diff --git "a/checkpoint-4750/trainer_state.json" "b/checkpoint-4750/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-4750/trainer_state.json" @@ -0,0 +1,33284 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.52, + "eval_steps": 500, + "global_step": 4750, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00032, + "grad_norm": 0.24713319540023804, + "learning_rate": 0.0, + "loss": 0.8884, + "step": 1 + }, + { + "epoch": 0.00064, + "grad_norm": 0.20603826642036438, + "learning_rate": 1.3333333333333334e-06, + "loss": 0.8479, + "step": 2 + }, + { + "epoch": 0.00096, + "grad_norm": 0.24543610215187073, + "learning_rate": 2.666666666666667e-06, + "loss": 0.9853, + "step": 3 + }, + { + "epoch": 0.00128, + "grad_norm": 0.2051621973514557, + "learning_rate": 4.000000000000001e-06, + "loss": 0.8293, + "step": 4 + }, + { + "epoch": 0.0016, + "grad_norm": 0.23469021916389465, + "learning_rate": 5.333333333333334e-06, + "loss": 0.9811, + "step": 5 + }, + { + "epoch": 0.00192, + "grad_norm": 0.25180304050445557, + "learning_rate": 6.666666666666667e-06, + "loss": 1.039, + "step": 6 + }, + { + "epoch": 0.00224, + "grad_norm": 0.23135970532894135, + "learning_rate": 8.000000000000001e-06, + "loss": 0.8688, + "step": 7 + }, + { + "epoch": 0.00256, + "grad_norm": 0.22394850850105286, + "learning_rate": 9.333333333333334e-06, + "loss": 0.9214, + "step": 8 + }, + { + "epoch": 0.00288, + "grad_norm": 0.25566354393959045, + "learning_rate": 1.0666666666666667e-05, + "loss": 0.8667, + "step": 9 + }, + { + "epoch": 0.0032, + "grad_norm": 0.2881408631801605, + "learning_rate": 1.2e-05, + "loss": 0.9544, + "step": 10 + }, + { + "epoch": 0.00352, + "grad_norm": 0.22555342316627502, + "learning_rate": 1.3333333333333333e-05, + "loss": 0.9383, + "step": 11 + }, + { + "epoch": 0.00384, + "grad_norm": 0.23086762428283691, + "learning_rate": 1.4666666666666668e-05, + "loss": 0.8595, + "step": 12 + }, + { + "epoch": 0.00416, + "grad_norm": 0.2487766444683075, + "learning_rate": 1.6000000000000003e-05, + "loss": 0.9469, + "step": 13 + }, + { + "epoch": 0.00448, + "grad_norm": 0.3970007598400116, + "learning_rate": 1.7333333333333336e-05, + "loss": 1.1923, + "step": 14 + }, + { + "epoch": 0.0048, + "grad_norm": 0.25204387307167053, + "learning_rate": 1.866666666666667e-05, + "loss": 0.8854, + "step": 15 + }, + { + "epoch": 0.00512, + "grad_norm": 0.23282122611999512, + "learning_rate": 2e-05, + "loss": 0.932, + "step": 16 + }, + { + "epoch": 0.00544, + "grad_norm": 0.2607545852661133, + "learning_rate": 2.1333333333333335e-05, + "loss": 0.9094, + "step": 17 + }, + { + "epoch": 0.00576, + "grad_norm": 0.21402348577976227, + "learning_rate": 2.2666666666666668e-05, + "loss": 0.7879, + "step": 18 + }, + { + "epoch": 0.00608, + "grad_norm": 0.23922377824783325, + "learning_rate": 2.4e-05, + "loss": 0.9193, + "step": 19 + }, + { + "epoch": 0.0064, + "grad_norm": 0.18835392594337463, + "learning_rate": 2.5333333333333337e-05, + "loss": 0.9786, + "step": 20 + }, + { + "epoch": 0.00672, + "grad_norm": 0.1775645613670349, + "learning_rate": 2.6666666666666667e-05, + "loss": 0.933, + "step": 21 + }, + { + "epoch": 0.00704, + "grad_norm": 0.20945614576339722, + "learning_rate": 2.8000000000000003e-05, + "loss": 0.8171, + "step": 22 + }, + { + "epoch": 0.00736, + "grad_norm": 0.16134795546531677, + "learning_rate": 2.9333333333333336e-05, + "loss": 0.9665, + "step": 23 + }, + { + "epoch": 0.00768, + "grad_norm": 0.1454283446073532, + "learning_rate": 3.066666666666667e-05, + "loss": 0.7238, + "step": 24 + }, + { + "epoch": 0.008, + "grad_norm": 0.15920202434062958, + "learning_rate": 3.2000000000000005e-05, + "loss": 0.7932, + "step": 25 + }, + { + "epoch": 0.00832, + "grad_norm": 0.16204868257045746, + "learning_rate": 3.3333333333333335e-05, + "loss": 0.8175, + "step": 26 + }, + { + "epoch": 0.00864, + "grad_norm": 0.15179912745952606, + "learning_rate": 3.466666666666667e-05, + "loss": 0.9067, + "step": 27 + }, + { + "epoch": 0.00896, + "grad_norm": 0.1287967711687088, + "learning_rate": 3.6e-05, + "loss": 0.8039, + "step": 28 + }, + { + "epoch": 0.00928, + "grad_norm": 0.13528791069984436, + "learning_rate": 3.733333333333334e-05, + "loss": 0.8892, + "step": 29 + }, + { + "epoch": 0.0096, + "grad_norm": 0.13059762120246887, + "learning_rate": 3.866666666666667e-05, + "loss": 0.7023, + "step": 30 + }, + { + "epoch": 0.00992, + "grad_norm": 0.13119487464427948, + "learning_rate": 4e-05, + "loss": 0.7607, + "step": 31 + }, + { + "epoch": 0.01024, + "grad_norm": 0.12565594911575317, + "learning_rate": 4.133333333333333e-05, + "loss": 0.808, + "step": 32 + }, + { + "epoch": 0.01056, + "grad_norm": 0.1259136199951172, + "learning_rate": 4.266666666666667e-05, + "loss": 0.8704, + "step": 33 + }, + { + "epoch": 0.01088, + "grad_norm": 0.1319507360458374, + "learning_rate": 4.4000000000000006e-05, + "loss": 0.6421, + "step": 34 + }, + { + "epoch": 0.0112, + "grad_norm": 0.1317446082830429, + "learning_rate": 4.5333333333333335e-05, + "loss": 0.688, + "step": 35 + }, + { + "epoch": 0.01152, + "grad_norm": 0.12360246479511261, + "learning_rate": 4.666666666666667e-05, + "loss": 0.798, + "step": 36 + }, + { + "epoch": 0.01184, + "grad_norm": 0.11974922567605972, + "learning_rate": 4.8e-05, + "loss": 0.829, + "step": 37 + }, + { + "epoch": 0.01216, + "grad_norm": 0.13918066024780273, + "learning_rate": 4.933333333333334e-05, + "loss": 0.8626, + "step": 38 + }, + { + "epoch": 0.01248, + "grad_norm": 0.11004938930273056, + "learning_rate": 5.0666666666666674e-05, + "loss": 0.8056, + "step": 39 + }, + { + "epoch": 0.0128, + "grad_norm": 0.10854440927505493, + "learning_rate": 5.2000000000000004e-05, + "loss": 0.8773, + "step": 40 + }, + { + "epoch": 0.01312, + "grad_norm": 0.10877286642789841, + "learning_rate": 5.333333333333333e-05, + "loss": 0.8947, + "step": 41 + }, + { + "epoch": 0.01344, + "grad_norm": 0.12767033278942108, + "learning_rate": 5.466666666666666e-05, + "loss": 0.7892, + "step": 42 + }, + { + "epoch": 0.01376, + "grad_norm": 0.11700518429279327, + "learning_rate": 5.6000000000000006e-05, + "loss": 0.8598, + "step": 43 + }, + { + "epoch": 0.01408, + "grad_norm": 0.12549364566802979, + "learning_rate": 5.7333333333333336e-05, + "loss": 0.8498, + "step": 44 + }, + { + "epoch": 0.0144, + "grad_norm": 0.11921705305576324, + "learning_rate": 5.866666666666667e-05, + "loss": 0.7979, + "step": 45 + }, + { + "epoch": 0.01472, + "grad_norm": 0.11621945351362228, + "learning_rate": 6e-05, + "loss": 0.9252, + "step": 46 + }, + { + "epoch": 0.01504, + "grad_norm": 0.10912485420703888, + "learning_rate": 6.133333333333334e-05, + "loss": 0.7674, + "step": 47 + }, + { + "epoch": 0.01536, + "grad_norm": 0.11039536446332932, + "learning_rate": 6.266666666666667e-05, + "loss": 0.841, + "step": 48 + }, + { + "epoch": 0.01568, + "grad_norm": 0.10852135717868805, + "learning_rate": 6.400000000000001e-05, + "loss": 0.7991, + "step": 49 + }, + { + "epoch": 0.016, + "grad_norm": 0.09653560072183609, + "learning_rate": 6.533333333333334e-05, + "loss": 0.6292, + "step": 50 + }, + { + "epoch": 0.01632, + "grad_norm": 0.11932919919490814, + "learning_rate": 6.666666666666667e-05, + "loss": 0.7327, + "step": 51 + }, + { + "epoch": 0.01664, + "grad_norm": 0.10563201457262039, + "learning_rate": 6.800000000000001e-05, + "loss": 0.7041, + "step": 52 + }, + { + "epoch": 0.01696, + "grad_norm": 0.1063474714756012, + "learning_rate": 6.933333333333334e-05, + "loss": 0.7953, + "step": 53 + }, + { + "epoch": 0.01728, + "grad_norm": 0.11396024376153946, + "learning_rate": 7.066666666666667e-05, + "loss": 0.899, + "step": 54 + }, + { + "epoch": 0.0176, + "grad_norm": 0.10861669480800629, + "learning_rate": 7.2e-05, + "loss": 0.9343, + "step": 55 + }, + { + "epoch": 0.01792, + "grad_norm": 0.12476427853107452, + "learning_rate": 7.333333333333333e-05, + "loss": 0.8145, + "step": 56 + }, + { + "epoch": 0.01824, + "grad_norm": 0.1240817978978157, + "learning_rate": 7.466666666666667e-05, + "loss": 0.7133, + "step": 57 + }, + { + "epoch": 0.01856, + "grad_norm": 0.10977518558502197, + "learning_rate": 7.6e-05, + "loss": 0.7576, + "step": 58 + }, + { + "epoch": 0.01888, + "grad_norm": 0.11463093012571335, + "learning_rate": 7.733333333333333e-05, + "loss": 0.8552, + "step": 59 + }, + { + "epoch": 0.0192, + "grad_norm": 0.1143040657043457, + "learning_rate": 7.866666666666666e-05, + "loss": 0.9198, + "step": 60 + }, + { + "epoch": 0.01952, + "grad_norm": 0.12839296460151672, + "learning_rate": 8e-05, + "loss": 0.6644, + "step": 61 + }, + { + "epoch": 0.01984, + "grad_norm": 0.11903133243322372, + "learning_rate": 8.133333333333334e-05, + "loss": 0.9053, + "step": 62 + }, + { + "epoch": 0.02016, + "grad_norm": 0.12296755611896515, + "learning_rate": 8.266666666666667e-05, + "loss": 0.7695, + "step": 63 + }, + { + "epoch": 0.02048, + "grad_norm": 0.11100097000598907, + "learning_rate": 8.4e-05, + "loss": 0.8909, + "step": 64 + }, + { + "epoch": 0.0208, + "grad_norm": 0.11848391592502594, + "learning_rate": 8.533333333333334e-05, + "loss": 0.8204, + "step": 65 + }, + { + "epoch": 0.02112, + "grad_norm": 0.11903996020555496, + "learning_rate": 8.666666666666667e-05, + "loss": 0.7778, + "step": 66 + }, + { + "epoch": 0.02144, + "grad_norm": 0.10964814573526382, + "learning_rate": 8.800000000000001e-05, + "loss": 0.6879, + "step": 67 + }, + { + "epoch": 0.02176, + "grad_norm": 0.11036152392625809, + "learning_rate": 8.933333333333334e-05, + "loss": 0.8051, + "step": 68 + }, + { + "epoch": 0.02208, + "grad_norm": 0.10888101160526276, + "learning_rate": 9.066666666666667e-05, + "loss": 0.759, + "step": 69 + }, + { + "epoch": 0.0224, + "grad_norm": 0.11158827692270279, + "learning_rate": 9.200000000000001e-05, + "loss": 0.7412, + "step": 70 + }, + { + "epoch": 0.02272, + "grad_norm": 0.12085185945034027, + "learning_rate": 9.333333333333334e-05, + "loss": 0.7978, + "step": 71 + }, + { + "epoch": 0.02304, + "grad_norm": 0.12325625121593475, + "learning_rate": 9.466666666666667e-05, + "loss": 0.7175, + "step": 72 + }, + { + "epoch": 0.02336, + "grad_norm": 0.10695890337228775, + "learning_rate": 9.6e-05, + "loss": 0.868, + "step": 73 + }, + { + "epoch": 0.02368, + "grad_norm": 0.11983097344636917, + "learning_rate": 9.733333333333335e-05, + "loss": 0.7965, + "step": 74 + }, + { + "epoch": 0.024, + "grad_norm": 0.11521439999341965, + "learning_rate": 9.866666666666668e-05, + "loss": 0.9305, + "step": 75 + }, + { + "epoch": 0.02432, + "grad_norm": 0.1252303123474121, + "learning_rate": 0.0001, + "loss": 0.6807, + "step": 76 + }, + { + "epoch": 0.02464, + "grad_norm": 0.1265154629945755, + "learning_rate": 0.00010133333333333335, + "loss": 0.718, + "step": 77 + }, + { + "epoch": 0.02496, + "grad_norm": 0.1109502986073494, + "learning_rate": 0.00010266666666666666, + "loss": 0.769, + "step": 78 + }, + { + "epoch": 0.02528, + "grad_norm": 0.11158560961484909, + "learning_rate": 0.00010400000000000001, + "loss": 0.8579, + "step": 79 + }, + { + "epoch": 0.0256, + "grad_norm": 0.13526469469070435, + "learning_rate": 0.00010533333333333332, + "loss": 0.7704, + "step": 80 + }, + { + "epoch": 0.02592, + "grad_norm": 0.12420251965522766, + "learning_rate": 0.00010666666666666667, + "loss": 0.673, + "step": 81 + }, + { + "epoch": 0.02624, + "grad_norm": 0.11838296055793762, + "learning_rate": 0.00010800000000000001, + "loss": 0.8576, + "step": 82 + }, + { + "epoch": 0.02656, + "grad_norm": 0.11958430707454681, + "learning_rate": 0.00010933333333333333, + "loss": 0.8521, + "step": 83 + }, + { + "epoch": 0.02688, + "grad_norm": 0.12049078941345215, + "learning_rate": 0.00011066666666666667, + "loss": 0.6008, + "step": 84 + }, + { + "epoch": 0.0272, + "grad_norm": 0.10999640822410583, + "learning_rate": 0.00011200000000000001, + "loss": 0.7351, + "step": 85 + }, + { + "epoch": 0.02752, + "grad_norm": 0.11500285565853119, + "learning_rate": 0.00011333333333333334, + "loss": 0.7841, + "step": 86 + }, + { + "epoch": 0.02784, + "grad_norm": 0.11917781829833984, + "learning_rate": 0.00011466666666666667, + "loss": 0.8275, + "step": 87 + }, + { + "epoch": 0.02816, + "grad_norm": 0.11745665222406387, + "learning_rate": 0.000116, + "loss": 0.6096, + "step": 88 + }, + { + "epoch": 0.02848, + "grad_norm": 0.1208786815404892, + "learning_rate": 0.00011733333333333334, + "loss": 0.6842, + "step": 89 + }, + { + "epoch": 0.0288, + "grad_norm": 0.12839952111244202, + "learning_rate": 0.00011866666666666669, + "loss": 0.7467, + "step": 90 + }, + { + "epoch": 0.02912, + "grad_norm": 0.12688350677490234, + "learning_rate": 0.00012, + "loss": 0.6507, + "step": 91 + }, + { + "epoch": 0.02944, + "grad_norm": 0.11489767581224442, + "learning_rate": 0.00012133333333333335, + "loss": 0.8403, + "step": 92 + }, + { + "epoch": 0.02976, + "grad_norm": 0.1305433064699173, + "learning_rate": 0.00012266666666666668, + "loss": 0.748, + "step": 93 + }, + { + "epoch": 0.03008, + "grad_norm": 0.13992320001125336, + "learning_rate": 0.000124, + "loss": 0.796, + "step": 94 + }, + { + "epoch": 0.0304, + "grad_norm": 0.12120400369167328, + "learning_rate": 0.00012533333333333334, + "loss": 0.6445, + "step": 95 + }, + { + "epoch": 0.03072, + "grad_norm": 0.13254240155220032, + "learning_rate": 0.00012666666666666666, + "loss": 0.7488, + "step": 96 + }, + { + "epoch": 0.03104, + "grad_norm": 0.13790103793144226, + "learning_rate": 0.00012800000000000002, + "loss": 0.8005, + "step": 97 + }, + { + "epoch": 0.03136, + "grad_norm": 0.12862953543663025, + "learning_rate": 0.00012933333333333332, + "loss": 0.7975, + "step": 98 + }, + { + "epoch": 0.03168, + "grad_norm": 0.14197330176830292, + "learning_rate": 0.00013066666666666668, + "loss": 0.9026, + "step": 99 + }, + { + "epoch": 0.032, + "grad_norm": 0.12494270503520966, + "learning_rate": 0.000132, + "loss": 0.7353, + "step": 100 + }, + { + "epoch": 0.03232, + "grad_norm": 0.12572987377643585, + "learning_rate": 0.00013333333333333334, + "loss": 0.9271, + "step": 101 + }, + { + "epoch": 0.03264, + "grad_norm": 0.1314753144979477, + "learning_rate": 0.00013466666666666667, + "loss": 0.7816, + "step": 102 + }, + { + "epoch": 0.03296, + "grad_norm": 0.12902331352233887, + "learning_rate": 0.00013600000000000003, + "loss": 0.7348, + "step": 103 + }, + { + "epoch": 0.03328, + "grad_norm": 0.14086449146270752, + "learning_rate": 0.00013733333333333333, + "loss": 0.831, + "step": 104 + }, + { + "epoch": 0.0336, + "grad_norm": 0.12464585155248642, + "learning_rate": 0.00013866666666666669, + "loss": 0.6945, + "step": 105 + }, + { + "epoch": 0.03392, + "grad_norm": 0.11436092108488083, + "learning_rate": 0.00014, + "loss": 0.8192, + "step": 106 + }, + { + "epoch": 0.03424, + "grad_norm": 0.11779410392045975, + "learning_rate": 0.00014133333333333334, + "loss": 0.7989, + "step": 107 + }, + { + "epoch": 0.03456, + "grad_norm": 0.14601927995681763, + "learning_rate": 0.00014266666666666667, + "loss": 0.6256, + "step": 108 + }, + { + "epoch": 0.03488, + "grad_norm": 0.13231013715267181, + "learning_rate": 0.000144, + "loss": 0.7445, + "step": 109 + }, + { + "epoch": 0.0352, + "grad_norm": 0.11249889433383942, + "learning_rate": 0.00014533333333333333, + "loss": 0.7338, + "step": 110 + }, + { + "epoch": 0.03552, + "grad_norm": 0.11275426298379898, + "learning_rate": 0.00014666666666666666, + "loss": 0.7993, + "step": 111 + }, + { + "epoch": 0.03584, + "grad_norm": 0.11678344756364822, + "learning_rate": 0.000148, + "loss": 0.9234, + "step": 112 + }, + { + "epoch": 0.03616, + "grad_norm": 0.10879474133253098, + "learning_rate": 0.00014933333333333335, + "loss": 0.8732, + "step": 113 + }, + { + "epoch": 0.03648, + "grad_norm": 0.13673344254493713, + "learning_rate": 0.00015066666666666668, + "loss": 0.8698, + "step": 114 + }, + { + "epoch": 0.0368, + "grad_norm": 0.12564410269260406, + "learning_rate": 0.000152, + "loss": 0.7913, + "step": 115 + }, + { + "epoch": 0.03712, + "grad_norm": 0.12272034585475922, + "learning_rate": 0.00015333333333333334, + "loss": 0.7244, + "step": 116 + }, + { + "epoch": 0.03744, + "grad_norm": 0.11841209977865219, + "learning_rate": 0.00015466666666666667, + "loss": 0.7648, + "step": 117 + }, + { + "epoch": 0.03776, + "grad_norm": 0.11938042938709259, + "learning_rate": 0.00015600000000000002, + "loss": 0.8381, + "step": 118 + }, + { + "epoch": 0.03808, + "grad_norm": 0.13476668298244476, + "learning_rate": 0.00015733333333333333, + "loss": 0.6931, + "step": 119 + }, + { + "epoch": 0.0384, + "grad_norm": 0.1274396926164627, + "learning_rate": 0.00015866666666666668, + "loss": 0.854, + "step": 120 + }, + { + "epoch": 0.03872, + "grad_norm": 0.11518494784832001, + "learning_rate": 0.00016, + "loss": 0.7591, + "step": 121 + }, + { + "epoch": 0.03904, + "grad_norm": 0.12739014625549316, + "learning_rate": 0.00016133333333333334, + "loss": 0.6956, + "step": 122 + }, + { + "epoch": 0.03936, + "grad_norm": 0.13384665548801422, + "learning_rate": 0.00016266666666666667, + "loss": 0.6487, + "step": 123 + }, + { + "epoch": 0.03968, + "grad_norm": 0.11407941579818726, + "learning_rate": 0.000164, + "loss": 0.6822, + "step": 124 + }, + { + "epoch": 0.04, + "grad_norm": 0.12307750433683395, + "learning_rate": 0.00016533333333333333, + "loss": 0.7038, + "step": 125 + }, + { + "epoch": 0.04032, + "grad_norm": 0.12420140206813812, + "learning_rate": 0.0001666666666666667, + "loss": 0.6589, + "step": 126 + }, + { + "epoch": 0.04064, + "grad_norm": 0.1327451914548874, + "learning_rate": 0.000168, + "loss": 0.6002, + "step": 127 + }, + { + "epoch": 0.04096, + "grad_norm": 0.1248873695731163, + "learning_rate": 0.00016933333333333335, + "loss": 0.8812, + "step": 128 + }, + { + "epoch": 0.04128, + "grad_norm": 0.11638613790273666, + "learning_rate": 0.00017066666666666668, + "loss": 0.7455, + "step": 129 + }, + { + "epoch": 0.0416, + "grad_norm": 0.12473749369382858, + "learning_rate": 0.000172, + "loss": 0.6218, + "step": 130 + }, + { + "epoch": 0.04192, + "grad_norm": 0.1339467465877533, + "learning_rate": 0.00017333333333333334, + "loss": 0.729, + "step": 131 + }, + { + "epoch": 0.04224, + "grad_norm": 0.11078035831451416, + "learning_rate": 0.00017466666666666667, + "loss": 0.8507, + "step": 132 + }, + { + "epoch": 0.04256, + "grad_norm": 0.11885277926921844, + "learning_rate": 0.00017600000000000002, + "loss": 0.8817, + "step": 133 + }, + { + "epoch": 0.04288, + "grad_norm": 0.13378757238388062, + "learning_rate": 0.00017733333333333335, + "loss": 0.705, + "step": 134 + }, + { + "epoch": 0.0432, + "grad_norm": 0.11696331948041916, + "learning_rate": 0.00017866666666666668, + "loss": 0.7089, + "step": 135 + }, + { + "epoch": 0.04352, + "grad_norm": 0.12444671988487244, + "learning_rate": 0.00018, + "loss": 0.699, + "step": 136 + }, + { + "epoch": 0.04384, + "grad_norm": 0.11136961728334427, + "learning_rate": 0.00018133333333333334, + "loss": 0.7805, + "step": 137 + }, + { + "epoch": 0.04416, + "grad_norm": 0.13461412489414215, + "learning_rate": 0.00018266666666666667, + "loss": 0.7581, + "step": 138 + }, + { + "epoch": 0.04448, + "grad_norm": 0.11344683915376663, + "learning_rate": 0.00018400000000000003, + "loss": 0.7234, + "step": 139 + }, + { + "epoch": 0.0448, + "grad_norm": 0.12247714400291443, + "learning_rate": 0.00018533333333333333, + "loss": 0.9326, + "step": 140 + }, + { + "epoch": 0.04512, + "grad_norm": 0.12991179525852203, + "learning_rate": 0.0001866666666666667, + "loss": 0.7644, + "step": 141 + }, + { + "epoch": 0.04544, + "grad_norm": 0.1182120144367218, + "learning_rate": 0.000188, + "loss": 0.8437, + "step": 142 + }, + { + "epoch": 0.04576, + "grad_norm": 0.10737155377864838, + "learning_rate": 0.00018933333333333335, + "loss": 0.8047, + "step": 143 + }, + { + "epoch": 0.04608, + "grad_norm": 0.11993440985679626, + "learning_rate": 0.00019066666666666668, + "loss": 0.7247, + "step": 144 + }, + { + "epoch": 0.0464, + "grad_norm": 0.11705442517995834, + "learning_rate": 0.000192, + "loss": 0.7104, + "step": 145 + }, + { + "epoch": 0.04672, + "grad_norm": 0.12065937370061874, + "learning_rate": 0.00019333333333333333, + "loss": 0.7243, + "step": 146 + }, + { + "epoch": 0.04704, + "grad_norm": 0.1178629919886589, + "learning_rate": 0.0001946666666666667, + "loss": 0.6946, + "step": 147 + }, + { + "epoch": 0.04736, + "grad_norm": 0.12744340300559998, + "learning_rate": 0.000196, + "loss": 0.6946, + "step": 148 + }, + { + "epoch": 0.04768, + "grad_norm": 0.14068365097045898, + "learning_rate": 0.00019733333333333335, + "loss": 0.7905, + "step": 149 + }, + { + "epoch": 0.048, + "grad_norm": 0.11114205420017242, + "learning_rate": 0.00019866666666666668, + "loss": 0.8352, + "step": 150 + }, + { + "epoch": 0.04832, + "grad_norm": 0.10463167726993561, + "learning_rate": 0.0002, + "loss": 0.6664, + "step": 151 + }, + { + "epoch": 0.04864, + "grad_norm": 0.12237433344125748, + "learning_rate": 0.00019995876288659794, + "loss": 0.8114, + "step": 152 + }, + { + "epoch": 0.04896, + "grad_norm": 0.12069843709468842, + "learning_rate": 0.0001999175257731959, + "loss": 0.7901, + "step": 153 + }, + { + "epoch": 0.04928, + "grad_norm": 0.11398322135210037, + "learning_rate": 0.00019987628865979383, + "loss": 0.7131, + "step": 154 + }, + { + "epoch": 0.0496, + "grad_norm": 0.12183695286512375, + "learning_rate": 0.00019983505154639176, + "loss": 0.7609, + "step": 155 + }, + { + "epoch": 0.04992, + "grad_norm": 0.12869080901145935, + "learning_rate": 0.0001997938144329897, + "loss": 0.8103, + "step": 156 + }, + { + "epoch": 0.05024, + "grad_norm": 0.1119738221168518, + "learning_rate": 0.00019975257731958762, + "loss": 0.6731, + "step": 157 + }, + { + "epoch": 0.05056, + "grad_norm": 0.12156844139099121, + "learning_rate": 0.00019971134020618558, + "loss": 0.7535, + "step": 158 + }, + { + "epoch": 0.05088, + "grad_norm": 0.12900514900684357, + "learning_rate": 0.00019967010309278351, + "loss": 0.7378, + "step": 159 + }, + { + "epoch": 0.0512, + "grad_norm": 0.12743592262268066, + "learning_rate": 0.00019962886597938147, + "loss": 0.6197, + "step": 160 + }, + { + "epoch": 0.05152, + "grad_norm": 0.128151997923851, + "learning_rate": 0.0001995876288659794, + "loss": 0.7812, + "step": 161 + }, + { + "epoch": 0.05184, + "grad_norm": 0.10753507912158966, + "learning_rate": 0.00019954639175257733, + "loss": 0.5774, + "step": 162 + }, + { + "epoch": 0.05216, + "grad_norm": 0.11813949793577194, + "learning_rate": 0.00019950515463917527, + "loss": 0.6894, + "step": 163 + }, + { + "epoch": 0.05248, + "grad_norm": 0.1340881586074829, + "learning_rate": 0.0001994639175257732, + "loss": 0.7193, + "step": 164 + }, + { + "epoch": 0.0528, + "grad_norm": 0.10451763868331909, + "learning_rate": 0.00019942268041237116, + "loss": 0.572, + "step": 165 + }, + { + "epoch": 0.05312, + "grad_norm": 0.10894517600536346, + "learning_rate": 0.0001993814432989691, + "loss": 0.7693, + "step": 166 + }, + { + "epoch": 0.05344, + "grad_norm": 0.1243942379951477, + "learning_rate": 0.00019934020618556702, + "loss": 0.7805, + "step": 167 + }, + { + "epoch": 0.05376, + "grad_norm": 0.14128854870796204, + "learning_rate": 0.00019929896907216498, + "loss": 0.8595, + "step": 168 + }, + { + "epoch": 0.05408, + "grad_norm": 0.12148380279541016, + "learning_rate": 0.00019925773195876288, + "loss": 0.8954, + "step": 169 + }, + { + "epoch": 0.0544, + "grad_norm": 0.10860492289066315, + "learning_rate": 0.0001992164948453608, + "loss": 0.8101, + "step": 170 + }, + { + "epoch": 0.05472, + "grad_norm": 0.12292741239070892, + "learning_rate": 0.00019917525773195877, + "loss": 0.7965, + "step": 171 + }, + { + "epoch": 0.05504, + "grad_norm": 0.13840395212173462, + "learning_rate": 0.0001991340206185567, + "loss": 0.7231, + "step": 172 + }, + { + "epoch": 0.05536, + "grad_norm": 0.11291699856519699, + "learning_rate": 0.00019909278350515466, + "loss": 0.8247, + "step": 173 + }, + { + "epoch": 0.05568, + "grad_norm": 0.11811841279268265, + "learning_rate": 0.0001990515463917526, + "loss": 0.8504, + "step": 174 + }, + { + "epoch": 0.056, + "grad_norm": 0.11040613055229187, + "learning_rate": 0.00019901030927835052, + "loss": 0.6464, + "step": 175 + }, + { + "epoch": 0.05632, + "grad_norm": 0.10370033234357834, + "learning_rate": 0.00019896907216494845, + "loss": 0.8016, + "step": 176 + }, + { + "epoch": 0.05664, + "grad_norm": 0.11868111789226532, + "learning_rate": 0.00019892783505154639, + "loss": 0.7267, + "step": 177 + }, + { + "epoch": 0.05696, + "grad_norm": 0.11893659085035324, + "learning_rate": 0.00019888659793814434, + "loss": 0.814, + "step": 178 + }, + { + "epoch": 0.05728, + "grad_norm": 0.11829444020986557, + "learning_rate": 0.00019884536082474227, + "loss": 0.7852, + "step": 179 + }, + { + "epoch": 0.0576, + "grad_norm": 0.10493418574333191, + "learning_rate": 0.0001988041237113402, + "loss": 0.6154, + "step": 180 + }, + { + "epoch": 0.05792, + "grad_norm": 0.10895421355962753, + "learning_rate": 0.00019876288659793816, + "loss": 0.8663, + "step": 181 + }, + { + "epoch": 0.05824, + "grad_norm": 0.11782484501600266, + "learning_rate": 0.0001987216494845361, + "loss": 0.7902, + "step": 182 + }, + { + "epoch": 0.05856, + "grad_norm": 0.12639066576957703, + "learning_rate": 0.00019868041237113403, + "loss": 0.7381, + "step": 183 + }, + { + "epoch": 0.05888, + "grad_norm": 0.1321602165699005, + "learning_rate": 0.00019863917525773196, + "loss": 0.9283, + "step": 184 + }, + { + "epoch": 0.0592, + "grad_norm": 0.12468945235013962, + "learning_rate": 0.0001985979381443299, + "loss": 0.6641, + "step": 185 + }, + { + "epoch": 0.05952, + "grad_norm": 0.12850762903690338, + "learning_rate": 0.00019855670103092785, + "loss": 0.8694, + "step": 186 + }, + { + "epoch": 0.05984, + "grad_norm": 0.10586417466402054, + "learning_rate": 0.00019851546391752578, + "loss": 0.806, + "step": 187 + }, + { + "epoch": 0.06016, + "grad_norm": 0.10202761739492416, + "learning_rate": 0.00019847422680412374, + "loss": 0.684, + "step": 188 + }, + { + "epoch": 0.06048, + "grad_norm": 0.11566773056983948, + "learning_rate": 0.00019843298969072167, + "loss": 0.6737, + "step": 189 + }, + { + "epoch": 0.0608, + "grad_norm": 0.12295495718717575, + "learning_rate": 0.0001983917525773196, + "loss": 0.6539, + "step": 190 + }, + { + "epoch": 0.06112, + "grad_norm": 0.14126336574554443, + "learning_rate": 0.00019835051546391753, + "loss": 0.8539, + "step": 191 + }, + { + "epoch": 0.06144, + "grad_norm": 0.1215134933590889, + "learning_rate": 0.00019830927835051546, + "loss": 0.7051, + "step": 192 + }, + { + "epoch": 0.06176, + "grad_norm": 0.10785145312547684, + "learning_rate": 0.00019826804123711342, + "loss": 0.7877, + "step": 193 + }, + { + "epoch": 0.06208, + "grad_norm": 0.1299588680267334, + "learning_rate": 0.00019822680412371135, + "loss": 0.7942, + "step": 194 + }, + { + "epoch": 0.0624, + "grad_norm": 0.1072564497590065, + "learning_rate": 0.00019818556701030928, + "loss": 0.7055, + "step": 195 + }, + { + "epoch": 0.06272, + "grad_norm": 0.12069325149059296, + "learning_rate": 0.00019814432989690724, + "loss": 0.9036, + "step": 196 + }, + { + "epoch": 0.06304, + "grad_norm": 0.14454863965511322, + "learning_rate": 0.00019810309278350517, + "loss": 0.7567, + "step": 197 + }, + { + "epoch": 0.06336, + "grad_norm": 0.11882466822862625, + "learning_rate": 0.0001980618556701031, + "loss": 0.5986, + "step": 198 + }, + { + "epoch": 0.06368, + "grad_norm": 0.12222633510828018, + "learning_rate": 0.00019802061855670104, + "loss": 0.8493, + "step": 199 + }, + { + "epoch": 0.064, + "grad_norm": 0.10818106681108475, + "learning_rate": 0.00019797938144329897, + "loss": 0.7932, + "step": 200 + }, + { + "epoch": 0.06432, + "grad_norm": 0.1225774884223938, + "learning_rate": 0.00019793814432989693, + "loss": 0.7501, + "step": 201 + }, + { + "epoch": 0.06464, + "grad_norm": 0.13827428221702576, + "learning_rate": 0.00019789690721649486, + "loss": 0.818, + "step": 202 + }, + { + "epoch": 0.06496, + "grad_norm": 0.1145445853471756, + "learning_rate": 0.0001978556701030928, + "loss": 0.6433, + "step": 203 + }, + { + "epoch": 0.06528, + "grad_norm": 0.11928769201040268, + "learning_rate": 0.00019781443298969075, + "loss": 0.7352, + "step": 204 + }, + { + "epoch": 0.0656, + "grad_norm": 0.11696071177721024, + "learning_rate": 0.00019777319587628865, + "loss": 0.8499, + "step": 205 + }, + { + "epoch": 0.06592, + "grad_norm": 0.1281200349330902, + "learning_rate": 0.0001977319587628866, + "loss": 0.7526, + "step": 206 + }, + { + "epoch": 0.06624, + "grad_norm": 0.1270955204963684, + "learning_rate": 0.00019769072164948454, + "loss": 0.8456, + "step": 207 + }, + { + "epoch": 0.06656, + "grad_norm": 0.12278109043836594, + "learning_rate": 0.00019764948453608247, + "loss": 0.5972, + "step": 208 + }, + { + "epoch": 0.06688, + "grad_norm": 0.10781408101320267, + "learning_rate": 0.00019760824742268043, + "loss": 0.5967, + "step": 209 + }, + { + "epoch": 0.0672, + "grad_norm": 0.12237494438886642, + "learning_rate": 0.00019756701030927836, + "loss": 0.7306, + "step": 210 + }, + { + "epoch": 0.06752, + "grad_norm": 0.1100955531001091, + "learning_rate": 0.00019752577319587632, + "loss": 0.7579, + "step": 211 + }, + { + "epoch": 0.06784, + "grad_norm": 0.11036559194326401, + "learning_rate": 0.00019748453608247422, + "loss": 0.6668, + "step": 212 + }, + { + "epoch": 0.06816, + "grad_norm": 0.14552150666713715, + "learning_rate": 0.00019744329896907216, + "loss": 0.6668, + "step": 213 + }, + { + "epoch": 0.06848, + "grad_norm": 0.11125028133392334, + "learning_rate": 0.0001974020618556701, + "loss": 0.7342, + "step": 214 + }, + { + "epoch": 0.0688, + "grad_norm": 0.14329232275485992, + "learning_rate": 0.00019736082474226804, + "loss": 0.8708, + "step": 215 + }, + { + "epoch": 0.06912, + "grad_norm": 0.11911016702651978, + "learning_rate": 0.000197319587628866, + "loss": 0.7224, + "step": 216 + }, + { + "epoch": 0.06944, + "grad_norm": 0.12713277339935303, + "learning_rate": 0.00019727835051546393, + "loss": 0.7222, + "step": 217 + }, + { + "epoch": 0.06976, + "grad_norm": 0.1140274852514267, + "learning_rate": 0.00019723711340206187, + "loss": 0.627, + "step": 218 + }, + { + "epoch": 0.07008, + "grad_norm": 0.1070765033364296, + "learning_rate": 0.0001971958762886598, + "loss": 0.7514, + "step": 219 + }, + { + "epoch": 0.0704, + "grad_norm": 0.09937503188848495, + "learning_rate": 0.00019715463917525773, + "loss": 0.7539, + "step": 220 + }, + { + "epoch": 0.07072, + "grad_norm": 0.12137589603662491, + "learning_rate": 0.0001971134020618557, + "loss": 0.68, + "step": 221 + }, + { + "epoch": 0.07104, + "grad_norm": 0.11091487109661102, + "learning_rate": 0.00019707216494845362, + "loss": 0.6904, + "step": 222 + }, + { + "epoch": 0.07136, + "grad_norm": 0.10996535420417786, + "learning_rate": 0.00019703092783505155, + "loss": 0.6854, + "step": 223 + }, + { + "epoch": 0.07168, + "grad_norm": 0.129950150847435, + "learning_rate": 0.0001969896907216495, + "loss": 0.7023, + "step": 224 + }, + { + "epoch": 0.072, + "grad_norm": 0.13028115034103394, + "learning_rate": 0.00019694845360824744, + "loss": 0.6479, + "step": 225 + }, + { + "epoch": 0.07232, + "grad_norm": 0.1166912168264389, + "learning_rate": 0.00019690721649484537, + "loss": 0.6498, + "step": 226 + }, + { + "epoch": 0.07264, + "grad_norm": 0.1148250624537468, + "learning_rate": 0.0001968659793814433, + "loss": 0.7461, + "step": 227 + }, + { + "epoch": 0.07296, + "grad_norm": 0.12639757990837097, + "learning_rate": 0.00019682474226804123, + "loss": 0.712, + "step": 228 + }, + { + "epoch": 0.07328, + "grad_norm": 0.13229860365390778, + "learning_rate": 0.0001967835051546392, + "loss": 0.7447, + "step": 229 + }, + { + "epoch": 0.0736, + "grad_norm": 0.11198769509792328, + "learning_rate": 0.00019674226804123712, + "loss": 0.7329, + "step": 230 + }, + { + "epoch": 0.07392, + "grad_norm": 0.11425165086984634, + "learning_rate": 0.00019670103092783505, + "loss": 0.868, + "step": 231 + }, + { + "epoch": 0.07424, + "grad_norm": 0.12034732848405838, + "learning_rate": 0.000196659793814433, + "loss": 0.7276, + "step": 232 + }, + { + "epoch": 0.07456, + "grad_norm": 0.10866102576255798, + "learning_rate": 0.00019661855670103094, + "loss": 0.7338, + "step": 233 + }, + { + "epoch": 0.07488, + "grad_norm": 0.12432979047298431, + "learning_rate": 0.00019657731958762887, + "loss": 0.8257, + "step": 234 + }, + { + "epoch": 0.0752, + "grad_norm": 0.12356464564800262, + "learning_rate": 0.0001965360824742268, + "loss": 0.6813, + "step": 235 + }, + { + "epoch": 0.07552, + "grad_norm": 0.1002034917473793, + "learning_rate": 0.00019649484536082474, + "loss": 0.7406, + "step": 236 + }, + { + "epoch": 0.07584, + "grad_norm": 0.11528842151165009, + "learning_rate": 0.0001964536082474227, + "loss": 0.7597, + "step": 237 + }, + { + "epoch": 0.07616, + "grad_norm": 0.11711521446704865, + "learning_rate": 0.00019641237113402063, + "loss": 0.6925, + "step": 238 + }, + { + "epoch": 0.07648, + "grad_norm": 0.09948590397834778, + "learning_rate": 0.00019637113402061859, + "loss": 0.8017, + "step": 239 + }, + { + "epoch": 0.0768, + "grad_norm": 0.11264682561159134, + "learning_rate": 0.00019632989690721652, + "loss": 0.6871, + "step": 240 + }, + { + "epoch": 0.07712, + "grad_norm": 0.12238943576812744, + "learning_rate": 0.00019628865979381442, + "loss": 0.7113, + "step": 241 + }, + { + "epoch": 0.07744, + "grad_norm": 0.11449657380580902, + "learning_rate": 0.00019624742268041238, + "loss": 0.7144, + "step": 242 + }, + { + "epoch": 0.07776, + "grad_norm": 0.12212596088647842, + "learning_rate": 0.0001962061855670103, + "loss": 0.697, + "step": 243 + }, + { + "epoch": 0.07808, + "grad_norm": 0.12231374531984329, + "learning_rate": 0.00019616494845360827, + "loss": 0.6955, + "step": 244 + }, + { + "epoch": 0.0784, + "grad_norm": 0.11105109006166458, + "learning_rate": 0.0001961237113402062, + "loss": 0.8251, + "step": 245 + }, + { + "epoch": 0.07872, + "grad_norm": 0.12369677424430847, + "learning_rate": 0.00019608247422680413, + "loss": 0.8814, + "step": 246 + }, + { + "epoch": 0.07904, + "grad_norm": 0.12226764857769012, + "learning_rate": 0.0001960412371134021, + "loss": 0.7408, + "step": 247 + }, + { + "epoch": 0.07936, + "grad_norm": 0.10316195338964462, + "learning_rate": 0.000196, + "loss": 0.792, + "step": 248 + }, + { + "epoch": 0.07968, + "grad_norm": 0.12036872655153275, + "learning_rate": 0.00019595876288659795, + "loss": 0.6, + "step": 249 + }, + { + "epoch": 0.08, + "grad_norm": 0.1140509843826294, + "learning_rate": 0.00019591752577319588, + "loss": 0.7599, + "step": 250 + }, + { + "epoch": 0.08032, + "grad_norm": 0.14200666546821594, + "learning_rate": 0.00019587628865979381, + "loss": 0.7723, + "step": 251 + }, + { + "epoch": 0.08064, + "grad_norm": 0.11529050767421722, + "learning_rate": 0.00019583505154639177, + "loss": 0.8305, + "step": 252 + }, + { + "epoch": 0.08096, + "grad_norm": 0.11324192583560944, + "learning_rate": 0.0001957938144329897, + "loss": 0.7913, + "step": 253 + }, + { + "epoch": 0.08128, + "grad_norm": 0.12280263006687164, + "learning_rate": 0.00019575257731958764, + "loss": 0.8388, + "step": 254 + }, + { + "epoch": 0.0816, + "grad_norm": 0.10810542851686478, + "learning_rate": 0.00019571134020618557, + "loss": 0.7025, + "step": 255 + }, + { + "epoch": 0.08192, + "grad_norm": 0.11314940452575684, + "learning_rate": 0.0001956701030927835, + "loss": 0.6696, + "step": 256 + }, + { + "epoch": 0.08224, + "grad_norm": 0.10368701070547104, + "learning_rate": 0.00019562886597938146, + "loss": 0.581, + "step": 257 + }, + { + "epoch": 0.08256, + "grad_norm": 0.11146122217178345, + "learning_rate": 0.0001955876288659794, + "loss": 0.8221, + "step": 258 + }, + { + "epoch": 0.08288, + "grad_norm": 0.11592487245798111, + "learning_rate": 0.00019554639175257732, + "loss": 0.8082, + "step": 259 + }, + { + "epoch": 0.0832, + "grad_norm": 0.11007897555828094, + "learning_rate": 0.00019550515463917528, + "loss": 0.8943, + "step": 260 + }, + { + "epoch": 0.08352, + "grad_norm": 0.11887793987989426, + "learning_rate": 0.0001954639175257732, + "loss": 0.7447, + "step": 261 + }, + { + "epoch": 0.08384, + "grad_norm": 0.10715821385383606, + "learning_rate": 0.00019542268041237114, + "loss": 0.6405, + "step": 262 + }, + { + "epoch": 0.08416, + "grad_norm": 0.11692032217979431, + "learning_rate": 0.00019538144329896907, + "loss": 0.863, + "step": 263 + }, + { + "epoch": 0.08448, + "grad_norm": 0.11917733401060104, + "learning_rate": 0.000195340206185567, + "loss": 0.6976, + "step": 264 + }, + { + "epoch": 0.0848, + "grad_norm": 0.1289709061384201, + "learning_rate": 0.00019529896907216496, + "loss": 0.9149, + "step": 265 + }, + { + "epoch": 0.08512, + "grad_norm": 0.10539647191762924, + "learning_rate": 0.0001952577319587629, + "loss": 0.8449, + "step": 266 + }, + { + "epoch": 0.08544, + "grad_norm": 0.10908807814121246, + "learning_rate": 0.00019521649484536085, + "loss": 0.7162, + "step": 267 + }, + { + "epoch": 0.08576, + "grad_norm": 0.10654058307409286, + "learning_rate": 0.00019517525773195878, + "loss": 0.6285, + "step": 268 + }, + { + "epoch": 0.08608, + "grad_norm": 0.111301951110363, + "learning_rate": 0.0001951340206185567, + "loss": 0.755, + "step": 269 + }, + { + "epoch": 0.0864, + "grad_norm": 0.10679474472999573, + "learning_rate": 0.00019509278350515464, + "loss": 0.7782, + "step": 270 + }, + { + "epoch": 0.08672, + "grad_norm": 0.10793033987283707, + "learning_rate": 0.00019505154639175258, + "loss": 0.7021, + "step": 271 + }, + { + "epoch": 0.08704, + "grad_norm": 0.13048402965068817, + "learning_rate": 0.00019501030927835053, + "loss": 0.7575, + "step": 272 + }, + { + "epoch": 0.08736, + "grad_norm": 0.11437942832708359, + "learning_rate": 0.00019496907216494847, + "loss": 0.7925, + "step": 273 + }, + { + "epoch": 0.08768, + "grad_norm": 0.1390543431043625, + "learning_rate": 0.0001949278350515464, + "loss": 0.8013, + "step": 274 + }, + { + "epoch": 0.088, + "grad_norm": 0.1289997398853302, + "learning_rate": 0.00019488659793814435, + "loss": 0.6053, + "step": 275 + }, + { + "epoch": 0.08832, + "grad_norm": 0.09682965278625488, + "learning_rate": 0.00019484536082474229, + "loss": 0.8756, + "step": 276 + }, + { + "epoch": 0.08864, + "grad_norm": 0.12752555310726166, + "learning_rate": 0.00019480412371134022, + "loss": 0.6314, + "step": 277 + }, + { + "epoch": 0.08896, + "grad_norm": 0.11673011630773544, + "learning_rate": 0.00019476288659793815, + "loss": 0.8842, + "step": 278 + }, + { + "epoch": 0.08928, + "grad_norm": 0.108588308095932, + "learning_rate": 0.00019472164948453608, + "loss": 0.6348, + "step": 279 + }, + { + "epoch": 0.0896, + "grad_norm": 0.1195591390132904, + "learning_rate": 0.00019468041237113404, + "loss": 0.8097, + "step": 280 + }, + { + "epoch": 0.08992, + "grad_norm": 0.11936333775520325, + "learning_rate": 0.00019463917525773197, + "loss": 0.7858, + "step": 281 + }, + { + "epoch": 0.09024, + "grad_norm": 0.11627551168203354, + "learning_rate": 0.00019459793814432993, + "loss": 0.7322, + "step": 282 + }, + { + "epoch": 0.09056, + "grad_norm": 0.10832615196704865, + "learning_rate": 0.00019455670103092786, + "loss": 0.6215, + "step": 283 + }, + { + "epoch": 0.09088, + "grad_norm": 0.11125192046165466, + "learning_rate": 0.00019451546391752576, + "loss": 0.8058, + "step": 284 + }, + { + "epoch": 0.0912, + "grad_norm": 0.11165905743837357, + "learning_rate": 0.00019447422680412372, + "loss": 0.7104, + "step": 285 + }, + { + "epoch": 0.09152, + "grad_norm": 0.10910595208406448, + "learning_rate": 0.00019443298969072165, + "loss": 0.707, + "step": 286 + }, + { + "epoch": 0.09184, + "grad_norm": 0.11235277354717255, + "learning_rate": 0.0001943917525773196, + "loss": 0.8161, + "step": 287 + }, + { + "epoch": 0.09216, + "grad_norm": 0.11668789386749268, + "learning_rate": 0.00019435051546391754, + "loss": 0.6641, + "step": 288 + }, + { + "epoch": 0.09248, + "grad_norm": 0.11274772882461548, + "learning_rate": 0.00019430927835051547, + "loss": 0.6987, + "step": 289 + }, + { + "epoch": 0.0928, + "grad_norm": 0.12356577813625336, + "learning_rate": 0.0001942680412371134, + "loss": 0.7051, + "step": 290 + }, + { + "epoch": 0.09312, + "grad_norm": 0.0987682193517685, + "learning_rate": 0.00019422680412371134, + "loss": 0.7218, + "step": 291 + }, + { + "epoch": 0.09344, + "grad_norm": 0.12354995310306549, + "learning_rate": 0.00019418556701030927, + "loss": 0.7379, + "step": 292 + }, + { + "epoch": 0.09376, + "grad_norm": 0.1022811159491539, + "learning_rate": 0.00019414432989690723, + "loss": 0.65, + "step": 293 + }, + { + "epoch": 0.09408, + "grad_norm": 0.11507881432771683, + "learning_rate": 0.00019410309278350516, + "loss": 0.8681, + "step": 294 + }, + { + "epoch": 0.0944, + "grad_norm": 0.12279893457889557, + "learning_rate": 0.00019406185567010312, + "loss": 0.6837, + "step": 295 + }, + { + "epoch": 0.09472, + "grad_norm": 0.10548900812864304, + "learning_rate": 0.00019402061855670105, + "loss": 0.669, + "step": 296 + }, + { + "epoch": 0.09504, + "grad_norm": 0.1302396059036255, + "learning_rate": 0.00019397938144329898, + "loss": 0.7969, + "step": 297 + }, + { + "epoch": 0.09536, + "grad_norm": 0.11492603272199631, + "learning_rate": 0.0001939381443298969, + "loss": 1.0184, + "step": 298 + }, + { + "epoch": 0.09568, + "grad_norm": 0.131149023771286, + "learning_rate": 0.00019389690721649484, + "loss": 0.8521, + "step": 299 + }, + { + "epoch": 0.096, + "grad_norm": 0.1252460777759552, + "learning_rate": 0.0001938556701030928, + "loss": 0.8791, + "step": 300 + }, + { + "epoch": 0.09632, + "grad_norm": 0.11322636157274246, + "learning_rate": 0.00019381443298969073, + "loss": 0.7305, + "step": 301 + }, + { + "epoch": 0.09664, + "grad_norm": 0.10373307764530182, + "learning_rate": 0.00019377319587628866, + "loss": 0.7844, + "step": 302 + }, + { + "epoch": 0.09696, + "grad_norm": 0.13182489573955536, + "learning_rate": 0.00019373195876288662, + "loss": 0.6581, + "step": 303 + }, + { + "epoch": 0.09728, + "grad_norm": 0.1276710480451584, + "learning_rate": 0.00019369072164948455, + "loss": 0.6664, + "step": 304 + }, + { + "epoch": 0.0976, + "grad_norm": 0.10770755261182785, + "learning_rate": 0.00019364948453608248, + "loss": 0.6315, + "step": 305 + }, + { + "epoch": 0.09792, + "grad_norm": 0.11755330115556717, + "learning_rate": 0.00019360824742268041, + "loss": 0.7562, + "step": 306 + }, + { + "epoch": 0.09824, + "grad_norm": 0.1252545267343521, + "learning_rate": 0.00019356701030927835, + "loss": 0.6606, + "step": 307 + }, + { + "epoch": 0.09856, + "grad_norm": 0.12284315377473831, + "learning_rate": 0.0001935257731958763, + "loss": 0.8037, + "step": 308 + }, + { + "epoch": 0.09888, + "grad_norm": 0.106113962829113, + "learning_rate": 0.00019348453608247424, + "loss": 0.6797, + "step": 309 + }, + { + "epoch": 0.0992, + "grad_norm": 0.09285733103752136, + "learning_rate": 0.0001934432989690722, + "loss": 0.6927, + "step": 310 + }, + { + "epoch": 0.09952, + "grad_norm": 0.10903103649616241, + "learning_rate": 0.00019340206185567012, + "loss": 0.7863, + "step": 311 + }, + { + "epoch": 0.09984, + "grad_norm": 0.12015190720558167, + "learning_rate": 0.00019336082474226806, + "loss": 0.6961, + "step": 312 + }, + { + "epoch": 0.10016, + "grad_norm": 0.12260061502456665, + "learning_rate": 0.000193319587628866, + "loss": 0.7965, + "step": 313 + }, + { + "epoch": 0.10048, + "grad_norm": 0.10297476500272751, + "learning_rate": 0.00019327835051546392, + "loss": 0.6935, + "step": 314 + }, + { + "epoch": 0.1008, + "grad_norm": 0.12614290416240692, + "learning_rate": 0.00019323711340206188, + "loss": 0.8751, + "step": 315 + }, + { + "epoch": 0.10112, + "grad_norm": 0.10241006314754486, + "learning_rate": 0.0001931958762886598, + "loss": 0.5642, + "step": 316 + }, + { + "epoch": 0.10144, + "grad_norm": 0.1053333505988121, + "learning_rate": 0.00019315463917525774, + "loss": 0.6425, + "step": 317 + }, + { + "epoch": 0.10176, + "grad_norm": 0.1024412289261818, + "learning_rate": 0.0001931134020618557, + "loss": 0.6414, + "step": 318 + }, + { + "epoch": 0.10208, + "grad_norm": 0.10628654062747955, + "learning_rate": 0.0001930721649484536, + "loss": 0.8259, + "step": 319 + }, + { + "epoch": 0.1024, + "grad_norm": 0.1291695237159729, + "learning_rate": 0.00019303092783505153, + "loss": 0.6887, + "step": 320 + }, + { + "epoch": 0.10272, + "grad_norm": 0.11315542459487915, + "learning_rate": 0.0001929896907216495, + "loss": 0.6316, + "step": 321 + }, + { + "epoch": 0.10304, + "grad_norm": 0.11541493982076645, + "learning_rate": 0.00019294845360824742, + "loss": 0.8023, + "step": 322 + }, + { + "epoch": 0.10336, + "grad_norm": 0.09769251197576523, + "learning_rate": 0.00019290721649484538, + "loss": 0.674, + "step": 323 + }, + { + "epoch": 0.10368, + "grad_norm": 0.10549357533454895, + "learning_rate": 0.0001928659793814433, + "loss": 0.7102, + "step": 324 + }, + { + "epoch": 0.104, + "grad_norm": 0.1344941407442093, + "learning_rate": 0.00019282474226804124, + "loss": 0.8861, + "step": 325 + }, + { + "epoch": 0.10432, + "grad_norm": 0.13625279068946838, + "learning_rate": 0.00019278350515463918, + "loss": 0.8333, + "step": 326 + }, + { + "epoch": 0.10464, + "grad_norm": 0.1136624813079834, + "learning_rate": 0.0001927422680412371, + "loss": 0.7187, + "step": 327 + }, + { + "epoch": 0.10496, + "grad_norm": 0.12351350486278534, + "learning_rate": 0.00019270103092783506, + "loss": 0.6361, + "step": 328 + }, + { + "epoch": 0.10528, + "grad_norm": 0.12927868962287903, + "learning_rate": 0.000192659793814433, + "loss": 0.6716, + "step": 329 + }, + { + "epoch": 0.1056, + "grad_norm": 0.11619272083044052, + "learning_rate": 0.00019261855670103093, + "loss": 0.8108, + "step": 330 + }, + { + "epoch": 0.10592, + "grad_norm": 0.12765595316886902, + "learning_rate": 0.00019257731958762889, + "loss": 0.77, + "step": 331 + }, + { + "epoch": 0.10624, + "grad_norm": 0.13606730103492737, + "learning_rate": 0.00019253608247422682, + "loss": 0.7572, + "step": 332 + }, + { + "epoch": 0.10656, + "grad_norm": 0.1074293777346611, + "learning_rate": 0.00019249484536082475, + "loss": 0.8137, + "step": 333 + }, + { + "epoch": 0.10688, + "grad_norm": 0.12371189147233963, + "learning_rate": 0.00019245360824742268, + "loss": 0.7414, + "step": 334 + }, + { + "epoch": 0.1072, + "grad_norm": 0.1253589391708374, + "learning_rate": 0.0001924123711340206, + "loss": 0.6796, + "step": 335 + }, + { + "epoch": 0.10752, + "grad_norm": 0.11865826696157455, + "learning_rate": 0.00019237113402061857, + "loss": 0.7051, + "step": 336 + }, + { + "epoch": 0.10784, + "grad_norm": 0.12180513888597488, + "learning_rate": 0.0001923298969072165, + "loss": 0.8126, + "step": 337 + }, + { + "epoch": 0.10816, + "grad_norm": 0.11744590848684311, + "learning_rate": 0.00019228865979381446, + "loss": 0.7553, + "step": 338 + }, + { + "epoch": 0.10848, + "grad_norm": 0.11796604096889496, + "learning_rate": 0.0001922474226804124, + "loss": 0.814, + "step": 339 + }, + { + "epoch": 0.1088, + "grad_norm": 0.125954270362854, + "learning_rate": 0.00019220618556701032, + "loss": 0.64, + "step": 340 + }, + { + "epoch": 0.10912, + "grad_norm": 0.12020507454872131, + "learning_rate": 0.00019216494845360825, + "loss": 0.6175, + "step": 341 + }, + { + "epoch": 0.10944, + "grad_norm": 0.10076280683279037, + "learning_rate": 0.00019212371134020618, + "loss": 0.6176, + "step": 342 + }, + { + "epoch": 0.10976, + "grad_norm": 0.11676807701587677, + "learning_rate": 0.00019208247422680414, + "loss": 0.677, + "step": 343 + }, + { + "epoch": 0.11008, + "grad_norm": 0.11364617198705673, + "learning_rate": 0.00019204123711340207, + "loss": 0.5619, + "step": 344 + }, + { + "epoch": 0.1104, + "grad_norm": 0.10261007398366928, + "learning_rate": 0.000192, + "loss": 0.614, + "step": 345 + }, + { + "epoch": 0.11072, + "grad_norm": 0.09645744413137436, + "learning_rate": 0.00019195876288659796, + "loss": 0.7274, + "step": 346 + }, + { + "epoch": 0.11104, + "grad_norm": 0.12451738864183426, + "learning_rate": 0.0001919175257731959, + "loss": 0.8735, + "step": 347 + }, + { + "epoch": 0.11136, + "grad_norm": 0.11062555015087128, + "learning_rate": 0.00019187628865979383, + "loss": 0.6268, + "step": 348 + }, + { + "epoch": 0.11168, + "grad_norm": 0.1201324462890625, + "learning_rate": 0.00019183505154639176, + "loss": 0.793, + "step": 349 + }, + { + "epoch": 0.112, + "grad_norm": 0.12797708809375763, + "learning_rate": 0.0001917938144329897, + "loss": 0.6075, + "step": 350 + }, + { + "epoch": 0.11232, + "grad_norm": 0.11388446390628815, + "learning_rate": 0.00019175257731958765, + "loss": 0.7343, + "step": 351 + }, + { + "epoch": 0.11264, + "grad_norm": 0.11379895359277725, + "learning_rate": 0.00019171134020618558, + "loss": 0.6767, + "step": 352 + }, + { + "epoch": 0.11296, + "grad_norm": 0.11296231299638748, + "learning_rate": 0.0001916701030927835, + "loss": 0.7473, + "step": 353 + }, + { + "epoch": 0.11328, + "grad_norm": 0.1344781517982483, + "learning_rate": 0.00019162886597938147, + "loss": 0.6533, + "step": 354 + }, + { + "epoch": 0.1136, + "grad_norm": 0.12438254803419113, + "learning_rate": 0.00019158762886597937, + "loss": 0.7511, + "step": 355 + }, + { + "epoch": 0.11392, + "grad_norm": 0.11343064159154892, + "learning_rate": 0.00019154639175257733, + "loss": 0.6876, + "step": 356 + }, + { + "epoch": 0.11424, + "grad_norm": 0.11751387268304825, + "learning_rate": 0.00019150515463917526, + "loss": 0.7198, + "step": 357 + }, + { + "epoch": 0.11456, + "grad_norm": 0.1194872185587883, + "learning_rate": 0.0001914639175257732, + "loss": 0.6149, + "step": 358 + }, + { + "epoch": 0.11488, + "grad_norm": 0.11200791597366333, + "learning_rate": 0.00019142268041237115, + "loss": 0.8605, + "step": 359 + }, + { + "epoch": 0.1152, + "grad_norm": 0.11708471924066544, + "learning_rate": 0.00019138144329896908, + "loss": 0.6603, + "step": 360 + }, + { + "epoch": 0.11552, + "grad_norm": 0.116248220205307, + "learning_rate": 0.00019134020618556704, + "loss": 0.7141, + "step": 361 + }, + { + "epoch": 0.11584, + "grad_norm": 0.11516683548688889, + "learning_rate": 0.00019129896907216494, + "loss": 0.7316, + "step": 362 + }, + { + "epoch": 0.11616, + "grad_norm": 0.10961434990167618, + "learning_rate": 0.00019125773195876288, + "loss": 0.6173, + "step": 363 + }, + { + "epoch": 0.11648, + "grad_norm": 0.12037073075771332, + "learning_rate": 0.00019121649484536083, + "loss": 0.6659, + "step": 364 + }, + { + "epoch": 0.1168, + "grad_norm": 0.12053418904542923, + "learning_rate": 0.00019117525773195877, + "loss": 0.6544, + "step": 365 + }, + { + "epoch": 0.11712, + "grad_norm": 0.12387484312057495, + "learning_rate": 0.00019113402061855672, + "loss": 0.6406, + "step": 366 + }, + { + "epoch": 0.11744, + "grad_norm": 0.12213553488254547, + "learning_rate": 0.00019109278350515466, + "loss": 0.6592, + "step": 367 + }, + { + "epoch": 0.11776, + "grad_norm": 0.1237708106637001, + "learning_rate": 0.0001910515463917526, + "loss": 0.779, + "step": 368 + }, + { + "epoch": 0.11808, + "grad_norm": 0.10254320502281189, + "learning_rate": 0.00019101030927835052, + "loss": 0.7242, + "step": 369 + }, + { + "epoch": 0.1184, + "grad_norm": 0.11179213225841522, + "learning_rate": 0.00019096907216494845, + "loss": 0.765, + "step": 370 + }, + { + "epoch": 0.11872, + "grad_norm": 0.11824121326208115, + "learning_rate": 0.0001909278350515464, + "loss": 0.6971, + "step": 371 + }, + { + "epoch": 0.11904, + "grad_norm": 0.13179510831832886, + "learning_rate": 0.00019088659793814434, + "loss": 0.7179, + "step": 372 + }, + { + "epoch": 0.11936, + "grad_norm": 0.1161201074719429, + "learning_rate": 0.00019084536082474227, + "loss": 0.6249, + "step": 373 + }, + { + "epoch": 0.11968, + "grad_norm": 0.12930230796337128, + "learning_rate": 0.00019080412371134023, + "loss": 0.6636, + "step": 374 + }, + { + "epoch": 0.12, + "grad_norm": 0.12602809071540833, + "learning_rate": 0.00019076288659793816, + "loss": 0.7044, + "step": 375 + }, + { + "epoch": 0.12032, + "grad_norm": 0.1393459141254425, + "learning_rate": 0.0001907216494845361, + "loss": 0.6518, + "step": 376 + }, + { + "epoch": 0.12064, + "grad_norm": 0.11460117250680923, + "learning_rate": 0.00019068041237113402, + "loss": 0.7716, + "step": 377 + }, + { + "epoch": 0.12096, + "grad_norm": 0.14296174049377441, + "learning_rate": 0.00019063917525773195, + "loss": 0.7899, + "step": 378 + }, + { + "epoch": 0.12128, + "grad_norm": 0.14445006847381592, + "learning_rate": 0.0001905979381443299, + "loss": 0.7617, + "step": 379 + }, + { + "epoch": 0.1216, + "grad_norm": 0.12754163146018982, + "learning_rate": 0.00019055670103092784, + "loss": 0.7888, + "step": 380 + }, + { + "epoch": 0.12192, + "grad_norm": 0.11828102916479111, + "learning_rate": 0.00019051546391752577, + "loss": 0.6641, + "step": 381 + }, + { + "epoch": 0.12224, + "grad_norm": 0.11433565616607666, + "learning_rate": 0.00019047422680412373, + "loss": 0.5984, + "step": 382 + }, + { + "epoch": 0.12256, + "grad_norm": 0.11538711935281754, + "learning_rate": 0.00019043298969072166, + "loss": 0.7464, + "step": 383 + }, + { + "epoch": 0.12288, + "grad_norm": 0.10493431240320206, + "learning_rate": 0.0001903917525773196, + "loss": 0.7104, + "step": 384 + }, + { + "epoch": 0.1232, + "grad_norm": 0.13114306330680847, + "learning_rate": 0.00019035051546391753, + "loss": 0.9104, + "step": 385 + }, + { + "epoch": 0.12352, + "grad_norm": 0.11584331095218658, + "learning_rate": 0.00019030927835051546, + "loss": 0.7205, + "step": 386 + }, + { + "epoch": 0.12384, + "grad_norm": 0.1123526394367218, + "learning_rate": 0.00019026804123711342, + "loss": 0.7061, + "step": 387 + }, + { + "epoch": 0.12416, + "grad_norm": 0.11471172422170639, + "learning_rate": 0.00019022680412371135, + "loss": 0.7491, + "step": 388 + }, + { + "epoch": 0.12448, + "grad_norm": 0.1361629068851471, + "learning_rate": 0.0001901855670103093, + "loss": 0.8168, + "step": 389 + }, + { + "epoch": 0.1248, + "grad_norm": 0.1272808462381363, + "learning_rate": 0.00019014432989690724, + "loss": 0.6907, + "step": 390 + }, + { + "epoch": 0.12512, + "grad_norm": 0.12672847509384155, + "learning_rate": 0.00019010309278350514, + "loss": 0.7511, + "step": 391 + }, + { + "epoch": 0.12544, + "grad_norm": 0.12205267697572708, + "learning_rate": 0.0001900618556701031, + "loss": 0.8164, + "step": 392 + }, + { + "epoch": 0.12576, + "grad_norm": 0.12806954979896545, + "learning_rate": 0.00019002061855670103, + "loss": 0.8247, + "step": 393 + }, + { + "epoch": 0.12608, + "grad_norm": 0.11856722086668015, + "learning_rate": 0.000189979381443299, + "loss": 0.7324, + "step": 394 + }, + { + "epoch": 0.1264, + "grad_norm": 0.11742661893367767, + "learning_rate": 0.00018993814432989692, + "loss": 0.8151, + "step": 395 + }, + { + "epoch": 0.12672, + "grad_norm": 0.11036640405654907, + "learning_rate": 0.00018989690721649485, + "loss": 0.5336, + "step": 396 + }, + { + "epoch": 0.12704, + "grad_norm": 0.11546391993761063, + "learning_rate": 0.0001898556701030928, + "loss": 0.8078, + "step": 397 + }, + { + "epoch": 0.12736, + "grad_norm": 0.11293994635343552, + "learning_rate": 0.00018981443298969071, + "loss": 0.7818, + "step": 398 + }, + { + "epoch": 0.12768, + "grad_norm": 0.11465566605329514, + "learning_rate": 0.00018977319587628867, + "loss": 0.8291, + "step": 399 + }, + { + "epoch": 0.128, + "grad_norm": 0.1251228302717209, + "learning_rate": 0.0001897319587628866, + "loss": 0.7199, + "step": 400 + }, + { + "epoch": 0.12832, + "grad_norm": 0.11204738169908524, + "learning_rate": 0.00018969072164948454, + "loss": 0.7193, + "step": 401 + }, + { + "epoch": 0.12864, + "grad_norm": 0.1072632372379303, + "learning_rate": 0.0001896494845360825, + "loss": 0.7763, + "step": 402 + }, + { + "epoch": 0.12896, + "grad_norm": 0.11516649276018143, + "learning_rate": 0.00018960824742268043, + "loss": 0.6853, + "step": 403 + }, + { + "epoch": 0.12928, + "grad_norm": 0.1176360473036766, + "learning_rate": 0.00018956701030927836, + "loss": 0.6184, + "step": 404 + }, + { + "epoch": 0.1296, + "grad_norm": 0.11547468602657318, + "learning_rate": 0.0001895257731958763, + "loss": 0.8048, + "step": 405 + }, + { + "epoch": 0.12992, + "grad_norm": 0.11308492720127106, + "learning_rate": 0.00018948453608247422, + "loss": 0.7213, + "step": 406 + }, + { + "epoch": 0.13024, + "grad_norm": 0.11973299831151962, + "learning_rate": 0.00018944329896907218, + "loss": 0.8809, + "step": 407 + }, + { + "epoch": 0.13056, + "grad_norm": 0.1213008239865303, + "learning_rate": 0.0001894020618556701, + "loss": 0.7794, + "step": 408 + }, + { + "epoch": 0.13088, + "grad_norm": 0.13353341817855835, + "learning_rate": 0.00018936082474226804, + "loss": 0.6674, + "step": 409 + }, + { + "epoch": 0.1312, + "grad_norm": 0.1155238002538681, + "learning_rate": 0.000189319587628866, + "loss": 0.7668, + "step": 410 + }, + { + "epoch": 0.13152, + "grad_norm": 0.13428455591201782, + "learning_rate": 0.00018927835051546393, + "loss": 0.6864, + "step": 411 + }, + { + "epoch": 0.13184, + "grad_norm": 0.13777495920658112, + "learning_rate": 0.00018923711340206186, + "loss": 0.8271, + "step": 412 + }, + { + "epoch": 0.13216, + "grad_norm": 0.12846726179122925, + "learning_rate": 0.0001891958762886598, + "loss": 0.9027, + "step": 413 + }, + { + "epoch": 0.13248, + "grad_norm": 0.11394570022821426, + "learning_rate": 0.00018915463917525772, + "loss": 0.7531, + "step": 414 + }, + { + "epoch": 0.1328, + "grad_norm": 0.11797945201396942, + "learning_rate": 0.00018911340206185568, + "loss": 0.7393, + "step": 415 + }, + { + "epoch": 0.13312, + "grad_norm": 0.11778296530246735, + "learning_rate": 0.0001890721649484536, + "loss": 0.8172, + "step": 416 + }, + { + "epoch": 0.13344, + "grad_norm": 0.12857432663440704, + "learning_rate": 0.00018903092783505157, + "loss": 0.5693, + "step": 417 + }, + { + "epoch": 0.13376, + "grad_norm": 0.1395915150642395, + "learning_rate": 0.0001889896907216495, + "loss": 0.7187, + "step": 418 + }, + { + "epoch": 0.13408, + "grad_norm": 0.12571577727794647, + "learning_rate": 0.00018894845360824743, + "loss": 0.7347, + "step": 419 + }, + { + "epoch": 0.1344, + "grad_norm": 0.11062180995941162, + "learning_rate": 0.00018890721649484537, + "loss": 0.8797, + "step": 420 + }, + { + "epoch": 0.13472, + "grad_norm": 0.12011471390724182, + "learning_rate": 0.0001888659793814433, + "loss": 0.7638, + "step": 421 + }, + { + "epoch": 0.13504, + "grad_norm": 0.11402884870767593, + "learning_rate": 0.00018882474226804126, + "loss": 0.6198, + "step": 422 + }, + { + "epoch": 0.13536, + "grad_norm": 0.11654789745807648, + "learning_rate": 0.00018878350515463919, + "loss": 0.8276, + "step": 423 + }, + { + "epoch": 0.13568, + "grad_norm": 0.110030896961689, + "learning_rate": 0.00018874226804123712, + "loss": 0.7828, + "step": 424 + }, + { + "epoch": 0.136, + "grad_norm": 0.1173168271780014, + "learning_rate": 0.00018870103092783508, + "loss": 0.6445, + "step": 425 + }, + { + "epoch": 0.13632, + "grad_norm": 0.11195468157529831, + "learning_rate": 0.000188659793814433, + "loss": 0.7336, + "step": 426 + }, + { + "epoch": 0.13664, + "grad_norm": 0.1138012707233429, + "learning_rate": 0.00018861855670103094, + "loss": 0.7085, + "step": 427 + }, + { + "epoch": 0.13696, + "grad_norm": 0.11757896840572357, + "learning_rate": 0.00018857731958762887, + "loss": 0.593, + "step": 428 + }, + { + "epoch": 0.13728, + "grad_norm": 0.13230633735656738, + "learning_rate": 0.0001885360824742268, + "loss": 0.705, + "step": 429 + }, + { + "epoch": 0.1376, + "grad_norm": 0.13503113389015198, + "learning_rate": 0.00018849484536082476, + "loss": 0.7134, + "step": 430 + }, + { + "epoch": 0.13792, + "grad_norm": 0.11426805704832077, + "learning_rate": 0.0001884536082474227, + "loss": 0.6781, + "step": 431 + }, + { + "epoch": 0.13824, + "grad_norm": 0.11934298276901245, + "learning_rate": 0.00018841237113402065, + "loss": 0.7727, + "step": 432 + }, + { + "epoch": 0.13856, + "grad_norm": 0.12700718641281128, + "learning_rate": 0.00018837113402061858, + "loss": 0.7153, + "step": 433 + }, + { + "epoch": 0.13888, + "grad_norm": 0.15089921653270721, + "learning_rate": 0.00018832989690721648, + "loss": 0.7326, + "step": 434 + }, + { + "epoch": 0.1392, + "grad_norm": 0.10799750685691833, + "learning_rate": 0.00018828865979381444, + "loss": 0.7738, + "step": 435 + }, + { + "epoch": 0.13952, + "grad_norm": 0.1076594665646553, + "learning_rate": 0.00018824742268041237, + "loss": 0.545, + "step": 436 + }, + { + "epoch": 0.13984, + "grad_norm": 0.1204753890633583, + "learning_rate": 0.00018820618556701033, + "loss": 0.73, + "step": 437 + }, + { + "epoch": 0.14016, + "grad_norm": 0.11598300188779831, + "learning_rate": 0.00018816494845360826, + "loss": 0.635, + "step": 438 + }, + { + "epoch": 0.14048, + "grad_norm": 0.11010400205850601, + "learning_rate": 0.0001881237113402062, + "loss": 0.7891, + "step": 439 + }, + { + "epoch": 0.1408, + "grad_norm": 0.12184228003025055, + "learning_rate": 0.00018808247422680413, + "loss": 0.7332, + "step": 440 + }, + { + "epoch": 0.14112, + "grad_norm": 0.11696261167526245, + "learning_rate": 0.00018804123711340206, + "loss": 0.8299, + "step": 441 + }, + { + "epoch": 0.14144, + "grad_norm": 0.11464802920818329, + "learning_rate": 0.000188, + "loss": 0.7107, + "step": 442 + }, + { + "epoch": 0.14176, + "grad_norm": 0.12120334804058075, + "learning_rate": 0.00018795876288659795, + "loss": 0.7729, + "step": 443 + }, + { + "epoch": 0.14208, + "grad_norm": 0.11571822315454483, + "learning_rate": 0.00018791752577319588, + "loss": 0.6551, + "step": 444 + }, + { + "epoch": 0.1424, + "grad_norm": 0.11872228980064392, + "learning_rate": 0.00018787628865979384, + "loss": 0.698, + "step": 445 + }, + { + "epoch": 0.14272, + "grad_norm": 0.12505149841308594, + "learning_rate": 0.00018783505154639177, + "loss": 0.751, + "step": 446 + }, + { + "epoch": 0.14304, + "grad_norm": 0.10618813335895538, + "learning_rate": 0.0001877938144329897, + "loss": 0.7299, + "step": 447 + }, + { + "epoch": 0.14336, + "grad_norm": 0.11824540048837662, + "learning_rate": 0.00018775257731958763, + "loss": 0.6796, + "step": 448 + }, + { + "epoch": 0.14368, + "grad_norm": 0.10695328563451767, + "learning_rate": 0.00018771134020618556, + "loss": 0.7424, + "step": 449 + }, + { + "epoch": 0.144, + "grad_norm": 0.12764234840869904, + "learning_rate": 0.00018767010309278352, + "loss": 0.7301, + "step": 450 + }, + { + "epoch": 0.14432, + "grad_norm": 0.12354765832424164, + "learning_rate": 0.00018762886597938145, + "loss": 0.7945, + "step": 451 + }, + { + "epoch": 0.14464, + "grad_norm": 0.12733551859855652, + "learning_rate": 0.00018758762886597938, + "loss": 0.7565, + "step": 452 + }, + { + "epoch": 0.14496, + "grad_norm": 0.1533960998058319, + "learning_rate": 0.00018754639175257734, + "loss": 0.7964, + "step": 453 + }, + { + "epoch": 0.14528, + "grad_norm": 0.11748246848583221, + "learning_rate": 0.00018750515463917527, + "loss": 0.7715, + "step": 454 + }, + { + "epoch": 0.1456, + "grad_norm": 0.12309489399194717, + "learning_rate": 0.0001874639175257732, + "loss": 0.7758, + "step": 455 + }, + { + "epoch": 0.14592, + "grad_norm": 0.10743677616119385, + "learning_rate": 0.00018742268041237114, + "loss": 0.8084, + "step": 456 + }, + { + "epoch": 0.14624, + "grad_norm": 0.15399467945098877, + "learning_rate": 0.00018738144329896907, + "loss": 0.7689, + "step": 457 + }, + { + "epoch": 0.14656, + "grad_norm": 0.1381152868270874, + "learning_rate": 0.00018734020618556702, + "loss": 0.7902, + "step": 458 + }, + { + "epoch": 0.14688, + "grad_norm": 0.12086453288793564, + "learning_rate": 0.00018729896907216496, + "loss": 0.739, + "step": 459 + }, + { + "epoch": 0.1472, + "grad_norm": 0.14193426072597504, + "learning_rate": 0.00018725773195876291, + "loss": 0.8245, + "step": 460 + }, + { + "epoch": 0.14752, + "grad_norm": 0.1121336966753006, + "learning_rate": 0.00018721649484536085, + "loss": 0.6877, + "step": 461 + }, + { + "epoch": 0.14784, + "grad_norm": 0.12167156487703323, + "learning_rate": 0.00018717525773195878, + "loss": 0.7017, + "step": 462 + }, + { + "epoch": 0.14816, + "grad_norm": 0.1287817507982254, + "learning_rate": 0.0001871340206185567, + "loss": 0.6229, + "step": 463 + }, + { + "epoch": 0.14848, + "grad_norm": 0.128828227519989, + "learning_rate": 0.00018709278350515464, + "loss": 0.7917, + "step": 464 + }, + { + "epoch": 0.1488, + "grad_norm": 0.13771773874759674, + "learning_rate": 0.0001870515463917526, + "loss": 0.6883, + "step": 465 + }, + { + "epoch": 0.14912, + "grad_norm": 0.10570216923952103, + "learning_rate": 0.00018701030927835053, + "loss": 0.7791, + "step": 466 + }, + { + "epoch": 0.14944, + "grad_norm": 0.1250445395708084, + "learning_rate": 0.00018696907216494846, + "loss": 0.7435, + "step": 467 + }, + { + "epoch": 0.14976, + "grad_norm": 0.1064952090382576, + "learning_rate": 0.00018692783505154642, + "loss": 0.7621, + "step": 468 + }, + { + "epoch": 0.15008, + "grad_norm": 0.10514791309833527, + "learning_rate": 0.00018688659793814432, + "loss": 0.6403, + "step": 469 + }, + { + "epoch": 0.1504, + "grad_norm": 0.10278081893920898, + "learning_rate": 0.00018684536082474225, + "loss": 0.8246, + "step": 470 + }, + { + "epoch": 0.15072, + "grad_norm": 0.11813073605298996, + "learning_rate": 0.0001868041237113402, + "loss": 0.8403, + "step": 471 + }, + { + "epoch": 0.15104, + "grad_norm": 0.10654576867818832, + "learning_rate": 0.00018676288659793814, + "loss": 0.7483, + "step": 472 + }, + { + "epoch": 0.15136, + "grad_norm": 0.11642228811979294, + "learning_rate": 0.0001867216494845361, + "loss": 0.6114, + "step": 473 + }, + { + "epoch": 0.15168, + "grad_norm": 0.12933100759983063, + "learning_rate": 0.00018668041237113403, + "loss": 0.7902, + "step": 474 + }, + { + "epoch": 0.152, + "grad_norm": 0.1200195848941803, + "learning_rate": 0.00018663917525773196, + "loss": 0.7298, + "step": 475 + }, + { + "epoch": 0.15232, + "grad_norm": 0.10747142881155014, + "learning_rate": 0.0001865979381443299, + "loss": 0.7376, + "step": 476 + }, + { + "epoch": 0.15264, + "grad_norm": 0.10868332535028458, + "learning_rate": 0.00018655670103092783, + "loss": 0.7013, + "step": 477 + }, + { + "epoch": 0.15296, + "grad_norm": 0.12750379741191864, + "learning_rate": 0.00018651546391752579, + "loss": 0.7718, + "step": 478 + }, + { + "epoch": 0.15328, + "grad_norm": 0.10860800743103027, + "learning_rate": 0.00018647422680412372, + "loss": 0.5887, + "step": 479 + }, + { + "epoch": 0.1536, + "grad_norm": 0.11445030570030212, + "learning_rate": 0.00018643298969072165, + "loss": 0.7955, + "step": 480 + }, + { + "epoch": 0.15392, + "grad_norm": 0.11774427443742752, + "learning_rate": 0.0001863917525773196, + "loss": 0.7095, + "step": 481 + }, + { + "epoch": 0.15424, + "grad_norm": 0.13853546977043152, + "learning_rate": 0.00018635051546391754, + "loss": 0.9395, + "step": 482 + }, + { + "epoch": 0.15456, + "grad_norm": 0.120772585272789, + "learning_rate": 0.00018630927835051547, + "loss": 0.8, + "step": 483 + }, + { + "epoch": 0.15488, + "grad_norm": 0.13263949751853943, + "learning_rate": 0.0001862680412371134, + "loss": 0.9118, + "step": 484 + }, + { + "epoch": 0.1552, + "grad_norm": 0.12179119139909744, + "learning_rate": 0.00018622680412371133, + "loss": 0.6567, + "step": 485 + }, + { + "epoch": 0.15552, + "grad_norm": 0.1244128867983818, + "learning_rate": 0.0001861855670103093, + "loss": 0.7702, + "step": 486 + }, + { + "epoch": 0.15584, + "grad_norm": 0.12456309050321579, + "learning_rate": 0.00018614432989690722, + "loss": 0.7687, + "step": 487 + }, + { + "epoch": 0.15616, + "grad_norm": 0.12738095223903656, + "learning_rate": 0.00018610309278350518, + "loss": 0.7489, + "step": 488 + }, + { + "epoch": 0.15648, + "grad_norm": 0.11148428916931152, + "learning_rate": 0.0001860618556701031, + "loss": 0.7362, + "step": 489 + }, + { + "epoch": 0.1568, + "grad_norm": 0.1029975637793541, + "learning_rate": 0.00018602061855670104, + "loss": 0.5808, + "step": 490 + }, + { + "epoch": 0.15712, + "grad_norm": 0.12028798460960388, + "learning_rate": 0.00018597938144329897, + "loss": 0.8541, + "step": 491 + }, + { + "epoch": 0.15744, + "grad_norm": 0.12013960629701614, + "learning_rate": 0.0001859381443298969, + "loss": 0.6814, + "step": 492 + }, + { + "epoch": 0.15776, + "grad_norm": 0.12026184052228928, + "learning_rate": 0.00018589690721649486, + "loss": 0.8444, + "step": 493 + }, + { + "epoch": 0.15808, + "grad_norm": 0.11788841336965561, + "learning_rate": 0.0001858556701030928, + "loss": 0.5718, + "step": 494 + }, + { + "epoch": 0.1584, + "grad_norm": 0.11122492700815201, + "learning_rate": 0.00018581443298969073, + "loss": 0.5172, + "step": 495 + }, + { + "epoch": 0.15872, + "grad_norm": 0.12228988111019135, + "learning_rate": 0.00018577319587628868, + "loss": 0.724, + "step": 496 + }, + { + "epoch": 0.15904, + "grad_norm": 0.12170293927192688, + "learning_rate": 0.00018573195876288662, + "loss": 0.6863, + "step": 497 + }, + { + "epoch": 0.15936, + "grad_norm": 0.12947559356689453, + "learning_rate": 0.00018569072164948455, + "loss": 0.897, + "step": 498 + }, + { + "epoch": 0.15968, + "grad_norm": 0.13217338919639587, + "learning_rate": 0.00018564948453608248, + "loss": 0.7, + "step": 499 + }, + { + "epoch": 0.16, + "grad_norm": 0.12307247519493103, + "learning_rate": 0.0001856082474226804, + "loss": 0.6686, + "step": 500 + }, + { + "epoch": 0.16032, + "grad_norm": 0.12643639743328094, + "learning_rate": 0.00018556701030927837, + "loss": 0.6166, + "step": 501 + }, + { + "epoch": 0.16064, + "grad_norm": 0.10645804554224014, + "learning_rate": 0.0001855257731958763, + "loss": 0.785, + "step": 502 + }, + { + "epoch": 0.16096, + "grad_norm": 0.12195251882076263, + "learning_rate": 0.00018548453608247423, + "loss": 0.6931, + "step": 503 + }, + { + "epoch": 0.16128, + "grad_norm": 0.10586483776569366, + "learning_rate": 0.0001854432989690722, + "loss": 0.6172, + "step": 504 + }, + { + "epoch": 0.1616, + "grad_norm": 0.12337654829025269, + "learning_rate": 0.0001854020618556701, + "loss": 0.7811, + "step": 505 + }, + { + "epoch": 0.16192, + "grad_norm": 0.11356624215841293, + "learning_rate": 0.00018536082474226805, + "loss": 0.7117, + "step": 506 + }, + { + "epoch": 0.16224, + "grad_norm": 0.11793960630893707, + "learning_rate": 0.00018531958762886598, + "loss": 0.7778, + "step": 507 + }, + { + "epoch": 0.16256, + "grad_norm": 0.12677061557769775, + "learning_rate": 0.00018527835051546391, + "loss": 0.6465, + "step": 508 + }, + { + "epoch": 0.16288, + "grad_norm": 0.12064032256603241, + "learning_rate": 0.00018523711340206187, + "loss": 0.7658, + "step": 509 + }, + { + "epoch": 0.1632, + "grad_norm": 0.12942875921726227, + "learning_rate": 0.0001851958762886598, + "loss": 0.7735, + "step": 510 + }, + { + "epoch": 0.16352, + "grad_norm": 0.1350182294845581, + "learning_rate": 0.00018515463917525776, + "loss": 0.7512, + "step": 511 + }, + { + "epoch": 0.16384, + "grad_norm": 0.12585826218128204, + "learning_rate": 0.00018511340206185567, + "loss": 0.5842, + "step": 512 + }, + { + "epoch": 0.16416, + "grad_norm": 0.1233866736292839, + "learning_rate": 0.0001850721649484536, + "loss": 0.6707, + "step": 513 + }, + { + "epoch": 0.16448, + "grad_norm": 0.13590243458747864, + "learning_rate": 0.00018503092783505156, + "loss": 0.645, + "step": 514 + }, + { + "epoch": 0.1648, + "grad_norm": 0.11508861184120178, + "learning_rate": 0.0001849896907216495, + "loss": 0.698, + "step": 515 + }, + { + "epoch": 0.16512, + "grad_norm": 0.11506541818380356, + "learning_rate": 0.00018494845360824745, + "loss": 0.6834, + "step": 516 + }, + { + "epoch": 0.16544, + "grad_norm": 0.1177835762500763, + "learning_rate": 0.00018490721649484538, + "loss": 0.8327, + "step": 517 + }, + { + "epoch": 0.16576, + "grad_norm": 0.12268577516078949, + "learning_rate": 0.0001848659793814433, + "loss": 0.7925, + "step": 518 + }, + { + "epoch": 0.16608, + "grad_norm": 0.11970438808202744, + "learning_rate": 0.00018482474226804124, + "loss": 0.9855, + "step": 519 + }, + { + "epoch": 0.1664, + "grad_norm": 0.11444945633411407, + "learning_rate": 0.00018478350515463917, + "loss": 0.6759, + "step": 520 + }, + { + "epoch": 0.16672, + "grad_norm": 0.1250583827495575, + "learning_rate": 0.00018474226804123713, + "loss": 0.7223, + "step": 521 + }, + { + "epoch": 0.16704, + "grad_norm": 0.128667950630188, + "learning_rate": 0.00018470103092783506, + "loss": 0.7038, + "step": 522 + }, + { + "epoch": 0.16736, + "grad_norm": 0.1111631840467453, + "learning_rate": 0.000184659793814433, + "loss": 0.8037, + "step": 523 + }, + { + "epoch": 0.16768, + "grad_norm": 0.10735559463500977, + "learning_rate": 0.00018461855670103095, + "loss": 0.7687, + "step": 524 + }, + { + "epoch": 0.168, + "grad_norm": 0.12646177411079407, + "learning_rate": 0.00018457731958762888, + "loss": 0.7467, + "step": 525 + }, + { + "epoch": 0.16832, + "grad_norm": 0.11589492112398148, + "learning_rate": 0.0001845360824742268, + "loss": 0.5764, + "step": 526 + }, + { + "epoch": 0.16864, + "grad_norm": 0.11384910345077515, + "learning_rate": 0.00018449484536082474, + "loss": 0.7205, + "step": 527 + }, + { + "epoch": 0.16896, + "grad_norm": 0.11103667318820953, + "learning_rate": 0.00018445360824742267, + "loss": 0.7596, + "step": 528 + }, + { + "epoch": 0.16928, + "grad_norm": 0.11777044087648392, + "learning_rate": 0.00018441237113402063, + "loss": 0.8103, + "step": 529 + }, + { + "epoch": 0.1696, + "grad_norm": 0.13625359535217285, + "learning_rate": 0.00018437113402061856, + "loss": 0.8057, + "step": 530 + }, + { + "epoch": 0.16992, + "grad_norm": 0.12384586781263351, + "learning_rate": 0.0001843298969072165, + "loss": 0.6888, + "step": 531 + }, + { + "epoch": 0.17024, + "grad_norm": 0.12003818899393082, + "learning_rate": 0.00018428865979381445, + "loss": 0.6938, + "step": 532 + }, + { + "epoch": 0.17056, + "grad_norm": 0.11006835103034973, + "learning_rate": 0.00018424742268041239, + "loss": 0.8389, + "step": 533 + }, + { + "epoch": 0.17088, + "grad_norm": 0.1308056265115738, + "learning_rate": 0.00018420618556701032, + "loss": 0.6838, + "step": 534 + }, + { + "epoch": 0.1712, + "grad_norm": 0.13694193959236145, + "learning_rate": 0.00018416494845360825, + "loss": 0.7372, + "step": 535 + }, + { + "epoch": 0.17152, + "grad_norm": 0.12474847584962845, + "learning_rate": 0.00018412371134020618, + "loss": 0.7316, + "step": 536 + }, + { + "epoch": 0.17184, + "grad_norm": 0.12851357460021973, + "learning_rate": 0.00018408247422680414, + "loss": 0.8742, + "step": 537 + }, + { + "epoch": 0.17216, + "grad_norm": 0.12235444784164429, + "learning_rate": 0.00018404123711340207, + "loss": 0.798, + "step": 538 + }, + { + "epoch": 0.17248, + "grad_norm": 0.13520556688308716, + "learning_rate": 0.00018400000000000003, + "loss": 0.8435, + "step": 539 + }, + { + "epoch": 0.1728, + "grad_norm": 0.10702179372310638, + "learning_rate": 0.00018395876288659796, + "loss": 0.5605, + "step": 540 + }, + { + "epoch": 0.17312, + "grad_norm": 0.13347023725509644, + "learning_rate": 0.00018391752577319586, + "loss": 0.7539, + "step": 541 + }, + { + "epoch": 0.17344, + "grad_norm": 0.12964722514152527, + "learning_rate": 0.00018387628865979382, + "loss": 0.7302, + "step": 542 + }, + { + "epoch": 0.17376, + "grad_norm": 0.1171535775065422, + "learning_rate": 0.00018383505154639175, + "loss": 0.6735, + "step": 543 + }, + { + "epoch": 0.17408, + "grad_norm": 0.11912510544061661, + "learning_rate": 0.0001837938144329897, + "loss": 0.7924, + "step": 544 + }, + { + "epoch": 0.1744, + "grad_norm": 0.13499023020267487, + "learning_rate": 0.00018375257731958764, + "loss": 0.6884, + "step": 545 + }, + { + "epoch": 0.17472, + "grad_norm": 0.11606519669294357, + "learning_rate": 0.00018371134020618557, + "loss": 0.6533, + "step": 546 + }, + { + "epoch": 0.17504, + "grad_norm": 0.11444912850856781, + "learning_rate": 0.00018367010309278353, + "loss": 0.6173, + "step": 547 + }, + { + "epoch": 0.17536, + "grad_norm": 0.12102066725492477, + "learning_rate": 0.00018362886597938144, + "loss": 0.7732, + "step": 548 + }, + { + "epoch": 0.17568, + "grad_norm": 0.12623833119869232, + "learning_rate": 0.0001835876288659794, + "loss": 0.8299, + "step": 549 + }, + { + "epoch": 0.176, + "grad_norm": 0.13104446232318878, + "learning_rate": 0.00018354639175257733, + "loss": 0.6679, + "step": 550 + }, + { + "epoch": 0.17632, + "grad_norm": 0.12092466652393341, + "learning_rate": 0.00018350515463917526, + "loss": 0.7737, + "step": 551 + }, + { + "epoch": 0.17664, + "grad_norm": 0.11604610830545425, + "learning_rate": 0.00018346391752577322, + "loss": 0.8411, + "step": 552 + }, + { + "epoch": 0.17696, + "grad_norm": 0.11258059740066528, + "learning_rate": 0.00018342268041237115, + "loss": 0.7016, + "step": 553 + }, + { + "epoch": 0.17728, + "grad_norm": 0.11493982374668121, + "learning_rate": 0.00018338144329896908, + "loss": 0.7129, + "step": 554 + }, + { + "epoch": 0.1776, + "grad_norm": 0.13074208796024323, + "learning_rate": 0.000183340206185567, + "loss": 0.7522, + "step": 555 + }, + { + "epoch": 0.17792, + "grad_norm": 0.1120564267039299, + "learning_rate": 0.00018329896907216494, + "loss": 0.6846, + "step": 556 + }, + { + "epoch": 0.17824, + "grad_norm": 0.1323213428258896, + "learning_rate": 0.0001832577319587629, + "loss": 0.8367, + "step": 557 + }, + { + "epoch": 0.17856, + "grad_norm": 0.14067520201206207, + "learning_rate": 0.00018321649484536083, + "loss": 0.9111, + "step": 558 + }, + { + "epoch": 0.17888, + "grad_norm": 0.12386526167392731, + "learning_rate": 0.0001831752577319588, + "loss": 0.7513, + "step": 559 + }, + { + "epoch": 0.1792, + "grad_norm": 0.11267178505659103, + "learning_rate": 0.00018313402061855672, + "loss": 0.7561, + "step": 560 + }, + { + "epoch": 0.17952, + "grad_norm": 0.12407040596008301, + "learning_rate": 0.00018309278350515465, + "loss": 0.9464, + "step": 561 + }, + { + "epoch": 0.17984, + "grad_norm": 0.12141172587871552, + "learning_rate": 0.00018305154639175258, + "loss": 0.8564, + "step": 562 + }, + { + "epoch": 0.18016, + "grad_norm": 0.11525171250104904, + "learning_rate": 0.0001830103092783505, + "loss": 0.7085, + "step": 563 + }, + { + "epoch": 0.18048, + "grad_norm": 0.1141604334115982, + "learning_rate": 0.00018296907216494844, + "loss": 0.89, + "step": 564 + }, + { + "epoch": 0.1808, + "grad_norm": 0.117414191365242, + "learning_rate": 0.0001829278350515464, + "loss": 0.6754, + "step": 565 + }, + { + "epoch": 0.18112, + "grad_norm": 0.13193415105342865, + "learning_rate": 0.00018288659793814433, + "loss": 0.8465, + "step": 566 + }, + { + "epoch": 0.18144, + "grad_norm": 0.12107145041227341, + "learning_rate": 0.0001828453608247423, + "loss": 0.7464, + "step": 567 + }, + { + "epoch": 0.18176, + "grad_norm": 0.11851692944765091, + "learning_rate": 0.00018280412371134022, + "loss": 0.6608, + "step": 568 + }, + { + "epoch": 0.18208, + "grad_norm": 0.11890149861574173, + "learning_rate": 0.00018276288659793816, + "loss": 0.9179, + "step": 569 + }, + { + "epoch": 0.1824, + "grad_norm": 0.11981701850891113, + "learning_rate": 0.0001827216494845361, + "loss": 0.7348, + "step": 570 + }, + { + "epoch": 0.18272, + "grad_norm": 0.13188837468624115, + "learning_rate": 0.00018268041237113402, + "loss": 0.8143, + "step": 571 + }, + { + "epoch": 0.18304, + "grad_norm": 0.12455637753009796, + "learning_rate": 0.00018263917525773198, + "loss": 0.5797, + "step": 572 + }, + { + "epoch": 0.18336, + "grad_norm": 0.12442684173583984, + "learning_rate": 0.0001825979381443299, + "loss": 0.8271, + "step": 573 + }, + { + "epoch": 0.18368, + "grad_norm": 0.12401217222213745, + "learning_rate": 0.00018255670103092784, + "loss": 0.6963, + "step": 574 + }, + { + "epoch": 0.184, + "grad_norm": 0.12561100721359253, + "learning_rate": 0.0001825154639175258, + "loss": 0.7436, + "step": 575 + }, + { + "epoch": 0.18432, + "grad_norm": 0.1348034292459488, + "learning_rate": 0.00018247422680412373, + "loss": 0.677, + "step": 576 + }, + { + "epoch": 0.18464, + "grad_norm": 0.12489503622055054, + "learning_rate": 0.00018243298969072166, + "loss": 0.7475, + "step": 577 + }, + { + "epoch": 0.18496, + "grad_norm": 0.11799902468919754, + "learning_rate": 0.0001823917525773196, + "loss": 0.6301, + "step": 578 + }, + { + "epoch": 0.18528, + "grad_norm": 0.12443645298480988, + "learning_rate": 0.00018235051546391752, + "loss": 0.8342, + "step": 579 + }, + { + "epoch": 0.1856, + "grad_norm": 0.1418801099061966, + "learning_rate": 0.00018230927835051548, + "loss": 0.7459, + "step": 580 + }, + { + "epoch": 0.18592, + "grad_norm": 0.13329613208770752, + "learning_rate": 0.0001822680412371134, + "loss": 0.8183, + "step": 581 + }, + { + "epoch": 0.18624, + "grad_norm": 0.11498089879751205, + "learning_rate": 0.00018222680412371137, + "loss": 0.586, + "step": 582 + }, + { + "epoch": 0.18656, + "grad_norm": 0.11994834244251251, + "learning_rate": 0.0001821855670103093, + "loss": 0.6292, + "step": 583 + }, + { + "epoch": 0.18688, + "grad_norm": 0.11894108355045319, + "learning_rate": 0.0001821443298969072, + "loss": 0.79, + "step": 584 + }, + { + "epoch": 0.1872, + "grad_norm": 0.13187873363494873, + "learning_rate": 0.00018210309278350516, + "loss": 0.6799, + "step": 585 + }, + { + "epoch": 0.18752, + "grad_norm": 0.12553033232688904, + "learning_rate": 0.0001820618556701031, + "loss": 0.7286, + "step": 586 + }, + { + "epoch": 0.18784, + "grad_norm": 0.1164742261171341, + "learning_rate": 0.00018202061855670105, + "loss": 0.6721, + "step": 587 + }, + { + "epoch": 0.18816, + "grad_norm": 0.10284867882728577, + "learning_rate": 0.00018197938144329898, + "loss": 0.5752, + "step": 588 + }, + { + "epoch": 0.18848, + "grad_norm": 0.14761145412921906, + "learning_rate": 0.00018193814432989692, + "loss": 0.8899, + "step": 589 + }, + { + "epoch": 0.1888, + "grad_norm": 0.11473073065280914, + "learning_rate": 0.00018189690721649485, + "loss": 0.856, + "step": 590 + }, + { + "epoch": 0.18912, + "grad_norm": 0.12335352599620819, + "learning_rate": 0.00018185567010309278, + "loss": 0.7842, + "step": 591 + }, + { + "epoch": 0.18944, + "grad_norm": 0.10649198293685913, + "learning_rate": 0.0001818144329896907, + "loss": 0.7395, + "step": 592 + }, + { + "epoch": 0.18976, + "grad_norm": 0.11945925652980804, + "learning_rate": 0.00018177319587628867, + "loss": 0.6553, + "step": 593 + }, + { + "epoch": 0.19008, + "grad_norm": 0.1164558157324791, + "learning_rate": 0.0001817319587628866, + "loss": 0.6454, + "step": 594 + }, + { + "epoch": 0.1904, + "grad_norm": 0.12806792557239532, + "learning_rate": 0.00018169072164948456, + "loss": 0.7041, + "step": 595 + }, + { + "epoch": 0.19072, + "grad_norm": 0.12284845858812332, + "learning_rate": 0.0001816494845360825, + "loss": 0.8651, + "step": 596 + }, + { + "epoch": 0.19104, + "grad_norm": 0.12949353456497192, + "learning_rate": 0.00018160824742268042, + "loss": 0.761, + "step": 597 + }, + { + "epoch": 0.19136, + "grad_norm": 0.11860032379627228, + "learning_rate": 0.00018156701030927835, + "loss": 0.7484, + "step": 598 + }, + { + "epoch": 0.19168, + "grad_norm": 0.12866483628749847, + "learning_rate": 0.00018152577319587628, + "loss": 0.7962, + "step": 599 + }, + { + "epoch": 0.192, + "grad_norm": 0.1290092021226883, + "learning_rate": 0.00018148453608247424, + "loss": 0.6357, + "step": 600 + }, + { + "epoch": 0.19232, + "grad_norm": 0.12456108629703522, + "learning_rate": 0.00018144329896907217, + "loss": 0.8116, + "step": 601 + }, + { + "epoch": 0.19264, + "grad_norm": 0.11560390144586563, + "learning_rate": 0.0001814020618556701, + "loss": 0.6152, + "step": 602 + }, + { + "epoch": 0.19296, + "grad_norm": 0.1171826645731926, + "learning_rate": 0.00018136082474226806, + "loss": 0.7314, + "step": 603 + }, + { + "epoch": 0.19328, + "grad_norm": 0.11141692847013474, + "learning_rate": 0.000181319587628866, + "loss": 0.831, + "step": 604 + }, + { + "epoch": 0.1936, + "grad_norm": 0.12226671725511551, + "learning_rate": 0.00018127835051546393, + "loss": 0.6946, + "step": 605 + }, + { + "epoch": 0.19392, + "grad_norm": 0.12684857845306396, + "learning_rate": 0.00018123711340206186, + "loss": 0.7299, + "step": 606 + }, + { + "epoch": 0.19424, + "grad_norm": 0.12957026064395905, + "learning_rate": 0.0001811958762886598, + "loss": 0.7208, + "step": 607 + }, + { + "epoch": 0.19456, + "grad_norm": 0.11711089313030243, + "learning_rate": 0.00018115463917525775, + "loss": 0.6967, + "step": 608 + }, + { + "epoch": 0.19488, + "grad_norm": 0.12728916108608246, + "learning_rate": 0.00018111340206185568, + "loss": 0.6866, + "step": 609 + }, + { + "epoch": 0.1952, + "grad_norm": 0.11292556673288345, + "learning_rate": 0.00018107216494845364, + "loss": 0.6493, + "step": 610 + }, + { + "epoch": 0.19552, + "grad_norm": 0.11602400243282318, + "learning_rate": 0.00018103092783505157, + "loss": 0.7802, + "step": 611 + }, + { + "epoch": 0.19584, + "grad_norm": 0.12061775475740433, + "learning_rate": 0.0001809896907216495, + "loss": 0.7336, + "step": 612 + }, + { + "epoch": 0.19616, + "grad_norm": 0.13388070464134216, + "learning_rate": 0.00018094845360824743, + "loss": 0.6978, + "step": 613 + }, + { + "epoch": 0.19648, + "grad_norm": 0.11619807779788971, + "learning_rate": 0.00018090721649484536, + "loss": 0.7327, + "step": 614 + }, + { + "epoch": 0.1968, + "grad_norm": 0.12670999765396118, + "learning_rate": 0.00018086597938144332, + "loss": 0.7561, + "step": 615 + }, + { + "epoch": 0.19712, + "grad_norm": 0.11523107439279556, + "learning_rate": 0.00018082474226804125, + "loss": 0.7343, + "step": 616 + }, + { + "epoch": 0.19744, + "grad_norm": 0.13320648670196533, + "learning_rate": 0.00018078350515463918, + "loss": 0.8011, + "step": 617 + }, + { + "epoch": 0.19776, + "grad_norm": 0.1562347114086151, + "learning_rate": 0.00018074226804123714, + "loss": 0.8, + "step": 618 + }, + { + "epoch": 0.19808, + "grad_norm": 0.13111111521720886, + "learning_rate": 0.00018070103092783507, + "loss": 0.7242, + "step": 619 + }, + { + "epoch": 0.1984, + "grad_norm": 0.13651488721370697, + "learning_rate": 0.000180659793814433, + "loss": 0.7653, + "step": 620 + }, + { + "epoch": 0.19872, + "grad_norm": 0.12414918839931488, + "learning_rate": 0.00018061855670103093, + "loss": 0.803, + "step": 621 + }, + { + "epoch": 0.19904, + "grad_norm": 0.11437831819057465, + "learning_rate": 0.00018057731958762887, + "loss": 0.6214, + "step": 622 + }, + { + "epoch": 0.19936, + "grad_norm": 0.10947408527135849, + "learning_rate": 0.00018053608247422682, + "loss": 0.6069, + "step": 623 + }, + { + "epoch": 0.19968, + "grad_norm": 0.11896886676549911, + "learning_rate": 0.00018049484536082475, + "loss": 0.8071, + "step": 624 + }, + { + "epoch": 0.2, + "grad_norm": 0.10557591170072556, + "learning_rate": 0.00018045360824742269, + "loss": 0.6428, + "step": 625 + }, + { + "epoch": 0.20032, + "grad_norm": 0.11610836535692215, + "learning_rate": 0.00018041237113402062, + "loss": 0.6034, + "step": 626 + }, + { + "epoch": 0.20064, + "grad_norm": 0.11742173135280609, + "learning_rate": 0.00018037113402061855, + "loss": 0.7368, + "step": 627 + }, + { + "epoch": 0.20096, + "grad_norm": 0.11513591557741165, + "learning_rate": 0.0001803298969072165, + "loss": 0.5753, + "step": 628 + }, + { + "epoch": 0.20128, + "grad_norm": 0.12790197134017944, + "learning_rate": 0.00018028865979381444, + "loss": 0.6109, + "step": 629 + }, + { + "epoch": 0.2016, + "grad_norm": 0.12305860221385956, + "learning_rate": 0.00018024742268041237, + "loss": 0.6894, + "step": 630 + }, + { + "epoch": 0.20192, + "grad_norm": 0.1050051599740982, + "learning_rate": 0.00018020618556701033, + "loss": 0.7507, + "step": 631 + }, + { + "epoch": 0.20224, + "grad_norm": 0.12011159211397171, + "learning_rate": 0.00018016494845360826, + "loss": 0.6234, + "step": 632 + }, + { + "epoch": 0.20256, + "grad_norm": 0.13481462001800537, + "learning_rate": 0.0001801237113402062, + "loss": 0.7408, + "step": 633 + }, + { + "epoch": 0.20288, + "grad_norm": 0.11771773546934128, + "learning_rate": 0.00018008247422680412, + "loss": 0.7283, + "step": 634 + }, + { + "epoch": 0.2032, + "grad_norm": 0.12805534899234772, + "learning_rate": 0.00018004123711340205, + "loss": 0.6377, + "step": 635 + }, + { + "epoch": 0.20352, + "grad_norm": 0.1451328545808792, + "learning_rate": 0.00018, + "loss": 0.6929, + "step": 636 + }, + { + "epoch": 0.20384, + "grad_norm": 0.116309255361557, + "learning_rate": 0.00017995876288659794, + "loss": 0.8133, + "step": 637 + }, + { + "epoch": 0.20416, + "grad_norm": 0.11883087456226349, + "learning_rate": 0.0001799175257731959, + "loss": 0.7036, + "step": 638 + }, + { + "epoch": 0.20448, + "grad_norm": 0.11787105351686478, + "learning_rate": 0.00017987628865979383, + "loss": 0.7313, + "step": 639 + }, + { + "epoch": 0.2048, + "grad_norm": 0.10839900374412537, + "learning_rate": 0.00017983505154639176, + "loss": 0.5725, + "step": 640 + }, + { + "epoch": 0.20512, + "grad_norm": 0.12540479004383087, + "learning_rate": 0.0001797938144329897, + "loss": 0.6349, + "step": 641 + }, + { + "epoch": 0.20544, + "grad_norm": 0.1231970563530922, + "learning_rate": 0.00017975257731958763, + "loss": 0.7068, + "step": 642 + }, + { + "epoch": 0.20576, + "grad_norm": 0.1263824850320816, + "learning_rate": 0.00017971134020618558, + "loss": 0.861, + "step": 643 + }, + { + "epoch": 0.20608, + "grad_norm": 0.134733647108078, + "learning_rate": 0.00017967010309278352, + "loss": 0.67, + "step": 644 + }, + { + "epoch": 0.2064, + "grad_norm": 0.14626337587833405, + "learning_rate": 0.00017962886597938145, + "loss": 0.6905, + "step": 645 + }, + { + "epoch": 0.20672, + "grad_norm": 0.11077899485826492, + "learning_rate": 0.0001795876288659794, + "loss": 0.6866, + "step": 646 + }, + { + "epoch": 0.20704, + "grad_norm": 0.12598447501659393, + "learning_rate": 0.00017954639175257734, + "loss": 0.7583, + "step": 647 + }, + { + "epoch": 0.20736, + "grad_norm": 0.12656626105308533, + "learning_rate": 0.00017950515463917527, + "loss": 0.679, + "step": 648 + }, + { + "epoch": 0.20768, + "grad_norm": 0.1403527706861496, + "learning_rate": 0.0001794639175257732, + "loss": 0.8081, + "step": 649 + }, + { + "epoch": 0.208, + "grad_norm": 0.11713801324367523, + "learning_rate": 0.00017942268041237113, + "loss": 0.7322, + "step": 650 + }, + { + "epoch": 0.20832, + "grad_norm": 0.1281721144914627, + "learning_rate": 0.0001793814432989691, + "loss": 0.8457, + "step": 651 + }, + { + "epoch": 0.20864, + "grad_norm": 0.11331729590892792, + "learning_rate": 0.00017934020618556702, + "loss": 0.6479, + "step": 652 + }, + { + "epoch": 0.20896, + "grad_norm": 0.11600172519683838, + "learning_rate": 0.00017929896907216495, + "loss": 0.7144, + "step": 653 + }, + { + "epoch": 0.20928, + "grad_norm": 0.13482129573822021, + "learning_rate": 0.0001792577319587629, + "loss": 0.7639, + "step": 654 + }, + { + "epoch": 0.2096, + "grad_norm": 0.12466888129711151, + "learning_rate": 0.00017921649484536081, + "loss": 0.6413, + "step": 655 + }, + { + "epoch": 0.20992, + "grad_norm": 0.11935215443372726, + "learning_rate": 0.00017917525773195877, + "loss": 0.7551, + "step": 656 + }, + { + "epoch": 0.21024, + "grad_norm": 0.12447444349527359, + "learning_rate": 0.0001791340206185567, + "loss": 0.7171, + "step": 657 + }, + { + "epoch": 0.21056, + "grad_norm": 0.1334541141986847, + "learning_rate": 0.00017909278350515463, + "loss": 0.9159, + "step": 658 + }, + { + "epoch": 0.21088, + "grad_norm": 0.111027792096138, + "learning_rate": 0.0001790515463917526, + "loss": 0.6894, + "step": 659 + }, + { + "epoch": 0.2112, + "grad_norm": 0.12192652374505997, + "learning_rate": 0.00017901030927835052, + "loss": 0.7354, + "step": 660 + }, + { + "epoch": 0.21152, + "grad_norm": 0.12837322056293488, + "learning_rate": 0.00017896907216494848, + "loss": 0.7535, + "step": 661 + }, + { + "epoch": 0.21184, + "grad_norm": 0.1143856793642044, + "learning_rate": 0.0001789278350515464, + "loss": 0.6189, + "step": 662 + }, + { + "epoch": 0.21216, + "grad_norm": 0.12996436655521393, + "learning_rate": 0.00017888659793814432, + "loss": 0.7194, + "step": 663 + }, + { + "epoch": 0.21248, + "grad_norm": 0.13088072836399078, + "learning_rate": 0.00017884536082474228, + "loss": 0.6356, + "step": 664 + }, + { + "epoch": 0.2128, + "grad_norm": 0.11904153227806091, + "learning_rate": 0.0001788041237113402, + "loss": 0.5998, + "step": 665 + }, + { + "epoch": 0.21312, + "grad_norm": 0.11976154148578644, + "learning_rate": 0.00017876288659793817, + "loss": 0.7417, + "step": 666 + }, + { + "epoch": 0.21344, + "grad_norm": 0.1175193041563034, + "learning_rate": 0.0001787216494845361, + "loss": 0.741, + "step": 667 + }, + { + "epoch": 0.21376, + "grad_norm": 0.1151885837316513, + "learning_rate": 0.00017868041237113403, + "loss": 0.7779, + "step": 668 + }, + { + "epoch": 0.21408, + "grad_norm": 0.1385083645582199, + "learning_rate": 0.00017863917525773196, + "loss": 0.7296, + "step": 669 + }, + { + "epoch": 0.2144, + "grad_norm": 0.1304243803024292, + "learning_rate": 0.0001785979381443299, + "loss": 0.7085, + "step": 670 + }, + { + "epoch": 0.21472, + "grad_norm": 0.11901744455099106, + "learning_rate": 0.00017855670103092785, + "loss": 0.7005, + "step": 671 + }, + { + "epoch": 0.21504, + "grad_norm": 0.11635670065879822, + "learning_rate": 0.00017851546391752578, + "loss": 0.7824, + "step": 672 + }, + { + "epoch": 0.21536, + "grad_norm": 0.14305877685546875, + "learning_rate": 0.0001784742268041237, + "loss": 0.6263, + "step": 673 + }, + { + "epoch": 0.21568, + "grad_norm": 0.11984144151210785, + "learning_rate": 0.00017843298969072167, + "loss": 0.6518, + "step": 674 + }, + { + "epoch": 0.216, + "grad_norm": 0.10570491850376129, + "learning_rate": 0.0001783917525773196, + "loss": 0.6311, + "step": 675 + }, + { + "epoch": 0.21632, + "grad_norm": 0.13266734778881073, + "learning_rate": 0.00017835051546391753, + "loss": 0.7975, + "step": 676 + }, + { + "epoch": 0.21664, + "grad_norm": 0.12022270262241364, + "learning_rate": 0.00017830927835051546, + "loss": 0.6691, + "step": 677 + }, + { + "epoch": 0.21696, + "grad_norm": 0.1266927272081375, + "learning_rate": 0.0001782680412371134, + "loss": 0.7211, + "step": 678 + }, + { + "epoch": 0.21728, + "grad_norm": 0.13459637761116028, + "learning_rate": 0.00017822680412371135, + "loss": 0.7911, + "step": 679 + }, + { + "epoch": 0.2176, + "grad_norm": 0.14075347781181335, + "learning_rate": 0.00017818556701030929, + "loss": 0.8012, + "step": 680 + }, + { + "epoch": 0.21792, + "grad_norm": 0.12798281013965607, + "learning_rate": 0.00017814432989690724, + "loss": 0.6135, + "step": 681 + }, + { + "epoch": 0.21824, + "grad_norm": 0.14813600480556488, + "learning_rate": 0.00017810309278350518, + "loss": 0.7988, + "step": 682 + }, + { + "epoch": 0.21856, + "grad_norm": 0.14158020913600922, + "learning_rate": 0.0001780618556701031, + "loss": 0.7725, + "step": 683 + }, + { + "epoch": 0.21888, + "grad_norm": 0.10573446750640869, + "learning_rate": 0.00017802061855670104, + "loss": 0.7956, + "step": 684 + }, + { + "epoch": 0.2192, + "grad_norm": 0.14443108439445496, + "learning_rate": 0.00017797938144329897, + "loss": 0.8615, + "step": 685 + }, + { + "epoch": 0.21952, + "grad_norm": 0.1228407770395279, + "learning_rate": 0.0001779381443298969, + "loss": 0.706, + "step": 686 + }, + { + "epoch": 0.21984, + "grad_norm": 0.1218361034989357, + "learning_rate": 0.00017789690721649486, + "loss": 0.6525, + "step": 687 + }, + { + "epoch": 0.22016, + "grad_norm": 0.12609677016735077, + "learning_rate": 0.0001778556701030928, + "loss": 0.5638, + "step": 688 + }, + { + "epoch": 0.22048, + "grad_norm": 0.1489102840423584, + "learning_rate": 0.00017781443298969075, + "loss": 0.641, + "step": 689 + }, + { + "epoch": 0.2208, + "grad_norm": 0.11327382177114487, + "learning_rate": 0.00017777319587628868, + "loss": 0.7372, + "step": 690 + }, + { + "epoch": 0.22112, + "grad_norm": 0.12614741921424866, + "learning_rate": 0.00017773195876288658, + "loss": 0.7819, + "step": 691 + }, + { + "epoch": 0.22144, + "grad_norm": 0.11078184098005295, + "learning_rate": 0.00017769072164948454, + "loss": 0.8376, + "step": 692 + }, + { + "epoch": 0.22176, + "grad_norm": 0.13330969214439392, + "learning_rate": 0.00017764948453608247, + "loss": 0.771, + "step": 693 + }, + { + "epoch": 0.22208, + "grad_norm": 0.14074283838272095, + "learning_rate": 0.00017760824742268043, + "loss": 0.7675, + "step": 694 + }, + { + "epoch": 0.2224, + "grad_norm": 0.12653911113739014, + "learning_rate": 0.00017756701030927836, + "loss": 0.7578, + "step": 695 + }, + { + "epoch": 0.22272, + "grad_norm": 0.13229238986968994, + "learning_rate": 0.0001775257731958763, + "loss": 0.5843, + "step": 696 + }, + { + "epoch": 0.22304, + "grad_norm": 0.14447958767414093, + "learning_rate": 0.00017748453608247425, + "loss": 0.8251, + "step": 697 + }, + { + "epoch": 0.22336, + "grad_norm": 0.12437491863965988, + "learning_rate": 0.00017744329896907216, + "loss": 0.7624, + "step": 698 + }, + { + "epoch": 0.22368, + "grad_norm": 0.13665013015270233, + "learning_rate": 0.00017740206185567012, + "loss": 0.6546, + "step": 699 + }, + { + "epoch": 0.224, + "grad_norm": 0.12415925413370132, + "learning_rate": 0.00017736082474226805, + "loss": 0.5989, + "step": 700 + }, + { + "epoch": 0.22432, + "grad_norm": 0.13429298996925354, + "learning_rate": 0.00017731958762886598, + "loss": 0.7283, + "step": 701 + }, + { + "epoch": 0.22464, + "grad_norm": 0.11220762878656387, + "learning_rate": 0.00017727835051546394, + "loss": 0.844, + "step": 702 + }, + { + "epoch": 0.22496, + "grad_norm": 0.15322859585285187, + "learning_rate": 0.00017723711340206187, + "loss": 0.744, + "step": 703 + }, + { + "epoch": 0.22528, + "grad_norm": 0.10617350786924362, + "learning_rate": 0.0001771958762886598, + "loss": 0.7462, + "step": 704 + }, + { + "epoch": 0.2256, + "grad_norm": 0.12569446861743927, + "learning_rate": 0.00017715463917525773, + "loss": 0.7465, + "step": 705 + }, + { + "epoch": 0.22592, + "grad_norm": 0.13956156373023987, + "learning_rate": 0.00017711340206185566, + "loss": 0.7465, + "step": 706 + }, + { + "epoch": 0.22624, + "grad_norm": 0.13465113937854767, + "learning_rate": 0.00017707216494845362, + "loss": 0.8502, + "step": 707 + }, + { + "epoch": 0.22656, + "grad_norm": 0.11806920915842056, + "learning_rate": 0.00017703092783505155, + "loss": 0.6246, + "step": 708 + }, + { + "epoch": 0.22688, + "grad_norm": 0.1319315880537033, + "learning_rate": 0.0001769896907216495, + "loss": 0.6639, + "step": 709 + }, + { + "epoch": 0.2272, + "grad_norm": 0.14387674629688263, + "learning_rate": 0.00017694845360824744, + "loss": 0.8698, + "step": 710 + }, + { + "epoch": 0.22752, + "grad_norm": 0.12522293627262115, + "learning_rate": 0.00017690721649484537, + "loss": 0.7884, + "step": 711 + }, + { + "epoch": 0.22784, + "grad_norm": 0.11482946574687958, + "learning_rate": 0.0001768659793814433, + "loss": 0.7562, + "step": 712 + }, + { + "epoch": 0.22816, + "grad_norm": 0.11533328890800476, + "learning_rate": 0.00017682474226804123, + "loss": 0.6228, + "step": 713 + }, + { + "epoch": 0.22848, + "grad_norm": 0.12137199193239212, + "learning_rate": 0.00017678350515463917, + "loss": 0.8835, + "step": 714 + }, + { + "epoch": 0.2288, + "grad_norm": 0.11885856837034225, + "learning_rate": 0.00017674226804123712, + "loss": 0.6241, + "step": 715 + }, + { + "epoch": 0.22912, + "grad_norm": 0.1262994110584259, + "learning_rate": 0.00017670103092783506, + "loss": 0.715, + "step": 716 + }, + { + "epoch": 0.22944, + "grad_norm": 0.1313248574733734, + "learning_rate": 0.00017665979381443301, + "loss": 0.8368, + "step": 717 + }, + { + "epoch": 0.22976, + "grad_norm": 0.15882809460163116, + "learning_rate": 0.00017661855670103095, + "loss": 0.8688, + "step": 718 + }, + { + "epoch": 0.23008, + "grad_norm": 0.13610002398490906, + "learning_rate": 0.00017657731958762888, + "loss": 0.7367, + "step": 719 + }, + { + "epoch": 0.2304, + "grad_norm": 0.11154987663030624, + "learning_rate": 0.0001765360824742268, + "loss": 0.8026, + "step": 720 + }, + { + "epoch": 0.23072, + "grad_norm": 0.14375466108322144, + "learning_rate": 0.00017649484536082474, + "loss": 0.8038, + "step": 721 + }, + { + "epoch": 0.23104, + "grad_norm": 0.13212132453918457, + "learning_rate": 0.0001764536082474227, + "loss": 0.9008, + "step": 722 + }, + { + "epoch": 0.23136, + "grad_norm": 0.1395123153924942, + "learning_rate": 0.00017641237113402063, + "loss": 0.7491, + "step": 723 + }, + { + "epoch": 0.23168, + "grad_norm": 0.11533952504396439, + "learning_rate": 0.00017637113402061856, + "loss": 0.6714, + "step": 724 + }, + { + "epoch": 0.232, + "grad_norm": 0.10333327949047089, + "learning_rate": 0.00017632989690721652, + "loss": 0.6643, + "step": 725 + }, + { + "epoch": 0.23232, + "grad_norm": 0.12025081366300583, + "learning_rate": 0.00017628865979381445, + "loss": 0.6639, + "step": 726 + }, + { + "epoch": 0.23264, + "grad_norm": 0.13901305198669434, + "learning_rate": 0.00017624742268041238, + "loss": 0.7138, + "step": 727 + }, + { + "epoch": 0.23296, + "grad_norm": 0.14389030635356903, + "learning_rate": 0.0001762061855670103, + "loss": 0.7785, + "step": 728 + }, + { + "epoch": 0.23328, + "grad_norm": 0.11535743623971939, + "learning_rate": 0.00017616494845360824, + "loss": 0.715, + "step": 729 + }, + { + "epoch": 0.2336, + "grad_norm": 0.11448106914758682, + "learning_rate": 0.0001761237113402062, + "loss": 0.7739, + "step": 730 + }, + { + "epoch": 0.23392, + "grad_norm": 0.1345527321100235, + "learning_rate": 0.00017608247422680413, + "loss": 0.6493, + "step": 731 + }, + { + "epoch": 0.23424, + "grad_norm": 0.12693467736244202, + "learning_rate": 0.0001760412371134021, + "loss": 0.7164, + "step": 732 + }, + { + "epoch": 0.23456, + "grad_norm": 0.12556055188179016, + "learning_rate": 0.00017600000000000002, + "loss": 0.7162, + "step": 733 + }, + { + "epoch": 0.23488, + "grad_norm": 0.13427414000034332, + "learning_rate": 0.00017595876288659793, + "loss": 0.7169, + "step": 734 + }, + { + "epoch": 0.2352, + "grad_norm": 0.1384212076663971, + "learning_rate": 0.00017591752577319589, + "loss": 0.8806, + "step": 735 + }, + { + "epoch": 0.23552, + "grad_norm": 0.1685527116060257, + "learning_rate": 0.00017587628865979382, + "loss": 0.8755, + "step": 736 + }, + { + "epoch": 0.23584, + "grad_norm": 0.12363595515489578, + "learning_rate": 0.00017583505154639177, + "loss": 0.7556, + "step": 737 + }, + { + "epoch": 0.23616, + "grad_norm": 0.121871218085289, + "learning_rate": 0.0001757938144329897, + "loss": 0.7021, + "step": 738 + }, + { + "epoch": 0.23648, + "grad_norm": 0.133682981133461, + "learning_rate": 0.00017575257731958764, + "loss": 0.8063, + "step": 739 + }, + { + "epoch": 0.2368, + "grad_norm": 0.10897425562143326, + "learning_rate": 0.00017571134020618557, + "loss": 0.7153, + "step": 740 + }, + { + "epoch": 0.23712, + "grad_norm": 0.1330791562795639, + "learning_rate": 0.0001756701030927835, + "loss": 0.7005, + "step": 741 + }, + { + "epoch": 0.23744, + "grad_norm": 0.13002124428749084, + "learning_rate": 0.00017562886597938146, + "loss": 0.7188, + "step": 742 + }, + { + "epoch": 0.23776, + "grad_norm": 0.11166704446077347, + "learning_rate": 0.0001755876288659794, + "loss": 0.6922, + "step": 743 + }, + { + "epoch": 0.23808, + "grad_norm": 0.1195744052529335, + "learning_rate": 0.00017554639175257732, + "loss": 0.6846, + "step": 744 + }, + { + "epoch": 0.2384, + "grad_norm": 0.12304572016000748, + "learning_rate": 0.00017550515463917528, + "loss": 0.7575, + "step": 745 + }, + { + "epoch": 0.23872, + "grad_norm": 0.15100136399269104, + "learning_rate": 0.0001754639175257732, + "loss": 0.9129, + "step": 746 + }, + { + "epoch": 0.23904, + "grad_norm": 0.11369301378726959, + "learning_rate": 0.00017542268041237114, + "loss": 0.8557, + "step": 747 + }, + { + "epoch": 0.23936, + "grad_norm": 0.14192970097064972, + "learning_rate": 0.00017538144329896907, + "loss": 0.7254, + "step": 748 + }, + { + "epoch": 0.23968, + "grad_norm": 0.11873924732208252, + "learning_rate": 0.000175340206185567, + "loss": 0.7028, + "step": 749 + }, + { + "epoch": 0.24, + "grad_norm": 0.1388227790594101, + "learning_rate": 0.00017529896907216496, + "loss": 0.8156, + "step": 750 + }, + { + "epoch": 0.24032, + "grad_norm": 0.1263178288936615, + "learning_rate": 0.0001752577319587629, + "loss": 0.7294, + "step": 751 + }, + { + "epoch": 0.24064, + "grad_norm": 0.1199144572019577, + "learning_rate": 0.00017521649484536083, + "loss": 0.6154, + "step": 752 + }, + { + "epoch": 0.24096, + "grad_norm": 0.1196129247546196, + "learning_rate": 0.00017517525773195878, + "loss": 0.6223, + "step": 753 + }, + { + "epoch": 0.24128, + "grad_norm": 0.1296255737543106, + "learning_rate": 0.00017513402061855671, + "loss": 0.7155, + "step": 754 + }, + { + "epoch": 0.2416, + "grad_norm": 0.13744045794010162, + "learning_rate": 0.00017509278350515465, + "loss": 0.7899, + "step": 755 + }, + { + "epoch": 0.24192, + "grad_norm": 0.136393204331398, + "learning_rate": 0.00017505154639175258, + "loss": 0.7931, + "step": 756 + }, + { + "epoch": 0.24224, + "grad_norm": 0.12587614357471466, + "learning_rate": 0.0001750103092783505, + "loss": 0.7043, + "step": 757 + }, + { + "epoch": 0.24256, + "grad_norm": 0.12651780247688293, + "learning_rate": 0.00017496907216494847, + "loss": 0.711, + "step": 758 + }, + { + "epoch": 0.24288, + "grad_norm": 0.12235745042562485, + "learning_rate": 0.0001749278350515464, + "loss": 0.8063, + "step": 759 + }, + { + "epoch": 0.2432, + "grad_norm": 0.14242300391197205, + "learning_rate": 0.00017488659793814436, + "loss": 0.6157, + "step": 760 + }, + { + "epoch": 0.24352, + "grad_norm": 0.12730719149112701, + "learning_rate": 0.0001748453608247423, + "loss": 0.8741, + "step": 761 + }, + { + "epoch": 0.24384, + "grad_norm": 0.12228512763977051, + "learning_rate": 0.00017480412371134022, + "loss": 0.6424, + "step": 762 + }, + { + "epoch": 0.24416, + "grad_norm": 0.12545928359031677, + "learning_rate": 0.00017476288659793815, + "loss": 0.7231, + "step": 763 + }, + { + "epoch": 0.24448, + "grad_norm": 0.14945553243160248, + "learning_rate": 0.00017472164948453608, + "loss": 0.7147, + "step": 764 + }, + { + "epoch": 0.2448, + "grad_norm": 0.12359154969453812, + "learning_rate": 0.00017468041237113404, + "loss": 0.689, + "step": 765 + }, + { + "epoch": 0.24512, + "grad_norm": 0.1245129331946373, + "learning_rate": 0.00017463917525773197, + "loss": 0.7739, + "step": 766 + }, + { + "epoch": 0.24544, + "grad_norm": 0.12101677060127258, + "learning_rate": 0.0001745979381443299, + "loss": 0.6389, + "step": 767 + }, + { + "epoch": 0.24576, + "grad_norm": 0.12827037274837494, + "learning_rate": 0.00017455670103092786, + "loss": 0.7213, + "step": 768 + }, + { + "epoch": 0.24608, + "grad_norm": 0.11057371646165848, + "learning_rate": 0.0001745154639175258, + "loss": 0.5549, + "step": 769 + }, + { + "epoch": 0.2464, + "grad_norm": 0.10602620244026184, + "learning_rate": 0.00017447422680412372, + "loss": 0.663, + "step": 770 + }, + { + "epoch": 0.24672, + "grad_norm": 0.1476941555738449, + "learning_rate": 0.00017443298969072165, + "loss": 0.6113, + "step": 771 + }, + { + "epoch": 0.24704, + "grad_norm": 0.1098361536860466, + "learning_rate": 0.00017439175257731959, + "loss": 0.6189, + "step": 772 + }, + { + "epoch": 0.24736, + "grad_norm": 0.12340857088565826, + "learning_rate": 0.00017435051546391754, + "loss": 0.6615, + "step": 773 + }, + { + "epoch": 0.24768, + "grad_norm": 0.1320737898349762, + "learning_rate": 0.00017430927835051548, + "loss": 0.6293, + "step": 774 + }, + { + "epoch": 0.248, + "grad_norm": 0.12213713675737381, + "learning_rate": 0.0001742680412371134, + "loss": 0.6703, + "step": 775 + }, + { + "epoch": 0.24832, + "grad_norm": 0.13890238106250763, + "learning_rate": 0.00017422680412371134, + "loss": 0.7381, + "step": 776 + }, + { + "epoch": 0.24864, + "grad_norm": 0.13525603711605072, + "learning_rate": 0.00017418556701030927, + "loss": 0.8244, + "step": 777 + }, + { + "epoch": 0.24896, + "grad_norm": 0.1357133686542511, + "learning_rate": 0.00017414432989690723, + "loss": 0.6428, + "step": 778 + }, + { + "epoch": 0.24928, + "grad_norm": 0.13172458112239838, + "learning_rate": 0.00017410309278350516, + "loss": 0.7103, + "step": 779 + }, + { + "epoch": 0.2496, + "grad_norm": 0.14605680108070374, + "learning_rate": 0.0001740618556701031, + "loss": 0.7102, + "step": 780 + }, + { + "epoch": 0.24992, + "grad_norm": 0.14995813369750977, + "learning_rate": 0.00017402061855670105, + "loss": 0.9, + "step": 781 + }, + { + "epoch": 0.25024, + "grad_norm": 0.13214722275733948, + "learning_rate": 0.00017397938144329898, + "loss": 0.8016, + "step": 782 + }, + { + "epoch": 0.25056, + "grad_norm": 0.12392943352460861, + "learning_rate": 0.0001739381443298969, + "loss": 0.5207, + "step": 783 + }, + { + "epoch": 0.25088, + "grad_norm": 0.13737192749977112, + "learning_rate": 0.00017389690721649484, + "loss": 0.7711, + "step": 784 + }, + { + "epoch": 0.2512, + "grad_norm": 0.1214001253247261, + "learning_rate": 0.00017385567010309277, + "loss": 0.6856, + "step": 785 + }, + { + "epoch": 0.25152, + "grad_norm": 0.11796026676893234, + "learning_rate": 0.00017381443298969073, + "loss": 0.6661, + "step": 786 + }, + { + "epoch": 0.25184, + "grad_norm": 0.11840114742517471, + "learning_rate": 0.00017377319587628866, + "loss": 0.6285, + "step": 787 + }, + { + "epoch": 0.25216, + "grad_norm": 0.12398465722799301, + "learning_rate": 0.00017373195876288662, + "loss": 0.6363, + "step": 788 + }, + { + "epoch": 0.25248, + "grad_norm": 0.12396920472383499, + "learning_rate": 0.00017369072164948455, + "loss": 0.8027, + "step": 789 + }, + { + "epoch": 0.2528, + "grad_norm": 0.10319101810455322, + "learning_rate": 0.00017364948453608248, + "loss": 0.5851, + "step": 790 + }, + { + "epoch": 0.25312, + "grad_norm": 0.14459432661533356, + "learning_rate": 0.00017360824742268042, + "loss": 0.9323, + "step": 791 + }, + { + "epoch": 0.25344, + "grad_norm": 0.1221664696931839, + "learning_rate": 0.00017356701030927835, + "loss": 0.6041, + "step": 792 + }, + { + "epoch": 0.25376, + "grad_norm": 0.118996761739254, + "learning_rate": 0.0001735257731958763, + "loss": 0.8632, + "step": 793 + }, + { + "epoch": 0.25408, + "grad_norm": 0.13408184051513672, + "learning_rate": 0.00017348453608247424, + "loss": 0.7143, + "step": 794 + }, + { + "epoch": 0.2544, + "grad_norm": 0.13923995196819305, + "learning_rate": 0.00017344329896907217, + "loss": 0.7232, + "step": 795 + }, + { + "epoch": 0.25472, + "grad_norm": 0.12153901159763336, + "learning_rate": 0.00017340206185567013, + "loss": 0.6634, + "step": 796 + }, + { + "epoch": 0.25504, + "grad_norm": 0.14275431632995605, + "learning_rate": 0.00017336082474226806, + "loss": 0.6741, + "step": 797 + }, + { + "epoch": 0.25536, + "grad_norm": 0.12189670652151108, + "learning_rate": 0.000173319587628866, + "loss": 0.776, + "step": 798 + }, + { + "epoch": 0.25568, + "grad_norm": 0.12858492136001587, + "learning_rate": 0.00017327835051546392, + "loss": 0.7927, + "step": 799 + }, + { + "epoch": 0.256, + "grad_norm": 0.13049420714378357, + "learning_rate": 0.00017323711340206185, + "loss": 0.717, + "step": 800 + }, + { + "epoch": 0.25632, + "grad_norm": 0.13621200621128082, + "learning_rate": 0.0001731958762886598, + "loss": 0.6823, + "step": 801 + }, + { + "epoch": 0.25664, + "grad_norm": 0.13778063654899597, + "learning_rate": 0.00017315463917525774, + "loss": 0.7343, + "step": 802 + }, + { + "epoch": 0.25696, + "grad_norm": 0.11168331652879715, + "learning_rate": 0.00017311340206185567, + "loss": 0.6899, + "step": 803 + }, + { + "epoch": 0.25728, + "grad_norm": 0.11649805307388306, + "learning_rate": 0.00017307216494845363, + "loss": 0.7278, + "step": 804 + }, + { + "epoch": 0.2576, + "grad_norm": 0.13680049777030945, + "learning_rate": 0.00017303092783505154, + "loss": 0.59, + "step": 805 + }, + { + "epoch": 0.25792, + "grad_norm": 0.1319751888513565, + "learning_rate": 0.0001729896907216495, + "loss": 0.7789, + "step": 806 + }, + { + "epoch": 0.25824, + "grad_norm": 0.15065547823905945, + "learning_rate": 0.00017294845360824742, + "loss": 0.7411, + "step": 807 + }, + { + "epoch": 0.25856, + "grad_norm": 0.13833020627498627, + "learning_rate": 0.00017290721649484536, + "loss": 0.7369, + "step": 808 + }, + { + "epoch": 0.25888, + "grad_norm": 0.12196467071771622, + "learning_rate": 0.00017286597938144331, + "loss": 0.9695, + "step": 809 + }, + { + "epoch": 0.2592, + "grad_norm": 0.117258720099926, + "learning_rate": 0.00017282474226804125, + "loss": 0.6831, + "step": 810 + }, + { + "epoch": 0.25952, + "grad_norm": 0.13261094689369202, + "learning_rate": 0.0001727835051546392, + "loss": 0.6081, + "step": 811 + }, + { + "epoch": 0.25984, + "grad_norm": 0.12902230024337769, + "learning_rate": 0.0001727422680412371, + "loss": 0.6816, + "step": 812 + }, + { + "epoch": 0.26016, + "grad_norm": 0.12592282891273499, + "learning_rate": 0.00017270103092783504, + "loss": 0.7926, + "step": 813 + }, + { + "epoch": 0.26048, + "grad_norm": 0.11538536846637726, + "learning_rate": 0.000172659793814433, + "loss": 0.6386, + "step": 814 + }, + { + "epoch": 0.2608, + "grad_norm": 0.14629842340946198, + "learning_rate": 0.00017261855670103093, + "loss": 0.6989, + "step": 815 + }, + { + "epoch": 0.26112, + "grad_norm": 0.12829239666461945, + "learning_rate": 0.0001725773195876289, + "loss": 0.5876, + "step": 816 + }, + { + "epoch": 0.26144, + "grad_norm": 0.13584017753601074, + "learning_rate": 0.00017253608247422682, + "loss": 0.8126, + "step": 817 + }, + { + "epoch": 0.26176, + "grad_norm": 0.13379888236522675, + "learning_rate": 0.00017249484536082475, + "loss": 0.661, + "step": 818 + }, + { + "epoch": 0.26208, + "grad_norm": 0.14031822979450226, + "learning_rate": 0.00017245360824742268, + "loss": 0.8392, + "step": 819 + }, + { + "epoch": 0.2624, + "grad_norm": 0.12089665979146957, + "learning_rate": 0.0001724123711340206, + "loss": 0.7065, + "step": 820 + }, + { + "epoch": 0.26272, + "grad_norm": 0.11851520836353302, + "learning_rate": 0.00017237113402061857, + "loss": 0.5666, + "step": 821 + }, + { + "epoch": 0.26304, + "grad_norm": 0.11357667297124863, + "learning_rate": 0.0001723298969072165, + "loss": 0.9064, + "step": 822 + }, + { + "epoch": 0.26336, + "grad_norm": 0.12621043622493744, + "learning_rate": 0.00017228865979381443, + "loss": 0.6377, + "step": 823 + }, + { + "epoch": 0.26368, + "grad_norm": 0.12150184065103531, + "learning_rate": 0.0001722474226804124, + "loss": 0.5876, + "step": 824 + }, + { + "epoch": 0.264, + "grad_norm": 0.14796634018421173, + "learning_rate": 0.00017220618556701032, + "loss": 0.7244, + "step": 825 + }, + { + "epoch": 0.26432, + "grad_norm": 0.1227513924241066, + "learning_rate": 0.00017216494845360825, + "loss": 0.5743, + "step": 826 + }, + { + "epoch": 0.26464, + "grad_norm": 0.13068315386772156, + "learning_rate": 0.00017212371134020619, + "loss": 0.6706, + "step": 827 + }, + { + "epoch": 0.26496, + "grad_norm": 0.136739581823349, + "learning_rate": 0.00017208247422680412, + "loss": 0.6284, + "step": 828 + }, + { + "epoch": 0.26528, + "grad_norm": 0.11144915968179703, + "learning_rate": 0.00017204123711340208, + "loss": 0.8231, + "step": 829 + }, + { + "epoch": 0.2656, + "grad_norm": 0.11677107214927673, + "learning_rate": 0.000172, + "loss": 0.7092, + "step": 830 + }, + { + "epoch": 0.26592, + "grad_norm": 0.10681087523698807, + "learning_rate": 0.00017195876288659796, + "loss": 0.6751, + "step": 831 + }, + { + "epoch": 0.26624, + "grad_norm": 0.1254301369190216, + "learning_rate": 0.0001719175257731959, + "loss": 0.8069, + "step": 832 + }, + { + "epoch": 0.26656, + "grad_norm": 0.12878485023975372, + "learning_rate": 0.00017187628865979383, + "loss": 0.7752, + "step": 833 + }, + { + "epoch": 0.26688, + "grad_norm": 0.1316794753074646, + "learning_rate": 0.00017183505154639176, + "loss": 0.6108, + "step": 834 + }, + { + "epoch": 0.2672, + "grad_norm": 0.1294477880001068, + "learning_rate": 0.0001717938144329897, + "loss": 0.8802, + "step": 835 + }, + { + "epoch": 0.26752, + "grad_norm": 0.1193527951836586, + "learning_rate": 0.00017175257731958762, + "loss": 0.7253, + "step": 836 + }, + { + "epoch": 0.26784, + "grad_norm": 0.13821986317634583, + "learning_rate": 0.00017171134020618558, + "loss": 0.8121, + "step": 837 + }, + { + "epoch": 0.26816, + "grad_norm": 0.12366246432065964, + "learning_rate": 0.0001716701030927835, + "loss": 0.7343, + "step": 838 + }, + { + "epoch": 0.26848, + "grad_norm": 0.12942105531692505, + "learning_rate": 0.00017162886597938147, + "loss": 0.7198, + "step": 839 + }, + { + "epoch": 0.2688, + "grad_norm": 0.1249607726931572, + "learning_rate": 0.0001715876288659794, + "loss": 0.7917, + "step": 840 + }, + { + "epoch": 0.26912, + "grad_norm": 0.13555577397346497, + "learning_rate": 0.0001715463917525773, + "loss": 0.7047, + "step": 841 + }, + { + "epoch": 0.26944, + "grad_norm": 0.10971818119287491, + "learning_rate": 0.00017150515463917526, + "loss": 0.68, + "step": 842 + }, + { + "epoch": 0.26976, + "grad_norm": 0.1281420886516571, + "learning_rate": 0.0001714639175257732, + "loss": 0.765, + "step": 843 + }, + { + "epoch": 0.27008, + "grad_norm": 0.13114331662654877, + "learning_rate": 0.00017142268041237115, + "loss": 0.7864, + "step": 844 + }, + { + "epoch": 0.2704, + "grad_norm": 0.13292206823825836, + "learning_rate": 0.00017138144329896908, + "loss": 0.8031, + "step": 845 + }, + { + "epoch": 0.27072, + "grad_norm": 0.11316345632076263, + "learning_rate": 0.00017134020618556702, + "loss": 0.7657, + "step": 846 + }, + { + "epoch": 0.27104, + "grad_norm": 0.12936796247959137, + "learning_rate": 0.00017129896907216497, + "loss": 0.6948, + "step": 847 + }, + { + "epoch": 0.27136, + "grad_norm": 0.14884187281131744, + "learning_rate": 0.00017125773195876288, + "loss": 0.5843, + "step": 848 + }, + { + "epoch": 0.27168, + "grad_norm": 0.1351289451122284, + "learning_rate": 0.00017121649484536084, + "loss": 0.6618, + "step": 849 + }, + { + "epoch": 0.272, + "grad_norm": 0.12176259607076645, + "learning_rate": 0.00017117525773195877, + "loss": 0.5761, + "step": 850 + }, + { + "epoch": 0.27232, + "grad_norm": 0.15906400978565216, + "learning_rate": 0.0001711340206185567, + "loss": 0.8627, + "step": 851 + }, + { + "epoch": 0.27264, + "grad_norm": 0.1319553554058075, + "learning_rate": 0.00017109278350515466, + "loss": 0.6524, + "step": 852 + }, + { + "epoch": 0.27296, + "grad_norm": 0.13609087467193604, + "learning_rate": 0.0001710515463917526, + "loss": 0.666, + "step": 853 + }, + { + "epoch": 0.27328, + "grad_norm": 0.13189756870269775, + "learning_rate": 0.00017101030927835055, + "loss": 0.7922, + "step": 854 + }, + { + "epoch": 0.2736, + "grad_norm": 0.13593921065330505, + "learning_rate": 0.00017096907216494845, + "loss": 0.7556, + "step": 855 + }, + { + "epoch": 0.27392, + "grad_norm": 0.13250242173671722, + "learning_rate": 0.00017092783505154638, + "loss": 0.732, + "step": 856 + }, + { + "epoch": 0.27424, + "grad_norm": 0.12137633562088013, + "learning_rate": 0.00017088659793814434, + "loss": 0.7496, + "step": 857 + }, + { + "epoch": 0.27456, + "grad_norm": 0.1237613707780838, + "learning_rate": 0.00017084536082474227, + "loss": 0.5882, + "step": 858 + }, + { + "epoch": 0.27488, + "grad_norm": 0.13498494029045105, + "learning_rate": 0.00017080412371134023, + "loss": 0.7405, + "step": 859 + }, + { + "epoch": 0.2752, + "grad_norm": 0.11016131192445755, + "learning_rate": 0.00017076288659793816, + "loss": 0.9321, + "step": 860 + }, + { + "epoch": 0.27552, + "grad_norm": 0.14061477780342102, + "learning_rate": 0.0001707216494845361, + "loss": 0.7053, + "step": 861 + }, + { + "epoch": 0.27584, + "grad_norm": 0.14012236893177032, + "learning_rate": 0.00017068041237113402, + "loss": 0.7159, + "step": 862 + }, + { + "epoch": 0.27616, + "grad_norm": 0.13183481991291046, + "learning_rate": 0.00017063917525773196, + "loss": 0.7325, + "step": 863 + }, + { + "epoch": 0.27648, + "grad_norm": 0.11476800590753555, + "learning_rate": 0.0001705979381443299, + "loss": 0.7055, + "step": 864 + }, + { + "epoch": 0.2768, + "grad_norm": 0.10209871083498001, + "learning_rate": 0.00017055670103092785, + "loss": 0.7001, + "step": 865 + }, + { + "epoch": 0.27712, + "grad_norm": 0.11558199673891068, + "learning_rate": 0.00017051546391752578, + "loss": 0.6917, + "step": 866 + }, + { + "epoch": 0.27744, + "grad_norm": 0.12456093728542328, + "learning_rate": 0.00017047422680412373, + "loss": 0.7678, + "step": 867 + }, + { + "epoch": 0.27776, + "grad_norm": 0.1426069140434265, + "learning_rate": 0.00017043298969072167, + "loss": 0.6215, + "step": 868 + }, + { + "epoch": 0.27808, + "grad_norm": 0.12650014460086823, + "learning_rate": 0.0001703917525773196, + "loss": 0.7265, + "step": 869 + }, + { + "epoch": 0.2784, + "grad_norm": 0.12991954386234283, + "learning_rate": 0.00017035051546391753, + "loss": 0.7828, + "step": 870 + }, + { + "epoch": 0.27872, + "grad_norm": 0.1391548216342926, + "learning_rate": 0.00017030927835051546, + "loss": 0.6953, + "step": 871 + }, + { + "epoch": 0.27904, + "grad_norm": 0.12694767117500305, + "learning_rate": 0.00017026804123711342, + "loss": 0.6607, + "step": 872 + }, + { + "epoch": 0.27936, + "grad_norm": 0.10944822430610657, + "learning_rate": 0.00017022680412371135, + "loss": 0.6494, + "step": 873 + }, + { + "epoch": 0.27968, + "grad_norm": 0.12811000645160675, + "learning_rate": 0.00017018556701030928, + "loss": 0.5901, + "step": 874 + }, + { + "epoch": 0.28, + "grad_norm": 0.12633275985717773, + "learning_rate": 0.00017014432989690724, + "loss": 0.7171, + "step": 875 + }, + { + "epoch": 0.28032, + "grad_norm": 0.13156813383102417, + "learning_rate": 0.00017010309278350517, + "loss": 0.7358, + "step": 876 + }, + { + "epoch": 0.28064, + "grad_norm": 0.11998021602630615, + "learning_rate": 0.0001700618556701031, + "loss": 0.7273, + "step": 877 + }, + { + "epoch": 0.28096, + "grad_norm": 0.11613285541534424, + "learning_rate": 0.00017002061855670103, + "loss": 0.7788, + "step": 878 + }, + { + "epoch": 0.28128, + "grad_norm": 0.14196628332138062, + "learning_rate": 0.00016997938144329896, + "loss": 0.8546, + "step": 879 + }, + { + "epoch": 0.2816, + "grad_norm": 0.11939578503370285, + "learning_rate": 0.00016993814432989692, + "loss": 0.7764, + "step": 880 + }, + { + "epoch": 0.28192, + "grad_norm": 0.13980519771575928, + "learning_rate": 0.00016989690721649485, + "loss": 0.6063, + "step": 881 + }, + { + "epoch": 0.28224, + "grad_norm": 0.12722016870975494, + "learning_rate": 0.0001698556701030928, + "loss": 0.6469, + "step": 882 + }, + { + "epoch": 0.28256, + "grad_norm": 0.12645535171031952, + "learning_rate": 0.00016981443298969074, + "loss": 0.9321, + "step": 883 + }, + { + "epoch": 0.28288, + "grad_norm": 0.13195961713790894, + "learning_rate": 0.00016977319587628865, + "loss": 0.7791, + "step": 884 + }, + { + "epoch": 0.2832, + "grad_norm": 0.1446891874074936, + "learning_rate": 0.0001697319587628866, + "loss": 0.651, + "step": 885 + }, + { + "epoch": 0.28352, + "grad_norm": 0.1256592720746994, + "learning_rate": 0.00016969072164948454, + "loss": 0.7208, + "step": 886 + }, + { + "epoch": 0.28384, + "grad_norm": 0.11931982636451721, + "learning_rate": 0.0001696494845360825, + "loss": 0.6168, + "step": 887 + }, + { + "epoch": 0.28416, + "grad_norm": 0.11525401473045349, + "learning_rate": 0.00016960824742268043, + "loss": 0.6315, + "step": 888 + }, + { + "epoch": 0.28448, + "grad_norm": 0.12549898028373718, + "learning_rate": 0.00016956701030927836, + "loss": 0.636, + "step": 889 + }, + { + "epoch": 0.2848, + "grad_norm": 0.1158369705080986, + "learning_rate": 0.0001695257731958763, + "loss": 0.7766, + "step": 890 + }, + { + "epoch": 0.28512, + "grad_norm": 0.12144404649734497, + "learning_rate": 0.00016948453608247422, + "loss": 0.6679, + "step": 891 + }, + { + "epoch": 0.28544, + "grad_norm": 0.13141587376594543, + "learning_rate": 0.00016944329896907218, + "loss": 0.7929, + "step": 892 + }, + { + "epoch": 0.28576, + "grad_norm": 0.11869249492883682, + "learning_rate": 0.0001694020618556701, + "loss": 0.7054, + "step": 893 + }, + { + "epoch": 0.28608, + "grad_norm": 0.1343916803598404, + "learning_rate": 0.00016936082474226804, + "loss": 0.7307, + "step": 894 + }, + { + "epoch": 0.2864, + "grad_norm": 0.12864436209201813, + "learning_rate": 0.000169319587628866, + "loss": 0.7858, + "step": 895 + }, + { + "epoch": 0.28672, + "grad_norm": 0.11874640733003616, + "learning_rate": 0.00016927835051546393, + "loss": 0.7739, + "step": 896 + }, + { + "epoch": 0.28704, + "grad_norm": 0.1262228637933731, + "learning_rate": 0.00016923711340206186, + "loss": 0.8163, + "step": 897 + }, + { + "epoch": 0.28736, + "grad_norm": 0.1348988115787506, + "learning_rate": 0.0001691958762886598, + "loss": 0.8096, + "step": 898 + }, + { + "epoch": 0.28768, + "grad_norm": 0.11836285144090652, + "learning_rate": 0.00016915463917525773, + "loss": 0.8158, + "step": 899 + }, + { + "epoch": 0.288, + "grad_norm": 0.12167097628116608, + "learning_rate": 0.00016911340206185568, + "loss": 0.7138, + "step": 900 + }, + { + "epoch": 0.28832, + "grad_norm": 0.13505558669567108, + "learning_rate": 0.00016907216494845361, + "loss": 0.9069, + "step": 901 + }, + { + "epoch": 0.28864, + "grad_norm": 0.13912267982959747, + "learning_rate": 0.00016903092783505155, + "loss": 0.7116, + "step": 902 + }, + { + "epoch": 0.28896, + "grad_norm": 0.15731939673423767, + "learning_rate": 0.0001689896907216495, + "loss": 0.6876, + "step": 903 + }, + { + "epoch": 0.28928, + "grad_norm": 0.12537138164043427, + "learning_rate": 0.00016894845360824744, + "loss": 0.6171, + "step": 904 + }, + { + "epoch": 0.2896, + "grad_norm": 0.13225343823432922, + "learning_rate": 0.00016890721649484537, + "loss": 0.6375, + "step": 905 + }, + { + "epoch": 0.28992, + "grad_norm": 0.12211292237043381, + "learning_rate": 0.0001688659793814433, + "loss": 0.732, + "step": 906 + }, + { + "epoch": 0.29024, + "grad_norm": 0.12123337388038635, + "learning_rate": 0.00016882474226804123, + "loss": 0.8403, + "step": 907 + }, + { + "epoch": 0.29056, + "grad_norm": 0.11805325746536255, + "learning_rate": 0.0001687835051546392, + "loss": 0.7806, + "step": 908 + }, + { + "epoch": 0.29088, + "grad_norm": 0.12412726879119873, + "learning_rate": 0.00016874226804123712, + "loss": 0.5854, + "step": 909 + }, + { + "epoch": 0.2912, + "grad_norm": 0.12389811873435974, + "learning_rate": 0.00016870103092783508, + "loss": 0.6711, + "step": 910 + }, + { + "epoch": 0.29152, + "grad_norm": 0.12261383980512619, + "learning_rate": 0.000168659793814433, + "loss": 0.7261, + "step": 911 + }, + { + "epoch": 0.29184, + "grad_norm": 0.14904655516147614, + "learning_rate": 0.00016861855670103094, + "loss": 0.6976, + "step": 912 + }, + { + "epoch": 0.29216, + "grad_norm": 0.11271169781684875, + "learning_rate": 0.00016857731958762887, + "loss": 0.7185, + "step": 913 + }, + { + "epoch": 0.29248, + "grad_norm": 0.11681025475263596, + "learning_rate": 0.0001685360824742268, + "loss": 0.7155, + "step": 914 + }, + { + "epoch": 0.2928, + "grad_norm": 0.12385556101799011, + "learning_rate": 0.00016849484536082476, + "loss": 0.7431, + "step": 915 + }, + { + "epoch": 0.29312, + "grad_norm": 0.13570012152194977, + "learning_rate": 0.0001684536082474227, + "loss": 0.8334, + "step": 916 + }, + { + "epoch": 0.29344, + "grad_norm": 0.14303314685821533, + "learning_rate": 0.00016841237113402062, + "loss": 0.7757, + "step": 917 + }, + { + "epoch": 0.29376, + "grad_norm": 0.13650397956371307, + "learning_rate": 0.00016837113402061858, + "loss": 0.6664, + "step": 918 + }, + { + "epoch": 0.29408, + "grad_norm": 0.14622308313846588, + "learning_rate": 0.0001683298969072165, + "loss": 0.6503, + "step": 919 + }, + { + "epoch": 0.2944, + "grad_norm": 0.1545690894126892, + "learning_rate": 0.00016828865979381444, + "loss": 0.6802, + "step": 920 + }, + { + "epoch": 0.29472, + "grad_norm": 0.12770286202430725, + "learning_rate": 0.00016824742268041238, + "loss": 0.5676, + "step": 921 + }, + { + "epoch": 0.29504, + "grad_norm": 0.1302863508462906, + "learning_rate": 0.0001682061855670103, + "loss": 0.7108, + "step": 922 + }, + { + "epoch": 0.29536, + "grad_norm": 0.12539339065551758, + "learning_rate": 0.00016816494845360827, + "loss": 0.7109, + "step": 923 + }, + { + "epoch": 0.29568, + "grad_norm": 0.13586951792240143, + "learning_rate": 0.0001681237113402062, + "loss": 0.7819, + "step": 924 + }, + { + "epoch": 0.296, + "grad_norm": 0.13007107377052307, + "learning_rate": 0.00016808247422680413, + "loss": 0.7138, + "step": 925 + }, + { + "epoch": 0.29632, + "grad_norm": 0.11121190339326859, + "learning_rate": 0.00016804123711340206, + "loss": 0.6099, + "step": 926 + }, + { + "epoch": 0.29664, + "grad_norm": 0.1328391581773758, + "learning_rate": 0.000168, + "loss": 0.8115, + "step": 927 + }, + { + "epoch": 0.29696, + "grad_norm": 0.12316712737083435, + "learning_rate": 0.00016795876288659795, + "loss": 0.6743, + "step": 928 + }, + { + "epoch": 0.29728, + "grad_norm": 0.13188757002353668, + "learning_rate": 0.00016791752577319588, + "loss": 0.7733, + "step": 929 + }, + { + "epoch": 0.2976, + "grad_norm": 0.1288207471370697, + "learning_rate": 0.0001678762886597938, + "loss": 0.6478, + "step": 930 + }, + { + "epoch": 0.29792, + "grad_norm": 0.13399000465869904, + "learning_rate": 0.00016783505154639177, + "loss": 0.7069, + "step": 931 + }, + { + "epoch": 0.29824, + "grad_norm": 0.14570803940296173, + "learning_rate": 0.0001677938144329897, + "loss": 0.7701, + "step": 932 + }, + { + "epoch": 0.29856, + "grad_norm": 0.12334295362234116, + "learning_rate": 0.00016775257731958763, + "loss": 0.7331, + "step": 933 + }, + { + "epoch": 0.29888, + "grad_norm": 0.12136422842741013, + "learning_rate": 0.00016771134020618556, + "loss": 0.6297, + "step": 934 + }, + { + "epoch": 0.2992, + "grad_norm": 0.12953773140907288, + "learning_rate": 0.0001676701030927835, + "loss": 0.6675, + "step": 935 + }, + { + "epoch": 0.29952, + "grad_norm": 0.13175757229328156, + "learning_rate": 0.00016762886597938145, + "loss": 0.8289, + "step": 936 + }, + { + "epoch": 0.29984, + "grad_norm": 0.12782573699951172, + "learning_rate": 0.00016758762886597938, + "loss": 0.7401, + "step": 937 + }, + { + "epoch": 0.30016, + "grad_norm": 0.13902167975902557, + "learning_rate": 0.00016754639175257734, + "loss": 0.6949, + "step": 938 + }, + { + "epoch": 0.30048, + "grad_norm": 0.1445343792438507, + "learning_rate": 0.00016750515463917527, + "loss": 0.6779, + "step": 939 + }, + { + "epoch": 0.3008, + "grad_norm": 0.12114536762237549, + "learning_rate": 0.0001674639175257732, + "loss": 0.6658, + "step": 940 + }, + { + "epoch": 0.30112, + "grad_norm": 0.13642756640911102, + "learning_rate": 0.00016742268041237114, + "loss": 0.6336, + "step": 941 + }, + { + "epoch": 0.30144, + "grad_norm": 0.12124267965555191, + "learning_rate": 0.00016738144329896907, + "loss": 0.6917, + "step": 942 + }, + { + "epoch": 0.30176, + "grad_norm": 0.1574261337518692, + "learning_rate": 0.00016734020618556703, + "loss": 0.6723, + "step": 943 + }, + { + "epoch": 0.30208, + "grad_norm": 0.1327444165945053, + "learning_rate": 0.00016729896907216496, + "loss": 0.7765, + "step": 944 + }, + { + "epoch": 0.3024, + "grad_norm": 0.12544023990631104, + "learning_rate": 0.0001672577319587629, + "loss": 0.7213, + "step": 945 + }, + { + "epoch": 0.30272, + "grad_norm": 0.12831197679042816, + "learning_rate": 0.00016721649484536085, + "loss": 0.624, + "step": 946 + }, + { + "epoch": 0.30304, + "grad_norm": 0.13576100766658783, + "learning_rate": 0.00016717525773195878, + "loss": 0.6694, + "step": 947 + }, + { + "epoch": 0.30336, + "grad_norm": 0.14174100756645203, + "learning_rate": 0.0001671340206185567, + "loss": 0.7847, + "step": 948 + }, + { + "epoch": 0.30368, + "grad_norm": 0.14193561673164368, + "learning_rate": 0.00016709278350515464, + "loss": 0.638, + "step": 949 + }, + { + "epoch": 0.304, + "grad_norm": 0.12588457763195038, + "learning_rate": 0.00016705154639175257, + "loss": 0.7496, + "step": 950 + }, + { + "epoch": 0.30432, + "grad_norm": 0.14540141820907593, + "learning_rate": 0.00016701030927835053, + "loss": 0.7611, + "step": 951 + }, + { + "epoch": 0.30464, + "grad_norm": 0.1433524489402771, + "learning_rate": 0.00016696907216494846, + "loss": 0.7602, + "step": 952 + }, + { + "epoch": 0.30496, + "grad_norm": 0.12800182402133942, + "learning_rate": 0.00016692783505154642, + "loss": 0.7136, + "step": 953 + }, + { + "epoch": 0.30528, + "grad_norm": 0.14627426862716675, + "learning_rate": 0.00016688659793814435, + "loss": 0.7954, + "step": 954 + }, + { + "epoch": 0.3056, + "grad_norm": 0.11950758844614029, + "learning_rate": 0.00016684536082474226, + "loss": 0.6012, + "step": 955 + }, + { + "epoch": 0.30592, + "grad_norm": 0.14980708062648773, + "learning_rate": 0.00016680412371134021, + "loss": 0.6755, + "step": 956 + }, + { + "epoch": 0.30624, + "grad_norm": 0.1280095875263214, + "learning_rate": 0.00016676288659793815, + "loss": 0.691, + "step": 957 + }, + { + "epoch": 0.30656, + "grad_norm": 0.12384190410375595, + "learning_rate": 0.00016672164948453608, + "loss": 0.7097, + "step": 958 + }, + { + "epoch": 0.30688, + "grad_norm": 0.11612387746572495, + "learning_rate": 0.00016668041237113404, + "loss": 0.6998, + "step": 959 + }, + { + "epoch": 0.3072, + "grad_norm": 0.11533758789300919, + "learning_rate": 0.00016663917525773197, + "loss": 0.5825, + "step": 960 + }, + { + "epoch": 0.30752, + "grad_norm": 0.13979220390319824, + "learning_rate": 0.00016659793814432993, + "loss": 0.887, + "step": 961 + }, + { + "epoch": 0.30784, + "grad_norm": 0.12515367567539215, + "learning_rate": 0.00016655670103092783, + "loss": 0.7497, + "step": 962 + }, + { + "epoch": 0.30816, + "grad_norm": 0.11801932007074356, + "learning_rate": 0.00016651546391752576, + "loss": 0.6204, + "step": 963 + }, + { + "epoch": 0.30848, + "grad_norm": 0.1284782588481903, + "learning_rate": 0.00016647422680412372, + "loss": 0.8, + "step": 964 + }, + { + "epoch": 0.3088, + "grad_norm": 0.11822401732206345, + "learning_rate": 0.00016643298969072165, + "loss": 0.6733, + "step": 965 + }, + { + "epoch": 0.30912, + "grad_norm": 0.1323615163564682, + "learning_rate": 0.0001663917525773196, + "loss": 0.6603, + "step": 966 + }, + { + "epoch": 0.30944, + "grad_norm": 0.1403515785932541, + "learning_rate": 0.00016635051546391754, + "loss": 0.7228, + "step": 967 + }, + { + "epoch": 0.30976, + "grad_norm": 0.13995344936847687, + "learning_rate": 0.00016630927835051547, + "loss": 0.7326, + "step": 968 + }, + { + "epoch": 0.31008, + "grad_norm": 0.12779420614242554, + "learning_rate": 0.0001662680412371134, + "loss": 0.7643, + "step": 969 + }, + { + "epoch": 0.3104, + "grad_norm": 0.13067853450775146, + "learning_rate": 0.00016622680412371133, + "loss": 0.584, + "step": 970 + }, + { + "epoch": 0.31072, + "grad_norm": 0.1376170516014099, + "learning_rate": 0.0001661855670103093, + "loss": 0.7033, + "step": 971 + }, + { + "epoch": 0.31104, + "grad_norm": 0.13981421291828156, + "learning_rate": 0.00016614432989690722, + "loss": 0.6585, + "step": 972 + }, + { + "epoch": 0.31136, + "grad_norm": 0.13018488883972168, + "learning_rate": 0.00016610309278350515, + "loss": 0.7505, + "step": 973 + }, + { + "epoch": 0.31168, + "grad_norm": 0.13596870005130768, + "learning_rate": 0.0001660618556701031, + "loss": 0.6731, + "step": 974 + }, + { + "epoch": 0.312, + "grad_norm": 0.12909188866615295, + "learning_rate": 0.00016602061855670104, + "loss": 0.7431, + "step": 975 + }, + { + "epoch": 0.31232, + "grad_norm": 0.12600624561309814, + "learning_rate": 0.00016597938144329898, + "loss": 0.8064, + "step": 976 + }, + { + "epoch": 0.31264, + "grad_norm": 0.12609797716140747, + "learning_rate": 0.0001659381443298969, + "loss": 0.856, + "step": 977 + }, + { + "epoch": 0.31296, + "grad_norm": 0.1194164827466011, + "learning_rate": 0.00016589690721649484, + "loss": 0.7126, + "step": 978 + }, + { + "epoch": 0.31328, + "grad_norm": 0.12827666103839874, + "learning_rate": 0.0001658556701030928, + "loss": 0.6233, + "step": 979 + }, + { + "epoch": 0.3136, + "grad_norm": 0.12019889056682587, + "learning_rate": 0.00016581443298969073, + "loss": 0.5571, + "step": 980 + }, + { + "epoch": 0.31392, + "grad_norm": 0.12774929404258728, + "learning_rate": 0.00016577319587628869, + "loss": 0.6896, + "step": 981 + }, + { + "epoch": 0.31424, + "grad_norm": 0.13669483363628387, + "learning_rate": 0.00016573195876288662, + "loss": 0.6872, + "step": 982 + }, + { + "epoch": 0.31456, + "grad_norm": 0.12400145828723907, + "learning_rate": 0.00016569072164948455, + "loss": 0.8399, + "step": 983 + }, + { + "epoch": 0.31488, + "grad_norm": 0.14067703485488892, + "learning_rate": 0.00016564948453608248, + "loss": 0.7675, + "step": 984 + }, + { + "epoch": 0.3152, + "grad_norm": 0.13461582362651825, + "learning_rate": 0.0001656082474226804, + "loss": 0.7943, + "step": 985 + }, + { + "epoch": 0.31552, + "grad_norm": 0.13829675316810608, + "learning_rate": 0.00016556701030927834, + "loss": 0.8666, + "step": 986 + }, + { + "epoch": 0.31584, + "grad_norm": 0.13706457614898682, + "learning_rate": 0.0001655257731958763, + "loss": 0.757, + "step": 987 + }, + { + "epoch": 0.31616, + "grad_norm": 0.1262834221124649, + "learning_rate": 0.00016548453608247423, + "loss": 0.6941, + "step": 988 + }, + { + "epoch": 0.31648, + "grad_norm": 0.15362948179244995, + "learning_rate": 0.0001654432989690722, + "loss": 0.8656, + "step": 989 + }, + { + "epoch": 0.3168, + "grad_norm": 0.11257731169462204, + "learning_rate": 0.00016540206185567012, + "loss": 0.6426, + "step": 990 + }, + { + "epoch": 0.31712, + "grad_norm": 0.14325182139873505, + "learning_rate": 0.00016536082474226803, + "loss": 0.7389, + "step": 991 + }, + { + "epoch": 0.31744, + "grad_norm": 0.12336672842502594, + "learning_rate": 0.00016531958762886598, + "loss": 0.7493, + "step": 992 + }, + { + "epoch": 0.31776, + "grad_norm": 0.14396019279956818, + "learning_rate": 0.00016527835051546392, + "loss": 0.7598, + "step": 993 + }, + { + "epoch": 0.31808, + "grad_norm": 0.1333467960357666, + "learning_rate": 0.00016523711340206187, + "loss": 0.7863, + "step": 994 + }, + { + "epoch": 0.3184, + "grad_norm": 0.1304144263267517, + "learning_rate": 0.0001651958762886598, + "loss": 0.7304, + "step": 995 + }, + { + "epoch": 0.31872, + "grad_norm": 0.12357225269079208, + "learning_rate": 0.00016515463917525774, + "loss": 0.6275, + "step": 996 + }, + { + "epoch": 0.31904, + "grad_norm": 0.11936910450458527, + "learning_rate": 0.0001651134020618557, + "loss": 0.7266, + "step": 997 + }, + { + "epoch": 0.31936, + "grad_norm": 0.12361844629049301, + "learning_rate": 0.0001650721649484536, + "loss": 0.7156, + "step": 998 + }, + { + "epoch": 0.31968, + "grad_norm": 0.12392809987068176, + "learning_rate": 0.00016503092783505156, + "loss": 0.7804, + "step": 999 + }, + { + "epoch": 0.32, + "grad_norm": 0.13606688380241394, + "learning_rate": 0.0001649896907216495, + "loss": 0.9582, + "step": 1000 + }, + { + "epoch": 0.32032, + "grad_norm": 0.1291663497686386, + "learning_rate": 0.00016494845360824742, + "loss": 0.6795, + "step": 1001 + }, + { + "epoch": 0.32064, + "grad_norm": 0.13042348623275757, + "learning_rate": 0.00016490721649484538, + "loss": 0.729, + "step": 1002 + }, + { + "epoch": 0.32096, + "grad_norm": 0.12246596813201904, + "learning_rate": 0.0001648659793814433, + "loss": 0.7758, + "step": 1003 + }, + { + "epoch": 0.32128, + "grad_norm": 0.12324332445859909, + "learning_rate": 0.00016482474226804127, + "loss": 0.6839, + "step": 1004 + }, + { + "epoch": 0.3216, + "grad_norm": 0.13313184678554535, + "learning_rate": 0.00016478350515463917, + "loss": 0.7324, + "step": 1005 + }, + { + "epoch": 0.32192, + "grad_norm": 0.1304756999015808, + "learning_rate": 0.0001647422680412371, + "loss": 0.6432, + "step": 1006 + }, + { + "epoch": 0.32224, + "grad_norm": 0.12304502725601196, + "learning_rate": 0.00016470103092783506, + "loss": 0.722, + "step": 1007 + }, + { + "epoch": 0.32256, + "grad_norm": 0.13414014875888824, + "learning_rate": 0.000164659793814433, + "loss": 0.6226, + "step": 1008 + }, + { + "epoch": 0.32288, + "grad_norm": 0.1390489637851715, + "learning_rate": 0.00016461855670103095, + "loss": 0.7368, + "step": 1009 + }, + { + "epoch": 0.3232, + "grad_norm": 0.13376174867153168, + "learning_rate": 0.00016457731958762888, + "loss": 0.9653, + "step": 1010 + }, + { + "epoch": 0.32352, + "grad_norm": 0.13524703681468964, + "learning_rate": 0.00016453608247422681, + "loss": 0.7106, + "step": 1011 + }, + { + "epoch": 0.32384, + "grad_norm": 0.1215464323759079, + "learning_rate": 0.00016449484536082475, + "loss": 0.6942, + "step": 1012 + }, + { + "epoch": 0.32416, + "grad_norm": 0.13397525250911713, + "learning_rate": 0.00016445360824742268, + "loss": 0.684, + "step": 1013 + }, + { + "epoch": 0.32448, + "grad_norm": 0.1190156638622284, + "learning_rate": 0.00016441237113402063, + "loss": 0.7029, + "step": 1014 + }, + { + "epoch": 0.3248, + "grad_norm": 0.13238269090652466, + "learning_rate": 0.00016437113402061857, + "loss": 0.7614, + "step": 1015 + }, + { + "epoch": 0.32512, + "grad_norm": 0.1246226504445076, + "learning_rate": 0.0001643298969072165, + "loss": 0.7804, + "step": 1016 + }, + { + "epoch": 0.32544, + "grad_norm": 0.13241590559482574, + "learning_rate": 0.00016428865979381446, + "loss": 0.7454, + "step": 1017 + }, + { + "epoch": 0.32576, + "grad_norm": 0.12619127333164215, + "learning_rate": 0.0001642474226804124, + "loss": 0.6793, + "step": 1018 + }, + { + "epoch": 0.32608, + "grad_norm": 0.128580242395401, + "learning_rate": 0.00016420618556701032, + "loss": 0.7866, + "step": 1019 + }, + { + "epoch": 0.3264, + "grad_norm": 0.14162757992744446, + "learning_rate": 0.00016416494845360825, + "loss": 0.8144, + "step": 1020 + }, + { + "epoch": 0.32672, + "grad_norm": 0.15262079238891602, + "learning_rate": 0.00016412371134020618, + "loss": 0.7865, + "step": 1021 + }, + { + "epoch": 0.32704, + "grad_norm": 0.1325531303882599, + "learning_rate": 0.00016408247422680414, + "loss": 0.6776, + "step": 1022 + }, + { + "epoch": 0.32736, + "grad_norm": 0.12302196770906448, + "learning_rate": 0.00016404123711340207, + "loss": 0.8476, + "step": 1023 + }, + { + "epoch": 0.32768, + "grad_norm": 0.12806765735149384, + "learning_rate": 0.000164, + "loss": 0.8215, + "step": 1024 + }, + { + "epoch": 0.328, + "grad_norm": 0.11527955532073975, + "learning_rate": 0.00016395876288659796, + "loss": 0.8466, + "step": 1025 + }, + { + "epoch": 0.32832, + "grad_norm": 0.11988075077533722, + "learning_rate": 0.0001639175257731959, + "loss": 0.7193, + "step": 1026 + }, + { + "epoch": 0.32864, + "grad_norm": 0.11877843737602234, + "learning_rate": 0.00016387628865979382, + "loss": 0.6518, + "step": 1027 + }, + { + "epoch": 0.32896, + "grad_norm": 0.11726564168930054, + "learning_rate": 0.00016383505154639175, + "loss": 0.6005, + "step": 1028 + }, + { + "epoch": 0.32928, + "grad_norm": 0.1332593411207199, + "learning_rate": 0.00016379381443298969, + "loss": 0.7817, + "step": 1029 + }, + { + "epoch": 0.3296, + "grad_norm": 0.13856275379657745, + "learning_rate": 0.00016375257731958764, + "loss": 0.8255, + "step": 1030 + }, + { + "epoch": 0.32992, + "grad_norm": 0.10795895755290985, + "learning_rate": 0.00016371134020618558, + "loss": 0.5919, + "step": 1031 + }, + { + "epoch": 0.33024, + "grad_norm": 0.12465250492095947, + "learning_rate": 0.00016367010309278353, + "loss": 0.6625, + "step": 1032 + }, + { + "epoch": 0.33056, + "grad_norm": 0.12707003951072693, + "learning_rate": 0.00016362886597938146, + "loss": 0.8995, + "step": 1033 + }, + { + "epoch": 0.33088, + "grad_norm": 0.137173593044281, + "learning_rate": 0.00016358762886597937, + "loss": 0.726, + "step": 1034 + }, + { + "epoch": 0.3312, + "grad_norm": 0.12696483731269836, + "learning_rate": 0.00016354639175257733, + "loss": 0.5944, + "step": 1035 + }, + { + "epoch": 0.33152, + "grad_norm": 0.1236334815621376, + "learning_rate": 0.00016350515463917526, + "loss": 0.6641, + "step": 1036 + }, + { + "epoch": 0.33184, + "grad_norm": 0.143971785902977, + "learning_rate": 0.00016346391752577322, + "loss": 0.7842, + "step": 1037 + }, + { + "epoch": 0.33216, + "grad_norm": 0.12057577818632126, + "learning_rate": 0.00016342268041237115, + "loss": 0.6724, + "step": 1038 + }, + { + "epoch": 0.33248, + "grad_norm": 0.15648114681243896, + "learning_rate": 0.00016338144329896908, + "loss": 0.9973, + "step": 1039 + }, + { + "epoch": 0.3328, + "grad_norm": 0.1502295732498169, + "learning_rate": 0.000163340206185567, + "loss": 0.6742, + "step": 1040 + }, + { + "epoch": 0.33312, + "grad_norm": 0.1341315358877182, + "learning_rate": 0.00016329896907216494, + "loss": 0.7678, + "step": 1041 + }, + { + "epoch": 0.33344, + "grad_norm": 0.10982177406549454, + "learning_rate": 0.0001632577319587629, + "loss": 0.7545, + "step": 1042 + }, + { + "epoch": 0.33376, + "grad_norm": 0.14167900383472443, + "learning_rate": 0.00016321649484536083, + "loss": 0.7139, + "step": 1043 + }, + { + "epoch": 0.33408, + "grad_norm": 0.11281700432300568, + "learning_rate": 0.00016317525773195876, + "loss": 0.7498, + "step": 1044 + }, + { + "epoch": 0.3344, + "grad_norm": 0.13683807849884033, + "learning_rate": 0.00016313402061855672, + "loss": 0.6963, + "step": 1045 + }, + { + "epoch": 0.33472, + "grad_norm": 0.1305110901594162, + "learning_rate": 0.00016309278350515465, + "loss": 0.6053, + "step": 1046 + }, + { + "epoch": 0.33504, + "grad_norm": 0.136047825217247, + "learning_rate": 0.00016305154639175258, + "loss": 0.7007, + "step": 1047 + }, + { + "epoch": 0.33536, + "grad_norm": 0.1397719383239746, + "learning_rate": 0.00016301030927835052, + "loss": 0.8724, + "step": 1048 + }, + { + "epoch": 0.33568, + "grad_norm": 0.13805080950260162, + "learning_rate": 0.00016296907216494845, + "loss": 0.7124, + "step": 1049 + }, + { + "epoch": 0.336, + "grad_norm": 0.14922751486301422, + "learning_rate": 0.0001629278350515464, + "loss": 0.7945, + "step": 1050 + }, + { + "epoch": 0.33632, + "grad_norm": 0.1439272165298462, + "learning_rate": 0.00016288659793814434, + "loss": 0.7118, + "step": 1051 + }, + { + "epoch": 0.33664, + "grad_norm": 0.11644583195447922, + "learning_rate": 0.00016284536082474227, + "loss": 0.8287, + "step": 1052 + }, + { + "epoch": 0.33696, + "grad_norm": 0.13248111307621002, + "learning_rate": 0.00016280412371134023, + "loss": 0.7215, + "step": 1053 + }, + { + "epoch": 0.33728, + "grad_norm": 0.13736796379089355, + "learning_rate": 0.00016276288659793816, + "loss": 0.7327, + "step": 1054 + }, + { + "epoch": 0.3376, + "grad_norm": 0.13475042581558228, + "learning_rate": 0.0001627216494845361, + "loss": 0.8121, + "step": 1055 + }, + { + "epoch": 0.33792, + "grad_norm": 0.12516193091869354, + "learning_rate": 0.00016268041237113402, + "loss": 0.7947, + "step": 1056 + }, + { + "epoch": 0.33824, + "grad_norm": 0.1351221799850464, + "learning_rate": 0.00016263917525773195, + "loss": 0.5659, + "step": 1057 + }, + { + "epoch": 0.33856, + "grad_norm": 0.12663671374320984, + "learning_rate": 0.0001625979381443299, + "loss": 0.6308, + "step": 1058 + }, + { + "epoch": 0.33888, + "grad_norm": 0.13254615664482117, + "learning_rate": 0.00016255670103092784, + "loss": 0.6656, + "step": 1059 + }, + { + "epoch": 0.3392, + "grad_norm": 0.1433473527431488, + "learning_rate": 0.0001625154639175258, + "loss": 0.808, + "step": 1060 + }, + { + "epoch": 0.33952, + "grad_norm": 0.12944042682647705, + "learning_rate": 0.00016247422680412373, + "loss": 0.6709, + "step": 1061 + }, + { + "epoch": 0.33984, + "grad_norm": 0.13818645477294922, + "learning_rate": 0.00016243298969072166, + "loss": 0.7565, + "step": 1062 + }, + { + "epoch": 0.34016, + "grad_norm": 0.12289980053901672, + "learning_rate": 0.0001623917525773196, + "loss": 0.7579, + "step": 1063 + }, + { + "epoch": 0.34048, + "grad_norm": 0.1363401859998703, + "learning_rate": 0.00016235051546391752, + "loss": 0.7677, + "step": 1064 + }, + { + "epoch": 0.3408, + "grad_norm": 0.12801024317741394, + "learning_rate": 0.00016230927835051548, + "loss": 0.5706, + "step": 1065 + }, + { + "epoch": 0.34112, + "grad_norm": 0.1203409880399704, + "learning_rate": 0.00016226804123711341, + "loss": 0.5443, + "step": 1066 + }, + { + "epoch": 0.34144, + "grad_norm": 0.14376769959926605, + "learning_rate": 0.00016222680412371134, + "loss": 0.7202, + "step": 1067 + }, + { + "epoch": 0.34176, + "grad_norm": 0.11748041212558746, + "learning_rate": 0.0001621855670103093, + "loss": 0.8342, + "step": 1068 + }, + { + "epoch": 0.34208, + "grad_norm": 0.13433270156383514, + "learning_rate": 0.00016214432989690723, + "loss": 0.7167, + "step": 1069 + }, + { + "epoch": 0.3424, + "grad_norm": 0.12405699491500854, + "learning_rate": 0.00016210309278350517, + "loss": 0.6997, + "step": 1070 + }, + { + "epoch": 0.34272, + "grad_norm": 0.13520316779613495, + "learning_rate": 0.0001620618556701031, + "loss": 0.8303, + "step": 1071 + }, + { + "epoch": 0.34304, + "grad_norm": 0.12949883937835693, + "learning_rate": 0.00016202061855670103, + "loss": 0.5729, + "step": 1072 + }, + { + "epoch": 0.34336, + "grad_norm": 0.11795380711555481, + "learning_rate": 0.000161979381443299, + "loss": 0.7703, + "step": 1073 + }, + { + "epoch": 0.34368, + "grad_norm": 0.13385391235351562, + "learning_rate": 0.00016193814432989692, + "loss": 0.5315, + "step": 1074 + }, + { + "epoch": 0.344, + "grad_norm": 0.11290988326072693, + "learning_rate": 0.00016189690721649488, + "loss": 0.6147, + "step": 1075 + }, + { + "epoch": 0.34432, + "grad_norm": 0.14130225777626038, + "learning_rate": 0.00016185567010309278, + "loss": 0.7681, + "step": 1076 + }, + { + "epoch": 0.34464, + "grad_norm": 0.15834380686283112, + "learning_rate": 0.0001618144329896907, + "loss": 0.7883, + "step": 1077 + }, + { + "epoch": 0.34496, + "grad_norm": 0.13480958342552185, + "learning_rate": 0.00016177319587628867, + "loss": 0.7518, + "step": 1078 + }, + { + "epoch": 0.34528, + "grad_norm": 0.12487983703613281, + "learning_rate": 0.0001617319587628866, + "loss": 0.7012, + "step": 1079 + }, + { + "epoch": 0.3456, + "grad_norm": 0.12848694622516632, + "learning_rate": 0.00016169072164948453, + "loss": 0.862, + "step": 1080 + }, + { + "epoch": 0.34592, + "grad_norm": 0.13103242218494415, + "learning_rate": 0.0001616494845360825, + "loss": 0.7836, + "step": 1081 + }, + { + "epoch": 0.34624, + "grad_norm": 0.12427806109189987, + "learning_rate": 0.00016160824742268042, + "loss": 0.6948, + "step": 1082 + }, + { + "epoch": 0.34656, + "grad_norm": 0.1163354143500328, + "learning_rate": 0.00016156701030927835, + "loss": 0.638, + "step": 1083 + }, + { + "epoch": 0.34688, + "grad_norm": 0.11864982545375824, + "learning_rate": 0.00016152577319587628, + "loss": 0.7002, + "step": 1084 + }, + { + "epoch": 0.3472, + "grad_norm": 0.12704510986804962, + "learning_rate": 0.00016148453608247422, + "loss": 0.6058, + "step": 1085 + }, + { + "epoch": 0.34752, + "grad_norm": 0.1170935109257698, + "learning_rate": 0.00016144329896907217, + "loss": 0.7592, + "step": 1086 + }, + { + "epoch": 0.34784, + "grad_norm": 0.1298816204071045, + "learning_rate": 0.0001614020618556701, + "loss": 0.7716, + "step": 1087 + }, + { + "epoch": 0.34816, + "grad_norm": 0.12749038636684418, + "learning_rate": 0.00016136082474226806, + "loss": 0.6785, + "step": 1088 + }, + { + "epoch": 0.34848, + "grad_norm": 0.13424167037010193, + "learning_rate": 0.000161319587628866, + "loss": 0.6733, + "step": 1089 + }, + { + "epoch": 0.3488, + "grad_norm": 0.14593367278575897, + "learning_rate": 0.00016127835051546393, + "loss": 0.8175, + "step": 1090 + }, + { + "epoch": 0.34912, + "grad_norm": 0.13538609445095062, + "learning_rate": 0.00016123711340206186, + "loss": 0.8064, + "step": 1091 + }, + { + "epoch": 0.34944, + "grad_norm": 0.1283688247203827, + "learning_rate": 0.0001611958762886598, + "loss": 0.7415, + "step": 1092 + }, + { + "epoch": 0.34976, + "grad_norm": 0.11879759281873703, + "learning_rate": 0.00016115463917525775, + "loss": 0.6461, + "step": 1093 + }, + { + "epoch": 0.35008, + "grad_norm": 0.15270709991455078, + "learning_rate": 0.00016111340206185568, + "loss": 0.6833, + "step": 1094 + }, + { + "epoch": 0.3504, + "grad_norm": 0.14082631468772888, + "learning_rate": 0.0001610721649484536, + "loss": 0.8527, + "step": 1095 + }, + { + "epoch": 0.35072, + "grad_norm": 0.13917644321918488, + "learning_rate": 0.00016103092783505157, + "loss": 0.6855, + "step": 1096 + }, + { + "epoch": 0.35104, + "grad_norm": 0.12534797191619873, + "learning_rate": 0.0001609896907216495, + "loss": 0.7396, + "step": 1097 + }, + { + "epoch": 0.35136, + "grad_norm": 0.1246245950460434, + "learning_rate": 0.00016094845360824743, + "loss": 0.7373, + "step": 1098 + }, + { + "epoch": 0.35168, + "grad_norm": 0.1447897106409073, + "learning_rate": 0.00016090721649484536, + "loss": 0.7252, + "step": 1099 + }, + { + "epoch": 0.352, + "grad_norm": 0.1323324739933014, + "learning_rate": 0.0001608659793814433, + "loss": 0.6699, + "step": 1100 + }, + { + "epoch": 0.35232, + "grad_norm": 0.12323680520057678, + "learning_rate": 0.00016082474226804125, + "loss": 0.7758, + "step": 1101 + }, + { + "epoch": 0.35264, + "grad_norm": 0.1507241576910019, + "learning_rate": 0.00016078350515463918, + "loss": 0.7029, + "step": 1102 + }, + { + "epoch": 0.35296, + "grad_norm": 0.1255369633436203, + "learning_rate": 0.00016074226804123714, + "loss": 0.7446, + "step": 1103 + }, + { + "epoch": 0.35328, + "grad_norm": 0.1623598039150238, + "learning_rate": 0.00016070103092783507, + "loss": 0.8059, + "step": 1104 + }, + { + "epoch": 0.3536, + "grad_norm": 0.14316579699516296, + "learning_rate": 0.00016065979381443298, + "loss": 0.7224, + "step": 1105 + }, + { + "epoch": 0.35392, + "grad_norm": 0.12849082052707672, + "learning_rate": 0.00016061855670103094, + "loss": 0.5966, + "step": 1106 + }, + { + "epoch": 0.35424, + "grad_norm": 0.11844538897275925, + "learning_rate": 0.00016057731958762887, + "loss": 0.6075, + "step": 1107 + }, + { + "epoch": 0.35456, + "grad_norm": 0.14459797739982605, + "learning_rate": 0.0001605360824742268, + "loss": 0.641, + "step": 1108 + }, + { + "epoch": 0.35488, + "grad_norm": 0.12330617755651474, + "learning_rate": 0.00016049484536082476, + "loss": 0.6642, + "step": 1109 + }, + { + "epoch": 0.3552, + "grad_norm": 0.12356940656900406, + "learning_rate": 0.0001604536082474227, + "loss": 0.6615, + "step": 1110 + }, + { + "epoch": 0.35552, + "grad_norm": 0.1263560801744461, + "learning_rate": 0.00016041237113402065, + "loss": 0.832, + "step": 1111 + }, + { + "epoch": 0.35584, + "grad_norm": 0.13208667933940887, + "learning_rate": 0.00016037113402061855, + "loss": 0.8078, + "step": 1112 + }, + { + "epoch": 0.35616, + "grad_norm": 0.11938857287168503, + "learning_rate": 0.00016032989690721648, + "loss": 0.6706, + "step": 1113 + }, + { + "epoch": 0.35648, + "grad_norm": 0.15087862312793732, + "learning_rate": 0.00016028865979381444, + "loss": 0.6342, + "step": 1114 + }, + { + "epoch": 0.3568, + "grad_norm": 0.13852839171886444, + "learning_rate": 0.00016024742268041237, + "loss": 0.7312, + "step": 1115 + }, + { + "epoch": 0.35712, + "grad_norm": 0.123874731361866, + "learning_rate": 0.00016020618556701033, + "loss": 0.5679, + "step": 1116 + }, + { + "epoch": 0.35744, + "grad_norm": 0.1398034542798996, + "learning_rate": 0.00016016494845360826, + "loss": 0.7158, + "step": 1117 + }, + { + "epoch": 0.35776, + "grad_norm": 0.15118446946144104, + "learning_rate": 0.0001601237113402062, + "loss": 0.9009, + "step": 1118 + }, + { + "epoch": 0.35808, + "grad_norm": 0.13998092710971832, + "learning_rate": 0.00016008247422680412, + "loss": 0.7229, + "step": 1119 + }, + { + "epoch": 0.3584, + "grad_norm": 0.13464230298995972, + "learning_rate": 0.00016004123711340205, + "loss": 0.6686, + "step": 1120 + }, + { + "epoch": 0.35872, + "grad_norm": 0.1247800886631012, + "learning_rate": 0.00016, + "loss": 0.6007, + "step": 1121 + }, + { + "epoch": 0.35904, + "grad_norm": 0.15494264662265778, + "learning_rate": 0.00015995876288659794, + "loss": 0.7049, + "step": 1122 + }, + { + "epoch": 0.35936, + "grad_norm": 0.13575033843517303, + "learning_rate": 0.00015991752577319588, + "loss": 0.6467, + "step": 1123 + }, + { + "epoch": 0.35968, + "grad_norm": 0.13395622372627258, + "learning_rate": 0.00015987628865979383, + "loss": 0.8035, + "step": 1124 + }, + { + "epoch": 0.36, + "grad_norm": 0.12352868914604187, + "learning_rate": 0.00015983505154639177, + "loss": 0.5654, + "step": 1125 + }, + { + "epoch": 0.36032, + "grad_norm": 0.1200244277715683, + "learning_rate": 0.0001597938144329897, + "loss": 0.6702, + "step": 1126 + }, + { + "epoch": 0.36064, + "grad_norm": 0.1397937685251236, + "learning_rate": 0.00015975257731958763, + "loss": 0.6206, + "step": 1127 + }, + { + "epoch": 0.36096, + "grad_norm": 0.14540423452854156, + "learning_rate": 0.00015971134020618556, + "loss": 0.7432, + "step": 1128 + }, + { + "epoch": 0.36128, + "grad_norm": 0.136407732963562, + "learning_rate": 0.00015967010309278352, + "loss": 0.7838, + "step": 1129 + }, + { + "epoch": 0.3616, + "grad_norm": 0.12719576060771942, + "learning_rate": 0.00015962886597938145, + "loss": 0.7853, + "step": 1130 + }, + { + "epoch": 0.36192, + "grad_norm": 0.12858954071998596, + "learning_rate": 0.0001595876288659794, + "loss": 0.6824, + "step": 1131 + }, + { + "epoch": 0.36224, + "grad_norm": 0.1230321004986763, + "learning_rate": 0.00015954639175257734, + "loss": 0.7019, + "step": 1132 + }, + { + "epoch": 0.36256, + "grad_norm": 0.15037384629249573, + "learning_rate": 0.00015950515463917527, + "loss": 0.8284, + "step": 1133 + }, + { + "epoch": 0.36288, + "grad_norm": 0.12637773156166077, + "learning_rate": 0.0001594639175257732, + "loss": 0.5629, + "step": 1134 + }, + { + "epoch": 0.3632, + "grad_norm": 0.1314142346382141, + "learning_rate": 0.00015942268041237113, + "loss": 0.8029, + "step": 1135 + }, + { + "epoch": 0.36352, + "grad_norm": 0.1253247708082199, + "learning_rate": 0.0001593814432989691, + "loss": 0.8017, + "step": 1136 + }, + { + "epoch": 0.36384, + "grad_norm": 0.14196865260601044, + "learning_rate": 0.00015934020618556702, + "loss": 0.7405, + "step": 1137 + }, + { + "epoch": 0.36416, + "grad_norm": 0.13544997572898865, + "learning_rate": 0.00015929896907216495, + "loss": 0.8474, + "step": 1138 + }, + { + "epoch": 0.36448, + "grad_norm": 0.14127661287784576, + "learning_rate": 0.0001592577319587629, + "loss": 0.7043, + "step": 1139 + }, + { + "epoch": 0.3648, + "grad_norm": 0.12157486379146576, + "learning_rate": 0.00015921649484536084, + "loss": 0.822, + "step": 1140 + }, + { + "epoch": 0.36512, + "grad_norm": 0.1410183608531952, + "learning_rate": 0.00015917525773195875, + "loss": 0.5824, + "step": 1141 + }, + { + "epoch": 0.36544, + "grad_norm": 0.12342025339603424, + "learning_rate": 0.0001591340206185567, + "loss": 0.577, + "step": 1142 + }, + { + "epoch": 0.36576, + "grad_norm": 0.15177765488624573, + "learning_rate": 0.00015909278350515464, + "loss": 0.5881, + "step": 1143 + }, + { + "epoch": 0.36608, + "grad_norm": 0.140254408121109, + "learning_rate": 0.0001590515463917526, + "loss": 0.6406, + "step": 1144 + }, + { + "epoch": 0.3664, + "grad_norm": 0.11794767528772354, + "learning_rate": 0.00015901030927835053, + "loss": 0.7347, + "step": 1145 + }, + { + "epoch": 0.36672, + "grad_norm": 0.13113269209861755, + "learning_rate": 0.00015896907216494846, + "loss": 0.8625, + "step": 1146 + }, + { + "epoch": 0.36704, + "grad_norm": 0.13670428097248077, + "learning_rate": 0.00015892783505154642, + "loss": 0.6342, + "step": 1147 + }, + { + "epoch": 0.36736, + "grad_norm": 0.12990567088127136, + "learning_rate": 0.00015888659793814432, + "loss": 0.7242, + "step": 1148 + }, + { + "epoch": 0.36768, + "grad_norm": 0.14626851677894592, + "learning_rate": 0.00015884536082474228, + "loss": 0.7929, + "step": 1149 + }, + { + "epoch": 0.368, + "grad_norm": 0.1317480057477951, + "learning_rate": 0.0001588041237113402, + "loss": 0.7436, + "step": 1150 + }, + { + "epoch": 0.36832, + "grad_norm": 0.1666252166032791, + "learning_rate": 0.00015876288659793814, + "loss": 0.6675, + "step": 1151 + }, + { + "epoch": 0.36864, + "grad_norm": 0.12659478187561035, + "learning_rate": 0.0001587216494845361, + "loss": 0.7032, + "step": 1152 + }, + { + "epoch": 0.36896, + "grad_norm": 0.12916727364063263, + "learning_rate": 0.00015868041237113403, + "loss": 0.7782, + "step": 1153 + }, + { + "epoch": 0.36928, + "grad_norm": 0.11833591759204865, + "learning_rate": 0.000158639175257732, + "loss": 0.5818, + "step": 1154 + }, + { + "epoch": 0.3696, + "grad_norm": 0.1341896802186966, + "learning_rate": 0.0001585979381443299, + "loss": 0.5695, + "step": 1155 + }, + { + "epoch": 0.36992, + "grad_norm": 0.13139894604682922, + "learning_rate": 0.00015855670103092782, + "loss": 0.7195, + "step": 1156 + }, + { + "epoch": 0.37024, + "grad_norm": 0.12866564095020294, + "learning_rate": 0.00015851546391752578, + "loss": 0.5747, + "step": 1157 + }, + { + "epoch": 0.37056, + "grad_norm": 0.13645455241203308, + "learning_rate": 0.00015847422680412371, + "loss": 0.632, + "step": 1158 + }, + { + "epoch": 0.37088, + "grad_norm": 0.12615318596363068, + "learning_rate": 0.00015843298969072167, + "loss": 0.6667, + "step": 1159 + }, + { + "epoch": 0.3712, + "grad_norm": 0.13311846554279327, + "learning_rate": 0.0001583917525773196, + "loss": 0.5789, + "step": 1160 + }, + { + "epoch": 0.37152, + "grad_norm": 0.14214034378528595, + "learning_rate": 0.00015835051546391754, + "loss": 0.7526, + "step": 1161 + }, + { + "epoch": 0.37184, + "grad_norm": 0.14578349888324738, + "learning_rate": 0.00015830927835051547, + "loss": 0.8063, + "step": 1162 + }, + { + "epoch": 0.37216, + "grad_norm": 0.14859075844287872, + "learning_rate": 0.0001582680412371134, + "loss": 0.6875, + "step": 1163 + }, + { + "epoch": 0.37248, + "grad_norm": 0.1303577870130539, + "learning_rate": 0.00015822680412371136, + "loss": 0.6363, + "step": 1164 + }, + { + "epoch": 0.3728, + "grad_norm": 0.13182777166366577, + "learning_rate": 0.0001581855670103093, + "loss": 0.7822, + "step": 1165 + }, + { + "epoch": 0.37312, + "grad_norm": 0.11544372886419296, + "learning_rate": 0.00015814432989690722, + "loss": 0.7469, + "step": 1166 + }, + { + "epoch": 0.37344, + "grad_norm": 0.12924250960350037, + "learning_rate": 0.00015810309278350518, + "loss": 0.7625, + "step": 1167 + }, + { + "epoch": 0.37376, + "grad_norm": 0.126427561044693, + "learning_rate": 0.0001580618556701031, + "loss": 0.651, + "step": 1168 + }, + { + "epoch": 0.37408, + "grad_norm": 0.13961675763130188, + "learning_rate": 0.00015802061855670104, + "loss": 0.7959, + "step": 1169 + }, + { + "epoch": 0.3744, + "grad_norm": 0.14169563353061676, + "learning_rate": 0.00015797938144329897, + "loss": 0.8699, + "step": 1170 + }, + { + "epoch": 0.37472, + "grad_norm": 0.1441062092781067, + "learning_rate": 0.0001579381443298969, + "loss": 0.7856, + "step": 1171 + }, + { + "epoch": 0.37504, + "grad_norm": 0.15582816302776337, + "learning_rate": 0.00015789690721649486, + "loss": 0.7262, + "step": 1172 + }, + { + "epoch": 0.37536, + "grad_norm": 0.12826552987098694, + "learning_rate": 0.0001578556701030928, + "loss": 0.7309, + "step": 1173 + }, + { + "epoch": 0.37568, + "grad_norm": 0.15205571055412292, + "learning_rate": 0.00015781443298969072, + "loss": 0.6164, + "step": 1174 + }, + { + "epoch": 0.376, + "grad_norm": 0.14039137959480286, + "learning_rate": 0.00015777319587628868, + "loss": 0.6441, + "step": 1175 + }, + { + "epoch": 0.37632, + "grad_norm": 0.1307058483362198, + "learning_rate": 0.0001577319587628866, + "loss": 0.7143, + "step": 1176 + }, + { + "epoch": 0.37664, + "grad_norm": 0.14123345911502838, + "learning_rate": 0.00015769072164948454, + "loss": 0.7048, + "step": 1177 + }, + { + "epoch": 0.37696, + "grad_norm": 0.13117215037345886, + "learning_rate": 0.00015764948453608248, + "loss": 0.6066, + "step": 1178 + }, + { + "epoch": 0.37728, + "grad_norm": 0.119295634329319, + "learning_rate": 0.0001576082474226804, + "loss": 0.8624, + "step": 1179 + }, + { + "epoch": 0.3776, + "grad_norm": 0.15351873636245728, + "learning_rate": 0.00015756701030927836, + "loss": 0.6634, + "step": 1180 + }, + { + "epoch": 0.37792, + "grad_norm": 0.12913480401039124, + "learning_rate": 0.0001575257731958763, + "loss": 0.7033, + "step": 1181 + }, + { + "epoch": 0.37824, + "grad_norm": 0.11295458674430847, + "learning_rate": 0.00015748453608247425, + "loss": 0.5714, + "step": 1182 + }, + { + "epoch": 0.37856, + "grad_norm": 0.16187618672847748, + "learning_rate": 0.00015744329896907219, + "loss": 0.7198, + "step": 1183 + }, + { + "epoch": 0.37888, + "grad_norm": 0.1341710090637207, + "learning_rate": 0.0001574020618556701, + "loss": 0.7831, + "step": 1184 + }, + { + "epoch": 0.3792, + "grad_norm": 0.14047707617282867, + "learning_rate": 0.00015736082474226805, + "loss": 0.8205, + "step": 1185 + }, + { + "epoch": 0.37952, + "grad_norm": 0.12923875451087952, + "learning_rate": 0.00015731958762886598, + "loss": 0.6956, + "step": 1186 + }, + { + "epoch": 0.37984, + "grad_norm": 0.13387133181095123, + "learning_rate": 0.00015727835051546394, + "loss": 0.659, + "step": 1187 + }, + { + "epoch": 0.38016, + "grad_norm": 0.11878004670143127, + "learning_rate": 0.00015723711340206187, + "loss": 0.7992, + "step": 1188 + }, + { + "epoch": 0.38048, + "grad_norm": 0.13417847454547882, + "learning_rate": 0.0001571958762886598, + "loss": 0.5762, + "step": 1189 + }, + { + "epoch": 0.3808, + "grad_norm": 0.15006715059280396, + "learning_rate": 0.00015715463917525773, + "loss": 0.6173, + "step": 1190 + }, + { + "epoch": 0.38112, + "grad_norm": 0.14036807417869568, + "learning_rate": 0.00015711340206185566, + "loss": 0.7551, + "step": 1191 + }, + { + "epoch": 0.38144, + "grad_norm": 0.11697541177272797, + "learning_rate": 0.00015707216494845362, + "loss": 0.5426, + "step": 1192 + }, + { + "epoch": 0.38176, + "grad_norm": 0.11747653037309647, + "learning_rate": 0.00015703092783505155, + "loss": 0.7528, + "step": 1193 + }, + { + "epoch": 0.38208, + "grad_norm": 0.13758531212806702, + "learning_rate": 0.00015698969072164948, + "loss": 0.6126, + "step": 1194 + }, + { + "epoch": 0.3824, + "grad_norm": 0.13478727638721466, + "learning_rate": 0.00015694845360824744, + "loss": 0.6984, + "step": 1195 + }, + { + "epoch": 0.38272, + "grad_norm": 0.12369264662265778, + "learning_rate": 0.00015690721649484537, + "loss": 0.7509, + "step": 1196 + }, + { + "epoch": 0.38304, + "grad_norm": 0.13458921015262604, + "learning_rate": 0.0001568659793814433, + "loss": 0.7056, + "step": 1197 + }, + { + "epoch": 0.38336, + "grad_norm": 0.1378922164440155, + "learning_rate": 0.00015682474226804124, + "loss": 0.7409, + "step": 1198 + }, + { + "epoch": 0.38368, + "grad_norm": 0.1378144472837448, + "learning_rate": 0.00015678350515463917, + "loss": 0.5798, + "step": 1199 + }, + { + "epoch": 0.384, + "grad_norm": 0.12796853482723236, + "learning_rate": 0.00015674226804123713, + "loss": 0.6913, + "step": 1200 + }, + { + "epoch": 0.38432, + "grad_norm": 0.14590856432914734, + "learning_rate": 0.00015670103092783506, + "loss": 0.7123, + "step": 1201 + }, + { + "epoch": 0.38464, + "grad_norm": 0.1079840138554573, + "learning_rate": 0.000156659793814433, + "loss": 0.5882, + "step": 1202 + }, + { + "epoch": 0.38496, + "grad_norm": 0.13286516070365906, + "learning_rate": 0.00015661855670103095, + "loss": 0.7692, + "step": 1203 + }, + { + "epoch": 0.38528, + "grad_norm": 0.1428433209657669, + "learning_rate": 0.00015657731958762888, + "loss": 0.838, + "step": 1204 + }, + { + "epoch": 0.3856, + "grad_norm": 0.12319666147232056, + "learning_rate": 0.0001565360824742268, + "loss": 0.846, + "step": 1205 + }, + { + "epoch": 0.38592, + "grad_norm": 0.14117276668548584, + "learning_rate": 0.00015649484536082474, + "loss": 0.7958, + "step": 1206 + }, + { + "epoch": 0.38624, + "grad_norm": 0.11911432445049286, + "learning_rate": 0.00015645360824742267, + "loss": 0.6158, + "step": 1207 + }, + { + "epoch": 0.38656, + "grad_norm": 0.13786646723747253, + "learning_rate": 0.00015641237113402063, + "loss": 0.7601, + "step": 1208 + }, + { + "epoch": 0.38688, + "grad_norm": 0.11726567149162292, + "learning_rate": 0.00015637113402061856, + "loss": 0.6906, + "step": 1209 + }, + { + "epoch": 0.3872, + "grad_norm": 0.12047921866178513, + "learning_rate": 0.00015632989690721652, + "loss": 0.638, + "step": 1210 + }, + { + "epoch": 0.38752, + "grad_norm": 0.1304006725549698, + "learning_rate": 0.00015628865979381445, + "loss": 0.7369, + "step": 1211 + }, + { + "epoch": 0.38784, + "grad_norm": 0.12691730260849, + "learning_rate": 0.00015624742268041238, + "loss": 0.747, + "step": 1212 + }, + { + "epoch": 0.38816, + "grad_norm": 0.13422106206417084, + "learning_rate": 0.00015620618556701031, + "loss": 0.7903, + "step": 1213 + }, + { + "epoch": 0.38848, + "grad_norm": 0.11382526159286499, + "learning_rate": 0.00015616494845360824, + "loss": 0.5815, + "step": 1214 + }, + { + "epoch": 0.3888, + "grad_norm": 0.14089763164520264, + "learning_rate": 0.0001561237113402062, + "loss": 0.7882, + "step": 1215 + }, + { + "epoch": 0.38912, + "grad_norm": 0.12979736924171448, + "learning_rate": 0.00015608247422680413, + "loss": 0.7013, + "step": 1216 + }, + { + "epoch": 0.38944, + "grad_norm": 0.1168995201587677, + "learning_rate": 0.00015604123711340207, + "loss": 0.6702, + "step": 1217 + }, + { + "epoch": 0.38976, + "grad_norm": 0.1315268576145172, + "learning_rate": 0.00015600000000000002, + "loss": 0.6036, + "step": 1218 + }, + { + "epoch": 0.39008, + "grad_norm": 0.130551278591156, + "learning_rate": 0.00015595876288659796, + "loss": 0.7314, + "step": 1219 + }, + { + "epoch": 0.3904, + "grad_norm": 0.11757661402225494, + "learning_rate": 0.0001559175257731959, + "loss": 0.6066, + "step": 1220 + }, + { + "epoch": 0.39072, + "grad_norm": 0.12399845570325851, + "learning_rate": 0.00015587628865979382, + "loss": 0.695, + "step": 1221 + }, + { + "epoch": 0.39104, + "grad_norm": 0.1578100472688675, + "learning_rate": 0.00015583505154639175, + "loss": 0.6838, + "step": 1222 + }, + { + "epoch": 0.39136, + "grad_norm": 0.13307176530361176, + "learning_rate": 0.0001557938144329897, + "loss": 0.7501, + "step": 1223 + }, + { + "epoch": 0.39168, + "grad_norm": 0.13976946473121643, + "learning_rate": 0.00015575257731958764, + "loss": 0.7296, + "step": 1224 + }, + { + "epoch": 0.392, + "grad_norm": 0.1478758454322815, + "learning_rate": 0.0001557113402061856, + "loss": 0.6399, + "step": 1225 + }, + { + "epoch": 0.39232, + "grad_norm": 0.1445852369070053, + "learning_rate": 0.0001556701030927835, + "loss": 0.691, + "step": 1226 + }, + { + "epoch": 0.39264, + "grad_norm": 0.1240420863032341, + "learning_rate": 0.00015562886597938143, + "loss": 0.8225, + "step": 1227 + }, + { + "epoch": 0.39296, + "grad_norm": 0.1172756627202034, + "learning_rate": 0.0001555876288659794, + "loss": 0.7198, + "step": 1228 + }, + { + "epoch": 0.39328, + "grad_norm": 0.11416526883840561, + "learning_rate": 0.00015554639175257732, + "loss": 0.6399, + "step": 1229 + }, + { + "epoch": 0.3936, + "grad_norm": 0.1195296198129654, + "learning_rate": 0.00015550515463917525, + "loss": 0.7431, + "step": 1230 + }, + { + "epoch": 0.39392, + "grad_norm": 0.13839997351169586, + "learning_rate": 0.0001554639175257732, + "loss": 0.7989, + "step": 1231 + }, + { + "epoch": 0.39424, + "grad_norm": 0.15744848549365997, + "learning_rate": 0.00015542268041237114, + "loss": 0.7423, + "step": 1232 + }, + { + "epoch": 0.39456, + "grad_norm": 0.12856073677539825, + "learning_rate": 0.00015538144329896907, + "loss": 0.5825, + "step": 1233 + }, + { + "epoch": 0.39488, + "grad_norm": 0.1339053362607956, + "learning_rate": 0.000155340206185567, + "loss": 0.7938, + "step": 1234 + }, + { + "epoch": 0.3952, + "grad_norm": 0.13555975258350372, + "learning_rate": 0.00015529896907216494, + "loss": 0.6591, + "step": 1235 + }, + { + "epoch": 0.39552, + "grad_norm": 0.14164508879184723, + "learning_rate": 0.0001552577319587629, + "loss": 0.6992, + "step": 1236 + }, + { + "epoch": 0.39584, + "grad_norm": 0.11711747199296951, + "learning_rate": 0.00015521649484536083, + "loss": 0.6513, + "step": 1237 + }, + { + "epoch": 0.39616, + "grad_norm": 0.1403762400150299, + "learning_rate": 0.00015517525773195879, + "loss": 0.7089, + "step": 1238 + }, + { + "epoch": 0.39648, + "grad_norm": 0.125601127743721, + "learning_rate": 0.00015513402061855672, + "loss": 0.5521, + "step": 1239 + }, + { + "epoch": 0.3968, + "grad_norm": 0.12626031041145325, + "learning_rate": 0.00015509278350515465, + "loss": 0.6345, + "step": 1240 + }, + { + "epoch": 0.39712, + "grad_norm": 0.1359778344631195, + "learning_rate": 0.00015505154639175258, + "loss": 0.7158, + "step": 1241 + }, + { + "epoch": 0.39744, + "grad_norm": 0.15178370475769043, + "learning_rate": 0.0001550103092783505, + "loss": 0.7438, + "step": 1242 + }, + { + "epoch": 0.39776, + "grad_norm": 0.12930850684642792, + "learning_rate": 0.00015496907216494847, + "loss": 0.8341, + "step": 1243 + }, + { + "epoch": 0.39808, + "grad_norm": 0.12345188111066818, + "learning_rate": 0.0001549278350515464, + "loss": 0.709, + "step": 1244 + }, + { + "epoch": 0.3984, + "grad_norm": 0.14064045250415802, + "learning_rate": 0.00015488659793814433, + "loss": 0.7319, + "step": 1245 + }, + { + "epoch": 0.39872, + "grad_norm": 0.1376401036977768, + "learning_rate": 0.0001548453608247423, + "loss": 0.6587, + "step": 1246 + }, + { + "epoch": 0.39904, + "grad_norm": 0.13576684892177582, + "learning_rate": 0.00015480412371134022, + "loss": 0.8445, + "step": 1247 + }, + { + "epoch": 0.39936, + "grad_norm": 0.1358659416437149, + "learning_rate": 0.00015476288659793815, + "loss": 0.8681, + "step": 1248 + }, + { + "epoch": 0.39968, + "grad_norm": 0.14173083007335663, + "learning_rate": 0.00015472164948453608, + "loss": 0.6538, + "step": 1249 + }, + { + "epoch": 0.4, + "grad_norm": 0.12762658298015594, + "learning_rate": 0.00015468041237113401, + "loss": 0.7012, + "step": 1250 + }, + { + "epoch": 0.40032, + "grad_norm": 0.14135272800922394, + "learning_rate": 0.00015463917525773197, + "loss": 0.6523, + "step": 1251 + }, + { + "epoch": 0.40064, + "grad_norm": 0.12327791005373001, + "learning_rate": 0.0001545979381443299, + "loss": 0.7231, + "step": 1252 + }, + { + "epoch": 0.40096, + "grad_norm": 0.13688601553440094, + "learning_rate": 0.00015455670103092786, + "loss": 0.6666, + "step": 1253 + }, + { + "epoch": 0.40128, + "grad_norm": 0.13409943878650665, + "learning_rate": 0.0001545154639175258, + "loss": 0.7849, + "step": 1254 + }, + { + "epoch": 0.4016, + "grad_norm": 0.14107739925384521, + "learning_rate": 0.0001544742268041237, + "loss": 0.7455, + "step": 1255 + }, + { + "epoch": 0.40192, + "grad_norm": 0.12240707129240036, + "learning_rate": 0.00015443298969072166, + "loss": 0.7664, + "step": 1256 + }, + { + "epoch": 0.40224, + "grad_norm": 0.1431652456521988, + "learning_rate": 0.0001543917525773196, + "loss": 0.7215, + "step": 1257 + }, + { + "epoch": 0.40256, + "grad_norm": 0.14196695387363434, + "learning_rate": 0.00015435051546391752, + "loss": 0.7169, + "step": 1258 + }, + { + "epoch": 0.40288, + "grad_norm": 0.15556858479976654, + "learning_rate": 0.00015430927835051548, + "loss": 0.7989, + "step": 1259 + }, + { + "epoch": 0.4032, + "grad_norm": 0.1357007473707199, + "learning_rate": 0.0001542680412371134, + "loss": 0.6919, + "step": 1260 + }, + { + "epoch": 0.40352, + "grad_norm": 0.13433754444122314, + "learning_rate": 0.00015422680412371137, + "loss": 0.7744, + "step": 1261 + }, + { + "epoch": 0.40384, + "grad_norm": 0.1415691375732422, + "learning_rate": 0.00015418556701030927, + "loss": 0.7641, + "step": 1262 + }, + { + "epoch": 0.40416, + "grad_norm": 0.15165191888809204, + "learning_rate": 0.0001541443298969072, + "loss": 0.7048, + "step": 1263 + }, + { + "epoch": 0.40448, + "grad_norm": 0.15109646320343018, + "learning_rate": 0.00015410309278350516, + "loss": 0.7522, + "step": 1264 + }, + { + "epoch": 0.4048, + "grad_norm": 0.1453443020582199, + "learning_rate": 0.0001540618556701031, + "loss": 0.7627, + "step": 1265 + }, + { + "epoch": 0.40512, + "grad_norm": 0.13811153173446655, + "learning_rate": 0.00015402061855670105, + "loss": 0.8577, + "step": 1266 + }, + { + "epoch": 0.40544, + "grad_norm": 0.11404547840356827, + "learning_rate": 0.00015397938144329898, + "loss": 0.6577, + "step": 1267 + }, + { + "epoch": 0.40576, + "grad_norm": 0.11984894424676895, + "learning_rate": 0.0001539381443298969, + "loss": 0.735, + "step": 1268 + }, + { + "epoch": 0.40608, + "grad_norm": 0.12414862215518951, + "learning_rate": 0.00015389690721649484, + "loss": 0.5764, + "step": 1269 + }, + { + "epoch": 0.4064, + "grad_norm": 0.14393122494220734, + "learning_rate": 0.00015385567010309278, + "loss": 0.7502, + "step": 1270 + }, + { + "epoch": 0.40672, + "grad_norm": 0.12545761466026306, + "learning_rate": 0.00015381443298969073, + "loss": 0.5799, + "step": 1271 + }, + { + "epoch": 0.40704, + "grad_norm": 0.1390172392129898, + "learning_rate": 0.00015377319587628867, + "loss": 0.6935, + "step": 1272 + }, + { + "epoch": 0.40736, + "grad_norm": 0.13300630450248718, + "learning_rate": 0.0001537319587628866, + "loss": 0.6794, + "step": 1273 + }, + { + "epoch": 0.40768, + "grad_norm": 0.14342691004276276, + "learning_rate": 0.00015369072164948456, + "loss": 0.7296, + "step": 1274 + }, + { + "epoch": 0.408, + "grad_norm": 0.12729567289352417, + "learning_rate": 0.00015364948453608249, + "loss": 0.6043, + "step": 1275 + }, + { + "epoch": 0.40832, + "grad_norm": 0.12945295870304108, + "learning_rate": 0.00015360824742268042, + "loss": 0.6406, + "step": 1276 + }, + { + "epoch": 0.40864, + "grad_norm": 0.13834479451179504, + "learning_rate": 0.00015356701030927835, + "loss": 0.6925, + "step": 1277 + }, + { + "epoch": 0.40896, + "grad_norm": 0.13942793011665344, + "learning_rate": 0.00015352577319587628, + "loss": 0.5954, + "step": 1278 + }, + { + "epoch": 0.40928, + "grad_norm": 0.12837499380111694, + "learning_rate": 0.00015348453608247424, + "loss": 0.7063, + "step": 1279 + }, + { + "epoch": 0.4096, + "grad_norm": 0.13395194709300995, + "learning_rate": 0.00015344329896907217, + "loss": 0.88, + "step": 1280 + }, + { + "epoch": 0.40992, + "grad_norm": 0.12082265317440033, + "learning_rate": 0.00015340206185567013, + "loss": 0.7147, + "step": 1281 + }, + { + "epoch": 0.41024, + "grad_norm": 0.12534114718437195, + "learning_rate": 0.00015336082474226806, + "loss": 0.736, + "step": 1282 + }, + { + "epoch": 0.41056, + "grad_norm": 0.12417594343423843, + "learning_rate": 0.000153319587628866, + "loss": 0.7383, + "step": 1283 + }, + { + "epoch": 0.41088, + "grad_norm": 0.1353493630886078, + "learning_rate": 0.00015327835051546392, + "loss": 0.8153, + "step": 1284 + }, + { + "epoch": 0.4112, + "grad_norm": 0.12171890586614609, + "learning_rate": 0.00015323711340206185, + "loss": 0.6863, + "step": 1285 + }, + { + "epoch": 0.41152, + "grad_norm": 0.12119521200656891, + "learning_rate": 0.0001531958762886598, + "loss": 0.7354, + "step": 1286 + }, + { + "epoch": 0.41184, + "grad_norm": 0.14262773096561432, + "learning_rate": 0.00015315463917525774, + "loss": 0.7325, + "step": 1287 + }, + { + "epoch": 0.41216, + "grad_norm": 0.14777816832065582, + "learning_rate": 0.00015311340206185567, + "loss": 0.805, + "step": 1288 + }, + { + "epoch": 0.41248, + "grad_norm": 0.15291215479373932, + "learning_rate": 0.00015307216494845363, + "loss": 0.737, + "step": 1289 + }, + { + "epoch": 0.4128, + "grad_norm": 0.13089078664779663, + "learning_rate": 0.00015303092783505156, + "loss": 0.7067, + "step": 1290 + }, + { + "epoch": 0.41312, + "grad_norm": 0.12483957409858704, + "learning_rate": 0.00015298969072164947, + "loss": 0.6353, + "step": 1291 + }, + { + "epoch": 0.41344, + "grad_norm": 0.13014909625053406, + "learning_rate": 0.00015294845360824743, + "loss": 0.8446, + "step": 1292 + }, + { + "epoch": 0.41376, + "grad_norm": 0.13972054421901703, + "learning_rate": 0.00015290721649484536, + "loss": 0.7848, + "step": 1293 + }, + { + "epoch": 0.41408, + "grad_norm": 0.14988793432712555, + "learning_rate": 0.00015286597938144332, + "loss": 0.7206, + "step": 1294 + }, + { + "epoch": 0.4144, + "grad_norm": 0.14217117428779602, + "learning_rate": 0.00015282474226804125, + "loss": 0.6913, + "step": 1295 + }, + { + "epoch": 0.41472, + "grad_norm": 0.13317078351974487, + "learning_rate": 0.00015278350515463918, + "loss": 0.6613, + "step": 1296 + }, + { + "epoch": 0.41504, + "grad_norm": 0.12724672257900238, + "learning_rate": 0.00015274226804123714, + "loss": 0.7247, + "step": 1297 + }, + { + "epoch": 0.41536, + "grad_norm": 0.1318334937095642, + "learning_rate": 0.00015270103092783504, + "loss": 0.7964, + "step": 1298 + }, + { + "epoch": 0.41568, + "grad_norm": 0.13650526106357574, + "learning_rate": 0.000152659793814433, + "loss": 0.6923, + "step": 1299 + }, + { + "epoch": 0.416, + "grad_norm": 0.1494276076555252, + "learning_rate": 0.00015261855670103093, + "loss": 0.7033, + "step": 1300 + }, + { + "epoch": 0.41632, + "grad_norm": 0.1358669549226761, + "learning_rate": 0.00015257731958762886, + "loss": 0.6855, + "step": 1301 + }, + { + "epoch": 0.41664, + "grad_norm": 0.1421387791633606, + "learning_rate": 0.00015253608247422682, + "loss": 0.7236, + "step": 1302 + }, + { + "epoch": 0.41696, + "grad_norm": 0.14789563417434692, + "learning_rate": 0.00015249484536082475, + "loss": 0.6563, + "step": 1303 + }, + { + "epoch": 0.41728, + "grad_norm": 0.1348695158958435, + "learning_rate": 0.0001524536082474227, + "loss": 0.8174, + "step": 1304 + }, + { + "epoch": 0.4176, + "grad_norm": 0.1256149411201477, + "learning_rate": 0.00015241237113402061, + "loss": 0.8209, + "step": 1305 + }, + { + "epoch": 0.41792, + "grad_norm": 0.13752906024456024, + "learning_rate": 0.00015237113402061855, + "loss": 0.6479, + "step": 1306 + }, + { + "epoch": 0.41824, + "grad_norm": 0.1335868388414383, + "learning_rate": 0.0001523298969072165, + "loss": 0.8608, + "step": 1307 + }, + { + "epoch": 0.41856, + "grad_norm": 0.1334158182144165, + "learning_rate": 0.00015228865979381444, + "loss": 0.615, + "step": 1308 + }, + { + "epoch": 0.41888, + "grad_norm": 0.1413755565881729, + "learning_rate": 0.0001522474226804124, + "loss": 0.8455, + "step": 1309 + }, + { + "epoch": 0.4192, + "grad_norm": 0.11322362720966339, + "learning_rate": 0.00015220618556701032, + "loss": 0.7054, + "step": 1310 + }, + { + "epoch": 0.41952, + "grad_norm": 0.12619735300540924, + "learning_rate": 0.00015216494845360826, + "loss": 0.7141, + "step": 1311 + }, + { + "epoch": 0.41984, + "grad_norm": 0.13447314500808716, + "learning_rate": 0.0001521237113402062, + "loss": 0.6255, + "step": 1312 + }, + { + "epoch": 0.42016, + "grad_norm": 0.12858811020851135, + "learning_rate": 0.00015208247422680412, + "loss": 0.8257, + "step": 1313 + }, + { + "epoch": 0.42048, + "grad_norm": 0.14247672259807587, + "learning_rate": 0.00015204123711340208, + "loss": 0.696, + "step": 1314 + }, + { + "epoch": 0.4208, + "grad_norm": 0.15502195060253143, + "learning_rate": 0.000152, + "loss": 0.7275, + "step": 1315 + }, + { + "epoch": 0.42112, + "grad_norm": 0.11959568411111832, + "learning_rate": 0.00015195876288659794, + "loss": 0.5196, + "step": 1316 + }, + { + "epoch": 0.42144, + "grad_norm": 0.12191260606050491, + "learning_rate": 0.0001519175257731959, + "loss": 0.6726, + "step": 1317 + }, + { + "epoch": 0.42176, + "grad_norm": 0.12593212723731995, + "learning_rate": 0.00015187628865979383, + "loss": 0.7233, + "step": 1318 + }, + { + "epoch": 0.42208, + "grad_norm": 0.13905274868011475, + "learning_rate": 0.00015183505154639176, + "loss": 0.8305, + "step": 1319 + }, + { + "epoch": 0.4224, + "grad_norm": 0.11840659379959106, + "learning_rate": 0.0001517938144329897, + "loss": 0.6331, + "step": 1320 + }, + { + "epoch": 0.42272, + "grad_norm": 0.14931786060333252, + "learning_rate": 0.00015175257731958762, + "loss": 0.7547, + "step": 1321 + }, + { + "epoch": 0.42304, + "grad_norm": 0.13363072276115417, + "learning_rate": 0.00015171134020618558, + "loss": 0.6848, + "step": 1322 + }, + { + "epoch": 0.42336, + "grad_norm": 0.13666385412216187, + "learning_rate": 0.0001516701030927835, + "loss": 0.7105, + "step": 1323 + }, + { + "epoch": 0.42368, + "grad_norm": 0.14308300614356995, + "learning_rate": 0.00015162886597938144, + "loss": 0.6972, + "step": 1324 + }, + { + "epoch": 0.424, + "grad_norm": 0.10861511528491974, + "learning_rate": 0.0001515876288659794, + "loss": 0.6213, + "step": 1325 + }, + { + "epoch": 0.42432, + "grad_norm": 0.13606682419776917, + "learning_rate": 0.00015154639175257733, + "loss": 0.789, + "step": 1326 + }, + { + "epoch": 0.42464, + "grad_norm": 0.13982504606246948, + "learning_rate": 0.00015150515463917526, + "loss": 0.6831, + "step": 1327 + }, + { + "epoch": 0.42496, + "grad_norm": 0.1425737738609314, + "learning_rate": 0.0001514639175257732, + "loss": 0.7159, + "step": 1328 + }, + { + "epoch": 0.42528, + "grad_norm": 0.1338706612586975, + "learning_rate": 0.00015142268041237113, + "loss": 0.776, + "step": 1329 + }, + { + "epoch": 0.4256, + "grad_norm": 0.13453266024589539, + "learning_rate": 0.00015138144329896909, + "loss": 0.5924, + "step": 1330 + }, + { + "epoch": 0.42592, + "grad_norm": 0.12489192187786102, + "learning_rate": 0.00015134020618556702, + "loss": 0.7241, + "step": 1331 + }, + { + "epoch": 0.42624, + "grad_norm": 0.1499992311000824, + "learning_rate": 0.00015129896907216498, + "loss": 0.7659, + "step": 1332 + }, + { + "epoch": 0.42656, + "grad_norm": 0.13693708181381226, + "learning_rate": 0.0001512577319587629, + "loss": 0.7011, + "step": 1333 + }, + { + "epoch": 0.42688, + "grad_norm": 0.1315765529870987, + "learning_rate": 0.0001512164948453608, + "loss": 0.76, + "step": 1334 + }, + { + "epoch": 0.4272, + "grad_norm": 0.161947101354599, + "learning_rate": 0.00015117525773195877, + "loss": 0.8024, + "step": 1335 + }, + { + "epoch": 0.42752, + "grad_norm": 0.14304119348526, + "learning_rate": 0.0001511340206185567, + "loss": 0.7123, + "step": 1336 + }, + { + "epoch": 0.42784, + "grad_norm": 0.11994520574808121, + "learning_rate": 0.00015109278350515466, + "loss": 0.5823, + "step": 1337 + }, + { + "epoch": 0.42816, + "grad_norm": 0.13152840733528137, + "learning_rate": 0.0001510515463917526, + "loss": 0.754, + "step": 1338 + }, + { + "epoch": 0.42848, + "grad_norm": 0.1281958967447281, + "learning_rate": 0.00015101030927835052, + "loss": 0.7402, + "step": 1339 + }, + { + "epoch": 0.4288, + "grad_norm": 0.14896529912948608, + "learning_rate": 0.00015096907216494845, + "loss": 0.7494, + "step": 1340 + }, + { + "epoch": 0.42912, + "grad_norm": 0.1480063498020172, + "learning_rate": 0.00015092783505154638, + "loss": 0.8042, + "step": 1341 + }, + { + "epoch": 0.42944, + "grad_norm": 0.11985750496387482, + "learning_rate": 0.00015088659793814434, + "loss": 0.7831, + "step": 1342 + }, + { + "epoch": 0.42976, + "grad_norm": 0.14073733985424042, + "learning_rate": 0.00015084536082474227, + "loss": 0.7365, + "step": 1343 + }, + { + "epoch": 0.43008, + "grad_norm": 0.15199826657772064, + "learning_rate": 0.0001508041237113402, + "loss": 0.6514, + "step": 1344 + }, + { + "epoch": 0.4304, + "grad_norm": 0.15114089846611023, + "learning_rate": 0.00015076288659793816, + "loss": 0.8126, + "step": 1345 + }, + { + "epoch": 0.43072, + "grad_norm": 0.13839714229106903, + "learning_rate": 0.0001507216494845361, + "loss": 0.6972, + "step": 1346 + }, + { + "epoch": 0.43104, + "grad_norm": 0.1391737163066864, + "learning_rate": 0.00015068041237113403, + "loss": 0.666, + "step": 1347 + }, + { + "epoch": 0.43136, + "grad_norm": 0.14514316618442535, + "learning_rate": 0.00015063917525773196, + "loss": 0.6214, + "step": 1348 + }, + { + "epoch": 0.43168, + "grad_norm": 0.13457876443862915, + "learning_rate": 0.0001505979381443299, + "loss": 0.6535, + "step": 1349 + }, + { + "epoch": 0.432, + "grad_norm": 0.13254746794700623, + "learning_rate": 0.00015055670103092785, + "loss": 0.7395, + "step": 1350 + }, + { + "epoch": 0.43232, + "grad_norm": 0.14473433792591095, + "learning_rate": 0.00015051546391752578, + "loss": 0.6602, + "step": 1351 + }, + { + "epoch": 0.43264, + "grad_norm": 0.13290688395500183, + "learning_rate": 0.0001504742268041237, + "loss": 0.6509, + "step": 1352 + }, + { + "epoch": 0.43296, + "grad_norm": 0.13023175299167633, + "learning_rate": 0.00015043298969072167, + "loss": 0.7091, + "step": 1353 + }, + { + "epoch": 0.43328, + "grad_norm": 0.13794860243797302, + "learning_rate": 0.0001503917525773196, + "loss": 0.5858, + "step": 1354 + }, + { + "epoch": 0.4336, + "grad_norm": 0.14477092027664185, + "learning_rate": 0.00015035051546391753, + "loss": 0.6183, + "step": 1355 + }, + { + "epoch": 0.43392, + "grad_norm": 0.12636211514472961, + "learning_rate": 0.00015030927835051546, + "loss": 0.6809, + "step": 1356 + }, + { + "epoch": 0.43424, + "grad_norm": 0.11365452408790588, + "learning_rate": 0.0001502680412371134, + "loss": 0.7062, + "step": 1357 + }, + { + "epoch": 0.43456, + "grad_norm": 0.15799300372600555, + "learning_rate": 0.00015022680412371135, + "loss": 0.6066, + "step": 1358 + }, + { + "epoch": 0.43488, + "grad_norm": 0.1377616673707962, + "learning_rate": 0.00015018556701030928, + "loss": 0.676, + "step": 1359 + }, + { + "epoch": 0.4352, + "grad_norm": 0.11081568151712418, + "learning_rate": 0.00015014432989690724, + "loss": 0.7228, + "step": 1360 + }, + { + "epoch": 0.43552, + "grad_norm": 0.13361386954784393, + "learning_rate": 0.00015010309278350517, + "loss": 0.8244, + "step": 1361 + }, + { + "epoch": 0.43584, + "grad_norm": 0.13209503889083862, + "learning_rate": 0.0001500618556701031, + "loss": 0.6072, + "step": 1362 + }, + { + "epoch": 0.43616, + "grad_norm": 0.13853539526462555, + "learning_rate": 0.00015002061855670103, + "loss": 0.7314, + "step": 1363 + }, + { + "epoch": 0.43648, + "grad_norm": 0.13965541124343872, + "learning_rate": 0.00014997938144329897, + "loss": 0.6755, + "step": 1364 + }, + { + "epoch": 0.4368, + "grad_norm": 0.13788078725337982, + "learning_rate": 0.00014993814432989692, + "loss": 0.7235, + "step": 1365 + }, + { + "epoch": 0.43712, + "grad_norm": 0.15744584798812866, + "learning_rate": 0.00014989690721649486, + "loss": 0.719, + "step": 1366 + }, + { + "epoch": 0.43744, + "grad_norm": 0.12780450284481049, + "learning_rate": 0.0001498556701030928, + "loss": 0.6638, + "step": 1367 + }, + { + "epoch": 0.43776, + "grad_norm": 0.1370459794998169, + "learning_rate": 0.00014981443298969075, + "loss": 0.6931, + "step": 1368 + }, + { + "epoch": 0.43808, + "grad_norm": 0.1236288920044899, + "learning_rate": 0.00014977319587628868, + "loss": 0.6773, + "step": 1369 + }, + { + "epoch": 0.4384, + "grad_norm": 0.14718841016292572, + "learning_rate": 0.0001497319587628866, + "loss": 0.8331, + "step": 1370 + }, + { + "epoch": 0.43872, + "grad_norm": 0.11981282383203506, + "learning_rate": 0.00014969072164948454, + "loss": 0.8594, + "step": 1371 + }, + { + "epoch": 0.43904, + "grad_norm": 0.12866224348545074, + "learning_rate": 0.00014964948453608247, + "loss": 0.6174, + "step": 1372 + }, + { + "epoch": 0.43936, + "grad_norm": 0.13502919673919678, + "learning_rate": 0.00014960824742268043, + "loss": 0.6696, + "step": 1373 + }, + { + "epoch": 0.43968, + "grad_norm": 0.1343011111021042, + "learning_rate": 0.00014956701030927836, + "loss": 0.5349, + "step": 1374 + }, + { + "epoch": 0.44, + "grad_norm": 0.12182802706956863, + "learning_rate": 0.00014952577319587632, + "loss": 0.6973, + "step": 1375 + }, + { + "epoch": 0.44032, + "grad_norm": 0.13002310693264008, + "learning_rate": 0.00014948453608247422, + "loss": 0.6637, + "step": 1376 + }, + { + "epoch": 0.44064, + "grad_norm": 0.13899263739585876, + "learning_rate": 0.00014944329896907215, + "loss": 0.721, + "step": 1377 + }, + { + "epoch": 0.44096, + "grad_norm": 0.12251676619052887, + "learning_rate": 0.0001494020618556701, + "loss": 0.6787, + "step": 1378 + }, + { + "epoch": 0.44128, + "grad_norm": 0.10930760204792023, + "learning_rate": 0.00014936082474226804, + "loss": 0.7106, + "step": 1379 + }, + { + "epoch": 0.4416, + "grad_norm": 0.1537499874830246, + "learning_rate": 0.00014931958762886597, + "loss": 0.7221, + "step": 1380 + }, + { + "epoch": 0.44192, + "grad_norm": 0.12851859629154205, + "learning_rate": 0.00014927835051546393, + "loss": 0.6125, + "step": 1381 + }, + { + "epoch": 0.44224, + "grad_norm": 0.13918665051460266, + "learning_rate": 0.00014923711340206186, + "loss": 0.7332, + "step": 1382 + }, + { + "epoch": 0.44256, + "grad_norm": 0.1295069009065628, + "learning_rate": 0.0001491958762886598, + "loss": 0.6818, + "step": 1383 + }, + { + "epoch": 0.44288, + "grad_norm": 0.11744880676269531, + "learning_rate": 0.00014915463917525773, + "loss": 0.685, + "step": 1384 + }, + { + "epoch": 0.4432, + "grad_norm": 0.12638433277606964, + "learning_rate": 0.00014911340206185566, + "loss": 0.8496, + "step": 1385 + }, + { + "epoch": 0.44352, + "grad_norm": 0.1390245407819748, + "learning_rate": 0.00014907216494845362, + "loss": 0.7578, + "step": 1386 + }, + { + "epoch": 0.44384, + "grad_norm": 0.15041057765483856, + "learning_rate": 0.00014903092783505155, + "loss": 0.7562, + "step": 1387 + }, + { + "epoch": 0.44416, + "grad_norm": 0.15098267793655396, + "learning_rate": 0.0001489896907216495, + "loss": 0.7524, + "step": 1388 + }, + { + "epoch": 0.44448, + "grad_norm": 0.12846030294895172, + "learning_rate": 0.00014894845360824744, + "loss": 0.6334, + "step": 1389 + }, + { + "epoch": 0.4448, + "grad_norm": 0.12228365242481232, + "learning_rate": 0.00014890721649484537, + "loss": 0.6569, + "step": 1390 + }, + { + "epoch": 0.44512, + "grad_norm": 0.12928150594234467, + "learning_rate": 0.0001488659793814433, + "loss": 0.7362, + "step": 1391 + }, + { + "epoch": 0.44544, + "grad_norm": 0.13998471200466156, + "learning_rate": 0.00014882474226804123, + "loss": 0.5858, + "step": 1392 + }, + { + "epoch": 0.44576, + "grad_norm": 0.1343538761138916, + "learning_rate": 0.0001487835051546392, + "loss": 0.5379, + "step": 1393 + }, + { + "epoch": 0.44608, + "grad_norm": 0.12681473791599274, + "learning_rate": 0.00014874226804123712, + "loss": 0.6388, + "step": 1394 + }, + { + "epoch": 0.4464, + "grad_norm": 0.14395521581172943, + "learning_rate": 0.00014870103092783505, + "loss": 0.767, + "step": 1395 + }, + { + "epoch": 0.44672, + "grad_norm": 0.12973885238170624, + "learning_rate": 0.000148659793814433, + "loss": 0.6566, + "step": 1396 + }, + { + "epoch": 0.44704, + "grad_norm": 0.1297558695077896, + "learning_rate": 0.00014861855670103094, + "loss": 0.6941, + "step": 1397 + }, + { + "epoch": 0.44736, + "grad_norm": 0.12988127768039703, + "learning_rate": 0.00014857731958762887, + "loss": 0.6677, + "step": 1398 + }, + { + "epoch": 0.44768, + "grad_norm": 0.1514456570148468, + "learning_rate": 0.0001485360824742268, + "loss": 0.7391, + "step": 1399 + }, + { + "epoch": 0.448, + "grad_norm": 0.1323680728673935, + "learning_rate": 0.00014849484536082474, + "loss": 0.5817, + "step": 1400 + }, + { + "epoch": 0.44832, + "grad_norm": 0.11719689518213272, + "learning_rate": 0.0001484536082474227, + "loss": 0.5905, + "step": 1401 + }, + { + "epoch": 0.44864, + "grad_norm": 0.1345493197441101, + "learning_rate": 0.00014841237113402063, + "loss": 0.6735, + "step": 1402 + }, + { + "epoch": 0.44896, + "grad_norm": 0.1417340785264969, + "learning_rate": 0.00014837113402061858, + "loss": 0.6581, + "step": 1403 + }, + { + "epoch": 0.44928, + "grad_norm": 0.13357757031917572, + "learning_rate": 0.00014832989690721652, + "loss": 0.6473, + "step": 1404 + }, + { + "epoch": 0.4496, + "grad_norm": 0.14018036425113678, + "learning_rate": 0.00014828865979381442, + "loss": 0.7394, + "step": 1405 + }, + { + "epoch": 0.44992, + "grad_norm": 0.16627104580402374, + "learning_rate": 0.00014824742268041238, + "loss": 0.8408, + "step": 1406 + }, + { + "epoch": 0.45024, + "grad_norm": 0.12548863887786865, + "learning_rate": 0.0001482061855670103, + "loss": 0.589, + "step": 1407 + }, + { + "epoch": 0.45056, + "grad_norm": 0.12361471354961395, + "learning_rate": 0.00014816494845360827, + "loss": 0.5416, + "step": 1408 + }, + { + "epoch": 0.45088, + "grad_norm": 0.15174099802970886, + "learning_rate": 0.0001481237113402062, + "loss": 0.9295, + "step": 1409 + }, + { + "epoch": 0.4512, + "grad_norm": 0.1422296017408371, + "learning_rate": 0.00014808247422680413, + "loss": 0.6572, + "step": 1410 + }, + { + "epoch": 0.45152, + "grad_norm": 0.13866426050662994, + "learning_rate": 0.0001480412371134021, + "loss": 0.6791, + "step": 1411 + }, + { + "epoch": 0.45184, + "grad_norm": 0.1270676702260971, + "learning_rate": 0.000148, + "loss": 0.7297, + "step": 1412 + }, + { + "epoch": 0.45216, + "grad_norm": 0.16468572616577148, + "learning_rate": 0.00014795876288659792, + "loss": 0.6981, + "step": 1413 + }, + { + "epoch": 0.45248, + "grad_norm": 0.1560823917388916, + "learning_rate": 0.00014791752577319588, + "loss": 0.7232, + "step": 1414 + }, + { + "epoch": 0.4528, + "grad_norm": 0.13539525866508484, + "learning_rate": 0.0001478762886597938, + "loss": 0.6393, + "step": 1415 + }, + { + "epoch": 0.45312, + "grad_norm": 0.1415485292673111, + "learning_rate": 0.00014783505154639177, + "loss": 0.7464, + "step": 1416 + }, + { + "epoch": 0.45344, + "grad_norm": 0.1506825089454651, + "learning_rate": 0.0001477938144329897, + "loss": 0.7382, + "step": 1417 + }, + { + "epoch": 0.45376, + "grad_norm": 0.11746775358915329, + "learning_rate": 0.00014775257731958763, + "loss": 0.6446, + "step": 1418 + }, + { + "epoch": 0.45408, + "grad_norm": 0.16701194643974304, + "learning_rate": 0.00014771134020618557, + "loss": 0.717, + "step": 1419 + }, + { + "epoch": 0.4544, + "grad_norm": 0.11998537927865982, + "learning_rate": 0.0001476701030927835, + "loss": 0.6464, + "step": 1420 + }, + { + "epoch": 0.45472, + "grad_norm": 0.14511455595493317, + "learning_rate": 0.00014762886597938146, + "loss": 0.6387, + "step": 1421 + }, + { + "epoch": 0.45504, + "grad_norm": 0.13467402756214142, + "learning_rate": 0.0001475876288659794, + "loss": 0.6711, + "step": 1422 + }, + { + "epoch": 0.45536, + "grad_norm": 0.1327303797006607, + "learning_rate": 0.00014754639175257732, + "loss": 0.6128, + "step": 1423 + }, + { + "epoch": 0.45568, + "grad_norm": 0.15037193894386292, + "learning_rate": 0.00014750515463917528, + "loss": 0.6816, + "step": 1424 + }, + { + "epoch": 0.456, + "grad_norm": 0.11883807182312012, + "learning_rate": 0.0001474639175257732, + "loss": 0.7379, + "step": 1425 + }, + { + "epoch": 0.45632, + "grad_norm": 0.15287265181541443, + "learning_rate": 0.00014742268041237114, + "loss": 0.7345, + "step": 1426 + }, + { + "epoch": 0.45664, + "grad_norm": 0.12675881385803223, + "learning_rate": 0.00014738144329896907, + "loss": 0.6973, + "step": 1427 + }, + { + "epoch": 0.45696, + "grad_norm": 0.13408991694450378, + "learning_rate": 0.000147340206185567, + "loss": 0.6391, + "step": 1428 + }, + { + "epoch": 0.45728, + "grad_norm": 0.14130409061908722, + "learning_rate": 0.00014729896907216496, + "loss": 0.6632, + "step": 1429 + }, + { + "epoch": 0.4576, + "grad_norm": 0.14775030314922333, + "learning_rate": 0.0001472577319587629, + "loss": 0.6253, + "step": 1430 + }, + { + "epoch": 0.45792, + "grad_norm": 0.12623794376850128, + "learning_rate": 0.00014721649484536085, + "loss": 0.809, + "step": 1431 + }, + { + "epoch": 0.45824, + "grad_norm": 0.13884440064430237, + "learning_rate": 0.00014717525773195878, + "loss": 0.7387, + "step": 1432 + }, + { + "epoch": 0.45856, + "grad_norm": 0.1509568840265274, + "learning_rate": 0.0001471340206185567, + "loss": 0.7517, + "step": 1433 + }, + { + "epoch": 0.45888, + "grad_norm": 0.13382567465305328, + "learning_rate": 0.00014709278350515464, + "loss": 0.6297, + "step": 1434 + }, + { + "epoch": 0.4592, + "grad_norm": 0.136491060256958, + "learning_rate": 0.00014705154639175257, + "loss": 0.8006, + "step": 1435 + }, + { + "epoch": 0.45952, + "grad_norm": 0.13975565135478973, + "learning_rate": 0.00014701030927835053, + "loss": 0.7377, + "step": 1436 + }, + { + "epoch": 0.45984, + "grad_norm": 0.13055750727653503, + "learning_rate": 0.00014696907216494846, + "loss": 0.7487, + "step": 1437 + }, + { + "epoch": 0.46016, + "grad_norm": 0.14289508759975433, + "learning_rate": 0.0001469278350515464, + "loss": 0.9125, + "step": 1438 + }, + { + "epoch": 0.46048, + "grad_norm": 0.12028776854276657, + "learning_rate": 0.00014688659793814435, + "loss": 0.631, + "step": 1439 + }, + { + "epoch": 0.4608, + "grad_norm": 0.16747339069843292, + "learning_rate": 0.00014684536082474228, + "loss": 0.7855, + "step": 1440 + }, + { + "epoch": 0.46112, + "grad_norm": 0.1353611946105957, + "learning_rate": 0.0001468041237113402, + "loss": 0.7545, + "step": 1441 + }, + { + "epoch": 0.46144, + "grad_norm": 0.1368974894285202, + "learning_rate": 0.00014676288659793815, + "loss": 0.8353, + "step": 1442 + }, + { + "epoch": 0.46176, + "grad_norm": 0.13990259170532227, + "learning_rate": 0.00014672164948453608, + "loss": 0.6957, + "step": 1443 + }, + { + "epoch": 0.46208, + "grad_norm": 0.15001052618026733, + "learning_rate": 0.00014668041237113404, + "loss": 0.6809, + "step": 1444 + }, + { + "epoch": 0.4624, + "grad_norm": 0.1292935013771057, + "learning_rate": 0.00014663917525773197, + "loss": 0.6982, + "step": 1445 + }, + { + "epoch": 0.46272, + "grad_norm": 0.14710412919521332, + "learning_rate": 0.0001465979381443299, + "loss": 0.5293, + "step": 1446 + }, + { + "epoch": 0.46304, + "grad_norm": 0.1477661430835724, + "learning_rate": 0.00014655670103092786, + "loss": 0.8996, + "step": 1447 + }, + { + "epoch": 0.46336, + "grad_norm": 0.12355426698923111, + "learning_rate": 0.00014651546391752576, + "loss": 0.6654, + "step": 1448 + }, + { + "epoch": 0.46368, + "grad_norm": 0.13544946908950806, + "learning_rate": 0.00014647422680412372, + "loss": 0.77, + "step": 1449 + }, + { + "epoch": 0.464, + "grad_norm": 0.13810297846794128, + "learning_rate": 0.00014643298969072165, + "loss": 0.5933, + "step": 1450 + }, + { + "epoch": 0.46432, + "grad_norm": 0.1285339742898941, + "learning_rate": 0.00014639175257731958, + "loss": 0.6522, + "step": 1451 + }, + { + "epoch": 0.46464, + "grad_norm": 0.1257028728723526, + "learning_rate": 0.00014635051546391754, + "loss": 0.581, + "step": 1452 + }, + { + "epoch": 0.46496, + "grad_norm": 0.1406828761100769, + "learning_rate": 0.00014630927835051547, + "loss": 0.6475, + "step": 1453 + }, + { + "epoch": 0.46528, + "grad_norm": 0.14316070079803467, + "learning_rate": 0.00014626804123711343, + "loss": 0.7419, + "step": 1454 + }, + { + "epoch": 0.4656, + "grad_norm": 0.14277386665344238, + "learning_rate": 0.00014622680412371134, + "loss": 0.6507, + "step": 1455 + }, + { + "epoch": 0.46592, + "grad_norm": 0.144886776804924, + "learning_rate": 0.00014618556701030927, + "loss": 0.7173, + "step": 1456 + }, + { + "epoch": 0.46624, + "grad_norm": 0.1306687295436859, + "learning_rate": 0.00014614432989690723, + "loss": 0.8279, + "step": 1457 + }, + { + "epoch": 0.46656, + "grad_norm": 0.14299385249614716, + "learning_rate": 0.00014610309278350516, + "loss": 0.6481, + "step": 1458 + }, + { + "epoch": 0.46688, + "grad_norm": 0.12328580021858215, + "learning_rate": 0.00014606185567010311, + "loss": 0.6978, + "step": 1459 + }, + { + "epoch": 0.4672, + "grad_norm": 0.13827897608280182, + "learning_rate": 0.00014602061855670105, + "loss": 0.7415, + "step": 1460 + }, + { + "epoch": 0.46752, + "grad_norm": 0.1236303448677063, + "learning_rate": 0.00014597938144329898, + "loss": 0.628, + "step": 1461 + }, + { + "epoch": 0.46784, + "grad_norm": 0.14038272202014923, + "learning_rate": 0.0001459381443298969, + "loss": 0.7499, + "step": 1462 + }, + { + "epoch": 0.46816, + "grad_norm": 0.1291581094264984, + "learning_rate": 0.00014589690721649484, + "loss": 0.7137, + "step": 1463 + }, + { + "epoch": 0.46848, + "grad_norm": 0.1379321664571762, + "learning_rate": 0.0001458556701030928, + "loss": 0.7787, + "step": 1464 + }, + { + "epoch": 0.4688, + "grad_norm": 0.13971829414367676, + "learning_rate": 0.00014581443298969073, + "loss": 0.7112, + "step": 1465 + }, + { + "epoch": 0.46912, + "grad_norm": 0.1500317007303238, + "learning_rate": 0.00014577319587628866, + "loss": 0.6516, + "step": 1466 + }, + { + "epoch": 0.46944, + "grad_norm": 0.1315118670463562, + "learning_rate": 0.00014573195876288662, + "loss": 0.793, + "step": 1467 + }, + { + "epoch": 0.46976, + "grad_norm": 0.14588972926139832, + "learning_rate": 0.00014569072164948455, + "loss": 0.7037, + "step": 1468 + }, + { + "epoch": 0.47008, + "grad_norm": 0.14539368450641632, + "learning_rate": 0.00014564948453608248, + "loss": 0.8842, + "step": 1469 + }, + { + "epoch": 0.4704, + "grad_norm": 0.12320923805236816, + "learning_rate": 0.0001456082474226804, + "loss": 0.5365, + "step": 1470 + }, + { + "epoch": 0.47072, + "grad_norm": 0.1435680389404297, + "learning_rate": 0.00014556701030927834, + "loss": 0.6666, + "step": 1471 + }, + { + "epoch": 0.47104, + "grad_norm": 0.12959571182727814, + "learning_rate": 0.0001455257731958763, + "loss": 0.7514, + "step": 1472 + }, + { + "epoch": 0.47136, + "grad_norm": 0.14037887752056122, + "learning_rate": 0.00014548453608247423, + "loss": 0.6174, + "step": 1473 + }, + { + "epoch": 0.47168, + "grad_norm": 0.13851316273212433, + "learning_rate": 0.00014544329896907217, + "loss": 0.6604, + "step": 1474 + }, + { + "epoch": 0.472, + "grad_norm": 0.13352732360363007, + "learning_rate": 0.00014540206185567012, + "loss": 0.6804, + "step": 1475 + }, + { + "epoch": 0.47232, + "grad_norm": 0.13814695179462433, + "learning_rate": 0.00014536082474226805, + "loss": 0.7677, + "step": 1476 + }, + { + "epoch": 0.47264, + "grad_norm": 0.13144215941429138, + "learning_rate": 0.00014531958762886599, + "loss": 0.715, + "step": 1477 + }, + { + "epoch": 0.47296, + "grad_norm": 0.1366271823644638, + "learning_rate": 0.00014527835051546392, + "loss": 0.7411, + "step": 1478 + }, + { + "epoch": 0.47328, + "grad_norm": 0.1347101926803589, + "learning_rate": 0.00014523711340206185, + "loss": 0.7294, + "step": 1479 + }, + { + "epoch": 0.4736, + "grad_norm": 0.13796299695968628, + "learning_rate": 0.0001451958762886598, + "loss": 0.8298, + "step": 1480 + }, + { + "epoch": 0.47392, + "grad_norm": 0.1773163378238678, + "learning_rate": 0.00014515463917525774, + "loss": 0.8373, + "step": 1481 + }, + { + "epoch": 0.47424, + "grad_norm": 0.12339209020137787, + "learning_rate": 0.0001451134020618557, + "loss": 0.7864, + "step": 1482 + }, + { + "epoch": 0.47456, + "grad_norm": 0.1335781365633011, + "learning_rate": 0.00014507216494845363, + "loss": 0.7072, + "step": 1483 + }, + { + "epoch": 0.47488, + "grad_norm": 0.12188687175512314, + "learning_rate": 0.00014503092783505153, + "loss": 0.6166, + "step": 1484 + }, + { + "epoch": 0.4752, + "grad_norm": 0.14522816240787506, + "learning_rate": 0.0001449896907216495, + "loss": 0.7699, + "step": 1485 + }, + { + "epoch": 0.47552, + "grad_norm": 0.15175694227218628, + "learning_rate": 0.00014494845360824742, + "loss": 0.8623, + "step": 1486 + }, + { + "epoch": 0.47584, + "grad_norm": 0.11388963460922241, + "learning_rate": 0.00014490721649484538, + "loss": 0.583, + "step": 1487 + }, + { + "epoch": 0.47616, + "grad_norm": 0.13264136016368866, + "learning_rate": 0.0001448659793814433, + "loss": 0.7065, + "step": 1488 + }, + { + "epoch": 0.47648, + "grad_norm": 0.15064145624637604, + "learning_rate": 0.00014482474226804124, + "loss": 0.7306, + "step": 1489 + }, + { + "epoch": 0.4768, + "grad_norm": 0.11270318925380707, + "learning_rate": 0.00014478350515463917, + "loss": 0.6129, + "step": 1490 + }, + { + "epoch": 0.47712, + "grad_norm": 0.13935883343219757, + "learning_rate": 0.0001447422680412371, + "loss": 0.603, + "step": 1491 + }, + { + "epoch": 0.47744, + "grad_norm": 0.12119437009096146, + "learning_rate": 0.00014470103092783506, + "loss": 0.6056, + "step": 1492 + }, + { + "epoch": 0.47776, + "grad_norm": 0.13266988098621368, + "learning_rate": 0.000144659793814433, + "loss": 0.6984, + "step": 1493 + }, + { + "epoch": 0.47808, + "grad_norm": 0.13279670476913452, + "learning_rate": 0.00014461855670103093, + "loss": 0.7021, + "step": 1494 + }, + { + "epoch": 0.4784, + "grad_norm": 0.12158623337745667, + "learning_rate": 0.00014457731958762888, + "loss": 0.617, + "step": 1495 + }, + { + "epoch": 0.47872, + "grad_norm": 0.14636047184467316, + "learning_rate": 0.00014453608247422682, + "loss": 0.649, + "step": 1496 + }, + { + "epoch": 0.47904, + "grad_norm": 0.13223512470722198, + "learning_rate": 0.00014449484536082475, + "loss": 0.6275, + "step": 1497 + }, + { + "epoch": 0.47936, + "grad_norm": 0.14344871044158936, + "learning_rate": 0.00014445360824742268, + "loss": 0.6604, + "step": 1498 + }, + { + "epoch": 0.47968, + "grad_norm": 0.13986854255199432, + "learning_rate": 0.0001444123711340206, + "loss": 0.8142, + "step": 1499 + }, + { + "epoch": 0.48, + "grad_norm": 0.12066833674907684, + "learning_rate": 0.00014437113402061857, + "loss": 0.6837, + "step": 1500 + }, + { + "epoch": 0.48032, + "grad_norm": 0.137162446975708, + "learning_rate": 0.0001443298969072165, + "loss": 0.5547, + "step": 1501 + }, + { + "epoch": 0.48064, + "grad_norm": 0.1468525379896164, + "learning_rate": 0.00014428865979381443, + "loss": 0.7303, + "step": 1502 + }, + { + "epoch": 0.48096, + "grad_norm": 0.1561747044324875, + "learning_rate": 0.0001442474226804124, + "loss": 0.8114, + "step": 1503 + }, + { + "epoch": 0.48128, + "grad_norm": 0.1542164832353592, + "learning_rate": 0.00014420618556701032, + "loss": 0.7206, + "step": 1504 + }, + { + "epoch": 0.4816, + "grad_norm": 0.14567720890045166, + "learning_rate": 0.00014416494845360825, + "loss": 0.7906, + "step": 1505 + }, + { + "epoch": 0.48192, + "grad_norm": 0.14530521631240845, + "learning_rate": 0.00014412371134020618, + "loss": 0.7286, + "step": 1506 + }, + { + "epoch": 0.48224, + "grad_norm": 0.13600529730319977, + "learning_rate": 0.00014408247422680411, + "loss": 0.739, + "step": 1507 + }, + { + "epoch": 0.48256, + "grad_norm": 0.13068170845508575, + "learning_rate": 0.00014404123711340207, + "loss": 0.7286, + "step": 1508 + }, + { + "epoch": 0.48288, + "grad_norm": 0.13706938922405243, + "learning_rate": 0.000144, + "loss": 0.6003, + "step": 1509 + }, + { + "epoch": 0.4832, + "grad_norm": 0.16714246571063995, + "learning_rate": 0.00014395876288659796, + "loss": 0.7568, + "step": 1510 + }, + { + "epoch": 0.48352, + "grad_norm": 0.15205715596675873, + "learning_rate": 0.0001439175257731959, + "loss": 0.714, + "step": 1511 + }, + { + "epoch": 0.48384, + "grad_norm": 0.15466946363449097, + "learning_rate": 0.00014387628865979382, + "loss": 0.7392, + "step": 1512 + }, + { + "epoch": 0.48416, + "grad_norm": 0.14179885387420654, + "learning_rate": 0.00014383505154639176, + "loss": 0.7935, + "step": 1513 + }, + { + "epoch": 0.48448, + "grad_norm": 0.13034437596797943, + "learning_rate": 0.0001437938144329897, + "loss": 0.6451, + "step": 1514 + }, + { + "epoch": 0.4848, + "grad_norm": 0.1632513552904129, + "learning_rate": 0.00014375257731958765, + "loss": 0.7264, + "step": 1515 + }, + { + "epoch": 0.48512, + "grad_norm": 0.15103459358215332, + "learning_rate": 0.00014371134020618558, + "loss": 0.6607, + "step": 1516 + }, + { + "epoch": 0.48544, + "grad_norm": 0.13666647672653198, + "learning_rate": 0.0001436701030927835, + "loss": 0.6192, + "step": 1517 + }, + { + "epoch": 0.48576, + "grad_norm": 0.14958883821964264, + "learning_rate": 0.00014362886597938147, + "loss": 0.5868, + "step": 1518 + }, + { + "epoch": 0.48608, + "grad_norm": 0.13229940831661224, + "learning_rate": 0.0001435876288659794, + "loss": 0.5931, + "step": 1519 + }, + { + "epoch": 0.4864, + "grad_norm": 0.14606302976608276, + "learning_rate": 0.00014354639175257733, + "loss": 0.6243, + "step": 1520 + }, + { + "epoch": 0.48672, + "grad_norm": 0.14299476146697998, + "learning_rate": 0.00014350515463917526, + "loss": 0.6024, + "step": 1521 + }, + { + "epoch": 0.48704, + "grad_norm": 0.1333245038986206, + "learning_rate": 0.0001434639175257732, + "loss": 0.8185, + "step": 1522 + }, + { + "epoch": 0.48736, + "grad_norm": 0.15337637066841125, + "learning_rate": 0.00014342268041237115, + "loss": 0.7985, + "step": 1523 + }, + { + "epoch": 0.48768, + "grad_norm": 0.1316799521446228, + "learning_rate": 0.00014338144329896908, + "loss": 0.8541, + "step": 1524 + }, + { + "epoch": 0.488, + "grad_norm": 0.15217626094818115, + "learning_rate": 0.00014334020618556704, + "loss": 0.7402, + "step": 1525 + }, + { + "epoch": 0.48832, + "grad_norm": 0.12334585934877396, + "learning_rate": 0.00014329896907216494, + "loss": 0.725, + "step": 1526 + }, + { + "epoch": 0.48864, + "grad_norm": 0.1464010328054428, + "learning_rate": 0.00014325773195876287, + "loss": 0.5833, + "step": 1527 + }, + { + "epoch": 0.48896, + "grad_norm": 0.12967054545879364, + "learning_rate": 0.00014321649484536083, + "loss": 0.6763, + "step": 1528 + }, + { + "epoch": 0.48928, + "grad_norm": 0.12463313341140747, + "learning_rate": 0.00014317525773195876, + "loss": 0.5501, + "step": 1529 + }, + { + "epoch": 0.4896, + "grad_norm": 0.13481323421001434, + "learning_rate": 0.00014313402061855672, + "loss": 0.807, + "step": 1530 + }, + { + "epoch": 0.48992, + "grad_norm": 0.1528305560350418, + "learning_rate": 0.00014309278350515465, + "loss": 0.6632, + "step": 1531 + }, + { + "epoch": 0.49024, + "grad_norm": 0.1464243084192276, + "learning_rate": 0.00014305154639175259, + "loss": 0.7756, + "step": 1532 + }, + { + "epoch": 0.49056, + "grad_norm": 0.1248038038611412, + "learning_rate": 0.00014301030927835052, + "loss": 0.6521, + "step": 1533 + }, + { + "epoch": 0.49088, + "grad_norm": 0.14767804741859436, + "learning_rate": 0.00014296907216494845, + "loss": 0.7785, + "step": 1534 + }, + { + "epoch": 0.4912, + "grad_norm": 0.13207073509693146, + "learning_rate": 0.00014292783505154638, + "loss": 0.8221, + "step": 1535 + }, + { + "epoch": 0.49152, + "grad_norm": 0.14973971247673035, + "learning_rate": 0.00014288659793814434, + "loss": 0.6877, + "step": 1536 + }, + { + "epoch": 0.49184, + "grad_norm": 0.13894066214561462, + "learning_rate": 0.00014284536082474227, + "loss": 0.7506, + "step": 1537 + }, + { + "epoch": 0.49216, + "grad_norm": 0.12908771634101868, + "learning_rate": 0.00014280412371134023, + "loss": 0.7926, + "step": 1538 + }, + { + "epoch": 0.49248, + "grad_norm": 0.1435110867023468, + "learning_rate": 0.00014276288659793816, + "loss": 0.6879, + "step": 1539 + }, + { + "epoch": 0.4928, + "grad_norm": 0.1450876146554947, + "learning_rate": 0.0001427216494845361, + "loss": 0.6263, + "step": 1540 + }, + { + "epoch": 0.49312, + "grad_norm": 0.1568361520767212, + "learning_rate": 0.00014268041237113402, + "loss": 0.7474, + "step": 1541 + }, + { + "epoch": 0.49344, + "grad_norm": 0.11616674065589905, + "learning_rate": 0.00014263917525773195, + "loss": 0.5327, + "step": 1542 + }, + { + "epoch": 0.49376, + "grad_norm": 0.15635275840759277, + "learning_rate": 0.0001425979381443299, + "loss": 0.7125, + "step": 1543 + }, + { + "epoch": 0.49408, + "grad_norm": 0.12766999006271362, + "learning_rate": 0.00014255670103092784, + "loss": 0.697, + "step": 1544 + }, + { + "epoch": 0.4944, + "grad_norm": 0.12405005097389221, + "learning_rate": 0.00014251546391752577, + "loss": 0.6365, + "step": 1545 + }, + { + "epoch": 0.49472, + "grad_norm": 0.1484600007534027, + "learning_rate": 0.00014247422680412373, + "loss": 0.8282, + "step": 1546 + }, + { + "epoch": 0.49504, + "grad_norm": 0.16164913773536682, + "learning_rate": 0.00014243298969072166, + "loss": 0.6262, + "step": 1547 + }, + { + "epoch": 0.49536, + "grad_norm": 0.1343853771686554, + "learning_rate": 0.0001423917525773196, + "loss": 0.8293, + "step": 1548 + }, + { + "epoch": 0.49568, + "grad_norm": 0.15468348562717438, + "learning_rate": 0.00014235051546391753, + "loss": 0.6369, + "step": 1549 + }, + { + "epoch": 0.496, + "grad_norm": 0.1245027631521225, + "learning_rate": 0.00014230927835051546, + "loss": 0.6722, + "step": 1550 + }, + { + "epoch": 0.49632, + "grad_norm": 0.1333836168050766, + "learning_rate": 0.00014226804123711342, + "loss": 0.6849, + "step": 1551 + }, + { + "epoch": 0.49664, + "grad_norm": 0.14538751542568207, + "learning_rate": 0.00014222680412371135, + "loss": 0.6602, + "step": 1552 + }, + { + "epoch": 0.49696, + "grad_norm": 0.13172303140163422, + "learning_rate": 0.0001421855670103093, + "loss": 0.6736, + "step": 1553 + }, + { + "epoch": 0.49728, + "grad_norm": 0.12844419479370117, + "learning_rate": 0.00014214432989690724, + "loss": 0.5586, + "step": 1554 + }, + { + "epoch": 0.4976, + "grad_norm": 0.14242854714393616, + "learning_rate": 0.00014210309278350517, + "loss": 0.7817, + "step": 1555 + }, + { + "epoch": 0.49792, + "grad_norm": 0.13983237743377686, + "learning_rate": 0.0001420618556701031, + "loss": 0.7779, + "step": 1556 + }, + { + "epoch": 0.49824, + "grad_norm": 0.12832964956760406, + "learning_rate": 0.00014202061855670103, + "loss": 0.8568, + "step": 1557 + }, + { + "epoch": 0.49856, + "grad_norm": 0.139461487531662, + "learning_rate": 0.000141979381443299, + "loss": 0.639, + "step": 1558 + }, + { + "epoch": 0.49888, + "grad_norm": 0.15198762714862823, + "learning_rate": 0.00014193814432989692, + "loss": 0.7881, + "step": 1559 + }, + { + "epoch": 0.4992, + "grad_norm": 0.13192367553710938, + "learning_rate": 0.00014189690721649485, + "loss": 0.7608, + "step": 1560 + }, + { + "epoch": 0.49952, + "grad_norm": 0.15548580884933472, + "learning_rate": 0.0001418556701030928, + "loss": 0.6954, + "step": 1561 + }, + { + "epoch": 0.49984, + "grad_norm": 0.1270473599433899, + "learning_rate": 0.0001418144329896907, + "loss": 0.5897, + "step": 1562 + }, + { + "epoch": 0.50016, + "grad_norm": 0.13659198582172394, + "learning_rate": 0.00014177319587628864, + "loss": 0.6254, + "step": 1563 + }, + { + "epoch": 0.50048, + "grad_norm": 0.1459890455007553, + "learning_rate": 0.0001417319587628866, + "loss": 0.6082, + "step": 1564 + }, + { + "epoch": 0.5008, + "grad_norm": 0.14214038848876953, + "learning_rate": 0.00014169072164948453, + "loss": 0.6159, + "step": 1565 + }, + { + "epoch": 0.50112, + "grad_norm": 0.15616926550865173, + "learning_rate": 0.0001416494845360825, + "loss": 0.6342, + "step": 1566 + }, + { + "epoch": 0.50144, + "grad_norm": 0.1373022496700287, + "learning_rate": 0.00014160824742268042, + "loss": 0.7758, + "step": 1567 + }, + { + "epoch": 0.50176, + "grad_norm": 0.15784426033496857, + "learning_rate": 0.00014156701030927836, + "loss": 0.9453, + "step": 1568 + }, + { + "epoch": 0.50208, + "grad_norm": 0.12950581312179565, + "learning_rate": 0.0001415257731958763, + "loss": 0.5893, + "step": 1569 + }, + { + "epoch": 0.5024, + "grad_norm": 0.14546595513820648, + "learning_rate": 0.00014148453608247422, + "loss": 0.7556, + "step": 1570 + }, + { + "epoch": 0.50272, + "grad_norm": 0.12755361199378967, + "learning_rate": 0.00014144329896907218, + "loss": 0.6448, + "step": 1571 + }, + { + "epoch": 0.50304, + "grad_norm": 0.14262261986732483, + "learning_rate": 0.0001414020618556701, + "loss": 0.6608, + "step": 1572 + }, + { + "epoch": 0.50336, + "grad_norm": 0.13705115020275116, + "learning_rate": 0.00014136082474226804, + "loss": 0.7744, + "step": 1573 + }, + { + "epoch": 0.50368, + "grad_norm": 0.14189304411411285, + "learning_rate": 0.000141319587628866, + "loss": 0.7894, + "step": 1574 + }, + { + "epoch": 0.504, + "grad_norm": 0.15143121778964996, + "learning_rate": 0.00014127835051546393, + "loss": 0.7071, + "step": 1575 + }, + { + "epoch": 0.50432, + "grad_norm": 0.13547833263874054, + "learning_rate": 0.00014123711340206186, + "loss": 0.7542, + "step": 1576 + }, + { + "epoch": 0.50464, + "grad_norm": 0.144126757979393, + "learning_rate": 0.0001411958762886598, + "loss": 0.6932, + "step": 1577 + }, + { + "epoch": 0.50496, + "grad_norm": 0.13636307418346405, + "learning_rate": 0.00014115463917525772, + "loss": 0.7421, + "step": 1578 + }, + { + "epoch": 0.50528, + "grad_norm": 0.14498503506183624, + "learning_rate": 0.00014111340206185568, + "loss": 0.7263, + "step": 1579 + }, + { + "epoch": 0.5056, + "grad_norm": 0.14120568335056305, + "learning_rate": 0.0001410721649484536, + "loss": 0.7044, + "step": 1580 + }, + { + "epoch": 0.50592, + "grad_norm": 0.13528260588645935, + "learning_rate": 0.00014103092783505157, + "loss": 0.7811, + "step": 1581 + }, + { + "epoch": 0.50624, + "grad_norm": 0.13734883069992065, + "learning_rate": 0.0001409896907216495, + "loss": 0.6135, + "step": 1582 + }, + { + "epoch": 0.50656, + "grad_norm": 0.1357855200767517, + "learning_rate": 0.00014094845360824743, + "loss": 0.5976, + "step": 1583 + }, + { + "epoch": 0.50688, + "grad_norm": 0.140086829662323, + "learning_rate": 0.00014090721649484536, + "loss": 0.6395, + "step": 1584 + }, + { + "epoch": 0.5072, + "grad_norm": 0.13380053639411926, + "learning_rate": 0.0001408659793814433, + "loss": 0.8164, + "step": 1585 + }, + { + "epoch": 0.50752, + "grad_norm": 0.13138113915920258, + "learning_rate": 0.00014082474226804125, + "loss": 0.708, + "step": 1586 + }, + { + "epoch": 0.50784, + "grad_norm": 0.14903022348880768, + "learning_rate": 0.00014078350515463919, + "loss": 0.7525, + "step": 1587 + }, + { + "epoch": 0.50816, + "grad_norm": 0.13910505175590515, + "learning_rate": 0.00014074226804123712, + "loss": 0.6733, + "step": 1588 + }, + { + "epoch": 0.50848, + "grad_norm": 0.14415191113948822, + "learning_rate": 0.00014070103092783507, + "loss": 0.6756, + "step": 1589 + }, + { + "epoch": 0.5088, + "grad_norm": 0.1318030208349228, + "learning_rate": 0.000140659793814433, + "loss": 0.6555, + "step": 1590 + }, + { + "epoch": 0.50912, + "grad_norm": 0.12761345505714417, + "learning_rate": 0.0001406185567010309, + "loss": 0.8246, + "step": 1591 + }, + { + "epoch": 0.50944, + "grad_norm": 0.1260489523410797, + "learning_rate": 0.00014057731958762887, + "loss": 0.5207, + "step": 1592 + }, + { + "epoch": 0.50976, + "grad_norm": 0.13099458813667297, + "learning_rate": 0.0001405360824742268, + "loss": 0.6913, + "step": 1593 + }, + { + "epoch": 0.51008, + "grad_norm": 0.12148265540599823, + "learning_rate": 0.00014049484536082476, + "loss": 0.661, + "step": 1594 + }, + { + "epoch": 0.5104, + "grad_norm": 0.1384635716676712, + "learning_rate": 0.0001404536082474227, + "loss": 0.7435, + "step": 1595 + }, + { + "epoch": 0.51072, + "grad_norm": 0.1455785036087036, + "learning_rate": 0.00014041237113402062, + "loss": 0.7483, + "step": 1596 + }, + { + "epoch": 0.51104, + "grad_norm": 0.14684465527534485, + "learning_rate": 0.00014037113402061858, + "loss": 0.6953, + "step": 1597 + }, + { + "epoch": 0.51136, + "grad_norm": 0.1461796909570694, + "learning_rate": 0.00014032989690721648, + "loss": 0.7893, + "step": 1598 + }, + { + "epoch": 0.51168, + "grad_norm": 0.14683541655540466, + "learning_rate": 0.00014028865979381444, + "loss": 0.7007, + "step": 1599 + }, + { + "epoch": 0.512, + "grad_norm": 0.13216625154018402, + "learning_rate": 0.00014024742268041237, + "loss": 0.6182, + "step": 1600 + }, + { + "epoch": 0.51232, + "grad_norm": 0.12355850636959076, + "learning_rate": 0.0001402061855670103, + "loss": 0.759, + "step": 1601 + }, + { + "epoch": 0.51264, + "grad_norm": 0.12836319208145142, + "learning_rate": 0.00014016494845360826, + "loss": 0.5823, + "step": 1602 + }, + { + "epoch": 0.51296, + "grad_norm": 0.15208405256271362, + "learning_rate": 0.0001401237113402062, + "loss": 0.7535, + "step": 1603 + }, + { + "epoch": 0.51328, + "grad_norm": 0.1435183584690094, + "learning_rate": 0.00014008247422680415, + "loss": 0.5679, + "step": 1604 + }, + { + "epoch": 0.5136, + "grad_norm": 0.15758828818798065, + "learning_rate": 0.00014004123711340206, + "loss": 0.5941, + "step": 1605 + }, + { + "epoch": 0.51392, + "grad_norm": 0.14006471633911133, + "learning_rate": 0.00014, + "loss": 0.6687, + "step": 1606 + }, + { + "epoch": 0.51424, + "grad_norm": 0.14834052324295044, + "learning_rate": 0.00013995876288659795, + "loss": 0.6459, + "step": 1607 + }, + { + "epoch": 0.51456, + "grad_norm": 0.13367867469787598, + "learning_rate": 0.00013991752577319588, + "loss": 0.6245, + "step": 1608 + }, + { + "epoch": 0.51488, + "grad_norm": 0.15048961341381073, + "learning_rate": 0.00013987628865979384, + "loss": 0.7576, + "step": 1609 + }, + { + "epoch": 0.5152, + "grad_norm": 0.13006894290447235, + "learning_rate": 0.00013983505154639177, + "loss": 0.835, + "step": 1610 + }, + { + "epoch": 0.51552, + "grad_norm": 0.13258709013462067, + "learning_rate": 0.0001397938144329897, + "loss": 0.5882, + "step": 1611 + }, + { + "epoch": 0.51584, + "grad_norm": 0.15616407990455627, + "learning_rate": 0.00013975257731958763, + "loss": 0.6671, + "step": 1612 + }, + { + "epoch": 0.51616, + "grad_norm": 0.11868329346179962, + "learning_rate": 0.00013971134020618556, + "loss": 0.7288, + "step": 1613 + }, + { + "epoch": 0.51648, + "grad_norm": 0.13981391489505768, + "learning_rate": 0.00013967010309278352, + "loss": 0.8886, + "step": 1614 + }, + { + "epoch": 0.5168, + "grad_norm": 0.14195965230464935, + "learning_rate": 0.00013962886597938145, + "loss": 0.7257, + "step": 1615 + }, + { + "epoch": 0.51712, + "grad_norm": 0.11883968859910965, + "learning_rate": 0.00013958762886597938, + "loss": 0.7532, + "step": 1616 + }, + { + "epoch": 0.51744, + "grad_norm": 0.12168893963098526, + "learning_rate": 0.00013954639175257734, + "loss": 0.6678, + "step": 1617 + }, + { + "epoch": 0.51776, + "grad_norm": 0.11379075050354004, + "learning_rate": 0.00013950515463917527, + "loss": 0.5969, + "step": 1618 + }, + { + "epoch": 0.51808, + "grad_norm": 0.13150529563426971, + "learning_rate": 0.0001394639175257732, + "loss": 0.7442, + "step": 1619 + }, + { + "epoch": 0.5184, + "grad_norm": 0.13873353600502014, + "learning_rate": 0.00013942268041237113, + "loss": 0.6682, + "step": 1620 + }, + { + "epoch": 0.51872, + "grad_norm": 0.12265759706497192, + "learning_rate": 0.00013938144329896907, + "loss": 0.8013, + "step": 1621 + }, + { + "epoch": 0.51904, + "grad_norm": 0.14572475850582123, + "learning_rate": 0.00013934020618556702, + "loss": 0.5613, + "step": 1622 + }, + { + "epoch": 0.51936, + "grad_norm": 0.15529032051563263, + "learning_rate": 0.00013929896907216495, + "loss": 0.7016, + "step": 1623 + }, + { + "epoch": 0.51968, + "grad_norm": 0.1380968689918518, + "learning_rate": 0.00013925773195876289, + "loss": 0.7537, + "step": 1624 + }, + { + "epoch": 0.52, + "grad_norm": 0.13644950091838837, + "learning_rate": 0.00013921649484536084, + "loss": 0.6711, + "step": 1625 + }, + { + "epoch": 0.52032, + "grad_norm": 0.1295856237411499, + "learning_rate": 0.00013917525773195878, + "loss": 0.6545, + "step": 1626 + }, + { + "epoch": 0.52064, + "grad_norm": 0.14123371243476868, + "learning_rate": 0.0001391340206185567, + "loss": 0.6589, + "step": 1627 + }, + { + "epoch": 0.52096, + "grad_norm": 0.14248934388160706, + "learning_rate": 0.00013909278350515464, + "loss": 0.7579, + "step": 1628 + }, + { + "epoch": 0.52128, + "grad_norm": 0.1358531415462494, + "learning_rate": 0.00013905154639175257, + "loss": 0.6164, + "step": 1629 + }, + { + "epoch": 0.5216, + "grad_norm": 0.12175789475440979, + "learning_rate": 0.00013901030927835053, + "loss": 0.6545, + "step": 1630 + }, + { + "epoch": 0.52192, + "grad_norm": 0.15957528352737427, + "learning_rate": 0.00013896907216494846, + "loss": 0.6837, + "step": 1631 + }, + { + "epoch": 0.52224, + "grad_norm": 0.13381953537464142, + "learning_rate": 0.00013892783505154642, + "loss": 0.6428, + "step": 1632 + }, + { + "epoch": 0.52256, + "grad_norm": 0.16512586176395416, + "learning_rate": 0.00013888659793814435, + "loss": 0.7207, + "step": 1633 + }, + { + "epoch": 0.52288, + "grad_norm": 0.11967901140451431, + "learning_rate": 0.00013884536082474225, + "loss": 0.6968, + "step": 1634 + }, + { + "epoch": 0.5232, + "grad_norm": 0.14090409874916077, + "learning_rate": 0.0001388041237113402, + "loss": 0.6865, + "step": 1635 + }, + { + "epoch": 0.52352, + "grad_norm": 0.13441434502601624, + "learning_rate": 0.00013876288659793814, + "loss": 0.8455, + "step": 1636 + }, + { + "epoch": 0.52384, + "grad_norm": 0.13941827416419983, + "learning_rate": 0.0001387216494845361, + "loss": 0.6088, + "step": 1637 + }, + { + "epoch": 0.52416, + "grad_norm": 0.13021568953990936, + "learning_rate": 0.00013868041237113403, + "loss": 0.6578, + "step": 1638 + }, + { + "epoch": 0.52448, + "grad_norm": 0.165999636054039, + "learning_rate": 0.00013863917525773196, + "loss": 0.8015, + "step": 1639 + }, + { + "epoch": 0.5248, + "grad_norm": 0.14085112512111664, + "learning_rate": 0.0001385979381443299, + "loss": 0.703, + "step": 1640 + }, + { + "epoch": 0.52512, + "grad_norm": 0.1498437523841858, + "learning_rate": 0.00013855670103092783, + "loss": 0.7359, + "step": 1641 + }, + { + "epoch": 0.52544, + "grad_norm": 0.12368453294038773, + "learning_rate": 0.00013851546391752578, + "loss": 0.7872, + "step": 1642 + }, + { + "epoch": 0.52576, + "grad_norm": 0.15656830370426178, + "learning_rate": 0.00013847422680412372, + "loss": 0.8115, + "step": 1643 + }, + { + "epoch": 0.52608, + "grad_norm": 0.15052351355552673, + "learning_rate": 0.00013843298969072165, + "loss": 0.7281, + "step": 1644 + }, + { + "epoch": 0.5264, + "grad_norm": 0.12937717139720917, + "learning_rate": 0.0001383917525773196, + "loss": 0.7406, + "step": 1645 + }, + { + "epoch": 0.52672, + "grad_norm": 0.14770230650901794, + "learning_rate": 0.00013835051546391754, + "loss": 0.7535, + "step": 1646 + }, + { + "epoch": 0.52704, + "grad_norm": 0.1457260549068451, + "learning_rate": 0.00013830927835051547, + "loss": 0.7317, + "step": 1647 + }, + { + "epoch": 0.52736, + "grad_norm": 0.14773280918598175, + "learning_rate": 0.0001382680412371134, + "loss": 0.7076, + "step": 1648 + }, + { + "epoch": 0.52768, + "grad_norm": 0.13222400844097137, + "learning_rate": 0.00013822680412371133, + "loss": 0.6722, + "step": 1649 + }, + { + "epoch": 0.528, + "grad_norm": 0.14744020998477936, + "learning_rate": 0.0001381855670103093, + "loss": 0.7032, + "step": 1650 + }, + { + "epoch": 0.52832, + "grad_norm": 0.14622749388217926, + "learning_rate": 0.00013814432989690722, + "loss": 0.7313, + "step": 1651 + }, + { + "epoch": 0.52864, + "grad_norm": 0.1350373923778534, + "learning_rate": 0.00013810309278350515, + "loss": 0.8763, + "step": 1652 + }, + { + "epoch": 0.52896, + "grad_norm": 0.14311152696609497, + "learning_rate": 0.0001380618556701031, + "loss": 0.7024, + "step": 1653 + }, + { + "epoch": 0.52928, + "grad_norm": 0.1319030225276947, + "learning_rate": 0.00013802061855670104, + "loss": 0.9381, + "step": 1654 + }, + { + "epoch": 0.5296, + "grad_norm": 0.13256032764911652, + "learning_rate": 0.00013797938144329897, + "loss": 0.73, + "step": 1655 + }, + { + "epoch": 0.52992, + "grad_norm": 0.1313982903957367, + "learning_rate": 0.0001379381443298969, + "loss": 0.7527, + "step": 1656 + }, + { + "epoch": 0.53024, + "grad_norm": 0.12875613570213318, + "learning_rate": 0.00013789690721649484, + "loss": 0.785, + "step": 1657 + }, + { + "epoch": 0.53056, + "grad_norm": 0.1500009149312973, + "learning_rate": 0.0001378556701030928, + "loss": 0.647, + "step": 1658 + }, + { + "epoch": 0.53088, + "grad_norm": 0.13786911964416504, + "learning_rate": 0.00013781443298969072, + "loss": 0.6475, + "step": 1659 + }, + { + "epoch": 0.5312, + "grad_norm": 0.1468701958656311, + "learning_rate": 0.00013777319587628868, + "loss": 0.6968, + "step": 1660 + }, + { + "epoch": 0.53152, + "grad_norm": 0.135935440659523, + "learning_rate": 0.00013773195876288661, + "loss": 0.6248, + "step": 1661 + }, + { + "epoch": 0.53184, + "grad_norm": 0.12344254553318024, + "learning_rate": 0.00013769072164948455, + "loss": 0.5669, + "step": 1662 + }, + { + "epoch": 0.53216, + "grad_norm": 0.14883849024772644, + "learning_rate": 0.00013764948453608248, + "loss": 0.631, + "step": 1663 + }, + { + "epoch": 0.53248, + "grad_norm": 0.146417036652565, + "learning_rate": 0.0001376082474226804, + "loss": 0.7066, + "step": 1664 + }, + { + "epoch": 0.5328, + "grad_norm": 0.1755172163248062, + "learning_rate": 0.00013756701030927837, + "loss": 0.7664, + "step": 1665 + }, + { + "epoch": 0.53312, + "grad_norm": 0.15486696362495422, + "learning_rate": 0.0001375257731958763, + "loss": 0.7043, + "step": 1666 + }, + { + "epoch": 0.53344, + "grad_norm": 0.13096696138381958, + "learning_rate": 0.00013748453608247423, + "loss": 0.7577, + "step": 1667 + }, + { + "epoch": 0.53376, + "grad_norm": 0.15096555650234222, + "learning_rate": 0.0001374432989690722, + "loss": 0.7308, + "step": 1668 + }, + { + "epoch": 0.53408, + "grad_norm": 0.15723296999931335, + "learning_rate": 0.00013740206185567012, + "loss": 0.7045, + "step": 1669 + }, + { + "epoch": 0.5344, + "grad_norm": 0.15512202680110931, + "learning_rate": 0.00013736082474226805, + "loss": 0.755, + "step": 1670 + }, + { + "epoch": 0.53472, + "grad_norm": 0.15073993802070618, + "learning_rate": 0.00013731958762886598, + "loss": 0.7349, + "step": 1671 + }, + { + "epoch": 0.53504, + "grad_norm": 0.1436302363872528, + "learning_rate": 0.0001372783505154639, + "loss": 0.6763, + "step": 1672 + }, + { + "epoch": 0.53536, + "grad_norm": 0.11828181892633438, + "learning_rate": 0.00013723711340206187, + "loss": 0.6416, + "step": 1673 + }, + { + "epoch": 0.53568, + "grad_norm": 0.14397715032100677, + "learning_rate": 0.0001371958762886598, + "loss": 0.6905, + "step": 1674 + }, + { + "epoch": 0.536, + "grad_norm": 0.13953359425067902, + "learning_rate": 0.00013715463917525776, + "loss": 0.6202, + "step": 1675 + }, + { + "epoch": 0.53632, + "grad_norm": 0.12717866897583008, + "learning_rate": 0.00013711340206185566, + "loss": 0.5746, + "step": 1676 + }, + { + "epoch": 0.53664, + "grad_norm": 0.1348889023065567, + "learning_rate": 0.0001370721649484536, + "loss": 0.7732, + "step": 1677 + }, + { + "epoch": 0.53696, + "grad_norm": 0.12386288493871689, + "learning_rate": 0.00013703092783505155, + "loss": 0.5508, + "step": 1678 + }, + { + "epoch": 0.53728, + "grad_norm": 0.16130952537059784, + "learning_rate": 0.00013698969072164949, + "loss": 0.8291, + "step": 1679 + }, + { + "epoch": 0.5376, + "grad_norm": 0.13158445060253143, + "learning_rate": 0.00013694845360824744, + "loss": 0.6264, + "step": 1680 + }, + { + "epoch": 0.53792, + "grad_norm": 0.1518358737230301, + "learning_rate": 0.00013690721649484538, + "loss": 0.6718, + "step": 1681 + }, + { + "epoch": 0.53824, + "grad_norm": 0.1292293816804886, + "learning_rate": 0.0001368659793814433, + "loss": 0.5921, + "step": 1682 + }, + { + "epoch": 0.53856, + "grad_norm": 0.12687307596206665, + "learning_rate": 0.00013682474226804124, + "loss": 0.6668, + "step": 1683 + }, + { + "epoch": 0.53888, + "grad_norm": 0.12536568939685822, + "learning_rate": 0.00013678350515463917, + "loss": 0.6386, + "step": 1684 + }, + { + "epoch": 0.5392, + "grad_norm": 0.13863909244537354, + "learning_rate": 0.0001367422680412371, + "loss": 0.6364, + "step": 1685 + }, + { + "epoch": 0.53952, + "grad_norm": 0.1437811255455017, + "learning_rate": 0.00013670103092783506, + "loss": 0.6716, + "step": 1686 + }, + { + "epoch": 0.53984, + "grad_norm": 0.13246455788612366, + "learning_rate": 0.000136659793814433, + "loss": 0.6327, + "step": 1687 + }, + { + "epoch": 0.54016, + "grad_norm": 0.144587904214859, + "learning_rate": 0.00013661855670103095, + "loss": 0.7246, + "step": 1688 + }, + { + "epoch": 0.54048, + "grad_norm": 0.14768695831298828, + "learning_rate": 0.00013657731958762888, + "loss": 0.6216, + "step": 1689 + }, + { + "epoch": 0.5408, + "grad_norm": 0.14124509692192078, + "learning_rate": 0.0001365360824742268, + "loss": 0.6779, + "step": 1690 + }, + { + "epoch": 0.54112, + "grad_norm": 0.14516718685626984, + "learning_rate": 0.00013649484536082474, + "loss": 0.7312, + "step": 1691 + }, + { + "epoch": 0.54144, + "grad_norm": 0.12537337839603424, + "learning_rate": 0.00013645360824742267, + "loss": 0.6352, + "step": 1692 + }, + { + "epoch": 0.54176, + "grad_norm": 0.13072200119495392, + "learning_rate": 0.00013641237113402063, + "loss": 0.8758, + "step": 1693 + }, + { + "epoch": 0.54208, + "grad_norm": 0.13488595187664032, + "learning_rate": 0.00013637113402061856, + "loss": 0.6175, + "step": 1694 + }, + { + "epoch": 0.5424, + "grad_norm": 0.12652455270290375, + "learning_rate": 0.0001363298969072165, + "loss": 0.6515, + "step": 1695 + }, + { + "epoch": 0.54272, + "grad_norm": 0.13127075135707855, + "learning_rate": 0.00013628865979381445, + "loss": 0.6905, + "step": 1696 + }, + { + "epoch": 0.54304, + "grad_norm": 0.1282002478837967, + "learning_rate": 0.00013624742268041238, + "loss": 0.8249, + "step": 1697 + }, + { + "epoch": 0.54336, + "grad_norm": 0.1588786542415619, + "learning_rate": 0.00013620618556701032, + "loss": 0.5997, + "step": 1698 + }, + { + "epoch": 0.54368, + "grad_norm": 0.16623853147029877, + "learning_rate": 0.00013616494845360825, + "loss": 0.8093, + "step": 1699 + }, + { + "epoch": 0.544, + "grad_norm": 0.15679019689559937, + "learning_rate": 0.00013612371134020618, + "loss": 0.8038, + "step": 1700 + }, + { + "epoch": 0.54432, + "grad_norm": 0.13648444414138794, + "learning_rate": 0.00013608247422680414, + "loss": 0.7195, + "step": 1701 + }, + { + "epoch": 0.54464, + "grad_norm": 0.14090919494628906, + "learning_rate": 0.00013604123711340207, + "loss": 0.7692, + "step": 1702 + }, + { + "epoch": 0.54496, + "grad_norm": 0.12612387537956238, + "learning_rate": 0.00013600000000000003, + "loss": 0.7787, + "step": 1703 + }, + { + "epoch": 0.54528, + "grad_norm": 0.13551892340183258, + "learning_rate": 0.00013595876288659796, + "loss": 0.7509, + "step": 1704 + }, + { + "epoch": 0.5456, + "grad_norm": 0.1308637112379074, + "learning_rate": 0.0001359175257731959, + "loss": 0.6905, + "step": 1705 + }, + { + "epoch": 0.54592, + "grad_norm": 0.12324418127536774, + "learning_rate": 0.00013587628865979382, + "loss": 0.7413, + "step": 1706 + }, + { + "epoch": 0.54624, + "grad_norm": 0.13526131212711334, + "learning_rate": 0.00013583505154639175, + "loss": 0.5412, + "step": 1707 + }, + { + "epoch": 0.54656, + "grad_norm": 0.14957688748836517, + "learning_rate": 0.0001357938144329897, + "loss": 0.6902, + "step": 1708 + }, + { + "epoch": 0.54688, + "grad_norm": 0.13602221012115479, + "learning_rate": 0.00013575257731958764, + "loss": 0.7267, + "step": 1709 + }, + { + "epoch": 0.5472, + "grad_norm": 0.13706456124782562, + "learning_rate": 0.00013571134020618557, + "loss": 0.6815, + "step": 1710 + }, + { + "epoch": 0.54752, + "grad_norm": 0.14902132749557495, + "learning_rate": 0.00013567010309278353, + "loss": 0.7527, + "step": 1711 + }, + { + "epoch": 0.54784, + "grad_norm": 0.1630292683839798, + "learning_rate": 0.00013562886597938143, + "loss": 0.6755, + "step": 1712 + }, + { + "epoch": 0.54816, + "grad_norm": 0.13587939739227295, + "learning_rate": 0.00013558762886597937, + "loss": 0.7811, + "step": 1713 + }, + { + "epoch": 0.54848, + "grad_norm": 0.1295965015888214, + "learning_rate": 0.00013554639175257732, + "loss": 0.5197, + "step": 1714 + }, + { + "epoch": 0.5488, + "grad_norm": 0.12837980687618256, + "learning_rate": 0.00013550515463917526, + "loss": 0.7401, + "step": 1715 + }, + { + "epoch": 0.54912, + "grad_norm": 0.1482519954442978, + "learning_rate": 0.00013546391752577321, + "loss": 0.6882, + "step": 1716 + }, + { + "epoch": 0.54944, + "grad_norm": 0.13367551565170288, + "learning_rate": 0.00013542268041237115, + "loss": 0.7258, + "step": 1717 + }, + { + "epoch": 0.54976, + "grad_norm": 0.15452930331230164, + "learning_rate": 0.00013538144329896908, + "loss": 0.6638, + "step": 1718 + }, + { + "epoch": 0.55008, + "grad_norm": 0.12944737076759338, + "learning_rate": 0.000135340206185567, + "loss": 0.7219, + "step": 1719 + }, + { + "epoch": 0.5504, + "grad_norm": 0.13899493217468262, + "learning_rate": 0.00013529896907216494, + "loss": 0.924, + "step": 1720 + }, + { + "epoch": 0.55072, + "grad_norm": 0.12524177134037018, + "learning_rate": 0.0001352577319587629, + "loss": 0.7761, + "step": 1721 + }, + { + "epoch": 0.55104, + "grad_norm": 0.12263306230306625, + "learning_rate": 0.00013521649484536083, + "loss": 0.6369, + "step": 1722 + }, + { + "epoch": 0.55136, + "grad_norm": 0.14325588941574097, + "learning_rate": 0.00013517525773195876, + "loss": 0.6871, + "step": 1723 + }, + { + "epoch": 0.55168, + "grad_norm": 0.13194425404071808, + "learning_rate": 0.00013513402061855672, + "loss": 0.8146, + "step": 1724 + }, + { + "epoch": 0.552, + "grad_norm": 0.15382345020771027, + "learning_rate": 0.00013509278350515465, + "loss": 0.7097, + "step": 1725 + }, + { + "epoch": 0.55232, + "grad_norm": 0.1529049277305603, + "learning_rate": 0.00013505154639175258, + "loss": 0.5837, + "step": 1726 + }, + { + "epoch": 0.55264, + "grad_norm": 0.1539517641067505, + "learning_rate": 0.0001350103092783505, + "loss": 0.5286, + "step": 1727 + }, + { + "epoch": 0.55296, + "grad_norm": 0.1586897224187851, + "learning_rate": 0.00013496907216494844, + "loss": 0.7999, + "step": 1728 + }, + { + "epoch": 0.55328, + "grad_norm": 0.12118642777204514, + "learning_rate": 0.0001349278350515464, + "loss": 0.5811, + "step": 1729 + }, + { + "epoch": 0.5536, + "grad_norm": 0.1392558515071869, + "learning_rate": 0.00013488659793814433, + "loss": 0.8315, + "step": 1730 + }, + { + "epoch": 0.55392, + "grad_norm": 0.13815298676490784, + "learning_rate": 0.0001348453608247423, + "loss": 0.6403, + "step": 1731 + }, + { + "epoch": 0.55424, + "grad_norm": 0.1518428474664688, + "learning_rate": 0.00013480412371134022, + "loss": 0.8195, + "step": 1732 + }, + { + "epoch": 0.55456, + "grad_norm": 0.1450730264186859, + "learning_rate": 0.00013476288659793815, + "loss": 0.641, + "step": 1733 + }, + { + "epoch": 0.55488, + "grad_norm": 0.14829540252685547, + "learning_rate": 0.00013472164948453609, + "loss": 0.7184, + "step": 1734 + }, + { + "epoch": 0.5552, + "grad_norm": 0.13482412695884705, + "learning_rate": 0.00013468041237113402, + "loss": 0.7044, + "step": 1735 + }, + { + "epoch": 0.55552, + "grad_norm": 0.1404435932636261, + "learning_rate": 0.00013463917525773197, + "loss": 0.7029, + "step": 1736 + }, + { + "epoch": 0.55584, + "grad_norm": 0.14939363300800323, + "learning_rate": 0.0001345979381443299, + "loss": 0.8864, + "step": 1737 + }, + { + "epoch": 0.55616, + "grad_norm": 0.14376312494277954, + "learning_rate": 0.00013455670103092784, + "loss": 0.8516, + "step": 1738 + }, + { + "epoch": 0.55648, + "grad_norm": 0.1442050188779831, + "learning_rate": 0.0001345154639175258, + "loss": 0.7104, + "step": 1739 + }, + { + "epoch": 0.5568, + "grad_norm": 0.12962238490581512, + "learning_rate": 0.00013447422680412373, + "loss": 0.687, + "step": 1740 + }, + { + "epoch": 0.55712, + "grad_norm": 0.13987627625465393, + "learning_rate": 0.00013443298969072166, + "loss": 0.7208, + "step": 1741 + }, + { + "epoch": 0.55744, + "grad_norm": 0.14828431606292725, + "learning_rate": 0.0001343917525773196, + "loss": 0.6677, + "step": 1742 + }, + { + "epoch": 0.55776, + "grad_norm": 0.14022362232208252, + "learning_rate": 0.00013435051546391752, + "loss": 0.6182, + "step": 1743 + }, + { + "epoch": 0.55808, + "grad_norm": 0.13057346642017365, + "learning_rate": 0.00013430927835051548, + "loss": 0.5761, + "step": 1744 + }, + { + "epoch": 0.5584, + "grad_norm": 0.1349085420370102, + "learning_rate": 0.0001342680412371134, + "loss": 0.6868, + "step": 1745 + }, + { + "epoch": 0.55872, + "grad_norm": 0.13358305394649506, + "learning_rate": 0.00013422680412371134, + "loss": 0.4868, + "step": 1746 + }, + { + "epoch": 0.55904, + "grad_norm": 0.14226961135864258, + "learning_rate": 0.0001341855670103093, + "loss": 0.737, + "step": 1747 + }, + { + "epoch": 0.55936, + "grad_norm": 0.14248816668987274, + "learning_rate": 0.0001341443298969072, + "loss": 0.6367, + "step": 1748 + }, + { + "epoch": 0.55968, + "grad_norm": 0.13468438386917114, + "learning_rate": 0.00013410309278350516, + "loss": 0.9036, + "step": 1749 + }, + { + "epoch": 0.56, + "grad_norm": 0.14425882697105408, + "learning_rate": 0.0001340618556701031, + "loss": 0.8104, + "step": 1750 + }, + { + "epoch": 0.56032, + "grad_norm": 0.16469749808311462, + "learning_rate": 0.00013402061855670103, + "loss": 0.6554, + "step": 1751 + }, + { + "epoch": 0.56064, + "grad_norm": 0.1481115072965622, + "learning_rate": 0.00013397938144329898, + "loss": 0.7364, + "step": 1752 + }, + { + "epoch": 0.56096, + "grad_norm": 0.13165169954299927, + "learning_rate": 0.00013393814432989691, + "loss": 0.6251, + "step": 1753 + }, + { + "epoch": 0.56128, + "grad_norm": 0.13377028703689575, + "learning_rate": 0.00013389690721649487, + "loss": 0.6772, + "step": 1754 + }, + { + "epoch": 0.5616, + "grad_norm": 0.12232476472854614, + "learning_rate": 0.00013385567010309278, + "loss": 0.6245, + "step": 1755 + }, + { + "epoch": 0.56192, + "grad_norm": 0.1359662115573883, + "learning_rate": 0.0001338144329896907, + "loss": 0.645, + "step": 1756 + }, + { + "epoch": 0.56224, + "grad_norm": 0.1492493748664856, + "learning_rate": 0.00013377319587628867, + "loss": 0.6599, + "step": 1757 + }, + { + "epoch": 0.56256, + "grad_norm": 0.13330449163913727, + "learning_rate": 0.0001337319587628866, + "loss": 0.6645, + "step": 1758 + }, + { + "epoch": 0.56288, + "grad_norm": 0.15756262838840485, + "learning_rate": 0.00013369072164948456, + "loss": 0.7503, + "step": 1759 + }, + { + "epoch": 0.5632, + "grad_norm": 0.13502779603004456, + "learning_rate": 0.0001336494845360825, + "loss": 0.6723, + "step": 1760 + }, + { + "epoch": 0.56352, + "grad_norm": 0.12344586849212646, + "learning_rate": 0.00013360824742268042, + "loss": 0.6976, + "step": 1761 + }, + { + "epoch": 0.56384, + "grad_norm": 0.15826153755187988, + "learning_rate": 0.00013356701030927835, + "loss": 0.7032, + "step": 1762 + }, + { + "epoch": 0.56416, + "grad_norm": 0.14459273219108582, + "learning_rate": 0.00013352577319587628, + "loss": 0.7533, + "step": 1763 + }, + { + "epoch": 0.56448, + "grad_norm": 0.13797153532505035, + "learning_rate": 0.00013348453608247424, + "loss": 0.539, + "step": 1764 + }, + { + "epoch": 0.5648, + "grad_norm": 0.14102274179458618, + "learning_rate": 0.00013344329896907217, + "loss": 0.5844, + "step": 1765 + }, + { + "epoch": 0.56512, + "grad_norm": 0.12047169357538223, + "learning_rate": 0.0001334020618556701, + "loss": 0.5146, + "step": 1766 + }, + { + "epoch": 0.56544, + "grad_norm": 0.14720189571380615, + "learning_rate": 0.00013336082474226806, + "loss": 0.7236, + "step": 1767 + }, + { + "epoch": 0.56576, + "grad_norm": 0.1129353791475296, + "learning_rate": 0.000133319587628866, + "loss": 0.6582, + "step": 1768 + }, + { + "epoch": 0.56608, + "grad_norm": 0.1277375966310501, + "learning_rate": 0.00013327835051546392, + "loss": 0.6529, + "step": 1769 + }, + { + "epoch": 0.5664, + "grad_norm": 0.13147298991680145, + "learning_rate": 0.00013323711340206186, + "loss": 0.7737, + "step": 1770 + }, + { + "epoch": 0.56672, + "grad_norm": 0.15640847384929657, + "learning_rate": 0.00013319587628865979, + "loss": 0.5673, + "step": 1771 + }, + { + "epoch": 0.56704, + "grad_norm": 0.1401795595884323, + "learning_rate": 0.00013315463917525774, + "loss": 0.6257, + "step": 1772 + }, + { + "epoch": 0.56736, + "grad_norm": 0.13787108659744263, + "learning_rate": 0.00013311340206185568, + "loss": 0.7706, + "step": 1773 + }, + { + "epoch": 0.56768, + "grad_norm": 0.13369019329547882, + "learning_rate": 0.0001330721649484536, + "loss": 0.6816, + "step": 1774 + }, + { + "epoch": 0.568, + "grad_norm": 0.14988088607788086, + "learning_rate": 0.00013303092783505157, + "loss": 0.7451, + "step": 1775 + }, + { + "epoch": 0.56832, + "grad_norm": 0.14894211292266846, + "learning_rate": 0.0001329896907216495, + "loss": 0.6831, + "step": 1776 + }, + { + "epoch": 0.56864, + "grad_norm": 0.13918204605579376, + "learning_rate": 0.00013294845360824743, + "loss": 0.7758, + "step": 1777 + }, + { + "epoch": 0.56896, + "grad_norm": 0.1445845663547516, + "learning_rate": 0.00013290721649484536, + "loss": 0.8243, + "step": 1778 + }, + { + "epoch": 0.56928, + "grad_norm": 0.13548369705677032, + "learning_rate": 0.0001328659793814433, + "loss": 0.7407, + "step": 1779 + }, + { + "epoch": 0.5696, + "grad_norm": 0.14967234432697296, + "learning_rate": 0.00013282474226804125, + "loss": 0.6652, + "step": 1780 + }, + { + "epoch": 0.56992, + "grad_norm": 0.13391603529453278, + "learning_rate": 0.00013278350515463918, + "loss": 0.6124, + "step": 1781 + }, + { + "epoch": 0.57024, + "grad_norm": 0.1333674043416977, + "learning_rate": 0.00013274226804123714, + "loss": 0.6413, + "step": 1782 + }, + { + "epoch": 0.57056, + "grad_norm": 0.1482914239168167, + "learning_rate": 0.00013270103092783507, + "loss": 0.7787, + "step": 1783 + }, + { + "epoch": 0.57088, + "grad_norm": 0.1339278519153595, + "learning_rate": 0.00013265979381443297, + "loss": 0.5781, + "step": 1784 + }, + { + "epoch": 0.5712, + "grad_norm": 0.14695081114768982, + "learning_rate": 0.00013261855670103093, + "loss": 0.7708, + "step": 1785 + }, + { + "epoch": 0.57152, + "grad_norm": 0.14863723516464233, + "learning_rate": 0.00013257731958762886, + "loss": 0.7009, + "step": 1786 + }, + { + "epoch": 0.57184, + "grad_norm": 0.14896482229232788, + "learning_rate": 0.00013253608247422682, + "loss": 0.6738, + "step": 1787 + }, + { + "epoch": 0.57216, + "grad_norm": 0.1547907441854477, + "learning_rate": 0.00013249484536082475, + "loss": 0.5639, + "step": 1788 + }, + { + "epoch": 0.57248, + "grad_norm": 0.13742683827877045, + "learning_rate": 0.00013245360824742268, + "loss": 0.6558, + "step": 1789 + }, + { + "epoch": 0.5728, + "grad_norm": 0.15084274113178253, + "learning_rate": 0.00013241237113402064, + "loss": 0.6869, + "step": 1790 + }, + { + "epoch": 0.57312, + "grad_norm": 0.14779849350452423, + "learning_rate": 0.00013237113402061855, + "loss": 0.638, + "step": 1791 + }, + { + "epoch": 0.57344, + "grad_norm": 0.12779498100280762, + "learning_rate": 0.0001323298969072165, + "loss": 0.6937, + "step": 1792 + }, + { + "epoch": 0.57376, + "grad_norm": 0.14302028715610504, + "learning_rate": 0.00013228865979381444, + "loss": 0.7148, + "step": 1793 + }, + { + "epoch": 0.57408, + "grad_norm": 0.12912467122077942, + "learning_rate": 0.00013224742268041237, + "loss": 0.8158, + "step": 1794 + }, + { + "epoch": 0.5744, + "grad_norm": 0.1584804505109787, + "learning_rate": 0.00013220618556701033, + "loss": 0.7598, + "step": 1795 + }, + { + "epoch": 0.57472, + "grad_norm": 0.14452625811100006, + "learning_rate": 0.00013216494845360826, + "loss": 0.7039, + "step": 1796 + }, + { + "epoch": 0.57504, + "grad_norm": 0.12496565282344818, + "learning_rate": 0.0001321237113402062, + "loss": 0.9407, + "step": 1797 + }, + { + "epoch": 0.57536, + "grad_norm": 0.14201682806015015, + "learning_rate": 0.00013208247422680412, + "loss": 0.5505, + "step": 1798 + }, + { + "epoch": 0.57568, + "grad_norm": 0.14024336636066437, + "learning_rate": 0.00013204123711340205, + "loss": 0.7361, + "step": 1799 + }, + { + "epoch": 0.576, + "grad_norm": 0.1336219757795334, + "learning_rate": 0.000132, + "loss": 0.7854, + "step": 1800 + }, + { + "epoch": 0.57632, + "grad_norm": 0.13467498123645782, + "learning_rate": 0.00013195876288659794, + "loss": 0.7383, + "step": 1801 + }, + { + "epoch": 0.57664, + "grad_norm": 0.1330866515636444, + "learning_rate": 0.0001319175257731959, + "loss": 0.7307, + "step": 1802 + }, + { + "epoch": 0.57696, + "grad_norm": 0.11944994330406189, + "learning_rate": 0.00013187628865979383, + "loss": 0.6408, + "step": 1803 + }, + { + "epoch": 0.57728, + "grad_norm": 0.13701461255550385, + "learning_rate": 0.00013183505154639176, + "loss": 0.8054, + "step": 1804 + }, + { + "epoch": 0.5776, + "grad_norm": 0.12145336717367172, + "learning_rate": 0.0001317938144329897, + "loss": 0.6173, + "step": 1805 + }, + { + "epoch": 0.57792, + "grad_norm": 0.13406139612197876, + "learning_rate": 0.00013175257731958762, + "loss": 0.7681, + "step": 1806 + }, + { + "epoch": 0.57824, + "grad_norm": 0.1492442935705185, + "learning_rate": 0.00013171134020618556, + "loss": 0.6116, + "step": 1807 + }, + { + "epoch": 0.57856, + "grad_norm": 0.14872552454471588, + "learning_rate": 0.00013167010309278351, + "loss": 0.7095, + "step": 1808 + }, + { + "epoch": 0.57888, + "grad_norm": 0.12427672743797302, + "learning_rate": 0.00013162886597938145, + "loss": 0.6901, + "step": 1809 + }, + { + "epoch": 0.5792, + "grad_norm": 0.13664910197257996, + "learning_rate": 0.0001315876288659794, + "loss": 0.7601, + "step": 1810 + }, + { + "epoch": 0.57952, + "grad_norm": 0.14193250238895416, + "learning_rate": 0.00013154639175257734, + "loss": 0.7525, + "step": 1811 + }, + { + "epoch": 0.57984, + "grad_norm": 0.14674946665763855, + "learning_rate": 0.00013150515463917527, + "loss": 0.6901, + "step": 1812 + }, + { + "epoch": 0.58016, + "grad_norm": 0.14705410599708557, + "learning_rate": 0.0001314639175257732, + "loss": 0.6423, + "step": 1813 + }, + { + "epoch": 0.58048, + "grad_norm": 0.14765401184558868, + "learning_rate": 0.00013142268041237113, + "loss": 0.7426, + "step": 1814 + }, + { + "epoch": 0.5808, + "grad_norm": 0.1558821201324463, + "learning_rate": 0.0001313814432989691, + "loss": 0.7017, + "step": 1815 + }, + { + "epoch": 0.58112, + "grad_norm": 0.13213512301445007, + "learning_rate": 0.00013134020618556702, + "loss": 0.8138, + "step": 1816 + }, + { + "epoch": 0.58144, + "grad_norm": 0.13489393889904022, + "learning_rate": 0.00013129896907216495, + "loss": 0.6842, + "step": 1817 + }, + { + "epoch": 0.58176, + "grad_norm": 0.1441633552312851, + "learning_rate": 0.0001312577319587629, + "loss": 0.7445, + "step": 1818 + }, + { + "epoch": 0.58208, + "grad_norm": 0.1427292674779892, + "learning_rate": 0.00013121649484536084, + "loss": 0.7351, + "step": 1819 + }, + { + "epoch": 0.5824, + "grad_norm": 0.13226602971553802, + "learning_rate": 0.00013117525773195877, + "loss": 0.7178, + "step": 1820 + }, + { + "epoch": 0.58272, + "grad_norm": 0.16712085902690887, + "learning_rate": 0.0001311340206185567, + "loss": 0.7166, + "step": 1821 + }, + { + "epoch": 0.58304, + "grad_norm": 0.13856053352355957, + "learning_rate": 0.00013109278350515463, + "loss": 0.6683, + "step": 1822 + }, + { + "epoch": 0.58336, + "grad_norm": 0.16740012168884277, + "learning_rate": 0.0001310515463917526, + "loss": 0.8369, + "step": 1823 + }, + { + "epoch": 0.58368, + "grad_norm": 0.1308383196592331, + "learning_rate": 0.00013101030927835052, + "loss": 0.5037, + "step": 1824 + }, + { + "epoch": 0.584, + "grad_norm": 0.15628789365291595, + "learning_rate": 0.00013096907216494848, + "loss": 0.7379, + "step": 1825 + }, + { + "epoch": 0.58432, + "grad_norm": 0.1515168994665146, + "learning_rate": 0.00013092783505154639, + "loss": 0.6811, + "step": 1826 + }, + { + "epoch": 0.58464, + "grad_norm": 0.1565101146697998, + "learning_rate": 0.00013088659793814432, + "loss": 0.6737, + "step": 1827 + }, + { + "epoch": 0.58496, + "grad_norm": 0.15891097486019135, + "learning_rate": 0.00013084536082474228, + "loss": 0.7704, + "step": 1828 + }, + { + "epoch": 0.58528, + "grad_norm": 0.1445811688899994, + "learning_rate": 0.0001308041237113402, + "loss": 0.7628, + "step": 1829 + }, + { + "epoch": 0.5856, + "grad_norm": 0.12324066460132599, + "learning_rate": 0.00013076288659793817, + "loss": 0.5743, + "step": 1830 + }, + { + "epoch": 0.58592, + "grad_norm": 0.15465180575847626, + "learning_rate": 0.0001307216494845361, + "loss": 0.783, + "step": 1831 + }, + { + "epoch": 0.58624, + "grad_norm": 0.14205990731716156, + "learning_rate": 0.00013068041237113403, + "loss": 0.7149, + "step": 1832 + }, + { + "epoch": 0.58656, + "grad_norm": 0.14440414309501648, + "learning_rate": 0.00013063917525773196, + "loss": 0.6951, + "step": 1833 + }, + { + "epoch": 0.58688, + "grad_norm": 0.16611546277999878, + "learning_rate": 0.0001305979381443299, + "loss": 0.9304, + "step": 1834 + }, + { + "epoch": 0.5872, + "grad_norm": 0.13717123866081238, + "learning_rate": 0.00013055670103092782, + "loss": 0.5108, + "step": 1835 + }, + { + "epoch": 0.58752, + "grad_norm": 0.1380356401205063, + "learning_rate": 0.00013051546391752578, + "loss": 0.6698, + "step": 1836 + }, + { + "epoch": 0.58784, + "grad_norm": 0.1463797241449356, + "learning_rate": 0.0001304742268041237, + "loss": 0.7732, + "step": 1837 + }, + { + "epoch": 0.58816, + "grad_norm": 0.13822247087955475, + "learning_rate": 0.00013043298969072167, + "loss": 0.7733, + "step": 1838 + }, + { + "epoch": 0.58848, + "grad_norm": 0.12503483891487122, + "learning_rate": 0.0001303917525773196, + "loss": 0.7661, + "step": 1839 + }, + { + "epoch": 0.5888, + "grad_norm": 0.1147647500038147, + "learning_rate": 0.00013035051546391753, + "loss": 0.4546, + "step": 1840 + }, + { + "epoch": 0.58912, + "grad_norm": 0.1230325773358345, + "learning_rate": 0.00013030927835051546, + "loss": 0.5585, + "step": 1841 + }, + { + "epoch": 0.58944, + "grad_norm": 0.14122216403484344, + "learning_rate": 0.0001302680412371134, + "loss": 0.5758, + "step": 1842 + }, + { + "epoch": 0.58976, + "grad_norm": 0.15151569247245789, + "learning_rate": 0.00013022680412371135, + "loss": 0.6843, + "step": 1843 + }, + { + "epoch": 0.59008, + "grad_norm": 0.1292106658220291, + "learning_rate": 0.00013018556701030928, + "loss": 0.699, + "step": 1844 + }, + { + "epoch": 0.5904, + "grad_norm": 0.13869857788085938, + "learning_rate": 0.00013014432989690722, + "loss": 0.7525, + "step": 1845 + }, + { + "epoch": 0.59072, + "grad_norm": 0.15281198918819427, + "learning_rate": 0.00013010309278350517, + "loss": 0.7205, + "step": 1846 + }, + { + "epoch": 0.59104, + "grad_norm": 0.13255898654460907, + "learning_rate": 0.0001300618556701031, + "loss": 0.5767, + "step": 1847 + }, + { + "epoch": 0.59136, + "grad_norm": 0.11870738863945007, + "learning_rate": 0.00013002061855670104, + "loss": 0.5077, + "step": 1848 + }, + { + "epoch": 0.59168, + "grad_norm": 0.14117857813835144, + "learning_rate": 0.00012997938144329897, + "loss": 0.568, + "step": 1849 + }, + { + "epoch": 0.592, + "grad_norm": 0.14479069411754608, + "learning_rate": 0.0001299381443298969, + "loss": 0.5165, + "step": 1850 + }, + { + "epoch": 0.59232, + "grad_norm": 0.14627261459827423, + "learning_rate": 0.00012989690721649486, + "loss": 0.6057, + "step": 1851 + }, + { + "epoch": 0.59264, + "grad_norm": 0.13284321129322052, + "learning_rate": 0.0001298556701030928, + "loss": 0.8567, + "step": 1852 + }, + { + "epoch": 0.59296, + "grad_norm": 0.15283948183059692, + "learning_rate": 0.00012981443298969075, + "loss": 0.7619, + "step": 1853 + }, + { + "epoch": 0.59328, + "grad_norm": 0.13968519866466522, + "learning_rate": 0.00012977319587628868, + "loss": 0.7535, + "step": 1854 + }, + { + "epoch": 0.5936, + "grad_norm": 0.1302313208580017, + "learning_rate": 0.0001297319587628866, + "loss": 0.6574, + "step": 1855 + }, + { + "epoch": 0.59392, + "grad_norm": 0.13157455623149872, + "learning_rate": 0.00012969072164948454, + "loss": 0.704, + "step": 1856 + }, + { + "epoch": 0.59424, + "grad_norm": 0.15155114233493805, + "learning_rate": 0.00012964948453608247, + "loss": 0.7422, + "step": 1857 + }, + { + "epoch": 0.59456, + "grad_norm": 0.13552071154117584, + "learning_rate": 0.00012960824742268043, + "loss": 0.5676, + "step": 1858 + }, + { + "epoch": 0.59488, + "grad_norm": 0.149967223405838, + "learning_rate": 0.00012956701030927836, + "loss": 0.6742, + "step": 1859 + }, + { + "epoch": 0.5952, + "grad_norm": 0.15703719854354858, + "learning_rate": 0.0001295257731958763, + "loss": 0.7453, + "step": 1860 + }, + { + "epoch": 0.59552, + "grad_norm": 0.1725168377161026, + "learning_rate": 0.00012948453608247425, + "loss": 0.8449, + "step": 1861 + }, + { + "epoch": 0.59584, + "grad_norm": 0.13105261325836182, + "learning_rate": 0.00012944329896907216, + "loss": 0.767, + "step": 1862 + }, + { + "epoch": 0.59616, + "grad_norm": 0.1367502510547638, + "learning_rate": 0.00012940206185567011, + "loss": 0.5927, + "step": 1863 + }, + { + "epoch": 0.59648, + "grad_norm": 0.18537381291389465, + "learning_rate": 0.00012936082474226805, + "loss": 0.756, + "step": 1864 + }, + { + "epoch": 0.5968, + "grad_norm": 0.14291930198669434, + "learning_rate": 0.00012931958762886598, + "loss": 0.7345, + "step": 1865 + }, + { + "epoch": 0.59712, + "grad_norm": 0.14345696568489075, + "learning_rate": 0.00012927835051546393, + "loss": 0.7558, + "step": 1866 + }, + { + "epoch": 0.59744, + "grad_norm": 0.13259442150592804, + "learning_rate": 0.00012923711340206187, + "loss": 0.6752, + "step": 1867 + }, + { + "epoch": 0.59776, + "grad_norm": 0.12619280815124512, + "learning_rate": 0.0001291958762886598, + "loss": 0.6396, + "step": 1868 + }, + { + "epoch": 0.59808, + "grad_norm": 0.145010843873024, + "learning_rate": 0.00012915463917525773, + "loss": 0.7933, + "step": 1869 + }, + { + "epoch": 0.5984, + "grad_norm": 0.14956963062286377, + "learning_rate": 0.00012911340206185566, + "loss": 0.7931, + "step": 1870 + }, + { + "epoch": 0.59872, + "grad_norm": 0.1342199295759201, + "learning_rate": 0.00012907216494845362, + "loss": 0.6408, + "step": 1871 + }, + { + "epoch": 0.59904, + "grad_norm": 0.1416260153055191, + "learning_rate": 0.00012903092783505155, + "loss": 0.6813, + "step": 1872 + }, + { + "epoch": 0.59936, + "grad_norm": 0.1454564481973648, + "learning_rate": 0.00012898969072164948, + "loss": 0.5713, + "step": 1873 + }, + { + "epoch": 0.59968, + "grad_norm": 0.13765162229537964, + "learning_rate": 0.00012894845360824744, + "loss": 0.5956, + "step": 1874 + }, + { + "epoch": 0.6, + "grad_norm": 0.13723546266555786, + "learning_rate": 0.00012890721649484537, + "loss": 0.7485, + "step": 1875 + }, + { + "epoch": 0.60032, + "grad_norm": 0.12591178715229034, + "learning_rate": 0.0001288659793814433, + "loss": 0.6953, + "step": 1876 + }, + { + "epoch": 0.60064, + "grad_norm": 0.12784048914909363, + "learning_rate": 0.00012882474226804123, + "loss": 0.7045, + "step": 1877 + }, + { + "epoch": 0.60096, + "grad_norm": 0.13309244811534882, + "learning_rate": 0.00012878350515463916, + "loss": 0.5357, + "step": 1878 + }, + { + "epoch": 0.60128, + "grad_norm": 0.12655137479305267, + "learning_rate": 0.00012874226804123712, + "loss": 0.6734, + "step": 1879 + }, + { + "epoch": 0.6016, + "grad_norm": 0.14110948145389557, + "learning_rate": 0.00012870103092783505, + "loss": 0.6368, + "step": 1880 + }, + { + "epoch": 0.60192, + "grad_norm": 0.12137459218502045, + "learning_rate": 0.000128659793814433, + "loss": 0.6156, + "step": 1881 + }, + { + "epoch": 0.60224, + "grad_norm": 0.15284189581871033, + "learning_rate": 0.00012861855670103094, + "loss": 0.6539, + "step": 1882 + }, + { + "epoch": 0.60256, + "grad_norm": 0.1448545604944229, + "learning_rate": 0.00012857731958762888, + "loss": 0.5931, + "step": 1883 + }, + { + "epoch": 0.60288, + "grad_norm": 0.1656656712293625, + "learning_rate": 0.0001285360824742268, + "loss": 0.6722, + "step": 1884 + }, + { + "epoch": 0.6032, + "grad_norm": 0.15540426969528198, + "learning_rate": 0.00012849484536082474, + "loss": 0.6907, + "step": 1885 + }, + { + "epoch": 0.60352, + "grad_norm": 0.13939699530601501, + "learning_rate": 0.0001284536082474227, + "loss": 0.7573, + "step": 1886 + }, + { + "epoch": 0.60384, + "grad_norm": 0.14229856431484222, + "learning_rate": 0.00012841237113402063, + "loss": 0.6349, + "step": 1887 + }, + { + "epoch": 0.60416, + "grad_norm": 0.13974221050739288, + "learning_rate": 0.00012837113402061856, + "loss": 0.8005, + "step": 1888 + }, + { + "epoch": 0.60448, + "grad_norm": 0.14249245822429657, + "learning_rate": 0.00012832989690721652, + "loss": 0.6809, + "step": 1889 + }, + { + "epoch": 0.6048, + "grad_norm": 0.1409933716058731, + "learning_rate": 0.00012828865979381445, + "loss": 0.7416, + "step": 1890 + }, + { + "epoch": 0.60512, + "grad_norm": 0.1686551868915558, + "learning_rate": 0.00012824742268041238, + "loss": 0.749, + "step": 1891 + }, + { + "epoch": 0.60544, + "grad_norm": 0.12405146658420563, + "learning_rate": 0.0001282061855670103, + "loss": 0.7878, + "step": 1892 + }, + { + "epoch": 0.60576, + "grad_norm": 0.13999059796333313, + "learning_rate": 0.00012816494845360824, + "loss": 0.5989, + "step": 1893 + }, + { + "epoch": 0.60608, + "grad_norm": 0.1495513916015625, + "learning_rate": 0.0001281237113402062, + "loss": 0.6428, + "step": 1894 + }, + { + "epoch": 0.6064, + "grad_norm": 0.13758565485477448, + "learning_rate": 0.00012808247422680413, + "loss": 0.7109, + "step": 1895 + }, + { + "epoch": 0.60672, + "grad_norm": 0.14844781160354614, + "learning_rate": 0.00012804123711340206, + "loss": 0.6306, + "step": 1896 + }, + { + "epoch": 0.60704, + "grad_norm": 0.13203153014183044, + "learning_rate": 0.00012800000000000002, + "loss": 0.7102, + "step": 1897 + }, + { + "epoch": 0.60736, + "grad_norm": 0.14242084324359894, + "learning_rate": 0.00012795876288659793, + "loss": 0.8308, + "step": 1898 + }, + { + "epoch": 0.60768, + "grad_norm": 0.12235117703676224, + "learning_rate": 0.00012791752577319588, + "loss": 0.6965, + "step": 1899 + }, + { + "epoch": 0.608, + "grad_norm": 0.14560118317604065, + "learning_rate": 0.00012787628865979382, + "loss": 0.6338, + "step": 1900 + }, + { + "epoch": 0.60832, + "grad_norm": 0.14123523235321045, + "learning_rate": 0.00012783505154639175, + "loss": 0.7186, + "step": 1901 + }, + { + "epoch": 0.60864, + "grad_norm": 0.1220543310046196, + "learning_rate": 0.0001277938144329897, + "loss": 0.5828, + "step": 1902 + }, + { + "epoch": 0.60896, + "grad_norm": 0.1339149922132492, + "learning_rate": 0.00012775257731958764, + "loss": 0.6722, + "step": 1903 + }, + { + "epoch": 0.60928, + "grad_norm": 0.1346278339624405, + "learning_rate": 0.0001277113402061856, + "loss": 0.8396, + "step": 1904 + }, + { + "epoch": 0.6096, + "grad_norm": 0.13202722370624542, + "learning_rate": 0.0001276701030927835, + "loss": 0.9066, + "step": 1905 + }, + { + "epoch": 0.60992, + "grad_norm": 0.1463432013988495, + "learning_rate": 0.00012762886597938143, + "loss": 0.7598, + "step": 1906 + }, + { + "epoch": 0.61024, + "grad_norm": 0.14574944972991943, + "learning_rate": 0.0001275876288659794, + "loss": 0.621, + "step": 1907 + }, + { + "epoch": 0.61056, + "grad_norm": 0.14618104696273804, + "learning_rate": 0.00012754639175257732, + "loss": 0.5827, + "step": 1908 + }, + { + "epoch": 0.61088, + "grad_norm": 0.14066514372825623, + "learning_rate": 0.00012750515463917528, + "loss": 0.5984, + "step": 1909 + }, + { + "epoch": 0.6112, + "grad_norm": 0.14523790776729584, + "learning_rate": 0.0001274639175257732, + "loss": 0.6005, + "step": 1910 + }, + { + "epoch": 0.61152, + "grad_norm": 0.14804573357105255, + "learning_rate": 0.00012742268041237114, + "loss": 0.6568, + "step": 1911 + }, + { + "epoch": 0.61184, + "grad_norm": 0.1466348022222519, + "learning_rate": 0.00012738144329896907, + "loss": 0.8652, + "step": 1912 + }, + { + "epoch": 0.61216, + "grad_norm": 0.14785926043987274, + "learning_rate": 0.000127340206185567, + "loss": 0.6374, + "step": 1913 + }, + { + "epoch": 0.61248, + "grad_norm": 0.13066139817237854, + "learning_rate": 0.00012729896907216496, + "loss": 0.5485, + "step": 1914 + }, + { + "epoch": 0.6128, + "grad_norm": 0.14120493829250336, + "learning_rate": 0.0001272577319587629, + "loss": 0.7202, + "step": 1915 + }, + { + "epoch": 0.61312, + "grad_norm": 0.16122666001319885, + "learning_rate": 0.00012721649484536082, + "loss": 0.7973, + "step": 1916 + }, + { + "epoch": 0.61344, + "grad_norm": 0.12795791029930115, + "learning_rate": 0.00012717525773195878, + "loss": 0.5116, + "step": 1917 + }, + { + "epoch": 0.61376, + "grad_norm": 0.14723263680934906, + "learning_rate": 0.00012713402061855671, + "loss": 0.7565, + "step": 1918 + }, + { + "epoch": 0.61408, + "grad_norm": 0.13519616425037384, + "learning_rate": 0.00012709278350515464, + "loss": 0.6887, + "step": 1919 + }, + { + "epoch": 0.6144, + "grad_norm": 0.14142216742038727, + "learning_rate": 0.00012705154639175258, + "loss": 0.8053, + "step": 1920 + }, + { + "epoch": 0.61472, + "grad_norm": 0.14305110275745392, + "learning_rate": 0.0001270103092783505, + "loss": 0.6143, + "step": 1921 + }, + { + "epoch": 0.61504, + "grad_norm": 0.13851754367351532, + "learning_rate": 0.00012696907216494847, + "loss": 0.7167, + "step": 1922 + }, + { + "epoch": 0.61536, + "grad_norm": 0.12938192486763, + "learning_rate": 0.0001269278350515464, + "loss": 0.5404, + "step": 1923 + }, + { + "epoch": 0.61568, + "grad_norm": 0.13803830742835999, + "learning_rate": 0.00012688659793814436, + "loss": 0.6398, + "step": 1924 + }, + { + "epoch": 0.616, + "grad_norm": 0.14019587635993958, + "learning_rate": 0.0001268453608247423, + "loss": 0.8015, + "step": 1925 + }, + { + "epoch": 0.61632, + "grad_norm": 0.15434010326862335, + "learning_rate": 0.00012680412371134022, + "loss": 0.6866, + "step": 1926 + }, + { + "epoch": 0.61664, + "grad_norm": 0.130121111869812, + "learning_rate": 0.00012676288659793815, + "loss": 0.8558, + "step": 1927 + }, + { + "epoch": 0.61696, + "grad_norm": 0.11969605833292007, + "learning_rate": 0.00012672164948453608, + "loss": 0.7358, + "step": 1928 + }, + { + "epoch": 0.61728, + "grad_norm": 0.13726817071437836, + "learning_rate": 0.000126680412371134, + "loss": 0.7962, + "step": 1929 + }, + { + "epoch": 0.6176, + "grad_norm": 0.16535113751888275, + "learning_rate": 0.00012663917525773197, + "loss": 0.7808, + "step": 1930 + }, + { + "epoch": 0.61792, + "grad_norm": 0.1266782283782959, + "learning_rate": 0.0001265979381443299, + "loss": 0.6858, + "step": 1931 + }, + { + "epoch": 0.61824, + "grad_norm": 0.13304907083511353, + "learning_rate": 0.00012655670103092786, + "loss": 0.5996, + "step": 1932 + }, + { + "epoch": 0.61856, + "grad_norm": 0.1450585424900055, + "learning_rate": 0.0001265154639175258, + "loss": 0.7357, + "step": 1933 + }, + { + "epoch": 0.61888, + "grad_norm": 0.13656657934188843, + "learning_rate": 0.0001264742268041237, + "loss": 0.7243, + "step": 1934 + }, + { + "epoch": 0.6192, + "grad_norm": 0.12630528211593628, + "learning_rate": 0.00012643298969072165, + "loss": 0.6664, + "step": 1935 + }, + { + "epoch": 0.61952, + "grad_norm": 0.14062075316905975, + "learning_rate": 0.00012639175257731958, + "loss": 0.6144, + "step": 1936 + }, + { + "epoch": 0.61984, + "grad_norm": 0.12736253440380096, + "learning_rate": 0.00012635051546391754, + "loss": 0.6227, + "step": 1937 + }, + { + "epoch": 0.62016, + "grad_norm": 0.14937236905097961, + "learning_rate": 0.00012630927835051547, + "loss": 0.6411, + "step": 1938 + }, + { + "epoch": 0.62048, + "grad_norm": 0.1423880010843277, + "learning_rate": 0.0001262680412371134, + "loss": 0.7728, + "step": 1939 + }, + { + "epoch": 0.6208, + "grad_norm": 0.1501777172088623, + "learning_rate": 0.00012622680412371136, + "loss": 0.759, + "step": 1940 + }, + { + "epoch": 0.62112, + "grad_norm": 0.13527531921863556, + "learning_rate": 0.00012618556701030927, + "loss": 0.5421, + "step": 1941 + }, + { + "epoch": 0.62144, + "grad_norm": 0.14614367485046387, + "learning_rate": 0.00012614432989690723, + "loss": 0.6431, + "step": 1942 + }, + { + "epoch": 0.62176, + "grad_norm": 0.13307569921016693, + "learning_rate": 0.00012610309278350516, + "loss": 0.6767, + "step": 1943 + }, + { + "epoch": 0.62208, + "grad_norm": 0.13808198273181915, + "learning_rate": 0.0001260618556701031, + "loss": 0.6282, + "step": 1944 + }, + { + "epoch": 0.6224, + "grad_norm": 0.12743665277957916, + "learning_rate": 0.00012602061855670105, + "loss": 0.5825, + "step": 1945 + }, + { + "epoch": 0.62272, + "grad_norm": 0.1567707061767578, + "learning_rate": 0.00012597938144329898, + "loss": 0.7175, + "step": 1946 + }, + { + "epoch": 0.62304, + "grad_norm": 0.12966367602348328, + "learning_rate": 0.0001259381443298969, + "loss": 0.6444, + "step": 1947 + }, + { + "epoch": 0.62336, + "grad_norm": 0.12914657592773438, + "learning_rate": 0.00012589690721649484, + "loss": 0.5537, + "step": 1948 + }, + { + "epoch": 0.62368, + "grad_norm": 0.1428314447402954, + "learning_rate": 0.00012585567010309277, + "loss": 0.7166, + "step": 1949 + }, + { + "epoch": 0.624, + "grad_norm": 0.14030608534812927, + "learning_rate": 0.00012581443298969073, + "loss": 0.7611, + "step": 1950 + }, + { + "epoch": 0.62432, + "grad_norm": 0.13528037071228027, + "learning_rate": 0.00012577319587628866, + "loss": 0.665, + "step": 1951 + }, + { + "epoch": 0.62464, + "grad_norm": 0.15317410230636597, + "learning_rate": 0.00012573195876288662, + "loss": 0.6768, + "step": 1952 + }, + { + "epoch": 0.62496, + "grad_norm": 0.146495059132576, + "learning_rate": 0.00012569072164948455, + "loss": 0.596, + "step": 1953 + }, + { + "epoch": 0.62528, + "grad_norm": 0.15675994753837585, + "learning_rate": 0.00012564948453608248, + "loss": 0.7209, + "step": 1954 + }, + { + "epoch": 0.6256, + "grad_norm": 0.1454465538263321, + "learning_rate": 0.00012560824742268041, + "loss": 0.8546, + "step": 1955 + }, + { + "epoch": 0.62592, + "grad_norm": 0.11702533811330795, + "learning_rate": 0.00012556701030927835, + "loss": 0.6069, + "step": 1956 + }, + { + "epoch": 0.62624, + "grad_norm": 0.15057525038719177, + "learning_rate": 0.00012552577319587628, + "loss": 0.5768, + "step": 1957 + }, + { + "epoch": 0.62656, + "grad_norm": 0.1403677612543106, + "learning_rate": 0.00012548453608247424, + "loss": 0.7117, + "step": 1958 + }, + { + "epoch": 0.62688, + "grad_norm": 0.1540384590625763, + "learning_rate": 0.00012544329896907217, + "loss": 0.7182, + "step": 1959 + }, + { + "epoch": 0.6272, + "grad_norm": 0.16401363909244537, + "learning_rate": 0.00012540206185567013, + "loss": 0.8016, + "step": 1960 + }, + { + "epoch": 0.62752, + "grad_norm": 0.15659166872501373, + "learning_rate": 0.00012536082474226806, + "loss": 0.5818, + "step": 1961 + }, + { + "epoch": 0.62784, + "grad_norm": 0.1653420329093933, + "learning_rate": 0.000125319587628866, + "loss": 0.7166, + "step": 1962 + }, + { + "epoch": 0.62816, + "grad_norm": 0.14746549725532532, + "learning_rate": 0.00012527835051546392, + "loss": 0.6868, + "step": 1963 + }, + { + "epoch": 0.62848, + "grad_norm": 0.13338971138000488, + "learning_rate": 0.00012523711340206185, + "loss": 0.7502, + "step": 1964 + }, + { + "epoch": 0.6288, + "grad_norm": 0.16360381245613098, + "learning_rate": 0.0001251958762886598, + "loss": 0.7824, + "step": 1965 + }, + { + "epoch": 0.62912, + "grad_norm": 0.15600335597991943, + "learning_rate": 0.00012515463917525774, + "loss": 0.7415, + "step": 1966 + }, + { + "epoch": 0.62944, + "grad_norm": 0.1535622477531433, + "learning_rate": 0.00012511340206185567, + "loss": 0.8199, + "step": 1967 + }, + { + "epoch": 0.62976, + "grad_norm": 0.14531464874744415, + "learning_rate": 0.00012507216494845363, + "loss": 0.5838, + "step": 1968 + }, + { + "epoch": 0.63008, + "grad_norm": 0.14213936030864716, + "learning_rate": 0.00012503092783505156, + "loss": 0.9415, + "step": 1969 + }, + { + "epoch": 0.6304, + "grad_norm": 0.15627896785736084, + "learning_rate": 0.0001249896907216495, + "loss": 0.7646, + "step": 1970 + }, + { + "epoch": 0.63072, + "grad_norm": 0.14554524421691895, + "learning_rate": 0.00012494845360824742, + "loss": 0.793, + "step": 1971 + }, + { + "epoch": 0.63104, + "grad_norm": 0.14182400703430176, + "learning_rate": 0.00012490721649484535, + "loss": 0.7185, + "step": 1972 + }, + { + "epoch": 0.63136, + "grad_norm": 0.15984246134757996, + "learning_rate": 0.0001248659793814433, + "loss": 0.6947, + "step": 1973 + }, + { + "epoch": 0.63168, + "grad_norm": 0.143141508102417, + "learning_rate": 0.00012482474226804124, + "loss": 0.694, + "step": 1974 + }, + { + "epoch": 0.632, + "grad_norm": 0.1503412425518036, + "learning_rate": 0.0001247835051546392, + "loss": 0.5724, + "step": 1975 + }, + { + "epoch": 0.63232, + "grad_norm": 0.15339024364948273, + "learning_rate": 0.0001247422680412371, + "loss": 0.6489, + "step": 1976 + }, + { + "epoch": 0.63264, + "grad_norm": 0.14561650156974792, + "learning_rate": 0.00012470103092783504, + "loss": 0.6682, + "step": 1977 + }, + { + "epoch": 0.63296, + "grad_norm": 0.13642360270023346, + "learning_rate": 0.000124659793814433, + "loss": 0.7385, + "step": 1978 + }, + { + "epoch": 0.63328, + "grad_norm": 0.13638220727443695, + "learning_rate": 0.00012461855670103093, + "loss": 0.5501, + "step": 1979 + }, + { + "epoch": 0.6336, + "grad_norm": 0.14778625965118408, + "learning_rate": 0.00012457731958762889, + "loss": 0.6875, + "step": 1980 + }, + { + "epoch": 0.63392, + "grad_norm": 0.1807885617017746, + "learning_rate": 0.00012453608247422682, + "loss": 0.7934, + "step": 1981 + }, + { + "epoch": 0.63424, + "grad_norm": 0.1617950052022934, + "learning_rate": 0.00012449484536082475, + "loss": 0.7168, + "step": 1982 + }, + { + "epoch": 0.63456, + "grad_norm": 0.1469191014766693, + "learning_rate": 0.00012445360824742268, + "loss": 0.6633, + "step": 1983 + }, + { + "epoch": 0.63488, + "grad_norm": 0.14085760712623596, + "learning_rate": 0.0001244123711340206, + "loss": 0.6816, + "step": 1984 + }, + { + "epoch": 0.6352, + "grad_norm": 0.13655433058738708, + "learning_rate": 0.00012437113402061854, + "loss": 0.6451, + "step": 1985 + }, + { + "epoch": 0.63552, + "grad_norm": 0.1581992506980896, + "learning_rate": 0.0001243298969072165, + "loss": 0.7158, + "step": 1986 + }, + { + "epoch": 0.63584, + "grad_norm": 0.13187338411808014, + "learning_rate": 0.00012428865979381443, + "loss": 0.5549, + "step": 1987 + }, + { + "epoch": 0.63616, + "grad_norm": 0.13619250059127808, + "learning_rate": 0.0001242474226804124, + "loss": 0.9195, + "step": 1988 + }, + { + "epoch": 0.63648, + "grad_norm": 0.13870079815387726, + "learning_rate": 0.00012420618556701032, + "loss": 0.5944, + "step": 1989 + }, + { + "epoch": 0.6368, + "grad_norm": 0.12829461693763733, + "learning_rate": 0.00012416494845360825, + "loss": 0.7286, + "step": 1990 + }, + { + "epoch": 0.63712, + "grad_norm": 0.13943707942962646, + "learning_rate": 0.00012412371134020618, + "loss": 0.5972, + "step": 1991 + }, + { + "epoch": 0.63744, + "grad_norm": 0.13263818621635437, + "learning_rate": 0.00012408247422680412, + "loss": 0.797, + "step": 1992 + }, + { + "epoch": 0.63776, + "grad_norm": 0.13591541349887848, + "learning_rate": 0.00012404123711340207, + "loss": 0.7167, + "step": 1993 + }, + { + "epoch": 0.63808, + "grad_norm": 0.14553944766521454, + "learning_rate": 0.000124, + "loss": 0.649, + "step": 1994 + }, + { + "epoch": 0.6384, + "grad_norm": 0.15163764357566833, + "learning_rate": 0.00012395876288659794, + "loss": 0.7781, + "step": 1995 + }, + { + "epoch": 0.63872, + "grad_norm": 0.15245743095874786, + "learning_rate": 0.0001239175257731959, + "loss": 0.6715, + "step": 1996 + }, + { + "epoch": 0.63904, + "grad_norm": 0.1749904602766037, + "learning_rate": 0.00012387628865979383, + "loss": 0.692, + "step": 1997 + }, + { + "epoch": 0.63936, + "grad_norm": 0.13300080597400665, + "learning_rate": 0.00012383505154639176, + "loss": 0.5676, + "step": 1998 + }, + { + "epoch": 0.63968, + "grad_norm": 0.15774020552635193, + "learning_rate": 0.0001237938144329897, + "loss": 0.7337, + "step": 1999 + }, + { + "epoch": 0.64, + "grad_norm": 0.13856618106365204, + "learning_rate": 0.00012375257731958762, + "loss": 0.5507, + "step": 2000 + }, + { + "epoch": 0.64032, + "grad_norm": 0.14179150760173798, + "learning_rate": 0.00012371134020618558, + "loss": 0.727, + "step": 2001 + }, + { + "epoch": 0.64064, + "grad_norm": 0.1413705050945282, + "learning_rate": 0.0001236701030927835, + "loss": 0.5845, + "step": 2002 + }, + { + "epoch": 0.64096, + "grad_norm": 0.14309062063694, + "learning_rate": 0.00012362886597938147, + "loss": 0.5994, + "step": 2003 + }, + { + "epoch": 0.64128, + "grad_norm": 0.14631907641887665, + "learning_rate": 0.0001235876288659794, + "loss": 0.5985, + "step": 2004 + }, + { + "epoch": 0.6416, + "grad_norm": 0.13707837462425232, + "learning_rate": 0.00012354639175257733, + "loss": 0.7102, + "step": 2005 + }, + { + "epoch": 0.64192, + "grad_norm": 0.14122074842453003, + "learning_rate": 0.00012350515463917526, + "loss": 0.704, + "step": 2006 + }, + { + "epoch": 0.64224, + "grad_norm": 0.13491104543209076, + "learning_rate": 0.0001234639175257732, + "loss": 0.6504, + "step": 2007 + }, + { + "epoch": 0.64256, + "grad_norm": 0.13734230399131775, + "learning_rate": 0.00012342268041237115, + "loss": 0.6335, + "step": 2008 + }, + { + "epoch": 0.64288, + "grad_norm": 0.16213947534561157, + "learning_rate": 0.00012338144329896908, + "loss": 0.7584, + "step": 2009 + }, + { + "epoch": 0.6432, + "grad_norm": 0.13865743577480316, + "learning_rate": 0.00012334020618556701, + "loss": 0.727, + "step": 2010 + }, + { + "epoch": 0.64352, + "grad_norm": 0.1343052089214325, + "learning_rate": 0.00012329896907216497, + "loss": 0.6881, + "step": 2011 + }, + { + "epoch": 0.64384, + "grad_norm": 0.14671185612678528, + "learning_rate": 0.00012325773195876288, + "loss": 0.6755, + "step": 2012 + }, + { + "epoch": 0.64416, + "grad_norm": 0.1421107053756714, + "learning_rate": 0.00012321649484536084, + "loss": 0.6594, + "step": 2013 + }, + { + "epoch": 0.64448, + "grad_norm": 0.12019284069538116, + "learning_rate": 0.00012317525773195877, + "loss": 0.6073, + "step": 2014 + }, + { + "epoch": 0.6448, + "grad_norm": 0.14594294130802155, + "learning_rate": 0.0001231340206185567, + "loss": 0.5721, + "step": 2015 + }, + { + "epoch": 0.64512, + "grad_norm": 0.1317436248064041, + "learning_rate": 0.00012309278350515466, + "loss": 0.7781, + "step": 2016 + }, + { + "epoch": 0.64544, + "grad_norm": 0.16204434633255005, + "learning_rate": 0.0001230515463917526, + "loss": 0.6408, + "step": 2017 + }, + { + "epoch": 0.64576, + "grad_norm": 0.12162824720144272, + "learning_rate": 0.00012301030927835052, + "loss": 0.5489, + "step": 2018 + }, + { + "epoch": 0.64608, + "grad_norm": 0.14092159271240234, + "learning_rate": 0.00012296907216494845, + "loss": 0.5897, + "step": 2019 + }, + { + "epoch": 0.6464, + "grad_norm": 0.12317449599504471, + "learning_rate": 0.00012292783505154638, + "loss": 0.7026, + "step": 2020 + }, + { + "epoch": 0.64672, + "grad_norm": 0.13785576820373535, + "learning_rate": 0.00012288659793814434, + "loss": 0.634, + "step": 2021 + }, + { + "epoch": 0.64704, + "grad_norm": 0.12226948887109756, + "learning_rate": 0.00012284536082474227, + "loss": 0.7811, + "step": 2022 + }, + { + "epoch": 0.64736, + "grad_norm": 0.15511289238929749, + "learning_rate": 0.0001228041237113402, + "loss": 0.6711, + "step": 2023 + }, + { + "epoch": 0.64768, + "grad_norm": 0.13871407508850098, + "learning_rate": 0.00012276288659793816, + "loss": 0.5914, + "step": 2024 + }, + { + "epoch": 0.648, + "grad_norm": 0.14690019190311432, + "learning_rate": 0.0001227216494845361, + "loss": 0.7308, + "step": 2025 + }, + { + "epoch": 0.64832, + "grad_norm": 0.140092134475708, + "learning_rate": 0.00012268041237113402, + "loss": 0.4961, + "step": 2026 + }, + { + "epoch": 0.64864, + "grad_norm": 0.13826578855514526, + "learning_rate": 0.00012263917525773195, + "loss": 0.7158, + "step": 2027 + }, + { + "epoch": 0.64896, + "grad_norm": 0.13793012499809265, + "learning_rate": 0.00012259793814432989, + "loss": 0.6803, + "step": 2028 + }, + { + "epoch": 0.64928, + "grad_norm": 0.15876394510269165, + "learning_rate": 0.00012255670103092784, + "loss": 0.7771, + "step": 2029 + }, + { + "epoch": 0.6496, + "grad_norm": 0.1506621390581131, + "learning_rate": 0.00012251546391752578, + "loss": 0.5935, + "step": 2030 + }, + { + "epoch": 0.64992, + "grad_norm": 0.13091327250003815, + "learning_rate": 0.00012247422680412373, + "loss": 0.7151, + "step": 2031 + }, + { + "epoch": 0.65024, + "grad_norm": 0.1396481841802597, + "learning_rate": 0.00012243298969072166, + "loss": 0.765, + "step": 2032 + }, + { + "epoch": 0.65056, + "grad_norm": 0.135127454996109, + "learning_rate": 0.0001223917525773196, + "loss": 0.8282, + "step": 2033 + }, + { + "epoch": 0.65088, + "grad_norm": 0.1507946103811264, + "learning_rate": 0.00012235051546391753, + "loss": 0.7046, + "step": 2034 + }, + { + "epoch": 0.6512, + "grad_norm": 0.14329278469085693, + "learning_rate": 0.00012230927835051546, + "loss": 0.6782, + "step": 2035 + }, + { + "epoch": 0.65152, + "grad_norm": 0.14071671664714813, + "learning_rate": 0.00012226804123711342, + "loss": 0.7114, + "step": 2036 + }, + { + "epoch": 0.65184, + "grad_norm": 0.13704700767993927, + "learning_rate": 0.00012222680412371135, + "loss": 0.6778, + "step": 2037 + }, + { + "epoch": 0.65216, + "grad_norm": 0.14313676953315735, + "learning_rate": 0.00012218556701030928, + "loss": 0.6243, + "step": 2038 + }, + { + "epoch": 0.65248, + "grad_norm": 0.16129252314567566, + "learning_rate": 0.00012214432989690724, + "loss": 0.6778, + "step": 2039 + }, + { + "epoch": 0.6528, + "grad_norm": 0.15804162621498108, + "learning_rate": 0.00012210309278350517, + "loss": 0.7778, + "step": 2040 + }, + { + "epoch": 0.65312, + "grad_norm": 0.14212483167648315, + "learning_rate": 0.00012206185567010311, + "loss": 0.7429, + "step": 2041 + }, + { + "epoch": 0.65344, + "grad_norm": 0.14250007271766663, + "learning_rate": 0.00012202061855670103, + "loss": 0.7414, + "step": 2042 + }, + { + "epoch": 0.65376, + "grad_norm": 0.14517898857593536, + "learning_rate": 0.00012197938144329896, + "loss": 0.6837, + "step": 2043 + }, + { + "epoch": 0.65408, + "grad_norm": 0.15459880232810974, + "learning_rate": 0.00012193814432989692, + "loss": 0.5952, + "step": 2044 + }, + { + "epoch": 0.6544, + "grad_norm": 0.13581113517284393, + "learning_rate": 0.00012189690721649485, + "loss": 0.6004, + "step": 2045 + }, + { + "epoch": 0.65472, + "grad_norm": 0.1296541839838028, + "learning_rate": 0.00012185567010309278, + "loss": 0.6925, + "step": 2046 + }, + { + "epoch": 0.65504, + "grad_norm": 0.13644792139530182, + "learning_rate": 0.00012181443298969073, + "loss": 0.7142, + "step": 2047 + }, + { + "epoch": 0.65536, + "grad_norm": 0.14497695863246918, + "learning_rate": 0.00012177319587628866, + "loss": 0.7295, + "step": 2048 + }, + { + "epoch": 0.65568, + "grad_norm": 0.12832631170749664, + "learning_rate": 0.0001217319587628866, + "loss": 0.5929, + "step": 2049 + }, + { + "epoch": 0.656, + "grad_norm": 0.15335868299007416, + "learning_rate": 0.00012169072164948454, + "loss": 0.8115, + "step": 2050 + }, + { + "epoch": 0.65632, + "grad_norm": 0.14223381876945496, + "learning_rate": 0.00012164948453608247, + "loss": 0.5943, + "step": 2051 + }, + { + "epoch": 0.65664, + "grad_norm": 0.14100460708141327, + "learning_rate": 0.00012160824742268043, + "loss": 0.6994, + "step": 2052 + }, + { + "epoch": 0.65696, + "grad_norm": 0.1499370038509369, + "learning_rate": 0.00012156701030927836, + "loss": 0.8442, + "step": 2053 + }, + { + "epoch": 0.65728, + "grad_norm": 0.1352858692407608, + "learning_rate": 0.0001215257731958763, + "loss": 0.7045, + "step": 2054 + }, + { + "epoch": 0.6576, + "grad_norm": 0.13496248424053192, + "learning_rate": 0.00012148453608247423, + "loss": 0.6771, + "step": 2055 + }, + { + "epoch": 0.65792, + "grad_norm": 0.146116703748703, + "learning_rate": 0.00012144329896907216, + "loss": 0.628, + "step": 2056 + }, + { + "epoch": 0.65824, + "grad_norm": 0.12573496997356415, + "learning_rate": 0.00012140206185567011, + "loss": 0.6114, + "step": 2057 + }, + { + "epoch": 0.65856, + "grad_norm": 0.1435246616601944, + "learning_rate": 0.00012136082474226804, + "loss": 0.7657, + "step": 2058 + }, + { + "epoch": 0.65888, + "grad_norm": 0.1303691864013672, + "learning_rate": 0.000121319587628866, + "loss": 0.6495, + "step": 2059 + }, + { + "epoch": 0.6592, + "grad_norm": 0.13893800973892212, + "learning_rate": 0.00012127835051546392, + "loss": 0.6923, + "step": 2060 + }, + { + "epoch": 0.65952, + "grad_norm": 0.13547277450561523, + "learning_rate": 0.00012123711340206185, + "loss": 0.5702, + "step": 2061 + }, + { + "epoch": 0.65984, + "grad_norm": 0.13063783943653107, + "learning_rate": 0.0001211958762886598, + "loss": 0.6577, + "step": 2062 + }, + { + "epoch": 0.66016, + "grad_norm": 0.12719577550888062, + "learning_rate": 0.00012115463917525774, + "loss": 0.4265, + "step": 2063 + }, + { + "epoch": 0.66048, + "grad_norm": 0.1602306067943573, + "learning_rate": 0.00012111340206185568, + "loss": 0.7139, + "step": 2064 + }, + { + "epoch": 0.6608, + "grad_norm": 0.16059952974319458, + "learning_rate": 0.00012107216494845361, + "loss": 0.7856, + "step": 2065 + }, + { + "epoch": 0.66112, + "grad_norm": 0.1634436398744583, + "learning_rate": 0.00012103092783505155, + "loss": 0.7477, + "step": 2066 + }, + { + "epoch": 0.66144, + "grad_norm": 0.14774012565612793, + "learning_rate": 0.00012098969072164949, + "loss": 0.7642, + "step": 2067 + }, + { + "epoch": 0.66176, + "grad_norm": 0.1423511803150177, + "learning_rate": 0.00012094845360824742, + "loss": 0.6945, + "step": 2068 + }, + { + "epoch": 0.66208, + "grad_norm": 0.14613443613052368, + "learning_rate": 0.00012090721649484538, + "loss": 0.7399, + "step": 2069 + }, + { + "epoch": 0.6624, + "grad_norm": 0.13143277168273926, + "learning_rate": 0.00012086597938144331, + "loss": 0.6649, + "step": 2070 + }, + { + "epoch": 0.66272, + "grad_norm": 0.143715038895607, + "learning_rate": 0.00012082474226804124, + "loss": 0.7525, + "step": 2071 + }, + { + "epoch": 0.66304, + "grad_norm": 0.13922396302223206, + "learning_rate": 0.00012078350515463919, + "loss": 0.6464, + "step": 2072 + }, + { + "epoch": 0.66336, + "grad_norm": 0.14103561639785767, + "learning_rate": 0.00012074226804123712, + "loss": 0.6986, + "step": 2073 + }, + { + "epoch": 0.66368, + "grad_norm": 0.11990336328744888, + "learning_rate": 0.00012070103092783506, + "loss": 0.7182, + "step": 2074 + }, + { + "epoch": 0.664, + "grad_norm": 0.13981446623802185, + "learning_rate": 0.000120659793814433, + "loss": 0.5347, + "step": 2075 + }, + { + "epoch": 0.66432, + "grad_norm": 0.14048956334590912, + "learning_rate": 0.00012061855670103093, + "loss": 0.5711, + "step": 2076 + }, + { + "epoch": 0.66464, + "grad_norm": 0.14830324053764343, + "learning_rate": 0.00012057731958762888, + "loss": 0.7842, + "step": 2077 + }, + { + "epoch": 0.66496, + "grad_norm": 0.14833824336528778, + "learning_rate": 0.0001205360824742268, + "loss": 0.6799, + "step": 2078 + }, + { + "epoch": 0.66528, + "grad_norm": 0.15588924288749695, + "learning_rate": 0.00012049484536082473, + "loss": 0.7668, + "step": 2079 + }, + { + "epoch": 0.6656, + "grad_norm": 0.14304298162460327, + "learning_rate": 0.00012045360824742269, + "loss": 0.6254, + "step": 2080 + }, + { + "epoch": 0.66592, + "grad_norm": 0.13424982130527496, + "learning_rate": 0.00012041237113402062, + "loss": 0.7185, + "step": 2081 + }, + { + "epoch": 0.66624, + "grad_norm": 0.134463369846344, + "learning_rate": 0.00012037113402061857, + "loss": 0.8457, + "step": 2082 + }, + { + "epoch": 0.66656, + "grad_norm": 0.17415812611579895, + "learning_rate": 0.0001203298969072165, + "loss": 0.6671, + "step": 2083 + }, + { + "epoch": 0.66688, + "grad_norm": 0.12847645580768585, + "learning_rate": 0.00012028865979381443, + "loss": 0.753, + "step": 2084 + }, + { + "epoch": 0.6672, + "grad_norm": 0.13758262991905212, + "learning_rate": 0.00012024742268041237, + "loss": 0.6031, + "step": 2085 + }, + { + "epoch": 0.66752, + "grad_norm": 0.13060477375984192, + "learning_rate": 0.0001202061855670103, + "loss": 0.6655, + "step": 2086 + }, + { + "epoch": 0.66784, + "grad_norm": 0.1436629444360733, + "learning_rate": 0.00012016494845360826, + "loss": 0.6334, + "step": 2087 + }, + { + "epoch": 0.66816, + "grad_norm": 0.15198498964309692, + "learning_rate": 0.0001201237113402062, + "loss": 0.6901, + "step": 2088 + }, + { + "epoch": 0.66848, + "grad_norm": 0.1374773532152176, + "learning_rate": 0.00012008247422680413, + "loss": 0.6681, + "step": 2089 + }, + { + "epoch": 0.6688, + "grad_norm": 0.1516476422548294, + "learning_rate": 0.00012004123711340207, + "loss": 0.7825, + "step": 2090 + }, + { + "epoch": 0.66912, + "grad_norm": 0.1323966383934021, + "learning_rate": 0.00012, + "loss": 0.8294, + "step": 2091 + }, + { + "epoch": 0.66944, + "grad_norm": 0.1251801699399948, + "learning_rate": 0.00011995876288659795, + "loss": 0.6912, + "step": 2092 + }, + { + "epoch": 0.66976, + "grad_norm": 0.14561590552330017, + "learning_rate": 0.00011991752577319588, + "loss": 0.6674, + "step": 2093 + }, + { + "epoch": 0.67008, + "grad_norm": 0.13464321196079254, + "learning_rate": 0.00011987628865979381, + "loss": 0.5783, + "step": 2094 + }, + { + "epoch": 0.6704, + "grad_norm": 0.13263243436813354, + "learning_rate": 0.00011983505154639177, + "loss": 0.6283, + "step": 2095 + }, + { + "epoch": 0.67072, + "grad_norm": 0.14757391810417175, + "learning_rate": 0.00011979381443298969, + "loss": 0.677, + "step": 2096 + }, + { + "epoch": 0.67104, + "grad_norm": 0.13637281954288483, + "learning_rate": 0.00011975257731958764, + "loss": 0.6223, + "step": 2097 + }, + { + "epoch": 0.67136, + "grad_norm": 0.15317052602767944, + "learning_rate": 0.00011971134020618558, + "loss": 0.8012, + "step": 2098 + }, + { + "epoch": 0.67168, + "grad_norm": 0.1539645791053772, + "learning_rate": 0.00011967010309278351, + "loss": 0.6283, + "step": 2099 + }, + { + "epoch": 0.672, + "grad_norm": 0.13951300084590912, + "learning_rate": 0.00011962886597938145, + "loss": 0.7161, + "step": 2100 + }, + { + "epoch": 0.67232, + "grad_norm": 0.12808813154697418, + "learning_rate": 0.00011958762886597938, + "loss": 0.7182, + "step": 2101 + }, + { + "epoch": 0.67264, + "grad_norm": 0.14265356957912445, + "learning_rate": 0.00011954639175257734, + "loss": 0.7135, + "step": 2102 + }, + { + "epoch": 0.67296, + "grad_norm": 0.15047064423561096, + "learning_rate": 0.00011950515463917526, + "loss": 0.6332, + "step": 2103 + }, + { + "epoch": 0.67328, + "grad_norm": 0.1577232927083969, + "learning_rate": 0.00011946391752577319, + "loss": 0.6437, + "step": 2104 + }, + { + "epoch": 0.6736, + "grad_norm": 0.1499931365251541, + "learning_rate": 0.00011942268041237115, + "loss": 0.6061, + "step": 2105 + }, + { + "epoch": 0.67392, + "grad_norm": 0.147731214761734, + "learning_rate": 0.00011938144329896908, + "loss": 0.5464, + "step": 2106 + }, + { + "epoch": 0.67424, + "grad_norm": 0.14864850044250488, + "learning_rate": 0.000119340206185567, + "loss": 0.6343, + "step": 2107 + }, + { + "epoch": 0.67456, + "grad_norm": 0.1334935575723648, + "learning_rate": 0.00011929896907216496, + "loss": 0.6889, + "step": 2108 + }, + { + "epoch": 0.67488, + "grad_norm": 0.14304758608341217, + "learning_rate": 0.00011925773195876289, + "loss": 0.755, + "step": 2109 + }, + { + "epoch": 0.6752, + "grad_norm": 0.17663174867630005, + "learning_rate": 0.00011921649484536083, + "loss": 0.6119, + "step": 2110 + }, + { + "epoch": 0.67552, + "grad_norm": 0.1393556445837021, + "learning_rate": 0.00011917525773195876, + "loss": 0.7621, + "step": 2111 + }, + { + "epoch": 0.67584, + "grad_norm": 0.133917897939682, + "learning_rate": 0.0001191340206185567, + "loss": 0.6723, + "step": 2112 + }, + { + "epoch": 0.67616, + "grad_norm": 0.12147454172372818, + "learning_rate": 0.00011909278350515465, + "loss": 0.5869, + "step": 2113 + }, + { + "epoch": 0.67648, + "grad_norm": 0.1496019959449768, + "learning_rate": 0.00011905154639175257, + "loss": 0.6668, + "step": 2114 + }, + { + "epoch": 0.6768, + "grad_norm": 0.14451980590820312, + "learning_rate": 0.00011901030927835053, + "loss": 0.6461, + "step": 2115 + }, + { + "epoch": 0.67712, + "grad_norm": 0.1440231204032898, + "learning_rate": 0.00011896907216494846, + "loss": 0.5656, + "step": 2116 + }, + { + "epoch": 0.67744, + "grad_norm": 0.13051605224609375, + "learning_rate": 0.00011892783505154639, + "loss": 0.6012, + "step": 2117 + }, + { + "epoch": 0.67776, + "grad_norm": 0.14402255415916443, + "learning_rate": 0.00011888659793814434, + "loss": 0.6596, + "step": 2118 + }, + { + "epoch": 0.67808, + "grad_norm": 0.14876003563404083, + "learning_rate": 0.00011884536082474227, + "loss": 0.6454, + "step": 2119 + }, + { + "epoch": 0.6784, + "grad_norm": 0.13757188618183136, + "learning_rate": 0.00011880412371134023, + "loss": 0.6516, + "step": 2120 + }, + { + "epoch": 0.67872, + "grad_norm": 0.13989025354385376, + "learning_rate": 0.00011876288659793814, + "loss": 0.6098, + "step": 2121 + }, + { + "epoch": 0.67904, + "grad_norm": 0.13720832765102386, + "learning_rate": 0.00011872164948453608, + "loss": 0.725, + "step": 2122 + }, + { + "epoch": 0.67936, + "grad_norm": 0.14132976531982422, + "learning_rate": 0.00011868041237113403, + "loss": 0.8721, + "step": 2123 + }, + { + "epoch": 0.67968, + "grad_norm": 0.12852109968662262, + "learning_rate": 0.00011863917525773197, + "loss": 0.6251, + "step": 2124 + }, + { + "epoch": 0.68, + "grad_norm": 0.1427404284477234, + "learning_rate": 0.00011859793814432991, + "loss": 0.6935, + "step": 2125 + }, + { + "epoch": 0.68032, + "grad_norm": 0.14873456954956055, + "learning_rate": 0.00011855670103092784, + "loss": 0.7003, + "step": 2126 + }, + { + "epoch": 0.68064, + "grad_norm": 0.15125808119773865, + "learning_rate": 0.00011851546391752577, + "loss": 0.7124, + "step": 2127 + }, + { + "epoch": 0.68096, + "grad_norm": 0.15154671669006348, + "learning_rate": 0.00011847422680412372, + "loss": 0.7452, + "step": 2128 + }, + { + "epoch": 0.68128, + "grad_norm": 0.14374732971191406, + "learning_rate": 0.00011843298969072165, + "loss": 0.7728, + "step": 2129 + }, + { + "epoch": 0.6816, + "grad_norm": 0.15634573996067047, + "learning_rate": 0.00011839175257731961, + "loss": 0.6528, + "step": 2130 + }, + { + "epoch": 0.68192, + "grad_norm": 0.12566031515598297, + "learning_rate": 0.00011835051546391754, + "loss": 0.6598, + "step": 2131 + }, + { + "epoch": 0.68224, + "grad_norm": 0.16290156543254852, + "learning_rate": 0.00011830927835051546, + "loss": 0.7089, + "step": 2132 + }, + { + "epoch": 0.68256, + "grad_norm": 0.1287078559398651, + "learning_rate": 0.00011826804123711341, + "loss": 0.7569, + "step": 2133 + }, + { + "epoch": 0.68288, + "grad_norm": 0.15098051726818085, + "learning_rate": 0.00011822680412371135, + "loss": 0.6207, + "step": 2134 + }, + { + "epoch": 0.6832, + "grad_norm": 0.14738884568214417, + "learning_rate": 0.00011818556701030929, + "loss": 0.5538, + "step": 2135 + }, + { + "epoch": 0.68352, + "grad_norm": 0.14219225943088531, + "learning_rate": 0.00011814432989690722, + "loss": 0.5939, + "step": 2136 + }, + { + "epoch": 0.68384, + "grad_norm": 0.14703382551670074, + "learning_rate": 0.00011810309278350515, + "loss": 0.5749, + "step": 2137 + }, + { + "epoch": 0.68416, + "grad_norm": 0.1436077207326889, + "learning_rate": 0.00011806185567010311, + "loss": 0.5331, + "step": 2138 + }, + { + "epoch": 0.68448, + "grad_norm": 0.1349116861820221, + "learning_rate": 0.00011802061855670103, + "loss": 0.8227, + "step": 2139 + }, + { + "epoch": 0.6848, + "grad_norm": 0.15986260771751404, + "learning_rate": 0.00011797938144329896, + "loss": 0.7254, + "step": 2140 + }, + { + "epoch": 0.68512, + "grad_norm": 0.15659818053245544, + "learning_rate": 0.00011793814432989692, + "loss": 0.8652, + "step": 2141 + }, + { + "epoch": 0.68544, + "grad_norm": 0.12938763201236725, + "learning_rate": 0.00011789690721649485, + "loss": 0.707, + "step": 2142 + }, + { + "epoch": 0.68576, + "grad_norm": 0.14879067242145538, + "learning_rate": 0.0001178556701030928, + "loss": 0.6229, + "step": 2143 + }, + { + "epoch": 0.68608, + "grad_norm": 0.1321226805448532, + "learning_rate": 0.00011781443298969073, + "loss": 0.6679, + "step": 2144 + }, + { + "epoch": 0.6864, + "grad_norm": 0.14504100382328033, + "learning_rate": 0.00011777319587628866, + "loss": 0.6423, + "step": 2145 + }, + { + "epoch": 0.68672, + "grad_norm": 0.12305402755737305, + "learning_rate": 0.0001177319587628866, + "loss": 0.6218, + "step": 2146 + }, + { + "epoch": 0.68704, + "grad_norm": 0.15316817164421082, + "learning_rate": 0.00011769072164948453, + "loss": 0.7296, + "step": 2147 + }, + { + "epoch": 0.68736, + "grad_norm": 0.13577841222286224, + "learning_rate": 0.00011764948453608249, + "loss": 0.7849, + "step": 2148 + }, + { + "epoch": 0.68768, + "grad_norm": 0.1502908319234848, + "learning_rate": 0.00011760824742268042, + "loss": 0.7167, + "step": 2149 + }, + { + "epoch": 0.688, + "grad_norm": 0.15807226300239563, + "learning_rate": 0.00011756701030927834, + "loss": 0.7735, + "step": 2150 + }, + { + "epoch": 0.68832, + "grad_norm": 0.16051429510116577, + "learning_rate": 0.0001175257731958763, + "loss": 0.7426, + "step": 2151 + }, + { + "epoch": 0.68864, + "grad_norm": 0.14449506998062134, + "learning_rate": 0.00011748453608247423, + "loss": 0.6979, + "step": 2152 + }, + { + "epoch": 0.68896, + "grad_norm": 0.13942117989063263, + "learning_rate": 0.00011744329896907218, + "loss": 0.5842, + "step": 2153 + }, + { + "epoch": 0.68928, + "grad_norm": 0.1369529664516449, + "learning_rate": 0.00011740206185567011, + "loss": 0.7403, + "step": 2154 + }, + { + "epoch": 0.6896, + "grad_norm": 0.1495199203491211, + "learning_rate": 0.00011736082474226804, + "loss": 0.721, + "step": 2155 + }, + { + "epoch": 0.68992, + "grad_norm": 0.1454606056213379, + "learning_rate": 0.000117319587628866, + "loss": 0.6192, + "step": 2156 + }, + { + "epoch": 0.69024, + "grad_norm": 0.1509052962064743, + "learning_rate": 0.00011727835051546391, + "loss": 0.7074, + "step": 2157 + }, + { + "epoch": 0.69056, + "grad_norm": 0.1502697616815567, + "learning_rate": 0.00011723711340206187, + "loss": 0.664, + "step": 2158 + }, + { + "epoch": 0.69088, + "grad_norm": 0.1388171762228012, + "learning_rate": 0.0001171958762886598, + "loss": 0.6384, + "step": 2159 + }, + { + "epoch": 0.6912, + "grad_norm": 0.12090221792459488, + "learning_rate": 0.00011715463917525774, + "loss": 0.6415, + "step": 2160 + }, + { + "epoch": 0.69152, + "grad_norm": 0.16175298392772675, + "learning_rate": 0.00011711340206185568, + "loss": 0.71, + "step": 2161 + }, + { + "epoch": 0.69184, + "grad_norm": 0.13869957625865936, + "learning_rate": 0.00011707216494845361, + "loss": 0.6828, + "step": 2162 + }, + { + "epoch": 0.69216, + "grad_norm": 0.1311745047569275, + "learning_rate": 0.00011703092783505156, + "loss": 0.6653, + "step": 2163 + }, + { + "epoch": 0.69248, + "grad_norm": 0.15573452413082123, + "learning_rate": 0.00011698969072164949, + "loss": 0.7371, + "step": 2164 + }, + { + "epoch": 0.6928, + "grad_norm": 0.12786449491977692, + "learning_rate": 0.00011694845360824742, + "loss": 0.5788, + "step": 2165 + }, + { + "epoch": 0.69312, + "grad_norm": 0.15374130010604858, + "learning_rate": 0.00011690721649484538, + "loss": 0.7331, + "step": 2166 + }, + { + "epoch": 0.69344, + "grad_norm": 0.14082935452461243, + "learning_rate": 0.00011686597938144331, + "loss": 0.8885, + "step": 2167 + }, + { + "epoch": 0.69376, + "grad_norm": 0.13708099722862244, + "learning_rate": 0.00011682474226804123, + "loss": 0.6278, + "step": 2168 + }, + { + "epoch": 0.69408, + "grad_norm": 0.13642029464244843, + "learning_rate": 0.00011678350515463918, + "loss": 0.6652, + "step": 2169 + }, + { + "epoch": 0.6944, + "grad_norm": 0.14015568792819977, + "learning_rate": 0.00011674226804123712, + "loss": 0.6787, + "step": 2170 + }, + { + "epoch": 0.69472, + "grad_norm": 0.14423607289791107, + "learning_rate": 0.00011670103092783506, + "loss": 0.6913, + "step": 2171 + }, + { + "epoch": 0.69504, + "grad_norm": 0.14666514098644257, + "learning_rate": 0.00011665979381443299, + "loss": 0.7238, + "step": 2172 + }, + { + "epoch": 0.69536, + "grad_norm": 0.14900925755500793, + "learning_rate": 0.00011661855670103092, + "loss": 0.8619, + "step": 2173 + }, + { + "epoch": 0.69568, + "grad_norm": 0.14885465800762177, + "learning_rate": 0.00011657731958762888, + "loss": 0.5463, + "step": 2174 + }, + { + "epoch": 0.696, + "grad_norm": 0.12460866570472717, + "learning_rate": 0.0001165360824742268, + "loss": 0.6635, + "step": 2175 + }, + { + "epoch": 0.69632, + "grad_norm": 0.15186761319637299, + "learning_rate": 0.00011649484536082476, + "loss": 0.6846, + "step": 2176 + }, + { + "epoch": 0.69664, + "grad_norm": 0.13989900052547455, + "learning_rate": 0.00011645360824742269, + "loss": 0.6585, + "step": 2177 + }, + { + "epoch": 0.69696, + "grad_norm": 0.14354805648326874, + "learning_rate": 0.00011641237113402062, + "loss": 0.701, + "step": 2178 + }, + { + "epoch": 0.69728, + "grad_norm": 0.12834502756595612, + "learning_rate": 0.00011637113402061857, + "loss": 0.7102, + "step": 2179 + }, + { + "epoch": 0.6976, + "grad_norm": 0.15479061007499695, + "learning_rate": 0.0001163298969072165, + "loss": 0.6082, + "step": 2180 + }, + { + "epoch": 0.69792, + "grad_norm": 0.15635207295417786, + "learning_rate": 0.00011628865979381444, + "loss": 0.5701, + "step": 2181 + }, + { + "epoch": 0.69824, + "grad_norm": 0.1427372545003891, + "learning_rate": 0.00011624742268041237, + "loss": 0.8152, + "step": 2182 + }, + { + "epoch": 0.69856, + "grad_norm": 0.1362147182226181, + "learning_rate": 0.0001162061855670103, + "loss": 0.586, + "step": 2183 + }, + { + "epoch": 0.69888, + "grad_norm": 0.1717229038476944, + "learning_rate": 0.00011616494845360826, + "loss": 0.6253, + "step": 2184 + }, + { + "epoch": 0.6992, + "grad_norm": 0.14592789113521576, + "learning_rate": 0.0001161237113402062, + "loss": 0.7353, + "step": 2185 + }, + { + "epoch": 0.69952, + "grad_norm": 0.1312609314918518, + "learning_rate": 0.00011608247422680414, + "loss": 0.7015, + "step": 2186 + }, + { + "epoch": 0.69984, + "grad_norm": 0.1368933916091919, + "learning_rate": 0.00011604123711340207, + "loss": 0.6785, + "step": 2187 + }, + { + "epoch": 0.70016, + "grad_norm": 0.15068896114826202, + "learning_rate": 0.000116, + "loss": 0.6236, + "step": 2188 + }, + { + "epoch": 0.70048, + "grad_norm": 0.1282639354467392, + "learning_rate": 0.00011595876288659795, + "loss": 0.8072, + "step": 2189 + }, + { + "epoch": 0.7008, + "grad_norm": 0.15041980147361755, + "learning_rate": 0.00011591752577319588, + "loss": 0.6618, + "step": 2190 + }, + { + "epoch": 0.70112, + "grad_norm": 0.1336228996515274, + "learning_rate": 0.00011587628865979384, + "loss": 0.7483, + "step": 2191 + }, + { + "epoch": 0.70144, + "grad_norm": 0.12458977848291397, + "learning_rate": 0.00011583505154639175, + "loss": 0.7225, + "step": 2192 + }, + { + "epoch": 0.70176, + "grad_norm": 0.1439562737941742, + "learning_rate": 0.00011579381443298968, + "loss": 0.6446, + "step": 2193 + }, + { + "epoch": 0.70208, + "grad_norm": 0.1581539511680603, + "learning_rate": 0.00011575257731958764, + "loss": 0.7608, + "step": 2194 + }, + { + "epoch": 0.7024, + "grad_norm": 0.13486076891422272, + "learning_rate": 0.00011571134020618557, + "loss": 0.6041, + "step": 2195 + }, + { + "epoch": 0.70272, + "grad_norm": 0.14702895283699036, + "learning_rate": 0.00011567010309278352, + "loss": 0.7054, + "step": 2196 + }, + { + "epoch": 0.70304, + "grad_norm": 0.12575383484363556, + "learning_rate": 0.00011562886597938145, + "loss": 0.6147, + "step": 2197 + }, + { + "epoch": 0.70336, + "grad_norm": 0.14570002257823944, + "learning_rate": 0.00011558762886597938, + "loss": 0.637, + "step": 2198 + }, + { + "epoch": 0.70368, + "grad_norm": 0.1291791945695877, + "learning_rate": 0.00011554639175257733, + "loss": 0.8668, + "step": 2199 + }, + { + "epoch": 0.704, + "grad_norm": 0.16638752818107605, + "learning_rate": 0.00011550515463917526, + "loss": 0.6756, + "step": 2200 + }, + { + "epoch": 0.70432, + "grad_norm": 0.1422128975391388, + "learning_rate": 0.00011546391752577319, + "loss": 0.7454, + "step": 2201 + }, + { + "epoch": 0.70464, + "grad_norm": 0.15601807832717896, + "learning_rate": 0.00011542268041237115, + "loss": 0.6648, + "step": 2202 + }, + { + "epoch": 0.70496, + "grad_norm": 0.14389337599277496, + "learning_rate": 0.00011538144329896908, + "loss": 0.7125, + "step": 2203 + }, + { + "epoch": 0.70528, + "grad_norm": 0.14620384573936462, + "learning_rate": 0.00011534020618556702, + "loss": 0.7479, + "step": 2204 + }, + { + "epoch": 0.7056, + "grad_norm": 0.13690398633480072, + "learning_rate": 0.00011529896907216495, + "loss": 0.6567, + "step": 2205 + }, + { + "epoch": 0.70592, + "grad_norm": 0.13650929927825928, + "learning_rate": 0.00011525773195876289, + "loss": 0.7431, + "step": 2206 + }, + { + "epoch": 0.70624, + "grad_norm": 0.16303400695323944, + "learning_rate": 0.00011521649484536083, + "loss": 0.6476, + "step": 2207 + }, + { + "epoch": 0.70656, + "grad_norm": 0.15209178626537323, + "learning_rate": 0.00011517525773195876, + "loss": 0.7247, + "step": 2208 + }, + { + "epoch": 0.70688, + "grad_norm": 0.13735201954841614, + "learning_rate": 0.00011513402061855672, + "loss": 0.653, + "step": 2209 + }, + { + "epoch": 0.7072, + "grad_norm": 0.1269208937883377, + "learning_rate": 0.00011509278350515464, + "loss": 0.5854, + "step": 2210 + }, + { + "epoch": 0.70752, + "grad_norm": 0.1514381468296051, + "learning_rate": 0.00011505154639175257, + "loss": 0.7452, + "step": 2211 + }, + { + "epoch": 0.70784, + "grad_norm": 0.1655164211988449, + "learning_rate": 0.00011501030927835053, + "loss": 0.7522, + "step": 2212 + }, + { + "epoch": 0.70816, + "grad_norm": 0.11439188569784164, + "learning_rate": 0.00011496907216494846, + "loss": 0.7603, + "step": 2213 + }, + { + "epoch": 0.70848, + "grad_norm": 0.1402910351753235, + "learning_rate": 0.0001149278350515464, + "loss": 0.7819, + "step": 2214 + }, + { + "epoch": 0.7088, + "grad_norm": 0.15886886417865753, + "learning_rate": 0.00011488659793814433, + "loss": 0.6011, + "step": 2215 + }, + { + "epoch": 0.70912, + "grad_norm": 0.13617096841335297, + "learning_rate": 0.00011484536082474227, + "loss": 0.7467, + "step": 2216 + }, + { + "epoch": 0.70944, + "grad_norm": 0.1294909119606018, + "learning_rate": 0.00011480412371134021, + "loss": 0.5939, + "step": 2217 + }, + { + "epoch": 0.70976, + "grad_norm": 0.13047568500041962, + "learning_rate": 0.00011476288659793814, + "loss": 0.6882, + "step": 2218 + }, + { + "epoch": 0.71008, + "grad_norm": 0.13602742552757263, + "learning_rate": 0.0001147216494845361, + "loss": 0.7865, + "step": 2219 + }, + { + "epoch": 0.7104, + "grad_norm": 0.16178275644779205, + "learning_rate": 0.00011468041237113403, + "loss": 0.7375, + "step": 2220 + }, + { + "epoch": 0.71072, + "grad_norm": 0.1382664293050766, + "learning_rate": 0.00011463917525773196, + "loss": 0.8516, + "step": 2221 + }, + { + "epoch": 0.71104, + "grad_norm": 0.14488007128238678, + "learning_rate": 0.00011459793814432991, + "loss": 0.637, + "step": 2222 + }, + { + "epoch": 0.71136, + "grad_norm": 0.11767949908971786, + "learning_rate": 0.00011455670103092784, + "loss": 0.6785, + "step": 2223 + }, + { + "epoch": 0.71168, + "grad_norm": 0.14513443410396576, + "learning_rate": 0.00011451546391752578, + "loss": 0.6054, + "step": 2224 + }, + { + "epoch": 0.712, + "grad_norm": 0.12544426321983337, + "learning_rate": 0.00011447422680412372, + "loss": 0.7249, + "step": 2225 + }, + { + "epoch": 0.71232, + "grad_norm": 0.1444287896156311, + "learning_rate": 0.00011443298969072165, + "loss": 0.7373, + "step": 2226 + }, + { + "epoch": 0.71264, + "grad_norm": 0.1619909405708313, + "learning_rate": 0.0001143917525773196, + "loss": 0.7486, + "step": 2227 + }, + { + "epoch": 0.71296, + "grad_norm": 0.13678887486457825, + "learning_rate": 0.00011435051546391752, + "loss": 0.6513, + "step": 2228 + }, + { + "epoch": 0.71328, + "grad_norm": 0.13019737601280212, + "learning_rate": 0.00011430927835051545, + "loss": 0.6584, + "step": 2229 + }, + { + "epoch": 0.7136, + "grad_norm": 0.12649005651474, + "learning_rate": 0.00011426804123711341, + "loss": 0.7087, + "step": 2230 + }, + { + "epoch": 0.71392, + "grad_norm": 0.14586623013019562, + "learning_rate": 0.00011422680412371134, + "loss": 0.5534, + "step": 2231 + }, + { + "epoch": 0.71424, + "grad_norm": 0.14360938966274261, + "learning_rate": 0.00011418556701030929, + "loss": 0.6226, + "step": 2232 + }, + { + "epoch": 0.71456, + "grad_norm": 0.14981038868427277, + "learning_rate": 0.00011414432989690722, + "loss": 0.8909, + "step": 2233 + }, + { + "epoch": 0.71488, + "grad_norm": 0.15238483250141144, + "learning_rate": 0.00011410309278350515, + "loss": 0.6697, + "step": 2234 + }, + { + "epoch": 0.7152, + "grad_norm": 0.15770408511161804, + "learning_rate": 0.0001140618556701031, + "loss": 0.5397, + "step": 2235 + }, + { + "epoch": 0.71552, + "grad_norm": 0.11679151654243469, + "learning_rate": 0.00011402061855670103, + "loss": 0.5246, + "step": 2236 + }, + { + "epoch": 0.71584, + "grad_norm": 0.14550989866256714, + "learning_rate": 0.00011397938144329899, + "loss": 0.7921, + "step": 2237 + }, + { + "epoch": 0.71616, + "grad_norm": 0.14028450846672058, + "learning_rate": 0.00011393814432989692, + "loss": 0.5572, + "step": 2238 + }, + { + "epoch": 0.71648, + "grad_norm": 0.15086843073368073, + "learning_rate": 0.00011389690721649485, + "loss": 0.7394, + "step": 2239 + }, + { + "epoch": 0.7168, + "grad_norm": 0.14893896877765656, + "learning_rate": 0.00011385567010309279, + "loss": 0.7344, + "step": 2240 + }, + { + "epoch": 0.71712, + "grad_norm": 0.13001826405525208, + "learning_rate": 0.00011381443298969072, + "loss": 0.6694, + "step": 2241 + }, + { + "epoch": 0.71744, + "grad_norm": 0.15100553631782532, + "learning_rate": 0.00011377319587628867, + "loss": 0.7389, + "step": 2242 + }, + { + "epoch": 0.71776, + "grad_norm": 0.15597517788410187, + "learning_rate": 0.0001137319587628866, + "loss": 0.6547, + "step": 2243 + }, + { + "epoch": 0.71808, + "grad_norm": 0.1242024376988411, + "learning_rate": 0.00011369072164948453, + "loss": 0.637, + "step": 2244 + }, + { + "epoch": 0.7184, + "grad_norm": 0.14592300355434418, + "learning_rate": 0.00011364948453608249, + "loss": 0.748, + "step": 2245 + }, + { + "epoch": 0.71872, + "grad_norm": 0.17492128908634186, + "learning_rate": 0.00011360824742268041, + "loss": 0.6804, + "step": 2246 + }, + { + "epoch": 0.71904, + "grad_norm": 0.1370578110218048, + "learning_rate": 0.00011356701030927837, + "loss": 0.8033, + "step": 2247 + }, + { + "epoch": 0.71936, + "grad_norm": 0.12731322646141052, + "learning_rate": 0.0001135257731958763, + "loss": 0.5088, + "step": 2248 + }, + { + "epoch": 0.71968, + "grad_norm": 0.14562630653381348, + "learning_rate": 0.00011348453608247423, + "loss": 0.7133, + "step": 2249 + }, + { + "epoch": 0.72, + "grad_norm": 0.1422312706708908, + "learning_rate": 0.00011344329896907217, + "loss": 0.7384, + "step": 2250 + }, + { + "epoch": 0.72032, + "grad_norm": 0.15803846716880798, + "learning_rate": 0.0001134020618556701, + "loss": 0.7215, + "step": 2251 + }, + { + "epoch": 0.72064, + "grad_norm": 0.17728443443775177, + "learning_rate": 0.00011336082474226806, + "loss": 0.6665, + "step": 2252 + }, + { + "epoch": 0.72096, + "grad_norm": 0.1341867744922638, + "learning_rate": 0.00011331958762886598, + "loss": 0.897, + "step": 2253 + }, + { + "epoch": 0.72128, + "grad_norm": 0.15132027864456177, + "learning_rate": 0.00011327835051546391, + "loss": 0.7178, + "step": 2254 + }, + { + "epoch": 0.7216, + "grad_norm": 0.14291875064373016, + "learning_rate": 0.00011323711340206187, + "loss": 0.6607, + "step": 2255 + }, + { + "epoch": 0.72192, + "grad_norm": 0.13445764780044556, + "learning_rate": 0.0001131958762886598, + "loss": 0.7151, + "step": 2256 + }, + { + "epoch": 0.72224, + "grad_norm": 0.14409317076206207, + "learning_rate": 0.00011315463917525775, + "loss": 0.6631, + "step": 2257 + }, + { + "epoch": 0.72256, + "grad_norm": 0.12997932732105255, + "learning_rate": 0.00011311340206185568, + "loss": 0.6383, + "step": 2258 + }, + { + "epoch": 0.72288, + "grad_norm": 0.16306443512439728, + "learning_rate": 0.00011307216494845361, + "loss": 0.6761, + "step": 2259 + }, + { + "epoch": 0.7232, + "grad_norm": 0.13918858766555786, + "learning_rate": 0.00011303092783505155, + "loss": 0.6667, + "step": 2260 + }, + { + "epoch": 0.72352, + "grad_norm": 0.16144287586212158, + "learning_rate": 0.00011298969072164949, + "loss": 0.8291, + "step": 2261 + }, + { + "epoch": 0.72384, + "grad_norm": 0.1441497802734375, + "learning_rate": 0.00011294845360824742, + "loss": 0.7612, + "step": 2262 + }, + { + "epoch": 0.72416, + "grad_norm": 0.13145220279693604, + "learning_rate": 0.00011290721649484537, + "loss": 0.5559, + "step": 2263 + }, + { + "epoch": 0.72448, + "grad_norm": 0.13326162099838257, + "learning_rate": 0.00011286597938144329, + "loss": 0.6179, + "step": 2264 + }, + { + "epoch": 0.7248, + "grad_norm": 0.12555551528930664, + "learning_rate": 0.00011282474226804125, + "loss": 0.7364, + "step": 2265 + }, + { + "epoch": 0.72512, + "grad_norm": 0.15165653824806213, + "learning_rate": 0.00011278350515463918, + "loss": 0.5797, + "step": 2266 + }, + { + "epoch": 0.72544, + "grad_norm": 0.16068631410598755, + "learning_rate": 0.00011274226804123711, + "loss": 0.6195, + "step": 2267 + }, + { + "epoch": 0.72576, + "grad_norm": 0.14478428661823273, + "learning_rate": 0.00011270103092783506, + "loss": 0.6368, + "step": 2268 + }, + { + "epoch": 0.72608, + "grad_norm": 0.15544095635414124, + "learning_rate": 0.00011265979381443299, + "loss": 0.6089, + "step": 2269 + }, + { + "epoch": 0.7264, + "grad_norm": 0.12719611823558807, + "learning_rate": 0.00011261855670103095, + "loss": 0.5623, + "step": 2270 + }, + { + "epoch": 0.72672, + "grad_norm": 0.1499808430671692, + "learning_rate": 0.00011257731958762887, + "loss": 0.7311, + "step": 2271 + }, + { + "epoch": 0.72704, + "grad_norm": 0.15544234216213226, + "learning_rate": 0.0001125360824742268, + "loss": 0.4643, + "step": 2272 + }, + { + "epoch": 0.72736, + "grad_norm": 0.14273808896541595, + "learning_rate": 0.00011249484536082476, + "loss": 0.5513, + "step": 2273 + }, + { + "epoch": 0.72768, + "grad_norm": 0.16164454817771912, + "learning_rate": 0.00011245360824742269, + "loss": 0.795, + "step": 2274 + }, + { + "epoch": 0.728, + "grad_norm": 0.15521922707557678, + "learning_rate": 0.00011241237113402063, + "loss": 0.7152, + "step": 2275 + }, + { + "epoch": 0.72832, + "grad_norm": 0.12201813608407974, + "learning_rate": 0.00011237113402061856, + "loss": 0.7628, + "step": 2276 + }, + { + "epoch": 0.72864, + "grad_norm": 0.158720463514328, + "learning_rate": 0.0001123298969072165, + "loss": 0.6673, + "step": 2277 + }, + { + "epoch": 0.72896, + "grad_norm": 0.15062803030014038, + "learning_rate": 0.00011228865979381444, + "loss": 0.5737, + "step": 2278 + }, + { + "epoch": 0.72928, + "grad_norm": 0.1498451828956604, + "learning_rate": 0.00011224742268041237, + "loss": 0.794, + "step": 2279 + }, + { + "epoch": 0.7296, + "grad_norm": 0.14372706413269043, + "learning_rate": 0.00011220618556701033, + "loss": 0.7394, + "step": 2280 + }, + { + "epoch": 0.72992, + "grad_norm": 0.1313040852546692, + "learning_rate": 0.00011216494845360826, + "loss": 0.5308, + "step": 2281 + }, + { + "epoch": 0.73024, + "grad_norm": 0.13964460790157318, + "learning_rate": 0.00011212371134020618, + "loss": 0.7841, + "step": 2282 + }, + { + "epoch": 0.73056, + "grad_norm": 0.15472452342510223, + "learning_rate": 0.00011208247422680414, + "loss": 0.6291, + "step": 2283 + }, + { + "epoch": 0.73088, + "grad_norm": 0.16834712028503418, + "learning_rate": 0.00011204123711340207, + "loss": 0.7094, + "step": 2284 + }, + { + "epoch": 0.7312, + "grad_norm": 0.13699273765087128, + "learning_rate": 0.00011200000000000001, + "loss": 0.7415, + "step": 2285 + }, + { + "epoch": 0.73152, + "grad_norm": 0.15085478127002716, + "learning_rate": 0.00011195876288659794, + "loss": 0.6539, + "step": 2286 + }, + { + "epoch": 0.73184, + "grad_norm": 0.14628127217292786, + "learning_rate": 0.00011191752577319587, + "loss": 0.7225, + "step": 2287 + }, + { + "epoch": 0.73216, + "grad_norm": 0.1367931216955185, + "learning_rate": 0.00011187628865979383, + "loss": 0.5515, + "step": 2288 + }, + { + "epoch": 0.73248, + "grad_norm": 0.1428016573190689, + "learning_rate": 0.00011183505154639175, + "loss": 0.7495, + "step": 2289 + }, + { + "epoch": 0.7328, + "grad_norm": 0.1390329897403717, + "learning_rate": 0.00011179381443298968, + "loss": 0.6184, + "step": 2290 + }, + { + "epoch": 0.73312, + "grad_norm": 0.14085960388183594, + "learning_rate": 0.00011175257731958764, + "loss": 0.5038, + "step": 2291 + }, + { + "epoch": 0.73344, + "grad_norm": 0.1606917828321457, + "learning_rate": 0.00011171134020618557, + "loss": 0.6901, + "step": 2292 + }, + { + "epoch": 0.73376, + "grad_norm": 0.15191350877285004, + "learning_rate": 0.00011167010309278352, + "loss": 0.7707, + "step": 2293 + }, + { + "epoch": 0.73408, + "grad_norm": 0.15254057943820953, + "learning_rate": 0.00011162886597938145, + "loss": 0.578, + "step": 2294 + }, + { + "epoch": 0.7344, + "grad_norm": 0.14743052423000336, + "learning_rate": 0.00011158762886597938, + "loss": 0.5683, + "step": 2295 + }, + { + "epoch": 0.73472, + "grad_norm": 0.14751708507537842, + "learning_rate": 0.00011154639175257732, + "loss": 0.7253, + "step": 2296 + }, + { + "epoch": 0.73504, + "grad_norm": 0.13405734300613403, + "learning_rate": 0.00011150515463917525, + "loss": 0.7078, + "step": 2297 + }, + { + "epoch": 0.73536, + "grad_norm": 0.15192092955112457, + "learning_rate": 0.00011146391752577321, + "loss": 0.5661, + "step": 2298 + }, + { + "epoch": 0.73568, + "grad_norm": 0.16940361261367798, + "learning_rate": 0.00011142268041237114, + "loss": 0.6846, + "step": 2299 + }, + { + "epoch": 0.736, + "grad_norm": 0.13660922646522522, + "learning_rate": 0.00011138144329896906, + "loss": 0.7002, + "step": 2300 + }, + { + "epoch": 0.73632, + "grad_norm": 0.15135608613491058, + "learning_rate": 0.00011134020618556702, + "loss": 0.7794, + "step": 2301 + }, + { + "epoch": 0.73664, + "grad_norm": 0.1440742313861847, + "learning_rate": 0.00011129896907216495, + "loss": 0.5979, + "step": 2302 + }, + { + "epoch": 0.73696, + "grad_norm": 0.12972429394721985, + "learning_rate": 0.0001112577319587629, + "loss": 0.7557, + "step": 2303 + }, + { + "epoch": 0.73728, + "grad_norm": 0.1492079198360443, + "learning_rate": 0.00011121649484536083, + "loss": 0.654, + "step": 2304 + }, + { + "epoch": 0.7376, + "grad_norm": 0.14438387751579285, + "learning_rate": 0.00011117525773195876, + "loss": 0.6337, + "step": 2305 + }, + { + "epoch": 0.73792, + "grad_norm": 0.11794924736022949, + "learning_rate": 0.00011113402061855672, + "loss": 0.7572, + "step": 2306 + }, + { + "epoch": 0.73824, + "grad_norm": 0.1404753476381302, + "learning_rate": 0.00011109278350515464, + "loss": 0.7616, + "step": 2307 + }, + { + "epoch": 0.73856, + "grad_norm": 0.1644495278596878, + "learning_rate": 0.0001110515463917526, + "loss": 0.6669, + "step": 2308 + }, + { + "epoch": 0.73888, + "grad_norm": 0.14212843775749207, + "learning_rate": 0.00011101030927835053, + "loss": 0.8174, + "step": 2309 + }, + { + "epoch": 0.7392, + "grad_norm": 0.12656918168067932, + "learning_rate": 0.00011096907216494846, + "loss": 0.8361, + "step": 2310 + }, + { + "epoch": 0.73952, + "grad_norm": 0.13977016508579254, + "learning_rate": 0.0001109278350515464, + "loss": 0.6348, + "step": 2311 + }, + { + "epoch": 0.73984, + "grad_norm": 0.14475086331367493, + "learning_rate": 0.00011088659793814433, + "loss": 0.6885, + "step": 2312 + }, + { + "epoch": 0.74016, + "grad_norm": 0.164530411362648, + "learning_rate": 0.00011084536082474228, + "loss": 0.7909, + "step": 2313 + }, + { + "epoch": 0.74048, + "grad_norm": 0.15153254568576813, + "learning_rate": 0.00011080412371134021, + "loss": 0.6789, + "step": 2314 + }, + { + "epoch": 0.7408, + "grad_norm": 0.1412627249956131, + "learning_rate": 0.00011076288659793814, + "loss": 0.7338, + "step": 2315 + }, + { + "epoch": 0.74112, + "grad_norm": 0.137797012925148, + "learning_rate": 0.0001107216494845361, + "loss": 0.6821, + "step": 2316 + }, + { + "epoch": 0.74144, + "grad_norm": 0.13235370814800262, + "learning_rate": 0.00011068041237113403, + "loss": 0.7515, + "step": 2317 + }, + { + "epoch": 0.74176, + "grad_norm": 0.14515861868858337, + "learning_rate": 0.00011063917525773197, + "loss": 0.6453, + "step": 2318 + }, + { + "epoch": 0.74208, + "grad_norm": 0.15566951036453247, + "learning_rate": 0.0001105979381443299, + "loss": 0.6402, + "step": 2319 + }, + { + "epoch": 0.7424, + "grad_norm": 0.1529199182987213, + "learning_rate": 0.00011055670103092784, + "loss": 0.7535, + "step": 2320 + }, + { + "epoch": 0.74272, + "grad_norm": 0.13149023056030273, + "learning_rate": 0.00011051546391752578, + "loss": 0.857, + "step": 2321 + }, + { + "epoch": 0.74304, + "grad_norm": 0.14014273881912231, + "learning_rate": 0.00011047422680412371, + "loss": 0.8388, + "step": 2322 + }, + { + "epoch": 0.74336, + "grad_norm": 0.1424631029367447, + "learning_rate": 0.00011043298969072164, + "loss": 0.7625, + "step": 2323 + }, + { + "epoch": 0.74368, + "grad_norm": 0.1402021050453186, + "learning_rate": 0.0001103917525773196, + "loss": 0.8199, + "step": 2324 + }, + { + "epoch": 0.744, + "grad_norm": 0.15589484572410583, + "learning_rate": 0.00011035051546391752, + "loss": 0.686, + "step": 2325 + }, + { + "epoch": 0.74432, + "grad_norm": 0.144229918718338, + "learning_rate": 0.00011030927835051548, + "loss": 0.7043, + "step": 2326 + }, + { + "epoch": 0.74464, + "grad_norm": 0.15780605375766754, + "learning_rate": 0.00011026804123711341, + "loss": 0.7132, + "step": 2327 + }, + { + "epoch": 0.74496, + "grad_norm": 0.15335465967655182, + "learning_rate": 0.00011022680412371134, + "loss": 0.5854, + "step": 2328 + }, + { + "epoch": 0.74528, + "grad_norm": 0.12628668546676636, + "learning_rate": 0.00011018556701030929, + "loss": 0.5966, + "step": 2329 + }, + { + "epoch": 0.7456, + "grad_norm": 0.14625613391399384, + "learning_rate": 0.00011014432989690722, + "loss": 0.6822, + "step": 2330 + }, + { + "epoch": 0.74592, + "grad_norm": 0.1358364075422287, + "learning_rate": 0.00011010309278350516, + "loss": 0.7885, + "step": 2331 + }, + { + "epoch": 0.74624, + "grad_norm": 0.13193346560001373, + "learning_rate": 0.0001100618556701031, + "loss": 0.7022, + "step": 2332 + }, + { + "epoch": 0.74656, + "grad_norm": 0.14863784611225128, + "learning_rate": 0.00011002061855670102, + "loss": 0.6399, + "step": 2333 + }, + { + "epoch": 0.74688, + "grad_norm": 0.14190231263637543, + "learning_rate": 0.00010997938144329898, + "loss": 0.5172, + "step": 2334 + }, + { + "epoch": 0.7472, + "grad_norm": 0.12917064130306244, + "learning_rate": 0.00010993814432989691, + "loss": 0.6897, + "step": 2335 + }, + { + "epoch": 0.74752, + "grad_norm": 0.15621407330036163, + "learning_rate": 0.00010989690721649486, + "loss": 0.7316, + "step": 2336 + }, + { + "epoch": 0.74784, + "grad_norm": 0.14652156829833984, + "learning_rate": 0.00010985567010309279, + "loss": 0.6385, + "step": 2337 + }, + { + "epoch": 0.74816, + "grad_norm": 0.1458377093076706, + "learning_rate": 0.00010981443298969072, + "loss": 0.7182, + "step": 2338 + }, + { + "epoch": 0.74848, + "grad_norm": 0.14937365055084229, + "learning_rate": 0.00010977319587628867, + "loss": 0.5324, + "step": 2339 + }, + { + "epoch": 0.7488, + "grad_norm": 0.15816596150398254, + "learning_rate": 0.0001097319587628866, + "loss": 0.7443, + "step": 2340 + }, + { + "epoch": 0.74912, + "grad_norm": 0.12018290907144547, + "learning_rate": 0.00010969072164948456, + "loss": 0.5157, + "step": 2341 + }, + { + "epoch": 0.74944, + "grad_norm": 0.1299060881137848, + "learning_rate": 0.00010964948453608247, + "loss": 0.6367, + "step": 2342 + }, + { + "epoch": 0.74976, + "grad_norm": 0.1438489556312561, + "learning_rate": 0.0001096082474226804, + "loss": 0.6601, + "step": 2343 + }, + { + "epoch": 0.75008, + "grad_norm": 0.17499497532844543, + "learning_rate": 0.00010956701030927836, + "loss": 0.73, + "step": 2344 + }, + { + "epoch": 0.7504, + "grad_norm": 0.13686110079288483, + "learning_rate": 0.0001095257731958763, + "loss": 0.8438, + "step": 2345 + }, + { + "epoch": 0.75072, + "grad_norm": 0.13297463953495026, + "learning_rate": 0.00010948453608247424, + "loss": 0.8016, + "step": 2346 + }, + { + "epoch": 0.75104, + "grad_norm": 0.14051952958106995, + "learning_rate": 0.00010944329896907217, + "loss": 0.793, + "step": 2347 + }, + { + "epoch": 0.75136, + "grad_norm": 0.14615367352962494, + "learning_rate": 0.0001094020618556701, + "loss": 0.6921, + "step": 2348 + }, + { + "epoch": 0.75168, + "grad_norm": 0.1447516828775406, + "learning_rate": 0.00010936082474226805, + "loss": 0.6656, + "step": 2349 + }, + { + "epoch": 0.752, + "grad_norm": 0.15158802270889282, + "learning_rate": 0.00010931958762886598, + "loss": 0.6885, + "step": 2350 + }, + { + "epoch": 0.75232, + "grad_norm": 0.14549511671066284, + "learning_rate": 0.00010927835051546391, + "loss": 0.6203, + "step": 2351 + }, + { + "epoch": 0.75264, + "grad_norm": 0.15099123120307922, + "learning_rate": 0.00010923711340206187, + "loss": 0.7163, + "step": 2352 + }, + { + "epoch": 0.75296, + "grad_norm": 0.1343199461698532, + "learning_rate": 0.0001091958762886598, + "loss": 0.6036, + "step": 2353 + }, + { + "epoch": 0.75328, + "grad_norm": 0.1767696887254715, + "learning_rate": 0.00010915463917525774, + "loss": 0.6297, + "step": 2354 + }, + { + "epoch": 0.7536, + "grad_norm": 0.1462734043598175, + "learning_rate": 0.00010911340206185568, + "loss": 0.7097, + "step": 2355 + }, + { + "epoch": 0.75392, + "grad_norm": 0.1527024656534195, + "learning_rate": 0.0001090721649484536, + "loss": 0.8247, + "step": 2356 + }, + { + "epoch": 0.75424, + "grad_norm": 0.15044885873794556, + "learning_rate": 0.00010903092783505155, + "loss": 0.6832, + "step": 2357 + }, + { + "epoch": 0.75456, + "grad_norm": 0.12089329957962036, + "learning_rate": 0.00010898969072164948, + "loss": 0.6399, + "step": 2358 + }, + { + "epoch": 0.75488, + "grad_norm": 0.13430249691009521, + "learning_rate": 0.00010894845360824744, + "loss": 0.5749, + "step": 2359 + }, + { + "epoch": 0.7552, + "grad_norm": 0.15970252454280853, + "learning_rate": 0.00010890721649484536, + "loss": 0.7148, + "step": 2360 + }, + { + "epoch": 0.75552, + "grad_norm": 0.14559680223464966, + "learning_rate": 0.00010886597938144329, + "loss": 0.7018, + "step": 2361 + }, + { + "epoch": 0.75584, + "grad_norm": 0.14394278824329376, + "learning_rate": 0.00010882474226804125, + "loss": 0.7519, + "step": 2362 + }, + { + "epoch": 0.75616, + "grad_norm": 0.1382453590631485, + "learning_rate": 0.00010878350515463918, + "loss": 0.7233, + "step": 2363 + }, + { + "epoch": 0.75648, + "grad_norm": 0.13959437608718872, + "learning_rate": 0.00010874226804123712, + "loss": 0.6875, + "step": 2364 + }, + { + "epoch": 0.7568, + "grad_norm": 0.1402619630098343, + "learning_rate": 0.00010870103092783506, + "loss": 0.6634, + "step": 2365 + }, + { + "epoch": 0.75712, + "grad_norm": 0.13670587539672852, + "learning_rate": 0.00010865979381443299, + "loss": 0.7574, + "step": 2366 + }, + { + "epoch": 0.75744, + "grad_norm": 0.129637211561203, + "learning_rate": 0.00010861855670103093, + "loss": 0.7029, + "step": 2367 + }, + { + "epoch": 0.75776, + "grad_norm": 0.1394963413476944, + "learning_rate": 0.00010857731958762886, + "loss": 0.6374, + "step": 2368 + }, + { + "epoch": 0.75808, + "grad_norm": 0.1414107233285904, + "learning_rate": 0.00010853608247422682, + "loss": 0.711, + "step": 2369 + }, + { + "epoch": 0.7584, + "grad_norm": 0.1316916048526764, + "learning_rate": 0.00010849484536082475, + "loss": 0.7014, + "step": 2370 + }, + { + "epoch": 0.75872, + "grad_norm": 0.1269483119249344, + "learning_rate": 0.00010845360824742268, + "loss": 0.6515, + "step": 2371 + }, + { + "epoch": 0.75904, + "grad_norm": 0.1407054364681244, + "learning_rate": 0.00010841237113402063, + "loss": 0.6385, + "step": 2372 + }, + { + "epoch": 0.75936, + "grad_norm": 0.1404874622821808, + "learning_rate": 0.00010837113402061856, + "loss": 0.8714, + "step": 2373 + }, + { + "epoch": 0.75968, + "grad_norm": 0.1649414300918579, + "learning_rate": 0.0001083298969072165, + "loss": 0.7033, + "step": 2374 + }, + { + "epoch": 0.76, + "grad_norm": 0.1498667150735855, + "learning_rate": 0.00010828865979381444, + "loss": 0.5976, + "step": 2375 + }, + { + "epoch": 0.76032, + "grad_norm": 0.14429685473442078, + "learning_rate": 0.00010824742268041237, + "loss": 0.7536, + "step": 2376 + }, + { + "epoch": 0.76064, + "grad_norm": 0.1346881240606308, + "learning_rate": 0.00010820618556701033, + "loss": 0.6769, + "step": 2377 + }, + { + "epoch": 0.76096, + "grad_norm": 0.1487450897693634, + "learning_rate": 0.00010816494845360824, + "loss": 0.7446, + "step": 2378 + }, + { + "epoch": 0.76128, + "grad_norm": 0.14742706716060638, + "learning_rate": 0.00010812371134020618, + "loss": 0.7797, + "step": 2379 + }, + { + "epoch": 0.7616, + "grad_norm": 0.14181749522686005, + "learning_rate": 0.00010808247422680413, + "loss": 0.6323, + "step": 2380 + }, + { + "epoch": 0.76192, + "grad_norm": 0.15597188472747803, + "learning_rate": 0.00010804123711340206, + "loss": 0.678, + "step": 2381 + }, + { + "epoch": 0.76224, + "grad_norm": 0.14637978374958038, + "learning_rate": 0.00010800000000000001, + "loss": 0.7357, + "step": 2382 + }, + { + "epoch": 0.76256, + "grad_norm": 0.17027409374713898, + "learning_rate": 0.00010795876288659794, + "loss": 0.8047, + "step": 2383 + }, + { + "epoch": 0.76288, + "grad_norm": 0.12585978209972382, + "learning_rate": 0.00010791752577319587, + "loss": 0.7793, + "step": 2384 + }, + { + "epoch": 0.7632, + "grad_norm": 0.13518136739730835, + "learning_rate": 0.00010787628865979382, + "loss": 0.4938, + "step": 2385 + }, + { + "epoch": 0.76352, + "grad_norm": 0.1401270627975464, + "learning_rate": 0.00010783505154639175, + "loss": 0.6674, + "step": 2386 + }, + { + "epoch": 0.76384, + "grad_norm": 0.17298191785812378, + "learning_rate": 0.0001077938144329897, + "loss": 0.6554, + "step": 2387 + }, + { + "epoch": 0.76416, + "grad_norm": 0.16476158797740936, + "learning_rate": 0.00010775257731958764, + "loss": 0.65, + "step": 2388 + }, + { + "epoch": 0.76448, + "grad_norm": 0.15369954705238342, + "learning_rate": 0.00010771134020618557, + "loss": 0.6063, + "step": 2389 + }, + { + "epoch": 0.7648, + "grad_norm": 0.14306509494781494, + "learning_rate": 0.00010767010309278351, + "loss": 0.5195, + "step": 2390 + }, + { + "epoch": 0.76512, + "grad_norm": 0.1348339170217514, + "learning_rate": 0.00010762886597938145, + "loss": 0.6566, + "step": 2391 + }, + { + "epoch": 0.76544, + "grad_norm": 0.1437949389219284, + "learning_rate": 0.00010758762886597939, + "loss": 0.6813, + "step": 2392 + }, + { + "epoch": 0.76576, + "grad_norm": 0.13925068080425262, + "learning_rate": 0.00010754639175257732, + "loss": 0.6833, + "step": 2393 + }, + { + "epoch": 0.76608, + "grad_norm": 0.14318601787090302, + "learning_rate": 0.00010750515463917525, + "loss": 0.5984, + "step": 2394 + }, + { + "epoch": 0.7664, + "grad_norm": 0.12551507353782654, + "learning_rate": 0.00010746391752577321, + "loss": 0.694, + "step": 2395 + }, + { + "epoch": 0.76672, + "grad_norm": 0.14767882227897644, + "learning_rate": 0.00010742268041237113, + "loss": 0.8011, + "step": 2396 + }, + { + "epoch": 0.76704, + "grad_norm": 0.15946391224861145, + "learning_rate": 0.00010738144329896909, + "loss": 0.6767, + "step": 2397 + }, + { + "epoch": 0.76736, + "grad_norm": 0.14214830100536346, + "learning_rate": 0.00010734020618556702, + "loss": 0.8293, + "step": 2398 + }, + { + "epoch": 0.76768, + "grad_norm": 0.12547197937965393, + "learning_rate": 0.00010729896907216495, + "loss": 0.6726, + "step": 2399 + }, + { + "epoch": 0.768, + "grad_norm": 0.13485264778137207, + "learning_rate": 0.0001072577319587629, + "loss": 0.5848, + "step": 2400 + }, + { + "epoch": 0.76832, + "grad_norm": 0.16473545134067535, + "learning_rate": 0.00010721649484536083, + "loss": 0.8307, + "step": 2401 + }, + { + "epoch": 0.76864, + "grad_norm": 0.14048965275287628, + "learning_rate": 0.00010717525773195878, + "loss": 0.5982, + "step": 2402 + }, + { + "epoch": 0.76896, + "grad_norm": 0.15585550665855408, + "learning_rate": 0.0001071340206185567, + "loss": 0.5772, + "step": 2403 + }, + { + "epoch": 0.76928, + "grad_norm": 0.12860015034675598, + "learning_rate": 0.00010709278350515463, + "loss": 0.5183, + "step": 2404 + }, + { + "epoch": 0.7696, + "grad_norm": 0.15549108386039734, + "learning_rate": 0.00010705154639175259, + "loss": 0.7392, + "step": 2405 + }, + { + "epoch": 0.76992, + "grad_norm": 0.13876163959503174, + "learning_rate": 0.00010701030927835052, + "loss": 0.6674, + "step": 2406 + }, + { + "epoch": 0.77024, + "grad_norm": 0.1360311508178711, + "learning_rate": 0.00010696907216494847, + "loss": 0.7174, + "step": 2407 + }, + { + "epoch": 0.77056, + "grad_norm": 0.13627313077449799, + "learning_rate": 0.0001069278350515464, + "loss": 0.8186, + "step": 2408 + }, + { + "epoch": 0.77088, + "grad_norm": 0.13149705529212952, + "learning_rate": 0.00010688659793814433, + "loss": 0.5992, + "step": 2409 + }, + { + "epoch": 0.7712, + "grad_norm": 0.14151443541049957, + "learning_rate": 0.00010684536082474227, + "loss": 0.5674, + "step": 2410 + }, + { + "epoch": 0.77152, + "grad_norm": 0.13653911650180817, + "learning_rate": 0.0001068041237113402, + "loss": 0.6866, + "step": 2411 + }, + { + "epoch": 0.77184, + "grad_norm": 0.1367635577917099, + "learning_rate": 0.00010676288659793814, + "loss": 0.5893, + "step": 2412 + }, + { + "epoch": 0.77216, + "grad_norm": 0.14945204555988312, + "learning_rate": 0.0001067216494845361, + "loss": 0.8111, + "step": 2413 + }, + { + "epoch": 0.77248, + "grad_norm": 0.13530069589614868, + "learning_rate": 0.00010668041237113401, + "loss": 0.6299, + "step": 2414 + }, + { + "epoch": 0.7728, + "grad_norm": 0.1549733430147171, + "learning_rate": 0.00010663917525773197, + "loss": 0.8494, + "step": 2415 + }, + { + "epoch": 0.77312, + "grad_norm": 0.1442151665687561, + "learning_rate": 0.0001065979381443299, + "loss": 0.5988, + "step": 2416 + }, + { + "epoch": 0.77344, + "grad_norm": 0.14296896755695343, + "learning_rate": 0.00010655670103092783, + "loss": 0.6425, + "step": 2417 + }, + { + "epoch": 0.77376, + "grad_norm": 0.13524125516414642, + "learning_rate": 0.00010651546391752578, + "loss": 0.6928, + "step": 2418 + }, + { + "epoch": 0.77408, + "grad_norm": 0.16157269477844238, + "learning_rate": 0.00010647422680412371, + "loss": 0.6892, + "step": 2419 + }, + { + "epoch": 0.7744, + "grad_norm": 0.1271234154701233, + "learning_rate": 0.00010643298969072167, + "loss": 0.6437, + "step": 2420 + }, + { + "epoch": 0.77472, + "grad_norm": 0.15270353853702545, + "learning_rate": 0.00010639175257731959, + "loss": 0.7292, + "step": 2421 + }, + { + "epoch": 0.77504, + "grad_norm": 0.13716702163219452, + "learning_rate": 0.00010635051546391752, + "loss": 0.5941, + "step": 2422 + }, + { + "epoch": 0.77536, + "grad_norm": 0.14884956181049347, + "learning_rate": 0.00010630927835051548, + "loss": 0.6834, + "step": 2423 + }, + { + "epoch": 0.77568, + "grad_norm": 0.13044248521327972, + "learning_rate": 0.00010626804123711341, + "loss": 0.7029, + "step": 2424 + }, + { + "epoch": 0.776, + "grad_norm": 0.13581915199756622, + "learning_rate": 0.00010622680412371135, + "loss": 0.7878, + "step": 2425 + }, + { + "epoch": 0.77632, + "grad_norm": 0.15196485817432404, + "learning_rate": 0.00010618556701030928, + "loss": 0.6267, + "step": 2426 + }, + { + "epoch": 0.77664, + "grad_norm": 0.159225195646286, + "learning_rate": 0.00010614432989690721, + "loss": 0.6357, + "step": 2427 + }, + { + "epoch": 0.77696, + "grad_norm": 0.13871781527996063, + "learning_rate": 0.00010610309278350516, + "loss": 0.7349, + "step": 2428 + }, + { + "epoch": 0.77728, + "grad_norm": 0.15360791981220245, + "learning_rate": 0.00010606185567010309, + "loss": 0.6125, + "step": 2429 + }, + { + "epoch": 0.7776, + "grad_norm": 0.17124877870082855, + "learning_rate": 0.00010602061855670105, + "loss": 0.7045, + "step": 2430 + }, + { + "epoch": 0.77792, + "grad_norm": 0.15152905881404877, + "learning_rate": 0.00010597938144329898, + "loss": 0.751, + "step": 2431 + }, + { + "epoch": 0.77824, + "grad_norm": 0.16107770800590515, + "learning_rate": 0.0001059381443298969, + "loss": 0.6475, + "step": 2432 + }, + { + "epoch": 0.77856, + "grad_norm": 0.1534031182527542, + "learning_rate": 0.00010589690721649486, + "loss": 0.7319, + "step": 2433 + }, + { + "epoch": 0.77888, + "grad_norm": 0.15022999048233032, + "learning_rate": 0.00010585567010309279, + "loss": 0.7894, + "step": 2434 + }, + { + "epoch": 0.7792, + "grad_norm": 0.13740140199661255, + "learning_rate": 0.00010581443298969073, + "loss": 0.6024, + "step": 2435 + }, + { + "epoch": 0.77952, + "grad_norm": 0.15208189189434052, + "learning_rate": 0.00010577319587628866, + "loss": 0.7979, + "step": 2436 + }, + { + "epoch": 0.77984, + "grad_norm": 0.15616613626480103, + "learning_rate": 0.0001057319587628866, + "loss": 0.7916, + "step": 2437 + }, + { + "epoch": 0.78016, + "grad_norm": 0.15824349224567413, + "learning_rate": 0.00010569072164948455, + "loss": 0.616, + "step": 2438 + }, + { + "epoch": 0.78048, + "grad_norm": 0.1351231336593628, + "learning_rate": 0.00010564948453608247, + "loss": 0.6497, + "step": 2439 + }, + { + "epoch": 0.7808, + "grad_norm": 0.12585292756557465, + "learning_rate": 0.0001056082474226804, + "loss": 0.7402, + "step": 2440 + }, + { + "epoch": 0.78112, + "grad_norm": 0.14339880645275116, + "learning_rate": 0.00010556701030927836, + "loss": 0.7663, + "step": 2441 + }, + { + "epoch": 0.78144, + "grad_norm": 0.14296932518482208, + "learning_rate": 0.00010552577319587629, + "loss": 0.8375, + "step": 2442 + }, + { + "epoch": 0.78176, + "grad_norm": 0.16566330194473267, + "learning_rate": 0.00010548453608247424, + "loss": 0.6668, + "step": 2443 + }, + { + "epoch": 0.78208, + "grad_norm": 0.13261167705059052, + "learning_rate": 0.00010544329896907217, + "loss": 0.8584, + "step": 2444 + }, + { + "epoch": 0.7824, + "grad_norm": 0.13362634181976318, + "learning_rate": 0.0001054020618556701, + "loss": 0.7592, + "step": 2445 + }, + { + "epoch": 0.78272, + "grad_norm": 0.14439146220684052, + "learning_rate": 0.00010536082474226804, + "loss": 0.6763, + "step": 2446 + }, + { + "epoch": 0.78304, + "grad_norm": 0.14779646694660187, + "learning_rate": 0.00010531958762886598, + "loss": 0.6069, + "step": 2447 + }, + { + "epoch": 0.78336, + "grad_norm": 0.14571484923362732, + "learning_rate": 0.00010527835051546393, + "loss": 0.535, + "step": 2448 + }, + { + "epoch": 0.78368, + "grad_norm": 0.14461350440979004, + "learning_rate": 0.00010523711340206187, + "loss": 0.8809, + "step": 2449 + }, + { + "epoch": 0.784, + "grad_norm": 0.16604378819465637, + "learning_rate": 0.00010519587628865978, + "loss": 0.7368, + "step": 2450 + }, + { + "epoch": 0.78432, + "grad_norm": 0.1877472996711731, + "learning_rate": 0.00010515463917525774, + "loss": 0.8126, + "step": 2451 + }, + { + "epoch": 0.78464, + "grad_norm": 0.1255762279033661, + "learning_rate": 0.00010511340206185567, + "loss": 0.7672, + "step": 2452 + }, + { + "epoch": 0.78496, + "grad_norm": 0.11469117552042007, + "learning_rate": 0.00010507216494845362, + "loss": 0.6687, + "step": 2453 + }, + { + "epoch": 0.78528, + "grad_norm": 0.1558726727962494, + "learning_rate": 0.00010503092783505155, + "loss": 0.6673, + "step": 2454 + }, + { + "epoch": 0.7856, + "grad_norm": 0.12309260666370392, + "learning_rate": 0.00010498969072164948, + "loss": 0.5958, + "step": 2455 + }, + { + "epoch": 0.78592, + "grad_norm": 0.15069130063056946, + "learning_rate": 0.00010494845360824744, + "loss": 0.6273, + "step": 2456 + }, + { + "epoch": 0.78624, + "grad_norm": 0.15864059329032898, + "learning_rate": 0.00010490721649484536, + "loss": 0.6465, + "step": 2457 + }, + { + "epoch": 0.78656, + "grad_norm": 0.15379779040813446, + "learning_rate": 0.00010486597938144331, + "loss": 0.5771, + "step": 2458 + }, + { + "epoch": 0.78688, + "grad_norm": 0.14139831066131592, + "learning_rate": 0.00010482474226804125, + "loss": 0.7836, + "step": 2459 + }, + { + "epoch": 0.7872, + "grad_norm": 0.14854180812835693, + "learning_rate": 0.00010478350515463918, + "loss": 0.65, + "step": 2460 + }, + { + "epoch": 0.78752, + "grad_norm": 0.14634421467781067, + "learning_rate": 0.00010474226804123712, + "loss": 0.6568, + "step": 2461 + }, + { + "epoch": 0.78784, + "grad_norm": 0.14304807782173157, + "learning_rate": 0.00010470103092783505, + "loss": 0.7057, + "step": 2462 + }, + { + "epoch": 0.78816, + "grad_norm": 0.13940250873565674, + "learning_rate": 0.000104659793814433, + "loss": 0.7555, + "step": 2463 + }, + { + "epoch": 0.78848, + "grad_norm": 0.15564656257629395, + "learning_rate": 0.00010461855670103093, + "loss": 0.7178, + "step": 2464 + }, + { + "epoch": 0.7888, + "grad_norm": 0.13384011387825012, + "learning_rate": 0.00010457731958762886, + "loss": 0.756, + "step": 2465 + }, + { + "epoch": 0.78912, + "grad_norm": 0.15792788565158844, + "learning_rate": 0.00010453608247422682, + "loss": 0.6656, + "step": 2466 + }, + { + "epoch": 0.78944, + "grad_norm": 0.1439536213874817, + "learning_rate": 0.00010449484536082475, + "loss": 0.7431, + "step": 2467 + }, + { + "epoch": 0.78976, + "grad_norm": 0.1517084836959839, + "learning_rate": 0.0001044536082474227, + "loss": 0.6925, + "step": 2468 + }, + { + "epoch": 0.79008, + "grad_norm": 0.1691693514585495, + "learning_rate": 0.00010441237113402063, + "loss": 0.7437, + "step": 2469 + }, + { + "epoch": 0.7904, + "grad_norm": 0.13419830799102783, + "learning_rate": 0.00010437113402061856, + "loss": 0.5462, + "step": 2470 + }, + { + "epoch": 0.79072, + "grad_norm": 0.13424335420131683, + "learning_rate": 0.0001043298969072165, + "loss": 0.7143, + "step": 2471 + }, + { + "epoch": 0.79104, + "grad_norm": 0.16811063885688782, + "learning_rate": 0.00010428865979381443, + "loss": 0.7323, + "step": 2472 + }, + { + "epoch": 0.79136, + "grad_norm": 0.14126327633857727, + "learning_rate": 0.00010424742268041237, + "loss": 0.6691, + "step": 2473 + }, + { + "epoch": 0.79168, + "grad_norm": 0.1655503362417221, + "learning_rate": 0.00010420618556701032, + "loss": 0.7112, + "step": 2474 + }, + { + "epoch": 0.792, + "grad_norm": 0.16073282063007355, + "learning_rate": 0.00010416494845360824, + "loss": 0.707, + "step": 2475 + }, + { + "epoch": 0.79232, + "grad_norm": 0.12400400638580322, + "learning_rate": 0.0001041237113402062, + "loss": 0.7682, + "step": 2476 + }, + { + "epoch": 0.79264, + "grad_norm": 0.16412343084812164, + "learning_rate": 0.00010408247422680413, + "loss": 0.714, + "step": 2477 + }, + { + "epoch": 0.79296, + "grad_norm": 0.1472770869731903, + "learning_rate": 0.00010404123711340206, + "loss": 0.6269, + "step": 2478 + }, + { + "epoch": 0.79328, + "grad_norm": 0.1521236002445221, + "learning_rate": 0.00010400000000000001, + "loss": 0.7731, + "step": 2479 + }, + { + "epoch": 0.7936, + "grad_norm": 0.14874829351902008, + "learning_rate": 0.00010395876288659794, + "loss": 0.7051, + "step": 2480 + }, + { + "epoch": 0.79392, + "grad_norm": 0.1498081088066101, + "learning_rate": 0.00010391752577319588, + "loss": 0.6799, + "step": 2481 + }, + { + "epoch": 0.79424, + "grad_norm": 0.15765486657619476, + "learning_rate": 0.00010387628865979381, + "loss": 0.7266, + "step": 2482 + }, + { + "epoch": 0.79456, + "grad_norm": 0.1501636952161789, + "learning_rate": 0.00010383505154639175, + "loss": 0.6243, + "step": 2483 + }, + { + "epoch": 0.79488, + "grad_norm": 0.14952801167964935, + "learning_rate": 0.0001037938144329897, + "loss": 0.644, + "step": 2484 + }, + { + "epoch": 0.7952, + "grad_norm": 0.1611526608467102, + "learning_rate": 0.00010375257731958764, + "loss": 0.6239, + "step": 2485 + }, + { + "epoch": 0.79552, + "grad_norm": 0.15747351944446564, + "learning_rate": 0.00010371134020618558, + "loss": 0.5894, + "step": 2486 + }, + { + "epoch": 0.79584, + "grad_norm": 0.1683342456817627, + "learning_rate": 0.00010367010309278351, + "loss": 0.603, + "step": 2487 + }, + { + "epoch": 0.79616, + "grad_norm": 0.15063390135765076, + "learning_rate": 0.00010362886597938144, + "loss": 0.5963, + "step": 2488 + }, + { + "epoch": 0.79648, + "grad_norm": 0.15094248950481415, + "learning_rate": 0.00010358762886597939, + "loss": 0.653, + "step": 2489 + }, + { + "epoch": 0.7968, + "grad_norm": 0.14150899648666382, + "learning_rate": 0.00010354639175257732, + "loss": 0.6828, + "step": 2490 + }, + { + "epoch": 0.79712, + "grad_norm": 0.14421166479587555, + "learning_rate": 0.00010350515463917528, + "loss": 0.6399, + "step": 2491 + }, + { + "epoch": 0.79744, + "grad_norm": 0.1364155411720276, + "learning_rate": 0.00010346391752577321, + "loss": 0.7633, + "step": 2492 + }, + { + "epoch": 0.79776, + "grad_norm": 0.1619148850440979, + "learning_rate": 0.00010342268041237113, + "loss": 0.7217, + "step": 2493 + }, + { + "epoch": 0.79808, + "grad_norm": 0.15138627588748932, + "learning_rate": 0.00010338144329896908, + "loss": 0.6985, + "step": 2494 + }, + { + "epoch": 0.7984, + "grad_norm": 0.15581676363945007, + "learning_rate": 0.00010334020618556702, + "loss": 0.6655, + "step": 2495 + }, + { + "epoch": 0.79872, + "grad_norm": 0.17176972329616547, + "learning_rate": 0.00010329896907216496, + "loss": 0.7634, + "step": 2496 + }, + { + "epoch": 0.79904, + "grad_norm": 0.14661046862602234, + "learning_rate": 0.00010325773195876289, + "loss": 0.7091, + "step": 2497 + }, + { + "epoch": 0.79936, + "grad_norm": 0.1558842957019806, + "learning_rate": 0.00010321649484536082, + "loss": 0.7343, + "step": 2498 + }, + { + "epoch": 0.79968, + "grad_norm": 0.13796402513980865, + "learning_rate": 0.00010317525773195877, + "loss": 0.5937, + "step": 2499 + }, + { + "epoch": 0.8, + "grad_norm": 0.13113626837730408, + "learning_rate": 0.0001031340206185567, + "loss": 0.7336, + "step": 2500 + }, + { + "epoch": 0.80032, + "grad_norm": 0.16652272641658783, + "learning_rate": 0.00010309278350515463, + "loss": 0.7065, + "step": 2501 + }, + { + "epoch": 0.80064, + "grad_norm": 0.1407938301563263, + "learning_rate": 0.00010305154639175259, + "loss": 0.5984, + "step": 2502 + }, + { + "epoch": 0.80096, + "grad_norm": 0.13912548124790192, + "learning_rate": 0.00010301030927835052, + "loss": 0.7336, + "step": 2503 + }, + { + "epoch": 0.80128, + "grad_norm": 0.14547914266586304, + "learning_rate": 0.00010296907216494847, + "loss": 0.5667, + "step": 2504 + }, + { + "epoch": 0.8016, + "grad_norm": 0.14989931881427765, + "learning_rate": 0.0001029278350515464, + "loss": 0.6701, + "step": 2505 + }, + { + "epoch": 0.80192, + "grad_norm": 0.13537397980690002, + "learning_rate": 0.00010288659793814433, + "loss": 0.736, + "step": 2506 + }, + { + "epoch": 0.80224, + "grad_norm": 0.1657114326953888, + "learning_rate": 0.00010284536082474227, + "loss": 0.662, + "step": 2507 + }, + { + "epoch": 0.80256, + "grad_norm": 0.1416400969028473, + "learning_rate": 0.0001028041237113402, + "loss": 0.6694, + "step": 2508 + }, + { + "epoch": 0.80288, + "grad_norm": 0.1446150243282318, + "learning_rate": 0.00010276288659793816, + "loss": 0.6481, + "step": 2509 + }, + { + "epoch": 0.8032, + "grad_norm": 0.14864562451839447, + "learning_rate": 0.00010272164948453608, + "loss": 0.609, + "step": 2510 + }, + { + "epoch": 0.80352, + "grad_norm": 0.13135448098182678, + "learning_rate": 0.00010268041237113401, + "loss": 0.6172, + "step": 2511 + }, + { + "epoch": 0.80384, + "grad_norm": 0.14855985343456268, + "learning_rate": 0.00010263917525773197, + "loss": 0.6292, + "step": 2512 + }, + { + "epoch": 0.80416, + "grad_norm": 0.14971700310707092, + "learning_rate": 0.0001025979381443299, + "loss": 0.5729, + "step": 2513 + }, + { + "epoch": 0.80448, + "grad_norm": 0.12926143407821655, + "learning_rate": 0.00010255670103092785, + "loss": 0.668, + "step": 2514 + }, + { + "epoch": 0.8048, + "grad_norm": 0.14829619228839874, + "learning_rate": 0.00010251546391752578, + "loss": 0.8166, + "step": 2515 + }, + { + "epoch": 0.80512, + "grad_norm": 0.14759129285812378, + "learning_rate": 0.00010247422680412371, + "loss": 0.5483, + "step": 2516 + }, + { + "epoch": 0.80544, + "grad_norm": 0.14889541268348694, + "learning_rate": 0.00010243298969072165, + "loss": 0.6407, + "step": 2517 + }, + { + "epoch": 0.80576, + "grad_norm": 0.16227422654628754, + "learning_rate": 0.00010239175257731958, + "loss": 0.6973, + "step": 2518 + }, + { + "epoch": 0.80608, + "grad_norm": 0.14233717322349548, + "learning_rate": 0.00010235051546391754, + "loss": 0.708, + "step": 2519 + }, + { + "epoch": 0.8064, + "grad_norm": 0.1603488177061081, + "learning_rate": 0.00010230927835051547, + "loss": 0.6483, + "step": 2520 + }, + { + "epoch": 0.80672, + "grad_norm": 0.13383691012859344, + "learning_rate": 0.0001022680412371134, + "loss": 0.5473, + "step": 2521 + }, + { + "epoch": 0.80704, + "grad_norm": 0.1260443478822708, + "learning_rate": 0.00010222680412371135, + "loss": 0.7326, + "step": 2522 + }, + { + "epoch": 0.80736, + "grad_norm": 0.1488627940416336, + "learning_rate": 0.00010218556701030928, + "loss": 0.5467, + "step": 2523 + }, + { + "epoch": 0.80768, + "grad_norm": 0.1343524158000946, + "learning_rate": 0.00010214432989690723, + "loss": 0.6407, + "step": 2524 + }, + { + "epoch": 0.808, + "grad_norm": 0.1389237344264984, + "learning_rate": 0.00010210309278350516, + "loss": 0.6876, + "step": 2525 + }, + { + "epoch": 0.80832, + "grad_norm": 0.14986011385917664, + "learning_rate": 0.00010206185567010309, + "loss": 0.6157, + "step": 2526 + }, + { + "epoch": 0.80864, + "grad_norm": 0.13218912482261658, + "learning_rate": 0.00010202061855670105, + "loss": 0.5754, + "step": 2527 + }, + { + "epoch": 0.80896, + "grad_norm": 0.1326342225074768, + "learning_rate": 0.00010197938144329896, + "loss": 0.5936, + "step": 2528 + }, + { + "epoch": 0.80928, + "grad_norm": 0.15890119969844818, + "learning_rate": 0.00010193814432989692, + "loss": 0.8067, + "step": 2529 + }, + { + "epoch": 0.8096, + "grad_norm": 0.13245666027069092, + "learning_rate": 0.00010189690721649485, + "loss": 0.6158, + "step": 2530 + }, + { + "epoch": 0.80992, + "grad_norm": 0.1530684381723404, + "learning_rate": 0.00010185567010309279, + "loss": 0.6256, + "step": 2531 + }, + { + "epoch": 0.81024, + "grad_norm": 0.15198303759098053, + "learning_rate": 0.00010181443298969073, + "loss": 0.6196, + "step": 2532 + }, + { + "epoch": 0.81056, + "grad_norm": 0.1567327082157135, + "learning_rate": 0.00010177319587628866, + "loss": 0.6496, + "step": 2533 + }, + { + "epoch": 0.81088, + "grad_norm": 0.13306449353694916, + "learning_rate": 0.00010173195876288659, + "loss": 0.6618, + "step": 2534 + }, + { + "epoch": 0.8112, + "grad_norm": 0.13311044871807098, + "learning_rate": 0.00010169072164948454, + "loss": 0.7932, + "step": 2535 + }, + { + "epoch": 0.81152, + "grad_norm": 0.15687695145606995, + "learning_rate": 0.00010164948453608247, + "loss": 0.518, + "step": 2536 + }, + { + "epoch": 0.81184, + "grad_norm": 0.1325300633907318, + "learning_rate": 0.00010160824742268043, + "loss": 0.5681, + "step": 2537 + }, + { + "epoch": 0.81216, + "grad_norm": 0.1348574310541153, + "learning_rate": 0.00010156701030927836, + "loss": 0.5869, + "step": 2538 + }, + { + "epoch": 0.81248, + "grad_norm": 0.15548910200595856, + "learning_rate": 0.00010152577319587629, + "loss": 0.7158, + "step": 2539 + }, + { + "epoch": 0.8128, + "grad_norm": 0.1589280664920807, + "learning_rate": 0.00010148453608247423, + "loss": 0.6607, + "step": 2540 + }, + { + "epoch": 0.81312, + "grad_norm": 0.14693498611450195, + "learning_rate": 0.00010144329896907217, + "loss": 0.6366, + "step": 2541 + }, + { + "epoch": 0.81344, + "grad_norm": 0.1422511786222458, + "learning_rate": 0.00010140206185567011, + "loss": 0.8752, + "step": 2542 + }, + { + "epoch": 0.81376, + "grad_norm": 0.14554370939731598, + "learning_rate": 0.00010136082474226804, + "loss": 0.7101, + "step": 2543 + }, + { + "epoch": 0.81408, + "grad_norm": 0.1348835527896881, + "learning_rate": 0.00010131958762886597, + "loss": 0.7375, + "step": 2544 + }, + { + "epoch": 0.8144, + "grad_norm": 0.1707027405500412, + "learning_rate": 0.00010127835051546393, + "loss": 0.6624, + "step": 2545 + }, + { + "epoch": 0.81472, + "grad_norm": 0.13480974733829498, + "learning_rate": 0.00010123711340206185, + "loss": 0.5508, + "step": 2546 + }, + { + "epoch": 0.81504, + "grad_norm": 0.14284251630306244, + "learning_rate": 0.00010119587628865981, + "loss": 0.8519, + "step": 2547 + }, + { + "epoch": 0.81536, + "grad_norm": 0.13237915933132172, + "learning_rate": 0.00010115463917525774, + "loss": 0.522, + "step": 2548 + }, + { + "epoch": 0.81568, + "grad_norm": 0.1333848088979721, + "learning_rate": 0.00010111340206185567, + "loss": 0.852, + "step": 2549 + }, + { + "epoch": 0.816, + "grad_norm": 0.156629279255867, + "learning_rate": 0.00010107216494845362, + "loss": 0.6419, + "step": 2550 + }, + { + "epoch": 0.81632, + "grad_norm": 0.14928120374679565, + "learning_rate": 0.00010103092783505155, + "loss": 0.702, + "step": 2551 + }, + { + "epoch": 0.81664, + "grad_norm": 0.1526610404253006, + "learning_rate": 0.0001009896907216495, + "loss": 0.6222, + "step": 2552 + }, + { + "epoch": 0.81696, + "grad_norm": 0.15716953575611115, + "learning_rate": 0.00010094845360824742, + "loss": 0.67, + "step": 2553 + }, + { + "epoch": 0.81728, + "grad_norm": 0.16704876720905304, + "learning_rate": 0.00010090721649484535, + "loss": 0.7312, + "step": 2554 + }, + { + "epoch": 0.8176, + "grad_norm": 0.15181118249893188, + "learning_rate": 0.00010086597938144331, + "loss": 0.6918, + "step": 2555 + }, + { + "epoch": 0.81792, + "grad_norm": 0.14745359122753143, + "learning_rate": 0.00010082474226804124, + "loss": 0.7272, + "step": 2556 + }, + { + "epoch": 0.81824, + "grad_norm": 0.15232853591442108, + "learning_rate": 0.00010078350515463919, + "loss": 0.7333, + "step": 2557 + }, + { + "epoch": 0.81856, + "grad_norm": 0.14403194189071655, + "learning_rate": 0.00010074226804123712, + "loss": 0.7319, + "step": 2558 + }, + { + "epoch": 0.81888, + "grad_norm": 0.16210834681987762, + "learning_rate": 0.00010070103092783505, + "loss": 0.677, + "step": 2559 + }, + { + "epoch": 0.8192, + "grad_norm": 0.1689915806055069, + "learning_rate": 0.000100659793814433, + "loss": 0.6012, + "step": 2560 + }, + { + "epoch": 0.81952, + "grad_norm": 0.14746619760990143, + "learning_rate": 0.00010061855670103093, + "loss": 0.6945, + "step": 2561 + }, + { + "epoch": 0.81984, + "grad_norm": 0.13620714843273163, + "learning_rate": 0.00010057731958762886, + "loss": 0.5955, + "step": 2562 + }, + { + "epoch": 0.82016, + "grad_norm": 0.15559464693069458, + "learning_rate": 0.00010053608247422682, + "loss": 0.5927, + "step": 2563 + }, + { + "epoch": 0.82048, + "grad_norm": 0.17088450491428375, + "learning_rate": 0.00010049484536082473, + "loss": 0.7917, + "step": 2564 + }, + { + "epoch": 0.8208, + "grad_norm": 0.14598117768764496, + "learning_rate": 0.00010045360824742269, + "loss": 0.5507, + "step": 2565 + }, + { + "epoch": 0.82112, + "grad_norm": 0.1416582465171814, + "learning_rate": 0.00010041237113402062, + "loss": 0.5937, + "step": 2566 + }, + { + "epoch": 0.82144, + "grad_norm": 0.14313280582427979, + "learning_rate": 0.00010037113402061856, + "loss": 0.5794, + "step": 2567 + }, + { + "epoch": 0.82176, + "grad_norm": 0.1723516285419464, + "learning_rate": 0.0001003298969072165, + "loss": 0.6988, + "step": 2568 + }, + { + "epoch": 0.82208, + "grad_norm": 0.13833698630332947, + "learning_rate": 0.00010028865979381443, + "loss": 0.6219, + "step": 2569 + }, + { + "epoch": 0.8224, + "grad_norm": 0.15013179183006287, + "learning_rate": 0.00010024742268041239, + "loss": 0.5789, + "step": 2570 + }, + { + "epoch": 0.82272, + "grad_norm": 0.13466188311576843, + "learning_rate": 0.00010020618556701031, + "loss": 0.6723, + "step": 2571 + }, + { + "epoch": 0.82304, + "grad_norm": 0.14390186965465546, + "learning_rate": 0.00010016494845360824, + "loss": 0.5937, + "step": 2572 + }, + { + "epoch": 0.82336, + "grad_norm": 0.13747534155845642, + "learning_rate": 0.0001001237113402062, + "loss": 0.7331, + "step": 2573 + }, + { + "epoch": 0.82368, + "grad_norm": 0.1619003713130951, + "learning_rate": 0.00010008247422680413, + "loss": 0.6813, + "step": 2574 + }, + { + "epoch": 0.824, + "grad_norm": 0.1569238156080246, + "learning_rate": 0.00010004123711340207, + "loss": 0.7711, + "step": 2575 + }, + { + "epoch": 0.82432, + "grad_norm": 0.13170728087425232, + "learning_rate": 0.0001, + "loss": 0.8772, + "step": 2576 + }, + { + "epoch": 0.82464, + "grad_norm": 0.16701076924800873, + "learning_rate": 9.995876288659795e-05, + "loss": 0.7607, + "step": 2577 + }, + { + "epoch": 0.82496, + "grad_norm": 0.13784126937389374, + "learning_rate": 9.991752577319588e-05, + "loss": 0.6722, + "step": 2578 + }, + { + "epoch": 0.82528, + "grad_norm": 0.14382927119731903, + "learning_rate": 9.987628865979381e-05, + "loss": 0.5802, + "step": 2579 + }, + { + "epoch": 0.8256, + "grad_norm": 0.1624596118927002, + "learning_rate": 9.983505154639176e-05, + "loss": 0.6901, + "step": 2580 + }, + { + "epoch": 0.82592, + "grad_norm": 0.16600771248340607, + "learning_rate": 9.97938144329897e-05, + "loss": 0.7496, + "step": 2581 + }, + { + "epoch": 0.82624, + "grad_norm": 0.13714657723903656, + "learning_rate": 9.975257731958763e-05, + "loss": 0.7355, + "step": 2582 + }, + { + "epoch": 0.82656, + "grad_norm": 0.14041724801063538, + "learning_rate": 9.971134020618558e-05, + "loss": 0.6475, + "step": 2583 + }, + { + "epoch": 0.82688, + "grad_norm": 0.13969376683235168, + "learning_rate": 9.967010309278351e-05, + "loss": 0.4939, + "step": 2584 + }, + { + "epoch": 0.8272, + "grad_norm": 0.14724983274936676, + "learning_rate": 9.962886597938144e-05, + "loss": 0.7215, + "step": 2585 + }, + { + "epoch": 0.82752, + "grad_norm": 0.13816259801387787, + "learning_rate": 9.958762886597939e-05, + "loss": 0.6734, + "step": 2586 + }, + { + "epoch": 0.82784, + "grad_norm": 0.16520173847675323, + "learning_rate": 9.954639175257733e-05, + "loss": 0.681, + "step": 2587 + }, + { + "epoch": 0.82816, + "grad_norm": 0.14162090420722961, + "learning_rate": 9.950515463917526e-05, + "loss": 0.5926, + "step": 2588 + }, + { + "epoch": 0.82848, + "grad_norm": 0.1469050943851471, + "learning_rate": 9.946391752577319e-05, + "loss": 0.6909, + "step": 2589 + }, + { + "epoch": 0.8288, + "grad_norm": 0.1438760757446289, + "learning_rate": 9.942268041237114e-05, + "loss": 0.5542, + "step": 2590 + }, + { + "epoch": 0.82912, + "grad_norm": 0.14326460659503937, + "learning_rate": 9.938144329896908e-05, + "loss": 0.6382, + "step": 2591 + }, + { + "epoch": 0.82944, + "grad_norm": 0.1343224048614502, + "learning_rate": 9.934020618556701e-05, + "loss": 0.7166, + "step": 2592 + }, + { + "epoch": 0.82976, + "grad_norm": 0.14496579766273499, + "learning_rate": 9.929896907216494e-05, + "loss": 0.6291, + "step": 2593 + }, + { + "epoch": 0.83008, + "grad_norm": 0.1555815488100052, + "learning_rate": 9.925773195876289e-05, + "loss": 0.8567, + "step": 2594 + }, + { + "epoch": 0.8304, + "grad_norm": 0.13513390719890594, + "learning_rate": 9.921649484536083e-05, + "loss": 0.6101, + "step": 2595 + }, + { + "epoch": 0.83072, + "grad_norm": 0.14676405489444733, + "learning_rate": 9.917525773195877e-05, + "loss": 0.6351, + "step": 2596 + }, + { + "epoch": 0.83104, + "grad_norm": 0.13840948045253754, + "learning_rate": 9.913402061855671e-05, + "loss": 0.6104, + "step": 2597 + }, + { + "epoch": 0.83136, + "grad_norm": 0.13688334822654724, + "learning_rate": 9.909278350515464e-05, + "loss": 0.4863, + "step": 2598 + }, + { + "epoch": 0.83168, + "grad_norm": 0.13742093741893768, + "learning_rate": 9.905154639175259e-05, + "loss": 0.7051, + "step": 2599 + }, + { + "epoch": 0.832, + "grad_norm": 0.13216431438922882, + "learning_rate": 9.901030927835052e-05, + "loss": 0.6708, + "step": 2600 + }, + { + "epoch": 0.83232, + "grad_norm": 0.14820389449596405, + "learning_rate": 9.896907216494846e-05, + "loss": 0.5897, + "step": 2601 + }, + { + "epoch": 0.83264, + "grad_norm": 0.1325574815273285, + "learning_rate": 9.89278350515464e-05, + "loss": 0.7931, + "step": 2602 + }, + { + "epoch": 0.83296, + "grad_norm": 0.15471117198467255, + "learning_rate": 9.888659793814433e-05, + "loss": 0.6001, + "step": 2603 + }, + { + "epoch": 0.83328, + "grad_norm": 0.14263775944709778, + "learning_rate": 9.884536082474227e-05, + "loss": 0.5715, + "step": 2604 + }, + { + "epoch": 0.8336, + "grad_norm": 0.1264907568693161, + "learning_rate": 9.880412371134022e-05, + "loss": 0.6784, + "step": 2605 + }, + { + "epoch": 0.83392, + "grad_norm": 0.13780780136585236, + "learning_rate": 9.876288659793816e-05, + "loss": 0.7156, + "step": 2606 + }, + { + "epoch": 0.83424, + "grad_norm": 0.15576423704624176, + "learning_rate": 9.872164948453608e-05, + "loss": 0.749, + "step": 2607 + }, + { + "epoch": 0.83456, + "grad_norm": 0.15901878476142883, + "learning_rate": 9.868041237113402e-05, + "loss": 0.6532, + "step": 2608 + }, + { + "epoch": 0.83488, + "grad_norm": 0.14208783209323883, + "learning_rate": 9.863917525773197e-05, + "loss": 0.8918, + "step": 2609 + }, + { + "epoch": 0.8352, + "grad_norm": 0.1626291424036026, + "learning_rate": 9.85979381443299e-05, + "loss": 0.6394, + "step": 2610 + }, + { + "epoch": 0.83552, + "grad_norm": 0.1540312021970749, + "learning_rate": 9.855670103092784e-05, + "loss": 0.6813, + "step": 2611 + }, + { + "epoch": 0.83584, + "grad_norm": 0.157301127910614, + "learning_rate": 9.851546391752577e-05, + "loss": 0.6241, + "step": 2612 + }, + { + "epoch": 0.83616, + "grad_norm": 0.14560700953006744, + "learning_rate": 9.847422680412372e-05, + "loss": 0.6556, + "step": 2613 + }, + { + "epoch": 0.83648, + "grad_norm": 0.14283452928066254, + "learning_rate": 9.843298969072165e-05, + "loss": 0.7826, + "step": 2614 + }, + { + "epoch": 0.8368, + "grad_norm": 0.13824255764484406, + "learning_rate": 9.83917525773196e-05, + "loss": 0.6806, + "step": 2615 + }, + { + "epoch": 0.83712, + "grad_norm": 0.1630062609910965, + "learning_rate": 9.835051546391753e-05, + "loss": 0.6263, + "step": 2616 + }, + { + "epoch": 0.83744, + "grad_norm": 0.14712683856487274, + "learning_rate": 9.830927835051547e-05, + "loss": 0.6213, + "step": 2617 + }, + { + "epoch": 0.83776, + "grad_norm": 0.16427718102931976, + "learning_rate": 9.82680412371134e-05, + "loss": 0.6873, + "step": 2618 + }, + { + "epoch": 0.83808, + "grad_norm": 0.156505286693573, + "learning_rate": 9.822680412371135e-05, + "loss": 0.7456, + "step": 2619 + }, + { + "epoch": 0.8384, + "grad_norm": 0.15571194887161255, + "learning_rate": 9.818556701030929e-05, + "loss": 0.6271, + "step": 2620 + }, + { + "epoch": 0.83872, + "grad_norm": 0.12065141648054123, + "learning_rate": 9.814432989690721e-05, + "loss": 0.5922, + "step": 2621 + }, + { + "epoch": 0.83904, + "grad_norm": 0.1480621099472046, + "learning_rate": 9.810309278350516e-05, + "loss": 0.7655, + "step": 2622 + }, + { + "epoch": 0.83936, + "grad_norm": 0.15559715032577515, + "learning_rate": 9.80618556701031e-05, + "loss": 0.643, + "step": 2623 + }, + { + "epoch": 0.83968, + "grad_norm": 0.1559750884771347, + "learning_rate": 9.802061855670104e-05, + "loss": 0.7525, + "step": 2624 + }, + { + "epoch": 0.84, + "grad_norm": 0.11596807092428207, + "learning_rate": 9.797938144329898e-05, + "loss": 0.5884, + "step": 2625 + }, + { + "epoch": 0.84032, + "grad_norm": 0.18173126876354218, + "learning_rate": 9.793814432989691e-05, + "loss": 0.7998, + "step": 2626 + }, + { + "epoch": 0.84064, + "grad_norm": 0.14626763761043549, + "learning_rate": 9.789690721649485e-05, + "loss": 0.7507, + "step": 2627 + }, + { + "epoch": 0.84096, + "grad_norm": 0.1517423689365387, + "learning_rate": 9.785567010309278e-05, + "loss": 0.7615, + "step": 2628 + }, + { + "epoch": 0.84128, + "grad_norm": 0.14430741965770721, + "learning_rate": 9.781443298969073e-05, + "loss": 0.6387, + "step": 2629 + }, + { + "epoch": 0.8416, + "grad_norm": 0.15506482124328613, + "learning_rate": 9.777319587628866e-05, + "loss": 0.6688, + "step": 2630 + }, + { + "epoch": 0.84192, + "grad_norm": 0.1437118798494339, + "learning_rate": 9.77319587628866e-05, + "loss": 0.6933, + "step": 2631 + }, + { + "epoch": 0.84224, + "grad_norm": 0.1302749067544937, + "learning_rate": 9.769072164948454e-05, + "loss": 0.6382, + "step": 2632 + }, + { + "epoch": 0.84256, + "grad_norm": 0.14823035895824432, + "learning_rate": 9.764948453608248e-05, + "loss": 0.7004, + "step": 2633 + }, + { + "epoch": 0.84288, + "grad_norm": 0.1370924860239029, + "learning_rate": 9.760824742268043e-05, + "loss": 0.7528, + "step": 2634 + }, + { + "epoch": 0.8432, + "grad_norm": 0.1547056883573532, + "learning_rate": 9.756701030927836e-05, + "loss": 0.6377, + "step": 2635 + }, + { + "epoch": 0.84352, + "grad_norm": 0.16224320232868195, + "learning_rate": 9.752577319587629e-05, + "loss": 0.7833, + "step": 2636 + }, + { + "epoch": 0.84384, + "grad_norm": 0.155901700258255, + "learning_rate": 9.748453608247423e-05, + "loss": 0.5518, + "step": 2637 + }, + { + "epoch": 0.84416, + "grad_norm": 0.1412147432565689, + "learning_rate": 9.744329896907218e-05, + "loss": 0.596, + "step": 2638 + }, + { + "epoch": 0.84448, + "grad_norm": 0.1567072719335556, + "learning_rate": 9.740206185567011e-05, + "loss": 0.7302, + "step": 2639 + }, + { + "epoch": 0.8448, + "grad_norm": 0.15213274955749512, + "learning_rate": 9.736082474226804e-05, + "loss": 0.7733, + "step": 2640 + }, + { + "epoch": 0.84512, + "grad_norm": 0.145083948969841, + "learning_rate": 9.731958762886598e-05, + "loss": 0.7287, + "step": 2641 + }, + { + "epoch": 0.84544, + "grad_norm": 0.1471479833126068, + "learning_rate": 9.727835051546393e-05, + "loss": 0.6619, + "step": 2642 + }, + { + "epoch": 0.84576, + "grad_norm": 0.15284790098667145, + "learning_rate": 9.723711340206186e-05, + "loss": 0.6752, + "step": 2643 + }, + { + "epoch": 0.84608, + "grad_norm": 0.13611441850662231, + "learning_rate": 9.71958762886598e-05, + "loss": 0.6216, + "step": 2644 + }, + { + "epoch": 0.8464, + "grad_norm": 0.1514507234096527, + "learning_rate": 9.715463917525774e-05, + "loss": 0.7845, + "step": 2645 + }, + { + "epoch": 0.84672, + "grad_norm": 0.14236316084861755, + "learning_rate": 9.711340206185567e-05, + "loss": 0.7666, + "step": 2646 + }, + { + "epoch": 0.84704, + "grad_norm": 0.14750486612319946, + "learning_rate": 9.707216494845361e-05, + "loss": 0.6536, + "step": 2647 + }, + { + "epoch": 0.84736, + "grad_norm": 0.1588330864906311, + "learning_rate": 9.703092783505156e-05, + "loss": 0.69, + "step": 2648 + }, + { + "epoch": 0.84768, + "grad_norm": 0.15343515574932098, + "learning_rate": 9.698969072164949e-05, + "loss": 0.5738, + "step": 2649 + }, + { + "epoch": 0.848, + "grad_norm": 0.14453311264514923, + "learning_rate": 9.694845360824742e-05, + "loss": 0.7927, + "step": 2650 + }, + { + "epoch": 0.84832, + "grad_norm": 0.14192573726177216, + "learning_rate": 9.690721649484537e-05, + "loss": 0.6135, + "step": 2651 + }, + { + "epoch": 0.84864, + "grad_norm": 0.14386694133281708, + "learning_rate": 9.686597938144331e-05, + "loss": 0.6794, + "step": 2652 + }, + { + "epoch": 0.84896, + "grad_norm": 0.1503598541021347, + "learning_rate": 9.682474226804124e-05, + "loss": 0.7953, + "step": 2653 + }, + { + "epoch": 0.84928, + "grad_norm": 0.14505556225776672, + "learning_rate": 9.678350515463917e-05, + "loss": 0.7214, + "step": 2654 + }, + { + "epoch": 0.8496, + "grad_norm": 0.15027664601802826, + "learning_rate": 9.674226804123712e-05, + "loss": 0.61, + "step": 2655 + }, + { + "epoch": 0.84992, + "grad_norm": 0.13343876600265503, + "learning_rate": 9.670103092783506e-05, + "loss": 0.5895, + "step": 2656 + }, + { + "epoch": 0.85024, + "grad_norm": 0.15161848068237305, + "learning_rate": 9.6659793814433e-05, + "loss": 0.6966, + "step": 2657 + }, + { + "epoch": 0.85056, + "grad_norm": 0.1409144550561905, + "learning_rate": 9.661855670103094e-05, + "loss": 0.6107, + "step": 2658 + }, + { + "epoch": 0.85088, + "grad_norm": 0.1427966058254242, + "learning_rate": 9.657731958762887e-05, + "loss": 0.7343, + "step": 2659 + }, + { + "epoch": 0.8512, + "grad_norm": 0.154793381690979, + "learning_rate": 9.65360824742268e-05, + "loss": 0.6959, + "step": 2660 + }, + { + "epoch": 0.85152, + "grad_norm": 0.13592591881752014, + "learning_rate": 9.649484536082475e-05, + "loss": 0.7076, + "step": 2661 + }, + { + "epoch": 0.85184, + "grad_norm": 0.148332878947258, + "learning_rate": 9.645360824742269e-05, + "loss": 0.7775, + "step": 2662 + }, + { + "epoch": 0.85216, + "grad_norm": 0.1632942110300064, + "learning_rate": 9.641237113402062e-05, + "loss": 0.8591, + "step": 2663 + }, + { + "epoch": 0.85248, + "grad_norm": 0.15150511264801025, + "learning_rate": 9.637113402061855e-05, + "loss": 0.6348, + "step": 2664 + }, + { + "epoch": 0.8528, + "grad_norm": 0.1594093143939972, + "learning_rate": 9.63298969072165e-05, + "loss": 0.7732, + "step": 2665 + }, + { + "epoch": 0.85312, + "grad_norm": 0.15586240589618683, + "learning_rate": 9.628865979381444e-05, + "loss": 0.5762, + "step": 2666 + }, + { + "epoch": 0.85344, + "grad_norm": 0.17030638456344604, + "learning_rate": 9.624742268041237e-05, + "loss": 0.8345, + "step": 2667 + }, + { + "epoch": 0.85376, + "grad_norm": 0.17846634984016418, + "learning_rate": 9.62061855670103e-05, + "loss": 0.6701, + "step": 2668 + }, + { + "epoch": 0.85408, + "grad_norm": 0.13325104117393494, + "learning_rate": 9.616494845360825e-05, + "loss": 0.6174, + "step": 2669 + }, + { + "epoch": 0.8544, + "grad_norm": 0.175913468003273, + "learning_rate": 9.61237113402062e-05, + "loss": 0.6785, + "step": 2670 + }, + { + "epoch": 0.85472, + "grad_norm": 0.14931370317935944, + "learning_rate": 9.608247422680413e-05, + "loss": 0.6546, + "step": 2671 + }, + { + "epoch": 0.85504, + "grad_norm": 0.16227376461029053, + "learning_rate": 9.604123711340207e-05, + "loss": 0.7058, + "step": 2672 + }, + { + "epoch": 0.85536, + "grad_norm": 0.152308851480484, + "learning_rate": 9.6e-05, + "loss": 0.6612, + "step": 2673 + }, + { + "epoch": 0.85568, + "grad_norm": 0.1365320086479187, + "learning_rate": 9.595876288659795e-05, + "loss": 0.5276, + "step": 2674 + }, + { + "epoch": 0.856, + "grad_norm": 0.16843044757843018, + "learning_rate": 9.591752577319588e-05, + "loss": 0.8361, + "step": 2675 + }, + { + "epoch": 0.85632, + "grad_norm": 0.177180677652359, + "learning_rate": 9.587628865979382e-05, + "loss": 0.7201, + "step": 2676 + }, + { + "epoch": 0.85664, + "grad_norm": 0.13919010758399963, + "learning_rate": 9.583505154639175e-05, + "loss": 0.5619, + "step": 2677 + }, + { + "epoch": 0.85696, + "grad_norm": 0.14129416644573212, + "learning_rate": 9.579381443298969e-05, + "loss": 0.7119, + "step": 2678 + }, + { + "epoch": 0.85728, + "grad_norm": 0.1422802209854126, + "learning_rate": 9.575257731958763e-05, + "loss": 0.666, + "step": 2679 + }, + { + "epoch": 0.8576, + "grad_norm": 0.13324308395385742, + "learning_rate": 9.571134020618558e-05, + "loss": 0.549, + "step": 2680 + }, + { + "epoch": 0.85792, + "grad_norm": 0.15192937850952148, + "learning_rate": 9.567010309278352e-05, + "loss": 0.7808, + "step": 2681 + }, + { + "epoch": 0.85824, + "grad_norm": 0.13704366981983185, + "learning_rate": 9.562886597938144e-05, + "loss": 0.5737, + "step": 2682 + }, + { + "epoch": 0.85856, + "grad_norm": 0.13395696878433228, + "learning_rate": 9.558762886597938e-05, + "loss": 0.5683, + "step": 2683 + }, + { + "epoch": 0.85888, + "grad_norm": 0.1612052321434021, + "learning_rate": 9.554639175257733e-05, + "loss": 0.6138, + "step": 2684 + }, + { + "epoch": 0.8592, + "grad_norm": 0.1311410516500473, + "learning_rate": 9.550515463917526e-05, + "loss": 0.7514, + "step": 2685 + }, + { + "epoch": 0.85952, + "grad_norm": 0.1430673897266388, + "learning_rate": 9.54639175257732e-05, + "loss": 0.6427, + "step": 2686 + }, + { + "epoch": 0.85984, + "grad_norm": 0.15446272492408752, + "learning_rate": 9.542268041237114e-05, + "loss": 0.6767, + "step": 2687 + }, + { + "epoch": 0.86016, + "grad_norm": 0.13702060282230377, + "learning_rate": 9.538144329896908e-05, + "loss": 0.757, + "step": 2688 + }, + { + "epoch": 0.86048, + "grad_norm": 0.15486273169517517, + "learning_rate": 9.534020618556701e-05, + "loss": 0.6733, + "step": 2689 + }, + { + "epoch": 0.8608, + "grad_norm": 0.157430961728096, + "learning_rate": 9.529896907216496e-05, + "loss": 0.5853, + "step": 2690 + }, + { + "epoch": 0.86112, + "grad_norm": 0.14828483760356903, + "learning_rate": 9.525773195876289e-05, + "loss": 0.7017, + "step": 2691 + }, + { + "epoch": 0.86144, + "grad_norm": 0.14540791511535645, + "learning_rate": 9.521649484536083e-05, + "loss": 0.7107, + "step": 2692 + }, + { + "epoch": 0.86176, + "grad_norm": 0.13171698153018951, + "learning_rate": 9.517525773195876e-05, + "loss": 0.7336, + "step": 2693 + }, + { + "epoch": 0.86208, + "grad_norm": 0.1436295062303543, + "learning_rate": 9.513402061855671e-05, + "loss": 0.8073, + "step": 2694 + }, + { + "epoch": 0.8624, + "grad_norm": 0.12731266021728516, + "learning_rate": 9.509278350515465e-05, + "loss": 0.5541, + "step": 2695 + }, + { + "epoch": 0.86272, + "grad_norm": 0.13568390905857086, + "learning_rate": 9.505154639175257e-05, + "loss": 0.5773, + "step": 2696 + }, + { + "epoch": 0.86304, + "grad_norm": 0.14307625591754913, + "learning_rate": 9.501030927835052e-05, + "loss": 0.8994, + "step": 2697 + }, + { + "epoch": 0.86336, + "grad_norm": 0.1440960019826889, + "learning_rate": 9.496907216494846e-05, + "loss": 0.7012, + "step": 2698 + }, + { + "epoch": 0.86368, + "grad_norm": 0.13853365182876587, + "learning_rate": 9.49278350515464e-05, + "loss": 0.7082, + "step": 2699 + }, + { + "epoch": 0.864, + "grad_norm": 0.13834795355796814, + "learning_rate": 9.488659793814434e-05, + "loss": 0.7439, + "step": 2700 + }, + { + "epoch": 0.86432, + "grad_norm": 0.13410118222236633, + "learning_rate": 9.484536082474227e-05, + "loss": 0.9104, + "step": 2701 + }, + { + "epoch": 0.86464, + "grad_norm": 0.14079883694648743, + "learning_rate": 9.480412371134021e-05, + "loss": 0.6094, + "step": 2702 + }, + { + "epoch": 0.86496, + "grad_norm": 0.1489664614200592, + "learning_rate": 9.476288659793814e-05, + "loss": 0.6187, + "step": 2703 + }, + { + "epoch": 0.86528, + "grad_norm": 0.14065685868263245, + "learning_rate": 9.472164948453609e-05, + "loss": 0.6203, + "step": 2704 + }, + { + "epoch": 0.8656, + "grad_norm": 0.14575262367725372, + "learning_rate": 9.468041237113402e-05, + "loss": 0.6529, + "step": 2705 + }, + { + "epoch": 0.86592, + "grad_norm": 0.1393088549375534, + "learning_rate": 9.463917525773196e-05, + "loss": 0.6857, + "step": 2706 + }, + { + "epoch": 0.86624, + "grad_norm": 0.13013994693756104, + "learning_rate": 9.45979381443299e-05, + "loss": 0.598, + "step": 2707 + }, + { + "epoch": 0.86656, + "grad_norm": 0.13313470780849457, + "learning_rate": 9.455670103092784e-05, + "loss": 0.7222, + "step": 2708 + }, + { + "epoch": 0.86688, + "grad_norm": 0.13162772357463837, + "learning_rate": 9.451546391752579e-05, + "loss": 0.6744, + "step": 2709 + }, + { + "epoch": 0.8672, + "grad_norm": 0.13559633493423462, + "learning_rate": 9.447422680412372e-05, + "loss": 0.6544, + "step": 2710 + }, + { + "epoch": 0.86752, + "grad_norm": 0.15338216722011566, + "learning_rate": 9.443298969072165e-05, + "loss": 0.6614, + "step": 2711 + }, + { + "epoch": 0.86784, + "grad_norm": 0.15808825194835663, + "learning_rate": 9.439175257731959e-05, + "loss": 0.7301, + "step": 2712 + }, + { + "epoch": 0.86816, + "grad_norm": 0.14262308180332184, + "learning_rate": 9.435051546391754e-05, + "loss": 0.7132, + "step": 2713 + }, + { + "epoch": 0.86848, + "grad_norm": 0.19240713119506836, + "learning_rate": 9.430927835051547e-05, + "loss": 0.7663, + "step": 2714 + }, + { + "epoch": 0.8688, + "grad_norm": 0.13576410710811615, + "learning_rate": 9.42680412371134e-05, + "loss": 0.7144, + "step": 2715 + }, + { + "epoch": 0.86912, + "grad_norm": 0.15717877447605133, + "learning_rate": 9.422680412371135e-05, + "loss": 0.7313, + "step": 2716 + }, + { + "epoch": 0.86944, + "grad_norm": 0.14994347095489502, + "learning_rate": 9.418556701030929e-05, + "loss": 0.7495, + "step": 2717 + }, + { + "epoch": 0.86976, + "grad_norm": 0.15492890775203705, + "learning_rate": 9.414432989690722e-05, + "loss": 0.7245, + "step": 2718 + }, + { + "epoch": 0.87008, + "grad_norm": 0.1292119026184082, + "learning_rate": 9.410309278350517e-05, + "loss": 0.5504, + "step": 2719 + }, + { + "epoch": 0.8704, + "grad_norm": 0.15975598990917206, + "learning_rate": 9.40618556701031e-05, + "loss": 0.7898, + "step": 2720 + }, + { + "epoch": 0.87072, + "grad_norm": 0.13598224520683289, + "learning_rate": 9.402061855670103e-05, + "loss": 0.6178, + "step": 2721 + }, + { + "epoch": 0.87104, + "grad_norm": 0.1578167974948883, + "learning_rate": 9.397938144329897e-05, + "loss": 0.6417, + "step": 2722 + }, + { + "epoch": 0.87136, + "grad_norm": 0.15769025683403015, + "learning_rate": 9.393814432989692e-05, + "loss": 0.6374, + "step": 2723 + }, + { + "epoch": 0.87168, + "grad_norm": 0.133459210395813, + "learning_rate": 9.389690721649485e-05, + "loss": 0.598, + "step": 2724 + }, + { + "epoch": 0.872, + "grad_norm": 0.15592925250530243, + "learning_rate": 9.385567010309278e-05, + "loss": 0.7515, + "step": 2725 + }, + { + "epoch": 0.87232, + "grad_norm": 0.13010936975479126, + "learning_rate": 9.381443298969073e-05, + "loss": 0.6336, + "step": 2726 + }, + { + "epoch": 0.87264, + "grad_norm": 0.14686454832553864, + "learning_rate": 9.377319587628867e-05, + "loss": 0.5441, + "step": 2727 + }, + { + "epoch": 0.87296, + "grad_norm": 0.134808748960495, + "learning_rate": 9.37319587628866e-05, + "loss": 0.6895, + "step": 2728 + }, + { + "epoch": 0.87328, + "grad_norm": 0.13320966064929962, + "learning_rate": 9.369072164948453e-05, + "loss": 0.666, + "step": 2729 + }, + { + "epoch": 0.8736, + "grad_norm": 0.15304717421531677, + "learning_rate": 9.364948453608248e-05, + "loss": 0.8172, + "step": 2730 + }, + { + "epoch": 0.87392, + "grad_norm": 0.14853641390800476, + "learning_rate": 9.360824742268042e-05, + "loss": 0.6484, + "step": 2731 + }, + { + "epoch": 0.87424, + "grad_norm": 0.14005105197429657, + "learning_rate": 9.356701030927835e-05, + "loss": 0.8279, + "step": 2732 + }, + { + "epoch": 0.87456, + "grad_norm": 0.14155597984790802, + "learning_rate": 9.35257731958763e-05, + "loss": 0.6222, + "step": 2733 + }, + { + "epoch": 0.87488, + "grad_norm": 0.13842347264289856, + "learning_rate": 9.348453608247423e-05, + "loss": 0.5894, + "step": 2734 + }, + { + "epoch": 0.8752, + "grad_norm": 0.15161506831645966, + "learning_rate": 9.344329896907216e-05, + "loss": 0.6623, + "step": 2735 + }, + { + "epoch": 0.87552, + "grad_norm": 0.15756961703300476, + "learning_rate": 9.34020618556701e-05, + "loss": 0.643, + "step": 2736 + }, + { + "epoch": 0.87584, + "grad_norm": 0.1613890677690506, + "learning_rate": 9.336082474226805e-05, + "loss": 0.7629, + "step": 2737 + }, + { + "epoch": 0.87616, + "grad_norm": 0.16575272381305695, + "learning_rate": 9.331958762886598e-05, + "loss": 0.7219, + "step": 2738 + }, + { + "epoch": 0.87648, + "grad_norm": 0.15608328580856323, + "learning_rate": 9.327835051546391e-05, + "loss": 0.7275, + "step": 2739 + }, + { + "epoch": 0.8768, + "grad_norm": 0.1430474817752838, + "learning_rate": 9.323711340206186e-05, + "loss": 0.7308, + "step": 2740 + }, + { + "epoch": 0.87712, + "grad_norm": 0.15988099575042725, + "learning_rate": 9.31958762886598e-05, + "loss": 0.657, + "step": 2741 + }, + { + "epoch": 0.87744, + "grad_norm": 0.1660308539867401, + "learning_rate": 9.315463917525773e-05, + "loss": 0.808, + "step": 2742 + }, + { + "epoch": 0.87776, + "grad_norm": 0.13987122476100922, + "learning_rate": 9.311340206185567e-05, + "loss": 0.8051, + "step": 2743 + }, + { + "epoch": 0.87808, + "grad_norm": 0.15883585810661316, + "learning_rate": 9.307216494845361e-05, + "loss": 0.799, + "step": 2744 + }, + { + "epoch": 0.8784, + "grad_norm": 0.16046148538589478, + "learning_rate": 9.303092783505156e-05, + "loss": 0.7372, + "step": 2745 + }, + { + "epoch": 0.87872, + "grad_norm": 0.13301150500774384, + "learning_rate": 9.298969072164949e-05, + "loss": 0.8091, + "step": 2746 + }, + { + "epoch": 0.87904, + "grad_norm": 0.14882948994636536, + "learning_rate": 9.294845360824743e-05, + "loss": 0.642, + "step": 2747 + }, + { + "epoch": 0.87936, + "grad_norm": 0.1393263190984726, + "learning_rate": 9.290721649484536e-05, + "loss": 0.7235, + "step": 2748 + }, + { + "epoch": 0.87968, + "grad_norm": 0.13505789637565613, + "learning_rate": 9.286597938144331e-05, + "loss": 0.6789, + "step": 2749 + }, + { + "epoch": 0.88, + "grad_norm": 0.14970743656158447, + "learning_rate": 9.282474226804124e-05, + "loss": 0.8541, + "step": 2750 + }, + { + "epoch": 0.88032, + "grad_norm": 0.14981940388679504, + "learning_rate": 9.278350515463918e-05, + "loss": 0.6984, + "step": 2751 + }, + { + "epoch": 0.88064, + "grad_norm": 0.14987780153751373, + "learning_rate": 9.274226804123712e-05, + "loss": 0.8118, + "step": 2752 + }, + { + "epoch": 0.88096, + "grad_norm": 0.15890739858150482, + "learning_rate": 9.270103092783505e-05, + "loss": 0.729, + "step": 2753 + }, + { + "epoch": 0.88128, + "grad_norm": 0.14084629714488983, + "learning_rate": 9.265979381443299e-05, + "loss": 0.7342, + "step": 2754 + }, + { + "epoch": 0.8816, + "grad_norm": 0.16956742107868195, + "learning_rate": 9.261855670103094e-05, + "loss": 0.7388, + "step": 2755 + }, + { + "epoch": 0.88192, + "grad_norm": 0.15381470322608948, + "learning_rate": 9.257731958762888e-05, + "loss": 0.6771, + "step": 2756 + }, + { + "epoch": 0.88224, + "grad_norm": 0.14246757328510284, + "learning_rate": 9.25360824742268e-05, + "loss": 0.6568, + "step": 2757 + }, + { + "epoch": 0.88256, + "grad_norm": 0.17393605411052704, + "learning_rate": 9.249484536082474e-05, + "loss": 0.742, + "step": 2758 + }, + { + "epoch": 0.88288, + "grad_norm": 0.15587688982486725, + "learning_rate": 9.245360824742269e-05, + "loss": 0.6511, + "step": 2759 + }, + { + "epoch": 0.8832, + "grad_norm": 0.15707197785377502, + "learning_rate": 9.241237113402062e-05, + "loss": 0.5291, + "step": 2760 + }, + { + "epoch": 0.88352, + "grad_norm": 0.14165055751800537, + "learning_rate": 9.237113402061856e-05, + "loss": 0.6675, + "step": 2761 + }, + { + "epoch": 0.88384, + "grad_norm": 0.1638755351305008, + "learning_rate": 9.23298969072165e-05, + "loss": 0.7502, + "step": 2762 + }, + { + "epoch": 0.88416, + "grad_norm": 0.16041173040866852, + "learning_rate": 9.228865979381444e-05, + "loss": 0.7577, + "step": 2763 + }, + { + "epoch": 0.88448, + "grad_norm": 0.1551227867603302, + "learning_rate": 9.224742268041237e-05, + "loss": 0.8418, + "step": 2764 + }, + { + "epoch": 0.8848, + "grad_norm": 0.14460650086402893, + "learning_rate": 9.220618556701032e-05, + "loss": 0.7102, + "step": 2765 + }, + { + "epoch": 0.88512, + "grad_norm": 0.13814446330070496, + "learning_rate": 9.216494845360825e-05, + "loss": 0.6649, + "step": 2766 + }, + { + "epoch": 0.88544, + "grad_norm": 0.1511017233133316, + "learning_rate": 9.212371134020619e-05, + "loss": 0.7478, + "step": 2767 + }, + { + "epoch": 0.88576, + "grad_norm": 0.14079943299293518, + "learning_rate": 9.208247422680412e-05, + "loss": 0.6215, + "step": 2768 + }, + { + "epoch": 0.88608, + "grad_norm": 0.15942710638046265, + "learning_rate": 9.204123711340207e-05, + "loss": 0.7076, + "step": 2769 + }, + { + "epoch": 0.8864, + "grad_norm": 0.15644119679927826, + "learning_rate": 9.200000000000001e-05, + "loss": 0.6345, + "step": 2770 + }, + { + "epoch": 0.88672, + "grad_norm": 0.14400945603847504, + "learning_rate": 9.195876288659793e-05, + "loss": 0.6789, + "step": 2771 + }, + { + "epoch": 0.88704, + "grad_norm": 0.12622833251953125, + "learning_rate": 9.191752577319588e-05, + "loss": 0.5362, + "step": 2772 + }, + { + "epoch": 0.88736, + "grad_norm": 0.15853072702884674, + "learning_rate": 9.187628865979382e-05, + "loss": 0.5783, + "step": 2773 + }, + { + "epoch": 0.88768, + "grad_norm": 0.143870547413826, + "learning_rate": 9.183505154639177e-05, + "loss": 0.6164, + "step": 2774 + }, + { + "epoch": 0.888, + "grad_norm": 0.1473972052335739, + "learning_rate": 9.17938144329897e-05, + "loss": 0.6449, + "step": 2775 + }, + { + "epoch": 0.88832, + "grad_norm": 0.12446624040603638, + "learning_rate": 9.175257731958763e-05, + "loss": 0.5581, + "step": 2776 + }, + { + "epoch": 0.88864, + "grad_norm": 0.12791526317596436, + "learning_rate": 9.171134020618557e-05, + "loss": 0.5724, + "step": 2777 + }, + { + "epoch": 0.88896, + "grad_norm": 0.15102334320545197, + "learning_rate": 9.16701030927835e-05, + "loss": 0.5043, + "step": 2778 + }, + { + "epoch": 0.88928, + "grad_norm": 0.14298342168331146, + "learning_rate": 9.162886597938145e-05, + "loss": 0.7134, + "step": 2779 + }, + { + "epoch": 0.8896, + "grad_norm": 0.1541118174791336, + "learning_rate": 9.15876288659794e-05, + "loss": 0.6842, + "step": 2780 + }, + { + "epoch": 0.88992, + "grad_norm": 0.15074671804904938, + "learning_rate": 9.154639175257733e-05, + "loss": 0.7091, + "step": 2781 + }, + { + "epoch": 0.89024, + "grad_norm": 0.16808509826660156, + "learning_rate": 9.150515463917526e-05, + "loss": 0.7012, + "step": 2782 + }, + { + "epoch": 0.89056, + "grad_norm": 0.15717698633670807, + "learning_rate": 9.14639175257732e-05, + "loss": 0.6977, + "step": 2783 + }, + { + "epoch": 0.89088, + "grad_norm": 0.1693105399608612, + "learning_rate": 9.142268041237115e-05, + "loss": 0.7837, + "step": 2784 + }, + { + "epoch": 0.8912, + "grad_norm": 0.16331219673156738, + "learning_rate": 9.138144329896908e-05, + "loss": 0.7228, + "step": 2785 + }, + { + "epoch": 0.89152, + "grad_norm": 0.1594894975423813, + "learning_rate": 9.134020618556701e-05, + "loss": 0.7547, + "step": 2786 + }, + { + "epoch": 0.89184, + "grad_norm": 0.15970784425735474, + "learning_rate": 9.129896907216495e-05, + "loss": 0.7639, + "step": 2787 + }, + { + "epoch": 0.89216, + "grad_norm": 0.15534740686416626, + "learning_rate": 9.12577319587629e-05, + "loss": 0.5998, + "step": 2788 + }, + { + "epoch": 0.89248, + "grad_norm": 0.15085740387439728, + "learning_rate": 9.121649484536083e-05, + "loss": 0.7808, + "step": 2789 + }, + { + "epoch": 0.8928, + "grad_norm": 0.15100796520709991, + "learning_rate": 9.117525773195876e-05, + "loss": 0.7349, + "step": 2790 + }, + { + "epoch": 0.89312, + "grad_norm": 0.16718041896820068, + "learning_rate": 9.11340206185567e-05, + "loss": 0.7519, + "step": 2791 + }, + { + "epoch": 0.89344, + "grad_norm": 0.1614488661289215, + "learning_rate": 9.109278350515465e-05, + "loss": 0.7194, + "step": 2792 + }, + { + "epoch": 0.89376, + "grad_norm": 0.14992555975914001, + "learning_rate": 9.105154639175258e-05, + "loss": 0.7029, + "step": 2793 + }, + { + "epoch": 0.89408, + "grad_norm": 0.13879945874214172, + "learning_rate": 9.101030927835053e-05, + "loss": 0.5998, + "step": 2794 + }, + { + "epoch": 0.8944, + "grad_norm": 0.13521887362003326, + "learning_rate": 9.096907216494846e-05, + "loss": 0.7166, + "step": 2795 + }, + { + "epoch": 0.89472, + "grad_norm": 0.13023538887500763, + "learning_rate": 9.092783505154639e-05, + "loss": 0.6996, + "step": 2796 + }, + { + "epoch": 0.89504, + "grad_norm": 0.14177380502223969, + "learning_rate": 9.088659793814433e-05, + "loss": 0.6777, + "step": 2797 + }, + { + "epoch": 0.89536, + "grad_norm": 0.15992188453674316, + "learning_rate": 9.084536082474228e-05, + "loss": 0.6179, + "step": 2798 + }, + { + "epoch": 0.89568, + "grad_norm": 0.13800425827503204, + "learning_rate": 9.080412371134021e-05, + "loss": 0.7616, + "step": 2799 + }, + { + "epoch": 0.896, + "grad_norm": 0.12557177245616913, + "learning_rate": 9.076288659793814e-05, + "loss": 0.6808, + "step": 2800 + }, + { + "epoch": 0.89632, + "grad_norm": 0.13006342947483063, + "learning_rate": 9.072164948453609e-05, + "loss": 0.646, + "step": 2801 + }, + { + "epoch": 0.89664, + "grad_norm": 0.1585310995578766, + "learning_rate": 9.068041237113403e-05, + "loss": 0.7162, + "step": 2802 + }, + { + "epoch": 0.89696, + "grad_norm": 0.16637931764125824, + "learning_rate": 9.063917525773196e-05, + "loss": 0.7483, + "step": 2803 + }, + { + "epoch": 0.89728, + "grad_norm": 0.16359177231788635, + "learning_rate": 9.05979381443299e-05, + "loss": 0.7872, + "step": 2804 + }, + { + "epoch": 0.8976, + "grad_norm": 0.16272203624248505, + "learning_rate": 9.055670103092784e-05, + "loss": 0.8246, + "step": 2805 + }, + { + "epoch": 0.89792, + "grad_norm": 0.16015036404132843, + "learning_rate": 9.051546391752578e-05, + "loss": 0.7272, + "step": 2806 + }, + { + "epoch": 0.89824, + "grad_norm": 0.17095240950584412, + "learning_rate": 9.047422680412371e-05, + "loss": 0.678, + "step": 2807 + }, + { + "epoch": 0.89856, + "grad_norm": 0.13806094229221344, + "learning_rate": 9.043298969072166e-05, + "loss": 0.618, + "step": 2808 + }, + { + "epoch": 0.89888, + "grad_norm": 0.15678280591964722, + "learning_rate": 9.039175257731959e-05, + "loss": 0.7253, + "step": 2809 + }, + { + "epoch": 0.8992, + "grad_norm": 0.13497911393642426, + "learning_rate": 9.035051546391754e-05, + "loss": 0.5294, + "step": 2810 + }, + { + "epoch": 0.89952, + "grad_norm": 0.13017329573631287, + "learning_rate": 9.030927835051547e-05, + "loss": 0.8205, + "step": 2811 + }, + { + "epoch": 0.89984, + "grad_norm": 0.14628775417804718, + "learning_rate": 9.026804123711341e-05, + "loss": 0.6253, + "step": 2812 + }, + { + "epoch": 0.90016, + "grad_norm": 0.14524394273757935, + "learning_rate": 9.022680412371134e-05, + "loss": 0.6162, + "step": 2813 + }, + { + "epoch": 0.90048, + "grad_norm": 0.15810175240039825, + "learning_rate": 9.018556701030927e-05, + "loss": 0.7413, + "step": 2814 + }, + { + "epoch": 0.9008, + "grad_norm": 0.13436077535152435, + "learning_rate": 9.014432989690722e-05, + "loss": 0.593, + "step": 2815 + }, + { + "epoch": 0.90112, + "grad_norm": 0.15093646943569183, + "learning_rate": 9.010309278350516e-05, + "loss": 0.5409, + "step": 2816 + }, + { + "epoch": 0.90144, + "grad_norm": 0.14732801914215088, + "learning_rate": 9.00618556701031e-05, + "loss": 0.7713, + "step": 2817 + }, + { + "epoch": 0.90176, + "grad_norm": 0.1452672928571701, + "learning_rate": 9.002061855670103e-05, + "loss": 0.6286, + "step": 2818 + }, + { + "epoch": 0.90208, + "grad_norm": 0.1473238170146942, + "learning_rate": 8.997938144329897e-05, + "loss": 0.6818, + "step": 2819 + }, + { + "epoch": 0.9024, + "grad_norm": 0.17339883744716644, + "learning_rate": 8.993814432989692e-05, + "loss": 0.8623, + "step": 2820 + }, + { + "epoch": 0.90272, + "grad_norm": 0.15654996037483215, + "learning_rate": 8.989690721649485e-05, + "loss": 0.7188, + "step": 2821 + }, + { + "epoch": 0.90304, + "grad_norm": 0.15103815495967865, + "learning_rate": 8.985567010309279e-05, + "loss": 0.7202, + "step": 2822 + }, + { + "epoch": 0.90336, + "grad_norm": 0.1422942727804184, + "learning_rate": 8.981443298969072e-05, + "loss": 0.5856, + "step": 2823 + }, + { + "epoch": 0.90368, + "grad_norm": 0.15279853343963623, + "learning_rate": 8.977319587628867e-05, + "loss": 0.7886, + "step": 2824 + }, + { + "epoch": 0.904, + "grad_norm": 0.15019339323043823, + "learning_rate": 8.97319587628866e-05, + "loss": 0.764, + "step": 2825 + }, + { + "epoch": 0.90432, + "grad_norm": 0.12838192284107208, + "learning_rate": 8.969072164948454e-05, + "loss": 0.7322, + "step": 2826 + }, + { + "epoch": 0.90464, + "grad_norm": 0.15417592227458954, + "learning_rate": 8.964948453608248e-05, + "loss": 0.7441, + "step": 2827 + }, + { + "epoch": 0.90496, + "grad_norm": 0.12865400314331055, + "learning_rate": 8.960824742268041e-05, + "loss": 0.727, + "step": 2828 + }, + { + "epoch": 0.90528, + "grad_norm": 0.1653413623571396, + "learning_rate": 8.956701030927835e-05, + "loss": 0.6851, + "step": 2829 + }, + { + "epoch": 0.9056, + "grad_norm": 0.13718865811824799, + "learning_rate": 8.95257731958763e-05, + "loss": 0.6101, + "step": 2830 + }, + { + "epoch": 0.90592, + "grad_norm": 0.1428721845149994, + "learning_rate": 8.948453608247424e-05, + "loss": 0.6204, + "step": 2831 + }, + { + "epoch": 0.90624, + "grad_norm": 0.13881301879882812, + "learning_rate": 8.944329896907216e-05, + "loss": 0.8527, + "step": 2832 + }, + { + "epoch": 0.90656, + "grad_norm": 0.13179130852222443, + "learning_rate": 8.94020618556701e-05, + "loss": 0.7399, + "step": 2833 + }, + { + "epoch": 0.90688, + "grad_norm": 0.15606658160686493, + "learning_rate": 8.936082474226805e-05, + "loss": 0.682, + "step": 2834 + }, + { + "epoch": 0.9072, + "grad_norm": 0.15682071447372437, + "learning_rate": 8.931958762886598e-05, + "loss": 0.7092, + "step": 2835 + }, + { + "epoch": 0.90752, + "grad_norm": 0.14316652715206146, + "learning_rate": 8.927835051546392e-05, + "loss": 0.7535, + "step": 2836 + }, + { + "epoch": 0.90784, + "grad_norm": 0.1332533359527588, + "learning_rate": 8.923711340206186e-05, + "loss": 0.6872, + "step": 2837 + }, + { + "epoch": 0.90816, + "grad_norm": 0.16481122374534607, + "learning_rate": 8.91958762886598e-05, + "loss": 0.5852, + "step": 2838 + }, + { + "epoch": 0.90848, + "grad_norm": 0.1456765979528427, + "learning_rate": 8.915463917525773e-05, + "loss": 0.8146, + "step": 2839 + }, + { + "epoch": 0.9088, + "grad_norm": 0.15643854439258575, + "learning_rate": 8.911340206185568e-05, + "loss": 0.7624, + "step": 2840 + }, + { + "epoch": 0.90912, + "grad_norm": 0.13627059757709503, + "learning_rate": 8.907216494845362e-05, + "loss": 0.6942, + "step": 2841 + }, + { + "epoch": 0.90944, + "grad_norm": 0.15078982710838318, + "learning_rate": 8.903092783505155e-05, + "loss": 0.677, + "step": 2842 + }, + { + "epoch": 0.90976, + "grad_norm": 0.15020903944969177, + "learning_rate": 8.898969072164948e-05, + "loss": 0.639, + "step": 2843 + }, + { + "epoch": 0.91008, + "grad_norm": 0.14150528609752655, + "learning_rate": 8.894845360824743e-05, + "loss": 0.7668, + "step": 2844 + }, + { + "epoch": 0.9104, + "grad_norm": 0.13930398225784302, + "learning_rate": 8.890721649484537e-05, + "loss": 0.5654, + "step": 2845 + }, + { + "epoch": 0.91072, + "grad_norm": 0.12243703007698059, + "learning_rate": 8.886597938144329e-05, + "loss": 0.5841, + "step": 2846 + }, + { + "epoch": 0.91104, + "grad_norm": 0.14588524401187897, + "learning_rate": 8.882474226804124e-05, + "loss": 0.6592, + "step": 2847 + }, + { + "epoch": 0.91136, + "grad_norm": 0.1599550098180771, + "learning_rate": 8.878350515463918e-05, + "loss": 0.7375, + "step": 2848 + }, + { + "epoch": 0.91168, + "grad_norm": 0.16268660128116608, + "learning_rate": 8.874226804123713e-05, + "loss": 0.7312, + "step": 2849 + }, + { + "epoch": 0.912, + "grad_norm": 0.14294636249542236, + "learning_rate": 8.870103092783506e-05, + "loss": 0.724, + "step": 2850 + }, + { + "epoch": 0.91232, + "grad_norm": 0.13613921403884888, + "learning_rate": 8.865979381443299e-05, + "loss": 0.6943, + "step": 2851 + }, + { + "epoch": 0.91264, + "grad_norm": 0.13716544210910797, + "learning_rate": 8.861855670103093e-05, + "loss": 0.6351, + "step": 2852 + }, + { + "epoch": 0.91296, + "grad_norm": 0.16739431023597717, + "learning_rate": 8.857731958762887e-05, + "loss": 0.7466, + "step": 2853 + }, + { + "epoch": 0.91328, + "grad_norm": 0.13381338119506836, + "learning_rate": 8.853608247422681e-05, + "loss": 0.7109, + "step": 2854 + }, + { + "epoch": 0.9136, + "grad_norm": 0.15339596569538116, + "learning_rate": 8.849484536082475e-05, + "loss": 0.5338, + "step": 2855 + }, + { + "epoch": 0.91392, + "grad_norm": 0.13078036904335022, + "learning_rate": 8.845360824742269e-05, + "loss": 0.6833, + "step": 2856 + }, + { + "epoch": 0.91424, + "grad_norm": 0.13721908628940582, + "learning_rate": 8.841237113402062e-05, + "loss": 0.7937, + "step": 2857 + }, + { + "epoch": 0.91456, + "grad_norm": 0.15523791313171387, + "learning_rate": 8.837113402061856e-05, + "loss": 0.5827, + "step": 2858 + }, + { + "epoch": 0.91488, + "grad_norm": 0.13530102372169495, + "learning_rate": 8.832989690721651e-05, + "loss": 0.6186, + "step": 2859 + }, + { + "epoch": 0.9152, + "grad_norm": 0.15016475319862366, + "learning_rate": 8.828865979381444e-05, + "loss": 0.6108, + "step": 2860 + }, + { + "epoch": 0.91552, + "grad_norm": 0.14494651556015015, + "learning_rate": 8.824742268041237e-05, + "loss": 0.6858, + "step": 2861 + }, + { + "epoch": 0.91584, + "grad_norm": 0.14727284014225006, + "learning_rate": 8.820618556701031e-05, + "loss": 0.5987, + "step": 2862 + }, + { + "epoch": 0.91616, + "grad_norm": 0.14758120477199554, + "learning_rate": 8.816494845360826e-05, + "loss": 0.5986, + "step": 2863 + }, + { + "epoch": 0.91648, + "grad_norm": 0.15791207551956177, + "learning_rate": 8.812371134020619e-05, + "loss": 0.6865, + "step": 2864 + }, + { + "epoch": 0.9168, + "grad_norm": 0.14156733453273773, + "learning_rate": 8.808247422680412e-05, + "loss": 0.661, + "step": 2865 + }, + { + "epoch": 0.91712, + "grad_norm": 0.1448497772216797, + "learning_rate": 8.804123711340207e-05, + "loss": 0.638, + "step": 2866 + }, + { + "epoch": 0.91744, + "grad_norm": 0.14729934930801392, + "learning_rate": 8.800000000000001e-05, + "loss": 0.6679, + "step": 2867 + }, + { + "epoch": 0.91776, + "grad_norm": 0.1369476318359375, + "learning_rate": 8.795876288659794e-05, + "loss": 0.6514, + "step": 2868 + }, + { + "epoch": 0.91808, + "grad_norm": 0.144110769033432, + "learning_rate": 8.791752577319589e-05, + "loss": 0.6096, + "step": 2869 + }, + { + "epoch": 0.9184, + "grad_norm": 0.1575762927532196, + "learning_rate": 8.787628865979382e-05, + "loss": 0.7457, + "step": 2870 + }, + { + "epoch": 0.91872, + "grad_norm": 0.1544065624475479, + "learning_rate": 8.783505154639175e-05, + "loss": 0.7686, + "step": 2871 + }, + { + "epoch": 0.91904, + "grad_norm": 0.15800683200359344, + "learning_rate": 8.77938144329897e-05, + "loss": 0.7623, + "step": 2872 + }, + { + "epoch": 0.91936, + "grad_norm": 0.14778579771518707, + "learning_rate": 8.775257731958764e-05, + "loss": 0.7223, + "step": 2873 + }, + { + "epoch": 0.91968, + "grad_norm": 0.14265850186347961, + "learning_rate": 8.771134020618557e-05, + "loss": 0.6866, + "step": 2874 + }, + { + "epoch": 0.92, + "grad_norm": 0.15168777108192444, + "learning_rate": 8.76701030927835e-05, + "loss": 0.7601, + "step": 2875 + }, + { + "epoch": 0.92032, + "grad_norm": 0.13224171102046967, + "learning_rate": 8.762886597938145e-05, + "loss": 0.5933, + "step": 2876 + }, + { + "epoch": 0.92064, + "grad_norm": 0.1686430126428604, + "learning_rate": 8.758762886597939e-05, + "loss": 0.6758, + "step": 2877 + }, + { + "epoch": 0.92096, + "grad_norm": 0.1453283280134201, + "learning_rate": 8.754639175257732e-05, + "loss": 0.6654, + "step": 2878 + }, + { + "epoch": 0.92128, + "grad_norm": 0.14618268609046936, + "learning_rate": 8.750515463917525e-05, + "loss": 0.5241, + "step": 2879 + }, + { + "epoch": 0.9216, + "grad_norm": 0.13317014276981354, + "learning_rate": 8.74639175257732e-05, + "loss": 0.7084, + "step": 2880 + }, + { + "epoch": 0.92192, + "grad_norm": 0.12066260725259781, + "learning_rate": 8.742268041237114e-05, + "loss": 0.502, + "step": 2881 + }, + { + "epoch": 0.92224, + "grad_norm": 0.1370021253824234, + "learning_rate": 8.738144329896908e-05, + "loss": 0.7712, + "step": 2882 + }, + { + "epoch": 0.92256, + "grad_norm": 0.15341876447200775, + "learning_rate": 8.734020618556702e-05, + "loss": 0.7493, + "step": 2883 + }, + { + "epoch": 0.92288, + "grad_norm": 0.14589917659759521, + "learning_rate": 8.729896907216495e-05, + "loss": 0.6643, + "step": 2884 + }, + { + "epoch": 0.9232, + "grad_norm": 0.16226980090141296, + "learning_rate": 8.72577319587629e-05, + "loss": 0.5704, + "step": 2885 + }, + { + "epoch": 0.92352, + "grad_norm": 0.1561785340309143, + "learning_rate": 8.721649484536083e-05, + "loss": 0.7559, + "step": 2886 + }, + { + "epoch": 0.92384, + "grad_norm": 0.150752454996109, + "learning_rate": 8.717525773195877e-05, + "loss": 0.5968, + "step": 2887 + }, + { + "epoch": 0.92416, + "grad_norm": 0.1586134135723114, + "learning_rate": 8.71340206185567e-05, + "loss": 0.6168, + "step": 2888 + }, + { + "epoch": 0.92448, + "grad_norm": 0.14044491946697235, + "learning_rate": 8.709278350515463e-05, + "loss": 0.6376, + "step": 2889 + }, + { + "epoch": 0.9248, + "grad_norm": 0.12909279763698578, + "learning_rate": 8.705154639175258e-05, + "loss": 0.6651, + "step": 2890 + }, + { + "epoch": 0.92512, + "grad_norm": 0.13023202121257782, + "learning_rate": 8.701030927835052e-05, + "loss": 0.6057, + "step": 2891 + }, + { + "epoch": 0.92544, + "grad_norm": 0.12108038365840912, + "learning_rate": 8.696907216494846e-05, + "loss": 0.5793, + "step": 2892 + }, + { + "epoch": 0.92576, + "grad_norm": 0.13165992498397827, + "learning_rate": 8.692783505154639e-05, + "loss": 0.6539, + "step": 2893 + }, + { + "epoch": 0.92608, + "grad_norm": 0.15014542639255524, + "learning_rate": 8.688659793814433e-05, + "loss": 0.7467, + "step": 2894 + }, + { + "epoch": 0.9264, + "grad_norm": 0.13839580118656158, + "learning_rate": 8.684536082474228e-05, + "loss": 0.5201, + "step": 2895 + }, + { + "epoch": 0.92672, + "grad_norm": 0.1394243836402893, + "learning_rate": 8.680412371134021e-05, + "loss": 0.6468, + "step": 2896 + }, + { + "epoch": 0.92704, + "grad_norm": 0.15375931560993195, + "learning_rate": 8.676288659793815e-05, + "loss": 0.6822, + "step": 2897 + }, + { + "epoch": 0.92736, + "grad_norm": 0.15792232751846313, + "learning_rate": 8.672164948453608e-05, + "loss": 0.6384, + "step": 2898 + }, + { + "epoch": 0.92768, + "grad_norm": 0.15484371781349182, + "learning_rate": 8.668041237113403e-05, + "loss": 0.6768, + "step": 2899 + }, + { + "epoch": 0.928, + "grad_norm": 0.14625611901283264, + "learning_rate": 8.663917525773196e-05, + "loss": 0.6147, + "step": 2900 + }, + { + "epoch": 0.92832, + "grad_norm": 0.14015774428844452, + "learning_rate": 8.65979381443299e-05, + "loss": 0.9256, + "step": 2901 + }, + { + "epoch": 0.92864, + "grad_norm": 0.16088813543319702, + "learning_rate": 8.655670103092784e-05, + "loss": 0.7263, + "step": 2902 + }, + { + "epoch": 0.92896, + "grad_norm": 0.14505772292613983, + "learning_rate": 8.651546391752577e-05, + "loss": 0.6002, + "step": 2903 + }, + { + "epoch": 0.92928, + "grad_norm": 0.1385323405265808, + "learning_rate": 8.647422680412371e-05, + "loss": 0.7715, + "step": 2904 + }, + { + "epoch": 0.9296, + "grad_norm": 0.14180082082748413, + "learning_rate": 8.643298969072166e-05, + "loss": 0.759, + "step": 2905 + }, + { + "epoch": 0.92992, + "grad_norm": 0.15554724633693695, + "learning_rate": 8.63917525773196e-05, + "loss": 0.7443, + "step": 2906 + }, + { + "epoch": 0.93024, + "grad_norm": 0.150379478931427, + "learning_rate": 8.635051546391752e-05, + "loss": 0.7549, + "step": 2907 + }, + { + "epoch": 0.93056, + "grad_norm": 0.15231016278266907, + "learning_rate": 8.630927835051546e-05, + "loss": 0.7697, + "step": 2908 + }, + { + "epoch": 0.93088, + "grad_norm": 0.14706070721149445, + "learning_rate": 8.626804123711341e-05, + "loss": 0.6861, + "step": 2909 + }, + { + "epoch": 0.9312, + "grad_norm": 0.12860330939292908, + "learning_rate": 8.622680412371134e-05, + "loss": 0.5153, + "step": 2910 + }, + { + "epoch": 0.93152, + "grad_norm": 0.13647432625293732, + "learning_rate": 8.618556701030929e-05, + "loss": 0.7768, + "step": 2911 + }, + { + "epoch": 0.93184, + "grad_norm": 0.16316340863704681, + "learning_rate": 8.614432989690722e-05, + "loss": 0.7096, + "step": 2912 + }, + { + "epoch": 0.93216, + "grad_norm": 0.14672242105007172, + "learning_rate": 8.610309278350516e-05, + "loss": 0.8637, + "step": 2913 + }, + { + "epoch": 0.93248, + "grad_norm": 0.14667625725269318, + "learning_rate": 8.606185567010309e-05, + "loss": 0.6642, + "step": 2914 + }, + { + "epoch": 0.9328, + "grad_norm": 0.1386280208826065, + "learning_rate": 8.602061855670104e-05, + "loss": 0.8753, + "step": 2915 + }, + { + "epoch": 0.93312, + "grad_norm": 0.1403239220380783, + "learning_rate": 8.597938144329898e-05, + "loss": 0.6372, + "step": 2916 + }, + { + "epoch": 0.93344, + "grad_norm": 0.149571493268013, + "learning_rate": 8.593814432989691e-05, + "loss": 0.6792, + "step": 2917 + }, + { + "epoch": 0.93376, + "grad_norm": 0.13790306448936462, + "learning_rate": 8.589690721649485e-05, + "loss": 0.5437, + "step": 2918 + }, + { + "epoch": 0.93408, + "grad_norm": 0.1625206619501114, + "learning_rate": 8.585567010309279e-05, + "loss": 0.7483, + "step": 2919 + }, + { + "epoch": 0.9344, + "grad_norm": 0.1595042496919632, + "learning_rate": 8.581443298969073e-05, + "loss": 0.6597, + "step": 2920 + }, + { + "epoch": 0.93472, + "grad_norm": 0.1480918675661087, + "learning_rate": 8.577319587628865e-05, + "loss": 0.8178, + "step": 2921 + }, + { + "epoch": 0.93504, + "grad_norm": 0.17507483065128326, + "learning_rate": 8.57319587628866e-05, + "loss": 0.7335, + "step": 2922 + }, + { + "epoch": 0.93536, + "grad_norm": 0.15011480450630188, + "learning_rate": 8.569072164948454e-05, + "loss": 0.6977, + "step": 2923 + }, + { + "epoch": 0.93568, + "grad_norm": 0.1479652374982834, + "learning_rate": 8.564948453608249e-05, + "loss": 0.7549, + "step": 2924 + }, + { + "epoch": 0.936, + "grad_norm": 0.1854063868522644, + "learning_rate": 8.560824742268042e-05, + "loss": 0.7332, + "step": 2925 + }, + { + "epoch": 0.93632, + "grad_norm": 0.15966558456420898, + "learning_rate": 8.556701030927835e-05, + "loss": 0.6839, + "step": 2926 + }, + { + "epoch": 0.93664, + "grad_norm": 0.14656081795692444, + "learning_rate": 8.55257731958763e-05, + "loss": 0.5552, + "step": 2927 + }, + { + "epoch": 0.93696, + "grad_norm": 0.16189329326152802, + "learning_rate": 8.548453608247423e-05, + "loss": 0.7422, + "step": 2928 + }, + { + "epoch": 0.93728, + "grad_norm": 0.15637987852096558, + "learning_rate": 8.544329896907217e-05, + "loss": 0.6701, + "step": 2929 + }, + { + "epoch": 0.9376, + "grad_norm": 0.14931650459766388, + "learning_rate": 8.540206185567012e-05, + "loss": 0.7482, + "step": 2930 + }, + { + "epoch": 0.93792, + "grad_norm": 0.1682416796684265, + "learning_rate": 8.536082474226805e-05, + "loss": 0.7533, + "step": 2931 + }, + { + "epoch": 0.93824, + "grad_norm": 0.13084636628627777, + "learning_rate": 8.531958762886598e-05, + "loss": 0.7718, + "step": 2932 + }, + { + "epoch": 0.93856, + "grad_norm": 0.15308988094329834, + "learning_rate": 8.527835051546392e-05, + "loss": 0.7359, + "step": 2933 + }, + { + "epoch": 0.93888, + "grad_norm": 0.1251639723777771, + "learning_rate": 8.523711340206187e-05, + "loss": 0.6421, + "step": 2934 + }, + { + "epoch": 0.9392, + "grad_norm": 0.15198728442192078, + "learning_rate": 8.51958762886598e-05, + "loss": 0.5748, + "step": 2935 + }, + { + "epoch": 0.93952, + "grad_norm": 0.13700324296951294, + "learning_rate": 8.515463917525773e-05, + "loss": 0.6779, + "step": 2936 + }, + { + "epoch": 0.93984, + "grad_norm": 0.15500114858150482, + "learning_rate": 8.511340206185567e-05, + "loss": 0.562, + "step": 2937 + }, + { + "epoch": 0.94016, + "grad_norm": 0.12747320532798767, + "learning_rate": 8.507216494845362e-05, + "loss": 0.565, + "step": 2938 + }, + { + "epoch": 0.94048, + "grad_norm": 0.12580440938472748, + "learning_rate": 8.503092783505155e-05, + "loss": 0.6645, + "step": 2939 + }, + { + "epoch": 0.9408, + "grad_norm": 0.14606395363807678, + "learning_rate": 8.498969072164948e-05, + "loss": 0.8513, + "step": 2940 + }, + { + "epoch": 0.94112, + "grad_norm": 0.16376923024654388, + "learning_rate": 8.494845360824743e-05, + "loss": 0.7621, + "step": 2941 + }, + { + "epoch": 0.94144, + "grad_norm": 0.1481485664844513, + "learning_rate": 8.490721649484537e-05, + "loss": 0.7023, + "step": 2942 + }, + { + "epoch": 0.94176, + "grad_norm": 0.1654863953590393, + "learning_rate": 8.48659793814433e-05, + "loss": 0.6058, + "step": 2943 + }, + { + "epoch": 0.94208, + "grad_norm": 0.15494300425052643, + "learning_rate": 8.482474226804125e-05, + "loss": 0.7003, + "step": 2944 + }, + { + "epoch": 0.9424, + "grad_norm": 0.16636860370635986, + "learning_rate": 8.478350515463918e-05, + "loss": 0.7653, + "step": 2945 + }, + { + "epoch": 0.94272, + "grad_norm": 0.15185664594173431, + "learning_rate": 8.474226804123711e-05, + "loss": 0.7658, + "step": 2946 + }, + { + "epoch": 0.94304, + "grad_norm": 0.14327672123908997, + "learning_rate": 8.470103092783506e-05, + "loss": 0.6759, + "step": 2947 + }, + { + "epoch": 0.94336, + "grad_norm": 0.14658670127391815, + "learning_rate": 8.4659793814433e-05, + "loss": 0.7865, + "step": 2948 + }, + { + "epoch": 0.94368, + "grad_norm": 0.1468987911939621, + "learning_rate": 8.461855670103093e-05, + "loss": 0.6403, + "step": 2949 + }, + { + "epoch": 0.944, + "grad_norm": 0.15289437770843506, + "learning_rate": 8.457731958762886e-05, + "loss": 0.8873, + "step": 2950 + }, + { + "epoch": 0.94432, + "grad_norm": 0.17362968623638153, + "learning_rate": 8.453608247422681e-05, + "loss": 0.6575, + "step": 2951 + }, + { + "epoch": 0.94464, + "grad_norm": 0.16052356362342834, + "learning_rate": 8.449484536082475e-05, + "loss": 0.6534, + "step": 2952 + }, + { + "epoch": 0.94496, + "grad_norm": 0.151447594165802, + "learning_rate": 8.445360824742268e-05, + "loss": 0.6363, + "step": 2953 + }, + { + "epoch": 0.94528, + "grad_norm": 0.15030694007873535, + "learning_rate": 8.441237113402061e-05, + "loss": 0.6551, + "step": 2954 + }, + { + "epoch": 0.9456, + "grad_norm": 0.15353736281394958, + "learning_rate": 8.437113402061856e-05, + "loss": 0.7256, + "step": 2955 + }, + { + "epoch": 0.94592, + "grad_norm": 0.1556423008441925, + "learning_rate": 8.43298969072165e-05, + "loss": 0.6969, + "step": 2956 + }, + { + "epoch": 0.94624, + "grad_norm": 0.15923531353473663, + "learning_rate": 8.428865979381444e-05, + "loss": 0.7082, + "step": 2957 + }, + { + "epoch": 0.94656, + "grad_norm": 0.15418791770935059, + "learning_rate": 8.424742268041238e-05, + "loss": 0.6947, + "step": 2958 + }, + { + "epoch": 0.94688, + "grad_norm": 0.13062843680381775, + "learning_rate": 8.420618556701031e-05, + "loss": 0.4885, + "step": 2959 + }, + { + "epoch": 0.9472, + "grad_norm": 0.14804647862911224, + "learning_rate": 8.416494845360826e-05, + "loss": 0.7022, + "step": 2960 + }, + { + "epoch": 0.94752, + "grad_norm": 0.15559984743595123, + "learning_rate": 8.412371134020619e-05, + "loss": 0.7179, + "step": 2961 + }, + { + "epoch": 0.94784, + "grad_norm": 0.15591378509998322, + "learning_rate": 8.408247422680413e-05, + "loss": 0.7348, + "step": 2962 + }, + { + "epoch": 0.94816, + "grad_norm": 0.14983315765857697, + "learning_rate": 8.404123711340206e-05, + "loss": 0.8834, + "step": 2963 + }, + { + "epoch": 0.94848, + "grad_norm": 0.1544259786605835, + "learning_rate": 8.4e-05, + "loss": 0.6249, + "step": 2964 + }, + { + "epoch": 0.9488, + "grad_norm": 0.15844641625881195, + "learning_rate": 8.395876288659794e-05, + "loss": 0.8443, + "step": 2965 + }, + { + "epoch": 0.94912, + "grad_norm": 0.15102942287921906, + "learning_rate": 8.391752577319589e-05, + "loss": 0.7955, + "step": 2966 + }, + { + "epoch": 0.94944, + "grad_norm": 0.1417260766029358, + "learning_rate": 8.387628865979382e-05, + "loss": 0.7164, + "step": 2967 + }, + { + "epoch": 0.94976, + "grad_norm": 0.13948242366313934, + "learning_rate": 8.383505154639175e-05, + "loss": 0.7465, + "step": 2968 + }, + { + "epoch": 0.95008, + "grad_norm": 0.1401572972536087, + "learning_rate": 8.379381443298969e-05, + "loss": 0.7776, + "step": 2969 + }, + { + "epoch": 0.9504, + "grad_norm": 0.16004788875579834, + "learning_rate": 8.375257731958764e-05, + "loss": 0.6654, + "step": 2970 + }, + { + "epoch": 0.95072, + "grad_norm": 0.15651416778564453, + "learning_rate": 8.371134020618557e-05, + "loss": 0.7671, + "step": 2971 + }, + { + "epoch": 0.95104, + "grad_norm": 0.14648835361003876, + "learning_rate": 8.367010309278351e-05, + "loss": 0.5606, + "step": 2972 + }, + { + "epoch": 0.95136, + "grad_norm": 0.14291217923164368, + "learning_rate": 8.362886597938144e-05, + "loss": 0.5954, + "step": 2973 + }, + { + "epoch": 0.95168, + "grad_norm": 0.1441880762577057, + "learning_rate": 8.358762886597939e-05, + "loss": 0.6142, + "step": 2974 + }, + { + "epoch": 0.952, + "grad_norm": 0.15546128153800964, + "learning_rate": 8.354639175257732e-05, + "loss": 0.7998, + "step": 2975 + }, + { + "epoch": 0.95232, + "grad_norm": 0.12989069521427155, + "learning_rate": 8.350515463917527e-05, + "loss": 0.5071, + "step": 2976 + }, + { + "epoch": 0.95264, + "grad_norm": 0.14002223312854767, + "learning_rate": 8.346391752577321e-05, + "loss": 0.6455, + "step": 2977 + }, + { + "epoch": 0.95296, + "grad_norm": 0.1425321400165558, + "learning_rate": 8.342268041237113e-05, + "loss": 0.6249, + "step": 2978 + }, + { + "epoch": 0.95328, + "grad_norm": 0.13261176645755768, + "learning_rate": 8.338144329896907e-05, + "loss": 0.6219, + "step": 2979 + }, + { + "epoch": 0.9536, + "grad_norm": 0.1749666929244995, + "learning_rate": 8.334020618556702e-05, + "loss": 0.6513, + "step": 2980 + }, + { + "epoch": 0.95392, + "grad_norm": 0.1411246806383133, + "learning_rate": 8.329896907216496e-05, + "loss": 0.7347, + "step": 2981 + }, + { + "epoch": 0.95424, + "grad_norm": 0.14908410608768463, + "learning_rate": 8.325773195876288e-05, + "loss": 0.853, + "step": 2982 + }, + { + "epoch": 0.95456, + "grad_norm": 0.14453519880771637, + "learning_rate": 8.321649484536083e-05, + "loss": 0.7926, + "step": 2983 + }, + { + "epoch": 0.95488, + "grad_norm": 0.16018891334533691, + "learning_rate": 8.317525773195877e-05, + "loss": 0.6696, + "step": 2984 + }, + { + "epoch": 0.9552, + "grad_norm": 0.16021482646465302, + "learning_rate": 8.31340206185567e-05, + "loss": 0.611, + "step": 2985 + }, + { + "epoch": 0.95552, + "grad_norm": 0.14570678770542145, + "learning_rate": 8.309278350515465e-05, + "loss": 0.5845, + "step": 2986 + }, + { + "epoch": 0.95584, + "grad_norm": 0.15256036818027496, + "learning_rate": 8.305154639175258e-05, + "loss": 0.6384, + "step": 2987 + }, + { + "epoch": 0.95616, + "grad_norm": 0.13376909494400024, + "learning_rate": 8.301030927835052e-05, + "loss": 0.6959, + "step": 2988 + }, + { + "epoch": 0.95648, + "grad_norm": 0.1550801694393158, + "learning_rate": 8.296907216494845e-05, + "loss": 0.7242, + "step": 2989 + }, + { + "epoch": 0.9568, + "grad_norm": 0.13952943682670593, + "learning_rate": 8.29278350515464e-05, + "loss": 0.6731, + "step": 2990 + }, + { + "epoch": 0.95712, + "grad_norm": 0.13454942405223846, + "learning_rate": 8.288659793814434e-05, + "loss": 0.5457, + "step": 2991 + }, + { + "epoch": 0.95744, + "grad_norm": 0.15188299119472504, + "learning_rate": 8.284536082474227e-05, + "loss": 0.6226, + "step": 2992 + }, + { + "epoch": 0.95776, + "grad_norm": 0.13514840602874756, + "learning_rate": 8.28041237113402e-05, + "loss": 0.6141, + "step": 2993 + }, + { + "epoch": 0.95808, + "grad_norm": 0.1550796627998352, + "learning_rate": 8.276288659793815e-05, + "loss": 0.7065, + "step": 2994 + }, + { + "epoch": 0.9584, + "grad_norm": 0.14826713502407074, + "learning_rate": 8.27216494845361e-05, + "loss": 0.6477, + "step": 2995 + }, + { + "epoch": 0.95872, + "grad_norm": 0.1430700421333313, + "learning_rate": 8.268041237113401e-05, + "loss": 0.7095, + "step": 2996 + }, + { + "epoch": 0.95904, + "grad_norm": 0.13777174055576324, + "learning_rate": 8.263917525773196e-05, + "loss": 0.5355, + "step": 2997 + }, + { + "epoch": 0.95936, + "grad_norm": 0.15069033205509186, + "learning_rate": 8.25979381443299e-05, + "loss": 0.6939, + "step": 2998 + }, + { + "epoch": 0.95968, + "grad_norm": 0.1457386612892151, + "learning_rate": 8.255670103092785e-05, + "loss": 0.6316, + "step": 2999 + }, + { + "epoch": 0.96, + "grad_norm": 0.16817040741443634, + "learning_rate": 8.251546391752578e-05, + "loss": 0.5605, + "step": 3000 + }, + { + "epoch": 0.96032, + "grad_norm": 0.15189550817012787, + "learning_rate": 8.247422680412371e-05, + "loss": 0.6544, + "step": 3001 + }, + { + "epoch": 0.96064, + "grad_norm": 0.15629969537258148, + "learning_rate": 8.243298969072165e-05, + "loss": 0.8185, + "step": 3002 + }, + { + "epoch": 0.96096, + "grad_norm": 0.15958912670612335, + "learning_rate": 8.239175257731959e-05, + "loss": 0.8043, + "step": 3003 + }, + { + "epoch": 0.96128, + "grad_norm": 0.17902502417564392, + "learning_rate": 8.235051546391753e-05, + "loss": 0.7066, + "step": 3004 + }, + { + "epoch": 0.9616, + "grad_norm": 0.12533916532993317, + "learning_rate": 8.230927835051548e-05, + "loss": 0.5782, + "step": 3005 + }, + { + "epoch": 0.96192, + "grad_norm": 0.14869771897792816, + "learning_rate": 8.226804123711341e-05, + "loss": 0.6112, + "step": 3006 + }, + { + "epoch": 0.96224, + "grad_norm": 0.14875675737857819, + "learning_rate": 8.222680412371134e-05, + "loss": 0.7601, + "step": 3007 + }, + { + "epoch": 0.96256, + "grad_norm": 0.16549402475357056, + "learning_rate": 8.218556701030928e-05, + "loss": 0.6784, + "step": 3008 + }, + { + "epoch": 0.96288, + "grad_norm": 0.16122770309448242, + "learning_rate": 8.214432989690723e-05, + "loss": 0.6006, + "step": 3009 + }, + { + "epoch": 0.9632, + "grad_norm": 0.1351601481437683, + "learning_rate": 8.210309278350516e-05, + "loss": 0.5438, + "step": 3010 + }, + { + "epoch": 0.96352, + "grad_norm": 0.16786044836044312, + "learning_rate": 8.206185567010309e-05, + "loss": 0.6346, + "step": 3011 + }, + { + "epoch": 0.96384, + "grad_norm": 0.13561882078647614, + "learning_rate": 8.202061855670104e-05, + "loss": 0.5939, + "step": 3012 + }, + { + "epoch": 0.96416, + "grad_norm": 0.14583711326122284, + "learning_rate": 8.197938144329898e-05, + "loss": 0.6912, + "step": 3013 + }, + { + "epoch": 0.96448, + "grad_norm": 0.1648751199245453, + "learning_rate": 8.193814432989691e-05, + "loss": 0.7473, + "step": 3014 + }, + { + "epoch": 0.9648, + "grad_norm": 0.14872856438159943, + "learning_rate": 8.189690721649484e-05, + "loss": 0.6013, + "step": 3015 + }, + { + "epoch": 0.96512, + "grad_norm": 0.1421995311975479, + "learning_rate": 8.185567010309279e-05, + "loss": 0.7753, + "step": 3016 + }, + { + "epoch": 0.96544, + "grad_norm": 0.15350091457366943, + "learning_rate": 8.181443298969073e-05, + "loss": 0.7044, + "step": 3017 + }, + { + "epoch": 0.96576, + "grad_norm": 0.14795854687690735, + "learning_rate": 8.177319587628866e-05, + "loss": 0.811, + "step": 3018 + }, + { + "epoch": 0.96608, + "grad_norm": 0.15389619767665863, + "learning_rate": 8.173195876288661e-05, + "loss": 0.6684, + "step": 3019 + }, + { + "epoch": 0.9664, + "grad_norm": 0.12923696637153625, + "learning_rate": 8.169072164948454e-05, + "loss": 0.6262, + "step": 3020 + }, + { + "epoch": 0.96672, + "grad_norm": 0.15288755297660828, + "learning_rate": 8.164948453608247e-05, + "loss": 0.6209, + "step": 3021 + }, + { + "epoch": 0.96704, + "grad_norm": 0.15286767482757568, + "learning_rate": 8.160824742268042e-05, + "loss": 0.6094, + "step": 3022 + }, + { + "epoch": 0.96736, + "grad_norm": 0.14733457565307617, + "learning_rate": 8.156701030927836e-05, + "loss": 0.7117, + "step": 3023 + }, + { + "epoch": 0.96768, + "grad_norm": 0.13559021055698395, + "learning_rate": 8.152577319587629e-05, + "loss": 0.5629, + "step": 3024 + }, + { + "epoch": 0.968, + "grad_norm": 0.15902240574359894, + "learning_rate": 8.148453608247422e-05, + "loss": 0.5938, + "step": 3025 + }, + { + "epoch": 0.96832, + "grad_norm": 0.17316369712352753, + "learning_rate": 8.144329896907217e-05, + "loss": 0.7798, + "step": 3026 + }, + { + "epoch": 0.96864, + "grad_norm": 0.15652863681316376, + "learning_rate": 8.140206185567011e-05, + "loss": 0.6532, + "step": 3027 + }, + { + "epoch": 0.96896, + "grad_norm": 0.15096962451934814, + "learning_rate": 8.136082474226804e-05, + "loss": 0.7204, + "step": 3028 + }, + { + "epoch": 0.96928, + "grad_norm": 0.14163367450237274, + "learning_rate": 8.131958762886598e-05, + "loss": 0.6087, + "step": 3029 + }, + { + "epoch": 0.9696, + "grad_norm": 0.1565769612789154, + "learning_rate": 8.127835051546392e-05, + "loss": 0.6306, + "step": 3030 + }, + { + "epoch": 0.96992, + "grad_norm": 0.1648605316877365, + "learning_rate": 8.123711340206187e-05, + "loss": 0.7088, + "step": 3031 + }, + { + "epoch": 0.97024, + "grad_norm": 0.13801802694797516, + "learning_rate": 8.11958762886598e-05, + "loss": 0.6103, + "step": 3032 + }, + { + "epoch": 0.97056, + "grad_norm": 0.1570446640253067, + "learning_rate": 8.115463917525774e-05, + "loss": 0.6716, + "step": 3033 + }, + { + "epoch": 0.97088, + "grad_norm": 0.1504104882478714, + "learning_rate": 8.111340206185567e-05, + "loss": 0.6985, + "step": 3034 + }, + { + "epoch": 0.9712, + "grad_norm": 0.13304787874221802, + "learning_rate": 8.107216494845362e-05, + "loss": 0.5574, + "step": 3035 + }, + { + "epoch": 0.97152, + "grad_norm": 0.15358252823352814, + "learning_rate": 8.103092783505155e-05, + "loss": 0.7928, + "step": 3036 + }, + { + "epoch": 0.97184, + "grad_norm": 0.17936421930789948, + "learning_rate": 8.09896907216495e-05, + "loss": 0.7428, + "step": 3037 + }, + { + "epoch": 0.97216, + "grad_norm": 0.13897693157196045, + "learning_rate": 8.094845360824744e-05, + "loss": 0.6974, + "step": 3038 + }, + { + "epoch": 0.97248, + "grad_norm": 0.12852588295936584, + "learning_rate": 8.090721649484536e-05, + "loss": 0.5747, + "step": 3039 + }, + { + "epoch": 0.9728, + "grad_norm": 0.14601412415504456, + "learning_rate": 8.08659793814433e-05, + "loss": 0.7507, + "step": 3040 + }, + { + "epoch": 0.97312, + "grad_norm": 0.15637436509132385, + "learning_rate": 8.082474226804125e-05, + "loss": 0.6533, + "step": 3041 + }, + { + "epoch": 0.97344, + "grad_norm": 0.15918178856372833, + "learning_rate": 8.078350515463918e-05, + "loss": 0.6355, + "step": 3042 + }, + { + "epoch": 0.97376, + "grad_norm": 0.14796431362628937, + "learning_rate": 8.074226804123711e-05, + "loss": 0.6788, + "step": 3043 + }, + { + "epoch": 0.97408, + "grad_norm": 0.12824980914592743, + "learning_rate": 8.070103092783505e-05, + "loss": 0.6991, + "step": 3044 + }, + { + "epoch": 0.9744, + "grad_norm": 0.13883481919765472, + "learning_rate": 8.0659793814433e-05, + "loss": 0.6398, + "step": 3045 + }, + { + "epoch": 0.97472, + "grad_norm": 0.15688638389110565, + "learning_rate": 8.061855670103093e-05, + "loss": 0.6781, + "step": 3046 + }, + { + "epoch": 0.97504, + "grad_norm": 0.14267170429229736, + "learning_rate": 8.057731958762887e-05, + "loss": 0.5564, + "step": 3047 + }, + { + "epoch": 0.97536, + "grad_norm": 0.15961501002311707, + "learning_rate": 8.05360824742268e-05, + "loss": 0.7607, + "step": 3048 + }, + { + "epoch": 0.97568, + "grad_norm": 0.13472016155719757, + "learning_rate": 8.049484536082475e-05, + "loss": 0.767, + "step": 3049 + }, + { + "epoch": 0.976, + "grad_norm": 0.14500410854816437, + "learning_rate": 8.045360824742268e-05, + "loss": 0.7666, + "step": 3050 + }, + { + "epoch": 0.97632, + "grad_norm": 0.144297793507576, + "learning_rate": 8.041237113402063e-05, + "loss": 0.7517, + "step": 3051 + }, + { + "epoch": 0.97664, + "grad_norm": 0.1413676142692566, + "learning_rate": 8.037113402061857e-05, + "loss": 0.6779, + "step": 3052 + }, + { + "epoch": 0.97696, + "grad_norm": 0.13563159108161926, + "learning_rate": 8.032989690721649e-05, + "loss": 0.663, + "step": 3053 + }, + { + "epoch": 0.97728, + "grad_norm": 0.15525302290916443, + "learning_rate": 8.028865979381443e-05, + "loss": 0.6544, + "step": 3054 + }, + { + "epoch": 0.9776, + "grad_norm": 0.13788056373596191, + "learning_rate": 8.024742268041238e-05, + "loss": 0.7199, + "step": 3055 + }, + { + "epoch": 0.97792, + "grad_norm": 0.1347031444311142, + "learning_rate": 8.020618556701032e-05, + "loss": 0.7213, + "step": 3056 + }, + { + "epoch": 0.97824, + "grad_norm": 0.14042726159095764, + "learning_rate": 8.016494845360824e-05, + "loss": 0.6522, + "step": 3057 + }, + { + "epoch": 0.97856, + "grad_norm": 0.14343152940273285, + "learning_rate": 8.012371134020619e-05, + "loss": 0.6394, + "step": 3058 + }, + { + "epoch": 0.97888, + "grad_norm": 0.1271657645702362, + "learning_rate": 8.008247422680413e-05, + "loss": 0.5985, + "step": 3059 + }, + { + "epoch": 0.9792, + "grad_norm": 0.13452835381031036, + "learning_rate": 8.004123711340206e-05, + "loss": 0.7102, + "step": 3060 + }, + { + "epoch": 0.97952, + "grad_norm": 0.14690479636192322, + "learning_rate": 8e-05, + "loss": 0.6692, + "step": 3061 + }, + { + "epoch": 0.97984, + "grad_norm": 0.14471113681793213, + "learning_rate": 7.995876288659794e-05, + "loss": 0.7379, + "step": 3062 + }, + { + "epoch": 0.98016, + "grad_norm": 0.1589064747095108, + "learning_rate": 7.991752577319588e-05, + "loss": 0.7114, + "step": 3063 + }, + { + "epoch": 0.98048, + "grad_norm": 0.15782232582569122, + "learning_rate": 7.987628865979381e-05, + "loss": 0.7886, + "step": 3064 + }, + { + "epoch": 0.9808, + "grad_norm": 0.13233911991119385, + "learning_rate": 7.983505154639176e-05, + "loss": 0.4364, + "step": 3065 + }, + { + "epoch": 0.98112, + "grad_norm": 0.13761979341506958, + "learning_rate": 7.97938144329897e-05, + "loss": 0.5122, + "step": 3066 + }, + { + "epoch": 0.98144, + "grad_norm": 0.15268053114414215, + "learning_rate": 7.975257731958763e-05, + "loss": 0.6873, + "step": 3067 + }, + { + "epoch": 0.98176, + "grad_norm": 0.1478249728679657, + "learning_rate": 7.971134020618557e-05, + "loss": 0.7437, + "step": 3068 + }, + { + "epoch": 0.98208, + "grad_norm": 0.14960023760795593, + "learning_rate": 7.967010309278351e-05, + "loss": 0.6136, + "step": 3069 + }, + { + "epoch": 0.9824, + "grad_norm": 0.1375860720872879, + "learning_rate": 7.962886597938146e-05, + "loss": 0.6136, + "step": 3070 + }, + { + "epoch": 0.98272, + "grad_norm": 0.1476413905620575, + "learning_rate": 7.958762886597937e-05, + "loss": 0.8405, + "step": 3071 + }, + { + "epoch": 0.98304, + "grad_norm": 0.15509913861751556, + "learning_rate": 7.954639175257732e-05, + "loss": 0.8451, + "step": 3072 + }, + { + "epoch": 0.98336, + "grad_norm": 0.15197688341140747, + "learning_rate": 7.950515463917526e-05, + "loss": 0.6325, + "step": 3073 + }, + { + "epoch": 0.98368, + "grad_norm": 0.15589657425880432, + "learning_rate": 7.946391752577321e-05, + "loss": 0.8487, + "step": 3074 + }, + { + "epoch": 0.984, + "grad_norm": 0.15564191341400146, + "learning_rate": 7.942268041237114e-05, + "loss": 0.6643, + "step": 3075 + }, + { + "epoch": 0.98432, + "grad_norm": 0.14862757921218872, + "learning_rate": 7.938144329896907e-05, + "loss": 0.7741, + "step": 3076 + }, + { + "epoch": 0.98464, + "grad_norm": 0.16020022332668304, + "learning_rate": 7.934020618556702e-05, + "loss": 0.6438, + "step": 3077 + }, + { + "epoch": 0.98496, + "grad_norm": 0.1530873328447342, + "learning_rate": 7.929896907216495e-05, + "loss": 0.792, + "step": 3078 + }, + { + "epoch": 0.98528, + "grad_norm": 0.17370900511741638, + "learning_rate": 7.925773195876289e-05, + "loss": 0.8047, + "step": 3079 + }, + { + "epoch": 0.9856, + "grad_norm": 0.16415753960609436, + "learning_rate": 7.921649484536084e-05, + "loss": 0.6415, + "step": 3080 + }, + { + "epoch": 0.98592, + "grad_norm": 0.15333254635334015, + "learning_rate": 7.917525773195877e-05, + "loss": 0.6169, + "step": 3081 + }, + { + "epoch": 0.98624, + "grad_norm": 0.1595192402601242, + "learning_rate": 7.91340206185567e-05, + "loss": 0.7994, + "step": 3082 + }, + { + "epoch": 0.98656, + "grad_norm": 0.1665579080581665, + "learning_rate": 7.909278350515464e-05, + "loss": 0.7229, + "step": 3083 + }, + { + "epoch": 0.98688, + "grad_norm": 0.15203313529491425, + "learning_rate": 7.905154639175259e-05, + "loss": 0.7225, + "step": 3084 + }, + { + "epoch": 0.9872, + "grad_norm": 0.12581361830234528, + "learning_rate": 7.901030927835052e-05, + "loss": 0.5979, + "step": 3085 + }, + { + "epoch": 0.98752, + "grad_norm": 0.13150061666965485, + "learning_rate": 7.896907216494845e-05, + "loss": 0.5886, + "step": 3086 + }, + { + "epoch": 0.98784, + "grad_norm": 0.16568483412265778, + "learning_rate": 7.89278350515464e-05, + "loss": 0.6729, + "step": 3087 + }, + { + "epoch": 0.98816, + "grad_norm": 0.14777551591396332, + "learning_rate": 7.888659793814434e-05, + "loss": 0.644, + "step": 3088 + }, + { + "epoch": 0.98848, + "grad_norm": 0.1415538638830185, + "learning_rate": 7.884536082474227e-05, + "loss": 0.642, + "step": 3089 + }, + { + "epoch": 0.9888, + "grad_norm": 0.14243784546852112, + "learning_rate": 7.88041237113402e-05, + "loss": 0.5386, + "step": 3090 + }, + { + "epoch": 0.98912, + "grad_norm": 0.1717735379934311, + "learning_rate": 7.876288659793815e-05, + "loss": 0.7391, + "step": 3091 + }, + { + "epoch": 0.98944, + "grad_norm": 0.12800368666648865, + "learning_rate": 7.872164948453609e-05, + "loss": 0.601, + "step": 3092 + }, + { + "epoch": 0.98976, + "grad_norm": 0.14534269273281097, + "learning_rate": 7.868041237113402e-05, + "loss": 0.6653, + "step": 3093 + }, + { + "epoch": 0.99008, + "grad_norm": 0.139258474111557, + "learning_rate": 7.863917525773197e-05, + "loss": 0.713, + "step": 3094 + }, + { + "epoch": 0.9904, + "grad_norm": 0.13516992330551147, + "learning_rate": 7.85979381443299e-05, + "loss": 0.7027, + "step": 3095 + }, + { + "epoch": 0.99072, + "grad_norm": 0.17429058253765106, + "learning_rate": 7.855670103092783e-05, + "loss": 0.6939, + "step": 3096 + }, + { + "epoch": 0.99104, + "grad_norm": 0.14165206253528595, + "learning_rate": 7.851546391752578e-05, + "loss": 0.5702, + "step": 3097 + }, + { + "epoch": 0.99136, + "grad_norm": 0.14981848001480103, + "learning_rate": 7.847422680412372e-05, + "loss": 0.6587, + "step": 3098 + }, + { + "epoch": 0.99168, + "grad_norm": 0.15685059130191803, + "learning_rate": 7.843298969072165e-05, + "loss": 0.8594, + "step": 3099 + }, + { + "epoch": 0.992, + "grad_norm": 0.1411353498697281, + "learning_rate": 7.839175257731958e-05, + "loss": 0.8197, + "step": 3100 + }, + { + "epoch": 0.99232, + "grad_norm": 0.13378973305225372, + "learning_rate": 7.835051546391753e-05, + "loss": 0.7167, + "step": 3101 + }, + { + "epoch": 0.99264, + "grad_norm": 0.14139211177825928, + "learning_rate": 7.830927835051547e-05, + "loss": 0.6995, + "step": 3102 + }, + { + "epoch": 0.99296, + "grad_norm": 0.1355200707912445, + "learning_rate": 7.82680412371134e-05, + "loss": 0.7276, + "step": 3103 + }, + { + "epoch": 0.99328, + "grad_norm": 0.1653222143650055, + "learning_rate": 7.822680412371134e-05, + "loss": 0.7067, + "step": 3104 + }, + { + "epoch": 0.9936, + "grad_norm": 0.1560157835483551, + "learning_rate": 7.818556701030928e-05, + "loss": 0.7061, + "step": 3105 + }, + { + "epoch": 0.99392, + "grad_norm": 0.15817159414291382, + "learning_rate": 7.814432989690723e-05, + "loss": 0.6558, + "step": 3106 + }, + { + "epoch": 0.99424, + "grad_norm": 0.17162838578224182, + "learning_rate": 7.810309278350516e-05, + "loss": 0.6633, + "step": 3107 + }, + { + "epoch": 0.99456, + "grad_norm": 0.15470553934574127, + "learning_rate": 7.80618556701031e-05, + "loss": 0.6646, + "step": 3108 + }, + { + "epoch": 0.99488, + "grad_norm": 0.1391722559928894, + "learning_rate": 7.802061855670103e-05, + "loss": 0.6251, + "step": 3109 + }, + { + "epoch": 0.9952, + "grad_norm": 0.15808038413524628, + "learning_rate": 7.797938144329898e-05, + "loss": 0.7259, + "step": 3110 + }, + { + "epoch": 0.99552, + "grad_norm": 0.14654666185379028, + "learning_rate": 7.793814432989691e-05, + "loss": 0.6831, + "step": 3111 + }, + { + "epoch": 0.99584, + "grad_norm": 0.14982281625270844, + "learning_rate": 7.789690721649485e-05, + "loss": 0.7082, + "step": 3112 + }, + { + "epoch": 0.99616, + "grad_norm": 0.11979085952043533, + "learning_rate": 7.78556701030928e-05, + "loss": 0.6854, + "step": 3113 + }, + { + "epoch": 0.99648, + "grad_norm": 0.1531863957643509, + "learning_rate": 7.781443298969072e-05, + "loss": 0.6789, + "step": 3114 + }, + { + "epoch": 0.9968, + "grad_norm": 0.1712014377117157, + "learning_rate": 7.777319587628866e-05, + "loss": 0.7116, + "step": 3115 + }, + { + "epoch": 0.99712, + "grad_norm": 0.1507626622915268, + "learning_rate": 7.77319587628866e-05, + "loss": 0.7359, + "step": 3116 + }, + { + "epoch": 0.99744, + "grad_norm": 0.14623558521270752, + "learning_rate": 7.769072164948454e-05, + "loss": 0.7649, + "step": 3117 + }, + { + "epoch": 0.99776, + "grad_norm": 0.15552276372909546, + "learning_rate": 7.764948453608247e-05, + "loss": 0.6826, + "step": 3118 + }, + { + "epoch": 0.99808, + "grad_norm": 0.17550498247146606, + "learning_rate": 7.760824742268041e-05, + "loss": 0.784, + "step": 3119 + }, + { + "epoch": 0.9984, + "grad_norm": 0.14296650886535645, + "learning_rate": 7.756701030927836e-05, + "loss": 0.5746, + "step": 3120 + }, + { + "epoch": 0.99872, + "grad_norm": 0.16486896574497223, + "learning_rate": 7.752577319587629e-05, + "loss": 0.6644, + "step": 3121 + }, + { + "epoch": 0.99904, + "grad_norm": 0.13312818109989166, + "learning_rate": 7.748453608247423e-05, + "loss": 0.5503, + "step": 3122 + }, + { + "epoch": 0.99936, + "grad_norm": 0.1439092457294464, + "learning_rate": 7.744329896907217e-05, + "loss": 0.5625, + "step": 3123 + }, + { + "epoch": 0.99968, + "grad_norm": 0.1402236670255661, + "learning_rate": 7.740206185567011e-05, + "loss": 0.564, + "step": 3124 + }, + { + "epoch": 1.0, + "grad_norm": 0.16119587421417236, + "learning_rate": 7.736082474226804e-05, + "loss": 0.6986, + "step": 3125 + }, + { + "epoch": 1.00032, + "grad_norm": 0.14376506209373474, + "learning_rate": 7.731958762886599e-05, + "loss": 0.776, + "step": 3126 + }, + { + "epoch": 1.00064, + "grad_norm": 0.15161509811878204, + "learning_rate": 7.727835051546393e-05, + "loss": 0.5929, + "step": 3127 + }, + { + "epoch": 1.00096, + "grad_norm": 0.15430408716201782, + "learning_rate": 7.723711340206185e-05, + "loss": 0.6505, + "step": 3128 + }, + { + "epoch": 1.00128, + "grad_norm": 0.14096947014331818, + "learning_rate": 7.71958762886598e-05, + "loss": 0.5792, + "step": 3129 + }, + { + "epoch": 1.0016, + "grad_norm": 0.15877479314804077, + "learning_rate": 7.715463917525774e-05, + "loss": 0.6326, + "step": 3130 + }, + { + "epoch": 1.00192, + "grad_norm": 0.1141802966594696, + "learning_rate": 7.711340206185568e-05, + "loss": 0.7297, + "step": 3131 + }, + { + "epoch": 1.00224, + "grad_norm": 0.16603852808475494, + "learning_rate": 7.70721649484536e-05, + "loss": 0.6054, + "step": 3132 + }, + { + "epoch": 1.00256, + "grad_norm": 0.15080897510051727, + "learning_rate": 7.703092783505155e-05, + "loss": 0.7566, + "step": 3133 + }, + { + "epoch": 1.00288, + "grad_norm": 0.14458318054676056, + "learning_rate": 7.698969072164949e-05, + "loss": 0.6773, + "step": 3134 + }, + { + "epoch": 1.0032, + "grad_norm": 0.14412885904312134, + "learning_rate": 7.694845360824742e-05, + "loss": 0.6043, + "step": 3135 + }, + { + "epoch": 1.00352, + "grad_norm": 0.13434572517871857, + "learning_rate": 7.690721649484537e-05, + "loss": 0.5144, + "step": 3136 + }, + { + "epoch": 1.00384, + "grad_norm": 0.13563475012779236, + "learning_rate": 7.68659793814433e-05, + "loss": 0.6369, + "step": 3137 + }, + { + "epoch": 1.00416, + "grad_norm": 0.17308926582336426, + "learning_rate": 7.682474226804124e-05, + "loss": 0.6562, + "step": 3138 + }, + { + "epoch": 1.00448, + "grad_norm": 0.17677193880081177, + "learning_rate": 7.678350515463917e-05, + "loss": 0.6004, + "step": 3139 + }, + { + "epoch": 1.0048, + "grad_norm": 0.1367504894733429, + "learning_rate": 7.674226804123712e-05, + "loss": 0.4425, + "step": 3140 + }, + { + "epoch": 1.00512, + "grad_norm": 0.16985586285591125, + "learning_rate": 7.670103092783506e-05, + "loss": 0.5647, + "step": 3141 + }, + { + "epoch": 1.0054400000000001, + "grad_norm": 0.1637527197599411, + "learning_rate": 7.6659793814433e-05, + "loss": 0.5344, + "step": 3142 + }, + { + "epoch": 1.00576, + "grad_norm": 0.16778002679347992, + "learning_rate": 7.661855670103093e-05, + "loss": 0.5872, + "step": 3143 + }, + { + "epoch": 1.00608, + "grad_norm": 0.14737436175346375, + "learning_rate": 7.657731958762887e-05, + "loss": 0.6518, + "step": 3144 + }, + { + "epoch": 1.0064, + "grad_norm": 0.14423024654388428, + "learning_rate": 7.653608247422682e-05, + "loss": 0.6182, + "step": 3145 + }, + { + "epoch": 1.00672, + "grad_norm": 0.1473619043827057, + "learning_rate": 7.649484536082473e-05, + "loss": 0.5591, + "step": 3146 + }, + { + "epoch": 1.00704, + "grad_norm": 0.16232752799987793, + "learning_rate": 7.645360824742268e-05, + "loss": 0.678, + "step": 3147 + }, + { + "epoch": 1.00736, + "grad_norm": 0.16704897582530975, + "learning_rate": 7.641237113402062e-05, + "loss": 0.6101, + "step": 3148 + }, + { + "epoch": 1.00768, + "grad_norm": 0.14027665555477142, + "learning_rate": 7.637113402061857e-05, + "loss": 0.7055, + "step": 3149 + }, + { + "epoch": 1.008, + "grad_norm": 0.1610679179430008, + "learning_rate": 7.63298969072165e-05, + "loss": 0.5775, + "step": 3150 + }, + { + "epoch": 1.00832, + "grad_norm": 0.13646800816059113, + "learning_rate": 7.628865979381443e-05, + "loss": 0.496, + "step": 3151 + }, + { + "epoch": 1.00864, + "grad_norm": 0.18376730382442474, + "learning_rate": 7.624742268041238e-05, + "loss": 0.6249, + "step": 3152 + }, + { + "epoch": 1.00896, + "grad_norm": 0.1544225960969925, + "learning_rate": 7.620618556701031e-05, + "loss": 0.6182, + "step": 3153 + }, + { + "epoch": 1.00928, + "grad_norm": 0.16254231333732605, + "learning_rate": 7.616494845360825e-05, + "loss": 0.6787, + "step": 3154 + }, + { + "epoch": 1.0096, + "grad_norm": 0.1361733227968216, + "learning_rate": 7.61237113402062e-05, + "loss": 0.7957, + "step": 3155 + }, + { + "epoch": 1.00992, + "grad_norm": 0.15801657736301422, + "learning_rate": 7.608247422680413e-05, + "loss": 0.6499, + "step": 3156 + }, + { + "epoch": 1.01024, + "grad_norm": 0.15821288526058197, + "learning_rate": 7.604123711340206e-05, + "loss": 0.5367, + "step": 3157 + }, + { + "epoch": 1.01056, + "grad_norm": 0.1741386204957962, + "learning_rate": 7.6e-05, + "loss": 0.6651, + "step": 3158 + }, + { + "epoch": 1.01088, + "grad_norm": 0.16049474477767944, + "learning_rate": 7.595876288659795e-05, + "loss": 0.5797, + "step": 3159 + }, + { + "epoch": 1.0112, + "grad_norm": 0.18430651724338531, + "learning_rate": 7.591752577319588e-05, + "loss": 0.6687, + "step": 3160 + }, + { + "epoch": 1.01152, + "grad_norm": 0.13597476482391357, + "learning_rate": 7.587628865979381e-05, + "loss": 0.4701, + "step": 3161 + }, + { + "epoch": 1.01184, + "grad_norm": 0.1269109845161438, + "learning_rate": 7.583505154639176e-05, + "loss": 0.6077, + "step": 3162 + }, + { + "epoch": 1.01216, + "grad_norm": 0.15657393634319305, + "learning_rate": 7.57938144329897e-05, + "loss": 0.6735, + "step": 3163 + }, + { + "epoch": 1.01248, + "grad_norm": 0.1470077931880951, + "learning_rate": 7.575257731958763e-05, + "loss": 0.6422, + "step": 3164 + }, + { + "epoch": 1.0128, + "grad_norm": 0.16874688863754272, + "learning_rate": 7.571134020618556e-05, + "loss": 0.7393, + "step": 3165 + }, + { + "epoch": 1.01312, + "grad_norm": 0.15357494354248047, + "learning_rate": 7.567010309278351e-05, + "loss": 0.6896, + "step": 3166 + }, + { + "epoch": 1.01344, + "grad_norm": 0.13379424810409546, + "learning_rate": 7.562886597938145e-05, + "loss": 0.6984, + "step": 3167 + }, + { + "epoch": 1.01376, + "grad_norm": 0.13537608087062836, + "learning_rate": 7.558762886597938e-05, + "loss": 0.578, + "step": 3168 + }, + { + "epoch": 1.01408, + "grad_norm": 0.16570653021335602, + "learning_rate": 7.554639175257733e-05, + "loss": 0.6028, + "step": 3169 + }, + { + "epoch": 1.0144, + "grad_norm": 0.1635371744632721, + "learning_rate": 7.550515463917526e-05, + "loss": 0.7298, + "step": 3170 + }, + { + "epoch": 1.01472, + "grad_norm": 0.16552339494228363, + "learning_rate": 7.546391752577319e-05, + "loss": 0.7123, + "step": 3171 + }, + { + "epoch": 1.01504, + "grad_norm": 0.16300083696842194, + "learning_rate": 7.542268041237114e-05, + "loss": 0.6619, + "step": 3172 + }, + { + "epoch": 1.01536, + "grad_norm": 0.15255211293697357, + "learning_rate": 7.538144329896908e-05, + "loss": 0.6497, + "step": 3173 + }, + { + "epoch": 1.01568, + "grad_norm": 0.13859154284000397, + "learning_rate": 7.534020618556701e-05, + "loss": 0.6106, + "step": 3174 + }, + { + "epoch": 1.016, + "grad_norm": 0.1487518548965454, + "learning_rate": 7.529896907216494e-05, + "loss": 0.6098, + "step": 3175 + }, + { + "epoch": 1.01632, + "grad_norm": 0.1739204227924347, + "learning_rate": 7.525773195876289e-05, + "loss": 0.6819, + "step": 3176 + }, + { + "epoch": 1.01664, + "grad_norm": 0.1509557068347931, + "learning_rate": 7.521649484536083e-05, + "loss": 0.5012, + "step": 3177 + }, + { + "epoch": 1.01696, + "grad_norm": 0.15069057047367096, + "learning_rate": 7.517525773195877e-05, + "loss": 0.6097, + "step": 3178 + }, + { + "epoch": 1.01728, + "grad_norm": 0.15466809272766113, + "learning_rate": 7.51340206185567e-05, + "loss": 0.6174, + "step": 3179 + }, + { + "epoch": 1.0176, + "grad_norm": 0.17535479366779327, + "learning_rate": 7.509278350515464e-05, + "loss": 0.6474, + "step": 3180 + }, + { + "epoch": 1.01792, + "grad_norm": 0.14726130664348602, + "learning_rate": 7.505154639175259e-05, + "loss": 0.6839, + "step": 3181 + }, + { + "epoch": 1.01824, + "grad_norm": 0.1776508241891861, + "learning_rate": 7.501030927835052e-05, + "loss": 0.7955, + "step": 3182 + }, + { + "epoch": 1.01856, + "grad_norm": 0.15428778529167175, + "learning_rate": 7.496907216494846e-05, + "loss": 0.8096, + "step": 3183 + }, + { + "epoch": 1.01888, + "grad_norm": 0.1336577981710434, + "learning_rate": 7.49278350515464e-05, + "loss": 0.6354, + "step": 3184 + }, + { + "epoch": 1.0192, + "grad_norm": 0.16552120447158813, + "learning_rate": 7.488659793814434e-05, + "loss": 0.6482, + "step": 3185 + }, + { + "epoch": 1.01952, + "grad_norm": 0.14349189400672913, + "learning_rate": 7.484536082474227e-05, + "loss": 0.6769, + "step": 3186 + }, + { + "epoch": 1.01984, + "grad_norm": 0.18180744349956512, + "learning_rate": 7.480412371134021e-05, + "loss": 0.6688, + "step": 3187 + }, + { + "epoch": 1.02016, + "grad_norm": 0.1554415225982666, + "learning_rate": 7.476288659793816e-05, + "loss": 0.751, + "step": 3188 + }, + { + "epoch": 1.02048, + "grad_norm": 0.15756060183048248, + "learning_rate": 7.472164948453608e-05, + "loss": 0.6845, + "step": 3189 + }, + { + "epoch": 1.0208, + "grad_norm": 0.15304717421531677, + "learning_rate": 7.468041237113402e-05, + "loss": 0.7468, + "step": 3190 + }, + { + "epoch": 1.02112, + "grad_norm": 0.13660815358161926, + "learning_rate": 7.463917525773197e-05, + "loss": 0.428, + "step": 3191 + }, + { + "epoch": 1.02144, + "grad_norm": 0.15597715973854065, + "learning_rate": 7.45979381443299e-05, + "loss": 0.6284, + "step": 3192 + }, + { + "epoch": 1.02176, + "grad_norm": 0.17346197366714478, + "learning_rate": 7.455670103092783e-05, + "loss": 0.6528, + "step": 3193 + }, + { + "epoch": 1.02208, + "grad_norm": 0.1622658520936966, + "learning_rate": 7.451546391752577e-05, + "loss": 0.6016, + "step": 3194 + }, + { + "epoch": 1.0224, + "grad_norm": 0.1548348069190979, + "learning_rate": 7.447422680412372e-05, + "loss": 0.6342, + "step": 3195 + }, + { + "epoch": 1.02272, + "grad_norm": 0.14985868334770203, + "learning_rate": 7.443298969072165e-05, + "loss": 0.6966, + "step": 3196 + }, + { + "epoch": 1.02304, + "grad_norm": 0.17515185475349426, + "learning_rate": 7.43917525773196e-05, + "loss": 0.72, + "step": 3197 + }, + { + "epoch": 1.02336, + "grad_norm": 0.1670687347650528, + "learning_rate": 7.435051546391753e-05, + "loss": 0.5848, + "step": 3198 + }, + { + "epoch": 1.02368, + "grad_norm": 0.15915238857269287, + "learning_rate": 7.430927835051547e-05, + "loss": 0.5614, + "step": 3199 + }, + { + "epoch": 1.024, + "grad_norm": 0.14589542150497437, + "learning_rate": 7.42680412371134e-05, + "loss": 0.7186, + "step": 3200 + }, + { + "epoch": 1.02432, + "grad_norm": 0.15243183076381683, + "learning_rate": 7.422680412371135e-05, + "loss": 0.7588, + "step": 3201 + }, + { + "epoch": 1.02464, + "grad_norm": 0.1360102891921997, + "learning_rate": 7.418556701030929e-05, + "loss": 0.6344, + "step": 3202 + }, + { + "epoch": 1.02496, + "grad_norm": 0.13525979220867157, + "learning_rate": 7.414432989690721e-05, + "loss": 0.6815, + "step": 3203 + }, + { + "epoch": 1.02528, + "grad_norm": 0.17362073063850403, + "learning_rate": 7.410309278350515e-05, + "loss": 0.7282, + "step": 3204 + }, + { + "epoch": 1.0256, + "grad_norm": 0.15797258913516998, + "learning_rate": 7.40618556701031e-05, + "loss": 0.6082, + "step": 3205 + }, + { + "epoch": 1.02592, + "grad_norm": 0.13886618614196777, + "learning_rate": 7.402061855670104e-05, + "loss": 0.6886, + "step": 3206 + }, + { + "epoch": 1.02624, + "grad_norm": 0.14410024881362915, + "learning_rate": 7.397938144329896e-05, + "loss": 0.7866, + "step": 3207 + }, + { + "epoch": 1.02656, + "grad_norm": 0.16933733224868774, + "learning_rate": 7.39381443298969e-05, + "loss": 0.6122, + "step": 3208 + }, + { + "epoch": 1.02688, + "grad_norm": 0.14775492250919342, + "learning_rate": 7.389690721649485e-05, + "loss": 0.6663, + "step": 3209 + }, + { + "epoch": 1.0272, + "grad_norm": 0.15294332802295685, + "learning_rate": 7.385567010309278e-05, + "loss": 0.6329, + "step": 3210 + }, + { + "epoch": 1.02752, + "grad_norm": 0.1664023995399475, + "learning_rate": 7.381443298969073e-05, + "loss": 0.5899, + "step": 3211 + }, + { + "epoch": 1.02784, + "grad_norm": 0.15477019548416138, + "learning_rate": 7.377319587628866e-05, + "loss": 0.649, + "step": 3212 + }, + { + "epoch": 1.02816, + "grad_norm": 0.16100965440273285, + "learning_rate": 7.37319587628866e-05, + "loss": 0.5406, + "step": 3213 + }, + { + "epoch": 1.02848, + "grad_norm": 0.16894161701202393, + "learning_rate": 7.369072164948453e-05, + "loss": 0.5609, + "step": 3214 + }, + { + "epoch": 1.0288, + "grad_norm": 0.18415647745132446, + "learning_rate": 7.364948453608248e-05, + "loss": 0.571, + "step": 3215 + }, + { + "epoch": 1.02912, + "grad_norm": 0.13336199522018433, + "learning_rate": 7.360824742268042e-05, + "loss": 0.5143, + "step": 3216 + }, + { + "epoch": 1.02944, + "grad_norm": 0.15968799591064453, + "learning_rate": 7.356701030927836e-05, + "loss": 0.5672, + "step": 3217 + }, + { + "epoch": 1.02976, + "grad_norm": 0.16611017286777496, + "learning_rate": 7.352577319587629e-05, + "loss": 0.6084, + "step": 3218 + }, + { + "epoch": 1.03008, + "grad_norm": 0.1370030641555786, + "learning_rate": 7.348453608247423e-05, + "loss": 0.6123, + "step": 3219 + }, + { + "epoch": 1.0304, + "grad_norm": 0.1765139400959015, + "learning_rate": 7.344329896907218e-05, + "loss": 0.6307, + "step": 3220 + }, + { + "epoch": 1.03072, + "grad_norm": 0.14999984204769135, + "learning_rate": 7.34020618556701e-05, + "loss": 0.7204, + "step": 3221 + }, + { + "epoch": 1.03104, + "grad_norm": 0.13516376912593842, + "learning_rate": 7.336082474226804e-05, + "loss": 0.4545, + "step": 3222 + }, + { + "epoch": 1.03136, + "grad_norm": 0.15037202835083008, + "learning_rate": 7.331958762886598e-05, + "loss": 0.5407, + "step": 3223 + }, + { + "epoch": 1.03168, + "grad_norm": 0.16515859961509705, + "learning_rate": 7.327835051546393e-05, + "loss": 0.7902, + "step": 3224 + }, + { + "epoch": 1.032, + "grad_norm": 0.15820130705833435, + "learning_rate": 7.323711340206186e-05, + "loss": 0.5323, + "step": 3225 + }, + { + "epoch": 1.03232, + "grad_norm": 0.1813393384218216, + "learning_rate": 7.319587628865979e-05, + "loss": 0.5441, + "step": 3226 + }, + { + "epoch": 1.03264, + "grad_norm": 0.1599324643611908, + "learning_rate": 7.315463917525774e-05, + "loss": 0.7387, + "step": 3227 + }, + { + "epoch": 1.03296, + "grad_norm": 0.15043006837368011, + "learning_rate": 7.311340206185567e-05, + "loss": 0.6578, + "step": 3228 + }, + { + "epoch": 1.03328, + "grad_norm": 0.1628497838973999, + "learning_rate": 7.307216494845361e-05, + "loss": 0.5581, + "step": 3229 + }, + { + "epoch": 1.0336, + "grad_norm": 0.15535464882850647, + "learning_rate": 7.303092783505156e-05, + "loss": 0.4492, + "step": 3230 + }, + { + "epoch": 1.03392, + "grad_norm": 0.17164427042007446, + "learning_rate": 7.298969072164949e-05, + "loss": 0.6095, + "step": 3231 + }, + { + "epoch": 1.03424, + "grad_norm": 0.12054961919784546, + "learning_rate": 7.294845360824742e-05, + "loss": 0.4767, + "step": 3232 + }, + { + "epoch": 1.03456, + "grad_norm": 0.16578055918216705, + "learning_rate": 7.290721649484536e-05, + "loss": 0.6648, + "step": 3233 + }, + { + "epoch": 1.03488, + "grad_norm": 0.14720064401626587, + "learning_rate": 7.286597938144331e-05, + "loss": 0.6055, + "step": 3234 + }, + { + "epoch": 1.0352, + "grad_norm": 0.1783543974161148, + "learning_rate": 7.282474226804124e-05, + "loss": 0.6477, + "step": 3235 + }, + { + "epoch": 1.03552, + "grad_norm": 0.15802974998950958, + "learning_rate": 7.278350515463917e-05, + "loss": 0.5497, + "step": 3236 + }, + { + "epoch": 1.03584, + "grad_norm": 0.15459182858467102, + "learning_rate": 7.274226804123712e-05, + "loss": 0.5117, + "step": 3237 + }, + { + "epoch": 1.03616, + "grad_norm": 0.15035270154476166, + "learning_rate": 7.270103092783506e-05, + "loss": 0.5575, + "step": 3238 + }, + { + "epoch": 1.03648, + "grad_norm": 0.16026097536087036, + "learning_rate": 7.265979381443299e-05, + "loss": 0.6079, + "step": 3239 + }, + { + "epoch": 1.0368, + "grad_norm": 0.13876131176948547, + "learning_rate": 7.261855670103092e-05, + "loss": 0.5959, + "step": 3240 + }, + { + "epoch": 1.03712, + "grad_norm": 0.14803935587406158, + "learning_rate": 7.257731958762887e-05, + "loss": 0.5759, + "step": 3241 + }, + { + "epoch": 1.03744, + "grad_norm": 0.17045043408870697, + "learning_rate": 7.253608247422681e-05, + "loss": 0.7337, + "step": 3242 + }, + { + "epoch": 1.03776, + "grad_norm": 0.14425960183143616, + "learning_rate": 7.249484536082475e-05, + "loss": 0.5783, + "step": 3243 + }, + { + "epoch": 1.03808, + "grad_norm": 0.162851020693779, + "learning_rate": 7.245360824742269e-05, + "loss": 0.5868, + "step": 3244 + }, + { + "epoch": 1.0384, + "grad_norm": 0.14762966334819794, + "learning_rate": 7.241237113402062e-05, + "loss": 0.7355, + "step": 3245 + }, + { + "epoch": 1.03872, + "grad_norm": 0.14285781979560852, + "learning_rate": 7.237113402061855e-05, + "loss": 0.4501, + "step": 3246 + }, + { + "epoch": 1.03904, + "grad_norm": 0.17376355826854706, + "learning_rate": 7.23298969072165e-05, + "loss": 0.6795, + "step": 3247 + }, + { + "epoch": 1.03936, + "grad_norm": 0.18657740950584412, + "learning_rate": 7.228865979381444e-05, + "loss": 0.6802, + "step": 3248 + }, + { + "epoch": 1.03968, + "grad_norm": 0.15594154596328735, + "learning_rate": 7.224742268041237e-05, + "loss": 0.5651, + "step": 3249 + }, + { + "epoch": 1.04, + "grad_norm": 0.16138726472854614, + "learning_rate": 7.22061855670103e-05, + "loss": 0.8319, + "step": 3250 + }, + { + "epoch": 1.04032, + "grad_norm": 0.1524965614080429, + "learning_rate": 7.216494845360825e-05, + "loss": 0.5944, + "step": 3251 + }, + { + "epoch": 1.04064, + "grad_norm": 0.16720843315124512, + "learning_rate": 7.21237113402062e-05, + "loss": 0.6276, + "step": 3252 + }, + { + "epoch": 1.04096, + "grad_norm": 0.14803019165992737, + "learning_rate": 7.208247422680413e-05, + "loss": 0.6452, + "step": 3253 + }, + { + "epoch": 1.04128, + "grad_norm": 0.15392228960990906, + "learning_rate": 7.204123711340206e-05, + "loss": 0.7125, + "step": 3254 + }, + { + "epoch": 1.0416, + "grad_norm": 0.14675074815750122, + "learning_rate": 7.2e-05, + "loss": 0.6749, + "step": 3255 + }, + { + "epoch": 1.04192, + "grad_norm": 0.1780073493719101, + "learning_rate": 7.195876288659795e-05, + "loss": 0.7117, + "step": 3256 + }, + { + "epoch": 1.04224, + "grad_norm": 0.15796047449111938, + "learning_rate": 7.191752577319588e-05, + "loss": 0.6535, + "step": 3257 + }, + { + "epoch": 1.04256, + "grad_norm": 0.1566307246685028, + "learning_rate": 7.187628865979382e-05, + "loss": 0.7495, + "step": 3258 + }, + { + "epoch": 1.04288, + "grad_norm": 0.16615964472293854, + "learning_rate": 7.183505154639175e-05, + "loss": 0.5215, + "step": 3259 + }, + { + "epoch": 1.0432, + "grad_norm": 0.16231288015842438, + "learning_rate": 7.17938144329897e-05, + "loss": 0.7243, + "step": 3260 + }, + { + "epoch": 1.04352, + "grad_norm": 0.16023856401443481, + "learning_rate": 7.175257731958763e-05, + "loss": 0.6108, + "step": 3261 + }, + { + "epoch": 1.04384, + "grad_norm": 0.17148688435554504, + "learning_rate": 7.171134020618557e-05, + "loss": 0.6303, + "step": 3262 + }, + { + "epoch": 1.04416, + "grad_norm": 0.17184311151504517, + "learning_rate": 7.167010309278352e-05, + "loss": 0.6225, + "step": 3263 + }, + { + "epoch": 1.04448, + "grad_norm": 0.1528804749250412, + "learning_rate": 7.162886597938144e-05, + "loss": 0.5807, + "step": 3264 + }, + { + "epoch": 1.0448, + "grad_norm": 0.1631060540676117, + "learning_rate": 7.158762886597938e-05, + "loss": 0.6401, + "step": 3265 + }, + { + "epoch": 1.04512, + "grad_norm": 0.14579348266124725, + "learning_rate": 7.154639175257733e-05, + "loss": 0.6612, + "step": 3266 + }, + { + "epoch": 1.04544, + "grad_norm": 0.172668918967247, + "learning_rate": 7.150515463917526e-05, + "loss": 0.7047, + "step": 3267 + }, + { + "epoch": 1.04576, + "grad_norm": 0.16685406863689423, + "learning_rate": 7.146391752577319e-05, + "loss": 0.6436, + "step": 3268 + }, + { + "epoch": 1.04608, + "grad_norm": 0.17377406358718872, + "learning_rate": 7.142268041237113e-05, + "loss": 0.7068, + "step": 3269 + }, + { + "epoch": 1.0464, + "grad_norm": 0.15471984446048737, + "learning_rate": 7.138144329896908e-05, + "loss": 0.5277, + "step": 3270 + }, + { + "epoch": 1.04672, + "grad_norm": 0.1381632685661316, + "learning_rate": 7.134020618556701e-05, + "loss": 0.49, + "step": 3271 + }, + { + "epoch": 1.04704, + "grad_norm": 0.1431090384721756, + "learning_rate": 7.129896907216496e-05, + "loss": 0.5934, + "step": 3272 + }, + { + "epoch": 1.04736, + "grad_norm": 0.15086936950683594, + "learning_rate": 7.125773195876289e-05, + "loss": 0.6157, + "step": 3273 + }, + { + "epoch": 1.04768, + "grad_norm": 0.15616509318351746, + "learning_rate": 7.121649484536083e-05, + "loss": 0.527, + "step": 3274 + }, + { + "epoch": 1.048, + "grad_norm": 0.15351109206676483, + "learning_rate": 7.117525773195876e-05, + "loss": 0.6324, + "step": 3275 + }, + { + "epoch": 1.04832, + "grad_norm": 0.15163224935531616, + "learning_rate": 7.113402061855671e-05, + "loss": 0.5924, + "step": 3276 + }, + { + "epoch": 1.04864, + "grad_norm": 0.1682780534029007, + "learning_rate": 7.109278350515465e-05, + "loss": 0.6493, + "step": 3277 + }, + { + "epoch": 1.04896, + "grad_norm": 0.14303894340991974, + "learning_rate": 7.105154639175258e-05, + "loss": 0.5073, + "step": 3278 + }, + { + "epoch": 1.04928, + "grad_norm": 0.16888777911663055, + "learning_rate": 7.101030927835052e-05, + "loss": 0.6284, + "step": 3279 + }, + { + "epoch": 1.0496, + "grad_norm": 0.16914886236190796, + "learning_rate": 7.096907216494846e-05, + "loss": 0.7228, + "step": 3280 + }, + { + "epoch": 1.04992, + "grad_norm": 0.15475879609584808, + "learning_rate": 7.09278350515464e-05, + "loss": 0.7003, + "step": 3281 + }, + { + "epoch": 1.05024, + "grad_norm": 0.15933775901794434, + "learning_rate": 7.088659793814432e-05, + "loss": 0.6266, + "step": 3282 + }, + { + "epoch": 1.05056, + "grad_norm": 0.1561759114265442, + "learning_rate": 7.084536082474227e-05, + "loss": 0.5964, + "step": 3283 + }, + { + "epoch": 1.05088, + "grad_norm": 0.16218483448028564, + "learning_rate": 7.080412371134021e-05, + "loss": 0.6181, + "step": 3284 + }, + { + "epoch": 1.0512, + "grad_norm": 0.15902891755104065, + "learning_rate": 7.076288659793814e-05, + "loss": 0.6627, + "step": 3285 + }, + { + "epoch": 1.05152, + "grad_norm": 0.15477587282657623, + "learning_rate": 7.072164948453609e-05, + "loss": 0.7069, + "step": 3286 + }, + { + "epoch": 1.0518399999999999, + "grad_norm": 0.17137540876865387, + "learning_rate": 7.068041237113402e-05, + "loss": 0.6075, + "step": 3287 + }, + { + "epoch": 1.05216, + "grad_norm": 0.18496264517307281, + "learning_rate": 7.063917525773196e-05, + "loss": 0.599, + "step": 3288 + }, + { + "epoch": 1.05248, + "grad_norm": 0.15534865856170654, + "learning_rate": 7.05979381443299e-05, + "loss": 0.6313, + "step": 3289 + }, + { + "epoch": 1.0528, + "grad_norm": 0.16796620190143585, + "learning_rate": 7.055670103092784e-05, + "loss": 0.6253, + "step": 3290 + }, + { + "epoch": 1.05312, + "grad_norm": 0.15389969944953918, + "learning_rate": 7.051546391752579e-05, + "loss": 0.6457, + "step": 3291 + }, + { + "epoch": 1.05344, + "grad_norm": 0.14354264736175537, + "learning_rate": 7.047422680412372e-05, + "loss": 0.6155, + "step": 3292 + }, + { + "epoch": 1.05376, + "grad_norm": 0.1503264605998993, + "learning_rate": 7.043298969072165e-05, + "loss": 0.4718, + "step": 3293 + }, + { + "epoch": 1.05408, + "grad_norm": 0.16295716166496277, + "learning_rate": 7.039175257731959e-05, + "loss": 0.647, + "step": 3294 + }, + { + "epoch": 1.0544, + "grad_norm": 0.14746500551700592, + "learning_rate": 7.035051546391754e-05, + "loss": 0.4913, + "step": 3295 + }, + { + "epoch": 1.05472, + "grad_norm": 0.14877967536449432, + "learning_rate": 7.030927835051546e-05, + "loss": 0.5637, + "step": 3296 + }, + { + "epoch": 1.05504, + "grad_norm": 0.16050051152706146, + "learning_rate": 7.02680412371134e-05, + "loss": 0.7172, + "step": 3297 + }, + { + "epoch": 1.05536, + "grad_norm": 0.16275620460510254, + "learning_rate": 7.022680412371134e-05, + "loss": 0.817, + "step": 3298 + }, + { + "epoch": 1.05568, + "grad_norm": 0.17402403056621552, + "learning_rate": 7.018556701030929e-05, + "loss": 0.6539, + "step": 3299 + }, + { + "epoch": 1.056, + "grad_norm": 0.16328458487987518, + "learning_rate": 7.014432989690722e-05, + "loss": 0.6837, + "step": 3300 + }, + { + "epoch": 1.05632, + "grad_norm": 0.17684189975261688, + "learning_rate": 7.010309278350515e-05, + "loss": 0.6295, + "step": 3301 + }, + { + "epoch": 1.05664, + "grad_norm": 0.14302825927734375, + "learning_rate": 7.00618556701031e-05, + "loss": 0.6465, + "step": 3302 + }, + { + "epoch": 1.05696, + "grad_norm": 0.15395498275756836, + "learning_rate": 7.002061855670103e-05, + "loss": 0.7057, + "step": 3303 + }, + { + "epoch": 1.05728, + "grad_norm": 0.19036656618118286, + "learning_rate": 6.997938144329897e-05, + "loss": 0.6346, + "step": 3304 + }, + { + "epoch": 1.0576, + "grad_norm": 0.14019374549388885, + "learning_rate": 6.993814432989692e-05, + "loss": 0.6769, + "step": 3305 + }, + { + "epoch": 1.05792, + "grad_norm": 0.1598007082939148, + "learning_rate": 6.989690721649485e-05, + "loss": 0.6182, + "step": 3306 + }, + { + "epoch": 1.05824, + "grad_norm": 0.1780097633600235, + "learning_rate": 6.985567010309278e-05, + "loss": 0.6103, + "step": 3307 + }, + { + "epoch": 1.05856, + "grad_norm": 0.1597011387348175, + "learning_rate": 6.981443298969073e-05, + "loss": 0.6446, + "step": 3308 + }, + { + "epoch": 1.05888, + "grad_norm": 0.14037050306797028, + "learning_rate": 6.977319587628867e-05, + "loss": 0.7254, + "step": 3309 + }, + { + "epoch": 1.0592, + "grad_norm": 0.16348814964294434, + "learning_rate": 6.97319587628866e-05, + "loss": 0.61, + "step": 3310 + }, + { + "epoch": 1.05952, + "grad_norm": 0.15756426751613617, + "learning_rate": 6.969072164948453e-05, + "loss": 0.6166, + "step": 3311 + }, + { + "epoch": 1.05984, + "grad_norm": 0.18667972087860107, + "learning_rate": 6.964948453608248e-05, + "loss": 0.6178, + "step": 3312 + }, + { + "epoch": 1.06016, + "grad_norm": 0.18235279619693756, + "learning_rate": 6.960824742268042e-05, + "loss": 0.5492, + "step": 3313 + }, + { + "epoch": 1.06048, + "grad_norm": 0.14999398589134216, + "learning_rate": 6.956701030927835e-05, + "loss": 0.5152, + "step": 3314 + }, + { + "epoch": 1.0608, + "grad_norm": 0.13952407240867615, + "learning_rate": 6.952577319587628e-05, + "loss": 0.5705, + "step": 3315 + }, + { + "epoch": 1.06112, + "grad_norm": 0.16914765536785126, + "learning_rate": 6.948453608247423e-05, + "loss": 0.5869, + "step": 3316 + }, + { + "epoch": 1.06144, + "grad_norm": 0.16051514446735382, + "learning_rate": 6.944329896907217e-05, + "loss": 0.6457, + "step": 3317 + }, + { + "epoch": 1.06176, + "grad_norm": 0.1702883541584015, + "learning_rate": 6.94020618556701e-05, + "loss": 0.7861, + "step": 3318 + }, + { + "epoch": 1.06208, + "grad_norm": 0.16490407288074493, + "learning_rate": 6.936082474226805e-05, + "loss": 0.7518, + "step": 3319 + }, + { + "epoch": 1.0624, + "grad_norm": 0.15156599879264832, + "learning_rate": 6.931958762886598e-05, + "loss": 0.6225, + "step": 3320 + }, + { + "epoch": 1.06272, + "grad_norm": 0.16130469739437103, + "learning_rate": 6.927835051546391e-05, + "loss": 0.6075, + "step": 3321 + }, + { + "epoch": 1.06304, + "grad_norm": 0.17261925339698792, + "learning_rate": 6.923711340206186e-05, + "loss": 0.6219, + "step": 3322 + }, + { + "epoch": 1.06336, + "grad_norm": 0.15308493375778198, + "learning_rate": 6.91958762886598e-05, + "loss": 0.6939, + "step": 3323 + }, + { + "epoch": 1.06368, + "grad_norm": 0.1413232982158661, + "learning_rate": 6.915463917525773e-05, + "loss": 0.5196, + "step": 3324 + }, + { + "epoch": 1.064, + "grad_norm": 0.1675940304994583, + "learning_rate": 6.911340206185567e-05, + "loss": 0.6209, + "step": 3325 + }, + { + "epoch": 1.06432, + "grad_norm": 0.1978452056646347, + "learning_rate": 6.907216494845361e-05, + "loss": 0.6959, + "step": 3326 + }, + { + "epoch": 1.06464, + "grad_norm": 0.15386295318603516, + "learning_rate": 6.903092783505155e-05, + "loss": 0.6692, + "step": 3327 + }, + { + "epoch": 1.06496, + "grad_norm": 0.1569909006357193, + "learning_rate": 6.898969072164949e-05, + "loss": 0.596, + "step": 3328 + }, + { + "epoch": 1.06528, + "grad_norm": 0.16718021035194397, + "learning_rate": 6.894845360824742e-05, + "loss": 0.6349, + "step": 3329 + }, + { + "epoch": 1.0656, + "grad_norm": 0.1535327136516571, + "learning_rate": 6.890721649484536e-05, + "loss": 0.5497, + "step": 3330 + }, + { + "epoch": 1.06592, + "grad_norm": 0.14143691956996918, + "learning_rate": 6.886597938144331e-05, + "loss": 0.5274, + "step": 3331 + }, + { + "epoch": 1.06624, + "grad_norm": 0.1517411768436432, + "learning_rate": 6.882474226804124e-05, + "loss": 0.6962, + "step": 3332 + }, + { + "epoch": 1.06656, + "grad_norm": 0.14961174130439758, + "learning_rate": 6.878350515463918e-05, + "loss": 0.541, + "step": 3333 + }, + { + "epoch": 1.06688, + "grad_norm": 0.15757699310779572, + "learning_rate": 6.874226804123711e-05, + "loss": 0.6236, + "step": 3334 + }, + { + "epoch": 1.0672, + "grad_norm": 0.16411931812763214, + "learning_rate": 6.870103092783506e-05, + "loss": 0.4759, + "step": 3335 + }, + { + "epoch": 1.06752, + "grad_norm": 0.1594354212284088, + "learning_rate": 6.865979381443299e-05, + "loss": 0.6816, + "step": 3336 + }, + { + "epoch": 1.06784, + "grad_norm": 0.14267060160636902, + "learning_rate": 6.861855670103094e-05, + "loss": 0.5101, + "step": 3337 + }, + { + "epoch": 1.06816, + "grad_norm": 0.17744643986225128, + "learning_rate": 6.857731958762888e-05, + "loss": 0.743, + "step": 3338 + }, + { + "epoch": 1.06848, + "grad_norm": 0.16844414174556732, + "learning_rate": 6.85360824742268e-05, + "loss": 0.6629, + "step": 3339 + }, + { + "epoch": 1.0688, + "grad_norm": 0.17188021540641785, + "learning_rate": 6.849484536082474e-05, + "loss": 0.6924, + "step": 3340 + }, + { + "epoch": 1.06912, + "grad_norm": 0.1579052209854126, + "learning_rate": 6.845360824742269e-05, + "loss": 0.6971, + "step": 3341 + }, + { + "epoch": 1.06944, + "grad_norm": 0.1879045069217682, + "learning_rate": 6.841237113402062e-05, + "loss": 0.6643, + "step": 3342 + }, + { + "epoch": 1.06976, + "grad_norm": 0.1789153814315796, + "learning_rate": 6.837113402061855e-05, + "loss": 0.5788, + "step": 3343 + }, + { + "epoch": 1.07008, + "grad_norm": 0.1521565318107605, + "learning_rate": 6.83298969072165e-05, + "loss": 0.6729, + "step": 3344 + }, + { + "epoch": 1.0704, + "grad_norm": 0.1653779000043869, + "learning_rate": 6.828865979381444e-05, + "loss": 0.6847, + "step": 3345 + }, + { + "epoch": 1.0707200000000001, + "grad_norm": 0.16444486379623413, + "learning_rate": 6.824742268041237e-05, + "loss": 0.5883, + "step": 3346 + }, + { + "epoch": 1.07104, + "grad_norm": 0.1460576057434082, + "learning_rate": 6.820618556701032e-05, + "loss": 0.5988, + "step": 3347 + }, + { + "epoch": 1.07136, + "grad_norm": 0.17272990942001343, + "learning_rate": 6.816494845360825e-05, + "loss": 0.5395, + "step": 3348 + }, + { + "epoch": 1.07168, + "grad_norm": 0.16581089794635773, + "learning_rate": 6.812371134020619e-05, + "loss": 0.6476, + "step": 3349 + }, + { + "epoch": 1.072, + "grad_norm": 0.15463808178901672, + "learning_rate": 6.808247422680412e-05, + "loss": 0.576, + "step": 3350 + }, + { + "epoch": 1.07232, + "grad_norm": 0.13940781354904175, + "learning_rate": 6.804123711340207e-05, + "loss": 0.4483, + "step": 3351 + }, + { + "epoch": 1.07264, + "grad_norm": 0.17467738687992096, + "learning_rate": 6.800000000000001e-05, + "loss": 0.5656, + "step": 3352 + }, + { + "epoch": 1.07296, + "grad_norm": 0.1533840149641037, + "learning_rate": 6.795876288659794e-05, + "loss": 0.536, + "step": 3353 + }, + { + "epoch": 1.07328, + "grad_norm": 0.1473342329263687, + "learning_rate": 6.791752577319588e-05, + "loss": 0.5546, + "step": 3354 + }, + { + "epoch": 1.0735999999999999, + "grad_norm": 0.1752699464559555, + "learning_rate": 6.787628865979382e-05, + "loss": 0.7301, + "step": 3355 + }, + { + "epoch": 1.07392, + "grad_norm": 0.16849805414676666, + "learning_rate": 6.783505154639177e-05, + "loss": 0.5781, + "step": 3356 + }, + { + "epoch": 1.07424, + "grad_norm": 0.1517651230096817, + "learning_rate": 6.779381443298968e-05, + "loss": 0.5906, + "step": 3357 + }, + { + "epoch": 1.07456, + "grad_norm": 0.1732926368713379, + "learning_rate": 6.775257731958763e-05, + "loss": 0.5962, + "step": 3358 + }, + { + "epoch": 1.07488, + "grad_norm": 0.15156053006649017, + "learning_rate": 6.771134020618557e-05, + "loss": 0.5485, + "step": 3359 + }, + { + "epoch": 1.0752, + "grad_norm": 0.15993241965770721, + "learning_rate": 6.76701030927835e-05, + "loss": 0.5845, + "step": 3360 + }, + { + "epoch": 1.07552, + "grad_norm": 0.16065481305122375, + "learning_rate": 6.762886597938145e-05, + "loss": 0.5872, + "step": 3361 + }, + { + "epoch": 1.07584, + "grad_norm": 0.16931287944316864, + "learning_rate": 6.758762886597938e-05, + "loss": 0.6885, + "step": 3362 + }, + { + "epoch": 1.07616, + "grad_norm": 0.18291021883487701, + "learning_rate": 6.754639175257732e-05, + "loss": 0.7476, + "step": 3363 + }, + { + "epoch": 1.07648, + "grad_norm": 0.15794986486434937, + "learning_rate": 6.750515463917526e-05, + "loss": 0.7351, + "step": 3364 + }, + { + "epoch": 1.0768, + "grad_norm": 0.14923250675201416, + "learning_rate": 6.74639175257732e-05, + "loss": 0.4501, + "step": 3365 + }, + { + "epoch": 1.07712, + "grad_norm": 0.1498347669839859, + "learning_rate": 6.742268041237115e-05, + "loss": 0.5631, + "step": 3366 + }, + { + "epoch": 1.07744, + "grad_norm": 0.17282015085220337, + "learning_rate": 6.738144329896908e-05, + "loss": 0.4743, + "step": 3367 + }, + { + "epoch": 1.07776, + "grad_norm": 0.14651599526405334, + "learning_rate": 6.734020618556701e-05, + "loss": 0.458, + "step": 3368 + }, + { + "epoch": 1.07808, + "grad_norm": 0.1682032197713852, + "learning_rate": 6.729896907216495e-05, + "loss": 0.5893, + "step": 3369 + }, + { + "epoch": 1.0784, + "grad_norm": 0.15365701913833618, + "learning_rate": 6.72577319587629e-05, + "loss": 0.7047, + "step": 3370 + }, + { + "epoch": 1.07872, + "grad_norm": 0.16186408698558807, + "learning_rate": 6.721649484536083e-05, + "loss": 0.6762, + "step": 3371 + }, + { + "epoch": 1.07904, + "grad_norm": 0.14413242042064667, + "learning_rate": 6.717525773195876e-05, + "loss": 0.6622, + "step": 3372 + }, + { + "epoch": 1.07936, + "grad_norm": 0.14972664415836334, + "learning_rate": 6.71340206185567e-05, + "loss": 0.5811, + "step": 3373 + }, + { + "epoch": 1.07968, + "grad_norm": 0.16976608335971832, + "learning_rate": 6.709278350515465e-05, + "loss": 0.6253, + "step": 3374 + }, + { + "epoch": 1.08, + "grad_norm": 0.14484351873397827, + "learning_rate": 6.705154639175258e-05, + "loss": 0.5135, + "step": 3375 + }, + { + "epoch": 1.08032, + "grad_norm": 0.17746739089488983, + "learning_rate": 6.701030927835051e-05, + "loss": 0.6563, + "step": 3376 + }, + { + "epoch": 1.08064, + "grad_norm": 0.1905883252620697, + "learning_rate": 6.696907216494846e-05, + "loss": 0.6204, + "step": 3377 + }, + { + "epoch": 1.08096, + "grad_norm": 0.15244178473949432, + "learning_rate": 6.692783505154639e-05, + "loss": 0.5752, + "step": 3378 + }, + { + "epoch": 1.08128, + "grad_norm": 0.16876575350761414, + "learning_rate": 6.688659793814433e-05, + "loss": 0.5994, + "step": 3379 + }, + { + "epoch": 1.0816, + "grad_norm": 0.1612626165151596, + "learning_rate": 6.684536082474228e-05, + "loss": 0.6156, + "step": 3380 + }, + { + "epoch": 1.08192, + "grad_norm": 0.16236315667629242, + "learning_rate": 6.680412371134021e-05, + "loss": 0.5461, + "step": 3381 + }, + { + "epoch": 1.08224, + "grad_norm": 0.1438879519701004, + "learning_rate": 6.676288659793814e-05, + "loss": 0.5853, + "step": 3382 + }, + { + "epoch": 1.08256, + "grad_norm": 0.14512212574481964, + "learning_rate": 6.672164948453609e-05, + "loss": 0.5821, + "step": 3383 + }, + { + "epoch": 1.08288, + "grad_norm": 0.16177988052368164, + "learning_rate": 6.668041237113403e-05, + "loss": 0.6818, + "step": 3384 + }, + { + "epoch": 1.0832, + "grad_norm": 0.18595930933952332, + "learning_rate": 6.663917525773196e-05, + "loss": 0.5716, + "step": 3385 + }, + { + "epoch": 1.08352, + "grad_norm": 0.14260977506637573, + "learning_rate": 6.659793814432989e-05, + "loss": 0.5358, + "step": 3386 + }, + { + "epoch": 1.08384, + "grad_norm": 0.149885892868042, + "learning_rate": 6.655670103092784e-05, + "loss": 0.6061, + "step": 3387 + }, + { + "epoch": 1.08416, + "grad_norm": 0.1474473625421524, + "learning_rate": 6.651546391752578e-05, + "loss": 0.6419, + "step": 3388 + }, + { + "epoch": 1.08448, + "grad_norm": 0.16489221155643463, + "learning_rate": 6.647422680412371e-05, + "loss": 0.5687, + "step": 3389 + }, + { + "epoch": 1.0848, + "grad_norm": 0.1622447967529297, + "learning_rate": 6.643298969072165e-05, + "loss": 0.5397, + "step": 3390 + }, + { + "epoch": 1.08512, + "grad_norm": 0.1594894826412201, + "learning_rate": 6.639175257731959e-05, + "loss": 0.5874, + "step": 3391 + }, + { + "epoch": 1.08544, + "grad_norm": 0.1566556692123413, + "learning_rate": 6.635051546391754e-05, + "loss": 0.5449, + "step": 3392 + }, + { + "epoch": 1.08576, + "grad_norm": 0.17454101145267487, + "learning_rate": 6.630927835051547e-05, + "loss": 0.7113, + "step": 3393 + }, + { + "epoch": 1.08608, + "grad_norm": 0.1696423441171646, + "learning_rate": 6.626804123711341e-05, + "loss": 0.6359, + "step": 3394 + }, + { + "epoch": 1.0864, + "grad_norm": 0.14636343717575073, + "learning_rate": 6.622680412371134e-05, + "loss": 0.4797, + "step": 3395 + }, + { + "epoch": 1.08672, + "grad_norm": 0.17569296061992645, + "learning_rate": 6.618556701030927e-05, + "loss": 0.5855, + "step": 3396 + }, + { + "epoch": 1.08704, + "grad_norm": 0.15421299636363983, + "learning_rate": 6.614432989690722e-05, + "loss": 0.7242, + "step": 3397 + }, + { + "epoch": 1.0873599999999999, + "grad_norm": 0.17635434865951538, + "learning_rate": 6.610309278350516e-05, + "loss": 0.6291, + "step": 3398 + }, + { + "epoch": 1.08768, + "grad_norm": 0.16109392046928406, + "learning_rate": 6.60618556701031e-05, + "loss": 0.6173, + "step": 3399 + }, + { + "epoch": 1.088, + "grad_norm": 0.17844942212104797, + "learning_rate": 6.602061855670103e-05, + "loss": 0.6731, + "step": 3400 + }, + { + "epoch": 1.08832, + "grad_norm": 0.15459822118282318, + "learning_rate": 6.597938144329897e-05, + "loss": 0.5422, + "step": 3401 + }, + { + "epoch": 1.08864, + "grad_norm": 0.14848153293132782, + "learning_rate": 6.593814432989692e-05, + "loss": 0.6626, + "step": 3402 + }, + { + "epoch": 1.08896, + "grad_norm": 0.15798848867416382, + "learning_rate": 6.589690721649485e-05, + "loss": 0.6044, + "step": 3403 + }, + { + "epoch": 1.08928, + "grad_norm": 0.17471981048583984, + "learning_rate": 6.585567010309278e-05, + "loss": 0.7037, + "step": 3404 + }, + { + "epoch": 1.0896, + "grad_norm": 0.14452970027923584, + "learning_rate": 6.581443298969072e-05, + "loss": 0.4239, + "step": 3405 + }, + { + "epoch": 1.08992, + "grad_norm": 0.1710081547498703, + "learning_rate": 6.577319587628867e-05, + "loss": 0.5928, + "step": 3406 + }, + { + "epoch": 1.09024, + "grad_norm": 0.1780221164226532, + "learning_rate": 6.57319587628866e-05, + "loss": 0.7261, + "step": 3407 + }, + { + "epoch": 1.09056, + "grad_norm": 0.16006827354431152, + "learning_rate": 6.569072164948454e-05, + "loss": 0.4992, + "step": 3408 + }, + { + "epoch": 1.09088, + "grad_norm": 0.15606503188610077, + "learning_rate": 6.564948453608248e-05, + "loss": 0.5361, + "step": 3409 + }, + { + "epoch": 1.0912, + "grad_norm": 0.1612595170736313, + "learning_rate": 6.560824742268042e-05, + "loss": 0.6163, + "step": 3410 + }, + { + "epoch": 1.09152, + "grad_norm": 0.13821835815906525, + "learning_rate": 6.556701030927835e-05, + "loss": 0.5373, + "step": 3411 + }, + { + "epoch": 1.09184, + "grad_norm": 0.15007717907428741, + "learning_rate": 6.55257731958763e-05, + "loss": 0.6348, + "step": 3412 + }, + { + "epoch": 1.09216, + "grad_norm": 0.15153367817401886, + "learning_rate": 6.548453608247424e-05, + "loss": 0.7932, + "step": 3413 + }, + { + "epoch": 1.0924800000000001, + "grad_norm": 0.16655097901821136, + "learning_rate": 6.544329896907216e-05, + "loss": 0.5825, + "step": 3414 + }, + { + "epoch": 1.0928, + "grad_norm": 0.16567280888557434, + "learning_rate": 6.54020618556701e-05, + "loss": 0.613, + "step": 3415 + }, + { + "epoch": 1.09312, + "grad_norm": 0.17999379336833954, + "learning_rate": 6.536082474226805e-05, + "loss": 0.5579, + "step": 3416 + }, + { + "epoch": 1.09344, + "grad_norm": 0.15881694853305817, + "learning_rate": 6.531958762886598e-05, + "loss": 0.6668, + "step": 3417 + }, + { + "epoch": 1.09376, + "grad_norm": 0.16057859361171722, + "learning_rate": 6.527835051546391e-05, + "loss": 0.5255, + "step": 3418 + }, + { + "epoch": 1.09408, + "grad_norm": 0.17333050072193146, + "learning_rate": 6.523711340206186e-05, + "loss": 0.5321, + "step": 3419 + }, + { + "epoch": 1.0944, + "grad_norm": 0.16746041178703308, + "learning_rate": 6.51958762886598e-05, + "loss": 0.6933, + "step": 3420 + }, + { + "epoch": 1.09472, + "grad_norm": 0.15572938323020935, + "learning_rate": 6.515463917525773e-05, + "loss": 0.5119, + "step": 3421 + }, + { + "epoch": 1.09504, + "grad_norm": 0.1518268883228302, + "learning_rate": 6.511340206185568e-05, + "loss": 0.5601, + "step": 3422 + }, + { + "epoch": 1.09536, + "grad_norm": 0.15530791878700256, + "learning_rate": 6.507216494845361e-05, + "loss": 0.6417, + "step": 3423 + }, + { + "epoch": 1.09568, + "grad_norm": 0.15300092101097107, + "learning_rate": 6.503092783505155e-05, + "loss": 0.7745, + "step": 3424 + }, + { + "epoch": 1.096, + "grad_norm": 0.17361828684806824, + "learning_rate": 6.498969072164948e-05, + "loss": 0.6163, + "step": 3425 + }, + { + "epoch": 1.09632, + "grad_norm": 0.18202218413352966, + "learning_rate": 6.494845360824743e-05, + "loss": 0.6019, + "step": 3426 + }, + { + "epoch": 1.09664, + "grad_norm": 0.16239918768405914, + "learning_rate": 6.490721649484537e-05, + "loss": 0.5636, + "step": 3427 + }, + { + "epoch": 1.09696, + "grad_norm": 0.15811161696910858, + "learning_rate": 6.48659793814433e-05, + "loss": 0.5439, + "step": 3428 + }, + { + "epoch": 1.09728, + "grad_norm": 0.15355218946933746, + "learning_rate": 6.482474226804124e-05, + "loss": 0.6027, + "step": 3429 + }, + { + "epoch": 1.0976, + "grad_norm": 0.15723192691802979, + "learning_rate": 6.478350515463918e-05, + "loss": 0.5261, + "step": 3430 + }, + { + "epoch": 1.09792, + "grad_norm": 0.18477247655391693, + "learning_rate": 6.474226804123713e-05, + "loss": 0.6936, + "step": 3431 + }, + { + "epoch": 1.09824, + "grad_norm": 0.16078898310661316, + "learning_rate": 6.470103092783506e-05, + "loss": 0.7035, + "step": 3432 + }, + { + "epoch": 1.09856, + "grad_norm": 0.14289724826812744, + "learning_rate": 6.465979381443299e-05, + "loss": 0.5302, + "step": 3433 + }, + { + "epoch": 1.09888, + "grad_norm": 0.16278979182243347, + "learning_rate": 6.461855670103093e-05, + "loss": 0.5206, + "step": 3434 + }, + { + "epoch": 1.0992, + "grad_norm": 0.14219529926776886, + "learning_rate": 6.457731958762886e-05, + "loss": 0.7134, + "step": 3435 + }, + { + "epoch": 1.09952, + "grad_norm": 0.174630269408226, + "learning_rate": 6.453608247422681e-05, + "loss": 0.7124, + "step": 3436 + }, + { + "epoch": 1.09984, + "grad_norm": 0.17383918166160583, + "learning_rate": 6.449484536082474e-05, + "loss": 0.6557, + "step": 3437 + }, + { + "epoch": 1.10016, + "grad_norm": 0.14923472702503204, + "learning_rate": 6.445360824742269e-05, + "loss": 0.6288, + "step": 3438 + }, + { + "epoch": 1.10048, + "grad_norm": 0.17504321038722992, + "learning_rate": 6.441237113402062e-05, + "loss": 0.5316, + "step": 3439 + }, + { + "epoch": 1.1008, + "grad_norm": 0.166417196393013, + "learning_rate": 6.437113402061856e-05, + "loss": 0.613, + "step": 3440 + }, + { + "epoch": 1.10112, + "grad_norm": 0.16554030776023865, + "learning_rate": 6.43298969072165e-05, + "loss": 0.5904, + "step": 3441 + }, + { + "epoch": 1.10144, + "grad_norm": 0.15483027696609497, + "learning_rate": 6.428865979381444e-05, + "loss": 0.6835, + "step": 3442 + }, + { + "epoch": 1.10176, + "grad_norm": 0.16063041985034943, + "learning_rate": 6.424742268041237e-05, + "loss": 0.7193, + "step": 3443 + }, + { + "epoch": 1.10208, + "grad_norm": 0.1681942343711853, + "learning_rate": 6.420618556701031e-05, + "loss": 0.615, + "step": 3444 + }, + { + "epoch": 1.1024, + "grad_norm": 0.1514786034822464, + "learning_rate": 6.416494845360826e-05, + "loss": 0.4374, + "step": 3445 + }, + { + "epoch": 1.10272, + "grad_norm": 0.16579103469848633, + "learning_rate": 6.412371134020619e-05, + "loss": 0.5564, + "step": 3446 + }, + { + "epoch": 1.10304, + "grad_norm": 0.17870011925697327, + "learning_rate": 6.408247422680412e-05, + "loss": 0.5678, + "step": 3447 + }, + { + "epoch": 1.10336, + "grad_norm": 0.15543703734874725, + "learning_rate": 6.404123711340207e-05, + "loss": 0.6556, + "step": 3448 + }, + { + "epoch": 1.10368, + "grad_norm": 0.16825172305107117, + "learning_rate": 6.400000000000001e-05, + "loss": 0.5566, + "step": 3449 + }, + { + "epoch": 1.104, + "grad_norm": 0.15695036947727203, + "learning_rate": 6.395876288659794e-05, + "loss": 0.5451, + "step": 3450 + }, + { + "epoch": 1.10432, + "grad_norm": 0.15175402164459229, + "learning_rate": 6.391752577319587e-05, + "loss": 0.4875, + "step": 3451 + }, + { + "epoch": 1.10464, + "grad_norm": 0.14299476146697998, + "learning_rate": 6.387628865979382e-05, + "loss": 0.5291, + "step": 3452 + }, + { + "epoch": 1.10496, + "grad_norm": 0.16131307184696198, + "learning_rate": 6.383505154639175e-05, + "loss": 0.6666, + "step": 3453 + }, + { + "epoch": 1.10528, + "grad_norm": 0.16285930573940277, + "learning_rate": 6.37938144329897e-05, + "loss": 0.5511, + "step": 3454 + }, + { + "epoch": 1.1056, + "grad_norm": 0.161065474152565, + "learning_rate": 6.375257731958764e-05, + "loss": 0.5377, + "step": 3455 + }, + { + "epoch": 1.10592, + "grad_norm": 0.14697834849357605, + "learning_rate": 6.371134020618557e-05, + "loss": 0.6609, + "step": 3456 + }, + { + "epoch": 1.1062400000000001, + "grad_norm": 0.16829650104045868, + "learning_rate": 6.36701030927835e-05, + "loss": 0.6584, + "step": 3457 + }, + { + "epoch": 1.10656, + "grad_norm": 0.16058404743671417, + "learning_rate": 6.362886597938145e-05, + "loss": 0.6112, + "step": 3458 + }, + { + "epoch": 1.10688, + "grad_norm": 0.1664557009935379, + "learning_rate": 6.358762886597939e-05, + "loss": 0.5997, + "step": 3459 + }, + { + "epoch": 1.1072, + "grad_norm": 0.17357204854488373, + "learning_rate": 6.354639175257732e-05, + "loss": 0.6711, + "step": 3460 + }, + { + "epoch": 1.10752, + "grad_norm": 0.16929683089256287, + "learning_rate": 6.350515463917525e-05, + "loss": 0.6434, + "step": 3461 + }, + { + "epoch": 1.10784, + "grad_norm": 0.16111072897911072, + "learning_rate": 6.34639175257732e-05, + "loss": 0.7489, + "step": 3462 + }, + { + "epoch": 1.10816, + "grad_norm": 0.16690926253795624, + "learning_rate": 6.342268041237114e-05, + "loss": 0.6013, + "step": 3463 + }, + { + "epoch": 1.10848, + "grad_norm": 0.155247762799263, + "learning_rate": 6.338144329896907e-05, + "loss": 0.5761, + "step": 3464 + }, + { + "epoch": 1.1088, + "grad_norm": 0.16866454482078552, + "learning_rate": 6.3340206185567e-05, + "loss": 0.7203, + "step": 3465 + }, + { + "epoch": 1.1091199999999999, + "grad_norm": 0.14751704037189484, + "learning_rate": 6.329896907216495e-05, + "loss": 0.637, + "step": 3466 + }, + { + "epoch": 1.10944, + "grad_norm": 0.15739154815673828, + "learning_rate": 6.32577319587629e-05, + "loss": 0.674, + "step": 3467 + }, + { + "epoch": 1.10976, + "grad_norm": 0.16756172478199005, + "learning_rate": 6.321649484536083e-05, + "loss": 0.8138, + "step": 3468 + }, + { + "epoch": 1.11008, + "grad_norm": 0.16327467560768127, + "learning_rate": 6.317525773195877e-05, + "loss": 0.5426, + "step": 3469 + }, + { + "epoch": 1.1104, + "grad_norm": 0.17287665605545044, + "learning_rate": 6.31340206185567e-05, + "loss": 0.6392, + "step": 3470 + }, + { + "epoch": 1.11072, + "grad_norm": 0.15868349373340607, + "learning_rate": 6.309278350515463e-05, + "loss": 0.6425, + "step": 3471 + }, + { + "epoch": 1.11104, + "grad_norm": 0.17105525732040405, + "learning_rate": 6.305154639175258e-05, + "loss": 0.5363, + "step": 3472 + }, + { + "epoch": 1.11136, + "grad_norm": 0.16856800019741058, + "learning_rate": 6.301030927835052e-05, + "loss": 0.5968, + "step": 3473 + }, + { + "epoch": 1.11168, + "grad_norm": 0.16913408041000366, + "learning_rate": 6.296907216494846e-05, + "loss": 0.5653, + "step": 3474 + }, + { + "epoch": 1.112, + "grad_norm": 0.14649230241775513, + "learning_rate": 6.292783505154639e-05, + "loss": 0.6169, + "step": 3475 + }, + { + "epoch": 1.11232, + "grad_norm": 0.1424107849597931, + "learning_rate": 6.288659793814433e-05, + "loss": 0.5195, + "step": 3476 + }, + { + "epoch": 1.11264, + "grad_norm": 0.1415460705757141, + "learning_rate": 6.284536082474228e-05, + "loss": 0.6416, + "step": 3477 + }, + { + "epoch": 1.11296, + "grad_norm": 0.16070856153964996, + "learning_rate": 6.280412371134021e-05, + "loss": 0.6178, + "step": 3478 + }, + { + "epoch": 1.11328, + "grad_norm": 0.167328879237175, + "learning_rate": 6.276288659793814e-05, + "loss": 0.6186, + "step": 3479 + }, + { + "epoch": 1.1136, + "grad_norm": 0.18806082010269165, + "learning_rate": 6.272164948453608e-05, + "loss": 0.4992, + "step": 3480 + }, + { + "epoch": 1.11392, + "grad_norm": 0.16985145211219788, + "learning_rate": 6.268041237113403e-05, + "loss": 0.617, + "step": 3481 + }, + { + "epoch": 1.11424, + "grad_norm": 0.17039939761161804, + "learning_rate": 6.263917525773196e-05, + "loss": 0.5863, + "step": 3482 + }, + { + "epoch": 1.11456, + "grad_norm": 0.19796743988990784, + "learning_rate": 6.25979381443299e-05, + "loss": 0.7299, + "step": 3483 + }, + { + "epoch": 1.11488, + "grad_norm": 0.16527968645095825, + "learning_rate": 6.255670103092784e-05, + "loss": 0.6445, + "step": 3484 + }, + { + "epoch": 1.1152, + "grad_norm": 0.1780211329460144, + "learning_rate": 6.251546391752578e-05, + "loss": 0.6573, + "step": 3485 + }, + { + "epoch": 1.11552, + "grad_norm": 0.18306085467338562, + "learning_rate": 6.247422680412371e-05, + "loss": 0.5989, + "step": 3486 + }, + { + "epoch": 1.11584, + "grad_norm": 0.17549125850200653, + "learning_rate": 6.243298969072166e-05, + "loss": 0.6191, + "step": 3487 + }, + { + "epoch": 1.11616, + "grad_norm": 0.1758272647857666, + "learning_rate": 6.23917525773196e-05, + "loss": 0.6413, + "step": 3488 + }, + { + "epoch": 1.11648, + "grad_norm": 0.16608774662017822, + "learning_rate": 6.235051546391752e-05, + "loss": 0.6678, + "step": 3489 + }, + { + "epoch": 1.1168, + "grad_norm": 0.1666615754365921, + "learning_rate": 6.230927835051546e-05, + "loss": 0.7896, + "step": 3490 + }, + { + "epoch": 1.11712, + "grad_norm": 0.16971318423748016, + "learning_rate": 6.226804123711341e-05, + "loss": 0.5949, + "step": 3491 + }, + { + "epoch": 1.11744, + "grad_norm": 0.18260937929153442, + "learning_rate": 6.222680412371134e-05, + "loss": 0.7132, + "step": 3492 + }, + { + "epoch": 1.11776, + "grad_norm": 0.19655515253543854, + "learning_rate": 6.218556701030927e-05, + "loss": 0.7573, + "step": 3493 + }, + { + "epoch": 1.11808, + "grad_norm": 0.171479269862175, + "learning_rate": 6.214432989690722e-05, + "loss": 0.6412, + "step": 3494 + }, + { + "epoch": 1.1184, + "grad_norm": 0.16946713626384735, + "learning_rate": 6.210309278350516e-05, + "loss": 0.743, + "step": 3495 + }, + { + "epoch": 1.11872, + "grad_norm": 0.15758058428764343, + "learning_rate": 6.206185567010309e-05, + "loss": 0.5797, + "step": 3496 + }, + { + "epoch": 1.11904, + "grad_norm": 0.17547620832920074, + "learning_rate": 6.202061855670104e-05, + "loss": 0.5737, + "step": 3497 + }, + { + "epoch": 1.11936, + "grad_norm": 0.16721968352794647, + "learning_rate": 6.197938144329897e-05, + "loss": 0.6717, + "step": 3498 + }, + { + "epoch": 1.11968, + "grad_norm": 0.14399756491184235, + "learning_rate": 6.193814432989691e-05, + "loss": 0.6144, + "step": 3499 + }, + { + "epoch": 1.12, + "grad_norm": 0.17312820255756378, + "learning_rate": 6.189690721649484e-05, + "loss": 0.6096, + "step": 3500 + }, + { + "epoch": 1.12032, + "grad_norm": 0.1613154113292694, + "learning_rate": 6.185567010309279e-05, + "loss": 0.6418, + "step": 3501 + }, + { + "epoch": 1.12064, + "grad_norm": 0.15359307825565338, + "learning_rate": 6.181443298969073e-05, + "loss": 0.7029, + "step": 3502 + }, + { + "epoch": 1.12096, + "grad_norm": 0.14040406048297882, + "learning_rate": 6.177319587628867e-05, + "loss": 0.4781, + "step": 3503 + }, + { + "epoch": 1.12128, + "grad_norm": 0.1534678041934967, + "learning_rate": 6.17319587628866e-05, + "loss": 0.4722, + "step": 3504 + }, + { + "epoch": 1.1216, + "grad_norm": 0.17000633478164673, + "learning_rate": 6.169072164948454e-05, + "loss": 0.7079, + "step": 3505 + }, + { + "epoch": 1.12192, + "grad_norm": 0.15092356503009796, + "learning_rate": 6.164948453608249e-05, + "loss": 0.6665, + "step": 3506 + }, + { + "epoch": 1.12224, + "grad_norm": 0.14488859474658966, + "learning_rate": 6.160824742268042e-05, + "loss": 0.5083, + "step": 3507 + }, + { + "epoch": 1.12256, + "grad_norm": 0.18396364152431488, + "learning_rate": 6.156701030927835e-05, + "loss": 0.6294, + "step": 3508 + }, + { + "epoch": 1.12288, + "grad_norm": 0.17745038866996765, + "learning_rate": 6.15257731958763e-05, + "loss": 0.678, + "step": 3509 + }, + { + "epoch": 1.1232, + "grad_norm": 0.1605219841003418, + "learning_rate": 6.148453608247422e-05, + "loss": 0.5193, + "step": 3510 + }, + { + "epoch": 1.12352, + "grad_norm": 0.17112098634243011, + "learning_rate": 6.144329896907217e-05, + "loss": 0.7309, + "step": 3511 + }, + { + "epoch": 1.12384, + "grad_norm": 0.15104934573173523, + "learning_rate": 6.14020618556701e-05, + "loss": 0.5778, + "step": 3512 + }, + { + "epoch": 1.12416, + "grad_norm": 0.1908404529094696, + "learning_rate": 6.136082474226805e-05, + "loss": 0.5893, + "step": 3513 + }, + { + "epoch": 1.12448, + "grad_norm": 0.18122270703315735, + "learning_rate": 6.131958762886598e-05, + "loss": 0.5372, + "step": 3514 + }, + { + "epoch": 1.1248, + "grad_norm": 0.13836006820201874, + "learning_rate": 6.127835051546392e-05, + "loss": 0.5404, + "step": 3515 + }, + { + "epoch": 1.12512, + "grad_norm": 0.17441757023334503, + "learning_rate": 6.123711340206187e-05, + "loss": 0.6734, + "step": 3516 + }, + { + "epoch": 1.12544, + "grad_norm": 0.17483940720558167, + "learning_rate": 6.11958762886598e-05, + "loss": 0.6978, + "step": 3517 + }, + { + "epoch": 1.12576, + "grad_norm": 0.16496886312961578, + "learning_rate": 6.115463917525773e-05, + "loss": 0.5986, + "step": 3518 + }, + { + "epoch": 1.12608, + "grad_norm": 0.1796550303697586, + "learning_rate": 6.111340206185567e-05, + "loss": 0.5951, + "step": 3519 + }, + { + "epoch": 1.1264, + "grad_norm": 0.15103594958782196, + "learning_rate": 6.107216494845362e-05, + "loss": 0.5055, + "step": 3520 + }, + { + "epoch": 1.12672, + "grad_norm": 0.15241901576519012, + "learning_rate": 6.103092783505156e-05, + "loss": 0.6539, + "step": 3521 + }, + { + "epoch": 1.12704, + "grad_norm": 0.1474504917860031, + "learning_rate": 6.098969072164948e-05, + "loss": 0.6596, + "step": 3522 + }, + { + "epoch": 1.12736, + "grad_norm": 0.16461731493473053, + "learning_rate": 6.0948453608247426e-05, + "loss": 0.7345, + "step": 3523 + }, + { + "epoch": 1.12768, + "grad_norm": 0.17002519965171814, + "learning_rate": 6.0907216494845364e-05, + "loss": 0.762, + "step": 3524 + }, + { + "epoch": 1.1280000000000001, + "grad_norm": 0.1835489273071289, + "learning_rate": 6.08659793814433e-05, + "loss": 0.7556, + "step": 3525 + }, + { + "epoch": 1.12832, + "grad_norm": 0.15879112482070923, + "learning_rate": 6.0824742268041234e-05, + "loss": 0.8045, + "step": 3526 + }, + { + "epoch": 1.12864, + "grad_norm": 0.18141713738441467, + "learning_rate": 6.078350515463918e-05, + "loss": 0.661, + "step": 3527 + }, + { + "epoch": 1.12896, + "grad_norm": 0.1683078557252884, + "learning_rate": 6.074226804123712e-05, + "loss": 0.6638, + "step": 3528 + }, + { + "epoch": 1.12928, + "grad_norm": 0.17433901131153107, + "learning_rate": 6.0701030927835055e-05, + "loss": 0.7253, + "step": 3529 + }, + { + "epoch": 1.1296, + "grad_norm": 0.17473284900188446, + "learning_rate": 6.0659793814433e-05, + "loss": 0.6998, + "step": 3530 + }, + { + "epoch": 1.12992, + "grad_norm": 0.1805301010608673, + "learning_rate": 6.0618556701030924e-05, + "loss": 0.7052, + "step": 3531 + }, + { + "epoch": 1.13024, + "grad_norm": 0.18975281715393066, + "learning_rate": 6.057731958762887e-05, + "loss": 0.6432, + "step": 3532 + }, + { + "epoch": 1.13056, + "grad_norm": 0.18678808212280273, + "learning_rate": 6.053608247422681e-05, + "loss": 0.5769, + "step": 3533 + }, + { + "epoch": 1.1308799999999999, + "grad_norm": 0.17771640419960022, + "learning_rate": 6.0494845360824745e-05, + "loss": 0.518, + "step": 3534 + }, + { + "epoch": 1.1312, + "grad_norm": 0.15519271790981293, + "learning_rate": 6.045360824742269e-05, + "loss": 0.7297, + "step": 3535 + }, + { + "epoch": 1.13152, + "grad_norm": 0.15465021133422852, + "learning_rate": 6.041237113402062e-05, + "loss": 0.6395, + "step": 3536 + }, + { + "epoch": 1.13184, + "grad_norm": 0.14844176173210144, + "learning_rate": 6.037113402061856e-05, + "loss": 0.6287, + "step": 3537 + }, + { + "epoch": 1.13216, + "grad_norm": 0.15697647631168365, + "learning_rate": 6.03298969072165e-05, + "loss": 0.4088, + "step": 3538 + }, + { + "epoch": 1.13248, + "grad_norm": 0.15739664435386658, + "learning_rate": 6.028865979381444e-05, + "loss": 0.7711, + "step": 3539 + }, + { + "epoch": 1.1328, + "grad_norm": 0.16153065860271454, + "learning_rate": 6.0247422680412366e-05, + "loss": 0.5695, + "step": 3540 + }, + { + "epoch": 1.13312, + "grad_norm": 0.1722458302974701, + "learning_rate": 6.020618556701031e-05, + "loss": 0.7029, + "step": 3541 + }, + { + "epoch": 1.13344, + "grad_norm": 0.1673302948474884, + "learning_rate": 6.016494845360825e-05, + "loss": 0.5101, + "step": 3542 + }, + { + "epoch": 1.13376, + "grad_norm": 0.14588989317417145, + "learning_rate": 6.012371134020619e-05, + "loss": 0.5743, + "step": 3543 + }, + { + "epoch": 1.13408, + "grad_norm": 0.12573234736919403, + "learning_rate": 6.008247422680413e-05, + "loss": 0.7307, + "step": 3544 + }, + { + "epoch": 1.1344, + "grad_norm": 0.13937446475028992, + "learning_rate": 6.0041237113402063e-05, + "loss": 0.766, + "step": 3545 + }, + { + "epoch": 1.13472, + "grad_norm": 0.16740339994430542, + "learning_rate": 6e-05, + "loss": 0.6827, + "step": 3546 + }, + { + "epoch": 1.13504, + "grad_norm": 0.16018402576446533, + "learning_rate": 5.995876288659794e-05, + "loss": 0.5533, + "step": 3547 + }, + { + "epoch": 1.13536, + "grad_norm": 0.1377609670162201, + "learning_rate": 5.9917525773195884e-05, + "loss": 0.6043, + "step": 3548 + }, + { + "epoch": 1.13568, + "grad_norm": 0.1586773693561554, + "learning_rate": 5.987628865979382e-05, + "loss": 0.6405, + "step": 3549 + }, + { + "epoch": 1.1360000000000001, + "grad_norm": 0.17038887739181519, + "learning_rate": 5.9835051546391754e-05, + "loss": 0.722, + "step": 3550 + }, + { + "epoch": 1.13632, + "grad_norm": 0.15706969797611237, + "learning_rate": 5.979381443298969e-05, + "loss": 0.6613, + "step": 3551 + }, + { + "epoch": 1.13664, + "grad_norm": 0.15962283313274384, + "learning_rate": 5.975257731958763e-05, + "loss": 0.5581, + "step": 3552 + }, + { + "epoch": 1.13696, + "grad_norm": 0.17226724326610565, + "learning_rate": 5.9711340206185575e-05, + "loss": 0.5782, + "step": 3553 + }, + { + "epoch": 1.13728, + "grad_norm": 0.15993818640708923, + "learning_rate": 5.96701030927835e-05, + "loss": 0.5843, + "step": 3554 + }, + { + "epoch": 1.1376, + "grad_norm": 0.16454949975013733, + "learning_rate": 5.9628865979381444e-05, + "loss": 0.6649, + "step": 3555 + }, + { + "epoch": 1.13792, + "grad_norm": 0.16202202439308167, + "learning_rate": 5.958762886597938e-05, + "loss": 0.6539, + "step": 3556 + }, + { + "epoch": 1.13824, + "grad_norm": 0.1439141482114792, + "learning_rate": 5.954639175257733e-05, + "loss": 0.6292, + "step": 3557 + }, + { + "epoch": 1.13856, + "grad_norm": 0.1561082899570465, + "learning_rate": 5.9505154639175265e-05, + "loss": 0.6183, + "step": 3558 + }, + { + "epoch": 1.13888, + "grad_norm": 0.1817554533481598, + "learning_rate": 5.9463917525773196e-05, + "loss": 0.6544, + "step": 3559 + }, + { + "epoch": 1.1392, + "grad_norm": 0.18692998588085175, + "learning_rate": 5.9422680412371134e-05, + "loss": 0.5285, + "step": 3560 + }, + { + "epoch": 1.13952, + "grad_norm": 0.155509814620018, + "learning_rate": 5.938144329896907e-05, + "loss": 0.6321, + "step": 3561 + }, + { + "epoch": 1.13984, + "grad_norm": 0.1768050193786621, + "learning_rate": 5.934020618556702e-05, + "loss": 0.5535, + "step": 3562 + }, + { + "epoch": 1.14016, + "grad_norm": 0.160866841673851, + "learning_rate": 5.9298969072164955e-05, + "loss": 0.5882, + "step": 3563 + }, + { + "epoch": 1.14048, + "grad_norm": 0.16839323937892914, + "learning_rate": 5.9257731958762886e-05, + "loss": 0.6628, + "step": 3564 + }, + { + "epoch": 1.1408, + "grad_norm": 0.15851418673992157, + "learning_rate": 5.9216494845360824e-05, + "loss": 0.5649, + "step": 3565 + }, + { + "epoch": 1.14112, + "grad_norm": 0.14016911387443542, + "learning_rate": 5.917525773195877e-05, + "loss": 0.5267, + "step": 3566 + }, + { + "epoch": 1.14144, + "grad_norm": 0.16235263645648956, + "learning_rate": 5.913402061855671e-05, + "loss": 0.6352, + "step": 3567 + }, + { + "epoch": 1.14176, + "grad_norm": 0.18664886057376862, + "learning_rate": 5.9092783505154645e-05, + "loss": 0.6406, + "step": 3568 + }, + { + "epoch": 1.14208, + "grad_norm": 0.19914813339710236, + "learning_rate": 5.905154639175258e-05, + "loss": 0.5784, + "step": 3569 + }, + { + "epoch": 1.1424, + "grad_norm": 0.17438171803951263, + "learning_rate": 5.9010309278350515e-05, + "loss": 0.6664, + "step": 3570 + }, + { + "epoch": 1.14272, + "grad_norm": 0.1540410816669464, + "learning_rate": 5.896907216494846e-05, + "loss": 0.502, + "step": 3571 + }, + { + "epoch": 1.14304, + "grad_norm": 0.1916867345571518, + "learning_rate": 5.89278350515464e-05, + "loss": 0.6944, + "step": 3572 + }, + { + "epoch": 1.14336, + "grad_norm": 0.17228063941001892, + "learning_rate": 5.888659793814433e-05, + "loss": 0.6052, + "step": 3573 + }, + { + "epoch": 1.14368, + "grad_norm": 0.19526247680187225, + "learning_rate": 5.884536082474227e-05, + "loss": 0.5658, + "step": 3574 + }, + { + "epoch": 1.144, + "grad_norm": 0.16881060600280762, + "learning_rate": 5.880412371134021e-05, + "loss": 0.6662, + "step": 3575 + }, + { + "epoch": 1.14432, + "grad_norm": 0.17348535358905792, + "learning_rate": 5.876288659793815e-05, + "loss": 0.5953, + "step": 3576 + }, + { + "epoch": 1.1446399999999999, + "grad_norm": 0.17873069643974304, + "learning_rate": 5.872164948453609e-05, + "loss": 0.7145, + "step": 3577 + }, + { + "epoch": 1.14496, + "grad_norm": 0.148677796125412, + "learning_rate": 5.868041237113402e-05, + "loss": 0.6276, + "step": 3578 + }, + { + "epoch": 1.14528, + "grad_norm": 0.1800248622894287, + "learning_rate": 5.863917525773196e-05, + "loss": 0.5695, + "step": 3579 + }, + { + "epoch": 1.1456, + "grad_norm": 0.16717465221881866, + "learning_rate": 5.85979381443299e-05, + "loss": 0.5279, + "step": 3580 + }, + { + "epoch": 1.14592, + "grad_norm": 0.1862657070159912, + "learning_rate": 5.855670103092784e-05, + "loss": 0.6707, + "step": 3581 + }, + { + "epoch": 1.14624, + "grad_norm": 0.15623800456523895, + "learning_rate": 5.851546391752578e-05, + "loss": 0.7396, + "step": 3582 + }, + { + "epoch": 1.14656, + "grad_norm": 0.17744137346744537, + "learning_rate": 5.847422680412371e-05, + "loss": 0.6794, + "step": 3583 + }, + { + "epoch": 1.14688, + "grad_norm": 0.16237524151802063, + "learning_rate": 5.8432989690721654e-05, + "loss": 0.4599, + "step": 3584 + }, + { + "epoch": 1.1472, + "grad_norm": 0.17019815742969513, + "learning_rate": 5.839175257731959e-05, + "loss": 0.7746, + "step": 3585 + }, + { + "epoch": 1.14752, + "grad_norm": 0.15601864457130432, + "learning_rate": 5.835051546391753e-05, + "loss": 0.7111, + "step": 3586 + }, + { + "epoch": 1.14784, + "grad_norm": 0.17591580748558044, + "learning_rate": 5.830927835051546e-05, + "loss": 0.6799, + "step": 3587 + }, + { + "epoch": 1.14816, + "grad_norm": 0.18657545745372772, + "learning_rate": 5.82680412371134e-05, + "loss": 0.7156, + "step": 3588 + }, + { + "epoch": 1.14848, + "grad_norm": 0.17224997282028198, + "learning_rate": 5.8226804123711344e-05, + "loss": 0.7064, + "step": 3589 + }, + { + "epoch": 1.1488, + "grad_norm": 0.1719975471496582, + "learning_rate": 5.818556701030928e-05, + "loss": 0.5949, + "step": 3590 + }, + { + "epoch": 1.14912, + "grad_norm": 0.16977804899215698, + "learning_rate": 5.814432989690722e-05, + "loss": 0.6423, + "step": 3591 + }, + { + "epoch": 1.14944, + "grad_norm": 0.1905670166015625, + "learning_rate": 5.810309278350515e-05, + "loss": 0.5933, + "step": 3592 + }, + { + "epoch": 1.1497600000000001, + "grad_norm": 0.166902557015419, + "learning_rate": 5.80618556701031e-05, + "loss": 0.6478, + "step": 3593 + }, + { + "epoch": 1.15008, + "grad_norm": 0.16657216846942902, + "learning_rate": 5.8020618556701035e-05, + "loss": 0.4754, + "step": 3594 + }, + { + "epoch": 1.1504, + "grad_norm": 0.1694144606590271, + "learning_rate": 5.797938144329897e-05, + "loss": 0.5398, + "step": 3595 + }, + { + "epoch": 1.15072, + "grad_norm": 0.17381985485553741, + "learning_rate": 5.793814432989692e-05, + "loss": 0.5794, + "step": 3596 + }, + { + "epoch": 1.15104, + "grad_norm": 0.1662823110818863, + "learning_rate": 5.789690721649484e-05, + "loss": 0.6788, + "step": 3597 + }, + { + "epoch": 1.15136, + "grad_norm": 0.1627492904663086, + "learning_rate": 5.785567010309279e-05, + "loss": 0.5754, + "step": 3598 + }, + { + "epoch": 1.15168, + "grad_norm": 0.1806124895811081, + "learning_rate": 5.7814432989690725e-05, + "loss": 0.7006, + "step": 3599 + }, + { + "epoch": 1.152, + "grad_norm": 0.16741174459457397, + "learning_rate": 5.777319587628866e-05, + "loss": 0.7161, + "step": 3600 + }, + { + "epoch": 1.15232, + "grad_norm": 0.16751769185066223, + "learning_rate": 5.7731958762886594e-05, + "loss": 0.7641, + "step": 3601 + }, + { + "epoch": 1.1526399999999999, + "grad_norm": 0.1575736254453659, + "learning_rate": 5.769072164948454e-05, + "loss": 0.4739, + "step": 3602 + }, + { + "epoch": 1.15296, + "grad_norm": 0.1704677939414978, + "learning_rate": 5.764948453608248e-05, + "loss": 0.683, + "step": 3603 + }, + { + "epoch": 1.15328, + "grad_norm": 0.1841387152671814, + "learning_rate": 5.7608247422680415e-05, + "loss": 0.6003, + "step": 3604 + }, + { + "epoch": 1.1536, + "grad_norm": 0.18312706053256989, + "learning_rate": 5.756701030927836e-05, + "loss": 0.6434, + "step": 3605 + }, + { + "epoch": 1.15392, + "grad_norm": 0.1501491665840149, + "learning_rate": 5.7525773195876285e-05, + "loss": 0.7138, + "step": 3606 + }, + { + "epoch": 1.15424, + "grad_norm": 0.14632530510425568, + "learning_rate": 5.748453608247423e-05, + "loss": 0.4731, + "step": 3607 + }, + { + "epoch": 1.15456, + "grad_norm": 0.17078693211078644, + "learning_rate": 5.744329896907217e-05, + "loss": 0.647, + "step": 3608 + }, + { + "epoch": 1.15488, + "grad_norm": 0.14590023458003998, + "learning_rate": 5.7402061855670105e-05, + "loss": 0.6518, + "step": 3609 + }, + { + "epoch": 1.1552, + "grad_norm": 0.16748274862766266, + "learning_rate": 5.736082474226805e-05, + "loss": 0.6705, + "step": 3610 + }, + { + "epoch": 1.15552, + "grad_norm": 0.14955997467041016, + "learning_rate": 5.731958762886598e-05, + "loss": 0.6579, + "step": 3611 + }, + { + "epoch": 1.15584, + "grad_norm": 0.1452036201953888, + "learning_rate": 5.727835051546392e-05, + "loss": 0.5901, + "step": 3612 + }, + { + "epoch": 1.15616, + "grad_norm": 0.17077872157096863, + "learning_rate": 5.723711340206186e-05, + "loss": 0.6467, + "step": 3613 + }, + { + "epoch": 1.15648, + "grad_norm": 0.17951638996601105, + "learning_rate": 5.71958762886598e-05, + "loss": 0.7231, + "step": 3614 + }, + { + "epoch": 1.1568, + "grad_norm": 0.1834004819393158, + "learning_rate": 5.715463917525773e-05, + "loss": 0.6255, + "step": 3615 + }, + { + "epoch": 1.15712, + "grad_norm": 0.15475787222385406, + "learning_rate": 5.711340206185567e-05, + "loss": 0.6738, + "step": 3616 + }, + { + "epoch": 1.15744, + "grad_norm": 0.15446065366268158, + "learning_rate": 5.707216494845361e-05, + "loss": 0.4767, + "step": 3617 + }, + { + "epoch": 1.1577600000000001, + "grad_norm": 0.1676519215106964, + "learning_rate": 5.703092783505155e-05, + "loss": 0.6347, + "step": 3618 + }, + { + "epoch": 1.15808, + "grad_norm": 0.1446048766374588, + "learning_rate": 5.698969072164949e-05, + "loss": 0.7609, + "step": 3619 + }, + { + "epoch": 1.1584, + "grad_norm": 0.15303777158260345, + "learning_rate": 5.6948453608247424e-05, + "loss": 0.6211, + "step": 3620 + }, + { + "epoch": 1.15872, + "grad_norm": 0.14983202517032623, + "learning_rate": 5.690721649484536e-05, + "loss": 0.5687, + "step": 3621 + }, + { + "epoch": 1.15904, + "grad_norm": 0.17641636729240417, + "learning_rate": 5.68659793814433e-05, + "loss": 0.6525, + "step": 3622 + }, + { + "epoch": 1.15936, + "grad_norm": 0.18665477633476257, + "learning_rate": 5.6824742268041245e-05, + "loss": 0.579, + "step": 3623 + }, + { + "epoch": 1.15968, + "grad_norm": 0.1705567091703415, + "learning_rate": 5.678350515463918e-05, + "loss": 0.5249, + "step": 3624 + }, + { + "epoch": 1.16, + "grad_norm": 0.15909402072429657, + "learning_rate": 5.6742268041237114e-05, + "loss": 0.7561, + "step": 3625 + }, + { + "epoch": 1.16032, + "grad_norm": 0.14951229095458984, + "learning_rate": 5.670103092783505e-05, + "loss": 0.5719, + "step": 3626 + }, + { + "epoch": 1.16064, + "grad_norm": 0.16878777742385864, + "learning_rate": 5.665979381443299e-05, + "loss": 0.5716, + "step": 3627 + }, + { + "epoch": 1.16096, + "grad_norm": 0.1514187902212143, + "learning_rate": 5.6618556701030935e-05, + "loss": 0.6961, + "step": 3628 + }, + { + "epoch": 1.16128, + "grad_norm": 0.1745612472295761, + "learning_rate": 5.657731958762887e-05, + "loss": 0.621, + "step": 3629 + }, + { + "epoch": 1.1616, + "grad_norm": 0.16520576179027557, + "learning_rate": 5.6536082474226805e-05, + "loss": 0.5765, + "step": 3630 + }, + { + "epoch": 1.16192, + "grad_norm": 0.16162173449993134, + "learning_rate": 5.649484536082474e-05, + "loss": 0.6829, + "step": 3631 + }, + { + "epoch": 1.16224, + "grad_norm": 0.1352095752954483, + "learning_rate": 5.645360824742269e-05, + "loss": 0.6463, + "step": 3632 + }, + { + "epoch": 1.16256, + "grad_norm": 0.17642022669315338, + "learning_rate": 5.6412371134020625e-05, + "loss": 0.7582, + "step": 3633 + }, + { + "epoch": 1.16288, + "grad_norm": 0.17780931293964386, + "learning_rate": 5.637113402061856e-05, + "loss": 0.5782, + "step": 3634 + }, + { + "epoch": 1.1632, + "grad_norm": 0.14908325672149658, + "learning_rate": 5.6329896907216495e-05, + "loss": 0.4494, + "step": 3635 + }, + { + "epoch": 1.16352, + "grad_norm": 0.15111255645751953, + "learning_rate": 5.628865979381443e-05, + "loss": 0.6255, + "step": 3636 + }, + { + "epoch": 1.16384, + "grad_norm": 0.16990917921066284, + "learning_rate": 5.624742268041238e-05, + "loss": 0.6684, + "step": 3637 + }, + { + "epoch": 1.16416, + "grad_norm": 0.17413973808288574, + "learning_rate": 5.6206185567010316e-05, + "loss": 0.7579, + "step": 3638 + }, + { + "epoch": 1.16448, + "grad_norm": 0.1476525217294693, + "learning_rate": 5.616494845360825e-05, + "loss": 0.5109, + "step": 3639 + }, + { + "epoch": 1.1648, + "grad_norm": 0.15902721881866455, + "learning_rate": 5.6123711340206185e-05, + "loss": 0.7173, + "step": 3640 + }, + { + "epoch": 1.16512, + "grad_norm": 0.180513396859169, + "learning_rate": 5.608247422680413e-05, + "loss": 0.7186, + "step": 3641 + }, + { + "epoch": 1.16544, + "grad_norm": 0.184820756316185, + "learning_rate": 5.604123711340207e-05, + "loss": 0.7613, + "step": 3642 + }, + { + "epoch": 1.16576, + "grad_norm": 0.16980108618736267, + "learning_rate": 5.6000000000000006e-05, + "loss": 0.6525, + "step": 3643 + }, + { + "epoch": 1.16608, + "grad_norm": 0.18438464403152466, + "learning_rate": 5.595876288659794e-05, + "loss": 0.5875, + "step": 3644 + }, + { + "epoch": 1.1663999999999999, + "grad_norm": 0.16444136202335358, + "learning_rate": 5.5917525773195875e-05, + "loss": 0.5731, + "step": 3645 + }, + { + "epoch": 1.16672, + "grad_norm": 0.18215300142765045, + "learning_rate": 5.587628865979382e-05, + "loss": 0.5865, + "step": 3646 + }, + { + "epoch": 1.16704, + "grad_norm": 0.16210678219795227, + "learning_rate": 5.583505154639176e-05, + "loss": 0.6511, + "step": 3647 + }, + { + "epoch": 1.16736, + "grad_norm": 0.15061765909194946, + "learning_rate": 5.579381443298969e-05, + "loss": 0.6394, + "step": 3648 + }, + { + "epoch": 1.16768, + "grad_norm": 0.157633438706398, + "learning_rate": 5.575257731958763e-05, + "loss": 0.5604, + "step": 3649 + }, + { + "epoch": 1.168, + "grad_norm": 0.17478065192699432, + "learning_rate": 5.571134020618557e-05, + "loss": 0.6402, + "step": 3650 + }, + { + "epoch": 1.16832, + "grad_norm": 0.15840552747249603, + "learning_rate": 5.567010309278351e-05, + "loss": 0.5045, + "step": 3651 + }, + { + "epoch": 1.16864, + "grad_norm": 0.1672770082950592, + "learning_rate": 5.562886597938145e-05, + "loss": 0.5504, + "step": 3652 + }, + { + "epoch": 1.16896, + "grad_norm": 0.1742299199104309, + "learning_rate": 5.558762886597938e-05, + "loss": 0.5375, + "step": 3653 + }, + { + "epoch": 1.16928, + "grad_norm": 0.15288464725017548, + "learning_rate": 5.554639175257732e-05, + "loss": 0.5451, + "step": 3654 + }, + { + "epoch": 1.1696, + "grad_norm": 0.16450318694114685, + "learning_rate": 5.550515463917526e-05, + "loss": 0.5019, + "step": 3655 + }, + { + "epoch": 1.16992, + "grad_norm": 0.168344646692276, + "learning_rate": 5.54639175257732e-05, + "loss": 0.5672, + "step": 3656 + }, + { + "epoch": 1.17024, + "grad_norm": 0.16811034083366394, + "learning_rate": 5.542268041237114e-05, + "loss": 0.5273, + "step": 3657 + }, + { + "epoch": 1.17056, + "grad_norm": 0.14858685433864594, + "learning_rate": 5.538144329896907e-05, + "loss": 0.5822, + "step": 3658 + }, + { + "epoch": 1.17088, + "grad_norm": 0.15245234966278076, + "learning_rate": 5.5340206185567015e-05, + "loss": 0.6663, + "step": 3659 + }, + { + "epoch": 1.1712, + "grad_norm": 0.13078056275844574, + "learning_rate": 5.529896907216495e-05, + "loss": 0.5445, + "step": 3660 + }, + { + "epoch": 1.1715200000000001, + "grad_norm": 0.1642111986875534, + "learning_rate": 5.525773195876289e-05, + "loss": 0.6605, + "step": 3661 + }, + { + "epoch": 1.17184, + "grad_norm": 0.161249041557312, + "learning_rate": 5.521649484536082e-05, + "loss": 0.6439, + "step": 3662 + }, + { + "epoch": 1.17216, + "grad_norm": 0.16502822935581207, + "learning_rate": 5.517525773195876e-05, + "loss": 0.6384, + "step": 3663 + }, + { + "epoch": 1.17248, + "grad_norm": 0.17600131034851074, + "learning_rate": 5.5134020618556705e-05, + "loss": 0.5549, + "step": 3664 + }, + { + "epoch": 1.1728, + "grad_norm": 0.16402779519557953, + "learning_rate": 5.509278350515464e-05, + "loss": 0.5588, + "step": 3665 + }, + { + "epoch": 1.17312, + "grad_norm": 0.17299668490886688, + "learning_rate": 5.505154639175258e-05, + "loss": 0.5501, + "step": 3666 + }, + { + "epoch": 1.17344, + "grad_norm": 0.16849659383296967, + "learning_rate": 5.501030927835051e-05, + "loss": 0.5524, + "step": 3667 + }, + { + "epoch": 1.17376, + "grad_norm": 0.14647427201271057, + "learning_rate": 5.496907216494846e-05, + "loss": 0.6124, + "step": 3668 + }, + { + "epoch": 1.17408, + "grad_norm": 0.17482984066009521, + "learning_rate": 5.4927835051546395e-05, + "loss": 0.6342, + "step": 3669 + }, + { + "epoch": 1.1743999999999999, + "grad_norm": 0.1597827821969986, + "learning_rate": 5.488659793814433e-05, + "loss": 0.4344, + "step": 3670 + }, + { + "epoch": 1.17472, + "grad_norm": 0.15515553951263428, + "learning_rate": 5.484536082474228e-05, + "loss": 0.5512, + "step": 3671 + }, + { + "epoch": 1.17504, + "grad_norm": 0.15960104763507843, + "learning_rate": 5.48041237113402e-05, + "loss": 0.6392, + "step": 3672 + }, + { + "epoch": 1.17536, + "grad_norm": 0.15532353520393372, + "learning_rate": 5.476288659793815e-05, + "loss": 0.6365, + "step": 3673 + }, + { + "epoch": 1.17568, + "grad_norm": 0.17004959285259247, + "learning_rate": 5.4721649484536085e-05, + "loss": 0.8895, + "step": 3674 + }, + { + "epoch": 1.176, + "grad_norm": 0.1705465167760849, + "learning_rate": 5.4680412371134024e-05, + "loss": 0.518, + "step": 3675 + }, + { + "epoch": 1.17632, + "grad_norm": 0.1794930249452591, + "learning_rate": 5.4639175257731955e-05, + "loss": 0.5239, + "step": 3676 + }, + { + "epoch": 1.17664, + "grad_norm": 0.1630152314901352, + "learning_rate": 5.45979381443299e-05, + "loss": 0.5635, + "step": 3677 + }, + { + "epoch": 1.17696, + "grad_norm": 0.1548643857240677, + "learning_rate": 5.455670103092784e-05, + "loss": 0.5947, + "step": 3678 + }, + { + "epoch": 1.17728, + "grad_norm": 0.1931552141904831, + "learning_rate": 5.4515463917525776e-05, + "loss": 0.6177, + "step": 3679 + }, + { + "epoch": 1.1776, + "grad_norm": 0.15705828368663788, + "learning_rate": 5.447422680412372e-05, + "loss": 0.6259, + "step": 3680 + }, + { + "epoch": 1.17792, + "grad_norm": 0.1707017719745636, + "learning_rate": 5.4432989690721645e-05, + "loss": 0.6558, + "step": 3681 + }, + { + "epoch": 1.17824, + "grad_norm": 0.15011698007583618, + "learning_rate": 5.439175257731959e-05, + "loss": 0.5197, + "step": 3682 + }, + { + "epoch": 1.17856, + "grad_norm": 0.14990614354610443, + "learning_rate": 5.435051546391753e-05, + "loss": 0.5677, + "step": 3683 + }, + { + "epoch": 1.17888, + "grad_norm": 0.1807318925857544, + "learning_rate": 5.4309278350515466e-05, + "loss": 0.5668, + "step": 3684 + }, + { + "epoch": 1.1792, + "grad_norm": 0.1779181808233261, + "learning_rate": 5.426804123711341e-05, + "loss": 0.715, + "step": 3685 + }, + { + "epoch": 1.1795200000000001, + "grad_norm": 0.16461992263793945, + "learning_rate": 5.422680412371134e-05, + "loss": 0.4617, + "step": 3686 + }, + { + "epoch": 1.17984, + "grad_norm": 0.1625106781721115, + "learning_rate": 5.418556701030928e-05, + "loss": 0.5647, + "step": 3687 + }, + { + "epoch": 1.1801599999999999, + "grad_norm": 0.18564371764659882, + "learning_rate": 5.414432989690722e-05, + "loss": 0.6231, + "step": 3688 + }, + { + "epoch": 1.18048, + "grad_norm": 0.15431852638721466, + "learning_rate": 5.410309278350516e-05, + "loss": 0.5038, + "step": 3689 + }, + { + "epoch": 1.1808, + "grad_norm": 0.16544587910175323, + "learning_rate": 5.406185567010309e-05, + "loss": 0.6491, + "step": 3690 + }, + { + "epoch": 1.18112, + "grad_norm": 0.19189564883708954, + "learning_rate": 5.402061855670103e-05, + "loss": 0.7599, + "step": 3691 + }, + { + "epoch": 1.18144, + "grad_norm": 0.1700255423784256, + "learning_rate": 5.397938144329897e-05, + "loss": 0.7249, + "step": 3692 + }, + { + "epoch": 1.18176, + "grad_norm": 0.18982042372226715, + "learning_rate": 5.393814432989691e-05, + "loss": 0.5685, + "step": 3693 + }, + { + "epoch": 1.18208, + "grad_norm": 0.16703614592552185, + "learning_rate": 5.389690721649485e-05, + "loss": 0.7333, + "step": 3694 + }, + { + "epoch": 1.1824, + "grad_norm": 0.14930245280265808, + "learning_rate": 5.3855670103092785e-05, + "loss": 0.6942, + "step": 3695 + }, + { + "epoch": 1.18272, + "grad_norm": 0.17281466722488403, + "learning_rate": 5.381443298969072e-05, + "loss": 0.5172, + "step": 3696 + }, + { + "epoch": 1.18304, + "grad_norm": 0.15837645530700684, + "learning_rate": 5.377319587628866e-05, + "loss": 0.7164, + "step": 3697 + }, + { + "epoch": 1.18336, + "grad_norm": 0.1876179426908493, + "learning_rate": 5.3731958762886605e-05, + "loss": 0.6342, + "step": 3698 + }, + { + "epoch": 1.18368, + "grad_norm": 0.18273694813251495, + "learning_rate": 5.3690721649484544e-05, + "loss": 0.7043, + "step": 3699 + }, + { + "epoch": 1.184, + "grad_norm": 0.16980963945388794, + "learning_rate": 5.3649484536082475e-05, + "loss": 0.6014, + "step": 3700 + }, + { + "epoch": 1.18432, + "grad_norm": 0.16482768952846527, + "learning_rate": 5.360824742268041e-05, + "loss": 0.5947, + "step": 3701 + }, + { + "epoch": 1.18464, + "grad_norm": 0.16299095749855042, + "learning_rate": 5.356701030927835e-05, + "loss": 0.6648, + "step": 3702 + }, + { + "epoch": 1.18496, + "grad_norm": 0.16586996614933014, + "learning_rate": 5.3525773195876296e-05, + "loss": 0.7024, + "step": 3703 + }, + { + "epoch": 1.1852800000000001, + "grad_norm": 0.16753004491329193, + "learning_rate": 5.3484536082474234e-05, + "loss": 0.7137, + "step": 3704 + }, + { + "epoch": 1.1856, + "grad_norm": 0.1559315025806427, + "learning_rate": 5.3443298969072165e-05, + "loss": 0.4983, + "step": 3705 + }, + { + "epoch": 1.18592, + "grad_norm": 0.15365763008594513, + "learning_rate": 5.34020618556701e-05, + "loss": 0.5101, + "step": 3706 + }, + { + "epoch": 1.18624, + "grad_norm": 0.180158331990242, + "learning_rate": 5.336082474226805e-05, + "loss": 0.5849, + "step": 3707 + }, + { + "epoch": 1.18656, + "grad_norm": 0.18548215925693512, + "learning_rate": 5.3319587628865986e-05, + "loss": 0.6192, + "step": 3708 + }, + { + "epoch": 1.18688, + "grad_norm": 0.1429319828748703, + "learning_rate": 5.327835051546392e-05, + "loss": 0.6515, + "step": 3709 + }, + { + "epoch": 1.1872, + "grad_norm": 0.16380463540554047, + "learning_rate": 5.3237113402061855e-05, + "loss": 0.5839, + "step": 3710 + }, + { + "epoch": 1.18752, + "grad_norm": 0.1544746607542038, + "learning_rate": 5.319587628865979e-05, + "loss": 0.5701, + "step": 3711 + }, + { + "epoch": 1.18784, + "grad_norm": 0.18146944046020508, + "learning_rate": 5.315463917525774e-05, + "loss": 0.5135, + "step": 3712 + }, + { + "epoch": 1.1881599999999999, + "grad_norm": 0.16630977392196655, + "learning_rate": 5.3113402061855676e-05, + "loss": 0.807, + "step": 3713 + }, + { + "epoch": 1.18848, + "grad_norm": 0.20390062034130096, + "learning_rate": 5.307216494845361e-05, + "loss": 0.6364, + "step": 3714 + }, + { + "epoch": 1.1888, + "grad_norm": 0.17039437592029572, + "learning_rate": 5.3030927835051546e-05, + "loss": 0.559, + "step": 3715 + }, + { + "epoch": 1.18912, + "grad_norm": 0.1545945107936859, + "learning_rate": 5.298969072164949e-05, + "loss": 0.6274, + "step": 3716 + }, + { + "epoch": 1.18944, + "grad_norm": 0.1594141125679016, + "learning_rate": 5.294845360824743e-05, + "loss": 0.6523, + "step": 3717 + }, + { + "epoch": 1.18976, + "grad_norm": 0.17188358306884766, + "learning_rate": 5.2907216494845366e-05, + "loss": 0.5997, + "step": 3718 + }, + { + "epoch": 1.19008, + "grad_norm": 0.17760074138641357, + "learning_rate": 5.28659793814433e-05, + "loss": 0.5243, + "step": 3719 + }, + { + "epoch": 1.1904, + "grad_norm": 0.139976367354393, + "learning_rate": 5.2824742268041236e-05, + "loss": 0.5421, + "step": 3720 + }, + { + "epoch": 1.19072, + "grad_norm": 0.18369121849536896, + "learning_rate": 5.278350515463918e-05, + "loss": 0.6367, + "step": 3721 + }, + { + "epoch": 1.19104, + "grad_norm": 0.17412689328193665, + "learning_rate": 5.274226804123712e-05, + "loss": 0.6983, + "step": 3722 + }, + { + "epoch": 1.19136, + "grad_norm": 0.1687963455915451, + "learning_rate": 5.270103092783505e-05, + "loss": 0.7777, + "step": 3723 + }, + { + "epoch": 1.19168, + "grad_norm": 0.15787208080291748, + "learning_rate": 5.265979381443299e-05, + "loss": 0.6715, + "step": 3724 + }, + { + "epoch": 1.192, + "grad_norm": 0.17575743794441223, + "learning_rate": 5.261855670103093e-05, + "loss": 0.6506, + "step": 3725 + }, + { + "epoch": 1.19232, + "grad_norm": 0.19815565645694733, + "learning_rate": 5.257731958762887e-05, + "loss": 0.5428, + "step": 3726 + }, + { + "epoch": 1.19264, + "grad_norm": 0.1689406931400299, + "learning_rate": 5.253608247422681e-05, + "loss": 0.7198, + "step": 3727 + }, + { + "epoch": 1.19296, + "grad_norm": 0.1571311503648758, + "learning_rate": 5.249484536082474e-05, + "loss": 0.5267, + "step": 3728 + }, + { + "epoch": 1.1932800000000001, + "grad_norm": 0.17965692281723022, + "learning_rate": 5.245360824742268e-05, + "loss": 0.6016, + "step": 3729 + }, + { + "epoch": 1.1936, + "grad_norm": 0.16691434383392334, + "learning_rate": 5.241237113402062e-05, + "loss": 0.6555, + "step": 3730 + }, + { + "epoch": 1.19392, + "grad_norm": 0.16236045956611633, + "learning_rate": 5.237113402061856e-05, + "loss": 0.4557, + "step": 3731 + }, + { + "epoch": 1.19424, + "grad_norm": 0.18354065716266632, + "learning_rate": 5.23298969072165e-05, + "loss": 0.583, + "step": 3732 + }, + { + "epoch": 1.19456, + "grad_norm": 0.1882420927286148, + "learning_rate": 5.228865979381443e-05, + "loss": 0.7055, + "step": 3733 + }, + { + "epoch": 1.19488, + "grad_norm": 0.15704384446144104, + "learning_rate": 5.2247422680412375e-05, + "loss": 0.6053, + "step": 3734 + }, + { + "epoch": 1.1952, + "grad_norm": 0.1488439440727234, + "learning_rate": 5.220618556701031e-05, + "loss": 0.523, + "step": 3735 + }, + { + "epoch": 1.19552, + "grad_norm": 0.15782950818538666, + "learning_rate": 5.216494845360825e-05, + "loss": 0.5029, + "step": 3736 + }, + { + "epoch": 1.19584, + "grad_norm": 0.1471652388572693, + "learning_rate": 5.212371134020618e-05, + "loss": 0.6854, + "step": 3737 + }, + { + "epoch": 1.19616, + "grad_norm": 0.17158359289169312, + "learning_rate": 5.208247422680412e-05, + "loss": 0.5529, + "step": 3738 + }, + { + "epoch": 1.19648, + "grad_norm": 0.15647731721401215, + "learning_rate": 5.2041237113402066e-05, + "loss": 0.6184, + "step": 3739 + }, + { + "epoch": 1.1968, + "grad_norm": 0.1597009152173996, + "learning_rate": 5.2000000000000004e-05, + "loss": 0.6303, + "step": 3740 + }, + { + "epoch": 1.19712, + "grad_norm": 0.15556739270687103, + "learning_rate": 5.195876288659794e-05, + "loss": 0.6117, + "step": 3741 + }, + { + "epoch": 1.19744, + "grad_norm": 0.15222102403640747, + "learning_rate": 5.191752577319587e-05, + "loss": 0.6533, + "step": 3742 + }, + { + "epoch": 1.19776, + "grad_norm": 0.17943742871284485, + "learning_rate": 5.187628865979382e-05, + "loss": 0.6596, + "step": 3743 + }, + { + "epoch": 1.19808, + "grad_norm": 0.1553201973438263, + "learning_rate": 5.1835051546391756e-05, + "loss": 0.624, + "step": 3744 + }, + { + "epoch": 1.1984, + "grad_norm": 0.18606509268283844, + "learning_rate": 5.1793814432989694e-05, + "loss": 0.5836, + "step": 3745 + }, + { + "epoch": 1.19872, + "grad_norm": 0.18627767264842987, + "learning_rate": 5.175257731958764e-05, + "loss": 0.5614, + "step": 3746 + }, + { + "epoch": 1.19904, + "grad_norm": 0.1716335266828537, + "learning_rate": 5.171134020618556e-05, + "loss": 0.6572, + "step": 3747 + }, + { + "epoch": 1.19936, + "grad_norm": 0.1577850878238678, + "learning_rate": 5.167010309278351e-05, + "loss": 0.5032, + "step": 3748 + }, + { + "epoch": 1.19968, + "grad_norm": 0.180816650390625, + "learning_rate": 5.1628865979381446e-05, + "loss": 0.5665, + "step": 3749 + }, + { + "epoch": 1.2, + "grad_norm": 0.16079489886760712, + "learning_rate": 5.1587628865979384e-05, + "loss": 0.6695, + "step": 3750 + }, + { + "epoch": 1.20032, + "grad_norm": 0.15245407819747925, + "learning_rate": 5.1546391752577315e-05, + "loss": 0.7045, + "step": 3751 + }, + { + "epoch": 1.20064, + "grad_norm": 0.16249269247055054, + "learning_rate": 5.150515463917526e-05, + "loss": 0.6178, + "step": 3752 + }, + { + "epoch": 1.20096, + "grad_norm": 0.1549021452665329, + "learning_rate": 5.14639175257732e-05, + "loss": 0.652, + "step": 3753 + }, + { + "epoch": 1.20128, + "grad_norm": 0.20225249230861664, + "learning_rate": 5.1422680412371136e-05, + "loss": 0.6569, + "step": 3754 + }, + { + "epoch": 1.2016, + "grad_norm": 0.18245147168636322, + "learning_rate": 5.138144329896908e-05, + "loss": 0.514, + "step": 3755 + }, + { + "epoch": 1.2019199999999999, + "grad_norm": 0.1648096889257431, + "learning_rate": 5.1340206185567006e-05, + "loss": 0.4863, + "step": 3756 + }, + { + "epoch": 1.20224, + "grad_norm": 0.16288764774799347, + "learning_rate": 5.129896907216495e-05, + "loss": 0.8185, + "step": 3757 + }, + { + "epoch": 1.20256, + "grad_norm": 0.1517041176557541, + "learning_rate": 5.125773195876289e-05, + "loss": 0.5156, + "step": 3758 + }, + { + "epoch": 1.20288, + "grad_norm": 0.1876288652420044, + "learning_rate": 5.1216494845360827e-05, + "loss": 0.6224, + "step": 3759 + }, + { + "epoch": 1.2032, + "grad_norm": 0.17451636493206024, + "learning_rate": 5.117525773195877e-05, + "loss": 0.6312, + "step": 3760 + }, + { + "epoch": 1.20352, + "grad_norm": 0.15739643573760986, + "learning_rate": 5.11340206185567e-05, + "loss": 0.6543, + "step": 3761 + }, + { + "epoch": 1.20384, + "grad_norm": 0.1767333745956421, + "learning_rate": 5.109278350515464e-05, + "loss": 0.6876, + "step": 3762 + }, + { + "epoch": 1.20416, + "grad_norm": 0.17043988406658173, + "learning_rate": 5.105154639175258e-05, + "loss": 0.617, + "step": 3763 + }, + { + "epoch": 1.20448, + "grad_norm": 0.17414774000644684, + "learning_rate": 5.1010309278350524e-05, + "loss": 0.6953, + "step": 3764 + }, + { + "epoch": 1.2048, + "grad_norm": 0.16423740983009338, + "learning_rate": 5.096907216494846e-05, + "loss": 0.7349, + "step": 3765 + }, + { + "epoch": 1.20512, + "grad_norm": 0.17699885368347168, + "learning_rate": 5.092783505154639e-05, + "loss": 0.7025, + "step": 3766 + }, + { + "epoch": 1.20544, + "grad_norm": 0.15230748057365417, + "learning_rate": 5.088659793814433e-05, + "loss": 0.5816, + "step": 3767 + }, + { + "epoch": 1.20576, + "grad_norm": 0.17530299723148346, + "learning_rate": 5.084536082474227e-05, + "loss": 0.6346, + "step": 3768 + }, + { + "epoch": 1.20608, + "grad_norm": 0.17154143750667572, + "learning_rate": 5.0804123711340214e-05, + "loss": 0.5121, + "step": 3769 + }, + { + "epoch": 1.2064, + "grad_norm": 0.17974890768527985, + "learning_rate": 5.0762886597938145e-05, + "loss": 0.777, + "step": 3770 + }, + { + "epoch": 1.20672, + "grad_norm": 0.1793683022260666, + "learning_rate": 5.072164948453608e-05, + "loss": 0.6657, + "step": 3771 + }, + { + "epoch": 1.2070400000000001, + "grad_norm": 0.1667945235967636, + "learning_rate": 5.068041237113402e-05, + "loss": 0.5496, + "step": 3772 + }, + { + "epoch": 1.20736, + "grad_norm": 0.17806769907474518, + "learning_rate": 5.0639175257731966e-05, + "loss": 0.715, + "step": 3773 + }, + { + "epoch": 1.20768, + "grad_norm": 0.14589308202266693, + "learning_rate": 5.0597938144329904e-05, + "loss": 0.5223, + "step": 3774 + }, + { + "epoch": 1.208, + "grad_norm": 0.17671100795269012, + "learning_rate": 5.0556701030927835e-05, + "loss": 0.571, + "step": 3775 + }, + { + "epoch": 1.20832, + "grad_norm": 0.1784905046224594, + "learning_rate": 5.051546391752577e-05, + "loss": 0.6227, + "step": 3776 + }, + { + "epoch": 1.20864, + "grad_norm": 0.1575552374124527, + "learning_rate": 5.047422680412371e-05, + "loss": 0.7301, + "step": 3777 + }, + { + "epoch": 1.20896, + "grad_norm": 0.17384012043476105, + "learning_rate": 5.0432989690721656e-05, + "loss": 0.6096, + "step": 3778 + }, + { + "epoch": 1.20928, + "grad_norm": 0.17526081204414368, + "learning_rate": 5.0391752577319594e-05, + "loss": 0.5793, + "step": 3779 + }, + { + "epoch": 1.2096, + "grad_norm": 0.16879479587078094, + "learning_rate": 5.0350515463917526e-05, + "loss": 0.6474, + "step": 3780 + }, + { + "epoch": 1.2099199999999999, + "grad_norm": 0.18120743334293365, + "learning_rate": 5.0309278350515464e-05, + "loss": 0.6916, + "step": 3781 + }, + { + "epoch": 1.21024, + "grad_norm": 0.15673841536045074, + "learning_rate": 5.026804123711341e-05, + "loss": 0.6571, + "step": 3782 + }, + { + "epoch": 1.21056, + "grad_norm": 0.18390482664108276, + "learning_rate": 5.0226804123711347e-05, + "loss": 0.6658, + "step": 3783 + }, + { + "epoch": 1.21088, + "grad_norm": 0.15100419521331787, + "learning_rate": 5.018556701030928e-05, + "loss": 0.5682, + "step": 3784 + }, + { + "epoch": 1.2112, + "grad_norm": 0.19734734296798706, + "learning_rate": 5.0144329896907216e-05, + "loss": 0.6365, + "step": 3785 + }, + { + "epoch": 1.21152, + "grad_norm": 0.15656329691410065, + "learning_rate": 5.0103092783505154e-05, + "loss": 0.5647, + "step": 3786 + }, + { + "epoch": 1.21184, + "grad_norm": 0.16992604732513428, + "learning_rate": 5.00618556701031e-05, + "loss": 0.5712, + "step": 3787 + }, + { + "epoch": 1.21216, + "grad_norm": 0.16440635919570923, + "learning_rate": 5.002061855670104e-05, + "loss": 0.483, + "step": 3788 + }, + { + "epoch": 1.21248, + "grad_norm": 0.1666940301656723, + "learning_rate": 4.9979381443298975e-05, + "loss": 0.6141, + "step": 3789 + }, + { + "epoch": 1.2128, + "grad_norm": 0.18127204477787018, + "learning_rate": 4.9938144329896906e-05, + "loss": 0.5371, + "step": 3790 + }, + { + "epoch": 1.21312, + "grad_norm": 0.14934316277503967, + "learning_rate": 4.989690721649485e-05, + "loss": 0.6739, + "step": 3791 + }, + { + "epoch": 1.21344, + "grad_norm": 0.17686113715171814, + "learning_rate": 4.985567010309279e-05, + "loss": 0.6935, + "step": 3792 + }, + { + "epoch": 1.21376, + "grad_norm": 0.14645908772945404, + "learning_rate": 4.981443298969072e-05, + "loss": 0.7485, + "step": 3793 + }, + { + "epoch": 1.21408, + "grad_norm": 0.19578814506530762, + "learning_rate": 4.9773195876288665e-05, + "loss": 0.6158, + "step": 3794 + }, + { + "epoch": 1.2144, + "grad_norm": 0.18915681540966034, + "learning_rate": 4.9731958762886596e-05, + "loss": 0.724, + "step": 3795 + }, + { + "epoch": 1.21472, + "grad_norm": 0.15513186156749725, + "learning_rate": 4.969072164948454e-05, + "loss": 0.5343, + "step": 3796 + }, + { + "epoch": 1.2150400000000001, + "grad_norm": 0.17371588945388794, + "learning_rate": 4.964948453608247e-05, + "loss": 0.6776, + "step": 3797 + }, + { + "epoch": 1.21536, + "grad_norm": 0.18510079383850098, + "learning_rate": 4.960824742268042e-05, + "loss": 0.588, + "step": 3798 + }, + { + "epoch": 1.21568, + "grad_norm": 0.15795986354351044, + "learning_rate": 4.9567010309278355e-05, + "loss": 0.6268, + "step": 3799 + }, + { + "epoch": 1.216, + "grad_norm": 0.14822380244731903, + "learning_rate": 4.952577319587629e-05, + "loss": 0.5801, + "step": 3800 + }, + { + "epoch": 1.21632, + "grad_norm": 0.17550544440746307, + "learning_rate": 4.948453608247423e-05, + "loss": 0.5323, + "step": 3801 + }, + { + "epoch": 1.21664, + "grad_norm": 0.16116420924663544, + "learning_rate": 4.944329896907216e-05, + "loss": 0.6009, + "step": 3802 + }, + { + "epoch": 1.21696, + "grad_norm": 0.15281707048416138, + "learning_rate": 4.940206185567011e-05, + "loss": 0.5828, + "step": 3803 + }, + { + "epoch": 1.21728, + "grad_norm": 0.17293612658977509, + "learning_rate": 4.936082474226804e-05, + "loss": 0.7453, + "step": 3804 + }, + { + "epoch": 1.2176, + "grad_norm": 0.18403638899326324, + "learning_rate": 4.9319587628865984e-05, + "loss": 0.6568, + "step": 3805 + }, + { + "epoch": 1.21792, + "grad_norm": 0.17931637167930603, + "learning_rate": 4.927835051546392e-05, + "loss": 0.6428, + "step": 3806 + }, + { + "epoch": 1.21824, + "grad_norm": 0.17066961526870728, + "learning_rate": 4.923711340206186e-05, + "loss": 0.546, + "step": 3807 + }, + { + "epoch": 1.21856, + "grad_norm": 0.1442670375108719, + "learning_rate": 4.91958762886598e-05, + "loss": 0.6811, + "step": 3808 + }, + { + "epoch": 1.21888, + "grad_norm": 0.17498700320720673, + "learning_rate": 4.9154639175257736e-05, + "loss": 0.5, + "step": 3809 + }, + { + "epoch": 1.2192, + "grad_norm": 0.15666191279888153, + "learning_rate": 4.9113402061855674e-05, + "loss": 0.6302, + "step": 3810 + }, + { + "epoch": 1.21952, + "grad_norm": 0.19409668445587158, + "learning_rate": 4.9072164948453605e-05, + "loss": 0.728, + "step": 3811 + }, + { + "epoch": 1.21984, + "grad_norm": 0.15672099590301514, + "learning_rate": 4.903092783505155e-05, + "loss": 0.5403, + "step": 3812 + }, + { + "epoch": 1.22016, + "grad_norm": 0.1535264104604721, + "learning_rate": 4.898969072164949e-05, + "loss": 0.4451, + "step": 3813 + }, + { + "epoch": 1.22048, + "grad_norm": 0.1800055354833603, + "learning_rate": 4.8948453608247426e-05, + "loss": 0.5328, + "step": 3814 + }, + { + "epoch": 1.2208, + "grad_norm": 0.17271973192691803, + "learning_rate": 4.8907216494845364e-05, + "loss": 0.6196, + "step": 3815 + }, + { + "epoch": 1.22112, + "grad_norm": 0.18747521936893463, + "learning_rate": 4.88659793814433e-05, + "loss": 0.7363, + "step": 3816 + }, + { + "epoch": 1.22144, + "grad_norm": 0.16658920049667358, + "learning_rate": 4.882474226804124e-05, + "loss": 0.6566, + "step": 3817 + }, + { + "epoch": 1.22176, + "grad_norm": 0.1810402125120163, + "learning_rate": 4.878350515463918e-05, + "loss": 0.5574, + "step": 3818 + }, + { + "epoch": 1.22208, + "grad_norm": 0.17182011902332306, + "learning_rate": 4.8742268041237116e-05, + "loss": 0.7017, + "step": 3819 + }, + { + "epoch": 1.2224, + "grad_norm": 0.17430496215820312, + "learning_rate": 4.8701030927835054e-05, + "loss": 0.6131, + "step": 3820 + }, + { + "epoch": 1.22272, + "grad_norm": 0.19244657456874847, + "learning_rate": 4.865979381443299e-05, + "loss": 0.6974, + "step": 3821 + }, + { + "epoch": 1.22304, + "grad_norm": 0.15759216248989105, + "learning_rate": 4.861855670103093e-05, + "loss": 0.6275, + "step": 3822 + }, + { + "epoch": 1.22336, + "grad_norm": 0.16780027747154236, + "learning_rate": 4.857731958762887e-05, + "loss": 0.4873, + "step": 3823 + }, + { + "epoch": 1.2236799999999999, + "grad_norm": 0.15435408055782318, + "learning_rate": 4.8536082474226807e-05, + "loss": 0.4692, + "step": 3824 + }, + { + "epoch": 1.224, + "grad_norm": 0.15156111121177673, + "learning_rate": 4.8494845360824745e-05, + "loss": 0.5, + "step": 3825 + }, + { + "epoch": 1.22432, + "grad_norm": 0.14709028601646423, + "learning_rate": 4.845360824742268e-05, + "loss": 0.6588, + "step": 3826 + }, + { + "epoch": 1.22464, + "grad_norm": 0.16893576085567474, + "learning_rate": 4.841237113402062e-05, + "loss": 0.6586, + "step": 3827 + }, + { + "epoch": 1.22496, + "grad_norm": 0.1782447248697281, + "learning_rate": 4.837113402061856e-05, + "loss": 0.6119, + "step": 3828 + }, + { + "epoch": 1.22528, + "grad_norm": 0.1754632145166397, + "learning_rate": 4.83298969072165e-05, + "loss": 0.6027, + "step": 3829 + }, + { + "epoch": 1.2256, + "grad_norm": 0.17047438025474548, + "learning_rate": 4.8288659793814435e-05, + "loss": 0.489, + "step": 3830 + }, + { + "epoch": 1.22592, + "grad_norm": 0.17885352671146393, + "learning_rate": 4.824742268041237e-05, + "loss": 0.7439, + "step": 3831 + }, + { + "epoch": 1.22624, + "grad_norm": 0.15219414234161377, + "learning_rate": 4.820618556701031e-05, + "loss": 0.6006, + "step": 3832 + }, + { + "epoch": 1.22656, + "grad_norm": 0.16418811678886414, + "learning_rate": 4.816494845360825e-05, + "loss": 0.6278, + "step": 3833 + }, + { + "epoch": 1.22688, + "grad_norm": 0.15196864306926727, + "learning_rate": 4.812371134020619e-05, + "loss": 0.5269, + "step": 3834 + }, + { + "epoch": 1.2272, + "grad_norm": 0.17161177098751068, + "learning_rate": 4.8082474226804125e-05, + "loss": 0.5893, + "step": 3835 + }, + { + "epoch": 1.22752, + "grad_norm": 0.18113063275814056, + "learning_rate": 4.804123711340206e-05, + "loss": 0.618, + "step": 3836 + }, + { + "epoch": 1.22784, + "grad_norm": 0.1861868053674698, + "learning_rate": 4.8e-05, + "loss": 0.6125, + "step": 3837 + }, + { + "epoch": 1.22816, + "grad_norm": 0.17038686573505402, + "learning_rate": 4.795876288659794e-05, + "loss": 0.637, + "step": 3838 + }, + { + "epoch": 1.22848, + "grad_norm": 0.18184801936149597, + "learning_rate": 4.791752577319588e-05, + "loss": 0.5468, + "step": 3839 + }, + { + "epoch": 1.2288000000000001, + "grad_norm": 0.18890900909900665, + "learning_rate": 4.7876288659793815e-05, + "loss": 0.6845, + "step": 3840 + }, + { + "epoch": 1.22912, + "grad_norm": 0.16302047669887543, + "learning_rate": 4.783505154639176e-05, + "loss": 0.4133, + "step": 3841 + }, + { + "epoch": 1.22944, + "grad_norm": 0.16767318546772003, + "learning_rate": 4.779381443298969e-05, + "loss": 0.6177, + "step": 3842 + }, + { + "epoch": 1.22976, + "grad_norm": 0.15622444450855255, + "learning_rate": 4.775257731958763e-05, + "loss": 0.5242, + "step": 3843 + }, + { + "epoch": 1.23008, + "grad_norm": 0.16470003128051758, + "learning_rate": 4.771134020618557e-05, + "loss": 0.4994, + "step": 3844 + }, + { + "epoch": 1.2304, + "grad_norm": 0.18380793929100037, + "learning_rate": 4.7670103092783506e-05, + "loss": 0.5817, + "step": 3845 + }, + { + "epoch": 1.23072, + "grad_norm": 0.16757820546627045, + "learning_rate": 4.7628865979381444e-05, + "loss": 0.7935, + "step": 3846 + }, + { + "epoch": 1.23104, + "grad_norm": 0.18820933997631073, + "learning_rate": 4.758762886597938e-05, + "loss": 0.6396, + "step": 3847 + }, + { + "epoch": 1.23136, + "grad_norm": 0.180404931306839, + "learning_rate": 4.7546391752577327e-05, + "loss": 0.522, + "step": 3848 + }, + { + "epoch": 1.2316799999999999, + "grad_norm": 0.15326844155788422, + "learning_rate": 4.750515463917526e-05, + "loss": 0.7603, + "step": 3849 + }, + { + "epoch": 1.232, + "grad_norm": 0.18296387791633606, + "learning_rate": 4.74639175257732e-05, + "loss": 0.5619, + "step": 3850 + }, + { + "epoch": 1.23232, + "grad_norm": 0.1817701756954193, + "learning_rate": 4.7422680412371134e-05, + "loss": 0.6294, + "step": 3851 + }, + { + "epoch": 1.23264, + "grad_norm": 0.15849092602729797, + "learning_rate": 4.738144329896907e-05, + "loss": 0.4907, + "step": 3852 + }, + { + "epoch": 1.23296, + "grad_norm": 0.15904287993907928, + "learning_rate": 4.734020618556701e-05, + "loss": 0.5125, + "step": 3853 + }, + { + "epoch": 1.23328, + "grad_norm": 0.15400706231594086, + "learning_rate": 4.729896907216495e-05, + "loss": 0.8262, + "step": 3854 + }, + { + "epoch": 1.2336, + "grad_norm": 0.1631423532962799, + "learning_rate": 4.725773195876289e-05, + "loss": 0.592, + "step": 3855 + }, + { + "epoch": 1.23392, + "grad_norm": 0.15791569650173187, + "learning_rate": 4.7216494845360824e-05, + "loss": 0.6807, + "step": 3856 + }, + { + "epoch": 1.23424, + "grad_norm": 0.16565977036952972, + "learning_rate": 4.717525773195877e-05, + "loss": 0.51, + "step": 3857 + }, + { + "epoch": 1.23456, + "grad_norm": 0.18299846351146698, + "learning_rate": 4.71340206185567e-05, + "loss": 0.5861, + "step": 3858 + }, + { + "epoch": 1.23488, + "grad_norm": 0.21313650906085968, + "learning_rate": 4.7092783505154645e-05, + "loss": 0.7502, + "step": 3859 + }, + { + "epoch": 1.2352, + "grad_norm": 0.17559814453125, + "learning_rate": 4.705154639175258e-05, + "loss": 0.6097, + "step": 3860 + }, + { + "epoch": 1.23552, + "grad_norm": 0.15951624512672424, + "learning_rate": 4.7010309278350514e-05, + "loss": 0.566, + "step": 3861 + }, + { + "epoch": 1.23584, + "grad_norm": 0.14401386678218842, + "learning_rate": 4.696907216494846e-05, + "loss": 0.6604, + "step": 3862 + }, + { + "epoch": 1.23616, + "grad_norm": 0.1666601598262787, + "learning_rate": 4.692783505154639e-05, + "loss": 0.4801, + "step": 3863 + }, + { + "epoch": 1.23648, + "grad_norm": 0.17491166293621063, + "learning_rate": 4.6886597938144335e-05, + "loss": 0.5297, + "step": 3864 + }, + { + "epoch": 1.2368000000000001, + "grad_norm": 0.15945620834827423, + "learning_rate": 4.684536082474227e-05, + "loss": 0.6219, + "step": 3865 + }, + { + "epoch": 1.23712, + "grad_norm": 0.15754877030849457, + "learning_rate": 4.680412371134021e-05, + "loss": 0.5237, + "step": 3866 + }, + { + "epoch": 1.23744, + "grad_norm": 0.15032131969928741, + "learning_rate": 4.676288659793815e-05, + "loss": 0.6611, + "step": 3867 + }, + { + "epoch": 1.23776, + "grad_norm": 0.16981737315654755, + "learning_rate": 4.672164948453608e-05, + "loss": 0.5923, + "step": 3868 + }, + { + "epoch": 1.23808, + "grad_norm": 0.17241406440734863, + "learning_rate": 4.6680412371134026e-05, + "loss": 0.5006, + "step": 3869 + }, + { + "epoch": 1.2384, + "grad_norm": 0.1760787069797516, + "learning_rate": 4.663917525773196e-05, + "loss": 0.5645, + "step": 3870 + }, + { + "epoch": 1.23872, + "grad_norm": 0.17664313316345215, + "learning_rate": 4.65979381443299e-05, + "loss": 0.6841, + "step": 3871 + }, + { + "epoch": 1.23904, + "grad_norm": 0.15562033653259277, + "learning_rate": 4.655670103092783e-05, + "loss": 0.7131, + "step": 3872 + }, + { + "epoch": 1.23936, + "grad_norm": 0.17468054592609406, + "learning_rate": 4.651546391752578e-05, + "loss": 0.7039, + "step": 3873 + }, + { + "epoch": 1.23968, + "grad_norm": 0.16800588369369507, + "learning_rate": 4.6474226804123716e-05, + "loss": 0.6387, + "step": 3874 + }, + { + "epoch": 1.24, + "grad_norm": 0.17324158549308777, + "learning_rate": 4.6432989690721654e-05, + "loss": 0.6951, + "step": 3875 + }, + { + "epoch": 1.24032, + "grad_norm": 0.1691412478685379, + "learning_rate": 4.639175257731959e-05, + "loss": 0.5878, + "step": 3876 + }, + { + "epoch": 1.24064, + "grad_norm": 0.16370351612567902, + "learning_rate": 4.635051546391752e-05, + "loss": 0.5483, + "step": 3877 + }, + { + "epoch": 1.24096, + "grad_norm": 0.16407427191734314, + "learning_rate": 4.630927835051547e-05, + "loss": 0.6548, + "step": 3878 + }, + { + "epoch": 1.24128, + "grad_norm": 0.1946336030960083, + "learning_rate": 4.62680412371134e-05, + "loss": 0.6243, + "step": 3879 + }, + { + "epoch": 1.2416, + "grad_norm": 0.1654862016439438, + "learning_rate": 4.6226804123711344e-05, + "loss": 0.6018, + "step": 3880 + }, + { + "epoch": 1.24192, + "grad_norm": 0.1829860508441925, + "learning_rate": 4.618556701030928e-05, + "loss": 0.7526, + "step": 3881 + }, + { + "epoch": 1.24224, + "grad_norm": 0.15461423993110657, + "learning_rate": 4.614432989690722e-05, + "loss": 0.4809, + "step": 3882 + }, + { + "epoch": 1.24256, + "grad_norm": 0.17071287333965302, + "learning_rate": 4.610309278350516e-05, + "loss": 0.5123, + "step": 3883 + }, + { + "epoch": 1.24288, + "grad_norm": 0.17397645115852356, + "learning_rate": 4.6061855670103096e-05, + "loss": 0.5579, + "step": 3884 + }, + { + "epoch": 1.2432, + "grad_norm": 0.1666763573884964, + "learning_rate": 4.6020618556701034e-05, + "loss": 0.5783, + "step": 3885 + }, + { + "epoch": 1.24352, + "grad_norm": 0.1614886373281479, + "learning_rate": 4.5979381443298966e-05, + "loss": 0.6572, + "step": 3886 + }, + { + "epoch": 1.24384, + "grad_norm": 0.17060226202011108, + "learning_rate": 4.593814432989691e-05, + "loss": 0.5795, + "step": 3887 + }, + { + "epoch": 1.24416, + "grad_norm": 0.18600896000862122, + "learning_rate": 4.589690721649485e-05, + "loss": 0.608, + "step": 3888 + }, + { + "epoch": 1.24448, + "grad_norm": 0.16216692328453064, + "learning_rate": 4.5855670103092787e-05, + "loss": 0.5546, + "step": 3889 + }, + { + "epoch": 1.2448, + "grad_norm": 0.17214037477970123, + "learning_rate": 4.5814432989690725e-05, + "loss": 0.5771, + "step": 3890 + }, + { + "epoch": 1.24512, + "grad_norm": 0.15760190784931183, + "learning_rate": 4.577319587628866e-05, + "loss": 0.6904, + "step": 3891 + }, + { + "epoch": 1.2454399999999999, + "grad_norm": 0.1580679565668106, + "learning_rate": 4.57319587628866e-05, + "loss": 0.7266, + "step": 3892 + }, + { + "epoch": 1.24576, + "grad_norm": 0.1528448462486267, + "learning_rate": 4.569072164948454e-05, + "loss": 0.6766, + "step": 3893 + }, + { + "epoch": 1.24608, + "grad_norm": 0.1720484048128128, + "learning_rate": 4.564948453608248e-05, + "loss": 0.6727, + "step": 3894 + }, + { + "epoch": 1.2464, + "grad_norm": 0.17034435272216797, + "learning_rate": 4.5608247422680415e-05, + "loss": 0.5757, + "step": 3895 + }, + { + "epoch": 1.24672, + "grad_norm": 0.16528472304344177, + "learning_rate": 4.556701030927835e-05, + "loss": 0.5643, + "step": 3896 + }, + { + "epoch": 1.24704, + "grad_norm": 0.16062431037425995, + "learning_rate": 4.552577319587629e-05, + "loss": 0.7354, + "step": 3897 + }, + { + "epoch": 1.24736, + "grad_norm": 0.18101076781749725, + "learning_rate": 4.548453608247423e-05, + "loss": 0.6676, + "step": 3898 + }, + { + "epoch": 1.24768, + "grad_norm": 0.17771029472351074, + "learning_rate": 4.544329896907217e-05, + "loss": 0.6374, + "step": 3899 + }, + { + "epoch": 1.248, + "grad_norm": 0.1986619383096695, + "learning_rate": 4.5402061855670105e-05, + "loss": 0.7025, + "step": 3900 + }, + { + "epoch": 1.24832, + "grad_norm": 0.16548004746437073, + "learning_rate": 4.536082474226804e-05, + "loss": 0.7711, + "step": 3901 + }, + { + "epoch": 1.24864, + "grad_norm": 0.1644992083311081, + "learning_rate": 4.531958762886598e-05, + "loss": 0.6983, + "step": 3902 + }, + { + "epoch": 1.24896, + "grad_norm": 0.17320775985717773, + "learning_rate": 4.527835051546392e-05, + "loss": 0.4937, + "step": 3903 + }, + { + "epoch": 1.24928, + "grad_norm": 0.16013507544994354, + "learning_rate": 4.523711340206186e-05, + "loss": 0.6342, + "step": 3904 + }, + { + "epoch": 1.2496, + "grad_norm": 0.191853329539299, + "learning_rate": 4.5195876288659795e-05, + "loss": 0.5984, + "step": 3905 + }, + { + "epoch": 1.24992, + "grad_norm": 0.16541354358196259, + "learning_rate": 4.5154639175257733e-05, + "loss": 0.6025, + "step": 3906 + }, + { + "epoch": 1.25024, + "grad_norm": 0.1679941564798355, + "learning_rate": 4.511340206185567e-05, + "loss": 0.4835, + "step": 3907 + }, + { + "epoch": 1.2505600000000001, + "grad_norm": 0.18813015520572662, + "learning_rate": 4.507216494845361e-05, + "loss": 0.5149, + "step": 3908 + }, + { + "epoch": 1.25088, + "grad_norm": 0.17049118876457214, + "learning_rate": 4.503092783505155e-05, + "loss": 0.6463, + "step": 3909 + }, + { + "epoch": 1.2511999999999999, + "grad_norm": 0.1573217660188675, + "learning_rate": 4.4989690721649486e-05, + "loss": 0.5762, + "step": 3910 + }, + { + "epoch": 1.25152, + "grad_norm": 0.16096588969230652, + "learning_rate": 4.4948453608247424e-05, + "loss": 0.4802, + "step": 3911 + }, + { + "epoch": 1.25184, + "grad_norm": 0.17451870441436768, + "learning_rate": 4.490721649484536e-05, + "loss": 0.7303, + "step": 3912 + }, + { + "epoch": 1.25216, + "grad_norm": 0.16196846961975098, + "learning_rate": 4.48659793814433e-05, + "loss": 0.5547, + "step": 3913 + }, + { + "epoch": 1.25248, + "grad_norm": 0.17281825840473175, + "learning_rate": 4.482474226804124e-05, + "loss": 0.6186, + "step": 3914 + }, + { + "epoch": 1.2528000000000001, + "grad_norm": 0.18874908983707428, + "learning_rate": 4.4783505154639176e-05, + "loss": 0.6238, + "step": 3915 + }, + { + "epoch": 1.25312, + "grad_norm": 0.17231100797653198, + "learning_rate": 4.474226804123712e-05, + "loss": 0.5922, + "step": 3916 + }, + { + "epoch": 1.2534399999999999, + "grad_norm": 0.17551463842391968, + "learning_rate": 4.470103092783505e-05, + "loss": 0.5875, + "step": 3917 + }, + { + "epoch": 1.25376, + "grad_norm": 0.1654803454875946, + "learning_rate": 4.465979381443299e-05, + "loss": 0.5791, + "step": 3918 + }, + { + "epoch": 1.25408, + "grad_norm": 0.16714084148406982, + "learning_rate": 4.461855670103093e-05, + "loss": 0.6406, + "step": 3919 + }, + { + "epoch": 1.2544, + "grad_norm": 0.17218844592571259, + "learning_rate": 4.4577319587628866e-05, + "loss": 0.6911, + "step": 3920 + }, + { + "epoch": 1.25472, + "grad_norm": 0.16492901742458344, + "learning_rate": 4.453608247422681e-05, + "loss": 0.5404, + "step": 3921 + }, + { + "epoch": 1.25504, + "grad_norm": 0.17982691526412964, + "learning_rate": 4.449484536082474e-05, + "loss": 0.6456, + "step": 3922 + }, + { + "epoch": 1.25536, + "grad_norm": 0.17917500436306, + "learning_rate": 4.445360824742269e-05, + "loss": 0.6242, + "step": 3923 + }, + { + "epoch": 1.25568, + "grad_norm": 0.15194085240364075, + "learning_rate": 4.441237113402062e-05, + "loss": 0.7029, + "step": 3924 + }, + { + "epoch": 1.256, + "grad_norm": 0.18785709142684937, + "learning_rate": 4.437113402061856e-05, + "loss": 0.5725, + "step": 3925 + }, + { + "epoch": 1.25632, + "grad_norm": 0.2058616578578949, + "learning_rate": 4.4329896907216494e-05, + "loss": 0.7146, + "step": 3926 + }, + { + "epoch": 1.25664, + "grad_norm": 0.1847427785396576, + "learning_rate": 4.428865979381443e-05, + "loss": 0.6479, + "step": 3927 + }, + { + "epoch": 1.25696, + "grad_norm": 0.1676405817270279, + "learning_rate": 4.424742268041238e-05, + "loss": 0.5956, + "step": 3928 + }, + { + "epoch": 1.25728, + "grad_norm": 0.19056977331638336, + "learning_rate": 4.420618556701031e-05, + "loss": 0.5944, + "step": 3929 + }, + { + "epoch": 1.2576, + "grad_norm": 0.1578793227672577, + "learning_rate": 4.4164948453608253e-05, + "loss": 0.526, + "step": 3930 + }, + { + "epoch": 1.25792, + "grad_norm": 0.16699771583080292, + "learning_rate": 4.4123711340206185e-05, + "loss": 0.5595, + "step": 3931 + }, + { + "epoch": 1.25824, + "grad_norm": 0.16113623976707458, + "learning_rate": 4.408247422680413e-05, + "loss": 0.6801, + "step": 3932 + }, + { + "epoch": 1.2585600000000001, + "grad_norm": 0.1931246668100357, + "learning_rate": 4.404123711340206e-05, + "loss": 0.5999, + "step": 3933 + }, + { + "epoch": 1.25888, + "grad_norm": 0.15342223644256592, + "learning_rate": 4.4000000000000006e-05, + "loss": 0.5284, + "step": 3934 + }, + { + "epoch": 1.2591999999999999, + "grad_norm": 0.18424861133098602, + "learning_rate": 4.3958762886597944e-05, + "loss": 0.6125, + "step": 3935 + }, + { + "epoch": 1.25952, + "grad_norm": 0.1978152096271515, + "learning_rate": 4.3917525773195875e-05, + "loss": 0.7088, + "step": 3936 + }, + { + "epoch": 1.25984, + "grad_norm": 0.17900687456130981, + "learning_rate": 4.387628865979382e-05, + "loss": 0.5347, + "step": 3937 + }, + { + "epoch": 1.26016, + "grad_norm": 0.20973551273345947, + "learning_rate": 4.383505154639175e-05, + "loss": 0.7015, + "step": 3938 + }, + { + "epoch": 1.26048, + "grad_norm": 0.19654051959514618, + "learning_rate": 4.3793814432989696e-05, + "loss": 0.6467, + "step": 3939 + }, + { + "epoch": 1.2608, + "grad_norm": 0.17594395577907562, + "learning_rate": 4.375257731958763e-05, + "loss": 0.6664, + "step": 3940 + }, + { + "epoch": 1.26112, + "grad_norm": 0.153744637966156, + "learning_rate": 4.371134020618557e-05, + "loss": 0.6435, + "step": 3941 + }, + { + "epoch": 1.26144, + "grad_norm": 0.16148659586906433, + "learning_rate": 4.367010309278351e-05, + "loss": 0.5717, + "step": 3942 + }, + { + "epoch": 1.26176, + "grad_norm": 0.19449885189533234, + "learning_rate": 4.362886597938145e-05, + "loss": 0.6007, + "step": 3943 + }, + { + "epoch": 1.26208, + "grad_norm": 0.16126826405525208, + "learning_rate": 4.3587628865979386e-05, + "loss": 0.6586, + "step": 3944 + }, + { + "epoch": 1.2624, + "grad_norm": 0.17294864356517792, + "learning_rate": 4.354639175257732e-05, + "loss": 0.6349, + "step": 3945 + }, + { + "epoch": 1.26272, + "grad_norm": 0.18492111563682556, + "learning_rate": 4.350515463917526e-05, + "loss": 0.5572, + "step": 3946 + }, + { + "epoch": 1.26304, + "grad_norm": 0.1736954152584076, + "learning_rate": 4.3463917525773194e-05, + "loss": 0.5119, + "step": 3947 + }, + { + "epoch": 1.26336, + "grad_norm": 0.18799877166748047, + "learning_rate": 4.342268041237114e-05, + "loss": 0.6575, + "step": 3948 + }, + { + "epoch": 1.26368, + "grad_norm": 0.15451136231422424, + "learning_rate": 4.3381443298969076e-05, + "loss": 0.6687, + "step": 3949 + }, + { + "epoch": 1.264, + "grad_norm": 0.17211328446865082, + "learning_rate": 4.3340206185567014e-05, + "loss": 0.6392, + "step": 3950 + }, + { + "epoch": 1.26432, + "grad_norm": 0.1470901221036911, + "learning_rate": 4.329896907216495e-05, + "loss": 0.5624, + "step": 3951 + }, + { + "epoch": 1.26464, + "grad_norm": 0.1566469520330429, + "learning_rate": 4.3257731958762884e-05, + "loss": 0.5272, + "step": 3952 + }, + { + "epoch": 1.2649599999999999, + "grad_norm": 0.18103241920471191, + "learning_rate": 4.321649484536083e-05, + "loss": 0.6354, + "step": 3953 + }, + { + "epoch": 1.26528, + "grad_norm": 0.2093370109796524, + "learning_rate": 4.317525773195876e-05, + "loss": 0.5834, + "step": 3954 + }, + { + "epoch": 1.2656, + "grad_norm": 0.19479696452617645, + "learning_rate": 4.3134020618556705e-05, + "loss": 0.6102, + "step": 3955 + }, + { + "epoch": 1.26592, + "grad_norm": 0.1535751074552536, + "learning_rate": 4.309278350515464e-05, + "loss": 0.5103, + "step": 3956 + }, + { + "epoch": 1.26624, + "grad_norm": 0.1928999274969101, + "learning_rate": 4.305154639175258e-05, + "loss": 0.6294, + "step": 3957 + }, + { + "epoch": 1.2665600000000001, + "grad_norm": 0.17962078750133514, + "learning_rate": 4.301030927835052e-05, + "loss": 0.4886, + "step": 3958 + }, + { + "epoch": 1.26688, + "grad_norm": 0.1581602394580841, + "learning_rate": 4.296907216494846e-05, + "loss": 0.5038, + "step": 3959 + }, + { + "epoch": 1.2671999999999999, + "grad_norm": 0.17142514884471893, + "learning_rate": 4.2927835051546395e-05, + "loss": 0.7397, + "step": 3960 + }, + { + "epoch": 1.26752, + "grad_norm": 0.18619868159294128, + "learning_rate": 4.2886597938144326e-05, + "loss": 0.6841, + "step": 3961 + }, + { + "epoch": 1.26784, + "grad_norm": 0.18114466965198517, + "learning_rate": 4.284536082474227e-05, + "loss": 0.5635, + "step": 3962 + }, + { + "epoch": 1.26816, + "grad_norm": 0.17204061150550842, + "learning_rate": 4.280412371134021e-05, + "loss": 0.5523, + "step": 3963 + }, + { + "epoch": 1.26848, + "grad_norm": 0.16677473485469818, + "learning_rate": 4.276288659793815e-05, + "loss": 0.4777, + "step": 3964 + }, + { + "epoch": 1.2688, + "grad_norm": 0.14719942212104797, + "learning_rate": 4.2721649484536085e-05, + "loss": 0.6854, + "step": 3965 + }, + { + "epoch": 1.26912, + "grad_norm": 0.17097407579421997, + "learning_rate": 4.268041237113402e-05, + "loss": 0.7022, + "step": 3966 + }, + { + "epoch": 1.26944, + "grad_norm": 0.16300159692764282, + "learning_rate": 4.263917525773196e-05, + "loss": 0.5867, + "step": 3967 + }, + { + "epoch": 1.26976, + "grad_norm": 0.13272792100906372, + "learning_rate": 4.25979381443299e-05, + "loss": 0.5664, + "step": 3968 + }, + { + "epoch": 1.27008, + "grad_norm": 0.17930343747138977, + "learning_rate": 4.255670103092784e-05, + "loss": 0.6826, + "step": 3969 + }, + { + "epoch": 1.2704, + "grad_norm": 0.1556936651468277, + "learning_rate": 4.2515463917525775e-05, + "loss": 0.5835, + "step": 3970 + }, + { + "epoch": 1.27072, + "grad_norm": 0.16930441558361053, + "learning_rate": 4.2474226804123713e-05, + "loss": 0.6278, + "step": 3971 + }, + { + "epoch": 1.27104, + "grad_norm": 0.16733677685260773, + "learning_rate": 4.243298969072165e-05, + "loss": 0.5467, + "step": 3972 + }, + { + "epoch": 1.27136, + "grad_norm": 0.19166122376918793, + "learning_rate": 4.239175257731959e-05, + "loss": 0.5816, + "step": 3973 + }, + { + "epoch": 1.27168, + "grad_norm": 0.14965656399726868, + "learning_rate": 4.235051546391753e-05, + "loss": 0.5519, + "step": 3974 + }, + { + "epoch": 1.272, + "grad_norm": 0.1614885777235031, + "learning_rate": 4.2309278350515466e-05, + "loss": 0.6711, + "step": 3975 + }, + { + "epoch": 1.2723200000000001, + "grad_norm": 0.19913813471794128, + "learning_rate": 4.2268041237113404e-05, + "loss": 0.7513, + "step": 3976 + }, + { + "epoch": 1.27264, + "grad_norm": 0.1941332072019577, + "learning_rate": 4.222680412371134e-05, + "loss": 0.6783, + "step": 3977 + }, + { + "epoch": 1.2729599999999999, + "grad_norm": 0.16733598709106445, + "learning_rate": 4.218556701030928e-05, + "loss": 0.5604, + "step": 3978 + }, + { + "epoch": 1.27328, + "grad_norm": 0.17445360124111176, + "learning_rate": 4.214432989690722e-05, + "loss": 0.7288, + "step": 3979 + }, + { + "epoch": 1.2736, + "grad_norm": 0.16022154688835144, + "learning_rate": 4.2103092783505156e-05, + "loss": 0.549, + "step": 3980 + }, + { + "epoch": 1.27392, + "grad_norm": 0.17464672029018402, + "learning_rate": 4.2061855670103094e-05, + "loss": 0.6134, + "step": 3981 + }, + { + "epoch": 1.27424, + "grad_norm": 0.16548338532447815, + "learning_rate": 4.202061855670103e-05, + "loss": 0.6052, + "step": 3982 + }, + { + "epoch": 1.2745600000000001, + "grad_norm": 0.13781480491161346, + "learning_rate": 4.197938144329897e-05, + "loss": 0.4117, + "step": 3983 + }, + { + "epoch": 1.27488, + "grad_norm": 0.16291040182113647, + "learning_rate": 4.193814432989691e-05, + "loss": 0.5467, + "step": 3984 + }, + { + "epoch": 1.2752, + "grad_norm": 0.19372756779193878, + "learning_rate": 4.1896907216494846e-05, + "loss": 0.6466, + "step": 3985 + }, + { + "epoch": 1.27552, + "grad_norm": 0.15742923319339752, + "learning_rate": 4.1855670103092784e-05, + "loss": 0.5923, + "step": 3986 + }, + { + "epoch": 1.27584, + "grad_norm": 0.17011013627052307, + "learning_rate": 4.181443298969072e-05, + "loss": 0.7429, + "step": 3987 + }, + { + "epoch": 1.27616, + "grad_norm": 0.16942501068115234, + "learning_rate": 4.177319587628866e-05, + "loss": 0.6318, + "step": 3988 + }, + { + "epoch": 1.27648, + "grad_norm": 0.16076137125492096, + "learning_rate": 4.1731958762886605e-05, + "loss": 0.6837, + "step": 3989 + }, + { + "epoch": 1.2768, + "grad_norm": 0.16504047811031342, + "learning_rate": 4.1690721649484536e-05, + "loss": 0.5856, + "step": 3990 + }, + { + "epoch": 1.27712, + "grad_norm": 0.18571355938911438, + "learning_rate": 4.164948453608248e-05, + "loss": 0.7246, + "step": 3991 + }, + { + "epoch": 1.27744, + "grad_norm": 0.17110587656497955, + "learning_rate": 4.160824742268041e-05, + "loss": 0.5775, + "step": 3992 + }, + { + "epoch": 1.27776, + "grad_norm": 0.16663064062595367, + "learning_rate": 4.156701030927835e-05, + "loss": 0.5566, + "step": 3993 + }, + { + "epoch": 1.27808, + "grad_norm": 0.16578438878059387, + "learning_rate": 4.152577319587629e-05, + "loss": 0.6839, + "step": 3994 + }, + { + "epoch": 1.2784, + "grad_norm": 0.156260684132576, + "learning_rate": 4.148453608247423e-05, + "loss": 0.4791, + "step": 3995 + }, + { + "epoch": 1.27872, + "grad_norm": 0.16070456802845, + "learning_rate": 4.144329896907217e-05, + "loss": 0.4916, + "step": 3996 + }, + { + "epoch": 1.27904, + "grad_norm": 0.17698252201080322, + "learning_rate": 4.14020618556701e-05, + "loss": 0.692, + "step": 3997 + }, + { + "epoch": 1.27936, + "grad_norm": 0.16911941766738892, + "learning_rate": 4.136082474226805e-05, + "loss": 0.5837, + "step": 3998 + }, + { + "epoch": 1.27968, + "grad_norm": 0.16667480766773224, + "learning_rate": 4.131958762886598e-05, + "loss": 0.6516, + "step": 3999 + }, + { + "epoch": 1.28, + "grad_norm": 0.17482155561447144, + "learning_rate": 4.1278350515463924e-05, + "loss": 0.648, + "step": 4000 + }, + { + "epoch": 1.2803200000000001, + "grad_norm": 0.16110439598560333, + "learning_rate": 4.1237113402061855e-05, + "loss": 0.6567, + "step": 4001 + }, + { + "epoch": 1.28064, + "grad_norm": 0.17544062435626984, + "learning_rate": 4.119587628865979e-05, + "loss": 0.6747, + "step": 4002 + }, + { + "epoch": 1.2809599999999999, + "grad_norm": 0.16116860508918762, + "learning_rate": 4.115463917525774e-05, + "loss": 0.6163, + "step": 4003 + }, + { + "epoch": 1.28128, + "grad_norm": 0.18779298663139343, + "learning_rate": 4.111340206185567e-05, + "loss": 0.5204, + "step": 4004 + }, + { + "epoch": 1.2816, + "grad_norm": 0.1786164939403534, + "learning_rate": 4.1072164948453614e-05, + "loss": 0.5287, + "step": 4005 + }, + { + "epoch": 1.28192, + "grad_norm": 0.16390131413936615, + "learning_rate": 4.1030927835051545e-05, + "loss": 0.5572, + "step": 4006 + }, + { + "epoch": 1.28224, + "grad_norm": 0.15653590857982635, + "learning_rate": 4.098969072164949e-05, + "loss": 0.6765, + "step": 4007 + }, + { + "epoch": 1.28256, + "grad_norm": 0.15331792831420898, + "learning_rate": 4.094845360824742e-05, + "loss": 0.5962, + "step": 4008 + }, + { + "epoch": 1.28288, + "grad_norm": 0.16848453879356384, + "learning_rate": 4.0907216494845366e-05, + "loss": 0.6766, + "step": 4009 + }, + { + "epoch": 1.2832, + "grad_norm": 0.19180482625961304, + "learning_rate": 4.0865979381443304e-05, + "loss": 0.587, + "step": 4010 + }, + { + "epoch": 1.28352, + "grad_norm": 0.1502186805009842, + "learning_rate": 4.0824742268041235e-05, + "loss": 0.5117, + "step": 4011 + }, + { + "epoch": 1.28384, + "grad_norm": 0.17697405815124512, + "learning_rate": 4.078350515463918e-05, + "loss": 0.6874, + "step": 4012 + }, + { + "epoch": 1.28416, + "grad_norm": 0.19886483252048492, + "learning_rate": 4.074226804123711e-05, + "loss": 0.7485, + "step": 4013 + }, + { + "epoch": 1.28448, + "grad_norm": 0.14835217595100403, + "learning_rate": 4.0701030927835056e-05, + "loss": 0.6443, + "step": 4014 + }, + { + "epoch": 1.2848, + "grad_norm": 0.15456253290176392, + "learning_rate": 4.065979381443299e-05, + "loss": 0.6111, + "step": 4015 + }, + { + "epoch": 1.28512, + "grad_norm": 0.1409899741411209, + "learning_rate": 4.061855670103093e-05, + "loss": 0.5719, + "step": 4016 + }, + { + "epoch": 1.28544, + "grad_norm": 0.19460925459861755, + "learning_rate": 4.057731958762887e-05, + "loss": 0.6555, + "step": 4017 + }, + { + "epoch": 1.28576, + "grad_norm": 0.18679526448249817, + "learning_rate": 4.053608247422681e-05, + "loss": 0.5511, + "step": 4018 + }, + { + "epoch": 1.2860800000000001, + "grad_norm": 0.1671118289232254, + "learning_rate": 4.049484536082475e-05, + "loss": 0.6662, + "step": 4019 + }, + { + "epoch": 1.2864, + "grad_norm": 0.1618160456418991, + "learning_rate": 4.045360824742268e-05, + "loss": 0.6684, + "step": 4020 + }, + { + "epoch": 1.2867199999999999, + "grad_norm": 0.17376519739627838, + "learning_rate": 4.041237113402062e-05, + "loss": 0.57, + "step": 4021 + }, + { + "epoch": 1.28704, + "grad_norm": 0.1481582373380661, + "learning_rate": 4.0371134020618554e-05, + "loss": 0.5645, + "step": 4022 + }, + { + "epoch": 1.28736, + "grad_norm": 0.1912136673927307, + "learning_rate": 4.03298969072165e-05, + "loss": 0.6852, + "step": 4023 + }, + { + "epoch": 1.28768, + "grad_norm": 0.1885708123445511, + "learning_rate": 4.028865979381444e-05, + "loss": 0.663, + "step": 4024 + }, + { + "epoch": 1.288, + "grad_norm": 0.18485084176063538, + "learning_rate": 4.0247422680412375e-05, + "loss": 0.6202, + "step": 4025 + }, + { + "epoch": 1.2883200000000001, + "grad_norm": 0.15674202144145966, + "learning_rate": 4.020618556701031e-05, + "loss": 0.483, + "step": 4026 + }, + { + "epoch": 1.28864, + "grad_norm": 0.17164091765880585, + "learning_rate": 4.0164948453608244e-05, + "loss": 0.6123, + "step": 4027 + }, + { + "epoch": 1.2889599999999999, + "grad_norm": 0.16092656552791595, + "learning_rate": 4.012371134020619e-05, + "loss": 0.6212, + "step": 4028 + }, + { + "epoch": 1.28928, + "grad_norm": 0.16720622777938843, + "learning_rate": 4.008247422680412e-05, + "loss": 0.6533, + "step": 4029 + }, + { + "epoch": 1.2896, + "grad_norm": 0.18945324420928955, + "learning_rate": 4.0041237113402065e-05, + "loss": 0.608, + "step": 4030 + }, + { + "epoch": 1.28992, + "grad_norm": 0.18045999109745026, + "learning_rate": 4e-05, + "loss": 0.5521, + "step": 4031 + }, + { + "epoch": 1.29024, + "grad_norm": 0.17056629061698914, + "learning_rate": 3.995876288659794e-05, + "loss": 0.5367, + "step": 4032 + }, + { + "epoch": 1.29056, + "grad_norm": 0.15712274610996246, + "learning_rate": 3.991752577319588e-05, + "loss": 0.6284, + "step": 4033 + }, + { + "epoch": 1.29088, + "grad_norm": 0.19665701687335968, + "learning_rate": 3.987628865979382e-05, + "loss": 0.615, + "step": 4034 + }, + { + "epoch": 1.2912, + "grad_norm": 0.17210933566093445, + "learning_rate": 3.9835051546391755e-05, + "loss": 0.5992, + "step": 4035 + }, + { + "epoch": 1.29152, + "grad_norm": 0.20588752627372742, + "learning_rate": 3.979381443298969e-05, + "loss": 0.7485, + "step": 4036 + }, + { + "epoch": 1.29184, + "grad_norm": 0.1646113097667694, + "learning_rate": 3.975257731958763e-05, + "loss": 0.6168, + "step": 4037 + }, + { + "epoch": 1.29216, + "grad_norm": 0.16620448231697083, + "learning_rate": 3.971134020618557e-05, + "loss": 0.6622, + "step": 4038 + }, + { + "epoch": 1.29248, + "grad_norm": 0.163372203707695, + "learning_rate": 3.967010309278351e-05, + "loss": 0.5766, + "step": 4039 + }, + { + "epoch": 1.2928, + "grad_norm": 0.16962704062461853, + "learning_rate": 3.9628865979381446e-05, + "loss": 0.5082, + "step": 4040 + }, + { + "epoch": 1.29312, + "grad_norm": 0.1489821970462799, + "learning_rate": 3.9587628865979384e-05, + "loss": 0.6058, + "step": 4041 + }, + { + "epoch": 1.29344, + "grad_norm": 0.1639564335346222, + "learning_rate": 3.954639175257732e-05, + "loss": 0.5802, + "step": 4042 + }, + { + "epoch": 1.29376, + "grad_norm": 0.20464909076690674, + "learning_rate": 3.950515463917526e-05, + "loss": 0.6019, + "step": 4043 + }, + { + "epoch": 1.2940800000000001, + "grad_norm": 0.16458465158939362, + "learning_rate": 3.94639175257732e-05, + "loss": 0.5881, + "step": 4044 + }, + { + "epoch": 1.2944, + "grad_norm": 0.1728154718875885, + "learning_rate": 3.9422680412371136e-05, + "loss": 0.6521, + "step": 4045 + }, + { + "epoch": 1.2947199999999999, + "grad_norm": 0.16128352284431458, + "learning_rate": 3.9381443298969074e-05, + "loss": 0.5295, + "step": 4046 + }, + { + "epoch": 1.29504, + "grad_norm": 0.1698767989873886, + "learning_rate": 3.934020618556701e-05, + "loss": 0.5252, + "step": 4047 + }, + { + "epoch": 1.29536, + "grad_norm": 0.1480582058429718, + "learning_rate": 3.929896907216495e-05, + "loss": 0.7894, + "step": 4048 + }, + { + "epoch": 1.29568, + "grad_norm": 0.14291821420192719, + "learning_rate": 3.925773195876289e-05, + "loss": 0.6046, + "step": 4049 + }, + { + "epoch": 1.296, + "grad_norm": 0.1739010065793991, + "learning_rate": 3.9216494845360826e-05, + "loss": 0.662, + "step": 4050 + }, + { + "epoch": 1.29632, + "grad_norm": 0.17592869699001312, + "learning_rate": 3.9175257731958764e-05, + "loss": 0.5889, + "step": 4051 + }, + { + "epoch": 1.29664, + "grad_norm": 0.19233426451683044, + "learning_rate": 3.91340206185567e-05, + "loss": 0.8656, + "step": 4052 + }, + { + "epoch": 1.29696, + "grad_norm": 0.19060637056827545, + "learning_rate": 3.909278350515464e-05, + "loss": 0.734, + "step": 4053 + }, + { + "epoch": 1.29728, + "grad_norm": 0.17077085375785828, + "learning_rate": 3.905154639175258e-05, + "loss": 0.4903, + "step": 4054 + }, + { + "epoch": 1.2976, + "grad_norm": 0.16599635779857635, + "learning_rate": 3.9010309278350516e-05, + "loss": 0.7009, + "step": 4055 + }, + { + "epoch": 1.29792, + "grad_norm": 0.1570754051208496, + "learning_rate": 3.8969072164948455e-05, + "loss": 0.4946, + "step": 4056 + }, + { + "epoch": 1.29824, + "grad_norm": 0.1928950846195221, + "learning_rate": 3.89278350515464e-05, + "loss": 0.6717, + "step": 4057 + }, + { + "epoch": 1.29856, + "grad_norm": 0.1669059842824936, + "learning_rate": 3.888659793814433e-05, + "loss": 0.6465, + "step": 4058 + }, + { + "epoch": 1.29888, + "grad_norm": 0.14909622073173523, + "learning_rate": 3.884536082474227e-05, + "loss": 0.4864, + "step": 4059 + }, + { + "epoch": 1.2992, + "grad_norm": 0.18066217005252838, + "learning_rate": 3.880412371134021e-05, + "loss": 0.7621, + "step": 4060 + }, + { + "epoch": 1.29952, + "grad_norm": 0.1689022183418274, + "learning_rate": 3.8762886597938145e-05, + "loss": 0.5502, + "step": 4061 + }, + { + "epoch": 1.29984, + "grad_norm": 0.1850537806749344, + "learning_rate": 3.872164948453608e-05, + "loss": 0.6278, + "step": 4062 + }, + { + "epoch": 1.30016, + "grad_norm": 0.15273945033550262, + "learning_rate": 3.868041237113402e-05, + "loss": 0.6979, + "step": 4063 + }, + { + "epoch": 1.30048, + "grad_norm": 0.18240220844745636, + "learning_rate": 3.8639175257731966e-05, + "loss": 0.6054, + "step": 4064 + }, + { + "epoch": 1.3008, + "grad_norm": 0.177093043923378, + "learning_rate": 3.85979381443299e-05, + "loss": 0.4943, + "step": 4065 + }, + { + "epoch": 1.30112, + "grad_norm": 0.1645955741405487, + "learning_rate": 3.855670103092784e-05, + "loss": 0.5826, + "step": 4066 + }, + { + "epoch": 1.30144, + "grad_norm": 0.1600593775510788, + "learning_rate": 3.851546391752577e-05, + "loss": 0.6461, + "step": 4067 + }, + { + "epoch": 1.30176, + "grad_norm": 0.19090795516967773, + "learning_rate": 3.847422680412371e-05, + "loss": 0.6258, + "step": 4068 + }, + { + "epoch": 1.3020800000000001, + "grad_norm": 0.1854894459247589, + "learning_rate": 3.843298969072165e-05, + "loss": 0.6582, + "step": 4069 + }, + { + "epoch": 1.3024, + "grad_norm": 0.19072747230529785, + "learning_rate": 3.839175257731959e-05, + "loss": 0.6467, + "step": 4070 + }, + { + "epoch": 1.3027199999999999, + "grad_norm": 0.1792403906583786, + "learning_rate": 3.835051546391753e-05, + "loss": 0.6383, + "step": 4071 + }, + { + "epoch": 1.30304, + "grad_norm": 0.17212651669979095, + "learning_rate": 3.830927835051546e-05, + "loss": 0.5966, + "step": 4072 + }, + { + "epoch": 1.30336, + "grad_norm": 0.1508508026599884, + "learning_rate": 3.826804123711341e-05, + "loss": 0.5523, + "step": 4073 + }, + { + "epoch": 1.30368, + "grad_norm": 0.18643657863140106, + "learning_rate": 3.822680412371134e-05, + "loss": 0.6645, + "step": 4074 + }, + { + "epoch": 1.304, + "grad_norm": 0.16659589111804962, + "learning_rate": 3.8185567010309284e-05, + "loss": 0.5956, + "step": 4075 + }, + { + "epoch": 1.30432, + "grad_norm": 0.2031671106815338, + "learning_rate": 3.8144329896907216e-05, + "loss": 0.6493, + "step": 4076 + }, + { + "epoch": 1.30464, + "grad_norm": 0.16055773198604584, + "learning_rate": 3.8103092783505154e-05, + "loss": 0.7892, + "step": 4077 + }, + { + "epoch": 1.30496, + "grad_norm": 0.17206372320652008, + "learning_rate": 3.80618556701031e-05, + "loss": 0.6848, + "step": 4078 + }, + { + "epoch": 1.30528, + "grad_norm": 0.18009749054908752, + "learning_rate": 3.802061855670103e-05, + "loss": 0.594, + "step": 4079 + }, + { + "epoch": 1.3056, + "grad_norm": 0.18778519332408905, + "learning_rate": 3.7979381443298975e-05, + "loss": 0.5568, + "step": 4080 + }, + { + "epoch": 1.30592, + "grad_norm": 0.1657392680644989, + "learning_rate": 3.7938144329896906e-05, + "loss": 0.6497, + "step": 4081 + }, + { + "epoch": 1.30624, + "grad_norm": 0.17428772151470184, + "learning_rate": 3.789690721649485e-05, + "loss": 0.6514, + "step": 4082 + }, + { + "epoch": 1.30656, + "grad_norm": 0.17245791852474213, + "learning_rate": 3.785567010309278e-05, + "loss": 0.5973, + "step": 4083 + }, + { + "epoch": 1.30688, + "grad_norm": 0.15220580995082855, + "learning_rate": 3.781443298969073e-05, + "loss": 0.5808, + "step": 4084 + }, + { + "epoch": 1.3072, + "grad_norm": 0.18960003554821014, + "learning_rate": 3.7773195876288665e-05, + "loss": 0.6281, + "step": 4085 + }, + { + "epoch": 1.30752, + "grad_norm": 0.17941728234291077, + "learning_rate": 3.7731958762886596e-05, + "loss": 0.8363, + "step": 4086 + }, + { + "epoch": 1.3078400000000001, + "grad_norm": 0.159403458237648, + "learning_rate": 3.769072164948454e-05, + "loss": 0.6123, + "step": 4087 + }, + { + "epoch": 1.30816, + "grad_norm": 0.16201511025428772, + "learning_rate": 3.764948453608247e-05, + "loss": 0.5588, + "step": 4088 + }, + { + "epoch": 1.3084799999999999, + "grad_norm": 0.18773354589939117, + "learning_rate": 3.760824742268042e-05, + "loss": 0.5752, + "step": 4089 + }, + { + "epoch": 1.3088, + "grad_norm": 0.1755516529083252, + "learning_rate": 3.756701030927835e-05, + "loss": 0.673, + "step": 4090 + }, + { + "epoch": 1.30912, + "grad_norm": 0.1829020231962204, + "learning_rate": 3.752577319587629e-05, + "loss": 0.7573, + "step": 4091 + }, + { + "epoch": 1.30944, + "grad_norm": 0.16617918014526367, + "learning_rate": 3.748453608247423e-05, + "loss": 0.7944, + "step": 4092 + }, + { + "epoch": 1.30976, + "grad_norm": 0.1652265042066574, + "learning_rate": 3.744329896907217e-05, + "loss": 0.5177, + "step": 4093 + }, + { + "epoch": 1.3100800000000001, + "grad_norm": 0.19705916941165924, + "learning_rate": 3.740206185567011e-05, + "loss": 0.5876, + "step": 4094 + }, + { + "epoch": 1.3104, + "grad_norm": 0.1578940451145172, + "learning_rate": 3.736082474226804e-05, + "loss": 0.5168, + "step": 4095 + }, + { + "epoch": 1.3107199999999999, + "grad_norm": 0.1926892101764679, + "learning_rate": 3.731958762886598e-05, + "loss": 0.62, + "step": 4096 + }, + { + "epoch": 1.31104, + "grad_norm": 0.19476854801177979, + "learning_rate": 3.7278350515463915e-05, + "loss": 0.74, + "step": 4097 + }, + { + "epoch": 1.31136, + "grad_norm": 0.1641494333744049, + "learning_rate": 3.723711340206186e-05, + "loss": 0.4658, + "step": 4098 + }, + { + "epoch": 1.31168, + "grad_norm": 0.18733379244804382, + "learning_rate": 3.71958762886598e-05, + "loss": 0.5664, + "step": 4099 + }, + { + "epoch": 1.312, + "grad_norm": 0.19119028747081757, + "learning_rate": 3.7154639175257736e-05, + "loss": 0.6076, + "step": 4100 + }, + { + "epoch": 1.31232, + "grad_norm": 0.16721762716770172, + "learning_rate": 3.7113402061855674e-05, + "loss": 0.528, + "step": 4101 + }, + { + "epoch": 1.31264, + "grad_norm": 0.17330771684646606, + "learning_rate": 3.7072164948453605e-05, + "loss": 0.6279, + "step": 4102 + }, + { + "epoch": 1.31296, + "grad_norm": 0.18464361131191254, + "learning_rate": 3.703092783505155e-05, + "loss": 0.63, + "step": 4103 + }, + { + "epoch": 1.31328, + "grad_norm": 0.17970842123031616, + "learning_rate": 3.698969072164948e-05, + "loss": 0.5991, + "step": 4104 + }, + { + "epoch": 1.3136, + "grad_norm": 0.18501725792884827, + "learning_rate": 3.6948453608247426e-05, + "loss": 0.5466, + "step": 4105 + }, + { + "epoch": 1.31392, + "grad_norm": 0.1707962155342102, + "learning_rate": 3.6907216494845364e-05, + "loss": 0.4654, + "step": 4106 + }, + { + "epoch": 1.31424, + "grad_norm": 0.17342007160186768, + "learning_rate": 3.68659793814433e-05, + "loss": 0.6561, + "step": 4107 + }, + { + "epoch": 1.31456, + "grad_norm": 0.1687413603067398, + "learning_rate": 3.682474226804124e-05, + "loss": 0.5463, + "step": 4108 + }, + { + "epoch": 1.31488, + "grad_norm": 0.17356687784194946, + "learning_rate": 3.678350515463918e-05, + "loss": 0.7898, + "step": 4109 + }, + { + "epoch": 1.3152, + "grad_norm": 0.18300271034240723, + "learning_rate": 3.6742268041237116e-05, + "loss": 0.5643, + "step": 4110 + }, + { + "epoch": 1.31552, + "grad_norm": 0.2000061720609665, + "learning_rate": 3.670103092783505e-05, + "loss": 0.5134, + "step": 4111 + }, + { + "epoch": 1.3158400000000001, + "grad_norm": 0.1677892506122589, + "learning_rate": 3.665979381443299e-05, + "loss": 0.6827, + "step": 4112 + }, + { + "epoch": 1.31616, + "grad_norm": 0.1657765805721283, + "learning_rate": 3.661855670103093e-05, + "loss": 0.6199, + "step": 4113 + }, + { + "epoch": 1.3164799999999999, + "grad_norm": 0.16753371059894562, + "learning_rate": 3.657731958762887e-05, + "loss": 0.6993, + "step": 4114 + }, + { + "epoch": 1.3168, + "grad_norm": 0.17366673052310944, + "learning_rate": 3.6536082474226806e-05, + "loss": 0.747, + "step": 4115 + }, + { + "epoch": 1.31712, + "grad_norm": 0.1795147955417633, + "learning_rate": 3.6494845360824744e-05, + "loss": 0.559, + "step": 4116 + }, + { + "epoch": 1.31744, + "grad_norm": 0.15964820981025696, + "learning_rate": 3.645360824742268e-05, + "loss": 0.5659, + "step": 4117 + }, + { + "epoch": 1.31776, + "grad_norm": 0.15476112067699432, + "learning_rate": 3.641237113402062e-05, + "loss": 0.6143, + "step": 4118 + }, + { + "epoch": 1.31808, + "grad_norm": 0.15909650921821594, + "learning_rate": 3.637113402061856e-05, + "loss": 0.5513, + "step": 4119 + }, + { + "epoch": 1.3184, + "grad_norm": 0.17541798949241638, + "learning_rate": 3.6329896907216497e-05, + "loss": 0.5677, + "step": 4120 + }, + { + "epoch": 1.31872, + "grad_norm": 0.17759791016578674, + "learning_rate": 3.6288659793814435e-05, + "loss": 0.6718, + "step": 4121 + }, + { + "epoch": 1.31904, + "grad_norm": 0.1713830530643463, + "learning_rate": 3.624742268041237e-05, + "loss": 0.4369, + "step": 4122 + }, + { + "epoch": 1.31936, + "grad_norm": 0.16756726801395416, + "learning_rate": 3.620618556701031e-05, + "loss": 0.4215, + "step": 4123 + }, + { + "epoch": 1.31968, + "grad_norm": 0.17106227576732635, + "learning_rate": 3.616494845360825e-05, + "loss": 0.633, + "step": 4124 + }, + { + "epoch": 1.32, + "grad_norm": 0.15157227218151093, + "learning_rate": 3.612371134020619e-05, + "loss": 0.6088, + "step": 4125 + }, + { + "epoch": 1.32032, + "grad_norm": 0.16967083513736725, + "learning_rate": 3.6082474226804125e-05, + "loss": 0.5646, + "step": 4126 + }, + { + "epoch": 1.32064, + "grad_norm": 0.1786714494228363, + "learning_rate": 3.604123711340206e-05, + "loss": 0.7126, + "step": 4127 + }, + { + "epoch": 1.32096, + "grad_norm": 0.16891857981681824, + "learning_rate": 3.6e-05, + "loss": 0.6251, + "step": 4128 + }, + { + "epoch": 1.32128, + "grad_norm": 0.18559736013412476, + "learning_rate": 3.595876288659794e-05, + "loss": 0.6678, + "step": 4129 + }, + { + "epoch": 1.3216, + "grad_norm": 0.15548352897167206, + "learning_rate": 3.591752577319588e-05, + "loss": 0.5615, + "step": 4130 + }, + { + "epoch": 1.32192, + "grad_norm": 0.1877441257238388, + "learning_rate": 3.5876288659793815e-05, + "loss": 0.6453, + "step": 4131 + }, + { + "epoch": 1.32224, + "grad_norm": 0.15813611447811127, + "learning_rate": 3.583505154639176e-05, + "loss": 0.572, + "step": 4132 + }, + { + "epoch": 1.32256, + "grad_norm": 0.17463301122188568, + "learning_rate": 3.579381443298969e-05, + "loss": 0.498, + "step": 4133 + }, + { + "epoch": 1.32288, + "grad_norm": 0.17169231176376343, + "learning_rate": 3.575257731958763e-05, + "loss": 0.5617, + "step": 4134 + }, + { + "epoch": 1.3232, + "grad_norm": 0.1643225997686386, + "learning_rate": 3.571134020618557e-05, + "loss": 0.6245, + "step": 4135 + }, + { + "epoch": 1.32352, + "grad_norm": 0.1765742301940918, + "learning_rate": 3.5670103092783505e-05, + "loss": 0.8405, + "step": 4136 + }, + { + "epoch": 1.3238400000000001, + "grad_norm": 0.14187227189540863, + "learning_rate": 3.562886597938144e-05, + "loss": 0.5079, + "step": 4137 + }, + { + "epoch": 1.32416, + "grad_norm": 0.18652300536632538, + "learning_rate": 3.558762886597938e-05, + "loss": 0.6436, + "step": 4138 + }, + { + "epoch": 1.3244799999999999, + "grad_norm": 0.2035590559244156, + "learning_rate": 3.5546391752577326e-05, + "loss": 0.6157, + "step": 4139 + }, + { + "epoch": 1.3248, + "grad_norm": 0.17886567115783691, + "learning_rate": 3.550515463917526e-05, + "loss": 0.645, + "step": 4140 + }, + { + "epoch": 1.32512, + "grad_norm": 0.1650288999080658, + "learning_rate": 3.54639175257732e-05, + "loss": 0.5001, + "step": 4141 + }, + { + "epoch": 1.32544, + "grad_norm": 0.15829825401306152, + "learning_rate": 3.5422680412371134e-05, + "loss": 0.6108, + "step": 4142 + }, + { + "epoch": 1.32576, + "grad_norm": 0.16972468793392181, + "learning_rate": 3.538144329896907e-05, + "loss": 0.5422, + "step": 4143 + }, + { + "epoch": 1.32608, + "grad_norm": 0.15033535659313202, + "learning_rate": 3.534020618556701e-05, + "loss": 0.5336, + "step": 4144 + }, + { + "epoch": 1.3264, + "grad_norm": 0.1781577169895172, + "learning_rate": 3.529896907216495e-05, + "loss": 0.7749, + "step": 4145 + }, + { + "epoch": 1.32672, + "grad_norm": 0.17294608056545258, + "learning_rate": 3.525773195876289e-05, + "loss": 0.5337, + "step": 4146 + }, + { + "epoch": 1.32704, + "grad_norm": 0.20067951083183289, + "learning_rate": 3.5216494845360824e-05, + "loss": 0.6059, + "step": 4147 + }, + { + "epoch": 1.32736, + "grad_norm": 0.18750374019145966, + "learning_rate": 3.517525773195877e-05, + "loss": 0.6847, + "step": 4148 + }, + { + "epoch": 1.32768, + "grad_norm": 0.1702159196138382, + "learning_rate": 3.51340206185567e-05, + "loss": 0.4934, + "step": 4149 + }, + { + "epoch": 1.328, + "grad_norm": 0.15907424688339233, + "learning_rate": 3.5092783505154645e-05, + "loss": 0.5242, + "step": 4150 + }, + { + "epoch": 1.32832, + "grad_norm": 0.16572628915309906, + "learning_rate": 3.5051546391752576e-05, + "loss": 0.5186, + "step": 4151 + }, + { + "epoch": 1.32864, + "grad_norm": 0.15777361392974854, + "learning_rate": 3.5010309278350514e-05, + "loss": 0.7009, + "step": 4152 + }, + { + "epoch": 1.32896, + "grad_norm": 0.1557043343782425, + "learning_rate": 3.496907216494846e-05, + "loss": 0.6816, + "step": 4153 + }, + { + "epoch": 1.32928, + "grad_norm": 0.1635477989912033, + "learning_rate": 3.492783505154639e-05, + "loss": 0.6416, + "step": 4154 + }, + { + "epoch": 1.3296000000000001, + "grad_norm": 0.16396832466125488, + "learning_rate": 3.4886597938144335e-05, + "loss": 0.6172, + "step": 4155 + }, + { + "epoch": 1.32992, + "grad_norm": 0.17230965197086334, + "learning_rate": 3.4845360824742266e-05, + "loss": 0.6826, + "step": 4156 + }, + { + "epoch": 1.3302399999999999, + "grad_norm": 0.17433539032936096, + "learning_rate": 3.480412371134021e-05, + "loss": 0.693, + "step": 4157 + }, + { + "epoch": 1.33056, + "grad_norm": 0.16696156561374664, + "learning_rate": 3.476288659793814e-05, + "loss": 0.5943, + "step": 4158 + }, + { + "epoch": 1.33088, + "grad_norm": 0.2033376544713974, + "learning_rate": 3.472164948453609e-05, + "loss": 0.5569, + "step": 4159 + }, + { + "epoch": 1.3312, + "grad_norm": 0.18195177614688873, + "learning_rate": 3.4680412371134025e-05, + "loss": 0.6795, + "step": 4160 + }, + { + "epoch": 1.33152, + "grad_norm": 0.1706056147813797, + "learning_rate": 3.4639175257731957e-05, + "loss": 0.7335, + "step": 4161 + }, + { + "epoch": 1.3318400000000001, + "grad_norm": 0.16146744787693024, + "learning_rate": 3.45979381443299e-05, + "loss": 0.7302, + "step": 4162 + }, + { + "epoch": 1.33216, + "grad_norm": 0.18231101334095, + "learning_rate": 3.455670103092783e-05, + "loss": 0.8253, + "step": 4163 + }, + { + "epoch": 1.3324799999999999, + "grad_norm": 0.18089066445827484, + "learning_rate": 3.451546391752578e-05, + "loss": 0.5007, + "step": 4164 + }, + { + "epoch": 1.3328, + "grad_norm": 0.17174197733402252, + "learning_rate": 3.447422680412371e-05, + "loss": 0.7461, + "step": 4165 + }, + { + "epoch": 1.33312, + "grad_norm": 0.17788538336753845, + "learning_rate": 3.4432989690721654e-05, + "loss": 0.7959, + "step": 4166 + }, + { + "epoch": 1.33344, + "grad_norm": 0.19028179347515106, + "learning_rate": 3.439175257731959e-05, + "loss": 0.5956, + "step": 4167 + }, + { + "epoch": 1.33376, + "grad_norm": 0.1663159877061844, + "learning_rate": 3.435051546391753e-05, + "loss": 0.4975, + "step": 4168 + }, + { + "epoch": 1.33408, + "grad_norm": 0.16774684190750122, + "learning_rate": 3.430927835051547e-05, + "loss": 0.6463, + "step": 4169 + }, + { + "epoch": 1.3344, + "grad_norm": 0.1656305491924286, + "learning_rate": 3.42680412371134e-05, + "loss": 0.6841, + "step": 4170 + }, + { + "epoch": 1.33472, + "grad_norm": 0.19979441165924072, + "learning_rate": 3.4226804123711344e-05, + "loss": 0.6376, + "step": 4171 + }, + { + "epoch": 1.33504, + "grad_norm": 0.17172974348068237, + "learning_rate": 3.4185567010309275e-05, + "loss": 0.6668, + "step": 4172 + }, + { + "epoch": 1.33536, + "grad_norm": 0.16718411445617676, + "learning_rate": 3.414432989690722e-05, + "loss": 0.7422, + "step": 4173 + }, + { + "epoch": 1.33568, + "grad_norm": 0.17090563476085663, + "learning_rate": 3.410309278350516e-05, + "loss": 0.5942, + "step": 4174 + }, + { + "epoch": 1.336, + "grad_norm": 0.18768467009067535, + "learning_rate": 3.4061855670103096e-05, + "loss": 0.7272, + "step": 4175 + }, + { + "epoch": 1.33632, + "grad_norm": 0.15505658090114594, + "learning_rate": 3.4020618556701034e-05, + "loss": 0.4975, + "step": 4176 + }, + { + "epoch": 1.33664, + "grad_norm": 0.17905862629413605, + "learning_rate": 3.397938144329897e-05, + "loss": 0.6165, + "step": 4177 + }, + { + "epoch": 1.33696, + "grad_norm": 0.16990354657173157, + "learning_rate": 3.393814432989691e-05, + "loss": 0.5109, + "step": 4178 + }, + { + "epoch": 1.33728, + "grad_norm": 0.18383246660232544, + "learning_rate": 3.389690721649484e-05, + "loss": 0.6767, + "step": 4179 + }, + { + "epoch": 1.3376000000000001, + "grad_norm": 0.18244962394237518, + "learning_rate": 3.3855670103092786e-05, + "loss": 0.5524, + "step": 4180 + }, + { + "epoch": 1.33792, + "grad_norm": 0.16114479303359985, + "learning_rate": 3.3814432989690724e-05, + "loss": 0.5833, + "step": 4181 + }, + { + "epoch": 1.3382399999999999, + "grad_norm": 0.14359430968761444, + "learning_rate": 3.377319587628866e-05, + "loss": 0.6817, + "step": 4182 + }, + { + "epoch": 1.33856, + "grad_norm": 0.1708555519580841, + "learning_rate": 3.37319587628866e-05, + "loss": 0.6959, + "step": 4183 + }, + { + "epoch": 1.33888, + "grad_norm": 0.17812421917915344, + "learning_rate": 3.369072164948454e-05, + "loss": 0.6709, + "step": 4184 + }, + { + "epoch": 1.3392, + "grad_norm": 0.1644599586725235, + "learning_rate": 3.3649484536082477e-05, + "loss": 0.4741, + "step": 4185 + }, + { + "epoch": 1.33952, + "grad_norm": 0.17871704697608948, + "learning_rate": 3.3608247422680415e-05, + "loss": 0.6257, + "step": 4186 + }, + { + "epoch": 1.33984, + "grad_norm": 0.17532320320606232, + "learning_rate": 3.356701030927835e-05, + "loss": 0.5856, + "step": 4187 + }, + { + "epoch": 1.34016, + "grad_norm": 0.16364017128944397, + "learning_rate": 3.352577319587629e-05, + "loss": 0.5678, + "step": 4188 + }, + { + "epoch": 1.34048, + "grad_norm": 0.16578179597854614, + "learning_rate": 3.348453608247423e-05, + "loss": 0.6948, + "step": 4189 + }, + { + "epoch": 1.3408, + "grad_norm": 0.15510474145412445, + "learning_rate": 3.344329896907217e-05, + "loss": 0.5631, + "step": 4190 + }, + { + "epoch": 1.34112, + "grad_norm": 0.17132383584976196, + "learning_rate": 3.3402061855670105e-05, + "loss": 0.6614, + "step": 4191 + }, + { + "epoch": 1.34144, + "grad_norm": 0.1798267513513565, + "learning_rate": 3.336082474226804e-05, + "loss": 0.6966, + "step": 4192 + }, + { + "epoch": 1.34176, + "grad_norm": 0.16424375772476196, + "learning_rate": 3.331958762886598e-05, + "loss": 0.4671, + "step": 4193 + }, + { + "epoch": 1.34208, + "grad_norm": 0.14996325969696045, + "learning_rate": 3.327835051546392e-05, + "loss": 0.5735, + "step": 4194 + }, + { + "epoch": 1.3424, + "grad_norm": 0.18250346183776855, + "learning_rate": 3.323711340206186e-05, + "loss": 0.5045, + "step": 4195 + }, + { + "epoch": 1.34272, + "grad_norm": 0.18423089385032654, + "learning_rate": 3.3195876288659795e-05, + "loss": 0.5305, + "step": 4196 + }, + { + "epoch": 1.34304, + "grad_norm": 0.18243086338043213, + "learning_rate": 3.315463917525773e-05, + "loss": 0.6209, + "step": 4197 + }, + { + "epoch": 1.34336, + "grad_norm": 0.175373375415802, + "learning_rate": 3.311340206185567e-05, + "loss": 0.5767, + "step": 4198 + }, + { + "epoch": 1.34368, + "grad_norm": 0.1949029415845871, + "learning_rate": 3.307216494845361e-05, + "loss": 0.6774, + "step": 4199 + }, + { + "epoch": 1.3439999999999999, + "grad_norm": 0.16588599979877472, + "learning_rate": 3.303092783505155e-05, + "loss": 0.6619, + "step": 4200 + }, + { + "epoch": 1.34432, + "grad_norm": 0.16757924854755402, + "learning_rate": 3.2989690721649485e-05, + "loss": 0.6912, + "step": 4201 + }, + { + "epoch": 1.34464, + "grad_norm": 0.17532667517662048, + "learning_rate": 3.2948453608247423e-05, + "loss": 0.7024, + "step": 4202 + }, + { + "epoch": 1.34496, + "grad_norm": 0.18824657797813416, + "learning_rate": 3.290721649484536e-05, + "loss": 0.72, + "step": 4203 + }, + { + "epoch": 1.34528, + "grad_norm": 0.16210421919822693, + "learning_rate": 3.28659793814433e-05, + "loss": 0.8508, + "step": 4204 + }, + { + "epoch": 1.3456000000000001, + "grad_norm": 0.14879001677036285, + "learning_rate": 3.282474226804124e-05, + "loss": 0.5652, + "step": 4205 + }, + { + "epoch": 1.34592, + "grad_norm": 0.14522148668766022, + "learning_rate": 3.2783505154639176e-05, + "loss": 0.5159, + "step": 4206 + }, + { + "epoch": 1.3462399999999999, + "grad_norm": 0.1888684630393982, + "learning_rate": 3.274226804123712e-05, + "loss": 0.7346, + "step": 4207 + }, + { + "epoch": 1.34656, + "grad_norm": 0.14803984761238098, + "learning_rate": 3.270103092783505e-05, + "loss": 0.603, + "step": 4208 + }, + { + "epoch": 1.34688, + "grad_norm": 0.15938089787960052, + "learning_rate": 3.265979381443299e-05, + "loss": 0.5864, + "step": 4209 + }, + { + "epoch": 1.3472, + "grad_norm": 0.16499760746955872, + "learning_rate": 3.261855670103093e-05, + "loss": 0.5828, + "step": 4210 + }, + { + "epoch": 1.34752, + "grad_norm": 0.18936966359615326, + "learning_rate": 3.2577319587628866e-05, + "loss": 0.6419, + "step": 4211 + }, + { + "epoch": 1.34784, + "grad_norm": 0.16670139133930206, + "learning_rate": 3.2536082474226804e-05, + "loss": 0.5609, + "step": 4212 + }, + { + "epoch": 1.34816, + "grad_norm": 0.17973646521568298, + "learning_rate": 3.249484536082474e-05, + "loss": 0.636, + "step": 4213 + }, + { + "epoch": 1.34848, + "grad_norm": 0.16859476268291473, + "learning_rate": 3.245360824742269e-05, + "loss": 0.5958, + "step": 4214 + }, + { + "epoch": 1.3488, + "grad_norm": 0.18482960760593414, + "learning_rate": 3.241237113402062e-05, + "loss": 0.8492, + "step": 4215 + }, + { + "epoch": 1.34912, + "grad_norm": 0.17158403992652893, + "learning_rate": 3.237113402061856e-05, + "loss": 0.6404, + "step": 4216 + }, + { + "epoch": 1.34944, + "grad_norm": 0.1715967208147049, + "learning_rate": 3.2329896907216494e-05, + "loss": 0.6779, + "step": 4217 + }, + { + "epoch": 1.34976, + "grad_norm": 0.1666477620601654, + "learning_rate": 3.228865979381443e-05, + "loss": 0.6989, + "step": 4218 + }, + { + "epoch": 1.35008, + "grad_norm": 0.15565049648284912, + "learning_rate": 3.224742268041237e-05, + "loss": 0.6347, + "step": 4219 + }, + { + "epoch": 1.3504, + "grad_norm": 0.17680475115776062, + "learning_rate": 3.220618556701031e-05, + "loss": 0.6154, + "step": 4220 + }, + { + "epoch": 1.35072, + "grad_norm": 0.155681774020195, + "learning_rate": 3.216494845360825e-05, + "loss": 0.6126, + "step": 4221 + }, + { + "epoch": 1.35104, + "grad_norm": 0.15983374416828156, + "learning_rate": 3.2123711340206184e-05, + "loss": 0.508, + "step": 4222 + }, + { + "epoch": 1.3513600000000001, + "grad_norm": 0.15935534238815308, + "learning_rate": 3.208247422680413e-05, + "loss": 0.6581, + "step": 4223 + }, + { + "epoch": 1.35168, + "grad_norm": 0.22319461405277252, + "learning_rate": 3.204123711340206e-05, + "loss": 0.6892, + "step": 4224 + }, + { + "epoch": 1.3519999999999999, + "grad_norm": 0.18240396678447723, + "learning_rate": 3.2000000000000005e-05, + "loss": 0.708, + "step": 4225 + }, + { + "epoch": 1.35232, + "grad_norm": 0.19421124458312988, + "learning_rate": 3.1958762886597937e-05, + "loss": 0.6236, + "step": 4226 + }, + { + "epoch": 1.35264, + "grad_norm": 0.16163286566734314, + "learning_rate": 3.1917525773195875e-05, + "loss": 0.5281, + "step": 4227 + }, + { + "epoch": 1.35296, + "grad_norm": 0.156102254986763, + "learning_rate": 3.187628865979382e-05, + "loss": 0.7051, + "step": 4228 + }, + { + "epoch": 1.35328, + "grad_norm": 0.18062733113765717, + "learning_rate": 3.183505154639175e-05, + "loss": 0.5144, + "step": 4229 + }, + { + "epoch": 1.3536000000000001, + "grad_norm": 0.15566325187683105, + "learning_rate": 3.1793814432989696e-05, + "loss": 0.6094, + "step": 4230 + }, + { + "epoch": 1.35392, + "grad_norm": 0.151219442486763, + "learning_rate": 3.175257731958763e-05, + "loss": 0.5362, + "step": 4231 + }, + { + "epoch": 1.3542399999999999, + "grad_norm": 0.19807979464530945, + "learning_rate": 3.171134020618557e-05, + "loss": 0.6431, + "step": 4232 + }, + { + "epoch": 1.35456, + "grad_norm": 0.17742912471294403, + "learning_rate": 3.16701030927835e-05, + "loss": 0.6011, + "step": 4233 + }, + { + "epoch": 1.35488, + "grad_norm": 0.17424309253692627, + "learning_rate": 3.162886597938145e-05, + "loss": 0.5641, + "step": 4234 + }, + { + "epoch": 1.3552, + "grad_norm": 0.17553144693374634, + "learning_rate": 3.1587628865979386e-05, + "loss": 0.5275, + "step": 4235 + }, + { + "epoch": 1.35552, + "grad_norm": 0.16215544939041138, + "learning_rate": 3.154639175257732e-05, + "loss": 0.5293, + "step": 4236 + }, + { + "epoch": 1.35584, + "grad_norm": 0.1639627367258072, + "learning_rate": 3.150515463917526e-05, + "loss": 0.8251, + "step": 4237 + }, + { + "epoch": 1.35616, + "grad_norm": 0.19562141597270966, + "learning_rate": 3.146391752577319e-05, + "loss": 0.551, + "step": 4238 + }, + { + "epoch": 1.35648, + "grad_norm": 0.17020286619663239, + "learning_rate": 3.142268041237114e-05, + "loss": 0.6031, + "step": 4239 + }, + { + "epoch": 1.3568, + "grad_norm": 0.17878757417201996, + "learning_rate": 3.138144329896907e-05, + "loss": 0.5833, + "step": 4240 + }, + { + "epoch": 1.35712, + "grad_norm": 0.1822752207517624, + "learning_rate": 3.1340206185567014e-05, + "loss": 0.4789, + "step": 4241 + }, + { + "epoch": 1.35744, + "grad_norm": 0.1861959993839264, + "learning_rate": 3.129896907216495e-05, + "loss": 0.7227, + "step": 4242 + }, + { + "epoch": 1.35776, + "grad_norm": 0.1710781306028366, + "learning_rate": 3.125773195876289e-05, + "loss": 0.6319, + "step": 4243 + }, + { + "epoch": 1.35808, + "grad_norm": 0.191038578748703, + "learning_rate": 3.121649484536083e-05, + "loss": 0.5079, + "step": 4244 + }, + { + "epoch": 1.3584, + "grad_norm": 0.18434099853038788, + "learning_rate": 3.117525773195876e-05, + "loss": 0.5806, + "step": 4245 + }, + { + "epoch": 1.35872, + "grad_norm": 0.18103472888469696, + "learning_rate": 3.1134020618556704e-05, + "loss": 0.6967, + "step": 4246 + }, + { + "epoch": 1.35904, + "grad_norm": 0.15182404220104218, + "learning_rate": 3.1092783505154636e-05, + "loss": 0.6462, + "step": 4247 + }, + { + "epoch": 1.3593600000000001, + "grad_norm": 0.19107307493686676, + "learning_rate": 3.105154639175258e-05, + "loss": 0.6616, + "step": 4248 + }, + { + "epoch": 1.35968, + "grad_norm": 0.17701385915279388, + "learning_rate": 3.101030927835052e-05, + "loss": 0.6591, + "step": 4249 + }, + { + "epoch": 1.3599999999999999, + "grad_norm": 0.18347352743148804, + "learning_rate": 3.0969072164948457e-05, + "loss": 0.5296, + "step": 4250 + }, + { + "epoch": 1.36032, + "grad_norm": 0.16632042825222015, + "learning_rate": 3.0927835051546395e-05, + "loss": 0.5334, + "step": 4251 + }, + { + "epoch": 1.36064, + "grad_norm": 0.18581409752368927, + "learning_rate": 3.088659793814433e-05, + "loss": 0.6837, + "step": 4252 + }, + { + "epoch": 1.36096, + "grad_norm": 0.17160242795944214, + "learning_rate": 3.084536082474227e-05, + "loss": 0.6772, + "step": 4253 + }, + { + "epoch": 1.36128, + "grad_norm": 0.16472972929477692, + "learning_rate": 3.080412371134021e-05, + "loss": 0.6442, + "step": 4254 + }, + { + "epoch": 1.3616, + "grad_norm": 0.17449098825454712, + "learning_rate": 3.076288659793815e-05, + "loss": 0.7537, + "step": 4255 + }, + { + "epoch": 1.36192, + "grad_norm": 0.18821565806865692, + "learning_rate": 3.0721649484536085e-05, + "loss": 0.5904, + "step": 4256 + }, + { + "epoch": 1.36224, + "grad_norm": 0.17404179275035858, + "learning_rate": 3.068041237113402e-05, + "loss": 0.6397, + "step": 4257 + }, + { + "epoch": 1.36256, + "grad_norm": 0.16451196372509003, + "learning_rate": 3.063917525773196e-05, + "loss": 0.6863, + "step": 4258 + }, + { + "epoch": 1.36288, + "grad_norm": 0.166012242436409, + "learning_rate": 3.05979381443299e-05, + "loss": 0.631, + "step": 4259 + }, + { + "epoch": 1.3632, + "grad_norm": 0.1768709272146225, + "learning_rate": 3.055670103092784e-05, + "loss": 0.6173, + "step": 4260 + }, + { + "epoch": 1.36352, + "grad_norm": 0.17743991315364838, + "learning_rate": 3.051546391752578e-05, + "loss": 0.7039, + "step": 4261 + }, + { + "epoch": 1.36384, + "grad_norm": 0.17292630672454834, + "learning_rate": 3.0474226804123713e-05, + "loss": 0.5512, + "step": 4262 + }, + { + "epoch": 1.36416, + "grad_norm": 0.1877375692129135, + "learning_rate": 3.043298969072165e-05, + "loss": 0.6332, + "step": 4263 + }, + { + "epoch": 1.36448, + "grad_norm": 0.18252000212669373, + "learning_rate": 3.039175257731959e-05, + "loss": 0.5371, + "step": 4264 + }, + { + "epoch": 1.3648, + "grad_norm": 0.19626648724079132, + "learning_rate": 3.0350515463917527e-05, + "loss": 0.6827, + "step": 4265 + }, + { + "epoch": 1.3651200000000001, + "grad_norm": 0.18412652611732483, + "learning_rate": 3.0309278350515462e-05, + "loss": 0.5424, + "step": 4266 + }, + { + "epoch": 1.36544, + "grad_norm": 0.18170318007469177, + "learning_rate": 3.0268041237113403e-05, + "loss": 0.5287, + "step": 4267 + }, + { + "epoch": 1.3657599999999999, + "grad_norm": 0.18524254858493805, + "learning_rate": 3.0226804123711345e-05, + "loss": 0.6626, + "step": 4268 + }, + { + "epoch": 1.36608, + "grad_norm": 0.1940350979566574, + "learning_rate": 3.018556701030928e-05, + "loss": 0.5886, + "step": 4269 + }, + { + "epoch": 1.3664, + "grad_norm": 0.1866217404603958, + "learning_rate": 3.014432989690722e-05, + "loss": 0.6541, + "step": 4270 + }, + { + "epoch": 1.36672, + "grad_norm": 0.19403192400932312, + "learning_rate": 3.0103092783505156e-05, + "loss": 0.7418, + "step": 4271 + }, + { + "epoch": 1.36704, + "grad_norm": 0.19303010404109955, + "learning_rate": 3.0061855670103094e-05, + "loss": 0.7189, + "step": 4272 + }, + { + "epoch": 1.3673600000000001, + "grad_norm": 0.15204846858978271, + "learning_rate": 3.0020618556701032e-05, + "loss": 0.6086, + "step": 4273 + }, + { + "epoch": 1.36768, + "grad_norm": 0.20270998775959015, + "learning_rate": 2.997938144329897e-05, + "loss": 0.7093, + "step": 4274 + }, + { + "epoch": 1.3679999999999999, + "grad_norm": 0.2002064734697342, + "learning_rate": 2.993814432989691e-05, + "loss": 0.6294, + "step": 4275 + }, + { + "epoch": 1.36832, + "grad_norm": 0.14641259610652924, + "learning_rate": 2.9896907216494846e-05, + "loss": 0.4407, + "step": 4276 + }, + { + "epoch": 1.36864, + "grad_norm": 0.15353450179100037, + "learning_rate": 2.9855670103092787e-05, + "loss": 0.6805, + "step": 4277 + }, + { + "epoch": 1.36896, + "grad_norm": 0.17229221761226654, + "learning_rate": 2.9814432989690722e-05, + "loss": 0.473, + "step": 4278 + }, + { + "epoch": 1.36928, + "grad_norm": 0.16991494596004486, + "learning_rate": 2.9773195876288663e-05, + "loss": 0.6759, + "step": 4279 + }, + { + "epoch": 1.3696, + "grad_norm": 0.17608174681663513, + "learning_rate": 2.9731958762886598e-05, + "loss": 0.476, + "step": 4280 + }, + { + "epoch": 1.36992, + "grad_norm": 0.18986006081104279, + "learning_rate": 2.9690721649484536e-05, + "loss": 0.5784, + "step": 4281 + }, + { + "epoch": 1.37024, + "grad_norm": 0.16312289237976074, + "learning_rate": 2.9649484536082478e-05, + "loss": 0.6792, + "step": 4282 + }, + { + "epoch": 1.37056, + "grad_norm": 0.16459216177463531, + "learning_rate": 2.9608247422680412e-05, + "loss": 0.6568, + "step": 4283 + }, + { + "epoch": 1.37088, + "grad_norm": 0.15064825117588043, + "learning_rate": 2.9567010309278354e-05, + "loss": 0.6193, + "step": 4284 + }, + { + "epoch": 1.3712, + "grad_norm": 0.17576581239700317, + "learning_rate": 2.952577319587629e-05, + "loss": 0.5971, + "step": 4285 + }, + { + "epoch": 1.37152, + "grad_norm": 0.18081048130989075, + "learning_rate": 2.948453608247423e-05, + "loss": 0.5935, + "step": 4286 + }, + { + "epoch": 1.37184, + "grad_norm": 0.17843838036060333, + "learning_rate": 2.9443298969072164e-05, + "loss": 0.7204, + "step": 4287 + }, + { + "epoch": 1.37216, + "grad_norm": 0.17621497809886932, + "learning_rate": 2.9402061855670106e-05, + "loss": 0.494, + "step": 4288 + }, + { + "epoch": 1.37248, + "grad_norm": 0.17775355279445648, + "learning_rate": 2.9360824742268044e-05, + "loss": 0.577, + "step": 4289 + }, + { + "epoch": 1.3728, + "grad_norm": 0.1764974296092987, + "learning_rate": 2.931958762886598e-05, + "loss": 0.7755, + "step": 4290 + }, + { + "epoch": 1.3731200000000001, + "grad_norm": 0.17004409432411194, + "learning_rate": 2.927835051546392e-05, + "loss": 0.6087, + "step": 4291 + }, + { + "epoch": 1.37344, + "grad_norm": 0.18059895932674408, + "learning_rate": 2.9237113402061855e-05, + "loss": 0.6836, + "step": 4292 + }, + { + "epoch": 1.3737599999999999, + "grad_norm": 0.16760993003845215, + "learning_rate": 2.9195876288659796e-05, + "loss": 0.6435, + "step": 4293 + }, + { + "epoch": 1.37408, + "grad_norm": 0.16186395287513733, + "learning_rate": 2.915463917525773e-05, + "loss": 0.6052, + "step": 4294 + }, + { + "epoch": 1.3744, + "grad_norm": 0.1611078530550003, + "learning_rate": 2.9113402061855672e-05, + "loss": 0.6066, + "step": 4295 + }, + { + "epoch": 1.37472, + "grad_norm": 0.1753527969121933, + "learning_rate": 2.907216494845361e-05, + "loss": 0.6143, + "step": 4296 + }, + { + "epoch": 1.37504, + "grad_norm": 0.1766851395368576, + "learning_rate": 2.903092783505155e-05, + "loss": 0.5898, + "step": 4297 + }, + { + "epoch": 1.3753600000000001, + "grad_norm": 0.20553576946258545, + "learning_rate": 2.8989690721649486e-05, + "loss": 0.5711, + "step": 4298 + }, + { + "epoch": 1.37568, + "grad_norm": 0.20391979813575745, + "learning_rate": 2.894845360824742e-05, + "loss": 0.6069, + "step": 4299 + }, + { + "epoch": 1.376, + "grad_norm": 0.17891335487365723, + "learning_rate": 2.8907216494845362e-05, + "loss": 0.6083, + "step": 4300 + }, + { + "epoch": 1.37632, + "grad_norm": 0.18589816987514496, + "learning_rate": 2.8865979381443297e-05, + "loss": 0.4808, + "step": 4301 + }, + { + "epoch": 1.37664, + "grad_norm": 0.18277621269226074, + "learning_rate": 2.882474226804124e-05, + "loss": 0.5558, + "step": 4302 + }, + { + "epoch": 1.37696, + "grad_norm": 0.17378287017345428, + "learning_rate": 2.878350515463918e-05, + "loss": 0.662, + "step": 4303 + }, + { + "epoch": 1.37728, + "grad_norm": 0.18203498423099518, + "learning_rate": 2.8742268041237115e-05, + "loss": 0.5874, + "step": 4304 + }, + { + "epoch": 1.3776, + "grad_norm": 0.16593731939792633, + "learning_rate": 2.8701030927835053e-05, + "loss": 0.6153, + "step": 4305 + }, + { + "epoch": 1.37792, + "grad_norm": 0.18321040272712708, + "learning_rate": 2.865979381443299e-05, + "loss": 0.6815, + "step": 4306 + }, + { + "epoch": 1.37824, + "grad_norm": 0.1536594033241272, + "learning_rate": 2.861855670103093e-05, + "loss": 0.8207, + "step": 4307 + }, + { + "epoch": 1.37856, + "grad_norm": 0.17886117100715637, + "learning_rate": 2.8577319587628863e-05, + "loss": 0.5559, + "step": 4308 + }, + { + "epoch": 1.37888, + "grad_norm": 0.19552530348300934, + "learning_rate": 2.8536082474226805e-05, + "loss": 0.5507, + "step": 4309 + }, + { + "epoch": 1.3792, + "grad_norm": 0.19648703932762146, + "learning_rate": 2.8494845360824746e-05, + "loss": 0.5416, + "step": 4310 + }, + { + "epoch": 1.37952, + "grad_norm": 0.16202472150325775, + "learning_rate": 2.845360824742268e-05, + "loss": 0.6003, + "step": 4311 + }, + { + "epoch": 1.37984, + "grad_norm": 0.16799217462539673, + "learning_rate": 2.8412371134020622e-05, + "loss": 0.5699, + "step": 4312 + }, + { + "epoch": 1.38016, + "grad_norm": 0.17623008787631989, + "learning_rate": 2.8371134020618557e-05, + "loss": 0.5272, + "step": 4313 + }, + { + "epoch": 1.38048, + "grad_norm": 0.16393224895000458, + "learning_rate": 2.8329896907216495e-05, + "loss": 0.5886, + "step": 4314 + }, + { + "epoch": 1.3808, + "grad_norm": 0.19053421914577484, + "learning_rate": 2.8288659793814437e-05, + "loss": 0.5842, + "step": 4315 + }, + { + "epoch": 1.3811200000000001, + "grad_norm": 0.20031902194023132, + "learning_rate": 2.824742268041237e-05, + "loss": 0.5964, + "step": 4316 + }, + { + "epoch": 1.38144, + "grad_norm": 0.19026906788349152, + "learning_rate": 2.8206185567010313e-05, + "loss": 0.5961, + "step": 4317 + }, + { + "epoch": 1.3817599999999999, + "grad_norm": 0.15659475326538086, + "learning_rate": 2.8164948453608247e-05, + "loss": 0.6427, + "step": 4318 + }, + { + "epoch": 1.38208, + "grad_norm": 0.16410760581493378, + "learning_rate": 2.812371134020619e-05, + "loss": 0.7042, + "step": 4319 + }, + { + "epoch": 1.3824, + "grad_norm": 0.1812852919101715, + "learning_rate": 2.8082474226804123e-05, + "loss": 0.7134, + "step": 4320 + }, + { + "epoch": 1.38272, + "grad_norm": 0.16654786467552185, + "learning_rate": 2.8041237113402065e-05, + "loss": 0.4821, + "step": 4321 + }, + { + "epoch": 1.38304, + "grad_norm": 0.1845560073852539, + "learning_rate": 2.8000000000000003e-05, + "loss": 0.569, + "step": 4322 + }, + { + "epoch": 1.38336, + "grad_norm": 0.18596427142620087, + "learning_rate": 2.7958762886597938e-05, + "loss": 0.6451, + "step": 4323 + }, + { + "epoch": 1.38368, + "grad_norm": 0.1698964387178421, + "learning_rate": 2.791752577319588e-05, + "loss": 0.6292, + "step": 4324 + }, + { + "epoch": 1.384, + "grad_norm": 0.20047952234745026, + "learning_rate": 2.7876288659793814e-05, + "loss": 0.628, + "step": 4325 + }, + { + "epoch": 1.38432, + "grad_norm": 0.17218372225761414, + "learning_rate": 2.7835051546391755e-05, + "loss": 0.6018, + "step": 4326 + }, + { + "epoch": 1.38464, + "grad_norm": 0.20120351016521454, + "learning_rate": 2.779381443298969e-05, + "loss": 0.7471, + "step": 4327 + }, + { + "epoch": 1.38496, + "grad_norm": 0.17577135562896729, + "learning_rate": 2.775257731958763e-05, + "loss": 0.7045, + "step": 4328 + }, + { + "epoch": 1.38528, + "grad_norm": 0.16630318760871887, + "learning_rate": 2.771134020618557e-05, + "loss": 0.6045, + "step": 4329 + }, + { + "epoch": 1.3856, + "grad_norm": 0.1741822510957718, + "learning_rate": 2.7670103092783507e-05, + "loss": 0.6338, + "step": 4330 + }, + { + "epoch": 1.38592, + "grad_norm": 0.17429104447364807, + "learning_rate": 2.7628865979381445e-05, + "loss": 0.5909, + "step": 4331 + }, + { + "epoch": 1.38624, + "grad_norm": 0.18926171958446503, + "learning_rate": 2.758762886597938e-05, + "loss": 0.6358, + "step": 4332 + }, + { + "epoch": 1.38656, + "grad_norm": 0.16413870453834534, + "learning_rate": 2.754639175257732e-05, + "loss": 0.5498, + "step": 4333 + }, + { + "epoch": 1.3868800000000001, + "grad_norm": 0.19882211089134216, + "learning_rate": 2.7505154639175256e-05, + "loss": 0.7153, + "step": 4334 + }, + { + "epoch": 1.3872, + "grad_norm": 0.16295145452022552, + "learning_rate": 2.7463917525773198e-05, + "loss": 0.7657, + "step": 4335 + }, + { + "epoch": 1.3875199999999999, + "grad_norm": 0.1640796959400177, + "learning_rate": 2.742268041237114e-05, + "loss": 0.6422, + "step": 4336 + }, + { + "epoch": 1.38784, + "grad_norm": 0.16327081620693207, + "learning_rate": 2.7381443298969074e-05, + "loss": 0.5789, + "step": 4337 + }, + { + "epoch": 1.38816, + "grad_norm": 0.1494014710187912, + "learning_rate": 2.7340206185567012e-05, + "loss": 0.508, + "step": 4338 + }, + { + "epoch": 1.38848, + "grad_norm": 0.17914126813411713, + "learning_rate": 2.729896907216495e-05, + "loss": 0.6385, + "step": 4339 + }, + { + "epoch": 1.3888, + "grad_norm": 0.17945991456508636, + "learning_rate": 2.7257731958762888e-05, + "loss": 0.6148, + "step": 4340 + }, + { + "epoch": 1.3891200000000001, + "grad_norm": 0.18314231932163239, + "learning_rate": 2.7216494845360823e-05, + "loss": 0.7221, + "step": 4341 + }, + { + "epoch": 1.38944, + "grad_norm": 0.17675819993019104, + "learning_rate": 2.7175257731958764e-05, + "loss": 0.7385, + "step": 4342 + }, + { + "epoch": 1.3897599999999999, + "grad_norm": 0.2135961651802063, + "learning_rate": 2.7134020618556705e-05, + "loss": 0.5983, + "step": 4343 + }, + { + "epoch": 1.39008, + "grad_norm": 0.17229905724525452, + "learning_rate": 2.709278350515464e-05, + "loss": 0.6333, + "step": 4344 + }, + { + "epoch": 1.3904, + "grad_norm": 0.16083352267742157, + "learning_rate": 2.705154639175258e-05, + "loss": 0.611, + "step": 4345 + }, + { + "epoch": 1.39072, + "grad_norm": 0.1584169566631317, + "learning_rate": 2.7010309278350516e-05, + "loss": 0.5667, + "step": 4346 + }, + { + "epoch": 1.39104, + "grad_norm": 0.1920241266489029, + "learning_rate": 2.6969072164948454e-05, + "loss": 0.6279, + "step": 4347 + }, + { + "epoch": 1.39136, + "grad_norm": 0.17197296023368835, + "learning_rate": 2.6927835051546392e-05, + "loss": 0.657, + "step": 4348 + }, + { + "epoch": 1.39168, + "grad_norm": 0.16557329893112183, + "learning_rate": 2.688659793814433e-05, + "loss": 0.5137, + "step": 4349 + }, + { + "epoch": 1.392, + "grad_norm": 0.17517340183258057, + "learning_rate": 2.6845360824742272e-05, + "loss": 0.6286, + "step": 4350 + }, + { + "epoch": 1.39232, + "grad_norm": 0.17068150639533997, + "learning_rate": 2.6804123711340206e-05, + "loss": 0.5919, + "step": 4351 + }, + { + "epoch": 1.39264, + "grad_norm": 0.17032217979431152, + "learning_rate": 2.6762886597938148e-05, + "loss": 0.631, + "step": 4352 + }, + { + "epoch": 1.39296, + "grad_norm": 0.16726545989513397, + "learning_rate": 2.6721649484536083e-05, + "loss": 0.5804, + "step": 4353 + }, + { + "epoch": 1.39328, + "grad_norm": 0.1657009720802307, + "learning_rate": 2.6680412371134024e-05, + "loss": 0.5891, + "step": 4354 + }, + { + "epoch": 1.3936, + "grad_norm": 0.18014325201511383, + "learning_rate": 2.663917525773196e-05, + "loss": 0.5637, + "step": 4355 + }, + { + "epoch": 1.39392, + "grad_norm": 0.1603650450706482, + "learning_rate": 2.6597938144329897e-05, + "loss": 0.5137, + "step": 4356 + }, + { + "epoch": 1.39424, + "grad_norm": 0.1470848023891449, + "learning_rate": 2.6556701030927838e-05, + "loss": 0.6019, + "step": 4357 + }, + { + "epoch": 1.39456, + "grad_norm": 0.1808815896511078, + "learning_rate": 2.6515463917525773e-05, + "loss": 0.6342, + "step": 4358 + }, + { + "epoch": 1.3948800000000001, + "grad_norm": 0.19223208725452423, + "learning_rate": 2.6474226804123714e-05, + "loss": 0.7474, + "step": 4359 + }, + { + "epoch": 1.3952, + "grad_norm": 0.1613210290670395, + "learning_rate": 2.643298969072165e-05, + "loss": 0.5558, + "step": 4360 + }, + { + "epoch": 1.3955199999999999, + "grad_norm": 0.17510631680488586, + "learning_rate": 2.639175257731959e-05, + "loss": 0.5866, + "step": 4361 + }, + { + "epoch": 1.39584, + "grad_norm": 0.17747077345848083, + "learning_rate": 2.6350515463917525e-05, + "loss": 0.6886, + "step": 4362 + }, + { + "epoch": 1.39616, + "grad_norm": 0.19352540373802185, + "learning_rate": 2.6309278350515466e-05, + "loss": 0.6694, + "step": 4363 + }, + { + "epoch": 1.39648, + "grad_norm": 0.15468527376651764, + "learning_rate": 2.6268041237113404e-05, + "loss": 0.7, + "step": 4364 + }, + { + "epoch": 1.3968, + "grad_norm": 0.16042891144752502, + "learning_rate": 2.622680412371134e-05, + "loss": 0.5212, + "step": 4365 + }, + { + "epoch": 1.39712, + "grad_norm": 0.14912304282188416, + "learning_rate": 2.618556701030928e-05, + "loss": 0.5616, + "step": 4366 + }, + { + "epoch": 1.39744, + "grad_norm": 0.17087720334529877, + "learning_rate": 2.6144329896907215e-05, + "loss": 0.6041, + "step": 4367 + }, + { + "epoch": 1.39776, + "grad_norm": 0.18186837434768677, + "learning_rate": 2.6103092783505157e-05, + "loss": 0.6106, + "step": 4368 + }, + { + "epoch": 1.39808, + "grad_norm": 0.17126071453094482, + "learning_rate": 2.606185567010309e-05, + "loss": 0.6188, + "step": 4369 + }, + { + "epoch": 1.3984, + "grad_norm": 0.1915753185749054, + "learning_rate": 2.6020618556701033e-05, + "loss": 0.7836, + "step": 4370 + }, + { + "epoch": 1.39872, + "grad_norm": 0.1609487682580948, + "learning_rate": 2.597938144329897e-05, + "loss": 0.788, + "step": 4371 + }, + { + "epoch": 1.39904, + "grad_norm": 0.1546202450990677, + "learning_rate": 2.593814432989691e-05, + "loss": 0.7086, + "step": 4372 + }, + { + "epoch": 1.39936, + "grad_norm": 0.1709596812725067, + "learning_rate": 2.5896907216494847e-05, + "loss": 0.5639, + "step": 4373 + }, + { + "epoch": 1.39968, + "grad_norm": 0.17999254167079926, + "learning_rate": 2.585567010309278e-05, + "loss": 0.7805, + "step": 4374 + }, + { + "epoch": 1.4, + "grad_norm": 0.16979794204235077, + "learning_rate": 2.5814432989690723e-05, + "loss": 0.7071, + "step": 4375 + }, + { + "epoch": 1.40032, + "grad_norm": 0.18748719990253448, + "learning_rate": 2.5773195876288658e-05, + "loss": 0.523, + "step": 4376 + }, + { + "epoch": 1.40064, + "grad_norm": 0.18088340759277344, + "learning_rate": 2.57319587628866e-05, + "loss": 0.4994, + "step": 4377 + }, + { + "epoch": 1.40096, + "grad_norm": 0.19258950650691986, + "learning_rate": 2.569072164948454e-05, + "loss": 0.7524, + "step": 4378 + }, + { + "epoch": 1.40128, + "grad_norm": 0.182938814163208, + "learning_rate": 2.5649484536082475e-05, + "loss": 0.6855, + "step": 4379 + }, + { + "epoch": 1.4016, + "grad_norm": 0.17788933217525482, + "learning_rate": 2.5608247422680413e-05, + "loss": 0.4428, + "step": 4380 + }, + { + "epoch": 1.40192, + "grad_norm": 0.1610572636127472, + "learning_rate": 2.556701030927835e-05, + "loss": 0.4871, + "step": 4381 + }, + { + "epoch": 1.40224, + "grad_norm": 0.19587811827659607, + "learning_rate": 2.552577319587629e-05, + "loss": 0.7363, + "step": 4382 + }, + { + "epoch": 1.40256, + "grad_norm": 0.15192607045173645, + "learning_rate": 2.548453608247423e-05, + "loss": 0.4958, + "step": 4383 + }, + { + "epoch": 1.4028800000000001, + "grad_norm": 0.20579548180103302, + "learning_rate": 2.5443298969072165e-05, + "loss": 0.7295, + "step": 4384 + }, + { + "epoch": 1.4032, + "grad_norm": 0.1643255203962326, + "learning_rate": 2.5402061855670107e-05, + "loss": 0.652, + "step": 4385 + }, + { + "epoch": 1.4035199999999999, + "grad_norm": 0.15259264409542084, + "learning_rate": 2.536082474226804e-05, + "loss": 0.5589, + "step": 4386 + }, + { + "epoch": 1.40384, + "grad_norm": 0.17888373136520386, + "learning_rate": 2.5319587628865983e-05, + "loss": 0.6588, + "step": 4387 + }, + { + "epoch": 1.40416, + "grad_norm": 0.17543144524097443, + "learning_rate": 2.5278350515463918e-05, + "loss": 0.4327, + "step": 4388 + }, + { + "epoch": 1.40448, + "grad_norm": 0.1596023291349411, + "learning_rate": 2.5237113402061856e-05, + "loss": 0.5104, + "step": 4389 + }, + { + "epoch": 1.4048, + "grad_norm": 0.1649980992078781, + "learning_rate": 2.5195876288659797e-05, + "loss": 0.5856, + "step": 4390 + }, + { + "epoch": 1.40512, + "grad_norm": 0.15757492184638977, + "learning_rate": 2.5154639175257732e-05, + "loss": 0.5936, + "step": 4391 + }, + { + "epoch": 1.40544, + "grad_norm": 0.15895769000053406, + "learning_rate": 2.5113402061855673e-05, + "loss": 0.5209, + "step": 4392 + }, + { + "epoch": 1.40576, + "grad_norm": 0.21580560505390167, + "learning_rate": 2.5072164948453608e-05, + "loss": 0.6465, + "step": 4393 + }, + { + "epoch": 1.40608, + "grad_norm": 0.18793350458145142, + "learning_rate": 2.503092783505155e-05, + "loss": 0.5731, + "step": 4394 + }, + { + "epoch": 1.4064, + "grad_norm": 0.17977844178676605, + "learning_rate": 2.4989690721649487e-05, + "loss": 0.6501, + "step": 4395 + }, + { + "epoch": 1.40672, + "grad_norm": 0.15634623169898987, + "learning_rate": 2.4948453608247425e-05, + "loss": 0.6461, + "step": 4396 + }, + { + "epoch": 1.40704, + "grad_norm": 0.18700604140758514, + "learning_rate": 2.490721649484536e-05, + "loss": 0.6104, + "step": 4397 + }, + { + "epoch": 1.40736, + "grad_norm": 0.17636099457740784, + "learning_rate": 2.4865979381443298e-05, + "loss": 0.5816, + "step": 4398 + }, + { + "epoch": 1.40768, + "grad_norm": 0.19534146785736084, + "learning_rate": 2.4824742268041236e-05, + "loss": 0.7752, + "step": 4399 + }, + { + "epoch": 1.408, + "grad_norm": 0.17842817306518555, + "learning_rate": 2.4783505154639178e-05, + "loss": 0.5592, + "step": 4400 + }, + { + "epoch": 1.40832, + "grad_norm": 0.1935962587594986, + "learning_rate": 2.4742268041237116e-05, + "loss": 0.5676, + "step": 4401 + }, + { + "epoch": 1.4086400000000001, + "grad_norm": 0.14800995588302612, + "learning_rate": 2.4701030927835054e-05, + "loss": 0.4701, + "step": 4402 + }, + { + "epoch": 1.40896, + "grad_norm": 0.18950824439525604, + "learning_rate": 2.4659793814432992e-05, + "loss": 0.586, + "step": 4403 + }, + { + "epoch": 1.4092799999999999, + "grad_norm": 0.17583727836608887, + "learning_rate": 2.461855670103093e-05, + "loss": 0.6413, + "step": 4404 + }, + { + "epoch": 1.4096, + "grad_norm": 0.17908890545368195, + "learning_rate": 2.4577319587628868e-05, + "loss": 0.7364, + "step": 4405 + }, + { + "epoch": 1.40992, + "grad_norm": 0.1721504032611847, + "learning_rate": 2.4536082474226803e-05, + "loss": 0.7526, + "step": 4406 + }, + { + "epoch": 1.41024, + "grad_norm": 0.17353549599647522, + "learning_rate": 2.4494845360824744e-05, + "loss": 0.6148, + "step": 4407 + }, + { + "epoch": 1.41056, + "grad_norm": 0.18320727348327637, + "learning_rate": 2.4453608247422682e-05, + "loss": 0.5361, + "step": 4408 + }, + { + "epoch": 1.4108800000000001, + "grad_norm": 0.1586429625749588, + "learning_rate": 2.441237113402062e-05, + "loss": 0.6576, + "step": 4409 + }, + { + "epoch": 1.4112, + "grad_norm": 0.177444726228714, + "learning_rate": 2.4371134020618558e-05, + "loss": 0.6701, + "step": 4410 + }, + { + "epoch": 1.4115199999999999, + "grad_norm": 0.1882229596376419, + "learning_rate": 2.4329896907216496e-05, + "loss": 0.5861, + "step": 4411 + }, + { + "epoch": 1.41184, + "grad_norm": 0.16356085240840912, + "learning_rate": 2.4288659793814434e-05, + "loss": 0.7565, + "step": 4412 + }, + { + "epoch": 1.41216, + "grad_norm": 0.20509935915470123, + "learning_rate": 2.4247422680412372e-05, + "loss": 0.5959, + "step": 4413 + }, + { + "epoch": 1.41248, + "grad_norm": 0.18892300128936768, + "learning_rate": 2.420618556701031e-05, + "loss": 0.5699, + "step": 4414 + }, + { + "epoch": 1.4128, + "grad_norm": 0.1533893197774887, + "learning_rate": 2.416494845360825e-05, + "loss": 0.7096, + "step": 4415 + }, + { + "epoch": 1.41312, + "grad_norm": 0.1792561262845993, + "learning_rate": 2.4123711340206186e-05, + "loss": 0.581, + "step": 4416 + }, + { + "epoch": 1.41344, + "grad_norm": 0.16917340457439423, + "learning_rate": 2.4082474226804125e-05, + "loss": 0.6276, + "step": 4417 + }, + { + "epoch": 1.41376, + "grad_norm": 0.20043566823005676, + "learning_rate": 2.4041237113402063e-05, + "loss": 0.6087, + "step": 4418 + }, + { + "epoch": 1.41408, + "grad_norm": 0.17254877090454102, + "learning_rate": 2.4e-05, + "loss": 0.4671, + "step": 4419 + }, + { + "epoch": 1.4144, + "grad_norm": 0.17740890383720398, + "learning_rate": 2.395876288659794e-05, + "loss": 0.6065, + "step": 4420 + }, + { + "epoch": 1.41472, + "grad_norm": 0.18336673080921173, + "learning_rate": 2.391752577319588e-05, + "loss": 0.493, + "step": 4421 + }, + { + "epoch": 1.41504, + "grad_norm": 0.18442244827747345, + "learning_rate": 2.3876288659793815e-05, + "loss": 0.6338, + "step": 4422 + }, + { + "epoch": 1.41536, + "grad_norm": 0.18067099153995514, + "learning_rate": 2.3835051546391753e-05, + "loss": 0.6396, + "step": 4423 + }, + { + "epoch": 1.41568, + "grad_norm": 0.17792640626430511, + "learning_rate": 2.379381443298969e-05, + "loss": 0.5087, + "step": 4424 + }, + { + "epoch": 1.416, + "grad_norm": 0.21553468704223633, + "learning_rate": 2.375257731958763e-05, + "loss": 0.5801, + "step": 4425 + }, + { + "epoch": 1.41632, + "grad_norm": 0.18563038110733032, + "learning_rate": 2.3711340206185567e-05, + "loss": 0.5339, + "step": 4426 + }, + { + "epoch": 1.4166400000000001, + "grad_norm": 0.1411089301109314, + "learning_rate": 2.3670103092783505e-05, + "loss": 0.4637, + "step": 4427 + }, + { + "epoch": 1.41696, + "grad_norm": 0.16860581934452057, + "learning_rate": 2.3628865979381446e-05, + "loss": 0.575, + "step": 4428 + }, + { + "epoch": 1.4172799999999999, + "grad_norm": 0.16433651745319366, + "learning_rate": 2.3587628865979384e-05, + "loss": 0.6054, + "step": 4429 + }, + { + "epoch": 1.4176, + "grad_norm": 0.14408190548419952, + "learning_rate": 2.3546391752577323e-05, + "loss": 0.6063, + "step": 4430 + }, + { + "epoch": 1.41792, + "grad_norm": 0.1569800078868866, + "learning_rate": 2.3505154639175257e-05, + "loss": 0.6218, + "step": 4431 + }, + { + "epoch": 1.41824, + "grad_norm": 0.18279337882995605, + "learning_rate": 2.3463917525773195e-05, + "loss": 0.6847, + "step": 4432 + }, + { + "epoch": 1.41856, + "grad_norm": 0.1807945966720581, + "learning_rate": 2.3422680412371133e-05, + "loss": 0.6078, + "step": 4433 + }, + { + "epoch": 1.41888, + "grad_norm": 0.16426122188568115, + "learning_rate": 2.3381443298969075e-05, + "loss": 0.5119, + "step": 4434 + }, + { + "epoch": 1.4192, + "grad_norm": 0.17681321501731873, + "learning_rate": 2.3340206185567013e-05, + "loss": 0.625, + "step": 4435 + }, + { + "epoch": 1.41952, + "grad_norm": 0.17861130833625793, + "learning_rate": 2.329896907216495e-05, + "loss": 0.5518, + "step": 4436 + }, + { + "epoch": 1.41984, + "grad_norm": 0.16329824924468994, + "learning_rate": 2.325773195876289e-05, + "loss": 0.5402, + "step": 4437 + }, + { + "epoch": 1.42016, + "grad_norm": 0.1864185333251953, + "learning_rate": 2.3216494845360827e-05, + "loss": 0.5356, + "step": 4438 + }, + { + "epoch": 1.42048, + "grad_norm": 0.17507138848304749, + "learning_rate": 2.317525773195876e-05, + "loss": 0.6345, + "step": 4439 + }, + { + "epoch": 1.4208, + "grad_norm": 0.16646729409694672, + "learning_rate": 2.31340206185567e-05, + "loss": 0.6998, + "step": 4440 + }, + { + "epoch": 1.42112, + "grad_norm": 0.1759895533323288, + "learning_rate": 2.309278350515464e-05, + "loss": 0.5587, + "step": 4441 + }, + { + "epoch": 1.42144, + "grad_norm": 0.175198033452034, + "learning_rate": 2.305154639175258e-05, + "loss": 0.5139, + "step": 4442 + }, + { + "epoch": 1.42176, + "grad_norm": 0.17822480201721191, + "learning_rate": 2.3010309278350517e-05, + "loss": 0.6774, + "step": 4443 + }, + { + "epoch": 1.42208, + "grad_norm": 0.15383930504322052, + "learning_rate": 2.2969072164948455e-05, + "loss": 0.6864, + "step": 4444 + }, + { + "epoch": 1.4224, + "grad_norm": 0.17063893377780914, + "learning_rate": 2.2927835051546393e-05, + "loss": 0.6342, + "step": 4445 + }, + { + "epoch": 1.42272, + "grad_norm": 0.1612689048051834, + "learning_rate": 2.288659793814433e-05, + "loss": 0.4439, + "step": 4446 + }, + { + "epoch": 1.42304, + "grad_norm": 0.1724826991558075, + "learning_rate": 2.284536082474227e-05, + "loss": 0.5141, + "step": 4447 + }, + { + "epoch": 1.42336, + "grad_norm": 0.19809328019618988, + "learning_rate": 2.2804123711340207e-05, + "loss": 0.6443, + "step": 4448 + }, + { + "epoch": 1.42368, + "grad_norm": 0.1949416995048523, + "learning_rate": 2.2762886597938146e-05, + "loss": 0.6113, + "step": 4449 + }, + { + "epoch": 1.424, + "grad_norm": 0.15237440168857574, + "learning_rate": 2.2721649484536084e-05, + "loss": 0.6185, + "step": 4450 + }, + { + "epoch": 1.42432, + "grad_norm": 0.1798660010099411, + "learning_rate": 2.268041237113402e-05, + "loss": 0.6288, + "step": 4451 + }, + { + "epoch": 1.4246400000000001, + "grad_norm": 0.17536713182926178, + "learning_rate": 2.263917525773196e-05, + "loss": 0.6291, + "step": 4452 + }, + { + "epoch": 1.42496, + "grad_norm": 0.16868086159229279, + "learning_rate": 2.2597938144329898e-05, + "loss": 0.7726, + "step": 4453 + }, + { + "epoch": 1.4252799999999999, + "grad_norm": 0.1751038134098053, + "learning_rate": 2.2556701030927836e-05, + "loss": 0.762, + "step": 4454 + }, + { + "epoch": 1.4256, + "grad_norm": 0.1755775511264801, + "learning_rate": 2.2515463917525774e-05, + "loss": 0.7294, + "step": 4455 + }, + { + "epoch": 1.42592, + "grad_norm": 0.1897972971200943, + "learning_rate": 2.2474226804123712e-05, + "loss": 0.6359, + "step": 4456 + }, + { + "epoch": 1.42624, + "grad_norm": 0.16473814845085144, + "learning_rate": 2.243298969072165e-05, + "loss": 0.6347, + "step": 4457 + }, + { + "epoch": 1.42656, + "grad_norm": 0.17288616299629211, + "learning_rate": 2.2391752577319588e-05, + "loss": 0.5785, + "step": 4458 + }, + { + "epoch": 1.42688, + "grad_norm": 0.15073929727077484, + "learning_rate": 2.2350515463917526e-05, + "loss": 0.577, + "step": 4459 + }, + { + "epoch": 1.4272, + "grad_norm": 0.17179200053215027, + "learning_rate": 2.2309278350515464e-05, + "loss": 0.6072, + "step": 4460 + }, + { + "epoch": 1.42752, + "grad_norm": 0.16867785155773163, + "learning_rate": 2.2268041237113405e-05, + "loss": 0.672, + "step": 4461 + }, + { + "epoch": 1.42784, + "grad_norm": 0.17833711206912994, + "learning_rate": 2.2226804123711344e-05, + "loss": 0.8226, + "step": 4462 + }, + { + "epoch": 1.42816, + "grad_norm": 0.16898909211158752, + "learning_rate": 2.218556701030928e-05, + "loss": 0.6202, + "step": 4463 + }, + { + "epoch": 1.42848, + "grad_norm": 0.17136110365390778, + "learning_rate": 2.2144329896907216e-05, + "loss": 0.6826, + "step": 4464 + }, + { + "epoch": 1.4288, + "grad_norm": 0.14478296041488647, + "learning_rate": 2.2103092783505154e-05, + "loss": 0.4986, + "step": 4465 + }, + { + "epoch": 1.42912, + "grad_norm": 0.198905810713768, + "learning_rate": 2.2061855670103092e-05, + "loss": 0.5845, + "step": 4466 + }, + { + "epoch": 1.42944, + "grad_norm": 0.17724497616291046, + "learning_rate": 2.202061855670103e-05, + "loss": 0.5323, + "step": 4467 + }, + { + "epoch": 1.42976, + "grad_norm": 0.18807314336299896, + "learning_rate": 2.1979381443298972e-05, + "loss": 0.646, + "step": 4468 + }, + { + "epoch": 1.43008, + "grad_norm": 0.16043980419635773, + "learning_rate": 2.193814432989691e-05, + "loss": 0.6827, + "step": 4469 + }, + { + "epoch": 1.4304000000000001, + "grad_norm": 0.16378290951251984, + "learning_rate": 2.1896907216494848e-05, + "loss": 0.5164, + "step": 4470 + }, + { + "epoch": 1.43072, + "grad_norm": 0.17819416522979736, + "learning_rate": 2.1855670103092786e-05, + "loss": 0.6651, + "step": 4471 + }, + { + "epoch": 1.4310399999999999, + "grad_norm": 0.1797819286584854, + "learning_rate": 2.1814432989690724e-05, + "loss": 0.5612, + "step": 4472 + }, + { + "epoch": 1.43136, + "grad_norm": 0.17763163149356842, + "learning_rate": 2.177319587628866e-05, + "loss": 0.7294, + "step": 4473 + }, + { + "epoch": 1.43168, + "grad_norm": 0.1586121767759323, + "learning_rate": 2.1731958762886597e-05, + "loss": 0.671, + "step": 4474 + }, + { + "epoch": 1.432, + "grad_norm": 0.16610221564769745, + "learning_rate": 2.1690721649484538e-05, + "loss": 0.7883, + "step": 4475 + }, + { + "epoch": 1.43232, + "grad_norm": 0.16714385151863098, + "learning_rate": 2.1649484536082476e-05, + "loss": 0.7972, + "step": 4476 + }, + { + "epoch": 1.4326400000000001, + "grad_norm": 0.18851213157176971, + "learning_rate": 2.1608247422680414e-05, + "loss": 0.6342, + "step": 4477 + }, + { + "epoch": 1.43296, + "grad_norm": 0.1893315315246582, + "learning_rate": 2.1567010309278352e-05, + "loss": 0.5787, + "step": 4478 + }, + { + "epoch": 1.4332799999999999, + "grad_norm": 0.19839134812355042, + "learning_rate": 2.152577319587629e-05, + "loss": 0.5543, + "step": 4479 + }, + { + "epoch": 1.4336, + "grad_norm": 0.18004603683948517, + "learning_rate": 2.148453608247423e-05, + "loss": 0.5599, + "step": 4480 + }, + { + "epoch": 1.43392, + "grad_norm": 0.15089333057403564, + "learning_rate": 2.1443298969072163e-05, + "loss": 0.58, + "step": 4481 + }, + { + "epoch": 1.43424, + "grad_norm": 0.192101389169693, + "learning_rate": 2.1402061855670105e-05, + "loss": 0.6552, + "step": 4482 + }, + { + "epoch": 1.43456, + "grad_norm": 0.16039416193962097, + "learning_rate": 2.1360824742268043e-05, + "loss": 0.6481, + "step": 4483 + }, + { + "epoch": 1.43488, + "grad_norm": 0.17155155539512634, + "learning_rate": 2.131958762886598e-05, + "loss": 0.7602, + "step": 4484 + }, + { + "epoch": 1.4352, + "grad_norm": 0.16525647044181824, + "learning_rate": 2.127835051546392e-05, + "loss": 0.5369, + "step": 4485 + }, + { + "epoch": 1.43552, + "grad_norm": 0.18237461149692535, + "learning_rate": 2.1237113402061857e-05, + "loss": 0.5984, + "step": 4486 + }, + { + "epoch": 1.43584, + "grad_norm": 0.157010018825531, + "learning_rate": 2.1195876288659795e-05, + "loss": 0.6702, + "step": 4487 + }, + { + "epoch": 1.43616, + "grad_norm": 0.20875589549541473, + "learning_rate": 2.1154639175257733e-05, + "loss": 0.632, + "step": 4488 + }, + { + "epoch": 1.43648, + "grad_norm": 0.161864772439003, + "learning_rate": 2.111340206185567e-05, + "loss": 0.5965, + "step": 4489 + }, + { + "epoch": 1.4368, + "grad_norm": 0.16834796965122223, + "learning_rate": 2.107216494845361e-05, + "loss": 0.6823, + "step": 4490 + }, + { + "epoch": 1.43712, + "grad_norm": 0.16008593142032623, + "learning_rate": 2.1030927835051547e-05, + "loss": 0.5663, + "step": 4491 + }, + { + "epoch": 1.43744, + "grad_norm": 0.15777644515037537, + "learning_rate": 2.0989690721649485e-05, + "loss": 0.4818, + "step": 4492 + }, + { + "epoch": 1.43776, + "grad_norm": 0.16782741248607635, + "learning_rate": 2.0948453608247423e-05, + "loss": 0.6509, + "step": 4493 + }, + { + "epoch": 1.43808, + "grad_norm": 0.17786957323551178, + "learning_rate": 2.090721649484536e-05, + "loss": 0.4979, + "step": 4494 + }, + { + "epoch": 1.4384000000000001, + "grad_norm": 0.17591677606105804, + "learning_rate": 2.0865979381443303e-05, + "loss": 0.5888, + "step": 4495 + }, + { + "epoch": 1.43872, + "grad_norm": 0.1644124835729599, + "learning_rate": 2.082474226804124e-05, + "loss": 0.5142, + "step": 4496 + }, + { + "epoch": 1.4390399999999999, + "grad_norm": 0.15530361235141754, + "learning_rate": 2.0783505154639175e-05, + "loss": 0.5663, + "step": 4497 + }, + { + "epoch": 1.43936, + "grad_norm": 0.2031339854001999, + "learning_rate": 2.0742268041237113e-05, + "loss": 0.5611, + "step": 4498 + }, + { + "epoch": 1.43968, + "grad_norm": 0.16557371616363525, + "learning_rate": 2.070103092783505e-05, + "loss": 0.6625, + "step": 4499 + }, + { + "epoch": 1.44, + "grad_norm": 0.19209522008895874, + "learning_rate": 2.065979381443299e-05, + "loss": 0.6415, + "step": 4500 + }, + { + "epoch": 1.44032, + "grad_norm": 0.17193055152893066, + "learning_rate": 2.0618556701030927e-05, + "loss": 0.5731, + "step": 4501 + }, + { + "epoch": 1.44064, + "grad_norm": 0.1839675009250641, + "learning_rate": 2.057731958762887e-05, + "loss": 0.5857, + "step": 4502 + }, + { + "epoch": 1.44096, + "grad_norm": 0.193007692694664, + "learning_rate": 2.0536082474226807e-05, + "loss": 0.7039, + "step": 4503 + }, + { + "epoch": 1.44128, + "grad_norm": 0.18779848515987396, + "learning_rate": 2.0494845360824745e-05, + "loss": 0.6407, + "step": 4504 + }, + { + "epoch": 1.4416, + "grad_norm": 0.16336211562156677, + "learning_rate": 2.0453608247422683e-05, + "loss": 0.5827, + "step": 4505 + }, + { + "epoch": 1.44192, + "grad_norm": 0.16009855270385742, + "learning_rate": 2.0412371134020618e-05, + "loss": 0.5179, + "step": 4506 + }, + { + "epoch": 1.44224, + "grad_norm": 0.16804586350917816, + "learning_rate": 2.0371134020618556e-05, + "loss": 0.5331, + "step": 4507 + }, + { + "epoch": 1.44256, + "grad_norm": 0.15751104056835175, + "learning_rate": 2.0329896907216494e-05, + "loss": 0.7124, + "step": 4508 + }, + { + "epoch": 1.44288, + "grad_norm": 0.16373498737812042, + "learning_rate": 2.0288659793814435e-05, + "loss": 0.6036, + "step": 4509 + }, + { + "epoch": 1.4432, + "grad_norm": 0.19661669433116913, + "learning_rate": 2.0247422680412373e-05, + "loss": 0.6597, + "step": 4510 + }, + { + "epoch": 1.44352, + "grad_norm": 0.17094755172729492, + "learning_rate": 2.020618556701031e-05, + "loss": 0.6058, + "step": 4511 + }, + { + "epoch": 1.44384, + "grad_norm": 0.1878400593996048, + "learning_rate": 2.016494845360825e-05, + "loss": 0.5012, + "step": 4512 + }, + { + "epoch": 1.44416, + "grad_norm": 0.15647532045841217, + "learning_rate": 2.0123711340206187e-05, + "loss": 0.4984, + "step": 4513 + }, + { + "epoch": 1.44448, + "grad_norm": 0.1727263480424881, + "learning_rate": 2.0082474226804122e-05, + "loss": 0.6056, + "step": 4514 + }, + { + "epoch": 1.4447999999999999, + "grad_norm": 0.15227076411247253, + "learning_rate": 2.004123711340206e-05, + "loss": 0.5922, + "step": 4515 + }, + { + "epoch": 1.44512, + "grad_norm": 0.20009352266788483, + "learning_rate": 2e-05, + "loss": 0.5527, + "step": 4516 + }, + { + "epoch": 1.44544, + "grad_norm": 0.1784680038690567, + "learning_rate": 1.995876288659794e-05, + "loss": 0.516, + "step": 4517 + }, + { + "epoch": 1.44576, + "grad_norm": 0.16646035015583038, + "learning_rate": 1.9917525773195878e-05, + "loss": 0.4856, + "step": 4518 + }, + { + "epoch": 1.44608, + "grad_norm": 0.18773174285888672, + "learning_rate": 1.9876288659793816e-05, + "loss": 0.716, + "step": 4519 + }, + { + "epoch": 1.4464000000000001, + "grad_norm": 0.18832553923130035, + "learning_rate": 1.9835051546391754e-05, + "loss": 0.7538, + "step": 4520 + }, + { + "epoch": 1.44672, + "grad_norm": 0.16927365958690643, + "learning_rate": 1.9793814432989692e-05, + "loss": 0.6419, + "step": 4521 + }, + { + "epoch": 1.4470399999999999, + "grad_norm": 0.17106406390666962, + "learning_rate": 1.975257731958763e-05, + "loss": 0.6202, + "step": 4522 + }, + { + "epoch": 1.44736, + "grad_norm": 0.15740400552749634, + "learning_rate": 1.9711340206185568e-05, + "loss": 0.6307, + "step": 4523 + }, + { + "epoch": 1.44768, + "grad_norm": 0.17782945930957794, + "learning_rate": 1.9670103092783506e-05, + "loss": 0.6115, + "step": 4524 + }, + { + "epoch": 1.448, + "grad_norm": 0.16445596516132355, + "learning_rate": 1.9628865979381444e-05, + "loss": 0.5, + "step": 4525 + }, + { + "epoch": 1.44832, + "grad_norm": 0.16315779089927673, + "learning_rate": 1.9587628865979382e-05, + "loss": 0.7754, + "step": 4526 + }, + { + "epoch": 1.44864, + "grad_norm": 0.172194242477417, + "learning_rate": 1.954639175257732e-05, + "loss": 0.5375, + "step": 4527 + }, + { + "epoch": 1.44896, + "grad_norm": 0.17178206145763397, + "learning_rate": 1.9505154639175258e-05, + "loss": 0.6747, + "step": 4528 + }, + { + "epoch": 1.44928, + "grad_norm": 0.21615272760391235, + "learning_rate": 1.94639175257732e-05, + "loss": 0.6677, + "step": 4529 + }, + { + "epoch": 1.4496, + "grad_norm": 0.16820017993450165, + "learning_rate": 1.9422680412371134e-05, + "loss": 0.6123, + "step": 4530 + }, + { + "epoch": 1.44992, + "grad_norm": 0.20010846853256226, + "learning_rate": 1.9381443298969072e-05, + "loss": 0.697, + "step": 4531 + }, + { + "epoch": 1.45024, + "grad_norm": 0.17667677998542786, + "learning_rate": 1.934020618556701e-05, + "loss": 0.759, + "step": 4532 + }, + { + "epoch": 1.45056, + "grad_norm": 0.15165181457996368, + "learning_rate": 1.929896907216495e-05, + "loss": 0.5463, + "step": 4533 + }, + { + "epoch": 1.45088, + "grad_norm": 0.1865624338388443, + "learning_rate": 1.9257731958762887e-05, + "loss": 0.7551, + "step": 4534 + }, + { + "epoch": 1.4512, + "grad_norm": 0.16454720497131348, + "learning_rate": 1.9216494845360825e-05, + "loss": 0.7221, + "step": 4535 + }, + { + "epoch": 1.45152, + "grad_norm": 0.16546441614627838, + "learning_rate": 1.9175257731958766e-05, + "loss": 0.5775, + "step": 4536 + }, + { + "epoch": 1.45184, + "grad_norm": 0.1854339987039566, + "learning_rate": 1.9134020618556704e-05, + "loss": 0.5673, + "step": 4537 + }, + { + "epoch": 1.4521600000000001, + "grad_norm": 0.16372042894363403, + "learning_rate": 1.9092783505154642e-05, + "loss": 0.6427, + "step": 4538 + }, + { + "epoch": 1.45248, + "grad_norm": 0.18942777812480927, + "learning_rate": 1.9051546391752577e-05, + "loss": 0.5009, + "step": 4539 + }, + { + "epoch": 1.4527999999999999, + "grad_norm": 0.19089898467063904, + "learning_rate": 1.9010309278350515e-05, + "loss": 0.7089, + "step": 4540 + }, + { + "epoch": 1.45312, + "grad_norm": 0.1639929860830307, + "learning_rate": 1.8969072164948453e-05, + "loss": 0.6327, + "step": 4541 + }, + { + "epoch": 1.45344, + "grad_norm": 0.1806475669145584, + "learning_rate": 1.892783505154639e-05, + "loss": 0.563, + "step": 4542 + }, + { + "epoch": 1.45376, + "grad_norm": 0.163899227976799, + "learning_rate": 1.8886597938144332e-05, + "loss": 0.7248, + "step": 4543 + }, + { + "epoch": 1.45408, + "grad_norm": 0.18689587712287903, + "learning_rate": 1.884536082474227e-05, + "loss": 0.6725, + "step": 4544 + }, + { + "epoch": 1.4544000000000001, + "grad_norm": 0.17408332228660583, + "learning_rate": 1.880412371134021e-05, + "loss": 0.6239, + "step": 4545 + }, + { + "epoch": 1.45472, + "grad_norm": 0.18414701521396637, + "learning_rate": 1.8762886597938147e-05, + "loss": 0.6478, + "step": 4546 + }, + { + "epoch": 1.45504, + "grad_norm": 0.1769532561302185, + "learning_rate": 1.8721649484536085e-05, + "loss": 0.5683, + "step": 4547 + }, + { + "epoch": 1.45536, + "grad_norm": 0.17712929844856262, + "learning_rate": 1.868041237113402e-05, + "loss": 0.572, + "step": 4548 + }, + { + "epoch": 1.45568, + "grad_norm": 0.18141962587833405, + "learning_rate": 1.8639175257731957e-05, + "loss": 0.6215, + "step": 4549 + }, + { + "epoch": 1.456, + "grad_norm": 0.17753002047538757, + "learning_rate": 1.85979381443299e-05, + "loss": 0.6366, + "step": 4550 + }, + { + "epoch": 1.45632, + "grad_norm": 0.2035723477602005, + "learning_rate": 1.8556701030927837e-05, + "loss": 0.6388, + "step": 4551 + }, + { + "epoch": 1.45664, + "grad_norm": 0.16969697177410126, + "learning_rate": 1.8515463917525775e-05, + "loss": 0.6073, + "step": 4552 + }, + { + "epoch": 1.45696, + "grad_norm": 0.18247532844543457, + "learning_rate": 1.8474226804123713e-05, + "loss": 0.4628, + "step": 4553 + }, + { + "epoch": 1.45728, + "grad_norm": 0.1639695018529892, + "learning_rate": 1.843298969072165e-05, + "loss": 0.6223, + "step": 4554 + }, + { + "epoch": 1.4576, + "grad_norm": 0.15585893392562866, + "learning_rate": 1.839175257731959e-05, + "loss": 0.5191, + "step": 4555 + }, + { + "epoch": 1.45792, + "grad_norm": 0.16417180001735687, + "learning_rate": 1.8350515463917524e-05, + "loss": 0.7356, + "step": 4556 + }, + { + "epoch": 1.45824, + "grad_norm": 0.19836829602718353, + "learning_rate": 1.8309278350515465e-05, + "loss": 0.7012, + "step": 4557 + }, + { + "epoch": 1.45856, + "grad_norm": 0.17286843061447144, + "learning_rate": 1.8268041237113403e-05, + "loss": 0.6431, + "step": 4558 + }, + { + "epoch": 1.45888, + "grad_norm": 0.20046155154705048, + "learning_rate": 1.822680412371134e-05, + "loss": 0.6582, + "step": 4559 + }, + { + "epoch": 1.4592, + "grad_norm": 0.15706866979599, + "learning_rate": 1.818556701030928e-05, + "loss": 0.6532, + "step": 4560 + }, + { + "epoch": 1.45952, + "grad_norm": 0.19204123318195343, + "learning_rate": 1.8144329896907217e-05, + "loss": 0.6033, + "step": 4561 + }, + { + "epoch": 1.45984, + "grad_norm": 0.16414190828800201, + "learning_rate": 1.8103092783505155e-05, + "loss": 0.5518, + "step": 4562 + }, + { + "epoch": 1.4601600000000001, + "grad_norm": 0.16594646871089935, + "learning_rate": 1.8061855670103093e-05, + "loss": 0.5726, + "step": 4563 + }, + { + "epoch": 1.46048, + "grad_norm": 0.16067630052566528, + "learning_rate": 1.802061855670103e-05, + "loss": 0.6, + "step": 4564 + }, + { + "epoch": 1.4607999999999999, + "grad_norm": 0.16492028534412384, + "learning_rate": 1.797938144329897e-05, + "loss": 0.7366, + "step": 4565 + }, + { + "epoch": 1.46112, + "grad_norm": 0.16765184700489044, + "learning_rate": 1.7938144329896908e-05, + "loss": 0.6677, + "step": 4566 + }, + { + "epoch": 1.46144, + "grad_norm": 0.1718098223209381, + "learning_rate": 1.7896907216494846e-05, + "loss": 0.6503, + "step": 4567 + }, + { + "epoch": 1.46176, + "grad_norm": 0.17017945647239685, + "learning_rate": 1.7855670103092784e-05, + "loss": 0.7043, + "step": 4568 + }, + { + "epoch": 1.46208, + "grad_norm": 0.1680067926645279, + "learning_rate": 1.781443298969072e-05, + "loss": 0.5759, + "step": 4569 + }, + { + "epoch": 1.4624, + "grad_norm": 0.18614085018634796, + "learning_rate": 1.7773195876288663e-05, + "loss": 0.6201, + "step": 4570 + }, + { + "epoch": 1.46272, + "grad_norm": 0.16951598227024078, + "learning_rate": 1.77319587628866e-05, + "loss": 0.6858, + "step": 4571 + }, + { + "epoch": 1.46304, + "grad_norm": 0.16893792152404785, + "learning_rate": 1.7690721649484536e-05, + "loss": 0.6161, + "step": 4572 + }, + { + "epoch": 1.46336, + "grad_norm": 0.16981172561645508, + "learning_rate": 1.7649484536082474e-05, + "loss": 0.6922, + "step": 4573 + }, + { + "epoch": 1.46368, + "grad_norm": 0.18362779915332794, + "learning_rate": 1.7608247422680412e-05, + "loss": 0.5367, + "step": 4574 + }, + { + "epoch": 1.464, + "grad_norm": 0.16373388469219208, + "learning_rate": 1.756701030927835e-05, + "loss": 0.6176, + "step": 4575 + }, + { + "epoch": 1.46432, + "grad_norm": 0.18533435463905334, + "learning_rate": 1.7525773195876288e-05, + "loss": 0.5748, + "step": 4576 + }, + { + "epoch": 1.46464, + "grad_norm": 0.16817443072795868, + "learning_rate": 1.748453608247423e-05, + "loss": 0.649, + "step": 4577 + }, + { + "epoch": 1.46496, + "grad_norm": 0.19549067318439484, + "learning_rate": 1.7443298969072168e-05, + "loss": 0.5757, + "step": 4578 + }, + { + "epoch": 1.46528, + "grad_norm": 0.152230903506279, + "learning_rate": 1.7402061855670106e-05, + "loss": 0.5285, + "step": 4579 + }, + { + "epoch": 1.4656, + "grad_norm": 0.19168081879615784, + "learning_rate": 1.7360824742268044e-05, + "loss": 0.6085, + "step": 4580 + }, + { + "epoch": 1.4659200000000001, + "grad_norm": 0.1855437159538269, + "learning_rate": 1.7319587628865978e-05, + "loss": 0.5797, + "step": 4581 + }, + { + "epoch": 1.46624, + "grad_norm": 0.17283518612384796, + "learning_rate": 1.7278350515463916e-05, + "loss": 0.7065, + "step": 4582 + }, + { + "epoch": 1.4665599999999999, + "grad_norm": 0.1763613522052765, + "learning_rate": 1.7237113402061854e-05, + "loss": 0.4959, + "step": 4583 + }, + { + "epoch": 1.46688, + "grad_norm": 0.18272486329078674, + "learning_rate": 1.7195876288659796e-05, + "loss": 0.6203, + "step": 4584 + }, + { + "epoch": 1.4672, + "grad_norm": 0.1844801902770996, + "learning_rate": 1.7154639175257734e-05, + "loss": 0.637, + "step": 4585 + }, + { + "epoch": 1.46752, + "grad_norm": 0.20046298205852509, + "learning_rate": 1.7113402061855672e-05, + "loss": 0.5695, + "step": 4586 + }, + { + "epoch": 1.46784, + "grad_norm": 0.20062506198883057, + "learning_rate": 1.707216494845361e-05, + "loss": 0.6705, + "step": 4587 + }, + { + "epoch": 1.4681600000000001, + "grad_norm": 0.16499346494674683, + "learning_rate": 1.7030927835051548e-05, + "loss": 0.662, + "step": 4588 + }, + { + "epoch": 1.46848, + "grad_norm": 0.15878276526927948, + "learning_rate": 1.6989690721649486e-05, + "loss": 0.5002, + "step": 4589 + }, + { + "epoch": 1.4687999999999999, + "grad_norm": 0.18598951399326324, + "learning_rate": 1.694845360824742e-05, + "loss": 0.6034, + "step": 4590 + }, + { + "epoch": 1.46912, + "grad_norm": 0.18219804763793945, + "learning_rate": 1.6907216494845362e-05, + "loss": 0.635, + "step": 4591 + }, + { + "epoch": 1.46944, + "grad_norm": 0.19031621515750885, + "learning_rate": 1.68659793814433e-05, + "loss": 0.6682, + "step": 4592 + }, + { + "epoch": 1.46976, + "grad_norm": 0.18513929843902588, + "learning_rate": 1.6824742268041238e-05, + "loss": 0.6137, + "step": 4593 + }, + { + "epoch": 1.47008, + "grad_norm": 0.1707831770181656, + "learning_rate": 1.6783505154639176e-05, + "loss": 0.6812, + "step": 4594 + }, + { + "epoch": 1.4704, + "grad_norm": 0.14272406697273254, + "learning_rate": 1.6742268041237114e-05, + "loss": 0.6671, + "step": 4595 + }, + { + "epoch": 1.47072, + "grad_norm": 0.17156119644641876, + "learning_rate": 1.6701030927835052e-05, + "loss": 0.5932, + "step": 4596 + }, + { + "epoch": 1.47104, + "grad_norm": 0.15326081216335297, + "learning_rate": 1.665979381443299e-05, + "loss": 0.7048, + "step": 4597 + }, + { + "epoch": 1.47136, + "grad_norm": 0.15721282362937927, + "learning_rate": 1.661855670103093e-05, + "loss": 0.7421, + "step": 4598 + }, + { + "epoch": 1.47168, + "grad_norm": 0.1902402937412262, + "learning_rate": 1.6577319587628867e-05, + "loss": 0.5363, + "step": 4599 + }, + { + "epoch": 1.472, + "grad_norm": 0.17582277953624725, + "learning_rate": 1.6536082474226805e-05, + "loss": 0.6491, + "step": 4600 + }, + { + "epoch": 1.47232, + "grad_norm": 0.17665983736515045, + "learning_rate": 1.6494845360824743e-05, + "loss": 0.5526, + "step": 4601 + }, + { + "epoch": 1.47264, + "grad_norm": 0.20353327691555023, + "learning_rate": 1.645360824742268e-05, + "loss": 0.7934, + "step": 4602 + }, + { + "epoch": 1.47296, + "grad_norm": 0.19271573424339294, + "learning_rate": 1.641237113402062e-05, + "loss": 0.5649, + "step": 4603 + }, + { + "epoch": 1.47328, + "grad_norm": 0.17228847742080688, + "learning_rate": 1.637113402061856e-05, + "loss": 0.619, + "step": 4604 + }, + { + "epoch": 1.4736, + "grad_norm": 0.18383730947971344, + "learning_rate": 1.6329896907216495e-05, + "loss": 0.7272, + "step": 4605 + }, + { + "epoch": 1.4739200000000001, + "grad_norm": 0.14697887003421783, + "learning_rate": 1.6288659793814433e-05, + "loss": 0.7019, + "step": 4606 + }, + { + "epoch": 1.47424, + "grad_norm": 0.17559650540351868, + "learning_rate": 1.624742268041237e-05, + "loss": 0.5875, + "step": 4607 + }, + { + "epoch": 1.4745599999999999, + "grad_norm": 0.19883470237255096, + "learning_rate": 1.620618556701031e-05, + "loss": 0.5795, + "step": 4608 + }, + { + "epoch": 1.47488, + "grad_norm": 0.16464519500732422, + "learning_rate": 1.6164948453608247e-05, + "loss": 0.5222, + "step": 4609 + }, + { + "epoch": 1.4752, + "grad_norm": 0.17749834060668945, + "learning_rate": 1.6123711340206185e-05, + "loss": 0.5575, + "step": 4610 + }, + { + "epoch": 1.47552, + "grad_norm": 0.1572466641664505, + "learning_rate": 1.6082474226804127e-05, + "loss": 0.5722, + "step": 4611 + }, + { + "epoch": 1.47584, + "grad_norm": 0.1883009523153305, + "learning_rate": 1.6041237113402065e-05, + "loss": 0.6017, + "step": 4612 + }, + { + "epoch": 1.4761600000000001, + "grad_norm": 0.18180038034915924, + "learning_rate": 1.6000000000000003e-05, + "loss": 0.6317, + "step": 4613 + }, + { + "epoch": 1.47648, + "grad_norm": 0.17230986058712006, + "learning_rate": 1.5958762886597937e-05, + "loss": 0.7567, + "step": 4614 + }, + { + "epoch": 1.4768, + "grad_norm": 0.2114444524049759, + "learning_rate": 1.5917525773195875e-05, + "loss": 0.5315, + "step": 4615 + }, + { + "epoch": 1.47712, + "grad_norm": 0.18661636114120483, + "learning_rate": 1.5876288659793813e-05, + "loss": 0.6558, + "step": 4616 + }, + { + "epoch": 1.47744, + "grad_norm": 0.1765984743833542, + "learning_rate": 1.583505154639175e-05, + "loss": 0.6335, + "step": 4617 + }, + { + "epoch": 1.47776, + "grad_norm": 0.18617284297943115, + "learning_rate": 1.5793814432989693e-05, + "loss": 0.5798, + "step": 4618 + }, + { + "epoch": 1.47808, + "grad_norm": 0.18458376824855804, + "learning_rate": 1.575257731958763e-05, + "loss": 0.709, + "step": 4619 + }, + { + "epoch": 1.4784, + "grad_norm": 0.16929477453231812, + "learning_rate": 1.571134020618557e-05, + "loss": 0.5918, + "step": 4620 + }, + { + "epoch": 1.47872, + "grad_norm": 0.17071892321109772, + "learning_rate": 1.5670103092783507e-05, + "loss": 0.6978, + "step": 4621 + }, + { + "epoch": 1.47904, + "grad_norm": 0.18717169761657715, + "learning_rate": 1.5628865979381445e-05, + "loss": 0.7538, + "step": 4622 + }, + { + "epoch": 1.47936, + "grad_norm": 0.1651584953069687, + "learning_rate": 1.558762886597938e-05, + "loss": 0.5432, + "step": 4623 + }, + { + "epoch": 1.47968, + "grad_norm": 0.1736440360546112, + "learning_rate": 1.5546391752577318e-05, + "loss": 0.7363, + "step": 4624 + }, + { + "epoch": 1.48, + "grad_norm": 0.16841652989387512, + "learning_rate": 1.550515463917526e-05, + "loss": 0.6389, + "step": 4625 + }, + { + "epoch": 1.48032, + "grad_norm": 0.17438125610351562, + "learning_rate": 1.5463917525773197e-05, + "loss": 0.6808, + "step": 4626 + }, + { + "epoch": 1.48064, + "grad_norm": 0.15965159237384796, + "learning_rate": 1.5422680412371135e-05, + "loss": 0.6287, + "step": 4627 + }, + { + "epoch": 1.48096, + "grad_norm": 0.19508816301822662, + "learning_rate": 1.5381443298969073e-05, + "loss": 0.5811, + "step": 4628 + }, + { + "epoch": 1.48128, + "grad_norm": 0.18179328739643097, + "learning_rate": 1.534020618556701e-05, + "loss": 0.6461, + "step": 4629 + }, + { + "epoch": 1.4816, + "grad_norm": 0.19117939472198486, + "learning_rate": 1.529896907216495e-05, + "loss": 0.6947, + "step": 4630 + }, + { + "epoch": 1.4819200000000001, + "grad_norm": 0.18652775883674622, + "learning_rate": 1.525773195876289e-05, + "loss": 0.6051, + "step": 4631 + }, + { + "epoch": 1.48224, + "grad_norm": 0.188206285238266, + "learning_rate": 1.5216494845360826e-05, + "loss": 0.5966, + "step": 4632 + }, + { + "epoch": 1.4825599999999999, + "grad_norm": 0.17399485409259796, + "learning_rate": 1.5175257731958764e-05, + "loss": 0.4764, + "step": 4633 + }, + { + "epoch": 1.48288, + "grad_norm": 0.16545790433883667, + "learning_rate": 1.5134020618556702e-05, + "loss": 0.5795, + "step": 4634 + }, + { + "epoch": 1.4832, + "grad_norm": 0.17916645109653473, + "learning_rate": 1.509278350515464e-05, + "loss": 0.5783, + "step": 4635 + }, + { + "epoch": 1.48352, + "grad_norm": 0.16165047883987427, + "learning_rate": 1.5051546391752578e-05, + "loss": 0.464, + "step": 4636 + }, + { + "epoch": 1.48384, + "grad_norm": 0.17863832414150238, + "learning_rate": 1.5010309278350516e-05, + "loss": 0.6447, + "step": 4637 + }, + { + "epoch": 1.48416, + "grad_norm": 0.1833624541759491, + "learning_rate": 1.4969072164948456e-05, + "loss": 0.7273, + "step": 4638 + }, + { + "epoch": 1.48448, + "grad_norm": 0.15523645281791687, + "learning_rate": 1.4927835051546394e-05, + "loss": 0.6936, + "step": 4639 + }, + { + "epoch": 1.4848, + "grad_norm": 0.15162844955921173, + "learning_rate": 1.4886597938144332e-05, + "loss": 0.5723, + "step": 4640 + }, + { + "epoch": 1.48512, + "grad_norm": 0.18488742411136627, + "learning_rate": 1.4845360824742268e-05, + "loss": 0.513, + "step": 4641 + }, + { + "epoch": 1.48544, + "grad_norm": 0.18891602754592896, + "learning_rate": 1.4804123711340206e-05, + "loss": 0.5646, + "step": 4642 + }, + { + "epoch": 1.48576, + "grad_norm": 0.16917508840560913, + "learning_rate": 1.4762886597938144e-05, + "loss": 0.6922, + "step": 4643 + }, + { + "epoch": 1.48608, + "grad_norm": 0.17949147522449493, + "learning_rate": 1.4721649484536082e-05, + "loss": 0.5753, + "step": 4644 + }, + { + "epoch": 1.4864, + "grad_norm": 0.16505208611488342, + "learning_rate": 1.4680412371134022e-05, + "loss": 0.5216, + "step": 4645 + }, + { + "epoch": 1.48672, + "grad_norm": 0.15745437145233154, + "learning_rate": 1.463917525773196e-05, + "loss": 0.7123, + "step": 4646 + }, + { + "epoch": 1.48704, + "grad_norm": 0.16634735465049744, + "learning_rate": 1.4597938144329898e-05, + "loss": 0.5325, + "step": 4647 + }, + { + "epoch": 1.48736, + "grad_norm": 0.17992018163204193, + "learning_rate": 1.4556701030927836e-05, + "loss": 0.7058, + "step": 4648 + }, + { + "epoch": 1.4876800000000001, + "grad_norm": 0.14916884899139404, + "learning_rate": 1.4515463917525774e-05, + "loss": 0.699, + "step": 4649 + }, + { + "epoch": 1.488, + "grad_norm": 0.17348222434520721, + "learning_rate": 1.447422680412371e-05, + "loss": 0.4929, + "step": 4650 + }, + { + "epoch": 1.4883199999999999, + "grad_norm": 0.15738362073898315, + "learning_rate": 1.4432989690721649e-05, + "loss": 0.7357, + "step": 4651 + }, + { + "epoch": 1.48864, + "grad_norm": 0.16833046078681946, + "learning_rate": 1.439175257731959e-05, + "loss": 0.5832, + "step": 4652 + }, + { + "epoch": 1.48896, + "grad_norm": 0.1806214153766632, + "learning_rate": 1.4350515463917526e-05, + "loss": 0.7366, + "step": 4653 + }, + { + "epoch": 1.48928, + "grad_norm": 0.17372943460941315, + "learning_rate": 1.4309278350515464e-05, + "loss": 0.6623, + "step": 4654 + }, + { + "epoch": 1.4896, + "grad_norm": 0.18022876977920532, + "learning_rate": 1.4268041237113402e-05, + "loss": 0.5859, + "step": 4655 + }, + { + "epoch": 1.4899200000000001, + "grad_norm": 0.1807183027267456, + "learning_rate": 1.422680412371134e-05, + "loss": 0.7046, + "step": 4656 + }, + { + "epoch": 1.49024, + "grad_norm": 0.14902786910533905, + "learning_rate": 1.4185567010309279e-05, + "loss": 0.7177, + "step": 4657 + }, + { + "epoch": 1.4905599999999999, + "grad_norm": 0.18070220947265625, + "learning_rate": 1.4144329896907218e-05, + "loss": 0.5772, + "step": 4658 + }, + { + "epoch": 1.49088, + "grad_norm": 0.1632976531982422, + "learning_rate": 1.4103092783505156e-05, + "loss": 0.656, + "step": 4659 + }, + { + "epoch": 1.4912, + "grad_norm": 0.2046787440776825, + "learning_rate": 1.4061855670103094e-05, + "loss": 0.6241, + "step": 4660 + }, + { + "epoch": 1.49152, + "grad_norm": 0.19503989815711975, + "learning_rate": 1.4020618556701032e-05, + "loss": 0.7821, + "step": 4661 + }, + { + "epoch": 1.49184, + "grad_norm": 0.18005701899528503, + "learning_rate": 1.3979381443298969e-05, + "loss": 0.6362, + "step": 4662 + }, + { + "epoch": 1.49216, + "grad_norm": 0.16249877214431763, + "learning_rate": 1.3938144329896907e-05, + "loss": 0.6157, + "step": 4663 + }, + { + "epoch": 1.49248, + "grad_norm": 0.17732185125350952, + "learning_rate": 1.3896907216494845e-05, + "loss": 0.5169, + "step": 4664 + }, + { + "epoch": 1.4928, + "grad_norm": 0.17700780928134918, + "learning_rate": 1.3855670103092785e-05, + "loss": 0.5849, + "step": 4665 + }, + { + "epoch": 1.49312, + "grad_norm": 0.19261586666107178, + "learning_rate": 1.3814432989690723e-05, + "loss": 0.6363, + "step": 4666 + }, + { + "epoch": 1.49344, + "grad_norm": 0.19220416247844696, + "learning_rate": 1.377319587628866e-05, + "loss": 0.5663, + "step": 4667 + }, + { + "epoch": 1.49376, + "grad_norm": 0.19623558223247528, + "learning_rate": 1.3731958762886599e-05, + "loss": 0.6074, + "step": 4668 + }, + { + "epoch": 1.49408, + "grad_norm": 0.1794913411140442, + "learning_rate": 1.3690721649484537e-05, + "loss": 0.5485, + "step": 4669 + }, + { + "epoch": 1.4944, + "grad_norm": 0.17053069174289703, + "learning_rate": 1.3649484536082475e-05, + "loss": 0.6694, + "step": 4670 + }, + { + "epoch": 1.49472, + "grad_norm": 0.19091373682022095, + "learning_rate": 1.3608247422680411e-05, + "loss": 0.5504, + "step": 4671 + }, + { + "epoch": 1.49504, + "grad_norm": 0.17620711028575897, + "learning_rate": 1.3567010309278353e-05, + "loss": 0.7379, + "step": 4672 + }, + { + "epoch": 1.49536, + "grad_norm": 0.1669468879699707, + "learning_rate": 1.352577319587629e-05, + "loss": 0.647, + "step": 4673 + }, + { + "epoch": 1.4956800000000001, + "grad_norm": 0.18525299429893494, + "learning_rate": 1.3484536082474227e-05, + "loss": 0.5763, + "step": 4674 + }, + { + "epoch": 1.496, + "grad_norm": 0.1792762130498886, + "learning_rate": 1.3443298969072165e-05, + "loss": 0.5375, + "step": 4675 + }, + { + "epoch": 1.4963199999999999, + "grad_norm": 0.1769266277551651, + "learning_rate": 1.3402061855670103e-05, + "loss": 0.509, + "step": 4676 + }, + { + "epoch": 1.49664, + "grad_norm": 0.1655147224664688, + "learning_rate": 1.3360824742268041e-05, + "loss": 0.5292, + "step": 4677 + }, + { + "epoch": 1.49696, + "grad_norm": 0.19241438806056976, + "learning_rate": 1.331958762886598e-05, + "loss": 0.6189, + "step": 4678 + }, + { + "epoch": 1.49728, + "grad_norm": 0.19850747287273407, + "learning_rate": 1.3278350515463919e-05, + "loss": 0.6349, + "step": 4679 + }, + { + "epoch": 1.4976, + "grad_norm": 0.19150324165821075, + "learning_rate": 1.3237113402061857e-05, + "loss": 0.6994, + "step": 4680 + }, + { + "epoch": 1.49792, + "grad_norm": 0.14939112961292267, + "learning_rate": 1.3195876288659795e-05, + "loss": 0.5804, + "step": 4681 + }, + { + "epoch": 1.49824, + "grad_norm": 0.19074268639087677, + "learning_rate": 1.3154639175257733e-05, + "loss": 0.651, + "step": 4682 + }, + { + "epoch": 1.49856, + "grad_norm": 0.18596892058849335, + "learning_rate": 1.311340206185567e-05, + "loss": 0.5666, + "step": 4683 + }, + { + "epoch": 1.49888, + "grad_norm": 0.19014668464660645, + "learning_rate": 1.3072164948453608e-05, + "loss": 0.6417, + "step": 4684 + }, + { + "epoch": 1.4992, + "grad_norm": 0.1826917678117752, + "learning_rate": 1.3030927835051546e-05, + "loss": 0.7029, + "step": 4685 + }, + { + "epoch": 1.49952, + "grad_norm": 0.18265700340270996, + "learning_rate": 1.2989690721649485e-05, + "loss": 0.699, + "step": 4686 + }, + { + "epoch": 1.49984, + "grad_norm": 0.16028939187526703, + "learning_rate": 1.2948453608247423e-05, + "loss": 0.6096, + "step": 4687 + }, + { + "epoch": 1.5001600000000002, + "grad_norm": 0.18243515491485596, + "learning_rate": 1.2907216494845362e-05, + "loss": 0.5324, + "step": 4688 + }, + { + "epoch": 1.50048, + "grad_norm": 0.14852313697338104, + "learning_rate": 1.28659793814433e-05, + "loss": 0.5771, + "step": 4689 + }, + { + "epoch": 1.5008, + "grad_norm": 0.1924453228712082, + "learning_rate": 1.2824742268041238e-05, + "loss": 0.6629, + "step": 4690 + }, + { + "epoch": 1.50112, + "grad_norm": 0.17784123122692108, + "learning_rate": 1.2783505154639176e-05, + "loss": 0.6583, + "step": 4691 + }, + { + "epoch": 1.50144, + "grad_norm": 0.20440368354320526, + "learning_rate": 1.2742268041237115e-05, + "loss": 0.7032, + "step": 4692 + }, + { + "epoch": 1.50176, + "grad_norm": 0.16986489295959473, + "learning_rate": 1.2701030927835053e-05, + "loss": 0.5513, + "step": 4693 + }, + { + "epoch": 1.5020799999999999, + "grad_norm": 0.1602695882320404, + "learning_rate": 1.2659793814432992e-05, + "loss": 0.5584, + "step": 4694 + }, + { + "epoch": 1.5024, + "grad_norm": 0.1735362857580185, + "learning_rate": 1.2618556701030928e-05, + "loss": 0.7323, + "step": 4695 + }, + { + "epoch": 1.50272, + "grad_norm": 0.18932658433914185, + "learning_rate": 1.2577319587628866e-05, + "loss": 0.5392, + "step": 4696 + }, + { + "epoch": 1.50304, + "grad_norm": 0.17110684514045715, + "learning_rate": 1.2536082474226804e-05, + "loss": 0.6041, + "step": 4697 + }, + { + "epoch": 1.50336, + "grad_norm": 0.18068088591098785, + "learning_rate": 1.2494845360824744e-05, + "loss": 0.7183, + "step": 4698 + }, + { + "epoch": 1.5036800000000001, + "grad_norm": 0.16365152597427368, + "learning_rate": 1.245360824742268e-05, + "loss": 0.5123, + "step": 4699 + }, + { + "epoch": 1.504, + "grad_norm": 0.16705627739429474, + "learning_rate": 1.2412371134020618e-05, + "loss": 0.5954, + "step": 4700 + }, + { + "epoch": 1.5043199999999999, + "grad_norm": 0.1735004484653473, + "learning_rate": 1.2371134020618558e-05, + "loss": 0.5713, + "step": 4701 + }, + { + "epoch": 1.50464, + "grad_norm": 0.1846713125705719, + "learning_rate": 1.2329896907216496e-05, + "loss": 0.5398, + "step": 4702 + }, + { + "epoch": 1.50496, + "grad_norm": 0.1814558506011963, + "learning_rate": 1.2288659793814434e-05, + "loss": 0.6437, + "step": 4703 + }, + { + "epoch": 1.50528, + "grad_norm": 0.1665477156639099, + "learning_rate": 1.2247422680412372e-05, + "loss": 0.5232, + "step": 4704 + }, + { + "epoch": 1.5056, + "grad_norm": 0.18824131786823273, + "learning_rate": 1.220618556701031e-05, + "loss": 0.5523, + "step": 4705 + }, + { + "epoch": 1.5059200000000001, + "grad_norm": 0.19317838549613953, + "learning_rate": 1.2164948453608248e-05, + "loss": 0.666, + "step": 4706 + }, + { + "epoch": 1.50624, + "grad_norm": 0.1785905808210373, + "learning_rate": 1.2123711340206186e-05, + "loss": 0.5433, + "step": 4707 + }, + { + "epoch": 1.50656, + "grad_norm": 0.17405672371387482, + "learning_rate": 1.2082474226804124e-05, + "loss": 0.5513, + "step": 4708 + }, + { + "epoch": 1.50688, + "grad_norm": 0.16743826866149902, + "learning_rate": 1.2041237113402062e-05, + "loss": 0.6245, + "step": 4709 + }, + { + "epoch": 1.5072, + "grad_norm": 0.18242061138153076, + "learning_rate": 1.2e-05, + "loss": 0.5851, + "step": 4710 + }, + { + "epoch": 1.50752, + "grad_norm": 0.16676275432109833, + "learning_rate": 1.195876288659794e-05, + "loss": 0.5667, + "step": 4711 + }, + { + "epoch": 1.5078399999999998, + "grad_norm": 0.17418786883354187, + "learning_rate": 1.1917525773195876e-05, + "loss": 0.6678, + "step": 4712 + }, + { + "epoch": 1.50816, + "grad_norm": 0.1886887401342392, + "learning_rate": 1.1876288659793814e-05, + "loss": 0.6135, + "step": 4713 + }, + { + "epoch": 1.50848, + "grad_norm": 0.15913134813308716, + "learning_rate": 1.1835051546391753e-05, + "loss": 0.4378, + "step": 4714 + }, + { + "epoch": 1.5088, + "grad_norm": 0.1763584017753601, + "learning_rate": 1.1793814432989692e-05, + "loss": 0.6366, + "step": 4715 + }, + { + "epoch": 1.50912, + "grad_norm": 0.17088782787322998, + "learning_rate": 1.1752577319587629e-05, + "loss": 0.681, + "step": 4716 + }, + { + "epoch": 1.5094400000000001, + "grad_norm": 0.16227391362190247, + "learning_rate": 1.1711340206185567e-05, + "loss": 0.5542, + "step": 4717 + }, + { + "epoch": 1.50976, + "grad_norm": 0.17143289744853973, + "learning_rate": 1.1670103092783506e-05, + "loss": 0.616, + "step": 4718 + }, + { + "epoch": 1.5100799999999999, + "grad_norm": 0.19086670875549316, + "learning_rate": 1.1628865979381444e-05, + "loss": 0.6082, + "step": 4719 + }, + { + "epoch": 1.5104, + "grad_norm": 0.1866482049226761, + "learning_rate": 1.158762886597938e-05, + "loss": 0.6053, + "step": 4720 + }, + { + "epoch": 1.51072, + "grad_norm": 0.1726406067609787, + "learning_rate": 1.154639175257732e-05, + "loss": 0.6854, + "step": 4721 + }, + { + "epoch": 1.51104, + "grad_norm": 0.195876806974411, + "learning_rate": 1.1505154639175259e-05, + "loss": 0.7168, + "step": 4722 + }, + { + "epoch": 1.51136, + "grad_norm": 0.18437542021274567, + "learning_rate": 1.1463917525773197e-05, + "loss": 0.6902, + "step": 4723 + }, + { + "epoch": 1.5116800000000001, + "grad_norm": 0.19087348878383636, + "learning_rate": 1.1422680412371135e-05, + "loss": 0.7063, + "step": 4724 + }, + { + "epoch": 1.512, + "grad_norm": 0.18250443041324615, + "learning_rate": 1.1381443298969073e-05, + "loss": 0.5951, + "step": 4725 + }, + { + "epoch": 1.5123199999999999, + "grad_norm": 0.1897394210100174, + "learning_rate": 1.134020618556701e-05, + "loss": 0.6629, + "step": 4726 + }, + { + "epoch": 1.51264, + "grad_norm": 0.18095068633556366, + "learning_rate": 1.1298969072164949e-05, + "loss": 0.6649, + "step": 4727 + }, + { + "epoch": 1.51296, + "grad_norm": 0.18501906096935272, + "learning_rate": 1.1257731958762887e-05, + "loss": 0.5851, + "step": 4728 + }, + { + "epoch": 1.51328, + "grad_norm": 0.198545902967453, + "learning_rate": 1.1216494845360825e-05, + "loss": 0.6545, + "step": 4729 + }, + { + "epoch": 1.5135999999999998, + "grad_norm": 0.16625016927719116, + "learning_rate": 1.1175257731958763e-05, + "loss": 0.584, + "step": 4730 + }, + { + "epoch": 1.5139200000000002, + "grad_norm": 0.14414826035499573, + "learning_rate": 1.1134020618556703e-05, + "loss": 0.6183, + "step": 4731 + }, + { + "epoch": 1.51424, + "grad_norm": 0.18143734335899353, + "learning_rate": 1.109278350515464e-05, + "loss": 0.6692, + "step": 4732 + }, + { + "epoch": 1.51456, + "grad_norm": 0.16903363168239594, + "learning_rate": 1.1051546391752577e-05, + "loss": 0.6657, + "step": 4733 + }, + { + "epoch": 1.51488, + "grad_norm": 0.1567782461643219, + "learning_rate": 1.1010309278350515e-05, + "loss": 0.4992, + "step": 4734 + }, + { + "epoch": 1.5152, + "grad_norm": 0.13358497619628906, + "learning_rate": 1.0969072164948455e-05, + "loss": 0.6245, + "step": 4735 + }, + { + "epoch": 1.51552, + "grad_norm": 0.16180114448070526, + "learning_rate": 1.0927835051546393e-05, + "loss": 0.6558, + "step": 4736 + }, + { + "epoch": 1.5158399999999999, + "grad_norm": 0.19241027534008026, + "learning_rate": 1.088659793814433e-05, + "loss": 0.6272, + "step": 4737 + }, + { + "epoch": 1.51616, + "grad_norm": 0.18042820692062378, + "learning_rate": 1.0845360824742269e-05, + "loss": 0.602, + "step": 4738 + }, + { + "epoch": 1.51648, + "grad_norm": 0.17353427410125732, + "learning_rate": 1.0804123711340207e-05, + "loss": 0.5975, + "step": 4739 + }, + { + "epoch": 1.5168, + "grad_norm": 0.17690590023994446, + "learning_rate": 1.0762886597938145e-05, + "loss": 0.6779, + "step": 4740 + }, + { + "epoch": 1.51712, + "grad_norm": 0.17293596267700195, + "learning_rate": 1.0721649484536082e-05, + "loss": 0.8002, + "step": 4741 + }, + { + "epoch": 1.5174400000000001, + "grad_norm": 0.18248236179351807, + "learning_rate": 1.0680412371134021e-05, + "loss": 0.6419, + "step": 4742 + }, + { + "epoch": 1.51776, + "grad_norm": 0.16799217462539673, + "learning_rate": 1.063917525773196e-05, + "loss": 0.6296, + "step": 4743 + }, + { + "epoch": 1.5180799999999999, + "grad_norm": 0.1524914652109146, + "learning_rate": 1.0597938144329897e-05, + "loss": 0.724, + "step": 4744 + }, + { + "epoch": 1.5184, + "grad_norm": 0.18851979076862335, + "learning_rate": 1.0556701030927835e-05, + "loss": 0.636, + "step": 4745 + }, + { + "epoch": 1.51872, + "grad_norm": 0.18869662284851074, + "learning_rate": 1.0515463917525773e-05, + "loss": 0.667, + "step": 4746 + }, + { + "epoch": 1.51904, + "grad_norm": 0.18190014362335205, + "learning_rate": 1.0474226804123712e-05, + "loss": 0.6888, + "step": 4747 + }, + { + "epoch": 1.51936, + "grad_norm": 0.1778111606836319, + "learning_rate": 1.0432989690721651e-05, + "loss": 0.5665, + "step": 4748 + }, + { + "epoch": 1.5196800000000001, + "grad_norm": 0.17537662386894226, + "learning_rate": 1.0391752577319588e-05, + "loss": 0.6557, + "step": 4749 + }, + { + "epoch": 1.52, + "grad_norm": 0.17495281994342804, + "learning_rate": 1.0350515463917526e-05, + "loss": 0.66, + "step": 4750 + } + ], + "logging_steps": 1, + "max_steps": 5000, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 250, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.0401059485953597e+18, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}