| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.18776338379896987, | |
| "eval_steps": 500, | |
| "global_step": 1203, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.9418012499809265, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.0495, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.9342747330665588, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.1571, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.9167969822883606, | |
| "learning_rate": 6e-06, | |
| "loss": 1.2975, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.9712246656417847, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.7433, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.1027685403823853, | |
| "learning_rate": 1e-05, | |
| "loss": 2.218, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.143164873123169, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.9224, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.8237400650978088, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.767, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.2908293008804321, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.6552, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.9047379493713379, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.9778, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.8817495107650757, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9901, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.2938435077667236, | |
| "learning_rate": 1.9999998794084053e-05, | |
| "loss": 1.7041, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.2303466796875, | |
| "learning_rate": 1.999999517633649e-05, | |
| "loss": 1.6639, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.835405707359314, | |
| "learning_rate": 1.999998914675819e-05, | |
| "loss": 1.6789, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.340865135192871, | |
| "learning_rate": 1.999998070535061e-05, | |
| "loss": 1.7557, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.8470378518104553, | |
| "learning_rate": 1.9999969852115782e-05, | |
| "loss": 1.592, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.643478274345398, | |
| "learning_rate": 1.999995658705632e-05, | |
| "loss": 2.1006, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.2171984910964966, | |
| "learning_rate": 1.9999940910175432e-05, | |
| "loss": 1.9627, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.8794788718223572, | |
| "learning_rate": 1.9999922821476894e-05, | |
| "loss": 2.3324, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.7493646740913391, | |
| "learning_rate": 1.9999902320965067e-05, | |
| "loss": 1.6542, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.8097922205924988, | |
| "learning_rate": 1.99998794086449e-05, | |
| "loss": 2.4314, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.7926909327507019, | |
| "learning_rate": 1.9999854084521913e-05, | |
| "loss": 2.4093, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.8623242378234863, | |
| "learning_rate": 1.999982634860222e-05, | |
| "loss": 2.2189, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.8397309184074402, | |
| "learning_rate": 1.999979620089251e-05, | |
| "loss": 1.8798, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.703669548034668, | |
| "learning_rate": 1.9999763641400046e-05, | |
| "loss": 2.0517, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.7364016771316528, | |
| "learning_rate": 1.999972867013269e-05, | |
| "loss": 2.0208, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.7407152056694031, | |
| "learning_rate": 1.9999691287098876e-05, | |
| "loss": 1.0987, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.8759409785270691, | |
| "learning_rate": 1.999965149230761e-05, | |
| "loss": 1.2877, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.769755482673645, | |
| "learning_rate": 1.9999609285768505e-05, | |
| "loss": 2.0839, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.7315641641616821, | |
| "learning_rate": 1.999956466749173e-05, | |
| "loss": 1.8807, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.77801114320755, | |
| "learning_rate": 1.9999517637488045e-05, | |
| "loss": 2.0659, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.8167088627815247, | |
| "learning_rate": 1.9999468195768802e-05, | |
| "loss": 1.5178, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.7299829721450806, | |
| "learning_rate": 1.9999416342345917e-05, | |
| "loss": 2.1694, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.885439932346344, | |
| "learning_rate": 1.9999362077231898e-05, | |
| "loss": 1.6862, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8192880749702454, | |
| "learning_rate": 1.999930540043984e-05, | |
| "loss": 1.8014, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.1161061525344849, | |
| "learning_rate": 1.9999246311983403e-05, | |
| "loss": 1.9917, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.687031090259552, | |
| "learning_rate": 1.999918481187684e-05, | |
| "loss": 1.7064, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9123612642288208, | |
| "learning_rate": 1.999912090013499e-05, | |
| "loss": 2.1708, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.0114833116531372, | |
| "learning_rate": 1.999905457677326e-05, | |
| "loss": 1.8041, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9185965061187744, | |
| "learning_rate": 1.999898584180765e-05, | |
| "loss": 1.9668, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.0017578601837158, | |
| "learning_rate": 1.9998914695254737e-05, | |
| "loss": 1.8251, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.718401312828064, | |
| "learning_rate": 1.999884113713168e-05, | |
| "loss": 1.9886, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7518029808998108, | |
| "learning_rate": 1.999876516745622e-05, | |
| "loss": 2.0368, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8515375852584839, | |
| "learning_rate": 1.9998686786246682e-05, | |
| "loss": 1.9362, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8468660712242126, | |
| "learning_rate": 1.9998605993521965e-05, | |
| "loss": 1.4887, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7643191814422607, | |
| "learning_rate": 1.999852278930156e-05, | |
| "loss": 1.7073, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.1301417350769043, | |
| "learning_rate": 1.999843717360553e-05, | |
| "loss": 1.6808, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.728996217250824, | |
| "learning_rate": 1.999834914645453e-05, | |
| "loss": 1.8557, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8372696042060852, | |
| "learning_rate": 1.9998258707869787e-05, | |
| "loss": 1.8909, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.06674063205719, | |
| "learning_rate": 1.9998165857873108e-05, | |
| "loss": 1.7721, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.79304039478302, | |
| "learning_rate": 1.99980705964869e-05, | |
| "loss": 1.9567, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7896111607551575, | |
| "learning_rate": 1.9997972923734123e-05, | |
| "loss": 1.7498, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9860767126083374, | |
| "learning_rate": 1.9997872839638347e-05, | |
| "loss": 1.9903, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8512886166572571, | |
| "learning_rate": 1.9997770344223705e-05, | |
| "loss": 1.5083, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7645949721336365, | |
| "learning_rate": 1.9997665437514914e-05, | |
| "loss": 1.6338, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8389264941215515, | |
| "learning_rate": 1.999755811953728e-05, | |
| "loss": 1.6309, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9311525821685791, | |
| "learning_rate": 1.9997448390316686e-05, | |
| "loss": 1.4101, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.6994078755378723, | |
| "learning_rate": 1.9997336249879596e-05, | |
| "loss": 1.4245, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7228380441665649, | |
| "learning_rate": 1.999722169825306e-05, | |
| "loss": 1.4485, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7117276787757874, | |
| "learning_rate": 1.99971047354647e-05, | |
| "loss": 1.3035, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.6175885200500488, | |
| "learning_rate": 1.999698536154273e-05, | |
| "loss": 1.1573, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7536934018135071, | |
| "learning_rate": 1.9996863576515937e-05, | |
| "loss": 1.5875, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7840648889541626, | |
| "learning_rate": 1.9996739380413696e-05, | |
| "loss": 1.5524, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9118802547454834, | |
| "learning_rate": 1.9996612773265963e-05, | |
| "loss": 1.7814, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7639506459236145, | |
| "learning_rate": 1.9996483755103266e-05, | |
| "loss": 2.0718, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7495502233505249, | |
| "learning_rate": 1.9996352325956732e-05, | |
| "loss": 1.3045, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.83675217628479, | |
| "learning_rate": 1.9996218485858054e-05, | |
| "loss": 1.945, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9988659620285034, | |
| "learning_rate": 1.999608223483951e-05, | |
| "loss": 1.8062, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9340213537216187, | |
| "learning_rate": 1.9995943572933963e-05, | |
| "loss": 1.9097, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8286553621292114, | |
| "learning_rate": 1.999580250017486e-05, | |
| "loss": 1.7936, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8728215098381042, | |
| "learning_rate": 1.999565901659622e-05, | |
| "loss": 1.6033, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.5670807361602783, | |
| "learning_rate": 1.9995513122232656e-05, | |
| "loss": 2.462, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.859460711479187, | |
| "learning_rate": 1.999536481711935e-05, | |
| "loss": 2.2517, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.4409862756729126, | |
| "learning_rate": 1.9995214101292065e-05, | |
| "loss": 1.7278, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7657808661460876, | |
| "learning_rate": 1.999506097478716e-05, | |
| "loss": 1.7529, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7684109807014465, | |
| "learning_rate": 1.999490543764157e-05, | |
| "loss": 2.1585, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.765892744064331, | |
| "learning_rate": 1.9994747489892793e-05, | |
| "loss": 1.909, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8307526707649231, | |
| "learning_rate": 1.9994587131578936e-05, | |
| "loss": 1.5767, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.117607593536377, | |
| "learning_rate": 1.999442436273867e-05, | |
| "loss": 1.4128, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7918795347213745, | |
| "learning_rate": 1.9994259183411257e-05, | |
| "loss": 1.5989, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8969283699989319, | |
| "learning_rate": 1.9994091593636527e-05, | |
| "loss": 1.7017, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8903273940086365, | |
| "learning_rate": 1.9993921593454904e-05, | |
| "loss": 1.9551, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9065592885017395, | |
| "learning_rate": 1.9993749182907392e-05, | |
| "loss": 1.8349, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7999714016914368, | |
| "learning_rate": 1.999357436203557e-05, | |
| "loss": 1.39, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.914991557598114, | |
| "learning_rate": 1.9993397130881602e-05, | |
| "loss": 1.3705, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8639907836914062, | |
| "learning_rate": 1.999321748948823e-05, | |
| "loss": 1.694, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.80655437707901, | |
| "learning_rate": 1.9993035437898794e-05, | |
| "loss": 1.3385, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.4766621589660645, | |
| "learning_rate": 1.9992850976157187e-05, | |
| "loss": 1.5354, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.7555109262466431, | |
| "learning_rate": 1.9992664104307904e-05, | |
| "loss": 1.841, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.1581100225448608, | |
| "learning_rate": 1.9992474822396012e-05, | |
| "loss": 1.8417, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9877253770828247, | |
| "learning_rate": 1.999228313046717e-05, | |
| "loss": 2.2302, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.9449608325958252, | |
| "learning_rate": 1.9992089028567604e-05, | |
| "loss": 1.8298, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.0140725374221802, | |
| "learning_rate": 1.999189251674413e-05, | |
| "loss": 1.705, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8601987957954407, | |
| "learning_rate": 1.999169359504414e-05, | |
| "loss": 2.1924, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.1179931163787842, | |
| "learning_rate": 1.999149226351562e-05, | |
| "loss": 1.4521, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.8562244772911072, | |
| "learning_rate": 1.999128852220712e-05, | |
| "loss": 1.8118, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.2716445922851562, | |
| "learning_rate": 1.9991082371167784e-05, | |
| "loss": 1.6766, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.1563775539398193, | |
| "learning_rate": 1.9990873810447326e-05, | |
| "loss": 1.7789, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8975584506988525, | |
| "learning_rate": 1.999066284009605e-05, | |
| "loss": 1.935, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9221625924110413, | |
| "learning_rate": 1.999044946016484e-05, | |
| "loss": 1.5732, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8881098031997681, | |
| "learning_rate": 1.9990233670705157e-05, | |
| "loss": 2.4744, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.6764086484909058, | |
| "learning_rate": 1.9990015471769048e-05, | |
| "loss": 0.7813, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.7852362394332886, | |
| "learning_rate": 1.998979486340914e-05, | |
| "loss": 1.334, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.0338146686553955, | |
| "learning_rate": 1.9989571845678636e-05, | |
| "loss": 1.6507, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.660638689994812, | |
| "learning_rate": 1.998934641863133e-05, | |
| "loss": 1.6307, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9326343536376953, | |
| "learning_rate": 1.9989118582321586e-05, | |
| "loss": 1.8679, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.7414451241493225, | |
| "learning_rate": 1.9988888336804354e-05, | |
| "loss": 1.8633, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.6003090143203735, | |
| "learning_rate": 1.9988655682135173e-05, | |
| "loss": 1.2029, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.214863657951355, | |
| "learning_rate": 1.9988420618370143e-05, | |
| "loss": 1.6799, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8017918467521667, | |
| "learning_rate": 1.998818314556597e-05, | |
| "loss": 1.7695, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.7820467352867126, | |
| "learning_rate": 1.9987943263779917e-05, | |
| "loss": 2.1115, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.233542799949646, | |
| "learning_rate": 1.998770097306985e-05, | |
| "loss": 1.5579, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.075163722038269, | |
| "learning_rate": 1.9987456273494197e-05, | |
| "loss": 1.5154, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9855638742446899, | |
| "learning_rate": 1.9987209165111982e-05, | |
| "loss": 1.4482, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.1536937952041626, | |
| "learning_rate": 1.9986959647982798e-05, | |
| "loss": 1.5155, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.1218782663345337, | |
| "learning_rate": 1.9986707722166825e-05, | |
| "loss": 1.5752, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.0412499904632568, | |
| "learning_rate": 1.998645338772483e-05, | |
| "loss": 1.4374, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.0152816772460938, | |
| "learning_rate": 1.9986196644718143e-05, | |
| "loss": 1.9691, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.48252534866333, | |
| "learning_rate": 1.9985937493208697e-05, | |
| "loss": 1.9283, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9401952624320984, | |
| "learning_rate": 1.998567593325899e-05, | |
| "loss": 1.7933, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.0944910049438477, | |
| "learning_rate": 1.9985411964932104e-05, | |
| "loss": 1.7656, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8697406053543091, | |
| "learning_rate": 1.9985145588291706e-05, | |
| "loss": 2.3612, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.024667739868164, | |
| "learning_rate": 1.9984876803402043e-05, | |
| "loss": 1.6229, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.1907790899276733, | |
| "learning_rate": 1.998460561032794e-05, | |
| "loss": 1.8345, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9052940607070923, | |
| "learning_rate": 1.99843320091348e-05, | |
| "loss": 2.1532, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8829050660133362, | |
| "learning_rate": 1.998405599988862e-05, | |
| "loss": 1.0946, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8987332582473755, | |
| "learning_rate": 1.9983777582655964e-05, | |
| "loss": 1.7491, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8795452117919922, | |
| "learning_rate": 1.998349675750398e-05, | |
| "loss": 1.5047, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.7775564789772034, | |
| "learning_rate": 1.9983213524500396e-05, | |
| "loss": 1.7593, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9722967147827148, | |
| "learning_rate": 1.998292788371353e-05, | |
| "loss": 1.4058, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8291569948196411, | |
| "learning_rate": 1.998263983521227e-05, | |
| "loss": 1.7716, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8410966992378235, | |
| "learning_rate": 1.9982349379066086e-05, | |
| "loss": 2.0283, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9493787884712219, | |
| "learning_rate": 1.998205651534504e-05, | |
| "loss": 1.8034, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8751788139343262, | |
| "learning_rate": 1.9981761244119756e-05, | |
| "loss": 1.3362, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8287643194198608, | |
| "learning_rate": 1.9981463565461452e-05, | |
| "loss": 1.8359, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9401163458824158, | |
| "learning_rate": 1.9981163479441925e-05, | |
| "loss": 2.0059, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.7662250995635986, | |
| "learning_rate": 1.998086098613355e-05, | |
| "loss": 1.1899, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.7782216668128967, | |
| "learning_rate": 1.998055608560928e-05, | |
| "loss": 1.6007, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9448866844177246, | |
| "learning_rate": 1.9980248777942655e-05, | |
| "loss": 2.0743, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.796358585357666, | |
| "learning_rate": 1.9979939063207792e-05, | |
| "loss": 2.2182, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.7981236577033997, | |
| "learning_rate": 1.9979626941479386e-05, | |
| "loss": 2.0773, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8096224665641785, | |
| "learning_rate": 1.997931241283272e-05, | |
| "loss": 1.9287, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9485547542572021, | |
| "learning_rate": 1.9978995477343655e-05, | |
| "loss": 1.6551, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9155763387680054, | |
| "learning_rate": 1.9978676135088623e-05, | |
| "loss": 1.8185, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9059701561927795, | |
| "learning_rate": 1.997835438614465e-05, | |
| "loss": 1.8789, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8805462718009949, | |
| "learning_rate": 1.997803023058933e-05, | |
| "loss": 1.5399, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.797347903251648, | |
| "learning_rate": 1.9977703668500853e-05, | |
| "loss": 1.712, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9459641575813293, | |
| "learning_rate": 1.9977374699957973e-05, | |
| "loss": 1.5353, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.4370256662368774, | |
| "learning_rate": 1.9977043325040033e-05, | |
| "loss": 1.4925, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.3168740272521973, | |
| "learning_rate": 1.997670954382696e-05, | |
| "loss": 1.0978, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9876695275306702, | |
| "learning_rate": 1.9976373356399247e-05, | |
| "loss": 2.9874, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8283455967903137, | |
| "learning_rate": 1.9976034762837987e-05, | |
| "loss": 1.5847, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.004213571548462, | |
| "learning_rate": 1.9975693763224834e-05, | |
| "loss": 2.1154, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9788973331451416, | |
| "learning_rate": 1.997535035764204e-05, | |
| "loss": 1.7769, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8842994570732117, | |
| "learning_rate": 1.9975004546172423e-05, | |
| "loss": 1.4797, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9473729729652405, | |
| "learning_rate": 1.9974656328899387e-05, | |
| "loss": 1.0775, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.217759132385254, | |
| "learning_rate": 1.9974305705906918e-05, | |
| "loss": 1.665, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.258272886276245, | |
| "learning_rate": 1.997395267727958e-05, | |
| "loss": 1.3609, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.9515790343284607, | |
| "learning_rate": 1.9973597243102518e-05, | |
| "loss": 1.9433, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.8606133460998535, | |
| "learning_rate": 1.9973239403461455e-05, | |
| "loss": 1.4242, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.6642533540725708, | |
| "learning_rate": 1.9972879158442696e-05, | |
| "loss": 1.8489, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.7613729238510132, | |
| "learning_rate": 1.997251650813313e-05, | |
| "loss": 1.8288, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9006676077842712, | |
| "learning_rate": 1.9972151452620215e-05, | |
| "loss": 2.406, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.5719473361968994, | |
| "learning_rate": 1.9971783991992005e-05, | |
| "loss": 1.2342, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9287020564079285, | |
| "learning_rate": 1.9971414126337117e-05, | |
| "loss": 2.0854, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0734272003173828, | |
| "learning_rate": 1.9971041855744764e-05, | |
| "loss": 1.89, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8500266671180725, | |
| "learning_rate": 1.9970667180304727e-05, | |
| "loss": 1.8576, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8321697115898132, | |
| "learning_rate": 1.9970290100107372e-05, | |
| "loss": 2.0847, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.739933967590332, | |
| "learning_rate": 1.996991061524364e-05, | |
| "loss": 1.9846, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9269589185714722, | |
| "learning_rate": 1.9969528725805066e-05, | |
| "loss": 1.2928, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9688999652862549, | |
| "learning_rate": 1.996914443188375e-05, | |
| "loss": 1.1562, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8473265171051025, | |
| "learning_rate": 1.9968757733572372e-05, | |
| "loss": 1.3455, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0271553993225098, | |
| "learning_rate": 1.9968368630964206e-05, | |
| "loss": 1.7439, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.1127710342407227, | |
| "learning_rate": 1.9967977124153095e-05, | |
| "loss": 1.4497, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0394861698150635, | |
| "learning_rate": 1.996758321323346e-05, | |
| "loss": 1.8584, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.2190815210342407, | |
| "learning_rate": 1.9967186898300307e-05, | |
| "loss": 1.9396, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9696542024612427, | |
| "learning_rate": 1.9966788179449223e-05, | |
| "loss": 2.2817, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0213602781295776, | |
| "learning_rate": 1.996638705677637e-05, | |
| "loss": 1.559, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.031296730041504, | |
| "learning_rate": 1.9965983530378498e-05, | |
| "loss": 1.6529, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9905886650085449, | |
| "learning_rate": 1.996557760035292e-05, | |
| "loss": 1.7384, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8480849266052246, | |
| "learning_rate": 1.9965169266797545e-05, | |
| "loss": 2.0596, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.219632387161255, | |
| "learning_rate": 1.996475852981086e-05, | |
| "loss": 1.2584, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9293888211250305, | |
| "learning_rate": 1.996434538949192e-05, | |
| "loss": 2.3078, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9126909971237183, | |
| "learning_rate": 1.9963929845940376e-05, | |
| "loss": 1.517, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9666692018508911, | |
| "learning_rate": 1.996351189925644e-05, | |
| "loss": 1.9152, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8901292085647583, | |
| "learning_rate": 1.9963091549540926e-05, | |
| "loss": 1.3486, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.6530977487564087, | |
| "learning_rate": 1.9962668796895207e-05, | |
| "loss": 1.2948, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.3919610977172852, | |
| "learning_rate": 1.9962243641421243e-05, | |
| "loss": 1.3113, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8002163767814636, | |
| "learning_rate": 1.996181608322158e-05, | |
| "loss": 1.4723, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.7906051874160767, | |
| "learning_rate": 1.9961386122399334e-05, | |
| "loss": 1.1714, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.6313331127166748, | |
| "learning_rate": 1.9960953759058204e-05, | |
| "loss": 1.0104, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.325727105140686, | |
| "learning_rate": 1.9960518993302468e-05, | |
| "loss": 1.7907, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8740698099136353, | |
| "learning_rate": 1.996008182523699e-05, | |
| "loss": 1.9081, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.6601691246032715, | |
| "learning_rate": 1.9959642254967203e-05, | |
| "loss": 0.9907, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0757032632827759, | |
| "learning_rate": 1.9959200282599123e-05, | |
| "loss": 1.8676, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.3735053539276123, | |
| "learning_rate": 1.995875590823935e-05, | |
| "loss": 1.5341, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8186277151107788, | |
| "learning_rate": 1.9958309131995058e-05, | |
| "loss": 1.4037, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.7334154844284058, | |
| "learning_rate": 1.9957859953974e-05, | |
| "loss": 0.3896, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.7954867482185364, | |
| "learning_rate": 1.995740837428451e-05, | |
| "loss": 2.3116, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.6459976434707642, | |
| "learning_rate": 1.9956954393035505e-05, | |
| "loss": 1.2643, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8668067455291748, | |
| "learning_rate": 1.9956498010336478e-05, | |
| "loss": 1.106, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9993132948875427, | |
| "learning_rate": 1.9956039226297496e-05, | |
| "loss": 0.5023, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.7291522026062012, | |
| "learning_rate": 1.995557804102921e-05, | |
| "loss": 0.7306, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0429105758666992, | |
| "learning_rate": 1.9955114454642857e-05, | |
| "loss": 1.491, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0801547765731812, | |
| "learning_rate": 1.9954648467250243e-05, | |
| "loss": 1.3588, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8252794742584229, | |
| "learning_rate": 1.9954180078963753e-05, | |
| "loss": 1.0519, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.4998942613601685, | |
| "learning_rate": 1.9953709289896355e-05, | |
| "loss": 2.0603, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8373726606369019, | |
| "learning_rate": 1.9953236100161598e-05, | |
| "loss": 1.2329, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0468814373016357, | |
| "learning_rate": 1.9952760509873607e-05, | |
| "loss": 1.6357, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.2933052778244019, | |
| "learning_rate": 1.995228251914709e-05, | |
| "loss": 0.9753, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8154112100601196, | |
| "learning_rate": 1.995180212809732e-05, | |
| "loss": 1.7323, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.7924510836601257, | |
| "learning_rate": 1.995131933684017e-05, | |
| "loss": 1.3031, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.103473424911499, | |
| "learning_rate": 1.9950834145492073e-05, | |
| "loss": 1.7168, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.7922141551971436, | |
| "learning_rate": 1.9950346554170056e-05, | |
| "loss": 1.3946, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.7477668523788452, | |
| "learning_rate": 1.994985656299171e-05, | |
| "loss": 1.2364, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.860984206199646, | |
| "learning_rate": 1.994936417207522e-05, | |
| "loss": 1.6675, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8422707915306091, | |
| "learning_rate": 1.9948869381539343e-05, | |
| "loss": 1.7477, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.21062171459198, | |
| "learning_rate": 1.9948372191503406e-05, | |
| "loss": 1.0363, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8037005066871643, | |
| "learning_rate": 1.994787260208733e-05, | |
| "loss": 1.9047, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0549213886260986, | |
| "learning_rate": 1.99473706134116e-05, | |
| "loss": 1.8576, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.9254148602485657, | |
| "learning_rate": 1.99468662255973e-05, | |
| "loss": 1.3331, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0943273305892944, | |
| "learning_rate": 1.9946359438766066e-05, | |
| "loss": 2.0064, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.0273025035858154, | |
| "learning_rate": 1.9945850253040134e-05, | |
| "loss": 1.2028, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8689888715744019, | |
| "learning_rate": 1.994533866854231e-05, | |
| "loss": 1.7144, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.8768132925033569, | |
| "learning_rate": 1.994482468539598e-05, | |
| "loss": 2.116, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.795387327671051, | |
| "learning_rate": 1.9944308303725107e-05, | |
| "loss": 1.3354, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.7924376130104065, | |
| "learning_rate": 1.9943789523654237e-05, | |
| "loss": 1.027, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8999226689338684, | |
| "learning_rate": 1.9943268345308483e-05, | |
| "loss": 0.5476, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9620956778526306, | |
| "learning_rate": 1.9942744768813553e-05, | |
| "loss": 1.701, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.7766193151473999, | |
| "learning_rate": 1.994221879429572e-05, | |
| "loss": 1.2552, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1024415493011475, | |
| "learning_rate": 1.994169042188184e-05, | |
| "loss": 1.8727, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9882749319076538, | |
| "learning_rate": 1.994115965169935e-05, | |
| "loss": 1.8009, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.0686743259429932, | |
| "learning_rate": 1.9940626483876263e-05, | |
| "loss": 1.7133, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9760270118713379, | |
| "learning_rate": 1.9940090918541168e-05, | |
| "loss": 1.7522, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9487199187278748, | |
| "learning_rate": 1.9939552955823237e-05, | |
| "loss": 1.5481, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8242840766906738, | |
| "learning_rate": 1.993901259585221e-05, | |
| "loss": 1.4483, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 5.89918327331543, | |
| "learning_rate": 1.993846983875843e-05, | |
| "loss": 1.7293, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1786655187606812, | |
| "learning_rate": 1.993792468467278e-05, | |
| "loss": 1.8322, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9560681581497192, | |
| "learning_rate": 1.9937377133726753e-05, | |
| "loss": 1.6228, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.0292149782180786, | |
| "learning_rate": 1.993682718605241e-05, | |
| "loss": 1.418, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1595954895019531, | |
| "learning_rate": 1.9936274841782386e-05, | |
| "loss": 1.768, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9066284894943237, | |
| "learning_rate": 1.9935720101049897e-05, | |
| "loss": 1.5641, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1178666353225708, | |
| "learning_rate": 1.993516296398874e-05, | |
| "loss": 1.8692, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1353600025177002, | |
| "learning_rate": 1.9934603430733285e-05, | |
| "loss": 2.6621, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8215340375900269, | |
| "learning_rate": 1.9934041501418478e-05, | |
| "loss": 1.6577, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9386087656021118, | |
| "learning_rate": 1.9933477176179856e-05, | |
| "loss": 1.6772, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.2389590740203857, | |
| "learning_rate": 1.9932910455153518e-05, | |
| "loss": 1.9047, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8284682035446167, | |
| "learning_rate": 1.993234133847615e-05, | |
| "loss": 2.6177, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.399548053741455, | |
| "learning_rate": 1.993176982628501e-05, | |
| "loss": 1.9387, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.845256507396698, | |
| "learning_rate": 1.993119591871794e-05, | |
| "loss": 1.8301, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8806995153427124, | |
| "learning_rate": 1.9930619615913364e-05, | |
| "loss": 2.2612, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.7267456650733948, | |
| "learning_rate": 1.9930040918010263e-05, | |
| "loss": 1.38, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8338924050331116, | |
| "learning_rate": 1.9929459825148215e-05, | |
| "loss": 2.1012, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8365648984909058, | |
| "learning_rate": 1.9928876337467375e-05, | |
| "loss": 1.541, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.187801718711853, | |
| "learning_rate": 1.9928290455108464e-05, | |
| "loss": 1.541, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 2.6386361122131348, | |
| "learning_rate": 1.9927702178212786e-05, | |
| "loss": 1.1252, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8757442831993103, | |
| "learning_rate": 1.9927111506922227e-05, | |
| "loss": 1.4014, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.0362242460250854, | |
| "learning_rate": 1.992651844137925e-05, | |
| "loss": 2.171, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9764441847801208, | |
| "learning_rate": 1.9925922981726885e-05, | |
| "loss": 1.3608, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1725298166275024, | |
| "learning_rate": 1.992532512810875e-05, | |
| "loss": 1.8949, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.0441815853118896, | |
| "learning_rate": 1.992472488066904e-05, | |
| "loss": 1.5673, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9130929112434387, | |
| "learning_rate": 1.992412223955252e-05, | |
| "loss": 1.5355, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9515172243118286, | |
| "learning_rate": 1.9923517204904544e-05, | |
| "loss": 1.855, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 2.7631518840789795, | |
| "learning_rate": 1.9922909776871026e-05, | |
| "loss": 1.6034, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1321780681610107, | |
| "learning_rate": 1.9922299955598476e-05, | |
| "loss": 1.4475, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9716091752052307, | |
| "learning_rate": 1.992168774123397e-05, | |
| "loss": 1.0144, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9520627856254578, | |
| "learning_rate": 1.9921073133925163e-05, | |
| "loss": 1.3794, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8643708229064941, | |
| "learning_rate": 1.992045613382029e-05, | |
| "loss": 1.7873, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8701338768005371, | |
| "learning_rate": 1.9919836741068157e-05, | |
| "loss": 1.4195, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.938417911529541, | |
| "learning_rate": 1.9919214955818153e-05, | |
| "loss": 1.917, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.946891188621521, | |
| "learning_rate": 1.9918590778220243e-05, | |
| "loss": 1.6889, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9969150424003601, | |
| "learning_rate": 1.991796420842497e-05, | |
| "loss": 1.6985, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.7932005524635315, | |
| "learning_rate": 1.9917335246583443e-05, | |
| "loss": 1.8445, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.6052041053771973, | |
| "learning_rate": 1.991670389284737e-05, | |
| "loss": 1.6405, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1532299518585205, | |
| "learning_rate": 1.9916070147369017e-05, | |
| "loss": 1.5678, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9578217267990112, | |
| "learning_rate": 1.991543401030123e-05, | |
| "loss": 0.9819, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9796022176742554, | |
| "learning_rate": 1.991479548179744e-05, | |
| "loss": 2.0374, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.816317617893219, | |
| "learning_rate": 1.9914154562011643e-05, | |
| "loss": 1.2706, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1207562685012817, | |
| "learning_rate": 1.9913511251098426e-05, | |
| "loss": 1.4879, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.951187789440155, | |
| "learning_rate": 1.9912865549212935e-05, | |
| "loss": 1.1507, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9572225213050842, | |
| "learning_rate": 1.9912217456510913e-05, | |
| "loss": 1.5829, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8443856239318848, | |
| "learning_rate": 1.9911566973148663e-05, | |
| "loss": 0.9778, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.02583909034729, | |
| "learning_rate": 1.991091409928307e-05, | |
| "loss": 2.2217, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.6288923621177673, | |
| "learning_rate": 1.9910258835071595e-05, | |
| "loss": 1.0959, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.690157949924469, | |
| "learning_rate": 1.9909601180672283e-05, | |
| "loss": 0.5142, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.7905344367027283, | |
| "learning_rate": 1.9908941136243746e-05, | |
| "loss": 0.5352, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.8578112125396729, | |
| "learning_rate": 1.9908278701945175e-05, | |
| "loss": 1.7706, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.6474728584289551, | |
| "learning_rate": 1.990761387793634e-05, | |
| "loss": 0.6515, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.6532492637634277, | |
| "learning_rate": 1.9906946664377577e-05, | |
| "loss": 0.6143, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.868615448474884, | |
| "learning_rate": 1.9906277061429818e-05, | |
| "loss": 1.8124, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.1038082838058472, | |
| "learning_rate": 1.9905605069254554e-05, | |
| "loss": 1.9446, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9355915188789368, | |
| "learning_rate": 1.990493068801386e-05, | |
| "loss": 1.9609, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.1322965621948242, | |
| "learning_rate": 1.9904253917870387e-05, | |
| "loss": 1.4975, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.7613868713378906, | |
| "learning_rate": 1.9903574758987355e-05, | |
| "loss": 1.5561, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9120030403137207, | |
| "learning_rate": 1.9902893211528572e-05, | |
| "loss": 1.3147, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.416876196861267, | |
| "learning_rate": 1.990220927565841e-05, | |
| "loss": 0.9732, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.3078633546829224, | |
| "learning_rate": 1.9901522951541827e-05, | |
| "loss": 1.7904, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.0230833292007446, | |
| "learning_rate": 1.9900834239344352e-05, | |
| "loss": 1.7119, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.742758572101593, | |
| "learning_rate": 1.9900143139232095e-05, | |
| "loss": 1.7543, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9155807495117188, | |
| "learning_rate": 1.9899449651371727e-05, | |
| "loss": 1.5954, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.844660222530365, | |
| "learning_rate": 1.9898753775930513e-05, | |
| "loss": 1.6544, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9266340136528015, | |
| "learning_rate": 1.9898055513076288e-05, | |
| "loss": 1.7524, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.1642438173294067, | |
| "learning_rate": 1.9897354862977453e-05, | |
| "loss": 1.5596, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.894527018070221, | |
| "learning_rate": 1.9896651825803003e-05, | |
| "loss": 1.9041, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.207465648651123, | |
| "learning_rate": 1.9895946401722495e-05, | |
| "loss": 1.7203, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8541679978370667, | |
| "learning_rate": 1.989523859090606e-05, | |
| "loss": 1.3416, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9433109164237976, | |
| "learning_rate": 1.989452839352442e-05, | |
| "loss": 1.7141, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.0357227325439453, | |
| "learning_rate": 1.9893815809748854e-05, | |
| "loss": 1.2034, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.1258755922317505, | |
| "learning_rate": 1.9893100839751233e-05, | |
| "loss": 1.5666, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9345949292182922, | |
| "learning_rate": 1.989238348370399e-05, | |
| "loss": 1.4935, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.737742006778717, | |
| "learning_rate": 1.9891663741780135e-05, | |
| "loss": 1.4562, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9488717317581177, | |
| "learning_rate": 1.9890941614153268e-05, | |
| "loss": 1.8453, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.7368555068969727, | |
| "learning_rate": 1.989021710099755e-05, | |
| "loss": 1.4831, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8286874294281006, | |
| "learning_rate": 1.988949020248772e-05, | |
| "loss": 2.2786, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.999751627445221, | |
| "learning_rate": 1.9888760918799093e-05, | |
| "loss": 1.7505, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8931187391281128, | |
| "learning_rate": 1.9888029250107565e-05, | |
| "loss": 1.6505, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9812218546867371, | |
| "learning_rate": 1.98872951965896e-05, | |
| "loss": 1.5357, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9732285737991333, | |
| "learning_rate": 1.988655875842223e-05, | |
| "loss": 1.5624, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.3475276231765747, | |
| "learning_rate": 1.9885819935783085e-05, | |
| "loss": 1.6918, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.884621262550354, | |
| "learning_rate": 1.988507872885035e-05, | |
| "loss": 1.6485, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8911191821098328, | |
| "learning_rate": 1.9884335137802792e-05, | |
| "loss": 1.1337, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.1611531972885132, | |
| "learning_rate": 1.9883589162819757e-05, | |
| "loss": 1.7675, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.5208086967468262, | |
| "learning_rate": 1.9882840804081155e-05, | |
| "loss": 1.7403, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.881212592124939, | |
| "learning_rate": 1.9882090061767482e-05, | |
| "loss": 1.2506, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.6470460891723633, | |
| "learning_rate": 1.9881336936059797e-05, | |
| "loss": 1.0243, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.797415554523468, | |
| "learning_rate": 1.9880581427139754e-05, | |
| "loss": 0.6804, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.0682158470153809, | |
| "learning_rate": 1.987982353518956e-05, | |
| "loss": 1.6586, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.7121508121490479, | |
| "learning_rate": 1.9879063260392012e-05, | |
| "loss": 1.4205, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.7593161463737488, | |
| "learning_rate": 1.987830060293047e-05, | |
| "loss": 0.566, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9202454686164856, | |
| "learning_rate": 1.987753556298887e-05, | |
| "loss": 1.4541, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8837355971336365, | |
| "learning_rate": 1.9876768140751738e-05, | |
| "loss": 1.3312, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.2164474725723267, | |
| "learning_rate": 1.9875998336404155e-05, | |
| "loss": 1.8487, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8529666066169739, | |
| "learning_rate": 1.987522615013179e-05, | |
| "loss": 1.4369, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.0183709859848022, | |
| "learning_rate": 1.9874451582120883e-05, | |
| "loss": 1.9902, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8488042950630188, | |
| "learning_rate": 1.9873674632558237e-05, | |
| "loss": 1.5772, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9346114993095398, | |
| "learning_rate": 1.9872895301631245e-05, | |
| "loss": 1.9671, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9772021174430847, | |
| "learning_rate": 1.9872113589527872e-05, | |
| "loss": 1.6645, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.6140549182891846, | |
| "learning_rate": 1.9871329496436646e-05, | |
| "loss": 0.716, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8879808187484741, | |
| "learning_rate": 1.9870543022546685e-05, | |
| "loss": 1.7084, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8681232929229736, | |
| "learning_rate": 1.986975416804767e-05, | |
| "loss": 1.2311, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8040364980697632, | |
| "learning_rate": 1.9868962933129857e-05, | |
| "loss": 1.1052, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8864116668701172, | |
| "learning_rate": 1.9868169317984082e-05, | |
| "loss": 1.3987, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.2038077116012573, | |
| "learning_rate": 1.986737332280175e-05, | |
| "loss": 1.9375, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.7685527801513672, | |
| "learning_rate": 1.9866574947774844e-05, | |
| "loss": 1.8077, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9364449977874756, | |
| "learning_rate": 1.9865774193095913e-05, | |
| "loss": 1.7186, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.1445951461791992, | |
| "learning_rate": 1.9864971058958092e-05, | |
| "loss": 1.788, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9366347193717957, | |
| "learning_rate": 1.986416554555508e-05, | |
| "loss": 1.984, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.359132170677185, | |
| "learning_rate": 1.986335765308116e-05, | |
| "loss": 0.2784, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.961164116859436, | |
| "learning_rate": 1.9862547381731166e-05, | |
| "loss": 2.5258, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.5328058004379272, | |
| "learning_rate": 1.9861734731700534e-05, | |
| "loss": 1.5605, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.8162689208984375, | |
| "learning_rate": 1.9860919703185265e-05, | |
| "loss": 1.4508, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.7568703293800354, | |
| "learning_rate": 1.986010229638192e-05, | |
| "loss": 1.5941, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.7346351146697998, | |
| "learning_rate": 1.985928251148765e-05, | |
| "loss": 1.6749, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.1934709548950195, | |
| "learning_rate": 1.985846034870017e-05, | |
| "loss": 2.0963, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9638925194740295, | |
| "learning_rate": 1.9857635808217778e-05, | |
| "loss": 1.9696, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0379003286361694, | |
| "learning_rate": 1.9856808890239333e-05, | |
| "loss": 0.8772, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0562266111373901, | |
| "learning_rate": 1.985597959496427e-05, | |
| "loss": 1.7018, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0546568632125854, | |
| "learning_rate": 1.9855147922592614e-05, | |
| "loss": 0.9769, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.7254536151885986, | |
| "learning_rate": 1.985431387332494e-05, | |
| "loss": 0.9393, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8547775149345398, | |
| "learning_rate": 1.9853477447362412e-05, | |
| "loss": 1.3414, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.802185595035553, | |
| "learning_rate": 1.985263864490676e-05, | |
| "loss": 0.9645, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8310496807098389, | |
| "learning_rate": 1.9851797466160283e-05, | |
| "loss": 1.3528, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.5428261756896973, | |
| "learning_rate": 1.985095391132587e-05, | |
| "loss": 1.7086, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8697307109832764, | |
| "learning_rate": 1.9850107980606964e-05, | |
| "loss": 2.0801, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0970669984817505, | |
| "learning_rate": 1.98492596742076e-05, | |
| "loss": 1.7845, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.1335629224777222, | |
| "learning_rate": 1.984840899233236e-05, | |
| "loss": 0.4997, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9874480366706848, | |
| "learning_rate": 1.9847555935186424e-05, | |
| "loss": 1.9856, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8648535013198853, | |
| "learning_rate": 1.9846700502975527e-05, | |
| "loss": 1.7317, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.2264970541000366, | |
| "learning_rate": 1.9845842695905997e-05, | |
| "loss": 1.6865, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.876065194606781, | |
| "learning_rate": 1.9844982514184718e-05, | |
| "loss": 1.5677, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.7052832841873169, | |
| "learning_rate": 1.9844119958019146e-05, | |
| "loss": 1.6304, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.8926292657852173, | |
| "learning_rate": 1.984325502761732e-05, | |
| "loss": 1.4055, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9539690613746643, | |
| "learning_rate": 1.9842387723187846e-05, | |
| "loss": 1.4473, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.9869019985198975, | |
| "learning_rate": 1.98415180449399e-05, | |
| "loss": 1.0927, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.443432331085205, | |
| "learning_rate": 1.984064599308324e-05, | |
| "loss": 1.1167, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8694020509719849, | |
| "learning_rate": 1.9839771567828184e-05, | |
| "loss": 1.1221, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9755352139472961, | |
| "learning_rate": 1.9838894769385635e-05, | |
| "loss": 0.8227, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9441053867340088, | |
| "learning_rate": 1.9838015597967056e-05, | |
| "loss": 0.8816, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.7311797738075256, | |
| "learning_rate": 1.983713405378449e-05, | |
| "loss": 0.3256, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8313472867012024, | |
| "learning_rate": 1.9836250137050553e-05, | |
| "loss": 1.8942, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.5323596000671387, | |
| "learning_rate": 1.9835363847978428e-05, | |
| "loss": 1.5672, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0517991781234741, | |
| "learning_rate": 1.9834475186781877e-05, | |
| "loss": 0.76, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.4064319133758545, | |
| "learning_rate": 1.983358415367522e-05, | |
| "loss": 1.5922, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9361559748649597, | |
| "learning_rate": 1.9832690748873375e-05, | |
| "loss": 1.2397, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.6758370399475098, | |
| "learning_rate": 1.9831794972591805e-05, | |
| "loss": 0.5728, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9486380219459534, | |
| "learning_rate": 1.9830896825046563e-05, | |
| "loss": 1.5274, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9687122702598572, | |
| "learning_rate": 1.9829996306454258e-05, | |
| "loss": 0.9347, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.1283320188522339, | |
| "learning_rate": 1.9829093417032085e-05, | |
| "loss": 1.8164, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.7893568277359009, | |
| "learning_rate": 1.982818815699781e-05, | |
| "loss": 1.4499, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8257303237915039, | |
| "learning_rate": 1.9827280526569756e-05, | |
| "loss": 1.4763, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.2423681020736694, | |
| "learning_rate": 1.982637052596684e-05, | |
| "loss": 1.757, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0532588958740234, | |
| "learning_rate": 1.9825458155408533e-05, | |
| "loss": 1.7136, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0051872730255127, | |
| "learning_rate": 1.982454341511488e-05, | |
| "loss": 1.6012, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8963203430175781, | |
| "learning_rate": 1.9823626305306506e-05, | |
| "loss": 1.072, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.3821759223937988, | |
| "learning_rate": 1.9822706826204606e-05, | |
| "loss": 1.7187, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0313401222229004, | |
| "learning_rate": 1.9821784978030933e-05, | |
| "loss": 1.6027, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9623168706893921, | |
| "learning_rate": 1.982086076100783e-05, | |
| "loss": 1.4132, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9375492334365845, | |
| "learning_rate": 1.9819934175358195e-05, | |
| "loss": 1.7144, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.7186566591262817, | |
| "learning_rate": 1.9819005221305513e-05, | |
| "loss": 1.3332, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.7701579332351685, | |
| "learning_rate": 1.9818073899073827e-05, | |
| "loss": 1.1296, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8675704598426819, | |
| "learning_rate": 1.9817140208887754e-05, | |
| "loss": 1.7195, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9611474871635437, | |
| "learning_rate": 1.9816204150972492e-05, | |
| "loss": 1.4536, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.056624412536621, | |
| "learning_rate": 1.9815265725553797e-05, | |
| "loss": 1.9253, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.585964560508728, | |
| "learning_rate": 1.9814324932858005e-05, | |
| "loss": 2.0826, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9207966327667236, | |
| "learning_rate": 1.9813381773112016e-05, | |
| "loss": 1.4885, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9804804921150208, | |
| "learning_rate": 1.9812436246543304e-05, | |
| "loss": 1.3071, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.7058753371238708, | |
| "learning_rate": 1.981148835337991e-05, | |
| "loss": 0.8504, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.990956723690033, | |
| "learning_rate": 1.9810538093850465e-05, | |
| "loss": 2.1998, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9452444911003113, | |
| "learning_rate": 1.980958546818414e-05, | |
| "loss": 1.7364, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8717514276504517, | |
| "learning_rate": 1.98086304766107e-05, | |
| "loss": 0.8079, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.1225824356079102, | |
| "learning_rate": 1.980767311936047e-05, | |
| "loss": 1.7555, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0479341745376587, | |
| "learning_rate": 1.980671339666435e-05, | |
| "loss": 1.6204, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.9834145307540894, | |
| "learning_rate": 1.9805751308753808e-05, | |
| "loss": 1.5251, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.074107050895691, | |
| "learning_rate": 1.9804786855860886e-05, | |
| "loss": 1.8535, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.1673707962036133, | |
| "learning_rate": 1.980382003821819e-05, | |
| "loss": 1.8278, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.8728948831558228, | |
| "learning_rate": 1.9802850856058903e-05, | |
| "loss": 0.8509, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.0313911437988281, | |
| "learning_rate": 1.9801879309616777e-05, | |
| "loss": 1.9013, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0420602560043335, | |
| "learning_rate": 1.9800905399126125e-05, | |
| "loss": 1.4502, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.2369115352630615, | |
| "learning_rate": 1.9799929124821844e-05, | |
| "loss": 1.689, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.1010773181915283, | |
| "learning_rate": 1.97989504869394e-05, | |
| "loss": 0.6274, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.3504084348678589, | |
| "learning_rate": 1.979796948571481e-05, | |
| "loss": 1.2692, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8189222812652588, | |
| "learning_rate": 1.9796986121384685e-05, | |
| "loss": 1.5439, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9344761967658997, | |
| "learning_rate": 1.9796000394186198e-05, | |
| "loss": 1.5129, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0255612134933472, | |
| "learning_rate": 1.979501230435708e-05, | |
| "loss": 1.9559, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.449123501777649, | |
| "learning_rate": 1.979402185213565e-05, | |
| "loss": 1.7581, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0824624300003052, | |
| "learning_rate": 1.9793029037760784e-05, | |
| "loss": 2.1333, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0862467288970947, | |
| "learning_rate": 1.9792033861471935e-05, | |
| "loss": 1.6442, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.3037303686141968, | |
| "learning_rate": 1.9791036323509126e-05, | |
| "loss": 1.4648, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.037331461906433, | |
| "learning_rate": 1.9790036424112938e-05, | |
| "loss": 1.017, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8827648162841797, | |
| "learning_rate": 1.9789034163524535e-05, | |
| "loss": 1.659, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8996952176094055, | |
| "learning_rate": 1.9788029541985643e-05, | |
| "loss": 1.3892, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0246918201446533, | |
| "learning_rate": 1.978702255973856e-05, | |
| "loss": 1.51, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.9744868278503418, | |
| "learning_rate": 1.9786013217026157e-05, | |
| "loss": 1.3445, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9991737604141235, | |
| "learning_rate": 1.9785001514091865e-05, | |
| "loss": 1.6653, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.4088870286941528, | |
| "learning_rate": 1.9783987451179696e-05, | |
| "loss": 1.57, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.7513281106948853, | |
| "learning_rate": 1.9782971028534223e-05, | |
| "loss": 0.7615, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.09906005859375, | |
| "learning_rate": 1.9781952246400587e-05, | |
| "loss": 0.9732, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.4216969013214111, | |
| "learning_rate": 1.97809311050245e-05, | |
| "loss": 1.1183, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9488307237625122, | |
| "learning_rate": 1.977990760465225e-05, | |
| "loss": 1.002, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9960178136825562, | |
| "learning_rate": 1.977888174553068e-05, | |
| "loss": 1.7178, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0002471208572388, | |
| "learning_rate": 1.9777853527907224e-05, | |
| "loss": 1.7894, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.886707603931427, | |
| "learning_rate": 1.9776822952029854e-05, | |
| "loss": 1.039, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.7411413192749023, | |
| "learning_rate": 1.977579001814714e-05, | |
| "loss": 1.5746, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0032079219818115, | |
| "learning_rate": 1.9774754726508202e-05, | |
| "loss": 1.3436, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0437644720077515, | |
| "learning_rate": 1.9773717077362734e-05, | |
| "loss": 1.4815, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0581908226013184, | |
| "learning_rate": 1.9772677070961007e-05, | |
| "loss": 1.8118, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8130147457122803, | |
| "learning_rate": 1.9771634707553846e-05, | |
| "loss": 1.1538, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.009604573249817, | |
| "learning_rate": 1.9770589987392653e-05, | |
| "loss": 1.6894, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0172890424728394, | |
| "learning_rate": 1.9769542910729397e-05, | |
| "loss": 1.6646, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9655178189277649, | |
| "learning_rate": 1.9768493477816618e-05, | |
| "loss": 1.5982, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8684929013252258, | |
| "learning_rate": 1.976744168890742e-05, | |
| "loss": 2.0278, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0428229570388794, | |
| "learning_rate": 1.9766387544255473e-05, | |
| "loss": 0.4468, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.1995832920074463, | |
| "learning_rate": 1.9765331044115024e-05, | |
| "loss": 1.671, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.851990282535553, | |
| "learning_rate": 1.9764272188740885e-05, | |
| "loss": 0.9449, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.2639641761779785, | |
| "learning_rate": 1.9763210978388426e-05, | |
| "loss": 0.5276, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.1194183826446533, | |
| "learning_rate": 1.9762147413313596e-05, | |
| "loss": 1.7057, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 4.371021747589111, | |
| "learning_rate": 1.9761081493772914e-05, | |
| "loss": 0.6986, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8994089365005493, | |
| "learning_rate": 1.976001322002346e-05, | |
| "loss": 2.5059, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8265968561172485, | |
| "learning_rate": 1.975894259232288e-05, | |
| "loss": 1.5496, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0377991199493408, | |
| "learning_rate": 1.9757869610929396e-05, | |
| "loss": 1.8287, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8921816945075989, | |
| "learning_rate": 1.9756794276101784e-05, | |
| "loss": 1.664, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.402585506439209, | |
| "learning_rate": 1.975571658809941e-05, | |
| "loss": 0.5236, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9020755290985107, | |
| "learning_rate": 1.9754636547182183e-05, | |
| "loss": 1.7119, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.136975884437561, | |
| "learning_rate": 1.9753554153610598e-05, | |
| "loss": 1.4358, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8028199672698975, | |
| "learning_rate": 1.975246940764571e-05, | |
| "loss": 1.1221, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9304431676864624, | |
| "learning_rate": 1.9751382309549135e-05, | |
| "loss": 1.1922, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.1193586587905884, | |
| "learning_rate": 1.9750292859583066e-05, | |
| "loss": 0.5726, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8460230827331543, | |
| "learning_rate": 1.974920105801026e-05, | |
| "loss": 2.4396, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8110451698303223, | |
| "learning_rate": 1.9748106905094045e-05, | |
| "loss": 1.3223, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.7961413860321045, | |
| "learning_rate": 1.9747010401098307e-05, | |
| "loss": 1.2879, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.1107492446899414, | |
| "learning_rate": 1.974591154628751e-05, | |
| "loss": 1.6185, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.14578115940094, | |
| "learning_rate": 1.9744810340926674e-05, | |
| "loss": 1.4298, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.3011332750320435, | |
| "learning_rate": 1.974370678528139e-05, | |
| "loss": 0.9556, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8401873111724854, | |
| "learning_rate": 1.9742600879617823e-05, | |
| "loss": 2.3746, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8258370757102966, | |
| "learning_rate": 1.9741492624202698e-05, | |
| "loss": 1.4742, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.7155411243438721, | |
| "learning_rate": 1.97403820193033e-05, | |
| "loss": 0.9388, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8539145588874817, | |
| "learning_rate": 1.9739269065187497e-05, | |
| "loss": 1.4629, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8882036805152893, | |
| "learning_rate": 1.9738153762123713e-05, | |
| "loss": 1.889, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.810894250869751, | |
| "learning_rate": 1.9737036110380935e-05, | |
| "loss": 0.5582, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.8205024600028992, | |
| "learning_rate": 1.973591611022873e-05, | |
| "loss": 1.3398, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0441452264785767, | |
| "learning_rate": 1.9734793761937216e-05, | |
| "loss": 1.4455, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.1005403995513916, | |
| "learning_rate": 1.973366906577709e-05, | |
| "loss": 1.5484, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.6423813700675964, | |
| "learning_rate": 1.9732542022019603e-05, | |
| "loss": 0.5292, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.9499244689941406, | |
| "learning_rate": 1.973141263093659e-05, | |
| "loss": 1.3558, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9298450350761414, | |
| "learning_rate": 1.973028089280043e-05, | |
| "loss": 1.3631, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.7698273062705994, | |
| "learning_rate": 1.9729146807884084e-05, | |
| "loss": 0.5124, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8734893798828125, | |
| "learning_rate": 1.972801037646107e-05, | |
| "loss": 1.637, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.1014962196350098, | |
| "learning_rate": 1.972687159880549e-05, | |
| "loss": 1.3022, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.2191619873046875, | |
| "learning_rate": 1.972573047519198e-05, | |
| "loss": 0.7652, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.4204211235046387, | |
| "learning_rate": 1.972458700589577e-05, | |
| "loss": 1.8729, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8936397433280945, | |
| "learning_rate": 1.9723441191192647e-05, | |
| "loss": 2.155, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.3093836307525635, | |
| "learning_rate": 1.9722293031358955e-05, | |
| "loss": 1.8055, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.918250560760498, | |
| "learning_rate": 1.9721142526671614e-05, | |
| "loss": 1.1799, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.806936502456665, | |
| "learning_rate": 1.971998967740811e-05, | |
| "loss": 2.2243, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.198886752128601, | |
| "learning_rate": 1.9718834483846485e-05, | |
| "loss": 1.5366, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.6140164136886597, | |
| "learning_rate": 1.9717676946265358e-05, | |
| "loss": 1.5059, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.061775803565979, | |
| "learning_rate": 1.97165170649439e-05, | |
| "loss": 1.5537, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.120010495185852, | |
| "learning_rate": 1.9715354840161862e-05, | |
| "loss": 1.5032, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.4784014225006104, | |
| "learning_rate": 1.9714190272199555e-05, | |
| "loss": 2.0133, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.0621649026870728, | |
| "learning_rate": 1.971302336133784e-05, | |
| "loss": 1.5967, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.0540523529052734, | |
| "learning_rate": 1.971185410785817e-05, | |
| "loss": 1.447, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8714383840560913, | |
| "learning_rate": 1.9710682512042548e-05, | |
| "loss": 1.6458, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8371648192405701, | |
| "learning_rate": 1.9709508574173535e-05, | |
| "loss": 1.6041, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.397619605064392, | |
| "learning_rate": 1.9708332294534274e-05, | |
| "loss": 0.787, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9853465557098389, | |
| "learning_rate": 1.9707153673408454e-05, | |
| "loss": 1.5562, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.108499526977539, | |
| "learning_rate": 1.970597271108035e-05, | |
| "loss": 0.9748, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.7067657709121704, | |
| "learning_rate": 1.9704789407834778e-05, | |
| "loss": 1.4679, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9193729162216187, | |
| "learning_rate": 1.9703603763957142e-05, | |
| "loss": 1.8292, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.0290855169296265, | |
| "learning_rate": 1.9702415779733392e-05, | |
| "loss": 1.623, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.1568219661712646, | |
| "learning_rate": 1.970122545545006e-05, | |
| "loss": 1.6309, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.8143751621246338, | |
| "learning_rate": 1.9700032791394215e-05, | |
| "loss": 1.5616, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.996700644493103, | |
| "learning_rate": 1.9698837787853523e-05, | |
| "loss": 1.941, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.7473616003990173, | |
| "learning_rate": 1.9697640445116192e-05, | |
| "loss": 1.6, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.0816364288330078, | |
| "learning_rate": 1.9696440763471e-05, | |
| "loss": 1.764, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9372792840003967, | |
| "learning_rate": 1.9695238743207292e-05, | |
| "loss": 1.4362, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.5626298189163208, | |
| "learning_rate": 1.969403438461498e-05, | |
| "loss": 0.6701, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8877453804016113, | |
| "learning_rate": 1.9692827687984523e-05, | |
| "loss": 1.6476, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8707518577575684, | |
| "learning_rate": 1.969161865360697e-05, | |
| "loss": 1.0084, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.0139648914337158, | |
| "learning_rate": 1.9690407281773905e-05, | |
| "loss": 1.4887, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.9187926054000854, | |
| "learning_rate": 1.9689193572777505e-05, | |
| "loss": 1.5581, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8604023456573486, | |
| "learning_rate": 1.9687977526910488e-05, | |
| "loss": 1.6908, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.2782931327819824, | |
| "learning_rate": 1.9686759144466142e-05, | |
| "loss": 1.1137, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9516409635543823, | |
| "learning_rate": 1.9685538425738326e-05, | |
| "loss": 1.2642, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.1259483098983765, | |
| "learning_rate": 1.968431537102146e-05, | |
| "loss": 1.555, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.298754334449768, | |
| "learning_rate": 1.9683089980610512e-05, | |
| "loss": 1.5994, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.3935412168502808, | |
| "learning_rate": 1.9681862254801034e-05, | |
| "loss": 1.0852, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8852770924568176, | |
| "learning_rate": 1.968063219388913e-05, | |
| "loss": 1.9436, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.731374204158783, | |
| "learning_rate": 1.9679399798171476e-05, | |
| "loss": 0.7434, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.073830008506775, | |
| "learning_rate": 1.9678165067945295e-05, | |
| "loss": 2.2306, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.2147220373153687, | |
| "learning_rate": 1.967692800350839e-05, | |
| "loss": 2.0442, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8961025476455688, | |
| "learning_rate": 1.967568860515912e-05, | |
| "loss": 1.6517, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.1098088026046753, | |
| "learning_rate": 1.9674446873196408e-05, | |
| "loss": 1.5784, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8114632964134216, | |
| "learning_rate": 1.9673202807919733e-05, | |
| "loss": 0.5461, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8463197946548462, | |
| "learning_rate": 1.967195640962915e-05, | |
| "loss": 1.2396, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.5498254299163818, | |
| "learning_rate": 1.9670707678625264e-05, | |
| "loss": 1.8367, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.2922487258911133, | |
| "learning_rate": 1.9669456615209247e-05, | |
| "loss": 2.251, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8260951042175293, | |
| "learning_rate": 1.966820321968284e-05, | |
| "loss": 2.8105, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9870577454566956, | |
| "learning_rate": 1.9666947492348334e-05, | |
| "loss": 1.6524, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9321213960647583, | |
| "learning_rate": 1.9665689433508596e-05, | |
| "loss": 0.9929, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.7950844764709473, | |
| "learning_rate": 1.9664429043467044e-05, | |
| "loss": 2.1832, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9330363273620605, | |
| "learning_rate": 1.9663166322527668e-05, | |
| "loss": 1.1176, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.0382758378982544, | |
| "learning_rate": 1.9661901270995013e-05, | |
| "loss": 1.5673, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.7401167154312134, | |
| "learning_rate": 1.9660633889174185e-05, | |
| "loss": 1.3154, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.299643874168396, | |
| "learning_rate": 1.9659364177370856e-05, | |
| "loss": 1.7258, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8082520961761475, | |
| "learning_rate": 1.9658092135891256e-05, | |
| "loss": 1.5656, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.076417088508606, | |
| "learning_rate": 1.9656817765042192e-05, | |
| "loss": 1.6585, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.0347617864608765, | |
| "learning_rate": 1.9655541065131013e-05, | |
| "loss": 1.608, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.070332646369934, | |
| "learning_rate": 1.965426203646563e-05, | |
| "loss": 1.9664, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.838746190071106, | |
| "learning_rate": 1.965298067935454e-05, | |
| "loss": 2.1611, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9825697541236877, | |
| "learning_rate": 1.965169699410677e-05, | |
| "loss": 1.7168, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.5898070335388184, | |
| "learning_rate": 1.965041098103193e-05, | |
| "loss": 1.3112, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9250902533531189, | |
| "learning_rate": 1.9649122640440187e-05, | |
| "loss": 0.9961, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.733625054359436, | |
| "learning_rate": 1.964783197264226e-05, | |
| "loss": 1.7441, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.052191972732544, | |
| "learning_rate": 1.9646538977949444e-05, | |
| "loss": 1.3966, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9679017663002014, | |
| "learning_rate": 1.9645243656673582e-05, | |
| "loss": 0.7808, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7678504586219788, | |
| "learning_rate": 1.9643946009127088e-05, | |
| "loss": 1.7577, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.4840080738067627, | |
| "learning_rate": 1.9642646035622926e-05, | |
| "loss": 1.5794, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8202274441719055, | |
| "learning_rate": 1.9641343736474635e-05, | |
| "loss": 1.9119, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.4255080223083496, | |
| "learning_rate": 1.96400391119963e-05, | |
| "loss": 1.1435, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.1355290412902832, | |
| "learning_rate": 1.9638732162502586e-05, | |
| "loss": 1.6109, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.0749764442443848, | |
| "learning_rate": 1.9637422888308696e-05, | |
| "loss": 1.1917, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9144006967544556, | |
| "learning_rate": 1.963611128973041e-05, | |
| "loss": 1.9197, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7404701113700867, | |
| "learning_rate": 1.9634797367084065e-05, | |
| "loss": 1.1256, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8343794345855713, | |
| "learning_rate": 1.9633481120686554e-05, | |
| "loss": 1.8625, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9903147220611572, | |
| "learning_rate": 1.9632162550855335e-05, | |
| "loss": 1.4753, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9988419413566589, | |
| "learning_rate": 1.9630841657908424e-05, | |
| "loss": 1.4715, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7479739189147949, | |
| "learning_rate": 1.9629518442164397e-05, | |
| "loss": 1.6179, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.0459107160568237, | |
| "learning_rate": 1.9628192903942395e-05, | |
| "loss": 1.6056, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.6827442646026611, | |
| "learning_rate": 1.9626865043562113e-05, | |
| "loss": 1.2805, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9262802004814148, | |
| "learning_rate": 1.9625534861343806e-05, | |
| "loss": 1.9221, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.0962365865707397, | |
| "learning_rate": 1.96242023576083e-05, | |
| "loss": 1.1623, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8223599195480347, | |
| "learning_rate": 1.9622867532676965e-05, | |
| "loss": 1.4158, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8202844262123108, | |
| "learning_rate": 1.962153038687174e-05, | |
| "loss": 2.3395, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.769325315952301, | |
| "learning_rate": 1.9620190920515124e-05, | |
| "loss": 1.2261, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9759867787361145, | |
| "learning_rate": 1.961884913393017e-05, | |
| "loss": 1.4156, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8986252546310425, | |
| "learning_rate": 1.96175050274405e-05, | |
| "loss": 1.2528, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8722736239433289, | |
| "learning_rate": 1.9616158601370284e-05, | |
| "loss": 1.2981, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.0854696035385132, | |
| "learning_rate": 1.961480985604426e-05, | |
| "loss": 1.6531, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9684545993804932, | |
| "learning_rate": 1.9613458791787723e-05, | |
| "loss": 1.149, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.1259515285491943, | |
| "learning_rate": 1.961210540892653e-05, | |
| "loss": 1.6374, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.6296025514602661, | |
| "learning_rate": 1.9610749707787086e-05, | |
| "loss": 0.1722, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9159917235374451, | |
| "learning_rate": 1.960939168869637e-05, | |
| "loss": 2.4642, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7028860449790955, | |
| "learning_rate": 1.9608031351981913e-05, | |
| "loss": 0.8297, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.1244877576828003, | |
| "learning_rate": 1.9606668697971804e-05, | |
| "loss": 1.4262, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.1643065214157104, | |
| "learning_rate": 1.9605303726994692e-05, | |
| "loss": 1.8621, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.2138675451278687, | |
| "learning_rate": 1.960393643937978e-05, | |
| "loss": 0.7296, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.5474352836608887, | |
| "learning_rate": 1.9602566835456847e-05, | |
| "loss": 1.2063, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.1823064088821411, | |
| "learning_rate": 1.9601194915556207e-05, | |
| "loss": 0.1894, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.6637358069419861, | |
| "learning_rate": 1.959982068000875e-05, | |
| "loss": 1.4456, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.6221618056297302, | |
| "learning_rate": 1.959844412914592e-05, | |
| "loss": 0.5781, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9478740096092224, | |
| "learning_rate": 1.959706526329971e-05, | |
| "loss": 1.2688, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.0652697086334229, | |
| "learning_rate": 1.9595684082802687e-05, | |
| "loss": 1.7915, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9758893847465515, | |
| "learning_rate": 1.9594300587987964e-05, | |
| "loss": 1.6521, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.3758968114852905, | |
| "learning_rate": 1.9592914779189218e-05, | |
| "loss": 1.7467, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.0427806377410889, | |
| "learning_rate": 1.9591526656740685e-05, | |
| "loss": 1.3578, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.4715042114257812, | |
| "learning_rate": 1.9590136220977156e-05, | |
| "loss": 1.6052, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9868489503860474, | |
| "learning_rate": 1.958874347223398e-05, | |
| "loss": 1.7042, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8361290097236633, | |
| "learning_rate": 1.958734841084706e-05, | |
| "loss": 1.1737, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.1467775106430054, | |
| "learning_rate": 1.958595103715287e-05, | |
| "loss": 1.135, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7978134751319885, | |
| "learning_rate": 1.9584551351488425e-05, | |
| "loss": 1.305, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9538130760192871, | |
| "learning_rate": 1.9583149354191312e-05, | |
| "loss": 1.8212, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.121382236480713, | |
| "learning_rate": 1.9581745045599665e-05, | |
| "loss": 0.3183, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7262074947357178, | |
| "learning_rate": 1.9580338426052182e-05, | |
| "loss": 2.0245, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7842314839363098, | |
| "learning_rate": 1.9578929495888116e-05, | |
| "loss": 0.9158, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7360401153564453, | |
| "learning_rate": 1.9577518255447274e-05, | |
| "loss": 0.8348, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8156128525733948, | |
| "learning_rate": 1.9576104705070027e-05, | |
| "loss": 1.3912, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.7814883589744568, | |
| "learning_rate": 1.9574688845097302e-05, | |
| "loss": 1.0388, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.038300633430481, | |
| "learning_rate": 1.9573270675870573e-05, | |
| "loss": 0.751, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9030879735946655, | |
| "learning_rate": 1.9571850197731884e-05, | |
| "loss": 1.3984, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8930461406707764, | |
| "learning_rate": 1.9570427411023827e-05, | |
| "loss": 1.8717, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.5750823020935059, | |
| "learning_rate": 1.956900231608956e-05, | |
| "loss": 2.1555, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8204957246780396, | |
| "learning_rate": 1.9567574913272788e-05, | |
| "loss": 1.3371, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9168577790260315, | |
| "learning_rate": 1.9566145202917773e-05, | |
| "loss": 1.132, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9453327059745789, | |
| "learning_rate": 1.9564713185369344e-05, | |
| "loss": 0.852, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.316981554031372, | |
| "learning_rate": 1.9563278860972876e-05, | |
| "loss": 1.8398, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0372123718261719, | |
| "learning_rate": 1.9561842230074302e-05, | |
| "loss": 1.653, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.1009215116500854, | |
| "learning_rate": 1.956040329302012e-05, | |
| "loss": 1.6638, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9586663246154785, | |
| "learning_rate": 1.955896205015737e-05, | |
| "loss": 1.0886, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.3998528718948364, | |
| "learning_rate": 1.9557518501833658e-05, | |
| "loss": 1.9539, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9007513523101807, | |
| "learning_rate": 1.955607264839715e-05, | |
| "loss": 0.4139, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.7559785842895508, | |
| "learning_rate": 1.955462449019655e-05, | |
| "loss": 1.8494, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0630534887313843, | |
| "learning_rate": 1.9553174027581138e-05, | |
| "loss": 1.6309, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9597139358520508, | |
| "learning_rate": 1.955172126090074e-05, | |
| "loss": 0.854, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.003679871559143, | |
| "learning_rate": 1.9550266190505735e-05, | |
| "loss": 1.1375, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2234400510787964, | |
| "learning_rate": 1.954880881674706e-05, | |
| "loss": 1.0621, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8920372128486633, | |
| "learning_rate": 1.954734913997622e-05, | |
| "loss": 0.8149, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0686150789260864, | |
| "learning_rate": 1.9545887160545252e-05, | |
| "loss": 1.0463, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2131983041763306, | |
| "learning_rate": 1.9544422878806768e-05, | |
| "loss": 1.4061, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.1307955980300903, | |
| "learning_rate": 1.9542956295113925e-05, | |
| "loss": 1.6915, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.9545600414276123, | |
| "learning_rate": 1.9541487409820442e-05, | |
| "loss": 1.9788, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9386622905731201, | |
| "learning_rate": 1.954001622328059e-05, | |
| "loss": 2.3962, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0973879098892212, | |
| "learning_rate": 1.9538542735849188e-05, | |
| "loss": 1.9024, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0093085765838623, | |
| "learning_rate": 1.953706694788162e-05, | |
| "loss": 1.6269, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.729065477848053, | |
| "learning_rate": 1.953558885973382e-05, | |
| "loss": 0.7673, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9341952204704285, | |
| "learning_rate": 1.953410847176228e-05, | |
| "loss": 1.3542, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.6524180769920349, | |
| "learning_rate": 1.9532625784324047e-05, | |
| "loss": 1.4862, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0662424564361572, | |
| "learning_rate": 1.9531140797776714e-05, | |
| "loss": 1.8857, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.7807791233062744, | |
| "learning_rate": 1.952965351247844e-05, | |
| "loss": 1.6716, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0078587532043457, | |
| "learning_rate": 1.952816392878793e-05, | |
| "loss": 1.9718, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2724988460540771, | |
| "learning_rate": 1.9526672047064448e-05, | |
| "loss": 1.348, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.6617332696914673, | |
| "learning_rate": 1.952517786766781e-05, | |
| "loss": 1.6268, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9120773077011108, | |
| "learning_rate": 1.9523681390958393e-05, | |
| "loss": 1.9604, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9997332096099854, | |
| "learning_rate": 1.952218261729711e-05, | |
| "loss": 1.6832, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9092682003974915, | |
| "learning_rate": 1.952068154704545e-05, | |
| "loss": 1.9769, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9539963006973267, | |
| "learning_rate": 1.951917818056544e-05, | |
| "loss": 0.9615, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8873115181922913, | |
| "learning_rate": 1.951767251821967e-05, | |
| "loss": 1.4593, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8866798281669617, | |
| "learning_rate": 1.951616456037128e-05, | |
| "loss": 1.491, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.098756194114685, | |
| "learning_rate": 1.9514654307383966e-05, | |
| "loss": 0.8209, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.7631105184555054, | |
| "learning_rate": 1.9513141759621973e-05, | |
| "loss": 1.7128, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 12.292449951171875, | |
| "learning_rate": 1.95116269174501e-05, | |
| "loss": 3.2222, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9233523607254028, | |
| "learning_rate": 1.9510109781233706e-05, | |
| "loss": 1.2803, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8416237235069275, | |
| "learning_rate": 1.9508590351338693e-05, | |
| "loss": 1.1823, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.337841033935547, | |
| "learning_rate": 1.950706862813153e-05, | |
| "loss": 1.2577, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9062979221343994, | |
| "learning_rate": 1.9505544611979226e-05, | |
| "loss": 2.1593, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9518592953681946, | |
| "learning_rate": 1.950401830324935e-05, | |
| "loss": 1.1195, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.020942211151123, | |
| "learning_rate": 1.9502489702310018e-05, | |
| "loss": 1.5837, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.879868745803833, | |
| "learning_rate": 1.9500958809529903e-05, | |
| "loss": 2.0357, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.6751201152801514, | |
| "learning_rate": 1.949942562527824e-05, | |
| "loss": 1.4446, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0675795078277588, | |
| "learning_rate": 1.94978901499248e-05, | |
| "loss": 1.4454, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8025611639022827, | |
| "learning_rate": 1.949635238383991e-05, | |
| "loss": 1.212, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0021713972091675, | |
| "learning_rate": 1.9494812327394458e-05, | |
| "loss": 1.3265, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0278031826019287, | |
| "learning_rate": 1.9493269980959885e-05, | |
| "loss": 1.0983, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.649834394454956, | |
| "learning_rate": 1.9491725344908167e-05, | |
| "loss": 1.5833, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8162665963172913, | |
| "learning_rate": 1.9490178419611853e-05, | |
| "loss": 1.6149, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.1506603956222534, | |
| "learning_rate": 1.9488629205444034e-05, | |
| "loss": 1.0756, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.4848748445510864, | |
| "learning_rate": 1.9487077702778353e-05, | |
| "loss": 1.7649, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9423287510871887, | |
| "learning_rate": 1.948552391198901e-05, | |
| "loss": 1.5612, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.312781572341919, | |
| "learning_rate": 1.9483967833450745e-05, | |
| "loss": 1.8895, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.3580660820007324, | |
| "learning_rate": 1.9482409467538863e-05, | |
| "loss": 1.7204, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9630597233772278, | |
| "learning_rate": 1.9480848814629217e-05, | |
| "loss": 1.9857, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.7201079726219177, | |
| "learning_rate": 1.947928587509821e-05, | |
| "loss": 1.1478, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9898256063461304, | |
| "learning_rate": 1.94777206493228e-05, | |
| "loss": 1.5353, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.0257985591888428, | |
| "learning_rate": 1.9476153137680482e-05, | |
| "loss": 1.4052, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8972013592720032, | |
| "learning_rate": 1.9474583340549324e-05, | |
| "loss": 1.8627, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9348580837249756, | |
| "learning_rate": 1.947301125830793e-05, | |
| "loss": 1.8107, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8059454560279846, | |
| "learning_rate": 1.947143689133546e-05, | |
| "loss": 1.7113, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.1312888860702515, | |
| "learning_rate": 1.9469860240011626e-05, | |
| "loss": 1.8456, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.2592461109161377, | |
| "learning_rate": 1.946828130471669e-05, | |
| "loss": 1.7084, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8490139842033386, | |
| "learning_rate": 1.9466700085831466e-05, | |
| "loss": 1.7365, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.3371502161026, | |
| "learning_rate": 1.9465116583737314e-05, | |
| "loss": 1.7185, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.1870206594467163, | |
| "learning_rate": 1.946353079881615e-05, | |
| "loss": 1.5285, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.8405884504318237, | |
| "learning_rate": 1.9461942731450435e-05, | |
| "loss": 1.6512, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.3664089441299438, | |
| "learning_rate": 1.9460352382023192e-05, | |
| "loss": 1.649, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.7940143346786499, | |
| "learning_rate": 1.945875975091798e-05, | |
| "loss": 1.7897, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.7846907377243042, | |
| "learning_rate": 1.9457164838518916e-05, | |
| "loss": 1.5796, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.7551419734954834, | |
| "learning_rate": 1.9455567645210668e-05, | |
| "loss": 1.4008, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.0351200103759766, | |
| "learning_rate": 1.945396817137845e-05, | |
| "loss": 1.5813, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.0557612180709839, | |
| "learning_rate": 1.9452366417408032e-05, | |
| "loss": 1.3311, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.1308883428573608, | |
| "learning_rate": 1.9450762383685725e-05, | |
| "loss": 0.6483, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9998852610588074, | |
| "learning_rate": 1.9449156070598397e-05, | |
| "loss": 1.7178, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8182191848754883, | |
| "learning_rate": 1.9447547478533467e-05, | |
| "loss": 1.9668, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.0509693622589111, | |
| "learning_rate": 1.9445936607878898e-05, | |
| "loss": 0.9926, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.0919716358184814, | |
| "learning_rate": 1.94443234590232e-05, | |
| "loss": 1.2947, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.7361246943473816, | |
| "learning_rate": 1.9442708032355444e-05, | |
| "loss": 1.4167, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.714868426322937, | |
| "learning_rate": 1.9441090328265242e-05, | |
| "loss": 1.5424, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8188365697860718, | |
| "learning_rate": 1.9439470347142756e-05, | |
| "loss": 1.7068, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9500496983528137, | |
| "learning_rate": 1.94378480893787e-05, | |
| "loss": 1.1715, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.612969994544983, | |
| "learning_rate": 1.9436223555364333e-05, | |
| "loss": 1.8178, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.6630369424819946, | |
| "learning_rate": 1.943459674549147e-05, | |
| "loss": 1.5356, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.4056222438812256, | |
| "learning_rate": 1.9432967660152456e-05, | |
| "loss": 1.0915, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9931894540786743, | |
| "learning_rate": 1.943133629974022e-05, | |
| "loss": 1.3489, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.721828281879425, | |
| "learning_rate": 1.9429702664648204e-05, | |
| "loss": 0.9029, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.1730399131774902, | |
| "learning_rate": 1.942806675527042e-05, | |
| "loss": 1.7768, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9179857969284058, | |
| "learning_rate": 1.9426428572001418e-05, | |
| "loss": 2.3638, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.0844107866287231, | |
| "learning_rate": 1.9424788115236303e-05, | |
| "loss": 1.5415, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.0590695142745972, | |
| "learning_rate": 1.942314538537072e-05, | |
| "loss": 1.6702, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.7648866772651672, | |
| "learning_rate": 1.9421500382800877e-05, | |
| "loss": 1.8638, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.0730507373809814, | |
| "learning_rate": 1.9419853107923515e-05, | |
| "loss": 1.5271, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8001253604888916, | |
| "learning_rate": 1.9418203561135927e-05, | |
| "loss": 2.9047, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8725544214248657, | |
| "learning_rate": 1.9416551742835964e-05, | |
| "loss": 1.4623, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.1127315759658813, | |
| "learning_rate": 1.941489765342201e-05, | |
| "loss": 1.7306, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.1793948411941528, | |
| "learning_rate": 1.9413241293293006e-05, | |
| "loss": 0.3196, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8964352011680603, | |
| "learning_rate": 1.9411582662848435e-05, | |
| "loss": 1.053, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.3153489828109741, | |
| "learning_rate": 1.9409921762488335e-05, | |
| "loss": 1.3615, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9132351279258728, | |
| "learning_rate": 1.9408258592613286e-05, | |
| "loss": 1.5991, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.7754179239273071, | |
| "learning_rate": 1.9406593153624416e-05, | |
| "loss": 0.8552, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.7985204458236694, | |
| "learning_rate": 1.94049254459234e-05, | |
| "loss": 0.6634, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9285178780555725, | |
| "learning_rate": 1.9403255469912467e-05, | |
| "loss": 0.345, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.291887640953064, | |
| "learning_rate": 1.9401583225994378e-05, | |
| "loss": 2.0273, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.5887540578842163, | |
| "learning_rate": 1.9399908714572457e-05, | |
| "loss": 1.6046, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.165919542312622, | |
| "learning_rate": 1.9398231936050562e-05, | |
| "loss": 1.8375, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.0102756023406982, | |
| "learning_rate": 1.939655289083311e-05, | |
| "loss": 1.3056, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.0535539388656616, | |
| "learning_rate": 1.9394871579325057e-05, | |
| "loss": 0.9489, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.4022819995880127, | |
| "learning_rate": 1.9393188001931906e-05, | |
| "loss": 0.8971, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.8239139318466187, | |
| "learning_rate": 1.9391502159059704e-05, | |
| "loss": 1.7716, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.7645108103752136, | |
| "learning_rate": 1.9389814051115054e-05, | |
| "loss": 1.6232, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.692162036895752, | |
| "learning_rate": 1.9388123678505095e-05, | |
| "loss": 1.0834, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.2660837173461914, | |
| "learning_rate": 1.938643104163752e-05, | |
| "loss": 1.2877, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.6457018256187439, | |
| "learning_rate": 1.9384736140920563e-05, | |
| "loss": 0.912, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.6962850093841553, | |
| "learning_rate": 1.9383038976763004e-05, | |
| "loss": 1.4546, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.3156652450561523, | |
| "learning_rate": 1.938133954957417e-05, | |
| "loss": 0.3989, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8944481611251831, | |
| "learning_rate": 1.937963785976394e-05, | |
| "loss": 1.1755, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.6729207038879395, | |
| "learning_rate": 1.9377933907742722e-05, | |
| "loss": 1.5848, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8458144068717957, | |
| "learning_rate": 1.9376227693921493e-05, | |
| "loss": 0.6623, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8411895632743835, | |
| "learning_rate": 1.9374519218711758e-05, | |
| "loss": 1.0984, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.1506187915802002, | |
| "learning_rate": 1.937280848252557e-05, | |
| "loss": 1.6137, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9453433156013489, | |
| "learning_rate": 1.9371095485775534e-05, | |
| "loss": 0.9001, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8604451417922974, | |
| "learning_rate": 1.9369380228874794e-05, | |
| "loss": 1.6193, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.1416794061660767, | |
| "learning_rate": 1.9367662712237043e-05, | |
| "loss": 1.8017, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.0399991273880005, | |
| "learning_rate": 1.9365942936276513e-05, | |
| "loss": 1.7701, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7908676862716675, | |
| "learning_rate": 1.9364220901407985e-05, | |
| "loss": 1.7858, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8512731790542603, | |
| "learning_rate": 1.9362496608046788e-05, | |
| "loss": 1.0902, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.198036789894104, | |
| "learning_rate": 1.936077005660879e-05, | |
| "loss": 1.3758, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.9642990231513977, | |
| "learning_rate": 1.9359041247510414e-05, | |
| "loss": 1.1954, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.0655434131622314, | |
| "learning_rate": 1.935731018116861e-05, | |
| "loss": 1.5582, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.4649953842163086, | |
| "learning_rate": 1.9355576858000885e-05, | |
| "loss": 1.4024, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.3037927150726318, | |
| "learning_rate": 1.9353841278425293e-05, | |
| "loss": 1.5681, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.0754226446151733, | |
| "learning_rate": 1.9352103442860415e-05, | |
| "loss": 1.432, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2181533575057983, | |
| "learning_rate": 1.93503633517254e-05, | |
| "loss": 1.4216, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7769268751144409, | |
| "learning_rate": 1.9348621005439917e-05, | |
| "loss": 2.0223, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.0469359159469604, | |
| "learning_rate": 1.9346876404424202e-05, | |
| "loss": 2.1334, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.877581834793091, | |
| "learning_rate": 1.9345129549099016e-05, | |
| "loss": 1.2296, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8810202479362488, | |
| "learning_rate": 1.9343380439885674e-05, | |
| "loss": 1.4057, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2003124952316284, | |
| "learning_rate": 1.934162907720603e-05, | |
| "loss": 0.7902, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8182734251022339, | |
| "learning_rate": 1.933987546148249e-05, | |
| "loss": 0.7547, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.8032336235046387, | |
| "learning_rate": 1.9338119593137985e-05, | |
| "loss": 1.6296, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7411457300186157, | |
| "learning_rate": 1.933636147259601e-05, | |
| "loss": 1.8137, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.225765347480774, | |
| "learning_rate": 1.933460110028059e-05, | |
| "loss": 1.3601, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.3987914323806763, | |
| "learning_rate": 1.9332838476616298e-05, | |
| "loss": 1.4657, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8765820264816284, | |
| "learning_rate": 1.933107360202825e-05, | |
| "loss": 1.5071, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7324680685997009, | |
| "learning_rate": 1.9329306476942102e-05, | |
| "loss": 0.7867, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.091023325920105, | |
| "learning_rate": 1.932753710178406e-05, | |
| "loss": 1.6422, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.734923243522644, | |
| "learning_rate": 1.932576547698086e-05, | |
| "loss": 0.3846, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.9723742008209229, | |
| "learning_rate": 1.9323991602959796e-05, | |
| "loss": 0.6102, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8249735236167908, | |
| "learning_rate": 1.932221548014869e-05, | |
| "loss": 1.4364, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7934770584106445, | |
| "learning_rate": 1.9320437108975915e-05, | |
| "loss": 1.431, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.0751607418060303, | |
| "learning_rate": 1.9318656489870387e-05, | |
| "loss": 1.5596, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2927584648132324, | |
| "learning_rate": 1.9316873623261555e-05, | |
| "loss": 2.008, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2063711881637573, | |
| "learning_rate": 1.9315088509579427e-05, | |
| "loss": 1.9501, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7476193308830261, | |
| "learning_rate": 1.9313301149254533e-05, | |
| "loss": 1.7395, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.9957605004310608, | |
| "learning_rate": 1.9311511542717956e-05, | |
| "loss": 1.3785, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.9378927946090698, | |
| "learning_rate": 1.9309719690401323e-05, | |
| "loss": 1.189, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.8630698919296265, | |
| "learning_rate": 1.9307925592736797e-05, | |
| "loss": 1.6241, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8105760216712952, | |
| "learning_rate": 1.930612925015708e-05, | |
| "loss": 1.3733, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.809337854385376, | |
| "learning_rate": 1.9304330663095426e-05, | |
| "loss": 2.8394, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.237175464630127, | |
| "learning_rate": 1.930252983198562e-05, | |
| "loss": 1.5124, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.9589737057685852, | |
| "learning_rate": 1.930072675726199e-05, | |
| "loss": 1.9335, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.1641108989715576, | |
| "learning_rate": 1.929892143935942e-05, | |
| "loss": 1.5772, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.968217670917511, | |
| "learning_rate": 1.9297113878713302e-05, | |
| "loss": 1.6706, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.723006546497345, | |
| "learning_rate": 1.9295304075759605e-05, | |
| "loss": 1.3032, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7680689692497253, | |
| "learning_rate": 1.9293492030934814e-05, | |
| "loss": 1.2522, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.8543024063110352, | |
| "learning_rate": 1.9291677744675976e-05, | |
| "loss": 1.2821, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7917808890342712, | |
| "learning_rate": 1.9289861217420652e-05, | |
| "loss": 1.6358, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8850444555282593, | |
| "learning_rate": 1.9288042449606963e-05, | |
| "loss": 1.5017, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8327540755271912, | |
| "learning_rate": 1.9286221441673566e-05, | |
| "loss": 0.6834, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8329907655715942, | |
| "learning_rate": 1.9284398194059662e-05, | |
| "loss": 1.3125, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2092479467391968, | |
| "learning_rate": 1.928257270720498e-05, | |
| "loss": 0.3374, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.6056079864501953, | |
| "learning_rate": 1.9280744981549803e-05, | |
| "loss": 1.4002, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.9556896090507507, | |
| "learning_rate": 1.927891501753494e-05, | |
| "loss": 1.4812, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7827396392822266, | |
| "learning_rate": 1.9277082815601756e-05, | |
| "loss": 0.5404, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.7754768133163452, | |
| "learning_rate": 1.9275248376192146e-05, | |
| "loss": 1.2932, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.1375806331634521, | |
| "learning_rate": 1.9273411699748542e-05, | |
| "loss": 2.6472, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.649649977684021, | |
| "learning_rate": 1.9271572786713915e-05, | |
| "loss": 1.7296, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.3229631185531616, | |
| "learning_rate": 1.9269731637531793e-05, | |
| "loss": 0.9259, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8701810836791992, | |
| "learning_rate": 1.9267888252646224e-05, | |
| "loss": 1.4846, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2067527770996094, | |
| "learning_rate": 1.9266042632501795e-05, | |
| "loss": 1.5756, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.9912710785865784, | |
| "learning_rate": 1.9264194777543646e-05, | |
| "loss": 1.3123, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2506612539291382, | |
| "learning_rate": 1.926234468821745e-05, | |
| "loss": 1.7554, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2464430332183838, | |
| "learning_rate": 1.9260492364969412e-05, | |
| "loss": 0.4939, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.230602741241455, | |
| "learning_rate": 1.9258637808246282e-05, | |
| "loss": 1.2158, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.0682477951049805, | |
| "learning_rate": 1.925678101849535e-05, | |
| "loss": 1.8281, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.7103515267372131, | |
| "learning_rate": 1.9254921996164446e-05, | |
| "loss": 1.6215, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.81144118309021, | |
| "learning_rate": 1.9253060741701928e-05, | |
| "loss": 1.1984, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.924333930015564, | |
| "learning_rate": 1.9251197255556703e-05, | |
| "loss": 1.8937, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1847920417785645, | |
| "learning_rate": 1.924933153817821e-05, | |
| "loss": 1.7972, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.7825921773910522, | |
| "learning_rate": 1.924746359001643e-05, | |
| "loss": 1.6789, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.7942408323287964, | |
| "learning_rate": 1.9245593411521882e-05, | |
| "loss": 1.5462, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1404311656951904, | |
| "learning_rate": 1.924372100314562e-05, | |
| "loss": 1.7976, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.3196184635162354, | |
| "learning_rate": 1.924184636533924e-05, | |
| "loss": 1.6385, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8548213243484497, | |
| "learning_rate": 1.9239969498554872e-05, | |
| "loss": 1.3902, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9516008496284485, | |
| "learning_rate": 1.923809040324518e-05, | |
| "loss": 1.4765, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.7390388250350952, | |
| "learning_rate": 1.9236209079863377e-05, | |
| "loss": 1.1018, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 3.578474760055542, | |
| "learning_rate": 1.9234325528863203e-05, | |
| "loss": 2.2269, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8223057389259338, | |
| "learning_rate": 1.9232439750698937e-05, | |
| "loss": 1.3773, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 3.3656997680664062, | |
| "learning_rate": 1.9230551745825403e-05, | |
| "loss": 1.4219, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9170312285423279, | |
| "learning_rate": 1.9228661514697952e-05, | |
| "loss": 1.106, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1395020484924316, | |
| "learning_rate": 1.9226769057772474e-05, | |
| "loss": 1.3624, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1166236400604248, | |
| "learning_rate": 1.92248743755054e-05, | |
| "loss": 0.3633, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8267510533332825, | |
| "learning_rate": 1.92229774683537e-05, | |
| "loss": 2.5529, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1344395875930786, | |
| "learning_rate": 1.922107833677487e-05, | |
| "loss": 1.8821, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9341800212860107, | |
| "learning_rate": 1.921917698122695e-05, | |
| "loss": 1.585, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8808992505073547, | |
| "learning_rate": 1.9217273402168515e-05, | |
| "loss": 2.2976, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.7030925154685974, | |
| "learning_rate": 1.9215367600058677e-05, | |
| "loss": 0.6211, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.045882225036621, | |
| "learning_rate": 1.9213459575357087e-05, | |
| "loss": 1.4662, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.7651109099388123, | |
| "learning_rate": 1.921154932852392e-05, | |
| "loss": 1.1614, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.2259867191314697, | |
| "learning_rate": 1.9209636860019902e-05, | |
| "loss": 1.4791, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.4694905281066895, | |
| "learning_rate": 1.9207722170306287e-05, | |
| "loss": 2.187, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8034918904304504, | |
| "learning_rate": 1.9205805259844865e-05, | |
| "loss": 0.8615, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.3326762914657593, | |
| "learning_rate": 1.920388612909796e-05, | |
| "loss": 0.9316, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1230294704437256, | |
| "learning_rate": 1.9201964778528438e-05, | |
| "loss": 1.5006, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.5623067617416382, | |
| "learning_rate": 1.92000412085997e-05, | |
| "loss": 1.3294, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8490066528320312, | |
| "learning_rate": 1.919811541977567e-05, | |
| "loss": 2.0144, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.8078815937042236, | |
| "learning_rate": 1.9196187412520823e-05, | |
| "loss": 1.7694, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1831212043762207, | |
| "learning_rate": 1.919425718730016e-05, | |
| "loss": 1.5831, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.4200443029403687, | |
| "learning_rate": 1.919232474457921e-05, | |
| "loss": 1.0131, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.700128972530365, | |
| "learning_rate": 1.9190390084824062e-05, | |
| "loss": 1.048, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.3507153987884521, | |
| "learning_rate": 1.9188453208501313e-05, | |
| "loss": 1.5529, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.0312564373016357, | |
| "learning_rate": 1.9186514116078105e-05, | |
| "loss": 0.7712, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1110159158706665, | |
| "learning_rate": 1.918457280802212e-05, | |
| "loss": 1.285, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8182725310325623, | |
| "learning_rate": 1.918262928480156e-05, | |
| "loss": 1.8237, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9742406010627747, | |
| "learning_rate": 1.918068354688518e-05, | |
| "loss": 1.7238, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.01756751537323, | |
| "learning_rate": 1.9178735594742255e-05, | |
| "loss": 1.2185, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1241008043289185, | |
| "learning_rate": 1.9176785428842598e-05, | |
| "loss": 1.3599, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.0219953060150146, | |
| "learning_rate": 1.9174833049656553e-05, | |
| "loss": 1.2674, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.339846134185791, | |
| "learning_rate": 1.917287845765501e-05, | |
| "loss": 2.0251, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.4480564594268799, | |
| "learning_rate": 1.9170921653309374e-05, | |
| "loss": 1.9456, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9150168895721436, | |
| "learning_rate": 1.9168962637091598e-05, | |
| "loss": 2.0659, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8925274014472961, | |
| "learning_rate": 1.9167001409474162e-05, | |
| "loss": 0.7544, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8011856079101562, | |
| "learning_rate": 1.9165037970930084e-05, | |
| "loss": 1.179, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.881730854511261, | |
| "learning_rate": 1.9163072321932913e-05, | |
| "loss": 1.688, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.4410722255706787, | |
| "learning_rate": 1.9161104462956723e-05, | |
| "loss": 1.3046, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.5103591680526733, | |
| "learning_rate": 1.9159134394476136e-05, | |
| "loss": 1.042, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.7637693285942078, | |
| "learning_rate": 1.9157162116966298e-05, | |
| "loss": 1.7315, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.6379944086074829, | |
| "learning_rate": 1.915518763090289e-05, | |
| "loss": 0.6313, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9315120577812195, | |
| "learning_rate": 1.9153210936762124e-05, | |
| "loss": 1.2323, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.7832790017127991, | |
| "learning_rate": 1.9151232035020738e-05, | |
| "loss": 0.9292, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.0305804014205933, | |
| "learning_rate": 1.9149250926156022e-05, | |
| "loss": 1.9421, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9271543622016907, | |
| "learning_rate": 1.9147267610645783e-05, | |
| "loss": 1.5984, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.8097622990608215, | |
| "learning_rate": 1.9145282088968355e-05, | |
| "loss": 0.8206, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.85881507396698, | |
| "learning_rate": 1.9143294361602618e-05, | |
| "loss": 1.6609, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1855149269104004, | |
| "learning_rate": 1.914130442902798e-05, | |
| "loss": 1.5922, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1737018823623657, | |
| "learning_rate": 1.9139312291724382e-05, | |
| "loss": 1.5697, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.0893067121505737, | |
| "learning_rate": 1.9137317950172287e-05, | |
| "loss": 1.292, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.2634906768798828, | |
| "learning_rate": 1.9135321404852703e-05, | |
| "loss": 1.5669, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.6731455326080322, | |
| "learning_rate": 1.9133322656247157e-05, | |
| "loss": 1.6371, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.3328168392181396, | |
| "learning_rate": 1.9131321704837716e-05, | |
| "loss": 1.695, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.3499823808670044, | |
| "learning_rate": 1.9129318551106977e-05, | |
| "loss": 0.5024, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.962078869342804, | |
| "learning_rate": 1.9127313195538065e-05, | |
| "loss": 1.4353, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9021065831184387, | |
| "learning_rate": 1.912530563861464e-05, | |
| "loss": 0.8511, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7653117775917053, | |
| "learning_rate": 1.9123295880820895e-05, | |
| "loss": 0.9386, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7594632506370544, | |
| "learning_rate": 1.912128392264154e-05, | |
| "loss": 1.3504, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.0727031230926514, | |
| "learning_rate": 1.9119269764561835e-05, | |
| "loss": 1.1641, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.2634934186935425, | |
| "learning_rate": 1.911725340706755e-05, | |
| "loss": 1.4128, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8265236020088196, | |
| "learning_rate": 1.911523485064501e-05, | |
| "loss": 0.6277, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.973920464515686, | |
| "learning_rate": 1.9113214095781046e-05, | |
| "loss": 1.6988, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8227178454399109, | |
| "learning_rate": 1.9111191142963037e-05, | |
| "loss": 2.415, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9979706406593323, | |
| "learning_rate": 1.910916599267888e-05, | |
| "loss": 1.2074, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.6832967400550842, | |
| "learning_rate": 1.9107138645417015e-05, | |
| "loss": 0.9535, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.93013334274292, | |
| "learning_rate": 1.9105109101666396e-05, | |
| "loss": 1.6132, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9972243309020996, | |
| "learning_rate": 1.9103077361916515e-05, | |
| "loss": 1.5842, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7579304575920105, | |
| "learning_rate": 1.9101043426657397e-05, | |
| "loss": 0.7145, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.286287546157837, | |
| "learning_rate": 1.9099007296379593e-05, | |
| "loss": 1.6661, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.0532516241073608, | |
| "learning_rate": 1.9096968971574183e-05, | |
| "loss": 1.0974, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.0688934326171875, | |
| "learning_rate": 1.9094928452732773e-05, | |
| "loss": 1.6377, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.725925624370575, | |
| "learning_rate": 1.909288574034751e-05, | |
| "loss": 1.4133, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.035366177558899, | |
| "learning_rate": 1.909084083491105e-05, | |
| "loss": 1.7386, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.6439036726951599, | |
| "learning_rate": 1.9088793736916608e-05, | |
| "loss": 0.8204, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.2139475345611572, | |
| "learning_rate": 1.9086744446857888e-05, | |
| "loss": 1.9203, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7955788373947144, | |
| "learning_rate": 1.908469296522916e-05, | |
| "loss": 1.7382, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9080290198326111, | |
| "learning_rate": 1.90826392925252e-05, | |
| "loss": 1.1714, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.4704135656356812, | |
| "learning_rate": 1.908058342924132e-05, | |
| "loss": 1.4689, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.841637372970581, | |
| "learning_rate": 1.9078525375873363e-05, | |
| "loss": 1.7025, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9378938674926758, | |
| "learning_rate": 1.907646513291769e-05, | |
| "loss": 1.7619, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.1404495239257812, | |
| "learning_rate": 1.9074402700871205e-05, | |
| "loss": 2.3074, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.888276994228363, | |
| "learning_rate": 1.9072338080231328e-05, | |
| "loss": 1.237, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.159399390220642, | |
| "learning_rate": 1.907027127149601e-05, | |
| "loss": 2.0084, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9085727334022522, | |
| "learning_rate": 1.9068202275163732e-05, | |
| "loss": 1.6468, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8967246413230896, | |
| "learning_rate": 1.90661310917335e-05, | |
| "loss": 1.8098, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7092412710189819, | |
| "learning_rate": 1.906405772170485e-05, | |
| "loss": 0.9415, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8793618679046631, | |
| "learning_rate": 1.9061982165577847e-05, | |
| "loss": 1.0499, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.6404721140861511, | |
| "learning_rate": 1.9059904423853072e-05, | |
| "loss": 0.7339, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9500699639320374, | |
| "learning_rate": 1.9057824497031647e-05, | |
| "loss": 1.8225, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7272951006889343, | |
| "learning_rate": 1.9055742385615215e-05, | |
| "loss": 0.8337, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7181074619293213, | |
| "learning_rate": 1.9053658090105944e-05, | |
| "loss": 0.4147, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.2174080610275269, | |
| "learning_rate": 1.9051571611006535e-05, | |
| "loss": 1.583, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7992798686027527, | |
| "learning_rate": 1.9049482948820207e-05, | |
| "loss": 2.0796, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9754629731178284, | |
| "learning_rate": 1.9047392104050718e-05, | |
| "loss": 1.2478, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.610386848449707, | |
| "learning_rate": 1.9045299077202336e-05, | |
| "loss": 0.5414, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7736746668815613, | |
| "learning_rate": 1.9043203868779863e-05, | |
| "loss": 1.1127, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7072888612747192, | |
| "learning_rate": 1.904110647928864e-05, | |
| "loss": 1.635, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.0180017948150635, | |
| "learning_rate": 1.903900690923451e-05, | |
| "loss": 1.3334, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8084598183631897, | |
| "learning_rate": 1.9036905159123854e-05, | |
| "loss": 1.8238, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.270559310913086, | |
| "learning_rate": 1.9034801229463593e-05, | |
| "loss": 1.4445, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.6077079772949219, | |
| "learning_rate": 1.903269512076114e-05, | |
| "loss": 1.3752, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.167836308479309, | |
| "learning_rate": 1.9030586833524468e-05, | |
| "loss": 1.3144, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 2.1038427352905273, | |
| "learning_rate": 1.902847636826205e-05, | |
| "loss": 1.3031, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.6927371621131897, | |
| "learning_rate": 1.902636372548291e-05, | |
| "loss": 1.3657, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.1325058937072754, | |
| "learning_rate": 1.9024248905696563e-05, | |
| "loss": 0.8486, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.2377064228057861, | |
| "learning_rate": 1.902213190941308e-05, | |
| "loss": 1.6585, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9400323033332825, | |
| "learning_rate": 1.9020012737143037e-05, | |
| "loss": 1.2736, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.0813266038894653, | |
| "learning_rate": 1.9017891389397555e-05, | |
| "loss": 1.6457, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.62490314245224, | |
| "learning_rate": 1.9015767866688253e-05, | |
| "loss": 0.1744, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9824820160865784, | |
| "learning_rate": 1.9013642169527297e-05, | |
| "loss": 1.5652, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9847505688667297, | |
| "learning_rate": 1.901151429842737e-05, | |
| "loss": 1.6858, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.6813574433326721, | |
| "learning_rate": 1.9009384253901674e-05, | |
| "loss": 1.2886, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.2777854204177856, | |
| "learning_rate": 1.9007252036463947e-05, | |
| "loss": 1.9161, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.0861783027648926, | |
| "learning_rate": 1.9005117646628437e-05, | |
| "loss": 1.3908, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.5930280685424805, | |
| "learning_rate": 1.9002981084909926e-05, | |
| "loss": 0.6656, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8518651723861694, | |
| "learning_rate": 1.9000842351823718e-05, | |
| "loss": 1.2132, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7691752910614014, | |
| "learning_rate": 1.8998701447885636e-05, | |
| "loss": 1.291, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.99872887134552, | |
| "learning_rate": 1.8996558373612032e-05, | |
| "loss": 1.6131, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9111050963401794, | |
| "learning_rate": 1.8994413129519782e-05, | |
| "loss": 1.1784, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9740219712257385, | |
| "learning_rate": 1.8992265716126277e-05, | |
| "loss": 1.6812, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.797783374786377, | |
| "learning_rate": 1.899011613394944e-05, | |
| "loss": 0.9716, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0966910123825073, | |
| "learning_rate": 1.8987964383507716e-05, | |
| "loss": 1.7276, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9825577735900879, | |
| "learning_rate": 1.8985810465320075e-05, | |
| "loss": 1.5695, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.992905855178833, | |
| "learning_rate": 1.8983654379905993e-05, | |
| "loss": 1.3159, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.6799482107162476, | |
| "learning_rate": 1.898149612778549e-05, | |
| "loss": 1.3488, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.2967268228530884, | |
| "learning_rate": 1.8979335709479102e-05, | |
| "loss": 1.7885, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.8869383335113525, | |
| "learning_rate": 1.897717312550788e-05, | |
| "loss": 1.1608, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.2248413562774658, | |
| "learning_rate": 1.8975008376393407e-05, | |
| "loss": 1.6525, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9131786823272705, | |
| "learning_rate": 1.897284146265778e-05, | |
| "loss": 2.7723, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7632516026496887, | |
| "learning_rate": 1.8970672384823627e-05, | |
| "loss": 1.8212, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.2612926959991455, | |
| "learning_rate": 1.8968501143414092e-05, | |
| "loss": 1.5892, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0912814140319824, | |
| "learning_rate": 1.896632773895284e-05, | |
| "loss": 1.7799, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7082846164703369, | |
| "learning_rate": 1.8964152171964056e-05, | |
| "loss": 1.6759, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9780619144439697, | |
| "learning_rate": 1.896197444297246e-05, | |
| "loss": 0.5018, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0766080617904663, | |
| "learning_rate": 1.8959794552503275e-05, | |
| "loss": 1.6004, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8449153900146484, | |
| "learning_rate": 1.895761250108226e-05, | |
| "loss": 1.7407, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8410694599151611, | |
| "learning_rate": 1.8955428289235685e-05, | |
| "loss": 1.6449, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7496964335441589, | |
| "learning_rate": 1.8953241917490346e-05, | |
| "loss": 0.9365, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.7132420539855957, | |
| "learning_rate": 1.895105338637356e-05, | |
| "loss": 0.9064, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8895228505134583, | |
| "learning_rate": 1.8948862696413163e-05, | |
| "loss": 0.8894, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.1113084554672241, | |
| "learning_rate": 1.8946669848137517e-05, | |
| "loss": 1.5976, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7062950134277344, | |
| "learning_rate": 1.8944474842075487e-05, | |
| "loss": 1.6246, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.3720945119857788, | |
| "learning_rate": 1.894227767875649e-05, | |
| "loss": 1.5104, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.2698320150375366, | |
| "learning_rate": 1.8940078358710432e-05, | |
| "loss": 1.5199, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9175028800964355, | |
| "learning_rate": 1.8937876882467755e-05, | |
| "loss": 1.1727, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7275259494781494, | |
| "learning_rate": 1.8935673250559422e-05, | |
| "loss": 1.2606, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.2364203929901123, | |
| "learning_rate": 1.8933467463516907e-05, | |
| "loss": 1.3575, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8634992837905884, | |
| "learning_rate": 1.893125952187221e-05, | |
| "loss": 1.2004, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0641099214553833, | |
| "learning_rate": 1.892904942615785e-05, | |
| "loss": 0.7576, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.3215057849884033, | |
| "learning_rate": 1.8926837176906864e-05, | |
| "loss": 1.9773, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0717252492904663, | |
| "learning_rate": 1.8924622774652808e-05, | |
| "loss": 1.9276, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9780475497245789, | |
| "learning_rate": 1.8922406219929767e-05, | |
| "loss": 1.5388, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8033444285392761, | |
| "learning_rate": 1.8920187513272323e-05, | |
| "loss": 1.4589, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.2679040431976318, | |
| "learning_rate": 1.8917966655215605e-05, | |
| "loss": 1.3719, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.5831549167633057, | |
| "learning_rate": 1.8915743646295236e-05, | |
| "loss": 1.2717, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9389986991882324, | |
| "learning_rate": 1.8913518487047374e-05, | |
| "loss": 1.4284, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0208168029785156, | |
| "learning_rate": 1.8911291178008685e-05, | |
| "loss": 1.6228, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.781409740447998, | |
| "learning_rate": 1.8909061719716365e-05, | |
| "loss": 1.1485, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.1682028770446777, | |
| "learning_rate": 1.890683011270812e-05, | |
| "loss": 1.843, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7864546775817871, | |
| "learning_rate": 1.8904596357522173e-05, | |
| "loss": 1.9129, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8679714798927307, | |
| "learning_rate": 1.8902360454697272e-05, | |
| "loss": 0.6919, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7505008578300476, | |
| "learning_rate": 1.890012240477267e-05, | |
| "loss": 0.3101, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8917982578277588, | |
| "learning_rate": 1.889788220828816e-05, | |
| "loss": 1.1233, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8512649536132812, | |
| "learning_rate": 1.8895639865784033e-05, | |
| "loss": 1.3445, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7751880288124084, | |
| "learning_rate": 1.8893395377801103e-05, | |
| "loss": 1.5494, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9000064134597778, | |
| "learning_rate": 1.8891148744880707e-05, | |
| "loss": 1.3592, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9392471313476562, | |
| "learning_rate": 1.8888899967564694e-05, | |
| "loss": 1.9797, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 3.50569224357605, | |
| "learning_rate": 1.888664904639543e-05, | |
| "loss": 2.9495, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9282636046409607, | |
| "learning_rate": 1.88843959819158e-05, | |
| "loss": 1.2952, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8388988375663757, | |
| "learning_rate": 1.88821407746692e-05, | |
| "loss": 0.6173, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.784787654876709, | |
| "learning_rate": 1.887988342519956e-05, | |
| "loss": 1.3081, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.6143754720687866, | |
| "learning_rate": 1.8877623934051305e-05, | |
| "loss": 1.3903, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.812180757522583, | |
| "learning_rate": 1.8875362301769387e-05, | |
| "loss": 0.611, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.1880741119384766, | |
| "learning_rate": 1.887309852889928e-05, | |
| "loss": 1.5304, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0518957376480103, | |
| "learning_rate": 1.8870832615986957e-05, | |
| "loss": 1.6762, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.70653235912323, | |
| "learning_rate": 1.8868564563578934e-05, | |
| "loss": 1.3218, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.9028378129005432, | |
| "learning_rate": 1.8866294372222208e-05, | |
| "loss": 1.746, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7664947509765625, | |
| "learning_rate": 1.8864022042464328e-05, | |
| "loss": 0.5167, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.0561689138412476, | |
| "learning_rate": 1.8861747574853334e-05, | |
| "loss": 1.4776, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7585257887840271, | |
| "learning_rate": 1.8859470969937786e-05, | |
| "loss": 1.5532, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7594607472419739, | |
| "learning_rate": 1.885719222826677e-05, | |
| "loss": 2.076, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9581306576728821, | |
| "learning_rate": 1.8854911350389878e-05, | |
| "loss": 1.8628, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9360955357551575, | |
| "learning_rate": 1.8852628336857213e-05, | |
| "loss": 1.2438, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.3346569538116455, | |
| "learning_rate": 1.885034318821941e-05, | |
| "loss": 1.5498, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.3662978410720825, | |
| "learning_rate": 1.8848055905027606e-05, | |
| "loss": 1.4224, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8976941704750061, | |
| "learning_rate": 1.8845766487833446e-05, | |
| "loss": 1.4752, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.6716092228889465, | |
| "learning_rate": 1.884347493718911e-05, | |
| "loss": 1.6956, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7952461242675781, | |
| "learning_rate": 1.8841181253647275e-05, | |
| "loss": 0.8128, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.6002814173698425, | |
| "learning_rate": 1.883888543776114e-05, | |
| "loss": 0.5184, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8543980717658997, | |
| "learning_rate": 1.883658749008442e-05, | |
| "loss": 2.2974, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.188140630722046, | |
| "learning_rate": 1.883428741117134e-05, | |
| "loss": 1.582, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.0838887691497803, | |
| "learning_rate": 1.8831985201576637e-05, | |
| "loss": 1.1045, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7349693179130554, | |
| "learning_rate": 1.8829680861855572e-05, | |
| "loss": 0.1732, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8598170876502991, | |
| "learning_rate": 1.8827374392563905e-05, | |
| "loss": 1.6251, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8577150702476501, | |
| "learning_rate": 1.8825065794257927e-05, | |
| "loss": 1.2437, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7986853718757629, | |
| "learning_rate": 1.8822755067494423e-05, | |
| "loss": 1.3749, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9996465444564819, | |
| "learning_rate": 1.882044221283071e-05, | |
| "loss": 1.3139, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.817579448223114, | |
| "learning_rate": 1.8818127230824602e-05, | |
| "loss": 1.2938, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8787328004837036, | |
| "learning_rate": 1.8815810122034443e-05, | |
| "loss": 2.1786, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.2396942377090454, | |
| "learning_rate": 1.881349088701907e-05, | |
| "loss": 1.6647, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.1219725608825684, | |
| "learning_rate": 1.8811169526337853e-05, | |
| "loss": 0.9283, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9217698574066162, | |
| "learning_rate": 1.880884604055066e-05, | |
| "loss": 1.4609, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.590294122695923, | |
| "learning_rate": 1.8806520430217875e-05, | |
| "loss": 0.4758, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.9318342208862305, | |
| "learning_rate": 1.8804192695900403e-05, | |
| "loss": 3.0273, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8690475821495056, | |
| "learning_rate": 1.8801862838159647e-05, | |
| "loss": 1.6065, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7368331551551819, | |
| "learning_rate": 1.879953085755754e-05, | |
| "loss": 0.8452, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8902385830879211, | |
| "learning_rate": 1.87971967546565e-05, | |
| "loss": 1.6896, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9730597138404846, | |
| "learning_rate": 1.879486053001949e-05, | |
| "loss": 1.8621, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.0001157522201538, | |
| "learning_rate": 1.8792522184209954e-05, | |
| "loss": 1.3968, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.2058511972427368, | |
| "learning_rate": 1.8790181717791874e-05, | |
| "loss": 1.6713, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.051774501800537, | |
| "learning_rate": 1.8787839131329724e-05, | |
| "loss": 1.8957, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.4670974016189575, | |
| "learning_rate": 1.8785494425388496e-05, | |
| "loss": 1.2512, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7796526551246643, | |
| "learning_rate": 1.8783147600533696e-05, | |
| "loss": 1.3797, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9571066498756409, | |
| "learning_rate": 1.8780798657331343e-05, | |
| "loss": 0.5489, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.775084912776947, | |
| "learning_rate": 1.8778447596347954e-05, | |
| "loss": 1.8418, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.788058340549469, | |
| "learning_rate": 1.8776094418150572e-05, | |
| "loss": 1.3976, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9418846964836121, | |
| "learning_rate": 1.877373912330674e-05, | |
| "loss": 1.1004, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9724819660186768, | |
| "learning_rate": 1.877138171238452e-05, | |
| "loss": 1.3961, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.0653592348098755, | |
| "learning_rate": 1.8769022185952473e-05, | |
| "loss": 1.6768, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.876054048538208, | |
| "learning_rate": 1.8766660544579683e-05, | |
| "loss": 2.3073, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.6656685471534729, | |
| "learning_rate": 1.8764296788835738e-05, | |
| "loss": 1.5729, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.6759817600250244, | |
| "learning_rate": 1.8761930919290732e-05, | |
| "loss": 1.8386, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.2566912174224854, | |
| "learning_rate": 1.8759562936515275e-05, | |
| "loss": 1.4726, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.0504961013793945, | |
| "learning_rate": 1.8757192841080486e-05, | |
| "loss": 1.5222, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7841604948043823, | |
| "learning_rate": 1.875482063355799e-05, | |
| "loss": 0.6235, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.355515480041504, | |
| "learning_rate": 1.875244631451993e-05, | |
| "loss": 1.8389, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8458920121192932, | |
| "learning_rate": 1.8750069884538943e-05, | |
| "loss": 2.1873, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.3264554738998413, | |
| "learning_rate": 1.8747691344188184e-05, | |
| "loss": 1.623, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.117638111114502, | |
| "learning_rate": 1.8745310694041327e-05, | |
| "loss": 1.2929, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.6209139823913574, | |
| "learning_rate": 1.8742927934672536e-05, | |
| "loss": 1.768, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.699492871761322, | |
| "learning_rate": 1.8740543066656494e-05, | |
| "loss": 0.6922, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9515039324760437, | |
| "learning_rate": 1.8738156090568394e-05, | |
| "loss": 0.8085, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7004430294036865, | |
| "learning_rate": 1.873576700698393e-05, | |
| "loss": 1.2552, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.6907755136489868, | |
| "learning_rate": 1.873337581647931e-05, | |
| "loss": 1.6774, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8606186509132385, | |
| "learning_rate": 1.8730982519631252e-05, | |
| "loss": 2.0097, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.428614616394043, | |
| "learning_rate": 1.872858711701698e-05, | |
| "loss": 1.707, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.1673688888549805, | |
| "learning_rate": 1.8726189609214217e-05, | |
| "loss": 2.1202, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.9758751392364502, | |
| "learning_rate": 1.8723789996801212e-05, | |
| "loss": 1.3832, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.65642249584198, | |
| "learning_rate": 1.87213882803567e-05, | |
| "loss": 1.607, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.4135712385177612, | |
| "learning_rate": 1.8718984460459945e-05, | |
| "loss": 1.4204, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9728422164916992, | |
| "learning_rate": 1.87165785376907e-05, | |
| "loss": 1.4806, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8608641624450684, | |
| "learning_rate": 1.871417051262924e-05, | |
| "loss": 1.145, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.4494624137878418, | |
| "learning_rate": 1.871176038585634e-05, | |
| "loss": 0.9609, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7167436480522156, | |
| "learning_rate": 1.870934815795327e-05, | |
| "loss": 1.6768, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.0592143535614014, | |
| "learning_rate": 1.870693382950183e-05, | |
| "loss": 1.425, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8277440667152405, | |
| "learning_rate": 1.8704517401084317e-05, | |
| "loss": 1.5469, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.2155778408050537, | |
| "learning_rate": 1.8702098873283524e-05, | |
| "loss": 1.2781, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.9879074096679688, | |
| "learning_rate": 1.869967824668277e-05, | |
| "loss": 0.9678, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.0017049312591553, | |
| "learning_rate": 1.8697255521865857e-05, | |
| "loss": 1.4003, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.220854640007019, | |
| "learning_rate": 1.8694830699417115e-05, | |
| "loss": 1.0837, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.319135308265686, | |
| "learning_rate": 1.8692403779921367e-05, | |
| "loss": 1.3641, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.3654942512512207, | |
| "learning_rate": 1.8689974763963946e-05, | |
| "loss": 1.4096, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.6490089893341064, | |
| "learning_rate": 1.868754365213069e-05, | |
| "loss": 1.6177, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8616020679473877, | |
| "learning_rate": 1.868511044500794e-05, | |
| "loss": 1.8009, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.3632266521453857, | |
| "learning_rate": 1.8682675143182553e-05, | |
| "loss": 1.3606, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.4812208414077759, | |
| "learning_rate": 1.868023774724187e-05, | |
| "loss": 1.2816, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.3317694664001465, | |
| "learning_rate": 1.8677798257773756e-05, | |
| "loss": 1.4442, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.9029511213302612, | |
| "learning_rate": 1.867535667536658e-05, | |
| "loss": 2.3617, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6913512945175171, | |
| "learning_rate": 1.86729130006092e-05, | |
| "loss": 0.091, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7284278869628906, | |
| "learning_rate": 1.8670467234091e-05, | |
| "loss": 0.8225, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7485316395759583, | |
| "learning_rate": 1.866801937640185e-05, | |
| "loss": 0.5161, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7488187551498413, | |
| "learning_rate": 1.8665569428132134e-05, | |
| "loss": 0.9691, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.095867395401001, | |
| "learning_rate": 1.8663117389872742e-05, | |
| "loss": 1.6216, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7063773274421692, | |
| "learning_rate": 1.866066326221506e-05, | |
| "loss": 2.0378, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7489952445030212, | |
| "learning_rate": 1.8658207045750983e-05, | |
| "loss": 0.5529, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6625726222991943, | |
| "learning_rate": 1.8655748741072913e-05, | |
| "loss": 0.449, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8839542269706726, | |
| "learning_rate": 1.8653288348773748e-05, | |
| "loss": 0.9134, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.936468243598938, | |
| "learning_rate": 1.8650825869446892e-05, | |
| "loss": 0.9317, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.4338470697402954, | |
| "learning_rate": 1.8648361303686258e-05, | |
| "loss": 0.4629, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.2629202604293823, | |
| "learning_rate": 1.8645894652086255e-05, | |
| "loss": 1.8685, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8673874735832214, | |
| "learning_rate": 1.86434259152418e-05, | |
| "loss": 1.9886, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.558154344558716, | |
| "learning_rate": 1.864095509374831e-05, | |
| "loss": 1.596, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8110612630844116, | |
| "learning_rate": 1.86384821882017e-05, | |
| "loss": 2.1369, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.3165100812911987, | |
| "learning_rate": 1.86360071991984e-05, | |
| "loss": 1.5489, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.0234098434448242, | |
| "learning_rate": 1.8633530127335336e-05, | |
| "loss": 1.7336, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6208418607711792, | |
| "learning_rate": 1.863105097320993e-05, | |
| "loss": 1.2285, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6910108923912048, | |
| "learning_rate": 1.8628569737420123e-05, | |
| "loss": 0.3964, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6648021936416626, | |
| "learning_rate": 1.8626086420564336e-05, | |
| "loss": 1.7522, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7676867246627808, | |
| "learning_rate": 1.862360102324151e-05, | |
| "loss": 0.5692, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6718749403953552, | |
| "learning_rate": 1.8621113546051078e-05, | |
| "loss": 0.6526, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6326141357421875, | |
| "learning_rate": 1.861862398959298e-05, | |
| "loss": 0.8584, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.0392323732376099, | |
| "learning_rate": 1.861613235446765e-05, | |
| "loss": 1.2752, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6072232127189636, | |
| "learning_rate": 1.861363864127604e-05, | |
| "loss": 0.5459, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7164649367332458, | |
| "learning_rate": 1.8611142850619576e-05, | |
| "loss": 0.7226, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.3524224758148193, | |
| "learning_rate": 1.8608644983100212e-05, | |
| "loss": 1.2067, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.5413652658462524, | |
| "learning_rate": 1.860614503932039e-05, | |
| "loss": 1.8641, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6611543297767639, | |
| "learning_rate": 1.8603643019883052e-05, | |
| "loss": 1.4456, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.1208839416503906, | |
| "learning_rate": 1.860113892539164e-05, | |
| "loss": 0.6717, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8838128447532654, | |
| "learning_rate": 1.859863275645011e-05, | |
| "loss": 0.5599, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7092893123626709, | |
| "learning_rate": 1.85961245136629e-05, | |
| "loss": 1.5641, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7692007422447205, | |
| "learning_rate": 1.8593614197634954e-05, | |
| "loss": 1.5014, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6679458618164062, | |
| "learning_rate": 1.8591101808971723e-05, | |
| "loss": 0.7755, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.6068809032440186, | |
| "learning_rate": 1.8588587348279153e-05, | |
| "loss": 1.421, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.9017466902732849, | |
| "learning_rate": 1.8586070816163685e-05, | |
| "loss": 1.1438, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6666081547737122, | |
| "learning_rate": 1.858355221323227e-05, | |
| "loss": 1.5628, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8469696640968323, | |
| "learning_rate": 1.8581031540092346e-05, | |
| "loss": 1.4695, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.9517692923545837, | |
| "learning_rate": 1.8578508797351863e-05, | |
| "loss": 1.4327, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.223913550376892, | |
| "learning_rate": 1.857598398561926e-05, | |
| "loss": 1.5891, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7739971876144409, | |
| "learning_rate": 1.8573457105503486e-05, | |
| "loss": 1.2211, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.1399494409561157, | |
| "learning_rate": 1.8570928157613972e-05, | |
| "loss": 0.9459, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.0470751523971558, | |
| "learning_rate": 1.8568397142560665e-05, | |
| "loss": 1.3632, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7569494247436523, | |
| "learning_rate": 1.8565864060954e-05, | |
| "loss": 1.6855, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.8458075523376465, | |
| "learning_rate": 1.8563328913404916e-05, | |
| "loss": 1.557, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8224319815635681, | |
| "learning_rate": 1.856079170052485e-05, | |
| "loss": 2.2359, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.700923502445221, | |
| "learning_rate": 1.8558252422925726e-05, | |
| "loss": 0.6053, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.0185672044754028, | |
| "learning_rate": 1.8555711081219983e-05, | |
| "loss": 0.9981, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.6706048846244812, | |
| "learning_rate": 1.8553167676020546e-05, | |
| "loss": 1.5536, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8932064175605774, | |
| "learning_rate": 1.8550622207940847e-05, | |
| "loss": 0.3132, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7320132851600647, | |
| "learning_rate": 1.8548074677594806e-05, | |
| "loss": 1.7206, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.715156614780426, | |
| "learning_rate": 1.8545525085596842e-05, | |
| "loss": 1.7057, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.8898051381111145, | |
| "learning_rate": 1.854297343256188e-05, | |
| "loss": 0.6021, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.8200762867927551, | |
| "learning_rate": 1.8540419719105332e-05, | |
| "loss": 1.696, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1514039039611816, | |
| "learning_rate": 1.853786394584311e-05, | |
| "loss": 1.2406, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 2.053684949874878, | |
| "learning_rate": 1.8535306113391628e-05, | |
| "loss": 2.1967, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1564501523971558, | |
| "learning_rate": 1.8532746222367784e-05, | |
| "loss": 1.748, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.001197099685669, | |
| "learning_rate": 1.8530184273388988e-05, | |
| "loss": 1.7622, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9028900861740112, | |
| "learning_rate": 1.8527620267073137e-05, | |
| "loss": 2.6661, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9210809469223022, | |
| "learning_rate": 1.852505420403863e-05, | |
| "loss": 0.931, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.8821693062782288, | |
| "learning_rate": 1.852248608490435e-05, | |
| "loss": 0.8729, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.8626199960708618, | |
| "learning_rate": 1.8519915910289686e-05, | |
| "loss": 1.3621, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.5940004587173462, | |
| "learning_rate": 1.8517343680814526e-05, | |
| "loss": 1.4931, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1320147514343262, | |
| "learning_rate": 1.8514769397099247e-05, | |
| "loss": 1.6531, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.357432246208191, | |
| "learning_rate": 1.8512193059764723e-05, | |
| "loss": 1.5411, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7297133803367615, | |
| "learning_rate": 1.8509614669432315e-05, | |
| "loss": 1.8413, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.4587827920913696, | |
| "learning_rate": 1.8507034226723903e-05, | |
| "loss": 1.4942, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7997937798500061, | |
| "learning_rate": 1.8504451732261834e-05, | |
| "loss": 1.1806, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1809964179992676, | |
| "learning_rate": 1.8501867186668968e-05, | |
| "loss": 1.589, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.045691967010498, | |
| "learning_rate": 1.849928059056865e-05, | |
| "loss": 1.8313, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9272171854972839, | |
| "learning_rate": 1.8496691944584726e-05, | |
| "loss": 1.4712, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.4044514894485474, | |
| "learning_rate": 1.8494101249341535e-05, | |
| "loss": 1.3147, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2790825366973877, | |
| "learning_rate": 1.8491508505463905e-05, | |
| "loss": 2.2231, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1366796493530273, | |
| "learning_rate": 1.8488913713577168e-05, | |
| "loss": 1.5343, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1620509624481201, | |
| "learning_rate": 1.8486316874307137e-05, | |
| "loss": 1.8474, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7728492617607117, | |
| "learning_rate": 1.8483717988280133e-05, | |
| "loss": 2.1768, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1838016510009766, | |
| "learning_rate": 1.848111705612296e-05, | |
| "loss": 1.1919, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2793999910354614, | |
| "learning_rate": 1.847851407846292e-05, | |
| "loss": 1.4788, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.3071448802947998, | |
| "learning_rate": 1.847590905592781e-05, | |
| "loss": 1.6954, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7144519090652466, | |
| "learning_rate": 1.847330198914591e-05, | |
| "loss": 1.7284, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.030076026916504, | |
| "learning_rate": 1.8470692878746007e-05, | |
| "loss": 1.4507, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9923644065856934, | |
| "learning_rate": 1.846808172535737e-05, | |
| "loss": 1.337, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.0590994358062744, | |
| "learning_rate": 1.8465468529609777e-05, | |
| "loss": 0.0855, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2966792583465576, | |
| "learning_rate": 1.846285329213347e-05, | |
| "loss": 1.3271, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.8333293199539185, | |
| "learning_rate": 1.8460236013559214e-05, | |
| "loss": 0.6277, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.8831344842910767, | |
| "learning_rate": 1.8457616694518243e-05, | |
| "loss": 1.3243, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1659544706344604, | |
| "learning_rate": 1.84549953356423e-05, | |
| "loss": 1.7745, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.8321329951286316, | |
| "learning_rate": 1.845237193756361e-05, | |
| "loss": 1.8983, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7504107356071472, | |
| "learning_rate": 1.844974650091489e-05, | |
| "loss": 1.7026, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7341120839118958, | |
| "learning_rate": 1.8447119026329355e-05, | |
| "loss": 1.3399, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.8567965626716614, | |
| "learning_rate": 1.8444489514440706e-05, | |
| "loss": 1.2334, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.0737907886505127, | |
| "learning_rate": 1.8441857965883138e-05, | |
| "loss": 1.4039, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.0732694864273071, | |
| "learning_rate": 1.8439224381291333e-05, | |
| "loss": 1.5017, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.0392206907272339, | |
| "learning_rate": 1.843658876130047e-05, | |
| "loss": 1.313, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.708148717880249, | |
| "learning_rate": 1.8433951106546217e-05, | |
| "loss": 0.5313, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7516269087791443, | |
| "learning_rate": 1.843131141766473e-05, | |
| "loss": 0.7228, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9159711599349976, | |
| "learning_rate": 1.8428669695292657e-05, | |
| "loss": 1.3187, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2239705324172974, | |
| "learning_rate": 1.8426025940067144e-05, | |
| "loss": 0.8966, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1255375146865845, | |
| "learning_rate": 1.8423380152625812e-05, | |
| "loss": 1.5237, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.1636478900909424, | |
| "learning_rate": 1.842073233360678e-05, | |
| "loss": 1.5424, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7868866920471191, | |
| "learning_rate": 1.8418082483648664e-05, | |
| "loss": 2.0077, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2694423198699951, | |
| "learning_rate": 1.8415430603390563e-05, | |
| "loss": 1.906, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7376890182495117, | |
| "learning_rate": 1.841277669347206e-05, | |
| "loss": 0.9802, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.036073088645935, | |
| "learning_rate": 1.8410120754533232e-05, | |
| "loss": 1.0298, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.0895626544952393, | |
| "learning_rate": 1.840746278721466e-05, | |
| "loss": 1.0902, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.4412864446640015, | |
| "learning_rate": 1.8404802792157385e-05, | |
| "loss": 1.8526, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.0933758020401, | |
| "learning_rate": 1.8402140770002965e-05, | |
| "loss": 1.4493, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.0682562589645386, | |
| "learning_rate": 1.839947672139343e-05, | |
| "loss": 1.8207, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.8491328358650208, | |
| "learning_rate": 1.83968106469713e-05, | |
| "loss": 1.4242, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2570806741714478, | |
| "learning_rate": 1.8394142547379594e-05, | |
| "loss": 1.3713, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.6603129506111145, | |
| "learning_rate": 1.839147242326181e-05, | |
| "loss": 1.7517, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9093148708343506, | |
| "learning_rate": 1.838880027526194e-05, | |
| "loss": 1.6046, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.0733606815338135, | |
| "learning_rate": 1.8386126104024456e-05, | |
| "loss": 1.8193, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9572824239730835, | |
| "learning_rate": 1.8383449910194325e-05, | |
| "loss": 1.5539, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.024169683456421, | |
| "learning_rate": 1.8380771694417003e-05, | |
| "loss": 0.9217, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.3997833728790283, | |
| "learning_rate": 1.8378091457338426e-05, | |
| "loss": 1.5381, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9886646270751953, | |
| "learning_rate": 1.8375409199605027e-05, | |
| "loss": 0.8998, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.8814927339553833, | |
| "learning_rate": 1.8372724921863717e-05, | |
| "loss": 1.5606, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.8278955221176147, | |
| "learning_rate": 1.83700386247619e-05, | |
| "loss": 1.3638, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.7573871612548828, | |
| "learning_rate": 1.8367350308947468e-05, | |
| "loss": 1.99, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.8589005470275879, | |
| "learning_rate": 1.8364659975068793e-05, | |
| "loss": 1.7079, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.7100195288658142, | |
| "learning_rate": 1.8361967623774745e-05, | |
| "loss": 1.6136, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9710107445716858, | |
| "learning_rate": 1.8359273255714664e-05, | |
| "loss": 1.3782, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.6068673133850098, | |
| "learning_rate": 1.8356576871538397e-05, | |
| "loss": 1.2857, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9702603220939636, | |
| "learning_rate": 1.8353878471896262e-05, | |
| "loss": 1.9289, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.045238971710205, | |
| "learning_rate": 1.8351178057439064e-05, | |
| "loss": 1.3155, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.8111006617546082, | |
| "learning_rate": 1.8348475628818104e-05, | |
| "loss": 0.4238, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.6043601632118225, | |
| "learning_rate": 1.834577118668516e-05, | |
| "loss": 0.3998, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.6240248084068298, | |
| "learning_rate": 1.8343064731692493e-05, | |
| "loss": 0.8843, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9114373922348022, | |
| "learning_rate": 1.8340356264492863e-05, | |
| "loss": 0.311, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9861066937446594, | |
| "learning_rate": 1.83376457857395e-05, | |
| "loss": 1.6434, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9441492557525635, | |
| "learning_rate": 1.833493329608613e-05, | |
| "loss": 1.1924, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.8383681178092957, | |
| "learning_rate": 1.833221879618696e-05, | |
| "loss": 1.2938, | |
| "step": 1203 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 6407, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 401, | |
| "total_flos": 4.536854588078162e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |