| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.06258779459965663, | |
| "eval_steps": 500, | |
| "global_step": 401, | |
| "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 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 6407, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 401, | |
| "total_flos": 1.5122848626927206e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |