| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 467, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.246411323547363, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 1.7737, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.543394088745117, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 1.803, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 10.129024505615234, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.7969, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.124090671539307, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 1.7299, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 5.658255577087402, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.6504, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.513315200805664, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.652, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.358516216278076, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 1.5206, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.1449408531188965, | |
| "learning_rate": 1.0666666666666667e-05, | |
| "loss": 1.5219, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.3279998302459717, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.422, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.4790432453155518, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 1.4384, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.099670648574829, | |
| "learning_rate": 1.4666666666666666e-05, | |
| "loss": 1.4168, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.403562545776367, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.3644, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.540651559829712, | |
| "learning_rate": 1.7333333333333336e-05, | |
| "loss": 1.3784, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.3258016109466553, | |
| "learning_rate": 1.866666666666667e-05, | |
| "loss": 1.3197, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.8612205982208252, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3194, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.9114793539047241, | |
| "learning_rate": 1.9999758458848847e-05, | |
| "loss": 1.278, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.5149403810501099, | |
| "learning_rate": 1.9999033847063813e-05, | |
| "loss": 1.3074, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.7767683267593384, | |
| "learning_rate": 1.9997826199649607e-05, | |
| "loss": 1.2354, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.4256222248077393, | |
| "learning_rate": 1.9996135574945543e-05, | |
| "loss": 1.265, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.5344816446304321, | |
| "learning_rate": 1.9993962054622703e-05, | |
| "loss": 1.2435, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.8870617151260376, | |
| "learning_rate": 1.9991305743680013e-05, | |
| "loss": 1.21, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.159767508506775, | |
| "learning_rate": 1.9988166770439156e-05, | |
| "loss": 1.2392, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.2659238576889038, | |
| "learning_rate": 1.9984545286538362e-05, | |
| "loss": 1.1977, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.3020108938217163, | |
| "learning_rate": 1.9980441466925118e-05, | |
| "loss": 1.2185, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.37464439868927, | |
| "learning_rate": 1.9975855509847688e-05, | |
| "loss": 1.1944, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0609029531478882, | |
| "learning_rate": 1.9970787636845536e-05, | |
| "loss": 1.1783, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0385644435882568, | |
| "learning_rate": 1.9965238092738643e-05, | |
| "loss": 1.1499, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9814103841781616, | |
| "learning_rate": 1.9959207145615663e-05, | |
| "loss": 1.1743, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.2126802206039429, | |
| "learning_rate": 1.9952695086820975e-05, | |
| "loss": 1.1281, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8121795654296875, | |
| "learning_rate": 1.9945702230940616e-05, | |
| "loss": 1.1435, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9171407222747803, | |
| "learning_rate": 1.993822891578708e-05, | |
| "loss": 1.1498, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0829691886901855, | |
| "learning_rate": 1.9930275502382993e-05, | |
| "loss": 1.104, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.7339140176773071, | |
| "learning_rate": 1.9921842374943682e-05, | |
| "loss": 1.1559, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8814339637756348, | |
| "learning_rate": 1.9912929940858607e-05, | |
| "loss": 1.0989, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9078977704048157, | |
| "learning_rate": 1.9903538630671687e-05, | |
| "loss": 1.1453, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9740134477615356, | |
| "learning_rate": 1.9893668898060504e-05, | |
| "loss": 1.1131, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.7271884083747864, | |
| "learning_rate": 1.988332121981436e-05, | |
| "loss": 1.1191, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.7487226724624634, | |
| "learning_rate": 1.9872496095811287e-05, | |
| "loss": 1.1082, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.7832925319671631, | |
| "learning_rate": 1.9861194048993865e-05, | |
| "loss": 1.1118, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7679692506790161, | |
| "learning_rate": 1.9849415625343972e-05, | |
| "loss": 1.1003, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8569953441619873, | |
| "learning_rate": 1.9837161393856413e-05, | |
| "loss": 1.0846, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8195632696151733, | |
| "learning_rate": 1.982443194651142e-05, | |
| "loss": 1.139, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9336971640586853, | |
| "learning_rate": 1.9811227898246072e-05, | |
| "loss": 1.0658, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.6557732224464417, | |
| "learning_rate": 1.979754988692457e-05, | |
| "loss": 1.1448, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2183958292007446, | |
| "learning_rate": 1.978339857330743e-05, | |
| "loss": 1.043, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2832332849502563, | |
| "learning_rate": 1.976877464101957e-05, | |
| "loss": 1.0996, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 4.317144393920898, | |
| "learning_rate": 1.975367879651728e-05, | |
| "loss": 1.0425, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2678251266479492, | |
| "learning_rate": 1.9738111769054095e-05, | |
| "loss": 1.1227, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8790923953056335, | |
| "learning_rate": 1.9722074310645553e-05, | |
| "loss": 1.0295, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9889333248138428, | |
| "learning_rate": 1.9705567196032894e-05, | |
| "loss": 1.0623, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.1119431257247925, | |
| "learning_rate": 1.9688591222645607e-05, | |
| "loss": 1.0716, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8923778533935547, | |
| "learning_rate": 1.9671147210562925e-05, | |
| "loss": 1.0416, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.1617224216461182, | |
| "learning_rate": 1.9653236002474202e-05, | |
| "loss": 1.0603, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8286775946617126, | |
| "learning_rate": 1.96348584636382e-05, | |
| "loss": 1.0734, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2225749492645264, | |
| "learning_rate": 1.9616015481841293e-05, | |
| "loss": 1.073, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.981415867805481, | |
| "learning_rate": 1.9596707967354588e-05, | |
| "loss": 1.0318, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7881583571434021, | |
| "learning_rate": 1.9576936852889937e-05, | |
| "loss": 1.0578, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8419631719589233, | |
| "learning_rate": 1.955670309355489e-05, | |
| "loss": 1.0451, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.0151898860931396, | |
| "learning_rate": 1.9536007666806555e-05, | |
| "loss": 1.0186, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9713700413703918, | |
| "learning_rate": 1.951485157240437e-05, | |
| "loss": 1.0603, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9093838334083557, | |
| "learning_rate": 1.9493235832361812e-05, | |
| "loss": 1.0537, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.0966782569885254, | |
| "learning_rate": 1.9471161490897027e-05, | |
| "loss": 1.071, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.921055257320404, | |
| "learning_rate": 1.9448629614382394e-05, | |
| "loss": 1.0199, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.0806677341461182, | |
| "learning_rate": 1.942564129129298e-05, | |
| "loss": 1.114, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8954259157180786, | |
| "learning_rate": 1.940219763215399e-05, | |
| "loss": 1.0356, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9842892289161682, | |
| "learning_rate": 1.9378299769487116e-05, | |
| "loss": 1.0475, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9474238157272339, | |
| "learning_rate": 1.93539488577558e-05, | |
| "loss": 1.054, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0999383926391602, | |
| "learning_rate": 1.9329146073309502e-05, | |
| "loss": 1.0291, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9849498867988586, | |
| "learning_rate": 1.9303892614326835e-05, | |
| "loss": 1.0726, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.1335885524749756, | |
| "learning_rate": 1.9278189700757717e-05, | |
| "loss": 0.9772, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.6924199461936951, | |
| "learning_rate": 1.9252038574264403e-05, | |
| "loss": 1.0549, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8446598052978516, | |
| "learning_rate": 1.9225440498161544e-05, | |
| "loss": 1.0317, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.6602651476860046, | |
| "learning_rate": 1.9198396757355118e-05, | |
| "loss": 1.0552, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9809678196907043, | |
| "learning_rate": 1.9170908658280388e-05, | |
| "loss": 0.9938, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.5746096968650818, | |
| "learning_rate": 1.9142977528838763e-05, | |
| "loss": 1.0971, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.743587076663971, | |
| "learning_rate": 1.911460471833368e-05, | |
| "loss": 1.0112, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.6281198859214783, | |
| "learning_rate": 1.9085791597405404e-05, | |
| "loss": 1.0234, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6992913484573364, | |
| "learning_rate": 1.9056539557964814e-05, | |
| "loss": 1.0683, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8796486258506775, | |
| "learning_rate": 1.902685001312616e-05, | |
| "loss": 0.9845, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6520846486091614, | |
| "learning_rate": 1.8996724397138813e-05, | |
| "loss": 1.0268, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6597517728805542, | |
| "learning_rate": 1.8966164165317968e-05, | |
| "loss": 1.0358, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7612221240997314, | |
| "learning_rate": 1.8935170793974335e-05, | |
| "loss": 1.0223, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7459383606910706, | |
| "learning_rate": 1.8903745780342838e-05, | |
| "loss": 1.0267, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7625844478607178, | |
| "learning_rate": 1.887189064251027e-05, | |
| "loss": 0.9661, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.6465503573417664, | |
| "learning_rate": 1.883960691934196e-05, | |
| "loss": 1.0564, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9878563284873962, | |
| "learning_rate": 1.8806896170407437e-05, | |
| "loss": 0.9975, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.8770087957382202, | |
| "learning_rate": 1.8773759975905098e-05, | |
| "loss": 1.0521, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9063981175422668, | |
| "learning_rate": 1.8740199936585856e-05, | |
| "loss": 0.9738, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9595961570739746, | |
| "learning_rate": 1.8706217673675813e-05, | |
| "loss": 1.0112, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9411620497703552, | |
| "learning_rate": 1.867181482879795e-05, | |
| "loss": 0.977, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.0874736309051514, | |
| "learning_rate": 1.8636993063892822e-05, | |
| "loss": 1.0564, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.878400444984436, | |
| "learning_rate": 1.8601754061138258e-05, | |
| "loss": 0.9803, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.0831676721572876, | |
| "learning_rate": 1.8566099522868118e-05, | |
| "loss": 1.042, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.9198699593544006, | |
| "learning_rate": 1.8530031171490055e-05, | |
| "loss": 0.9682, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.1813023090362549, | |
| "learning_rate": 1.8493550749402278e-05, | |
| "loss": 1.0193, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.6807182431221008, | |
| "learning_rate": 1.8456660018909424e-05, | |
| "loss": 1.0518, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.0440760850906372, | |
| "learning_rate": 1.8419360762137395e-05, | |
| "loss": 0.96, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.6701475381851196, | |
| "learning_rate": 1.8381654780947272e-05, | |
| "loss": 1.0707, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.7763065695762634, | |
| "learning_rate": 1.8343543896848275e-05, | |
| "loss": 0.9945, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.7572762966156006, | |
| "learning_rate": 1.830502995090977e-05, | |
| "loss": 1.0143, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.7713978290557861, | |
| "learning_rate": 1.826611480367232e-05, | |
| "loss": 0.9784, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.7623790502548218, | |
| "learning_rate": 1.822680033505782e-05, | |
| "loss": 1.0076, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.6760876774787903, | |
| "learning_rate": 1.8187088444278675e-05, | |
| "loss": 1.0127, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.8094907402992249, | |
| "learning_rate": 1.814698104974604e-05, | |
| "loss": 0.9477, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.6718782186508179, | |
| "learning_rate": 1.8106480088977174e-05, | |
| "loss": 1.0371, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.8604088425636292, | |
| "learning_rate": 1.8065587518501806e-05, | |
| "loss": 0.9591, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.6633604764938354, | |
| "learning_rate": 1.8024305313767648e-05, | |
| "loss": 0.9884, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.6507821679115295, | |
| "learning_rate": 1.798263546904495e-05, | |
| "loss": 1.0436, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.8079816102981567, | |
| "learning_rate": 1.7940579997330167e-05, | |
| "loss": 0.9626, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.713695764541626, | |
| "learning_rate": 1.7898140930248703e-05, | |
| "loss": 0.9958, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.7410963177680969, | |
| "learning_rate": 1.7855320317956785e-05, | |
| "loss": 0.9718, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.6720312833786011, | |
| "learning_rate": 1.7812120229042415e-05, | |
| "loss": 1.0364, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.9324734210968018, | |
| "learning_rate": 1.7768542750425427e-05, | |
| "loss": 0.941, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.9109165668487549, | |
| "learning_rate": 1.7724589987256697e-05, | |
| "loss": 1.049, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.8574079275131226, | |
| "learning_rate": 1.768026406281642e-05, | |
| "loss": 0.9955, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.8273678421974182, | |
| "learning_rate": 1.7635567118411568e-05, | |
| "loss": 1.0186, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.0233052968978882, | |
| "learning_rate": 1.7590501313272415e-05, | |
| "loss": 0.9848, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.9921731352806091, | |
| "learning_rate": 1.7545068824448255e-05, | |
| "loss": 0.9973, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.9273267388343811, | |
| "learning_rate": 1.7499271846702216e-05, | |
| "loss": 0.9849, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.8230627179145813, | |
| "learning_rate": 1.7453112592405245e-05, | |
| "loss": 1.0165, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.2466211318969727, | |
| "learning_rate": 1.740659329142922e-05, | |
| "loss": 0.9238, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.9236559271812439, | |
| "learning_rate": 1.7359716191039248e-05, | |
| "loss": 1.0209, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.255590558052063, | |
| "learning_rate": 1.7312483555785087e-05, | |
| "loss": 0.9844, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 1.4753785133361816, | |
| "learning_rate": 1.7264897667391757e-05, | |
| "loss": 0.9471, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.6179955005645752, | |
| "learning_rate": 1.7216960824649304e-05, | |
| "loss": 0.9971, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.8780351877212524, | |
| "learning_rate": 1.7168675343301768e-05, | |
| "loss": 0.9777, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.8617651462554932, | |
| "learning_rate": 1.71200435559353e-05, | |
| "loss": 0.9988, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.7536336183547974, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 1.0015, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 1.1492656469345093, | |
| "learning_rate": 1.7021750477023823e-05, | |
| "loss": 0.9303, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 1.0156809091567993, | |
| "learning_rate": 1.69720939338435e-05, | |
| "loss": 1.0139, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.9024391770362854, | |
| "learning_rate": 1.6922100581144228e-05, | |
| "loss": 0.9492, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.6689717173576355, | |
| "learning_rate": 1.6871772834016406e-05, | |
| "loss": 1.0206, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 1.6621466875076294, | |
| "learning_rate": 1.6821113123704425e-05, | |
| "loss": 0.9274, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1.6041533946990967, | |
| "learning_rate": 1.677012389748923e-05, | |
| "loss": 0.9442, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.7168300747871399, | |
| "learning_rate": 1.671880761857011e-05, | |
| "loss": 1.0304, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.935062825679779, | |
| "learning_rate": 1.666716676594567e-05, | |
| "loss": 0.9373, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.6920915246009827, | |
| "learning_rate": 1.661520383429412e-05, | |
| "loss": 1.0088, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.7663747668266296, | |
| "learning_rate": 1.6562921333852714e-05, | |
| "loss": 0.9647, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.7997227311134338, | |
| "learning_rate": 1.6510321790296527e-05, | |
| "loss": 1.0063, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.8207136392593384, | |
| "learning_rate": 1.6457407744616417e-05, | |
| "loss": 0.9139, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.7844796776771545, | |
| "learning_rate": 1.6404181752996287e-05, | |
| "loss": 1.0004, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.9199810028076172, | |
| "learning_rate": 1.6350646386689593e-05, | |
| "loss": 0.9681, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.7962723970413208, | |
| "learning_rate": 1.629680423189514e-05, | |
| "loss": 0.9528, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.8203385472297668, | |
| "learning_rate": 1.6242657889632133e-05, | |
| "loss": 0.9788, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.565852701663971, | |
| "learning_rate": 1.618820997561454e-05, | |
| "loss": 1.0151, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.833648681640625, | |
| "learning_rate": 1.613346312012473e-05, | |
| "loss": 0.9924, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.7219722867012024, | |
| "learning_rate": 1.6078419967886402e-05, | |
| "loss": 0.9684, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.0238351821899414, | |
| "learning_rate": 1.6023083177936824e-05, | |
| "loss": 0.9269, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.6105848550796509, | |
| "learning_rate": 1.5967455423498387e-05, | |
| "loss": 1.0133, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.7665531039237976, | |
| "learning_rate": 1.591153939184946e-05, | |
| "loss": 0.9752, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.6332923769950867, | |
| "learning_rate": 1.5855337784194576e-05, | |
| "loss": 0.9646, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.6672260761260986, | |
| "learning_rate": 1.5798853315533932e-05, | |
| "loss": 0.9162, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.6032528281211853, | |
| "learning_rate": 1.5742088714532247e-05, | |
| "loss": 1.0412, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.6922709941864014, | |
| "learning_rate": 1.568504672338694e-05, | |
| "loss": 0.9336, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.5953813195228577, | |
| "learning_rate": 1.562773009769564e-05, | |
| "loss": 1.0065, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.8393539786338806, | |
| "learning_rate": 1.5570141606323105e-05, | |
| "loss": 0.9293, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.5988898277282715, | |
| "learning_rate": 1.551228403126744e-05, | |
| "loss": 0.9865, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.8385195732116699, | |
| "learning_rate": 1.5454160167525688e-05, | |
| "loss": 0.9353, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.5348495841026306, | |
| "learning_rate": 1.5395772822958844e-05, | |
| "loss": 1.0111, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.8932000994682312, | |
| "learning_rate": 1.5337124818156203e-05, | |
| "loss": 0.8873, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.6047860980033875, | |
| "learning_rate": 1.5278218986299074e-05, | |
| "loss": 0.9815, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.7165077924728394, | |
| "learning_rate": 1.5219058173023948e-05, | |
| "loss": 0.9793, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.6991870403289795, | |
| "learning_rate": 1.515964523628501e-05, | |
| "loss": 0.9357, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.7123513221740723, | |
| "learning_rate": 1.5099983046216089e-05, | |
| "loss": 0.973, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.7146787047386169, | |
| "learning_rate": 1.5040074484992e-05, | |
| "loss": 0.9489, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.6027345657348633, | |
| "learning_rate": 1.4979922446689308e-05, | |
| "loss": 1.0027, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.8369460701942444, | |
| "learning_rate": 1.4919529837146529e-05, | |
| "loss": 0.8835, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.5643919110298157, | |
| "learning_rate": 1.4858899573823752e-05, | |
| "loss": 0.9793, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.6556090712547302, | |
| "learning_rate": 1.4798034585661696e-05, | |
| "loss": 0.9801, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.6771051287651062, | |
| "learning_rate": 1.4736937812940217e-05, | |
| "loss": 0.955, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.6701157689094543, | |
| "learning_rate": 1.4675612207136283e-05, | |
| "loss": 0.9176, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.7202320098876953, | |
| "learning_rate": 1.4614060730781377e-05, | |
| "loss": 0.9585, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.5713448524475098, | |
| "learning_rate": 1.455228635731839e-05, | |
| "loss": 0.9742, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.7807315587997437, | |
| "learning_rate": 1.4490292070957978e-05, | |
| "loss": 0.9419, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.5524024963378906, | |
| "learning_rate": 1.4428080866534397e-05, | |
| "loss": 0.97, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.6541144251823425, | |
| "learning_rate": 1.4365655749360833e-05, | |
| "loss": 1.0058, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.7000547051429749, | |
| "learning_rate": 1.4303019735084225e-05, | |
| "loss": 0.9745, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.6041190028190613, | |
| "learning_rate": 1.4240175849539566e-05, | |
| "loss": 0.8899, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.5961765646934509, | |
| "learning_rate": 1.4177127128603748e-05, | |
| "loss": 1.0048, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.756955087184906, | |
| "learning_rate": 1.4113876618048896e-05, | |
| "loss": 0.9451, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.6756475567817688, | |
| "learning_rate": 1.4050427373395241e-05, | |
| "loss": 0.8928, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.6327915191650391, | |
| "learning_rate": 1.3986782459763499e-05, | |
| "loss": 0.9559, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.6087237596511841, | |
| "learning_rate": 1.3922944951726811e-05, | |
| "loss": 0.9261, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.5963519811630249, | |
| "learning_rate": 1.3858917933162212e-05, | |
| "loss": 0.9788, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.6132527589797974, | |
| "learning_rate": 1.3794704497101656e-05, | |
| "loss": 0.9418, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.5518594980239868, | |
| "learning_rate": 1.3730307745582594e-05, | |
| "loss": 0.9749, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.6775579452514648, | |
| "learning_rate": 1.366573078949813e-05, | |
| "loss": 0.8951, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.5314546227455139, | |
| "learning_rate": 1.3600976748446722e-05, | |
| "loss": 0.9779, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.6605689525604248, | |
| "learning_rate": 1.3536048750581494e-05, | |
| "loss": 0.9186, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.641006350517273, | |
| "learning_rate": 1.3470949932459116e-05, | |
| "loss": 0.9264, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.5755122900009155, | |
| "learning_rate": 1.3405683438888281e-05, | |
| "loss": 0.9537, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.5418959259986877, | |
| "learning_rate": 1.3340252422777788e-05, | |
| "loss": 0.9863, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.7049468755722046, | |
| "learning_rate": 1.3274660044984225e-05, | |
| "loss": 0.9764, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.7064807415008545, | |
| "learning_rate": 1.3208909474159279e-05, | |
| "loss": 0.9212, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.5477349758148193, | |
| "learning_rate": 1.314300388659667e-05, | |
| "loss": 0.9786, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.623717188835144, | |
| "learning_rate": 1.3076946466078691e-05, | |
| "loss": 0.9293, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.5747312307357788, | |
| "learning_rate": 1.301074040372242e-05, | |
| "loss": 0.9639, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.6373106241226196, | |
| "learning_rate": 1.2944388897825559e-05, | |
| "loss": 0.9142, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.6176166534423828, | |
| "learning_rate": 1.2877895153711935e-05, | |
| "loss": 0.939, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.593360960483551, | |
| "learning_rate": 1.2811262383576646e-05, | |
| "loss": 1.0074, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.7040067911148071, | |
| "learning_rate": 1.274449380633089e-05, | |
| "loss": 0.8878, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.5974382758140564, | |
| "learning_rate": 1.2677592647446472e-05, | |
| "loss": 0.9607, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.6244839429855347, | |
| "learning_rate": 1.2610562138799977e-05, | |
| "loss": 0.9581, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.6047521829605103, | |
| "learning_rate": 1.2543405518516651e-05, | |
| "loss": 0.9549, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.6225395798683167, | |
| "learning_rate": 1.2476126030813964e-05, | |
| "loss": 0.9173, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.6194489002227783, | |
| "learning_rate": 1.24087269258449e-05, | |
| "loss": 0.9653, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.6908699870109558, | |
| "learning_rate": 1.234121145954094e-05, | |
| "loss": 0.9235, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.6953298449516296, | |
| "learning_rate": 1.2273582893454774e-05, | |
| "loss": 0.9036, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.580132007598877, | |
| "learning_rate": 1.2205844494602741e-05, | |
| "loss": 0.9565, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.6095865368843079, | |
| "learning_rate": 1.213799953530701e-05, | |
| "loss": 0.9322, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.5975431799888611, | |
| "learning_rate": 1.2070051293037493e-05, | |
| "loss": 0.9694, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.9315632581710815, | |
| "learning_rate": 1.2002003050253524e-05, | |
| "loss": 0.8799, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.56451016664505, | |
| "learning_rate": 1.1933858094245281e-05, | |
| "loss": 0.9988, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.8228814601898193, | |
| "learning_rate": 1.1865619716974986e-05, | |
| "loss": 0.914, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.5807238221168518, | |
| "learning_rate": 1.1797291214917882e-05, | |
| "loss": 0.9723, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.6809439659118652, | |
| "learning_rate": 1.1728875888902975e-05, | |
| "loss": 0.8911, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.6201900243759155, | |
| "learning_rate": 1.1660377043953588e-05, | |
| "loss": 0.9172, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.5941144227981567, | |
| "learning_rate": 1.1591797989127691e-05, | |
| "loss": 0.9477, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.6052290797233582, | |
| "learning_rate": 1.152314203735805e-05, | |
| "loss": 0.9736, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.6165608763694763, | |
| "learning_rate": 1.14544125052922e-05, | |
| "loss": 0.9087, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.6176548600196838, | |
| "learning_rate": 1.1385612713132191e-05, | |
| "loss": 0.9396, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.5738443732261658, | |
| "learning_rate": 1.1316745984474227e-05, | |
| "loss": 0.94, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.8920817375183105, | |
| "learning_rate": 1.1247815646148088e-05, | |
| "loss": 0.9924, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6532885432243347, | |
| "learning_rate": 1.117882502805643e-05, | |
| "loss": 0.9591, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6352014541625977, | |
| "learning_rate": 1.1109777463013915e-05, | |
| "loss": 0.8678, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6214503645896912, | |
| "learning_rate": 1.1040676286586212e-05, | |
| "loss": 0.9255, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.5152882933616638, | |
| "learning_rate": 1.097152483692886e-05, | |
| "loss": 1.0121, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.6979736089706421, | |
| "learning_rate": 1.0902326454626012e-05, | |
| "loss": 0.873, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.626701831817627, | |
| "learning_rate": 1.0833084482529048e-05, | |
| "loss": 0.9935, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.6194130182266235, | |
| "learning_rate": 1.0763802265595103e-05, | |
| "loss": 0.9181, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.531891405582428, | |
| "learning_rate": 1.0694483150725458e-05, | |
| "loss": 1.0178, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.6141016483306885, | |
| "learning_rate": 1.0625130486603879e-05, | |
| "loss": 0.9174, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.6764979362487793, | |
| "learning_rate": 1.055574762353483e-05, | |
| "loss": 0.892, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.554815411567688, | |
| "learning_rate": 1.0486337913281633e-05, | |
| "loss": 0.9678, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.6804779171943665, | |
| "learning_rate": 1.041690470890455e-05, | |
| "loss": 0.877, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.549809455871582, | |
| "learning_rate": 1.0347451364598805e-05, | |
| "loss": 0.9467, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.5851323008537292, | |
| "learning_rate": 1.0277981235532541e-05, | |
| "loss": 0.946, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.6090149283409119, | |
| "learning_rate": 1.0208497677684755e-05, | |
| "loss": 0.9357, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.6555798649787903, | |
| "learning_rate": 1.0139004047683152e-05, | |
| "loss": 0.8908, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.614620566368103, | |
| "learning_rate": 1.0069503702642011e-05, | |
| "loss": 0.9503, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.642510712146759, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9285, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.6009228229522705, | |
| "learning_rate": 9.930496297357994e-06, | |
| "loss": 0.9688, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.5821659564971924, | |
| "learning_rate": 9.860995952316851e-06, | |
| "loss": 0.8912, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.6112207770347595, | |
| "learning_rate": 9.791502322315249e-06, | |
| "loss": 0.9411, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.6378392577171326, | |
| "learning_rate": 9.72201876446746e-06, | |
| "loss": 0.9362, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.601566731929779, | |
| "learning_rate": 9.6525486354012e-06, | |
| "loss": 0.8635, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.5216711163520813, | |
| "learning_rate": 9.583095291095454e-06, | |
| "loss": 1.014, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.6568646430969238, | |
| "learning_rate": 9.513662086718372e-06, | |
| "loss": 0.8845, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.58800208568573, | |
| "learning_rate": 9.444252376465171e-06, | |
| "loss": 0.9404, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.5654013752937317, | |
| "learning_rate": 9.374869513396123e-06, | |
| "loss": 0.9399, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.5590551495552063, | |
| "learning_rate": 9.305516849274542e-06, | |
| "loss": 0.946, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.5611296892166138, | |
| "learning_rate": 9.2361977344049e-06, | |
| "loss": 0.9554, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.6075342893600464, | |
| "learning_rate": 9.166915517470953e-06, | |
| "loss": 0.9458, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.5304029583930969, | |
| "learning_rate": 9.09767354537399e-06, | |
| "loss": 0.9857, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.714830756187439, | |
| "learning_rate": 9.028475163071142e-06, | |
| "loss": 0.88, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.552518904209137, | |
| "learning_rate": 8.959323713413792e-06, | |
| "loss": 0.904, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.6133166551589966, | |
| "learning_rate": 8.890222536986085e-06, | |
| "loss": 0.9453, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.6771183609962463, | |
| "learning_rate": 8.821174971943573e-06, | |
| "loss": 0.964, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.656585156917572, | |
| "learning_rate": 8.752184353851917e-06, | |
| "loss": 0.8962, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.6072052717208862, | |
| "learning_rate": 8.683254015525776e-06, | |
| "loss": 0.9071, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.6188064217567444, | |
| "learning_rate": 8.614387286867814e-06, | |
| "loss": 0.9336, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.5567412972450256, | |
| "learning_rate": 8.545587494707803e-06, | |
| "loss": 0.9206, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.5682706832885742, | |
| "learning_rate": 8.476857962641951e-06, | |
| "loss": 0.91, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.601051390171051, | |
| "learning_rate": 8.408202010872312e-06, | |
| "loss": 0.9073, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.5651359558105469, | |
| "learning_rate": 8.339622956046417e-06, | |
| "loss": 0.9497, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.6029815673828125, | |
| "learning_rate": 8.271124111097026e-06, | |
| "loss": 0.8832, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.6452885866165161, | |
| "learning_rate": 8.202708785082122e-06, | |
| "loss": 0.9312, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.6445124745368958, | |
| "learning_rate": 8.134380283025014e-06, | |
| "loss": 0.9526, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.6829676032066345, | |
| "learning_rate": 8.066141905754724e-06, | |
| "loss": 0.9279, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.6220987439155579, | |
| "learning_rate": 7.997996949746478e-06, | |
| "loss": 0.9267, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.6775074601173401, | |
| "learning_rate": 7.929948706962508e-06, | |
| "loss": 0.9473, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.6403180956840515, | |
| "learning_rate": 7.862000464692992e-06, | |
| "loss": 0.9184, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.5521963834762573, | |
| "learning_rate": 7.79415550539726e-06, | |
| "loss": 0.9035, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.5692378878593445, | |
| "learning_rate": 7.726417106545231e-06, | |
| "loss": 0.9504, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.727057695388794, | |
| "learning_rate": 7.658788540459063e-06, | |
| "loss": 0.9145, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.5626967549324036, | |
| "learning_rate": 7.5912730741551044e-06, | |
| "loss": 0.9331, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.6037367582321167, | |
| "learning_rate": 7.523873969186039e-06, | |
| "loss": 0.9423, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.6398015022277832, | |
| "learning_rate": 7.456594481483355e-06, | |
| "loss": 0.9245, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.5678814649581909, | |
| "learning_rate": 7.389437861200024e-06, | |
| "loss": 0.9295, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.6217985153198242, | |
| "learning_rate": 7.322407352553529e-06, | |
| "loss": 0.8704, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.5044295191764832, | |
| "learning_rate": 7.2555061936691104e-06, | |
| "loss": 1.0015, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.6049959063529968, | |
| "learning_rate": 7.188737616423357e-06, | |
| "loss": 0.8575, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.6413652896881104, | |
| "learning_rate": 7.122104846288065e-06, | |
| "loss": 0.9465, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.5489308834075928, | |
| "learning_rate": 7.055611102174442e-06, | |
| "loss": 0.957, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.6324492692947388, | |
| "learning_rate": 6.9892595962775826e-06, | |
| "loss": 0.8461, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.5411304235458374, | |
| "learning_rate": 6.923053533921312e-06, | |
| "loss": 0.9513, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.5741369724273682, | |
| "learning_rate": 6.85699611340333e-06, | |
| "loss": 0.9126, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.5294156074523926, | |
| "learning_rate": 6.791090525840722e-06, | |
| "loss": 0.9495, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.5881220698356628, | |
| "learning_rate": 6.725339955015777e-06, | |
| "loss": 0.8817, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.5774428844451904, | |
| "learning_rate": 6.659747577222215e-06, | |
| "loss": 0.9236, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.5695903301239014, | |
| "learning_rate": 6.5943165611117244e-06, | |
| "loss": 0.9682, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.5730464458465576, | |
| "learning_rate": 6.529050067540887e-06, | |
| "loss": 0.9206, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.5845257639884949, | |
| "learning_rate": 6.4639512494185104e-06, | |
| "loss": 0.8612, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.5975412130355835, | |
| "learning_rate": 6.39902325155328e-06, | |
| "loss": 0.949, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.5623159408569336, | |
| "learning_rate": 6.334269210501876e-06, | |
| "loss": 0.9112, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.5416470766067505, | |
| "learning_rate": 6.269692254417408e-06, | |
| "loss": 0.9329, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5556077361106873, | |
| "learning_rate": 6.205295502898348e-06, | |
| "loss": 0.9605, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.6438037753105164, | |
| "learning_rate": 6.141082066837791e-06, | |
| "loss": 0.8652, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5236561894416809, | |
| "learning_rate": 6.077055048273193e-06, | |
| "loss": 0.9613, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5974493622779846, | |
| "learning_rate": 6.013217540236503e-06, | |
| "loss": 0.9125, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5160361528396606, | |
| "learning_rate": 5.9495726266047605e-06, | |
| "loss": 0.9862, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.6102047562599182, | |
| "learning_rate": 5.886123381951103e-06, | |
| "loss": 0.8526, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.592778742313385, | |
| "learning_rate": 5.822872871396255e-06, | |
| "loss": 0.9067, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.5523326396942139, | |
| "learning_rate": 5.759824150460436e-06, | |
| "loss": 0.9467, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.5588356852531433, | |
| "learning_rate": 5.696980264915777e-06, | |
| "loss": 0.9123, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.5231946706771851, | |
| "learning_rate": 5.63434425063917e-06, | |
| "loss": 0.92, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.9351530075073242, | |
| "learning_rate": 5.571919133465605e-06, | |
| "loss": 0.9395, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.5406903624534607, | |
| "learning_rate": 5.50970792904203e-06, | |
| "loss": 0.9467, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.5286233425140381, | |
| "learning_rate": 5.447713642681612e-06, | |
| "loss": 0.9241, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.6577736735343933, | |
| "learning_rate": 5.3859392692186256e-06, | |
| "loss": 0.9526, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.576171875, | |
| "learning_rate": 5.324387792863719e-06, | |
| "loss": 0.8893, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.5532373785972595, | |
| "learning_rate": 5.263062187059785e-06, | |
| "loss": 0.9053, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.5663406848907471, | |
| "learning_rate": 5.201965414338308e-06, | |
| "loss": 0.9293, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.6376209855079651, | |
| "learning_rate": 5.14110042617625e-06, | |
| "loss": 0.9127, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.5396673679351807, | |
| "learning_rate": 5.080470162853473e-06, | |
| "loss": 0.9618, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.6800861358642578, | |
| "learning_rate": 5.020077553310694e-06, | |
| "loss": 0.8498, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.5277402997016907, | |
| "learning_rate": 4.959925515008003e-06, | |
| "loss": 0.9428, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.5845181345939636, | |
| "learning_rate": 4.9000169537839126e-06, | |
| "loss": 0.9225, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.5695692300796509, | |
| "learning_rate": 4.840354763714991e-06, | |
| "loss": 0.9236, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.596485435962677, | |
| "learning_rate": 4.780941826976054e-06, | |
| "loss": 0.8826, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.5882150530815125, | |
| "learning_rate": 4.721781013700928e-06, | |
| "loss": 0.907, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.573978066444397, | |
| "learning_rate": 4.662875181843799e-06, | |
| "loss": 0.9495, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.6377553343772888, | |
| "learning_rate": 4.604227177041156e-06, | |
| "loss": 0.8496, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.5344380736351013, | |
| "learning_rate": 4.545839832474318e-06, | |
| "loss": 0.9197, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.5722324252128601, | |
| "learning_rate": 4.487715968732568e-06, | |
| "loss": 0.9379, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.525770366191864, | |
| "learning_rate": 4.429858393676898e-06, | |
| "loss": 0.9349, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.5412399768829346, | |
| "learning_rate": 4.3722699023043634e-06, | |
| "loss": 0.8609, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.5131021738052368, | |
| "learning_rate": 4.314953276613066e-06, | |
| "loss": 0.9974, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.5758804678916931, | |
| "learning_rate": 4.257911285467754e-06, | |
| "loss": 0.9081, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.6420027613639832, | |
| "learning_rate": 4.201146684466065e-06, | |
| "loss": 0.9422, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.5016642808914185, | |
| "learning_rate": 4.144662215805426e-06, | |
| "loss": 0.9642, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.5681389570236206, | |
| "learning_rate": 4.088460608150537e-06, | |
| "loss": 0.8824, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.5395475029945374, | |
| "learning_rate": 4.0325445765016145e-06, | |
| "loss": 0.928, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.6365355253219604, | |
| "learning_rate": 3.9769168220631745e-06, | |
| "loss": 0.9289, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.543300986289978, | |
| "learning_rate": 3.921580032113602e-06, | |
| "loss": 0.9404, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5718982815742493, | |
| "learning_rate": 3.866536879875269e-06, | |
| "loss": 0.9252, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5609005689620972, | |
| "learning_rate": 3.81179002438546e-06, | |
| "loss": 0.929, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5835498571395874, | |
| "learning_rate": 3.7573421103678707e-06, | |
| "loss": 0.9106, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.6111095547676086, | |
| "learning_rate": 3.7031957681048604e-06, | |
| "loss": 0.8461, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.5231614112854004, | |
| "learning_rate": 3.649353613310409e-06, | |
| "loss": 0.9542, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.5488110780715942, | |
| "learning_rate": 3.5958182470037127e-06, | |
| "loss": 0.9111, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.5633453726768494, | |
| "learning_rate": 3.5425922553835866e-06, | |
| "loss": 0.9417, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.5549660325050354, | |
| "learning_rate": 3.4896782097034755e-06, | |
| "loss": 0.8683, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.4999479055404663, | |
| "learning_rate": 3.4370786661472922e-06, | |
| "loss": 0.9473, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.5905501246452332, | |
| "learning_rate": 3.384796165705885e-06, | |
| "loss": 0.8959, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.5118279457092285, | |
| "learning_rate": 3.3328332340543314e-06, | |
| "loss": 0.9547, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.5384517312049866, | |
| "learning_rate": 3.281192381429894e-06, | |
| "loss": 0.8385, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.5861096382141113, | |
| "learning_rate": 3.2298761025107707e-06, | |
| "loss": 0.9305, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.5211436152458191, | |
| "learning_rate": 3.178886876295578e-06, | |
| "loss": 0.9697, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.5872800350189209, | |
| "learning_rate": 3.128227165983595e-06, | |
| "loss": 0.8571, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.49042809009552, | |
| "learning_rate": 3.0778994188557722e-06, | |
| "loss": 0.974, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.5632616281509399, | |
| "learning_rate": 3.027906066156503e-06, | |
| "loss": 0.9143, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.502046525478363, | |
| "learning_rate": 2.978249522976181e-06, | |
| "loss": 0.922, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.5344391465187073, | |
| "learning_rate": 2.9289321881345257e-06, | |
| "loss": 0.9108, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.5567865371704102, | |
| "learning_rate": 2.879956444064703e-06, | |
| "loss": 0.9269, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.5362602472305298, | |
| "learning_rate": 2.8313246566982342e-06, | |
| "loss": 0.8992, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.5565657615661621, | |
| "learning_rate": 2.783039175350699e-06, | |
| "loss": 0.8796, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.590705394744873, | |
| "learning_rate": 2.735102332608247e-06, | |
| "loss": 0.9105, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.49303296208381653, | |
| "learning_rate": 2.6875164442149147e-06, | |
| "loss": 0.9494, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.5503098964691162, | |
| "learning_rate": 2.640283808960754e-06, | |
| "loss": 0.9275, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.5888702273368835, | |
| "learning_rate": 2.5934067085707835e-06, | |
| "loss": 0.907, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.5702516436576843, | |
| "learning_rate": 2.54688740759476e-06, | |
| "loss": 0.9394, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.5794321298599243, | |
| "learning_rate": 2.500728153297788e-06, | |
| "loss": 0.8737, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.5498908758163452, | |
| "learning_rate": 2.454931175551746e-06, | |
| "loss": 0.9329, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.6707480549812317, | |
| "learning_rate": 2.409498686727587e-06, | |
| "loss": 0.9386, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.6271098256111145, | |
| "learning_rate": 2.364432881588431e-06, | |
| "loss": 0.8743, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.5068037509918213, | |
| "learning_rate": 2.3197359371835802e-06, | |
| "loss": 0.9495, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.5746956467628479, | |
| "learning_rate": 2.2754100127433033e-06, | |
| "loss": 0.8866, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.5277212262153625, | |
| "learning_rate": 2.2314572495745746e-06, | |
| "loss": 0.9685, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.5652911067008972, | |
| "learning_rate": 2.187879770957585e-06, | |
| "loss": 0.8458, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.5308396220207214, | |
| "learning_rate": 2.144679682043217e-06, | |
| "loss": 0.9255, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.5504590272903442, | |
| "learning_rate": 2.1018590697513007e-06, | |
| "loss": 0.9181, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.5485373139381409, | |
| "learning_rate": 2.0594200026698363e-06, | |
| "loss": 0.9407, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.56805419921875, | |
| "learning_rate": 2.017364530955055e-06, | |
| "loss": 0.8464, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.529406726360321, | |
| "learning_rate": 1.9756946862323534e-06, | |
| "loss": 0.9606, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.5330740809440613, | |
| "learning_rate": 1.934412481498198e-06, | |
| "loss": 0.9426, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.5351034998893738, | |
| "learning_rate": 1.8935199110228274e-06, | |
| "loss": 0.8774, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.5133355259895325, | |
| "learning_rate": 1.8530189502539608e-06, | |
| "loss": 0.9277, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.5475199222564697, | |
| "learning_rate": 1.8129115557213262e-06, | |
| "loss": 0.922, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.5252749919891357, | |
| "learning_rate": 1.77319966494218e-06, | |
| "loss": 0.9593, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.5665723085403442, | |
| "learning_rate": 1.7338851963276827e-06, | |
| "loss": 0.864, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.49674439430236816, | |
| "learning_rate": 1.6949700490902344e-06, | |
| "loss": 0.9284, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.5066332221031189, | |
| "learning_rate": 1.6564561031517278e-06, | |
| "loss": 0.9416, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.4922288656234741, | |
| "learning_rate": 1.6183452190527317e-06, | |
| "loss": 0.9211, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.527193546295166, | |
| "learning_rate": 1.5806392378626079e-06, | |
| "loss": 0.915, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.5450595021247864, | |
| "learning_rate": 1.543339981090578e-06, | |
| "loss": 0.9326, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.5060311555862427, | |
| "learning_rate": 1.5064492505977234e-06, | |
| "loss": 0.9191, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.5316381454467773, | |
| "learning_rate": 1.4699688285099489e-06, | |
| "loss": 0.8768, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.5138689875602722, | |
| "learning_rate": 1.433900477131882e-06, | |
| "loss": 0.9311, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.5076268315315247, | |
| "learning_rate": 1.3982459388617453e-06, | |
| "loss": 0.9368, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.4838371276855469, | |
| "learning_rate": 1.363006936107183e-06, | |
| "loss": 0.9182, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.5362660884857178, | |
| "learning_rate": 1.3281851712020522e-06, | |
| "loss": 0.8731, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.4921913743019104, | |
| "learning_rate": 1.29378232632419e-06, | |
| "loss": 0.9548, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.5297012329101562, | |
| "learning_rate": 1.259800063414146e-06, | |
| "loss": 0.8701, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.5477970838546753, | |
| "learning_rate": 1.2262400240949023e-06, | |
| "loss": 0.8565, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.5233986377716064, | |
| "learning_rate": 1.1931038295925646e-06, | |
| "loss": 0.9665, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.5435713529586792, | |
| "learning_rate": 1.1603930806580443e-06, | |
| "loss": 0.8954, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.48413795232772827, | |
| "learning_rate": 1.128109357489734e-06, | |
| "loss": 0.9547, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.5370103716850281, | |
| "learning_rate": 1.0962542196571636e-06, | |
| "loss": 0.9077, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.5304175615310669, | |
| "learning_rate": 1.064829206025665e-06, | |
| "loss": 0.9014, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.5705199241638184, | |
| "learning_rate": 1.0338358346820355e-06, | |
| "loss": 0.8509, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.5038979649543762, | |
| "learning_rate": 1.003275602861188e-06, | |
| "loss": 0.9482, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.5279713869094849, | |
| "learning_rate": 9.731499868738448e-07, | |
| "loss": 0.9416, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.5886192321777344, | |
| "learning_rate": 9.434604420351912e-07, | |
| "loss": 0.9102, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.5839190483093262, | |
| "learning_rate": 9.142084025945986e-07, | |
| "loss": 0.8602, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.4787314236164093, | |
| "learning_rate": 8.853952816663214e-07, | |
| "loss": 0.984, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.673080563545227, | |
| "learning_rate": 8.570224711612385e-07, | |
| "loss": 0.9083, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.48418405652046204, | |
| "learning_rate": 8.290913417196178e-07, | |
| "loss": 0.9585, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.48270198702812195, | |
| "learning_rate": 8.016032426448816e-07, | |
| "loss": 0.9076, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.612147867679596, | |
| "learning_rate": 7.745595018384577e-07, | |
| "loss": 0.9241, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.4766865074634552, | |
| "learning_rate": 7.479614257355972e-07, | |
| "loss": 0.9069, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.56040358543396, | |
| "learning_rate": 7.218102992422882e-07, | |
| "loss": 0.8848, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.5449951887130737, | |
| "learning_rate": 6.961073856731648e-07, | |
| "loss": 0.8982, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.49720117449760437, | |
| "learning_rate": 6.708539266905e-07, | |
| "loss": 0.9877, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.6003532409667969, | |
| "learning_rate": 6.460511422441984e-07, | |
| "loss": 0.8369, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.569581151008606, | |
| "learning_rate": 6.21700230512885e-07, | |
| "loss": 0.9456, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.5281058549880981, | |
| "learning_rate": 5.978023678460099e-07, | |
| "loss": 0.8766, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.5012950301170349, | |
| "learning_rate": 5.743587087070235e-07, | |
| "loss": 0.9226, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.5251154899597168, | |
| "learning_rate": 5.513703856176112e-07, | |
| "loss": 0.9043, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.6647759079933167, | |
| "learning_rate": 5.288385091029724e-07, | |
| "loss": 0.9309, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.5760605335235596, | |
| "learning_rate": 5.067641676381918e-07, | |
| "loss": 0.9003, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.49836233258247375, | |
| "learning_rate": 4.851484275956331e-07, | |
| "loss": 0.8489, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.5523092150688171, | |
| "learning_rate": 4.6399233319344703e-07, | |
| "loss": 0.933, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.5302231907844543, | |
| "learning_rate": 4.432969064451109e-07, | |
| "loss": 0.918, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.7435190677642822, | |
| "learning_rate": 4.230631471100655e-07, | |
| "loss": 0.9095, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.6375725865364075, | |
| "learning_rate": 4.0329203264541594e-07, | |
| "loss": 0.9322, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.5424436330795288, | |
| "learning_rate": 3.8398451815870984e-07, | |
| "loss": 0.8929, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.5206416845321655, | |
| "learning_rate": 3.6514153636180384e-07, | |
| "loss": 0.9334, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.4849807918071747, | |
| "learning_rate": 3.467639975257997e-07, | |
| "loss": 0.9472, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.5070081949234009, | |
| "learning_rate": 3.2885278943707524e-07, | |
| "loss": 0.9374, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.6271994709968567, | |
| "learning_rate": 3.114087773543939e-07, | |
| "loss": 0.8506, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.48747432231903076, | |
| "learning_rate": 2.9443280396710847e-07, | |
| "loss": 0.96, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.499573677778244, | |
| "learning_rate": 2.7792568935444796e-07, | |
| "loss": 0.9149, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.5160935521125793, | |
| "learning_rate": 2.618882309459081e-07, | |
| "loss": 0.8342, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.5272057056427002, | |
| "learning_rate": 2.4632120348272e-07, | |
| "loss": 0.94, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.2326080799102783, | |
| "learning_rate": 2.312253589804314e-07, | |
| "loss": 0.8861, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.4588733911514282, | |
| "learning_rate": 2.166014266925731e-07, | |
| "loss": 0.9644, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.5415163636207581, | |
| "learning_rate": 2.0245011307543416e-07, | |
| "loss": 0.9155, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.5331488251686096, | |
| "learning_rate": 1.88772101753929e-07, | |
| "loss": 0.8814, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.579383134841919, | |
| "learning_rate": 1.7556805348858063e-07, | |
| "loss": 0.9047, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.5765702724456787, | |
| "learning_rate": 1.6283860614358936e-07, | |
| "loss": 0.9482, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.5090899467468262, | |
| "learning_rate": 1.5058437465602982e-07, | |
| "loss": 0.8338, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.5124567151069641, | |
| "learning_rate": 1.388059510061379e-07, | |
| "loss": 0.9727, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.5381713509559631, | |
| "learning_rate": 1.2750390418871605e-07, | |
| "loss": 0.8973, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.5176908373832703, | |
| "learning_rate": 1.1667878018564171e-07, | |
| "loss": 0.9471, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5220381617546082, | |
| "learning_rate": 1.063311019395008e-07, | |
| "loss": 0.8995, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5147991180419922, | |
| "learning_rate": 9.64613693283123e-08, | |
| "loss": 0.8977, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.47362008690834045, | |
| "learning_rate": 8.707005914139422e-08, | |
| "loss": 0.9144, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5703770518302917, | |
| "learning_rate": 7.815762505632096e-08, | |
| "loss": 0.8989, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5848727226257324, | |
| "learning_rate": 6.972449761700862e-08, | |
| "loss": 0.9313, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.5251525044441223, | |
| "learning_rate": 6.177108421292266e-08, | |
| "loss": 0.8647, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.5035085082054138, | |
| "learning_rate": 5.429776905938489e-08, | |
| "loss": 0.9014, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.45447394251823425, | |
| "learning_rate": 4.7304913179025967e-08, | |
| "loss": 0.9724, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.5824120044708252, | |
| "learning_rate": 4.0792854384338334e-08, | |
| "loss": 0.8544, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.4841687083244324, | |
| "learning_rate": 3.4761907261356976e-08, | |
| "loss": 0.9547, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.48440462350845337, | |
| "learning_rate": 2.9212363154463853e-08, | |
| "loss": 0.9335, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.495300829410553, | |
| "learning_rate": 2.4144490152313572e-08, | |
| "loss": 0.9332, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.5265560150146484, | |
| "learning_rate": 1.9558533074882646e-08, | |
| "loss": 0.864, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.5560371279716492, | |
| "learning_rate": 1.545471346164007e-08, | |
| "loss": 0.9295, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.5226584672927856, | |
| "learning_rate": 1.1833229560848092e-08, | |
| "loss": 0.907, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.4815405011177063, | |
| "learning_rate": 8.694256319987659e-09, | |
| "loss": 0.9004, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.48337435722351074, | |
| "learning_rate": 6.037945377297405e-09, | |
| "loss": 0.9179, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.5337357521057129, | |
| "learning_rate": 3.8644250544594975e-09, | |
| "loss": 0.9105, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.5046442151069641, | |
| "learning_rate": 2.173800350394606e-09, | |
| "loss": 0.9864, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5206977725028992, | |
| "learning_rate": 9.661529361892907e-10, | |
| "loss": 0.8549, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.48737195134162903, | |
| "learning_rate": 2.415411511536014e-10, | |
| "loss": 0.9553, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.48597776889801025, | |
| "learning_rate": 0.0, | |
| "loss": 0.9239, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 467, | |
| "total_flos": 0.0, | |
| "train_loss": 0.9806490919574681, | |
| "train_runtime": 8015.271, | |
| "train_samples_per_second": 135.547, | |
| "train_steps_per_second": 0.058 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 467, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |