| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 10.0, |
| "eval_steps": 500, |
| "global_step": 2560, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0390625, |
| "grad_norm": 8.492419242858887, |
| "learning_rate": 7.8125e-06, |
| "loss": 1.3342, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.078125, |
| "grad_norm": 3.8445992469787598, |
| "learning_rate": 1.5625e-05, |
| "loss": 1.0352, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.1171875, |
| "grad_norm": 2.211190700531006, |
| "learning_rate": 2.34375e-05, |
| "loss": 0.4873, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.15625, |
| "grad_norm": 1.9646821022033691, |
| "learning_rate": 3.125e-05, |
| "loss": 0.3898, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.1953125, |
| "grad_norm": 1.9914284944534302, |
| "learning_rate": 3.90625e-05, |
| "loss": 0.2554, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.234375, |
| "grad_norm": 0.9096956253051758, |
| "learning_rate": 4.6875e-05, |
| "loss": 0.1924, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.2734375, |
| "grad_norm": 0.6492406129837036, |
| "learning_rate": 5.46875e-05, |
| "loss": 0.1721, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.3125, |
| "grad_norm": 0.8035663962364197, |
| "learning_rate": 6.25e-05, |
| "loss": 0.1454, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.3515625, |
| "grad_norm": 0.5276547074317932, |
| "learning_rate": 7.031250000000001e-05, |
| "loss": 0.1343, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.390625, |
| "grad_norm": 0.735613226890564, |
| "learning_rate": 7.8125e-05, |
| "loss": 0.1318, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.4296875, |
| "grad_norm": 1.0286799669265747, |
| "learning_rate": 8.593750000000001e-05, |
| "loss": 0.111, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.46875, |
| "grad_norm": 0.6054307222366333, |
| "learning_rate": 9.375e-05, |
| "loss": 0.1102, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.5078125, |
| "grad_norm": 0.6528738737106323, |
| "learning_rate": 9.999983313226824e-05, |
| "loss": 0.1024, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.546875, |
| "grad_norm": 0.522591769695282, |
| "learning_rate": 9.999399287860388e-05, |
| "loss": 0.0946, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.5859375, |
| "grad_norm": 0.7753025889396667, |
| "learning_rate": 9.997981035211134e-05, |
| "loss": 0.093, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.625, |
| "grad_norm": 0.58796226978302, |
| "learning_rate": 9.995728791936504e-05, |
| "loss": 0.0881, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.6640625, |
| "grad_norm": 0.5107079744338989, |
| "learning_rate": 9.992642933858212e-05, |
| "loss": 0.0807, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.703125, |
| "grad_norm": 0.9606569409370422, |
| "learning_rate": 9.988723975899517e-05, |
| "loss": 0.0991, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.7421875, |
| "grad_norm": 0.7940914034843445, |
| "learning_rate": 9.983972571999318e-05, |
| "loss": 0.0858, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.78125, |
| "grad_norm": 0.805740237236023, |
| "learning_rate": 9.97838951500302e-05, |
| "loss": 0.0975, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.8203125, |
| "grad_norm": 0.8114387392997742, |
| "learning_rate": 9.971975736530243e-05, |
| "loss": 0.0815, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.859375, |
| "grad_norm": 0.5807178616523743, |
| "learning_rate": 9.96473230681937e-05, |
| "loss": 0.0768, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.8984375, |
| "grad_norm": 0.655323326587677, |
| "learning_rate": 9.956660434548948e-05, |
| "loss": 0.0858, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.9375, |
| "grad_norm": 0.513111412525177, |
| "learning_rate": 9.947761466636014e-05, |
| "loss": 0.0816, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.9765625, |
| "grad_norm": 0.7743618488311768, |
| "learning_rate": 9.938036888011337e-05, |
| "loss": 0.071, |
| "step": 250 |
| }, |
| { |
| "epoch": 1.015625, |
| "grad_norm": 0.45275968313217163, |
| "learning_rate": 9.92748832137163e-05, |
| "loss": 0.073, |
| "step": 260 |
| }, |
| { |
| "epoch": 1.0546875, |
| "grad_norm": 0.7138253450393677, |
| "learning_rate": 9.916117526908785e-05, |
| "loss": 0.0774, |
| "step": 270 |
| }, |
| { |
| "epoch": 1.09375, |
| "grad_norm": 0.8007267117500305, |
| "learning_rate": 9.903926402016153e-05, |
| "loss": 0.0721, |
| "step": 280 |
| }, |
| { |
| "epoch": 1.1328125, |
| "grad_norm": 0.6417205333709717, |
| "learning_rate": 9.890916980971936e-05, |
| "loss": 0.0692, |
| "step": 290 |
| }, |
| { |
| "epoch": 1.171875, |
| "grad_norm": 0.605466902256012, |
| "learning_rate": 9.877091434599734e-05, |
| "loss": 0.0728, |
| "step": 300 |
| }, |
| { |
| "epoch": 1.2109375, |
| "grad_norm": 0.6525840759277344, |
| "learning_rate": 9.862452069906317e-05, |
| "loss": 0.0758, |
| "step": 310 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 1.1812381744384766, |
| "learning_rate": 9.847001329696653e-05, |
| "loss": 0.0707, |
| "step": 320 |
| }, |
| { |
| "epoch": 1.2890625, |
| "grad_norm": 0.7314594388008118, |
| "learning_rate": 9.830741792166295e-05, |
| "loss": 0.0693, |
| "step": 330 |
| }, |
| { |
| "epoch": 1.328125, |
| "grad_norm": 0.721834659576416, |
| "learning_rate": 9.813676170471177e-05, |
| "loss": 0.0702, |
| "step": 340 |
| }, |
| { |
| "epoch": 1.3671875, |
| "grad_norm": 0.5326610207557678, |
| "learning_rate": 9.795807312274862e-05, |
| "loss": 0.0684, |
| "step": 350 |
| }, |
| { |
| "epoch": 1.40625, |
| "grad_norm": 0.6198597550392151, |
| "learning_rate": 9.777138199273384e-05, |
| "loss": 0.0675, |
| "step": 360 |
| }, |
| { |
| "epoch": 1.4453125, |
| "grad_norm": 0.5808058977127075, |
| "learning_rate": 9.757671946697698e-05, |
| "loss": 0.0665, |
| "step": 370 |
| }, |
| { |
| "epoch": 1.484375, |
| "grad_norm": 0.4104309678077698, |
| "learning_rate": 9.737411802793852e-05, |
| "loss": 0.0676, |
| "step": 380 |
| }, |
| { |
| "epoch": 1.5234375, |
| "grad_norm": 0.618438720703125, |
| "learning_rate": 9.716361148280977e-05, |
| "loss": 0.0615, |
| "step": 390 |
| }, |
| { |
| "epoch": 1.5625, |
| "grad_norm": 0.4504268765449524, |
| "learning_rate": 9.694523495787149e-05, |
| "loss": 0.0721, |
| "step": 400 |
| }, |
| { |
| "epoch": 1.6015625, |
| "grad_norm": 0.333695650100708, |
| "learning_rate": 9.671902489263261e-05, |
| "loss": 0.0579, |
| "step": 410 |
| }, |
| { |
| "epoch": 1.640625, |
| "grad_norm": 0.7360309362411499, |
| "learning_rate": 9.648501903374975e-05, |
| "loss": 0.0651, |
| "step": 420 |
| }, |
| { |
| "epoch": 1.6796875, |
| "grad_norm": 0.4267025291919708, |
| "learning_rate": 9.624325642872853e-05, |
| "loss": 0.057, |
| "step": 430 |
| }, |
| { |
| "epoch": 1.71875, |
| "grad_norm": 0.3154551684856415, |
| "learning_rate": 9.599377741940792e-05, |
| "loss": 0.0696, |
| "step": 440 |
| }, |
| { |
| "epoch": 1.7578125, |
| "grad_norm": 0.4576047658920288, |
| "learning_rate": 9.57366236352286e-05, |
| "loss": 0.0595, |
| "step": 450 |
| }, |
| { |
| "epoch": 1.796875, |
| "grad_norm": 0.5197241902351379, |
| "learning_rate": 9.547183798628641e-05, |
| "loss": 0.0661, |
| "step": 460 |
| }, |
| { |
| "epoch": 1.8359375, |
| "grad_norm": 0.42932412028312683, |
| "learning_rate": 9.519946465617218e-05, |
| "loss": 0.0609, |
| "step": 470 |
| }, |
| { |
| "epoch": 1.875, |
| "grad_norm": 0.4223524034023285, |
| "learning_rate": 9.491954909459895e-05, |
| "loss": 0.0574, |
| "step": 480 |
| }, |
| { |
| "epoch": 1.9140625, |
| "grad_norm": 0.47859442234039307, |
| "learning_rate": 9.463213800981805e-05, |
| "loss": 0.0601, |
| "step": 490 |
| }, |
| { |
| "epoch": 1.953125, |
| "grad_norm": 0.31995102763175964, |
| "learning_rate": 9.433727936082504e-05, |
| "loss": 0.0555, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.9921875, |
| "grad_norm": 0.3553142547607422, |
| "learning_rate": 9.403502234935706e-05, |
| "loss": 0.0582, |
| "step": 510 |
| }, |
| { |
| "epoch": 2.03125, |
| "grad_norm": 0.6797146201133728, |
| "learning_rate": 9.372541741168272e-05, |
| "loss": 0.0528, |
| "step": 520 |
| }, |
| { |
| "epoch": 2.0703125, |
| "grad_norm": 0.41948041319847107, |
| "learning_rate": 9.340851621018603e-05, |
| "loss": 0.0598, |
| "step": 530 |
| }, |
| { |
| "epoch": 2.109375, |
| "grad_norm": 0.5185118317604065, |
| "learning_rate": 9.308437162474579e-05, |
| "loss": 0.0573, |
| "step": 540 |
| }, |
| { |
| "epoch": 2.1484375, |
| "grad_norm": 0.5795381665229797, |
| "learning_rate": 9.275303774391165e-05, |
| "loss": 0.0582, |
| "step": 550 |
| }, |
| { |
| "epoch": 2.1875, |
| "grad_norm": 0.44867536425590515, |
| "learning_rate": 9.241456985587868e-05, |
| "loss": 0.0616, |
| "step": 560 |
| }, |
| { |
| "epoch": 2.2265625, |
| "grad_norm": 0.42551538348197937, |
| "learning_rate": 9.20690244392617e-05, |
| "loss": 0.0578, |
| "step": 570 |
| }, |
| { |
| "epoch": 2.265625, |
| "grad_norm": 0.6047021746635437, |
| "learning_rate": 9.171645915367088e-05, |
| "loss": 0.0523, |
| "step": 580 |
| }, |
| { |
| "epoch": 2.3046875, |
| "grad_norm": 0.5272088050842285, |
| "learning_rate": 9.135693283009035e-05, |
| "loss": 0.0648, |
| "step": 590 |
| }, |
| { |
| "epoch": 2.34375, |
| "grad_norm": 0.396763414144516, |
| "learning_rate": 9.099050546106133e-05, |
| "loss": 0.0543, |
| "step": 600 |
| }, |
| { |
| "epoch": 2.3828125, |
| "grad_norm": 0.4244937300682068, |
| "learning_rate": 9.061723819067154e-05, |
| "loss": 0.0506, |
| "step": 610 |
| }, |
| { |
| "epoch": 2.421875, |
| "grad_norm": 0.37396669387817383, |
| "learning_rate": 9.023719330435223e-05, |
| "loss": 0.0566, |
| "step": 620 |
| }, |
| { |
| "epoch": 2.4609375, |
| "grad_norm": 0.3504962623119354, |
| "learning_rate": 8.9850434218485e-05, |
| "loss": 0.0586, |
| "step": 630 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 0.5627210736274719, |
| "learning_rate": 8.945702546981969e-05, |
| "loss": 0.053, |
| "step": 640 |
| }, |
| { |
| "epoch": 2.5390625, |
| "grad_norm": 0.6733695268630981, |
| "learning_rate": 8.905703270470549e-05, |
| "loss": 0.0566, |
| "step": 650 |
| }, |
| { |
| "epoch": 2.578125, |
| "grad_norm": 0.28862297534942627, |
| "learning_rate": 8.865052266813685e-05, |
| "loss": 0.0484, |
| "step": 660 |
| }, |
| { |
| "epoch": 2.6171875, |
| "grad_norm": 0.9207034111022949, |
| "learning_rate": 8.823756319261597e-05, |
| "loss": 0.0556, |
| "step": 670 |
| }, |
| { |
| "epoch": 2.65625, |
| "grad_norm": 0.5370097756385803, |
| "learning_rate": 8.781822318683394e-05, |
| "loss": 0.0533, |
| "step": 680 |
| }, |
| { |
| "epoch": 2.6953125, |
| "grad_norm": 0.3493472635746002, |
| "learning_rate": 8.739257262417225e-05, |
| "loss": 0.0539, |
| "step": 690 |
| }, |
| { |
| "epoch": 2.734375, |
| "grad_norm": 0.36126288771629333, |
| "learning_rate": 8.696068253102665e-05, |
| "loss": 0.0557, |
| "step": 700 |
| }, |
| { |
| "epoch": 2.7734375, |
| "grad_norm": 0.331190824508667, |
| "learning_rate": 8.652262497495527e-05, |
| "loss": 0.0514, |
| "step": 710 |
| }, |
| { |
| "epoch": 2.8125, |
| "grad_norm": 0.4170578420162201, |
| "learning_rate": 8.60784730526531e-05, |
| "loss": 0.0469, |
| "step": 720 |
| }, |
| { |
| "epoch": 2.8515625, |
| "grad_norm": 0.35276809334754944, |
| "learning_rate": 8.56283008777546e-05, |
| "loss": 0.0541, |
| "step": 730 |
| }, |
| { |
| "epoch": 2.890625, |
| "grad_norm": 0.5037612318992615, |
| "learning_rate": 8.517218356846665e-05, |
| "loss": 0.0493, |
| "step": 740 |
| }, |
| { |
| "epoch": 2.9296875, |
| "grad_norm": 0.3002952039241791, |
| "learning_rate": 8.471019723503411e-05, |
| "loss": 0.0526, |
| "step": 750 |
| }, |
| { |
| "epoch": 2.96875, |
| "grad_norm": 0.33518242835998535, |
| "learning_rate": 8.424241896703936e-05, |
| "loss": 0.0522, |
| "step": 760 |
| }, |
| { |
| "epoch": 3.0078125, |
| "grad_norm": 0.4655015766620636, |
| "learning_rate": 8.376892682053898e-05, |
| "loss": 0.0504, |
| "step": 770 |
| }, |
| { |
| "epoch": 3.046875, |
| "grad_norm": 0.537617564201355, |
| "learning_rate": 8.328979980503873e-05, |
| "loss": 0.0543, |
| "step": 780 |
| }, |
| { |
| "epoch": 3.0859375, |
| "grad_norm": 0.38395699858665466, |
| "learning_rate": 8.280511787030963e-05, |
| "loss": 0.0534, |
| "step": 790 |
| }, |
| { |
| "epoch": 3.125, |
| "grad_norm": 0.44339320063591003, |
| "learning_rate": 8.231496189304704e-05, |
| "loss": 0.0474, |
| "step": 800 |
| }, |
| { |
| "epoch": 3.1640625, |
| "grad_norm": 0.49106666445732117, |
| "learning_rate": 8.18194136633753e-05, |
| "loss": 0.0523, |
| "step": 810 |
| }, |
| { |
| "epoch": 3.203125, |
| "grad_norm": 0.3647231161594391, |
| "learning_rate": 8.131855587119957e-05, |
| "loss": 0.0515, |
| "step": 820 |
| }, |
| { |
| "epoch": 3.2421875, |
| "grad_norm": 0.43639010190963745, |
| "learning_rate": 8.081247209240788e-05, |
| "loss": 0.0503, |
| "step": 830 |
| }, |
| { |
| "epoch": 3.28125, |
| "grad_norm": 0.3658983111381531, |
| "learning_rate": 8.030124677492512e-05, |
| "loss": 0.053, |
| "step": 840 |
| }, |
| { |
| "epoch": 3.3203125, |
| "grad_norm": 0.23427674174308777, |
| "learning_rate": 7.978496522462167e-05, |
| "loss": 0.0497, |
| "step": 850 |
| }, |
| { |
| "epoch": 3.359375, |
| "grad_norm": 0.3150970935821533, |
| "learning_rate": 7.926371359107871e-05, |
| "loss": 0.0468, |
| "step": 860 |
| }, |
| { |
| "epoch": 3.3984375, |
| "grad_norm": 0.4161214828491211, |
| "learning_rate": 7.873757885321291e-05, |
| "loss": 0.0488, |
| "step": 870 |
| }, |
| { |
| "epoch": 3.4375, |
| "grad_norm": 0.27835017442703247, |
| "learning_rate": 7.820664880476256e-05, |
| "loss": 0.0449, |
| "step": 880 |
| }, |
| { |
| "epoch": 3.4765625, |
| "grad_norm": 0.31080129742622375, |
| "learning_rate": 7.767101203963786e-05, |
| "loss": 0.0459, |
| "step": 890 |
| }, |
| { |
| "epoch": 3.515625, |
| "grad_norm": 0.32351627945899963, |
| "learning_rate": 7.713075793713774e-05, |
| "loss": 0.0414, |
| "step": 900 |
| }, |
| { |
| "epoch": 3.5546875, |
| "grad_norm": 0.4844232499599457, |
| "learning_rate": 7.658597664703533e-05, |
| "loss": 0.0456, |
| "step": 910 |
| }, |
| { |
| "epoch": 3.59375, |
| "grad_norm": 0.49269959330558777, |
| "learning_rate": 7.603675907453534e-05, |
| "loss": 0.0443, |
| "step": 920 |
| }, |
| { |
| "epoch": 3.6328125, |
| "grad_norm": 0.29903239011764526, |
| "learning_rate": 7.548319686510487e-05, |
| "loss": 0.0481, |
| "step": 930 |
| }, |
| { |
| "epoch": 3.671875, |
| "grad_norm": 0.34350693225860596, |
| "learning_rate": 7.492538238918116e-05, |
| "loss": 0.0456, |
| "step": 940 |
| }, |
| { |
| "epoch": 3.7109375, |
| "grad_norm": 0.2822878956794739, |
| "learning_rate": 7.436340872675792e-05, |
| "loss": 0.0439, |
| "step": 950 |
| }, |
| { |
| "epoch": 3.75, |
| "grad_norm": 0.3564528226852417, |
| "learning_rate": 7.379736965185368e-05, |
| "loss": 0.0495, |
| "step": 960 |
| }, |
| { |
| "epoch": 3.7890625, |
| "grad_norm": 0.44323813915252686, |
| "learning_rate": 7.322735961686409e-05, |
| "loss": 0.0497, |
| "step": 970 |
| }, |
| { |
| "epoch": 3.828125, |
| "grad_norm": 0.4935294985771179, |
| "learning_rate": 7.265347373680102e-05, |
| "loss": 0.0474, |
| "step": 980 |
| }, |
| { |
| "epoch": 3.8671875, |
| "grad_norm": 0.3851205110549927, |
| "learning_rate": 7.20758077734212e-05, |
| "loss": 0.0475, |
| "step": 990 |
| }, |
| { |
| "epoch": 3.90625, |
| "grad_norm": 0.4426199495792389, |
| "learning_rate": 7.149445811924685e-05, |
| "loss": 0.0467, |
| "step": 1000 |
| }, |
| { |
| "epoch": 3.9453125, |
| "grad_norm": 0.4967394769191742, |
| "learning_rate": 7.090952178148112e-05, |
| "loss": 0.0508, |
| "step": 1010 |
| }, |
| { |
| "epoch": 3.984375, |
| "grad_norm": 0.3541552722454071, |
| "learning_rate": 7.032109636582096e-05, |
| "loss": 0.0457, |
| "step": 1020 |
| }, |
| { |
| "epoch": 4.0234375, |
| "grad_norm": 0.4096279740333557, |
| "learning_rate": 6.972928006017001e-05, |
| "loss": 0.0446, |
| "step": 1030 |
| }, |
| { |
| "epoch": 4.0625, |
| "grad_norm": 0.34364351630210876, |
| "learning_rate": 6.91341716182545e-05, |
| "loss": 0.0474, |
| "step": 1040 |
| }, |
| { |
| "epoch": 4.1015625, |
| "grad_norm": 0.7376656532287598, |
| "learning_rate": 6.853587034314463e-05, |
| "loss": 0.0452, |
| "step": 1050 |
| }, |
| { |
| "epoch": 4.140625, |
| "grad_norm": 0.46325570344924927, |
| "learning_rate": 6.793447607068437e-05, |
| "loss": 0.0426, |
| "step": 1060 |
| }, |
| { |
| "epoch": 4.1796875, |
| "grad_norm": 0.44051945209503174, |
| "learning_rate": 6.733008915283216e-05, |
| "loss": 0.0385, |
| "step": 1070 |
| }, |
| { |
| "epoch": 4.21875, |
| "grad_norm": 0.4191512167453766, |
| "learning_rate": 6.672281044091582e-05, |
| "loss": 0.0468, |
| "step": 1080 |
| }, |
| { |
| "epoch": 4.2578125, |
| "grad_norm": 0.23939189314842224, |
| "learning_rate": 6.611274126880378e-05, |
| "loss": 0.0417, |
| "step": 1090 |
| }, |
| { |
| "epoch": 4.296875, |
| "grad_norm": 0.2203720659017563, |
| "learning_rate": 6.549998343599601e-05, |
| "loss": 0.0457, |
| "step": 1100 |
| }, |
| { |
| "epoch": 4.3359375, |
| "grad_norm": 0.4468579888343811, |
| "learning_rate": 6.488463919063727e-05, |
| "loss": 0.0459, |
| "step": 1110 |
| }, |
| { |
| "epoch": 4.375, |
| "grad_norm": 0.39386874437332153, |
| "learning_rate": 6.426681121245527e-05, |
| "loss": 0.0439, |
| "step": 1120 |
| }, |
| { |
| "epoch": 4.4140625, |
| "grad_norm": 0.29376569390296936, |
| "learning_rate": 6.364660259562713e-05, |
| "loss": 0.0439, |
| "step": 1130 |
| }, |
| { |
| "epoch": 4.453125, |
| "grad_norm": 0.24146665632724762, |
| "learning_rate": 6.302411683157637e-05, |
| "loss": 0.0429, |
| "step": 1140 |
| }, |
| { |
| "epoch": 4.4921875, |
| "grad_norm": 0.3997792899608612, |
| "learning_rate": 6.239945779170389e-05, |
| "loss": 0.0463, |
| "step": 1150 |
| }, |
| { |
| "epoch": 4.53125, |
| "grad_norm": 0.302913635969162, |
| "learning_rate": 6.177272971005528e-05, |
| "loss": 0.0503, |
| "step": 1160 |
| }, |
| { |
| "epoch": 4.5703125, |
| "grad_norm": 0.4844031035900116, |
| "learning_rate": 6.114403716592794e-05, |
| "loss": 0.0489, |
| "step": 1170 |
| }, |
| { |
| "epoch": 4.609375, |
| "grad_norm": 0.34038305282592773, |
| "learning_rate": 6.051348506642017e-05, |
| "loss": 0.0436, |
| "step": 1180 |
| }, |
| { |
| "epoch": 4.6484375, |
| "grad_norm": 0.5346148610115051, |
| "learning_rate": 5.988117862892601e-05, |
| "loss": 0.0456, |
| "step": 1190 |
| }, |
| { |
| "epoch": 4.6875, |
| "grad_norm": 0.28704309463500977, |
| "learning_rate": 5.924722336357793e-05, |
| "loss": 0.0421, |
| "step": 1200 |
| }, |
| { |
| "epoch": 4.7265625, |
| "grad_norm": 0.2789936065673828, |
| "learning_rate": 5.861172505564091e-05, |
| "loss": 0.0414, |
| "step": 1210 |
| }, |
| { |
| "epoch": 4.765625, |
| "grad_norm": 0.29666388034820557, |
| "learning_rate": 5.797478974786055e-05, |
| "loss": 0.0476, |
| "step": 1220 |
| }, |
| { |
| "epoch": 4.8046875, |
| "grad_norm": 0.520137369632721, |
| "learning_rate": 5.733652372276809e-05, |
| "loss": 0.0399, |
| "step": 1230 |
| }, |
| { |
| "epoch": 4.84375, |
| "grad_norm": 0.29389891028404236, |
| "learning_rate": 5.669703348494564e-05, |
| "loss": 0.0426, |
| "step": 1240 |
| }, |
| { |
| "epoch": 4.8828125, |
| "grad_norm": 0.4732462465763092, |
| "learning_rate": 5.605642574325413e-05, |
| "loss": 0.0436, |
| "step": 1250 |
| }, |
| { |
| "epoch": 4.921875, |
| "grad_norm": 0.6422872543334961, |
| "learning_rate": 5.5414807393027376e-05, |
| "loss": 0.0385, |
| "step": 1260 |
| }, |
| { |
| "epoch": 4.9609375, |
| "grad_norm": 0.31786808371543884, |
| "learning_rate": 5.47722854982349e-05, |
| "loss": 0.0449, |
| "step": 1270 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 0.7411214113235474, |
| "learning_rate": 5.4128967273616625e-05, |
| "loss": 0.0379, |
| "step": 1280 |
| }, |
| { |
| "epoch": 5.0390625, |
| "grad_norm": 0.440328449010849, |
| "learning_rate": 5.348496006679244e-05, |
| "loss": 0.0404, |
| "step": 1290 |
| }, |
| { |
| "epoch": 5.078125, |
| "grad_norm": 0.40569090843200684, |
| "learning_rate": 5.284037134034964e-05, |
| "loss": 0.0481, |
| "step": 1300 |
| }, |
| { |
| "epoch": 5.1171875, |
| "grad_norm": 0.23646467924118042, |
| "learning_rate": 5.2195308653911035e-05, |
| "loss": 0.0365, |
| "step": 1310 |
| }, |
| { |
| "epoch": 5.15625, |
| "grad_norm": 0.41665926575660706, |
| "learning_rate": 5.154987964618704e-05, |
| "loss": 0.0364, |
| "step": 1320 |
| }, |
| { |
| "epoch": 5.1953125, |
| "grad_norm": 0.24862350523471832, |
| "learning_rate": 5.090419201701445e-05, |
| "loss": 0.041, |
| "step": 1330 |
| }, |
| { |
| "epoch": 5.234375, |
| "grad_norm": 0.3748071491718292, |
| "learning_rate": 5.025835350938506e-05, |
| "loss": 0.0477, |
| "step": 1340 |
| }, |
| { |
| "epoch": 5.2734375, |
| "grad_norm": 0.40436264872550964, |
| "learning_rate": 4.961247189146707e-05, |
| "loss": 0.0348, |
| "step": 1350 |
| }, |
| { |
| "epoch": 5.3125, |
| "grad_norm": 0.4232139587402344, |
| "learning_rate": 4.8966654938622295e-05, |
| "loss": 0.0367, |
| "step": 1360 |
| }, |
| { |
| "epoch": 5.3515625, |
| "grad_norm": 0.26344922184944153, |
| "learning_rate": 4.832101041542219e-05, |
| "loss": 0.0355, |
| "step": 1370 |
| }, |
| { |
| "epoch": 5.390625, |
| "grad_norm": 0.4235900044441223, |
| "learning_rate": 4.7675646057665644e-05, |
| "loss": 0.0432, |
| "step": 1380 |
| }, |
| { |
| "epoch": 5.4296875, |
| "grad_norm": 0.5052417516708374, |
| "learning_rate": 4.7030669554401596e-05, |
| "loss": 0.0353, |
| "step": 1390 |
| }, |
| { |
| "epoch": 5.46875, |
| "grad_norm": 0.38285690546035767, |
| "learning_rate": 4.638618852995947e-05, |
| "loss": 0.0401, |
| "step": 1400 |
| }, |
| { |
| "epoch": 5.5078125, |
| "grad_norm": 0.6338162422180176, |
| "learning_rate": 4.574231052599034e-05, |
| "loss": 0.0418, |
| "step": 1410 |
| }, |
| { |
| "epoch": 5.546875, |
| "grad_norm": 0.3230713903903961, |
| "learning_rate": 4.509914298352197e-05, |
| "loss": 0.0372, |
| "step": 1420 |
| }, |
| { |
| "epoch": 5.5859375, |
| "grad_norm": 0.3894837498664856, |
| "learning_rate": 4.445679322503063e-05, |
| "loss": 0.0387, |
| "step": 1430 |
| }, |
| { |
| "epoch": 5.625, |
| "grad_norm": 0.3093172609806061, |
| "learning_rate": 4.381536843653262e-05, |
| "loss": 0.0363, |
| "step": 1440 |
| }, |
| { |
| "epoch": 5.6640625, |
| "grad_norm": 0.4512802064418793, |
| "learning_rate": 4.317497564969868e-05, |
| "loss": 0.0382, |
| "step": 1450 |
| }, |
| { |
| "epoch": 5.703125, |
| "grad_norm": 0.3944973051548004, |
| "learning_rate": 4.253572172399407e-05, |
| "loss": 0.0413, |
| "step": 1460 |
| }, |
| { |
| "epoch": 5.7421875, |
| "grad_norm": 0.4658920466899872, |
| "learning_rate": 4.18977133288474e-05, |
| "loss": 0.0363, |
| "step": 1470 |
| }, |
| { |
| "epoch": 5.78125, |
| "grad_norm": 0.3259652554988861, |
| "learning_rate": 4.126105692585121e-05, |
| "loss": 0.0372, |
| "step": 1480 |
| }, |
| { |
| "epoch": 5.8203125, |
| "grad_norm": 0.3776373863220215, |
| "learning_rate": 4.0625858750997184e-05, |
| "loss": 0.0366, |
| "step": 1490 |
| }, |
| { |
| "epoch": 5.859375, |
| "grad_norm": 0.22152118384838104, |
| "learning_rate": 3.999222479694901e-05, |
| "loss": 0.038, |
| "step": 1500 |
| }, |
| { |
| "epoch": 5.8984375, |
| "grad_norm": 0.3621926009654999, |
| "learning_rate": 3.936026079535593e-05, |
| "loss": 0.0402, |
| "step": 1510 |
| }, |
| { |
| "epoch": 5.9375, |
| "grad_norm": 0.3338565230369568, |
| "learning_rate": 3.87300721992097e-05, |
| "loss": 0.0417, |
| "step": 1520 |
| }, |
| { |
| "epoch": 5.9765625, |
| "grad_norm": 0.5546092391014099, |
| "learning_rate": 3.810176416524812e-05, |
| "loss": 0.0357, |
| "step": 1530 |
| }, |
| { |
| "epoch": 6.015625, |
| "grad_norm": 0.8308889865875244, |
| "learning_rate": 3.747544153640809e-05, |
| "loss": 0.0382, |
| "step": 1540 |
| }, |
| { |
| "epoch": 6.0546875, |
| "grad_norm": 0.24918757379055023, |
| "learning_rate": 3.6851208824330844e-05, |
| "loss": 0.0361, |
| "step": 1550 |
| }, |
| { |
| "epoch": 6.09375, |
| "grad_norm": 0.5025043487548828, |
| "learning_rate": 3.622917019192261e-05, |
| "loss": 0.0385, |
| "step": 1560 |
| }, |
| { |
| "epoch": 6.1328125, |
| "grad_norm": 0.30080971121788025, |
| "learning_rate": 3.560942943597344e-05, |
| "loss": 0.0378, |
| "step": 1570 |
| }, |
| { |
| "epoch": 6.171875, |
| "grad_norm": 0.2558280825614929, |
| "learning_rate": 3.4992089969837014e-05, |
| "loss": 0.0334, |
| "step": 1580 |
| }, |
| { |
| "epoch": 6.2109375, |
| "grad_norm": 0.3407670259475708, |
| "learning_rate": 3.437725480617462e-05, |
| "loss": 0.0319, |
| "step": 1590 |
| }, |
| { |
| "epoch": 6.25, |
| "grad_norm": 0.33768805861473083, |
| "learning_rate": 3.3765026539765834e-05, |
| "loss": 0.0372, |
| "step": 1600 |
| }, |
| { |
| "epoch": 6.2890625, |
| "grad_norm": 0.3476319909095764, |
| "learning_rate": 3.3155507330389e-05, |
| "loss": 0.0369, |
| "step": 1610 |
| }, |
| { |
| "epoch": 6.328125, |
| "grad_norm": 0.45832690596580505, |
| "learning_rate": 3.254879888577432e-05, |
| "loss": 0.0308, |
| "step": 1620 |
| }, |
| { |
| "epoch": 6.3671875, |
| "grad_norm": 0.3107905387878418, |
| "learning_rate": 3.194500244463229e-05, |
| "loss": 0.0344, |
| "step": 1630 |
| }, |
| { |
| "epoch": 6.40625, |
| "grad_norm": 0.30494946241378784, |
| "learning_rate": 3.13442187597605e-05, |
| "loss": 0.0365, |
| "step": 1640 |
| }, |
| { |
| "epoch": 6.4453125, |
| "grad_norm": 0.4013335406780243, |
| "learning_rate": 3.074654808123143e-05, |
| "loss": 0.0357, |
| "step": 1650 |
| }, |
| { |
| "epoch": 6.484375, |
| "grad_norm": 0.49742478132247925, |
| "learning_rate": 3.0152090139664174e-05, |
| "loss": 0.0339, |
| "step": 1660 |
| }, |
| { |
| "epoch": 6.5234375, |
| "grad_norm": 0.3577755391597748, |
| "learning_rate": 2.9560944129582896e-05, |
| "loss": 0.0399, |
| "step": 1670 |
| }, |
| { |
| "epoch": 6.5625, |
| "grad_norm": 0.40309587121009827, |
| "learning_rate": 2.8973208692864624e-05, |
| "loss": 0.0347, |
| "step": 1680 |
| }, |
| { |
| "epoch": 6.6015625, |
| "grad_norm": 0.2728106677532196, |
| "learning_rate": 2.838898190227931e-05, |
| "loss": 0.0333, |
| "step": 1690 |
| }, |
| { |
| "epoch": 6.640625, |
| "grad_norm": 0.2323320060968399, |
| "learning_rate": 2.780836124512494e-05, |
| "loss": 0.0362, |
| "step": 1700 |
| }, |
| { |
| "epoch": 6.6796875, |
| "grad_norm": 0.3762952387332916, |
| "learning_rate": 2.7231443606960238e-05, |
| "loss": 0.0352, |
| "step": 1710 |
| }, |
| { |
| "epoch": 6.71875, |
| "grad_norm": 0.31488674879074097, |
| "learning_rate": 2.6658325255437744e-05, |
| "loss": 0.0318, |
| "step": 1720 |
| }, |
| { |
| "epoch": 6.7578125, |
| "grad_norm": 0.3722898066043854, |
| "learning_rate": 2.6089101824240138e-05, |
| "loss": 0.0341, |
| "step": 1730 |
| }, |
| { |
| "epoch": 6.796875, |
| "grad_norm": 0.41492217779159546, |
| "learning_rate": 2.5523868297122265e-05, |
| "loss": 0.0446, |
| "step": 1740 |
| }, |
| { |
| "epoch": 6.8359375, |
| "grad_norm": 0.2935633361339569, |
| "learning_rate": 2.4962718992061634e-05, |
| "loss": 0.0381, |
| "step": 1750 |
| }, |
| { |
| "epoch": 6.875, |
| "grad_norm": 0.30372825264930725, |
| "learning_rate": 2.4405747545519963e-05, |
| "loss": 0.0369, |
| "step": 1760 |
| }, |
| { |
| "epoch": 6.9140625, |
| "grad_norm": 0.265733540058136, |
| "learning_rate": 2.385304689681847e-05, |
| "loss": 0.0319, |
| "step": 1770 |
| }, |
| { |
| "epoch": 6.953125, |
| "grad_norm": 0.22491760551929474, |
| "learning_rate": 2.3304709272629584e-05, |
| "loss": 0.0333, |
| "step": 1780 |
| }, |
| { |
| "epoch": 6.9921875, |
| "grad_norm": 0.3895363211631775, |
| "learning_rate": 2.276082617158743e-05, |
| "loss": 0.0367, |
| "step": 1790 |
| }, |
| { |
| "epoch": 7.03125, |
| "grad_norm": 0.24833568930625916, |
| "learning_rate": 2.2221488349019903e-05, |
| "loss": 0.0292, |
| "step": 1800 |
| }, |
| { |
| "epoch": 7.0703125, |
| "grad_norm": 0.3045443594455719, |
| "learning_rate": 2.1686785801804653e-05, |
| "loss": 0.0322, |
| "step": 1810 |
| }, |
| { |
| "epoch": 7.109375, |
| "grad_norm": 0.2556830644607544, |
| "learning_rate": 2.115680775335184e-05, |
| "loss": 0.0325, |
| "step": 1820 |
| }, |
| { |
| "epoch": 7.1484375, |
| "grad_norm": 0.28755396604537964, |
| "learning_rate": 2.063164263871573e-05, |
| "loss": 0.0333, |
| "step": 1830 |
| }, |
| { |
| "epoch": 7.1875, |
| "grad_norm": 0.4773646593093872, |
| "learning_rate": 2.0111378089837956e-05, |
| "loss": 0.0333, |
| "step": 1840 |
| }, |
| { |
| "epoch": 7.2265625, |
| "grad_norm": 0.375969260931015, |
| "learning_rate": 1.959610092092479e-05, |
| "loss": 0.0322, |
| "step": 1850 |
| }, |
| { |
| "epoch": 7.265625, |
| "grad_norm": 0.2566347122192383, |
| "learning_rate": 1.9085897113960793e-05, |
| "loss": 0.0312, |
| "step": 1860 |
| }, |
| { |
| "epoch": 7.3046875, |
| "grad_norm": 0.513367235660553, |
| "learning_rate": 1.858085180436147e-05, |
| "loss": 0.0337, |
| "step": 1870 |
| }, |
| { |
| "epoch": 7.34375, |
| "grad_norm": 0.31472358107566833, |
| "learning_rate": 1.8081049266767046e-05, |
| "loss": 0.0302, |
| "step": 1880 |
| }, |
| { |
| "epoch": 7.3828125, |
| "grad_norm": 0.31810522079467773, |
| "learning_rate": 1.758657290097997e-05, |
| "loss": 0.03, |
| "step": 1890 |
| }, |
| { |
| "epoch": 7.421875, |
| "grad_norm": 0.3877921998500824, |
| "learning_rate": 1.709750521804831e-05, |
| "loss": 0.0384, |
| "step": 1900 |
| }, |
| { |
| "epoch": 7.4609375, |
| "grad_norm": 0.45386990904808044, |
| "learning_rate": 1.661392782649755e-05, |
| "loss": 0.037, |
| "step": 1910 |
| }, |
| { |
| "epoch": 7.5, |
| "grad_norm": 0.33699849247932434, |
| "learning_rate": 1.6135921418712956e-05, |
| "loss": 0.0316, |
| "step": 1920 |
| }, |
| { |
| "epoch": 7.5390625, |
| "grad_norm": 0.31667983531951904, |
| "learning_rate": 1.5663565757474773e-05, |
| "loss": 0.029, |
| "step": 1930 |
| }, |
| { |
| "epoch": 7.578125, |
| "grad_norm": 0.3529506325721741, |
| "learning_rate": 1.5196939662648569e-05, |
| "loss": 0.0353, |
| "step": 1940 |
| }, |
| { |
| "epoch": 7.6171875, |
| "grad_norm": 0.3233952820301056, |
| "learning_rate": 1.4736120998032949e-05, |
| "loss": 0.0331, |
| "step": 1950 |
| }, |
| { |
| "epoch": 7.65625, |
| "grad_norm": 0.439943790435791, |
| "learning_rate": 1.428118665836668e-05, |
| "loss": 0.0296, |
| "step": 1960 |
| }, |
| { |
| "epoch": 7.6953125, |
| "grad_norm": 0.25135713815689087, |
| "learning_rate": 1.3832212556497748e-05, |
| "loss": 0.0312, |
| "step": 1970 |
| }, |
| { |
| "epoch": 7.734375, |
| "grad_norm": 0.4036009609699249, |
| "learning_rate": 1.338927361071603e-05, |
| "loss": 0.0324, |
| "step": 1980 |
| }, |
| { |
| "epoch": 7.7734375, |
| "grad_norm": 0.27127811312675476, |
| "learning_rate": 1.2952443732252057e-05, |
| "loss": 0.0305, |
| "step": 1990 |
| }, |
| { |
| "epoch": 7.8125, |
| "grad_norm": 0.569515585899353, |
| "learning_rate": 1.2521795812943704e-05, |
| "loss": 0.0286, |
| "step": 2000 |
| }, |
| { |
| "epoch": 7.8515625, |
| "grad_norm": 0.7388462424278259, |
| "learning_rate": 1.2097401713073175e-05, |
| "loss": 0.0329, |
| "step": 2010 |
| }, |
| { |
| "epoch": 7.890625, |
| "grad_norm": 0.35602420568466187, |
| "learning_rate": 1.1679332249375923e-05, |
| "loss": 0.032, |
| "step": 2020 |
| }, |
| { |
| "epoch": 7.9296875, |
| "grad_norm": 0.23437829315662384, |
| "learning_rate": 1.1267657183223812e-05, |
| "loss": 0.0325, |
| "step": 2030 |
| }, |
| { |
| "epoch": 7.96875, |
| "grad_norm": 0.8530910015106201, |
| "learning_rate": 1.086244520898428e-05, |
| "loss": 0.0365, |
| "step": 2040 |
| }, |
| { |
| "epoch": 8.0078125, |
| "grad_norm": 0.44752374291419983, |
| "learning_rate": 1.046376394255773e-05, |
| "loss": 0.0309, |
| "step": 2050 |
| }, |
| { |
| "epoch": 8.046875, |
| "grad_norm": 0.5457649230957031, |
| "learning_rate": 1.00716799100947e-05, |
| "loss": 0.0317, |
| "step": 2060 |
| }, |
| { |
| "epoch": 8.0859375, |
| "grad_norm": 0.4462254047393799, |
| "learning_rate": 9.686258536894948e-06, |
| "loss": 0.0286, |
| "step": 2070 |
| }, |
| { |
| "epoch": 8.125, |
| "grad_norm": 0.2501213252544403, |
| "learning_rate": 9.307564136490254e-06, |
| "loss": 0.0351, |
| "step": 2080 |
| }, |
| { |
| "epoch": 8.1640625, |
| "grad_norm": 0.49755340814590454, |
| "learning_rate": 8.935659899912658e-06, |
| "loss": 0.0328, |
| "step": 2090 |
| }, |
| { |
| "epoch": 8.203125, |
| "grad_norm": 0.4353477954864502, |
| "learning_rate": 8.5706078851501e-06, |
| "loss": 0.0306, |
| "step": 2100 |
| }, |
| { |
| "epoch": 8.2421875, |
| "grad_norm": 0.2827202379703522, |
| "learning_rate": 8.21246900679109e-06, |
| "loss": 0.0316, |
| "step": 2110 |
| }, |
| { |
| "epoch": 8.28125, |
| "grad_norm": 0.26606473326683044, |
| "learning_rate": 7.861303025860135e-06, |
| "loss": 0.0254, |
| "step": 2120 |
| }, |
| { |
| "epoch": 8.3203125, |
| "grad_norm": 0.251438170671463, |
| "learning_rate": 7.517168539845709e-06, |
| "loss": 0.0258, |
| "step": 2130 |
| }, |
| { |
| "epoch": 8.359375, |
| "grad_norm": 0.38992422819137573, |
| "learning_rate": 7.180122972922304e-06, |
| "loss": 0.03, |
| "step": 2140 |
| }, |
| { |
| "epoch": 8.3984375, |
| "grad_norm": 0.4747330844402313, |
| "learning_rate": 6.8502225663683975e-06, |
| "loss": 0.0305, |
| "step": 2150 |
| }, |
| { |
| "epoch": 8.4375, |
| "grad_norm": 0.29881390929222107, |
| "learning_rate": 6.527522369181655e-06, |
| "loss": 0.0255, |
| "step": 2160 |
| }, |
| { |
| "epoch": 8.4765625, |
| "grad_norm": 0.696563720703125, |
| "learning_rate": 6.212076228893205e-06, |
| "loss": 0.0286, |
| "step": 2170 |
| }, |
| { |
| "epoch": 8.515625, |
| "grad_norm": 0.26265567541122437, |
| "learning_rate": 5.903936782582253e-06, |
| "loss": 0.033, |
| "step": 2180 |
| }, |
| { |
| "epoch": 8.5546875, |
| "grad_norm": 0.42761778831481934, |
| "learning_rate": 5.603155448092895e-06, |
| "loss": 0.0296, |
| "step": 2190 |
| }, |
| { |
| "epoch": 8.59375, |
| "grad_norm": 0.2030072659254074, |
| "learning_rate": 5.309782415454151e-06, |
| "loss": 0.0289, |
| "step": 2200 |
| }, |
| { |
| "epoch": 8.6328125, |
| "grad_norm": 0.2712879776954651, |
| "learning_rate": 5.023866638505031e-06, |
| "loss": 0.0354, |
| "step": 2210 |
| }, |
| { |
| "epoch": 8.671875, |
| "grad_norm": 0.26241105794906616, |
| "learning_rate": 4.745455826725825e-06, |
| "loss": 0.0316, |
| "step": 2220 |
| }, |
| { |
| "epoch": 8.7109375, |
| "grad_norm": 0.2872694134712219, |
| "learning_rate": 4.474596437276996e-06, |
| "loss": 0.03, |
| "step": 2230 |
| }, |
| { |
| "epoch": 8.75, |
| "grad_norm": 0.28568607568740845, |
| "learning_rate": 4.2113336672471245e-06, |
| "loss": 0.0294, |
| "step": 2240 |
| }, |
| { |
| "epoch": 8.7890625, |
| "grad_norm": 0.3385384976863861, |
| "learning_rate": 3.955711446111082e-06, |
| "loss": 0.0304, |
| "step": 2250 |
| }, |
| { |
| "epoch": 8.828125, |
| "grad_norm": 0.675881564617157, |
| "learning_rate": 3.707772428399681e-06, |
| "loss": 0.0334, |
| "step": 2260 |
| }, |
| { |
| "epoch": 8.8671875, |
| "grad_norm": 0.5481550097465515, |
| "learning_rate": 3.467557986582104e-06, |
| "loss": 0.0297, |
| "step": 2270 |
| }, |
| { |
| "epoch": 8.90625, |
| "grad_norm": 0.26067742705345154, |
| "learning_rate": 3.2351082041623125e-06, |
| "loss": 0.0287, |
| "step": 2280 |
| }, |
| { |
| "epoch": 8.9453125, |
| "grad_norm": 0.5334944128990173, |
| "learning_rate": 3.0104618689904386e-06, |
| "loss": 0.0322, |
| "step": 2290 |
| }, |
| { |
| "epoch": 8.984375, |
| "grad_norm": 0.37815043330192566, |
| "learning_rate": 2.793656466790473e-06, |
| "loss": 0.0255, |
| "step": 2300 |
| }, |
| { |
| "epoch": 9.0234375, |
| "grad_norm": 0.1792028248310089, |
| "learning_rate": 2.584728174905171e-06, |
| "loss": 0.029, |
| "step": 2310 |
| }, |
| { |
| "epoch": 9.0625, |
| "grad_norm": 0.8325735926628113, |
| "learning_rate": 2.3837118562592797e-06, |
| "loss": 0.0301, |
| "step": 2320 |
| }, |
| { |
| "epoch": 9.1015625, |
| "grad_norm": 0.28852543234825134, |
| "learning_rate": 2.1906410535421697e-06, |
| "loss": 0.0303, |
| "step": 2330 |
| }, |
| { |
| "epoch": 9.140625, |
| "grad_norm": 0.5437615513801575, |
| "learning_rate": 2.005547983610684e-06, |
| "loss": 0.0319, |
| "step": 2340 |
| }, |
| { |
| "epoch": 9.1796875, |
| "grad_norm": 0.2050412893295288, |
| "learning_rate": 1.828463532113256e-06, |
| "loss": 0.029, |
| "step": 2350 |
| }, |
| { |
| "epoch": 9.21875, |
| "grad_norm": 0.38466140627861023, |
| "learning_rate": 1.6594172483361758e-06, |
| "loss": 0.0261, |
| "step": 2360 |
| }, |
| { |
| "epoch": 9.2578125, |
| "grad_norm": 0.30740851163864136, |
| "learning_rate": 1.4984373402728014e-06, |
| "loss": 0.0258, |
| "step": 2370 |
| }, |
| { |
| "epoch": 9.296875, |
| "grad_norm": 0.33520451188087463, |
| "learning_rate": 1.3455506699166732e-06, |
| "loss": 0.0305, |
| "step": 2380 |
| }, |
| { |
| "epoch": 9.3359375, |
| "grad_norm": 0.2344502955675125, |
| "learning_rate": 1.2007827487791112e-06, |
| "loss": 0.0307, |
| "step": 2390 |
| }, |
| { |
| "epoch": 9.375, |
| "grad_norm": 0.4900089502334595, |
| "learning_rate": 1.064157733632276e-06, |
| "loss": 0.0305, |
| "step": 2400 |
| }, |
| { |
| "epoch": 9.4140625, |
| "grad_norm": 0.34539440274238586, |
| "learning_rate": 9.35698422478204e-07, |
| "loss": 0.0272, |
| "step": 2410 |
| }, |
| { |
| "epoch": 9.453125, |
| "grad_norm": 0.4583103358745575, |
| "learning_rate": 8.154262507446109e-07, |
| "loss": 0.0279, |
| "step": 2420 |
| }, |
| { |
| "epoch": 9.4921875, |
| "grad_norm": 0.39252835512161255, |
| "learning_rate": 7.033612877080819e-07, |
| "loss": 0.0339, |
| "step": 2430 |
| }, |
| { |
| "epoch": 9.53125, |
| "grad_norm": 0.6091126799583435, |
| "learning_rate": 5.995222331451722e-07, |
| "loss": 0.028, |
| "step": 2440 |
| }, |
| { |
| "epoch": 9.5703125, |
| "grad_norm": 0.36491745710372925, |
| "learning_rate": 5.039264142120803e-07, |
| "loss": 0.0295, |
| "step": 2450 |
| }, |
| { |
| "epoch": 9.609375, |
| "grad_norm": 0.31240883469581604, |
| "learning_rate": 4.165897825533227e-07, |
| "loss": 0.028, |
| "step": 2460 |
| }, |
| { |
| "epoch": 9.6484375, |
| "grad_norm": 0.35602691769599915, |
| "learning_rate": 3.375269116399793e-07, |
| "loss": 0.0359, |
| "step": 2470 |
| }, |
| { |
| "epoch": 9.6875, |
| "grad_norm": 0.2611185312271118, |
| "learning_rate": 2.667509943378721e-07, |
| "loss": 0.0244, |
| "step": 2480 |
| }, |
| { |
| "epoch": 9.7265625, |
| "grad_norm": 0.3753460943698883, |
| "learning_rate": 2.0427384070612065e-07, |
| "loss": 0.0278, |
| "step": 2490 |
| }, |
| { |
| "epoch": 9.765625, |
| "grad_norm": 0.4795132577419281, |
| "learning_rate": 1.5010587602647397e-07, |
| "loss": 0.0324, |
| "step": 2500 |
| }, |
| { |
| "epoch": 9.8046875, |
| "grad_norm": 0.4804089069366455, |
| "learning_rate": 1.0425613906365761e-07, |
| "loss": 0.0289, |
| "step": 2510 |
| }, |
| { |
| "epoch": 9.84375, |
| "grad_norm": 0.3166308104991913, |
| "learning_rate": 6.673228055715241e-08, |
| "loss": 0.0289, |
| "step": 2520 |
| }, |
| { |
| "epoch": 9.8828125, |
| "grad_norm": 0.3317022919654846, |
| "learning_rate": 3.754056194452682e-08, |
| "loss": 0.0327, |
| "step": 2530 |
| }, |
| { |
| "epoch": 9.921875, |
| "grad_norm": 0.2820194363594055, |
| "learning_rate": 1.668585431663372e-08, |
| "loss": 0.0268, |
| "step": 2540 |
| }, |
| { |
| "epoch": 9.9609375, |
| "grad_norm": 0.3767685890197754, |
| "learning_rate": 4.171637604760692e-09, |
| "loss": 0.0298, |
| "step": 2550 |
| }, |
| { |
| "epoch": 10.0, |
| "grad_norm": 0.375841349363327, |
| "learning_rate": 0.0, |
| "loss": 0.0277, |
| "step": 2560 |
| }, |
| { |
| "epoch": 10.0, |
| "step": 2560, |
| "total_flos": 2.6088527768247245e+17, |
| "train_loss": 0.060091388958971946, |
| "train_runtime": 2712.9359, |
| "train_samples_per_second": 46.127, |
| "train_steps_per_second": 0.944 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 2560, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 10, |
| "save_steps": 10000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2.6088527768247245e+17, |
| "train_batch_size": 49, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
