| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.9904, | |
| "eval_steps": 500, | |
| "global_step": 1248, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 2.3852293863398297, | |
| "learning_rate": 2.631578947368421e-06, | |
| "loss": 3.0228, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 2.278305395121773, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 2.8948, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 2.432053456282922, | |
| "learning_rate": 7.894736842105263e-06, | |
| "loss": 2.9685, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.8462406525449728, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 2.8781, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.6679144296856878, | |
| "learning_rate": 1.3157894736842106e-05, | |
| "loss": 2.8323, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.5352730719636578, | |
| "learning_rate": 1.5789473684210526e-05, | |
| "loss": 2.8891, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.3656005018424684, | |
| "learning_rate": 1.8421052631578947e-05, | |
| "loss": 2.8461, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.1711736437076634, | |
| "learning_rate": 2.105263157894737e-05, | |
| "loss": 2.6232, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9532059996096075, | |
| "learning_rate": 2.368421052631579e-05, | |
| "loss": 2.7889, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9200056915508535, | |
| "learning_rate": 2.6315789473684212e-05, | |
| "loss": 2.6584, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.9892570713384861, | |
| "learning_rate": 2.8947368421052634e-05, | |
| "loss": 2.6855, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.1737032379418306, | |
| "learning_rate": 3.157894736842105e-05, | |
| "loss": 2.6815, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.1396207804258625, | |
| "learning_rate": 3.421052631578947e-05, | |
| "loss": 2.7002, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.026878595312307, | |
| "learning_rate": 3.6842105263157895e-05, | |
| "loss": 2.6317, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.0270589281505744, | |
| "learning_rate": 3.9473684210526316e-05, | |
| "loss": 2.7518, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.9332530658530782, | |
| "learning_rate": 4.210526315789474e-05, | |
| "loss": 2.6212, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8838859358375801, | |
| "learning_rate": 4.473684210526316e-05, | |
| "loss": 2.5728, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.9917284885842201, | |
| "learning_rate": 4.736842105263158e-05, | |
| "loss": 2.4781, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.8122674298840572, | |
| "learning_rate": 5e-05, | |
| "loss": 2.4781, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.9927284359317292, | |
| "learning_rate": 5.2631578947368424e-05, | |
| "loss": 2.5996, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.971876774249461, | |
| "learning_rate": 5.526315789473685e-05, | |
| "loss": 2.508, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9294725813075919, | |
| "learning_rate": 5.789473684210527e-05, | |
| "loss": 2.5437, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.9214762463109677, | |
| "learning_rate": 6.052631578947369e-05, | |
| "loss": 2.589, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8888801870082056, | |
| "learning_rate": 6.31578947368421e-05, | |
| "loss": 2.4147, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8402381742845089, | |
| "learning_rate": 6.578947368421054e-05, | |
| "loss": 2.4834, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.8781982415413805, | |
| "learning_rate": 6.842105263157895e-05, | |
| "loss": 2.5108, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9356297956079316, | |
| "learning_rate": 7.105263157894737e-05, | |
| "loss": 2.6323, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9127856670393594, | |
| "learning_rate": 7.368421052631579e-05, | |
| "loss": 2.5534, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8783502121089134, | |
| "learning_rate": 7.631578947368422e-05, | |
| "loss": 2.5085, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.9162872435067455, | |
| "learning_rate": 7.894736842105263e-05, | |
| "loss": 2.5263, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.863753866582432, | |
| "learning_rate": 8.157894736842105e-05, | |
| "loss": 2.5423, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.964988290243888, | |
| "learning_rate": 8.421052631578948e-05, | |
| "loss": 2.5148, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.0089358136228708, | |
| "learning_rate": 8.68421052631579e-05, | |
| "loss": 2.3565, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.983875971830578, | |
| "learning_rate": 8.947368421052632e-05, | |
| "loss": 2.3883, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.0336533271496344, | |
| "learning_rate": 9.210526315789474e-05, | |
| "loss": 2.5371, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.09953333695354, | |
| "learning_rate": 9.473684210526316e-05, | |
| "loss": 2.5711, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.9814647100150796, | |
| "learning_rate": 9.736842105263158e-05, | |
| "loss": 2.437, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.065696585050419, | |
| "learning_rate": 0.0001, | |
| "loss": 2.5983, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.9764593606764957, | |
| "learning_rate": 9.999983147327955e-05, | |
| "loss": 2.4674, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.0220198340776872, | |
| "learning_rate": 9.999932589425423e-05, | |
| "loss": 2.4248, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.9402124234062932, | |
| "learning_rate": 9.999848326633219e-05, | |
| "loss": 2.4359, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.9533672274828753, | |
| "learning_rate": 9.999730359519366e-05, | |
| "loss": 2.4616, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.9164865299764797, | |
| "learning_rate": 9.999578688879086e-05, | |
| "loss": 2.4265, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.8978585337852701, | |
| "learning_rate": 9.999393315734801e-05, | |
| "loss": 2.3858, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.886248191859053, | |
| "learning_rate": 9.999174241336126e-05, | |
| "loss": 2.441, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.9253068801523042, | |
| "learning_rate": 9.998921467159855e-05, | |
| "loss": 2.3882, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.9279981761271726, | |
| "learning_rate": 9.998634994909958e-05, | |
| "loss": 2.4059, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.9772412791266225, | |
| "learning_rate": 9.998314826517563e-05, | |
| "loss": 2.4608, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.0028380201307552, | |
| "learning_rate": 9.997960964140947e-05, | |
| "loss": 2.4477, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.9715337426005088, | |
| "learning_rate": 9.99757341016552e-05, | |
| "loss": 2.3818, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.9363938098973695, | |
| "learning_rate": 9.997152167203811e-05, | |
| "loss": 2.3897, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.9035155554465317, | |
| "learning_rate": 9.99669723809545e-05, | |
| "loss": 2.3618, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.9117521223203663, | |
| "learning_rate": 9.996208625907141e-05, | |
| "loss": 2.354, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.9593841316816577, | |
| "learning_rate": 9.995686333932655e-05, | |
| "loss": 2.3347, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.9383421158383985, | |
| "learning_rate": 9.995130365692798e-05, | |
| "loss": 2.3991, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.9671406117527945, | |
| "learning_rate": 9.994540724935389e-05, | |
| "loss": 2.4494, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.9594395035175591, | |
| "learning_rate": 9.993917415635238e-05, | |
| "loss": 2.3753, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.8794895418385533, | |
| "learning_rate": 9.993260441994116e-05, | |
| "loss": 2.3254, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.9408127413040817, | |
| "learning_rate": 9.992569808440726e-05, | |
| "loss": 2.3317, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.9719890580562476, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 2.3178, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.9995603022158662, | |
| "learning_rate": 9.991087580446454e-05, | |
| "loss": 2.5114, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.0351382467081842, | |
| "learning_rate": 9.99029599599737e-05, | |
| "loss": 2.3309, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.9178979552142348, | |
| "learning_rate": 9.989470771619555e-05, | |
| "loss": 2.325, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.104498989767433, | |
| "learning_rate": 9.988611912875901e-05, | |
| "loss": 2.4103, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.9631085684946267, | |
| "learning_rate": 9.987719425556035e-05, | |
| "loss": 2.3055, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.3883014369025373, | |
| "learning_rate": 9.986793315676276e-05, | |
| "loss": 2.3918, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.9422792446755142, | |
| "learning_rate": 9.985833589479592e-05, | |
| "loss": 2.3102, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.22737962239379, | |
| "learning_rate": 9.984840253435568e-05, | |
| "loss": 2.4137, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.0607449730695824, | |
| "learning_rate": 9.983813314240345e-05, | |
| "loss": 2.417, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.1835115210845712, | |
| "learning_rate": 9.982752778816595e-05, | |
| "loss": 2.3438, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.0152973368545788, | |
| "learning_rate": 9.981658654313457e-05, | |
| "loss": 2.3453, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.0057378936408388, | |
| "learning_rate": 9.980530948106504e-05, | |
| "loss": 2.3467, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.018565392623426, | |
| "learning_rate": 9.979369667797675e-05, | |
| "loss": 2.3852, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.9367971990132274, | |
| "learning_rate": 9.978174821215247e-05, | |
| "loss": 2.2556, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.909806110488264, | |
| "learning_rate": 9.976946416413759e-05, | |
| "loss": 2.2677, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.0095582102908085, | |
| "learning_rate": 9.975684461673972e-05, | |
| "loss": 2.3358, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.0078815064331534, | |
| "learning_rate": 9.974388965502812e-05, | |
| "loss": 2.3772, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.9768517574758191, | |
| "learning_rate": 9.973059936633306e-05, | |
| "loss": 2.3693, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.010216250596747, | |
| "learning_rate": 9.971697384024532e-05, | |
| "loss": 2.2971, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.949870069819407, | |
| "learning_rate": 9.970301316861548e-05, | |
| "loss": 2.3445, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.0178203545305047, | |
| "learning_rate": 9.968871744555339e-05, | |
| "loss": 2.4457, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.0236669812619412, | |
| "learning_rate": 9.967408676742751e-05, | |
| "loss": 2.4371, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.9561722299721087, | |
| "learning_rate": 9.965912123286424e-05, | |
| "loss": 2.4237, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.9833242329352696, | |
| "learning_rate": 9.964382094274732e-05, | |
| "loss": 2.3846, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.1594025069167722, | |
| "learning_rate": 9.9628186000217e-05, | |
| "loss": 2.3911, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.9964157788231106, | |
| "learning_rate": 9.961221651066952e-05, | |
| "loss": 2.3597, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.0566672057925608, | |
| "learning_rate": 9.959591258175634e-05, | |
| "loss": 2.1981, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.0121551703226008, | |
| "learning_rate": 9.957927432338332e-05, | |
| "loss": 2.2248, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.2173775059369842, | |
| "learning_rate": 9.956230184771014e-05, | |
| "loss": 2.3425, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.058302563110291, | |
| "learning_rate": 9.954499526914941e-05, | |
| "loss": 2.3719, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.3810829968676661, | |
| "learning_rate": 9.952735470436597e-05, | |
| "loss": 2.4029, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.4489882109876326, | |
| "learning_rate": 9.950938027227608e-05, | |
| "loss": 2.3809, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.9650873666997658, | |
| "learning_rate": 9.949107209404665e-05, | |
| "loss": 2.2972, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.2515068910613174, | |
| "learning_rate": 9.947243029309433e-05, | |
| "loss": 2.2471, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.0214373685081224, | |
| "learning_rate": 9.945345499508478e-05, | |
| "loss": 2.2677, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.0826134822343876, | |
| "learning_rate": 9.943414632793184e-05, | |
| "loss": 2.3239, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.090816532297076, | |
| "learning_rate": 9.941450442179652e-05, | |
| "loss": 2.316, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.416272570679686, | |
| "learning_rate": 9.939452940908626e-05, | |
| "loss": 2.2249, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.9714992591271596, | |
| "learning_rate": 9.937422142445402e-05, | |
| "loss": 2.2731, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.312846923158138, | |
| "learning_rate": 9.935358060479731e-05, | |
| "loss": 2.4491, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.6944863862875934, | |
| "learning_rate": 9.933260708925731e-05, | |
| "loss": 2.2938, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.6393464723835427, | |
| "learning_rate": 9.931130101921795e-05, | |
| "loss": 2.292, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.3207704736826629, | |
| "learning_rate": 9.92896625383049e-05, | |
| "loss": 2.2317, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.9685616054346899, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 2.2072, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.3607156390739419, | |
| "learning_rate": 9.924538892956353e-05, | |
| "loss": 2.3215, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.2862667459615003, | |
| "learning_rate": 9.922275410018663e-05, | |
| "loss": 2.3466, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.0158906591091563, | |
| "learning_rate": 9.919978745683692e-05, | |
| "loss": 2.3257, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.3240666951465665, | |
| "learning_rate": 9.917648915433413e-05, | |
| "loss": 2.3781, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.236043665028954, | |
| "learning_rate": 9.915285934973371e-05, | |
| "loss": 2.3234, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.9542864317914936, | |
| "learning_rate": 9.912889820232578e-05, | |
| "loss": 2.1875, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 1.091723622672655, | |
| "learning_rate": 9.910460587363413e-05, | |
| "loss": 2.3579, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.2239891390789717, | |
| "learning_rate": 9.907998252741498e-05, | |
| "loss": 2.2612, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.079092118412001, | |
| "learning_rate": 9.905502832965602e-05, | |
| "loss": 2.2921, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.0974948746629458, | |
| "learning_rate": 9.902974344857521e-05, | |
| "loss": 2.3771, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.1589894787549435, | |
| "learning_rate": 9.900412805461967e-05, | |
| "loss": 2.2556, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.2256066240444334, | |
| "learning_rate": 9.897818232046454e-05, | |
| "loss": 2.4017, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.9928621462327584, | |
| "learning_rate": 9.89519064210118e-05, | |
| "loss": 2.2751, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.163183522674101, | |
| "learning_rate": 9.892530053338909e-05, | |
| "loss": 2.2977, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.146118395476374, | |
| "learning_rate": 9.889836483694852e-05, | |
| "loss": 2.2255, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.0292517144197497, | |
| "learning_rate": 9.88710995132655e-05, | |
| "loss": 2.2453, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.3845070195619045, | |
| "learning_rate": 9.884350474613741e-05, | |
| "loss": 2.3061, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.1160681368015506, | |
| "learning_rate": 9.881558072158252e-05, | |
| "loss": 2.187, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.22842890802036, | |
| "learning_rate": 9.87873276278386e-05, | |
| "loss": 2.2481, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.2341062203837596, | |
| "learning_rate": 9.875874565536167e-05, | |
| "loss": 2.2139, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.080638552567379, | |
| "learning_rate": 9.872983499682477e-05, | |
| "loss": 2.221, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.123485828237814, | |
| "learning_rate": 9.870059584711668e-05, | |
| "loss": 2.1541, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 1.0338118032477248, | |
| "learning_rate": 9.867102840334049e-05, | |
| "loss": 2.2662, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 1.0143481230927869, | |
| "learning_rate": 9.864113286481237e-05, | |
| "loss": 2.1758, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.043490896589895, | |
| "learning_rate": 9.861090943306021e-05, | |
| "loss": 2.1783, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.0634106872172329, | |
| "learning_rate": 9.858035831182226e-05, | |
| "loss": 2.2788, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 1.004859863673821, | |
| "learning_rate": 9.85494797070457e-05, | |
| "loss": 2.1507, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 1.0068717432212184, | |
| "learning_rate": 9.851827382688535e-05, | |
| "loss": 2.1571, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.0193352330073893, | |
| "learning_rate": 9.84867408817022e-05, | |
| "loss": 2.205, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.118517270848358, | |
| "learning_rate": 9.845488108406198e-05, | |
| "loss": 2.136, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.0679863001942589, | |
| "learning_rate": 9.84226946487338e-05, | |
| "loss": 2.2706, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.0428030556858152, | |
| "learning_rate": 9.839018179268862e-05, | |
| "loss": 2.1824, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.0523700663041782, | |
| "learning_rate": 9.835734273509786e-05, | |
| "loss": 2.1046, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.2267686120994454, | |
| "learning_rate": 9.832417769733185e-05, | |
| "loss": 2.2311, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 1.2168571502002221, | |
| "learning_rate": 9.829068690295839e-05, | |
| "loss": 2.1478, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.9837056302455408, | |
| "learning_rate": 9.825687057774126e-05, | |
| "loss": 2.1607, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.3862316882794774, | |
| "learning_rate": 9.82227289496386e-05, | |
| "loss": 2.282, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.2582790056629096, | |
| "learning_rate": 9.81882622488015e-05, | |
| "loss": 2.2172, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.070299873215969, | |
| "learning_rate": 9.815347070757232e-05, | |
| "loss": 2.1387, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.1673976883150066, | |
| "learning_rate": 9.811835456048328e-05, | |
| "loss": 2.2499, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.4169709644935056, | |
| "learning_rate": 9.808291404425473e-05, | |
| "loss": 2.2064, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.0682092504810614, | |
| "learning_rate": 9.804714939779362e-05, | |
| "loss": 2.0947, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1.955941498041186, | |
| "learning_rate": 9.80110608621919e-05, | |
| "loss": 2.1462, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1.1944463639093175, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 2.0653, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.915930420457324, | |
| "learning_rate": 9.793791309884956e-05, | |
| "loss": 2.1094, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.257430446986982, | |
| "learning_rate": 9.790085436420304e-05, | |
| "loss": 2.2623, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.6956621554088878, | |
| "learning_rate": 9.786347272660082e-05, | |
| "loss": 2.1837, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.1640548281872594, | |
| "learning_rate": 9.782576843803504e-05, | |
| "loss": 2.1858, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.2085838179846153, | |
| "learning_rate": 9.778774175267295e-05, | |
| "loss": 2.0926, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.392864302454965, | |
| "learning_rate": 9.774939292685504e-05, | |
| "loss": 2.1463, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.107619930692779, | |
| "learning_rate": 9.771072221909337e-05, | |
| "loss": 2.144, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.1363140828142162, | |
| "learning_rate": 9.767172989006985e-05, | |
| "loss": 2.2211, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.1743444531131, | |
| "learning_rate": 9.763241620263447e-05, | |
| "loss": 2.22, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.07129578039127, | |
| "learning_rate": 9.759278142180348e-05, | |
| "loss": 2.0728, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.2133700569692323, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.1745, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.1119512401535325, | |
| "learning_rate": 9.751254965084056e-05, | |
| "loss": 2.1675, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.134393420684843, | |
| "learning_rate": 9.747195320155651e-05, | |
| "loss": 2.1736, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.3717466432589143, | |
| "learning_rate": 9.7431036740569e-05, | |
| "loss": 2.1831, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.1021306987374873, | |
| "learning_rate": 9.73898005436987e-05, | |
| "loss": 2.098, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.4161462177589432, | |
| "learning_rate": 9.734824488892164e-05, | |
| "loss": 2.1452, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.1435513652600235, | |
| "learning_rate": 9.730637005636738e-05, | |
| "loss": 2.1496, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.2985999802404784, | |
| "learning_rate": 9.726417632831701e-05, | |
| "loss": 2.0894, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.4152240547879988, | |
| "learning_rate": 9.72216639892014e-05, | |
| "loss": 2.152, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 1.305762863593827, | |
| "learning_rate": 9.71788333255991e-05, | |
| "loss": 2.1193, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 1.35035097885144, | |
| "learning_rate": 9.713568462623461e-05, | |
| "loss": 2.2088, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.1536778762582054, | |
| "learning_rate": 9.709221818197624e-05, | |
| "loss": 2.0303, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.0435202624027013, | |
| "learning_rate": 9.704843428583431e-05, | |
| "loss": 2.1607, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.1271503153567661, | |
| "learning_rate": 9.700433323295907e-05, | |
| "loss": 2.0843, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.03689893232153, | |
| "learning_rate": 9.695991532063875e-05, | |
| "loss": 2.0812, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.1067829679650272, | |
| "learning_rate": 9.691518084829756e-05, | |
| "loss": 2.1112, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.0689063335989921, | |
| "learning_rate": 9.687013011749364e-05, | |
| "loss": 2.1808, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.1756020215612761, | |
| "learning_rate": 9.682476343191708e-05, | |
| "loss": 2.1938, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.128094303106894, | |
| "learning_rate": 9.677908109738784e-05, | |
| "loss": 2.2148, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.1487631280959074, | |
| "learning_rate": 9.673308342185365e-05, | |
| "loss": 2.0904, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.0770720795884738, | |
| "learning_rate": 9.668677071538805e-05, | |
| "loss": 2.0042, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.1524178714707023, | |
| "learning_rate": 9.664014329018813e-05, | |
| "loss": 2.0993, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.3540053847451625, | |
| "learning_rate": 9.659320146057262e-05, | |
| "loss": 2.2639, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.682982692329932, | |
| "learning_rate": 9.65459455429796e-05, | |
| "loss": 2.079, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.238575071772689, | |
| "learning_rate": 9.649837585596444e-05, | |
| "loss": 2.2053, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.5751413702608084, | |
| "learning_rate": 9.645049272019773e-05, | |
| "loss": 1.9713, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.597985519705181, | |
| "learning_rate": 9.640229645846294e-05, | |
| "loss": 2.1062, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.1483980505962625, | |
| "learning_rate": 9.635378739565439e-05, | |
| "loss": 2.026, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.5550147218639505, | |
| "learning_rate": 9.630496585877504e-05, | |
| "loss": 2.0588, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.1728865641403023, | |
| "learning_rate": 9.62558321769342e-05, | |
| "loss": 2.1807, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.6202415645668293, | |
| "learning_rate": 9.620638668134541e-05, | |
| "loss": 2.2627, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.1636236508074238, | |
| "learning_rate": 9.615662970532416e-05, | |
| "loss": 2.01, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.3936148798128163, | |
| "learning_rate": 9.610656158428564e-05, | |
| "loss": 2.2421, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.1711660051031476, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 2.0344, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.1639199621824277, | |
| "learning_rate": 9.60054932593026e-05, | |
| "loss": 2.1793, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.0749543681262987, | |
| "learning_rate": 9.59544937366666e-05, | |
| "loss": 2.121, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.1788650531272336, | |
| "learning_rate": 9.590318443162582e-05, | |
| "loss": 2.0115, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.2633632126956826, | |
| "learning_rate": 9.585156569005982e-05, | |
| "loss": 2.0947, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.223649275806162, | |
| "learning_rate": 9.579963785993407e-05, | |
| "loss": 1.9416, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.2368195481290762, | |
| "learning_rate": 9.574740129129767e-05, | |
| "loss": 1.9588, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.2844824463647442, | |
| "learning_rate": 9.569485633628089e-05, | |
| "loss": 2.145, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.3257177579195516, | |
| "learning_rate": 9.564200334909292e-05, | |
| "loss": 2.1133, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.2649495214958049, | |
| "learning_rate": 9.558884268601936e-05, | |
| "loss": 1.9536, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.207154649790508, | |
| "learning_rate": 9.553537470541992e-05, | |
| "loss": 2.0501, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.226854960810907, | |
| "learning_rate": 9.548159976772592e-05, | |
| "loss": 2.1357, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.3962848028974717, | |
| "learning_rate": 9.542751823543793e-05, | |
| "loss": 2.0916, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.6906177805833782, | |
| "learning_rate": 9.537313047312327e-05, | |
| "loss": 2.072, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.0886699544395027, | |
| "learning_rate": 9.53184368474136e-05, | |
| "loss": 1.9649, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.4298747306777986, | |
| "learning_rate": 9.526343772700239e-05, | |
| "loss": 1.9251, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.2833394908795364, | |
| "learning_rate": 9.520813348264252e-05, | |
| "loss": 1.9304, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.1929607057105187, | |
| "learning_rate": 9.515252448714368e-05, | |
| "loss": 1.9338, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.3086597078679512, | |
| "learning_rate": 9.509661111536998e-05, | |
| "loss": 1.9337, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.3132008291775636, | |
| "learning_rate": 9.504039374423729e-05, | |
| "loss": 1.9191, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.300627574978304, | |
| "learning_rate": 9.498387275271074e-05, | |
| "loss": 1.9594, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.3247252542344983, | |
| "learning_rate": 9.492704852180228e-05, | |
| "loss": 1.8434, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 1.2151097739991084, | |
| "learning_rate": 9.486992143456792e-05, | |
| "loss": 1.7906, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 1.4058803238984519, | |
| "learning_rate": 9.48124918761053e-05, | |
| "loss": 1.8975, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 1.6109794449780737, | |
| "learning_rate": 9.475476023355103e-05, | |
| "loss": 2.0031, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 1.5279537347535015, | |
| "learning_rate": 9.46967268960781e-05, | |
| "loss": 1.7328, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 1.3646898955920579, | |
| "learning_rate": 9.46383922548932e-05, | |
| "loss": 1.9444, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 1.5987166715335974, | |
| "learning_rate": 9.457975670323416e-05, | |
| "loss": 1.8069, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 1.3209892956459395, | |
| "learning_rate": 9.452082063636729e-05, | |
| "loss": 1.7247, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 1.803347817097376, | |
| "learning_rate": 9.446158445158468e-05, | |
| "loss": 1.8944, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 2.1748681501794174, | |
| "learning_rate": 9.440204854820149e-05, | |
| "loss": 1.904, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.4970788496336827, | |
| "learning_rate": 9.434221332755339e-05, | |
| "loss": 1.9035, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.7984008193748446, | |
| "learning_rate": 9.428207919299368e-05, | |
| "loss": 1.8869, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.5122903569344504, | |
| "learning_rate": 9.422164654989072e-05, | |
| "loss": 1.9453, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.3291388907563695, | |
| "learning_rate": 9.416091580562512e-05, | |
| "loss": 1.7636, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 1.2950768960374879, | |
| "learning_rate": 9.4099887369587e-05, | |
| "loss": 1.8532, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 1.4517480321525016, | |
| "learning_rate": 9.403856165317321e-05, | |
| "loss": 1.907, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.4135962053548707, | |
| "learning_rate": 9.397693906978468e-05, | |
| "loss": 1.8239, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.6847642690637237, | |
| "learning_rate": 9.391502003482349e-05, | |
| "loss": 1.949, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 1.7586875735963552, | |
| "learning_rate": 9.385280496569006e-05, | |
| "loss": 1.9202, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 1.5460579755051689, | |
| "learning_rate": 9.37902942817805e-05, | |
| "loss": 1.9326, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.9181361062843023, | |
| "learning_rate": 9.372748840448361e-05, | |
| "loss": 1.9873, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.646537282403314, | |
| "learning_rate": 9.366438775717814e-05, | |
| "loss": 1.8314, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 1.8348278120589312, | |
| "learning_rate": 9.360099276522989e-05, | |
| "loss": 1.8685, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 1.8401538788965073, | |
| "learning_rate": 9.353730385598887e-05, | |
| "loss": 2.0042, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 2.0549233121980564, | |
| "learning_rate": 9.347332145878638e-05, | |
| "loss": 1.975, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 1.4833224131753604, | |
| "learning_rate": 9.34090460049322e-05, | |
| "loss": 1.791, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 1.614967484872237, | |
| "learning_rate": 9.334447792771155e-05, | |
| "loss": 1.8568, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 1.4552614688126433, | |
| "learning_rate": 9.327961766238231e-05, | |
| "loss": 1.7902, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.481637692393112, | |
| "learning_rate": 9.321446564617198e-05, | |
| "loss": 1.7595, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.417832774008434, | |
| "learning_rate": 9.314902231827478e-05, | |
| "loss": 1.7908, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 1.510298898158219, | |
| "learning_rate": 9.30832881198487e-05, | |
| "loss": 1.8595, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 1.4357460814625236, | |
| "learning_rate": 9.301726349401249e-05, | |
| "loss": 1.9393, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 1.5366655993308636, | |
| "learning_rate": 9.295094888584269e-05, | |
| "loss": 1.7696, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 1.4987290243265812, | |
| "learning_rate": 9.288434474237064e-05, | |
| "loss": 1.8646, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 1.4086562310656776, | |
| "learning_rate": 9.281745151257946e-05, | |
| "loss": 1.7895, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 1.3997144721225205, | |
| "learning_rate": 9.275026964740101e-05, | |
| "loss": 1.7373, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.504848670193301, | |
| "learning_rate": 9.268279959971286e-05, | |
| "loss": 1.854, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.5235684170064667, | |
| "learning_rate": 9.261504182433528e-05, | |
| "loss": 1.7758, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.5346047487353527, | |
| "learning_rate": 9.254699677802803e-05, | |
| "loss": 1.7958, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.5822284685688552, | |
| "learning_rate": 9.247866491948752e-05, | |
| "loss": 1.8794, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.5968687313602794, | |
| "learning_rate": 9.241004670934348e-05, | |
| "loss": 1.8429, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 1.586810780688857, | |
| "learning_rate": 9.234114261015597e-05, | |
| "loss": 1.7057, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 1.5052767784201073, | |
| "learning_rate": 9.22719530864123e-05, | |
| "loss": 1.7909, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 1.37613725279528, | |
| "learning_rate": 9.220247860452378e-05, | |
| "loss": 1.6222, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 2.1068518462107115, | |
| "learning_rate": 9.21327196328227e-05, | |
| "loss": 1.7406, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.6292680139796645, | |
| "learning_rate": 9.206267664155907e-05, | |
| "loss": 1.7475, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.61623177875573, | |
| "learning_rate": 9.19923501028975e-05, | |
| "loss": 1.8179, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.501240236686501, | |
| "learning_rate": 9.192174049091407e-05, | |
| "loss": 1.8483, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.417464422028843, | |
| "learning_rate": 9.185084828159301e-05, | |
| "loss": 1.6606, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 1.5638371443632166, | |
| "learning_rate": 9.177967395282359e-05, | |
| "loss": 1.7282, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 1.3900047370711577, | |
| "learning_rate": 9.170821798439683e-05, | |
| "loss": 1.8847, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 1.3433227359618032, | |
| "learning_rate": 9.163648085800236e-05, | |
| "loss": 1.7456, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 1.531440728964947, | |
| "learning_rate": 9.156446305722508e-05, | |
| "loss": 1.7164, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.454071434933275, | |
| "learning_rate": 9.149216506754192e-05, | |
| "loss": 1.7632, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.4217995999244628, | |
| "learning_rate": 9.141958737631864e-05, | |
| "loss": 1.7045, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.4516290375634704, | |
| "learning_rate": 9.134673047280645e-05, | |
| "loss": 1.6271, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.8847885939595728, | |
| "learning_rate": 9.12735948481387e-05, | |
| "loss": 1.7511, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 1.5386549781019352, | |
| "learning_rate": 9.120018099532773e-05, | |
| "loss": 1.7609, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 1.7117117148493712, | |
| "learning_rate": 9.112648940926132e-05, | |
| "loss": 1.6948, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 1.6749257064571572, | |
| "learning_rate": 9.105252058669957e-05, | |
| "loss": 1.8191, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 1.6117317295437434, | |
| "learning_rate": 9.097827502627136e-05, | |
| "loss": 1.7689, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.5033785932097372, | |
| "learning_rate": 9.090375322847118e-05, | |
| "loss": 1.7251, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.7866264034825081, | |
| "learning_rate": 9.082895569565554e-05, | |
| "loss": 1.7619, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.9781011069119638, | |
| "learning_rate": 9.075388293203978e-05, | |
| "loss": 1.7155, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 1.4615980851304928, | |
| "learning_rate": 9.067853544369457e-05, | |
| "loss": 1.7907, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 2.0007876107435676, | |
| "learning_rate": 9.060291373854251e-05, | |
| "loss": 1.7825, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 1.6384769175784961, | |
| "learning_rate": 9.052701832635471e-05, | |
| "loss": 1.6804, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 2.0599174513983147, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.6904, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.5662842107292945, | |
| "learning_rate": 9.037440842917834e-05, | |
| "loss": 1.8675, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.7107030874773999, | |
| "learning_rate": 9.029769497294358e-05, | |
| "loss": 1.7887, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.4731513376744692, | |
| "learning_rate": 9.022070986717379e-05, | |
| "loss": 1.7093, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.4993862777242914, | |
| "learning_rate": 9.014345363083086e-05, | |
| "loss": 1.6706, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 1.563437308102064, | |
| "learning_rate": 9.006592678470443e-05, | |
| "loss": 1.7706, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 1.8794149563978206, | |
| "learning_rate": 8.998812985140825e-05, | |
| "loss": 1.719, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 1.8829845552801692, | |
| "learning_rate": 8.991006335537684e-05, | |
| "loss": 1.7171, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 2.1156104006085124, | |
| "learning_rate": 8.98317278228618e-05, | |
| "loss": 1.7734, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 2.442631405974673, | |
| "learning_rate": 8.975312378192837e-05, | |
| "loss": 1.7274, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 1.6915501393635948, | |
| "learning_rate": 8.967425176245178e-05, | |
| "loss": 1.7179, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.404580018571904, | |
| "learning_rate": 8.959511229611376e-05, | |
| "loss": 1.7432, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 3.333137910455401, | |
| "learning_rate": 8.951570591639889e-05, | |
| "loss": 1.7969, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 1.48041794418801, | |
| "learning_rate": 8.943603315859101e-05, | |
| "loss": 1.7002, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 1.930744743646977, | |
| "learning_rate": 8.93560945597697e-05, | |
| "loss": 1.8734, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 1.7263585594923776, | |
| "learning_rate": 8.927589065880656e-05, | |
| "loss": 1.7049, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 1.9926289875861087, | |
| "learning_rate": 8.919542199636158e-05, | |
| "loss": 1.8248, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 1.76134542441635, | |
| "learning_rate": 8.911468911487954e-05, | |
| "loss": 1.7953, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 1.7467213180841994, | |
| "learning_rate": 8.90336925585864e-05, | |
| "loss": 1.7202, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.7277759342302612, | |
| "learning_rate": 8.89524328734855e-05, | |
| "loss": 1.5597, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.7093511557938714, | |
| "learning_rate": 8.887091060735395e-05, | |
| "loss": 1.6057, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 2.082397440792827, | |
| "learning_rate": 8.878912630973897e-05, | |
| "loss": 1.6007, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 1.6383928868041817, | |
| "learning_rate": 8.870708053195413e-05, | |
| "loss": 1.7003, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 2.176555768389211, | |
| "learning_rate": 8.862477382707568e-05, | |
| "loss": 1.8684, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.55540411920029, | |
| "learning_rate": 8.854220674993876e-05, | |
| "loss": 1.6201, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 2.2046090494061046, | |
| "learning_rate": 8.845937985713375e-05, | |
| "loss": 1.7399, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 1.9556668530055292, | |
| "learning_rate": 8.83762937070024e-05, | |
| "loss": 1.7704, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 1.6710388245941092, | |
| "learning_rate": 8.829294885963419e-05, | |
| "loss": 1.7269, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 1.8191831827549731, | |
| "learning_rate": 8.820934587686247e-05, | |
| "loss": 1.5905, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 1.4217390270525492, | |
| "learning_rate": 8.812548532226068e-05, | |
| "loss": 1.7362, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.8153968846016577, | |
| "learning_rate": 8.80413677611386e-05, | |
| "loss": 1.7309, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.7146733473628513, | |
| "learning_rate": 8.795699376053852e-05, | |
| "loss": 1.6261, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.591439802389271, | |
| "learning_rate": 8.787236388923137e-05, | |
| "loss": 1.6449, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.716038939405748, | |
| "learning_rate": 8.778747871771292e-05, | |
| "loss": 1.6828, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 2.0077680436159753, | |
| "learning_rate": 8.770233881819997e-05, | |
| "loss": 1.6376, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 2.2775492331381915, | |
| "learning_rate": 8.761694476462645e-05, | |
| "loss": 1.7102, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 1.7472535284569068, | |
| "learning_rate": 8.753129713263951e-05, | |
| "loss": 1.5514, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 2.1058629743708677, | |
| "learning_rate": 8.744539649959578e-05, | |
| "loss": 1.7601, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 1.7173809697164537, | |
| "learning_rate": 8.735924344455732e-05, | |
| "loss": 1.5444, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 1.982345849551334, | |
| "learning_rate": 8.72728385482878e-05, | |
| "loss": 1.6174, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 1.9136730004725138, | |
| "learning_rate": 8.718618239324858e-05, | |
| "loss": 1.6849, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 2.1116990742600708, | |
| "learning_rate": 8.709927556359476e-05, | |
| "loss": 1.6695, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 2.0488063550758775, | |
| "learning_rate": 8.701211864517126e-05, | |
| "loss": 1.6461, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 2.022802941321216, | |
| "learning_rate": 8.692471222550886e-05, | |
| "loss": 1.7081, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.8526395070300479, | |
| "learning_rate": 8.683705689382024e-05, | |
| "loss": 1.5796, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.6577123900861979, | |
| "learning_rate": 8.674915324099606e-05, | |
| "loss": 1.7203, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 2.194377411789526, | |
| "learning_rate": 8.666100185960087e-05, | |
| "loss": 1.7998, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.7198491303485337, | |
| "learning_rate": 8.657260334386918e-05, | |
| "loss": 1.6838, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.903053663512861, | |
| "learning_rate": 8.64839582897015e-05, | |
| "loss": 1.5624, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 1.7096354901887252, | |
| "learning_rate": 8.639506729466023e-05, | |
| "loss": 1.6499, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 1.7826939143664446, | |
| "learning_rate": 8.630593095796567e-05, | |
| "loss": 1.6089, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.7647962095089293, | |
| "learning_rate": 8.621654988049203e-05, | |
| "loss": 1.5572, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.55746199166791, | |
| "learning_rate": 8.612692466476328e-05, | |
| "loss": 1.5801, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.7619777402169174, | |
| "learning_rate": 8.603705591494916e-05, | |
| "loss": 1.6376, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.1737997419907598, | |
| "learning_rate": 8.594694423686112e-05, | |
| "loss": 1.6056, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 2.299248894851507, | |
| "learning_rate": 8.585659023794818e-05, | |
| "loss": 1.6231, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 2.451871871872707, | |
| "learning_rate": 8.576599452729287e-05, | |
| "loss": 1.5327, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 2.225164953545606, | |
| "learning_rate": 8.567515771560706e-05, | |
| "loss": 1.5523, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 1.9543488700652831, | |
| "learning_rate": 8.558408041522801e-05, | |
| "loss": 1.5363, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.886477599133871, | |
| "learning_rate": 8.549276324011406e-05, | |
| "loss": 1.6014, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.592624781428746, | |
| "learning_rate": 8.540120680584054e-05, | |
| "loss": 1.4036, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 1.6835146775379843, | |
| "learning_rate": 8.53094117295957e-05, | |
| "loss": 1.6554, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 2.5006431620641227, | |
| "learning_rate": 8.521737863017644e-05, | |
| "loss": 1.7409, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 1.8712857362794038, | |
| "learning_rate": 8.512510812798426e-05, | |
| "loss": 1.6594, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 2.35024943019957, | |
| "learning_rate": 8.503260084502094e-05, | |
| "loss": 1.661, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 2.3859714271819596, | |
| "learning_rate": 8.493985740488444e-05, | |
| "loss": 1.6869, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 3.730804556677943, | |
| "learning_rate": 8.484687843276469e-05, | |
| "loss": 1.6895, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 3.401444185223409, | |
| "learning_rate": 8.475366455543929e-05, | |
| "loss": 1.6232, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 2.936751834059242, | |
| "learning_rate": 8.466021640126945e-05, | |
| "loss": 1.559, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 2.3341742140687223, | |
| "learning_rate": 8.45665346001956e-05, | |
| "loss": 1.4884, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 2.5293531816082786, | |
| "learning_rate": 8.447261978373319e-05, | |
| "loss": 1.6024, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 2.065188181154739, | |
| "learning_rate": 8.43784725849685e-05, | |
| "loss": 1.616, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 2.573776972883085, | |
| "learning_rate": 8.428409363855423e-05, | |
| "loss": 1.5546, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 1.70266974318547, | |
| "learning_rate": 8.418948358070535e-05, | |
| "loss": 1.5945, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 2.2783543679887166, | |
| "learning_rate": 8.409464304919484e-05, | |
| "loss": 1.5214, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.7993998742820507, | |
| "learning_rate": 8.399957268334915e-05, | |
| "loss": 1.4325, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 1.9184711153439358, | |
| "learning_rate": 8.390427312404426e-05, | |
| "loss": 1.5695, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 2.7817088284146823, | |
| "learning_rate": 8.380874501370097e-05, | |
| "loss": 1.6329, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 1.9395373662028492, | |
| "learning_rate": 8.371298899628091e-05, | |
| "loss": 1.4374, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 2.1823508498843465, | |
| "learning_rate": 8.361700571728195e-05, | |
| "loss": 1.4376, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 1.9441045910036505, | |
| "learning_rate": 8.352079582373398e-05, | |
| "loss": 1.5261, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 2.6680215972330483, | |
| "learning_rate": 8.342435996419453e-05, | |
| "loss": 1.5437, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 1.5341843239220305, | |
| "learning_rate": 8.332769878874434e-05, | |
| "loss": 1.5105, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 2.064899822972901, | |
| "learning_rate": 8.323081294898308e-05, | |
| "loss": 1.4466, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 2.1635779238981674, | |
| "learning_rate": 8.313370309802483e-05, | |
| "loss": 1.5804, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.5913181677729609, | |
| "learning_rate": 8.30363698904938e-05, | |
| "loss": 1.369, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.8661112716137087, | |
| "learning_rate": 8.293881398251984e-05, | |
| "loss": 1.4641, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.6880309988396638, | |
| "learning_rate": 8.284103603173402e-05, | |
| "loss": 1.4163, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.7598674336737188, | |
| "learning_rate": 8.274303669726426e-05, | |
| "loss": 1.4407, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 2.223066643321913, | |
| "learning_rate": 8.264481663973079e-05, | |
| "loss": 1.5348, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 1.961444354246269, | |
| "learning_rate": 8.254637652124182e-05, | |
| "loss": 1.5079, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 1.9212105252554303, | |
| "learning_rate": 8.244771700538892e-05, | |
| "loss": 1.5176, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 2.385157536103143, | |
| "learning_rate": 8.234883875724269e-05, | |
| "loss": 1.5019, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 2.4426799290713985, | |
| "learning_rate": 8.224974244334819e-05, | |
| "loss": 1.5466, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.409843251519227, | |
| "learning_rate": 8.215042873172054e-05, | |
| "loss": 1.5295, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.8305121574631777, | |
| "learning_rate": 8.205089829184025e-05, | |
| "loss": 1.5378, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.171794845016919, | |
| "learning_rate": 8.19511517946489e-05, | |
| "loss": 1.4222, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 2.3215037748178333, | |
| "learning_rate": 8.185118991254448e-05, | |
| "loss": 1.4534, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 1.8316113817877149, | |
| "learning_rate": 8.175101331937693e-05, | |
| "loss": 1.4417, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 2.1946076531379464, | |
| "learning_rate": 8.165062269044353e-05, | |
| "loss": 1.5138, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 2.1907681156016516, | |
| "learning_rate": 8.155001870248443e-05, | |
| "loss": 1.4666, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 2.125095559400482, | |
| "learning_rate": 8.144920203367805e-05, | |
| "loss": 1.6482, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 2.45374817090128, | |
| "learning_rate": 8.134817336363647e-05, | |
| "loss": 1.5389, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 3.327670627449134, | |
| "learning_rate": 8.124693337340092e-05, | |
| "loss": 1.5057, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 1.7605999703516142, | |
| "learning_rate": 8.114548274543717e-05, | |
| "loss": 1.4962, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 2.5390980109776518, | |
| "learning_rate": 8.104382216363083e-05, | |
| "loss": 1.4682, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 1.796778995698997, | |
| "learning_rate": 8.09419523132829e-05, | |
| "loss": 1.4226, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 2.6450673226885884, | |
| "learning_rate": 8.083987388110506e-05, | |
| "loss": 1.4184, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 1.7418724263237955, | |
| "learning_rate": 8.073758755521505e-05, | |
| "loss": 1.5211, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 2.8442143515382683, | |
| "learning_rate": 8.063509402513202e-05, | |
| "loss": 1.563, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 1.7751877109232908, | |
| "learning_rate": 8.053239398177191e-05, | |
| "loss": 1.5094, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 2.3572545909439153, | |
| "learning_rate": 8.042948811744279e-05, | |
| "loss": 1.4501, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.7325129850559884, | |
| "learning_rate": 8.032637712584016e-05, | |
| "loss": 1.4899, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 1.7140609126170652, | |
| "learning_rate": 8.022306170204233e-05, | |
| "loss": 1.3113, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 1.954964733079678, | |
| "learning_rate": 8.011954254250564e-05, | |
| "loss": 1.4376, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.7244124021341851, | |
| "learning_rate": 8.001582034505992e-05, | |
| "loss": 1.3247, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.7787360360696058, | |
| "learning_rate": 7.991189580890362e-05, | |
| "loss": 1.4327, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 1.9997882548948591, | |
| "learning_rate": 7.980776963459918e-05, | |
| "loss": 1.4778, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 2.1221502897354845, | |
| "learning_rate": 7.970344252406831e-05, | |
| "loss": 1.4963, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.993386408185386, | |
| "learning_rate": 7.959891518058726e-05, | |
| "loss": 1.6095, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.932317142829957, | |
| "learning_rate": 7.949418830878203e-05, | |
| "loss": 1.4098, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 2.082446688957067, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.5336, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 1.9454242914487379, | |
| "learning_rate": 7.928413880542349e-05, | |
| "loss": 1.5305, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 1.99506133297581, | |
| "learning_rate": 7.917881758982837e-05, | |
| "loss": 1.5216, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 2.254023966913613, | |
| "learning_rate": 7.907329967781581e-05, | |
| "loss": 1.5393, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 1.9464211505340732, | |
| "learning_rate": 7.896758578068938e-05, | |
| "loss": 1.3219, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 2.362112480268074, | |
| "learning_rate": 7.886167661107369e-05, | |
| "loss": 1.4254, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 2.113169681211877, | |
| "learning_rate": 7.875557288290976e-05, | |
| "loss": 1.5297, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 2.819420952915015, | |
| "learning_rate": 7.864927531145011e-05, | |
| "loss": 1.5851, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 2.0096679012243275, | |
| "learning_rate": 7.8542784613254e-05, | |
| "loss": 1.5072, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 2.113866615243935, | |
| "learning_rate": 7.843610150618255e-05, | |
| "loss": 1.45, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 2.712117674912262, | |
| "learning_rate": 7.832922670939391e-05, | |
| "loss": 1.561, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 2.2564450730972054, | |
| "learning_rate": 7.822216094333847e-05, | |
| "loss": 1.3621, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 2.2003646964404395, | |
| "learning_rate": 7.81149049297539e-05, | |
| "loss": 1.4346, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 2.445037967760889, | |
| "learning_rate": 7.800745939166039e-05, | |
| "loss": 1.3095, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 1.9665903892711243, | |
| "learning_rate": 7.78998250533557e-05, | |
| "loss": 1.3709, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.5663020248425346, | |
| "learning_rate": 7.779200264041029e-05, | |
| "loss": 1.2765, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.939213611440194, | |
| "learning_rate": 7.76839928796625e-05, | |
| "loss": 1.3327, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 2.6026275340664484, | |
| "learning_rate": 7.757579649921354e-05, | |
| "loss": 1.2207, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 2.238214826445917, | |
| "learning_rate": 7.746741422842265e-05, | |
| "loss": 1.2125, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 2.66806457212071, | |
| "learning_rate": 7.73588467979022e-05, | |
| "loss": 1.29, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 2.584096398757311, | |
| "learning_rate": 7.72500949395127e-05, | |
| "loss": 1.1871, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 2.4311420186260175, | |
| "learning_rate": 7.714115938635791e-05, | |
| "loss": 1.2876, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 3.77557316145633, | |
| "learning_rate": 7.703204087277988e-05, | |
| "loss": 1.2485, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 1.821953691360643, | |
| "learning_rate": 7.692274013435403e-05, | |
| "loss": 1.1479, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 3.201342095326416, | |
| "learning_rate": 7.681325790788416e-05, | |
| "loss": 1.2582, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 2.7464736714373545, | |
| "learning_rate": 7.670359493139751e-05, | |
| "loss": 1.1313, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 2.9009868596624075, | |
| "learning_rate": 7.659375194413972e-05, | |
| "loss": 1.2187, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 3.052337163805675, | |
| "learning_rate": 7.648372968656993e-05, | |
| "loss": 1.3169, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 2.1429819676835673, | |
| "learning_rate": 7.637352890035576e-05, | |
| "loss": 1.2177, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 3.402335808460986, | |
| "learning_rate": 7.626315032836831e-05, | |
| "loss": 1.3171, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 2.470139404158673, | |
| "learning_rate": 7.615259471467711e-05, | |
| "loss": 1.1625, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 2.6632734157067186, | |
| "learning_rate": 7.604186280454515e-05, | |
| "loss": 1.2448, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 3.2386066201077153, | |
| "learning_rate": 7.593095534442388e-05, | |
| "loss": 1.2964, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.8887801977922518, | |
| "learning_rate": 7.58198730819481e-05, | |
| "loss": 1.1535, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 2.6561480904516936, | |
| "learning_rate": 7.570861676593101e-05, | |
| "loss": 1.1787, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 2.9503555474300738, | |
| "learning_rate": 7.559718714635907e-05, | |
| "loss": 1.252, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.0393889897111874, | |
| "learning_rate": 7.548558497438702e-05, | |
| "loss": 1.2115, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 3.6105351565864656, | |
| "learning_rate": 7.537381100233278e-05, | |
| "loss": 1.276, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 1.9484218148686905, | |
| "learning_rate": 7.526186598367242e-05, | |
| "loss": 1.157, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 2.579375437419657, | |
| "learning_rate": 7.514975067303496e-05, | |
| "loss": 1.1504, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 2.9320519655563406, | |
| "learning_rate": 7.503746582619746e-05, | |
| "loss": 1.2654, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 1.9057779312607441, | |
| "learning_rate": 7.492501220007979e-05, | |
| "loss": 1.1034, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 2.776688972314242, | |
| "learning_rate": 7.481239055273959e-05, | |
| "loss": 1.2567, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 2.099806023719329, | |
| "learning_rate": 7.469960164336711e-05, | |
| "loss": 1.2164, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 2.0792937479825544, | |
| "learning_rate": 7.45866462322802e-05, | |
| "loss": 1.1962, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 2.8329508294292762, | |
| "learning_rate": 7.447352508091902e-05, | |
| "loss": 1.155, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 1.9292205279805203, | |
| "learning_rate": 7.436023895184102e-05, | |
| "loss": 1.0866, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 2.2760375198196017, | |
| "learning_rate": 7.424678860871584e-05, | |
| "loss": 1.1146, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 2.5165820726353645, | |
| "learning_rate": 7.413317481632002e-05, | |
| "loss": 1.2126, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 1.9170988142441296, | |
| "learning_rate": 7.401939834053197e-05, | |
| "loss": 1.1024, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 2.315980121195765, | |
| "learning_rate": 7.390545994832672e-05, | |
| "loss": 1.2745, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 2.3814944212079134, | |
| "learning_rate": 7.379136040777083e-05, | |
| "loss": 1.2071, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 1.9782914121729676, | |
| "learning_rate": 7.367710048801715e-05, | |
| "loss": 1.0375, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 2.8207703340888797, | |
| "learning_rate": 7.356268095929966e-05, | |
| "loss": 1.1569, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 2.0428976792280995, | |
| "learning_rate": 7.344810259292829e-05, | |
| "loss": 1.1255, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 2.3280327206192055, | |
| "learning_rate": 7.333336616128369e-05, | |
| "loss": 1.1719, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 2.30854568329358, | |
| "learning_rate": 7.321847243781202e-05, | |
| "loss": 1.059, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 2.01469829840449, | |
| "learning_rate": 7.310342219701981e-05, | |
| "loss": 1.0694, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 2.4051870652897813, | |
| "learning_rate": 7.298821621446862e-05, | |
| "loss": 1.1172, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 2.397042730095134, | |
| "learning_rate": 7.287285526676994e-05, | |
| "loss": 1.0745, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 2.09738199921565, | |
| "learning_rate": 7.275734013157981e-05, | |
| "loss": 1.1541, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 1.8971514141852077, | |
| "learning_rate": 7.264167158759374e-05, | |
| "loss": 1.0936, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 2.1251621373701735, | |
| "learning_rate": 7.252585041454135e-05, | |
| "loss": 1.1499, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 2.5685341355572535, | |
| "learning_rate": 7.24098773931811e-05, | |
| "loss": 1.2119, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 2.2445872219136986, | |
| "learning_rate": 7.229375330529517e-05, | |
| "loss": 1.0626, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 2.142309249994188, | |
| "learning_rate": 7.217747893368397e-05, | |
| "loss": 1.1265, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 2.134182273068111, | |
| "learning_rate": 7.206105506216106e-05, | |
| "loss": 1.1575, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 2.220588967630846, | |
| "learning_rate": 7.19444824755478e-05, | |
| "loss": 1.1859, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 2.1750251516302312, | |
| "learning_rate": 7.182776195966796e-05, | |
| "loss": 1.1047, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 2.022638604825416, | |
| "learning_rate": 7.171089430134262e-05, | |
| "loss": 1.1084, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 2.3743719611396323, | |
| "learning_rate": 7.159388028838467e-05, | |
| "loss": 1.1561, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 2.166524139650481, | |
| "learning_rate": 7.147672070959367e-05, | |
| "loss": 1.1017, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 2.0944271029444685, | |
| "learning_rate": 7.135941635475036e-05, | |
| "loss": 0.9974, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 1.7185154761086074, | |
| "learning_rate": 7.12419680146115e-05, | |
| "loss": 1.0276, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 2.137309748612882, | |
| "learning_rate": 7.11243764809044e-05, | |
| "loss": 1.0702, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 1.9027222853696886, | |
| "learning_rate": 7.10066425463217e-05, | |
| "loss": 1.0934, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 2.531230429576785, | |
| "learning_rate": 7.088876700451597e-05, | |
| "loss": 1.2123, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 1.858065075971331, | |
| "learning_rate": 7.077075065009433e-05, | |
| "loss": 1.0934, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.8858607204253484, | |
| "learning_rate": 7.065259427861315e-05, | |
| "loss": 1.103, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 2.19179446624781, | |
| "learning_rate": 7.053429868657265e-05, | |
| "loss": 1.094, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 2.4746530572536316, | |
| "learning_rate": 7.041586467141159e-05, | |
| "loss": 1.0045, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 1.9916500805485537, | |
| "learning_rate": 7.029729303150178e-05, | |
| "loss": 1.041, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 2.700559567585137, | |
| "learning_rate": 7.017858456614284e-05, | |
| "loss": 1.169, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 2.028105002642923, | |
| "learning_rate": 7.005974007555667e-05, | |
| "loss": 1.0877, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 2.7109137146024604, | |
| "learning_rate": 6.994076036088218e-05, | |
| "loss": 0.9836, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 2.626642663205964, | |
| "learning_rate": 6.98216462241698e-05, | |
| "loss": 1.1233, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 1.909111404091151, | |
| "learning_rate": 6.970239846837614e-05, | |
| "loss": 1.0951, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 2.776976335760313, | |
| "learning_rate": 6.958301789735852e-05, | |
| "loss": 1.1482, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 2.300253577372805, | |
| "learning_rate": 6.946350531586959e-05, | |
| "loss": 1.1585, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 2.3309592650922935, | |
| "learning_rate": 6.934386152955189e-05, | |
| "loss": 1.2012, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 2.369741746524773, | |
| "learning_rate": 6.922408734493238e-05, | |
| "loss": 1.152, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 2.0241198597304852, | |
| "learning_rate": 6.910418356941711e-05, | |
| "loss": 1.2441, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 2.2368491287524437, | |
| "learning_rate": 6.89841510112857e-05, | |
| "loss": 1.0747, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 2.0809096315967683, | |
| "learning_rate": 6.886399047968585e-05, | |
| "loss": 1.1075, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 1.8853633460503207, | |
| "learning_rate": 6.874370278462803e-05, | |
| "loss": 1.0238, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 2.2512885382272576, | |
| "learning_rate": 6.862328873697978e-05, | |
| "loss": 1.1333, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 2.0111670741240073, | |
| "learning_rate": 6.850274914846058e-05, | |
| "loss": 1.1767, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 2.045878818200496, | |
| "learning_rate": 6.8382084831636e-05, | |
| "loss": 1.0164, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 2.462234387953122, | |
| "learning_rate": 6.826129659991259e-05, | |
| "loss": 1.224, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 2.5212056031238284, | |
| "learning_rate": 6.814038526753205e-05, | |
| "loss": 1.0609, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 2.275687998159092, | |
| "learning_rate": 6.801935164956606e-05, | |
| "loss": 1.0613, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 2.3519107835461086, | |
| "learning_rate": 6.789819656191053e-05, | |
| "loss": 1.2186, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 2.6267930875729726, | |
| "learning_rate": 6.777692082128024e-05, | |
| "loss": 1.0959, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 2.1137431323399283, | |
| "learning_rate": 6.765552524520333e-05, | |
| "loss": 1.1186, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 2.3352459521629636, | |
| "learning_rate": 6.753401065201572e-05, | |
| "loss": 1.0901, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 2.698461645646085, | |
| "learning_rate": 6.741237786085565e-05, | |
| "loss": 1.0986, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 2.3143924074898745, | |
| "learning_rate": 6.729062769165813e-05, | |
| "loss": 1.0429, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 2.448558108168179, | |
| "learning_rate": 6.716876096514943e-05, | |
| "loss": 1.1144, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 1.9577459215715654, | |
| "learning_rate": 6.704677850284153e-05, | |
| "loss": 0.9685, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 2.119991221572539, | |
| "learning_rate": 6.692468112702663e-05, | |
| "loss": 1.2351, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 2.649211621943293, | |
| "learning_rate": 6.680246966077151e-05, | |
| "loss": 1.0469, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 2.159159938354453, | |
| "learning_rate": 6.66801449279121e-05, | |
| "loss": 1.0609, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 2.309516482894794, | |
| "learning_rate": 6.655770775304783e-05, | |
| "loss": 1.1226, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 2.3406654339822057, | |
| "learning_rate": 6.643515896153612e-05, | |
| "loss": 1.0319, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 2.1265533768826606, | |
| "learning_rate": 6.631249937948683e-05, | |
| "loss": 0.99, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 2.1626668572204855, | |
| "learning_rate": 6.61897298337566e-05, | |
| "loss": 1.0612, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 2.5061319831634137, | |
| "learning_rate": 6.606685115194341e-05, | |
| "loss": 1.1213, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 1.9730571009320355, | |
| "learning_rate": 6.594386416238094e-05, | |
| "loss": 0.9766, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 2.380203734848038, | |
| "learning_rate": 6.582076969413291e-05, | |
| "loss": 1.1466, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.4579437943060958, | |
| "learning_rate": 6.569756857698761e-05, | |
| "loss": 1.0207, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.172588349823695, | |
| "learning_rate": 6.557426164145226e-05, | |
| "loss": 1.0883, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 2.491028408971267, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.1004, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 2.6041109063917345, | |
| "learning_rate": 6.532733364080126e-05, | |
| "loss": 1.0826, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 2.2824163011548273, | |
| "learning_rate": 6.520371424024425e-05, | |
| "loss": 1.1966, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 2.1216014895806476, | |
| "learning_rate": 6.507999235040327e-05, | |
| "loss": 1.0633, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 2.5187998065016783, | |
| "learning_rate": 6.495616880529607e-05, | |
| "loss": 1.0776, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 2.2958553015888574, | |
| "learning_rate": 6.483224443962569e-05, | |
| "loss": 0.9498, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 2.033929496628245, | |
| "learning_rate": 6.470822008877482e-05, | |
| "loss": 1.0443, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 2.6394776694992492, | |
| "learning_rate": 6.458409658880014e-05, | |
| "loss": 1.1617, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 2.6019441956420417, | |
| "learning_rate": 6.44598747764267e-05, | |
| "loss": 1.0233, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 2.4721186265423274, | |
| "learning_rate": 6.433555548904228e-05, | |
| "loss": 1.1522, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 2.6122933227946405, | |
| "learning_rate": 6.421113956469179e-05, | |
| "loss": 1.1163, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 2.10006870450658, | |
| "learning_rate": 6.408662784207149e-05, | |
| "loss": 0.9744, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 2.6382488481470467, | |
| "learning_rate": 6.396202116052348e-05, | |
| "loss": 1.1331, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 2.5057892374375523, | |
| "learning_rate": 6.383732036003e-05, | |
| "loss": 1.1327, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 2.547354869103696, | |
| "learning_rate": 6.371252628120772e-05, | |
| "loss": 1.1525, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 2.326241142633525, | |
| "learning_rate": 6.358763976530208e-05, | |
| "loss": 1.1238, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 1.9377670783989376, | |
| "learning_rate": 6.346266165418173e-05, | |
| "loss": 1.0251, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 2.2613719862983066, | |
| "learning_rate": 6.333759279033268e-05, | |
| "loss": 0.9448, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 2.1794997941251215, | |
| "learning_rate": 6.321243401685276e-05, | |
| "loss": 1.0504, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 2.466226368107345, | |
| "learning_rate": 6.308718617744588e-05, | |
| "loss": 1.0916, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 2.1742039942571965, | |
| "learning_rate": 6.296185011641634e-05, | |
| "loss": 1.1007, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 2.040258935713953, | |
| "learning_rate": 6.283642667866316e-05, | |
| "loss": 1.0868, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 2.406767020922857, | |
| "learning_rate": 6.271091670967436e-05, | |
| "loss": 1.0912, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 1.973291950359408, | |
| "learning_rate": 6.258532105552126e-05, | |
| "loss": 0.9685, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 2.4201118522587324, | |
| "learning_rate": 6.245964056285283e-05, | |
| "loss": 1.0032, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 2.0671271483824394, | |
| "learning_rate": 6.233387607888994e-05, | |
| "loss": 0.9443, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 2.2638066875220235, | |
| "learning_rate": 6.220802845141958e-05, | |
| "loss": 1.0463, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 1.979751969912936, | |
| "learning_rate": 6.208209852878929e-05, | |
| "loss": 1.0013, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 1.97052458676751, | |
| "learning_rate": 6.195608715990137e-05, | |
| "loss": 0.8921, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 2.1359146414425223, | |
| "learning_rate": 6.182999519420709e-05, | |
| "loss": 0.9126, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 2.7443373359155863, | |
| "learning_rate": 6.17038234817011e-05, | |
| "loss": 1.0791, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 2.164665226856409, | |
| "learning_rate": 6.157757287291557e-05, | |
| "loss": 1.0569, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 2.4311659767949663, | |
| "learning_rate": 6.145124421891457e-05, | |
| "loss": 0.915, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 2.4925543422048846, | |
| "learning_rate": 6.132483837128823e-05, | |
| "loss": 0.9641, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 2.2752385244658475, | |
| "learning_rate": 6.119835618214707e-05, | |
| "loss": 0.9996, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 2.623016577042139, | |
| "learning_rate": 6.107179850411623e-05, | |
| "loss": 1.0951, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 2.125306679310265, | |
| "learning_rate": 6.094516619032975e-05, | |
| "loss": 1.091, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 2.360177620598592, | |
| "learning_rate": 6.081846009442474e-05, | |
| "loss": 1.1818, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 2.5149767270595986, | |
| "learning_rate": 6.0691681070535735e-05, | |
| "loss": 1.002, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 2.089686984930275, | |
| "learning_rate": 6.056482997328884e-05, | |
| "loss": 1.0515, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 2.0549862773735197, | |
| "learning_rate": 6.0437907657796034e-05, | |
| "loss": 0.9499, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 1.9234961671956081, | |
| "learning_rate": 6.031091497964941e-05, | |
| "loss": 1.0201, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 2.2143228673447672, | |
| "learning_rate": 6.018385279491533e-05, | |
| "loss": 1.0318, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 1.9389429285300634, | |
| "learning_rate": 6.0056721960128706e-05, | |
| "loss": 0.9639, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 2.1462225665383023, | |
| "learning_rate": 5.992952333228728e-05, | |
| "loss": 1.071, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 1.9215230295647012, | |
| "learning_rate": 5.9802257768845725e-05, | |
| "loss": 0.943, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 2.3941033919952166, | |
| "learning_rate": 5.967492612770999e-05, | |
| "loss": 1.0517, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 2.232755303714322, | |
| "learning_rate": 5.9547529267231405e-05, | |
| "loss": 1.0703, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 2.3842216297298586, | |
| "learning_rate": 5.9420068046200995e-05, | |
| "loss": 0.9261, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 2.0816155307568645, | |
| "learning_rate": 5.92925433238436e-05, | |
| "loss": 1.0535, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 2.167412994768514, | |
| "learning_rate": 5.9164955959812176e-05, | |
| "loss": 0.9592, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 2.9863715822293564, | |
| "learning_rate": 5.90373068141819e-05, | |
| "loss": 1.1386, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 2.3076847313139224, | |
| "learning_rate": 5.8909596747444477e-05, | |
| "loss": 1.0339, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 2.2380455567287654, | |
| "learning_rate": 5.8781826620502223e-05, | |
| "loss": 1.0531, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 2.51600693580925, | |
| "learning_rate": 5.865399729466237e-05, | |
| "loss": 1.039, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 1.9149651267080585, | |
| "learning_rate": 5.85261096316312e-05, | |
| "loss": 0.9387, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 2.653097407966605, | |
| "learning_rate": 5.8398164493508244e-05, | |
| "loss": 1.0379, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 1.9821281385596556, | |
| "learning_rate": 5.827016274278051e-05, | |
| "loss": 1.0068, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 2.039932043531972, | |
| "learning_rate": 5.814210524231657e-05, | |
| "loss": 0.8729, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 2.0716898073237875, | |
| "learning_rate": 5.8013992855360876e-05, | |
| "loss": 0.9907, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 2.107950874658358, | |
| "learning_rate": 5.788582644552782e-05, | |
| "loss": 0.9477, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.4948785349577216, | |
| "learning_rate": 5.775760687679603e-05, | |
| "loss": 0.9646, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.096933319167589, | |
| "learning_rate": 5.762933501350242e-05, | |
| "loss": 1.1164, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 1.8664721075360444, | |
| "learning_rate": 5.750101172033643e-05, | |
| "loss": 1.0111, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 2.0478243231610986, | |
| "learning_rate": 5.737263786233423e-05, | |
| "loss": 0.9597, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 2.006878992868448, | |
| "learning_rate": 5.724421430487282e-05, | |
| "loss": 0.9658, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 1.9664569236393388, | |
| "learning_rate": 5.7115741913664264e-05, | |
| "loss": 0.9627, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 2.3845447947867338, | |
| "learning_rate": 5.698722155474977e-05, | |
| "loss": 1.0821, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 2.248729684550033, | |
| "learning_rate": 5.6858654094493924e-05, | |
| "loss": 0.9435, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 2.033462978991578, | |
| "learning_rate": 5.673004039957882e-05, | |
| "loss": 0.9197, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 2.8004619095778063, | |
| "learning_rate": 5.660138133699825e-05, | |
| "loss": 0.9732, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 2.4219270648233646, | |
| "learning_rate": 5.647267777405177e-05, | |
| "loss": 1.0443, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 2.4573802349852407, | |
| "learning_rate": 5.634393057833899e-05, | |
| "loss": 0.9246, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 2.368257545780357, | |
| "learning_rate": 5.62151406177536e-05, | |
| "loss": 0.8357, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 2.5935487092070155, | |
| "learning_rate": 5.608630876047759e-05, | |
| "loss": 1.0068, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 2.6858032479162395, | |
| "learning_rate": 5.5957435874975395e-05, | |
| "loss": 0.9448, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 1.9761310479981815, | |
| "learning_rate": 5.5828522829987964e-05, | |
| "loss": 0.9014, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 3.557209105858733, | |
| "learning_rate": 5.569957049452703e-05, | |
| "loss": 1.0596, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 2.2170383737498356, | |
| "learning_rate": 5.5570579737869166e-05, | |
| "loss": 0.9958, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 2.1058095173724887, | |
| "learning_rate": 5.544155142954993e-05, | |
| "loss": 0.9846, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 2.1614839680468325, | |
| "learning_rate": 5.531248643935803e-05, | |
| "loss": 0.9172, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 2.120189858636912, | |
| "learning_rate": 5.5183385637329446e-05, | |
| "loss": 0.9487, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 2.0398912365174957, | |
| "learning_rate": 5.505424989374157e-05, | |
| "loss": 0.9022, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 2.3133595550614454, | |
| "learning_rate": 5.4925080079107326e-05, | |
| "loss": 1.0098, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 1.8814903929708948, | |
| "learning_rate": 5.4795877064169345e-05, | |
| "loss": 0.8668, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 2.153833002237324, | |
| "learning_rate": 5.466664171989402e-05, | |
| "loss": 0.9755, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 2.1606050704930273, | |
| "learning_rate": 5.453737491746572e-05, | |
| "loss": 0.911, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 2.336110910443884, | |
| "learning_rate": 5.4408077528280845e-05, | |
| "loss": 1.2218, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 2.1962170986292398, | |
| "learning_rate": 5.427875042394199e-05, | |
| "loss": 0.9806, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 1.9379444323949253, | |
| "learning_rate": 5.414939447625208e-05, | |
| "loss": 0.8914, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 2.100994717309111, | |
| "learning_rate": 5.402001055720844e-05, | |
| "loss": 0.9924, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 2.6568438499740825, | |
| "learning_rate": 5.389059953899699e-05, | |
| "loss": 0.9826, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 1.991812400079013, | |
| "learning_rate": 5.376116229398631e-05, | |
| "loss": 0.9798, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 2.468642248027418, | |
| "learning_rate": 5.3631699694721774e-05, | |
| "loss": 0.912, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 2.070023765778072, | |
| "learning_rate": 5.350221261391966e-05, | |
| "loss": 0.9243, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 2.0803911188980835, | |
| "learning_rate": 5.33727019244613e-05, | |
| "loss": 0.9768, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 2.7349637263970012, | |
| "learning_rate": 5.324316849938715e-05, | |
| "loss": 1.0446, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 1.968524608852828, | |
| "learning_rate": 5.311361321189098e-05, | |
| "loss": 0.8527, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 2.1488599620853983, | |
| "learning_rate": 5.298403693531385e-05, | |
| "loss": 0.89, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 2.1546767098536073, | |
| "learning_rate": 5.2854440543138406e-05, | |
| "loss": 0.9992, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 2.0251580669714455, | |
| "learning_rate": 5.2724824908982815e-05, | |
| "loss": 0.8361, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 1.856050365868989, | |
| "learning_rate": 5.2595190906594996e-05, | |
| "loss": 0.8315, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.0152486049532143, | |
| "learning_rate": 5.246553940984668e-05, | |
| "loss": 0.8762, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.02704690154883, | |
| "learning_rate": 5.233587129272752e-05, | |
| "loss": 0.8419, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.8992003893809617, | |
| "learning_rate": 5.220618742933923e-05, | |
| "loss": 0.6959, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 2.118677689268931, | |
| "learning_rate": 5.207648869388966e-05, | |
| "loss": 0.7975, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 2.06788920167534, | |
| "learning_rate": 5.1946775960686887e-05, | |
| "loss": 0.65, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 2.2268943126063583, | |
| "learning_rate": 5.18170501041334e-05, | |
| "loss": 0.7543, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 2.193900397607207, | |
| "learning_rate": 5.168731199872012e-05, | |
| "loss": 0.7416, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 1.8734959320983164, | |
| "learning_rate": 5.155756251902053e-05, | |
| "loss": 0.6924, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 2.6997559978819337, | |
| "learning_rate": 5.142780253968481e-05, | |
| "loss": 0.7713, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 2.1989988315886055, | |
| "learning_rate": 5.129803293543392e-05, | |
| "loss": 0.7538, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 2.2979031738619433, | |
| "learning_rate": 5.1168254581053675e-05, | |
| "loss": 0.741, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 2.882997880304532, | |
| "learning_rate": 5.103846835138891e-05, | |
| "loss": 0.7797, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 2.2160752742329795, | |
| "learning_rate": 5.0908675121337525e-05, | |
| "loss": 0.6394, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 2.45562608714171, | |
| "learning_rate": 5.07788757658446e-05, | |
| "loss": 0.6232, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 2.4883925008705083, | |
| "learning_rate": 5.064907115989655e-05, | |
| "loss": 0.7472, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 2.0146997329035825, | |
| "learning_rate": 5.051926217851513e-05, | |
| "loss": 0.7265, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 2.1393534296121306, | |
| "learning_rate": 5.038944969675165e-05, | |
| "loss": 0.7298, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 2.835644321610313, | |
| "learning_rate": 5.025963458968095e-05, | |
| "loss": 0.8086, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 2.093797069377126, | |
| "learning_rate": 5.01298177323956e-05, | |
| "loss": 0.6875, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 2.2546841149384513, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6654, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 2.436692330380529, | |
| "learning_rate": 4.9870182267604406e-05, | |
| "loss": 0.7452, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 2.1594944882219154, | |
| "learning_rate": 4.974036541031907e-05, | |
| "loss": 0.6145, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 1.912406969668213, | |
| "learning_rate": 4.961055030324836e-05, | |
| "loss": 0.7441, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 2.4471931165946215, | |
| "learning_rate": 4.948073782148487e-05, | |
| "loss": 0.7242, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 2.589519589038129, | |
| "learning_rate": 4.9350928840103464e-05, | |
| "loss": 0.7709, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 2.3449056187209534, | |
| "learning_rate": 4.9221124234155415e-05, | |
| "loss": 0.7106, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 2.805227606393924, | |
| "learning_rate": 4.909132487866248e-05, | |
| "loss": 0.7273, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 2.3409170876252117, | |
| "learning_rate": 4.8961531648611095e-05, | |
| "loss": 0.8471, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 2.477116428537875, | |
| "learning_rate": 4.883174541894633e-05, | |
| "loss": 0.7897, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 2.0926407030635246, | |
| "learning_rate": 4.8701967064566095e-05, | |
| "loss": 0.736, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 1.865516970056475, | |
| "learning_rate": 4.85721974603152e-05, | |
| "loss": 0.6193, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 1.9457304071194101, | |
| "learning_rate": 4.844243748097948e-05, | |
| "loss": 0.6391, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 2.282584735103655, | |
| "learning_rate": 4.831268800127989e-05, | |
| "loss": 0.722, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 2.548936276258133, | |
| "learning_rate": 4.8182949895866616e-05, | |
| "loss": 0.621, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 2.239461513092809, | |
| "learning_rate": 4.8053224039313125e-05, | |
| "loss": 0.7313, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 2.2103196138399857, | |
| "learning_rate": 4.7923511306110354e-05, | |
| "loss": 0.8155, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 2.4727327700188435, | |
| "learning_rate": 4.779381257066078e-05, | |
| "loss": 0.7856, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 2.140548717498076, | |
| "learning_rate": 4.766412870727249e-05, | |
| "loss": 0.692, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 2.2603245532398115, | |
| "learning_rate": 4.753446059015334e-05, | |
| "loss": 0.7835, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 2.106369804198163, | |
| "learning_rate": 4.740480909340502e-05, | |
| "loss": 0.7118, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 2.6374917789933527, | |
| "learning_rate": 4.727517509101719e-05, | |
| "loss": 0.8211, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 2.004032135600095, | |
| "learning_rate": 4.71455594568616e-05, | |
| "loss": 0.679, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 2.4329743207201338, | |
| "learning_rate": 4.7015963064686155e-05, | |
| "loss": 0.7338, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 2.419605130989746, | |
| "learning_rate": 4.688638678810905e-05, | |
| "loss": 0.8148, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 2.1035811893620817, | |
| "learning_rate": 4.675683150061285e-05, | |
| "loss": 0.6361, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 2.010576143639259, | |
| "learning_rate": 4.662729807553871e-05, | |
| "loss": 0.7664, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 2.4289060573335832, | |
| "learning_rate": 4.649778738608036e-05, | |
| "loss": 0.8349, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 2.1833824476159704, | |
| "learning_rate": 4.636830030527825e-05, | |
| "loss": 0.7255, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 2.02475067872114, | |
| "learning_rate": 4.623883770601369e-05, | |
| "loss": 0.6892, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 1.9859391895889293, | |
| "learning_rate": 4.6109400461003004e-05, | |
| "loss": 0.7444, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 2.0970611312425835, | |
| "learning_rate": 4.597998944279156e-05, | |
| "loss": 0.7623, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 1.9833194210797214, | |
| "learning_rate": 4.5850605523747934e-05, | |
| "loss": 0.6836, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 1.9798199365759683, | |
| "learning_rate": 4.5721249576058027e-05, | |
| "loss": 0.632, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 2.0842941934660963, | |
| "learning_rate": 4.559192247171916e-05, | |
| "loss": 0.7344, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 1.8340259954594293, | |
| "learning_rate": 4.546262508253429e-05, | |
| "loss": 0.7121, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "grad_norm": 2.2468709518264833, | |
| "learning_rate": 4.533335828010599e-05, | |
| "loss": 0.6639, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "grad_norm": 2.56185669237772, | |
| "learning_rate": 4.520412293583068e-05, | |
| "loss": 0.7308, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 1.9322345911512993, | |
| "learning_rate": 4.507491992089268e-05, | |
| "loss": 0.6255, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 3.0808262548087035, | |
| "learning_rate": 4.494575010625844e-05, | |
| "loss": 0.7357, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 2.10449650834673, | |
| "learning_rate": 4.481661436267057e-05, | |
| "loss": 0.6401, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 2.2957409307271432, | |
| "learning_rate": 4.468751356064199e-05, | |
| "loss": 0.6895, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 2.2970436775943743, | |
| "learning_rate": 4.455844857045009e-05, | |
| "loss": 0.6296, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 1.8615972387021242, | |
| "learning_rate": 4.442942026213084e-05, | |
| "loss": 0.5905, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 2.1688657277997136, | |
| "learning_rate": 4.4300429505472976e-05, | |
| "loss": 0.7506, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 2.5888436590177166, | |
| "learning_rate": 4.417147717001205e-05, | |
| "loss": 0.7552, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 2.607687618189682, | |
| "learning_rate": 4.404256412502462e-05, | |
| "loss": 0.8212, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 2.100252774809288, | |
| "learning_rate": 4.3913691239522406e-05, | |
| "loss": 0.7902, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 2.5160351863170285, | |
| "learning_rate": 4.37848593822464e-05, | |
| "loss": 0.6859, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 2.3194718244384034, | |
| "learning_rate": 4.365606942166102e-05, | |
| "loss": 0.6765, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 2.062386297832033, | |
| "learning_rate": 4.3527322225948234e-05, | |
| "loss": 0.7386, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 1.9956370090901974, | |
| "learning_rate": 4.339861866300177e-05, | |
| "loss": 0.6714, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 1.9976058474345035, | |
| "learning_rate": 4.326995960042119e-05, | |
| "loss": 0.6099, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 2.125791856628619, | |
| "learning_rate": 4.314134590550608e-05, | |
| "loss": 0.7361, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 2.276748391695722, | |
| "learning_rate": 4.3012778445250244e-05, | |
| "loss": 0.678, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 2.314882360822672, | |
| "learning_rate": 4.288425808633575e-05, | |
| "loss": 0.7869, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 2.237000606507243, | |
| "learning_rate": 4.2755785695127184e-05, | |
| "loss": 0.7641, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 2.0669481447014917, | |
| "learning_rate": 4.262736213766578e-05, | |
| "loss": 0.6835, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 2.2317746152113846, | |
| "learning_rate": 4.249898827966358e-05, | |
| "loss": 0.6147, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 2.2319467759544778, | |
| "learning_rate": 4.2370664986497605e-05, | |
| "loss": 0.7463, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 2.392563091136776, | |
| "learning_rate": 4.2242393123203986e-05, | |
| "loss": 0.7206, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 2.0059939737465156, | |
| "learning_rate": 4.211417355447217e-05, | |
| "loss": 0.6668, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 2.161188995058832, | |
| "learning_rate": 4.1986007144639136e-05, | |
| "loss": 0.7027, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 1.83128553589811, | |
| "learning_rate": 4.185789475768344e-05, | |
| "loss": 0.6377, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 2.282597346472662, | |
| "learning_rate": 4.1729837257219514e-05, | |
| "loss": 0.7318, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 2.035444881540161, | |
| "learning_rate": 4.1601835506491755e-05, | |
| "loss": 0.7077, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 2.0501096703297765, | |
| "learning_rate": 4.147389036836881e-05, | |
| "loss": 0.5982, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 1.965000717839894, | |
| "learning_rate": 4.134600270533765e-05, | |
| "loss": 0.6734, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 2.1223940821917266, | |
| "learning_rate": 4.12181733794978e-05, | |
| "loss": 0.6099, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 1.9865841452591566, | |
| "learning_rate": 4.109040325255553e-05, | |
| "loss": 0.5848, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 2.142992583525927, | |
| "learning_rate": 4.0962693185818104e-05, | |
| "loss": 0.6249, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 2.4280078740583853, | |
| "learning_rate": 4.0835044040187836e-05, | |
| "loss": 0.7771, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 1.9393783040131458, | |
| "learning_rate": 4.0707456676156406e-05, | |
| "loss": 0.5472, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 2.164660768743472, | |
| "learning_rate": 4.057993195379903e-05, | |
| "loss": 0.7352, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 1.9869557346679565, | |
| "learning_rate": 4.04524707327686e-05, | |
| "loss": 0.675, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 2.4071174226037506, | |
| "learning_rate": 4.032507387229002e-05, | |
| "loss": 0.7416, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 2.450927058089603, | |
| "learning_rate": 4.0197742231154286e-05, | |
| "loss": 0.6837, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 2.39247371385272, | |
| "learning_rate": 4.007047666771274e-05, | |
| "loss": 0.7581, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 2.0059718846434356, | |
| "learning_rate": 3.994327803987129e-05, | |
| "loss": 0.6258, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 2.0456883739444858, | |
| "learning_rate": 3.981614720508468e-05, | |
| "loss": 0.617, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 2.4060973662885026, | |
| "learning_rate": 3.96890850203506e-05, | |
| "loss": 0.7118, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 2.303058741642879, | |
| "learning_rate": 3.956209234220397e-05, | |
| "loss": 0.6896, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 2.189381915231591, | |
| "learning_rate": 3.943517002671118e-05, | |
| "loss": 0.7205, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 2.1375043120815023, | |
| "learning_rate": 3.930831892946428e-05, | |
| "loss": 0.6661, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 2.193712287502315, | |
| "learning_rate": 3.9181539905575275e-05, | |
| "loss": 0.7194, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 2.3537739174795593, | |
| "learning_rate": 3.9054833809670264e-05, | |
| "loss": 0.6943, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 1.988797286388587, | |
| "learning_rate": 3.892820149588378e-05, | |
| "loss": 0.7331, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 2.0064291676940376, | |
| "learning_rate": 3.880164381785294e-05, | |
| "loss": 0.636, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 2.2012243909282065, | |
| "learning_rate": 3.8675161628711776e-05, | |
| "loss": 0.6783, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 2.0622187034225217, | |
| "learning_rate": 3.8548755781085445e-05, | |
| "loss": 0.7299, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 2.110083204488387, | |
| "learning_rate": 3.842242712708444e-05, | |
| "loss": 0.7061, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 1.9806746021456059, | |
| "learning_rate": 3.829617651829892e-05, | |
| "loss": 0.6697, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 2.430116437944247, | |
| "learning_rate": 3.8170004805792905e-05, | |
| "loss": 0.5439, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 2.1307704410419297, | |
| "learning_rate": 3.8043912840098644e-05, | |
| "loss": 0.6434, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 2.2590003480937333, | |
| "learning_rate": 3.791790147121072e-05, | |
| "loss": 0.7212, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 2.0964754102570438, | |
| "learning_rate": 3.779197154858044e-05, | |
| "loss": 0.6782, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 2.238770907528468, | |
| "learning_rate": 3.7666123921110076e-05, | |
| "loss": 0.5911, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 2.2812990640923014, | |
| "learning_rate": 3.7540359437147163e-05, | |
| "loss": 0.7862, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 2.1537187687437584, | |
| "learning_rate": 3.741467894447875e-05, | |
| "loss": 0.7089, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 2.4652366490821414, | |
| "learning_rate": 3.728908329032567e-05, | |
| "loss": 0.6799, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 2.2500359217712504, | |
| "learning_rate": 3.716357332133687e-05, | |
| "loss": 0.6881, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 2.01882178877652, | |
| "learning_rate": 3.703814988358366e-05, | |
| "loss": 0.6626, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 2.393760914695007, | |
| "learning_rate": 3.691281382255413e-05, | |
| "loss": 0.742, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 2.076745734235737, | |
| "learning_rate": 3.678756598314725e-05, | |
| "loss": 0.6669, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 2.017635460385742, | |
| "learning_rate": 3.6662407209667335e-05, | |
| "loss": 0.6938, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 2.329463585492503, | |
| "learning_rate": 3.6537338345818274e-05, | |
| "loss": 0.5765, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 2.3504695521757353, | |
| "learning_rate": 3.6412360234697916e-05, | |
| "loss": 0.611, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 1.9911693665723682, | |
| "learning_rate": 3.6287473718792295e-05, | |
| "loss": 0.6292, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 2.12464200554563, | |
| "learning_rate": 3.616267963997001e-05, | |
| "loss": 0.5917, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 2.0636069815357545, | |
| "learning_rate": 3.603797883947651e-05, | |
| "loss": 0.666, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 2.2350418028455374, | |
| "learning_rate": 3.591337215792852e-05, | |
| "loss": 0.6555, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 2.0597303782371017, | |
| "learning_rate": 3.5788860435308225e-05, | |
| "loss": 0.6796, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 2.210711231029459, | |
| "learning_rate": 3.5664444510957715e-05, | |
| "loss": 0.7473, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 2.0635669106331997, | |
| "learning_rate": 3.554012522357331e-05, | |
| "loss": 0.7166, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 2.5600212673239873, | |
| "learning_rate": 3.5415903411199865e-05, | |
| "loss": 0.6629, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 2.564959735174983, | |
| "learning_rate": 3.5291779911225186e-05, | |
| "loss": 0.7182, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 2.0651962891449114, | |
| "learning_rate": 3.516775556037433e-05, | |
| "loss": 0.5532, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 1.9407913382413917, | |
| "learning_rate": 3.504383119470395e-05, | |
| "loss": 0.6525, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 2.211318876037454, | |
| "learning_rate": 3.492000764959674e-05, | |
| "loss": 0.6568, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 2.1837153635736715, | |
| "learning_rate": 3.4796285759755767e-05, | |
| "loss": 0.5812, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 2.1868732760151763, | |
| "learning_rate": 3.467266635919876e-05, | |
| "loss": 0.7444, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 2.1688479601387525, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.6943, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 2.3721571426472345, | |
| "learning_rate": 3.442573835854777e-05, | |
| "loss": 0.652, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 2.478417007067636, | |
| "learning_rate": 3.430243142301239e-05, | |
| "loss": 0.6494, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 2.3749113695547552, | |
| "learning_rate": 3.41792303058671e-05, | |
| "loss": 0.7334, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 2.369611846692306, | |
| "learning_rate": 3.4056135837619074e-05, | |
| "loss": 0.672, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 2.1693484467365254, | |
| "learning_rate": 3.39331488480566e-05, | |
| "loss": 0.6793, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 2.1010633960729574, | |
| "learning_rate": 3.3810270166243404e-05, | |
| "loss": 0.6972, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 2.2147785425675033, | |
| "learning_rate": 3.368750062051319e-05, | |
| "loss": 0.6428, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 2.173581879160362, | |
| "learning_rate": 3.356484103846389e-05, | |
| "loss": 0.708, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 1.9008372745605175, | |
| "learning_rate": 3.3442292246952186e-05, | |
| "loss": 0.5385, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 2.112019942567123, | |
| "learning_rate": 3.331985507208793e-05, | |
| "loss": 0.637, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 2.199591229490892, | |
| "learning_rate": 3.3197530339228487e-05, | |
| "loss": 0.7516, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 2.2089584377933513, | |
| "learning_rate": 3.307531887297338e-05, | |
| "loss": 0.7104, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 2.0369749274272215, | |
| "learning_rate": 3.295322149715848e-05, | |
| "loss": 0.5698, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 2.3605545680449285, | |
| "learning_rate": 3.283123903485059e-05, | |
| "loss": 0.6716, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 2.1809149939613737, | |
| "learning_rate": 3.2709372308341866e-05, | |
| "loss": 0.6313, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 2.0469760303849434, | |
| "learning_rate": 3.258762213914435e-05, | |
| "loss": 0.6836, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 2.3838779114236153, | |
| "learning_rate": 3.246598934798428e-05, | |
| "loss": 0.7282, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 2.0385753808522473, | |
| "learning_rate": 3.234447475479668e-05, | |
| "loss": 0.7013, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 2.117207725356933, | |
| "learning_rate": 3.222307917871977e-05, | |
| "loss": 0.5627, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 2.3634042403426827, | |
| "learning_rate": 3.2101803438089485e-05, | |
| "loss": 0.6385, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 2.09755716301372, | |
| "learning_rate": 3.198064835043396e-05, | |
| "loss": 0.6767, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 1.8931116392152203, | |
| "learning_rate": 3.1859614732467954e-05, | |
| "loss": 0.6645, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 2.257936268385384, | |
| "learning_rate": 3.1738703400087444e-05, | |
| "loss": 0.7328, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 2.1580267825441704, | |
| "learning_rate": 3.1617915168363996e-05, | |
| "loss": 0.6316, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 2.2050655069807186, | |
| "learning_rate": 3.149725085153944e-05, | |
| "loss": 0.6916, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 2.070569386524003, | |
| "learning_rate": 3.137671126302022e-05, | |
| "loss": 0.6311, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 2.0826516482412187, | |
| "learning_rate": 3.125629721537199e-05, | |
| "loss": 0.711, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 2.0933632049690867, | |
| "learning_rate": 3.1136009520314145e-05, | |
| "loss": 0.6323, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 1.6609569506949207, | |
| "learning_rate": 3.1015848988714305e-05, | |
| "loss": 0.6321, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 2.0549267991720668, | |
| "learning_rate": 3.0895816430582904e-05, | |
| "loss": 0.5915, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 2.207424630875684, | |
| "learning_rate": 3.077591265506764e-05, | |
| "loss": 0.6289, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 2.193270355371512, | |
| "learning_rate": 3.065613847044815e-05, | |
| "loss": 0.7403, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 2.1131431920915857, | |
| "learning_rate": 3.053649468413043e-05, | |
| "loss": 0.7265, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 2.783788265817973, | |
| "learning_rate": 3.041698210264149e-05, | |
| "loss": 0.6422, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 2.434472852039058, | |
| "learning_rate": 3.029760153162387e-05, | |
| "loss": 0.6835, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 1.9996265887527647, | |
| "learning_rate": 3.0178353775830216e-05, | |
| "loss": 0.6091, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 2.518329542583787, | |
| "learning_rate": 3.0059239639117836e-05, | |
| "loss": 0.6226, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 3.2146232312446887, | |
| "learning_rate": 2.9940259924443336e-05, | |
| "loss": 0.618, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 2.1061426372823675, | |
| "learning_rate": 2.9821415433857174e-05, | |
| "loss": 0.5277, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 2.1387534854830546, | |
| "learning_rate": 2.970270696849823e-05, | |
| "loss": 0.7032, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 2.97109207840636, | |
| "learning_rate": 2.958413532858843e-05, | |
| "loss": 0.6769, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 3.3197093097220973, | |
| "learning_rate": 2.9465701313427345e-05, | |
| "loss": 0.6824, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 1.9489958632157642, | |
| "learning_rate": 2.934740572138686e-05, | |
| "loss": 0.5847, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 2.1852942614668325, | |
| "learning_rate": 2.9229249349905684e-05, | |
| "loss": 0.6441, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 2.54416325761753, | |
| "learning_rate": 2.9111232995484037e-05, | |
| "loss": 0.6588, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 2.0813293747492274, | |
| "learning_rate": 2.8993357453678305e-05, | |
| "loss": 0.6586, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 2.204230025743336, | |
| "learning_rate": 2.887562351909562e-05, | |
| "loss": 0.6567, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 2.129187975631819, | |
| "learning_rate": 2.875803198538851e-05, | |
| "loss": 0.665, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 2.1855157141713177, | |
| "learning_rate": 2.864058364524964e-05, | |
| "loss": 0.6007, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 2.5831144031351805, | |
| "learning_rate": 2.852327929040634e-05, | |
| "loss": 0.727, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 2.201632351048249, | |
| "learning_rate": 2.840611971161533e-05, | |
| "loss": 0.7096, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 2.0653223924790494, | |
| "learning_rate": 2.828910569865738e-05, | |
| "loss": 0.6123, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 2.2812923953832076, | |
| "learning_rate": 2.8172238040332045e-05, | |
| "loss": 0.6177, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 2.1755173279612574, | |
| "learning_rate": 2.805551752445222e-05, | |
| "loss": 0.5508, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 2.612389594839949, | |
| "learning_rate": 2.7938944937838923e-05, | |
| "loss": 0.7081, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 2.441961279311164, | |
| "learning_rate": 2.7822521066316026e-05, | |
| "loss": 0.6661, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 2.5407672693469525, | |
| "learning_rate": 2.770624669470484e-05, | |
| "loss": 0.6925, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 2.300667989756706, | |
| "learning_rate": 2.759012260681889e-05, | |
| "loss": 0.6204, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 2.0056661633450372, | |
| "learning_rate": 2.7474149585458662e-05, | |
| "loss": 0.5709, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 2.1508527299772235, | |
| "learning_rate": 2.7358328412406253e-05, | |
| "loss": 0.6595, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 2.420434463751638, | |
| "learning_rate": 2.7242659868420194e-05, | |
| "loss": 0.6097, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 2.1692706917556146, | |
| "learning_rate": 2.712714473323008e-05, | |
| "loss": 0.6637, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 2.2317285112833214, | |
| "learning_rate": 2.7011783785531387e-05, | |
| "loss": 0.5152, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 2.4835258013309582, | |
| "learning_rate": 2.689657780298019e-05, | |
| "loss": 0.6462, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 2.008194903895354, | |
| "learning_rate": 2.678152756218797e-05, | |
| "loss": 0.5645, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 2.117621258923309, | |
| "learning_rate": 2.6666633838716314e-05, | |
| "loss": 0.6431, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 2.404672216794891, | |
| "learning_rate": 2.6551897407071713e-05, | |
| "loss": 0.714, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 2.7256270559990328, | |
| "learning_rate": 2.6437319040700347e-05, | |
| "loss": 0.7147, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.1681894338596712, | |
| "learning_rate": 2.632289951198285e-05, | |
| "loss": 0.5624, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.234934431261228, | |
| "learning_rate": 2.620863959222918e-05, | |
| "loss": 0.6534, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 1.6326816997072477, | |
| "learning_rate": 2.60945400516733e-05, | |
| "loss": 0.4432, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 1.9810106089462702, | |
| "learning_rate": 2.5980601659468058e-05, | |
| "loss": 0.4776, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 2.300335155888505, | |
| "learning_rate": 2.5866825183679982e-05, | |
| "loss": 0.5293, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 2.414648045080412, | |
| "learning_rate": 2.5753211391284172e-05, | |
| "loss": 0.4727, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 1.7790876943461056, | |
| "learning_rate": 2.5639761048158985e-05, | |
| "loss": 0.4424, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 1.84115083875142, | |
| "learning_rate": 2.552647491908101e-05, | |
| "loss": 0.5687, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 2.1116631314528425, | |
| "learning_rate": 2.5413353767719805e-05, | |
| "loss": 0.5255, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 2.505859132018095, | |
| "learning_rate": 2.5300398356632882e-05, | |
| "loss": 0.5369, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 2.0727434151563062, | |
| "learning_rate": 2.5187609447260417e-05, | |
| "loss": 0.5113, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 2.079695018714233, | |
| "learning_rate": 2.5074987799920223e-05, | |
| "loss": 0.5754, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 1.9571358074091139, | |
| "learning_rate": 2.4962534173802558e-05, | |
| "loss": 0.4756, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 2.1348135626297147, | |
| "learning_rate": 2.485024932696504e-05, | |
| "loss": 0.4258, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "grad_norm": 2.0616692704334487, | |
| "learning_rate": 2.4738134016327596e-05, | |
| "loss": 0.5165, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "grad_norm": 1.9547669475683698, | |
| "learning_rate": 2.4626188997667222e-05, | |
| "loss": 0.5104, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 2.045306060866488, | |
| "learning_rate": 2.4514415025612997e-05, | |
| "loss": 0.493, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 1.8430827358634285, | |
| "learning_rate": 2.4402812853640938e-05, | |
| "loss": 0.4327, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 2.4682602801503752, | |
| "learning_rate": 2.4291383234069003e-05, | |
| "loss": 0.5579, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "grad_norm": 2.044393895989231, | |
| "learning_rate": 2.418012691805191e-05, | |
| "loss": 0.5456, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "grad_norm": 1.8533727826833761, | |
| "learning_rate": 2.406904465557614e-05, | |
| "loss": 0.457, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 2.1374211896990314, | |
| "learning_rate": 2.3958137195454873e-05, | |
| "loss": 0.4545, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 2.4732501207928332, | |
| "learning_rate": 2.3847405285322898e-05, | |
| "loss": 0.5064, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 1.9115434501438884, | |
| "learning_rate": 2.37368496716317e-05, | |
| "loss": 0.4627, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 1.8747998484145278, | |
| "learning_rate": 2.3626471099644243e-05, | |
| "loss": 0.4819, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 2.5033155554597117, | |
| "learning_rate": 2.3516270313430083e-05, | |
| "loss": 0.5741, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 2.04963411735937, | |
| "learning_rate": 2.3406248055860292e-05, | |
| "loss": 0.5199, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 1.9495564999516717, | |
| "learning_rate": 2.32964050686025e-05, | |
| "loss": 0.5086, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 2.200660127761931, | |
| "learning_rate": 2.3186742092115844e-05, | |
| "loss": 0.5096, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 2.3489921731300196, | |
| "learning_rate": 2.307725986564598e-05, | |
| "loss": 0.4554, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 1.9797945328056459, | |
| "learning_rate": 2.296795912722014e-05, | |
| "loss": 0.4597, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "grad_norm": 2.262811898900508, | |
| "learning_rate": 2.28588406136421e-05, | |
| "loss": 0.4827, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "grad_norm": 1.846396086176893, | |
| "learning_rate": 2.2749905060487308e-05, | |
| "loss": 0.5482, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 2.046865942347122, | |
| "learning_rate": 2.264115320209781e-05, | |
| "loss": 0.541, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 1.9547867227244085, | |
| "learning_rate": 2.2532585771577362e-05, | |
| "loss": 0.5899, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 1.994670120187159, | |
| "learning_rate": 2.2424203500786474e-05, | |
| "loss": 0.4878, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 1.834520078454246, | |
| "learning_rate": 2.2316007120337517e-05, | |
| "loss": 0.4401, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 1.657069668133102, | |
| "learning_rate": 2.2207997359589728e-05, | |
| "loss": 0.401, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 2.2174438734919115, | |
| "learning_rate": 2.210017494664433e-05, | |
| "loss": 0.4723, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 2.0695899567780502, | |
| "learning_rate": 2.1992540608339624e-05, | |
| "loss": 0.5119, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 2.24562436093221, | |
| "learning_rate": 2.1885095070246114e-05, | |
| "loss": 0.5329, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 1.86235223704696, | |
| "learning_rate": 2.1777839056661554e-05, | |
| "loss": 0.4565, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 2.0972619356324875, | |
| "learning_rate": 2.167077329060611e-05, | |
| "loss": 0.4907, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 2.0083867329369856, | |
| "learning_rate": 2.1563898493817464e-05, | |
| "loss": 0.5133, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 2.1759905914178357, | |
| "learning_rate": 2.1457215386746017e-05, | |
| "loss": 0.5126, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 1.9306888576124372, | |
| "learning_rate": 2.1350724688549906e-05, | |
| "loss": 0.4405, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "grad_norm": 1.9982035551407271, | |
| "learning_rate": 2.124442711709025e-05, | |
| "loss": 0.5417, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "grad_norm": 1.9237161202785842, | |
| "learning_rate": 2.1138323388926318e-05, | |
| "loss": 0.4933, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "grad_norm": 2.225529832476932, | |
| "learning_rate": 2.103241421931063e-05, | |
| "loss": 0.5084, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "grad_norm": 1.9174115859902279, | |
| "learning_rate": 2.0926700322184195e-05, | |
| "loss": 0.554, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 2.0622818636892695, | |
| "learning_rate": 2.0821182410171636e-05, | |
| "loss": 0.48, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 2.0526022047360692, | |
| "learning_rate": 2.0715861194576508e-05, | |
| "loss": 0.5, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 1.814445279862053, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.4392, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 2.1252623704056806, | |
| "learning_rate": 2.0505811691217973e-05, | |
| "loss": 0.5603, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "grad_norm": 2.0134612884283056, | |
| "learning_rate": 2.040108481941274e-05, | |
| "loss": 0.4952, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "grad_norm": 2.1511716542102235, | |
| "learning_rate": 2.029655747593169e-05, | |
| "loss": 0.4685, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 1.9326878235595093, | |
| "learning_rate": 2.0192230365400833e-05, | |
| "loss": 0.5696, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 1.9439010201934361, | |
| "learning_rate": 2.008810419109638e-05, | |
| "loss": 0.4383, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "grad_norm": 2.0000864192563865, | |
| "learning_rate": 1.9984179654940078e-05, | |
| "loss": 0.4967, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "grad_norm": 1.959567583264176, | |
| "learning_rate": 1.9880457457494356e-05, | |
| "loss": 0.5063, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 1.8875732483924463, | |
| "learning_rate": 1.9776938297957687e-05, | |
| "loss": 0.4715, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 1.9894582015678914, | |
| "learning_rate": 1.9673622874159853e-05, | |
| "loss": 0.4627, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 1.823726250793184, | |
| "learning_rate": 1.9570511882557213e-05, | |
| "loss": 0.471, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 2.043781863739054, | |
| "learning_rate": 1.946760601822809e-05, | |
| "loss": 0.5498, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 1.7905397586258414, | |
| "learning_rate": 1.9364905974867985e-05, | |
| "loss": 0.4487, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 1.6534527703483726, | |
| "learning_rate": 1.926241244478496e-05, | |
| "loss": 0.3667, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 2.044292024016082, | |
| "learning_rate": 1.9160126118894932e-05, | |
| "loss": 0.4846, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 1.9830715158173426, | |
| "learning_rate": 1.90580476867171e-05, | |
| "loss": 0.4992, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 1.8772251283435037, | |
| "learning_rate": 1.8956177836369182e-05, | |
| "loss": 0.5737, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 1.9017310460150294, | |
| "learning_rate": 1.8854517254562854e-05, | |
| "loss": 0.4287, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 1.9692864793327678, | |
| "learning_rate": 1.8753066626599086e-05, | |
| "loss": 0.511, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 1.9242146110496645, | |
| "learning_rate": 1.8651826636363528e-05, | |
| "loss": 0.5187, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 2.0720790383322303, | |
| "learning_rate": 1.855079796632196e-05, | |
| "loss": 0.5519, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "grad_norm": 1.7679082982749694, | |
| "learning_rate": 1.844998129751558e-05, | |
| "loss": 0.4804, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "grad_norm": 1.9304243642619154, | |
| "learning_rate": 1.8349377309556486e-05, | |
| "loss": 0.4775, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 1.8701732120735095, | |
| "learning_rate": 1.8248986680623075e-05, | |
| "loss": 0.441, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 1.6984810886999355, | |
| "learning_rate": 1.814881008745552e-05, | |
| "loss": 0.404, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 1.882075585310639, | |
| "learning_rate": 1.8048848205351104e-05, | |
| "loss": 0.4854, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 1.9626246247141248, | |
| "learning_rate": 1.7949101708159754e-05, | |
| "loss": 0.3867, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 2.3019504138435423, | |
| "learning_rate": 1.784957126827948e-05, | |
| "loss": 0.5518, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 1.9375744227606633, | |
| "learning_rate": 1.7750257556651805e-05, | |
| "loss": 0.4208, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 2.0501785986237517, | |
| "learning_rate": 1.765116124275733e-05, | |
| "loss": 0.4843, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 1.8751017354762511, | |
| "learning_rate": 1.7552282994611103e-05, | |
| "loss": 0.4594, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 2.0108881707704893, | |
| "learning_rate": 1.745362347875821e-05, | |
| "loss": 0.4477, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 1.7984684530072164, | |
| "learning_rate": 1.7355183360269207e-05, | |
| "loss": 0.4544, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 1.8679920749277188, | |
| "learning_rate": 1.725696330273575e-05, | |
| "loss": 0.5151, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 1.7364097045110307, | |
| "learning_rate": 1.715896396826599e-05, | |
| "loss": 0.4237, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "grad_norm": 2.0156308619388774, | |
| "learning_rate": 1.706118601748018e-05, | |
| "loss": 0.4912, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "grad_norm": 1.7801016142385238, | |
| "learning_rate": 1.696363010950622e-05, | |
| "loss": 0.488, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "grad_norm": 2.0501960957232352, | |
| "learning_rate": 1.6866296901975177e-05, | |
| "loss": 0.5007, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "grad_norm": 1.9341617179477686, | |
| "learning_rate": 1.6769187051016933e-05, | |
| "loss": 0.4342, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 1.758022541585096, | |
| "learning_rate": 1.667230121125567e-05, | |
| "loss": 0.5322, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 2.073233318826544, | |
| "learning_rate": 1.6575640035805496e-05, | |
| "loss": 0.5558, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 1.906208484897761, | |
| "learning_rate": 1.647920417626603e-05, | |
| "loss": 0.4854, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 1.8351847589849273, | |
| "learning_rate": 1.638299428271807e-05, | |
| "loss": 0.4746, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 2.0434885140049306, | |
| "learning_rate": 1.6287011003719103e-05, | |
| "loss": 0.5984, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 1.9973035189702482, | |
| "learning_rate": 1.619125498629904e-05, | |
| "loss": 0.5286, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 1.9898044805452346, | |
| "learning_rate": 1.609572687595575e-05, | |
| "loss": 0.4942, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "grad_norm": 2.1927288151163014, | |
| "learning_rate": 1.6000427316650847e-05, | |
| "loss": 0.5357, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "grad_norm": 1.8432013466376262, | |
| "learning_rate": 1.5905356950805185e-05, | |
| "loss": 0.4305, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 1.9739110981267056, | |
| "learning_rate": 1.5810516419294652e-05, | |
| "loss": 0.4971, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 1.832157513674343, | |
| "learning_rate": 1.5715906361445797e-05, | |
| "loss": 0.4404, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "grad_norm": 2.1230593487651075, | |
| "learning_rate": 1.5621527415031518e-05, | |
| "loss": 0.6079, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "grad_norm": 2.3652898839456844, | |
| "learning_rate": 1.5527380216266812e-05, | |
| "loss": 0.4596, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 2.1855158329783793, | |
| "learning_rate": 1.5433465399804415e-05, | |
| "loss": 0.5082, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 2.031995290399084, | |
| "learning_rate": 1.5339783598730567e-05, | |
| "loss": 0.5205, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 1.8684737297963356, | |
| "learning_rate": 1.5246335444560717e-05, | |
| "loss": 0.4017, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 2.2025808379530987, | |
| "learning_rate": 1.5153121567235335e-05, | |
| "loss": 0.4787, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 1.8254695187636378, | |
| "learning_rate": 1.5060142595115572e-05, | |
| "loss": 0.457, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 2.146681369602577, | |
| "learning_rate": 1.4967399154979073e-05, | |
| "loss": 0.489, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "grad_norm": 1.9613490067554153, | |
| "learning_rate": 1.4874891872015734e-05, | |
| "loss": 0.5024, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "grad_norm": 1.9206330128596096, | |
| "learning_rate": 1.4782621369823557e-05, | |
| "loss": 0.4612, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 1.969009910012194, | |
| "learning_rate": 1.4690588270404315e-05, | |
| "loss": 0.4663, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 2.06968780028015, | |
| "learning_rate": 1.4598793194159466e-05, | |
| "loss": 0.4712, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 1.85781444183358, | |
| "learning_rate": 1.4507236759885956e-05, | |
| "loss": 0.4849, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 1.862804780529419, | |
| "learning_rate": 1.4415919584771998e-05, | |
| "loss": 0.408, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "grad_norm": 2.0203987739951144, | |
| "learning_rate": 1.432484228439293e-05, | |
| "loss": 0.4789, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "grad_norm": 2.0515654338208815, | |
| "learning_rate": 1.4234005472707151e-05, | |
| "loss": 0.4836, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "grad_norm": 1.8544927774995303, | |
| "learning_rate": 1.414340976205183e-05, | |
| "loss": 0.354, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 1.9740989436111676, | |
| "learning_rate": 1.4053055763138884e-05, | |
| "loss": 0.447, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 1.8589057384569714, | |
| "learning_rate": 1.3962944085050832e-05, | |
| "loss": 0.4364, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 1.5906115148360191, | |
| "learning_rate": 1.3873075335236729e-05, | |
| "loss": 0.4183, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 1.7536914709928637, | |
| "learning_rate": 1.378345011950798e-05, | |
| "loss": 0.4759, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 1.7828585339362397, | |
| "learning_rate": 1.369406904203433e-05, | |
| "loss": 0.4811, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 2.1848109750855524, | |
| "learning_rate": 1.3604932705339768e-05, | |
| "loss": 0.5203, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 2.047705525874548, | |
| "learning_rate": 1.35160417102985e-05, | |
| "loss": 0.4261, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 1.98609723819213, | |
| "learning_rate": 1.342739665613082e-05, | |
| "loss": 0.5035, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "grad_norm": 2.15151962930573, | |
| "learning_rate": 1.3338998140399145e-05, | |
| "loss": 0.4945, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "grad_norm": 1.9641556835393486, | |
| "learning_rate": 1.3250846759003948e-05, | |
| "loss": 0.5406, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 2.055942811292963, | |
| "learning_rate": 1.3162943106179749e-05, | |
| "loss": 0.4603, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 1.9783512836246524, | |
| "learning_rate": 1.3075287774491146e-05, | |
| "loss": 0.4794, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "grad_norm": 2.137509739452884, | |
| "learning_rate": 1.2987881354828752e-05, | |
| "loss": 0.4734, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "grad_norm": 1.8818523834950385, | |
| "learning_rate": 1.2900724436405255e-05, | |
| "loss": 0.4274, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 1.7500869426242696, | |
| "learning_rate": 1.2813817606751415e-05, | |
| "loss": 0.418, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 1.7799842699517143, | |
| "learning_rate": 1.2727161451712199e-05, | |
| "loss": 0.4233, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 1.8443236113999752, | |
| "learning_rate": 1.2640756555442684e-05, | |
| "loss": 0.4665, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 2.015207412246041, | |
| "learning_rate": 1.2554603500404227e-05, | |
| "loss": 0.5234, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "grad_norm": 1.8134292217947385, | |
| "learning_rate": 1.2468702867360483e-05, | |
| "loss": 0.4123, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "grad_norm": 1.854961434705403, | |
| "learning_rate": 1.2383055235373565e-05, | |
| "loss": 0.4787, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 1.8712364085149127, | |
| "learning_rate": 1.2297661181800036e-05, | |
| "loss": 0.3825, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 2.0055289484117687, | |
| "learning_rate": 1.2212521282287092e-05, | |
| "loss": 0.522, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 2.075414587473912, | |
| "learning_rate": 1.2127636110768647e-05, | |
| "loss": 0.5555, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 1.8946795115564716, | |
| "learning_rate": 1.2043006239461479e-05, | |
| "loss": 0.4712, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 1.904692938477857, | |
| "learning_rate": 1.1958632238861394e-05, | |
| "loss": 0.463, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 1.9560832171498352, | |
| "learning_rate": 1.187451467773933e-05, | |
| "loss": 0.5153, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 1.8950498904269404, | |
| "learning_rate": 1.1790654123137552e-05, | |
| "loss": 0.4224, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 1.8400724649644247, | |
| "learning_rate": 1.1707051140365815e-05, | |
| "loss": 0.4876, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 1.8816504773293643, | |
| "learning_rate": 1.1623706292997605e-05, | |
| "loss": 0.4219, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 1.7940036623276212, | |
| "learning_rate": 1.1540620142866265e-05, | |
| "loss": 0.4421, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 1.751218485739817, | |
| "learning_rate": 1.145779325006125e-05, | |
| "loss": 0.3982, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "grad_norm": 1.9810976616288671, | |
| "learning_rate": 1.1375226172924341e-05, | |
| "loss": 0.4366, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "grad_norm": 1.9883872381553116, | |
| "learning_rate": 1.1292919468045877e-05, | |
| "loss": 0.5206, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 1.8620225976698854, | |
| "learning_rate": 1.1210873690261048e-05, | |
| "loss": 0.4613, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 2.074989972752041, | |
| "learning_rate": 1.112908939264607e-05, | |
| "loss": 0.472, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "grad_norm": 1.8809326710520073, | |
| "learning_rate": 1.1047567126514519e-05, | |
| "loss": 0.4546, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "grad_norm": 1.9165025684516834, | |
| "learning_rate": 1.0966307441413598e-05, | |
| "loss": 0.4528, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 1.9181304746376673, | |
| "learning_rate": 1.0885310885120453e-05, | |
| "loss": 0.4868, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 1.7351002616936348, | |
| "learning_rate": 1.0804578003638438e-05, | |
| "loss": 0.4349, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 1.905057337243823, | |
| "learning_rate": 1.0724109341193461e-05, | |
| "loss": 0.4261, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 1.8645772558442577, | |
| "learning_rate": 1.0643905440230318e-05, | |
| "loss": 0.4795, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "grad_norm": 2.0699372296258476, | |
| "learning_rate": 1.0563966841408995e-05, | |
| "loss": 0.5004, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "grad_norm": 1.9386071384404275, | |
| "learning_rate": 1.0484294083601131e-05, | |
| "loss": 0.5371, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "grad_norm": 1.725364075379709, | |
| "learning_rate": 1.0404887703886251e-05, | |
| "loss": 0.476, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "grad_norm": 1.8403754077428023, | |
| "learning_rate": 1.0325748237548226e-05, | |
| "loss": 0.4728, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 2.028942839285731, | |
| "learning_rate": 1.0246876218071632e-05, | |
| "loss": 0.5218, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 2.131518474570134, | |
| "learning_rate": 1.01682721771382e-05, | |
| "loss": 0.4229, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 2.002233156797534, | |
| "learning_rate": 1.0089936644623177e-05, | |
| "loss": 0.5236, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.9003711201201666, | |
| "learning_rate": 1.0011870148591767e-05, | |
| "loss": 0.4078, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 1.9053760578392858, | |
| "learning_rate": 9.934073215295597e-06, | |
| "loss": 0.4979, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "grad_norm": 1.7429791977863358, | |
| "learning_rate": 9.85654636916914e-06, | |
| "loss": 0.4528, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "grad_norm": 1.8731548716421391, | |
| "learning_rate": 9.779290132826224e-06, | |
| "loss": 0.4297, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "grad_norm": 2.0213232197341933, | |
| "learning_rate": 9.70230502705644e-06, | |
| "loss": 0.469, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "grad_norm": 1.7895786568184158, | |
| "learning_rate": 9.625591570821685e-06, | |
| "loss": 0.3583, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "grad_norm": 1.8274220140488615, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.4544, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "grad_norm": 1.5707168143363093, | |
| "learning_rate": 9.472981673645303e-06, | |
| "loss": 0.4324, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 2.033961931843386, | |
| "learning_rate": 9.39708626145751e-06, | |
| "loss": 0.5063, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 1.7480455928523881, | |
| "learning_rate": 9.321464556305438e-06, | |
| "loss": 0.4762, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 1.8617881244451975, | |
| "learning_rate": 9.246117067960231e-06, | |
| "loss": 0.4917, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 1.9089581634285688, | |
| "learning_rate": 9.171044304344472e-06, | |
| "loss": 0.4781, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 1.9554342666200304, | |
| "learning_rate": 9.096246771528837e-06, | |
| "loss": 0.4924, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 1.9323888542980714, | |
| "learning_rate": 9.021724973728635e-06, | |
| "loss": 0.5014, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "grad_norm": 1.9994328015585918, | |
| "learning_rate": 8.947479413300441e-06, | |
| "loss": 0.5364, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "grad_norm": 1.859257068217401, | |
| "learning_rate": 8.873510590738687e-06, | |
| "loss": 0.4641, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "grad_norm": 2.282293743278961, | |
| "learning_rate": 8.799819004672283e-06, | |
| "loss": 0.4939, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "grad_norm": 1.918571839911202, | |
| "learning_rate": 8.7264051518613e-06, | |
| "loss": 0.5155, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 1.837000460078746, | |
| "learning_rate": 8.65326952719357e-06, | |
| "loss": 0.4258, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 1.902093769062293, | |
| "learning_rate": 8.580412623681362e-06, | |
| "loss": 0.5096, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 1.8484936072949363, | |
| "learning_rate": 8.50783493245807e-06, | |
| "loss": 0.4556, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 2.105720733687499, | |
| "learning_rate": 8.435536942774924e-06, | |
| "loss": 0.5139, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 1.9753646345240035, | |
| "learning_rate": 8.363519141997644e-06, | |
| "loss": 0.4935, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 1.9863002159004608, | |
| "learning_rate": 8.291782015603179e-06, | |
| "loss": 0.543, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "grad_norm": 1.8557042986416377, | |
| "learning_rate": 8.22032604717642e-06, | |
| "loss": 0.4915, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "grad_norm": 1.8579045586565002, | |
| "learning_rate": 8.149151718406989e-06, | |
| "loss": 0.444, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "grad_norm": 1.8221042346582486, | |
| "learning_rate": 8.078259509085929e-06, | |
| "loss": 0.4429, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "grad_norm": 1.7497718645112934, | |
| "learning_rate": 8.007649897102498e-06, | |
| "loss": 0.3797, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "grad_norm": 1.7591870833787677, | |
| "learning_rate": 7.937323358440935e-06, | |
| "loss": 0.4167, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 1.8645287154393613, | |
| "learning_rate": 7.867280367177309e-06, | |
| "loss": 0.4633, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 1.8573688730040008, | |
| "learning_rate": 7.79752139547622e-06, | |
| "loss": 0.419, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 1.8568137869366368, | |
| "learning_rate": 7.728046913587711e-06, | |
| "loss": 0.3978, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 1.8640782821387256, | |
| "learning_rate": 7.658857389844037e-06, | |
| "loss": 0.4359, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 2.0389647850268484, | |
| "learning_rate": 7.589953290656532e-06, | |
| "loss": 0.5372, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 1.741227374435571, | |
| "learning_rate": 7.521335080512487e-06, | |
| "loss": 0.4477, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 1.7994841787007152, | |
| "learning_rate": 7.453003221971971e-06, | |
| "loss": 0.4079, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 1.852370172408188, | |
| "learning_rate": 7.384958175664746e-06, | |
| "loss": 0.462, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "grad_norm": 2.0292787879013114, | |
| "learning_rate": 7.317200400287139e-06, | |
| "loss": 0.5174, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "grad_norm": 1.8537064474120468, | |
| "learning_rate": 7.249730352598999e-06, | |
| "loss": 0.446, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "grad_norm": 1.8708672410207159, | |
| "learning_rate": 7.182548487420554e-06, | |
| "loss": 0.396, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "grad_norm": 1.934112659748951, | |
| "learning_rate": 7.115655257629372e-06, | |
| "loss": 0.5196, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.8672296319802737, | |
| "learning_rate": 7.049051114157329e-06, | |
| "loss": 0.4432, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.9466607279339219, | |
| "learning_rate": 6.982736505987519e-06, | |
| "loss": 0.489, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 1.5642489823618089, | |
| "learning_rate": 6.916711880151306e-06, | |
| "loss": 0.3434, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 1.6544244336083729, | |
| "learning_rate": 6.850977681725224e-06, | |
| "loss": 0.3648, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 1.6091246857033576, | |
| "learning_rate": 6.785534353828032e-06, | |
| "loss": 0.3296, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 1.58159162886627, | |
| "learning_rate": 6.720382337617692e-06, | |
| "loss": 0.3579, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 1.7835298243260125, | |
| "learning_rate": 6.65552207228845e-06, | |
| "loss": 0.4094, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 1.8751666734426788, | |
| "learning_rate": 6.590953995067811e-06, | |
| "loss": 0.4853, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "grad_norm": 1.7193683005589295, | |
| "learning_rate": 6.5266785412136265e-06, | |
| "loss": 0.363, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "grad_norm": 1.959499918108107, | |
| "learning_rate": 6.462696144011149e-06, | |
| "loss": 0.422, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "grad_norm": 1.7348741245634662, | |
| "learning_rate": 6.399007234770115e-06, | |
| "loss": 0.4053, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "grad_norm": 1.8293973697262131, | |
| "learning_rate": 6.33561224282187e-06, | |
| "loss": 0.4233, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "grad_norm": 1.8120673620233005, | |
| "learning_rate": 6.272511595516401e-06, | |
| "loss": 0.4405, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "grad_norm": 1.585746078257199, | |
| "learning_rate": 6.209705718219511e-06, | |
| "loss": 0.3371, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "grad_norm": 1.7731099026667199, | |
| "learning_rate": 6.1471950343099384e-06, | |
| "loss": 0.4234, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "grad_norm": 1.7687102987740553, | |
| "learning_rate": 6.084979965176524e-06, | |
| "loss": 0.425, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "grad_norm": 1.695471243288155, | |
| "learning_rate": 6.023060930215319e-06, | |
| "loss": 0.3848, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "grad_norm": 1.590849688229686, | |
| "learning_rate": 5.9614383468267924e-06, | |
| "loss": 0.3582, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "grad_norm": 1.555839723236568, | |
| "learning_rate": 5.900112630413018e-06, | |
| "loss": 0.3316, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "grad_norm": 1.7109121164072945, | |
| "learning_rate": 5.839084194374894e-06, | |
| "loss": 0.3743, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "grad_norm": 1.9484290161070803, | |
| "learning_rate": 5.778353450109286e-06, | |
| "loss": 0.4515, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 1.806656087806492, | |
| "learning_rate": 5.717920807006333e-06, | |
| "loss": 0.4769, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 1.7261025660346443, | |
| "learning_rate": 5.65778667244663e-06, | |
| "loss": 0.4145, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "grad_norm": 1.839203712881466, | |
| "learning_rate": 5.597951451798511e-06, | |
| "loss": 0.3771, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "grad_norm": 1.7103794991208325, | |
| "learning_rate": 5.538415548415338e-06, | |
| "loss": 0.4228, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "grad_norm": 1.8724044133808104, | |
| "learning_rate": 5.479179363632719e-06, | |
| "loss": 0.501, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "grad_norm": 1.854140002632786, | |
| "learning_rate": 5.420243296765853e-06, | |
| "loss": 0.4423, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "grad_norm": 1.7498537806158494, | |
| "learning_rate": 5.361607745106817e-06, | |
| "loss": 0.3795, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "grad_norm": 1.7421303199372071, | |
| "learning_rate": 5.303273103921918e-06, | |
| "loss": 0.4337, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "grad_norm": 1.6004980168874006, | |
| "learning_rate": 5.245239766448973e-06, | |
| "loss": 0.3646, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "grad_norm": 1.729915801446613, | |
| "learning_rate": 5.187508123894702e-06, | |
| "loss": 0.3886, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 5.15, | |
| "grad_norm": 1.705760976326433, | |
| "learning_rate": 5.13007856543209e-06, | |
| "loss": 0.4236, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 5.15, | |
| "grad_norm": 1.8697529919543623, | |
| "learning_rate": 5.072951478197724e-06, | |
| "loss": 0.4194, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "grad_norm": 1.8438018060011427, | |
| "learning_rate": 5.01612724728926e-06, | |
| "loss": 0.4439, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "grad_norm": 1.8528102678660445, | |
| "learning_rate": 4.959606255762728e-06, | |
| "loss": 0.46, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "grad_norm": 1.787182779423874, | |
| "learning_rate": 4.903388884630017e-06, | |
| "loss": 0.3479, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "grad_norm": 1.8059845910359302, | |
| "learning_rate": 4.847475512856315e-06, | |
| "loss": 0.4051, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "grad_norm": 1.6264223615390982, | |
| "learning_rate": 4.7918665173574905e-06, | |
| "loss": 0.3657, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "grad_norm": 1.585841860935979, | |
| "learning_rate": 4.7365622729976244e-06, | |
| "loss": 0.3659, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "grad_norm": 1.7836098994050769, | |
| "learning_rate": 4.681563152586416e-06, | |
| "loss": 0.4432, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "grad_norm": 1.8565575327503716, | |
| "learning_rate": 4.626869526876737e-06, | |
| "loss": 0.3796, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "grad_norm": 1.5395976746701239, | |
| "learning_rate": 4.572481764562081e-06, | |
| "loss": 0.3226, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 1.900174949075142, | |
| "learning_rate": 4.5184002322740785e-06, | |
| "loss": 0.4631, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 1.9983865732380948, | |
| "learning_rate": 4.464625294580083e-06, | |
| "loss": 0.4455, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 1.885862857110732, | |
| "learning_rate": 4.411157313980646e-06, | |
| "loss": 0.4063, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 1.9026668025471913, | |
| "learning_rate": 4.357996650907087e-06, | |
| "loss": 0.4331, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "grad_norm": 1.6778221783535614, | |
| "learning_rate": 4.305143663719113e-06, | |
| "loss": 0.3552, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "grad_norm": 1.579157722925671, | |
| "learning_rate": 4.252598708702343e-06, | |
| "loss": 0.3509, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 5.23, | |
| "grad_norm": 1.5708052868916198, | |
| "learning_rate": 4.200362140065933e-06, | |
| "loss": 0.3754, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 5.23, | |
| "grad_norm": 1.7000917349659141, | |
| "learning_rate": 4.148434309940186e-06, | |
| "loss": 0.3952, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 5.24, | |
| "grad_norm": 1.5796222145472663, | |
| "learning_rate": 4.0968155683741735e-06, | |
| "loss": 0.3575, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 5.24, | |
| "grad_norm": 1.8603544933214582, | |
| "learning_rate": 4.045506263333398e-06, | |
| "loss": 0.3997, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 5.25, | |
| "grad_norm": 1.6491719163529657, | |
| "learning_rate": 3.994506740697407e-06, | |
| "loss": 0.3206, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 5.25, | |
| "grad_norm": 1.7192642126800726, | |
| "learning_rate": 3.9438173442575e-06, | |
| "loss": 0.3873, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 5.26, | |
| "grad_norm": 1.7156715766602297, | |
| "learning_rate": 3.893438415714368e-06, | |
| "loss": 0.3953, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 5.26, | |
| "grad_norm": 1.9833514179934424, | |
| "learning_rate": 3.843370294675852e-06, | |
| "loss": 0.3848, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "grad_norm": 1.7686276263693377, | |
| "learning_rate": 3.7936133186546042e-06, | |
| "loss": 0.376, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "grad_norm": 1.7863193788527514, | |
| "learning_rate": 3.744167823065814e-06, | |
| "loss": 0.416, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 5.28, | |
| "grad_norm": 2.08881115185044, | |
| "learning_rate": 3.695034141224968e-06, | |
| "loss": 0.486, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 5.28, | |
| "grad_norm": 1.6747322757908776, | |
| "learning_rate": 3.646212604345606e-06, | |
| "loss": 0.3304, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.28, | |
| "grad_norm": 1.6153433561865993, | |
| "learning_rate": 3.5977035415370675e-06, | |
| "loss": 0.383, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "grad_norm": 1.948843583231187, | |
| "learning_rate": 3.5495072798022777e-06, | |
| "loss": 0.3636, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "grad_norm": 1.9757277031593108, | |
| "learning_rate": 3.501624144035559e-06, | |
| "loss": 0.4764, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 1.8195223564786063, | |
| "learning_rate": 3.4540544570204113e-06, | |
| "loss": 0.4221, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 1.991210186667102, | |
| "learning_rate": 3.406798539427386e-06, | |
| "loss": 0.4107, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "grad_norm": 1.7365656987884344, | |
| "learning_rate": 3.3598567098118648e-06, | |
| "loss": 0.3886, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "grad_norm": 1.7371829352919934, | |
| "learning_rate": 3.313229284611963e-06, | |
| "loss": 0.2957, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 5.32, | |
| "grad_norm": 1.7329836065091726, | |
| "learning_rate": 3.2669165781463396e-06, | |
| "loss": 0.375, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 5.32, | |
| "grad_norm": 1.7026642940376162, | |
| "learning_rate": 3.2209189026121635e-06, | |
| "loss": 0.4093, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "grad_norm": 1.826103900885128, | |
| "learning_rate": 3.1752365680829175e-06, | |
| "loss": 0.4239, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "grad_norm": 1.8729544425655098, | |
| "learning_rate": 3.12986988250637e-06, | |
| "loss": 0.5039, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "grad_norm": 1.910235871230281, | |
| "learning_rate": 3.084819151702456e-06, | |
| "loss": 0.4969, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "grad_norm": 2.30947631905642, | |
| "learning_rate": 3.040084679361255e-06, | |
| "loss": 0.3673, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "grad_norm": 1.9003867763006077, | |
| "learning_rate": 2.995666767040939e-06, | |
| "loss": 0.3869, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "grad_norm": 1.7270574102910248, | |
| "learning_rate": 2.9515657141656993e-06, | |
| "loss": 0.3954, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "grad_norm": 1.7454832170861654, | |
| "learning_rate": 2.9077818180237693e-06, | |
| "loss": 0.4273, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "grad_norm": 1.8225971333389457, | |
| "learning_rate": 2.8643153737654027e-06, | |
| "loss": 0.3851, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 5.37, | |
| "grad_norm": 1.8766488243303348, | |
| "learning_rate": 2.8211666744009047e-06, | |
| "loss": 0.4408, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 5.37, | |
| "grad_norm": 1.7552382029202105, | |
| "learning_rate": 2.778336010798616e-06, | |
| "loss": 0.4075, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 5.38, | |
| "grad_norm": 1.731668817579114, | |
| "learning_rate": 2.735823671682991e-06, | |
| "loss": 0.4126, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 5.38, | |
| "grad_norm": 1.8146628409967658, | |
| "learning_rate": 2.6936299436326305e-06, | |
| "loss": 0.3215, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 5.39, | |
| "grad_norm": 1.6056930864525185, | |
| "learning_rate": 2.651755111078358e-06, | |
| "loss": 0.3629, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 5.39, | |
| "grad_norm": 1.6078320455406179, | |
| "learning_rate": 2.61019945630131e-06, | |
| "loss": 0.4298, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 1.834231662377504, | |
| "learning_rate": 2.5689632594310065e-06, | |
| "loss": 0.4749, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 1.7500523453597836, | |
| "learning_rate": 2.5280467984434953e-06, | |
| "loss": 0.362, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 1.6198669767244074, | |
| "learning_rate": 2.4874503491594426e-06, | |
| "loss": 0.3672, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "grad_norm": 1.5427921307871089, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.3257, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "grad_norm": 2.017520614842527, | |
| "learning_rate": 2.4072185781965238e-06, | |
| "loss": 0.4051, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "grad_norm": 1.7916974259141798, | |
| "learning_rate": 2.3675837973655413e-06, | |
| "loss": 0.3854, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "grad_norm": 1.7283761414644547, | |
| "learning_rate": 2.3282701099301583e-06, | |
| "loss": 0.3677, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "grad_norm": 1.9369814377838661, | |
| "learning_rate": 2.28927778090664e-06, | |
| "loss": 0.4454, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "grad_norm": 1.7594007165743824, | |
| "learning_rate": 2.250607073144978e-06, | |
| "loss": 0.4041, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "grad_norm": 1.579492207812733, | |
| "learning_rate": 2.212258247327059e-06, | |
| "loss": 0.3765, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "grad_norm": 1.8600771758383663, | |
| "learning_rate": 2.1742315619649678e-06, | |
| "loss": 0.4138, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "grad_norm": 1.715787747296006, | |
| "learning_rate": 2.136527273399197e-06, | |
| "loss": 0.3434, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "grad_norm": 1.6078075910681187, | |
| "learning_rate": 2.099145635796962e-06, | |
| "loss": 0.2971, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 5.46, | |
| "grad_norm": 1.7101609091723584, | |
| "learning_rate": 2.0620869011504508e-06, | |
| "loss": 0.3561, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 5.46, | |
| "grad_norm": 1.755319755784633, | |
| "learning_rate": 2.0253513192751373e-06, | |
| "loss": 0.4157, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 5.47, | |
| "grad_norm": 1.9860693595376244, | |
| "learning_rate": 1.98893913780811e-06, | |
| "loss": 0.4383, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 5.47, | |
| "grad_norm": 1.8549329680023345, | |
| "learning_rate": 1.9528506022063907e-06, | |
| "loss": 0.4325, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "grad_norm": 1.7598983864505284, | |
| "learning_rate": 1.9170859557452846e-06, | |
| "loss": 0.4323, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "grad_norm": 1.6814371975401559, | |
| "learning_rate": 1.8816454395167328e-06, | |
| "loss": 0.3696, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "grad_norm": 1.675635736880616, | |
| "learning_rate": 1.8465292924276845e-06, | |
| "loss": 0.3818, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "grad_norm": 1.8068893605651373, | |
| "learning_rate": 1.8117377511985223e-06, | |
| "loss": 0.4381, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 1.7031720300351896, | |
| "learning_rate": 1.7772710503614122e-06, | |
| "loss": 0.4074, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 1.8945289964665175, | |
| "learning_rate": 1.7431294222587492e-06, | |
| "loss": 0.4662, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "grad_norm": 1.6723232333788873, | |
| "learning_rate": 1.7093130970416038e-06, | |
| "loss": 0.3435, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "grad_norm": 1.5575163788888324, | |
| "learning_rate": 1.6758223026681507e-06, | |
| "loss": 0.3621, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "grad_norm": 1.504432747275806, | |
| "learning_rate": 1.6426572649021476e-06, | |
| "loss": 0.2935, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "grad_norm": 1.7977724567770013, | |
| "learning_rate": 1.6098182073113754e-06, | |
| "loss": 0.4337, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "grad_norm": 1.8336728231598682, | |
| "learning_rate": 1.577305351266206e-06, | |
| "loss": 0.4527, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 5.53, | |
| "grad_norm": 1.844872569431052, | |
| "learning_rate": 1.5451189159380265e-06, | |
| "loss": 0.3774, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 5.53, | |
| "grad_norm": 1.76905301805892, | |
| "learning_rate": 1.5132591182978106e-06, | |
| "loss": 0.3965, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "grad_norm": 1.9969844754418222, | |
| "learning_rate": 1.4817261731146503e-06, | |
| "loss": 0.4193, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "grad_norm": 1.7627044062481438, | |
| "learning_rate": 1.4505202929543048e-06, | |
| "loss": 0.4027, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 5.55, | |
| "grad_norm": 1.7984942850258447, | |
| "learning_rate": 1.419641688177753e-06, | |
| "loss": 0.4443, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 5.55, | |
| "grad_norm": 1.7758681262918605, | |
| "learning_rate": 1.389090566939788e-06, | |
| "loss": 0.452, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "grad_norm": 1.860667890199162, | |
| "learning_rate": 1.358867135187636e-06, | |
| "loss": 0.3609, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "grad_norm": 1.8267685871465127, | |
| "learning_rate": 1.3289715966595174e-06, | |
| "loss": 0.4358, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "grad_norm": 1.6583202942625692, | |
| "learning_rate": 1.2994041528833266e-06, | |
| "loss": 0.4166, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "grad_norm": 1.8459469829349806, | |
| "learning_rate": 1.2701650031752277e-06, | |
| "loss": 0.4494, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 5.58, | |
| "grad_norm": 1.9690079994685203, | |
| "learning_rate": 1.2412543446383374e-06, | |
| "loss": 0.4927, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 5.58, | |
| "grad_norm": 2.0087709565549017, | |
| "learning_rate": 1.2126723721614053e-06, | |
| "loss": 0.4526, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 5.59, | |
| "grad_norm": 1.9537745315937955, | |
| "learning_rate": 1.1844192784174757e-06, | |
| "loss": 0.4479, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 5.59, | |
| "grad_norm": 1.8085786330659137, | |
| "learning_rate": 1.1564952538625884e-06, | |
| "loss": 0.3945, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 1.7956788996022015, | |
| "learning_rate": 1.1289004867345133e-06, | |
| "loss": 0.4453, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 1.724756990685903, | |
| "learning_rate": 1.1016351630514843e-06, | |
| "loss": 0.4213, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 5.61, | |
| "grad_norm": 1.7751733848596096, | |
| "learning_rate": 1.0746994666109234e-06, | |
| "loss": 0.3934, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 5.61, | |
| "grad_norm": 1.7645702681843625, | |
| "learning_rate": 1.048093578988213e-06, | |
| "loss": 0.3446, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "grad_norm": 1.7120396990099136, | |
| "learning_rate": 1.021817679535464e-06, | |
| "loss": 0.3698, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "grad_norm": 1.6633910576418705, | |
| "learning_rate": 9.958719453803278e-07, | |
| "loss": 0.3855, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 5.63, | |
| "grad_norm": 1.7077072115177097, | |
| "learning_rate": 9.702565514247918e-07, | |
| "loss": 0.3724, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 5.63, | |
| "grad_norm": 1.7601996690253545, | |
| "learning_rate": 9.449716703439804e-07, | |
| "loss": 0.4168, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "grad_norm": 1.8996797026474692, | |
| "learning_rate": 9.200174725850219e-07, | |
| "loss": 0.4677, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "grad_norm": 1.8232526565489129, | |
| "learning_rate": 8.953941263658727e-07, | |
| "loss": 0.4281, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "grad_norm": 1.6971924726323753, | |
| "learning_rate": 8.711017976742175e-07, | |
| "loss": 0.4344, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "grad_norm": 1.7296412238507917, | |
| "learning_rate": 8.471406502663037e-07, | |
| "loss": 0.4066, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "grad_norm": 1.6341194104420105, | |
| "learning_rate": 8.235108456658813e-07, | |
| "loss": 0.3872, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "grad_norm": 1.6950686680423468, | |
| "learning_rate": 8.002125431630869e-07, | |
| "loss": 0.3364, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "grad_norm": 1.7504675921360768, | |
| "learning_rate": 7.77245899813378e-07, | |
| "loss": 0.4278, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 5.67, | |
| "grad_norm": 1.8374321726809235, | |
| "learning_rate": 7.546110704364895e-07, | |
| "loss": 0.4415, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 5.67, | |
| "grad_norm": 1.6230477566019639, | |
| "learning_rate": 7.323082076153509e-07, | |
| "loss": 0.4153, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "grad_norm": 1.7736333759334915, | |
| "learning_rate": 7.103374616951042e-07, | |
| "loss": 0.4324, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "grad_norm": 1.437327390433701, | |
| "learning_rate": 6.886989807820488e-07, | |
| "loss": 0.3094, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 5.69, | |
| "grad_norm": 1.7291165196693685, | |
| "learning_rate": 6.673929107426868e-07, | |
| "loss": 0.4134, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 5.69, | |
| "grad_norm": 1.7519849248088708, | |
| "learning_rate": 6.464193952026965e-07, | |
| "loss": 0.362, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 1.6470515914547659, | |
| "learning_rate": 6.257785755459877e-07, | |
| "loss": 0.3736, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 1.7253306383504192, | |
| "learning_rate": 6.054705909137426e-07, | |
| "loss": 0.4416, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "grad_norm": 1.936865854193816, | |
| "learning_rate": 5.854955782034988e-07, | |
| "loss": 0.4335, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "grad_norm": 1.6943216112791677, | |
| "learning_rate": 5.658536720681728e-07, | |
| "loss": 0.456, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 5.72, | |
| "grad_norm": 1.632015930965036, | |
| "learning_rate": 5.465450049152221e-07, | |
| "loss": 0.4139, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 5.72, | |
| "grad_norm": 1.678437736133256, | |
| "learning_rate": 5.275697069056895e-07, | |
| "loss": 0.3857, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "grad_norm": 1.8207520891231839, | |
| "learning_rate": 5.089279059533658e-07, | |
| "loss": 0.4214, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "grad_norm": 1.6142638217862484, | |
| "learning_rate": 4.906197277239178e-07, | |
| "loss": 0.3383, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 5.74, | |
| "grad_norm": 1.8111968361900292, | |
| "learning_rate": 4.7264529563403324e-07, | |
| "loss": 0.4011, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 5.74, | |
| "grad_norm": 1.9272585160398414, | |
| "learning_rate": 4.550047308505945e-07, | |
| "loss": 0.4285, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "grad_norm": 1.8288494375046014, | |
| "learning_rate": 4.376981522898671e-07, | |
| "loss": 0.3567, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "grad_norm": 2.3157716530481878, | |
| "learning_rate": 4.207256766166845e-07, | |
| "loss": 0.5008, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "grad_norm": 1.574563992208348, | |
| "learning_rate": 4.0408741824367604e-07, | |
| "loss": 0.3265, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "grad_norm": 1.7838485958441967, | |
| "learning_rate": 3.8778348933048436e-07, | |
| "loss": 0.3415, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "grad_norm": 1.8564777757353124, | |
| "learning_rate": 3.718139997830106e-07, | |
| "loss": 0.4093, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "grad_norm": 1.6620301708737435, | |
| "learning_rate": 3.561790572526924e-07, | |
| "loss": 0.3586, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "grad_norm": 1.928302195715843, | |
| "learning_rate": 3.408787671357494e-07, | |
| "loss": 0.4215, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 5.78, | |
| "grad_norm": 1.8242036534719173, | |
| "learning_rate": 3.2591323257248893e-07, | |
| "loss": 0.4448, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 5.78, | |
| "grad_norm": 1.7571467327121855, | |
| "learning_rate": 3.1128255444661805e-07, | |
| "loss": 0.4766, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 5.79, | |
| "grad_norm": 1.7194165217331403, | |
| "learning_rate": 2.9698683138452697e-07, | |
| "loss": 0.346, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 5.79, | |
| "grad_norm": 1.868762938346331, | |
| "learning_rate": 2.830261597546846e-07, | |
| "loss": 0.4348, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 1.588310688177164, | |
| "learning_rate": 2.6940063366693303e-07, | |
| "loss": 0.3081, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 1.498217939911864, | |
| "learning_rate": 2.561103449718827e-07, | |
| "loss": 0.3004, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "grad_norm": 1.7944299742124368, | |
| "learning_rate": 2.4315538326028507e-07, | |
| "loss": 0.438, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "grad_norm": 1.6627821471652164, | |
| "learning_rate": 2.305358358624221e-07, | |
| "loss": 0.4249, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 5.82, | |
| "grad_norm": 1.6989673632236972, | |
| "learning_rate": 2.1825178784753985e-07, | |
| "loss": 0.3876, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 5.82, | |
| "grad_norm": 1.6849056002434288, | |
| "learning_rate": 2.063033220232491e-07, | |
| "loss": 0.3711, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 1.6558156644610367, | |
| "learning_rate": 1.946905189349757e-07, | |
| "loss": 0.3559, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 1.711368104238552, | |
| "learning_rate": 1.8341345686543332e-07, | |
| "loss": 0.4078, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "grad_norm": 1.9634733011668026, | |
| "learning_rate": 1.724722118340627e-07, | |
| "loss": 0.4308, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "grad_norm": 1.9123458061067942, | |
| "learning_rate": 1.6186685759655428e-07, | |
| "loss": 0.427, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 5.85, | |
| "grad_norm": 1.7034588097045482, | |
| "learning_rate": 1.5159746564433198e-07, | |
| "loss": 0.3741, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 5.85, | |
| "grad_norm": 1.7292657979658022, | |
| "learning_rate": 1.416641052040757e-07, | |
| "loss": 0.3963, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "grad_norm": 1.8915053528613424, | |
| "learning_rate": 1.3206684323724961e-07, | |
| "loss": 0.3722, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "grad_norm": 1.8642059643739726, | |
| "learning_rate": 1.2280574443965242e-07, | |
| "loss": 0.4073, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "grad_norm": 1.7812049935425671, | |
| "learning_rate": 1.1388087124099556e-07, | |
| "loss": 0.3828, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "grad_norm": 1.9268846934389878, | |
| "learning_rate": 1.0529228380445899e-07, | |
| "loss": 0.4493, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 5.88, | |
| "grad_norm": 1.6724454049419895, | |
| "learning_rate": 9.704004002630274e-08, | |
| "loss": 0.409, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 5.88, | |
| "grad_norm": 1.8933262536296047, | |
| "learning_rate": 8.912419553547269e-08, | |
| "loss": 0.3906, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 5.88, | |
| "grad_norm": 1.8997119827971012, | |
| "learning_rate": 8.15448036932176e-08, | |
| "loss": 0.4052, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "grad_norm": 1.610607557905884, | |
| "learning_rate": 7.430191559273936e-08, | |
| "loss": 0.3859, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "grad_norm": 1.8048033072575713, | |
| "learning_rate": 6.739558005884883e-08, | |
| "loss": 0.3439, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 1.6649072403261496, | |
| "learning_rate": 6.082584364762722e-08, | |
| "loss": 0.3658, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 1.9094687067026925, | |
| "learning_rate": 5.459275064611524e-08, | |
| "loss": 0.4879, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 5.91, | |
| "grad_norm": 1.8260693835668214, | |
| "learning_rate": 4.869634307202997e-08, | |
| "loss": 0.4475, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 5.91, | |
| "grad_norm": 1.6977001739048292, | |
| "learning_rate": 4.313666067345401e-08, | |
| "loss": 0.3918, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "grad_norm": 1.6095040936302651, | |
| "learning_rate": 3.791374092859679e-08, | |
| "loss": 0.3745, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "grad_norm": 1.8893539624969005, | |
| "learning_rate": 3.3027619045511435e-08, | |
| "loss": 0.4356, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 5.93, | |
| "grad_norm": 1.7955489356168806, | |
| "learning_rate": 2.847832796188388e-08, | |
| "loss": 0.4511, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.93, | |
| "grad_norm": 1.8047868077005986, | |
| "learning_rate": 2.426589834479964e-08, | |
| "loss": 0.3773, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "grad_norm": 1.882375952307432, | |
| "learning_rate": 2.0390358590538504e-08, | |
| "loss": 0.4503, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "grad_norm": 1.687944650059007, | |
| "learning_rate": 1.685173482438018e-08, | |
| "loss": 0.3282, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 5.95, | |
| "grad_norm": 1.6937250028919937, | |
| "learning_rate": 1.3650050900426702e-08, | |
| "loss": 0.4567, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 5.95, | |
| "grad_norm": 1.6985143471210042, | |
| "learning_rate": 1.0785328401452522e-08, | |
| "loss": 0.4132, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 5.96, | |
| "grad_norm": 1.865882484601669, | |
| "learning_rate": 8.257586638749093e-09, | |
| "loss": 0.4312, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 5.96, | |
| "grad_norm": 1.7593475277650907, | |
| "learning_rate": 6.0668426519971955e-09, | |
| "loss": 0.415, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 5.97, | |
| "grad_norm": 1.7445633518118073, | |
| "learning_rate": 4.213111209155907e-09, | |
| "loss": 0.45, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 5.97, | |
| "grad_norm": 1.581260484943774, | |
| "learning_rate": 2.6964048063515822e-09, | |
| "loss": 0.3601, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "grad_norm": 1.739106535520495, | |
| "learning_rate": 1.516733667811243e-09, | |
| "loss": 0.3955, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "grad_norm": 1.7190964613757493, | |
| "learning_rate": 6.741057457737566e-10, | |
| "loss": 0.4375, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 5.99, | |
| "grad_norm": 1.6442777615342268, | |
| "learning_rate": 1.6852672045652995e-10, | |
| "loss": 0.3787, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 5.99, | |
| "grad_norm": 1.7221089917417118, | |
| "learning_rate": 0.0, | |
| "loss": 0.3859, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 5.99, | |
| "step": 1248, | |
| "total_flos": 131510528548864.0, | |
| "train_loss": 1.1018599691824653, | |
| "train_runtime": 20850.851, | |
| "train_samples_per_second": 2.878, | |
| "train_steps_per_second": 0.06 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 1248, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 200, | |
| "total_flos": 131510528548864.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |