| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9990426041168023, | |
| "eval_steps": 500, | |
| "global_step": 1044, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0019147917663954045, | |
| "grad_norm": 61.0, | |
| "learning_rate": 3.773584905660378e-07, | |
| "loss": 3.0718, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003829583532790809, | |
| "grad_norm": 62.0, | |
| "learning_rate": 7.547169811320755e-07, | |
| "loss": 3.1052, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0057443752991862135, | |
| "grad_norm": 62.5, | |
| "learning_rate": 1.1320754716981133e-06, | |
| "loss": 3.1376, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007659167065581618, | |
| "grad_norm": 103.5, | |
| "learning_rate": 1.509433962264151e-06, | |
| "loss": 3.0909, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009573958831977022, | |
| "grad_norm": 62.0, | |
| "learning_rate": 1.8867924528301889e-06, | |
| "loss": 3.1064, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.011488750598372427, | |
| "grad_norm": 60.5, | |
| "learning_rate": 2.2641509433962266e-06, | |
| "loss": 3.0511, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.013403542364767831, | |
| "grad_norm": 58.75, | |
| "learning_rate": 2.6415094339622644e-06, | |
| "loss": 3.0435, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.015318334131163236, | |
| "grad_norm": 54.0, | |
| "learning_rate": 3.018867924528302e-06, | |
| "loss": 2.9226, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01723312589755864, | |
| "grad_norm": 56.0, | |
| "learning_rate": 3.3962264150943395e-06, | |
| "loss": 3.0539, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.019147917663954045, | |
| "grad_norm": 50.5, | |
| "learning_rate": 3.7735849056603777e-06, | |
| "loss": 2.8775, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02106270943034945, | |
| "grad_norm": 46.75, | |
| "learning_rate": 4.150943396226416e-06, | |
| "loss": 2.8774, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.022977501196744854, | |
| "grad_norm": 42.5, | |
| "learning_rate": 4.528301886792453e-06, | |
| "loss": 2.6977, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02489229296314026, | |
| "grad_norm": 35.5, | |
| "learning_rate": 4.905660377358491e-06, | |
| "loss": 2.6109, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.026807084729535663, | |
| "grad_norm": 30.0, | |
| "learning_rate": 5.283018867924529e-06, | |
| "loss": 2.5524, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.028721876495931067, | |
| "grad_norm": 24.125, | |
| "learning_rate": 5.660377358490566e-06, | |
| "loss": 2.5467, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.030636668262326472, | |
| "grad_norm": 19.625, | |
| "learning_rate": 6.037735849056604e-06, | |
| "loss": 2.3976, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.032551460028721876, | |
| "grad_norm": 17.25, | |
| "learning_rate": 6.415094339622642e-06, | |
| "loss": 2.349, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03446625179511728, | |
| "grad_norm": 13.625, | |
| "learning_rate": 6.792452830188679e-06, | |
| "loss": 2.3495, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.036381043561512685, | |
| "grad_norm": 10.375, | |
| "learning_rate": 7.169811320754717e-06, | |
| "loss": 2.3404, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.03829583532790809, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 7.5471698113207555e-06, | |
| "loss": 2.2341, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.040210627094303494, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 7.924528301886793e-06, | |
| "loss": 2.1499, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0421254188606989, | |
| "grad_norm": 7.125, | |
| "learning_rate": 8.301886792452832e-06, | |
| "loss": 2.1879, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0440402106270943, | |
| "grad_norm": 5.875, | |
| "learning_rate": 8.67924528301887e-06, | |
| "loss": 2.147, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04595500239348971, | |
| "grad_norm": 15.375, | |
| "learning_rate": 9.056603773584907e-06, | |
| "loss": 2.2175, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04786979415988511, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 9.433962264150944e-06, | |
| "loss": 2.1928, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04978458592628052, | |
| "grad_norm": 4.125, | |
| "learning_rate": 9.811320754716981e-06, | |
| "loss": 2.1355, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.05169937769267592, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 1.018867924528302e-05, | |
| "loss": 2.117, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.053614169459071326, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 1.0566037735849058e-05, | |
| "loss": 2.0604, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.05552896122546673, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.0943396226415095e-05, | |
| "loss": 2.0949, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.057443752991862135, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 1.1320754716981132e-05, | |
| "loss": 2.0517, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05935854475825754, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 1.169811320754717e-05, | |
| "loss": 2.1012, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.061273336524652944, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 1.2075471698113209e-05, | |
| "loss": 2.0485, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.06318812829104835, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 1.2452830188679246e-05, | |
| "loss": 1.9389, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06510292005744375, | |
| "grad_norm": 3.0, | |
| "learning_rate": 1.2830188679245283e-05, | |
| "loss": 2.0672, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06701771182383916, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 1.320754716981132e-05, | |
| "loss": 2.0884, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06893250359023456, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 1.3584905660377358e-05, | |
| "loss": 2.0209, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.07084729535662997, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 1.3962264150943397e-05, | |
| "loss": 1.952, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.07276208712302537, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 1.4339622641509435e-05, | |
| "loss": 1.9835, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07467687888942078, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 1.4716981132075472e-05, | |
| "loss": 1.962, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07659167065581618, | |
| "grad_norm": 2.625, | |
| "learning_rate": 1.5094339622641511e-05, | |
| "loss": 1.9884, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07850646242221158, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 1.547169811320755e-05, | |
| "loss": 2.0301, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.08042125418860699, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.5849056603773586e-05, | |
| "loss": 1.9727, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0823360459550024, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 1.6226415094339625e-05, | |
| "loss": 2.0236, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0842508377213978, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 1.6603773584905664e-05, | |
| "loss": 1.9287, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0861656294877932, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 1.69811320754717e-05, | |
| "loss": 1.9009, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0880804212541886, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.735849056603774e-05, | |
| "loss": 1.9898, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08999521302058401, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 1.7735849056603774e-05, | |
| "loss": 1.9054, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.09191000478697942, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 1.8113207547169813e-05, | |
| "loss": 1.9189, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.09382479655337482, | |
| "grad_norm": 2.78125, | |
| "learning_rate": 1.8490566037735852e-05, | |
| "loss": 1.9284, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09573958831977022, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 1.8867924528301888e-05, | |
| "loss": 1.8944, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09765438008616563, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 1.9245283018867927e-05, | |
| "loss": 1.9709, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.09956917185256103, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 1.9622641509433963e-05, | |
| "loss": 1.9177, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.10148396361895644, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9661, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.10339875538535184, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 1.9999949751618577e-05, | |
| "loss": 1.8821, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.10531354715174725, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 1.9999799006979282e-05, | |
| "loss": 1.922, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.10722833891814265, | |
| "grad_norm": 2.875, | |
| "learning_rate": 1.9999547767597055e-05, | |
| "loss": 2.01, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10914313068453806, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 1.9999196035996768e-05, | |
| "loss": 1.9654, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.11105792245093346, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 1.999874381571321e-05, | |
| "loss": 1.9159, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.11297271421732887, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 1.999819111129105e-05, | |
| "loss": 1.9388, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.11488750598372427, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 1.9997537928284783e-05, | |
| "loss": 1.8953, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11680229775011967, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 1.9996784273258688e-05, | |
| "loss": 1.8523, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.11871708951651508, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 1.999593015378676e-05, | |
| "loss": 1.9153, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.12063188128291048, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 1.9994975578452626e-05, | |
| "loss": 1.817, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.12254667304930589, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 1.9993920556849448e-05, | |
| "loss": 1.845, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.12446146481570129, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 1.9992765099579853e-05, | |
| "loss": 1.8886, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.1263762565820967, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.999150921825582e-05, | |
| "loss": 1.8955, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.12829104834849211, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.9990152925498545e-05, | |
| "loss": 1.8308, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1302058401148875, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9988696234938333e-05, | |
| "loss": 1.9141, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.13212063188128292, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.998713916121445e-05, | |
| "loss": 1.9814, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.13403542364767831, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.9985481719974985e-05, | |
| "loss": 1.7735, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13595021541407373, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.9983723927876685e-05, | |
| "loss": 1.8418, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.13786500718046912, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9981865802584795e-05, | |
| "loss": 1.8867, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13977979894686454, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.9979907362772865e-05, | |
| "loss": 1.8945, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.14169459071325993, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.997784862812259e-05, | |
| "loss": 1.8312, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.14360938247965535, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.997568961932358e-05, | |
| "loss": 1.8979, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.14552417424605074, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.997343035807318e-05, | |
| "loss": 1.8536, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.14743896601244616, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.9971070867076227e-05, | |
| "loss": 1.814, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.14935375777884155, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.996861117004485e-05, | |
| "loss": 1.8618, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.15126854954523697, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.9966051291698202e-05, | |
| "loss": 1.8376, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.15318334131163236, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.9963391257762234e-05, | |
| "loss": 1.8344, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.15509813307802778, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.9960631094969424e-05, | |
| "loss": 1.8546, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.15701292484442317, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.9957770831058518e-05, | |
| "loss": 1.8692, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.1589277166108186, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.995481049477424e-05, | |
| "loss": 1.8659, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.16084250837721398, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.9951750115867013e-05, | |
| "loss": 1.8511, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.1627573001436094, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.994858972509265e-05, | |
| "loss": 1.8137, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1646720919100048, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.994532935421206e-05, | |
| "loss": 1.9099, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.1665868836764002, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.9941969035990913e-05, | |
| "loss": 1.8185, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1685016754427956, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.9938508804199322e-05, | |
| "loss": 1.8309, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.170416467209191, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.9934948693611495e-05, | |
| "loss": 1.7971, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.1723312589755864, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9931288740005388e-05, | |
| "loss": 1.8312, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.17424605074198182, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.992752898016235e-05, | |
| "loss": 1.7379, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.1761608425083772, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 1.9923669451866753e-05, | |
| "loss": 1.8071, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.17807563427477263, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9919710193905607e-05, | |
| "loss": 1.8601, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.17999042604116802, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.9915651246068176e-05, | |
| "loss": 1.8539, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.18190521780756344, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.991149264914556e-05, | |
| "loss": 1.8179, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.18382000957395883, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.9907234444930328e-05, | |
| "loss": 1.8305, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.18573480134035425, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.9902876676216044e-05, | |
| "loss": 1.8367, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.18764959310674964, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.989841938679687e-05, | |
| "loss": 1.8328, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.18956438487314506, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.989386262146713e-05, | |
| "loss": 1.8388, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.19147917663954045, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.9889206426020837e-05, | |
| "loss": 1.834, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.19339396840593587, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.988445084725125e-05, | |
| "loss": 1.8004, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.19530876017233126, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.987959593295039e-05, | |
| "loss": 1.8231, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.19722355193872668, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.987464173190858e-05, | |
| "loss": 1.8879, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.19913834370512207, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.9869588293913932e-05, | |
| "loss": 1.8918, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.20105313547151749, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.986443566975187e-05, | |
| "loss": 1.8015, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.20296792723791288, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.9859183911204588e-05, | |
| "loss": 1.8785, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2048827190043083, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.9853833071050567e-05, | |
| "loss": 1.8675, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.20679751077070369, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.9848383203064018e-05, | |
| "loss": 1.8918, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.2087123025370991, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.984283436201435e-05, | |
| "loss": 1.8418, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2106270943034945, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9837186603665615e-05, | |
| "loss": 1.8626, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2125418860698899, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.983143998477596e-05, | |
| "loss": 1.858, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2144566778362853, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9825594563097043e-05, | |
| "loss": 1.8028, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.21637146960268072, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.981965039737346e-05, | |
| "loss": 1.7789, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2182862613690761, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.9813607547342152e-05, | |
| "loss": 1.7911, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.22020105313547153, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.9807466073731806e-05, | |
| "loss": 1.7719, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.22211584490186692, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9801226038262244e-05, | |
| "loss": 1.8903, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.22403063666826234, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.9794887503643805e-05, | |
| "loss": 1.8288, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.22594542843465773, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.9788450533576708e-05, | |
| "loss": 1.7571, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.22786022020105315, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9781915192750413e-05, | |
| "loss": 1.8572, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.22977501196744854, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.9775281546842985e-05, | |
| "loss": 1.7588, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.23168980373384396, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.976854966252042e-05, | |
| "loss": 1.8695, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.23360459550023935, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9761719607435973e-05, | |
| "loss": 1.8384, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.23551938726663477, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.9754791450229485e-05, | |
| "loss": 1.8699, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.23743417903303016, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.9747765260526696e-05, | |
| "loss": 1.8323, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.23934897079942558, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9740641108938538e-05, | |
| "loss": 1.7859, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.24126376256582097, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.9733419067060427e-05, | |
| "loss": 1.8269, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.24317855433221638, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.972609920747155e-05, | |
| "loss": 1.8094, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.24509334609861178, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.9718681603734124e-05, | |
| "loss": 1.8463, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.2470081378650072, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.971116633039266e-05, | |
| "loss": 1.9069, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.24892292963140258, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.9703553462973227e-05, | |
| "loss": 1.7865, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.250837721397798, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.9695843077982677e-05, | |
| "loss": 1.8369, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.2527525131641934, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.9688035252907888e-05, | |
| "loss": 1.823, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.2546673049305888, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9680130066214968e-05, | |
| "loss": 1.8506, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.25658209669698423, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.9672127597348486e-05, | |
| "loss": 1.8261, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.2584968884633796, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.9664027926730664e-05, | |
| "loss": 1.7482, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.260411680229775, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.965583113576057e-05, | |
| "loss": 1.8663, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.26232647199617043, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.9647537306813303e-05, | |
| "loss": 1.8714, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.26424126376256585, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.9639146523239156e-05, | |
| "loss": 1.7767, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.2661560555289612, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9630658869362786e-05, | |
| "loss": 1.7756, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.26807084729535663, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.9622074430482366e-05, | |
| "loss": 1.8538, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.26998563906175205, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.9613393292868735e-05, | |
| "loss": 1.8367, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.27190043082814747, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.9604615543764506e-05, | |
| "loss": 1.8716, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.27381522259454283, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.9595741271383225e-05, | |
| "loss": 1.8014, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.27573001436093825, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9586770564908453e-05, | |
| "loss": 1.7858, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.27764480612733367, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.9577703514492886e-05, | |
| "loss": 1.7948, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.2795595978937291, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.9568540211257444e-05, | |
| "loss": 1.8298, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.28147438966012445, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9559280747290362e-05, | |
| "loss": 1.8202, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.28338918142651986, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.954992521564625e-05, | |
| "loss": 1.8162, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2853039731929153, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.9540473710345177e-05, | |
| "loss": 1.8356, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.2872187649593107, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.9530926326371712e-05, | |
| "loss": 1.8004, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.28913355672570606, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.952128315967397e-05, | |
| "loss": 1.8221, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2910483484921015, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.9511544307162656e-05, | |
| "loss": 1.8628, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2929631402584969, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.9501709866710086e-05, | |
| "loss": 1.7806, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.2948779320248923, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.9491779937149204e-05, | |
| "loss": 1.8573, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2967927237912877, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9481754618272585e-05, | |
| "loss": 1.8113, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2987075155576831, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.947163401083144e-05, | |
| "loss": 1.8203, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3006223073240785, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.9461418216534594e-05, | |
| "loss": 1.8106, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.30253709909047394, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.9451107338047478e-05, | |
| "loss": 1.792, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3044518908568693, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.9440701478991078e-05, | |
| "loss": 1.8266, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.3063666826232647, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.9430200743940915e-05, | |
| "loss": 1.862, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.30828147438966014, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.9419605238425975e-05, | |
| "loss": 1.8587, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.31019626615605556, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9408915068927653e-05, | |
| "loss": 1.7967, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3121110579224509, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.93981303428787e-05, | |
| "loss": 1.7357, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.31402584968884634, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9387251168662114e-05, | |
| "loss": 1.8246, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.31594064145524176, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.937627765561008e-05, | |
| "loss": 1.8022, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3178554332216372, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.936520991400285e-05, | |
| "loss": 1.8312, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.31977022498803254, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.935404805506764e-05, | |
| "loss": 1.7942, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.32168501675442795, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.9342792190977532e-05, | |
| "loss": 1.8574, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3235998085208234, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.93314424348503e-05, | |
| "loss": 1.78, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3255146002872188, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.9319998900747327e-05, | |
| "loss": 1.8314, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.32742939205361415, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.930846170367243e-05, | |
| "loss": 1.8051, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.3293441838200096, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.9296830959570697e-05, | |
| "loss": 1.9012, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.331258975586405, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.9285106785327345e-05, | |
| "loss": 1.8383, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3331737673528004, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.9273289298766532e-05, | |
| "loss": 1.808, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3350885591191958, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.9261378618650166e-05, | |
| "loss": 1.7936, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3370033508855912, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9249374864676733e-05, | |
| "loss": 1.8398, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.3389181426519866, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9237278157480074e-05, | |
| "loss": 1.7898, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.340832934418382, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.922508861862818e-05, | |
| "loss": 1.8345, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.3427477261847774, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9212806370621963e-05, | |
| "loss": 1.8619, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.3446625179511728, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.920043153689405e-05, | |
| "loss": 1.8286, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3465773097175682, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.9187964241807508e-05, | |
| "loss": 1.8079, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.34849210148396365, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.917540461065462e-05, | |
| "loss": 1.7952, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.350406893250359, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.916275276965561e-05, | |
| "loss": 1.8172, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.3523216850167544, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.9150008845957388e-05, | |
| "loss": 1.7986, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.35423647678314985, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9137172967632262e-05, | |
| "loss": 1.7909, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.35615126854954526, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.912424526367665e-05, | |
| "loss": 1.7713, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.3580660603159406, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9111225864009794e-05, | |
| "loss": 1.8165, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.35998085208233604, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9098114899472443e-05, | |
| "loss": 1.8618, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.36189564384873146, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.9084912501825554e-05, | |
| "loss": 1.8634, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.3638104356151269, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.9071618803748945e-05, | |
| "loss": 1.8141, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.36572522738152224, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9058233938839975e-05, | |
| "loss": 1.7962, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.36764001914791766, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.9044758041612207e-05, | |
| "loss": 1.8472, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.3695548109143131, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.9031191247494046e-05, | |
| "loss": 1.7234, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.3714696026807085, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9017533692827383e-05, | |
| "loss": 1.8239, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.37338439444710386, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.9003785514866214e-05, | |
| "loss": 1.7942, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3752991862134993, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.8989946851775287e-05, | |
| "loss": 1.8161, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.3772139779798947, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.8976017842628677e-05, | |
| "loss": 1.8224, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.3791287697462901, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.8961998627408424e-05, | |
| "loss": 1.8211, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.3810435615126855, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.89478893470031e-05, | |
| "loss": 1.8292, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.3829583532790809, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.89336901432064e-05, | |
| "loss": 1.8522, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3848731450454763, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.891940115871574e-05, | |
| "loss": 1.8274, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.38678793681187174, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.8905022537130774e-05, | |
| "loss": 1.8349, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.3887027285782671, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.8890554422952e-05, | |
| "loss": 1.8058, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.3906175203446625, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.8875996961579282e-05, | |
| "loss": 1.8383, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.39253231211105793, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.8861350299310397e-05, | |
| "loss": 1.7721, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.39444710387745335, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.8846614583339555e-05, | |
| "loss": 1.8389, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.3963618956438487, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.883178996175593e-05, | |
| "loss": 1.8014, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.39827668741024413, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.8816876583542177e-05, | |
| "loss": 1.8612, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.40019147917663955, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.8801874598572918e-05, | |
| "loss": 1.7983, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.40210627094303497, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.8786784157613242e-05, | |
| "loss": 1.8277, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.40402106270943033, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.8771605412317192e-05, | |
| "loss": 1.815, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.40593585447582575, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.875633851522625e-05, | |
| "loss": 1.7652, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.40785064624222117, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.874098361976779e-05, | |
| "loss": 1.8605, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.4097654380086166, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.872554088025354e-05, | |
| "loss": 1.7953, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.41168022977501195, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.8710010451878036e-05, | |
| "loss": 1.7491, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.41359502154140737, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.869439249071705e-05, | |
| "loss": 1.7363, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.4155098133078028, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 1.8678687153726037e-05, | |
| "loss": 1.8131, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.4174246050741982, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.866289459873855e-05, | |
| "loss": 1.8631, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.41933939684059357, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.8647014984464657e-05, | |
| "loss": 1.8336, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.421254188606989, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.8631048470489343e-05, | |
| "loss": 1.7929, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4231689803733844, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.8614995217270893e-05, | |
| "loss": 1.7582, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.4250837721397798, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.859885538613932e-05, | |
| "loss": 1.8399, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.4269985639061752, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.8582629139294693e-05, | |
| "loss": 1.8379, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.4289133556725706, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.8566316639805543e-05, | |
| "loss": 1.711, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.430828147438966, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.854991805160721e-05, | |
| "loss": 1.8693, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.43274293920536144, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.8533433539500195e-05, | |
| "loss": 1.7657, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.4346577309717568, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.851686326914851e-05, | |
| "loss": 1.7919, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.4365725227381522, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.850020740707801e-05, | |
| "loss": 1.7852, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.43848731450454764, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.8483466120674708e-05, | |
| "loss": 1.7779, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.44040210627094306, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.846663957818312e-05, | |
| "loss": 1.8108, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4423168980373384, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.8449727948704556e-05, | |
| "loss": 1.7799, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.44423168980373384, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.843273140219541e-05, | |
| "loss": 1.7874, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.44614648157012926, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.8415650109465473e-05, | |
| "loss": 1.7549, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.4480612733365247, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.8398484242176206e-05, | |
| "loss": 1.7661, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.44997606510292004, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.8381233972839027e-05, | |
| "loss": 1.8091, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.45189085686931546, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.836389947481355e-05, | |
| "loss": 1.7503, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.4538056486357109, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.834648092230587e-05, | |
| "loss": 1.829, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.4557204404021063, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.8328978490366804e-05, | |
| "loss": 1.8481, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.45763523216850166, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.831139235489013e-05, | |
| "loss": 1.8335, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.4595500239348971, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.8293722692610805e-05, | |
| "loss": 1.7653, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4614648157012925, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.8275969681103226e-05, | |
| "loss": 1.8049, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.4633796074676879, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.8258133498779407e-05, | |
| "loss": 1.8257, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.4652943992340833, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.8240214324887212e-05, | |
| "loss": 1.8437, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.4672091910004787, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.8222212339508528e-05, | |
| "loss": 1.8295, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.4691239827668741, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.820412772355749e-05, | |
| "loss": 1.7919, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.47103877453326953, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.818596065877863e-05, | |
| "loss": 1.8298, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.4729535662996649, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.816771132774507e-05, | |
| "loss": 1.7665, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.4748683580660603, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.814937991385667e-05, | |
| "loss": 1.8966, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.47678314983245573, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.813096660133822e-05, | |
| "loss": 1.7429, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.47869794159885115, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.811247157523754e-05, | |
| "loss": 1.8369, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4806127333652465, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.8093895021423658e-05, | |
| "loss": 1.7983, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.48252752513164193, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.807523712658493e-05, | |
| "loss": 1.764, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.48444231689803735, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 1.8056498078227152e-05, | |
| "loss": 1.8022, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.48635710866443277, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.8037678064671702e-05, | |
| "loss": 1.7991, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.48827190043082813, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.801877727505362e-05, | |
| "loss": 1.8253, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.49018669219722355, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.7999795899319718e-05, | |
| "loss": 1.8086, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.49210148396361897, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.7980734128226685e-05, | |
| "loss": 1.8171, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.4940162757300144, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.7961592153339146e-05, | |
| "loss": 1.8484, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.49593106749640975, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.7942370167027756e-05, | |
| "loss": 1.8499, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.49784585926280517, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.7923068362467252e-05, | |
| "loss": 1.8069, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4997606510292006, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.7903686933634522e-05, | |
| "loss": 1.8241, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.501675442795596, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.7884226075306652e-05, | |
| "loss": 1.7349, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.5035902345619914, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.786468598305897e-05, | |
| "loss": 1.8709, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.5055050263283868, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.7845066853263068e-05, | |
| "loss": 1.7636, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.5074198180947822, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.7825368883084865e-05, | |
| "loss": 1.8139, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5093346098611776, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.7805592270482576e-05, | |
| "loss": 1.7859, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.511249401627573, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.7785737214204752e-05, | |
| "loss": 1.7915, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.5131641933939685, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.776580391378829e-05, | |
| "loss": 1.7832, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.5150789851603638, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.7745792569556402e-05, | |
| "loss": 1.7741, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.5169937769267592, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.772570338261662e-05, | |
| "loss": 1.8348, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5189085686931546, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.7705536554858768e-05, | |
| "loss": 1.768, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.52082336045955, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.768529228895294e-05, | |
| "loss": 1.8223, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.5227381522259454, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.7664970788347454e-05, | |
| "loss": 1.8459, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.5246529439923409, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.7644572257266807e-05, | |
| "loss": 1.7417, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.5265677357587363, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.762409690070964e-05, | |
| "loss": 1.8374, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5284825275251317, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.7603544924446655e-05, | |
| "loss": 1.8115, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.530397319291527, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.7582916535018567e-05, | |
| "loss": 1.8349, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.5323121110579224, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.7562211939734e-05, | |
| "loss": 1.7624, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.5342269028243178, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.7541431346667447e-05, | |
| "loss": 1.765, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.5361416945907133, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.752057496465713e-05, | |
| "loss": 1.8085, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5380564863571087, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.7499643003302943e-05, | |
| "loss": 1.7753, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.5399712781235041, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.7478635672964324e-05, | |
| "loss": 1.8409, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.5418860698898995, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.7457553184758135e-05, | |
| "loss": 1.7598, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.5438008616562949, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.7436395750556563e-05, | |
| "loss": 1.7612, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5457156534226902, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.7415163582984972e-05, | |
| "loss": 1.8075, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5476304451890857, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.739385689541977e-05, | |
| "loss": 1.775, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5495452369554811, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.7372475901986275e-05, | |
| "loss": 1.8337, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.5514600287218765, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.7351020817556548e-05, | |
| "loss": 1.8364, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.5533748204882719, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.732949185774724e-05, | |
| "loss": 1.7762, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5552896122546673, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.7307889238917424e-05, | |
| "loss": 1.8312, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5572044040210627, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.7286213178166434e-05, | |
| "loss": 1.7795, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.5591191957874582, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.726446389333166e-05, | |
| "loss": 1.7856, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5610339875538535, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.7242641602986376e-05, | |
| "loss": 1.7926, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.5629487793202489, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.7220746526437535e-05, | |
| "loss": 1.7724, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5648635710866443, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.7198778883723567e-05, | |
| "loss": 1.7524, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5667783628530397, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.717673889561217e-05, | |
| "loss": 1.7831, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.5686931546194351, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.7154626783598092e-05, | |
| "loss": 1.7599, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5706079463858306, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.7132442769900898e-05, | |
| "loss": 1.8121, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.572522738152226, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.7110187077462747e-05, | |
| "loss": 1.7459, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5744375299186214, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.7087859929946144e-05, | |
| "loss": 1.7934, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5763523216850167, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.706546155173169e-05, | |
| "loss": 1.7771, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.5782671134514121, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.7042992167915836e-05, | |
| "loss": 1.8167, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.5801819052178075, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.7020452004308622e-05, | |
| "loss": 1.7243, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.582096696984203, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.699784128743139e-05, | |
| "loss": 1.7949, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.5840114887505984, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.6975160244514522e-05, | |
| "loss": 1.8373, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.5859262805169938, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.6952409103495163e-05, | |
| "loss": 1.8333, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.5878410722833892, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.6929588093014916e-05, | |
| "loss": 1.8455, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.5897558640497846, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.6906697442417547e-05, | |
| "loss": 1.8232, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.59167065581618, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.688373738174668e-05, | |
| "loss": 1.7521, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.5935854475825754, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.6860708141743498e-05, | |
| "loss": 1.7724, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5955002393489708, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.6837609953844406e-05, | |
| "loss": 1.7663, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.5974150311153662, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.6814443050178713e-05, | |
| "loss": 1.8029, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5993298228817616, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.6791207663566307e-05, | |
| "loss": 1.805, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.601244614648157, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.67679040275153e-05, | |
| "loss": 1.7871, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.6031594064145525, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.6744532376219688e-05, | |
| "loss": 1.7786, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6050741981809479, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.6721092944557003e-05, | |
| "loss": 1.7525, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.6069889899473432, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.6697585968085942e-05, | |
| "loss": 1.7923, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.6089037817137386, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.667401168304401e-05, | |
| "loss": 1.7879, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.610818573480134, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.6650370326345146e-05, | |
| "loss": 1.797, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.6127333652465294, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.6626662135577324e-05, | |
| "loss": 1.7889, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6146481570129249, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.660288734900019e-05, | |
| "loss": 1.7892, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.6165629487793203, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.6579046205542656e-05, | |
| "loss": 1.8961, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.6184777405457157, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.655513894480049e-05, | |
| "loss": 1.7397, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.6203925323121111, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.653116580703393e-05, | |
| "loss": 1.8193, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.6223073240785064, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.6507127033165243e-05, | |
| "loss": 1.8165, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6242221158449018, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.6483022864776327e-05, | |
| "loss": 1.7779, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.6261369076112973, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.645885354410627e-05, | |
| "loss": 1.8215, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.6280516993776927, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.643461931404893e-05, | |
| "loss": 1.8383, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.6299664911440881, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.641032041815046e-05, | |
| "loss": 1.7761, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.6318812829104835, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.6385957100606912e-05, | |
| "loss": 1.7577, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6337960746768789, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.636152960626173e-05, | |
| "loss": 1.7726, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.6357108664432743, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.6337038180603332e-05, | |
| "loss": 1.7556, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.6376256582096697, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.631248306976261e-05, | |
| "loss": 1.8344, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.6395404499760651, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.6287864520510483e-05, | |
| "loss": 1.7506, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.6414552417424605, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.62631827802554e-05, | |
| "loss": 1.84, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6433700335088559, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.6238438097040866e-05, | |
| "loss": 1.6987, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.6452848252752513, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.621363071954293e-05, | |
| "loss": 1.7537, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.6471996170416467, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.6188760897067712e-05, | |
| "loss": 1.7974, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.6491144088080422, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.616382887954887e-05, | |
| "loss": 1.7815, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.6510292005744376, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.6138834917545115e-05, | |
| "loss": 1.8334, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6529439923408329, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.611377926223767e-05, | |
| "loss": 1.7674, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.6548587841072283, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.6088662165427767e-05, | |
| "loss": 1.789, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.6567735758736237, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.6063483879534092e-05, | |
| "loss": 1.7917, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.6586883676400191, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.6038244657590266e-05, | |
| "loss": 1.7639, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.6606031594064146, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.6012944753242304e-05, | |
| "loss": 1.7936, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.66251795117281, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.598758442074605e-05, | |
| "loss": 1.7632, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.6644327429392054, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.5962163914964644e-05, | |
| "loss": 1.7101, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.6663475347056008, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.5936683491365933e-05, | |
| "loss": 1.7726, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.6682623264719961, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.5911143406019926e-05, | |
| "loss": 1.7999, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.6701771182383915, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.5885543915596215e-05, | |
| "loss": 1.863, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.672091910004787, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.5859885277361395e-05, | |
| "loss": 1.8356, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.6740067017711824, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.583416774917647e-05, | |
| "loss": 1.7902, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.6759214935375778, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.580839158949427e-05, | |
| "loss": 1.7567, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.6778362853039732, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.5782557057356866e-05, | |
| "loss": 1.7907, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.6797510770703686, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.575666441239293e-05, | |
| "loss": 1.7513, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.681665868836764, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.5730713914815174e-05, | |
| "loss": 1.853, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.6835806606031594, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.570470582541769e-05, | |
| "loss": 1.829, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.6854954523695548, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.5678640405573357e-05, | |
| "loss": 1.8129, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.6874102441359502, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.5652517917231212e-05, | |
| "loss": 1.8097, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.6893250359023456, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.5626338622913807e-05, | |
| "loss": 1.7673, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.691239827668741, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.5600102785714567e-05, | |
| "loss": 1.7461, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.6931546194351365, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.5573810669295176e-05, | |
| "loss": 1.7969, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.6950694112015319, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.554746253788288e-05, | |
| "loss": 1.7485, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.6969842029679273, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.5521058656267874e-05, | |
| "loss": 1.7933, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.6988989947343226, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.5494599289800627e-05, | |
| "loss": 1.8053, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.700813786500718, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.54680847043892e-05, | |
| "loss": 1.7844, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.7027285782671134, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.5441515166496593e-05, | |
| "loss": 1.8336, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.7046433700335089, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.5414890943138068e-05, | |
| "loss": 1.8394, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.7065581617999043, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.5388212301878445e-05, | |
| "loss": 1.7923, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.7084729535662997, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.5361479510829434e-05, | |
| "loss": 1.7798, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7103877453326951, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.5334692838646927e-05, | |
| "loss": 1.7606, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.7123025370990905, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.5307852554528318e-05, | |
| "loss": 1.7701, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.7142173288654858, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.5280958928209763e-05, | |
| "loss": 1.7918, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.7161321206318813, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.5254012229963509e-05, | |
| "loss": 1.7718, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.7180469123982767, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.5227012730595146e-05, | |
| "loss": 1.7738, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7199617041646721, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.5199960701440902e-05, | |
| "loss": 1.7932, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.7218764959310675, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.5172856414364916e-05, | |
| "loss": 1.7774, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.7237912876974629, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.5145700141756496e-05, | |
| "loss": 1.7727, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.7257060794638583, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.5118492156527395e-05, | |
| "loss": 1.8104, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.7276208712302538, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.5091232732109053e-05, | |
| "loss": 1.7703, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7295356629966491, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.5063922142449857e-05, | |
| "loss": 1.7688, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.7314504547630445, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.5036560662012405e-05, | |
| "loss": 1.7204, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.7333652465294399, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.5009148565770707e-05, | |
| "loss": 1.7462, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.7352800382958353, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.498168612920746e-05, | |
| "loss": 1.7404, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.7371948300622307, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.4954173628311262e-05, | |
| "loss": 1.7549, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.7391096218286262, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.492661133957384e-05, | |
| "loss": 1.8004, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.7410244135950216, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.4898999539987273e-05, | |
| "loss": 1.7897, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.742939205361417, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.487133850704121e-05, | |
| "loss": 1.7573, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.7448539971278123, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.4843628518720076e-05, | |
| "loss": 1.8918, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.7467687888942077, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.4815869853500286e-05, | |
| "loss": 1.7759, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7486835806606031, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.4788062790347437e-05, | |
| "loss": 1.8615, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.7505983724269986, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.4760207608713515e-05, | |
| "loss": 1.7365, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.752513164193394, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.4732304588534073e-05, | |
| "loss": 1.783, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.7544279559597894, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.4704354010225436e-05, | |
| "loss": 1.7981, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.7563427477261848, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.4676356154681867e-05, | |
| "loss": 1.8287, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7582575394925802, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.464831130327275e-05, | |
| "loss": 1.7761, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.7601723312589755, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.4620219737839766e-05, | |
| "loss": 1.8013, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.762087123025371, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.4592081740694051e-05, | |
| "loss": 1.8035, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.7640019147917664, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.4563897594613368e-05, | |
| "loss": 1.8472, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.7659167065581618, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.453566758283926e-05, | |
| "loss": 1.7201, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7678314983245572, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.450739198907421e-05, | |
| "loss": 1.7396, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.7697462900909526, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.4479071097478778e-05, | |
| "loss": 1.8413, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.771661081857348, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.4450705192668763e-05, | |
| "loss": 1.8091, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.7735758736237435, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.442229455971232e-05, | |
| "loss": 1.7785, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.7754906653901388, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.4393839484127117e-05, | |
| "loss": 1.7627, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.7774054571565342, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.4365340251877446e-05, | |
| "loss": 1.7348, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.7793202489229296, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.4336797149371377e-05, | |
| "loss": 1.8185, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.781235040689325, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.4308210463457842e-05, | |
| "loss": 1.7564, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.7831498324557205, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.4279580481423778e-05, | |
| "loss": 1.7314, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.7850646242221159, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.4250907490991244e-05, | |
| "loss": 1.7787, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7869794159885113, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.4222191780314508e-05, | |
| "loss": 1.7753, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.7888942077549067, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.4193433637977165e-05, | |
| "loss": 1.7371, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.790808999521302, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.416463335298924e-05, | |
| "loss": 1.7923, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.7927237912876974, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.4135791214784272e-05, | |
| "loss": 1.8241, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.7946385830540929, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.4106907513216412e-05, | |
| "loss": 1.7248, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.7965533748204883, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.4077982538557511e-05, | |
| "loss": 1.8105, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.7984681665868837, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.4049016581494204e-05, | |
| "loss": 1.8315, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.8003829583532791, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.402000993312498e-05, | |
| "loss": 1.7119, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.8022977501196745, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.3990962884957267e-05, | |
| "loss": 1.7673, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.8042125418860699, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.3961875728904495e-05, | |
| "loss": 1.8066, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8061273336524652, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.3932748757283165e-05, | |
| "loss": 1.7593, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.8080421254188607, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.3903582262809918e-05, | |
| "loss": 1.8116, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.8099569171852561, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.3874376538598574e-05, | |
| "loss": 1.789, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.8118717089516515, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.3845131878157214e-05, | |
| "loss": 1.7977, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.8137865007180469, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.3815848575385207e-05, | |
| "loss": 1.8037, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.8157012924844423, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.3786526924570262e-05, | |
| "loss": 1.7475, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.8176160842508378, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.3757167220385483e-05, | |
| "loss": 1.7709, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.8195308760172332, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.3727769757886388e-05, | |
| "loss": 1.7576, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.8214456677836285, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.3698334832507962e-05, | |
| "loss": 1.8164, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.8233604595500239, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.366886274006168e-05, | |
| "loss": 1.7753, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8252752513164193, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.3639353776732523e-05, | |
| "loss": 1.7188, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.8271900430828147, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.3609808239076025e-05, | |
| "loss": 1.7456, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.8291048348492102, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.3580226424015273e-05, | |
| "loss": 1.7591, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.8310196266156056, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.3550608628837933e-05, | |
| "loss": 1.8017, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.832934418382001, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.352095515119326e-05, | |
| "loss": 1.7858, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.8348492101483964, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.3491266289089107e-05, | |
| "loss": 1.7865, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.8367640019147917, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.3461542340888921e-05, | |
| "loss": 1.7595, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.8386787936811871, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.3431783605308761e-05, | |
| "loss": 1.8015, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.8405935854475826, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.3401990381414287e-05, | |
| "loss": 1.7645, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.842508377213978, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.3372162968617757e-05, | |
| "loss": 1.7324, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8444231689803734, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.3342301666675013e-05, | |
| "loss": 1.7304, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.8463379607467688, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.3312406775682471e-05, | |
| "loss": 1.814, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.8482527525131642, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.328247859607411e-05, | |
| "loss": 1.7776, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.8501675442795597, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.3252517428618448e-05, | |
| "loss": 1.7183, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.852082336045955, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.3222523574415516e-05, | |
| "loss": 1.7699, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.8539971278123504, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.3192497334893842e-05, | |
| "loss": 1.8766, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.8559119195787458, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.316243901180741e-05, | |
| "loss": 1.7422, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.8578267113451412, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.3132348907232639e-05, | |
| "loss": 1.7802, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.8597415031115366, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.310222732356534e-05, | |
| "loss": 1.7016, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.861656294877932, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.3072074563517676e-05, | |
| "loss": 1.7917, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8635710866443275, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.3041890930115125e-05, | |
| "loss": 1.7344, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.8654858784107229, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.3011676726693432e-05, | |
| "loss": 1.7282, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.8674006701771182, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.298143225689556e-05, | |
| "loss": 1.7817, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.8693154619435136, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.2951157824668645e-05, | |
| "loss": 1.7465, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.871230253709909, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.2920853734260925e-05, | |
| "loss": 1.7836, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.8731450454763044, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.2890520290218698e-05, | |
| "loss": 1.762, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.8750598372426999, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.286015779738326e-05, | |
| "loss": 1.7934, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.8769746290090953, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.2829766560887837e-05, | |
| "loss": 1.7687, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.8788894207754907, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.2799346886154513e-05, | |
| "loss": 1.8095, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.8808042125418861, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.2768899078891174e-05, | |
| "loss": 1.7651, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8827190043082814, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.2738423445088429e-05, | |
| "loss": 1.7002, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.8846337960746768, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.2707920291016526e-05, | |
| "loss": 1.804, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.8865485878410723, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.2677389923222297e-05, | |
| "loss": 1.7597, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.8884633796074677, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.2646832648526048e-05, | |
| "loss": 1.7853, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.8903781713738631, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.2616248774018503e-05, | |
| "loss": 1.7303, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.8922929631402585, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.2585638607057698e-05, | |
| "loss": 1.7358, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.8942077549066539, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.25550024552659e-05, | |
| "loss": 1.8593, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.8961225466730494, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.2524340626526521e-05, | |
| "loss": 1.8011, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.8980373384394447, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.2493653428981014e-05, | |
| "loss": 1.7807, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.8999521302058401, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.2462941171025777e-05, | |
| "loss": 1.7492, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9018669219722355, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.2432204161309063e-05, | |
| "loss": 1.7557, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.9037817137386309, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.2401442708727869e-05, | |
| "loss": 1.779, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.9056965055050263, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.2370657122424835e-05, | |
| "loss": 1.7955, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.9076112972714218, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.2339847711785139e-05, | |
| "loss": 1.7179, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.9095260890378172, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.2309014786433381e-05, | |
| "loss": 1.8023, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.9114408808042126, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.2278158656230486e-05, | |
| "loss": 1.8111, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.9133556725706079, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.224727963127057e-05, | |
| "loss": 1.7841, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.9152704643370033, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.2216378021877835e-05, | |
| "loss": 1.775, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.9171852561033987, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.2185454138603458e-05, | |
| "loss": 1.7291, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.9191000478697942, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.215450829222245e-05, | |
| "loss": 1.7993, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9210148396361896, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.2123540793730554e-05, | |
| "loss": 1.7748, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.922929631402585, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.2092551954341104e-05, | |
| "loss": 1.7842, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.9248444231689804, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.2061542085481904e-05, | |
| "loss": 1.7519, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.9267592149353758, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.2030511498792095e-05, | |
| "loss": 1.8114, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.9286740067017711, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.199946050611903e-05, | |
| "loss": 1.7339, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.9305887984681666, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.1968389419515134e-05, | |
| "loss": 1.8057, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.932503590234562, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.1937298551234769e-05, | |
| "loss": 1.7458, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.9344183820009574, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.1906188213731099e-05, | |
| "loss": 1.7539, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.9363331737673528, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 1.187505871965294e-05, | |
| "loss": 1.8346, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.9382479655337482, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.1843910381841637e-05, | |
| "loss": 1.7451, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9401627573001436, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.1812743513327896e-05, | |
| "loss": 1.7659, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.9420775490665391, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.1781558427328662e-05, | |
| "loss": 1.8026, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.9439923408329344, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.1750355437243947e-05, | |
| "loss": 1.7427, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.9459071325993298, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.1719134856653704e-05, | |
| "loss": 1.7436, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.9478219243657252, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.1687896999314663e-05, | |
| "loss": 1.8219, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.9497367161321206, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.1656642179157173e-05, | |
| "loss": 1.757, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.951651507898516, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.1625370710282067e-05, | |
| "loss": 1.7482, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.9535662996649115, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.1594082906957478e-05, | |
| "loss": 1.8083, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.9554810914313069, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.1562779083615702e-05, | |
| "loss": 1.8082, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.9573958831977023, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.153145955485003e-05, | |
| "loss": 1.7942, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9593106749640976, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.1500124635411592e-05, | |
| "loss": 1.7538, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.961225466730493, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.146877464020618e-05, | |
| "loss": 1.7099, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.9631402584968884, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.1437409884291097e-05, | |
| "loss": 1.821, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.9650550502632839, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.140603068287199e-05, | |
| "loss": 1.8018, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.9669698420296793, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.1374637351299672e-05, | |
| "loss": 1.8382, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.9688846337960747, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.1343230205066963e-05, | |
| "loss": 1.8094, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.9707994255624701, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.131180955980552e-05, | |
| "loss": 1.7568, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.9727142173288655, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.128037573128265e-05, | |
| "loss": 1.743, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.9746290090952608, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.124892903539816e-05, | |
| "loss": 1.8106, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.9765438008616563, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.1217469788181158e-05, | |
| "loss": 1.7705, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.9784585926280517, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.1185998305786902e-05, | |
| "loss": 1.8163, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.9803733843944471, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.1154514904493599e-05, | |
| "loss": 1.8028, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.9822881761608425, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.1123019900699239e-05, | |
| "loss": 1.8154, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.9842029679272379, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.1091513610918415e-05, | |
| "loss": 1.7655, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.9861177596936334, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.1059996351779139e-05, | |
| "loss": 1.7482, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.9880325514600288, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.1028468440019666e-05, | |
| "loss": 1.7829, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.9899473432264241, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.0996930192485302e-05, | |
| "loss": 1.7006, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.9918621349928195, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.0965381926125224e-05, | |
| "loss": 1.7737, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.9937769267592149, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.0933823957989298e-05, | |
| "loss": 1.7464, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.9956917185256103, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.0902256605224885e-05, | |
| "loss": 1.8521, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.9976065102920058, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.0870680185073666e-05, | |
| "loss": 1.7895, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.9995213020584012, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.083909501486844e-05, | |
| "loss": 1.7476, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.0014360938247966, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 1.080750141202994e-05, | |
| "loss": 3.1048, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.003350885591192, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.0775899694063649e-05, | |
| "loss": 1.6357, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.0052656773575874, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.0744290178556604e-05, | |
| "loss": 1.7283, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.0071804691239827, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.0712673183174205e-05, | |
| "loss": 1.7224, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.0090952608903783, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.0681049025657015e-05, | |
| "loss": 1.6712, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0110100526567736, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.0649418023817583e-05, | |
| "loss": 1.7204, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.0129248444231689, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.061778049553724e-05, | |
| "loss": 1.6806, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.0148396361895644, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.0586136758762902e-05, | |
| "loss": 1.7295, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.0167544279559597, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.0554487131503874e-05, | |
| "loss": 1.7294, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.0186692197223552, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.0522831931828677e-05, | |
| "loss": 1.674, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.0205840114887506, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.049117147786181e-05, | |
| "loss": 1.6801, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.022498803255146, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.0459506087780593e-05, | |
| "loss": 1.6725, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.0244135950215414, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.042783607981194e-05, | |
| "loss": 1.6826, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.026328386787937, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.0396161772229185e-05, | |
| "loss": 1.6825, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.0282431785543322, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.0364483483348859e-05, | |
| "loss": 1.7038, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.0301579703207275, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.0332801531527516e-05, | |
| "loss": 1.7607, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.032072762087123, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.0301116235158516e-05, | |
| "loss": 1.7785, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.0339875538535184, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.0269427912668826e-05, | |
| "loss": 1.6801, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.035902345619914, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.0237736882515832e-05, | |
| "loss": 1.6535, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.0378171373863092, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.0206043463184127e-05, | |
| "loss": 1.7229, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.0397319291527047, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.0174347973182318e-05, | |
| "loss": 1.6847, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.0416467209191, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.0142650731039815e-05, | |
| "loss": 1.6263, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.0435615126854954, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.0110952055303647e-05, | |
| "loss": 1.6602, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.0454763044518909, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.0079252264535237e-05, | |
| "loss": 1.6483, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.0473910962182862, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.0047551677307226e-05, | |
| "loss": 1.7304, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.0493058879846817, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.0015850612200249e-05, | |
| "loss": 1.7147, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.051220679751077, | |
| "grad_norm": 1.75, | |
| "learning_rate": 9.984149387799754e-06, | |
| "loss": 1.655, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.0531354715174726, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 9.952448322692776e-06, | |
| "loss": 1.7411, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.0550502632838679, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 9.920747735464766e-06, | |
| "loss": 1.658, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.0569650550502634, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.889047944696354e-06, | |
| "loss": 1.6857, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.0588798468166587, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 9.857349268960186e-06, | |
| "loss": 1.7655, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.060794638583054, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 9.825652026817683e-06, | |
| "loss": 1.6613, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.0627094303494495, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 9.793956536815874e-06, | |
| "loss": 1.7267, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.0646242221158448, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 9.76226311748417e-06, | |
| "loss": 1.788, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.0665390138822404, | |
| "grad_norm": 1.75, | |
| "learning_rate": 9.730572087331177e-06, | |
| "loss": 1.6566, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.0684538056486357, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 9.698883764841484e-06, | |
| "loss": 1.707, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.0703685974150312, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 9.667198468472485e-06, | |
| "loss": 1.7831, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.0722833891814265, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 9.63551651665114e-06, | |
| "loss": 1.7095, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.0741981809478218, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 9.603838227770819e-06, | |
| "loss": 1.7534, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.0761129727142174, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 9.57216392018806e-06, | |
| "loss": 1.5954, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.0780277644806127, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 9.54049391221941e-06, | |
| "loss": 1.6686, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.0799425562470082, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 9.508828522138191e-06, | |
| "loss": 1.7209, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.0818573480134035, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 9.477168068171326e-06, | |
| "loss": 1.6337, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.083772139779799, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 9.445512868496128e-06, | |
| "loss": 1.6905, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.0856869315461943, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 9.413863241237101e-06, | |
| "loss": 1.6972, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.0876017233125896, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.382219504462766e-06, | |
| "loss": 1.7069, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.0895165150789852, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 9.350581976182418e-06, | |
| "loss": 1.6554, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.0914313068453805, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.31895097434299e-06, | |
| "loss": 1.6556, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.093346098611776, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 9.287326816825799e-06, | |
| "loss": 1.7767, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.0952608903781713, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 9.255709821443399e-06, | |
| "loss": 1.6971, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.0971756821445668, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 9.224100305936353e-06, | |
| "loss": 1.6232, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.0990904739109622, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 9.192498587970065e-06, | |
| "loss": 1.6941, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.1010052656773577, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 9.160904985131564e-06, | |
| "loss": 1.6419, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.102920057443753, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 9.129319814926339e-06, | |
| "loss": 1.7512, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.1048348492101483, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 9.097743394775116e-06, | |
| "loss": 1.749, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.1067496409765438, | |
| "grad_norm": 1.75, | |
| "learning_rate": 9.066176042010705e-06, | |
| "loss": 1.723, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.1086644327429391, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 9.034618073874777e-06, | |
| "loss": 1.7473, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.1105792245093347, | |
| "grad_norm": 1.875, | |
| "learning_rate": 9.003069807514702e-06, | |
| "loss": 1.6308, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.11249401627573, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 8.971531559980334e-06, | |
| "loss": 1.7228, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.1144088080421255, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 8.940003648220863e-06, | |
| "loss": 1.6956, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.1163235998085208, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 8.908486389081587e-06, | |
| "loss": 1.7137, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.1182383915749163, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 8.876980099300764e-06, | |
| "loss": 1.7155, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.1201531833413116, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 8.845485095506401e-06, | |
| "loss": 1.6971, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.122067975107707, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 8.8140016942131e-06, | |
| "loss": 1.7349, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.1239827668741025, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 8.78253021181884e-06, | |
| "loss": 1.7291, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.1258975586404978, | |
| "grad_norm": 1.75, | |
| "learning_rate": 8.751070964601845e-06, | |
| "loss": 1.6818, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.1278123504068933, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 8.719624268717351e-06, | |
| "loss": 1.6302, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.1297271421732886, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 8.688190440194483e-06, | |
| "loss": 1.7592, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.1316419339396842, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 8.656769794933042e-06, | |
| "loss": 1.7508, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.1335567257060795, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 8.625362648700332e-06, | |
| "loss": 1.6968, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.135471517472475, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 8.593969317128015e-06, | |
| "loss": 1.7048, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.1373863092388703, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 8.562590115708906e-06, | |
| "loss": 1.6671, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.1393011010052656, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 8.531225359793825e-06, | |
| "loss": 1.7416, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.1412158927716611, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 8.49987536458841e-06, | |
| "loss": 1.7597, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.1431306845380564, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 8.468540445149973e-06, | |
| "loss": 1.6507, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.145045476304452, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 8.437220916384301e-06, | |
| "loss": 1.6958, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.1469602680708473, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 8.405917093042526e-06, | |
| "loss": 1.72, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.1488750598372426, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 8.374629289717937e-06, | |
| "loss": 1.7362, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.1507898516036381, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 8.343357820842829e-06, | |
| "loss": 1.6909, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.1527046433700334, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 8.31210300068534e-06, | |
| "loss": 1.641, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.154619435136429, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 8.280865143346301e-06, | |
| "loss": 1.7413, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.1565342269028243, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 8.249644562756056e-06, | |
| "loss": 1.6597, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.1584490186692198, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 8.218441572671343e-06, | |
| "loss": 1.6769, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.160363810435615, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 8.187256486672106e-06, | |
| "loss": 1.6976, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.1622786022020106, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 8.156089618158366e-06, | |
| "loss": 1.6065, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.164193393968406, | |
| "grad_norm": 1.75, | |
| "learning_rate": 8.12494128034706e-06, | |
| "loss": 1.7632, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.1661081857348012, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 8.093811786268905e-06, | |
| "loss": 1.6849, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.1680229775011968, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 8.06270144876523e-06, | |
| "loss": 1.7192, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.169937769267592, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 8.03161058048487e-06, | |
| "loss": 1.6864, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.1718525610339876, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 8.000539493880972e-06, | |
| "loss": 1.704, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.173767352800383, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7.969488501207909e-06, | |
| "loss": 1.6947, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.1756821445667784, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 7.938457914518098e-06, | |
| "loss": 1.7126, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.1775969363331737, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7.907448045658899e-06, | |
| "loss": 1.7502, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.1795117280995693, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 7.876459206269446e-06, | |
| "loss": 1.7348, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.1814265198659646, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 7.845491707777551e-06, | |
| "loss": 1.6578, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.18334131163236, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.814545861396543e-06, | |
| "loss": 1.7479, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.1852561033987554, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 7.783621978122167e-06, | |
| "loss": 1.7027, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.1871708951651507, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 7.752720368729436e-06, | |
| "loss": 1.6828, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.1890856869315463, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 7.721841343769518e-06, | |
| "loss": 1.689, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.1910004786979416, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 7.69098521356662e-06, | |
| "loss": 1.6498, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.192915270464337, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 7.660152288214865e-06, | |
| "loss": 1.6855, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.1948300622307324, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 7.629342877575169e-06, | |
| "loss": 1.7234, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.196744853997128, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 7.598557291272133e-06, | |
| "loss": 1.7746, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.1986596457635232, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7.567795838690941e-06, | |
| "loss": 1.7283, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.2005744375299185, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 7.537058828974226e-06, | |
| "loss": 1.7204, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.202489229296314, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 7.506346571018992e-06, | |
| "loss": 1.7417, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.2044040210627094, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 7.475659373473481e-06, | |
| "loss": 1.6638, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.206318812829105, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 7.444997544734105e-06, | |
| "loss": 1.7001, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.2082336045955002, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 7.414361392942307e-06, | |
| "loss": 1.7401, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.2101483963618955, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7.383751225981503e-06, | |
| "loss": 1.7466, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.212063188128291, | |
| "grad_norm": 1.875, | |
| "learning_rate": 7.353167351473955e-06, | |
| "loss": 1.6681, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.2139779798946864, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 7.322610076777707e-06, | |
| "loss": 1.7014, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.215892771661082, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.292079708983475e-06, | |
| "loss": 1.7218, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.2178075634274772, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 7.261576554911575e-06, | |
| "loss": 1.7206, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.2197223551938727, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 7.2311009211088255e-06, | |
| "loss": 1.6895, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.221637146960268, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.20065311384549e-06, | |
| "loss": 1.744, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.2235519387266636, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 7.170233439112164e-06, | |
| "loss": 1.7217, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.2254667304930589, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 7.139842202616741e-06, | |
| "loss": 1.6799, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.2273815222594542, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 7.109479709781302e-06, | |
| "loss": 1.7117, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.2292963140258497, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.079146265739079e-06, | |
| "loss": 1.6948, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.231211105792245, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 7.048842175331356e-06, | |
| "loss": 1.7343, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.2331258975586405, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 7.0185677431044404e-06, | |
| "loss": 1.7078, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.2350406893250359, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 6.988323273306569e-06, | |
| "loss": 1.7168, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.2369554810914314, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 6.958109069884879e-06, | |
| "loss": 1.6997, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.2388702728578267, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 6.9279254364823265e-06, | |
| "loss": 1.6204, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.2407850646242222, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 6.897772676434663e-06, | |
| "loss": 1.727, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.2426998563906175, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 6.867651092767366e-06, | |
| "loss": 1.7278, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.2446146481570128, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 6.837560988192593e-06, | |
| "loss": 1.7087, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.2465294399234084, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 6.807502665106164e-06, | |
| "loss": 1.6614, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.2484442316898037, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 6.777476425584486e-06, | |
| "loss": 1.7264, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.2503590234561992, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 6.747482571381556e-06, | |
| "loss": 1.701, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.2522738152225945, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 6.717521403925892e-06, | |
| "loss": 1.7061, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.2541886069889898, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 6.687593224317533e-06, | |
| "loss": 1.6551, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.2561033987553853, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 6.657698333324991e-06, | |
| "loss": 1.731, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.2580181905217809, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 6.627837031382246e-06, | |
| "loss": 1.7075, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.2599329822881762, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 6.598009618585717e-06, | |
| "loss": 1.6765, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.2618477740545715, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 6.568216394691245e-06, | |
| "loss": 1.7378, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.263762565820967, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 6.538457659111084e-06, | |
| "loss": 1.7609, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.2656773575873623, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 6.5087337109109e-06, | |
| "loss": 1.7036, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.2675921493537579, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 6.479044848806739e-06, | |
| "loss": 1.6546, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.2695069411201532, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 6.4493913711620685e-06, | |
| "loss": 1.7018, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.2714217328865485, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 6.419773575984727e-06, | |
| "loss": 1.7357, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.273336524652944, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 6.390191760923978e-06, | |
| "loss": 1.6928, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.2752513164193395, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 6.360646223267477e-06, | |
| "loss": 1.6623, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.2771661081857348, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 6.3311372599383245e-06, | |
| "loss": 1.6921, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.2790808999521301, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 6.301665167492037e-06, | |
| "loss": 1.7036, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.2809956917185257, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 6.272230242113613e-06, | |
| "loss": 1.7099, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.282910483484921, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 6.242832779614521e-06, | |
| "loss": 1.6826, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.2848252752513165, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 6.213473075429741e-06, | |
| "loss": 1.7058, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.2867400670177118, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 6.184151424614795e-06, | |
| "loss": 1.6677, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.2886548587841071, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 6.154868121842788e-06, | |
| "loss": 1.7125, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.2905696505505027, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 6.1256234614014256e-06, | |
| "loss": 1.6581, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.292484442316898, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 6.096417737190085e-06, | |
| "loss": 1.7104, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.2943992340832935, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 6.067251242716838e-06, | |
| "loss": 1.6612, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.2963140258496888, | |
| "grad_norm": 1.75, | |
| "learning_rate": 6.038124271095507e-06, | |
| "loss": 1.6501, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.2982288176160843, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 6.0090371150427375e-06, | |
| "loss": 1.7283, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.3001436093824796, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 5.979990066875022e-06, | |
| "loss": 1.699, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.3020584011488752, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 5.950983418505799e-06, | |
| "loss": 1.7458, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.3039731929152705, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 5.922017461442492e-06, | |
| "loss": 1.6889, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.3058879846816658, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 5.893092486783594e-06, | |
| "loss": 1.6935, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.3078027764480613, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 5.864208785215732e-06, | |
| "loss": 1.6641, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.3097175682144566, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 5.835366647010767e-06, | |
| "loss": 1.7062, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.3116323599808521, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 5.8065663620228404e-06, | |
| "loss": 1.7008, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.3135471517472475, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 5.777808219685496e-06, | |
| "loss": 1.7002, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.3154619435136428, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 5.749092509008761e-06, | |
| "loss": 1.6896, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.3173767352800383, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 5.720419518576223e-06, | |
| "loss": 1.7014, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.3192915270464338, | |
| "grad_norm": 1.75, | |
| "learning_rate": 5.691789536542161e-06, | |
| "loss": 1.6799, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.3212063188128291, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 5.6632028506286266e-06, | |
| "loss": 1.6558, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.3231211105792244, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 5.634659748122552e-06, | |
| "loss": 1.6286, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.32503590234562, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 5.606160515872886e-06, | |
| "loss": 1.6983, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.3269506941120153, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 5.57770544028768e-06, | |
| "loss": 1.7115, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.3288654858784108, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 5.5492948073312406e-06, | |
| "loss": 1.719, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.330780277644806, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 5.520928902521221e-06, | |
| "loss": 1.7074, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.3326950694112014, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 5.492608010925793e-06, | |
| "loss": 1.7135, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.334609861177597, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 5.46433241716074e-06, | |
| "loss": 1.6395, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.3365246529439925, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 5.436102405386636e-06, | |
| "loss": 1.7543, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.3384394447103878, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 5.407918259305951e-06, | |
| "loss": 1.6431, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.340354236476783, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 5.379780262160237e-06, | |
| "loss": 1.7222, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.3422690282431786, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 5.3516886967272485e-06, | |
| "loss": 1.7227, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.344183820009574, | |
| "grad_norm": 1.75, | |
| "learning_rate": 5.323643845318135e-06, | |
| "loss": 1.7426, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.3460986117759695, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 5.295645989774565e-06, | |
| "loss": 1.7319, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.3480134035423648, | |
| "grad_norm": 1.75, | |
| "learning_rate": 5.26769541146593e-06, | |
| "loss": 1.6958, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.34992819530876, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 5.239792391286492e-06, | |
| "loss": 1.682, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.3518429870751556, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 5.211937209652567e-06, | |
| "loss": 1.6474, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.353757778841551, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 5.1841301464997206e-06, | |
| "loss": 1.6762, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.3556725706079464, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 5.156371481279928e-06, | |
| "loss": 1.6729, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.3575873623743417, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 5.128661492958793e-06, | |
| "loss": 1.7092, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.3595021541407373, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 5.101000460012731e-06, | |
| "loss": 1.6943, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.3614169459071326, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 5.073388660426164e-06, | |
| "loss": 1.6758, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.363331737673528, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 5.04582637168874e-06, | |
| "loss": 1.671, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.3652465294399234, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 5.018313870792544e-06, | |
| "loss": 1.6675, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.3671613212063187, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 4.990851434229295e-06, | |
| "loss": 1.7074, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.3690761129727143, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 4.9634393379875986e-06, | |
| "loss": 1.6558, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.3709909047391096, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 4.936077857550141e-06, | |
| "loss": 1.7064, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.372905696505505, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 4.908767267890952e-06, | |
| "loss": 1.646, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.3748204882719004, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 4.8815078434726075e-06, | |
| "loss": 1.7207, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.3767352800382957, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 4.854299858243505e-06, | |
| "loss": 1.7459, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.3786500718046912, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 4.827143585635085e-06, | |
| "loss": 1.677, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.3805648635710868, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 4.800039298559101e-06, | |
| "loss": 1.682, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.382479655337482, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 4.772987269404855e-06, | |
| "loss": 1.6784, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.3843944471038774, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 4.745987770036494e-06, | |
| "loss": 1.7062, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.386309238870273, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 4.719041071790238e-06, | |
| "loss": 1.6879, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.3882240306366682, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 4.692147445471687e-06, | |
| "loss": 1.7091, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.3901388224030637, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 4.665307161353073e-06, | |
| "loss": 1.721, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.392053614169459, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 4.638520489170572e-06, | |
| "loss": 1.6974, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.3939684059358544, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 4.611787698121558e-06, | |
| "loss": 1.7036, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.39588319770225, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 4.585109056861936e-06, | |
| "loss": 1.6918, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.3977979894686454, | |
| "grad_norm": 1.75, | |
| "learning_rate": 4.558484833503407e-06, | |
| "loss": 1.6522, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.3997127812350407, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 4.531915295610805e-06, | |
| "loss": 1.682, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.401627573001436, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 4.505400710199376e-06, | |
| "loss": 1.7143, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.4035423647678316, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 4.478941343732125e-06, | |
| "loss": 1.7184, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.4054571565342269, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 4.452537462117123e-06, | |
| "loss": 1.7365, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.4073719483006224, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 4.426189330704826e-06, | |
| "loss": 1.6457, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.4092867400670177, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 4.3998972142854334e-06, | |
| "loss": 1.6866, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.411201531833413, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 4.373661377086195e-06, | |
| "loss": 1.6959, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.4131163235998085, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 4.3474820827687894e-06, | |
| "loss": 1.7323, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.4150311153662039, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 4.321359594426644e-06, | |
| "loss": 1.6982, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.4169459071325994, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 4.295294174582315e-06, | |
| "loss": 1.6859, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.4188606988989947, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 4.2692860851848295e-06, | |
| "loss": 1.6616, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.4207754906653902, | |
| "grad_norm": 1.75, | |
| "learning_rate": 4.243335587607074e-06, | |
| "loss": 1.6983, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.4226902824317855, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 4.217442942643138e-06, | |
| "loss": 1.6405, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.424605074198181, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 4.191608410505732e-06, | |
| "loss": 1.6706, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.4265198659645764, | |
| "grad_norm": 1.875, | |
| "learning_rate": 4.165832250823534e-06, | |
| "loss": 1.7247, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.4284346577309717, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 4.140114722638609e-06, | |
| "loss": 1.713, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.4303494494973672, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 4.114456084403785e-06, | |
| "loss": 1.7519, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.4322642412637625, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 4.088856593980078e-06, | |
| "loss": 1.7403, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.434179033030158, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 4.06331650863407e-06, | |
| "loss": 1.6786, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.4360938247965533, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 4.03783608503536e-06, | |
| "loss": 1.6476, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.4380086165629486, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 4.0124155792539496e-06, | |
| "loss": 1.8036, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.4399234083293442, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.987055246757701e-06, | |
| "loss": 1.7387, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.4418382000957397, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 3.961755342409737e-06, | |
| "loss": 1.7148, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.443752991862135, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 3.936516120465914e-06, | |
| "loss": 1.621, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.4456677836285303, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 3.911337834572235e-06, | |
| "loss": 1.6647, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.4475825753949259, | |
| "grad_norm": 1.75, | |
| "learning_rate": 3.886220737762328e-06, | |
| "loss": 1.6833, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.4494973671613212, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.861165082454888e-06, | |
| "loss": 1.7302, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.4514121589277167, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 3.836171120451131e-06, | |
| "loss": 1.7396, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.453326950694112, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 3.811239102932289e-06, | |
| "loss": 1.7763, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.4552417424605073, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 3.7863692804570707e-06, | |
| "loss": 1.734, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.4571565342269028, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.761561902959139e-06, | |
| "loss": 1.6783, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.4590713259932984, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.7368172197446007e-06, | |
| "loss": 1.6689, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.4609861177596937, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 3.7121354794895216e-06, | |
| "loss": 1.6886, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.462900909526089, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 3.6875169302373938e-06, | |
| "loss": 1.6309, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.4648157012924845, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 3.6629618193966744e-06, | |
| "loss": 1.7063, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.4667304930588798, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 3.6384703937382714e-06, | |
| "loss": 1.7162, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.4686452848252753, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 3.6140428993930922e-06, | |
| "loss": 1.7338, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.4705600765916707, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.589679581849539e-06, | |
| "loss": 1.6977, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.472474868358066, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 3.5653806859510743e-06, | |
| "loss": 1.6789, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.4743896601244615, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 3.5411464558937302e-06, | |
| "loss": 1.6898, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.4763044518908568, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.5169771352236782e-06, | |
| "loss": 1.7292, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.4782192436572523, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.4928729668347616e-06, | |
| "loss": 1.7046, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.4801340354236476, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.4688341929660776e-06, | |
| "loss": 1.7028, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.4820488271900432, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 3.444861055199512e-06, | |
| "loss": 1.6834, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.4839636189564385, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 3.420953794457349e-06, | |
| "loss": 1.7299, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.485878410722834, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 3.397112650999811e-06, | |
| "loss": 1.6711, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.4877932024892293, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 3.37333786442268e-06, | |
| "loss": 1.6332, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.4897079942556246, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 3.349629673654858e-06, | |
| "loss": 1.673, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.4916227860220201, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 3.32598831695599e-06, | |
| "loss": 1.6594, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.4935375777884154, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 3.3024140319140617e-06, | |
| "loss": 1.6547, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.495452369554811, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 3.2789070554430003e-06, | |
| "loss": 1.6371, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.4973671613212063, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 3.2554676237803117e-06, | |
| "loss": 1.7326, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.4992819530876016, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 3.2320959724847e-06, | |
| "loss": 1.7118, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.5011967448539971, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 3.2087923364336904e-06, | |
| "loss": 1.8037, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.5031115366203927, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.1855569498212857e-06, | |
| "loss": 1.7526, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.505026328386788, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.1623900461555933e-06, | |
| "loss": 1.665, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.5069411201531833, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 3.1392918582565037e-06, | |
| "loss": 1.6528, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.5088559119195788, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.1162626182533207e-06, | |
| "loss": 1.7152, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.510770703685974, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.093302557582457e-06, | |
| "loss": 1.661, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.5126854954523696, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 3.070411906985088e-06, | |
| "loss": 1.7286, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.514600287218765, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.0475908965048374e-06, | |
| "loss": 1.7394, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.5165150789851602, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.0248397554854813e-06, | |
| "loss": 1.6925, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.5184298707515558, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 3.002158712568615e-06, | |
| "loss": 1.6607, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.5203446625179513, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.979547995691383e-06, | |
| "loss": 1.6779, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.5222594542843466, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 2.9570078320841644e-06, | |
| "loss": 1.7353, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.524174246050742, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.9345384482683148e-06, | |
| "loss": 1.7012, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.5260890378171372, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 2.9121400700538593e-06, | |
| "loss": 1.6287, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.5280038295835328, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.8898129225372564e-06, | |
| "loss": 1.6926, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.5299186213499283, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.867557230099104e-06, | |
| "loss": 1.6822, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.5318334131163236, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 2.845373216401913e-06, | |
| "loss": 1.7203, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.533748204882719, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.823261104387833e-06, | |
| "loss": 1.7103, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.5356629966491144, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.801221116276436e-06, | |
| "loss": 1.6772, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.53757778841551, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.7792534735624687e-06, | |
| "loss": 1.7132, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.5394925801819053, | |
| "grad_norm": 1.75, | |
| "learning_rate": 2.757358397013625e-06, | |
| "loss": 1.8025, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.5414073719483006, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.7355361066683393e-06, | |
| "loss": 1.6785, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.5433221637146959, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.7137868218335674e-06, | |
| "loss": 1.6791, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.5452369554810914, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 2.692110761082577e-06, | |
| "loss": 1.7253, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.547151747247487, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.670508142252766e-06, | |
| "loss": 1.7035, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.5490665390138822, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 2.648979182443454e-06, | |
| "loss": 1.7488, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.5509813307802776, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.6275240980137272e-06, | |
| "loss": 1.704, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.552896122546673, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 2.6061431045802286e-06, | |
| "loss": 1.7235, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.5548109143130686, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.5848364170150307e-06, | |
| "loss": 1.6652, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.556725706079464, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.563604249443438e-06, | |
| "loss": 1.6524, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.5586404978458592, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.542446815241867e-06, | |
| "loss": 1.697, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.5605552896122545, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 2.521364327035678e-06, | |
| "loss": 1.6973, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.56247008137865, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.5003569966970574e-06, | |
| "loss": 1.7513, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.5643848731450456, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.4794250353428707e-06, | |
| "loss": 1.6145, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.566299664911441, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 2.458568653332557e-06, | |
| "loss": 1.7153, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.5682144566778362, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.437788060266002e-06, | |
| "loss": 1.6331, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.5701292484442317, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.4170834649814366e-06, | |
| "loss": 1.6747, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.572044040210627, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.3964550755533468e-06, | |
| "loss": 1.6055, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.5739588319770226, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.375903099290362e-06, | |
| "loss": 1.6992, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.5758736237434179, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 2.355427742733197e-06, | |
| "loss": 1.6433, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.5777884155098132, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.335029211652552e-06, | |
| "loss": 1.7133, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.5797032072762087, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.314707711047063e-06, | |
| "loss": 1.7327, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.5816179990426043, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.294463445141233e-06, | |
| "loss": 1.7631, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.5835327908089996, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.2742966173833835e-06, | |
| "loss": 1.7577, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.5854475825753949, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.254207430443599e-06, | |
| "loss": 1.7896, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.5873623743417902, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 2.2341960862117118e-06, | |
| "loss": 1.6763, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.5892771661081857, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 2.214262785795248e-06, | |
| "loss": 1.6878, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.5911919578745812, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 2.1944077295174284e-06, | |
| "loss": 1.6532, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.5931067496409765, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 2.174631116915137e-06, | |
| "loss": 1.7702, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.5950215414073718, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.1549331467369327e-06, | |
| "loss": 1.7115, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.5969363331737674, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 2.1353140169410347e-06, | |
| "loss": 1.6486, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.598851124940163, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 2.1157739246933507e-06, | |
| "loss": 1.7097, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.6007659167065582, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.0963130663654785e-06, | |
| "loss": 1.7174, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.6026807084729535, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.0769316375327497e-06, | |
| "loss": 1.6954, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.6045955002393488, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.0576298329722445e-06, | |
| "loss": 1.6773, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.6065102920057444, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 2.038407846660855e-06, | |
| "loss": 1.6202, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.6084250837721399, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.019265871773316e-06, | |
| "loss": 1.7177, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.6103398755385352, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.0002041006802843e-06, | |
| "loss": 1.7824, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.6122546673049305, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.981222724946383e-06, | |
| "loss": 1.7061, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.614169459071326, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9623219353283005e-06, | |
| "loss": 1.7551, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.6160842508377216, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.943501921772848e-06, | |
| "loss": 1.7165, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.6179990426041169, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.9247628734150725e-06, | |
| "loss": 1.6687, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.6199138343705122, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.9061049785763419e-06, | |
| "loss": 1.6888, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.6218286261369075, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.8875284247624625e-06, | |
| "loss": 1.6674, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.623743417903303, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.8690333986617827e-06, | |
| "loss": 1.7384, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.6256582096696985, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.8506200861433287e-06, | |
| "loss": 1.6367, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.6275730014360938, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.832288672254936e-06, | |
| "loss": 1.6592, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.6294877932024892, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.8140393412213719e-06, | |
| "loss": 1.7263, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.6314025849688847, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.7958722764425119e-06, | |
| "loss": 1.6543, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.63331737673528, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.7777876604914712e-06, | |
| "loss": 1.7082, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.6352321685016755, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.7597856751127919e-06, | |
| "loss": 1.7153, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.6371469602680708, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.7418665012205927e-06, | |
| "loss": 1.65, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.6390617520344661, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.7240303188967767e-06, | |
| "loss": 1.6985, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.6409765438008617, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.7062773073891958e-06, | |
| "loss": 1.6766, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.6428913355672572, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.6886076451098766e-06, | |
| "loss": 1.6786, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.6448061273336525, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.6710215096331971e-06, | |
| "loss": 1.7329, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.6467209191000478, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.6535190776941323e-06, | |
| "loss": 1.7428, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.6486357108664431, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.6361005251864525e-06, | |
| "loss": 1.6936, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.6505505026328386, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.6187660271609773e-06, | |
| "loss": 1.7386, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.6524652943992342, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.6015157578237939e-06, | |
| "loss": 1.7213, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.6543800861656295, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.584349890534531e-06, | |
| "loss": 1.6877, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.6562948779320248, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.5672685978045931e-06, | |
| "loss": 1.7153, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.6582096696984203, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.5502720512954472e-06, | |
| "loss": 1.7155, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.6601244614648158, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.5333604218168785e-06, | |
| "loss": 1.7235, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.6620392532312112, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.5165338793252937e-06, | |
| "loss": 1.6423, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.6639540449976065, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.4997925929219937e-06, | |
| "loss": 1.7088, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.6658688367640018, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.483136730851492e-06, | |
| "loss": 1.6486, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.6677836285303973, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.4665664604998053e-06, | |
| "loss": 1.6938, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.6696984202967928, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.4500819483927898e-06, | |
| "loss": 1.6819, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.6716132120631881, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.4336833601944577e-06, | |
| "loss": 1.6385, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.6735280038295834, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.4173708607053071e-06, | |
| "loss": 1.6798, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.675442795595979, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.4011446138606822e-06, | |
| "loss": 1.7299, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.6773575873623745, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.3850047827291057e-06, | |
| "loss": 1.715, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.6792723791287698, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.3689515295106626e-06, | |
| "loss": 1.6405, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.6811871708951651, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.352985015535344e-06, | |
| "loss": 1.7398, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.6831019626615604, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.3371054012614527e-06, | |
| "loss": 1.6731, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.685016754427956, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.3213128462739656e-06, | |
| "loss": 1.6651, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.6869315461943515, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.3056075092829546e-06, | |
| "loss": 1.6424, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.6888463379607468, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.2899895481219672e-06, | |
| "loss": 1.6476, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.690761129727142, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.2744591197464618e-06, | |
| "loss": 1.7422, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.6926759214935376, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.2590163802322108e-06, | |
| "loss": 1.6761, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.694590713259933, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.2436614847737526e-06, | |
| "loss": 1.7296, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.6965055050263285, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.2283945876828107e-06, | |
| "loss": 1.671, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.6984202967927238, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.2132158423867645e-06, | |
| "loss": 1.7288, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.700335088559119, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.198125401427085e-06, | |
| "loss": 1.7769, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.7022498803255146, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.1831234164578242e-06, | |
| "loss": 1.7237, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.7041646720919101, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.1682100382440686e-06, | |
| "loss": 1.7282, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.7060794638583054, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.1533854166604486e-06, | |
| "loss": 1.7385, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.7079942556247008, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.1386497006896058e-06, | |
| "loss": 1.6813, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.709909047391096, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.1240030384207202e-06, | |
| "loss": 1.7269, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.7118238391574916, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.1094455770480017e-06, | |
| "loss": 1.6812, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.7137386309238871, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.0949774628692278e-06, | |
| "loss": 1.6795, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.7156534226902824, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.0805988412842638e-06, | |
| "loss": 1.7112, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.7175682144566777, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.0663098567935981e-06, | |
| "loss": 1.7174, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.7194830062230733, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.0521106529969016e-06, | |
| "loss": 1.6405, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.7213977979894688, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.0380013725915783e-06, | |
| "loss": 1.7008, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.723312589755864, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.0239821573713228e-06, | |
| "loss": 1.6936, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.7252273815222594, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.0100531482247155e-06, | |
| "loss": 1.7914, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.7271421732886547, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 9.962144851337863e-07, | |
| "loss": 1.654, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 1.7290569650550502, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 9.824663071726204e-07, | |
| "loss": 1.7272, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.7309717568214458, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 9.688087525059575e-07, | |
| "loss": 1.7288, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.732886548587841, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 9.55241958387796e-07, | |
| "loss": 1.6932, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.7348013403542364, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 9.417660611600299e-07, | |
| "loss": 1.6952, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 1.736716132120632, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 9.283811962510603e-07, | |
| "loss": 1.7722, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 1.7386309238870274, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.150874981744507e-07, | |
| "loss": 1.6527, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.7405457156534228, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 9.018851005275586e-07, | |
| "loss": 1.7071, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 1.742460507419818, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 8.887741359902113e-07, | |
| "loss": 1.7559, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.7443752991862134, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 8.757547363233543e-07, | |
| "loss": 1.6998, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 1.746290090952609, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 8.628270323677424e-07, | |
| "loss": 1.593, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.7482048827190044, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 8.499911540426131e-07, | |
| "loss": 1.7532, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 1.7501196744853997, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 8.372472303443924e-07, | |
| "loss": 1.696, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 1.752034466251795, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 8.245953893453829e-07, | |
| "loss": 1.7278, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.7539492580181906, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 8.120357581924931e-07, | |
| "loss": 1.7215, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.7558640497845859, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 7.99568463105953e-07, | |
| "loss": 1.6702, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.7577788415509814, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 7.87193629378038e-07, | |
| "loss": 1.6721, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 1.7596936333173767, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 7.749113813718234e-07, | |
| "loss": 1.7008, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 1.761608425083772, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 7.627218425199278e-07, | |
| "loss": 1.6697, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.7635232168501676, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.506251353232663e-07, | |
| "loss": 1.7305, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 1.765438008616563, | |
| "grad_norm": 1.75, | |
| "learning_rate": 7.386213813498344e-07, | |
| "loss": 1.7425, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 1.7673528003829584, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 7.267107012334707e-07, | |
| "loss": 1.7188, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 1.7692675921493537, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 7.148932146726572e-07, | |
| "loss": 1.668, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.771182383915749, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 7.031690404293046e-07, | |
| "loss": 1.6772, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.7730971756821445, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 6.915382963275741e-07, | |
| "loss": 1.704, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 1.77501196744854, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 6.800010992526729e-07, | |
| "loss": 1.6764, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 1.7769267592149354, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 6.685575651497022e-07, | |
| "loss": 1.7017, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 1.7788415509813307, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 6.572078090224721e-07, | |
| "loss": 1.7396, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 1.7807563427477262, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 6.459519449323592e-07, | |
| "loss": 1.7057, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.7826711345141217, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 6.347900859971534e-07, | |
| "loss": 1.6723, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.784585926280517, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 6.237223443899221e-07, | |
| "loss": 1.7167, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 1.7865007180469124, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 6.127488313378894e-07, | |
| "loss": 1.7802, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 1.7884155098133077, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 6.018696571213045e-07, | |
| "loss": 1.6742, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 1.7903303015797032, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 5.910849310723499e-07, | |
| "loss": 1.7288, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.7922450933460987, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 5.803947615740291e-07, | |
| "loss": 1.6607, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 1.794159885112494, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 5.697992560590882e-07, | |
| "loss": 1.7431, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 1.7960746768788893, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 5.59298521008923e-07, | |
| "loss": 1.6916, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 1.7979894686452849, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 5.488926619525248e-07, | |
| "loss": 1.6678, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 1.7999042604116804, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 5.385817834654095e-07, | |
| "loss": 1.673, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.8018190521780757, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 5.283659891685656e-07, | |
| "loss": 1.7494, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 1.803733843944471, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 5.18245381727418e-07, | |
| "loss": 1.667, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 1.8056486357108663, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 5.082200628507994e-07, | |
| "loss": 1.7009, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 1.8075634274772618, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 4.982901332899159e-07, | |
| "loss": 1.681, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 1.8094782192436574, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 4.884556928373462e-07, | |
| "loss": 1.6666, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.8113930110100527, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 4.787168403260323e-07, | |
| "loss": 1.6944, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 1.813307802776448, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 4.690736736282908e-07, | |
| "loss": 1.6373, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 1.8152225945428435, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 4.595262896548236e-07, | |
| "loss": 1.709, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 1.8171373863092388, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 4.500747843537523e-07, | |
| "loss": 1.7181, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 1.8190521780756344, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 4.407192527096404e-07, | |
| "loss": 1.6296, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.8209669698420297, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 4.3145978874255757e-07, | |
| "loss": 1.6935, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 1.822881761608425, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 4.222964855071154e-07, | |
| "loss": 1.6726, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.8247965533748205, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 4.1322943509154887e-07, | |
| "loss": 1.6841, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 1.826711345141216, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 4.042587286167754e-07, | |
| "loss": 1.6808, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 1.8286261369076113, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 3.953844562354936e-07, | |
| "loss": 1.7918, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.8305409286740066, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 3.8660670713126735e-07, | |
| "loss": 1.7129, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 1.832455720440402, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.7792556951763424e-07, | |
| "loss": 1.6843, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 1.8343705122067975, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 3.6934113063721634e-07, | |
| "loss": 1.7275, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 1.836285303973193, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 3.6085347676084736e-07, | |
| "loss": 1.7158, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 1.8382000957395883, | |
| "grad_norm": 1.625, | |
| "learning_rate": 3.5246269318669924e-07, | |
| "loss": 1.6048, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.8401148875059836, | |
| "grad_norm": 1.875, | |
| "learning_rate": 3.441688642394292e-07, | |
| "loss": 1.6986, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 1.8420296792723792, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 3.359720732693361e-07, | |
| "loss": 1.6519, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 1.8439444710387747, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.2787240265151674e-07, | |
| "loss": 1.7094, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 1.84585926280517, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 3.1986993378503526e-07, | |
| "loss": 1.678, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 1.8477740545715653, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 3.11964747092115e-07, | |
| "loss": 1.6928, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.8496888463379606, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.041569220173235e-07, | |
| "loss": 1.7484, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.8516036381043561, | |
| "grad_norm": 1.75, | |
| "learning_rate": 2.9644653702677553e-07, | |
| "loss": 1.7354, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 1.8535184298707517, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.888336696073435e-07, | |
| "loss": 1.6876, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.855433221637147, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.8131839626588056e-07, | |
| "loss": 1.6774, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 1.8573480134035423, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.7390079252845205e-07, | |
| "loss": 1.6936, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.8592628051699378, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.6658093293957187e-07, | |
| "loss": 1.7162, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 1.8611775969363333, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.5935889106146305e-07, | |
| "loss": 1.7387, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 1.8630923887027286, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.522347394733049e-07, | |
| "loss": 1.6896, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 1.865007180469124, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.45208549770517e-07, | |
| "loss": 1.7292, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 1.8669219722355193, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.382803925640309e-07, | |
| "loss": 1.7433, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.8688367640019148, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 2.314503374795829e-07, | |
| "loss": 1.6975, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.8707515557683103, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.247184531570168e-07, | |
| "loss": 1.7252, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 1.8726663475347056, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.1808480724959004e-07, | |
| "loss": 1.7407, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 1.874581139301101, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.1154946642329644e-07, | |
| "loss": 1.7345, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 1.8764959310674965, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.051124963561979e-07, | |
| "loss": 1.7497, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.8784107228338918, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.9877396173775598e-07, | |
| "loss": 1.7163, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 1.8803255146002873, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9253392626819468e-07, | |
| "loss": 1.6973, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 1.8822403063666826, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.8639245265784866e-07, | |
| "loss": 1.6974, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 1.884155098133078, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.8034960262654276e-07, | |
| "loss": 1.7159, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.8860698898994734, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.744054369029591e-07, | |
| "loss": 1.6499, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.887984681665869, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.6856001522404296e-07, | |
| "loss": 1.7192, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 1.8898994734322643, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.6281339633438698e-07, | |
| "loss": 1.7233, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.8918142651986596, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.5716563798565232e-07, | |
| "loss": 1.7212, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.893729056965055, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.5161679693598274e-07, | |
| "loss": 1.6976, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 1.8956438487314504, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.4616692894943274e-07, | |
| "loss": 1.7309, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.897558640497846, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.4081608879541241e-07, | |
| "loss": 1.7336, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 1.8994734322642413, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.3556433024813353e-07, | |
| "loss": 1.7123, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 1.9013882240306366, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.304117060860688e-07, | |
| "loss": 1.6961, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 1.903303015797032, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.2535826809142339e-07, | |
| "loss": 1.6591, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 1.9052178075634276, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.2040406704961316e-07, | |
| "loss": 1.6986, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.907132599329823, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.15549152748754e-07, | |
| "loss": 1.6602, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 1.9090473910962182, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.1079357397916435e-07, | |
| "loss": 1.6829, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 1.9109621828626135, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.061373785328701e-07, | |
| "loss": 1.688, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 1.912876974629009, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.015806132031305e-07, | |
| "loss": 1.7246, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 1.9147917663954046, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 9.712332378395861e-08, | |
| "loss": 1.7222, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9167065581618, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 9.276555506967378e-08, | |
| "loss": 1.7414, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 1.9186213499281952, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 8.850735085443763e-08, | |
| "loss": 1.6952, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 1.9205361416945907, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 8.434875393182662e-08, | |
| "loss": 1.671, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 1.9224509334609863, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 8.028980609439241e-08, | |
| "loss": 1.6717, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 1.9243657252273816, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 7.633054813324769e-08, | |
| "loss": 1.6772, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.926280516993777, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 7.247101983765104e-08, | |
| "loss": 1.7314, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 1.9281953087601722, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 6.871125999461604e-08, | |
| "loss": 1.5927, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 1.9301101005265677, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 6.505130638850831e-08, | |
| "loss": 1.6729, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 1.9320248922929633, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 6.14911958006792e-08, | |
| "loss": 1.7042, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 1.9339396840593586, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 5.803096400908703e-08, | |
| "loss": 1.8112, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.9358544758257539, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 5.46706457879409e-08, | |
| "loss": 1.7101, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 1.9377692675921494, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 5.141027490735195e-08, | |
| "loss": 1.782, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 1.9396840593585447, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 4.824988413299037e-08, | |
| "loss": 1.6928, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 1.9415988511249402, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 4.5189505225762266e-08, | |
| "loss": 1.7131, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 1.9435136428913355, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 4.2229168941484434e-08, | |
| "loss": 1.7265, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.9454284346577309, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 3.9368905030578994e-08, | |
| "loss": 1.6424, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 1.9473432264241264, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 3.6608742237769227e-08, | |
| "loss": 1.7135, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 1.949258018190522, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 3.394870830180197e-08, | |
| "loss": 1.7292, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 1.9511728099569172, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 3.1388829955153466e-08, | |
| "loss": 1.6791, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 1.9530876017233125, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.892913292377508e-08, | |
| "loss": 1.6906, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.9550023934897078, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.656964192682354e-08, | |
| "loss": 1.698, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 1.9569171852561034, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 2.431038067642111e-08, | |
| "loss": 1.6888, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 1.958831977022499, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.2151371877412452e-08, | |
| "loss": 1.6964, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 1.9607467687888942, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.0092637227134836e-08, | |
| "loss": 1.6933, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 1.9626615605552895, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.8134197415207165e-08, | |
| "loss": 1.689, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.964576352321685, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.627607212331572e-08, | |
| "loss": 1.6823, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 1.9664911440880806, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.451828002501654e-08, | |
| "loss": 1.7396, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 1.9684059358544759, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.286083878555111e-08, | |
| "loss": 1.6459, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 1.9703207276208712, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.1303765061668748e-08, | |
| "loss": 1.6741, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 1.9722355193872665, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 9.847074501456722e-09, | |
| "loss": 1.6616, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.974150311153662, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 8.490781744181498e-09, | |
| "loss": 1.7448, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 1.9760651029200575, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 7.234900420147739e-09, | |
| "loss": 1.6673, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 1.9779798946864529, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 6.079443150556197e-09, | |
| "loss": 1.7032, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 1.9798946864528482, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 5.0244215473782556e-09, | |
| "loss": 1.6774, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 1.9818094782192437, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 4.069846213238249e-09, | |
| "loss": 1.7234, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.9837242699856392, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.2157267413113203e-09, | |
| "loss": 1.6599, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 1.9856390617520345, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.4620717152201713e-09, | |
| "loss": 1.7449, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 1.9875538535184298, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.8088887089551255e-09, | |
| "loss": 1.6684, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 1.9894686452848251, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.256184286793083e-09, | |
| "loss": 1.641, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 1.9913834370512207, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 8.039640032342366e-10, | |
| "loss": 1.7045, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.9932982288176162, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 4.522324029465619e-10, | |
| "loss": 1.7222, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 1.9952130205840115, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2.0099302071807658e-10, | |
| "loss": 1.6946, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 1.9971278123504068, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 5.024838142464461e-11, | |
| "loss": 1.7026, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 1.9990426041168023, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 0.0, | |
| "loss": 1.7507, | |
| "step": 1044 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1044, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.28007906539733e+17, | |
| "train_batch_size": 10, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |