| { | |
| "best_metric": NaN, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.5675368898978433, | |
| "eval_steps": 200, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0005675368898978433, | |
| "grad_norm": 4.242635250091553, | |
| "learning_rate": 1e-05, | |
| "loss": 2.0895, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0011350737797956867, | |
| "grad_norm": 3.8852005004882812, | |
| "learning_rate": 2e-05, | |
| "loss": 2.5238, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00170261066969353, | |
| "grad_norm": 4.170104026794434, | |
| "learning_rate": 3e-05, | |
| "loss": 2.7007, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0022701475595913734, | |
| "grad_norm": 4.369339942932129, | |
| "learning_rate": 4e-05, | |
| "loss": 2.7838, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0028376844494892167, | |
| "grad_norm": 4.925931453704834, | |
| "learning_rate": 5e-05, | |
| "loss": 2.8797, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00340522133938706, | |
| "grad_norm": 5.264482021331787, | |
| "learning_rate": 6e-05, | |
| "loss": 2.971, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.003972758229284903, | |
| "grad_norm": 5.356241703033447, | |
| "learning_rate": 7e-05, | |
| "loss": 3.1023, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.004540295119182747, | |
| "grad_norm": 5.722445964813232, | |
| "learning_rate": 8e-05, | |
| "loss": 2.7219, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00510783200908059, | |
| "grad_norm": 10.222125053405762, | |
| "learning_rate": 9e-05, | |
| "loss": 2.7866, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0056753688989784334, | |
| "grad_norm": 6.646042346954346, | |
| "learning_rate": 0.0001, | |
| "loss": 2.4618, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.006242905788876277, | |
| "grad_norm": 8.881997108459473, | |
| "learning_rate": 9.999974825027756e-05, | |
| "loss": 2.2149, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.00681044267877412, | |
| "grad_norm": 7.798340797424316, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 1.9245, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0073779795686719635, | |
| "grad_norm": 8.679018020629883, | |
| "learning_rate": 9.999773426770865e-05, | |
| "loss": 1.4955, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.007945516458569807, | |
| "grad_norm": 7.172414302825928, | |
| "learning_rate": 9.999597205514297e-05, | |
| "loss": 1.0912, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.00851305334846765, | |
| "grad_norm": 9.331069946289062, | |
| "learning_rate": 9.999370638369377e-05, | |
| "loss": 0.9289, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.009080590238365494, | |
| "grad_norm": 9.002270698547363, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.657, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.009648127128263337, | |
| "grad_norm": 9.825034141540527, | |
| "learning_rate": 9.998766476047547e-05, | |
| "loss": 0.6728, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.01021566401816118, | |
| "grad_norm": 14.936773300170898, | |
| "learning_rate": 9.998388886954547e-05, | |
| "loss": 0.6303, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.010783200908059024, | |
| "grad_norm": 14.347192764282227, | |
| "learning_rate": 9.997960964140947e-05, | |
| "loss": 0.6379, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.011350737797956867, | |
| "grad_norm": 12.535562515258789, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.5843, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01191827468785471, | |
| "grad_norm": 12.070816993713379, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 0.5889, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.012485811577752554, | |
| "grad_norm": 10.580400466918945, | |
| "learning_rate": 9.996375239002369e-05, | |
| "loss": 0.601, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.013053348467650397, | |
| "grad_norm": 8.476105690002441, | |
| "learning_rate": 9.995746029466071e-05, | |
| "loss": 0.6079, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01362088535754824, | |
| "grad_norm": 9.052068710327148, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 0.6068, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.014188422247446084, | |
| "grad_norm": 7.772611141204834, | |
| "learning_rate": 9.99433669591504e-05, | |
| "loss": 0.7142, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.014755959137343927, | |
| "grad_norm": 5.843691349029541, | |
| "learning_rate": 9.993556586092281e-05, | |
| "loss": 0.4146, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.01532349602724177, | |
| "grad_norm": 6.3077392578125, | |
| "learning_rate": 9.992726191210138e-05, | |
| "loss": 0.4692, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.015891032917139614, | |
| "grad_norm": 5.899518013000488, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.3898, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.016458569807037457, | |
| "grad_norm": 7.513155460357666, | |
| "learning_rate": 9.990914580222257e-05, | |
| "loss": 0.4173, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0170261066969353, | |
| "grad_norm": 7.608649253845215, | |
| "learning_rate": 9.989933382359422e-05, | |
| "loss": 0.3552, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.017593643586833144, | |
| "grad_norm": 6.533905029296875, | |
| "learning_rate": 9.988901935922826e-05, | |
| "loss": 0.3371, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.018161180476730987, | |
| "grad_norm": 5.706025123596191, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.2471, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.01872871736662883, | |
| "grad_norm": 4.540302276611328, | |
| "learning_rate": 9.986688339380862e-05, | |
| "loss": 0.2083, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.019296254256526674, | |
| "grad_norm": 3.610647439956665, | |
| "learning_rate": 9.985506211566388e-05, | |
| "loss": 0.1795, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.019863791146424517, | |
| "grad_norm": 5.012790203094482, | |
| "learning_rate": 9.984273879759713e-05, | |
| "loss": 0.2031, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02043132803632236, | |
| "grad_norm": 3.4943277835845947, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.1355, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.020998864926220204, | |
| "grad_norm": 4.174023628234863, | |
| "learning_rate": 9.981658654313457e-05, | |
| "loss": 0.1841, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.021566401816118047, | |
| "grad_norm": 2.6815176010131836, | |
| "learning_rate": 9.98027578700917e-05, | |
| "loss": 0.1052, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.02213393870601589, | |
| "grad_norm": 1.8553539514541626, | |
| "learning_rate": 9.978842768382998e-05, | |
| "loss": 0.0862, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.022701475595913734, | |
| "grad_norm": 2.5569417476654053, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.0708, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.023269012485811577, | |
| "grad_norm": 4.1771979331970215, | |
| "learning_rate": 9.975826335391808e-05, | |
| "loss": 0.0879, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.02383654937570942, | |
| "grad_norm": 3.163653612136841, | |
| "learning_rate": 9.974242951402235e-05, | |
| "loss": 0.1033, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.024404086265607264, | |
| "grad_norm": 4.081658840179443, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.1103, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.024971623155505107, | |
| "grad_norm": 5.564581871032715, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.1136, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.02553916004540295, | |
| "grad_norm": 86.91041564941406, | |
| "learning_rate": 9.969192322306271e-05, | |
| "loss": 1.3437, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.026106696935300794, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.967408676742751e-05, | |
| "loss": 0.0, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.026674233825198637, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.965575009429006e-05, | |
| "loss": 0.0, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.02724177071509648, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.0, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.027809307604994324, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.961757683914406e-05, | |
| "loss": 0.0, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.028376844494892167, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.959774064153977e-05, | |
| "loss": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02894438138479001, | |
| "grad_norm": 13.452136993408203, | |
| "learning_rate": 9.957740499523787e-05, | |
| "loss": 1.2541, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.029511918274687854, | |
| "grad_norm": 10.72777271270752, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 1.1451, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.030079455164585697, | |
| "grad_norm": 5.7461466789245605, | |
| "learning_rate": 9.953523618068749e-05, | |
| "loss": 0.7454, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.03064699205448354, | |
| "grad_norm": 4.224442481994629, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.6189, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.031214528944381384, | |
| "grad_norm": 5.231300354003906, | |
| "learning_rate": 9.949107209404665e-05, | |
| "loss": 0.5754, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03178206583427923, | |
| "grad_norm": 4.194768905639648, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.4918, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.03234960272417707, | |
| "grad_norm": 3.1545193195343018, | |
| "learning_rate": 9.944491451423828e-05, | |
| "loss": 0.3889, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.032917139614074914, | |
| "grad_norm": 4.240617275238037, | |
| "learning_rate": 9.942108874226811e-05, | |
| "loss": 0.383, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.03348467650397276, | |
| "grad_norm": 3.6889731884002686, | |
| "learning_rate": 9.939676530048301e-05, | |
| "loss": 0.3308, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0340522133938706, | |
| "grad_norm": 7.476080894470215, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.2755, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.034619750283768444, | |
| "grad_norm": 3.6329305171966553, | |
| "learning_rate": 9.934662639222412e-05, | |
| "loss": 0.2908, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.03518728717366629, | |
| "grad_norm": 3.839686870574951, | |
| "learning_rate": 9.93208114306486e-05, | |
| "loss": 0.317, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.03575482406356413, | |
| "grad_norm": 2.5644800662994385, | |
| "learning_rate": 9.929449980904952e-05, | |
| "loss": 0.2524, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.036322360953461974, | |
| "grad_norm": 2.268704652786255, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.2686, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.03688989784335982, | |
| "grad_norm": 3.147296905517578, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.2311, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.03745743473325766, | |
| "grad_norm": 2.1133084297180176, | |
| "learning_rate": 9.921258765867919e-05, | |
| "loss": 0.2072, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.038024971623155504, | |
| "grad_norm": 1.422222375869751, | |
| "learning_rate": 9.918429209653662e-05, | |
| "loss": 0.2128, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.03859250851305335, | |
| "grad_norm": 1.157293438911438, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.1427, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.03916004540295119, | |
| "grad_norm": 1.4492322206497192, | |
| "learning_rate": 9.912621540634887e-05, | |
| "loss": 0.1704, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.039727582292849034, | |
| "grad_norm": 1.1632133722305298, | |
| "learning_rate": 9.909643486313533e-05, | |
| "loss": 0.1299, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04029511918274688, | |
| "grad_norm": 3.0194833278656006, | |
| "learning_rate": 9.90661599193678e-05, | |
| "loss": 0.1088, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.04086265607264472, | |
| "grad_norm": 1.3391764163970947, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.1105, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.041430192962542564, | |
| "grad_norm": 1.761568307876587, | |
| "learning_rate": 9.900412805461967e-05, | |
| "loss": 0.1449, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.04199772985244041, | |
| "grad_norm": 1.8852134943008423, | |
| "learning_rate": 9.897237175829926e-05, | |
| "loss": 0.1381, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.04256526674233825, | |
| "grad_norm": 1.8318125009536743, | |
| "learning_rate": 9.894012231073894e-05, | |
| "loss": 0.0953, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.043132803632236094, | |
| "grad_norm": 1.9700140953063965, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.1002, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.04370034052213394, | |
| "grad_norm": 2.201488971710205, | |
| "learning_rate": 9.887414526586763e-05, | |
| "loss": 0.0935, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.04426787741203178, | |
| "grad_norm": 2.130439281463623, | |
| "learning_rate": 9.884041833294476e-05, | |
| "loss": 0.1128, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.044835414301929624, | |
| "grad_norm": 1.2699000835418701, | |
| "learning_rate": 9.880619957755151e-05, | |
| "loss": 0.0704, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.04540295119182747, | |
| "grad_norm": 1.2597098350524902, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.0913, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04597048808172531, | |
| "grad_norm": 1.9284838438034058, | |
| "learning_rate": 9.873628798263296e-05, | |
| "loss": 0.1002, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.046538024971623154, | |
| "grad_norm": 2.915311574935913, | |
| "learning_rate": 9.870059584711668e-05, | |
| "loss": 0.0878, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.047105561861521, | |
| "grad_norm": 2.436211347579956, | |
| "learning_rate": 9.866441329714088e-05, | |
| "loss": 0.0733, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.04767309875141884, | |
| "grad_norm": 1.696310043334961, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.0852, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.048240635641316684, | |
| "grad_norm": 0.9896736145019531, | |
| "learning_rate": 9.859057841617709e-05, | |
| "loss": 0.0163, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.04880817253121453, | |
| "grad_norm": 1.1014189720153809, | |
| "learning_rate": 9.855292682870551e-05, | |
| "loss": 0.0783, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.04937570942111237, | |
| "grad_norm": 1.2181459665298462, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 0.0454, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.049943246311010214, | |
| "grad_norm": 1.8675448894500732, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.062, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.05051078320090806, | |
| "grad_norm": 2.457249641418457, | |
| "learning_rate": 9.843704004290392e-05, | |
| "loss": 0.0158, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.0510783200908059, | |
| "grad_norm": 1.7994331121444702, | |
| "learning_rate": 9.839743506981782e-05, | |
| "loss": 0.0463, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.051645856980703744, | |
| "grad_norm": 1.6931531429290771, | |
| "learning_rate": 9.835734273509786e-05, | |
| "loss": 0.0308, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.05221339387060159, | |
| "grad_norm": 0.2792555093765259, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.0043, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.05278093076049943, | |
| "grad_norm": 1.1043611764907837, | |
| "learning_rate": 9.827569760057755e-05, | |
| "loss": 0.0266, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.053348467650397274, | |
| "grad_norm": 2.844864845275879, | |
| "learning_rate": 9.82341456229428e-05, | |
| "loss": 0.0724, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.05391600454029512, | |
| "grad_norm": 47.320552825927734, | |
| "learning_rate": 9.819210792799712e-05, | |
| "loss": 0.9477, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.05448354143019296, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.0, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.055051078320090804, | |
| "grad_norm": 8.552749633789062, | |
| "learning_rate": 9.810657708433637e-05, | |
| "loss": 0.0776, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.05561861520998865, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.0, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.05618615209988649, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.801910851476523e-05, | |
| "loss": 0.0, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.056753688989784334, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05732122587968218, | |
| "grad_norm": 7.161002159118652, | |
| "learning_rate": 9.792970574250493e-05, | |
| "loss": 0.57, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.05788876276958002, | |
| "grad_norm": 5.865129470825195, | |
| "learning_rate": 9.788428015268027e-05, | |
| "loss": 0.4715, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.058456299659477864, | |
| "grad_norm": 4.487433433532715, | |
| "learning_rate": 9.783837236868609e-05, | |
| "loss": 0.3068, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.05902383654937571, | |
| "grad_norm": 4.0235209465026855, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.233, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.05959137343927355, | |
| "grad_norm": 5.06863260269165, | |
| "learning_rate": 9.77451120722037e-05, | |
| "loss": 0.2298, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.060158910329171394, | |
| "grad_norm": 2.5704989433288574, | |
| "learning_rate": 9.769776049884563e-05, | |
| "loss": 0.156, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.06072644721906924, | |
| "grad_norm": 3.1615333557128906, | |
| "learning_rate": 9.764992860956889e-05, | |
| "loss": 0.2069, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.06129398410896708, | |
| "grad_norm": 2.513225555419922, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 0.129, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.061861520998864925, | |
| "grad_norm": 2.5690817832946777, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.1146, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.06242905788876277, | |
| "grad_norm": 3.713148593902588, | |
| "learning_rate": 9.750355588704727e-05, | |
| "loss": 0.1187, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06299659477866061, | |
| "grad_norm": 2.496454954147339, | |
| "learning_rate": 9.745380759905647e-05, | |
| "loss": 0.1376, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.06356413166855845, | |
| "grad_norm": 1.6230063438415527, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.0658, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0641316685584563, | |
| "grad_norm": 1.4402037858963013, | |
| "learning_rate": 9.735287795090455e-05, | |
| "loss": 0.067, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.06469920544835414, | |
| "grad_norm": 1.7551097869873047, | |
| "learning_rate": 9.730169760710386e-05, | |
| "loss": 0.0546, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.06526674233825198, | |
| "grad_norm": 9.121005058288574, | |
| "learning_rate": 9.725004093573342e-05, | |
| "loss": 0.0719, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.06583427922814983, | |
| "grad_norm": 1.0054181814193726, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.04, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.06640181611804767, | |
| "grad_norm": 1.0197290182113647, | |
| "learning_rate": 9.714530069580309e-05, | |
| "loss": 0.0465, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.06696935300794551, | |
| "grad_norm": 1.8132661581039429, | |
| "learning_rate": 9.709221818197624e-05, | |
| "loss": 0.0315, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.06753688989784336, | |
| "grad_norm": 1.3299674987792969, | |
| "learning_rate": 9.703866145003511e-05, | |
| "loss": 0.0777, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.0681044267877412, | |
| "grad_norm": 0.7436254024505615, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0274, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06867196367763904, | |
| "grad_norm": 0.8951746821403503, | |
| "learning_rate": 9.693012749384279e-05, | |
| "loss": 0.0349, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.06923950056753689, | |
| "grad_norm": 1.018670678138733, | |
| "learning_rate": 9.687515136252731e-05, | |
| "loss": 0.0209, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.06980703745743473, | |
| "grad_norm": 0.5818009972572327, | |
| "learning_rate": 9.681970319895803e-05, | |
| "loss": 0.0151, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.07037457434733257, | |
| "grad_norm": 0.49394237995147705, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.0149, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.07094211123723042, | |
| "grad_norm": 1.2261513471603394, | |
| "learning_rate": 9.670739301325534e-05, | |
| "loss": 0.0279, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.07150964812712826, | |
| "grad_norm": 0.8551621437072754, | |
| "learning_rate": 9.665053212208426e-05, | |
| "loss": 0.0307, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.0720771850170261, | |
| "grad_norm": 0.7473931312561035, | |
| "learning_rate": 9.659320146057262e-05, | |
| "loss": 0.0258, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.07264472190692395, | |
| "grad_norm": 1.058668851852417, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.0333, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.07321225879682179, | |
| "grad_norm": 0.4712604284286499, | |
| "learning_rate": 9.647713314052896e-05, | |
| "loss": 0.0194, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.07377979568671963, | |
| "grad_norm": 0.554572343826294, | |
| "learning_rate": 9.641839665080363e-05, | |
| "loss": 0.012, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07434733257661748, | |
| "grad_norm": 0.6603901386260986, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.0223, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.07491486946651532, | |
| "grad_norm": 0.8694583773612976, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.0455, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.07548240635641316, | |
| "grad_norm": 1.202289342880249, | |
| "learning_rate": 9.623938497462646e-05, | |
| "loss": 0.0438, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.07604994324631101, | |
| "grad_norm": 3.9090631008148193, | |
| "learning_rate": 9.617878234984055e-05, | |
| "loss": 0.0531, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.07661748013620885, | |
| "grad_norm": 1.0893713235855103, | |
| "learning_rate": 9.611771470522908e-05, | |
| "loss": 0.0357, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.0771850170261067, | |
| "grad_norm": 0.5505892634391785, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.0116, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.07775255391600454, | |
| "grad_norm": 3.2899348735809326, | |
| "learning_rate": 9.599418682100793e-05, | |
| "loss": 0.028, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.07832009080590238, | |
| "grad_norm": 1.5124907493591309, | |
| "learning_rate": 9.593172782532268e-05, | |
| "loss": 0.0374, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.07888762769580022, | |
| "grad_norm": 1.1038376092910767, | |
| "learning_rate": 9.586880629764817e-05, | |
| "loss": 0.0433, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.07945516458569807, | |
| "grad_norm": 0.2565675675868988, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.0037, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.08002270147559591, | |
| "grad_norm": 1.869051456451416, | |
| "learning_rate": 9.574157818545901e-05, | |
| "loss": 0.0232, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.08059023836549375, | |
| "grad_norm": 0.21874603629112244, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.004, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.0811577752553916, | |
| "grad_norm": 0.47795429825782776, | |
| "learning_rate": 9.561250760917027e-05, | |
| "loss": 0.0062, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.08172531214528944, | |
| "grad_norm": 1.0321537256240845, | |
| "learning_rate": 9.554728301876526e-05, | |
| "loss": 0.0152, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.08229284903518728, | |
| "grad_norm": 10.375513076782227, | |
| "learning_rate": 9.548159976772592e-05, | |
| "loss": 0.5914, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.08286038592508513, | |
| "grad_norm": 10.878011703491211, | |
| "learning_rate": 9.541545851748186e-05, | |
| "loss": 0.0679, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.08342792281498297, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.534885993407474e-05, | |
| "loss": 0.0, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.08399545970488081, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.528180468815155e-05, | |
| "loss": 0.0, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.08456299659477866, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.521429345495787e-05, | |
| "loss": 0.0, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.0851305334846765, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.514632691433107e-05, | |
| "loss": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08569807037457434, | |
| "grad_norm": 5.398491859436035, | |
| "learning_rate": 9.507790575069347e-05, | |
| "loss": 0.2644, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.08626560726447219, | |
| "grad_norm": 3.6026408672332764, | |
| "learning_rate": 9.50090306530454e-05, | |
| "loss": 0.1911, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.08683314415437003, | |
| "grad_norm": 3.1517863273620605, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.0834, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.08740068104426787, | |
| "grad_norm": 3.0677297115325928, | |
| "learning_rate": 9.486992143456792e-05, | |
| "loss": 0.0744, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.08796821793416572, | |
| "grad_norm": 2.8431897163391113, | |
| "learning_rate": 9.479968871456679e-05, | |
| "loss": 0.0717, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.08853575482406356, | |
| "grad_norm": 1.704907774925232, | |
| "learning_rate": 9.472900486219769e-05, | |
| "loss": 0.0569, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.0891032917139614, | |
| "grad_norm": 1.7260456085205078, | |
| "learning_rate": 9.46578705892462e-05, | |
| "loss": 0.0341, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.08967082860385925, | |
| "grad_norm": 2.4934730529785156, | |
| "learning_rate": 9.458628661203367e-05, | |
| "loss": 0.0329, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.09023836549375709, | |
| "grad_norm": 2.195416212081909, | |
| "learning_rate": 9.451425365140996e-05, | |
| "loss": 0.0476, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.09080590238365494, | |
| "grad_norm": 0.4546680748462677, | |
| "learning_rate": 9.444177243274618e-05, | |
| "loss": 0.0112, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.09137343927355278, | |
| "grad_norm": 1.1639807224273682, | |
| "learning_rate": 9.43688436859274e-05, | |
| "loss": 0.0313, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.09194097616345062, | |
| "grad_norm": 2.623199462890625, | |
| "learning_rate": 9.429546814534529e-05, | |
| "loss": 0.0568, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.09250851305334847, | |
| "grad_norm": 0.5077713131904602, | |
| "learning_rate": 9.422164654989072e-05, | |
| "loss": 0.0125, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.09307604994324631, | |
| "grad_norm": 0.37310367822647095, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.0067, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.09364358683314415, | |
| "grad_norm": 1.5596776008605957, | |
| "learning_rate": 9.407266817237911e-05, | |
| "loss": 0.0392, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.094211123723042, | |
| "grad_norm": 1.5760366916656494, | |
| "learning_rate": 9.399751289053267e-05, | |
| "loss": 0.0296, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.09477866061293984, | |
| "grad_norm": 0.7743924856185913, | |
| "learning_rate": 9.392191455421988e-05, | |
| "loss": 0.0112, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.09534619750283768, | |
| "grad_norm": 0.9516783356666565, | |
| "learning_rate": 9.384587392471515e-05, | |
| "loss": 0.0109, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.09591373439273553, | |
| "grad_norm": 0.824831485748291, | |
| "learning_rate": 9.376939176774679e-05, | |
| "loss": 0.0133, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.09648127128263337, | |
| "grad_norm": 0.7576391696929932, | |
| "learning_rate": 9.369246885348926e-05, | |
| "loss": 0.0229, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.09704880817253121, | |
| "grad_norm": 0.7438343167304993, | |
| "learning_rate": 9.361510595655545e-05, | |
| "loss": 0.0148, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.09761634506242906, | |
| "grad_norm": 1.1500699520111084, | |
| "learning_rate": 9.353730385598887e-05, | |
| "loss": 0.0842, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.0981838819523269, | |
| "grad_norm": 0.8265398740768433, | |
| "learning_rate": 9.345906333525581e-05, | |
| "loss": 0.0161, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.09875141884222474, | |
| "grad_norm": 0.7592639923095703, | |
| "learning_rate": 9.338038518223747e-05, | |
| "loss": 0.0276, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.09931895573212259, | |
| "grad_norm": 0.9068871736526489, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.0173, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.09988649262202043, | |
| "grad_norm": 3.255566120147705, | |
| "learning_rate": 9.322171915289635e-05, | |
| "loss": 0.0371, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.10045402951191827, | |
| "grad_norm": 0.607418954372406, | |
| "learning_rate": 9.314173287433873e-05, | |
| "loss": 0.0184, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.10102156640181612, | |
| "grad_norm": 1.0031543970108032, | |
| "learning_rate": 9.306131215901003e-05, | |
| "loss": 0.0184, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.10158910329171396, | |
| "grad_norm": 0.9904316663742065, | |
| "learning_rate": 9.298045781674596e-05, | |
| "loss": 0.0403, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.1021566401816118, | |
| "grad_norm": 0.6840120553970337, | |
| "learning_rate": 9.289917066174886e-05, | |
| "loss": 0.012, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.10272417707150965, | |
| "grad_norm": 1.294670820236206, | |
| "learning_rate": 9.281745151257946e-05, | |
| "loss": 0.0204, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.10329171396140749, | |
| "grad_norm": 0.5927445888519287, | |
| "learning_rate": 9.273530119214868e-05, | |
| "loss": 0.0166, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.10385925085130533, | |
| "grad_norm": 1.159449815750122, | |
| "learning_rate": 9.265272052770936e-05, | |
| "loss": 0.0258, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.10442678774120318, | |
| "grad_norm": 0.5215921401977539, | |
| "learning_rate": 9.256971035084785e-05, | |
| "loss": 0.0121, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.10499432463110102, | |
| "grad_norm": 0.537041962146759, | |
| "learning_rate": 9.248627149747573e-05, | |
| "loss": 0.0125, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.10556186152099886, | |
| "grad_norm": 0.3171516954898834, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.0058, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.1061293984108967, | |
| "grad_norm": 0.13635091483592987, | |
| "learning_rate": 9.231811112642121e-05, | |
| "loss": 0.0028, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.10669693530079455, | |
| "grad_norm": 0.38101646304130554, | |
| "learning_rate": 9.223339130211192e-05, | |
| "loss": 0.0174, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.10726447219069239, | |
| "grad_norm": 0.6654653549194336, | |
| "learning_rate": 9.214824618802109e-05, | |
| "loss": 0.0101, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.10783200908059024, | |
| "grad_norm": 0.3335312008857727, | |
| "learning_rate": 9.206267664155907e-05, | |
| "loss": 0.0065, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.10839954597048808, | |
| "grad_norm": 0.4047980308532715, | |
| "learning_rate": 9.197668352441025e-05, | |
| "loss": 0.0035, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.10896708286038592, | |
| "grad_norm": 1.2880914211273193, | |
| "learning_rate": 9.189026770252436e-05, | |
| "loss": 0.0283, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.10953461975028377, | |
| "grad_norm": 0.276368111371994, | |
| "learning_rate": 9.18034300461078e-05, | |
| "loss": 0.0044, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.11010215664018161, | |
| "grad_norm": 0.5553678274154663, | |
| "learning_rate": 9.171617142961477e-05, | |
| "loss": 0.0107, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.11066969353007945, | |
| "grad_norm": 2.4395854473114014, | |
| "learning_rate": 9.162849273173857e-05, | |
| "loss": 0.0229, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.1112372304199773, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.154039483540273e-05, | |
| "loss": 0.0, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.11180476730987514, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 0.0, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.11237230419977298, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.136294500014386e-05, | |
| "loss": 0.0, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.11293984108967083, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.12735948481387e-05, | |
| "loss": 0.0, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.11350737797956867, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.118382907149165e-05, | |
| "loss": 0.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11350737797956867, | |
| "eval_loss": NaN, | |
| "eval_runtime": 106.2144, | |
| "eval_samples_per_second": 27.943, | |
| "eval_steps_per_second": 6.986, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11407491486946651, | |
| "grad_norm": 2.6148464679718018, | |
| "learning_rate": 9.109364857414306e-05, | |
| "loss": 0.1986, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.11464245175936436, | |
| "grad_norm": 1.466818928718567, | |
| "learning_rate": 9.100305426420956e-05, | |
| "loss": 0.0807, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.1152099886492622, | |
| "grad_norm": 0.8603935241699219, | |
| "learning_rate": 9.091204705397484e-05, | |
| "loss": 0.0535, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.11577752553916004, | |
| "grad_norm": 1.8044847249984741, | |
| "learning_rate": 9.082062785988049e-05, | |
| "loss": 0.0691, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.11634506242905789, | |
| "grad_norm": 0.43121302127838135, | |
| "learning_rate": 9.072879760251679e-05, | |
| "loss": 0.0214, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.11691259931895573, | |
| "grad_norm": 0.8256431221961975, | |
| "learning_rate": 9.06365572066134e-05, | |
| "loss": 0.0738, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.11748013620885357, | |
| "grad_norm": 0.938698410987854, | |
| "learning_rate": 9.05439076010301e-05, | |
| "loss": 0.0192, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.11804767309875142, | |
| "grad_norm": 0.551313579082489, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.0205, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.11861520998864926, | |
| "grad_norm": 1.3714677095413208, | |
| "learning_rate": 9.035738449685707e-05, | |
| "loss": 0.0479, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.1191827468785471, | |
| "grad_norm": 0.35658133029937744, | |
| "learning_rate": 9.026351287655294e-05, | |
| "loss": 0.0098, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.11975028376844495, | |
| "grad_norm": 0.6199024319648743, | |
| "learning_rate": 9.016923580312113e-05, | |
| "loss": 0.036, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.12031782065834279, | |
| "grad_norm": 0.315612256526947, | |
| "learning_rate": 9.007455422593077e-05, | |
| "loss": 0.0065, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.12088535754824063, | |
| "grad_norm": 0.4234760105609894, | |
| "learning_rate": 8.997946909842425e-05, | |
| "loss": 0.0294, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.12145289443813848, | |
| "grad_norm": 0.6859511137008667, | |
| "learning_rate": 8.988398137810777e-05, | |
| "loss": 0.0263, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.12202043132803632, | |
| "grad_norm": 0.41570374369621277, | |
| "learning_rate": 8.978809202654162e-05, | |
| "loss": 0.0102, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.12258796821793416, | |
| "grad_norm": 0.5462591052055359, | |
| "learning_rate": 8.969180200933047e-05, | |
| "loss": 0.027, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.123155505107832, | |
| "grad_norm": 1.4283958673477173, | |
| "learning_rate": 8.959511229611376e-05, | |
| "loss": 0.0177, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.12372304199772985, | |
| "grad_norm": 0.2500029504299164, | |
| "learning_rate": 8.949802386055581e-05, | |
| "loss": 0.0027, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.12429057888762769, | |
| "grad_norm": 0.452014297246933, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 0.0182, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.12485811577752554, | |
| "grad_norm": 0.896853506565094, | |
| "learning_rate": 8.930265473713938e-05, | |
| "loss": 0.0119, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.1254256526674234, | |
| "grad_norm": 0.2784510552883148, | |
| "learning_rate": 8.92043760166458e-05, | |
| "loss": 0.0081, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.12599318955732122, | |
| "grad_norm": 0.6255968809127808, | |
| "learning_rate": 8.910570250852097e-05, | |
| "loss": 0.0148, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.12656072644721908, | |
| "grad_norm": 0.7261110544204712, | |
| "learning_rate": 8.900663520640604e-05, | |
| "loss": 0.022, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.1271282633371169, | |
| "grad_norm": 1.0769401788711548, | |
| "learning_rate": 8.890717510790763e-05, | |
| "loss": 0.0061, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.12769580022701477, | |
| "grad_norm": 0.48627111315727234, | |
| "learning_rate": 8.880732321458784e-05, | |
| "loss": 0.0277, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1282633371169126, | |
| "grad_norm": 0.23434560000896454, | |
| "learning_rate": 8.870708053195413e-05, | |
| "loss": 0.0031, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.12883087400681045, | |
| "grad_norm": 0.5444976091384888, | |
| "learning_rate": 8.860644806944918e-05, | |
| "loss": 0.0096, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.12939841089670828, | |
| "grad_norm": 1.0603594779968262, | |
| "learning_rate": 8.850542684044078e-05, | |
| "loss": 0.0314, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.12996594778660614, | |
| "grad_norm": 0.5144844651222229, | |
| "learning_rate": 8.840401786221159e-05, | |
| "loss": 0.013, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.13053348467650397, | |
| "grad_norm": 0.1403108835220337, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.0024, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.13110102156640183, | |
| "grad_norm": 0.24111324548721313, | |
| "learning_rate": 8.820004074673433e-05, | |
| "loss": 0.0034, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.13166855845629966, | |
| "grad_norm": 0.789040744304657, | |
| "learning_rate": 8.809747466353356e-05, | |
| "loss": 0.0295, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.1322360953461975, | |
| "grad_norm": 0.7863220572471619, | |
| "learning_rate": 8.799452493918585e-05, | |
| "loss": 0.0139, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.13280363223609534, | |
| "grad_norm": 0.5174320340156555, | |
| "learning_rate": 8.789119261039385e-05, | |
| "loss": 0.0097, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.1333711691259932, | |
| "grad_norm": 0.23707064986228943, | |
| "learning_rate": 8.778747871771292e-05, | |
| "loss": 0.0029, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.13393870601589103, | |
| "grad_norm": 0.20634303987026215, | |
| "learning_rate": 8.768338430554082e-05, | |
| "loss": 0.0036, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.1345062429057889, | |
| "grad_norm": 0.34142839908599854, | |
| "learning_rate": 8.757891042210714e-05, | |
| "loss": 0.0041, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.13507377979568672, | |
| "grad_norm": 0.3984658420085907, | |
| "learning_rate": 8.74740581194627e-05, | |
| "loss": 0.0062, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.13564131668558457, | |
| "grad_norm": 0.39278241991996765, | |
| "learning_rate": 8.736882845346906e-05, | |
| "loss": 0.011, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.1362088535754824, | |
| "grad_norm": 0.49513405561447144, | |
| "learning_rate": 8.726322248378775e-05, | |
| "loss": 0.0138, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13677639046538026, | |
| "grad_norm": 0.1306513100862503, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 0.0009, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.1373439273552781, | |
| "grad_norm": 0.04676857590675354, | |
| "learning_rate": 8.705088589094459e-05, | |
| "loss": 0.0007, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.13791146424517595, | |
| "grad_norm": 0.1314803659915924, | |
| "learning_rate": 8.694415740600988e-05, | |
| "loss": 0.0018, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.13847900113507378, | |
| "grad_norm": 0.1374429613351822, | |
| "learning_rate": 8.683705689382024e-05, | |
| "loss": 0.0011, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.13904653802497163, | |
| "grad_norm": 0.0954061895608902, | |
| "learning_rate": 8.672958543287666e-05, | |
| "loss": 0.0008, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.13961407491486946, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.662174410541555e-05, | |
| "loss": 0.0, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.14018161180476732, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.651353399739787e-05, | |
| "loss": 0.0, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.14074914869466515, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.640495619849821e-05, | |
| "loss": 0.0, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.141316685584563, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.629601180209381e-05, | |
| "loss": 0.0, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.14188422247446084, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.618670190525352e-05, | |
| "loss": 0.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.1424517593643587, | |
| "grad_norm": 5.090747356414795, | |
| "learning_rate": 8.607702760872678e-05, | |
| "loss": 0.3383, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.14301929625425652, | |
| "grad_norm": 3.2074856758117676, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.1635, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.14358683314415438, | |
| "grad_norm": 1.3404688835144043, | |
| "learning_rate": 8.585659023794818e-05, | |
| "loss": 0.0633, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.1441543700340522, | |
| "grad_norm": 0.9414183497428894, | |
| "learning_rate": 8.574582938349817e-05, | |
| "loss": 0.0453, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.14472190692395007, | |
| "grad_norm": 1.2090741395950317, | |
| "learning_rate": 8.563470856894316e-05, | |
| "loss": 0.0521, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.1452894438138479, | |
| "grad_norm": 0.47880247235298157, | |
| "learning_rate": 8.552322891326846e-05, | |
| "loss": 0.0178, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.14585698070374575, | |
| "grad_norm": 0.9624150991439819, | |
| "learning_rate": 8.541139153907296e-05, | |
| "loss": 0.0165, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.14642451759364358, | |
| "grad_norm": 0.33472761511802673, | |
| "learning_rate": 8.529919757255783e-05, | |
| "loss": 0.0167, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.14699205448354144, | |
| "grad_norm": 0.44358474016189575, | |
| "learning_rate": 8.518664814351502e-05, | |
| "loss": 0.0156, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.14755959137343927, | |
| "grad_norm": 0.7658271789550781, | |
| "learning_rate": 8.507374438531607e-05, | |
| "loss": 0.0382, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14812712826333713, | |
| "grad_norm": 0.30857032537460327, | |
| "learning_rate": 8.496048743490053e-05, | |
| "loss": 0.0075, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.14869466515323496, | |
| "grad_norm": 1.9148088693618774, | |
| "learning_rate": 8.484687843276469e-05, | |
| "loss": 0.0156, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.1492622020431328, | |
| "grad_norm": 0.8631612658500671, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 0.0441, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.14982973893303064, | |
| "grad_norm": 0.5475025773048401, | |
| "learning_rate": 8.461860885303114e-05, | |
| "loss": 0.0406, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.1503972758229285, | |
| "grad_norm": 0.12846483290195465, | |
| "learning_rate": 8.450395057410561e-05, | |
| "loss": 0.003, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.15096481271282633, | |
| "grad_norm": 0.275508850812912, | |
| "learning_rate": 8.438894484078086e-05, | |
| "loss": 0.0133, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.1515323496027242, | |
| "grad_norm": 0.5227663516998291, | |
| "learning_rate": 8.427359281116334e-05, | |
| "loss": 0.0177, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.15209988649262202, | |
| "grad_norm": 0.6141021847724915, | |
| "learning_rate": 8.415789564684673e-05, | |
| "loss": 0.0131, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.15266742338251987, | |
| "grad_norm": 0.04998692125082016, | |
| "learning_rate": 8.404185451290018e-05, | |
| "loss": 0.0013, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.1532349602724177, | |
| "grad_norm": 0.46891334652900696, | |
| "learning_rate": 8.392547057785661e-05, | |
| "loss": 0.0273, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.15380249716231556, | |
| "grad_norm": 0.5121367573738098, | |
| "learning_rate": 8.380874501370097e-05, | |
| "loss": 0.019, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.1543700340522134, | |
| "grad_norm": 0.12058154493570328, | |
| "learning_rate": 8.369167899585841e-05, | |
| "loss": 0.0022, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.15493757094211125, | |
| "grad_norm": 0.31559038162231445, | |
| "learning_rate": 8.357427370318239e-05, | |
| "loss": 0.0164, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.15550510783200908, | |
| "grad_norm": 0.7393925189971924, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.0179, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.15607264472190693, | |
| "grad_norm": 0.4647248089313507, | |
| "learning_rate": 8.333845002581458e-05, | |
| "loss": 0.008, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.15664018161180476, | |
| "grad_norm": 0.21942315995693207, | |
| "learning_rate": 8.322003401586462e-05, | |
| "loss": 0.0028, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.15720771850170262, | |
| "grad_norm": 0.4762805104255676, | |
| "learning_rate": 8.310128348054094e-05, | |
| "loss": 0.0254, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.15777525539160045, | |
| "grad_norm": 0.3500935435295105, | |
| "learning_rate": 8.298219961566009e-05, | |
| "loss": 0.0085, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.1583427922814983, | |
| "grad_norm": 0.3653341233730316, | |
| "learning_rate": 8.286278362039528e-05, | |
| "loss": 0.0074, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.15891032917139614, | |
| "grad_norm": 0.2545139193534851, | |
| "learning_rate": 8.274303669726426e-05, | |
| "loss": 0.0049, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.159477866061294, | |
| "grad_norm": 0.0680074691772461, | |
| "learning_rate": 8.262296005211721e-05, | |
| "loss": 0.0014, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.16004540295119182, | |
| "grad_norm": 0.2128506749868393, | |
| "learning_rate": 8.250255489412463e-05, | |
| "loss": 0.0049, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.16061293984108968, | |
| "grad_norm": 0.0370691753923893, | |
| "learning_rate": 8.238182243576512e-05, | |
| "loss": 0.0008, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.1611804767309875, | |
| "grad_norm": 0.05422423034906387, | |
| "learning_rate": 8.226076389281316e-05, | |
| "loss": 0.0009, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.16174801362088537, | |
| "grad_norm": 0.25359049439430237, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 0.002, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.1623155505107832, | |
| "grad_norm": 0.12178357690572739, | |
| "learning_rate": 8.201767343263612e-05, | |
| "loss": 0.0022, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.16288308740068105, | |
| "grad_norm": 0.06412477791309357, | |
| "learning_rate": 8.189564396332928e-05, | |
| "loss": 0.0012, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.16345062429057888, | |
| "grad_norm": 0.04413165897130966, | |
| "learning_rate": 8.177329330524182e-05, | |
| "loss": 0.0009, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.16401816118047674, | |
| "grad_norm": 0.1375110000371933, | |
| "learning_rate": 8.165062269044353e-05, | |
| "loss": 0.0019, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.16458569807037457, | |
| "grad_norm": 0.04058424010872841, | |
| "learning_rate": 8.152763335422613e-05, | |
| "loss": 0.0006, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.16515323496027243, | |
| "grad_norm": 0.46348121762275696, | |
| "learning_rate": 8.140432653509089e-05, | |
| "loss": 0.0038, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.16572077185017026, | |
| "grad_norm": 0.18360164761543274, | |
| "learning_rate": 8.128070347473609e-05, | |
| "loss": 0.0018, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.1662883087400681, | |
| "grad_norm": 0.6381713151931763, | |
| "learning_rate": 8.115676541804456e-05, | |
| "loss": 0.0069, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.16685584562996594, | |
| "grad_norm": 0.11540421843528748, | |
| "learning_rate": 8.103251361307119e-05, | |
| "loss": 0.0011, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.1674233825198638, | |
| "grad_norm": 0.5324001908302307, | |
| "learning_rate": 8.090794931103026e-05, | |
| "loss": 0.0035, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.16799091940976163, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.0, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.1685584562996595, | |
| "grad_norm": 0.12677313387393951, | |
| "learning_rate": 8.065788823632451e-05, | |
| "loss": 0.0018, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.16912599318955732, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.053239398177191e-05, | |
| "loss": 0.0, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.16969353007945517, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.04065922663509e-05, | |
| "loss": 0.0, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.170261066969353, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.028048435688333e-05, | |
| "loss": 0.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.17082860385925086, | |
| "grad_norm": 3.0016751289367676, | |
| "learning_rate": 8.015407152327448e-05, | |
| "loss": 0.2288, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.1713961407491487, | |
| "grad_norm": 1.9297741651535034, | |
| "learning_rate": 8.002735503850016e-05, | |
| "loss": 0.0538, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.17196367763904655, | |
| "grad_norm": 0.6649733185768127, | |
| "learning_rate": 7.990033617859396e-05, | |
| "loss": 0.0204, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.17253121452894438, | |
| "grad_norm": 1.4802879095077515, | |
| "learning_rate": 7.97730162226344e-05, | |
| "loss": 0.0355, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.17309875141884223, | |
| "grad_norm": 1.2510706186294556, | |
| "learning_rate": 7.964539645273204e-05, | |
| "loss": 0.0712, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.17366628830874006, | |
| "grad_norm": 0.45705509185791016, | |
| "learning_rate": 7.95174781540165e-05, | |
| "loss": 0.0259, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.17423382519863792, | |
| "grad_norm": 0.5345933437347412, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.02, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.17480136208853575, | |
| "grad_norm": 0.3882739841938019, | |
| "learning_rate": 7.926075112568259e-05, | |
| "loss": 0.014, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.1753688989784336, | |
| "grad_norm": 0.07783171534538269, | |
| "learning_rate": 7.913194498130252e-05, | |
| "loss": 0.0029, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.17593643586833144, | |
| "grad_norm": 0.39135614037513733, | |
| "learning_rate": 7.900284547855991e-05, | |
| "loss": 0.0127, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.1765039727582293, | |
| "grad_norm": 0.6026532053947449, | |
| "learning_rate": 7.887345391748533e-05, | |
| "loss": 0.003, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.17707150964812712, | |
| "grad_norm": 0.4667406380176544, | |
| "learning_rate": 7.874377160105036e-05, | |
| "loss": 0.008, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.17763904653802498, | |
| "grad_norm": 0.4185452461242676, | |
| "learning_rate": 7.861379983515449e-05, | |
| "loss": 0.0121, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.1782065834279228, | |
| "grad_norm": 0.28240469098091125, | |
| "learning_rate": 7.848353992861195e-05, | |
| "loss": 0.007, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.17877412031782067, | |
| "grad_norm": 0.2165137082338333, | |
| "learning_rate": 7.835299319313853e-05, | |
| "loss": 0.0057, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.1793416572077185, | |
| "grad_norm": 0.4501195549964905, | |
| "learning_rate": 7.822216094333847e-05, | |
| "loss": 0.016, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.17990919409761635, | |
| "grad_norm": 0.21665939688682556, | |
| "learning_rate": 7.809104449669101e-05, | |
| "loss": 0.0053, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.18047673098751418, | |
| "grad_norm": 0.45562973618507385, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.0399, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.18104426787741204, | |
| "grad_norm": 0.2064024657011032, | |
| "learning_rate": 7.78279642970672e-05, | |
| "loss": 0.0033, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.18161180476730987, | |
| "grad_norm": 0.06030461564660072, | |
| "learning_rate": 7.769600319330552e-05, | |
| "loss": 0.0015, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.18217934165720773, | |
| "grad_norm": 0.12329499423503876, | |
| "learning_rate": 7.756376319109917e-05, | |
| "loss": 0.002, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.18274687854710556, | |
| "grad_norm": 0.09843454509973526, | |
| "learning_rate": 7.74312456221035e-05, | |
| "loss": 0.0024, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.18331441543700341, | |
| "grad_norm": 0.3691561222076416, | |
| "learning_rate": 7.729845182076895e-05, | |
| "loss": 0.0087, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.18388195232690124, | |
| "grad_norm": 0.18092408776283264, | |
| "learning_rate": 7.716538312432766e-05, | |
| "loss": 0.0042, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.1844494892167991, | |
| "grad_norm": 0.2996932566165924, | |
| "learning_rate": 7.703204087277988e-05, | |
| "loss": 0.0221, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.18501702610669693, | |
| "grad_norm": 0.12783432006835938, | |
| "learning_rate": 7.689842640888063e-05, | |
| "loss": 0.0033, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.1855845629965948, | |
| "grad_norm": 0.26556456089019775, | |
| "learning_rate": 7.676454107812607e-05, | |
| "loss": 0.0056, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.18615209988649262, | |
| "grad_norm": 0.1686052829027176, | |
| "learning_rate": 7.663038622873999e-05, | |
| "loss": 0.0029, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.18671963677639047, | |
| "grad_norm": 0.04450399801135063, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 0.0007, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.1872871736662883, | |
| "grad_norm": 0.2206043303012848, | |
| "learning_rate": 7.636127338052512e-05, | |
| "loss": 0.0019, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.18785471055618616, | |
| "grad_norm": 0.050599176436662674, | |
| "learning_rate": 7.622631809165973e-05, | |
| "loss": 0.0006, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.188422247446084, | |
| "grad_norm": 1.3266581296920776, | |
| "learning_rate": 7.60910987040623e-05, | |
| "loss": 0.0502, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.18898978433598185, | |
| "grad_norm": 0.012006393633782864, | |
| "learning_rate": 7.595561657939061e-05, | |
| "loss": 0.0004, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.18955732122587968, | |
| "grad_norm": 0.24034585058689117, | |
| "learning_rate": 7.58198730819481e-05, | |
| "loss": 0.0033, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.19012485811577753, | |
| "grad_norm": 0.378845751285553, | |
| "learning_rate": 7.568386957867033e-05, | |
| "loss": 0.0189, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.19069239500567536, | |
| "grad_norm": 0.019944118335843086, | |
| "learning_rate": 7.554760743911103e-05, | |
| "loss": 0.0004, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.19125993189557322, | |
| "grad_norm": 1.7960922718048096, | |
| "learning_rate": 7.541108803542846e-05, | |
| "loss": 0.0138, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.19182746878547105, | |
| "grad_norm": 0.015294855460524559, | |
| "learning_rate": 7.52743127423715e-05, | |
| "loss": 0.0004, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.1923950056753689, | |
| "grad_norm": 0.08567917346954346, | |
| "learning_rate": 7.51372829372658e-05, | |
| "loss": 0.0013, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.19296254256526674, | |
| "grad_norm": 0.07724782824516296, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.0008, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.1935300794551646, | |
| "grad_norm": 0.009706157259643078, | |
| "learning_rate": 7.486246531301177e-05, | |
| "loss": 0.0003, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.19409761634506242, | |
| "grad_norm": 0.04667770117521286, | |
| "learning_rate": 7.472468026127385e-05, | |
| "loss": 0.0004, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.19466515323496028, | |
| "grad_norm": 0.01700473390519619, | |
| "learning_rate": 7.45866462322802e-05, | |
| "loss": 0.0004, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.1952326901248581, | |
| "grad_norm": 0.6124288439750671, | |
| "learning_rate": 7.444836461603195e-05, | |
| "loss": 0.0087, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.19580022701475597, | |
| "grad_norm": 0.4518308937549591, | |
| "learning_rate": 7.430983680502344e-05, | |
| "loss": 0.0028, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.1963677639046538, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.417106419422819e-05, | |
| "loss": 0.0, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.19693530079455165, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.403204818108487e-05, | |
| "loss": 0.0, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.19750283768444948, | |
| "grad_norm": 0.14064697921276093, | |
| "learning_rate": 7.389279016548316e-05, | |
| "loss": 0.0015, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.19807037457434734, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.375329154974975e-05, | |
| "loss": 0.0, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.19863791146424517, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.361355373863414e-05, | |
| "loss": 0.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.19920544835414303, | |
| "grad_norm": 2.25301194190979, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 0.0754, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.19977298524404086, | |
| "grad_norm": 1.7841074466705322, | |
| "learning_rate": 7.333336616128369e-05, | |
| "loss": 0.051, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.20034052213393871, | |
| "grad_norm": 0.9712184071540833, | |
| "learning_rate": 7.319291921653464e-05, | |
| "loss": 0.016, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.20090805902383654, | |
| "grad_norm": 0.5902833342552185, | |
| "learning_rate": 7.305223871934657e-05, | |
| "loss": 0.0106, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.2014755959137344, | |
| "grad_norm": 0.2831375002861023, | |
| "learning_rate": 7.291132608637052e-05, | |
| "loss": 0.0077, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.20204313280363223, | |
| "grad_norm": 0.5281710624694824, | |
| "learning_rate": 7.277018273659517e-05, | |
| "loss": 0.0236, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.2026106696935301, | |
| "grad_norm": 0.7202128767967224, | |
| "learning_rate": 7.262881009133242e-05, | |
| "loss": 0.0256, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.20317820658342792, | |
| "grad_norm": 0.2578269839286804, | |
| "learning_rate": 7.24872095742033e-05, | |
| "loss": 0.0061, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.20374574347332577, | |
| "grad_norm": 0.21329842507839203, | |
| "learning_rate": 7.23453826111234e-05, | |
| "loss": 0.0029, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.2043132803632236, | |
| "grad_norm": 0.3269996643066406, | |
| "learning_rate": 7.220333063028872e-05, | |
| "loss": 0.0042, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.20488081725312146, | |
| "grad_norm": 0.5432631373405457, | |
| "learning_rate": 7.206105506216106e-05, | |
| "loss": 0.0272, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.2054483541430193, | |
| "grad_norm": 1.9401220083236694, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.0163, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.20601589103291715, | |
| "grad_norm": 0.24572275578975677, | |
| "learning_rate": 7.177583889711762e-05, | |
| "loss": 0.006, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.20658342792281498, | |
| "grad_norm": 0.13575679063796997, | |
| "learning_rate": 7.163290117232542e-05, | |
| "loss": 0.0026, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.20715096481271283, | |
| "grad_norm": 0.4401944577693939, | |
| "learning_rate": 7.148974560445859e-05, | |
| "loss": 0.0187, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.20771850170261066, | |
| "grad_norm": 0.45200228691101074, | |
| "learning_rate": 7.13463736350921e-05, | |
| "loss": 0.0083, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.20828603859250852, | |
| "grad_norm": 0.5528292655944824, | |
| "learning_rate": 7.120278670798009e-05, | |
| "loss": 0.0092, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.20885357548240635, | |
| "grad_norm": 0.5644862651824951, | |
| "learning_rate": 7.105898626904134e-05, | |
| "loss": 0.0147, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.2094211123723042, | |
| "grad_norm": 0.7960838675498962, | |
| "learning_rate": 7.091497376634464e-05, | |
| "loss": 0.0252, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.20998864926220204, | |
| "grad_norm": 0.26013273000717163, | |
| "learning_rate": 7.077075065009433e-05, | |
| "loss": 0.003, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2105561861520999, | |
| "grad_norm": 0.44845283031463623, | |
| "learning_rate": 7.062631837261557e-05, | |
| "loss": 0.005, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.21112372304199772, | |
| "grad_norm": 0.48455584049224854, | |
| "learning_rate": 7.048167838833977e-05, | |
| "loss": 0.0084, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.21169125993189558, | |
| "grad_norm": 0.48875439167022705, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 0.0034, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.2122587968217934, | |
| "grad_norm": 1.5464515686035156, | |
| "learning_rate": 7.019178112756624e-05, | |
| "loss": 0.0422, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.21282633371169127, | |
| "grad_norm": 0.08226021379232407, | |
| "learning_rate": 7.004652677033068e-05, | |
| "loss": 0.0014, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.2133938706015891, | |
| "grad_norm": 0.9428783655166626, | |
| "learning_rate": 6.990107054479312e-05, | |
| "loss": 0.0085, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.21396140749148695, | |
| "grad_norm": 0.5083039999008179, | |
| "learning_rate": 6.97554139156961e-05, | |
| "loss": 0.0054, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.21452894438138478, | |
| "grad_norm": 0.7152851819992065, | |
| "learning_rate": 6.960955834980028e-05, | |
| "loss": 0.0118, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.21509648127128264, | |
| "grad_norm": 0.7423697113990784, | |
| "learning_rate": 6.946350531586959e-05, | |
| "loss": 0.0196, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.21566401816118047, | |
| "grad_norm": 0.19148842990398407, | |
| "learning_rate": 6.931725628465643e-05, | |
| "loss": 0.0029, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.21623155505107833, | |
| "grad_norm": 0.16525211930274963, | |
| "learning_rate": 6.917081272888697e-05, | |
| "loss": 0.0018, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.21679909194097616, | |
| "grad_norm": 0.7403731942176819, | |
| "learning_rate": 6.902417612324615e-05, | |
| "loss": 0.0156, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.21736662883087401, | |
| "grad_norm": 0.7522996068000793, | |
| "learning_rate": 6.8877347944363e-05, | |
| "loss": 0.0106, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.21793416572077184, | |
| "grad_norm": 0.35488778352737427, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.0266, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.2185017026106697, | |
| "grad_norm": 0.5787685513496399, | |
| "learning_rate": 6.858312278301637e-05, | |
| "loss": 0.0044, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.21906923950056753, | |
| "grad_norm": 0.5935757756233215, | |
| "learning_rate": 6.843572876339705e-05, | |
| "loss": 0.0101, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.2196367763904654, | |
| "grad_norm": 0.0742402896285057, | |
| "learning_rate": 6.828814909619373e-05, | |
| "loss": 0.0006, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.22020431328036322, | |
| "grad_norm": 0.08480936288833618, | |
| "learning_rate": 6.814038526753205e-05, | |
| "loss": 0.0014, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.22077185017026107, | |
| "grad_norm": 0.20840072631835938, | |
| "learning_rate": 6.799243876539212e-05, | |
| "loss": 0.0017, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.2213393870601589, | |
| "grad_norm": 0.050797827541828156, | |
| "learning_rate": 6.784431107959359e-05, | |
| "loss": 0.0007, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.22190692395005676, | |
| "grad_norm": 0.0777006596326828, | |
| "learning_rate": 6.769600370178059e-05, | |
| "loss": 0.0013, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.2224744608399546, | |
| "grad_norm": 0.04812987521290779, | |
| "learning_rate": 6.754751812540679e-05, | |
| "loss": 0.0008, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.22304199772985245, | |
| "grad_norm": 0.07565217465162277, | |
| "learning_rate": 6.739885584572026e-05, | |
| "loss": 0.0006, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.22360953461975028, | |
| "grad_norm": 0.03945764899253845, | |
| "learning_rate": 6.725001835974853e-05, | |
| "loss": 0.0005, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.22417707150964813, | |
| "grad_norm": 1.1225963830947876, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 0.0077, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.22474460839954596, | |
| "grad_norm": 0.24701376259326935, | |
| "learning_rate": 6.695182376586603e-05, | |
| "loss": 0.0024, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.22531214528944382, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.680246966077151e-05, | |
| "loss": 0.0, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.22587968217934165, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.665294635499404e-05, | |
| "loss": 0.0, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.2264472190692395, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.650325535423167e-05, | |
| "loss": 0.0, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.22701475595913734, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.635339816587109e-05, | |
| "loss": 0.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.22701475595913734, | |
| "eval_loss": NaN, | |
| "eval_runtime": 106.1666, | |
| "eval_samples_per_second": 27.956, | |
| "eval_steps_per_second": 6.989, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2275822928490352, | |
| "grad_norm": 1.7875468730926514, | |
| "learning_rate": 6.620337629897254e-05, | |
| "loss": 0.1379, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.22814982973893302, | |
| "grad_norm": 1.6237751245498657, | |
| "learning_rate": 6.605319126425454e-05, | |
| "loss": 0.02, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.22871736662883088, | |
| "grad_norm": 0.16924582421779633, | |
| "learning_rate": 6.590284457407876e-05, | |
| "loss": 0.0052, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.2292849035187287, | |
| "grad_norm": 0.3651062250137329, | |
| "learning_rate": 6.575233774243465e-05, | |
| "loss": 0.0265, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.22985244040862657, | |
| "grad_norm": 0.13029779493808746, | |
| "learning_rate": 6.560167228492436e-05, | |
| "loss": 0.0032, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.2304199772985244, | |
| "grad_norm": 0.11845195293426514, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.0039, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.23098751418842225, | |
| "grad_norm": 0.344365656375885, | |
| "learning_rate": 6.529987156268526e-05, | |
| "loss": 0.0043, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.23155505107832008, | |
| "grad_norm": 0.8041085004806519, | |
| "learning_rate": 6.514873933708638e-05, | |
| "loss": 0.0077, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.23212258796821794, | |
| "grad_norm": 0.05168134719133377, | |
| "learning_rate": 6.499745456385054e-05, | |
| "loss": 0.0016, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.23269012485811577, | |
| "grad_norm": 0.27078044414520264, | |
| "learning_rate": 6.484601876641375e-05, | |
| "loss": 0.0126, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.23325766174801363, | |
| "grad_norm": 0.28879401087760925, | |
| "learning_rate": 6.46944334697328e-05, | |
| "loss": 0.0115, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.23382519863791146, | |
| "grad_norm": 0.9801868200302124, | |
| "learning_rate": 6.454270020026995e-05, | |
| "loss": 0.0165, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.23439273552780931, | |
| "grad_norm": 0.28464144468307495, | |
| "learning_rate": 6.439082048597755e-05, | |
| "loss": 0.0039, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.23496027241770714, | |
| "grad_norm": 0.07441152632236481, | |
| "learning_rate": 6.423879585628261e-05, | |
| "loss": 0.0013, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.235527809307605, | |
| "grad_norm": 0.4811006188392639, | |
| "learning_rate": 6.408662784207149e-05, | |
| "loss": 0.0168, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.23609534619750283, | |
| "grad_norm": 0.15365412831306458, | |
| "learning_rate": 6.39343179756744e-05, | |
| "loss": 0.0032, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.2366628830874007, | |
| "grad_norm": 0.1440768986940384, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 0.0026, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.23723041997729852, | |
| "grad_norm": 0.12057225406169891, | |
| "learning_rate": 6.36292788227699e-05, | |
| "loss": 0.0019, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.23779795686719638, | |
| "grad_norm": 0.41369637846946716, | |
| "learning_rate": 6.34765526080034e-05, | |
| "loss": 0.0114, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.2383654937570942, | |
| "grad_norm": 0.06439776718616486, | |
| "learning_rate": 6.332369068450174e-05, | |
| "loss": 0.0015, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.23893303064699206, | |
| "grad_norm": 0.05166054517030716, | |
| "learning_rate": 6.317069459158284e-05, | |
| "loss": 0.0012, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.2395005675368899, | |
| "grad_norm": 0.040280867367982864, | |
| "learning_rate": 6.30175658699156e-05, | |
| "loss": 0.0011, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.24006810442678775, | |
| "grad_norm": 0.2597777843475342, | |
| "learning_rate": 6.286430606150459e-05, | |
| "loss": 0.0034, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.24063564131668558, | |
| "grad_norm": 0.34109053015708923, | |
| "learning_rate": 6.271091670967436e-05, | |
| "loss": 0.0055, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.24120317820658344, | |
| "grad_norm": 0.6494819521903992, | |
| "learning_rate": 6.255739935905396e-05, | |
| "loss": 0.0092, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.24177071509648126, | |
| "grad_norm": 0.6631916165351868, | |
| "learning_rate": 6.240375555556145e-05, | |
| "loss": 0.016, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.24233825198637912, | |
| "grad_norm": 0.23462681472301483, | |
| "learning_rate": 6.22499868463882e-05, | |
| "loss": 0.0034, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.24290578887627695, | |
| "grad_norm": 0.03203234449028969, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.0007, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.2434733257661748, | |
| "grad_norm": 0.08686164021492004, | |
| "learning_rate": 6.194208090603844e-05, | |
| "loss": 0.0013, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.24404086265607264, | |
| "grad_norm": 0.2361176759004593, | |
| "learning_rate": 6.178794677547137e-05, | |
| "loss": 0.0025, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.2446083995459705, | |
| "grad_norm": 0.30464842915534973, | |
| "learning_rate": 6.163369394041111e-05, | |
| "loss": 0.0026, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.24517593643586832, | |
| "grad_norm": 1.3310155868530273, | |
| "learning_rate": 6.147932395418205e-05, | |
| "loss": 0.0677, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.24574347332576618, | |
| "grad_norm": 0.006984487175941467, | |
| "learning_rate": 6.132483837128823e-05, | |
| "loss": 0.0002, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.246311010215664, | |
| "grad_norm": 0.7241768836975098, | |
| "learning_rate": 6.117023874739772e-05, | |
| "loss": 0.0213, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.24687854710556187, | |
| "grad_norm": 0.06802449375391006, | |
| "learning_rate": 6.1015526639327035e-05, | |
| "loss": 0.0007, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.2474460839954597, | |
| "grad_norm": 0.22861234843730927, | |
| "learning_rate": 6.0860703605025395e-05, | |
| "loss": 0.0027, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.24801362088535756, | |
| "grad_norm": 0.09795883297920227, | |
| "learning_rate": 6.0705771203559024e-05, | |
| "loss": 0.0008, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.24858115777525538, | |
| "grad_norm": 0.9199258685112, | |
| "learning_rate": 6.05507309950955e-05, | |
| "loss": 0.0459, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.24914869466515324, | |
| "grad_norm": 1.980196475982666, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 0.0015, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.24971623155505107, | |
| "grad_norm": 0.2596800625324249, | |
| "learning_rate": 6.024033340325954e-05, | |
| "loss": 0.0033, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.25028376844494893, | |
| "grad_norm": 0.9777965545654297, | |
| "learning_rate": 6.008497914558744e-05, | |
| "loss": 0.017, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.2508513053348468, | |
| "grad_norm": 0.36498022079467773, | |
| "learning_rate": 5.992952333228728e-05, | |
| "loss": 0.0042, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.2514188422247446, | |
| "grad_norm": 0.022741034626960754, | |
| "learning_rate": 5.9773967528797414e-05, | |
| "loss": 0.0004, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.25198637911464244, | |
| "grad_norm": 0.27772486209869385, | |
| "learning_rate": 5.9618313301563055e-05, | |
| "loss": 0.0032, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.2525539160045403, | |
| "grad_norm": 0.0808940976858139, | |
| "learning_rate": 5.946256221802051e-05, | |
| "loss": 0.0002, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.25312145289443816, | |
| "grad_norm": 0.27334359288215637, | |
| "learning_rate": 5.9306715846581506e-05, | |
| "loss": 0.0004, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.25368898978433596, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.915077575661723e-05, | |
| "loss": 0.0, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.2542565266742338, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.8994743518442694e-05, | |
| "loss": 0.0, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.2548240635641317, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.8838620703300784e-05, | |
| "loss": 0.0, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.25539160045402953, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.25595913734392733, | |
| "grad_norm": 1.7918459177017212, | |
| "learning_rate": 5.85261096316312e-05, | |
| "loss": 0.1334, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.2565266742338252, | |
| "grad_norm": 1.2250529527664185, | |
| "learning_rate": 5.836972452208654e-05, | |
| "loss": 0.0459, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.25709421112372305, | |
| "grad_norm": 0.15675216913223267, | |
| "learning_rate": 5.821325512950886e-05, | |
| "loss": 0.0036, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.2576617480136209, | |
| "grad_norm": 0.18163926899433136, | |
| "learning_rate": 5.805670302954321e-05, | |
| "loss": 0.0018, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.2582292849035187, | |
| "grad_norm": 0.2342452108860016, | |
| "learning_rate": 5.79000697986675e-05, | |
| "loss": 0.0119, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.25879682179341656, | |
| "grad_norm": 0.7253749370574951, | |
| "learning_rate": 5.7743357014176624e-05, | |
| "loss": 0.0193, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.2593643586833144, | |
| "grad_norm": 0.45949435234069824, | |
| "learning_rate": 5.7586566254166583e-05, | |
| "loss": 0.0124, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.2599318955732123, | |
| "grad_norm": 0.05333389714360237, | |
| "learning_rate": 5.7429699097518585e-05, | |
| "loss": 0.0021, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.2604994324631101, | |
| "grad_norm": 0.22825823724269867, | |
| "learning_rate": 5.7272757123883184e-05, | |
| "loss": 0.0032, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.26106696935300794, | |
| "grad_norm": 0.35596853494644165, | |
| "learning_rate": 5.7115741913664264e-05, | |
| "loss": 0.0059, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.2616345062429058, | |
| "grad_norm": 0.4787977635860443, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 0.0123, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.26220204313280365, | |
| "grad_norm": 0.15637889504432678, | |
| "learning_rate": 5.680149810876322e-05, | |
| "loss": 0.0022, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.26276958002270145, | |
| "grad_norm": 0.1934584677219391, | |
| "learning_rate": 5.664427267851271e-05, | |
| "loss": 0.0058, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.2633371169125993, | |
| "grad_norm": 0.08029989898204803, | |
| "learning_rate": 5.6486980340510086e-05, | |
| "loss": 0.0021, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.26390465380249717, | |
| "grad_norm": 0.21124523878097534, | |
| "learning_rate": 5.6329622678687463e-05, | |
| "loss": 0.0061, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.264472190692395, | |
| "grad_norm": 0.05409558117389679, | |
| "learning_rate": 5.617220127763474e-05, | |
| "loss": 0.0013, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.26503972758229283, | |
| "grad_norm": 0.5069450736045837, | |
| "learning_rate": 5.601471772258368e-05, | |
| "loss": 0.0084, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.2656072644721907, | |
| "grad_norm": 0.30205589532852173, | |
| "learning_rate": 5.585717359939192e-05, | |
| "loss": 0.0114, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.26617480136208854, | |
| "grad_norm": 0.13869255781173706, | |
| "learning_rate": 5.569957049452703e-05, | |
| "loss": 0.0011, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.2667423382519864, | |
| "grad_norm": 0.048545245081186295, | |
| "learning_rate": 5.5541909995050554e-05, | |
| "loss": 0.001, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2673098751418842, | |
| "grad_norm": 0.01697002351284027, | |
| "learning_rate": 5.538419368860196e-05, | |
| "loss": 0.0006, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.26787741203178206, | |
| "grad_norm": 0.3081216812133789, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.0097, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.2684449489216799, | |
| "grad_norm": 0.1498224139213562, | |
| "learning_rate": 5.506860000814017e-05, | |
| "loss": 0.0024, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.2690124858115778, | |
| "grad_norm": 0.039433132857084274, | |
| "learning_rate": 5.4910725812151864e-05, | |
| "loss": 0.0007, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.2695800227014756, | |
| "grad_norm": 0.0603884682059288, | |
| "learning_rate": 5.475280216520913e-05, | |
| "loss": 0.0012, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.27014755959137343, | |
| "grad_norm": 0.8099268078804016, | |
| "learning_rate": 5.4594830657601384e-05, | |
| "loss": 0.0319, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.2707150964812713, | |
| "grad_norm": 0.23851421475410461, | |
| "learning_rate": 5.443681288009991e-05, | |
| "loss": 0.007, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.27128263337116915, | |
| "grad_norm": 0.09010318666696548, | |
| "learning_rate": 5.427875042394199e-05, | |
| "loss": 0.0015, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.27185017026106695, | |
| "grad_norm": 0.29702991247177124, | |
| "learning_rate": 5.412064488081482e-05, | |
| "loss": 0.0017, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.2724177071509648, | |
| "grad_norm": 0.46306195855140686, | |
| "learning_rate": 5.396249784283942e-05, | |
| "loss": 0.0063, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.27298524404086266, | |
| "grad_norm": 0.40036702156066895, | |
| "learning_rate": 5.3804310902554754e-05, | |
| "loss": 0.0078, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.2735527809307605, | |
| "grad_norm": 0.014721478335559368, | |
| "learning_rate": 5.364608565290155e-05, | |
| "loss": 0.0003, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.2741203178206583, | |
| "grad_norm": 0.43365278840065, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 0.0272, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.2746878547105562, | |
| "grad_norm": 0.8489099144935608, | |
| "learning_rate": 5.3329526599165204e-05, | |
| "loss": 0.0232, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.27525539160045404, | |
| "grad_norm": 0.22350762784481049, | |
| "learning_rate": 5.317119598282823e-05, | |
| "loss": 0.005, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.2758229284903519, | |
| "grad_norm": 0.028759444132447243, | |
| "learning_rate": 5.301283343258293e-05, | |
| "loss": 0.0007, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.2763904653802497, | |
| "grad_norm": 0.0732959434390068, | |
| "learning_rate": 5.2854440543138406e-05, | |
| "loss": 0.0015, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.27695800227014755, | |
| "grad_norm": 0.19646039605140686, | |
| "learning_rate": 5.2696018909509306e-05, | |
| "loss": 0.0036, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.2775255391600454, | |
| "grad_norm": 0.19537141919136047, | |
| "learning_rate": 5.253757012699972e-05, | |
| "loss": 0.0038, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.27809307604994327, | |
| "grad_norm": 0.18001757562160492, | |
| "learning_rate": 5.2379095791187124e-05, | |
| "loss": 0.0029, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.27866061293984107, | |
| "grad_norm": 0.05839482694864273, | |
| "learning_rate": 5.2220597497906307e-05, | |
| "loss": 0.0013, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.2792281498297389, | |
| "grad_norm": 0.0680989921092987, | |
| "learning_rate": 5.2062076843233366e-05, | |
| "loss": 0.0016, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.2797956867196368, | |
| "grad_norm": 0.03857843577861786, | |
| "learning_rate": 5.1903535423469505e-05, | |
| "loss": 0.0009, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.28036322360953464, | |
| "grad_norm": 0.2856152355670929, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.0027, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.28093076049943244, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.158639667490339e-05, | |
| "loss": 0.0, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.2814982973893303, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.142780253968481e-05, | |
| "loss": 0.0, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.28206583427922816, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.126919402651052e-05, | |
| "loss": 0.0, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.282633371169126, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.1110572732566475e-05, | |
| "loss": 0.0, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.2832009080590238, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.095194025516733e-05, | |
| "loss": 0.0, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.28376844494892167, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.0793298191740404e-05, | |
| "loss": 0.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.28433598183881953, | |
| "grad_norm": 3.819697856903076, | |
| "learning_rate": 5.063464813980948e-05, | |
| "loss": 0.271, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.2849035187287174, | |
| "grad_norm": 0.8054677844047546, | |
| "learning_rate": 5.047599169697884e-05, | |
| "loss": 0.0246, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.2854710556186152, | |
| "grad_norm": 0.3912200629711151, | |
| "learning_rate": 5.03173304609171e-05, | |
| "loss": 0.0167, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.28603859250851305, | |
| "grad_norm": 0.3080594539642334, | |
| "learning_rate": 5.015866602934112e-05, | |
| "loss": 0.0099, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.2866061293984109, | |
| "grad_norm": 0.6711567044258118, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0091, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.28717366628830876, | |
| "grad_norm": 0.41851040720939636, | |
| "learning_rate": 4.984133397065889e-05, | |
| "loss": 0.0136, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.28774120317820656, | |
| "grad_norm": 0.15193375945091248, | |
| "learning_rate": 4.968266953908292e-05, | |
| "loss": 0.0029, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.2883087400681044, | |
| "grad_norm": 0.27360770106315613, | |
| "learning_rate": 4.952400830302117e-05, | |
| "loss": 0.0049, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.2888762769580023, | |
| "grad_norm": 0.06797119230031967, | |
| "learning_rate": 4.9365351860190526e-05, | |
| "loss": 0.0012, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.28944381384790013, | |
| "grad_norm": 0.42943570017814636, | |
| "learning_rate": 4.92067018082596e-05, | |
| "loss": 0.007, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.29001135073779793, | |
| "grad_norm": 0.41933485865592957, | |
| "learning_rate": 4.9048059744832666e-05, | |
| "loss": 0.024, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.2905788876276958, | |
| "grad_norm": 0.07279060781002045, | |
| "learning_rate": 4.888942726743353e-05, | |
| "loss": 0.0014, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.29114642451759365, | |
| "grad_norm": 0.022339163348078728, | |
| "learning_rate": 4.8730805973489476e-05, | |
| "loss": 0.0007, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.2917139614074915, | |
| "grad_norm": 0.033492498099803925, | |
| "learning_rate": 4.85721974603152e-05, | |
| "loss": 0.0011, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.2922814982973893, | |
| "grad_norm": 0.15747897326946259, | |
| "learning_rate": 4.841360332509663e-05, | |
| "loss": 0.0023, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.29284903518728717, | |
| "grad_norm": 0.4097544252872467, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.0324, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.293416572077185, | |
| "grad_norm": 0.39055225253105164, | |
| "learning_rate": 4.8096464576530507e-05, | |
| "loss": 0.0043, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.2939841089670829, | |
| "grad_norm": 0.011669347062706947, | |
| "learning_rate": 4.7937923156766646e-05, | |
| "loss": 0.0005, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.2945516458569807, | |
| "grad_norm": 0.15535907447338104, | |
| "learning_rate": 4.77794025020937e-05, | |
| "loss": 0.0019, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.29511918274687854, | |
| "grad_norm": 0.042138513177633286, | |
| "learning_rate": 4.762090420881289e-05, | |
| "loss": 0.0009, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.2956867196367764, | |
| "grad_norm": 0.24504394829273224, | |
| "learning_rate": 4.7462429873000295e-05, | |
| "loss": 0.0047, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.29625425652667425, | |
| "grad_norm": 0.08995606750249863, | |
| "learning_rate": 4.730398109049071e-05, | |
| "loss": 0.0018, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.29682179341657206, | |
| "grad_norm": 0.020904161036014557, | |
| "learning_rate": 4.71455594568616e-05, | |
| "loss": 0.0004, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.2973893303064699, | |
| "grad_norm": 0.028654640540480614, | |
| "learning_rate": 4.698716656741708e-05, | |
| "loss": 0.0007, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.29795686719636777, | |
| "grad_norm": 0.05848681926727295, | |
| "learning_rate": 4.6828804017171776e-05, | |
| "loss": 0.0009, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.2985244040862656, | |
| "grad_norm": 0.3643423020839691, | |
| "learning_rate": 4.667047340083481e-05, | |
| "loss": 0.0033, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.29909194097616343, | |
| "grad_norm": 0.12494904547929764, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 0.0018, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.2996594778660613, | |
| "grad_norm": 0.03870720416307449, | |
| "learning_rate": 4.635391434709847e-05, | |
| "loss": 0.0008, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.30022701475595914, | |
| "grad_norm": 0.8066434264183044, | |
| "learning_rate": 4.619568909744524e-05, | |
| "loss": 0.0234, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.300794551645857, | |
| "grad_norm": 0.04658526927232742, | |
| "learning_rate": 4.603750215716057e-05, | |
| "loss": 0.0006, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.3013620885357548, | |
| "grad_norm": 0.362132728099823, | |
| "learning_rate": 4.587935511918521e-05, | |
| "loss": 0.04, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.30192962542565266, | |
| "grad_norm": 0.06521368026733398, | |
| "learning_rate": 4.5721249576058027e-05, | |
| "loss": 0.0009, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.3024971623155505, | |
| "grad_norm": 0.029287142679095268, | |
| "learning_rate": 4.5563187119900104e-05, | |
| "loss": 0.0005, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.3030646992054484, | |
| "grad_norm": 0.1905515044927597, | |
| "learning_rate": 4.5405169342398634e-05, | |
| "loss": 0.0018, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.3036322360953462, | |
| "grad_norm": 0.3800831735134125, | |
| "learning_rate": 4.5247197834790876e-05, | |
| "loss": 0.0213, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.30419977298524403, | |
| "grad_norm": 0.02702210657298565, | |
| "learning_rate": 4.508927418784815e-05, | |
| "loss": 0.0005, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.3047673098751419, | |
| "grad_norm": 0.021253688260912895, | |
| "learning_rate": 4.493139999185983e-05, | |
| "loss": 0.0004, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.30533484676503975, | |
| "grad_norm": 0.019332874566316605, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.0004, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.30590238365493755, | |
| "grad_norm": 0.20063234865665436, | |
| "learning_rate": 4.461580631139805e-05, | |
| "loss": 0.0033, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.3064699205448354, | |
| "grad_norm": 0.031181707978248596, | |
| "learning_rate": 4.445809000494946e-05, | |
| "loss": 0.0005, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.30703745743473326, | |
| "grad_norm": 0.6691449284553528, | |
| "learning_rate": 4.4300429505472976e-05, | |
| "loss": 0.0271, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.3076049943246311, | |
| "grad_norm": 0.009593677707016468, | |
| "learning_rate": 4.4142826400608086e-05, | |
| "loss": 0.0002, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.3081725312145289, | |
| "grad_norm": 0.009516783058643341, | |
| "learning_rate": 4.398528227741633e-05, | |
| "loss": 0.0003, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.3087400681044268, | |
| "grad_norm": 0.05990798771381378, | |
| "learning_rate": 4.3827798722365264e-05, | |
| "loss": 0.001, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.30930760499432464, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.3670377321312535e-05, | |
| "loss": 0.0, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.3098751418842225, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.351301965948991e-05, | |
| "loss": 0.0, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.3104426787741203, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.33557273214873e-05, | |
| "loss": 0.0, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.31101021566401815, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.3198501891236804e-05, | |
| "loss": 0.0, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.311577752553916, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 0.0, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.31214528944381387, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.288425808633575e-05, | |
| "loss": 0.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.31271282633371167, | |
| "grad_norm": 0.8620555400848389, | |
| "learning_rate": 4.272724287611684e-05, | |
| "loss": 0.0271, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.3132803632236095, | |
| "grad_norm": 0.25762560963630676, | |
| "learning_rate": 4.2570300902481426e-05, | |
| "loss": 0.0042, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.3138479001135074, | |
| "grad_norm": 0.15001147985458374, | |
| "learning_rate": 4.241343374583343e-05, | |
| "loss": 0.0047, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.31441543700340524, | |
| "grad_norm": 0.2072601318359375, | |
| "learning_rate": 4.2256642985823395e-05, | |
| "loss": 0.0076, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.31498297389330304, | |
| "grad_norm": 0.19115956127643585, | |
| "learning_rate": 4.20999302013325e-05, | |
| "loss": 0.004, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.3155505107832009, | |
| "grad_norm": 0.19318291544914246, | |
| "learning_rate": 4.19432969704568e-05, | |
| "loss": 0.0067, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.31611804767309876, | |
| "grad_norm": 0.0738341435790062, | |
| "learning_rate": 4.178674487049116e-05, | |
| "loss": 0.0015, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.3166855845629966, | |
| "grad_norm": 0.3997495174407959, | |
| "learning_rate": 4.163027547791347e-05, | |
| "loss": 0.013, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.3172531214528944, | |
| "grad_norm": 0.10490912199020386, | |
| "learning_rate": 4.147389036836881e-05, | |
| "loss": 0.002, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.3178206583427923, | |
| "grad_norm": 0.3020564615726471, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.0184, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.31838819523269013, | |
| "grad_norm": 0.1874658763408661, | |
| "learning_rate": 4.116137929669921e-05, | |
| "loss": 0.0056, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.318955732122588, | |
| "grad_norm": 0.1652912199497223, | |
| "learning_rate": 4.100525648155731e-05, | |
| "loss": 0.0029, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.3195232690124858, | |
| "grad_norm": 0.2240392118692398, | |
| "learning_rate": 4.084922424338277e-05, | |
| "loss": 0.0084, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.32009080590238365, | |
| "grad_norm": 0.3925991654396057, | |
| "learning_rate": 4.06932841534185e-05, | |
| "loss": 0.0035, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.3206583427922815, | |
| "grad_norm": 0.23100757598876953, | |
| "learning_rate": 4.0537437781979506e-05, | |
| "loss": 0.0054, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.32122587968217936, | |
| "grad_norm": 0.05905711650848389, | |
| "learning_rate": 4.038168669843697e-05, | |
| "loss": 0.0007, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.32179341657207716, | |
| "grad_norm": 0.26876482367515564, | |
| "learning_rate": 4.0226032471202604e-05, | |
| "loss": 0.0166, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.322360953461975, | |
| "grad_norm": 0.5311969518661499, | |
| "learning_rate": 4.007047666771274e-05, | |
| "loss": 0.0041, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.3229284903518729, | |
| "grad_norm": 0.0217901561409235, | |
| "learning_rate": 3.991502085441259e-05, | |
| "loss": 0.0006, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.32349602724177073, | |
| "grad_norm": 0.04868500307202339, | |
| "learning_rate": 3.9759666596740476e-05, | |
| "loss": 0.0009, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.32406356413166854, | |
| "grad_norm": 0.01934129185974598, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 0.0005, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.3246311010215664, | |
| "grad_norm": 0.3002278208732605, | |
| "learning_rate": 3.944926900490452e-05, | |
| "loss": 0.006, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.32519863791146425, | |
| "grad_norm": 0.08359342068433762, | |
| "learning_rate": 3.929422879644099e-05, | |
| "loss": 0.0008, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.3257661748013621, | |
| "grad_norm": 0.05703277885913849, | |
| "learning_rate": 3.913929639497462e-05, | |
| "loss": 0.0006, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.3263337116912599, | |
| "grad_norm": 0.1300325244665146, | |
| "learning_rate": 3.898447336067297e-05, | |
| "loss": 0.0007, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.32690124858115777, | |
| "grad_norm": 0.2558203935623169, | |
| "learning_rate": 3.882976125260229e-05, | |
| "loss": 0.0116, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.3274687854710556, | |
| "grad_norm": 0.12041133642196655, | |
| "learning_rate": 3.8675161628711776e-05, | |
| "loss": 0.0024, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.3280363223609535, | |
| "grad_norm": 0.01596921868622303, | |
| "learning_rate": 3.852067604581794e-05, | |
| "loss": 0.0005, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.3286038592508513, | |
| "grad_norm": 0.35162538290023804, | |
| "learning_rate": 3.836630605958888e-05, | |
| "loss": 0.017, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.32917139614074914, | |
| "grad_norm": 0.43673884868621826, | |
| "learning_rate": 3.821205322452863e-05, | |
| "loss": 0.0116, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.329738933030647, | |
| "grad_norm": 0.9133800268173218, | |
| "learning_rate": 3.8057919093961553e-05, | |
| "loss": 0.0122, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.33030646992054485, | |
| "grad_norm": 0.024000134319067, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.0007, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.33087400681044266, | |
| "grad_norm": 0.934701681137085, | |
| "learning_rate": 3.775001315361183e-05, | |
| "loss": 0.0058, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.3314415437003405, | |
| "grad_norm": 0.904114842414856, | |
| "learning_rate": 3.759624444443858e-05, | |
| "loss": 0.036, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.33200908059023837, | |
| "grad_norm": 0.07346749305725098, | |
| "learning_rate": 3.744260064094604e-05, | |
| "loss": 0.0009, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.3325766174801362, | |
| "grad_norm": 0.03927430883049965, | |
| "learning_rate": 3.728908329032567e-05, | |
| "loss": 0.0008, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.33314415437003403, | |
| "grad_norm": 0.42941439151763916, | |
| "learning_rate": 3.713569393849543e-05, | |
| "loss": 0.015, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.3337116912599319, | |
| "grad_norm": 0.03521761670708656, | |
| "learning_rate": 3.69824341300844e-05, | |
| "loss": 0.0008, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.33427922814982974, | |
| "grad_norm": 0.019192036241292953, | |
| "learning_rate": 3.6829305408417166e-05, | |
| "loss": 0.0005, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.3348467650397276, | |
| "grad_norm": 0.07493746280670166, | |
| "learning_rate": 3.6676309315498256e-05, | |
| "loss": 0.0011, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.3354143019296254, | |
| "grad_norm": 0.43895918130874634, | |
| "learning_rate": 3.6523447391996614e-05, | |
| "loss": 0.0059, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.33598183881952326, | |
| "grad_norm": 0.6594648361206055, | |
| "learning_rate": 3.6370721177230116e-05, | |
| "loss": 0.0177, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.3365493757094211, | |
| "grad_norm": 0.12148375064134598, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 0.0016, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.337116912599319, | |
| "grad_norm": 0.033791981637477875, | |
| "learning_rate": 3.606568202432562e-05, | |
| "loss": 0.0006, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.3376844494892168, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.591337215792852e-05, | |
| "loss": 0.0, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.33825198637911463, | |
| "grad_norm": 0.12862201035022736, | |
| "learning_rate": 3.5761204143717385e-05, | |
| "loss": 0.0021, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.3388195232690125, | |
| "grad_norm": 0.05550703406333923, | |
| "learning_rate": 3.560917951402245e-05, | |
| "loss": 0.0005, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.33938706015891035, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.545729979973005e-05, | |
| "loss": 0.0, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.33995459704880815, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.530556653026721e-05, | |
| "loss": 0.0, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.340522133938706, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.515398123358627e-05, | |
| "loss": 0.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.340522133938706, | |
| "eval_loss": NaN, | |
| "eval_runtime": 107.34, | |
| "eval_samples_per_second": 27.65, | |
| "eval_steps_per_second": 6.913, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.34108967082860386, | |
| "grad_norm": 0.9571943283081055, | |
| "learning_rate": 3.5002545436149474e-05, | |
| "loss": 0.0473, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.3416572077185017, | |
| "grad_norm": 0.480672150850296, | |
| "learning_rate": 3.485126066291364e-05, | |
| "loss": 0.0109, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.3422247446083995, | |
| "grad_norm": 0.6047912836074829, | |
| "learning_rate": 3.470012843731476e-05, | |
| "loss": 0.0188, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.3427922814982974, | |
| "grad_norm": 0.02252427488565445, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.0008, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.34335981838819524, | |
| "grad_norm": 0.1912216693162918, | |
| "learning_rate": 3.439832771507565e-05, | |
| "loss": 0.0036, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.3439273552780931, | |
| "grad_norm": 0.18382948637008667, | |
| "learning_rate": 3.424766225756537e-05, | |
| "loss": 0.0077, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.3444948921679909, | |
| "grad_norm": 0.10974773019552231, | |
| "learning_rate": 3.4097155425921254e-05, | |
| "loss": 0.0018, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.34506242905788875, | |
| "grad_norm": 0.32770535349845886, | |
| "learning_rate": 3.394680873574546e-05, | |
| "loss": 0.0061, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.3456299659477866, | |
| "grad_norm": 0.389616996049881, | |
| "learning_rate": 3.3796623701027476e-05, | |
| "loss": 0.0097, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.34619750283768447, | |
| "grad_norm": 0.11117340624332428, | |
| "learning_rate": 3.364660183412892e-05, | |
| "loss": 0.0024, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.34676503972758227, | |
| "grad_norm": 0.11243616044521332, | |
| "learning_rate": 3.349674464576834e-05, | |
| "loss": 0.0022, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.3473325766174801, | |
| "grad_norm": 0.05497328191995621, | |
| "learning_rate": 3.334705364500596e-05, | |
| "loss": 0.0012, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.347900113507378, | |
| "grad_norm": 0.03620595484972, | |
| "learning_rate": 3.3197530339228487e-05, | |
| "loss": 0.0011, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.34846765039727584, | |
| "grad_norm": 0.029850907623767853, | |
| "learning_rate": 3.304817623413397e-05, | |
| "loss": 0.0007, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.34903518728717364, | |
| "grad_norm": 0.4595206677913666, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 0.0065, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.3496027241770715, | |
| "grad_norm": 0.3553248941898346, | |
| "learning_rate": 3.274998164025148e-05, | |
| "loss": 0.006, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.35017026106696936, | |
| "grad_norm": 0.06678071618080139, | |
| "learning_rate": 3.260114415427975e-05, | |
| "loss": 0.0012, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.3507377979568672, | |
| "grad_norm": 0.08883315324783325, | |
| "learning_rate": 3.2452481874593234e-05, | |
| "loss": 0.0022, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.351305334846765, | |
| "grad_norm": 0.04673172906041145, | |
| "learning_rate": 3.230399629821942e-05, | |
| "loss": 0.001, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.3518728717366629, | |
| "grad_norm": 0.03680291026830673, | |
| "learning_rate": 3.215568892040641e-05, | |
| "loss": 0.0009, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.35244040862656073, | |
| "grad_norm": 0.08393888175487518, | |
| "learning_rate": 3.200756123460788e-05, | |
| "loss": 0.0013, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.3530079455164586, | |
| "grad_norm": 0.3266814649105072, | |
| "learning_rate": 3.1859614732467954e-05, | |
| "loss": 0.0146, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.3535754824063564, | |
| "grad_norm": 0.38569343090057373, | |
| "learning_rate": 3.171185090380628e-05, | |
| "loss": 0.0099, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.35414301929625425, | |
| "grad_norm": 0.02289619669318199, | |
| "learning_rate": 3.156427123660297e-05, | |
| "loss": 0.0005, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.3547105561861521, | |
| "grad_norm": 0.3161522448062897, | |
| "learning_rate": 3.141687721698363e-05, | |
| "loss": 0.0036, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.35527809307604996, | |
| "grad_norm": 0.30018478631973267, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.0065, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.35584562996594776, | |
| "grad_norm": 0.41302579641342163, | |
| "learning_rate": 3.1122652055637015e-05, | |
| "loss": 0.0059, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.3564131668558456, | |
| "grad_norm": 0.5455114245414734, | |
| "learning_rate": 3.097582387675385e-05, | |
| "loss": 0.0085, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.3569807037457435, | |
| "grad_norm": 0.028173979371786118, | |
| "learning_rate": 3.082918727111304e-05, | |
| "loss": 0.0006, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.35754824063564133, | |
| "grad_norm": 0.035780180245637894, | |
| "learning_rate": 3.0682743715343564e-05, | |
| "loss": 0.0005, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.35811577752553914, | |
| "grad_norm": 0.031802088022232056, | |
| "learning_rate": 3.053649468413043e-05, | |
| "loss": 0.0008, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.358683314415437, | |
| "grad_norm": 0.5345308184623718, | |
| "learning_rate": 3.0390441650199724e-05, | |
| "loss": 0.0111, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.35925085130533485, | |
| "grad_norm": 0.2168167382478714, | |
| "learning_rate": 3.0244586084303905e-05, | |
| "loss": 0.0022, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.3598183881952327, | |
| "grad_norm": 0.36197105050086975, | |
| "learning_rate": 3.0098929455206904e-05, | |
| "loss": 0.003, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.3603859250851305, | |
| "grad_norm": 0.31182852387428284, | |
| "learning_rate": 2.9953473229669328e-05, | |
| "loss": 0.0045, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.36095346197502837, | |
| "grad_norm": 0.013852439820766449, | |
| "learning_rate": 2.9808218872433767e-05, | |
| "loss": 0.0003, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.3615209988649262, | |
| "grad_norm": 0.09768744558095932, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 0.0016, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.3620885357548241, | |
| "grad_norm": 0.8393372297286987, | |
| "learning_rate": 2.9518321611660237e-05, | |
| "loss": 0.0051, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.3626560726447219, | |
| "grad_norm": 0.03129115700721741, | |
| "learning_rate": 2.9373681627384447e-05, | |
| "loss": 0.0006, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.36322360953461974, | |
| "grad_norm": 0.20783281326293945, | |
| "learning_rate": 2.9229249349905684e-05, | |
| "loss": 0.0013, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3637911464245176, | |
| "grad_norm": 0.42036134004592896, | |
| "learning_rate": 2.9085026233655365e-05, | |
| "loss": 0.0089, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.36435868331441545, | |
| "grad_norm": 0.18196117877960205, | |
| "learning_rate": 2.894101373095867e-05, | |
| "loss": 0.002, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.36492622020431326, | |
| "grad_norm": 0.008144269697368145, | |
| "learning_rate": 2.8797213292019926e-05, | |
| "loss": 0.0002, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.3654937570942111, | |
| "grad_norm": 4.53141450881958, | |
| "learning_rate": 2.8653626364907917e-05, | |
| "loss": 0.0394, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.36606129398410897, | |
| "grad_norm": 0.011452808044850826, | |
| "learning_rate": 2.851025439554142e-05, | |
| "loss": 0.0002, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.36662883087400683, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.8367098827674578e-05, | |
| "loss": 0.0, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.36719636776390463, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.8224161102882397e-05, | |
| "loss": 0.0, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.3677639046538025, | |
| "grad_norm": 0.11076200008392334, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 0.0003, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.36833144154370034, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.7938944937838923e-05, | |
| "loss": 0.0, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.3688989784335982, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.7796669369711294e-05, | |
| "loss": 0.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.369466515323496, | |
| "grad_norm": 0.6546966433525085, | |
| "learning_rate": 2.7654617388876615e-05, | |
| "loss": 0.0211, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.37003405221339386, | |
| "grad_norm": 0.1961146891117096, | |
| "learning_rate": 2.7512790425796718e-05, | |
| "loss": 0.004, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.3706015891032917, | |
| "grad_norm": 0.7561377882957458, | |
| "learning_rate": 2.7371189908667604e-05, | |
| "loss": 0.0221, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.3711691259931896, | |
| "grad_norm": 0.2948407530784607, | |
| "learning_rate": 2.7229817263404866e-05, | |
| "loss": 0.0067, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.3717366628830874, | |
| "grad_norm": 0.0565456859767437, | |
| "learning_rate": 2.708867391362948e-05, | |
| "loss": 0.0011, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.37230419977298523, | |
| "grad_norm": 0.13881909847259521, | |
| "learning_rate": 2.694776128065345e-05, | |
| "loss": 0.0015, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.3728717366628831, | |
| "grad_norm": 0.3793433904647827, | |
| "learning_rate": 2.6807080783465376e-05, | |
| "loss": 0.0015, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.37343927355278095, | |
| "grad_norm": 0.03849627077579498, | |
| "learning_rate": 2.6666633838716314e-05, | |
| "loss": 0.0009, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.37400681044267875, | |
| "grad_norm": 0.5067241787910461, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 0.0078, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.3745743473325766, | |
| "grad_norm": 0.2140672206878662, | |
| "learning_rate": 2.638644626136587e-05, | |
| "loss": 0.0036, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.37514188422247446, | |
| "grad_norm": 0.051408469676971436, | |
| "learning_rate": 2.6246708450250256e-05, | |
| "loss": 0.0011, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.3757094211123723, | |
| "grad_norm": 0.3721614480018616, | |
| "learning_rate": 2.6107209834516854e-05, | |
| "loss": 0.0084, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.3762769580022701, | |
| "grad_norm": 0.21189674735069275, | |
| "learning_rate": 2.596795181891514e-05, | |
| "loss": 0.0024, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.376844494892168, | |
| "grad_norm": 0.46903058886528015, | |
| "learning_rate": 2.5828935805771802e-05, | |
| "loss": 0.0079, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.37741203178206584, | |
| "grad_norm": 0.03936934471130371, | |
| "learning_rate": 2.5690163194976575e-05, | |
| "loss": 0.0007, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.3779795686719637, | |
| "grad_norm": 0.034673310816287994, | |
| "learning_rate": 2.5551635383968065e-05, | |
| "loss": 0.0006, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.3785471055618615, | |
| "grad_norm": 0.01133895106613636, | |
| "learning_rate": 2.5413353767719805e-05, | |
| "loss": 0.0004, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.37911464245175935, | |
| "grad_norm": 0.48205289244651794, | |
| "learning_rate": 2.5275319738726165e-05, | |
| "loss": 0.0069, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.3796821793416572, | |
| "grad_norm": 0.24426761269569397, | |
| "learning_rate": 2.513753468698826e-05, | |
| "loss": 0.0019, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.38024971623155507, | |
| "grad_norm": 0.03907699137926102, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.0003, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.38081725312145287, | |
| "grad_norm": 0.2291565239429474, | |
| "learning_rate": 2.486271706273421e-05, | |
| "loss": 0.0106, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.3813847900113507, | |
| "grad_norm": 0.019995173439383507, | |
| "learning_rate": 2.4725687257628534e-05, | |
| "loss": 0.0004, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.3819523269012486, | |
| "grad_norm": 0.10260719805955887, | |
| "learning_rate": 2.4588911964571553e-05, | |
| "loss": 0.0011, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.38251986379114644, | |
| "grad_norm": 0.024873068556189537, | |
| "learning_rate": 2.4452392560888976e-05, | |
| "loss": 0.0004, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.38308740068104424, | |
| "grad_norm": 0.5002231001853943, | |
| "learning_rate": 2.4316130421329697e-05, | |
| "loss": 0.0108, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.3836549375709421, | |
| "grad_norm": 0.037127118557691574, | |
| "learning_rate": 2.418012691805191e-05, | |
| "loss": 0.0005, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.38422247446083996, | |
| "grad_norm": 0.2599027752876282, | |
| "learning_rate": 2.4044383420609406e-05, | |
| "loss": 0.0083, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.3847900113507378, | |
| "grad_norm": 0.018559547141194344, | |
| "learning_rate": 2.3908901295937713e-05, | |
| "loss": 0.0004, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.3853575482406356, | |
| "grad_norm": 0.13668963313102722, | |
| "learning_rate": 2.3773681908340284e-05, | |
| "loss": 0.0018, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.3859250851305335, | |
| "grad_norm": 0.21998494863510132, | |
| "learning_rate": 2.363872661947488e-05, | |
| "loss": 0.0011, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.38649262202043133, | |
| "grad_norm": 0.12420105934143066, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 0.0014, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.3870601589103292, | |
| "grad_norm": 0.2006537914276123, | |
| "learning_rate": 2.336961377126001e-05, | |
| "loss": 0.0045, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.387627695800227, | |
| "grad_norm": 0.09202957153320312, | |
| "learning_rate": 2.3235458921873925e-05, | |
| "loss": 0.0011, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.38819523269012485, | |
| "grad_norm": 0.1019575372338295, | |
| "learning_rate": 2.310157359111938e-05, | |
| "loss": 0.0009, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.3887627695800227, | |
| "grad_norm": 0.01464917603880167, | |
| "learning_rate": 2.296795912722014e-05, | |
| "loss": 0.0003, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.38933030646992056, | |
| "grad_norm": 0.024914277717471123, | |
| "learning_rate": 2.283461687567236e-05, | |
| "loss": 0.0004, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.38989784335981836, | |
| "grad_norm": 0.03489803895354271, | |
| "learning_rate": 2.2701548179231048e-05, | |
| "loss": 0.0005, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.3904653802497162, | |
| "grad_norm": 0.0056303925812244415, | |
| "learning_rate": 2.2568754377896516e-05, | |
| "loss": 0.0002, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.3910329171396141, | |
| "grad_norm": 1.0861475467681885, | |
| "learning_rate": 2.2436236808900844e-05, | |
| "loss": 0.0469, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.39160045402951194, | |
| "grad_norm": 0.7234563231468201, | |
| "learning_rate": 2.2303996806694488e-05, | |
| "loss": 0.0229, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.39216799091940974, | |
| "grad_norm": 0.01756235770881176, | |
| "learning_rate": 2.2172035702932825e-05, | |
| "loss": 0.0004, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.3927355278093076, | |
| "grad_norm": 0.007982458919286728, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 0.0002, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.39330306469920545, | |
| "grad_norm": 0.004965408705174923, | |
| "learning_rate": 2.1908955503308993e-05, | |
| "loss": 0.0001, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.3938706015891033, | |
| "grad_norm": 1.1320465803146362, | |
| "learning_rate": 2.1777839056661554e-05, | |
| "loss": 0.0237, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.3944381384790011, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.164700680686147e-05, | |
| "loss": 0.0, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.39500567536889897, | |
| "grad_norm": 0.05166594311594963, | |
| "learning_rate": 2.1516460071388062e-05, | |
| "loss": 0.0004, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.3955732122587968, | |
| "grad_norm": 0.013484718278050423, | |
| "learning_rate": 2.1386200164845526e-05, | |
| "loss": 0.0002, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.3961407491486947, | |
| "grad_norm": 26.556129455566406, | |
| "learning_rate": 2.125622839894964e-05, | |
| "loss": 0.4464, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.3967082860385925, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.1126546082514664e-05, | |
| "loss": 0.0, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.39727582292849034, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.09971545214401e-05, | |
| "loss": 0.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3978433598183882, | |
| "grad_norm": 1.003913402557373, | |
| "learning_rate": 2.086805501869749e-05, | |
| "loss": 0.0426, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.39841089670828606, | |
| "grad_norm": 0.2675124406814575, | |
| "learning_rate": 2.073924887431744e-05, | |
| "loss": 0.0221, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.39897843359818386, | |
| "grad_norm": 0.4695562720298767, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.0276, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.3995459704880817, | |
| "grad_norm": 0.8297376036643982, | |
| "learning_rate": 2.048252184598352e-05, | |
| "loss": 0.0142, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.40011350737797957, | |
| "grad_norm": 0.12075504660606384, | |
| "learning_rate": 2.0354603547267985e-05, | |
| "loss": 0.0023, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.40068104426787743, | |
| "grad_norm": 0.36436015367507935, | |
| "learning_rate": 2.0226983777365604e-05, | |
| "loss": 0.0067, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.40124858115777523, | |
| "grad_norm": 0.21536526083946228, | |
| "learning_rate": 2.0099663821406056e-05, | |
| "loss": 0.0066, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.4018161180476731, | |
| "grad_norm": 0.1440303772687912, | |
| "learning_rate": 1.9972644961499854e-05, | |
| "loss": 0.0027, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.40238365493757094, | |
| "grad_norm": 0.31195855140686035, | |
| "learning_rate": 1.9845928476725524e-05, | |
| "loss": 0.0039, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.4029511918274688, | |
| "grad_norm": 0.19405895471572876, | |
| "learning_rate": 1.9719515643116674e-05, | |
| "loss": 0.0015, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.4035187287173666, | |
| "grad_norm": 0.118756964802742, | |
| "learning_rate": 1.959340773364911e-05, | |
| "loss": 0.0014, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.40408626560726446, | |
| "grad_norm": 0.651408851146698, | |
| "learning_rate": 1.946760601822809e-05, | |
| "loss": 0.0064, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.4046538024971623, | |
| "grad_norm": 0.09903181344270706, | |
| "learning_rate": 1.9342111763675512e-05, | |
| "loss": 0.0012, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.4052213393870602, | |
| "grad_norm": 0.09159818291664124, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 0.0012, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.405788876276958, | |
| "grad_norm": 0.4659949839115143, | |
| "learning_rate": 1.9092050688969738e-05, | |
| "loss": 0.0086, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.40635641316685583, | |
| "grad_norm": 0.0276198647916317, | |
| "learning_rate": 1.8967486386928817e-05, | |
| "loss": 0.0006, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.4069239500567537, | |
| "grad_norm": 0.11385304480791092, | |
| "learning_rate": 1.8843234581955442e-05, | |
| "loss": 0.0015, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.40749148694665155, | |
| "grad_norm": 0.30209067463874817, | |
| "learning_rate": 1.8719296525263922e-05, | |
| "loss": 0.0127, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.40805902383654935, | |
| "grad_norm": 0.24259290099143982, | |
| "learning_rate": 1.859567346490913e-05, | |
| "loss": 0.0012, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.4086265607264472, | |
| "grad_norm": 0.2819889783859253, | |
| "learning_rate": 1.847236664577389e-05, | |
| "loss": 0.0059, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.40919409761634506, | |
| "grad_norm": 0.5837430953979492, | |
| "learning_rate": 1.8349377309556486e-05, | |
| "loss": 0.0041, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.4097616345062429, | |
| "grad_norm": 0.05538428574800491, | |
| "learning_rate": 1.8226706694758195e-05, | |
| "loss": 0.001, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.4103291713961407, | |
| "grad_norm": 0.08718933165073395, | |
| "learning_rate": 1.810435603667075e-05, | |
| "loss": 0.0009, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.4108967082860386, | |
| "grad_norm": 0.09461364895105362, | |
| "learning_rate": 1.7982326567363888e-05, | |
| "loss": 0.0017, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.41146424517593644, | |
| "grad_norm": 0.43470796942710876, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 0.0176, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.4120317820658343, | |
| "grad_norm": 0.04546342045068741, | |
| "learning_rate": 1.773923610718686e-05, | |
| "loss": 0.0009, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.4125993189557321, | |
| "grad_norm": 0.014890948310494423, | |
| "learning_rate": 1.7618177564234905e-05, | |
| "loss": 0.0003, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.41316685584562995, | |
| "grad_norm": 0.030182119458913803, | |
| "learning_rate": 1.7497445105875377e-05, | |
| "loss": 0.0005, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.4137343927355278, | |
| "grad_norm": 0.05278665944933891, | |
| "learning_rate": 1.73770399478828e-05, | |
| "loss": 0.0008, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.41430192962542567, | |
| "grad_norm": 0.4548901915550232, | |
| "learning_rate": 1.725696330273575e-05, | |
| "loss": 0.0118, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.41486946651532347, | |
| "grad_norm": 0.33028581738471985, | |
| "learning_rate": 1.7137216379604727e-05, | |
| "loss": 0.0071, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.41543700340522133, | |
| "grad_norm": 0.0861010029911995, | |
| "learning_rate": 1.7017800384339928e-05, | |
| "loss": 0.001, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.4160045402951192, | |
| "grad_norm": 0.438575804233551, | |
| "learning_rate": 1.6898716519459074e-05, | |
| "loss": 0.0071, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.41657207718501704, | |
| "grad_norm": 0.05357427895069122, | |
| "learning_rate": 1.6779965984135377e-05, | |
| "loss": 0.0006, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.41713961407491484, | |
| "grad_norm": 0.022530531510710716, | |
| "learning_rate": 1.6661549974185424e-05, | |
| "loss": 0.0003, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.4177071509648127, | |
| "grad_norm": 0.4733809232711792, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 0.0127, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.41827468785471056, | |
| "grad_norm": 0.12539038062095642, | |
| "learning_rate": 1.6425726296817633e-05, | |
| "loss": 0.0014, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.4188422247446084, | |
| "grad_norm": 0.4548875689506531, | |
| "learning_rate": 1.6308321004141607e-05, | |
| "loss": 0.0076, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.4194097616345062, | |
| "grad_norm": 0.0097389817237854, | |
| "learning_rate": 1.619125498629904e-05, | |
| "loss": 0.0002, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.4199772985244041, | |
| "grad_norm": 0.019004186615347862, | |
| "learning_rate": 1.60745294221434e-05, | |
| "loss": 0.0004, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.42054483541430193, | |
| "grad_norm": 0.03138939291238785, | |
| "learning_rate": 1.595814548709983e-05, | |
| "loss": 0.0004, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.4211123723041998, | |
| "grad_norm": 0.5367324948310852, | |
| "learning_rate": 1.5842104353153287e-05, | |
| "loss": 0.0249, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.4216799091940976, | |
| "grad_norm": 1.0344882011413574, | |
| "learning_rate": 1.5726407188836673e-05, | |
| "loss": 0.0335, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.42224744608399545, | |
| "grad_norm": 0.0143516156822443, | |
| "learning_rate": 1.5611055159219152e-05, | |
| "loss": 0.0003, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.4228149829738933, | |
| "grad_norm": 31.227991104125977, | |
| "learning_rate": 1.549604942589441e-05, | |
| "loss": 1.1122, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.42338251986379116, | |
| "grad_norm": 0.10639174282550812, | |
| "learning_rate": 1.5381391146968866e-05, | |
| "loss": 0.0009, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.42395005675368896, | |
| "grad_norm": 0.05225389450788498, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 0.0003, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.4245175936435868, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.5153121567235335e-05, | |
| "loss": 0.0, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.4250851305334847, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.5039512565099467e-05, | |
| "loss": 0.0, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.42565266742338254, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.4926255614683932e-05, | |
| "loss": 0.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.42622020431328034, | |
| "grad_norm": 0.5881960988044739, | |
| "learning_rate": 1.481335185648498e-05, | |
| "loss": 0.0209, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.4267877412031782, | |
| "grad_norm": 0.46018141508102417, | |
| "learning_rate": 1.4700802427442179e-05, | |
| "loss": 0.009, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.42735527809307605, | |
| "grad_norm": 0.40768754482269287, | |
| "learning_rate": 1.458860846092705e-05, | |
| "loss": 0.0032, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.4279228149829739, | |
| "grad_norm": 0.2659337520599365, | |
| "learning_rate": 1.4476771086731567e-05, | |
| "loss": 0.0061, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.4284903518728717, | |
| "grad_norm": 0.10703348368406296, | |
| "learning_rate": 1.4365291431056871e-05, | |
| "loss": 0.0017, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.42905788876276957, | |
| "grad_norm": 0.3611052930355072, | |
| "learning_rate": 1.4254170616501827e-05, | |
| "loss": 0.0034, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.4296254256526674, | |
| "grad_norm": 0.05121847242116928, | |
| "learning_rate": 1.414340976205183e-05, | |
| "loss": 0.001, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.4301929625425653, | |
| "grad_norm": 0.31806862354278564, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 0.0059, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.4307604994324631, | |
| "grad_norm": 0.05238351970911026, | |
| "learning_rate": 1.3922972391273226e-05, | |
| "loss": 0.0011, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.43132803632236094, | |
| "grad_norm": 0.17556647956371307, | |
| "learning_rate": 1.3813298094746491e-05, | |
| "loss": 0.0029, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.4318955732122588, | |
| "grad_norm": 0.1977948248386383, | |
| "learning_rate": 1.3703988197906209e-05, | |
| "loss": 0.0043, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.43246311010215666, | |
| "grad_norm": 0.058601368218660355, | |
| "learning_rate": 1.3595043801501794e-05, | |
| "loss": 0.0013, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.43303064699205446, | |
| "grad_norm": 0.2709505558013916, | |
| "learning_rate": 1.3486466002602133e-05, | |
| "loss": 0.0022, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.4335981838819523, | |
| "grad_norm": 0.04408566281199455, | |
| "learning_rate": 1.3378255894584463e-05, | |
| "loss": 0.0008, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.43416572077185017, | |
| "grad_norm": 0.034973569214344025, | |
| "learning_rate": 1.327041456712334e-05, | |
| "loss": 0.0006, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.43473325766174803, | |
| "grad_norm": 0.32786574959754944, | |
| "learning_rate": 1.3162943106179749e-05, | |
| "loss": 0.018, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.43530079455164583, | |
| "grad_norm": 0.05485441908240318, | |
| "learning_rate": 1.3055842593990131e-05, | |
| "loss": 0.0005, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.4358683314415437, | |
| "grad_norm": 0.07898583263158798, | |
| "learning_rate": 1.2949114109055415e-05, | |
| "loss": 0.0013, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.43643586833144155, | |
| "grad_norm": 0.03237922489643097, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 0.0007, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.4370034052213394, | |
| "grad_norm": 0.05610362067818642, | |
| "learning_rate": 1.2736777516212266e-05, | |
| "loss": 0.0008, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.4375709421112372, | |
| "grad_norm": 0.24216708540916443, | |
| "learning_rate": 1.2631171546530968e-05, | |
| "loss": 0.0037, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.43813847900113506, | |
| "grad_norm": 0.07961627095937729, | |
| "learning_rate": 1.2525941880537307e-05, | |
| "loss": 0.0013, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.4387060158910329, | |
| "grad_norm": 0.11050142347812653, | |
| "learning_rate": 1.2421089577892869e-05, | |
| "loss": 0.0012, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.4392735527809308, | |
| "grad_norm": 0.04272003099322319, | |
| "learning_rate": 1.2316615694459189e-05, | |
| "loss": 0.0006, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.4398410896708286, | |
| "grad_norm": 0.02150142751634121, | |
| "learning_rate": 1.2212521282287092e-05, | |
| "loss": 0.0004, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.44040862656072643, | |
| "grad_norm": 0.08354512602090836, | |
| "learning_rate": 1.2108807389606158e-05, | |
| "loss": 0.001, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.4409761634506243, | |
| "grad_norm": 0.1265098601579666, | |
| "learning_rate": 1.2005475060814159e-05, | |
| "loss": 0.0018, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.44154370034052215, | |
| "grad_norm": 0.18211375176906586, | |
| "learning_rate": 1.1902525336466464e-05, | |
| "loss": 0.002, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.44211123723041995, | |
| "grad_norm": 0.017222406342625618, | |
| "learning_rate": 1.1799959253265668e-05, | |
| "loss": 0.0004, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.4426787741203178, | |
| "grad_norm": 0.030056394636631012, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 0.0007, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.44324631101021567, | |
| "grad_norm": 0.2576983571052551, | |
| "learning_rate": 1.1595982137788403e-05, | |
| "loss": 0.002, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.4438138479001135, | |
| "grad_norm": 0.20658354461193085, | |
| "learning_rate": 1.1494573159559213e-05, | |
| "loss": 0.0021, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.4443813847900113, | |
| "grad_norm": 0.324457049369812, | |
| "learning_rate": 1.1393551930550828e-05, | |
| "loss": 0.0023, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.4449489216799092, | |
| "grad_norm": 0.2382335364818573, | |
| "learning_rate": 1.1292919468045877e-05, | |
| "loss": 0.0017, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.44551645856980704, | |
| "grad_norm": 0.3122727572917938, | |
| "learning_rate": 1.1192676785412154e-05, | |
| "loss": 0.0041, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.4460839954597049, | |
| "grad_norm": 0.06047174334526062, | |
| "learning_rate": 1.1092824892092373e-05, | |
| "loss": 0.0011, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.4466515323496027, | |
| "grad_norm": 0.12172012776136398, | |
| "learning_rate": 1.099336479359398e-05, | |
| "loss": 0.002, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.44721906923950056, | |
| "grad_norm": 0.05569892004132271, | |
| "learning_rate": 1.0894297491479045e-05, | |
| "loss": 0.0008, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.4477866061293984, | |
| "grad_norm": 0.02484039030969143, | |
| "learning_rate": 1.0795623983354215e-05, | |
| "loss": 0.0004, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.44835414301929627, | |
| "grad_norm": 0.0289757139980793, | |
| "learning_rate": 1.0697345262860636e-05, | |
| "loss": 0.0005, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.44892167990919407, | |
| "grad_norm": 0.025877099484205246, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 0.0004, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.44948921679909193, | |
| "grad_norm": 0.011032159440219402, | |
| "learning_rate": 1.0501976139444191e-05, | |
| "loss": 0.0002, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.4500567536889898, | |
| "grad_norm": 0.008858336135745049, | |
| "learning_rate": 1.0404887703886251e-05, | |
| "loss": 0.0001, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.45062429057888764, | |
| "grad_norm": 0.022364582866430283, | |
| "learning_rate": 1.0308197990669538e-05, | |
| "loss": 0.0003, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.45119182746878544, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.021190797345839e-05, | |
| "loss": 0.0, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.4517593643586833, | |
| "grad_norm": 0.1561431884765625, | |
| "learning_rate": 1.0116018621892237e-05, | |
| "loss": 0.0011, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.45232690124858116, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.0020530901575754e-05, | |
| "loss": 0.0, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.452894438138479, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.92544577406923e-06, | |
| "loss": 0.0, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.4534619750283768, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.830764196878872e-06, | |
| "loss": 0.0, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.4540295119182747, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.73648712344707e-06, | |
| "loss": 0.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4540295119182747, | |
| "eval_loss": NaN, | |
| "eval_runtime": 105.9039, | |
| "eval_samples_per_second": 28.025, | |
| "eval_steps_per_second": 7.006, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.45459704880817253, | |
| "grad_norm": 0.623717725276947, | |
| "learning_rate": 9.642615503142926e-06, | |
| "loss": 0.0176, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.4551645856980704, | |
| "grad_norm": 0.24405649304389954, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.0043, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.4557321225879682, | |
| "grad_norm": 0.04640405625104904, | |
| "learning_rate": 9.456092398969902e-06, | |
| "loss": 0.0013, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.45629965947786605, | |
| "grad_norm": 0.018343493342399597, | |
| "learning_rate": 9.363442793386606e-06, | |
| "loss": 0.0005, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.4568671963677639, | |
| "grad_norm": 0.1869243085384369, | |
| "learning_rate": 9.271202397483215e-06, | |
| "loss": 0.0109, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.45743473325766176, | |
| "grad_norm": 1.0756665468215942, | |
| "learning_rate": 9.179372140119525e-06, | |
| "loss": 0.0142, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.45800227014755956, | |
| "grad_norm": 0.23151446878910065, | |
| "learning_rate": 9.087952946025175e-06, | |
| "loss": 0.0147, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.4585698070374574, | |
| "grad_norm": 0.01738697662949562, | |
| "learning_rate": 8.996945735790447e-06, | |
| "loss": 0.0004, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.4591373439273553, | |
| "grad_norm": 0.45961514115333557, | |
| "learning_rate": 8.906351425856952e-06, | |
| "loss": 0.0045, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.45970488081725314, | |
| "grad_norm": 0.3801596760749817, | |
| "learning_rate": 8.816170928508365e-06, | |
| "loss": 0.009, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.46027241770715094, | |
| "grad_norm": 0.03849168121814728, | |
| "learning_rate": 8.7264051518613e-06, | |
| "loss": 0.0008, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.4608399545970488, | |
| "grad_norm": 0.2326851785182953, | |
| "learning_rate": 8.637054999856148e-06, | |
| "loss": 0.0093, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.46140749148694665, | |
| "grad_norm": 0.048841919749975204, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 0.0007, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.4619750283768445, | |
| "grad_norm": 0.02097911760210991, | |
| "learning_rate": 8.459605164597267e-06, | |
| "loss": 0.0005, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.4625425652667423, | |
| "grad_norm": 0.3414818346500397, | |
| "learning_rate": 8.371507268261437e-06, | |
| "loss": 0.0023, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.46311010215664017, | |
| "grad_norm": 0.08118417859077454, | |
| "learning_rate": 8.283828570385238e-06, | |
| "loss": 0.0008, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.463677639046538, | |
| "grad_norm": 0.41794729232788086, | |
| "learning_rate": 8.196569953892202e-06, | |
| "loss": 0.0112, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.4642451759364359, | |
| "grad_norm": 0.3035317063331604, | |
| "learning_rate": 8.109732297475635e-06, | |
| "loss": 0.0123, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.4648127128263337, | |
| "grad_norm": 0.03364351764321327, | |
| "learning_rate": 8.023316475589754e-06, | |
| "loss": 0.0007, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.46538024971623154, | |
| "grad_norm": 0.48411476612091064, | |
| "learning_rate": 7.937323358440935e-06, | |
| "loss": 0.0161, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4659477866061294, | |
| "grad_norm": 0.020044121891260147, | |
| "learning_rate": 7.851753811978924e-06, | |
| "loss": 0.0005, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.46651532349602726, | |
| "grad_norm": 0.011223547160625458, | |
| "learning_rate": 7.766608697888095e-06, | |
| "loss": 0.0002, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.46708286038592506, | |
| "grad_norm": 0.16546539962291718, | |
| "learning_rate": 7.681888873578786e-06, | |
| "loss": 0.0025, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.4676503972758229, | |
| "grad_norm": 0.009762106463313103, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 0.0003, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.4682179341657208, | |
| "grad_norm": 0.023298079147934914, | |
| "learning_rate": 7.513728502524286e-06, | |
| "loss": 0.0004, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.46878547105561863, | |
| "grad_norm": 0.18507546186447144, | |
| "learning_rate": 7.430289649152156e-06, | |
| "loss": 0.0015, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.46935300794551643, | |
| "grad_norm": 0.37739697098731995, | |
| "learning_rate": 7.347279472290647e-06, | |
| "loss": 0.0096, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.4699205448354143, | |
| "grad_norm": 0.10507706552743912, | |
| "learning_rate": 7.264698807851328e-06, | |
| "loss": 0.0015, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.47048808172531215, | |
| "grad_norm": 0.049794506281614304, | |
| "learning_rate": 7.182548487420554e-06, | |
| "loss": 0.0009, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.47105561861521, | |
| "grad_norm": 0.061873581260442734, | |
| "learning_rate": 7.100829338251147e-06, | |
| "loss": 0.001, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4716231555051078, | |
| "grad_norm": 0.02593647502362728, | |
| "learning_rate": 7.019542183254046e-06, | |
| "loss": 0.0004, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.47219069239500566, | |
| "grad_norm": 0.6185386776924133, | |
| "learning_rate": 6.9386878409899715e-06, | |
| "loss": 0.0073, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.4727582292849035, | |
| "grad_norm": 0.666622519493103, | |
| "learning_rate": 6.858267125661272e-06, | |
| "loss": 0.0093, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.4733257661748014, | |
| "grad_norm": 0.05342670530080795, | |
| "learning_rate": 6.778280847103669e-06, | |
| "loss": 0.0008, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.4738933030646992, | |
| "grad_norm": 1.0183546543121338, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 0.0078, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.47446083995459704, | |
| "grad_norm": 0.1871764212846756, | |
| "learning_rate": 6.619614817762537e-06, | |
| "loss": 0.0014, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.4750283768444949, | |
| "grad_norm": 0.19295842945575714, | |
| "learning_rate": 6.540936664744196e-06, | |
| "loss": 0.002, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.47559591373439275, | |
| "grad_norm": 0.7019402384757996, | |
| "learning_rate": 6.462696144011149e-06, | |
| "loss": 0.0207, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.47616345062429055, | |
| "grad_norm": 0.02959679253399372, | |
| "learning_rate": 6.384894043444567e-06, | |
| "loss": 0.0004, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.4767309875141884, | |
| "grad_norm": 0.1982329785823822, | |
| "learning_rate": 6.3075311465107535e-06, | |
| "loss": 0.0008, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.47729852440408627, | |
| "grad_norm": 0.03872201591730118, | |
| "learning_rate": 6.230608232253227e-06, | |
| "loss": 0.0004, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.4778660612939841, | |
| "grad_norm": 0.08221829682588577, | |
| "learning_rate": 6.154126075284855e-06, | |
| "loss": 0.0006, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.4784335981838819, | |
| "grad_norm": 0.007916197180747986, | |
| "learning_rate": 6.078085445780129e-06, | |
| "loss": 0.0002, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.4790011350737798, | |
| "grad_norm": 0.008437985554337502, | |
| "learning_rate": 6.002487109467347e-06, | |
| "loss": 0.0002, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.47956867196367764, | |
| "grad_norm": 0.005791019182652235, | |
| "learning_rate": 5.927331827620903e-06, | |
| "loss": 0.0001, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.4801362088535755, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 0.0, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.4807037457434733, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.778353450109286e-06, | |
| "loss": 0.0, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.48127128263337116, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.704531854654721e-06, | |
| "loss": 0.0, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.481838819523269, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.631156314072605e-06, | |
| "loss": 0.0, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.48240635641316687, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.558227567253832e-06, | |
| "loss": 0.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.48297389330306467, | |
| "grad_norm": 0.3478143811225891, | |
| "learning_rate": 5.485746348590048e-06, | |
| "loss": 0.0098, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.48354143019296253, | |
| "grad_norm": 0.38309186697006226, | |
| "learning_rate": 5.413713387966329e-06, | |
| "loss": 0.0101, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.4841089670828604, | |
| "grad_norm": 0.28978919982910156, | |
| "learning_rate": 5.34212941075381e-06, | |
| "loss": 0.0048, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.48467650397275824, | |
| "grad_norm": 0.016624854877591133, | |
| "learning_rate": 5.270995137802315e-06, | |
| "loss": 0.0004, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.48524404086265605, | |
| "grad_norm": 0.03391743823885918, | |
| "learning_rate": 5.200311285433213e-06, | |
| "loss": 0.0006, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.4858115777525539, | |
| "grad_norm": 0.1863984912633896, | |
| "learning_rate": 5.13007856543209e-06, | |
| "loss": 0.0104, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.48637911464245176, | |
| "grad_norm": 0.01976456306874752, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 0.0004, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.4869466515323496, | |
| "grad_norm": 0.21336333453655243, | |
| "learning_rate": 4.99096934695461e-06, | |
| "loss": 0.0115, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.4875141884222474, | |
| "grad_norm": 0.0573849081993103, | |
| "learning_rate": 4.922094249306558e-06, | |
| "loss": 0.0007, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.4880817253121453, | |
| "grad_norm": 0.0516238808631897, | |
| "learning_rate": 4.853673085668947e-06, | |
| "loss": 0.0008, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.48864926220204313, | |
| "grad_norm": 0.548546314239502, | |
| "learning_rate": 4.78570654504214e-06, | |
| "loss": 0.0216, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.489216799091941, | |
| "grad_norm": 0.03412328287959099, | |
| "learning_rate": 4.7181953118484556e-06, | |
| "loss": 0.0006, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.4897843359818388, | |
| "grad_norm": 0.03500758484005928, | |
| "learning_rate": 4.651140065925269e-06, | |
| "loss": 0.0008, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.49035187287173665, | |
| "grad_norm": 0.019669918343424797, | |
| "learning_rate": 4.58454148251814e-06, | |
| "loss": 0.0004, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.4909194097616345, | |
| "grad_norm": 0.3114485442638397, | |
| "learning_rate": 4.5184002322740785e-06, | |
| "loss": 0.0028, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.49148694665153236, | |
| "grad_norm": 0.5113534331321716, | |
| "learning_rate": 4.452716981234744e-06, | |
| "loss": 0.0185, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.49205448354143017, | |
| "grad_norm": 0.07297579944133759, | |
| "learning_rate": 4.387492390829734e-06, | |
| "loss": 0.0009, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.492622020431328, | |
| "grad_norm": 0.4376738667488098, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 0.0091, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.4931895573212259, | |
| "grad_norm": 0.011865437030792236, | |
| "learning_rate": 4.258421814540992e-06, | |
| "loss": 0.0002, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.49375709421112374, | |
| "grad_norm": 0.16000713407993317, | |
| "learning_rate": 4.19457712839652e-06, | |
| "loss": 0.0013, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.49432463110102154, | |
| "grad_norm": 0.15423361957073212, | |
| "learning_rate": 4.131193702351827e-06, | |
| "loss": 0.0024, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.4948921679909194, | |
| "grad_norm": 0.022781820967793465, | |
| "learning_rate": 4.068272174677335e-06, | |
| "loss": 0.0004, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.49545970488081725, | |
| "grad_norm": 0.025359636172652245, | |
| "learning_rate": 4.005813178992091e-06, | |
| "loss": 0.0005, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.4960272417707151, | |
| "grad_norm": 0.029715919867157936, | |
| "learning_rate": 3.9438173442575e-06, | |
| "loss": 0.0006, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.4965947786606129, | |
| "grad_norm": 0.019626963883638382, | |
| "learning_rate": 3.8822852947709375e-06, | |
| "loss": 0.0003, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.49716231555051077, | |
| "grad_norm": 0.05726097524166107, | |
| "learning_rate": 3.821217650159453e-06, | |
| "loss": 0.0008, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.4977298524404086, | |
| "grad_norm": 0.2366546243429184, | |
| "learning_rate": 3.760615025373543e-06, | |
| "loss": 0.0064, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.4982973893303065, | |
| "grad_norm": 0.10293308645486832, | |
| "learning_rate": 3.700478030680987e-06, | |
| "loss": 0.0014, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.4988649262202043, | |
| "grad_norm": 0.01754389889538288, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 0.0003, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.49943246311010214, | |
| "grad_norm": 0.8322834968566895, | |
| "learning_rate": 3.581603349196372e-06, | |
| "loss": 0.0213, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.2963551878929138, | |
| "learning_rate": 3.522866859471047e-06, | |
| "loss": 0.0038, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.5005675368898979, | |
| "grad_norm": 0.47917166352272034, | |
| "learning_rate": 3.4645983939604496e-06, | |
| "loss": 0.008, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.5011350737797957, | |
| "grad_norm": 0.04047662764787674, | |
| "learning_rate": 3.406798539427386e-06, | |
| "loss": 0.0004, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.5017026106696936, | |
| "grad_norm": 0.03451113775372505, | |
| "learning_rate": 3.349467877915746e-06, | |
| "loss": 0.0002, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.5022701475595914, | |
| "grad_norm": 0.04126652702689171, | |
| "learning_rate": 3.2926069867446675e-06, | |
| "loss": 0.0005, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.5028376844494892, | |
| "grad_norm": 0.5223488211631775, | |
| "learning_rate": 3.2362164385026706e-06, | |
| "loss": 0.0093, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.503405221339387, | |
| "grad_norm": 0.06192615255713463, | |
| "learning_rate": 3.180296801041971e-06, | |
| "loss": 0.0007, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.5039727582292849, | |
| "grad_norm": 0.022352147847414017, | |
| "learning_rate": 3.1248486374726883e-06, | |
| "loss": 0.0003, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.5045402951191827, | |
| "grad_norm": 0.3611301779747009, | |
| "learning_rate": 3.069872506157212e-06, | |
| "loss": 0.0028, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.5051078320090806, | |
| "grad_norm": 0.4576888978481293, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 0.0058, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.5056753688989785, | |
| "grad_norm": 0.22780518233776093, | |
| "learning_rate": 2.961338549964893e-06, | |
| "loss": 0.0019, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.5062429057888763, | |
| "grad_norm": 0.05007459968328476, | |
| "learning_rate": 2.9077818180237693e-06, | |
| "loss": 0.0005, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.5068104426787742, | |
| "grad_norm": 0.012189110741019249, | |
| "learning_rate": 2.8546993041969173e-06, | |
| "loss": 0.0002, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.5073779795686719, | |
| "grad_norm": 0.009090066887438297, | |
| "learning_rate": 2.802091543024671e-06, | |
| "loss": 0.0002, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.5079455164585698, | |
| "grad_norm": 0.6339737176895142, | |
| "learning_rate": 2.7499590642665774e-06, | |
| "loss": 0.0025, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.5085130533484676, | |
| "grad_norm": 0.06957720965147018, | |
| "learning_rate": 2.6983023928961404e-06, | |
| "loss": 0.0005, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.5090805902383655, | |
| "grad_norm": 7.535338878631592, | |
| "learning_rate": 2.647122049095463e-06, | |
| "loss": 0.0468, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.5096481271282634, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.596418548250029e-06, | |
| "loss": 0.0, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.5102156640181612, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.546192400943537e-06, | |
| "loss": 0.0, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.5107832009080591, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.496444112952734e-06, | |
| "loss": 0.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.5113507377979569, | |
| "grad_norm": 0.33791133761405945, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.01, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.5119182746878547, | |
| "grad_norm": 0.10187462717294693, | |
| "learning_rate": 2.3983831139599287e-06, | |
| "loss": 0.002, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.5124858115777525, | |
| "grad_norm": 0.02221851982176304, | |
| "learning_rate": 2.3500713904311024e-06, | |
| "loss": 0.0007, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.5130533484676504, | |
| "grad_norm": 0.014832521788775921, | |
| "learning_rate": 2.3022395011543686e-06, | |
| "loss": 0.0005, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.5136208853575482, | |
| "grad_norm": 0.26342910528182983, | |
| "learning_rate": 2.2548879277963064e-06, | |
| "loss": 0.0045, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.5141884222474461, | |
| "grad_norm": 0.08771803975105286, | |
| "learning_rate": 2.208017147186736e-06, | |
| "loss": 0.0009, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.514755959137344, | |
| "grad_norm": 0.2147466242313385, | |
| "learning_rate": 2.161627631313923e-06, | |
| "loss": 0.001, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.5153234960272418, | |
| "grad_norm": 0.09128167480230331, | |
| "learning_rate": 2.1157198473197414e-06, | |
| "loss": 0.0013, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.5158910329171397, | |
| "grad_norm": 0.01900799199938774, | |
| "learning_rate": 2.070294257495081e-06, | |
| "loss": 0.0004, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.5164585698070374, | |
| "grad_norm": 0.040807388722896576, | |
| "learning_rate": 2.0253513192751373e-06, | |
| "loss": 0.0006, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.5170261066969353, | |
| "grad_norm": 0.19579406082630157, | |
| "learning_rate": 1.9808914852347813e-06, | |
| "loss": 0.0024, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.5175936435868331, | |
| "grad_norm": 0.22570084035396576, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 0.011, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.518161180476731, | |
| "grad_norm": 0.16357550024986267, | |
| "learning_rate": 1.8934229156636452e-06, | |
| "loss": 0.0018, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.5187287173666288, | |
| "grad_norm": 0.024271734058856964, | |
| "learning_rate": 1.8504150609403858e-06, | |
| "loss": 0.0004, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.5192962542565267, | |
| "grad_norm": 0.4943500757217407, | |
| "learning_rate": 1.807892072002898e-06, | |
| "loss": 0.0099, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.5198637911464246, | |
| "grad_norm": 0.22430936992168427, | |
| "learning_rate": 1.7658543770572189e-06, | |
| "loss": 0.0062, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.5204313280363224, | |
| "grad_norm": 0.02454625442624092, | |
| "learning_rate": 1.724302399422456e-06, | |
| "loss": 0.0006, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.5209988649262202, | |
| "grad_norm": 0.11323986947536469, | |
| "learning_rate": 1.6832365575265741e-06, | |
| "loss": 0.0014, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.521566401816118, | |
| "grad_norm": 0.03107013925909996, | |
| "learning_rate": 1.6426572649021476e-06, | |
| "loss": 0.0006, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.5221339387060159, | |
| "grad_norm": 0.04345984384417534, | |
| "learning_rate": 1.6025649301821876e-06, | |
| "loss": 0.0004, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.5227014755959137, | |
| "grad_norm": 0.4001345932483673, | |
| "learning_rate": 1.5629599570960718e-06, | |
| "loss": 0.0115, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.5232690124858116, | |
| "grad_norm": 0.09491916000843048, | |
| "learning_rate": 1.523842744465437e-06, | |
| "loss": 0.0006, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.5238365493757094, | |
| "grad_norm": 0.17167732119560242, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 0.0021, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.5244040862656073, | |
| "grad_norm": 0.09016118198633194, | |
| "learning_rate": 1.4470731712944884e-06, | |
| "loss": 0.0004, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.5249716231555052, | |
| "grad_norm": 0.10207764804363251, | |
| "learning_rate": 1.4094215838229176e-06, | |
| "loss": 0.001, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.5255391600454029, | |
| "grad_norm": 0.009659104980528355, | |
| "learning_rate": 1.372259302936546e-06, | |
| "loss": 0.0002, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.5261066969353008, | |
| "grad_norm": 0.022223835811018944, | |
| "learning_rate": 1.3355867028591208e-06, | |
| "loss": 0.0004, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.5266742338251986, | |
| "grad_norm": 0.3852534890174866, | |
| "learning_rate": 1.2994041528833266e-06, | |
| "loss": 0.0176, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.5272417707150965, | |
| "grad_norm": 0.032568030059337616, | |
| "learning_rate": 1.2637120173670358e-06, | |
| "loss": 0.0005, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.5278093076049943, | |
| "grad_norm": 0.12185148894786835, | |
| "learning_rate": 1.2285106557296477e-06, | |
| "loss": 0.0014, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5283768444948922, | |
| "grad_norm": 0.610008716583252, | |
| "learning_rate": 1.1938004224484988e-06, | |
| "loss": 0.0051, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.52894438138479, | |
| "grad_norm": 0.06961622089147568, | |
| "learning_rate": 1.1595816670552428e-06, | |
| "loss": 0.0011, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.5295119182746879, | |
| "grad_norm": 0.055191271007061005, | |
| "learning_rate": 1.1258547341323699e-06, | |
| "loss": 0.0005, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.5300794551645857, | |
| "grad_norm": 0.43228384852409363, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.0173, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.5306469920544835, | |
| "grad_norm": 0.4304471015930176, | |
| "learning_rate": 1.0598776892610685e-06, | |
| "loss": 0.0156, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.5312145289443814, | |
| "grad_norm": 0.02435118891298771, | |
| "learning_rate": 1.02762824170074e-06, | |
| "loss": 0.0003, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.5317820658342792, | |
| "grad_norm": 0.18826240301132202, | |
| "learning_rate": 9.958719453803278e-07, | |
| "loss": 0.002, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.5323496027241771, | |
| "grad_norm": 0.5460970401763916, | |
| "learning_rate": 9.646091200853802e-07, | |
| "loss": 0.0294, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.5329171396140749, | |
| "grad_norm": 0.013325286097824574, | |
| "learning_rate": 9.338400806321978e-07, | |
| "loss": 0.0002, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.5334846765039728, | |
| "grad_norm": 0.010715479031205177, | |
| "learning_rate": 9.035651368646648e-07, | |
| "loss": 0.0002, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5340522133938707, | |
| "grad_norm": 0.05249933898448944, | |
| "learning_rate": 8.737845936511335e-07, | |
| "loss": 0.0007, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.5346197502837684, | |
| "grad_norm": 0.6176870465278625, | |
| "learning_rate": 8.444987508813451e-07, | |
| "loss": 0.0114, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.5351872871736663, | |
| "grad_norm": 0.05264892801642418, | |
| "learning_rate": 8.157079034633974e-07, | |
| "loss": 0.0003, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.5357548240635641, | |
| "grad_norm": 0.022312408313155174, | |
| "learning_rate": 7.874123413208145e-07, | |
| "loss": 0.0002, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.536322360953462, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 0.0, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.5368898978433598, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.323082076153509e-07, | |
| "loss": 0.0, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.5374574347332577, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.055001909504755e-07, | |
| "loss": 0.0, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.5380249716231555, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.791885693514133e-07, | |
| "loss": 0.0, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.5385925085130534, | |
| "grad_norm": 29.0640926361084, | |
| "learning_rate": 6.533736077758868e-07, | |
| "loss": 0.3484, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.5391600454029511, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.280555661802856e-07, | |
| "loss": 0.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.539727582292849, | |
| "grad_norm": 0.7819038033485413, | |
| "learning_rate": 6.032346995169968e-07, | |
| "loss": 0.0194, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.5402951191827469, | |
| "grad_norm": 0.15347006916999817, | |
| "learning_rate": 5.78911257731879e-07, | |
| "loss": 0.0101, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.5408626560726447, | |
| "grad_norm": 0.2514842748641968, | |
| "learning_rate": 5.550854857617193e-07, | |
| "loss": 0.0147, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.5414301929625426, | |
| "grad_norm": 0.018654726445674896, | |
| "learning_rate": 5.317576235317756e-07, | |
| "loss": 0.0004, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.5419977298524404, | |
| "grad_norm": 0.017752651125192642, | |
| "learning_rate": 5.089279059533658e-07, | |
| "loss": 0.0005, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.5425652667423383, | |
| "grad_norm": 0.010392699390649796, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 0.0003, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.5431328036322361, | |
| "grad_norm": 0.183840811252594, | |
| "learning_rate": 4.647638193125137e-07, | |
| "loss": 0.009, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.5437003405221339, | |
| "grad_norm": 0.22836080193519592, | |
| "learning_rate": 4.434298949819449e-07, | |
| "loss": 0.0039, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.5442678774120318, | |
| "grad_norm": 0.19406422972679138, | |
| "learning_rate": 4.2259500476214407e-07, | |
| "loss": 0.0091, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.5448354143019296, | |
| "grad_norm": 0.015072612091898918, | |
| "learning_rate": 4.02259358460233e-07, | |
| "loss": 0.0004, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5454029511918275, | |
| "grad_norm": 0.09216684103012085, | |
| "learning_rate": 3.824231608559492e-07, | |
| "loss": 0.0006, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.5459704880817253, | |
| "grad_norm": 0.35321223735809326, | |
| "learning_rate": 3.630866116995757e-07, | |
| "loss": 0.006, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.5465380249716232, | |
| "grad_norm": 0.1484595090150833, | |
| "learning_rate": 3.4424990570994797e-07, | |
| "loss": 0.0039, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.547105561861521, | |
| "grad_norm": 0.25822874903678894, | |
| "learning_rate": 3.2591323257248893e-07, | |
| "loss": 0.0036, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.5476730987514189, | |
| "grad_norm": 0.11979275941848755, | |
| "learning_rate": 3.080767769372939e-07, | |
| "loss": 0.0012, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.5482406356413166, | |
| "grad_norm": 0.013929195702075958, | |
| "learning_rate": 2.907407184172706e-07, | |
| "loss": 0.0003, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.5488081725312145, | |
| "grad_norm": 0.017956310883164406, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 0.0004, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.5493757094211124, | |
| "grad_norm": 0.010877071879804134, | |
| "learning_rate": 2.5757048597765396e-07, | |
| "loss": 0.0003, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.5499432463110102, | |
| "grad_norm": 0.016699308529496193, | |
| "learning_rate": 2.4173664608193593e-07, | |
| "loss": 0.0004, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.5505107832009081, | |
| "grad_norm": 0.014117494225502014, | |
| "learning_rate": 2.2640387134577058e-07, | |
| "loss": 0.0003, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.5510783200908059, | |
| "grad_norm": 0.3465084135532379, | |
| "learning_rate": 2.1157231617002783e-07, | |
| "loss": 0.0134, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.5516458569807038, | |
| "grad_norm": 0.098577119410038, | |
| "learning_rate": 1.9724212990830938e-07, | |
| "loss": 0.0015, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.5522133938706016, | |
| "grad_norm": 0.008477923460304737, | |
| "learning_rate": 1.8341345686543332e-07, | |
| "loss": 0.0002, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.5527809307604994, | |
| "grad_norm": 0.31904277205467224, | |
| "learning_rate": 1.7008643629596866e-07, | |
| "loss": 0.0125, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.5533484676503972, | |
| "grad_norm": 0.02196822501718998, | |
| "learning_rate": 1.5726120240288634e-07, | |
| "loss": 0.0005, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.5539160045402951, | |
| "grad_norm": 0.03046775981783867, | |
| "learning_rate": 1.449378843361271e-07, | |
| "loss": 0.0004, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.554483541430193, | |
| "grad_norm": 0.021513327956199646, | |
| "learning_rate": 1.3311660619138578e-07, | |
| "loss": 0.0004, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.5550510783200908, | |
| "grad_norm": 0.01769246906042099, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 0.0004, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.5556186152099887, | |
| "grad_norm": 0.12877629697322845, | |
| "learning_rate": 1.109806407717462e-07, | |
| "loss": 0.0014, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.5561861520998865, | |
| "grad_norm": 0.3173483610153198, | |
| "learning_rate": 1.0066617640578368e-07, | |
| "loss": 0.0027, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5567536889897844, | |
| "grad_norm": 0.045826442539691925, | |
| "learning_rate": 9.085419777743465e-08, | |
| "loss": 0.0007, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.5573212258796821, | |
| "grad_norm": 0.01739875227212906, | |
| "learning_rate": 8.15448036932176e-08, | |
| "loss": 0.0002, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.55788876276958, | |
| "grad_norm": 0.04034648835659027, | |
| "learning_rate": 7.273808789862724e-08, | |
| "loss": 0.0005, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.5584562996594779, | |
| "grad_norm": 0.02119840681552887, | |
| "learning_rate": 6.443413907720186e-08, | |
| "loss": 0.0003, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.5590238365493757, | |
| "grad_norm": 0.006225614342838526, | |
| "learning_rate": 5.663304084960186e-08, | |
| "loss": 0.0001, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.5595913734392736, | |
| "grad_norm": 0.07903040200471878, | |
| "learning_rate": 4.933487177280482e-08, | |
| "loss": 0.0005, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.5601589103291714, | |
| "grad_norm": 0.5586972832679749, | |
| "learning_rate": 4.253970533929508e-08, | |
| "loss": 0.011, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.5607264472190693, | |
| "grad_norm": 0.9101510047912598, | |
| "learning_rate": 3.624760997631982e-08, | |
| "loss": 0.0241, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.5612939841089671, | |
| "grad_norm": 0.04058132693171501, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 0.0004, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.5618615209988649, | |
| "grad_norm": 0.017602860927581787, | |
| "learning_rate": 2.5172880840745873e-08, | |
| "loss": 0.0003, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5624290578887627, | |
| "grad_norm": 0.008799027651548386, | |
| "learning_rate": 2.0390358590538504e-08, | |
| "loss": 0.0002, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.5629965947786606, | |
| "grad_norm": 0.022075001150369644, | |
| "learning_rate": 1.6111130454543598e-08, | |
| "loss": 0.0003, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.5635641316685585, | |
| "grad_norm": 0.00847385823726654, | |
| "learning_rate": 1.2335239524541299e-08, | |
| "loss": 0.0002, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.5641316685584563, | |
| "grad_norm": 0.039732273668050766, | |
| "learning_rate": 9.06272382371065e-09, | |
| "loss": 0.0004, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.5646992054483542, | |
| "grad_norm": 0.022346949204802513, | |
| "learning_rate": 6.293616306246586e-09, | |
| "loss": 0.0003, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.565266742338252, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.0279448570323954e-09, | |
| "loss": 0.0, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.5658342792281499, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.265732291356626e-09, | |
| "loss": 0.0, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.5664018161180476, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.0069963546743832e-09, | |
| "loss": 0.0, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.5669693530079455, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.5174972244634833e-10, | |
| "loss": 0.0, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.5675368898978433, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5675368898978433, | |
| "eval_loss": NaN, | |
| "eval_runtime": 105.9662, | |
| "eval_samples_per_second": 28.009, | |
| "eval_steps_per_second": 7.002, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 4 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.3001540599808e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |