| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.0, |
| "eval_steps": 500, |
| "global_step": 252, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.003968253968253968, |
| "grad_norm": 0.8016456961631775, |
| "learning_rate": 7.692307692307694e-07, |
| "loss": 0.1572, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.007936507936507936, |
| "grad_norm": 0.5781773924827576, |
| "learning_rate": 1.5384615384615387e-06, |
| "loss": 0.0899, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.011904761904761904, |
| "grad_norm": 0.5898916721343994, |
| "learning_rate": 2.307692307692308e-06, |
| "loss": 0.0867, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.015873015873015872, |
| "grad_norm": 0.7932422161102295, |
| "learning_rate": 3.0769230769230774e-06, |
| "loss": 0.1254, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.01984126984126984, |
| "grad_norm": 0.8549998998641968, |
| "learning_rate": 3.846153846153847e-06, |
| "loss": 0.1682, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.023809523809523808, |
| "grad_norm": 0.6565646529197693, |
| "learning_rate": 4.615384615384616e-06, |
| "loss": 0.0863, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.027777777777777776, |
| "grad_norm": 0.6592453718185425, |
| "learning_rate": 5.384615384615385e-06, |
| "loss": 0.092, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.031746031746031744, |
| "grad_norm": 0.613888144493103, |
| "learning_rate": 6.153846153846155e-06, |
| "loss": 0.0893, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.03571428571428571, |
| "grad_norm": 0.7832797765731812, |
| "learning_rate": 6.923076923076923e-06, |
| "loss": 0.1094, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.03968253968253968, |
| "grad_norm": 0.6986587643623352, |
| "learning_rate": 7.692307692307694e-06, |
| "loss": 0.0972, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.04365079365079365, |
| "grad_norm": 0.7623270153999329, |
| "learning_rate": 8.461538461538462e-06, |
| "loss": 0.1219, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.047619047619047616, |
| "grad_norm": 0.7717706561088562, |
| "learning_rate": 9.230769230769232e-06, |
| "loss": 0.1111, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.051587301587301584, |
| "grad_norm": 0.6941069960594177, |
| "learning_rate": 1e-05, |
| "loss": 0.0913, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.05555555555555555, |
| "grad_norm": 0.9567614793777466, |
| "learning_rate": 9.999568045802216e-06, |
| "loss": 0.1244, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.05952380952380952, |
| "grad_norm": 0.9038094282150269, |
| "learning_rate": 9.99827225784264e-06, |
| "loss": 0.1261, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.06349206349206349, |
| "grad_norm": 0.942318320274353, |
| "learning_rate": 9.996112860009689e-06, |
| "loss": 0.1154, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.06746031746031746, |
| "grad_norm": 1.2963838577270508, |
| "learning_rate": 9.993090225407743e-06, |
| "loss": 0.1179, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.07142857142857142, |
| "grad_norm": 0.8918065428733826, |
| "learning_rate": 9.98920487629269e-06, |
| "loss": 0.1136, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.07539682539682539, |
| "grad_norm": 0.8215576410293579, |
| "learning_rate": 9.98445748398167e-06, |
| "loss": 0.0897, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.07936507936507936, |
| "grad_norm": 0.9790538549423218, |
| "learning_rate": 9.978848868737099e-06, |
| "loss": 0.0951, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.08333333333333333, |
| "grad_norm": 0.9144031405448914, |
| "learning_rate": 9.972379999624935e-06, |
| "loss": 0.1015, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.0873015873015873, |
| "grad_norm": 2.1270599365234375, |
| "learning_rate": 9.96505199434725e-06, |
| "loss": 0.1525, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.09126984126984126, |
| "grad_norm": 0.9955826997756958, |
| "learning_rate": 9.956866119049095e-06, |
| "loss": 0.1295, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.09523809523809523, |
| "grad_norm": 1.271159052848816, |
| "learning_rate": 9.947823788099754e-06, |
| "loss": 0.1149, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.0992063492063492, |
| "grad_norm": 0.8611926436424255, |
| "learning_rate": 9.937926563848345e-06, |
| "loss": 0.109, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.10317460317460317, |
| "grad_norm": 0.9697916507720947, |
| "learning_rate": 9.9271761563539e-06, |
| "loss": 0.1274, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.10714285714285714, |
| "grad_norm": 1.031341791152954, |
| "learning_rate": 9.915574423089872e-06, |
| "loss": 0.1389, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.1111111111111111, |
| "grad_norm": 0.9682615399360657, |
| "learning_rate": 9.903123368623216e-06, |
| "loss": 0.1257, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.11507936507936507, |
| "grad_norm": 0.8228491544723511, |
| "learning_rate": 9.889825144268029e-06, |
| "loss": 0.0996, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.11904761904761904, |
| "grad_norm": 0.8693366646766663, |
| "learning_rate": 9.875682047713847e-06, |
| "loss": 0.1193, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.12301587301587301, |
| "grad_norm": 0.8517415523529053, |
| "learning_rate": 9.860696522628638e-06, |
| "loss": 0.1002, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.12698412698412698, |
| "grad_norm": 0.8783800005912781, |
| "learning_rate": 9.84487115823659e-06, |
| "loss": 0.1055, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.13095238095238096, |
| "grad_norm": 0.8121059536933899, |
| "learning_rate": 9.828208688870736e-06, |
| "loss": 0.0906, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.1349206349206349, |
| "grad_norm": 0.9300619959831238, |
| "learning_rate": 9.810711993500506e-06, |
| "loss": 0.132, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.1388888888888889, |
| "grad_norm": 0.7378125190734863, |
| "learning_rate": 9.792384095234312e-06, |
| "loss": 0.0821, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.14285714285714285, |
| "grad_norm": 0.9593575596809387, |
| "learning_rate": 9.773228160797187e-06, |
| "loss": 0.1413, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.14682539682539683, |
| "grad_norm": 0.8884281516075134, |
| "learning_rate": 9.753247499983649e-06, |
| "loss": 0.1197, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.15079365079365079, |
| "grad_norm": 0.9438914656639099, |
| "learning_rate": 9.732445565085823e-06, |
| "loss": 0.1173, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.15476190476190477, |
| "grad_norm": 0.9140347838401794, |
| "learning_rate": 9.71082595029695e-06, |
| "loss": 0.1369, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.15873015873015872, |
| "grad_norm": 0.9356802105903625, |
| "learning_rate": 9.688392391090374e-06, |
| "loss": 0.1345, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.1626984126984127, |
| "grad_norm": 0.9041910767555237, |
| "learning_rate": 9.665148763574123e-06, |
| "loss": 0.1288, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.16666666666666666, |
| "grad_norm": 1.086296558380127, |
| "learning_rate": 9.64109908382119e-06, |
| "loss": 0.1717, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.17063492063492064, |
| "grad_norm": 0.81577467918396, |
| "learning_rate": 9.616247507175624e-06, |
| "loss": 0.0939, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.1746031746031746, |
| "grad_norm": 0.7911767363548279, |
| "learning_rate": 9.590598327534563e-06, |
| "loss": 0.1011, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.17857142857142858, |
| "grad_norm": 1.0346065759658813, |
| "learning_rate": 9.56415597660634e-06, |
| "loss": 0.1359, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.18253968253968253, |
| "grad_norm": 0.9680626392364502, |
| "learning_rate": 9.536925023144742e-06, |
| "loss": 0.1411, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.1865079365079365, |
| "grad_norm": 1.1418447494506836, |
| "learning_rate": 9.508910172159635e-06, |
| "loss": 0.1447, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.19047619047619047, |
| "grad_norm": 1.0825251340866089, |
| "learning_rate": 9.48011626410401e-06, |
| "loss": 0.1203, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.19444444444444445, |
| "grad_norm": 0.8085933923721313, |
| "learning_rate": 9.450548274037652e-06, |
| "loss": 0.1118, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.1984126984126984, |
| "grad_norm": 1.0938574075698853, |
| "learning_rate": 9.420211310767534e-06, |
| "loss": 0.131, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.20238095238095238, |
| "grad_norm": 0.9817928075790405, |
| "learning_rate": 9.389110615965102e-06, |
| "loss": 0.1272, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.20634920634920634, |
| "grad_norm": 0.7266258597373962, |
| "learning_rate": 9.35725156326063e-06, |
| "loss": 0.1021, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.21031746031746032, |
| "grad_norm": 0.9259639382362366, |
| "learning_rate": 9.324639657314742e-06, |
| "loss": 0.1637, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.21428571428571427, |
| "grad_norm": 0.7401134371757507, |
| "learning_rate": 9.291280532867301e-06, |
| "loss": 0.1058, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.21825396825396826, |
| "grad_norm": 0.7972776889801025, |
| "learning_rate": 9.257179953763846e-06, |
| "loss": 0.119, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.2222222222222222, |
| "grad_norm": 0.7029431462287903, |
| "learning_rate": 9.222343811959694e-06, |
| "loss": 0.0905, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.2261904761904762, |
| "grad_norm": 0.9574322700500488, |
| "learning_rate": 9.186778126501916e-06, |
| "loss": 0.1559, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.23015873015873015, |
| "grad_norm": 0.8770625591278076, |
| "learning_rate": 9.150489042489368e-06, |
| "loss": 0.1117, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.23412698412698413, |
| "grad_norm": 0.798253059387207, |
| "learning_rate": 9.113482830010918e-06, |
| "loss": 0.0985, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.23809523809523808, |
| "grad_norm": 0.780923068523407, |
| "learning_rate": 9.075765883062093e-06, |
| "loss": 0.1185, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.24206349206349206, |
| "grad_norm": 0.9421917796134949, |
| "learning_rate": 9.037344718440321e-06, |
| "loss": 0.1397, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.24603174603174602, |
| "grad_norm": 0.9296698570251465, |
| "learning_rate": 8.99822597461894e-06, |
| "loss": 0.1439, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 0.989691972732544, |
| "learning_rate": 8.958416410600188e-06, |
| "loss": 0.1417, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.25396825396825395, |
| "grad_norm": 0.8695488572120667, |
| "learning_rate": 8.917922904747385e-06, |
| "loss": 0.1231, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.25793650793650796, |
| "grad_norm": 0.8557742834091187, |
| "learning_rate": 8.876752453596462e-06, |
| "loss": 0.1129, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.2619047619047619, |
| "grad_norm": 0.8572667837142944, |
| "learning_rate": 8.834912170647102e-06, |
| "loss": 0.1355, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.26587301587301587, |
| "grad_norm": 1.0003724098205566, |
| "learning_rate": 8.792409285133644e-06, |
| "loss": 0.1601, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.2698412698412698, |
| "grad_norm": 0.9211122989654541, |
| "learning_rate": 8.749251140776016e-06, |
| "loss": 0.0913, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.27380952380952384, |
| "grad_norm": 0.8586229681968689, |
| "learning_rate": 8.705445194510868e-06, |
| "loss": 0.1041, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.2777777777777778, |
| "grad_norm": 0.8599536418914795, |
| "learning_rate": 8.660999015203152e-06, |
| "loss": 0.1141, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.28174603174603174, |
| "grad_norm": 0.7690649032592773, |
| "learning_rate": 8.615920282338355e-06, |
| "loss": 0.0876, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.2857142857142857, |
| "grad_norm": 0.9678345918655396, |
| "learning_rate": 8.570216784695637e-06, |
| "loss": 0.1313, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.2896825396825397, |
| "grad_norm": 1.0103791952133179, |
| "learning_rate": 8.52389641900206e-06, |
| "loss": 0.1396, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.29365079365079366, |
| "grad_norm": 0.8074073195457458, |
| "learning_rate": 8.476967188568187e-06, |
| "loss": 0.1069, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.2976190476190476, |
| "grad_norm": 0.8165928721427917, |
| "learning_rate": 8.429437201905254e-06, |
| "loss": 0.0957, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.30158730158730157, |
| "grad_norm": 0.8850393891334534, |
| "learning_rate": 8.38131467132416e-06, |
| "loss": 0.1139, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.3055555555555556, |
| "grad_norm": 0.7706073522567749, |
| "learning_rate": 8.332607911516545e-06, |
| "loss": 0.0906, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.30952380952380953, |
| "grad_norm": 0.8396443128585815, |
| "learning_rate": 8.283325338118154e-06, |
| "loss": 0.1223, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.3134920634920635, |
| "grad_norm": 0.6697443127632141, |
| "learning_rate": 8.233475466254766e-06, |
| "loss": 0.0934, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.31746031746031744, |
| "grad_norm": 0.931776762008667, |
| "learning_rate": 8.183066909070946e-06, |
| "loss": 0.126, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.32142857142857145, |
| "grad_norm": 0.8448562026023865, |
| "learning_rate": 8.132108376241849e-06, |
| "loss": 0.1074, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.3253968253968254, |
| "grad_norm": 0.7601461410522461, |
| "learning_rate": 8.08060867246834e-06, |
| "loss": 0.1016, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.32936507936507936, |
| "grad_norm": 0.8407372832298279, |
| "learning_rate": 8.028576695955711e-06, |
| "loss": 0.0987, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.3333333333333333, |
| "grad_norm": 0.7016237378120422, |
| "learning_rate": 7.976021436876232e-06, |
| "loss": 0.0854, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.3373015873015873, |
| "grad_norm": 0.8386403322219849, |
| "learning_rate": 7.92295197581581e-06, |
| "loss": 0.1308, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.3412698412698413, |
| "grad_norm": 0.8006751537322998, |
| "learning_rate": 7.869377482205042e-06, |
| "loss": 0.1136, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.34523809523809523, |
| "grad_norm": 0.9047397375106812, |
| "learning_rate": 7.815307212734888e-06, |
| "loss": 0.1307, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.3492063492063492, |
| "grad_norm": 0.8037883043289185, |
| "learning_rate": 7.7607505097573e-06, |
| "loss": 0.1057, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.3531746031746032, |
| "grad_norm": 0.7339609861373901, |
| "learning_rate": 7.705716799671019e-06, |
| "loss": 0.0983, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.35714285714285715, |
| "grad_norm": 0.9146028757095337, |
| "learning_rate": 7.650215591292888e-06, |
| "loss": 0.1195, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.3611111111111111, |
| "grad_norm": 0.857414960861206, |
| "learning_rate": 7.594256474214883e-06, |
| "loss": 0.1124, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.36507936507936506, |
| "grad_norm": 0.775814414024353, |
| "learning_rate": 7.537849117147212e-06, |
| "loss": 0.0912, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.36904761904761907, |
| "grad_norm": 0.8781107664108276, |
| "learning_rate": 7.481003266247745e-06, |
| "loss": 0.1272, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.373015873015873, |
| "grad_norm": 0.7291442155838013, |
| "learning_rate": 7.4237287434380485e-06, |
| "loss": 0.1009, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.376984126984127, |
| "grad_norm": 0.8335267901420593, |
| "learning_rate": 7.366035444706346e-06, |
| "loss": 0.0911, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.38095238095238093, |
| "grad_norm": 0.9337305426597595, |
| "learning_rate": 7.307933338397667e-06, |
| "loss": 0.1116, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.38492063492063494, |
| "grad_norm": 0.9217402338981628, |
| "learning_rate": 7.249432463491498e-06, |
| "loss": 0.1299, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.3888888888888889, |
| "grad_norm": 0.9241219758987427, |
| "learning_rate": 7.190542927867234e-06, |
| "loss": 0.1247, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.39285714285714285, |
| "grad_norm": 0.9741147756576538, |
| "learning_rate": 7.131274906557725e-06, |
| "loss": 0.1123, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.3968253968253968, |
| "grad_norm": 0.9606366157531738, |
| "learning_rate": 7.0716386399912075e-06, |
| "loss": 0.1176, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.4007936507936508, |
| "grad_norm": 0.7203896641731262, |
| "learning_rate": 7.0116444322219575e-06, |
| "loss": 0.1118, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.40476190476190477, |
| "grad_norm": 0.8363652229309082, |
| "learning_rate": 6.95130264914993e-06, |
| "loss": 0.12, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.4087301587301587, |
| "grad_norm": 0.6669372916221619, |
| "learning_rate": 6.890623716729724e-06, |
| "loss": 0.0751, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.4126984126984127, |
| "grad_norm": 0.7728121280670166, |
| "learning_rate": 6.829618119169169e-06, |
| "loss": 0.1016, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.4166666666666667, |
| "grad_norm": 0.8097180724143982, |
| "learning_rate": 6.768296397117848e-06, |
| "loss": 0.1163, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.42063492063492064, |
| "grad_norm": 0.935376763343811, |
| "learning_rate": 6.706669145845863e-06, |
| "loss": 0.1407, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.4246031746031746, |
| "grad_norm": 0.7769891619682312, |
| "learning_rate": 6.6447470134131685e-06, |
| "loss": 0.0998, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.42857142857142855, |
| "grad_norm": 0.7420790195465088, |
| "learning_rate": 6.5825406988297815e-06, |
| "loss": 0.0974, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.43253968253968256, |
| "grad_norm": 0.8356537818908691, |
| "learning_rate": 6.520060950207186e-06, |
| "loss": 0.1368, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.4365079365079365, |
| "grad_norm": 0.8886961340904236, |
| "learning_rate": 6.457318562901257e-06, |
| "loss": 0.1258, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.44047619047619047, |
| "grad_norm": 0.8278570771217346, |
| "learning_rate": 6.394324377647028e-06, |
| "loss": 0.1207, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.4444444444444444, |
| "grad_norm": 0.7972725033760071, |
| "learning_rate": 6.331089278685599e-06, |
| "loss": 0.1233, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.44841269841269843, |
| "grad_norm": 0.7953070402145386, |
| "learning_rate": 6.267624191883551e-06, |
| "loss": 0.0925, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.4523809523809524, |
| "grad_norm": 0.8089295029640198, |
| "learning_rate": 6.203940082845144e-06, |
| "loss": 0.1058, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.45634920634920634, |
| "grad_norm": 0.9612205028533936, |
| "learning_rate": 6.140047955017672e-06, |
| "loss": 0.1254, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.4603174603174603, |
| "grad_norm": 0.7713960409164429, |
| "learning_rate": 6.075958847790262e-06, |
| "loss": 0.0923, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.4642857142857143, |
| "grad_norm": 0.8480079174041748, |
| "learning_rate": 6.011683834586474e-06, |
| "loss": 0.1245, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.46825396825396826, |
| "grad_norm": 0.7191230654716492, |
| "learning_rate": 5.947234020951015e-06, |
| "loss": 0.0957, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.4722222222222222, |
| "grad_norm": 0.8899836540222168, |
| "learning_rate": 5.882620542630901e-06, |
| "loss": 0.1155, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.47619047619047616, |
| "grad_norm": 0.813059389591217, |
| "learning_rate": 5.817854563651415e-06, |
| "loss": 0.1106, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.4801587301587302, |
| "grad_norm": 0.8137439489364624, |
| "learning_rate": 5.752947274387147e-06, |
| "loss": 0.1042, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.48412698412698413, |
| "grad_norm": 0.9558090567588806, |
| "learning_rate": 5.687909889628529e-06, |
| "loss": 0.1197, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.4880952380952381, |
| "grad_norm": 0.9763339757919312, |
| "learning_rate": 5.622753646644102e-06, |
| "loss": 0.1273, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.49206349206349204, |
| "grad_norm": 0.7899503111839294, |
| "learning_rate": 5.557489803238934e-06, |
| "loss": 0.0896, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.49603174603174605, |
| "grad_norm": 0.8102907538414001, |
| "learning_rate": 5.492129635809473e-06, |
| "loss": 0.0992, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 0.7076608538627625, |
| "learning_rate": 5.426684437395196e-06, |
| "loss": 0.123, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.503968253968254, |
| "grad_norm": 0.7774274945259094, |
| "learning_rate": 5.361165515727374e-06, |
| "loss": 0.1035, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.5079365079365079, |
| "grad_norm": 0.7213121056556702, |
| "learning_rate": 5.295584191275308e-06, |
| "loss": 0.0872, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.5119047619047619, |
| "grad_norm": 0.8430085182189941, |
| "learning_rate": 5.229951795290353e-06, |
| "loss": 0.1232, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.5158730158730159, |
| "grad_norm": 0.9626345634460449, |
| "learning_rate": 5.164279667848094e-06, |
| "loss": 0.1609, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.5198412698412699, |
| "grad_norm": 0.8764887452125549, |
| "learning_rate": 5.0985791558889785e-06, |
| "loss": 0.135, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.5238095238095238, |
| "grad_norm": 0.8301170468330383, |
| "learning_rate": 5.032861611257783e-06, |
| "loss": 0.1246, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.5277777777777778, |
| "grad_norm": 0.7681728601455688, |
| "learning_rate": 4.967138388742218e-06, |
| "loss": 0.103, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.5317460317460317, |
| "grad_norm": 0.7660509347915649, |
| "learning_rate": 4.9014208441110215e-06, |
| "loss": 0.1055, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.5357142857142857, |
| "grad_norm": 0.8104261755943298, |
| "learning_rate": 4.835720332151907e-06, |
| "loss": 0.1237, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.5396825396825397, |
| "grad_norm": 0.7689002156257629, |
| "learning_rate": 4.770048204709648e-06, |
| "loss": 0.1049, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.5436507936507936, |
| "grad_norm": 0.833803653717041, |
| "learning_rate": 4.7044158087246926e-06, |
| "loss": 0.1099, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.5476190476190477, |
| "grad_norm": 0.7744324207305908, |
| "learning_rate": 4.6388344842726266e-06, |
| "loss": 0.1009, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.5515873015873016, |
| "grad_norm": 0.8948577642440796, |
| "learning_rate": 4.573315562604804e-06, |
| "loss": 0.1226, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.5555555555555556, |
| "grad_norm": 0.8321415781974792, |
| "learning_rate": 4.5078703641905275e-06, |
| "loss": 0.0984, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.5595238095238095, |
| "grad_norm": 0.8380492925643921, |
| "learning_rate": 4.442510196761068e-06, |
| "loss": 0.1052, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.5634920634920635, |
| "grad_norm": 0.8537865877151489, |
| "learning_rate": 4.377246353355899e-06, |
| "loss": 0.1145, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.5674603174603174, |
| "grad_norm": 0.9566012620925903, |
| "learning_rate": 4.312090110371473e-06, |
| "loss": 0.1453, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.5714285714285714, |
| "grad_norm": 0.8509257435798645, |
| "learning_rate": 4.247052725612853e-06, |
| "loss": 0.0998, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.5753968253968254, |
| "grad_norm": 0.7241506576538086, |
| "learning_rate": 4.182145436348587e-06, |
| "loss": 0.0929, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.5793650793650794, |
| "grad_norm": 0.9016909003257751, |
| "learning_rate": 4.1173794573691e-06, |
| "loss": 0.1061, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.5833333333333334, |
| "grad_norm": 0.846988320350647, |
| "learning_rate": 4.052765979048986e-06, |
| "loss": 0.1237, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.5873015873015873, |
| "grad_norm": 0.9077289700508118, |
| "learning_rate": 3.988316165413528e-06, |
| "loss": 0.0911, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.5912698412698413, |
| "grad_norm": 0.8484727740287781, |
| "learning_rate": 3.924041152209739e-06, |
| "loss": 0.1028, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.5952380952380952, |
| "grad_norm": 0.7354507446289062, |
| "learning_rate": 3.859952044982329e-06, |
| "loss": 0.0847, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.5992063492063492, |
| "grad_norm": 0.644497275352478, |
| "learning_rate": 3.7960599171548572e-06, |
| "loss": 0.09, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.6031746031746031, |
| "grad_norm": 0.7260298728942871, |
| "learning_rate": 3.732375808116451e-06, |
| "loss": 0.0788, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.6071428571428571, |
| "grad_norm": 0.7378935217857361, |
| "learning_rate": 3.6689107213144025e-06, |
| "loss": 0.1021, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.6111111111111112, |
| "grad_norm": 0.644940197467804, |
| "learning_rate": 3.6056756223529734e-06, |
| "loss": 0.0769, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.6150793650793651, |
| "grad_norm": 0.8905737400054932, |
| "learning_rate": 3.542681437098745e-06, |
| "loss": 0.1694, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.6190476190476191, |
| "grad_norm": 0.7038963437080383, |
| "learning_rate": 3.479939049792817e-06, |
| "loss": 0.0968, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.623015873015873, |
| "grad_norm": 0.7893255352973938, |
| "learning_rate": 3.4174593011702197e-06, |
| "loss": 0.1158, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.626984126984127, |
| "grad_norm": 0.7545840740203857, |
| "learning_rate": 3.3552529865868323e-06, |
| "loss": 0.0929, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.6309523809523809, |
| "grad_norm": 0.7931913733482361, |
| "learning_rate": 3.2933308541541365e-06, |
| "loss": 0.1034, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.6349206349206349, |
| "grad_norm": 0.7249116897583008, |
| "learning_rate": 3.2317036028821523e-06, |
| "loss": 0.0874, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.6388888888888888, |
| "grad_norm": 0.706990122795105, |
| "learning_rate": 3.1703818808308327e-06, |
| "loss": 0.0882, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.6428571428571429, |
| "grad_norm": 0.6917218565940857, |
| "learning_rate": 3.1093762832702775e-06, |
| "loss": 0.0817, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.6468253968253969, |
| "grad_norm": 0.8852331042289734, |
| "learning_rate": 3.048697350850073e-06, |
| "loss": 0.1279, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.6507936507936508, |
| "grad_norm": 0.8701708316802979, |
| "learning_rate": 2.988355567778043e-06, |
| "loss": 0.1171, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.6547619047619048, |
| "grad_norm": 0.8259684443473816, |
| "learning_rate": 2.9283613600087933e-06, |
| "loss": 0.1141, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.6587301587301587, |
| "grad_norm": 0.8396156430244446, |
| "learning_rate": 2.8687250934422774e-06, |
| "loss": 0.1178, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.6626984126984127, |
| "grad_norm": 0.7612534761428833, |
| "learning_rate": 2.809457072132766e-06, |
| "loss": 0.1037, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.6666666666666666, |
| "grad_norm": 0.7613610029220581, |
| "learning_rate": 2.750567536508504e-06, |
| "loss": 0.0943, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.6706349206349206, |
| "grad_norm": 0.8023819923400879, |
| "learning_rate": 2.692066661602333e-06, |
| "loss": 0.0949, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.6746031746031746, |
| "grad_norm": 0.7530853748321533, |
| "learning_rate": 2.633964555293654e-06, |
| "loss": 0.0852, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.6785714285714286, |
| "grad_norm": 0.9524847269058228, |
| "learning_rate": 2.576271256561953e-06, |
| "loss": 0.0924, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.6825396825396826, |
| "grad_norm": 0.8371558785438538, |
| "learning_rate": 2.5189967337522574e-06, |
| "loss": 0.1211, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.6865079365079365, |
| "grad_norm": 0.676647961139679, |
| "learning_rate": 2.46215088285279e-06, |
| "loss": 0.0919, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.6904761904761905, |
| "grad_norm": 0.6709980368614197, |
| "learning_rate": 2.4057435257851173e-06, |
| "loss": 0.0745, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.6944444444444444, |
| "grad_norm": 0.7309594750404358, |
| "learning_rate": 2.349784408707112e-06, |
| "loss": 0.0967, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.6984126984126984, |
| "grad_norm": 0.7355422377586365, |
| "learning_rate": 2.2942832003289823e-06, |
| "loss": 0.1096, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.7023809523809523, |
| "grad_norm": 0.6394128203392029, |
| "learning_rate": 2.2392494902427027e-06, |
| "loss": 0.0876, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.7063492063492064, |
| "grad_norm": 0.7420768141746521, |
| "learning_rate": 2.1846927872651135e-06, |
| "loss": 0.1039, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.7103174603174603, |
| "grad_norm": 0.7211629152297974, |
| "learning_rate": 2.1306225177949584e-06, |
| "loss": 0.0875, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.7142857142857143, |
| "grad_norm": 0.7187921404838562, |
| "learning_rate": 2.07704802418419e-06, |
| "loss": 0.0885, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.7182539682539683, |
| "grad_norm": 0.7532177567481995, |
| "learning_rate": 2.023978563123771e-06, |
| "loss": 0.106, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.7222222222222222, |
| "grad_norm": 0.726718544960022, |
| "learning_rate": 1.9714233040442915e-06, |
| "loss": 0.092, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.7261904761904762, |
| "grad_norm": 0.6967416405677795, |
| "learning_rate": 1.919391327531663e-06, |
| "loss": 0.0771, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.7301587301587301, |
| "grad_norm": 0.7169931530952454, |
| "learning_rate": 1.8678916237581524e-06, |
| "loss": 0.1009, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.7341269841269841, |
| "grad_norm": 0.662554919719696, |
| "learning_rate": 1.816933090929055e-06, |
| "loss": 0.0727, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.7380952380952381, |
| "grad_norm": 0.9608215093612671, |
| "learning_rate": 1.7665245337452368e-06, |
| "loss": 0.1224, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.7420634920634921, |
| "grad_norm": 0.9227773547172546, |
| "learning_rate": 1.716674661881848e-06, |
| "loss": 0.1166, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.746031746031746, |
| "grad_norm": 0.8112769722938538, |
| "learning_rate": 1.667392088483456e-06, |
| "loss": 0.1101, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 0.8895912766456604, |
| "learning_rate": 1.6186853286758397e-06, |
| "loss": 0.1162, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.753968253968254, |
| "grad_norm": 0.6724901795387268, |
| "learning_rate": 1.570562798094747e-06, |
| "loss": 0.0819, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.7579365079365079, |
| "grad_norm": 0.8350103497505188, |
| "learning_rate": 1.5230328114318127e-06, |
| "loss": 0.1021, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.7619047619047619, |
| "grad_norm": 0.9384281039237976, |
| "learning_rate": 1.4761035809979395e-06, |
| "loss": 0.135, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.7658730158730159, |
| "grad_norm": 0.894897997379303, |
| "learning_rate": 1.4297832153043657e-06, |
| "loss": 0.1075, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.7698412698412699, |
| "grad_norm": 0.7277584671974182, |
| "learning_rate": 1.3840797176616467e-06, |
| "loss": 0.0788, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.7738095238095238, |
| "grad_norm": 0.809172511100769, |
| "learning_rate": 1.3390009847968505e-06, |
| "loss": 0.0923, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.7777777777777778, |
| "grad_norm": 0.8635011911392212, |
| "learning_rate": 1.2945548054891322e-06, |
| "loss": 0.0895, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.7817460317460317, |
| "grad_norm": 0.8458127975463867, |
| "learning_rate": 1.2507488592239848e-06, |
| "loss": 0.1392, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.7857142857142857, |
| "grad_norm": 0.7588448524475098, |
| "learning_rate": 1.2075907148663579e-06, |
| "loss": 0.1068, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.7896825396825397, |
| "grad_norm": 0.6866970062255859, |
| "learning_rate": 1.1650878293528994e-06, |
| "loss": 0.0875, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.7936507936507936, |
| "grad_norm": 0.7142922878265381, |
| "learning_rate": 1.1232475464035386e-06, |
| "loss": 0.0892, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.7976190476190477, |
| "grad_norm": 0.6719024777412415, |
| "learning_rate": 1.0820770952526155e-06, |
| "loss": 0.0904, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.8015873015873016, |
| "grad_norm": 0.6220860481262207, |
| "learning_rate": 1.0415835893998116e-06, |
| "loss": 0.0773, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.8055555555555556, |
| "grad_norm": 0.5566734075546265, |
| "learning_rate": 1.0017740253810608e-06, |
| "loss": 0.0704, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.8095238095238095, |
| "grad_norm": 0.6570054888725281, |
| "learning_rate": 9.62655281559679e-07, |
| "loss": 0.0852, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.8134920634920635, |
| "grad_norm": 0.6448202729225159, |
| "learning_rate": 9.242341169379077e-07, |
| "loss": 0.0703, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.8174603174603174, |
| "grad_norm": 0.7391513586044312, |
| "learning_rate": 8.865171699890835e-07, |
| "loss": 0.1028, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.8214285714285714, |
| "grad_norm": 0.7353842854499817, |
| "learning_rate": 8.495109575106331e-07, |
| "loss": 0.0968, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.8253968253968254, |
| "grad_norm": 0.8170803785324097, |
| "learning_rate": 8.132218734980852e-07, |
| "loss": 0.113, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.8293650793650794, |
| "grad_norm": 0.7208415269851685, |
| "learning_rate": 7.776561880403072e-07, |
| "loss": 0.0937, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.8333333333333334, |
| "grad_norm": 0.792476236820221, |
| "learning_rate": 7.42820046236154e-07, |
| "loss": 0.117, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.8373015873015873, |
| "grad_norm": 0.7033000588417053, |
| "learning_rate": 7.087194671326986e-07, |
| "loss": 0.0836, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.8412698412698413, |
| "grad_norm": 0.6652337908744812, |
| "learning_rate": 6.753603426852589e-07, |
| "loss": 0.0754, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.8452380952380952, |
| "grad_norm": 0.9486038684844971, |
| "learning_rate": 6.427484367393699e-07, |
| "loss": 0.1304, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.8492063492063492, |
| "grad_norm": 0.8958883285522461, |
| "learning_rate": 6.108893840348995e-07, |
| "loss": 0.0917, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.8531746031746031, |
| "grad_norm": 0.6590295433998108, |
| "learning_rate": 5.797886892324695e-07, |
| "loss": 0.0759, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.8571428571428571, |
| "grad_norm": 0.789568305015564, |
| "learning_rate": 5.494517259623478e-07, |
| "loss": 0.0966, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.8611111111111112, |
| "grad_norm": 0.8645297288894653, |
| "learning_rate": 5.198837358959901e-07, |
| "loss": 0.1092, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.8650793650793651, |
| "grad_norm": 0.8162879347801208, |
| "learning_rate": 4.91089827840367e-07, |
| "loss": 0.1082, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.8690476190476191, |
| "grad_norm": 0.9198590517044067, |
| "learning_rate": 4.6307497685525894e-07, |
| "loss": 0.1278, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.873015873015873, |
| "grad_norm": 0.6241840124130249, |
| "learning_rate": 4.3584402339366174e-07, |
| "loss": 0.0662, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.876984126984127, |
| "grad_norm": 0.8222904801368713, |
| "learning_rate": 4.0940167246543595e-07, |
| "loss": 0.1049, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.8809523809523809, |
| "grad_norm": 0.7959262132644653, |
| "learning_rate": 3.8375249282437743e-07, |
| "loss": 0.0972, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.8849206349206349, |
| "grad_norm": 0.6456685066223145, |
| "learning_rate": 3.589009161788104e-07, |
| "loss": 0.0774, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.8888888888888888, |
| "grad_norm": 0.9123170375823975, |
| "learning_rate": 3.3485123642587657e-07, |
| "loss": 0.1471, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.8928571428571429, |
| "grad_norm": 0.6460578441619873, |
| "learning_rate": 3.116076089096265e-07, |
| "loss": 0.0684, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.8968253968253969, |
| "grad_norm": 0.6745744347572327, |
| "learning_rate": 2.8917404970305096e-07, |
| "loss": 0.0848, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.9007936507936508, |
| "grad_norm": 0.7377632260322571, |
| "learning_rate": 2.6755443491417786e-07, |
| "loss": 0.0897, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.9047619047619048, |
| "grad_norm": 0.7380819916725159, |
| "learning_rate": 2.467525000163523e-07, |
| "loss": 0.1127, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.9087301587301587, |
| "grad_norm": 0.6314770579338074, |
| "learning_rate": 2.2677183920281342e-07, |
| "loss": 0.0775, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.9126984126984127, |
| "grad_norm": 0.7447143197059631, |
| "learning_rate": 2.0761590476568893e-07, |
| "loss": 0.1098, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.9166666666666666, |
| "grad_norm": 0.6674031615257263, |
| "learning_rate": 1.892880064994934e-07, |
| "loss": 0.0729, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.9206349206349206, |
| "grad_norm": 0.7818775773048401, |
| "learning_rate": 1.7179131112926628e-07, |
| "loss": 0.0965, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.9246031746031746, |
| "grad_norm": 0.6574113368988037, |
| "learning_rate": 1.551288417634106e-07, |
| "loss": 0.0833, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.9285714285714286, |
| "grad_norm": 0.7490445971488953, |
| "learning_rate": 1.3930347737136195e-07, |
| "loss": 0.0865, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.9325396825396826, |
| "grad_norm": 0.8491647839546204, |
| "learning_rate": 1.2431795228615372e-07, |
| "loss": 0.1268, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.9365079365079365, |
| "grad_norm": 0.7804898023605347, |
| "learning_rate": 1.1017485573197151e-07, |
| "loss": 0.0975, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.9404761904761905, |
| "grad_norm": 0.8244890570640564, |
| "learning_rate": 9.687663137678605e-08, |
| "loss": 0.1113, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.9444444444444444, |
| "grad_norm": 0.803353488445282, |
| "learning_rate": 8.442557691013042e-08, |
| "loss": 0.1053, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.9484126984126984, |
| "grad_norm": 0.7855432629585266, |
| "learning_rate": 7.282384364610207e-08, |
| "loss": 0.1012, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.9523809523809523, |
| "grad_norm": 0.80164635181427, |
| "learning_rate": 6.207343615165562e-08, |
| "loss": 0.1138, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.9563492063492064, |
| "grad_norm": 0.6952555179595947, |
| "learning_rate": 5.21762119002478e-08, |
| "loss": 0.0786, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.9603174603174603, |
| "grad_norm": 0.7810098528862, |
| "learning_rate": 4.31338809509052e-08, |
| "loss": 0.0923, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.9642857142857143, |
| "grad_norm": 0.8230253458023071, |
| "learning_rate": 3.494800565275125e-08, |
| "loss": 0.1149, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.9682539682539683, |
| "grad_norm": 0.8179909586906433, |
| "learning_rate": 2.7620000375064848e-08, |
| "loss": 0.0936, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.9722222222222222, |
| "grad_norm": 0.9115637540817261, |
| "learning_rate": 2.115113126290258e-08, |
| "loss": 0.1169, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.9761904761904762, |
| "grad_norm": 0.6795814633369446, |
| "learning_rate": 1.554251601833201e-08, |
| "loss": 0.0689, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.9801587301587301, |
| "grad_norm": 0.7916927337646484, |
| "learning_rate": 1.0795123707312283e-08, |
| "loss": 0.1116, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.9841269841269841, |
| "grad_norm": 0.6779837012290955, |
| "learning_rate": 6.9097745922580564e-09, |
| "loss": 0.0853, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.9880952380952381, |
| "grad_norm": 0.6872785091400146, |
| "learning_rate": 3.887139990313427e-09, |
| "loss": 0.0926, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.9920634920634921, |
| "grad_norm": 0.6764498949050903, |
| "learning_rate": 1.7277421573608234e-09, |
| "loss": 0.0831, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.996031746031746, |
| "grad_norm": 0.7526364326477051, |
| "learning_rate": 4.3195419778319095e-10, |
| "loss": 0.0966, |
| "step": 251 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.7119873762130737, |
| "learning_rate": 0.0, |
| "loss": 0.1041, |
| "step": 252 |
| }, |
| { |
| "epoch": 1.0, |
| "step": 252, |
| "total_flos": 27436118114304.0, |
| "train_loss": 0.1079725922040996, |
| "train_runtime": 1946.0648, |
| "train_samples_per_second": 2.067, |
| "train_steps_per_second": 0.129 |
| } |
| ], |
| "logging_steps": 1.0, |
| "max_steps": 252, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": false, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 27436118114304.0, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
