| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 300, | |
| "global_step": 1443, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002079002079002079, | |
| "grad_norm": 21.0, | |
| "learning_rate": 1e-06, | |
| "loss": 1.4443, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.010395010395010396, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 1.9999786113807535e-06, | |
| "loss": 1.4249, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02079002079002079, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 1.999847906465395e-06, | |
| "loss": 1.3779, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031185031185031187, | |
| "grad_norm": 4.125, | |
| "learning_rate": 1.999598394713099e-06, | |
| "loss": 1.3446, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04158004158004158, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 1.9992301057720783e-06, | |
| "loss": 1.3445, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05197505197505198, | |
| "grad_norm": 3.75, | |
| "learning_rate": 1.9987430834042318e-06, | |
| "loss": 1.3221, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.062370062370062374, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 1.9981373854799487e-06, | |
| "loss": 1.3312, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07276507276507277, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.99741308397123e-06, | |
| "loss": 1.3108, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08316008316008316, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 1.9965702649431362e-06, | |
| "loss": 1.3148, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09355509355509356, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.9956090285435628e-06, | |
| "loss": 1.3075, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10395010395010396, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 1.9945294889913377e-06, | |
| "loss": 1.3029, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11434511434511435, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 1.9933317745626504e-06, | |
| "loss": 1.3022, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12474012474012475, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 1.9920160275758096e-06, | |
| "loss": 1.2885, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13513513513513514, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.9905824043743316e-06, | |
| "loss": 1.2944, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14553014553014554, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 1.9890310753083633e-06, | |
| "loss": 1.2993, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15592515592515593, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 1.987362224714441e-06, | |
| "loss": 1.2885, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16632016632016633, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.985576050893585e-06, | |
| "loss": 1.2862, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17671517671517672, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.983672766087738e-06, | |
| "loss": 1.2854, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.18711018711018712, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.9816525964545446e-06, | |
| "loss": 1.2765, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19750519750519752, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.979515782040478e-06, | |
| "loss": 1.2874, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2079002079002079, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.977262576752319e-06, | |
| "loss": 1.2808, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2182952182952183, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.97489324832698e-06, | |
| "loss": 1.2768, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2286902286902287, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.972408078299698e-06, | |
| "loss": 1.2757, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2390852390852391, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.969807361970575e-06, | |
| "loss": 1.2897, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2494802494802495, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.9670914083694948e-06, | |
| "loss": 1.2796, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2598752598752599, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.9642605402193976e-06, | |
| "loss": 1.2684, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2702702702702703, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.9613150938979343e-06, | |
| "loss": 1.2694, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2806652806652807, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.958255419397498e-06, | |
| "loss": 1.263, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2910602910602911, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.9550818802836337e-06, | |
| "loss": 1.271, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30145530145530147, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.9517948536518397e-06, | |
| "loss": 1.2638, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31185031185031187, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.9483947300827577e-06, | |
| "loss": 1.2633, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.32224532224532226, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 1.944881913595762e-06, | |
| "loss": 1.2648, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33264033264033266, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 1.941256821600955e-06, | |
| "loss": 1.2588, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34303534303534305, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.9375198848495637e-06, | |
| "loss": 1.258, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.35343035343035345, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.9336715473827594e-06, | |
| "loss": 1.2588, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36382536382536385, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.929712266478893e-06, | |
| "loss": 1.2538, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.37422037422037424, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 1.9256425125991596e-06, | |
| "loss": 1.2539, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.9214627693316952e-06, | |
| "loss": 1.2556, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39501039501039503, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.9171735333341148e-06, | |
| "loss": 1.2524, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.40540540540540543, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.9127753142744977e-06, | |
| "loss": 1.2471, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4158004158004158, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.9082686347708253e-06, | |
| "loss": 1.246, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4261954261954262, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 1.9036540303288815e-06, | |
| "loss": 1.247, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4365904365904366, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.898932049278621e-06, | |
| "loss": 1.2496, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.446985446985447, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 1.8941032527090148e-06, | |
| "loss": 1.2492, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4573804573804574, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.8891682144013777e-06, | |
| "loss": 1.2476, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4677754677754678, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 1.8841275207611895e-06, | |
| "loss": 1.2426, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4781704781704782, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.8789817707484162e-06, | |
| "loss": 1.2388, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4885654885654886, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.8737315758063364e-06, | |
| "loss": 1.2374, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.498960498960499, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 1.8683775597888886e-06, | |
| "loss": 1.2437, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5093555093555093, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.8629203588865419e-06, | |
| "loss": 1.2335, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5197505197505198, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.8573606215506985e-06, | |
| "loss": 1.2355, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5301455301455301, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 1.8516990084166443e-06, | |
| "loss": 1.2384, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5405405405405406, | |
| "grad_norm": 3.625, | |
| "learning_rate": 1.8459361922250469e-06, | |
| "loss": 1.232, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5509355509355509, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.8400728577420187e-06, | |
| "loss": 1.2356, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5613305613305614, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.8341097016777484e-06, | |
| "loss": 1.2408, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5717255717255717, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 1.8280474326037155e-06, | |
| "loss": 1.2465, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5821205821205822, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.8218867708684937e-06, | |
| "loss": 1.2331, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5925155925155925, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 1.8156284485121556e-06, | |
| "loss": 1.2323, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6029106029106029, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 1.8092732091792884e-06, | |
| "loss": 1.2365, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6133056133056133, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.8028218080306302e-06, | |
| "loss": 1.2394, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6237006237006237, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.7962750116533387e-06, | |
| "loss": 1.2263, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6237006237006237, | |
| "eval_loss": 1.2331745624542236, | |
| "eval_runtime": 10.8501, | |
| "eval_samples_per_second": 85.713, | |
| "eval_steps_per_second": 2.765, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6340956340956341, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.7896335979699001e-06, | |
| "loss": 1.2311, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6444906444906445, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 1.782898356145694e-06, | |
| "loss": 1.2236, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6548856548856549, | |
| "grad_norm": 4.25, | |
| "learning_rate": 1.7760700864952205e-06, | |
| "loss": 1.2376, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6652806652806653, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 1.7691496003870018e-06, | |
| "loss": 1.2239, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6756756756756757, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.7621377201471735e-06, | |
| "loss": 1.2311, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6860706860706861, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 1.75503527896177e-06, | |
| "loss": 1.226, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6964656964656964, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 1.7478431207777215e-06, | |
| "loss": 1.228, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7068607068607069, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 1.7405621002025735e-06, | |
| "loss": 1.2216, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7172557172557172, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 1.733193082402936e-06, | |
| "loss": 1.2223, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7276507276507277, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.7257369430016817e-06, | |
| "loss": 1.2158, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.738045738045738, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 1.7181945679739003e-06, | |
| "loss": 1.2206, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7484407484407485, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.7105668535416205e-06, | |
| "loss": 1.2292, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7588357588357588, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.7028547060673197e-06, | |
| "loss": 1.2272, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.6950590419462229e-06, | |
| "loss": 1.2264, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7796257796257796, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.687180787497413e-06, | |
| "loss": 1.2155, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7900207900207901, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 1.6792208788537617e-06, | |
| "loss": 1.2154, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8004158004158004, | |
| "grad_norm": 3.625, | |
| "learning_rate": 1.6711802618506926e-06, | |
| "loss": 1.2134, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8108108108108109, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.663059891913793e-06, | |
| "loss": 1.2211, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8212058212058212, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 1.6548607339452852e-06, | |
| "loss": 1.2222, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8316008316008316, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.6465837622093722e-06, | |
| "loss": 1.2188, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.841995841995842, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.6382299602164706e-06, | |
| "loss": 1.2189, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8523908523908524, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.6298003206063466e-06, | |
| "loss": 1.2145, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8627858627858628, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.6212958450301625e-06, | |
| "loss": 1.2168, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8731808731808732, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.6127175440314594e-06, | |
| "loss": 1.2083, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8835758835758836, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 1.6040664369260758e-06, | |
| "loss": 1.2183, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.893970893970894, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 1.5953435516810303e-06, | |
| "loss": 1.2134, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9043659043659044, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.586549924792372e-06, | |
| "loss": 1.2145, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9147609147609148, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.5776866011620198e-06, | |
| "loss": 1.2148, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9251559251559252, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 1.5687546339736013e-06, | |
| "loss": 1.212, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9355509355509356, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.559755084567309e-06, | |
| "loss": 1.2183, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9459459459459459, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.5506890223137857e-06, | |
| "loss": 1.2123, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9563409563409564, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.5415575244870578e-06, | |
| "loss": 1.214, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9667359667359667, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.5323616761365278e-06, | |
| "loss": 1.2081, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9771309771309772, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.5231025699580427e-06, | |
| "loss": 1.1995, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9875259875259875, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 1.513781306164056e-06, | |
| "loss": 1.2226, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.997920997920998, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.5043989923528937e-06, | |
| "loss": 1.2183, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0083160083160083, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 1.4949567433771448e-06, | |
| "loss": 1.1826, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.0187110187110187, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.4854556812111887e-06, | |
| "loss": 1.1925, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0291060291060292, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.4758969348178766e-06, | |
| "loss": 1.19, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.0395010395010396, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.4662816400143836e-06, | |
| "loss": 1.1963, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.04989604989605, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.4566109393372433e-06, | |
| "loss": 1.1872, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0602910602910602, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.4468859819065882e-06, | |
| "loss": 1.1833, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0706860706860706, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.4371079232896044e-06, | |
| "loss": 1.1815, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.0810810810810811, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.4272779253632212e-06, | |
| "loss": 1.1855, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0914760914760915, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.4173971561760518e-06, | |
| "loss": 1.188, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.1018711018711018, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.4074667898096009e-06, | |
| "loss": 1.1873, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.1122661122661124, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.397488006238752e-06, | |
| "loss": 1.1945, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1226611226611227, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.387461991191559e-06, | |
| "loss": 1.1856, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.133056133056133, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.3773899360083524e-06, | |
| "loss": 1.1868, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.1434511434511434, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.3672730375001773e-06, | |
| "loss": 1.1791, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 1.357112497806582e-06, | |
| "loss": 1.1969, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1642411642411643, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.3469095242527764e-06, | |
| "loss": 1.1828, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.1746361746361746, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.3366653292061682e-06, | |
| "loss": 1.1803, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.185031185031185, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 1.3263811299323063e-06, | |
| "loss": 1.1803, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.1954261954261955, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 1.3160581484502382e-06, | |
| "loss": 1.1823, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.2058212058212059, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 1.3056976113873037e-06, | |
| "loss": 1.1832, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2162162162162162, | |
| "grad_norm": 3.625, | |
| "learning_rate": 1.2953007498333807e-06, | |
| "loss": 1.1841, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.2266112266112266, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.284868799194602e-06, | |
| "loss": 1.1804, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.237006237006237, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.2744029990465574e-06, | |
| "loss": 1.1811, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.2474012474012475, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.2639045929870018e-06, | |
| "loss": 1.1794, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2474012474012475, | |
| "eval_loss": 1.2036519050598145, | |
| "eval_runtime": 10.8624, | |
| "eval_samples_per_second": 85.616, | |
| "eval_steps_per_second": 2.762, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2577962577962578, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.2533748284880842e-06, | |
| "loss": 1.1905, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.2681912681912682, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.2428149567481184e-06, | |
| "loss": 1.1836, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.2785862785862787, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.2322262325429063e-06, | |
| "loss": 1.1823, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.288981288981289, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.2216099140766436e-06, | |
| "loss": 1.1836, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.2993762993762994, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 1.2109672628324104e-06, | |
| "loss": 1.1837, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.3097713097713097, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.2002995434222767e-06, | |
| "loss": 1.1827, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.32016632016632, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 1.1896080234370355e-06, | |
| "loss": 1.1803, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.3305613305613306, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 1.178893973295581e-06, | |
| "loss": 1.1788, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.340956340956341, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.1681586660939504e-06, | |
| "loss": 1.1918, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.3513513513513513, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.1574033774540505e-06, | |
| "loss": 1.1796, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.3617463617463619, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 1.1466293853720795e-06, | |
| "loss": 1.1837, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.3721413721413722, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.1358379700666703e-06, | |
| "loss": 1.1776, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.3825363825363826, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.1250304138267701e-06, | |
| "loss": 1.181, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.392931392931393, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 1.11420800085927e-06, | |
| "loss": 1.1764, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.4033264033264032, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 1.1033720171364108e-06, | |
| "loss": 1.1792, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.4137214137214138, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 1.092523750242977e-06, | |
| "loss": 1.1784, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4241164241164241, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 1.0816644892232997e-06, | |
| "loss": 1.1855, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.4345114345114345, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.070795524428086e-06, | |
| "loss": 1.1782, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.444906444906445, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.0599181473610938e-06, | |
| "loss": 1.1837, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.4553014553014554, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.049033650525668e-06, | |
| "loss": 1.1786, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.4656964656964657, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.0381433272711585e-06, | |
| "loss": 1.1747, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.476091476091476, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 1.0272484716392408e-06, | |
| "loss": 1.1854, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.4864864864864864, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.0163503782101484e-06, | |
| "loss": 1.1755, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.496881496881497, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.0054503419488454e-06, | |
| "loss": 1.1795, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.5072765072765073, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 9.945496580511543e-07, | |
| "loss": 1.1846, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.5176715176715176, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 9.836496217898518e-07, | |
| "loss": 1.1806, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5280665280665282, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 9.72751528360759e-07, | |
| "loss": 1.1667, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 9.618566727288414e-07, | |
| "loss": 1.185, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5488565488565489, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 9.509663494743321e-07, | |
| "loss": 1.191, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.5592515592515592, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 9.400818526389062e-07, | |
| "loss": 1.1769, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.5696465696465696, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 9.292044755719138e-07, | |
| "loss": 1.1741, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.5800415800415801, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 9.183355107767003e-07, | |
| "loss": 1.1771, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.5904365904365905, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 9.07476249757023e-07, | |
| "loss": 1.1725, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.6008316008316008, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 8.966279828635894e-07, | |
| "loss": 1.1801, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.6112266112266114, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 8.8579199914073e-07, | |
| "loss": 1.1697, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.6216216216216215, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 8.749695861732299e-07, | |
| "loss": 1.175, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.632016632016632, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 8.641620299333295e-07, | |
| "loss": 1.1818, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.6424116424116424, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 8.533706146279207e-07, | |
| "loss": 1.1697, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.6528066528066527, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 8.425966225459493e-07, | |
| "loss": 1.1702, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.6632016632016633, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 8.318413339060495e-07, | |
| "loss": 1.1793, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.6735966735966736, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 8.21106026704419e-07, | |
| "loss": 1.1801, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.683991683991684, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 8.103919765629645e-07, | |
| "loss": 1.1797, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.6943866943866945, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 7.997004565777233e-07, | |
| "loss": 1.1659, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.7047817047817047, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 7.890327371675895e-07, | |
| "loss": 1.1743, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7151767151767152, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 7.783900859233562e-07, | |
| "loss": 1.1849, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.7255717255717256, | |
| "grad_norm": 3.625, | |
| "learning_rate": 7.677737674570936e-07, | |
| "loss": 1.1791, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.735966735966736, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 7.571850432518819e-07, | |
| "loss": 1.1789, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.7463617463617465, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 7.466251715119156e-07, | |
| "loss": 1.1773, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.7567567567567568, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 7.360954070129981e-07, | |
| "loss": 1.175, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.7671517671517671, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 7.255970009534425e-07, | |
| "loss": 1.1816, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.7775467775467777, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 7.151312008053979e-07, | |
| "loss": 1.1848, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.7879417879417878, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 7.046992501666195e-07, | |
| "loss": 1.175, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.7983367983367984, | |
| "grad_norm": 3.625, | |
| "learning_rate": 6.943023886126965e-07, | |
| "loss": 1.1753, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.8087318087318087, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 6.839418515497618e-07, | |
| "loss": 1.1718, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.819126819126819, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 6.736188700676935e-07, | |
| "loss": 1.164, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.8295218295218296, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 6.633346707938319e-07, | |
| "loss": 1.1761, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.83991683991684, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 6.530904757472236e-07, | |
| "loss": 1.1869, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.8503118503118503, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 6.42887502193418e-07, | |
| "loss": 1.1836, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.8607068607068609, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 6.327269624998227e-07, | |
| "loss": 1.1699, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.871101871101871, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 6.226100639916474e-07, | |
| "loss": 1.1743, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.871101871101871, | |
| "eval_loss": 1.1942965984344482, | |
| "eval_runtime": 10.786, | |
| "eval_samples_per_second": 86.223, | |
| "eval_steps_per_second": 2.781, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8814968814968815, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 6.125380088084408e-07, | |
| "loss": 1.1797, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.8918918918918919, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 6.025119937612481e-07, | |
| "loss": 1.1758, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.9022869022869022, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 5.925332101903994e-07, | |
| "loss": 1.1783, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.9126819126819128, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 5.826028438239479e-07, | |
| "loss": 1.1763, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 5.727220746367791e-07, | |
| "loss": 1.18, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.9334719334719335, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 5.628920767103957e-07, | |
| "loss": 1.1782, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.943866943866944, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 5.531140180934119e-07, | |
| "loss": 1.1772, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.9542619542619541, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 5.433890606627568e-07, | |
| "loss": 1.175, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.9646569646569647, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 5.337183599856164e-07, | |
| "loss": 1.1745, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.975051975051975, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 5.241030651821231e-07, | |
| "loss": 1.1662, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.9854469854469854, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 5.145443187888114e-07, | |
| "loss": 1.171, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.995841995841996, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 5.050432566228552e-07, | |
| "loss": 1.1831, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.006237006237006, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 4.956010076471065e-07, | |
| "loss": 1.1592, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.0166320166320166, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 4.862186938359441e-07, | |
| "loss": 1.1667, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.027027027027027, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 4.768974300419573e-07, | |
| "loss": 1.1678, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.0374220374220373, | |
| "grad_norm": 3.625, | |
| "learning_rate": 4.6763832386347214e-07, | |
| "loss": 1.1723, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.047817047817048, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 4.5844247551294224e-07, | |
| "loss": 1.1642, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.0582120582120584, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 4.493109776862143e-07, | |
| "loss": 1.17, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.0686070686070686, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 4.402449154326913e-07, | |
| "loss": 1.1651, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.079002079002079, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 4.312453660263987e-07, | |
| "loss": 1.1783, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0893970893970892, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 4.2231339883798025e-07, | |
| "loss": 1.1711, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.0997920997921, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 4.13450075207628e-07, | |
| "loss": 1.163, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.1101871101871104, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 4.0465644831897006e-07, | |
| "loss": 1.17, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.1205821205821205, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 3.9593356307392436e-07, | |
| "loss": 1.1733, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.130977130977131, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 3.872824559685409e-07, | |
| "loss": 1.1762, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.141372141372141, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 3.7870415496983743e-07, | |
| "loss": 1.1734, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.1517671517671517, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 3.701996793936535e-07, | |
| "loss": 1.1724, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.1621621621621623, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 3.6177003978352917e-07, | |
| "loss": 1.1718, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.1725571725571724, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 3.5341623779062813e-07, | |
| "loss": 1.1688, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.182952182952183, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 3.45139266054715e-07, | |
| "loss": 1.1732, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.1933471933471935, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 3.3694010808620733e-07, | |
| "loss": 1.1619, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.2037422037422036, | |
| "grad_norm": 3.625, | |
| "learning_rate": 3.288197381493075e-07, | |
| "loss": 1.1673, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.214137214137214, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 3.207791211462383e-07, | |
| "loss": 1.1725, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.2245322245322248, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 3.128192125025869e-07, | |
| "loss": 1.1673, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.234927234927235, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 3.049409580537773e-07, | |
| "loss": 1.1735, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.2453222453222454, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 2.9714529393268016e-07, | |
| "loss": 1.1583, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.2557172557172556, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 2.8943314645837955e-07, | |
| "loss": 1.1715, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.266112266112266, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 2.8180543202609984e-07, | |
| "loss": 1.164, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.2765072765072767, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 2.742630569983182e-07, | |
| "loss": 1.1695, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.286902286902287, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 2.66806917597064e-07, | |
| "loss": 1.169, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.2972972972972974, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 2.594378997974267e-07, | |
| "loss": 1.1615, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 2.5215687922227845e-07, | |
| "loss": 1.1712, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.318087318087318, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 2.4496472103823027e-07, | |
| "loss": 1.1688, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.3284823284823286, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 2.378622798528266e-07, | |
| "loss": 1.1631, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.3388773388773387, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 2.3085039961299814e-07, | |
| "loss": 1.1671, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.3492723492723493, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 2.239299135047794e-07, | |
| "loss": 1.1623, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.35966735966736, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 2.1710164385430585e-07, | |
| "loss": 1.1716, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.37006237006237, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 2.103664020300997e-07, | |
| "loss": 1.1674, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.3804573804573805, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 2.037249883466614e-07, | |
| "loss": 1.1623, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.390852390852391, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 1.971781919693697e-07, | |
| "loss": 1.1808, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.401247401247401, | |
| "grad_norm": 3.625, | |
| "learning_rate": 1.9072679082071163e-07, | |
| "loss": 1.169, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.4116424116424118, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 1.8437155148784433e-07, | |
| "loss": 1.1717, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.422037422037422, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.781132291315064e-07, | |
| "loss": 1.1706, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.4324324324324325, | |
| "grad_norm": 3.625, | |
| "learning_rate": 1.7195256739628439e-07, | |
| "loss": 1.1722, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.442827442827443, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 1.6589029832225155e-07, | |
| "loss": 1.1615, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.453222453222453, | |
| "grad_norm": 3.625, | |
| "learning_rate": 1.599271422579812e-07, | |
| "loss": 1.1691, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.4636174636174637, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.5406380777495297e-07, | |
| "loss": 1.1647, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.474012474012474, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.4830099158335563e-07, | |
| "loss": 1.1707, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.4844074844074844, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.426393784493015e-07, | |
| "loss": 1.1564, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.494802494802495, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.3707964111345805e-07, | |
| "loss": 1.1721, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.494802494802495, | |
| "eval_loss": 1.1930803060531616, | |
| "eval_runtime": 10.8355, | |
| "eval_samples_per_second": 85.829, | |
| "eval_steps_per_second": 2.769, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.505197505197505, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 1.3162244021111123e-07, | |
| "loss": 1.1677, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.5155925155925156, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.2626842419366369e-07, | |
| "loss": 1.1551, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.525987525987526, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.2101822925158378e-07, | |
| "loss": 1.1678, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.5363825363825363, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 1.1587247923881016e-07, | |
| "loss": 1.1731, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.546777546777547, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 1.1083178559862227e-07, | |
| "loss": 1.1707, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.5571725571725574, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 1.0589674729098507e-07, | |
| "loss": 1.1733, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.5675675675675675, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.0106795072137896e-07, | |
| "loss": 1.1741, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.577962577962578, | |
| "grad_norm": 3.625, | |
| "learning_rate": 9.634596967111853e-08, | |
| "loss": 1.1704, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.5883575883575882, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 9.173136522917457e-08, | |
| "loss": 1.1679, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.598752598752599, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 8.722468572550213e-08, | |
| "loss": 1.1682, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.609147609147609, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 8.28264666658851e-08, | |
| "loss": 1.1653, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.6195426195426195, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 7.853723066830486e-08, | |
| "loss": 1.1672, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.62993762993763, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 7.435748740084046e-08, | |
| "loss": 1.1606, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.64033264033264, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 7.028773352110684e-08, | |
| "loss": 1.1634, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.6507276507276507, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 6.632845261724051e-08, | |
| "loss": 1.1635, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.6611226611226613, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 6.248011515043617e-08, | |
| "loss": 1.1641, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.6715176715176714, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 5.8743178399044966e-08, | |
| "loss": 1.1642, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.681912681912682, | |
| "grad_norm": 3.625, | |
| "learning_rate": 5.511808640423765e-08, | |
| "loss": 1.1727, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 5.160526991724246e-08, | |
| "loss": 1.1732, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.7027027027027026, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 4.8205146348160195e-08, | |
| "loss": 1.1699, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.713097713097713, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 4.491811971636605e-08, | |
| "loss": 1.166, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.7234927234927238, | |
| "grad_norm": 3.625, | |
| "learning_rate": 4.174458060250208e-08, | |
| "loss": 1.1712, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.733887733887734, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 3.868490610206565e-08, | |
| "loss": 1.1595, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.7442827442827444, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 3.5739459780602665e-08, | |
| "loss": 1.1684, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.7546777546777546, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 3.290859163050508e-08, | |
| "loss": 1.1744, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.765072765072765, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 3.0192638029424735e-08, | |
| "loss": 1.1664, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.7754677754677752, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 2.7591921700302222e-08, | |
| "loss": 1.1612, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.785862785862786, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 2.5106751673020012e-08, | |
| "loss": 1.174, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.7962577962577964, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 2.273742324768124e-08, | |
| "loss": 1.1602, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.8066528066528065, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 2.048421795952171e-08, | |
| "loss": 1.1708, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.817047817047817, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.8347403545455497e-08, | |
| "loss": 1.1622, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.8274428274428276, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.6327233912261984e-08, | |
| "loss": 1.1668, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.8378378378378377, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 1.4423949106414868e-08, | |
| "loss": 1.1708, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.8482328482328483, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.2637775285558983e-08, | |
| "loss": 1.1663, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.858627858627859, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 1.0968924691636572e-08, | |
| "loss": 1.1621, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.869022869022869, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 9.417595625668462e-09, | |
| "loss": 1.1769, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.8794178794178795, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 7.983972424190354e-09, | |
| "loss": 1.1784, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.88981288981289, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 6.668225437349351e-09, | |
| "loss": 1.1734, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.9002079002079, | |
| "grad_norm": 3.75, | |
| "learning_rate": 5.470511008662026e-09, | |
| "loss": 1.1747, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.9106029106029108, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 4.390971456437076e-09, | |
| "loss": 1.171, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.920997920997921, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 3.429735056863725e-09, | |
| "loss": 1.1667, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 2.9313929313929314, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 2.5869160287702586e-09, | |
| "loss": 1.1683, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.9417879417879416, | |
| "grad_norm": 3.625, | |
| "learning_rate": 1.8626145200513199e-09, | |
| "loss": 1.1702, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 2.952182952182952, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.2569165957680983e-09, | |
| "loss": 1.1712, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.9625779625779627, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 7.698942279216192e-10, | |
| "loss": 1.1725, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 2.972972972972973, | |
| "grad_norm": 3.5, | |
| "learning_rate": 4.016052869005859e-10, | |
| "loss": 1.1693, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.9833679833679834, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 1.520935346051022e-10, | |
| "loss": 1.1682, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 2.993762993762994, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 2.1388619246498486e-11, | |
| "loss": 1.1627, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1443, | |
| "total_flos": 8.072899219804914e+18, | |
| "train_loss": 1.2002355462548142, | |
| "train_runtime": 6547.9575, | |
| "train_samples_per_second": 14.077, | |
| "train_steps_per_second": 0.22 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1443, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.072899219804914e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |