| { |
| "dataset_name": "Mu-Math", |
| "group_index": 9, |
| "prompt_group_file": "/hkfs/work/workspace/scratch/tum_fmp0582-dndworkspace/自己训练lora/train_lora/prompt_groups/Mu-Math/group_09.json", |
| "output_dir": "/hkfs/work/workspace/scratch/tum_fmp0582-dndworkspace/自己训练lora/train_lora/outputs/Mu-Math/group_09", |
| "checkpoint_root": "/hkfs/work/workspace/scratch/tum_fmp0582-dndworkspace/自己训练lora/train_lora/outputs/Mu-Math/group_09/checkpoints", |
| "generated_at": "2025-11-06T12:08:18Z", |
| "train_loss": 0.051218728774951566, |
| "metrics": { |
| "train_runtime": 30712.13, |
| "train_samples_per_second": 1.875, |
| "train_steps_per_second": 0.059, |
| "total_flos": 1.5308141101056e+18, |
| "train_loss": 0.051218728774951566, |
| "epoch": 450.0 |
| }, |
| "trainer_state": [ |
| { |
| "loss": 1.8592, |
| "grad_norm": 12.20382308959961, |
| "learning_rate": 0.0, |
| "epoch": 0.32, |
| "step": 1 |
| }, |
| { |
| "loss": 1.711, |
| "grad_norm": 7.592686653137207, |
| "learning_rate": 7.4074074074074075e-06, |
| "epoch": 1.32, |
| "step": 5 |
| }, |
| { |
| "loss": 1.5701, |
| "grad_norm": 3.648606777191162, |
| "learning_rate": 1.6666666666666667e-05, |
| "epoch": 2.64, |
| "step": 10 |
| }, |
| { |
| "loss": 1.1998, |
| "grad_norm": 1.6780602931976318, |
| "learning_rate": 2.5925925925925925e-05, |
| "epoch": 3.96, |
| "step": 15 |
| }, |
| { |
| "loss": 0.9021, |
| "grad_norm": 1.223113775253296, |
| "learning_rate": 3.518518518518519e-05, |
| "epoch": 5.0, |
| "step": 20 |
| }, |
| { |
| "loss": 0.7777, |
| "grad_norm": 0.530234694480896, |
| "learning_rate": 4.4444444444444447e-05, |
| "epoch": 6.32, |
| "step": 25 |
| }, |
| { |
| "loss": 0.614, |
| "grad_norm": 0.5733339786529541, |
| "learning_rate": 5.370370370370371e-05, |
| "epoch": 7.64, |
| "step": 30 |
| }, |
| { |
| "loss": 0.5389, |
| "grad_norm": 0.6008118391036987, |
| "learning_rate": 6.296296296296296e-05, |
| "epoch": 8.96, |
| "step": 35 |
| }, |
| { |
| "loss": 0.4315, |
| "grad_norm": 1.1401246786117554, |
| "learning_rate": 7.222222222222222e-05, |
| "epoch": 10.0, |
| "step": 40 |
| }, |
| { |
| "loss": 0.3393, |
| "grad_norm": 0.5528788566589355, |
| "learning_rate": 8.148148148148148e-05, |
| "epoch": 11.32, |
| "step": 45 |
| }, |
| { |
| "loss": 0.233, |
| "grad_norm": 0.7162251472473145, |
| "learning_rate": 9.074074074074075e-05, |
| "epoch": 12.64, |
| "step": 50 |
| }, |
| { |
| "loss": 0.1434, |
| "grad_norm": 0.6179470419883728, |
| "learning_rate": 0.0001, |
| "epoch": 13.96, |
| "step": 55 |
| }, |
| { |
| "loss": 0.111, |
| "grad_norm": 1.2848559617996216, |
| "learning_rate": 9.971363115693013e-05, |
| "epoch": 15.0, |
| "step": 60 |
| }, |
| { |
| "loss": 0.0602, |
| "grad_norm": 0.4586828351020813, |
| "learning_rate": 9.942726231386026e-05, |
| "epoch": 16.32, |
| "step": 65 |
| }, |
| { |
| "loss": 0.0501, |
| "grad_norm": 0.6835700869560242, |
| "learning_rate": 9.914089347079038e-05, |
| "epoch": 17.64, |
| "step": 70 |
| }, |
| { |
| "loss": 0.0514, |
| "grad_norm": 0.9215507507324219, |
| "learning_rate": 9.885452462772051e-05, |
| "epoch": 18.96, |
| "step": 75 |
| }, |
| { |
| "loss": 0.0479, |
| "grad_norm": 1.617738962173462, |
| "learning_rate": 9.856815578465064e-05, |
| "epoch": 20.0, |
| "step": 80 |
| }, |
| { |
| "loss": 0.0423, |
| "grad_norm": 0.2121235877275467, |
| "learning_rate": 9.828178694158075e-05, |
| "epoch": 21.32, |
| "step": 85 |
| }, |
| { |
| "loss": 0.0435, |
| "grad_norm": 0.41456228494644165, |
| "learning_rate": 9.799541809851088e-05, |
| "epoch": 22.64, |
| "step": 90 |
| }, |
| { |
| "loss": 0.0351, |
| "grad_norm": 0.3220762610435486, |
| "learning_rate": 9.7709049255441e-05, |
| "epoch": 23.96, |
| "step": 95 |
| }, |
| { |
| "loss": 0.0336, |
| "grad_norm": 0.5470985174179077, |
| "learning_rate": 9.742268041237114e-05, |
| "epoch": 25.0, |
| "step": 100 |
| }, |
| { |
| "loss": 0.0375, |
| "grad_norm": 0.166081041097641, |
| "learning_rate": 9.713631156930127e-05, |
| "epoch": 26.32, |
| "step": 105 |
| }, |
| { |
| "loss": 0.0346, |
| "grad_norm": 0.1535070687532425, |
| "learning_rate": 9.68499427262314e-05, |
| "epoch": 27.64, |
| "step": 110 |
| }, |
| { |
| "loss": 0.0375, |
| "grad_norm": 0.14369714260101318, |
| "learning_rate": 9.656357388316152e-05, |
| "epoch": 28.96, |
| "step": 115 |
| }, |
| { |
| "loss": 0.0354, |
| "grad_norm": 0.41668763756752014, |
| "learning_rate": 9.627720504009165e-05, |
| "epoch": 30.0, |
| "step": 120 |
| }, |
| { |
| "loss": 0.0354, |
| "grad_norm": 0.12425274401903152, |
| "learning_rate": 9.599083619702178e-05, |
| "epoch": 31.32, |
| "step": 125 |
| }, |
| { |
| "loss": 0.0318, |
| "grad_norm": 0.10805577784776688, |
| "learning_rate": 9.57044673539519e-05, |
| "epoch": 32.64, |
| "step": 130 |
| }, |
| { |
| "loss": 0.0336, |
| "grad_norm": 0.16544418036937714, |
| "learning_rate": 9.541809851088203e-05, |
| "epoch": 33.96, |
| "step": 135 |
| }, |
| { |
| "loss": 0.0288, |
| "grad_norm": 0.21030113101005554, |
| "learning_rate": 9.513172966781214e-05, |
| "epoch": 35.0, |
| "step": 140 |
| }, |
| { |
| "loss": 0.03, |
| "grad_norm": 0.08930998295545578, |
| "learning_rate": 9.484536082474227e-05, |
| "epoch": 36.32, |
| "step": 145 |
| }, |
| { |
| "loss": 0.0309, |
| "grad_norm": 0.07835185527801514, |
| "learning_rate": 9.45589919816724e-05, |
| "epoch": 37.64, |
| "step": 150 |
| }, |
| { |
| "loss": 0.0331, |
| "grad_norm": 0.29740723967552185, |
| "learning_rate": 9.427262313860252e-05, |
| "epoch": 38.96, |
| "step": 155 |
| }, |
| { |
| "loss": 0.0341, |
| "grad_norm": 0.1962287724018097, |
| "learning_rate": 9.398625429553265e-05, |
| "epoch": 40.0, |
| "step": 160 |
| }, |
| { |
| "loss": 0.0292, |
| "grad_norm": 0.08103634417057037, |
| "learning_rate": 9.369988545246277e-05, |
| "epoch": 41.32, |
| "step": 165 |
| }, |
| { |
| "loss": 0.0288, |
| "grad_norm": 0.08957203477621078, |
| "learning_rate": 9.34135166093929e-05, |
| "epoch": 42.64, |
| "step": 170 |
| }, |
| { |
| "loss": 0.0291, |
| "grad_norm": 0.13071388006210327, |
| "learning_rate": 9.312714776632303e-05, |
| "epoch": 43.96, |
| "step": 175 |
| }, |
| { |
| "loss": 0.0334, |
| "grad_norm": 0.38164010643959045, |
| "learning_rate": 9.284077892325315e-05, |
| "epoch": 45.0, |
| "step": 180 |
| }, |
| { |
| "loss": 0.038, |
| "grad_norm": 0.09823141247034073, |
| "learning_rate": 9.255441008018328e-05, |
| "epoch": 46.32, |
| "step": 185 |
| }, |
| { |
| "loss": 0.0307, |
| "grad_norm": 0.1312790811061859, |
| "learning_rate": 9.22680412371134e-05, |
| "epoch": 47.64, |
| "step": 190 |
| }, |
| { |
| "loss": 0.0312, |
| "grad_norm": 0.07468929886817932, |
| "learning_rate": 9.198167239404353e-05, |
| "epoch": 48.96, |
| "step": 195 |
| }, |
| { |
| "loss": 0.0314, |
| "grad_norm": 0.4955481290817261, |
| "learning_rate": 9.169530355097366e-05, |
| "epoch": 50.0, |
| "step": 200 |
| }, |
| { |
| "loss": 0.0294, |
| "grad_norm": 0.1239602342247963, |
| "learning_rate": 9.140893470790379e-05, |
| "epoch": 51.32, |
| "step": 205 |
| }, |
| { |
| "loss": 0.0321, |
| "grad_norm": 0.09653571248054504, |
| "learning_rate": 9.112256586483391e-05, |
| "epoch": 52.64, |
| "step": 210 |
| }, |
| { |
| "loss": 0.0281, |
| "grad_norm": 0.07385823130607605, |
| "learning_rate": 9.083619702176404e-05, |
| "epoch": 53.96, |
| "step": 215 |
| }, |
| { |
| "loss": 0.0335, |
| "grad_norm": 0.3176237642765045, |
| "learning_rate": 9.054982817869416e-05, |
| "epoch": 55.0, |
| "step": 220 |
| }, |
| { |
| "loss": 0.0308, |
| "grad_norm": 0.09472192823886871, |
| "learning_rate": 9.026345933562429e-05, |
| "epoch": 56.32, |
| "step": 225 |
| }, |
| { |
| "loss": 0.0309, |
| "grad_norm": 0.06622374057769775, |
| "learning_rate": 8.997709049255442e-05, |
| "epoch": 57.64, |
| "step": 230 |
| }, |
| { |
| "loss": 0.0302, |
| "grad_norm": 0.07544614374637604, |
| "learning_rate": 8.969072164948454e-05, |
| "epoch": 58.96, |
| "step": 235 |
| }, |
| { |
| "loss": 0.0291, |
| "grad_norm": 0.2342057228088379, |
| "learning_rate": 8.940435280641467e-05, |
| "epoch": 60.0, |
| "step": 240 |
| }, |
| { |
| "loss": 0.0285, |
| "grad_norm": 0.08162841200828552, |
| "learning_rate": 8.91179839633448e-05, |
| "epoch": 61.32, |
| "step": 245 |
| }, |
| { |
| "loss": 0.0299, |
| "grad_norm": 0.05843600630760193, |
| "learning_rate": 8.883161512027491e-05, |
| "epoch": 62.64, |
| "step": 250 |
| }, |
| { |
| "loss": 0.0285, |
| "grad_norm": 0.3217257261276245, |
| "learning_rate": 8.854524627720504e-05, |
| "epoch": 63.96, |
| "step": 255 |
| }, |
| { |
| "loss": 0.0291, |
| "grad_norm": 0.13366186618804932, |
| "learning_rate": 8.825887743413516e-05, |
| "epoch": 65.0, |
| "step": 260 |
| }, |
| { |
| "loss": 0.0283, |
| "grad_norm": 0.08992842584848404, |
| "learning_rate": 8.797250859106529e-05, |
| "epoch": 66.32, |
| "step": 265 |
| }, |
| { |
| "loss": 0.0275, |
| "grad_norm": 0.4722216725349426, |
| "learning_rate": 8.768613974799542e-05, |
| "epoch": 67.64, |
| "step": 270 |
| }, |
| { |
| "loss": 0.0284, |
| "grad_norm": 0.10617433488368988, |
| "learning_rate": 8.739977090492554e-05, |
| "epoch": 68.96, |
| "step": 275 |
| }, |
| { |
| "loss": 0.0315, |
| "grad_norm": 4.244348049163818, |
| "learning_rate": 8.711340206185567e-05, |
| "epoch": 70.0, |
| "step": 280 |
| }, |
| { |
| "loss": 0.0282, |
| "grad_norm": 0.32534635066986084, |
| "learning_rate": 8.682703321878581e-05, |
| "epoch": 71.32, |
| "step": 285 |
| }, |
| { |
| "loss": 0.0343, |
| "grad_norm": 0.10109279304742813, |
| "learning_rate": 8.654066437571594e-05, |
| "epoch": 72.64, |
| "step": 290 |
| }, |
| { |
| "loss": 0.0301, |
| "grad_norm": 0.10409004241228104, |
| "learning_rate": 8.625429553264606e-05, |
| "epoch": 73.96, |
| "step": 295 |
| }, |
| { |
| "loss": 0.0316, |
| "grad_norm": 0.1833920031785965, |
| "learning_rate": 8.596792668957619e-05, |
| "epoch": 75.0, |
| "step": 300 |
| }, |
| { |
| "loss": 0.03, |
| "grad_norm": 0.06169729679822922, |
| "learning_rate": 8.56815578465063e-05, |
| "epoch": 76.32, |
| "step": 305 |
| }, |
| { |
| "loss": 0.0331, |
| "grad_norm": 0.0845058485865593, |
| "learning_rate": 8.539518900343643e-05, |
| "epoch": 77.64, |
| "step": 310 |
| }, |
| { |
| "loss": 0.0321, |
| "grad_norm": 0.14785389602184296, |
| "learning_rate": 8.510882016036655e-05, |
| "epoch": 78.96, |
| "step": 315 |
| }, |
| { |
| "loss": 0.0293, |
| "grad_norm": 0.2871634066104889, |
| "learning_rate": 8.482245131729668e-05, |
| "epoch": 80.0, |
| "step": 320 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.0759759396314621, |
| "learning_rate": 8.453608247422681e-05, |
| "epoch": 81.32, |
| "step": 325 |
| }, |
| { |
| "loss": 0.0302, |
| "grad_norm": 0.14092276990413666, |
| "learning_rate": 8.424971363115693e-05, |
| "epoch": 82.64, |
| "step": 330 |
| }, |
| { |
| "loss": 0.0322, |
| "grad_norm": 0.513141393661499, |
| "learning_rate": 8.396334478808706e-05, |
| "epoch": 83.96, |
| "step": 335 |
| }, |
| { |
| "loss": 0.0297, |
| "grad_norm": 0.16899359226226807, |
| "learning_rate": 8.367697594501719e-05, |
| "epoch": 85.0, |
| "step": 340 |
| }, |
| { |
| "loss": 0.0294, |
| "grad_norm": 0.059015680104494095, |
| "learning_rate": 8.339060710194731e-05, |
| "epoch": 86.32, |
| "step": 345 |
| }, |
| { |
| "loss": 0.0283, |
| "grad_norm": 0.40950626134872437, |
| "learning_rate": 8.310423825887744e-05, |
| "epoch": 87.64, |
| "step": 350 |
| }, |
| { |
| "loss": 0.0343, |
| "grad_norm": 0.4559670090675354, |
| "learning_rate": 8.281786941580757e-05, |
| "epoch": 88.96, |
| "step": 355 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.14732497930526733, |
| "learning_rate": 8.253150057273768e-05, |
| "epoch": 90.0, |
| "step": 360 |
| }, |
| { |
| "loss": 0.029, |
| "grad_norm": 0.09411142766475677, |
| "learning_rate": 8.224513172966782e-05, |
| "epoch": 91.32, |
| "step": 365 |
| }, |
| { |
| "loss": 0.0288, |
| "grad_norm": 0.1458071768283844, |
| "learning_rate": 8.195876288659795e-05, |
| "epoch": 92.64, |
| "step": 370 |
| }, |
| { |
| "loss": 0.0287, |
| "grad_norm": 0.10315226763486862, |
| "learning_rate": 8.167239404352807e-05, |
| "epoch": 93.96, |
| "step": 375 |
| }, |
| { |
| "loss": 0.0294, |
| "grad_norm": 0.21710525453090668, |
| "learning_rate": 8.13860252004582e-05, |
| "epoch": 95.0, |
| "step": 380 |
| }, |
| { |
| "loss": 0.0277, |
| "grad_norm": 0.1466694325208664, |
| "learning_rate": 8.109965635738833e-05, |
| "epoch": 96.32, |
| "step": 385 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.0593799464404583, |
| "learning_rate": 8.081328751431845e-05, |
| "epoch": 97.64, |
| "step": 390 |
| }, |
| { |
| "loss": 0.0311, |
| "grad_norm": 0.2900218963623047, |
| "learning_rate": 8.052691867124858e-05, |
| "epoch": 98.96, |
| "step": 395 |
| }, |
| { |
| "loss": 0.0325, |
| "grad_norm": 0.2579890787601471, |
| "learning_rate": 8.02405498281787e-05, |
| "epoch": 100.0, |
| "step": 400 |
| }, |
| { |
| "loss": 0.0341, |
| "grad_norm": 0.07449392974376678, |
| "learning_rate": 7.995418098510883e-05, |
| "epoch": 101.32, |
| "step": 405 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.08262854814529419, |
| "learning_rate": 7.966781214203894e-05, |
| "epoch": 102.64, |
| "step": 410 |
| }, |
| { |
| "loss": 0.0289, |
| "grad_norm": 0.05616652965545654, |
| "learning_rate": 7.938144329896907e-05, |
| "epoch": 103.96, |
| "step": 415 |
| }, |
| { |
| "loss": 0.0308, |
| "grad_norm": 0.20183321833610535, |
| "learning_rate": 7.90950744558992e-05, |
| "epoch": 105.0, |
| "step": 420 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.054062746465206146, |
| "learning_rate": 7.880870561282932e-05, |
| "epoch": 106.32, |
| "step": 425 |
| }, |
| { |
| "loss": 0.0299, |
| "grad_norm": 0.23319438099861145, |
| "learning_rate": 7.852233676975945e-05, |
| "epoch": 107.64, |
| "step": 430 |
| }, |
| { |
| "loss": 0.0341, |
| "grad_norm": 0.42012178897857666, |
| "learning_rate": 7.823596792668958e-05, |
| "epoch": 108.96, |
| "step": 435 |
| }, |
| { |
| "loss": 0.0302, |
| "grad_norm": 0.13727161288261414, |
| "learning_rate": 7.79495990836197e-05, |
| "epoch": 110.0, |
| "step": 440 |
| }, |
| { |
| "loss": 0.0297, |
| "grad_norm": 0.06950726360082626, |
| "learning_rate": 7.766323024054983e-05, |
| "epoch": 111.32, |
| "step": 445 |
| }, |
| { |
| "loss": 0.0301, |
| "grad_norm": 0.10826419293880463, |
| "learning_rate": 7.737686139747996e-05, |
| "epoch": 112.64, |
| "step": 450 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.06952483206987381, |
| "learning_rate": 7.709049255441008e-05, |
| "epoch": 113.96, |
| "step": 455 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.16315822303295135, |
| "learning_rate": 7.680412371134021e-05, |
| "epoch": 115.0, |
| "step": 460 |
| }, |
| { |
| "loss": 0.0276, |
| "grad_norm": 0.056907810270786285, |
| "learning_rate": 7.651775486827034e-05, |
| "epoch": 116.32, |
| "step": 465 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.05950301140546799, |
| "learning_rate": 7.623138602520046e-05, |
| "epoch": 117.64, |
| "step": 470 |
| }, |
| { |
| "loss": 0.029, |
| "grad_norm": 0.04811059311032295, |
| "learning_rate": 7.594501718213059e-05, |
| "epoch": 118.96, |
| "step": 475 |
| }, |
| { |
| "loss": 0.0301, |
| "grad_norm": 0.19055458903312683, |
| "learning_rate": 7.565864833906071e-05, |
| "epoch": 120.0, |
| "step": 480 |
| }, |
| { |
| "loss": 0.0254, |
| "grad_norm": 0.0495222732424736, |
| "learning_rate": 7.537227949599084e-05, |
| "epoch": 121.32, |
| "step": 485 |
| }, |
| { |
| "loss": 0.0277, |
| "grad_norm": 0.05043775215744972, |
| "learning_rate": 7.508591065292097e-05, |
| "epoch": 122.64, |
| "step": 490 |
| }, |
| { |
| "loss": 0.0289, |
| "grad_norm": 0.05455939844250679, |
| "learning_rate": 7.47995418098511e-05, |
| "epoch": 123.96, |
| "step": 495 |
| }, |
| { |
| "loss": 0.0268, |
| "grad_norm": 0.11698618531227112, |
| "learning_rate": 7.451317296678122e-05, |
| "epoch": 125.0, |
| "step": 500 |
| }, |
| { |
| "loss": 0.0285, |
| "grad_norm": 0.04787712171673775, |
| "learning_rate": 7.422680412371135e-05, |
| "epoch": 126.32, |
| "step": 505 |
| }, |
| { |
| "loss": 0.0258, |
| "grad_norm": 0.04081322252750397, |
| "learning_rate": 7.394043528064147e-05, |
| "epoch": 127.64, |
| "step": 510 |
| }, |
| { |
| "loss": 0.0282, |
| "grad_norm": 0.04783250391483307, |
| "learning_rate": 7.36540664375716e-05, |
| "epoch": 128.96, |
| "step": 515 |
| }, |
| { |
| "loss": 0.0293, |
| "grad_norm": 0.21604378521442413, |
| "learning_rate": 7.336769759450171e-05, |
| "epoch": 130.0, |
| "step": 520 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.05380740389227867, |
| "learning_rate": 7.308132875143184e-05, |
| "epoch": 131.32, |
| "step": 525 |
| }, |
| { |
| "loss": 0.025, |
| "grad_norm": 0.050852205604314804, |
| "learning_rate": 7.279495990836197e-05, |
| "epoch": 132.64, |
| "step": 530 |
| }, |
| { |
| "loss": 0.028, |
| "grad_norm": 0.056515030562877655, |
| "learning_rate": 7.250859106529209e-05, |
| "epoch": 133.96, |
| "step": 535 |
| }, |
| { |
| "loss": 0.0247, |
| "grad_norm": 0.11390186846256256, |
| "learning_rate": 7.222222222222222e-05, |
| "epoch": 135.0, |
| "step": 540 |
| }, |
| { |
| "loss": 0.0298, |
| "grad_norm": 0.04874785616993904, |
| "learning_rate": 7.193585337915235e-05, |
| "epoch": 136.32, |
| "step": 545 |
| }, |
| { |
| "loss": 0.0266, |
| "grad_norm": 0.0457790307700634, |
| "learning_rate": 7.164948453608247e-05, |
| "epoch": 137.64, |
| "step": 550 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.04593640938401222, |
| "learning_rate": 7.136311569301261e-05, |
| "epoch": 138.96, |
| "step": 555 |
| }, |
| { |
| "loss": 0.0286, |
| "grad_norm": 0.1468404084444046, |
| "learning_rate": 7.107674684994274e-05, |
| "epoch": 140.0, |
| "step": 560 |
| }, |
| { |
| "loss": 0.0262, |
| "grad_norm": 0.03566594421863556, |
| "learning_rate": 7.079037800687286e-05, |
| "epoch": 141.32, |
| "step": 565 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.04443350061774254, |
| "learning_rate": 7.050400916380299e-05, |
| "epoch": 142.64, |
| "step": 570 |
| }, |
| { |
| "loss": 0.0307, |
| "grad_norm": 0.04099564626812935, |
| "learning_rate": 7.02176403207331e-05, |
| "epoch": 143.96, |
| "step": 575 |
| }, |
| { |
| "loss": 0.0301, |
| "grad_norm": 0.1269964724779129, |
| "learning_rate": 6.993127147766323e-05, |
| "epoch": 145.0, |
| "step": 580 |
| }, |
| { |
| "loss": 0.0289, |
| "grad_norm": 0.049064043909311295, |
| "learning_rate": 6.964490263459336e-05, |
| "epoch": 146.32, |
| "step": 585 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.035255834460258484, |
| "learning_rate": 6.935853379152348e-05, |
| "epoch": 147.64, |
| "step": 590 |
| }, |
| { |
| "loss": 0.0282, |
| "grad_norm": 0.04822870343923569, |
| "learning_rate": 6.907216494845361e-05, |
| "epoch": 148.96, |
| "step": 595 |
| }, |
| { |
| "loss": 0.0293, |
| "grad_norm": 0.11265220493078232, |
| "learning_rate": 6.878579610538374e-05, |
| "epoch": 150.0, |
| "step": 600 |
| }, |
| { |
| "loss": 0.0268, |
| "grad_norm": 0.03577497974038124, |
| "learning_rate": 6.849942726231386e-05, |
| "epoch": 151.32, |
| "step": 605 |
| }, |
| { |
| "loss": 0.0298, |
| "grad_norm": 0.029043061658740044, |
| "learning_rate": 6.821305841924399e-05, |
| "epoch": 152.64, |
| "step": 610 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.03485685959458351, |
| "learning_rate": 6.792668957617412e-05, |
| "epoch": 153.96, |
| "step": 615 |
| }, |
| { |
| "loss": 0.0274, |
| "grad_norm": 0.1452644020318985, |
| "learning_rate": 6.764032073310424e-05, |
| "epoch": 155.0, |
| "step": 620 |
| }, |
| { |
| "loss": 0.0284, |
| "grad_norm": 0.034407999366521835, |
| "learning_rate": 6.735395189003437e-05, |
| "epoch": 156.32, |
| "step": 625 |
| }, |
| { |
| "loss": 0.0285, |
| "grad_norm": 0.03767184168100357, |
| "learning_rate": 6.706758304696448e-05, |
| "epoch": 157.64, |
| "step": 630 |
| }, |
| { |
| "loss": 0.0249, |
| "grad_norm": 0.043731339275836945, |
| "learning_rate": 6.678121420389462e-05, |
| "epoch": 158.96, |
| "step": 635 |
| }, |
| { |
| "loss": 0.0279, |
| "grad_norm": 0.12056618928909302, |
| "learning_rate": 6.649484536082475e-05, |
| "epoch": 160.0, |
| "step": 640 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.03441184386610985, |
| "learning_rate": 6.620847651775487e-05, |
| "epoch": 161.32, |
| "step": 645 |
| }, |
| { |
| "loss": 0.0292, |
| "grad_norm": 0.03794039785861969, |
| "learning_rate": 6.5922107674685e-05, |
| "epoch": 162.64, |
| "step": 650 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.04225137084722519, |
| "learning_rate": 6.563573883161513e-05, |
| "epoch": 163.96, |
| "step": 655 |
| }, |
| { |
| "loss": 0.0262, |
| "grad_norm": 0.11336100846529007, |
| "learning_rate": 6.534936998854525e-05, |
| "epoch": 165.0, |
| "step": 660 |
| }, |
| { |
| "loss": 0.0283, |
| "grad_norm": 0.04133683070540428, |
| "learning_rate": 6.506300114547538e-05, |
| "epoch": 166.32, |
| "step": 665 |
| }, |
| { |
| "loss": 0.0279, |
| "grad_norm": 0.037548743188381195, |
| "learning_rate": 6.477663230240551e-05, |
| "epoch": 167.64, |
| "step": 670 |
| }, |
| { |
| "loss": 0.0291, |
| "grad_norm": 0.038737714290618896, |
| "learning_rate": 6.449026345933563e-05, |
| "epoch": 168.96, |
| "step": 675 |
| }, |
| { |
| "loss": 0.0261, |
| "grad_norm": 0.1200108751654625, |
| "learning_rate": 6.420389461626576e-05, |
| "epoch": 170.0, |
| "step": 680 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.03670820966362953, |
| "learning_rate": 6.391752577319587e-05, |
| "epoch": 171.32, |
| "step": 685 |
| }, |
| { |
| "loss": 0.0277, |
| "grad_norm": 0.038459792733192444, |
| "learning_rate": 6.3631156930126e-05, |
| "epoch": 172.64, |
| "step": 690 |
| }, |
| { |
| "loss": 0.0256, |
| "grad_norm": 0.04245300218462944, |
| "learning_rate": 6.334478808705613e-05, |
| "epoch": 173.96, |
| "step": 695 |
| }, |
| { |
| "loss": 0.0239, |
| "grad_norm": 0.08714708685874939, |
| "learning_rate": 6.305841924398625e-05, |
| "epoch": 175.0, |
| "step": 700 |
| }, |
| { |
| "loss": 0.026, |
| "grad_norm": 0.03563086688518524, |
| "learning_rate": 6.277205040091638e-05, |
| "epoch": 176.32, |
| "step": 705 |
| }, |
| { |
| "loss": 0.029, |
| "grad_norm": 0.03432348370552063, |
| "learning_rate": 6.24856815578465e-05, |
| "epoch": 177.64, |
| "step": 710 |
| }, |
| { |
| "loss": 0.0299, |
| "grad_norm": 0.03847336396574974, |
| "learning_rate": 6.219931271477663e-05, |
| "epoch": 178.96, |
| "step": 715 |
| }, |
| { |
| "loss": 0.0249, |
| "grad_norm": 0.0912068635225296, |
| "learning_rate": 6.191294387170676e-05, |
| "epoch": 180.0, |
| "step": 720 |
| }, |
| { |
| "loss": 0.0296, |
| "grad_norm": 0.03907967358827591, |
| "learning_rate": 6.162657502863689e-05, |
| "epoch": 181.32, |
| "step": 725 |
| }, |
| { |
| "loss": 0.0266, |
| "grad_norm": 0.03369147703051567, |
| "learning_rate": 6.134020618556701e-05, |
| "epoch": 182.64, |
| "step": 730 |
| }, |
| { |
| "loss": 0.0291, |
| "grad_norm": 0.03768694028258324, |
| "learning_rate": 6.105383734249714e-05, |
| "epoch": 183.96, |
| "step": 735 |
| }, |
| { |
| "loss": 0.0276, |
| "grad_norm": 0.15149357914924622, |
| "learning_rate": 6.076746849942726e-05, |
| "epoch": 185.0, |
| "step": 740 |
| }, |
| { |
| "loss": 0.0276, |
| "grad_norm": 0.038406625390052795, |
| "learning_rate": 6.0481099656357384e-05, |
| "epoch": 186.32, |
| "step": 745 |
| }, |
| { |
| "loss": 0.0284, |
| "grad_norm": 0.03291546180844307, |
| "learning_rate": 6.019473081328752e-05, |
| "epoch": 187.64, |
| "step": 750 |
| }, |
| { |
| "loss": 0.0266, |
| "grad_norm": 0.046228110790252686, |
| "learning_rate": 5.9908361970217644e-05, |
| "epoch": 188.96, |
| "step": 755 |
| }, |
| { |
| "loss": 0.0263, |
| "grad_norm": 0.12910252809524536, |
| "learning_rate": 5.962199312714777e-05, |
| "epoch": 190.0, |
| "step": 760 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.03608384728431702, |
| "learning_rate": 5.93356242840779e-05, |
| "epoch": 191.32, |
| "step": 765 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.03974054008722305, |
| "learning_rate": 5.904925544100802e-05, |
| "epoch": 192.64, |
| "step": 770 |
| }, |
| { |
| "loss": 0.0292, |
| "grad_norm": 0.037297558039426804, |
| "learning_rate": 5.876288659793815e-05, |
| "epoch": 193.96, |
| "step": 775 |
| }, |
| { |
| "loss": 0.0264, |
| "grad_norm": 0.09680958837270737, |
| "learning_rate": 5.8476517754868276e-05, |
| "epoch": 195.0, |
| "step": 780 |
| }, |
| { |
| "loss": 0.0262, |
| "grad_norm": 0.041529204696416855, |
| "learning_rate": 5.81901489117984e-05, |
| "epoch": 196.32, |
| "step": 785 |
| }, |
| { |
| "loss": 0.0275, |
| "grad_norm": 0.034658145159482956, |
| "learning_rate": 5.790378006872853e-05, |
| "epoch": 197.64, |
| "step": 790 |
| }, |
| { |
| "loss": 0.0265, |
| "grad_norm": 0.035674963146448135, |
| "learning_rate": 5.761741122565865e-05, |
| "epoch": 198.96, |
| "step": 795 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.10445430874824524, |
| "learning_rate": 5.7331042382588775e-05, |
| "epoch": 200.0, |
| "step": 800 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.038163043558597565, |
| "learning_rate": 5.70446735395189e-05, |
| "epoch": 201.32, |
| "step": 805 |
| }, |
| { |
| "loss": 0.0257, |
| "grad_norm": 0.028517192229628563, |
| "learning_rate": 5.675830469644903e-05, |
| "epoch": 202.64, |
| "step": 810 |
| }, |
| { |
| "loss": 0.0288, |
| "grad_norm": 0.02864096872508526, |
| "learning_rate": 5.6471935853379155e-05, |
| "epoch": 203.96, |
| "step": 815 |
| }, |
| { |
| "loss": 0.0296, |
| "grad_norm": 0.1180969849228859, |
| "learning_rate": 5.618556701030928e-05, |
| "epoch": 205.0, |
| "step": 820 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.032436177134513855, |
| "learning_rate": 5.589919816723941e-05, |
| "epoch": 206.32, |
| "step": 825 |
| }, |
| { |
| "loss": 0.03, |
| "grad_norm": 0.04156746715307236, |
| "learning_rate": 5.5612829324169534e-05, |
| "epoch": 207.64, |
| "step": 830 |
| }, |
| { |
| "loss": 0.0259, |
| "grad_norm": 0.03368216007947922, |
| "learning_rate": 5.532646048109966e-05, |
| "epoch": 208.96, |
| "step": 835 |
| }, |
| { |
| "loss": 0.0275, |
| "grad_norm": 0.13619345426559448, |
| "learning_rate": 5.504009163802979e-05, |
| "epoch": 210.0, |
| "step": 840 |
| }, |
| { |
| "loss": 0.0292, |
| "grad_norm": 0.0326935313642025, |
| "learning_rate": 5.4753722794959914e-05, |
| "epoch": 211.32, |
| "step": 845 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.03230908140540123, |
| "learning_rate": 5.4467353951890033e-05, |
| "epoch": 212.64, |
| "step": 850 |
| }, |
| { |
| "loss": 0.0276, |
| "grad_norm": 0.034922704100608826, |
| "learning_rate": 5.418098510882016e-05, |
| "epoch": 213.96, |
| "step": 855 |
| }, |
| { |
| "loss": 0.0267, |
| "grad_norm": 0.07679939270019531, |
| "learning_rate": 5.3894616265750286e-05, |
| "epoch": 215.0, |
| "step": 860 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.037064969539642334, |
| "learning_rate": 5.360824742268041e-05, |
| "epoch": 216.32, |
| "step": 865 |
| }, |
| { |
| "loss": 0.0301, |
| "grad_norm": 0.04513270780444145, |
| "learning_rate": 5.332187857961054e-05, |
| "epoch": 217.64, |
| "step": 870 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.03031047247350216, |
| "learning_rate": 5.3035509736540666e-05, |
| "epoch": 218.96, |
| "step": 875 |
| }, |
| { |
| "loss": 0.0265, |
| "grad_norm": 0.1258755624294281, |
| "learning_rate": 5.274914089347079e-05, |
| "epoch": 220.0, |
| "step": 880 |
| }, |
| { |
| "loss": 0.0299, |
| "grad_norm": 0.03660378232598305, |
| "learning_rate": 5.246277205040092e-05, |
| "epoch": 221.32, |
| "step": 885 |
| }, |
| { |
| "loss": 0.0247, |
| "grad_norm": 0.028527770191431046, |
| "learning_rate": 5.2176403207331045e-05, |
| "epoch": 222.64, |
| "step": 890 |
| }, |
| { |
| "loss": 0.0285, |
| "grad_norm": 0.03860415890812874, |
| "learning_rate": 5.189003436426118e-05, |
| "epoch": 223.96, |
| "step": 895 |
| }, |
| { |
| "loss": 0.0265, |
| "grad_norm": 0.10100441426038742, |
| "learning_rate": 5.1603665521191305e-05, |
| "epoch": 225.0, |
| "step": 900 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.03017512708902359, |
| "learning_rate": 5.131729667812142e-05, |
| "epoch": 226.32, |
| "step": 905 |
| }, |
| { |
| "loss": 0.029, |
| "grad_norm": 0.0309721902012825, |
| "learning_rate": 5.1030927835051544e-05, |
| "epoch": 227.64, |
| "step": 910 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.046862490475177765, |
| "learning_rate": 5.074455899198167e-05, |
| "epoch": 228.96, |
| "step": 915 |
| }, |
| { |
| "loss": 0.0268, |
| "grad_norm": 0.1213354617357254, |
| "learning_rate": 5.04581901489118e-05, |
| "epoch": 230.0, |
| "step": 920 |
| }, |
| { |
| "loss": 0.0296, |
| "grad_norm": 0.02722300961613655, |
| "learning_rate": 5.0171821305841924e-05, |
| "epoch": 231.32, |
| "step": 925 |
| }, |
| { |
| "loss": 0.0262, |
| "grad_norm": 0.03292243555188179, |
| "learning_rate": 4.988545246277205e-05, |
| "epoch": 232.64, |
| "step": 930 |
| }, |
| { |
| "loss": 0.026, |
| "grad_norm": 0.04071458429098129, |
| "learning_rate": 4.9599083619702184e-05, |
| "epoch": 233.96, |
| "step": 935 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.09393114596605301, |
| "learning_rate": 4.931271477663231e-05, |
| "epoch": 235.0, |
| "step": 940 |
| }, |
| { |
| "loss": 0.0262, |
| "grad_norm": 0.031333085149526596, |
| "learning_rate": 4.902634593356243e-05, |
| "epoch": 236.32, |
| "step": 945 |
| }, |
| { |
| "loss": 0.0274, |
| "grad_norm": 0.03641931340098381, |
| "learning_rate": 4.8739977090492556e-05, |
| "epoch": 237.64, |
| "step": 950 |
| }, |
| { |
| "loss": 0.0303, |
| "grad_norm": 0.03518975153565407, |
| "learning_rate": 4.845360824742268e-05, |
| "epoch": 238.96, |
| "step": 955 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.08912181109189987, |
| "learning_rate": 4.816723940435281e-05, |
| "epoch": 240.0, |
| "step": 960 |
| }, |
| { |
| "loss": 0.0268, |
| "grad_norm": 0.03128821775317192, |
| "learning_rate": 4.7880870561282936e-05, |
| "epoch": 241.32, |
| "step": 965 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.03294266760349274, |
| "learning_rate": 4.7594501718213055e-05, |
| "epoch": 242.64, |
| "step": 970 |
| }, |
| { |
| "loss": 0.0242, |
| "grad_norm": 0.03294999524950981, |
| "learning_rate": 4.730813287514318e-05, |
| "epoch": 243.96, |
| "step": 975 |
| }, |
| { |
| "loss": 0.0266, |
| "grad_norm": 0.09809399396181107, |
| "learning_rate": 4.7021764032073315e-05, |
| "epoch": 245.0, |
| "step": 980 |
| }, |
| { |
| "loss": 0.0276, |
| "grad_norm": 0.034879233688116074, |
| "learning_rate": 4.673539518900344e-05, |
| "epoch": 246.32, |
| "step": 985 |
| }, |
| { |
| "loss": 0.0281, |
| "grad_norm": 0.031896039843559265, |
| "learning_rate": 4.644902634593357e-05, |
| "epoch": 247.64, |
| "step": 990 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.03620505332946777, |
| "learning_rate": 4.6162657502863694e-05, |
| "epoch": 248.96, |
| "step": 995 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.12210187315940857, |
| "learning_rate": 4.5876288659793814e-05, |
| "epoch": 250.0, |
| "step": 1000 |
| }, |
| { |
| "loss": 0.0262, |
| "grad_norm": 0.03563583269715309, |
| "learning_rate": 4.558991981672394e-05, |
| "epoch": 251.32, |
| "step": 1005 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.037155359983444214, |
| "learning_rate": 4.530355097365407e-05, |
| "epoch": 252.64, |
| "step": 1010 |
| }, |
| { |
| "loss": 0.0262, |
| "grad_norm": 0.031501322984695435, |
| "learning_rate": 4.5017182130584194e-05, |
| "epoch": 253.96, |
| "step": 1015 |
| }, |
| { |
| "loss": 0.0308, |
| "grad_norm": 0.12778820097446442, |
| "learning_rate": 4.473081328751432e-05, |
| "epoch": 255.0, |
| "step": 1020 |
| }, |
| { |
| "loss": 0.0283, |
| "grad_norm": 0.0294523723423481, |
| "learning_rate": 4.4444444444444447e-05, |
| "epoch": 256.32, |
| "step": 1025 |
| }, |
| { |
| "loss": 0.0259, |
| "grad_norm": 0.02713526226580143, |
| "learning_rate": 4.415807560137457e-05, |
| "epoch": 257.64, |
| "step": 1030 |
| }, |
| { |
| "loss": 0.0263, |
| "grad_norm": 0.03244701400399208, |
| "learning_rate": 4.38717067583047e-05, |
| "epoch": 258.96, |
| "step": 1035 |
| }, |
| { |
| "loss": 0.0291, |
| "grad_norm": 0.1053709089756012, |
| "learning_rate": 4.3585337915234826e-05, |
| "epoch": 260.0, |
| "step": 1040 |
| }, |
| { |
| "loss": 0.0256, |
| "grad_norm": 0.03464820608496666, |
| "learning_rate": 4.329896907216495e-05, |
| "epoch": 261.32, |
| "step": 1045 |
| }, |
| { |
| "loss": 0.0282, |
| "grad_norm": 0.031316548585891724, |
| "learning_rate": 4.301260022909508e-05, |
| "epoch": 262.64, |
| "step": 1050 |
| }, |
| { |
| "loss": 0.0277, |
| "grad_norm": 0.032593999058008194, |
| "learning_rate": 4.27262313860252e-05, |
| "epoch": 263.96, |
| "step": 1055 |
| }, |
| { |
| "loss": 0.0311, |
| "grad_norm": 0.09594771265983582, |
| "learning_rate": 4.2439862542955325e-05, |
| "epoch": 265.0, |
| "step": 1060 |
| }, |
| { |
| "loss": 0.0274, |
| "grad_norm": 0.028000134974718094, |
| "learning_rate": 4.215349369988545e-05, |
| "epoch": 266.32, |
| "step": 1065 |
| }, |
| { |
| "loss": 0.0252, |
| "grad_norm": 0.03135870397090912, |
| "learning_rate": 4.1867124856815585e-05, |
| "epoch": 267.64, |
| "step": 1070 |
| }, |
| { |
| "loss": 0.0295, |
| "grad_norm": 0.029970373958349228, |
| "learning_rate": 4.158075601374571e-05, |
| "epoch": 268.96, |
| "step": 1075 |
| }, |
| { |
| "loss": 0.0288, |
| "grad_norm": 0.11500892788171768, |
| "learning_rate": 4.129438717067583e-05, |
| "epoch": 270.0, |
| "step": 1080 |
| }, |
| { |
| "loss": 0.0264, |
| "grad_norm": 0.03145052120089531, |
| "learning_rate": 4.100801832760596e-05, |
| "epoch": 271.32, |
| "step": 1085 |
| }, |
| { |
| "loss": 0.026, |
| "grad_norm": 0.03734217584133148, |
| "learning_rate": 4.0721649484536084e-05, |
| "epoch": 272.64, |
| "step": 1090 |
| }, |
| { |
| "loss": 0.0265, |
| "grad_norm": 0.035455215722322464, |
| "learning_rate": 4.043528064146621e-05, |
| "epoch": 273.96, |
| "step": 1095 |
| }, |
| { |
| "loss": 0.0318, |
| "grad_norm": 0.13708935678005219, |
| "learning_rate": 4.014891179839634e-05, |
| "epoch": 275.0, |
| "step": 1100 |
| }, |
| { |
| "loss": 0.0282, |
| "grad_norm": 0.036748431622982025, |
| "learning_rate": 3.9862542955326463e-05, |
| "epoch": 276.32, |
| "step": 1105 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.03389858826994896, |
| "learning_rate": 3.957617411225659e-05, |
| "epoch": 277.64, |
| "step": 1110 |
| }, |
| { |
| "loss": 0.0276, |
| "grad_norm": 0.03763600438833237, |
| "learning_rate": 3.9289805269186716e-05, |
| "epoch": 278.96, |
| "step": 1115 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.07438627630472183, |
| "learning_rate": 3.900343642611684e-05, |
| "epoch": 280.0, |
| "step": 1120 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.031286075711250305, |
| "learning_rate": 3.871706758304697e-05, |
| "epoch": 281.32, |
| "step": 1125 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.03399110957980156, |
| "learning_rate": 3.8430698739977096e-05, |
| "epoch": 282.64, |
| "step": 1130 |
| }, |
| { |
| "loss": 0.0255, |
| "grad_norm": 0.03141126036643982, |
| "learning_rate": 3.8144329896907216e-05, |
| "epoch": 283.96, |
| "step": 1135 |
| }, |
| { |
| "loss": 0.0296, |
| "grad_norm": 0.11621753126382828, |
| "learning_rate": 3.785796105383734e-05, |
| "epoch": 285.0, |
| "step": 1140 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.03157641738653183, |
| "learning_rate": 3.757159221076747e-05, |
| "epoch": 286.32, |
| "step": 1145 |
| }, |
| { |
| "loss": 0.0256, |
| "grad_norm": 0.025382421910762787, |
| "learning_rate": 3.7285223367697595e-05, |
| "epoch": 287.64, |
| "step": 1150 |
| }, |
| { |
| "loss": 0.0285, |
| "grad_norm": 0.028903618454933167, |
| "learning_rate": 3.699885452462772e-05, |
| "epoch": 288.96, |
| "step": 1155 |
| }, |
| { |
| "loss": 0.0283, |
| "grad_norm": 0.10580641031265259, |
| "learning_rate": 3.671248568155785e-05, |
| "epoch": 290.0, |
| "step": 1160 |
| }, |
| { |
| "loss": 0.026, |
| "grad_norm": 0.03191475197672844, |
| "learning_rate": 3.6426116838487974e-05, |
| "epoch": 291.32, |
| "step": 1165 |
| }, |
| { |
| "loss": 0.0274, |
| "grad_norm": 0.02625187113881111, |
| "learning_rate": 3.61397479954181e-05, |
| "epoch": 292.64, |
| "step": 1170 |
| }, |
| { |
| "loss": 0.0275, |
| "grad_norm": 0.034425340592861176, |
| "learning_rate": 3.585337915234823e-05, |
| "epoch": 293.96, |
| "step": 1175 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.10750774294137955, |
| "learning_rate": 3.5567010309278354e-05, |
| "epoch": 295.0, |
| "step": 1180 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.02907622791826725, |
| "learning_rate": 3.528064146620848e-05, |
| "epoch": 296.32, |
| "step": 1185 |
| }, |
| { |
| "loss": 0.0251, |
| "grad_norm": 0.02848324552178383, |
| "learning_rate": 3.49942726231386e-05, |
| "epoch": 297.64, |
| "step": 1190 |
| }, |
| { |
| "loss": 0.0294, |
| "grad_norm": 0.03020431660115719, |
| "learning_rate": 3.4707903780068726e-05, |
| "epoch": 298.96, |
| "step": 1195 |
| }, |
| { |
| "loss": 0.0258, |
| "grad_norm": 0.10761333256959915, |
| "learning_rate": 3.442153493699885e-05, |
| "epoch": 300.0, |
| "step": 1200 |
| }, |
| { |
| "loss": 0.0261, |
| "grad_norm": 0.027282001450657845, |
| "learning_rate": 3.4135166093928986e-05, |
| "epoch": 301.32, |
| "step": 1205 |
| }, |
| { |
| "loss": 0.025, |
| "grad_norm": 0.03536438196897507, |
| "learning_rate": 3.384879725085911e-05, |
| "epoch": 302.64, |
| "step": 1210 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.027092602103948593, |
| "learning_rate": 3.356242840778923e-05, |
| "epoch": 303.96, |
| "step": 1215 |
| }, |
| { |
| "loss": 0.025, |
| "grad_norm": 0.0726863294839859, |
| "learning_rate": 3.327605956471936e-05, |
| "epoch": 305.0, |
| "step": 1220 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.034631919115781784, |
| "learning_rate": 3.2989690721649485e-05, |
| "epoch": 306.32, |
| "step": 1225 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.036680981516838074, |
| "learning_rate": 3.270332187857961e-05, |
| "epoch": 307.64, |
| "step": 1230 |
| }, |
| { |
| "loss": 0.0268, |
| "grad_norm": 0.02548901177942753, |
| "learning_rate": 3.241695303550974e-05, |
| "epoch": 308.96, |
| "step": 1235 |
| }, |
| { |
| "loss": 0.0286, |
| "grad_norm": 0.12884041666984558, |
| "learning_rate": 3.2130584192439865e-05, |
| "epoch": 310.0, |
| "step": 1240 |
| }, |
| { |
| "loss": 0.0266, |
| "grad_norm": 0.02462596632540226, |
| "learning_rate": 3.184421534936999e-05, |
| "epoch": 311.32, |
| "step": 1245 |
| }, |
| { |
| "loss": 0.0276, |
| "grad_norm": 0.02773185260593891, |
| "learning_rate": 3.155784650630012e-05, |
| "epoch": 312.64, |
| "step": 1250 |
| }, |
| { |
| "loss": 0.0285, |
| "grad_norm": 0.027679359540343285, |
| "learning_rate": 3.1271477663230244e-05, |
| "epoch": 313.96, |
| "step": 1255 |
| }, |
| { |
| "loss": 0.0277, |
| "grad_norm": 0.120714470744133, |
| "learning_rate": 3.098510882016037e-05, |
| "epoch": 315.0, |
| "step": 1260 |
| }, |
| { |
| "loss": 0.0275, |
| "grad_norm": 0.03183070197701454, |
| "learning_rate": 3.06987399770905e-05, |
| "epoch": 316.32, |
| "step": 1265 |
| }, |
| { |
| "loss": 0.0279, |
| "grad_norm": 0.034614916890859604, |
| "learning_rate": 3.0412371134020617e-05, |
| "epoch": 317.64, |
| "step": 1270 |
| }, |
| { |
| "loss": 0.025, |
| "grad_norm": 0.03779444843530655, |
| "learning_rate": 3.0126002290950743e-05, |
| "epoch": 318.96, |
| "step": 1275 |
| }, |
| { |
| "loss": 0.0279, |
| "grad_norm": 0.10465690493583679, |
| "learning_rate": 2.983963344788087e-05, |
| "epoch": 320.0, |
| "step": 1280 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.027243228629231453, |
| "learning_rate": 2.9553264604811e-05, |
| "epoch": 321.32, |
| "step": 1285 |
| }, |
| { |
| "loss": 0.0283, |
| "grad_norm": 0.029929351061582565, |
| "learning_rate": 2.9266895761741126e-05, |
| "epoch": 322.64, |
| "step": 1290 |
| }, |
| { |
| "loss": 0.0262, |
| "grad_norm": 0.030301930382847786, |
| "learning_rate": 2.8980526918671253e-05, |
| "epoch": 323.96, |
| "step": 1295 |
| }, |
| { |
| "loss": 0.0294, |
| "grad_norm": 0.09200041741132736, |
| "learning_rate": 2.8694158075601372e-05, |
| "epoch": 325.0, |
| "step": 1300 |
| }, |
| { |
| "loss": 0.0283, |
| "grad_norm": 0.030280061066150665, |
| "learning_rate": 2.8407789232531502e-05, |
| "epoch": 326.32, |
| "step": 1305 |
| }, |
| { |
| "loss": 0.0257, |
| "grad_norm": 0.02707235887646675, |
| "learning_rate": 2.812142038946163e-05, |
| "epoch": 327.64, |
| "step": 1310 |
| }, |
| { |
| "loss": 0.029, |
| "grad_norm": 0.0276896134018898, |
| "learning_rate": 2.7835051546391755e-05, |
| "epoch": 328.96, |
| "step": 1315 |
| }, |
| { |
| "loss": 0.0267, |
| "grad_norm": 0.10159367322921753, |
| "learning_rate": 2.754868270332188e-05, |
| "epoch": 330.0, |
| "step": 1320 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.035023126751184464, |
| "learning_rate": 2.7262313860252005e-05, |
| "epoch": 331.32, |
| "step": 1325 |
| }, |
| { |
| "loss": 0.0265, |
| "grad_norm": 0.02885899506509304, |
| "learning_rate": 2.697594501718213e-05, |
| "epoch": 332.64, |
| "step": 1330 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.0393955297768116, |
| "learning_rate": 2.6689576174112258e-05, |
| "epoch": 333.96, |
| "step": 1335 |
| }, |
| { |
| "loss": 0.0236, |
| "grad_norm": 0.09097255021333694, |
| "learning_rate": 2.6403207331042384e-05, |
| "epoch": 335.0, |
| "step": 1340 |
| }, |
| { |
| "loss": 0.0264, |
| "grad_norm": 0.02950783632695675, |
| "learning_rate": 2.611683848797251e-05, |
| "epoch": 336.32, |
| "step": 1345 |
| }, |
| { |
| "loss": 0.0242, |
| "grad_norm": 0.03414856642484665, |
| "learning_rate": 2.5830469644902637e-05, |
| "epoch": 337.64, |
| "step": 1350 |
| }, |
| { |
| "loss": 0.0258, |
| "grad_norm": 0.0349162220954895, |
| "learning_rate": 2.554410080183276e-05, |
| "epoch": 338.96, |
| "step": 1355 |
| }, |
| { |
| "loss": 0.0233, |
| "grad_norm": 0.07821868360042572, |
| "learning_rate": 2.5257731958762887e-05, |
| "epoch": 340.0, |
| "step": 1360 |
| }, |
| { |
| "loss": 0.0241, |
| "grad_norm": 0.026167893782258034, |
| "learning_rate": 2.4971363115693013e-05, |
| "epoch": 341.32, |
| "step": 1365 |
| }, |
| { |
| "loss": 0.0288, |
| "grad_norm": 0.04508533701300621, |
| "learning_rate": 2.468499427262314e-05, |
| "epoch": 342.64, |
| "step": 1370 |
| }, |
| { |
| "loss": 0.0261, |
| "grad_norm": 0.027592236176133156, |
| "learning_rate": 2.4398625429553266e-05, |
| "epoch": 343.96, |
| "step": 1375 |
| }, |
| { |
| "loss": 0.0246, |
| "grad_norm": 0.0881950631737709, |
| "learning_rate": 2.4112256586483393e-05, |
| "epoch": 345.0, |
| "step": 1380 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.027164354920387268, |
| "learning_rate": 2.3825887743413516e-05, |
| "epoch": 346.32, |
| "step": 1385 |
| }, |
| { |
| "loss": 0.0254, |
| "grad_norm": 0.028504472225904465, |
| "learning_rate": 2.3539518900343642e-05, |
| "epoch": 347.64, |
| "step": 1390 |
| }, |
| { |
| "loss": 0.0258, |
| "grad_norm": 0.02704612910747528, |
| "learning_rate": 2.3253150057273772e-05, |
| "epoch": 348.96, |
| "step": 1395 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.10028278827667236, |
| "learning_rate": 2.2966781214203895e-05, |
| "epoch": 350.0, |
| "step": 1400 |
| }, |
| { |
| "loss": 0.0241, |
| "grad_norm": 0.02599525637924671, |
| "learning_rate": 2.268041237113402e-05, |
| "epoch": 351.32, |
| "step": 1405 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.041026365011930466, |
| "learning_rate": 2.2394043528064148e-05, |
| "epoch": 352.64, |
| "step": 1410 |
| }, |
| { |
| "loss": 0.0252, |
| "grad_norm": 0.03604757413268089, |
| "learning_rate": 2.210767468499427e-05, |
| "epoch": 353.96, |
| "step": 1415 |
| }, |
| { |
| "loss": 0.0257, |
| "grad_norm": 0.08558797836303711, |
| "learning_rate": 2.18213058419244e-05, |
| "epoch": 355.0, |
| "step": 1420 |
| }, |
| { |
| "loss": 0.0287, |
| "grad_norm": 0.03581983968615532, |
| "learning_rate": 2.1534936998854528e-05, |
| "epoch": 356.32, |
| "step": 1425 |
| }, |
| { |
| "loss": 0.0256, |
| "grad_norm": 0.028226159512996674, |
| "learning_rate": 2.124856815578465e-05, |
| "epoch": 357.64, |
| "step": 1430 |
| }, |
| { |
| "loss": 0.0256, |
| "grad_norm": 0.03068063221871853, |
| "learning_rate": 2.0962199312714777e-05, |
| "epoch": 358.96, |
| "step": 1435 |
| }, |
| { |
| "loss": 0.0288, |
| "grad_norm": 0.10401294380426407, |
| "learning_rate": 2.0675830469644904e-05, |
| "epoch": 360.0, |
| "step": 1440 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.029932934790849686, |
| "learning_rate": 2.038946162657503e-05, |
| "epoch": 361.32, |
| "step": 1445 |
| }, |
| { |
| "loss": 0.0251, |
| "grad_norm": 0.03073325753211975, |
| "learning_rate": 2.0103092783505157e-05, |
| "epoch": 362.64, |
| "step": 1450 |
| }, |
| { |
| "loss": 0.0279, |
| "grad_norm": 0.028144460171461105, |
| "learning_rate": 1.981672394043528e-05, |
| "epoch": 363.96, |
| "step": 1455 |
| }, |
| { |
| "loss": 0.0263, |
| "grad_norm": 0.10323811322450638, |
| "learning_rate": 1.9530355097365406e-05, |
| "epoch": 365.0, |
| "step": 1460 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.02789253555238247, |
| "learning_rate": 1.9243986254295536e-05, |
| "epoch": 366.32, |
| "step": 1465 |
| }, |
| { |
| "loss": 0.0264, |
| "grad_norm": 0.027953842654824257, |
| "learning_rate": 1.895761741122566e-05, |
| "epoch": 367.64, |
| "step": 1470 |
| }, |
| { |
| "loss": 0.0267, |
| "grad_norm": 0.033589918166399, |
| "learning_rate": 1.8671248568155786e-05, |
| "epoch": 368.96, |
| "step": 1475 |
| }, |
| { |
| "loss": 0.025, |
| "grad_norm": 0.08682391792535782, |
| "learning_rate": 1.8384879725085912e-05, |
| "epoch": 370.0, |
| "step": 1480 |
| }, |
| { |
| "loss": 0.0253, |
| "grad_norm": 0.027875252068042755, |
| "learning_rate": 1.809851088201604e-05, |
| "epoch": 371.32, |
| "step": 1485 |
| }, |
| { |
| "loss": 0.0251, |
| "grad_norm": 0.02853395976126194, |
| "learning_rate": 1.7812142038946165e-05, |
| "epoch": 372.64, |
| "step": 1490 |
| }, |
| { |
| "loss": 0.0263, |
| "grad_norm": 0.03265855088829994, |
| "learning_rate": 1.7525773195876288e-05, |
| "epoch": 373.96, |
| "step": 1495 |
| }, |
| { |
| "loss": 0.0281, |
| "grad_norm": 0.14788594841957092, |
| "learning_rate": 1.7239404352806415e-05, |
| "epoch": 375.0, |
| "step": 1500 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.032310985028743744, |
| "learning_rate": 1.695303550973654e-05, |
| "epoch": 376.32, |
| "step": 1505 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.026569277048110962, |
| "learning_rate": 1.6666666666666667e-05, |
| "epoch": 377.64, |
| "step": 1510 |
| }, |
| { |
| "loss": 0.0254, |
| "grad_norm": 0.032844848930835724, |
| "learning_rate": 1.6380297823596794e-05, |
| "epoch": 378.96, |
| "step": 1515 |
| }, |
| { |
| "loss": 0.0284, |
| "grad_norm": 0.1191343441605568, |
| "learning_rate": 1.609392898052692e-05, |
| "epoch": 380.0, |
| "step": 1520 |
| }, |
| { |
| "loss": 0.026, |
| "grad_norm": 0.040164776146411896, |
| "learning_rate": 1.5807560137457044e-05, |
| "epoch": 381.32, |
| "step": 1525 |
| }, |
| { |
| "loss": 0.0284, |
| "grad_norm": 0.03208829462528229, |
| "learning_rate": 1.5521191294387173e-05, |
| "epoch": 382.64, |
| "step": 1530 |
| }, |
| { |
| "loss": 0.0279, |
| "grad_norm": 0.030518101528286934, |
| "learning_rate": 1.5234822451317298e-05, |
| "epoch": 383.96, |
| "step": 1535 |
| }, |
| { |
| "loss": 0.0282, |
| "grad_norm": 0.09137098491191864, |
| "learning_rate": 1.4948453608247423e-05, |
| "epoch": 385.0, |
| "step": 1540 |
| }, |
| { |
| "loss": 0.0259, |
| "grad_norm": 0.029169032350182533, |
| "learning_rate": 1.466208476517755e-05, |
| "epoch": 386.32, |
| "step": 1545 |
| }, |
| { |
| "loss": 0.0256, |
| "grad_norm": 0.02922474592924118, |
| "learning_rate": 1.4375715922107674e-05, |
| "epoch": 387.64, |
| "step": 1550 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.03208035230636597, |
| "learning_rate": 1.40893470790378e-05, |
| "epoch": 388.96, |
| "step": 1555 |
| }, |
| { |
| "loss": 0.0262, |
| "grad_norm": 0.11002888530492783, |
| "learning_rate": 1.3802978235967929e-05, |
| "epoch": 390.0, |
| "step": 1560 |
| }, |
| { |
| "loss": 0.0253, |
| "grad_norm": 0.02571057341992855, |
| "learning_rate": 1.3516609392898052e-05, |
| "epoch": 391.32, |
| "step": 1565 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.033123720437288284, |
| "learning_rate": 1.323024054982818e-05, |
| "epoch": 392.64, |
| "step": 1570 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.0276569165289402, |
| "learning_rate": 1.2943871706758307e-05, |
| "epoch": 393.96, |
| "step": 1575 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.09020277112722397, |
| "learning_rate": 1.2657502863688431e-05, |
| "epoch": 395.0, |
| "step": 1580 |
| }, |
| { |
| "loss": 0.0246, |
| "grad_norm": 0.02700515277683735, |
| "learning_rate": 1.2371134020618558e-05, |
| "epoch": 396.32, |
| "step": 1585 |
| }, |
| { |
| "loss": 0.0277, |
| "grad_norm": 0.030765585601329803, |
| "learning_rate": 1.2084765177548683e-05, |
| "epoch": 397.64, |
| "step": 1590 |
| }, |
| { |
| "loss": 0.0282, |
| "grad_norm": 0.028774507343769073, |
| "learning_rate": 1.1798396334478809e-05, |
| "epoch": 398.96, |
| "step": 1595 |
| }, |
| { |
| "loss": 0.0267, |
| "grad_norm": 0.09411438554525375, |
| "learning_rate": 1.1512027491408934e-05, |
| "epoch": 400.0, |
| "step": 1600 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.030230769887566566, |
| "learning_rate": 1.1225658648339062e-05, |
| "epoch": 401.32, |
| "step": 1605 |
| }, |
| { |
| "loss": 0.0257, |
| "grad_norm": 0.02960030548274517, |
| "learning_rate": 1.0939289805269187e-05, |
| "epoch": 402.64, |
| "step": 1610 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.03375796973705292, |
| "learning_rate": 1.0652920962199313e-05, |
| "epoch": 403.96, |
| "step": 1615 |
| }, |
| { |
| "loss": 0.0224, |
| "grad_norm": 0.0628446713089943, |
| "learning_rate": 1.036655211912944e-05, |
| "epoch": 405.0, |
| "step": 1620 |
| }, |
| { |
| "loss": 0.0256, |
| "grad_norm": 0.028491677716374397, |
| "learning_rate": 1.0080183276059566e-05, |
| "epoch": 406.32, |
| "step": 1625 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.027530612424016, |
| "learning_rate": 9.793814432989691e-06, |
| "epoch": 407.64, |
| "step": 1630 |
| }, |
| { |
| "loss": 0.0286, |
| "grad_norm": 0.02860579453408718, |
| "learning_rate": 9.507445589919818e-06, |
| "epoch": 408.96, |
| "step": 1635 |
| }, |
| { |
| "loss": 0.0274, |
| "grad_norm": 0.1014900878071785, |
| "learning_rate": 9.221076746849944e-06, |
| "epoch": 410.0, |
| "step": 1640 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.032212644815444946, |
| "learning_rate": 8.934707903780069e-06, |
| "epoch": 411.32, |
| "step": 1645 |
| }, |
| { |
| "loss": 0.0275, |
| "grad_norm": 0.028369037434458733, |
| "learning_rate": 8.648339060710195e-06, |
| "epoch": 412.64, |
| "step": 1650 |
| }, |
| { |
| "loss": 0.0247, |
| "grad_norm": 0.03228297829627991, |
| "learning_rate": 8.36197021764032e-06, |
| "epoch": 413.96, |
| "step": 1655 |
| }, |
| { |
| "loss": 0.0271, |
| "grad_norm": 0.10698091983795166, |
| "learning_rate": 8.075601374570448e-06, |
| "epoch": 415.0, |
| "step": 1660 |
| }, |
| { |
| "loss": 0.0253, |
| "grad_norm": 0.031531911343336105, |
| "learning_rate": 7.789232531500573e-06, |
| "epoch": 416.32, |
| "step": 1665 |
| }, |
| { |
| "loss": 0.0265, |
| "grad_norm": 0.029902342706918716, |
| "learning_rate": 7.502863688430699e-06, |
| "epoch": 417.64, |
| "step": 1670 |
| }, |
| { |
| "loss": 0.0275, |
| "grad_norm": 0.027522563934326172, |
| "learning_rate": 7.216494845360824e-06, |
| "epoch": 418.96, |
| "step": 1675 |
| }, |
| { |
| "loss": 0.0269, |
| "grad_norm": 0.11039866507053375, |
| "learning_rate": 6.930126002290952e-06, |
| "epoch": 420.0, |
| "step": 1680 |
| }, |
| { |
| "loss": 0.0283, |
| "grad_norm": 0.028080735355615616, |
| "learning_rate": 6.643757159221077e-06, |
| "epoch": 421.32, |
| "step": 1685 |
| }, |
| { |
| "loss": 0.0276, |
| "grad_norm": 0.033488210290670395, |
| "learning_rate": 6.357388316151203e-06, |
| "epoch": 422.64, |
| "step": 1690 |
| }, |
| { |
| "loss": 0.0257, |
| "grad_norm": 0.029904184862971306, |
| "learning_rate": 6.071019473081329e-06, |
| "epoch": 423.96, |
| "step": 1695 |
| }, |
| { |
| "loss": 0.0267, |
| "grad_norm": 0.11767726391553879, |
| "learning_rate": 5.784650630011455e-06, |
| "epoch": 425.0, |
| "step": 1700 |
| }, |
| { |
| "loss": 0.0272, |
| "grad_norm": 0.03559400141239166, |
| "learning_rate": 5.498281786941581e-06, |
| "epoch": 426.32, |
| "step": 1705 |
| }, |
| { |
| "loss": 0.0242, |
| "grad_norm": 0.02916187420487404, |
| "learning_rate": 5.211912943871707e-06, |
| "epoch": 427.64, |
| "step": 1710 |
| }, |
| { |
| "loss": 0.0265, |
| "grad_norm": 0.029780741780996323, |
| "learning_rate": 4.925544100801833e-06, |
| "epoch": 428.96, |
| "step": 1715 |
| }, |
| { |
| "loss": 0.0261, |
| "grad_norm": 0.0798487439751625, |
| "learning_rate": 4.639175257731959e-06, |
| "epoch": 430.0, |
| "step": 1720 |
| }, |
| { |
| "loss": 0.0258, |
| "grad_norm": 0.03286178782582283, |
| "learning_rate": 4.352806414662085e-06, |
| "epoch": 431.32, |
| "step": 1725 |
| }, |
| { |
| "loss": 0.028, |
| "grad_norm": 0.02974529378116131, |
| "learning_rate": 4.066437571592211e-06, |
| "epoch": 432.64, |
| "step": 1730 |
| }, |
| { |
| "loss": 0.0241, |
| "grad_norm": 0.030569862574338913, |
| "learning_rate": 3.7800687285223365e-06, |
| "epoch": 433.96, |
| "step": 1735 |
| }, |
| { |
| "loss": 0.0278, |
| "grad_norm": 0.08211632072925568, |
| "learning_rate": 3.493699885452463e-06, |
| "epoch": 435.0, |
| "step": 1740 |
| }, |
| { |
| "loss": 0.0295, |
| "grad_norm": 0.03577543795108795, |
| "learning_rate": 3.2073310423825886e-06, |
| "epoch": 436.32, |
| "step": 1745 |
| }, |
| { |
| "loss": 0.0245, |
| "grad_norm": 0.02749614790081978, |
| "learning_rate": 2.920962199312715e-06, |
| "epoch": 437.64, |
| "step": 1750 |
| }, |
| { |
| "loss": 0.027, |
| "grad_norm": 0.031330712139606476, |
| "learning_rate": 2.6345933562428407e-06, |
| "epoch": 438.96, |
| "step": 1755 |
| }, |
| { |
| "loss": 0.0273, |
| "grad_norm": 0.10281779617071152, |
| "learning_rate": 2.3482245131729668e-06, |
| "epoch": 440.0, |
| "step": 1760 |
| }, |
| { |
| "loss": 0.0287, |
| "grad_norm": 0.038072314113378525, |
| "learning_rate": 2.061855670103093e-06, |
| "epoch": 441.32, |
| "step": 1765 |
| }, |
| { |
| "loss": 0.0251, |
| "grad_norm": 0.02888447605073452, |
| "learning_rate": 1.7754868270332189e-06, |
| "epoch": 442.64, |
| "step": 1770 |
| }, |
| { |
| "loss": 0.0254, |
| "grad_norm": 0.034519221633672714, |
| "learning_rate": 1.4891179839633447e-06, |
| "epoch": 443.96, |
| "step": 1775 |
| }, |
| { |
| "loss": 0.0245, |
| "grad_norm": 0.07151661068201065, |
| "learning_rate": 1.202749140893471e-06, |
| "epoch": 445.0, |
| "step": 1780 |
| }, |
| { |
| "loss": 0.0248, |
| "grad_norm": 0.023701045662164688, |
| "learning_rate": 9.163802978235968e-07, |
| "epoch": 446.32, |
| "step": 1785 |
| }, |
| { |
| "loss": 0.031, |
| "grad_norm": 0.02814486064016819, |
| "learning_rate": 6.300114547537229e-07, |
| "epoch": 447.64, |
| "step": 1790 |
| }, |
| { |
| "loss": 0.0255, |
| "grad_norm": 0.028605803847312927, |
| "learning_rate": 3.436426116838488e-07, |
| "epoch": 448.96, |
| "step": 1795 |
| }, |
| { |
| "loss": 0.0257, |
| "grad_norm": 0.08770228177309036, |
| "learning_rate": 5.72737686139748e-08, |
| "epoch": 450.0, |
| "step": 1800 |
| }, |
| { |
| "train_runtime": 30712.13, |
| "train_samples_per_second": 1.875, |
| "train_steps_per_second": 0.059, |
| "total_flos": 1.5308141101056e+18, |
| "train_loss": 0.051218728774951566, |
| "epoch": 450.0, |
| "step": 1800 |
| } |
| ], |
| "training_args": { |
| "output_dir": "/hkfs/work/workspace/scratch/tum_fmp0582-dndworkspace/自己训练lora/train_lora/outputs/Mu-Math/group_09/checkpoints", |
| "overwrite_output_dir": false, |
| "do_train": false, |
| "do_eval": false, |
| "do_predict": false, |
| "eval_strategy": "no", |
| "prediction_loss_only": false, |
| "per_device_train_batch_size": 2, |
| "per_device_eval_batch_size": 8, |
| "per_gpu_train_batch_size": null, |
| "per_gpu_eval_batch_size": null, |
| "gradient_accumulation_steps": 16, |
| "eval_accumulation_steps": null, |
| "eval_delay": 0, |
| "torch_empty_cache_steps": null, |
| "learning_rate": 0.0001, |
| "weight_decay": 0.01, |
| "adam_beta1": 0.9, |
| "adam_beta2": 0.999, |
| "adam_epsilon": 1e-08, |
| "max_grad_norm": 1.0, |
| "num_train_epochs": 12, |
| "max_steps": 1800, |
| "lr_scheduler_type": "linear", |
| "lr_scheduler_kwargs": {}, |
| "warmup_ratio": 0.03, |
| "warmup_steps": 0, |
| "log_level": "passive", |
| "log_level_replica": "warning", |
| "log_on_each_node": true, |
| "logging_dir": "/hkfs/work/workspace/scratch/tum_fmp0582-dndworkspace/自己训练lora/train_lora/logs/Mu-Math/group_09", |
| "logging_strategy": "steps", |
| "logging_first_step": true, |
| "logging_steps": 5, |
| "logging_nan_inf_filter": true, |
| "save_strategy": "steps", |
| "save_steps": 300, |
| "save_total_limit": 6, |
| "save_safetensors": true, |
| "save_on_each_node": false, |
| "save_only_model": false, |
| "restore_callback_states_from_checkpoint": false, |
| "no_cuda": false, |
| "use_cpu": false, |
| "use_mps_device": false, |
| "seed": 42, |
| "data_seed": null, |
| "jit_mode_eval": false, |
| "bf16": true, |
| "fp16": false, |
| "fp16_opt_level": "O1", |
| "half_precision_backend": "auto", |
| "bf16_full_eval": false, |
| "fp16_full_eval": false, |
| "tf32": null, |
| "local_rank": 0, |
| "ddp_backend": null, |
| "tpu_num_cores": null, |
| "tpu_metrics_debug": false, |
| "debug": [], |
| "dataloader_drop_last": false, |
| "eval_steps": null, |
| "dataloader_num_workers": 0, |
| "dataloader_prefetch_factor": null, |
| "past_index": -1, |
| "run_name": null, |
| "disable_tqdm": false, |
| "remove_unused_columns": true, |
| "label_names": null, |
| "load_best_model_at_end": false, |
| "metric_for_best_model": null, |
| "greater_is_better": null, |
| "ignore_data_skip": false, |
| "fsdp": [], |
| "fsdp_min_num_params": 0, |
| "fsdp_config": { |
| "min_num_params": 0, |
| "xla": false, |
| "xla_fsdp_v2": false, |
| "xla_fsdp_grad_ckpt": false |
| }, |
| "fsdp_transformer_layer_cls_to_wrap": null, |
| "accelerator_config": { |
| "split_batches": false, |
| "dispatch_batches": null, |
| "even_batches": true, |
| "use_seedable_sampler": true, |
| "non_blocking": false, |
| "gradient_accumulation_kwargs": null |
| }, |
| "parallelism_config": null, |
| "deepspeed": null, |
| "label_smoothing_factor": 0.0, |
| "optim": "adamw_torch", |
| "optim_args": null, |
| "adafactor": false, |
| "group_by_length": false, |
| "length_column_name": "length", |
| "report_to": [], |
| "project": "huggingface", |
| "trackio_space_id": "trackio", |
| "ddp_find_unused_parameters": null, |
| "ddp_bucket_cap_mb": null, |
| "ddp_broadcast_buffers": null, |
| "dataloader_pin_memory": true, |
| "dataloader_persistent_workers": false, |
| "skip_memory_metrics": true, |
| "use_legacy_prediction_loop": false, |
| "push_to_hub": false, |
| "resume_from_checkpoint": null, |
| "hub_model_id": null, |
| "hub_strategy": "every_save", |
| "hub_token": "<HUB_TOKEN>", |
| "hub_private_repo": null, |
| "hub_always_push": false, |
| "hub_revision": null, |
| "gradient_checkpointing": true, |
| "gradient_checkpointing_kwargs": null, |
| "include_inputs_for_metrics": false, |
| "include_for_metrics": [], |
| "eval_do_concat_batches": true, |
| "fp16_backend": "auto", |
| "push_to_hub_model_id": null, |
| "push_to_hub_organization": null, |
| "push_to_hub_token": "<PUSH_TO_HUB_TOKEN>", |
| "mp_parameters": "", |
| "auto_find_batch_size": false, |
| "full_determinism": false, |
| "torchdynamo": null, |
| "ray_scope": "last", |
| "ddp_timeout": 1800, |
| "torch_compile": false, |
| "torch_compile_backend": null, |
| "torch_compile_mode": null, |
| "include_tokens_per_second": false, |
| "include_num_input_tokens_seen": "no", |
| "neftune_noise_alpha": null, |
| "optim_target_modules": null, |
| "batch_eval_metrics": false, |
| "eval_on_start": false, |
| "use_liger_kernel": false, |
| "liger_kernel_config": null, |
| "eval_use_gather_object": false, |
| "average_tokens_across_devices": true |
| }, |
| "lora_config": { |
| "r": 64, |
| "alpha": 128, |
| "dropout": 0.05, |
| "target_modules": [ |
| "q_proj", |
| "k_proj", |
| "v_proj", |
| "o_proj", |
| "gate_proj", |
| "up_proj", |
| "down_proj" |
| ] |
| }, |
| "effective_batch_size": 32, |
| "world_size": 1, |
| "git_commit": "" |
| } |
