| { | |
| "best_metric": 1.9330840110778809, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.45454545454545453, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0022727272727272726, | |
| "grad_norm": 3.007187604904175, | |
| "learning_rate": 1e-05, | |
| "loss": 9.3059, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0022727272727272726, | |
| "eval_loss": 2.4756627082824707, | |
| "eval_runtime": 15.1202, | |
| "eval_samples_per_second": 49.007, | |
| "eval_steps_per_second": 12.301, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004545454545454545, | |
| "grad_norm": 4.499298572540283, | |
| "learning_rate": 2e-05, | |
| "loss": 6.5353, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.006818181818181818, | |
| "grad_norm": 3.624962568283081, | |
| "learning_rate": 3e-05, | |
| "loss": 7.0474, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00909090909090909, | |
| "grad_norm": 4.367627143859863, | |
| "learning_rate": 4e-05, | |
| "loss": 8.2617, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.011363636363636364, | |
| "grad_norm": 4.702261447906494, | |
| "learning_rate": 5e-05, | |
| "loss": 7.4427, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.013636363636363636, | |
| "grad_norm": 3.6211180686950684, | |
| "learning_rate": 6e-05, | |
| "loss": 7.619, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.015909090909090907, | |
| "grad_norm": 4.267794609069824, | |
| "learning_rate": 7e-05, | |
| "loss": 8.5787, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01818181818181818, | |
| "grad_norm": 4.701915264129639, | |
| "learning_rate": 8e-05, | |
| "loss": 7.9755, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.020454545454545454, | |
| "grad_norm": 4.555323123931885, | |
| "learning_rate": 9e-05, | |
| "loss": 8.499, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.022727272727272728, | |
| "grad_norm": 5.117600440979004, | |
| "learning_rate": 0.0001, | |
| "loss": 8.4335, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 5.2592692375183105, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 8.5359, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02727272727272727, | |
| "grad_norm": 4.408563137054443, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 8.3489, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.029545454545454545, | |
| "grad_norm": 4.8282623291015625, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 8.0155, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.031818181818181815, | |
| "grad_norm": 4.181012153625488, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 8.8293, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03409090909090909, | |
| "grad_norm": 4.475191116333008, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 9.5334, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03636363636363636, | |
| "grad_norm": 4.718602657318115, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 8.0556, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.038636363636363635, | |
| "grad_norm": 4.524474620819092, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 7.7689, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04090909090909091, | |
| "grad_norm": 3.6022841930389404, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 9.0193, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04318181818181818, | |
| "grad_norm": 4.323879241943359, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 7.9922, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.045454545454545456, | |
| "grad_norm": 3.5238919258117676, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 8.7278, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04772727272727273, | |
| "grad_norm": 4.561350345611572, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 8.9544, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.8655569553375244, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 9.0573, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05227272727272727, | |
| "grad_norm": 4.048493385314941, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 9.6017, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.05454545454545454, | |
| "grad_norm": 4.086993217468262, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 8.9635, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.056818181818181816, | |
| "grad_norm": 4.17971658706665, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 8.7347, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05909090909090909, | |
| "grad_norm": 3.864226818084717, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 8.2089, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06136363636363636, | |
| "grad_norm": 4.086110591888428, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 8.3532, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06363636363636363, | |
| "grad_norm": 4.302659511566162, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 8.5754, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0659090909090909, | |
| "grad_norm": 4.012546539306641, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 8.7473, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.06818181818181818, | |
| "grad_norm": 4.200735092163086, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 9.2413, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07045454545454545, | |
| "grad_norm": 4.538753032684326, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 8.645, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07272727272727272, | |
| "grad_norm": 4.760058403015137, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 8.3604, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 5.103679656982422, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 8.0496, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.07727272727272727, | |
| "grad_norm": 4.99975061416626, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 7.7049, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.07954545454545454, | |
| "grad_norm": 5.385135173797607, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 7.8227, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08181818181818182, | |
| "grad_norm": 5.87006950378418, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 7.9425, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08409090909090909, | |
| "grad_norm": 6.1299028396606445, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 7.9917, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.08636363636363636, | |
| "grad_norm": 6.624438285827637, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 8.1363, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.08863636363636364, | |
| "grad_norm": 6.535163402557373, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 8.147, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 7.127198219299316, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 7.6708, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09318181818181819, | |
| "grad_norm": 8.905800819396973, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 9.4024, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.09545454545454546, | |
| "grad_norm": 10.360119819641113, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 8.3818, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.09772727272727273, | |
| "grad_norm": 10.646662712097168, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 6.6648, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 11.907293319702148, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 7.1416, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.10227272727272728, | |
| "grad_norm": 12.129688262939453, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 6.6378, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10454545454545454, | |
| "grad_norm": 11.963746070861816, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 7.3376, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.10681818181818181, | |
| "grad_norm": 14.229978561401367, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 7.0126, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.10909090909090909, | |
| "grad_norm": 13.639841079711914, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 8.1755, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.11136363636363636, | |
| "grad_norm": 25.13767433166504, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 10.0208, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.11363636363636363, | |
| "grad_norm": 46.8985710144043, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 13.1513, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11363636363636363, | |
| "eval_loss": 2.336198329925537, | |
| "eval_runtime": 14.8906, | |
| "eval_samples_per_second": 49.763, | |
| "eval_steps_per_second": 12.491, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1159090909090909, | |
| "grad_norm": 14.870892524719238, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 8.5908, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.11818181818181818, | |
| "grad_norm": 12.544276237487793, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 8.5961, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12045454545454545, | |
| "grad_norm": 9.471765518188477, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 7.833, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.12272727272727273, | |
| "grad_norm": 7.225938320159912, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 8.6043, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 4.557694435119629, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 7.3944, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12727272727272726, | |
| "grad_norm": 4.5267791748046875, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 8.2113, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.12954545454545455, | |
| "grad_norm": 3.35784912109375, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 8.3994, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.1318181818181818, | |
| "grad_norm": 3.500852584838867, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 8.8882, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1340909090909091, | |
| "grad_norm": 3.8013248443603516, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 9.0467, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.13636363636363635, | |
| "grad_norm": 3.442084789276123, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 7.7046, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13863636363636364, | |
| "grad_norm": 3.7370309829711914, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 7.4389, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1409090909090909, | |
| "grad_norm": 4.021424293518066, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 8.5011, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1431818181818182, | |
| "grad_norm": 3.0823864936828613, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 9.1405, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 3.261443614959717, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 8.3323, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.14772727272727273, | |
| "grad_norm": 2.704866647720337, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 8.4853, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 3.0572941303253174, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 8.1483, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.15227272727272728, | |
| "grad_norm": 3.489821672439575, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 7.6902, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.15454545454545454, | |
| "grad_norm": 2.7429654598236084, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 9.3882, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.15681818181818183, | |
| "grad_norm": 2.6050376892089844, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 9.12, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1590909090909091, | |
| "grad_norm": 2.8787131309509277, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 9.3259, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16136363636363638, | |
| "grad_norm": 3.618349552154541, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 9.0869, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.16363636363636364, | |
| "grad_norm": 2.98835825920105, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 8.2804, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.16590909090909092, | |
| "grad_norm": 2.9625604152679443, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 9.0852, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.16818181818181818, | |
| "grad_norm": 3.588578701019287, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 9.4061, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.17045454545454544, | |
| "grad_norm": 3.0199036598205566, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 8.5761, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17272727272727273, | |
| "grad_norm": 4.410862445831299, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 8.2581, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 5.893017768859863, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 8.4073, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.17727272727272728, | |
| "grad_norm": 3.6259286403656006, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 8.7203, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.17954545454545454, | |
| "grad_norm": 3.892850399017334, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 8.2902, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 4.450218200683594, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 8.1158, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18409090909090908, | |
| "grad_norm": 4.157801151275635, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 8.0612, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.18636363636363637, | |
| "grad_norm": 10.402759552001953, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 7.9552, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.18863636363636363, | |
| "grad_norm": 4.338803291320801, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 8.191, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.19090909090909092, | |
| "grad_norm": 5.273690700531006, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 8.3091, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.19318181818181818, | |
| "grad_norm": 4.366064548492432, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 9.5072, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19545454545454546, | |
| "grad_norm": 5.908369541168213, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 7.6029, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.19772727272727272, | |
| "grad_norm": 6.150649547576904, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 7.7976, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 5.759665012359619, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 8.311, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.20227272727272727, | |
| "grad_norm": 6.425780296325684, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 8.1277, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.20454545454545456, | |
| "grad_norm": 7.0809102058410645, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 8.1241, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20681818181818182, | |
| "grad_norm": 5.829425811767578, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 7.7421, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.20909090909090908, | |
| "grad_norm": 6.820650100708008, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 7.762, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.21136363636363636, | |
| "grad_norm": 9.303520202636719, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 6.3288, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.21363636363636362, | |
| "grad_norm": 35.969783782958984, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 8.2725, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2159090909090909, | |
| "grad_norm": 11.365795135498047, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 6.9468, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 8.189657211303711, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 7.6747, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.22045454545454546, | |
| "grad_norm": 12.022112846374512, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 6.8685, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.22272727272727272, | |
| "grad_norm": 12.496430397033691, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 7.6345, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 21.50300407409668, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 9.5732, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 36.77299118041992, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 11.5283, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "eval_loss": 2.0454964637756348, | |
| "eval_runtime": 14.884, | |
| "eval_samples_per_second": 49.785, | |
| "eval_steps_per_second": 12.497, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22954545454545455, | |
| "grad_norm": 4.069813251495361, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 8.8511, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.2318181818181818, | |
| "grad_norm": 4.448278903961182, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 8.2243, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2340909090909091, | |
| "grad_norm": 4.594381332397461, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 9.07, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.23636363636363636, | |
| "grad_norm": 5.057316303253174, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 6.588, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.23863636363636365, | |
| "grad_norm": 3.6976261138916016, | |
| "learning_rate": 5e-05, | |
| "loss": 8.4177, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2409090909090909, | |
| "grad_norm": 3.10335373878479, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 8.3845, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2431818181818182, | |
| "grad_norm": 2.9570393562316895, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 7.4729, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.24545454545454545, | |
| "grad_norm": 2.7596237659454346, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 8.6437, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.24772727272727274, | |
| "grad_norm": 3.071605920791626, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 8.0106, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.4575226306915283, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 7.0526, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25227272727272726, | |
| "grad_norm": 2.5973315238952637, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 7.9603, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2545454545454545, | |
| "grad_norm": 3.112067461013794, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 8.4722, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.25681818181818183, | |
| "grad_norm": 2.7353177070617676, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 7.8179, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2590909090909091, | |
| "grad_norm": 3.256028413772583, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 8.2511, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.26136363636363635, | |
| "grad_norm": 2.3543553352355957, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 8.7977, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2636363636363636, | |
| "grad_norm": 2.310845136642456, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 8.1625, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.26590909090909093, | |
| "grad_norm": 2.6301939487457275, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 8.0373, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2681818181818182, | |
| "grad_norm": 2.970533847808838, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 8.8501, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.27045454545454545, | |
| "grad_norm": 2.5005667209625244, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 8.1597, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 2.960244655609131, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 7.977, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 2.9278712272644043, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 8.7007, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2772727272727273, | |
| "grad_norm": 2.6727843284606934, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 9.5243, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.27954545454545454, | |
| "grad_norm": 2.4607348442077637, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 8.8664, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2818181818181818, | |
| "grad_norm": 2.757230043411255, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 8.7458, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.2840909090909091, | |
| "grad_norm": 2.8296964168548584, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 8.8684, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2863636363636364, | |
| "grad_norm": 2.813182830810547, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 8.3542, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.28863636363636364, | |
| "grad_norm": 3.2281596660614014, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 7.6691, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 3.085167646408081, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 7.918, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.29318181818181815, | |
| "grad_norm": 2.9861419200897217, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 8.2211, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.29545454545454547, | |
| "grad_norm": 3.203407049179077, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 8.1782, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29772727272727273, | |
| "grad_norm": 3.2800235748291016, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 8.2206, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 3.5517466068267822, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 7.7706, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.30227272727272725, | |
| "grad_norm": 3.4505789279937744, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 8.1443, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.30454545454545456, | |
| "grad_norm": 3.553616762161255, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 8.4562, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3068181818181818, | |
| "grad_norm": 3.695871591567993, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 7.4584, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3090909090909091, | |
| "grad_norm": 3.6847994327545166, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 7.4983, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.31136363636363634, | |
| "grad_norm": 3.95638108253479, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 7.4831, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.31363636363636366, | |
| "grad_norm": 4.275068283081055, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 8.7212, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3159090909090909, | |
| "grad_norm": 5.396293640136719, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 8.2101, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3181818181818182, | |
| "grad_norm": 5.945644855499268, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 8.7379, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.32045454545454544, | |
| "grad_norm": 5.538638591766357, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 7.6408, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.32272727272727275, | |
| "grad_norm": 5.8112664222717285, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 7.614, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 6.020136833190918, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 6.5472, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.32727272727272727, | |
| "grad_norm": 6.261348724365234, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 5.11, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.32954545454545453, | |
| "grad_norm": 7.140927314758301, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 6.003, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.33181818181818185, | |
| "grad_norm": 8.335502624511719, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 6.588, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3340909090909091, | |
| "grad_norm": 9.745265007019043, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 8.8445, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.33636363636363636, | |
| "grad_norm": 14.067954063415527, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 9.889, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3386363636363636, | |
| "grad_norm": 15.827913284301758, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 8.592, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3409090909090909, | |
| "grad_norm": 31.608592987060547, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 12.4333, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3409090909090909, | |
| "eval_loss": 1.9408866167068481, | |
| "eval_runtime": 14.8918, | |
| "eval_samples_per_second": 49.759, | |
| "eval_steps_per_second": 12.49, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3431818181818182, | |
| "grad_norm": 1.9436229467391968, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 6.3519, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.34545454545454546, | |
| "grad_norm": 1.5854361057281494, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 5.6877, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3477272727272727, | |
| "grad_norm": 2.804558515548706, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 6.9716, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.9725067615509033, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 5.9957, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.3522727272727273, | |
| "grad_norm": 2.6551668643951416, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 6.1764, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.35454545454545455, | |
| "grad_norm": 2.4639272689819336, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 8.0812, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3568181818181818, | |
| "grad_norm": 2.1735146045684814, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 8.0641, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.35909090909090907, | |
| "grad_norm": 2.647315740585327, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 8.0203, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3613636363636364, | |
| "grad_norm": 2.662860631942749, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 5.9229, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 3.6951229572296143, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 6.88, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3659090909090909, | |
| "grad_norm": 2.3631911277770996, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 9.4944, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.36818181818181817, | |
| "grad_norm": 2.3874800205230713, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 8.993, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3704545454545455, | |
| "grad_norm": 2.6109044551849365, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 8.17, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.37272727272727274, | |
| "grad_norm": 2.769998788833618, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 7.6741, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 2.9831223487854004, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 8.0147, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.37727272727272726, | |
| "grad_norm": 2.7223970890045166, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 7.1708, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3795454545454545, | |
| "grad_norm": 2.840115785598755, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 8.9037, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.38181818181818183, | |
| "grad_norm": 2.896212100982666, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 7.5585, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3840909090909091, | |
| "grad_norm": 2.6225814819335938, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 9.1249, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.38636363636363635, | |
| "grad_norm": 2.7717206478118896, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 7.6802, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3886363636363636, | |
| "grad_norm": 2.5629074573516846, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 8.4209, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.39090909090909093, | |
| "grad_norm": 2.909229278564453, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 8.2913, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3931818181818182, | |
| "grad_norm": 3.118156671524048, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 9.2818, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.39545454545454545, | |
| "grad_norm": 3.183821201324463, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 8.5311, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3977272727272727, | |
| "grad_norm": 2.814068078994751, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 8.9429, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.1080613136291504, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 8.2314, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.4022727272727273, | |
| "grad_norm": 3.2898058891296387, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 8.6311, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.40454545454545454, | |
| "grad_norm": 3.2309272289276123, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 8.6578, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.4068181818181818, | |
| "grad_norm": 3.076035737991333, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 9.1452, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4090909090909091, | |
| "grad_norm": 3.6452112197875977, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 7.5169, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4113636363636364, | |
| "grad_norm": 3.2491209506988525, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 8.4995, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.41363636363636364, | |
| "grad_norm": 4.010043144226074, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 7.771, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4159090909090909, | |
| "grad_norm": 3.741434097290039, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 8.4316, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.41818181818181815, | |
| "grad_norm": 3.5997982025146484, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 7.4769, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.42045454545454547, | |
| "grad_norm": 3.847963571548462, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 7.4228, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.42272727272727273, | |
| "grad_norm": 4.234038829803467, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 7.764, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 4.80214786529541, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 7.178, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.42727272727272725, | |
| "grad_norm": 4.395010948181152, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 7.8632, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.42954545454545456, | |
| "grad_norm": 4.980628490447998, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 7.6437, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4318181818181818, | |
| "grad_norm": 5.122274398803711, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 8.0314, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4340909090909091, | |
| "grad_norm": 5.353920936584473, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 8.9338, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 5.870506286621094, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 7.3561, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.43863636363636366, | |
| "grad_norm": 6.991383075714111, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 7.2168, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.4409090909090909, | |
| "grad_norm": 6.39097785949707, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 5.389, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.4431818181818182, | |
| "grad_norm": 7.921006202697754, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 7.6233, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.44545454545454544, | |
| "grad_norm": 9.876401901245117, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 6.2759, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.44772727272727275, | |
| "grad_norm": 10.19648551940918, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 8.0472, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 12.335653305053711, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 8.6216, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.45227272727272727, | |
| "grad_norm": 16.507097244262695, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 8.1291, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 32.99649429321289, | |
| "learning_rate": 0.0, | |
| "loss": 11.9441, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "eval_loss": 1.9330840110778809, | |
| "eval_runtime": 14.9468, | |
| "eval_samples_per_second": 49.576, | |
| "eval_steps_per_second": 12.444, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.055914886089933e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |