| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 193455, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007753741180119408, | |
| "grad_norm": 2.8055944442749023, | |
| "learning_rate": 4.9870770980331344e-05, | |
| "loss": 8.6951, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.015507482360238816, | |
| "grad_norm": 2.7864086627960205, | |
| "learning_rate": 4.974154196066269e-05, | |
| "loss": 6.4397, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.023261223540358224, | |
| "grad_norm": 2.835425853729248, | |
| "learning_rate": 4.961231294099403e-05, | |
| "loss": 5.8474, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.03101496472047763, | |
| "grad_norm": 3.0123701095581055, | |
| "learning_rate": 4.9483083921325374e-05, | |
| "loss": 5.5019, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.038768705900597035, | |
| "grad_norm": 2.9796297550201416, | |
| "learning_rate": 4.935385490165672e-05, | |
| "loss": 5.2444, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.04652244708071645, | |
| "grad_norm": 3.042416572570801, | |
| "learning_rate": 4.9224625881988064e-05, | |
| "loss": 5.0392, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.054276188260835856, | |
| "grad_norm": 2.9367804527282715, | |
| "learning_rate": 4.9095396862319405e-05, | |
| "loss": 4.8613, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.06202992944095526, | |
| "grad_norm": 2.9740755558013916, | |
| "learning_rate": 4.8966167842650746e-05, | |
| "loss": 4.7246, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.06978367062107467, | |
| "grad_norm": 3.364341974258423, | |
| "learning_rate": 4.8836938822982094e-05, | |
| "loss": 4.5969, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.07753741180119407, | |
| "grad_norm": 3.4089107513427734, | |
| "learning_rate": 4.8707709803313436e-05, | |
| "loss": 4.4929, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.08529115298131348, | |
| "grad_norm": 3.324126720428467, | |
| "learning_rate": 4.857848078364478e-05, | |
| "loss": 4.3966, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.0930448941614329, | |
| "grad_norm": 3.3835084438323975, | |
| "learning_rate": 4.8449251763976125e-05, | |
| "loss": 4.3176, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.1007986353415523, | |
| "grad_norm": 3.6219263076782227, | |
| "learning_rate": 4.8320022744307466e-05, | |
| "loss": 4.2318, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.10855237652167171, | |
| "grad_norm": 3.2887094020843506, | |
| "learning_rate": 4.819079372463881e-05, | |
| "loss": 4.161, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.11630611770179111, | |
| "grad_norm": 3.6443707942962646, | |
| "learning_rate": 4.806156470497015e-05, | |
| "loss": 4.0995, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.12405985888191053, | |
| "grad_norm": 3.623699903488159, | |
| "learning_rate": 4.79323356853015e-05, | |
| "loss": 4.0583, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.13181360006202994, | |
| "grad_norm": 3.6302268505096436, | |
| "learning_rate": 4.780310666563284e-05, | |
| "loss": 4.0, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.13956734124214934, | |
| "grad_norm": 3.6418278217315674, | |
| "learning_rate": 4.767387764596418e-05, | |
| "loss": 3.9486, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.14732108242226874, | |
| "grad_norm": 3.635089159011841, | |
| "learning_rate": 4.754464862629552e-05, | |
| "loss": 3.896, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.15507482360238814, | |
| "grad_norm": 3.8315062522888184, | |
| "learning_rate": 4.741541960662686e-05, | |
| "loss": 3.8629, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.16282856478250757, | |
| "grad_norm": 3.7439329624176025, | |
| "learning_rate": 4.728619058695821e-05, | |
| "loss": 3.8255, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.17058230596262697, | |
| "grad_norm": 3.732562303543091, | |
| "learning_rate": 4.715696156728955e-05, | |
| "loss": 3.7887, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.17833604714274637, | |
| "grad_norm": 3.7784509658813477, | |
| "learning_rate": 4.702773254762089e-05, | |
| "loss": 3.749, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.1860897883228658, | |
| "grad_norm": 3.9984617233276367, | |
| "learning_rate": 4.6898503527952234e-05, | |
| "loss": 3.7065, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.1938435295029852, | |
| "grad_norm": 3.954652786254883, | |
| "learning_rate": 4.676927450828358e-05, | |
| "loss": 3.6931, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.2015972706831046, | |
| "grad_norm": 3.9958720207214355, | |
| "learning_rate": 4.6640045488614924e-05, | |
| "loss": 3.6545, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.209351011863224, | |
| "grad_norm": 3.922849178314209, | |
| "learning_rate": 4.6510816468946265e-05, | |
| "loss": 3.6207, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.21710475304334342, | |
| "grad_norm": 4.092968463897705, | |
| "learning_rate": 4.638158744927761e-05, | |
| "loss": 3.5974, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.22485849422346282, | |
| "grad_norm": 3.8097622394561768, | |
| "learning_rate": 4.6252358429608954e-05, | |
| "loss": 3.5775, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.23261223540358222, | |
| "grad_norm": 4.094768047332764, | |
| "learning_rate": 4.6123129409940296e-05, | |
| "loss": 3.5543, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.24036597658370162, | |
| "grad_norm": 3.8095757961273193, | |
| "learning_rate": 4.5993900390271644e-05, | |
| "loss": 3.5392, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.24811971776382105, | |
| "grad_norm": 3.8520195484161377, | |
| "learning_rate": 4.5864671370602985e-05, | |
| "loss": 3.5047, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.2558734589439404, | |
| "grad_norm": 4.280955791473389, | |
| "learning_rate": 4.5735442350934326e-05, | |
| "loss": 3.495, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.2636272001240599, | |
| "grad_norm": 4.455172538757324, | |
| "learning_rate": 4.5606213331265674e-05, | |
| "loss": 3.4756, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.2713809413041793, | |
| "grad_norm": 4.1695733070373535, | |
| "learning_rate": 4.5476984311597016e-05, | |
| "loss": 3.4505, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.2791346824842987, | |
| "grad_norm": 4.223578453063965, | |
| "learning_rate": 4.534775529192836e-05, | |
| "loss": 3.4446, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.2868884236644181, | |
| "grad_norm": 4.616567134857178, | |
| "learning_rate": 4.52185262722597e-05, | |
| "loss": 3.4202, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.2946421648445375, | |
| "grad_norm": 4.264205455780029, | |
| "learning_rate": 4.5089297252591046e-05, | |
| "loss": 3.4024, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.3023959060246569, | |
| "grad_norm": 4.493732929229736, | |
| "learning_rate": 4.496006823292239e-05, | |
| "loss": 3.397, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.3101496472047763, | |
| "grad_norm": 4.321922779083252, | |
| "learning_rate": 4.483083921325373e-05, | |
| "loss": 3.3678, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.31790338838489574, | |
| "grad_norm": 4.248241424560547, | |
| "learning_rate": 4.470161019358508e-05, | |
| "loss": 3.3634, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.32565712956501514, | |
| "grad_norm": 4.490056037902832, | |
| "learning_rate": 4.457238117391642e-05, | |
| "loss": 3.3529, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.33341087074513454, | |
| "grad_norm": 4.652819633483887, | |
| "learning_rate": 4.444315215424776e-05, | |
| "loss": 3.3491, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.34116461192525394, | |
| "grad_norm": 4.65127420425415, | |
| "learning_rate": 4.431392313457911e-05, | |
| "loss": 3.3239, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.34891835310537334, | |
| "grad_norm": 4.279469013214111, | |
| "learning_rate": 4.418469411491045e-05, | |
| "loss": 3.3061, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.35667209428549274, | |
| "grad_norm": 4.562509059906006, | |
| "learning_rate": 4.405546509524179e-05, | |
| "loss": 3.3014, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.36442583546561214, | |
| "grad_norm": 4.664194583892822, | |
| "learning_rate": 4.392623607557314e-05, | |
| "loss": 3.2963, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.3721795766457316, | |
| "grad_norm": 4.531868934631348, | |
| "learning_rate": 4.379700705590448e-05, | |
| "loss": 3.2776, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.379933317825851, | |
| "grad_norm": 4.553177833557129, | |
| "learning_rate": 4.366777803623582e-05, | |
| "loss": 3.2559, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.3876870590059704, | |
| "grad_norm": 4.358414649963379, | |
| "learning_rate": 4.353854901656716e-05, | |
| "loss": 3.253, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.3954408001860898, | |
| "grad_norm": 4.921834468841553, | |
| "learning_rate": 4.3409319996898503e-05, | |
| "loss": 3.257, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.4031945413662092, | |
| "grad_norm": 4.523054599761963, | |
| "learning_rate": 4.3280090977229845e-05, | |
| "loss": 3.2368, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.4109482825463286, | |
| "grad_norm": 4.663913249969482, | |
| "learning_rate": 4.315086195756119e-05, | |
| "loss": 3.2314, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.418702023726448, | |
| "grad_norm": 4.529434680938721, | |
| "learning_rate": 4.3021632937892534e-05, | |
| "loss": 3.2216, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.4264557649065674, | |
| "grad_norm": 4.806913375854492, | |
| "learning_rate": 4.2892403918223875e-05, | |
| "loss": 3.2176, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.43420950608668685, | |
| "grad_norm": 4.8940019607543945, | |
| "learning_rate": 4.276317489855522e-05, | |
| "loss": 3.2061, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.44196324726680625, | |
| "grad_norm": 4.507465839385986, | |
| "learning_rate": 4.2633945878886565e-05, | |
| "loss": 3.1957, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.44971698844692565, | |
| "grad_norm": 5.141822338104248, | |
| "learning_rate": 4.2504716859217906e-05, | |
| "loss": 3.1849, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.45747072962704505, | |
| "grad_norm": 4.68912935256958, | |
| "learning_rate": 4.237548783954925e-05, | |
| "loss": 3.1707, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.46522447080716445, | |
| "grad_norm": 5.046304225921631, | |
| "learning_rate": 4.2246258819880595e-05, | |
| "loss": 3.1705, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.47297821198728385, | |
| "grad_norm": 4.986753940582275, | |
| "learning_rate": 4.211702980021194e-05, | |
| "loss": 3.1641, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.48073195316740325, | |
| "grad_norm": 5.041975498199463, | |
| "learning_rate": 4.198780078054328e-05, | |
| "loss": 3.1545, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.4884856943475227, | |
| "grad_norm": 5.255220413208008, | |
| "learning_rate": 4.1858571760874626e-05, | |
| "loss": 3.1566, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.4962394355276421, | |
| "grad_norm": 5.2884697914123535, | |
| "learning_rate": 4.172934274120597e-05, | |
| "loss": 3.1497, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.5039931767077614, | |
| "grad_norm": 4.990237712860107, | |
| "learning_rate": 4.160011372153731e-05, | |
| "loss": 3.1417, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.5117469178878808, | |
| "grad_norm": 5.165491104125977, | |
| "learning_rate": 4.147088470186866e-05, | |
| "loss": 3.1227, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.5195006590680004, | |
| "grad_norm": 5.079947471618652, | |
| "learning_rate": 4.13416556822e-05, | |
| "loss": 3.1276, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 0.5272544002481198, | |
| "grad_norm": 4.830618858337402, | |
| "learning_rate": 4.121242666253134e-05, | |
| "loss": 3.1195, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.5350081414282392, | |
| "grad_norm": 5.067671775817871, | |
| "learning_rate": 4.108319764286268e-05, | |
| "loss": 3.1168, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 0.5427618826083586, | |
| "grad_norm": 5.153751373291016, | |
| "learning_rate": 4.095396862319403e-05, | |
| "loss": 3.1087, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.550515623788478, | |
| "grad_norm": 5.0933027267456055, | |
| "learning_rate": 4.082473960352537e-05, | |
| "loss": 3.0945, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 0.5582693649685974, | |
| "grad_norm": 5.372694492340088, | |
| "learning_rate": 4.069551058385671e-05, | |
| "loss": 3.0906, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.5660231061487168, | |
| "grad_norm": 5.414623737335205, | |
| "learning_rate": 4.056628156418806e-05, | |
| "loss": 3.0894, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 0.5737768473288362, | |
| "grad_norm": 5.126060962677002, | |
| "learning_rate": 4.04370525445194e-05, | |
| "loss": 3.0848, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.5815305885089556, | |
| "grad_norm": 4.877682209014893, | |
| "learning_rate": 4.030782352485074e-05, | |
| "loss": 3.0752, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 0.589284329689075, | |
| "grad_norm": 5.10503625869751, | |
| "learning_rate": 4.017859450518209e-05, | |
| "loss": 3.0678, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.5970380708691944, | |
| "grad_norm": 5.342639923095703, | |
| "learning_rate": 4.004936548551343e-05, | |
| "loss": 3.074, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 0.6047918120493138, | |
| "grad_norm": 5.377997875213623, | |
| "learning_rate": 3.992013646584477e-05, | |
| "loss": 3.0633, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.6125455532294332, | |
| "grad_norm": 5.535395622253418, | |
| "learning_rate": 3.979090744617612e-05, | |
| "loss": 3.0667, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 0.6202992944095526, | |
| "grad_norm": 5.380632400512695, | |
| "learning_rate": 3.966167842650746e-05, | |
| "loss": 3.042, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.6280530355896721, | |
| "grad_norm": 5.2183027267456055, | |
| "learning_rate": 3.95324494068388e-05, | |
| "loss": 3.0488, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 0.6358067767697915, | |
| "grad_norm": 5.565989971160889, | |
| "learning_rate": 3.9403220387170145e-05, | |
| "loss": 3.0477, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.6435605179499109, | |
| "grad_norm": 5.33280611038208, | |
| "learning_rate": 3.9273991367501486e-05, | |
| "loss": 3.0429, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 0.6513142591300303, | |
| "grad_norm": 5.605506896972656, | |
| "learning_rate": 3.9144762347832834e-05, | |
| "loss": 3.0423, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.6590680003101497, | |
| "grad_norm": 5.62191915512085, | |
| "learning_rate": 3.9015533328164175e-05, | |
| "loss": 3.0205, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 0.6668217414902691, | |
| "grad_norm": 5.318583965301514, | |
| "learning_rate": 3.888630430849552e-05, | |
| "loss": 3.0277, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.6745754826703885, | |
| "grad_norm": 5.763892650604248, | |
| "learning_rate": 3.875707528882686e-05, | |
| "loss": 3.0342, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 0.6823292238505079, | |
| "grad_norm": 5.43319034576416, | |
| "learning_rate": 3.86278462691582e-05, | |
| "loss": 3.0134, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.6900829650306273, | |
| "grad_norm": 5.857462406158447, | |
| "learning_rate": 3.849861724948955e-05, | |
| "loss": 3.0091, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 0.6978367062107467, | |
| "grad_norm": 5.6617021560668945, | |
| "learning_rate": 3.836938822982089e-05, | |
| "loss": 3.0117, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.7055904473908661, | |
| "grad_norm": 5.857166290283203, | |
| "learning_rate": 3.824015921015223e-05, | |
| "loss": 3.0189, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 0.7133441885709855, | |
| "grad_norm": 5.425033092498779, | |
| "learning_rate": 3.811093019048358e-05, | |
| "loss": 3.019, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.7210979297511049, | |
| "grad_norm": 5.529470920562744, | |
| "learning_rate": 3.798170117081492e-05, | |
| "loss": 3.002, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 0.7288516709312243, | |
| "grad_norm": 5.534538745880127, | |
| "learning_rate": 3.785247215114626e-05, | |
| "loss": 2.995, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.7366054121113437, | |
| "grad_norm": 5.707602024078369, | |
| "learning_rate": 3.772324313147761e-05, | |
| "loss": 2.9942, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 0.7443591532914632, | |
| "grad_norm": 5.565040588378906, | |
| "learning_rate": 3.759401411180895e-05, | |
| "loss": 2.9947, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.7521128944715826, | |
| "grad_norm": 5.714329242706299, | |
| "learning_rate": 3.746478509214029e-05, | |
| "loss": 2.9941, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 0.759866635651702, | |
| "grad_norm": 5.562424182891846, | |
| "learning_rate": 3.733555607247163e-05, | |
| "loss": 2.9827, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 0.7676203768318214, | |
| "grad_norm": 5.685086250305176, | |
| "learning_rate": 3.720632705280298e-05, | |
| "loss": 2.9812, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 0.7753741180119408, | |
| "grad_norm": 5.563987731933594, | |
| "learning_rate": 3.707709803313432e-05, | |
| "loss": 2.9781, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.7831278591920602, | |
| "grad_norm": 5.545105934143066, | |
| "learning_rate": 3.694786901346566e-05, | |
| "loss": 2.9586, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 0.7908816003721796, | |
| "grad_norm": 5.9238386154174805, | |
| "learning_rate": 3.681863999379701e-05, | |
| "loss": 2.9713, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 0.798635341552299, | |
| "grad_norm": 5.929417133331299, | |
| "learning_rate": 3.668941097412835e-05, | |
| "loss": 2.9623, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 0.8063890827324184, | |
| "grad_norm": 5.734675884246826, | |
| "learning_rate": 3.6560181954459694e-05, | |
| "loss": 2.9659, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 0.8141428239125378, | |
| "grad_norm": 5.726919651031494, | |
| "learning_rate": 3.643095293479104e-05, | |
| "loss": 2.9624, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 0.8218965650926572, | |
| "grad_norm": 5.474340438842773, | |
| "learning_rate": 3.630172391512238e-05, | |
| "loss": 2.9597, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 0.8296503062727766, | |
| "grad_norm": 5.879449367523193, | |
| "learning_rate": 3.6172494895453725e-05, | |
| "loss": 2.9601, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 0.837404047452896, | |
| "grad_norm": 6.051661491394043, | |
| "learning_rate": 3.604326587578507e-05, | |
| "loss": 2.9606, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 0.8451577886330154, | |
| "grad_norm": 6.062263488769531, | |
| "learning_rate": 3.5914036856116414e-05, | |
| "loss": 2.9573, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 0.8529115298131348, | |
| "grad_norm": 5.804770469665527, | |
| "learning_rate": 3.5784807836447755e-05, | |
| "loss": 2.9512, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.8606652709932543, | |
| "grad_norm": 5.899106025695801, | |
| "learning_rate": 3.5655578816779097e-05, | |
| "loss": 2.9563, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 0.8684190121733737, | |
| "grad_norm": 5.6610894203186035, | |
| "learning_rate": 3.5526349797110445e-05, | |
| "loss": 2.9468, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 0.8761727533534931, | |
| "grad_norm": 5.676564693450928, | |
| "learning_rate": 3.5397120777441786e-05, | |
| "loss": 2.9442, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 0.8839264945336125, | |
| "grad_norm": 6.026761531829834, | |
| "learning_rate": 3.526789175777313e-05, | |
| "loss": 2.945, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 0.8916802357137319, | |
| "grad_norm": 6.285312652587891, | |
| "learning_rate": 3.5138662738104475e-05, | |
| "loss": 2.935, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 0.8994339768938513, | |
| "grad_norm": 5.787561416625977, | |
| "learning_rate": 3.5009433718435817e-05, | |
| "loss": 2.9395, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 0.9071877180739707, | |
| "grad_norm": 5.658621311187744, | |
| "learning_rate": 3.488020469876716e-05, | |
| "loss": 2.9288, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 0.9149414592540901, | |
| "grad_norm": 5.896640300750732, | |
| "learning_rate": 3.47509756790985e-05, | |
| "loss": 2.9351, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 0.9226952004342095, | |
| "grad_norm": 6.219537734985352, | |
| "learning_rate": 3.462174665942984e-05, | |
| "loss": 2.9257, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 0.9304489416143289, | |
| "grad_norm": 6.03794527053833, | |
| "learning_rate": 3.449251763976118e-05, | |
| "loss": 2.924, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 0.9382026827944483, | |
| "grad_norm": 6.291288375854492, | |
| "learning_rate": 3.436328862009253e-05, | |
| "loss": 2.9141, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 0.9459564239745677, | |
| "grad_norm": 6.239747524261475, | |
| "learning_rate": 3.423405960042387e-05, | |
| "loss": 2.9212, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 0.9537101651546871, | |
| "grad_norm": 5.852220058441162, | |
| "learning_rate": 3.410483058075521e-05, | |
| "loss": 2.9187, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 0.9614639063348065, | |
| "grad_norm": 5.890344142913818, | |
| "learning_rate": 3.397560156108656e-05, | |
| "loss": 2.9167, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 0.969217647514926, | |
| "grad_norm": 6.417314529418945, | |
| "learning_rate": 3.38463725414179e-05, | |
| "loss": 2.9147, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 0.9769713886950454, | |
| "grad_norm": 6.077672481536865, | |
| "learning_rate": 3.371714352174924e-05, | |
| "loss": 2.9144, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 0.9847251298751648, | |
| "grad_norm": 6.114253520965576, | |
| "learning_rate": 3.3587914502080584e-05, | |
| "loss": 2.9025, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 0.9924788710552842, | |
| "grad_norm": 6.094882965087891, | |
| "learning_rate": 3.345868548241193e-05, | |
| "loss": 2.9058, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 1.0002326122354035, | |
| "grad_norm": 5.9865498542785645, | |
| "learning_rate": 3.3329456462743274e-05, | |
| "loss": 2.9067, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 1.007986353415523, | |
| "grad_norm": 5.904115200042725, | |
| "learning_rate": 3.3200227443074615e-05, | |
| "loss": 2.8973, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 1.0157400945956423, | |
| "grad_norm": 6.1497392654418945, | |
| "learning_rate": 3.307099842340596e-05, | |
| "loss": 2.897, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 1.0234938357757617, | |
| "grad_norm": 6.136323928833008, | |
| "learning_rate": 3.2941769403737304e-05, | |
| "loss": 2.9023, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 1.0312475769558813, | |
| "grad_norm": 6.581076145172119, | |
| "learning_rate": 3.2812540384068646e-05, | |
| "loss": 2.8962, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 1.0390013181360007, | |
| "grad_norm": 6.236713886260986, | |
| "learning_rate": 3.2683311364399994e-05, | |
| "loss": 2.8953, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 1.0467550593161201, | |
| "grad_norm": 6.375150680541992, | |
| "learning_rate": 3.2554082344731335e-05, | |
| "loss": 2.8951, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 1.0545088004962395, | |
| "grad_norm": 6.454219818115234, | |
| "learning_rate": 3.2424853325062676e-05, | |
| "loss": 2.8902, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 1.062262541676359, | |
| "grad_norm": 6.4552226066589355, | |
| "learning_rate": 3.2295624305394024e-05, | |
| "loss": 2.9028, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 1.0700162828564783, | |
| "grad_norm": 6.323915481567383, | |
| "learning_rate": 3.2166395285725366e-05, | |
| "loss": 2.8793, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 1.0777700240365977, | |
| "grad_norm": 6.413472652435303, | |
| "learning_rate": 3.203716626605671e-05, | |
| "loss": 2.8704, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 1.0855237652167171, | |
| "grad_norm": 6.230509281158447, | |
| "learning_rate": 3.190793724638805e-05, | |
| "loss": 2.8867, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 1.0932775063968365, | |
| "grad_norm": 6.334902286529541, | |
| "learning_rate": 3.1778708226719396e-05, | |
| "loss": 2.8851, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 1.101031247576956, | |
| "grad_norm": 6.225689888000488, | |
| "learning_rate": 3.164947920705074e-05, | |
| "loss": 2.8783, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 1.1087849887570753, | |
| "grad_norm": 6.088467597961426, | |
| "learning_rate": 3.152025018738208e-05, | |
| "loss": 2.8837, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 1.1165387299371947, | |
| "grad_norm": 6.089876174926758, | |
| "learning_rate": 3.139102116771343e-05, | |
| "loss": 2.8681, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 1.1242924711173141, | |
| "grad_norm": 6.693448066711426, | |
| "learning_rate": 3.126179214804477e-05, | |
| "loss": 2.8678, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 1.1320462122974335, | |
| "grad_norm": 6.45464563369751, | |
| "learning_rate": 3.113256312837611e-05, | |
| "loss": 2.8717, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 1.139799953477553, | |
| "grad_norm": 6.7853803634643555, | |
| "learning_rate": 3.100333410870746e-05, | |
| "loss": 2.8607, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 1.1475536946576723, | |
| "grad_norm": 6.709279537200928, | |
| "learning_rate": 3.08741050890388e-05, | |
| "loss": 2.862, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 1.1553074358377917, | |
| "grad_norm": 6.519172668457031, | |
| "learning_rate": 3.074487606937014e-05, | |
| "loss": 2.8704, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 1.1630611770179111, | |
| "grad_norm": 6.5100507736206055, | |
| "learning_rate": 3.061564704970148e-05, | |
| "loss": 2.8636, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 1.1708149181980305, | |
| "grad_norm": 6.233548164367676, | |
| "learning_rate": 3.0486418030032826e-05, | |
| "loss": 2.8773, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 1.17856865937815, | |
| "grad_norm": 6.704037666320801, | |
| "learning_rate": 3.0357189010364168e-05, | |
| "loss": 2.8704, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 1.1863224005582693, | |
| "grad_norm": 6.638470649719238, | |
| "learning_rate": 3.0227959990695516e-05, | |
| "loss": 2.8708, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 1.1940761417383887, | |
| "grad_norm": 6.931632995605469, | |
| "learning_rate": 3.0098730971026857e-05, | |
| "loss": 2.8619, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 1.2018298829185081, | |
| "grad_norm": 6.729213714599609, | |
| "learning_rate": 2.99695019513582e-05, | |
| "loss": 2.8558, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 1.2095836240986275, | |
| "grad_norm": 6.931024551391602, | |
| "learning_rate": 2.984027293168954e-05, | |
| "loss": 2.8559, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 1.217337365278747, | |
| "grad_norm": 6.658525466918945, | |
| "learning_rate": 2.9711043912020888e-05, | |
| "loss": 2.8573, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 1.2250911064588663, | |
| "grad_norm": 6.135016918182373, | |
| "learning_rate": 2.958181489235223e-05, | |
| "loss": 2.8604, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 1.2328448476389857, | |
| "grad_norm": 6.685146331787109, | |
| "learning_rate": 2.945258587268357e-05, | |
| "loss": 2.844, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 1.2405985888191053, | |
| "grad_norm": 6.7349982261657715, | |
| "learning_rate": 2.9323356853014915e-05, | |
| "loss": 2.8451, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 1.2483523299992245, | |
| "grad_norm": 6.538317680358887, | |
| "learning_rate": 2.9194127833346256e-05, | |
| "loss": 2.8531, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 1.2561060711793441, | |
| "grad_norm": 6.23037576675415, | |
| "learning_rate": 2.9064898813677598e-05, | |
| "loss": 2.8454, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 1.2638598123594633, | |
| "grad_norm": 6.324411392211914, | |
| "learning_rate": 2.8935669794008946e-05, | |
| "loss": 2.8462, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 1.271613553539583, | |
| "grad_norm": 6.693195343017578, | |
| "learning_rate": 2.8806440774340287e-05, | |
| "loss": 2.8382, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 1.2793672947197021, | |
| "grad_norm": 6.47921085357666, | |
| "learning_rate": 2.8677211754671628e-05, | |
| "loss": 2.8492, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 1.2871210358998217, | |
| "grad_norm": 6.768202304840088, | |
| "learning_rate": 2.8547982735002976e-05, | |
| "loss": 2.8344, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 1.2948747770799411, | |
| "grad_norm": 6.594978332519531, | |
| "learning_rate": 2.8418753715334318e-05, | |
| "loss": 2.8337, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 1.3026285182600605, | |
| "grad_norm": 6.703185081481934, | |
| "learning_rate": 2.828952469566566e-05, | |
| "loss": 2.834, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 1.31038225944018, | |
| "grad_norm": 6.614627838134766, | |
| "learning_rate": 2.8160295675997e-05, | |
| "loss": 2.835, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 1.3181360006202993, | |
| "grad_norm": 6.38785457611084, | |
| "learning_rate": 2.8031066656328348e-05, | |
| "loss": 2.8439, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 1.3258897418004187, | |
| "grad_norm": 6.479560852050781, | |
| "learning_rate": 2.790183763665969e-05, | |
| "loss": 2.8338, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 1.3336434829805381, | |
| "grad_norm": 6.241596698760986, | |
| "learning_rate": 2.777260861699103e-05, | |
| "loss": 2.8304, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 1.3413972241606575, | |
| "grad_norm": 7.731629371643066, | |
| "learning_rate": 2.764337959732238e-05, | |
| "loss": 2.828, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 1.349150965340777, | |
| "grad_norm": 6.706104278564453, | |
| "learning_rate": 2.751415057765372e-05, | |
| "loss": 2.8297, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 1.3569047065208963, | |
| "grad_norm": 6.772350311279297, | |
| "learning_rate": 2.738492155798506e-05, | |
| "loss": 2.8331, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 1.3646584477010157, | |
| "grad_norm": 6.890201091766357, | |
| "learning_rate": 2.7255692538316406e-05, | |
| "loss": 2.8286, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 1.3724121888811351, | |
| "grad_norm": 6.540558815002441, | |
| "learning_rate": 2.7126463518647748e-05, | |
| "loss": 2.8306, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 1.3801659300612545, | |
| "grad_norm": 6.890176773071289, | |
| "learning_rate": 2.699723449897909e-05, | |
| "loss": 2.827, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 1.387919671241374, | |
| "grad_norm": 6.634540557861328, | |
| "learning_rate": 2.6868005479310437e-05, | |
| "loss": 2.8263, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 1.3956734124214933, | |
| "grad_norm": 7.228022575378418, | |
| "learning_rate": 2.6738776459641778e-05, | |
| "loss": 2.8365, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 1.4034271536016127, | |
| "grad_norm": 6.7347869873046875, | |
| "learning_rate": 2.660954743997312e-05, | |
| "loss": 2.825, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 1.4111808947817321, | |
| "grad_norm": 6.936458110809326, | |
| "learning_rate": 2.6480318420304468e-05, | |
| "loss": 2.8266, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 1.4189346359618515, | |
| "grad_norm": 6.489315509796143, | |
| "learning_rate": 2.635108940063581e-05, | |
| "loss": 2.8173, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 1.426688377141971, | |
| "grad_norm": 7.1031012535095215, | |
| "learning_rate": 2.622186038096715e-05, | |
| "loss": 2.8114, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 1.4344421183220903, | |
| "grad_norm": 6.918934345245361, | |
| "learning_rate": 2.609263136129849e-05, | |
| "loss": 2.8252, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 1.4421958595022097, | |
| "grad_norm": 6.686205863952637, | |
| "learning_rate": 2.596340234162984e-05, | |
| "loss": 2.8138, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 1.4499496006823291, | |
| "grad_norm": 6.464860439300537, | |
| "learning_rate": 2.583417332196118e-05, | |
| "loss": 2.8284, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 1.4577033418624485, | |
| "grad_norm": 6.871826171875, | |
| "learning_rate": 2.5704944302292522e-05, | |
| "loss": 2.8215, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 1.4654570830425682, | |
| "grad_norm": 6.555510520935059, | |
| "learning_rate": 2.557571528262387e-05, | |
| "loss": 2.8166, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 1.4732108242226873, | |
| "grad_norm": 6.931303977966309, | |
| "learning_rate": 2.544648626295521e-05, | |
| "loss": 2.8218, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 1.480964565402807, | |
| "grad_norm": 6.583662509918213, | |
| "learning_rate": 2.5317257243286553e-05, | |
| "loss": 2.8075, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 1.4887183065829261, | |
| "grad_norm": 6.624995231628418, | |
| "learning_rate": 2.5188028223617897e-05, | |
| "loss": 2.8281, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 1.4964720477630458, | |
| "grad_norm": 6.899562835693359, | |
| "learning_rate": 2.505879920394924e-05, | |
| "loss": 2.8195, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 1.504225788943165, | |
| "grad_norm": 6.846054553985596, | |
| "learning_rate": 2.4929570184280583e-05, | |
| "loss": 2.8149, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 1.5119795301232846, | |
| "grad_norm": 6.590377330780029, | |
| "learning_rate": 2.4800341164611925e-05, | |
| "loss": 2.7984, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 1.5197332713034037, | |
| "grad_norm": 6.771044731140137, | |
| "learning_rate": 2.467111214494327e-05, | |
| "loss": 2.82, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 1.5274870124835234, | |
| "grad_norm": 6.996868133544922, | |
| "learning_rate": 2.454188312527461e-05, | |
| "loss": 2.803, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 1.5352407536636425, | |
| "grad_norm": 6.782078742980957, | |
| "learning_rate": 2.4412654105605955e-05, | |
| "loss": 2.8156, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 1.5429944948437622, | |
| "grad_norm": 7.141603469848633, | |
| "learning_rate": 2.42834250859373e-05, | |
| "loss": 2.8081, | |
| "step": 99500 | |
| }, | |
| { | |
| "epoch": 1.5507482360238816, | |
| "grad_norm": 7.204538822174072, | |
| "learning_rate": 2.415419606626864e-05, | |
| "loss": 2.7939, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 1.558501977204001, | |
| "grad_norm": 7.218080043792725, | |
| "learning_rate": 2.4024967046599986e-05, | |
| "loss": 2.8011, | |
| "step": 100500 | |
| }, | |
| { | |
| "epoch": 1.5662557183841204, | |
| "grad_norm": 6.774955749511719, | |
| "learning_rate": 2.3895738026931327e-05, | |
| "loss": 2.8086, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 1.5740094595642398, | |
| "grad_norm": 6.7942657470703125, | |
| "learning_rate": 2.3766509007262672e-05, | |
| "loss": 2.7981, | |
| "step": 101500 | |
| }, | |
| { | |
| "epoch": 1.5817632007443592, | |
| "grad_norm": 6.575582027435303, | |
| "learning_rate": 2.3637279987594017e-05, | |
| "loss": 2.8048, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 1.5895169419244786, | |
| "grad_norm": 6.921658515930176, | |
| "learning_rate": 2.3508050967925358e-05, | |
| "loss": 2.8001, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 1.597270683104598, | |
| "grad_norm": 7.207976341247559, | |
| "learning_rate": 2.3378821948256703e-05, | |
| "loss": 2.8112, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 1.6050244242847174, | |
| "grad_norm": 7.6573710441589355, | |
| "learning_rate": 2.3249592928588047e-05, | |
| "loss": 2.8044, | |
| "step": 103500 | |
| }, | |
| { | |
| "epoch": 1.6127781654648368, | |
| "grad_norm": 7.072439670562744, | |
| "learning_rate": 2.312036390891939e-05, | |
| "loss": 2.7871, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 1.6205319066449562, | |
| "grad_norm": 7.3211259841918945, | |
| "learning_rate": 2.299113488925073e-05, | |
| "loss": 2.7921, | |
| "step": 104500 | |
| }, | |
| { | |
| "epoch": 1.6282856478250756, | |
| "grad_norm": 7.107245445251465, | |
| "learning_rate": 2.2861905869582075e-05, | |
| "loss": 2.8031, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 1.636039389005195, | |
| "grad_norm": 6.947020530700684, | |
| "learning_rate": 2.2732676849913416e-05, | |
| "loss": 2.7869, | |
| "step": 105500 | |
| }, | |
| { | |
| "epoch": 1.6437931301853144, | |
| "grad_norm": 7.2329301834106445, | |
| "learning_rate": 2.260344783024476e-05, | |
| "loss": 2.7876, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 1.6515468713654338, | |
| "grad_norm": 6.861079692840576, | |
| "learning_rate": 2.2474218810576102e-05, | |
| "loss": 2.7875, | |
| "step": 106500 | |
| }, | |
| { | |
| "epoch": 1.6593006125455534, | |
| "grad_norm": 7.405232906341553, | |
| "learning_rate": 2.2344989790907447e-05, | |
| "loss": 2.7955, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 1.6670543537256726, | |
| "grad_norm": 7.370352745056152, | |
| "learning_rate": 2.221576077123879e-05, | |
| "loss": 2.7931, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 1.6748080949057922, | |
| "grad_norm": 7.008327960968018, | |
| "learning_rate": 2.2086531751570133e-05, | |
| "loss": 2.7936, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 1.6825618360859114, | |
| "grad_norm": 7.268221378326416, | |
| "learning_rate": 2.1957302731901477e-05, | |
| "loss": 2.7969, | |
| "step": 108500 | |
| }, | |
| { | |
| "epoch": 1.690315577266031, | |
| "grad_norm": 6.869812488555908, | |
| "learning_rate": 2.182807371223282e-05, | |
| "loss": 2.7879, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 1.6980693184461502, | |
| "grad_norm": 6.86647891998291, | |
| "learning_rate": 2.1698844692564163e-05, | |
| "loss": 2.7992, | |
| "step": 109500 | |
| }, | |
| { | |
| "epoch": 1.7058230596262698, | |
| "grad_norm": 7.082624435424805, | |
| "learning_rate": 2.1569615672895508e-05, | |
| "loss": 2.7829, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 1.713576800806389, | |
| "grad_norm": 6.880459308624268, | |
| "learning_rate": 2.144038665322685e-05, | |
| "loss": 2.7797, | |
| "step": 110500 | |
| }, | |
| { | |
| "epoch": 1.7213305419865086, | |
| "grad_norm": 7.15917444229126, | |
| "learning_rate": 2.1311157633558194e-05, | |
| "loss": 2.7952, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 1.7290842831666278, | |
| "grad_norm": 7.239593982696533, | |
| "learning_rate": 2.118192861388954e-05, | |
| "loss": 2.7836, | |
| "step": 111500 | |
| }, | |
| { | |
| "epoch": 1.7368380243467474, | |
| "grad_norm": 6.907558441162109, | |
| "learning_rate": 2.105269959422088e-05, | |
| "loss": 2.7974, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 1.7445917655268666, | |
| "grad_norm": 7.07895040512085, | |
| "learning_rate": 2.0923470574552225e-05, | |
| "loss": 2.7847, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 1.7523455067069862, | |
| "grad_norm": 6.944314956665039, | |
| "learning_rate": 2.0794241554883566e-05, | |
| "loss": 2.7875, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 1.7600992478871054, | |
| "grad_norm": 6.936674118041992, | |
| "learning_rate": 2.0665012535214907e-05, | |
| "loss": 2.7743, | |
| "step": 113500 | |
| }, | |
| { | |
| "epoch": 1.767852989067225, | |
| "grad_norm": 7.579113960266113, | |
| "learning_rate": 2.0535783515546252e-05, | |
| "loss": 2.7748, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 1.7756067302473444, | |
| "grad_norm": 6.939824104309082, | |
| "learning_rate": 2.0406554495877593e-05, | |
| "loss": 2.7813, | |
| "step": 114500 | |
| }, | |
| { | |
| "epoch": 1.7833604714274638, | |
| "grad_norm": 6.831909656524658, | |
| "learning_rate": 2.0277325476208938e-05, | |
| "loss": 2.7755, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 1.7911142126075832, | |
| "grad_norm": 6.841889381408691, | |
| "learning_rate": 2.0148096456540283e-05, | |
| "loss": 2.7833, | |
| "step": 115500 | |
| }, | |
| { | |
| "epoch": 1.7988679537877026, | |
| "grad_norm": 6.934596538543701, | |
| "learning_rate": 2.0018867436871624e-05, | |
| "loss": 2.7816, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 1.806621694967822, | |
| "grad_norm": 7.232493877410889, | |
| "learning_rate": 1.988963841720297e-05, | |
| "loss": 2.7718, | |
| "step": 116500 | |
| }, | |
| { | |
| "epoch": 1.8143754361479414, | |
| "grad_norm": 6.8913421630859375, | |
| "learning_rate": 1.976040939753431e-05, | |
| "loss": 2.7806, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 1.8221291773280608, | |
| "grad_norm": 7.189756393432617, | |
| "learning_rate": 1.9631180377865655e-05, | |
| "loss": 2.7696, | |
| "step": 117500 | |
| }, | |
| { | |
| "epoch": 1.8298829185081802, | |
| "grad_norm": 7.165264129638672, | |
| "learning_rate": 1.9501951358197e-05, | |
| "loss": 2.772, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 1.8376366596882996, | |
| "grad_norm": 7.418449878692627, | |
| "learning_rate": 1.937272233852834e-05, | |
| "loss": 2.7673, | |
| "step": 118500 | |
| }, | |
| { | |
| "epoch": 1.845390400868419, | |
| "grad_norm": 7.016151428222656, | |
| "learning_rate": 1.9243493318859685e-05, | |
| "loss": 2.7721, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 1.8531441420485384, | |
| "grad_norm": 7.69176721572876, | |
| "learning_rate": 1.9114264299191027e-05, | |
| "loss": 2.7712, | |
| "step": 119500 | |
| }, | |
| { | |
| "epoch": 1.8608978832286578, | |
| "grad_norm": 7.078608512878418, | |
| "learning_rate": 1.898503527952237e-05, | |
| "loss": 2.7626, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 1.8686516244087772, | |
| "grad_norm": 7.167757034301758, | |
| "learning_rate": 1.8855806259853716e-05, | |
| "loss": 2.7757, | |
| "step": 120500 | |
| }, | |
| { | |
| "epoch": 1.8764053655888966, | |
| "grad_norm": 7.261277675628662, | |
| "learning_rate": 1.8726577240185057e-05, | |
| "loss": 2.7638, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 1.8841591067690162, | |
| "grad_norm": 7.32155179977417, | |
| "learning_rate": 1.85973482205164e-05, | |
| "loss": 2.7743, | |
| "step": 121500 | |
| }, | |
| { | |
| "epoch": 1.8919128479491354, | |
| "grad_norm": 7.3756103515625, | |
| "learning_rate": 1.8468119200847743e-05, | |
| "loss": 2.7751, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 1.899666589129255, | |
| "grad_norm": 7.422321796417236, | |
| "learning_rate": 1.8338890181179085e-05, | |
| "loss": 2.7766, | |
| "step": 122500 | |
| }, | |
| { | |
| "epoch": 1.9074203303093742, | |
| "grad_norm": 7.091059684753418, | |
| "learning_rate": 1.820966116151043e-05, | |
| "loss": 2.7643, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 1.9151740714894938, | |
| "grad_norm": 6.582401275634766, | |
| "learning_rate": 1.808043214184177e-05, | |
| "loss": 2.7665, | |
| "step": 123500 | |
| }, | |
| { | |
| "epoch": 1.922927812669613, | |
| "grad_norm": 7.574552536010742, | |
| "learning_rate": 1.7951203122173115e-05, | |
| "loss": 2.7548, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 1.9306815538497326, | |
| "grad_norm": 7.082491874694824, | |
| "learning_rate": 1.782197410250446e-05, | |
| "loss": 2.7577, | |
| "step": 124500 | |
| }, | |
| { | |
| "epoch": 1.9384352950298518, | |
| "grad_norm": 7.546943187713623, | |
| "learning_rate": 1.76927450828358e-05, | |
| "loss": 2.7709, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 1.9461890362099714, | |
| "grad_norm": 7.498143196105957, | |
| "learning_rate": 1.7563516063167146e-05, | |
| "loss": 2.7674, | |
| "step": 125500 | |
| }, | |
| { | |
| "epoch": 1.9539427773900906, | |
| "grad_norm": 7.182895660400391, | |
| "learning_rate": 1.743428704349849e-05, | |
| "loss": 2.7595, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 1.9616965185702102, | |
| "grad_norm": 7.754599094390869, | |
| "learning_rate": 1.7305058023829832e-05, | |
| "loss": 2.7586, | |
| "step": 126500 | |
| }, | |
| { | |
| "epoch": 1.9694502597503294, | |
| "grad_norm": 7.348043918609619, | |
| "learning_rate": 1.7175829004161177e-05, | |
| "loss": 2.7671, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 1.977204000930449, | |
| "grad_norm": 7.5025835037231445, | |
| "learning_rate": 1.7046599984492518e-05, | |
| "loss": 2.7596, | |
| "step": 127500 | |
| }, | |
| { | |
| "epoch": 1.9849577421105682, | |
| "grad_norm": 7.277556896209717, | |
| "learning_rate": 1.6917370964823863e-05, | |
| "loss": 2.7554, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 1.9927114832906878, | |
| "grad_norm": 7.15894079208374, | |
| "learning_rate": 1.6788141945155207e-05, | |
| "loss": 2.758, | |
| "step": 128500 | |
| }, | |
| { | |
| "epoch": 2.000465224470807, | |
| "grad_norm": 7.221950054168701, | |
| "learning_rate": 1.665891292548655e-05, | |
| "loss": 2.7564, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 2.0082189656509266, | |
| "grad_norm": 7.185346603393555, | |
| "learning_rate": 1.6529683905817893e-05, | |
| "loss": 2.7671, | |
| "step": 129500 | |
| }, | |
| { | |
| "epoch": 2.015972706831046, | |
| "grad_norm": 7.411344528198242, | |
| "learning_rate": 1.6400454886149234e-05, | |
| "loss": 2.7698, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 2.0237264480111654, | |
| "grad_norm": 7.3418498039245605, | |
| "learning_rate": 1.6271225866480576e-05, | |
| "loss": 2.7508, | |
| "step": 130500 | |
| }, | |
| { | |
| "epoch": 2.0314801891912846, | |
| "grad_norm": 7.749533176422119, | |
| "learning_rate": 1.614199684681192e-05, | |
| "loss": 2.7643, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 2.039233930371404, | |
| "grad_norm": 7.400169372558594, | |
| "learning_rate": 1.6012767827143262e-05, | |
| "loss": 2.7523, | |
| "step": 131500 | |
| }, | |
| { | |
| "epoch": 2.0469876715515234, | |
| "grad_norm": 6.999739646911621, | |
| "learning_rate": 1.5883538807474606e-05, | |
| "loss": 2.7525, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 2.054741412731643, | |
| "grad_norm": 7.423766613006592, | |
| "learning_rate": 1.575430978780595e-05, | |
| "loss": 2.7497, | |
| "step": 132500 | |
| }, | |
| { | |
| "epoch": 2.0624951539117626, | |
| "grad_norm": 7.121034622192383, | |
| "learning_rate": 1.5625080768137292e-05, | |
| "loss": 2.7644, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 2.070248895091882, | |
| "grad_norm": 7.697927951812744, | |
| "learning_rate": 1.5495851748468637e-05, | |
| "loss": 2.7599, | |
| "step": 133500 | |
| }, | |
| { | |
| "epoch": 2.0780026362720014, | |
| "grad_norm": 7.259957313537598, | |
| "learning_rate": 1.5366622728799982e-05, | |
| "loss": 2.7569, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 2.0857563774521206, | |
| "grad_norm": 7.2549943923950195, | |
| "learning_rate": 1.5237393709131323e-05, | |
| "loss": 2.764, | |
| "step": 134500 | |
| }, | |
| { | |
| "epoch": 2.0935101186322402, | |
| "grad_norm": 7.1079535484313965, | |
| "learning_rate": 1.5108164689462668e-05, | |
| "loss": 2.7463, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 2.1012638598123594, | |
| "grad_norm": 7.20269250869751, | |
| "learning_rate": 1.4978935669794009e-05, | |
| "loss": 2.7464, | |
| "step": 135500 | |
| }, | |
| { | |
| "epoch": 2.109017600992479, | |
| "grad_norm": 7.686685085296631, | |
| "learning_rate": 1.4849706650125352e-05, | |
| "loss": 2.7546, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 2.116771342172598, | |
| "grad_norm": 6.865842342376709, | |
| "learning_rate": 1.4720477630456697e-05, | |
| "loss": 2.743, | |
| "step": 136500 | |
| }, | |
| { | |
| "epoch": 2.124525083352718, | |
| "grad_norm": 6.892743110656738, | |
| "learning_rate": 1.4591248610788038e-05, | |
| "loss": 2.7597, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 2.132278824532837, | |
| "grad_norm": 7.216090679168701, | |
| "learning_rate": 1.4462019591119383e-05, | |
| "loss": 2.7638, | |
| "step": 137500 | |
| }, | |
| { | |
| "epoch": 2.1400325657129566, | |
| "grad_norm": 7.859537601470947, | |
| "learning_rate": 1.4332790571450724e-05, | |
| "loss": 2.7522, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 2.147786306893076, | |
| "grad_norm": 7.197884559631348, | |
| "learning_rate": 1.4203561551782069e-05, | |
| "loss": 2.7524, | |
| "step": 138500 | |
| }, | |
| { | |
| "epoch": 2.1555400480731954, | |
| "grad_norm": 7.401318550109863, | |
| "learning_rate": 1.4074332532113413e-05, | |
| "loss": 2.7629, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 2.1632937892533146, | |
| "grad_norm": 7.095146656036377, | |
| "learning_rate": 1.3945103512444755e-05, | |
| "loss": 2.7571, | |
| "step": 139500 | |
| }, | |
| { | |
| "epoch": 2.1710475304334342, | |
| "grad_norm": 7.200826168060303, | |
| "learning_rate": 1.38158744927761e-05, | |
| "loss": 2.7411, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 2.1788012716135534, | |
| "grad_norm": 7.727132797241211, | |
| "learning_rate": 1.3686645473107442e-05, | |
| "loss": 2.7518, | |
| "step": 140500 | |
| }, | |
| { | |
| "epoch": 2.186555012793673, | |
| "grad_norm": 7.624775409698486, | |
| "learning_rate": 1.3557416453438784e-05, | |
| "loss": 2.7461, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 2.194308753973792, | |
| "grad_norm": 7.7125935554504395, | |
| "learning_rate": 1.3428187433770128e-05, | |
| "loss": 2.7506, | |
| "step": 141500 | |
| }, | |
| { | |
| "epoch": 2.202062495153912, | |
| "grad_norm": 7.944740295410156, | |
| "learning_rate": 1.329895841410147e-05, | |
| "loss": 2.7435, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 2.209816236334031, | |
| "grad_norm": 7.168126106262207, | |
| "learning_rate": 1.3169729394432814e-05, | |
| "loss": 2.7577, | |
| "step": 142500 | |
| }, | |
| { | |
| "epoch": 2.2175699775141506, | |
| "grad_norm": 7.608551979064941, | |
| "learning_rate": 1.3040500374764159e-05, | |
| "loss": 2.7514, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 2.22532371869427, | |
| "grad_norm": 7.155666351318359, | |
| "learning_rate": 1.29112713550955e-05, | |
| "loss": 2.756, | |
| "step": 143500 | |
| }, | |
| { | |
| "epoch": 2.2330774598743894, | |
| "grad_norm": 7.49126672744751, | |
| "learning_rate": 1.2782042335426845e-05, | |
| "loss": 2.7452, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 2.2408312010545086, | |
| "grad_norm": 7.515799522399902, | |
| "learning_rate": 1.2652813315758188e-05, | |
| "loss": 2.7485, | |
| "step": 144500 | |
| }, | |
| { | |
| "epoch": 2.2485849422346282, | |
| "grad_norm": 7.652871608734131, | |
| "learning_rate": 1.252358429608953e-05, | |
| "loss": 2.7519, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 2.2563386834147474, | |
| "grad_norm": 7.006313800811768, | |
| "learning_rate": 1.2394355276420874e-05, | |
| "loss": 2.7401, | |
| "step": 145500 | |
| }, | |
| { | |
| "epoch": 2.264092424594867, | |
| "grad_norm": 7.337978839874268, | |
| "learning_rate": 1.2265126256752217e-05, | |
| "loss": 2.7458, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 2.2718461657749867, | |
| "grad_norm": 7.185283184051514, | |
| "learning_rate": 1.213589723708356e-05, | |
| "loss": 2.7546, | |
| "step": 146500 | |
| }, | |
| { | |
| "epoch": 2.279599906955106, | |
| "grad_norm": 7.885451316833496, | |
| "learning_rate": 1.2006668217414903e-05, | |
| "loss": 2.738, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 2.287353648135225, | |
| "grad_norm": 7.163339138031006, | |
| "learning_rate": 1.1877439197746248e-05, | |
| "loss": 2.7403, | |
| "step": 147500 | |
| }, | |
| { | |
| "epoch": 2.2951073893153446, | |
| "grad_norm": 7.566407680511475, | |
| "learning_rate": 1.174821017807759e-05, | |
| "loss": 2.7441, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 2.3028611304954643, | |
| "grad_norm": 7.626791477203369, | |
| "learning_rate": 1.1618981158408934e-05, | |
| "loss": 2.7442, | |
| "step": 148500 | |
| }, | |
| { | |
| "epoch": 2.3106148716755834, | |
| "grad_norm": 7.609415054321289, | |
| "learning_rate": 1.1489752138740275e-05, | |
| "loss": 2.7442, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 2.318368612855703, | |
| "grad_norm": 7.549880504608154, | |
| "learning_rate": 1.136052311907162e-05, | |
| "loss": 2.7398, | |
| "step": 149500 | |
| }, | |
| { | |
| "epoch": 2.3261223540358222, | |
| "grad_norm": 7.753575325012207, | |
| "learning_rate": 1.1231294099402963e-05, | |
| "loss": 2.7376, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 2.333876095215942, | |
| "grad_norm": 7.579866886138916, | |
| "learning_rate": 1.1102065079734306e-05, | |
| "loss": 2.7449, | |
| "step": 150500 | |
| }, | |
| { | |
| "epoch": 2.341629836396061, | |
| "grad_norm": 7.787561893463135, | |
| "learning_rate": 1.0972836060065649e-05, | |
| "loss": 2.7418, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 2.3493835775761807, | |
| "grad_norm": 7.163692474365234, | |
| "learning_rate": 1.0843607040396992e-05, | |
| "loss": 2.7459, | |
| "step": 151500 | |
| }, | |
| { | |
| "epoch": 2.3571373187563, | |
| "grad_norm": 8.124524116516113, | |
| "learning_rate": 1.0714378020728336e-05, | |
| "loss": 2.7472, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 2.3648910599364195, | |
| "grad_norm": 7.68442964553833, | |
| "learning_rate": 1.058514900105968e-05, | |
| "loss": 2.7454, | |
| "step": 152500 | |
| }, | |
| { | |
| "epoch": 2.3726448011165386, | |
| "grad_norm": 7.561328887939453, | |
| "learning_rate": 1.045591998139102e-05, | |
| "loss": 2.7314, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 2.3803985422966583, | |
| "grad_norm": 7.480719566345215, | |
| "learning_rate": 1.0326690961722364e-05, | |
| "loss": 2.7385, | |
| "step": 153500 | |
| }, | |
| { | |
| "epoch": 2.3881522834767774, | |
| "grad_norm": 7.676718235015869, | |
| "learning_rate": 1.0197461942053708e-05, | |
| "loss": 2.7363, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 2.395906024656897, | |
| "grad_norm": 7.30204963684082, | |
| "learning_rate": 1.0068232922385051e-05, | |
| "loss": 2.7469, | |
| "step": 154500 | |
| }, | |
| { | |
| "epoch": 2.4036597658370162, | |
| "grad_norm": 7.684243202209473, | |
| "learning_rate": 9.939003902716394e-06, | |
| "loss": 2.7404, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 2.411413507017136, | |
| "grad_norm": 7.187122821807861, | |
| "learning_rate": 9.809774883047737e-06, | |
| "loss": 2.7364, | |
| "step": 155500 | |
| }, | |
| { | |
| "epoch": 2.419167248197255, | |
| "grad_norm": 7.586068153381348, | |
| "learning_rate": 9.680545863379082e-06, | |
| "loss": 2.7331, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 2.4269209893773747, | |
| "grad_norm": 7.374856948852539, | |
| "learning_rate": 9.551316843710425e-06, | |
| "loss": 2.739, | |
| "step": 156500 | |
| }, | |
| { | |
| "epoch": 2.434674730557494, | |
| "grad_norm": 7.3092474937438965, | |
| "learning_rate": 9.422087824041768e-06, | |
| "loss": 2.7412, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 2.4424284717376135, | |
| "grad_norm": 7.9172844886779785, | |
| "learning_rate": 9.29285880437311e-06, | |
| "loss": 2.7384, | |
| "step": 157500 | |
| }, | |
| { | |
| "epoch": 2.4501822129177326, | |
| "grad_norm": 7.155998706817627, | |
| "learning_rate": 9.163629784704454e-06, | |
| "loss": 2.7324, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 2.4579359540978523, | |
| "grad_norm": 7.371484756469727, | |
| "learning_rate": 9.034400765035797e-06, | |
| "loss": 2.7382, | |
| "step": 158500 | |
| }, | |
| { | |
| "epoch": 2.4656896952779714, | |
| "grad_norm": 7.271082401275635, | |
| "learning_rate": 8.90517174536714e-06, | |
| "loss": 2.7238, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 2.473443436458091, | |
| "grad_norm": 7.525820255279541, | |
| "learning_rate": 8.775942725698483e-06, | |
| "loss": 2.7353, | |
| "step": 159500 | |
| }, | |
| { | |
| "epoch": 2.4811971776382107, | |
| "grad_norm": 7.422860622406006, | |
| "learning_rate": 8.646713706029828e-06, | |
| "loss": 2.7312, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 2.48895091881833, | |
| "grad_norm": 7.786092758178711, | |
| "learning_rate": 8.51748468636117e-06, | |
| "loss": 2.729, | |
| "step": 160500 | |
| }, | |
| { | |
| "epoch": 2.496704659998449, | |
| "grad_norm": 7.733543872833252, | |
| "learning_rate": 8.388255666692514e-06, | |
| "loss": 2.7377, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 2.5044584011785687, | |
| "grad_norm": 7.477449417114258, | |
| "learning_rate": 8.259026647023855e-06, | |
| "loss": 2.7431, | |
| "step": 161500 | |
| }, | |
| { | |
| "epoch": 2.5122121423586883, | |
| "grad_norm": 7.466070652008057, | |
| "learning_rate": 8.1297976273552e-06, | |
| "loss": 2.743, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 2.5199658835388075, | |
| "grad_norm": 7.578529357910156, | |
| "learning_rate": 8.000568607686542e-06, | |
| "loss": 2.738, | |
| "step": 162500 | |
| }, | |
| { | |
| "epoch": 2.5277196247189266, | |
| "grad_norm": 7.481320381164551, | |
| "learning_rate": 7.871339588017885e-06, | |
| "loss": 2.7332, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 2.5354733658990463, | |
| "grad_norm": 8.073503494262695, | |
| "learning_rate": 7.742110568349228e-06, | |
| "loss": 2.7517, | |
| "step": 163500 | |
| }, | |
| { | |
| "epoch": 2.543227107079166, | |
| "grad_norm": 7.4196457862854, | |
| "learning_rate": 7.612881548680572e-06, | |
| "loss": 2.7231, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 2.550980848259285, | |
| "grad_norm": 7.558558940887451, | |
| "learning_rate": 7.483652529011915e-06, | |
| "loss": 2.7384, | |
| "step": 164500 | |
| }, | |
| { | |
| "epoch": 2.5587345894394042, | |
| "grad_norm": 7.38846492767334, | |
| "learning_rate": 7.354423509343258e-06, | |
| "loss": 2.7348, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 2.566488330619524, | |
| "grad_norm": 7.8365864753723145, | |
| "learning_rate": 7.225194489674601e-06, | |
| "loss": 2.7254, | |
| "step": 165500 | |
| }, | |
| { | |
| "epoch": 2.5742420717996435, | |
| "grad_norm": 7.362669944763184, | |
| "learning_rate": 7.095965470005945e-06, | |
| "loss": 2.729, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 2.5819958129797627, | |
| "grad_norm": 7.646996974945068, | |
| "learning_rate": 6.966736450337288e-06, | |
| "loss": 2.7333, | |
| "step": 166500 | |
| }, | |
| { | |
| "epoch": 2.5897495541598823, | |
| "grad_norm": 7.944218158721924, | |
| "learning_rate": 6.837507430668631e-06, | |
| "loss": 2.7423, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 2.5975032953400015, | |
| "grad_norm": 7.502200603485107, | |
| "learning_rate": 6.708278410999974e-06, | |
| "loss": 2.7225, | |
| "step": 167500 | |
| }, | |
| { | |
| "epoch": 2.605257036520121, | |
| "grad_norm": 7.175666809082031, | |
| "learning_rate": 6.579049391331319e-06, | |
| "loss": 2.7366, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 2.6130107777002403, | |
| "grad_norm": 7.814846992492676, | |
| "learning_rate": 6.449820371662661e-06, | |
| "loss": 2.7421, | |
| "step": 168500 | |
| }, | |
| { | |
| "epoch": 2.62076451888036, | |
| "grad_norm": 7.270232677459717, | |
| "learning_rate": 6.320591351994004e-06, | |
| "loss": 2.7335, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 2.628518260060479, | |
| "grad_norm": 7.920383930206299, | |
| "learning_rate": 6.191362332325347e-06, | |
| "loss": 2.7352, | |
| "step": 169500 | |
| }, | |
| { | |
| "epoch": 2.6362720012405987, | |
| "grad_norm": 7.142765998840332, | |
| "learning_rate": 6.062133312656691e-06, | |
| "loss": 2.7318, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 2.644025742420718, | |
| "grad_norm": 8.12151050567627, | |
| "learning_rate": 5.932904292988034e-06, | |
| "loss": 2.7168, | |
| "step": 170500 | |
| }, | |
| { | |
| "epoch": 2.6517794836008375, | |
| "grad_norm": 7.717370510101318, | |
| "learning_rate": 5.803675273319377e-06, | |
| "loss": 2.7145, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 2.6595332247809567, | |
| "grad_norm": 7.359320640563965, | |
| "learning_rate": 5.67444625365072e-06, | |
| "loss": 2.7382, | |
| "step": 171500 | |
| }, | |
| { | |
| "epoch": 2.6672869659610763, | |
| "grad_norm": 7.525691509246826, | |
| "learning_rate": 5.5452172339820636e-06, | |
| "loss": 2.7228, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 2.6750407071411955, | |
| "grad_norm": 7.967082500457764, | |
| "learning_rate": 5.4159882143134066e-06, | |
| "loss": 2.7258, | |
| "step": 172500 | |
| }, | |
| { | |
| "epoch": 2.682794448321315, | |
| "grad_norm": 7.760034561157227, | |
| "learning_rate": 5.2867591946447495e-06, | |
| "loss": 2.7385, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 2.6905481895014347, | |
| "grad_norm": 7.141742706298828, | |
| "learning_rate": 5.1575301749760925e-06, | |
| "loss": 2.7282, | |
| "step": 173500 | |
| }, | |
| { | |
| "epoch": 2.698301930681554, | |
| "grad_norm": 7.685527801513672, | |
| "learning_rate": 5.028301155307436e-06, | |
| "loss": 2.7216, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 2.706055671861673, | |
| "grad_norm": 7.3134236335754395, | |
| "learning_rate": 4.899072135638779e-06, | |
| "loss": 2.7188, | |
| "step": 174500 | |
| }, | |
| { | |
| "epoch": 2.7138094130417927, | |
| "grad_norm": 7.750508785247803, | |
| "learning_rate": 4.769843115970122e-06, | |
| "loss": 2.7277, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 2.7215631542219123, | |
| "grad_norm": 7.504671096801758, | |
| "learning_rate": 4.640614096301465e-06, | |
| "loss": 2.738, | |
| "step": 175500 | |
| }, | |
| { | |
| "epoch": 2.7293168954020315, | |
| "grad_norm": 7.484751224517822, | |
| "learning_rate": 4.511385076632809e-06, | |
| "loss": 2.7148, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 2.7370706365821507, | |
| "grad_norm": 7.809044361114502, | |
| "learning_rate": 4.382156056964152e-06, | |
| "loss": 2.7388, | |
| "step": 176500 | |
| }, | |
| { | |
| "epoch": 2.7448243777622703, | |
| "grad_norm": 7.876001834869385, | |
| "learning_rate": 4.252927037295495e-06, | |
| "loss": 2.7214, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 2.75257811894239, | |
| "grad_norm": 7.753846645355225, | |
| "learning_rate": 4.123698017626838e-06, | |
| "loss": 2.719, | |
| "step": 177500 | |
| }, | |
| { | |
| "epoch": 2.760331860122509, | |
| "grad_norm": 7.285833358764648, | |
| "learning_rate": 3.994468997958182e-06, | |
| "loss": 2.7347, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 2.7680856013026283, | |
| "grad_norm": 7.894680023193359, | |
| "learning_rate": 3.865239978289525e-06, | |
| "loss": 2.7376, | |
| "step": 178500 | |
| }, | |
| { | |
| "epoch": 2.775839342482748, | |
| "grad_norm": 7.850667953491211, | |
| "learning_rate": 3.7360109586208684e-06, | |
| "loss": 2.7289, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 2.7835930836628675, | |
| "grad_norm": 7.380823135375977, | |
| "learning_rate": 3.606781938952211e-06, | |
| "loss": 2.7289, | |
| "step": 179500 | |
| }, | |
| { | |
| "epoch": 2.7913468248429867, | |
| "grad_norm": 7.752573490142822, | |
| "learning_rate": 3.477552919283555e-06, | |
| "loss": 2.724, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 2.7991005660231063, | |
| "grad_norm": 7.117413520812988, | |
| "learning_rate": 3.3483238996148974e-06, | |
| "loss": 2.7332, | |
| "step": 180500 | |
| }, | |
| { | |
| "epoch": 2.8068543072032255, | |
| "grad_norm": 7.8615522384643555, | |
| "learning_rate": 3.2190948799462412e-06, | |
| "loss": 2.7316, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 2.814608048383345, | |
| "grad_norm": 7.938878059387207, | |
| "learning_rate": 3.0898658602775842e-06, | |
| "loss": 2.7223, | |
| "step": 181500 | |
| }, | |
| { | |
| "epoch": 2.8223617895634643, | |
| "grad_norm": 7.760583877563477, | |
| "learning_rate": 2.9606368406089272e-06, | |
| "loss": 2.7222, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 2.830115530743584, | |
| "grad_norm": 7.352213382720947, | |
| "learning_rate": 2.8314078209402706e-06, | |
| "loss": 2.7233, | |
| "step": 182500 | |
| }, | |
| { | |
| "epoch": 2.837869271923703, | |
| "grad_norm": 7.541159629821777, | |
| "learning_rate": 2.7021788012716136e-06, | |
| "loss": 2.7225, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 2.8456230131038227, | |
| "grad_norm": 7.890182018280029, | |
| "learning_rate": 2.572949781602957e-06, | |
| "loss": 2.7219, | |
| "step": 183500 | |
| }, | |
| { | |
| "epoch": 2.853376754283942, | |
| "grad_norm": 7.695311546325684, | |
| "learning_rate": 2.4437207619343e-06, | |
| "loss": 2.7172, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 2.8611304954640615, | |
| "grad_norm": 7.7702317237854, | |
| "learning_rate": 2.3144917422656434e-06, | |
| "loss": 2.728, | |
| "step": 184500 | |
| }, | |
| { | |
| "epoch": 2.8688842366441807, | |
| "grad_norm": 7.646172046661377, | |
| "learning_rate": 2.1852627225969864e-06, | |
| "loss": 2.7312, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 2.8766379778243003, | |
| "grad_norm": 7.06711483001709, | |
| "learning_rate": 2.05603370292833e-06, | |
| "loss": 2.7175, | |
| "step": 185500 | |
| }, | |
| { | |
| "epoch": 2.8843917190044195, | |
| "grad_norm": 7.974971294403076, | |
| "learning_rate": 1.926804683259673e-06, | |
| "loss": 2.7244, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 2.892145460184539, | |
| "grad_norm": 7.5829315185546875, | |
| "learning_rate": 1.797575663591016e-06, | |
| "loss": 2.7298, | |
| "step": 186500 | |
| }, | |
| { | |
| "epoch": 2.8998992013646583, | |
| "grad_norm": 7.224939823150635, | |
| "learning_rate": 1.6683466439223592e-06, | |
| "loss": 2.7227, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 2.907652942544778, | |
| "grad_norm": 8.057891845703125, | |
| "learning_rate": 1.5391176242537025e-06, | |
| "loss": 2.7393, | |
| "step": 187500 | |
| }, | |
| { | |
| "epoch": 2.915406683724897, | |
| "grad_norm": 7.886134624481201, | |
| "learning_rate": 1.4098886045850457e-06, | |
| "loss": 2.7264, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 2.9231604249050167, | |
| "grad_norm": 7.65654993057251, | |
| "learning_rate": 1.2806595849163889e-06, | |
| "loss": 2.7277, | |
| "step": 188500 | |
| }, | |
| { | |
| "epoch": 2.9309141660851363, | |
| "grad_norm": 7.524332046508789, | |
| "learning_rate": 1.1514305652477323e-06, | |
| "loss": 2.7281, | |
| "step": 189000 | |
| }, | |
| { | |
| "epoch": 2.9386679072652555, | |
| "grad_norm": 7.878385543823242, | |
| "learning_rate": 1.0222015455790753e-06, | |
| "loss": 2.7275, | |
| "step": 189500 | |
| }, | |
| { | |
| "epoch": 2.9464216484453747, | |
| "grad_norm": 7.491950035095215, | |
| "learning_rate": 8.929725259104185e-07, | |
| "loss": 2.7225, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 2.9541753896254943, | |
| "grad_norm": 8.570446968078613, | |
| "learning_rate": 7.637435062417617e-07, | |
| "loss": 2.722, | |
| "step": 190500 | |
| }, | |
| { | |
| "epoch": 2.961929130805614, | |
| "grad_norm": 7.909883975982666, | |
| "learning_rate": 6.345144865731049e-07, | |
| "loss": 2.7175, | |
| "step": 191000 | |
| }, | |
| { | |
| "epoch": 2.969682871985733, | |
| "grad_norm": 7.273110389709473, | |
| "learning_rate": 5.052854669044481e-07, | |
| "loss": 2.7167, | |
| "step": 191500 | |
| }, | |
| { | |
| "epoch": 2.9774366131658523, | |
| "grad_norm": 7.78535270690918, | |
| "learning_rate": 3.760564472357913e-07, | |
| "loss": 2.7301, | |
| "step": 192000 | |
| }, | |
| { | |
| "epoch": 2.985190354345972, | |
| "grad_norm": 7.7455973625183105, | |
| "learning_rate": 2.468274275671345e-07, | |
| "loss": 2.7192, | |
| "step": 192500 | |
| }, | |
| { | |
| "epoch": 2.9929440955260915, | |
| "grad_norm": 7.988417148590088, | |
| "learning_rate": 1.1759840789847768e-07, | |
| "loss": 2.7195, | |
| "step": 193000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 193455, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.6392157746049843e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |