| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.0, | |
| "eval_steps": 500, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 5.456483190001166, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 0.8054, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 5.759462172287989, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.8566, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 5.647493114157845, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.8604, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 5.062774089894794, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.8926, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.742178622247567, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.8096, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.094214554696164, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.8214, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 4.578250564308243, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.8371, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.914224760901885, | |
| "learning_rate": 1.0666666666666667e-05, | |
| "loss": 0.8386, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.104636711377158, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.768, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.741500054760724, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.7601, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.0673965583798566, | |
| "learning_rate": 1.4666666666666666e-05, | |
| "loss": 0.6711, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.3130829408950597, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.6996, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.4521476942644096, | |
| "learning_rate": 1.7333333333333336e-05, | |
| "loss": 0.6883, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.060188555365567, | |
| "learning_rate": 1.866666666666667e-05, | |
| "loss": 0.6712, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.5880322665772053, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6361, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.5065788447188164, | |
| "learning_rate": 1.999729241179462e-05, | |
| "loss": 0.6229, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.846665331936718, | |
| "learning_rate": 1.998917111338525e-05, | |
| "loss": 0.6138, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.397736554007089, | |
| "learning_rate": 1.9975640502598243e-05, | |
| "loss": 0.6224, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.250044207797159, | |
| "learning_rate": 1.9956707906498046e-05, | |
| "loss": 0.5978, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.1978733008887823, | |
| "learning_rate": 1.9932383577419432e-05, | |
| "loss": 0.5691, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.3616776529906953, | |
| "learning_rate": 1.9902680687415704e-05, | |
| "loss": 0.5554, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.097602231035553, | |
| "learning_rate": 1.9867615321125796e-05, | |
| "loss": 0.5255, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.0111196181375406, | |
| "learning_rate": 1.9827206467064133e-05, | |
| "loss": 0.4776, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.9955327255126635, | |
| "learning_rate": 1.9781476007338058e-05, | |
| "loss": 0.4818, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.9140689471373215, | |
| "learning_rate": 1.973044870579824e-05, | |
| "loss": 0.4888, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.8950005541922588, | |
| "learning_rate": 1.967415219462864e-05, | |
| "loss": 0.4678, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.9264924413390109, | |
| "learning_rate": 1.961261695938319e-05, | |
| "loss": 0.4916, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.9016767081799504, | |
| "learning_rate": 1.954587632247732e-05, | |
| "loss": 0.4914, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.8711206171973317, | |
| "learning_rate": 1.9473966425143292e-05, | |
| "loss": 0.4279, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.965284052630944, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 0.4769, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.8710629657905165, | |
| "learning_rate": 1.9314797389261426e-05, | |
| "loss": 0.3557, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.7829249544686165, | |
| "learning_rate": 1.9227624443554425e-05, | |
| "loss": 0.3159, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 1.1372489228974627, | |
| "learning_rate": 1.913545457642601e-05, | |
| "loss": 0.3496, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.8970662832168831, | |
| "learning_rate": 1.9038337699485207e-05, | |
| "loss": 0.3805, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 1.0612014407477506, | |
| "learning_rate": 1.8936326403234125e-05, | |
| "loss": 0.3481, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.8370235615255618, | |
| "learning_rate": 1.8829475928589272e-05, | |
| "loss": 0.3333, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 1.128753299885166, | |
| "learning_rate": 1.8717844136967626e-05, | |
| "loss": 0.3398, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 1.0163020931362265, | |
| "learning_rate": 1.860149147895366e-05, | |
| "loss": 0.3192, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.7411794509431812, | |
| "learning_rate": 1.848048096156426e-05, | |
| "loss": 0.2694, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.9250164203908562, | |
| "learning_rate": 1.8354878114129368e-05, | |
| "loss": 0.2863, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.8417377767403931, | |
| "learning_rate": 1.8224750952806626e-05, | |
| "loss": 0.2318, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.8880509994076226, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 0.2463, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.8319508153001826, | |
| "learning_rate": 1.795120796494848e-05, | |
| "loss": 0.2142, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.9069642204701174, | |
| "learning_rate": 1.7807940266766595e-05, | |
| "loss": 0.1861, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.886648766459474, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.1981, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.7878272604729076, | |
| "learning_rate": 1.7508800329814993e-05, | |
| "loss": 0.1927, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.859960892022644, | |
| "learning_rate": 1.735309008059829e-05, | |
| "loss": 0.2185, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.8198360072603958, | |
| "learning_rate": 1.7193398003386514e-05, | |
| "loss": 0.1932, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.8197303574303446, | |
| "learning_rate": 1.702981057425662e-05, | |
| "loss": 0.1896, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.7095776631735655, | |
| "learning_rate": 1.686241637868734e-05, | |
| "loss": 0.1808, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.782000326482277, | |
| "learning_rate": 1.6691306063588583e-05, | |
| "loss": 0.1546, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.757071977981014, | |
| "learning_rate": 1.6516572288214555e-05, | |
| "loss": 0.1219, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.7489429204163558, | |
| "learning_rate": 1.63383096739871e-05, | |
| "loss": 0.1047, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.7352612139748114, | |
| "learning_rate": 1.6156614753256583e-05, | |
| "loss": 0.1179, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.7350662711218962, | |
| "learning_rate": 1.5971585917027864e-05, | |
| "loss": 0.1137, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.7039275993910796, | |
| "learning_rate": 1.5783323361679865e-05, | |
| "loss": 0.1172, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.6910516646490666, | |
| "learning_rate": 1.5591929034707468e-05, | |
| "loss": 0.1024, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.7236787235958125, | |
| "learning_rate": 1.539750657951513e-05, | |
| "loss": 0.0989, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.6420220198866008, | |
| "learning_rate": 1.5200161279292154e-05, | |
| "loss": 0.1241, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6203363523773471, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.0817, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.6104409465029903, | |
| "learning_rate": 1.4797131132502464e-05, | |
| "loss": 0.0601, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.5844775253639324, | |
| "learning_rate": 1.4591664533870118e-05, | |
| "loss": 0.0627, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.5069649786509103, | |
| "learning_rate": 1.4383711467890776e-05, | |
| "loss": 0.0587, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.8916378878545967, | |
| "learning_rate": 1.417338454481818e-05, | |
| "loss": 0.0836, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.6085732114627656, | |
| "learning_rate": 1.396079766039157e-05, | |
| "loss": 0.0444, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.8766545204586843, | |
| "learning_rate": 1.3746065934159123e-05, | |
| "loss": 0.0671, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.5289923674586601, | |
| "learning_rate": 1.3529305647138689e-05, | |
| "loss": 0.0544, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.5631884598314233, | |
| "learning_rate": 1.3310634178849583e-05, | |
| "loss": 0.0532, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.5869614627369972, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "loss": 0.0594, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.49999369328894544, | |
| "learning_rate": 1.2868032327110904e-05, | |
| "loss": 0.0547, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.443457075418473, | |
| "learning_rate": 1.2644341620372025e-05, | |
| "loss": 0.0329, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.4438435027066986, | |
| "learning_rate": 1.2419218955996677e-05, | |
| "loss": 0.0284, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.4800778119424111, | |
| "learning_rate": 1.2192786241879033e-05, | |
| "loss": 0.0404, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.4379474529157175, | |
| "learning_rate": 1.1965166095328302e-05, | |
| "loss": 0.0225, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.4429369566894146, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.0223, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.40432933172297614, | |
| "learning_rate": 1.1506857122494832e-05, | |
| "loss": 0.0244, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.48073981290705153, | |
| "learning_rate": 1.127641647860595e-05, | |
| "loss": 0.0319, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.5001498731871603, | |
| "learning_rate": 1.1045284632676535e-05, | |
| "loss": 0.0307, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.3920879064027223, | |
| "learning_rate": 1.0813586746678584e-05, | |
| "loss": 0.0254, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.4091992943404657, | |
| "learning_rate": 1.0581448289104759e-05, | |
| "loss": 0.0259, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.24591634835172133, | |
| "learning_rate": 1.0348994967025012e-05, | |
| "loss": 0.0129, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.27732329015357815, | |
| "learning_rate": 1.0116352658013973e-05, | |
| "loss": 0.0118, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.2737809179611153, | |
| "learning_rate": 9.883647341986032e-06, | |
| "loss": 0.0125, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.30306810527445566, | |
| "learning_rate": 9.651005032974994e-06, | |
| "loss": 0.0154, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.30379855260514727, | |
| "learning_rate": 9.418551710895243e-06, | |
| "loss": 0.0118, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.4811355659404676, | |
| "learning_rate": 9.18641325332142e-06, | |
| "loss": 0.0155, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.37455252522236027, | |
| "learning_rate": 8.954715367323468e-06, | |
| "loss": 0.0109, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.3382983231236897, | |
| "learning_rate": 8.723583521394054e-06, | |
| "loss": 0.0152, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.3297109614209282, | |
| "learning_rate": 8.49314287750517e-06, | |
| "loss": 0.0119, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.3116900824217769, | |
| "learning_rate": 8.263518223330698e-06, | |
| "loss": 0.0099, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.2048883072228153, | |
| "learning_rate": 8.034833904671698e-06, | |
| "loss": 0.0068, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.16372430312031436, | |
| "learning_rate": 7.807213758120965e-06, | |
| "loss": 0.0046, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.17126616465729205, | |
| "learning_rate": 7.580781044003324e-06, | |
| "loss": 0.0045, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.17140249123203252, | |
| "learning_rate": 7.355658379627981e-06, | |
| "loss": 0.0054, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.21145963284827615, | |
| "learning_rate": 7.131967672889101e-06, | |
| "loss": 0.0064, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.16836913651183916, | |
| "learning_rate": 6.909830056250527e-06, | |
| "loss": 0.0038, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.18115546955306835, | |
| "learning_rate": 6.689365821150421e-06, | |
| "loss": 0.0052, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.2577468160241198, | |
| "learning_rate": 6.4706943528613135e-06, | |
| "loss": 0.0056, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.19659235565523647, | |
| "learning_rate": 6.25393406584088e-06, | |
| "loss": 0.0042, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.23535354069396494, | |
| "learning_rate": 6.039202339608432e-06, | |
| "loss": 0.0043, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.08757863007098991, | |
| "learning_rate": 5.8266154551818225e-06, | |
| "loss": 0.0022, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.11715962745878235, | |
| "learning_rate": 5.616288532109225e-06, | |
| "loss": 0.0022, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.12361827660694866, | |
| "learning_rate": 5.4083354661298816e-06, | |
| "loss": 0.002, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.13828988254954486, | |
| "learning_rate": 5.202868867497542e-06, | |
| "loss": 0.0025, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.1215485981095267, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.002, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.10696031651620225, | |
| "learning_rate": 4.799838720707847e-06, | |
| "loss": 0.0024, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.11537485121928975, | |
| "learning_rate": 4.6024934204848745e-06, | |
| "loss": 0.0019, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.11477645379676313, | |
| "learning_rate": 4.408070965292534e-06, | |
| "loss": 0.0023, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.09407959791293473, | |
| "learning_rate": 4.216676638320135e-06, | |
| "loss": 0.0017, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.10539572481392215, | |
| "learning_rate": 4.028414082972141e-06, | |
| "loss": 0.0022, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.04526592878878114, | |
| "learning_rate": 3.8433852467434175e-06, | |
| "loss": 0.0013, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.04747157304079231, | |
| "learning_rate": 3.661690326012897e-06, | |
| "loss": 0.0012, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.05989856905170414, | |
| "learning_rate": 3.483427711785449e-06, | |
| "loss": 0.0014, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.07405131299600318, | |
| "learning_rate": 3.308693936411421e-06, | |
| "loss": 0.0012, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.036105812123230004, | |
| "learning_rate": 3.1375836213126653e-06, | |
| "loss": 0.0012, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.04336874203309867, | |
| "learning_rate": 2.970189425743383e-06, | |
| "loss": 0.0015, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.03911723515161731, | |
| "learning_rate": 2.8066019966134907e-06, | |
| "loss": 0.0014, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.06319472371081573, | |
| "learning_rate": 2.6469099194017144e-06, | |
| "loss": 0.0013, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.06203490181593157, | |
| "learning_rate": 2.4911996701850083e-06, | |
| "loss": 0.0012, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.038461603051341636, | |
| "learning_rate": 2.339555568810221e-06, | |
| "loss": 0.0011, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.03151892994828768, | |
| "learning_rate": 2.192059733233408e-06, | |
| "loss": 0.0011, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.026069986960984427, | |
| "learning_rate": 2.048792035051521e-06, | |
| "loss": 0.001, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.029353134476379206, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 0.0011, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.026449026260907393, | |
| "learning_rate": 1.7752490471933769e-06, | |
| "loss": 0.001, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.024494407561961196, | |
| "learning_rate": 1.6451218858706374e-06, | |
| "loss": 0.001, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.02435497987035409, | |
| "learning_rate": 1.5195190384357405e-06, | |
| "loss": 0.0009, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.026479892781776853, | |
| "learning_rate": 1.3985085210463479e-06, | |
| "loss": 0.0008, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.02103523842744242, | |
| "learning_rate": 1.282155863032377e-06, | |
| "loss": 0.0009, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.026913938950377716, | |
| "learning_rate": 1.1705240714107301e-06, | |
| "loss": 0.0008, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.021497261239699632, | |
| "learning_rate": 1.0636735967658785e-06, | |
| "loss": 0.0009, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.1, | |
| "grad_norm": 0.01744937560052008, | |
| "learning_rate": 9.616623005147952e-07, | |
| "loss": 0.0008, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "grad_norm": 0.022830147389962408, | |
| "learning_rate": 8.645454235739903e-07, | |
| "loss": 0.001, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 13.3, | |
| "grad_norm": 0.018549892449922493, | |
| "learning_rate": 7.723755564455771e-07, | |
| "loss": 0.0008, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 13.4, | |
| "grad_norm": 0.018499963246780587, | |
| "learning_rate": 6.852026107385756e-07, | |
| "loss": 0.0007, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 13.5, | |
| "grad_norm": 0.020186180806129512, | |
| "learning_rate": 6.030737921409169e-07, | |
| "loss": 0.0008, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "grad_norm": 0.02160863937701566, | |
| "learning_rate": 5.26033574856708e-07, | |
| "loss": 0.0008, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 13.7, | |
| "grad_norm": 0.021380770597770406, | |
| "learning_rate": 4.5412367752268094e-07, | |
| "loss": 0.0009, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 13.8, | |
| "grad_norm": 0.017139391436868278, | |
| "learning_rate": 3.8738304061681107e-07, | |
| "loss": 0.0008, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 13.9, | |
| "grad_norm": 0.015543842724841007, | |
| "learning_rate": 3.2584780537136206e-07, | |
| "loss": 0.0006, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.023669058624946445, | |
| "learning_rate": 2.6955129420176193e-07, | |
| "loss": 0.0008, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 14.1, | |
| "grad_norm": 0.017273400705971477, | |
| "learning_rate": 2.1852399266194312e-07, | |
| "loss": 0.0007, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 14.2, | |
| "grad_norm": 0.017728316737897203, | |
| "learning_rate": 1.7279353293586765e-07, | |
| "loss": 0.0008, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 14.3, | |
| "grad_norm": 0.019827798313299654, | |
| "learning_rate": 1.323846788742078e-07, | |
| "loss": 0.0009, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "grad_norm": 0.016917306292850116, | |
| "learning_rate": 9.731931258429638e-08, | |
| "loss": 0.0007, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 14.5, | |
| "grad_norm": 0.01653817553682287, | |
| "learning_rate": 6.761642258056977e-08, | |
| "loss": 0.0007, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 14.6, | |
| "grad_norm": 0.03480683137636861, | |
| "learning_rate": 4.329209350195651e-08, | |
| "loss": 0.0007, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 14.7, | |
| "grad_norm": 0.020559555488437672, | |
| "learning_rate": 2.4359497401758026e-08, | |
| "loss": 0.001, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "grad_norm": 0.018105533298564763, | |
| "learning_rate": 1.0828886614754342e-08, | |
| "loss": 0.0008, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 14.9, | |
| "grad_norm": 0.019042313810839078, | |
| "learning_rate": 2.7075882053828605e-09, | |
| "loss": 0.0008, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.018967228362383, | |
| "learning_rate": 0.0, | |
| "loss": 0.0008, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "step": 150, | |
| "total_flos": 8.481547119440691e+16, | |
| "train_loss": 0.1808963591993476, | |
| "train_runtime": 6243.0048, | |
| "train_samples_per_second": 0.759, | |
| "train_steps_per_second": 0.024 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 150, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.481547119440691e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |