112, + "grad_norm": 0.7202199101448059, + "learning_rate": 0.0009374845093974878, + "loss": 2.916, + "step": 1454 + }, + { + "epoch": 0.18624, + "grad_norm": 0.6304972767829895, + "learning_rate": 0.0009373840961259619, + "loss": 2.6923, + "step": 1455 + }, + { + "epoch": 0.186368, + "grad_norm": 0.6607401371002197, + "learning_rate": 0.0009372836076630392, + "loss": 2.778, + "step": 1456 + }, + { + "epoch": 0.186496, + "grad_norm": 0.5729928612709045, + "learning_rate": 0.0009371830440259947, + "loss": 2.7787, + "step": 1457 + }, + { + "epoch": 0.186624, + "grad_norm": 0.5789631605148315, + "learning_rate": 0.0009370824052321168, + "loss": 2.6999, + "step": 1458 + }, + { + "epoch": 0.186752, + "grad_norm": 0.6823089718818665, + "learning_rate": 0.0009369816912987065, + "loss": 2.7818, + "step": 1459 + }, + { + "epoch": 0.18688, + "grad_norm": 0.6037008762359619, + "learning_rate": 0.0009368809022430773, + "loss": 2.7643, + "step": 1460 + }, + { + "epoch": 0.187008, + "grad_norm": 1.2001616954803467, + "learning_rate": 0.0009367800380825564, + "loss": 2.7358, + "step": 1461 + }, + { + "epoch": 0.187136, + "grad_norm": 0.6732541918754578, + "learning_rate": 0.0009366790988344834, + "loss": 2.7285, + "step": 1462 + }, + { + "epoch": 0.187264, + "grad_norm": 0.6780279874801636, + "learning_rate": 0.0009365780845162108, + "loss": 2.7931, + "step": 1463 + }, + { + "epoch": 0.187392, + "grad_norm": 0.6173373460769653, + "learning_rate": 0.0009364769951451044, + "loss": 2.731, + "step": 1464 + }, + { + "epoch": 0.18752, + "grad_norm": 0.6026294231414795, + "learning_rate": 0.0009363758307385422, + "loss": 2.7196, + "step": 1465 + }, + { + "epoch": 0.187648, + "grad_norm": 0.5897713899612427, + "learning_rate": 0.0009362745913139159, + "loss": 2.8895, + "step": 1466 + }, + { + "epoch": 0.187776, + "grad_norm": 0.8264421224594116, + "learning_rate": 0.0009361732768886298, + "loss": 2.8378, + "step": 1467 + }, + { + "epoch": 0.187904, + "grad_norm": 0.6034918427467346, + "learning_rate": 0.0009360718874801006, + "loss": 2.6364, + "step": 1468 + }, + { + "epoch": 0.188032, + "grad_norm": 0.5612911581993103, + "learning_rate": 0.0009359704231057586, + "loss": 2.7084, + "step": 1469 + }, + { + "epoch": 0.18816, + "grad_norm": 1.121798038482666, + "learning_rate": 0.0009358688837830468, + "loss": 2.8691, + "step": 1470 + }, + { + "epoch": 0.188288, + "grad_norm": 0.768196165561676, + "learning_rate": 0.0009357672695294208, + "loss": 2.7893, + "step": 1471 + }, + { + "epoch": 0.188416, + "grad_norm": 0.6220189332962036, + "learning_rate": 0.0009356655803623492, + "loss": 2.7707, + "step": 1472 + }, + { + "epoch": 0.188544, + "grad_norm": 0.662118136882782, + "learning_rate": 0.0009355638162993139, + "loss": 2.7824, + "step": 1473 + }, + { + "epoch": 0.188672, + "grad_norm": 0.6290492415428162, + "learning_rate": 0.0009354619773578088, + "loss": 2.7904, + "step": 1474 + }, + { + "epoch": 0.1888, + "grad_norm": 0.6271003484725952, + "learning_rate": 0.0009353600635553416, + "loss": 2.8123, + "step": 1475 + }, + { + "epoch": 0.188928, + "grad_norm": 0.7680790424346924, + "learning_rate": 0.0009352580749094324, + "loss": 2.7329, + "step": 1476 + }, + { + "epoch": 0.189056, + "grad_norm": 0.745980441570282, + "learning_rate": 0.0009351560114376142, + "loss": 2.7437, + "step": 1477 + }, + { + "epoch": 0.189184, + "grad_norm": 1.3958970308303833, + "learning_rate": 0.0009350538731574329, + "loss": 2.7875, + "step": 1478 + }, + { + "epoch": 0.189312, + "grad_norm": 0.6578157544136047, + "learning_rate": 0.0009349516600864471, + "loss": 2.8874, + "step": 1479 + }, + { + "epoch": 0.18944, + "grad_norm": 0.6597474813461304, + "learning_rate": 0.0009348493722422287, + "loss": 2.8496, + "step": 1480 + }, + { + "epoch": 0.189568, + "grad_norm": 0.673967182636261, + "learning_rate": 0.0009347470096423618, + "loss": 2.8557, + "step": 1481 + }, + { + "epoch": 0.189696, + "grad_norm": 0.5717607140541077, + "learning_rate": 0.0009346445723044441, + "loss": 2.7742, + "step": 1482 + }, + { + "epoch": 0.189824, + "grad_norm": 0.5886583924293518, + "learning_rate": 0.0009345420602460855, + "loss": 2.6799, + "step": 1483 + }, + { + "epoch": 0.189952, + "grad_norm": 0.6268536448478699, + "learning_rate": 0.0009344394734849091, + "loss": 2.751, + "step": 1484 + }, + { + "epoch": 0.19008, + "grad_norm": 0.6699987053871155, + "learning_rate": 0.0009343368120385507, + "loss": 2.8007, + "step": 1485 + }, + { + "epoch": 0.190208, + "grad_norm": 0.7414336800575256, + "learning_rate": 0.0009342340759246591, + "loss": 2.7743, + "step": 1486 + }, + { + "epoch": 0.190336, + "grad_norm": 0.5739485025405884, + "learning_rate": 0.0009341312651608958, + "loss": 2.7604, + "step": 1487 + }, + { + "epoch": 0.190464, + "grad_norm": 0.5848697423934937, + "learning_rate": 0.0009340283797649352, + "loss": 2.7614, + "step": 1488 + }, + { + "epoch": 0.190592, + "grad_norm": 0.5967357158660889, + "learning_rate": 0.0009339254197544642, + "loss": 2.8327, + "step": 1489 + }, + { + "epoch": 0.19072, + "grad_norm": 0.5756387710571289, + "learning_rate": 0.0009338223851471832, + "loss": 2.7316, + "step": 1490 + }, + { + "epoch": 0.190848, + "grad_norm": 0.5862936973571777, + "learning_rate": 0.0009337192759608049, + "loss": 2.6845, + "step": 1491 + }, + { + "epoch": 0.190976, + "grad_norm": 0.6189854741096497, + "learning_rate": 0.000933616092213055, + "loss": 2.7767, + "step": 1492 + }, + { + "epoch": 0.191104, + "grad_norm": 0.7291973829269409, + "learning_rate": 0.0009335128339216719, + "loss": 2.8247, + "step": 1493 + }, + { + "epoch": 0.191232, + "grad_norm": 0.6859250068664551, + "learning_rate": 0.0009334095011044069, + "loss": 2.7146, + "step": 1494 + }, + { + "epoch": 0.19136, + "grad_norm": 0.6348487138748169, + "learning_rate": 0.0009333060937790242, + "loss": 2.7228, + "step": 1495 + }, + { + "epoch": 0.191488, + "grad_norm": 0.6064632534980774, + "learning_rate": 0.0009332026119633006, + "loss": 2.7609, + "step": 1496 + }, + { + "epoch": 0.191616, + "grad_norm": 0.6052952408790588, + "learning_rate": 0.000933099055675026, + "loss": 2.7431, + "step": 1497 + }, + { + "epoch": 0.191744, + "grad_norm": 0.5797833204269409, + "learning_rate": 0.0009329954249320027, + "loss": 2.693, + "step": 1498 + }, + { + "epoch": 0.191872, + "grad_norm": 0.6107142567634583, + "learning_rate": 0.0009328917197520461, + "loss": 2.6514, + "step": 1499 + }, + { + "epoch": 0.192, + "grad_norm": 0.5816287398338318, + "learning_rate": 0.0009327879401529846, + "loss": 2.783, + "step": 1500 + }, + { + "epoch": 0.192128, + "grad_norm": 0.568423330783844, + "learning_rate": 0.0009326840861526586, + "loss": 2.6456, + "step": 1501 + }, + { + "epoch": 0.192256, + "grad_norm": 0.5701746344566345, + "learning_rate": 0.0009325801577689223, + "loss": 2.7118, + "step": 1502 + }, + { + "epoch": 0.192384, + "grad_norm": 0.6388358473777771, + "learning_rate": 0.0009324761550196419, + "loss": 2.7162, + "step": 1503 + }, + { + "epoch": 0.192512, + "grad_norm": 0.6285187602043152, + "learning_rate": 0.0009323720779226967, + "loss": 2.852, + "step": 1504 + }, + { + "epoch": 0.19264, + "grad_norm": 0.6314083933830261, + "learning_rate": 0.0009322679264959789, + "loss": 2.6788, + "step": 1505 + }, + { + "epoch": 0.192768, + "grad_norm": 0.6231140494346619, + "learning_rate": 0.000932163700757393, + "loss": 2.7398, + "step": 1506 + }, + { + "epoch": 0.192896, + "grad_norm": 0.5752736926078796, + "learning_rate": 0.0009320594007248573, + "loss": 2.8896, + "step": 1507 + }, + { + "epoch": 0.193024, + "grad_norm": 0.6121694445610046, + "learning_rate": 0.0009319550264163015, + "loss": 2.7484, + "step": 1508 + }, + { + "epoch": 0.193152, + "grad_norm": 0.6337934136390686, + "learning_rate": 0.0009318505778496691, + "loss": 2.7727, + "step": 1509 + }, + { + "epoch": 0.19328, + "grad_norm": 0.6899873614311218, + "learning_rate": 0.0009317460550429159, + "loss": 2.7602, + "step": 1510 + }, + { + "epoch": 0.193408, + "grad_norm": 0.6460145115852356, + "learning_rate": 0.0009316414580140105, + "loss": 2.7371, + "step": 1511 + }, + { + "epoch": 0.193536, + "grad_norm": 0.7547436356544495, + "learning_rate": 0.0009315367867809345, + "loss": 2.793, + "step": 1512 + }, + { + "epoch": 0.193664, + "grad_norm": 0.5792000889778137, + "learning_rate": 0.0009314320413616821, + "loss": 2.7309, + "step": 1513 + }, + { + "epoch": 0.193792, + "grad_norm": 0.630005419254303, + "learning_rate": 0.0009313272217742603, + "loss": 2.7602, + "step": 1514 + }, + { + "epoch": 0.19392, + "grad_norm": 0.5753366947174072, + "learning_rate": 0.0009312223280366885, + "loss": 2.696, + "step": 1515 + }, + { + "epoch": 0.194048, + "grad_norm": 0.5648632049560547, + "learning_rate": 0.0009311173601669996, + "loss": 2.7308, + "step": 1516 + }, + { + "epoch": 0.194176, + "grad_norm": 1.680699348449707, + "learning_rate": 0.0009310123181832382, + "loss": 2.7363, + "step": 1517 + }, + { + "epoch": 0.194304, + "grad_norm": 0.537129282951355, + "learning_rate": 0.0009309072021034628, + "loss": 2.7931, + "step": 1518 + }, + { + "epoch": 0.194432, + "grad_norm": 0.5468804836273193, + "learning_rate": 0.0009308020119457439, + "loss": 2.7673, + "step": 1519 + }, + { + "epoch": 0.19456, + "grad_norm": 0.5661011338233948, + "learning_rate": 0.0009306967477281647, + "loss": 2.6418, + "step": 1520 + }, + { + "epoch": 0.194688, + "grad_norm": 0.6523237824440002, + "learning_rate": 0.0009305914094688216, + "loss": 2.7435, + "step": 1521 + }, + { + "epoch": 0.194816, + "grad_norm": 0.5781131386756897, + "learning_rate": 0.0009304859971858233, + "loss": 2.7699, + "step": 1522 + }, + { + "epoch": 0.194944, + "grad_norm": 0.6037434339523315, + "learning_rate": 0.0009303805108972916, + "loss": 2.7373, + "step": 1523 + }, + { + "epoch": 0.195072, + "grad_norm": 0.9032896757125854, + "learning_rate": 0.0009302749506213606, + "loss": 2.8776, + "step": 1524 + }, + { + "epoch": 0.1952, + "grad_norm": 0.6844773292541504, + "learning_rate": 0.0009301693163761773, + "loss": 2.7557, + "step": 1525 + }, + { + "epoch": 0.195328, + "grad_norm": 0.6290201544761658, + "learning_rate": 0.0009300636081799017, + "loss": 2.8424, + "step": 1526 + }, + { + "epoch": 0.195456, + "grad_norm": 0.6723546981811523, + "learning_rate": 0.0009299578260507061, + "loss": 2.8418, + "step": 1527 + }, + { + "epoch": 0.195584, + "grad_norm": 0.7425581812858582, + "learning_rate": 0.0009298519700067757, + "loss": 2.7787, + "step": 1528 + }, + { + "epoch": 0.195712, + "grad_norm": 0.6425061821937561, + "learning_rate": 0.0009297460400663084, + "loss": 2.7936, + "step": 1529 + }, + { + "epoch": 0.19584, + "grad_norm": 0.6449588537216187, + "learning_rate": 0.0009296400362475148, + "loss": 2.6188, + "step": 1530 + }, + { + "epoch": 0.195968, + "grad_norm": 0.7468236088752747, + "learning_rate": 0.000929533958568618, + "loss": 2.8578, + "step": 1531 + }, + { + "epoch": 0.196096, + "grad_norm": 0.9885931611061096, + "learning_rate": 0.0009294278070478544, + "loss": 2.8432, + "step": 1532 + }, + { + "epoch": 0.196224, + "grad_norm": 0.6888399720191956, + "learning_rate": 0.0009293215817034722, + "loss": 2.8127, + "step": 1533 + }, + { + "epoch": 0.196352, + "grad_norm": 0.622332751750946, + "learning_rate": 0.000929215282553733, + "loss": 2.8664, + "step": 1534 + }, + { + "epoch": 0.19648, + "grad_norm": 0.6289708018302917, + "learning_rate": 0.0009291089096169109, + "loss": 2.7752, + "step": 1535 + }, + { + "epoch": 0.196608, + "grad_norm": 0.5639466643333435, + "learning_rate": 0.0009290024629112926, + "loss": 2.83, + "step": 1536 + }, + { + "epoch": 0.196736, + "grad_norm": 1.1604632139205933, + "learning_rate": 0.0009288959424551774, + "loss": 2.7996, + "step": 1537 + }, + { + "epoch": 0.196864, + "grad_norm": 0.5813000798225403, + "learning_rate": 0.0009287893482668774, + "loss": 2.7442, + "step": 1538 + }, + { + "epoch": 0.196992, + "grad_norm": 1.3354380130767822, + "learning_rate": 0.0009286826803647177, + "loss": 2.8519, + "step": 1539 + }, + { + "epoch": 0.19712, + "grad_norm": 0.5739391446113586, + "learning_rate": 0.0009285759387670356, + "loss": 2.8187, + "step": 1540 + }, + { + "epoch": 0.197248, + "grad_norm": 0.6101009249687195, + "learning_rate": 0.0009284691234921809, + "loss": 2.8414, + "step": 1541 + }, + { + "epoch": 0.197376, + "grad_norm": 0.5772199630737305, + "learning_rate": 0.0009283622345585168, + "loss": 2.726, + "step": 1542 + }, + { + "epoch": 0.197504, + "grad_norm": 0.7045386433601379, + "learning_rate": 0.0009282552719844185, + "loss": 2.909, + "step": 1543 + }, + { + "epoch": 0.197632, + "grad_norm": 0.608944296836853, + "learning_rate": 0.0009281482357882743, + "loss": 2.8594, + "step": 1544 + }, + { + "epoch": 0.19776, + "grad_norm": 0.6451389789581299, + "learning_rate": 0.0009280411259884846, + "loss": 2.6874, + "step": 1545 + }, + { + "epoch": 0.197888, + "grad_norm": 0.5961341857910156, + "learning_rate": 0.0009279339426034633, + "loss": 2.7211, + "step": 1546 + }, + { + "epoch": 0.198016, + "grad_norm": 1.1845489740371704, + "learning_rate": 0.0009278266856516362, + "loss": 2.7744, + "step": 1547 + }, + { + "epoch": 0.198144, + "grad_norm": 0.8207604289054871, + "learning_rate": 0.000927719355151442, + "loss": 2.7326, + "step": 1548 + }, + { + "epoch": 0.198272, + "grad_norm": 0.7027949094772339, + "learning_rate": 0.0009276119511213321, + "loss": 2.7415, + "step": 1549 + }, + { + "epoch": 0.1984, + "grad_norm": 1.2326611280441284, + "learning_rate": 0.0009275044735797705, + "loss": 2.8019, + "step": 1550 + }, + { + "epoch": 0.198528, + "grad_norm": 0.6740733981132507, + "learning_rate": 0.0009273969225452339, + "loss": 2.8231, + "step": 1551 + }, + { + "epoch": 0.198656, + "grad_norm": 0.6482301950454712, + "learning_rate": 0.0009272892980362114, + "loss": 2.8449, + "step": 1552 + }, + { + "epoch": 0.198784, + "grad_norm": 0.5775904059410095, + "learning_rate": 0.000927181600071205, + "loss": 2.8385, + "step": 1553 + }, + { + "epoch": 0.198912, + "grad_norm": 0.5908972024917603, + "learning_rate": 0.0009270738286687293, + "loss": 2.807, + "step": 1554 + }, + { + "epoch": 0.19904, + "grad_norm": 0.585676372051239, + "learning_rate": 0.0009269659838473113, + "loss": 2.8708, + "step": 1555 + }, + { + "epoch": 0.199168, + "grad_norm": 0.5811192989349365, + "learning_rate": 0.0009268580656254911, + "loss": 2.7106, + "step": 1556 + }, + { + "epoch": 0.199296, + "grad_norm": 0.6256431937217712, + "learning_rate": 0.0009267500740218205, + "loss": 2.7637, + "step": 1557 + }, + { + "epoch": 0.199424, + "grad_norm": 0.5644899606704712, + "learning_rate": 0.000926642009054865, + "loss": 2.8672, + "step": 1558 + }, + { + "epoch": 0.199552, + "grad_norm": 0.5999881029129028, + "learning_rate": 0.0009265338707432023, + "loss": 2.75, + "step": 1559 + }, + { + "epoch": 0.19968, + "grad_norm": 2.273519515991211, + "learning_rate": 0.0009264256591054222, + "loss": 2.7891, + "step": 1560 + }, + { + "epoch": 0.199808, + "grad_norm": 0.703692615032196, + "learning_rate": 0.0009263173741601279, + "loss": 2.6662, + "step": 1561 + }, + { + "epoch": 0.199936, + "grad_norm": 0.6346200108528137, + "learning_rate": 0.0009262090159259344, + "loss": 2.7391, + "step": 1562 + }, + { + "epoch": 0.200064, + "grad_norm": 0.6193463206291199, + "learning_rate": 0.0009261005844214703, + "loss": 2.8519, + "step": 1563 + }, + { + "epoch": 0.200192, + "grad_norm": 0.5907310247421265, + "learning_rate": 0.0009259920796653758, + "loss": 2.8709, + "step": 1564 + }, + { + "epoch": 0.20032, + "grad_norm": 0.5878954529762268, + "learning_rate": 0.0009258835016763043, + "loss": 2.7774, + "step": 1565 + }, + { + "epoch": 0.200448, + "grad_norm": 0.611607015132904, + "learning_rate": 0.0009257748504729218, + "loss": 2.8023, + "step": 1566 + }, + { + "epoch": 0.200576, + "grad_norm": 0.5898959636688232, + "learning_rate": 0.0009256661260739063, + "loss": 2.8498, + "step": 1567 + }, + { + "epoch": 0.200704, + "grad_norm": 0.5500224828720093, + "learning_rate": 0.000925557328497949, + "loss": 2.7885, + "step": 1568 + }, + { + "epoch": 0.200832, + "grad_norm": 0.5630072951316833, + "learning_rate": 0.0009254484577637534, + "loss": 2.6984, + "step": 1569 + }, + { + "epoch": 0.20096, + "grad_norm": 0.684489905834198, + "learning_rate": 0.0009253395138900358, + "loss": 2.6353, + "step": 1570 + }, + { + "epoch": 0.201088, + "grad_norm": 0.6115654110908508, + "learning_rate": 0.0009252304968955247, + "loss": 2.7066, + "step": 1571 + }, + { + "epoch": 0.201216, + "grad_norm": 0.6081086993217468, + "learning_rate": 0.0009251214067989616, + "loss": 2.8109, + "step": 1572 + }, + { + "epoch": 0.201344, + "grad_norm": 0.6056236028671265, + "learning_rate": 0.0009250122436191002, + "loss": 2.7249, + "step": 1573 + }, + { + "epoch": 0.201472, + "grad_norm": 0.5989468693733215, + "learning_rate": 0.0009249030073747069, + "loss": 2.788, + "step": 1574 + }, + { + "epoch": 0.2016, + "grad_norm": 0.5688483119010925, + "learning_rate": 0.0009247936980845608, + "loss": 2.7337, + "step": 1575 + }, + { + "epoch": 0.201728, + "grad_norm": 0.641139805316925, + "learning_rate": 0.0009246843157674534, + "loss": 2.7351, + "step": 1576 + }, + { + "epoch": 0.201856, + "grad_norm": 0.5944218039512634, + "learning_rate": 0.0009245748604421886, + "loss": 2.8009, + "step": 1577 + }, + { + "epoch": 0.201984, + "grad_norm": 0.6251586675643921, + "learning_rate": 0.0009244653321275833, + "loss": 2.7959, + "step": 1578 + }, + { + "epoch": 0.202112, + "grad_norm": 0.6268388628959656, + "learning_rate": 0.0009243557308424666, + "loss": 2.8453, + "step": 1579 + }, + { + "epoch": 0.20224, + "grad_norm": 0.6086769700050354, + "learning_rate": 0.0009242460566056803, + "loss": 2.7324, + "step": 1580 + }, + { + "epoch": 0.202368, + "grad_norm": 0.5927457809448242, + "learning_rate": 0.0009241363094360784, + "loss": 2.7569, + "step": 1581 + }, + { + "epoch": 0.202496, + "grad_norm": 0.6339706182479858, + "learning_rate": 0.0009240264893525281, + "loss": 2.8761, + "step": 1582 + }, + { + "epoch": 0.202624, + "grad_norm": 0.6309786438941956, + "learning_rate": 0.0009239165963739084, + "loss": 2.8367, + "step": 1583 + }, + { + "epoch": 0.202752, + "grad_norm": 0.531125545501709, + "learning_rate": 0.0009238066305191113, + "loss": 2.7737, + "step": 1584 + }, + { + "epoch": 0.20288, + "grad_norm": 0.6070235967636108, + "learning_rate": 0.0009236965918070412, + "loss": 2.7846, + "step": 1585 + }, + { + "epoch": 0.203008, + "grad_norm": 0.5815805792808533, + "learning_rate": 0.0009235864802566152, + "loss": 2.7472, + "step": 1586 + }, + { + "epoch": 0.203136, + "grad_norm": 0.5229941010475159, + "learning_rate": 0.0009234762958867626, + "loss": 2.7007, + "step": 1587 + }, + { + "epoch": 0.203264, + "grad_norm": 0.6663217544555664, + "learning_rate": 0.0009233660387164254, + "loss": 2.7146, + "step": 1588 + }, + { + "epoch": 0.203392, + "grad_norm": 0.6146247386932373, + "learning_rate": 0.0009232557087645579, + "loss": 2.6935, + "step": 1589 + }, + { + "epoch": 0.20352, + "grad_norm": 0.6247140765190125, + "learning_rate": 0.0009231453060501274, + "loss": 2.7576, + "step": 1590 + }, + { + "epoch": 0.203648, + "grad_norm": 0.7261404395103455, + "learning_rate": 0.0009230348305921131, + "loss": 2.8287, + "step": 1591 + }, + { + "epoch": 0.203776, + "grad_norm": 0.6601182222366333, + "learning_rate": 0.0009229242824095073, + "loss": 2.8417, + "step": 1592 + }, + { + "epoch": 0.203904, + "grad_norm": 0.816545307636261, + "learning_rate": 0.0009228136615213144, + "loss": 2.7785, + "step": 1593 + }, + { + "epoch": 0.204032, + "grad_norm": 0.6936466097831726, + "learning_rate": 0.0009227029679465514, + "loss": 2.7696, + "step": 1594 + }, + { + "epoch": 0.20416, + "grad_norm": 0.5794689059257507, + "learning_rate": 0.0009225922017042479, + "loss": 2.7184, + "step": 1595 + }, + { + "epoch": 0.204288, + "grad_norm": 0.5918311476707458, + "learning_rate": 0.0009224813628134457, + "loss": 2.7906, + "step": 1596 + }, + { + "epoch": 0.204416, + "grad_norm": 0.625354528427124, + "learning_rate": 0.0009223704512931994, + "loss": 2.7957, + "step": 1597 + }, + { + "epoch": 0.204544, + "grad_norm": 0.559731662273407, + "learning_rate": 0.0009222594671625759, + "loss": 2.7127, + "step": 1598 + }, + { + "epoch": 0.204672, + "grad_norm": 0.5262173414230347, + "learning_rate": 0.000922148410440655, + "loss": 2.8383, + "step": 1599 + }, + { + "epoch": 0.2048, + "grad_norm": 0.6046397089958191, + "learning_rate": 0.0009220372811465282, + "loss": 2.79, + "step": 1600 + }, + { + "epoch": 0.204928, + "grad_norm": 0.6836240887641907, + "learning_rate": 0.0009219260792993001, + "loss": 2.757, + "step": 1601 + }, + { + "epoch": 0.205056, + "grad_norm": 0.6307790279388428, + "learning_rate": 0.0009218148049180877, + "loss": 2.7505, + "step": 1602 + }, + { + "epoch": 0.205184, + "grad_norm": 0.6118954420089722, + "learning_rate": 0.0009217034580220201, + "loss": 2.7632, + "step": 1603 + }, + { + "epoch": 0.205312, + "grad_norm": 0.5859875679016113, + "learning_rate": 0.0009215920386302394, + "loss": 2.7941, + "step": 1604 + }, + { + "epoch": 0.20544, + "grad_norm": 0.5770704746246338, + "learning_rate": 0.0009214805467618997, + "loss": 2.768, + "step": 1605 + }, + { + "epoch": 0.205568, + "grad_norm": 0.6055383682250977, + "learning_rate": 0.0009213689824361678, + "loss": 2.8517, + "step": 1606 + }, + { + "epoch": 0.205696, + "grad_norm": 0.5955507159233093, + "learning_rate": 0.0009212573456722226, + "loss": 2.8709, + "step": 1607 + }, + { + "epoch": 0.205824, + "grad_norm": 0.5940427780151367, + "learning_rate": 0.0009211456364892564, + "loss": 2.6999, + "step": 1608 + }, + { + "epoch": 0.205952, + "grad_norm": 0.7476893067359924, + "learning_rate": 0.0009210338549064728, + "loss": 2.6755, + "step": 1609 + }, + { + "epoch": 0.20608, + "grad_norm": 0.585966169834137, + "learning_rate": 0.0009209220009430885, + "loss": 2.8625, + "step": 1610 + }, + { + "epoch": 0.206208, + "grad_norm": 0.7102552056312561, + "learning_rate": 0.0009208100746183326, + "loss": 2.6416, + "step": 1611 + }, + { + "epoch": 0.206336, + "grad_norm": 0.744121253490448, + "learning_rate": 0.0009206980759514464, + "loss": 2.8209, + "step": 1612 + }, + { + "epoch": 0.206464, + "grad_norm": 0.5921427607536316, + "learning_rate": 0.0009205860049616837, + "loss": 2.6643, + "step": 1613 + }, + { + "epoch": 0.206592, + "grad_norm": 0.6341274380683899, + "learning_rate": 0.000920473861668311, + "loss": 2.6944, + "step": 1614 + }, + { + "epoch": 0.20672, + "grad_norm": 0.5827640891075134, + "learning_rate": 0.0009203616460906068, + "loss": 2.7927, + "step": 1615 + }, + { + "epoch": 0.206848, + "grad_norm": 0.5493629574775696, + "learning_rate": 0.0009202493582478625, + "loss": 2.7356, + "step": 1616 + }, + { + "epoch": 0.206976, + "grad_norm": 0.5624290108680725, + "learning_rate": 0.0009201369981593814, + "loss": 2.7785, + "step": 1617 + }, + { + "epoch": 0.207104, + "grad_norm": 0.5683034062385559, + "learning_rate": 0.0009200245658444799, + "loss": 2.8448, + "step": 1618 + }, + { + "epoch": 0.207232, + "grad_norm": 0.5526583194732666, + "learning_rate": 0.0009199120613224861, + "loss": 2.629, + "step": 1619 + }, + { + "epoch": 0.20736, + "grad_norm": 0.5419575572013855, + "learning_rate": 0.0009197994846127409, + "loss": 2.7474, + "step": 1620 + }, + { + "epoch": 0.207488, + "grad_norm": 0.564158022403717, + "learning_rate": 0.0009196868357345976, + "loss": 2.9059, + "step": 1621 + }, + { + "epoch": 0.207616, + "grad_norm": 0.5982062816619873, + "learning_rate": 0.0009195741147074217, + "loss": 2.8405, + "step": 1622 + }, + { + "epoch": 0.207744, + "grad_norm": 0.5874093174934387, + "learning_rate": 0.0009194613215505915, + "loss": 2.7303, + "step": 1623 + }, + { + "epoch": 0.207872, + "grad_norm": 0.5794885158538818, + "learning_rate": 0.0009193484562834973, + "loss": 2.7397, + "step": 1624 + }, + { + "epoch": 0.208, + "grad_norm": 0.6253930926322937, + "learning_rate": 0.0009192355189255418, + "loss": 2.7759, + "step": 1625 + }, + { + "epoch": 0.208128, + "grad_norm": 0.5434867739677429, + "learning_rate": 0.0009191225094961407, + "loss": 2.6589, + "step": 1626 + }, + { + "epoch": 0.208256, + "grad_norm": 0.5666244029998779, + "learning_rate": 0.000919009428014721, + "loss": 2.753, + "step": 1627 + }, + { + "epoch": 0.208384, + "grad_norm": 0.5513975024223328, + "learning_rate": 0.0009188962745007233, + "loss": 2.7631, + "step": 1628 + }, + { + "epoch": 0.208512, + "grad_norm": 0.5388301014900208, + "learning_rate": 0.0009187830489735995, + "loss": 2.74, + "step": 1629 + }, + { + "epoch": 0.20864, + "grad_norm": 0.5168362855911255, + "learning_rate": 0.0009186697514528148, + "loss": 2.7776, + "step": 1630 + }, + { + "epoch": 0.208768, + "grad_norm": 0.5800379514694214, + "learning_rate": 0.0009185563819578462, + "loss": 2.7635, + "step": 1631 + }, + { + "epoch": 0.208896, + "grad_norm": 0.5516932606697083, + "learning_rate": 0.0009184429405081832, + "loss": 2.7846, + "step": 1632 + }, + { + "epoch": 0.209024, + "grad_norm": 0.611897349357605, + "learning_rate": 0.0009183294271233277, + "loss": 2.8096, + "step": 1633 + }, + { + "epoch": 0.209152, + "grad_norm": 0.5401003956794739, + "learning_rate": 0.000918215841822794, + "loss": 2.8011, + "step": 1634 + }, + { + "epoch": 0.20928, + "grad_norm": 0.5412010550498962, + "learning_rate": 0.0009181021846261087, + "loss": 2.6164, + "step": 1635 + }, + { + "epoch": 0.209408, + "grad_norm": 0.5777937173843384, + "learning_rate": 0.0009179884555528109, + "loss": 2.7101, + "step": 1636 + }, + { + "epoch": 0.209536, + "grad_norm": 0.5814940929412842, + "learning_rate": 0.0009178746546224517, + "loss": 2.7617, + "step": 1637 + }, + { + "epoch": 0.209664, + "grad_norm": 0.5631812810897827, + "learning_rate": 0.000917760781854595, + "loss": 2.7524, + "step": 1638 + }, + { + "epoch": 0.209792, + "grad_norm": 0.556922972202301, + "learning_rate": 0.0009176468372688168, + "loss": 2.8131, + "step": 1639 + }, + { + "epoch": 0.20992, + "grad_norm": 0.5579834580421448, + "learning_rate": 0.0009175328208847056, + "loss": 2.7663, + "step": 1640 + }, + { + "epoch": 0.210048, + "grad_norm": 0.5752408504486084, + "learning_rate": 0.000917418732721862, + "loss": 2.8604, + "step": 1641 + }, + { + "epoch": 0.210176, + "grad_norm": 0.5482182502746582, + "learning_rate": 0.000917304572799899, + "loss": 2.8557, + "step": 1642 + }, + { + "epoch": 0.210304, + "grad_norm": 0.6275211572647095, + "learning_rate": 0.0009171903411384422, + "loss": 2.824, + "step": 1643 + }, + { + "epoch": 0.210432, + "grad_norm": 0.6048553586006165, + "learning_rate": 0.0009170760377571292, + "loss": 2.612, + "step": 1644 + }, + { + "epoch": 0.21056, + "grad_norm": 0.6032798886299133, + "learning_rate": 0.0009169616626756102, + "loss": 2.7881, + "step": 1645 + }, + { + "epoch": 0.210688, + "grad_norm": 0.5460773706436157, + "learning_rate": 0.0009168472159135476, + "loss": 2.825, + "step": 1646 + }, + { + "epoch": 0.210816, + "grad_norm": 0.5550834536552429, + "learning_rate": 0.000916732697490616, + "loss": 2.734, + "step": 1647 + }, + { + "epoch": 0.210944, + "grad_norm": 0.6077369451522827, + "learning_rate": 0.0009166181074265025, + "loss": 2.7172, + "step": 1648 + }, + { + "epoch": 0.211072, + "grad_norm": 0.6191293597221375, + "learning_rate": 0.0009165034457409066, + "loss": 2.6761, + "step": 1649 + }, + { + "epoch": 0.2112, + "grad_norm": 0.5971262454986572, + "learning_rate": 0.0009163887124535397, + "loss": 2.8055, + "step": 1650 + }, + { + "epoch": 0.211328, + "grad_norm": 0.6649320125579834, + "learning_rate": 0.0009162739075841262, + "loss": 2.7275, + "step": 1651 + }, + { + "epoch": 0.211456, + "grad_norm": 0.6934270858764648, + "learning_rate": 0.000916159031152402, + "loss": 2.8427, + "step": 1652 + }, + { + "epoch": 0.211584, + "grad_norm": 0.6173274517059326, + "learning_rate": 0.0009160440831781157, + "loss": 2.7422, + "step": 1653 + }, + { + "epoch": 0.211712, + "grad_norm": 1.019655704498291, + "learning_rate": 0.0009159290636810283, + "loss": 2.8288, + "step": 1654 + }, + { + "epoch": 0.21184, + "grad_norm": 0.5642037391662598, + "learning_rate": 0.0009158139726809132, + "loss": 2.8126, + "step": 1655 + }, + { + "epoch": 0.211968, + "grad_norm": 0.5538053512573242, + "learning_rate": 0.0009156988101975554, + "loss": 2.8336, + "step": 1656 + }, + { + "epoch": 0.212096, + "grad_norm": 0.5769753456115723, + "learning_rate": 0.0009155835762507531, + "loss": 2.8141, + "step": 1657 + }, + { + "epoch": 0.212224, + "grad_norm": 0.5608971118927002, + "learning_rate": 0.0009154682708603162, + "loss": 2.6936, + "step": 1658 + }, + { + "epoch": 0.212352, + "grad_norm": 0.6361830234527588, + "learning_rate": 0.0009153528940460669, + "loss": 2.7991, + "step": 1659 + }, + { + "epoch": 0.21248, + "grad_norm": 0.6143143773078918, + "learning_rate": 0.0009152374458278401, + "loss": 2.6693, + "step": 1660 + }, + { + "epoch": 0.212608, + "grad_norm": 0.5177375078201294, + "learning_rate": 0.0009151219262254825, + "loss": 2.7288, + "step": 1661 + }, + { + "epoch": 0.212736, + "grad_norm": 0.7556158304214478, + "learning_rate": 0.0009150063352588531, + "loss": 2.7922, + "step": 1662 + }, + { + "epoch": 0.212864, + "grad_norm": 0.5196354985237122, + "learning_rate": 0.0009148906729478236, + "loss": 2.6251, + "step": 1663 + }, + { + "epoch": 0.212992, + "grad_norm": 0.5831946730613708, + "learning_rate": 0.0009147749393122775, + "loss": 2.7826, + "step": 1664 + }, + { + "epoch": 0.21312, + "grad_norm": 0.5745211243629456, + "learning_rate": 0.000914659134372111, + "loss": 2.8022, + "step": 1665 + }, + { + "epoch": 0.213248, + "grad_norm": 0.5908459424972534, + "learning_rate": 0.0009145432581472322, + "loss": 2.7725, + "step": 1666 + }, + { + "epoch": 0.213376, + "grad_norm": 0.5920864939689636, + "learning_rate": 0.0009144273106575613, + "loss": 2.7241, + "step": 1667 + }, + { + "epoch": 0.213504, + "grad_norm": 0.5600253939628601, + "learning_rate": 0.0009143112919230314, + "loss": 2.783, + "step": 1668 + }, + { + "epoch": 0.213632, + "grad_norm": 0.5919373035430908, + "learning_rate": 0.0009141952019635873, + "loss": 2.8413, + "step": 1669 + }, + { + "epoch": 0.21376, + "grad_norm": 0.6038680672645569, + "learning_rate": 0.0009140790407991861, + "loss": 2.8378, + "step": 1670 + }, + { + "epoch": 0.213888, + "grad_norm": 0.6116589307785034, + "learning_rate": 0.0009139628084497975, + "loss": 2.7675, + "step": 1671 + }, + { + "epoch": 0.214016, + "grad_norm": 0.5660625100135803, + "learning_rate": 0.0009138465049354031, + "loss": 2.7423, + "step": 1672 + }, + { + "epoch": 0.214144, + "grad_norm": 0.5730261206626892, + "learning_rate": 0.0009137301302759968, + "loss": 2.7625, + "step": 1673 + }, + { + "epoch": 0.214272, + "grad_norm": 0.5283810496330261, + "learning_rate": 0.0009136136844915846, + "loss": 2.8403, + "step": 1674 + }, + { + "epoch": 0.2144, + "grad_norm": 0.5318596959114075, + "learning_rate": 0.000913497167602185, + "loss": 2.8199, + "step": 1675 + }, + { + "epoch": 0.214528, + "grad_norm": 0.5382590293884277, + "learning_rate": 0.0009133805796278287, + "loss": 2.7469, + "step": 1676 + }, + { + "epoch": 0.214656, + "grad_norm": 0.5691607594490051, + "learning_rate": 0.0009132639205885585, + "loss": 2.7562, + "step": 1677 + }, + { + "epoch": 0.214784, + "grad_norm": 0.6125684380531311, + "learning_rate": 0.0009131471905044293, + "loss": 2.7408, + "step": 1678 + }, + { + "epoch": 0.214912, + "grad_norm": 0.6050195097923279, + "learning_rate": 0.0009130303893955084, + "loss": 2.7627, + "step": 1679 + }, + { + "epoch": 0.21504, + "grad_norm": 0.5911500453948975, + "learning_rate": 0.0009129135172818753, + "loss": 2.8156, + "step": 1680 + }, + { + "epoch": 0.215168, + "grad_norm": 0.5351465940475464, + "learning_rate": 0.0009127965741836218, + "loss": 2.7178, + "step": 1681 + }, + { + "epoch": 0.215296, + "grad_norm": 0.5542508959770203, + "learning_rate": 0.0009126795601208515, + "loss": 2.7752, + "step": 1682 + }, + { + "epoch": 0.215424, + "grad_norm": 0.5392981171607971, + "learning_rate": 0.0009125624751136807, + "loss": 2.767, + "step": 1683 + }, + { + "epoch": 0.215552, + "grad_norm": 0.8082560896873474, + "learning_rate": 0.0009124453191822376, + "loss": 2.7364, + "step": 1684 + }, + { + "epoch": 0.21568, + "grad_norm": 0.5678607821464539, + "learning_rate": 0.0009123280923466626, + "loss": 2.693, + "step": 1685 + }, + { + "epoch": 0.215808, + "grad_norm": 0.6290626525878906, + "learning_rate": 0.0009122107946271086, + "loss": 2.7074, + "step": 1686 + }, + { + "epoch": 0.215936, + "grad_norm": 0.5808846354484558, + "learning_rate": 0.00091209342604374, + "loss": 2.827, + "step": 1687 + }, + { + "epoch": 0.216064, + "grad_norm": 0.5632963180541992, + "learning_rate": 0.000911975986616734, + "loss": 2.7965, + "step": 1688 + }, + { + "epoch": 0.216192, + "grad_norm": 0.7044224143028259, + "learning_rate": 0.0009118584763662803, + "loss": 2.707, + "step": 1689 + }, + { + "epoch": 0.21632, + "grad_norm": 0.553856372833252, + "learning_rate": 0.0009117408953125795, + "loss": 2.6769, + "step": 1690 + }, + { + "epoch": 0.216448, + "grad_norm": 0.6009950637817383, + "learning_rate": 0.0009116232434758456, + "loss": 2.7807, + "step": 1691 + }, + { + "epoch": 0.216576, + "grad_norm": 0.5508719682693481, + "learning_rate": 0.0009115055208763041, + "loss": 2.7873, + "step": 1692 + }, + { + "epoch": 0.216704, + "grad_norm": 0.5741768479347229, + "learning_rate": 0.0009113877275341931, + "loss": 2.7295, + "step": 1693 + }, + { + "epoch": 0.216832, + "grad_norm": 0.5896250605583191, + "learning_rate": 0.0009112698634697624, + "loss": 2.6778, + "step": 1694 + }, + { + "epoch": 0.21696, + "grad_norm": 0.533506453037262, + "learning_rate": 0.0009111519287032743, + "loss": 2.7928, + "step": 1695 + }, + { + "epoch": 0.217088, + "grad_norm": 0.6024271249771118, + "learning_rate": 0.0009110339232550034, + "loss": 2.76, + "step": 1696 + }, + { + "epoch": 0.217216, + "grad_norm": 0.5668559074401855, + "learning_rate": 0.0009109158471452357, + "loss": 2.8063, + "step": 1697 + }, + { + "epoch": 0.217344, + "grad_norm": 0.5946394205093384, + "learning_rate": 0.0009107977003942702, + "loss": 2.7246, + "step": 1698 + }, + { + "epoch": 0.217472, + "grad_norm": 0.6549597382545471, + "learning_rate": 0.0009106794830224179, + "loss": 2.6907, + "step": 1699 + }, + { + "epoch": 0.2176, + "grad_norm": 0.526060163974762, + "learning_rate": 0.0009105611950500012, + "loss": 2.8077, + "step": 1700 + }, + { + "epoch": 0.217728, + "grad_norm": 0.5894448757171631, + "learning_rate": 0.0009104428364973553, + "loss": 2.6298, + "step": 1701 + }, + { + "epoch": 0.217856, + "grad_norm": 0.5502864122390747, + "learning_rate": 0.0009103244073848278, + "loss": 2.9057, + "step": 1702 + }, + { + "epoch": 0.217984, + "grad_norm": 0.5495195984840393, + "learning_rate": 0.0009102059077327778, + "loss": 2.8701, + "step": 1703 + }, + { + "epoch": 0.218112, + "grad_norm": 0.6184273958206177, + "learning_rate": 0.0009100873375615765, + "loss": 2.6966, + "step": 1704 + }, + { + "epoch": 0.21824, + "grad_norm": 0.5599693655967712, + "learning_rate": 0.0009099686968916082, + "loss": 2.6946, + "step": 1705 + }, + { + "epoch": 0.218368, + "grad_norm": 0.5855576395988464, + "learning_rate": 0.0009098499857432676, + "loss": 2.7181, + "step": 1706 + }, + { + "epoch": 0.218496, + "grad_norm": 0.5576251745223999, + "learning_rate": 0.0009097312041369634, + "loss": 2.707, + "step": 1707 + }, + { + "epoch": 0.218624, + "grad_norm": 0.5349119901657104, + "learning_rate": 0.000909612352093115, + "loss": 2.6928, + "step": 1708 + }, + { + "epoch": 0.218752, + "grad_norm": 0.5703234076499939, + "learning_rate": 0.0009094934296321549, + "loss": 2.7747, + "step": 1709 + }, + { + "epoch": 0.21888, + "grad_norm": 0.5703900456428528, + "learning_rate": 0.0009093744367745267, + "loss": 2.7938, + "step": 1710 + }, + { + "epoch": 0.219008, + "grad_norm": 0.5430485606193542, + "learning_rate": 0.0009092553735406871, + "loss": 2.7982, + "step": 1711 + }, + { + "epoch": 0.219136, + "grad_norm": 0.5827349424362183, + "learning_rate": 0.0009091362399511043, + "loss": 2.6996, + "step": 1712 + }, + { + "epoch": 0.219264, + "grad_norm": 0.734889030456543, + "learning_rate": 0.0009090170360262586, + "loss": 2.7443, + "step": 1713 + }, + { + "epoch": 0.219392, + "grad_norm": 0.5829759836196899, + "learning_rate": 0.0009088977617866428, + "loss": 2.7583, + "step": 1714 + }, + { + "epoch": 0.21952, + "grad_norm": 0.521007239818573, + "learning_rate": 0.0009087784172527614, + "loss": 2.7402, + "step": 1715 + }, + { + "epoch": 0.219648, + "grad_norm": 0.6352012753486633, + "learning_rate": 0.0009086590024451311, + "loss": 2.7028, + "step": 1716 + }, + { + "epoch": 0.219776, + "grad_norm": 0.5814325213432312, + "learning_rate": 0.0009085395173842807, + "loss": 2.775, + "step": 1717 + }, + { + "epoch": 0.219904, + "grad_norm": 0.5633854866027832, + "learning_rate": 0.000908419962090751, + "loss": 2.7999, + "step": 1718 + }, + { + "epoch": 0.220032, + "grad_norm": 0.5677870512008667, + "learning_rate": 0.000908300336585095, + "loss": 2.7434, + "step": 1719 + }, + { + "epoch": 0.22016, + "grad_norm": 0.6360035538673401, + "learning_rate": 0.0009081806408878778, + "loss": 2.7977, + "step": 1720 + }, + { + "epoch": 0.220288, + "grad_norm": 0.6812296509742737, + "learning_rate": 0.0009080608750196764, + "loss": 2.7647, + "step": 1721 + }, + { + "epoch": 0.220416, + "grad_norm": 0.5436292290687561, + "learning_rate": 0.0009079410390010798, + "loss": 2.7262, + "step": 1722 + }, + { + "epoch": 0.220544, + "grad_norm": 0.5323370099067688, + "learning_rate": 0.0009078211328526895, + "loss": 2.6156, + "step": 1723 + }, + { + "epoch": 0.220672, + "grad_norm": 0.566717803478241, + "learning_rate": 0.0009077011565951185, + "loss": 2.7004, + "step": 1724 + }, + { + "epoch": 0.2208, + "grad_norm": 0.6425420641899109, + "learning_rate": 0.0009075811102489923, + "loss": 2.7924, + "step": 1725 + }, + { + "epoch": 0.220928, + "grad_norm": 0.5369394421577454, + "learning_rate": 0.0009074609938349481, + "loss": 2.8685, + "step": 1726 + }, + { + "epoch": 0.221056, + "grad_norm": 0.6010847687721252, + "learning_rate": 0.0009073408073736354, + "loss": 2.7303, + "step": 1727 + }, + { + "epoch": 0.221184, + "grad_norm": 0.6874864101409912, + "learning_rate": 0.0009072205508857154, + "loss": 2.7838, + "step": 1728 + }, + { + "epoch": 0.221312, + "grad_norm": 0.5255059599876404, + "learning_rate": 0.0009071002243918621, + "loss": 2.7385, + "step": 1729 + }, + { + "epoch": 0.22144, + "grad_norm": 0.5583516359329224, + "learning_rate": 0.0009069798279127605, + "loss": 2.8614, + "step": 1730 + }, + { + "epoch": 0.221568, + "grad_norm": 0.5732067823410034, + "learning_rate": 0.0009068593614691085, + "loss": 2.7953, + "step": 1731 + }, + { + "epoch": 0.221696, + "grad_norm": 0.6962567567825317, + "learning_rate": 0.0009067388250816156, + "loss": 2.7443, + "step": 1732 + }, + { + "epoch": 0.221824, + "grad_norm": 0.587809145450592, + "learning_rate": 0.0009066182187710031, + "loss": 2.8575, + "step": 1733 + }, + { + "epoch": 0.221952, + "grad_norm": 0.5434462428092957, + "learning_rate": 0.0009064975425580051, + "loss": 2.6171, + "step": 1734 + }, + { + "epoch": 0.22208, + "grad_norm": 0.5850188732147217, + "learning_rate": 0.0009063767964633669, + "loss": 2.7316, + "step": 1735 + }, + { + "epoch": 0.222208, + "grad_norm": 0.6243098974227905, + "learning_rate": 0.0009062559805078463, + "loss": 2.7267, + "step": 1736 + }, + { + "epoch": 0.222336, + "grad_norm": 0.5583423972129822, + "learning_rate": 0.0009061350947122129, + "loss": 2.7218, + "step": 1737 + }, + { + "epoch": 0.222464, + "grad_norm": 0.5160875916481018, + "learning_rate": 0.0009060141390972485, + "loss": 2.6418, + "step": 1738 + }, + { + "epoch": 0.222592, + "grad_norm": 0.6006955504417419, + "learning_rate": 0.0009058931136837465, + "loss": 2.8299, + "step": 1739 + }, + { + "epoch": 0.22272, + "grad_norm": 0.5853919386863708, + "learning_rate": 0.0009057720184925129, + "loss": 2.7568, + "step": 1740 + }, + { + "epoch": 0.222848, + "grad_norm": 0.6064701676368713, + "learning_rate": 0.0009056508535443651, + "loss": 2.7056, + "step": 1741 + }, + { + "epoch": 0.222976, + "grad_norm": 0.6047890186309814, + "learning_rate": 0.0009055296188601329, + "loss": 2.8738, + "step": 1742 + }, + { + "epoch": 0.223104, + "grad_norm": 0.5475422739982605, + "learning_rate": 0.0009054083144606579, + "loss": 2.6745, + "step": 1743 + }, + { + "epoch": 0.223232, + "grad_norm": 0.5560581088066101, + "learning_rate": 0.0009052869403667938, + "loss": 2.8222, + "step": 1744 + }, + { + "epoch": 0.22336, + "grad_norm": 0.533816933631897, + "learning_rate": 0.000905165496599406, + "loss": 2.7471, + "step": 1745 + }, + { + "epoch": 0.223488, + "grad_norm": 0.6208500266075134, + "learning_rate": 0.0009050439831793726, + "loss": 2.7562, + "step": 1746 + }, + { + "epoch": 0.223616, + "grad_norm": 0.5637892484664917, + "learning_rate": 0.0009049224001275825, + "loss": 2.7696, + "step": 1747 + }, + { + "epoch": 0.223744, + "grad_norm": 0.7924732565879822, + "learning_rate": 0.0009048007474649377, + "loss": 2.7625, + "step": 1748 + }, + { + "epoch": 0.223872, + "grad_norm": 0.6413007974624634, + "learning_rate": 0.0009046790252123514, + "loss": 2.7311, + "step": 1749 + }, + { + "epoch": 0.224, + "grad_norm": 0.5868000984191895, + "learning_rate": 0.0009045572333907495, + "loss": 2.8201, + "step": 1750 + }, + { + "epoch": 0.224128, + "grad_norm": 0.5709046721458435, + "learning_rate": 0.000904435372021069, + "loss": 2.7413, + "step": 1751 + }, + { + "epoch": 0.224256, + "grad_norm": 0.5980023741722107, + "learning_rate": 0.0009043134411242593, + "loss": 2.736, + "step": 1752 + }, + { + "epoch": 0.224384, + "grad_norm": 2.0653412342071533, + "learning_rate": 0.0009041914407212821, + "loss": 2.8058, + "step": 1753 + }, + { + "epoch": 0.224512, + "grad_norm": 0.6288266777992249, + "learning_rate": 0.0009040693708331103, + "loss": 2.6544, + "step": 1754 + }, + { + "epoch": 0.22464, + "grad_norm": 0.6511600613594055, + "learning_rate": 0.0009039472314807293, + "loss": 2.7809, + "step": 1755 + }, + { + "epoch": 0.224768, + "grad_norm": 0.6175117492675781, + "learning_rate": 0.0009038250226851362, + "loss": 2.8528, + "step": 1756 + }, + { + "epoch": 0.224896, + "grad_norm": 0.6223093271255493, + "learning_rate": 0.0009037027444673402, + "loss": 2.7069, + "step": 1757 + }, + { + "epoch": 0.225024, + "grad_norm": 1.1270447969436646, + "learning_rate": 0.0009035803968483625, + "loss": 2.813, + "step": 1758 + }, + { + "epoch": 0.225152, + "grad_norm": 0.5581690669059753, + "learning_rate": 0.0009034579798492356, + "loss": 2.7265, + "step": 1759 + }, + { + "epoch": 0.22528, + "grad_norm": 1.2689167261123657, + "learning_rate": 0.0009033354934910048, + "loss": 2.7771, + "step": 1760 + }, + { + "epoch": 0.225408, + "grad_norm": 0.6231227517127991, + "learning_rate": 0.0009032129377947267, + "loss": 2.7715, + "step": 1761 + }, + { + "epoch": 0.225536, + "grad_norm": 0.577476978302002, + "learning_rate": 0.0009030903127814703, + "loss": 2.7126, + "step": 1762 + }, + { + "epoch": 0.225664, + "grad_norm": 0.6021720767021179, + "learning_rate": 0.0009029676184723161, + "loss": 2.6932, + "step": 1763 + }, + { + "epoch": 0.225792, + "grad_norm": 0.6132115721702576, + "learning_rate": 0.0009028448548883566, + "loss": 2.7793, + "step": 1764 + }, + { + "epoch": 0.22592, + "grad_norm": 0.6078073978424072, + "learning_rate": 0.0009027220220506965, + "loss": 2.669, + "step": 1765 + }, + { + "epoch": 0.226048, + "grad_norm": 0.6389878988265991, + "learning_rate": 0.0009025991199804517, + "loss": 2.7057, + "step": 1766 + }, + { + "epoch": 0.226176, + "grad_norm": 1.8456603288650513, + "learning_rate": 0.0009024761486987512, + "loss": 2.797, + "step": 1767 + }, + { + "epoch": 0.226304, + "grad_norm": 0.6979783773422241, + "learning_rate": 0.0009023531082267347, + "loss": 2.7083, + "step": 1768 + }, + { + "epoch": 0.226432, + "grad_norm": 0.6199493408203125, + "learning_rate": 0.0009022299985855545, + "loss": 2.8265, + "step": 1769 + }, + { + "epoch": 0.22656, + "grad_norm": 0.5692549347877502, + "learning_rate": 0.0009021068197963744, + "loss": 2.6706, + "step": 1770 + }, + { + "epoch": 0.226688, + "grad_norm": 0.579572319984436, + "learning_rate": 0.0009019835718803704, + "loss": 2.6981, + "step": 1771 + }, + { + "epoch": 0.226816, + "grad_norm": 0.6162997484207153, + "learning_rate": 0.0009018602548587302, + "loss": 2.7533, + "step": 1772 + }, + { + "epoch": 0.226944, + "grad_norm": 0.576424777507782, + "learning_rate": 0.0009017368687526535, + "loss": 2.7075, + "step": 1773 + }, + { + "epoch": 0.227072, + "grad_norm": 0.5504729747772217, + "learning_rate": 0.0009016134135833516, + "loss": 2.6589, + "step": 1774 + }, + { + "epoch": 0.2272, + "grad_norm": 0.5981659889221191, + "learning_rate": 0.000901489889372048, + "loss": 2.8223, + "step": 1775 + }, + { + "epoch": 0.227328, + "grad_norm": 0.5784311294555664, + "learning_rate": 0.000901366296139978, + "loss": 2.8014, + "step": 1776 + }, + { + "epoch": 0.227456, + "grad_norm": 0.5793903470039368, + "learning_rate": 0.0009012426339083887, + "loss": 2.7536, + "step": 1777 + }, + { + "epoch": 0.227584, + "grad_norm": 0.5997087955474854, + "learning_rate": 0.0009011189026985389, + "loss": 2.7446, + "step": 1778 + }, + { + "epoch": 0.227712, + "grad_norm": 0.5495092868804932, + "learning_rate": 0.0009009951025316997, + "loss": 2.727, + "step": 1779 + }, + { + "epoch": 0.22784, + "grad_norm": 0.6049352884292603, + "learning_rate": 0.0009008712334291535, + "loss": 2.8759, + "step": 1780 + }, + { + "epoch": 0.227968, + "grad_norm": 0.5964339971542358, + "learning_rate": 0.0009007472954121952, + "loss": 2.8372, + "step": 1781 + }, + { + "epoch": 0.228096, + "grad_norm": 0.5625279545783997, + "learning_rate": 0.0009006232885021308, + "loss": 2.8381, + "step": 1782 + }, + { + "epoch": 0.228224, + "grad_norm": 0.5701865553855896, + "learning_rate": 0.0009004992127202787, + "loss": 2.7373, + "step": 1783 + }, + { + "epoch": 0.228352, + "grad_norm": 0.5600448846817017, + "learning_rate": 0.0009003750680879689, + "loss": 2.7418, + "step": 1784 + }, + { + "epoch": 0.22848, + "grad_norm": 0.5849079489707947, + "learning_rate": 0.0009002508546265434, + "loss": 2.8037, + "step": 1785 + }, + { + "epoch": 0.228608, + "grad_norm": 0.6102093458175659, + "learning_rate": 0.0009001265723573559, + "loss": 2.8055, + "step": 1786 + }, + { + "epoch": 0.228736, + "grad_norm": 0.6025654673576355, + "learning_rate": 0.0009000022213017719, + "loss": 2.7096, + "step": 1787 + }, + { + "epoch": 0.228864, + "grad_norm": 0.6363075971603394, + "learning_rate": 0.0008998778014811689, + "loss": 2.7919, + "step": 1788 + }, + { + "epoch": 0.228992, + "grad_norm": 0.6316297054290771, + "learning_rate": 0.000899753312916936, + "loss": 2.7514, + "step": 1789 + }, + { + "epoch": 0.22912, + "grad_norm": 0.5344892144203186, + "learning_rate": 0.0008996287556304743, + "loss": 2.7374, + "step": 1790 + }, + { + "epoch": 0.229248, + "grad_norm": 0.5569682717323303, + "learning_rate": 0.0008995041296431965, + "loss": 2.7175, + "step": 1791 + }, + { + "epoch": 0.229376, + "grad_norm": 0.5568434000015259, + "learning_rate": 0.0008993794349765274, + "loss": 2.6722, + "step": 1792 + }, + { + "epoch": 0.229504, + "grad_norm": 0.5787746906280518, + "learning_rate": 0.0008992546716519034, + "loss": 2.7721, + "step": 1793 + }, + { + "epoch": 0.229632, + "grad_norm": 0.5321656465530396, + "learning_rate": 0.000899129839690773, + "loss": 2.7626, + "step": 1794 + }, + { + "epoch": 0.22976, + "grad_norm": 0.8510437607765198, + "learning_rate": 0.0008990049391145959, + "loss": 2.8595, + "step": 1795 + }, + { + "epoch": 0.229888, + "grad_norm": 0.5333331823348999, + "learning_rate": 0.0008988799699448441, + "loss": 2.7715, + "step": 1796 + }, + { + "epoch": 0.230016, + "grad_norm": 0.5505545139312744, + "learning_rate": 0.0008987549322030013, + "loss": 2.5957, + "step": 1797 + }, + { + "epoch": 0.230144, + "grad_norm": 0.6309477686882019, + "learning_rate": 0.0008986298259105629, + "loss": 2.7767, + "step": 1798 + }, + { + "epoch": 0.230272, + "grad_norm": 0.6191679239273071, + "learning_rate": 0.000898504651089036, + "loss": 2.6799, + "step": 1799 + }, + { + "epoch": 0.2304, + "grad_norm": 0.593274712562561, + "learning_rate": 0.0008983794077599398, + "loss": 2.8038, + "step": 1800 + }, + { + "epoch": 0.230528, + "grad_norm": 0.5892409086227417, + "learning_rate": 0.0008982540959448049, + "loss": 2.6642, + "step": 1801 + }, + { + "epoch": 0.230656, + "grad_norm": 0.5426162481307983, + "learning_rate": 0.000898128715665174, + "loss": 2.7403, + "step": 1802 + }, + { + "epoch": 0.230784, + "grad_norm": 0.5729665756225586, + "learning_rate": 0.0008980032669426015, + "loss": 2.7667, + "step": 1803 + }, + { + "epoch": 0.230912, + "grad_norm": 0.5569630265235901, + "learning_rate": 0.0008978777497986532, + "loss": 2.6993, + "step": 1804 + }, + { + "epoch": 0.23104, + "grad_norm": 0.5645332336425781, + "learning_rate": 0.0008977521642549072, + "loss": 2.7108, + "step": 1805 + }, + { + "epoch": 0.231168, + "grad_norm": 0.5812719464302063, + "learning_rate": 0.0008976265103329531, + "loss": 2.8377, + "step": 1806 + }, + { + "epoch": 0.231296, + "grad_norm": 0.5995230078697205, + "learning_rate": 0.0008975007880543921, + "loss": 2.7553, + "step": 1807 + }, + { + "epoch": 0.231424, + "grad_norm": 0.774755597114563, + "learning_rate": 0.0008973749974408374, + "loss": 2.793, + "step": 1808 + }, + { + "epoch": 0.231552, + "grad_norm": 0.7100176215171814, + "learning_rate": 0.0008972491385139138, + "loss": 2.6578, + "step": 1809 + }, + { + "epoch": 0.23168, + "grad_norm": 0.5912706851959229, + "learning_rate": 0.0008971232112952581, + "loss": 2.7144, + "step": 1810 + }, + { + "epoch": 0.231808, + "grad_norm": 0.6149972677230835, + "learning_rate": 0.0008969972158065185, + "loss": 2.8228, + "step": 1811 + }, + { + "epoch": 0.231936, + "grad_norm": 1.1960097551345825, + "learning_rate": 0.0008968711520693551, + "loss": 2.6056, + "step": 1812 + }, + { + "epoch": 0.232064, + "grad_norm": 0.5933524966239929, + "learning_rate": 0.0008967450201054397, + "loss": 2.644, + "step": 1813 + }, + { + "epoch": 0.232192, + "grad_norm": 0.5270127058029175, + "learning_rate": 0.000896618819936456, + "loss": 2.8606, + "step": 1814 + }, + { + "epoch": 0.23232, + "grad_norm": 0.5209812521934509, + "learning_rate": 0.0008964925515840991, + "loss": 2.6984, + "step": 1815 + }, + { + "epoch": 0.232448, + "grad_norm": 0.5721277594566345, + "learning_rate": 0.000896366215070076, + "loss": 2.6521, + "step": 1816 + }, + { + "epoch": 0.232576, + "grad_norm": 0.5251753330230713, + "learning_rate": 0.0008962398104161054, + "loss": 2.7257, + "step": 1817 + }, + { + "epoch": 0.232704, + "grad_norm": 0.6888055801391602, + "learning_rate": 0.0008961133376439178, + "loss": 2.7465, + "step": 1818 + }, + { + "epoch": 0.232832, + "grad_norm": 0.5247756838798523, + "learning_rate": 0.0008959867967752556, + "loss": 2.7998, + "step": 1819 + }, + { + "epoch": 0.23296, + "grad_norm": 0.5991374850273132, + "learning_rate": 0.000895860187831872, + "loss": 2.7562, + "step": 1820 + }, + { + "epoch": 0.233088, + "grad_norm": 0.5818384289741516, + "learning_rate": 0.0008957335108355331, + "loss": 2.7627, + "step": 1821 + }, + { + "epoch": 0.233216, + "grad_norm": 0.5520640015602112, + "learning_rate": 0.0008956067658080158, + "loss": 2.7632, + "step": 1822 + }, + { + "epoch": 0.233344, + "grad_norm": 0.5700559616088867, + "learning_rate": 0.0008954799527711094, + "loss": 2.7476, + "step": 1823 + }, + { + "epoch": 0.233472, + "grad_norm": 0.576516330242157, + "learning_rate": 0.0008953530717466142, + "loss": 2.5946, + "step": 1824 + }, + { + "epoch": 0.2336, + "grad_norm": 0.5596787333488464, + "learning_rate": 0.0008952261227563426, + "loss": 2.6655, + "step": 1825 + }, + { + "epoch": 0.233728, + "grad_norm": 0.696294903755188, + "learning_rate": 0.0008950991058221188, + "loss": 2.8382, + "step": 1826 + }, + { + "epoch": 0.233856, + "grad_norm": 0.5801590085029602, + "learning_rate": 0.0008949720209657781, + "loss": 2.7301, + "step": 1827 + }, + { + "epoch": 0.233984, + "grad_norm": 0.5753062963485718, + "learning_rate": 0.0008948448682091683, + "loss": 2.7499, + "step": 1828 + }, + { + "epoch": 0.234112, + "grad_norm": 0.5120306611061096, + "learning_rate": 0.0008947176475741481, + "loss": 2.7313, + "step": 1829 + }, + { + "epoch": 0.23424, + "grad_norm": 0.543670654296875, + "learning_rate": 0.0008945903590825883, + "loss": 2.7197, + "step": 1830 + }, + { + "epoch": 0.234368, + "grad_norm": 0.4928607642650604, + "learning_rate": 0.0008944630027563714, + "loss": 2.8583, + "step": 1831 + }, + { + "epoch": 0.234496, + "grad_norm": 0.527778148651123, + "learning_rate": 0.0008943355786173912, + "loss": 2.8003, + "step": 1832 + }, + { + "epoch": 0.234624, + "grad_norm": 0.6268999576568604, + "learning_rate": 0.0008942080866875537, + "loss": 2.6912, + "step": 1833 + }, + { + "epoch": 0.234752, + "grad_norm": 0.5320695042610168, + "learning_rate": 0.000894080526988776, + "loss": 2.8791, + "step": 1834 + }, + { + "epoch": 0.23488, + "grad_norm": 0.5674242377281189, + "learning_rate": 0.0008939528995429869, + "loss": 2.7762, + "step": 1835 + }, + { + "epoch": 0.235008, + "grad_norm": 0.5534884929656982, + "learning_rate": 0.0008938252043721274, + "loss": 2.7151, + "step": 1836 + }, + { + "epoch": 0.235136, + "grad_norm": 0.5493801832199097, + "learning_rate": 0.0008936974414981498, + "loss": 2.7745, + "step": 1837 + }, + { + "epoch": 0.235264, + "grad_norm": 0.5700369477272034, + "learning_rate": 0.0008935696109430177, + "loss": 2.7568, + "step": 1838 + }, + { + "epoch": 0.235392, + "grad_norm": 0.6391476392745972, + "learning_rate": 0.0008934417127287069, + "loss": 2.8081, + "step": 1839 + }, + { + "epoch": 0.23552, + "grad_norm": 0.5627278089523315, + "learning_rate": 0.0008933137468772046, + "loss": 2.7187, + "step": 1840 + }, + { + "epoch": 0.235648, + "grad_norm": 0.4870466887950897, + "learning_rate": 0.0008931857134105093, + "loss": 2.7287, + "step": 1841 + }, + { + "epoch": 0.235776, + "grad_norm": 0.5405287742614746, + "learning_rate": 0.0008930576123506317, + "loss": 2.7431, + "step": 1842 + }, + { + "epoch": 0.235904, + "grad_norm": 0.5109707713127136, + "learning_rate": 0.000892929443719594, + "loss": 2.7858, + "step": 1843 + }, + { + "epoch": 0.236032, + "grad_norm": 1.1445302963256836, + "learning_rate": 0.0008928012075394296, + "loss": 2.6329, + "step": 1844 + }, + { + "epoch": 0.23616, + "grad_norm": 0.5646989941596985, + "learning_rate": 0.000892672903832184, + "loss": 2.6906, + "step": 1845 + }, + { + "epoch": 0.236288, + "grad_norm": 0.5140470862388611, + "learning_rate": 0.0008925445326199138, + "loss": 2.7203, + "step": 1846 + }, + { + "epoch": 0.236416, + "grad_norm": 0.5473006963729858, + "learning_rate": 0.0008924160939246877, + "loss": 2.6792, + "step": 1847 + }, + { + "epoch": 0.236544, + "grad_norm": 0.5588535070419312, + "learning_rate": 0.0008922875877685858, + "loss": 2.7931, + "step": 1848 + }, + { + "epoch": 0.236672, + "grad_norm": 0.5372774600982666, + "learning_rate": 0.0008921590141737, + "loss": 2.7049, + "step": 1849 + }, + { + "epoch": 0.2368, + "grad_norm": 0.568909764289856, + "learning_rate": 0.0008920303731621331, + "loss": 2.7698, + "step": 1850 + }, + { + "epoch": 0.236928, + "grad_norm": 0.5480197668075562, + "learning_rate": 0.0008919016647560002, + "loss": 2.8619, + "step": 1851 + }, + { + "epoch": 0.237056, + "grad_norm": 1.0304176807403564, + "learning_rate": 0.0008917728889774279, + "loss": 2.6992, + "step": 1852 + }, + { + "epoch": 0.237184, + "grad_norm": 0.645991325378418, + "learning_rate": 0.0008916440458485542, + "loss": 2.8467, + "step": 1853 + }, + { + "epoch": 0.237312, + "grad_norm": 0.5643646717071533, + "learning_rate": 0.0008915151353915288, + "loss": 2.7491, + "step": 1854 + }, + { + "epoch": 0.23744, + "grad_norm": 0.8061437010765076, + "learning_rate": 0.0008913861576285127, + "loss": 2.698, + "step": 1855 + }, + { + "epoch": 0.237568, + "grad_norm": 0.5621592402458191, + "learning_rate": 0.0008912571125816786, + "loss": 2.7421, + "step": 1856 + }, + { + "epoch": 0.237696, + "grad_norm": 0.5776521563529968, + "learning_rate": 0.0008911280002732112, + "loss": 2.8334, + "step": 1857 + }, + { + "epoch": 0.237824, + "grad_norm": 0.5514097809791565, + "learning_rate": 0.0008909988207253061, + "loss": 2.8386, + "step": 1858 + }, + { + "epoch": 0.237952, + "grad_norm": 0.5023911595344543, + "learning_rate": 0.0008908695739601709, + "loss": 2.7495, + "step": 1859 + }, + { + "epoch": 0.23808, + "grad_norm": 0.537032425403595, + "learning_rate": 0.0008907402600000248, + "loss": 2.7346, + "step": 1860 + }, + { + "epoch": 0.238208, + "grad_norm": 0.5302927494049072, + "learning_rate": 0.0008906108788670979, + "loss": 2.7598, + "step": 1861 + }, + { + "epoch": 0.238336, + "grad_norm": 0.540716290473938, + "learning_rate": 0.0008904814305836326, + "loss": 2.757, + "step": 1862 + }, + { + "epoch": 0.238464, + "grad_norm": 0.5407140254974365, + "learning_rate": 0.0008903519151718826, + "loss": 2.6691, + "step": 1863 + }, + { + "epoch": 0.238592, + "grad_norm": 0.558936595916748, + "learning_rate": 0.000890222332654113, + "loss": 2.7372, + "step": 1864 + }, + { + "epoch": 0.23872, + "grad_norm": 0.5928846597671509, + "learning_rate": 0.0008900926830526005, + "loss": 2.8391, + "step": 1865 + }, + { + "epoch": 0.238848, + "grad_norm": 0.5569897294044495, + "learning_rate": 0.0008899629663896336, + "loss": 2.729, + "step": 1866 + }, + { + "epoch": 0.238976, + "grad_norm": 0.5800123810768127, + "learning_rate": 0.0008898331826875117, + "loss": 2.887, + "step": 1867 + }, + { + "epoch": 0.239104, + "grad_norm": 0.535680890083313, + "learning_rate": 0.0008897033319685466, + "loss": 2.7403, + "step": 1868 + }, + { + "epoch": 0.239232, + "grad_norm": 0.5999345779418945, + "learning_rate": 0.0008895734142550609, + "loss": 2.6385, + "step": 1869 + }, + { + "epoch": 0.23936, + "grad_norm": 0.519296407699585, + "learning_rate": 0.0008894434295693888, + "loss": 2.7139, + "step": 1870 + }, + { + "epoch": 0.239488, + "grad_norm": 0.5852693319320679, + "learning_rate": 0.0008893133779338764, + "loss": 2.8226, + "step": 1871 + }, + { + "epoch": 0.239616, + "grad_norm": 0.6064324378967285, + "learning_rate": 0.0008891832593708811, + "loss": 2.7331, + "step": 1872 + }, + { + "epoch": 0.239744, + "grad_norm": 0.5887659788131714, + "learning_rate": 0.0008890530739027717, + "loss": 2.7805, + "step": 1873 + }, + { + "epoch": 0.239872, + "grad_norm": 0.5569069981575012, + "learning_rate": 0.0008889228215519285, + "loss": 2.8086, + "step": 1874 + }, + { + "epoch": 0.24, + "grad_norm": 0.5670425891876221, + "learning_rate": 0.0008887925023407437, + "loss": 2.6267, + "step": 1875 + }, + { + "epoch": 0.240128, + "grad_norm": 0.5443927645683289, + "learning_rate": 0.0008886621162916203, + "loss": 2.7267, + "step": 1876 + }, + { + "epoch": 0.240256, + "grad_norm": 0.534038245677948, + "learning_rate": 0.0008885316634269733, + "loss": 2.6854, + "step": 1877 + }, + { + "epoch": 0.240384, + "grad_norm": 0.6081476807594299, + "learning_rate": 0.0008884011437692294, + "loss": 2.8065, + "step": 1878 + }, + { + "epoch": 0.240512, + "grad_norm": 0.5584090948104858, + "learning_rate": 0.0008882705573408259, + "loss": 2.7178, + "step": 1879 + }, + { + "epoch": 0.24064, + "grad_norm": 0.5038118362426758, + "learning_rate": 0.0008881399041642124, + "loss": 2.699, + "step": 1880 + }, + { + "epoch": 0.240768, + "grad_norm": 0.503760576248169, + "learning_rate": 0.0008880091842618497, + "loss": 2.7053, + "step": 1881 + }, + { + "epoch": 0.240896, + "grad_norm": 0.5326517224311829, + "learning_rate": 0.0008878783976562102, + "loss": 2.8035, + "step": 1882 + }, + { + "epoch": 0.241024, + "grad_norm": 0.5007380247116089, + "learning_rate": 0.0008877475443697773, + "loss": 2.6923, + "step": 1883 + }, + { + "epoch": 0.241152, + "grad_norm": 0.5539369583129883, + "learning_rate": 0.0008876166244250463, + "loss": 2.6909, + "step": 1884 + }, + { + "epoch": 0.24128, + "grad_norm": 0.5132195949554443, + "learning_rate": 0.0008874856378445241, + "loss": 2.682, + "step": 1885 + }, + { + "epoch": 0.241408, + "grad_norm": 0.5732895731925964, + "learning_rate": 0.0008873545846507285, + "loss": 2.8623, + "step": 1886 + }, + { + "epoch": 0.241536, + "grad_norm": 0.5636366009712219, + "learning_rate": 0.0008872234648661892, + "loss": 2.7995, + "step": 1887 + }, + { + "epoch": 0.241664, + "grad_norm": 0.5853220224380493, + "learning_rate": 0.0008870922785134472, + "loss": 2.8668, + "step": 1888 + }, + { + "epoch": 0.241792, + "grad_norm": 0.50035560131073, + "learning_rate": 0.0008869610256150551, + "loss": 2.734, + "step": 1889 + }, + { + "epoch": 0.24192, + "grad_norm": 0.5374618172645569, + "learning_rate": 0.0008868297061935766, + "loss": 2.6069, + "step": 1890 + }, + { + "epoch": 0.242048, + "grad_norm": 0.5749822854995728, + "learning_rate": 0.000886698320271587, + "loss": 2.7383, + "step": 1891 + }, + { + "epoch": 0.242176, + "grad_norm": 0.48450371623039246, + "learning_rate": 0.0008865668678716734, + "loss": 2.6363, + "step": 1892 + }, + { + "epoch": 0.242304, + "grad_norm": 0.54808509349823, + "learning_rate": 0.0008864353490164335, + "loss": 2.7769, + "step": 1893 + }, + { + "epoch": 0.242432, + "grad_norm": 0.5256767868995667, + "learning_rate": 0.0008863037637284772, + "loss": 2.6759, + "step": 1894 + }, + { + "epoch": 0.24256, + "grad_norm": 0.5679090619087219, + "learning_rate": 0.0008861721120304256, + "loss": 2.7243, + "step": 1895 + }, + { + "epoch": 0.242688, + "grad_norm": 0.5560072064399719, + "learning_rate": 0.0008860403939449109, + "loss": 2.7664, + "step": 1896 + }, + { + "epoch": 0.242816, + "grad_norm": 0.551827609539032, + "learning_rate": 0.0008859086094945772, + "loss": 2.7652, + "step": 1897 + }, + { + "epoch": 0.242944, + "grad_norm": 0.5052963495254517, + "learning_rate": 0.0008857767587020798, + "loss": 2.7726, + "step": 1898 + }, + { + "epoch": 0.243072, + "grad_norm": 0.5256214141845703, + "learning_rate": 0.000885644841590085, + "loss": 2.7464, + "step": 1899 + }, + { + "epoch": 0.2432, + "grad_norm": 0.5747609734535217, + "learning_rate": 0.0008855128581812714, + "loss": 2.7424, + "step": 1900 + }, + { + "epoch": 0.243328, + "grad_norm": 0.4850216805934906, + "learning_rate": 0.0008853808084983281, + "loss": 2.7923, + "step": 1901 + }, + { + "epoch": 0.243456, + "grad_norm": 0.5352384448051453, + "learning_rate": 0.0008852486925639561, + "loss": 2.8033, + "step": 1902 + }, + { + "epoch": 0.243584, + "grad_norm": 0.5037577152252197, + "learning_rate": 0.0008851165104008679, + "loss": 2.6401, + "step": 1903 + }, + { + "epoch": 0.243712, + "grad_norm": 0.5165992379188538, + "learning_rate": 0.0008849842620317865, + "loss": 2.6501, + "step": 1904 + }, + { + "epoch": 0.24384, + "grad_norm": 0.509350061416626, + "learning_rate": 0.0008848519474794476, + "loss": 2.7592, + "step": 1905 + }, + { + "epoch": 0.243968, + "grad_norm": 0.4816300868988037, + "learning_rate": 0.0008847195667665973, + "loss": 2.7477, + "step": 1906 + }, + { + "epoch": 0.244096, + "grad_norm": 0.5296264886856079, + "learning_rate": 0.0008845871199159934, + "loss": 2.6747, + "step": 1907 + }, + { + "epoch": 0.244224, + "grad_norm": 0.5780068635940552, + "learning_rate": 0.0008844546069504051, + "loss": 2.7379, + "step": 1908 + }, + { + "epoch": 0.244352, + "grad_norm": 0.5602445602416992, + "learning_rate": 0.0008843220278926128, + "loss": 2.6034, + "step": 1909 + }, + { + "epoch": 0.24448, + "grad_norm": 0.6009156703948975, + "learning_rate": 0.0008841893827654086, + "loss": 2.7385, + "step": 1910 + }, + { + "epoch": 0.244608, + "grad_norm": 0.6112554669380188, + "learning_rate": 0.0008840566715915955, + "loss": 2.6319, + "step": 1911 + }, + { + "epoch": 0.244736, + "grad_norm": 0.5159949660301208, + "learning_rate": 0.0008839238943939882, + "loss": 2.6188, + "step": 1912 + }, + { + "epoch": 0.244864, + "grad_norm": 0.5704315304756165, + "learning_rate": 0.0008837910511954127, + "loss": 2.7647, + "step": 1913 + }, + { + "epoch": 0.244992, + "grad_norm": 0.5631330609321594, + "learning_rate": 0.0008836581420187062, + "loss": 2.7594, + "step": 1914 + }, + { + "epoch": 0.24512, + "grad_norm": 0.5872315764427185, + "learning_rate": 0.0008835251668867172, + "loss": 2.8399, + "step": 1915 + }, + { + "epoch": 0.245248, + "grad_norm": 0.5084890127182007, + "learning_rate": 0.0008833921258223059, + "loss": 2.624, + "step": 1916 + }, + { + "epoch": 0.245376, + "grad_norm": 0.7557908296585083, + "learning_rate": 0.0008832590188483436, + "loss": 2.6391, + "step": 1917 + }, + { + "epoch": 0.245504, + "grad_norm": 0.537010908126831, + "learning_rate": 0.0008831258459877127, + "loss": 2.7672, + "step": 1918 + }, + { + "epoch": 0.245632, + "grad_norm": 0.5237004160881042, + "learning_rate": 0.0008829926072633074, + "loss": 2.7724, + "step": 1919 + }, + { + "epoch": 0.24576, + "grad_norm": 0.577792763710022, + "learning_rate": 0.0008828593026980327, + "loss": 2.7575, + "step": 1920 + }, + { + "epoch": 0.245888, + "grad_norm": 0.5184248089790344, + "learning_rate": 0.0008827259323148055, + "loss": 2.5627, + "step": 1921 + }, + { + "epoch": 0.246016, + "grad_norm": 0.572047770023346, + "learning_rate": 0.0008825924961365537, + "loss": 2.7313, + "step": 1922 + }, + { + "epoch": 0.246144, + "grad_norm": 0.5284725427627563, + "learning_rate": 0.0008824589941862163, + "loss": 2.764, + "step": 1923 + }, + { + "epoch": 0.246272, + "grad_norm": 0.50603848695755, + "learning_rate": 0.0008823254264867439, + "loss": 2.6762, + "step": 1924 + }, + { + "epoch": 0.2464, + "grad_norm": 0.49609073996543884, + "learning_rate": 0.0008821917930610985, + "loss": 2.7844, + "step": 1925 + }, + { + "epoch": 0.246528, + "grad_norm": 0.5195318460464478, + "learning_rate": 0.0008820580939322532, + "loss": 2.7654, + "step": 1926 + }, + { + "epoch": 0.246656, + "grad_norm": 0.5433355569839478, + "learning_rate": 0.0008819243291231922, + "loss": 2.6808, + "step": 1927 + }, + { + "epoch": 0.246784, + "grad_norm": 0.6128140687942505, + "learning_rate": 0.0008817904986569114, + "loss": 2.8232, + "step": 1928 + }, + { + "epoch": 0.246912, + "grad_norm": 0.5121793746948242, + "learning_rate": 0.0008816566025564179, + "loss": 2.7952, + "step": 1929 + }, + { + "epoch": 0.24704, + "grad_norm": 0.5472586750984192, + "learning_rate": 0.0008815226408447296, + "loss": 2.7215, + "step": 1930 + }, + { + "epoch": 0.247168, + "grad_norm": 0.5293869376182556, + "learning_rate": 0.0008813886135448767, + "loss": 2.6996, + "step": 1931 + }, + { + "epoch": 0.247296, + "grad_norm": 0.5513208508491516, + "learning_rate": 0.0008812545206798994, + "loss": 2.7501, + "step": 1932 + }, + { + "epoch": 0.247424, + "grad_norm": 0.5020402073860168, + "learning_rate": 0.0008811203622728503, + "loss": 2.695, + "step": 1933 + }, + { + "epoch": 0.247552, + "grad_norm": 1.4214844703674316, + "learning_rate": 0.0008809861383467925, + "loss": 2.6719, + "step": 1934 + }, + { + "epoch": 0.24768, + "grad_norm": 0.5379670262336731, + "learning_rate": 0.0008808518489248009, + "loss": 2.708, + "step": 1935 + }, + { + "epoch": 0.247808, + "grad_norm": 0.5257925987243652, + "learning_rate": 0.0008807174940299612, + "loss": 2.7136, + "step": 1936 + }, + { + "epoch": 0.247936, + "grad_norm": 0.4856511652469635, + "learning_rate": 0.0008805830736853705, + "loss": 2.6005, + "step": 1937 + }, + { + "epoch": 0.248064, + "grad_norm": 0.49013012647628784, + "learning_rate": 0.0008804485879141374, + "loss": 2.6188, + "step": 1938 + }, + { + "epoch": 0.248192, + "grad_norm": 0.5924715995788574, + "learning_rate": 0.0008803140367393815, + "loss": 2.748, + "step": 1939 + }, + { + "epoch": 0.24832, + "grad_norm": 0.53048175573349, + "learning_rate": 0.0008801794201842337, + "loss": 2.7667, + "step": 1940 + }, + { + "epoch": 0.248448, + "grad_norm": 0.615045964717865, + "learning_rate": 0.0008800447382718362, + "loss": 2.6917, + "step": 1941 + }, + { + "epoch": 0.248576, + "grad_norm": 0.5835052132606506, + "learning_rate": 0.0008799099910253423, + "loss": 2.7286, + "step": 1942 + }, + { + "epoch": 0.248704, + "grad_norm": 0.5474470257759094, + "learning_rate": 0.0008797751784679166, + "loss": 2.7898, + "step": 1943 + }, + { + "epoch": 0.248832, + "grad_norm": 0.5027730464935303, + "learning_rate": 0.000879640300622735, + "loss": 2.7885, + "step": 1944 + }, + { + "epoch": 0.24896, + "grad_norm": 0.4856603145599365, + "learning_rate": 0.0008795053575129846, + "loss": 2.8444, + "step": 1945 + }, + { + "epoch": 0.249088, + "grad_norm": 0.5373024344444275, + "learning_rate": 0.0008793703491618637, + "loss": 2.7374, + "step": 1946 + }, + { + "epoch": 0.249216, + "grad_norm": 0.5420591831207275, + "learning_rate": 0.0008792352755925817, + "loss": 2.7519, + "step": 1947 + }, + { + "epoch": 0.249344, + "grad_norm": 0.4880862236022949, + "learning_rate": 0.0008791001368283594, + "loss": 2.6944, + "step": 1948 + }, + { + "epoch": 0.249472, + "grad_norm": 0.5385234355926514, + "learning_rate": 0.0008789649328924285, + "loss": 2.7737, + "step": 1949 + }, + { + "epoch": 0.2496, + "grad_norm": 0.4868691861629486, + "learning_rate": 0.0008788296638080324, + "loss": 2.725, + "step": 1950 + }, + { + "epoch": 0.249728, + "grad_norm": 0.46566295623779297, + "learning_rate": 0.0008786943295984253, + "loss": 2.686, + "step": 1951 + }, + { + "epoch": 0.249856, + "grad_norm": 0.5268548727035522, + "learning_rate": 0.0008785589302868727, + "loss": 2.7443, + "step": 1952 + }, + { + "epoch": 0.249984, + "grad_norm": 0.501522421836853, + "learning_rate": 0.0008784234658966515, + "loss": 2.6336, + "step": 1953 + }, + { + "epoch": 0.250112, + "grad_norm": 0.6232520937919617, + "learning_rate": 0.0008782879364510494, + "loss": 2.6611, + "step": 1954 + }, + { + "epoch": 0.25024, + "grad_norm": 0.5842361450195312, + "learning_rate": 0.0008781523419733654, + "loss": 2.7569, + "step": 1955 + }, + { + "epoch": 0.250368, + "grad_norm": 0.6412742733955383, + "learning_rate": 0.00087801668248691, + "loss": 2.7931, + "step": 1956 + }, + { + "epoch": 0.250496, + "grad_norm": 0.5604439377784729, + "learning_rate": 0.0008778809580150045, + "loss": 2.7138, + "step": 1957 + }, + { + "epoch": 0.250624, + "grad_norm": 0.7050967812538147, + "learning_rate": 0.0008777451685809816, + "loss": 2.7913, + "step": 1958 + }, + { + "epoch": 0.250752, + "grad_norm": 0.7495609521865845, + "learning_rate": 0.0008776093142081849, + "loss": 2.7611, + "step": 1959 + }, + { + "epoch": 0.25088, + "grad_norm": 0.5464465618133545, + "learning_rate": 0.0008774733949199696, + "loss": 2.7192, + "step": 1960 + }, + { + "epoch": 0.251008, + "grad_norm": 0.573903501033783, + "learning_rate": 0.0008773374107397016, + "loss": 2.7001, + "step": 1961 + }, + { + "epoch": 0.251136, + "grad_norm": 0.5515214800834656, + "learning_rate": 0.0008772013616907584, + "loss": 2.6717, + "step": 1962 + }, + { + "epoch": 0.251264, + "grad_norm": 0.4912697672843933, + "learning_rate": 0.000877065247796528, + "loss": 2.7011, + "step": 1963 + }, + { + "epoch": 0.251392, + "grad_norm": 0.5166741013526917, + "learning_rate": 0.0008769290690804103, + "loss": 2.7439, + "step": 1964 + }, + { + "epoch": 0.25152, + "grad_norm": 0.692838728427887, + "learning_rate": 0.0008767928255658157, + "loss": 2.8256, + "step": 1965 + }, + { + "epoch": 0.251648, + "grad_norm": 0.5331072211265564, + "learning_rate": 0.0008766565172761664, + "loss": 2.7714, + "step": 1966 + }, + { + "epoch": 0.251776, + "grad_norm": 0.6245594024658203, + "learning_rate": 0.0008765201442348954, + "loss": 2.6777, + "step": 1967 + }, + { + "epoch": 0.251904, + "grad_norm": 0.5381416082382202, + "learning_rate": 0.0008763837064654463, + "loss": 2.7502, + "step": 1968 + }, + { + "epoch": 0.252032, + "grad_norm": 0.5605438947677612, + "learning_rate": 0.0008762472039912747, + "loss": 2.6992, + "step": 1969 + }, + { + "epoch": 0.25216, + "grad_norm": 0.5128378868103027, + "learning_rate": 0.0008761106368358469, + "loss": 2.7805, + "step": 1970 + }, + { + "epoch": 0.252288, + "grad_norm": 0.5406197309494019, + "learning_rate": 0.0008759740050226406, + "loss": 2.7453, + "step": 1971 + }, + { + "epoch": 0.252416, + "grad_norm": 0.5340100526809692, + "learning_rate": 0.0008758373085751439, + "loss": 2.6268, + "step": 1972 + }, + { + "epoch": 0.252544, + "grad_norm": 0.5095389485359192, + "learning_rate": 0.0008757005475168569, + "loss": 2.6931, + "step": 1973 + }, + { + "epoch": 0.252672, + "grad_norm": 0.7908546924591064, + "learning_rate": 0.0008755637218712902, + "loss": 2.7508, + "step": 1974 + }, + { + "epoch": 0.2528, + "grad_norm": 0.4738413989543915, + "learning_rate": 0.0008754268316619659, + "loss": 2.7239, + "step": 1975 + }, + { + "epoch": 0.252928, + "grad_norm": 0.9236796498298645, + "learning_rate": 0.0008752898769124169, + "loss": 2.8192, + "step": 1976 + }, + { + "epoch": 0.253056, + "grad_norm": 0.582481324672699, + "learning_rate": 0.0008751528576461873, + "loss": 2.7449, + "step": 1977 + }, + { + "epoch": 0.253184, + "grad_norm": 0.5827543139457703, + "learning_rate": 0.0008750157738868323, + "loss": 2.6783, + "step": 1978 + }, + { + "epoch": 0.253312, + "grad_norm": 0.4837055802345276, + "learning_rate": 0.0008748786256579182, + "loss": 2.7724, + "step": 1979 + }, + { + "epoch": 0.25344, + "grad_norm": 0.4983767867088318, + "learning_rate": 0.0008747414129830224, + "loss": 2.6582, + "step": 1980 + }, + { + "epoch": 0.253568, + "grad_norm": 0.5329318046569824, + "learning_rate": 0.0008746041358857334, + "loss": 2.7501, + "step": 1981 + }, + { + "epoch": 0.253696, + "grad_norm": 0.6032710075378418, + "learning_rate": 0.0008744667943896506, + "loss": 2.8017, + "step": 1982 + }, + { + "epoch": 0.253824, + "grad_norm": 0.5409277081489563, + "learning_rate": 0.0008743293885183848, + "loss": 2.6969, + "step": 1983 + }, + { + "epoch": 0.253952, + "grad_norm": 0.5728859305381775, + "learning_rate": 0.0008741919182955573, + "loss": 2.6272, + "step": 1984 + }, + { + "epoch": 0.25408, + "grad_norm": 0.6665219664573669, + "learning_rate": 0.0008740543837448012, + "loss": 2.8619, + "step": 1985 + }, + { + "epoch": 0.254208, + "grad_norm": 0.5601611137390137, + "learning_rate": 0.00087391678488976, + "loss": 2.6859, + "step": 1986 + }, + { + "epoch": 0.254336, + "grad_norm": 0.5875957012176514, + "learning_rate": 0.0008737791217540887, + "loss": 2.7487, + "step": 1987 + }, + { + "epoch": 0.254464, + "grad_norm": 0.555705189704895, + "learning_rate": 0.0008736413943614532, + "loss": 2.7375, + "step": 1988 + }, + { + "epoch": 0.254592, + "grad_norm": 0.669825553894043, + "learning_rate": 0.0008735036027355303, + "loss": 2.7017, + "step": 1989 + }, + { + "epoch": 0.25472, + "grad_norm": 0.5810889601707458, + "learning_rate": 0.0008733657469000081, + "loss": 2.7791, + "step": 1990 + }, + { + "epoch": 0.254848, + "grad_norm": 0.809769868850708, + "learning_rate": 0.0008732278268785856, + "loss": 2.6972, + "step": 1991 + }, + { + "epoch": 0.254976, + "grad_norm": 0.5098397731781006, + "learning_rate": 0.0008730898426949727, + "loss": 2.6991, + "step": 1992 + }, + { + "epoch": 0.255104, + "grad_norm": 0.5319783687591553, + "learning_rate": 0.0008729517943728909, + "loss": 2.7044, + "step": 1993 + }, + { + "epoch": 0.255232, + "grad_norm": 0.5790089964866638, + "learning_rate": 0.0008728136819360717, + "loss": 2.6487, + "step": 1994 + }, + { + "epoch": 0.25536, + "grad_norm": 0.5487154126167297, + "learning_rate": 0.0008726755054082588, + "loss": 2.7103, + "step": 1995 + }, + { + "epoch": 0.255488, + "grad_norm": 0.5697524547576904, + "learning_rate": 0.0008725372648132059, + "loss": 2.8932, + "step": 1996 + }, + { + "epoch": 0.255616, + "grad_norm": 0.5938974022865295, + "learning_rate": 0.0008723989601746784, + "loss": 2.6722, + "step": 1997 + }, + { + "epoch": 0.255744, + "grad_norm": 0.5522858500480652, + "learning_rate": 0.0008722605915164525, + "loss": 2.7722, + "step": 1998 + }, + { + "epoch": 0.255872, + "grad_norm": 0.566238522529602, + "learning_rate": 0.0008721221588623154, + "loss": 2.7333, + "step": 1999 + }, + { + "epoch": 0.256, + "grad_norm": 0.5676962733268738, + "learning_rate": 0.000871983662236065, + "loss": 2.8802, + "step": 2000 + }, + { + "epoch": 0.256128, + "grad_norm": 0.5711868405342102, + "learning_rate": 0.0008718451016615108, + "loss": 2.7441, + "step": 2001 + }, + { + "epoch": 0.256256, + "grad_norm": 0.5521535873413086, + "learning_rate": 0.0008717064771624727, + "loss": 2.7983, + "step": 2002 + }, + { + "epoch": 0.256384, + "grad_norm": 0.5133667588233948, + "learning_rate": 0.000871567788762782, + "loss": 2.7645, + "step": 2003 + }, + { + "epoch": 0.256512, + "grad_norm": 1.1753185987472534, + "learning_rate": 0.000871429036486281, + "loss": 2.7107, + "step": 2004 + }, + { + "epoch": 0.25664, + "grad_norm": 0.5383815765380859, + "learning_rate": 0.0008712902203568225, + "loss": 2.6407, + "step": 2005 + }, + { + "epoch": 0.256768, + "grad_norm": 0.5102605819702148, + "learning_rate": 0.0008711513403982707, + "loss": 2.796, + "step": 2006 + }, + { + "epoch": 0.256896, + "grad_norm": 0.5042199492454529, + "learning_rate": 0.0008710123966345007, + "loss": 2.7055, + "step": 2007 + }, + { + "epoch": 0.257024, + "grad_norm": 0.6168984174728394, + "learning_rate": 0.0008708733890893987, + "loss": 2.6519, + "step": 2008 + }, + { + "epoch": 0.257152, + "grad_norm": 0.6673797965049744, + "learning_rate": 0.0008707343177868617, + "loss": 2.7312, + "step": 2009 + }, + { + "epoch": 0.25728, + "grad_norm": 0.5742576122283936, + "learning_rate": 0.0008705951827507974, + "loss": 2.7946, + "step": 2010 + }, + { + "epoch": 0.257408, + "grad_norm": 0.5656010508537292, + "learning_rate": 0.0008704559840051248, + "loss": 2.747, + "step": 2011 + }, + { + "epoch": 0.257536, + "grad_norm": 0.5876798629760742, + "learning_rate": 0.0008703167215737739, + "loss": 2.648, + "step": 2012 + }, + { + "epoch": 0.257664, + "grad_norm": 0.6567627787590027, + "learning_rate": 0.0008701773954806855, + "loss": 2.7944, + "step": 2013 + }, + { + "epoch": 0.257792, + "grad_norm": 0.6031860709190369, + "learning_rate": 0.0008700380057498114, + "loss": 2.7005, + "step": 2014 + }, + { + "epoch": 0.25792, + "grad_norm": 0.5228147506713867, + "learning_rate": 0.0008698985524051143, + "loss": 2.7208, + "step": 2015 + }, + { + "epoch": 0.258048, + "grad_norm": 0.5203722715377808, + "learning_rate": 0.0008697590354705678, + "loss": 2.7684, + "step": 2016 + }, + { + "epoch": 0.258176, + "grad_norm": 0.5995572805404663, + "learning_rate": 0.0008696194549701564, + "loss": 2.7619, + "step": 2017 + }, + { + "epoch": 0.258304, + "grad_norm": 0.4906824827194214, + "learning_rate": 0.0008694798109278757, + "loss": 2.8186, + "step": 2018 + }, + { + "epoch": 0.258432, + "grad_norm": 0.5740574598312378, + "learning_rate": 0.0008693401033677321, + "loss": 2.7504, + "step": 2019 + }, + { + "epoch": 0.25856, + "grad_norm": 0.5433348417282104, + "learning_rate": 0.000869200332313743, + "loss": 2.792, + "step": 2020 + }, + { + "epoch": 0.258688, + "grad_norm": 0.6155961155891418, + "learning_rate": 0.0008690604977899368, + "loss": 2.7135, + "step": 2021 + }, + { + "epoch": 0.258816, + "grad_norm": 0.5380985736846924, + "learning_rate": 0.0008689205998203522, + "loss": 2.8413, + "step": 2022 + }, + { + "epoch": 0.258944, + "grad_norm": 0.6063902378082275, + "learning_rate": 0.0008687806384290399, + "loss": 2.6277, + "step": 2023 + }, + { + "epoch": 0.259072, + "grad_norm": 0.5541757345199585, + "learning_rate": 0.0008686406136400604, + "loss": 2.6547, + "step": 2024 + }, + { + "epoch": 0.2592, + "grad_norm": 0.5061906576156616, + "learning_rate": 0.0008685005254774858, + "loss": 2.7378, + "step": 2025 + }, + { + "epoch": 0.259328, + "grad_norm": 0.8128677606582642, + "learning_rate": 0.0008683603739653989, + "loss": 2.6717, + "step": 2026 + }, + { + "epoch": 0.259456, + "grad_norm": 0.4875728189945221, + "learning_rate": 0.0008682201591278933, + "loss": 2.6647, + "step": 2027 + }, + { + "epoch": 0.259584, + "grad_norm": 0.5105767250061035, + "learning_rate": 0.0008680798809890736, + "loss": 2.7493, + "step": 2028 + }, + { + "epoch": 0.259712, + "grad_norm": 0.5507586002349854, + "learning_rate": 0.0008679395395730552, + "loss": 2.6908, + "step": 2029 + }, + { + "epoch": 0.25984, + "grad_norm": 0.5040123462677002, + "learning_rate": 0.0008677991349039643, + "loss": 2.8191, + "step": 2030 + }, + { + "epoch": 0.259968, + "grad_norm": 0.7114841938018799, + "learning_rate": 0.0008676586670059383, + "loss": 2.711, + "step": 2031 + }, + { + "epoch": 0.260096, + "grad_norm": 0.553690493106842, + "learning_rate": 0.0008675181359031252, + "loss": 2.7327, + "step": 2032 + }, + { + "epoch": 0.260224, + "grad_norm": 0.5457369685173035, + "learning_rate": 0.0008673775416196839, + "loss": 2.7564, + "step": 2033 + }, + { + "epoch": 0.260352, + "grad_norm": 0.6097334027290344, + "learning_rate": 0.0008672368841797841, + "loss": 2.7033, + "step": 2034 + }, + { + "epoch": 0.26048, + "grad_norm": 0.5570827126502991, + "learning_rate": 0.0008670961636076067, + "loss": 2.7765, + "step": 2035 + }, + { + "epoch": 0.260608, + "grad_norm": 0.6282195448875427, + "learning_rate": 0.0008669553799273428, + "loss": 2.8019, + "step": 2036 + }, + { + "epoch": 0.260736, + "grad_norm": 0.5623509883880615, + "learning_rate": 0.0008668145331631953, + "loss": 2.6807, + "step": 2037 + }, + { + "epoch": 0.260864, + "grad_norm": 0.516298770904541, + "learning_rate": 0.0008666736233393768, + "loss": 2.7094, + "step": 2038 + }, + { + "epoch": 0.260992, + "grad_norm": 0.5591663122177124, + "learning_rate": 0.0008665326504801117, + "loss": 2.7467, + "step": 2039 + }, + { + "epoch": 0.26112, + "grad_norm": 0.5458998084068298, + "learning_rate": 0.0008663916146096348, + "loss": 2.8276, + "step": 2040 + }, + { + "epoch": 0.261248, + "grad_norm": 0.5692448616027832, + "learning_rate": 0.0008662505157521918, + "loss": 2.8547, + "step": 2041 + }, + { + "epoch": 0.261376, + "grad_norm": 0.5307191014289856, + "learning_rate": 0.0008661093539320391, + "loss": 2.7458, + "step": 2042 + }, + { + "epoch": 0.261504, + "grad_norm": 0.5887875556945801, + "learning_rate": 0.0008659681291734441, + "loss": 2.7677, + "step": 2043 + }, + { + "epoch": 0.261632, + "grad_norm": 0.8129265308380127, + "learning_rate": 0.0008658268415006851, + "loss": 2.6459, + "step": 2044 + }, + { + "epoch": 0.26176, + "grad_norm": 0.7609989643096924, + "learning_rate": 0.0008656854909380511, + "loss": 2.7404, + "step": 2045 + }, + { + "epoch": 0.261888, + "grad_norm": 0.546855092048645, + "learning_rate": 0.0008655440775098417, + "loss": 2.7221, + "step": 2046 + }, + { + "epoch": 0.262016, + "grad_norm": 0.5967245101928711, + "learning_rate": 0.0008654026012403677, + "loss": 2.7527, + "step": 2047 + }, + { + "epoch": 0.262144, + "grad_norm": 0.6165740489959717, + "learning_rate": 0.0008652610621539505, + "loss": 2.8512, + "step": 2048 + }, + { + "epoch": 0.262272, + "grad_norm": 0.5135408639907837, + "learning_rate": 0.0008651194602749221, + "loss": 2.7924, + "step": 2049 + }, + { + "epoch": 0.2624, + "grad_norm": 0.5222053527832031, + "learning_rate": 0.0008649777956276257, + "loss": 2.6621, + "step": 2050 + }, + { + "epoch": 0.262528, + "grad_norm": 0.5899808406829834, + "learning_rate": 0.0008648360682364151, + "loss": 2.6055, + "step": 2051 + }, + { + "epoch": 0.262656, + "grad_norm": 0.7222493290901184, + "learning_rate": 0.0008646942781256548, + "loss": 2.5974, + "step": 2052 + }, + { + "epoch": 0.262784, + "grad_norm": 0.5929338335990906, + "learning_rate": 0.0008645524253197201, + "loss": 2.782, + "step": 2053 + }, + { + "epoch": 0.262912, + "grad_norm": 0.572364330291748, + "learning_rate": 0.0008644105098429974, + "loss": 2.6727, + "step": 2054 + }, + { + "epoch": 0.26304, + "grad_norm": 0.6602236032485962, + "learning_rate": 0.0008642685317198833, + "loss": 2.7388, + "step": 2055 + }, + { + "epoch": 0.263168, + "grad_norm": 0.7184756398200989, + "learning_rate": 0.0008641264909747858, + "loss": 2.6357, + "step": 2056 + }, + { + "epoch": 0.263296, + "grad_norm": 0.7605315446853638, + "learning_rate": 0.0008639843876321231, + "loss": 2.7733, + "step": 2057 + }, + { + "epoch": 0.263424, + "grad_norm": 0.8635470271110535, + "learning_rate": 0.0008638422217163244, + "loss": 2.7274, + "step": 2058 + }, + { + "epoch": 0.263552, + "grad_norm": 0.5438524484634399, + "learning_rate": 0.00086369999325183, + "loss": 2.769, + "step": 2059 + }, + { + "epoch": 0.26368, + "grad_norm": 0.9255624413490295, + "learning_rate": 0.0008635577022630901, + "loss": 2.7409, + "step": 2060 + }, + { + "epoch": 0.263808, + "grad_norm": 0.5303696990013123, + "learning_rate": 0.0008634153487745666, + "loss": 2.713, + "step": 2061 + }, + { + "epoch": 0.263936, + "grad_norm": 0.7363379001617432, + "learning_rate": 0.0008632729328107317, + "loss": 2.6983, + "step": 2062 + }, + { + "epoch": 0.264064, + "grad_norm": 0.5880509614944458, + "learning_rate": 0.000863130454396068, + "loss": 2.7457, + "step": 2063 + }, + { + "epoch": 0.264192, + "grad_norm": 0.5148838758468628, + "learning_rate": 0.0008629879135550694, + "loss": 2.7786, + "step": 2064 + }, + { + "epoch": 0.26432, + "grad_norm": 0.5029786825180054, + "learning_rate": 0.0008628453103122404, + "loss": 2.759, + "step": 2065 + }, + { + "epoch": 0.264448, + "grad_norm": 0.8820178508758545, + "learning_rate": 0.0008627026446920963, + "loss": 2.7527, + "step": 2066 + }, + { + "epoch": 0.264576, + "grad_norm": 1.9845834970474243, + "learning_rate": 0.0008625599167191626, + "loss": 2.7881, + "step": 2067 + }, + { + "epoch": 0.264704, + "grad_norm": 0.5779441595077515, + "learning_rate": 0.0008624171264179761, + "loss": 2.7178, + "step": 2068 + }, + { + "epoch": 0.264832, + "grad_norm": 0.5626403093338013, + "learning_rate": 0.000862274273813084, + "loss": 2.7475, + "step": 2069 + }, + { + "epoch": 0.26496, + "grad_norm": 0.5690915584564209, + "learning_rate": 0.0008621313589290445, + "loss": 2.8884, + "step": 2070 + }, + { + "epoch": 0.265088, + "grad_norm": 0.5364595651626587, + "learning_rate": 0.0008619883817904262, + "loss": 2.7797, + "step": 2071 + }, + { + "epoch": 0.265216, + "grad_norm": 0.6102113127708435, + "learning_rate": 0.0008618453424218084, + "loss": 2.687, + "step": 2072 + }, + { + "epoch": 0.265344, + "grad_norm": 0.5840786695480347, + "learning_rate": 0.0008617022408477815, + "loss": 2.9046, + "step": 2073 + }, + { + "epoch": 0.265472, + "grad_norm": 0.581188440322876, + "learning_rate": 0.000861559077092946, + "loss": 2.8247, + "step": 2074 + }, + { + "epoch": 0.2656, + "grad_norm": 0.5462239980697632, + "learning_rate": 0.0008614158511819138, + "loss": 2.7113, + "step": 2075 + }, + { + "epoch": 0.265728, + "grad_norm": 0.5585112571716309, + "learning_rate": 0.0008612725631393068, + "loss": 2.8027, + "step": 2076 + }, + { + "epoch": 0.265856, + "grad_norm": 0.5565884709358215, + "learning_rate": 0.0008611292129897581, + "loss": 2.7272, + "step": 2077 + }, + { + "epoch": 0.265984, + "grad_norm": 0.5338641405105591, + "learning_rate": 0.000860985800757911, + "loss": 2.6894, + "step": 2078 + }, + { + "epoch": 0.266112, + "grad_norm": 0.5015736222267151, + "learning_rate": 0.0008608423264684198, + "loss": 2.6124, + "step": 2079 + }, + { + "epoch": 0.26624, + "grad_norm": 0.6066538095474243, + "learning_rate": 0.0008606987901459496, + "loss": 2.6111, + "step": 2080 + }, + { + "epoch": 0.266368, + "grad_norm": 0.5827579498291016, + "learning_rate": 0.0008605551918151755, + "loss": 2.6753, + "step": 2081 + }, + { + "epoch": 0.266496, + "grad_norm": 0.5312411785125732, + "learning_rate": 0.0008604115315007843, + "loss": 2.7404, + "step": 2082 + }, + { + "epoch": 0.266624, + "grad_norm": 0.5452515482902527, + "learning_rate": 0.0008602678092274725, + "loss": 2.6823, + "step": 2083 + }, + { + "epoch": 0.266752, + "grad_norm": 0.5418721437454224, + "learning_rate": 0.0008601240250199476, + "loss": 2.7015, + "step": 2084 + }, + { + "epoch": 0.26688, + "grad_norm": 0.5737361311912537, + "learning_rate": 0.000859980178902928, + "loss": 2.743, + "step": 2085 + }, + { + "epoch": 0.267008, + "grad_norm": 0.5255559086799622, + "learning_rate": 0.0008598362709011424, + "loss": 2.7129, + "step": 2086 + }, + { + "epoch": 0.267136, + "grad_norm": 0.5412623882293701, + "learning_rate": 0.0008596923010393303, + "loss": 2.7338, + "step": 2087 + }, + { + "epoch": 0.267264, + "grad_norm": 0.5262688994407654, + "learning_rate": 0.0008595482693422417, + "loss": 2.8859, + "step": 2088 + }, + { + "epoch": 0.267392, + "grad_norm": 0.5399100184440613, + "learning_rate": 0.0008594041758346371, + "loss": 2.669, + "step": 2089 + }, + { + "epoch": 0.26752, + "grad_norm": 0.4655872583389282, + "learning_rate": 0.0008592600205412883, + "loss": 2.7368, + "step": 2090 + }, + { + "epoch": 0.267648, + "grad_norm": 0.49932432174682617, + "learning_rate": 0.0008591158034869773, + "loss": 2.7823, + "step": 2091 + }, + { + "epoch": 0.267776, + "grad_norm": 0.4971008896827698, + "learning_rate": 0.0008589715246964963, + "loss": 2.7492, + "step": 2092 + }, + { + "epoch": 0.267904, + "grad_norm": 0.5891103148460388, + "learning_rate": 0.0008588271841946484, + "loss": 2.7484, + "step": 2093 + }, + { + "epoch": 0.268032, + "grad_norm": 0.516826331615448, + "learning_rate": 0.000858682782006248, + "loss": 2.7914, + "step": 2094 + }, + { + "epoch": 0.26816, + "grad_norm": 0.5383368134498596, + "learning_rate": 0.000858538318156119, + "loss": 2.714, + "step": 2095 + }, + { + "epoch": 0.268288, + "grad_norm": 0.5893653035163879, + "learning_rate": 0.0008583937926690965, + "loss": 2.6925, + "step": 2096 + }, + { + "epoch": 0.268416, + "grad_norm": 0.5459854602813721, + "learning_rate": 0.0008582492055700264, + "loss": 2.7893, + "step": 2097 + }, + { + "epoch": 0.268544, + "grad_norm": 1.451204776763916, + "learning_rate": 0.0008581045568837645, + "loss": 2.6642, + "step": 2098 + }, + { + "epoch": 0.268672, + "grad_norm": 0.5352036356925964, + "learning_rate": 0.0008579598466351779, + "loss": 2.7392, + "step": 2099 + }, + { + "epoch": 0.2688, + "grad_norm": 0.5973572134971619, + "learning_rate": 0.0008578150748491438, + "loss": 2.6354, + "step": 2100 + }, + { + "epoch": 0.268928, + "grad_norm": 0.8076955080032349, + "learning_rate": 0.0008576702415505501, + "loss": 2.8232, + "step": 2101 + }, + { + "epoch": 0.269056, + "grad_norm": 0.5685824155807495, + "learning_rate": 0.0008575253467642954, + "loss": 2.7448, + "step": 2102 + }, + { + "epoch": 0.269184, + "grad_norm": 0.7295274138450623, + "learning_rate": 0.000857380390515289, + "loss": 2.7431, + "step": 2103 + }, + { + "epoch": 0.269312, + "grad_norm": 0.5350815653800964, + "learning_rate": 0.0008572353728284501, + "loss": 2.7019, + "step": 2104 + }, + { + "epoch": 0.26944, + "grad_norm": 0.5247540473937988, + "learning_rate": 0.0008570902937287092, + "loss": 2.8055, + "step": 2105 + }, + { + "epoch": 0.269568, + "grad_norm": 0.542457103729248, + "learning_rate": 0.0008569451532410074, + "loss": 2.6874, + "step": 2106 + }, + { + "epoch": 0.269696, + "grad_norm": 0.6246141791343689, + "learning_rate": 0.0008567999513902953, + "loss": 2.7564, + "step": 2107 + }, + { + "epoch": 0.269824, + "grad_norm": 0.5577154159545898, + "learning_rate": 0.0008566546882015354, + "loss": 2.5852, + "step": 2108 + }, + { + "epoch": 0.269952, + "grad_norm": 1.1153438091278076, + "learning_rate": 0.0008565093636996999, + "loss": 2.6685, + "step": 2109 + }, + { + "epoch": 0.27008, + "grad_norm": 0.5495931506156921, + "learning_rate": 0.0008563639779097716, + "loss": 2.7498, + "step": 2110 + }, + { + "epoch": 0.270208, + "grad_norm": 0.6265218257904053, + "learning_rate": 0.0008562185308567443, + "loss": 2.7854, + "step": 2111 + }, + { + "epoch": 0.270336, + "grad_norm": 0.5860053896903992, + "learning_rate": 0.000856073022565622, + "loss": 2.714, + "step": 2112 + }, + { + "epoch": 0.270464, + "grad_norm": 0.5405157208442688, + "learning_rate": 0.0008559274530614189, + "loss": 2.7615, + "step": 2113 + }, + { + "epoch": 0.270592, + "grad_norm": 0.5539007782936096, + "learning_rate": 0.0008557818223691607, + "loss": 2.7512, + "step": 2114 + }, + { + "epoch": 0.27072, + "grad_norm": 0.49888473749160767, + "learning_rate": 0.0008556361305138825, + "loss": 2.8248, + "step": 2115 + }, + { + "epoch": 0.270848, + "grad_norm": 0.5659744143486023, + "learning_rate": 0.0008554903775206304, + "loss": 2.7231, + "step": 2116 + }, + { + "epoch": 0.270976, + "grad_norm": 0.5525941848754883, + "learning_rate": 0.0008553445634144614, + "loss": 2.7072, + "step": 2117 + }, + { + "epoch": 0.271104, + "grad_norm": 0.5901686549186707, + "learning_rate": 0.0008551986882204424, + "loss": 2.7423, + "step": 2118 + }, + { + "epoch": 0.271232, + "grad_norm": 0.5768177509307861, + "learning_rate": 0.0008550527519636511, + "loss": 2.7605, + "step": 2119 + }, + { + "epoch": 0.27136, + "grad_norm": 0.4890199601650238, + "learning_rate": 0.0008549067546691756, + "loss": 2.7437, + "step": 2120 + }, + { + "epoch": 0.271488, + "grad_norm": 0.5464968681335449, + "learning_rate": 0.0008547606963621145, + "loss": 2.7079, + "step": 2121 + }, + { + "epoch": 0.271616, + "grad_norm": 0.5759803652763367, + "learning_rate": 0.000854614577067577, + "loss": 2.7839, + "step": 2122 + }, + { + "epoch": 0.271744, + "grad_norm": 0.5245680212974548, + "learning_rate": 0.0008544683968106826, + "loss": 2.6912, + "step": 2123 + }, + { + "epoch": 0.271872, + "grad_norm": 0.5321072936058044, + "learning_rate": 0.0008543221556165614, + "loss": 2.7707, + "step": 2124 + }, + { + "epoch": 0.272, + "grad_norm": 0.5317277312278748, + "learning_rate": 0.0008541758535103541, + "loss": 2.7679, + "step": 2125 + }, + { + "epoch": 0.272128, + "grad_norm": 0.5957646369934082, + "learning_rate": 0.0008540294905172117, + "loss": 2.776, + "step": 2126 + }, + { + "epoch": 0.272256, + "grad_norm": 0.5257060527801514, + "learning_rate": 0.0008538830666622954, + "loss": 2.6879, + "step": 2127 + }, + { + "epoch": 0.272384, + "grad_norm": 0.5462499856948853, + "learning_rate": 0.0008537365819707776, + "loss": 2.7069, + "step": 2128 + }, + { + "epoch": 0.272512, + "grad_norm": 0.5775899291038513, + "learning_rate": 0.0008535900364678402, + "loss": 2.7586, + "step": 2129 + }, + { + "epoch": 0.27264, + "grad_norm": 0.5264943242073059, + "learning_rate": 0.0008534434301786766, + "loss": 2.6966, + "step": 2130 + }, + { + "epoch": 0.272768, + "grad_norm": 0.50956791639328, + "learning_rate": 0.0008532967631284898, + "loss": 2.6611, + "step": 2131 + }, + { + "epoch": 0.272896, + "grad_norm": 0.5075507164001465, + "learning_rate": 0.0008531500353424937, + "loss": 2.6989, + "step": 2132 + }, + { + "epoch": 0.273024, + "grad_norm": 0.5973472595214844, + "learning_rate": 0.0008530032468459126, + "loss": 2.6627, + "step": 2133 + }, + { + "epoch": 0.273152, + "grad_norm": 0.5526325702667236, + "learning_rate": 0.0008528563976639807, + "loss": 2.7352, + "step": 2134 + }, + { + "epoch": 0.27328, + "grad_norm": 0.5524231195449829, + "learning_rate": 0.0008527094878219434, + "loss": 2.6697, + "step": 2135 + }, + { + "epoch": 0.273408, + "grad_norm": 0.4792745113372803, + "learning_rate": 0.0008525625173450564, + "loss": 2.7937, + "step": 2136 + }, + { + "epoch": 0.273536, + "grad_norm": 0.521050214767456, + "learning_rate": 0.0008524154862585853, + "loss": 2.8144, + "step": 2137 + }, + { + "epoch": 0.273664, + "grad_norm": 0.5374805331230164, + "learning_rate": 0.0008522683945878066, + "loss": 2.6759, + "step": 2138 + }, + { + "epoch": 0.273792, + "grad_norm": 0.49986332654953003, + "learning_rate": 0.000852121242358007, + "loss": 2.678, + "step": 2139 + }, + { + "epoch": 0.27392, + "grad_norm": 0.5627812147140503, + "learning_rate": 0.0008519740295944838, + "loss": 2.7831, + "step": 2140 + }, + { + "epoch": 0.274048, + "grad_norm": 0.5308534502983093, + "learning_rate": 0.0008518267563225444, + "loss": 2.8108, + "step": 2141 + }, + { + "epoch": 0.274176, + "grad_norm": 0.5385130643844604, + "learning_rate": 0.0008516794225675069, + "loss": 2.7812, + "step": 2142 + }, + { + "epoch": 0.274304, + "grad_norm": 0.5822041630744934, + "learning_rate": 0.0008515320283546995, + "loss": 2.8096, + "step": 2143 + }, + { + "epoch": 0.274432, + "grad_norm": 0.5466216802597046, + "learning_rate": 0.0008513845737094612, + "loss": 2.6554, + "step": 2144 + }, + { + "epoch": 0.27456, + "grad_norm": 0.5679618716239929, + "learning_rate": 0.0008512370586571409, + "loss": 2.7489, + "step": 2145 + }, + { + "epoch": 0.274688, + "grad_norm": 0.5938407778739929, + "learning_rate": 0.0008510894832230987, + "loss": 2.737, + "step": 2146 + }, + { + "epoch": 0.274816, + "grad_norm": 0.48485463857650757, + "learning_rate": 0.000850941847432704, + "loss": 2.7208, + "step": 2147 + }, + { + "epoch": 0.274944, + "grad_norm": 0.5022928714752197, + "learning_rate": 0.000850794151311337, + "loss": 2.6942, + "step": 2148 + }, + { + "epoch": 0.275072, + "grad_norm": 0.49790874123573303, + "learning_rate": 0.000850646394884389, + "loss": 2.7523, + "step": 2149 + }, + { + "epoch": 0.2752, + "grad_norm": 0.5469455718994141, + "learning_rate": 0.0008504985781772605, + "loss": 2.7232, + "step": 2150 + }, + { + "epoch": 0.275328, + "grad_norm": 0.5304444432258606, + "learning_rate": 0.0008503507012153632, + "loss": 2.748, + "step": 2151 + }, + { + "epoch": 0.275456, + "grad_norm": 0.5316221117973328, + "learning_rate": 0.0008502027640241187, + "loss": 2.8533, + "step": 2152 + }, + { + "epoch": 0.275584, + "grad_norm": 0.5415834188461304, + "learning_rate": 0.000850054766628959, + "loss": 2.7636, + "step": 2153 + }, + { + "epoch": 0.275712, + "grad_norm": 0.5462274551391602, + "learning_rate": 0.0008499067090553269, + "loss": 2.7985, + "step": 2154 + }, + { + "epoch": 0.27584, + "grad_norm": 0.5090065598487854, + "learning_rate": 0.0008497585913286752, + "loss": 2.7975, + "step": 2155 + }, + { + "epoch": 0.275968, + "grad_norm": 0.5620598196983337, + "learning_rate": 0.0008496104134744667, + "loss": 2.7759, + "step": 2156 + }, + { + "epoch": 0.276096, + "grad_norm": 0.5368714928627014, + "learning_rate": 0.000849462175518175, + "loss": 2.6583, + "step": 2157 + }, + { + "epoch": 0.276224, + "grad_norm": 0.5545941591262817, + "learning_rate": 0.0008493138774852841, + "loss": 2.7243, + "step": 2158 + }, + { + "epoch": 0.276352, + "grad_norm": 0.5175477266311646, + "learning_rate": 0.0008491655194012881, + "loss": 2.6613, + "step": 2159 + }, + { + "epoch": 0.27648, + "grad_norm": 0.4776386320590973, + "learning_rate": 0.0008490171012916914, + "loss": 2.7482, + "step": 2160 + }, + { + "epoch": 0.276608, + "grad_norm": 0.5403317809104919, + "learning_rate": 0.0008488686231820088, + "loss": 2.7232, + "step": 2161 + }, + { + "epoch": 0.276736, + "grad_norm": 0.5945667028427124, + "learning_rate": 0.0008487200850977657, + "loss": 2.7174, + "step": 2162 + }, + { + "epoch": 0.276864, + "grad_norm": 0.5367197394371033, + "learning_rate": 0.0008485714870644969, + "loss": 2.6281, + "step": 2163 + }, + { + "epoch": 0.276992, + "grad_norm": 0.49431636929512024, + "learning_rate": 0.0008484228291077488, + "loss": 2.6785, + "step": 2164 + }, + { + "epoch": 0.27712, + "grad_norm": 1.4602652788162231, + "learning_rate": 0.0008482741112530769, + "loss": 2.6681, + "step": 2165 + }, + { + "epoch": 0.277248, + "grad_norm": 0.5216421484947205, + "learning_rate": 0.0008481253335260478, + "loss": 2.7381, + "step": 2166 + }, + { + "epoch": 0.277376, + "grad_norm": 0.505403995513916, + "learning_rate": 0.000847976495952238, + "loss": 2.7431, + "step": 2167 + }, + { + "epoch": 0.277504, + "grad_norm": 0.6463417410850525, + "learning_rate": 0.0008478275985572346, + "loss": 2.6558, + "step": 2168 + }, + { + "epoch": 0.277632, + "grad_norm": 0.5575148463249207, + "learning_rate": 0.0008476786413666345, + "loss": 2.6795, + "step": 2169 + }, + { + "epoch": 0.27776, + "grad_norm": 0.5822579860687256, + "learning_rate": 0.0008475296244060454, + "loss": 2.6181, + "step": 2170 + }, + { + "epoch": 0.277888, + "grad_norm": 0.574121356010437, + "learning_rate": 0.0008473805477010849, + "loss": 2.7176, + "step": 2171 + }, + { + "epoch": 0.278016, + "grad_norm": 0.5547507405281067, + "learning_rate": 0.0008472314112773811, + "loss": 2.7771, + "step": 2172 + }, + { + "epoch": 0.278144, + "grad_norm": 0.59307461977005, + "learning_rate": 0.0008470822151605723, + "loss": 2.7463, + "step": 2173 + }, + { + "epoch": 0.278272, + "grad_norm": 0.48492565751075745, + "learning_rate": 0.0008469329593763069, + "loss": 2.6321, + "step": 2174 + }, + { + "epoch": 0.2784, + "grad_norm": 0.6304513216018677, + "learning_rate": 0.0008467836439502439, + "loss": 2.8099, + "step": 2175 + }, + { + "epoch": 0.278528, + "grad_norm": 0.5091226100921631, + "learning_rate": 0.000846634268908052, + "loss": 2.7805, + "step": 2176 + }, + { + "epoch": 0.278656, + "grad_norm": 0.5389769077301025, + "learning_rate": 0.000846484834275411, + "loss": 2.5795, + "step": 2177 + }, + { + "epoch": 0.278784, + "grad_norm": 0.5702961683273315, + "learning_rate": 0.0008463353400780101, + "loss": 2.8064, + "step": 2178 + }, + { + "epoch": 0.278912, + "grad_norm": 0.5212931036949158, + "learning_rate": 0.0008461857863415492, + "loss": 2.7546, + "step": 2179 + }, + { + "epoch": 0.27904, + "grad_norm": 0.541519820690155, + "learning_rate": 0.0008460361730917383, + "loss": 2.7417, + "step": 2180 + }, + { + "epoch": 0.279168, + "grad_norm": 0.7998132109642029, + "learning_rate": 0.0008458865003542975, + "loss": 2.8394, + "step": 2181 + }, + { + "epoch": 0.279296, + "grad_norm": 0.5797017812728882, + "learning_rate": 0.0008457367681549575, + "loss": 2.8063, + "step": 2182 + }, + { + "epoch": 0.279424, + "grad_norm": 0.513437807559967, + "learning_rate": 0.0008455869765194591, + "loss": 2.7008, + "step": 2183 + }, + { + "epoch": 0.279552, + "grad_norm": 0.7775626182556152, + "learning_rate": 0.000845437125473553, + "loss": 2.6971, + "step": 2184 + }, + { + "epoch": 0.27968, + "grad_norm": 0.5727411508560181, + "learning_rate": 0.0008452872150430002, + "loss": 2.6882, + "step": 2185 + }, + { + "epoch": 0.279808, + "grad_norm": 0.5552663207054138, + "learning_rate": 0.0008451372452535724, + "loss": 2.6989, + "step": 2186 + }, + { + "epoch": 0.279936, + "grad_norm": 0.5015628337860107, + "learning_rate": 0.000844987216131051, + "loss": 2.7587, + "step": 2187 + }, + { + "epoch": 0.280064, + "grad_norm": 0.5751020908355713, + "learning_rate": 0.0008448371277012276, + "loss": 2.7277, + "step": 2188 + }, + { + "epoch": 0.280192, + "grad_norm": 0.5536028742790222, + "learning_rate": 0.0008446869799899042, + "loss": 2.7023, + "step": 2189 + }, + { + "epoch": 0.28032, + "grad_norm": 0.4952138662338257, + "learning_rate": 0.000844536773022893, + "loss": 2.7581, + "step": 2190 + }, + { + "epoch": 0.280448, + "grad_norm": 0.4865911900997162, + "learning_rate": 0.0008443865068260164, + "loss": 2.7633, + "step": 2191 + }, + { + "epoch": 0.280576, + "grad_norm": 0.509646475315094, + "learning_rate": 0.0008442361814251069, + "loss": 2.6441, + "step": 2192 + }, + { + "epoch": 0.280704, + "grad_norm": 0.5118752121925354, + "learning_rate": 0.0008440857968460069, + "loss": 2.7458, + "step": 2193 + }, + { + "epoch": 0.280832, + "grad_norm": 0.5439507961273193, + "learning_rate": 0.0008439353531145695, + "loss": 2.8314, + "step": 2194 + }, + { + "epoch": 0.28096, + "grad_norm": 0.5212950706481934, + "learning_rate": 0.0008437848502566575, + "loss": 2.8071, + "step": 2195 + }, + { + "epoch": 0.281088, + "grad_norm": 0.5042040944099426, + "learning_rate": 0.0008436342882981445, + "loss": 2.7478, + "step": 2196 + }, + { + "epoch": 0.281216, + "grad_norm": 0.538835883140564, + "learning_rate": 0.0008434836672649135, + "loss": 2.7484, + "step": 2197 + }, + { + "epoch": 0.281344, + "grad_norm": 0.560782790184021, + "learning_rate": 0.0008433329871828582, + "loss": 2.7871, + "step": 2198 + }, + { + "epoch": 0.281472, + "grad_norm": 0.5447134375572205, + "learning_rate": 0.0008431822480778819, + "loss": 2.6482, + "step": 2199 + }, + { + "epoch": 0.2816, + "grad_norm": 0.505682110786438, + "learning_rate": 0.0008430314499758986, + "loss": 2.6842, + "step": 2200 + }, + { + "epoch": 0.281728, + "grad_norm": 0.5328322052955627, + "learning_rate": 0.0008428805929028326, + "loss": 2.7011, + "step": 2201 + }, + { + "epoch": 0.281856, + "grad_norm": 0.5166845917701721, + "learning_rate": 0.0008427296768846175, + "loss": 2.7559, + "step": 2202 + }, + { + "epoch": 0.281984, + "grad_norm": 0.5435641407966614, + "learning_rate": 0.0008425787019471978, + "loss": 2.6703, + "step": 2203 + }, + { + "epoch": 0.282112, + "grad_norm": 0.5395270586013794, + "learning_rate": 0.0008424276681165275, + "loss": 2.7412, + "step": 2204 + }, + { + "epoch": 0.28224, + "grad_norm": 0.5111108422279358, + "learning_rate": 0.0008422765754185715, + "loss": 2.7033, + "step": 2205 + }, + { + "epoch": 0.282368, + "grad_norm": 0.554808497428894, + "learning_rate": 0.000842125423879304, + "loss": 2.6524, + "step": 2206 + }, + { + "epoch": 0.282496, + "grad_norm": 0.5661003589630127, + "learning_rate": 0.00084197421352471, + "loss": 2.795, + "step": 2207 + }, + { + "epoch": 0.282624, + "grad_norm": 0.49313610792160034, + "learning_rate": 0.0008418229443807841, + "loss": 2.6698, + "step": 2208 + }, + { + "epoch": 0.282752, + "grad_norm": 0.47432395815849304, + "learning_rate": 0.0008416716164735314, + "loss": 2.7345, + "step": 2209 + }, + { + "epoch": 0.28288, + "grad_norm": 0.6603214144706726, + "learning_rate": 0.0008415202298289668, + "loss": 2.792, + "step": 2210 + }, + { + "epoch": 0.283008, + "grad_norm": 0.5673984289169312, + "learning_rate": 0.0008413687844731155, + "loss": 2.7402, + "step": 2211 + }, + { + "epoch": 0.283136, + "grad_norm": 0.5082360506057739, + "learning_rate": 0.0008412172804320126, + "loss": 2.6704, + "step": 2212 + }, + { + "epoch": 0.283264, + "grad_norm": 0.6179699301719666, + "learning_rate": 0.0008410657177317034, + "loss": 2.7528, + "step": 2213 + }, + { + "epoch": 0.283392, + "grad_norm": 0.5587566494941711, + "learning_rate": 0.0008409140963982435, + "loss": 2.6895, + "step": 2214 + }, + { + "epoch": 0.28352, + "grad_norm": 0.5643453598022461, + "learning_rate": 0.0008407624164576981, + "loss": 2.6554, + "step": 2215 + }, + { + "epoch": 0.283648, + "grad_norm": 0.525655210018158, + "learning_rate": 0.0008406106779361429, + "loss": 2.7391, + "step": 2216 + }, + { + "epoch": 0.283776, + "grad_norm": 0.605749785900116, + "learning_rate": 0.0008404588808596634, + "loss": 2.7498, + "step": 2217 + }, + { + "epoch": 0.283904, + "grad_norm": 0.5916219353675842, + "learning_rate": 0.0008403070252543555, + "loss": 2.7515, + "step": 2218 + }, + { + "epoch": 0.284032, + "grad_norm": 0.5468689203262329, + "learning_rate": 0.0008401551111463245, + "loss": 2.7477, + "step": 2219 + }, + { + "epoch": 0.28416, + "grad_norm": 0.5176936984062195, + "learning_rate": 0.0008400031385616867, + "loss": 2.7447, + "step": 2220 + }, + { + "epoch": 0.284288, + "grad_norm": 0.48176297545433044, + "learning_rate": 0.0008398511075265677, + "loss": 2.7559, + "step": 2221 + }, + { + "epoch": 0.284416, + "grad_norm": 0.5551603436470032, + "learning_rate": 0.0008396990180671033, + "loss": 2.7289, + "step": 2222 + }, + { + "epoch": 0.284544, + "grad_norm": 0.533989429473877, + "learning_rate": 0.0008395468702094398, + "loss": 2.6434, + "step": 2223 + }, + { + "epoch": 0.284672, + "grad_norm": 0.5383233428001404, + "learning_rate": 0.0008393946639797328, + "loss": 2.6922, + "step": 2224 + }, + { + "epoch": 0.2848, + "grad_norm": 0.4892423450946808, + "learning_rate": 0.0008392423994041485, + "loss": 2.7823, + "step": 2225 + }, + { + "epoch": 0.284928, + "grad_norm": 0.46781355142593384, + "learning_rate": 0.000839090076508863, + "loss": 2.6893, + "step": 2226 + }, + { + "epoch": 0.285056, + "grad_norm": 0.5247506499290466, + "learning_rate": 0.000838937695320062, + "loss": 2.6588, + "step": 2227 + }, + { + "epoch": 0.285184, + "grad_norm": 0.9848877787590027, + "learning_rate": 0.0008387852558639421, + "loss": 2.6945, + "step": 2228 + }, + { + "epoch": 0.285312, + "grad_norm": 0.5907170176506042, + "learning_rate": 0.0008386327581667091, + "loss": 2.7281, + "step": 2229 + }, + { + "epoch": 0.28544, + "grad_norm": 0.5427205562591553, + "learning_rate": 0.0008384802022545793, + "loss": 2.5913, + "step": 2230 + }, + { + "epoch": 0.285568, + "grad_norm": 0.5133828520774841, + "learning_rate": 0.0008383275881537786, + "loss": 2.7753, + "step": 2231 + }, + { + "epoch": 0.285696, + "grad_norm": 0.5270614624023438, + "learning_rate": 0.0008381749158905432, + "loss": 2.6496, + "step": 2232 + }, + { + "epoch": 0.285824, + "grad_norm": 0.4909033179283142, + "learning_rate": 0.0008380221854911194, + "loss": 2.7189, + "step": 2233 + }, + { + "epoch": 0.285952, + "grad_norm": 0.548814058303833, + "learning_rate": 0.0008378693969817632, + "loss": 2.5521, + "step": 2234 + }, + { + "epoch": 0.28608, + "grad_norm": 1.0360703468322754, + "learning_rate": 0.0008377165503887409, + "loss": 2.6672, + "step": 2235 + }, + { + "epoch": 0.286208, + "grad_norm": 0.48052120208740234, + "learning_rate": 0.0008375636457383282, + "loss": 2.7783, + "step": 2236 + }, + { + "epoch": 0.286336, + "grad_norm": 0.4967365860939026, + "learning_rate": 0.0008374106830568117, + "loss": 2.6803, + "step": 2237 + }, + { + "epoch": 0.286464, + "grad_norm": 0.5122537016868591, + "learning_rate": 0.000837257662370487, + "loss": 2.6601, + "step": 2238 + }, + { + "epoch": 0.286592, + "grad_norm": 0.5718188285827637, + "learning_rate": 0.0008371045837056604, + "loss": 2.7524, + "step": 2239 + }, + { + "epoch": 0.28672, + "grad_norm": 0.6057481169700623, + "learning_rate": 0.0008369514470886476, + "loss": 2.6687, + "step": 2240 + }, + { + "epoch": 0.286848, + "grad_norm": 0.5871496796607971, + "learning_rate": 0.000836798252545775, + "loss": 2.6731, + "step": 2241 + }, + { + "epoch": 0.286976, + "grad_norm": 0.5098207592964172, + "learning_rate": 0.0008366450001033783, + "loss": 2.7044, + "step": 2242 + }, + { + "epoch": 0.287104, + "grad_norm": 0.5176671743392944, + "learning_rate": 0.0008364916897878032, + "loss": 2.7162, + "step": 2243 + }, + { + "epoch": 0.287232, + "grad_norm": 0.5018636584281921, + "learning_rate": 0.0008363383216254057, + "loss": 2.8327, + "step": 2244 + }, + { + "epoch": 0.28736, + "grad_norm": 0.537283718585968, + "learning_rate": 0.0008361848956425515, + "loss": 2.6656, + "step": 2245 + }, + { + "epoch": 0.287488, + "grad_norm": 0.5077382326126099, + "learning_rate": 0.0008360314118656165, + "loss": 2.6802, + "step": 2246 + }, + { + "epoch": 0.287616, + "grad_norm": 0.4888544976711273, + "learning_rate": 0.0008358778703209861, + "loss": 2.7498, + "step": 2247 + }, + { + "epoch": 0.287744, + "grad_norm": 0.4774535000324249, + "learning_rate": 0.0008357242710350561, + "loss": 2.6677, + "step": 2248 + }, + { + "epoch": 0.287872, + "grad_norm": 0.49129414558410645, + "learning_rate": 0.0008355706140342318, + "loss": 2.6418, + "step": 2249 + }, + { + "epoch": 0.288, + "grad_norm": 0.49008697271347046, + "learning_rate": 0.0008354168993449285, + "loss": 2.6889, + "step": 2250 + }, + { + "epoch": 0.288128, + "grad_norm": 0.49576401710510254, + "learning_rate": 0.0008352631269935718, + "loss": 2.7871, + "step": 2251 + }, + { + "epoch": 0.288256, + "grad_norm": 0.5070188641548157, + "learning_rate": 0.0008351092970065969, + "loss": 2.7461, + "step": 2252 + }, + { + "epoch": 0.288384, + "grad_norm": 0.515635073184967, + "learning_rate": 0.000834955409410449, + "loss": 2.6618, + "step": 2253 + }, + { + "epoch": 0.288512, + "grad_norm": 0.4969954490661621, + "learning_rate": 0.0008348014642315831, + "loss": 2.7613, + "step": 2254 + }, + { + "epoch": 0.28864, + "grad_norm": 0.4822400212287903, + "learning_rate": 0.0008346474614964641, + "loss": 2.8199, + "step": 2255 + }, + { + "epoch": 0.288768, + "grad_norm": 0.4669189155101776, + "learning_rate": 0.0008344934012315668, + "loss": 2.6907, + "step": 2256 + }, + { + "epoch": 0.288896, + "grad_norm": 0.5238752365112305, + "learning_rate": 0.0008343392834633761, + "loss": 2.5927, + "step": 2257 + }, + { + "epoch": 0.289024, + "grad_norm": 0.4554906487464905, + "learning_rate": 0.0008341851082183867, + "loss": 2.76, + "step": 2258 + }, + { + "epoch": 0.289152, + "grad_norm": 0.45607656240463257, + "learning_rate": 0.0008340308755231027, + "loss": 2.7556, + "step": 2259 + }, + { + "epoch": 0.28928, + "grad_norm": 0.4897055923938751, + "learning_rate": 0.0008338765854040392, + "loss": 2.6945, + "step": 2260 + }, + { + "epoch": 0.289408, + "grad_norm": 0.5000702738761902, + "learning_rate": 0.0008337222378877196, + "loss": 2.7574, + "step": 2261 + }, + { + "epoch": 0.289536, + "grad_norm": 0.5078631043434143, + "learning_rate": 0.0008335678330006789, + "loss": 2.6889, + "step": 2262 + }, + { + "epoch": 0.289664, + "grad_norm": 0.5353304147720337, + "learning_rate": 0.0008334133707694604, + "loss": 2.8124, + "step": 2263 + }, + { + "epoch": 0.289792, + "grad_norm": 0.5144590735435486, + "learning_rate": 0.0008332588512206181, + "loss": 2.8661, + "step": 2264 + }, + { + "epoch": 0.28992, + "grad_norm": 0.5031466484069824, + "learning_rate": 0.000833104274380716, + "loss": 2.792, + "step": 2265 + }, + { + "epoch": 0.290048, + "grad_norm": 0.5017592906951904, + "learning_rate": 0.0008329496402763273, + "loss": 2.7305, + "step": 2266 + }, + { + "epoch": 0.290176, + "grad_norm": 0.5044152140617371, + "learning_rate": 0.0008327949489340358, + "loss": 2.7013, + "step": 2267 + }, + { + "epoch": 0.290304, + "grad_norm": 0.5020714402198792, + "learning_rate": 0.0008326402003804343, + "loss": 2.628, + "step": 2268 + }, + { + "epoch": 0.290432, + "grad_norm": 0.5423691868782043, + "learning_rate": 0.000832485394642126, + "loss": 2.6252, + "step": 2269 + }, + { + "epoch": 0.29056, + "grad_norm": 0.5040082335472107, + "learning_rate": 0.000832330531745724, + "loss": 2.6812, + "step": 2270 + }, + { + "epoch": 0.290688, + "grad_norm": 0.5063585042953491, + "learning_rate": 0.000832175611717851, + "loss": 2.7528, + "step": 2271 + }, + { + "epoch": 0.290816, + "grad_norm": 0.5018650889396667, + "learning_rate": 0.0008320206345851393, + "loss": 2.6289, + "step": 2272 + }, + { + "epoch": 0.290944, + "grad_norm": 0.5043584704399109, + "learning_rate": 0.0008318656003742313, + "loss": 2.7862, + "step": 2273 + }, + { + "epoch": 0.291072, + "grad_norm": 0.5444108843803406, + "learning_rate": 0.0008317105091117794, + "loss": 2.7126, + "step": 2274 + }, + { + "epoch": 0.2912, + "grad_norm": 0.4808315932750702, + "learning_rate": 0.0008315553608244453, + "loss": 2.6549, + "step": 2275 + }, + { + "epoch": 0.291328, + "grad_norm": 0.47408202290534973, + "learning_rate": 0.0008314001555389012, + "loss": 2.5767, + "step": 2276 + }, + { + "epoch": 0.291456, + "grad_norm": 0.46221351623535156, + "learning_rate": 0.0008312448932818283, + "loss": 2.7057, + "step": 2277 + }, + { + "epoch": 0.291584, + "grad_norm": 0.4605005383491516, + "learning_rate": 0.0008310895740799181, + "loss": 2.8542, + "step": 2278 + }, + { + "epoch": 0.291712, + "grad_norm": 0.464360773563385, + "learning_rate": 0.0008309341979598719, + "loss": 2.6865, + "step": 2279 + }, + { + "epoch": 0.29184, + "grad_norm": 0.49868881702423096, + "learning_rate": 0.0008307787649484004, + "loss": 2.828, + "step": 2280 + }, + { + "epoch": 0.291968, + "grad_norm": 0.47383245825767517, + "learning_rate": 0.0008306232750722247, + "loss": 2.7489, + "step": 2281 + }, + { + "epoch": 0.292096, + "grad_norm": 0.5117741227149963, + "learning_rate": 0.0008304677283580749, + "loss": 2.7642, + "step": 2282 + }, + { + "epoch": 0.292224, + "grad_norm": 0.532818615436554, + "learning_rate": 0.0008303121248326915, + "loss": 2.6846, + "step": 2283 + }, + { + "epoch": 0.292352, + "grad_norm": 0.5157540440559387, + "learning_rate": 0.0008301564645228248, + "loss": 2.7469, + "step": 2284 + }, + { + "epoch": 0.29248, + "grad_norm": 0.5038409233093262, + "learning_rate": 0.0008300007474552341, + "loss": 2.7216, + "step": 2285 + }, + { + "epoch": 0.292608, + "grad_norm": 0.49100759625434875, + "learning_rate": 0.0008298449736566893, + "loss": 2.6789, + "step": 2286 + }, + { + "epoch": 0.292736, + "grad_norm": 0.5991317629814148, + "learning_rate": 0.0008296891431539695, + "loss": 2.819, + "step": 2287 + }, + { + "epoch": 0.292864, + "grad_norm": 0.501103937625885, + "learning_rate": 0.0008295332559738641, + "loss": 2.7783, + "step": 2288 + }, + { + "epoch": 0.292992, + "grad_norm": 0.6345279216766357, + "learning_rate": 0.0008293773121431716, + "loss": 2.8409, + "step": 2289 + }, + { + "epoch": 0.29312, + "grad_norm": 0.5161519646644592, + "learning_rate": 0.0008292213116887008, + "loss": 2.7095, + "step": 2290 + }, + { + "epoch": 0.293248, + "grad_norm": 0.623309850692749, + "learning_rate": 0.0008290652546372696, + "loss": 2.6848, + "step": 2291 + }, + { + "epoch": 0.293376, + "grad_norm": 0.5313959121704102, + "learning_rate": 0.0008289091410157066, + "loss": 2.8393, + "step": 2292 + }, + { + "epoch": 0.293504, + "grad_norm": 0.4671935439109802, + "learning_rate": 0.000828752970850849, + "loss": 2.5912, + "step": 2293 + }, + { + "epoch": 0.293632, + "grad_norm": 0.607555627822876, + "learning_rate": 0.0008285967441695446, + "loss": 2.7052, + "step": 2294 + }, + { + "epoch": 0.29376, + "grad_norm": 0.5399308800697327, + "learning_rate": 0.0008284404609986504, + "loss": 2.7332, + "step": 2295 + }, + { + "epoch": 0.293888, + "grad_norm": 0.5239366292953491, + "learning_rate": 0.0008282841213650332, + "loss": 2.6831, + "step": 2296 + }, + { + "epoch": 0.294016, + "grad_norm": 0.49240121245384216, + "learning_rate": 0.0008281277252955699, + "loss": 2.7084, + "step": 2297 + }, + { + "epoch": 0.294144, + "grad_norm": 0.5213314890861511, + "learning_rate": 0.0008279712728171466, + "loss": 2.7922, + "step": 2298 + }, + { + "epoch": 0.294272, + "grad_norm": 0.5120644569396973, + "learning_rate": 0.0008278147639566595, + "loss": 2.7762, + "step": 2299 + }, + { + "epoch": 0.2944, + "grad_norm": 0.4986029863357544, + "learning_rate": 0.0008276581987410139, + "loss": 2.7346, + "step": 2300 + }, + { + "epoch": 0.294528, + "grad_norm": 0.5103858113288879, + "learning_rate": 0.0008275015771971255, + "loss": 2.6046, + "step": 2301 + }, + { + "epoch": 0.294656, + "grad_norm": 0.4776620864868164, + "learning_rate": 0.0008273448993519193, + "loss": 2.7077, + "step": 2302 + }, + { + "epoch": 0.294784, + "grad_norm": 0.5250994563102722, + "learning_rate": 0.0008271881652323298, + "loss": 2.7212, + "step": 2303 + }, + { + "epoch": 0.294912, + "grad_norm": 0.488767147064209, + "learning_rate": 0.0008270313748653018, + "loss": 2.6667, + "step": 2304 + }, + { + "epoch": 0.29504, + "grad_norm": 0.5135625004768372, + "learning_rate": 0.0008268745282777891, + "loss": 2.7946, + "step": 2305 + }, + { + "epoch": 0.295168, + "grad_norm": 0.4546082317829132, + "learning_rate": 0.0008267176254967556, + "loss": 2.7468, + "step": 2306 + }, + { + "epoch": 0.295296, + "grad_norm": 0.5520173907279968, + "learning_rate": 0.0008265606665491745, + "loss": 2.6167, + "step": 2307 + }, + { + "epoch": 0.295424, + "grad_norm": 0.49607744812965393, + "learning_rate": 0.0008264036514620293, + "loss": 2.605, + "step": 2308 + }, + { + "epoch": 0.295552, + "grad_norm": 0.5683253407478333, + "learning_rate": 0.0008262465802623121, + "loss": 2.6695, + "step": 2309 + }, + { + "epoch": 0.29568, + "grad_norm": 0.5206356048583984, + "learning_rate": 0.0008260894529770257, + "loss": 2.6277, + "step": 2310 + }, + { + "epoch": 0.295808, + "grad_norm": 0.8860145807266235, + "learning_rate": 0.0008259322696331819, + "loss": 2.639, + "step": 2311 + }, + { + "epoch": 0.295936, + "grad_norm": 0.5042305588722229, + "learning_rate": 0.0008257750302578026, + "loss": 2.8003, + "step": 2312 + }, + { + "epoch": 0.296064, + "grad_norm": 0.5310149788856506, + "learning_rate": 0.0008256177348779188, + "loss": 2.7714, + "step": 2313 + }, + { + "epoch": 0.296192, + "grad_norm": 0.5376003980636597, + "learning_rate": 0.0008254603835205714, + "loss": 2.7219, + "step": 2314 + }, + { + "epoch": 0.29632, + "grad_norm": 0.49583539366722107, + "learning_rate": 0.000825302976212811, + "loss": 2.7705, + "step": 2315 + }, + { + "epoch": 0.296448, + "grad_norm": 0.48704463243484497, + "learning_rate": 0.0008251455129816979, + "loss": 2.7267, + "step": 2316 + }, + { + "epoch": 0.296576, + "grad_norm": 0.5617477893829346, + "learning_rate": 0.0008249879938543017, + "loss": 2.6073, + "step": 2317 + }, + { + "epoch": 0.296704, + "grad_norm": 0.5217834115028381, + "learning_rate": 0.0008248304188577017, + "loss": 2.7325, + "step": 2318 + }, + { + "epoch": 0.296832, + "grad_norm": 0.5688753724098206, + "learning_rate": 0.0008246727880189869, + "loss": 2.7647, + "step": 2319 + }, + { + "epoch": 0.29696, + "grad_norm": 0.5704891085624695, + "learning_rate": 0.0008245151013652561, + "loss": 2.8575, + "step": 2320 + }, + { + "epoch": 0.297088, + "grad_norm": 0.5163342952728271, + "learning_rate": 0.0008243573589236171, + "loss": 2.7649, + "step": 2321 + }, + { + "epoch": 0.297216, + "grad_norm": 0.46955928206443787, + "learning_rate": 0.0008241995607211879, + "loss": 2.6498, + "step": 2322 + }, + { + "epoch": 0.297344, + "grad_norm": 0.5684748291969299, + "learning_rate": 0.0008240417067850957, + "loss": 2.7875, + "step": 2323 + }, + { + "epoch": 0.297472, + "grad_norm": 0.6250737905502319, + "learning_rate": 0.0008238837971424775, + "loss": 2.7467, + "step": 2324 + }, + { + "epoch": 0.2976, + "grad_norm": 0.5321910977363586, + "learning_rate": 0.0008237258318204799, + "loss": 2.8176, + "step": 2325 + }, + { + "epoch": 0.297728, + "grad_norm": 0.5078482031822205, + "learning_rate": 0.0008235678108462587, + "loss": 2.641, + "step": 2326 + }, + { + "epoch": 0.297856, + "grad_norm": 0.9147288799285889, + "learning_rate": 0.0008234097342469798, + "loss": 2.6683, + "step": 2327 + }, + { + "epoch": 0.297984, + "grad_norm": 0.5506005883216858, + "learning_rate": 0.0008232516020498183, + "loss": 2.6579, + "step": 2328 + }, + { + "epoch": 0.298112, + "grad_norm": 0.6050474643707275, + "learning_rate": 0.0008230934142819588, + "loss": 2.7101, + "step": 2329 + }, + { + "epoch": 0.29824, + "grad_norm": 0.5142304301261902, + "learning_rate": 0.000822935170970596, + "loss": 2.6038, + "step": 2330 + }, + { + "epoch": 0.298368, + "grad_norm": 0.5501793622970581, + "learning_rate": 0.0008227768721429334, + "loss": 2.6585, + "step": 2331 + }, + { + "epoch": 0.298496, + "grad_norm": 0.5116293430328369, + "learning_rate": 0.0008226185178261846, + "loss": 2.6975, + "step": 2332 + }, + { + "epoch": 0.298624, + "grad_norm": 0.5132080316543579, + "learning_rate": 0.0008224601080475723, + "loss": 2.715, + "step": 2333 + }, + { + "epoch": 0.298752, + "grad_norm": 0.5260328650474548, + "learning_rate": 0.0008223016428343293, + "loss": 2.7651, + "step": 2334 + }, + { + "epoch": 0.29888, + "grad_norm": 0.5765454769134521, + "learning_rate": 0.0008221431222136975, + "loss": 2.694, + "step": 2335 + }, + { + "epoch": 0.299008, + "grad_norm": 0.4954605996608734, + "learning_rate": 0.0008219845462129283, + "loss": 2.7918, + "step": 2336 + }, + { + "epoch": 0.299136, + "grad_norm": 0.47261354327201843, + "learning_rate": 0.0008218259148592828, + "loss": 2.6093, + "step": 2337 + }, + { + "epoch": 0.299264, + "grad_norm": 0.5914682149887085, + "learning_rate": 0.0008216672281800316, + "loss": 2.7078, + "step": 2338 + }, + { + "epoch": 0.299392, + "grad_norm": 0.4728040397167206, + "learning_rate": 0.0008215084862024549, + "loss": 2.6916, + "step": 2339 + }, + { + "epoch": 0.29952, + "grad_norm": 0.46365886926651, + "learning_rate": 0.0008213496889538421, + "loss": 2.7819, + "step": 2340 + }, + { + "epoch": 0.299648, + "grad_norm": 0.5113892555236816, + "learning_rate": 0.0008211908364614924, + "loss": 2.7315, + "step": 2341 + }, + { + "epoch": 0.299776, + "grad_norm": 0.5816705822944641, + "learning_rate": 0.0008210319287527143, + "loss": 2.6715, + "step": 2342 + }, + { + "epoch": 0.299904, + "grad_norm": 0.5438860654830933, + "learning_rate": 0.000820872965854826, + "loss": 2.7429, + "step": 2343 + }, + { + "epoch": 0.300032, + "grad_norm": 0.49123257398605347, + "learning_rate": 0.0008207139477951549, + "loss": 2.691, + "step": 2344 + }, + { + "epoch": 0.30016, + "grad_norm": 0.49952125549316406, + "learning_rate": 0.0008205548746010382, + "loss": 2.6686, + "step": 2345 + }, + { + "epoch": 0.300288, + "grad_norm": 0.5482579469680786, + "learning_rate": 0.0008203957462998224, + "loss": 2.7593, + "step": 2346 + }, + { + "epoch": 0.300416, + "grad_norm": 0.5390737056732178, + "learning_rate": 0.0008202365629188633, + "loss": 2.7123, + "step": 2347 + }, + { + "epoch": 0.300544, + "grad_norm": 0.5487398505210876, + "learning_rate": 0.0008200773244855267, + "loss": 2.7673, + "step": 2348 + }, + { + "epoch": 0.300672, + "grad_norm": 0.4889863431453705, + "learning_rate": 0.0008199180310271873, + "loss": 2.7316, + "step": 2349 + }, + { + "epoch": 0.3008, + "grad_norm": 0.5008310079574585, + "learning_rate": 0.0008197586825712295, + "loss": 2.7271, + "step": 2350 + }, + { + "epoch": 0.300928, + "grad_norm": 0.4707774519920349, + "learning_rate": 0.0008195992791450474, + "loss": 2.6231, + "step": 2351 + }, + { + "epoch": 0.301056, + "grad_norm": 0.5093923211097717, + "learning_rate": 0.0008194398207760439, + "loss": 2.6353, + "step": 2352 + }, + { + "epoch": 0.301184, + "grad_norm": 0.5727475881576538, + "learning_rate": 0.0008192803074916321, + "loss": 2.7785, + "step": 2353 + }, + { + "epoch": 0.301312, + "grad_norm": 0.5063148736953735, + "learning_rate": 0.0008191207393192339, + "loss": 2.7974, + "step": 2354 + }, + { + "epoch": 0.30144, + "grad_norm": 0.5637764930725098, + "learning_rate": 0.0008189611162862812, + "loss": 2.7907, + "step": 2355 + }, + { + "epoch": 0.301568, + "grad_norm": 0.5030006170272827, + "learning_rate": 0.0008188014384202147, + "loss": 2.6512, + "step": 2356 + }, + { + "epoch": 0.301696, + "grad_norm": 0.5257261991500854, + "learning_rate": 0.0008186417057484852, + "loss": 2.6588, + "step": 2357 + }, + { + "epoch": 0.301824, + "grad_norm": 0.5035802721977234, + "learning_rate": 0.0008184819182985524, + "loss": 2.7026, + "step": 2358 + }, + { + "epoch": 0.301952, + "grad_norm": 0.46964186429977417, + "learning_rate": 0.0008183220760978858, + "loss": 2.6363, + "step": 2359 + }, + { + "epoch": 0.30208, + "grad_norm": 0.5465080738067627, + "learning_rate": 0.000818162179173964, + "loss": 2.792, + "step": 2360 + }, + { + "epoch": 0.302208, + "grad_norm": 0.6074954867362976, + "learning_rate": 0.0008180022275542749, + "loss": 2.835, + "step": 2361 + }, + { + "epoch": 0.302336, + "grad_norm": 0.5213726758956909, + "learning_rate": 0.0008178422212663166, + "loss": 2.6654, + "step": 2362 + }, + { + "epoch": 0.302464, + "grad_norm": 0.4711441099643707, + "learning_rate": 0.0008176821603375954, + "loss": 2.7109, + "step": 2363 + }, + { + "epoch": 0.302592, + "grad_norm": 0.5097600221633911, + "learning_rate": 0.0008175220447956281, + "loss": 2.7086, + "step": 2364 + }, + { + "epoch": 0.30272, + "grad_norm": 0.524763822555542, + "learning_rate": 0.0008173618746679406, + "loss": 2.7419, + "step": 2365 + }, + { + "epoch": 0.302848, + "grad_norm": 0.48518967628479004, + "learning_rate": 0.0008172016499820671, + "loss": 2.8077, + "step": 2366 + }, + { + "epoch": 0.302976, + "grad_norm": 0.5653429627418518, + "learning_rate": 0.000817041370765553, + "loss": 2.7644, + "step": 2367 + }, + { + "epoch": 0.303104, + "grad_norm": 0.5319207310676575, + "learning_rate": 0.0008168810370459518, + "loss": 2.6889, + "step": 2368 + }, + { + "epoch": 0.303232, + "grad_norm": 0.5532823801040649, + "learning_rate": 0.0008167206488508267, + "loss": 2.76, + "step": 2369 + }, + { + "epoch": 0.30336, + "grad_norm": 0.49760231375694275, + "learning_rate": 0.0008165602062077502, + "loss": 2.7026, + "step": 2370 + }, + { + "epoch": 0.303488, + "grad_norm": 0.5163533091545105, + "learning_rate": 0.0008163997091443045, + "loss": 2.588, + "step": 2371 + }, + { + "epoch": 0.303616, + "grad_norm": 0.5411005020141602, + "learning_rate": 0.0008162391576880808, + "loss": 2.7847, + "step": 2372 + }, + { + "epoch": 0.303744, + "grad_norm": 0.5488636493682861, + "learning_rate": 0.0008160785518666796, + "loss": 2.6618, + "step": 2373 + }, + { + "epoch": 0.303872, + "grad_norm": 0.5334295630455017, + "learning_rate": 0.0008159178917077111, + "loss": 2.7489, + "step": 2374 + }, + { + "epoch": 0.304, + "grad_norm": 0.4845445454120636, + "learning_rate": 0.0008157571772387946, + "loss": 2.6326, + "step": 2375 + }, + { + "epoch": 0.304128, + "grad_norm": 0.46909546852111816, + "learning_rate": 0.0008155964084875587, + "loss": 2.7344, + "step": 2376 + }, + { + "epoch": 0.304256, + "grad_norm": 0.5020903944969177, + "learning_rate": 0.0008154355854816416, + "loss": 2.7995, + "step": 2377 + }, + { + "epoch": 0.304384, + "grad_norm": 0.540817141532898, + "learning_rate": 0.0008152747082486904, + "loss": 2.7706, + "step": 2378 + }, + { + "epoch": 0.304512, + "grad_norm": 0.5027971267700195, + "learning_rate": 0.0008151137768163619, + "loss": 2.7099, + "step": 2379 + }, + { + "epoch": 0.30464, + "grad_norm": 0.5109881162643433, + "learning_rate": 0.000814952791212322, + "loss": 2.7361, + "step": 2380 + }, + { + "epoch": 0.304768, + "grad_norm": 0.5136557817459106, + "learning_rate": 0.0008147917514642462, + "loss": 2.6424, + "step": 2381 + }, + { + "epoch": 0.304896, + "grad_norm": 0.48797905445098877, + "learning_rate": 0.0008146306575998188, + "loss": 2.7603, + "step": 2382 + }, + { + "epoch": 0.305024, + "grad_norm": 0.49893662333488464, + "learning_rate": 0.0008144695096467341, + "loss": 2.7661, + "step": 2383 + }, + { + "epoch": 0.305152, + "grad_norm": 0.680966854095459, + "learning_rate": 0.0008143083076326947, + "loss": 2.7585, + "step": 2384 + }, + { + "epoch": 0.30528, + "grad_norm": 0.4788847267627716, + "learning_rate": 0.0008141470515854135, + "loss": 2.6077, + "step": 2385 + }, + { + "epoch": 0.305408, + "grad_norm": 0.5391897559165955, + "learning_rate": 0.0008139857415326124, + "loss": 2.734, + "step": 2386 + }, + { + "epoch": 0.305536, + "grad_norm": 0.49233973026275635, + "learning_rate": 0.0008138243775020223, + "loss": 2.7004, + "step": 2387 + }, + { + "epoch": 0.305664, + "grad_norm": 0.5187211036682129, + "learning_rate": 0.0008136629595213834, + "loss": 2.6304, + "step": 2388 + }, + { + "epoch": 0.305792, + "grad_norm": 0.4985654652118683, + "learning_rate": 0.0008135014876184454, + "loss": 2.7275, + "step": 2389 + }, + { + "epoch": 0.30592, + "grad_norm": 0.5419280529022217, + "learning_rate": 0.0008133399618209674, + "loss": 2.7457, + "step": 2390 + }, + { + "epoch": 0.306048, + "grad_norm": 0.6023381352424622, + "learning_rate": 0.0008131783821567174, + "loss": 2.6814, + "step": 2391 + }, + { + "epoch": 0.306176, + "grad_norm": 0.4937226176261902, + "learning_rate": 0.0008130167486534729, + "loss": 2.7144, + "step": 2392 + }, + { + "epoch": 0.306304, + "grad_norm": 0.5069573521614075, + "learning_rate": 0.0008128550613390205, + "loss": 2.7599, + "step": 2393 + }, + { + "epoch": 0.306432, + "grad_norm": 0.5452480316162109, + "learning_rate": 0.0008126933202411559, + "loss": 2.6819, + "step": 2394 + }, + { + "epoch": 0.30656, + "grad_norm": 0.6676955819129944, + "learning_rate": 0.0008125315253876849, + "loss": 2.6828, + "step": 2395 + }, + { + "epoch": 0.306688, + "grad_norm": 0.4887922704219818, + "learning_rate": 0.0008123696768064212, + "loss": 2.6373, + "step": 2396 + }, + { + "epoch": 0.306816, + "grad_norm": 0.5046238899230957, + "learning_rate": 0.0008122077745251888, + "loss": 2.6773, + "step": 2397 + }, + { + "epoch": 0.306944, + "grad_norm": 0.5116103291511536, + "learning_rate": 0.0008120458185718205, + "loss": 2.7136, + "step": 2398 + }, + { + "epoch": 0.307072, + "grad_norm": 0.5339289307594299, + "learning_rate": 0.0008118838089741585, + "loss": 2.581, + "step": 2399 + }, + { + "epoch": 0.3072, + "grad_norm": 0.5624111294746399, + "learning_rate": 0.0008117217457600541, + "loss": 2.6805, + "step": 2400 + }, + { + "epoch": 0.307328, + "grad_norm": 0.47977617383003235, + "learning_rate": 0.0008115596289573678, + "loss": 2.7334, + "step": 2401 + }, + { + "epoch": 0.307456, + "grad_norm": 0.4993445575237274, + "learning_rate": 0.0008113974585939694, + "loss": 2.7309, + "step": 2402 + }, + { + "epoch": 0.307584, + "grad_norm": 0.687139093875885, + "learning_rate": 0.0008112352346977377, + "loss": 2.7061, + "step": 2403 + }, + { + "epoch": 0.307712, + "grad_norm": 0.4758971035480499, + "learning_rate": 0.0008110729572965612, + "loss": 2.6866, + "step": 2404 + }, + { + "epoch": 0.30784, + "grad_norm": 0.49293866753578186, + "learning_rate": 0.0008109106264183369, + "loss": 2.7218, + "step": 2405 + }, + { + "epoch": 0.307968, + "grad_norm": 0.5503746271133423, + "learning_rate": 0.0008107482420909718, + "loss": 2.651, + "step": 2406 + }, + { + "epoch": 0.308096, + "grad_norm": 0.5575663447380066, + "learning_rate": 0.0008105858043423811, + "loss": 2.6935, + "step": 2407 + }, + { + "epoch": 0.308224, + "grad_norm": 0.49155205488204956, + "learning_rate": 0.0008104233132004901, + "loss": 2.803, + "step": 2408 + }, + { + "epoch": 0.308352, + "grad_norm": 0.5458154082298279, + "learning_rate": 0.0008102607686932328, + "loss": 2.7432, + "step": 2409 + }, + { + "epoch": 0.30848, + "grad_norm": 0.4682959020137787, + "learning_rate": 0.0008100981708485526, + "loss": 2.587, + "step": 2410 + }, + { + "epoch": 0.308608, + "grad_norm": 0.7160769701004028, + "learning_rate": 0.0008099355196944018, + "loss": 2.6488, + "step": 2411 + }, + { + "epoch": 0.308736, + "grad_norm": 0.6005364060401917, + "learning_rate": 0.000809772815258742, + "loss": 2.7945, + "step": 2412 + }, + { + "epoch": 0.308864, + "grad_norm": 0.4972257614135742, + "learning_rate": 0.0008096100575695442, + "loss": 2.7045, + "step": 2413 + }, + { + "epoch": 0.308992, + "grad_norm": 0.5468704700469971, + "learning_rate": 0.0008094472466547882, + "loss": 2.8125, + "step": 2414 + }, + { + "epoch": 0.30912, + "grad_norm": 0.5141345262527466, + "learning_rate": 0.0008092843825424629, + "loss": 2.7469, + "step": 2415 + }, + { + "epoch": 0.309248, + "grad_norm": 0.5765616297721863, + "learning_rate": 0.0008091214652605669, + "loss": 2.794, + "step": 2416 + }, + { + "epoch": 0.309376, + "grad_norm": 0.4803778827190399, + "learning_rate": 0.0008089584948371074, + "loss": 2.7838, + "step": 2417 + }, + { + "epoch": 0.309504, + "grad_norm": 0.4671151041984558, + "learning_rate": 0.0008087954713001007, + "loss": 2.6792, + "step": 2418 + }, + { + "epoch": 0.309632, + "grad_norm": 0.7678715586662292, + "learning_rate": 0.0008086323946775727, + "loss": 2.6948, + "step": 2419 + }, + { + "epoch": 0.30976, + "grad_norm": 0.5633267760276794, + "learning_rate": 0.0008084692649975582, + "loss": 2.5718, + "step": 2420 + }, + { + "epoch": 0.309888, + "grad_norm": 0.4930620491504669, + "learning_rate": 0.0008083060822881008, + "loss": 2.7499, + "step": 2421 + }, + { + "epoch": 0.310016, + "grad_norm": 0.48592904210090637, + "learning_rate": 0.0008081428465772538, + "loss": 2.7751, + "step": 2422 + }, + { + "epoch": 0.310144, + "grad_norm": 0.6134775280952454, + "learning_rate": 0.0008079795578930791, + "loss": 2.6161, + "step": 2423 + }, + { + "epoch": 0.310272, + "grad_norm": 0.540285050868988, + "learning_rate": 0.000807816216263648, + "loss": 2.7438, + "step": 2424 + }, + { + "epoch": 0.3104, + "grad_norm": 0.5360494256019592, + "learning_rate": 0.0008076528217170408, + "loss": 2.6855, + "step": 2425 + }, + { + "epoch": 0.310528, + "grad_norm": 0.5166777968406677, + "learning_rate": 0.0008074893742813469, + "loss": 2.7405, + "step": 2426 + }, + { + "epoch": 0.310656, + "grad_norm": 0.522057294845581, + "learning_rate": 0.0008073258739846649, + "loss": 2.6758, + "step": 2427 + }, + { + "epoch": 0.310784, + "grad_norm": 0.5074620246887207, + "learning_rate": 0.0008071623208551023, + "loss": 2.6549, + "step": 2428 + }, + { + "epoch": 0.310912, + "grad_norm": 0.4936791956424713, + "learning_rate": 0.0008069987149207759, + "loss": 2.7257, + "step": 2429 + }, + { + "epoch": 0.31104, + "grad_norm": 0.45561349391937256, + "learning_rate": 0.0008068350562098112, + "loss": 2.6814, + "step": 2430 + }, + { + "epoch": 0.311168, + "grad_norm": 0.5450307726860046, + "learning_rate": 0.0008066713447503433, + "loss": 2.6994, + "step": 2431 + }, + { + "epoch": 0.311296, + "grad_norm": 0.4999459683895111, + "learning_rate": 0.000806507580570516, + "loss": 2.7397, + "step": 2432 + }, + { + "epoch": 0.311424, + "grad_norm": 0.469186395406723, + "learning_rate": 0.0008063437636984823, + "loss": 2.6317, + "step": 2433 + }, + { + "epoch": 0.311552, + "grad_norm": 0.517356276512146, + "learning_rate": 0.0008061798941624041, + "loss": 2.6685, + "step": 2434 + }, + { + "epoch": 0.31168, + "grad_norm": 0.7971348166465759, + "learning_rate": 0.0008060159719904525, + "loss": 2.5808, + "step": 2435 + }, + { + "epoch": 0.311808, + "grad_norm": 0.5929805636405945, + "learning_rate": 0.0008058519972108079, + "loss": 2.7412, + "step": 2436 + }, + { + "epoch": 0.311936, + "grad_norm": 0.54878830909729, + "learning_rate": 0.0008056879698516591, + "loss": 2.7162, + "step": 2437 + }, + { + "epoch": 0.312064, + "grad_norm": 0.5049181580543518, + "learning_rate": 0.0008055238899412045, + "loss": 2.6745, + "step": 2438 + }, + { + "epoch": 0.312192, + "grad_norm": 0.7295874357223511, + "learning_rate": 0.0008053597575076512, + "loss": 2.6823, + "step": 2439 + }, + { + "epoch": 0.31232, + "grad_norm": 0.4962950348854065, + "learning_rate": 0.0008051955725792155, + "loss": 2.6849, + "step": 2440 + }, + { + "epoch": 0.312448, + "grad_norm": 0.5048396587371826, + "learning_rate": 0.0008050313351841228, + "loss": 2.7588, + "step": 2441 + }, + { + "epoch": 0.312576, + "grad_norm": 0.5057218670845032, + "learning_rate": 0.0008048670453506074, + "loss": 2.7, + "step": 2442 + }, + { + "epoch": 0.312704, + "grad_norm": 0.5434165000915527, + "learning_rate": 0.0008047027031069126, + "loss": 2.6141, + "step": 2443 + }, + { + "epoch": 0.312832, + "grad_norm": 0.4824839234352112, + "learning_rate": 0.0008045383084812906, + "loss": 2.785, + "step": 2444 + }, + { + "epoch": 0.31296, + "grad_norm": 0.5279079675674438, + "learning_rate": 0.0008043738615020029, + "loss": 2.6811, + "step": 2445 + }, + { + "epoch": 0.313088, + "grad_norm": 0.5532738566398621, + "learning_rate": 0.0008042093621973198, + "loss": 2.716, + "step": 2446 + }, + { + "epoch": 0.313216, + "grad_norm": 0.5089097619056702, + "learning_rate": 0.0008040448105955208, + "loss": 2.7685, + "step": 2447 + }, + { + "epoch": 0.313344, + "grad_norm": 0.4816966652870178, + "learning_rate": 0.0008038802067248939, + "loss": 2.7074, + "step": 2448 + }, + { + "epoch": 0.313472, + "grad_norm": 0.4838141202926636, + "learning_rate": 0.0008037155506137366, + "loss": 2.5601, + "step": 2449 + }, + { + "epoch": 0.3136, + "grad_norm": 0.5404328107833862, + "learning_rate": 0.0008035508422903553, + "loss": 2.8101, + "step": 2450 + }, + { + "epoch": 0.313728, + "grad_norm": 0.5346477627754211, + "learning_rate": 0.000803386081783065, + "loss": 2.6905, + "step": 2451 + }, + { + "epoch": 0.313856, + "grad_norm": 0.530609667301178, + "learning_rate": 0.0008032212691201901, + "loss": 2.6495, + "step": 2452 + }, + { + "epoch": 0.313984, + "grad_norm": 0.5314395427703857, + "learning_rate": 0.0008030564043300639, + "loss": 2.7879, + "step": 2453 + }, + { + "epoch": 0.314112, + "grad_norm": 0.6026572585105896, + "learning_rate": 0.0008028914874410283, + "loss": 2.7266, + "step": 2454 + }, + { + "epoch": 0.31424, + "grad_norm": 0.49589499831199646, + "learning_rate": 0.0008027265184814348, + "loss": 2.7033, + "step": 2455 + }, + { + "epoch": 0.314368, + "grad_norm": 0.5314655303955078, + "learning_rate": 0.0008025614974796431, + "loss": 2.7357, + "step": 2456 + }, + { + "epoch": 0.314496, + "grad_norm": 0.5936437249183655, + "learning_rate": 0.0008023964244640224, + "loss": 2.5885, + "step": 2457 + }, + { + "epoch": 0.314624, + "grad_norm": 0.545367956161499, + "learning_rate": 0.0008022312994629508, + "loss": 2.6301, + "step": 2458 + }, + { + "epoch": 0.314752, + "grad_norm": 0.5255335569381714, + "learning_rate": 0.0008020661225048149, + "loss": 2.8378, + "step": 2459 + }, + { + "epoch": 0.31488, + "grad_norm": 0.5140583515167236, + "learning_rate": 0.0008019008936180108, + "loss": 2.7351, + "step": 2460 + }, + { + "epoch": 0.315008, + "grad_norm": 0.5122703909873962, + "learning_rate": 0.0008017356128309431, + "loss": 2.6829, + "step": 2461 + }, + { + "epoch": 0.315136, + "grad_norm": 0.5080159306526184, + "learning_rate": 0.0008015702801720255, + "loss": 2.7177, + "step": 2462 + }, + { + "epoch": 0.315264, + "grad_norm": 0.5291783809661865, + "learning_rate": 0.0008014048956696807, + "loss": 2.6141, + "step": 2463 + }, + { + "epoch": 0.315392, + "grad_norm": 0.49185362458229065, + "learning_rate": 0.0008012394593523403, + "loss": 2.6964, + "step": 2464 + }, + { + "epoch": 0.31552, + "grad_norm": 0.5045166015625, + "learning_rate": 0.0008010739712484443, + "loss": 2.7504, + "step": 2465 + }, + { + "epoch": 0.315648, + "grad_norm": 0.4936505854129791, + "learning_rate": 0.0008009084313864423, + "loss": 2.6691, + "step": 2466 + }, + { + "epoch": 0.315776, + "grad_norm": 0.4659939408302307, + "learning_rate": 0.0008007428397947925, + "loss": 2.7129, + "step": 2467 + }, + { + "epoch": 0.315904, + "grad_norm": 0.5057143568992615, + "learning_rate": 0.0008005771965019622, + "loss": 2.6181, + "step": 2468 + }, + { + "epoch": 0.316032, + "grad_norm": 0.6470767259597778, + "learning_rate": 0.0008004115015364272, + "loss": 2.8044, + "step": 2469 + }, + { + "epoch": 0.31616, + "grad_norm": 0.5448716282844543, + "learning_rate": 0.0008002457549266724, + "loss": 2.7101, + "step": 2470 + }, + { + "epoch": 0.316288, + "grad_norm": 0.5227517485618591, + "learning_rate": 0.0008000799567011914, + "loss": 2.7095, + "step": 2471 + }, + { + "epoch": 0.316416, + "grad_norm": 0.6266842484474182, + "learning_rate": 0.0007999141068884872, + "loss": 2.7027, + "step": 2472 + }, + { + "epoch": 0.316544, + "grad_norm": 0.49957144260406494, + "learning_rate": 0.0007997482055170712, + "loss": 2.7203, + "step": 2473 + }, + { + "epoch": 0.316672, + "grad_norm": 0.7435759902000427, + "learning_rate": 0.0007995822526154636, + "loss": 2.6645, + "step": 2474 + }, + { + "epoch": 0.3168, + "grad_norm": 0.5999026298522949, + "learning_rate": 0.0007994162482121936, + "loss": 2.7535, + "step": 2475 + }, + { + "epoch": 0.316928, + "grad_norm": 0.4780690371990204, + "learning_rate": 0.0007992501923357997, + "loss": 2.5626, + "step": 2476 + }, + { + "epoch": 0.317056, + "grad_norm": 0.5613746643066406, + "learning_rate": 0.0007990840850148282, + "loss": 2.6589, + "step": 2477 + }, + { + "epoch": 0.317184, + "grad_norm": 0.4944455921649933, + "learning_rate": 0.0007989179262778355, + "loss": 2.603, + "step": 2478 + }, + { + "epoch": 0.317312, + "grad_norm": 0.4796963334083557, + "learning_rate": 0.0007987517161533857, + "loss": 2.5784, + "step": 2479 + }, + { + "epoch": 0.31744, + "grad_norm": 0.5259159803390503, + "learning_rate": 0.0007985854546700525, + "loss": 2.7514, + "step": 2480 + }, + { + "epoch": 0.317568, + "grad_norm": 0.5473962426185608, + "learning_rate": 0.0007984191418564182, + "loss": 2.683, + "step": 2481 + }, + { + "epoch": 0.317696, + "grad_norm": 0.5132822394371033, + "learning_rate": 0.0007982527777410737, + "loss": 2.7057, + "step": 2482 + }, + { + "epoch": 0.317824, + "grad_norm": 0.5078954696655273, + "learning_rate": 0.0007980863623526193, + "loss": 2.8688, + "step": 2483 + }, + { + "epoch": 0.317952, + "grad_norm": 0.5008555054664612, + "learning_rate": 0.0007979198957196634, + "loss": 2.6299, + "step": 2484 + }, + { + "epoch": 0.31808, + "grad_norm": 0.4791910648345947, + "learning_rate": 0.0007977533778708236, + "loss": 2.6953, + "step": 2485 + }, + { + "epoch": 0.318208, + "grad_norm": 0.5985043048858643, + "learning_rate": 0.0007975868088347263, + "loss": 2.7153, + "step": 2486 + }, + { + "epoch": 0.318336, + "grad_norm": 0.46569350361824036, + "learning_rate": 0.0007974201886400066, + "loss": 2.7577, + "step": 2487 + }, + { + "epoch": 0.318464, + "grad_norm": 0.5320923924446106, + "learning_rate": 0.0007972535173153086, + "loss": 2.6762, + "step": 2488 + }, + { + "epoch": 0.318592, + "grad_norm": 0.5294308662414551, + "learning_rate": 0.0007970867948892848, + "loss": 2.7517, + "step": 2489 + }, + { + "epoch": 0.31872, + "grad_norm": 0.517565131187439, + "learning_rate": 0.000796920021390597, + "loss": 2.5613, + "step": 2490 + }, + { + "epoch": 0.318848, + "grad_norm": 0.5872734785079956, + "learning_rate": 0.0007967531968479151, + "loss": 2.772, + "step": 2491 + }, + { + "epoch": 0.318976, + "grad_norm": 0.5379122495651245, + "learning_rate": 0.0007965863212899185, + "loss": 2.787, + "step": 2492 + }, + { + "epoch": 0.319104, + "grad_norm": 0.5204347968101501, + "learning_rate": 0.0007964193947452949, + "loss": 2.6975, + "step": 2493 + }, + { + "epoch": 0.319232, + "grad_norm": 0.48446860909461975, + "learning_rate": 0.0007962524172427411, + "loss": 2.6886, + "step": 2494 + }, + { + "epoch": 0.31936, + "grad_norm": 0.6012802124023438, + "learning_rate": 0.0007960853888109622, + "loss": 2.6159, + "step": 2495 + }, + { + "epoch": 0.319488, + "grad_norm": 0.802354633808136, + "learning_rate": 0.0007959183094786725, + "loss": 2.7728, + "step": 2496 + }, + { + "epoch": 0.319616, + "grad_norm": 0.49367815256118774, + "learning_rate": 0.0007957511792745949, + "loss": 2.6284, + "step": 2497 + }, + { + "epoch": 0.319744, + "grad_norm": 0.5232271552085876, + "learning_rate": 0.0007955839982274607, + "loss": 2.6814, + "step": 2498 + }, + { + "epoch": 0.319872, + "grad_norm": 0.54829341173172, + "learning_rate": 0.0007954167663660108, + "loss": 2.6264, + "step": 2499 + }, + { + "epoch": 0.32, + "grad_norm": 0.5162155628204346, + "learning_rate": 0.0007952494837189937, + "loss": 2.7409, + "step": 2500 + }, + { + "epoch": 0.320128, + "grad_norm": 0.4739733338356018, + "learning_rate": 0.0007950821503151678, + "loss": 2.7708, + "step": 2501 + }, + { + "epoch": 0.320256, + "grad_norm": 0.5036630630493164, + "learning_rate": 0.0007949147661832992, + "loss": 2.7749, + "step": 2502 + }, + { + "epoch": 0.320384, + "grad_norm": 0.4932006299495697, + "learning_rate": 0.0007947473313521634, + "loss": 2.8594, + "step": 2503 + }, + { + "epoch": 0.320512, + "grad_norm": 0.5399882793426514, + "learning_rate": 0.0007945798458505442, + "loss": 2.7124, + "step": 2504 + }, + { + "epoch": 0.32064, + "grad_norm": 0.48951268196105957, + "learning_rate": 0.0007944123097072347, + "loss": 2.7879, + "step": 2505 + }, + { + "epoch": 0.320768, + "grad_norm": 0.4845176041126251, + "learning_rate": 0.0007942447229510359, + "loss": 2.6927, + "step": 2506 + }, + { + "epoch": 0.320896, + "grad_norm": 0.4523446559906006, + "learning_rate": 0.000794077085610758, + "loss": 2.7689, + "step": 2507 + }, + { + "epoch": 0.321024, + "grad_norm": 0.47698935866355896, + "learning_rate": 0.00079390939771522, + "loss": 2.6594, + "step": 2508 + }, + { + "epoch": 0.321152, + "grad_norm": 0.4626786410808563, + "learning_rate": 0.000793741659293249, + "loss": 2.6505, + "step": 2509 + }, + { + "epoch": 0.32128, + "grad_norm": 0.5631040930747986, + "learning_rate": 0.0007935738703736817, + "loss": 2.6689, + "step": 2510 + }, + { + "epoch": 0.321408, + "grad_norm": 0.5421676635742188, + "learning_rate": 0.0007934060309853626, + "loss": 2.7051, + "step": 2511 + }, + { + "epoch": 0.321536, + "grad_norm": 0.49667859077453613, + "learning_rate": 0.0007932381411571453, + "loss": 2.7062, + "step": 2512 + }, + { + "epoch": 0.321664, + "grad_norm": 0.5250574350357056, + "learning_rate": 0.0007930702009178918, + "loss": 2.6615, + "step": 2513 + }, + { + "epoch": 0.321792, + "grad_norm": 0.5898976922035217, + "learning_rate": 0.0007929022102964734, + "loss": 2.6715, + "step": 2514 + }, + { + "epoch": 0.32192, + "grad_norm": 0.5876429677009583, + "learning_rate": 0.0007927341693217695, + "loss": 2.7376, + "step": 2515 + }, + { + "epoch": 0.322048, + "grad_norm": 0.48137733340263367, + "learning_rate": 0.0007925660780226683, + "loss": 2.7268, + "step": 2516 + }, + { + "epoch": 0.322176, + "grad_norm": 0.5001925826072693, + "learning_rate": 0.0007923979364280663, + "loss": 2.678, + "step": 2517 + }, + { + "epoch": 0.322304, + "grad_norm": 0.4613078236579895, + "learning_rate": 0.0007922297445668694, + "loss": 2.6443, + "step": 2518 + }, + { + "epoch": 0.322432, + "grad_norm": 0.5213683247566223, + "learning_rate": 0.0007920615024679916, + "loss": 2.7095, + "step": 2519 + }, + { + "epoch": 0.32256, + "grad_norm": 0.44228264689445496, + "learning_rate": 0.0007918932101603555, + "loss": 2.7282, + "step": 2520 + }, + { + "epoch": 0.322688, + "grad_norm": 0.5399432182312012, + "learning_rate": 0.0007917248676728929, + "loss": 2.635, + "step": 2521 + }, + { + "epoch": 0.322816, + "grad_norm": 0.6094278693199158, + "learning_rate": 0.0007915564750345435, + "loss": 2.7921, + "step": 2522 + }, + { + "epoch": 0.322944, + "grad_norm": 0.5689902901649475, + "learning_rate": 0.0007913880322742559, + "loss": 2.7075, + "step": 2523 + }, + { + "epoch": 0.323072, + "grad_norm": 0.5367972254753113, + "learning_rate": 0.0007912195394209877, + "loss": 2.592, + "step": 2524 + }, + { + "epoch": 0.3232, + "grad_norm": 0.512249767780304, + "learning_rate": 0.0007910509965037045, + "loss": 2.6985, + "step": 2525 + }, + { + "epoch": 0.323328, + "grad_norm": 0.4707354009151459, + "learning_rate": 0.0007908824035513808, + "loss": 2.7005, + "step": 2526 + }, + { + "epoch": 0.323456, + "grad_norm": 0.46003326773643494, + "learning_rate": 0.0007907137605929997, + "loss": 2.6574, + "step": 2527 + }, + { + "epoch": 0.323584, + "grad_norm": 0.4980950951576233, + "learning_rate": 0.000790545067657553, + "loss": 2.6543, + "step": 2528 + }, + { + "epoch": 0.323712, + "grad_norm": 0.4777020514011383, + "learning_rate": 0.0007903763247740407, + "loss": 2.7028, + "step": 2529 + }, + { + "epoch": 0.32384, + "grad_norm": 0.4793339669704437, + "learning_rate": 0.0007902075319714721, + "loss": 2.718, + "step": 2530 + }, + { + "epoch": 0.323968, + "grad_norm": 0.4862494170665741, + "learning_rate": 0.0007900386892788642, + "loss": 2.7776, + "step": 2531 + }, + { + "epoch": 0.324096, + "grad_norm": 0.4915720224380493, + "learning_rate": 0.0007898697967252433, + "loss": 2.7153, + "step": 2532 + }, + { + "epoch": 0.324224, + "grad_norm": 0.4848230183124542, + "learning_rate": 0.0007897008543396438, + "loss": 2.7283, + "step": 2533 + }, + { + "epoch": 0.324352, + "grad_norm": 0.46038520336151123, + "learning_rate": 0.000789531862151109, + "loss": 2.6792, + "step": 2534 + }, + { + "epoch": 0.32448, + "grad_norm": 0.46778157353401184, + "learning_rate": 0.0007893628201886907, + "loss": 2.7309, + "step": 2535 + }, + { + "epoch": 0.324608, + "grad_norm": 0.6603022217750549, + "learning_rate": 0.0007891937284814488, + "loss": 2.6398, + "step": 2536 + }, + { + "epoch": 0.324736, + "grad_norm": 0.44336551427841187, + "learning_rate": 0.0007890245870584522, + "loss": 2.6586, + "step": 2537 + }, + { + "epoch": 0.324864, + "grad_norm": 0.5862497687339783, + "learning_rate": 0.0007888553959487787, + "loss": 2.7698, + "step": 2538 + }, + { + "epoch": 0.324992, + "grad_norm": 0.49440208077430725, + "learning_rate": 0.0007886861551815138, + "loss": 2.6335, + "step": 2539 + }, + { + "epoch": 0.32512, + "grad_norm": 0.48265665769577026, + "learning_rate": 0.0007885168647857524, + "loss": 2.6881, + "step": 2540 + }, + { + "epoch": 0.325248, + "grad_norm": 0.5119762420654297, + "learning_rate": 0.0007883475247905969, + "loss": 2.7068, + "step": 2541 + }, + { + "epoch": 0.325376, + "grad_norm": 0.5339782238006592, + "learning_rate": 0.0007881781352251591, + "loss": 2.7109, + "step": 2542 + }, + { + "epoch": 0.325504, + "grad_norm": 0.5108572840690613, + "learning_rate": 0.0007880086961185589, + "loss": 2.7859, + "step": 2543 + }, + { + "epoch": 0.325632, + "grad_norm": 0.4533543884754181, + "learning_rate": 0.0007878392074999253, + "loss": 2.7016, + "step": 2544 + }, + { + "epoch": 0.32576, + "grad_norm": 0.5028905868530273, + "learning_rate": 0.0007876696693983946, + "loss": 2.7811, + "step": 2545 + }, + { + "epoch": 0.325888, + "grad_norm": 0.45935091376304626, + "learning_rate": 0.000787500081843113, + "loss": 2.6421, + "step": 2546 + }, + { + "epoch": 0.326016, + "grad_norm": 0.42180514335632324, + "learning_rate": 0.0007873304448632344, + "loss": 2.6188, + "step": 2547 + }, + { + "epoch": 0.326144, + "grad_norm": 0.4369472563266754, + "learning_rate": 0.0007871607584879211, + "loss": 2.6881, + "step": 2548 + }, + { + "epoch": 0.326272, + "grad_norm": 0.4790958762168884, + "learning_rate": 0.0007869910227463446, + "loss": 2.6948, + "step": 2549 + }, + { + "epoch": 0.3264, + "grad_norm": 0.467967689037323, + "learning_rate": 0.000786821237667684, + "loss": 2.7671, + "step": 2550 + }, + { + "epoch": 0.326528, + "grad_norm": 0.44105201959609985, + "learning_rate": 0.0007866514032811276, + "loss": 2.6311, + "step": 2551 + }, + { + "epoch": 0.326656, + "grad_norm": 0.4860559403896332, + "learning_rate": 0.0007864815196158718, + "loss": 2.68, + "step": 2552 + }, + { + "epoch": 0.326784, + "grad_norm": 0.44505929946899414, + "learning_rate": 0.0007863115867011216, + "loss": 2.6067, + "step": 2553 + }, + { + "epoch": 0.326912, + "grad_norm": 0.4458320438861847, + "learning_rate": 0.0007861416045660905, + "loss": 2.7622, + "step": 2554 + }, + { + "epoch": 0.32704, + "grad_norm": 0.44874417781829834, + "learning_rate": 0.0007859715732400004, + "loss": 2.6405, + "step": 2555 + }, + { + "epoch": 0.327168, + "grad_norm": 0.43373560905456543, + "learning_rate": 0.0007858014927520814, + "loss": 2.7157, + "step": 2556 + }, + { + "epoch": 0.327296, + "grad_norm": 0.47387319803237915, + "learning_rate": 0.0007856313631315726, + "loss": 2.7252, + "step": 2557 + }, + { + "epoch": 0.327424, + "grad_norm": 0.9276442527770996, + "learning_rate": 0.0007854611844077213, + "loss": 2.8169, + "step": 2558 + }, + { + "epoch": 0.327552, + "grad_norm": 0.6153337359428406, + "learning_rate": 0.0007852909566097828, + "loss": 2.6813, + "step": 2559 + }, + { + "epoch": 0.32768, + "grad_norm": 0.46642306447029114, + "learning_rate": 0.0007851206797670216, + "loss": 2.6197, + "step": 2560 + }, + { + "epoch": 0.327808, + "grad_norm": 0.5660092830657959, + "learning_rate": 0.0007849503539087101, + "loss": 2.6113, + "step": 2561 + }, + { + "epoch": 0.327936, + "grad_norm": 0.4524734318256378, + "learning_rate": 0.0007847799790641295, + "loss": 2.6584, + "step": 2562 + }, + { + "epoch": 0.328064, + "grad_norm": 0.4929679036140442, + "learning_rate": 0.0007846095552625688, + "loss": 2.7567, + "step": 2563 + }, + { + "epoch": 0.328192, + "grad_norm": 0.49571582674980164, + "learning_rate": 0.0007844390825333263, + "loss": 2.8022, + "step": 2564 + }, + { + "epoch": 0.32832, + "grad_norm": 0.5147833824157715, + "learning_rate": 0.0007842685609057077, + "loss": 2.6844, + "step": 2565 + }, + { + "epoch": 0.328448, + "grad_norm": 0.50923752784729, + "learning_rate": 0.0007840979904090282, + "loss": 2.6056, + "step": 2566 + }, + { + "epoch": 0.328576, + "grad_norm": 0.5054170489311218, + "learning_rate": 0.0007839273710726107, + "loss": 2.5898, + "step": 2567 + }, + { + "epoch": 0.328704, + "grad_norm": 0.5314496159553528, + "learning_rate": 0.0007837567029257864, + "loss": 2.6636, + "step": 2568 + }, + { + "epoch": 0.328832, + "grad_norm": 0.4752858579158783, + "learning_rate": 0.0007835859859978951, + "loss": 2.6581, + "step": 2569 + }, + { + "epoch": 0.32896, + "grad_norm": 0.5540331602096558, + "learning_rate": 0.0007834152203182854, + "loss": 2.8323, + "step": 2570 + }, + { + "epoch": 0.329088, + "grad_norm": 0.4749324321746826, + "learning_rate": 0.0007832444059163137, + "loss": 2.7364, + "step": 2571 + }, + { + "epoch": 0.329216, + "grad_norm": 0.5133570432662964, + "learning_rate": 0.0007830735428213451, + "loss": 2.65, + "step": 2572 + }, + { + "epoch": 0.329344, + "grad_norm": 0.6132932901382446, + "learning_rate": 0.0007829026310627527, + "loss": 2.7927, + "step": 2573 + }, + { + "epoch": 0.329472, + "grad_norm": 0.5227040648460388, + "learning_rate": 0.0007827316706699181, + "loss": 2.6355, + "step": 2574 + }, + { + "epoch": 0.3296, + "grad_norm": 0.5074606537818909, + "learning_rate": 0.0007825606616722318, + "loss": 2.865, + "step": 2575 + }, + { + "epoch": 0.329728, + "grad_norm": 0.5151696801185608, + "learning_rate": 0.0007823896040990921, + "loss": 2.7669, + "step": 2576 + }, + { + "epoch": 0.329856, + "grad_norm": 0.5212640762329102, + "learning_rate": 0.0007822184979799057, + "loss": 2.7299, + "step": 2577 + }, + { + "epoch": 0.329984, + "grad_norm": 0.5219990611076355, + "learning_rate": 0.0007820473433440876, + "loss": 2.6893, + "step": 2578 + }, + { + "epoch": 0.330112, + "grad_norm": 0.45889779925346375, + "learning_rate": 0.0007818761402210615, + "loss": 2.7343, + "step": 2579 + }, + { + "epoch": 0.33024, + "grad_norm": 0.5362652540206909, + "learning_rate": 0.0007817048886402592, + "loss": 2.6369, + "step": 2580 + }, + { + "epoch": 0.330368, + "grad_norm": 0.4811416566371918, + "learning_rate": 0.0007815335886311206, + "loss": 2.6809, + "step": 2581 + }, + { + "epoch": 0.330496, + "grad_norm": 0.6013070344924927, + "learning_rate": 0.0007813622402230943, + "loss": 2.7249, + "step": 2582 + }, + { + "epoch": 0.330624, + "grad_norm": 0.5002033114433289, + "learning_rate": 0.0007811908434456372, + "loss": 2.5596, + "step": 2583 + }, + { + "epoch": 0.330752, + "grad_norm": 0.49829399585723877, + "learning_rate": 0.0007810193983282141, + "loss": 2.6774, + "step": 2584 + }, + { + "epoch": 0.33088, + "grad_norm": 0.45730990171432495, + "learning_rate": 0.0007808479049002987, + "loss": 2.7241, + "step": 2585 + }, + { + "epoch": 0.331008, + "grad_norm": 0.46753621101379395, + "learning_rate": 0.0007806763631913725, + "loss": 2.6549, + "step": 2586 + }, + { + "epoch": 0.331136, + "grad_norm": 0.5013433694839478, + "learning_rate": 0.0007805047732309256, + "loss": 2.6444, + "step": 2587 + }, + { + "epoch": 0.331264, + "grad_norm": 0.6305160522460938, + "learning_rate": 0.0007803331350484563, + "loss": 2.6569, + "step": 2588 + }, + { + "epoch": 0.331392, + "grad_norm": 0.45029962062835693, + "learning_rate": 0.0007801614486734711, + "loss": 2.6307, + "step": 2589 + }, + { + "epoch": 0.33152, + "grad_norm": 0.44586989283561707, + "learning_rate": 0.0007799897141354853, + "loss": 2.687, + "step": 2590 + }, + { + "epoch": 0.331648, + "grad_norm": 0.4781543016433716, + "learning_rate": 0.0007798179314640214, + "loss": 2.7061, + "step": 2591 + }, + { + "epoch": 0.331776, + "grad_norm": 0.5743341445922852, + "learning_rate": 0.000779646100688611, + "loss": 2.6692, + "step": 2592 + }, + { + "epoch": 0.331904, + "grad_norm": 0.47281157970428467, + "learning_rate": 0.0007794742218387943, + "loss": 2.7471, + "step": 2593 + }, + { + "epoch": 0.332032, + "grad_norm": 0.4553368091583252, + "learning_rate": 0.0007793022949441189, + "loss": 2.6146, + "step": 2594 + }, + { + "epoch": 0.33216, + "grad_norm": 0.5296782851219177, + "learning_rate": 0.000779130320034141, + "loss": 2.7634, + "step": 2595 + }, + { + "epoch": 0.332288, + "grad_norm": 0.5079725384712219, + "learning_rate": 0.000778958297138425, + "loss": 2.7516, + "step": 2596 + }, + { + "epoch": 0.332416, + "grad_norm": 0.5941053628921509, + "learning_rate": 0.0007787862262865438, + "loss": 2.6413, + "step": 2597 + }, + { + "epoch": 0.332544, + "grad_norm": 0.4671876132488251, + "learning_rate": 0.0007786141075080785, + "loss": 2.7212, + "step": 2598 + }, + { + "epoch": 0.332672, + "grad_norm": 0.4861707389354706, + "learning_rate": 0.000778441940832618, + "loss": 2.7192, + "step": 2599 + }, + { + "epoch": 0.3328, + "grad_norm": 0.4995121955871582, + "learning_rate": 0.00077826972628976, + "loss": 2.6971, + "step": 2600 + }, + { + "epoch": 0.332928, + "grad_norm": 0.49643996357917786, + "learning_rate": 0.00077809746390911, + "loss": 2.7891, + "step": 2601 + }, + { + "epoch": 0.333056, + "grad_norm": 0.4545004069805145, + "learning_rate": 0.0007779251537202818, + "loss": 2.6766, + "step": 2602 + }, + { + "epoch": 0.333184, + "grad_norm": 0.5374282598495483, + "learning_rate": 0.0007777527957528981, + "loss": 2.7706, + "step": 2603 + }, + { + "epoch": 0.333312, + "grad_norm": 0.481202095746994, + "learning_rate": 0.0007775803900365885, + "loss": 2.6335, + "step": 2604 + }, + { + "epoch": 0.33344, + "grad_norm": 0.4587520658969879, + "learning_rate": 0.0007774079366009919, + "loss": 2.744, + "step": 2605 + }, + { + "epoch": 0.333568, + "grad_norm": 0.5247642397880554, + "learning_rate": 0.0007772354354757549, + "loss": 2.7318, + "step": 2606 + }, + { + "epoch": 0.333696, + "grad_norm": 0.5117645263671875, + "learning_rate": 0.0007770628866905329, + "loss": 2.7279, + "step": 2607 + }, + { + "epoch": 0.333824, + "grad_norm": 0.49331048130989075, + "learning_rate": 0.0007768902902749884, + "loss": 2.6772, + "step": 2608 + }, + { + "epoch": 0.333952, + "grad_norm": 0.4739198684692383, + "learning_rate": 0.0007767176462587931, + "loss": 2.783, + "step": 2609 + }, + { + "epoch": 0.33408, + "grad_norm": 0.4322013556957245, + "learning_rate": 0.0007765449546716266, + "loss": 2.6094, + "step": 2610 + }, + { + "epoch": 0.334208, + "grad_norm": 0.4827602207660675, + "learning_rate": 0.0007763722155431761, + "loss": 2.6358, + "step": 2611 + }, + { + "epoch": 0.334336, + "grad_norm": 0.5184532403945923, + "learning_rate": 0.000776199428903138, + "loss": 2.6821, + "step": 2612 + }, + { + "epoch": 0.334464, + "grad_norm": 0.46903955936431885, + "learning_rate": 0.000776026594781216, + "loss": 2.694, + "step": 2613 + }, + { + "epoch": 0.334592, + "grad_norm": 0.4707714915275574, + "learning_rate": 0.0007758537132071224, + "loss": 2.6106, + "step": 2614 + }, + { + "epoch": 0.33472, + "grad_norm": 0.6403118371963501, + "learning_rate": 0.0007756807842105774, + "loss": 2.6225, + "step": 2615 + }, + { + "epoch": 0.334848, + "grad_norm": 0.5491101741790771, + "learning_rate": 0.0007755078078213098, + "loss": 2.6564, + "step": 2616 + }, + { + "epoch": 0.334976, + "grad_norm": 0.4702625572681427, + "learning_rate": 0.000775334784069056, + "loss": 2.7025, + "step": 2617 + }, + { + "epoch": 0.335104, + "grad_norm": 0.5069715976715088, + "learning_rate": 0.0007751617129835609, + "loss": 2.7155, + "step": 2618 + }, + { + "epoch": 0.335232, + "grad_norm": 0.4462975859642029, + "learning_rate": 0.0007749885945945774, + "loss": 2.6879, + "step": 2619 + }, + { + "epoch": 0.33536, + "grad_norm": 0.685248613357544, + "learning_rate": 0.0007748154289318665, + "loss": 2.7684, + "step": 2620 + }, + { + "epoch": 0.335488, + "grad_norm": 0.4953244626522064, + "learning_rate": 0.0007746422160251975, + "loss": 2.6464, + "step": 2621 + }, + { + "epoch": 0.335616, + "grad_norm": 0.44611385464668274, + "learning_rate": 0.0007744689559043475, + "loss": 2.7243, + "step": 2622 + }, + { + "epoch": 0.335744, + "grad_norm": 0.4487125277519226, + "learning_rate": 0.0007742956485991022, + "loss": 2.6877, + "step": 2623 + }, + { + "epoch": 0.335872, + "grad_norm": 0.42250576615333557, + "learning_rate": 0.000774122294139255, + "loss": 2.5941, + "step": 2624 + }, + { + "epoch": 0.336, + "grad_norm": 0.45041605830192566, + "learning_rate": 0.0007739488925546075, + "loss": 2.7458, + "step": 2625 + }, + { + "epoch": 0.336128, + "grad_norm": 0.596748411655426, + "learning_rate": 0.0007737754438749693, + "loss": 2.6385, + "step": 2626 + }, + { + "epoch": 0.336256, + "grad_norm": 0.5343237519264221, + "learning_rate": 0.0007736019481301588, + "loss": 2.7455, + "step": 2627 + }, + { + "epoch": 0.336384, + "grad_norm": 0.4919133186340332, + "learning_rate": 0.0007734284053500013, + "loss": 2.6794, + "step": 2628 + }, + { + "epoch": 0.336512, + "grad_norm": 0.4921506345272064, + "learning_rate": 0.000773254815564331, + "loss": 2.7911, + "step": 2629 + }, + { + "epoch": 0.33664, + "grad_norm": 0.6178864240646362, + "learning_rate": 0.0007730811788029902, + "loss": 2.7142, + "step": 2630 + }, + { + "epoch": 0.336768, + "grad_norm": 0.6205880045890808, + "learning_rate": 0.0007729074950958289, + "loss": 2.7513, + "step": 2631 + }, + { + "epoch": 0.336896, + "grad_norm": 0.5030384063720703, + "learning_rate": 0.0007727337644727055, + "loss": 2.7857, + "step": 2632 + }, + { + "epoch": 0.337024, + "grad_norm": 0.5359963774681091, + "learning_rate": 0.000772559986963486, + "loss": 2.6643, + "step": 2633 + }, + { + "epoch": 0.337152, + "grad_norm": 0.6612207889556885, + "learning_rate": 0.000772386162598045, + "loss": 2.7161, + "step": 2634 + }, + { + "epoch": 0.33728, + "grad_norm": 0.44476062059402466, + "learning_rate": 0.0007722122914062649, + "loss": 2.6772, + "step": 2635 + }, + { + "epoch": 0.337408, + "grad_norm": 0.4950471520423889, + "learning_rate": 0.0007720383734180361, + "loss": 2.7155, + "step": 2636 + }, + { + "epoch": 0.337536, + "grad_norm": 0.4797663688659668, + "learning_rate": 0.0007718644086632573, + "loss": 2.7579, + "step": 2637 + }, + { + "epoch": 0.337664, + "grad_norm": 0.48678120970726013, + "learning_rate": 0.000771690397171835, + "loss": 2.7467, + "step": 2638 + }, + { + "epoch": 0.337792, + "grad_norm": 0.5050081610679626, + "learning_rate": 0.0007715163389736834, + "loss": 2.641, + "step": 2639 + }, + { + "epoch": 0.33792, + "grad_norm": 0.5242562294006348, + "learning_rate": 0.0007713422340987257, + "loss": 2.61, + "step": 2640 + }, + { + "epoch": 0.338048, + "grad_norm": 0.5045949220657349, + "learning_rate": 0.0007711680825768921, + "loss": 2.7068, + "step": 2641 + }, + { + "epoch": 0.338176, + "grad_norm": 0.5316445231437683, + "learning_rate": 0.0007709938844381217, + "loss": 2.7661, + "step": 2642 + }, + { + "epoch": 0.338304, + "grad_norm": 0.48030027747154236, + "learning_rate": 0.0007708196397123607, + "loss": 2.6772, + "step": 2643 + }, + { + "epoch": 0.338432, + "grad_norm": 0.5007248520851135, + "learning_rate": 0.0007706453484295641, + "loss": 2.7537, + "step": 2644 + }, + { + "epoch": 0.33856, + "grad_norm": 0.5464178323745728, + "learning_rate": 0.0007704710106196946, + "loss": 2.6194, + "step": 2645 + }, + { + "epoch": 0.338688, + "grad_norm": 0.46075439453125, + "learning_rate": 0.0007702966263127227, + "loss": 2.6888, + "step": 2646 + }, + { + "epoch": 0.338816, + "grad_norm": 0.468636691570282, + "learning_rate": 0.0007701221955386273, + "loss": 2.6177, + "step": 2647 + }, + { + "epoch": 0.338944, + "grad_norm": 0.4459661841392517, + "learning_rate": 0.0007699477183273948, + "loss": 2.6989, + "step": 2648 + }, + { + "epoch": 0.339072, + "grad_norm": 0.4772389233112335, + "learning_rate": 0.00076977319470902, + "loss": 2.6981, + "step": 2649 + }, + { + "epoch": 0.3392, + "grad_norm": 0.47440263628959656, + "learning_rate": 0.0007695986247135059, + "loss": 2.7362, + "step": 2650 + }, + { + "epoch": 0.339328, + "grad_norm": 0.47389811277389526, + "learning_rate": 0.0007694240083708624, + "loss": 2.7294, + "step": 2651 + }, + { + "epoch": 0.339456, + "grad_norm": 0.4681686758995056, + "learning_rate": 0.0007692493457111085, + "loss": 2.7348, + "step": 2652 + }, + { + "epoch": 0.339584, + "grad_norm": 2.0587267875671387, + "learning_rate": 0.0007690746367642707, + "loss": 2.6596, + "step": 2653 + }, + { + "epoch": 0.339712, + "grad_norm": 0.4848611652851105, + "learning_rate": 0.0007688998815603836, + "loss": 2.6621, + "step": 2654 + }, + { + "epoch": 0.33984, + "grad_norm": 0.4478171467781067, + "learning_rate": 0.0007687250801294892, + "loss": 2.8525, + "step": 2655 + }, + { + "epoch": 0.339968, + "grad_norm": 0.7593162655830383, + "learning_rate": 0.0007685502325016385, + "loss": 2.7568, + "step": 2656 + }, + { + "epoch": 0.340096, + "grad_norm": 0.45905524492263794, + "learning_rate": 0.0007683753387068893, + "loss": 2.8062, + "step": 2657 + }, + { + "epoch": 0.340224, + "grad_norm": 0.4981978237628937, + "learning_rate": 0.0007682003987753083, + "loss": 2.7639, + "step": 2658 + }, + { + "epoch": 0.340352, + "grad_norm": 0.47199493646621704, + "learning_rate": 0.0007680254127369695, + "loss": 2.7768, + "step": 2659 + }, + { + "epoch": 0.34048, + "grad_norm": 0.49306967854499817, + "learning_rate": 0.0007678503806219549, + "loss": 2.6508, + "step": 2660 + }, + { + "epoch": 0.340608, + "grad_norm": 0.4497023820877075, + "learning_rate": 0.0007676753024603546, + "loss": 2.6985, + "step": 2661 + }, + { + "epoch": 0.340736, + "grad_norm": 0.49964553117752075, + "learning_rate": 0.0007675001782822668, + "loss": 2.7493, + "step": 2662 + }, + { + "epoch": 0.340864, + "grad_norm": 0.541168212890625, + "learning_rate": 0.0007673250081177972, + "loss": 2.72, + "step": 2663 + }, + { + "epoch": 0.340992, + "grad_norm": 0.5703802704811096, + "learning_rate": 0.0007671497919970597, + "loss": 2.7155, + "step": 2664 + }, + { + "epoch": 0.34112, + "grad_norm": 0.5080617666244507, + "learning_rate": 0.0007669745299501758, + "loss": 2.6071, + "step": 2665 + }, + { + "epoch": 0.341248, + "grad_norm": 0.46402254700660706, + "learning_rate": 0.0007667992220072751, + "loss": 2.6664, + "step": 2666 + }, + { + "epoch": 0.341376, + "grad_norm": 0.4558883309364319, + "learning_rate": 0.0007666238681984952, + "loss": 2.7611, + "step": 2667 + }, + { + "epoch": 0.341504, + "grad_norm": 0.5211532711982727, + "learning_rate": 0.0007664484685539814, + "loss": 2.6822, + "step": 2668 + }, + { + "epoch": 0.341632, + "grad_norm": 0.4725772738456726, + "learning_rate": 0.0007662730231038869, + "loss": 2.6316, + "step": 2669 + }, + { + "epoch": 0.34176, + "grad_norm": 0.45442813634872437, + "learning_rate": 0.0007660975318783729, + "loss": 2.6495, + "step": 2670 + }, + { + "epoch": 0.341888, + "grad_norm": 0.4776241183280945, + "learning_rate": 0.0007659219949076081, + "loss": 2.7442, + "step": 2671 + }, + { + "epoch": 0.342016, + "grad_norm": 0.6457310914993286, + "learning_rate": 0.0007657464122217697, + "loss": 2.6693, + "step": 2672 + }, + { + "epoch": 0.342144, + "grad_norm": 0.46596112847328186, + "learning_rate": 0.0007655707838510423, + "loss": 2.6865, + "step": 2673 + }, + { + "epoch": 0.342272, + "grad_norm": 0.42607131600379944, + "learning_rate": 0.0007653951098256183, + "loss": 2.6375, + "step": 2674 + }, + { + "epoch": 0.3424, + "grad_norm": 0.5538437366485596, + "learning_rate": 0.0007652193901756983, + "loss": 2.7232, + "step": 2675 + }, + { + "epoch": 0.342528, + "grad_norm": 0.4219939410686493, + "learning_rate": 0.0007650436249314903, + "loss": 2.6951, + "step": 2676 + }, + { + "epoch": 0.342656, + "grad_norm": 0.4713102877140045, + "learning_rate": 0.0007648678141232106, + "loss": 2.7435, + "step": 2677 + }, + { + "epoch": 0.342784, + "grad_norm": 0.4305635988712311, + "learning_rate": 0.000764691957781083, + "loss": 2.8059, + "step": 2678 + }, + { + "epoch": 0.342912, + "grad_norm": 0.43878090381622314, + "learning_rate": 0.0007645160559353393, + "loss": 2.8836, + "step": 2679 + }, + { + "epoch": 0.34304, + "grad_norm": 0.5294603109359741, + "learning_rate": 0.000764340108616219, + "loss": 2.7789, + "step": 2680 + }, + { + "epoch": 0.343168, + "grad_norm": 0.45951688289642334, + "learning_rate": 0.0007641641158539697, + "loss": 2.6353, + "step": 2681 + }, + { + "epoch": 0.343296, + "grad_norm": 0.5120280385017395, + "learning_rate": 0.0007639880776788464, + "loss": 2.7032, + "step": 2682 + }, + { + "epoch": 0.343424, + "grad_norm": 0.42373278737068176, + "learning_rate": 0.0007638119941211119, + "loss": 2.7288, + "step": 2683 + }, + { + "epoch": 0.343552, + "grad_norm": 0.4415596127510071, + "learning_rate": 0.0007636358652110375, + "loss": 2.7103, + "step": 2684 + }, + { + "epoch": 0.34368, + "grad_norm": 0.42392629384994507, + "learning_rate": 0.0007634596909789013, + "loss": 2.6327, + "step": 2685 + }, + { + "epoch": 0.343808, + "grad_norm": 0.507201611995697, + "learning_rate": 0.0007632834714549901, + "loss": 2.6755, + "step": 2686 + }, + { + "epoch": 0.343936, + "grad_norm": 0.45773783326148987, + "learning_rate": 0.0007631072066695978, + "loss": 2.7881, + "step": 2687 + }, + { + "epoch": 0.344064, + "grad_norm": 0.4418371319770813, + "learning_rate": 0.0007629308966530268, + "loss": 2.6202, + "step": 2688 + }, + { + "epoch": 0.344192, + "grad_norm": 0.6277620792388916, + "learning_rate": 0.0007627545414355861, + "loss": 2.8098, + "step": 2689 + }, + { + "epoch": 0.34432, + "grad_norm": 0.4525109827518463, + "learning_rate": 0.0007625781410475938, + "loss": 2.787, + "step": 2690 + }, + { + "epoch": 0.344448, + "grad_norm": 0.7210133075714111, + "learning_rate": 0.000762401695519375, + "loss": 2.7368, + "step": 2691 + }, + { + "epoch": 0.344576, + "grad_norm": 0.473358154296875, + "learning_rate": 0.0007622252048812628, + "loss": 2.6464, + "step": 2692 + }, + { + "epoch": 0.344704, + "grad_norm": 0.4680347442626953, + "learning_rate": 0.0007620486691635977, + "loss": 2.7556, + "step": 2693 + }, + { + "epoch": 0.344832, + "grad_norm": 0.5034725666046143, + "learning_rate": 0.0007618720883967285, + "loss": 2.6337, + "step": 2694 + }, + { + "epoch": 0.34496, + "grad_norm": 0.46354690194129944, + "learning_rate": 0.0007616954626110115, + "loss": 2.5892, + "step": 2695 + }, + { + "epoch": 0.345088, + "grad_norm": 0.4546692669391632, + "learning_rate": 0.0007615187918368106, + "loss": 2.7989, + "step": 2696 + }, + { + "epoch": 0.345216, + "grad_norm": 0.5051708221435547, + "learning_rate": 0.0007613420761044976, + "loss": 2.6423, + "step": 2697 + }, + { + "epoch": 0.345344, + "grad_norm": 0.5238072872161865, + "learning_rate": 0.000761165315444452, + "loss": 2.7804, + "step": 2698 + }, + { + "epoch": 0.345472, + "grad_norm": 0.4573189914226532, + "learning_rate": 0.0007609885098870609, + "loss": 2.7675, + "step": 2699 + }, + { + "epoch": 0.3456, + "grad_norm": 0.48363369703292847, + "learning_rate": 0.0007608116594627195, + "loss": 2.7534, + "step": 2700 + }, + { + "epoch": 0.345728, + "grad_norm": 0.451181560754776, + "learning_rate": 0.0007606347642018301, + "loss": 2.6823, + "step": 2701 + }, + { + "epoch": 0.345856, + "grad_norm": 0.4982965290546417, + "learning_rate": 0.0007604578241348032, + "loss": 2.7087, + "step": 2702 + }, + { + "epoch": 0.345984, + "grad_norm": 0.46901488304138184, + "learning_rate": 0.0007602808392920569, + "loss": 2.7094, + "step": 2703 + }, + { + "epoch": 0.346112, + "grad_norm": 0.4354756772518158, + "learning_rate": 0.0007601038097040169, + "loss": 2.8394, + "step": 2704 + }, + { + "epoch": 0.34624, + "grad_norm": 0.4903188645839691, + "learning_rate": 0.0007599267354011166, + "loss": 2.729, + "step": 2705 + }, + { + "epoch": 0.346368, + "grad_norm": 0.4693494737148285, + "learning_rate": 0.0007597496164137972, + "loss": 2.7331, + "step": 2706 + }, + { + "epoch": 0.346496, + "grad_norm": 0.44422784447669983, + "learning_rate": 0.0007595724527725074, + "loss": 2.8175, + "step": 2707 + }, + { + "epoch": 0.346624, + "grad_norm": 0.4750046133995056, + "learning_rate": 0.0007593952445077035, + "loss": 2.8309, + "step": 2708 + }, + { + "epoch": 0.346752, + "grad_norm": 0.43334031105041504, + "learning_rate": 0.0007592179916498503, + "loss": 2.6518, + "step": 2709 + }, + { + "epoch": 0.34688, + "grad_norm": 0.7594212889671326, + "learning_rate": 0.0007590406942294191, + "loss": 2.6831, + "step": 2710 + }, + { + "epoch": 0.347008, + "grad_norm": 0.4918264150619507, + "learning_rate": 0.0007588633522768895, + "loss": 2.7349, + "step": 2711 + }, + { + "epoch": 0.347136, + "grad_norm": 0.44594627618789673, + "learning_rate": 0.0007586859658227486, + "loss": 2.618, + "step": 2712 + }, + { + "epoch": 0.347264, + "grad_norm": 0.446895569562912, + "learning_rate": 0.0007585085348974912, + "loss": 2.7388, + "step": 2713 + }, + { + "epoch": 0.347392, + "grad_norm": 0.46774575114250183, + "learning_rate": 0.0007583310595316199, + "loss": 2.7242, + "step": 2714 + }, + { + "epoch": 0.34752, + "grad_norm": 0.4894832372665405, + "learning_rate": 0.0007581535397556445, + "loss": 2.6574, + "step": 2715 + }, + { + "epoch": 0.347648, + "grad_norm": 0.4994119107723236, + "learning_rate": 0.000757975975600083, + "loss": 2.6174, + "step": 2716 + }, + { + "epoch": 0.347776, + "grad_norm": 0.5232442021369934, + "learning_rate": 0.0007577983670954603, + "loss": 2.5294, + "step": 2717 + }, + { + "epoch": 0.347904, + "grad_norm": 0.4395061135292053, + "learning_rate": 0.0007576207142723099, + "loss": 2.6015, + "step": 2718 + }, + { + "epoch": 0.348032, + "grad_norm": 0.4295394718647003, + "learning_rate": 0.000757443017161172, + "loss": 2.5255, + "step": 2719 + }, + { + "epoch": 0.34816, + "grad_norm": 0.4649607241153717, + "learning_rate": 0.0007572652757925951, + "loss": 2.6918, + "step": 2720 + }, + { + "epoch": 0.348288, + "grad_norm": 0.5237522125244141, + "learning_rate": 0.0007570874901971347, + "loss": 2.7246, + "step": 2721 + }, + { + "epoch": 0.348416, + "grad_norm": 0.48668423295021057, + "learning_rate": 0.0007569096604053543, + "loss": 2.8309, + "step": 2722 + }, + { + "epoch": 0.348544, + "grad_norm": 0.46027258038520813, + "learning_rate": 0.0007567317864478247, + "loss": 2.5963, + "step": 2723 + }, + { + "epoch": 0.348672, + "grad_norm": 0.4759104251861572, + "learning_rate": 0.0007565538683551251, + "loss": 2.6374, + "step": 2724 + }, + { + "epoch": 0.3488, + "grad_norm": 0.46329259872436523, + "learning_rate": 0.0007563759061578411, + "loss": 2.7088, + "step": 2725 + }, + { + "epoch": 0.348928, + "grad_norm": 0.4713778793811798, + "learning_rate": 0.0007561978998865666, + "loss": 2.6335, + "step": 2726 + }, + { + "epoch": 0.349056, + "grad_norm": 0.44144144654273987, + "learning_rate": 0.0007560198495719032, + "loss": 2.7192, + "step": 2727 + }, + { + "epoch": 0.349184, + "grad_norm": 0.5319382548332214, + "learning_rate": 0.0007558417552444596, + "loss": 2.6129, + "step": 2728 + }, + { + "epoch": 0.349312, + "grad_norm": 0.4931134879589081, + "learning_rate": 0.0007556636169348523, + "loss": 2.6622, + "step": 2729 + }, + { + "epoch": 0.34944, + "grad_norm": 0.4713144302368164, + "learning_rate": 0.0007554854346737052, + "loss": 2.6801, + "step": 2730 + }, + { + "epoch": 0.349568, + "grad_norm": 0.43408435583114624, + "learning_rate": 0.00075530720849165, + "loss": 2.7721, + "step": 2731 + }, + { + "epoch": 0.349696, + "grad_norm": 0.45249509811401367, + "learning_rate": 0.0007551289384193261, + "loss": 2.6521, + "step": 2732 + }, + { + "epoch": 0.349824, + "grad_norm": 0.5061367154121399, + "learning_rate": 0.0007549506244873799, + "loss": 2.6913, + "step": 2733 + }, + { + "epoch": 0.349952, + "grad_norm": 0.5255594849586487, + "learning_rate": 0.0007547722667264656, + "loss": 2.7759, + "step": 2734 + }, + { + "epoch": 0.35008, + "grad_norm": 0.46052080392837524, + "learning_rate": 0.0007545938651672453, + "loss": 2.6585, + "step": 2735 + }, + { + "epoch": 0.350208, + "grad_norm": 0.4631306231021881, + "learning_rate": 0.0007544154198403879, + "loss": 2.6749, + "step": 2736 + }, + { + "epoch": 0.350336, + "grad_norm": 0.4549815356731415, + "learning_rate": 0.0007542369307765704, + "loss": 2.7571, + "step": 2737 + }, + { + "epoch": 0.350464, + "grad_norm": 0.4157131016254425, + "learning_rate": 0.0007540583980064771, + "loss": 2.7198, + "step": 2738 + }, + { + "epoch": 0.350592, + "grad_norm": 0.4643344581127167, + "learning_rate": 0.0007538798215608, + "loss": 2.649, + "step": 2739 + }, + { + "epoch": 0.35072, + "grad_norm": 0.41892045736312866, + "learning_rate": 0.0007537012014702382, + "loss": 2.6784, + "step": 2740 + }, + { + "epoch": 0.350848, + "grad_norm": 0.48754769563674927, + "learning_rate": 0.0007535225377654989, + "loss": 2.6767, + "step": 2741 + }, + { + "epoch": 0.350976, + "grad_norm": 0.48485440015792847, + "learning_rate": 0.0007533438304772962, + "loss": 2.7094, + "step": 2742 + }, + { + "epoch": 0.351104, + "grad_norm": 0.4926983714103699, + "learning_rate": 0.0007531650796363521, + "loss": 2.6643, + "step": 2743 + }, + { + "epoch": 0.351232, + "grad_norm": 0.4316820204257965, + "learning_rate": 0.0007529862852733957, + "loss": 2.7682, + "step": 2744 + }, + { + "epoch": 0.35136, + "grad_norm": 0.46376681327819824, + "learning_rate": 0.000752807447419164, + "loss": 2.6825, + "step": 2745 + }, + { + "epoch": 0.351488, + "grad_norm": 0.477416068315506, + "learning_rate": 0.0007526285661044014, + "loss": 2.6361, + "step": 2746 + }, + { + "epoch": 0.351616, + "grad_norm": 0.5014001727104187, + "learning_rate": 0.0007524496413598595, + "loss": 2.7426, + "step": 2747 + }, + { + "epoch": 0.351744, + "grad_norm": 0.46568822860717773, + "learning_rate": 0.0007522706732162976, + "loss": 2.7881, + "step": 2748 + }, + { + "epoch": 0.351872, + "grad_norm": 0.4857662618160248, + "learning_rate": 0.0007520916617044825, + "loss": 2.8255, + "step": 2749 + }, + { + "epoch": 0.352, + "grad_norm": 0.4515175223350525, + "learning_rate": 0.0007519126068551882, + "loss": 2.7291, + "step": 2750 + }, + { + "epoch": 0.352128, + "grad_norm": 0.48131126165390015, + "learning_rate": 0.0007517335086991964, + "loss": 2.8029, + "step": 2751 + }, + { + "epoch": 0.352256, + "grad_norm": 0.4160010516643524, + "learning_rate": 0.000751554367267296, + "loss": 2.7968, + "step": 2752 + }, + { + "epoch": 0.352384, + "grad_norm": 0.4348675310611725, + "learning_rate": 0.0007513751825902837, + "loss": 2.6241, + "step": 2753 + }, + { + "epoch": 0.352512, + "grad_norm": 0.458451509475708, + "learning_rate": 0.0007511959546989632, + "loss": 2.6186, + "step": 2754 + }, + { + "epoch": 0.35264, + "grad_norm": 0.4544239640235901, + "learning_rate": 0.000751016683624146, + "loss": 2.7692, + "step": 2755 + }, + { + "epoch": 0.352768, + "grad_norm": 0.4764367640018463, + "learning_rate": 0.0007508373693966509, + "loss": 2.6912, + "step": 2756 + }, + { + "epoch": 0.352896, + "grad_norm": 0.4374392330646515, + "learning_rate": 0.0007506580120473041, + "loss": 2.7078, + "step": 2757 + }, + { + "epoch": 0.353024, + "grad_norm": 0.4727518558502197, + "learning_rate": 0.0007504786116069391, + "loss": 2.757, + "step": 2758 + }, + { + "epoch": 0.353152, + "grad_norm": 0.49557122588157654, + "learning_rate": 0.0007502991681063969, + "loss": 2.7705, + "step": 2759 + }, + { + "epoch": 0.35328, + "grad_norm": 0.4641502797603607, + "learning_rate": 0.0007501196815765261, + "loss": 2.6938, + "step": 2760 + }, + { + "epoch": 0.353408, + "grad_norm": 0.4675647020339966, + "learning_rate": 0.0007499401520481824, + "loss": 2.6718, + "step": 2761 + }, + { + "epoch": 0.353536, + "grad_norm": 0.47594618797302246, + "learning_rate": 0.0007497605795522288, + "loss": 2.7111, + "step": 2762 + }, + { + "epoch": 0.353664, + "grad_norm": 0.5657196640968323, + "learning_rate": 0.0007495809641195364, + "loss": 2.6221, + "step": 2763 + }, + { + "epoch": 0.353792, + "grad_norm": 0.47059890627861023, + "learning_rate": 0.0007494013057809827, + "loss": 2.7252, + "step": 2764 + }, + { + "epoch": 0.35392, + "grad_norm": 0.45090481638908386, + "learning_rate": 0.0007492216045674532, + "loss": 2.7337, + "step": 2765 + }, + { + "epoch": 0.354048, + "grad_norm": 0.4720679819583893, + "learning_rate": 0.0007490418605098408, + "loss": 2.649, + "step": 2766 + }, + { + "epoch": 0.354176, + "grad_norm": 0.4530324339866638, + "learning_rate": 0.0007488620736390454, + "loss": 2.6895, + "step": 2767 + }, + { + "epoch": 0.354304, + "grad_norm": 0.4221791923046112, + "learning_rate": 0.0007486822439859744, + "loss": 2.5929, + "step": 2768 + }, + { + "epoch": 0.354432, + "grad_norm": 0.5041205286979675, + "learning_rate": 0.0007485023715815426, + "loss": 2.653, + "step": 2769 + }, + { + "epoch": 0.35456, + "grad_norm": 0.4709184765815735, + "learning_rate": 0.0007483224564566724, + "loss": 2.7625, + "step": 2770 + }, + { + "epoch": 0.354688, + "grad_norm": 0.48218750953674316, + "learning_rate": 0.0007481424986422932, + "loss": 2.67, + "step": 2771 + }, + { + "epoch": 0.354816, + "grad_norm": 0.43823832273483276, + "learning_rate": 0.0007479624981693416, + "loss": 2.727, + "step": 2772 + }, + { + "epoch": 0.354944, + "grad_norm": 0.43520388007164, + "learning_rate": 0.000747782455068762, + "loss": 2.5611, + "step": 2773 + }, + { + "epoch": 0.355072, + "grad_norm": 0.42899924516677856, + "learning_rate": 0.0007476023693715058, + "loss": 2.6533, + "step": 2774 + }, + { + "epoch": 0.3552, + "grad_norm": 0.45936495065689087, + "learning_rate": 0.000747422241108532, + "loss": 2.7512, + "step": 2775 + }, + { + "epoch": 0.355328, + "grad_norm": 0.4750352203845978, + "learning_rate": 0.0007472420703108064, + "loss": 2.751, + "step": 2776 + }, + { + "epoch": 0.355456, + "grad_norm": 0.4708814024925232, + "learning_rate": 0.0007470618570093026, + "loss": 2.8057, + "step": 2777 + }, + { + "epoch": 0.355584, + "grad_norm": 0.44537246227264404, + "learning_rate": 0.0007468816012350016, + "loss": 2.5847, + "step": 2778 + }, + { + "epoch": 0.355712, + "grad_norm": 0.4353965222835541, + "learning_rate": 0.0007467013030188911, + "loss": 2.6567, + "step": 2779 + }, + { + "epoch": 0.35584, + "grad_norm": 0.44848254323005676, + "learning_rate": 0.0007465209623919668, + "loss": 2.7586, + "step": 2780 + }, + { + "epoch": 0.355968, + "grad_norm": 0.44026562571525574, + "learning_rate": 0.0007463405793852309, + "loss": 2.6547, + "step": 2781 + }, + { + "epoch": 0.356096, + "grad_norm": 0.4524605870246887, + "learning_rate": 0.0007461601540296937, + "loss": 2.6529, + "step": 2782 + }, + { + "epoch": 0.356224, + "grad_norm": 0.46165981888771057, + "learning_rate": 0.0007459796863563723, + "loss": 2.8151, + "step": 2783 + }, + { + "epoch": 0.356352, + "grad_norm": 0.44827839732170105, + "learning_rate": 0.0007457991763962911, + "loss": 2.6943, + "step": 2784 + }, + { + "epoch": 0.35648, + "grad_norm": 0.5175144076347351, + "learning_rate": 0.0007456186241804821, + "loss": 2.7825, + "step": 2785 + }, + { + "epoch": 0.356608, + "grad_norm": 0.5228431224822998, + "learning_rate": 0.000745438029739984, + "loss": 2.6116, + "step": 2786 + }, + { + "epoch": 0.356736, + "grad_norm": 0.5849762558937073, + "learning_rate": 0.0007452573931058432, + "loss": 2.6885, + "step": 2787 + }, + { + "epoch": 0.356864, + "grad_norm": 0.5490732789039612, + "learning_rate": 0.0007450767143091133, + "loss": 2.7458, + "step": 2788 + }, + { + "epoch": 0.356992, + "grad_norm": 0.539766252040863, + "learning_rate": 0.0007448959933808551, + "loss": 2.6806, + "step": 2789 + }, + { + "epoch": 0.35712, + "grad_norm": 0.43395742774009705, + "learning_rate": 0.0007447152303521365, + "loss": 2.6514, + "step": 2790 + }, + { + "epoch": 0.357248, + "grad_norm": 0.5158015489578247, + "learning_rate": 0.0007445344252540327, + "loss": 2.6913, + "step": 2791 + }, + { + "epoch": 0.357376, + "grad_norm": 0.440975159406662, + "learning_rate": 0.0007443535781176265, + "loss": 2.7475, + "step": 2792 + }, + { + "epoch": 0.357504, + "grad_norm": 0.4303780198097229, + "learning_rate": 0.0007441726889740075, + "loss": 2.6311, + "step": 2793 + }, + { + "epoch": 0.357632, + "grad_norm": 0.4619155526161194, + "learning_rate": 0.0007439917578542723, + "loss": 2.6435, + "step": 2794 + }, + { + "epoch": 0.35776, + "grad_norm": 0.5777912735939026, + "learning_rate": 0.0007438107847895256, + "loss": 2.6191, + "step": 2795 + }, + { + "epoch": 0.357888, + "grad_norm": 0.448974609375, + "learning_rate": 0.0007436297698108782, + "loss": 2.6403, + "step": 2796 + }, + { + "epoch": 0.358016, + "grad_norm": 0.4426156282424927, + "learning_rate": 0.0007434487129494493, + "loss": 2.6792, + "step": 2797 + }, + { + "epoch": 0.358144, + "grad_norm": 0.4050840735435486, + "learning_rate": 0.0007432676142363642, + "loss": 2.7929, + "step": 2798 + }, + { + "epoch": 0.358272, + "grad_norm": 0.5335375070571899, + "learning_rate": 0.0007430864737027561, + "loss": 2.7025, + "step": 2799 + }, + { + "epoch": 0.3584, + "grad_norm": 0.4465997517108917, + "learning_rate": 0.0007429052913797648, + "loss": 2.7013, + "step": 2800 + }, + { + "epoch": 0.358528, + "grad_norm": 0.4201470911502838, + "learning_rate": 0.000742724067298538, + "loss": 2.6824, + "step": 2801 + }, + { + "epoch": 0.358656, + "grad_norm": 0.4278305172920227, + "learning_rate": 0.0007425428014902303, + "loss": 2.7463, + "step": 2802 + }, + { + "epoch": 0.358784, + "grad_norm": 0.4747074842453003, + "learning_rate": 0.0007423614939860032, + "loss": 2.6983, + "step": 2803 + }, + { + "epoch": 0.358912, + "grad_norm": 0.4183098077774048, + "learning_rate": 0.0007421801448170256, + "loss": 2.6137, + "step": 2804 + }, + { + "epoch": 0.35904, + "grad_norm": 0.4281679093837738, + "learning_rate": 0.0007419987540144733, + "loss": 2.5894, + "step": 2805 + }, + { + "epoch": 0.359168, + "grad_norm": 0.4583606421947479, + "learning_rate": 0.0007418173216095299, + "loss": 2.7147, + "step": 2806 + }, + { + "epoch": 0.359296, + "grad_norm": 0.4678187370300293, + "learning_rate": 0.0007416358476333855, + "loss": 2.661, + "step": 2807 + }, + { + "epoch": 0.359424, + "grad_norm": 0.4583361744880676, + "learning_rate": 0.0007414543321172376, + "loss": 2.6901, + "step": 2808 + }, + { + "epoch": 0.359552, + "grad_norm": 0.47278088331222534, + "learning_rate": 0.0007412727750922908, + "loss": 2.691, + "step": 2809 + }, + { + "epoch": 0.35968, + "grad_norm": 0.46271762251853943, + "learning_rate": 0.0007410911765897567, + "loss": 2.7033, + "step": 2810 + }, + { + "epoch": 0.359808, + "grad_norm": 0.44931742548942566, + "learning_rate": 0.0007409095366408547, + "loss": 2.6165, + "step": 2811 + }, + { + "epoch": 0.359936, + "grad_norm": 0.5250083804130554, + "learning_rate": 0.0007407278552768103, + "loss": 2.71, + "step": 2812 + }, + { + "epoch": 0.360064, + "grad_norm": 0.4430692195892334, + "learning_rate": 0.0007405461325288569, + "loss": 2.6482, + "step": 2813 + }, + { + "epoch": 0.360192, + "grad_norm": 0.473458468914032, + "learning_rate": 0.0007403643684282346, + "loss": 2.7319, + "step": 2814 + }, + { + "epoch": 0.36032, + "grad_norm": 0.4768596887588501, + "learning_rate": 0.000740182563006191, + "loss": 2.6655, + "step": 2815 + }, + { + "epoch": 0.360448, + "grad_norm": 0.45600271224975586, + "learning_rate": 0.0007400007162939802, + "loss": 2.7284, + "step": 2816 + }, + { + "epoch": 0.360576, + "grad_norm": 0.4411280155181885, + "learning_rate": 0.000739818828322864, + "loss": 2.6928, + "step": 2817 + }, + { + "epoch": 0.360704, + "grad_norm": 0.4802837371826172, + "learning_rate": 0.000739636899124111, + "loss": 2.758, + "step": 2818 + }, + { + "epoch": 0.360832, + "grad_norm": 0.42581894993782043, + "learning_rate": 0.0007394549287289969, + "loss": 2.7035, + "step": 2819 + }, + { + "epoch": 0.36096, + "grad_norm": 0.43644294142723083, + "learning_rate": 0.0007392729171688046, + "loss": 2.7568, + "step": 2820 + }, + { + "epoch": 0.361088, + "grad_norm": 0.4549701511859894, + "learning_rate": 0.000739090864474824, + "loss": 2.7541, + "step": 2821 + }, + { + "epoch": 0.361216, + "grad_norm": 0.40372517704963684, + "learning_rate": 0.0007389087706783518, + "loss": 2.6135, + "step": 2822 + }, + { + "epoch": 0.361344, + "grad_norm": 0.4421142637729645, + "learning_rate": 0.0007387266358106924, + "loss": 2.6087, + "step": 2823 + }, + { + "epoch": 0.361472, + "grad_norm": 0.4718833863735199, + "learning_rate": 0.0007385444599031567, + "loss": 2.7115, + "step": 2824 + }, + { + "epoch": 0.3616, + "grad_norm": 0.4323756992816925, + "learning_rate": 0.000738362242987063, + "loss": 2.6262, + "step": 2825 + }, + { + "epoch": 0.361728, + "grad_norm": 0.4571376442909241, + "learning_rate": 0.0007381799850937363, + "loss": 2.8174, + "step": 2826 + }, + { + "epoch": 0.361856, + "grad_norm": 0.4889936149120331, + "learning_rate": 0.0007379976862545091, + "loss": 2.6822, + "step": 2827 + }, + { + "epoch": 0.361984, + "grad_norm": 0.476522296667099, + "learning_rate": 0.0007378153465007202, + "loss": 2.7816, + "step": 2828 + }, + { + "epoch": 0.362112, + "grad_norm": 0.4727150499820709, + "learning_rate": 0.0007376329658637166, + "loss": 2.6857, + "step": 2829 + }, + { + "epoch": 0.36224, + "grad_norm": 0.4578971266746521, + "learning_rate": 0.0007374505443748511, + "loss": 2.5923, + "step": 2830 + }, + { + "epoch": 0.362368, + "grad_norm": 0.49818405508995056, + "learning_rate": 0.0007372680820654843, + "loss": 2.6482, + "step": 2831 + }, + { + "epoch": 0.362496, + "grad_norm": 0.43027588725090027, + "learning_rate": 0.0007370855789669837, + "loss": 2.6884, + "step": 2832 + }, + { + "epoch": 0.362624, + "grad_norm": 0.410414457321167, + "learning_rate": 0.0007369030351107232, + "loss": 2.6465, + "step": 2833 + }, + { + "epoch": 0.362752, + "grad_norm": 0.47063958644866943, + "learning_rate": 0.0007367204505280847, + "loss": 2.655, + "step": 2834 + }, + { + "epoch": 0.36288, + "grad_norm": 0.4370226562023163, + "learning_rate": 0.0007365378252504566, + "loss": 2.6831, + "step": 2835 + }, + { + "epoch": 0.363008, + "grad_norm": 0.42440956830978394, + "learning_rate": 0.0007363551593092342, + "loss": 2.7492, + "step": 2836 + }, + { + "epoch": 0.363136, + "grad_norm": 0.46704912185668945, + "learning_rate": 0.0007361724527358194, + "loss": 2.6491, + "step": 2837 + }, + { + "epoch": 0.363264, + "grad_norm": 0.42691555619239807, + "learning_rate": 0.0007359897055616225, + "loss": 2.7905, + "step": 2838 + }, + { + "epoch": 0.363392, + "grad_norm": 0.44082528352737427, + "learning_rate": 0.0007358069178180592, + "loss": 2.6843, + "step": 2839 + }, + { + "epoch": 0.36352, + "grad_norm": 0.4492018520832062, + "learning_rate": 0.000735624089536553, + "loss": 2.5746, + "step": 2840 + }, + { + "epoch": 0.363648, + "grad_norm": 0.488733172416687, + "learning_rate": 0.0007354412207485342, + "loss": 2.7385, + "step": 2841 + }, + { + "epoch": 0.363776, + "grad_norm": 0.48301631212234497, + "learning_rate": 0.0007352583114854402, + "loss": 2.6548, + "step": 2842 + }, + { + "epoch": 0.363904, + "grad_norm": 0.4363223612308502, + "learning_rate": 0.000735075361778715, + "loss": 2.8096, + "step": 2843 + }, + { + "epoch": 0.364032, + "grad_norm": 0.4508669972419739, + "learning_rate": 0.0007348923716598099, + "loss": 2.5912, + "step": 2844 + }, + { + "epoch": 0.36416, + "grad_norm": 0.4479140639305115, + "learning_rate": 0.0007347093411601831, + "loss": 2.7602, + "step": 2845 + }, + { + "epoch": 0.364288, + "grad_norm": 0.45990341901779175, + "learning_rate": 0.0007345262703112996, + "loss": 2.6334, + "step": 2846 + }, + { + "epoch": 0.364416, + "grad_norm": 0.437555730342865, + "learning_rate": 0.0007343431591446312, + "loss": 2.7309, + "step": 2847 + }, + { + "epoch": 0.364544, + "grad_norm": 0.43353086709976196, + "learning_rate": 0.0007341600076916571, + "loss": 2.5682, + "step": 2848 + }, + { + "epoch": 0.364672, + "grad_norm": 0.4558120369911194, + "learning_rate": 0.0007339768159838631, + "loss": 2.7688, + "step": 2849 + }, + { + "epoch": 0.3648, + "grad_norm": 0.43889161944389343, + "learning_rate": 0.0007337935840527419, + "loss": 2.7453, + "step": 2850 + }, + { + "epoch": 0.364928, + "grad_norm": 0.42096880078315735, + "learning_rate": 0.000733610311929793, + "loss": 2.7217, + "step": 2851 + }, + { + "epoch": 0.365056, + "grad_norm": 0.4732609987258911, + "learning_rate": 0.0007334269996465235, + "loss": 2.6335, + "step": 2852 + }, + { + "epoch": 0.365184, + "grad_norm": 0.42193546891212463, + "learning_rate": 0.0007332436472344467, + "loss": 2.5948, + "step": 2853 + }, + { + "epoch": 0.365312, + "grad_norm": 0.42109912633895874, + "learning_rate": 0.0007330602547250827, + "loss": 2.7396, + "step": 2854 + }, + { + "epoch": 0.36544, + "grad_norm": 0.46621036529541016, + "learning_rate": 0.0007328768221499594, + "loss": 2.6781, + "step": 2855 + }, + { + "epoch": 0.365568, + "grad_norm": 0.4868764877319336, + "learning_rate": 0.0007326933495406103, + "loss": 2.6224, + "step": 2856 + }, + { + "epoch": 0.365696, + "grad_norm": 0.4586702883243561, + "learning_rate": 0.000732509836928577, + "loss": 2.6978, + "step": 2857 + }, + { + "epoch": 0.365824, + "grad_norm": 0.411920428276062, + "learning_rate": 0.0007323262843454071, + "loss": 2.6936, + "step": 2858 + }, + { + "epoch": 0.365952, + "grad_norm": 0.39197516441345215, + "learning_rate": 0.0007321426918226557, + "loss": 2.6188, + "step": 2859 + }, + { + "epoch": 0.36608, + "grad_norm": 0.43411049246788025, + "learning_rate": 0.0007319590593918843, + "loss": 2.7581, + "step": 2860 + }, + { + "epoch": 0.366208, + "grad_norm": 0.42723992466926575, + "learning_rate": 0.0007317753870846614, + "loss": 2.6543, + "step": 2861 + }, + { + "epoch": 0.366336, + "grad_norm": 0.4451085925102234, + "learning_rate": 0.0007315916749325625, + "loss": 2.7178, + "step": 2862 + }, + { + "epoch": 0.366464, + "grad_norm": 0.4549846351146698, + "learning_rate": 0.0007314079229671699, + "loss": 2.6689, + "step": 2863 + }, + { + "epoch": 0.366592, + "grad_norm": 0.4868393838405609, + "learning_rate": 0.0007312241312200727, + "loss": 2.7028, + "step": 2864 + }, + { + "epoch": 0.36672, + "grad_norm": 0.458278626203537, + "learning_rate": 0.0007310402997228667, + "loss": 2.7057, + "step": 2865 + }, + { + "epoch": 0.366848, + "grad_norm": 0.4463958442211151, + "learning_rate": 0.0007308564285071546, + "loss": 2.7198, + "step": 2866 + }, + { + "epoch": 0.366976, + "grad_norm": 0.4077272415161133, + "learning_rate": 0.0007306725176045464, + "loss": 2.6789, + "step": 2867 + }, + { + "epoch": 0.367104, + "grad_norm": 0.45734286308288574, + "learning_rate": 0.0007304885670466582, + "loss": 2.5775, + "step": 2868 + }, + { + "epoch": 0.367232, + "grad_norm": 0.45704734325408936, + "learning_rate": 0.000730304576865113, + "loss": 2.559, + "step": 2869 + }, + { + "epoch": 0.36736, + "grad_norm": 0.45791861414909363, + "learning_rate": 0.0007301205470915412, + "loss": 2.6165, + "step": 2870 + }, + { + "epoch": 0.367488, + "grad_norm": 0.45598602294921875, + "learning_rate": 0.0007299364777575796, + "loss": 2.6874, + "step": 2871 + }, + { + "epoch": 0.367616, + "grad_norm": 0.4427374303340912, + "learning_rate": 0.000729752368894872, + "loss": 2.6473, + "step": 2872 + }, + { + "epoch": 0.367744, + "grad_norm": 0.47263506054878235, + "learning_rate": 0.0007295682205350684, + "loss": 2.63, + "step": 2873 + }, + { + "epoch": 0.367872, + "grad_norm": 0.44672220945358276, + "learning_rate": 0.0007293840327098264, + "loss": 2.7183, + "step": 2874 + }, + { + "epoch": 0.368, + "grad_norm": 0.5286298990249634, + "learning_rate": 0.0007291998054508101, + "loss": 2.7039, + "step": 2875 + }, + { + "epoch": 0.368128, + "grad_norm": 0.4476739764213562, + "learning_rate": 0.0007290155387896899, + "loss": 2.7231, + "step": 2876 + }, + { + "epoch": 0.368256, + "grad_norm": 0.4423424005508423, + "learning_rate": 0.0007288312327581438, + "loss": 2.7611, + "step": 2877 + }, + { + "epoch": 0.368384, + "grad_norm": 0.4724349081516266, + "learning_rate": 0.0007286468873878559, + "loss": 2.6522, + "step": 2878 + }, + { + "epoch": 0.368512, + "grad_norm": 0.4661543071269989, + "learning_rate": 0.0007284625027105172, + "loss": 2.6822, + "step": 2879 + }, + { + "epoch": 0.36864, + "grad_norm": 0.4397793412208557, + "learning_rate": 0.0007282780787578258, + "loss": 2.6306, + "step": 2880 + }, + { + "epoch": 0.368768, + "grad_norm": 0.4324410855770111, + "learning_rate": 0.0007280936155614863, + "loss": 2.6658, + "step": 2881 + }, + { + "epoch": 0.368896, + "grad_norm": 0.4742600619792938, + "learning_rate": 0.0007279091131532101, + "loss": 2.6708, + "step": 2882 + }, + { + "epoch": 0.369024, + "grad_norm": 0.4657389223575592, + "learning_rate": 0.000727724571564715, + "loss": 2.7403, + "step": 2883 + }, + { + "epoch": 0.369152, + "grad_norm": 0.4656267762184143, + "learning_rate": 0.000727539990827726, + "loss": 2.74, + "step": 2884 + }, + { + "epoch": 0.36928, + "grad_norm": 0.47150757908821106, + "learning_rate": 0.0007273553709739749, + "loss": 2.7115, + "step": 2885 + }, + { + "epoch": 0.369408, + "grad_norm": 0.46897831559181213, + "learning_rate": 0.0007271707120351996, + "loss": 2.7115, + "step": 2886 + }, + { + "epoch": 0.369536, + "grad_norm": 0.4475855827331543, + "learning_rate": 0.0007269860140431454, + "loss": 2.7315, + "step": 2887 + }, + { + "epoch": 0.369664, + "grad_norm": 0.44628772139549255, + "learning_rate": 0.000726801277029564, + "loss": 2.6269, + "step": 2888 + }, + { + "epoch": 0.369792, + "grad_norm": 0.4667726159095764, + "learning_rate": 0.0007266165010262138, + "loss": 2.6426, + "step": 2889 + }, + { + "epoch": 0.36992, + "grad_norm": 0.4968041181564331, + "learning_rate": 0.0007264316860648598, + "loss": 2.6259, + "step": 2890 + }, + { + "epoch": 0.370048, + "grad_norm": 0.43608489632606506, + "learning_rate": 0.000726246832177274, + "loss": 2.6368, + "step": 2891 + }, + { + "epoch": 0.370176, + "grad_norm": 0.6808265447616577, + "learning_rate": 0.000726061939395235, + "loss": 2.7631, + "step": 2892 + }, + { + "epoch": 0.370304, + "grad_norm": 0.4517097771167755, + "learning_rate": 0.0007258770077505275, + "loss": 2.7017, + "step": 2893 + }, + { + "epoch": 0.370432, + "grad_norm": 0.4336215555667877, + "learning_rate": 0.000725692037274944, + "loss": 2.6377, + "step": 2894 + }, + { + "epoch": 0.37056, + "grad_norm": 0.44542887806892395, + "learning_rate": 0.0007255070280002829, + "loss": 2.6271, + "step": 2895 + }, + { + "epoch": 0.370688, + "grad_norm": 0.4608234167098999, + "learning_rate": 0.0007253219799583495, + "loss": 2.6437, + "step": 2896 + }, + { + "epoch": 0.370816, + "grad_norm": 0.46421608328819275, + "learning_rate": 0.0007251368931809553, + "loss": 2.6923, + "step": 2897 + }, + { + "epoch": 0.370944, + "grad_norm": 0.49241477251052856, + "learning_rate": 0.0007249517676999192, + "loss": 2.8135, + "step": 2898 + }, + { + "epoch": 0.371072, + "grad_norm": 0.4541221261024475, + "learning_rate": 0.0007247666035470666, + "loss": 2.618, + "step": 2899 + }, + { + "epoch": 0.3712, + "grad_norm": 0.5043959021568298, + "learning_rate": 0.0007245814007542289, + "loss": 2.5829, + "step": 2900 + }, + { + "epoch": 0.371328, + "grad_norm": 0.4726247191429138, + "learning_rate": 0.000724396159353245, + "loss": 2.6839, + "step": 2901 + }, + { + "epoch": 0.371456, + "grad_norm": 0.41919615864753723, + "learning_rate": 0.0007242108793759596, + "loss": 2.6465, + "step": 2902 + }, + { + "epoch": 0.371584, + "grad_norm": 0.4420366585254669, + "learning_rate": 0.0007240255608542251, + "loss": 2.7934, + "step": 2903 + }, + { + "epoch": 0.371712, + "grad_norm": 0.4282279312610626, + "learning_rate": 0.0007238402038198994, + "loss": 2.7897, + "step": 2904 + }, + { + "epoch": 0.37184, + "grad_norm": 0.49175578355789185, + "learning_rate": 0.0007236548083048478, + "loss": 2.7117, + "step": 2905 + }, + { + "epoch": 0.371968, + "grad_norm": 0.4456961750984192, + "learning_rate": 0.0007234693743409417, + "loss": 2.7211, + "step": 2906 + }, + { + "epoch": 0.372096, + "grad_norm": 0.43551579117774963, + "learning_rate": 0.0007232839019600594, + "loss": 2.6908, + "step": 2907 + }, + { + "epoch": 0.372224, + "grad_norm": 0.4621182382106781, + "learning_rate": 0.0007230983911940861, + "loss": 2.5761, + "step": 2908 + }, + { + "epoch": 0.372352, + "grad_norm": 0.4367345869541168, + "learning_rate": 0.0007229128420749127, + "loss": 2.6502, + "step": 2909 + }, + { + "epoch": 0.37248, + "grad_norm": 0.4206143617630005, + "learning_rate": 0.0007227272546344377, + "loss": 2.6542, + "step": 2910 + }, + { + "epoch": 0.372608, + "grad_norm": 0.4587569534778595, + "learning_rate": 0.0007225416289045654, + "loss": 2.796, + "step": 2911 + }, + { + "epoch": 0.372736, + "grad_norm": 0.5044446587562561, + "learning_rate": 0.0007223559649172073, + "loss": 2.5776, + "step": 2912 + }, + { + "epoch": 0.372864, + "grad_norm": 0.5307785868644714, + "learning_rate": 0.000722170262704281, + "loss": 2.7513, + "step": 2913 + }, + { + "epoch": 0.372992, + "grad_norm": 0.4770559072494507, + "learning_rate": 0.0007219845222977109, + "loss": 2.6889, + "step": 2914 + }, + { + "epoch": 0.37312, + "grad_norm": 0.45791053771972656, + "learning_rate": 0.000721798743729428, + "loss": 2.7711, + "step": 2915 + }, + { + "epoch": 0.373248, + "grad_norm": 0.42664042115211487, + "learning_rate": 0.0007216129270313697, + "loss": 2.6315, + "step": 2916 + }, + { + "epoch": 0.373376, + "grad_norm": 0.44343462586402893, + "learning_rate": 0.0007214270722354801, + "loss": 2.638, + "step": 2917 + }, + { + "epoch": 0.373504, + "grad_norm": 0.5459274649620056, + "learning_rate": 0.00072124117937371, + "loss": 2.6766, + "step": 2918 + }, + { + "epoch": 0.373632, + "grad_norm": 0.459365576505661, + "learning_rate": 0.000721055248478016, + "loss": 2.6698, + "step": 2919 + }, + { + "epoch": 0.37376, + "grad_norm": 0.45560142397880554, + "learning_rate": 0.0007208692795803622, + "loss": 2.7458, + "step": 2920 + }, + { + "epoch": 0.373888, + "grad_norm": 0.4459371566772461, + "learning_rate": 0.0007206832727127186, + "loss": 2.719, + "step": 2921 + }, + { + "epoch": 0.374016, + "grad_norm": 0.4461308717727661, + "learning_rate": 0.0007204972279070622, + "loss": 2.698, + "step": 2922 + }, + { + "epoch": 0.374144, + "grad_norm": 0.5080463290214539, + "learning_rate": 0.000720311145195376, + "loss": 2.6792, + "step": 2923 + }, + { + "epoch": 0.374272, + "grad_norm": 0.4494166076183319, + "learning_rate": 0.00072012502460965, + "loss": 2.7024, + "step": 2924 + }, + { + "epoch": 0.3744, + "grad_norm": 0.482312947511673, + "learning_rate": 0.0007199388661818802, + "loss": 2.6888, + "step": 2925 + }, + { + "epoch": 0.374528, + "grad_norm": 0.46038976311683655, + "learning_rate": 0.0007197526699440697, + "loss": 2.5363, + "step": 2926 + }, + { + "epoch": 0.374656, + "grad_norm": 0.4706070125102997, + "learning_rate": 0.0007195664359282275, + "loss": 2.676, + "step": 2927 + }, + { + "epoch": 0.374784, + "grad_norm": 0.47953522205352783, + "learning_rate": 0.0007193801641663697, + "loss": 2.6536, + "step": 2928 + }, + { + "epoch": 0.374912, + "grad_norm": 0.44674932956695557, + "learning_rate": 0.0007191938546905183, + "loss": 2.5701, + "step": 2929 + }, + { + "epoch": 0.37504, + "grad_norm": 0.41350120306015015, + "learning_rate": 0.0007190075075327021, + "loss": 2.7163, + "step": 2930 + }, + { + "epoch": 0.375168, + "grad_norm": 0.42401111125946045, + "learning_rate": 0.0007188211227249565, + "loss": 2.6569, + "step": 2931 + }, + { + "epoch": 0.375296, + "grad_norm": 0.44833990931510925, + "learning_rate": 0.0007186347002993232, + "loss": 2.6717, + "step": 2932 + }, + { + "epoch": 0.375424, + "grad_norm": 0.5048654675483704, + "learning_rate": 0.0007184482402878501, + "loss": 2.6927, + "step": 2933 + }, + { + "epoch": 0.375552, + "grad_norm": 0.4625113904476166, + "learning_rate": 0.0007182617427225922, + "loss": 2.7947, + "step": 2934 + }, + { + "epoch": 0.37568, + "grad_norm": 0.4666441082954407, + "learning_rate": 0.0007180752076356102, + "loss": 2.6553, + "step": 2935 + }, + { + "epoch": 0.375808, + "grad_norm": 0.4773557484149933, + "learning_rate": 0.0007178886350589721, + "loss": 2.7463, + "step": 2936 + }, + { + "epoch": 0.375936, + "grad_norm": 0.46159473061561584, + "learning_rate": 0.0007177020250247514, + "loss": 2.7508, + "step": 2937 + }, + { + "epoch": 0.376064, + "grad_norm": 0.41907113790512085, + "learning_rate": 0.0007175153775650289, + "loss": 2.7482, + "step": 2938 + }, + { + "epoch": 0.376192, + "grad_norm": 0.4789009988307953, + "learning_rate": 0.0007173286927118914, + "loss": 2.7249, + "step": 2939 + }, + { + "epoch": 0.37632, + "grad_norm": 0.4240191876888275, + "learning_rate": 0.000717141970497432, + "loss": 2.7629, + "step": 2940 + }, + { + "epoch": 0.376448, + "grad_norm": 0.42141878604888916, + "learning_rate": 0.0007169552109537507, + "loss": 2.6059, + "step": 2941 + }, + { + "epoch": 0.376576, + "grad_norm": 0.4773353040218353, + "learning_rate": 0.0007167684141129536, + "loss": 2.7464, + "step": 2942 + }, + { + "epoch": 0.376704, + "grad_norm": 0.4504489302635193, + "learning_rate": 0.0007165815800071529, + "loss": 2.7749, + "step": 2943 + }, + { + "epoch": 0.376832, + "grad_norm": 0.504988968372345, + "learning_rate": 0.0007163947086684676, + "loss": 2.6694, + "step": 2944 + }, + { + "epoch": 0.37696, + "grad_norm": 0.46867406368255615, + "learning_rate": 0.0007162078001290235, + "loss": 2.6159, + "step": 2945 + }, + { + "epoch": 0.377088, + "grad_norm": 0.489372193813324, + "learning_rate": 0.0007160208544209521, + "loss": 2.7308, + "step": 2946 + }, + { + "epoch": 0.377216, + "grad_norm": 0.4027650058269501, + "learning_rate": 0.0007158338715763913, + "loss": 2.6571, + "step": 2947 + }, + { + "epoch": 0.377344, + "grad_norm": 0.4530901312828064, + "learning_rate": 0.0007156468516274859, + "loss": 2.6746, + "step": 2948 + }, + { + "epoch": 0.377472, + "grad_norm": 0.4197593033313751, + "learning_rate": 0.0007154597946063867, + "loss": 2.6079, + "step": 2949 + }, + { + "epoch": 0.3776, + "grad_norm": 0.43104904890060425, + "learning_rate": 0.000715272700545251, + "loss": 2.739, + "step": 2950 + }, + { + "epoch": 0.377728, + "grad_norm": 0.44477078318595886, + "learning_rate": 0.0007150855694762424, + "loss": 2.6912, + "step": 2951 + }, + { + "epoch": 0.377856, + "grad_norm": 0.42638999223709106, + "learning_rate": 0.000714898401431531, + "loss": 2.6572, + "step": 2952 + }, + { + "epoch": 0.377984, + "grad_norm": 0.4674924612045288, + "learning_rate": 0.000714711196443293, + "loss": 2.7486, + "step": 2953 + }, + { + "epoch": 0.378112, + "grad_norm": 0.43492379784584045, + "learning_rate": 0.0007145239545437112, + "loss": 2.7881, + "step": 2954 + }, + { + "epoch": 0.37824, + "grad_norm": 0.435375839471817, + "learning_rate": 0.0007143366757649746, + "loss": 2.603, + "step": 2955 + }, + { + "epoch": 0.378368, + "grad_norm": 0.4790843427181244, + "learning_rate": 0.0007141493601392787, + "loss": 2.67, + "step": 2956 + }, + { + "epoch": 0.378496, + "grad_norm": 0.47786176204681396, + "learning_rate": 0.0007139620076988251, + "loss": 2.6035, + "step": 2957 + }, + { + "epoch": 0.378624, + "grad_norm": 0.45329487323760986, + "learning_rate": 0.0007137746184758218, + "loss": 2.7102, + "step": 2958 + }, + { + "epoch": 0.378752, + "grad_norm": 0.42998236417770386, + "learning_rate": 0.0007135871925024835, + "loss": 2.6512, + "step": 2959 + }, + { + "epoch": 0.37888, + "grad_norm": 0.4542911648750305, + "learning_rate": 0.0007133997298110307, + "loss": 2.7214, + "step": 2960 + }, + { + "epoch": 0.379008, + "grad_norm": 0.4886983036994934, + "learning_rate": 0.0007132122304336903, + "loss": 2.778, + "step": 2961 + }, + { + "epoch": 0.379136, + "grad_norm": 0.4561481177806854, + "learning_rate": 0.0007130246944026956, + "loss": 2.6253, + "step": 2962 + }, + { + "epoch": 0.379264, + "grad_norm": 0.47425973415374756, + "learning_rate": 0.0007128371217502867, + "loss": 2.6121, + "step": 2963 + }, + { + "epoch": 0.379392, + "grad_norm": 0.45755621790885925, + "learning_rate": 0.000712649512508709, + "loss": 2.6053, + "step": 2964 + }, + { + "epoch": 0.37952, + "grad_norm": 0.4175988733768463, + "learning_rate": 0.0007124618667102147, + "loss": 2.589, + "step": 2965 + }, + { + "epoch": 0.379648, + "grad_norm": 0.43995943665504456, + "learning_rate": 0.0007122741843870626, + "loss": 2.5877, + "step": 2966 + }, + { + "epoch": 0.379776, + "grad_norm": 0.5148723721504211, + "learning_rate": 0.0007120864655715171, + "loss": 2.6434, + "step": 2967 + }, + { + "epoch": 0.379904, + "grad_norm": 0.4607153832912445, + "learning_rate": 0.0007118987102958498, + "loss": 2.7572, + "step": 2968 + }, + { + "epoch": 0.380032, + "grad_norm": 0.4861503541469574, + "learning_rate": 0.0007117109185923373, + "loss": 2.6574, + "step": 2969 + }, + { + "epoch": 0.38016, + "grad_norm": 0.4536001682281494, + "learning_rate": 0.0007115230904932639, + "loss": 2.7858, + "step": 2970 + }, + { + "epoch": 0.380288, + "grad_norm": 0.44380247592926025, + "learning_rate": 0.0007113352260309188, + "loss": 2.6443, + "step": 2971 + }, + { + "epoch": 0.380416, + "grad_norm": 0.44570493698120117, + "learning_rate": 0.0007111473252375983, + "loss": 2.6224, + "step": 2972 + }, + { + "epoch": 0.380544, + "grad_norm": 0.40681371092796326, + "learning_rate": 0.0007109593881456048, + "loss": 2.6423, + "step": 2973 + }, + { + "epoch": 0.380672, + "grad_norm": 0.5353396534919739, + "learning_rate": 0.000710771414787247, + "loss": 2.721, + "step": 2974 + }, + { + "epoch": 0.3808, + "grad_norm": 0.4317675232887268, + "learning_rate": 0.0007105834051948395, + "loss": 2.6092, + "step": 2975 + }, + { + "epoch": 0.380928, + "grad_norm": 0.54416424036026, + "learning_rate": 0.0007103953594007029, + "loss": 2.6467, + "step": 2976 + }, + { + "epoch": 0.381056, + "grad_norm": 0.4674355685710907, + "learning_rate": 0.0007102072774371654, + "loss": 2.5982, + "step": 2977 + }, + { + "epoch": 0.381184, + "grad_norm": 0.5025956034660339, + "learning_rate": 0.0007100191593365599, + "loss": 2.8008, + "step": 2978 + }, + { + "epoch": 0.381312, + "grad_norm": 0.48335424065589905, + "learning_rate": 0.000709831005131226, + "loss": 2.729, + "step": 2979 + }, + { + "epoch": 0.38144, + "grad_norm": 0.44937407970428467, + "learning_rate": 0.00070964281485351, + "loss": 2.6083, + "step": 2980 + }, + { + "epoch": 0.381568, + "grad_norm": 0.4208657443523407, + "learning_rate": 0.0007094545885357635, + "loss": 2.6746, + "step": 2981 + }, + { + "epoch": 0.381696, + "grad_norm": 0.46431466937065125, + "learning_rate": 0.0007092663262103451, + "loss": 2.5818, + "step": 2982 + }, + { + "epoch": 0.381824, + "grad_norm": 0.4556119441986084, + "learning_rate": 0.0007090780279096192, + "loss": 2.6836, + "step": 2983 + }, + { + "epoch": 0.381952, + "grad_norm": 0.4427129030227661, + "learning_rate": 0.0007088896936659566, + "loss": 2.6835, + "step": 2984 + }, + { + "epoch": 0.38208, + "grad_norm": 0.4567842185497284, + "learning_rate": 0.0007087013235117338, + "loss": 2.6828, + "step": 2985 + }, + { + "epoch": 0.382208, + "grad_norm": 0.4658835530281067, + "learning_rate": 0.0007085129174793342, + "loss": 2.6008, + "step": 2986 + }, + { + "epoch": 0.382336, + "grad_norm": 0.47157159447669983, + "learning_rate": 0.0007083244756011465, + "loss": 2.6735, + "step": 2987 + }, + { + "epoch": 0.382464, + "grad_norm": 0.5626171827316284, + "learning_rate": 0.0007081359979095666, + "loss": 2.5689, + "step": 2988 + }, + { + "epoch": 0.382592, + "grad_norm": 0.49118125438690186, + "learning_rate": 0.0007079474844369957, + "loss": 2.7756, + "step": 2989 + }, + { + "epoch": 0.38272, + "grad_norm": 0.47579678893089294, + "learning_rate": 0.0007077589352158415, + "loss": 2.808, + "step": 2990 + }, + { + "epoch": 0.382848, + "grad_norm": 0.4123162627220154, + "learning_rate": 0.0007075703502785177, + "loss": 2.5969, + "step": 2991 + }, + { + "epoch": 0.382976, + "grad_norm": 0.4646452069282532, + "learning_rate": 0.0007073817296574444, + "loss": 2.7418, + "step": 2992 + }, + { + "epoch": 0.383104, + "grad_norm": 0.42933544516563416, + "learning_rate": 0.0007071930733850476, + "loss": 2.7268, + "step": 2993 + }, + { + "epoch": 0.383232, + "grad_norm": 0.4115656614303589, + "learning_rate": 0.0007070043814937594, + "loss": 2.5622, + "step": 2994 + }, + { + "epoch": 0.38336, + "grad_norm": 0.4414701461791992, + "learning_rate": 0.0007068156540160182, + "loss": 2.6317, + "step": 2995 + }, + { + "epoch": 0.383488, + "grad_norm": 0.4508860111236572, + "learning_rate": 0.0007066268909842687, + "loss": 2.6221, + "step": 2996 + }, + { + "epoch": 0.383616, + "grad_norm": 0.4824790954589844, + "learning_rate": 0.000706438092430961, + "loss": 2.7187, + "step": 2997 + }, + { + "epoch": 0.383744, + "grad_norm": 0.44146856665611267, + "learning_rate": 0.000706249258388552, + "loss": 2.6404, + "step": 2998 + }, + { + "epoch": 0.383872, + "grad_norm": 0.47207847237586975, + "learning_rate": 0.0007060603888895045, + "loss": 2.7084, + "step": 2999 + }, + { + "epoch": 0.384, + "grad_norm": 0.4444594383239746, + "learning_rate": 0.0007058714839662874, + "loss": 2.6619, + "step": 3000 + }, + { + "epoch": 0.384128, + "grad_norm": 0.49057066440582275, + "learning_rate": 0.0007056825436513753, + "loss": 2.6441, + "step": 3001 + }, + { + "epoch": 0.384256, + "grad_norm": 0.44169387221336365, + "learning_rate": 0.0007054935679772496, + "loss": 2.6072, + "step": 3002 + }, + { + "epoch": 0.384384, + "grad_norm": 0.4131368100643158, + "learning_rate": 0.0007053045569763973, + "loss": 2.7261, + "step": 3003 + }, + { + "epoch": 0.384512, + "grad_norm": 0.3924567699432373, + "learning_rate": 0.0007051155106813114, + "loss": 2.7067, + "step": 3004 + }, + { + "epoch": 0.38464, + "grad_norm": 0.7526036500930786, + "learning_rate": 0.0007049264291244914, + "loss": 2.6417, + "step": 3005 + }, + { + "epoch": 0.384768, + "grad_norm": 0.4586635231971741, + "learning_rate": 0.0007047373123384426, + "loss": 2.7561, + "step": 3006 + }, + { + "epoch": 0.384896, + "grad_norm": 0.4140767455101013, + "learning_rate": 0.0007045481603556763, + "loss": 2.6417, + "step": 3007 + }, + { + "epoch": 0.385024, + "grad_norm": 0.503156840801239, + "learning_rate": 0.0007043589732087098, + "loss": 2.7114, + "step": 3008 + }, + { + "epoch": 0.385152, + "grad_norm": 0.4833620488643646, + "learning_rate": 0.0007041697509300667, + "loss": 2.6877, + "step": 3009 + }, + { + "epoch": 0.38528, + "grad_norm": 0.45955905318260193, + "learning_rate": 0.0007039804935522765, + "loss": 2.7732, + "step": 3010 + }, + { + "epoch": 0.385408, + "grad_norm": 0.42978236079216003, + "learning_rate": 0.0007037912011078749, + "loss": 2.6871, + "step": 3011 + }, + { + "epoch": 0.385536, + "grad_norm": 0.4695398211479187, + "learning_rate": 0.000703601873629403, + "loss": 2.7181, + "step": 3012 + }, + { + "epoch": 0.385664, + "grad_norm": 0.45631322264671326, + "learning_rate": 0.0007034125111494087, + "loss": 2.7046, + "step": 3013 + }, + { + "epoch": 0.385792, + "grad_norm": 0.48728179931640625, + "learning_rate": 0.0007032231137004457, + "loss": 2.661, + "step": 3014 + }, + { + "epoch": 0.38592, + "grad_norm": 0.4507933259010315, + "learning_rate": 0.0007030336813150733, + "loss": 2.6437, + "step": 3015 + }, + { + "epoch": 0.386048, + "grad_norm": 0.5023179054260254, + "learning_rate": 0.0007028442140258575, + "loss": 2.765, + "step": 3016 + }, + { + "epoch": 0.386176, + "grad_norm": 0.46155527234077454, + "learning_rate": 0.0007026547118653696, + "loss": 2.5889, + "step": 3017 + }, + { + "epoch": 0.386304, + "grad_norm": 0.48626676201820374, + "learning_rate": 0.0007024651748661875, + "loss": 2.6087, + "step": 3018 + }, + { + "epoch": 0.386432, + "grad_norm": 0.4624722898006439, + "learning_rate": 0.0007022756030608946, + "loss": 2.7739, + "step": 3019 + }, + { + "epoch": 0.38656, + "grad_norm": 0.45596858859062195, + "learning_rate": 0.0007020859964820805, + "loss": 2.6465, + "step": 3020 + }, + { + "epoch": 0.386688, + "grad_norm": 0.47300103306770325, + "learning_rate": 0.000701896355162341, + "loss": 2.705, + "step": 3021 + }, + { + "epoch": 0.386816, + "grad_norm": 0.490530401468277, + "learning_rate": 0.0007017066791342772, + "loss": 2.7155, + "step": 3022 + }, + { + "epoch": 0.386944, + "grad_norm": 0.45950818061828613, + "learning_rate": 0.0007015169684304969, + "loss": 2.6555, + "step": 3023 + }, + { + "epoch": 0.387072, + "grad_norm": 0.4722858667373657, + "learning_rate": 0.0007013272230836138, + "loss": 2.7296, + "step": 3024 + }, + { + "epoch": 0.3872, + "grad_norm": 0.45955756306648254, + "learning_rate": 0.000701137443126247, + "loss": 2.7558, + "step": 3025 + }, + { + "epoch": 0.387328, + "grad_norm": 0.5439246296882629, + "learning_rate": 0.0007009476285910218, + "loss": 2.6745, + "step": 3026 + }, + { + "epoch": 0.387456, + "grad_norm": 0.44840607047080994, + "learning_rate": 0.0007007577795105696, + "loss": 2.6904, + "step": 3027 + }, + { + "epoch": 0.387584, + "grad_norm": 0.5287879705429077, + "learning_rate": 0.0007005678959175279, + "loss": 2.7669, + "step": 3028 + }, + { + "epoch": 0.387712, + "grad_norm": 0.431357204914093, + "learning_rate": 0.0007003779778445397, + "loss": 2.6911, + "step": 3029 + }, + { + "epoch": 0.38784, + "grad_norm": 0.48556748032569885, + "learning_rate": 0.0007001880253242541, + "loss": 2.684, + "step": 3030 + }, + { + "epoch": 0.387968, + "grad_norm": 0.4537012577056885, + "learning_rate": 0.0006999980383893261, + "loss": 2.6251, + "step": 3031 + }, + { + "epoch": 0.388096, + "grad_norm": 0.48545727133750916, + "learning_rate": 0.0006998080170724167, + "loss": 2.7734, + "step": 3032 + }, + { + "epoch": 0.388224, + "grad_norm": 0.4682837426662445, + "learning_rate": 0.0006996179614061929, + "loss": 2.6965, + "step": 3033 + }, + { + "epoch": 0.388352, + "grad_norm": 0.4412662088871002, + "learning_rate": 0.0006994278714233273, + "loss": 2.6237, + "step": 3034 + }, + { + "epoch": 0.38848, + "grad_norm": 0.42491602897644043, + "learning_rate": 0.0006992377471564987, + "loss": 2.6818, + "step": 3035 + }, + { + "epoch": 0.388608, + "grad_norm": 0.44887933135032654, + "learning_rate": 0.0006990475886383915, + "loss": 2.6274, + "step": 3036 + }, + { + "epoch": 0.388736, + "grad_norm": 0.45977821946144104, + "learning_rate": 0.0006988573959016963, + "loss": 2.7543, + "step": 3037 + }, + { + "epoch": 0.388864, + "grad_norm": 0.44426262378692627, + "learning_rate": 0.0006986671689791094, + "loss": 2.603, + "step": 3038 + }, + { + "epoch": 0.388992, + "grad_norm": 0.46926644444465637, + "learning_rate": 0.0006984769079033331, + "loss": 2.7566, + "step": 3039 + }, + { + "epoch": 0.38912, + "grad_norm": 0.4359223246574402, + "learning_rate": 0.0006982866127070753, + "loss": 2.6363, + "step": 3040 + }, + { + "epoch": 0.389248, + "grad_norm": 0.5087350606918335, + "learning_rate": 0.0006980962834230499, + "loss": 2.732, + "step": 3041 + }, + { + "epoch": 0.389376, + "grad_norm": 0.48416972160339355, + "learning_rate": 0.0006979059200839771, + "loss": 2.6949, + "step": 3042 + }, + { + "epoch": 0.389504, + "grad_norm": 0.4825270473957062, + "learning_rate": 0.0006977155227225822, + "loss": 2.691, + "step": 3043 + }, + { + "epoch": 0.389632, + "grad_norm": 0.4746887981891632, + "learning_rate": 0.0006975250913715968, + "loss": 2.646, + "step": 3044 + }, + { + "epoch": 0.38976, + "grad_norm": 0.4818806052207947, + "learning_rate": 0.0006973346260637583, + "loss": 2.7292, + "step": 3045 + }, + { + "epoch": 0.389888, + "grad_norm": 0.5424981117248535, + "learning_rate": 0.0006971441268318099, + "loss": 2.5768, + "step": 3046 + }, + { + "epoch": 0.390016, + "grad_norm": 0.5124492645263672, + "learning_rate": 0.0006969535937085005, + "loss": 2.7908, + "step": 3047 + }, + { + "epoch": 0.390144, + "grad_norm": 0.47376492619514465, + "learning_rate": 0.0006967630267265853, + "loss": 2.8087, + "step": 3048 + }, + { + "epoch": 0.390272, + "grad_norm": 0.47444671392440796, + "learning_rate": 0.0006965724259188246, + "loss": 2.6741, + "step": 3049 + }, + { + "epoch": 0.3904, + "grad_norm": 0.4937334954738617, + "learning_rate": 0.0006963817913179849, + "loss": 2.6158, + "step": 3050 + }, + { + "epoch": 0.390528, + "grad_norm": 0.43661990761756897, + "learning_rate": 0.0006961911229568387, + "loss": 2.7384, + "step": 3051 + }, + { + "epoch": 0.390656, + "grad_norm": 0.43692028522491455, + "learning_rate": 0.0006960004208681642, + "loss": 2.7314, + "step": 3052 + }, + { + "epoch": 0.390784, + "grad_norm": 0.4787447154521942, + "learning_rate": 0.000695809685084745, + "loss": 2.7608, + "step": 3053 + }, + { + "epoch": 0.390912, + "grad_norm": 0.4433353543281555, + "learning_rate": 0.0006956189156393708, + "loss": 2.5653, + "step": 3054 + }, + { + "epoch": 0.39104, + "grad_norm": 0.6088268756866455, + "learning_rate": 0.0006954281125648373, + "loss": 2.7287, + "step": 3055 + }, + { + "epoch": 0.391168, + "grad_norm": 0.4373394250869751, + "learning_rate": 0.0006952372758939457, + "loss": 2.7053, + "step": 3056 + }, + { + "epoch": 0.391296, + "grad_norm": 0.4539661109447479, + "learning_rate": 0.0006950464056595029, + "loss": 2.6521, + "step": 3057 + }, + { + "epoch": 0.391424, + "grad_norm": 0.5094650983810425, + "learning_rate": 0.0006948555018943219, + "loss": 2.7503, + "step": 3058 + }, + { + "epoch": 0.391552, + "grad_norm": 0.4541882872581482, + "learning_rate": 0.0006946645646312211, + "loss": 2.5551, + "step": 3059 + }, + { + "epoch": 0.39168, + "grad_norm": 0.44633573293685913, + "learning_rate": 0.0006944735939030249, + "loss": 2.7314, + "step": 3060 + }, + { + "epoch": 0.391808, + "grad_norm": 0.44832292199134827, + "learning_rate": 0.0006942825897425633, + "loss": 2.69, + "step": 3061 + }, + { + "epoch": 0.391936, + "grad_norm": 0.9240698218345642, + "learning_rate": 0.0006940915521826723, + "loss": 2.6757, + "step": 3062 + }, + { + "epoch": 0.392064, + "grad_norm": 0.4975711703300476, + "learning_rate": 0.0006939004812561933, + "loss": 2.6912, + "step": 3063 + }, + { + "epoch": 0.392192, + "grad_norm": 0.5434896349906921, + "learning_rate": 0.0006937093769959736, + "loss": 2.6477, + "step": 3064 + }, + { + "epoch": 0.39232, + "grad_norm": 0.4612567126750946, + "learning_rate": 0.0006935182394348664, + "loss": 2.5692, + "step": 3065 + }, + { + "epoch": 0.392448, + "grad_norm": 0.4792902171611786, + "learning_rate": 0.0006933270686057307, + "loss": 2.6376, + "step": 3066 + }, + { + "epoch": 0.392576, + "grad_norm": 0.4379271864891052, + "learning_rate": 0.0006931358645414304, + "loss": 2.7968, + "step": 3067 + }, + { + "epoch": 0.392704, + "grad_norm": 0.45561280846595764, + "learning_rate": 0.0006929446272748359, + "loss": 2.5993, + "step": 3068 + }, + { + "epoch": 0.392832, + "grad_norm": 0.45777419209480286, + "learning_rate": 0.0006927533568388231, + "loss": 2.7295, + "step": 3069 + }, + { + "epoch": 0.39296, + "grad_norm": 0.47148066759109497, + "learning_rate": 0.000692562053266274, + "loss": 2.6419, + "step": 3070 + }, + { + "epoch": 0.393088, + "grad_norm": 0.551578164100647, + "learning_rate": 0.0006923707165900754, + "loss": 2.6372, + "step": 3071 + }, + { + "epoch": 0.393216, + "grad_norm": 0.41483014822006226, + "learning_rate": 0.0006921793468431203, + "loss": 2.6217, + "step": 3072 + }, + { + "epoch": 0.393344, + "grad_norm": 0.4473557770252228, + "learning_rate": 0.0006919879440583077, + "loss": 2.5983, + "step": 3073 + }, + { + "epoch": 0.393472, + "grad_norm": 0.4438343346118927, + "learning_rate": 0.0006917965082685418, + "loss": 2.6659, + "step": 3074 + }, + { + "epoch": 0.3936, + "grad_norm": 0.48738205432891846, + "learning_rate": 0.0006916050395067327, + "loss": 2.6699, + "step": 3075 + }, + { + "epoch": 0.393728, + "grad_norm": 0.427270770072937, + "learning_rate": 0.0006914135378057959, + "loss": 2.6565, + "step": 3076 + }, + { + "epoch": 0.393856, + "grad_norm": 0.47675907611846924, + "learning_rate": 0.0006912220031986531, + "loss": 2.7341, + "step": 3077 + }, + { + "epoch": 0.393984, + "grad_norm": 0.5815351009368896, + "learning_rate": 0.0006910304357182307, + "loss": 2.7111, + "step": 3078 + }, + { + "epoch": 0.394112, + "grad_norm": 0.44046759605407715, + "learning_rate": 0.0006908388353974622, + "loss": 2.5889, + "step": 3079 + }, + { + "epoch": 0.39424, + "grad_norm": 0.45098191499710083, + "learning_rate": 0.0006906472022692854, + "loss": 2.6483, + "step": 3080 + }, + { + "epoch": 0.394368, + "grad_norm": 0.4653068780899048, + "learning_rate": 0.0006904555363666442, + "loss": 2.6705, + "step": 3081 + }, + { + "epoch": 0.394496, + "grad_norm": 0.5028269290924072, + "learning_rate": 0.0006902638377224886, + "loss": 2.7503, + "step": 3082 + }, + { + "epoch": 0.394624, + "grad_norm": 0.436442106962204, + "learning_rate": 0.0006900721063697732, + "loss": 2.6599, + "step": 3083 + }, + { + "epoch": 0.394752, + "grad_norm": 0.43633177876472473, + "learning_rate": 0.0006898803423414595, + "loss": 2.627, + "step": 3084 + }, + { + "epoch": 0.39488, + "grad_norm": 0.4394446611404419, + "learning_rate": 0.0006896885456705137, + "loss": 2.7117, + "step": 3085 + }, + { + "epoch": 0.395008, + "grad_norm": 0.4389834702014923, + "learning_rate": 0.0006894967163899076, + "loss": 2.7361, + "step": 3086 + }, + { + "epoch": 0.395136, + "grad_norm": 0.4698423743247986, + "learning_rate": 0.0006893048545326192, + "loss": 2.7584, + "step": 3087 + }, + { + "epoch": 0.395264, + "grad_norm": 0.4919544458389282, + "learning_rate": 0.0006891129601316318, + "loss": 2.7499, + "step": 3088 + }, + { + "epoch": 0.395392, + "grad_norm": 0.4746415317058563, + "learning_rate": 0.0006889210332199343, + "loss": 2.8341, + "step": 3089 + }, + { + "epoch": 0.39552, + "grad_norm": 0.4418022334575653, + "learning_rate": 0.0006887290738305207, + "loss": 2.7618, + "step": 3090 + }, + { + "epoch": 0.395648, + "grad_norm": 0.48406752943992615, + "learning_rate": 0.0006885370819963917, + "loss": 2.6371, + "step": 3091 + }, + { + "epoch": 0.395776, + "grad_norm": 0.4679573178291321, + "learning_rate": 0.0006883450577505523, + "loss": 2.6296, + "step": 3092 + }, + { + "epoch": 0.395904, + "grad_norm": 0.4658784568309784, + "learning_rate": 0.0006881530011260142, + "loss": 2.7328, + "step": 3093 + }, + { + "epoch": 0.396032, + "grad_norm": 0.4444080591201782, + "learning_rate": 0.0006879609121557937, + "loss": 2.6079, + "step": 3094 + }, + { + "epoch": 0.39616, + "grad_norm": 0.5271738171577454, + "learning_rate": 0.0006877687908729137, + "loss": 2.7502, + "step": 3095 + }, + { + "epoch": 0.396288, + "grad_norm": 0.47427046298980713, + "learning_rate": 0.0006875766373104016, + "loss": 2.6855, + "step": 3096 + }, + { + "epoch": 0.396416, + "grad_norm": 0.45691362023353577, + "learning_rate": 0.0006873844515012908, + "loss": 2.5709, + "step": 3097 + }, + { + "epoch": 0.396544, + "grad_norm": 0.4308910071849823, + "learning_rate": 0.0006871922334786205, + "loss": 2.6256, + "step": 3098 + }, + { + "epoch": 0.396672, + "grad_norm": 0.5334233045578003, + "learning_rate": 0.0006869999832754355, + "loss": 2.7692, + "step": 3099 + }, + { + "epoch": 0.3968, + "grad_norm": 0.45469143986701965, + "learning_rate": 0.0006868077009247852, + "loss": 2.718, + "step": 3100 + }, + { + "epoch": 0.396928, + "grad_norm": 0.4555945098400116, + "learning_rate": 0.0006866153864597253, + "loss": 2.6914, + "step": 3101 + }, + { + "epoch": 0.397056, + "grad_norm": 0.47432318329811096, + "learning_rate": 0.0006864230399133173, + "loss": 2.6651, + "step": 3102 + }, + { + "epoch": 0.397184, + "grad_norm": 0.4350561201572418, + "learning_rate": 0.0006862306613186275, + "loss": 2.768, + "step": 3103 + }, + { + "epoch": 0.397312, + "grad_norm": 0.4356270134449005, + "learning_rate": 0.0006860382507087279, + "loss": 2.7308, + "step": 3104 + }, + { + "epoch": 0.39744, + "grad_norm": 0.46051275730133057, + "learning_rate": 0.0006858458081166964, + "loss": 2.7458, + "step": 3105 + }, + { + "epoch": 0.397568, + "grad_norm": 0.4567035734653473, + "learning_rate": 0.0006856533335756158, + "loss": 2.6572, + "step": 3106 + }, + { + "epoch": 0.397696, + "grad_norm": 0.4817090630531311, + "learning_rate": 0.0006854608271185752, + "loss": 2.6621, + "step": 3107 + }, + { + "epoch": 0.397824, + "grad_norm": 0.5238298773765564, + "learning_rate": 0.0006852682887786681, + "loss": 2.7617, + "step": 3108 + }, + { + "epoch": 0.397952, + "grad_norm": 0.4467407763004303, + "learning_rate": 0.0006850757185889945, + "loss": 2.7268, + "step": 3109 + }, + { + "epoch": 0.39808, + "grad_norm": 0.46863090991973877, + "learning_rate": 0.0006848831165826591, + "loss": 2.7144, + "step": 3110 + }, + { + "epoch": 0.398208, + "grad_norm": 0.5075850486755371, + "learning_rate": 0.0006846904827927728, + "loss": 2.6608, + "step": 3111 + }, + { + "epoch": 0.398336, + "grad_norm": 0.42297032475471497, + "learning_rate": 0.0006844978172524513, + "loss": 2.7404, + "step": 3112 + }, + { + "epoch": 0.398464, + "grad_norm": 0.4591881036758423, + "learning_rate": 0.0006843051199948161, + "loss": 2.6706, + "step": 3113 + }, + { + "epoch": 0.398592, + "grad_norm": 0.4692125618457794, + "learning_rate": 0.0006841123910529943, + "loss": 2.7811, + "step": 3114 + }, + { + "epoch": 0.39872, + "grad_norm": 0.41288280487060547, + "learning_rate": 0.0006839196304601179, + "loss": 2.6966, + "step": 3115 + }, + { + "epoch": 0.398848, + "grad_norm": 0.42315611243247986, + "learning_rate": 0.0006837268382493248, + "loss": 2.7039, + "step": 3116 + }, + { + "epoch": 0.398976, + "grad_norm": 0.44952934980392456, + "learning_rate": 0.0006835340144537584, + "loss": 2.7395, + "step": 3117 + }, + { + "epoch": 0.399104, + "grad_norm": 0.43882644176483154, + "learning_rate": 0.0006833411591065673, + "loss": 2.772, + "step": 3118 + }, + { + "epoch": 0.399232, + "grad_norm": 0.43627554178237915, + "learning_rate": 0.0006831482722409052, + "loss": 2.7421, + "step": 3119 + }, + { + "epoch": 0.39936, + "grad_norm": 0.4241209626197815, + "learning_rate": 0.0006829553538899321, + "loss": 2.687, + "step": 3120 + }, + { + "epoch": 0.399488, + "grad_norm": 0.45065757632255554, + "learning_rate": 0.0006827624040868125, + "loss": 2.6928, + "step": 3121 + }, + { + "epoch": 0.399616, + "grad_norm": 0.42040497064590454, + "learning_rate": 0.0006825694228647169, + "loss": 2.7798, + "step": 3122 + }, + { + "epoch": 0.399744, + "grad_norm": 0.45668095350265503, + "learning_rate": 0.0006823764102568211, + "loss": 2.6796, + "step": 3123 + }, + { + "epoch": 0.399872, + "grad_norm": 0.5343291759490967, + "learning_rate": 0.000682183366296306, + "loss": 2.6856, + "step": 3124 + }, + { + "epoch": 0.4, + "grad_norm": 0.4991387724876404, + "learning_rate": 0.0006819902910163581, + "loss": 2.7092, + "step": 3125 + }, + { + "epoch": 0.400128, + "grad_norm": 0.4771918058395386, + "learning_rate": 0.0006817971844501695, + "loss": 2.7749, + "step": 3126 + }, + { + "epoch": 0.400256, + "grad_norm": 0.43733036518096924, + "learning_rate": 0.0006816040466309372, + "loss": 2.6654, + "step": 3127 + }, + { + "epoch": 0.400384, + "grad_norm": 0.5161102414131165, + "learning_rate": 0.0006814108775918641, + "loss": 2.6251, + "step": 3128 + }, + { + "epoch": 0.400512, + "grad_norm": 0.47682568430900574, + "learning_rate": 0.0006812176773661579, + "loss": 2.5705, + "step": 3129 + }, + { + "epoch": 0.40064, + "grad_norm": 0.5132142901420593, + "learning_rate": 0.0006810244459870322, + "loss": 2.6429, + "step": 3130 + }, + { + "epoch": 0.400768, + "grad_norm": 0.516671359539032, + "learning_rate": 0.0006808311834877056, + "loss": 2.7309, + "step": 3131 + }, + { + "epoch": 0.400896, + "grad_norm": 0.44640886783599854, + "learning_rate": 0.0006806378899014023, + "loss": 2.6606, + "step": 3132 + }, + { + "epoch": 0.401024, + "grad_norm": 0.4475449323654175, + "learning_rate": 0.0006804445652613514, + "loss": 2.6903, + "step": 3133 + }, + { + "epoch": 0.401152, + "grad_norm": 0.5599510669708252, + "learning_rate": 0.0006802512096007878, + "loss": 2.6635, + "step": 3134 + }, + { + "epoch": 0.40128, + "grad_norm": 0.5951410531997681, + "learning_rate": 0.0006800578229529518, + "loss": 2.7789, + "step": 3135 + }, + { + "epoch": 0.401408, + "grad_norm": 0.4984765350818634, + "learning_rate": 0.0006798644053510886, + "loss": 2.7694, + "step": 3136 + }, + { + "epoch": 0.401536, + "grad_norm": 0.4956558048725128, + "learning_rate": 0.0006796709568284488, + "loss": 2.7605, + "step": 3137 + }, + { + "epoch": 0.401664, + "grad_norm": 0.45814016461372375, + "learning_rate": 0.0006794774774182887, + "loss": 2.7111, + "step": 3138 + }, + { + "epoch": 0.401792, + "grad_norm": 0.4692833125591278, + "learning_rate": 0.0006792839671538695, + "loss": 2.7554, + "step": 3139 + }, + { + "epoch": 0.40192, + "grad_norm": 0.5296211242675781, + "learning_rate": 0.0006790904260684581, + "loss": 2.6778, + "step": 3140 + }, + { + "epoch": 0.402048, + "grad_norm": 0.5093342661857605, + "learning_rate": 0.0006788968541953261, + "loss": 2.7135, + "step": 3141 + }, + { + "epoch": 0.402176, + "grad_norm": 0.43689438700675964, + "learning_rate": 0.0006787032515677509, + "loss": 2.6697, + "step": 3142 + }, + { + "epoch": 0.402304, + "grad_norm": 0.4528839588165283, + "learning_rate": 0.0006785096182190151, + "loss": 2.74, + "step": 3143 + }, + { + "epoch": 0.402432, + "grad_norm": 0.6912872195243835, + "learning_rate": 0.0006783159541824065, + "loss": 2.7831, + "step": 3144 + }, + { + "epoch": 0.40256, + "grad_norm": 0.5493539571762085, + "learning_rate": 0.0006781222594912181, + "loss": 2.6899, + "step": 3145 + }, + { + "epoch": 0.402688, + "grad_norm": 0.4472182095050812, + "learning_rate": 0.0006779285341787484, + "loss": 2.7043, + "step": 3146 + }, + { + "epoch": 0.402816, + "grad_norm": 0.6814253330230713, + "learning_rate": 0.0006777347782783008, + "loss": 2.5738, + "step": 3147 + }, + { + "epoch": 0.402944, + "grad_norm": 0.5127127766609192, + "learning_rate": 0.0006775409918231843, + "loss": 2.6396, + "step": 3148 + }, + { + "epoch": 0.403072, + "grad_norm": 0.4704344868659973, + "learning_rate": 0.0006773471748467133, + "loss": 2.6562, + "step": 3149 + }, + { + "epoch": 0.4032, + "grad_norm": 0.5241386890411377, + "learning_rate": 0.000677153327382207, + "loss": 2.6525, + "step": 3150 + }, + { + "epoch": 0.403328, + "grad_norm": 0.5215307474136353, + "learning_rate": 0.0006769594494629897, + "loss": 2.7353, + "step": 3151 + }, + { + "epoch": 0.403456, + "grad_norm": 0.47325026988983154, + "learning_rate": 0.0006767655411223916, + "loss": 2.5867, + "step": 3152 + }, + { + "epoch": 0.403584, + "grad_norm": 0.476476788520813, + "learning_rate": 0.0006765716023937479, + "loss": 2.7852, + "step": 3153 + }, + { + "epoch": 0.403712, + "grad_norm": 0.43523988127708435, + "learning_rate": 0.0006763776333103988, + "loss": 2.6875, + "step": 3154 + }, + { + "epoch": 0.40384, + "grad_norm": 0.7361124753952026, + "learning_rate": 0.0006761836339056896, + "loss": 2.697, + "step": 3155 + }, + { + "epoch": 0.403968, + "grad_norm": 0.4545031487941742, + "learning_rate": 0.0006759896042129713, + "loss": 2.6399, + "step": 3156 + }, + { + "epoch": 0.404096, + "grad_norm": 0.4307551980018616, + "learning_rate": 0.0006757955442655998, + "loss": 2.6844, + "step": 3157 + }, + { + "epoch": 0.404224, + "grad_norm": 0.4737016260623932, + "learning_rate": 0.0006756014540969362, + "loss": 2.7739, + "step": 3158 + }, + { + "epoch": 0.404352, + "grad_norm": 2.2147984504699707, + "learning_rate": 0.0006754073337403469, + "loss": 2.7207, + "step": 3159 + }, + { + "epoch": 0.40448, + "grad_norm": 0.487289696931839, + "learning_rate": 0.0006752131832292035, + "loss": 2.6729, + "step": 3160 + }, + { + "epoch": 0.404608, + "grad_norm": 0.4602402150630951, + "learning_rate": 0.0006750190025968827, + "loss": 2.6724, + "step": 3161 + }, + { + "epoch": 0.404736, + "grad_norm": 0.4387877583503723, + "learning_rate": 0.0006748247918767662, + "loss": 2.6605, + "step": 3162 + }, + { + "epoch": 0.404864, + "grad_norm": 0.4432268738746643, + "learning_rate": 0.0006746305511022414, + "loss": 2.7272, + "step": 3163 + }, + { + "epoch": 0.404992, + "grad_norm": 0.43925800919532776, + "learning_rate": 0.0006744362803067002, + "loss": 2.6855, + "step": 3164 + }, + { + "epoch": 0.40512, + "grad_norm": 0.5621846318244934, + "learning_rate": 0.0006742419795235403, + "loss": 2.7754, + "step": 3165 + }, + { + "epoch": 0.405248, + "grad_norm": 0.5356775522232056, + "learning_rate": 0.000674047648786164, + "loss": 2.6945, + "step": 3166 + }, + { + "epoch": 0.405376, + "grad_norm": 0.45477956533432007, + "learning_rate": 0.0006738532881279793, + "loss": 2.7157, + "step": 3167 + }, + { + "epoch": 0.405504, + "grad_norm": 0.458711177110672, + "learning_rate": 0.000673658897582399, + "loss": 2.6565, + "step": 3168 + }, + { + "epoch": 0.405632, + "grad_norm": 0.435838520526886, + "learning_rate": 0.0006734644771828407, + "loss": 2.6409, + "step": 3169 + }, + { + "epoch": 0.40576, + "grad_norm": 0.45911622047424316, + "learning_rate": 0.0006732700269627279, + "loss": 2.7347, + "step": 3170 + }, + { + "epoch": 0.405888, + "grad_norm": 0.41894179582595825, + "learning_rate": 0.0006730755469554888, + "loss": 2.679, + "step": 3171 + }, + { + "epoch": 0.406016, + "grad_norm": 0.45418664813041687, + "learning_rate": 0.0006728810371945566, + "loss": 2.7193, + "step": 3172 + }, + { + "epoch": 0.406144, + "grad_norm": 0.41420260071754456, + "learning_rate": 0.00067268649771337, + "loss": 2.7356, + "step": 3173 + }, + { + "epoch": 0.406272, + "grad_norm": 0.4962754547595978, + "learning_rate": 0.0006724919285453725, + "loss": 2.7485, + "step": 3174 + }, + { + "epoch": 0.4064, + "grad_norm": 0.5047679543495178, + "learning_rate": 0.0006722973297240127, + "loss": 2.7534, + "step": 3175 + }, + { + "epoch": 0.406528, + "grad_norm": 0.4558008909225464, + "learning_rate": 0.0006721027012827444, + "loss": 2.6604, + "step": 3176 + }, + { + "epoch": 0.406656, + "grad_norm": 0.5403389930725098, + "learning_rate": 0.0006719080432550269, + "loss": 2.6606, + "step": 3177 + }, + { + "epoch": 0.406784, + "grad_norm": 0.4607144892215729, + "learning_rate": 0.0006717133556743235, + "loss": 2.7608, + "step": 3178 + }, + { + "epoch": 0.406912, + "grad_norm": 0.45087486505508423, + "learning_rate": 0.0006715186385741037, + "loss": 2.7064, + "step": 3179 + }, + { + "epoch": 0.40704, + "grad_norm": 0.4888212978839874, + "learning_rate": 0.0006713238919878415, + "loss": 2.6863, + "step": 3180 + }, + { + "epoch": 0.407168, + "grad_norm": 0.467678040266037, + "learning_rate": 0.0006711291159490161, + "loss": 2.735, + "step": 3181 + }, + { + "epoch": 0.407296, + "grad_norm": 0.4147484600543976, + "learning_rate": 0.0006709343104911119, + "loss": 2.5894, + "step": 3182 + }, + { + "epoch": 0.407424, + "grad_norm": 0.46527352929115295, + "learning_rate": 0.0006707394756476179, + "loss": 2.6344, + "step": 3183 + }, + { + "epoch": 0.407552, + "grad_norm": 0.4450977146625519, + "learning_rate": 0.0006705446114520288, + "loss": 2.7335, + "step": 3184 + }, + { + "epoch": 0.40768, + "grad_norm": 0.4420625567436218, + "learning_rate": 0.0006703497179378438, + "loss": 2.6305, + "step": 3185 + }, + { + "epoch": 0.407808, + "grad_norm": 0.4202772080898285, + "learning_rate": 0.0006701547951385675, + "loss": 2.6835, + "step": 3186 + }, + { + "epoch": 0.407936, + "grad_norm": 0.44483891129493713, + "learning_rate": 0.0006699598430877092, + "loss": 2.7599, + "step": 3187 + }, + { + "epoch": 0.408064, + "grad_norm": 0.44306060671806335, + "learning_rate": 0.0006697648618187836, + "loss": 2.6671, + "step": 3188 + }, + { + "epoch": 0.408192, + "grad_norm": 0.43513786792755127, + "learning_rate": 0.0006695698513653101, + "loss": 2.7346, + "step": 3189 + }, + { + "epoch": 0.40832, + "grad_norm": 0.45813924074172974, + "learning_rate": 0.0006693748117608133, + "loss": 2.6709, + "step": 3190 + }, + { + "epoch": 0.408448, + "grad_norm": 0.4489612877368927, + "learning_rate": 0.000669179743038823, + "loss": 2.6699, + "step": 3191 + }, + { + "epoch": 0.408576, + "grad_norm": 0.6308565735816956, + "learning_rate": 0.0006689846452328735, + "loss": 2.5748, + "step": 3192 + }, + { + "epoch": 0.408704, + "grad_norm": 0.4322335720062256, + "learning_rate": 0.0006687895183765043, + "loss": 2.5826, + "step": 3193 + }, + { + "epoch": 0.408832, + "grad_norm": 0.43782946467399597, + "learning_rate": 0.0006685943625032603, + "loss": 2.7333, + "step": 3194 + }, + { + "epoch": 0.40896, + "grad_norm": 0.43110591173171997, + "learning_rate": 0.0006683991776466906, + "loss": 2.7263, + "step": 3195 + }, + { + "epoch": 0.409088, + "grad_norm": 0.421236127614975, + "learning_rate": 0.0006682039638403503, + "loss": 2.6126, + "step": 3196 + }, + { + "epoch": 0.409216, + "grad_norm": 0.43923133611679077, + "learning_rate": 0.0006680087211177985, + "loss": 2.6129, + "step": 3197 + }, + { + "epoch": 0.409344, + "grad_norm": 0.448255330324173, + "learning_rate": 0.0006678134495125998, + "loss": 2.6832, + "step": 3198 + }, + { + "epoch": 0.409472, + "grad_norm": 0.4213913381099701, + "learning_rate": 0.0006676181490583238, + "loss": 2.569, + "step": 3199 + }, + { + "epoch": 0.4096, + "grad_norm": 0.4395485520362854, + "learning_rate": 0.0006674228197885446, + "loss": 2.7419, + "step": 3200 + }, + { + "epoch": 0.409728, + "grad_norm": 0.44541576504707336, + "learning_rate": 0.0006672274617368418, + "loss": 2.5613, + "step": 3201 + }, + { + "epoch": 0.409856, + "grad_norm": 0.4636181890964508, + "learning_rate": 0.0006670320749367997, + "loss": 2.6944, + "step": 3202 + }, + { + "epoch": 0.409984, + "grad_norm": 0.429030179977417, + "learning_rate": 0.0006668366594220075, + "loss": 2.6527, + "step": 3203 + }, + { + "epoch": 0.410112, + "grad_norm": 0.4832770526409149, + "learning_rate": 0.0006666412152260595, + "loss": 2.7316, + "step": 3204 + }, + { + "epoch": 0.41024, + "grad_norm": 0.44255244731903076, + "learning_rate": 0.0006664457423825546, + "loss": 2.8024, + "step": 3205 + }, + { + "epoch": 0.410368, + "grad_norm": 0.4415391683578491, + "learning_rate": 0.000666250240925097, + "loss": 2.6766, + "step": 3206 + }, + { + "epoch": 0.410496, + "grad_norm": 0.4333421587944031, + "learning_rate": 0.0006660547108872959, + "loss": 2.6382, + "step": 3207 + }, + { + "epoch": 0.410624, + "grad_norm": 0.4594539701938629, + "learning_rate": 0.0006658591523027643, + "loss": 2.737, + "step": 3208 + }, + { + "epoch": 0.410752, + "grad_norm": 0.4202982485294342, + "learning_rate": 0.0006656635652051222, + "loss": 2.6027, + "step": 3209 + }, + { + "epoch": 0.41088, + "grad_norm": 0.478049635887146, + "learning_rate": 0.0006654679496279926, + "loss": 2.6836, + "step": 3210 + }, + { + "epoch": 0.411008, + "grad_norm": 0.44769778847694397, + "learning_rate": 0.000665272305605004, + "loss": 2.6871, + "step": 3211 + }, + { + "epoch": 0.411136, + "grad_norm": 0.4171440303325653, + "learning_rate": 0.0006650766331697902, + "loss": 2.7153, + "step": 3212 + }, + { + "epoch": 0.411264, + "grad_norm": 0.43034762144088745, + "learning_rate": 0.0006648809323559895, + "loss": 2.5843, + "step": 3213 + }, + { + "epoch": 0.411392, + "grad_norm": 0.4240395724773407, + "learning_rate": 0.0006646852031972449, + "loss": 2.6049, + "step": 3214 + }, + { + "epoch": 0.41152, + "grad_norm": 0.4087779223918915, + "learning_rate": 0.0006644894457272047, + "loss": 2.5678, + "step": 3215 + }, + { + "epoch": 0.411648, + "grad_norm": 0.42350083589553833, + "learning_rate": 0.0006642936599795219, + "loss": 2.7201, + "step": 3216 + }, + { + "epoch": 0.411776, + "grad_norm": 0.43299293518066406, + "learning_rate": 0.0006640978459878543, + "loss": 2.6579, + "step": 3217 + }, + { + "epoch": 0.411904, + "grad_norm": 0.4023003876209259, + "learning_rate": 0.0006639020037858646, + "loss": 2.6492, + "step": 3218 + }, + { + "epoch": 0.412032, + "grad_norm": 0.4347284436225891, + "learning_rate": 0.0006637061334072203, + "loss": 2.6309, + "step": 3219 + }, + { + "epoch": 0.41216, + "grad_norm": 0.5061505436897278, + "learning_rate": 0.0006635102348855939, + "loss": 2.6537, + "step": 3220 + }, + { + "epoch": 0.412288, + "grad_norm": 0.5077633261680603, + "learning_rate": 0.0006633143082546624, + "loss": 2.6164, + "step": 3221 + }, + { + "epoch": 0.412416, + "grad_norm": 0.39584609866142273, + "learning_rate": 0.0006631183535481081, + "loss": 2.6271, + "step": 3222 + }, + { + "epoch": 0.412544, + "grad_norm": 0.4372694492340088, + "learning_rate": 0.0006629223707996177, + "loss": 2.6662, + "step": 3223 + }, + { + "epoch": 0.412672, + "grad_norm": 0.41504383087158203, + "learning_rate": 0.0006627263600428832, + "loss": 2.6936, + "step": 3224 + }, + { + "epoch": 0.4128, + "grad_norm": 0.6956790089607239, + "learning_rate": 0.0006625303213116008, + "loss": 2.6927, + "step": 3225 + }, + { + "epoch": 0.412928, + "grad_norm": 0.4266282618045807, + "learning_rate": 0.0006623342546394719, + "loss": 2.5919, + "step": 3226 + }, + { + "epoch": 0.413056, + "grad_norm": 0.47486695647239685, + "learning_rate": 0.0006621381600602028, + "loss": 2.7659, + "step": 3227 + }, + { + "epoch": 0.413184, + "grad_norm": 0.43533650040626526, + "learning_rate": 0.0006619420376075042, + "loss": 2.5724, + "step": 3228 + }, + { + "epoch": 0.413312, + "grad_norm": 0.4538191258907318, + "learning_rate": 0.0006617458873150921, + "loss": 2.7866, + "step": 3229 + }, + { + "epoch": 0.41344, + "grad_norm": 0.4156574308872223, + "learning_rate": 0.0006615497092166868, + "loss": 2.8024, + "step": 3230 + }, + { + "epoch": 0.413568, + "grad_norm": 0.4739934206008911, + "learning_rate": 0.0006613535033460135, + "loss": 2.6533, + "step": 3231 + }, + { + "epoch": 0.413696, + "grad_norm": 0.434194415807724, + "learning_rate": 0.0006611572697368026, + "loss": 2.7648, + "step": 3232 + }, + { + "epoch": 0.413824, + "grad_norm": 0.4059939980506897, + "learning_rate": 0.0006609610084227886, + "loss": 2.6301, + "step": 3233 + }, + { + "epoch": 0.413952, + "grad_norm": 0.41219398379325867, + "learning_rate": 0.0006607647194377113, + "loss": 2.7562, + "step": 3234 + }, + { + "epoch": 0.41408, + "grad_norm": 0.454926073551178, + "learning_rate": 0.0006605684028153149, + "loss": 2.6921, + "step": 3235 + }, + { + "epoch": 0.414208, + "grad_norm": 0.42336592078208923, + "learning_rate": 0.0006603720585893486, + "loss": 2.7875, + "step": 3236 + }, + { + "epoch": 0.414336, + "grad_norm": 0.4229947328567505, + "learning_rate": 0.0006601756867935664, + "loss": 2.6421, + "step": 3237 + }, + { + "epoch": 0.414464, + "grad_norm": 0.4526359438896179, + "learning_rate": 0.0006599792874617266, + "loss": 2.6294, + "step": 3238 + }, + { + "epoch": 0.414592, + "grad_norm": 0.4559069573879242, + "learning_rate": 0.0006597828606275928, + "loss": 2.6282, + "step": 3239 + }, + { + "epoch": 0.41472, + "grad_norm": 0.4515373706817627, + "learning_rate": 0.0006595864063249326, + "loss": 2.651, + "step": 3240 + }, + { + "epoch": 0.414848, + "grad_norm": 0.4717891216278076, + "learning_rate": 0.0006593899245875192, + "loss": 2.6389, + "step": 3241 + }, + { + "epoch": 0.414976, + "grad_norm": 0.43515175580978394, + "learning_rate": 0.00065919341544913, + "loss": 2.6617, + "step": 3242 + }, + { + "epoch": 0.415104, + "grad_norm": 0.5243332386016846, + "learning_rate": 0.0006589968789435472, + "loss": 2.6804, + "step": 3243 + }, + { + "epoch": 0.415232, + "grad_norm": 0.46102023124694824, + "learning_rate": 0.0006588003151045576, + "loss": 2.5873, + "step": 3244 + }, + { + "epoch": 0.41536, + "grad_norm": 0.4167424142360687, + "learning_rate": 0.0006586037239659528, + "loss": 2.6555, + "step": 3245 + }, + { + "epoch": 0.415488, + "grad_norm": 0.41300827264785767, + "learning_rate": 0.0006584071055615293, + "loss": 2.6151, + "step": 3246 + }, + { + "epoch": 0.415616, + "grad_norm": 0.46057477593421936, + "learning_rate": 0.0006582104599250878, + "loss": 2.784, + "step": 3247 + }, + { + "epoch": 0.415744, + "grad_norm": 0.44236165285110474, + "learning_rate": 0.0006580137870904341, + "loss": 2.6345, + "step": 3248 + }, + { + "epoch": 0.415872, + "grad_norm": 0.4245651662349701, + "learning_rate": 0.0006578170870913787, + "loss": 2.6832, + "step": 3249 + }, + { + "epoch": 0.416, + "grad_norm": 0.4622315466403961, + "learning_rate": 0.0006576203599617363, + "loss": 2.6692, + "step": 3250 + }, + { + "epoch": 0.416128, + "grad_norm": 0.48566800355911255, + "learning_rate": 0.0006574236057353268, + "loss": 2.6791, + "step": 3251 + }, + { + "epoch": 0.416256, + "grad_norm": 0.44099023938179016, + "learning_rate": 0.0006572268244459745, + "loss": 2.6075, + "step": 3252 + }, + { + "epoch": 0.416384, + "grad_norm": 0.4571684002876282, + "learning_rate": 0.0006570300161275084, + "loss": 2.6949, + "step": 3253 + }, + { + "epoch": 0.416512, + "grad_norm": 0.41769513487815857, + "learning_rate": 0.0006568331808137619, + "loss": 2.6066, + "step": 3254 + }, + { + "epoch": 0.41664, + "grad_norm": 0.46616673469543457, + "learning_rate": 0.0006566363185385735, + "loss": 2.7063, + "step": 3255 + }, + { + "epoch": 0.416768, + "grad_norm": 0.44651731848716736, + "learning_rate": 0.0006564394293357861, + "loss": 2.6433, + "step": 3256 + }, + { + "epoch": 0.416896, + "grad_norm": 0.46666648983955383, + "learning_rate": 0.0006562425132392472, + "loss": 2.7943, + "step": 3257 + }, + { + "epoch": 0.417024, + "grad_norm": 0.4180584251880646, + "learning_rate": 0.0006560455702828089, + "loss": 2.7487, + "step": 3258 + }, + { + "epoch": 0.417152, + "grad_norm": 0.4604591727256775, + "learning_rate": 0.0006558486005003282, + "loss": 2.7417, + "step": 3259 + }, + { + "epoch": 0.41728, + "grad_norm": 0.4263269603252411, + "learning_rate": 0.0006556516039256661, + "loss": 2.7055, + "step": 3260 + }, + { + "epoch": 0.417408, + "grad_norm": 0.5124565958976746, + "learning_rate": 0.000655454580592689, + "loss": 2.8019, + "step": 3261 + }, + { + "epoch": 0.417536, + "grad_norm": 0.3917422890663147, + "learning_rate": 0.0006552575305352671, + "loss": 2.6719, + "step": 3262 + }, + { + "epoch": 0.417664, + "grad_norm": 0.4305773675441742, + "learning_rate": 0.0006550604537872759, + "loss": 2.5685, + "step": 3263 + }, + { + "epoch": 0.417792, + "grad_norm": 0.43464550375938416, + "learning_rate": 0.0006548633503825952, + "loss": 2.6974, + "step": 3264 + }, + { + "epoch": 0.41792, + "grad_norm": 0.4311438798904419, + "learning_rate": 0.0006546662203551092, + "loss": 2.5965, + "step": 3265 + }, + { + "epoch": 0.418048, + "grad_norm": 0.4359222948551178, + "learning_rate": 0.0006544690637387068, + "loss": 2.7066, + "step": 3266 + }, + { + "epoch": 0.418176, + "grad_norm": 0.5098830461502075, + "learning_rate": 0.0006542718805672816, + "loss": 2.6353, + "step": 3267 + }, + { + "epoch": 0.418304, + "grad_norm": 0.4811002016067505, + "learning_rate": 0.0006540746708747317, + "loss": 2.718, + "step": 3268 + }, + { + "epoch": 0.418432, + "grad_norm": 0.46382275223731995, + "learning_rate": 0.0006538774346949597, + "loss": 2.6426, + "step": 3269 + }, + { + "epoch": 0.41856, + "grad_norm": 0.45982626080513, + "learning_rate": 0.0006536801720618727, + "loss": 2.6258, + "step": 3270 + }, + { + "epoch": 0.418688, + "grad_norm": 0.4671774208545685, + "learning_rate": 0.0006534828830093827, + "loss": 2.7139, + "step": 3271 + }, + { + "epoch": 0.418816, + "grad_norm": 0.4230813980102539, + "learning_rate": 0.0006532855675714054, + "loss": 2.5583, + "step": 3272 + }, + { + "epoch": 0.418944, + "grad_norm": 0.45539167523384094, + "learning_rate": 0.0006530882257818622, + "loss": 2.7118, + "step": 3273 + }, + { + "epoch": 0.419072, + "grad_norm": 0.44494444131851196, + "learning_rate": 0.0006528908576746784, + "loss": 2.798, + "step": 3274 + }, + { + "epoch": 0.4192, + "grad_norm": 0.4947274327278137, + "learning_rate": 0.0006526934632837835, + "loss": 2.719, + "step": 3275 + }, + { + "epoch": 0.419328, + "grad_norm": 0.4438377320766449, + "learning_rate": 0.0006524960426431122, + "loss": 2.6572, + "step": 3276 + }, + { + "epoch": 0.419456, + "grad_norm": 0.41650500893592834, + "learning_rate": 0.0006522985957866032, + "loss": 2.682, + "step": 3277 + }, + { + "epoch": 0.419584, + "grad_norm": 0.3961542248725891, + "learning_rate": 0.0006521011227482, + "loss": 2.7294, + "step": 3278 + }, + { + "epoch": 0.419712, + "grad_norm": 0.432506799697876, + "learning_rate": 0.0006519036235618505, + "loss": 2.6973, + "step": 3279 + }, + { + "epoch": 0.41984, + "grad_norm": 0.4351501166820526, + "learning_rate": 0.0006517060982615071, + "loss": 2.8052, + "step": 3280 + }, + { + "epoch": 0.419968, + "grad_norm": 0.40583375096321106, + "learning_rate": 0.0006515085468811268, + "loss": 2.7134, + "step": 3281 + }, + { + "epoch": 0.420096, + "grad_norm": 0.39669328927993774, + "learning_rate": 0.0006513109694546706, + "loss": 2.663, + "step": 3282 + }, + { + "epoch": 0.420224, + "grad_norm": 0.42452430725097656, + "learning_rate": 0.0006511133660161047, + "loss": 2.727, + "step": 3283 + }, + { + "epoch": 0.420352, + "grad_norm": 0.4356725811958313, + "learning_rate": 0.0006509157365993994, + "loss": 2.575, + "step": 3284 + }, + { + "epoch": 0.42048, + "grad_norm": 0.4386571943759918, + "learning_rate": 0.0006507180812385295, + "loss": 2.7412, + "step": 3285 + }, + { + "epoch": 0.420608, + "grad_norm": 0.4526103734970093, + "learning_rate": 0.000650520399967474, + "loss": 2.6764, + "step": 3286 + }, + { + "epoch": 0.420736, + "grad_norm": 0.4444957375526428, + "learning_rate": 0.0006503226928202167, + "loss": 2.6348, + "step": 3287 + }, + { + "epoch": 0.420864, + "grad_norm": 0.41257622838020325, + "learning_rate": 0.000650124959830746, + "loss": 2.6668, + "step": 3288 + }, + { + "epoch": 0.420992, + "grad_norm": 0.4419594407081604, + "learning_rate": 0.0006499272010330543, + "loss": 2.7401, + "step": 3289 + }, + { + "epoch": 0.42112, + "grad_norm": 0.42492491006851196, + "learning_rate": 0.0006497294164611385, + "loss": 2.6091, + "step": 3290 + }, + { + "epoch": 0.421248, + "grad_norm": 0.46179303526878357, + "learning_rate": 0.0006495316061490002, + "loss": 2.6277, + "step": 3291 + }, + { + "epoch": 0.421376, + "grad_norm": 0.4277006983757019, + "learning_rate": 0.0006493337701306454, + "loss": 2.7745, + "step": 3292 + }, + { + "epoch": 0.421504, + "grad_norm": 0.4020775556564331, + "learning_rate": 0.0006491359084400843, + "loss": 2.6948, + "step": 3293 + }, + { + "epoch": 0.421632, + "grad_norm": 0.460849791765213, + "learning_rate": 0.0006489380211113315, + "loss": 2.7703, + "step": 3294 + }, + { + "epoch": 0.42176, + "grad_norm": 0.43379780650138855, + "learning_rate": 0.0006487401081784062, + "loss": 2.6873, + "step": 3295 + }, + { + "epoch": 0.421888, + "grad_norm": 0.42068982124328613, + "learning_rate": 0.000648542169675332, + "loss": 2.7047, + "step": 3296 + }, + { + "epoch": 0.422016, + "grad_norm": 0.41824421286582947, + "learning_rate": 0.0006483442056361368, + "loss": 2.7041, + "step": 3297 + }, + { + "epoch": 0.422144, + "grad_norm": 0.4550422132015228, + "learning_rate": 0.000648146216094853, + "loss": 2.5712, + "step": 3298 + }, + { + "epoch": 0.422272, + "grad_norm": 0.42593613266944885, + "learning_rate": 0.0006479482010855172, + "loss": 2.7031, + "step": 3299 + }, + { + "epoch": 0.4224, + "grad_norm": 0.4613417088985443, + "learning_rate": 0.0006477501606421702, + "loss": 2.7479, + "step": 3300 + }, + { + "epoch": 0.422528, + "grad_norm": 0.4399949014186859, + "learning_rate": 0.000647552094798858, + "loss": 2.6725, + "step": 3301 + }, + { + "epoch": 0.422656, + "grad_norm": 0.4620365500450134, + "learning_rate": 0.00064735400358963, + "loss": 2.7112, + "step": 3302 + }, + { + "epoch": 0.422784, + "grad_norm": 0.41163724660873413, + "learning_rate": 0.0006471558870485407, + "loss": 2.5384, + "step": 3303 + }, + { + "epoch": 0.422912, + "grad_norm": 0.4223048686981201, + "learning_rate": 0.0006469577452096482, + "loss": 2.6576, + "step": 3304 + }, + { + "epoch": 0.42304, + "grad_norm": 0.4487987160682678, + "learning_rate": 0.0006467595781070157, + "loss": 2.6047, + "step": 3305 + }, + { + "epoch": 0.423168, + "grad_norm": 0.4062969982624054, + "learning_rate": 0.0006465613857747104, + "loss": 2.55, + "step": 3306 + }, + { + "epoch": 0.423296, + "grad_norm": 0.4332147240638733, + "learning_rate": 0.0006463631682468037, + "loss": 2.6513, + "step": 3307 + }, + { + "epoch": 0.423424, + "grad_norm": 0.3948557376861572, + "learning_rate": 0.0006461649255573716, + "loss": 2.7904, + "step": 3308 + }, + { + "epoch": 0.423552, + "grad_norm": 0.38719817996025085, + "learning_rate": 0.0006459666577404943, + "loss": 2.6887, + "step": 3309 + }, + { + "epoch": 0.42368, + "grad_norm": 0.39434048533439636, + "learning_rate": 0.0006457683648302565, + "loss": 2.6392, + "step": 3310 + }, + { + "epoch": 0.423808, + "grad_norm": 0.4169418215751648, + "learning_rate": 0.0006455700468607469, + "loss": 2.6233, + "step": 3311 + }, + { + "epoch": 0.423936, + "grad_norm": 0.42901280522346497, + "learning_rate": 0.0006453717038660584, + "loss": 2.7011, + "step": 3312 + }, + { + "epoch": 0.424064, + "grad_norm": 0.4049619436264038, + "learning_rate": 0.0006451733358802888, + "loss": 2.6169, + "step": 3313 + }, + { + "epoch": 0.424192, + "grad_norm": 0.41347041726112366, + "learning_rate": 0.0006449749429375398, + "loss": 2.6281, + "step": 3314 + }, + { + "epoch": 0.42432, + "grad_norm": 0.4229666292667389, + "learning_rate": 0.0006447765250719174, + "loss": 2.6258, + "step": 3315 + }, + { + "epoch": 0.424448, + "grad_norm": 0.41616299748420715, + "learning_rate": 0.0006445780823175321, + "loss": 2.7265, + "step": 3316 + }, + { + "epoch": 0.424576, + "grad_norm": 0.4153330624103546, + "learning_rate": 0.0006443796147084981, + "loss": 2.7557, + "step": 3317 + }, + { + "epoch": 0.424704, + "grad_norm": 0.4180048704147339, + "learning_rate": 0.0006441811222789347, + "loss": 2.7599, + "step": 3318 + }, + { + "epoch": 0.424832, + "grad_norm": 0.44427061080932617, + "learning_rate": 0.0006439826050629648, + "loss": 2.5963, + "step": 3319 + }, + { + "epoch": 0.42496, + "grad_norm": 0.4242607355117798, + "learning_rate": 0.0006437840630947161, + "loss": 2.6355, + "step": 3320 + }, + { + "epoch": 0.425088, + "grad_norm": 0.42353934049606323, + "learning_rate": 0.0006435854964083198, + "loss": 2.6016, + "step": 3321 + }, + { + "epoch": 0.425216, + "grad_norm": 0.42068615555763245, + "learning_rate": 0.0006433869050379121, + "loss": 2.6851, + "step": 3322 + }, + { + "epoch": 0.425344, + "grad_norm": 0.4306877553462982, + "learning_rate": 0.0006431882890176333, + "loss": 2.7693, + "step": 3323 + }, + { + "epoch": 0.425472, + "grad_norm": 0.415218323469162, + "learning_rate": 0.0006429896483816276, + "loss": 2.7751, + "step": 3324 + }, + { + "epoch": 0.4256, + "grad_norm": 0.45054367184638977, + "learning_rate": 0.0006427909831640437, + "loss": 2.7479, + "step": 3325 + }, + { + "epoch": 0.425728, + "grad_norm": 0.4466172754764557, + "learning_rate": 0.0006425922933990343, + "loss": 2.745, + "step": 3326 + }, + { + "epoch": 0.425856, + "grad_norm": 0.4466860592365265, + "learning_rate": 0.0006423935791207568, + "loss": 2.7798, + "step": 3327 + }, + { + "epoch": 0.425984, + "grad_norm": 0.38836267590522766, + "learning_rate": 0.0006421948403633721, + "loss": 2.5643, + "step": 3328 + }, + { + "epoch": 0.426112, + "grad_norm": 0.430422842502594, + "learning_rate": 0.000641996077161046, + "loss": 2.6741, + "step": 3329 + }, + { + "epoch": 0.42624, + "grad_norm": 0.4170595705509186, + "learning_rate": 0.000641797289547948, + "loss": 2.5819, + "step": 3330 + }, + { + "epoch": 0.426368, + "grad_norm": 0.4708280861377716, + "learning_rate": 0.0006415984775582521, + "loss": 2.6748, + "step": 3331 + }, + { + "epoch": 0.426496, + "grad_norm": 0.42773568630218506, + "learning_rate": 0.0006413996412261363, + "loss": 2.6982, + "step": 3332 + }, + { + "epoch": 0.426624, + "grad_norm": 0.4588960111141205, + "learning_rate": 0.0006412007805857828, + "loss": 2.5785, + "step": 3333 + }, + { + "epoch": 0.426752, + "grad_norm": 0.416199266910553, + "learning_rate": 0.0006410018956713783, + "loss": 2.5565, + "step": 3334 + }, + { + "epoch": 0.42688, + "grad_norm": 0.43983960151672363, + "learning_rate": 0.0006408029865171134, + "loss": 2.5981, + "step": 3335 + }, + { + "epoch": 0.427008, + "grad_norm": 0.4907297194004059, + "learning_rate": 0.0006406040531571828, + "loss": 2.6821, + "step": 3336 + }, + { + "epoch": 0.427136, + "grad_norm": 0.43239015340805054, + "learning_rate": 0.0006404050956257852, + "loss": 2.6441, + "step": 3337 + }, + { + "epoch": 0.427264, + "grad_norm": 0.44523000717163086, + "learning_rate": 0.0006402061139571243, + "loss": 2.6762, + "step": 3338 + }, + { + "epoch": 0.427392, + "grad_norm": 0.4542066752910614, + "learning_rate": 0.0006400071081854068, + "loss": 2.7693, + "step": 3339 + }, + { + "epoch": 0.42752, + "grad_norm": 0.47359341382980347, + "learning_rate": 0.0006398080783448442, + "loss": 2.6307, + "step": 3340 + }, + { + "epoch": 0.427648, + "grad_norm": 0.4746011793613434, + "learning_rate": 0.0006396090244696523, + "loss": 2.5641, + "step": 3341 + }, + { + "epoch": 0.427776, + "grad_norm": 0.4117324948310852, + "learning_rate": 0.0006394099465940505, + "loss": 2.6515, + "step": 3342 + }, + { + "epoch": 0.427904, + "grad_norm": 0.42863738536834717, + "learning_rate": 0.0006392108447522629, + "loss": 2.7133, + "step": 3343 + }, + { + "epoch": 0.428032, + "grad_norm": 0.49543696641921997, + "learning_rate": 0.000639011718978517, + "loss": 2.6184, + "step": 3344 + }, + { + "epoch": 0.42816, + "grad_norm": 0.4391080141067505, + "learning_rate": 0.0006388125693070451, + "loss": 2.6642, + "step": 3345 + }, + { + "epoch": 0.428288, + "grad_norm": 0.7400546669960022, + "learning_rate": 0.0006386133957720832, + "loss": 2.6737, + "step": 3346 + }, + { + "epoch": 0.428416, + "grad_norm": 0.3917359411716461, + "learning_rate": 0.0006384141984078719, + "loss": 2.6127, + "step": 3347 + }, + { + "epoch": 0.428544, + "grad_norm": 0.41547077894210815, + "learning_rate": 0.000638214977248655, + "loss": 2.625, + "step": 3348 + }, + { + "epoch": 0.428672, + "grad_norm": 0.40934839844703674, + "learning_rate": 0.0006380157323286812, + "loss": 2.6937, + "step": 3349 + }, + { + "epoch": 0.4288, + "grad_norm": 0.40447890758514404, + "learning_rate": 0.0006378164636822032, + "loss": 2.7266, + "step": 3350 + }, + { + "epoch": 0.428928, + "grad_norm": 0.41714683175086975, + "learning_rate": 0.0006376171713434771, + "loss": 2.7295, + "step": 3351 + }, + { + "epoch": 0.429056, + "grad_norm": 0.4348315894603729, + "learning_rate": 0.0006374178553467639, + "loss": 2.742, + "step": 3352 + }, + { + "epoch": 0.429184, + "grad_norm": 0.43367066979408264, + "learning_rate": 0.0006372185157263286, + "loss": 2.6492, + "step": 3353 + }, + { + "epoch": 0.429312, + "grad_norm": 0.40848004817962646, + "learning_rate": 0.0006370191525164394, + "loss": 2.7317, + "step": 3354 + }, + { + "epoch": 0.42944, + "grad_norm": 0.43032941222190857, + "learning_rate": 0.0006368197657513695, + "loss": 2.6708, + "step": 3355 + }, + { + "epoch": 0.429568, + "grad_norm": 0.4492982029914856, + "learning_rate": 0.0006366203554653957, + "loss": 2.6732, + "step": 3356 + }, + { + "epoch": 0.429696, + "grad_norm": 0.4378896653652191, + "learning_rate": 0.000636420921692799, + "loss": 2.7521, + "step": 3357 + }, + { + "epoch": 0.429824, + "grad_norm": 0.4118106961250305, + "learning_rate": 0.0006362214644678641, + "loss": 2.6644, + "step": 3358 + }, + { + "epoch": 0.429952, + "grad_norm": 0.4338974952697754, + "learning_rate": 0.0006360219838248806, + "loss": 2.7006, + "step": 3359 + }, + { + "epoch": 0.43008, + "grad_norm": 0.509498655796051, + "learning_rate": 0.0006358224797981409, + "loss": 2.5007, + "step": 3360 + }, + { + "epoch": 0.430208, + "grad_norm": 0.47988927364349365, + "learning_rate": 0.0006356229524219425, + "loss": 2.7458, + "step": 3361 + }, + { + "epoch": 0.430336, + "grad_norm": 0.45221787691116333, + "learning_rate": 0.0006354234017305862, + "loss": 2.6849, + "step": 3362 + }, + { + "epoch": 0.430464, + "grad_norm": 0.45204591751098633, + "learning_rate": 0.0006352238277583773, + "loss": 2.6474, + "step": 3363 + }, + { + "epoch": 0.430592, + "grad_norm": 0.4269125461578369, + "learning_rate": 0.0006350242305396247, + "loss": 2.6993, + "step": 3364 + }, + { + "epoch": 0.43072, + "grad_norm": 0.4190605878829956, + "learning_rate": 0.0006348246101086413, + "loss": 2.6957, + "step": 3365 + }, + { + "epoch": 0.430848, + "grad_norm": 0.41927221417427063, + "learning_rate": 0.0006346249664997448, + "loss": 2.6737, + "step": 3366 + }, + { + "epoch": 0.430976, + "grad_norm": 0.4251423478126526, + "learning_rate": 0.0006344252997472556, + "loss": 2.6924, + "step": 3367 + }, + { + "epoch": 0.431104, + "grad_norm": 0.4349012076854706, + "learning_rate": 0.0006342256098854992, + "loss": 2.6029, + "step": 3368 + }, + { + "epoch": 0.431232, + "grad_norm": 0.45315271615982056, + "learning_rate": 0.0006340258969488039, + "loss": 2.6734, + "step": 3369 + }, + { + "epoch": 0.43136, + "grad_norm": 0.4251817762851715, + "learning_rate": 0.0006338261609715037, + "loss": 2.7057, + "step": 3370 + }, + { + "epoch": 0.431488, + "grad_norm": 0.4300902783870697, + "learning_rate": 0.0006336264019879348, + "loss": 2.676, + "step": 3371 + }, + { + "epoch": 0.431616, + "grad_norm": 0.4188481867313385, + "learning_rate": 0.0006334266200324381, + "loss": 2.678, + "step": 3372 + }, + { + "epoch": 0.431744, + "grad_norm": 0.4486893117427826, + "learning_rate": 0.0006332268151393589, + "loss": 2.7339, + "step": 3373 + }, + { + "epoch": 0.431872, + "grad_norm": 0.43334639072418213, + "learning_rate": 0.0006330269873430454, + "loss": 2.5948, + "step": 3374 + }, + { + "epoch": 0.432, + "grad_norm": 0.42759305238723755, + "learning_rate": 0.0006328271366778509, + "loss": 2.6662, + "step": 3375 + }, + { + "epoch": 0.432128, + "grad_norm": 0.4469779431819916, + "learning_rate": 0.0006326272631781314, + "loss": 2.6279, + "step": 3376 + }, + { + "epoch": 0.432256, + "grad_norm": 0.4017120599746704, + "learning_rate": 0.000632427366878248, + "loss": 2.5945, + "step": 3377 + }, + { + "epoch": 0.432384, + "grad_norm": 0.4050412178039551, + "learning_rate": 0.000632227447812565, + "loss": 2.5869, + "step": 3378 + }, + { + "epoch": 0.432512, + "grad_norm": 0.8083236813545227, + "learning_rate": 0.0006320275060154507, + "loss": 2.7146, + "step": 3379 + }, + { + "epoch": 0.43264, + "grad_norm": 0.44127780199050903, + "learning_rate": 0.0006318275415212776, + "loss": 2.6129, + "step": 3380 + }, + { + "epoch": 0.432768, + "grad_norm": 0.4602248966693878, + "learning_rate": 0.0006316275543644218, + "loss": 2.6579, + "step": 3381 + }, + { + "epoch": 0.432896, + "grad_norm": 0.443558931350708, + "learning_rate": 0.0006314275445792637, + "loss": 2.6741, + "step": 3382 + }, + { + "epoch": 0.433024, + "grad_norm": 0.4636945426464081, + "learning_rate": 0.0006312275122001866, + "loss": 2.7393, + "step": 3383 + }, + { + "epoch": 0.433152, + "grad_norm": 0.40727105736732483, + "learning_rate": 0.0006310274572615791, + "loss": 2.5937, + "step": 3384 + }, + { + "epoch": 0.43328, + "grad_norm": 0.40816035866737366, + "learning_rate": 0.0006308273797978328, + "loss": 2.74, + "step": 3385 + }, + { + "epoch": 0.433408, + "grad_norm": 0.5189038515090942, + "learning_rate": 0.000630627279843343, + "loss": 2.5505, + "step": 3386 + }, + { + "epoch": 0.433536, + "grad_norm": 0.40875667333602905, + "learning_rate": 0.0006304271574325095, + "loss": 2.53, + "step": 3387 + }, + { + "epoch": 0.433664, + "grad_norm": 0.41026049852371216, + "learning_rate": 0.0006302270125997356, + "loss": 2.6343, + "step": 3388 + }, + { + "epoch": 0.433792, + "grad_norm": 0.44734427332878113, + "learning_rate": 0.0006300268453794287, + "loss": 2.6697, + "step": 3389 + }, + { + "epoch": 0.43392, + "grad_norm": 0.4356188476085663, + "learning_rate": 0.0006298266558059995, + "loss": 2.6196, + "step": 3390 + }, + { + "epoch": 0.434048, + "grad_norm": 0.47979116439819336, + "learning_rate": 0.000629626443913863, + "loss": 2.7245, + "step": 3391 + }, + { + "epoch": 0.434176, + "grad_norm": 0.4491216242313385, + "learning_rate": 0.0006294262097374382, + "loss": 2.6778, + "step": 3392 + }, + { + "epoch": 0.434304, + "grad_norm": 0.4373437762260437, + "learning_rate": 0.0006292259533111474, + "loss": 2.6741, + "step": 3393 + }, + { + "epoch": 0.434432, + "grad_norm": 0.4897361099720001, + "learning_rate": 0.0006290256746694172, + "loss": 2.7173, + "step": 3394 + }, + { + "epoch": 0.43456, + "grad_norm": 0.4202924370765686, + "learning_rate": 0.0006288253738466777, + "loss": 2.6025, + "step": 3395 + }, + { + "epoch": 0.434688, + "grad_norm": 0.4378717541694641, + "learning_rate": 0.000628625050877363, + "loss": 2.7919, + "step": 3396 + }, + { + "epoch": 0.434816, + "grad_norm": 0.4350748360157013, + "learning_rate": 0.0006284247057959107, + "loss": 2.827, + "step": 3397 + }, + { + "epoch": 0.434944, + "grad_norm": 0.44158047437667847, + "learning_rate": 0.0006282243386367628, + "loss": 2.7086, + "step": 3398 + }, + { + "epoch": 0.435072, + "grad_norm": 0.4948982000350952, + "learning_rate": 0.0006280239494343647, + "loss": 2.5704, + "step": 3399 + }, + { + "epoch": 0.4352, + "grad_norm": 0.43888410925865173, + "learning_rate": 0.0006278235382231654, + "loss": 2.6833, + "step": 3400 + }, + { + "epoch": 0.435328, + "grad_norm": 0.4323723316192627, + "learning_rate": 0.000627623105037618, + "loss": 2.6853, + "step": 3401 + }, + { + "epoch": 0.435456, + "grad_norm": 0.4543459713459015, + "learning_rate": 0.0006274226499121792, + "loss": 2.6463, + "step": 3402 + }, + { + "epoch": 0.435584, + "grad_norm": 0.40493711829185486, + "learning_rate": 0.0006272221728813099, + "loss": 2.6197, + "step": 3403 + }, + { + "epoch": 0.435712, + "grad_norm": 0.4561637341976166, + "learning_rate": 0.0006270216739794741, + "loss": 2.5759, + "step": 3404 + }, + { + "epoch": 0.43584, + "grad_norm": 0.5108547806739807, + "learning_rate": 0.00062682115324114, + "loss": 2.6906, + "step": 3405 + }, + { + "epoch": 0.435968, + "grad_norm": 0.4499706029891968, + "learning_rate": 0.0006266206107007793, + "loss": 2.6868, + "step": 3406 + }, + { + "epoch": 0.436096, + "grad_norm": 0.41568249464035034, + "learning_rate": 0.0006264200463928679, + "loss": 2.5893, + "step": 3407 + }, + { + "epoch": 0.436224, + "grad_norm": 0.47048285603523254, + "learning_rate": 0.0006262194603518847, + "loss": 2.7373, + "step": 3408 + }, + { + "epoch": 0.436352, + "grad_norm": 0.442351758480072, + "learning_rate": 0.0006260188526123133, + "loss": 2.6686, + "step": 3409 + }, + { + "epoch": 0.43648, + "grad_norm": 0.4359511733055115, + "learning_rate": 0.00062581822320864, + "loss": 2.7326, + "step": 3410 + }, + { + "epoch": 0.436608, + "grad_norm": 0.47016915678977966, + "learning_rate": 0.0006256175721753556, + "loss": 2.5591, + "step": 3411 + }, + { + "epoch": 0.436736, + "grad_norm": 0.4121721684932709, + "learning_rate": 0.0006254168995469541, + "loss": 2.7095, + "step": 3412 + }, + { + "epoch": 0.436864, + "grad_norm": 0.4678231179714203, + "learning_rate": 0.0006252162053579337, + "loss": 2.5992, + "step": 3413 + }, + { + "epoch": 0.436992, + "grad_norm": 0.42315465211868286, + "learning_rate": 0.0006250154896427962, + "loss": 2.7552, + "step": 3414 + }, + { + "epoch": 0.43712, + "grad_norm": 0.45252934098243713, + "learning_rate": 0.0006248147524360464, + "loss": 2.7048, + "step": 3415 + }, + { + "epoch": 0.437248, + "grad_norm": 0.44867971539497375, + "learning_rate": 0.0006246139937721939, + "loss": 2.6264, + "step": 3416 + }, + { + "epoch": 0.437376, + "grad_norm": 0.4689997434616089, + "learning_rate": 0.0006244132136857512, + "loss": 2.6783, + "step": 3417 + }, + { + "epoch": 0.437504, + "grad_norm": 0.4854201376438141, + "learning_rate": 0.0006242124122112347, + "loss": 2.6981, + "step": 3418 + }, + { + "epoch": 0.437632, + "grad_norm": 0.7346668243408203, + "learning_rate": 0.0006240115893831644, + "loss": 2.674, + "step": 3419 + }, + { + "epoch": 0.43776, + "grad_norm": 0.7936125993728638, + "learning_rate": 0.0006238107452360643, + "loss": 2.6822, + "step": 3420 + }, + { + "epoch": 0.437888, + "grad_norm": 0.4783092737197876, + "learning_rate": 0.0006236098798044618, + "loss": 2.6709, + "step": 3421 + }, + { + "epoch": 0.438016, + "grad_norm": 0.5199064612388611, + "learning_rate": 0.000623408993122888, + "loss": 2.6423, + "step": 3422 + }, + { + "epoch": 0.438144, + "grad_norm": 0.4634522795677185, + "learning_rate": 0.0006232080852258776, + "loss": 2.6645, + "step": 3423 + }, + { + "epoch": 0.438272, + "grad_norm": 0.43513014912605286, + "learning_rate": 0.000623007156147969, + "loss": 2.6582, + "step": 3424 + }, + { + "epoch": 0.4384, + "grad_norm": 0.43726152181625366, + "learning_rate": 0.000622806205923704, + "loss": 2.6334, + "step": 3425 + }, + { + "epoch": 0.438528, + "grad_norm": 0.4493562877178192, + "learning_rate": 0.0006226052345876288, + "loss": 2.6862, + "step": 3426 + }, + { + "epoch": 0.438656, + "grad_norm": 0.45185214281082153, + "learning_rate": 0.0006224042421742924, + "loss": 2.5992, + "step": 3427 + }, + { + "epoch": 0.438784, + "grad_norm": 0.4370933771133423, + "learning_rate": 0.0006222032287182477, + "loss": 2.7442, + "step": 3428 + }, + { + "epoch": 0.438912, + "grad_norm": 0.4530796408653259, + "learning_rate": 0.0006220021942540512, + "loss": 2.6354, + "step": 3429 + }, + { + "epoch": 0.43904, + "grad_norm": 0.39763426780700684, + "learning_rate": 0.000621801138816263, + "loss": 2.5078, + "step": 3430 + }, + { + "epoch": 0.439168, + "grad_norm": 0.4525744616985321, + "learning_rate": 0.0006216000624394472, + "loss": 2.7007, + "step": 3431 + }, + { + "epoch": 0.439296, + "grad_norm": 0.47901681065559387, + "learning_rate": 0.000621398965158171, + "loss": 2.7361, + "step": 3432 + }, + { + "epoch": 0.439424, + "grad_norm": 0.46646848320961, + "learning_rate": 0.0006211978470070052, + "loss": 2.6585, + "step": 3433 + }, + { + "epoch": 0.439552, + "grad_norm": 0.425365149974823, + "learning_rate": 0.0006209967080205244, + "loss": 2.6388, + "step": 3434 + }, + { + "epoch": 0.43968, + "grad_norm": 0.4274127185344696, + "learning_rate": 0.000620795548233307, + "loss": 2.5919, + "step": 3435 + }, + { + "epoch": 0.439808, + "grad_norm": 0.4152558743953705, + "learning_rate": 0.0006205943676799344, + "loss": 2.7253, + "step": 3436 + }, + { + "epoch": 0.439936, + "grad_norm": 0.44425493478775024, + "learning_rate": 0.0006203931663949918, + "loss": 2.5555, + "step": 3437 + }, + { + "epoch": 0.440064, + "grad_norm": 0.43028995394706726, + "learning_rate": 0.0006201919444130683, + "loss": 2.6765, + "step": 3438 + }, + { + "epoch": 0.440192, + "grad_norm": 0.4320198595523834, + "learning_rate": 0.0006199907017687561, + "loss": 2.6739, + "step": 3439 + }, + { + "epoch": 0.44032, + "grad_norm": 0.41169169545173645, + "learning_rate": 0.0006197894384966512, + "loss": 2.6045, + "step": 3440 + }, + { + "epoch": 0.440448, + "grad_norm": 0.449472576379776, + "learning_rate": 0.0006195881546313532, + "loss": 2.7153, + "step": 3441 + }, + { + "epoch": 0.440576, + "grad_norm": 0.4166293442249298, + "learning_rate": 0.0006193868502074651, + "loss": 2.6659, + "step": 3442 + }, + { + "epoch": 0.440704, + "grad_norm": 0.48259440064430237, + "learning_rate": 0.0006191855252595933, + "loss": 2.6969, + "step": 3443 + }, + { + "epoch": 0.440832, + "grad_norm": 0.40903523564338684, + "learning_rate": 0.0006189841798223479, + "loss": 2.6292, + "step": 3444 + }, + { + "epoch": 0.44096, + "grad_norm": 0.48224544525146484, + "learning_rate": 0.0006187828139303427, + "loss": 2.5605, + "step": 3445 + }, + { + "epoch": 0.441088, + "grad_norm": 0.4171522557735443, + "learning_rate": 0.0006185814276181948, + "loss": 2.6635, + "step": 3446 + }, + { + "epoch": 0.441216, + "grad_norm": 0.446623831987381, + "learning_rate": 0.0006183800209205246, + "loss": 2.6513, + "step": 3447 + }, + { + "epoch": 0.441344, + "grad_norm": 0.4088413119316101, + "learning_rate": 0.0006181785938719565, + "loss": 2.6708, + "step": 3448 + }, + { + "epoch": 0.441472, + "grad_norm": 0.3858242630958557, + "learning_rate": 0.0006179771465071181, + "loss": 2.7569, + "step": 3449 + }, + { + "epoch": 0.4416, + "grad_norm": 0.4366661608219147, + "learning_rate": 0.0006177756788606405, + "loss": 2.5872, + "step": 3450 + }, + { + "epoch": 0.441728, + "grad_norm": 0.41199368238449097, + "learning_rate": 0.0006175741909671584, + "loss": 2.6971, + "step": 3451 + }, + { + "epoch": 0.441856, + "grad_norm": 0.41862690448760986, + "learning_rate": 0.0006173726828613097, + "loss": 2.6665, + "step": 3452 + }, + { + "epoch": 0.441984, + "grad_norm": 0.4698803722858429, + "learning_rate": 0.0006171711545777363, + "loss": 2.6704, + "step": 3453 + }, + { + "epoch": 0.442112, + "grad_norm": 0.44222116470336914, + "learning_rate": 0.000616969606151083, + "loss": 2.6558, + "step": 3454 + }, + { + "epoch": 0.44224, + "grad_norm": 0.44089874625205994, + "learning_rate": 0.0006167680376159982, + "loss": 2.7073, + "step": 3455 + }, + { + "epoch": 0.442368, + "grad_norm": 0.3864535689353943, + "learning_rate": 0.0006165664490071342, + "loss": 2.7432, + "step": 3456 + }, + { + "epoch": 0.442496, + "grad_norm": 0.5355798006057739, + "learning_rate": 0.0006163648403591462, + "loss": 2.5932, + "step": 3457 + }, + { + "epoch": 0.442624, + "grad_norm": 0.47481071949005127, + "learning_rate": 0.0006161632117066931, + "loss": 2.7347, + "step": 3458 + }, + { + "epoch": 0.442752, + "grad_norm": 0.46041786670684814, + "learning_rate": 0.0006159615630844374, + "loss": 2.711, + "step": 3459 + }, + { + "epoch": 0.44288, + "grad_norm": 0.4408402144908905, + "learning_rate": 0.0006157598945270446, + "loss": 2.6565, + "step": 3460 + }, + { + "epoch": 0.443008, + "grad_norm": 0.41998228430747986, + "learning_rate": 0.0006155582060691837, + "loss": 2.6351, + "step": 3461 + }, + { + "epoch": 0.443136, + "grad_norm": 0.4139338731765747, + "learning_rate": 0.0006153564977455277, + "loss": 2.5588, + "step": 3462 + }, + { + "epoch": 0.443264, + "grad_norm": 0.466104656457901, + "learning_rate": 0.0006151547695907525, + "loss": 2.7905, + "step": 3463 + }, + { + "epoch": 0.443392, + "grad_norm": 0.42699718475341797, + "learning_rate": 0.0006149530216395374, + "loss": 2.6253, + "step": 3464 + }, + { + "epoch": 0.44352, + "grad_norm": 0.46177321672439575, + "learning_rate": 0.0006147512539265651, + "loss": 2.7537, + "step": 3465 + }, + { + "epoch": 0.443648, + "grad_norm": 0.4464803636074066, + "learning_rate": 0.000614549466486522, + "loss": 2.6699, + "step": 3466 + }, + { + "epoch": 0.443776, + "grad_norm": 0.452265202999115, + "learning_rate": 0.0006143476593540976, + "loss": 2.7528, + "step": 3467 + }, + { + "epoch": 0.443904, + "grad_norm": 1.226440668106079, + "learning_rate": 0.000614145832563985, + "loss": 2.6916, + "step": 3468 + }, + { + "epoch": 0.444032, + "grad_norm": 0.6060749292373657, + "learning_rate": 0.0006139439861508803, + "loss": 2.6212, + "step": 3469 + }, + { + "epoch": 0.44416, + "grad_norm": 0.4561951458454132, + "learning_rate": 0.0006137421201494836, + "loss": 2.7134, + "step": 3470 + }, + { + "epoch": 0.444288, + "grad_norm": 0.42577803134918213, + "learning_rate": 0.0006135402345944978, + "loss": 2.6883, + "step": 3471 + }, + { + "epoch": 0.444416, + "grad_norm": 0.513135552406311, + "learning_rate": 0.0006133383295206294, + "loss": 2.6411, + "step": 3472 + }, + { + "epoch": 0.444544, + "grad_norm": 0.498402863740921, + "learning_rate": 0.0006131364049625882, + "loss": 2.6614, + "step": 3473 + }, + { + "epoch": 0.444672, + "grad_norm": 0.4870561361312866, + "learning_rate": 0.0006129344609550876, + "loss": 2.666, + "step": 3474 + }, + { + "epoch": 0.4448, + "grad_norm": 0.40029385685920715, + "learning_rate": 0.0006127324975328436, + "loss": 2.7143, + "step": 3475 + }, + { + "epoch": 0.444928, + "grad_norm": 0.42133551836013794, + "learning_rate": 0.0006125305147305764, + "loss": 2.6269, + "step": 3476 + }, + { + "epoch": 0.445056, + "grad_norm": 0.4832616150379181, + "learning_rate": 0.0006123285125830093, + "loss": 2.6583, + "step": 3477 + }, + { + "epoch": 0.445184, + "grad_norm": 0.5843334197998047, + "learning_rate": 0.0006121264911248687, + "loss": 2.6166, + "step": 3478 + }, + { + "epoch": 0.445312, + "grad_norm": 0.522104799747467, + "learning_rate": 0.0006119244503908843, + "loss": 2.7184, + "step": 3479 + }, + { + "epoch": 0.44544, + "grad_norm": 0.43012726306915283, + "learning_rate": 0.0006117223904157893, + "loss": 2.6969, + "step": 3480 + }, + { + "epoch": 0.445568, + "grad_norm": 0.44549480080604553, + "learning_rate": 0.0006115203112343203, + "loss": 2.5251, + "step": 3481 + }, + { + "epoch": 0.445696, + "grad_norm": 0.5584995150566101, + "learning_rate": 0.0006113182128812169, + "loss": 2.793, + "step": 3482 + }, + { + "epoch": 0.445824, + "grad_norm": 0.5712236166000366, + "learning_rate": 0.0006111160953912221, + "loss": 2.6291, + "step": 3483 + }, + { + "epoch": 0.445952, + "grad_norm": 0.43198341131210327, + "learning_rate": 0.0006109139587990825, + "loss": 2.5949, + "step": 3484 + }, + { + "epoch": 0.44608, + "grad_norm": 0.4386935532093048, + "learning_rate": 0.0006107118031395475, + "loss": 2.6165, + "step": 3485 + }, + { + "epoch": 0.446208, + "grad_norm": 0.4076373279094696, + "learning_rate": 0.0006105096284473701, + "loss": 2.7062, + "step": 3486 + }, + { + "epoch": 0.446336, + "grad_norm": 0.4728737771511078, + "learning_rate": 0.0006103074347573063, + "loss": 2.6988, + "step": 3487 + }, + { + "epoch": 0.446464, + "grad_norm": 0.558007538318634, + "learning_rate": 0.0006101052221041157, + "loss": 2.7443, + "step": 3488 + }, + { + "epoch": 0.446592, + "grad_norm": 0.46800366044044495, + "learning_rate": 0.0006099029905225611, + "loss": 2.6633, + "step": 3489 + }, + { + "epoch": 0.44672, + "grad_norm": 0.45893508195877075, + "learning_rate": 0.000609700740047408, + "loss": 2.5996, + "step": 3490 + }, + { + "epoch": 0.446848, + "grad_norm": 0.5587778091430664, + "learning_rate": 0.0006094984707134262, + "loss": 2.7029, + "step": 3491 + }, + { + "epoch": 0.446976, + "grad_norm": 0.3972693979740143, + "learning_rate": 0.000609296182555388, + "loss": 2.5295, + "step": 3492 + }, + { + "epoch": 0.447104, + "grad_norm": 0.45898205041885376, + "learning_rate": 0.0006090938756080688, + "loss": 2.6739, + "step": 3493 + }, + { + "epoch": 0.447232, + "grad_norm": 0.44471311569213867, + "learning_rate": 0.0006088915499062475, + "loss": 2.6486, + "step": 3494 + }, + { + "epoch": 0.44736, + "grad_norm": 0.4146456718444824, + "learning_rate": 0.0006086892054847065, + "loss": 2.6161, + "step": 3495 + }, + { + "epoch": 0.447488, + "grad_norm": 0.4358496367931366, + "learning_rate": 0.0006084868423782312, + "loss": 2.74, + "step": 3496 + }, + { + "epoch": 0.447616, + "grad_norm": 0.3947336673736572, + "learning_rate": 0.0006082844606216098, + "loss": 2.7969, + "step": 3497 + }, + { + "epoch": 0.447744, + "grad_norm": 0.433628112077713, + "learning_rate": 0.0006080820602496344, + "loss": 2.6514, + "step": 3498 + }, + { + "epoch": 0.447872, + "grad_norm": 0.43359604477882385, + "learning_rate": 0.0006078796412970997, + "loss": 2.5689, + "step": 3499 + }, + { + "epoch": 0.448, + "grad_norm": 0.4407494366168976, + "learning_rate": 0.0006076772037988042, + "loss": 2.695, + "step": 3500 + }, + { + "epoch": 0.448128, + "grad_norm": 0.43490663170814514, + "learning_rate": 0.0006074747477895489, + "loss": 2.7516, + "step": 3501 + }, + { + "epoch": 0.448256, + "grad_norm": 0.4399477541446686, + "learning_rate": 0.0006072722733041387, + "loss": 2.7535, + "step": 3502 + }, + { + "epoch": 0.448384, + "grad_norm": 0.426708847284317, + "learning_rate": 0.000607069780377381, + "loss": 2.6923, + "step": 3503 + }, + { + "epoch": 0.448512, + "grad_norm": 0.4081552028656006, + "learning_rate": 0.0006068672690440867, + "loss": 2.582, + "step": 3504 + }, + { + "epoch": 0.44864, + "grad_norm": 0.422916054725647, + "learning_rate": 0.00060666473933907, + "loss": 2.6167, + "step": 3505 + }, + { + "epoch": 0.448768, + "grad_norm": 0.432437539100647, + "learning_rate": 0.0006064621912971481, + "loss": 2.6604, + "step": 3506 + }, + { + "epoch": 0.448896, + "grad_norm": 0.44323721528053284, + "learning_rate": 0.0006062596249531414, + "loss": 2.6849, + "step": 3507 + }, + { + "epoch": 0.449024, + "grad_norm": 0.4319612383842468, + "learning_rate": 0.0006060570403418731, + "loss": 2.7732, + "step": 3508 + }, + { + "epoch": 0.449152, + "grad_norm": 0.46495163440704346, + "learning_rate": 0.00060585443749817, + "loss": 2.6691, + "step": 3509 + }, + { + "epoch": 0.44928, + "grad_norm": 0.4313790798187256, + "learning_rate": 0.0006056518164568622, + "loss": 2.613, + "step": 3510 + }, + { + "epoch": 0.449408, + "grad_norm": 0.43348926305770874, + "learning_rate": 0.0006054491772527821, + "loss": 2.6751, + "step": 3511 + }, + { + "epoch": 0.449536, + "grad_norm": 0.46647322177886963, + "learning_rate": 0.0006052465199207661, + "loss": 2.5929, + "step": 3512 + }, + { + "epoch": 0.449664, + "grad_norm": 0.4679429531097412, + "learning_rate": 0.0006050438444956531, + "loss": 2.5532, + "step": 3513 + }, + { + "epoch": 0.449792, + "grad_norm": 0.4132622480392456, + "learning_rate": 0.0006048411510122855, + "loss": 2.6818, + "step": 3514 + }, + { + "epoch": 0.44992, + "grad_norm": 0.4442862570285797, + "learning_rate": 0.0006046384395055086, + "loss": 2.6728, + "step": 3515 + }, + { + "epoch": 0.450048, + "grad_norm": 0.44336575269699097, + "learning_rate": 0.000604435710010171, + "loss": 2.6118, + "step": 3516 + }, + { + "epoch": 0.450176, + "grad_norm": 0.41580548882484436, + "learning_rate": 0.0006042329625611239, + "loss": 2.7139, + "step": 3517 + }, + { + "epoch": 0.450304, + "grad_norm": 0.45374441146850586, + "learning_rate": 0.0006040301971932222, + "loss": 2.6915, + "step": 3518 + }, + { + "epoch": 0.450432, + "grad_norm": 0.4193647503852844, + "learning_rate": 0.0006038274139413237, + "loss": 2.7238, + "step": 3519 + }, + { + "epoch": 0.45056, + "grad_norm": 0.45160380005836487, + "learning_rate": 0.0006036246128402891, + "loss": 2.6538, + "step": 3520 + }, + { + "epoch": 0.450688, + "grad_norm": 0.44265976548194885, + "learning_rate": 0.0006034217939249822, + "loss": 2.7432, + "step": 3521 + }, + { + "epoch": 0.450816, + "grad_norm": 0.471637099981308, + "learning_rate": 0.00060321895723027, + "loss": 2.649, + "step": 3522 + }, + { + "epoch": 0.450944, + "grad_norm": 0.4324197769165039, + "learning_rate": 0.0006030161027910223, + "loss": 2.5995, + "step": 3523 + }, + { + "epoch": 0.451072, + "grad_norm": 0.42540398240089417, + "learning_rate": 0.0006028132306421124, + "loss": 2.6597, + "step": 3524 + }, + { + "epoch": 0.4512, + "grad_norm": 0.48885324597358704, + "learning_rate": 0.0006026103408184162, + "loss": 2.6637, + "step": 3525 + }, + { + "epoch": 0.451328, + "grad_norm": 0.4165911078453064, + "learning_rate": 0.0006024074333548127, + "loss": 2.7281, + "step": 3526 + }, + { + "epoch": 0.451456, + "grad_norm": 0.4622178077697754, + "learning_rate": 0.0006022045082861843, + "loss": 2.6516, + "step": 3527 + }, + { + "epoch": 0.451584, + "grad_norm": 0.468044251203537, + "learning_rate": 0.0006020015656474161, + "loss": 2.6498, + "step": 3528 + }, + { + "epoch": 0.451712, + "grad_norm": 0.4295586049556732, + "learning_rate": 0.0006017986054733962, + "loss": 2.5864, + "step": 3529 + }, + { + "epoch": 0.45184, + "grad_norm": 0.4716932475566864, + "learning_rate": 0.000601595627799016, + "loss": 2.5885, + "step": 3530 + }, + { + "epoch": 0.451968, + "grad_norm": 0.4211593270301819, + "learning_rate": 0.0006013926326591695, + "loss": 2.6458, + "step": 3531 + }, + { + "epoch": 0.452096, + "grad_norm": 0.43002307415008545, + "learning_rate": 0.0006011896200887539, + "loss": 2.6343, + "step": 3532 + }, + { + "epoch": 0.452224, + "grad_norm": 0.42186614871025085, + "learning_rate": 0.0006009865901226697, + "loss": 2.6071, + "step": 3533 + }, + { + "epoch": 0.452352, + "grad_norm": 0.4477013051509857, + "learning_rate": 0.0006007835427958198, + "loss": 2.827, + "step": 3534 + }, + { + "epoch": 0.45248, + "grad_norm": 0.4372682571411133, + "learning_rate": 0.0006005804781431106, + "loss": 2.6776, + "step": 3535 + }, + { + "epoch": 0.452608, + "grad_norm": 0.5314607620239258, + "learning_rate": 0.0006003773961994511, + "loss": 2.6402, + "step": 3536 + }, + { + "epoch": 0.452736, + "grad_norm": 0.458514541387558, + "learning_rate": 0.0006001742969997534, + "loss": 2.6672, + "step": 3537 + }, + { + "epoch": 0.452864, + "grad_norm": 0.42673179507255554, + "learning_rate": 0.0005999711805789329, + "loss": 2.6372, + "step": 3538 + }, + { + "epoch": 0.452992, + "grad_norm": 0.48301783204078674, + "learning_rate": 0.0005997680469719076, + "loss": 2.6327, + "step": 3539 + }, + { + "epoch": 0.45312, + "grad_norm": 0.5077583193778992, + "learning_rate": 0.0005995648962135983, + "loss": 2.7408, + "step": 3540 + }, + { + "epoch": 0.453248, + "grad_norm": 0.4512609839439392, + "learning_rate": 0.0005993617283389289, + "loss": 2.6609, + "step": 3541 + }, + { + "epoch": 0.453376, + "grad_norm": 0.4425329864025116, + "learning_rate": 0.0005991585433828267, + "loss": 2.5847, + "step": 3542 + }, + { + "epoch": 0.453504, + "grad_norm": 0.5638167858123779, + "learning_rate": 0.0005989553413802214, + "loss": 2.6281, + "step": 3543 + }, + { + "epoch": 0.453632, + "grad_norm": 0.4378117024898529, + "learning_rate": 0.0005987521223660456, + "loss": 2.6016, + "step": 3544 + }, + { + "epoch": 0.45376, + "grad_norm": 0.49881839752197266, + "learning_rate": 0.0005985488863752351, + "loss": 2.5619, + "step": 3545 + }, + { + "epoch": 0.453888, + "grad_norm": 0.5203617215156555, + "learning_rate": 0.0005983456334427286, + "loss": 2.5802, + "step": 3546 + }, + { + "epoch": 0.454016, + "grad_norm": 0.4292161166667938, + "learning_rate": 0.0005981423636034676, + "loss": 2.602, + "step": 3547 + }, + { + "epoch": 0.454144, + "grad_norm": 0.5293667316436768, + "learning_rate": 0.0005979390768923967, + "loss": 2.7564, + "step": 3548 + }, + { + "epoch": 0.454272, + "grad_norm": 0.44395872950553894, + "learning_rate": 0.000597735773344463, + "loss": 2.5763, + "step": 3549 + }, + { + "epoch": 0.4544, + "grad_norm": 0.4539519250392914, + "learning_rate": 0.0005975324529946165, + "loss": 2.7497, + "step": 3550 + }, + { + "epoch": 0.454528, + "grad_norm": 0.44959402084350586, + "learning_rate": 0.0005973291158778109, + "loss": 2.6359, + "step": 3551 + }, + { + "epoch": 0.454656, + "grad_norm": 0.45422396063804626, + "learning_rate": 0.000597125762029002, + "loss": 2.7312, + "step": 3552 + }, + { + "epoch": 0.454784, + "grad_norm": 0.5480499863624573, + "learning_rate": 0.0005969223914831486, + "loss": 2.6753, + "step": 3553 + }, + { + "epoch": 0.454912, + "grad_norm": 0.46321946382522583, + "learning_rate": 0.0005967190042752123, + "loss": 2.6398, + "step": 3554 + }, + { + "epoch": 0.45504, + "grad_norm": 0.4175094962120056, + "learning_rate": 0.000596515600440158, + "loss": 2.6711, + "step": 3555 + }, + { + "epoch": 0.455168, + "grad_norm": 0.5400997996330261, + "learning_rate": 0.0005963121800129532, + "loss": 2.7117, + "step": 3556 + }, + { + "epoch": 0.455296, + "grad_norm": 0.44600537419319153, + "learning_rate": 0.0005961087430285681, + "loss": 2.6966, + "step": 3557 + }, + { + "epoch": 0.455424, + "grad_norm": 0.4471571445465088, + "learning_rate": 0.0005959052895219757, + "loss": 2.6718, + "step": 3558 + }, + { + "epoch": 0.455552, + "grad_norm": 0.42906850576400757, + "learning_rate": 0.0005957018195281523, + "loss": 2.6312, + "step": 3559 + }, + { + "epoch": 0.45568, + "grad_norm": 0.435300350189209, + "learning_rate": 0.0005954983330820766, + "loss": 2.7237, + "step": 3560 + }, + { + "epoch": 0.455808, + "grad_norm": 0.48888835310935974, + "learning_rate": 0.0005952948302187306, + "loss": 2.6575, + "step": 3561 + }, + { + "epoch": 0.455936, + "grad_norm": 0.41141611337661743, + "learning_rate": 0.0005950913109730982, + "loss": 2.6101, + "step": 3562 + }, + { + "epoch": 0.456064, + "grad_norm": 0.44142019748687744, + "learning_rate": 0.0005948877753801673, + "loss": 2.722, + "step": 3563 + }, + { + "epoch": 0.456192, + "grad_norm": 0.453522264957428, + "learning_rate": 0.0005946842234749274, + "loss": 2.6174, + "step": 3564 + }, + { + "epoch": 0.45632, + "grad_norm": 0.430726557970047, + "learning_rate": 0.0005944806552923721, + "loss": 2.7046, + "step": 3565 + }, + { + "epoch": 0.456448, + "grad_norm": 0.4429830312728882, + "learning_rate": 0.0005942770708674969, + "loss": 2.7415, + "step": 3566 + }, + { + "epoch": 0.456576, + "grad_norm": 0.43722376227378845, + "learning_rate": 0.0005940734702353002, + "loss": 2.7064, + "step": 3567 + }, + { + "epoch": 0.456704, + "grad_norm": 0.42542433738708496, + "learning_rate": 0.0005938698534307833, + "loss": 2.6179, + "step": 3568 + }, + { + "epoch": 0.456832, + "grad_norm": 0.3991949260234833, + "learning_rate": 0.0005936662204889504, + "loss": 2.6196, + "step": 3569 + }, + { + "epoch": 0.45696, + "grad_norm": 0.45752012729644775, + "learning_rate": 0.0005934625714448084, + "loss": 2.6704, + "step": 3570 + }, + { + "epoch": 0.457088, + "grad_norm": 0.40307122468948364, + "learning_rate": 0.0005932589063333668, + "loss": 2.7211, + "step": 3571 + }, + { + "epoch": 0.457216, + "grad_norm": 0.5042967796325684, + "learning_rate": 0.000593055225189638, + "loss": 2.7455, + "step": 3572 + }, + { + "epoch": 0.457344, + "grad_norm": 0.4047251045703888, + "learning_rate": 0.0005928515280486372, + "loss": 2.6744, + "step": 3573 + }, + { + "epoch": 0.457472, + "grad_norm": 0.46663719415664673, + "learning_rate": 0.0005926478149453823, + "loss": 2.6549, + "step": 3574 + }, + { + "epoch": 0.4576, + "grad_norm": 0.41524648666381836, + "learning_rate": 0.000592444085914894, + "loss": 2.6683, + "step": 3575 + }, + { + "epoch": 0.457728, + "grad_norm": 0.4046042561531067, + "learning_rate": 0.0005922403409921956, + "loss": 2.76, + "step": 3576 + }, + { + "epoch": 0.457856, + "grad_norm": 0.43207526206970215, + "learning_rate": 0.0005920365802123133, + "loss": 2.7735, + "step": 3577 + }, + { + "epoch": 0.457984, + "grad_norm": 0.40478605031967163, + "learning_rate": 0.0005918328036102758, + "loss": 2.6837, + "step": 3578 + }, + { + "epoch": 0.458112, + "grad_norm": 0.42747628688812256, + "learning_rate": 0.000591629011221115, + "loss": 2.6875, + "step": 3579 + }, + { + "epoch": 0.45824, + "grad_norm": 0.4189658463001251, + "learning_rate": 0.0005914252030798645, + "loss": 2.5584, + "step": 3580 + }, + { + "epoch": 0.458368, + "grad_norm": 0.4563167691230774, + "learning_rate": 0.000591221379221562, + "loss": 2.5511, + "step": 3581 + }, + { + "epoch": 0.458496, + "grad_norm": 0.4272587299346924, + "learning_rate": 0.0005910175396812467, + "loss": 2.6216, + "step": 3582 + }, + { + "epoch": 0.458624, + "grad_norm": 0.43637731671333313, + "learning_rate": 0.0005908136844939611, + "loss": 2.7281, + "step": 3583 + }, + { + "epoch": 0.458752, + "grad_norm": 0.45516642928123474, + "learning_rate": 0.0005906098136947507, + "loss": 2.6138, + "step": 3584 + }, + { + "epoch": 0.45888, + "grad_norm": 0.4063355028629303, + "learning_rate": 0.0005904059273186626, + "loss": 2.6154, + "step": 3585 + }, + { + "epoch": 0.459008, + "grad_norm": 0.4430711269378662, + "learning_rate": 0.0005902020254007475, + "loss": 2.6609, + "step": 3586 + }, + { + "epoch": 0.459136, + "grad_norm": 0.44740745425224304, + "learning_rate": 0.0005899981079760585, + "loss": 2.7105, + "step": 3587 + }, + { + "epoch": 0.459264, + "grad_norm": 0.48125383257865906, + "learning_rate": 0.0005897941750796516, + "loss": 2.721, + "step": 3588 + }, + { + "epoch": 0.459392, + "grad_norm": 0.41304150223731995, + "learning_rate": 0.000589590226746585, + "loss": 2.6001, + "step": 3589 + }, + { + "epoch": 0.45952, + "grad_norm": 0.43877699971199036, + "learning_rate": 0.0005893862630119197, + "loss": 2.7628, + "step": 3590 + }, + { + "epoch": 0.459648, + "grad_norm": 0.41296303272247314, + "learning_rate": 0.0005891822839107195, + "loss": 2.7231, + "step": 3591 + }, + { + "epoch": 0.459776, + "grad_norm": 0.41989144682884216, + "learning_rate": 0.0005889782894780509, + "loss": 2.6355, + "step": 3592 + }, + { + "epoch": 0.459904, + "grad_norm": 0.4233319163322449, + "learning_rate": 0.0005887742797489828, + "loss": 2.6961, + "step": 3593 + }, + { + "epoch": 0.460032, + "grad_norm": 0.44329315423965454, + "learning_rate": 0.0005885702547585867, + "loss": 2.6218, + "step": 3594 + }, + { + "epoch": 0.46016, + "grad_norm": 0.48605257272720337, + "learning_rate": 0.0005883662145419373, + "loss": 2.6755, + "step": 3595 + }, + { + "epoch": 0.460288, + "grad_norm": 0.44402265548706055, + "learning_rate": 0.0005881621591341109, + "loss": 2.443, + "step": 3596 + }, + { + "epoch": 0.460416, + "grad_norm": 0.5939376950263977, + "learning_rate": 0.0005879580885701874, + "loss": 2.6519, + "step": 3597 + }, + { + "epoch": 0.460544, + "grad_norm": 0.4200781285762787, + "learning_rate": 0.0005877540028852488, + "loss": 2.6323, + "step": 3598 + }, + { + "epoch": 0.460672, + "grad_norm": 0.39743635058403015, + "learning_rate": 0.0005875499021143799, + "loss": 2.7415, + "step": 3599 + }, + { + "epoch": 0.4608, + "grad_norm": 0.403612345457077, + "learning_rate": 0.0005873457862926677, + "loss": 2.6308, + "step": 3600 + }, + { + "epoch": 0.460928, + "grad_norm": 0.45728176832199097, + "learning_rate": 0.0005871416554552021, + "loss": 2.628, + "step": 3601 + }, + { + "epoch": 0.461056, + "grad_norm": 0.4094712734222412, + "learning_rate": 0.000586937509637076, + "loss": 2.5819, + "step": 3602 + }, + { + "epoch": 0.461184, + "grad_norm": 0.5142059326171875, + "learning_rate": 0.0005867333488733839, + "loss": 2.6352, + "step": 3603 + }, + { + "epoch": 0.461312, + "grad_norm": 0.4781353175640106, + "learning_rate": 0.0005865291731992238, + "loss": 2.6298, + "step": 3604 + }, + { + "epoch": 0.46144, + "grad_norm": 0.47610265016555786, + "learning_rate": 0.0005863249826496955, + "loss": 2.7479, + "step": 3605 + }, + { + "epoch": 0.461568, + "grad_norm": 0.42319995164871216, + "learning_rate": 0.000586120777259902, + "loss": 2.6397, + "step": 3606 + }, + { + "epoch": 0.461696, + "grad_norm": 0.4460820257663727, + "learning_rate": 0.0005859165570649486, + "loss": 2.7028, + "step": 3607 + }, + { + "epoch": 0.461824, + "grad_norm": 0.44550517201423645, + "learning_rate": 0.0005857123220999428, + "loss": 2.6594, + "step": 3608 + }, + { + "epoch": 0.461952, + "grad_norm": 0.4461895227432251, + "learning_rate": 0.0005855080723999954, + "loss": 2.7193, + "step": 3609 + }, + { + "epoch": 0.46208, + "grad_norm": 0.4278015196323395, + "learning_rate": 0.0005853038080002188, + "loss": 2.5316, + "step": 3610 + }, + { + "epoch": 0.462208, + "grad_norm": 0.4606892764568329, + "learning_rate": 0.0005850995289357289, + "loss": 2.6142, + "step": 3611 + }, + { + "epoch": 0.462336, + "grad_norm": 0.45052728056907654, + "learning_rate": 0.0005848952352416433, + "loss": 2.6282, + "step": 3612 + }, + { + "epoch": 0.462464, + "grad_norm": 0.5000635385513306, + "learning_rate": 0.0005846909269530827, + "loss": 2.5806, + "step": 3613 + }, + { + "epoch": 0.462592, + "grad_norm": 0.47088709473609924, + "learning_rate": 0.0005844866041051698, + "loss": 2.7163, + "step": 3614 + }, + { + "epoch": 0.46272, + "grad_norm": 0.5213805437088013, + "learning_rate": 0.0005842822667330302, + "loss": 2.7166, + "step": 3615 + }, + { + "epoch": 0.462848, + "grad_norm": 0.5701237916946411, + "learning_rate": 0.0005840779148717919, + "loss": 2.6774, + "step": 3616 + }, + { + "epoch": 0.462976, + "grad_norm": 0.40302032232284546, + "learning_rate": 0.0005838735485565855, + "loss": 2.641, + "step": 3617 + }, + { + "epoch": 0.463104, + "grad_norm": 0.5306447148323059, + "learning_rate": 0.0005836691678225434, + "loss": 2.5983, + "step": 3618 + }, + { + "epoch": 0.463232, + "grad_norm": 0.4661039113998413, + "learning_rate": 0.0005834647727048016, + "loss": 2.6266, + "step": 3619 + }, + { + "epoch": 0.46336, + "grad_norm": 0.4363112151622772, + "learning_rate": 0.0005832603632384977, + "loss": 2.6868, + "step": 3620 + }, + { + "epoch": 0.463488, + "grad_norm": 0.4355423152446747, + "learning_rate": 0.0005830559394587724, + "loss": 2.559, + "step": 3621 + }, + { + "epoch": 0.463616, + "grad_norm": 0.4596025347709656, + "learning_rate": 0.0005828515014007678, + "loss": 2.648, + "step": 3622 + }, + { + "epoch": 0.463744, + "grad_norm": 0.47258317470550537, + "learning_rate": 0.0005826470490996298, + "loss": 2.6634, + "step": 3623 + }, + { + "epoch": 0.463872, + "grad_norm": 0.41454631090164185, + "learning_rate": 0.000582442582590506, + "loss": 2.6486, + "step": 3624 + }, + { + "epoch": 0.464, + "grad_norm": 0.48408401012420654, + "learning_rate": 0.0005822381019085465, + "loss": 2.6065, + "step": 3625 + }, + { + "epoch": 0.464128, + "grad_norm": 0.47781991958618164, + "learning_rate": 0.0005820336070889037, + "loss": 2.6131, + "step": 3626 + }, + { + "epoch": 0.464256, + "grad_norm": 0.4564192295074463, + "learning_rate": 0.0005818290981667332, + "loss": 2.6354, + "step": 3627 + }, + { + "epoch": 0.464384, + "grad_norm": 0.5042110681533813, + "learning_rate": 0.0005816245751771917, + "loss": 2.6841, + "step": 3628 + }, + { + "epoch": 0.464512, + "grad_norm": 0.49353185296058655, + "learning_rate": 0.0005814200381554397, + "loss": 2.7768, + "step": 3629 + }, + { + "epoch": 0.46464, + "grad_norm": 0.5043033957481384, + "learning_rate": 0.0005812154871366389, + "loss": 2.6408, + "step": 3630 + }, + { + "epoch": 0.464768, + "grad_norm": 0.4309419095516205, + "learning_rate": 0.0005810109221559548, + "loss": 2.6778, + "step": 3631 + }, + { + "epoch": 0.464896, + "grad_norm": 0.4768683612346649, + "learning_rate": 0.0005808063432485537, + "loss": 2.6911, + "step": 3632 + }, + { + "epoch": 0.465024, + "grad_norm": 0.4615538716316223, + "learning_rate": 0.0005806017504496054, + "loss": 2.6901, + "step": 3633 + }, + { + "epoch": 0.465152, + "grad_norm": 0.471219927072525, + "learning_rate": 0.0005803971437942819, + "loss": 2.6688, + "step": 3634 + }, + { + "epoch": 0.46528, + "grad_norm": 0.46263858675956726, + "learning_rate": 0.0005801925233177574, + "loss": 2.6296, + "step": 3635 + }, + { + "epoch": 0.465408, + "grad_norm": 0.4238431453704834, + "learning_rate": 0.0005799878890552083, + "loss": 2.6385, + "step": 3636 + }, + { + "epoch": 0.465536, + "grad_norm": 0.5052914023399353, + "learning_rate": 0.000579783241041814, + "loss": 2.6308, + "step": 3637 + }, + { + "epoch": 0.465664, + "grad_norm": 0.5161786079406738, + "learning_rate": 0.0005795785793127554, + "loss": 2.7343, + "step": 3638 + }, + { + "epoch": 0.465792, + "grad_norm": 0.43728360533714294, + "learning_rate": 0.0005793739039032165, + "loss": 2.726, + "step": 3639 + }, + { + "epoch": 0.46592, + "grad_norm": 0.45656251907348633, + "learning_rate": 0.0005791692148483834, + "loss": 2.7087, + "step": 3640 + }, + { + "epoch": 0.466048, + "grad_norm": 0.4170609414577484, + "learning_rate": 0.0005789645121834445, + "loss": 2.6259, + "step": 3641 + }, + { + "epoch": 0.466176, + "grad_norm": 0.47406303882598877, + "learning_rate": 0.0005787597959435904, + "loss": 2.6925, + "step": 3642 + }, + { + "epoch": 0.466304, + "grad_norm": 0.45932844281196594, + "learning_rate": 0.0005785550661640144, + "loss": 2.6546, + "step": 3643 + }, + { + "epoch": 0.466432, + "grad_norm": 0.4843679666519165, + "learning_rate": 0.0005783503228799119, + "loss": 2.6963, + "step": 3644 + }, + { + "epoch": 0.46656, + "grad_norm": 0.43630293011665344, + "learning_rate": 0.0005781455661264805, + "loss": 2.6931, + "step": 3645 + }, + { + "epoch": 0.466688, + "grad_norm": 0.48868098855018616, + "learning_rate": 0.0005779407959389204, + "loss": 2.7167, + "step": 3646 + }, + { + "epoch": 0.466816, + "grad_norm": 0.49096399545669556, + "learning_rate": 0.0005777360123524338, + "loss": 2.654, + "step": 3647 + }, + { + "epoch": 0.466944, + "grad_norm": 0.40336236357688904, + "learning_rate": 0.0005775312154022255, + "loss": 2.6429, + "step": 3648 + }, + { + "epoch": 0.467072, + "grad_norm": 0.45266270637512207, + "learning_rate": 0.0005773264051235025, + "loss": 2.6879, + "step": 3649 + }, + { + "epoch": 0.4672, + "grad_norm": 0.435447096824646, + "learning_rate": 0.0005771215815514741, + "loss": 2.676, + "step": 3650 + }, + { + "epoch": 0.467328, + "grad_norm": 0.418267160654068, + "learning_rate": 0.0005769167447213517, + "loss": 2.5809, + "step": 3651 + }, + { + "epoch": 0.467456, + "grad_norm": 0.4243088662624359, + "learning_rate": 0.0005767118946683491, + "loss": 2.7319, + "step": 3652 + }, + { + "epoch": 0.467584, + "grad_norm": 0.5652197599411011, + "learning_rate": 0.0005765070314276826, + "loss": 2.6412, + "step": 3653 + }, + { + "epoch": 0.467712, + "grad_norm": 0.4587794542312622, + "learning_rate": 0.0005763021550345703, + "loss": 2.6819, + "step": 3654 + }, + { + "epoch": 0.46784, + "grad_norm": 0.4222044050693512, + "learning_rate": 0.0005760972655242332, + "loss": 2.7243, + "step": 3655 + }, + { + "epoch": 0.467968, + "grad_norm": 0.4461591839790344, + "learning_rate": 0.0005758923629318937, + "loss": 2.7423, + "step": 3656 + }, + { + "epoch": 0.468096, + "grad_norm": 0.486573725938797, + "learning_rate": 0.0005756874472927774, + "loss": 2.6446, + "step": 3657 + }, + { + "epoch": 0.468224, + "grad_norm": 0.4273239076137543, + "learning_rate": 0.0005754825186421113, + "loss": 2.737, + "step": 3658 + }, + { + "epoch": 0.468352, + "grad_norm": 0.4456970691680908, + "learning_rate": 0.0005752775770151254, + "loss": 2.6116, + "step": 3659 + }, + { + "epoch": 0.46848, + "grad_norm": 0.48849642276763916, + "learning_rate": 0.0005750726224470513, + "loss": 2.7072, + "step": 3660 + }, + { + "epoch": 0.468608, + "grad_norm": 0.442396342754364, + "learning_rate": 0.0005748676549731229, + "loss": 2.6892, + "step": 3661 + }, + { + "epoch": 0.468736, + "grad_norm": 0.4361538290977478, + "learning_rate": 0.0005746626746285767, + "loss": 2.6594, + "step": 3662 + }, + { + "epoch": 0.468864, + "grad_norm": 0.41567790508270264, + "learning_rate": 0.0005744576814486512, + "loss": 2.7202, + "step": 3663 + }, + { + "epoch": 0.468992, + "grad_norm": 0.4098895788192749, + "learning_rate": 0.0005742526754685871, + "loss": 2.6144, + "step": 3664 + }, + { + "epoch": 0.46912, + "grad_norm": 0.43717727065086365, + "learning_rate": 0.0005740476567236272, + "loss": 2.7071, + "step": 3665 + }, + { + "epoch": 0.469248, + "grad_norm": 0.4887903034687042, + "learning_rate": 0.0005738426252490168, + "loss": 2.6952, + "step": 3666 + }, + { + "epoch": 0.469376, + "grad_norm": 0.38119539618492126, + "learning_rate": 0.000573637581080003, + "loss": 2.5969, + "step": 3667 + }, + { + "epoch": 0.469504, + "grad_norm": 0.4192467927932739, + "learning_rate": 0.0005734325242518354, + "loss": 2.6234, + "step": 3668 + }, + { + "epoch": 0.469632, + "grad_norm": 0.4024077355861664, + "learning_rate": 0.0005732274547997655, + "loss": 2.6893, + "step": 3669 + }, + { + "epoch": 0.46976, + "grad_norm": 0.4626774489879608, + "learning_rate": 0.0005730223727590471, + "loss": 2.6409, + "step": 3670 + }, + { + "epoch": 0.469888, + "grad_norm": 0.4465026259422302, + "learning_rate": 0.0005728172781649365, + "loss": 2.6853, + "step": 3671 + }, + { + "epoch": 0.470016, + "grad_norm": 0.4947108328342438, + "learning_rate": 0.0005726121710526913, + "loss": 2.7328, + "step": 3672 + }, + { + "epoch": 0.470144, + "grad_norm": 0.4315232038497925, + "learning_rate": 0.0005724070514575724, + "loss": 2.6478, + "step": 3673 + }, + { + "epoch": 0.470272, + "grad_norm": 0.4100453853607178, + "learning_rate": 0.0005722019194148419, + "loss": 2.6825, + "step": 3674 + }, + { + "epoch": 0.4704, + "grad_norm": 0.45318523049354553, + "learning_rate": 0.0005719967749597643, + "loss": 2.6238, + "step": 3675 + }, + { + "epoch": 0.470528, + "grad_norm": 0.47459253668785095, + "learning_rate": 0.0005717916181276065, + "loss": 2.6932, + "step": 3676 + }, + { + "epoch": 0.470656, + "grad_norm": 0.46613195538520813, + "learning_rate": 0.0005715864489536372, + "loss": 2.7968, + "step": 3677 + }, + { + "epoch": 0.470784, + "grad_norm": 0.4268389046192169, + "learning_rate": 0.0005713812674731276, + "loss": 2.6443, + "step": 3678 + }, + { + "epoch": 0.470912, + "grad_norm": 0.4265710115432739, + "learning_rate": 0.0005711760737213504, + "loss": 2.5913, + "step": 3679 + }, + { + "epoch": 0.47104, + "grad_norm": 0.46362075209617615, + "learning_rate": 0.000570970867733581, + "loss": 2.7535, + "step": 3680 + }, + { + "epoch": 0.471168, + "grad_norm": 0.48826438188552856, + "learning_rate": 0.0005707656495450968, + "loss": 2.6166, + "step": 3681 + }, + { + "epoch": 0.471296, + "grad_norm": 0.4304620623588562, + "learning_rate": 0.000570560419191177, + "loss": 2.729, + "step": 3682 + }, + { + "epoch": 0.471424, + "grad_norm": 0.4621700644493103, + "learning_rate": 0.000570355176707103, + "loss": 2.6526, + "step": 3683 + }, + { + "epoch": 0.471552, + "grad_norm": 0.4161078929901123, + "learning_rate": 0.0005701499221281586, + "loss": 2.6403, + "step": 3684 + }, + { + "epoch": 0.47168, + "grad_norm": 0.4544752240180969, + "learning_rate": 0.0005699446554896293, + "loss": 2.807, + "step": 3685 + }, + { + "epoch": 0.471808, + "grad_norm": 0.45239487290382385, + "learning_rate": 0.0005697393768268031, + "loss": 2.7332, + "step": 3686 + }, + { + "epoch": 0.471936, + "grad_norm": 0.7077343463897705, + "learning_rate": 0.0005695340861749693, + "loss": 2.6124, + "step": 3687 + }, + { + "epoch": 0.472064, + "grad_norm": 0.4165453314781189, + "learning_rate": 0.00056932878356942, + "loss": 2.5729, + "step": 3688 + }, + { + "epoch": 0.472192, + "grad_norm": 0.43241196870803833, + "learning_rate": 0.0005691234690454491, + "loss": 2.7393, + "step": 3689 + }, + { + "epoch": 0.47232, + "grad_norm": 0.42498812079429626, + "learning_rate": 0.0005689181426383524, + "loss": 2.6883, + "step": 3690 + }, + { + "epoch": 0.472448, + "grad_norm": 0.5331746339797974, + "learning_rate": 0.0005687128043834282, + "loss": 2.5817, + "step": 3691 + }, + { + "epoch": 0.472576, + "grad_norm": 0.4804598093032837, + "learning_rate": 0.0005685074543159763, + "loss": 2.7492, + "step": 3692 + }, + { + "epoch": 0.472704, + "grad_norm": 0.43090009689331055, + "learning_rate": 0.0005683020924712988, + "loss": 2.6554, + "step": 3693 + }, + { + "epoch": 0.472832, + "grad_norm": 0.45021384954452515, + "learning_rate": 0.0005680967188846996, + "loss": 2.7361, + "step": 3694 + }, + { + "epoch": 0.47296, + "grad_norm": 0.44792014360427856, + "learning_rate": 0.0005678913335914851, + "loss": 2.7838, + "step": 3695 + }, + { + "epoch": 0.473088, + "grad_norm": 0.44593751430511475, + "learning_rate": 0.0005676859366269634, + "loss": 2.6219, + "step": 3696 + }, + { + "epoch": 0.473216, + "grad_norm": 0.48828551173210144, + "learning_rate": 0.0005674805280264445, + "loss": 2.6332, + "step": 3697 + }, + { + "epoch": 0.473344, + "grad_norm": 0.4695759415626526, + "learning_rate": 0.0005672751078252403, + "loss": 2.5918, + "step": 3698 + }, + { + "epoch": 0.473472, + "grad_norm": 0.47193557024002075, + "learning_rate": 0.0005670696760586653, + "loss": 2.7297, + "step": 3699 + }, + { + "epoch": 0.4736, + "grad_norm": 0.41559553146362305, + "learning_rate": 0.0005668642327620356, + "loss": 2.5577, + "step": 3700 + }, + { + "epoch": 0.473728, + "grad_norm": 0.4561074674129486, + "learning_rate": 0.000566658777970669, + "loss": 2.5934, + "step": 3701 + }, + { + "epoch": 0.473856, + "grad_norm": 0.4355768859386444, + "learning_rate": 0.0005664533117198855, + "loss": 2.5968, + "step": 3702 + }, + { + "epoch": 0.473984, + "grad_norm": 0.5653644800186157, + "learning_rate": 0.0005662478340450076, + "loss": 2.5291, + "step": 3703 + }, + { + "epoch": 0.474112, + "grad_norm": 0.3954998552799225, + "learning_rate": 0.0005660423449813588, + "loss": 2.7224, + "step": 3704 + }, + { + "epoch": 0.47424, + "grad_norm": 0.45250001549720764, + "learning_rate": 0.0005658368445642654, + "loss": 2.6019, + "step": 3705 + }, + { + "epoch": 0.474368, + "grad_norm": 0.42180928587913513, + "learning_rate": 0.0005656313328290553, + "loss": 2.6628, + "step": 3706 + }, + { + "epoch": 0.474496, + "grad_norm": 0.4360772967338562, + "learning_rate": 0.000565425809811058, + "loss": 2.69, + "step": 3707 + }, + { + "epoch": 0.474624, + "grad_norm": 0.43011775612831116, + "learning_rate": 0.0005652202755456055, + "loss": 2.7879, + "step": 3708 + }, + { + "epoch": 0.474752, + "grad_norm": 0.40996360778808594, + "learning_rate": 0.0005650147300680317, + "loss": 2.6108, + "step": 3709 + }, + { + "epoch": 0.47488, + "grad_norm": 0.4257133901119232, + "learning_rate": 0.000564809173413672, + "loss": 2.5384, + "step": 3710 + }, + { + "epoch": 0.475008, + "grad_norm": 0.4248911738395691, + "learning_rate": 0.000564603605617864, + "loss": 2.7264, + "step": 3711 + }, + { + "epoch": 0.475136, + "grad_norm": 0.48547297716140747, + "learning_rate": 0.0005643980267159473, + "loss": 2.7402, + "step": 3712 + }, + { + "epoch": 0.475264, + "grad_norm": 0.435365229845047, + "learning_rate": 0.0005641924367432633, + "loss": 2.7706, + "step": 3713 + }, + { + "epoch": 0.475392, + "grad_norm": 0.46224191784858704, + "learning_rate": 0.0005639868357351553, + "loss": 2.5336, + "step": 3714 + }, + { + "epoch": 0.47552, + "grad_norm": 0.45761701464653015, + "learning_rate": 0.0005637812237269682, + "loss": 2.8145, + "step": 3715 + }, + { + "epoch": 0.475648, + "grad_norm": 0.40603500604629517, + "learning_rate": 0.0005635756007540494, + "loss": 2.7762, + "step": 3716 + }, + { + "epoch": 0.475776, + "grad_norm": 0.4583989679813385, + "learning_rate": 0.000563369966851748, + "loss": 2.6259, + "step": 3717 + }, + { + "epoch": 0.475904, + "grad_norm": 0.4010428786277771, + "learning_rate": 0.0005631643220554146, + "loss": 2.6942, + "step": 3718 + }, + { + "epoch": 0.476032, + "grad_norm": 0.431246817111969, + "learning_rate": 0.0005629586664004017, + "loss": 2.6307, + "step": 3719 + }, + { + "epoch": 0.47616, + "grad_norm": 0.4512070417404175, + "learning_rate": 0.0005627529999220644, + "loss": 2.6238, + "step": 3720 + }, + { + "epoch": 0.476288, + "grad_norm": 0.4383637309074402, + "learning_rate": 0.0005625473226557588, + "loss": 2.6489, + "step": 3721 + }, + { + "epoch": 0.476416, + "grad_norm": 0.4428677260875702, + "learning_rate": 0.0005623416346368434, + "loss": 2.6454, + "step": 3722 + }, + { + "epoch": 0.476544, + "grad_norm": 0.49374184012413025, + "learning_rate": 0.0005621359359006783, + "loss": 2.5758, + "step": 3723 + }, + { + "epoch": 0.476672, + "grad_norm": 0.4245469868183136, + "learning_rate": 0.0005619302264826255, + "loss": 2.6523, + "step": 3724 + }, + { + "epoch": 0.4768, + "grad_norm": 0.44209834933280945, + "learning_rate": 0.0005617245064180488, + "loss": 2.7807, + "step": 3725 + }, + { + "epoch": 0.476928, + "grad_norm": 0.4383705258369446, + "learning_rate": 0.0005615187757423137, + "loss": 2.5814, + "step": 3726 + }, + { + "epoch": 0.477056, + "grad_norm": 0.74104905128479, + "learning_rate": 0.0005613130344907881, + "loss": 2.631, + "step": 3727 + }, + { + "epoch": 0.477184, + "grad_norm": 0.4400256276130676, + "learning_rate": 0.0005611072826988409, + "loss": 2.5999, + "step": 3728 + }, + { + "epoch": 0.477312, + "grad_norm": 0.4158361554145813, + "learning_rate": 0.0005609015204018435, + "loss": 2.5939, + "step": 3729 + }, + { + "epoch": 0.47744, + "grad_norm": 0.394648015499115, + "learning_rate": 0.0005606957476351687, + "loss": 2.6997, + "step": 3730 + }, + { + "epoch": 0.477568, + "grad_norm": 0.5470306277275085, + "learning_rate": 0.0005604899644341911, + "loss": 2.6077, + "step": 3731 + }, + { + "epoch": 0.477696, + "grad_norm": 0.4924063980579376, + "learning_rate": 0.0005602841708342876, + "loss": 2.6509, + "step": 3732 + }, + { + "epoch": 0.477824, + "grad_norm": 0.4641261100769043, + "learning_rate": 0.000560078366870836, + "loss": 2.7691, + "step": 3733 + }, + { + "epoch": 0.477952, + "grad_norm": 0.46346789598464966, + "learning_rate": 0.0005598725525792168, + "loss": 2.7131, + "step": 3734 + }, + { + "epoch": 0.47808, + "grad_norm": 0.4877304434776306, + "learning_rate": 0.0005596667279948117, + "loss": 2.7515, + "step": 3735 + }, + { + "epoch": 0.478208, + "grad_norm": 0.43101781606674194, + "learning_rate": 0.0005594608931530043, + "loss": 2.6322, + "step": 3736 + }, + { + "epoch": 0.478336, + "grad_norm": 0.44474083185195923, + "learning_rate": 0.00055925504808918, + "loss": 2.6658, + "step": 3737 + }, + { + "epoch": 0.478464, + "grad_norm": 0.46803393959999084, + "learning_rate": 0.0005590491928387262, + "loss": 2.6353, + "step": 3738 + }, + { + "epoch": 0.478592, + "grad_norm": 0.41632434725761414, + "learning_rate": 0.0005588433274370313, + "loss": 2.6738, + "step": 3739 + }, + { + "epoch": 0.47872, + "grad_norm": 0.4114239513874054, + "learning_rate": 0.0005586374519194864, + "loss": 2.6882, + "step": 3740 + }, + { + "epoch": 0.478848, + "grad_norm": 0.845369279384613, + "learning_rate": 0.0005584315663214836, + "loss": 2.7244, + "step": 3741 + }, + { + "epoch": 0.478976, + "grad_norm": 0.4515933096408844, + "learning_rate": 0.0005582256706784173, + "loss": 2.6818, + "step": 3742 + }, + { + "epoch": 0.479104, + "grad_norm": 0.6133928894996643, + "learning_rate": 0.0005580197650256832, + "loss": 2.6426, + "step": 3743 + }, + { + "epoch": 0.479232, + "grad_norm": 0.46311184763908386, + "learning_rate": 0.0005578138493986788, + "loss": 2.6328, + "step": 3744 + }, + { + "epoch": 0.47936, + "grad_norm": 0.4225883185863495, + "learning_rate": 0.0005576079238328034, + "loss": 2.7125, + "step": 3745 + }, + { + "epoch": 0.479488, + "grad_norm": 0.4527849853038788, + "learning_rate": 0.0005574019883634583, + "loss": 2.7921, + "step": 3746 + }, + { + "epoch": 0.479616, + "grad_norm": 0.4324473738670349, + "learning_rate": 0.0005571960430260457, + "loss": 2.6422, + "step": 3747 + }, + { + "epoch": 0.479744, + "grad_norm": 0.4414703845977783, + "learning_rate": 0.0005569900878559704, + "loss": 2.7028, + "step": 3748 + }, + { + "epoch": 0.479872, + "grad_norm": 0.43103477358818054, + "learning_rate": 0.0005567841228886381, + "loss": 2.6607, + "step": 3749 + }, + { + "epoch": 0.48, + "grad_norm": 0.4744291603565216, + "learning_rate": 0.000556578148159457, + "loss": 2.6762, + "step": 3750 + }, + { + "epoch": 0.480128, + "grad_norm": 0.4204949736595154, + "learning_rate": 0.0005563721637038362, + "loss": 2.7709, + "step": 3751 + }, + { + "epoch": 0.480256, + "grad_norm": 0.4230082035064697, + "learning_rate": 0.000556166169557187, + "loss": 2.7344, + "step": 3752 + }, + { + "epoch": 0.480384, + "grad_norm": 0.39522430300712585, + "learning_rate": 0.0005559601657549219, + "loss": 2.6479, + "step": 3753 + }, + { + "epoch": 0.480512, + "grad_norm": 0.44803449511528015, + "learning_rate": 0.0005557541523324558, + "loss": 2.6754, + "step": 3754 + }, + { + "epoch": 0.48064, + "grad_norm": 0.44900137186050415, + "learning_rate": 0.0005555481293252044, + "loss": 2.6969, + "step": 3755 + }, + { + "epoch": 0.480768, + "grad_norm": 0.4304867088794708, + "learning_rate": 0.0005553420967685857, + "loss": 2.7179, + "step": 3756 + }, + { + "epoch": 0.480896, + "grad_norm": 0.4481363296508789, + "learning_rate": 0.0005551360546980191, + "loss": 2.7201, + "step": 3757 + }, + { + "epoch": 0.481024, + "grad_norm": 0.43246254324913025, + "learning_rate": 0.0005549300031489252, + "loss": 2.6608, + "step": 3758 + }, + { + "epoch": 0.481152, + "grad_norm": 0.3961625397205353, + "learning_rate": 0.0005547239421567271, + "loss": 2.6688, + "step": 3759 + }, + { + "epoch": 0.48128, + "grad_norm": 0.42394301295280457, + "learning_rate": 0.0005545178717568491, + "loss": 2.6428, + "step": 3760 + }, + { + "epoch": 0.481408, + "grad_norm": 0.4212658107280731, + "learning_rate": 0.0005543117919847167, + "loss": 2.7128, + "step": 3761 + }, + { + "epoch": 0.481536, + "grad_norm": 0.40259850025177, + "learning_rate": 0.0005541057028757576, + "loss": 2.6921, + "step": 3762 + }, + { + "epoch": 0.481664, + "grad_norm": 0.4516702890396118, + "learning_rate": 0.0005538996044654011, + "loss": 2.6807, + "step": 3763 + }, + { + "epoch": 0.481792, + "grad_norm": 0.422404944896698, + "learning_rate": 0.0005536934967890777, + "loss": 2.7633, + "step": 3764 + }, + { + "epoch": 0.48192, + "grad_norm": 0.5426852703094482, + "learning_rate": 0.0005534873798822196, + "loss": 2.6847, + "step": 3765 + }, + { + "epoch": 0.482048, + "grad_norm": 0.41695764660835266, + "learning_rate": 0.0005532812537802611, + "loss": 2.6525, + "step": 3766 + }, + { + "epoch": 0.482176, + "grad_norm": 0.4352109730243683, + "learning_rate": 0.0005530751185186372, + "loss": 2.6285, + "step": 3767 + }, + { + "epoch": 0.482304, + "grad_norm": 0.46236085891723633, + "learning_rate": 0.000552868974132785, + "loss": 2.6484, + "step": 3768 + }, + { + "epoch": 0.482432, + "grad_norm": 0.4753769338130951, + "learning_rate": 0.0005526628206581433, + "loss": 2.7974, + "step": 3769 + }, + { + "epoch": 0.48256, + "grad_norm": 0.40449774265289307, + "learning_rate": 0.0005524566581301525, + "loss": 2.5901, + "step": 3770 + }, + { + "epoch": 0.482688, + "grad_norm": 0.42356690764427185, + "learning_rate": 0.0005522504865842537, + "loss": 2.6528, + "step": 3771 + }, + { + "epoch": 0.482816, + "grad_norm": 0.462542325258255, + "learning_rate": 0.0005520443060558906, + "loss": 2.7082, + "step": 3772 + }, + { + "epoch": 0.482944, + "grad_norm": 0.444864422082901, + "learning_rate": 0.000551838116580508, + "loss": 2.7848, + "step": 3773 + }, + { + "epoch": 0.483072, + "grad_norm": 0.4483335614204407, + "learning_rate": 0.0005516319181935521, + "loss": 2.6369, + "step": 3774 + }, + { + "epoch": 0.4832, + "grad_norm": 0.42951545119285583, + "learning_rate": 0.0005514257109304711, + "loss": 2.6023, + "step": 3775 + }, + { + "epoch": 0.483328, + "grad_norm": 0.45386552810668945, + "learning_rate": 0.0005512194948267139, + "loss": 2.6618, + "step": 3776 + }, + { + "epoch": 0.483456, + "grad_norm": 0.5735847353935242, + "learning_rate": 0.000551013269917732, + "loss": 2.7035, + "step": 3777 + }, + { + "epoch": 0.483584, + "grad_norm": 0.417343407869339, + "learning_rate": 0.0005508070362389776, + "loss": 2.6878, + "step": 3778 + }, + { + "epoch": 0.483712, + "grad_norm": 0.43150776624679565, + "learning_rate": 0.0005506007938259045, + "loss": 2.602, + "step": 3779 + }, + { + "epoch": 0.48384, + "grad_norm": 0.4000837802886963, + "learning_rate": 0.0005503945427139684, + "loss": 2.5557, + "step": 3780 + }, + { + "epoch": 0.483968, + "grad_norm": 0.427904337644577, + "learning_rate": 0.0005501882829386263, + "loss": 2.6391, + "step": 3781 + }, + { + "epoch": 0.484096, + "grad_norm": 0.4546709954738617, + "learning_rate": 0.0005499820145353363, + "loss": 2.6327, + "step": 3782 + }, + { + "epoch": 0.484224, + "grad_norm": 0.41821029782295227, + "learning_rate": 0.0005497757375395588, + "loss": 2.6203, + "step": 3783 + }, + { + "epoch": 0.484352, + "grad_norm": 0.6692801713943481, + "learning_rate": 0.0005495694519867549, + "loss": 2.7452, + "step": 3784 + }, + { + "epoch": 0.48448, + "grad_norm": 0.668999433517456, + "learning_rate": 0.0005493631579123876, + "loss": 2.7481, + "step": 3785 + }, + { + "epoch": 0.484608, + "grad_norm": 0.4453621208667755, + "learning_rate": 0.000549156855351921, + "loss": 2.7032, + "step": 3786 + }, + { + "epoch": 0.484736, + "grad_norm": 0.433443158864975, + "learning_rate": 0.0005489505443408211, + "loss": 2.6776, + "step": 3787 + }, + { + "epoch": 0.484864, + "grad_norm": 0.465967059135437, + "learning_rate": 0.0005487442249145553, + "loss": 2.7108, + "step": 3788 + }, + { + "epoch": 0.484992, + "grad_norm": 0.4396260380744934, + "learning_rate": 0.0005485378971085921, + "loss": 2.7084, + "step": 3789 + }, + { + "epoch": 0.48512, + "grad_norm": 0.43707039952278137, + "learning_rate": 0.0005483315609584016, + "loss": 2.517, + "step": 3790 + }, + { + "epoch": 0.485248, + "grad_norm": 0.4021312892436981, + "learning_rate": 0.0005481252164994554, + "loss": 2.6134, + "step": 3791 + }, + { + "epoch": 0.485376, + "grad_norm": 0.4633592665195465, + "learning_rate": 0.0005479188637672269, + "loss": 2.7646, + "step": 3792 + }, + { + "epoch": 0.485504, + "grad_norm": 0.4592098593711853, + "learning_rate": 0.00054771250279719, + "loss": 2.716, + "step": 3793 + }, + { + "epoch": 0.485632, + "grad_norm": 0.4307038486003876, + "learning_rate": 0.0005475061336248207, + "loss": 2.6887, + "step": 3794 + }, + { + "epoch": 0.48576, + "grad_norm": 0.4110162854194641, + "learning_rate": 0.0005472997562855962, + "loss": 2.5349, + "step": 3795 + }, + { + "epoch": 0.485888, + "grad_norm": 0.43847543001174927, + "learning_rate": 0.0005470933708149955, + "loss": 2.6471, + "step": 3796 + }, + { + "epoch": 0.486016, + "grad_norm": 0.42973846197128296, + "learning_rate": 0.0005468869772484982, + "loss": 2.6206, + "step": 3797 + }, + { + "epoch": 0.486144, + "grad_norm": 0.4953716993331909, + "learning_rate": 0.000546680575621586, + "loss": 2.5784, + "step": 3798 + }, + { + "epoch": 0.486272, + "grad_norm": 0.47715187072753906, + "learning_rate": 0.0005464741659697419, + "loss": 2.6558, + "step": 3799 + }, + { + "epoch": 0.4864, + "grad_norm": 0.41248583793640137, + "learning_rate": 0.0005462677483284496, + "loss": 2.6441, + "step": 3800 + }, + { + "epoch": 0.486528, + "grad_norm": 0.42318928241729736, + "learning_rate": 0.000546061322733195, + "loss": 2.6904, + "step": 3801 + }, + { + "epoch": 0.486656, + "grad_norm": 0.5063456296920776, + "learning_rate": 0.0005458548892194651, + "loss": 2.64, + "step": 3802 + }, + { + "epoch": 0.486784, + "grad_norm": 0.4324662685394287, + "learning_rate": 0.0005456484478227482, + "loss": 2.6548, + "step": 3803 + }, + { + "epoch": 0.486912, + "grad_norm": 0.3969917297363281, + "learning_rate": 0.0005454419985785339, + "loss": 2.656, + "step": 3804 + }, + { + "epoch": 0.48704, + "grad_norm": 0.4054245352745056, + "learning_rate": 0.0005452355415223131, + "loss": 2.677, + "step": 3805 + }, + { + "epoch": 0.487168, + "grad_norm": 0.42510950565338135, + "learning_rate": 0.0005450290766895786, + "loss": 2.7108, + "step": 3806 + }, + { + "epoch": 0.487296, + "grad_norm": 0.42784297466278076, + "learning_rate": 0.0005448226041158236, + "loss": 2.5825, + "step": 3807 + }, + { + "epoch": 0.487424, + "grad_norm": 0.44128742814064026, + "learning_rate": 0.0005446161238365433, + "loss": 2.696, + "step": 3808 + }, + { + "epoch": 0.487552, + "grad_norm": 0.4970986843109131, + "learning_rate": 0.0005444096358872341, + "loss": 2.6793, + "step": 3809 + }, + { + "epoch": 0.48768, + "grad_norm": 0.5205861926078796, + "learning_rate": 0.0005442031403033937, + "loss": 2.5326, + "step": 3810 + }, + { + "epoch": 0.487808, + "grad_norm": 0.49707409739494324, + "learning_rate": 0.0005439966371205209, + "loss": 2.5351, + "step": 3811 + }, + { + "epoch": 0.487936, + "grad_norm": 0.4640769064426422, + "learning_rate": 0.0005437901263741163, + "loss": 2.6195, + "step": 3812 + }, + { + "epoch": 0.488064, + "grad_norm": 0.4480901062488556, + "learning_rate": 0.0005435836080996812, + "loss": 2.6134, + "step": 3813 + }, + { + "epoch": 0.488192, + "grad_norm": 0.4137508273124695, + "learning_rate": 0.0005433770823327186, + "loss": 2.7493, + "step": 3814 + }, + { + "epoch": 0.48832, + "grad_norm": 0.5089665651321411, + "learning_rate": 0.0005431705491087326, + "loss": 2.6853, + "step": 3815 + }, + { + "epoch": 0.488448, + "grad_norm": 0.6090633273124695, + "learning_rate": 0.000542964008463229, + "loss": 2.5759, + "step": 3816 + }, + { + "epoch": 0.488576, + "grad_norm": 0.46131548285484314, + "learning_rate": 0.0005427574604317142, + "loss": 2.6073, + "step": 3817 + }, + { + "epoch": 0.488704, + "grad_norm": 0.4706808030605316, + "learning_rate": 0.000542550905049696, + "loss": 2.6006, + "step": 3818 + }, + { + "epoch": 0.488832, + "grad_norm": 0.4734962284564972, + "learning_rate": 0.0005423443423526841, + "loss": 2.6672, + "step": 3819 + }, + { + "epoch": 0.48896, + "grad_norm": 0.5078383684158325, + "learning_rate": 0.000542137772376189, + "loss": 2.7525, + "step": 3820 + }, + { + "epoch": 0.489088, + "grad_norm": 0.4780387878417969, + "learning_rate": 0.0005419311951557221, + "loss": 2.6853, + "step": 3821 + }, + { + "epoch": 0.489216, + "grad_norm": 0.4783133268356323, + "learning_rate": 0.0005417246107267968, + "loss": 2.6751, + "step": 3822 + }, + { + "epoch": 0.489344, + "grad_norm": 0.43995487689971924, + "learning_rate": 0.0005415180191249271, + "loss": 2.6414, + "step": 3823 + }, + { + "epoch": 0.489472, + "grad_norm": 0.503720760345459, + "learning_rate": 0.0005413114203856286, + "loss": 2.615, + "step": 3824 + }, + { + "epoch": 0.4896, + "grad_norm": 0.4918985664844513, + "learning_rate": 0.0005411048145444182, + "loss": 2.6948, + "step": 3825 + }, + { + "epoch": 0.489728, + "grad_norm": 0.4510829448699951, + "learning_rate": 0.0005408982016368133, + "loss": 2.6424, + "step": 3826 + }, + { + "epoch": 0.489856, + "grad_norm": 0.4295160174369812, + "learning_rate": 0.0005406915816983336, + "loss": 2.6385, + "step": 3827 + }, + { + "epoch": 0.489984, + "grad_norm": 0.6206302642822266, + "learning_rate": 0.0005404849547644994, + "loss": 2.6244, + "step": 3828 + }, + { + "epoch": 0.490112, + "grad_norm": 0.45527786016464233, + "learning_rate": 0.0005402783208708319, + "loss": 2.5978, + "step": 3829 + }, + { + "epoch": 0.49024, + "grad_norm": 0.4228364825248718, + "learning_rate": 0.0005400716800528542, + "loss": 2.6866, + "step": 3830 + }, + { + "epoch": 0.490368, + "grad_norm": 0.469095915555954, + "learning_rate": 0.0005398650323460903, + "loss": 2.7129, + "step": 3831 + }, + { + "epoch": 0.490496, + "grad_norm": 0.46259865164756775, + "learning_rate": 0.000539658377786065, + "loss": 2.7099, + "step": 3832 + }, + { + "epoch": 0.490624, + "grad_norm": 0.46431514620780945, + "learning_rate": 0.0005394517164083047, + "loss": 2.6788, + "step": 3833 + }, + { + "epoch": 0.490752, + "grad_norm": 0.47434794902801514, + "learning_rate": 0.0005392450482483372, + "loss": 2.6426, + "step": 3834 + }, + { + "epoch": 0.49088, + "grad_norm": 0.4072410464286804, + "learning_rate": 0.0005390383733416909, + "loss": 2.5934, + "step": 3835 + }, + { + "epoch": 0.491008, + "grad_norm": 0.4412747919559479, + "learning_rate": 0.0005388316917238955, + "loss": 2.6869, + "step": 3836 + }, + { + "epoch": 0.491136, + "grad_norm": 0.414314329624176, + "learning_rate": 0.0005386250034304823, + "loss": 2.7098, + "step": 3837 + }, + { + "epoch": 0.491264, + "grad_norm": 0.43442773818969727, + "learning_rate": 0.0005384183084969831, + "loss": 2.7516, + "step": 3838 + }, + { + "epoch": 0.491392, + "grad_norm": 0.48390865325927734, + "learning_rate": 0.0005382116069589315, + "loss": 2.7078, + "step": 3839 + }, + { + "epoch": 0.49152, + "grad_norm": 0.4427437484264374, + "learning_rate": 0.0005380048988518613, + "loss": 2.6231, + "step": 3840 + }, + { + "epoch": 0.491648, + "grad_norm": 0.4694170355796814, + "learning_rate": 0.0005377981842113087, + "loss": 2.7306, + "step": 3841 + }, + { + "epoch": 0.491776, + "grad_norm": 0.43507522344589233, + "learning_rate": 0.0005375914630728099, + "loss": 2.7115, + "step": 3842 + }, + { + "epoch": 0.491904, + "grad_norm": 0.39285460114479065, + "learning_rate": 0.0005373847354719029, + "loss": 2.6627, + "step": 3843 + }, + { + "epoch": 0.492032, + "grad_norm": 0.45853787660598755, + "learning_rate": 0.0005371780014441264, + "loss": 2.635, + "step": 3844 + }, + { + "epoch": 0.49216, + "grad_norm": 0.42445603013038635, + "learning_rate": 0.0005369712610250206, + "loss": 2.5841, + "step": 3845 + }, + { + "epoch": 0.492288, + "grad_norm": 0.3723571002483368, + "learning_rate": 0.0005367645142501262, + "loss": 2.6785, + "step": 3846 + }, + { + "epoch": 0.492416, + "grad_norm": 0.41740357875823975, + "learning_rate": 0.0005365577611549855, + "loss": 2.5751, + "step": 3847 + }, + { + "epoch": 0.492544, + "grad_norm": 0.4131864905357361, + "learning_rate": 0.0005363510017751422, + "loss": 2.6288, + "step": 3848 + }, + { + "epoch": 0.492672, + "grad_norm": 0.39444902539253235, + "learning_rate": 0.0005361442361461402, + "loss": 2.6556, + "step": 3849 + }, + { + "epoch": 0.4928, + "grad_norm": 0.4297245442867279, + "learning_rate": 0.0005359374643035248, + "loss": 2.6998, + "step": 3850 + }, + { + "epoch": 0.492928, + "grad_norm": 0.4482232332229614, + "learning_rate": 0.0005357306862828427, + "loss": 2.7273, + "step": 3851 + }, + { + "epoch": 0.493056, + "grad_norm": 0.44462934136390686, + "learning_rate": 0.0005355239021196416, + "loss": 2.6451, + "step": 3852 + }, + { + "epoch": 0.493184, + "grad_norm": 0.4155009090900421, + "learning_rate": 0.0005353171118494698, + "loss": 2.5957, + "step": 3853 + }, + { + "epoch": 0.493312, + "grad_norm": 0.41241106390953064, + "learning_rate": 0.0005351103155078768, + "loss": 2.5487, + "step": 3854 + }, + { + "epoch": 0.49344, + "grad_norm": 0.44541457295417786, + "learning_rate": 0.0005349035131304138, + "loss": 2.4901, + "step": 3855 + }, + { + "epoch": 0.493568, + "grad_norm": 0.4480997622013092, + "learning_rate": 0.0005346967047526321, + "loss": 2.7034, + "step": 3856 + }, + { + "epoch": 0.493696, + "grad_norm": 0.4124755263328552, + "learning_rate": 0.0005344898904100848, + "loss": 2.6571, + "step": 3857 + }, + { + "epoch": 0.493824, + "grad_norm": 0.4042321443557739, + "learning_rate": 0.0005342830701383254, + "loss": 2.6112, + "step": 3858 + }, + { + "epoch": 0.493952, + "grad_norm": 0.44386109709739685, + "learning_rate": 0.0005340762439729088, + "loss": 2.6823, + "step": 3859 + }, + { + "epoch": 0.49408, + "grad_norm": 0.4014067053794861, + "learning_rate": 0.0005338694119493908, + "loss": 2.6308, + "step": 3860 + }, + { + "epoch": 0.494208, + "grad_norm": 0.41920560598373413, + "learning_rate": 0.0005336625741033283, + "loss": 2.5912, + "step": 3861 + }, + { + "epoch": 0.494336, + "grad_norm": 0.5951371788978577, + "learning_rate": 0.0005334557304702789, + "loss": 2.6521, + "step": 3862 + }, + { + "epoch": 0.494464, + "grad_norm": 0.4338981509208679, + "learning_rate": 0.0005332488810858017, + "loss": 2.6722, + "step": 3863 + }, + { + "epoch": 0.494592, + "grad_norm": 0.4750713109970093, + "learning_rate": 0.0005330420259854564, + "loss": 2.6857, + "step": 3864 + }, + { + "epoch": 0.49472, + "grad_norm": 1.1047511100769043, + "learning_rate": 0.0005328351652048037, + "loss": 2.7432, + "step": 3865 + }, + { + "epoch": 0.494848, + "grad_norm": 0.4662556052207947, + "learning_rate": 0.0005326282987794056, + "loss": 2.709, + "step": 3866 + }, + { + "epoch": 0.494976, + "grad_norm": 0.72990882396698, + "learning_rate": 0.0005324214267448245, + "loss": 2.6558, + "step": 3867 + }, + { + "epoch": 0.495104, + "grad_norm": 0.5773710608482361, + "learning_rate": 0.0005322145491366243, + "loss": 2.5975, + "step": 3868 + }, + { + "epoch": 0.495232, + "grad_norm": 0.44200319051742554, + "learning_rate": 0.0005320076659903698, + "loss": 2.6685, + "step": 3869 + }, + { + "epoch": 0.49536, + "grad_norm": 0.4211951792240143, + "learning_rate": 0.0005318007773416263, + "loss": 2.6208, + "step": 3870 + }, + { + "epoch": 0.495488, + "grad_norm": 0.46826452016830444, + "learning_rate": 0.0005315938832259606, + "loss": 2.6377, + "step": 3871 + }, + { + "epoch": 0.495616, + "grad_norm": 0.5078384876251221, + "learning_rate": 0.00053138698367894, + "loss": 2.6169, + "step": 3872 + }, + { + "epoch": 0.495744, + "grad_norm": 0.43319636583328247, + "learning_rate": 0.0005311800787361332, + "loss": 2.7108, + "step": 3873 + }, + { + "epoch": 0.495872, + "grad_norm": 0.440754234790802, + "learning_rate": 0.0005309731684331093, + "loss": 2.7109, + "step": 3874 + }, + { + "epoch": 0.496, + "grad_norm": 0.43541428446769714, + "learning_rate": 0.0005307662528054387, + "loss": 2.7261, + "step": 3875 + }, + { + "epoch": 0.496128, + "grad_norm": 0.43146899342536926, + "learning_rate": 0.0005305593318886924, + "loss": 2.7484, + "step": 3876 + }, + { + "epoch": 0.496256, + "grad_norm": 0.40964609384536743, + "learning_rate": 0.0005303524057184429, + "loss": 2.6847, + "step": 3877 + }, + { + "epoch": 0.496384, + "grad_norm": 0.4310256242752075, + "learning_rate": 0.0005301454743302629, + "loss": 2.6583, + "step": 3878 + }, + { + "epoch": 0.496512, + "grad_norm": 0.39927583932876587, + "learning_rate": 0.0005299385377597265, + "loss": 2.6526, + "step": 3879 + }, + { + "epoch": 0.49664, + "grad_norm": 0.48663222789764404, + "learning_rate": 0.0005297315960424081, + "loss": 2.6504, + "step": 3880 + }, + { + "epoch": 0.496768, + "grad_norm": 0.49432533979415894, + "learning_rate": 0.0005295246492138839, + "loss": 2.6243, + "step": 3881 + }, + { + "epoch": 0.496896, + "grad_norm": 0.9528817534446716, + "learning_rate": 0.0005293176973097302, + "loss": 2.673, + "step": 3882 + }, + { + "epoch": 0.497024, + "grad_norm": 0.4488283693790436, + "learning_rate": 0.0005291107403655242, + "loss": 2.6863, + "step": 3883 + }, + { + "epoch": 0.497152, + "grad_norm": 0.45814424753189087, + "learning_rate": 0.0005289037784168448, + "loss": 2.7212, + "step": 3884 + }, + { + "epoch": 0.49728, + "grad_norm": 0.45292603969573975, + "learning_rate": 0.0005286968114992706, + "loss": 2.6407, + "step": 3885 + }, + { + "epoch": 0.497408, + "grad_norm": 0.5074853897094727, + "learning_rate": 0.0005284898396483819, + "loss": 2.7059, + "step": 3886 + }, + { + "epoch": 0.497536, + "grad_norm": 0.4621811807155609, + "learning_rate": 0.0005282828628997595, + "loss": 2.799, + "step": 3887 + }, + { + "epoch": 0.497664, + "grad_norm": 0.45863792300224304, + "learning_rate": 0.000528075881288985, + "loss": 2.6383, + "step": 3888 + }, + { + "epoch": 0.497792, + "grad_norm": 0.4181120693683624, + "learning_rate": 0.0005278688948516412, + "loss": 2.5482, + "step": 3889 + }, + { + "epoch": 0.49792, + "grad_norm": 0.4085069000720978, + "learning_rate": 0.0005276619036233111, + "loss": 2.5981, + "step": 3890 + }, + { + "epoch": 0.498048, + "grad_norm": 0.44205352663993835, + "learning_rate": 0.0005274549076395792, + "loss": 2.6483, + "step": 3891 + }, + { + "epoch": 0.498176, + "grad_norm": 0.4058488607406616, + "learning_rate": 0.0005272479069360302, + "loss": 2.6613, + "step": 3892 + }, + { + "epoch": 0.498304, + "grad_norm": 0.5268768668174744, + "learning_rate": 0.0005270409015482503, + "loss": 2.7821, + "step": 3893 + }, + { + "epoch": 0.498432, + "grad_norm": 0.40639176964759827, + "learning_rate": 0.0005268338915118258, + "loss": 2.7475, + "step": 3894 + }, + { + "epoch": 0.49856, + "grad_norm": 0.454645037651062, + "learning_rate": 0.0005266268768623444, + "loss": 2.6594, + "step": 3895 + }, + { + "epoch": 0.498688, + "grad_norm": 0.4685520529747009, + "learning_rate": 0.0005264198576353942, + "loss": 2.7097, + "step": 3896 + }, + { + "epoch": 0.498816, + "grad_norm": 0.4289955198764801, + "learning_rate": 0.0005262128338665641, + "loss": 2.7411, + "step": 3897 + }, + { + "epoch": 0.498944, + "grad_norm": 0.4607875645160675, + "learning_rate": 0.000526005805591444, + "loss": 2.5897, + "step": 3898 + }, + { + "epoch": 0.499072, + "grad_norm": 0.49246251583099365, + "learning_rate": 0.0005257987728456244, + "loss": 2.6905, + "step": 3899 + }, + { + "epoch": 0.4992, + "grad_norm": 0.38600799441337585, + "learning_rate": 0.0005255917356646968, + "loss": 2.5355, + "step": 3900 + }, + { + "epoch": 0.499328, + "grad_norm": 0.43069031834602356, + "learning_rate": 0.0005253846940842528, + "loss": 2.7546, + "step": 3901 + }, + { + "epoch": 0.499456, + "grad_norm": 0.42126476764678955, + "learning_rate": 0.000525177648139886, + "loss": 2.7258, + "step": 3902 + }, + { + "epoch": 0.499584, + "grad_norm": 0.4198932945728302, + "learning_rate": 0.0005249705978671895, + "loss": 2.6621, + "step": 3903 + }, + { + "epoch": 0.499712, + "grad_norm": 0.4117319583892822, + "learning_rate": 0.0005247635433017576, + "loss": 2.7039, + "step": 3904 + }, + { + "epoch": 0.49984, + "grad_norm": 0.4263758361339569, + "learning_rate": 0.0005245564844791857, + "loss": 2.6702, + "step": 3905 + }, + { + "epoch": 0.499968, + "grad_norm": 0.4072818458080292, + "learning_rate": 0.0005243494214350693, + "loss": 2.5431, + "step": 3906 + }, + { + "epoch": 0.500096, + "grad_norm": 0.3844020366668701, + "learning_rate": 0.0005241423542050051, + "loss": 2.644, + "step": 3907 + }, + { + "epoch": 0.500224, + "grad_norm": 0.44682496786117554, + "learning_rate": 0.0005239352828245904, + "loss": 2.7607, + "step": 3908 + }, + { + "epoch": 0.500352, + "grad_norm": 0.43886733055114746, + "learning_rate": 0.000523728207329423, + "loss": 2.6553, + "step": 3909 + }, + { + "epoch": 0.50048, + "grad_norm": 0.4739525318145752, + "learning_rate": 0.0005235211277551018, + "loss": 2.6626, + "step": 3910 + }, + { + "epoch": 0.500608, + "grad_norm": 0.40852072834968567, + "learning_rate": 0.000523314044137226, + "loss": 2.6759, + "step": 3911 + }, + { + "epoch": 0.500736, + "grad_norm": 0.4266478717327118, + "learning_rate": 0.0005231069565113957, + "loss": 2.6162, + "step": 3912 + }, + { + "epoch": 0.500864, + "grad_norm": 0.40479058027267456, + "learning_rate": 0.0005228998649132118, + "loss": 2.6851, + "step": 3913 + }, + { + "epoch": 0.500992, + "grad_norm": 0.5709474682807922, + "learning_rate": 0.0005226927693782754, + "loss": 2.6923, + "step": 3914 + }, + { + "epoch": 0.50112, + "grad_norm": 0.44608768820762634, + "learning_rate": 0.0005224856699421889, + "loss": 2.8049, + "step": 3915 + }, + { + "epoch": 0.501248, + "grad_norm": 0.4100112020969391, + "learning_rate": 0.0005222785666405553, + "loss": 2.5785, + "step": 3916 + }, + { + "epoch": 0.501376, + "grad_norm": 0.3949468731880188, + "learning_rate": 0.0005220714595089777, + "loss": 2.5817, + "step": 3917 + }, + { + "epoch": 0.501504, + "grad_norm": 0.4276130497455597, + "learning_rate": 0.0005218643485830605, + "loss": 2.6424, + "step": 3918 + }, + { + "epoch": 0.501632, + "grad_norm": 0.44696810841560364, + "learning_rate": 0.000521657233898408, + "loss": 2.6103, + "step": 3919 + }, + { + "epoch": 0.50176, + "grad_norm": 0.3952341377735138, + "learning_rate": 0.0005214501154906261, + "loss": 2.6057, + "step": 3920 + }, + { + "epoch": 0.501888, + "grad_norm": 0.389127254486084, + "learning_rate": 0.0005212429933953206, + "loss": 2.6684, + "step": 3921 + }, + { + "epoch": 0.502016, + "grad_norm": 0.4341872036457062, + "learning_rate": 0.0005210358676480983, + "loss": 2.6773, + "step": 3922 + }, + { + "epoch": 0.502144, + "grad_norm": 0.4407554566860199, + "learning_rate": 0.0005208287382845666, + "loss": 2.5916, + "step": 3923 + }, + { + "epoch": 0.502272, + "grad_norm": 0.4845099449157715, + "learning_rate": 0.0005206216053403333, + "loss": 2.6728, + "step": 3924 + }, + { + "epoch": 0.5024, + "grad_norm": 0.41734105348587036, + "learning_rate": 0.0005204144688510068, + "loss": 2.6129, + "step": 3925 + }, + { + "epoch": 0.502528, + "grad_norm": 0.41436612606048584, + "learning_rate": 0.0005202073288521964, + "loss": 2.6284, + "step": 3926 + }, + { + "epoch": 0.502656, + "grad_norm": 0.3756304383277893, + "learning_rate": 0.0005200001853795123, + "loss": 2.6988, + "step": 3927 + }, + { + "epoch": 0.502784, + "grad_norm": 0.37071526050567627, + "learning_rate": 0.0005197930384685642, + "loss": 2.5071, + "step": 3928 + }, + { + "epoch": 0.502912, + "grad_norm": 0.39747411012649536, + "learning_rate": 0.0005195858881549632, + "loss": 2.7985, + "step": 3929 + }, + { + "epoch": 0.50304, + "grad_norm": 0.43258538842201233, + "learning_rate": 0.000519378734474321, + "loss": 2.6137, + "step": 3930 + }, + { + "epoch": 0.503168, + "grad_norm": 0.43841633200645447, + "learning_rate": 0.0005191715774622499, + "loss": 2.7778, + "step": 3931 + }, + { + "epoch": 0.503296, + "grad_norm": 0.41932666301727295, + "learning_rate": 0.0005189644171543624, + "loss": 2.6326, + "step": 3932 + }, + { + "epoch": 0.503424, + "grad_norm": 0.3787570595741272, + "learning_rate": 0.0005187572535862715, + "loss": 2.6714, + "step": 3933 + }, + { + "epoch": 0.503552, + "grad_norm": 0.4544612467288971, + "learning_rate": 0.0005185500867935913, + "loss": 2.5545, + "step": 3934 + }, + { + "epoch": 0.50368, + "grad_norm": 0.4061710238456726, + "learning_rate": 0.0005183429168119362, + "loss": 2.6552, + "step": 3935 + }, + { + "epoch": 0.503808, + "grad_norm": 0.43091699481010437, + "learning_rate": 0.000518135743676921, + "loss": 2.6606, + "step": 3936 + }, + { + "epoch": 0.503936, + "grad_norm": 0.4195031225681305, + "learning_rate": 0.0005179285674241614, + "loss": 2.6837, + "step": 3937 + }, + { + "epoch": 0.504064, + "grad_norm": 0.39230048656463623, + "learning_rate": 0.0005177213880892729, + "loss": 2.6854, + "step": 3938 + }, + { + "epoch": 0.504192, + "grad_norm": 0.44561275839805603, + "learning_rate": 0.0005175142057078727, + "loss": 2.6593, + "step": 3939 + }, + { + "epoch": 0.50432, + "grad_norm": 0.5589061975479126, + "learning_rate": 0.0005173070203155772, + "loss": 2.679, + "step": 3940 + }, + { + "epoch": 0.504448, + "grad_norm": 0.4336915910243988, + "learning_rate": 0.0005170998319480043, + "loss": 2.5968, + "step": 3941 + }, + { + "epoch": 0.504576, + "grad_norm": 0.4155622124671936, + "learning_rate": 0.0005168926406407722, + "loss": 2.6227, + "step": 3942 + }, + { + "epoch": 0.504704, + "grad_norm": 0.4537605345249176, + "learning_rate": 0.0005166854464294993, + "loss": 2.6322, + "step": 3943 + }, + { + "epoch": 0.504832, + "grad_norm": 0.4417303204536438, + "learning_rate": 0.0005164782493498046, + "loss": 2.6954, + "step": 3944 + }, + { + "epoch": 0.50496, + "grad_norm": 0.453670471906662, + "learning_rate": 0.000516271049437308, + "loss": 2.5739, + "step": 3945 + }, + { + "epoch": 0.505088, + "grad_norm": 0.43824470043182373, + "learning_rate": 0.0005160638467276293, + "loss": 2.5704, + "step": 3946 + }, + { + "epoch": 0.505216, + "grad_norm": 0.46332788467407227, + "learning_rate": 0.000515856641256389, + "loss": 2.6935, + "step": 3947 + }, + { + "epoch": 0.505344, + "grad_norm": 0.4481138288974762, + "learning_rate": 0.0005156494330592083, + "loss": 2.6556, + "step": 3948 + }, + { + "epoch": 0.505472, + "grad_norm": 0.4427787661552429, + "learning_rate": 0.0005154422221717086, + "loss": 2.771, + "step": 3949 + }, + { + "epoch": 0.5056, + "grad_norm": 0.40499767661094666, + "learning_rate": 0.0005152350086295121, + "loss": 2.5441, + "step": 3950 + }, + { + "epoch": 0.505728, + "grad_norm": 0.38289305567741394, + "learning_rate": 0.0005150277924682408, + "loss": 2.6781, + "step": 3951 + }, + { + "epoch": 0.505856, + "grad_norm": 0.4336622953414917, + "learning_rate": 0.0005148205737235177, + "loss": 2.6315, + "step": 3952 + }, + { + "epoch": 0.505984, + "grad_norm": 0.4525275230407715, + "learning_rate": 0.0005146133524309664, + "loss": 2.6787, + "step": 3953 + }, + { + "epoch": 0.506112, + "grad_norm": 0.4053783714771271, + "learning_rate": 0.0005144061286262102, + "loss": 2.6947, + "step": 3954 + }, + { + "epoch": 0.50624, + "grad_norm": 0.44160932302474976, + "learning_rate": 0.0005141989023448735, + "loss": 2.5798, + "step": 3955 + }, + { + "epoch": 0.506368, + "grad_norm": 0.39990928769111633, + "learning_rate": 0.000513991673622581, + "loss": 2.7084, + "step": 3956 + }, + { + "epoch": 0.506496, + "grad_norm": 0.4035250246524811, + "learning_rate": 0.0005137844424949575, + "loss": 2.6534, + "step": 3957 + }, + { + "epoch": 0.506624, + "grad_norm": 0.40214812755584717, + "learning_rate": 0.0005135772089976285, + "loss": 2.5488, + "step": 3958 + }, + { + "epoch": 0.506752, + "grad_norm": 0.42011764645576477, + "learning_rate": 0.00051336997316622, + "loss": 2.6597, + "step": 3959 + }, + { + "epoch": 0.50688, + "grad_norm": 0.420714408159256, + "learning_rate": 0.0005131627350363581, + "loss": 2.6936, + "step": 3960 + }, + { + "epoch": 0.507008, + "grad_norm": 0.4046907126903534, + "learning_rate": 0.0005129554946436694, + "loss": 2.5585, + "step": 3961 + }, + { + "epoch": 0.507136, + "grad_norm": 0.41296398639678955, + "learning_rate": 0.0005127482520237808, + "loss": 2.6857, + "step": 3962 + }, + { + "epoch": 0.507264, + "grad_norm": 0.4998627007007599, + "learning_rate": 0.0005125410072123202, + "loss": 2.7587, + "step": 3963 + }, + { + "epoch": 0.507392, + "grad_norm": 0.4240685701370239, + "learning_rate": 0.000512333760244915, + "loss": 2.782, + "step": 3964 + }, + { + "epoch": 0.50752, + "grad_norm": 0.4455365836620331, + "learning_rate": 0.0005121265111571933, + "loss": 2.7218, + "step": 3965 + }, + { + "epoch": 0.507648, + "grad_norm": 0.41035640239715576, + "learning_rate": 0.0005119192599847837, + "loss": 2.6314, + "step": 3966 + }, + { + "epoch": 0.507776, + "grad_norm": 0.42151129245758057, + "learning_rate": 0.0005117120067633153, + "loss": 2.8258, + "step": 3967 + }, + { + "epoch": 0.507904, + "grad_norm": 0.40187808871269226, + "learning_rate": 0.0005115047515284173, + "loss": 2.6692, + "step": 3968 + }, + { + "epoch": 0.508032, + "grad_norm": 0.38517531752586365, + "learning_rate": 0.0005112974943157189, + "loss": 2.5534, + "step": 3969 + }, + { + "epoch": 0.50816, + "grad_norm": 0.42202627658843994, + "learning_rate": 0.0005110902351608503, + "loss": 2.4519, + "step": 3970 + }, + { + "epoch": 0.508288, + "grad_norm": 0.39083683490753174, + "learning_rate": 0.0005108829740994417, + "loss": 2.6117, + "step": 3971 + }, + { + "epoch": 0.508416, + "grad_norm": 0.42481496930122375, + "learning_rate": 0.0005106757111671235, + "loss": 2.6151, + "step": 3972 + }, + { + "epoch": 0.508544, + "grad_norm": 0.39872315526008606, + "learning_rate": 0.0005104684463995271, + "loss": 2.6873, + "step": 3973 + }, + { + "epoch": 0.508672, + "grad_norm": 0.4124671220779419, + "learning_rate": 0.0005102611798322833, + "loss": 2.7509, + "step": 3974 + }, + { + "epoch": 0.5088, + "grad_norm": 0.3670516908168793, + "learning_rate": 0.0005100539115010236, + "loss": 2.5553, + "step": 3975 + }, + { + "epoch": 0.508928, + "grad_norm": 0.4196142852306366, + "learning_rate": 0.00050984664144138, + "loss": 2.6157, + "step": 3976 + }, + { + "epoch": 0.509056, + "grad_norm": 0.4042658805847168, + "learning_rate": 0.0005096393696889847, + "loss": 2.6181, + "step": 3977 + }, + { + "epoch": 0.509184, + "grad_norm": 0.5241581201553345, + "learning_rate": 0.0005094320962794701, + "loss": 2.6249, + "step": 3978 + }, + { + "epoch": 0.509312, + "grad_norm": 0.4300318658351898, + "learning_rate": 0.0005092248212484686, + "loss": 2.7282, + "step": 3979 + }, + { + "epoch": 0.50944, + "grad_norm": 0.4211404025554657, + "learning_rate": 0.0005090175446316133, + "loss": 2.6968, + "step": 3980 + }, + { + "epoch": 0.509568, + "grad_norm": 0.45848745107650757, + "learning_rate": 0.0005088102664645378, + "loss": 2.6882, + "step": 3981 + }, + { + "epoch": 0.509696, + "grad_norm": 0.42045632004737854, + "learning_rate": 0.0005086029867828752, + "loss": 2.6834, + "step": 3982 + }, + { + "epoch": 0.509824, + "grad_norm": 0.4007071852684021, + "learning_rate": 0.0005083957056222594, + "loss": 2.6785, + "step": 3983 + }, + { + "epoch": 0.509952, + "grad_norm": 0.39284345507621765, + "learning_rate": 0.0005081884230183247, + "loss": 2.5733, + "step": 3984 + }, + { + "epoch": 0.51008, + "grad_norm": 0.39687618613243103, + "learning_rate": 0.0005079811390067048, + "loss": 2.5637, + "step": 3985 + }, + { + "epoch": 0.510208, + "grad_norm": 0.39252978563308716, + "learning_rate": 0.000507773853623035, + "loss": 2.6012, + "step": 3986 + }, + { + "epoch": 0.510336, + "grad_norm": 0.43734800815582275, + "learning_rate": 0.0005075665669029494, + "loss": 2.6211, + "step": 3987 + }, + { + "epoch": 0.510464, + "grad_norm": 0.44745996594429016, + "learning_rate": 0.0005073592788820832, + "loss": 2.6498, + "step": 3988 + }, + { + "epoch": 0.510592, + "grad_norm": 0.4072677493095398, + "learning_rate": 0.0005071519895960717, + "loss": 2.6316, + "step": 3989 + }, + { + "epoch": 0.51072, + "grad_norm": 0.41475772857666016, + "learning_rate": 0.0005069446990805501, + "loss": 2.5861, + "step": 3990 + }, + { + "epoch": 0.510848, + "grad_norm": 0.4990641176700592, + "learning_rate": 0.0005067374073711544, + "loss": 2.6897, + "step": 3991 + }, + { + "epoch": 0.510976, + "grad_norm": 0.40266168117523193, + "learning_rate": 0.0005065301145035204, + "loss": 2.6162, + "step": 3992 + }, + { + "epoch": 0.511104, + "grad_norm": 0.44434821605682373, + "learning_rate": 0.0005063228205132839, + "loss": 2.6079, + "step": 3993 + }, + { + "epoch": 0.511232, + "grad_norm": 0.42097529768943787, + "learning_rate": 0.000506115525436081, + "loss": 2.7356, + "step": 3994 + }, + { + "epoch": 0.51136, + "grad_norm": 0.39932605624198914, + "learning_rate": 0.0005059082293075487, + "loss": 2.6273, + "step": 3995 + }, + { + "epoch": 0.511488, + "grad_norm": 0.40529486536979675, + "learning_rate": 0.0005057009321633233, + "loss": 2.7133, + "step": 3996 + }, + { + "epoch": 0.511616, + "grad_norm": 0.3790394365787506, + "learning_rate": 0.0005054936340390414, + "loss": 2.6067, + "step": 3997 + }, + { + "epoch": 0.511744, + "grad_norm": 0.4078729450702667, + "learning_rate": 0.0005052863349703402, + "loss": 2.6839, + "step": 3998 + }, + { + "epoch": 0.511872, + "grad_norm": 0.4046071767807007, + "learning_rate": 0.0005050790349928566, + "loss": 2.77, + "step": 3999 + }, + { + "epoch": 0.512, + "grad_norm": 0.3799568712711334, + "learning_rate": 0.0005048717341422282, + "loss": 2.7695, + "step": 4000 + }, + { + "epoch": 0.512128, + "grad_norm": 0.399759441614151, + "learning_rate": 0.0005046644324540921, + "loss": 2.6236, + "step": 4001 + }, + { + "epoch": 0.512256, + "grad_norm": 0.3925498425960541, + "learning_rate": 0.000504457129964086, + "loss": 2.6476, + "step": 4002 + }, + { + "epoch": 0.512384, + "grad_norm": 0.40816423296928406, + "learning_rate": 0.0005042498267078476, + "loss": 2.6954, + "step": 4003 + }, + { + "epoch": 0.512512, + "grad_norm": 0.40665721893310547, + "learning_rate": 0.0005040425227210148, + "loss": 2.5916, + "step": 4004 + }, + { + "epoch": 0.51264, + "grad_norm": 0.3761599361896515, + "learning_rate": 0.0005038352180392254, + "loss": 2.5954, + "step": 4005 + }, + { + "epoch": 0.512768, + "grad_norm": 0.43202078342437744, + "learning_rate": 0.0005036279126981176, + "loss": 2.59, + "step": 4006 + }, + { + "epoch": 0.512896, + "grad_norm": 0.3927147388458252, + "learning_rate": 0.0005034206067333296, + "loss": 2.621, + "step": 4007 + }, + { + "epoch": 0.513024, + "grad_norm": 0.40252193808555603, + "learning_rate": 0.0005032133001804993, + "loss": 2.5281, + "step": 4008 + }, + { + "epoch": 0.513152, + "grad_norm": 0.3846836984157562, + "learning_rate": 0.0005030059930752658, + "loss": 2.4783, + "step": 4009 + }, + { + "epoch": 0.51328, + "grad_norm": 0.4078712463378906, + "learning_rate": 0.0005027986854532672, + "loss": 2.701, + "step": 4010 + }, + { + "epoch": 0.513408, + "grad_norm": 0.4123474359512329, + "learning_rate": 0.0005025913773501421, + "loss": 2.7332, + "step": 4011 + }, + { + "epoch": 0.513536, + "grad_norm": 0.4135470390319824, + "learning_rate": 0.0005023840688015291, + "loss": 2.6579, + "step": 4012 + }, + { + "epoch": 0.513664, + "grad_norm": 0.4171890914440155, + "learning_rate": 0.0005021767598430671, + "loss": 2.7423, + "step": 4013 + }, + { + "epoch": 0.513792, + "grad_norm": 0.38908499479293823, + "learning_rate": 0.0005019694505103948, + "loss": 2.6703, + "step": 4014 + }, + { + "epoch": 0.51392, + "grad_norm": 0.4309137165546417, + "learning_rate": 0.0005017621408391513, + "loss": 2.6905, + "step": 4015 + }, + { + "epoch": 0.514048, + "grad_norm": 0.4100412428379059, + "learning_rate": 0.0005015548308649752, + "loss": 2.6662, + "step": 4016 + }, + { + "epoch": 0.514176, + "grad_norm": 0.4000186324119568, + "learning_rate": 0.0005013475206235056, + "loss": 2.6758, + "step": 4017 + }, + { + "epoch": 0.514304, + "grad_norm": 0.378866046667099, + "learning_rate": 0.0005011402101503818, + "loss": 2.5533, + "step": 4018 + }, + { + "epoch": 0.514432, + "grad_norm": 0.378206342458725, + "learning_rate": 0.0005009328994812425, + "loss": 2.6005, + "step": 4019 + }, + { + "epoch": 0.51456, + "grad_norm": 0.7311468124389648, + "learning_rate": 0.000500725588651727, + "loss": 2.5477, + "step": 4020 + }, + { + "epoch": 0.514688, + "grad_norm": 0.4048193693161011, + "learning_rate": 0.0005005182776974745, + "loss": 2.6789, + "step": 4021 + }, + { + "epoch": 0.514816, + "grad_norm": 0.4024932086467743, + "learning_rate": 0.0005003109666541238, + "loss": 2.7148, + "step": 4022 + }, + { + "epoch": 0.514944, + "grad_norm": 0.4188704788684845, + "learning_rate": 0.0005001036555573145, + "loss": 2.5922, + "step": 4023 + }, + { + "epoch": 0.515072, + "grad_norm": 0.4515575170516968, + "learning_rate": 0.0004998963444426855, + "loss": 2.6029, + "step": 4024 + }, + { + "epoch": 0.5152, + "grad_norm": 0.4547857940196991, + "learning_rate": 0.0004996890333458762, + "loss": 2.5765, + "step": 4025 + }, + { + "epoch": 0.515328, + "grad_norm": 0.39009520411491394, + "learning_rate": 0.0004994817223025256, + "loss": 2.6858, + "step": 4026 + }, + { + "epoch": 0.515456, + "grad_norm": 0.3988005518913269, + "learning_rate": 0.000499274411348273, + "loss": 2.782, + "step": 4027 + }, + { + "epoch": 0.515584, + "grad_norm": 0.41151976585388184, + "learning_rate": 0.0004990671005187576, + "loss": 2.6403, + "step": 4028 + }, + { + "epoch": 0.515712, + "grad_norm": 0.4034416377544403, + "learning_rate": 0.0004988597898496183, + "loss": 2.659, + "step": 4029 + }, + { + "epoch": 0.51584, + "grad_norm": 0.4047340452671051, + "learning_rate": 0.0004986524793764945, + "loss": 2.577, + "step": 4030 + }, + { + "epoch": 0.515968, + "grad_norm": 0.39854365587234497, + "learning_rate": 0.0004984451691350249, + "loss": 2.6418, + "step": 4031 + }, + { + "epoch": 0.516096, + "grad_norm": 0.4442670941352844, + "learning_rate": 0.0004982378591608488, + "loss": 2.7179, + "step": 4032 + }, + { + "epoch": 0.516224, + "grad_norm": 0.41467252373695374, + "learning_rate": 0.0004980305494896052, + "loss": 2.6729, + "step": 4033 + }, + { + "epoch": 0.516352, + "grad_norm": 0.3805702328681946, + "learning_rate": 0.0004978232401569329, + "loss": 2.6117, + "step": 4034 + }, + { + "epoch": 0.51648, + "grad_norm": 0.38779574632644653, + "learning_rate": 0.000497615931198471, + "loss": 2.6637, + "step": 4035 + }, + { + "epoch": 0.516608, + "grad_norm": 0.415041446685791, + "learning_rate": 0.0004974086226498581, + "loss": 2.643, + "step": 4036 + }, + { + "epoch": 0.516736, + "grad_norm": 0.4005393981933594, + "learning_rate": 0.0004972013145467329, + "loss": 2.7391, + "step": 4037 + }, + { + "epoch": 0.516864, + "grad_norm": 0.44793012738227844, + "learning_rate": 0.0004969940069247343, + "loss": 2.6044, + "step": 4038 + }, + { + "epoch": 0.516992, + "grad_norm": 0.40833261609077454, + "learning_rate": 0.0004967866998195008, + "loss": 2.6147, + "step": 4039 + }, + { + "epoch": 0.51712, + "grad_norm": 0.41289201378822327, + "learning_rate": 0.0004965793932666707, + "loss": 2.6444, + "step": 4040 + }, + { + "epoch": 0.517248, + "grad_norm": 0.4314199686050415, + "learning_rate": 0.0004963720873018827, + "loss": 2.7931, + "step": 4041 + }, + { + "epoch": 0.517376, + "grad_norm": 0.42754456400871277, + "learning_rate": 0.0004961647819607749, + "loss": 2.7412, + "step": 4042 + }, + { + "epoch": 0.517504, + "grad_norm": 0.42214858531951904, + "learning_rate": 0.0004959574772789853, + "loss": 2.637, + "step": 4043 + }, + { + "epoch": 0.517632, + "grad_norm": 0.4341021180152893, + "learning_rate": 0.0004957501732921524, + "loss": 2.6375, + "step": 4044 + }, + { + "epoch": 0.51776, + "grad_norm": 0.408645361661911, + "learning_rate": 0.000495542870035914, + "loss": 2.6934, + "step": 4045 + }, + { + "epoch": 0.517888, + "grad_norm": 0.4056028425693512, + "learning_rate": 0.000495335567545908, + "loss": 2.6608, + "step": 4046 + }, + { + "epoch": 0.518016, + "grad_norm": 0.4157329797744751, + "learning_rate": 0.0004951282658577718, + "loss": 2.6351, + "step": 4047 + }, + { + "epoch": 0.518144, + "grad_norm": 0.40591564774513245, + "learning_rate": 0.0004949209650071435, + "loss": 2.6134, + "step": 4048 + }, + { + "epoch": 0.518272, + "grad_norm": 0.4737180471420288, + "learning_rate": 0.0004947136650296601, + "loss": 2.6613, + "step": 4049 + }, + { + "epoch": 0.5184, + "grad_norm": 0.4482850730419159, + "learning_rate": 0.0004945063659609587, + "loss": 2.6168, + "step": 4050 + }, + { + "epoch": 0.518528, + "grad_norm": 0.4057692289352417, + "learning_rate": 0.0004942990678366769, + "loss": 2.6155, + "step": 4051 + }, + { + "epoch": 0.518656, + "grad_norm": 0.4006463289260864, + "learning_rate": 0.0004940917706924513, + "loss": 2.6807, + "step": 4052 + }, + { + "epoch": 0.518784, + "grad_norm": 0.4178115427494049, + "learning_rate": 0.000493884474563919, + "loss": 2.7427, + "step": 4053 + }, + { + "epoch": 0.518912, + "grad_norm": 0.4632314443588257, + "learning_rate": 0.0004936771794867163, + "loss": 2.5982, + "step": 4054 + }, + { + "epoch": 0.51904, + "grad_norm": 0.38948336243629456, + "learning_rate": 0.0004934698854964797, + "loss": 2.6655, + "step": 4055 + }, + { + "epoch": 0.519168, + "grad_norm": 0.4233257472515106, + "learning_rate": 0.0004932625926288456, + "loss": 2.7923, + "step": 4056 + }, + { + "epoch": 0.519296, + "grad_norm": 0.443207323551178, + "learning_rate": 0.0004930553009194499, + "loss": 2.7249, + "step": 4057 + }, + { + "epoch": 0.519424, + "grad_norm": 0.3937855064868927, + "learning_rate": 0.0004928480104039284, + "loss": 2.6961, + "step": 4058 + }, + { + "epoch": 0.519552, + "grad_norm": 0.419141948223114, + "learning_rate": 0.000492640721117917, + "loss": 2.6008, + "step": 4059 + }, + { + "epoch": 0.51968, + "grad_norm": 0.4858699142932892, + "learning_rate": 0.0004924334330970509, + "loss": 2.6093, + "step": 4060 + }, + { + "epoch": 0.519808, + "grad_norm": 0.39398401975631714, + "learning_rate": 0.0004922261463769651, + "loss": 2.7745, + "step": 4061 + }, + { + "epoch": 0.519936, + "grad_norm": 0.374613493680954, + "learning_rate": 0.0004920188609932951, + "loss": 2.5367, + "step": 4062 + }, + { + "epoch": 0.520064, + "grad_norm": 0.41044381260871887, + "learning_rate": 0.0004918115769816754, + "loss": 2.5862, + "step": 4063 + }, + { + "epoch": 0.520192, + "grad_norm": 0.41931572556495667, + "learning_rate": 0.0004916042943777406, + "loss": 2.6819, + "step": 4064 + }, + { + "epoch": 0.52032, + "grad_norm": 0.37891700863838196, + "learning_rate": 0.0004913970132171248, + "loss": 2.6521, + "step": 4065 + }, + { + "epoch": 0.520448, + "grad_norm": 0.4320984184741974, + "learning_rate": 0.0004911897335354624, + "loss": 2.6791, + "step": 4066 + }, + { + "epoch": 0.520576, + "grad_norm": 0.4022679030895233, + "learning_rate": 0.0004909824553683867, + "loss": 2.6887, + "step": 4067 + }, + { + "epoch": 0.520704, + "grad_norm": 0.396770179271698, + "learning_rate": 0.0004907751787515315, + "loss": 2.6103, + "step": 4068 + }, + { + "epoch": 0.520832, + "grad_norm": 0.4266141653060913, + "learning_rate": 0.0004905679037205302, + "loss": 2.604, + "step": 4069 + }, + { + "epoch": 0.52096, + "grad_norm": 0.4201093912124634, + "learning_rate": 0.0004903606303110152, + "loss": 2.6406, + "step": 4070 + }, + { + "epoch": 0.521088, + "grad_norm": 0.49153074622154236, + "learning_rate": 0.0004901533585586199, + "loss": 2.5602, + "step": 4071 + }, + { + "epoch": 0.521216, + "grad_norm": 0.39519134163856506, + "learning_rate": 0.0004899460884989764, + "loss": 2.6469, + "step": 4072 + }, + { + "epoch": 0.521344, + "grad_norm": 0.4154376983642578, + "learning_rate": 0.0004897388201677168, + "loss": 2.6027, + "step": 4073 + }, + { + "epoch": 0.521472, + "grad_norm": 0.4107426702976227, + "learning_rate": 0.000489531553600473, + "loss": 2.62, + "step": 4074 + }, + { + "epoch": 0.5216, + "grad_norm": 0.40223878622055054, + "learning_rate": 0.0004893242888328765, + "loss": 2.6568, + "step": 4075 + }, + { + "epoch": 0.521728, + "grad_norm": 0.43982499837875366, + "learning_rate": 0.0004891170259005585, + "loss": 2.7427, + "step": 4076 + }, + { + "epoch": 0.521856, + "grad_norm": 0.4365213215351105, + "learning_rate": 0.00048890976483915, + "loss": 2.6268, + "step": 4077 + }, + { + "epoch": 0.521984, + "grad_norm": 0.40938010811805725, + "learning_rate": 0.0004887025056842814, + "loss": 2.6852, + "step": 4078 + }, + { + "epoch": 0.522112, + "grad_norm": 0.41548168659210205, + "learning_rate": 0.000488495248471583, + "loss": 2.6384, + "step": 4079 + }, + { + "epoch": 0.52224, + "grad_norm": 0.47277066111564636, + "learning_rate": 0.0004882879932366846, + "loss": 2.6138, + "step": 4080 + }, + { + "epoch": 0.522368, + "grad_norm": 0.40378767251968384, + "learning_rate": 0.0004880807400152162, + "loss": 2.6946, + "step": 4081 + }, + { + "epoch": 0.522496, + "grad_norm": 0.3841347396373749, + "learning_rate": 0.0004878734888428068, + "loss": 2.7078, + "step": 4082 + }, + { + "epoch": 0.522624, + "grad_norm": 0.4460586607456207, + "learning_rate": 0.0004876662397550851, + "loss": 2.6743, + "step": 4083 + }, + { + "epoch": 0.522752, + "grad_norm": 0.3998332917690277, + "learning_rate": 0.00048745899278767983, + "loss": 2.691, + "step": 4084 + }, + { + "epoch": 0.52288, + "grad_norm": 0.39342212677001953, + "learning_rate": 0.00048725174797621917, + "loss": 2.6329, + "step": 4085 + }, + { + "epoch": 0.523008, + "grad_norm": 0.4453089237213135, + "learning_rate": 0.00048704450535633074, + "loss": 2.6, + "step": 4086 + }, + { + "epoch": 0.523136, + "grad_norm": 0.3963582515716553, + "learning_rate": 0.00048683726496364207, + "loss": 2.5882, + "step": 4087 + }, + { + "epoch": 0.523264, + "grad_norm": 0.39797183871269226, + "learning_rate": 0.00048663002683378015, + "loss": 2.6235, + "step": 4088 + }, + { + "epoch": 0.523392, + "grad_norm": 0.38129478693008423, + "learning_rate": 0.00048642279100237143, + "loss": 2.5674, + "step": 4089 + }, + { + "epoch": 0.52352, + "grad_norm": 0.38640400767326355, + "learning_rate": 0.00048621555750504253, + "loss": 2.5762, + "step": 4090 + }, + { + "epoch": 0.523648, + "grad_norm": 0.40850093960762024, + "learning_rate": 0.000486008326377419, + "loss": 2.5951, + "step": 4091 + }, + { + "epoch": 0.523776, + "grad_norm": 0.3988695442676544, + "learning_rate": 0.0004858010976551265, + "loss": 2.6464, + "step": 4092 + }, + { + "epoch": 0.523904, + "grad_norm": 0.43225958943367004, + "learning_rate": 0.0004855938713737899, + "loss": 2.7302, + "step": 4093 + }, + { + "epoch": 0.524032, + "grad_norm": 0.6538171768188477, + "learning_rate": 0.00048538664756903375, + "loss": 2.6999, + "step": 4094 + }, + { + "epoch": 0.52416, + "grad_norm": 0.440096378326416, + "learning_rate": 0.0004851794262764823, + "loss": 2.6512, + "step": 4095 + }, + { + "epoch": 0.524288, + "grad_norm": 0.3967292606830597, + "learning_rate": 0.00048497220753175943, + "loss": 2.5926, + "step": 4096 + }, + { + "epoch": 0.524416, + "grad_norm": 0.38446322083473206, + "learning_rate": 0.0004847649913704881, + "loss": 2.5909, + "step": 4097 + }, + { + "epoch": 0.524544, + "grad_norm": 0.3926900327205658, + "learning_rate": 0.00048455777782829124, + "loss": 2.5241, + "step": 4098 + }, + { + "epoch": 0.524672, + "grad_norm": 0.41444721817970276, + "learning_rate": 0.00048435056694079166, + "loss": 2.745, + "step": 4099 + }, + { + "epoch": 0.5248, + "grad_norm": 0.3920603394508362, + "learning_rate": 0.000484143358743611, + "loss": 2.6194, + "step": 4100 + }, + { + "epoch": 0.524928, + "grad_norm": 0.41159191727638245, + "learning_rate": 0.0004839361532723708, + "loss": 2.6081, + "step": 4101 + }, + { + "epoch": 0.525056, + "grad_norm": 0.4525657296180725, + "learning_rate": 0.0004837289505626921, + "loss": 2.6962, + "step": 4102 + }, + { + "epoch": 0.525184, + "grad_norm": 0.6175769567489624, + "learning_rate": 0.0004835217506501954, + "loss": 2.7088, + "step": 4103 + }, + { + "epoch": 0.525312, + "grad_norm": 0.4355625510215759, + "learning_rate": 0.0004833145535705008, + "loss": 2.6905, + "step": 4104 + }, + { + "epoch": 0.52544, + "grad_norm": 0.41116493940353394, + "learning_rate": 0.0004831073593592279, + "loss": 2.7067, + "step": 4105 + }, + { + "epoch": 0.525568, + "grad_norm": 0.4791308045387268, + "learning_rate": 0.0004829001680519958, + "loss": 2.6417, + "step": 4106 + }, + { + "epoch": 0.525696, + "grad_norm": 0.39732620120048523, + "learning_rate": 0.0004826929796844228, + "loss": 2.5145, + "step": 4107 + }, + { + "epoch": 0.525824, + "grad_norm": 0.4136200249195099, + "learning_rate": 0.00048248579429212735, + "loss": 2.6572, + "step": 4108 + }, + { + "epoch": 0.525952, + "grad_norm": 0.40039223432540894, + "learning_rate": 0.00048227861191072705, + "loss": 2.7054, + "step": 4109 + }, + { + "epoch": 0.52608, + "grad_norm": 0.46404942870140076, + "learning_rate": 0.00048207143257583874, + "loss": 2.6599, + "step": 4110 + }, + { + "epoch": 0.526208, + "grad_norm": 0.4081782400608063, + "learning_rate": 0.00048186425632307906, + "loss": 2.7326, + "step": 4111 + }, + { + "epoch": 0.526336, + "grad_norm": 0.4442542791366577, + "learning_rate": 0.00048165708318806383, + "loss": 2.7355, + "step": 4112 + }, + { + "epoch": 0.526464, + "grad_norm": 0.41375914216041565, + "learning_rate": 0.00048144991320640877, + "loss": 2.634, + "step": 4113 + }, + { + "epoch": 0.526592, + "grad_norm": 0.40012285113334656, + "learning_rate": 0.00048124274641372873, + "loss": 2.6685, + "step": 4114 + }, + { + "epoch": 0.52672, + "grad_norm": 0.4232390522956848, + "learning_rate": 0.00048103558284563785, + "loss": 2.6914, + "step": 4115 + }, + { + "epoch": 0.526848, + "grad_norm": 0.38594648241996765, + "learning_rate": 0.00048082842253775027, + "loss": 2.5901, + "step": 4116 + }, + { + "epoch": 0.526976, + "grad_norm": 0.3836800754070282, + "learning_rate": 0.0004806212655256789, + "loss": 2.626, + "step": 4117 + }, + { + "epoch": 0.527104, + "grad_norm": 0.3987427353858948, + "learning_rate": 0.00048041411184503684, + "loss": 2.6117, + "step": 4118 + }, + { + "epoch": 0.527232, + "grad_norm": 0.4038834273815155, + "learning_rate": 0.00048020696153143596, + "loss": 2.6381, + "step": 4119 + }, + { + "epoch": 0.52736, + "grad_norm": 0.4189172089099884, + "learning_rate": 0.0004799998146204879, + "loss": 2.647, + "step": 4120 + }, + { + "epoch": 0.527488, + "grad_norm": 0.4199029505252838, + "learning_rate": 0.0004797926711478036, + "loss": 2.6292, + "step": 4121 + }, + { + "epoch": 0.527616, + "grad_norm": 0.42810550332069397, + "learning_rate": 0.00047958553114899326, + "loss": 2.5542, + "step": 4122 + }, + { + "epoch": 0.527744, + "grad_norm": 0.42069557309150696, + "learning_rate": 0.0004793783946596669, + "loss": 2.6273, + "step": 4123 + }, + { + "epoch": 0.527872, + "grad_norm": 0.40454092621803284, + "learning_rate": 0.0004791712617154336, + "loss": 2.5299, + "step": 4124 + }, + { + "epoch": 0.528, + "grad_norm": 0.4205642640590668, + "learning_rate": 0.00047896413235190175, + "loss": 2.4989, + "step": 4125 + }, + { + "epoch": 0.528128, + "grad_norm": 0.4665340781211853, + "learning_rate": 0.0004787570066046793, + "loss": 2.7011, + "step": 4126 + }, + { + "epoch": 0.528256, + "grad_norm": 0.39806878566741943, + "learning_rate": 0.0004785498845093739, + "loss": 2.6507, + "step": 4127 + }, + { + "epoch": 0.528384, + "grad_norm": 0.48281142115592957, + "learning_rate": 0.000478342766101592, + "loss": 2.7528, + "step": 4128 + }, + { + "epoch": 0.528512, + "grad_norm": 0.3980749845504761, + "learning_rate": 0.0004781356514169396, + "loss": 2.6213, + "step": 4129 + }, + { + "epoch": 0.52864, + "grad_norm": 0.36846616864204407, + "learning_rate": 0.0004779285404910223, + "loss": 2.7307, + "step": 4130 + }, + { + "epoch": 0.528768, + "grad_norm": 0.42093613743782043, + "learning_rate": 0.0004777214333594448, + "loss": 2.6818, + "step": 4131 + }, + { + "epoch": 0.528896, + "grad_norm": 0.40272077918052673, + "learning_rate": 0.0004775143300578111, + "loss": 2.6798, + "step": 4132 + }, + { + "epoch": 0.529024, + "grad_norm": 0.536262571811676, + "learning_rate": 0.0004773072306217247, + "loss": 2.7647, + "step": 4133 + }, + { + "epoch": 0.529152, + "grad_norm": 0.42738062143325806, + "learning_rate": 0.0004771001350867885, + "loss": 2.6591, + "step": 4134 + }, + { + "epoch": 0.52928, + "grad_norm": 0.3984385132789612, + "learning_rate": 0.00047689304348860436, + "loss": 2.7178, + "step": 4135 + }, + { + "epoch": 0.529408, + "grad_norm": 0.4264759123325348, + "learning_rate": 0.00047668595586277415, + "loss": 2.6772, + "step": 4136 + }, + { + "epoch": 0.529536, + "grad_norm": 0.4121776223182678, + "learning_rate": 0.0004764788722448983, + "loss": 2.6721, + "step": 4137 + }, + { + "epoch": 0.529664, + "grad_norm": 0.39996811747550964, + "learning_rate": 0.0004762717926705771, + "loss": 2.619, + "step": 4138 + }, + { + "epoch": 0.529792, + "grad_norm": 0.442778617143631, + "learning_rate": 0.00047606471717540977, + "loss": 2.5915, + "step": 4139 + }, + { + "epoch": 0.52992, + "grad_norm": 0.391592800617218, + "learning_rate": 0.00047585764579499495, + "loss": 2.6286, + "step": 4140 + }, + { + "epoch": 0.530048, + "grad_norm": 0.42282572388648987, + "learning_rate": 0.0004756505785649308, + "loss": 2.533, + "step": 4141 + }, + { + "epoch": 0.530176, + "grad_norm": 0.4546029269695282, + "learning_rate": 0.00047544351552081455, + "loss": 2.7358, + "step": 4142 + }, + { + "epoch": 0.530304, + "grad_norm": 0.49877631664276123, + "learning_rate": 0.0004752364566982425, + "loss": 2.4949, + "step": 4143 + }, + { + "epoch": 0.530432, + "grad_norm": 0.4177952706813812, + "learning_rate": 0.0004750294021328105, + "loss": 2.5222, + "step": 4144 + }, + { + "epoch": 0.53056, + "grad_norm": 0.4101257026195526, + "learning_rate": 0.00047482235186011403, + "loss": 2.7143, + "step": 4145 + }, + { + "epoch": 0.530688, + "grad_norm": 0.42484837770462036, + "learning_rate": 0.0004746153059157471, + "loss": 2.5394, + "step": 4146 + }, + { + "epoch": 0.530816, + "grad_norm": 0.40051063895225525, + "learning_rate": 0.0004744082643353033, + "loss": 2.7217, + "step": 4147 + }, + { + "epoch": 0.530944, + "grad_norm": 0.4330638349056244, + "learning_rate": 0.00047420122715437565, + "loss": 2.698, + "step": 4148 + }, + { + "epoch": 0.531072, + "grad_norm": 0.3996522128582001, + "learning_rate": 0.00047399419440855614, + "loss": 2.5876, + "step": 4149 + }, + { + "epoch": 0.5312, + "grad_norm": 0.3962436616420746, + "learning_rate": 0.000473787166133436, + "loss": 2.5939, + "step": 4150 + }, + { + "epoch": 0.531328, + "grad_norm": 0.3768753111362457, + "learning_rate": 0.000473580142364606, + "loss": 2.7107, + "step": 4151 + }, + { + "epoch": 0.531456, + "grad_norm": 0.4448409974575043, + "learning_rate": 0.00047337312313765575, + "loss": 2.7094, + "step": 4152 + }, + { + "epoch": 0.531584, + "grad_norm": 0.40867820382118225, + "learning_rate": 0.00047316610848817435, + "loss": 2.6927, + "step": 4153 + }, + { + "epoch": 0.531712, + "grad_norm": 0.39568090438842773, + "learning_rate": 0.0004729590984517498, + "loss": 2.6413, + "step": 4154 + }, + { + "epoch": 0.53184, + "grad_norm": 0.4381014108657837, + "learning_rate": 0.0004727520930639698, + "loss": 2.6772, + "step": 4155 + }, + { + "epoch": 0.531968, + "grad_norm": 0.43804341554641724, + "learning_rate": 0.00047254509236042096, + "loss": 2.6358, + "step": 4156 + }, + { + "epoch": 0.532096, + "grad_norm": 0.45965608954429626, + "learning_rate": 0.00047233809637668906, + "loss": 2.673, + "step": 4157 + }, + { + "epoch": 0.532224, + "grad_norm": 0.4337191879749298, + "learning_rate": 0.000472131105148359, + "loss": 2.636, + "step": 4158 + }, + { + "epoch": 0.532352, + "grad_norm": 0.41232365369796753, + "learning_rate": 0.0004719241187110151, + "loss": 2.7031, + "step": 4159 + }, + { + "epoch": 0.53248, + "grad_norm": 0.4005645215511322, + "learning_rate": 0.00047171713710024074, + "loss": 2.6872, + "step": 4160 + }, + { + "epoch": 0.532608, + "grad_norm": 0.41328543424606323, + "learning_rate": 0.0004715101603516182, + "loss": 2.6321, + "step": 4161 + }, + { + "epoch": 0.532736, + "grad_norm": 0.3938266932964325, + "learning_rate": 0.0004713031885007296, + "loss": 2.5472, + "step": 4162 + }, + { + "epoch": 0.532864, + "grad_norm": 0.4321519732475281, + "learning_rate": 0.00047109622158315526, + "loss": 2.7256, + "step": 4163 + }, + { + "epoch": 0.532992, + "grad_norm": 0.4507468640804291, + "learning_rate": 0.00047088925963447575, + "loss": 2.6777, + "step": 4164 + }, + { + "epoch": 0.53312, + "grad_norm": 0.3950223922729492, + "learning_rate": 0.0004706823026902699, + "loss": 2.6693, + "step": 4165 + }, + { + "epoch": 0.533248, + "grad_norm": 0.40953874588012695, + "learning_rate": 0.0004704753507861162, + "loss": 2.6005, + "step": 4166 + }, + { + "epoch": 0.533376, + "grad_norm": 0.40956470370292664, + "learning_rate": 0.00047026840395759196, + "loss": 2.4929, + "step": 4167 + }, + { + "epoch": 0.533504, + "grad_norm": 0.40485185384750366, + "learning_rate": 0.0004700614622402737, + "loss": 2.6403, + "step": 4168 + }, + { + "epoch": 0.533632, + "grad_norm": 0.4775274097919464, + "learning_rate": 0.00046985452566973725, + "loss": 2.7461, + "step": 4169 + }, + { + "epoch": 0.53376, + "grad_norm": 0.4422805905342102, + "learning_rate": 0.0004696475942815573, + "loss": 2.7723, + "step": 4170 + }, + { + "epoch": 0.533888, + "grad_norm": 0.43448737263679504, + "learning_rate": 0.00046944066811130776, + "loss": 2.6099, + "step": 4171 + }, + { + "epoch": 0.534016, + "grad_norm": 0.4521656632423401, + "learning_rate": 0.00046923374719456143, + "loss": 2.6091, + "step": 4172 + }, + { + "epoch": 0.534144, + "grad_norm": 0.4412339925765991, + "learning_rate": 0.00046902683156689075, + "loss": 2.6379, + "step": 4173 + }, + { + "epoch": 0.534272, + "grad_norm": 0.4054372012615204, + "learning_rate": 0.0004688199212638669, + "loss": 2.5675, + "step": 4174 + }, + { + "epoch": 0.5344, + "grad_norm": 0.3982574939727783, + "learning_rate": 0.00046861301632106, + "loss": 2.6443, + "step": 4175 + }, + { + "epoch": 0.534528, + "grad_norm": 0.4294828176498413, + "learning_rate": 0.00046840611677403943, + "loss": 2.7813, + "step": 4176 + }, + { + "epoch": 0.534656, + "grad_norm": 0.4122430384159088, + "learning_rate": 0.00046819922265837376, + "loss": 2.6053, + "step": 4177 + }, + { + "epoch": 0.534784, + "grad_norm": 0.425021767616272, + "learning_rate": 0.00046799233400963037, + "loss": 2.7121, + "step": 4178 + }, + { + "epoch": 0.534912, + "grad_norm": 0.4190857410430908, + "learning_rate": 0.0004677854508633757, + "loss": 2.6822, + "step": 4179 + }, + { + "epoch": 0.53504, + "grad_norm": 0.5558974742889404, + "learning_rate": 0.0004675785732551756, + "loss": 2.6545, + "step": 4180 + }, + { + "epoch": 0.535168, + "grad_norm": 0.40109187364578247, + "learning_rate": 0.00046737170122059446, + "loss": 2.6472, + "step": 4181 + }, + { + "epoch": 0.535296, + "grad_norm": 0.4574723243713379, + "learning_rate": 0.0004671648347951963, + "loss": 2.6169, + "step": 4182 + }, + { + "epoch": 0.535424, + "grad_norm": 0.442525178194046, + "learning_rate": 0.0004669579740145436, + "loss": 2.6709, + "step": 4183 + }, + { + "epoch": 0.535552, + "grad_norm": 0.7303462028503418, + "learning_rate": 0.0004667511189141983, + "loss": 2.5951, + "step": 4184 + }, + { + "epoch": 0.53568, + "grad_norm": 0.43746834993362427, + "learning_rate": 0.0004665442695297212, + "loss": 2.6936, + "step": 4185 + }, + { + "epoch": 0.535808, + "grad_norm": 0.43732231855392456, + "learning_rate": 0.0004663374258966718, + "loss": 2.5986, + "step": 4186 + }, + { + "epoch": 0.535936, + "grad_norm": 0.44925811886787415, + "learning_rate": 0.00046613058805060933, + "loss": 2.6624, + "step": 4187 + }, + { + "epoch": 0.536064, + "grad_norm": 0.4211912453174591, + "learning_rate": 0.0004659237560270914, + "loss": 2.6536, + "step": 4188 + }, + { + "epoch": 0.536192, + "grad_norm": 0.392376571893692, + "learning_rate": 0.0004657169298616747, + "loss": 2.6571, + "step": 4189 + }, + { + "epoch": 0.53632, + "grad_norm": 0.5080148577690125, + "learning_rate": 0.00046551010958991536, + "loss": 2.7602, + "step": 4190 + }, + { + "epoch": 0.536448, + "grad_norm": 0.41217944025993347, + "learning_rate": 0.0004653032952473678, + "loss": 2.5978, + "step": 4191 + }, + { + "epoch": 0.536576, + "grad_norm": 0.3953767418861389, + "learning_rate": 0.0004650964868695863, + "loss": 2.6649, + "step": 4192 + }, + { + "epoch": 0.536704, + "grad_norm": 0.41394487023353577, + "learning_rate": 0.0004648896844921232, + "loss": 2.7213, + "step": 4193 + }, + { + "epoch": 0.536832, + "grad_norm": 0.4235329031944275, + "learning_rate": 0.0004646828881505304, + "loss": 2.6035, + "step": 4194 + }, + { + "epoch": 0.53696, + "grad_norm": 0.3955673277378082, + "learning_rate": 0.0004644760978803586, + "loss": 2.557, + "step": 4195 + }, + { + "epoch": 0.537088, + "grad_norm": 0.3986234962940216, + "learning_rate": 0.0004642693137171574, + "loss": 2.6056, + "step": 4196 + }, + { + "epoch": 0.537216, + "grad_norm": 0.3882398009300232, + "learning_rate": 0.0004640625356964753, + "loss": 2.5467, + "step": 4197 + }, + { + "epoch": 0.537344, + "grad_norm": 0.4052629768848419, + "learning_rate": 0.00046385576385386004, + "loss": 2.6565, + "step": 4198 + }, + { + "epoch": 0.537472, + "grad_norm": 0.46349629759788513, + "learning_rate": 0.000463648998224858, + "loss": 2.5017, + "step": 4199 + }, + { + "epoch": 0.5376, + "grad_norm": 0.4529445171356201, + "learning_rate": 0.00046344223884501435, + "loss": 2.5639, + "step": 4200 + }, + { + "epoch": 0.537728, + "grad_norm": 0.38720396161079407, + "learning_rate": 0.0004632354857498738, + "loss": 2.6947, + "step": 4201 + }, + { + "epoch": 0.537856, + "grad_norm": 0.43821078538894653, + "learning_rate": 0.00046302873897497956, + "loss": 2.6677, + "step": 4202 + }, + { + "epoch": 0.537984, + "grad_norm": 0.43608081340789795, + "learning_rate": 0.00046282199855587365, + "loss": 2.7359, + "step": 4203 + }, + { + "epoch": 0.538112, + "grad_norm": 0.426552414894104, + "learning_rate": 0.00046261526452809717, + "loss": 2.7136, + "step": 4204 + }, + { + "epoch": 0.53824, + "grad_norm": 0.4335874021053314, + "learning_rate": 0.00046240853692719014, + "loss": 2.6714, + "step": 4205 + }, + { + "epoch": 0.538368, + "grad_norm": 0.4228070080280304, + "learning_rate": 0.00046220181578869146, + "loss": 2.6126, + "step": 4206 + }, + { + "epoch": 0.538496, + "grad_norm": 0.43420127034187317, + "learning_rate": 0.0004619951011481387, + "loss": 2.6427, + "step": 4207 + }, + { + "epoch": 0.538624, + "grad_norm": 0.5214553475379944, + "learning_rate": 0.0004617883930410688, + "loss": 2.7528, + "step": 4208 + }, + { + "epoch": 0.538752, + "grad_norm": 0.6020479798316956, + "learning_rate": 0.0004615816915030169, + "loss": 2.5465, + "step": 4209 + }, + { + "epoch": 0.53888, + "grad_norm": 0.44846150279045105, + "learning_rate": 0.00046137499656951774, + "loss": 2.6148, + "step": 4210 + }, + { + "epoch": 0.539008, + "grad_norm": 0.3904094398021698, + "learning_rate": 0.00046116830827610454, + "loss": 2.7435, + "step": 4211 + }, + { + "epoch": 0.539136, + "grad_norm": 0.4037339985370636, + "learning_rate": 0.0004609616266583092, + "loss": 2.6893, + "step": 4212 + }, + { + "epoch": 0.539264, + "grad_norm": 0.41285622119903564, + "learning_rate": 0.0004607549517516629, + "loss": 2.7, + "step": 4213 + }, + { + "epoch": 0.539392, + "grad_norm": 0.4040307104587555, + "learning_rate": 0.0004605482835916954, + "loss": 2.6624, + "step": 4214 + }, + { + "epoch": 0.53952, + "grad_norm": 0.39195188879966736, + "learning_rate": 0.0004603416222139352, + "loss": 2.6667, + "step": 4215 + }, + { + "epoch": 0.539648, + "grad_norm": 0.4226936995983124, + "learning_rate": 0.0004601349676539099, + "loss": 2.735, + "step": 4216 + }, + { + "epoch": 0.539776, + "grad_norm": 0.43115362524986267, + "learning_rate": 0.00045992831994714594, + "loss": 2.6197, + "step": 4217 + }, + { + "epoch": 0.539904, + "grad_norm": 0.38780203461647034, + "learning_rate": 0.00045972167912916806, + "loss": 2.5345, + "step": 4218 + }, + { + "epoch": 0.540032, + "grad_norm": 0.4660673439502716, + "learning_rate": 0.0004595150452355006, + "loss": 2.6522, + "step": 4219 + }, + { + "epoch": 0.54016, + "grad_norm": 0.41473764181137085, + "learning_rate": 0.0004593084183016664, + "loss": 2.6874, + "step": 4220 + }, + { + "epoch": 0.540288, + "grad_norm": 0.42332813143730164, + "learning_rate": 0.00045910179836318666, + "loss": 2.7056, + "step": 4221 + }, + { + "epoch": 0.540416, + "grad_norm": 0.4514668881893158, + "learning_rate": 0.00045889518545558195, + "loss": 2.6634, + "step": 4222 + }, + { + "epoch": 0.540544, + "grad_norm": 0.40660709142684937, + "learning_rate": 0.00045868857961437143, + "loss": 2.5717, + "step": 4223 + }, + { + "epoch": 0.540672, + "grad_norm": 0.4411821663379669, + "learning_rate": 0.00045848198087507295, + "loss": 2.781, + "step": 4224 + }, + { + "epoch": 0.5408, + "grad_norm": 0.4213494062423706, + "learning_rate": 0.00045827538927320326, + "loss": 2.6919, + "step": 4225 + }, + { + "epoch": 0.540928, + "grad_norm": 0.43693289160728455, + "learning_rate": 0.000458068804844278, + "loss": 2.6279, + "step": 4226 + }, + { + "epoch": 0.541056, + "grad_norm": 0.39040952920913696, + "learning_rate": 0.00045786222762381125, + "loss": 2.5738, + "step": 4227 + }, + { + "epoch": 0.541184, + "grad_norm": 0.42950204014778137, + "learning_rate": 0.0004576556576473158, + "loss": 2.6167, + "step": 4228 + }, + { + "epoch": 0.541312, + "grad_norm": 0.3848821222782135, + "learning_rate": 0.00045744909495030393, + "loss": 2.5585, + "step": 4229 + }, + { + "epoch": 0.54144, + "grad_norm": 0.4366667568683624, + "learning_rate": 0.0004572425395682859, + "loss": 2.6974, + "step": 4230 + }, + { + "epoch": 0.541568, + "grad_norm": 0.4414156973361969, + "learning_rate": 0.00045703599153677104, + "loss": 2.6176, + "step": 4231 + }, + { + "epoch": 0.541696, + "grad_norm": 0.39379191398620605, + "learning_rate": 0.00045682945089126737, + "loss": 2.5659, + "step": 4232 + }, + { + "epoch": 0.541824, + "grad_norm": 0.43996334075927734, + "learning_rate": 0.0004566229176672814, + "loss": 2.5298, + "step": 4233 + }, + { + "epoch": 0.541952, + "grad_norm": 0.412761926651001, + "learning_rate": 0.0004564163919003189, + "loss": 2.6093, + "step": 4234 + }, + { + "epoch": 0.54208, + "grad_norm": 0.4374488294124603, + "learning_rate": 0.00045620987362588394, + "loss": 2.6952, + "step": 4235 + }, + { + "epoch": 0.542208, + "grad_norm": 0.4663166403770447, + "learning_rate": 0.0004560033628794792, + "loss": 2.6123, + "step": 4236 + }, + { + "epoch": 0.542336, + "grad_norm": 0.5396358370780945, + "learning_rate": 0.00045579685969660634, + "loss": 2.6324, + "step": 4237 + }, + { + "epoch": 0.542464, + "grad_norm": 0.42057085037231445, + "learning_rate": 0.00045559036411276595, + "loss": 2.6196, + "step": 4238 + }, + { + "epoch": 0.542592, + "grad_norm": 0.43615540862083435, + "learning_rate": 0.00045538387616345685, + "loss": 2.6949, + "step": 4239 + }, + { + "epoch": 0.54272, + "grad_norm": 0.4377860426902771, + "learning_rate": 0.0004551773958841765, + "loss": 2.5985, + "step": 4240 + }, + { + "epoch": 0.542848, + "grad_norm": 0.4090745151042938, + "learning_rate": 0.00045497092331042157, + "loss": 2.7874, + "step": 4241 + }, + { + "epoch": 0.542976, + "grad_norm": 0.4682491421699524, + "learning_rate": 0.00045476445847768694, + "loss": 2.5661, + "step": 4242 + }, + { + "epoch": 0.543104, + "grad_norm": 0.3855051100254059, + "learning_rate": 0.00045455800142146617, + "loss": 2.6613, + "step": 4243 + }, + { + "epoch": 0.543232, + "grad_norm": 0.43203580379486084, + "learning_rate": 0.0004543515521772519, + "loss": 2.7455, + "step": 4244 + }, + { + "epoch": 0.54336, + "grad_norm": 0.44733330607414246, + "learning_rate": 0.00045414511078053506, + "loss": 2.6786, + "step": 4245 + }, + { + "epoch": 0.543488, + "grad_norm": 0.43769004940986633, + "learning_rate": 0.000453938677266805, + "loss": 2.7029, + "step": 4246 + }, + { + "epoch": 0.543616, + "grad_norm": 0.43163836002349854, + "learning_rate": 0.0004537322516715505, + "loss": 2.67, + "step": 4247 + }, + { + "epoch": 0.543744, + "grad_norm": 0.4416983425617218, + "learning_rate": 0.00045352583403025825, + "loss": 2.6619, + "step": 4248 + }, + { + "epoch": 0.543872, + "grad_norm": 0.4717807173728943, + "learning_rate": 0.000453319424378414, + "loss": 2.5828, + "step": 4249 + }, + { + "epoch": 0.544, + "grad_norm": 0.39596521854400635, + "learning_rate": 0.0004531130227515019, + "loss": 2.636, + "step": 4250 + }, + { + "epoch": 0.544128, + "grad_norm": 0.43067842721939087, + "learning_rate": 0.00045290662918500465, + "loss": 2.6851, + "step": 4251 + }, + { + "epoch": 0.544256, + "grad_norm": 0.4234975278377533, + "learning_rate": 0.00045270024371440385, + "loss": 2.6514, + "step": 4252 + }, + { + "epoch": 0.544384, + "grad_norm": 0.4443461298942566, + "learning_rate": 0.00045249386637517954, + "loss": 2.8048, + "step": 4253 + }, + { + "epoch": 0.544512, + "grad_norm": 0.4245295524597168, + "learning_rate": 0.00045228749720281023, + "loss": 2.6549, + "step": 4254 + }, + { + "epoch": 0.54464, + "grad_norm": 0.4242624342441559, + "learning_rate": 0.0004520811362327731, + "loss": 2.6232, + "step": 4255 + }, + { + "epoch": 0.544768, + "grad_norm": 0.44125840067863464, + "learning_rate": 0.00045187478350054445, + "loss": 2.7043, + "step": 4256 + }, + { + "epoch": 0.544896, + "grad_norm": 0.38107308745384216, + "learning_rate": 0.0004516684390415985, + "loss": 2.6152, + "step": 4257 + }, + { + "epoch": 0.545024, + "grad_norm": 0.40125125646591187, + "learning_rate": 0.000451462102891408, + "loss": 2.6341, + "step": 4258 + }, + { + "epoch": 0.545152, + "grad_norm": 0.5528990030288696, + "learning_rate": 0.0004512557750854448, + "loss": 2.7583, + "step": 4259 + }, + { + "epoch": 0.54528, + "grad_norm": 0.42129960656166077, + "learning_rate": 0.000451049455659179, + "loss": 2.7478, + "step": 4260 + }, + { + "epoch": 0.545408, + "grad_norm": 0.46789219975471497, + "learning_rate": 0.00045084314464807914, + "loss": 2.6472, + "step": 4261 + }, + { + "epoch": 0.545536, + "grad_norm": 0.40193507075309753, + "learning_rate": 0.00045063684208761267, + "loss": 2.5903, + "step": 4262 + }, + { + "epoch": 0.545664, + "grad_norm": 0.3961617350578308, + "learning_rate": 0.00045043054801324533, + "loss": 2.6655, + "step": 4263 + }, + { + "epoch": 0.545792, + "grad_norm": 0.4226107597351074, + "learning_rate": 0.00045022426246044134, + "loss": 2.61, + "step": 4264 + }, + { + "epoch": 0.54592, + "grad_norm": 0.374549925327301, + "learning_rate": 0.0004500179854646637, + "loss": 2.6918, + "step": 4265 + }, + { + "epoch": 0.546048, + "grad_norm": 0.4217566251754761, + "learning_rate": 0.00044981171706137376, + "loss": 2.6168, + "step": 4266 + }, + { + "epoch": 0.546176, + "grad_norm": 0.46322739124298096, + "learning_rate": 0.00044960545728603163, + "loss": 2.6294, + "step": 4267 + }, + { + "epoch": 0.546304, + "grad_norm": 0.4148804247379303, + "learning_rate": 0.0004493992061740956, + "loss": 2.7794, + "step": 4268 + }, + { + "epoch": 0.546432, + "grad_norm": 0.4465634524822235, + "learning_rate": 0.00044919296376102254, + "loss": 2.6666, + "step": 4269 + }, + { + "epoch": 0.54656, + "grad_norm": 0.4575641453266144, + "learning_rate": 0.0004489867300822681, + "loss": 2.5716, + "step": 4270 + }, + { + "epoch": 0.546688, + "grad_norm": 0.43791863322257996, + "learning_rate": 0.0004487805051732862, + "loss": 2.6325, + "step": 4271 + }, + { + "epoch": 0.546816, + "grad_norm": 0.44494277238845825, + "learning_rate": 0.00044857428906952914, + "loss": 2.6364, + "step": 4272 + }, + { + "epoch": 0.546944, + "grad_norm": 0.4246644079685211, + "learning_rate": 0.00044836808180644805, + "loss": 2.5611, + "step": 4273 + }, + { + "epoch": 0.547072, + "grad_norm": 0.4174667000770569, + "learning_rate": 0.00044816188341949205, + "loss": 2.7088, + "step": 4274 + }, + { + "epoch": 0.5472, + "grad_norm": 0.3971532881259918, + "learning_rate": 0.00044795569394410947, + "loss": 2.6704, + "step": 4275 + }, + { + "epoch": 0.547328, + "grad_norm": 0.4119352102279663, + "learning_rate": 0.00044774951341574633, + "loss": 2.4442, + "step": 4276 + }, + { + "epoch": 0.547456, + "grad_norm": 0.4336684048175812, + "learning_rate": 0.0004475433418698477, + "loss": 2.7242, + "step": 4277 + }, + { + "epoch": 0.547584, + "grad_norm": 0.4275575876235962, + "learning_rate": 0.0004473371793418567, + "loss": 2.659, + "step": 4278 + }, + { + "epoch": 0.547712, + "grad_norm": 0.4036449193954468, + "learning_rate": 0.000447131025867215, + "loss": 2.6647, + "step": 4279 + }, + { + "epoch": 0.54784, + "grad_norm": 0.4958394765853882, + "learning_rate": 0.000446924881481363, + "loss": 2.7073, + "step": 4280 + }, + { + "epoch": 0.547968, + "grad_norm": 0.42525574564933777, + "learning_rate": 0.0004467187462197392, + "loss": 2.5051, + "step": 4281 + }, + { + "epoch": 0.548096, + "grad_norm": 0.41916921734809875, + "learning_rate": 0.00044651262011778046, + "loss": 2.6804, + "step": 4282 + }, + { + "epoch": 0.548224, + "grad_norm": 0.3653654158115387, + "learning_rate": 0.00044630650321092236, + "loss": 2.5635, + "step": 4283 + }, + { + "epoch": 0.548352, + "grad_norm": 0.39375102519989014, + "learning_rate": 0.00044610039553459884, + "loss": 2.5482, + "step": 4284 + }, + { + "epoch": 0.54848, + "grad_norm": 0.5402255654335022, + "learning_rate": 0.0004458942971242423, + "loss": 2.6718, + "step": 4285 + }, + { + "epoch": 0.548608, + "grad_norm": 0.4305102825164795, + "learning_rate": 0.0004456882080152834, + "loss": 2.693, + "step": 4286 + }, + { + "epoch": 0.548736, + "grad_norm": 0.4190831482410431, + "learning_rate": 0.00044548212824315104, + "loss": 2.7008, + "step": 4287 + }, + { + "epoch": 0.548864, + "grad_norm": 0.41537871956825256, + "learning_rate": 0.0004452760578432729, + "loss": 2.6946, + "step": 4288 + }, + { + "epoch": 0.548992, + "grad_norm": 1.0402803421020508, + "learning_rate": 0.0004450699968510749, + "loss": 2.6707, + "step": 4289 + }, + { + "epoch": 0.54912, + "grad_norm": 0.38201043009757996, + "learning_rate": 0.00044486394530198116, + "loss": 2.6646, + "step": 4290 + }, + { + "epoch": 0.549248, + "grad_norm": 0.4149767756462097, + "learning_rate": 0.00044465790323141446, + "loss": 2.6901, + "step": 4291 + }, + { + "epoch": 0.549376, + "grad_norm": 0.4285163879394531, + "learning_rate": 0.0004444518706747956, + "loss": 2.6704, + "step": 4292 + }, + { + "epoch": 0.549504, + "grad_norm": 0.41227686405181885, + "learning_rate": 0.0004442458476675443, + "loss": 2.5764, + "step": 4293 + }, + { + "epoch": 0.549632, + "grad_norm": 0.38418400287628174, + "learning_rate": 0.0004440398342450781, + "loss": 2.5588, + "step": 4294 + }, + { + "epoch": 0.54976, + "grad_norm": 0.38549965620040894, + "learning_rate": 0.0004438338304428132, + "loss": 2.606, + "step": 4295 + }, + { + "epoch": 0.549888, + "grad_norm": 0.41701170802116394, + "learning_rate": 0.000443627836296164, + "loss": 2.635, + "step": 4296 + }, + { + "epoch": 0.550016, + "grad_norm": 0.4463912844657898, + "learning_rate": 0.00044342185184054314, + "loss": 2.6129, + "step": 4297 + }, + { + "epoch": 0.550144, + "grad_norm": 0.4698549211025238, + "learning_rate": 0.00044321587711136195, + "loss": 2.7494, + "step": 4298 + }, + { + "epoch": 0.550272, + "grad_norm": 0.3931661546230316, + "learning_rate": 0.00044300991214402986, + "loss": 2.6626, + "step": 4299 + }, + { + "epoch": 0.5504, + "grad_norm": 0.38947176933288574, + "learning_rate": 0.00044280395697395437, + "loss": 2.5264, + "step": 4300 + }, + { + "epoch": 0.550528, + "grad_norm": 0.4240279495716095, + "learning_rate": 0.0004425980116365419, + "loss": 2.6708, + "step": 4301 + }, + { + "epoch": 0.550656, + "grad_norm": 0.4482777714729309, + "learning_rate": 0.00044239207616719647, + "loss": 2.668, + "step": 4302 + }, + { + "epoch": 0.550784, + "grad_norm": 0.3898100256919861, + "learning_rate": 0.00044218615060132125, + "loss": 2.7665, + "step": 4303 + }, + { + "epoch": 0.550912, + "grad_norm": 0.40828004479408264, + "learning_rate": 0.0004419802349743169, + "loss": 2.6597, + "step": 4304 + }, + { + "epoch": 0.55104, + "grad_norm": 0.402142196893692, + "learning_rate": 0.0004417743293215827, + "loss": 2.7037, + "step": 4305 + }, + { + "epoch": 0.551168, + "grad_norm": 0.4336748421192169, + "learning_rate": 0.0004415684336785164, + "loss": 2.6398, + "step": 4306 + }, + { + "epoch": 0.551296, + "grad_norm": 0.4081558585166931, + "learning_rate": 0.0004413625480805138, + "loss": 2.5922, + "step": 4307 + }, + { + "epoch": 0.551424, + "grad_norm": 0.42224255204200745, + "learning_rate": 0.0004411566725629688, + "loss": 2.577, + "step": 4308 + }, + { + "epoch": 0.551552, + "grad_norm": 0.4280494749546051, + "learning_rate": 0.00044095080716127414, + "loss": 2.6432, + "step": 4309 + }, + { + "epoch": 0.55168, + "grad_norm": 0.4311339855194092, + "learning_rate": 0.0004407449519108202, + "loss": 2.4976, + "step": 4310 + }, + { + "epoch": 0.551808, + "grad_norm": 0.48127809166908264, + "learning_rate": 0.00044053910684699576, + "loss": 2.6107, + "step": 4311 + }, + { + "epoch": 0.551936, + "grad_norm": 0.4212475121021271, + "learning_rate": 0.00044033327200518834, + "loss": 2.6965, + "step": 4312 + }, + { + "epoch": 0.552064, + "grad_norm": 0.45013412833213806, + "learning_rate": 0.00044012744742078327, + "loss": 2.6574, + "step": 4313 + }, + { + "epoch": 0.552192, + "grad_norm": 0.44309404492378235, + "learning_rate": 0.0004399216331291641, + "loss": 2.6574, + "step": 4314 + }, + { + "epoch": 0.55232, + "grad_norm": 0.41280123591423035, + "learning_rate": 0.0004397158291657125, + "loss": 2.7408, + "step": 4315 + }, + { + "epoch": 0.552448, + "grad_norm": 0.44655194878578186, + "learning_rate": 0.0004395100355658089, + "loss": 2.6657, + "step": 4316 + }, + { + "epoch": 0.552576, + "grad_norm": 0.44350218772888184, + "learning_rate": 0.0004393042523648315, + "loss": 2.6392, + "step": 4317 + }, + { + "epoch": 0.552704, + "grad_norm": 0.4003751277923584, + "learning_rate": 0.0004390984795981566, + "loss": 2.7031, + "step": 4318 + }, + { + "epoch": 0.552832, + "grad_norm": 0.38005539774894714, + "learning_rate": 0.0004388927173011592, + "loss": 2.5834, + "step": 4319 + }, + { + "epoch": 0.55296, + "grad_norm": 0.46752050518989563, + "learning_rate": 0.00043868696550921195, + "loss": 2.6316, + "step": 4320 + }, + { + "epoch": 0.553088, + "grad_norm": 0.43450993299484253, + "learning_rate": 0.00043848122425768627, + "loss": 2.651, + "step": 4321 + }, + { + "epoch": 0.553216, + "grad_norm": 0.4073009192943573, + "learning_rate": 0.00043827549358195137, + "loss": 2.6577, + "step": 4322 + }, + { + "epoch": 0.553344, + "grad_norm": 0.38162532448768616, + "learning_rate": 0.0004380697735173745, + "loss": 2.5327, + "step": 4323 + }, + { + "epoch": 0.553472, + "grad_norm": 0.40096354484558105, + "learning_rate": 0.00043786406409932176, + "loss": 2.7364, + "step": 4324 + }, + { + "epoch": 0.5536, + "grad_norm": 0.418475478887558, + "learning_rate": 0.00043765836536315667, + "loss": 2.6775, + "step": 4325 + }, + { + "epoch": 0.553728, + "grad_norm": 0.41699033975601196, + "learning_rate": 0.0004374526773442412, + "loss": 2.6871, + "step": 4326 + }, + { + "epoch": 0.553856, + "grad_norm": 0.4541328549385071, + "learning_rate": 0.00043724700007793574, + "loss": 2.5741, + "step": 4327 + }, + { + "epoch": 0.553984, + "grad_norm": 0.46125859022140503, + "learning_rate": 0.00043704133359959846, + "loss": 2.6889, + "step": 4328 + }, + { + "epoch": 0.554112, + "grad_norm": 0.4188210070133209, + "learning_rate": 0.00043683567794458556, + "loss": 2.6359, + "step": 4329 + }, + { + "epoch": 0.55424, + "grad_norm": 0.42874079942703247, + "learning_rate": 0.000436630033148252, + "loss": 2.6302, + "step": 4330 + }, + { + "epoch": 0.554368, + "grad_norm": 0.40102797746658325, + "learning_rate": 0.00043642439924595057, + "loss": 2.5523, + "step": 4331 + }, + { + "epoch": 0.554496, + "grad_norm": 0.4420813024044037, + "learning_rate": 0.00043621877627303185, + "loss": 2.6315, + "step": 4332 + }, + { + "epoch": 0.554624, + "grad_norm": 0.39273613691329956, + "learning_rate": 0.00043601316426484485, + "loss": 2.6795, + "step": 4333 + }, + { + "epoch": 0.554752, + "grad_norm": 0.41839203238487244, + "learning_rate": 0.00043580756325673674, + "loss": 2.6264, + "step": 4334 + }, + { + "epoch": 0.55488, + "grad_norm": 0.7538331747055054, + "learning_rate": 0.0004356019732840528, + "loss": 2.618, + "step": 4335 + }, + { + "epoch": 0.555008, + "grad_norm": 0.4403936266899109, + "learning_rate": 0.000435396394382136, + "loss": 2.6765, + "step": 4336 + }, + { + "epoch": 0.555136, + "grad_norm": 0.41596105694770813, + "learning_rate": 0.00043519082658632815, + "loss": 2.7205, + "step": 4337 + }, + { + "epoch": 0.555264, + "grad_norm": 0.447325736284256, + "learning_rate": 0.0004349852699319685, + "loss": 2.7757, + "step": 4338 + }, + { + "epoch": 0.555392, + "grad_norm": 0.4144016206264496, + "learning_rate": 0.00043477972445439447, + "loss": 2.6156, + "step": 4339 + }, + { + "epoch": 0.55552, + "grad_norm": 0.5207722187042236, + "learning_rate": 0.00043457419018894203, + "loss": 2.6028, + "step": 4340 + }, + { + "epoch": 0.555648, + "grad_norm": 0.3891447186470032, + "learning_rate": 0.0004343686671709448, + "loss": 2.5927, + "step": 4341 + }, + { + "epoch": 0.555776, + "grad_norm": 0.4142095148563385, + "learning_rate": 0.00043416315543573465, + "loss": 2.7032, + "step": 4342 + }, + { + "epoch": 0.555904, + "grad_norm": 0.49225977063179016, + "learning_rate": 0.0004339576550186413, + "loss": 2.6776, + "step": 4343 + }, + { + "epoch": 0.556032, + "grad_norm": 0.4231899678707123, + "learning_rate": 0.0004337521659549925, + "loss": 2.5551, + "step": 4344 + }, + { + "epoch": 0.55616, + "grad_norm": 0.4501100182533264, + "learning_rate": 0.0004335466882801146, + "loss": 2.5433, + "step": 4345 + }, + { + "epoch": 0.556288, + "grad_norm": 0.4487120807170868, + "learning_rate": 0.0004333412220293313, + "loss": 2.7101, + "step": 4346 + }, + { + "epoch": 0.556416, + "grad_norm": 0.4717293381690979, + "learning_rate": 0.0004331357672379646, + "loss": 2.6119, + "step": 4347 + }, + { + "epoch": 0.556544, + "grad_norm": 0.4149990379810333, + "learning_rate": 0.0004329303239413346, + "loss": 2.6576, + "step": 4348 + }, + { + "epoch": 0.556672, + "grad_norm": 0.4696747362613678, + "learning_rate": 0.00043272489217475963, + "loss": 2.5976, + "step": 4349 + }, + { + "epoch": 0.5568, + "grad_norm": 0.4237804710865021, + "learning_rate": 0.0004325194719735557, + "loss": 2.5521, + "step": 4350 + }, + { + "epoch": 0.556928, + "grad_norm": 0.4077187478542328, + "learning_rate": 0.00043231406337303663, + "loss": 2.7407, + "step": 4351 + }, + { + "epoch": 0.557056, + "grad_norm": 0.415274441242218, + "learning_rate": 0.00043210866640851485, + "loss": 2.6448, + "step": 4352 + }, + { + "epoch": 0.557184, + "grad_norm": 0.4562598466873169, + "learning_rate": 0.00043190328111530045, + "loss": 2.6519, + "step": 4353 + }, + { + "epoch": 0.557312, + "grad_norm": 0.39282017946243286, + "learning_rate": 0.00043169790752870137, + "loss": 2.6128, + "step": 4354 + }, + { + "epoch": 0.55744, + "grad_norm": 0.44993484020233154, + "learning_rate": 0.00043149254568402385, + "loss": 2.596, + "step": 4355 + }, + { + "epoch": 0.557568, + "grad_norm": 0.4991808533668518, + "learning_rate": 0.00043128719561657196, + "loss": 2.5581, + "step": 4356 + }, + { + "epoch": 0.557696, + "grad_norm": 0.45125052332878113, + "learning_rate": 0.0004310818573616476, + "loss": 2.6555, + "step": 4357 + }, + { + "epoch": 0.557824, + "grad_norm": 0.46367236971855164, + "learning_rate": 0.0004308765309545509, + "loss": 2.5744, + "step": 4358 + }, + { + "epoch": 0.557952, + "grad_norm": 0.6806403994560242, + "learning_rate": 0.00043067121643058004, + "loss": 2.5664, + "step": 4359 + }, + { + "epoch": 0.55808, + "grad_norm": 1.5325465202331543, + "learning_rate": 0.0004304659138250309, + "loss": 2.6494, + "step": 4360 + }, + { + "epoch": 0.558208, + "grad_norm": 0.4915200471878052, + "learning_rate": 0.00043026062317319704, + "loss": 2.75, + "step": 4361 + }, + { + "epoch": 0.558336, + "grad_norm": 0.6957546472549438, + "learning_rate": 0.00043005534451037063, + "loss": 2.6401, + "step": 4362 + }, + { + "epoch": 0.558464, + "grad_norm": 0.489810585975647, + "learning_rate": 0.0004298500778718415, + "loss": 2.6181, + "step": 4363 + }, + { + "epoch": 0.558592, + "grad_norm": 0.4262291193008423, + "learning_rate": 0.0004296448232928971, + "loss": 2.6257, + "step": 4364 + }, + { + "epoch": 0.55872, + "grad_norm": 0.5574291348457336, + "learning_rate": 0.0004294395808088232, + "loss": 2.5469, + "step": 4365 + }, + { + "epoch": 0.558848, + "grad_norm": 0.45909884572029114, + "learning_rate": 0.00042923435045490323, + "loss": 2.7547, + "step": 4366 + }, + { + "epoch": 0.558976, + "grad_norm": 0.7189844250679016, + "learning_rate": 0.000429029132266419, + "loss": 2.6127, + "step": 4367 + }, + { + "epoch": 0.559104, + "grad_norm": 0.6522992253303528, + "learning_rate": 0.0004288239262786497, + "loss": 2.7038, + "step": 4368 + }, + { + "epoch": 0.559232, + "grad_norm": 0.4044354557991028, + "learning_rate": 0.0004286187325268725, + "loss": 2.6026, + "step": 4369 + }, + { + "epoch": 0.55936, + "grad_norm": 0.4353001117706299, + "learning_rate": 0.00042841355104636283, + "loss": 2.7388, + "step": 4370 + }, + { + "epoch": 0.559488, + "grad_norm": 0.40752774477005005, + "learning_rate": 0.00042820838187239364, + "loss": 2.5424, + "step": 4371 + }, + { + "epoch": 0.559616, + "grad_norm": 0.4630663990974426, + "learning_rate": 0.00042800322504023573, + "loss": 2.6093, + "step": 4372 + }, + { + "epoch": 0.559744, + "grad_norm": 0.4413747787475586, + "learning_rate": 0.00042779808058515824, + "loss": 2.6603, + "step": 4373 + }, + { + "epoch": 0.559872, + "grad_norm": 0.4324740767478943, + "learning_rate": 0.00042759294854242773, + "loss": 2.6452, + "step": 4374 + }, + { + "epoch": 0.56, + "grad_norm": 0.4675288796424866, + "learning_rate": 0.00042738782894730874, + "loss": 2.7075, + "step": 4375 + }, + { + "epoch": 0.560128, + "grad_norm": 0.42481178045272827, + "learning_rate": 0.00042718272183506353, + "loss": 2.6157, + "step": 4376 + }, + { + "epoch": 0.560256, + "grad_norm": 0.4133565425872803, + "learning_rate": 0.00042697762724095285, + "loss": 2.5943, + "step": 4377 + }, + { + "epoch": 0.560384, + "grad_norm": 0.44176700711250305, + "learning_rate": 0.00042677254520023463, + "loss": 2.5991, + "step": 4378 + }, + { + "epoch": 0.560512, + "grad_norm": 0.37365683913230896, + "learning_rate": 0.0004265674757481647, + "loss": 2.759, + "step": 4379 + }, + { + "epoch": 0.56064, + "grad_norm": 0.4658011794090271, + "learning_rate": 0.00042636241891999706, + "loss": 2.6484, + "step": 4380 + }, + { + "epoch": 0.560768, + "grad_norm": 0.3821597099304199, + "learning_rate": 0.0004261573747509833, + "loss": 2.5272, + "step": 4381 + }, + { + "epoch": 0.560896, + "grad_norm": 0.4194265604019165, + "learning_rate": 0.00042595234327637277, + "loss": 2.5673, + "step": 4382 + }, + { + "epoch": 0.561024, + "grad_norm": 0.4008798599243164, + "learning_rate": 0.000425747324531413, + "loss": 2.5987, + "step": 4383 + }, + { + "epoch": 0.561152, + "grad_norm": 0.3662230968475342, + "learning_rate": 0.0004255423185513489, + "loss": 2.578, + "step": 4384 + }, + { + "epoch": 0.56128, + "grad_norm": 0.5020869374275208, + "learning_rate": 0.00042533732537142335, + "loss": 2.6582, + "step": 4385 + }, + { + "epoch": 0.561408, + "grad_norm": 0.5966584086418152, + "learning_rate": 0.00042513234502687723, + "loss": 2.6386, + "step": 4386 + }, + { + "epoch": 0.561536, + "grad_norm": 3.153876543045044, + "learning_rate": 0.00042492737755294883, + "loss": 2.5947, + "step": 4387 + }, + { + "epoch": 0.561664, + "grad_norm": 0.4683152735233307, + "learning_rate": 0.00042472242298487466, + "loss": 2.6202, + "step": 4388 + }, + { + "epoch": 0.561792, + "grad_norm": 0.6384204626083374, + "learning_rate": 0.0004245174813578887, + "loss": 2.5381, + "step": 4389 + }, + { + "epoch": 0.56192, + "grad_norm": 0.4361715316772461, + "learning_rate": 0.0004243125527072227, + "loss": 2.7064, + "step": 4390 + }, + { + "epoch": 0.562048, + "grad_norm": 0.4679601192474365, + "learning_rate": 0.00042410763706810634, + "loss": 2.7442, + "step": 4391 + }, + { + "epoch": 0.562176, + "grad_norm": 0.4377609193325043, + "learning_rate": 0.000423902734475767, + "loss": 2.6025, + "step": 4392 + }, + { + "epoch": 0.562304, + "grad_norm": 0.4008437991142273, + "learning_rate": 0.00042369784496542977, + "loss": 2.6354, + "step": 4393 + }, + { + "epoch": 0.562432, + "grad_norm": 0.42706501483917236, + "learning_rate": 0.0004234929685723174, + "loss": 2.7054, + "step": 4394 + }, + { + "epoch": 0.56256, + "grad_norm": 0.40548884868621826, + "learning_rate": 0.0004232881053316509, + "loss": 2.7334, + "step": 4395 + }, + { + "epoch": 0.562688, + "grad_norm": 0.3884022533893585, + "learning_rate": 0.0004230832552786484, + "loss": 2.603, + "step": 4396 + }, + { + "epoch": 0.562816, + "grad_norm": 0.48705196380615234, + "learning_rate": 0.00042287841844852595, + "loss": 2.6097, + "step": 4397 + }, + { + "epoch": 0.562944, + "grad_norm": 0.5045711994171143, + "learning_rate": 0.00042267359487649755, + "loss": 2.8118, + "step": 4398 + }, + { + "epoch": 0.563072, + "grad_norm": 0.47055584192276, + "learning_rate": 0.0004224687845977746, + "loss": 2.6852, + "step": 4399 + }, + { + "epoch": 0.5632, + "grad_norm": 0.44241926074028015, + "learning_rate": 0.00042226398764756625, + "loss": 2.7026, + "step": 4400 + }, + { + "epoch": 0.563328, + "grad_norm": 0.4363991618156433, + "learning_rate": 0.0004220592040610798, + "loss": 2.587, + "step": 4401 + }, + { + "epoch": 0.563456, + "grad_norm": 0.4156983196735382, + "learning_rate": 0.0004218544338735197, + "loss": 2.6346, + "step": 4402 + }, + { + "epoch": 0.563584, + "grad_norm": 0.4293241500854492, + "learning_rate": 0.0004216496771200881, + "loss": 2.7764, + "step": 4403 + }, + { + "epoch": 0.563712, + "grad_norm": 0.4990403950214386, + "learning_rate": 0.0004214449338359856, + "loss": 2.6644, + "step": 4404 + }, + { + "epoch": 0.56384, + "grad_norm": 0.5096971988677979, + "learning_rate": 0.00042124020405640955, + "loss": 2.5338, + "step": 4405 + }, + { + "epoch": 0.563968, + "grad_norm": 0.5568214654922485, + "learning_rate": 0.00042103548781655554, + "loss": 2.5547, + "step": 4406 + }, + { + "epoch": 0.564096, + "grad_norm": 0.4320657253265381, + "learning_rate": 0.00042083078515161665, + "loss": 2.685, + "step": 4407 + }, + { + "epoch": 0.564224, + "grad_norm": 0.4259560704231262, + "learning_rate": 0.00042062609609678353, + "loss": 2.6622, + "step": 4408 + }, + { + "epoch": 0.564352, + "grad_norm": 0.4091716706752777, + "learning_rate": 0.0004204214206872448, + "loss": 2.6919, + "step": 4409 + }, + { + "epoch": 0.56448, + "grad_norm": 0.43004927039146423, + "learning_rate": 0.0004202167589581863, + "loss": 2.5507, + "step": 4410 + }, + { + "epoch": 0.564608, + "grad_norm": 0.41948190331459045, + "learning_rate": 0.0004200121109447918, + "loss": 2.6696, + "step": 4411 + }, + { + "epoch": 0.564736, + "grad_norm": 0.4086318016052246, + "learning_rate": 0.00041980747668224285, + "loss": 2.5544, + "step": 4412 + }, + { + "epoch": 0.564864, + "grad_norm": 0.49518775939941406, + "learning_rate": 0.00041960285620571807, + "loss": 2.6702, + "step": 4413 + }, + { + "epoch": 0.564992, + "grad_norm": 0.3951432704925537, + "learning_rate": 0.00041939824955039464, + "loss": 2.6565, + "step": 4414 + }, + { + "epoch": 0.56512, + "grad_norm": 0.4453786313533783, + "learning_rate": 0.00041919365675144635, + "loss": 2.7177, + "step": 4415 + }, + { + "epoch": 0.565248, + "grad_norm": 0.5075834393501282, + "learning_rate": 0.0004189890778440454, + "loss": 2.6477, + "step": 4416 + }, + { + "epoch": 0.565376, + "grad_norm": 0.4543869197368622, + "learning_rate": 0.0004187845128633611, + "loss": 2.6334, + "step": 4417 + }, + { + "epoch": 0.565504, + "grad_norm": 0.47770199179649353, + "learning_rate": 0.00041857996184456047, + "loss": 2.5204, + "step": 4418 + }, + { + "epoch": 0.565632, + "grad_norm": 0.41104042530059814, + "learning_rate": 0.00041837542482280846, + "loss": 2.6648, + "step": 4419 + }, + { + "epoch": 0.56576, + "grad_norm": 0.8459033966064453, + "learning_rate": 0.0004181709018332671, + "loss": 2.6647, + "step": 4420 + }, + { + "epoch": 0.565888, + "grad_norm": 0.43203043937683105, + "learning_rate": 0.0004179663929110963, + "loss": 2.5777, + "step": 4421 + }, + { + "epoch": 0.566016, + "grad_norm": 0.4356876611709595, + "learning_rate": 0.0004177618980914536, + "loss": 2.5663, + "step": 4422 + }, + { + "epoch": 0.566144, + "grad_norm": 0.4615427553653717, + "learning_rate": 0.000417557417409494, + "loss": 2.5344, + "step": 4423 + }, + { + "epoch": 0.566272, + "grad_norm": 0.3988760709762573, + "learning_rate": 0.0004173529509003702, + "loss": 2.7279, + "step": 4424 + }, + { + "epoch": 0.5664, + "grad_norm": 0.4394991993904114, + "learning_rate": 0.00041714849859923227, + "loss": 2.688, + "step": 4425 + }, + { + "epoch": 0.566528, + "grad_norm": 0.3941129446029663, + "learning_rate": 0.0004169440605412278, + "loss": 2.6555, + "step": 4426 + }, + { + "epoch": 0.566656, + "grad_norm": 0.4342931807041168, + "learning_rate": 0.00041673963676150226, + "loss": 2.7715, + "step": 4427 + }, + { + "epoch": 0.566784, + "grad_norm": 0.4034636318683624, + "learning_rate": 0.0004165352272951985, + "loss": 2.6612, + "step": 4428 + }, + { + "epoch": 0.566912, + "grad_norm": 0.4255024790763855, + "learning_rate": 0.0004163308321774566, + "loss": 2.6131, + "step": 4429 + }, + { + "epoch": 0.56704, + "grad_norm": 0.3889661133289337, + "learning_rate": 0.0004161264514434148, + "loss": 2.6809, + "step": 4430 + }, + { + "epoch": 0.567168, + "grad_norm": 0.4314446449279785, + "learning_rate": 0.000415922085128208, + "loss": 2.7158, + "step": 4431 + }, + { + "epoch": 0.567296, + "grad_norm": 0.4497658908367157, + "learning_rate": 0.00041571773326696984, + "loss": 2.7419, + "step": 4432 + }, + { + "epoch": 0.567424, + "grad_norm": 0.4090615510940552, + "learning_rate": 0.00041551339589483016, + "loss": 2.631, + "step": 4433 + }, + { + "epoch": 0.567552, + "grad_norm": 0.40961143374443054, + "learning_rate": 0.0004153090730469174, + "loss": 2.5546, + "step": 4434 + }, + { + "epoch": 0.56768, + "grad_norm": 0.4044528901576996, + "learning_rate": 0.0004151047647583568, + "loss": 2.653, + "step": 4435 + }, + { + "epoch": 0.567808, + "grad_norm": 0.43563565611839294, + "learning_rate": 0.00041490047106427114, + "loss": 2.5539, + "step": 4436 + }, + { + "epoch": 0.567936, + "grad_norm": 0.41192811727523804, + "learning_rate": 0.00041469619199978124, + "loss": 2.6096, + "step": 4437 + }, + { + "epoch": 0.568064, + "grad_norm": 0.47918978333473206, + "learning_rate": 0.0004144919276000048, + "loss": 2.6876, + "step": 4438 + }, + { + "epoch": 0.568192, + "grad_norm": 0.4784252643585205, + "learning_rate": 0.00041428767790005727, + "loss": 2.7555, + "step": 4439 + }, + { + "epoch": 0.56832, + "grad_norm": 0.47463634610176086, + "learning_rate": 0.0004140834429350515, + "loss": 2.7061, + "step": 4440 + }, + { + "epoch": 0.568448, + "grad_norm": 0.41794246435165405, + "learning_rate": 0.000413879222740098, + "loss": 2.638, + "step": 4441 + }, + { + "epoch": 0.568576, + "grad_norm": 0.42930611968040466, + "learning_rate": 0.00041367501735030453, + "loss": 2.6698, + "step": 4442 + }, + { + "epoch": 0.568704, + "grad_norm": 0.4008221924304962, + "learning_rate": 0.0004134708268007764, + "loss": 2.8051, + "step": 4443 + }, + { + "epoch": 0.568832, + "grad_norm": 0.39214009046554565, + "learning_rate": 0.00041326665112661614, + "loss": 2.574, + "step": 4444 + }, + { + "epoch": 0.56896, + "grad_norm": 0.38475900888442993, + "learning_rate": 0.0004130624903629242, + "loss": 2.6525, + "step": 4445 + }, + { + "epoch": 0.569088, + "grad_norm": 0.3967072367668152, + "learning_rate": 0.00041285834454479796, + "loss": 2.5411, + "step": 4446 + }, + { + "epoch": 0.569216, + "grad_norm": 0.39994752407073975, + "learning_rate": 0.00041265421370733247, + "loss": 2.6176, + "step": 4447 + }, + { + "epoch": 0.569344, + "grad_norm": 0.41329383850097656, + "learning_rate": 0.00041245009788562037, + "loss": 2.7309, + "step": 4448 + }, + { + "epoch": 0.569472, + "grad_norm": 0.40564101934432983, + "learning_rate": 0.0004122459971147514, + "loss": 2.6726, + "step": 4449 + }, + { + "epoch": 0.5696, + "grad_norm": 0.43901336193084717, + "learning_rate": 0.0004120419114298126, + "loss": 2.6394, + "step": 4450 + }, + { + "epoch": 0.569728, + "grad_norm": 0.4354161024093628, + "learning_rate": 0.00041183784086588904, + "loss": 2.6324, + "step": 4451 + }, + { + "epoch": 0.569856, + "grad_norm": 0.44197598099708557, + "learning_rate": 0.00041163378545806286, + "loss": 2.7324, + "step": 4452 + }, + { + "epoch": 0.569984, + "grad_norm": 0.42677977681159973, + "learning_rate": 0.0004114297452414133, + "loss": 2.5929, + "step": 4453 + }, + { + "epoch": 0.570112, + "grad_norm": 0.43691718578338623, + "learning_rate": 0.0004112257202510173, + "loss": 2.642, + "step": 4454 + }, + { + "epoch": 0.57024, + "grad_norm": 0.5055019855499268, + "learning_rate": 0.0004110217105219492, + "loss": 2.5601, + "step": 4455 + }, + { + "epoch": 0.570368, + "grad_norm": 0.42564842104911804, + "learning_rate": 0.00041081771608928065, + "loss": 2.6255, + "step": 4456 + }, + { + "epoch": 0.570496, + "grad_norm": 0.3633372485637665, + "learning_rate": 0.0004106137369880804, + "loss": 2.5407, + "step": 4457 + }, + { + "epoch": 0.570624, + "grad_norm": 0.4407828450202942, + "learning_rate": 0.0004104097732534152, + "loss": 2.6319, + "step": 4458 + }, + { + "epoch": 0.570752, + "grad_norm": 0.4707808196544647, + "learning_rate": 0.0004102058249203483, + "loss": 2.7247, + "step": 4459 + }, + { + "epoch": 0.57088, + "grad_norm": 0.407308965921402, + "learning_rate": 0.0004100018920239414, + "loss": 2.6791, + "step": 4460 + }, + { + "epoch": 0.571008, + "grad_norm": 0.4175347685813904, + "learning_rate": 0.00040979797459925253, + "loss": 2.7332, + "step": 4461 + }, + { + "epoch": 0.571136, + "grad_norm": 0.4024442136287689, + "learning_rate": 0.00040959407268133745, + "loss": 2.6089, + "step": 4462 + }, + { + "epoch": 0.571264, + "grad_norm": 0.41807976365089417, + "learning_rate": 0.0004093901863052495, + "loss": 2.6405, + "step": 4463 + }, + { + "epoch": 0.571392, + "grad_norm": 0.4408838748931885, + "learning_rate": 0.0004091863155060389, + "loss": 2.6751, + "step": 4464 + }, + { + "epoch": 0.57152, + "grad_norm": 0.4227537214756012, + "learning_rate": 0.0004089824603187534, + "loss": 2.6844, + "step": 4465 + }, + { + "epoch": 0.571648, + "grad_norm": 0.9231544733047485, + "learning_rate": 0.00040877862077843825, + "loss": 2.6839, + "step": 4466 + }, + { + "epoch": 0.571776, + "grad_norm": 0.41226059198379517, + "learning_rate": 0.00040857479692013566, + "loss": 2.6891, + "step": 4467 + }, + { + "epoch": 0.571904, + "grad_norm": 0.4957022964954376, + "learning_rate": 0.00040837098877888523, + "loss": 2.6358, + "step": 4468 + }, + { + "epoch": 0.572032, + "grad_norm": 0.4228988289833069, + "learning_rate": 0.0004081671963897241, + "loss": 2.6478, + "step": 4469 + }, + { + "epoch": 0.57216, + "grad_norm": 0.475028932094574, + "learning_rate": 0.0004079634197876867, + "loss": 2.5683, + "step": 4470 + }, + { + "epoch": 0.572288, + "grad_norm": 0.3951622545719147, + "learning_rate": 0.00040775965900780443, + "loss": 2.6028, + "step": 4471 + }, + { + "epoch": 0.572416, + "grad_norm": 0.4329795837402344, + "learning_rate": 0.000407555914085106, + "loss": 2.5016, + "step": 4472 + }, + { + "epoch": 0.572544, + "grad_norm": 0.4128694236278534, + "learning_rate": 0.0004073521850546177, + "loss": 2.8648, + "step": 4473 + }, + { + "epoch": 0.572672, + "grad_norm": 0.5038319826126099, + "learning_rate": 0.00040714847195136294, + "loss": 2.6438, + "step": 4474 + }, + { + "epoch": 0.5728, + "grad_norm": 0.4314952492713928, + "learning_rate": 0.00040694477481036215, + "loss": 2.5992, + "step": 4475 + }, + { + "epoch": 0.572928, + "grad_norm": 0.3775739073753357, + "learning_rate": 0.00040674109366663344, + "loss": 2.5699, + "step": 4476 + }, + { + "epoch": 0.573056, + "grad_norm": 0.4038936197757721, + "learning_rate": 0.00040653742855519166, + "loss": 2.6321, + "step": 4477 + }, + { + "epoch": 0.573184, + "grad_norm": 0.4481957256793976, + "learning_rate": 0.00040633377951104966, + "loss": 2.5542, + "step": 4478 + }, + { + "epoch": 0.573312, + "grad_norm": 0.43616339564323425, + "learning_rate": 0.00040613014656921677, + "loss": 2.6217, + "step": 4479 + }, + { + "epoch": 0.57344, + "grad_norm": 0.42005473375320435, + "learning_rate": 0.0004059265297646999, + "loss": 2.5761, + "step": 4480 + }, + { + "epoch": 0.573568, + "grad_norm": 0.4149060845375061, + "learning_rate": 0.0004057229291325032, + "loss": 2.6485, + "step": 4481 + }, + { + "epoch": 0.573696, + "grad_norm": 0.8321806192398071, + "learning_rate": 0.00040551934470762797, + "loss": 2.6484, + "step": 4482 + }, + { + "epoch": 0.573824, + "grad_norm": 0.40239831805229187, + "learning_rate": 0.00040531577652507255, + "loss": 2.6604, + "step": 4483 + }, + { + "epoch": 0.573952, + "grad_norm": 0.4007665514945984, + "learning_rate": 0.000405112224619833, + "loss": 2.6132, + "step": 4484 + }, + { + "epoch": 0.57408, + "grad_norm": 0.4488524794578552, + "learning_rate": 0.00040490868902690204, + "loss": 2.5755, + "step": 4485 + }, + { + "epoch": 0.574208, + "grad_norm": 0.3744049370288849, + "learning_rate": 0.0004047051697812697, + "loss": 2.7224, + "step": 4486 + }, + { + "epoch": 0.574336, + "grad_norm": 0.4213537871837616, + "learning_rate": 0.00040450166691792327, + "loss": 2.6006, + "step": 4487 + }, + { + "epoch": 0.574464, + "grad_norm": 0.41649994254112244, + "learning_rate": 0.00040429818047184774, + "loss": 2.5925, + "step": 4488 + }, + { + "epoch": 0.574592, + "grad_norm": 0.44413402676582336, + "learning_rate": 0.00040409471047802436, + "loss": 2.6975, + "step": 4489 + }, + { + "epoch": 0.57472, + "grad_norm": 0.43579837679862976, + "learning_rate": 0.000403891256971432, + "loss": 2.6925, + "step": 4490 + }, + { + "epoch": 0.574848, + "grad_norm": 0.41490525007247925, + "learning_rate": 0.0004036878199870469, + "loss": 2.6276, + "step": 4491 + }, + { + "epoch": 0.574976, + "grad_norm": 0.39915966987609863, + "learning_rate": 0.00040348439955984204, + "loss": 2.7189, + "step": 4492 + }, + { + "epoch": 0.575104, + "grad_norm": 0.4155600070953369, + "learning_rate": 0.0004032809957247877, + "loss": 2.6175, + "step": 4493 + }, + { + "epoch": 0.575232, + "grad_norm": 0.4613172709941864, + "learning_rate": 0.0004030776085168516, + "loss": 2.5508, + "step": 4494 + }, + { + "epoch": 0.57536, + "grad_norm": 0.4489295482635498, + "learning_rate": 0.0004028742379709982, + "loss": 2.6733, + "step": 4495 + }, + { + "epoch": 0.575488, + "grad_norm": 0.3953936994075775, + "learning_rate": 0.000402670884122189, + "loss": 2.5731, + "step": 4496 + }, + { + "epoch": 0.575616, + "grad_norm": 0.4574202299118042, + "learning_rate": 0.00040246754700538353, + "loss": 2.7201, + "step": 4497 + }, + { + "epoch": 0.575744, + "grad_norm": 0.46241557598114014, + "learning_rate": 0.0004022642266555372, + "loss": 2.7424, + "step": 4498 + }, + { + "epoch": 0.575872, + "grad_norm": 0.44105958938598633, + "learning_rate": 0.00040206092310760347, + "loss": 2.7393, + "step": 4499 + }, + { + "epoch": 0.576, + "grad_norm": 0.4570227861404419, + "learning_rate": 0.00040185763639653246, + "loss": 2.622, + "step": 4500 + }, + { + "epoch": 0.576128, + "grad_norm": 0.42703261971473694, + "learning_rate": 0.0004016543665572714, + "loss": 2.6278, + "step": 4501 + }, + { + "epoch": 0.576256, + "grad_norm": 0.5067856907844543, + "learning_rate": 0.000401451113624765, + "loss": 2.6909, + "step": 4502 + }, + { + "epoch": 0.576384, + "grad_norm": 0.4446900486946106, + "learning_rate": 0.00040124787763395463, + "loss": 2.6235, + "step": 4503 + }, + { + "epoch": 0.576512, + "grad_norm": 0.42020493745803833, + "learning_rate": 0.0004010446586197788, + "loss": 2.7128, + "step": 4504 + }, + { + "epoch": 0.57664, + "grad_norm": 0.43034467101097107, + "learning_rate": 0.00040084145661717325, + "loss": 2.6608, + "step": 4505 + }, + { + "epoch": 0.576768, + "grad_norm": 1.7534255981445312, + "learning_rate": 0.000400638271661071, + "loss": 2.7218, + "step": 4506 + }, + { + "epoch": 0.576896, + "grad_norm": 0.4501379728317261, + "learning_rate": 0.00040043510378640193, + "loss": 2.7238, + "step": 4507 + }, + { + "epoch": 0.577024, + "grad_norm": 0.4675261378288269, + "learning_rate": 0.0004002319530280925, + "loss": 2.6255, + "step": 4508 + }, + { + "epoch": 0.577152, + "grad_norm": 0.3905614912509918, + "learning_rate": 0.0004000288194210671, + "loss": 2.5679, + "step": 4509 + }, + { + "epoch": 0.57728, + "grad_norm": 0.3986473083496094, + "learning_rate": 0.00039982570300024664, + "loss": 2.6071, + "step": 4510 + }, + { + "epoch": 0.577408, + "grad_norm": 0.4018094837665558, + "learning_rate": 0.00039962260380054903, + "loss": 2.7142, + "step": 4511 + }, + { + "epoch": 0.577536, + "grad_norm": 0.4327203035354614, + "learning_rate": 0.0003994195218568896, + "loss": 2.5753, + "step": 4512 + }, + { + "epoch": 0.577664, + "grad_norm": 0.45541366934776306, + "learning_rate": 0.0003992164572041804, + "loss": 2.6497, + "step": 4513 + }, + { + "epoch": 0.577792, + "grad_norm": 0.4198867976665497, + "learning_rate": 0.00039901340987733036, + "loss": 2.7427, + "step": 4514 + }, + { + "epoch": 0.57792, + "grad_norm": 0.4960704743862152, + "learning_rate": 0.0003988103799112461, + "loss": 2.6792, + "step": 4515 + }, + { + "epoch": 0.578048, + "grad_norm": 0.41351351141929626, + "learning_rate": 0.0003986073673408306, + "loss": 2.7017, + "step": 4516 + }, + { + "epoch": 0.578176, + "grad_norm": 0.4027905762195587, + "learning_rate": 0.0003984043722009841, + "loss": 2.5964, + "step": 4517 + }, + { + "epoch": 0.578304, + "grad_norm": 0.4385596811771393, + "learning_rate": 0.0003982013945266039, + "loss": 2.8152, + "step": 4518 + }, + { + "epoch": 0.578432, + "grad_norm": 0.4144969582557678, + "learning_rate": 0.000397998434352584, + "loss": 2.5908, + "step": 4519 + }, + { + "epoch": 0.57856, + "grad_norm": 0.4675523042678833, + "learning_rate": 0.0003977954917138159, + "loss": 2.5974, + "step": 4520 + }, + { + "epoch": 0.578688, + "grad_norm": 0.4294047951698303, + "learning_rate": 0.00039759256664518753, + "loss": 2.5455, + "step": 4521 + }, + { + "epoch": 0.578816, + "grad_norm": 0.40751364827156067, + "learning_rate": 0.0003973896591815841, + "loss": 2.6463, + "step": 4522 + }, + { + "epoch": 0.578944, + "grad_norm": 0.4519767761230469, + "learning_rate": 0.0003971867693578879, + "loss": 2.6364, + "step": 4523 + }, + { + "epoch": 0.579072, + "grad_norm": 0.46659737825393677, + "learning_rate": 0.00039698389720897774, + "loss": 2.6185, + "step": 4524 + }, + { + "epoch": 0.5792, + "grad_norm": 0.44820505380630493, + "learning_rate": 0.00039678104276973016, + "loss": 2.6432, + "step": 4525 + }, + { + "epoch": 0.579328, + "grad_norm": 0.5014375448226929, + "learning_rate": 0.0003965782060750178, + "loss": 2.6004, + "step": 4526 + }, + { + "epoch": 0.579456, + "grad_norm": 0.3943502902984619, + "learning_rate": 0.00039637538715971095, + "loss": 2.5355, + "step": 4527 + }, + { + "epoch": 0.579584, + "grad_norm": 0.41826850175857544, + "learning_rate": 0.00039617258605867633, + "loss": 2.6775, + "step": 4528 + }, + { + "epoch": 0.579712, + "grad_norm": 0.9254022240638733, + "learning_rate": 0.00039596980280677774, + "loss": 2.7249, + "step": 4529 + }, + { + "epoch": 0.57984, + "grad_norm": 0.44706347584724426, + "learning_rate": 0.0003957670374388762, + "loss": 2.7586, + "step": 4530 + }, + { + "epoch": 0.579968, + "grad_norm": 0.3979160189628601, + "learning_rate": 0.00039556428998982924, + "loss": 2.5962, + "step": 4531 + }, + { + "epoch": 0.580096, + "grad_norm": 0.5676912665367126, + "learning_rate": 0.00039536156049449153, + "loss": 2.5606, + "step": 4532 + }, + { + "epoch": 0.580224, + "grad_norm": 0.4059109687805176, + "learning_rate": 0.0003951588489877145, + "loss": 2.7163, + "step": 4533 + }, + { + "epoch": 0.580352, + "grad_norm": 0.4341028928756714, + "learning_rate": 0.0003949561555043469, + "loss": 2.7566, + "step": 4534 + }, + { + "epoch": 0.58048, + "grad_norm": 0.4015989303588867, + "learning_rate": 0.000394753480079234, + "loss": 2.6554, + "step": 4535 + }, + { + "epoch": 0.580608, + "grad_norm": 0.3944129943847656, + "learning_rate": 0.000394550822747218, + "loss": 2.6458, + "step": 4536 + }, + { + "epoch": 0.580736, + "grad_norm": 0.440176784992218, + "learning_rate": 0.00039434818354313787, + "loss": 2.5956, + "step": 4537 + }, + { + "epoch": 0.580864, + "grad_norm": 0.45402494072914124, + "learning_rate": 0.00039414556250183, + "loss": 2.5895, + "step": 4538 + }, + { + "epoch": 0.580992, + "grad_norm": 0.4163173735141754, + "learning_rate": 0.0003939429596581271, + "loss": 2.6212, + "step": 4539 + }, + { + "epoch": 0.58112, + "grad_norm": 0.46821799874305725, + "learning_rate": 0.00039374037504685876, + "loss": 2.4781, + "step": 4540 + }, + { + "epoch": 0.581248, + "grad_norm": 0.4191175103187561, + "learning_rate": 0.00039353780870285204, + "loss": 2.6607, + "step": 4541 + }, + { + "epoch": 0.581376, + "grad_norm": 0.4198140501976013, + "learning_rate": 0.00039333526066092993, + "loss": 2.6467, + "step": 4542 + }, + { + "epoch": 0.581504, + "grad_norm": 0.4122590124607086, + "learning_rate": 0.0003931327309559133, + "loss": 2.6512, + "step": 4543 + }, + { + "epoch": 0.581632, + "grad_norm": 0.8250200152397156, + "learning_rate": 0.000392930219622619, + "loss": 2.6441, + "step": 4544 + }, + { + "epoch": 0.58176, + "grad_norm": 0.8919051289558411, + "learning_rate": 0.0003927277266958614, + "loss": 2.6567, + "step": 4545 + }, + { + "epoch": 0.581888, + "grad_norm": 0.48951300978660583, + "learning_rate": 0.00039252525221045116, + "loss": 2.5935, + "step": 4546 + }, + { + "epoch": 0.582016, + "grad_norm": 0.44534552097320557, + "learning_rate": 0.00039232279620119586, + "loss": 2.7393, + "step": 4547 + }, + { + "epoch": 0.582144, + "grad_norm": 0.5100317597389221, + "learning_rate": 0.00039212035870290037, + "loss": 2.5862, + "step": 4548 + }, + { + "epoch": 0.582272, + "grad_norm": 0.4347313940525055, + "learning_rate": 0.00039191793975036583, + "loss": 2.7727, + "step": 4549 + }, + { + "epoch": 0.5824, + "grad_norm": 0.4631744921207428, + "learning_rate": 0.0003917155393783903, + "loss": 2.5681, + "step": 4550 + }, + { + "epoch": 0.582528, + "grad_norm": 0.457998126745224, + "learning_rate": 0.0003915131576217688, + "loss": 2.5385, + "step": 4551 + }, + { + "epoch": 0.582656, + "grad_norm": 0.4286135733127594, + "learning_rate": 0.0003913107945152934, + "loss": 2.6293, + "step": 4552 + }, + { + "epoch": 0.582784, + "grad_norm": 0.43017053604125977, + "learning_rate": 0.0003911084500937525, + "loss": 2.5954, + "step": 4553 + }, + { + "epoch": 0.582912, + "grad_norm": 0.47550931572914124, + "learning_rate": 0.00039090612439193127, + "loss": 2.6254, + "step": 4554 + }, + { + "epoch": 0.58304, + "grad_norm": 0.49304598569869995, + "learning_rate": 0.00039070381744461217, + "loss": 2.6223, + "step": 4555 + }, + { + "epoch": 0.583168, + "grad_norm": 0.4577657878398895, + "learning_rate": 0.0003905015292865738, + "loss": 2.747, + "step": 4556 + }, + { + "epoch": 0.583296, + "grad_norm": 1.0024709701538086, + "learning_rate": 0.00039029925995259204, + "loss": 2.5728, + "step": 4557 + }, + { + "epoch": 0.583424, + "grad_norm": 0.5268073678016663, + "learning_rate": 0.00039009700947743907, + "loss": 2.5917, + "step": 4558 + }, + { + "epoch": 0.583552, + "grad_norm": 0.5072221755981445, + "learning_rate": 0.00038989477789588443, + "loss": 2.6665, + "step": 4559 + }, + { + "epoch": 0.58368, + "grad_norm": 0.5007774829864502, + "learning_rate": 0.00038969256524269393, + "loss": 2.6868, + "step": 4560 + }, + { + "epoch": 0.583808, + "grad_norm": 3.416715621948242, + "learning_rate": 0.00038949037155263007, + "loss": 2.6013, + "step": 4561 + }, + { + "epoch": 0.583936, + "grad_norm": 0.45993706583976746, + "learning_rate": 0.00038928819686045253, + "loss": 2.6636, + "step": 4562 + }, + { + "epoch": 0.584064, + "grad_norm": 0.5862695574760437, + "learning_rate": 0.00038908604120091754, + "loss": 2.6018, + "step": 4563 + }, + { + "epoch": 0.584192, + "grad_norm": 0.4078100621700287, + "learning_rate": 0.00038888390460877794, + "loss": 2.6816, + "step": 4564 + }, + { + "epoch": 0.58432, + "grad_norm": 0.47273048758506775, + "learning_rate": 0.00038868178711878317, + "loss": 2.6286, + "step": 4565 + }, + { + "epoch": 0.584448, + "grad_norm": 0.5302479267120361, + "learning_rate": 0.0003884796887656798, + "loss": 2.6588, + "step": 4566 + }, + { + "epoch": 0.584576, + "grad_norm": 0.4754522442817688, + "learning_rate": 0.00038827760958421077, + "loss": 2.654, + "step": 4567 + }, + { + "epoch": 0.584704, + "grad_norm": 0.4260980486869812, + "learning_rate": 0.0003880755496091158, + "loss": 2.6497, + "step": 4568 + }, + { + "epoch": 0.584832, + "grad_norm": 0.4464724361896515, + "learning_rate": 0.00038787350887513144, + "loss": 2.6776, + "step": 4569 + }, + { + "epoch": 0.58496, + "grad_norm": 0.46338146924972534, + "learning_rate": 0.00038767148741699063, + "loss": 2.7026, + "step": 4570 + }, + { + "epoch": 0.585088, + "grad_norm": 0.5527905225753784, + "learning_rate": 0.00038746948526942353, + "loss": 2.622, + "step": 4571 + }, + { + "epoch": 0.585216, + "grad_norm": 0.43018317222595215, + "learning_rate": 0.0003872675024671564, + "loss": 2.6132, + "step": 4572 + }, + { + "epoch": 0.585344, + "grad_norm": 0.4376511573791504, + "learning_rate": 0.00038706553904491257, + "loss": 2.6652, + "step": 4573 + }, + { + "epoch": 0.585472, + "grad_norm": 0.4877801239490509, + "learning_rate": 0.00038686359503741185, + "loss": 2.577, + "step": 4574 + }, + { + "epoch": 0.5856, + "grad_norm": 0.6508839130401611, + "learning_rate": 0.0003866616704793706, + "loss": 2.6142, + "step": 4575 + }, + { + "epoch": 0.585728, + "grad_norm": 0.4064885973930359, + "learning_rate": 0.0003864597654055023, + "loss": 2.5893, + "step": 4576 + }, + { + "epoch": 0.585856, + "grad_norm": 0.5320581197738647, + "learning_rate": 0.00038625787985051654, + "loss": 2.7026, + "step": 4577 + }, + { + "epoch": 0.585984, + "grad_norm": 0.8278800845146179, + "learning_rate": 0.0003860560138491199, + "loss": 2.7302, + "step": 4578 + }, + { + "epoch": 0.586112, + "grad_norm": 0.4488581418991089, + "learning_rate": 0.0003858541674360151, + "loss": 2.6847, + "step": 4579 + }, + { + "epoch": 0.58624, + "grad_norm": 0.48947781324386597, + "learning_rate": 0.00038565234064590244, + "loss": 2.6188, + "step": 4580 + }, + { + "epoch": 0.586368, + "grad_norm": 0.3944290578365326, + "learning_rate": 0.00038545053351347815, + "loss": 2.6265, + "step": 4581 + }, + { + "epoch": 0.586496, + "grad_norm": 1.8691896200180054, + "learning_rate": 0.00038524874607343506, + "loss": 2.5956, + "step": 4582 + }, + { + "epoch": 0.586624, + "grad_norm": 0.4278256297111511, + "learning_rate": 0.00038504697836046274, + "loss": 2.6475, + "step": 4583 + }, + { + "epoch": 0.586752, + "grad_norm": 0.395406574010849, + "learning_rate": 0.00038484523040924765, + "loss": 2.5993, + "step": 4584 + }, + { + "epoch": 0.58688, + "grad_norm": 0.42678356170654297, + "learning_rate": 0.0003846435022544724, + "loss": 2.5685, + "step": 4585 + }, + { + "epoch": 0.587008, + "grad_norm": 0.40452152490615845, + "learning_rate": 0.0003844417939308163, + "loss": 2.6273, + "step": 4586 + }, + { + "epoch": 0.587136, + "grad_norm": 0.3910859525203705, + "learning_rate": 0.0003842401054729556, + "loss": 2.6983, + "step": 4587 + }, + { + "epoch": 0.587264, + "grad_norm": 0.3960724472999573, + "learning_rate": 0.0003840384369155626, + "loss": 2.595, + "step": 4588 + }, + { + "epoch": 0.587392, + "grad_norm": 0.43013903498649597, + "learning_rate": 0.0003838367882933068, + "loss": 2.6885, + "step": 4589 + }, + { + "epoch": 0.58752, + "grad_norm": 0.49028903245925903, + "learning_rate": 0.0003836351596408537, + "loss": 2.6273, + "step": 4590 + }, + { + "epoch": 0.587648, + "grad_norm": 0.40792447328567505, + "learning_rate": 0.00038343355099286583, + "loss": 2.5954, + "step": 4591 + }, + { + "epoch": 0.587776, + "grad_norm": 0.39642763137817383, + "learning_rate": 0.0003832319623840018, + "loss": 2.6313, + "step": 4592 + }, + { + "epoch": 0.587904, + "grad_norm": 0.39889827370643616, + "learning_rate": 0.00038303039384891715, + "loss": 2.676, + "step": 4593 + }, + { + "epoch": 0.588032, + "grad_norm": 0.39837318658828735, + "learning_rate": 0.0003828288454222638, + "loss": 2.5662, + "step": 4594 + }, + { + "epoch": 0.58816, + "grad_norm": 0.43782860040664673, + "learning_rate": 0.0003826273171386904, + "loss": 2.6322, + "step": 4595 + }, + { + "epoch": 0.588288, + "grad_norm": 0.4381295442581177, + "learning_rate": 0.0003824258090328418, + "loss": 2.6943, + "step": 4596 + }, + { + "epoch": 0.588416, + "grad_norm": 0.41415128111839294, + "learning_rate": 0.0003822243211393596, + "loss": 2.6112, + "step": 4597 + }, + { + "epoch": 0.588544, + "grad_norm": 0.4414929449558258, + "learning_rate": 0.0003820228534928818, + "loss": 2.6696, + "step": 4598 + }, + { + "epoch": 0.588672, + "grad_norm": 0.416482537984848, + "learning_rate": 0.0003818214061280435, + "loss": 2.6133, + "step": 4599 + }, + { + "epoch": 0.5888, + "grad_norm": 0.42350107431411743, + "learning_rate": 0.0003816199790794754, + "loss": 2.6588, + "step": 4600 + }, + { + "epoch": 0.588928, + "grad_norm": 0.41795843839645386, + "learning_rate": 0.0003814185723818053, + "loss": 2.6283, + "step": 4601 + }, + { + "epoch": 0.589056, + "grad_norm": 0.43084847927093506, + "learning_rate": 0.0003812171860696574, + "loss": 2.7004, + "step": 4602 + }, + { + "epoch": 0.589184, + "grad_norm": 0.41763192415237427, + "learning_rate": 0.00038101582017765226, + "loss": 2.5113, + "step": 4603 + }, + { + "epoch": 0.589312, + "grad_norm": 0.44403186440467834, + "learning_rate": 0.00038081447474040686, + "loss": 2.5671, + "step": 4604 + }, + { + "epoch": 0.58944, + "grad_norm": 0.43676307797431946, + "learning_rate": 0.00038061314979253506, + "loss": 2.5859, + "step": 4605 + }, + { + "epoch": 0.589568, + "grad_norm": 0.38509833812713623, + "learning_rate": 0.0003804118453686469, + "loss": 2.6066, + "step": 4606 + }, + { + "epoch": 0.589696, + "grad_norm": 0.46046754717826843, + "learning_rate": 0.0003802105615033488, + "loss": 2.8315, + "step": 4607 + }, + { + "epoch": 0.589824, + "grad_norm": 0.40069779753685, + "learning_rate": 0.00038000929823124385, + "loss": 2.5909, + "step": 4608 + }, + { + "epoch": 0.589952, + "grad_norm": 0.44431740045547485, + "learning_rate": 0.0003798080555869318, + "loss": 2.5805, + "step": 4609 + }, + { + "epoch": 0.59008, + "grad_norm": 0.3958064019680023, + "learning_rate": 0.00037960683360500826, + "loss": 2.6147, + "step": 4610 + }, + { + "epoch": 0.590208, + "grad_norm": 0.42153939604759216, + "learning_rate": 0.0003794056323200657, + "loss": 2.5988, + "step": 4611 + }, + { + "epoch": 0.590336, + "grad_norm": 0.4688696563243866, + "learning_rate": 0.00037920445176669313, + "loss": 2.5559, + "step": 4612 + }, + { + "epoch": 0.590464, + "grad_norm": 0.4057025611400604, + "learning_rate": 0.00037900329197947563, + "loss": 2.5828, + "step": 4613 + }, + { + "epoch": 0.590592, + "grad_norm": 0.42620667815208435, + "learning_rate": 0.00037880215299299487, + "loss": 2.5948, + "step": 4614 + }, + { + "epoch": 0.59072, + "grad_norm": 0.5258532762527466, + "learning_rate": 0.0003786010348418292, + "loss": 2.6884, + "step": 4615 + }, + { + "epoch": 0.590848, + "grad_norm": 0.45526957511901855, + "learning_rate": 0.0003783999375605527, + "loss": 2.6449, + "step": 4616 + }, + { + "epoch": 0.590976, + "grad_norm": 0.5264735221862793, + "learning_rate": 0.0003781988611837369, + "loss": 2.7754, + "step": 4617 + }, + { + "epoch": 0.591104, + "grad_norm": 0.42343711853027344, + "learning_rate": 0.000377997805745949, + "loss": 2.7202, + "step": 4618 + }, + { + "epoch": 0.591232, + "grad_norm": 0.43366125226020813, + "learning_rate": 0.00037779677128175245, + "loss": 2.6803, + "step": 4619 + }, + { + "epoch": 0.59136, + "grad_norm": 0.7336463928222656, + "learning_rate": 0.0003775957578257078, + "loss": 2.6225, + "step": 4620 + }, + { + "epoch": 0.591488, + "grad_norm": 0.4624364972114563, + "learning_rate": 0.0003773947654123713, + "loss": 2.6379, + "step": 4621 + }, + { + "epoch": 0.591616, + "grad_norm": 0.40690353512763977, + "learning_rate": 0.00037719379407629593, + "loss": 2.6902, + "step": 4622 + }, + { + "epoch": 0.591744, + "grad_norm": 1.7006784677505493, + "learning_rate": 0.0003769928438520312, + "loss": 2.63, + "step": 4623 + }, + { + "epoch": 0.591872, + "grad_norm": 0.39683932065963745, + "learning_rate": 0.0003767919147741226, + "loss": 2.7855, + "step": 4624 + }, + { + "epoch": 0.592, + "grad_norm": 0.47553667426109314, + "learning_rate": 0.000376591006877112, + "loss": 2.5948, + "step": 4625 + }, + { + "epoch": 0.592128, + "grad_norm": 0.7201041579246521, + "learning_rate": 0.0003763901201955381, + "loss": 2.6505, + "step": 4626 + }, + { + "epoch": 0.592256, + "grad_norm": 0.4402835965156555, + "learning_rate": 0.00037618925476393567, + "loss": 2.7023, + "step": 4627 + }, + { + "epoch": 0.592384, + "grad_norm": 0.41544002294540405, + "learning_rate": 0.0003759884106168356, + "loss": 2.7439, + "step": 4628 + }, + { + "epoch": 0.592512, + "grad_norm": 0.43705064058303833, + "learning_rate": 0.00037578758778876545, + "loss": 2.604, + "step": 4629 + }, + { + "epoch": 0.59264, + "grad_norm": 0.8710454702377319, + "learning_rate": 0.000375586786314249, + "loss": 2.696, + "step": 4630 + }, + { + "epoch": 0.592768, + "grad_norm": 0.4348949193954468, + "learning_rate": 0.00037538600622780626, + "loss": 2.6858, + "step": 4631 + }, + { + "epoch": 0.592896, + "grad_norm": 0.4133608341217041, + "learning_rate": 0.0003751852475639536, + "loss": 2.6647, + "step": 4632 + }, + { + "epoch": 0.593024, + "grad_norm": 0.40428128838539124, + "learning_rate": 0.000374984510357204, + "loss": 2.7243, + "step": 4633 + }, + { + "epoch": 0.593152, + "grad_norm": 0.4019679129123688, + "learning_rate": 0.00037478379464206635, + "loss": 2.6208, + "step": 4634 + }, + { + "epoch": 0.59328, + "grad_norm": 0.4381609559059143, + "learning_rate": 0.00037458310045304584, + "loss": 2.6977, + "step": 4635 + }, + { + "epoch": 0.593408, + "grad_norm": 0.5094555616378784, + "learning_rate": 0.0003743824278246445, + "loss": 2.6435, + "step": 4636 + }, + { + "epoch": 0.593536, + "grad_norm": 0.4461229145526886, + "learning_rate": 0.00037418177679136, + "loss": 2.5867, + "step": 4637 + }, + { + "epoch": 0.593664, + "grad_norm": 0.5176783204078674, + "learning_rate": 0.00037398114738768686, + "loss": 2.5316, + "step": 4638 + }, + { + "epoch": 0.593792, + "grad_norm": 0.4257391691207886, + "learning_rate": 0.0003737805396481153, + "loss": 2.607, + "step": 4639 + }, + { + "epoch": 0.59392, + "grad_norm": 0.43047088384628296, + "learning_rate": 0.0003735799536071322, + "loss": 2.6116, + "step": 4640 + }, + { + "epoch": 0.594048, + "grad_norm": 0.38390403985977173, + "learning_rate": 0.00037337938929922075, + "loss": 2.4819, + "step": 4641 + }, + { + "epoch": 0.594176, + "grad_norm": 0.4016527235507965, + "learning_rate": 0.00037317884675886024, + "loss": 2.5765, + "step": 4642 + }, + { + "epoch": 0.594304, + "grad_norm": 0.41984063386917114, + "learning_rate": 0.00037297832602052605, + "loss": 2.6297, + "step": 4643 + }, + { + "epoch": 0.594432, + "grad_norm": 0.38718441128730774, + "learning_rate": 0.0003727778271186901, + "loss": 2.6246, + "step": 4644 + }, + { + "epoch": 0.59456, + "grad_norm": 0.41334038972854614, + "learning_rate": 0.00037257735008782073, + "loss": 2.6938, + "step": 4645 + }, + { + "epoch": 0.594688, + "grad_norm": 0.4360121786594391, + "learning_rate": 0.00037237689496238214, + "loss": 2.6891, + "step": 4646 + }, + { + "epoch": 0.594816, + "grad_norm": 0.48514243960380554, + "learning_rate": 0.00037217646177683465, + "loss": 2.6626, + "step": 4647 + }, + { + "epoch": 0.594944, + "grad_norm": 0.5790982246398926, + "learning_rate": 0.00037197605056563544, + "loss": 2.7304, + "step": 4648 + }, + { + "epoch": 0.595072, + "grad_norm": 0.3931313455104828, + "learning_rate": 0.00037177566136323726, + "loss": 2.6597, + "step": 4649 + }, + { + "epoch": 0.5952, + "grad_norm": 0.3664645254611969, + "learning_rate": 0.00037157529420408933, + "loss": 2.7641, + "step": 4650 + }, + { + "epoch": 0.595328, + "grad_norm": 0.3898569345474243, + "learning_rate": 0.0003713749491226372, + "loss": 2.6783, + "step": 4651 + }, + { + "epoch": 0.595456, + "grad_norm": 0.4366317689418793, + "learning_rate": 0.00037117462615332254, + "loss": 2.6232, + "step": 4652 + }, + { + "epoch": 0.595584, + "grad_norm": 0.3913480043411255, + "learning_rate": 0.00037097432533058283, + "loss": 2.4965, + "step": 4653 + }, + { + "epoch": 0.595712, + "grad_norm": 0.3825232982635498, + "learning_rate": 0.00037077404668885264, + "loss": 2.6196, + "step": 4654 + }, + { + "epoch": 0.59584, + "grad_norm": 0.4291878044605255, + "learning_rate": 0.00037057379026256184, + "loss": 2.6429, + "step": 4655 + }, + { + "epoch": 0.595968, + "grad_norm": 0.4807449281215668, + "learning_rate": 0.000370373556086137, + "loss": 2.7524, + "step": 4656 + }, + { + "epoch": 0.596096, + "grad_norm": 0.40942618250846863, + "learning_rate": 0.0003701733441940007, + "loss": 2.6469, + "step": 4657 + }, + { + "epoch": 0.596224, + "grad_norm": 0.41929247975349426, + "learning_rate": 0.00036997315462057145, + "loss": 2.5889, + "step": 4658 + }, + { + "epoch": 0.596352, + "grad_norm": 0.3748518228530884, + "learning_rate": 0.00036977298740026445, + "loss": 2.7046, + "step": 4659 + }, + { + "epoch": 0.59648, + "grad_norm": 0.4213089942932129, + "learning_rate": 0.0003695728425674906, + "loss": 2.4683, + "step": 4660 + }, + { + "epoch": 0.596608, + "grad_norm": 0.37997621297836304, + "learning_rate": 0.00036937272015665705, + "loss": 2.5724, + "step": 4661 + }, + { + "epoch": 0.596736, + "grad_norm": 0.40500155091285706, + "learning_rate": 0.00036917262020216726, + "loss": 2.6703, + "step": 4662 + }, + { + "epoch": 0.596864, + "grad_norm": 0.40431156754493713, + "learning_rate": 0.00036897254273842085, + "loss": 2.6559, + "step": 4663 + }, + { + "epoch": 0.596992, + "grad_norm": 0.42184266448020935, + "learning_rate": 0.0003687724877998133, + "loss": 2.672, + "step": 4664 + }, + { + "epoch": 0.59712, + "grad_norm": 0.41119688749313354, + "learning_rate": 0.00036857245542073645, + "loss": 2.687, + "step": 4665 + }, + { + "epoch": 0.597248, + "grad_norm": 0.6714110374450684, + "learning_rate": 0.00036837244563557814, + "loss": 2.6156, + "step": 4666 + }, + { + "epoch": 0.597376, + "grad_norm": 0.36600422859191895, + "learning_rate": 0.0003681724584787225, + "loss": 2.5697, + "step": 4667 + }, + { + "epoch": 0.597504, + "grad_norm": 0.4398244619369507, + "learning_rate": 0.00036797249398454933, + "loss": 2.6702, + "step": 4668 + }, + { + "epoch": 0.597632, + "grad_norm": 0.39946359395980835, + "learning_rate": 0.00036777255218743514, + "loss": 2.72, + "step": 4669 + }, + { + "epoch": 0.59776, + "grad_norm": 0.3912281095981598, + "learning_rate": 0.00036757263312175217, + "loss": 2.664, + "step": 4670 + }, + { + "epoch": 0.597888, + "grad_norm": 0.38609614968299866, + "learning_rate": 0.0003673727368218687, + "loss": 2.684, + "step": 4671 + }, + { + "epoch": 0.598016, + "grad_norm": 0.38429170846939087, + "learning_rate": 0.00036717286332214926, + "loss": 2.6223, + "step": 4672 + }, + { + "epoch": 0.598144, + "grad_norm": 0.3874562382698059, + "learning_rate": 0.0003669730126569545, + "loss": 2.6144, + "step": 4673 + }, + { + "epoch": 0.598272, + "grad_norm": 0.39705923199653625, + "learning_rate": 0.00036677318486064126, + "loss": 2.62, + "step": 4674 + }, + { + "epoch": 0.5984, + "grad_norm": 0.40542060136795044, + "learning_rate": 0.0003665733799675619, + "loss": 2.5704, + "step": 4675 + }, + { + "epoch": 0.598528, + "grad_norm": 0.459948867559433, + "learning_rate": 0.0003663735980120653, + "loss": 2.8126, + "step": 4676 + }, + { + "epoch": 0.598656, + "grad_norm": 0.37912046909332275, + "learning_rate": 0.00036617383902849644, + "loss": 2.6219, + "step": 4677 + }, + { + "epoch": 0.598784, + "grad_norm": 0.5157961249351501, + "learning_rate": 0.000365974103051196, + "loss": 2.5931, + "step": 4678 + }, + { + "epoch": 0.598912, + "grad_norm": 0.3904270529747009, + "learning_rate": 0.0003657743901145011, + "loss": 2.6295, + "step": 4679 + }, + { + "epoch": 0.59904, + "grad_norm": 0.4050494134426117, + "learning_rate": 0.00036557470025274454, + "loss": 2.6178, + "step": 4680 + }, + { + "epoch": 0.599168, + "grad_norm": 0.36732497811317444, + "learning_rate": 0.0003653750335002552, + "loss": 2.755, + "step": 4681 + }, + { + "epoch": 0.599296, + "grad_norm": 0.3847680389881134, + "learning_rate": 0.00036517538989135865, + "loss": 2.6607, + "step": 4682 + }, + { + "epoch": 0.599424, + "grad_norm": 0.42124906182289124, + "learning_rate": 0.0003649757694603754, + "loss": 2.6238, + "step": 4683 + }, + { + "epoch": 0.599552, + "grad_norm": 0.429205983877182, + "learning_rate": 0.00036477617224162285, + "loss": 2.736, + "step": 4684 + }, + { + "epoch": 0.59968, + "grad_norm": 0.3862900137901306, + "learning_rate": 0.0003645765982694139, + "loss": 2.5713, + "step": 4685 + }, + { + "epoch": 0.599808, + "grad_norm": 0.4693947434425354, + "learning_rate": 0.0003643770475780576, + "loss": 2.6162, + "step": 4686 + }, + { + "epoch": 0.599936, + "grad_norm": 0.4024186134338379, + "learning_rate": 0.00036417752020185916, + "loss": 2.6323, + "step": 4687 + }, + { + "epoch": 0.600064, + "grad_norm": 0.48671308159828186, + "learning_rate": 0.0003639780161751196, + "loss": 2.5894, + "step": 4688 + }, + { + "epoch": 0.600192, + "grad_norm": 0.39929717779159546, + "learning_rate": 0.00036377853553213593, + "loss": 2.6727, + "step": 4689 + }, + { + "epoch": 0.60032, + "grad_norm": 0.39655447006225586, + "learning_rate": 0.00036357907830720116, + "loss": 2.6231, + "step": 4690 + }, + { + "epoch": 0.600448, + "grad_norm": 0.39157775044441223, + "learning_rate": 0.0003633796445346044, + "loss": 2.5814, + "step": 4691 + }, + { + "epoch": 0.600576, + "grad_norm": 0.41847023367881775, + "learning_rate": 0.00036318023424863053, + "loss": 2.6993, + "step": 4692 + }, + { + "epoch": 0.600704, + "grad_norm": 0.41996878385543823, + "learning_rate": 0.00036298084748356076, + "loss": 2.6299, + "step": 4693 + }, + { + "epoch": 0.600832, + "grad_norm": 0.3840958774089813, + "learning_rate": 0.0003627814842736715, + "loss": 2.5562, + "step": 4694 + }, + { + "epoch": 0.60096, + "grad_norm": 0.4376620352268219, + "learning_rate": 0.000362582144653236, + "loss": 2.5968, + "step": 4695 + }, + { + "epoch": 0.601088, + "grad_norm": 0.720391571521759, + "learning_rate": 0.000362382828656523, + "loss": 2.7674, + "step": 4696 + }, + { + "epoch": 0.601216, + "grad_norm": 0.38719284534454346, + "learning_rate": 0.000362183536317797, + "loss": 2.7008, + "step": 4697 + }, + { + "epoch": 0.601344, + "grad_norm": 0.3905613124370575, + "learning_rate": 0.0003619842676713189, + "loss": 2.6646, + "step": 4698 + }, + { + "epoch": 0.601472, + "grad_norm": 0.38424450159072876, + "learning_rate": 0.00036178502275134507, + "loss": 2.6057, + "step": 4699 + }, + { + "epoch": 0.6016, + "grad_norm": 0.4266892969608307, + "learning_rate": 0.0003615858015921283, + "loss": 2.622, + "step": 4700 + }, + { + "epoch": 0.601728, + "grad_norm": 0.38126054406166077, + "learning_rate": 0.00036138660422791667, + "loss": 2.6179, + "step": 4701 + }, + { + "epoch": 0.601856, + "grad_norm": 0.3952687382698059, + "learning_rate": 0.000361187430692955, + "loss": 2.6178, + "step": 4702 + }, + { + "epoch": 0.601984, + "grad_norm": 0.37331879138946533, + "learning_rate": 0.0003609882810214832, + "loss": 2.6696, + "step": 4703 + }, + { + "epoch": 0.602112, + "grad_norm": 0.3996114134788513, + "learning_rate": 0.0003607891552477373, + "loss": 2.6594, + "step": 4704 + }, + { + "epoch": 0.60224, + "grad_norm": 0.3956165015697479, + "learning_rate": 0.00036059005340594957, + "loss": 2.625, + "step": 4705 + }, + { + "epoch": 0.602368, + "grad_norm": 2.815762519836426, + "learning_rate": 0.0003603909755303479, + "loss": 2.5728, + "step": 4706 + }, + { + "epoch": 0.602496, + "grad_norm": 0.4609861969947815, + "learning_rate": 0.0003601919216551559, + "loss": 2.6086, + "step": 4707 + }, + { + "epoch": 0.602624, + "grad_norm": 0.4448070526123047, + "learning_rate": 0.0003599928918145934, + "loss": 2.5844, + "step": 4708 + }, + { + "epoch": 0.602752, + "grad_norm": 0.404475599527359, + "learning_rate": 0.00035979388604287575, + "loss": 2.6892, + "step": 4709 + }, + { + "epoch": 0.60288, + "grad_norm": 0.5179335474967957, + "learning_rate": 0.0003595949043742147, + "loss": 2.6538, + "step": 4710 + }, + { + "epoch": 0.603008, + "grad_norm": 0.5004408955574036, + "learning_rate": 0.00035939594684281735, + "loss": 2.646, + "step": 4711 + }, + { + "epoch": 0.603136, + "grad_norm": 0.4922592341899872, + "learning_rate": 0.0003591970134828866, + "loss": 2.6866, + "step": 4712 + }, + { + "epoch": 0.603264, + "grad_norm": 0.4687597453594208, + "learning_rate": 0.00035899810432862166, + "loss": 2.6896, + "step": 4713 + }, + { + "epoch": 0.603392, + "grad_norm": 0.5932851433753967, + "learning_rate": 0.0003587992194142173, + "loss": 2.6766, + "step": 4714 + }, + { + "epoch": 0.60352, + "grad_norm": 0.5022562742233276, + "learning_rate": 0.00035860035877386385, + "loss": 2.7279, + "step": 4715 + }, + { + "epoch": 0.603648, + "grad_norm": 0.4977573752403259, + "learning_rate": 0.0003584015224417482, + "loss": 2.7174, + "step": 4716 + }, + { + "epoch": 0.603776, + "grad_norm": 0.5035923719406128, + "learning_rate": 0.0003582027104520523, + "loss": 2.7161, + "step": 4717 + }, + { + "epoch": 0.603904, + "grad_norm": 0.47569069266319275, + "learning_rate": 0.0003580039228389541, + "loss": 2.6361, + "step": 4718 + }, + { + "epoch": 0.604032, + "grad_norm": 0.39854127168655396, + "learning_rate": 0.00035780515963662786, + "loss": 2.4843, + "step": 4719 + }, + { + "epoch": 0.60416, + "grad_norm": 0.4387221932411194, + "learning_rate": 0.0003576064208792433, + "loss": 2.6689, + "step": 4720 + }, + { + "epoch": 0.604288, + "grad_norm": 0.3983166217803955, + "learning_rate": 0.0003574077066009657, + "loss": 2.6283, + "step": 4721 + }, + { + "epoch": 0.604416, + "grad_norm": 0.4021855890750885, + "learning_rate": 0.0003572090168359563, + "loss": 2.5869, + "step": 4722 + }, + { + "epoch": 0.604544, + "grad_norm": 0.39995506405830383, + "learning_rate": 0.0003570103516183724, + "loss": 2.6414, + "step": 4723 + }, + { + "epoch": 0.604672, + "grad_norm": 0.3866463899612427, + "learning_rate": 0.00035681171098236674, + "loss": 2.6922, + "step": 4724 + }, + { + "epoch": 0.6048, + "grad_norm": 0.3708978593349457, + "learning_rate": 0.0003566130949620878, + "loss": 2.5214, + "step": 4725 + }, + { + "epoch": 0.604928, + "grad_norm": 0.44341087341308594, + "learning_rate": 0.0003564145035916803, + "loss": 2.7834, + "step": 4726 + }, + { + "epoch": 0.605056, + "grad_norm": 0.4812960922718048, + "learning_rate": 0.000356215936905284, + "loss": 2.6938, + "step": 4727 + }, + { + "epoch": 0.605184, + "grad_norm": 0.40861812233924866, + "learning_rate": 0.00035601739493703515, + "loss": 2.654, + "step": 4728 + }, + { + "epoch": 0.605312, + "grad_norm": 0.4085811972618103, + "learning_rate": 0.00035581887772106536, + "loss": 2.6636, + "step": 4729 + }, + { + "epoch": 0.60544, + "grad_norm": 0.42695221304893494, + "learning_rate": 0.00035562038529150184, + "loss": 2.6669, + "step": 4730 + }, + { + "epoch": 0.605568, + "grad_norm": 0.4325527250766754, + "learning_rate": 0.00035542191768246807, + "loss": 2.6615, + "step": 4731 + }, + { + "epoch": 0.605696, + "grad_norm": 0.3970029354095459, + "learning_rate": 0.0003552234749280827, + "loss": 2.4821, + "step": 4732 + }, + { + "epoch": 0.605824, + "grad_norm": 0.3982072174549103, + "learning_rate": 0.00035502505706246025, + "loss": 2.6311, + "step": 4733 + }, + { + "epoch": 0.605952, + "grad_norm": 0.4579387307167053, + "learning_rate": 0.0003548266641197113, + "loss": 2.5267, + "step": 4734 + }, + { + "epoch": 0.60608, + "grad_norm": 0.39069488644599915, + "learning_rate": 0.0003546282961339418, + "loss": 2.5639, + "step": 4735 + }, + { + "epoch": 0.606208, + "grad_norm": 0.4101778566837311, + "learning_rate": 0.0003544299531392533, + "loss": 2.6321, + "step": 4736 + }, + { + "epoch": 0.606336, + "grad_norm": 0.4233366549015045, + "learning_rate": 0.00035423163516974356, + "loss": 2.5965, + "step": 4737 + }, + { + "epoch": 0.606464, + "grad_norm": 0.3689652383327484, + "learning_rate": 0.0003540333422595057, + "loss": 2.5986, + "step": 4738 + }, + { + "epoch": 0.606592, + "grad_norm": 0.42743685841560364, + "learning_rate": 0.0003538350744426285, + "loss": 2.6651, + "step": 4739 + }, + { + "epoch": 0.60672, + "grad_norm": 0.3799985945224762, + "learning_rate": 0.0003536368317531963, + "loss": 2.6255, + "step": 4740 + }, + { + "epoch": 0.606848, + "grad_norm": 0.40586522221565247, + "learning_rate": 0.00035343861422528974, + "loss": 2.5969, + "step": 4741 + }, + { + "epoch": 0.606976, + "grad_norm": 0.41935232281684875, + "learning_rate": 0.00035324042189298447, + "loss": 2.6179, + "step": 4742 + }, + { + "epoch": 0.607104, + "grad_norm": 0.3872668743133545, + "learning_rate": 0.00035304225479035193, + "loss": 2.6174, + "step": 4743 + }, + { + "epoch": 0.607232, + "grad_norm": 0.39406752586364746, + "learning_rate": 0.0003528441129514595, + "loss": 2.629, + "step": 4744 + }, + { + "epoch": 0.60736, + "grad_norm": 0.5479130148887634, + "learning_rate": 0.0003526459964103701, + "loss": 2.5914, + "step": 4745 + }, + { + "epoch": 0.607488, + "grad_norm": 0.39131563901901245, + "learning_rate": 0.000352447905201142, + "loss": 2.5226, + "step": 4746 + }, + { + "epoch": 0.607616, + "grad_norm": 0.430257648229599, + "learning_rate": 0.0003522498393578297, + "loss": 2.7716, + "step": 4747 + }, + { + "epoch": 0.607744, + "grad_norm": 0.40319183468818665, + "learning_rate": 0.0003520517989144829, + "loss": 2.6544, + "step": 4748 + }, + { + "epoch": 0.607872, + "grad_norm": 0.3877379894256592, + "learning_rate": 0.0003518537839051471, + "loss": 2.6205, + "step": 4749 + }, + { + "epoch": 0.608, + "grad_norm": 0.4050053358078003, + "learning_rate": 0.0003516557943638632, + "loss": 2.6525, + "step": 4750 + }, + { + "epoch": 0.608128, + "grad_norm": 0.4213573634624481, + "learning_rate": 0.00035145783032466797, + "loss": 2.5258, + "step": 4751 + }, + { + "epoch": 0.608256, + "grad_norm": 0.3587777614593506, + "learning_rate": 0.00035125989182159393, + "loss": 2.7447, + "step": 4752 + }, + { + "epoch": 0.608384, + "grad_norm": 0.4206939935684204, + "learning_rate": 0.0003510619788886687, + "loss": 2.6029, + "step": 4753 + }, + { + "epoch": 0.608512, + "grad_norm": 0.4196014404296875, + "learning_rate": 0.0003508640915599159, + "loss": 2.6694, + "step": 4754 + }, + { + "epoch": 0.60864, + "grad_norm": 0.4459465742111206, + "learning_rate": 0.0003506662298693546, + "loss": 2.5809, + "step": 4755 + }, + { + "epoch": 0.608768, + "grad_norm": 0.43471622467041016, + "learning_rate": 0.00035046839385099977, + "loss": 2.5721, + "step": 4756 + }, + { + "epoch": 0.608896, + "grad_norm": 0.4099656939506531, + "learning_rate": 0.0003502705835388616, + "loss": 2.6829, + "step": 4757 + }, + { + "epoch": 0.609024, + "grad_norm": 0.47615769505500793, + "learning_rate": 0.00035007279896694576, + "loss": 2.6359, + "step": 4758 + }, + { + "epoch": 0.609152, + "grad_norm": 0.40208810567855835, + "learning_rate": 0.0003498750401692541, + "loss": 2.6573, + "step": 4759 + }, + { + "epoch": 0.60928, + "grad_norm": 0.4290124177932739, + "learning_rate": 0.00034967730717978335, + "loss": 2.6781, + "step": 4760 + }, + { + "epoch": 0.609408, + "grad_norm": 0.37656599283218384, + "learning_rate": 0.0003494796000325261, + "loss": 2.6841, + "step": 4761 + }, + { + "epoch": 0.609536, + "grad_norm": 0.38396593928337097, + "learning_rate": 0.0003492819187614707, + "loss": 2.6629, + "step": 4762 + }, + { + "epoch": 0.609664, + "grad_norm": 0.40243422985076904, + "learning_rate": 0.00034908426340060075, + "loss": 2.5653, + "step": 4763 + }, + { + "epoch": 0.609792, + "grad_norm": 0.4116131067276001, + "learning_rate": 0.00034888663398389524, + "loss": 2.6441, + "step": 4764 + }, + { + "epoch": 0.60992, + "grad_norm": 0.43904736638069153, + "learning_rate": 0.0003486890305453293, + "loss": 2.7133, + "step": 4765 + }, + { + "epoch": 0.610048, + "grad_norm": 0.49391013383865356, + "learning_rate": 0.0003484914531188733, + "loss": 2.6296, + "step": 4766 + }, + { + "epoch": 0.610176, + "grad_norm": 0.599597156047821, + "learning_rate": 0.00034829390173849297, + "loss": 2.6662, + "step": 4767 + }, + { + "epoch": 0.610304, + "grad_norm": 0.3860556483268738, + "learning_rate": 0.00034809637643814964, + "loss": 2.4887, + "step": 4768 + }, + { + "epoch": 0.610432, + "grad_norm": 0.38421693444252014, + "learning_rate": 0.0003478988772518001, + "loss": 2.7621, + "step": 4769 + }, + { + "epoch": 0.61056, + "grad_norm": 0.4244097173213959, + "learning_rate": 0.000347701404213397, + "loss": 2.5966, + "step": 4770 + }, + { + "epoch": 0.610688, + "grad_norm": 0.4079534709453583, + "learning_rate": 0.00034750395735688803, + "loss": 2.6236, + "step": 4771 + }, + { + "epoch": 0.610816, + "grad_norm": 0.4488288164138794, + "learning_rate": 0.00034730653671621663, + "loss": 2.6459, + "step": 4772 + }, + { + "epoch": 0.610944, + "grad_norm": 0.37888985872268677, + "learning_rate": 0.0003471091423253216, + "loss": 2.6445, + "step": 4773 + }, + { + "epoch": 0.611072, + "grad_norm": 0.41185683012008667, + "learning_rate": 0.0003469117742181378, + "loss": 2.6039, + "step": 4774 + }, + { + "epoch": 0.6112, + "grad_norm": 0.40209707617759705, + "learning_rate": 0.00034671443242859464, + "loss": 2.6946, + "step": 4775 + }, + { + "epoch": 0.611328, + "grad_norm": 0.46361494064331055, + "learning_rate": 0.0003465171169906175, + "loss": 2.5749, + "step": 4776 + }, + { + "epoch": 0.611456, + "grad_norm": 0.45178982615470886, + "learning_rate": 0.00034631982793812744, + "loss": 2.6369, + "step": 4777 + }, + { + "epoch": 0.611584, + "grad_norm": 0.4557405710220337, + "learning_rate": 0.0003461225653050405, + "loss": 2.6265, + "step": 4778 + }, + { + "epoch": 0.611712, + "grad_norm": 0.38750723004341125, + "learning_rate": 0.0003459253291252684, + "loss": 2.5949, + "step": 4779 + }, + { + "epoch": 0.61184, + "grad_norm": 0.4472619593143463, + "learning_rate": 0.00034572811943271855, + "loss": 2.612, + "step": 4780 + }, + { + "epoch": 0.611968, + "grad_norm": 0.39342132210731506, + "learning_rate": 0.0003455309362612934, + "loss": 2.5604, + "step": 4781 + }, + { + "epoch": 0.612096, + "grad_norm": 0.5255709886550903, + "learning_rate": 0.000345333779644891, + "loss": 2.6794, + "step": 4782 + }, + { + "epoch": 0.612224, + "grad_norm": 0.42997393012046814, + "learning_rate": 0.00034513664961740475, + "loss": 2.6566, + "step": 4783 + }, + { + "epoch": 0.612352, + "grad_norm": 0.4055844843387604, + "learning_rate": 0.00034493954621272403, + "loss": 2.6461, + "step": 4784 + }, + { + "epoch": 0.61248, + "grad_norm": 0.4663175046443939, + "learning_rate": 0.00034474246946473286, + "loss": 2.607, + "step": 4785 + }, + { + "epoch": 0.612608, + "grad_norm": 0.44858643412590027, + "learning_rate": 0.0003445454194073111, + "loss": 2.7474, + "step": 4786 + }, + { + "epoch": 0.612736, + "grad_norm": 0.3969683349132538, + "learning_rate": 0.00034434839607433394, + "loss": 2.665, + "step": 4787 + }, + { + "epoch": 0.612864, + "grad_norm": 0.3831048309803009, + "learning_rate": 0.000344151399499672, + "loss": 2.6833, + "step": 4788 + }, + { + "epoch": 0.612992, + "grad_norm": 0.5297566056251526, + "learning_rate": 0.00034395442971719124, + "loss": 2.7634, + "step": 4789 + }, + { + "epoch": 0.61312, + "grad_norm": 0.3879701495170593, + "learning_rate": 0.0003437574867607529, + "loss": 2.6231, + "step": 4790 + }, + { + "epoch": 0.613248, + "grad_norm": 0.462617427110672, + "learning_rate": 0.0003435605706642141, + "loss": 2.6466, + "step": 4791 + }, + { + "epoch": 0.613376, + "grad_norm": 0.3726422190666199, + "learning_rate": 0.00034336368146142647, + "loss": 2.6447, + "step": 4792 + }, + { + "epoch": 0.613504, + "grad_norm": 0.3506596088409424, + "learning_rate": 0.0003431668191862382, + "loss": 2.5488, + "step": 4793 + }, + { + "epoch": 0.613632, + "grad_norm": 0.39872580766677856, + "learning_rate": 0.0003429699838724917, + "loss": 2.5818, + "step": 4794 + }, + { + "epoch": 0.61376, + "grad_norm": 0.40842631459236145, + "learning_rate": 0.0003427731755540256, + "loss": 2.6131, + "step": 4795 + }, + { + "epoch": 0.613888, + "grad_norm": 0.4400881230831146, + "learning_rate": 0.0003425763942646733, + "loss": 2.5638, + "step": 4796 + }, + { + "epoch": 0.614016, + "grad_norm": 0.40201425552368164, + "learning_rate": 0.00034237964003826373, + "loss": 2.6083, + "step": 4797 + }, + { + "epoch": 0.614144, + "grad_norm": 0.44499120116233826, + "learning_rate": 0.0003421829129086215, + "loss": 2.6226, + "step": 4798 + }, + { + "epoch": 0.614272, + "grad_norm": 0.5029693245887756, + "learning_rate": 0.000341986212909566, + "loss": 2.7009, + "step": 4799 + }, + { + "epoch": 0.6144, + "grad_norm": 0.4298045039176941, + "learning_rate": 0.00034178954007491237, + "loss": 2.7009, + "step": 4800 + }, + { + "epoch": 0.614528, + "grad_norm": 0.4559882879257202, + "learning_rate": 0.00034159289443847076, + "loss": 2.6577, + "step": 4801 + }, + { + "epoch": 0.614656, + "grad_norm": 0.4057038724422455, + "learning_rate": 0.00034139627603404723, + "loss": 2.7401, + "step": 4802 + }, + { + "epoch": 0.614784, + "grad_norm": 0.4399392008781433, + "learning_rate": 0.0003411996848954425, + "loss": 2.6126, + "step": 4803 + }, + { + "epoch": 0.614912, + "grad_norm": 0.3977852165699005, + "learning_rate": 0.0003410031210564528, + "loss": 2.6477, + "step": 4804 + }, + { + "epoch": 0.61504, + "grad_norm": 0.43043190240859985, + "learning_rate": 0.00034080658455087, + "loss": 2.6562, + "step": 4805 + }, + { + "epoch": 0.615168, + "grad_norm": 0.3858686089515686, + "learning_rate": 0.0003406100754124809, + "loss": 2.6225, + "step": 4806 + }, + { + "epoch": 0.615296, + "grad_norm": 0.35044586658477783, + "learning_rate": 0.0003404135936750674, + "loss": 2.6128, + "step": 4807 + }, + { + "epoch": 0.615424, + "grad_norm": 0.6066361665725708, + "learning_rate": 0.00034021713937240747, + "loss": 2.718, + "step": 4808 + }, + { + "epoch": 0.615552, + "grad_norm": 0.38304585218429565, + "learning_rate": 0.00034002071253827357, + "loss": 2.5758, + "step": 4809 + }, + { + "epoch": 0.61568, + "grad_norm": 0.39926326274871826, + "learning_rate": 0.00033982431320643357, + "loss": 2.7441, + "step": 4810 + }, + { + "epoch": 0.615808, + "grad_norm": 0.4184373617172241, + "learning_rate": 0.0003396279414106513, + "loss": 2.7011, + "step": 4811 + }, + { + "epoch": 0.615936, + "grad_norm": 0.341718852519989, + "learning_rate": 0.000339431597184685, + "loss": 2.5666, + "step": 4812 + }, + { + "epoch": 0.616064, + "grad_norm": 0.41389232873916626, + "learning_rate": 0.00033923528056228875, + "loss": 2.6749, + "step": 4813 + }, + { + "epoch": 0.616192, + "grad_norm": 0.48231637477874756, + "learning_rate": 0.0003390389915772115, + "loss": 2.5549, + "step": 4814 + }, + { + "epoch": 0.61632, + "grad_norm": 0.3807574212551117, + "learning_rate": 0.00033884273026319747, + "loss": 2.453, + "step": 4815 + }, + { + "epoch": 0.616448, + "grad_norm": 0.3829277753829956, + "learning_rate": 0.0003386464966539865, + "loss": 2.7068, + "step": 4816 + }, + { + "epoch": 0.616576, + "grad_norm": 0.3867637515068054, + "learning_rate": 0.0003384502907833134, + "loss": 2.6317, + "step": 4817 + }, + { + "epoch": 0.616704, + "grad_norm": 0.4140404462814331, + "learning_rate": 0.000338254112684908, + "loss": 2.622, + "step": 4818 + }, + { + "epoch": 0.616832, + "grad_norm": 0.4017888009548187, + "learning_rate": 0.0003380579623924959, + "loss": 2.5605, + "step": 4819 + }, + { + "epoch": 0.61696, + "grad_norm": 0.5005500912666321, + "learning_rate": 0.0003378618399397972, + "loss": 2.4968, + "step": 4820 + }, + { + "epoch": 0.617088, + "grad_norm": 0.39686450362205505, + "learning_rate": 0.0003376657453605281, + "loss": 2.6958, + "step": 4821 + }, + { + "epoch": 0.617216, + "grad_norm": 0.39940145611763, + "learning_rate": 0.00033746967868839917, + "loss": 2.6715, + "step": 4822 + }, + { + "epoch": 0.617344, + "grad_norm": 1.0125236511230469, + "learning_rate": 0.0003372736399571169, + "loss": 2.6198, + "step": 4823 + }, + { + "epoch": 0.617472, + "grad_norm": 0.4108070135116577, + "learning_rate": 0.0003370776292003823, + "loss": 2.6874, + "step": 4824 + }, + { + "epoch": 0.6176, + "grad_norm": 0.374907523393631, + "learning_rate": 0.00033688164645189194, + "loss": 2.639, + "step": 4825 + }, + { + "epoch": 0.617728, + "grad_norm": 0.4346195161342621, + "learning_rate": 0.0003366856917453377, + "loss": 2.6905, + "step": 4826 + }, + { + "epoch": 0.617856, + "grad_norm": 0.3731248080730438, + "learning_rate": 0.00033648976511440633, + "loss": 2.6096, + "step": 4827 + }, + { + "epoch": 0.617984, + "grad_norm": 0.8448394536972046, + "learning_rate": 0.00033629386659277984, + "loss": 2.4977, + "step": 4828 + }, + { + "epoch": 0.618112, + "grad_norm": 0.41050541400909424, + "learning_rate": 0.0003360979962141355, + "loss": 2.6163, + "step": 4829 + }, + { + "epoch": 0.61824, + "grad_norm": 0.4701104760169983, + "learning_rate": 0.0003359021540121457, + "loss": 2.6209, + "step": 4830 + }, + { + "epoch": 0.618368, + "grad_norm": 0.578869640827179, + "learning_rate": 0.0003357063400204781, + "loss": 2.5715, + "step": 4831 + }, + { + "epoch": 0.618496, + "grad_norm": 0.4304218888282776, + "learning_rate": 0.0003355105542727954, + "loss": 2.6809, + "step": 4832 + }, + { + "epoch": 0.618624, + "grad_norm": 1.8357881307601929, + "learning_rate": 0.00033531479680275515, + "loss": 2.5449, + "step": 4833 + }, + { + "epoch": 0.618752, + "grad_norm": 0.39309293031692505, + "learning_rate": 0.00033511906764401066, + "loss": 2.6146, + "step": 4834 + }, + { + "epoch": 0.61888, + "grad_norm": 0.542018711566925, + "learning_rate": 0.00033492336683020997, + "loss": 2.7063, + "step": 4835 + }, + { + "epoch": 0.619008, + "grad_norm": 0.3887154757976532, + "learning_rate": 0.0003347276943949961, + "loss": 2.6435, + "step": 4836 + }, + { + "epoch": 0.619136, + "grad_norm": 0.3701033294200897, + "learning_rate": 0.00033453205037200763, + "loss": 2.5665, + "step": 4837 + }, + { + "epoch": 0.619264, + "grad_norm": 0.40817591547966003, + "learning_rate": 0.00033433643479487775, + "loss": 2.7482, + "step": 4838 + }, + { + "epoch": 0.619392, + "grad_norm": 1.2336585521697998, + "learning_rate": 0.0003341408476972355, + "loss": 2.5379, + "step": 4839 + }, + { + "epoch": 0.61952, + "grad_norm": 0.4031967222690582, + "learning_rate": 0.00033394528911270425, + "loss": 2.5764, + "step": 4840 + }, + { + "epoch": 0.619648, + "grad_norm": 0.37733951210975647, + "learning_rate": 0.000333749759074903, + "loss": 2.5646, + "step": 4841 + }, + { + "epoch": 0.619776, + "grad_norm": 0.3798186779022217, + "learning_rate": 0.0003335542576174455, + "loss": 2.6042, + "step": 4842 + }, + { + "epoch": 0.619904, + "grad_norm": 0.4107207655906677, + "learning_rate": 0.00033335878477394056, + "loss": 2.6611, + "step": 4843 + }, + { + "epoch": 0.620032, + "grad_norm": 0.42099112272262573, + "learning_rate": 0.0003331633405779926, + "loss": 2.6297, + "step": 4844 + }, + { + "epoch": 0.62016, + "grad_norm": 0.4595227837562561, + "learning_rate": 0.0003329679250632004, + "loss": 2.5912, + "step": 4845 + }, + { + "epoch": 0.620288, + "grad_norm": 0.5064132809638977, + "learning_rate": 0.0003327725382631582, + "loss": 2.613, + "step": 4846 + }, + { + "epoch": 0.620416, + "grad_norm": 0.4422188401222229, + "learning_rate": 0.00033257718021145543, + "loss": 2.6214, + "step": 4847 + }, + { + "epoch": 0.620544, + "grad_norm": 0.5194345116615295, + "learning_rate": 0.00033238185094167625, + "loss": 2.6932, + "step": 4848 + }, + { + "epoch": 0.620672, + "grad_norm": 0.5011755228042603, + "learning_rate": 0.00033218655048740023, + "loss": 2.5626, + "step": 4849 + }, + { + "epoch": 0.6208, + "grad_norm": 0.43143072724342346, + "learning_rate": 0.0003319912788822016, + "loss": 2.5607, + "step": 4850 + }, + { + "epoch": 0.620928, + "grad_norm": 0.40711066126823425, + "learning_rate": 0.00033179603615964974, + "loss": 2.632, + "step": 4851 + }, + { + "epoch": 0.621056, + "grad_norm": 0.5010096430778503, + "learning_rate": 0.00033160082235330935, + "loss": 2.7054, + "step": 4852 + }, + { + "epoch": 0.621184, + "grad_norm": 0.45251160860061646, + "learning_rate": 0.0003314056374967399, + "loss": 2.6117, + "step": 4853 + }, + { + "epoch": 0.621312, + "grad_norm": 0.4231073558330536, + "learning_rate": 0.0003312104816234958, + "loss": 2.6969, + "step": 4854 + }, + { + "epoch": 0.62144, + "grad_norm": 0.41083824634552, + "learning_rate": 0.00033101535476712674, + "loss": 2.5903, + "step": 4855 + }, + { + "epoch": 0.621568, + "grad_norm": 0.4704928398132324, + "learning_rate": 0.0003308202569611772, + "loss": 2.5227, + "step": 4856 + }, + { + "epoch": 0.621696, + "grad_norm": 0.36865493655204773, + "learning_rate": 0.00033062518823918663, + "loss": 2.5806, + "step": 4857 + }, + { + "epoch": 0.621824, + "grad_norm": 0.45893728733062744, + "learning_rate": 0.00033043014863468983, + "loss": 2.6452, + "step": 4858 + }, + { + "epoch": 0.621952, + "grad_norm": 0.44199368357658386, + "learning_rate": 0.0003302351381812164, + "loss": 2.6185, + "step": 4859 + }, + { + "epoch": 0.62208, + "grad_norm": 0.39167502522468567, + "learning_rate": 0.0003300401569122908, + "loss": 2.5552, + "step": 4860 + }, + { + "epoch": 0.622208, + "grad_norm": 0.39865168929100037, + "learning_rate": 0.00032984520486143256, + "loss": 2.6254, + "step": 4861 + }, + { + "epoch": 0.622336, + "grad_norm": 0.39841228723526, + "learning_rate": 0.00032965028206215625, + "loss": 2.6026, + "step": 4862 + }, + { + "epoch": 0.622464, + "grad_norm": 0.4952615797519684, + "learning_rate": 0.0003294553885479713, + "loss": 2.8218, + "step": 4863 + }, + { + "epoch": 0.622592, + "grad_norm": 0.41880905628204346, + "learning_rate": 0.0003292605243523821, + "loss": 2.5744, + "step": 4864 + }, + { + "epoch": 0.62272, + "grad_norm": 0.36745545268058777, + "learning_rate": 0.00032906568950888827, + "loss": 2.5652, + "step": 4865 + }, + { + "epoch": 0.622848, + "grad_norm": 0.4446851909160614, + "learning_rate": 0.00032887088405098377, + "loss": 2.7008, + "step": 4866 + }, + { + "epoch": 0.622976, + "grad_norm": 0.3867964744567871, + "learning_rate": 0.0003286761080121585, + "loss": 2.5757, + "step": 4867 + }, + { + "epoch": 0.623104, + "grad_norm": 0.3892044425010681, + "learning_rate": 0.00032848136142589636, + "loss": 2.6433, + "step": 4868 + }, + { + "epoch": 0.623232, + "grad_norm": 0.5423216223716736, + "learning_rate": 0.00032828664432567653, + "loss": 2.6496, + "step": 4869 + }, + { + "epoch": 0.62336, + "grad_norm": 0.40245917439460754, + "learning_rate": 0.00032809195674497326, + "loss": 2.6534, + "step": 4870 + }, + { + "epoch": 0.623488, + "grad_norm": 0.42153117060661316, + "learning_rate": 0.0003278972987172556, + "loss": 2.6257, + "step": 4871 + }, + { + "epoch": 0.623616, + "grad_norm": 0.4553181231021881, + "learning_rate": 0.00032770267027598736, + "loss": 2.5769, + "step": 4872 + }, + { + "epoch": 0.623744, + "grad_norm": 0.4457576274871826, + "learning_rate": 0.0003275080714546277, + "loss": 2.6299, + "step": 4873 + }, + { + "epoch": 0.623872, + "grad_norm": 0.4185037910938263, + "learning_rate": 0.00032731350228663023, + "loss": 2.7774, + "step": 4874 + }, + { + "epoch": 0.624, + "grad_norm": 0.4105425775051117, + "learning_rate": 0.0003271189628054434, + "loss": 2.6302, + "step": 4875 + }, + { + "epoch": 0.624128, + "grad_norm": 0.4377095103263855, + "learning_rate": 0.0003269244530445112, + "loss": 2.6003, + "step": 4876 + }, + { + "epoch": 0.624256, + "grad_norm": 0.5406677722930908, + "learning_rate": 0.00032672997303727217, + "loss": 2.7313, + "step": 4877 + }, + { + "epoch": 0.624384, + "grad_norm": 0.4224395751953125, + "learning_rate": 0.0003265355228171594, + "loss": 2.663, + "step": 4878 + }, + { + "epoch": 0.624512, + "grad_norm": 0.41205114126205444, + "learning_rate": 0.00032634110241760116, + "loss": 2.6533, + "step": 4879 + }, + { + "epoch": 0.62464, + "grad_norm": 0.45046132802963257, + "learning_rate": 0.0003261467118720207, + "loss": 2.7648, + "step": 4880 + }, + { + "epoch": 0.624768, + "grad_norm": 0.40709933638572693, + "learning_rate": 0.000325952351213836, + "loss": 2.7692, + "step": 4881 + }, + { + "epoch": 0.624896, + "grad_norm": 0.4143389165401459, + "learning_rate": 0.00032575802047645977, + "loss": 2.6619, + "step": 4882 + }, + { + "epoch": 0.625024, + "grad_norm": 0.4210661053657532, + "learning_rate": 0.0003255637196932999, + "loss": 2.695, + "step": 4883 + }, + { + "epoch": 0.625152, + "grad_norm": 0.5516756772994995, + "learning_rate": 0.00032536944889775864, + "loss": 2.6457, + "step": 4884 + }, + { + "epoch": 0.62528, + "grad_norm": 0.3773483633995056, + "learning_rate": 0.0003251752081232338, + "loss": 2.7121, + "step": 4885 + }, + { + "epoch": 0.625408, + "grad_norm": 0.43465355038642883, + "learning_rate": 0.0003249809974031175, + "loss": 2.6806, + "step": 4886 + }, + { + "epoch": 0.625536, + "grad_norm": 0.42175793647766113, + "learning_rate": 0.00032478681677079654, + "loss": 2.5341, + "step": 4887 + }, + { + "epoch": 0.625664, + "grad_norm": 0.40694624185562134, + "learning_rate": 0.0003245926662596531, + "loss": 2.6754, + "step": 4888 + }, + { + "epoch": 0.625792, + "grad_norm": 0.6228442192077637, + "learning_rate": 0.00032439854590306396, + "loss": 2.6252, + "step": 4889 + }, + { + "epoch": 0.62592, + "grad_norm": 0.41557615995407104, + "learning_rate": 0.0003242044557344003, + "loss": 2.6222, + "step": 4890 + }, + { + "epoch": 0.626048, + "grad_norm": 0.4171470105648041, + "learning_rate": 0.00032401039578702887, + "loss": 2.6462, + "step": 4891 + }, + { + "epoch": 0.626176, + "grad_norm": 0.4258725643157959, + "learning_rate": 0.0003238163660943106, + "loss": 2.6004, + "step": 4892 + }, + { + "epoch": 0.626304, + "grad_norm": 0.42643052339553833, + "learning_rate": 0.00032362236668960145, + "loss": 2.6914, + "step": 4893 + }, + { + "epoch": 0.626432, + "grad_norm": 0.454049676656723, + "learning_rate": 0.0003234283976062521, + "loss": 2.7729, + "step": 4894 + }, + { + "epoch": 0.62656, + "grad_norm": 0.40321531891822815, + "learning_rate": 0.0003232344588776083, + "loss": 2.5944, + "step": 4895 + }, + { + "epoch": 0.626688, + "grad_norm": 0.4182411730289459, + "learning_rate": 0.00032304055053701037, + "loss": 2.6322, + "step": 4896 + }, + { + "epoch": 0.626816, + "grad_norm": 0.4161665141582489, + "learning_rate": 0.0003228466726177932, + "loss": 2.5842, + "step": 4897 + }, + { + "epoch": 0.626944, + "grad_norm": 0.4011986255645752, + "learning_rate": 0.0003226528251532868, + "loss": 2.6187, + "step": 4898 + }, + { + "epoch": 0.627072, + "grad_norm": 0.3755515515804291, + "learning_rate": 0.0003224590081768157, + "loss": 2.5978, + "step": 4899 + }, + { + "epoch": 0.6272, + "grad_norm": 0.3918296992778778, + "learning_rate": 0.00032226522172169923, + "loss": 2.6718, + "step": 4900 + }, + { + "epoch": 0.627328, + "grad_norm": 0.8004422187805176, + "learning_rate": 0.0003220714658212518, + "loss": 2.7459, + "step": 4901 + }, + { + "epoch": 0.627456, + "grad_norm": 0.7452102899551392, + "learning_rate": 0.0003218777405087821, + "loss": 2.6645, + "step": 4902 + }, + { + "epoch": 0.627584, + "grad_norm": 0.42995211482048035, + "learning_rate": 0.00032168404581759357, + "loss": 2.6978, + "step": 4903 + }, + { + "epoch": 0.627712, + "grad_norm": 0.4097798466682434, + "learning_rate": 0.000321490381780985, + "loss": 2.5478, + "step": 4904 + }, + { + "epoch": 0.62784, + "grad_norm": 0.4526125192642212, + "learning_rate": 0.0003212967484322491, + "loss": 2.539, + "step": 4905 + }, + { + "epoch": 0.627968, + "grad_norm": 0.4011882245540619, + "learning_rate": 0.00032110314580467404, + "loss": 2.6946, + "step": 4906 + }, + { + "epoch": 0.628096, + "grad_norm": 0.40289947390556335, + "learning_rate": 0.0003209095739315421, + "loss": 2.681, + "step": 4907 + }, + { + "epoch": 0.628224, + "grad_norm": 0.4128550589084625, + "learning_rate": 0.0003207160328461305, + "loss": 2.6561, + "step": 4908 + }, + { + "epoch": 0.628352, + "grad_norm": 0.4450598359107971, + "learning_rate": 0.00032052252258171136, + "loss": 2.5738, + "step": 4909 + }, + { + "epoch": 0.62848, + "grad_norm": 0.4333166480064392, + "learning_rate": 0.0003203290431715513, + "loss": 2.694, + "step": 4910 + }, + { + "epoch": 0.628608, + "grad_norm": 0.4178408086299896, + "learning_rate": 0.00032013559464891164, + "loss": 2.6444, + "step": 4911 + }, + { + "epoch": 0.628736, + "grad_norm": 0.4203293025493622, + "learning_rate": 0.0003199421770470482, + "loss": 2.6741, + "step": 4912 + }, + { + "epoch": 0.628864, + "grad_norm": 0.40221890807151794, + "learning_rate": 0.00031974879039921213, + "loss": 2.6156, + "step": 4913 + }, + { + "epoch": 0.628992, + "grad_norm": 0.39228835701942444, + "learning_rate": 0.0003195554347386487, + "loss": 2.5209, + "step": 4914 + }, + { + "epoch": 0.62912, + "grad_norm": 0.38258886337280273, + "learning_rate": 0.00031936211009859784, + "loss": 2.6185, + "step": 4915 + }, + { + "epoch": 0.629248, + "grad_norm": 0.4204352796077728, + "learning_rate": 0.00031916881651229445, + "loss": 2.5587, + "step": 4916 + }, + { + "epoch": 0.629376, + "grad_norm": 0.3788236081600189, + "learning_rate": 0.0003189755540129679, + "loss": 2.5596, + "step": 4917 + }, + { + "epoch": 0.629504, + "grad_norm": 0.39973577857017517, + "learning_rate": 0.0003187823226338421, + "loss": 2.5544, + "step": 4918 + }, + { + "epoch": 0.629632, + "grad_norm": 0.3895542323589325, + "learning_rate": 0.000318589122408136, + "loss": 2.6896, + "step": 4919 + }, + { + "epoch": 0.62976, + "grad_norm": 0.4679749310016632, + "learning_rate": 0.0003183959533690629, + "loss": 2.6635, + "step": 4920 + }, + { + "epoch": 0.629888, + "grad_norm": 0.3810689449310303, + "learning_rate": 0.00031820281554983054, + "loss": 2.6593, + "step": 4921 + }, + { + "epoch": 0.630016, + "grad_norm": 0.37772443890571594, + "learning_rate": 0.0003180097089836419, + "loss": 2.597, + "step": 4922 + }, + { + "epoch": 0.630144, + "grad_norm": 0.5288244485855103, + "learning_rate": 0.0003178166337036941, + "loss": 2.6628, + "step": 4923 + }, + { + "epoch": 0.630272, + "grad_norm": 0.38509735465049744, + "learning_rate": 0.00031762358974317905, + "loss": 2.5009, + "step": 4924 + }, + { + "epoch": 0.6304, + "grad_norm": 0.46172428131103516, + "learning_rate": 0.00031743057713528325, + "loss": 2.7019, + "step": 4925 + }, + { + "epoch": 0.630528, + "grad_norm": 0.5475366711616516, + "learning_rate": 0.00031723759591318755, + "loss": 2.5863, + "step": 4926 + }, + { + "epoch": 0.630656, + "grad_norm": 0.39572814106941223, + "learning_rate": 0.0003170446461100681, + "loss": 2.6767, + "step": 4927 + }, + { + "epoch": 0.630784, + "grad_norm": 0.39948973059654236, + "learning_rate": 0.000316851727759095, + "loss": 2.615, + "step": 4928 + }, + { + "epoch": 0.630912, + "grad_norm": 0.38225337862968445, + "learning_rate": 0.00031665884089343295, + "loss": 2.5995, + "step": 4929 + }, + { + "epoch": 0.63104, + "grad_norm": 0.3867644667625427, + "learning_rate": 0.0003164659855462417, + "loss": 2.5642, + "step": 4930 + }, + { + "epoch": 0.631168, + "grad_norm": 0.40918439626693726, + "learning_rate": 0.00031627316175067514, + "loss": 2.7359, + "step": 4931 + }, + { + "epoch": 0.631296, + "grad_norm": 0.37383684515953064, + "learning_rate": 0.0003160803695398822, + "loss": 2.6573, + "step": 4932 + }, + { + "epoch": 0.631424, + "grad_norm": 0.3943917751312256, + "learning_rate": 0.00031588760894700577, + "loss": 2.638, + "step": 4933 + }, + { + "epoch": 0.631552, + "grad_norm": 0.5955628752708435, + "learning_rate": 0.0003156948800051839, + "loss": 2.675, + "step": 4934 + }, + { + "epoch": 0.63168, + "grad_norm": 0.41756826639175415, + "learning_rate": 0.0003155021827475488, + "loss": 2.5879, + "step": 4935 + }, + { + "epoch": 0.631808, + "grad_norm": 0.40705496072769165, + "learning_rate": 0.0003153095172072273, + "loss": 2.6949, + "step": 4936 + }, + { + "epoch": 0.631936, + "grad_norm": 0.3750820755958557, + "learning_rate": 0.00031511688341734096, + "loss": 2.5547, + "step": 4937 + }, + { + "epoch": 0.632064, + "grad_norm": 0.3859703838825226, + "learning_rate": 0.0003149242814110057, + "loss": 2.7005, + "step": 4938 + }, + { + "epoch": 0.632192, + "grad_norm": 0.380958616733551, + "learning_rate": 0.00031473171122133205, + "loss": 2.5742, + "step": 4939 + }, + { + "epoch": 0.63232, + "grad_norm": 0.3730742931365967, + "learning_rate": 0.00031453917288142496, + "loss": 2.6024, + "step": 4940 + }, + { + "epoch": 0.632448, + "grad_norm": 0.4393983483314514, + "learning_rate": 0.0003143466664243841, + "loss": 2.672, + "step": 4941 + }, + { + "epoch": 0.632576, + "grad_norm": 0.3826356828212738, + "learning_rate": 0.0003141541918833037, + "loss": 2.6714, + "step": 4942 + }, + { + "epoch": 0.632704, + "grad_norm": 0.4304339587688446, + "learning_rate": 0.0003139617492912722, + "loss": 2.5505, + "step": 4943 + }, + { + "epoch": 0.632832, + "grad_norm": 0.500690758228302, + "learning_rate": 0.00031376933868137264, + "loss": 2.5795, + "step": 4944 + }, + { + "epoch": 0.63296, + "grad_norm": 0.5018519163131714, + "learning_rate": 0.0003135769600866828, + "loss": 2.6041, + "step": 4945 + }, + { + "epoch": 0.633088, + "grad_norm": 0.4204028844833374, + "learning_rate": 0.0003133846135402748, + "loss": 2.5454, + "step": 4946 + }, + { + "epoch": 0.633216, + "grad_norm": 0.40732109546661377, + "learning_rate": 0.000313192299075215, + "loss": 2.63, + "step": 4947 + }, + { + "epoch": 0.633344, + "grad_norm": 0.4451805353164673, + "learning_rate": 0.0003130000167245648, + "loss": 2.5808, + "step": 4948 + }, + { + "epoch": 0.633472, + "grad_norm": 0.6103656888008118, + "learning_rate": 0.00031280776652137937, + "loss": 2.6127, + "step": 4949 + }, + { + "epoch": 0.6336, + "grad_norm": 0.4511677920818329, + "learning_rate": 0.0003126155484987092, + "loss": 2.7016, + "step": 4950 + }, + { + "epoch": 0.633728, + "grad_norm": 0.4362303614616394, + "learning_rate": 0.0003124233626895985, + "loss": 2.6384, + "step": 4951 + }, + { + "epoch": 0.633856, + "grad_norm": 0.4364250898361206, + "learning_rate": 0.00031223120912708647, + "loss": 2.574, + "step": 4952 + }, + { + "epoch": 0.633984, + "grad_norm": 0.4641512930393219, + "learning_rate": 0.0003120390878442063, + "loss": 2.5504, + "step": 4953 + }, + { + "epoch": 0.634112, + "grad_norm": 0.4676347076892853, + "learning_rate": 0.0003118469988739859, + "loss": 2.5846, + "step": 4954 + }, + { + "epoch": 0.63424, + "grad_norm": 0.3820212781429291, + "learning_rate": 0.0003116549422494478, + "loss": 2.5937, + "step": 4955 + }, + { + "epoch": 0.634368, + "grad_norm": 0.4202719032764435, + "learning_rate": 0.0003114629180036086, + "loss": 2.5423, + "step": 4956 + }, + { + "epoch": 0.634496, + "grad_norm": 0.5103210806846619, + "learning_rate": 0.0003112709261694793, + "loss": 2.6185, + "step": 4957 + }, + { + "epoch": 0.634624, + "grad_norm": 0.39567282795906067, + "learning_rate": 0.00031107896678006573, + "loss": 2.6093, + "step": 4958 + }, + { + "epoch": 0.634752, + "grad_norm": 0.4440123736858368, + "learning_rate": 0.0003108870398683681, + "loss": 2.6896, + "step": 4959 + }, + { + "epoch": 0.63488, + "grad_norm": 0.3976078927516937, + "learning_rate": 0.00031069514546738076, + "loss": 2.631, + "step": 4960 + }, + { + "epoch": 0.635008, + "grad_norm": 0.43415728211402893, + "learning_rate": 0.00031050328361009245, + "loss": 2.7022, + "step": 4961 + }, + { + "epoch": 0.635136, + "grad_norm": 0.4059361517429352, + "learning_rate": 0.00031031145432948643, + "loss": 2.6575, + "step": 4962 + }, + { + "epoch": 0.635264, + "grad_norm": 0.3996849060058594, + "learning_rate": 0.00031011965765854056, + "loss": 2.5331, + "step": 4963 + }, + { + "epoch": 0.635392, + "grad_norm": 0.4555598497390747, + "learning_rate": 0.0003099278936302269, + "loss": 2.5035, + "step": 4964 + }, + { + "epoch": 0.63552, + "grad_norm": 0.41005003452301025, + "learning_rate": 0.00030973616227751165, + "loss": 2.6907, + "step": 4965 + }, + { + "epoch": 0.635648, + "grad_norm": 0.4296392798423767, + "learning_rate": 0.0003095444636333559, + "loss": 2.6214, + "step": 4966 + }, + { + "epoch": 0.635776, + "grad_norm": 0.3879089057445526, + "learning_rate": 0.0003093527977307149, + "loss": 2.5706, + "step": 4967 + }, + { + "epoch": 0.635904, + "grad_norm": 0.3884294033050537, + "learning_rate": 0.00030916116460253783, + "loss": 2.6929, + "step": 4968 + }, + { + "epoch": 0.636032, + "grad_norm": 0.4038540720939636, + "learning_rate": 0.0003089695642817691, + "loss": 2.7751, + "step": 4969 + }, + { + "epoch": 0.63616, + "grad_norm": 0.39935103058815, + "learning_rate": 0.0003087779968013471, + "loss": 2.5996, + "step": 4970 + }, + { + "epoch": 0.636288, + "grad_norm": 0.3980906009674072, + "learning_rate": 0.00030858646219420415, + "loss": 2.5885, + "step": 4971 + }, + { + "epoch": 0.636416, + "grad_norm": 0.37314289808273315, + "learning_rate": 0.0003083949604932674, + "loss": 2.6576, + "step": 4972 + }, + { + "epoch": 0.636544, + "grad_norm": 0.4349898397922516, + "learning_rate": 0.00030820349173145824, + "loss": 2.5694, + "step": 4973 + }, + { + "epoch": 0.636672, + "grad_norm": 0.3949562609195709, + "learning_rate": 0.0003080120559416924, + "loss": 2.6758, + "step": 4974 + }, + { + "epoch": 0.6368, + "grad_norm": 4.28139066696167, + "learning_rate": 0.00030782065315687975, + "loss": 2.5596, + "step": 4975 + }, + { + "epoch": 0.636928, + "grad_norm": 0.4528507590293884, + "learning_rate": 0.0003076292834099248, + "loss": 2.6574, + "step": 4976 + }, + { + "epoch": 0.637056, + "grad_norm": 0.4297684133052826, + "learning_rate": 0.00030743794673372606, + "loss": 2.6513, + "step": 4977 + }, + { + "epoch": 0.637184, + "grad_norm": 0.3860400319099426, + "learning_rate": 0.0003072466431611768, + "loss": 2.6073, + "step": 4978 + }, + { + "epoch": 0.637312, + "grad_norm": 0.41148290038108826, + "learning_rate": 0.000307055372725164, + "loss": 2.5342, + "step": 4979 + }, + { + "epoch": 0.63744, + "grad_norm": 0.39482444524765015, + "learning_rate": 0.0003068641354585697, + "loss": 2.7681, + "step": 4980 + }, + { + "epoch": 0.637568, + "grad_norm": 0.4294346570968628, + "learning_rate": 0.00030667293139426946, + "loss": 2.6145, + "step": 4981 + }, + { + "epoch": 0.637696, + "grad_norm": 0.4673868417739868, + "learning_rate": 0.00030648176056513355, + "loss": 2.6874, + "step": 4982 + }, + { + "epoch": 0.637824, + "grad_norm": 0.43259063363075256, + "learning_rate": 0.0003062906230040263, + "loss": 2.646, + "step": 4983 + }, + { + "epoch": 0.637952, + "grad_norm": 0.446490615606308, + "learning_rate": 0.0003060995187438068, + "loss": 2.6658, + "step": 4984 + }, + { + "epoch": 0.63808, + "grad_norm": 0.4297359585762024, + "learning_rate": 0.00030590844781732795, + "loss": 2.5836, + "step": 4985 + }, + { + "epoch": 0.638208, + "grad_norm": 0.43281710147857666, + "learning_rate": 0.0003057174102574367, + "loss": 2.6636, + "step": 4986 + }, + { + "epoch": 0.638336, + "grad_norm": 0.40517011284828186, + "learning_rate": 0.00030552640609697514, + "loss": 2.6055, + "step": 4987 + }, + { + "epoch": 0.638464, + "grad_norm": 0.5939275622367859, + "learning_rate": 0.00030533543536877895, + "loss": 2.6809, + "step": 4988 + }, + { + "epoch": 0.638592, + "grad_norm": 0.4347609579563141, + "learning_rate": 0.00030514449810567826, + "loss": 2.5607, + "step": 4989 + }, + { + "epoch": 0.63872, + "grad_norm": 0.4844886362552643, + "learning_rate": 0.00030495359434049707, + "loss": 2.6211, + "step": 4990 + }, + { + "epoch": 0.638848, + "grad_norm": 0.49755093455314636, + "learning_rate": 0.0003047627241060544, + "loss": 2.6232, + "step": 4991 + }, + { + "epoch": 0.638976, + "grad_norm": 0.40589287877082825, + "learning_rate": 0.00030457188743516274, + "loss": 2.5958, + "step": 4992 + }, + { + "epoch": 0.639104, + "grad_norm": 0.44368693232536316, + "learning_rate": 0.00030438108436062916, + "loss": 2.6143, + "step": 4993 + }, + { + "epoch": 0.639232, + "grad_norm": 0.42409124970436096, + "learning_rate": 0.0003041903149152552, + "loss": 2.6067, + "step": 4994 + }, + { + "epoch": 0.63936, + "grad_norm": 0.4271189570426941, + "learning_rate": 0.0003039995791318358, + "loss": 2.5482, + "step": 4995 + }, + { + "epoch": 0.639488, + "grad_norm": 0.3936440348625183, + "learning_rate": 0.00030380887704316116, + "loss": 2.5719, + "step": 4996 + }, + { + "epoch": 0.639616, + "grad_norm": 0.4662407338619232, + "learning_rate": 0.00030361820868201494, + "loss": 2.6598, + "step": 4997 + }, + { + "epoch": 0.639744, + "grad_norm": 0.4693392515182495, + "learning_rate": 0.00030342757408117544, + "loss": 2.5765, + "step": 4998 + }, + { + "epoch": 0.639872, + "grad_norm": 0.5223625898361206, + "learning_rate": 0.0003032369732734148, + "loss": 2.5784, + "step": 4999 + }, + { + "epoch": 0.64, + "grad_norm": 0.47401267290115356, + "learning_rate": 0.0003030464062914995, + "loss": 2.5827, + "step": 5000 + }, + { + "epoch": 0.640128, + "grad_norm": 0.4315483868122101, + "learning_rate": 0.0003028558731681902, + "loss": 2.6034, + "step": 5001 + }, + { + "epoch": 0.640256, + "grad_norm": 0.45059508085250854, + "learning_rate": 0.0003026653739362418, + "loss": 2.5953, + "step": 5002 + }, + { + "epoch": 0.640384, + "grad_norm": 0.38210222125053406, + "learning_rate": 0.00030247490862840337, + "loss": 2.6777, + "step": 5003 + }, + { + "epoch": 0.640512, + "grad_norm": 0.42323222756385803, + "learning_rate": 0.000302284477277418, + "loss": 2.6065, + "step": 5004 + }, + { + "epoch": 0.64064, + "grad_norm": 0.5485125184059143, + "learning_rate": 0.0003020940799160229, + "loss": 2.5955, + "step": 5005 + }, + { + "epoch": 0.640768, + "grad_norm": 0.4442387521266937, + "learning_rate": 0.00030190371657695, + "loss": 2.6937, + "step": 5006 + }, + { + "epoch": 0.640896, + "grad_norm": 0.42173466086387634, + "learning_rate": 0.00030171338729292484, + "loss": 2.6014, + "step": 5007 + }, + { + "epoch": 0.641024, + "grad_norm": 0.47701504826545715, + "learning_rate": 0.000301523092096667, + "loss": 2.5496, + "step": 5008 + }, + { + "epoch": 0.641152, + "grad_norm": 0.4222245514392853, + "learning_rate": 0.00030133283102089063, + "loss": 2.6297, + "step": 5009 + }, + { + "epoch": 0.64128, + "grad_norm": 0.41402706503868103, + "learning_rate": 0.00030114260409830384, + "loss": 2.7182, + "step": 5010 + }, + { + "epoch": 0.641408, + "grad_norm": 0.42676910758018494, + "learning_rate": 0.00030095241136160857, + "loss": 2.5563, + "step": 5011 + }, + { + "epoch": 0.641536, + "grad_norm": 0.5077771544456482, + "learning_rate": 0.00030076225284350144, + "loss": 2.5425, + "step": 5012 + }, + { + "epoch": 0.641664, + "grad_norm": 0.4178498387336731, + "learning_rate": 0.0003005721285766729, + "loss": 2.6804, + "step": 5013 + }, + { + "epoch": 0.641792, + "grad_norm": 0.4084574580192566, + "learning_rate": 0.00030038203859380716, + "loss": 2.645, + "step": 5014 + }, + { + "epoch": 0.64192, + "grad_norm": 0.40574634075164795, + "learning_rate": 0.00030019198292758324, + "loss": 2.6849, + "step": 5015 + }, + { + "epoch": 0.642048, + "grad_norm": 0.448537677526474, + "learning_rate": 0.00030000196161067394, + "loss": 2.6555, + "step": 5016 + }, + { + "epoch": 0.642176, + "grad_norm": 0.42789772152900696, + "learning_rate": 0.00029981197467574607, + "loss": 2.6032, + "step": 5017 + }, + { + "epoch": 0.642304, + "grad_norm": 0.41724637150764465, + "learning_rate": 0.0002996220221554604, + "loss": 2.5181, + "step": 5018 + }, + { + "epoch": 0.642432, + "grad_norm": 0.4035194516181946, + "learning_rate": 0.00029943210408247216, + "loss": 2.6373, + "step": 5019 + }, + { + "epoch": 0.64256, + "grad_norm": 0.5747184157371521, + "learning_rate": 0.00029924222048943044, + "loss": 2.6643, + "step": 5020 + }, + { + "epoch": 0.642688, + "grad_norm": 0.4227426052093506, + "learning_rate": 0.00029905237140897845, + "loss": 2.6318, + "step": 5021 + }, + { + "epoch": 0.642816, + "grad_norm": 0.4272056221961975, + "learning_rate": 0.00029886255687375316, + "loss": 2.6083, + "step": 5022 + }, + { + "epoch": 0.642944, + "grad_norm": 0.482057124376297, + "learning_rate": 0.0002986727769163862, + "loss": 2.7092, + "step": 5023 + }, + { + "epoch": 0.643072, + "grad_norm": 0.450277179479599, + "learning_rate": 0.000298483031569503, + "loss": 2.6281, + "step": 5024 + }, + { + "epoch": 0.6432, + "grad_norm": 0.4113381505012512, + "learning_rate": 0.0002982933208657228, + "loss": 2.6277, + "step": 5025 + }, + { + "epoch": 0.643328, + "grad_norm": 0.42849254608154297, + "learning_rate": 0.0002981036448376591, + "loss": 2.6184, + "step": 5026 + }, + { + "epoch": 0.643456, + "grad_norm": 0.386930376291275, + "learning_rate": 0.00029791400351791955, + "loss": 2.611, + "step": 5027 + }, + { + "epoch": 0.643584, + "grad_norm": 0.40451061725616455, + "learning_rate": 0.0002977243969391055, + "loss": 2.549, + "step": 5028 + }, + { + "epoch": 0.643712, + "grad_norm": 0.38628438115119934, + "learning_rate": 0.0002975348251338126, + "loss": 2.6968, + "step": 5029 + }, + { + "epoch": 0.64384, + "grad_norm": 0.42121440172195435, + "learning_rate": 0.0002973452881346305, + "loss": 2.708, + "step": 5030 + }, + { + "epoch": 0.643968, + "grad_norm": 0.43240809440612793, + "learning_rate": 0.00029715578597414266, + "loss": 2.5893, + "step": 5031 + }, + { + "epoch": 0.644096, + "grad_norm": 0.6236486434936523, + "learning_rate": 0.0002969663186849266, + "loss": 2.4476, + "step": 5032 + }, + { + "epoch": 0.644224, + "grad_norm": 0.4210284650325775, + "learning_rate": 0.00029677688629955437, + "loss": 2.7788, + "step": 5033 + }, + { + "epoch": 0.644352, + "grad_norm": 0.4072295129299164, + "learning_rate": 0.00029658748885059126, + "loss": 2.6876, + "step": 5034 + }, + { + "epoch": 0.64448, + "grad_norm": 0.40987884998321533, + "learning_rate": 0.00029639812637059716, + "loss": 2.6977, + "step": 5035 + }, + { + "epoch": 0.644608, + "grad_norm": 0.40888941287994385, + "learning_rate": 0.00029620879889212526, + "loss": 2.5428, + "step": 5036 + }, + { + "epoch": 0.644736, + "grad_norm": 0.43272513151168823, + "learning_rate": 0.0002960195064477235, + "loss": 2.6078, + "step": 5037 + }, + { + "epoch": 0.644864, + "grad_norm": 0.4326850473880768, + "learning_rate": 0.00029583024906993337, + "loss": 2.6475, + "step": 5038 + }, + { + "epoch": 0.644992, + "grad_norm": 0.40309759974479675, + "learning_rate": 0.00029564102679129024, + "loss": 2.7338, + "step": 5039 + }, + { + "epoch": 0.64512, + "grad_norm": 0.3874347507953644, + "learning_rate": 0.00029545183964432386, + "loss": 2.6001, + "step": 5040 + }, + { + "epoch": 0.645248, + "grad_norm": 0.5488985776901245, + "learning_rate": 0.0002952626876615575, + "loss": 2.6558, + "step": 5041 + }, + { + "epoch": 0.645376, + "grad_norm": 0.4920961856842041, + "learning_rate": 0.00029507357087550857, + "loss": 2.6141, + "step": 5042 + }, + { + "epoch": 0.645504, + "grad_norm": 0.4118904173374176, + "learning_rate": 0.0002948844893186886, + "loss": 2.7338, + "step": 5043 + }, + { + "epoch": 0.645632, + "grad_norm": 0.4534509479999542, + "learning_rate": 0.0002946954430236028, + "loss": 2.6286, + "step": 5044 + }, + { + "epoch": 0.64576, + "grad_norm": 0.3804827332496643, + "learning_rate": 0.0002945064320227505, + "loss": 2.6279, + "step": 5045 + }, + { + "epoch": 0.645888, + "grad_norm": 0.4253818094730377, + "learning_rate": 0.00029431745634862484, + "loss": 2.6082, + "step": 5046 + }, + { + "epoch": 0.646016, + "grad_norm": 0.4196168780326843, + "learning_rate": 0.0002941285160337128, + "loss": 2.6721, + "step": 5047 + }, + { + "epoch": 0.646144, + "grad_norm": 0.4151251018047333, + "learning_rate": 0.0002939396111104956, + "loss": 2.6741, + "step": 5048 + }, + { + "epoch": 0.646272, + "grad_norm": 0.3826475441455841, + "learning_rate": 0.0002937507416114481, + "loss": 2.5703, + "step": 5049 + }, + { + "epoch": 0.6464, + "grad_norm": 0.41563984751701355, + "learning_rate": 0.0002935619075690391, + "loss": 2.7098, + "step": 5050 + }, + { + "epoch": 0.646528, + "grad_norm": 0.5672480463981628, + "learning_rate": 0.0002933731090157313, + "loss": 2.6303, + "step": 5051 + }, + { + "epoch": 0.646656, + "grad_norm": 0.379607230424881, + "learning_rate": 0.0002931843459839817, + "loss": 2.6309, + "step": 5052 + }, + { + "epoch": 0.646784, + "grad_norm": 0.39004719257354736, + "learning_rate": 0.0002929956185062407, + "loss": 2.6825, + "step": 5053 + }, + { + "epoch": 0.646912, + "grad_norm": 0.4022815525531769, + "learning_rate": 0.00029280692661495247, + "loss": 2.679, + "step": 5054 + }, + { + "epoch": 0.64704, + "grad_norm": 0.4305501878261566, + "learning_rate": 0.0002926182703425557, + "loss": 2.6148, + "step": 5055 + }, + { + "epoch": 0.647168, + "grad_norm": 0.3660087585449219, + "learning_rate": 0.00029242964972148243, + "loss": 2.492, + "step": 5056 + }, + { + "epoch": 0.647296, + "grad_norm": 0.42564159631729126, + "learning_rate": 0.00029224106478415856, + "loss": 2.599, + "step": 5057 + }, + { + "epoch": 0.647424, + "grad_norm": 0.4006332457065582, + "learning_rate": 0.0002920525155630044, + "loss": 2.5003, + "step": 5058 + }, + { + "epoch": 0.647552, + "grad_norm": 0.4425037205219269, + "learning_rate": 0.0002918640020904334, + "loss": 2.5651, + "step": 5059 + }, + { + "epoch": 0.64768, + "grad_norm": 0.449706107378006, + "learning_rate": 0.0002916755243988533, + "loss": 2.6219, + "step": 5060 + }, + { + "epoch": 0.647808, + "grad_norm": 0.4204936921596527, + "learning_rate": 0.00029148708252066586, + "loss": 2.5626, + "step": 5061 + }, + { + "epoch": 0.647936, + "grad_norm": 0.40150532126426697, + "learning_rate": 0.0002912986764882662, + "loss": 2.5665, + "step": 5062 + }, + { + "epoch": 0.648064, + "grad_norm": 0.44313934445381165, + "learning_rate": 0.0002911103063340435, + "loss": 2.5956, + "step": 5063 + }, + { + "epoch": 0.648192, + "grad_norm": 0.41141214966773987, + "learning_rate": 0.0002909219720903808, + "loss": 2.6692, + "step": 5064 + }, + { + "epoch": 0.64832, + "grad_norm": 0.3871181607246399, + "learning_rate": 0.000290733673789655, + "loss": 2.5213, + "step": 5065 + }, + { + "epoch": 0.648448, + "grad_norm": 0.42137569189071655, + "learning_rate": 0.00029054541146423645, + "loss": 2.626, + "step": 5066 + }, + { + "epoch": 0.648576, + "grad_norm": 0.3602226972579956, + "learning_rate": 0.00029035718514649014, + "loss": 2.5379, + "step": 5067 + }, + { + "epoch": 0.648704, + "grad_norm": 0.47417470812797546, + "learning_rate": 0.000290168994868774, + "loss": 2.5638, + "step": 5068 + }, + { + "epoch": 0.648832, + "grad_norm": 0.3757677674293518, + "learning_rate": 0.00028998084066344, + "loss": 2.5915, + "step": 5069 + }, + { + "epoch": 0.64896, + "grad_norm": 0.3889467716217041, + "learning_rate": 0.0002897927225628345, + "loss": 2.6135, + "step": 5070 + }, + { + "epoch": 0.649088, + "grad_norm": 0.40808436274528503, + "learning_rate": 0.00028960464059929697, + "loss": 2.639, + "step": 5071 + }, + { + "epoch": 0.649216, + "grad_norm": 0.37872031331062317, + "learning_rate": 0.0002894165948051608, + "loss": 2.5787, + "step": 5072 + }, + { + "epoch": 0.649344, + "grad_norm": 0.4174529016017914, + "learning_rate": 0.0002892285852127532, + "loss": 2.5947, + "step": 5073 + }, + { + "epoch": 0.649472, + "grad_norm": 0.3979347348213196, + "learning_rate": 0.0002890406118543951, + "loss": 2.5495, + "step": 5074 + }, + { + "epoch": 0.6496, + "grad_norm": 0.4040169417858124, + "learning_rate": 0.00028885267476240175, + "loss": 2.6428, + "step": 5075 + }, + { + "epoch": 0.649728, + "grad_norm": 0.41675323247909546, + "learning_rate": 0.0002886647739690814, + "loss": 2.6794, + "step": 5076 + }, + { + "epoch": 0.649856, + "grad_norm": 0.4573207497596741, + "learning_rate": 0.0002884769095067364, + "loss": 2.6404, + "step": 5077 + }, + { + "epoch": 0.649984, + "grad_norm": 0.37298938632011414, + "learning_rate": 0.00028828908140766284, + "loss": 2.5633, + "step": 5078 + }, + { + "epoch": 0.650112, + "grad_norm": 0.38577982783317566, + "learning_rate": 0.00028810128970415034, + "loss": 2.4803, + "step": 5079 + }, + { + "epoch": 0.65024, + "grad_norm": 0.4402933418750763, + "learning_rate": 0.0002879135344284829, + "loss": 2.6942, + "step": 5080 + }, + { + "epoch": 0.650368, + "grad_norm": 0.4197927713394165, + "learning_rate": 0.00028772581561293745, + "loss": 2.6836, + "step": 5081 + }, + { + "epoch": 0.650496, + "grad_norm": 0.41522523760795593, + "learning_rate": 0.0002875381332897853, + "loss": 2.6785, + "step": 5082 + }, + { + "epoch": 0.650624, + "grad_norm": 0.4105958342552185, + "learning_rate": 0.0002873504874912911, + "loss": 2.5484, + "step": 5083 + }, + { + "epoch": 0.650752, + "grad_norm": 0.4315531849861145, + "learning_rate": 0.00028716287824971345, + "loss": 2.6397, + "step": 5084 + }, + { + "epoch": 0.65088, + "grad_norm": 0.43016037344932556, + "learning_rate": 0.0002869753055973044, + "loss": 2.617, + "step": 5085 + }, + { + "epoch": 0.651008, + "grad_norm": 0.5828379988670349, + "learning_rate": 0.0002867877695663099, + "loss": 2.5844, + "step": 5086 + }, + { + "epoch": 0.651136, + "grad_norm": 0.3809945285320282, + "learning_rate": 0.00028660027018896937, + "loss": 2.6815, + "step": 5087 + }, + { + "epoch": 0.651264, + "grad_norm": 0.39329859614372253, + "learning_rate": 0.00028641280749751655, + "loss": 2.5988, + "step": 5088 + }, + { + "epoch": 0.651392, + "grad_norm": 0.3837498128414154, + "learning_rate": 0.00028622538152417806, + "loss": 2.6942, + "step": 5089 + }, + { + "epoch": 0.65152, + "grad_norm": 0.4459095299243927, + "learning_rate": 0.00028603799230117493, + "loss": 2.6057, + "step": 5090 + }, + { + "epoch": 0.651648, + "grad_norm": 0.37754112482070923, + "learning_rate": 0.0002858506398607214, + "loss": 2.6985, + "step": 5091 + }, + { + "epoch": 0.651776, + "grad_norm": 0.39762479066848755, + "learning_rate": 0.0002856633242350255, + "loss": 2.6106, + "step": 5092 + }, + { + "epoch": 0.651904, + "grad_norm": 0.41446158289909363, + "learning_rate": 0.000285476045456289, + "loss": 2.4891, + "step": 5093 + }, + { + "epoch": 0.652032, + "grad_norm": 0.4107595682144165, + "learning_rate": 0.00028528880355670725, + "loss": 2.5566, + "step": 5094 + }, + { + "epoch": 0.65216, + "grad_norm": 0.3998103439807892, + "learning_rate": 0.0002851015985684691, + "loss": 2.6403, + "step": 5095 + }, + { + "epoch": 0.652288, + "grad_norm": 0.4067171514034271, + "learning_rate": 0.00028491443052375764, + "loss": 2.7062, + "step": 5096 + }, + { + "epoch": 0.652416, + "grad_norm": 0.3974125385284424, + "learning_rate": 0.00028472729945474895, + "loss": 2.6391, + "step": 5097 + }, + { + "epoch": 0.652544, + "grad_norm": 0.3943491280078888, + "learning_rate": 0.00028454020539361326, + "loss": 2.6204, + "step": 5098 + }, + { + "epoch": 0.652672, + "grad_norm": 0.3912612199783325, + "learning_rate": 0.00028435314837251414, + "loss": 2.5576, + "step": 5099 + }, + { + "epoch": 0.6528, + "grad_norm": 0.4222390353679657, + "learning_rate": 0.0002841661284236088, + "loss": 2.6349, + "step": 5100 + }, + { + "epoch": 0.652928, + "grad_norm": 0.4083568751811981, + "learning_rate": 0.0002839791455790481, + "loss": 2.6292, + "step": 5101 + }, + { + "epoch": 0.653056, + "grad_norm": 0.4341954290866852, + "learning_rate": 0.0002837921998709765, + "loss": 2.5965, + "step": 5102 + }, + { + "epoch": 0.653184, + "grad_norm": 0.38992512226104736, + "learning_rate": 0.00028360529133153234, + "loss": 2.611, + "step": 5103 + }, + { + "epoch": 0.653312, + "grad_norm": 0.4117458462715149, + "learning_rate": 0.0002834184199928472, + "loss": 2.6477, + "step": 5104 + }, + { + "epoch": 0.65344, + "grad_norm": 0.45699983835220337, + "learning_rate": 0.00028323158588704667, + "loss": 2.6309, + "step": 5105 + }, + { + "epoch": 0.653568, + "grad_norm": 0.4621064364910126, + "learning_rate": 0.0002830447890462492, + "loss": 2.658, + "step": 5106 + }, + { + "epoch": 0.653696, + "grad_norm": 0.38909244537353516, + "learning_rate": 0.0002828580295025679, + "loss": 2.5896, + "step": 5107 + }, + { + "epoch": 0.653824, + "grad_norm": 0.38079971075057983, + "learning_rate": 0.00028267130728810866, + "loss": 2.6219, + "step": 5108 + }, + { + "epoch": 0.653952, + "grad_norm": 0.37485018372535706, + "learning_rate": 0.0002824846224349712, + "loss": 2.5816, + "step": 5109 + }, + { + "epoch": 0.65408, + "grad_norm": 0.3728744089603424, + "learning_rate": 0.0002822979749752486, + "loss": 2.6701, + "step": 5110 + }, + { + "epoch": 0.654208, + "grad_norm": 0.4004908800125122, + "learning_rate": 0.00028211136494102806, + "loss": 2.6333, + "step": 5111 + }, + { + "epoch": 0.654336, + "grad_norm": 0.42415234446525574, + "learning_rate": 0.00028192479236438983, + "loss": 2.6949, + "step": 5112 + }, + { + "epoch": 0.654464, + "grad_norm": 0.4085251986980438, + "learning_rate": 0.00028173825727740806, + "loss": 2.5302, + "step": 5113 + }, + { + "epoch": 0.654592, + "grad_norm": 0.3850867450237274, + "learning_rate": 0.00028155175971215007, + "loss": 2.6597, + "step": 5114 + }, + { + "epoch": 0.65472, + "grad_norm": 0.4112549126148224, + "learning_rate": 0.00028136529970067714, + "loss": 2.6549, + "step": 5115 + }, + { + "epoch": 0.654848, + "grad_norm": 0.3911561965942383, + "learning_rate": 0.00028117887727504355, + "loss": 2.5147, + "step": 5116 + }, + { + "epoch": 0.654976, + "grad_norm": 0.416885107755661, + "learning_rate": 0.0002809924924672978, + "loss": 2.5938, + "step": 5117 + }, + { + "epoch": 0.655104, + "grad_norm": 0.4035748541355133, + "learning_rate": 0.0002808061453094817, + "loss": 2.5379, + "step": 5118 + }, + { + "epoch": 0.655232, + "grad_norm": 0.3958868086338043, + "learning_rate": 0.00028061983583363035, + "loss": 2.6559, + "step": 5119 + }, + { + "epoch": 0.65536, + "grad_norm": 0.5719947218894958, + "learning_rate": 0.00028043356407177254, + "loss": 2.6075, + "step": 5120 + }, + { + "epoch": 0.655488, + "grad_norm": 0.4238130748271942, + "learning_rate": 0.0002802473300559305, + "loss": 2.5255, + "step": 5121 + }, + { + "epoch": 0.655616, + "grad_norm": 0.38510483503341675, + "learning_rate": 0.00028006113381812, + "loss": 2.5425, + "step": 5122 + }, + { + "epoch": 0.655744, + "grad_norm": 0.40483731031417847, + "learning_rate": 0.00027987497539035014, + "loss": 2.746, + "step": 5123 + }, + { + "epoch": 0.655872, + "grad_norm": 0.38529467582702637, + "learning_rate": 0.00027968885480462405, + "loss": 2.6, + "step": 5124 + }, + { + "epoch": 0.656, + "grad_norm": 0.47398319840431213, + "learning_rate": 0.00027950277209293776, + "loss": 2.729, + "step": 5125 + }, + { + "epoch": 0.656128, + "grad_norm": 0.3863236904144287, + "learning_rate": 0.00027931672728728134, + "loss": 2.5424, + "step": 5126 + }, + { + "epoch": 0.656256, + "grad_norm": 0.4463825821876526, + "learning_rate": 0.00027913072041963795, + "loss": 2.5598, + "step": 5127 + }, + { + "epoch": 0.656384, + "grad_norm": 0.3788616359233856, + "learning_rate": 0.0002789447515219842, + "loss": 2.7591, + "step": 5128 + }, + { + "epoch": 0.656512, + "grad_norm": 0.4006454050540924, + "learning_rate": 0.0002787588206262903, + "loss": 2.6129, + "step": 5129 + }, + { + "epoch": 0.65664, + "grad_norm": 0.38037431240081787, + "learning_rate": 0.00027857292776452004, + "loss": 2.5871, + "step": 5130 + }, + { + "epoch": 0.656768, + "grad_norm": 0.40153583884239197, + "learning_rate": 0.00027838707296863033, + "loss": 2.6062, + "step": 5131 + }, + { + "epoch": 0.656896, + "grad_norm": 0.3690142035484314, + "learning_rate": 0.0002782012562705721, + "loss": 2.5944, + "step": 5132 + }, + { + "epoch": 0.657024, + "grad_norm": 0.39835840463638306, + "learning_rate": 0.0002780154777022892, + "loss": 2.6983, + "step": 5133 + }, + { + "epoch": 0.657152, + "grad_norm": 0.38105812668800354, + "learning_rate": 0.000277829737295719, + "loss": 2.6381, + "step": 5134 + }, + { + "epoch": 0.65728, + "grad_norm": 0.40341728925704956, + "learning_rate": 0.00027764403508279275, + "loss": 2.6102, + "step": 5135 + }, + { + "epoch": 0.657408, + "grad_norm": 0.3882504403591156, + "learning_rate": 0.0002774583710954346, + "loss": 2.5321, + "step": 5136 + }, + { + "epoch": 0.657536, + "grad_norm": 0.391020268201828, + "learning_rate": 0.00027727274536556247, + "loss": 2.5593, + "step": 5137 + }, + { + "epoch": 0.657664, + "grad_norm": 0.45479056239128113, + "learning_rate": 0.0002770871579250872, + "loss": 2.6263, + "step": 5138 + }, + { + "epoch": 0.657792, + "grad_norm": 0.35670095682144165, + "learning_rate": 0.00027690160880591396, + "loss": 2.6208, + "step": 5139 + }, + { + "epoch": 0.65792, + "grad_norm": 0.42366430163383484, + "learning_rate": 0.00027671609803994057, + "loss": 2.7739, + "step": 5140 + }, + { + "epoch": 0.658048, + "grad_norm": 0.44365161657333374, + "learning_rate": 0.0002765306256590584, + "loss": 2.5137, + "step": 5141 + }, + { + "epoch": 0.658176, + "grad_norm": 0.38311609625816345, + "learning_rate": 0.00027634519169515237, + "loss": 2.6241, + "step": 5142 + }, + { + "epoch": 0.658304, + "grad_norm": 0.3804953694343567, + "learning_rate": 0.0002761597961801006, + "loss": 2.6549, + "step": 5143 + }, + { + "epoch": 0.658432, + "grad_norm": 0.3725756108760834, + "learning_rate": 0.00027597443914577493, + "loss": 2.6595, + "step": 5144 + }, + { + "epoch": 0.65856, + "grad_norm": 0.423648476600647, + "learning_rate": 0.0002757891206240404, + "loss": 2.6438, + "step": 5145 + }, + { + "epoch": 0.658688, + "grad_norm": 0.420483261346817, + "learning_rate": 0.00027560384064675503, + "loss": 2.5788, + "step": 5146 + }, + { + "epoch": 0.658816, + "grad_norm": 0.3768726885318756, + "learning_rate": 0.00027541859924577104, + "loss": 2.6301, + "step": 5147 + }, + { + "epoch": 0.658944, + "grad_norm": 0.45112189650535583, + "learning_rate": 0.0002752333964529335, + "loss": 2.5257, + "step": 5148 + }, + { + "epoch": 0.659072, + "grad_norm": 0.39266690611839294, + "learning_rate": 0.0002750482323000808, + "loss": 2.5477, + "step": 5149 + }, + { + "epoch": 0.6592, + "grad_norm": 0.39556869864463806, + "learning_rate": 0.00027486310681904485, + "loss": 2.5823, + "step": 5150 + }, + { + "epoch": 0.659328, + "grad_norm": 0.3874141275882721, + "learning_rate": 0.0002746780200416509, + "loss": 2.6054, + "step": 5151 + }, + { + "epoch": 0.659456, + "grad_norm": 0.4181019961833954, + "learning_rate": 0.00027449297199971714, + "loss": 2.6499, + "step": 5152 + }, + { + "epoch": 0.659584, + "grad_norm": 0.3915250301361084, + "learning_rate": 0.00027430796272505587, + "loss": 2.6574, + "step": 5153 + }, + { + "epoch": 0.659712, + "grad_norm": 0.37995296716690063, + "learning_rate": 0.0002741229922494724, + "loss": 2.6775, + "step": 5154 + }, + { + "epoch": 0.65984, + "grad_norm": 0.41828039288520813, + "learning_rate": 0.00027393806060476523, + "loss": 2.58, + "step": 5155 + }, + { + "epoch": 0.659968, + "grad_norm": 0.5012006163597107, + "learning_rate": 0.0002737531678227261, + "loss": 2.7157, + "step": 5156 + }, + { + "epoch": 0.660096, + "grad_norm": 0.3837071359157562, + "learning_rate": 0.00027356831393514033, + "loss": 2.5449, + "step": 5157 + }, + { + "epoch": 0.660224, + "grad_norm": 0.4453413188457489, + "learning_rate": 0.00027338349897378637, + "loss": 2.5264, + "step": 5158 + }, + { + "epoch": 0.660352, + "grad_norm": 0.40073221921920776, + "learning_rate": 0.000273198722970436, + "loss": 2.6835, + "step": 5159 + }, + { + "epoch": 0.66048, + "grad_norm": 0.4121040105819702, + "learning_rate": 0.0002730139859568546, + "loss": 2.5614, + "step": 5160 + }, + { + "epoch": 0.660608, + "grad_norm": 0.4655535817146301, + "learning_rate": 0.00027282928796480043, + "loss": 2.6412, + "step": 5161 + }, + { + "epoch": 0.660736, + "grad_norm": 0.4078647792339325, + "learning_rate": 0.0002726446290260251, + "loss": 2.647, + "step": 5162 + }, + { + "epoch": 0.660864, + "grad_norm": 0.38353976607322693, + "learning_rate": 0.000272460009172274, + "loss": 2.6904, + "step": 5163 + }, + { + "epoch": 0.660992, + "grad_norm": 0.40887290239334106, + "learning_rate": 0.00027227542843528507, + "loss": 2.669, + "step": 5164 + }, + { + "epoch": 0.66112, + "grad_norm": 0.3800019919872284, + "learning_rate": 0.00027209088684679013, + "loss": 2.6061, + "step": 5165 + }, + { + "epoch": 0.661248, + "grad_norm": 0.5341488718986511, + "learning_rate": 0.0002719063844385138, + "loss": 2.5636, + "step": 5166 + }, + { + "epoch": 0.661376, + "grad_norm": 0.3620031774044037, + "learning_rate": 0.0002717219212421742, + "loss": 2.6546, + "step": 5167 + }, + { + "epoch": 0.661504, + "grad_norm": 0.3749382793903351, + "learning_rate": 0.00027153749728948284, + "loss": 2.6032, + "step": 5168 + }, + { + "epoch": 0.661632, + "grad_norm": 0.4382811188697815, + "learning_rate": 0.0002713531126121442, + "loss": 2.6358, + "step": 5169 + }, + { + "epoch": 0.66176, + "grad_norm": 0.40987297892570496, + "learning_rate": 0.00027116876724185633, + "loss": 2.6941, + "step": 5170 + }, + { + "epoch": 0.661888, + "grad_norm": 0.5307432413101196, + "learning_rate": 0.00027098446121031006, + "loss": 2.6158, + "step": 5171 + }, + { + "epoch": 0.662016, + "grad_norm": 0.3862095773220062, + "learning_rate": 0.00027080019454919, + "loss": 2.6647, + "step": 5172 + }, + { + "epoch": 0.662144, + "grad_norm": 0.4369296133518219, + "learning_rate": 0.00027061596729017356, + "loss": 2.7441, + "step": 5173 + }, + { + "epoch": 0.662272, + "grad_norm": 0.3724866211414337, + "learning_rate": 0.00027043177946493147, + "loss": 2.448, + "step": 5174 + }, + { + "epoch": 0.6624, + "grad_norm": 0.7063791751861572, + "learning_rate": 0.00027024763110512805, + "loss": 2.5749, + "step": 5175 + }, + { + "epoch": 0.662528, + "grad_norm": 0.3985101282596588, + "learning_rate": 0.0002700635222424204, + "loss": 2.5789, + "step": 5176 + }, + { + "epoch": 0.662656, + "grad_norm": 0.4674212336540222, + "learning_rate": 0.00026987945290845883, + "loss": 2.6104, + "step": 5177 + }, + { + "epoch": 0.662784, + "grad_norm": 0.4533572494983673, + "learning_rate": 0.0002696954231348871, + "loss": 2.716, + "step": 5178 + }, + { + "epoch": 0.662912, + "grad_norm": 0.4372638165950775, + "learning_rate": 0.0002695114329533422, + "loss": 2.6303, + "step": 5179 + }, + { + "epoch": 0.66304, + "grad_norm": 0.492294043302536, + "learning_rate": 0.0002693274823954537, + "loss": 2.609, + "step": 5180 + }, + { + "epoch": 0.663168, + "grad_norm": 0.3917854130268097, + "learning_rate": 0.0002691435714928452, + "loss": 2.6316, + "step": 5181 + }, + { + "epoch": 0.663296, + "grad_norm": 0.40750691294670105, + "learning_rate": 0.0002689597002771333, + "loss": 2.6307, + "step": 5182 + }, + { + "epoch": 0.663424, + "grad_norm": 0.39508137106895447, + "learning_rate": 0.0002687758687799273, + "loss": 2.5818, + "step": 5183 + }, + { + "epoch": 0.663552, + "grad_norm": 0.40290066599845886, + "learning_rate": 0.0002685920770328301, + "loss": 2.6496, + "step": 5184 + }, + { + "epoch": 0.66368, + "grad_norm": 0.37970519065856934, + "learning_rate": 0.0002684083250674376, + "loss": 2.4386, + "step": 5185 + }, + { + "epoch": 0.663808, + "grad_norm": 0.3917458951473236, + "learning_rate": 0.0002682246129153387, + "loss": 2.5083, + "step": 5186 + }, + { + "epoch": 0.663936, + "grad_norm": 0.4080035388469696, + "learning_rate": 0.000268040940608116, + "loss": 2.7051, + "step": 5187 + }, + { + "epoch": 0.664064, + "grad_norm": 0.4950394928455353, + "learning_rate": 0.00026785730817734446, + "loss": 2.5709, + "step": 5188 + }, + { + "epoch": 0.664192, + "grad_norm": 0.38304099440574646, + "learning_rate": 0.000267673715654593, + "loss": 2.6189, + "step": 5189 + }, + { + "epoch": 0.66432, + "grad_norm": 0.43491318821907043, + "learning_rate": 0.00026749016307142307, + "loss": 2.5012, + "step": 5190 + }, + { + "epoch": 0.664448, + "grad_norm": 0.4001062512397766, + "learning_rate": 0.0002673066504593897, + "loss": 2.6716, + "step": 5191 + }, + { + "epoch": 0.664576, + "grad_norm": 0.39897820353507996, + "learning_rate": 0.00026712317785004076, + "loss": 2.6462, + "step": 5192 + }, + { + "epoch": 0.664704, + "grad_norm": 0.4019775390625, + "learning_rate": 0.00026693974527491725, + "loss": 2.5474, + "step": 5193 + }, + { + "epoch": 0.664832, + "grad_norm": 0.4140518009662628, + "learning_rate": 0.00026675635276555347, + "loss": 2.6599, + "step": 5194 + }, + { + "epoch": 0.66496, + "grad_norm": 0.41413938999176025, + "learning_rate": 0.0002665730003534764, + "loss": 2.6008, + "step": 5195 + }, + { + "epoch": 0.665088, + "grad_norm": 0.38132086396217346, + "learning_rate": 0.0002663896880702069, + "loss": 2.5804, + "step": 5196 + }, + { + "epoch": 0.665216, + "grad_norm": 0.35278382897377014, + "learning_rate": 0.0002662064159472583, + "loss": 2.6556, + "step": 5197 + }, + { + "epoch": 0.665344, + "grad_norm": 0.3980943262577057, + "learning_rate": 0.0002660231840161371, + "loss": 2.6372, + "step": 5198 + }, + { + "epoch": 0.665472, + "grad_norm": 0.39521658420562744, + "learning_rate": 0.00026583999230834294, + "loss": 2.68, + "step": 5199 + }, + { + "epoch": 0.6656, + "grad_norm": 0.4064350724220276, + "learning_rate": 0.0002656568408553689, + "loss": 2.5757, + "step": 5200 + }, + { + "epoch": 0.665728, + "grad_norm": 0.40529003739356995, + "learning_rate": 0.0002654737296887006, + "loss": 2.6062, + "step": 5201 + }, + { + "epoch": 0.665856, + "grad_norm": 0.7186464071273804, + "learning_rate": 0.00026529065883981707, + "loss": 2.6254, + "step": 5202 + }, + { + "epoch": 0.665984, + "grad_norm": 0.3919147551059723, + "learning_rate": 0.00026510762834019004, + "loss": 2.5496, + "step": 5203 + }, + { + "epoch": 0.666112, + "grad_norm": 0.36975550651550293, + "learning_rate": 0.000264924638221285, + "loss": 2.5121, + "step": 5204 + }, + { + "epoch": 0.66624, + "grad_norm": 0.5004108548164368, + "learning_rate": 0.0002647416885145599, + "loss": 2.6952, + "step": 5205 + }, + { + "epoch": 0.666368, + "grad_norm": 0.38370251655578613, + "learning_rate": 0.0002645587792514659, + "loss": 2.578, + "step": 5206 + }, + { + "epoch": 0.666496, + "grad_norm": 0.3962876796722412, + "learning_rate": 0.0002643759104634472, + "loss": 2.5954, + "step": 5207 + }, + { + "epoch": 0.666624, + "grad_norm": 0.3682904541492462, + "learning_rate": 0.00026419308218194085, + "loss": 2.644, + "step": 5208 + }, + { + "epoch": 0.666752, + "grad_norm": 0.37494441866874695, + "learning_rate": 0.00026401029443837767, + "loss": 2.4844, + "step": 5209 + }, + { + "epoch": 0.66688, + "grad_norm": 0.4283830225467682, + "learning_rate": 0.00026382754726418043, + "loss": 2.6061, + "step": 5210 + }, + { + "epoch": 0.667008, + "grad_norm": 0.4135827124118805, + "learning_rate": 0.000263644840690766, + "loss": 2.6034, + "step": 5211 + }, + { + "epoch": 0.667136, + "grad_norm": 0.3682136833667755, + "learning_rate": 0.00026346217474954346, + "loss": 2.6607, + "step": 5212 + }, + { + "epoch": 0.667264, + "grad_norm": 0.3946116268634796, + "learning_rate": 0.00026327954947191534, + "loss": 2.6301, + "step": 5213 + }, + { + "epoch": 0.667392, + "grad_norm": 0.3795643448829651, + "learning_rate": 0.0002630969648892769, + "loss": 2.6435, + "step": 5214 + }, + { + "epoch": 0.66752, + "grad_norm": 0.40038546919822693, + "learning_rate": 0.00026291442103301666, + "loss": 2.6663, + "step": 5215 + }, + { + "epoch": 0.667648, + "grad_norm": 0.37560898065567017, + "learning_rate": 0.0002627319179345157, + "loss": 2.5958, + "step": 5216 + }, + { + "epoch": 0.667776, + "grad_norm": 0.3793976306915283, + "learning_rate": 0.00026254945562514877, + "loss": 2.6503, + "step": 5217 + }, + { + "epoch": 0.667904, + "grad_norm": 0.3943001925945282, + "learning_rate": 0.0002623670341362834, + "loss": 2.6033, + "step": 5218 + }, + { + "epoch": 0.668032, + "grad_norm": 0.4187106192111969, + "learning_rate": 0.00026218465349927964, + "loss": 2.643, + "step": 5219 + }, + { + "epoch": 0.66816, + "grad_norm": 0.4091593325138092, + "learning_rate": 0.000262002313745491, + "loss": 2.616, + "step": 5220 + }, + { + "epoch": 0.668288, + "grad_norm": 0.3878062963485718, + "learning_rate": 0.00026182001490626373, + "loss": 2.565, + "step": 5221 + }, + { + "epoch": 0.668416, + "grad_norm": 0.410752534866333, + "learning_rate": 0.00026163775701293715, + "loss": 2.5885, + "step": 5222 + }, + { + "epoch": 0.668544, + "grad_norm": 0.3864632844924927, + "learning_rate": 0.0002614555400968434, + "loss": 2.5937, + "step": 5223 + }, + { + "epoch": 0.668672, + "grad_norm": 0.39001286029815674, + "learning_rate": 0.0002612733641893076, + "loss": 2.6535, + "step": 5224 + }, + { + "epoch": 0.6688, + "grad_norm": 0.40915483236312866, + "learning_rate": 0.00026109122932164823, + "loss": 2.6035, + "step": 5225 + }, + { + "epoch": 0.668928, + "grad_norm": 0.4118044376373291, + "learning_rate": 0.00026090913552517624, + "loss": 2.7845, + "step": 5226 + }, + { + "epoch": 0.669056, + "grad_norm": 0.4550974667072296, + "learning_rate": 0.0002607270828311954, + "loss": 2.6243, + "step": 5227 + }, + { + "epoch": 0.669184, + "grad_norm": 0.41496187448501587, + "learning_rate": 0.00026054507127100314, + "loss": 2.6561, + "step": 5228 + }, + { + "epoch": 0.669312, + "grad_norm": 0.4162474572658539, + "learning_rate": 0.0002603631008758891, + "loss": 2.5173, + "step": 5229 + }, + { + "epoch": 0.66944, + "grad_norm": 0.4632064700126648, + "learning_rate": 0.00026018117167713617, + "loss": 2.6421, + "step": 5230 + }, + { + "epoch": 0.669568, + "grad_norm": 0.377778023481369, + "learning_rate": 0.0002599992837060198, + "loss": 2.6384, + "step": 5231 + }, + { + "epoch": 0.669696, + "grad_norm": 0.3968907296657562, + "learning_rate": 0.0002598174369938091, + "loss": 2.5964, + "step": 5232 + }, + { + "epoch": 0.669824, + "grad_norm": 0.39502325654029846, + "learning_rate": 0.0002596356315717654, + "loss": 2.623, + "step": 5233 + }, + { + "epoch": 0.669952, + "grad_norm": 0.40071040391921997, + "learning_rate": 0.0002594538674711432, + "loss": 2.5859, + "step": 5234 + }, + { + "epoch": 0.67008, + "grad_norm": 0.4093089997768402, + "learning_rate": 0.0002592721447231898, + "loss": 2.6769, + "step": 5235 + }, + { + "epoch": 0.670208, + "grad_norm": 0.3827778398990631, + "learning_rate": 0.00025909046335914534, + "loss": 2.7091, + "step": 5236 + }, + { + "epoch": 0.670336, + "grad_norm": 0.3861181437969208, + "learning_rate": 0.0002589088234102432, + "loss": 2.6778, + "step": 5237 + }, + { + "epoch": 0.670464, + "grad_norm": 0.387295663356781, + "learning_rate": 0.00025872722490770915, + "loss": 2.5743, + "step": 5238 + }, + { + "epoch": 0.670592, + "grad_norm": 0.36986446380615234, + "learning_rate": 0.00025854566788276244, + "loss": 2.5999, + "step": 5239 + }, + { + "epoch": 0.67072, + "grad_norm": 0.4120786786079407, + "learning_rate": 0.0002583641523666146, + "loss": 2.5062, + "step": 5240 + }, + { + "epoch": 0.670848, + "grad_norm": 0.41202569007873535, + "learning_rate": 0.0002581826783904702, + "loss": 2.5734, + "step": 5241 + }, + { + "epoch": 0.670976, + "grad_norm": 0.5148423314094543, + "learning_rate": 0.00025800124598552676, + "loss": 2.6533, + "step": 5242 + }, + { + "epoch": 0.671104, + "grad_norm": 0.41330772638320923, + "learning_rate": 0.0002578198551829747, + "loss": 2.7358, + "step": 5243 + }, + { + "epoch": 0.671232, + "grad_norm": 0.47050169110298157, + "learning_rate": 0.00025763850601399706, + "loss": 2.626, + "step": 5244 + }, + { + "epoch": 0.67136, + "grad_norm": 0.38003620505332947, + "learning_rate": 0.00025745719850976976, + "loss": 2.5988, + "step": 5245 + }, + { + "epoch": 0.671488, + "grad_norm": 0.4140872657299042, + "learning_rate": 0.0002572759327014618, + "loss": 2.5223, + "step": 5246 + }, + { + "epoch": 0.671616, + "grad_norm": 0.42146480083465576, + "learning_rate": 0.00025709470862023516, + "loss": 2.5561, + "step": 5247 + }, + { + "epoch": 0.671744, + "grad_norm": 0.4432838261127472, + "learning_rate": 0.00025691352629724406, + "loss": 2.7031, + "step": 5248 + }, + { + "epoch": 0.671872, + "grad_norm": 0.4066985547542572, + "learning_rate": 0.0002567323857636359, + "loss": 2.5489, + "step": 5249 + }, + { + "epoch": 0.672, + "grad_norm": 0.435158908367157, + "learning_rate": 0.00025655128705055085, + "loss": 2.6508, + "step": 5250 + }, + { + "epoch": 0.672128, + "grad_norm": 0.37268921732902527, + "learning_rate": 0.00025637023018912185, + "loss": 2.6021, + "step": 5251 + }, + { + "epoch": 0.672256, + "grad_norm": 0.4060450494289398, + "learning_rate": 0.0002561892152104744, + "loss": 2.6307, + "step": 5252 + }, + { + "epoch": 0.672384, + "grad_norm": 0.36908289790153503, + "learning_rate": 0.0002560082421457277, + "loss": 2.5995, + "step": 5253 + }, + { + "epoch": 0.672512, + "grad_norm": 0.4006135165691376, + "learning_rate": 0.0002558273110259925, + "loss": 2.6522, + "step": 5254 + }, + { + "epoch": 0.67264, + "grad_norm": 0.38714247941970825, + "learning_rate": 0.00025564642188237344, + "loss": 2.6368, + "step": 5255 + }, + { + "epoch": 0.672768, + "grad_norm": 0.39564505219459534, + "learning_rate": 0.0002554655747459672, + "loss": 2.5862, + "step": 5256 + }, + { + "epoch": 0.672896, + "grad_norm": 0.36665773391723633, + "learning_rate": 0.0002552847696478636, + "loss": 2.6065, + "step": 5257 + }, + { + "epoch": 0.673024, + "grad_norm": 0.3964935541152954, + "learning_rate": 0.000255104006619145, + "loss": 2.586, + "step": 5258 + }, + { + "epoch": 0.673152, + "grad_norm": 0.42348843812942505, + "learning_rate": 0.00025492328569088664, + "loss": 2.6096, + "step": 5259 + }, + { + "epoch": 0.67328, + "grad_norm": 0.37963181734085083, + "learning_rate": 0.00025474260689415686, + "loss": 2.6806, + "step": 5260 + }, + { + "epoch": 0.673408, + "grad_norm": 0.4240288734436035, + "learning_rate": 0.0002545619702600161, + "loss": 2.5448, + "step": 5261 + }, + { + "epoch": 0.673536, + "grad_norm": 0.36575382947921753, + "learning_rate": 0.0002543813758195181, + "loss": 2.6449, + "step": 5262 + }, + { + "epoch": 0.673664, + "grad_norm": 0.42919155955314636, + "learning_rate": 0.000254200823603709, + "loss": 2.5934, + "step": 5263 + }, + { + "epoch": 0.673792, + "grad_norm": 0.44469746947288513, + "learning_rate": 0.0002540203136436277, + "loss": 2.6047, + "step": 5264 + }, + { + "epoch": 0.67392, + "grad_norm": 0.4368082880973816, + "learning_rate": 0.00025383984597030635, + "loss": 2.6145, + "step": 5265 + }, + { + "epoch": 0.674048, + "grad_norm": 0.4023550748825073, + "learning_rate": 0.0002536594206147692, + "loss": 2.6422, + "step": 5266 + }, + { + "epoch": 0.674176, + "grad_norm": 0.40943852066993713, + "learning_rate": 0.00025347903760803325, + "loss": 2.6311, + "step": 5267 + }, + { + "epoch": 0.674304, + "grad_norm": 0.3796302378177643, + "learning_rate": 0.00025329869698110884, + "loss": 2.677, + "step": 5268 + }, + { + "epoch": 0.674432, + "grad_norm": 0.43228548765182495, + "learning_rate": 0.00025311839876499844, + "loss": 2.5058, + "step": 5269 + }, + { + "epoch": 0.67456, + "grad_norm": 0.44568175077438354, + "learning_rate": 0.0002529381429906974, + "loss": 2.6756, + "step": 5270 + }, + { + "epoch": 0.674688, + "grad_norm": 0.3645046055316925, + "learning_rate": 0.00025275792968919373, + "loss": 2.5985, + "step": 5271 + }, + { + "epoch": 0.674816, + "grad_norm": 0.3818860948085785, + "learning_rate": 0.00025257775889146825, + "loss": 2.6103, + "step": 5272 + }, + { + "epoch": 0.674944, + "grad_norm": 0.3848963975906372, + "learning_rate": 0.00025239763062849427, + "loss": 2.5989, + "step": 5273 + }, + { + "epoch": 0.675072, + "grad_norm": 0.3937123417854309, + "learning_rate": 0.00025221754493123794, + "loss": 2.5979, + "step": 5274 + }, + { + "epoch": 0.6752, + "grad_norm": 0.3734511733055115, + "learning_rate": 0.0002520375018306584, + "loss": 2.5283, + "step": 5275 + }, + { + "epoch": 0.675328, + "grad_norm": 0.41701263189315796, + "learning_rate": 0.0002518575013577069, + "loss": 2.6141, + "step": 5276 + }, + { + "epoch": 0.675456, + "grad_norm": 0.38944172859191895, + "learning_rate": 0.0002516775435433276, + "loss": 2.5833, + "step": 5277 + }, + { + "epoch": 0.675584, + "grad_norm": 0.4079633355140686, + "learning_rate": 0.00025149762841845746, + "loss": 2.5517, + "step": 5278 + }, + { + "epoch": 0.675712, + "grad_norm": 0.4900324046611786, + "learning_rate": 0.00025131775601402583, + "loss": 2.5515, + "step": 5279 + }, + { + "epoch": 0.67584, + "grad_norm": 0.3572828769683838, + "learning_rate": 0.00025113792636095465, + "loss": 2.6167, + "step": 5280 + }, + { + "epoch": 0.675968, + "grad_norm": 0.3777770400047302, + "learning_rate": 0.00025095813949015935, + "loss": 2.626, + "step": 5281 + }, + { + "epoch": 0.676096, + "grad_norm": 0.41958168148994446, + "learning_rate": 0.0002507783954325467, + "loss": 2.6008, + "step": 5282 + }, + { + "epoch": 0.676224, + "grad_norm": 0.3967720568180084, + "learning_rate": 0.0002505986942190173, + "loss": 2.6211, + "step": 5283 + }, + { + "epoch": 0.676352, + "grad_norm": 0.6515386700630188, + "learning_rate": 0.0002504190358804637, + "loss": 2.6481, + "step": 5284 + }, + { + "epoch": 0.67648, + "grad_norm": 0.45086121559143066, + "learning_rate": 0.00025023942044777114, + "loss": 2.5695, + "step": 5285 + }, + { + "epoch": 0.676608, + "grad_norm": 0.4137996733188629, + "learning_rate": 0.0002500598479518178, + "loss": 2.7222, + "step": 5286 + }, + { + "epoch": 0.676736, + "grad_norm": 0.3994307518005371, + "learning_rate": 0.00024988031842347407, + "loss": 2.6124, + "step": 5287 + }, + { + "epoch": 0.676864, + "grad_norm": 0.38529929518699646, + "learning_rate": 0.0002497008318936031, + "loss": 2.6348, + "step": 5288 + }, + { + "epoch": 0.676992, + "grad_norm": 0.38449084758758545, + "learning_rate": 0.000249521388393061, + "loss": 2.6979, + "step": 5289 + }, + { + "epoch": 0.67712, + "grad_norm": 0.38079801201820374, + "learning_rate": 0.00024934198795269604, + "loss": 2.643, + "step": 5290 + }, + { + "epoch": 0.677248, + "grad_norm": 0.37083399295806885, + "learning_rate": 0.00024916263060334904, + "loss": 2.6237, + "step": 5291 + }, + { + "epoch": 0.677376, + "grad_norm": 0.4245010316371918, + "learning_rate": 0.000248983316375854, + "loss": 2.6452, + "step": 5292 + }, + { + "epoch": 0.677504, + "grad_norm": 0.3888593018054962, + "learning_rate": 0.0002488040453010369, + "loss": 2.632, + "step": 5293 + }, + { + "epoch": 0.677632, + "grad_norm": 0.4005393087863922, + "learning_rate": 0.0002486248174097165, + "loss": 2.6361, + "step": 5294 + }, + { + "epoch": 0.67776, + "grad_norm": 0.4021790623664856, + "learning_rate": 0.00024844563273270404, + "loss": 2.5945, + "step": 5295 + }, + { + "epoch": 0.677888, + "grad_norm": 0.4169047474861145, + "learning_rate": 0.0002482664913008037, + "loss": 2.6303, + "step": 5296 + }, + { + "epoch": 0.678016, + "grad_norm": 0.3867737352848053, + "learning_rate": 0.0002480873931448119, + "loss": 2.6929, + "step": 5297 + }, + { + "epoch": 0.678144, + "grad_norm": 0.37985959649086, + "learning_rate": 0.0002479083382955176, + "loss": 2.6188, + "step": 5298 + }, + { + "epoch": 0.678272, + "grad_norm": 0.38549497723579407, + "learning_rate": 0.0002477293267837025, + "loss": 2.7501, + "step": 5299 + }, + { + "epoch": 0.6784, + "grad_norm": 0.3828747272491455, + "learning_rate": 0.00024755035864014067, + "loss": 2.6481, + "step": 5300 + }, + { + "epoch": 0.678528, + "grad_norm": 0.3690762519836426, + "learning_rate": 0.0002473714338955987, + "loss": 2.6697, + "step": 5301 + }, + { + "epoch": 0.678656, + "grad_norm": 0.3891821801662445, + "learning_rate": 0.0002471925525808361, + "loss": 2.5382, + "step": 5302 + }, + { + "epoch": 0.678784, + "grad_norm": 0.46617451310157776, + "learning_rate": 0.0002470137147266043, + "loss": 2.6421, + "step": 5303 + }, + { + "epoch": 0.678912, + "grad_norm": 0.3878360092639923, + "learning_rate": 0.000246834920363648, + "loss": 2.5466, + "step": 5304 + }, + { + "epoch": 0.67904, + "grad_norm": 0.9007443189620972, + "learning_rate": 0.00024665616952270386, + "loss": 2.5599, + "step": 5305 + }, + { + "epoch": 0.679168, + "grad_norm": 0.3993183970451355, + "learning_rate": 0.00024647746223450126, + "loss": 2.6879, + "step": 5306 + }, + { + "epoch": 0.679296, + "grad_norm": 0.40214061737060547, + "learning_rate": 0.0002462987985297619, + "loss": 2.6438, + "step": 5307 + }, + { + "epoch": 0.679424, + "grad_norm": 0.3603222966194153, + "learning_rate": 0.0002461201784392002, + "loss": 2.6291, + "step": 5308 + }, + { + "epoch": 0.679552, + "grad_norm": 0.39639052748680115, + "learning_rate": 0.0002459416019935229, + "loss": 2.6493, + "step": 5309 + }, + { + "epoch": 0.67968, + "grad_norm": 0.36325621604919434, + "learning_rate": 0.0002457630692234296, + "loss": 2.5363, + "step": 5310 + }, + { + "epoch": 0.679808, + "grad_norm": 0.4159042239189148, + "learning_rate": 0.0002455845801596121, + "loss": 2.7334, + "step": 5311 + }, + { + "epoch": 0.679936, + "grad_norm": 0.38538694381713867, + "learning_rate": 0.0002454061348327548, + "loss": 2.6269, + "step": 5312 + }, + { + "epoch": 0.680064, + "grad_norm": 0.35998186469078064, + "learning_rate": 0.00024522773327353444, + "loss": 2.6094, + "step": 5313 + }, + { + "epoch": 0.680192, + "grad_norm": 0.46342217922210693, + "learning_rate": 0.00024504937551262024, + "loss": 2.5925, + "step": 5314 + }, + { + "epoch": 0.68032, + "grad_norm": 0.3912767767906189, + "learning_rate": 0.00024487106158067407, + "loss": 2.6276, + "step": 5315 + }, + { + "epoch": 0.680448, + "grad_norm": 0.3866388499736786, + "learning_rate": 0.00024469279150834993, + "loss": 2.6949, + "step": 5316 + }, + { + "epoch": 0.680576, + "grad_norm": 0.3818334639072418, + "learning_rate": 0.00024451456532629493, + "loss": 2.6339, + "step": 5317 + }, + { + "epoch": 0.680704, + "grad_norm": 0.4885517358779907, + "learning_rate": 0.00024433638306514796, + "loss": 2.536, + "step": 5318 + }, + { + "epoch": 0.680832, + "grad_norm": 0.38744550943374634, + "learning_rate": 0.00024415824475554045, + "loss": 2.6761, + "step": 5319 + }, + { + "epoch": 0.68096, + "grad_norm": 0.40208789706230164, + "learning_rate": 0.0002439801504280968, + "loss": 2.5334, + "step": 5320 + }, + { + "epoch": 0.681088, + "grad_norm": 0.4112870693206787, + "learning_rate": 0.0002438021001134334, + "loss": 2.6315, + "step": 5321 + }, + { + "epoch": 0.681216, + "grad_norm": 0.3894655704498291, + "learning_rate": 0.00024362409384215904, + "loss": 2.6579, + "step": 5322 + }, + { + "epoch": 0.681344, + "grad_norm": 0.3734358847141266, + "learning_rate": 0.00024344613164487512, + "loss": 2.6154, + "step": 5323 + }, + { + "epoch": 0.681472, + "grad_norm": 0.37331709265708923, + "learning_rate": 0.0002432682135521752, + "loss": 2.5718, + "step": 5324 + }, + { + "epoch": 0.6816, + "grad_norm": 0.38891661167144775, + "learning_rate": 0.0002430903395946459, + "loss": 2.4885, + "step": 5325 + }, + { + "epoch": 0.681728, + "grad_norm": 0.3664712905883789, + "learning_rate": 0.0002429125098028655, + "loss": 2.596, + "step": 5326 + }, + { + "epoch": 0.681856, + "grad_norm": 0.3787769675254822, + "learning_rate": 0.0002427347242074051, + "loss": 2.5016, + "step": 5327 + }, + { + "epoch": 0.681984, + "grad_norm": 0.4327602684497833, + "learning_rate": 0.00024255698283882793, + "loss": 2.6055, + "step": 5328 + }, + { + "epoch": 0.682112, + "grad_norm": 0.39133667945861816, + "learning_rate": 0.00024237928572769013, + "loss": 2.6428, + "step": 5329 + }, + { + "epoch": 0.68224, + "grad_norm": 0.37107619643211365, + "learning_rate": 0.00024220163290453972, + "loss": 2.5286, + "step": 5330 + }, + { + "epoch": 0.682368, + "grad_norm": 0.5632586479187012, + "learning_rate": 0.00024202402439991706, + "loss": 2.5359, + "step": 5331 + }, + { + "epoch": 0.682496, + "grad_norm": 0.3835335671901703, + "learning_rate": 0.00024184646024435547, + "loss": 2.6763, + "step": 5332 + }, + { + "epoch": 0.682624, + "grad_norm": 0.4023592174053192, + "learning_rate": 0.00024166894046838018, + "loss": 2.6321, + "step": 5333 + }, + { + "epoch": 0.682752, + "grad_norm": 0.4059971868991852, + "learning_rate": 0.00024149146510250884, + "loss": 2.5377, + "step": 5334 + }, + { + "epoch": 0.68288, + "grad_norm": 0.4482147693634033, + "learning_rate": 0.00024131403417725156, + "loss": 2.6393, + "step": 5335 + }, + { + "epoch": 0.683008, + "grad_norm": 0.41405385732650757, + "learning_rate": 0.00024113664772311072, + "loss": 2.5809, + "step": 5336 + }, + { + "epoch": 0.683136, + "grad_norm": 0.4077891707420349, + "learning_rate": 0.000240959305770581, + "loss": 2.6082, + "step": 5337 + }, + { + "epoch": 0.683264, + "grad_norm": 0.3854316473007202, + "learning_rate": 0.0002407820083501498, + "loss": 2.6591, + "step": 5338 + }, + { + "epoch": 0.683392, + "grad_norm": 0.4656102955341339, + "learning_rate": 0.00024060475549229633, + "loss": 2.7274, + "step": 5339 + }, + { + "epoch": 0.68352, + "grad_norm": 0.40924736857414246, + "learning_rate": 0.0002404275472274927, + "loss": 2.6012, + "step": 5340 + }, + { + "epoch": 0.683648, + "grad_norm": 0.3960039019584656, + "learning_rate": 0.000240250383586203, + "loss": 2.6483, + "step": 5341 + }, + { + "epoch": 0.683776, + "grad_norm": 0.3897576332092285, + "learning_rate": 0.00024007326459888357, + "loss": 2.6799, + "step": 5342 + }, + { + "epoch": 0.683904, + "grad_norm": 0.4071364402770996, + "learning_rate": 0.0002398961902959833, + "loss": 2.6348, + "step": 5343 + }, + { + "epoch": 0.684032, + "grad_norm": 0.40977510809898376, + "learning_rate": 0.00023971916070794337, + "loss": 2.5616, + "step": 5344 + }, + { + "epoch": 0.68416, + "grad_norm": 0.38848447799682617, + "learning_rate": 0.0002395421758651969, + "loss": 2.5763, + "step": 5345 + }, + { + "epoch": 0.684288, + "grad_norm": 0.47347959876060486, + "learning_rate": 0.00023936523579817003, + "loss": 2.578, + "step": 5346 + }, + { + "epoch": 0.684416, + "grad_norm": 0.4182603657245636, + "learning_rate": 0.0002391883405372805, + "loss": 2.632, + "step": 5347 + }, + { + "epoch": 0.684544, + "grad_norm": 0.41833582520484924, + "learning_rate": 0.00023901149011293905, + "loss": 2.7022, + "step": 5348 + }, + { + "epoch": 0.684672, + "grad_norm": 0.4208545982837677, + "learning_rate": 0.00023883468455554807, + "loss": 2.6294, + "step": 5349 + }, + { + "epoch": 0.6848, + "grad_norm": 0.4403442442417145, + "learning_rate": 0.0002386579238955025, + "loss": 2.5978, + "step": 5350 + }, + { + "epoch": 0.684928, + "grad_norm": 0.4147782623767853, + "learning_rate": 0.00023848120816318954, + "loss": 2.7614, + "step": 5351 + }, + { + "epoch": 0.685056, + "grad_norm": 0.430778443813324, + "learning_rate": 0.0002383045373889885, + "loss": 2.572, + "step": 5352 + }, + { + "epoch": 0.685184, + "grad_norm": 0.4295988976955414, + "learning_rate": 0.00023812791160327153, + "loss": 2.5772, + "step": 5353 + }, + { + "epoch": 0.685312, + "grad_norm": 0.42285963892936707, + "learning_rate": 0.0002379513308364024, + "loss": 2.6522, + "step": 5354 + }, + { + "epoch": 0.68544, + "grad_norm": 0.4423457682132721, + "learning_rate": 0.00023777479511873746, + "loss": 2.7266, + "step": 5355 + }, + { + "epoch": 0.685568, + "grad_norm": 0.3944604694843292, + "learning_rate": 0.00023759830448062498, + "loss": 2.6593, + "step": 5356 + }, + { + "epoch": 0.685696, + "grad_norm": 0.4161713421344757, + "learning_rate": 0.0002374218589524062, + "loss": 2.5668, + "step": 5357 + }, + { + "epoch": 0.685824, + "grad_norm": 0.4068816304206848, + "learning_rate": 0.00023724545856441392, + "loss": 2.5894, + "step": 5358 + }, + { + "epoch": 0.685952, + "grad_norm": 0.37721022963523865, + "learning_rate": 0.00023706910334697347, + "loss": 2.5418, + "step": 5359 + }, + { + "epoch": 0.68608, + "grad_norm": 0.4367161989212036, + "learning_rate": 0.0002368927933304021, + "loss": 2.6242, + "step": 5360 + }, + { + "epoch": 0.686208, + "grad_norm": 0.3834114372730255, + "learning_rate": 0.0002367165285450099, + "loss": 2.6907, + "step": 5361 + }, + { + "epoch": 0.686336, + "grad_norm": 0.3976205885410309, + "learning_rate": 0.00023654030902109873, + "loss": 2.5993, + "step": 5362 + }, + { + "epoch": 0.686464, + "grad_norm": 0.3984668254852295, + "learning_rate": 0.00023636413478896274, + "loss": 2.6097, + "step": 5363 + }, + { + "epoch": 0.686592, + "grad_norm": 0.4570087492465973, + "learning_rate": 0.00023618800587888827, + "loss": 2.4318, + "step": 5364 + }, + { + "epoch": 0.68672, + "grad_norm": 0.38049113750457764, + "learning_rate": 0.00023601192232115375, + "loss": 2.4816, + "step": 5365 + }, + { + "epoch": 0.686848, + "grad_norm": 0.370777428150177, + "learning_rate": 0.00023583588414603047, + "loss": 2.6833, + "step": 5366 + }, + { + "epoch": 0.686976, + "grad_norm": 0.40403929352760315, + "learning_rate": 0.0002356598913837809, + "loss": 2.5871, + "step": 5367 + }, + { + "epoch": 0.687104, + "grad_norm": 0.44271817803382874, + "learning_rate": 0.00023548394406466074, + "loss": 2.6547, + "step": 5368 + }, + { + "epoch": 0.687232, + "grad_norm": 0.44872596859931946, + "learning_rate": 0.00023530804221891705, + "loss": 2.5643, + "step": 5369 + }, + { + "epoch": 0.68736, + "grad_norm": 0.36122024059295654, + "learning_rate": 0.00023513218587678953, + "loss": 2.5837, + "step": 5370 + }, + { + "epoch": 0.687488, + "grad_norm": 0.4133700430393219, + "learning_rate": 0.00023495637506850987, + "loss": 2.5156, + "step": 5371 + }, + { + "epoch": 0.687616, + "grad_norm": 0.39579516649246216, + "learning_rate": 0.00023478060982430195, + "loss": 2.562, + "step": 5372 + }, + { + "epoch": 0.687744, + "grad_norm": 0.40363189578056335, + "learning_rate": 0.00023460489017438174, + "loss": 2.6626, + "step": 5373 + }, + { + "epoch": 0.687872, + "grad_norm": 0.3823588788509369, + "learning_rate": 0.00023442921614895778, + "loss": 2.5571, + "step": 5374 + }, + { + "epoch": 0.688, + "grad_norm": 0.38040804862976074, + "learning_rate": 0.0002342535877782302, + "loss": 2.5697, + "step": 5375 + }, + { + "epoch": 0.688128, + "grad_norm": 0.41419199109077454, + "learning_rate": 0.0002340780050923918, + "loss": 2.6696, + "step": 5376 + }, + { + "epoch": 0.688256, + "grad_norm": 0.40255481004714966, + "learning_rate": 0.00023390246812162718, + "loss": 2.645, + "step": 5377 + }, + { + "epoch": 0.688384, + "grad_norm": 0.46150800585746765, + "learning_rate": 0.00023372697689611317, + "loss": 2.6288, + "step": 5378 + }, + { + "epoch": 0.688512, + "grad_norm": 0.4139755368232727, + "learning_rate": 0.0002335515314460187, + "loss": 2.6165, + "step": 5379 + }, + { + "epoch": 0.68864, + "grad_norm": 0.4696686565876007, + "learning_rate": 0.00023337613180150497, + "loss": 2.6331, + "step": 5380 + }, + { + "epoch": 0.688768, + "grad_norm": 0.4530939757823944, + "learning_rate": 0.00023320077799272493, + "loss": 2.6575, + "step": 5381 + }, + { + "epoch": 0.688896, + "grad_norm": 0.39919939637184143, + "learning_rate": 0.00023302547004982433, + "loss": 2.6515, + "step": 5382 + }, + { + "epoch": 0.689024, + "grad_norm": 0.43258780241012573, + "learning_rate": 0.0002328502080029405, + "loss": 2.5821, + "step": 5383 + }, + { + "epoch": 0.689152, + "grad_norm": 0.384708046913147, + "learning_rate": 0.00023267499188220277, + "loss": 2.5473, + "step": 5384 + }, + { + "epoch": 0.68928, + "grad_norm": 0.37090280652046204, + "learning_rate": 0.00023249982171773322, + "loss": 2.5944, + "step": 5385 + }, + { + "epoch": 0.689408, + "grad_norm": 0.4394456446170807, + "learning_rate": 0.0002323246975396454, + "loss": 2.621, + "step": 5386 + }, + { + "epoch": 0.689536, + "grad_norm": 0.39605075120925903, + "learning_rate": 0.0002321496193780453, + "loss": 2.5275, + "step": 5387 + }, + { + "epoch": 0.689664, + "grad_norm": 0.5400271415710449, + "learning_rate": 0.0002319745872630306, + "loss": 2.6198, + "step": 5388 + }, + { + "epoch": 0.689792, + "grad_norm": 0.42227548360824585, + "learning_rate": 0.00023179960122469173, + "loss": 2.5439, + "step": 5389 + }, + { + "epoch": 0.68992, + "grad_norm": 0.40714338421821594, + "learning_rate": 0.00023162466129311073, + "loss": 2.6067, + "step": 5390 + }, + { + "epoch": 0.690048, + "grad_norm": 0.6505782604217529, + "learning_rate": 0.00023144976749836166, + "loss": 2.6345, + "step": 5391 + }, + { + "epoch": 0.690176, + "grad_norm": 0.4194187521934509, + "learning_rate": 0.00023127491987051085, + "loss": 2.702, + "step": 5392 + }, + { + "epoch": 0.690304, + "grad_norm": 0.4079093635082245, + "learning_rate": 0.0002311001184396165, + "loss": 2.6011, + "step": 5393 + }, + { + "epoch": 0.690432, + "grad_norm": 0.422248899936676, + "learning_rate": 0.00023092536323572933, + "loss": 2.5542, + "step": 5394 + }, + { + "epoch": 0.69056, + "grad_norm": 0.3794178068637848, + "learning_rate": 0.0002307506542888916, + "loss": 2.6265, + "step": 5395 + }, + { + "epoch": 0.690688, + "grad_norm": 0.452921599149704, + "learning_rate": 0.00023057599162913756, + "loss": 2.5611, + "step": 5396 + }, + { + "epoch": 0.690816, + "grad_norm": 0.39621347188949585, + "learning_rate": 0.00023040137528649424, + "loss": 2.6956, + "step": 5397 + }, + { + "epoch": 0.690944, + "grad_norm": 0.4603031575679779, + "learning_rate": 0.00023022680529097994, + "loss": 2.5975, + "step": 5398 + }, + { + "epoch": 0.691072, + "grad_norm": 0.42886486649513245, + "learning_rate": 0.00023005228167260528, + "loss": 2.7765, + "step": 5399 + }, + { + "epoch": 0.6912, + "grad_norm": 0.4351581037044525, + "learning_rate": 0.00022987780446137292, + "loss": 2.6847, + "step": 5400 + }, + { + "epoch": 0.691328, + "grad_norm": 0.42801740765571594, + "learning_rate": 0.0002297033736872775, + "loss": 2.6158, + "step": 5401 + }, + { + "epoch": 0.691456, + "grad_norm": 0.38079455494880676, + "learning_rate": 0.00022952898938030552, + "loss": 2.6638, + "step": 5402 + }, + { + "epoch": 0.691584, + "grad_norm": 0.43478846549987793, + "learning_rate": 0.0002293546515704358, + "loss": 2.5386, + "step": 5403 + }, + { + "epoch": 0.691712, + "grad_norm": 0.39340740442276, + "learning_rate": 0.00022918036028763927, + "loss": 2.6724, + "step": 5404 + }, + { + "epoch": 0.69184, + "grad_norm": 0.46107327938079834, + "learning_rate": 0.0002290061155618784, + "loss": 2.6322, + "step": 5405 + }, + { + "epoch": 0.691968, + "grad_norm": 0.418556272983551, + "learning_rate": 0.00022883191742310795, + "loss": 2.6175, + "step": 5406 + }, + { + "epoch": 0.692096, + "grad_norm": 0.3955962359905243, + "learning_rate": 0.00022865776590127445, + "loss": 2.5618, + "step": 5407 + }, + { + "epoch": 0.692224, + "grad_norm": 0.40389180183410645, + "learning_rate": 0.00022848366102631675, + "loss": 2.5531, + "step": 5408 + }, + { + "epoch": 0.692352, + "grad_norm": 0.40488961338996887, + "learning_rate": 0.00022830960282816515, + "loss": 2.6185, + "step": 5409 + }, + { + "epoch": 0.69248, + "grad_norm": 0.46863025426864624, + "learning_rate": 0.00022813559133674278, + "loss": 2.6355, + "step": 5410 + }, + { + "epoch": 0.692608, + "grad_norm": 0.3847232162952423, + "learning_rate": 0.00022796162658196396, + "loss": 2.7147, + "step": 5411 + }, + { + "epoch": 0.692736, + "grad_norm": 0.3768986761569977, + "learning_rate": 0.00022778770859373503, + "loss": 2.4858, + "step": 5412 + }, + { + "epoch": 0.692864, + "grad_norm": 0.40156102180480957, + "learning_rate": 0.000227613837401955, + "loss": 2.6116, + "step": 5413 + }, + { + "epoch": 0.692992, + "grad_norm": 0.41063565015792847, + "learning_rate": 0.00022744001303651407, + "loss": 2.6596, + "step": 5414 + }, + { + "epoch": 0.69312, + "grad_norm": 0.39824721217155457, + "learning_rate": 0.0002272662355272947, + "loss": 2.5798, + "step": 5415 + }, + { + "epoch": 0.693248, + "grad_norm": 0.4192044734954834, + "learning_rate": 0.0002270925049041712, + "loss": 2.5894, + "step": 5416 + }, + { + "epoch": 0.693376, + "grad_norm": 0.3851228952407837, + "learning_rate": 0.0002269188211970098, + "loss": 2.5895, + "step": 5417 + }, + { + "epoch": 0.693504, + "grad_norm": 0.4149327576160431, + "learning_rate": 0.00022674518443566906, + "loss": 2.6931, + "step": 5418 + }, + { + "epoch": 0.693632, + "grad_norm": 0.4141116738319397, + "learning_rate": 0.00022657159464999888, + "loss": 2.5977, + "step": 5419 + }, + { + "epoch": 0.69376, + "grad_norm": 0.4528082609176636, + "learning_rate": 0.00022639805186984148, + "loss": 2.6344, + "step": 5420 + }, + { + "epoch": 0.693888, + "grad_norm": 0.3947140574455261, + "learning_rate": 0.00022622455612503062, + "loss": 2.5344, + "step": 5421 + }, + { + "epoch": 0.694016, + "grad_norm": 0.3789835274219513, + "learning_rate": 0.00022605110744539265, + "loss": 2.5926, + "step": 5422 + }, + { + "epoch": 0.694144, + "grad_norm": 0.38878387212753296, + "learning_rate": 0.00022587770586074519, + "loss": 2.6823, + "step": 5423 + }, + { + "epoch": 0.694272, + "grad_norm": 0.3979751765727997, + "learning_rate": 0.0002257043514008978, + "loss": 2.546, + "step": 5424 + }, + { + "epoch": 0.6944, + "grad_norm": 0.48152676224708557, + "learning_rate": 0.0002255310440956525, + "loss": 2.7112, + "step": 5425 + }, + { + "epoch": 0.694528, + "grad_norm": 0.3841072916984558, + "learning_rate": 0.00022535778397480262, + "loss": 2.6187, + "step": 5426 + }, + { + "epoch": 0.694656, + "grad_norm": 0.3617481291294098, + "learning_rate": 0.0002251845710681336, + "loss": 2.5459, + "step": 5427 + }, + { + "epoch": 0.694784, + "grad_norm": 0.42861950397491455, + "learning_rate": 0.0002250114054054228, + "loss": 2.4651, + "step": 5428 + }, + { + "epoch": 0.694912, + "grad_norm": 0.9192508459091187, + "learning_rate": 0.00022483828701643933, + "loss": 2.7222, + "step": 5429 + }, + { + "epoch": 0.69504, + "grad_norm": 0.40111014246940613, + "learning_rate": 0.00022466521593094403, + "loss": 2.6134, + "step": 5430 + }, + { + "epoch": 0.695168, + "grad_norm": 0.38886722922325134, + "learning_rate": 0.0002244921921786901, + "loss": 2.6046, + "step": 5431 + }, + { + "epoch": 0.695296, + "grad_norm": 0.3768211007118225, + "learning_rate": 0.0002243192157894225, + "loss": 2.5403, + "step": 5432 + }, + { + "epoch": 0.695424, + "grad_norm": 0.4512787163257599, + "learning_rate": 0.00022414628679287769, + "loss": 2.7581, + "step": 5433 + }, + { + "epoch": 0.695552, + "grad_norm": 0.37775588035583496, + "learning_rate": 0.00022397340521878412, + "loss": 2.537, + "step": 5434 + }, + { + "epoch": 0.69568, + "grad_norm": 0.3766360878944397, + "learning_rate": 0.00022380057109686215, + "loss": 2.5131, + "step": 5435 + }, + { + "epoch": 0.695808, + "grad_norm": 0.43872085213661194, + "learning_rate": 0.00022362778445682402, + "loss": 2.6016, + "step": 5436 + }, + { + "epoch": 0.695936, + "grad_norm": 0.3957422971725464, + "learning_rate": 0.00022345504532837368, + "loss": 2.6298, + "step": 5437 + }, + { + "epoch": 0.696064, + "grad_norm": 0.4077834486961365, + "learning_rate": 0.00022328235374120692, + "loss": 2.6457, + "step": 5438 + }, + { + "epoch": 0.696192, + "grad_norm": 0.3739756941795349, + "learning_rate": 0.0002231097097250115, + "loss": 2.5547, + "step": 5439 + }, + { + "epoch": 0.69632, + "grad_norm": 0.42328914999961853, + "learning_rate": 0.0002229371133094671, + "loss": 2.6809, + "step": 5440 + }, + { + "epoch": 0.696448, + "grad_norm": 0.4241255521774292, + "learning_rate": 0.00022276456452424497, + "loss": 2.6785, + "step": 5441 + }, + { + "epoch": 0.696576, + "grad_norm": 0.41238391399383545, + "learning_rate": 0.00022259206339900818, + "loss": 2.6598, + "step": 5442 + }, + { + "epoch": 0.696704, + "grad_norm": 0.38952186703681946, + "learning_rate": 0.00022241960996341166, + "loss": 2.6177, + "step": 5443 + }, + { + "epoch": 0.696832, + "grad_norm": 0.3800019919872284, + "learning_rate": 0.0002222472042471022, + "loss": 2.6473, + "step": 5444 + }, + { + "epoch": 0.69696, + "grad_norm": 0.4116501212120056, + "learning_rate": 0.00022207484627971813, + "loss": 2.675, + "step": 5445 + }, + { + "epoch": 0.697088, + "grad_norm": 0.9280096888542175, + "learning_rate": 0.0002219025360908901, + "loss": 2.507, + "step": 5446 + }, + { + "epoch": 0.697216, + "grad_norm": 0.38703256845474243, + "learning_rate": 0.0002217302737102402, + "loss": 2.7521, + "step": 5447 + }, + { + "epoch": 0.697344, + "grad_norm": 0.4286862015724182, + "learning_rate": 0.00022155805916738213, + "loss": 2.5844, + "step": 5448 + }, + { + "epoch": 0.697472, + "grad_norm": 0.398480623960495, + "learning_rate": 0.00022138589249192154, + "loss": 2.5593, + "step": 5449 + }, + { + "epoch": 0.6976, + "grad_norm": 0.394742488861084, + "learning_rate": 0.0002212137737134562, + "loss": 2.613, + "step": 5450 + }, + { + "epoch": 0.697728, + "grad_norm": 0.3776836395263672, + "learning_rate": 0.00022104170286157504, + "loss": 2.567, + "step": 5451 + }, + { + "epoch": 0.697856, + "grad_norm": 0.3957948684692383, + "learning_rate": 0.00022086967996585898, + "loss": 2.5273, + "step": 5452 + }, + { + "epoch": 0.697984, + "grad_norm": 0.48328065872192383, + "learning_rate": 0.00022069770505588111, + "loss": 2.5711, + "step": 5453 + }, + { + "epoch": 0.698112, + "grad_norm": 0.3995787501335144, + "learning_rate": 0.00022052577816120567, + "loss": 2.578, + "step": 5454 + }, + { + "epoch": 0.69824, + "grad_norm": 0.4803442060947418, + "learning_rate": 0.00022035389931138893, + "loss": 2.6084, + "step": 5455 + }, + { + "epoch": 0.698368, + "grad_norm": 0.3923967480659485, + "learning_rate": 0.00022018206853597879, + "loss": 2.6479, + "step": 5456 + }, + { + "epoch": 0.698496, + "grad_norm": 0.3931106925010681, + "learning_rate": 0.000220010285864515, + "loss": 2.6342, + "step": 5457 + }, + { + "epoch": 0.698624, + "grad_norm": 0.44224533438682556, + "learning_rate": 0.00021983855132652887, + "loss": 2.54, + "step": 5458 + }, + { + "epoch": 0.698752, + "grad_norm": 0.41962993144989014, + "learning_rate": 0.00021966686495154375, + "loss": 2.6264, + "step": 5459 + }, + { + "epoch": 0.69888, + "grad_norm": 0.4201401174068451, + "learning_rate": 0.00021949522676907436, + "loss": 2.5148, + "step": 5460 + }, + { + "epoch": 0.699008, + "grad_norm": 0.4217771887779236, + "learning_rate": 0.0002193236368086275, + "loss": 2.6797, + "step": 5461 + }, + { + "epoch": 0.699136, + "grad_norm": 0.40869107842445374, + "learning_rate": 0.0002191520950997014, + "loss": 2.6766, + "step": 5462 + }, + { + "epoch": 0.699264, + "grad_norm": 0.4000718295574188, + "learning_rate": 0.000218980601671786, + "loss": 2.5849, + "step": 5463 + }, + { + "epoch": 0.699392, + "grad_norm": 0.38915130496025085, + "learning_rate": 0.00021880915655436302, + "loss": 2.5441, + "step": 5464 + }, + { + "epoch": 0.69952, + "grad_norm": 0.3986451327800751, + "learning_rate": 0.00021863775977690587, + "loss": 2.6101, + "step": 5465 + }, + { + "epoch": 0.699648, + "grad_norm": 0.4235024154186249, + "learning_rate": 0.00021846641136887945, + "loss": 2.5997, + "step": 5466 + }, + { + "epoch": 0.699776, + "grad_norm": 0.3871089816093445, + "learning_rate": 0.0002182951113597408, + "loss": 2.6241, + "step": 5467 + }, + { + "epoch": 0.699904, + "grad_norm": 0.4296649098396301, + "learning_rate": 0.00021812385977893845, + "loss": 2.6116, + "step": 5468 + }, + { + "epoch": 0.700032, + "grad_norm": 0.39932528138160706, + "learning_rate": 0.00021795265665591236, + "loss": 2.5659, + "step": 5469 + }, + { + "epoch": 0.70016, + "grad_norm": 0.48799699544906616, + "learning_rate": 0.0002177815020200944, + "loss": 2.5983, + "step": 5470 + }, + { + "epoch": 0.700288, + "grad_norm": 0.3850672245025635, + "learning_rate": 0.00021761039590090798, + "loss": 2.5713, + "step": 5471 + }, + { + "epoch": 0.700416, + "grad_norm": 0.3760695457458496, + "learning_rate": 0.0002174393383277683, + "loss": 2.6012, + "step": 5472 + }, + { + "epoch": 0.700544, + "grad_norm": 0.39714059233665466, + "learning_rate": 0.00021726832933008185, + "loss": 2.6549, + "step": 5473 + }, + { + "epoch": 0.700672, + "grad_norm": 0.3993338346481323, + "learning_rate": 0.00021709736893724745, + "loss": 2.6935, + "step": 5474 + }, + { + "epoch": 0.7008, + "grad_norm": 0.39300891757011414, + "learning_rate": 0.0002169264571786551, + "loss": 2.5123, + "step": 5475 + }, + { + "epoch": 0.700928, + "grad_norm": 0.40830352902412415, + "learning_rate": 0.00021675559408368622, + "loss": 2.6155, + "step": 5476 + }, + { + "epoch": 0.701056, + "grad_norm": 0.4105035662651062, + "learning_rate": 0.00021658477968171453, + "loss": 2.6121, + "step": 5477 + }, + { + "epoch": 0.701184, + "grad_norm": 0.41517171263694763, + "learning_rate": 0.00021641401400210487, + "loss": 2.5938, + "step": 5478 + }, + { + "epoch": 0.701312, + "grad_norm": 0.3771858811378479, + "learning_rate": 0.00021624329707421375, + "loss": 2.5264, + "step": 5479 + }, + { + "epoch": 0.70144, + "grad_norm": 0.46912243962287903, + "learning_rate": 0.00021607262892738954, + "loss": 2.5928, + "step": 5480 + }, + { + "epoch": 0.701568, + "grad_norm": 0.3913196921348572, + "learning_rate": 0.00021590200959097173, + "loss": 2.6008, + "step": 5481 + }, + { + "epoch": 0.701696, + "grad_norm": 0.39126673340797424, + "learning_rate": 0.00021573143909429227, + "loss": 2.5918, + "step": 5482 + }, + { + "epoch": 0.701824, + "grad_norm": 0.374636709690094, + "learning_rate": 0.0002155609174666739, + "loss": 2.5807, + "step": 5483 + }, + { + "epoch": 0.701952, + "grad_norm": 0.43187135457992554, + "learning_rate": 0.00021539044473743135, + "loss": 2.5294, + "step": 5484 + }, + { + "epoch": 0.70208, + "grad_norm": 0.44585877656936646, + "learning_rate": 0.00021522002093587078, + "loss": 2.6399, + "step": 5485 + }, + { + "epoch": 0.702208, + "grad_norm": 0.4281296730041504, + "learning_rate": 0.0002150496460912899, + "loss": 2.6051, + "step": 5486 + }, + { + "epoch": 0.702336, + "grad_norm": 0.40635374188423157, + "learning_rate": 0.00021487932023297852, + "loss": 2.5637, + "step": 5487 + }, + { + "epoch": 0.702464, + "grad_norm": 0.4660395681858063, + "learning_rate": 0.0002147090433902172, + "loss": 2.5995, + "step": 5488 + }, + { + "epoch": 0.702592, + "grad_norm": 0.3930402100086212, + "learning_rate": 0.00021453881559227884, + "loss": 2.5647, + "step": 5489 + }, + { + "epoch": 0.70272, + "grad_norm": 0.3938429355621338, + "learning_rate": 0.0002143686368684274, + "loss": 2.6719, + "step": 5490 + }, + { + "epoch": 0.702848, + "grad_norm": 0.41851919889450073, + "learning_rate": 0.00021419850724791866, + "loss": 2.5319, + "step": 5491 + }, + { + "epoch": 0.702976, + "grad_norm": 0.3934067189693451, + "learning_rate": 0.0002140284267599998, + "loss": 2.604, + "step": 5492 + }, + { + "epoch": 0.703104, + "grad_norm": 0.3869519531726837, + "learning_rate": 0.00021385839543390967, + "loss": 2.5261, + "step": 5493 + }, + { + "epoch": 0.703232, + "grad_norm": 0.4350231885910034, + "learning_rate": 0.0002136884132988784, + "loss": 2.6785, + "step": 5494 + }, + { + "epoch": 0.70336, + "grad_norm": 0.42622247338294983, + "learning_rate": 0.0002135184803841283, + "loss": 2.6245, + "step": 5495 + }, + { + "epoch": 0.703488, + "grad_norm": 0.41607335209846497, + "learning_rate": 0.00021334859671887234, + "loss": 2.7217, + "step": 5496 + }, + { + "epoch": 0.703616, + "grad_norm": 0.40987446904182434, + "learning_rate": 0.00021317876233231605, + "loss": 2.6752, + "step": 5497 + }, + { + "epoch": 0.703744, + "grad_norm": 0.38541772961616516, + "learning_rate": 0.00021300897725365554, + "loss": 2.681, + "step": 5498 + }, + { + "epoch": 0.703872, + "grad_norm": 0.4118821322917938, + "learning_rate": 0.00021283924151207896, + "loss": 2.4887, + "step": 5499 + }, + { + "epoch": 0.704, + "grad_norm": 0.41327235102653503, + "learning_rate": 0.00021266955513676584, + "loss": 2.5716, + "step": 5500 + }, + { + "epoch": 0.704128, + "grad_norm": 0.4164629876613617, + "learning_rate": 0.0002124999181568872, + "loss": 2.65, + "step": 5501 + }, + { + "epoch": 0.704256, + "grad_norm": 0.39222902059555054, + "learning_rate": 0.00021233033060160539, + "loss": 2.5934, + "step": 5502 + }, + { + "epoch": 0.704384, + "grad_norm": 0.3694218397140503, + "learning_rate": 0.00021216079250007492, + "loss": 2.5362, + "step": 5503 + }, + { + "epoch": 0.704512, + "grad_norm": 0.4683862328529358, + "learning_rate": 0.00021199130388144095, + "loss": 2.6602, + "step": 5504 + }, + { + "epoch": 0.70464, + "grad_norm": 0.3857412338256836, + "learning_rate": 0.00021182186477484093, + "loss": 2.5693, + "step": 5505 + }, + { + "epoch": 0.704768, + "grad_norm": 0.39681416749954224, + "learning_rate": 0.00021165247520940319, + "loss": 2.6732, + "step": 5506 + }, + { + "epoch": 0.704896, + "grad_norm": 0.48928868770599365, + "learning_rate": 0.00021148313521424778, + "loss": 2.5695, + "step": 5507 + }, + { + "epoch": 0.705024, + "grad_norm": 0.3959125578403473, + "learning_rate": 0.00021131384481848614, + "loss": 2.6149, + "step": 5508 + }, + { + "epoch": 0.705152, + "grad_norm": 0.7935575842857361, + "learning_rate": 0.0002111446040512212, + "loss": 2.6225, + "step": 5509 + }, + { + "epoch": 0.70528, + "grad_norm": 0.3939632475376129, + "learning_rate": 0.0002109754129415477, + "loss": 2.5438, + "step": 5510 + }, + { + "epoch": 0.705408, + "grad_norm": 0.3814971148967743, + "learning_rate": 0.00021080627151855136, + "loss": 2.5835, + "step": 5511 + }, + { + "epoch": 0.705536, + "grad_norm": 0.38651588559150696, + "learning_rate": 0.00021063717981130954, + "loss": 2.5658, + "step": 5512 + }, + { + "epoch": 0.705664, + "grad_norm": 0.41764622926712036, + "learning_rate": 0.00021046813784889097, + "loss": 2.6195, + "step": 5513 + }, + { + "epoch": 0.705792, + "grad_norm": 0.35676059126853943, + "learning_rate": 0.00021029914566035617, + "loss": 2.4956, + "step": 5514 + }, + { + "epoch": 0.70592, + "grad_norm": 0.40451115369796753, + "learning_rate": 0.0002101302032747568, + "loss": 2.599, + "step": 5515 + }, + { + "epoch": 0.706048, + "grad_norm": 0.38899508118629456, + "learning_rate": 0.00020996131072113588, + "loss": 2.6937, + "step": 5516 + }, + { + "epoch": 0.706176, + "grad_norm": 0.3853878676891327, + "learning_rate": 0.00020979246802852792, + "loss": 2.59, + "step": 5517 + }, + { + "epoch": 0.706304, + "grad_norm": 0.38784313201904297, + "learning_rate": 0.00020962367522595922, + "loss": 2.6654, + "step": 5518 + }, + { + "epoch": 0.706432, + "grad_norm": 0.4120856523513794, + "learning_rate": 0.00020945493234244712, + "loss": 2.6106, + "step": 5519 + }, + { + "epoch": 0.70656, + "grad_norm": 0.38091906905174255, + "learning_rate": 0.00020928623940700042, + "loss": 2.4365, + "step": 5520 + }, + { + "epoch": 0.706688, + "grad_norm": 0.4012247622013092, + "learning_rate": 0.00020911759644861937, + "loss": 2.6035, + "step": 5521 + }, + { + "epoch": 0.706816, + "grad_norm": 0.3840239644050598, + "learning_rate": 0.00020894900349629575, + "loss": 2.5996, + "step": 5522 + }, + { + "epoch": 0.706944, + "grad_norm": 0.4150223731994629, + "learning_rate": 0.0002087804605790124, + "loss": 2.5449, + "step": 5523 + }, + { + "epoch": 0.707072, + "grad_norm": 0.43391430377960205, + "learning_rate": 0.00020861196772574398, + "loss": 2.4888, + "step": 5524 + }, + { + "epoch": 0.7072, + "grad_norm": 0.39689671993255615, + "learning_rate": 0.0002084435249654565, + "loss": 2.6373, + "step": 5525 + }, + { + "epoch": 0.707328, + "grad_norm": 0.41512566804885864, + "learning_rate": 0.00020827513232710715, + "loss": 2.6673, + "step": 5526 + }, + { + "epoch": 0.707456, + "grad_norm": 0.41904377937316895, + "learning_rate": 0.00020810678983964448, + "loss": 2.5517, + "step": 5527 + }, + { + "epoch": 0.707584, + "grad_norm": 0.39016464352607727, + "learning_rate": 0.00020793849753200855, + "loss": 2.6759, + "step": 5528 + }, + { + "epoch": 0.707712, + "grad_norm": 0.4268808364868164, + "learning_rate": 0.00020777025543313078, + "loss": 2.6041, + "step": 5529 + }, + { + "epoch": 0.70784, + "grad_norm": 0.4542088508605957, + "learning_rate": 0.00020760206357193372, + "loss": 2.6462, + "step": 5530 + }, + { + "epoch": 0.707968, + "grad_norm": 0.41179972887039185, + "learning_rate": 0.00020743392197733191, + "loss": 2.6469, + "step": 5531 + }, + { + "epoch": 0.708096, + "grad_norm": 0.4247884154319763, + "learning_rate": 0.00020726583067823045, + "loss": 2.6291, + "step": 5532 + }, + { + "epoch": 0.708224, + "grad_norm": 0.3888337016105652, + "learning_rate": 0.00020709778970352656, + "loss": 2.6064, + "step": 5533 + }, + { + "epoch": 0.708352, + "grad_norm": 0.3989221453666687, + "learning_rate": 0.0002069297990821082, + "loss": 2.5819, + "step": 5534 + }, + { + "epoch": 0.70848, + "grad_norm": 0.4205979108810425, + "learning_rate": 0.00020676185884285493, + "loss": 2.6795, + "step": 5535 + }, + { + "epoch": 0.708608, + "grad_norm": 0.409648060798645, + "learning_rate": 0.0002065939690146376, + "loss": 2.5969, + "step": 5536 + }, + { + "epoch": 0.708736, + "grad_norm": 0.37926360964775085, + "learning_rate": 0.00020642612962631853, + "loss": 2.6182, + "step": 5537 + }, + { + "epoch": 0.708864, + "grad_norm": 0.5333307385444641, + "learning_rate": 0.00020625834070675091, + "loss": 2.6588, + "step": 5538 + }, + { + "epoch": 0.708992, + "grad_norm": 0.38575634360313416, + "learning_rate": 0.00020609060228478015, + "loss": 2.6163, + "step": 5539 + }, + { + "epoch": 0.70912, + "grad_norm": 0.36855223774909973, + "learning_rate": 0.0002059229143892421, + "loss": 2.5242, + "step": 5540 + }, + { + "epoch": 0.709248, + "grad_norm": 0.4018947184085846, + "learning_rate": 0.00020575527704896413, + "loss": 2.6062, + "step": 5541 + }, + { + "epoch": 0.709376, + "grad_norm": 0.3912242650985718, + "learning_rate": 0.00020558769029276535, + "loss": 2.5432, + "step": 5542 + }, + { + "epoch": 0.709504, + "grad_norm": 0.4214605391025543, + "learning_rate": 0.00020542015414945575, + "loss": 2.6268, + "step": 5543 + }, + { + "epoch": 0.709632, + "grad_norm": 0.4195290803909302, + "learning_rate": 0.00020525266864783676, + "loss": 2.5573, + "step": 5544 + }, + { + "epoch": 0.70976, + "grad_norm": 0.39769142866134644, + "learning_rate": 0.0002050852338167008, + "loss": 2.569, + "step": 5545 + }, + { + "epoch": 0.709888, + "grad_norm": 0.44172781705856323, + "learning_rate": 0.0002049178496848323, + "loss": 2.5765, + "step": 5546 + }, + { + "epoch": 0.710016, + "grad_norm": 0.5012351870536804, + "learning_rate": 0.00020475051628100633, + "loss": 2.5798, + "step": 5547 + }, + { + "epoch": 0.710144, + "grad_norm": 0.40907061100006104, + "learning_rate": 0.0002045832336339894, + "loss": 2.5153, + "step": 5548 + }, + { + "epoch": 0.710272, + "grad_norm": 0.5575770735740662, + "learning_rate": 0.00020441600177253938, + "loss": 2.56, + "step": 5549 + }, + { + "epoch": 0.7104, + "grad_norm": 0.4388718008995056, + "learning_rate": 0.00020424882072540542, + "loss": 2.6304, + "step": 5550 + }, + { + "epoch": 0.710528, + "grad_norm": 0.40956273674964905, + "learning_rate": 0.0002040816905213276, + "loss": 2.5636, + "step": 5551 + }, + { + "epoch": 0.710656, + "grad_norm": 0.46992728114128113, + "learning_rate": 0.00020391461118903786, + "loss": 2.5472, + "step": 5552 + }, + { + "epoch": 0.710784, + "grad_norm": 0.39368072152137756, + "learning_rate": 0.0002037475827572589, + "loss": 2.5555, + "step": 5553 + }, + { + "epoch": 0.710912, + "grad_norm": 0.43890872597694397, + "learning_rate": 0.00020358060525470508, + "loss": 2.6006, + "step": 5554 + }, + { + "epoch": 0.71104, + "grad_norm": 0.4829445481300354, + "learning_rate": 0.00020341367871008154, + "loss": 2.641, + "step": 5555 + }, + { + "epoch": 0.711168, + "grad_norm": 0.39020419120788574, + "learning_rate": 0.00020324680315208504, + "loss": 2.5591, + "step": 5556 + }, + { + "epoch": 0.711296, + "grad_norm": 0.43189215660095215, + "learning_rate": 0.0002030799786094033, + "loss": 2.5776, + "step": 5557 + }, + { + "epoch": 0.711424, + "grad_norm": 0.41074737906455994, + "learning_rate": 0.00020291320511071543, + "loss": 2.6391, + "step": 5558 + }, + { + "epoch": 0.711552, + "grad_norm": 0.3691082000732422, + "learning_rate": 0.00020274648268469155, + "loss": 2.6767, + "step": 5559 + }, + { + "epoch": 0.71168, + "grad_norm": 0.4347698390483856, + "learning_rate": 0.0002025798113599933, + "loss": 2.5458, + "step": 5560 + }, + { + "epoch": 0.711808, + "grad_norm": 0.3984563946723938, + "learning_rate": 0.00020241319116527374, + "loss": 2.6458, + "step": 5561 + }, + { + "epoch": 0.711936, + "grad_norm": 0.39710745215415955, + "learning_rate": 0.0002022466221291765, + "loss": 2.5612, + "step": 5562 + }, + { + "epoch": 0.712064, + "grad_norm": 0.9772101044654846, + "learning_rate": 0.00020208010428033674, + "loss": 2.5926, + "step": 5563 + }, + { + "epoch": 0.712192, + "grad_norm": 0.3991886377334595, + "learning_rate": 0.00020191363764738085, + "loss": 2.7418, + "step": 5564 + }, + { + "epoch": 0.71232, + "grad_norm": 0.4166642725467682, + "learning_rate": 0.00020174722225892638, + "loss": 2.603, + "step": 5565 + }, + { + "epoch": 0.712448, + "grad_norm": 0.46654224395751953, + "learning_rate": 0.00020158085814358183, + "loss": 2.6144, + "step": 5566 + }, + { + "epoch": 0.712576, + "grad_norm": 0.3992750644683838, + "learning_rate": 0.0002014145453299476, + "loss": 2.6323, + "step": 5567 + }, + { + "epoch": 0.712704, + "grad_norm": 0.4025958180427551, + "learning_rate": 0.0002012482838466145, + "loss": 2.6316, + "step": 5568 + }, + { + "epoch": 0.712832, + "grad_norm": 0.4456075131893158, + "learning_rate": 0.00020108207372216457, + "loss": 2.6215, + "step": 5569 + }, + { + "epoch": 0.71296, + "grad_norm": 0.3915743827819824, + "learning_rate": 0.0002009159149851718, + "loss": 2.5834, + "step": 5570 + }, + { + "epoch": 0.713088, + "grad_norm": 0.5044238567352295, + "learning_rate": 0.00020074980766420048, + "loss": 2.6884, + "step": 5571 + }, + { + "epoch": 0.713216, + "grad_norm": 0.4196140468120575, + "learning_rate": 0.0002005837517878064, + "loss": 2.6694, + "step": 5572 + }, + { + "epoch": 0.713344, + "grad_norm": 0.4449235200881958, + "learning_rate": 0.0002004177473845366, + "loss": 2.5741, + "step": 5573 + }, + { + "epoch": 0.713472, + "grad_norm": 0.4125233292579651, + "learning_rate": 0.00020025179448292886, + "loss": 2.6088, + "step": 5574 + }, + { + "epoch": 0.7136, + "grad_norm": 0.5051617622375488, + "learning_rate": 0.00020008589311151277, + "loss": 2.6377, + "step": 5575 + }, + { + "epoch": 0.713728, + "grad_norm": 0.4317322373390198, + "learning_rate": 0.00019992004329880857, + "loss": 2.718, + "step": 5576 + }, + { + "epoch": 0.713856, + "grad_norm": 0.41306787729263306, + "learning_rate": 0.0001997542450733278, + "loss": 2.6491, + "step": 5577 + }, + { + "epoch": 0.713984, + "grad_norm": 0.39971819519996643, + "learning_rate": 0.00019958849846357285, + "loss": 2.5911, + "step": 5578 + }, + { + "epoch": 0.714112, + "grad_norm": 0.3949612081050873, + "learning_rate": 0.00019942280349803781, + "loss": 2.5824, + "step": 5579 + }, + { + "epoch": 0.71424, + "grad_norm": 0.41206273436546326, + "learning_rate": 0.00019925716020520752, + "loss": 2.5354, + "step": 5580 + }, + { + "epoch": 0.714368, + "grad_norm": 0.3945072889328003, + "learning_rate": 0.00019909156861355764, + "loss": 2.615, + "step": 5581 + }, + { + "epoch": 0.714496, + "grad_norm": 0.4279208481311798, + "learning_rate": 0.0001989260287515558, + "loss": 2.6926, + "step": 5582 + }, + { + "epoch": 0.714624, + "grad_norm": 0.46780747175216675, + "learning_rate": 0.00019876054064765992, + "loss": 2.6862, + "step": 5583 + }, + { + "epoch": 0.714752, + "grad_norm": 0.4388517439365387, + "learning_rate": 0.0001985951043303194, + "loss": 2.6777, + "step": 5584 + }, + { + "epoch": 0.71488, + "grad_norm": 0.5995991230010986, + "learning_rate": 0.0001984297198279746, + "loss": 2.6277, + "step": 5585 + }, + { + "epoch": 0.715008, + "grad_norm": 0.4060506820678711, + "learning_rate": 0.0001982643871690571, + "loss": 2.6374, + "step": 5586 + }, + { + "epoch": 0.715136, + "grad_norm": 0.4729832410812378, + "learning_rate": 0.00019809910638198924, + "loss": 2.4946, + "step": 5587 + }, + { + "epoch": 0.715264, + "grad_norm": 0.41935276985168457, + "learning_rate": 0.00019793387749518515, + "loss": 2.6775, + "step": 5588 + }, + { + "epoch": 0.715392, + "grad_norm": 0.42137959599494934, + "learning_rate": 0.00019776870053704915, + "loss": 2.5806, + "step": 5589 + }, + { + "epoch": 0.71552, + "grad_norm": 0.4164171814918518, + "learning_rate": 0.00019760357553597752, + "loss": 2.5961, + "step": 5590 + }, + { + "epoch": 0.715648, + "grad_norm": 0.42261162400245667, + "learning_rate": 0.0001974385025203569, + "loss": 2.4958, + "step": 5591 + }, + { + "epoch": 0.715776, + "grad_norm": 0.46748414635658264, + "learning_rate": 0.00019727348151856534, + "loss": 2.4816, + "step": 5592 + }, + { + "epoch": 0.715904, + "grad_norm": 0.40825700759887695, + "learning_rate": 0.00019710851255897173, + "loss": 2.703, + "step": 5593 + }, + { + "epoch": 0.716032, + "grad_norm": 0.5003231167793274, + "learning_rate": 0.0001969435956699363, + "loss": 2.5619, + "step": 5594 + }, + { + "epoch": 0.71616, + "grad_norm": 0.4404820501804352, + "learning_rate": 0.0001967787308798099, + "loss": 2.6521, + "step": 5595 + }, + { + "epoch": 0.716288, + "grad_norm": 0.40609708428382874, + "learning_rate": 0.00019661391821693507, + "loss": 2.5822, + "step": 5596 + }, + { + "epoch": 0.716416, + "grad_norm": 0.44542622566223145, + "learning_rate": 0.00019644915770964473, + "loss": 2.5777, + "step": 5597 + }, + { + "epoch": 0.716544, + "grad_norm": 0.662138044834137, + "learning_rate": 0.00019628444938626333, + "loss": 2.5541, + "step": 5598 + }, + { + "epoch": 0.716672, + "grad_norm": 0.36510270833969116, + "learning_rate": 0.00019611979327510616, + "loss": 2.6328, + "step": 5599 + }, + { + "epoch": 0.7168, + "grad_norm": 0.3970123529434204, + "learning_rate": 0.00019595518940447931, + "loss": 2.6299, + "step": 5600 + }, + { + "epoch": 0.716928, + "grad_norm": 0.42104703187942505, + "learning_rate": 0.00019579063780268024, + "loss": 2.5933, + "step": 5601 + }, + { + "epoch": 0.717056, + "grad_norm": 0.5142852663993835, + "learning_rate": 0.00019562613849799705, + "loss": 2.6275, + "step": 5602 + }, + { + "epoch": 0.717184, + "grad_norm": 0.41391414403915405, + "learning_rate": 0.00019546169151870942, + "loss": 2.4686, + "step": 5603 + }, + { + "epoch": 0.717312, + "grad_norm": 0.4186552166938782, + "learning_rate": 0.00019529729689308757, + "loss": 2.6479, + "step": 5604 + }, + { + "epoch": 0.71744, + "grad_norm": 0.42728474736213684, + "learning_rate": 0.00019513295464939273, + "loss": 2.4977, + "step": 5605 + }, + { + "epoch": 0.717568, + "grad_norm": 0.4068087339401245, + "learning_rate": 0.00019496866481587716, + "loss": 2.7216, + "step": 5606 + }, + { + "epoch": 0.717696, + "grad_norm": 0.4317927062511444, + "learning_rate": 0.00019480442742078453, + "loss": 2.6277, + "step": 5607 + }, + { + "epoch": 0.717824, + "grad_norm": 0.40240705013275146, + "learning_rate": 0.00019464024249234897, + "loss": 2.6349, + "step": 5608 + }, + { + "epoch": 0.717952, + "grad_norm": 0.4521007537841797, + "learning_rate": 0.00019447611005879572, + "loss": 2.6806, + "step": 5609 + }, + { + "epoch": 0.71808, + "grad_norm": 1.056782841682434, + "learning_rate": 0.00019431203014834092, + "loss": 2.6328, + "step": 5610 + }, + { + "epoch": 0.718208, + "grad_norm": 0.3850770890712738, + "learning_rate": 0.0001941480027891922, + "loss": 2.5261, + "step": 5611 + }, + { + "epoch": 0.718336, + "grad_norm": 0.948060929775238, + "learning_rate": 0.00019398402800954745, + "loss": 2.5638, + "step": 5612 + }, + { + "epoch": 0.718464, + "grad_norm": 0.40895116329193115, + "learning_rate": 0.00019382010583759602, + "loss": 2.5846, + "step": 5613 + }, + { + "epoch": 0.718592, + "grad_norm": 0.4134816527366638, + "learning_rate": 0.0001936562363015179, + "loss": 2.5271, + "step": 5614 + }, + { + "epoch": 0.71872, + "grad_norm": 0.3856849670410156, + "learning_rate": 0.00019349241942948403, + "loss": 2.4619, + "step": 5615 + }, + { + "epoch": 0.718848, + "grad_norm": 0.4369450807571411, + "learning_rate": 0.00019332865524965677, + "loss": 2.6151, + "step": 5616 + }, + { + "epoch": 0.718976, + "grad_norm": 0.43936192989349365, + "learning_rate": 0.00019316494379018872, + "loss": 2.5229, + "step": 5617 + }, + { + "epoch": 0.719104, + "grad_norm": 0.4000258147716522, + "learning_rate": 0.00019300128507922416, + "loss": 2.6935, + "step": 5618 + }, + { + "epoch": 0.719232, + "grad_norm": 0.3972648084163666, + "learning_rate": 0.00019283767914489775, + "loss": 2.5606, + "step": 5619 + }, + { + "epoch": 0.71936, + "grad_norm": 0.3830040693283081, + "learning_rate": 0.00019267412601533518, + "loss": 2.7268, + "step": 5620 + }, + { + "epoch": 0.719488, + "grad_norm": 0.43260741233825684, + "learning_rate": 0.0001925106257186532, + "loss": 2.6711, + "step": 5621 + }, + { + "epoch": 0.719616, + "grad_norm": 0.4471967816352844, + "learning_rate": 0.00019234717828295938, + "loss": 2.559, + "step": 5622 + }, + { + "epoch": 0.719744, + "grad_norm": 0.40191662311553955, + "learning_rate": 0.0001921837837363521, + "loss": 2.5531, + "step": 5623 + }, + { + "epoch": 0.719872, + "grad_norm": 0.43488001823425293, + "learning_rate": 0.000192020442106921, + "loss": 2.6359, + "step": 5624 + }, + { + "epoch": 0.72, + "grad_norm": 0.41912841796875, + "learning_rate": 0.0001918571534227462, + "loss": 2.6981, + "step": 5625 + }, + { + "epoch": 0.720128, + "grad_norm": 0.55359947681427, + "learning_rate": 0.00019169391771189914, + "loss": 2.6652, + "step": 5626 + }, + { + "epoch": 0.720256, + "grad_norm": 0.3634112477302551, + "learning_rate": 0.0001915307350024419, + "loss": 2.5585, + "step": 5627 + }, + { + "epoch": 0.720384, + "grad_norm": 0.4926748275756836, + "learning_rate": 0.00019136760532242731, + "loss": 2.5337, + "step": 5628 + }, + { + "epoch": 0.720512, + "grad_norm": 0.38539978861808777, + "learning_rate": 0.0001912045286998994, + "loss": 2.5485, + "step": 5629 + }, + { + "epoch": 0.72064, + "grad_norm": 0.37538382411003113, + "learning_rate": 0.0001910415051628928, + "loss": 2.6324, + "step": 5630 + }, + { + "epoch": 0.720768, + "grad_norm": 0.3632480502128601, + "learning_rate": 0.0001908785347394331, + "loss": 2.5602, + "step": 5631 + }, + { + "epoch": 0.720896, + "grad_norm": 0.4019763767719269, + "learning_rate": 0.00019071561745753713, + "loss": 2.6572, + "step": 5632 + }, + { + "epoch": 0.721024, + "grad_norm": 0.4033469259738922, + "learning_rate": 0.000190552753345212, + "loss": 2.6117, + "step": 5633 + }, + { + "epoch": 0.721152, + "grad_norm": 0.41112446784973145, + "learning_rate": 0.00019038994243045581, + "loss": 2.6412, + "step": 5634 + }, + { + "epoch": 0.72128, + "grad_norm": 0.3935225307941437, + "learning_rate": 0.000190227184741258, + "loss": 2.6828, + "step": 5635 + }, + { + "epoch": 0.721408, + "grad_norm": 0.39504683017730713, + "learning_rate": 0.00019006448030559831, + "loss": 2.5768, + "step": 5636 + }, + { + "epoch": 0.721536, + "grad_norm": 0.3960243761539459, + "learning_rate": 0.0001899018291514476, + "loss": 2.5669, + "step": 5637 + }, + { + "epoch": 0.721664, + "grad_norm": 0.4352892339229584, + "learning_rate": 0.0001897392313067672, + "loss": 2.609, + "step": 5638 + }, + { + "epoch": 0.721792, + "grad_norm": 0.4191034734249115, + "learning_rate": 0.00018957668679950995, + "loss": 2.6131, + "step": 5639 + }, + { + "epoch": 0.72192, + "grad_norm": 0.42938289046287537, + "learning_rate": 0.00018941419565761904, + "loss": 2.5851, + "step": 5640 + }, + { + "epoch": 0.722048, + "grad_norm": 0.3982309401035309, + "learning_rate": 0.0001892517579090285, + "loss": 2.5771, + "step": 5641 + }, + { + "epoch": 0.722176, + "grad_norm": 0.4499538838863373, + "learning_rate": 0.00018908937358166322, + "loss": 2.6724, + "step": 5642 + }, + { + "epoch": 0.722304, + "grad_norm": 0.4155412018299103, + "learning_rate": 0.00018892704270343885, + "loss": 2.477, + "step": 5643 + }, + { + "epoch": 0.722432, + "grad_norm": 0.42175689339637756, + "learning_rate": 0.00018876476530226233, + "loss": 2.5701, + "step": 5644 + }, + { + "epoch": 0.72256, + "grad_norm": 0.43344205617904663, + "learning_rate": 0.0001886025414060306, + "loss": 2.6412, + "step": 5645 + }, + { + "epoch": 0.722688, + "grad_norm": 0.38894182443618774, + "learning_rate": 0.00018844037104263223, + "loss": 2.6842, + "step": 5646 + }, + { + "epoch": 0.722816, + "grad_norm": 0.38698670268058777, + "learning_rate": 0.00018827825423994594, + "loss": 2.6992, + "step": 5647 + }, + { + "epoch": 0.722944, + "grad_norm": 0.4386952817440033, + "learning_rate": 0.00018811619102584155, + "loss": 2.4949, + "step": 5648 + }, + { + "epoch": 0.723072, + "grad_norm": 0.40530478954315186, + "learning_rate": 0.0001879541814281796, + "loss": 2.6814, + "step": 5649 + }, + { + "epoch": 0.7232, + "grad_norm": 0.48911651968955994, + "learning_rate": 0.0001877922254748114, + "loss": 2.6461, + "step": 5650 + }, + { + "epoch": 0.723328, + "grad_norm": 0.43096715211868286, + "learning_rate": 0.0001876303231935791, + "loss": 2.5659, + "step": 5651 + }, + { + "epoch": 0.723456, + "grad_norm": 0.39251187443733215, + "learning_rate": 0.0001874684746123153, + "loss": 2.6942, + "step": 5652 + }, + { + "epoch": 0.723584, + "grad_norm": 0.4394238293170929, + "learning_rate": 0.00018730667975884398, + "loss": 2.5482, + "step": 5653 + }, + { + "epoch": 0.723712, + "grad_norm": 0.4592877924442291, + "learning_rate": 0.00018714493866097955, + "loss": 2.7522, + "step": 5654 + }, + { + "epoch": 0.72384, + "grad_norm": 0.40442290902137756, + "learning_rate": 0.0001869832513465271, + "loss": 2.5738, + "step": 5655 + }, + { + "epoch": 0.723968, + "grad_norm": 0.4325430691242218, + "learning_rate": 0.0001868216178432826, + "loss": 2.6571, + "step": 5656 + }, + { + "epoch": 0.724096, + "grad_norm": 0.3563113808631897, + "learning_rate": 0.00018666003817903267, + "loss": 2.6412, + "step": 5657 + }, + { + "epoch": 0.724224, + "grad_norm": 0.4156321883201599, + "learning_rate": 0.00018649851238155464, + "loss": 2.7536, + "step": 5658 + }, + { + "epoch": 0.724352, + "grad_norm": 0.3868090808391571, + "learning_rate": 0.00018633704047861665, + "loss": 2.4916, + "step": 5659 + }, + { + "epoch": 0.72448, + "grad_norm": 0.3755502700805664, + "learning_rate": 0.00018617562249797787, + "loss": 2.5855, + "step": 5660 + }, + { + "epoch": 0.724608, + "grad_norm": 0.3723786473274231, + "learning_rate": 0.00018601425846738774, + "loss": 2.5612, + "step": 5661 + }, + { + "epoch": 0.724736, + "grad_norm": 0.37210047245025635, + "learning_rate": 0.0001858529484145864, + "loss": 2.4279, + "step": 5662 + }, + { + "epoch": 0.724864, + "grad_norm": 0.41525211930274963, + "learning_rate": 0.00018569169236730532, + "loss": 2.6427, + "step": 5663 + }, + { + "epoch": 0.724992, + "grad_norm": 0.4350430369377136, + "learning_rate": 0.00018553049035326612, + "loss": 2.5436, + "step": 5664 + }, + { + "epoch": 0.72512, + "grad_norm": 0.3811248540878296, + "learning_rate": 0.00018536934240018127, + "loss": 2.6274, + "step": 5665 + }, + { + "epoch": 0.725248, + "grad_norm": 0.38136932253837585, + "learning_rate": 0.0001852082485357538, + "loss": 2.5744, + "step": 5666 + }, + { + "epoch": 0.725376, + "grad_norm": 0.4000515341758728, + "learning_rate": 0.00018504720878767793, + "loss": 2.5736, + "step": 5667 + }, + { + "epoch": 0.725504, + "grad_norm": 0.4402845501899719, + "learning_rate": 0.00018488622318363812, + "loss": 2.6366, + "step": 5668 + }, + { + "epoch": 0.725632, + "grad_norm": 0.37968435883522034, + "learning_rate": 0.0001847252917513097, + "loss": 2.6877, + "step": 5669 + }, + { + "epoch": 0.72576, + "grad_norm": 0.41380131244659424, + "learning_rate": 0.0001845644145183586, + "loss": 2.5676, + "step": 5670 + }, + { + "epoch": 0.725888, + "grad_norm": 0.3909755051136017, + "learning_rate": 0.0001844035915124413, + "loss": 2.7026, + "step": 5671 + }, + { + "epoch": 0.726016, + "grad_norm": 0.40326982736587524, + "learning_rate": 0.00018424282276120545, + "loss": 2.6532, + "step": 5672 + }, + { + "epoch": 0.726144, + "grad_norm": 0.36154472827911377, + "learning_rate": 0.000184082108292289, + "loss": 2.6555, + "step": 5673 + }, + { + "epoch": 0.726272, + "grad_norm": 0.4011732041835785, + "learning_rate": 0.0001839214481333204, + "loss": 2.5294, + "step": 5674 + }, + { + "epoch": 0.7264, + "grad_norm": 0.428805410861969, + "learning_rate": 0.0001837608423119193, + "loss": 2.647, + "step": 5675 + }, + { + "epoch": 0.726528, + "grad_norm": 0.38126882910728455, + "learning_rate": 0.00018360029085569558, + "loss": 2.6066, + "step": 5676 + }, + { + "epoch": 0.726656, + "grad_norm": 0.4033830463886261, + "learning_rate": 0.0001834397937922499, + "loss": 2.684, + "step": 5677 + }, + { + "epoch": 0.726784, + "grad_norm": 0.4001924991607666, + "learning_rate": 0.0001832793511491735, + "loss": 2.515, + "step": 5678 + }, + { + "epoch": 0.726912, + "grad_norm": 0.3668014705181122, + "learning_rate": 0.00018311896295404839, + "loss": 2.5099, + "step": 5679 + }, + { + "epoch": 0.72704, + "grad_norm": 0.36662301421165466, + "learning_rate": 0.000182958629234447, + "loss": 2.6788, + "step": 5680 + }, + { + "epoch": 0.727168, + "grad_norm": 0.3968120813369751, + "learning_rate": 0.0001827983500179327, + "loss": 2.6221, + "step": 5681 + }, + { + "epoch": 0.727296, + "grad_norm": 0.4028300940990448, + "learning_rate": 0.00018263812533205953, + "loss": 2.6334, + "step": 5682 + }, + { + "epoch": 0.727424, + "grad_norm": 0.3940477967262268, + "learning_rate": 0.00018247795520437177, + "loss": 2.686, + "step": 5683 + }, + { + "epoch": 0.727552, + "grad_norm": 0.4261523485183716, + "learning_rate": 0.00018231783966240456, + "loss": 2.6457, + "step": 5684 + }, + { + "epoch": 0.72768, + "grad_norm": 0.3971397280693054, + "learning_rate": 0.00018215777873368361, + "loss": 2.6672, + "step": 5685 + }, + { + "epoch": 0.727808, + "grad_norm": 0.4555475115776062, + "learning_rate": 0.00018199777244572523, + "loss": 2.6066, + "step": 5686 + }, + { + "epoch": 0.727936, + "grad_norm": 0.5278118252754211, + "learning_rate": 0.00018183782082603617, + "loss": 2.5894, + "step": 5687 + }, + { + "epoch": 0.728064, + "grad_norm": 0.371559739112854, + "learning_rate": 0.00018167792390211436, + "loss": 2.5522, + "step": 5688 + }, + { + "epoch": 0.728192, + "grad_norm": 0.4704395830631256, + "learning_rate": 0.0001815180817014475, + "loss": 2.636, + "step": 5689 + }, + { + "epoch": 0.72832, + "grad_norm": 0.4417004883289337, + "learning_rate": 0.0001813582942515148, + "loss": 2.5826, + "step": 5690 + }, + { + "epoch": 0.728448, + "grad_norm": 0.5798066854476929, + "learning_rate": 0.00018119856157978531, + "loss": 2.6713, + "step": 5691 + }, + { + "epoch": 0.728576, + "grad_norm": 0.38684818148612976, + "learning_rate": 0.00018103888371371897, + "loss": 2.5846, + "step": 5692 + }, + { + "epoch": 0.728704, + "grad_norm": 0.42824625968933105, + "learning_rate": 0.0001808792606807662, + "loss": 2.6141, + "step": 5693 + }, + { + "epoch": 0.728832, + "grad_norm": 0.4001918435096741, + "learning_rate": 0.00018071969250836812, + "loss": 2.6822, + "step": 5694 + }, + { + "epoch": 0.72896, + "grad_norm": 0.379299521446228, + "learning_rate": 0.00018056017922395606, + "loss": 2.6225, + "step": 5695 + }, + { + "epoch": 0.729088, + "grad_norm": 0.37631675601005554, + "learning_rate": 0.00018040072085495274, + "loss": 2.6229, + "step": 5696 + }, + { + "epoch": 0.729216, + "grad_norm": 0.37896573543548584, + "learning_rate": 0.00018024131742877054, + "loss": 2.6981, + "step": 5697 + }, + { + "epoch": 0.729344, + "grad_norm": 0.3953535854816437, + "learning_rate": 0.00018008196897281286, + "loss": 2.5447, + "step": 5698 + }, + { + "epoch": 0.729472, + "grad_norm": 0.3987883925437927, + "learning_rate": 0.00017992267551447333, + "loss": 2.6567, + "step": 5699 + }, + { + "epoch": 0.7296, + "grad_norm": 0.36690258979797363, + "learning_rate": 0.00017976343708113674, + "loss": 2.6227, + "step": 5700 + }, + { + "epoch": 0.729728, + "grad_norm": 0.346601277589798, + "learning_rate": 0.0001796042537001778, + "loss": 2.5423, + "step": 5701 + }, + { + "epoch": 0.729856, + "grad_norm": 0.38506993651390076, + "learning_rate": 0.0001794451253989618, + "loss": 2.6838, + "step": 5702 + }, + { + "epoch": 0.729984, + "grad_norm": 0.4124627411365509, + "learning_rate": 0.00017928605220484512, + "loss": 2.5546, + "step": 5703 + }, + { + "epoch": 0.730112, + "grad_norm": 0.3983548879623413, + "learning_rate": 0.0001791270341451741, + "loss": 2.5412, + "step": 5704 + }, + { + "epoch": 0.73024, + "grad_norm": 0.4128880202770233, + "learning_rate": 0.00017896807124728582, + "loss": 2.6459, + "step": 5705 + }, + { + "epoch": 0.730368, + "grad_norm": 0.380547434091568, + "learning_rate": 0.00017880916353850778, + "loss": 2.6835, + "step": 5706 + }, + { + "epoch": 0.730496, + "grad_norm": 0.53420490026474, + "learning_rate": 0.00017865031104615807, + "loss": 2.628, + "step": 5707 + }, + { + "epoch": 0.730624, + "grad_norm": 0.43399351835250854, + "learning_rate": 0.0001784915137975452, + "loss": 2.504, + "step": 5708 + }, + { + "epoch": 0.730752, + "grad_norm": 0.4074702262878418, + "learning_rate": 0.0001783327718199685, + "loss": 2.5865, + "step": 5709 + }, + { + "epoch": 0.73088, + "grad_norm": 0.39005911350250244, + "learning_rate": 0.0001781740851407172, + "loss": 2.6198, + "step": 5710 + }, + { + "epoch": 0.731008, + "grad_norm": 0.40494564175605774, + "learning_rate": 0.00017801545378707178, + "loss": 2.6025, + "step": 5711 + }, + { + "epoch": 0.731136, + "grad_norm": 0.38487035036087036, + "learning_rate": 0.00017785687778630266, + "loss": 2.626, + "step": 5712 + }, + { + "epoch": 0.731264, + "grad_norm": 0.4293110966682434, + "learning_rate": 0.0001776983571656708, + "loss": 2.6915, + "step": 5713 + }, + { + "epoch": 0.731392, + "grad_norm": 0.7254016995429993, + "learning_rate": 0.00017753989195242782, + "loss": 2.6064, + "step": 5714 + }, + { + "epoch": 0.73152, + "grad_norm": 0.40167227387428284, + "learning_rate": 0.00017738148217381566, + "loss": 2.6354, + "step": 5715 + }, + { + "epoch": 0.731648, + "grad_norm": 0.38170892000198364, + "learning_rate": 0.00017722312785706668, + "loss": 2.6242, + "step": 5716 + }, + { + "epoch": 0.731776, + "grad_norm": 0.42717406153678894, + "learning_rate": 0.00017706482902940396, + "loss": 2.5017, + "step": 5717 + }, + { + "epoch": 0.731904, + "grad_norm": 0.7764276266098022, + "learning_rate": 0.00017690658571804107, + "loss": 2.5252, + "step": 5718 + }, + { + "epoch": 0.732032, + "grad_norm": 0.39688533544540405, + "learning_rate": 0.00017674839795018171, + "loss": 2.6287, + "step": 5719 + }, + { + "epoch": 0.73216, + "grad_norm": 0.4220132827758789, + "learning_rate": 0.00017659026575302022, + "loss": 2.6884, + "step": 5720 + }, + { + "epoch": 0.732288, + "grad_norm": 0.5468571782112122, + "learning_rate": 0.00017643218915374138, + "loss": 2.5645, + "step": 5721 + }, + { + "epoch": 0.732416, + "grad_norm": 0.3845243453979492, + "learning_rate": 0.0001762741681795203, + "loss": 2.7539, + "step": 5722 + }, + { + "epoch": 0.732544, + "grad_norm": 0.4138295650482178, + "learning_rate": 0.00017611620285752245, + "loss": 2.4934, + "step": 5723 + }, + { + "epoch": 0.732672, + "grad_norm": 0.3945460319519043, + "learning_rate": 0.00017595829321490435, + "loss": 2.6004, + "step": 5724 + }, + { + "epoch": 0.7328, + "grad_norm": 0.3701542913913727, + "learning_rate": 0.00017580043927881224, + "loss": 2.61, + "step": 5725 + }, + { + "epoch": 0.732928, + "grad_norm": 0.43493154644966125, + "learning_rate": 0.0001756426410763829, + "loss": 2.5944, + "step": 5726 + }, + { + "epoch": 0.733056, + "grad_norm": 0.4103633165359497, + "learning_rate": 0.00017548489863474393, + "loss": 2.5139, + "step": 5727 + }, + { + "epoch": 0.733184, + "grad_norm": 0.413441926240921, + "learning_rate": 0.0001753272119810131, + "loss": 2.725, + "step": 5728 + }, + { + "epoch": 0.733312, + "grad_norm": 0.42868924140930176, + "learning_rate": 0.00017516958114229835, + "loss": 2.526, + "step": 5729 + }, + { + "epoch": 0.73344, + "grad_norm": 0.40177279710769653, + "learning_rate": 0.00017501200614569845, + "loss": 2.6545, + "step": 5730 + }, + { + "epoch": 0.733568, + "grad_norm": 0.38972657918930054, + "learning_rate": 0.00017485448701830205, + "loss": 2.7001, + "step": 5731 + }, + { + "epoch": 0.733696, + "grad_norm": 0.4057590961456299, + "learning_rate": 0.00017469702378718893, + "loss": 2.6224, + "step": 5732 + }, + { + "epoch": 0.733824, + "grad_norm": 0.38667032122612, + "learning_rate": 0.00017453961647942868, + "loss": 2.5768, + "step": 5733 + }, + { + "epoch": 0.733952, + "grad_norm": 0.4014623761177063, + "learning_rate": 0.00017438226512208138, + "loss": 2.6168, + "step": 5734 + }, + { + "epoch": 0.73408, + "grad_norm": 0.3654799163341522, + "learning_rate": 0.0001742249697421976, + "loss": 2.5695, + "step": 5735 + }, + { + "epoch": 0.734208, + "grad_norm": 0.42252686619758606, + "learning_rate": 0.00017406773036681805, + "loss": 2.7389, + "step": 5736 + }, + { + "epoch": 0.734336, + "grad_norm": 0.3758196532726288, + "learning_rate": 0.00017391054702297438, + "loss": 2.6976, + "step": 5737 + }, + { + "epoch": 0.734464, + "grad_norm": 0.37559598684310913, + "learning_rate": 0.00017375341973768787, + "loss": 2.548, + "step": 5738 + }, + { + "epoch": 0.734592, + "grad_norm": 0.44506940245628357, + "learning_rate": 0.0001735963485379708, + "loss": 2.6324, + "step": 5739 + }, + { + "epoch": 0.73472, + "grad_norm": 0.42468324303627014, + "learning_rate": 0.00017343933345082547, + "loss": 2.6242, + "step": 5740 + }, + { + "epoch": 0.734848, + "grad_norm": 0.43599823117256165, + "learning_rate": 0.00017328237450324452, + "loss": 2.6437, + "step": 5741 + }, + { + "epoch": 0.734976, + "grad_norm": 0.4200100004673004, + "learning_rate": 0.00017312547172221098, + "loss": 2.7136, + "step": 5742 + }, + { + "epoch": 0.735104, + "grad_norm": 0.3851967453956604, + "learning_rate": 0.00017296862513469836, + "loss": 2.5746, + "step": 5743 + }, + { + "epoch": 0.735232, + "grad_norm": 0.4301871657371521, + "learning_rate": 0.00017281183476767015, + "loss": 2.6657, + "step": 5744 + }, + { + "epoch": 0.73536, + "grad_norm": 0.3977126181125641, + "learning_rate": 0.00017265510064808082, + "loss": 2.6281, + "step": 5745 + }, + { + "epoch": 0.735488, + "grad_norm": 0.3824109435081482, + "learning_rate": 0.00017249842280287443, + "loss": 2.5073, + "step": 5746 + }, + { + "epoch": 0.735616, + "grad_norm": 0.36062368750572205, + "learning_rate": 0.00017234180125898606, + "loss": 2.608, + "step": 5747 + }, + { + "epoch": 0.735744, + "grad_norm": 0.3704891800880432, + "learning_rate": 0.00017218523604334057, + "loss": 2.6532, + "step": 5748 + }, + { + "epoch": 0.735872, + "grad_norm": 0.42177271842956543, + "learning_rate": 0.0001720287271828534, + "loss": 2.6052, + "step": 5749 + }, + { + "epoch": 0.736, + "grad_norm": 0.4593290686607361, + "learning_rate": 0.00017187227470443013, + "loss": 2.708, + "step": 5750 + }, + { + "epoch": 0.736128, + "grad_norm": 0.3896675109863281, + "learning_rate": 0.00017171587863496686, + "loss": 2.574, + "step": 5751 + }, + { + "epoch": 0.736256, + "grad_norm": 0.3901676833629608, + "learning_rate": 0.00017155953900134969, + "loss": 2.6333, + "step": 5752 + }, + { + "epoch": 0.736384, + "grad_norm": 0.3728947043418884, + "learning_rate": 0.00017140325583045552, + "loss": 2.7956, + "step": 5753 + }, + { + "epoch": 0.736512, + "grad_norm": 0.3886569142341614, + "learning_rate": 0.00017124702914915097, + "loss": 2.6927, + "step": 5754 + }, + { + "epoch": 0.73664, + "grad_norm": 0.40396779775619507, + "learning_rate": 0.00017109085898429343, + "loss": 2.5741, + "step": 5755 + }, + { + "epoch": 0.736768, + "grad_norm": 0.4112150967121124, + "learning_rate": 0.00017093474536273036, + "loss": 2.6826, + "step": 5756 + }, + { + "epoch": 0.736896, + "grad_norm": 0.3871515691280365, + "learning_rate": 0.00017077868831129933, + "loss": 2.6329, + "step": 5757 + }, + { + "epoch": 0.737024, + "grad_norm": 0.43098366260528564, + "learning_rate": 0.00017062268785682848, + "loss": 2.6439, + "step": 5758 + }, + { + "epoch": 0.737152, + "grad_norm": 0.405316025018692, + "learning_rate": 0.00017046674402613593, + "loss": 2.6202, + "step": 5759 + }, + { + "epoch": 0.73728, + "grad_norm": 0.4005146622657776, + "learning_rate": 0.00017031085684603048, + "loss": 2.4695, + "step": 5760 + }, + { + "epoch": 0.737408, + "grad_norm": 0.3764992356300354, + "learning_rate": 0.00017015502634331082, + "loss": 2.6707, + "step": 5761 + }, + { + "epoch": 0.737536, + "grad_norm": 0.4180753231048584, + "learning_rate": 0.00016999925254476605, + "loss": 2.5752, + "step": 5762 + }, + { + "epoch": 0.737664, + "grad_norm": 0.4170711934566498, + "learning_rate": 0.0001698435354771753, + "loss": 2.6178, + "step": 5763 + }, + { + "epoch": 0.737792, + "grad_norm": 0.3923589289188385, + "learning_rate": 0.00016968787516730843, + "loss": 2.6475, + "step": 5764 + }, + { + "epoch": 0.73792, + "grad_norm": 0.4039965867996216, + "learning_rate": 0.00016953227164192514, + "loss": 2.6218, + "step": 5765 + }, + { + "epoch": 0.738048, + "grad_norm": 0.41160935163497925, + "learning_rate": 0.00016937672492777544, + "loss": 2.6986, + "step": 5766 + }, + { + "epoch": 0.738176, + "grad_norm": 0.40074071288108826, + "learning_rate": 0.00016922123505159952, + "loss": 2.6136, + "step": 5767 + }, + { + "epoch": 0.738304, + "grad_norm": 0.4559706151485443, + "learning_rate": 0.00016906580204012817, + "loss": 2.5226, + "step": 5768 + }, + { + "epoch": 0.738432, + "grad_norm": 0.3968808948993683, + "learning_rate": 0.00016891042592008198, + "loss": 2.5685, + "step": 5769 + }, + { + "epoch": 0.73856, + "grad_norm": 0.4044816195964813, + "learning_rate": 0.00016875510671817185, + "loss": 2.5658, + "step": 5770 + }, + { + "epoch": 0.738688, + "grad_norm": 0.4256884753704071, + "learning_rate": 0.00016859984446109905, + "loss": 2.661, + "step": 5771 + }, + { + "epoch": 0.738816, + "grad_norm": 0.4486631155014038, + "learning_rate": 0.00016844463917555487, + "loss": 2.6538, + "step": 5772 + }, + { + "epoch": 0.738944, + "grad_norm": 0.3946046829223633, + "learning_rate": 0.00016828949088822077, + "loss": 2.6437, + "step": 5773 + }, + { + "epoch": 0.739072, + "grad_norm": 0.412197470664978, + "learning_rate": 0.00016813439962576872, + "loss": 2.6204, + "step": 5774 + }, + { + "epoch": 0.7392, + "grad_norm": 0.4021168351173401, + "learning_rate": 0.00016797936541486082, + "loss": 2.6435, + "step": 5775 + }, + { + "epoch": 0.739328, + "grad_norm": 0.36580538749694824, + "learning_rate": 0.00016782438828214912, + "loss": 2.622, + "step": 5776 + }, + { + "epoch": 0.739456, + "grad_norm": 0.3935343027114868, + "learning_rate": 0.00016766946825427605, + "loss": 2.606, + "step": 5777 + }, + { + "epoch": 0.739584, + "grad_norm": 0.39893555641174316, + "learning_rate": 0.00016751460535787405, + "loss": 2.4783, + "step": 5778 + }, + { + "epoch": 0.739712, + "grad_norm": 0.4013819098472595, + "learning_rate": 0.00016735979961956588, + "loss": 2.6896, + "step": 5779 + }, + { + "epoch": 0.73984, + "grad_norm": 0.37883052229881287, + "learning_rate": 0.00016720505106596428, + "loss": 2.569, + "step": 5780 + }, + { + "epoch": 0.739968, + "grad_norm": 0.4031547009944916, + "learning_rate": 0.00016705035972367272, + "loss": 2.6499, + "step": 5781 + }, + { + "epoch": 0.740096, + "grad_norm": 0.40660199522972107, + "learning_rate": 0.00016689572561928395, + "loss": 2.5879, + "step": 5782 + }, + { + "epoch": 0.740224, + "grad_norm": 0.41501787304878235, + "learning_rate": 0.00016674114877938183, + "loss": 2.581, + "step": 5783 + }, + { + "epoch": 0.740352, + "grad_norm": 0.36564627289772034, + "learning_rate": 0.00016658662923053975, + "loss": 2.6779, + "step": 5784 + }, + { + "epoch": 0.74048, + "grad_norm": 0.3867315948009491, + "learning_rate": 0.0001664321669993213, + "loss": 2.5976, + "step": 5785 + }, + { + "epoch": 0.740608, + "grad_norm": 0.37608423829078674, + "learning_rate": 0.0001662777621122804, + "loss": 2.5239, + "step": 5786 + }, + { + "epoch": 0.740736, + "grad_norm": 0.3591589629650116, + "learning_rate": 0.0001661234145959611, + "loss": 2.6131, + "step": 5787 + }, + { + "epoch": 0.740864, + "grad_norm": 0.8444374799728394, + "learning_rate": 0.0001659691244768972, + "loss": 2.5042, + "step": 5788 + }, + { + "epoch": 0.740992, + "grad_norm": 0.3825283348560333, + "learning_rate": 0.00016581489178161347, + "loss": 2.7005, + "step": 5789 + }, + { + "epoch": 0.74112, + "grad_norm": 0.3716539144515991, + "learning_rate": 0.00016566071653662402, + "loss": 2.5692, + "step": 5790 + }, + { + "epoch": 0.741248, + "grad_norm": 0.38441047072410583, + "learning_rate": 0.00016550659876843316, + "loss": 2.6684, + "step": 5791 + }, + { + "epoch": 0.741376, + "grad_norm": 0.3823108375072479, + "learning_rate": 0.000165352538503536, + "loss": 2.5966, + "step": 5792 + }, + { + "epoch": 0.741504, + "grad_norm": 0.3644655644893646, + "learning_rate": 0.000165198535768417, + "loss": 2.5139, + "step": 5793 + }, + { + "epoch": 0.741632, + "grad_norm": 0.3828287124633789, + "learning_rate": 0.00016504459058955107, + "loss": 2.5026, + "step": 5794 + }, + { + "epoch": 0.74176, + "grad_norm": 0.394607275724411, + "learning_rate": 0.00016489070299340297, + "loss": 2.5366, + "step": 5795 + }, + { + "epoch": 0.741888, + "grad_norm": 0.39209476113319397, + "learning_rate": 0.00016473687300642815, + "loss": 2.5714, + "step": 5796 + }, + { + "epoch": 0.742016, + "grad_norm": 0.3787729740142822, + "learning_rate": 0.00016458310065507154, + "loss": 2.6028, + "step": 5797 + }, + { + "epoch": 0.742144, + "grad_norm": 0.43597257137298584, + "learning_rate": 0.00016442938596576839, + "loss": 2.6725, + "step": 5798 + }, + { + "epoch": 0.742272, + "grad_norm": 0.4148973226547241, + "learning_rate": 0.0001642757289649441, + "loss": 2.6325, + "step": 5799 + }, + { + "epoch": 0.7424, + "grad_norm": 0.37813833355903625, + "learning_rate": 0.00016412212967901386, + "loss": 2.6862, + "step": 5800 + }, + { + "epoch": 0.742528, + "grad_norm": 0.3916599154472351, + "learning_rate": 0.00016396858813438355, + "loss": 2.5339, + "step": 5801 + }, + { + "epoch": 0.742656, + "grad_norm": 0.4168313443660736, + "learning_rate": 0.0001638151043574485, + "loss": 2.7583, + "step": 5802 + }, + { + "epoch": 0.742784, + "grad_norm": 0.4079570770263672, + "learning_rate": 0.00016366167837459427, + "loss": 2.677, + "step": 5803 + }, + { + "epoch": 0.742912, + "grad_norm": 0.38518476486206055, + "learning_rate": 0.00016350831021219687, + "loss": 2.5947, + "step": 5804 + }, + { + "epoch": 0.74304, + "grad_norm": 0.39561226963996887, + "learning_rate": 0.00016335499989662185, + "loss": 2.5274, + "step": 5805 + }, + { + "epoch": 0.743168, + "grad_norm": 0.383658766746521, + "learning_rate": 0.0001632017474542251, + "loss": 2.6072, + "step": 5806 + }, + { + "epoch": 0.743296, + "grad_norm": 0.40299639105796814, + "learning_rate": 0.00016304855291135245, + "loss": 2.609, + "step": 5807 + }, + { + "epoch": 0.743424, + "grad_norm": 0.399600088596344, + "learning_rate": 0.00016289541629433985, + "loss": 2.5075, + "step": 5808 + }, + { + "epoch": 0.743552, + "grad_norm": 0.4094698429107666, + "learning_rate": 0.00016274233762951306, + "loss": 2.656, + "step": 5809 + }, + { + "epoch": 0.74368, + "grad_norm": 0.37717321515083313, + "learning_rate": 0.00016258931694318829, + "loss": 2.5482, + "step": 5810 + }, + { + "epoch": 0.743808, + "grad_norm": 0.3897896111011505, + "learning_rate": 0.0001624363542616717, + "loss": 2.548, + "step": 5811 + }, + { + "epoch": 0.743936, + "grad_norm": 0.3924122154712677, + "learning_rate": 0.00016228344961125913, + "loss": 2.6528, + "step": 5812 + }, + { + "epoch": 0.744064, + "grad_norm": 0.38635414838790894, + "learning_rate": 0.00016213060301823674, + "loss": 2.623, + "step": 5813 + }, + { + "epoch": 0.744192, + "grad_norm": 0.3903887867927551, + "learning_rate": 0.00016197781450888067, + "loss": 2.6196, + "step": 5814 + }, + { + "epoch": 0.74432, + "grad_norm": 0.36960816383361816, + "learning_rate": 0.0001618250841094569, + "loss": 2.5352, + "step": 5815 + }, + { + "epoch": 0.744448, + "grad_norm": 0.4529739022254944, + "learning_rate": 0.0001616724118462215, + "loss": 2.581, + "step": 5816 + }, + { + "epoch": 0.744576, + "grad_norm": 0.45853322744369507, + "learning_rate": 0.00016151979774542085, + "loss": 2.6522, + "step": 5817 + }, + { + "epoch": 0.744704, + "grad_norm": 0.40609949827194214, + "learning_rate": 0.00016136724183329104, + "loss": 2.5544, + "step": 5818 + }, + { + "epoch": 0.744832, + "grad_norm": 0.3793293237686157, + "learning_rate": 0.0001612147441360579, + "loss": 2.5748, + "step": 5819 + }, + { + "epoch": 0.74496, + "grad_norm": 0.3733328580856323, + "learning_rate": 0.000161062304679938, + "loss": 2.6249, + "step": 5820 + }, + { + "epoch": 0.745088, + "grad_norm": 0.41219085454940796, + "learning_rate": 0.0001609099234911372, + "loss": 2.5789, + "step": 5821 + }, + { + "epoch": 0.745216, + "grad_norm": 0.40360596776008606, + "learning_rate": 0.00016075760059585165, + "loss": 2.6161, + "step": 5822 + }, + { + "epoch": 0.745344, + "grad_norm": 0.41088923811912537, + "learning_rate": 0.00016060533602026733, + "loss": 2.6451, + "step": 5823 + }, + { + "epoch": 0.745472, + "grad_norm": 0.4225967526435852, + "learning_rate": 0.00016045312979056026, + "loss": 2.7054, + "step": 5824 + }, + { + "epoch": 0.7456, + "grad_norm": 0.4073977768421173, + "learning_rate": 0.00016030098193289667, + "loss": 2.6749, + "step": 5825 + }, + { + "epoch": 0.745728, + "grad_norm": 0.3648526072502136, + "learning_rate": 0.00016014889247343238, + "loss": 2.6373, + "step": 5826 + }, + { + "epoch": 0.745856, + "grad_norm": 0.44379615783691406, + "learning_rate": 0.00015999686143831343, + "loss": 2.5497, + "step": 5827 + }, + { + "epoch": 0.745984, + "grad_norm": 0.4142730236053467, + "learning_rate": 0.00015984488885367542, + "loss": 2.6209, + "step": 5828 + }, + { + "epoch": 0.746112, + "grad_norm": 0.39415809512138367, + "learning_rate": 0.00015969297474564465, + "loss": 2.6669, + "step": 5829 + }, + { + "epoch": 0.74624, + "grad_norm": 0.4292858839035034, + "learning_rate": 0.0001595411191403367, + "loss": 2.603, + "step": 5830 + }, + { + "epoch": 0.746368, + "grad_norm": 0.35461243987083435, + "learning_rate": 0.00015938932206385715, + "loss": 2.5409, + "step": 5831 + }, + { + "epoch": 0.746496, + "grad_norm": 0.38418495655059814, + "learning_rate": 0.00015923758354230194, + "loss": 2.5376, + "step": 5832 + }, + { + "epoch": 0.746624, + "grad_norm": 0.4851783812046051, + "learning_rate": 0.00015908590360175662, + "loss": 2.5666, + "step": 5833 + }, + { + "epoch": 0.746752, + "grad_norm": 0.4069935083389282, + "learning_rate": 0.00015893428226829671, + "loss": 2.595, + "step": 5834 + }, + { + "epoch": 0.74688, + "grad_norm": 0.4040979743003845, + "learning_rate": 0.00015878271956798763, + "loss": 2.6426, + "step": 5835 + }, + { + "epoch": 0.747008, + "grad_norm": 0.4921731650829315, + "learning_rate": 0.00015863121552688476, + "loss": 2.6354, + "step": 5836 + }, + { + "epoch": 0.747136, + "grad_norm": 0.46137523651123047, + "learning_rate": 0.00015847977017103326, + "loss": 2.6342, + "step": 5837 + }, + { + "epoch": 0.747264, + "grad_norm": 0.41960954666137695, + "learning_rate": 0.00015832838352646856, + "loss": 2.656, + "step": 5838 + }, + { + "epoch": 0.747392, + "grad_norm": 0.3828534185886383, + "learning_rate": 0.00015817705561921587, + "loss": 2.6309, + "step": 5839 + }, + { + "epoch": 0.74752, + "grad_norm": 0.4002903997898102, + "learning_rate": 0.00015802578647529004, + "loss": 2.6001, + "step": 5840 + }, + { + "epoch": 0.747648, + "grad_norm": 0.36892834305763245, + "learning_rate": 0.00015787457612069606, + "loss": 2.4584, + "step": 5841 + }, + { + "epoch": 0.747776, + "grad_norm": 0.38230422139167786, + "learning_rate": 0.00015772342458142868, + "loss": 2.6055, + "step": 5842 + }, + { + "epoch": 0.747904, + "grad_norm": 0.3625714182853699, + "learning_rate": 0.00015757233188347264, + "loss": 2.5981, + "step": 5843 + }, + { + "epoch": 0.748032, + "grad_norm": 0.3972434401512146, + "learning_rate": 0.00015742129805280247, + "loss": 2.5205, + "step": 5844 + }, + { + "epoch": 0.74816, + "grad_norm": 0.3919067978858948, + "learning_rate": 0.00015727032311538258, + "loss": 2.5124, + "step": 5845 + }, + { + "epoch": 0.748288, + "grad_norm": 0.43027353286743164, + "learning_rate": 0.00015711940709716755, + "loss": 2.5684, + "step": 5846 + }, + { + "epoch": 0.748416, + "grad_norm": 0.3753076493740082, + "learning_rate": 0.00015696855002410128, + "loss": 2.5718, + "step": 5847 + }, + { + "epoch": 0.748544, + "grad_norm": 0.37018468976020813, + "learning_rate": 0.00015681775192211817, + "loss": 2.5393, + "step": 5848 + }, + { + "epoch": 0.748672, + "grad_norm": 0.38455408811569214, + "learning_rate": 0.00015666701281714202, + "loss": 2.6659, + "step": 5849 + }, + { + "epoch": 0.7488, + "grad_norm": 0.4417814612388611, + "learning_rate": 0.00015651633273508665, + "loss": 2.6486, + "step": 5850 + }, + { + "epoch": 0.748928, + "grad_norm": 0.4333549737930298, + "learning_rate": 0.00015636571170185565, + "loss": 2.5101, + "step": 5851 + }, + { + "epoch": 0.749056, + "grad_norm": 0.3966858386993408, + "learning_rate": 0.00015621514974334245, + "loss": 2.5471, + "step": 5852 + }, + { + "epoch": 0.749184, + "grad_norm": 0.4119015634059906, + "learning_rate": 0.00015606464688543064, + "loss": 2.615, + "step": 5853 + }, + { + "epoch": 0.749312, + "grad_norm": 0.44868725538253784, + "learning_rate": 0.00015591420315399325, + "loss": 2.5488, + "step": 5854 + }, + { + "epoch": 0.74944, + "grad_norm": 0.4201100468635559, + "learning_rate": 0.00015576381857489337, + "loss": 2.6397, + "step": 5855 + }, + { + "epoch": 0.749568, + "grad_norm": 0.3741871416568756, + "learning_rate": 0.0001556134931739836, + "loss": 2.7098, + "step": 5856 + }, + { + "epoch": 0.749696, + "grad_norm": 0.4414110481739044, + "learning_rate": 0.00015546322697710698, + "loss": 2.6074, + "step": 5857 + }, + { + "epoch": 0.749824, + "grad_norm": 0.4266470968723297, + "learning_rate": 0.0001553130200100959, + "loss": 2.5612, + "step": 5858 + }, + { + "epoch": 0.749952, + "grad_norm": 0.7046429514884949, + "learning_rate": 0.00015516287229877242, + "loss": 2.6814, + "step": 5859 + }, + { + "epoch": 0.75008, + "grad_norm": 0.7137123942375183, + "learning_rate": 0.00015501278386894908, + "loss": 2.7408, + "step": 5860 + }, + { + "epoch": 0.750208, + "grad_norm": 0.6542083024978638, + "learning_rate": 0.00015486275474642762, + "loss": 2.5149, + "step": 5861 + }, + { + "epoch": 0.750336, + "grad_norm": 0.4116375744342804, + "learning_rate": 0.0001547127849569998, + "loss": 2.5672, + "step": 5862 + }, + { + "epoch": 0.750464, + "grad_norm": 0.4030555188655853, + "learning_rate": 0.00015456287452644718, + "loss": 2.6412, + "step": 5863 + }, + { + "epoch": 0.750592, + "grad_norm": 0.47841012477874756, + "learning_rate": 0.00015441302348054103, + "loss": 2.5606, + "step": 5864 + }, + { + "epoch": 0.75072, + "grad_norm": 0.4387114942073822, + "learning_rate": 0.00015426323184504243, + "loss": 2.5809, + "step": 5865 + }, + { + "epoch": 0.750848, + "grad_norm": 0.4327980577945709, + "learning_rate": 0.00015411349964570254, + "loss": 2.5596, + "step": 5866 + }, + { + "epoch": 0.750976, + "grad_norm": 0.4231038987636566, + "learning_rate": 0.00015396382690826172, + "loss": 2.5335, + "step": 5867 + }, + { + "epoch": 0.751104, + "grad_norm": 0.47892871499061584, + "learning_rate": 0.00015381421365845082, + "loss": 2.7172, + "step": 5868 + }, + { + "epoch": 0.751232, + "grad_norm": 0.4134797155857086, + "learning_rate": 0.00015366465992198996, + "loss": 2.5279, + "step": 5869 + }, + { + "epoch": 0.75136, + "grad_norm": 0.381535142660141, + "learning_rate": 0.00015351516572458912, + "loss": 2.6034, + "step": 5870 + }, + { + "epoch": 0.751488, + "grad_norm": 0.38747987151145935, + "learning_rate": 0.00015336573109194808, + "loss": 2.5548, + "step": 5871 + }, + { + "epoch": 0.751616, + "grad_norm": 0.4089334011077881, + "learning_rate": 0.00015321635604975637, + "loss": 2.7165, + "step": 5872 + }, + { + "epoch": 0.751744, + "grad_norm": 0.41963014006614685, + "learning_rate": 0.0001530670406236932, + "loss": 2.6667, + "step": 5873 + }, + { + "epoch": 0.751872, + "grad_norm": 0.39433878660202026, + "learning_rate": 0.0001529177848394277, + "loss": 2.4818, + "step": 5874 + }, + { + "epoch": 0.752, + "grad_norm": 0.3936058580875397, + "learning_rate": 0.0001527685887226189, + "loss": 2.5362, + "step": 5875 + }, + { + "epoch": 0.752128, + "grad_norm": 0.46208903193473816, + "learning_rate": 0.00015261945229891511, + "loss": 2.6756, + "step": 5876 + }, + { + "epoch": 0.752256, + "grad_norm": 0.4543020725250244, + "learning_rate": 0.00015247037559395466, + "loss": 2.7396, + "step": 5877 + }, + { + "epoch": 0.752384, + "grad_norm": 0.4577682316303253, + "learning_rate": 0.00015232135863336554, + "loss": 2.5736, + "step": 5878 + }, + { + "epoch": 0.752512, + "grad_norm": 0.4466424584388733, + "learning_rate": 0.00015217240144276556, + "loss": 2.6429, + "step": 5879 + }, + { + "epoch": 0.75264, + "grad_norm": 0.38940608501434326, + "learning_rate": 0.00015202350404776193, + "loss": 2.625, + "step": 5880 + }, + { + "epoch": 0.752768, + "grad_norm": 0.42364394664764404, + "learning_rate": 0.00015187466647395226, + "loss": 2.617, + "step": 5881 + }, + { + "epoch": 0.752896, + "grad_norm": 0.3996695280075073, + "learning_rate": 0.00015172588874692316, + "loss": 2.5801, + "step": 5882 + }, + { + "epoch": 0.753024, + "grad_norm": 0.4202781021595001, + "learning_rate": 0.00015157717089225143, + "loss": 2.506, + "step": 5883 + }, + { + "epoch": 0.753152, + "grad_norm": 0.3977370262145996, + "learning_rate": 0.00015142851293550302, + "loss": 2.6261, + "step": 5884 + }, + { + "epoch": 0.75328, + "grad_norm": 0.45759350061416626, + "learning_rate": 0.00015127991490223447, + "loss": 2.6155, + "step": 5885 + }, + { + "epoch": 0.753408, + "grad_norm": 0.4245941638946533, + "learning_rate": 0.0001511313768179912, + "loss": 2.558, + "step": 5886 + }, + { + "epoch": 0.753536, + "grad_norm": 0.40532588958740234, + "learning_rate": 0.00015098289870830867, + "loss": 2.6269, + "step": 5887 + }, + { + "epoch": 0.753664, + "grad_norm": 0.42097657918930054, + "learning_rate": 0.00015083448059871192, + "loss": 2.5675, + "step": 5888 + }, + { + "epoch": 0.753792, + "grad_norm": 0.40669089555740356, + "learning_rate": 0.00015068612251471593, + "loss": 2.5474, + "step": 5889 + }, + { + "epoch": 0.75392, + "grad_norm": 0.42677465081214905, + "learning_rate": 0.00015053782448182508, + "loss": 2.5954, + "step": 5890 + }, + { + "epoch": 0.754048, + "grad_norm": 0.41200876235961914, + "learning_rate": 0.00015038958652553353, + "loss": 2.6728, + "step": 5891 + }, + { + "epoch": 0.754176, + "grad_norm": 0.372958779335022, + "learning_rate": 0.00015024140867132507, + "loss": 2.6479, + "step": 5892 + }, + { + "epoch": 0.754304, + "grad_norm": 0.42624250054359436, + "learning_rate": 0.0001500932909446731, + "loss": 2.5809, + "step": 5893 + }, + { + "epoch": 0.754432, + "grad_norm": 0.3926074206829071, + "learning_rate": 0.00014994523337104099, + "loss": 2.5749, + "step": 5894 + }, + { + "epoch": 0.75456, + "grad_norm": 0.37479808926582336, + "learning_rate": 0.0001497972359758813, + "loss": 2.645, + "step": 5895 + }, + { + "epoch": 0.754688, + "grad_norm": 0.4227044880390167, + "learning_rate": 0.00014964929878463683, + "loss": 2.6566, + "step": 5896 + }, + { + "epoch": 0.754816, + "grad_norm": 0.4616771638393402, + "learning_rate": 0.00014950142182273946, + "loss": 2.619, + "step": 5897 + }, + { + "epoch": 0.754944, + "grad_norm": 0.37610822916030884, + "learning_rate": 0.0001493536051156111, + "loss": 2.5097, + "step": 5898 + }, + { + "epoch": 0.755072, + "grad_norm": 0.4525318741798401, + "learning_rate": 0.00014920584868866293, + "loss": 2.632, + "step": 5899 + }, + { + "epoch": 0.7552, + "grad_norm": 0.37713319063186646, + "learning_rate": 0.0001490581525672962, + "loss": 2.6277, + "step": 5900 + }, + { + "epoch": 0.755328, + "grad_norm": 0.44049257040023804, + "learning_rate": 0.00014891051677690153, + "loss": 2.5412, + "step": 5901 + }, + { + "epoch": 0.755456, + "grad_norm": 0.39488622546195984, + "learning_rate": 0.000148762941342859, + "loss": 2.5194, + "step": 5902 + }, + { + "epoch": 0.755584, + "grad_norm": 0.3814573287963867, + "learning_rate": 0.0001486154262905388, + "loss": 2.6571, + "step": 5903 + }, + { + "epoch": 0.755712, + "grad_norm": 0.4185042083263397, + "learning_rate": 0.00014846797164530051, + "loss": 2.5972, + "step": 5904 + }, + { + "epoch": 0.75584, + "grad_norm": 0.43625205755233765, + "learning_rate": 0.00014832057743249328, + "loss": 2.6014, + "step": 5905 + }, + { + "epoch": 0.755968, + "grad_norm": 0.4717802405357361, + "learning_rate": 0.00014817324367745572, + "loss": 2.6608, + "step": 5906 + }, + { + "epoch": 0.756096, + "grad_norm": 0.45968466997146606, + "learning_rate": 0.00014802597040551634, + "loss": 2.5933, + "step": 5907 + }, + { + "epoch": 0.756224, + "grad_norm": 0.3915333151817322, + "learning_rate": 0.0001478787576419931, + "loss": 2.6613, + "step": 5908 + }, + { + "epoch": 0.756352, + "grad_norm": 0.5003970861434937, + "learning_rate": 0.00014773160541219337, + "loss": 2.5147, + "step": 5909 + }, + { + "epoch": 0.75648, + "grad_norm": 0.4384174048900604, + "learning_rate": 0.00014758451374141467, + "loss": 2.6411, + "step": 5910 + }, + { + "epoch": 0.756608, + "grad_norm": 0.40679052472114563, + "learning_rate": 0.0001474374826549435, + "loss": 2.6406, + "step": 5911 + }, + { + "epoch": 0.756736, + "grad_norm": 0.46530285477638245, + "learning_rate": 0.00014729051217805646, + "loss": 2.5628, + "step": 5912 + }, + { + "epoch": 0.756864, + "grad_norm": 0.4556387662887573, + "learning_rate": 0.0001471436023360193, + "loss": 2.5803, + "step": 5913 + }, + { + "epoch": 0.756992, + "grad_norm": 0.3659857511520386, + "learning_rate": 0.00014699675315408756, + "loss": 2.6208, + "step": 5914 + }, + { + "epoch": 0.75712, + "grad_norm": 0.41011282801628113, + "learning_rate": 0.00014684996465750638, + "loss": 2.6672, + "step": 5915 + }, + { + "epoch": 0.757248, + "grad_norm": 0.3920772969722748, + "learning_rate": 0.0001467032368715101, + "loss": 2.5998, + "step": 5916 + }, + { + "epoch": 0.757376, + "grad_norm": 0.4057745635509491, + "learning_rate": 0.00014655656982132337, + "loss": 2.6309, + "step": 5917 + }, + { + "epoch": 0.757504, + "grad_norm": 0.3947126567363739, + "learning_rate": 0.0001464099635321598, + "loss": 2.5921, + "step": 5918 + }, + { + "epoch": 0.757632, + "grad_norm": 0.37214604020118713, + "learning_rate": 0.0001462634180292226, + "loss": 2.5544, + "step": 5919 + }, + { + "epoch": 0.75776, + "grad_norm": 0.3633916676044464, + "learning_rate": 0.0001461169333377047, + "loss": 2.6357, + "step": 5920 + }, + { + "epoch": 0.757888, + "grad_norm": 0.4399898946285248, + "learning_rate": 0.00014597050948278838, + "loss": 2.6398, + "step": 5921 + }, + { + "epoch": 0.758016, + "grad_norm": 0.377533495426178, + "learning_rate": 0.00014582414648964593, + "loss": 2.6138, + "step": 5922 + }, + { + "epoch": 0.758144, + "grad_norm": 0.39986783266067505, + "learning_rate": 0.00014567784438343867, + "loss": 2.5835, + "step": 5923 + }, + { + "epoch": 0.758272, + "grad_norm": 0.4069202244281769, + "learning_rate": 0.0001455316031893174, + "loss": 2.4988, + "step": 5924 + }, + { + "epoch": 0.7584, + "grad_norm": 0.3827359974384308, + "learning_rate": 0.00014538542293242307, + "loss": 2.58, + "step": 5925 + }, + { + "epoch": 0.758528, + "grad_norm": 0.42368394136428833, + "learning_rate": 0.0001452393036378856, + "loss": 2.6458, + "step": 5926 + }, + { + "epoch": 0.758656, + "grad_norm": 0.39790305495262146, + "learning_rate": 0.00014509324533082459, + "loss": 2.5512, + "step": 5927 + }, + { + "epoch": 0.758784, + "grad_norm": 0.3629143536090851, + "learning_rate": 0.00014494724803634912, + "loss": 2.7262, + "step": 5928 + }, + { + "epoch": 0.758912, + "grad_norm": 0.4353103041648865, + "learning_rate": 0.0001448013117795578, + "loss": 2.6108, + "step": 5929 + }, + { + "epoch": 0.75904, + "grad_norm": 0.3640861213207245, + "learning_rate": 0.0001446554365855387, + "loss": 2.5761, + "step": 5930 + }, + { + "epoch": 0.759168, + "grad_norm": 0.4349060356616974, + "learning_rate": 0.0001445096224793695, + "loss": 2.6601, + "step": 5931 + }, + { + "epoch": 0.759296, + "grad_norm": 0.3720160126686096, + "learning_rate": 0.00014436386948611763, + "loss": 2.6283, + "step": 5932 + }, + { + "epoch": 0.759424, + "grad_norm": 0.40644440054893494, + "learning_rate": 0.0001442181776308394, + "loss": 2.664, + "step": 5933 + }, + { + "epoch": 0.759552, + "grad_norm": 0.4154106676578522, + "learning_rate": 0.00014407254693858107, + "loss": 2.6094, + "step": 5934 + }, + { + "epoch": 0.75968, + "grad_norm": 0.4404156804084778, + "learning_rate": 0.00014392697743437815, + "loss": 2.6457, + "step": 5935 + }, + { + "epoch": 0.759808, + "grad_norm": 0.39627066254615784, + "learning_rate": 0.0001437814691432558, + "loss": 2.5383, + "step": 5936 + }, + { + "epoch": 0.759936, + "grad_norm": 0.5011621117591858, + "learning_rate": 0.00014363602209022835, + "loss": 2.695, + "step": 5937 + }, + { + "epoch": 0.760064, + "grad_norm": 0.3889596164226532, + "learning_rate": 0.00014349063630030018, + "loss": 2.5388, + "step": 5938 + }, + { + "epoch": 0.760192, + "grad_norm": 0.4076254367828369, + "learning_rate": 0.00014334531179846456, + "loss": 2.5247, + "step": 5939 + }, + { + "epoch": 0.76032, + "grad_norm": 0.3907029330730438, + "learning_rate": 0.0001432000486097046, + "loss": 2.587, + "step": 5940 + }, + { + "epoch": 0.760448, + "grad_norm": 0.3768913447856903, + "learning_rate": 0.00014305484675899272, + "loss": 2.5982, + "step": 5941 + }, + { + "epoch": 0.760576, + "grad_norm": 0.41510510444641113, + "learning_rate": 0.00014290970627129073, + "loss": 2.6038, + "step": 5942 + }, + { + "epoch": 0.760704, + "grad_norm": 0.4152619540691376, + "learning_rate": 0.00014276462717154997, + "loss": 2.499, + "step": 5943 + }, + { + "epoch": 0.760832, + "grad_norm": 0.4099377691745758, + "learning_rate": 0.0001426196094847112, + "loss": 2.5898, + "step": 5944 + }, + { + "epoch": 0.76096, + "grad_norm": 0.39811888337135315, + "learning_rate": 0.0001424746532357046, + "loss": 2.6292, + "step": 5945 + }, + { + "epoch": 0.761088, + "grad_norm": 0.378768652677536, + "learning_rate": 0.00014232975844944996, + "loss": 2.5302, + "step": 5946 + }, + { + "epoch": 0.761216, + "grad_norm": 0.43514731526374817, + "learning_rate": 0.00014218492515085635, + "loss": 2.6423, + "step": 5947 + }, + { + "epoch": 0.761344, + "grad_norm": 0.45551493763923645, + "learning_rate": 0.00014204015336482213, + "loss": 2.7541, + "step": 5948 + }, + { + "epoch": 0.761472, + "grad_norm": 0.4251973330974579, + "learning_rate": 0.00014189544311623547, + "loss": 2.6295, + "step": 5949 + }, + { + "epoch": 0.7616, + "grad_norm": 0.3958838880062103, + "learning_rate": 0.00014175079442997369, + "loss": 2.5371, + "step": 5950 + }, + { + "epoch": 0.761728, + "grad_norm": 0.4063677191734314, + "learning_rate": 0.00014160620733090352, + "loss": 2.5339, + "step": 5951 + }, + { + "epoch": 0.761856, + "grad_norm": 0.41704097390174866, + "learning_rate": 0.00014146168184388098, + "loss": 2.617, + "step": 5952 + }, + { + "epoch": 0.761984, + "grad_norm": 0.5185837149620056, + "learning_rate": 0.00014131721799375208, + "loss": 2.6517, + "step": 5953 + }, + { + "epoch": 0.762112, + "grad_norm": 0.39858102798461914, + "learning_rate": 0.0001411728158053516, + "loss": 2.5756, + "step": 5954 + }, + { + "epoch": 0.76224, + "grad_norm": 0.48899194598197937, + "learning_rate": 0.00014102847530350388, + "loss": 2.5991, + "step": 5955 + }, + { + "epoch": 0.762368, + "grad_norm": 0.40714865922927856, + "learning_rate": 0.00014088419651302287, + "loss": 2.593, + "step": 5956 + }, + { + "epoch": 0.762496, + "grad_norm": 0.37778663635253906, + "learning_rate": 0.0001407399794587117, + "loss": 2.6118, + "step": 5957 + }, + { + "epoch": 0.762624, + "grad_norm": 0.6372710466384888, + "learning_rate": 0.0001405958241653628, + "loss": 2.6085, + "step": 5958 + }, + { + "epoch": 0.762752, + "grad_norm": 0.44536885619163513, + "learning_rate": 0.0001404517306577585, + "loss": 2.7307, + "step": 5959 + }, + { + "epoch": 0.76288, + "grad_norm": 0.3561967611312866, + "learning_rate": 0.00014030769896066975, + "loss": 2.5618, + "step": 5960 + }, + { + "epoch": 0.763008, + "grad_norm": 0.38628026843070984, + "learning_rate": 0.0001401637290988576, + "loss": 2.6248, + "step": 5961 + }, + { + "epoch": 0.763136, + "grad_norm": 0.3767784833908081, + "learning_rate": 0.000140019821097072, + "loss": 2.4377, + "step": 5962 + }, + { + "epoch": 0.763264, + "grad_norm": 0.40352579951286316, + "learning_rate": 0.00013987597498005245, + "loss": 2.6968, + "step": 5963 + }, + { + "epoch": 0.763392, + "grad_norm": 0.35581278800964355, + "learning_rate": 0.00013973219077252768, + "loss": 2.5945, + "step": 5964 + }, + { + "epoch": 0.76352, + "grad_norm": 0.3732057809829712, + "learning_rate": 0.0001395884684992159, + "loss": 2.5674, + "step": 5965 + }, + { + "epoch": 0.763648, + "grad_norm": 0.39238375425338745, + "learning_rate": 0.00013944480818482447, + "loss": 2.5485, + "step": 5966 + }, + { + "epoch": 0.763776, + "grad_norm": 0.41421717405319214, + "learning_rate": 0.0001393012098540505, + "loss": 2.6366, + "step": 5967 + }, + { + "epoch": 0.763904, + "grad_norm": 0.39487192034721375, + "learning_rate": 0.0001391576735315802, + "loss": 2.6261, + "step": 5968 + }, + { + "epoch": 0.764032, + "grad_norm": 0.3753373622894287, + "learning_rate": 0.00013901419924208908, + "loss": 2.604, + "step": 5969 + }, + { + "epoch": 0.76416, + "grad_norm": 0.38527151942253113, + "learning_rate": 0.00013887078701024203, + "loss": 2.5705, + "step": 5970 + }, + { + "epoch": 0.764288, + "grad_norm": 0.40749484300613403, + "learning_rate": 0.00013872743686069327, + "loss": 2.6085, + "step": 5971 + }, + { + "epoch": 0.764416, + "grad_norm": 0.4113428592681885, + "learning_rate": 0.00013858414881808635, + "loss": 2.5924, + "step": 5972 + }, + { + "epoch": 0.764544, + "grad_norm": 0.3800300359725952, + "learning_rate": 0.00013844092290705396, + "loss": 2.5529, + "step": 5973 + }, + { + "epoch": 0.764672, + "grad_norm": 0.3963069021701813, + "learning_rate": 0.0001382977591522186, + "loss": 2.5887, + "step": 5974 + }, + { + "epoch": 0.7648, + "grad_norm": 0.39833906292915344, + "learning_rate": 0.00013815465757819175, + "loss": 2.6884, + "step": 5975 + }, + { + "epoch": 0.764928, + "grad_norm": 0.40300074219703674, + "learning_rate": 0.00013801161820957387, + "loss": 2.5954, + "step": 5976 + }, + { + "epoch": 0.765056, + "grad_norm": 0.39359840750694275, + "learning_rate": 0.0001378686410709556, + "loss": 2.5013, + "step": 5977 + }, + { + "epoch": 0.765184, + "grad_norm": 0.3697671592235565, + "learning_rate": 0.00013772572618691602, + "loss": 2.5682, + "step": 5978 + }, + { + "epoch": 0.765312, + "grad_norm": 0.4263605773448944, + "learning_rate": 0.00013758287358202405, + "loss": 2.7345, + "step": 5979 + }, + { + "epoch": 0.76544, + "grad_norm": 0.3890385329723358, + "learning_rate": 0.00013744008328083756, + "loss": 2.5638, + "step": 5980 + }, + { + "epoch": 0.765568, + "grad_norm": 0.397869348526001, + "learning_rate": 0.00013729735530790377, + "loss": 2.662, + "step": 5981 + }, + { + "epoch": 0.765696, + "grad_norm": 0.42335546016693115, + "learning_rate": 0.00013715468968775952, + "loss": 2.6342, + "step": 5982 + }, + { + "epoch": 0.765824, + "grad_norm": 2.0847132205963135, + "learning_rate": 0.00013701208644493064, + "loss": 2.5201, + "step": 5983 + }, + { + "epoch": 0.765952, + "grad_norm": 0.400134414434433, + "learning_rate": 0.00013686954560393217, + "loss": 2.5731, + "step": 5984 + }, + { + "epoch": 0.76608, + "grad_norm": 0.49944111704826355, + "learning_rate": 0.00013672706718926848, + "loss": 2.6327, + "step": 5985 + }, + { + "epoch": 0.766208, + "grad_norm": 0.3908420503139496, + "learning_rate": 0.00013658465122543345, + "loss": 2.6291, + "step": 5986 + }, + { + "epoch": 0.766336, + "grad_norm": 0.4272955656051636, + "learning_rate": 0.00013644229773690996, + "loss": 2.492, + "step": 5987 + }, + { + "epoch": 0.766464, + "grad_norm": 0.3924238979816437, + "learning_rate": 0.00013630000674817012, + "loss": 2.6368, + "step": 5988 + }, + { + "epoch": 0.766592, + "grad_norm": 0.5695684552192688, + "learning_rate": 0.0001361577782836756, + "loss": 2.721, + "step": 5989 + }, + { + "epoch": 0.76672, + "grad_norm": 0.3947003185749054, + "learning_rate": 0.00013601561236787702, + "loss": 2.6861, + "step": 5990 + }, + { + "epoch": 0.766848, + "grad_norm": 0.40070661902427673, + "learning_rate": 0.00013587350902521435, + "loss": 2.5595, + "step": 5991 + }, + { + "epoch": 0.766976, + "grad_norm": 0.44523224234580994, + "learning_rate": 0.0001357314682801168, + "loss": 2.6108, + "step": 5992 + }, + { + "epoch": 0.767104, + "grad_norm": 0.3930947482585907, + "learning_rate": 0.00013558949015700277, + "loss": 2.6232, + "step": 5993 + }, + { + "epoch": 0.767232, + "grad_norm": 0.48281559348106384, + "learning_rate": 0.00013544757468027984, + "loss": 2.543, + "step": 5994 + }, + { + "epoch": 0.76736, + "grad_norm": 0.37959811091423035, + "learning_rate": 0.0001353057218743453, + "loss": 2.6654, + "step": 5995 + }, + { + "epoch": 0.767488, + "grad_norm": 0.3995136320590973, + "learning_rate": 0.00013516393176358487, + "loss": 2.6361, + "step": 5996 + }, + { + "epoch": 0.767616, + "grad_norm": 0.3621743619441986, + "learning_rate": 0.00013502220437237427, + "loss": 2.514, + "step": 5997 + }, + { + "epoch": 0.767744, + "grad_norm": 0.432722806930542, + "learning_rate": 0.0001348805397250779, + "loss": 2.6254, + "step": 5998 + }, + { + "epoch": 0.767872, + "grad_norm": 0.4696858823299408, + "learning_rate": 0.00013473893784604962, + "loss": 2.5646, + "step": 5999 + }, + { + "epoch": 0.768, + "grad_norm": 0.38990601897239685, + "learning_rate": 0.00013459739875963238, + "loss": 2.6142, + "step": 6000 + }, + { + "epoch": 0.768128, + "grad_norm": 0.37455669045448303, + "learning_rate": 0.0001344559224901584, + "loss": 2.5207, + "step": 6001 + }, + { + "epoch": 0.768256, + "grad_norm": 0.41965043544769287, + "learning_rate": 0.00013431450906194891, + "loss": 2.5763, + "step": 6002 + }, + { + "epoch": 0.768384, + "grad_norm": 0.44685274362564087, + "learning_rate": 0.00013417315849931493, + "loss": 2.6001, + "step": 6003 + }, + { + "epoch": 0.768512, + "grad_norm": 0.40778180956840515, + "learning_rate": 0.00013403187082655584, + "loss": 2.545, + "step": 6004 + }, + { + "epoch": 0.76864, + "grad_norm": 0.4224303364753723, + "learning_rate": 0.00013389064606796096, + "loss": 2.6623, + "step": 6005 + }, + { + "epoch": 0.768768, + "grad_norm": 0.4353911280632019, + "learning_rate": 0.00013374948424780837, + "loss": 2.5873, + "step": 6006 + }, + { + "epoch": 0.768896, + "grad_norm": 0.5653361678123474, + "learning_rate": 0.0001336083853903653, + "loss": 2.6556, + "step": 6007 + }, + { + "epoch": 0.769024, + "grad_norm": 0.38783666491508484, + "learning_rate": 0.00013346734951988842, + "loss": 2.5966, + "step": 6008 + }, + { + "epoch": 0.769152, + "grad_norm": 0.4894041419029236, + "learning_rate": 0.0001333263766606232, + "loss": 2.7302, + "step": 6009 + }, + { + "epoch": 0.76928, + "grad_norm": 0.39207038283348083, + "learning_rate": 0.00013318546683680483, + "loss": 2.5903, + "step": 6010 + }, + { + "epoch": 0.769408, + "grad_norm": 0.4287959337234497, + "learning_rate": 0.00013304462007265717, + "loss": 2.7138, + "step": 6011 + }, + { + "epoch": 0.769536, + "grad_norm": 0.3942119777202606, + "learning_rate": 0.00013290383639239344, + "loss": 2.584, + "step": 6012 + }, + { + "epoch": 0.769664, + "grad_norm": 0.37009304761886597, + "learning_rate": 0.00013276311582021582, + "loss": 2.538, + "step": 6013 + }, + { + "epoch": 0.769792, + "grad_norm": 0.4277808964252472, + "learning_rate": 0.00013262245838031617, + "loss": 2.5587, + "step": 6014 + }, + { + "epoch": 0.76992, + "grad_norm": 0.39824414253234863, + "learning_rate": 0.0001324818640968749, + "loss": 2.5348, + "step": 6015 + }, + { + "epoch": 0.770048, + "grad_norm": 0.45082005858421326, + "learning_rate": 0.0001323413329940618, + "loss": 2.6779, + "step": 6016 + }, + { + "epoch": 0.770176, + "grad_norm": 0.45401063561439514, + "learning_rate": 0.00013220086509603567, + "loss": 2.6322, + "step": 6017 + }, + { + "epoch": 0.770304, + "grad_norm": 0.42173635959625244, + "learning_rate": 0.00013206046042694492, + "loss": 2.4997, + "step": 6018 + }, + { + "epoch": 0.770432, + "grad_norm": 0.4384644031524658, + "learning_rate": 0.00013192011901092655, + "loss": 2.5749, + "step": 6019 + }, + { + "epoch": 0.77056, + "grad_norm": 1.459173321723938, + "learning_rate": 0.0001317798408721068, + "loss": 2.4842, + "step": 6020 + }, + { + "epoch": 0.770688, + "grad_norm": 0.4311870038509369, + "learning_rate": 0.0001316396260346012, + "loss": 2.6772, + "step": 6021 + }, + { + "epoch": 0.770816, + "grad_norm": 0.41069552302360535, + "learning_rate": 0.0001314994745225142, + "loss": 2.6156, + "step": 6022 + }, + { + "epoch": 0.770944, + "grad_norm": 0.3610895872116089, + "learning_rate": 0.00013135938635993965, + "loss": 2.4605, + "step": 6023 + }, + { + "epoch": 0.771072, + "grad_norm": 0.4874170124530792, + "learning_rate": 0.0001312193615709601, + "loss": 2.6532, + "step": 6024 + }, + { + "epoch": 0.7712, + "grad_norm": 0.3981195092201233, + "learning_rate": 0.0001310794001796477, + "loss": 2.5126, + "step": 6025 + }, + { + "epoch": 0.771328, + "grad_norm": 0.40881311893463135, + "learning_rate": 0.00013093950221006328, + "loss": 2.476, + "step": 6026 + }, + { + "epoch": 0.771456, + "grad_norm": 0.39507657289505005, + "learning_rate": 0.000130799667686257, + "loss": 2.675, + "step": 6027 + }, + { + "epoch": 0.771584, + "grad_norm": 0.4423034191131592, + "learning_rate": 0.00013065989663226795, + "loss": 2.7164, + "step": 6028 + }, + { + "epoch": 0.771712, + "grad_norm": 0.3992152214050293, + "learning_rate": 0.00013052018907212448, + "loss": 2.6141, + "step": 6029 + }, + { + "epoch": 0.77184, + "grad_norm": 0.6502902507781982, + "learning_rate": 0.0001303805450298437, + "loss": 2.7038, + "step": 6030 + }, + { + "epoch": 0.771968, + "grad_norm": 0.41722235083580017, + "learning_rate": 0.00013024096452943234, + "loss": 2.5915, + "step": 6031 + }, + { + "epoch": 0.772096, + "grad_norm": 0.42524388432502747, + "learning_rate": 0.00013010144759488567, + "loss": 2.589, + "step": 6032 + }, + { + "epoch": 0.772224, + "grad_norm": 0.3829645812511444, + "learning_rate": 0.00012996199425018855, + "loss": 2.5695, + "step": 6033 + }, + { + "epoch": 0.772352, + "grad_norm": 0.4347381591796875, + "learning_rate": 0.00012982260451931444, + "loss": 2.6105, + "step": 6034 + }, + { + "epoch": 0.77248, + "grad_norm": 0.40310269594192505, + "learning_rate": 0.00012968327842622612, + "loss": 2.599, + "step": 6035 + }, + { + "epoch": 0.772608, + "grad_norm": 0.4387594759464264, + "learning_rate": 0.0001295440159948753, + "loss": 2.5764, + "step": 6036 + }, + { + "epoch": 0.772736, + "grad_norm": 0.4947674572467804, + "learning_rate": 0.00012940481724920284, + "loss": 2.5464, + "step": 6037 + }, + { + "epoch": 0.772864, + "grad_norm": 0.39959707856178284, + "learning_rate": 0.0001292656822131384, + "loss": 2.6558, + "step": 6038 + }, + { + "epoch": 0.772992, + "grad_norm": 0.4070192873477936, + "learning_rate": 0.0001291266109106013, + "loss": 2.7143, + "step": 6039 + }, + { + "epoch": 0.77312, + "grad_norm": 0.4066345989704132, + "learning_rate": 0.0001289876033654993, + "loss": 2.5375, + "step": 6040 + }, + { + "epoch": 0.773248, + "grad_norm": 0.39163923263549805, + "learning_rate": 0.0001288486596017293, + "loss": 2.6229, + "step": 6041 + }, + { + "epoch": 0.773376, + "grad_norm": 0.36789408326148987, + "learning_rate": 0.0001287097796431776, + "loss": 2.5121, + "step": 6042 + }, + { + "epoch": 0.773504, + "grad_norm": 0.44676944613456726, + "learning_rate": 0.00012857096351371917, + "loss": 2.6655, + "step": 6043 + }, + { + "epoch": 0.773632, + "grad_norm": 0.4007432162761688, + "learning_rate": 0.00012843221123721803, + "loss": 2.6109, + "step": 6044 + }, + { + "epoch": 0.77376, + "grad_norm": 0.42772090435028076, + "learning_rate": 0.00012829352283752726, + "loss": 2.6084, + "step": 6045 + }, + { + "epoch": 0.773888, + "grad_norm": 0.4334762394428253, + "learning_rate": 0.00012815489833848927, + "loss": 2.6473, + "step": 6046 + }, + { + "epoch": 0.774016, + "grad_norm": 0.4699229598045349, + "learning_rate": 0.00012801633776393501, + "loss": 2.4626, + "step": 6047 + }, + { + "epoch": 0.774144, + "grad_norm": 0.4210101068019867, + "learning_rate": 0.0001278778411376847, + "loss": 2.5666, + "step": 6048 + }, + { + "epoch": 0.774272, + "grad_norm": 0.41926348209381104, + "learning_rate": 0.00012773940848354753, + "loss": 2.5959, + "step": 6049 + }, + { + "epoch": 0.7744, + "grad_norm": 0.4410623610019684, + "learning_rate": 0.0001276010398253215, + "loss": 2.6689, + "step": 6050 + }, + { + "epoch": 0.774528, + "grad_norm": 0.3991164565086365, + "learning_rate": 0.00012746273518679407, + "loss": 2.6387, + "step": 6051 + }, + { + "epoch": 0.774656, + "grad_norm": 0.42062366008758545, + "learning_rate": 0.00012732449459174134, + "loss": 2.4776, + "step": 6052 + }, + { + "epoch": 0.774784, + "grad_norm": 0.4154279828071594, + "learning_rate": 0.00012718631806392822, + "loss": 2.5461, + "step": 6053 + }, + { + "epoch": 0.774912, + "grad_norm": 0.4180595278739929, + "learning_rate": 0.00012704820562710922, + "loss": 2.5671, + "step": 6054 + }, + { + "epoch": 0.77504, + "grad_norm": 0.44801971316337585, + "learning_rate": 0.0001269101573050273, + "loss": 2.6001, + "step": 6055 + }, + { + "epoch": 0.775168, + "grad_norm": 0.3490329682826996, + "learning_rate": 0.00012677217312141454, + "loss": 2.543, + "step": 6056 + }, + { + "epoch": 0.775296, + "grad_norm": 0.4102097153663635, + "learning_rate": 0.00012663425309999204, + "loss": 2.5972, + "step": 6057 + }, + { + "epoch": 0.775424, + "grad_norm": 0.40814492106437683, + "learning_rate": 0.00012649639726446994, + "loss": 2.6344, + "step": 6058 + }, + { + "epoch": 0.775552, + "grad_norm": 0.42263761162757874, + "learning_rate": 0.00012635860563854696, + "loss": 2.6185, + "step": 6059 + }, + { + "epoch": 0.77568, + "grad_norm": 0.46357741951942444, + "learning_rate": 0.00012622087824591127, + "loss": 2.5598, + "step": 6060 + }, + { + "epoch": 0.775808, + "grad_norm": 0.4117509424686432, + "learning_rate": 0.00012608321511024006, + "loss": 2.6204, + "step": 6061 + }, + { + "epoch": 0.775936, + "grad_norm": 0.3964688777923584, + "learning_rate": 0.0001259456162551989, + "loss": 2.5746, + "step": 6062 + }, + { + "epoch": 0.776064, + "grad_norm": 0.42731478810310364, + "learning_rate": 0.0001258080817044428, + "loss": 2.621, + "step": 6063 + }, + { + "epoch": 0.776192, + "grad_norm": 0.3736017048358917, + "learning_rate": 0.0001256706114816154, + "loss": 2.6466, + "step": 6064 + }, + { + "epoch": 0.77632, + "grad_norm": 0.41515132784843445, + "learning_rate": 0.00012553320561034954, + "loss": 2.5366, + "step": 6065 + }, + { + "epoch": 0.776448, + "grad_norm": 0.4259808361530304, + "learning_rate": 0.00012539586411426662, + "loss": 2.5695, + "step": 6066 + }, + { + "epoch": 0.776576, + "grad_norm": 0.39585670828819275, + "learning_rate": 0.00012525858701697761, + "loss": 2.4992, + "step": 6067 + }, + { + "epoch": 0.776704, + "grad_norm": 0.41548046469688416, + "learning_rate": 0.00012512137434208187, + "loss": 2.6208, + "step": 6068 + }, + { + "epoch": 0.776832, + "grad_norm": 0.445583701133728, + "learning_rate": 0.00012498422611316768, + "loss": 2.4992, + "step": 6069 + }, + { + "epoch": 0.77696, + "grad_norm": 0.4260144531726837, + "learning_rate": 0.00012484714235381277, + "loss": 2.5115, + "step": 6070 + }, + { + "epoch": 0.777088, + "grad_norm": 0.37480854988098145, + "learning_rate": 0.00012471012308758322, + "loss": 2.5682, + "step": 6071 + }, + { + "epoch": 0.777216, + "grad_norm": 0.3887835443019867, + "learning_rate": 0.00012457316833803421, + "loss": 2.5231, + "step": 6072 + }, + { + "epoch": 0.777344, + "grad_norm": 0.6635140776634216, + "learning_rate": 0.00012443627812870988, + "loss": 2.5612, + "step": 6073 + }, + { + "epoch": 0.777472, + "grad_norm": 0.40268394351005554, + "learning_rate": 0.00012429945248314318, + "loss": 2.6871, + "step": 6074 + }, + { + "epoch": 0.7776, + "grad_norm": 0.40202364325523376, + "learning_rate": 0.00012416269142485615, + "loss": 2.4837, + "step": 6075 + }, + { + "epoch": 0.777728, + "grad_norm": 0.3684506118297577, + "learning_rate": 0.0001240259949773596, + "loss": 2.516, + "step": 6076 + }, + { + "epoch": 0.777856, + "grad_norm": 0.4249964952468872, + "learning_rate": 0.00012388936316415317, + "loss": 2.5721, + "step": 6077 + }, + { + "epoch": 0.777984, + "grad_norm": 0.41694092750549316, + "learning_rate": 0.0001237527960087253, + "loss": 2.558, + "step": 6078 + }, + { + "epoch": 0.778112, + "grad_norm": 0.3742409944534302, + "learning_rate": 0.0001236162935345538, + "loss": 2.5948, + "step": 6079 + }, + { + "epoch": 0.77824, + "grad_norm": 0.3880797326564789, + "learning_rate": 0.0001234798557651048, + "loss": 2.6868, + "step": 6080 + }, + { + "epoch": 0.778368, + "grad_norm": 0.4104139506816864, + "learning_rate": 0.0001233434827238335, + "loss": 2.5968, + "step": 6081 + }, + { + "epoch": 0.778496, + "grad_norm": 0.41453614830970764, + "learning_rate": 0.00012320717443418423, + "loss": 2.5422, + "step": 6082 + }, + { + "epoch": 0.778624, + "grad_norm": 0.434508353471756, + "learning_rate": 0.0001230709309195898, + "loss": 2.6479, + "step": 6083 + }, + { + "epoch": 0.778752, + "grad_norm": 0.4860893785953522, + "learning_rate": 0.00012293475220347212, + "loss": 2.5838, + "step": 6084 + }, + { + "epoch": 0.77888, + "grad_norm": 0.3768160045146942, + "learning_rate": 0.0001227986383092418, + "loss": 2.509, + "step": 6085 + }, + { + "epoch": 0.779008, + "grad_norm": 0.39229536056518555, + "learning_rate": 0.0001226625892602985, + "loss": 2.5807, + "step": 6086 + }, + { + "epoch": 0.779136, + "grad_norm": 0.4675126373767853, + "learning_rate": 0.00012252660508003043, + "loss": 2.5926, + "step": 6087 + }, + { + "epoch": 0.779264, + "grad_norm": 0.4180111885070801, + "learning_rate": 0.00012239068579181494, + "loss": 2.4469, + "step": 6088 + }, + { + "epoch": 0.779392, + "grad_norm": 0.358153760433197, + "learning_rate": 0.00012225483141901837, + "loss": 2.6102, + "step": 6089 + }, + { + "epoch": 0.77952, + "grad_norm": 0.4402056336402893, + "learning_rate": 0.00012211904198499552, + "loss": 2.5655, + "step": 6090 + }, + { + "epoch": 0.779648, + "grad_norm": 0.38750237226486206, + "learning_rate": 0.00012198331751309005, + "loss": 2.6252, + "step": 6091 + }, + { + "epoch": 0.779776, + "grad_norm": 0.39174678921699524, + "learning_rate": 0.00012184765802663466, + "loss": 2.541, + "step": 6092 + }, + { + "epoch": 0.779904, + "grad_norm": 0.4234752058982849, + "learning_rate": 0.00012171206354895081, + "loss": 2.6459, + "step": 6093 + }, + { + "epoch": 0.780032, + "grad_norm": 0.37854552268981934, + "learning_rate": 0.00012157653410334874, + "loss": 2.541, + "step": 6094 + }, + { + "epoch": 0.78016, + "grad_norm": 0.3849351108074188, + "learning_rate": 0.0001214410697131273, + "loss": 2.5979, + "step": 6095 + }, + { + "epoch": 0.780288, + "grad_norm": 0.42502111196517944, + "learning_rate": 0.00012130567040157465, + "loss": 2.5929, + "step": 6096 + }, + { + "epoch": 0.780416, + "grad_norm": 0.4367362856864929, + "learning_rate": 0.0001211703361919676, + "loss": 2.6548, + "step": 6097 + }, + { + "epoch": 0.780544, + "grad_norm": 0.45486143231391907, + "learning_rate": 0.00012103506710757156, + "loss": 2.5115, + "step": 6098 + }, + { + "epoch": 0.780672, + "grad_norm": 0.38279542326927185, + "learning_rate": 0.00012089986317164075, + "loss": 2.5333, + "step": 6099 + }, + { + "epoch": 0.7808, + "grad_norm": 0.46876493096351624, + "learning_rate": 0.00012076472440741842, + "loss": 2.6656, + "step": 6100 + }, + { + "epoch": 0.780928, + "grad_norm": 0.42411431670188904, + "learning_rate": 0.00012062965083813642, + "loss": 2.714, + "step": 6101 + }, + { + "epoch": 0.781056, + "grad_norm": 0.3856874704360962, + "learning_rate": 0.00012049464248701536, + "loss": 2.6518, + "step": 6102 + }, + { + "epoch": 0.781184, + "grad_norm": 0.4524640142917633, + "learning_rate": 0.00012035969937726498, + "loss": 2.6513, + "step": 6103 + }, + { + "epoch": 0.781312, + "grad_norm": 0.37456026673316956, + "learning_rate": 0.00012022482153208347, + "loss": 2.5192, + "step": 6104 + }, + { + "epoch": 0.78144, + "grad_norm": 0.40767011046409607, + "learning_rate": 0.00012009000897465783, + "loss": 2.5338, + "step": 6105 + }, + { + "epoch": 0.781568, + "grad_norm": 0.4316698908805847, + "learning_rate": 0.00011995526172816384, + "loss": 2.5836, + "step": 6106 + }, + { + "epoch": 0.781696, + "grad_norm": 0.38462188839912415, + "learning_rate": 0.00011982057981576632, + "loss": 2.5887, + "step": 6107 + }, + { + "epoch": 0.781824, + "grad_norm": 0.4326455891132355, + "learning_rate": 0.00011968596326061858, + "loss": 2.5405, + "step": 6108 + }, + { + "epoch": 0.781952, + "grad_norm": 0.40543439984321594, + "learning_rate": 0.00011955141208586257, + "loss": 2.5763, + "step": 6109 + }, + { + "epoch": 0.78208, + "grad_norm": 0.4252394735813141, + "learning_rate": 0.00011941692631462953, + "loss": 2.6506, + "step": 6110 + }, + { + "epoch": 0.782208, + "grad_norm": 0.4076583683490753, + "learning_rate": 0.00011928250597003892, + "loss": 2.5649, + "step": 6111 + }, + { + "epoch": 0.782336, + "grad_norm": 0.41954389214515686, + "learning_rate": 0.00011914815107519922, + "loss": 2.574, + "step": 6112 + }, + { + "epoch": 0.782464, + "grad_norm": 0.40812399983406067, + "learning_rate": 0.00011901386165320755, + "loss": 2.6252, + "step": 6113 + }, + { + "epoch": 0.782592, + "grad_norm": 0.4459620714187622, + "learning_rate": 0.00011887963772714983, + "loss": 2.501, + "step": 6114 + }, + { + "epoch": 0.78272, + "grad_norm": 0.453479140996933, + "learning_rate": 0.00011874547932010055, + "loss": 2.6336, + "step": 6115 + }, + { + "epoch": 0.782848, + "grad_norm": 0.4174945056438446, + "learning_rate": 0.00011861138645512343, + "loss": 2.5459, + "step": 6116 + }, + { + "epoch": 0.782976, + "grad_norm": 0.40557554364204407, + "learning_rate": 0.00011847735915527025, + "loss": 2.5735, + "step": 6117 + }, + { + "epoch": 0.783104, + "grad_norm": 0.3878689110279083, + "learning_rate": 0.00011834339744358218, + "loss": 2.6048, + "step": 6118 + }, + { + "epoch": 0.783232, + "grad_norm": 0.4497729241847992, + "learning_rate": 0.0001182095013430886, + "loss": 2.5933, + "step": 6119 + }, + { + "epoch": 0.78336, + "grad_norm": 0.4724157154560089, + "learning_rate": 0.00011807567087680787, + "loss": 2.5397, + "step": 6120 + }, + { + "epoch": 0.783488, + "grad_norm": 0.39902248978614807, + "learning_rate": 0.00011794190606774697, + "loss": 2.4629, + "step": 6121 + }, + { + "epoch": 0.783616, + "grad_norm": 0.39661210775375366, + "learning_rate": 0.0001178082069389016, + "loss": 2.6175, + "step": 6122 + }, + { + "epoch": 0.783744, + "grad_norm": 0.3682713508605957, + "learning_rate": 0.00011767457351325605, + "loss": 2.5086, + "step": 6123 + }, + { + "epoch": 0.783872, + "grad_norm": 0.368360310792923, + "learning_rate": 0.00011754100581378364, + "loss": 2.5417, + "step": 6124 + }, + { + "epoch": 0.784, + "grad_norm": 0.39216670393943787, + "learning_rate": 0.00011740750386344628, + "loss": 2.6079, + "step": 6125 + }, + { + "epoch": 0.784128, + "grad_norm": 0.4417192339897156, + "learning_rate": 0.00011727406768519445, + "loss": 2.5864, + "step": 6126 + }, + { + "epoch": 0.784256, + "grad_norm": 0.41959866881370544, + "learning_rate": 0.0001171406973019673, + "loss": 2.5532, + "step": 6127 + }, + { + "epoch": 0.784384, + "grad_norm": 0.46297940611839294, + "learning_rate": 0.00011700739273669275, + "loss": 2.5196, + "step": 6128 + }, + { + "epoch": 0.784512, + "grad_norm": 0.5041238069534302, + "learning_rate": 0.0001168741540122874, + "loss": 2.5943, + "step": 6129 + }, + { + "epoch": 0.78464, + "grad_norm": 0.44284334778785706, + "learning_rate": 0.00011674098115165643, + "loss": 2.5374, + "step": 6130 + }, + { + "epoch": 0.784768, + "grad_norm": 0.4281841516494751, + "learning_rate": 0.00011660787417769409, + "loss": 2.5738, + "step": 6131 + }, + { + "epoch": 0.784896, + "grad_norm": 0.4711056351661682, + "learning_rate": 0.00011647483311328283, + "loss": 2.6766, + "step": 6132 + }, + { + "epoch": 0.785024, + "grad_norm": 0.5705511569976807, + "learning_rate": 0.00011634185798129382, + "loss": 2.6927, + "step": 6133 + }, + { + "epoch": 0.785152, + "grad_norm": 0.7353594899177551, + "learning_rate": 0.00011620894880458732, + "loss": 2.6067, + "step": 6134 + }, + { + "epoch": 0.78528, + "grad_norm": 1.2990829944610596, + "learning_rate": 0.00011607610560601179, + "loss": 2.5848, + "step": 6135 + }, + { + "epoch": 0.785408, + "grad_norm": 0.5837934017181396, + "learning_rate": 0.00011594332840840455, + "loss": 2.6133, + "step": 6136 + }, + { + "epoch": 0.785536, + "grad_norm": 0.41487887501716614, + "learning_rate": 0.00011581061723459153, + "loss": 2.493, + "step": 6137 + }, + { + "epoch": 0.785664, + "grad_norm": 0.3844457268714905, + "learning_rate": 0.00011567797210738712, + "loss": 2.6079, + "step": 6138 + }, + { + "epoch": 0.785792, + "grad_norm": 0.44623324275016785, + "learning_rate": 0.00011554539304959494, + "loss": 2.5307, + "step": 6139 + }, + { + "epoch": 0.78592, + "grad_norm": 0.8843861222267151, + "learning_rate": 0.00011541288008400663, + "loss": 2.5407, + "step": 6140 + }, + { + "epoch": 0.786048, + "grad_norm": 0.5175185799598694, + "learning_rate": 0.00011528043323340281, + "loss": 2.5654, + "step": 6141 + }, + { + "epoch": 0.786176, + "grad_norm": 0.38398343324661255, + "learning_rate": 0.0001151480525205525, + "loss": 2.5806, + "step": 6142 + }, + { + "epoch": 0.786304, + "grad_norm": 0.3955792188644409, + "learning_rate": 0.00011501573796821346, + "loss": 2.6059, + "step": 6143 + }, + { + "epoch": 0.786432, + "grad_norm": 0.40941405296325684, + "learning_rate": 0.0001148834895991323, + "loss": 2.6732, + "step": 6144 + }, + { + "epoch": 0.78656, + "grad_norm": 0.4019754230976105, + "learning_rate": 0.00011475130743604379, + "loss": 2.5454, + "step": 6145 + }, + { + "epoch": 0.786688, + "grad_norm": 0.4227709174156189, + "learning_rate": 0.00011461919150167188, + "loss": 2.5984, + "step": 6146 + }, + { + "epoch": 0.786816, + "grad_norm": 0.40379685163497925, + "learning_rate": 0.00011448714181872865, + "loss": 2.4909, + "step": 6147 + }, + { + "epoch": 0.786944, + "grad_norm": 0.4228176176548004, + "learning_rate": 0.000114355158409915, + "loss": 2.4767, + "step": 6148 + }, + { + "epoch": 0.787072, + "grad_norm": 0.369876891374588, + "learning_rate": 0.00011422324129792039, + "loss": 2.598, + "step": 6149 + }, + { + "epoch": 0.7872, + "grad_norm": 0.40793362259864807, + "learning_rate": 0.00011409139050542295, + "loss": 2.5376, + "step": 6150 + }, + { + "epoch": 0.787328, + "grad_norm": 0.4096187949180603, + "learning_rate": 0.00011395960605508915, + "loss": 2.6154, + "step": 6151 + }, + { + "epoch": 0.787456, + "grad_norm": 0.37260255217552185, + "learning_rate": 0.00011382788796957454, + "loss": 2.5821, + "step": 6152 + }, + { + "epoch": 0.787584, + "grad_norm": 0.3761984705924988, + "learning_rate": 0.00011369623627152276, + "loss": 2.6707, + "step": 6153 + }, + { + "epoch": 0.787712, + "grad_norm": 0.42467889189720154, + "learning_rate": 0.00011356465098356656, + "loss": 2.7302, + "step": 6154 + }, + { + "epoch": 0.78784, + "grad_norm": 0.393588125705719, + "learning_rate": 0.00011343313212832674, + "loss": 2.522, + "step": 6155 + }, + { + "epoch": 0.787968, + "grad_norm": 0.435018390417099, + "learning_rate": 0.00011330167972841299, + "loss": 2.614, + "step": 6156 + }, + { + "epoch": 0.788096, + "grad_norm": 0.41986948251724243, + "learning_rate": 0.00011317029380642352, + "loss": 2.5883, + "step": 6157 + }, + { + "epoch": 0.788224, + "grad_norm": 0.4088584780693054, + "learning_rate": 0.00011303897438494504, + "loss": 2.4108, + "step": 6158 + }, + { + "epoch": 0.788352, + "grad_norm": 0.39886975288391113, + "learning_rate": 0.00011290772148655276, + "loss": 2.5597, + "step": 6159 + }, + { + "epoch": 0.78848, + "grad_norm": 0.4684580862522125, + "learning_rate": 0.0001127765351338108, + "loss": 2.6274, + "step": 6160 + }, + { + "epoch": 0.788608, + "grad_norm": 0.3745497465133667, + "learning_rate": 0.00011264541534927147, + "loss": 2.6764, + "step": 6161 + }, + { + "epoch": 0.788736, + "grad_norm": 0.4433757960796356, + "learning_rate": 0.00011251436215547595, + "loss": 2.5414, + "step": 6162 + }, + { + "epoch": 0.788864, + "grad_norm": 0.4787682592868805, + "learning_rate": 0.0001123833755749537, + "loss": 2.5988, + "step": 6163 + }, + { + "epoch": 0.788992, + "grad_norm": 0.4084990620613098, + "learning_rate": 0.00011225245563022285, + "loss": 2.6468, + "step": 6164 + }, + { + "epoch": 0.78912, + "grad_norm": 0.6452922821044922, + "learning_rate": 0.00011212160234378999, + "loss": 2.5464, + "step": 6165 + }, + { + "epoch": 0.789248, + "grad_norm": 0.4354986250400543, + "learning_rate": 0.00011199081573815024, + "loss": 2.6661, + "step": 6166 + }, + { + "epoch": 0.789376, + "grad_norm": 0.36521950364112854, + "learning_rate": 0.0001118600958357876, + "loss": 2.5917, + "step": 6167 + }, + { + "epoch": 0.789504, + "grad_norm": 0.4024215042591095, + "learning_rate": 0.00011172944265917417, + "loss": 2.5067, + "step": 6168 + }, + { + "epoch": 0.789632, + "grad_norm": 0.3836224377155304, + "learning_rate": 0.00011159885623077076, + "loss": 2.6186, + "step": 6169 + }, + { + "epoch": 0.78976, + "grad_norm": 0.39075860381126404, + "learning_rate": 0.00011146833657302658, + "loss": 2.667, + "step": 6170 + }, + { + "epoch": 0.789888, + "grad_norm": 0.431636780500412, + "learning_rate": 0.00011133788370837972, + "loss": 2.6856, + "step": 6171 + }, + { + "epoch": 0.790016, + "grad_norm": 0.4108085036277771, + "learning_rate": 0.00011120749765925642, + "loss": 2.5703, + "step": 6172 + }, + { + "epoch": 0.790144, + "grad_norm": 0.4094730019569397, + "learning_rate": 0.00011107717844807153, + "loss": 2.5466, + "step": 6173 + }, + { + "epoch": 0.790272, + "grad_norm": 0.4315975308418274, + "learning_rate": 0.00011094692609722829, + "loss": 2.6566, + "step": 6174 + }, + { + "epoch": 0.7904, + "grad_norm": 0.38948121666908264, + "learning_rate": 0.0001108167406291189, + "loss": 2.6243, + "step": 6175 + }, + { + "epoch": 0.790528, + "grad_norm": 0.43086501955986023, + "learning_rate": 0.00011068662206612362, + "loss": 2.5946, + "step": 6176 + }, + { + "epoch": 0.790656, + "grad_norm": 0.3917463421821594, + "learning_rate": 0.00011055657043061124, + "loss": 2.6926, + "step": 6177 + }, + { + "epoch": 0.790784, + "grad_norm": 0.5056282877922058, + "learning_rate": 0.0001104265857449393, + "loss": 2.5089, + "step": 6178 + }, + { + "epoch": 0.790912, + "grad_norm": 0.3781290352344513, + "learning_rate": 0.00011029666803145355, + "loss": 2.5305, + "step": 6179 + }, + { + "epoch": 0.79104, + "grad_norm": 0.4113095998764038, + "learning_rate": 0.00011016681731248829, + "loss": 2.4985, + "step": 6180 + }, + { + "epoch": 0.791168, + "grad_norm": 0.44862794876098633, + "learning_rate": 0.00011003703361036643, + "loss": 2.6886, + "step": 6181 + }, + { + "epoch": 0.791296, + "grad_norm": 0.4014025032520294, + "learning_rate": 0.00010990731694739947, + "loss": 2.6505, + "step": 6182 + }, + { + "epoch": 0.791424, + "grad_norm": 0.39642563462257385, + "learning_rate": 0.00010977766734588706, + "loss": 2.5966, + "step": 6183 + }, + { + "epoch": 0.791552, + "grad_norm": 0.398286372423172, + "learning_rate": 0.0001096480848281175, + "loss": 2.5619, + "step": 6184 + }, + { + "epoch": 0.79168, + "grad_norm": 0.37700319290161133, + "learning_rate": 0.00010951856941636751, + "loss": 2.5545, + "step": 6185 + }, + { + "epoch": 0.791808, + "grad_norm": 0.5551208853721619, + "learning_rate": 0.00010938912113290229, + "loss": 2.5873, + "step": 6186 + }, + { + "epoch": 0.791936, + "grad_norm": 0.3730669617652893, + "learning_rate": 0.00010925973999997535, + "loss": 2.5771, + "step": 6187 + }, + { + "epoch": 0.792064, + "grad_norm": 0.4099901020526886, + "learning_rate": 0.00010913042603982909, + "loss": 2.6195, + "step": 6188 + }, + { + "epoch": 0.792192, + "grad_norm": 0.39384856820106506, + "learning_rate": 0.00010900117927469383, + "loss": 2.6733, + "step": 6189 + }, + { + "epoch": 0.79232, + "grad_norm": 0.4566054046154022, + "learning_rate": 0.0001088719997267888, + "loss": 2.6953, + "step": 6190 + }, + { + "epoch": 0.792448, + "grad_norm": 0.40621981024742126, + "learning_rate": 0.00010874288741832139, + "loss": 2.6396, + "step": 6191 + }, + { + "epoch": 0.792576, + "grad_norm": 0.39121052622795105, + "learning_rate": 0.00010861384237148748, + "loss": 2.651, + "step": 6192 + }, + { + "epoch": 0.792704, + "grad_norm": 0.4509212076663971, + "learning_rate": 0.00010848486460847134, + "loss": 2.579, + "step": 6193 + }, + { + "epoch": 0.792832, + "grad_norm": 0.4166094958782196, + "learning_rate": 0.00010835595415144588, + "loss": 2.485, + "step": 6194 + }, + { + "epoch": 0.79296, + "grad_norm": 0.41125476360321045, + "learning_rate": 0.00010822711102257204, + "loss": 2.5514, + "step": 6195 + }, + { + "epoch": 0.793088, + "grad_norm": 0.43993499875068665, + "learning_rate": 0.00010809833524399982, + "loss": 2.6849, + "step": 6196 + }, + { + "epoch": 0.793216, + "grad_norm": 0.4208594262599945, + "learning_rate": 0.00010796962683786709, + "loss": 2.6484, + "step": 6197 + }, + { + "epoch": 0.793344, + "grad_norm": 0.44704702496528625, + "learning_rate": 0.00010784098582630015, + "loss": 2.4935, + "step": 6198 + }, + { + "epoch": 0.793472, + "grad_norm": 0.40129876136779785, + "learning_rate": 0.00010771241223141415, + "loss": 2.5282, + "step": 6199 + }, + { + "epoch": 0.7936, + "grad_norm": 0.39394524693489075, + "learning_rate": 0.00010758390607531232, + "loss": 2.5556, + "step": 6200 + }, + { + "epoch": 0.793728, + "grad_norm": 0.4036063253879547, + "learning_rate": 0.00010745546738008627, + "loss": 2.4914, + "step": 6201 + }, + { + "epoch": 0.793856, + "grad_norm": 0.473423033952713, + "learning_rate": 0.00010732709616781606, + "loss": 2.5779, + "step": 6202 + }, + { + "epoch": 0.793984, + "grad_norm": 0.40421062707901, + "learning_rate": 0.0001071987924605704, + "loss": 2.5134, + "step": 6203 + }, + { + "epoch": 0.794112, + "grad_norm": 0.40572938323020935, + "learning_rate": 0.00010707055628040602, + "loss": 2.6059, + "step": 6204 + }, + { + "epoch": 0.79424, + "grad_norm": 0.436573326587677, + "learning_rate": 0.00010694238764936826, + "loss": 2.5578, + "step": 6205 + }, + { + "epoch": 0.794368, + "grad_norm": 0.4652629494667053, + "learning_rate": 0.00010681428658949082, + "loss": 2.506, + "step": 6206 + }, + { + "epoch": 0.794496, + "grad_norm": 0.38754796981811523, + "learning_rate": 0.00010668625312279551, + "loss": 2.5581, + "step": 6207 + }, + { + "epoch": 0.794624, + "grad_norm": 0.44858866930007935, + "learning_rate": 0.00010655828727129318, + "loss": 2.6965, + "step": 6208 + }, + { + "epoch": 0.794752, + "grad_norm": 0.37972113490104675, + "learning_rate": 0.00010643038905698237, + "loss": 2.5429, + "step": 6209 + }, + { + "epoch": 0.79488, + "grad_norm": 0.42003586888313293, + "learning_rate": 0.00010630255850185022, + "loss": 2.6109, + "step": 6210 + }, + { + "epoch": 0.795008, + "grad_norm": 0.42145484685897827, + "learning_rate": 0.00010617479562787258, + "loss": 2.4636, + "step": 6211 + }, + { + "epoch": 0.795136, + "grad_norm": 0.41739118099212646, + "learning_rate": 0.0001060471004570131, + "loss": 2.5788, + "step": 6212 + }, + { + "epoch": 0.795264, + "grad_norm": 0.4431496560573578, + "learning_rate": 0.00010591947301122424, + "loss": 2.6488, + "step": 6213 + }, + { + "epoch": 0.795392, + "grad_norm": 0.39537879824638367, + "learning_rate": 0.00010579191331244648, + "loss": 2.6327, + "step": 6214 + }, + { + "epoch": 0.79552, + "grad_norm": 0.39421266317367554, + "learning_rate": 0.00010566442138260884, + "loss": 2.6525, + "step": 6215 + }, + { + "epoch": 0.795648, + "grad_norm": 0.4374276101589203, + "learning_rate": 0.00010553699724362865, + "loss": 2.6845, + "step": 6216 + }, + { + "epoch": 0.795776, + "grad_norm": 0.41279447078704834, + "learning_rate": 0.00010540964091741157, + "loss": 2.5261, + "step": 6217 + }, + { + "epoch": 0.795904, + "grad_norm": 0.41928228735923767, + "learning_rate": 0.00010528235242585187, + "loss": 2.4742, + "step": 6218 + }, + { + "epoch": 0.796032, + "grad_norm": 0.38649114966392517, + "learning_rate": 0.00010515513179083175, + "loss": 2.5985, + "step": 6219 + }, + { + "epoch": 0.79616, + "grad_norm": 0.43657806515693665, + "learning_rate": 0.0001050279790342219, + "loss": 2.7265, + "step": 6220 + }, + { + "epoch": 0.796288, + "grad_norm": 0.4380522072315216, + "learning_rate": 0.00010490089417788134, + "loss": 2.581, + "step": 6221 + }, + { + "epoch": 0.796416, + "grad_norm": 0.36412376165390015, + "learning_rate": 0.0001047738772436575, + "loss": 2.4909, + "step": 6222 + }, + { + "epoch": 0.796544, + "grad_norm": 0.3958243429660797, + "learning_rate": 0.00010464692825338585, + "loss": 2.6718, + "step": 6223 + }, + { + "epoch": 0.796672, + "grad_norm": 0.4250694513320923, + "learning_rate": 0.00010452004722889069, + "loss": 2.7076, + "step": 6224 + }, + { + "epoch": 0.7968, + "grad_norm": 0.4062851667404175, + "learning_rate": 0.00010439323419198421, + "loss": 2.5692, + "step": 6225 + }, + { + "epoch": 0.796928, + "grad_norm": 0.3936968445777893, + "learning_rate": 0.0001042664891644669, + "loss": 2.5785, + "step": 6226 + }, + { + "epoch": 0.797056, + "grad_norm": 0.44450563192367554, + "learning_rate": 0.00010413981216812801, + "loss": 2.6699, + "step": 6227 + }, + { + "epoch": 0.797184, + "grad_norm": 0.42522838711738586, + "learning_rate": 0.00010401320322474456, + "loss": 2.5348, + "step": 6228 + }, + { + "epoch": 0.797312, + "grad_norm": 0.4095156192779541, + "learning_rate": 0.00010388666235608219, + "loss": 2.5856, + "step": 6229 + }, + { + "epoch": 0.79744, + "grad_norm": 0.40073803067207336, + "learning_rate": 0.00010376018958389472, + "loss": 2.6046, + "step": 6230 + }, + { + "epoch": 0.797568, + "grad_norm": 0.3974088132381439, + "learning_rate": 0.00010363378492992409, + "loss": 2.6178, + "step": 6231 + }, + { + "epoch": 0.797696, + "grad_norm": 0.4009820520877838, + "learning_rate": 0.00010350744841590103, + "loss": 2.5752, + "step": 6232 + }, + { + "epoch": 0.797824, + "grad_norm": 0.4158313572406769, + "learning_rate": 0.00010338118006354413, + "loss": 2.6108, + "step": 6233 + }, + { + "epoch": 0.797952, + "grad_norm": 0.3915190100669861, + "learning_rate": 0.0001032549798945604, + "loss": 2.6264, + "step": 6234 + }, + { + "epoch": 0.79808, + "grad_norm": 0.542098343372345, + "learning_rate": 0.00010312884793064492, + "loss": 2.5886, + "step": 6235 + }, + { + "epoch": 0.798208, + "grad_norm": 0.3943377435207367, + "learning_rate": 0.00010300278419348158, + "loss": 2.654, + "step": 6236 + }, + { + "epoch": 0.798336, + "grad_norm": 0.3855968415737152, + "learning_rate": 0.00010287678870474199, + "loss": 2.6594, + "step": 6237 + }, + { + "epoch": 0.798464, + "grad_norm": 0.3650366961956024, + "learning_rate": 0.00010275086148608614, + "loss": 2.7225, + "step": 6238 + }, + { + "epoch": 0.798592, + "grad_norm": 0.9391686320304871, + "learning_rate": 0.00010262500255916268, + "loss": 2.5141, + "step": 6239 + }, + { + "epoch": 0.79872, + "grad_norm": 0.4542687237262726, + "learning_rate": 0.00010249921194560802, + "loss": 2.6045, + "step": 6240 + }, + { + "epoch": 0.798848, + "grad_norm": 0.39455199241638184, + "learning_rate": 0.00010237348966704707, + "loss": 2.6348, + "step": 6241 + }, + { + "epoch": 0.798976, + "grad_norm": 0.3904871642589569, + "learning_rate": 0.0001022478357450929, + "loss": 2.6075, + "step": 6242 + }, + { + "epoch": 0.799104, + "grad_norm": 0.401242733001709, + "learning_rate": 0.00010212225020134691, + "loss": 2.5551, + "step": 6243 + }, + { + "epoch": 0.799232, + "grad_norm": 0.43371614813804626, + "learning_rate": 0.00010199673305739853, + "loss": 2.5965, + "step": 6244 + }, + { + "epoch": 0.79936, + "grad_norm": 0.4289807975292206, + "learning_rate": 0.000101871284334826, + "loss": 2.7333, + "step": 6245 + }, + { + "epoch": 0.799488, + "grad_norm": 0.4139355421066284, + "learning_rate": 0.00010174590405519501, + "loss": 2.6641, + "step": 6246 + }, + { + "epoch": 0.799616, + "grad_norm": 0.3957567811012268, + "learning_rate": 0.00010162059224006026, + "loss": 2.5948, + "step": 6247 + }, + { + "epoch": 0.799744, + "grad_norm": 0.40135255455970764, + "learning_rate": 0.00010149534891096408, + "loss": 2.678, + "step": 6248 + }, + { + "epoch": 0.799872, + "grad_norm": 0.39553776383399963, + "learning_rate": 0.00010137017408943727, + "loss": 2.6048, + "step": 6249 + }, + { + "epoch": 0.8, + "grad_norm": 0.3996278941631317, + "learning_rate": 0.00010124506779699882, + "loss": 2.5392, + "step": 6250 + }, + { + "epoch": 0.800128, + "grad_norm": 0.3785918653011322, + "learning_rate": 0.00010112003005515602, + "loss": 2.6059, + "step": 6251 + }, + { + "epoch": 0.800256, + "grad_norm": 0.38481271266937256, + "learning_rate": 0.00010099506088540417, + "loss": 2.6288, + "step": 6252 + }, + { + "epoch": 0.800384, + "grad_norm": 0.4032129943370819, + "learning_rate": 0.00010087016030922708, + "loss": 2.5711, + "step": 6253 + }, + { + "epoch": 0.800512, + "grad_norm": 0.4152248799800873, + "learning_rate": 0.00010074532834809646, + "loss": 2.5048, + "step": 6254 + }, + { + "epoch": 0.80064, + "grad_norm": 0.48478755354881287, + "learning_rate": 0.00010062056502347256, + "loss": 2.47, + "step": 6255 + }, + { + "epoch": 0.800768, + "grad_norm": 0.4142160415649414, + "learning_rate": 0.00010049587035680352, + "loss": 2.5782, + "step": 6256 + }, + { + "epoch": 0.800896, + "grad_norm": 0.37973690032958984, + "learning_rate": 0.00010037124436952588, + "loss": 2.5552, + "step": 6257 + }, + { + "epoch": 0.801024, + "grad_norm": 0.38268962502479553, + "learning_rate": 0.00010024668708306417, + "loss": 2.5503, + "step": 6258 + }, + { + "epoch": 0.801152, + "grad_norm": 0.36374276876449585, + "learning_rate": 0.0001001221985188312, + "loss": 2.6831, + "step": 6259 + }, + { + "epoch": 0.80128, + "grad_norm": 0.3838098645210266, + "learning_rate": 9.999777869822818e-05, + "loss": 2.5195, + "step": 6260 + }, + { + "epoch": 0.801408, + "grad_norm": 0.3994699716567993, + "learning_rate": 9.987342764264423e-05, + "loss": 2.559, + "step": 6261 + }, + { + "epoch": 0.801536, + "grad_norm": 0.41610050201416016, + "learning_rate": 9.974914537345675e-05, + "loss": 2.6577, + "step": 6262 + }, + { + "epoch": 0.801664, + "grad_norm": 0.39575299620628357, + "learning_rate": 9.96249319120311e-05, + "loss": 2.584, + "step": 6263 + }, + { + "epoch": 0.801792, + "grad_norm": 1.007204294204712, + "learning_rate": 9.950078727972139e-05, + "loss": 2.6026, + "step": 6264 + }, + { + "epoch": 0.80192, + "grad_norm": 0.40505295991897583, + "learning_rate": 9.937671149786931e-05, + "loss": 2.5854, + "step": 6265 + }, + { + "epoch": 0.802048, + "grad_norm": 0.43168526887893677, + "learning_rate": 9.925270458780495e-05, + "loss": 2.7391, + "step": 6266 + }, + { + "epoch": 0.802176, + "grad_norm": 0.38799527287483215, + "learning_rate": 9.912876657084641e-05, + "loss": 2.6287, + "step": 6267 + }, + { + "epoch": 0.802304, + "grad_norm": 0.37564823031425476, + "learning_rate": 9.900489746830032e-05, + "loss": 2.538, + "step": 6268 + }, + { + "epoch": 0.802432, + "grad_norm": 0.43947792053222656, + "learning_rate": 9.888109730146111e-05, + "loss": 2.5797, + "step": 6269 + }, + { + "epoch": 0.80256, + "grad_norm": 0.3881331980228424, + "learning_rate": 9.87573660916114e-05, + "loss": 2.6044, + "step": 6270 + }, + { + "epoch": 0.802688, + "grad_norm": 0.5659694075584412, + "learning_rate": 9.863370386002212e-05, + "loss": 2.5749, + "step": 6271 + }, + { + "epoch": 0.802816, + "grad_norm": 0.4602586328983307, + "learning_rate": 9.851011062795201e-05, + "loss": 2.6728, + "step": 6272 + }, + { + "epoch": 0.802944, + "grad_norm": 0.42805588245391846, + "learning_rate": 9.838658641664849e-05, + "loss": 2.6752, + "step": 6273 + }, + { + "epoch": 0.803072, + "grad_norm": 0.40637239813804626, + "learning_rate": 9.826313124734653e-05, + "loss": 2.6685, + "step": 6274 + }, + { + "epoch": 0.8032, + "grad_norm": 0.39065513014793396, + "learning_rate": 9.813974514126978e-05, + "loss": 2.6264, + "step": 6275 + }, + { + "epoch": 0.803328, + "grad_norm": 0.41184261441230774, + "learning_rate": 9.801642811962964e-05, + "loss": 2.6141, + "step": 6276 + }, + { + "epoch": 0.803456, + "grad_norm": 0.41264235973358154, + "learning_rate": 9.789318020362564e-05, + "loss": 2.5102, + "step": 6277 + }, + { + "epoch": 0.803584, + "grad_norm": 0.45627647638320923, + "learning_rate": 9.777000141444564e-05, + "loss": 2.5978, + "step": 6278 + }, + { + "epoch": 0.803712, + "grad_norm": 0.4191996455192566, + "learning_rate": 9.764689177326541e-05, + "loss": 2.5676, + "step": 6279 + }, + { + "epoch": 0.80384, + "grad_norm": 0.4326649010181427, + "learning_rate": 9.75238513012488e-05, + "loss": 2.5406, + "step": 6280 + }, + { + "epoch": 0.803968, + "grad_norm": 0.4116477966308594, + "learning_rate": 9.740088001954811e-05, + "loss": 2.5024, + "step": 6281 + }, + { + "epoch": 0.804096, + "grad_norm": 0.3644542694091797, + "learning_rate": 9.72779779493036e-05, + "loss": 2.4393, + "step": 6282 + }, + { + "epoch": 0.804224, + "grad_norm": 0.4060465395450592, + "learning_rate": 9.715514511164342e-05, + "loss": 2.4953, + "step": 6283 + }, + { + "epoch": 0.804352, + "grad_norm": 0.40174949169158936, + "learning_rate": 9.703238152768401e-05, + "loss": 2.6104, + "step": 6284 + }, + { + "epoch": 0.80448, + "grad_norm": 0.3788057267665863, + "learning_rate": 9.690968721852978e-05, + "loss": 2.5654, + "step": 6285 + }, + { + "epoch": 0.804608, + "grad_norm": 0.44650670886039734, + "learning_rate": 9.678706220527333e-05, + "loss": 2.6263, + "step": 6286 + }, + { + "epoch": 0.804736, + "grad_norm": 0.3977706730365753, + "learning_rate": 9.666450650899533e-05, + "loss": 2.6063, + "step": 6287 + }, + { + "epoch": 0.804864, + "grad_norm": 0.37962281703948975, + "learning_rate": 9.654202015076441e-05, + "loss": 2.4763, + "step": 6288 + }, + { + "epoch": 0.804992, + "grad_norm": 0.38277146220207214, + "learning_rate": 9.641960315163767e-05, + "loss": 2.5417, + "step": 6289 + }, + { + "epoch": 0.80512, + "grad_norm": 0.46654602885246277, + "learning_rate": 9.629725553265983e-05, + "loss": 2.7493, + "step": 6290 + }, + { + "epoch": 0.805248, + "grad_norm": 0.4443380534648895, + "learning_rate": 9.617497731486374e-05, + "loss": 2.6164, + "step": 6291 + }, + { + "epoch": 0.805376, + "grad_norm": 0.40439870953559875, + "learning_rate": 9.605276851927075e-05, + "loss": 2.5971, + "step": 6292 + }, + { + "epoch": 0.805504, + "grad_norm": 0.39641648530960083, + "learning_rate": 9.593062916688982e-05, + "loss": 2.5876, + "step": 6293 + }, + { + "epoch": 0.805632, + "grad_norm": 0.41149190068244934, + "learning_rate": 9.580855927871807e-05, + "loss": 2.5788, + "step": 6294 + }, + { + "epoch": 0.80576, + "grad_norm": 0.4134078323841095, + "learning_rate": 9.568655887574068e-05, + "loss": 2.5854, + "step": 6295 + }, + { + "epoch": 0.805888, + "grad_norm": 0.46514183282852173, + "learning_rate": 9.556462797893111e-05, + "loss": 2.5829, + "step": 6296 + }, + { + "epoch": 0.806016, + "grad_norm": 0.4071011543273926, + "learning_rate": 9.544276660925066e-05, + "loss": 2.431, + "step": 6297 + }, + { + "epoch": 0.806144, + "grad_norm": 0.37151530385017395, + "learning_rate": 9.532097478764862e-05, + "loss": 2.6134, + "step": 6298 + }, + { + "epoch": 0.806272, + "grad_norm": 0.4449272155761719, + "learning_rate": 9.519925253506245e-05, + "loss": 2.6214, + "step": 6299 + }, + { + "epoch": 0.8064, + "grad_norm": 0.4132419526576996, + "learning_rate": 9.507759987241754e-05, + "loss": 2.5272, + "step": 6300 + }, + { + "epoch": 0.806528, + "grad_norm": 0.39396196603775024, + "learning_rate": 9.495601682062755e-05, + "loss": 2.4529, + "step": 6301 + }, + { + "epoch": 0.806656, + "grad_norm": 0.39563828706741333, + "learning_rate": 9.483450340059385e-05, + "loss": 2.6081, + "step": 6302 + }, + { + "epoch": 0.806784, + "grad_norm": 0.48571741580963135, + "learning_rate": 9.47130596332062e-05, + "loss": 2.6484, + "step": 6303 + }, + { + "epoch": 0.806912, + "grad_norm": 0.3864864408969879, + "learning_rate": 9.459168553934211e-05, + "loss": 2.5596, + "step": 6304 + }, + { + "epoch": 0.80704, + "grad_norm": 0.380285382270813, + "learning_rate": 9.447038113986717e-05, + "loss": 2.5817, + "step": 6305 + }, + { + "epoch": 0.807168, + "grad_norm": 0.3914862275123596, + "learning_rate": 9.434914645563497e-05, + "loss": 2.5642, + "step": 6306 + }, + { + "epoch": 0.807296, + "grad_norm": 0.44341814517974854, + "learning_rate": 9.422798150748724e-05, + "loss": 2.499, + "step": 6307 + }, + { + "epoch": 0.807424, + "grad_norm": 0.39854612946510315, + "learning_rate": 9.410688631625364e-05, + "loss": 2.5425, + "step": 6308 + }, + { + "epoch": 0.807552, + "grad_norm": 0.4639688730239868, + "learning_rate": 9.398586090275164e-05, + "loss": 2.6254, + "step": 6309 + }, + { + "epoch": 0.80768, + "grad_norm": 0.38877391815185547, + "learning_rate": 9.386490528778702e-05, + "loss": 2.679, + "step": 6310 + }, + { + "epoch": 0.807808, + "grad_norm": 0.42077451944351196, + "learning_rate": 9.374401949215366e-05, + "loss": 2.6456, + "step": 6311 + }, + { + "epoch": 0.807936, + "grad_norm": 0.4299446940422058, + "learning_rate": 9.362320353663311e-05, + "loss": 2.7313, + "step": 6312 + }, + { + "epoch": 0.808064, + "grad_norm": 0.4195094406604767, + "learning_rate": 9.350245744199498e-05, + "loss": 2.619, + "step": 6313 + }, + { + "epoch": 0.808192, + "grad_norm": 0.3983438313007355, + "learning_rate": 9.338178122899693e-05, + "loss": 2.5151, + "step": 6314 + }, + { + "epoch": 0.80832, + "grad_norm": 0.4156458079814911, + "learning_rate": 9.32611749183846e-05, + "loss": 2.6296, + "step": 6315 + }, + { + "epoch": 0.808448, + "grad_norm": 0.410634845495224, + "learning_rate": 9.31406385308915e-05, + "loss": 2.616, + "step": 6316 + }, + { + "epoch": 0.808576, + "grad_norm": 0.4119603633880615, + "learning_rate": 9.30201720872395e-05, + "loss": 2.5884, + "step": 6317 + }, + { + "epoch": 0.808704, + "grad_norm": 0.3995600640773773, + "learning_rate": 9.289977560813789e-05, + "loss": 2.6277, + "step": 6318 + }, + { + "epoch": 0.808832, + "grad_norm": 0.4291670620441437, + "learning_rate": 9.277944911428449e-05, + "loss": 2.5679, + "step": 6319 + }, + { + "epoch": 0.80896, + "grad_norm": 0.38543960452079773, + "learning_rate": 9.265919262636468e-05, + "loss": 2.5379, + "step": 6320 + }, + { + "epoch": 0.809088, + "grad_norm": 0.4298980236053467, + "learning_rate": 9.2539006165052e-05, + "loss": 2.5324, + "step": 6321 + }, + { + "epoch": 0.809216, + "grad_norm": 0.40078073740005493, + "learning_rate": 9.241888975100782e-05, + "loss": 2.6573, + "step": 6322 + }, + { + "epoch": 0.809344, + "grad_norm": 0.4289223253726959, + "learning_rate": 9.229884340488149e-05, + "loss": 2.5319, + "step": 6323 + }, + { + "epoch": 0.809472, + "grad_norm": 0.41495782136917114, + "learning_rate": 9.217886714731055e-05, + "loss": 2.5492, + "step": 6324 + }, + { + "epoch": 0.8096, + "grad_norm": 0.41053664684295654, + "learning_rate": 9.205896099892019e-05, + "loss": 2.5998, + "step": 6325 + }, + { + "epoch": 0.809728, + "grad_norm": 0.374761164188385, + "learning_rate": 9.193912498032375e-05, + "loss": 2.5292, + "step": 6326 + }, + { + "epoch": 0.809856, + "grad_norm": 0.38548800349235535, + "learning_rate": 9.18193591121223e-05, + "loss": 2.5025, + "step": 6327 + }, + { + "epoch": 0.809984, + "grad_norm": 0.431510329246521, + "learning_rate": 9.169966341490499e-05, + "loss": 2.508, + "step": 6328 + }, + { + "epoch": 0.810112, + "grad_norm": 0.3962239623069763, + "learning_rate": 9.158003790924906e-05, + "loss": 2.5804, + "step": 6329 + }, + { + "epoch": 0.81024, + "grad_norm": 1.0521199703216553, + "learning_rate": 9.146048261571944e-05, + "loss": 2.5079, + "step": 6330 + }, + { + "epoch": 0.810368, + "grad_norm": 0.5074988007545471, + "learning_rate": 9.134099755486891e-05, + "loss": 2.5239, + "step": 6331 + }, + { + "epoch": 0.810496, + "grad_norm": 0.37636256217956543, + "learning_rate": 9.122158274723863e-05, + "loss": 2.5695, + "step": 6332 + }, + { + "epoch": 0.810624, + "grad_norm": 0.4092382788658142, + "learning_rate": 9.110223821335723e-05, + "loss": 2.7559, + "step": 6333 + }, + { + "epoch": 0.810752, + "grad_norm": 0.37143802642822266, + "learning_rate": 9.098296397374145e-05, + "loss": 2.4314, + "step": 6334 + }, + { + "epoch": 0.81088, + "grad_norm": 0.6269592046737671, + "learning_rate": 9.08637600488959e-05, + "loss": 2.6054, + "step": 6335 + }, + { + "epoch": 0.811008, + "grad_norm": 0.4276467561721802, + "learning_rate": 9.074462645931308e-05, + "loss": 2.6443, + "step": 6336 + }, + { + "epoch": 0.811136, + "grad_norm": 0.4109061658382416, + "learning_rate": 9.062556322547333e-05, + "loss": 2.5917, + "step": 6337 + }, + { + "epoch": 0.811264, + "grad_norm": 0.4013563394546509, + "learning_rate": 9.050657036784515e-05, + "loss": 2.719, + "step": 6338 + }, + { + "epoch": 0.811392, + "grad_norm": 0.4017780125141144, + "learning_rate": 9.038764790688492e-05, + "loss": 2.5593, + "step": 6339 + }, + { + "epoch": 0.81152, + "grad_norm": 0.40477582812309265, + "learning_rate": 9.026879586303666e-05, + "loss": 2.6392, + "step": 6340 + }, + { + "epoch": 0.811648, + "grad_norm": 0.5405519008636475, + "learning_rate": 9.015001425673241e-05, + "loss": 2.5695, + "step": 6341 + }, + { + "epoch": 0.811776, + "grad_norm": 0.435845285654068, + "learning_rate": 9.003130310839203e-05, + "loss": 2.6347, + "step": 6342 + }, + { + "epoch": 0.811904, + "grad_norm": 0.3656361997127533, + "learning_rate": 8.99126624384235e-05, + "loss": 2.606, + "step": 6343 + }, + { + "epoch": 0.812032, + "grad_norm": 0.47047221660614014, + "learning_rate": 8.979409226722224e-05, + "loss": 2.6826, + "step": 6344 + }, + { + "epoch": 0.81216, + "grad_norm": 0.9867327213287354, + "learning_rate": 8.967559261517217e-05, + "loss": 2.6939, + "step": 6345 + }, + { + "epoch": 0.812288, + "grad_norm": 0.40477967262268066, + "learning_rate": 8.955716350264454e-05, + "loss": 2.5025, + "step": 6346 + }, + { + "epoch": 0.812416, + "grad_norm": 0.39622217416763306, + "learning_rate": 8.943880494999885e-05, + "loss": 2.5552, + "step": 6347 + }, + { + "epoch": 0.812544, + "grad_norm": 0.4136972725391388, + "learning_rate": 8.932051697758225e-05, + "loss": 2.6405, + "step": 6348 + }, + { + "epoch": 0.812672, + "grad_norm": 0.3801695704460144, + "learning_rate": 8.920229960572973e-05, + "loss": 2.64, + "step": 6349 + }, + { + "epoch": 0.8128, + "grad_norm": 0.3849983513355255, + "learning_rate": 8.908415285476434e-05, + "loss": 2.5503, + "step": 6350 + }, + { + "epoch": 0.812928, + "grad_norm": 0.41388508677482605, + "learning_rate": 8.896607674499679e-05, + "loss": 2.6573, + "step": 6351 + }, + { + "epoch": 0.813056, + "grad_norm": 0.4311745762825012, + "learning_rate": 8.884807129672567e-05, + "loss": 2.5933, + "step": 6352 + }, + { + "epoch": 0.813184, + "grad_norm": 0.39478152990341187, + "learning_rate": 8.873013653023765e-05, + "loss": 2.6019, + "step": 6353 + }, + { + "epoch": 0.813312, + "grad_norm": 0.40727511048316956, + "learning_rate": 8.861227246580705e-05, + "loss": 2.5849, + "step": 6354 + }, + { + "epoch": 0.81344, + "grad_norm": 0.4204227924346924, + "learning_rate": 8.84944791236959e-05, + "loss": 2.5499, + "step": 6355 + }, + { + "epoch": 0.813568, + "grad_norm": 0.38610294461250305, + "learning_rate": 8.83767565241545e-05, + "loss": 2.6065, + "step": 6356 + }, + { + "epoch": 0.813696, + "grad_norm": 0.3849245607852936, + "learning_rate": 8.82591046874206e-05, + "loss": 2.6262, + "step": 6357 + }, + { + "epoch": 0.813824, + "grad_norm": 0.36378127336502075, + "learning_rate": 8.814152363371992e-05, + "loss": 2.5143, + "step": 6358 + }, + { + "epoch": 0.813952, + "grad_norm": 0.37478986382484436, + "learning_rate": 8.802401338326582e-05, + "loss": 2.6708, + "step": 6359 + }, + { + "epoch": 0.81408, + "grad_norm": 0.41959092020988464, + "learning_rate": 8.790657395626001e-05, + "loss": 2.5405, + "step": 6360 + }, + { + "epoch": 0.814208, + "grad_norm": 0.4024343490600586, + "learning_rate": 8.778920537289154e-05, + "loss": 2.5893, + "step": 6361 + }, + { + "epoch": 0.814336, + "grad_norm": 0.39370670914649963, + "learning_rate": 8.767190765333744e-05, + "loss": 2.5295, + "step": 6362 + }, + { + "epoch": 0.814464, + "grad_norm": 0.3913019895553589, + "learning_rate": 8.755468081776253e-05, + "loss": 2.6169, + "step": 6363 + }, + { + "epoch": 0.814592, + "grad_norm": 0.4205024540424347, + "learning_rate": 8.743752488631945e-05, + "loss": 2.6511, + "step": 6364 + }, + { + "epoch": 0.81472, + "grad_norm": 0.4016045033931732, + "learning_rate": 8.732043987914857e-05, + "loss": 2.5697, + "step": 6365 + }, + { + "epoch": 0.814848, + "grad_norm": 0.40390029549598694, + "learning_rate": 8.720342581637836e-05, + "loss": 2.6204, + "step": 6366 + }, + { + "epoch": 0.814976, + "grad_norm": 0.4087165594100952, + "learning_rate": 8.70864827181247e-05, + "loss": 2.5788, + "step": 6367 + }, + { + "epoch": 0.815104, + "grad_norm": 0.39381295442581177, + "learning_rate": 8.696961060449165e-05, + "loss": 2.6009, + "step": 6368 + }, + { + "epoch": 0.815232, + "grad_norm": 0.3870253562927246, + "learning_rate": 8.685280949557084e-05, + "loss": 2.5513, + "step": 6369 + }, + { + "epoch": 0.81536, + "grad_norm": 0.4070032835006714, + "learning_rate": 8.673607941144169e-05, + "loss": 2.609, + "step": 6370 + }, + { + "epoch": 0.815488, + "grad_norm": 0.3760526478290558, + "learning_rate": 8.66194203721714e-05, + "loss": 2.5918, + "step": 6371 + }, + { + "epoch": 0.815616, + "grad_norm": 0.47568079829216003, + "learning_rate": 8.650283239781515e-05, + "loss": 2.6156, + "step": 6372 + }, + { + "epoch": 0.815744, + "grad_norm": 0.43657931685447693, + "learning_rate": 8.638631550841552e-05, + "loss": 2.552, + "step": 6373 + }, + { + "epoch": 0.815872, + "grad_norm": 0.3782326877117157, + "learning_rate": 8.626986972400325e-05, + "loss": 2.5199, + "step": 6374 + }, + { + "epoch": 0.816, + "grad_norm": 0.42667946219444275, + "learning_rate": 8.615349506459692e-05, + "loss": 2.7188, + "step": 6375 + }, + { + "epoch": 0.816128, + "grad_norm": 0.40211278200149536, + "learning_rate": 8.603719155020246e-05, + "loss": 2.5124, + "step": 6376 + }, + { + "epoch": 0.816256, + "grad_norm": 0.39331918954849243, + "learning_rate": 8.592095920081383e-05, + "loss": 2.6275, + "step": 6377 + }, + { + "epoch": 0.816384, + "grad_norm": 0.4036480486392975, + "learning_rate": 8.580479803641278e-05, + "loss": 2.6037, + "step": 6378 + }, + { + "epoch": 0.816512, + "grad_norm": 0.3832785487174988, + "learning_rate": 8.568870807696872e-05, + "loss": 2.5563, + "step": 6379 + }, + { + "epoch": 0.81664, + "grad_norm": 0.37265101075172424, + "learning_rate": 8.557268934243867e-05, + "loss": 2.5235, + "step": 6380 + }, + { + "epoch": 0.816768, + "grad_norm": 0.48915359377861023, + "learning_rate": 8.545674185276791e-05, + "loss": 2.5495, + "step": 6381 + }, + { + "epoch": 0.816896, + "grad_norm": 0.41964802145957947, + "learning_rate": 8.534086562788906e-05, + "loss": 2.6262, + "step": 6382 + }, + { + "epoch": 0.817024, + "grad_norm": 0.41210928559303284, + "learning_rate": 8.52250606877224e-05, + "loss": 2.6601, + "step": 6383 + }, + { + "epoch": 0.817152, + "grad_norm": 0.4077272117137909, + "learning_rate": 8.510932705217644e-05, + "loss": 2.6121, + "step": 6384 + }, + { + "epoch": 0.81728, + "grad_norm": 0.467304527759552, + "learning_rate": 8.499366474114695e-05, + "loss": 2.6119, + "step": 6385 + }, + { + "epoch": 0.817408, + "grad_norm": 0.4466140866279602, + "learning_rate": 8.487807377451767e-05, + "loss": 2.499, + "step": 6386 + }, + { + "epoch": 0.817536, + "grad_norm": 0.3973773121833801, + "learning_rate": 8.476255417216005e-05, + "loss": 2.5344, + "step": 6387 + }, + { + "epoch": 0.817664, + "grad_norm": 0.38463225960731506, + "learning_rate": 8.464710595393305e-05, + "loss": 2.6628, + "step": 6388 + }, + { + "epoch": 0.817792, + "grad_norm": 0.4099874198436737, + "learning_rate": 8.453172913968382e-05, + "loss": 2.5134, + "step": 6389 + }, + { + "epoch": 0.81792, + "grad_norm": 0.4047398567199707, + "learning_rate": 8.44164237492469e-05, + "loss": 2.4735, + "step": 6390 + }, + { + "epoch": 0.818048, + "grad_norm": 0.46522706747055054, + "learning_rate": 8.430118980244461e-05, + "loss": 2.5692, + "step": 6391 + }, + { + "epoch": 0.818176, + "grad_norm": 0.3666349947452545, + "learning_rate": 8.418602731908686e-05, + "loss": 2.5027, + "step": 6392 + }, + { + "epoch": 0.818304, + "grad_norm": 0.427106648683548, + "learning_rate": 8.407093631897167e-05, + "loss": 2.6064, + "step": 6393 + }, + { + "epoch": 0.818432, + "grad_norm": 0.3813337981700897, + "learning_rate": 8.395591682188442e-05, + "loss": 2.5614, + "step": 6394 + }, + { + "epoch": 0.81856, + "grad_norm": 0.43556275963783264, + "learning_rate": 8.384096884759807e-05, + "loss": 2.6473, + "step": 6395 + }, + { + "epoch": 0.818688, + "grad_norm": 0.38397276401519775, + "learning_rate": 8.372609241587387e-05, + "loss": 2.6343, + "step": 6396 + }, + { + "epoch": 0.818816, + "grad_norm": 0.4170434772968292, + "learning_rate": 8.361128754646024e-05, + "loss": 2.5163, + "step": 6397 + }, + { + "epoch": 0.818944, + "grad_norm": 0.41919684410095215, + "learning_rate": 8.349655425909347e-05, + "loss": 2.6993, + "step": 6398 + }, + { + "epoch": 0.819072, + "grad_norm": 0.40234068036079407, + "learning_rate": 8.338189257349754e-05, + "loss": 2.5724, + "step": 6399 + }, + { + "epoch": 0.8192, + "grad_norm": 0.3749343156814575, + "learning_rate": 8.326730250938413e-05, + "loss": 2.5042, + "step": 6400 + }, + { + "epoch": 0.819328, + "grad_norm": 0.4401456117630005, + "learning_rate": 8.31527840864525e-05, + "loss": 2.6616, + "step": 6401 + }, + { + "epoch": 0.819456, + "grad_norm": 0.39094167947769165, + "learning_rate": 8.303833732438987e-05, + "loss": 2.5868, + "step": 6402 + }, + { + "epoch": 0.819584, + "grad_norm": 0.38304996490478516, + "learning_rate": 8.29239622428708e-05, + "loss": 2.6164, + "step": 6403 + }, + { + "epoch": 0.819712, + "grad_norm": 0.4311363101005554, + "learning_rate": 8.28096588615579e-05, + "loss": 2.6489, + "step": 6404 + }, + { + "epoch": 0.81984, + "grad_norm": 0.3754993677139282, + "learning_rate": 8.269542720010109e-05, + "loss": 2.4754, + "step": 6405 + }, + { + "epoch": 0.819968, + "grad_norm": 0.4373638331890106, + "learning_rate": 8.25812672781382e-05, + "loss": 2.5985, + "step": 6406 + }, + { + "epoch": 0.820096, + "grad_norm": 0.5673739910125732, + "learning_rate": 8.246717911529456e-05, + "loss": 2.6392, + "step": 6407 + }, + { + "epoch": 0.820224, + "grad_norm": 0.4513801038265228, + "learning_rate": 8.235316273118332e-05, + "loss": 2.5755, + "step": 6408 + }, + { + "epoch": 0.820352, + "grad_norm": 0.4090612530708313, + "learning_rate": 8.223921814540503e-05, + "loss": 2.5368, + "step": 6409 + }, + { + "epoch": 0.82048, + "grad_norm": 0.38459083437919617, + "learning_rate": 8.21253453775484e-05, + "loss": 2.5142, + "step": 6410 + }, + { + "epoch": 0.820608, + "grad_norm": 0.41318950057029724, + "learning_rate": 8.201154444718917e-05, + "loss": 2.5991, + "step": 6411 + }, + { + "epoch": 0.820736, + "grad_norm": 0.388629287481308, + "learning_rate": 8.189781537389135e-05, + "loss": 2.6497, + "step": 6412 + }, + { + "epoch": 0.820864, + "grad_norm": 0.39804786443710327, + "learning_rate": 8.178415817720608e-05, + "loss": 2.5742, + "step": 6413 + }, + { + "epoch": 0.820992, + "grad_norm": 0.395658016204834, + "learning_rate": 8.16705728766724e-05, + "loss": 2.6267, + "step": 6414 + }, + { + "epoch": 0.82112, + "grad_norm": 0.4503498673439026, + "learning_rate": 8.155705949181691e-05, + "loss": 2.6124, + "step": 6415 + }, + { + "epoch": 0.821248, + "grad_norm": 0.42711111903190613, + "learning_rate": 8.144361804215383e-05, + "loss": 2.6388, + "step": 6416 + }, + { + "epoch": 0.821376, + "grad_norm": 0.48490849137306213, + "learning_rate": 8.133024854718524e-05, + "loss": 2.5613, + "step": 6417 + }, + { + "epoch": 0.821504, + "grad_norm": 0.40668535232543945, + "learning_rate": 8.121695102640054e-05, + "loss": 2.5849, + "step": 6418 + }, + { + "epoch": 0.821632, + "grad_norm": 0.41154050827026367, + "learning_rate": 8.110372549927692e-05, + "loss": 2.5218, + "step": 6419 + }, + { + "epoch": 0.82176, + "grad_norm": 0.3855734169483185, + "learning_rate": 8.099057198527899e-05, + "loss": 2.6639, + "step": 6420 + }, + { + "epoch": 0.821888, + "grad_norm": 0.44118064641952515, + "learning_rate": 8.087749050385951e-05, + "loss": 2.5819, + "step": 6421 + }, + { + "epoch": 0.822016, + "grad_norm": 0.39301788806915283, + "learning_rate": 8.076448107445822e-05, + "loss": 2.6014, + "step": 6422 + }, + { + "epoch": 0.822144, + "grad_norm": 0.41799113154411316, + "learning_rate": 8.065154371650285e-05, + "loss": 2.5472, + "step": 6423 + }, + { + "epoch": 0.822272, + "grad_norm": 0.3994382321834564, + "learning_rate": 8.053867844940854e-05, + "loss": 2.5804, + "step": 6424 + }, + { + "epoch": 0.8224, + "grad_norm": 0.40445831418037415, + "learning_rate": 8.042588529257828e-05, + "loss": 2.6311, + "step": 6425 + }, + { + "epoch": 0.822528, + "grad_norm": 0.5269573926925659, + "learning_rate": 8.031316426540252e-05, + "loss": 2.6145, + "step": 6426 + }, + { + "epoch": 0.822656, + "grad_norm": 0.41829046607017517, + "learning_rate": 8.02005153872592e-05, + "loss": 2.5372, + "step": 6427 + }, + { + "epoch": 0.822784, + "grad_norm": 0.4000975489616394, + "learning_rate": 8.008793867751402e-05, + "loss": 2.464, + "step": 6428 + }, + { + "epoch": 0.822912, + "grad_norm": 0.38993456959724426, + "learning_rate": 7.997543415552011e-05, + "loss": 2.6323, + "step": 6429 + }, + { + "epoch": 0.82304, + "grad_norm": 0.3988874852657318, + "learning_rate": 7.986300184061856e-05, + "loss": 2.5855, + "step": 6430 + }, + { + "epoch": 0.823168, + "grad_norm": 0.48986274003982544, + "learning_rate": 7.975064175213748e-05, + "loss": 2.6427, + "step": 6431 + }, + { + "epoch": 0.823296, + "grad_norm": 0.4080543518066406, + "learning_rate": 7.963835390939316e-05, + "loss": 2.5547, + "step": 6432 + }, + { + "epoch": 0.823424, + "grad_norm": 0.36345747113227844, + "learning_rate": 7.952613833168909e-05, + "loss": 2.6382, + "step": 6433 + }, + { + "epoch": 0.823552, + "grad_norm": 0.36700552701950073, + "learning_rate": 7.941399503831637e-05, + "loss": 2.5412, + "step": 6434 + }, + { + "epoch": 0.82368, + "grad_norm": 0.39359161257743835, + "learning_rate": 7.930192404855374e-05, + "loss": 2.5047, + "step": 6435 + }, + { + "epoch": 0.823808, + "grad_norm": 0.42142897844314575, + "learning_rate": 7.918992538166752e-05, + "loss": 2.7336, + "step": 6436 + }, + { + "epoch": 0.823936, + "grad_norm": 0.4173029363155365, + "learning_rate": 7.907799905691143e-05, + "loss": 2.5545, + "step": 6437 + }, + { + "epoch": 0.824064, + "grad_norm": 0.4100871980190277, + "learning_rate": 7.896614509352723e-05, + "loss": 2.5305, + "step": 6438 + }, + { + "epoch": 0.824192, + "grad_norm": 0.46236032247543335, + "learning_rate": 7.885436351074355e-05, + "loss": 2.5393, + "step": 6439 + }, + { + "epoch": 0.82432, + "grad_norm": 0.3984875977039337, + "learning_rate": 7.874265432777728e-05, + "loss": 2.5814, + "step": 6440 + }, + { + "epoch": 0.824448, + "grad_norm": 0.4051078259944916, + "learning_rate": 7.863101756383235e-05, + "loss": 2.5406, + "step": 6441 + }, + { + "epoch": 0.824576, + "grad_norm": 0.4475669860839844, + "learning_rate": 7.851945323810045e-05, + "loss": 2.5855, + "step": 6442 + }, + { + "epoch": 0.824704, + "grad_norm": 0.4651617705821991, + "learning_rate": 7.840796136976075e-05, + "loss": 2.5572, + "step": 6443 + }, + { + "epoch": 0.824832, + "grad_norm": 0.38562366366386414, + "learning_rate": 7.829654197797998e-05, + "loss": 2.623, + "step": 6444 + }, + { + "epoch": 0.82496, + "grad_norm": 0.4425043761730194, + "learning_rate": 7.818519508191235e-05, + "loss": 2.632, + "step": 6445 + }, + { + "epoch": 0.825088, + "grad_norm": 0.37948885560035706, + "learning_rate": 7.807392070069991e-05, + "loss": 2.6491, + "step": 6446 + }, + { + "epoch": 0.825216, + "grad_norm": 0.39727070927619934, + "learning_rate": 7.796271885347189e-05, + "loss": 2.5743, + "step": 6447 + }, + { + "epoch": 0.825344, + "grad_norm": 0.3991154432296753, + "learning_rate": 7.785158955934507e-05, + "loss": 2.6253, + "step": 6448 + }, + { + "epoch": 0.825472, + "grad_norm": 0.45039257407188416, + "learning_rate": 7.774053283742406e-05, + "loss": 2.5599, + "step": 6449 + }, + { + "epoch": 0.8256, + "grad_norm": 0.41769176721572876, + "learning_rate": 7.762954870680067e-05, + "loss": 2.7114, + "step": 6450 + }, + { + "epoch": 0.825728, + "grad_norm": 0.4114800989627838, + "learning_rate": 7.751863718655444e-05, + "loss": 2.6452, + "step": 6451 + }, + { + "epoch": 0.825856, + "grad_norm": 0.4108750820159912, + "learning_rate": 7.740779829575217e-05, + "loss": 2.6223, + "step": 6452 + }, + { + "epoch": 0.825984, + "grad_norm": 0.4180569350719452, + "learning_rate": 7.729703205344863e-05, + "loss": 2.6332, + "step": 6453 + }, + { + "epoch": 0.826112, + "grad_norm": 0.41905611753463745, + "learning_rate": 7.718633847868567e-05, + "loss": 2.6147, + "step": 6454 + }, + { + "epoch": 0.82624, + "grad_norm": 0.4837961494922638, + "learning_rate": 7.707571759049281e-05, + "loss": 2.5441, + "step": 6455 + }, + { + "epoch": 0.826368, + "grad_norm": 0.4170343279838562, + "learning_rate": 7.6965169407887e-05, + "loss": 2.5836, + "step": 6456 + }, + { + "epoch": 0.826496, + "grad_norm": 0.43713346123695374, + "learning_rate": 7.685469394987271e-05, + "loss": 2.5908, + "step": 6457 + }, + { + "epoch": 0.826624, + "grad_norm": 0.4291347563266754, + "learning_rate": 7.674429123544218e-05, + "loss": 2.6703, + "step": 6458 + }, + { + "epoch": 0.826752, + "grad_norm": 0.45089593529701233, + "learning_rate": 7.66339612835748e-05, + "loss": 2.6072, + "step": 6459 + }, + { + "epoch": 0.82688, + "grad_norm": 0.4503784775733948, + "learning_rate": 7.652370411323744e-05, + "loss": 2.5877, + "step": 6460 + }, + { + "epoch": 0.827008, + "grad_norm": 0.3664728105068207, + "learning_rate": 7.641351974338478e-05, + "loss": 2.4704, + "step": 6461 + }, + { + "epoch": 0.827136, + "grad_norm": 0.4293111562728882, + "learning_rate": 7.630340819295878e-05, + "loss": 2.607, + "step": 6462 + }, + { + "epoch": 0.827264, + "grad_norm": 0.38720396161079407, + "learning_rate": 7.619336948088879e-05, + "loss": 2.5736, + "step": 6463 + }, + { + "epoch": 0.827392, + "grad_norm": 0.3794127106666565, + "learning_rate": 7.608340362609173e-05, + "loss": 2.5924, + "step": 6464 + }, + { + "epoch": 0.82752, + "grad_norm": 0.40776291489601135, + "learning_rate": 7.59735106474721e-05, + "loss": 2.6006, + "step": 6465 + }, + { + "epoch": 0.827648, + "grad_norm": 0.3499011695384979, + "learning_rate": 7.586369056392161e-05, + "loss": 2.5424, + "step": 6466 + }, + { + "epoch": 0.827776, + "grad_norm": 0.41117000579833984, + "learning_rate": 7.575394339431969e-05, + "loss": 2.6694, + "step": 6467 + }, + { + "epoch": 0.827904, + "grad_norm": 0.3894113302230835, + "learning_rate": 7.564426915753331e-05, + "loss": 2.5904, + "step": 6468 + }, + { + "epoch": 0.828032, + "grad_norm": 0.40780338644981384, + "learning_rate": 7.553466787241664e-05, + "loss": 2.5192, + "step": 6469 + }, + { + "epoch": 0.82816, + "grad_norm": 0.4123285114765167, + "learning_rate": 7.542513955781138e-05, + "loss": 2.5592, + "step": 6470 + }, + { + "epoch": 0.828288, + "grad_norm": 0.49615800380706787, + "learning_rate": 7.53156842325467e-05, + "loss": 2.5812, + "step": 6471 + }, + { + "epoch": 0.828416, + "grad_norm": 0.36531856656074524, + "learning_rate": 7.520630191543932e-05, + "loss": 2.5807, + "step": 6472 + }, + { + "epoch": 0.828544, + "grad_norm": 0.3708794414997101, + "learning_rate": 7.509699262529307e-05, + "loss": 2.63, + "step": 6473 + }, + { + "epoch": 0.828672, + "grad_norm": 0.4070297181606293, + "learning_rate": 7.498775638089988e-05, + "loss": 2.6862, + "step": 6474 + }, + { + "epoch": 0.8288, + "grad_norm": 0.4492786228656769, + "learning_rate": 7.487859320103846e-05, + "loss": 2.5382, + "step": 6475 + }, + { + "epoch": 0.828928, + "grad_norm": 0.4517577886581421, + "learning_rate": 7.476950310447522e-05, + "loss": 2.5522, + "step": 6476 + }, + { + "epoch": 0.829056, + "grad_norm": 0.393142431974411, + "learning_rate": 7.466048610996423e-05, + "loss": 2.618, + "step": 6477 + }, + { + "epoch": 0.829184, + "grad_norm": 0.41158485412597656, + "learning_rate": 7.455154223624661e-05, + "loss": 2.5128, + "step": 6478 + }, + { + "epoch": 0.829312, + "grad_norm": 0.4893052279949188, + "learning_rate": 7.444267150205108e-05, + "loss": 2.6229, + "step": 6479 + }, + { + "epoch": 0.82944, + "grad_norm": 0.3598281741142273, + "learning_rate": 7.433387392609387e-05, + "loss": 2.5662, + "step": 6480 + }, + { + "epoch": 0.829568, + "grad_norm": 0.35018032789230347, + "learning_rate": 7.422514952707832e-05, + "loss": 2.6274, + "step": 6481 + }, + { + "epoch": 0.829696, + "grad_norm": 0.4122823476791382, + "learning_rate": 7.411649832369566e-05, + "loss": 2.5575, + "step": 6482 + }, + { + "epoch": 0.829824, + "grad_norm": 0.3637377619743347, + "learning_rate": 7.400792033462428e-05, + "loss": 2.5105, + "step": 6483 + }, + { + "epoch": 0.829952, + "grad_norm": 0.42767468094825745, + "learning_rate": 7.389941557852986e-05, + "loss": 2.5504, + "step": 6484 + }, + { + "epoch": 0.83008, + "grad_norm": 0.3628970682621002, + "learning_rate": 7.379098407406553e-05, + "loss": 2.498, + "step": 6485 + }, + { + "epoch": 0.830208, + "grad_norm": 0.37143900990486145, + "learning_rate": 7.368262583987228e-05, + "loss": 2.5946, + "step": 6486 + }, + { + "epoch": 0.830336, + "grad_norm": 0.40600940585136414, + "learning_rate": 7.357434089457788e-05, + "loss": 2.564, + "step": 6487 + }, + { + "epoch": 0.830464, + "grad_norm": 0.3826906085014343, + "learning_rate": 7.346612925679775e-05, + "loss": 2.5551, + "step": 6488 + }, + { + "epoch": 0.830592, + "grad_norm": 0.4537860155105591, + "learning_rate": 7.33579909451349e-05, + "loss": 2.6629, + "step": 6489 + }, + { + "epoch": 0.83072, + "grad_norm": 0.3961121737957001, + "learning_rate": 7.324992597817947e-05, + "loss": 2.6793, + "step": 6490 + }, + { + "epoch": 0.830848, + "grad_norm": 0.38836586475372314, + "learning_rate": 7.314193437450912e-05, + "loss": 2.5815, + "step": 6491 + }, + { + "epoch": 0.830976, + "grad_norm": 0.4021748900413513, + "learning_rate": 7.303401615268878e-05, + "loss": 2.6182, + "step": 6492 + }, + { + "epoch": 0.831104, + "grad_norm": 0.4404360055923462, + "learning_rate": 7.292617133127083e-05, + "loss": 2.6011, + "step": 6493 + }, + { + "epoch": 0.831232, + "grad_norm": 0.4379208981990814, + "learning_rate": 7.281839992879502e-05, + "loss": 2.6833, + "step": 6494 + }, + { + "epoch": 0.83136, + "grad_norm": 0.38900527358055115, + "learning_rate": 7.27107019637886e-05, + "loss": 2.4967, + "step": 6495 + }, + { + "epoch": 0.831488, + "grad_norm": 0.38160789012908936, + "learning_rate": 7.260307745476619e-05, + "loss": 2.6915, + "step": 6496 + }, + { + "epoch": 0.831616, + "grad_norm": 0.39136388897895813, + "learning_rate": 7.249552642022955e-05, + "loss": 2.5851, + "step": 6497 + }, + { + "epoch": 0.831744, + "grad_norm": 0.39539605379104614, + "learning_rate": 7.238804887866796e-05, + "loss": 2.4483, + "step": 6498 + }, + { + "epoch": 0.831872, + "grad_norm": 0.4097229540348053, + "learning_rate": 7.228064484855806e-05, + "loss": 2.6097, + "step": 6499 + }, + { + "epoch": 0.832, + "grad_norm": 0.3920186460018158, + "learning_rate": 7.217331434836394e-05, + "loss": 2.6092, + "step": 6500 + }, + { + "epoch": 0.832128, + "grad_norm": 0.3531385064125061, + "learning_rate": 7.206605739653682e-05, + "loss": 2.6002, + "step": 6501 + }, + { + "epoch": 0.832256, + "grad_norm": 0.4552976191043854, + "learning_rate": 7.195887401151535e-05, + "loss": 2.4925, + "step": 6502 + }, + { + "epoch": 0.832384, + "grad_norm": 0.3987303674221039, + "learning_rate": 7.185176421172573e-05, + "loss": 2.6295, + "step": 6503 + }, + { + "epoch": 0.832512, + "grad_norm": 0.4279789626598358, + "learning_rate": 7.174472801558146e-05, + "loss": 2.4866, + "step": 6504 + }, + { + "epoch": 0.83264, + "grad_norm": 0.41291743516921997, + "learning_rate": 7.16377654414832e-05, + "loss": 2.5774, + "step": 6505 + }, + { + "epoch": 0.832768, + "grad_norm": 0.41168007254600525, + "learning_rate": 7.153087650781909e-05, + "loss": 2.6217, + "step": 6506 + }, + { + "epoch": 0.832896, + "grad_norm": 0.42552945017814636, + "learning_rate": 7.142406123296453e-05, + "loss": 2.6352, + "step": 6507 + }, + { + "epoch": 0.833024, + "grad_norm": 0.4128701090812683, + "learning_rate": 7.131731963528232e-05, + "loss": 2.561, + "step": 6508 + }, + { + "epoch": 0.833152, + "grad_norm": 0.40044286847114563, + "learning_rate": 7.121065173312253e-05, + "loss": 2.6572, + "step": 6509 + }, + { + "epoch": 0.83328, + "grad_norm": 0.4492650032043457, + "learning_rate": 7.110405754482268e-05, + "loss": 2.493, + "step": 6510 + }, + { + "epoch": 0.833408, + "grad_norm": 0.46955016255378723, + "learning_rate": 7.09975370887076e-05, + "loss": 2.6206, + "step": 6511 + }, + { + "epoch": 0.833536, + "grad_norm": 0.4083474278450012, + "learning_rate": 7.089109038308927e-05, + "loss": 2.5506, + "step": 6512 + }, + { + "epoch": 0.833664, + "grad_norm": 0.4565685987472534, + "learning_rate": 7.078471744626707e-05, + "loss": 2.636, + "step": 6513 + }, + { + "epoch": 0.833792, + "grad_norm": 0.3975681662559509, + "learning_rate": 7.067841829652794e-05, + "loss": 2.5897, + "step": 6514 + }, + { + "epoch": 0.83392, + "grad_norm": 0.4207529127597809, + "learning_rate": 7.057219295214579e-05, + "loss": 2.5524, + "step": 6515 + }, + { + "epoch": 0.834048, + "grad_norm": 0.4134220778942108, + "learning_rate": 7.046604143138197e-05, + "loss": 2.5152, + "step": 6516 + }, + { + "epoch": 0.834176, + "grad_norm": 0.3958606719970703, + "learning_rate": 7.035996375248527e-05, + "loss": 2.6193, + "step": 6517 + }, + { + "epoch": 0.834304, + "grad_norm": 0.393385112285614, + "learning_rate": 7.025395993369165e-05, + "loss": 2.5974, + "step": 6518 + }, + { + "epoch": 0.834432, + "grad_norm": 0.588342010974884, + "learning_rate": 7.01480299932244e-05, + "loss": 2.6335, + "step": 6519 + }, + { + "epoch": 0.83456, + "grad_norm": 0.4284498393535614, + "learning_rate": 7.004217394929402e-05, + "loss": 2.5217, + "step": 6520 + }, + { + "epoch": 0.834688, + "grad_norm": 0.3789389133453369, + "learning_rate": 6.993639182009842e-05, + "loss": 2.6047, + "step": 6521 + }, + { + "epoch": 0.834816, + "grad_norm": 0.3803437650203705, + "learning_rate": 6.983068362382273e-05, + "loss": 2.5617, + "step": 6522 + }, + { + "epoch": 0.834944, + "grad_norm": 0.40047764778137207, + "learning_rate": 6.972504937863955e-05, + "loss": 2.5986, + "step": 6523 + }, + { + "epoch": 0.835072, + "grad_norm": 0.4191794991493225, + "learning_rate": 6.961948910270843e-05, + "loss": 2.4867, + "step": 6524 + }, + { + "epoch": 0.8352, + "grad_norm": 0.397727906703949, + "learning_rate": 6.951400281417669e-05, + "loss": 2.5355, + "step": 6525 + }, + { + "epoch": 0.835328, + "grad_norm": 0.38811954855918884, + "learning_rate": 6.940859053117842e-05, + "loss": 2.555, + "step": 6526 + }, + { + "epoch": 0.835456, + "grad_norm": 0.36514905095100403, + "learning_rate": 6.930325227183538e-05, + "loss": 2.6324, + "step": 6527 + }, + { + "epoch": 0.835584, + "grad_norm": 0.3733247220516205, + "learning_rate": 6.919798805425626e-05, + "loss": 2.5005, + "step": 6528 + }, + { + "epoch": 0.835712, + "grad_norm": 0.37960657477378845, + "learning_rate": 6.909279789653733e-05, + "loss": 2.6065, + "step": 6529 + }, + { + "epoch": 0.83584, + "grad_norm": 0.39351025223731995, + "learning_rate": 6.898768181676179e-05, + "loss": 2.5826, + "step": 6530 + }, + { + "epoch": 0.835968, + "grad_norm": 0.3910727798938751, + "learning_rate": 6.888263983300047e-05, + "loss": 2.5199, + "step": 6531 + }, + { + "epoch": 0.836096, + "grad_norm": 0.376213401556015, + "learning_rate": 6.877767196331147e-05, + "loss": 2.5646, + "step": 6532 + }, + { + "epoch": 0.836224, + "grad_norm": 0.3769659996032715, + "learning_rate": 6.867277822573975e-05, + "loss": 2.5526, + "step": 6533 + }, + { + "epoch": 0.836352, + "grad_norm": 0.42861324548721313, + "learning_rate": 6.856795863831788e-05, + "loss": 2.5422, + "step": 6534 + }, + { + "epoch": 0.83648, + "grad_norm": 0.3831743896007538, + "learning_rate": 6.846321321906552e-05, + "loss": 2.5287, + "step": 6535 + }, + { + "epoch": 0.836608, + "grad_norm": 0.4057539403438568, + "learning_rate": 6.835854198598956e-05, + "loss": 2.5162, + "step": 6536 + }, + { + "epoch": 0.836736, + "grad_norm": 0.4198070764541626, + "learning_rate": 6.825394495708415e-05, + "loss": 2.6799, + "step": 6537 + }, + { + "epoch": 0.836864, + "grad_norm": 0.4451785683631897, + "learning_rate": 6.814942215033098e-05, + "loss": 2.5805, + "step": 6538 + }, + { + "epoch": 0.836992, + "grad_norm": 0.3858621120452881, + "learning_rate": 6.804497358369854e-05, + "loss": 2.5177, + "step": 6539 + }, + { + "epoch": 0.83712, + "grad_norm": 0.3733470141887665, + "learning_rate": 6.794059927514268e-05, + "loss": 2.5902, + "step": 6540 + }, + { + "epoch": 0.837248, + "grad_norm": 0.42792120575904846, + "learning_rate": 6.783629924260681e-05, + "loss": 2.4336, + "step": 6541 + }, + { + "epoch": 0.837376, + "grad_norm": 0.4686349928379059, + "learning_rate": 6.773207350402117e-05, + "loss": 2.5873, + "step": 6542 + }, + { + "epoch": 0.837504, + "grad_norm": 0.40974143147468567, + "learning_rate": 6.762792207730334e-05, + "loss": 2.4941, + "step": 6543 + }, + { + "epoch": 0.837632, + "grad_norm": 0.4259672164916992, + "learning_rate": 6.752384498035824e-05, + "loss": 2.6559, + "step": 6544 + }, + { + "epoch": 0.83776, + "grad_norm": 0.3768787086009979, + "learning_rate": 6.741984223107773e-05, + "loss": 2.5021, + "step": 6545 + }, + { + "epoch": 0.837888, + "grad_norm": 0.3675452470779419, + "learning_rate": 6.731591384734137e-05, + "loss": 2.4493, + "step": 6546 + }, + { + "epoch": 0.838016, + "grad_norm": 0.3838387131690979, + "learning_rate": 6.72120598470155e-05, + "loss": 2.5941, + "step": 6547 + }, + { + "epoch": 0.838144, + "grad_norm": 0.43246781826019287, + "learning_rate": 6.710828024795391e-05, + "loss": 2.6317, + "step": 6548 + }, + { + "epoch": 0.838272, + "grad_norm": 0.44331759214401245, + "learning_rate": 6.70045750679974e-05, + "loss": 2.5714, + "step": 6549 + }, + { + "epoch": 0.8384, + "grad_norm": 0.38848474621772766, + "learning_rate": 6.690094432497407e-05, + "loss": 2.565, + "step": 6550 + }, + { + "epoch": 0.838528, + "grad_norm": 0.4013170599937439, + "learning_rate": 6.679738803669943e-05, + "loss": 2.5719, + "step": 6551 + }, + { + "epoch": 0.838656, + "grad_norm": 0.35733574628829956, + "learning_rate": 6.669390622097576e-05, + "loss": 2.6359, + "step": 6552 + }, + { + "epoch": 0.838784, + "grad_norm": 0.4070725440979004, + "learning_rate": 6.65904988955931e-05, + "loss": 2.6258, + "step": 6553 + }, + { + "epoch": 0.838912, + "grad_norm": 0.46662479639053345, + "learning_rate": 6.648716607832811e-05, + "loss": 2.6342, + "step": 6554 + }, + { + "epoch": 0.83904, + "grad_norm": 0.3906477391719818, + "learning_rate": 6.638390778694503e-05, + "loss": 2.5665, + "step": 6555 + }, + { + "epoch": 0.839168, + "grad_norm": 0.4124968349933624, + "learning_rate": 6.62807240391951e-05, + "loss": 2.5966, + "step": 6556 + }, + { + "epoch": 0.839296, + "grad_norm": 0.40746796131134033, + "learning_rate": 6.617761485281687e-05, + "loss": 2.462, + "step": 6557 + }, + { + "epoch": 0.839424, + "grad_norm": 0.4573360085487366, + "learning_rate": 6.607458024553575e-05, + "loss": 2.5789, + "step": 6558 + }, + { + "epoch": 0.839552, + "grad_norm": 0.4276686906814575, + "learning_rate": 6.597162023506492e-05, + "loss": 2.6712, + "step": 6559 + }, + { + "epoch": 0.83968, + "grad_norm": 0.40804415941238403, + "learning_rate": 6.586873483910416e-05, + "loss": 2.5475, + "step": 6560 + }, + { + "epoch": 0.839808, + "grad_norm": 0.38586005568504333, + "learning_rate": 6.576592407534088e-05, + "loss": 2.5905, + "step": 6561 + }, + { + "epoch": 0.839936, + "grad_norm": 0.3927323818206787, + "learning_rate": 6.566318796144933e-05, + "loss": 2.7352, + "step": 6562 + }, + { + "epoch": 0.840064, + "grad_norm": 0.3756961226463318, + "learning_rate": 6.556052651509104e-05, + "loss": 2.5352, + "step": 6563 + }, + { + "epoch": 0.840192, + "grad_norm": 0.4893258810043335, + "learning_rate": 6.545793975391468e-05, + "loss": 2.6588, + "step": 6564 + }, + { + "epoch": 0.84032, + "grad_norm": 0.40187686681747437, + "learning_rate": 6.535542769555608e-05, + "loss": 2.7081, + "step": 6565 + }, + { + "epoch": 0.840448, + "grad_norm": 0.37256669998168945, + "learning_rate": 6.525299035763827e-05, + "loss": 2.5012, + "step": 6566 + }, + { + "epoch": 0.840576, + "grad_norm": 0.46019044518470764, + "learning_rate": 6.515062775777148e-05, + "loss": 2.63, + "step": 6567 + }, + { + "epoch": 0.840704, + "grad_norm": 0.390085369348526, + "learning_rate": 6.504833991355292e-05, + "loss": 2.5725, + "step": 6568 + }, + { + "epoch": 0.840832, + "grad_norm": 0.3959929645061493, + "learning_rate": 6.494612684256718e-05, + "loss": 2.5247, + "step": 6569 + }, + { + "epoch": 0.84096, + "grad_norm": 0.3848709166049957, + "learning_rate": 6.484398856238583e-05, + "loss": 2.6663, + "step": 6570 + }, + { + "epoch": 0.841088, + "grad_norm": 0.3847634196281433, + "learning_rate": 6.47419250905676e-05, + "loss": 2.5269, + "step": 6571 + }, + { + "epoch": 0.841216, + "grad_norm": 0.3914564549922943, + "learning_rate": 6.463993644465843e-05, + "loss": 2.4894, + "step": 6572 + }, + { + "epoch": 0.841344, + "grad_norm": 0.41058674454689026, + "learning_rate": 6.453802264219117e-05, + "loss": 2.5551, + "step": 6573 + }, + { + "epoch": 0.841472, + "grad_norm": 0.439300537109375, + "learning_rate": 6.443618370068622e-05, + "loss": 2.6256, + "step": 6574 + }, + { + "epoch": 0.8416, + "grad_norm": 0.42485150694847107, + "learning_rate": 6.43344196376508e-05, + "loss": 2.587, + "step": 6575 + }, + { + "epoch": 0.841728, + "grad_norm": 0.4196860194206238, + "learning_rate": 6.423273047057932e-05, + "loss": 2.5118, + "step": 6576 + }, + { + "epoch": 0.841856, + "grad_norm": 0.3730558753013611, + "learning_rate": 6.413111621695323e-05, + "loss": 2.6455, + "step": 6577 + }, + { + "epoch": 0.841984, + "grad_norm": 0.40899986028671265, + "learning_rate": 6.402957689424138e-05, + "loss": 2.6611, + "step": 6578 + }, + { + "epoch": 0.842112, + "grad_norm": 0.43016737699508667, + "learning_rate": 6.392811251989945e-05, + "loss": 2.576, + "step": 6579 + }, + { + "epoch": 0.84224, + "grad_norm": 0.3752392828464508, + "learning_rate": 6.382672311137039e-05, + "loss": 2.545, + "step": 6580 + }, + { + "epoch": 0.842368, + "grad_norm": 0.48014745116233826, + "learning_rate": 6.3725408686084e-05, + "loss": 2.5919, + "step": 6581 + }, + { + "epoch": 0.842496, + "grad_norm": 0.4221203327178955, + "learning_rate": 6.362416926145775e-05, + "loss": 2.6381, + "step": 6582 + }, + { + "epoch": 0.842624, + "grad_norm": 0.39720970392227173, + "learning_rate": 6.352300485489571e-05, + "loss": 2.6442, + "step": 6583 + }, + { + "epoch": 0.842752, + "grad_norm": 0.375165194272995, + "learning_rate": 6.342191548378923e-05, + "loss": 2.5303, + "step": 6584 + }, + { + "epoch": 0.84288, + "grad_norm": 0.4187237024307251, + "learning_rate": 6.33209011655167e-05, + "loss": 2.5594, + "step": 6585 + }, + { + "epoch": 0.843008, + "grad_norm": 0.5330372452735901, + "learning_rate": 6.321996191744367e-05, + "loss": 2.624, + "step": 6586 + }, + { + "epoch": 0.843136, + "grad_norm": 0.3944437801837921, + "learning_rate": 6.311909775692265e-05, + "loss": 2.4246, + "step": 6587 + }, + { + "epoch": 0.843264, + "grad_norm": 0.4092988669872284, + "learning_rate": 6.301830870129349e-05, + "loss": 2.6326, + "step": 6588 + }, + { + "epoch": 0.843392, + "grad_norm": 0.4333173930644989, + "learning_rate": 6.291759476788312e-05, + "loss": 2.5805, + "step": 6589 + }, + { + "epoch": 0.84352, + "grad_norm": 0.3961753249168396, + "learning_rate": 6.281695597400534e-05, + "loss": 2.4024, + "step": 6590 + }, + { + "epoch": 0.843648, + "grad_norm": 0.4012063145637512, + "learning_rate": 6.271639233696103e-05, + "loss": 2.6078, + "step": 6591 + }, + { + "epoch": 0.843776, + "grad_norm": 0.4615851640701294, + "learning_rate": 6.261590387403832e-05, + "loss": 2.5333, + "step": 6592 + }, + { + "epoch": 0.843904, + "grad_norm": 0.4336642324924469, + "learning_rate": 6.251549060251233e-05, + "loss": 2.6099, + "step": 6593 + }, + { + "epoch": 0.844032, + "grad_norm": 0.4094434380531311, + "learning_rate": 6.241515253964513e-05, + "loss": 2.5672, + "step": 6594 + }, + { + "epoch": 0.84416, + "grad_norm": 0.40127405524253845, + "learning_rate": 6.231488970268628e-05, + "loss": 2.4914, + "step": 6595 + }, + { + "epoch": 0.844288, + "grad_norm": 0.40969613194465637, + "learning_rate": 6.221470210887181e-05, + "loss": 2.6574, + "step": 6596 + }, + { + "epoch": 0.844416, + "grad_norm": 0.4183996617794037, + "learning_rate": 6.211458977542545e-05, + "loss": 2.7572, + "step": 6597 + }, + { + "epoch": 0.844544, + "grad_norm": 0.4099719822406769, + "learning_rate": 6.201455271955747e-05, + "loss": 2.6757, + "step": 6598 + }, + { + "epoch": 0.844672, + "grad_norm": 0.4136928915977478, + "learning_rate": 6.191459095846546e-05, + "loss": 2.5363, + "step": 6599 + }, + { + "epoch": 0.8448, + "grad_norm": 0.38702648878097534, + "learning_rate": 6.181470450933397e-05, + "loss": 2.6117, + "step": 6600 + }, + { + "epoch": 0.844928, + "grad_norm": 0.41234090924263, + "learning_rate": 6.171489338933466e-05, + "loss": 2.482, + "step": 6601 + }, + { + "epoch": 0.845056, + "grad_norm": 0.4077073931694031, + "learning_rate": 6.161515761562614e-05, + "loss": 2.6248, + "step": 6602 + }, + { + "epoch": 0.845184, + "grad_norm": 0.37565475702285767, + "learning_rate": 6.151549720535432e-05, + "loss": 2.5795, + "step": 6603 + }, + { + "epoch": 0.845312, + "grad_norm": 0.41597482562065125, + "learning_rate": 6.141591217565185e-05, + "loss": 2.5345, + "step": 6604 + }, + { + "epoch": 0.84544, + "grad_norm": 0.41309311985969543, + "learning_rate": 6.131640254363846e-05, + "loss": 2.5562, + "step": 6605 + }, + { + "epoch": 0.845568, + "grad_norm": 0.4693642556667328, + "learning_rate": 6.121696832642126e-05, + "loss": 2.6063, + "step": 6606 + }, + { + "epoch": 0.845696, + "grad_norm": 0.4153495132923126, + "learning_rate": 6.111760954109402e-05, + "loss": 2.6011, + "step": 6607 + }, + { + "epoch": 0.845824, + "grad_norm": 0.9296197295188904, + "learning_rate": 6.101832620473763e-05, + "loss": 2.6613, + "step": 6608 + }, + { + "epoch": 0.845952, + "grad_norm": 0.44810396432876587, + "learning_rate": 6.0919118334419985e-05, + "loss": 2.6835, + "step": 6609 + }, + { + "epoch": 0.84608, + "grad_norm": 0.4078626036643982, + "learning_rate": 6.081998594719629e-05, + "loss": 2.5799, + "step": 6610 + }, + { + "epoch": 0.846208, + "grad_norm": 0.4163365066051483, + "learning_rate": 6.07209290601084e-05, + "loss": 2.6382, + "step": 6611 + }, + { + "epoch": 0.846336, + "grad_norm": 0.41507449746131897, + "learning_rate": 6.062194769018542e-05, + "loss": 2.6643, + "step": 6612 + }, + { + "epoch": 0.846464, + "grad_norm": 0.4325169026851654, + "learning_rate": 6.0523041854443295e-05, + "loss": 2.6268, + "step": 6613 + }, + { + "epoch": 0.846592, + "grad_norm": 0.39414486289024353, + "learning_rate": 6.042421156988498e-05, + "loss": 2.6053, + "step": 6614 + }, + { + "epoch": 0.84672, + "grad_norm": 0.4516761600971222, + "learning_rate": 6.032545685350088e-05, + "loss": 2.5914, + "step": 6615 + }, + { + "epoch": 0.846848, + "grad_norm": 0.518183171749115, + "learning_rate": 6.0226777722267814e-05, + "loss": 2.6399, + "step": 6616 + }, + { + "epoch": 0.846976, + "grad_norm": 0.46503201127052307, + "learning_rate": 6.012817419314992e-05, + "loss": 2.4805, + "step": 6617 + }, + { + "epoch": 0.847104, + "grad_norm": 0.4206988513469696, + "learning_rate": 6.002964628309837e-05, + "loss": 2.5678, + "step": 6618 + }, + { + "epoch": 0.847232, + "grad_norm": 0.4013086259365082, + "learning_rate": 5.993119400905123e-05, + "loss": 2.5701, + "step": 6619 + }, + { + "epoch": 0.84736, + "grad_norm": 0.41361841559410095, + "learning_rate": 5.98328173879335e-05, + "loss": 2.6108, + "step": 6620 + }, + { + "epoch": 0.847488, + "grad_norm": 0.3851803243160248, + "learning_rate": 5.9734516436657335e-05, + "loss": 2.6175, + "step": 6621 + }, + { + "epoch": 0.847616, + "grad_norm": 0.3925992548465729, + "learning_rate": 5.9636291172121824e-05, + "loss": 2.6348, + "step": 6622 + }, + { + "epoch": 0.847744, + "grad_norm": 0.4725612998008728, + "learning_rate": 5.95381416112129e-05, + "loss": 2.4342, + "step": 6623 + }, + { + "epoch": 0.847872, + "grad_norm": 0.40100958943367004, + "learning_rate": 5.944006777080363e-05, + "loss": 2.5856, + "step": 6624 + }, + { + "epoch": 0.848, + "grad_norm": 0.4318883717060089, + "learning_rate": 5.9342069667754295e-05, + "loss": 2.5099, + "step": 6625 + }, + { + "epoch": 0.848128, + "grad_norm": 0.3778785169124603, + "learning_rate": 5.92441473189117e-05, + "loss": 2.5888, + "step": 6626 + }, + { + "epoch": 0.848256, + "grad_norm": 0.4491979777812958, + "learning_rate": 5.9146300741109885e-05, + "loss": 2.5572, + "step": 6627 + }, + { + "epoch": 0.848384, + "grad_norm": 0.41188931465148926, + "learning_rate": 5.9048529951169775e-05, + "loss": 2.5604, + "step": 6628 + }, + { + "epoch": 0.848512, + "grad_norm": 0.3925970494747162, + "learning_rate": 5.8950834965899315e-05, + "loss": 2.5706, + "step": 6629 + }, + { + "epoch": 0.84864, + "grad_norm": 0.4252620339393616, + "learning_rate": 5.8853215802093244e-05, + "loss": 2.7122, + "step": 6630 + }, + { + "epoch": 0.848768, + "grad_norm": 0.38431841135025024, + "learning_rate": 5.875567247653374e-05, + "loss": 2.5236, + "step": 6631 + }, + { + "epoch": 0.848896, + "grad_norm": 0.4339466392993927, + "learning_rate": 5.8658205005989506e-05, + "loss": 2.6435, + "step": 6632 + }, + { + "epoch": 0.849024, + "grad_norm": 0.4465898275375366, + "learning_rate": 5.85608134072162e-05, + "loss": 2.684, + "step": 6633 + }, + { + "epoch": 0.849152, + "grad_norm": 0.42482665181159973, + "learning_rate": 5.8463497696956746e-05, + "loss": 2.5131, + "step": 6634 + }, + { + "epoch": 0.84928, + "grad_norm": 0.43159255385398865, + "learning_rate": 5.836625789194078e-05, + "loss": 2.5969, + "step": 6635 + }, + { + "epoch": 0.849408, + "grad_norm": 0.4397905170917511, + "learning_rate": 5.826909400888491e-05, + "loss": 2.5799, + "step": 6636 + }, + { + "epoch": 0.849536, + "grad_norm": 0.4228934347629547, + "learning_rate": 5.81720060644928e-05, + "loss": 2.6459, + "step": 6637 + }, + { + "epoch": 0.849664, + "grad_norm": 0.39497336745262146, + "learning_rate": 5.807499407545486e-05, + "loss": 2.5635, + "step": 6638 + }, + { + "epoch": 0.849792, + "grad_norm": 0.38154956698417664, + "learning_rate": 5.797805805844875e-05, + "loss": 2.669, + "step": 6639 + }, + { + "epoch": 0.84992, + "grad_norm": 0.43370723724365234, + "learning_rate": 5.7881198030138806e-05, + "loss": 2.4978, + "step": 6640 + }, + { + "epoch": 0.850048, + "grad_norm": 0.40133240818977356, + "learning_rate": 5.7784414007176436e-05, + "loss": 2.4695, + "step": 6641 + }, + { + "epoch": 0.850176, + "grad_norm": 0.42155149579048157, + "learning_rate": 5.768770600619977e-05, + "loss": 2.612, + "step": 6642 + }, + { + "epoch": 0.850304, + "grad_norm": 0.4143686592578888, + "learning_rate": 5.759107404383429e-05, + "loss": 2.5307, + "step": 6643 + }, + { + "epoch": 0.850432, + "grad_norm": 0.38390135765075684, + "learning_rate": 5.749451813669204e-05, + "loss": 2.5416, + "step": 6644 + }, + { + "epoch": 0.85056, + "grad_norm": 0.4028919041156769, + "learning_rate": 5.7398038301371916e-05, + "loss": 2.5538, + "step": 6645 + }, + { + "epoch": 0.850688, + "grad_norm": 0.46274346113204956, + "learning_rate": 5.7301634554460246e-05, + "loss": 2.6069, + "step": 6646 + }, + { + "epoch": 0.850816, + "grad_norm": 0.3701043725013733, + "learning_rate": 5.7205306912529786e-05, + "loss": 2.5341, + "step": 6647 + }, + { + "epoch": 0.850944, + "grad_norm": 0.3708423674106598, + "learning_rate": 5.71090553921404e-05, + "loss": 2.5744, + "step": 6648 + }, + { + "epoch": 0.851072, + "grad_norm": 0.4202256500720978, + "learning_rate": 5.701288000983884e-05, + "loss": 2.5567, + "step": 6649 + }, + { + "epoch": 0.8512, + "grad_norm": 0.40638408064842224, + "learning_rate": 5.691678078215873e-05, + "loss": 2.6017, + "step": 6650 + }, + { + "epoch": 0.851328, + "grad_norm": 0.3689452111721039, + "learning_rate": 5.682075772562051e-05, + "loss": 2.5636, + "step": 6651 + }, + { + "epoch": 0.851456, + "grad_norm": 0.49226322770118713, + "learning_rate": 5.672481085673198e-05, + "loss": 2.6842, + "step": 6652 + }, + { + "epoch": 0.851584, + "grad_norm": 0.3862845301628113, + "learning_rate": 5.662894019198722e-05, + "loss": 2.481, + "step": 6653 + }, + { + "epoch": 0.851712, + "grad_norm": 0.4918995797634125, + "learning_rate": 5.653314574786778e-05, + "loss": 2.5667, + "step": 6654 + }, + { + "epoch": 0.85184, + "grad_norm": 0.39168012142181396, + "learning_rate": 5.643742754084163e-05, + "loss": 2.6251, + "step": 6655 + }, + { + "epoch": 0.851968, + "grad_norm": 0.4656273424625397, + "learning_rate": 5.634178558736397e-05, + "loss": 2.5966, + "step": 6656 + }, + { + "epoch": 0.852096, + "grad_norm": 0.5611627697944641, + "learning_rate": 5.624621990387668e-05, + "loss": 2.5585, + "step": 6657 + }, + { + "epoch": 0.852224, + "grad_norm": 0.37679561972618103, + "learning_rate": 5.615073050680858e-05, + "loss": 2.5398, + "step": 6658 + }, + { + "epoch": 0.852352, + "grad_norm": 0.39829158782958984, + "learning_rate": 5.6055317412575354e-05, + "loss": 2.6805, + "step": 6659 + }, + { + "epoch": 0.85248, + "grad_norm": 0.3865181505680084, + "learning_rate": 5.5959980637579844e-05, + "loss": 2.4354, + "step": 6660 + }, + { + "epoch": 0.852608, + "grad_norm": 0.43774914741516113, + "learning_rate": 5.5864720198211316e-05, + "loss": 2.5962, + "step": 6661 + }, + { + "epoch": 0.852736, + "grad_norm": 0.3983827531337738, + "learning_rate": 5.5769536110846344e-05, + "loss": 2.5805, + "step": 6662 + }, + { + "epoch": 0.852864, + "grad_norm": 0.43789899349212646, + "learning_rate": 5.567442839184805e-05, + "loss": 2.5546, + "step": 6663 + }, + { + "epoch": 0.852992, + "grad_norm": 0.39641985297203064, + "learning_rate": 5.5579397057566646e-05, + "loss": 2.5614, + "step": 6664 + }, + { + "epoch": 0.85312, + "grad_norm": 0.4369643032550812, + "learning_rate": 5.548444212433901e-05, + "loss": 2.6189, + "step": 6665 + }, + { + "epoch": 0.853248, + "grad_norm": 0.4274299144744873, + "learning_rate": 5.53895636084889e-05, + "loss": 2.5694, + "step": 6666 + }, + { + "epoch": 0.853376, + "grad_norm": 0.532335102558136, + "learning_rate": 5.5294761526327345e-05, + "loss": 2.539, + "step": 6667 + }, + { + "epoch": 0.853504, + "grad_norm": 0.3811858594417572, + "learning_rate": 5.520003589415168e-05, + "loss": 2.4569, + "step": 6668 + }, + { + "epoch": 0.853632, + "grad_norm": 0.4006649851799011, + "learning_rate": 5.510538672824644e-05, + "loss": 2.5181, + "step": 6669 + }, + { + "epoch": 0.85376, + "grad_norm": 0.4060637056827545, + "learning_rate": 5.501081404488278e-05, + "loss": 2.5084, + "step": 6670 + }, + { + "epoch": 0.853888, + "grad_norm": 0.4315345287322998, + "learning_rate": 5.491631786031903e-05, + "loss": 2.5707, + "step": 6671 + }, + { + "epoch": 0.854016, + "grad_norm": 0.42627450823783875, + "learning_rate": 5.482189819080013e-05, + "loss": 2.4615, + "step": 6672 + }, + { + "epoch": 0.854144, + "grad_norm": 0.4380694329738617, + "learning_rate": 5.472755505255783e-05, + "loss": 2.5943, + "step": 6673 + }, + { + "epoch": 0.854272, + "grad_norm": 0.4655149579048157, + "learning_rate": 5.463328846181082e-05, + "loss": 2.5837, + "step": 6674 + }, + { + "epoch": 0.8544, + "grad_norm": 0.43338871002197266, + "learning_rate": 5.453909843476468e-05, + "loss": 2.6077, + "step": 6675 + }, + { + "epoch": 0.854528, + "grad_norm": 0.4532991051673889, + "learning_rate": 5.44449849876118e-05, + "loss": 2.5457, + "step": 6676 + }, + { + "epoch": 0.854656, + "grad_norm": 0.3955659866333008, + "learning_rate": 5.4350948136531296e-05, + "loss": 2.5906, + "step": 6677 + }, + { + "epoch": 0.854784, + "grad_norm": 0.5145876407623291, + "learning_rate": 5.4256987897689234e-05, + "loss": 2.7347, + "step": 6678 + }, + { + "epoch": 0.854912, + "grad_norm": 0.4313179552555084, + "learning_rate": 5.416310428723836e-05, + "loss": 2.4535, + "step": 6679 + }, + { + "epoch": 0.85504, + "grad_norm": 0.4393027126789093, + "learning_rate": 5.406929732131849e-05, + "loss": 2.6446, + "step": 6680 + }, + { + "epoch": 0.855168, + "grad_norm": 0.3964560329914093, + "learning_rate": 5.397556701605605e-05, + "loss": 2.531, + "step": 6681 + }, + { + "epoch": 0.855296, + "grad_norm": 0.4264828860759735, + "learning_rate": 5.388191338756443e-05, + "loss": 2.6149, + "step": 6682 + }, + { + "epoch": 0.855424, + "grad_norm": 0.4092752933502197, + "learning_rate": 5.378833645194375e-05, + "loss": 2.5556, + "step": 6683 + }, + { + "epoch": 0.855552, + "grad_norm": 0.44376590847969055, + "learning_rate": 5.369483622528104e-05, + "loss": 2.4862, + "step": 6684 + }, + { + "epoch": 0.85568, + "grad_norm": 0.4358241856098175, + "learning_rate": 5.360141272364993e-05, + "loss": 2.5298, + "step": 6685 + }, + { + "epoch": 0.855808, + "grad_norm": 0.38280075788497925, + "learning_rate": 5.350806596311109e-05, + "loss": 2.5073, + "step": 6686 + }, + { + "epoch": 0.855936, + "grad_norm": 0.39145979285240173, + "learning_rate": 5.341479595971177e-05, + "loss": 2.5525, + "step": 6687 + }, + { + "epoch": 0.856064, + "grad_norm": 0.42794057726860046, + "learning_rate": 5.3321602729486275e-05, + "loss": 2.6556, + "step": 6688 + }, + { + "epoch": 0.856192, + "grad_norm": 0.40619802474975586, + "learning_rate": 5.322848628845578e-05, + "loss": 2.5348, + "step": 6689 + }, + { + "epoch": 0.85632, + "grad_norm": 0.37558096647262573, + "learning_rate": 5.313544665262782e-05, + "loss": 2.6622, + "step": 6690 + }, + { + "epoch": 0.856448, + "grad_norm": 0.48782026767730713, + "learning_rate": 5.3042483837997145e-05, + "loss": 2.5141, + "step": 6691 + }, + { + "epoch": 0.856576, + "grad_norm": 0.41296449303627014, + "learning_rate": 5.294959786054504e-05, + "loss": 2.4893, + "step": 6692 + }, + { + "epoch": 0.856704, + "grad_norm": 0.4259023070335388, + "learning_rate": 5.285678873623972e-05, + "loss": 2.4507, + "step": 6693 + }, + { + "epoch": 0.856832, + "grad_norm": 0.37933433055877686, + "learning_rate": 5.276405648103616e-05, + "loss": 2.6636, + "step": 6694 + }, + { + "epoch": 0.85696, + "grad_norm": 0.568436324596405, + "learning_rate": 5.267140111087604e-05, + "loss": 2.5941, + "step": 6695 + }, + { + "epoch": 0.857088, + "grad_norm": 0.4006655514240265, + "learning_rate": 5.257882264168795e-05, + "loss": 2.5139, + "step": 6696 + }, + { + "epoch": 0.857216, + "grad_norm": 0.4243081212043762, + "learning_rate": 5.248632108938728e-05, + "loss": 2.5198, + "step": 6697 + }, + { + "epoch": 0.857344, + "grad_norm": 0.4189431369304657, + "learning_rate": 5.239389646987591e-05, + "loss": 2.5866, + "step": 6698 + }, + { + "epoch": 0.857472, + "grad_norm": 0.40218445658683777, + "learning_rate": 5.230154879904292e-05, + "loss": 2.5514, + "step": 6699 + }, + { + "epoch": 0.8576, + "grad_norm": 0.3797632157802582, + "learning_rate": 5.2209278092763824e-05, + "loss": 2.5804, + "step": 6700 + }, + { + "epoch": 0.857728, + "grad_norm": 0.45747485756874084, + "learning_rate": 5.21170843669011e-05, + "loss": 2.5489, + "step": 6701 + }, + { + "epoch": 0.857856, + "grad_norm": 0.3891725242137909, + "learning_rate": 5.20249676373038e-05, + "loss": 2.5476, + "step": 6702 + }, + { + "epoch": 0.857984, + "grad_norm": 0.4233708381652832, + "learning_rate": 5.1932927919808025e-05, + "loss": 2.6693, + "step": 6703 + }, + { + "epoch": 0.858112, + "grad_norm": 0.5266421437263489, + "learning_rate": 5.184096523023635e-05, + "loss": 2.6449, + "step": 6704 + }, + { + "epoch": 0.85824, + "grad_norm": 0.41411030292510986, + "learning_rate": 5.1749079584398284e-05, + "loss": 2.5588, + "step": 6705 + }, + { + "epoch": 0.858368, + "grad_norm": 0.39509135484695435, + "learning_rate": 5.165727099808998e-05, + "loss": 2.5031, + "step": 6706 + }, + { + "epoch": 0.858496, + "grad_norm": 0.4143739640712738, + "learning_rate": 5.156553948709436e-05, + "loss": 2.5266, + "step": 6707 + }, + { + "epoch": 0.858624, + "grad_norm": 0.3950138986110687, + "learning_rate": 5.147388506718126e-05, + "loss": 2.4397, + "step": 6708 + }, + { + "epoch": 0.858752, + "grad_norm": 0.4681456387042999, + "learning_rate": 5.138230775410713e-05, + "loss": 2.5294, + "step": 6709 + }, + { + "epoch": 0.85888, + "grad_norm": 0.41264742612838745, + "learning_rate": 5.1290807563614984e-05, + "loss": 2.5927, + "step": 6710 + }, + { + "epoch": 0.859008, + "grad_norm": 0.47311505675315857, + "learning_rate": 5.1199384511435023e-05, + "loss": 2.5073, + "step": 6711 + }, + { + "epoch": 0.859136, + "grad_norm": 0.4491392970085144, + "learning_rate": 5.110803861328378e-05, + "loss": 2.619, + "step": 6712 + }, + { + "epoch": 0.859264, + "grad_norm": 0.39232349395751953, + "learning_rate": 5.10167698848647e-05, + "loss": 2.6169, + "step": 6713 + }, + { + "epoch": 0.859392, + "grad_norm": 0.4404979646205902, + "learning_rate": 5.092557834186795e-05, + "loss": 2.5275, + "step": 6714 + }, + { + "epoch": 0.85952, + "grad_norm": 0.4441821575164795, + "learning_rate": 5.083446399997044e-05, + "loss": 2.7279, + "step": 6715 + }, + { + "epoch": 0.859648, + "grad_norm": 0.47439196705818176, + "learning_rate": 5.0743426874835574e-05, + "loss": 2.5189, + "step": 6716 + }, + { + "epoch": 0.859776, + "grad_norm": 0.42082712054252625, + "learning_rate": 5.0652466982113896e-05, + "loss": 2.5709, + "step": 6717 + }, + { + "epoch": 0.859904, + "grad_norm": 0.4279492199420929, + "learning_rate": 5.0561584337442514e-05, + "loss": 2.5718, + "step": 6718 + }, + { + "epoch": 0.860032, + "grad_norm": 0.3799394965171814, + "learning_rate": 5.047077895644514e-05, + "loss": 2.6586, + "step": 6719 + }, + { + "epoch": 0.86016, + "grad_norm": 0.4114617705345154, + "learning_rate": 5.038005085473224e-05, + "loss": 2.4848, + "step": 6720 + }, + { + "epoch": 0.860288, + "grad_norm": 0.41541942954063416, + "learning_rate": 5.0289400047901055e-05, + "loss": 2.5064, + "step": 6721 + }, + { + "epoch": 0.860416, + "grad_norm": 0.3723580837249756, + "learning_rate": 5.019882655153546e-05, + "loss": 2.5643, + "step": 6722 + }, + { + "epoch": 0.860544, + "grad_norm": 0.4416189193725586, + "learning_rate": 5.0108330381206044e-05, + "loss": 2.5363, + "step": 6723 + }, + { + "epoch": 0.860672, + "grad_norm": 0.3872107267379761, + "learning_rate": 5.001791155247032e-05, + "loss": 2.4798, + "step": 6724 + }, + { + "epoch": 0.8608, + "grad_norm": 0.4042311906814575, + "learning_rate": 4.9927570080872176e-05, + "loss": 2.7057, + "step": 6725 + }, + { + "epoch": 0.860928, + "grad_norm": 0.36621078848838806, + "learning_rate": 4.983730598194247e-05, + "loss": 2.493, + "step": 6726 + }, + { + "epoch": 0.861056, + "grad_norm": 0.36256143450737, + "learning_rate": 4.974711927119863e-05, + "loss": 2.5353, + "step": 6727 + }, + { + "epoch": 0.861184, + "grad_norm": 0.44287779927253723, + "learning_rate": 4.9657009964144794e-05, + "loss": 2.538, + "step": 6728 + }, + { + "epoch": 0.861312, + "grad_norm": 0.376102089881897, + "learning_rate": 4.956697807627169e-05, + "loss": 2.5598, + "step": 6729 + }, + { + "epoch": 0.86144, + "grad_norm": 0.3954898715019226, + "learning_rate": 4.9477023623057e-05, + "loss": 2.6053, + "step": 6730 + }, + { + "epoch": 0.861568, + "grad_norm": 0.4088953733444214, + "learning_rate": 4.9387146619964684e-05, + "loss": 2.6639, + "step": 6731 + }, + { + "epoch": 0.861696, + "grad_norm": 0.39725250005722046, + "learning_rate": 4.929734708244599e-05, + "loss": 2.6339, + "step": 6732 + }, + { + "epoch": 0.861824, + "grad_norm": 0.4083765149116516, + "learning_rate": 4.9207625025938285e-05, + "loss": 2.6215, + "step": 6733 + }, + { + "epoch": 0.861952, + "grad_norm": 0.4175489842891693, + "learning_rate": 4.911798046586591e-05, + "loss": 2.5671, + "step": 6734 + }, + { + "epoch": 0.86208, + "grad_norm": 0.39047038555145264, + "learning_rate": 4.902841341763964e-05, + "loss": 2.6226, + "step": 6735 + }, + { + "epoch": 0.862208, + "grad_norm": 0.4189351797103882, + "learning_rate": 4.893892389665727e-05, + "loss": 2.5136, + "step": 6736 + }, + { + "epoch": 0.862336, + "grad_norm": 0.4215710461139679, + "learning_rate": 4.8849511918303117e-05, + "loss": 2.579, + "step": 6737 + }, + { + "epoch": 0.862464, + "grad_norm": 0.4035865068435669, + "learning_rate": 4.8760177497947866e-05, + "loss": 2.6113, + "step": 6738 + }, + { + "epoch": 0.862592, + "grad_norm": 0.3991084396839142, + "learning_rate": 4.867092065094947e-05, + "loss": 2.5741, + "step": 6739 + }, + { + "epoch": 0.86272, + "grad_norm": 0.4358431100845337, + "learning_rate": 4.858174139265209e-05, + "loss": 2.7063, + "step": 6740 + }, + { + "epoch": 0.862848, + "grad_norm": 0.7047058343887329, + "learning_rate": 4.849263973838664e-05, + "loss": 2.7167, + "step": 6741 + }, + { + "epoch": 0.862976, + "grad_norm": 0.5041335821151733, + "learning_rate": 4.840361570347068e-05, + "loss": 2.6587, + "step": 6742 + }, + { + "epoch": 0.863104, + "grad_norm": 0.41915205121040344, + "learning_rate": 4.8314669303208605e-05, + "loss": 2.5958, + "step": 6743 + }, + { + "epoch": 0.863232, + "grad_norm": 0.3762286901473999, + "learning_rate": 4.822580055289116e-05, + "loss": 2.6018, + "step": 6744 + }, + { + "epoch": 0.86336, + "grad_norm": 0.3664076328277588, + "learning_rate": 4.8137009467795976e-05, + "loss": 2.5765, + "step": 6745 + }, + { + "epoch": 0.863488, + "grad_norm": 0.42949652671813965, + "learning_rate": 4.8048296063187446e-05, + "loss": 2.5303, + "step": 6746 + }, + { + "epoch": 0.863616, + "grad_norm": 0.3729037344455719, + "learning_rate": 4.795966035431626e-05, + "loss": 2.5166, + "step": 6747 + }, + { + "epoch": 0.863744, + "grad_norm": 0.4241585433483124, + "learning_rate": 4.787110235642001e-05, + "loss": 2.5164, + "step": 6748 + }, + { + "epoch": 0.863872, + "grad_norm": 0.3959144651889801, + "learning_rate": 4.778262208472273e-05, + "loss": 2.5743, + "step": 6749 + }, + { + "epoch": 0.864, + "grad_norm": 0.47864043712615967, + "learning_rate": 4.7694219554435316e-05, + "loss": 2.641, + "step": 6750 + }, + { + "epoch": 0.864128, + "grad_norm": 0.35452380776405334, + "learning_rate": 4.760589478075511e-05, + "loss": 2.4383, + "step": 6751 + }, + { + "epoch": 0.864256, + "grad_norm": 0.4857720732688904, + "learning_rate": 4.751764777886614e-05, + "loss": 2.5656, + "step": 6752 + }, + { + "epoch": 0.864384, + "grad_norm": 0.39362087845802307, + "learning_rate": 4.742947856393909e-05, + "loss": 2.6816, + "step": 6753 + }, + { + "epoch": 0.864512, + "grad_norm": 0.40556901693344116, + "learning_rate": 4.734138715113145e-05, + "loss": 2.5663, + "step": 6754 + }, + { + "epoch": 0.86464, + "grad_norm": 0.369126558303833, + "learning_rate": 4.7253373555587006e-05, + "loss": 2.521, + "step": 6755 + }, + { + "epoch": 0.864768, + "grad_norm": 0.5318395495414734, + "learning_rate": 4.716543779243637e-05, + "loss": 2.5577, + "step": 6756 + }, + { + "epoch": 0.864896, + "grad_norm": 0.4389074146747589, + "learning_rate": 4.707757987679667e-05, + "loss": 2.733, + "step": 6757 + }, + { + "epoch": 0.865024, + "grad_norm": 0.40680480003356934, + "learning_rate": 4.6989799823771714e-05, + "loss": 2.5771, + "step": 6758 + }, + { + "epoch": 0.865152, + "grad_norm": 0.45674943923950195, + "learning_rate": 4.6902097648451824e-05, + "loss": 2.5631, + "step": 6759 + }, + { + "epoch": 0.86528, + "grad_norm": 0.41087332367897034, + "learning_rate": 4.681447336591421e-05, + "loss": 2.5071, + "step": 6760 + }, + { + "epoch": 0.865408, + "grad_norm": 0.3605395257472992, + "learning_rate": 4.672692699122238e-05, + "loss": 2.4962, + "step": 6761 + }, + { + "epoch": 0.865536, + "grad_norm": 0.4103125035762787, + "learning_rate": 4.663945853942653e-05, + "loss": 2.5134, + "step": 6762 + }, + { + "epoch": 0.865664, + "grad_norm": 0.4060099422931671, + "learning_rate": 4.655206802556361e-05, + "loss": 2.5388, + "step": 6763 + }, + { + "epoch": 0.865792, + "grad_norm": 0.40063557028770447, + "learning_rate": 4.646475546465706e-05, + "loss": 2.5856, + "step": 6764 + }, + { + "epoch": 0.86592, + "grad_norm": 0.4228542447090149, + "learning_rate": 4.637752087171687e-05, + "loss": 2.698, + "step": 6765 + }, + { + "epoch": 0.866048, + "grad_norm": 0.4328107237815857, + "learning_rate": 4.6290364261739545e-05, + "loss": 2.7182, + "step": 6766 + }, + { + "epoch": 0.866176, + "grad_norm": 0.391055166721344, + "learning_rate": 4.6203285649708603e-05, + "loss": 2.6487, + "step": 6767 + }, + { + "epoch": 0.866304, + "grad_norm": 0.3719969689846039, + "learning_rate": 4.611628505059373e-05, + "loss": 2.4797, + "step": 6768 + }, + { + "epoch": 0.866432, + "grad_norm": 0.40859177708625793, + "learning_rate": 4.60293624793513e-05, + "loss": 2.5784, + "step": 6769 + }, + { + "epoch": 0.86656, + "grad_norm": 0.404646098613739, + "learning_rate": 4.59425179509243e-05, + "loss": 2.565, + "step": 6770 + }, + { + "epoch": 0.866688, + "grad_norm": 0.3828140199184418, + "learning_rate": 4.58557514802424e-05, + "loss": 2.5321, + "step": 6771 + }, + { + "epoch": 0.866816, + "grad_norm": 0.40080761909484863, + "learning_rate": 4.576906308222162e-05, + "loss": 2.5842, + "step": 6772 + }, + { + "epoch": 0.866944, + "grad_norm": 0.41179677844047546, + "learning_rate": 4.568245277176486e-05, + "loss": 2.7195, + "step": 6773 + }, + { + "epoch": 0.867072, + "grad_norm": 0.39691776037216187, + "learning_rate": 4.559592056376133e-05, + "loss": 2.6026, + "step": 6774 + }, + { + "epoch": 0.8672, + "grad_norm": 0.43765413761138916, + "learning_rate": 4.550946647308701e-05, + "loss": 2.5854, + "step": 6775 + }, + { + "epoch": 0.867328, + "grad_norm": 0.39732328057289124, + "learning_rate": 4.5423090514604304e-05, + "loss": 2.4924, + "step": 6776 + }, + { + "epoch": 0.867456, + "grad_norm": 0.4488281309604645, + "learning_rate": 4.533679270316227e-05, + "loss": 2.6426, + "step": 6777 + }, + { + "epoch": 0.867584, + "grad_norm": 0.37922757863998413, + "learning_rate": 4.525057305359648e-05, + "loss": 2.5558, + "step": 6778 + }, + { + "epoch": 0.867712, + "grad_norm": 0.514557421207428, + "learning_rate": 4.5164431580729084e-05, + "loss": 2.7129, + "step": 6779 + }, + { + "epoch": 0.86784, + "grad_norm": 0.38466519117355347, + "learning_rate": 4.5078368299368675e-05, + "loss": 2.5464, + "step": 6780 + }, + { + "epoch": 0.867968, + "grad_norm": 0.38552820682525635, + "learning_rate": 4.499238322431071e-05, + "loss": 2.6104, + "step": 6781 + }, + { + "epoch": 0.868096, + "grad_norm": 0.42582404613494873, + "learning_rate": 4.490647637033701e-05, + "loss": 2.6498, + "step": 6782 + }, + { + "epoch": 0.868224, + "grad_norm": 0.4080196022987366, + "learning_rate": 4.482064775221595e-05, + "loss": 2.6051, + "step": 6783 + }, + { + "epoch": 0.868352, + "grad_norm": 0.36494824290275574, + "learning_rate": 4.473489738470243e-05, + "loss": 2.5392, + "step": 6784 + }, + { + "epoch": 0.86848, + "grad_norm": 0.4244120419025421, + "learning_rate": 4.464922528253795e-05, + "loss": 2.5264, + "step": 6785 + }, + { + "epoch": 0.868608, + "grad_norm": 0.4095558226108551, + "learning_rate": 4.4563631460450505e-05, + "loss": 2.6463, + "step": 6786 + }, + { + "epoch": 0.868736, + "grad_norm": 0.43133020401000977, + "learning_rate": 4.447811593315459e-05, + "loss": 2.5325, + "step": 6787 + }, + { + "epoch": 0.868864, + "grad_norm": 0.43680840730667114, + "learning_rate": 4.439267871535152e-05, + "loss": 2.5821, + "step": 6788 + }, + { + "epoch": 0.868992, + "grad_norm": 0.38642457127571106, + "learning_rate": 4.430731982172886e-05, + "loss": 2.5186, + "step": 6789 + }, + { + "epoch": 0.86912, + "grad_norm": 0.45484527945518494, + "learning_rate": 4.422203926696061e-05, + "loss": 2.6419, + "step": 6790 + }, + { + "epoch": 0.869248, + "grad_norm": 0.5452561974525452, + "learning_rate": 4.4136837065707756e-05, + "loss": 2.5557, + "step": 6791 + }, + { + "epoch": 0.869376, + "grad_norm": 0.47702136635780334, + "learning_rate": 4.405171323261742e-05, + "loss": 2.4608, + "step": 6792 + }, + { + "epoch": 0.869504, + "grad_norm": 0.3964433968067169, + "learning_rate": 4.396666778232344e-05, + "loss": 2.5761, + "step": 6793 + }, + { + "epoch": 0.869632, + "grad_norm": 0.4446137249469757, + "learning_rate": 4.388170072944603e-05, + "loss": 2.5686, + "step": 6794 + }, + { + "epoch": 0.86976, + "grad_norm": 0.4878055453300476, + "learning_rate": 4.379681208859193e-05, + "loss": 2.5147, + "step": 6795 + }, + { + "epoch": 0.869888, + "grad_norm": 0.41036278009414673, + "learning_rate": 4.37120018743547e-05, + "loss": 2.6261, + "step": 6796 + }, + { + "epoch": 0.870016, + "grad_norm": 0.5276798605918884, + "learning_rate": 4.3627270101314064e-05, + "loss": 2.5094, + "step": 6797 + }, + { + "epoch": 0.870144, + "grad_norm": 0.4958159029483795, + "learning_rate": 4.3542616784036426e-05, + "loss": 2.5892, + "step": 6798 + }, + { + "epoch": 0.870272, + "grad_norm": 0.4209704101085663, + "learning_rate": 4.345804193707459e-05, + "loss": 2.5706, + "step": 6799 + }, + { + "epoch": 0.8704, + "grad_norm": 0.42813539505004883, + "learning_rate": 4.3373545574968144e-05, + "loss": 2.5799, + "step": 6800 + }, + { + "epoch": 0.870528, + "grad_norm": 0.38678833842277527, + "learning_rate": 4.328912771224286e-05, + "loss": 2.5148, + "step": 6801 + }, + { + "epoch": 0.870656, + "grad_norm": 0.43750107288360596, + "learning_rate": 4.320478836341113e-05, + "loss": 2.6134, + "step": 6802 + }, + { + "epoch": 0.870784, + "grad_norm": 0.40695276856422424, + "learning_rate": 4.312052754297197e-05, + "loss": 2.6151, + "step": 6803 + }, + { + "epoch": 0.870912, + "grad_norm": 0.47345998883247375, + "learning_rate": 4.303634526541067e-05, + "loss": 2.6117, + "step": 6804 + }, + { + "epoch": 0.87104, + "grad_norm": 0.4355403482913971, + "learning_rate": 4.2952241545199225e-05, + "loss": 2.5937, + "step": 6805 + }, + { + "epoch": 0.871168, + "grad_norm": 0.3892994523048401, + "learning_rate": 4.2868216396796055e-05, + "loss": 2.645, + "step": 6806 + }, + { + "epoch": 0.871296, + "grad_norm": 0.43915364146232605, + "learning_rate": 4.278426983464595e-05, + "loss": 2.5219, + "step": 6807 + }, + { + "epoch": 0.871424, + "grad_norm": 0.40870800614356995, + "learning_rate": 4.2700401873180303e-05, + "loss": 2.5702, + "step": 6808 + }, + { + "epoch": 0.871552, + "grad_norm": 0.37690913677215576, + "learning_rate": 4.261661252681714e-05, + "loss": 2.5441, + "step": 6809 + }, + { + "epoch": 0.87168, + "grad_norm": 0.40519505739212036, + "learning_rate": 4.253290180996067e-05, + "loss": 2.5685, + "step": 6810 + }, + { + "epoch": 0.871808, + "grad_norm": 0.41273388266563416, + "learning_rate": 4.2449269737001804e-05, + "loss": 2.5914, + "step": 6811 + }, + { + "epoch": 0.871936, + "grad_norm": 0.44314226508140564, + "learning_rate": 4.2365716322317894e-05, + "loss": 2.6713, + "step": 6812 + }, + { + "epoch": 0.872064, + "grad_norm": 0.4258072078227997, + "learning_rate": 4.228224158027272e-05, + "loss": 2.6334, + "step": 6813 + }, + { + "epoch": 0.872192, + "grad_norm": 0.39315855503082275, + "learning_rate": 4.2198845525216524e-05, + "loss": 2.5652, + "step": 6814 + }, + { + "epoch": 0.87232, + "grad_norm": 0.4112686514854431, + "learning_rate": 4.211552817148606e-05, + "loss": 2.5448, + "step": 6815 + }, + { + "epoch": 0.872448, + "grad_norm": 0.4186755418777466, + "learning_rate": 4.2032289533404434e-05, + "loss": 2.5814, + "step": 6816 + }, + { + "epoch": 0.872576, + "grad_norm": 0.42755600810050964, + "learning_rate": 4.194912962528158e-05, + "loss": 2.5392, + "step": 6817 + }, + { + "epoch": 0.872704, + "grad_norm": 0.4528592526912689, + "learning_rate": 4.186604846141345e-05, + "loss": 2.5463, + "step": 6818 + }, + { + "epoch": 0.872832, + "grad_norm": 0.400479793548584, + "learning_rate": 4.17830460560828e-05, + "loss": 2.6317, + "step": 6819 + }, + { + "epoch": 0.87296, + "grad_norm": 0.5182257890701294, + "learning_rate": 4.170012242355864e-05, + "loss": 2.5713, + "step": 6820 + }, + { + "epoch": 0.873088, + "grad_norm": 0.38835665583610535, + "learning_rate": 4.1617277578096526e-05, + "loss": 2.5091, + "step": 6821 + }, + { + "epoch": 0.873216, + "grad_norm": 0.4115679860115051, + "learning_rate": 4.153451153393839e-05, + "loss": 2.5978, + "step": 6822 + }, + { + "epoch": 0.873344, + "grad_norm": 0.39763835072517395, + "learning_rate": 4.1451824305312634e-05, + "loss": 2.5376, + "step": 6823 + }, + { + "epoch": 0.873472, + "grad_norm": 0.40352004766464233, + "learning_rate": 4.136921590643433e-05, + "loss": 2.6805, + "step": 6824 + }, + { + "epoch": 0.8736, + "grad_norm": 0.4316037893295288, + "learning_rate": 4.1286686351504735e-05, + "loss": 2.5924, + "step": 6825 + }, + { + "epoch": 0.873728, + "grad_norm": 0.3915325403213501, + "learning_rate": 4.120423565471154e-05, + "loss": 2.6231, + "step": 6826 + }, + { + "epoch": 0.873856, + "grad_norm": 0.38879960775375366, + "learning_rate": 4.1121863830229036e-05, + "loss": 2.6827, + "step": 6827 + }, + { + "epoch": 0.873984, + "grad_norm": 0.4510926604270935, + "learning_rate": 4.103957089221799e-05, + "loss": 2.5784, + "step": 6828 + }, + { + "epoch": 0.874112, + "grad_norm": 0.41842174530029297, + "learning_rate": 4.0957356854825435e-05, + "loss": 2.6253, + "step": 6829 + }, + { + "epoch": 0.87424, + "grad_norm": 0.4046889841556549, + "learning_rate": 4.0875221732184884e-05, + "loss": 2.6261, + "step": 6830 + }, + { + "epoch": 0.874368, + "grad_norm": 0.41355860233306885, + "learning_rate": 4.0793165538416275e-05, + "loss": 2.5848, + "step": 6831 + }, + { + "epoch": 0.874496, + "grad_norm": 0.4312601685523987, + "learning_rate": 4.0711188287626156e-05, + "loss": 2.5213, + "step": 6832 + }, + { + "epoch": 0.874624, + "grad_norm": 0.3909367322921753, + "learning_rate": 4.062928999390736e-05, + "loss": 2.5376, + "step": 6833 + }, + { + "epoch": 0.874752, + "grad_norm": 0.39087873697280884, + "learning_rate": 4.054747067133907e-05, + "loss": 2.6154, + "step": 6834 + }, + { + "epoch": 0.87488, + "grad_norm": 0.40541112422943115, + "learning_rate": 4.0465730333986985e-05, + "loss": 2.5295, + "step": 6835 + }, + { + "epoch": 0.875008, + "grad_norm": 2.0730929374694824, + "learning_rate": 4.038406899590313e-05, + "loss": 2.5756, + "step": 6836 + }, + { + "epoch": 0.875136, + "grad_norm": 0.4692312777042389, + "learning_rate": 4.030248667112629e-05, + "loss": 2.6064, + "step": 6837 + }, + { + "epoch": 0.875264, + "grad_norm": 0.420119047164917, + "learning_rate": 4.022098337368113e-05, + "loss": 2.6276, + "step": 6838 + }, + { + "epoch": 0.875392, + "grad_norm": 0.4298732578754425, + "learning_rate": 4.013955911757922e-05, + "loss": 2.5324, + "step": 6839 + }, + { + "epoch": 0.87552, + "grad_norm": 0.4344952702522278, + "learning_rate": 4.0058213916818295e-05, + "loss": 2.5098, + "step": 6840 + }, + { + "epoch": 0.875648, + "grad_norm": 0.4682067930698395, + "learning_rate": 3.997694778538247e-05, + "loss": 2.5617, + "step": 6841 + }, + { + "epoch": 0.875776, + "grad_norm": 0.46585604548454285, + "learning_rate": 3.989576073724238e-05, + "loss": 2.5654, + "step": 6842 + }, + { + "epoch": 0.875904, + "grad_norm": 0.421794593334198, + "learning_rate": 3.981465278635504e-05, + "loss": 2.5431, + "step": 6843 + }, + { + "epoch": 0.876032, + "grad_norm": 0.4392314553260803, + "learning_rate": 3.973362394666369e-05, + "loss": 2.5879, + "step": 6844 + }, + { + "epoch": 0.87616, + "grad_norm": 0.3697851002216339, + "learning_rate": 3.965267423209834e-05, + "loss": 2.6787, + "step": 6845 + }, + { + "epoch": 0.876288, + "grad_norm": 0.42557045817375183, + "learning_rate": 3.957180365657498e-05, + "loss": 2.6878, + "step": 6846 + }, + { + "epoch": 0.876416, + "grad_norm": 0.3642940819263458, + "learning_rate": 3.949101223399643e-05, + "loss": 2.5219, + "step": 6847 + }, + { + "epoch": 0.876544, + "grad_norm": 0.40211576223373413, + "learning_rate": 3.941029997825152e-05, + "loss": 2.5139, + "step": 6848 + }, + { + "epoch": 0.876672, + "grad_norm": 0.39950722455978394, + "learning_rate": 3.932966690321571e-05, + "loss": 2.5008, + "step": 6849 + }, + { + "epoch": 0.8768, + "grad_norm": 0.418538898229599, + "learning_rate": 3.9249113022750694e-05, + "loss": 2.54, + "step": 6850 + }, + { + "epoch": 0.876928, + "grad_norm": 0.4048285484313965, + "learning_rate": 3.9168638350704614e-05, + "loss": 2.531, + "step": 6851 + }, + { + "epoch": 0.877056, + "grad_norm": 0.4364299178123474, + "learning_rate": 3.908824290091195e-05, + "loss": 2.4672, + "step": 6852 + }, + { + "epoch": 0.877184, + "grad_norm": 0.38029173016548157, + "learning_rate": 3.900792668719372e-05, + "loss": 2.6084, + "step": 6853 + }, + { + "epoch": 0.877312, + "grad_norm": 0.3971855938434601, + "learning_rate": 3.892768972335725e-05, + "loss": 2.5613, + "step": 6854 + }, + { + "epoch": 0.87744, + "grad_norm": 0.3860521912574768, + "learning_rate": 3.884753202319596e-05, + "loss": 2.5102, + "step": 6855 + }, + { + "epoch": 0.877568, + "grad_norm": 0.3760021924972534, + "learning_rate": 3.8767453600490167e-05, + "loss": 2.5463, + "step": 6856 + }, + { + "epoch": 0.877696, + "grad_norm": 0.44875624775886536, + "learning_rate": 3.868745446900618e-05, + "loss": 2.687, + "step": 6857 + }, + { + "epoch": 0.877824, + "grad_norm": 0.42923399806022644, + "learning_rate": 3.8607534642496777e-05, + "loss": 2.5121, + "step": 6858 + }, + { + "epoch": 0.877952, + "grad_norm": 0.4482659697532654, + "learning_rate": 3.852769413470098e-05, + "loss": 2.6427, + "step": 6859 + }, + { + "epoch": 0.87808, + "grad_norm": 0.4599689245223999, + "learning_rate": 3.844793295934451e-05, + "loss": 2.6404, + "step": 6860 + }, + { + "epoch": 0.878208, + "grad_norm": 0.4565597474575043, + "learning_rate": 3.8368251130139086e-05, + "loss": 2.5182, + "step": 6861 + }, + { + "epoch": 0.878336, + "grad_norm": 0.42007848620414734, + "learning_rate": 3.828864866078302e-05, + "loss": 2.5427, + "step": 6862 + }, + { + "epoch": 0.878464, + "grad_norm": 0.6499045491218567, + "learning_rate": 3.820912556496087e-05, + "loss": 2.5509, + "step": 6863 + }, + { + "epoch": 0.878592, + "grad_norm": 0.4505326449871063, + "learning_rate": 3.812968185634352e-05, + "loss": 2.5758, + "step": 6864 + }, + { + "epoch": 0.87872, + "grad_norm": 0.4494789242744446, + "learning_rate": 3.805031754858834e-05, + "loss": 2.5187, + "step": 6865 + }, + { + "epoch": 0.878848, + "grad_norm": 0.4104771018028259, + "learning_rate": 3.7971032655339024e-05, + "loss": 2.5989, + "step": 6866 + }, + { + "epoch": 0.878976, + "grad_norm": 0.39748090505599976, + "learning_rate": 3.789182719022538e-05, + "loss": 2.5153, + "step": 6867 + }, + { + "epoch": 0.879104, + "grad_norm": 0.38362938165664673, + "learning_rate": 3.781270116686392e-05, + "loss": 2.4467, + "step": 6868 + }, + { + "epoch": 0.879232, + "grad_norm": 0.446426123380661, + "learning_rate": 3.77336545988573e-05, + "loss": 2.6225, + "step": 6869 + }, + { + "epoch": 0.87936, + "grad_norm": 0.42837122082710266, + "learning_rate": 3.765468749979445e-05, + "loss": 2.6308, + "step": 6870 + }, + { + "epoch": 0.879488, + "grad_norm": 0.3850862979888916, + "learning_rate": 3.757579988325083e-05, + "loss": 2.5445, + "step": 6871 + }, + { + "epoch": 0.879616, + "grad_norm": 0.4541570544242859, + "learning_rate": 3.749699176278809e-05, + "loss": 2.4993, + "step": 6872 + }, + { + "epoch": 0.879744, + "grad_norm": 0.3983314335346222, + "learning_rate": 3.741826315195418e-05, + "loss": 2.5434, + "step": 6873 + }, + { + "epoch": 0.879872, + "grad_norm": 0.37768784165382385, + "learning_rate": 3.733961406428354e-05, + "loss": 2.562, + "step": 6874 + }, + { + "epoch": 0.88, + "grad_norm": 0.38468313217163086, + "learning_rate": 3.7261044513296926e-05, + "loss": 2.4925, + "step": 6875 + }, + { + "epoch": 0.880128, + "grad_norm": 0.4006417691707611, + "learning_rate": 3.7182554512501296e-05, + "loss": 2.4722, + "step": 6876 + }, + { + "epoch": 0.880256, + "grad_norm": 0.5043922662734985, + "learning_rate": 3.710414407538998e-05, + "loss": 2.5828, + "step": 6877 + }, + { + "epoch": 0.880384, + "grad_norm": 1.031905174255371, + "learning_rate": 3.7025813215442684e-05, + "loss": 2.6225, + "step": 6878 + }, + { + "epoch": 0.880512, + "grad_norm": 0.4329602122306824, + "learning_rate": 3.694756194612531e-05, + "loss": 2.5133, + "step": 6879 + }, + { + "epoch": 0.88064, + "grad_norm": 0.3770529329776764, + "learning_rate": 3.686939028089015e-05, + "loss": 2.4859, + "step": 6880 + }, + { + "epoch": 0.880768, + "grad_norm": 0.41699764132499695, + "learning_rate": 3.679129823317595e-05, + "loss": 2.6282, + "step": 6881 + }, + { + "epoch": 0.880896, + "grad_norm": 0.5274009704589844, + "learning_rate": 3.6713285816407535e-05, + "loss": 2.5241, + "step": 6882 + }, + { + "epoch": 0.881024, + "grad_norm": 0.4716354310512543, + "learning_rate": 3.663535304399607e-05, + "loss": 2.619, + "step": 6883 + }, + { + "epoch": 0.881152, + "grad_norm": 0.4058501422405243, + "learning_rate": 3.655749992933932e-05, + "loss": 2.6874, + "step": 6884 + }, + { + "epoch": 0.88128, + "grad_norm": 0.4634402394294739, + "learning_rate": 3.647972648582104e-05, + "loss": 2.6232, + "step": 6885 + }, + { + "epoch": 0.881408, + "grad_norm": 0.38579291105270386, + "learning_rate": 3.640203272681131e-05, + "loss": 2.6466, + "step": 6886 + }, + { + "epoch": 0.881536, + "grad_norm": 0.39434754848480225, + "learning_rate": 3.632441866566672e-05, + "loss": 2.6067, + "step": 6887 + }, + { + "epoch": 0.881664, + "grad_norm": 0.45147454738616943, + "learning_rate": 3.624688431572981e-05, + "loss": 2.5019, + "step": 6888 + }, + { + "epoch": 0.881792, + "grad_norm": 0.4012867510318756, + "learning_rate": 3.6169429690329825e-05, + "loss": 2.5683, + "step": 6889 + }, + { + "epoch": 0.88192, + "grad_norm": 0.41718173027038574, + "learning_rate": 3.60920548027821e-05, + "loss": 2.6001, + "step": 6890 + }, + { + "epoch": 0.882048, + "grad_norm": 0.4959055185317993, + "learning_rate": 3.6014759666388284e-05, + "loss": 2.4734, + "step": 6891 + }, + { + "epoch": 0.882176, + "grad_norm": 0.46119117736816406, + "learning_rate": 3.593754429443613e-05, + "loss": 2.5934, + "step": 6892 + }, + { + "epoch": 0.882304, + "grad_norm": 0.4371633529663086, + "learning_rate": 3.586040870020008e-05, + "loss": 2.7163, + "step": 6893 + }, + { + "epoch": 0.882432, + "grad_norm": 0.44072234630584717, + "learning_rate": 3.578335289694051e-05, + "loss": 2.5944, + "step": 6894 + }, + { + "epoch": 0.88256, + "grad_norm": 0.5617063641548157, + "learning_rate": 3.570637689790418e-05, + "loss": 2.5562, + "step": 6895 + }, + { + "epoch": 0.882688, + "grad_norm": 0.4351447820663452, + "learning_rate": 3.562948071632427e-05, + "loss": 2.6313, + "step": 6896 + }, + { + "epoch": 0.882816, + "grad_norm": 0.4277151823043823, + "learning_rate": 3.5552664365420094e-05, + "loss": 2.6548, + "step": 6897 + }, + { + "epoch": 0.882944, + "grad_norm": 0.432465136051178, + "learning_rate": 3.547592785839726e-05, + "loss": 2.5651, + "step": 6898 + }, + { + "epoch": 0.883072, + "grad_norm": 0.414711058139801, + "learning_rate": 3.539927120844766e-05, + "loss": 2.6761, + "step": 6899 + }, + { + "epoch": 0.8832, + "grad_norm": 0.43013978004455566, + "learning_rate": 3.532269442874941e-05, + "loss": 2.6311, + "step": 6900 + }, + { + "epoch": 0.883328, + "grad_norm": 0.4673464596271515, + "learning_rate": 3.524619753246694e-05, + "loss": 2.5117, + "step": 6901 + }, + { + "epoch": 0.883456, + "grad_norm": 0.4189634919166565, + "learning_rate": 3.5169780532751105e-05, + "loss": 2.6359, + "step": 6902 + }, + { + "epoch": 0.883584, + "grad_norm": 0.425032377243042, + "learning_rate": 3.509344344273863e-05, + "loss": 2.7044, + "step": 6903 + }, + { + "epoch": 0.883712, + "grad_norm": 0.43681973218917847, + "learning_rate": 3.501718627555301e-05, + "loss": 2.6193, + "step": 6904 + }, + { + "epoch": 0.88384, + "grad_norm": 0.39850154519081116, + "learning_rate": 3.4941009044303585e-05, + "loss": 2.6615, + "step": 6905 + }, + { + "epoch": 0.883968, + "grad_norm": 0.39460912346839905, + "learning_rate": 3.486491176208617e-05, + "loss": 2.5914, + "step": 6906 + }, + { + "epoch": 0.884096, + "grad_norm": 0.43398532271385193, + "learning_rate": 3.4788894441982685e-05, + "loss": 2.7112, + "step": 6907 + }, + { + "epoch": 0.884224, + "grad_norm": 0.3895828127861023, + "learning_rate": 3.471295709706146e-05, + "loss": 2.5678, + "step": 6908 + }, + { + "epoch": 0.884352, + "grad_norm": 0.4364405870437622, + "learning_rate": 3.4637099740376935e-05, + "loss": 2.5849, + "step": 6909 + }, + { + "epoch": 0.88448, + "grad_norm": 0.42558732628822327, + "learning_rate": 3.456132238496984e-05, + "loss": 2.6129, + "step": 6910 + }, + { + "epoch": 0.884608, + "grad_norm": 0.4278978109359741, + "learning_rate": 3.448562504386738e-05, + "loss": 2.5409, + "step": 6911 + }, + { + "epoch": 0.884736, + "grad_norm": 0.4496183693408966, + "learning_rate": 3.4410007730082685e-05, + "loss": 2.5106, + "step": 6912 + }, + { + "epoch": 0.884864, + "grad_norm": 0.46993786096572876, + "learning_rate": 3.4334470456615254e-05, + "loss": 2.4799, + "step": 6913 + }, + { + "epoch": 0.884992, + "grad_norm": 0.4005064070224762, + "learning_rate": 3.425901323645075e-05, + "loss": 2.619, + "step": 6914 + }, + { + "epoch": 0.88512, + "grad_norm": 0.4034572243690491, + "learning_rate": 3.418363608256125e-05, + "loss": 2.6565, + "step": 6915 + }, + { + "epoch": 0.885248, + "grad_norm": 0.41250884532928467, + "learning_rate": 3.410833900790483e-05, + "loss": 2.618, + "step": 6916 + }, + { + "epoch": 0.885376, + "grad_norm": 0.4279601275920868, + "learning_rate": 3.403312202542608e-05, + "loss": 2.6278, + "step": 6917 + }, + { + "epoch": 0.885504, + "grad_norm": 0.43504786491394043, + "learning_rate": 3.395798514805565e-05, + "loss": 2.6376, + "step": 6918 + }, + { + "epoch": 0.885632, + "grad_norm": 0.4389624297618866, + "learning_rate": 3.3882928388710376e-05, + "loss": 2.6538, + "step": 6919 + }, + { + "epoch": 0.88576, + "grad_norm": 0.39755213260650635, + "learning_rate": 3.3807951760293275e-05, + "loss": 2.5931, + "step": 6920 + }, + { + "epoch": 0.885888, + "grad_norm": 0.49977728724479675, + "learning_rate": 3.373305527569398e-05, + "loss": 2.4941, + "step": 6921 + }, + { + "epoch": 0.886016, + "grad_norm": 0.44500967860221863, + "learning_rate": 3.3658238947787855e-05, + "loss": 2.6498, + "step": 6922 + }, + { + "epoch": 0.886144, + "grad_norm": 0.5136641263961792, + "learning_rate": 3.358350278943678e-05, + "loss": 2.5189, + "step": 6923 + }, + { + "epoch": 0.886272, + "grad_norm": 0.3639911115169525, + "learning_rate": 3.3508846813488646e-05, + "loss": 2.6781, + "step": 6924 + }, + { + "epoch": 0.8864, + "grad_norm": 0.4341471195220947, + "learning_rate": 3.343427103277785e-05, + "loss": 2.5852, + "step": 6925 + }, + { + "epoch": 0.886528, + "grad_norm": 0.45463085174560547, + "learning_rate": 3.335977546012481e-05, + "loss": 2.5194, + "step": 6926 + }, + { + "epoch": 0.886656, + "grad_norm": 0.4551713764667511, + "learning_rate": 3.328536010833605e-05, + "loss": 2.6301, + "step": 6927 + }, + { + "epoch": 0.886784, + "grad_norm": 0.4511566758155823, + "learning_rate": 3.3211024990204565e-05, + "loss": 2.5486, + "step": 6928 + }, + { + "epoch": 0.886912, + "grad_norm": 0.39343398809432983, + "learning_rate": 3.313677011850924e-05, + "loss": 2.5694, + "step": 6929 + }, + { + "epoch": 0.88704, + "grad_norm": 0.3876035511493683, + "learning_rate": 3.306259550601559e-05, + "loss": 2.5515, + "step": 6930 + }, + { + "epoch": 0.887168, + "grad_norm": 0.3861950635910034, + "learning_rate": 3.298850116547486e-05, + "loss": 2.549, + "step": 6931 + }, + { + "epoch": 0.887296, + "grad_norm": 0.4213889539241791, + "learning_rate": 3.291448710962497e-05, + "loss": 2.5368, + "step": 6932 + }, + { + "epoch": 0.887424, + "grad_norm": 0.43368643522262573, + "learning_rate": 3.284055335118963e-05, + "loss": 2.6862, + "step": 6933 + }, + { + "epoch": 0.887552, + "grad_norm": 0.41561615467071533, + "learning_rate": 3.276669990287895e-05, + "loss": 2.5925, + "step": 6934 + }, + { + "epoch": 0.88768, + "grad_norm": 0.4606480300426483, + "learning_rate": 3.269292677738922e-05, + "loss": 2.5398, + "step": 6935 + }, + { + "epoch": 0.887808, + "grad_norm": 0.41022780537605286, + "learning_rate": 3.261923398740285e-05, + "loss": 2.6397, + "step": 6936 + }, + { + "epoch": 0.887936, + "grad_norm": 0.4742579460144043, + "learning_rate": 3.2545621545588436e-05, + "loss": 2.5469, + "step": 6937 + }, + { + "epoch": 0.888064, + "grad_norm": 0.4164696931838989, + "learning_rate": 3.247208946460084e-05, + "loss": 2.5994, + "step": 6938 + }, + { + "epoch": 0.888192, + "grad_norm": 0.4113706946372986, + "learning_rate": 3.2398637757081185e-05, + "loss": 2.592, + "step": 6939 + }, + { + "epoch": 0.88832, + "grad_norm": 0.4039529860019684, + "learning_rate": 3.232526643565664e-05, + "loss": 2.6165, + "step": 6940 + }, + { + "epoch": 0.888448, + "grad_norm": 0.393559068441391, + "learning_rate": 3.2251975512940514e-05, + "loss": 2.5106, + "step": 6941 + }, + { + "epoch": 0.888576, + "grad_norm": 0.3986364006996155, + "learning_rate": 3.2178765001532386e-05, + "loss": 2.583, + "step": 6942 + }, + { + "epoch": 0.888704, + "grad_norm": 0.4024338722229004, + "learning_rate": 3.210563491401808e-05, + "loss": 2.5384, + "step": 6943 + }, + { + "epoch": 0.888832, + "grad_norm": 0.4403318166732788, + "learning_rate": 3.203258526296937e-05, + "loss": 2.5962, + "step": 6944 + }, + { + "epoch": 0.88896, + "grad_norm": 0.4163675308227539, + "learning_rate": 3.195961606094433e-05, + "loss": 2.4682, + "step": 6945 + }, + { + "epoch": 0.889088, + "grad_norm": 0.41123533248901367, + "learning_rate": 3.188672732048731e-05, + "loss": 2.5749, + "step": 6946 + }, + { + "epoch": 0.889216, + "grad_norm": 0.4098280668258667, + "learning_rate": 3.181391905412867e-05, + "loss": 2.5631, + "step": 6947 + }, + { + "epoch": 0.889344, + "grad_norm": 0.4212675094604492, + "learning_rate": 3.174119127438507e-05, + "loss": 2.5539, + "step": 6948 + }, + { + "epoch": 0.889472, + "grad_norm": 0.46314945816993713, + "learning_rate": 3.166854399375929e-05, + "loss": 2.4432, + "step": 6949 + }, + { + "epoch": 0.8896, + "grad_norm": 0.5121071338653564, + "learning_rate": 3.159597722474006e-05, + "loss": 2.5204, + "step": 6950 + }, + { + "epoch": 0.889728, + "grad_norm": 0.4045240581035614, + "learning_rate": 3.1523490979802614e-05, + "loss": 2.5908, + "step": 6951 + }, + { + "epoch": 0.889856, + "grad_norm": 0.3962552845478058, + "learning_rate": 3.145108527140805e-05, + "loss": 2.5044, + "step": 6952 + }, + { + "epoch": 0.889984, + "grad_norm": 0.5439667701721191, + "learning_rate": 3.1378760112003853e-05, + "loss": 2.6146, + "step": 6953 + }, + { + "epoch": 0.890112, + "grad_norm": 0.44630077481269836, + "learning_rate": 3.1306515514023527e-05, + "loss": 2.6716, + "step": 6954 + }, + { + "epoch": 0.89024, + "grad_norm": 0.44963333010673523, + "learning_rate": 3.1234351489886736e-05, + "loss": 2.6237, + "step": 6955 + }, + { + "epoch": 0.890368, + "grad_norm": 0.3818088471889496, + "learning_rate": 3.116226805199929e-05, + "loss": 2.5935, + "step": 6956 + }, + { + "epoch": 0.890496, + "grad_norm": 0.42514657974243164, + "learning_rate": 3.109026521275321e-05, + "loss": 2.6262, + "step": 6957 + }, + { + "epoch": 0.890624, + "grad_norm": 0.4332951605319977, + "learning_rate": 3.1018342984526604e-05, + "loss": 2.5394, + "step": 6958 + }, + { + "epoch": 0.890752, + "grad_norm": 0.5330499410629272, + "learning_rate": 3.0946501379683686e-05, + "loss": 2.6195, + "step": 6959 + }, + { + "epoch": 0.89088, + "grad_norm": 0.40308520197868347, + "learning_rate": 3.087474041057503e-05, + "loss": 2.6254, + "step": 6960 + }, + { + "epoch": 0.891008, + "grad_norm": 0.42062023282051086, + "learning_rate": 3.080306008953704e-05, + "loss": 2.4853, + "step": 6961 + }, + { + "epoch": 0.891136, + "grad_norm": 0.4929019808769226, + "learning_rate": 3.073146042889241e-05, + "loss": 2.5162, + "step": 6962 + }, + { + "epoch": 0.891264, + "grad_norm": 0.43891003727912903, + "learning_rate": 3.065994144094997e-05, + "loss": 2.6535, + "step": 6963 + }, + { + "epoch": 0.891392, + "grad_norm": 0.43548834323883057, + "learning_rate": 3.05885031380046e-05, + "loss": 2.6058, + "step": 6964 + }, + { + "epoch": 0.89152, + "grad_norm": 0.4155910909175873, + "learning_rate": 3.0517145532337476e-05, + "loss": 2.6659, + "step": 6965 + }, + { + "epoch": 0.891648, + "grad_norm": 0.4263012111186981, + "learning_rate": 3.0445868636215678e-05, + "loss": 2.6848, + "step": 6966 + }, + { + "epoch": 0.891776, + "grad_norm": 0.4191797077655792, + "learning_rate": 3.037467246189257e-05, + "loss": 2.5931, + "step": 6967 + }, + { + "epoch": 0.891904, + "grad_norm": 0.4403403401374817, + "learning_rate": 3.0303557021607754e-05, + "loss": 2.5764, + "step": 6968 + }, + { + "epoch": 0.892032, + "grad_norm": 0.4062803387641907, + "learning_rate": 3.0232522327586664e-05, + "loss": 2.4545, + "step": 6969 + }, + { + "epoch": 0.89216, + "grad_norm": 0.3817160129547119, + "learning_rate": 3.0161568392040982e-05, + "loss": 2.5314, + "step": 6970 + }, + { + "epoch": 0.892288, + "grad_norm": 0.4245813190937042, + "learning_rate": 3.0090695227168507e-05, + "loss": 2.6592, + "step": 6971 + }, + { + "epoch": 0.892416, + "grad_norm": 0.40507131814956665, + "learning_rate": 3.0019902845153214e-05, + "loss": 2.569, + "step": 6972 + }, + { + "epoch": 0.892544, + "grad_norm": 0.48020800948143005, + "learning_rate": 2.9949191258164976e-05, + "loss": 2.5769, + "step": 6973 + }, + { + "epoch": 0.892672, + "grad_norm": 0.4422432780265808, + "learning_rate": 2.9878560478360183e-05, + "loss": 2.5978, + "step": 6974 + }, + { + "epoch": 0.8928, + "grad_norm": 0.4369790256023407, + "learning_rate": 2.9808010517880847e-05, + "loss": 2.6306, + "step": 6975 + }, + { + "epoch": 0.892928, + "grad_norm": 0.4148487448692322, + "learning_rate": 2.9737541388855536e-05, + "loss": 2.5591, + "step": 6976 + }, + { + "epoch": 0.893056, + "grad_norm": 0.3695615231990814, + "learning_rate": 2.966715310339857e-05, + "loss": 2.5925, + "step": 6977 + }, + { + "epoch": 0.893184, + "grad_norm": 0.40112268924713135, + "learning_rate": 2.9596845673610596e-05, + "loss": 2.6326, + "step": 6978 + }, + { + "epoch": 0.893312, + "grad_norm": 0.44328999519348145, + "learning_rate": 2.9526619111578222e-05, + "loss": 2.5604, + "step": 6979 + }, + { + "epoch": 0.89344, + "grad_norm": 0.41346532106399536, + "learning_rate": 2.945647342937413e-05, + "loss": 2.7343, + "step": 6980 + }, + { + "epoch": 0.893568, + "grad_norm": 0.4050333499908447, + "learning_rate": 2.9386408639057392e-05, + "loss": 2.4633, + "step": 6981 + }, + { + "epoch": 0.893696, + "grad_norm": 0.3942220211029053, + "learning_rate": 2.9316424752672767e-05, + "loss": 2.5492, + "step": 6982 + }, + { + "epoch": 0.893824, + "grad_norm": 0.41197559237480164, + "learning_rate": 2.92465217822514e-05, + "loss": 2.6771, + "step": 6983 + }, + { + "epoch": 0.893952, + "grad_norm": 0.41432061791419983, + "learning_rate": 2.9176699739810298e-05, + "loss": 2.6308, + "step": 6984 + }, + { + "epoch": 0.89408, + "grad_norm": 0.4048239290714264, + "learning_rate": 2.9106958637352797e-05, + "loss": 2.638, + "step": 6985 + }, + { + "epoch": 0.894208, + "grad_norm": 0.45177897810935974, + "learning_rate": 2.90372984868682e-05, + "loss": 2.6498, + "step": 6986 + }, + { + "epoch": 0.894336, + "grad_norm": 0.4531558156013489, + "learning_rate": 2.8967719300331875e-05, + "loss": 2.6631, + "step": 6987 + }, + { + "epoch": 0.894464, + "grad_norm": 0.39852118492126465, + "learning_rate": 2.889822108970519e-05, + "loss": 2.5749, + "step": 6988 + }, + { + "epoch": 0.894592, + "grad_norm": 0.4419059157371521, + "learning_rate": 2.8828803866935815e-05, + "loss": 2.6576, + "step": 6989 + }, + { + "epoch": 0.89472, + "grad_norm": 0.4697657525539398, + "learning_rate": 2.8759467643957372e-05, + "loss": 2.6546, + "step": 6990 + }, + { + "epoch": 0.894848, + "grad_norm": 0.3899262547492981, + "learning_rate": 2.8690212432689544e-05, + "loss": 2.5331, + "step": 6991 + }, + { + "epoch": 0.894976, + "grad_norm": 0.4631907343864441, + "learning_rate": 2.862103824503809e-05, + "loss": 2.6066, + "step": 6992 + }, + { + "epoch": 0.895104, + "grad_norm": 0.4432423412799835, + "learning_rate": 2.8551945092894825e-05, + "loss": 2.6339, + "step": 6993 + }, + { + "epoch": 0.895232, + "grad_norm": 0.39350375533103943, + "learning_rate": 2.8482932988137644e-05, + "loss": 2.4799, + "step": 6994 + }, + { + "epoch": 0.89536, + "grad_norm": 0.5604450106620789, + "learning_rate": 2.841400194263055e-05, + "loss": 2.5632, + "step": 6995 + }, + { + "epoch": 0.895488, + "grad_norm": 0.4648224413394928, + "learning_rate": 2.8345151968223738e-05, + "loss": 2.6101, + "step": 6996 + }, + { + "epoch": 0.895616, + "grad_norm": 0.4327242076396942, + "learning_rate": 2.8276383076753175e-05, + "loss": 2.6116, + "step": 6997 + }, + { + "epoch": 0.895744, + "grad_norm": 0.4102991223335266, + "learning_rate": 2.8207695280041023e-05, + "loss": 2.4971, + "step": 6998 + }, + { + "epoch": 0.895872, + "grad_norm": 0.43903836607933044, + "learning_rate": 2.8139088589895557e-05, + "loss": 2.5068, + "step": 6999 + }, + { + "epoch": 0.896, + "grad_norm": 0.47437161207199097, + "learning_rate": 2.8070563018111063e-05, + "loss": 2.5825, + "step": 7000 + }, + { + "epoch": 0.896128, + "grad_norm": 0.4623834788799286, + "learning_rate": 2.8002118576467785e-05, + "loss": 2.5673, + "step": 7001 + }, + { + "epoch": 0.896256, + "grad_norm": 0.46222320199012756, + "learning_rate": 2.7933755276732254e-05, + "loss": 2.6937, + "step": 7002 + }, + { + "epoch": 0.896384, + "grad_norm": 0.40356069803237915, + "learning_rate": 2.7865473130656793e-05, + "loss": 2.499, + "step": 7003 + }, + { + "epoch": 0.896512, + "grad_norm": 0.4505128562450409, + "learning_rate": 2.779727214997996e-05, + "loss": 2.6659, + "step": 7004 + }, + { + "epoch": 0.89664, + "grad_norm": 0.4731083810329437, + "learning_rate": 2.7729152346426366e-05, + "loss": 2.5795, + "step": 7005 + }, + { + "epoch": 0.896768, + "grad_norm": 0.3978686034679413, + "learning_rate": 2.7661113731706488e-05, + "loss": 2.5658, + "step": 7006 + }, + { + "epoch": 0.896896, + "grad_norm": 0.38640427589416504, + "learning_rate": 2.7593156317516966e-05, + "loss": 2.5242, + "step": 7007 + }, + { + "epoch": 0.897024, + "grad_norm": 0.40536025166511536, + "learning_rate": 2.7525280115540506e-05, + "loss": 2.6133, + "step": 7008 + }, + { + "epoch": 0.897152, + "grad_norm": 0.37359723448753357, + "learning_rate": 2.7457485137445727e-05, + "loss": 2.5006, + "step": 7009 + }, + { + "epoch": 0.89728, + "grad_norm": 0.4056099057197571, + "learning_rate": 2.738977139488752e-05, + "loss": 2.5246, + "step": 7010 + }, + { + "epoch": 0.897408, + "grad_norm": 0.37766510248184204, + "learning_rate": 2.732213889950652e-05, + "loss": 2.5944, + "step": 7011 + }, + { + "epoch": 0.897536, + "grad_norm": 0.3976452350616455, + "learning_rate": 2.7254587662929543e-05, + "loss": 2.5921, + "step": 7012 + }, + { + "epoch": 0.897664, + "grad_norm": 0.4169355630874634, + "learning_rate": 2.718711769676957e-05, + "loss": 2.6083, + "step": 7013 + }, + { + "epoch": 0.897792, + "grad_norm": 0.37271827459335327, + "learning_rate": 2.7119729012625382e-05, + "loss": 2.4439, + "step": 7014 + }, + { + "epoch": 0.89792, + "grad_norm": 0.4326093792915344, + "learning_rate": 2.705242162208188e-05, + "loss": 2.5462, + "step": 7015 + }, + { + "epoch": 0.898048, + "grad_norm": 0.38507676124572754, + "learning_rate": 2.6985195536709918e-05, + "loss": 2.6518, + "step": 7016 + }, + { + "epoch": 0.898176, + "grad_norm": 0.41357406973838806, + "learning_rate": 2.6918050768066528e-05, + "loss": 2.4643, + "step": 7017 + }, + { + "epoch": 0.898304, + "grad_norm": 0.3871475160121918, + "learning_rate": 2.6850987327694698e-05, + "loss": 2.4413, + "step": 7018 + }, + { + "epoch": 0.898432, + "grad_norm": 0.4288724660873413, + "learning_rate": 2.678400522712332e-05, + "loss": 2.5578, + "step": 7019 + }, + { + "epoch": 0.89856, + "grad_norm": 0.40597841143608093, + "learning_rate": 2.671710447786746e-05, + "loss": 2.545, + "step": 7020 + }, + { + "epoch": 0.898688, + "grad_norm": 0.44435402750968933, + "learning_rate": 2.665028509142803e-05, + "loss": 2.5465, + "step": 7021 + }, + { + "epoch": 0.898816, + "grad_norm": 0.3971388638019562, + "learning_rate": 2.6583547079292224e-05, + "loss": 2.5015, + "step": 7022 + }, + { + "epoch": 0.898944, + "grad_norm": 0.4287407100200653, + "learning_rate": 2.6516890452932928e-05, + "loss": 2.4851, + "step": 7023 + }, + { + "epoch": 0.899072, + "grad_norm": 0.47564467787742615, + "learning_rate": 2.645031522380925e-05, + "loss": 2.6806, + "step": 7024 + }, + { + "epoch": 0.8992, + "grad_norm": 0.4320276081562042, + "learning_rate": 2.6383821403366258e-05, + "loss": 2.6372, + "step": 7025 + }, + { + "epoch": 0.899328, + "grad_norm": 0.4384422302246094, + "learning_rate": 2.6317409003035032e-05, + "loss": 2.58, + "step": 7026 + }, + { + "epoch": 0.899456, + "grad_norm": 0.43857282400131226, + "learning_rate": 2.62510780342326e-05, + "loss": 2.5977, + "step": 7027 + }, + { + "epoch": 0.899584, + "grad_norm": 0.4031519591808319, + "learning_rate": 2.6184828508362013e-05, + "loss": 2.5786, + "step": 7028 + }, + { + "epoch": 0.899712, + "grad_norm": 0.40919992327690125, + "learning_rate": 2.6118660436812324e-05, + "loss": 2.576, + "step": 7029 + }, + { + "epoch": 0.89984, + "grad_norm": 0.3913564085960388, + "learning_rate": 2.6052573830958547e-05, + "loss": 2.6613, + "step": 7030 + }, + { + "epoch": 0.899968, + "grad_norm": 0.4475085139274597, + "learning_rate": 2.5986568702161818e-05, + "loss": 2.561, + "step": 7031 + }, + { + "epoch": 0.900096, + "grad_norm": 0.3969626724720001, + "learning_rate": 2.592064506176922e-05, + "loss": 2.6083, + "step": 7032 + }, + { + "epoch": 0.900224, + "grad_norm": 0.38928210735321045, + "learning_rate": 2.585480292111375e-05, + "loss": 2.5589, + "step": 7033 + }, + { + "epoch": 0.900352, + "grad_norm": 0.4310283660888672, + "learning_rate": 2.578904229151441e-05, + "loss": 2.6002, + "step": 7034 + }, + { + "epoch": 0.90048, + "grad_norm": 0.40630021691322327, + "learning_rate": 2.5723363184276206e-05, + "loss": 2.4792, + "step": 7035 + }, + { + "epoch": 0.900608, + "grad_norm": 0.3960680365562439, + "learning_rate": 2.565776561069022e-05, + "loss": 2.6095, + "step": 7036 + }, + { + "epoch": 0.900736, + "grad_norm": 0.3864751160144806, + "learning_rate": 2.559224958203321e-05, + "loss": 2.4936, + "step": 7037 + }, + { + "epoch": 0.900864, + "grad_norm": 0.42540043592453003, + "learning_rate": 2.5526815109568447e-05, + "loss": 2.5525, + "step": 7038 + }, + { + "epoch": 0.900992, + "grad_norm": 0.42042988538742065, + "learning_rate": 2.5461462204544704e-05, + "loss": 2.6169, + "step": 7039 + }, + { + "epoch": 0.90112, + "grad_norm": 0.4371265172958374, + "learning_rate": 2.539619087819689e-05, + "loss": 2.5275, + "step": 7040 + }, + { + "epoch": 0.901248, + "grad_norm": 0.45207351446151733, + "learning_rate": 2.5331001141745968e-05, + "loss": 2.5513, + "step": 7041 + }, + { + "epoch": 0.901376, + "grad_norm": 0.4150327742099762, + "learning_rate": 2.526589300639881e-05, + "loss": 2.632, + "step": 7042 + }, + { + "epoch": 0.901504, + "grad_norm": 0.44317862391471863, + "learning_rate": 2.520086648334824e-05, + "loss": 2.5923, + "step": 7043 + }, + { + "epoch": 0.901632, + "grad_norm": 0.47187453508377075, + "learning_rate": 2.5135921583773036e-05, + "loss": 2.6819, + "step": 7044 + }, + { + "epoch": 0.90176, + "grad_norm": 0.420290470123291, + "learning_rate": 2.5071058318837936e-05, + "loss": 2.609, + "step": 7045 + }, + { + "epoch": 0.901888, + "grad_norm": 0.42972370982170105, + "learning_rate": 2.5006276699693852e-05, + "loss": 2.6302, + "step": 7046 + }, + { + "epoch": 0.902016, + "grad_norm": 0.4836251437664032, + "learning_rate": 2.4941576737477434e-05, + "loss": 2.4503, + "step": 7047 + }, + { + "epoch": 0.902144, + "grad_norm": 0.4245152473449707, + "learning_rate": 2.4876958443311282e-05, + "loss": 2.5877, + "step": 7048 + }, + { + "epoch": 0.902272, + "grad_norm": 0.4373956620693207, + "learning_rate": 2.481242182830401e-05, + "loss": 2.42, + "step": 7049 + }, + { + "epoch": 0.9024, + "grad_norm": 0.38731127977371216, + "learning_rate": 2.4747966903550358e-05, + "loss": 2.4722, + "step": 7050 + }, + { + "epoch": 0.902528, + "grad_norm": 0.48443251848220825, + "learning_rate": 2.4683593680130732e-05, + "loss": 2.5488, + "step": 7051 + }, + { + "epoch": 0.902656, + "grad_norm": 0.522452175617218, + "learning_rate": 2.461930216911168e-05, + "loss": 2.5868, + "step": 7052 + }, + { + "epoch": 0.902784, + "grad_norm": 0.40974804759025574, + "learning_rate": 2.455509238154574e-05, + "loss": 2.6298, + "step": 7053 + }, + { + "epoch": 0.902912, + "grad_norm": 0.41881483793258667, + "learning_rate": 2.4490964328471254e-05, + "loss": 2.5801, + "step": 7054 + }, + { + "epoch": 0.90304, + "grad_norm": 0.40929001569747925, + "learning_rate": 2.442691802091257e-05, + "loss": 2.4817, + "step": 7055 + }, + { + "epoch": 0.903168, + "grad_norm": 0.4251731038093567, + "learning_rate": 2.4362953469879935e-05, + "loss": 2.5901, + "step": 7056 + }, + { + "epoch": 0.903296, + "grad_norm": 0.4689820110797882, + "learning_rate": 2.4299070686369717e-05, + "loss": 2.5662, + "step": 7057 + }, + { + "epoch": 0.903424, + "grad_norm": 0.41066092252731323, + "learning_rate": 2.423526968136397e-05, + "loss": 2.51, + "step": 7058 + }, + { + "epoch": 0.903552, + "grad_norm": 0.408984512090683, + "learning_rate": 2.417155046583097e-05, + "loss": 2.6624, + "step": 7059 + }, + { + "epoch": 0.90368, + "grad_norm": 0.4086671769618988, + "learning_rate": 2.4107913050724628e-05, + "loss": 2.6384, + "step": 7060 + }, + { + "epoch": 0.903808, + "grad_norm": 0.40890640020370483, + "learning_rate": 2.4044357446985133e-05, + "loss": 2.6484, + "step": 7061 + }, + { + "epoch": 0.903936, + "grad_norm": 0.39242711663246155, + "learning_rate": 2.3980883665538357e-05, + "loss": 2.6213, + "step": 7062 + }, + { + "epoch": 0.904064, + "grad_norm": 0.4147701561450958, + "learning_rate": 2.391749171729618e-05, + "loss": 2.5467, + "step": 7063 + }, + { + "epoch": 0.904192, + "grad_norm": 0.4058387875556946, + "learning_rate": 2.385418161315639e-05, + "loss": 2.5826, + "step": 7064 + }, + { + "epoch": 0.90432, + "grad_norm": 0.487770140171051, + "learning_rate": 2.379095336400272e-05, + "loss": 2.5455, + "step": 7065 + }, + { + "epoch": 0.904448, + "grad_norm": 0.45004674792289734, + "learning_rate": 2.3727806980704758e-05, + "loss": 2.6839, + "step": 7066 + }, + { + "epoch": 0.904576, + "grad_norm": 0.4449792206287384, + "learning_rate": 2.3664742474118316e-05, + "loss": 2.5592, + "step": 7067 + }, + { + "epoch": 0.904704, + "grad_norm": 0.40458351373672485, + "learning_rate": 2.360175985508467e-05, + "loss": 2.5635, + "step": 7068 + }, + { + "epoch": 0.904832, + "grad_norm": 0.3948645293712616, + "learning_rate": 2.3538859134431546e-05, + "loss": 2.4729, + "step": 7069 + }, + { + "epoch": 0.90496, + "grad_norm": 0.40291860699653625, + "learning_rate": 2.347604032297207e-05, + "loss": 2.4896, + "step": 7070 + }, + { + "epoch": 0.905088, + "grad_norm": 0.4307396709918976, + "learning_rate": 2.3413303431505607e-05, + "loss": 2.5632, + "step": 7071 + }, + { + "epoch": 0.905216, + "grad_norm": 0.4040849208831787, + "learning_rate": 2.3350648470817413e-05, + "loss": 2.6222, + "step": 7072 + }, + { + "epoch": 0.905344, + "grad_norm": 0.5466479063034058, + "learning_rate": 2.328807545167838e-05, + "loss": 2.6064, + "step": 7073 + }, + { + "epoch": 0.905472, + "grad_norm": 0.37593433260917664, + "learning_rate": 2.3225584384845845e-05, + "loss": 2.5983, + "step": 7074 + }, + { + "epoch": 0.9056, + "grad_norm": 0.38275042176246643, + "learning_rate": 2.3163175281062544e-05, + "loss": 2.5539, + "step": 7075 + }, + { + "epoch": 0.905728, + "grad_norm": 0.46024343371391296, + "learning_rate": 2.3100848151057395e-05, + "loss": 2.553, + "step": 7076 + }, + { + "epoch": 0.905856, + "grad_norm": 0.41526082158088684, + "learning_rate": 2.3038603005545113e-05, + "loss": 2.6169, + "step": 7077 + }, + { + "epoch": 0.905984, + "grad_norm": 0.5105660557746887, + "learning_rate": 2.2976439855226404e-05, + "loss": 2.5512, + "step": 7078 + }, + { + "epoch": 0.906112, + "grad_norm": 0.4137076139450073, + "learning_rate": 2.291435871078784e-05, + "loss": 2.6581, + "step": 7079 + }, + { + "epoch": 0.90624, + "grad_norm": 0.40044474601745605, + "learning_rate": 2.2852359582901826e-05, + "loss": 2.6041, + "step": 7080 + }, + { + "epoch": 0.906368, + "grad_norm": 0.4602932631969452, + "learning_rate": 2.2790442482226725e-05, + "loss": 2.553, + "step": 7081 + }, + { + "epoch": 0.906496, + "grad_norm": 0.4525594413280487, + "learning_rate": 2.2728607419406965e-05, + "loss": 2.6351, + "step": 7082 + }, + { + "epoch": 0.906624, + "grad_norm": 0.45553579926490784, + "learning_rate": 2.2666854405072547e-05, + "loss": 2.5106, + "step": 7083 + }, + { + "epoch": 0.906752, + "grad_norm": 0.40650489926338196, + "learning_rate": 2.2605183449839583e-05, + "loss": 2.6401, + "step": 7084 + }, + { + "epoch": 0.90688, + "grad_norm": 0.44661247730255127, + "learning_rate": 2.2543594564309988e-05, + "loss": 2.5817, + "step": 7085 + }, + { + "epoch": 0.907008, + "grad_norm": 0.5607438087463379, + "learning_rate": 2.2482087759071622e-05, + "loss": 2.467, + "step": 7086 + }, + { + "epoch": 0.907136, + "grad_norm": 0.5245310664176941, + "learning_rate": 2.2420663044698253e-05, + "loss": 2.5679, + "step": 7087 + }, + { + "epoch": 0.907264, + "grad_norm": 0.40726903080940247, + "learning_rate": 2.235932043174943e-05, + "loss": 2.5346, + "step": 7088 + }, + { + "epoch": 0.907392, + "grad_norm": 0.45651647448539734, + "learning_rate": 2.229805993077083e-05, + "loss": 2.49, + "step": 7089 + }, + { + "epoch": 0.90752, + "grad_norm": 0.412074476480484, + "learning_rate": 2.2236881552293642e-05, + "loss": 2.499, + "step": 7090 + }, + { + "epoch": 0.907648, + "grad_norm": 0.43797221779823303, + "learning_rate": 2.2175785306835284e-05, + "loss": 2.5524, + "step": 7091 + }, + { + "epoch": 0.907776, + "grad_norm": 0.4176832139492035, + "learning_rate": 2.21147712048988e-05, + "loss": 2.5078, + "step": 7092 + }, + { + "epoch": 0.907904, + "grad_norm": 0.42163145542144775, + "learning_rate": 2.2053839256973295e-05, + "loss": 2.576, + "step": 7093 + }, + { + "epoch": 0.908032, + "grad_norm": 0.44483426213264465, + "learning_rate": 2.1992989473533563e-05, + "loss": 2.5567, + "step": 7094 + }, + { + "epoch": 0.90816, + "grad_norm": 0.3907797336578369, + "learning_rate": 2.193222186504057e-05, + "loss": 2.4901, + "step": 7095 + }, + { + "epoch": 0.908288, + "grad_norm": 0.42946600914001465, + "learning_rate": 2.1871536441940786e-05, + "loss": 2.5221, + "step": 7096 + }, + { + "epoch": 0.908416, + "grad_norm": 0.42940109968185425, + "learning_rate": 2.181093321466693e-05, + "loss": 2.5678, + "step": 7097 + }, + { + "epoch": 0.908544, + "grad_norm": 0.40373146533966064, + "learning_rate": 2.1750412193637214e-05, + "loss": 2.4792, + "step": 7098 + }, + { + "epoch": 0.908672, + "grad_norm": 0.47745585441589355, + "learning_rate": 2.1689973389256044e-05, + "loss": 2.6127, + "step": 7099 + }, + { + "epoch": 0.9088, + "grad_norm": 0.38897281885147095, + "learning_rate": 2.16296168119135e-05, + "loss": 2.5895, + "step": 7100 + }, + { + "epoch": 0.908928, + "grad_norm": 0.41830745339393616, + "learning_rate": 2.1569342471985554e-05, + "loss": 2.5983, + "step": 7101 + }, + { + "epoch": 0.909056, + "grad_norm": 0.5524827241897583, + "learning_rate": 2.150915037983403e-05, + "loss": 2.5521, + "step": 7102 + }, + { + "epoch": 0.909184, + "grad_norm": 0.4808143079280853, + "learning_rate": 2.1449040545806764e-05, + "loss": 2.6119, + "step": 7103 + }, + { + "epoch": 0.909312, + "grad_norm": 0.5137273669242859, + "learning_rate": 2.1389012980237265e-05, + "loss": 2.5091, + "step": 7104 + }, + { + "epoch": 0.90944, + "grad_norm": 0.4308184087276459, + "learning_rate": 2.1329067693444893e-05, + "loss": 2.5465, + "step": 7105 + }, + { + "epoch": 0.909568, + "grad_norm": 0.44478270411491394, + "learning_rate": 2.1269204695735066e-05, + "loss": 2.5283, + "step": 7106 + }, + { + "epoch": 0.909696, + "grad_norm": 0.4838577210903168, + "learning_rate": 2.120942399739889e-05, + "loss": 2.4755, + "step": 7107 + }, + { + "epoch": 0.909824, + "grad_norm": 0.44317829608917236, + "learning_rate": 2.1149725608713367e-05, + "loss": 2.6383, + "step": 7108 + }, + { + "epoch": 0.909952, + "grad_norm": 0.3964625895023346, + "learning_rate": 2.1090109539941228e-05, + "loss": 2.7199, + "step": 7109 + }, + { + "epoch": 0.91008, + "grad_norm": 0.42857441306114197, + "learning_rate": 2.1030575801331332e-05, + "loss": 2.5063, + "step": 7110 + }, + { + "epoch": 0.910208, + "grad_norm": 0.4047713279724121, + "learning_rate": 2.0971124403118157e-05, + "loss": 2.6018, + "step": 7111 + }, + { + "epoch": 0.910336, + "grad_norm": 0.3970494270324707, + "learning_rate": 2.091175535552209e-05, + "loss": 2.5615, + "step": 7112 + }, + { + "epoch": 0.910464, + "grad_norm": 0.5889071226119995, + "learning_rate": 2.0852468668749293e-05, + "loss": 2.5389, + "step": 7113 + }, + { + "epoch": 0.910592, + "grad_norm": 0.5422903895378113, + "learning_rate": 2.079326435299189e-05, + "loss": 2.5418, + "step": 7114 + }, + { + "epoch": 0.91072, + "grad_norm": 0.4924256503582001, + "learning_rate": 2.0734142418427805e-05, + "loss": 2.6808, + "step": 7115 + }, + { + "epoch": 0.910848, + "grad_norm": 0.38007405400276184, + "learning_rate": 2.067510287522073e-05, + "loss": 2.6034, + "step": 7116 + }, + { + "epoch": 0.910976, + "grad_norm": 0.44840019941329956, + "learning_rate": 2.0616145733520274e-05, + "loss": 2.5391, + "step": 7117 + }, + { + "epoch": 0.911104, + "grad_norm": 0.5027469396591187, + "learning_rate": 2.055727100346194e-05, + "loss": 2.6359, + "step": 7118 + }, + { + "epoch": 0.911232, + "grad_norm": 0.43941593170166016, + "learning_rate": 2.0498478695166855e-05, + "loss": 2.5624, + "step": 7119 + }, + { + "epoch": 0.91136, + "grad_norm": 0.4545767307281494, + "learning_rate": 2.0439768818742155e-05, + "loss": 2.5565, + "step": 7120 + }, + { + "epoch": 0.911488, + "grad_norm": 0.44169387221336365, + "learning_rate": 2.038114138428071e-05, + "loss": 2.5591, + "step": 7121 + }, + { + "epoch": 0.911616, + "grad_norm": 0.39177387952804565, + "learning_rate": 2.0322596401861294e-05, + "loss": 2.6446, + "step": 7122 + }, + { + "epoch": 0.911744, + "grad_norm": 0.3678559958934784, + "learning_rate": 2.0264133881548406e-05, + "loss": 2.5696, + "step": 7123 + }, + { + "epoch": 0.911872, + "grad_norm": 0.4656907320022583, + "learning_rate": 2.020575383339246e-05, + "loss": 2.5701, + "step": 7124 + }, + { + "epoch": 0.912, + "grad_norm": 0.4271152913570404, + "learning_rate": 2.0147456267429755e-05, + "loss": 2.5388, + "step": 7125 + }, + { + "epoch": 0.912128, + "grad_norm": 0.37745705246925354, + "learning_rate": 2.0089241193682272e-05, + "loss": 2.4744, + "step": 7126 + }, + { + "epoch": 0.912256, + "grad_norm": 0.4124067723751068, + "learning_rate": 2.003110862215779e-05, + "loss": 2.6793, + "step": 7127 + }, + { + "epoch": 0.912384, + "grad_norm": 0.39795780181884766, + "learning_rate": 1.997305856285003e-05, + "loss": 2.6263, + "step": 7128 + }, + { + "epoch": 0.912512, + "grad_norm": 0.4625019133090973, + "learning_rate": 1.9915091025738462e-05, + "loss": 2.5613, + "step": 7129 + }, + { + "epoch": 0.91264, + "grad_norm": 0.4147728383541107, + "learning_rate": 1.9857206020788276e-05, + "loss": 2.5434, + "step": 7130 + }, + { + "epoch": 0.912768, + "grad_norm": 0.3919908106327057, + "learning_rate": 1.979940355795079e-05, + "loss": 2.4624, + "step": 7131 + }, + { + "epoch": 0.912896, + "grad_norm": 0.5011029243469238, + "learning_rate": 1.9741683647162724e-05, + "loss": 2.4958, + "step": 7132 + }, + { + "epoch": 0.913024, + "grad_norm": 0.4243015646934509, + "learning_rate": 1.9684046298346857e-05, + "loss": 2.6348, + "step": 7133 + }, + { + "epoch": 0.913152, + "grad_norm": 0.381253719329834, + "learning_rate": 1.9626491521411772e-05, + "loss": 2.6143, + "step": 7134 + }, + { + "epoch": 0.91328, + "grad_norm": 0.4037047326564789, + "learning_rate": 1.956901932625177e-05, + "loss": 2.5938, + "step": 7135 + }, + { + "epoch": 0.913408, + "grad_norm": 0.38956552743911743, + "learning_rate": 1.9511629722747004e-05, + "loss": 2.4509, + "step": 7136 + }, + { + "epoch": 0.913536, + "grad_norm": 0.4162704646587372, + "learning_rate": 1.9454322720763362e-05, + "loss": 2.6231, + "step": 7137 + }, + { + "epoch": 0.913664, + "grad_norm": 0.4161348342895508, + "learning_rate": 1.939709833015252e-05, + "loss": 2.5569, + "step": 7138 + }, + { + "epoch": 0.913792, + "grad_norm": 0.43291059136390686, + "learning_rate": 1.9339956560752214e-05, + "loss": 2.5425, + "step": 7139 + }, + { + "epoch": 0.91392, + "grad_norm": 0.419244647026062, + "learning_rate": 1.9282897422385593e-05, + "loss": 2.4652, + "step": 7140 + }, + { + "epoch": 0.914048, + "grad_norm": 0.36543017625808716, + "learning_rate": 1.922592092486192e-05, + "loss": 2.4929, + "step": 7141 + }, + { + "epoch": 0.914176, + "grad_norm": 0.4239806532859802, + "learning_rate": 1.9169027077975965e-05, + "loss": 2.5511, + "step": 7142 + }, + { + "epoch": 0.914304, + "grad_norm": 0.3944949805736542, + "learning_rate": 1.9112215891508632e-05, + "loss": 2.5894, + "step": 7143 + }, + { + "epoch": 0.914432, + "grad_norm": 0.4182380139827728, + "learning_rate": 1.9055487375226277e-05, + "loss": 2.567, + "step": 7144 + }, + { + "epoch": 0.91456, + "grad_norm": 0.45609402656555176, + "learning_rate": 1.8998841538881152e-05, + "loss": 2.5468, + "step": 7145 + }, + { + "epoch": 0.914688, + "grad_norm": 0.4508400857448578, + "learning_rate": 1.8942278392211464e-05, + "loss": 2.5963, + "step": 7146 + }, + { + "epoch": 0.914816, + "grad_norm": 0.4298573136329651, + "learning_rate": 1.8885797944941052e-05, + "loss": 2.5725, + "step": 7147 + }, + { + "epoch": 0.914944, + "grad_norm": 0.3986316919326782, + "learning_rate": 1.8829400206779536e-05, + "loss": 2.5174, + "step": 7148 + }, + { + "epoch": 0.915072, + "grad_norm": 0.48708823323249817, + "learning_rate": 1.8773085187422323e-05, + "loss": 2.6501, + "step": 7149 + }, + { + "epoch": 0.9152, + "grad_norm": 0.40571650862693787, + "learning_rate": 1.8716852896550618e-05, + "loss": 2.4919, + "step": 7150 + }, + { + "epoch": 0.915328, + "grad_norm": 0.42349478602409363, + "learning_rate": 1.866070334383135e-05, + "loss": 2.5982, + "step": 7151 + }, + { + "epoch": 0.915456, + "grad_norm": 0.46220168471336365, + "learning_rate": 1.8604636538917364e-05, + "loss": 2.4953, + "step": 7152 + }, + { + "epoch": 0.915584, + "grad_norm": 0.3904590606689453, + "learning_rate": 1.8548652491447214e-05, + "loss": 2.4862, + "step": 7153 + }, + { + "epoch": 0.915712, + "grad_norm": 0.4993683695793152, + "learning_rate": 1.8492751211045157e-05, + "loss": 2.6719, + "step": 7154 + }, + { + "epoch": 0.91584, + "grad_norm": 0.3975001573562622, + "learning_rate": 1.8436932707321273e-05, + "loss": 2.5908, + "step": 7155 + }, + { + "epoch": 0.915968, + "grad_norm": 0.44054633378982544, + "learning_rate": 1.8381196989871452e-05, + "loss": 2.4684, + "step": 7156 + }, + { + "epoch": 0.916096, + "grad_norm": 0.4088689088821411, + "learning_rate": 1.8325544068277245e-05, + "loss": 2.6369, + "step": 7157 + }, + { + "epoch": 0.916224, + "grad_norm": 0.4143781363964081, + "learning_rate": 1.8269973952106055e-05, + "loss": 2.4842, + "step": 7158 + }, + { + "epoch": 0.916352, + "grad_norm": 0.38713449239730835, + "learning_rate": 1.821448665091102e-05, + "loss": 2.5787, + "step": 7159 + }, + { + "epoch": 0.91648, + "grad_norm": 0.3993268609046936, + "learning_rate": 1.8159082174231012e-05, + "loss": 2.6134, + "step": 7160 + }, + { + "epoch": 0.916608, + "grad_norm": 0.42989811301231384, + "learning_rate": 1.8103760531590852e-05, + "loss": 2.5761, + "step": 7161 + }, + { + "epoch": 0.916736, + "grad_norm": 0.4158007502555847, + "learning_rate": 1.8048521732500878e-05, + "loss": 2.5258, + "step": 7162 + }, + { + "epoch": 0.916864, + "grad_norm": 0.48868149518966675, + "learning_rate": 1.7993365786457217e-05, + "loss": 2.4379, + "step": 7163 + }, + { + "epoch": 0.916992, + "grad_norm": 0.40952232480049133, + "learning_rate": 1.793829270294195e-05, + "loss": 2.6132, + "step": 7164 + }, + { + "epoch": 0.91712, + "grad_norm": 0.38538268208503723, + "learning_rate": 1.7883302491422672e-05, + "loss": 2.4933, + "step": 7165 + }, + { + "epoch": 0.917248, + "grad_norm": 0.3964797258377075, + "learning_rate": 1.7828395161352817e-05, + "loss": 2.5993, + "step": 7166 + }, + { + "epoch": 0.917376, + "grad_norm": 0.4230397045612335, + "learning_rate": 1.777357072217173e-05, + "loss": 2.5619, + "step": 7167 + }, + { + "epoch": 0.917504, + "grad_norm": 0.3907894194126129, + "learning_rate": 1.7718829183304252e-05, + "loss": 2.5035, + "step": 7168 + }, + { + "epoch": 0.917632, + "grad_norm": 0.43706634640693665, + "learning_rate": 1.7664170554161142e-05, + "loss": 2.563, + "step": 7169 + }, + { + "epoch": 0.91776, + "grad_norm": 0.7130138874053955, + "learning_rate": 1.7609594844138766e-05, + "loss": 2.5795, + "step": 7170 + }, + { + "epoch": 0.917888, + "grad_norm": 0.41963106393814087, + "learning_rate": 1.7555102062619455e-05, + "loss": 2.5014, + "step": 7171 + }, + { + "epoch": 0.918016, + "grad_norm": 0.4021182656288147, + "learning_rate": 1.750069221897105e-05, + "loss": 2.5317, + "step": 7172 + }, + { + "epoch": 0.918144, + "grad_norm": 0.46719157695770264, + "learning_rate": 1.744636532254723e-05, + "loss": 2.6525, + "step": 7173 + }, + { + "epoch": 0.918272, + "grad_norm": 0.38142529129981995, + "learning_rate": 1.739212138268753e-05, + "loss": 2.5021, + "step": 7174 + }, + { + "epoch": 0.9184, + "grad_norm": 0.3974112272262573, + "learning_rate": 1.733796040871699e-05, + "loss": 2.6756, + "step": 7175 + }, + { + "epoch": 0.918528, + "grad_norm": 0.40443652868270874, + "learning_rate": 1.7283882409946548e-05, + "loss": 2.605, + "step": 7176 + }, + { + "epoch": 0.918656, + "grad_norm": 0.464616984128952, + "learning_rate": 1.722988739567288e-05, + "loss": 2.6307, + "step": 7177 + }, + { + "epoch": 0.918784, + "grad_norm": 0.4505669176578522, + "learning_rate": 1.7175975375178344e-05, + "loss": 2.5978, + "step": 7178 + }, + { + "epoch": 0.918912, + "grad_norm": 0.4281698167324066, + "learning_rate": 1.7122146357730907e-05, + "loss": 2.6762, + "step": 7179 + }, + { + "epoch": 0.91904, + "grad_norm": 0.4053265452384949, + "learning_rate": 1.706840035258461e-05, + "loss": 2.6142, + "step": 7180 + }, + { + "epoch": 0.919168, + "grad_norm": 0.42141085863113403, + "learning_rate": 1.7014737368978793e-05, + "loss": 2.5071, + "step": 7181 + }, + { + "epoch": 0.919296, + "grad_norm": 0.3837496042251587, + "learning_rate": 1.696115741613902e-05, + "loss": 2.5938, + "step": 7182 + }, + { + "epoch": 0.919424, + "grad_norm": 0.46189379692077637, + "learning_rate": 1.690766050327608e-05, + "loss": 2.6364, + "step": 7183 + }, + { + "epoch": 0.919552, + "grad_norm": 0.4716891348361969, + "learning_rate": 1.685424663958679e-05, + "loss": 2.6442, + "step": 7184 + }, + { + "epoch": 0.91968, + "grad_norm": 0.3987894356250763, + "learning_rate": 1.6800915834253637e-05, + "loss": 2.4725, + "step": 7185 + }, + { + "epoch": 0.919808, + "grad_norm": 0.4708530008792877, + "learning_rate": 1.674766809644479e-05, + "loss": 2.594, + "step": 7186 + }, + { + "epoch": 0.919936, + "grad_norm": 0.4059840142726898, + "learning_rate": 1.6694503435314036e-05, + "loss": 2.6076, + "step": 7187 + }, + { + "epoch": 0.920064, + "grad_norm": 0.3933749794960022, + "learning_rate": 1.664142186000117e-05, + "loss": 2.5559, + "step": 7188 + }, + { + "epoch": 0.920192, + "grad_norm": 0.37262317538261414, + "learning_rate": 1.6588423379631458e-05, + "loss": 2.6328, + "step": 7189 + }, + { + "epoch": 0.92032, + "grad_norm": 0.3989917039871216, + "learning_rate": 1.6535508003315937e-05, + "loss": 2.5551, + "step": 7190 + }, + { + "epoch": 0.920448, + "grad_norm": 0.3929620683193207, + "learning_rate": 1.6482675740151442e-05, + "loss": 2.5593, + "step": 7191 + }, + { + "epoch": 0.920576, + "grad_norm": 0.3943556845188141, + "learning_rate": 1.6429926599220378e-05, + "loss": 2.6975, + "step": 7192 + }, + { + "epoch": 0.920704, + "grad_norm": 0.41368553042411804, + "learning_rate": 1.6377260589590938e-05, + "loss": 2.4892, + "step": 7193 + }, + { + "epoch": 0.920832, + "grad_norm": 0.5935844779014587, + "learning_rate": 1.632467772031704e-05, + "loss": 2.5787, + "step": 7194 + }, + { + "epoch": 0.92096, + "grad_norm": 0.3837582468986511, + "learning_rate": 1.627217800043829e-05, + "loss": 2.5404, + "step": 7195 + }, + { + "epoch": 0.921088, + "grad_norm": 0.4189823865890503, + "learning_rate": 1.621976143898002e-05, + "loss": 2.5105, + "step": 7196 + }, + { + "epoch": 0.921216, + "grad_norm": 0.4165870249271393, + "learning_rate": 1.6167428044953137e-05, + "loss": 2.5506, + "step": 7197 + }, + { + "epoch": 0.921344, + "grad_norm": 0.43843474984169006, + "learning_rate": 1.6115177827354554e-05, + "loss": 2.4275, + "step": 7198 + }, + { + "epoch": 0.921472, + "grad_norm": 0.39749637246131897, + "learning_rate": 1.6063010795166534e-05, + "loss": 2.5351, + "step": 7199 + }, + { + "epoch": 0.9216, + "grad_norm": 0.3396679162979126, + "learning_rate": 1.601092695735723e-05, + "loss": 2.624, + "step": 7200 + }, + { + "epoch": 0.921728, + "grad_norm": 0.4060334265232086, + "learning_rate": 1.5958926322880484e-05, + "loss": 2.4325, + "step": 7201 + }, + { + "epoch": 0.921856, + "grad_norm": 0.41730445623397827, + "learning_rate": 1.5907008900675758e-05, + "loss": 2.5789, + "step": 7202 + }, + { + "epoch": 0.921984, + "grad_norm": 0.47442376613616943, + "learning_rate": 1.5855174699668296e-05, + "loss": 2.5909, + "step": 7203 + }, + { + "epoch": 0.922112, + "grad_norm": 0.3716631531715393, + "learning_rate": 1.580342372876903e-05, + "loss": 2.5498, + "step": 7204 + }, + { + "epoch": 0.92224, + "grad_norm": 0.45109280943870544, + "learning_rate": 1.5751755996874452e-05, + "loss": 2.4873, + "step": 7205 + }, + { + "epoch": 0.922368, + "grad_norm": 0.44068366289138794, + "learning_rate": 1.5700171512866957e-05, + "loss": 2.5775, + "step": 7206 + }, + { + "epoch": 0.922496, + "grad_norm": 0.42454221844673157, + "learning_rate": 1.5648670285614396e-05, + "loss": 2.5811, + "step": 7207 + }, + { + "epoch": 0.922624, + "grad_norm": 0.45816537737846375, + "learning_rate": 1.559725232397058e-05, + "loss": 2.5491, + "step": 7208 + }, + { + "epoch": 0.922752, + "grad_norm": 0.3920009732246399, + "learning_rate": 1.5545917636774653e-05, + "loss": 2.7065, + "step": 7209 + }, + { + "epoch": 0.92288, + "grad_norm": 0.41835349798202515, + "learning_rate": 1.5494666232851894e-05, + "loss": 2.6629, + "step": 7210 + }, + { + "epoch": 0.923008, + "grad_norm": 0.4094226062297821, + "learning_rate": 1.5443498121012812e-05, + "loss": 2.4218, + "step": 7211 + }, + { + "epoch": 0.923136, + "grad_norm": 0.3945467472076416, + "learning_rate": 1.5392413310053867e-05, + "loss": 2.6306, + "step": 7212 + }, + { + "epoch": 0.923264, + "grad_norm": 0.40114641189575195, + "learning_rate": 1.5341411808757143e-05, + "loss": 2.5213, + "step": 7213 + }, + { + "epoch": 0.923392, + "grad_norm": 0.405740886926651, + "learning_rate": 1.529049362589041e-05, + "loss": 2.5835, + "step": 7214 + }, + { + "epoch": 0.92352, + "grad_norm": 0.4317033588886261, + "learning_rate": 1.5239658770206943e-05, + "loss": 2.3927, + "step": 7215 + }, + { + "epoch": 0.923648, + "grad_norm": 0.37236690521240234, + "learning_rate": 1.5188907250446027e-05, + "loss": 2.4892, + "step": 7216 + }, + { + "epoch": 0.923776, + "grad_norm": 0.3983495533466339, + "learning_rate": 1.513823907533235e-05, + "loss": 2.5744, + "step": 7217 + }, + { + "epoch": 0.923904, + "grad_norm": 0.4099683463573456, + "learning_rate": 1.5087654253576388e-05, + "loss": 2.564, + "step": 7218 + }, + { + "epoch": 0.924032, + "grad_norm": 0.41086888313293457, + "learning_rate": 1.5037152793874243e-05, + "loss": 2.4443, + "step": 7219 + }, + { + "epoch": 0.92416, + "grad_norm": 0.39392542839050293, + "learning_rate": 1.4986734704907745e-05, + "loss": 2.6061, + "step": 7220 + }, + { + "epoch": 0.924288, + "grad_norm": 0.4264104664325714, + "learning_rate": 1.4936399995344297e-05, + "loss": 2.5209, + "step": 7221 + }, + { + "epoch": 0.924416, + "grad_norm": 0.45430323481559753, + "learning_rate": 1.4886148673836974e-05, + "loss": 2.537, + "step": 7222 + }, + { + "epoch": 0.924544, + "grad_norm": 0.3836105167865753, + "learning_rate": 1.4835980749024591e-05, + "loss": 2.5657, + "step": 7223 + }, + { + "epoch": 0.924672, + "grad_norm": 0.40656813979148865, + "learning_rate": 1.4785896229531693e-05, + "loss": 2.5737, + "step": 7224 + }, + { + "epoch": 0.9248, + "grad_norm": 0.43133068084716797, + "learning_rate": 1.473589512396828e-05, + "loss": 2.6078, + "step": 7225 + }, + { + "epoch": 0.924928, + "grad_norm": 0.419893741607666, + "learning_rate": 1.4685977440930199e-05, + "loss": 2.5858, + "step": 7226 + }, + { + "epoch": 0.925056, + "grad_norm": 0.4268682301044464, + "learning_rate": 1.4636143188998807e-05, + "loss": 2.6127, + "step": 7227 + }, + { + "epoch": 0.925184, + "grad_norm": 0.3703017234802246, + "learning_rate": 1.4586392376741254e-05, + "loss": 2.5581, + "step": 7228 + }, + { + "epoch": 0.925312, + "grad_norm": 0.40217265486717224, + "learning_rate": 1.453672501271025e-05, + "loss": 2.6321, + "step": 7229 + }, + { + "epoch": 0.92544, + "grad_norm": 0.4201507866382599, + "learning_rate": 1.4487141105444134e-05, + "loss": 2.513, + "step": 7230 + }, + { + "epoch": 0.925568, + "grad_norm": 0.39665839076042175, + "learning_rate": 1.4437640663467033e-05, + "loss": 2.5942, + "step": 7231 + }, + { + "epoch": 0.925696, + "grad_norm": 0.3985322117805481, + "learning_rate": 1.4388223695288694e-05, + "loss": 2.4721, + "step": 7232 + }, + { + "epoch": 0.925824, + "grad_norm": 0.3932397961616516, + "learning_rate": 1.433889020940432e-05, + "loss": 2.5292, + "step": 7233 + }, + { + "epoch": 0.925952, + "grad_norm": 0.48039552569389343, + "learning_rate": 1.4289640214294964e-05, + "loss": 2.5885, + "step": 7234 + }, + { + "epoch": 0.92608, + "grad_norm": 0.47686952352523804, + "learning_rate": 1.4240473718427348e-05, + "loss": 2.6812, + "step": 7235 + }, + { + "epoch": 0.926208, + "grad_norm": 0.40371447801589966, + "learning_rate": 1.4191390730253716e-05, + "loss": 2.4783, + "step": 7236 + }, + { + "epoch": 0.926336, + "grad_norm": 0.41442030668258667, + "learning_rate": 1.4142391258211984e-05, + "loss": 2.5177, + "step": 7237 + }, + { + "epoch": 0.926464, + "grad_norm": 0.440430223941803, + "learning_rate": 1.409347531072569e-05, + "loss": 2.6437, + "step": 7238 + }, + { + "epoch": 0.926592, + "grad_norm": 0.41276225447654724, + "learning_rate": 1.4044642896204107e-05, + "loss": 2.5787, + "step": 7239 + }, + { + "epoch": 0.92672, + "grad_norm": 0.4018295407295227, + "learning_rate": 1.3995894023042133e-05, + "loss": 2.5703, + "step": 7240 + }, + { + "epoch": 0.926848, + "grad_norm": 0.4535297155380249, + "learning_rate": 1.394722869962023e-05, + "loss": 2.4888, + "step": 7241 + }, + { + "epoch": 0.926976, + "grad_norm": 0.3816322386264801, + "learning_rate": 1.3898646934304538e-05, + "loss": 2.529, + "step": 7242 + }, + { + "epoch": 0.927104, + "grad_norm": 0.44566610455513, + "learning_rate": 1.3850148735446766e-05, + "loss": 2.5674, + "step": 7243 + }, + { + "epoch": 0.927232, + "grad_norm": 0.4206902086734772, + "learning_rate": 1.38017341113843e-05, + "loss": 2.5568, + "step": 7244 + }, + { + "epoch": 0.92736, + "grad_norm": 0.5049059391021729, + "learning_rate": 1.3753403070440262e-05, + "loss": 2.4888, + "step": 7245 + }, + { + "epoch": 0.927488, + "grad_norm": 0.42197924852371216, + "learning_rate": 1.3705155620923337e-05, + "loss": 2.6927, + "step": 7246 + }, + { + "epoch": 0.927616, + "grad_norm": 0.40403568744659424, + "learning_rate": 1.365699177112778e-05, + "loss": 2.4937, + "step": 7247 + }, + { + "epoch": 0.927744, + "grad_norm": 0.4047708213329315, + "learning_rate": 1.3608911529333467e-05, + "loss": 2.6805, + "step": 7248 + }, + { + "epoch": 0.927872, + "grad_norm": 0.5405136346817017, + "learning_rate": 1.3560914903806065e-05, + "loss": 2.5644, + "step": 7249 + }, + { + "epoch": 0.928, + "grad_norm": 0.42183369398117065, + "learning_rate": 1.351300190279664e-05, + "loss": 2.6234, + "step": 7250 + }, + { + "epoch": 0.928128, + "grad_norm": 0.4542330801486969, + "learning_rate": 1.3465172534541937e-05, + "loss": 2.6072, + "step": 7251 + }, + { + "epoch": 0.928256, + "grad_norm": 0.4270299971103668, + "learning_rate": 1.3417426807264599e-05, + "loss": 2.5848, + "step": 7252 + }, + { + "epoch": 0.928384, + "grad_norm": 0.46278488636016846, + "learning_rate": 1.3369764729172451e-05, + "loss": 2.5876, + "step": 7253 + }, + { + "epoch": 0.928512, + "grad_norm": 0.4094250202178955, + "learning_rate": 1.3322186308459271e-05, + "loss": 2.5389, + "step": 7254 + }, + { + "epoch": 0.92864, + "grad_norm": 0.5805171728134155, + "learning_rate": 1.3274691553304353e-05, + "loss": 2.6243, + "step": 7255 + }, + { + "epoch": 0.928768, + "grad_norm": 0.446397066116333, + "learning_rate": 1.322728047187255e-05, + "loss": 2.7155, + "step": 7256 + }, + { + "epoch": 0.928896, + "grad_norm": 0.4056369960308075, + "learning_rate": 1.3179953072314398e-05, + "loss": 2.5086, + "step": 7257 + }, + { + "epoch": 0.929024, + "grad_norm": 0.42251092195510864, + "learning_rate": 1.3132709362766004e-05, + "loss": 2.5884, + "step": 7258 + }, + { + "epoch": 0.929152, + "grad_norm": 0.4047323167324066, + "learning_rate": 1.3085549351349091e-05, + "loss": 2.4779, + "step": 7259 + }, + { + "epoch": 0.92928, + "grad_norm": 0.4671434164047241, + "learning_rate": 1.3038473046171061e-05, + "loss": 2.5793, + "step": 7260 + }, + { + "epoch": 0.929408, + "grad_norm": 0.38276737928390503, + "learning_rate": 1.299148045532489e-05, + "loss": 2.6442, + "step": 7261 + }, + { + "epoch": 0.929536, + "grad_norm": 0.49448198080062866, + "learning_rate": 1.2944571586888997e-05, + "loss": 2.5516, + "step": 7262 + }, + { + "epoch": 0.929664, + "grad_norm": 0.4366230070590973, + "learning_rate": 1.2897746448927827e-05, + "loss": 2.4888, + "step": 7263 + }, + { + "epoch": 0.929792, + "grad_norm": 0.49051496386528015, + "learning_rate": 1.2851005049490939e-05, + "loss": 2.5094, + "step": 7264 + }, + { + "epoch": 0.92992, + "grad_norm": 0.4572928547859192, + "learning_rate": 1.2804347396613847e-05, + "loss": 2.6774, + "step": 7265 + }, + { + "epoch": 0.930048, + "grad_norm": 0.42147698998451233, + "learning_rate": 1.2757773498317416e-05, + "loss": 2.5372, + "step": 7266 + }, + { + "epoch": 0.930176, + "grad_norm": 0.42955780029296875, + "learning_rate": 1.271128336260835e-05, + "loss": 2.626, + "step": 7267 + }, + { + "epoch": 0.930304, + "grad_norm": 0.4276795983314514, + "learning_rate": 1.266487699747887e-05, + "loss": 2.6199, + "step": 7268 + }, + { + "epoch": 0.930432, + "grad_norm": 0.44419267773628235, + "learning_rate": 1.2618554410906646e-05, + "loss": 2.5075, + "step": 7269 + }, + { + "epoch": 0.93056, + "grad_norm": 0.41769522428512573, + "learning_rate": 1.25723156108552e-05, + "loss": 2.6078, + "step": 7270 + }, + { + "epoch": 0.930688, + "grad_norm": 0.41800788044929504, + "learning_rate": 1.252616060527334e-05, + "loss": 2.5594, + "step": 7271 + }, + { + "epoch": 0.930816, + "grad_norm": 0.43055352568626404, + "learning_rate": 1.248008940209583e-05, + "loss": 2.5819, + "step": 7272 + }, + { + "epoch": 0.930944, + "grad_norm": 0.4433128535747528, + "learning_rate": 1.243410200924272e-05, + "loss": 2.6771, + "step": 7273 + }, + { + "epoch": 0.931072, + "grad_norm": 0.4333994388580322, + "learning_rate": 1.2388198434619802e-05, + "loss": 2.5329, + "step": 7274 + }, + { + "epoch": 0.9312, + "grad_norm": 0.4949740171432495, + "learning_rate": 1.2342378686118538e-05, + "loss": 2.6109, + "step": 7275 + }, + { + "epoch": 0.931328, + "grad_norm": 0.4048687517642975, + "learning_rate": 1.229664277161574e-05, + "loss": 2.6668, + "step": 7276 + }, + { + "epoch": 0.931456, + "grad_norm": 0.4539468586444855, + "learning_rate": 1.2250990698974007e-05, + "loss": 2.6235, + "step": 7277 + }, + { + "epoch": 0.931584, + "grad_norm": 0.40177011489868164, + "learning_rate": 1.220542247604145e-05, + "loss": 2.4547, + "step": 7278 + }, + { + "epoch": 0.931712, + "grad_norm": 0.45285356044769287, + "learning_rate": 1.2159938110651803e-05, + "loss": 2.5257, + "step": 7279 + }, + { + "epoch": 0.93184, + "grad_norm": 1.194417119026184, + "learning_rate": 1.2114537610624254e-05, + "loss": 2.4104, + "step": 7280 + }, + { + "epoch": 0.931968, + "grad_norm": 0.40595313906669617, + "learning_rate": 1.2069220983763729e-05, + "loss": 2.6084, + "step": 7281 + }, + { + "epoch": 0.932096, + "grad_norm": 0.38193726539611816, + "learning_rate": 1.2023988237860717e-05, + "loss": 2.5759, + "step": 7282 + }, + { + "epoch": 0.932224, + "grad_norm": 0.4369750916957855, + "learning_rate": 1.1978839380691275e-05, + "loss": 2.5773, + "step": 7283 + }, + { + "epoch": 0.932352, + "grad_norm": 0.40815407037734985, + "learning_rate": 1.1933774420016974e-05, + "loss": 2.5769, + "step": 7284 + }, + { + "epoch": 0.93248, + "grad_norm": 0.38337674736976624, + "learning_rate": 1.1888793363584948e-05, + "loss": 2.6251, + "step": 7285 + }, + { + "epoch": 0.932608, + "grad_norm": 0.404139906167984, + "learning_rate": 1.184389621912807e-05, + "loss": 2.6318, + "step": 7286 + }, + { + "epoch": 0.932736, + "grad_norm": 0.4245070815086365, + "learning_rate": 1.1799082994364551e-05, + "loss": 2.4863, + "step": 7287 + }, + { + "epoch": 0.932864, + "grad_norm": 0.40501803159713745, + "learning_rate": 1.17543536969984e-05, + "loss": 2.5778, + "step": 7288 + }, + { + "epoch": 0.932992, + "grad_norm": 0.48955070972442627, + "learning_rate": 1.1709708334719127e-05, + "loss": 2.529, + "step": 7289 + }, + { + "epoch": 0.93312, + "grad_norm": 0.41297686100006104, + "learning_rate": 1.1665146915201653e-05, + "loss": 2.551, + "step": 7290 + }, + { + "epoch": 0.933248, + "grad_norm": 0.4225448966026306, + "learning_rate": 1.1620669446106735e-05, + "loss": 2.7046, + "step": 7291 + }, + { + "epoch": 0.933376, + "grad_norm": 0.40051066875457764, + "learning_rate": 1.1576275935080538e-05, + "loss": 2.4873, + "step": 7292 + }, + { + "epoch": 0.933504, + "grad_norm": 0.4681970775127411, + "learning_rate": 1.1531966389754788e-05, + "loss": 2.5252, + "step": 7293 + }, + { + "epoch": 0.933632, + "grad_norm": 0.4141117036342621, + "learning_rate": 1.1487740817746839e-05, + "loss": 2.5353, + "step": 7294 + }, + { + "epoch": 0.93376, + "grad_norm": 0.465634286403656, + "learning_rate": 1.1443599226659495e-05, + "loss": 2.46, + "step": 7295 + }, + { + "epoch": 0.933888, + "grad_norm": 0.4559068977832794, + "learning_rate": 1.1399541624081355e-05, + "loss": 2.5561, + "step": 7296 + }, + { + "epoch": 0.934016, + "grad_norm": 0.4160729944705963, + "learning_rate": 1.1355568017586304e-05, + "loss": 2.5474, + "step": 7297 + }, + { + "epoch": 0.934144, + "grad_norm": 0.43605276942253113, + "learning_rate": 1.1311678414734016e-05, + "loss": 2.6237, + "step": 7298 + }, + { + "epoch": 0.934272, + "grad_norm": 0.3892807960510254, + "learning_rate": 1.1267872823069458e-05, + "loss": 2.5073, + "step": 7299 + }, + { + "epoch": 0.9344, + "grad_norm": 0.40114539861679077, + "learning_rate": 1.1224151250123548e-05, + "loss": 2.4471, + "step": 7300 + }, + { + "epoch": 0.934528, + "grad_norm": 0.3832181692123413, + "learning_rate": 1.1180513703412387e-05, + "loss": 2.5943, + "step": 7301 + }, + { + "epoch": 0.934656, + "grad_norm": 0.3984559178352356, + "learning_rate": 1.1136960190437751e-05, + "loss": 2.5881, + "step": 7302 + }, + { + "epoch": 0.934784, + "grad_norm": 0.4167960584163666, + "learning_rate": 1.1093490718687094e-05, + "loss": 2.4369, + "step": 7303 + }, + { + "epoch": 0.934912, + "grad_norm": 0.40504327416419983, + "learning_rate": 1.1050105295633272e-05, + "loss": 2.7034, + "step": 7304 + }, + { + "epoch": 0.93504, + "grad_norm": 0.4598621129989624, + "learning_rate": 1.100680392873471e-05, + "loss": 2.6593, + "step": 7305 + }, + { + "epoch": 0.935168, + "grad_norm": 0.593015193939209, + "learning_rate": 1.0963586625435506e-05, + "loss": 2.6213, + "step": 7306 + }, + { + "epoch": 0.935296, + "grad_norm": 0.39840957522392273, + "learning_rate": 1.092045339316511e-05, + "loss": 2.5904, + "step": 7307 + }, + { + "epoch": 0.935424, + "grad_norm": 0.44056832790374756, + "learning_rate": 1.087740423933864e-05, + "loss": 2.5508, + "step": 7308 + }, + { + "epoch": 0.935552, + "grad_norm": 0.5301606059074402, + "learning_rate": 1.0834439171356846e-05, + "loss": 2.6634, + "step": 7309 + }, + { + "epoch": 0.93568, + "grad_norm": 0.39574146270751953, + "learning_rate": 1.0791558196605822e-05, + "loss": 2.6499, + "step": 7310 + }, + { + "epoch": 0.935808, + "grad_norm": 0.40742936730384827, + "learning_rate": 1.0748761322457335e-05, + "loss": 2.496, + "step": 7311 + }, + { + "epoch": 0.935936, + "grad_norm": 0.4357624053955078, + "learning_rate": 1.0706048556268666e-05, + "loss": 2.5966, + "step": 7312 + }, + { + "epoch": 0.936064, + "grad_norm": 0.4151630103588104, + "learning_rate": 1.0663419905382666e-05, + "loss": 2.5131, + "step": 7313 + }, + { + "epoch": 0.936192, + "grad_norm": 0.3948584198951721, + "learning_rate": 1.0620875377127637e-05, + "loss": 2.6552, + "step": 7314 + }, + { + "epoch": 0.93632, + "grad_norm": 0.4036500155925751, + "learning_rate": 1.0578414978817508e-05, + "loss": 2.5266, + "step": 7315 + }, + { + "epoch": 0.936448, + "grad_norm": 0.41104063391685486, + "learning_rate": 1.0536038717751606e-05, + "loss": 2.6501, + "step": 7316 + }, + { + "epoch": 0.936576, + "grad_norm": 0.40640750527381897, + "learning_rate": 1.0493746601215104e-05, + "loss": 2.5774, + "step": 7317 + }, + { + "epoch": 0.936704, + "grad_norm": 0.46106746792793274, + "learning_rate": 1.0451538636478353e-05, + "loss": 2.617, + "step": 7318 + }, + { + "epoch": 0.936832, + "grad_norm": 0.40800198912620544, + "learning_rate": 1.0409414830797493e-05, + "loss": 2.5016, + "step": 7319 + }, + { + "epoch": 0.93696, + "grad_norm": 0.4067517817020416, + "learning_rate": 1.0367375191414064e-05, + "loss": 2.514, + "step": 7320 + }, + { + "epoch": 0.937088, + "grad_norm": 0.41281867027282715, + "learning_rate": 1.0325419725555119e-05, + "loss": 2.5809, + "step": 7321 + }, + { + "epoch": 0.937216, + "grad_norm": 0.440372109413147, + "learning_rate": 1.0283548440433333e-05, + "loss": 2.6929, + "step": 7322 + }, + { + "epoch": 0.937344, + "grad_norm": 0.4149656593799591, + "learning_rate": 1.024176134324678e-05, + "loss": 2.5243, + "step": 7323 + }, + { + "epoch": 0.937472, + "grad_norm": 0.43490371108055115, + "learning_rate": 1.0200058441179271e-05, + "loss": 2.6154, + "step": 7324 + }, + { + "epoch": 0.9376, + "grad_norm": 0.4151422679424286, + "learning_rate": 1.015843974139996e-05, + "loss": 2.5922, + "step": 7325 + }, + { + "epoch": 0.937728, + "grad_norm": 0.4314764440059662, + "learning_rate": 1.0116905251063623e-05, + "loss": 2.5701, + "step": 7326 + }, + { + "epoch": 0.937856, + "grad_norm": 0.4217415750026703, + "learning_rate": 1.0075454977310383e-05, + "loss": 2.6648, + "step": 7327 + }, + { + "epoch": 0.937984, + "grad_norm": 0.4381561577320099, + "learning_rate": 1.0034088927266205e-05, + "loss": 2.5829, + "step": 7328 + }, + { + "epoch": 0.938112, + "grad_norm": 0.3950148820877075, + "learning_rate": 9.99280710804229e-06, + "loss": 2.5384, + "step": 7329 + }, + { + "epoch": 0.93824, + "grad_norm": 0.41851601004600525, + "learning_rate": 9.951609526735517e-06, + "loss": 2.6193, + "step": 7330 + }, + { + "epoch": 0.938368, + "grad_norm": 0.41929322481155396, + "learning_rate": 9.910496190428163e-06, + "loss": 2.5149, + "step": 7331 + }, + { + "epoch": 0.938496, + "grad_norm": 0.40609264373779297, + "learning_rate": 9.869467106188135e-06, + "loss": 2.6435, + "step": 7332 + }, + { + "epoch": 0.938624, + "grad_norm": 0.47448524832725525, + "learning_rate": 9.828522281068786e-06, + "loss": 2.6233, + "step": 7333 + }, + { + "epoch": 0.938752, + "grad_norm": 0.40476131439208984, + "learning_rate": 9.787661722108987e-06, + "loss": 2.6246, + "step": 7334 + }, + { + "epoch": 0.93888, + "grad_norm": 0.4117335081100464, + "learning_rate": 9.746885436333231e-06, + "loss": 2.4654, + "step": 7335 + }, + { + "epoch": 0.939008, + "grad_norm": 0.44802790880203247, + "learning_rate": 9.706193430751298e-06, + "loss": 2.5825, + "step": 7336 + }, + { + "epoch": 0.939136, + "grad_norm": 0.42264145612716675, + "learning_rate": 9.665585712358705e-06, + "loss": 2.7474, + "step": 7337 + }, + { + "epoch": 0.939264, + "grad_norm": 0.37594422698020935, + "learning_rate": 9.625062288136366e-06, + "loss": 2.4451, + "step": 7338 + }, + { + "epoch": 0.939392, + "grad_norm": 0.3982522487640381, + "learning_rate": 9.584623165050766e-06, + "loss": 2.5256, + "step": 7339 + }, + { + "epoch": 0.93952, + "grad_norm": 0.43084684014320374, + "learning_rate": 9.544268350053842e-06, + "loss": 2.5995, + "step": 7340 + }, + { + "epoch": 0.939648, + "grad_norm": 0.3749980628490448, + "learning_rate": 9.50399785008299e-06, + "loss": 2.541, + "step": 7341 + }, + { + "epoch": 0.939776, + "grad_norm": 0.6122822761535645, + "learning_rate": 9.463811672061284e-06, + "loss": 2.5578, + "step": 7342 + }, + { + "epoch": 0.939904, + "grad_norm": 0.42344003915786743, + "learning_rate": 9.423709822897086e-06, + "loss": 2.6289, + "step": 7343 + }, + { + "epoch": 0.940032, + "grad_norm": 0.5347748398780823, + "learning_rate": 9.383692309484381e-06, + "loss": 2.6601, + "step": 7344 + }, + { + "epoch": 0.94016, + "grad_norm": 0.4243503510951996, + "learning_rate": 9.343759138702724e-06, + "loss": 2.4796, + "step": 7345 + }, + { + "epoch": 0.940288, + "grad_norm": 0.3900662362575531, + "learning_rate": 9.303910317417064e-06, + "loss": 2.4789, + "step": 7346 + }, + { + "epoch": 0.940416, + "grad_norm": 0.42213526368141174, + "learning_rate": 9.264145852477868e-06, + "loss": 2.5572, + "step": 7347 + }, + { + "epoch": 0.940544, + "grad_norm": 0.39184248447418213, + "learning_rate": 9.224465750721056e-06, + "loss": 2.5761, + "step": 7348 + }, + { + "epoch": 0.940672, + "grad_norm": 0.472854882478714, + "learning_rate": 9.184870018968172e-06, + "loss": 2.7077, + "step": 7349 + }, + { + "epoch": 0.9408, + "grad_norm": 0.4059925973415375, + "learning_rate": 9.14535866402616e-06, + "loss": 2.7253, + "step": 7350 + }, + { + "epoch": 0.940928, + "grad_norm": 0.3940680921077728, + "learning_rate": 9.105931692687419e-06, + "loss": 2.6203, + "step": 7351 + }, + { + "epoch": 0.941056, + "grad_norm": 0.4841678738594055, + "learning_rate": 9.066589111729972e-06, + "loss": 2.6935, + "step": 7352 + }, + { + "epoch": 0.941184, + "grad_norm": 0.4817845821380615, + "learning_rate": 9.027330927917243e-06, + "loss": 2.5463, + "step": 7353 + }, + { + "epoch": 0.941312, + "grad_norm": 0.47015953063964844, + "learning_rate": 8.988157147998222e-06, + "loss": 2.5739, + "step": 7354 + }, + { + "epoch": 0.94144, + "grad_norm": 0.3959668278694153, + "learning_rate": 8.949067778707187e-06, + "loss": 2.5522, + "step": 7355 + }, + { + "epoch": 0.941568, + "grad_norm": 0.3927760422229767, + "learning_rate": 8.910062826764208e-06, + "loss": 2.5614, + "step": 7356 + }, + { + "epoch": 0.941696, + "grad_norm": 0.4417262673377991, + "learning_rate": 8.871142298874646e-06, + "loss": 2.6295, + "step": 7357 + }, + { + "epoch": 0.941824, + "grad_norm": 0.4545871317386627, + "learning_rate": 8.832306201729368e-06, + "loss": 2.5511, + "step": 7358 + }, + { + "epoch": 0.941952, + "grad_norm": 0.4545605778694153, + "learning_rate": 8.793554542004756e-06, + "loss": 2.5402, + "step": 7359 + }, + { + "epoch": 0.94208, + "grad_norm": 0.4790133535861969, + "learning_rate": 8.754887326362648e-06, + "loss": 2.6583, + "step": 7360 + }, + { + "epoch": 0.942208, + "grad_norm": 0.602649986743927, + "learning_rate": 8.71630456145045e-06, + "loss": 2.5632, + "step": 7361 + }, + { + "epoch": 0.942336, + "grad_norm": 0.39419546723365784, + "learning_rate": 8.677806253900967e-06, + "loss": 2.5341, + "step": 7362 + }, + { + "epoch": 0.942464, + "grad_norm": 0.4342040717601776, + "learning_rate": 8.639392410332403e-06, + "loss": 2.439, + "step": 7363 + }, + { + "epoch": 0.942592, + "grad_norm": 0.40779435634613037, + "learning_rate": 8.601063037348644e-06, + "loss": 2.5178, + "step": 7364 + }, + { + "epoch": 0.94272, + "grad_norm": 0.4825285077095032, + "learning_rate": 8.562818141538975e-06, + "loss": 2.6454, + "step": 7365 + }, + { + "epoch": 0.942848, + "grad_norm": 0.3724840581417084, + "learning_rate": 8.52465772947808e-06, + "loss": 2.5923, + "step": 7366 + }, + { + "epoch": 0.942976, + "grad_norm": 0.44499698281288147, + "learning_rate": 8.486581807726157e-06, + "loss": 2.4998, + "step": 7367 + }, + { + "epoch": 0.943104, + "grad_norm": 0.3638100028038025, + "learning_rate": 8.448590382829024e-06, + "loss": 2.5236, + "step": 7368 + }, + { + "epoch": 0.943232, + "grad_norm": 0.4093799591064453, + "learning_rate": 8.410683461317681e-06, + "loss": 2.6143, + "step": 7369 + }, + { + "epoch": 0.94336, + "grad_norm": 0.39207983016967773, + "learning_rate": 8.372861049708858e-06, + "loss": 2.5394, + "step": 7370 + }, + { + "epoch": 0.943488, + "grad_norm": 0.4646625518798828, + "learning_rate": 8.335123154504687e-06, + "loss": 2.6268, + "step": 7371 + }, + { + "epoch": 0.943616, + "grad_norm": 0.40788304805755615, + "learning_rate": 8.297469782192702e-06, + "loss": 2.5934, + "step": 7372 + }, + { + "epoch": 0.943744, + "grad_norm": 0.3974551558494568, + "learning_rate": 8.25990093924589e-06, + "loss": 2.7146, + "step": 7373 + }, + { + "epoch": 0.943872, + "grad_norm": 0.409431517124176, + "learning_rate": 8.222416632122865e-06, + "loss": 2.4886, + "step": 7374 + }, + { + "epoch": 0.944, + "grad_norm": 0.40804290771484375, + "learning_rate": 8.185016867267691e-06, + "loss": 2.4787, + "step": 7375 + }, + { + "epoch": 0.944128, + "grad_norm": 0.43623384833335876, + "learning_rate": 8.147701651109674e-06, + "loss": 2.642, + "step": 7376 + }, + { + "epoch": 0.944256, + "grad_norm": 0.40936043858528137, + "learning_rate": 8.110470990063789e-06, + "loss": 2.4647, + "step": 7377 + }, + { + "epoch": 0.944384, + "grad_norm": 0.39065447449684143, + "learning_rate": 8.07332489053042e-06, + "loss": 2.6087, + "step": 7378 + }, + { + "epoch": 0.944512, + "grad_norm": 0.40297016501426697, + "learning_rate": 8.036263358895402e-06, + "loss": 2.5414, + "step": 7379 + }, + { + "epoch": 0.94464, + "grad_norm": 0.41840556263923645, + "learning_rate": 7.99928640152997e-06, + "loss": 2.6455, + "step": 7380 + }, + { + "epoch": 0.944768, + "grad_norm": 0.4409067630767822, + "learning_rate": 7.962394024791042e-06, + "loss": 2.5909, + "step": 7381 + }, + { + "epoch": 0.944896, + "grad_norm": 0.4121255874633789, + "learning_rate": 7.925586235020766e-06, + "loss": 2.5838, + "step": 7382 + }, + { + "epoch": 0.945024, + "grad_norm": 0.38883277773857117, + "learning_rate": 7.888863038546801e-06, + "loss": 2.641, + "step": 7383 + }, + { + "epoch": 0.945152, + "grad_norm": 0.4104900062084198, + "learning_rate": 7.852224441682376e-06, + "loss": 2.5689, + "step": 7384 + }, + { + "epoch": 0.94528, + "grad_norm": 0.4174940288066864, + "learning_rate": 7.81567045072601e-06, + "loss": 2.4114, + "step": 7385 + }, + { + "epoch": 0.945408, + "grad_norm": 0.40685224533081055, + "learning_rate": 7.779201071961784e-06, + "loss": 2.6186, + "step": 7386 + }, + { + "epoch": 0.945536, + "grad_norm": 0.43184760212898254, + "learning_rate": 7.742816311659185e-06, + "loss": 2.5929, + "step": 7387 + }, + { + "epoch": 0.945664, + "grad_norm": 0.9431355595588684, + "learning_rate": 7.706516176073208e-06, + "loss": 2.5897, + "step": 7388 + }, + { + "epoch": 0.945792, + "grad_norm": 0.5511864423751831, + "learning_rate": 7.670300671444308e-06, + "loss": 2.471, + "step": 7389 + }, + { + "epoch": 0.94592, + "grad_norm": 0.45242443680763245, + "learning_rate": 7.634169803998226e-06, + "loss": 2.5949, + "step": 7390 + }, + { + "epoch": 0.946048, + "grad_norm": 0.45784449577331543, + "learning_rate": 7.598123579946381e-06, + "loss": 2.536, + "step": 7391 + }, + { + "epoch": 0.946176, + "grad_norm": 0.47127655148506165, + "learning_rate": 7.562162005485485e-06, + "loss": 2.7155, + "step": 7392 + }, + { + "epoch": 0.946304, + "grad_norm": 0.3975053131580353, + "learning_rate": 7.5262850867978124e-06, + "loss": 2.6608, + "step": 7393 + }, + { + "epoch": 0.946432, + "grad_norm": 0.3999113142490387, + "learning_rate": 7.490492830050932e-06, + "loss": 2.6769, + "step": 7394 + }, + { + "epoch": 0.94656, + "grad_norm": 0.45532381534576416, + "learning_rate": 7.454785241398032e-06, + "loss": 2.5453, + "step": 7395 + }, + { + "epoch": 0.946688, + "grad_norm": 0.40079328417778015, + "learning_rate": 7.419162326977591e-06, + "loss": 2.4895, + "step": 7396 + }, + { + "epoch": 0.946816, + "grad_norm": 0.4200393259525299, + "learning_rate": 7.383624092913654e-06, + "loss": 2.5513, + "step": 7397 + }, + { + "epoch": 0.946944, + "grad_norm": 0.3949698209762573, + "learning_rate": 7.348170545315613e-06, + "loss": 2.5347, + "step": 7398 + }, + { + "epoch": 0.947072, + "grad_norm": 0.4416053593158722, + "learning_rate": 7.312801690278425e-06, + "loss": 2.5669, + "step": 7399 + }, + { + "epoch": 0.9472, + "grad_norm": 0.4510653018951416, + "learning_rate": 7.277517533882283e-06, + "loss": 2.5414, + "step": 7400 + }, + { + "epoch": 0.947328, + "grad_norm": 0.44276735186576843, + "learning_rate": 7.242318082193e-06, + "loss": 2.6111, + "step": 7401 + }, + { + "epoch": 0.947456, + "grad_norm": 0.45593729615211487, + "learning_rate": 7.207203341261792e-06, + "loss": 2.6475, + "step": 7402 + }, + { + "epoch": 0.947584, + "grad_norm": 0.4440819025039673, + "learning_rate": 7.172173317125275e-06, + "loss": 2.6295, + "step": 7403 + }, + { + "epoch": 0.947712, + "grad_norm": 0.4182509183883667, + "learning_rate": 7.1372280158055196e-06, + "loss": 2.5524, + "step": 7404 + }, + { + "epoch": 0.94784, + "grad_norm": 0.3802037537097931, + "learning_rate": 7.102367443310054e-06, + "loss": 2.5788, + "step": 7405 + }, + { + "epoch": 0.947968, + "grad_norm": 0.42085155844688416, + "learning_rate": 7.067591605631751e-06, + "loss": 2.521, + "step": 7406 + }, + { + "epoch": 0.948096, + "grad_norm": 0.36734452843666077, + "learning_rate": 7.032900508749051e-06, + "loss": 2.4626, + "step": 7407 + }, + { + "epoch": 0.948224, + "grad_norm": 0.426626980304718, + "learning_rate": 6.998294158625684e-06, + "loss": 2.6137, + "step": 7408 + }, + { + "epoch": 0.948352, + "grad_norm": 0.3879397511482239, + "learning_rate": 6.96377256121089e-06, + "loss": 2.6165, + "step": 7409 + }, + { + "epoch": 0.94848, + "grad_norm": 0.41603636741638184, + "learning_rate": 6.9293357224393675e-06, + "loss": 2.5733, + "step": 7410 + }, + { + "epoch": 0.948608, + "grad_norm": 0.4313593804836273, + "learning_rate": 6.894983648231212e-06, + "loss": 2.6329, + "step": 7411 + }, + { + "epoch": 0.948736, + "grad_norm": 0.4180525541305542, + "learning_rate": 6.86071634449198e-06, + "loss": 2.6072, + "step": 7412 + }, + { + "epoch": 0.948864, + "grad_norm": 0.38967227935791016, + "learning_rate": 6.826533817112512e-06, + "loss": 2.6108, + "step": 7413 + }, + { + "epoch": 0.948992, + "grad_norm": 0.4292363226413727, + "learning_rate": 6.792436071969277e-06, + "loss": 2.6003, + "step": 7414 + }, + { + "epoch": 0.94912, + "grad_norm": 0.40691372752189636, + "learning_rate": 6.758423114924028e-06, + "loss": 2.5855, + "step": 7415 + }, + { + "epoch": 0.949248, + "grad_norm": 0.388538658618927, + "learning_rate": 6.724494951823978e-06, + "loss": 2.4713, + "step": 7416 + }, + { + "epoch": 0.949376, + "grad_norm": 0.4140101373195648, + "learning_rate": 6.6906515885017945e-06, + "loss": 2.6107, + "step": 7417 + }, + { + "epoch": 0.949504, + "grad_norm": 0.3837708830833435, + "learning_rate": 6.656893030775546e-06, + "loss": 2.5945, + "step": 7418 + }, + { + "epoch": 0.949632, + "grad_norm": 0.4404456913471222, + "learning_rate": 6.623219284448645e-06, + "loss": 2.5029, + "step": 7419 + }, + { + "epoch": 0.94976, + "grad_norm": 0.4209921658039093, + "learning_rate": 6.589630355310128e-06, + "loss": 2.4787, + "step": 7420 + }, + { + "epoch": 0.949888, + "grad_norm": 0.417693555355072, + "learning_rate": 6.556126249134209e-06, + "loss": 2.5108, + "step": 7421 + }, + { + "epoch": 0.950016, + "grad_norm": 0.4259869456291199, + "learning_rate": 6.522706971680725e-06, + "loss": 2.6029, + "step": 7422 + }, + { + "epoch": 0.950144, + "grad_norm": 0.4510781764984131, + "learning_rate": 6.489372528694748e-06, + "loss": 2.6406, + "step": 7423 + }, + { + "epoch": 0.950272, + "grad_norm": 0.4215375781059265, + "learning_rate": 6.456122925906971e-06, + "loss": 2.6881, + "step": 7424 + }, + { + "epoch": 0.9504, + "grad_norm": 0.4131709933280945, + "learning_rate": 6.422958169033266e-06, + "loss": 2.6248, + "step": 7425 + }, + { + "epoch": 0.950528, + "grad_norm": 0.7811562418937683, + "learning_rate": 6.389878263775128e-06, + "loss": 2.4976, + "step": 7426 + }, + { + "epoch": 0.950656, + "grad_norm": 0.3772207796573639, + "learning_rate": 6.3568832158192865e-06, + "loss": 2.5511, + "step": 7427 + }, + { + "epoch": 0.950784, + "grad_norm": 0.38898158073425293, + "learning_rate": 6.323973030838037e-06, + "loss": 2.5964, + "step": 7428 + }, + { + "epoch": 0.950912, + "grad_norm": 0.42359817028045654, + "learning_rate": 6.291147714488965e-06, + "loss": 2.676, + "step": 7429 + }, + { + "epoch": 0.95104, + "grad_norm": 0.4182545840740204, + "learning_rate": 6.258407272415223e-06, + "loss": 2.5747, + "step": 7430 + }, + { + "epoch": 0.951168, + "grad_norm": 0.4069494605064392, + "learning_rate": 6.2257517102451975e-06, + "loss": 2.5156, + "step": 7431 + }, + { + "epoch": 0.951296, + "grad_norm": 0.4123496115207672, + "learning_rate": 6.193181033592787e-06, + "loss": 2.5798, + "step": 7432 + }, + { + "epoch": 0.951424, + "grad_norm": 0.41443145275115967, + "learning_rate": 6.1606952480572356e-06, + "loss": 2.508, + "step": 7433 + }, + { + "epoch": 0.951552, + "grad_norm": 0.3969399034976959, + "learning_rate": 6.128294359223296e-06, + "loss": 2.5092, + "step": 7434 + }, + { + "epoch": 0.95168, + "grad_norm": 0.38661834597587585, + "learning_rate": 6.09597837266096e-06, + "loss": 2.6024, + "step": 7435 + }, + { + "epoch": 0.951808, + "grad_norm": 0.40505123138427734, + "learning_rate": 6.063747293925781e-06, + "loss": 2.5106, + "step": 7436 + }, + { + "epoch": 0.951936, + "grad_norm": 0.3996022343635559, + "learning_rate": 6.031601128558606e-06, + "loss": 2.5285, + "step": 7437 + }, + { + "epoch": 0.952064, + "grad_norm": 0.457783967256546, + "learning_rate": 5.999539882085791e-06, + "loss": 2.5393, + "step": 7438 + }, + { + "epoch": 0.952192, + "grad_norm": 0.40005365014076233, + "learning_rate": 5.967563560019041e-06, + "loss": 2.789, + "step": 7439 + }, + { + "epoch": 0.95232, + "grad_norm": 0.6548488736152649, + "learning_rate": 5.935672167855455e-06, + "loss": 2.5145, + "step": 7440 + }, + { + "epoch": 0.952448, + "grad_norm": 0.40430691838264465, + "learning_rate": 5.903865711077483e-06, + "loss": 2.6346, + "step": 7441 + }, + { + "epoch": 0.952576, + "grad_norm": 0.4722895920276642, + "learning_rate": 5.872144195153029e-06, + "loss": 2.5711, + "step": 7442 + }, + { + "epoch": 0.952704, + "grad_norm": 0.40410324931144714, + "learning_rate": 5.840507625535396e-06, + "loss": 2.6585, + "step": 7443 + }, + { + "epoch": 0.952832, + "grad_norm": 0.41587501764297485, + "learning_rate": 5.808956007663291e-06, + "loss": 2.5779, + "step": 7444 + }, + { + "epoch": 0.95296, + "grad_norm": 0.44112247228622437, + "learning_rate": 5.777489346960874e-06, + "loss": 2.5509, + "step": 7445 + }, + { + "epoch": 0.953088, + "grad_norm": 0.38403207063674927, + "learning_rate": 5.746107648837484e-06, + "loss": 2.6084, + "step": 7446 + }, + { + "epoch": 0.953216, + "grad_norm": 0.3847346007823944, + "learning_rate": 5.714810918688085e-06, + "loss": 2.53, + "step": 7447 + }, + { + "epoch": 0.953344, + "grad_norm": 0.38632306456565857, + "learning_rate": 5.683599161892928e-06, + "loss": 2.5492, + "step": 7448 + }, + { + "epoch": 0.953472, + "grad_norm": 0.5180634260177612, + "learning_rate": 5.65247238381772e-06, + "loss": 2.5046, + "step": 7449 + }, + { + "epoch": 0.9536, + "grad_norm": 0.5995205044746399, + "learning_rate": 5.621430589813459e-06, + "loss": 2.7147, + "step": 7450 + }, + { + "epoch": 0.953728, + "grad_norm": 0.4032134413719177, + "learning_rate": 5.590473785216654e-06, + "loss": 2.4996, + "step": 7451 + }, + { + "epoch": 0.953856, + "grad_norm": 0.4551421105861664, + "learning_rate": 5.559601975348994e-06, + "loss": 2.5588, + "step": 7452 + }, + { + "epoch": 0.953984, + "grad_norm": 0.37471768260002136, + "learning_rate": 5.528815165517842e-06, + "loss": 2.5665, + "step": 7453 + }, + { + "epoch": 0.954112, + "grad_norm": 0.4299696981906891, + "learning_rate": 5.498113361015799e-06, + "loss": 2.5701, + "step": 7454 + }, + { + "epoch": 0.95424, + "grad_norm": 0.4367274045944214, + "learning_rate": 5.467496567120811e-06, + "loss": 2.6245, + "step": 7455 + }, + { + "epoch": 0.954368, + "grad_norm": 0.44042113423347473, + "learning_rate": 5.436964789096221e-06, + "loss": 2.5254, + "step": 7456 + }, + { + "epoch": 0.954496, + "grad_norm": 0.41162917017936707, + "learning_rate": 5.406518032190943e-06, + "loss": 2.5586, + "step": 7457 + }, + { + "epoch": 0.954624, + "grad_norm": 0.39941108226776123, + "learning_rate": 5.376156301638957e-06, + "loss": 2.5462, + "step": 7458 + }, + { + "epoch": 0.954752, + "grad_norm": 0.4423048198223114, + "learning_rate": 5.34587960265992e-06, + "loss": 2.5706, + "step": 7459 + }, + { + "epoch": 0.95488, + "grad_norm": 0.42700108885765076, + "learning_rate": 5.315687940458669e-06, + "loss": 2.5123, + "step": 7460 + }, + { + "epoch": 0.955008, + "grad_norm": 0.418856680393219, + "learning_rate": 5.285581320225552e-06, + "loss": 2.4962, + "step": 7461 + }, + { + "epoch": 0.955136, + "grad_norm": 0.4519667625427246, + "learning_rate": 5.255559747136263e-06, + "loss": 2.5312, + "step": 7462 + }, + { + "epoch": 0.955264, + "grad_norm": 0.43154242634773254, + "learning_rate": 5.2256232263517835e-06, + "loss": 2.5548, + "step": 7463 + }, + { + "epoch": 0.955392, + "grad_norm": 0.38657546043395996, + "learning_rate": 5.195771763018609e-06, + "loss": 2.5681, + "step": 7464 + }, + { + "epoch": 0.95552, + "grad_norm": 0.4735799729824066, + "learning_rate": 5.166005362268467e-06, + "loss": 2.5076, + "step": 7465 + }, + { + "epoch": 0.955648, + "grad_norm": 0.44894692301750183, + "learning_rate": 5.136324029218653e-06, + "loss": 2.6287, + "step": 7466 + }, + { + "epoch": 0.955776, + "grad_norm": 0.40836286544799805, + "learning_rate": 5.106727768971697e-06, + "loss": 2.4821, + "step": 7467 + }, + { + "epoch": 0.955904, + "grad_norm": 0.4501597583293915, + "learning_rate": 5.07721658661553e-06, + "loss": 2.5923, + "step": 7468 + }, + { + "epoch": 0.956032, + "grad_norm": 0.5756939649581909, + "learning_rate": 5.04779048722348e-06, + "loss": 2.6724, + "step": 7469 + }, + { + "epoch": 0.95616, + "grad_norm": 0.4361560344696045, + "learning_rate": 5.018449475854225e-06, + "loss": 2.5928, + "step": 7470 + }, + { + "epoch": 0.956288, + "grad_norm": 0.36855873465538025, + "learning_rate": 4.989193557551786e-06, + "loss": 2.5883, + "step": 7471 + }, + { + "epoch": 0.956416, + "grad_norm": 0.4002513587474823, + "learning_rate": 4.960022737345693e-06, + "loss": 2.6461, + "step": 7472 + }, + { + "epoch": 0.956544, + "grad_norm": 0.40370407700538635, + "learning_rate": 4.930937020250603e-06, + "loss": 2.619, + "step": 7473 + }, + { + "epoch": 0.956672, + "grad_norm": 0.408393919467926, + "learning_rate": 4.90193641126685e-06, + "loss": 2.6096, + "step": 7474 + }, + { + "epoch": 0.9568, + "grad_norm": 0.45503541827201843, + "learning_rate": 4.873020915379834e-06, + "loss": 2.5714, + "step": 7475 + }, + { + "epoch": 0.956928, + "grad_norm": 0.7753427624702454, + "learning_rate": 4.8441905375605774e-06, + "loss": 2.6459, + "step": 7476 + }, + { + "epoch": 0.957056, + "grad_norm": 0.49220314621925354, + "learning_rate": 4.815445282765285e-06, + "loss": 2.542, + "step": 7477 + }, + { + "epoch": 0.957184, + "grad_norm": 0.4129416346549988, + "learning_rate": 4.786785155935669e-06, + "loss": 2.5966, + "step": 7478 + }, + { + "epoch": 0.957312, + "grad_norm": 0.4524403512477875, + "learning_rate": 4.758210161998622e-06, + "loss": 2.5659, + "step": 7479 + }, + { + "epoch": 0.95744, + "grad_norm": 0.4526710510253906, + "learning_rate": 4.729720305866603e-06, + "loss": 2.5189, + "step": 7480 + }, + { + "epoch": 0.957568, + "grad_norm": 0.4111582338809967, + "learning_rate": 4.70131559243736e-06, + "loss": 2.6254, + "step": 7481 + }, + { + "epoch": 0.957696, + "grad_norm": 0.41709357500076294, + "learning_rate": 4.672996026593934e-06, + "loss": 2.5315, + "step": 7482 + }, + { + "epoch": 0.957824, + "grad_norm": 0.45707786083221436, + "learning_rate": 4.644761613204817e-06, + "loss": 2.5191, + "step": 7483 + }, + { + "epoch": 0.957952, + "grad_norm": 0.42622917890548706, + "learning_rate": 4.616612357123795e-06, + "loss": 2.6089, + "step": 7484 + }, + { + "epoch": 0.95808, + "grad_norm": 0.4243945777416229, + "learning_rate": 4.588548263190107e-06, + "loss": 2.6025, + "step": 7485 + }, + { + "epoch": 0.958208, + "grad_norm": 0.386727899312973, + "learning_rate": 4.560569336228337e-06, + "loss": 2.5031, + "step": 7486 + }, + { + "epoch": 0.958336, + "grad_norm": 0.4000343382358551, + "learning_rate": 4.532675581048251e-06, + "loss": 2.552, + "step": 7487 + }, + { + "epoch": 0.958464, + "grad_norm": 0.3857247829437256, + "learning_rate": 4.504867002445179e-06, + "loss": 2.4875, + "step": 7488 + }, + { + "epoch": 0.958592, + "grad_norm": 0.4140677750110626, + "learning_rate": 4.477143605199796e-06, + "loss": 2.6346, + "step": 7489 + }, + { + "epoch": 0.95872, + "grad_norm": 0.4734921455383301, + "learning_rate": 4.449505394078013e-06, + "loss": 2.5362, + "step": 7490 + }, + { + "epoch": 0.958848, + "grad_norm": 0.42030689120292664, + "learning_rate": 4.421952373831139e-06, + "loss": 2.4874, + "step": 7491 + }, + { + "epoch": 0.958976, + "grad_norm": 0.3927699029445648, + "learning_rate": 4.394484549195887e-06, + "loss": 2.5995, + "step": 7492 + }, + { + "epoch": 0.959104, + "grad_norm": 0.4467799961566925, + "learning_rate": 4.3671019248943125e-06, + "loss": 2.6947, + "step": 7493 + }, + { + "epoch": 0.959232, + "grad_norm": 0.4297340512275696, + "learning_rate": 4.3398045056337594e-06, + "loss": 2.4824, + "step": 7494 + }, + { + "epoch": 0.95936, + "grad_norm": 0.4193342924118042, + "learning_rate": 4.312592296106976e-06, + "loss": 2.6046, + "step": 7495 + }, + { + "epoch": 0.959488, + "grad_norm": 0.4486200511455536, + "learning_rate": 4.285465300992164e-06, + "loss": 2.4908, + "step": 7496 + }, + { + "epoch": 0.959616, + "grad_norm": 0.4247700572013855, + "learning_rate": 4.258423524952649e-06, + "loss": 2.5564, + "step": 7497 + }, + { + "epoch": 0.959744, + "grad_norm": 0.41457831859588623, + "learning_rate": 4.23146697263721e-06, + "loss": 2.4742, + "step": 7498 + }, + { + "epoch": 0.959872, + "grad_norm": 0.4399730861186981, + "learning_rate": 4.204595648680088e-06, + "loss": 2.515, + "step": 7499 + }, + { + "epoch": 0.96, + "grad_norm": 0.4235430061817169, + "learning_rate": 4.177809557700751e-06, + "loss": 2.5643, + "step": 7500 + }, + { + "epoch": 0.960128, + "grad_norm": 0.43129876255989075, + "learning_rate": 4.151108704303963e-06, + "loss": 2.5947, + "step": 7501 + }, + { + "epoch": 0.960256, + "grad_norm": 0.3935773968696594, + "learning_rate": 4.12449309307994e-06, + "loss": 2.6288, + "step": 7502 + }, + { + "epoch": 0.960384, + "grad_norm": 0.379271000623703, + "learning_rate": 4.097962728604299e-06, + "loss": 2.537, + "step": 7503 + }, + { + "epoch": 0.960512, + "grad_norm": 0.5057717561721802, + "learning_rate": 4.071517615437836e-06, + "loss": 2.613, + "step": 7504 + }, + { + "epoch": 0.96064, + "grad_norm": 0.4207730293273926, + "learning_rate": 4.045157758126749e-06, + "loss": 2.5377, + "step": 7505 + }, + { + "epoch": 0.960768, + "grad_norm": 0.43355464935302734, + "learning_rate": 4.018883161202691e-06, + "loss": 2.7616, + "step": 7506 + }, + { + "epoch": 0.960896, + "grad_norm": 0.4981098175048828, + "learning_rate": 3.992693829182548e-06, + "loss": 2.5313, + "step": 7507 + }, + { + "epoch": 0.961024, + "grad_norm": 0.4136715531349182, + "learning_rate": 3.966589766568496e-06, + "loss": 2.6173, + "step": 7508 + }, + { + "epoch": 0.961152, + "grad_norm": 0.409449964761734, + "learning_rate": 3.940570977848168e-06, + "loss": 2.7487, + "step": 7509 + }, + { + "epoch": 0.96128, + "grad_norm": 0.4522875249385834, + "learning_rate": 3.914637467494542e-06, + "loss": 2.5711, + "step": 7510 + }, + { + "epoch": 0.961408, + "grad_norm": 0.41443565487861633, + "learning_rate": 3.888789239965884e-06, + "loss": 2.5145, + "step": 7511 + }, + { + "epoch": 0.961536, + "grad_norm": 0.4519394040107727, + "learning_rate": 3.863026299705696e-06, + "loss": 2.4542, + "step": 7512 + }, + { + "epoch": 0.961664, + "grad_norm": 0.4138665497303009, + "learning_rate": 3.837348651142991e-06, + "loss": 2.5089, + "step": 7513 + }, + { + "epoch": 0.961792, + "grad_norm": 0.4067786633968353, + "learning_rate": 3.811756298692126e-06, + "loss": 2.5706, + "step": 7514 + }, + { + "epoch": 0.96192, + "grad_norm": 0.45242324471473694, + "learning_rate": 3.786249246752638e-06, + "loss": 2.5699, + "step": 7515 + }, + { + "epoch": 0.962048, + "grad_norm": 0.4086751639842987, + "learning_rate": 3.7608274997095184e-06, + "loss": 2.5347, + "step": 7516 + }, + { + "epoch": 0.962176, + "grad_norm": 0.4364364445209503, + "learning_rate": 3.7354910619329942e-06, + "loss": 2.5838, + "step": 7517 + }, + { + "epoch": 0.962304, + "grad_norm": 0.41598188877105713, + "learning_rate": 3.710239937778803e-06, + "loss": 2.6167, + "step": 7518 + }, + { + "epoch": 0.962432, + "grad_norm": 0.4109190106391907, + "learning_rate": 3.685074131587862e-06, + "loss": 2.5141, + "step": 7519 + }, + { + "epoch": 0.96256, + "grad_norm": 0.40340539813041687, + "learning_rate": 3.6599936476864326e-06, + "loss": 2.5945, + "step": 7520 + }, + { + "epoch": 0.962688, + "grad_norm": 0.3834770917892456, + "learning_rate": 3.634998490386121e-06, + "loss": 2.6029, + "step": 7521 + }, + { + "epoch": 0.962816, + "grad_norm": 0.38975247740745544, + "learning_rate": 3.61008866398399e-06, + "loss": 2.6234, + "step": 7522 + }, + { + "epoch": 0.962944, + "grad_norm": 0.42569485306739807, + "learning_rate": 3.5852641727622257e-06, + "loss": 2.5023, + "step": 7523 + }, + { + "epoch": 0.963072, + "grad_norm": 0.39874956011772156, + "learning_rate": 3.5605250209885253e-06, + "loss": 2.5234, + "step": 7524 + }, + { + "epoch": 0.9632, + "grad_norm": 0.40695804357528687, + "learning_rate": 3.5358712129157642e-06, + "loss": 2.5393, + "step": 7525 + }, + { + "epoch": 0.963328, + "grad_norm": 0.4179682433605194, + "learning_rate": 3.5113027527822192e-06, + "loss": 2.5262, + "step": 7526 + }, + { + "epoch": 0.963456, + "grad_norm": 0.4181496500968933, + "learning_rate": 3.4868196448115676e-06, + "loss": 2.556, + "step": 7527 + }, + { + "epoch": 0.963584, + "grad_norm": 0.4088679552078247, + "learning_rate": 3.4624218932127194e-06, + "loss": 2.4935, + "step": 7528 + }, + { + "epoch": 0.963712, + "grad_norm": 0.444168359041214, + "learning_rate": 3.43810950217982e-06, + "loss": 2.6677, + "step": 7529 + }, + { + "epoch": 0.96384, + "grad_norm": 0.4424348771572113, + "learning_rate": 3.4138824758925822e-06, + "loss": 2.5095, + "step": 7530 + }, + { + "epoch": 0.963968, + "grad_norm": 0.4290226697921753, + "learning_rate": 3.389740818515841e-06, + "loss": 2.5648, + "step": 7531 + }, + { + "epoch": 0.964096, + "grad_norm": 0.40473493933677673, + "learning_rate": 3.3656845341998865e-06, + "loss": 2.5375, + "step": 7532 + }, + { + "epoch": 0.964224, + "grad_norm": 0.4235174357891083, + "learning_rate": 3.341713627080245e-06, + "loss": 2.5563, + "step": 7533 + }, + { + "epoch": 0.964352, + "grad_norm": 0.38970229029655457, + "learning_rate": 3.3178281012777866e-06, + "loss": 2.604, + "step": 7534 + }, + { + "epoch": 0.96448, + "grad_norm": 0.4420566260814667, + "learning_rate": 3.2940279608986714e-06, + "loss": 2.6947, + "step": 7535 + }, + { + "epoch": 0.964608, + "grad_norm": 0.4145389497280121, + "learning_rate": 3.2703132100345146e-06, + "loss": 2.5237, + "step": 7536 + }, + { + "epoch": 0.964736, + "grad_norm": 0.39020127058029175, + "learning_rate": 3.2466838527620557e-06, + "loss": 2.5145, + "step": 7537 + }, + { + "epoch": 0.964864, + "grad_norm": 0.405472993850708, + "learning_rate": 3.22313989314349e-06, + "loss": 2.5691, + "step": 7538 + }, + { + "epoch": 0.964992, + "grad_norm": 0.5110159516334534, + "learning_rate": 3.1996813352263566e-06, + "loss": 2.6045, + "step": 7539 + }, + { + "epoch": 0.96512, + "grad_norm": 0.4119569659233093, + "learning_rate": 3.176308183043375e-06, + "loss": 2.386, + "step": 7540 + }, + { + "epoch": 0.965248, + "grad_norm": 0.4300902783870697, + "learning_rate": 3.1530204406127192e-06, + "loss": 2.5549, + "step": 7541 + }, + { + "epoch": 0.965376, + "grad_norm": 0.3879724144935608, + "learning_rate": 3.1298181119377432e-06, + "loss": 2.5365, + "step": 7542 + }, + { + "epoch": 0.965504, + "grad_norm": 0.4030568599700928, + "learning_rate": 3.1067012010073114e-06, + "loss": 2.6601, + "step": 7543 + }, + { + "epoch": 0.965632, + "grad_norm": 0.41895174980163574, + "learning_rate": 3.0836697117954115e-06, + "loss": 2.5436, + "step": 7544 + }, + { + "epoch": 0.96576, + "grad_norm": 0.44717803597450256, + "learning_rate": 3.0607236482613766e-06, + "loss": 2.5336, + "step": 7545 + }, + { + "epoch": 0.965888, + "grad_norm": 0.4777510464191437, + "learning_rate": 3.0378630143499953e-06, + "loss": 2.5239, + "step": 7546 + }, + { + "epoch": 0.966016, + "grad_norm": 0.3865121901035309, + "learning_rate": 3.0150878139912906e-06, + "loss": 2.4706, + "step": 7547 + }, + { + "epoch": 0.966144, + "grad_norm": 0.43048495054244995, + "learning_rate": 2.9923980511004646e-06, + "loss": 2.5814, + "step": 7548 + }, + { + "epoch": 0.966272, + "grad_norm": 0.4510347545146942, + "learning_rate": 2.9697937295782295e-06, + "loss": 2.6045, + "step": 7549 + }, + { + "epoch": 0.9664, + "grad_norm": 0.47126853466033936, + "learning_rate": 2.9472748533105886e-06, + "loss": 2.5059, + "step": 7550 + }, + { + "epoch": 0.966528, + "grad_norm": 0.47745025157928467, + "learning_rate": 2.924841426168667e-06, + "loss": 2.5429, + "step": 7551 + }, + { + "epoch": 0.966656, + "grad_norm": 0.38515186309814453, + "learning_rate": 2.9024934520091028e-06, + "loss": 2.6064, + "step": 7552 + }, + { + "epoch": 0.966784, + "grad_norm": 0.40486273169517517, + "learning_rate": 2.8802309346737666e-06, + "loss": 2.68, + "step": 7553 + }, + { + "epoch": 0.966912, + "grad_norm": 0.42380794882774353, + "learning_rate": 2.858053877989819e-06, + "loss": 2.5442, + "step": 7554 + }, + { + "epoch": 0.96704, + "grad_norm": 0.4022480845451355, + "learning_rate": 2.835962285769822e-06, + "loss": 2.5879, + "step": 7555 + }, + { + "epoch": 0.967168, + "grad_norm": 0.3881256580352783, + "learning_rate": 2.813956161811515e-06, + "loss": 2.5463, + "step": 7556 + }, + { + "epoch": 0.967296, + "grad_norm": 0.3859476149082184, + "learning_rate": 2.7920355098979833e-06, + "loss": 2.6705, + "step": 7557 + }, + { + "epoch": 0.967424, + "grad_norm": 0.40897300839424133, + "learning_rate": 2.7702003337977123e-06, + "loss": 2.6134, + "step": 7558 + }, + { + "epoch": 0.967552, + "grad_norm": 0.4170248210430145, + "learning_rate": 2.7484506372643658e-06, + "loss": 2.5625, + "step": 7559 + }, + { + "epoch": 0.96768, + "grad_norm": 0.40354302525520325, + "learning_rate": 2.7267864240369533e-06, + "loss": 2.6491, + "step": 7560 + }, + { + "epoch": 0.967808, + "grad_norm": 0.4332180917263031, + "learning_rate": 2.7052076978398842e-06, + "loss": 2.6001, + "step": 7561 + }, + { + "epoch": 0.967936, + "grad_norm": 0.4034114480018616, + "learning_rate": 2.683714462382747e-06, + "loss": 2.6007, + "step": 7562 + }, + { + "epoch": 0.968064, + "grad_norm": 0.4232821762561798, + "learning_rate": 2.6623067213605303e-06, + "loss": 2.6059, + "step": 7563 + }, + { + "epoch": 0.968192, + "grad_norm": 0.4318840503692627, + "learning_rate": 2.6409844784533456e-06, + "loss": 2.5534, + "step": 7564 + }, + { + "epoch": 0.96832, + "grad_norm": 0.4028768241405487, + "learning_rate": 2.619747737326872e-06, + "loss": 2.5998, + "step": 7565 + }, + { + "epoch": 0.968448, + "grad_norm": 0.4226822853088379, + "learning_rate": 2.5985965016318e-06, + "loss": 2.5917, + "step": 7566 + }, + { + "epoch": 0.968576, + "grad_norm": 0.39708036184310913, + "learning_rate": 2.5775307750043887e-06, + "loss": 2.6534, + "step": 7567 + }, + { + "epoch": 0.968704, + "grad_norm": 0.39461594820022583, + "learning_rate": 2.556550561066073e-06, + "loss": 2.5348, + "step": 7568 + }, + { + "epoch": 0.968832, + "grad_norm": 0.42343178391456604, + "learning_rate": 2.5356558634235806e-06, + "loss": 2.4889, + "step": 7569 + }, + { + "epoch": 0.96896, + "grad_norm": 0.41625237464904785, + "learning_rate": 2.514846685668981e-06, + "loss": 2.6841, + "step": 7570 + }, + { + "epoch": 0.969088, + "grad_norm": 0.4002101719379425, + "learning_rate": 2.4941230313795247e-06, + "loss": 2.6005, + "step": 7571 + }, + { + "epoch": 0.969216, + "grad_norm": 0.45506608486175537, + "learning_rate": 2.4734849041179176e-06, + "loss": 2.6362, + "step": 7572 + }, + { + "epoch": 0.969344, + "grad_norm": 0.5337355732917786, + "learning_rate": 2.4529323074320984e-06, + "loss": 2.5513, + "step": 7573 + }, + { + "epoch": 0.969472, + "grad_norm": 0.4344041049480438, + "learning_rate": 2.432465244855242e-06, + "loss": 2.6117, + "step": 7574 + }, + { + "epoch": 0.9696, + "grad_norm": 0.4113321304321289, + "learning_rate": 2.412083719905922e-06, + "loss": 2.6083, + "step": 7575 + }, + { + "epoch": 0.969728, + "grad_norm": 0.4304996132850647, + "learning_rate": 2.391787736087947e-06, + "loss": 2.4752, + "step": 7576 + }, + { + "epoch": 0.969856, + "grad_norm": 0.37941327691078186, + "learning_rate": 2.3715772968904147e-06, + "loss": 2.549, + "step": 7577 + }, + { + "epoch": 0.969984, + "grad_norm": 0.4010094702243805, + "learning_rate": 2.3514524057877685e-06, + "loss": 2.5953, + "step": 7578 + }, + { + "epoch": 0.970112, + "grad_norm": 0.3828171193599701, + "learning_rate": 2.331413066239685e-06, + "loss": 2.4595, + "step": 7579 + }, + { + "epoch": 0.97024, + "grad_norm": 0.4162217378616333, + "learning_rate": 2.3114592816911863e-06, + "loss": 2.5858, + "step": 7580 + }, + { + "epoch": 0.970368, + "grad_norm": 0.9453883767127991, + "learning_rate": 2.2915910555725283e-06, + "loss": 2.6033, + "step": 7581 + }, + { + "epoch": 0.970496, + "grad_norm": 0.46689698100090027, + "learning_rate": 2.2718083912992572e-06, + "loss": 2.5128, + "step": 7582 + }, + { + "epoch": 0.970624, + "grad_norm": 0.3702486455440521, + "learning_rate": 2.2521112922723184e-06, + "loss": 2.601, + "step": 7583 + }, + { + "epoch": 0.970752, + "grad_norm": 0.42867031693458557, + "learning_rate": 2.2324997618778375e-06, + "loss": 2.5585, + "step": 7584 + }, + { + "epoch": 0.97088, + "grad_norm": 0.40826404094696045, + "learning_rate": 2.212973803487284e-06, + "loss": 2.5837, + "step": 7585 + }, + { + "epoch": 0.971008, + "grad_norm": 0.3799974024295807, + "learning_rate": 2.1935334204573054e-06, + "loss": 2.5389, + "step": 7586 + }, + { + "epoch": 0.971136, + "grad_norm": 0.43944844603538513, + "learning_rate": 2.1741786161300626e-06, + "loss": 2.4713, + "step": 7587 + }, + { + "epoch": 0.971264, + "grad_norm": 0.40445923805236816, + "learning_rate": 2.1549093938327825e-06, + "loss": 2.5617, + "step": 7588 + }, + { + "epoch": 0.971392, + "grad_norm": 0.3911871612071991, + "learning_rate": 2.135725756878093e-06, + "loss": 2.6172, + "step": 7589 + }, + { + "epoch": 0.97152, + "grad_norm": 0.3939182758331299, + "learning_rate": 2.1166277085639117e-06, + "loss": 2.6159, + "step": 7590 + }, + { + "epoch": 0.971648, + "grad_norm": 0.3966546952724457, + "learning_rate": 2.0976152521733905e-06, + "loss": 2.6196, + "step": 7591 + }, + { + "epoch": 0.971776, + "grad_norm": 0.4972662627696991, + "learning_rate": 2.0786883909749698e-06, + "loss": 2.6254, + "step": 7592 + }, + { + "epoch": 0.971904, + "grad_norm": 0.4093163311481476, + "learning_rate": 2.059847128222381e-06, + "loss": 2.5547, + "step": 7593 + }, + { + "epoch": 0.972032, + "grad_norm": 0.4084579050540924, + "learning_rate": 2.0410914671546986e-06, + "loss": 2.6226, + "step": 7594 + }, + { + "epoch": 0.97216, + "grad_norm": 0.4025396704673767, + "learning_rate": 2.0224214109962334e-06, + "loss": 2.5702, + "step": 7595 + }, + { + "epoch": 0.972288, + "grad_norm": 0.4274108111858368, + "learning_rate": 2.0038369629565846e-06, + "loss": 2.5836, + "step": 7596 + }, + { + "epoch": 0.972416, + "grad_norm": 0.3959181010723114, + "learning_rate": 1.98533812623064e-06, + "loss": 2.6656, + "step": 7597 + }, + { + "epoch": 0.972544, + "grad_norm": 0.4176934063434601, + "learning_rate": 1.9669249039985794e-06, + "loss": 2.5591, + "step": 7598 + }, + { + "epoch": 0.972672, + "grad_norm": 0.42159682512283325, + "learning_rate": 1.9485972994257585e-06, + "loss": 2.5756, + "step": 7599 + }, + { + "epoch": 0.9728, + "grad_norm": 0.41468843817710876, + "learning_rate": 1.930355315663046e-06, + "loss": 2.4446, + "step": 7600 + }, + { + "epoch": 0.972928, + "grad_norm": 0.4367999732494354, + "learning_rate": 1.9121989558463225e-06, + "loss": 2.5568, + "step": 7601 + }, + { + "epoch": 0.973056, + "grad_norm": 0.40781354904174805, + "learning_rate": 1.8941282230969247e-06, + "loss": 2.5735, + "step": 7602 + }, + { + "epoch": 0.973184, + "grad_norm": 0.4281311333179474, + "learning_rate": 1.876143120521423e-06, + "loss": 2.562, + "step": 7603 + }, + { + "epoch": 0.973312, + "grad_norm": 0.43072745203971863, + "learning_rate": 1.8582436512116775e-06, + "loss": 2.6725, + "step": 7604 + }, + { + "epoch": 0.97344, + "grad_norm": 0.3937316834926605, + "learning_rate": 1.840429818244782e-06, + "loss": 2.4102, + "step": 7605 + }, + { + "epoch": 0.973568, + "grad_norm": 0.4085983335971832, + "learning_rate": 1.8227016246831762e-06, + "loss": 2.4809, + "step": 7606 + }, + { + "epoch": 0.973696, + "grad_norm": 0.4226935803890228, + "learning_rate": 1.805059073574533e-06, + "loss": 2.6111, + "step": 7607 + }, + { + "epoch": 0.973824, + "grad_norm": 0.42723459005355835, + "learning_rate": 1.7875021679518155e-06, + "loss": 2.5079, + "step": 7608 + }, + { + "epoch": 0.973952, + "grad_norm": 0.3943471908569336, + "learning_rate": 1.7700309108332203e-06, + "loss": 2.5294, + "step": 7609 + }, + { + "epoch": 0.97408, + "grad_norm": 0.43922796845436096, + "learning_rate": 1.7526453052223445e-06, + "loss": 2.5237, + "step": 7610 + }, + { + "epoch": 0.974208, + "grad_norm": 0.45675045251846313, + "learning_rate": 1.7353453541078535e-06, + "loss": 2.6084, + "step": 7611 + }, + { + "epoch": 0.974336, + "grad_norm": 0.39316803216934204, + "learning_rate": 1.7181310604639233e-06, + "loss": 2.5364, + "step": 7612 + }, + { + "epoch": 0.974464, + "grad_norm": 0.39234259724617004, + "learning_rate": 1.7010024272498537e-06, + "loss": 2.5756, + "step": 7613 + }, + { + "epoch": 0.974592, + "grad_norm": 0.38692614436149597, + "learning_rate": 1.683959457410178e-06, + "loss": 2.5206, + "step": 7614 + }, + { + "epoch": 0.97472, + "grad_norm": 0.4014062285423279, + "learning_rate": 1.6670021538748859e-06, + "loss": 2.6013, + "step": 7615 + }, + { + "epoch": 0.974848, + "grad_norm": 0.47871342301368713, + "learning_rate": 1.65013051955909e-06, + "loss": 2.5532, + "step": 7616 + }, + { + "epoch": 0.974976, + "grad_norm": 0.6936469674110413, + "learning_rate": 1.6333445573632478e-06, + "loss": 2.6053, + "step": 7617 + }, + { + "epoch": 0.975104, + "grad_norm": 0.593148946762085, + "learning_rate": 1.6166442701730511e-06, + "loss": 2.5178, + "step": 7618 + }, + { + "epoch": 0.975232, + "grad_norm": 0.42703285813331604, + "learning_rate": 1.6000296608594256e-06, + "loss": 2.5967, + "step": 7619 + }, + { + "epoch": 0.97536, + "grad_norm": 0.6202399134635925, + "learning_rate": 1.5835007322786421e-06, + "loss": 2.5458, + "step": 7620 + }, + { + "epoch": 0.975488, + "grad_norm": 0.434815376996994, + "learning_rate": 1.567057487272261e-06, + "loss": 2.5949, + "step": 7621 + }, + { + "epoch": 0.975616, + "grad_norm": 0.412482887506485, + "learning_rate": 1.5506999286669654e-06, + "loss": 2.5421, + "step": 7622 + }, + { + "epoch": 0.975744, + "grad_norm": 0.5819879174232483, + "learning_rate": 1.534428059274895e-06, + "loss": 2.4786, + "step": 7623 + }, + { + "epoch": 0.975872, + "grad_norm": 0.4113178253173828, + "learning_rate": 1.5182418818933675e-06, + "loss": 2.5388, + "step": 7624 + }, + { + "epoch": 0.976, + "grad_norm": 0.42280441522598267, + "learning_rate": 1.502141399304935e-06, + "loss": 2.5608, + "step": 7625 + }, + { + "epoch": 0.976128, + "grad_norm": 0.3922262489795685, + "learning_rate": 1.4861266142775499e-06, + "loss": 2.6533, + "step": 7626 + }, + { + "epoch": 0.976256, + "grad_norm": 0.41943466663360596, + "learning_rate": 1.4701975295641768e-06, + "loss": 2.5934, + "step": 7627 + }, + { + "epoch": 0.976384, + "grad_norm": 0.5183844566345215, + "learning_rate": 1.4543541479033473e-06, + "loss": 2.5034, + "step": 7628 + }, + { + "epoch": 0.976512, + "grad_norm": 0.4098818600177765, + "learning_rate": 1.438596472018716e-06, + "loss": 2.6172, + "step": 7629 + }, + { + "epoch": 0.97664, + "grad_norm": 0.41440823674201965, + "learning_rate": 1.422924504619061e-06, + "loss": 2.4008, + "step": 7630 + }, + { + "epoch": 0.976768, + "grad_norm": 0.42212551832199097, + "learning_rate": 1.4073382483987817e-06, + "loss": 2.6098, + "step": 7631 + }, + { + "epoch": 0.976896, + "grad_norm": 0.4193539619445801, + "learning_rate": 1.39183770603718e-06, + "loss": 2.6308, + "step": 7632 + }, + { + "epoch": 0.977024, + "grad_norm": 0.4363216757774353, + "learning_rate": 1.3764228801990687e-06, + "loss": 2.5685, + "step": 7633 + }, + { + "epoch": 0.977152, + "grad_norm": 0.4579240381717682, + "learning_rate": 1.3610937735344386e-06, + "loss": 2.4833, + "step": 7634 + }, + { + "epoch": 0.97728, + "grad_norm": 0.44959786534309387, + "learning_rate": 1.3458503886784602e-06, + "loss": 2.5956, + "step": 7635 + }, + { + "epoch": 0.977408, + "grad_norm": 0.3996305763721466, + "learning_rate": 1.3306927282517034e-06, + "loss": 2.5393, + "step": 7636 + }, + { + "epoch": 0.977536, + "grad_norm": 0.3845694959163666, + "learning_rate": 1.3156207948599176e-06, + "loss": 2.5237, + "step": 7637 + }, + { + "epoch": 0.977664, + "grad_norm": 0.43127405643463135, + "learning_rate": 1.3006345910941963e-06, + "loss": 2.6086, + "step": 7638 + }, + { + "epoch": 0.977792, + "grad_norm": 0.43851780891418457, + "learning_rate": 1.2857341195308125e-06, + "loss": 2.5689, + "step": 7639 + }, + { + "epoch": 0.97792, + "grad_norm": 0.4074998199939728, + "learning_rate": 1.2709193827312725e-06, + "loss": 2.545, + "step": 7640 + }, + { + "epoch": 0.978048, + "grad_norm": 0.39829012751579285, + "learning_rate": 1.2561903832424836e-06, + "loss": 2.5601, + "step": 7641 + }, + { + "epoch": 0.978176, + "grad_norm": 0.42944851517677307, + "learning_rate": 1.2415471235965315e-06, + "loss": 2.6522, + "step": 7642 + }, + { + "epoch": 0.978304, + "grad_norm": 0.5755578875541687, + "learning_rate": 1.2269896063107356e-06, + "loss": 2.4972, + "step": 7643 + }, + { + "epoch": 0.978432, + "grad_norm": 0.4230700135231018, + "learning_rate": 1.2125178338876497e-06, + "loss": 2.6397, + "step": 7644 + }, + { + "epoch": 0.97856, + "grad_norm": 0.4422700107097626, + "learning_rate": 1.1981318088152281e-06, + "loss": 2.5097, + "step": 7645 + }, + { + "epoch": 0.978688, + "grad_norm": 0.4128256142139435, + "learning_rate": 1.1838315335664928e-06, + "loss": 2.556, + "step": 7646 + }, + { + "epoch": 0.978816, + "grad_norm": 0.40612727403640747, + "learning_rate": 1.169617010599977e-06, + "loss": 2.582, + "step": 7647 + }, + { + "epoch": 0.978944, + "grad_norm": 0.4578205347061157, + "learning_rate": 1.1554882423591707e-06, + "loss": 2.6212, + "step": 7648 + }, + { + "epoch": 0.979072, + "grad_norm": 0.4124743342399597, + "learning_rate": 1.141445231273075e-06, + "loss": 2.5877, + "step": 7649 + }, + { + "epoch": 0.9792, + "grad_norm": 0.42608746886253357, + "learning_rate": 1.1274879797558146e-06, + "loss": 2.5812, + "step": 7650 + }, + { + "epoch": 0.979328, + "grad_norm": 0.40750160813331604, + "learning_rate": 1.113616490206748e-06, + "loss": 2.623, + "step": 7651 + }, + { + "epoch": 0.979456, + "grad_norm": 0.41186824440956116, + "learning_rate": 1.0998307650106342e-06, + "loss": 2.6349, + "step": 7652 + }, + { + "epoch": 0.979584, + "grad_norm": 0.471891313791275, + "learning_rate": 1.0861308065373554e-06, + "loss": 2.5026, + "step": 7653 + }, + { + "epoch": 0.979712, + "grad_norm": 0.4193877875804901, + "learning_rate": 1.0725166171420831e-06, + "loss": 2.5, + "step": 7654 + }, + { + "epoch": 0.97984, + "grad_norm": 0.44056087732315063, + "learning_rate": 1.0589881991652784e-06, + "loss": 2.6395, + "step": 7655 + }, + { + "epoch": 0.979968, + "grad_norm": 0.42842358350753784, + "learning_rate": 1.0455455549326365e-06, + "loss": 2.4727, + "step": 7656 + }, + { + "epoch": 0.980096, + "grad_norm": 0.4461008608341217, + "learning_rate": 1.0321886867550868e-06, + "loss": 2.608, + "step": 7657 + }, + { + "epoch": 0.980224, + "grad_norm": 0.4245521128177643, + "learning_rate": 1.0189175969287924e-06, + "loss": 2.5598, + "step": 7658 + }, + { + "epoch": 0.980352, + "grad_norm": 0.4050993025302887, + "learning_rate": 1.0057322877352615e-06, + "loss": 2.6848, + "step": 7659 + }, + { + "epoch": 0.98048, + "grad_norm": 0.3928740918636322, + "learning_rate": 9.926327614411811e-07, + "loss": 2.5842, + "step": 7660 + }, + { + "epoch": 0.980608, + "grad_norm": 0.9688799977302551, + "learning_rate": 9.796190202985832e-07, + "loss": 2.5576, + "step": 7661 + }, + { + "epoch": 0.980736, + "grad_norm": 0.36992180347442627, + "learning_rate": 9.666910665445672e-07, + "loss": 2.6438, + "step": 7662 + }, + { + "epoch": 0.980864, + "grad_norm": 0.39990776777267456, + "learning_rate": 9.538489024016328e-07, + "loss": 2.5588, + "step": 7663 + }, + { + "epoch": 0.980992, + "grad_norm": 0.4056423008441925, + "learning_rate": 9.410925300775697e-07, + "loss": 2.7129, + "step": 7664 + }, + { + "epoch": 0.98112, + "grad_norm": 0.43105509877204895, + "learning_rate": 9.284219517652349e-07, + "loss": 2.6193, + "step": 7665 + }, + { + "epoch": 0.981248, + "grad_norm": 0.3895041346549988, + "learning_rate": 9.158371696428858e-07, + "loss": 2.5637, + "step": 7666 + }, + { + "epoch": 0.981376, + "grad_norm": 0.4242360591888428, + "learning_rate": 9.033381858740697e-07, + "loss": 2.5937, + "step": 7667 + }, + { + "epoch": 0.981504, + "grad_norm": 0.3951515853404999, + "learning_rate": 8.909250026074012e-07, + "loss": 2.4134, + "step": 7668 + }, + { + "epoch": 0.981632, + "grad_norm": 0.48890045285224915, + "learning_rate": 8.785976219768954e-07, + "loss": 2.6935, + "step": 7669 + }, + { + "epoch": 0.98176, + "grad_norm": 0.6143409609794617, + "learning_rate": 8.663560461018016e-07, + "loss": 2.5858, + "step": 7670 + }, + { + "epoch": 0.981888, + "grad_norm": 0.41634756326675415, + "learning_rate": 8.542002770865476e-07, + "loss": 2.6669, + "step": 7671 + }, + { + "epoch": 0.982016, + "grad_norm": 0.40904489159584045, + "learning_rate": 8.421303170208505e-07, + "loss": 2.4316, + "step": 7672 + }, + { + "epoch": 0.982144, + "grad_norm": 0.3874613046646118, + "learning_rate": 8.301461679796618e-07, + "loss": 2.5575, + "step": 7673 + }, + { + "epoch": 0.982272, + "grad_norm": 0.41015979647636414, + "learning_rate": 8.182478320232223e-07, + "loss": 2.5715, + "step": 7674 + }, + { + "epoch": 0.9824, + "grad_norm": 0.458454430103302, + "learning_rate": 8.064353111970068e-07, + "loss": 2.5154, + "step": 7675 + }, + { + "epoch": 0.982528, + "grad_norm": 0.40367835760116577, + "learning_rate": 7.947086075317245e-07, + "loss": 2.5314, + "step": 7676 + }, + { + "epoch": 0.982656, + "grad_norm": 0.3911985158920288, + "learning_rate": 7.830677230433181e-07, + "loss": 2.4441, + "step": 7677 + }, + { + "epoch": 0.982784, + "grad_norm": 0.4333050847053528, + "learning_rate": 7.715126597329647e-07, + "loss": 2.5223, + "step": 7678 + }, + { + "epoch": 0.982912, + "grad_norm": 0.40235844254493713, + "learning_rate": 7.600434195871864e-07, + "loss": 2.5462, + "step": 7679 + }, + { + "epoch": 0.98304, + "grad_norm": 0.4274805188179016, + "learning_rate": 7.486600045775727e-07, + "loss": 2.5653, + "step": 7680 + }, + { + "epoch": 0.983168, + "grad_norm": 0.3893744945526123, + "learning_rate": 7.373624166611137e-07, + "loss": 2.5392, + "step": 7681 + }, + { + "epoch": 0.983296, + "grad_norm": 0.448949933052063, + "learning_rate": 7.261506577800892e-07, + "loss": 2.628, + "step": 7682 + }, + { + "epoch": 0.983424, + "grad_norm": 0.4743226170539856, + "learning_rate": 7.150247298617907e-07, + "loss": 2.6714, + "step": 7683 + }, + { + "epoch": 0.983552, + "grad_norm": 0.6317682266235352, + "learning_rate": 7.039846348189105e-07, + "loss": 2.6044, + "step": 7684 + }, + { + "epoch": 0.98368, + "grad_norm": 0.40756648778915405, + "learning_rate": 6.930303745494859e-07, + "loss": 2.5857, + "step": 7685 + }, + { + "epoch": 0.983808, + "grad_norm": 0.4294353425502777, + "learning_rate": 6.821619509365662e-07, + "loss": 2.5699, + "step": 7686 + }, + { + "epoch": 0.983936, + "grad_norm": 0.392442911863327, + "learning_rate": 6.713793658486012e-07, + "loss": 2.47, + "step": 7687 + }, + { + "epoch": 0.984064, + "grad_norm": 0.44355887174606323, + "learning_rate": 6.606826211392192e-07, + "loss": 2.5652, + "step": 7688 + }, + { + "epoch": 0.984192, + "grad_norm": 0.5149610638618469, + "learning_rate": 6.500717186473382e-07, + "loss": 2.5883, + "step": 7689 + }, + { + "epoch": 0.98432, + "grad_norm": 0.4410872757434845, + "learning_rate": 6.3954666019711e-07, + "loss": 2.5612, + "step": 7690 + }, + { + "epoch": 0.984448, + "grad_norm": 0.38350918889045715, + "learning_rate": 6.291074475978653e-07, + "loss": 2.4212, + "step": 7691 + }, + { + "epoch": 0.984576, + "grad_norm": 0.41192880272865295, + "learning_rate": 6.187540826442795e-07, + "loss": 2.4873, + "step": 7692 + }, + { + "epoch": 0.984704, + "grad_norm": 0.3953574001789093, + "learning_rate": 6.084865671162065e-07, + "loss": 2.59, + "step": 7693 + }, + { + "epoch": 0.984832, + "grad_norm": 0.40229541063308716, + "learning_rate": 5.983049027786791e-07, + "loss": 2.6429, + "step": 7694 + }, + { + "epoch": 0.98496, + "grad_norm": 0.39004287123680115, + "learning_rate": 5.882090913821303e-07, + "loss": 2.5907, + "step": 7695 + }, + { + "epoch": 0.985088, + "grad_norm": 0.40671682357788086, + "learning_rate": 5.781991346621163e-07, + "loss": 2.5789, + "step": 7696 + }, + { + "epoch": 0.985216, + "grad_norm": 0.38154056668281555, + "learning_rate": 5.682750343394826e-07, + "loss": 2.5881, + "step": 7697 + }, + { + "epoch": 0.985344, + "grad_norm": 0.41573047637939453, + "learning_rate": 5.584367921202538e-07, + "loss": 2.4547, + "step": 7698 + }, + { + "epoch": 0.985472, + "grad_norm": 0.439553439617157, + "learning_rate": 5.486844096957988e-07, + "loss": 2.6532, + "step": 7699 + }, + { + "epoch": 0.9856, + "grad_norm": 0.42800694704055786, + "learning_rate": 5.3901788874261e-07, + "loss": 2.584, + "step": 7700 + }, + { + "epoch": 0.985728, + "grad_norm": 0.4257173538208008, + "learning_rate": 5.294372309224693e-07, + "loss": 2.4708, + "step": 7701 + }, + { + "epoch": 0.985856, + "grad_norm": 0.4204472303390503, + "learning_rate": 5.19942437882448e-07, + "loss": 2.5773, + "step": 7702 + }, + { + "epoch": 0.985984, + "grad_norm": 0.4059085547924042, + "learning_rate": 5.10533511254796e-07, + "loss": 2.5555, + "step": 7703 + }, + { + "epoch": 0.986112, + "grad_norm": 0.4596484899520874, + "learning_rate": 5.012104526569972e-07, + "loss": 2.6713, + "step": 7704 + }, + { + "epoch": 0.98624, + "grad_norm": 0.4165598452091217, + "learning_rate": 4.919732636918251e-07, + "loss": 2.6005, + "step": 7705 + }, + { + "epoch": 0.986368, + "grad_norm": 0.40604734420776367, + "learning_rate": 4.828219459472316e-07, + "loss": 2.7097, + "step": 7706 + }, + { + "epoch": 0.986496, + "grad_norm": 0.46335098147392273, + "learning_rate": 4.737565009964584e-07, + "loss": 2.6424, + "step": 7707 + }, + { + "epoch": 0.986624, + "grad_norm": 0.40809494256973267, + "learning_rate": 4.647769303979255e-07, + "loss": 2.5817, + "step": 7708 + }, + { + "epoch": 0.986752, + "grad_norm": 0.41696181893348694, + "learning_rate": 4.558832356953979e-07, + "loss": 2.492, + "step": 7709 + }, + { + "epoch": 0.98688, + "grad_norm": 0.3914263844490051, + "learning_rate": 4.470754184177084e-07, + "loss": 2.6538, + "step": 7710 + }, + { + "epoch": 0.987008, + "grad_norm": 0.4100235402584076, + "learning_rate": 4.3835348007909004e-07, + "loss": 2.6261, + "step": 7711 + }, + { + "epoch": 0.987136, + "grad_norm": 0.4010683298110962, + "learning_rate": 4.297174221788991e-07, + "loss": 2.5196, + "step": 7712 + }, + { + "epoch": 0.987264, + "grad_norm": 0.3987716734409332, + "learning_rate": 4.2116724620183675e-07, + "loss": 2.524, + "step": 7713 + }, + { + "epoch": 0.987392, + "grad_norm": 0.452481210231781, + "learning_rate": 4.1270295361772737e-07, + "loss": 2.5012, + "step": 7714 + }, + { + "epoch": 0.98752, + "grad_norm": 0.44848570227622986, + "learning_rate": 4.043245458816847e-07, + "loss": 2.6013, + "step": 7715 + }, + { + "epoch": 0.987648, + "grad_norm": 0.37905052304267883, + "learning_rate": 3.960320244340565e-07, + "loss": 2.5376, + "step": 7716 + }, + { + "epoch": 0.987776, + "grad_norm": 0.4249488413333893, + "learning_rate": 3.878253907004248e-07, + "loss": 2.5396, + "step": 7717 + }, + { + "epoch": 0.987904, + "grad_norm": 0.4086533784866333, + "learning_rate": 3.7970464609160535e-07, + "loss": 2.6052, + "step": 7718 + }, + { + "epoch": 0.988032, + "grad_norm": 0.4322079122066498, + "learning_rate": 3.7166979200364826e-07, + "loss": 2.5427, + "step": 7719 + }, + { + "epoch": 0.98816, + "grad_norm": 0.39620599150657654, + "learning_rate": 3.637208298178929e-07, + "loss": 2.5561, + "step": 7720 + }, + { + "epoch": 0.988288, + "grad_norm": 0.38227006793022156, + "learning_rate": 3.558577609007463e-07, + "loss": 2.6864, + "step": 7721 + }, + { + "epoch": 0.988416, + "grad_norm": 0.40953972935676575, + "learning_rate": 3.480805866040715e-07, + "loss": 2.5039, + "step": 7722 + }, + { + "epoch": 0.988544, + "grad_norm": 0.42143556475639343, + "learning_rate": 3.403893082647991e-07, + "loss": 2.6415, + "step": 7723 + }, + { + "epoch": 0.988672, + "grad_norm": 0.4974040687084198, + "learning_rate": 3.3278392720514915e-07, + "loss": 2.6113, + "step": 7724 + }, + { + "epoch": 0.9888, + "grad_norm": 0.3998076319694519, + "learning_rate": 3.252644447325759e-07, + "loss": 2.5595, + "step": 7725 + }, + { + "epoch": 0.988928, + "grad_norm": 0.4052791893482208, + "learning_rate": 3.178308621397674e-07, + "loss": 2.6066, + "step": 7726 + }, + { + "epoch": 0.989056, + "grad_norm": 0.41366228461265564, + "learning_rate": 3.10483180704646e-07, + "loss": 2.553, + "step": 7727 + }, + { + "epoch": 0.989184, + "grad_norm": 0.47696906328201294, + "learning_rate": 3.0322140169036784e-07, + "loss": 2.602, + "step": 7728 + }, + { + "epoch": 0.989312, + "grad_norm": 0.45343658328056335, + "learning_rate": 2.9604552634532324e-07, + "loss": 2.5946, + "step": 7729 + }, + { + "epoch": 0.98944, + "grad_norm": 0.43656083941459656, + "learning_rate": 2.88955555903081e-07, + "loss": 2.5568, + "step": 7730 + }, + { + "epoch": 0.989568, + "grad_norm": 0.4064961075782776, + "learning_rate": 2.8195149158255497e-07, + "loss": 2.5195, + "step": 7731 + }, + { + "epoch": 0.989696, + "grad_norm": 0.39533132314682007, + "learning_rate": 2.75033334587782e-07, + "loss": 2.5355, + "step": 7732 + }, + { + "epoch": 0.989824, + "grad_norm": 0.37238696217536926, + "learning_rate": 2.6820108610808855e-07, + "loss": 2.6184, + "step": 7733 + }, + { + "epoch": 0.989952, + "grad_norm": 0.41293662786483765, + "learning_rate": 2.614547473180351e-07, + "loss": 2.587, + "step": 7734 + }, + { + "epoch": 0.99008, + "grad_norm": 0.44657954573631287, + "learning_rate": 2.5479431937736055e-07, + "loss": 2.5535, + "step": 7735 + }, + { + "epoch": 0.990208, + "grad_norm": 0.4218708574771881, + "learning_rate": 2.482198034310934e-07, + "loss": 2.5511, + "step": 7736 + }, + { + "epoch": 0.990336, + "grad_norm": 0.42913421988487244, + "learning_rate": 2.4173120060949626e-07, + "loss": 2.577, + "step": 7737 + }, + { + "epoch": 0.990464, + "grad_norm": 0.43025708198547363, + "learning_rate": 2.353285120279547e-07, + "loss": 2.5552, + "step": 7738 + }, + { + "epoch": 0.990592, + "grad_norm": 0.3936161994934082, + "learning_rate": 2.2901173878725478e-07, + "loss": 2.5659, + "step": 7739 + }, + { + "epoch": 0.99072, + "grad_norm": 0.4089466333389282, + "learning_rate": 2.2278088197325018e-07, + "loss": 2.5571, + "step": 7740 + }, + { + "epoch": 0.990848, + "grad_norm": 0.39933764934539795, + "learning_rate": 2.1663594265708407e-07, + "loss": 2.647, + "step": 7741 + }, + { + "epoch": 0.990976, + "grad_norm": 0.41837194561958313, + "learning_rate": 2.1057692189524468e-07, + "loss": 2.567, + "step": 7742 + }, + { + "epoch": 0.991104, + "grad_norm": 0.4884002208709717, + "learning_rate": 2.0460382072928775e-07, + "loss": 2.5211, + "step": 7743 + }, + { + "epoch": 0.991232, + "grad_norm": 0.43809768557548523, + "learning_rate": 1.98716640186003e-07, + "loss": 2.5954, + "step": 7744 + }, + { + "epoch": 0.99136, + "grad_norm": 0.4037536382675171, + "learning_rate": 1.9291538127758078e-07, + "loss": 2.5838, + "step": 7745 + }, + { + "epoch": 0.991488, + "grad_norm": 0.4625838100910187, + "learning_rate": 1.8720004500122345e-07, + "loss": 2.5869, + "step": 7746 + }, + { + "epoch": 0.991616, + "grad_norm": 0.39551764726638794, + "learning_rate": 1.8157063233953387e-07, + "loss": 2.5951, + "step": 7747 + }, + { + "epoch": 0.991744, + "grad_norm": 0.41296568512916565, + "learning_rate": 1.7602714426023791e-07, + "loss": 2.538, + "step": 7748 + }, + { + "epoch": 0.991872, + "grad_norm": 0.42505696415901184, + "learning_rate": 1.7056958171635108e-07, + "loss": 2.5165, + "step": 7749 + }, + { + "epoch": 0.992, + "grad_norm": 0.5562304854393005, + "learning_rate": 1.6519794564606727e-07, + "loss": 2.6318, + "step": 7750 + }, + { + "epoch": 0.992128, + "grad_norm": 0.41729649901390076, + "learning_rate": 1.5991223697281455e-07, + "loss": 2.5581, + "step": 7751 + }, + { + "epoch": 0.992256, + "grad_norm": 0.39844727516174316, + "learning_rate": 1.5471245660531042e-07, + "loss": 2.5138, + "step": 7752 + }, + { + "epoch": 0.992384, + "grad_norm": 0.3881789743900299, + "learning_rate": 1.4959860543739546e-07, + "loss": 2.5003, + "step": 7753 + }, + { + "epoch": 0.992512, + "grad_norm": 0.4086701273918152, + "learning_rate": 1.4457068434831077e-07, + "loss": 2.5743, + "step": 7754 + }, + { + "epoch": 0.99264, + "grad_norm": 0.4005666673183441, + "learning_rate": 1.396286942023095e-07, + "loss": 2.5282, + "step": 7755 + }, + { + "epoch": 0.992768, + "grad_norm": 0.4330980181694031, + "learning_rate": 1.3477263584904532e-07, + "loss": 2.4889, + "step": 7756 + }, + { + "epoch": 0.992896, + "grad_norm": 0.45946115255355835, + "learning_rate": 1.3000251012323938e-07, + "loss": 2.6225, + "step": 7757 + }, + { + "epoch": 0.993024, + "grad_norm": 0.4067467153072357, + "learning_rate": 1.2531831784506897e-07, + "loss": 2.4457, + "step": 7758 + }, + { + "epoch": 0.993152, + "grad_norm": 0.44573575258255005, + "learning_rate": 1.2072005981966781e-07, + "loss": 2.4782, + "step": 7759 + }, + { + "epoch": 0.99328, + "grad_norm": 0.3963452875614166, + "learning_rate": 1.1620773683757024e-07, + "loss": 2.5677, + "step": 7760 + }, + { + "epoch": 0.993408, + "grad_norm": 0.4831584095954895, + "learning_rate": 1.1178134967454457e-07, + "loss": 2.6369, + "step": 7761 + }, + { + "epoch": 0.993536, + "grad_norm": 0.40597647428512573, + "learning_rate": 1.0744089909153765e-07, + "loss": 2.6103, + "step": 7762 + }, + { + "epoch": 0.993664, + "grad_norm": 0.4025424122810364, + "learning_rate": 1.0318638583467488e-07, + "loss": 2.521, + "step": 7763 + }, + { + "epoch": 0.993792, + "grad_norm": 0.40648508071899414, + "learning_rate": 9.901781063531568e-08, + "loss": 2.5765, + "step": 7764 + }, + { + "epoch": 0.99392, + "grad_norm": 0.42939871549606323, + "learning_rate": 9.493517421022002e-08, + "loss": 2.6259, + "step": 7765 + }, + { + "epoch": 0.994048, + "grad_norm": 0.43415457010269165, + "learning_rate": 9.093847726110438e-08, + "loss": 2.587, + "step": 7766 + }, + { + "epoch": 0.994176, + "grad_norm": 0.8042909502983093, + "learning_rate": 8.702772047514129e-08, + "loss": 2.6391, + "step": 7767 + }, + { + "epoch": 0.994304, + "grad_norm": 0.42947402596473694, + "learning_rate": 8.320290452462631e-08, + "loss": 2.5772, + "step": 7768 + }, + { + "epoch": 0.994432, + "grad_norm": 0.374708354473114, + "learning_rate": 7.94640300670335e-08, + "loss": 2.6501, + "step": 7769 + }, + { + "epoch": 0.99456, + "grad_norm": 0.38886433839797974, + "learning_rate": 7.581109774512652e-08, + "loss": 2.5742, + "step": 7770 + }, + { + "epoch": 0.994688, + "grad_norm": 0.40963608026504517, + "learning_rate": 7.224410818695848e-08, + "loss": 2.5621, + "step": 7771 + }, + { + "epoch": 0.994816, + "grad_norm": 0.5574448704719543, + "learning_rate": 6.876306200565009e-08, + "loss": 2.623, + "step": 7772 + }, + { + "epoch": 0.994944, + "grad_norm": 0.42132848501205444, + "learning_rate": 6.536795979966703e-08, + "loss": 2.5171, + "step": 7773 + }, + { + "epoch": 0.995072, + "grad_norm": 0.45626896619796753, + "learning_rate": 6.205880215270909e-08, + "loss": 2.5304, + "step": 7774 + }, + { + "epoch": 0.9952, + "grad_norm": 0.38392314314842224, + "learning_rate": 5.883558963359903e-08, + "loss": 2.516, + "step": 7775 + }, + { + "epoch": 0.995328, + "grad_norm": 0.3983561396598816, + "learning_rate": 5.569832279644915e-08, + "loss": 2.5005, + "step": 7776 + }, + { + "epoch": 0.995456, + "grad_norm": 0.5089921355247498, + "learning_rate": 5.264700218066132e-08, + "loss": 2.5338, + "step": 7777 + }, + { + "epoch": 0.995584, + "grad_norm": 0.3942891061306, + "learning_rate": 4.968162831070488e-08, + "loss": 2.5224, + "step": 7778 + }, + { + "epoch": 0.995712, + "grad_norm": 0.42520105838775635, + "learning_rate": 4.6802201696394264e-08, + "loss": 2.5215, + "step": 7779 + }, + { + "epoch": 0.99584, + "grad_norm": 0.3953329622745514, + "learning_rate": 4.400872283272239e-08, + "loss": 2.541, + "step": 7780 + }, + { + "epoch": 0.995968, + "grad_norm": 0.397418349981308, + "learning_rate": 4.130119219997175e-08, + "loss": 2.5945, + "step": 7781 + }, + { + "epoch": 0.996096, + "grad_norm": 0.46265798807144165, + "learning_rate": 3.8679610263547824e-08, + "loss": 2.6313, + "step": 7782 + }, + { + "epoch": 0.996224, + "grad_norm": 0.3903265595436096, + "learning_rate": 3.614397747420117e-08, + "loss": 2.5399, + "step": 7783 + }, + { + "epoch": 0.996352, + "grad_norm": 0.4148658514022827, + "learning_rate": 3.3694294267694326e-08, + "loss": 2.5176, + "step": 7784 + }, + { + "epoch": 0.99648, + "grad_norm": 0.49215662479400635, + "learning_rate": 3.133056106530141e-08, + "loss": 2.5783, + "step": 7785 + }, + { + "epoch": 0.996608, + "grad_norm": 0.4316207468509674, + "learning_rate": 2.9052778273308545e-08, + "loss": 2.5401, + "step": 7786 + }, + { + "epoch": 0.996736, + "grad_norm": 0.4922514259815216, + "learning_rate": 2.6860946283291388e-08, + "loss": 2.4998, + "step": 7787 + }, + { + "epoch": 0.996864, + "grad_norm": 0.3953917622566223, + "learning_rate": 2.4755065472115144e-08, + "loss": 2.5941, + "step": 7788 + }, + { + "epoch": 0.996992, + "grad_norm": 0.3907940685749054, + "learning_rate": 2.2735136201712524e-08, + "loss": 2.5503, + "step": 7789 + }, + { + "epoch": 0.99712, + "grad_norm": 0.4295826852321625, + "learning_rate": 2.0801158819361287e-08, + "loss": 2.5501, + "step": 7790 + }, + { + "epoch": 0.997248, + "grad_norm": 0.3892579674720764, + "learning_rate": 1.895313365757323e-08, + "loss": 2.6197, + "step": 7791 + }, + { + "epoch": 0.997376, + "grad_norm": 0.44979408383369446, + "learning_rate": 1.7191061033983158e-08, + "loss": 2.4721, + "step": 7792 + }, + { + "epoch": 0.997504, + "grad_norm": 0.4299708604812622, + "learning_rate": 1.5514941251570934e-08, + "loss": 2.5629, + "step": 7793 + }, + { + "epoch": 0.997632, + "grad_norm": 0.49729976058006287, + "learning_rate": 1.3924774598439438e-08, + "loss": 2.5827, + "step": 7794 + }, + { + "epoch": 0.99776, + "grad_norm": 0.41326233744621277, + "learning_rate": 1.2420561348036597e-08, + "loss": 2.48, + "step": 7795 + }, + { + "epoch": 0.997888, + "grad_norm": 0.4581976532936096, + "learning_rate": 1.100230175882233e-08, + "loss": 2.6375, + "step": 7796 + }, + { + "epoch": 0.998016, + "grad_norm": 0.41912850737571716, + "learning_rate": 9.669996074712639e-09, + "loss": 2.5744, + "step": 7797 + }, + { + "epoch": 0.998144, + "grad_norm": 0.4488714337348938, + "learning_rate": 8.423644524691021e-09, + "loss": 2.4852, + "step": 7798 + }, + { + "epoch": 0.998272, + "grad_norm": 0.45455828309059143, + "learning_rate": 7.263247323086031e-09, + "loss": 2.5732, + "step": 7799 + }, + { + "epoch": 0.9984, + "grad_norm": 0.4155208468437195, + "learning_rate": 6.188804669349235e-09, + "loss": 2.4788, + "step": 7800 + }, + { + "epoch": 0.998528, + "grad_norm": 0.4247376024723053, + "learning_rate": 5.200316748110723e-09, + "loss": 2.6096, + "step": 7801 + }, + { + "epoch": 0.998656, + "grad_norm": 0.43610551953315735, + "learning_rate": 4.297783729456661e-09, + "loss": 2.6087, + "step": 7802 + }, + { + "epoch": 0.998784, + "grad_norm": 0.39986518025398254, + "learning_rate": 3.4812057684296962e-09, + "loss": 2.6362, + "step": 7803 + }, + { + "epoch": 0.998912, + "grad_norm": 0.4368244409561157, + "learning_rate": 2.7505830054175284e-09, + "loss": 2.5639, + "step": 7804 + }, + { + "epoch": 0.99904, + "grad_norm": 0.4288599491119385, + "learning_rate": 2.1059155660418936e-09, + "loss": 2.5435, + "step": 7805 + }, + { + "epoch": 0.999168, + "grad_norm": 0.47066229581832886, + "learning_rate": 1.547203561158561e-09, + "loss": 2.5608, + "step": 7806 + }, + { + "epoch": 0.999296, + "grad_norm": 0.38986000418663025, + "learning_rate": 1.074447086801822e-09, + "loss": 2.5032, + "step": 7807 + }, + { + "epoch": 0.999424, + "grad_norm": 0.42730167508125305, + "learning_rate": 6.876462241844905e-10, + "loss": 2.6605, + "step": 7808 + }, + { + "epoch": 0.999552, + "grad_norm": 0.7117275595664978, + "learning_rate": 3.8680103991994844e-10, + "loss": 2.492, + "step": 7809 + }, + { + "epoch": 0.99968, + "grad_norm": 0.4488013684749603, + "learning_rate": 1.7191158563356623e-10, + "loss": 2.5854, + "step": 7810 + }, + { + "epoch": 0.999808, + "grad_norm": 0.40580034255981445, + "learning_rate": 4.297789824025955e-11, + "loss": 2.5055, + "step": 7811 + }, + { + "epoch": 0.999936, + "grad_norm": 0.41080743074417114, + "learning_rate": 0.0, + "loss": 2.6596, + "step": 7812 + }, + { + "epoch": 0.999936, + "step": 7812, + "total_flos": 2.6494366569767895e+18, + "train_loss": 2.70675185876691, + "train_runtime": 90500.5744, + "train_samples_per_second": 22.099, + "train_steps_per_second": 0.086 + } + ], + "logging_steps": 1.0, + "max_steps": 7812, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 24000, + "total_flos": 2.6494366569767895e+18, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}