| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.968, | |
| "eval_steps": 500, | |
| "global_step": 123, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.024, | |
| "grad_norm": 6.8362836837768555, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.2519, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 6.583799362182617, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.23, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.072, | |
| "grad_norm": 6.952097415924072, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.2785, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 6.371545314788818, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.2499, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 4.713391304016113, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 1.2259, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 3.5113883018493652, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 1.1784, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.168, | |
| "grad_norm": 2.6190996170043945, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 1.1433, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 4.008308410644531, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.1378, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.216, | |
| "grad_norm": 4.964306831359863, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 1.1729, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 4.513827323913574, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 1.1446, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.264, | |
| "grad_norm": 3.6050338745117188, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 1.0779, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 3.471524715423584, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.0671, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.312, | |
| "grad_norm": 2.42974591255188, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0223, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 1.8660697937011719, | |
| "learning_rate": 9.997960964140946e-06, | |
| "loss": 1.0106, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.853407621383667, | |
| "learning_rate": 9.991845519630679e-06, | |
| "loss": 1.0557, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.5663851499557495, | |
| "learning_rate": 9.981658654313458e-06, | |
| "loss": 0.9931, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.408, | |
| "grad_norm": 1.4195172786712646, | |
| "learning_rate": 9.96740867674275e-06, | |
| "loss": 0.993, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 1.153138518333435, | |
| "learning_rate": 9.949107209404664e-06, | |
| "loss": 1.0047, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.456, | |
| "grad_norm": 1.4434843063354492, | |
| "learning_rate": 9.926769179238467e-06, | |
| "loss": 0.9929, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.1956053972244263, | |
| "learning_rate": 9.900412805461968e-06, | |
| "loss": 0.9942, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.504, | |
| "grad_norm": 1.0062505006790161, | |
| "learning_rate": 9.870059584711668e-06, | |
| "loss": 0.9831, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 1.0054383277893066, | |
| "learning_rate": 9.835734273509787e-06, | |
| "loss": 0.9784, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.552, | |
| "grad_norm": 0.9153076410293579, | |
| "learning_rate": 9.797464868072489e-06, | |
| "loss": 0.9665, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.9883344769477844, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.9633, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.9623836874961853, | |
| "learning_rate": 9.709221818197626e-06, | |
| "loss": 0.9732, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.8564513325691223, | |
| "learning_rate": 9.659320146057263e-06, | |
| "loss": 0.9188, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.648, | |
| "grad_norm": 0.7836757302284241, | |
| "learning_rate": 9.60561826557425e-06, | |
| "loss": 0.9723, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.8154799938201904, | |
| "learning_rate": 9.548159976772593e-06, | |
| "loss": 0.9103, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.696, | |
| "grad_norm": 0.8699765205383301, | |
| "learning_rate": 9.486992143456792e-06, | |
| "loss": 0.9487, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.6717429757118225, | |
| "learning_rate": 9.422164654989073e-06, | |
| "loss": 0.9657, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.744, | |
| "grad_norm": 0.7567694187164307, | |
| "learning_rate": 9.353730385598887e-06, | |
| "loss": 0.9785, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.7300078868865967, | |
| "learning_rate": 9.281745151257946e-06, | |
| "loss": 0.9331, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.792, | |
| "grad_norm": 0.8414798378944397, | |
| "learning_rate": 9.206267664155906e-06, | |
| "loss": 0.9369, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.7341222763061523, | |
| "learning_rate": 9.12735948481387e-06, | |
| "loss": 0.9246, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.7759191393852234, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.935, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.9772646427154541, | |
| "learning_rate": 8.959511229611377e-06, | |
| "loss": 0.9506, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.888, | |
| "grad_norm": 0.7034265995025635, | |
| "learning_rate": 8.870708053195414e-06, | |
| "loss": 0.9332, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.6552270650863647, | |
| "learning_rate": 8.778747871771293e-06, | |
| "loss": 0.8891, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.936, | |
| "grad_norm": 0.7311168313026428, | |
| "learning_rate": 8.683705689382025e-06, | |
| "loss": 0.9656, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6239702701568604, | |
| "learning_rate": 8.585659023794818e-06, | |
| "loss": 0.9594, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.984, | |
| "grad_norm": 0.7404515743255615, | |
| "learning_rate": 8.48468784327647e-06, | |
| "loss": 0.9467, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.016, | |
| "grad_norm": 1.015267252922058, | |
| "learning_rate": 8.380874501370098e-06, | |
| "loss": 1.4435, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.7437450885772705, | |
| "learning_rate": 8.274303669726427e-06, | |
| "loss": 0.9858, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.064, | |
| "grad_norm": 0.5530978441238403, | |
| "learning_rate": 8.165062269044353e-06, | |
| "loss": 0.7881, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6483109593391418, | |
| "learning_rate": 8.053239398177191e-06, | |
| "loss": 0.8529, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.112, | |
| "grad_norm": 0.6724779009819031, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.8778, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.1360000000000001, | |
| "grad_norm": 0.7351576089859009, | |
| "learning_rate": 7.822216094333847e-06, | |
| "loss": 0.9103, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.6864327192306519, | |
| "learning_rate": 7.703204087277989e-06, | |
| "loss": 0.848, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.184, | |
| "grad_norm": 0.8479780554771423, | |
| "learning_rate": 7.5819873081948105e-06, | |
| "loss": 0.896, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.208, | |
| "grad_norm": 0.6257553100585938, | |
| "learning_rate": 7.45866462322802e-06, | |
| "loss": 0.8881, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.232, | |
| "grad_norm": 0.6846205592155457, | |
| "learning_rate": 7.333336616128369e-06, | |
| "loss": 0.9257, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.256, | |
| "grad_norm": 0.970287561416626, | |
| "learning_rate": 7.206105506216107e-06, | |
| "loss": 0.8823, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.6447298526763916, | |
| "learning_rate": 7.0770750650094335e-06, | |
| "loss": 0.822, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.304, | |
| "grad_norm": 2.3928756713867188, | |
| "learning_rate": 6.946350531586959e-06, | |
| "loss": 0.7433, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.328, | |
| "grad_norm": 0.8941048979759216, | |
| "learning_rate": 6.814038526753205e-06, | |
| "loss": 0.9697, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.3519999999999999, | |
| "grad_norm": 0.6773728728294373, | |
| "learning_rate": 6.680246966077151e-06, | |
| "loss": 0.8388, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.376, | |
| "grad_norm": 0.6578003764152527, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.898, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.674293041229248, | |
| "learning_rate": 6.408662784207149e-06, | |
| "loss": 0.9156, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.424, | |
| "grad_norm": 0.6401296854019165, | |
| "learning_rate": 6.271091670967437e-06, | |
| "loss": 0.8781, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.448, | |
| "grad_norm": 0.6086285710334778, | |
| "learning_rate": 6.132483837128823e-06, | |
| "loss": 0.8464, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.6196046471595764, | |
| "learning_rate": 5.9929523332287275e-06, | |
| "loss": 0.8604, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.496, | |
| "grad_norm": 0.6998230814933777, | |
| "learning_rate": 5.85261096316312e-06, | |
| "loss": 0.9152, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.645214855670929, | |
| "learning_rate": 5.711574191366427e-06, | |
| "loss": 0.8257, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.544, | |
| "grad_norm": 0.7682178616523743, | |
| "learning_rate": 5.569957049452703e-06, | |
| "loss": 0.893, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.568, | |
| "grad_norm": 0.5541701912879944, | |
| "learning_rate": 5.4278750423942e-06, | |
| "loss": 0.8719, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.592, | |
| "grad_norm": 0.5373249053955078, | |
| "learning_rate": 5.285444054313841e-06, | |
| "loss": 0.8374, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.616, | |
| "grad_norm": 0.6957194805145264, | |
| "learning_rate": 5.142780253968481e-06, | |
| "loss": 0.9636, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 0.4742963910102844, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7453, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.6149660348892212, | |
| "learning_rate": 4.85721974603152e-06, | |
| "loss": 0.8727, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.688, | |
| "grad_norm": 0.6649055480957031, | |
| "learning_rate": 4.71455594568616e-06, | |
| "loss": 0.8197, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.712, | |
| "grad_norm": 0.5363019704818726, | |
| "learning_rate": 4.572124957605803e-06, | |
| "loss": 0.8977, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.736, | |
| "grad_norm": 0.5486201643943787, | |
| "learning_rate": 4.430042950547298e-06, | |
| "loss": 0.8524, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.7598503232002258, | |
| "learning_rate": 4.2884258086335755e-06, | |
| "loss": 0.8718, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.784, | |
| "grad_norm": 0.5470713973045349, | |
| "learning_rate": 4.147389036836881e-06, | |
| "loss": 0.8436, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.808, | |
| "grad_norm": 0.5422508120536804, | |
| "learning_rate": 4.007047666771274e-06, | |
| "loss": 0.9016, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.8319999999999999, | |
| "grad_norm": 0.563028872013092, | |
| "learning_rate": 3.867516162871177e-06, | |
| "loss": 0.8374, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.5209132432937622, | |
| "learning_rate": 3.7289083290325668e-06, | |
| "loss": 0.833, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.5739846229553223, | |
| "learning_rate": 3.5913372157928515e-06, | |
| "loss": 0.9076, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.904, | |
| "grad_norm": 0.5330556035041809, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.8753, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.928, | |
| "grad_norm": 0.4843089282512665, | |
| "learning_rate": 3.319753033922849e-06, | |
| "loss": 0.8169, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.952, | |
| "grad_norm": 0.5041735172271729, | |
| "learning_rate": 3.1859614732467957e-06, | |
| "loss": 0.8877, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.976, | |
| "grad_norm": 0.5104906558990479, | |
| "learning_rate": 3.053649468413043e-06, | |
| "loss": 0.7909, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.008, | |
| "grad_norm": 1.0316483974456787, | |
| "learning_rate": 2.9229249349905686e-06, | |
| "loss": 1.3689, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 0.5285851955413818, | |
| "learning_rate": 2.7938944937838924e-06, | |
| "loss": 0.8718, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.056, | |
| "grad_norm": 0.5638172626495361, | |
| "learning_rate": 2.6666633838716317e-06, | |
| "loss": 0.8144, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.4385935962200165, | |
| "learning_rate": 2.5413353767719805e-06, | |
| "loss": 0.7497, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.104, | |
| "grad_norm": 0.5555351376533508, | |
| "learning_rate": 2.418012691805191e-06, | |
| "loss": 0.9051, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 0.5720311999320984, | |
| "learning_rate": 2.296795912722014e-06, | |
| "loss": 0.8908, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.152, | |
| "grad_norm": 0.46188703179359436, | |
| "learning_rate": 2.1777839056661555e-06, | |
| "loss": 0.7216, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.48220935463905334, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.8029, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.49016672372817993, | |
| "learning_rate": 1.946760601822809e-06, | |
| "loss": 0.9015, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.224, | |
| "grad_norm": 0.44178417325019836, | |
| "learning_rate": 1.8349377309556487e-06, | |
| "loss": 0.7919, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.248, | |
| "grad_norm": 0.5433825850486755, | |
| "learning_rate": 1.7256963302735752e-06, | |
| "loss": 0.7993, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.2720000000000002, | |
| "grad_norm": 0.39549779891967773, | |
| "learning_rate": 1.6191254986299044e-06, | |
| "loss": 0.7369, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.296, | |
| "grad_norm": 0.48636099696159363, | |
| "learning_rate": 1.5153121567235334e-06, | |
| "loss": 0.9214, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.4830661118030548, | |
| "learning_rate": 1.4143409762051829e-06, | |
| "loss": 0.8425, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.344, | |
| "grad_norm": 0.42800870537757874, | |
| "learning_rate": 1.3162943106179748e-06, | |
| "loss": 0.7417, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.46301019191741943, | |
| "learning_rate": 1.2212521282287093e-06, | |
| "loss": 0.8874, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.392, | |
| "grad_norm": 0.4442095458507538, | |
| "learning_rate": 1.1292919468045876e-06, | |
| "loss": 0.8052, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.416, | |
| "grad_norm": 0.49116918444633484, | |
| "learning_rate": 1.0404887703886252e-06, | |
| "loss": 0.8252, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.45165371894836426, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.7867, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.464, | |
| "grad_norm": 0.4509302079677582, | |
| "learning_rate": 8.7264051518613e-07, | |
| "loss": 0.9062, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.488, | |
| "grad_norm": 0.45219430327415466, | |
| "learning_rate": 7.937323358440935e-07, | |
| "loss": 0.7844, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 0.39615580439567566, | |
| "learning_rate": 7.182548487420555e-07, | |
| "loss": 0.7702, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.536, | |
| "grad_norm": 0.5083522796630859, | |
| "learning_rate": 6.462696144011149e-07, | |
| "loss": 0.9273, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.40505877137184143, | |
| "learning_rate": 5.778353450109286e-07, | |
| "loss": 0.8128, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.584, | |
| "grad_norm": 0.4423372447490692, | |
| "learning_rate": 5.130078565432089e-07, | |
| "loss": 0.7965, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.608, | |
| "grad_norm": 0.448724627494812, | |
| "learning_rate": 4.5184002322740784e-07, | |
| "loss": 0.7753, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.632, | |
| "grad_norm": 0.4123077690601349, | |
| "learning_rate": 3.9438173442575e-07, | |
| "loss": 0.8232, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.656, | |
| "grad_norm": 0.4331120550632477, | |
| "learning_rate": 3.406798539427386e-07, | |
| "loss": 0.8859, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.463981956243515, | |
| "learning_rate": 2.9077818180237693e-07, | |
| "loss": 0.81, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.7039999999999997, | |
| "grad_norm": 0.4227185547351837, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.7866, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.7279999999999998, | |
| "grad_norm": 0.4594978392124176, | |
| "learning_rate": 2.0253513192751374e-07, | |
| "loss": 0.8428, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.4640582799911499, | |
| "learning_rate": 1.6426572649021477e-07, | |
| "loss": 0.8939, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.776, | |
| "grad_norm": 0.410780668258667, | |
| "learning_rate": 1.2994041528833267e-07, | |
| "loss": 0.7411, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.4083250164985657, | |
| "learning_rate": 9.958719453803278e-08, | |
| "loss": 0.8683, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.824, | |
| "grad_norm": 0.36987847089767456, | |
| "learning_rate": 7.32308207615351e-08, | |
| "loss": 0.7558, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.848, | |
| "grad_norm": 0.3830074965953827, | |
| "learning_rate": 5.089279059533658e-08, | |
| "loss": 0.8281, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.872, | |
| "grad_norm": 0.4141235947608948, | |
| "learning_rate": 3.25913232572489e-08, | |
| "loss": 0.8886, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 0.3976007103919983, | |
| "learning_rate": 1.834134568654333e-08, | |
| "loss": 0.7874, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.447281152009964, | |
| "learning_rate": 8.15448036932176e-09, | |
| "loss": 0.8983, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3919880986213684, | |
| "learning_rate": 2.0390358590538507e-09, | |
| "loss": 0.6968, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.968, | |
| "grad_norm": 0.42730164527893066, | |
| "learning_rate": 0.0, | |
| "loss": 0.8298, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.968, | |
| "step": 123, | |
| "total_flos": 121724209528832.0, | |
| "train_loss": 0.9151467346563572, | |
| "train_runtime": 7021.3239, | |
| "train_samples_per_second": 1.709, | |
| "train_steps_per_second": 0.018 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 123, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 121724209528832.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |