prxy5607's picture
Training in progress, step 200, checkpoint
e9851f5 verified
{
"best_metric": 10.3406343460083,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 1.4035087719298245,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.007017543859649123,
"grad_norm": 0.058438096195459366,
"learning_rate": 1e-05,
"loss": 10.3801,
"step": 1
},
{
"epoch": 0.007017543859649123,
"eval_loss": 10.380678176879883,
"eval_runtime": 0.5556,
"eval_samples_per_second": 431.928,
"eval_steps_per_second": 107.982,
"step": 1
},
{
"epoch": 0.014035087719298246,
"grad_norm": 0.06063023954629898,
"learning_rate": 2e-05,
"loss": 10.3792,
"step": 2
},
{
"epoch": 0.021052631578947368,
"grad_norm": 0.06841549277305603,
"learning_rate": 3e-05,
"loss": 10.3804,
"step": 3
},
{
"epoch": 0.028070175438596492,
"grad_norm": 0.0681336522102356,
"learning_rate": 4e-05,
"loss": 10.3805,
"step": 4
},
{
"epoch": 0.03508771929824561,
"grad_norm": 0.06004761531949043,
"learning_rate": 5e-05,
"loss": 10.3808,
"step": 5
},
{
"epoch": 0.042105263157894736,
"grad_norm": 0.061728086322546005,
"learning_rate": 6e-05,
"loss": 10.3788,
"step": 6
},
{
"epoch": 0.04912280701754386,
"grad_norm": 0.06197639927268028,
"learning_rate": 7e-05,
"loss": 10.381,
"step": 7
},
{
"epoch": 0.056140350877192984,
"grad_norm": 0.055631447583436966,
"learning_rate": 8e-05,
"loss": 10.3793,
"step": 8
},
{
"epoch": 0.06315789473684211,
"grad_norm": 0.06178618595004082,
"learning_rate": 9e-05,
"loss": 10.3788,
"step": 9
},
{
"epoch": 0.07017543859649122,
"grad_norm": 0.06878446787595749,
"learning_rate": 0.0001,
"loss": 10.3803,
"step": 10
},
{
"epoch": 0.07719298245614035,
"grad_norm": 0.07208960503339767,
"learning_rate": 9.999316524962345e-05,
"loss": 10.3764,
"step": 11
},
{
"epoch": 0.08421052631578947,
"grad_norm": 0.0600082203745842,
"learning_rate": 9.997266286704631e-05,
"loss": 10.3814,
"step": 12
},
{
"epoch": 0.0912280701754386,
"grad_norm": 0.05622672289609909,
"learning_rate": 9.993849845741524e-05,
"loss": 10.3759,
"step": 13
},
{
"epoch": 0.09824561403508772,
"grad_norm": 0.061736516654491425,
"learning_rate": 9.989068136093873e-05,
"loss": 10.3774,
"step": 14
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.07511258125305176,
"learning_rate": 9.98292246503335e-05,
"loss": 10.3772,
"step": 15
},
{
"epoch": 0.11228070175438597,
"grad_norm": 0.0732622891664505,
"learning_rate": 9.975414512725057e-05,
"loss": 10.3779,
"step": 16
},
{
"epoch": 0.11929824561403508,
"grad_norm": 0.08091577887535095,
"learning_rate": 9.966546331768191e-05,
"loss": 10.3772,
"step": 17
},
{
"epoch": 0.12631578947368421,
"grad_norm": 0.08580437302589417,
"learning_rate": 9.956320346634876e-05,
"loss": 10.3753,
"step": 18
},
{
"epoch": 0.13333333333333333,
"grad_norm": 0.08107656985521317,
"learning_rate": 9.944739353007344e-05,
"loss": 10.3727,
"step": 19
},
{
"epoch": 0.14035087719298245,
"grad_norm": 0.0906912162899971,
"learning_rate": 9.931806517013612e-05,
"loss": 10.3738,
"step": 20
},
{
"epoch": 0.14736842105263157,
"grad_norm": 0.0931132361292839,
"learning_rate": 9.917525374361912e-05,
"loss": 10.3734,
"step": 21
},
{
"epoch": 0.1543859649122807,
"grad_norm": 0.1076606959104538,
"learning_rate": 9.901899829374047e-05,
"loss": 10.3719,
"step": 22
},
{
"epoch": 0.16140350877192983,
"grad_norm": 0.12018569558858871,
"learning_rate": 9.884934153917997e-05,
"loss": 10.3713,
"step": 23
},
{
"epoch": 0.16842105263157894,
"grad_norm": 0.10436299443244934,
"learning_rate": 9.86663298624003e-05,
"loss": 10.373,
"step": 24
},
{
"epoch": 0.17543859649122806,
"grad_norm": 0.11032664775848389,
"learning_rate": 9.847001329696653e-05,
"loss": 10.3712,
"step": 25
},
{
"epoch": 0.1824561403508772,
"grad_norm": 0.11277417093515396,
"learning_rate": 9.826044551386744e-05,
"loss": 10.3753,
"step": 26
},
{
"epoch": 0.18947368421052632,
"grad_norm": 0.15820494294166565,
"learning_rate": 9.803768380684242e-05,
"loss": 10.3676,
"step": 27
},
{
"epoch": 0.19649122807017544,
"grad_norm": 0.14549218118190765,
"learning_rate": 9.780178907671789e-05,
"loss": 10.3708,
"step": 28
},
{
"epoch": 0.20350877192982456,
"grad_norm": 0.17387129366397858,
"learning_rate": 9.755282581475769e-05,
"loss": 10.3713,
"step": 29
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.15915431082248688,
"learning_rate": 9.729086208503174e-05,
"loss": 10.3647,
"step": 30
},
{
"epoch": 0.21754385964912282,
"grad_norm": 0.18707400560379028,
"learning_rate": 9.701596950580806e-05,
"loss": 10.3629,
"step": 31
},
{
"epoch": 0.22456140350877193,
"grad_norm": 0.1707993447780609,
"learning_rate": 9.672822322997305e-05,
"loss": 10.3639,
"step": 32
},
{
"epoch": 0.23157894736842105,
"grad_norm": 0.20038415491580963,
"learning_rate": 9.642770192448536e-05,
"loss": 10.3629,
"step": 33
},
{
"epoch": 0.23859649122807017,
"grad_norm": 0.21611055731773376,
"learning_rate": 9.611448774886924e-05,
"loss": 10.3656,
"step": 34
},
{
"epoch": 0.24561403508771928,
"grad_norm": 0.30490511655807495,
"learning_rate": 9.578866633275288e-05,
"loss": 10.3556,
"step": 35
},
{
"epoch": 0.25263157894736843,
"grad_norm": 0.12674015760421753,
"learning_rate": 9.545032675245813e-05,
"loss": 10.3663,
"step": 36
},
{
"epoch": 0.2596491228070175,
"grad_norm": 0.14866743981838226,
"learning_rate": 9.509956150664796e-05,
"loss": 10.3612,
"step": 37
},
{
"epoch": 0.26666666666666666,
"grad_norm": 0.15407007932662964,
"learning_rate": 9.473646649103818e-05,
"loss": 10.3618,
"step": 38
},
{
"epoch": 0.2736842105263158,
"grad_norm": 0.14178234338760376,
"learning_rate": 9.43611409721806e-05,
"loss": 10.3603,
"step": 39
},
{
"epoch": 0.2807017543859649,
"grad_norm": 0.14356359839439392,
"learning_rate": 9.397368756032445e-05,
"loss": 10.3616,
"step": 40
},
{
"epoch": 0.28771929824561404,
"grad_norm": 0.13513581454753876,
"learning_rate": 9.357421218136386e-05,
"loss": 10.3578,
"step": 41
},
{
"epoch": 0.29473684210526313,
"grad_norm": 0.1360790729522705,
"learning_rate": 9.316282404787871e-05,
"loss": 10.3565,
"step": 42
},
{
"epoch": 0.3017543859649123,
"grad_norm": 0.13210955262184143,
"learning_rate": 9.273963562927695e-05,
"loss": 10.3586,
"step": 43
},
{
"epoch": 0.3087719298245614,
"grad_norm": 0.1250680387020111,
"learning_rate": 9.230476262104677e-05,
"loss": 10.3573,
"step": 44
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.10175206512212753,
"learning_rate": 9.185832391312644e-05,
"loss": 10.3552,
"step": 45
},
{
"epoch": 0.32280701754385965,
"grad_norm": 0.12674783170223236,
"learning_rate": 9.140044155740101e-05,
"loss": 10.3553,
"step": 46
},
{
"epoch": 0.3298245614035088,
"grad_norm": 0.14150883257389069,
"learning_rate": 9.093124073433463e-05,
"loss": 10.3538,
"step": 47
},
{
"epoch": 0.3368421052631579,
"grad_norm": 0.09747511148452759,
"learning_rate": 9.045084971874738e-05,
"loss": 10.3481,
"step": 48
},
{
"epoch": 0.34385964912280703,
"grad_norm": 0.11724815517663956,
"learning_rate": 8.995939984474624e-05,
"loss": 10.3541,
"step": 49
},
{
"epoch": 0.3508771929824561,
"grad_norm": 0.1142168715596199,
"learning_rate": 8.945702546981969e-05,
"loss": 10.3529,
"step": 50
},
{
"epoch": 0.3508771929824561,
"eval_loss": 10.349671363830566,
"eval_runtime": 0.55,
"eval_samples_per_second": 436.342,
"eval_steps_per_second": 109.086,
"step": 50
},
{
"epoch": 0.35789473684210527,
"grad_norm": 0.13000968098640442,
"learning_rate": 8.894386393810563e-05,
"loss": 10.35,
"step": 51
},
{
"epoch": 0.3649122807017544,
"grad_norm": 0.11280800402164459,
"learning_rate": 8.842005554284296e-05,
"loss": 10.3477,
"step": 52
},
{
"epoch": 0.3719298245614035,
"grad_norm": 0.10914858430624008,
"learning_rate": 8.788574348801675e-05,
"loss": 10.3519,
"step": 53
},
{
"epoch": 0.37894736842105264,
"grad_norm": 0.11431189626455307,
"learning_rate": 8.73410738492077e-05,
"loss": 10.3505,
"step": 54
},
{
"epoch": 0.38596491228070173,
"grad_norm": 0.12891604006290436,
"learning_rate": 8.678619553365659e-05,
"loss": 10.3476,
"step": 55
},
{
"epoch": 0.3929824561403509,
"grad_norm": 0.1287131905555725,
"learning_rate": 8.622126023955446e-05,
"loss": 10.3501,
"step": 56
},
{
"epoch": 0.4,
"grad_norm": 0.11735491454601288,
"learning_rate": 8.564642241456986e-05,
"loss": 10.3542,
"step": 57
},
{
"epoch": 0.4070175438596491,
"grad_norm": 0.11743725091218948,
"learning_rate": 8.506183921362443e-05,
"loss": 10.3495,
"step": 58
},
{
"epoch": 0.41403508771929826,
"grad_norm": 0.10528979450464249,
"learning_rate": 8.44676704559283e-05,
"loss": 10.3494,
"step": 59
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.11193346232175827,
"learning_rate": 8.386407858128706e-05,
"loss": 10.3488,
"step": 60
},
{
"epoch": 0.4280701754385965,
"grad_norm": 0.1297432780265808,
"learning_rate": 8.32512286056924e-05,
"loss": 10.3467,
"step": 61
},
{
"epoch": 0.43508771929824563,
"grad_norm": 0.09085945785045624,
"learning_rate": 8.262928807620843e-05,
"loss": 10.3503,
"step": 62
},
{
"epoch": 0.4421052631578947,
"grad_norm": 0.10125606507062912,
"learning_rate": 8.199842702516583e-05,
"loss": 10.3498,
"step": 63
},
{
"epoch": 0.44912280701754387,
"grad_norm": 0.08393903076648712,
"learning_rate": 8.135881792367686e-05,
"loss": 10.3449,
"step": 64
},
{
"epoch": 0.45614035087719296,
"grad_norm": 0.09036874771118164,
"learning_rate": 8.07106356344834e-05,
"loss": 10.3421,
"step": 65
},
{
"epoch": 0.4631578947368421,
"grad_norm": 0.07801595330238342,
"learning_rate": 8.005405736415126e-05,
"loss": 10.3427,
"step": 66
},
{
"epoch": 0.47017543859649125,
"grad_norm": 0.0749220997095108,
"learning_rate": 7.938926261462366e-05,
"loss": 10.3468,
"step": 67
},
{
"epoch": 0.47719298245614034,
"grad_norm": 0.09423329681158066,
"learning_rate": 7.871643313414718e-05,
"loss": 10.347,
"step": 68
},
{
"epoch": 0.4842105263157895,
"grad_norm": 0.11977524310350418,
"learning_rate": 7.803575286758364e-05,
"loss": 10.3468,
"step": 69
},
{
"epoch": 0.49122807017543857,
"grad_norm": 0.12595734000205994,
"learning_rate": 7.734740790612136e-05,
"loss": 10.3427,
"step": 70
},
{
"epoch": 0.4982456140350877,
"grad_norm": 0.13295474648475647,
"learning_rate": 7.66515864363997e-05,
"loss": 10.3554,
"step": 71
},
{
"epoch": 0.5052631578947369,
"grad_norm": 0.1230146735906601,
"learning_rate": 7.594847868906076e-05,
"loss": 10.3497,
"step": 72
},
{
"epoch": 0.512280701754386,
"grad_norm": 0.11127111315727234,
"learning_rate": 7.52382768867422e-05,
"loss": 10.3591,
"step": 73
},
{
"epoch": 0.519298245614035,
"grad_norm": 0.11810820549726486,
"learning_rate": 7.452117519152542e-05,
"loss": 10.3515,
"step": 74
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.12868225574493408,
"learning_rate": 7.379736965185368e-05,
"loss": 10.3481,
"step": 75
},
{
"epoch": 0.5333333333333333,
"grad_norm": 0.06768758594989777,
"learning_rate": 7.30670581489344e-05,
"loss": 10.3518,
"step": 76
},
{
"epoch": 0.5403508771929825,
"grad_norm": 0.0636783093214035,
"learning_rate": 7.233044034264034e-05,
"loss": 10.351,
"step": 77
},
{
"epoch": 0.5473684210526316,
"grad_norm": 0.08322663605213165,
"learning_rate": 7.158771761692464e-05,
"loss": 10.3484,
"step": 78
},
{
"epoch": 0.5543859649122806,
"grad_norm": 0.10304541140794754,
"learning_rate": 7.083909302476453e-05,
"loss": 10.3504,
"step": 79
},
{
"epoch": 0.5614035087719298,
"grad_norm": 0.0707196518778801,
"learning_rate": 7.008477123264848e-05,
"loss": 10.3456,
"step": 80
},
{
"epoch": 0.5684210526315789,
"grad_norm": 0.06259272247552872,
"learning_rate": 6.932495846462261e-05,
"loss": 10.3492,
"step": 81
},
{
"epoch": 0.5754385964912281,
"grad_norm": 0.10322371125221252,
"learning_rate": 6.855986244591104e-05,
"loss": 10.3414,
"step": 82
},
{
"epoch": 0.5824561403508772,
"grad_norm": 0.06719489395618439,
"learning_rate": 6.778969234612584e-05,
"loss": 10.3529,
"step": 83
},
{
"epoch": 0.5894736842105263,
"grad_norm": 0.06424535065889359,
"learning_rate": 6.701465872208216e-05,
"loss": 10.3474,
"step": 84
},
{
"epoch": 0.5964912280701754,
"grad_norm": 0.07862396538257599,
"learning_rate": 6.623497346023418e-05,
"loss": 10.3485,
"step": 85
},
{
"epoch": 0.6035087719298246,
"grad_norm": 0.08303350955247879,
"learning_rate": 6.545084971874738e-05,
"loss": 10.3498,
"step": 86
},
{
"epoch": 0.6105263157894737,
"grad_norm": 0.05822298675775528,
"learning_rate": 6.466250186922325e-05,
"loss": 10.3519,
"step": 87
},
{
"epoch": 0.6175438596491228,
"grad_norm": 0.0716489925980568,
"learning_rate": 6.387014543809223e-05,
"loss": 10.3456,
"step": 88
},
{
"epoch": 0.624561403508772,
"grad_norm": 0.07383385300636292,
"learning_rate": 6.307399704769099e-05,
"loss": 10.35,
"step": 89
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.03729870170354843,
"learning_rate": 6.227427435703997e-05,
"loss": 10.3457,
"step": 90
},
{
"epoch": 0.6385964912280702,
"grad_norm": 0.0629424974322319,
"learning_rate": 6.147119600233758e-05,
"loss": 10.3491,
"step": 91
},
{
"epoch": 0.6456140350877193,
"grad_norm": 0.061789434403181076,
"learning_rate": 6.066498153718735e-05,
"loss": 10.3454,
"step": 92
},
{
"epoch": 0.6526315789473685,
"grad_norm": 0.06357581913471222,
"learning_rate": 5.985585137257401e-05,
"loss": 10.3463,
"step": 93
},
{
"epoch": 0.6596491228070176,
"grad_norm": 0.04997308924794197,
"learning_rate": 5.90440267166055e-05,
"loss": 10.3471,
"step": 94
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.09659120440483093,
"learning_rate": 5.8229729514036705e-05,
"loss": 10.3509,
"step": 95
},
{
"epoch": 0.6736842105263158,
"grad_norm": 0.07209810614585876,
"learning_rate": 5.74131823855921e-05,
"loss": 10.3398,
"step": 96
},
{
"epoch": 0.6807017543859649,
"grad_norm": 0.07157547026872635,
"learning_rate": 5.6594608567103456e-05,
"loss": 10.3457,
"step": 97
},
{
"epoch": 0.6877192982456141,
"grad_norm": 0.06521572172641754,
"learning_rate": 5.577423184847932e-05,
"loss": 10.347,
"step": 98
},
{
"epoch": 0.6947368421052632,
"grad_norm": 0.0904020220041275,
"learning_rate": 5.495227651252315e-05,
"loss": 10.3432,
"step": 99
},
{
"epoch": 0.7017543859649122,
"grad_norm": 0.1014203205704689,
"learning_rate": 5.4128967273616625e-05,
"loss": 10.3452,
"step": 100
},
{
"epoch": 0.7017543859649122,
"eval_loss": 10.344407081604004,
"eval_runtime": 0.5389,
"eval_samples_per_second": 445.323,
"eval_steps_per_second": 111.331,
"step": 100
},
{
"epoch": 0.7087719298245614,
"grad_norm": 0.09822150319814682,
"learning_rate": 5.330452921628497e-05,
"loss": 10.3361,
"step": 101
},
{
"epoch": 0.7157894736842105,
"grad_norm": 0.11128773540258408,
"learning_rate": 5.247918773366112e-05,
"loss": 10.3422,
"step": 102
},
{
"epoch": 0.7228070175438597,
"grad_norm": 0.09118737280368805,
"learning_rate": 5.165316846586541e-05,
"loss": 10.3438,
"step": 103
},
{
"epoch": 0.7298245614035088,
"grad_norm": 0.0816028043627739,
"learning_rate": 5.0826697238317935e-05,
"loss": 10.3466,
"step": 104
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.1651792824268341,
"learning_rate": 5e-05,
"loss": 10.337,
"step": 105
},
{
"epoch": 0.743859649122807,
"grad_norm": 0.15269403159618378,
"learning_rate": 4.917330276168208e-05,
"loss": 10.354,
"step": 106
},
{
"epoch": 0.7508771929824561,
"grad_norm": 0.12254855781793594,
"learning_rate": 4.834683153413459e-05,
"loss": 10.3473,
"step": 107
},
{
"epoch": 0.7578947368421053,
"grad_norm": 0.08288544416427612,
"learning_rate": 4.7520812266338885e-05,
"loss": 10.3517,
"step": 108
},
{
"epoch": 0.7649122807017544,
"grad_norm": 0.09320428222417831,
"learning_rate": 4.669547078371504e-05,
"loss": 10.3507,
"step": 109
},
{
"epoch": 0.7719298245614035,
"grad_norm": 0.09738554060459137,
"learning_rate": 4.5871032726383386e-05,
"loss": 10.3499,
"step": 110
},
{
"epoch": 0.7789473684210526,
"grad_norm": 0.0636056587100029,
"learning_rate": 4.504772348747687e-05,
"loss": 10.3465,
"step": 111
},
{
"epoch": 0.7859649122807018,
"grad_norm": 0.07954997569322586,
"learning_rate": 4.4225768151520694e-05,
"loss": 10.349,
"step": 112
},
{
"epoch": 0.7929824561403509,
"grad_norm": 0.09548051655292511,
"learning_rate": 4.3405391432896555e-05,
"loss": 10.3488,
"step": 113
},
{
"epoch": 0.8,
"grad_norm": 0.06435037404298782,
"learning_rate": 4.2586817614407895e-05,
"loss": 10.3494,
"step": 114
},
{
"epoch": 0.8070175438596491,
"grad_norm": 0.0702732726931572,
"learning_rate": 4.17702704859633e-05,
"loss": 10.3458,
"step": 115
},
{
"epoch": 0.8140350877192982,
"grad_norm": 0.07422645390033722,
"learning_rate": 4.095597328339452e-05,
"loss": 10.3467,
"step": 116
},
{
"epoch": 0.8210526315789474,
"grad_norm": 0.042868562042713165,
"learning_rate": 4.0144148627425993e-05,
"loss": 10.3465,
"step": 117
},
{
"epoch": 0.8280701754385965,
"grad_norm": 0.07119426876306534,
"learning_rate": 3.933501846281267e-05,
"loss": 10.343,
"step": 118
},
{
"epoch": 0.8350877192982457,
"grad_norm": 0.044547829777002335,
"learning_rate": 3.852880399766243e-05,
"loss": 10.3475,
"step": 119
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.06999172270298004,
"learning_rate": 3.772572564296005e-05,
"loss": 10.346,
"step": 120
},
{
"epoch": 0.8491228070175438,
"grad_norm": 0.051015742123126984,
"learning_rate": 3.6926002952309016e-05,
"loss": 10.345,
"step": 121
},
{
"epoch": 0.856140350877193,
"grad_norm": 0.06585199385881424,
"learning_rate": 3.612985456190778e-05,
"loss": 10.3462,
"step": 122
},
{
"epoch": 0.8631578947368421,
"grad_norm": 0.05538675934076309,
"learning_rate": 3.533749813077677e-05,
"loss": 10.347,
"step": 123
},
{
"epoch": 0.8701754385964913,
"grad_norm": 0.0970248132944107,
"learning_rate": 3.4549150281252636e-05,
"loss": 10.3447,
"step": 124
},
{
"epoch": 0.8771929824561403,
"grad_norm": 0.07111281156539917,
"learning_rate": 3.3765026539765834e-05,
"loss": 10.3453,
"step": 125
},
{
"epoch": 0.8842105263157894,
"grad_norm": 0.07990825921297073,
"learning_rate": 3.298534127791785e-05,
"loss": 10.3469,
"step": 126
},
{
"epoch": 0.8912280701754386,
"grad_norm": 0.059036485850811005,
"learning_rate": 3.221030765387417e-05,
"loss": 10.3409,
"step": 127
},
{
"epoch": 0.8982456140350877,
"grad_norm": 0.06820007413625717,
"learning_rate": 3.144013755408895e-05,
"loss": 10.3455,
"step": 128
},
{
"epoch": 0.9052631578947369,
"grad_norm": 0.08070363104343414,
"learning_rate": 3.0675041535377405e-05,
"loss": 10.3413,
"step": 129
},
{
"epoch": 0.9122807017543859,
"grad_norm": 0.08679619431495667,
"learning_rate": 2.991522876735154e-05,
"loss": 10.3451,
"step": 130
},
{
"epoch": 0.9192982456140351,
"grad_norm": 0.08083868026733398,
"learning_rate": 2.916090697523549e-05,
"loss": 10.3447,
"step": 131
},
{
"epoch": 0.9263157894736842,
"grad_norm": 0.07657700031995773,
"learning_rate": 2.8412282383075363e-05,
"loss": 10.3456,
"step": 132
},
{
"epoch": 0.9333333333333333,
"grad_norm": 0.08390800654888153,
"learning_rate": 2.766955965735968e-05,
"loss": 10.347,
"step": 133
},
{
"epoch": 0.9403508771929825,
"grad_norm": 0.07944929599761963,
"learning_rate": 2.693294185106562e-05,
"loss": 10.3427,
"step": 134
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.10431376844644547,
"learning_rate": 2.6202630348146324e-05,
"loss": 10.34,
"step": 135
},
{
"epoch": 0.9543859649122807,
"grad_norm": 0.08040665835142136,
"learning_rate": 2.547882480847461e-05,
"loss": 10.3384,
"step": 136
},
{
"epoch": 0.9614035087719298,
"grad_norm": 0.08806628733873367,
"learning_rate": 2.476172311325783e-05,
"loss": 10.3457,
"step": 137
},
{
"epoch": 0.968421052631579,
"grad_norm": 0.10316865891218185,
"learning_rate": 2.405152131093926e-05,
"loss": 10.3383,
"step": 138
},
{
"epoch": 0.9754385964912281,
"grad_norm": 0.15002769231796265,
"learning_rate": 2.3348413563600325e-05,
"loss": 10.3355,
"step": 139
},
{
"epoch": 0.9824561403508771,
"grad_norm": 0.17146770656108856,
"learning_rate": 2.2652592093878666e-05,
"loss": 10.3355,
"step": 140
},
{
"epoch": 0.9894736842105263,
"grad_norm": 0.09206395596265793,
"learning_rate": 2.196424713241637e-05,
"loss": 10.3488,
"step": 141
},
{
"epoch": 0.9964912280701754,
"grad_norm": 0.055690743029117584,
"learning_rate": 2.128356686585282e-05,
"loss": 10.3418,
"step": 142
},
{
"epoch": 1.0035087719298246,
"grad_norm": 0.08770684897899628,
"learning_rate": 2.061073738537635e-05,
"loss": 15.5021,
"step": 143
},
{
"epoch": 1.0105263157894737,
"grad_norm": 0.11512074619531631,
"learning_rate": 1.9945942635848748e-05,
"loss": 10.4473,
"step": 144
},
{
"epoch": 1.0175438596491229,
"grad_norm": 0.12913571298122406,
"learning_rate": 1.928936436551661e-05,
"loss": 10.5235,
"step": 145
},
{
"epoch": 1.024561403508772,
"grad_norm": 0.0952976867556572,
"learning_rate": 1.8641182076323148e-05,
"loss": 10.151,
"step": 146
},
{
"epoch": 1.0315789473684212,
"grad_norm": 0.11134811490774155,
"learning_rate": 1.800157297483417e-05,
"loss": 9.9626,
"step": 147
},
{
"epoch": 1.03859649122807,
"grad_norm": 0.08841732144355774,
"learning_rate": 1.7370711923791567e-05,
"loss": 10.7647,
"step": 148
},
{
"epoch": 1.0456140350877192,
"grad_norm": 0.09394381940364838,
"learning_rate": 1.6748771394307585e-05,
"loss": 10.2504,
"step": 149
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.09958937764167786,
"learning_rate": 1.6135921418712956e-05,
"loss": 10.3807,
"step": 150
},
{
"epoch": 1.0526315789473684,
"eval_loss": 10.341802597045898,
"eval_runtime": 0.5476,
"eval_samples_per_second": 438.287,
"eval_steps_per_second": 109.572,
"step": 150
},
{
"epoch": 1.0596491228070175,
"grad_norm": 0.08683663606643677,
"learning_rate": 1.553232954407171e-05,
"loss": 10.8776,
"step": 151
},
{
"epoch": 1.0666666666666667,
"grad_norm": 0.05788689851760864,
"learning_rate": 1.4938160786375572e-05,
"loss": 9.4835,
"step": 152
},
{
"epoch": 1.0736842105263158,
"grad_norm": 0.09865240007638931,
"learning_rate": 1.435357758543015e-05,
"loss": 11.0772,
"step": 153
},
{
"epoch": 1.080701754385965,
"grad_norm": 0.08386930823326111,
"learning_rate": 1.3778739760445552e-05,
"loss": 9.8618,
"step": 154
},
{
"epoch": 1.087719298245614,
"grad_norm": 0.06436026841402054,
"learning_rate": 1.3213804466343421e-05,
"loss": 10.1651,
"step": 155
},
{
"epoch": 1.0947368421052632,
"grad_norm": 0.06652949750423431,
"learning_rate": 1.2658926150792322e-05,
"loss": 10.8007,
"step": 156
},
{
"epoch": 1.1017543859649124,
"grad_norm": 0.07348314672708511,
"learning_rate": 1.2114256511983274e-05,
"loss": 10.2204,
"step": 157
},
{
"epoch": 1.1087719298245613,
"grad_norm": 0.090261310338974,
"learning_rate": 1.157994445715706e-05,
"loss": 10.6777,
"step": 158
},
{
"epoch": 1.1157894736842104,
"grad_norm": 0.06834334135055542,
"learning_rate": 1.1056136061894384e-05,
"loss": 9.7461,
"step": 159
},
{
"epoch": 1.1228070175438596,
"grad_norm": 0.06692725419998169,
"learning_rate": 1.0542974530180327e-05,
"loss": 10.4553,
"step": 160
},
{
"epoch": 1.1298245614035087,
"grad_norm": 0.08627898246049881,
"learning_rate": 1.0040600155253765e-05,
"loss": 10.2857,
"step": 161
},
{
"epoch": 1.1368421052631579,
"grad_norm": 0.08676744997501373,
"learning_rate": 9.549150281252633e-06,
"loss": 10.6293,
"step": 162
},
{
"epoch": 1.143859649122807,
"grad_norm": 0.0774228647351265,
"learning_rate": 9.068759265665384e-06,
"loss": 10.278,
"step": 163
},
{
"epoch": 1.1508771929824562,
"grad_norm": 0.07185292989015579,
"learning_rate": 8.599558442598998e-06,
"loss": 10.1883,
"step": 164
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.07643651217222214,
"learning_rate": 8.141676086873572e-06,
"loss": 10.4077,
"step": 165
},
{
"epoch": 1.1649122807017545,
"grad_norm": 0.08384363353252411,
"learning_rate": 7.695237378953223e-06,
"loss": 10.1208,
"step": 166
},
{
"epoch": 1.1719298245614036,
"grad_norm": 0.07069496810436249,
"learning_rate": 7.260364370723044e-06,
"loss": 10.6734,
"step": 167
},
{
"epoch": 1.1789473684210527,
"grad_norm": 0.1068500503897667,
"learning_rate": 6.837175952121306e-06,
"loss": 10.247,
"step": 168
},
{
"epoch": 1.1859649122807017,
"grad_norm": 0.07882976531982422,
"learning_rate": 6.425787818636131e-06,
"loss": 10.2507,
"step": 169
},
{
"epoch": 1.1929824561403508,
"grad_norm": 0.09044394642114639,
"learning_rate": 6.026312439675552e-06,
"loss": 10.3481,
"step": 170
},
{
"epoch": 1.2,
"grad_norm": 0.09455709159374237,
"learning_rate": 5.6388590278194096e-06,
"loss": 10.6524,
"step": 171
},
{
"epoch": 1.207017543859649,
"grad_norm": 0.09445133060216904,
"learning_rate": 5.263533508961827e-06,
"loss": 10.0846,
"step": 172
},
{
"epoch": 1.2140350877192982,
"grad_norm": 0.09570515155792236,
"learning_rate": 4.900438493352055e-06,
"loss": 10.3241,
"step": 173
},
{
"epoch": 1.2210526315789474,
"grad_norm": 0.09260547906160355,
"learning_rate": 4.549673247541875e-06,
"loss": 10.5036,
"step": 174
},
{
"epoch": 1.2280701754385965,
"grad_norm": 0.10068077594041824,
"learning_rate": 4.2113336672471245e-06,
"loss": 10.2396,
"step": 175
},
{
"epoch": 1.2350877192982457,
"grad_norm": 0.15771834552288055,
"learning_rate": 3.885512251130763e-06,
"loss": 10.6784,
"step": 176
},
{
"epoch": 1.2421052631578948,
"grad_norm": 0.1511031687259674,
"learning_rate": 3.5722980755146517e-06,
"loss": 10.6633,
"step": 177
},
{
"epoch": 1.2491228070175437,
"grad_norm": 0.11353182792663574,
"learning_rate": 3.271776770026963e-06,
"loss": 10.1151,
"step": 178
},
{
"epoch": 1.256140350877193,
"grad_norm": 0.14163772761821747,
"learning_rate": 2.9840304941919415e-06,
"loss": 9.8749,
"step": 179
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.10430266708135605,
"learning_rate": 2.7091379149682685e-06,
"loss": 10.7884,
"step": 180
},
{
"epoch": 1.2701754385964912,
"grad_norm": 0.11592936515808105,
"learning_rate": 2.4471741852423237e-06,
"loss": 10.0337,
"step": 181
},
{
"epoch": 1.2771929824561403,
"grad_norm": 0.11478496342897415,
"learning_rate": 2.1982109232821178e-06,
"loss": 10.4883,
"step": 182
},
{
"epoch": 1.2842105263157895,
"grad_norm": 0.0955367162823677,
"learning_rate": 1.962316193157593e-06,
"loss": 10.0631,
"step": 183
},
{
"epoch": 1.2912280701754386,
"grad_norm": 0.09962975978851318,
"learning_rate": 1.7395544861325718e-06,
"loss": 10.5269,
"step": 184
},
{
"epoch": 1.2982456140350878,
"grad_norm": 0.09933590888977051,
"learning_rate": 1.5299867030334814e-06,
"loss": 10.472,
"step": 185
},
{
"epoch": 1.305263157894737,
"grad_norm": 0.07964471727609634,
"learning_rate": 1.333670137599713e-06,
"loss": 10.1427,
"step": 186
},
{
"epoch": 1.312280701754386,
"grad_norm": 0.059783875942230225,
"learning_rate": 1.1506584608200367e-06,
"loss": 10.4737,
"step": 187
},
{
"epoch": 1.3192982456140352,
"grad_norm": 0.08105704933404922,
"learning_rate": 9.810017062595322e-07,
"loss": 10.2115,
"step": 188
},
{
"epoch": 1.3263157894736843,
"grad_norm": 0.07160281389951706,
"learning_rate": 8.247462563808817e-07,
"loss": 10.4014,
"step": 189
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.08081609010696411,
"learning_rate": 6.819348298638839e-07,
"loss": 10.251,
"step": 190
},
{
"epoch": 1.3403508771929824,
"grad_norm": 0.054519325494766235,
"learning_rate": 5.526064699265753e-07,
"loss": 10.1473,
"step": 191
},
{
"epoch": 1.3473684210526315,
"grad_norm": 0.08812150359153748,
"learning_rate": 4.367965336512403e-07,
"loss": 10.4325,
"step": 192
},
{
"epoch": 1.3543859649122807,
"grad_norm": 0.06724565476179123,
"learning_rate": 3.3453668231809286e-07,
"loss": 10.2351,
"step": 193
},
{
"epoch": 1.3614035087719298,
"grad_norm": 0.07397038489580154,
"learning_rate": 2.458548727494292e-07,
"loss": 10.4147,
"step": 194
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.08534782379865646,
"learning_rate": 1.7077534966650766e-07,
"loss": 10.6654,
"step": 195
},
{
"epoch": 1.3754385964912281,
"grad_norm": 0.08197546750307083,
"learning_rate": 1.0931863906127327e-07,
"loss": 10.0753,
"step": 196
},
{
"epoch": 1.3824561403508773,
"grad_norm": 0.08120612800121307,
"learning_rate": 6.150154258476315e-08,
"loss": 10.3052,
"step": 197
},
{
"epoch": 1.3894736842105262,
"grad_norm": 0.08832953125238419,
"learning_rate": 2.7337132953697554e-08,
"loss": 10.493,
"step": 198
},
{
"epoch": 1.3964912280701753,
"grad_norm": 0.07817672938108444,
"learning_rate": 6.834750376549792e-09,
"loss": 10.4036,
"step": 199
},
{
"epoch": 1.4035087719298245,
"grad_norm": 0.07051555067300797,
"learning_rate": 0.0,
"loss": 10.2162,
"step": 200
},
{
"epoch": 1.4035087719298245,
"eval_loss": 10.3406343460083,
"eval_runtime": 0.5479,
"eval_samples_per_second": 438.003,
"eval_steps_per_second": 109.501,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"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": 22895773089792.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}