| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.964200477326969, | |
| "eval_steps": 52, | |
| "global_step": 520, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00954653937947494, | |
| "grad_norm": 0.541497528553009, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.8864, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00954653937947494, | |
| "eval_loss": 0.975593626499176, | |
| "eval_runtime": 12.7833, | |
| "eval_samples_per_second": 13.846, | |
| "eval_steps_per_second": 1.799, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01909307875894988, | |
| "grad_norm": 0.7477102875709534, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.9382, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.028639618138424822, | |
| "grad_norm": 0.8567990064620972, | |
| "learning_rate": 6e-06, | |
| "loss": 1.0294, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03818615751789976, | |
| "grad_norm": 0.737090528011322, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.9398, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0477326968973747, | |
| "grad_norm": 0.6384573578834534, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9345, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.057279236276849645, | |
| "grad_norm": 0.7606269121170044, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.941, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06682577565632458, | |
| "grad_norm": 0.7251742482185364, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 0.958, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07637231503579953, | |
| "grad_norm": 0.6773186922073364, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.9318, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08591885441527446, | |
| "grad_norm": 0.8379804491996765, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.9919, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0954653937947494, | |
| "grad_norm": 0.9668668508529663, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1819, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10501193317422435, | |
| "grad_norm": 0.5556192398071289, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 0.9474, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11455847255369929, | |
| "grad_norm": 0.6156389117240906, | |
| "learning_rate": 2.4e-05, | |
| "loss": 0.9212, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12410501193317422, | |
| "grad_norm": 0.6029731035232544, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 0.8982, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.13365155131264916, | |
| "grad_norm": 0.6002732515335083, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 0.9832, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.1431980906921241, | |
| "grad_norm": 0.6266693472862244, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9485, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15274463007159905, | |
| "grad_norm": 0.5493901371955872, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.9013, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.162291169451074, | |
| "grad_norm": 0.5467607378959656, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 0.8975, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1718377088305489, | |
| "grad_norm": 0.556983232498169, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.1175, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.18138424821002386, | |
| "grad_norm": 0.4399558901786804, | |
| "learning_rate": 3.8e-05, | |
| "loss": 0.8975, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.1909307875894988, | |
| "grad_norm": 0.4225325882434845, | |
| "learning_rate": 4e-05, | |
| "loss": 0.8186, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.20047732696897375, | |
| "grad_norm": 0.38200807571411133, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.82, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2100238663484487, | |
| "grad_norm": 0.3150412440299988, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.844, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.21957040572792363, | |
| "grad_norm": 0.3351333737373352, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.9004, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.22911694510739858, | |
| "grad_norm": 0.3316827416419983, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.8965, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2386634844868735, | |
| "grad_norm": 0.31210237741470337, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9084, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24821002386634844, | |
| "grad_norm": 0.33221927285194397, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 0.9448, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2577565632458234, | |
| "grad_norm": 0.35407039523124695, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 0.82, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.26730310262529833, | |
| "grad_norm": 0.34163302183151245, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.8843, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.27684964200477324, | |
| "grad_norm": 0.3265129029750824, | |
| "learning_rate": 5.8e-05, | |
| "loss": 0.8376, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2863961813842482, | |
| "grad_norm": 0.38681215047836304, | |
| "learning_rate": 6e-05, | |
| "loss": 0.9903, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.29594272076372313, | |
| "grad_norm": 0.3341940641403198, | |
| "learning_rate": 6.2e-05, | |
| "loss": 0.8054, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3054892601431981, | |
| "grad_norm": 0.3521149158477783, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.8931, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.315035799522673, | |
| "grad_norm": 0.3392002284526825, | |
| "learning_rate": 6.6e-05, | |
| "loss": 0.7776, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.324582338902148, | |
| "grad_norm": 0.3627275824546814, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 0.7253, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3341288782816229, | |
| "grad_norm": 0.33696770668029785, | |
| "learning_rate": 7e-05, | |
| "loss": 0.9011, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3436754176610978, | |
| "grad_norm": 0.3550478518009186, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.8064, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3532219570405728, | |
| "grad_norm": 0.3183474838733673, | |
| "learning_rate": 7.4e-05, | |
| "loss": 0.8094, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3627684964200477, | |
| "grad_norm": 0.33637434244155884, | |
| "learning_rate": 7.6e-05, | |
| "loss": 0.8637, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3723150357995227, | |
| "grad_norm": 0.34323257207870483, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 0.8717, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.3818615751789976, | |
| "grad_norm": 0.290461003780365, | |
| "learning_rate": 8e-05, | |
| "loss": 0.8501, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3914081145584726, | |
| "grad_norm": 0.35435885190963745, | |
| "learning_rate": 8.2e-05, | |
| "loss": 0.8446, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4009546539379475, | |
| "grad_norm": 0.30319270491600037, | |
| "learning_rate": 8.4e-05, | |
| "loss": 0.8175, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4105011933174224, | |
| "grad_norm": 0.28563714027404785, | |
| "learning_rate": 8.6e-05, | |
| "loss": 0.7933, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4200477326968974, | |
| "grad_norm": 0.26857540011405945, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.7177, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4295942720763723, | |
| "grad_norm": 0.27898675203323364, | |
| "learning_rate": 9e-05, | |
| "loss": 0.7506, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.43914081145584727, | |
| "grad_norm": 0.2959842383861542, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 0.859, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4486873508353222, | |
| "grad_norm": 0.2996789515018463, | |
| "learning_rate": 9.4e-05, | |
| "loss": 0.7429, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.45823389021479716, | |
| "grad_norm": 0.2459433376789093, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.6911, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4677804295942721, | |
| "grad_norm": 0.318551629781723, | |
| "learning_rate": 9.8e-05, | |
| "loss": 0.8618, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.477326968973747, | |
| "grad_norm": 0.30586713552474976, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8546, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48687350835322196, | |
| "grad_norm": 0.33441683650016785, | |
| "learning_rate": 9.999888302765345e-05, | |
| "loss": 0.8177, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4964200477326969, | |
| "grad_norm": 0.3031998872756958, | |
| "learning_rate": 9.99955321605189e-05, | |
| "loss": 0.8585, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.4964200477326969, | |
| "eval_loss": 0.8117080926895142, | |
| "eval_runtime": 12.9734, | |
| "eval_samples_per_second": 13.643, | |
| "eval_steps_per_second": 1.773, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5059665871121718, | |
| "grad_norm": 0.2897348403930664, | |
| "learning_rate": 9.99899475483094e-05, | |
| "loss": 0.731, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5155131264916468, | |
| "grad_norm": 0.29636526107788086, | |
| "learning_rate": 9.99821294405392e-05, | |
| "loss": 0.8064, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5250596658711217, | |
| "grad_norm": 0.3027283847332001, | |
| "learning_rate": 9.997207818651274e-05, | |
| "loss": 0.7737, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5346062052505967, | |
| "grad_norm": 0.2626173794269562, | |
| "learning_rate": 9.995979423530892e-05, | |
| "loss": 0.793, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5441527446300716, | |
| "grad_norm": 0.2591354548931122, | |
| "learning_rate": 9.99452781357611e-05, | |
| "loss": 0.7764, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5536992840095465, | |
| "grad_norm": 0.3225250244140625, | |
| "learning_rate": 9.992853053643257e-05, | |
| "loss": 0.8997, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5632458233890215, | |
| "grad_norm": 0.29806965589523315, | |
| "learning_rate": 9.99095521855875e-05, | |
| "loss": 0.73, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5727923627684964, | |
| "grad_norm": 0.28511664271354675, | |
| "learning_rate": 9.988834393115767e-05, | |
| "loss": 0.7074, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5823389021479713, | |
| "grad_norm": 0.2897747755050659, | |
| "learning_rate": 9.986490672070437e-05, | |
| "loss": 0.8004, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5918854415274463, | |
| "grad_norm": 0.31389617919921875, | |
| "learning_rate": 9.983924160137625e-05, | |
| "loss": 0.7397, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6014319809069213, | |
| "grad_norm": 0.2554086744785309, | |
| "learning_rate": 9.98113497198625e-05, | |
| "loss": 0.8052, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6109785202863962, | |
| "grad_norm": 0.2977796494960785, | |
| "learning_rate": 9.978123232234147e-05, | |
| "loss": 0.7742, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6205250596658711, | |
| "grad_norm": 0.2979027330875397, | |
| "learning_rate": 9.974889075442521e-05, | |
| "loss": 0.7169, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.630071599045346, | |
| "grad_norm": 0.29229047894477844, | |
| "learning_rate": 9.971432646109919e-05, | |
| "loss": 0.8076, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.639618138424821, | |
| "grad_norm": 0.29552674293518066, | |
| "learning_rate": 9.967754098665778e-05, | |
| "loss": 0.7533, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.649164677804296, | |
| "grad_norm": 0.2626473605632782, | |
| "learning_rate": 9.963853597463533e-05, | |
| "loss": 0.7669, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6587112171837709, | |
| "grad_norm": 0.28999584913253784, | |
| "learning_rate": 9.959731316773259e-05, | |
| "loss": 0.8175, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.6682577565632458, | |
| "grad_norm": 0.3006598949432373, | |
| "learning_rate": 9.9553874407739e-05, | |
| "loss": 0.822, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6778042959427207, | |
| "grad_norm": 0.3051898181438446, | |
| "learning_rate": 9.950822163545032e-05, | |
| "loss": 0.8513, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6873508353221957, | |
| "grad_norm": 0.3299737572669983, | |
| "learning_rate": 9.946035689058188e-05, | |
| "loss": 0.8559, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6968973747016707, | |
| "grad_norm": 0.2913152575492859, | |
| "learning_rate": 9.941028231167756e-05, | |
| "loss": 0.7742, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7064439140811456, | |
| "grad_norm": 0.2927692234516144, | |
| "learning_rate": 9.935800013601414e-05, | |
| "loss": 0.8335, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7159904534606205, | |
| "grad_norm": 0.28141114115715027, | |
| "learning_rate": 9.930351269950143e-05, | |
| "loss": 0.7976, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7255369928400954, | |
| "grad_norm": 0.2617853581905365, | |
| "learning_rate": 9.924682243657779e-05, | |
| "loss": 0.7707, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7350835322195705, | |
| "grad_norm": 0.3233207166194916, | |
| "learning_rate": 9.918793188010147e-05, | |
| "loss": 0.8698, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7446300715990454, | |
| "grad_norm": 0.29364439845085144, | |
| "learning_rate": 9.91268436612374e-05, | |
| "loss": 0.79, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7541766109785203, | |
| "grad_norm": 0.3051166832447052, | |
| "learning_rate": 9.906356050933961e-05, | |
| "loss": 0.864, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7637231503579952, | |
| "grad_norm": 0.2952063977718353, | |
| "learning_rate": 9.899808525182935e-05, | |
| "loss": 0.7312, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7732696897374701, | |
| "grad_norm": 0.2865234911441803, | |
| "learning_rate": 9.893042081406867e-05, | |
| "loss": 0.7051, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7828162291169452, | |
| "grad_norm": 0.3515304625034332, | |
| "learning_rate": 9.886057021922982e-05, | |
| "loss": 0.917, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7923627684964201, | |
| "grad_norm": 0.2626941204071045, | |
| "learning_rate": 9.878853658816014e-05, | |
| "loss": 0.7131, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.801909307875895, | |
| "grad_norm": 0.276617169380188, | |
| "learning_rate": 9.871432313924255e-05, | |
| "loss": 0.7683, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8114558472553699, | |
| "grad_norm": 0.2965279519557953, | |
| "learning_rate": 9.863793318825186e-05, | |
| "loss": 0.686, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8210023866348448, | |
| "grad_norm": 0.3024349510669708, | |
| "learning_rate": 9.85593701482066e-05, | |
| "loss": 0.8203, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8305489260143198, | |
| "grad_norm": 0.31996044516563416, | |
| "learning_rate": 9.847863752921649e-05, | |
| "loss": 0.8381, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8400954653937948, | |
| "grad_norm": 0.2804871201515198, | |
| "learning_rate": 9.839573893832563e-05, | |
| "loss": 0.8378, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8496420047732697, | |
| "grad_norm": 0.3115004301071167, | |
| "learning_rate": 9.831067807935137e-05, | |
| "loss": 0.7382, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8591885441527446, | |
| "grad_norm": 0.3251977264881134, | |
| "learning_rate": 9.822345875271883e-05, | |
| "loss": 0.8506, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8687350835322196, | |
| "grad_norm": 0.32227852940559387, | |
| "learning_rate": 9.813408485529103e-05, | |
| "loss": 0.7768, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8782816229116945, | |
| "grad_norm": 0.33089524507522583, | |
| "learning_rate": 9.804256038019481e-05, | |
| "loss": 0.7763, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8878281622911695, | |
| "grad_norm": 0.3890259563922882, | |
| "learning_rate": 9.794888941664253e-05, | |
| "loss": 0.9264, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8973747016706444, | |
| "grad_norm": 0.31058862805366516, | |
| "learning_rate": 9.785307614974921e-05, | |
| "loss": 0.8139, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9069212410501193, | |
| "grad_norm": 0.2780233323574066, | |
| "learning_rate": 9.775512486034563e-05, | |
| "loss": 0.8274, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9164677804295943, | |
| "grad_norm": 0.3168707489967346, | |
| "learning_rate": 9.765503992478704e-05, | |
| "loss": 0.8425, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9260143198090692, | |
| "grad_norm": 0.3951367139816284, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.6909, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9355608591885441, | |
| "grad_norm": 0.3271735608577728, | |
| "learning_rate": 9.744848709707091e-05, | |
| "loss": 0.7677, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9451073985680191, | |
| "grad_norm": 0.26784732937812805, | |
| "learning_rate": 9.734202843346522e-05, | |
| "loss": 0.6579, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.954653937947494, | |
| "grad_norm": 0.3045744001865387, | |
| "learning_rate": 9.723345458039594e-05, | |
| "loss": 0.738, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.964200477326969, | |
| "grad_norm": 0.30207037925720215, | |
| "learning_rate": 9.712277038882273e-05, | |
| "loss": 0.7435, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9737470167064439, | |
| "grad_norm": 0.26012739539146423, | |
| "learning_rate": 9.700998080399287e-05, | |
| "loss": 0.6577, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9832935560859188, | |
| "grad_norm": 0.3377532660961151, | |
| "learning_rate": 9.689509086522019e-05, | |
| "loss": 0.8357, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9928400954653938, | |
| "grad_norm": 0.2794972360134125, | |
| "learning_rate": 9.67781057056601e-05, | |
| "loss": 0.7169, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.9928400954653938, | |
| "eval_loss": 0.7859669327735901, | |
| "eval_runtime": 13.0799, | |
| "eval_samples_per_second": 13.532, | |
| "eval_steps_per_second": 1.758, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0023866348448687, | |
| "grad_norm": 0.34849369525909424, | |
| "learning_rate": 9.665903055208014e-05, | |
| "loss": 0.7555, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0119331742243436, | |
| "grad_norm": 0.32931214570999146, | |
| "learning_rate": 9.653787072462643e-05, | |
| "loss": 0.8744, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0214797136038185, | |
| "grad_norm": 0.2845034897327423, | |
| "learning_rate": 9.641463163658605e-05, | |
| "loss": 0.7014, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0310262529832936, | |
| "grad_norm": 0.3317318260669708, | |
| "learning_rate": 9.628931879414517e-05, | |
| "loss": 0.6797, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0405727923627686, | |
| "grad_norm": 0.3143399655818939, | |
| "learning_rate": 9.616193779614294e-05, | |
| "loss": 0.7855, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0501193317422435, | |
| "grad_norm": 0.3227895498275757, | |
| "learning_rate": 9.603249433382144e-05, | |
| "loss": 0.8448, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0596658711217184, | |
| "grad_norm": 0.29557734727859497, | |
| "learning_rate": 9.590099419057141e-05, | |
| "loss": 0.7718, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0692124105011933, | |
| "grad_norm": 0.2954796254634857, | |
| "learning_rate": 9.57674432416738e-05, | |
| "loss": 0.7213, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0787589498806682, | |
| "grad_norm": 0.33134913444519043, | |
| "learning_rate": 9.563184745403724e-05, | |
| "loss": 0.7673, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0883054892601431, | |
| "grad_norm": 0.2836027145385742, | |
| "learning_rate": 9.549421288593157e-05, | |
| "loss": 0.6618, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.097852028639618, | |
| "grad_norm": 0.27743393182754517, | |
| "learning_rate": 9.535454568671704e-05, | |
| "loss": 0.5814, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.107398568019093, | |
| "grad_norm": 0.30395206809043884, | |
| "learning_rate": 9.521285209656962e-05, | |
| "loss": 0.7574, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1169451073985681, | |
| "grad_norm": 0.3082098364830017, | |
| "learning_rate": 9.506913844620218e-05, | |
| "loss": 0.6795, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.126491646778043, | |
| "grad_norm": 0.34204941987991333, | |
| "learning_rate": 9.492341115658167e-05, | |
| "loss": 0.7607, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.136038186157518, | |
| "grad_norm": 0.2831597924232483, | |
| "learning_rate": 9.477567673864216e-05, | |
| "loss": 0.7202, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1455847255369929, | |
| "grad_norm": 0.31007006764411926, | |
| "learning_rate": 9.462594179299406e-05, | |
| "loss": 0.7508, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1551312649164678, | |
| "grad_norm": 0.36465707421302795, | |
| "learning_rate": 9.44742130096291e-05, | |
| "loss": 0.7767, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1646778042959427, | |
| "grad_norm": 0.30296140909194946, | |
| "learning_rate": 9.43204971676215e-05, | |
| "loss": 0.7016, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1742243436754176, | |
| "grad_norm": 0.30213695764541626, | |
| "learning_rate": 9.416480113482504e-05, | |
| "loss": 0.7981, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1837708830548925, | |
| "grad_norm": 0.3844444155693054, | |
| "learning_rate": 9.400713186756625e-05, | |
| "loss": 0.8803, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1933174224343674, | |
| "grad_norm": 0.35131993889808655, | |
| "learning_rate": 9.384749641033359e-05, | |
| "loss": 0.7987, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2028639618138426, | |
| "grad_norm": 0.32640498876571655, | |
| "learning_rate": 9.368590189546269e-05, | |
| "loss": 0.7141, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2124105011933175, | |
| "grad_norm": 0.33494001626968384, | |
| "learning_rate": 9.352235554281774e-05, | |
| "loss": 0.7661, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2219570405727924, | |
| "grad_norm": 0.3069153130054474, | |
| "learning_rate": 9.335686465946887e-05, | |
| "loss": 0.7727, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2315035799522673, | |
| "grad_norm": 0.34017837047576904, | |
| "learning_rate": 9.31894366393657e-05, | |
| "loss": 0.712, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2410501193317423, | |
| "grad_norm": 0.3549365997314453, | |
| "learning_rate": 9.302007896300698e-05, | |
| "loss": 0.7778, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2505966587112172, | |
| "grad_norm": 0.36418265104293823, | |
| "learning_rate": 9.284879919710632e-05, | |
| "loss": 0.7858, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.260143198090692, | |
| "grad_norm": 0.35619720816612244, | |
| "learning_rate": 9.267560499425424e-05, | |
| "loss": 0.7254, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.269689737470167, | |
| "grad_norm": 0.3609873652458191, | |
| "learning_rate": 9.250050409257611e-05, | |
| "loss": 0.702, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.279236276849642, | |
| "grad_norm": 0.33549803495407104, | |
| "learning_rate": 9.232350431538656e-05, | |
| "loss": 0.753, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.288782816229117, | |
| "grad_norm": 0.3622001111507416, | |
| "learning_rate": 9.214461357083985e-05, | |
| "loss": 0.7751, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2983293556085918, | |
| "grad_norm": 0.36122390627861023, | |
| "learning_rate": 9.196383985157656e-05, | |
| "loss": 0.7293, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.307875894988067, | |
| "grad_norm": 0.3313043713569641, | |
| "learning_rate": 9.17811912343665e-05, | |
| "loss": 0.7719, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3174224343675418, | |
| "grad_norm": 0.420242041349411, | |
| "learning_rate": 9.159667587974785e-05, | |
| "loss": 0.8639, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3269689737470167, | |
| "grad_norm": 0.3835316598415375, | |
| "learning_rate": 9.141030203166256e-05, | |
| "loss": 0.7759, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3365155131264916, | |
| "grad_norm": 0.3210572898387909, | |
| "learning_rate": 9.122207801708802e-05, | |
| "loss": 0.7146, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3460620525059666, | |
| "grad_norm": 0.3777942955493927, | |
| "learning_rate": 9.103201224566498e-05, | |
| "loss": 0.6785, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3556085918854415, | |
| "grad_norm": 0.36900511384010315, | |
| "learning_rate": 9.084011320932189e-05, | |
| "loss": 0.7334, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3651551312649164, | |
| "grad_norm": 0.34692367911338806, | |
| "learning_rate": 9.064638948189538e-05, | |
| "loss": 0.8239, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3747016706443915, | |
| "grad_norm": 0.3698657155036926, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.816, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.3842482100238662, | |
| "grad_norm": 0.3782055974006653, | |
| "learning_rate": 9.025350265637815e-05, | |
| "loss": 0.7642, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3937947494033414, | |
| "grad_norm": 0.32594063878059387, | |
| "learning_rate": 9.005435711203618e-05, | |
| "loss": 0.6679, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4033412887828163, | |
| "grad_norm": 0.31659746170043945, | |
| "learning_rate": 8.985342198332407e-05, | |
| "loss": 0.6733, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4128878281622912, | |
| "grad_norm": 0.3700142502784729, | |
| "learning_rate": 8.965070624780116e-05, | |
| "loss": 0.7628, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.422434367541766, | |
| "grad_norm": 0.36097973585128784, | |
| "learning_rate": 8.944621896258225e-05, | |
| "loss": 0.823, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.431980906921241, | |
| "grad_norm": 0.34358176589012146, | |
| "learning_rate": 8.923996926393305e-05, | |
| "loss": 0.8027, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.441527446300716, | |
| "grad_norm": 0.3484276831150055, | |
| "learning_rate": 8.903196636686197e-05, | |
| "loss": 0.7654, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4510739856801909, | |
| "grad_norm": 0.3297461271286011, | |
| "learning_rate": 8.882221956470836e-05, | |
| "loss": 0.7309, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.460620525059666, | |
| "grad_norm": 0.337780237197876, | |
| "learning_rate": 8.861073822872734e-05, | |
| "loss": 0.6041, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4701670644391407, | |
| "grad_norm": 0.4128996431827545, | |
| "learning_rate": 8.839753180767108e-05, | |
| "loss": 0.755, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4797136038186158, | |
| "grad_norm": 0.39525777101516724, | |
| "learning_rate": 8.818260982736661e-05, | |
| "loss": 0.893, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4892601431980907, | |
| "grad_norm": 0.364863783121109, | |
| "learning_rate": 8.79659818902903e-05, | |
| "loss": 0.7558, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.4892601431980907, | |
| "eval_loss": 0.7723400592803955, | |
| "eval_runtime": 12.9849, | |
| "eval_samples_per_second": 13.631, | |
| "eval_steps_per_second": 1.771, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.4988066825775657, | |
| "grad_norm": 0.3498753309249878, | |
| "learning_rate": 8.774765767513875e-05, | |
| "loss": 0.6726, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5083532219570406, | |
| "grad_norm": 0.3801003694534302, | |
| "learning_rate": 8.752764693639638e-05, | |
| "loss": 0.8054, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5178997613365155, | |
| "grad_norm": 0.33022618293762207, | |
| "learning_rate": 8.730595950389968e-05, | |
| "loss": 0.6711, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.5274463007159904, | |
| "grad_norm": 0.3503524363040924, | |
| "learning_rate": 8.708260528239788e-05, | |
| "loss": 0.695, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5369928400954653, | |
| "grad_norm": 0.3124283254146576, | |
| "learning_rate": 8.685759425111056e-05, | |
| "loss": 0.6797, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5465393794749405, | |
| "grad_norm": 0.3534424602985382, | |
| "learning_rate": 8.663093646328166e-05, | |
| "loss": 0.7181, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5560859188544152, | |
| "grad_norm": 0.40379753708839417, | |
| "learning_rate": 8.640264204573047e-05, | |
| "loss": 0.7923, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5656324582338903, | |
| "grad_norm": 0.4083033502101898, | |
| "learning_rate": 8.617272119839903e-05, | |
| "loss": 0.7997, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.575178997613365, | |
| "grad_norm": 0.3316013216972351, | |
| "learning_rate": 8.594118419389647e-05, | |
| "loss": 0.7135, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5847255369928401, | |
| "grad_norm": 0.3352660834789276, | |
| "learning_rate": 8.570804137704003e-05, | |
| "loss": 0.6358, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.594272076372315, | |
| "grad_norm": 0.34296879172325134, | |
| "learning_rate": 8.547330316439291e-05, | |
| "loss": 0.683, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.60381861575179, | |
| "grad_norm": 0.362801730632782, | |
| "learning_rate": 8.523698004379877e-05, | |
| "loss": 0.7637, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6133651551312649, | |
| "grad_norm": 0.3877812623977661, | |
| "learning_rate": 8.499908257391323e-05, | |
| "loss": 0.6848, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6229116945107398, | |
| "grad_norm": 0.3417890965938568, | |
| "learning_rate": 8.475962138373213e-05, | |
| "loss": 0.6743, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.632458233890215, | |
| "grad_norm": 0.3690805435180664, | |
| "learning_rate": 8.451860717211653e-05, | |
| "loss": 0.717, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6420047732696896, | |
| "grad_norm": 0.39610129594802856, | |
| "learning_rate": 8.427605070731482e-05, | |
| "loss": 0.831, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6515513126491648, | |
| "grad_norm": 0.3383614718914032, | |
| "learning_rate": 8.403196282648156e-05, | |
| "loss": 0.6713, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6610978520286395, | |
| "grad_norm": 0.4015936553478241, | |
| "learning_rate": 8.378635443519327e-05, | |
| "loss": 0.8089, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6706443914081146, | |
| "grad_norm": 0.34833744168281555, | |
| "learning_rate": 8.353923650696118e-05, | |
| "loss": 0.6678, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6801909307875895, | |
| "grad_norm": 0.35956260561943054, | |
| "learning_rate": 8.329062008274098e-05, | |
| "loss": 0.751, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6897374701670644, | |
| "grad_norm": 0.35701537132263184, | |
| "learning_rate": 8.304051627043953e-05, | |
| "loss": 0.6618, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.6992840095465394, | |
| "grad_norm": 0.2929876446723938, | |
| "learning_rate": 8.278893624441848e-05, | |
| "loss": 0.647, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7088305489260143, | |
| "grad_norm": 0.3412924110889435, | |
| "learning_rate": 8.253589124499512e-05, | |
| "loss": 0.6297, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.7183770883054894, | |
| "grad_norm": 0.3739149272441864, | |
| "learning_rate": 8.228139257794012e-05, | |
| "loss": 0.7222, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.727923627684964, | |
| "grad_norm": 0.327097624540329, | |
| "learning_rate": 8.202545161397242e-05, | |
| "loss": 0.7254, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7374701670644392, | |
| "grad_norm": 0.4074615240097046, | |
| "learning_rate": 8.176807978825118e-05, | |
| "loss": 0.8142, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.747016706443914, | |
| "grad_norm": 0.36452674865722656, | |
| "learning_rate": 8.150928859986488e-05, | |
| "loss": 0.732, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.756563245823389, | |
| "grad_norm": 0.41651418805122375, | |
| "learning_rate": 8.124908961131759e-05, | |
| "loss": 0.8458, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.766109785202864, | |
| "grad_norm": 0.3859712779521942, | |
| "learning_rate": 8.098749444801224e-05, | |
| "loss": 0.852, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.775656324582339, | |
| "grad_norm": 0.3324730694293976, | |
| "learning_rate": 8.072451479773143e-05, | |
| "loss": 0.6524, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7852028639618138, | |
| "grad_norm": 0.35326018929481506, | |
| "learning_rate": 8.0460162410115e-05, | |
| "loss": 0.6616, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.7947494033412887, | |
| "grad_norm": 0.3914497494697571, | |
| "learning_rate": 8.019444909613522e-05, | |
| "loss": 0.8023, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8042959427207639, | |
| "grad_norm": 0.35949957370758057, | |
| "learning_rate": 7.992738672756909e-05, | |
| "loss": 0.6108, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8138424821002386, | |
| "grad_norm": 0.366787314414978, | |
| "learning_rate": 7.965898723646776e-05, | |
| "loss": 0.8388, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8233890214797137, | |
| "grad_norm": 0.3832429349422455, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.7082, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8329355608591884, | |
| "grad_norm": 0.3898661732673645, | |
| "learning_rate": 7.911822491303452e-05, | |
| "loss": 0.7633, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8424821002386635, | |
| "grad_norm": 0.3598467707633972, | |
| "learning_rate": 7.884588624136504e-05, | |
| "loss": 0.6604, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8520286396181385, | |
| "grad_norm": 0.33222588896751404, | |
| "learning_rate": 7.857225876740584e-05, | |
| "loss": 0.7332, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8615751789976134, | |
| "grad_norm": 0.410185307264328, | |
| "learning_rate": 7.829735471652978e-05, | |
| "loss": 0.8056, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8711217183770883, | |
| "grad_norm": 0.41406041383743286, | |
| "learning_rate": 7.802118637114573e-05, | |
| "loss": 0.8366, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8806682577565632, | |
| "grad_norm": 0.4121192693710327, | |
| "learning_rate": 7.774376607014995e-05, | |
| "loss": 0.8089, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8902147971360383, | |
| "grad_norm": 0.3749670684337616, | |
| "learning_rate": 7.746510620837459e-05, | |
| "loss": 0.7294, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.899761336515513, | |
| "grad_norm": 0.42995452880859375, | |
| "learning_rate": 7.718521923603404e-05, | |
| "loss": 0.7776, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.9093078758949882, | |
| "grad_norm": 0.36848926544189453, | |
| "learning_rate": 7.690411765816864e-05, | |
| "loss": 0.7549, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9188544152744629, | |
| "grad_norm": 0.40055760741233826, | |
| "learning_rate": 7.662181403408593e-05, | |
| "loss": 0.6901, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.928400954653938, | |
| "grad_norm": 0.38324177265167236, | |
| "learning_rate": 7.633832097679958e-05, | |
| "loss": 0.8203, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.937947494033413, | |
| "grad_norm": 0.31426626443862915, | |
| "learning_rate": 7.605365115246581e-05, | |
| "loss": 0.7143, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9474940334128878, | |
| "grad_norm": 0.3708311915397644, | |
| "learning_rate": 7.576781727981749e-05, | |
| "loss": 0.7329, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.9570405727923628, | |
| "grad_norm": 0.3401066064834595, | |
| "learning_rate": 7.548083212959588e-05, | |
| "loss": 0.6827, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9665871121718377, | |
| "grad_norm": 0.32935282588005066, | |
| "learning_rate": 7.519270852398001e-05, | |
| "loss": 0.6928, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9761336515513126, | |
| "grad_norm": 0.30615073442459106, | |
| "learning_rate": 7.490345933601395e-05, | |
| "loss": 0.6326, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9856801909307875, | |
| "grad_norm": 0.37611639499664307, | |
| "learning_rate": 7.461309748903139e-05, | |
| "loss": 0.7814, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.9856801909307875, | |
| "eval_loss": 0.7655227780342102, | |
| "eval_runtime": 12.9629, | |
| "eval_samples_per_second": 13.654, | |
| "eval_steps_per_second": 1.774, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.9952267303102627, | |
| "grad_norm": 0.4345564842224121, | |
| "learning_rate": 7.432163595607851e-05, | |
| "loss": 0.8073, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.0047732696897373, | |
| "grad_norm": 0.3374365270137787, | |
| "learning_rate": 7.402908775933419e-05, | |
| "loss": 0.6948, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0143198090692125, | |
| "grad_norm": 0.34650278091430664, | |
| "learning_rate": 7.373546596952829e-05, | |
| "loss": 0.6353, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.023866348448687, | |
| "grad_norm": 0.40124908089637756, | |
| "learning_rate": 7.344078370535757e-05, | |
| "loss": 0.7312, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0334128878281623, | |
| "grad_norm": 0.3434641361236572, | |
| "learning_rate": 7.314505413289964e-05, | |
| "loss": 0.6418, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.042959427207637, | |
| "grad_norm": 0.36048009991645813, | |
| "learning_rate": 7.284829046502468e-05, | |
| "loss": 0.7244, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.052505966587112, | |
| "grad_norm": 0.37354418635368347, | |
| "learning_rate": 7.255050596080509e-05, | |
| "loss": 0.7364, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0620525059665873, | |
| "grad_norm": 0.3802691698074341, | |
| "learning_rate": 7.225171392492316e-05, | |
| "loss": 0.7152, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.071599045346062, | |
| "grad_norm": 0.3605879545211792, | |
| "learning_rate": 7.195192770707654e-05, | |
| "loss": 0.6506, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.081145584725537, | |
| "grad_norm": 0.3946097791194916, | |
| "learning_rate": 7.165116070138183e-05, | |
| "loss": 0.715, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.090692124105012, | |
| "grad_norm": 0.35205793380737305, | |
| "learning_rate": 7.134942634577614e-05, | |
| "loss": 0.7169, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.100238663484487, | |
| "grad_norm": 0.4253597557544708, | |
| "learning_rate": 7.104673812141675e-05, | |
| "loss": 0.7926, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1097852028639617, | |
| "grad_norm": 0.43062886595726013, | |
| "learning_rate": 7.07431095520787e-05, | |
| "loss": 0.6923, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.119331742243437, | |
| "grad_norm": 0.38732174038887024, | |
| "learning_rate": 7.04385542035506e-05, | |
| "loss": 0.6641, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.128878281622912, | |
| "grad_norm": 0.37736237049102783, | |
| "learning_rate": 7.013308568302854e-05, | |
| "loss": 0.6602, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1384248210023866, | |
| "grad_norm": 0.3910059928894043, | |
| "learning_rate": 6.982671763850814e-05, | |
| "loss": 0.5671, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.1479713603818618, | |
| "grad_norm": 0.4084903597831726, | |
| "learning_rate": 6.951946375817474e-05, | |
| "loss": 0.7144, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1575178997613365, | |
| "grad_norm": 0.46185556054115295, | |
| "learning_rate": 6.921133776979186e-05, | |
| "loss": 0.7373, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1670644391408116, | |
| "grad_norm": 0.38187095522880554, | |
| "learning_rate": 6.890235344008781e-05, | |
| "loss": 0.6753, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1766109785202863, | |
| "grad_norm": 0.43176108598709106, | |
| "learning_rate": 6.859252457414067e-05, | |
| "loss": 0.683, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1861575178997614, | |
| "grad_norm": 0.43367186188697815, | |
| "learning_rate": 6.828186501476144e-05, | |
| "loss": 0.7548, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.195704057279236, | |
| "grad_norm": 0.3524869978427887, | |
| "learning_rate": 6.797038864187564e-05, | |
| "loss": 0.6032, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2052505966587113, | |
| "grad_norm": 0.40967708826065063, | |
| "learning_rate": 6.765810937190306e-05, | |
| "loss": 0.6378, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.214797136038186, | |
| "grad_norm": 0.3816625773906708, | |
| "learning_rate": 6.734504115713604e-05, | |
| "loss": 0.623, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.224343675417661, | |
| "grad_norm": 0.5187081694602966, | |
| "learning_rate": 6.703119798511612e-05, | |
| "loss": 0.838, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2338902147971362, | |
| "grad_norm": 0.4348510503768921, | |
| "learning_rate": 6.67165938780091e-05, | |
| "loss": 0.7311, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.243436754176611, | |
| "grad_norm": 0.3842396140098572, | |
| "learning_rate": 6.640124289197845e-05, | |
| "loss": 0.6527, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.252983293556086, | |
| "grad_norm": 0.4316810965538025, | |
| "learning_rate": 6.608515911655744e-05, | |
| "loss": 0.7794, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2625298329355608, | |
| "grad_norm": 0.3772410452365875, | |
| "learning_rate": 6.576835667401953e-05, | |
| "loss": 0.6369, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.272076372315036, | |
| "grad_norm": 0.4141497015953064, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.6993, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2816229116945106, | |
| "grad_norm": 0.41354265809059143, | |
| "learning_rate": 6.513265243660057e-05, | |
| "loss": 0.6389, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.2911694510739857, | |
| "grad_norm": 0.3867475986480713, | |
| "learning_rate": 6.481377904428171e-05, | |
| "loss": 0.6306, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.300715990453461, | |
| "grad_norm": 0.40222805738449097, | |
| "learning_rate": 6.449424378870123e-05, | |
| "loss": 0.6857, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3102625298329356, | |
| "grad_norm": 0.4215107858181, | |
| "learning_rate": 6.41740609463409e-05, | |
| "loss": 0.7309, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.3198090692124103, | |
| "grad_norm": 0.4149893820285797, | |
| "learning_rate": 6.385324482261597e-05, | |
| "loss": 0.6562, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3293556085918854, | |
| "grad_norm": 0.4119661748409271, | |
| "learning_rate": 6.353180975123595e-05, | |
| "loss": 0.7544, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.3389021479713605, | |
| "grad_norm": 0.32324427366256714, | |
| "learning_rate": 6.320977009356431e-05, | |
| "loss": 0.5994, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3484486873508352, | |
| "grad_norm": 0.4508344531059265, | |
| "learning_rate": 6.288714023797672e-05, | |
| "loss": 0.7047, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3579952267303104, | |
| "grad_norm": 0.3957417905330658, | |
| "learning_rate": 6.256393459921824e-05, | |
| "loss": 0.6364, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.367541766109785, | |
| "grad_norm": 0.4180348813533783, | |
| "learning_rate": 6.224016761775933e-05, | |
| "loss": 0.6511, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.37708830548926, | |
| "grad_norm": 0.46107926964759827, | |
| "learning_rate": 6.191585375915055e-05, | |
| "loss": 0.6736, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.386634844868735, | |
| "grad_norm": 0.43949881196022034, | |
| "learning_rate": 6.159100751337642e-05, | |
| "loss": 0.6639, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.39618138424821, | |
| "grad_norm": 0.4427139461040497, | |
| "learning_rate": 6.126564339420784e-05, | |
| "loss": 0.6581, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.405727923627685, | |
| "grad_norm": 0.4241901636123657, | |
| "learning_rate": 6.093977593855375e-05, | |
| "loss": 0.6738, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.41527446300716, | |
| "grad_norm": 0.41828441619873047, | |
| "learning_rate": 6.061341970581165e-05, | |
| "loss": 0.685, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.424821002386635, | |
| "grad_norm": 0.4712134599685669, | |
| "learning_rate": 6.028658927721697e-05, | |
| "loss": 0.6853, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.4343675417661097, | |
| "grad_norm": 0.47678568959236145, | |
| "learning_rate": 5.99592992551918e-05, | |
| "loss": 0.673, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.443914081145585, | |
| "grad_norm": 0.46318480372428894, | |
| "learning_rate": 5.9631564262692274e-05, | |
| "loss": 0.688, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.4534606205250595, | |
| "grad_norm": 0.4256531000137329, | |
| "learning_rate": 5.930339894255532e-05, | |
| "loss": 0.6521, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4630071599045347, | |
| "grad_norm": 0.39636510610580444, | |
| "learning_rate": 5.897481795684446e-05, | |
| "loss": 0.6713, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4725536992840094, | |
| "grad_norm": 0.497344434261322, | |
| "learning_rate": 5.8645835986194676e-05, | |
| "loss": 0.7745, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.4821002386634845, | |
| "grad_norm": 0.4814034104347229, | |
| "learning_rate": 5.831646772915651e-05, | |
| "loss": 0.6849, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.4821002386634845, | |
| "eval_loss": 0.7669724225997925, | |
| "eval_runtime": 13.0227, | |
| "eval_samples_per_second": 13.592, | |
| "eval_steps_per_second": 1.766, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.491646778042959, | |
| "grad_norm": 0.4661683142185211, | |
| "learning_rate": 5.7986727901539374e-05, | |
| "loss": 0.7284, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.5011933174224343, | |
| "grad_norm": 0.4617115557193756, | |
| "learning_rate": 5.7656631235754014e-05, | |
| "loss": 0.7026, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5107398568019095, | |
| "grad_norm": 0.48742860555648804, | |
| "learning_rate": 5.732619248015434e-05, | |
| "loss": 0.6964, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.520286396181384, | |
| "grad_norm": 0.4442538917064667, | |
| "learning_rate": 5.699542639837844e-05, | |
| "loss": 0.6804, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.5298329355608593, | |
| "grad_norm": 0.49106183648109436, | |
| "learning_rate": 5.666434776868895e-05, | |
| "loss": 0.6865, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.539379474940334, | |
| "grad_norm": 0.4551374316215515, | |
| "learning_rate": 5.633297138331285e-05, | |
| "loss": 0.7748, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.548926014319809, | |
| "grad_norm": 0.413185715675354, | |
| "learning_rate": 5.6001312047780486e-05, | |
| "loss": 0.5999, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.558472553699284, | |
| "grad_norm": 0.44866228103637695, | |
| "learning_rate": 5.5669384580264104e-05, | |
| "loss": 0.6102, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.568019093078759, | |
| "grad_norm": 0.45962679386138916, | |
| "learning_rate": 5.533720381091582e-05, | |
| "loss": 0.7214, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.577565632458234, | |
| "grad_norm": 0.45130231976509094, | |
| "learning_rate": 5.5004784581204927e-05, | |
| "loss": 0.7123, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.587112171837709, | |
| "grad_norm": 0.44015398621559143, | |
| "learning_rate": 5.467214174325493e-05, | |
| "loss": 0.6339, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.5966587112171835, | |
| "grad_norm": 0.45853525400161743, | |
| "learning_rate": 5.4339290159179875e-05, | |
| "loss": 0.7224, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6062052505966586, | |
| "grad_norm": 0.45820721983909607, | |
| "learning_rate": 5.400624470042037e-05, | |
| "loss": 0.638, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.615751789976134, | |
| "grad_norm": 0.4199580252170563, | |
| "learning_rate": 5.367302024707911e-05, | |
| "loss": 0.6548, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.6252983293556085, | |
| "grad_norm": 0.5202221870422363, | |
| "learning_rate": 5.3339631687256084e-05, | |
| "loss": 0.7328, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6348448687350836, | |
| "grad_norm": 0.41012313961982727, | |
| "learning_rate": 5.300609391638336e-05, | |
| "loss": 0.6315, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6443914081145583, | |
| "grad_norm": 0.5391274094581604, | |
| "learning_rate": 5.267242183655961e-05, | |
| "loss": 0.7476, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6539379474940334, | |
| "grad_norm": 0.4461412727832794, | |
| "learning_rate": 5.233863035588426e-05, | |
| "loss": 0.6223, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.663484486873508, | |
| "grad_norm": 0.4335575997829437, | |
| "learning_rate": 5.200473438779146e-05, | |
| "loss": 0.6816, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6730310262529833, | |
| "grad_norm": 0.44996485114097595, | |
| "learning_rate": 5.167074885038373e-05, | |
| "loss": 0.6554, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6825775656324584, | |
| "grad_norm": 0.4875689744949341, | |
| "learning_rate": 5.133668866576544e-05, | |
| "loss": 0.7889, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.692124105011933, | |
| "grad_norm": 0.45980238914489746, | |
| "learning_rate": 5.1002568759376134e-05, | |
| "loss": 0.6615, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7016706443914082, | |
| "grad_norm": 0.43575319647789, | |
| "learning_rate": 5.0668404059323635e-05, | |
| "loss": 0.6595, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.711217183770883, | |
| "grad_norm": 0.438365638256073, | |
| "learning_rate": 5.033420949571712e-05, | |
| "loss": 0.6354, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.720763723150358, | |
| "grad_norm": 0.4110269546508789, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6814, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7303102625298328, | |
| "grad_norm": 0.4653089642524719, | |
| "learning_rate": 4.96657905042829e-05, | |
| "loss": 0.7375, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.739856801909308, | |
| "grad_norm": 0.45703983306884766, | |
| "learning_rate": 4.933159594067637e-05, | |
| "loss": 0.7464, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.749403341288783, | |
| "grad_norm": 0.40393638610839844, | |
| "learning_rate": 4.899743124062388e-05, | |
| "loss": 0.6091, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7589498806682577, | |
| "grad_norm": 0.468176007270813, | |
| "learning_rate": 4.866331133423456e-05, | |
| "loss": 0.7052, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.7684964200477324, | |
| "grad_norm": 0.4771003723144531, | |
| "learning_rate": 4.832925114961629e-05, | |
| "loss": 0.7071, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7780429594272076, | |
| "grad_norm": 0.4809763729572296, | |
| "learning_rate": 4.799526561220855e-05, | |
| "loss": 0.7289, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.7875894988066827, | |
| "grad_norm": 0.46262210607528687, | |
| "learning_rate": 4.7661369644115755e-05, | |
| "loss": 0.6031, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.7971360381861574, | |
| "grad_norm": 0.48082613945007324, | |
| "learning_rate": 4.73275781634404e-05, | |
| "loss": 0.7008, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.8066825775656326, | |
| "grad_norm": 0.5203619003295898, | |
| "learning_rate": 4.6993906083616654e-05, | |
| "loss": 0.6972, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.8162291169451072, | |
| "grad_norm": 0.5060526728630066, | |
| "learning_rate": 4.666036831274392e-05, | |
| "loss": 0.844, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8257756563245824, | |
| "grad_norm": 0.5173178911209106, | |
| "learning_rate": 4.63269797529209e-05, | |
| "loss": 0.711, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.835322195704057, | |
| "grad_norm": 0.40481331944465637, | |
| "learning_rate": 4.5993755299579626e-05, | |
| "loss": 0.6849, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.844868735083532, | |
| "grad_norm": 0.5248334407806396, | |
| "learning_rate": 4.566070984082013e-05, | |
| "loss": 0.8126, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8544152744630074, | |
| "grad_norm": 0.469752699136734, | |
| "learning_rate": 4.5327858256745073e-05, | |
| "loss": 0.7265, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.863961813842482, | |
| "grad_norm": 0.4676867425441742, | |
| "learning_rate": 4.4995215418795085e-05, | |
| "loss": 0.7859, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.873508353221957, | |
| "grad_norm": 0.43738093972206116, | |
| "learning_rate": 4.466279618908419e-05, | |
| "loss": 0.5927, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.883054892601432, | |
| "grad_norm": 0.4657769799232483, | |
| "learning_rate": 4.433061541973591e-05, | |
| "loss": 0.6947, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.892601431980907, | |
| "grad_norm": 0.44069626927375793, | |
| "learning_rate": 4.399868795221951e-05, | |
| "loss": 0.6984, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.9021479713603817, | |
| "grad_norm": 0.3833423852920532, | |
| "learning_rate": 4.366702861668716e-05, | |
| "loss": 0.5624, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.911694510739857, | |
| "grad_norm": 0.42630308866500854, | |
| "learning_rate": 4.333565223131107e-05, | |
| "loss": 0.5505, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.921241050119332, | |
| "grad_norm": 0.46541622281074524, | |
| "learning_rate": 4.300457360162158e-05, | |
| "loss": 0.7071, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9307875894988067, | |
| "grad_norm": 0.41140827536582947, | |
| "learning_rate": 4.267380751984568e-05, | |
| "loss": 0.5786, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9403341288782814, | |
| "grad_norm": 0.42655548453330994, | |
| "learning_rate": 4.2343368764246e-05, | |
| "loss": 0.5977, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9498806682577565, | |
| "grad_norm": 0.45397818088531494, | |
| "learning_rate": 4.201327209846065e-05, | |
| "loss": 0.6592, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.9594272076372317, | |
| "grad_norm": 0.5159749984741211, | |
| "learning_rate": 4.1683532270843504e-05, | |
| "loss": 0.7322, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9689737470167064, | |
| "grad_norm": 0.4518764615058899, | |
| "learning_rate": 4.135416401380535e-05, | |
| "loss": 0.796, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9785202863961815, | |
| "grad_norm": 0.47323092818260193, | |
| "learning_rate": 4.102518204315555e-05, | |
| "loss": 0.7456, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9785202863961815, | |
| "eval_loss": 0.7645925879478455, | |
| "eval_runtime": 13.0059, | |
| "eval_samples_per_second": 13.609, | |
| "eval_steps_per_second": 1.768, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.988066825775656, | |
| "grad_norm": 0.49233347177505493, | |
| "learning_rate": 4.069660105744469e-05, | |
| "loss": 0.6878, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.9976133651551313, | |
| "grad_norm": 0.4664769470691681, | |
| "learning_rate": 4.036843573730774e-05, | |
| "loss": 0.7379, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 3.007159904534606, | |
| "grad_norm": 0.39737001061439514, | |
| "learning_rate": 4.0040700744808204e-05, | |
| "loss": 0.6706, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 3.016706443914081, | |
| "grad_norm": 0.46519792079925537, | |
| "learning_rate": 3.971341072278302e-05, | |
| "loss": 0.6626, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 3.0262529832935563, | |
| "grad_norm": 0.4189288914203644, | |
| "learning_rate": 3.938658029418837e-05, | |
| "loss": 0.6465, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 3.035799522673031, | |
| "grad_norm": 0.5005217790603638, | |
| "learning_rate": 3.9060224061446245e-05, | |
| "loss": 0.6552, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.045346062052506, | |
| "grad_norm": 0.4551185965538025, | |
| "learning_rate": 3.873435660579217e-05, | |
| "loss": 0.5767, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 3.054892601431981, | |
| "grad_norm": 0.44225507974624634, | |
| "learning_rate": 3.840899248662358e-05, | |
| "loss": 0.6387, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.064439140811456, | |
| "grad_norm": 0.5424272418022156, | |
| "learning_rate": 3.808414624084946e-05, | |
| "loss": 0.6887, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 3.0739856801909307, | |
| "grad_norm": 0.4525783658027649, | |
| "learning_rate": 3.77598323822407e-05, | |
| "loss": 0.6612, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 3.083532219570406, | |
| "grad_norm": 0.4461750388145447, | |
| "learning_rate": 3.7436065400781774e-05, | |
| "loss": 0.6418, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 3.0930787589498805, | |
| "grad_norm": 0.524474561214447, | |
| "learning_rate": 3.7112859762023314e-05, | |
| "loss": 0.6662, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 3.1026252983293556, | |
| "grad_norm": 0.5077919960021973, | |
| "learning_rate": 3.6790229906435705e-05, | |
| "loss": 0.6368, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.1121718377088303, | |
| "grad_norm": 0.5111605525016785, | |
| "learning_rate": 3.646819024876406e-05, | |
| "loss": 0.6471, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 3.1217183770883055, | |
| "grad_norm": 0.47017544507980347, | |
| "learning_rate": 3.614675517738405e-05, | |
| "loss": 0.6412, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 3.1312649164677806, | |
| "grad_norm": 0.5017347931861877, | |
| "learning_rate": 3.5825939053659116e-05, | |
| "loss": 0.6719, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 3.1408114558472553, | |
| "grad_norm": 0.4742693305015564, | |
| "learning_rate": 3.550575621129878e-05, | |
| "loss": 0.6389, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 3.1503579952267304, | |
| "grad_norm": 0.48239150643348694, | |
| "learning_rate": 3.5186220955718306e-05, | |
| "loss": 0.6071, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.159904534606205, | |
| "grad_norm": 0.461012065410614, | |
| "learning_rate": 3.486734756339943e-05, | |
| "loss": 0.5409, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 3.1694510739856803, | |
| "grad_norm": 0.5519300103187561, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.6617, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.178997613365155, | |
| "grad_norm": 0.4272817373275757, | |
| "learning_rate": 3.423164332598049e-05, | |
| "loss": 0.4943, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.18854415274463, | |
| "grad_norm": 0.5129975080490112, | |
| "learning_rate": 3.391484088344257e-05, | |
| "loss": 0.6569, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.1980906921241052, | |
| "grad_norm": 0.5243316292762756, | |
| "learning_rate": 3.3598757108021546e-05, | |
| "loss": 0.6659, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.20763723150358, | |
| "grad_norm": 0.4862256348133087, | |
| "learning_rate": 3.3283406121990915e-05, | |
| "loss": 0.6414, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.217183770883055, | |
| "grad_norm": 0.5197688937187195, | |
| "learning_rate": 3.2968802014883874e-05, | |
| "loss": 0.6751, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.2267303102625298, | |
| "grad_norm": 0.5474854111671448, | |
| "learning_rate": 3.265495884286397e-05, | |
| "loss": 0.7531, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.236276849642005, | |
| "grad_norm": 0.47088584303855896, | |
| "learning_rate": 3.234189062809695e-05, | |
| "loss": 0.6319, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.2458233890214796, | |
| "grad_norm": 0.4922322630882263, | |
| "learning_rate": 3.202961135812437e-05, | |
| "loss": 0.6297, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.2553699284009547, | |
| "grad_norm": 0.5747278332710266, | |
| "learning_rate": 3.1718134985238567e-05, | |
| "loss": 0.6269, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.2649164677804294, | |
| "grad_norm": 0.5492677092552185, | |
| "learning_rate": 3.1407475425859345e-05, | |
| "loss": 0.7295, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.2744630071599046, | |
| "grad_norm": 0.487875759601593, | |
| "learning_rate": 3.109764655991221e-05, | |
| "loss": 0.6013, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.2840095465393793, | |
| "grad_norm": 0.5148051381111145, | |
| "learning_rate": 3.078866223020815e-05, | |
| "loss": 0.7156, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.2935560859188544, | |
| "grad_norm": 0.4531029760837555, | |
| "learning_rate": 3.0480536241825263e-05, | |
| "loss": 0.581, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.3031026252983295, | |
| "grad_norm": 0.4673330783843994, | |
| "learning_rate": 3.0173282361491868e-05, | |
| "loss": 0.5355, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.3126491646778042, | |
| "grad_norm": 0.4933515191078186, | |
| "learning_rate": 2.9866914316971477e-05, | |
| "loss": 0.5584, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.3221957040572794, | |
| "grad_norm": 0.5033287405967712, | |
| "learning_rate": 2.9561445796449415e-05, | |
| "loss": 0.6598, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.331742243436754, | |
| "grad_norm": 0.5585161447525024, | |
| "learning_rate": 2.925689044792132e-05, | |
| "loss": 0.7268, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.341288782816229, | |
| "grad_norm": 0.4780008792877197, | |
| "learning_rate": 2.895326187858326e-05, | |
| "loss": 0.6363, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.350835322195704, | |
| "grad_norm": 0.5018965005874634, | |
| "learning_rate": 2.865057365422386e-05, | |
| "loss": 0.6098, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.360381861575179, | |
| "grad_norm": 0.4768741726875305, | |
| "learning_rate": 2.8348839298618178e-05, | |
| "loss": 0.6504, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.369928400954654, | |
| "grad_norm": 0.47006916999816895, | |
| "learning_rate": 2.8048072292923465e-05, | |
| "loss": 0.7122, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.379474940334129, | |
| "grad_norm": 0.5839781165122986, | |
| "learning_rate": 2.7748286075076835e-05, | |
| "loss": 0.688, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.389021479713604, | |
| "grad_norm": 0.48642516136169434, | |
| "learning_rate": 2.74494940391949e-05, | |
| "loss": 0.5968, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.3985680190930787, | |
| "grad_norm": 0.5563519597053528, | |
| "learning_rate": 2.7151709534975324e-05, | |
| "loss": 0.6904, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.408114558472554, | |
| "grad_norm": 0.489467591047287, | |
| "learning_rate": 2.685494586710038e-05, | |
| "loss": 0.6965, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.4176610978520285, | |
| "grad_norm": 0.5098894238471985, | |
| "learning_rate": 2.655921629464245e-05, | |
| "loss": 0.5593, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.4272076372315037, | |
| "grad_norm": 0.5460035800933838, | |
| "learning_rate": 2.626453403047172e-05, | |
| "loss": 0.7281, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.4367541766109784, | |
| "grad_norm": 0.5174149870872498, | |
| "learning_rate": 2.5970912240665813e-05, | |
| "loss": 0.6505, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.4463007159904535, | |
| "grad_norm": 0.5922746658325195, | |
| "learning_rate": 2.5678364043921504e-05, | |
| "loss": 0.7072, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.455847255369928, | |
| "grad_norm": 0.501711368560791, | |
| "learning_rate": 2.5386902510968625e-05, | |
| "loss": 0.7021, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.4653937947494033, | |
| "grad_norm": 0.551275372505188, | |
| "learning_rate": 2.5096540663986067e-05, | |
| "loss": 0.6635, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.4749403341288785, | |
| "grad_norm": 0.510543942451477, | |
| "learning_rate": 2.4807291476019995e-05, | |
| "loss": 0.6037, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.4749403341288785, | |
| "eval_loss": 0.7706080079078674, | |
| "eval_runtime": 12.9841, | |
| "eval_samples_per_second": 13.632, | |
| "eval_steps_per_second": 1.771, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.484486873508353, | |
| "grad_norm": 0.5204513072967529, | |
| "learning_rate": 2.4519167870404125e-05, | |
| "loss": 0.6473, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.4940334128878283, | |
| "grad_norm": 0.5144868493080139, | |
| "learning_rate": 2.4232182720182522e-05, | |
| "loss": 0.5738, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.503579952267303, | |
| "grad_norm": 0.4935360848903656, | |
| "learning_rate": 2.3946348847534194e-05, | |
| "loss": 0.6041, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.513126491646778, | |
| "grad_norm": 0.5450278520584106, | |
| "learning_rate": 2.3661679023200422e-05, | |
| "loss": 0.6976, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.522673031026253, | |
| "grad_norm": 0.5161455869674683, | |
| "learning_rate": 2.337818596591408e-05, | |
| "loss": 0.7566, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.532219570405728, | |
| "grad_norm": 0.47960028052330017, | |
| "learning_rate": 2.3095882341831372e-05, | |
| "loss": 0.6354, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.541766109785203, | |
| "grad_norm": 0.46854427456855774, | |
| "learning_rate": 2.281478076396596e-05, | |
| "loss": 0.6636, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.551312649164678, | |
| "grad_norm": 0.49230995774269104, | |
| "learning_rate": 2.2534893791625406e-05, | |
| "loss": 0.5336, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.5608591885441525, | |
| "grad_norm": 0.475827693939209, | |
| "learning_rate": 2.2256233929850044e-05, | |
| "loss": 0.6397, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.5704057279236276, | |
| "grad_norm": 0.5518814921379089, | |
| "learning_rate": 2.197881362885426e-05, | |
| "loss": 0.6751, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.579952267303103, | |
| "grad_norm": 0.5297905802726746, | |
| "learning_rate": 2.1702645283470236e-05, | |
| "loss": 0.6379, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.5894988066825775, | |
| "grad_norm": 0.5582184195518494, | |
| "learning_rate": 2.1427741232594184e-05, | |
| "loss": 0.6456, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.5990453460620526, | |
| "grad_norm": 0.6382442712783813, | |
| "learning_rate": 2.115411375863497e-05, | |
| "loss": 0.7669, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.6085918854415273, | |
| "grad_norm": 0.5114546418190002, | |
| "learning_rate": 2.0881775086965495e-05, | |
| "loss": 0.6209, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.6181384248210025, | |
| "grad_norm": 0.5362977981567383, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.5689, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.627684964200477, | |
| "grad_norm": 0.5577702522277832, | |
| "learning_rate": 2.0341012763532243e-05, | |
| "loss": 0.68, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.6372315035799523, | |
| "grad_norm": 0.4712672233581543, | |
| "learning_rate": 2.0072613272430923e-05, | |
| "loss": 0.6012, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.6467780429594274, | |
| "grad_norm": 0.5355048179626465, | |
| "learning_rate": 1.9805550903864774e-05, | |
| "loss": 0.6409, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.656324582338902, | |
| "grad_norm": 0.5459690093994141, | |
| "learning_rate": 1.9539837589885024e-05, | |
| "loss": 0.615, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.665871121718377, | |
| "grad_norm": 0.5169499516487122, | |
| "learning_rate": 1.9275485202268572e-05, | |
| "loss": 0.5431, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.675417661097852, | |
| "grad_norm": 0.4908435344696045, | |
| "learning_rate": 1.9012505551987765e-05, | |
| "loss": 0.5995, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.684964200477327, | |
| "grad_norm": 0.5475894808769226, | |
| "learning_rate": 1.875091038868243e-05, | |
| "loss": 0.5709, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.694510739856802, | |
| "grad_norm": 0.4647465944290161, | |
| "learning_rate": 1.8490711400135118e-05, | |
| "loss": 0.6245, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.704057279236277, | |
| "grad_norm": 0.5150587558746338, | |
| "learning_rate": 1.823192021174882e-05, | |
| "loss": 0.6999, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.713603818615752, | |
| "grad_norm": 0.5281825661659241, | |
| "learning_rate": 1.7974548386027585e-05, | |
| "loss": 0.6953, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.7231503579952268, | |
| "grad_norm": 0.5862619876861572, | |
| "learning_rate": 1.771860742205988e-05, | |
| "loss": 0.7626, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.7326968973747015, | |
| "grad_norm": 0.5895105600357056, | |
| "learning_rate": 1.746410875500488e-05, | |
| "loss": 0.7215, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.7422434367541766, | |
| "grad_norm": 0.5397690534591675, | |
| "learning_rate": 1.7211063755581525e-05, | |
| "loss": 0.6534, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.7517899761336517, | |
| "grad_norm": 0.5140255689620972, | |
| "learning_rate": 1.695948372956047e-05, | |
| "loss": 0.6316, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.7613365155131264, | |
| "grad_norm": 0.555587887763977, | |
| "learning_rate": 1.6709379917259028e-05, | |
| "loss": 0.5928, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.7708830548926016, | |
| "grad_norm": 0.5376729965209961, | |
| "learning_rate": 1.646076349303884e-05, | |
| "loss": 0.7066, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.7804295942720763, | |
| "grad_norm": 0.5037944912910461, | |
| "learning_rate": 1.621364556480675e-05, | |
| "loss": 0.7138, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.7899761336515514, | |
| "grad_norm": 0.5480453372001648, | |
| "learning_rate": 1.596803717351845e-05, | |
| "loss": 0.6835, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.799522673031026, | |
| "grad_norm": 0.5427130460739136, | |
| "learning_rate": 1.5723949292685192e-05, | |
| "loss": 0.645, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.8090692124105012, | |
| "grad_norm": 0.5332582592964172, | |
| "learning_rate": 1.548139282788349e-05, | |
| "loss": 0.6536, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 3.8186157517899764, | |
| "grad_norm": 0.45773863792419434, | |
| "learning_rate": 1.5240378616267886e-05, | |
| "loss": 0.4132, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.828162291169451, | |
| "grad_norm": 0.5269774794578552, | |
| "learning_rate": 1.5000917426086768e-05, | |
| "loss": 0.709, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 3.8377088305489258, | |
| "grad_norm": 0.5353370904922485, | |
| "learning_rate": 1.4763019956201252e-05, | |
| "loss": 0.6245, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.847255369928401, | |
| "grad_norm": 0.5651338696479797, | |
| "learning_rate": 1.452669683560709e-05, | |
| "loss": 0.7354, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 3.856801909307876, | |
| "grad_norm": 0.5225863456726074, | |
| "learning_rate": 1.4291958622959973e-05, | |
| "loss": 0.6269, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.8663484486873507, | |
| "grad_norm": 0.4741249680519104, | |
| "learning_rate": 1.4058815806103542e-05, | |
| "loss": 0.6367, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.875894988066826, | |
| "grad_norm": 0.4927036166191101, | |
| "learning_rate": 1.3827278801600979e-05, | |
| "loss": 0.5917, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 3.8854415274463006, | |
| "grad_norm": 0.528948187828064, | |
| "learning_rate": 1.3597357954269535e-05, | |
| "loss": 0.6245, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 3.8949880668257757, | |
| "grad_norm": 0.499723881483078, | |
| "learning_rate": 1.3369063536718345e-05, | |
| "loss": 0.6612, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.9045346062052504, | |
| "grad_norm": 0.5520623326301575, | |
| "learning_rate": 1.3142405748889457e-05, | |
| "loss": 0.6276, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 3.9140811455847255, | |
| "grad_norm": 0.519705593585968, | |
| "learning_rate": 1.2917394717602121e-05, | |
| "loss": 0.6639, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.9236276849642007, | |
| "grad_norm": 0.5106028318405151, | |
| "learning_rate": 1.2694040496100318e-05, | |
| "loss": 0.6402, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 3.9331742243436754, | |
| "grad_norm": 0.5074647665023804, | |
| "learning_rate": 1.2472353063603625e-05, | |
| "loss": 0.6424, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 3.9427207637231505, | |
| "grad_norm": 0.594458281993866, | |
| "learning_rate": 1.2252342324861272e-05, | |
| "loss": 0.6677, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 3.952267303102625, | |
| "grad_norm": 0.5148235559463501, | |
| "learning_rate": 1.2034018109709716e-05, | |
| "loss": 0.5508, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 3.9618138424821003, | |
| "grad_norm": 0.5786213278770447, | |
| "learning_rate": 1.1817390172633403e-05, | |
| "loss": 0.5669, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.971360381861575, | |
| "grad_norm": 0.5178529024124146, | |
| "learning_rate": 1.1602468192328936e-05, | |
| "loss": 0.6335, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.971360381861575, | |
| "eval_loss": 0.7703084945678711, | |
| "eval_runtime": 13.1275, | |
| "eval_samples_per_second": 13.483, | |
| "eval_steps_per_second": 1.752, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.98090692124105, | |
| "grad_norm": 0.6362994313240051, | |
| "learning_rate": 1.1389261771272663e-05, | |
| "loss": 0.667, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 3.9904534606205253, | |
| "grad_norm": 0.46682727336883545, | |
| "learning_rate": 1.117778043529164e-05, | |
| "loss": 0.5675, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5248848795890808, | |
| "learning_rate": 1.096803363313803e-05, | |
| "loss": 0.6431, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 4.009546539379475, | |
| "grad_norm": 0.4764450192451477, | |
| "learning_rate": 1.0760030736066951e-05, | |
| "loss": 0.5765, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.01909307875895, | |
| "grad_norm": 0.5138590335845947, | |
| "learning_rate": 1.055378103741777e-05, | |
| "loss": 0.5938, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 4.028639618138425, | |
| "grad_norm": 0.560528039932251, | |
| "learning_rate": 1.034929375219884e-05, | |
| "loss": 0.7372, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 4.0381861575179, | |
| "grad_norm": 0.5144191384315491, | |
| "learning_rate": 1.0146578016675934e-05, | |
| "loss": 0.6935, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 4.047732696897374, | |
| "grad_norm": 0.4730660915374756, | |
| "learning_rate": 9.945642887963841e-06, | |
| "loss": 0.5619, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 4.05727923627685, | |
| "grad_norm": 0.5379086136817932, | |
| "learning_rate": 9.746497343621857e-06, | |
| "loss": 0.5955, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 4.066825775656325, | |
| "grad_norm": 0.46496352553367615, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.5427, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 4.076372315035799, | |
| "grad_norm": 0.4938548803329468, | |
| "learning_rate": 9.353610518104611e-06, | |
| "loss": 0.6966, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 4.085918854415274, | |
| "grad_norm": 0.5221308469772339, | |
| "learning_rate": 9.159886790678124e-06, | |
| "loss": 0.6781, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 4.09546539379475, | |
| "grad_norm": 0.4903409779071808, | |
| "learning_rate": 8.967987754335022e-06, | |
| "loss": 0.6055, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 4.105011933174224, | |
| "grad_norm": 0.5302107930183411, | |
| "learning_rate": 8.777921982911996e-06, | |
| "loss": 0.5903, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 4.114558472553699, | |
| "grad_norm": 0.4461494982242584, | |
| "learning_rate": 8.589697968337446e-06, | |
| "loss": 0.5499, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 4.124105011933175, | |
| "grad_norm": 0.6021783351898193, | |
| "learning_rate": 8.40332412025216e-06, | |
| "loss": 0.6363, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 4.133651551312649, | |
| "grad_norm": 0.5821279883384705, | |
| "learning_rate": 8.218808765633512e-06, | |
| "loss": 0.5479, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 4.143198090692124, | |
| "grad_norm": 0.5231468677520752, | |
| "learning_rate": 8.036160148423449e-06, | |
| "loss": 0.6066, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 4.152744630071599, | |
| "grad_norm": 0.5418100357055664, | |
| "learning_rate": 7.85538642916015e-06, | |
| "loss": 0.6153, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 4.162291169451074, | |
| "grad_norm": 0.5310184359550476, | |
| "learning_rate": 7.676495684613432e-06, | |
| "loss": 0.541, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 4.171837708830549, | |
| "grad_norm": 0.5160778164863586, | |
| "learning_rate": 7.499495907423887e-06, | |
| "loss": 0.6041, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 4.181384248210024, | |
| "grad_norm": 0.5340070128440857, | |
| "learning_rate": 7.324395005745771e-06, | |
| "loss": 0.6006, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 4.190930787589499, | |
| "grad_norm": 0.5116534233093262, | |
| "learning_rate": 7.151200802893682e-06, | |
| "loss": 0.6093, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 4.200477326968974, | |
| "grad_norm": 0.5754937529563904, | |
| "learning_rate": 6.979921036993042e-06, | |
| "loss": 0.6219, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 4.210023866348449, | |
| "grad_norm": 0.6205330491065979, | |
| "learning_rate": 6.810563360634298e-06, | |
| "loss": 0.6091, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 4.219570405727923, | |
| "grad_norm": 0.47774946689605713, | |
| "learning_rate": 6.643135340531137e-06, | |
| "loss": 0.5943, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 4.229116945107399, | |
| "grad_norm": 0.5198448896408081, | |
| "learning_rate": 6.477644457182275e-06, | |
| "loss": 0.711, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 4.238663484486874, | |
| "grad_norm": 0.548129141330719, | |
| "learning_rate": 6.314098104537325e-06, | |
| "loss": 0.5444, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 4.248210023866348, | |
| "grad_norm": 0.549915611743927, | |
| "learning_rate": 6.152503589666425e-06, | |
| "loss": 0.6128, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 4.257756563245824, | |
| "grad_norm": 0.473899245262146, | |
| "learning_rate": 5.992868132433754e-06, | |
| "loss": 0.5761, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 4.2673031026252985, | |
| "grad_norm": 0.6241645216941833, | |
| "learning_rate": 5.835198865174956e-06, | |
| "loss": 0.6581, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 4.276849642004773, | |
| "grad_norm": 0.5191392302513123, | |
| "learning_rate": 5.679502832378497e-06, | |
| "loss": 0.6428, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 4.286396181384248, | |
| "grad_norm": 0.5194590091705322, | |
| "learning_rate": 5.5257869903709015e-06, | |
| "loss": 0.6437, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 4.2959427207637235, | |
| "grad_norm": 0.565933108329773, | |
| "learning_rate": 5.374058207005944e-06, | |
| "loss": 0.6357, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.305489260143198, | |
| "grad_norm": 0.5410757064819336, | |
| "learning_rate": 5.224323261357844e-06, | |
| "loss": 0.6178, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 4.315035799522673, | |
| "grad_norm": 0.5989289283752441, | |
| "learning_rate": 5.0765888434183454e-06, | |
| "loss": 0.6333, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 4.324582338902148, | |
| "grad_norm": 0.48727279901504517, | |
| "learning_rate": 4.930861553797822e-06, | |
| "loss": 0.5238, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 4.334128878281623, | |
| "grad_norm": 0.5582761764526367, | |
| "learning_rate": 4.7871479034303835e-06, | |
| "loss": 0.5644, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 4.343675417661098, | |
| "grad_norm": 0.5856003165245056, | |
| "learning_rate": 4.645454313282965e-06, | |
| "loss": 0.7528, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 4.353221957040573, | |
| "grad_norm": 0.5278400778770447, | |
| "learning_rate": 4.505787114068432e-06, | |
| "loss": 0.5257, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 4.362768496420047, | |
| "grad_norm": 0.555724024772644, | |
| "learning_rate": 4.3681525459627614e-06, | |
| "loss": 0.6286, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 4.372315035799523, | |
| "grad_norm": 0.5980591773986816, | |
| "learning_rate": 4.232556758326212e-06, | |
| "loss": 0.7107, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 4.3818615751789975, | |
| "grad_norm": 0.5622268915176392, | |
| "learning_rate": 4.099005809428596e-06, | |
| "loss": 0.6305, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 4.391408114558472, | |
| "grad_norm": 0.5708304643630981, | |
| "learning_rate": 3.967505666178556e-06, | |
| "loss": 0.5354, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.400954653937948, | |
| "grad_norm": 0.5751599073410034, | |
| "learning_rate": 3.838062203857074e-06, | |
| "loss": 0.5867, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 4.4105011933174225, | |
| "grad_norm": 0.5867375135421753, | |
| "learning_rate": 3.7106812058548377e-06, | |
| "loss": 0.5848, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 4.420047732696897, | |
| "grad_norm": 0.5069125294685364, | |
| "learning_rate": 3.5853683634139434e-06, | |
| "loss": 0.5737, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 4.429594272076372, | |
| "grad_norm": 0.5603401064872742, | |
| "learning_rate": 3.462129275373577e-06, | |
| "loss": 0.6771, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 4.4391408114558475, | |
| "grad_norm": 0.4407503306865692, | |
| "learning_rate": 3.340969447919873e-06, | |
| "loss": 0.4828, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 4.448687350835322, | |
| "grad_norm": 0.5868078470230103, | |
| "learning_rate": 3.2218942943399112e-06, | |
| "loss": 0.6198, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 4.458233890214797, | |
| "grad_norm": 0.5815831422805786, | |
| "learning_rate": 3.104909134779821e-06, | |
| "loss": 0.6983, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 4.4677804295942725, | |
| "grad_norm": 0.5096269845962524, | |
| "learning_rate": 2.9900191960071544e-06, | |
| "loss": 0.5835, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.4677804295942725, | |
| "eval_loss": 0.774871826171875, | |
| "eval_runtime": 13.0266, | |
| "eval_samples_per_second": 13.588, | |
| "eval_steps_per_second": 1.766, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.477326968973747, | |
| "grad_norm": 0.5428768396377563, | |
| "learning_rate": 2.877229611177268e-06, | |
| "loss": 0.6312, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 4.486873508353222, | |
| "grad_norm": 0.6652801632881165, | |
| "learning_rate": 2.7665454196040664e-06, | |
| "loss": 0.564, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.4964200477326965, | |
| "grad_norm": 0.6429077982902527, | |
| "learning_rate": 2.6579715665347893e-06, | |
| "loss": 0.6679, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 4.505966587112172, | |
| "grad_norm": 0.5690730810165405, | |
| "learning_rate": 2.5515129029290986e-06, | |
| "loss": 0.615, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 4.515513126491647, | |
| "grad_norm": 0.6044589877128601, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.6089, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 4.5250596658711215, | |
| "grad_norm": 0.5761138200759888, | |
| "learning_rate": 2.34496007521296e-06, | |
| "loss": 0.6444, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 4.534606205250597, | |
| "grad_norm": 0.6119807958602905, | |
| "learning_rate": 2.2448751396543787e-06, | |
| "loss": 0.7381, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.544152744630072, | |
| "grad_norm": 0.5442492961883545, | |
| "learning_rate": 2.1469238502507925e-06, | |
| "loss": 0.6084, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 4.5536992840095465, | |
| "grad_norm": 0.5115097761154175, | |
| "learning_rate": 2.0511105833574683e-06, | |
| "loss": 0.5036, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 4.563245823389021, | |
| "grad_norm": 0.5502949953079224, | |
| "learning_rate": 1.957439619805196e-06, | |
| "loss": 0.6587, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 4.572792362768497, | |
| "grad_norm": 0.5296341776847839, | |
| "learning_rate": 1.865915144708985e-06, | |
| "loss": 0.6043, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 4.5823389021479715, | |
| "grad_norm": 0.5244660377502441, | |
| "learning_rate": 1.7765412472811771e-06, | |
| "loss": 0.7407, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.591885441527446, | |
| "grad_norm": 0.4937599301338196, | |
| "learning_rate": 1.6893219206486233e-06, | |
| "loss": 0.653, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 4.601431980906922, | |
| "grad_norm": 0.5165035128593445, | |
| "learning_rate": 1.6042610616743781e-06, | |
| "loss": 0.6745, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 4.610978520286396, | |
| "grad_norm": 0.5757156014442444, | |
| "learning_rate": 1.5213624707835273e-06, | |
| "loss": 0.7117, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 4.620525059665871, | |
| "grad_norm": 0.5235234498977661, | |
| "learning_rate": 1.4406298517934069e-06, | |
| "loss": 0.5699, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 4.630071599045346, | |
| "grad_norm": 0.5536040663719177, | |
| "learning_rate": 1.3620668117481472e-06, | |
| "loss": 0.6527, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 4.6396181384248205, | |
| "grad_norm": 0.5768601894378662, | |
| "learning_rate": 1.2856768607574566e-06, | |
| "loss": 0.7039, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 4.649164677804296, | |
| "grad_norm": 0.5293567776679993, | |
| "learning_rate": 1.2114634118398638e-06, | |
| "loss": 0.5016, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 4.658711217183771, | |
| "grad_norm": 0.4922030568122864, | |
| "learning_rate": 1.1394297807701737e-06, | |
| "loss": 0.6378, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.6682577565632455, | |
| "grad_norm": 0.5279848575592041, | |
| "learning_rate": 1.0695791859313298e-06, | |
| "loss": 0.5854, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 4.677804295942721, | |
| "grad_norm": 0.5893853902816772, | |
| "learning_rate": 1.0019147481706625e-06, | |
| "loss": 0.6481, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.687350835322196, | |
| "grad_norm": 0.5379951000213623, | |
| "learning_rate": 9.364394906603901e-07, | |
| "loss": 0.5932, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 4.6968973747016705, | |
| "grad_norm": 0.5492742657661438, | |
| "learning_rate": 8.731563387626096e-07, | |
| "loss": 0.6475, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 4.706443914081145, | |
| "grad_norm": 0.596684455871582, | |
| "learning_rate": 8.120681198985292e-07, | |
| "loss": 0.6257, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 4.715990453460621, | |
| "grad_norm": 0.5613247156143188, | |
| "learning_rate": 7.531775634222138e-07, | |
| "loss": 0.6233, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 4.725536992840095, | |
| "grad_norm": 0.5937960743904114, | |
| "learning_rate": 6.964873004985717e-07, | |
| "loss": 0.6734, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 4.73508353221957, | |
| "grad_norm": 0.5525433421134949, | |
| "learning_rate": 6.419998639858538e-07, | |
| "loss": 0.5601, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 4.744630071599046, | |
| "grad_norm": 0.5674152374267578, | |
| "learning_rate": 5.897176883224442e-07, | |
| "loss": 0.6594, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 4.75417661097852, | |
| "grad_norm": 0.5498500466346741, | |
| "learning_rate": 5.396431094181198e-07, | |
| "loss": 0.616, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 4.763723150357995, | |
| "grad_norm": 0.5305196046829224, | |
| "learning_rate": 4.917783645496888e-07, | |
| "loss": 0.4995, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 4.77326968973747, | |
| "grad_norm": 0.5790063738822937, | |
| "learning_rate": 4.461255922609986e-07, | |
| "loss": 0.5518, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.782816229116945, | |
| "grad_norm": 0.548302412033081, | |
| "learning_rate": 4.0268683226741265e-07, | |
| "loss": 0.6202, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 4.79236276849642, | |
| "grad_norm": 0.548893928527832, | |
| "learning_rate": 3.6146402536468283e-07, | |
| "loss": 0.6218, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 4.801909307875895, | |
| "grad_norm": 0.5890070199966431, | |
| "learning_rate": 3.2245901334221895e-07, | |
| "loss": 0.6638, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 4.81145584725537, | |
| "grad_norm": 0.4871584475040436, | |
| "learning_rate": 2.856735389008269e-07, | |
| "loss": 0.6107, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 4.821002386634845, | |
| "grad_norm": 0.5432624816894531, | |
| "learning_rate": 2.511092455747932e-07, | |
| "loss": 0.583, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 4.83054892601432, | |
| "grad_norm": 0.5359986424446106, | |
| "learning_rate": 2.1876767765853234e-07, | |
| "loss": 0.5368, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 4.840095465393794, | |
| "grad_norm": 0.5359886288642883, | |
| "learning_rate": 1.8865028013751452e-07, | |
| "loss": 0.6259, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 4.84964200477327, | |
| "grad_norm": 0.5111921429634094, | |
| "learning_rate": 1.6075839862374488e-07, | |
| "loss": 0.5609, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 4.859188544152745, | |
| "grad_norm": 0.6437258124351501, | |
| "learning_rate": 1.3509327929563942e-07, | |
| "loss": 0.6395, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 4.868735083532219, | |
| "grad_norm": 0.5992398262023926, | |
| "learning_rate": 1.1165606884234181e-07, | |
| "loss": 0.6546, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.878281622911695, | |
| "grad_norm": 0.5831811428070068, | |
| "learning_rate": 9.044781441249207e-08, | |
| "loss": 0.609, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 4.88782816229117, | |
| "grad_norm": 0.5561614632606506, | |
| "learning_rate": 7.146946356743067e-08, | |
| "loss": 0.6699, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 4.897374701670644, | |
| "grad_norm": 0.5337750911712646, | |
| "learning_rate": 5.472186423889358e-08, | |
| "loss": 0.5614, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 4.906921241050119, | |
| "grad_norm": 0.41179969906806946, | |
| "learning_rate": 4.020576469108139e-08, | |
| "loss": 0.4451, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 4.916467780429595, | |
| "grad_norm": 0.4820442795753479, | |
| "learning_rate": 2.792181348726941e-08, | |
| "loss": 0.5897, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 4.926014319809069, | |
| "grad_norm": 0.5927594900131226, | |
| "learning_rate": 1.7870559460814173e-08, | |
| "loss": 0.6788, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 4.935560859188544, | |
| "grad_norm": 0.5302107334136963, | |
| "learning_rate": 1.0052451690617527e-08, | |
| "loss": 0.6105, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 4.945107398568019, | |
| "grad_norm": 0.5596168041229248, | |
| "learning_rate": 4.46783948109819e-09, | |
| "loss": 0.5537, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 4.954653937947494, | |
| "grad_norm": 0.5655501484870911, | |
| "learning_rate": 1.1169723465487281e-09, | |
| "loss": 0.6915, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 4.964200477326969, | |
| "grad_norm": 0.5537259578704834, | |
| "learning_rate": 0.0, | |
| "loss": 0.7157, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.964200477326969, | |
| "eval_loss": 0.7753176689147949, | |
| "eval_runtime": 12.9676, | |
| "eval_samples_per_second": 13.649, | |
| "eval_steps_per_second": 1.774, | |
| "step": 520 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 520, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 20, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.0257258893869056e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |