| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 870, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017241379310344827, | |
| "grad_norm": 9.331307328744412, | |
| "learning_rate": 5.747126436781609e-07, | |
| "loss": 1.2873, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.034482758620689655, | |
| "grad_norm": 5.0843555710900326, | |
| "learning_rate": 1.1494252873563219e-06, | |
| "loss": 1.3464, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05172413793103448, | |
| "grad_norm": 3.2704303545266757, | |
| "learning_rate": 1.724137931034483e-06, | |
| "loss": 1.2482, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06896551724137931, | |
| "grad_norm": 3.4622234413756647, | |
| "learning_rate": 2.2988505747126437e-06, | |
| "loss": 1.2549, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08620689655172414, | |
| "grad_norm": 2.903152829120139, | |
| "learning_rate": 2.8735632183908046e-06, | |
| "loss": 1.1964, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10344827586206896, | |
| "grad_norm": 3.054985905010679, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 1.1774, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1206896551724138, | |
| "grad_norm": 2.4975716112394486, | |
| "learning_rate": 4.022988505747127e-06, | |
| "loss": 1.1393, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 2.959802442307543, | |
| "learning_rate": 4.5977011494252875e-06, | |
| "loss": 1.1998, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15517241379310345, | |
| "grad_norm": 3.27289191086378, | |
| "learning_rate": 5.172413793103449e-06, | |
| "loss": 1.205, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 3.07678334771907, | |
| "learning_rate": 5.747126436781609e-06, | |
| "loss": 1.1961, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1896551724137931, | |
| "grad_norm": 4.039994080396709, | |
| "learning_rate": 6.321839080459771e-06, | |
| "loss": 1.1906, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.20689655172413793, | |
| "grad_norm": 3.0137562214933613, | |
| "learning_rate": 6.896551724137932e-06, | |
| "loss": 1.1844, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22413793103448276, | |
| "grad_norm": 2.845134487512356, | |
| "learning_rate": 7.4712643678160925e-06, | |
| "loss": 1.1145, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2413793103448276, | |
| "grad_norm": 2.5241979501478036, | |
| "learning_rate": 8.045977011494253e-06, | |
| "loss": 1.1685, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25862068965517243, | |
| "grad_norm": 2.722719328804337, | |
| "learning_rate": 8.620689655172414e-06, | |
| "loss": 1.2057, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 2.396157449528953, | |
| "learning_rate": 9.195402298850575e-06, | |
| "loss": 1.1636, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.29310344827586204, | |
| "grad_norm": 2.7098204318960986, | |
| "learning_rate": 9.770114942528738e-06, | |
| "loss": 1.1612, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3103448275862069, | |
| "grad_norm": 2.2895872071801495, | |
| "learning_rate": 9.999637795788383e-06, | |
| "loss": 1.1645, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3275862068965517, | |
| "grad_norm": 2.6598445818580263, | |
| "learning_rate": 9.997424515642709e-06, | |
| "loss": 1.1547, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 2.102393309002244, | |
| "learning_rate": 9.993200069547117e-06, | |
| "loss": 1.1738, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3620689655172414, | |
| "grad_norm": 2.6915767876023913, | |
| "learning_rate": 9.986966157589751e-06, | |
| "loss": 1.1797, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3793103448275862, | |
| "grad_norm": 2.4200791763638647, | |
| "learning_rate": 9.978725288549161e-06, | |
| "loss": 1.2318, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.39655172413793105, | |
| "grad_norm": 2.5921861155145307, | |
| "learning_rate": 9.968480778884693e-06, | |
| "loss": 1.205, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.41379310344827586, | |
| "grad_norm": 3.7006769047874486, | |
| "learning_rate": 9.95623675140179e-06, | |
| "loss": 1.2033, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.43103448275862066, | |
| "grad_norm": 2.2257771592614737, | |
| "learning_rate": 9.941998133592825e-06, | |
| "loss": 1.1718, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4482758620689655, | |
| "grad_norm": 2.7692105786463483, | |
| "learning_rate": 9.925770655654061e-06, | |
| "loss": 1.1035, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.46551724137931033, | |
| "grad_norm": 2.9856964243297894, | |
| "learning_rate": 9.907560848179607e-06, | |
| "loss": 1.1374, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4827586206896552, | |
| "grad_norm": 2.3833997110519785, | |
| "learning_rate": 9.887376039533227e-06, | |
| "loss": 1.2123, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.441770609627606, | |
| "learning_rate": 9.86522435289912e-06, | |
| "loss": 1.2104, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 2.647051507442294, | |
| "learning_rate": 9.841114703012817e-06, | |
| "loss": 1.1937, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5344827586206896, | |
| "grad_norm": 2.7754614517017533, | |
| "learning_rate": 9.815056792573531e-06, | |
| "loss": 1.2122, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 2.691549218743732, | |
| "learning_rate": 9.7870611083394e-06, | |
| "loss": 1.2088, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5689655172413793, | |
| "grad_norm": 2.3179060468120296, | |
| "learning_rate": 9.757138916907184e-06, | |
| "loss": 1.1936, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5862068965517241, | |
| "grad_norm": 2.4542264174801143, | |
| "learning_rate": 9.725302260178145e-06, | |
| "loss": 1.1571, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.603448275862069, | |
| "grad_norm": 2.341677736854312, | |
| "learning_rate": 9.69156395051188e-06, | |
| "loss": 1.1808, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6206896551724138, | |
| "grad_norm": 2.3508234314371794, | |
| "learning_rate": 9.655937565570124e-06, | |
| "loss": 1.1737, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6379310344827587, | |
| "grad_norm": 2.5885491423781937, | |
| "learning_rate": 9.618437442852539e-06, | |
| "loss": 1.1372, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6551724137931034, | |
| "grad_norm": 2.512485093543423, | |
| "learning_rate": 9.579078673926729e-06, | |
| "loss": 1.188, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6724137931034483, | |
| "grad_norm": 2.230805472437592, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 1.1737, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 2.362407181996345, | |
| "learning_rate": 9.494849297318795e-06, | |
| "loss": 1.1371, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7068965517241379, | |
| "grad_norm": 2.4353970458455634, | |
| "learning_rate": 9.450012586947912e-06, | |
| "loss": 1.1921, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7241379310344828, | |
| "grad_norm": 2.222742138433333, | |
| "learning_rate": 9.40338501134964e-06, | |
| "loss": 1.1577, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7413793103448276, | |
| "grad_norm": 2.6747288559282074, | |
| "learning_rate": 9.354985335348155e-06, | |
| "loss": 1.0958, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7586206896551724, | |
| "grad_norm": 2.5845221311863606, | |
| "learning_rate": 9.30483303693258e-06, | |
| "loss": 1.1759, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7758620689655172, | |
| "grad_norm": 2.65063728191238, | |
| "learning_rate": 9.252948299418255e-06, | |
| "loss": 1.187, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7931034482758621, | |
| "grad_norm": 2.546015283230143, | |
| "learning_rate": 9.199352003324151e-06, | |
| "loss": 1.1741, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.8103448275862069, | |
| "grad_norm": 2.6278320735083227, | |
| "learning_rate": 9.144065717969707e-06, | |
| "loss": 1.14, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 2.6907971238529407, | |
| "learning_rate": 9.08711169279446e-06, | |
| "loss": 1.2098, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8448275862068966, | |
| "grad_norm": 2.6549220831544607, | |
| "learning_rate": 9.028512848403971e-06, | |
| "loss": 1.1585, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 3.9743418742434775, | |
| "learning_rate": 8.968292767345646e-06, | |
| "loss": 1.199, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8793103448275862, | |
| "grad_norm": 2.3215747943265823, | |
| "learning_rate": 8.90647568461816e-06, | |
| "loss": 1.1734, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.896551724137931, | |
| "grad_norm": 2.076014885969078, | |
| "learning_rate": 8.843086477918317e-06, | |
| "loss": 1.1296, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9137931034482759, | |
| "grad_norm": 2.1514269175918312, | |
| "learning_rate": 8.778150657629258e-06, | |
| "loss": 1.1756, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.9310344827586207, | |
| "grad_norm": 2.225293146269018, | |
| "learning_rate": 8.71169435655405e-06, | |
| "loss": 1.159, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9482758620689655, | |
| "grad_norm": 2.2545083946242905, | |
| "learning_rate": 8.643744319398781e-06, | |
| "loss": 1.1451, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9655172413793104, | |
| "grad_norm": 2.5425358625334824, | |
| "learning_rate": 8.574327892009415e-06, | |
| "loss": 1.1701, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9827586206896551, | |
| "grad_norm": 2.292596274404439, | |
| "learning_rate": 8.503473010366713e-06, | |
| "loss": 1.175, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.7932559109425275, | |
| "learning_rate": 8.43120818934367e-06, | |
| "loss": 1.1198, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.0172413793103448, | |
| "grad_norm": 2.154259952280766, | |
| "learning_rate": 8.357562511229961e-06, | |
| "loss": 0.7509, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.0344827586206897, | |
| "grad_norm": 2.9402880338561617, | |
| "learning_rate": 8.282565614028068e-06, | |
| "loss": 0.6958, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0517241379310345, | |
| "grad_norm": 2.292404675437898, | |
| "learning_rate": 8.206247679525736e-06, | |
| "loss": 0.6641, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.0689655172413792, | |
| "grad_norm": 2.4068253005632836, | |
| "learning_rate": 8.1286394211496e-06, | |
| "loss": 0.6752, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0862068965517242, | |
| "grad_norm": 2.5678671148349395, | |
| "learning_rate": 8.049772071604864e-06, | |
| "loss": 0.6926, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.103448275862069, | |
| "grad_norm": 2.290285133604941, | |
| "learning_rate": 7.969677370306e-06, | |
| "loss": 0.6654, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.1206896551724137, | |
| "grad_norm": 2.8990212853626725, | |
| "learning_rate": 7.888387550603505e-06, | |
| "loss": 0.6767, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.1379310344827587, | |
| "grad_norm": 2.3313227761853272, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.6968, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1551724137931034, | |
| "grad_norm": 2.3456535599013835, | |
| "learning_rate": 7.722353881044223e-06, | |
| "loss": 0.6625, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.1724137931034484, | |
| "grad_norm": 2.221283082287151, | |
| "learning_rate": 7.637676849858077e-06, | |
| "loss": 0.7104, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1896551724137931, | |
| "grad_norm": 2.29086724469859, | |
| "learning_rate": 7.551938310719043e-06, | |
| "loss": 0.6944, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.206896551724138, | |
| "grad_norm": 2.127721775862383, | |
| "learning_rate": 7.465172768286463e-06, | |
| "loss": 0.6516, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.2241379310344827, | |
| "grad_norm": 2.433252747676718, | |
| "learning_rate": 7.377415140527388e-06, | |
| "loss": 0.6446, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.2413793103448276, | |
| "grad_norm": 2.222940961609735, | |
| "learning_rate": 7.288700744664167e-06, | |
| "loss": 0.679, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.2586206896551724, | |
| "grad_norm": 2.2460931536316697, | |
| "learning_rate": 7.199065282961372e-06, | |
| "loss": 0.7054, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.2758620689655173, | |
| "grad_norm": 2.1519776354853444, | |
| "learning_rate": 7.1085448283577556e-06, | |
| "loss": 0.6966, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.293103448275862, | |
| "grad_norm": 2.460141861751599, | |
| "learning_rate": 7.017175809949044e-06, | |
| "loss": 0.678, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.3103448275862069, | |
| "grad_norm": 2.6814215239672308, | |
| "learning_rate": 6.924994998327395e-06, | |
| "loss": 0.6925, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.3275862068965516, | |
| "grad_norm": 2.6153556243725533, | |
| "learning_rate": 6.832039490783422e-06, | |
| "loss": 0.7251, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.3448275862068966, | |
| "grad_norm": 2.0448430816189362, | |
| "learning_rate": 6.7383466963767386e-06, | |
| "loss": 0.6822, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3620689655172413, | |
| "grad_norm": 2.2459257516184636, | |
| "learning_rate": 6.643954320881045e-06, | |
| "loss": 0.6862, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 2.9435945925232176, | |
| "learning_rate": 6.548900351609794e-06, | |
| "loss": 0.6842, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.396551724137931, | |
| "grad_norm": 2.053653397494253, | |
| "learning_rate": 6.453223042128556e-06, | |
| "loss": 0.6621, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.4137931034482758, | |
| "grad_norm": 2.478369477656992, | |
| "learning_rate": 6.3569608968602415e-06, | |
| "loss": 0.6399, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.4310344827586206, | |
| "grad_norm": 2.3692300011890963, | |
| "learning_rate": 6.260152655589358e-06, | |
| "loss": 0.6625, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.4482758620689655, | |
| "grad_norm": 2.6561080123885787, | |
| "learning_rate": 6.162837277871553e-06, | |
| "loss": 0.7111, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4655172413793103, | |
| "grad_norm": 2.2662300356037712, | |
| "learning_rate": 6.0650539273547145e-06, | |
| "loss": 0.6805, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.4827586206896552, | |
| "grad_norm": 2.3923914621173976, | |
| "learning_rate": 5.966841956017928e-06, | |
| "loss": 0.7073, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.2024157308125827, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.6719, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.5172413793103448, | |
| "grad_norm": 2.485013221721386, | |
| "learning_rate": 5.769290405366469e-06, | |
| "loss": 0.7115, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.5344827586206895, | |
| "grad_norm": 2.266211971536318, | |
| "learning_rate": 5.670030328793812e-06, | |
| "loss": 0.6679, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.5517241379310345, | |
| "grad_norm": 2.540019685554583, | |
| "learning_rate": 5.570500604890124e-06, | |
| "loss": 0.6434, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.5689655172413794, | |
| "grad_norm": 2.1889073162299524, | |
| "learning_rate": 5.470741288445844e-06, | |
| "loss": 0.6791, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.5862068965517242, | |
| "grad_norm": 2.072082472571136, | |
| "learning_rate": 5.370792526648747e-06, | |
| "loss": 0.6392, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.603448275862069, | |
| "grad_norm": 2.597868449195596, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.6675, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.6206896551724137, | |
| "grad_norm": 2.1226456555287125, | |
| "learning_rate": 5.170487620762066e-06, | |
| "loss": 0.6485, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.6379310344827587, | |
| "grad_norm": 2.1087011375964275, | |
| "learning_rate": 5.070212087476116e-06, | |
| "loss": 0.734, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.6551724137931034, | |
| "grad_norm": 2.3428994556890523, | |
| "learning_rate": 4.9699082980035735e-06, | |
| "loss": 0.696, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.6724137931034484, | |
| "grad_norm": 2.1819328061566243, | |
| "learning_rate": 4.869616618650201e-06, | |
| "loss": 0.6611, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.6896551724137931, | |
| "grad_norm": 2.355181731987328, | |
| "learning_rate": 4.769377410848162e-06, | |
| "loss": 0.6828, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.706896551724138, | |
| "grad_norm": 2.26524912434979, | |
| "learning_rate": 4.6692310149129425e-06, | |
| "loss": 0.7233, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 2.5499608122876882, | |
| "learning_rate": 4.569217733808774e-06, | |
| "loss": 0.6543, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.7413793103448276, | |
| "grad_norm": 2.0257734264386764, | |
| "learning_rate": 4.4693778169290934e-06, | |
| "loss": 0.6763, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.7586206896551724, | |
| "grad_norm": 2.307245881012447, | |
| "learning_rate": 4.369751443898554e-06, | |
| "loss": 0.6468, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.7758620689655173, | |
| "grad_norm": 2.5757186999798347, | |
| "learning_rate": 4.2703787084031175e-06, | |
| "loss": 0.6951, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.793103448275862, | |
| "grad_norm": 2.1142543692594513, | |
| "learning_rate": 4.171299602054736e-06, | |
| "loss": 0.651, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.8103448275862069, | |
| "grad_norm": 2.4680914278700827, | |
| "learning_rate": 4.072553998297103e-06, | |
| "loss": 0.6584, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.8275862068965516, | |
| "grad_norm": 2.1232313646170557, | |
| "learning_rate": 3.974181636358963e-06, | |
| "loss": 0.6862, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.8448275862068966, | |
| "grad_norm": 2.4782758201734665, | |
| "learning_rate": 3.87622210526145e-06, | |
| "loss": 0.6821, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.8620689655172413, | |
| "grad_norm": 2.0024977625075207, | |
| "learning_rate": 3.7787148278858453e-06, | |
| "loss": 0.6854, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.8793103448275863, | |
| "grad_norm": 2.4881786144910247, | |
| "learning_rate": 3.6816990451082297e-06, | |
| "loss": 0.6488, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.896551724137931, | |
| "grad_norm": 2.501852490981883, | |
| "learning_rate": 3.5852138000073566e-06, | |
| "loss": 0.7261, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.9137931034482758, | |
| "grad_norm": 2.3190367462390613, | |
| "learning_rate": 3.489297922152136e-06, | |
| "loss": 0.6829, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.9310344827586206, | |
| "grad_norm": 2.1282123456653923, | |
| "learning_rate": 3.3939900119750543e-06, | |
| "loss": 0.6567, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.9482758620689655, | |
| "grad_norm": 2.1578335106758515, | |
| "learning_rate": 3.299328425237781e-06, | |
| "loss": 0.6965, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.9655172413793105, | |
| "grad_norm": 2.3185722355182716, | |
| "learning_rate": 3.205351257595272e-06, | |
| "loss": 0.6066, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.9827586206896552, | |
| "grad_norm": 2.387747701227867, | |
| "learning_rate": 3.11209632926453e-06, | |
| "loss": 0.6526, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.9269580371676338, | |
| "learning_rate": 3.019601169804216e-06, | |
| "loss": 0.6352, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.0172413793103448, | |
| "grad_norm": 2.0236058838383397, | |
| "learning_rate": 2.927903003011241e-06, | |
| "loss": 0.321, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.0344827586206895, | |
| "grad_norm": 2.2133504635385095, | |
| "learning_rate": 2.837038731940397e-06, | |
| "loss": 0.2667, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.0517241379310347, | |
| "grad_norm": 2.413464630013685, | |
| "learning_rate": 2.7470449240530785e-06, | |
| "loss": 0.2632, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.0689655172413794, | |
| "grad_norm": 2.458598609118027, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.288, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.086206896551724, | |
| "grad_norm": 2.053211229266344, | |
| "learning_rate": 2.569813201551205e-06, | |
| "loss": 0.2973, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.103448275862069, | |
| "grad_norm": 1.770644646455138, | |
| "learning_rate": 2.4826466121571575e-06, | |
| "loss": 0.2813, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.1206896551724137, | |
| "grad_norm": 2.489693320306748, | |
| "learning_rate": 2.396493107683488e-06, | |
| "loss": 0.265, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.1379310344827585, | |
| "grad_norm": 2.249234016593084, | |
| "learning_rate": 2.311387359788395e-06, | |
| "loss": 0.258, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.1551724137931036, | |
| "grad_norm": 2.0866027899593633, | |
| "learning_rate": 2.2273636184704074e-06, | |
| "loss": 0.2817, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.1724137931034484, | |
| "grad_norm": 2.2562342362257737, | |
| "learning_rate": 2.1444556982847996e-06, | |
| "loss": 0.2595, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.189655172413793, | |
| "grad_norm": 2.04363251716917, | |
| "learning_rate": 2.06269696473525e-06, | |
| "loss": 0.2805, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.206896551724138, | |
| "grad_norm": 1.9682620834990578, | |
| "learning_rate": 1.982120320846208e-06, | |
| "loss": 0.2689, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.2241379310344827, | |
| "grad_norm": 2.125252810092142, | |
| "learning_rate": 1.9027581939213852e-06, | |
| "loss": 0.2696, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.2413793103448274, | |
| "grad_norm": 2.0257458044704197, | |
| "learning_rate": 1.8246425224936986e-06, | |
| "loss": 0.2727, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.2586206896551726, | |
| "grad_norm": 2.2847004577300734, | |
| "learning_rate": 1.747804743471907e-06, | |
| "loss": 0.2664, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.2758620689655173, | |
| "grad_norm": 4.622508078600453, | |
| "learning_rate": 1.6722757794891287e-06, | |
| "loss": 0.2969, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.293103448275862, | |
| "grad_norm": 2.300647420998415, | |
| "learning_rate": 1.598086026458322e-06, | |
| "loss": 0.2547, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.310344827586207, | |
| "grad_norm": 2.5311183593832105, | |
| "learning_rate": 1.52526534133974e-06, | |
| "loss": 0.3067, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.3275862068965516, | |
| "grad_norm": 2.23412202590258, | |
| "learning_rate": 1.4538430301252783e-06, | |
| "loss": 0.2798, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.344827586206897, | |
| "grad_norm": 2.49448795601641, | |
| "learning_rate": 1.3838478360445616e-06, | |
| "loss": 0.2911, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.3620689655172415, | |
| "grad_norm": 1.9660989748507585, | |
| "learning_rate": 1.3153079279975011e-06, | |
| "loss": 0.2916, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.3793103448275863, | |
| "grad_norm": 2.4364982240422797, | |
| "learning_rate": 1.2482508892179884e-06, | |
| "loss": 0.2583, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.396551724137931, | |
| "grad_norm": 2.2114322402406548, | |
| "learning_rate": 1.1827037061732877e-06, | |
| "loss": 0.2769, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.413793103448276, | |
| "grad_norm": 2.1754962055440807, | |
| "learning_rate": 1.1186927577035867e-06, | |
| "loss": 0.2567, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.4310344827586206, | |
| "grad_norm": 2.13797171831918, | |
| "learning_rate": 1.0562438044060846e-06, | |
| "loss": 0.2668, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.4482758620689653, | |
| "grad_norm": 3.0013612019154463, | |
| "learning_rate": 9.953819782678887e-07, | |
| "loss": 0.256, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.4655172413793105, | |
| "grad_norm": 2.3878327191424935, | |
| "learning_rate": 9.361317725518749e-07, | |
| "loss": 0.3042, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.4827586206896552, | |
| "grad_norm": 2.1197648068461814, | |
| "learning_rate": 8.785170319396174e-07, | |
| "loss": 0.2751, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.383204062872722, | |
| "learning_rate": 8.225609429353187e-07, | |
| "loss": 0.2849, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.5172413793103448, | |
| "grad_norm": 2.4136436239961165, | |
| "learning_rate": 7.682860245346213e-07, | |
| "loss": 0.2763, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.5344827586206895, | |
| "grad_norm": 2.102876793971456, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.2737, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.5517241379310347, | |
| "grad_norm": 2.1010828551207497, | |
| "learning_rate": 6.648663838807562e-07, | |
| "loss": 0.2636, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.5689655172413794, | |
| "grad_norm": 2.1468501823656365, | |
| "learning_rate": 6.157632818780179e-07, | |
| "loss": 0.2942, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.586206896551724, | |
| "grad_norm": 2.123594278339021, | |
| "learning_rate": 5.684245742300625e-07, | |
| "loss": 0.2632, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.603448275862069, | |
| "grad_norm": 2.220781323028423, | |
| "learning_rate": 5.228693119493955e-07, | |
| "loss": 0.2742, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.6206896551724137, | |
| "grad_norm": 2.0591574448891152, | |
| "learning_rate": 4.791158283178999e-07, | |
| "loss": 0.2598, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.637931034482759, | |
| "grad_norm": 2.747851806203062, | |
| "learning_rate": 4.371817315087845e-07, | |
| "loss": 0.2476, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.655172413793103, | |
| "grad_norm": 2.5857626546141974, | |
| "learning_rate": 3.9708389750034105e-07, | |
| "loss": 0.2914, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.6724137931034484, | |
| "grad_norm": 2.3230337862508534, | |
| "learning_rate": 3.5883846328436943e-07, | |
| "loss": 0.2648, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.689655172413793, | |
| "grad_norm": 2.324111936612965, | |
| "learning_rate": 3.224608203719953e-07, | |
| "loss": 0.2959, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.706896551724138, | |
| "grad_norm": 2.2596124717885893, | |
| "learning_rate": 2.879656085995042e-07, | |
| "loss": 0.2852, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.7241379310344827, | |
| "grad_norm": 2.0490467708210516, | |
| "learning_rate": 2.5536671023668305e-07, | |
| "loss": 0.2783, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.7413793103448274, | |
| "grad_norm": 1.9570990731879616, | |
| "learning_rate": 2.2467724440002336e-07, | |
| "loss": 0.2732, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 2.7586206896551726, | |
| "grad_norm": 2.0413692175737257, | |
| "learning_rate": 1.9590956177306665e-07, | |
| "loss": 0.2724, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.7758620689655173, | |
| "grad_norm": 1.8263133287669178, | |
| "learning_rate": 1.690752396359857e-07, | |
| "loss": 0.2992, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 2.793103448275862, | |
| "grad_norm": 2.0310974537395623, | |
| "learning_rate": 1.4418507720641794e-07, | |
| "loss": 0.2684, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.810344827586207, | |
| "grad_norm": 2.102811024193508, | |
| "learning_rate": 1.2124909129342332e-07, | |
| "loss": 0.2835, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 2.8275862068965516, | |
| "grad_norm": 1.9121801119342077, | |
| "learning_rate": 1.0027651226631463e-07, | |
| "loss": 0.2831, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.844827586206897, | |
| "grad_norm": 2.021176166908057, | |
| "learning_rate": 8.127578033998663e-08, | |
| "loss": 0.2958, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.862068965517241, | |
| "grad_norm": 2.0559845622615414, | |
| "learning_rate": 6.425454217822425e-08, | |
| "loss": 0.2929, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.8793103448275863, | |
| "grad_norm": 2.404915988640184, | |
| "learning_rate": 4.9219647816383666e-08, | |
| "loss": 0.2711, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.896551724137931, | |
| "grad_norm": 2.129160759410838, | |
| "learning_rate": 3.617714790465576e-08, | |
| "loss": 0.278, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.913793103448276, | |
| "grad_norm": 1.9571879472542826, | |
| "learning_rate": 2.5132291273042288e-08, | |
| "loss": 0.2778, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.9310344827586206, | |
| "grad_norm": 1.9523875015010932, | |
| "learning_rate": 1.608952281901055e-08, | |
| "loss": 0.2846, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.9482758620689653, | |
| "grad_norm": 1.935690040865698, | |
| "learning_rate": 9.052481718690998e-09, | |
| "loss": 0.284, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.9655172413793105, | |
| "grad_norm": 2.2744658928124317, | |
| "learning_rate": 4.0239999623226115e-09, | |
| "loss": 0.2666, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.9827586206896552, | |
| "grad_norm": 2.021706645465733, | |
| "learning_rate": 1.006101214545696e-09, | |
| "loss": 0.2556, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.7378856554061182, | |
| "learning_rate": 0.0, | |
| "loss": 0.2884, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 870, | |
| "total_flos": 274577504862208.0, | |
| "train_loss": 0.2316871819825008, | |
| "train_runtime": 5750.3528, | |
| "train_samples_per_second": 19.35, | |
| "train_steps_per_second": 0.151 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 870, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 274577504862208.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |