prxy5605's picture
Training in progress, step 200, checkpoint
98432dc verified
{
"best_metric": 5.247466564178467,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 0.5475701574264202,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0027378507871321013,
"grad_norm": 2869.560302734375,
"learning_rate": 1e-05,
"loss": 18.9407,
"step": 1
},
{
"epoch": 0.0027378507871321013,
"eval_loss": 5.184251308441162,
"eval_runtime": 2.1423,
"eval_samples_per_second": 287.075,
"eval_steps_per_second": 71.886,
"step": 1
},
{
"epoch": 0.0054757015742642025,
"grad_norm": 3299.45068359375,
"learning_rate": 2e-05,
"loss": 17.9501,
"step": 2
},
{
"epoch": 0.008213552361396304,
"grad_norm": 2877.494140625,
"learning_rate": 3e-05,
"loss": 18.3079,
"step": 3
},
{
"epoch": 0.010951403148528405,
"grad_norm": 3389.11767578125,
"learning_rate": 4e-05,
"loss": 19.2225,
"step": 4
},
{
"epoch": 0.013689253935660506,
"grad_norm": 4282.4208984375,
"learning_rate": 5e-05,
"loss": 19.2442,
"step": 5
},
{
"epoch": 0.01642710472279261,
"grad_norm": 3912.53466796875,
"learning_rate": 6e-05,
"loss": 18.8629,
"step": 6
},
{
"epoch": 0.019164955509924708,
"grad_norm": 4039.3466796875,
"learning_rate": 7e-05,
"loss": 18.9577,
"step": 7
},
{
"epoch": 0.02190280629705681,
"grad_norm": 5384.4765625,
"learning_rate": 8e-05,
"loss": 18.5148,
"step": 8
},
{
"epoch": 0.024640657084188913,
"grad_norm": 3950.795654296875,
"learning_rate": 9e-05,
"loss": 18.7446,
"step": 9
},
{
"epoch": 0.02737850787132101,
"grad_norm": 4083.84619140625,
"learning_rate": 0.0001,
"loss": 17.6605,
"step": 10
},
{
"epoch": 0.030116358658453114,
"grad_norm": 3851.057861328125,
"learning_rate": 9.999316524962345e-05,
"loss": 19.1159,
"step": 11
},
{
"epoch": 0.03285420944558522,
"grad_norm": 4973.96875,
"learning_rate": 9.997266286704631e-05,
"loss": 18.5552,
"step": 12
},
{
"epoch": 0.03559206023271732,
"grad_norm": 3997.625732421875,
"learning_rate": 9.993849845741524e-05,
"loss": 18.5181,
"step": 13
},
{
"epoch": 0.038329911019849415,
"grad_norm": 4821.0908203125,
"learning_rate": 9.989068136093873e-05,
"loss": 18.9077,
"step": 14
},
{
"epoch": 0.04106776180698152,
"grad_norm": 7009.69091796875,
"learning_rate": 9.98292246503335e-05,
"loss": 18.1293,
"step": 15
},
{
"epoch": 0.04380561259411362,
"grad_norm": 5173.1787109375,
"learning_rate": 9.975414512725057e-05,
"loss": 18.3298,
"step": 16
},
{
"epoch": 0.04654346338124572,
"grad_norm": 5275.22705078125,
"learning_rate": 9.966546331768191e-05,
"loss": 17.488,
"step": 17
},
{
"epoch": 0.049281314168377825,
"grad_norm": 6161.54345703125,
"learning_rate": 9.956320346634876e-05,
"loss": 18.125,
"step": 18
},
{
"epoch": 0.05201916495550993,
"grad_norm": 5380.9541015625,
"learning_rate": 9.944739353007344e-05,
"loss": 19.6235,
"step": 19
},
{
"epoch": 0.05475701574264202,
"grad_norm": 7076.9326171875,
"learning_rate": 9.931806517013612e-05,
"loss": 19.6551,
"step": 20
},
{
"epoch": 0.057494866529774126,
"grad_norm": 9411.6328125,
"learning_rate": 9.917525374361912e-05,
"loss": 19.1774,
"step": 21
},
{
"epoch": 0.06023271731690623,
"grad_norm": 5219.16845703125,
"learning_rate": 9.901899829374047e-05,
"loss": 17.0145,
"step": 22
},
{
"epoch": 0.06297056810403832,
"grad_norm": 5389.611328125,
"learning_rate": 9.884934153917997e-05,
"loss": 18.715,
"step": 23
},
{
"epoch": 0.06570841889117043,
"grad_norm": 4909.1533203125,
"learning_rate": 9.86663298624003e-05,
"loss": 18.0538,
"step": 24
},
{
"epoch": 0.06844626967830253,
"grad_norm": 4762.9892578125,
"learning_rate": 9.847001329696653e-05,
"loss": 18.122,
"step": 25
},
{
"epoch": 0.07118412046543464,
"grad_norm": 5046.6279296875,
"learning_rate": 9.826044551386744e-05,
"loss": 16.9833,
"step": 26
},
{
"epoch": 0.07392197125256673,
"grad_norm": 5887.14208984375,
"learning_rate": 9.803768380684242e-05,
"loss": 18.4829,
"step": 27
},
{
"epoch": 0.07665982203969883,
"grad_norm": 7543.8603515625,
"learning_rate": 9.780178907671789e-05,
"loss": 19.2867,
"step": 28
},
{
"epoch": 0.07939767282683094,
"grad_norm": 4809.97705078125,
"learning_rate": 9.755282581475769e-05,
"loss": 18.0179,
"step": 29
},
{
"epoch": 0.08213552361396304,
"grad_norm": 4606.07666015625,
"learning_rate": 9.729086208503174e-05,
"loss": 18.5569,
"step": 30
},
{
"epoch": 0.08487337440109514,
"grad_norm": 4569.7158203125,
"learning_rate": 9.701596950580806e-05,
"loss": 18.9656,
"step": 31
},
{
"epoch": 0.08761122518822724,
"grad_norm": 4704.12060546875,
"learning_rate": 9.672822322997305e-05,
"loss": 18.4432,
"step": 32
},
{
"epoch": 0.09034907597535935,
"grad_norm": 5654.9912109375,
"learning_rate": 9.642770192448536e-05,
"loss": 18.5611,
"step": 33
},
{
"epoch": 0.09308692676249145,
"grad_norm": 6667.42626953125,
"learning_rate": 9.611448774886924e-05,
"loss": 19.0092,
"step": 34
},
{
"epoch": 0.09582477754962354,
"grad_norm": 9212.4580078125,
"learning_rate": 9.578866633275288e-05,
"loss": 19.2947,
"step": 35
},
{
"epoch": 0.09856262833675565,
"grad_norm": 5465.796875,
"learning_rate": 9.545032675245813e-05,
"loss": 18.93,
"step": 36
},
{
"epoch": 0.10130047912388775,
"grad_norm": 7496.95654296875,
"learning_rate": 9.509956150664796e-05,
"loss": 21.2021,
"step": 37
},
{
"epoch": 0.10403832991101986,
"grad_norm": 8689.849609375,
"learning_rate": 9.473646649103818e-05,
"loss": 20.4252,
"step": 38
},
{
"epoch": 0.10677618069815195,
"grad_norm": 7595.74365234375,
"learning_rate": 9.43611409721806e-05,
"loss": 21.0276,
"step": 39
},
{
"epoch": 0.10951403148528405,
"grad_norm": 7750.20166015625,
"learning_rate": 9.397368756032445e-05,
"loss": 21.0263,
"step": 40
},
{
"epoch": 0.11225188227241616,
"grad_norm": 9827.46875,
"learning_rate": 9.357421218136386e-05,
"loss": 21.9128,
"step": 41
},
{
"epoch": 0.11498973305954825,
"grad_norm": 6157.37744140625,
"learning_rate": 9.316282404787871e-05,
"loss": 20.8637,
"step": 42
},
{
"epoch": 0.11772758384668036,
"grad_norm": 7301.2265625,
"learning_rate": 9.273963562927695e-05,
"loss": 21.7574,
"step": 43
},
{
"epoch": 0.12046543463381246,
"grad_norm": 7452.45654296875,
"learning_rate": 9.230476262104677e-05,
"loss": 21.4877,
"step": 44
},
{
"epoch": 0.12320328542094455,
"grad_norm": 6658.0244140625,
"learning_rate": 9.185832391312644e-05,
"loss": 21.5988,
"step": 45
},
{
"epoch": 0.12594113620807665,
"grad_norm": 9002.0439453125,
"learning_rate": 9.140044155740101e-05,
"loss": 21.8134,
"step": 46
},
{
"epoch": 0.12867898699520877,
"grad_norm": 8484.044921875,
"learning_rate": 9.093124073433463e-05,
"loss": 22.6028,
"step": 47
},
{
"epoch": 0.13141683778234087,
"grad_norm": 5393.41650390625,
"learning_rate": 9.045084971874738e-05,
"loss": 21.4441,
"step": 48
},
{
"epoch": 0.13415468856947296,
"grad_norm": 7081.9052734375,
"learning_rate": 8.995939984474624e-05,
"loss": 21.2952,
"step": 49
},
{
"epoch": 0.13689253935660506,
"grad_norm": 8458.6767578125,
"learning_rate": 8.945702546981969e-05,
"loss": 21.0256,
"step": 50
},
{
"epoch": 0.13689253935660506,
"eval_loss": 5.5685906410217285,
"eval_runtime": 2.1364,
"eval_samples_per_second": 287.871,
"eval_steps_per_second": 72.085,
"step": 50
},
{
"epoch": 0.13963039014373715,
"grad_norm": 3334.87939453125,
"learning_rate": 8.894386393810563e-05,
"loss": 23.6486,
"step": 51
},
{
"epoch": 0.14236824093086928,
"grad_norm": 2512.7841796875,
"learning_rate": 8.842005554284296e-05,
"loss": 22.699,
"step": 52
},
{
"epoch": 0.14510609171800137,
"grad_norm": 1850.7974853515625,
"learning_rate": 8.788574348801675e-05,
"loss": 21.8799,
"step": 53
},
{
"epoch": 0.14784394250513347,
"grad_norm": 3691.093505859375,
"learning_rate": 8.73410738492077e-05,
"loss": 21.4601,
"step": 54
},
{
"epoch": 0.15058179329226556,
"grad_norm": 2152.451904296875,
"learning_rate": 8.678619553365659e-05,
"loss": 21.5812,
"step": 55
},
{
"epoch": 0.15331964407939766,
"grad_norm": 3066.42529296875,
"learning_rate": 8.622126023955446e-05,
"loss": 22.5719,
"step": 56
},
{
"epoch": 0.15605749486652978,
"grad_norm": 2297.962158203125,
"learning_rate": 8.564642241456986e-05,
"loss": 21.2545,
"step": 57
},
{
"epoch": 0.15879534565366188,
"grad_norm": 2484.205810546875,
"learning_rate": 8.506183921362443e-05,
"loss": 22.9001,
"step": 58
},
{
"epoch": 0.16153319644079397,
"grad_norm": 3208.600830078125,
"learning_rate": 8.44676704559283e-05,
"loss": 21.5575,
"step": 59
},
{
"epoch": 0.16427104722792607,
"grad_norm": 3078.96630859375,
"learning_rate": 8.386407858128706e-05,
"loss": 21.7485,
"step": 60
},
{
"epoch": 0.16700889801505817,
"grad_norm": 3359.209228515625,
"learning_rate": 8.32512286056924e-05,
"loss": 21.8616,
"step": 61
},
{
"epoch": 0.1697467488021903,
"grad_norm": 2449.829833984375,
"learning_rate": 8.262928807620843e-05,
"loss": 22.6381,
"step": 62
},
{
"epoch": 0.17248459958932238,
"grad_norm": 4092.363037109375,
"learning_rate": 8.199842702516583e-05,
"loss": 22.4161,
"step": 63
},
{
"epoch": 0.17522245037645448,
"grad_norm": 3124.9306640625,
"learning_rate": 8.135881792367686e-05,
"loss": 22.3952,
"step": 64
},
{
"epoch": 0.17796030116358658,
"grad_norm": 2817.8486328125,
"learning_rate": 8.07106356344834e-05,
"loss": 22.8886,
"step": 65
},
{
"epoch": 0.1806981519507187,
"grad_norm": 2720.580078125,
"learning_rate": 8.005405736415126e-05,
"loss": 21.3383,
"step": 66
},
{
"epoch": 0.1834360027378508,
"grad_norm": 3549.72607421875,
"learning_rate": 7.938926261462366e-05,
"loss": 21.5768,
"step": 67
},
{
"epoch": 0.1861738535249829,
"grad_norm": 3672.636962890625,
"learning_rate": 7.871643313414718e-05,
"loss": 22.2083,
"step": 68
},
{
"epoch": 0.188911704312115,
"grad_norm": 2854.32177734375,
"learning_rate": 7.803575286758364e-05,
"loss": 22.5027,
"step": 69
},
{
"epoch": 0.19164955509924708,
"grad_norm": 3031.279052734375,
"learning_rate": 7.734740790612136e-05,
"loss": 22.2783,
"step": 70
},
{
"epoch": 0.1943874058863792,
"grad_norm": 3828.364501953125,
"learning_rate": 7.66515864363997e-05,
"loss": 22.2839,
"step": 71
},
{
"epoch": 0.1971252566735113,
"grad_norm": 4242.29443359375,
"learning_rate": 7.594847868906076e-05,
"loss": 22.0545,
"step": 72
},
{
"epoch": 0.1998631074606434,
"grad_norm": 3641.48291015625,
"learning_rate": 7.52382768867422e-05,
"loss": 22.1318,
"step": 73
},
{
"epoch": 0.2026009582477755,
"grad_norm": 3086.799560546875,
"learning_rate": 7.452117519152542e-05,
"loss": 22.5933,
"step": 74
},
{
"epoch": 0.2053388090349076,
"grad_norm": 6644.65478515625,
"learning_rate": 7.379736965185368e-05,
"loss": 21.4046,
"step": 75
},
{
"epoch": 0.2080766598220397,
"grad_norm": 4785.70703125,
"learning_rate": 7.30670581489344e-05,
"loss": 22.3585,
"step": 76
},
{
"epoch": 0.2108145106091718,
"grad_norm": 3184.20654296875,
"learning_rate": 7.233044034264034e-05,
"loss": 21.6925,
"step": 77
},
{
"epoch": 0.2135523613963039,
"grad_norm": 4529.197265625,
"learning_rate": 7.158771761692464e-05,
"loss": 21.0992,
"step": 78
},
{
"epoch": 0.216290212183436,
"grad_norm": 5615.53759765625,
"learning_rate": 7.083909302476453e-05,
"loss": 21.2176,
"step": 79
},
{
"epoch": 0.2190280629705681,
"grad_norm": 4465.0966796875,
"learning_rate": 7.008477123264848e-05,
"loss": 20.9185,
"step": 80
},
{
"epoch": 0.22176591375770022,
"grad_norm": 4487.5703125,
"learning_rate": 6.932495846462261e-05,
"loss": 21.5685,
"step": 81
},
{
"epoch": 0.2245037645448323,
"grad_norm": 3949.593994140625,
"learning_rate": 6.855986244591104e-05,
"loss": 21.399,
"step": 82
},
{
"epoch": 0.2272416153319644,
"grad_norm": 5499.59228515625,
"learning_rate": 6.778969234612584e-05,
"loss": 22.0713,
"step": 83
},
{
"epoch": 0.2299794661190965,
"grad_norm": 4665.52734375,
"learning_rate": 6.701465872208216e-05,
"loss": 21.8885,
"step": 84
},
{
"epoch": 0.2327173169062286,
"grad_norm": 3186.498779296875,
"learning_rate": 6.623497346023418e-05,
"loss": 21.87,
"step": 85
},
{
"epoch": 0.23545516769336072,
"grad_norm": 5759.07763671875,
"learning_rate": 6.545084971874738e-05,
"loss": 21.9995,
"step": 86
},
{
"epoch": 0.23819301848049282,
"grad_norm": 5038.76025390625,
"learning_rate": 6.466250186922325e-05,
"loss": 22.0786,
"step": 87
},
{
"epoch": 0.24093086926762491,
"grad_norm": 4387.96923828125,
"learning_rate": 6.387014543809223e-05,
"loss": 21.6851,
"step": 88
},
{
"epoch": 0.243668720054757,
"grad_norm": 2766.3203125,
"learning_rate": 6.307399704769099e-05,
"loss": 21.6898,
"step": 89
},
{
"epoch": 0.2464065708418891,
"grad_norm": 3970.920654296875,
"learning_rate": 6.227427435703997e-05,
"loss": 22.7555,
"step": 90
},
{
"epoch": 0.24914442162902123,
"grad_norm": 3838.981201171875,
"learning_rate": 6.147119600233758e-05,
"loss": 22.3365,
"step": 91
},
{
"epoch": 0.2518822724161533,
"grad_norm": 3907.515380859375,
"learning_rate": 6.066498153718735e-05,
"loss": 20.7592,
"step": 92
},
{
"epoch": 0.2546201232032854,
"grad_norm": 5557.63232421875,
"learning_rate": 5.985585137257401e-05,
"loss": 21.8049,
"step": 93
},
{
"epoch": 0.25735797399041754,
"grad_norm": 4137.37744140625,
"learning_rate": 5.90440267166055e-05,
"loss": 22.4302,
"step": 94
},
{
"epoch": 0.2600958247775496,
"grad_norm": 3966.4287109375,
"learning_rate": 5.8229729514036705e-05,
"loss": 21.5639,
"step": 95
},
{
"epoch": 0.26283367556468173,
"grad_norm": 3949.333984375,
"learning_rate": 5.74131823855921e-05,
"loss": 21.8001,
"step": 96
},
{
"epoch": 0.2655715263518138,
"grad_norm": 4312.3271484375,
"learning_rate": 5.6594608567103456e-05,
"loss": 21.6497,
"step": 97
},
{
"epoch": 0.2683093771389459,
"grad_norm": 4580.8916015625,
"learning_rate": 5.577423184847932e-05,
"loss": 22.6536,
"step": 98
},
{
"epoch": 0.27104722792607805,
"grad_norm": 6759.92724609375,
"learning_rate": 5.495227651252315e-05,
"loss": 22.1202,
"step": 99
},
{
"epoch": 0.2737850787132101,
"grad_norm": 7867.4287109375,
"learning_rate": 5.4128967273616625e-05,
"loss": 21.4744,
"step": 100
},
{
"epoch": 0.2737850787132101,
"eval_loss": 5.565637111663818,
"eval_runtime": 2.1311,
"eval_samples_per_second": 288.582,
"eval_steps_per_second": 72.263,
"step": 100
},
{
"epoch": 0.27652292950034224,
"grad_norm": 1168.2845458984375,
"learning_rate": 5.330452921628497e-05,
"loss": 25.9074,
"step": 101
},
{
"epoch": 0.2792607802874743,
"grad_norm": 942.9157104492188,
"learning_rate": 5.247918773366112e-05,
"loss": 23.996,
"step": 102
},
{
"epoch": 0.28199863107460643,
"grad_norm": 1322.2786865234375,
"learning_rate": 5.165316846586541e-05,
"loss": 22.5146,
"step": 103
},
{
"epoch": 0.28473648186173856,
"grad_norm": 1830.467529296875,
"learning_rate": 5.0826697238317935e-05,
"loss": 22.4435,
"step": 104
},
{
"epoch": 0.2874743326488706,
"grad_norm": 1752.7239990234375,
"learning_rate": 5e-05,
"loss": 21.6148,
"step": 105
},
{
"epoch": 0.29021218343600275,
"grad_norm": 1231.7613525390625,
"learning_rate": 4.917330276168208e-05,
"loss": 22.1886,
"step": 106
},
{
"epoch": 0.2929500342231348,
"grad_norm": 1728.1563720703125,
"learning_rate": 4.834683153413459e-05,
"loss": 22.3693,
"step": 107
},
{
"epoch": 0.29568788501026694,
"grad_norm": 1733.2125244140625,
"learning_rate": 4.7520812266338885e-05,
"loss": 22.247,
"step": 108
},
{
"epoch": 0.29842573579739906,
"grad_norm": 1734.533935546875,
"learning_rate": 4.669547078371504e-05,
"loss": 21.8739,
"step": 109
},
{
"epoch": 0.30116358658453113,
"grad_norm": 1873.6116943359375,
"learning_rate": 4.5871032726383386e-05,
"loss": 21.815,
"step": 110
},
{
"epoch": 0.30390143737166325,
"grad_norm": 1351.445556640625,
"learning_rate": 4.504772348747687e-05,
"loss": 21.1957,
"step": 111
},
{
"epoch": 0.3066392881587953,
"grad_norm": 2103.620361328125,
"learning_rate": 4.4225768151520694e-05,
"loss": 21.7416,
"step": 112
},
{
"epoch": 0.30937713894592744,
"grad_norm": 1383.822021484375,
"learning_rate": 4.3405391432896555e-05,
"loss": 21.6563,
"step": 113
},
{
"epoch": 0.31211498973305957,
"grad_norm": 1721.58740234375,
"learning_rate": 4.2586817614407895e-05,
"loss": 22.5202,
"step": 114
},
{
"epoch": 0.31485284052019163,
"grad_norm": 1968.678955078125,
"learning_rate": 4.17702704859633e-05,
"loss": 21.7438,
"step": 115
},
{
"epoch": 0.31759069130732376,
"grad_norm": 1647.9808349609375,
"learning_rate": 4.095597328339452e-05,
"loss": 21.0037,
"step": 116
},
{
"epoch": 0.3203285420944558,
"grad_norm": 1761.572509765625,
"learning_rate": 4.0144148627425993e-05,
"loss": 21.8027,
"step": 117
},
{
"epoch": 0.32306639288158795,
"grad_norm": 2769.5302734375,
"learning_rate": 3.933501846281267e-05,
"loss": 21.9951,
"step": 118
},
{
"epoch": 0.3258042436687201,
"grad_norm": 2277.421875,
"learning_rate": 3.852880399766243e-05,
"loss": 21.12,
"step": 119
},
{
"epoch": 0.32854209445585214,
"grad_norm": 2341.66650390625,
"learning_rate": 3.772572564296005e-05,
"loss": 22.025,
"step": 120
},
{
"epoch": 0.33127994524298426,
"grad_norm": 1773.951416015625,
"learning_rate": 3.6926002952309016e-05,
"loss": 21.2627,
"step": 121
},
{
"epoch": 0.33401779603011633,
"grad_norm": 1828.259521484375,
"learning_rate": 3.612985456190778e-05,
"loss": 22.3545,
"step": 122
},
{
"epoch": 0.33675564681724846,
"grad_norm": 2535.53076171875,
"learning_rate": 3.533749813077677e-05,
"loss": 21.6817,
"step": 123
},
{
"epoch": 0.3394934976043806,
"grad_norm": 2379.452880859375,
"learning_rate": 3.4549150281252636e-05,
"loss": 21.8373,
"step": 124
},
{
"epoch": 0.34223134839151265,
"grad_norm": 3527.873046875,
"learning_rate": 3.3765026539765834e-05,
"loss": 21.5692,
"step": 125
},
{
"epoch": 0.34496919917864477,
"grad_norm": 2357.097900390625,
"learning_rate": 3.298534127791785e-05,
"loss": 21.7367,
"step": 126
},
{
"epoch": 0.34770704996577684,
"grad_norm": 1879.040771484375,
"learning_rate": 3.221030765387417e-05,
"loss": 20.2953,
"step": 127
},
{
"epoch": 0.35044490075290896,
"grad_norm": 3364.228271484375,
"learning_rate": 3.144013755408895e-05,
"loss": 22.1799,
"step": 128
},
{
"epoch": 0.3531827515400411,
"grad_norm": 2147.4453125,
"learning_rate": 3.0675041535377405e-05,
"loss": 21.9013,
"step": 129
},
{
"epoch": 0.35592060232717315,
"grad_norm": 2385.769775390625,
"learning_rate": 2.991522876735154e-05,
"loss": 21.7978,
"step": 130
},
{
"epoch": 0.3586584531143053,
"grad_norm": 2511.2177734375,
"learning_rate": 2.916090697523549e-05,
"loss": 22.1258,
"step": 131
},
{
"epoch": 0.3613963039014374,
"grad_norm": 2592.1796875,
"learning_rate": 2.8412282383075363e-05,
"loss": 21.5315,
"step": 132
},
{
"epoch": 0.36413415468856947,
"grad_norm": 2833.41064453125,
"learning_rate": 2.766955965735968e-05,
"loss": 21.7618,
"step": 133
},
{
"epoch": 0.3668720054757016,
"grad_norm": 5354.45654296875,
"learning_rate": 2.693294185106562e-05,
"loss": 22.2704,
"step": 134
},
{
"epoch": 0.36960985626283366,
"grad_norm": 2638.498779296875,
"learning_rate": 2.6202630348146324e-05,
"loss": 20.916,
"step": 135
},
{
"epoch": 0.3723477070499658,
"grad_norm": 2750.00537109375,
"learning_rate": 2.547882480847461e-05,
"loss": 21.635,
"step": 136
},
{
"epoch": 0.3750855578370979,
"grad_norm": 3375.379638671875,
"learning_rate": 2.476172311325783e-05,
"loss": 21.3345,
"step": 137
},
{
"epoch": 0.37782340862423,
"grad_norm": 3189.904296875,
"learning_rate": 2.405152131093926e-05,
"loss": 21.4638,
"step": 138
},
{
"epoch": 0.3805612594113621,
"grad_norm": 5572.9091796875,
"learning_rate": 2.3348413563600325e-05,
"loss": 21.5838,
"step": 139
},
{
"epoch": 0.38329911019849416,
"grad_norm": 3437.836669921875,
"learning_rate": 2.2652592093878666e-05,
"loss": 21.4991,
"step": 140
},
{
"epoch": 0.3860369609856263,
"grad_norm": 4173.70947265625,
"learning_rate": 2.196424713241637e-05,
"loss": 21.0687,
"step": 141
},
{
"epoch": 0.3887748117727584,
"grad_norm": 4284.75732421875,
"learning_rate": 2.128356686585282e-05,
"loss": 20.7872,
"step": 142
},
{
"epoch": 0.3915126625598905,
"grad_norm": 4492.8427734375,
"learning_rate": 2.061073738537635e-05,
"loss": 21.2808,
"step": 143
},
{
"epoch": 0.3942505133470226,
"grad_norm": 5982.513671875,
"learning_rate": 1.9945942635848748e-05,
"loss": 21.4427,
"step": 144
},
{
"epoch": 0.39698836413415467,
"grad_norm": 4533.22900390625,
"learning_rate": 1.928936436551661e-05,
"loss": 21.1165,
"step": 145
},
{
"epoch": 0.3997262149212868,
"grad_norm": 3245.87353515625,
"learning_rate": 1.8641182076323148e-05,
"loss": 20.9156,
"step": 146
},
{
"epoch": 0.4024640657084189,
"grad_norm": 5858.158203125,
"learning_rate": 1.800157297483417e-05,
"loss": 21.3943,
"step": 147
},
{
"epoch": 0.405201916495551,
"grad_norm": 8529.74609375,
"learning_rate": 1.7370711923791567e-05,
"loss": 20.7764,
"step": 148
},
{
"epoch": 0.4079397672826831,
"grad_norm": 5922.708984375,
"learning_rate": 1.6748771394307585e-05,
"loss": 21.1832,
"step": 149
},
{
"epoch": 0.4106776180698152,
"grad_norm": 6118.82958984375,
"learning_rate": 1.6135921418712956e-05,
"loss": 20.7851,
"step": 150
},
{
"epoch": 0.4106776180698152,
"eval_loss": 5.317856788635254,
"eval_runtime": 2.1337,
"eval_samples_per_second": 288.235,
"eval_steps_per_second": 72.176,
"step": 150
},
{
"epoch": 0.4134154688569473,
"grad_norm": 1872.77685546875,
"learning_rate": 1.553232954407171e-05,
"loss": 22.2569,
"step": 151
},
{
"epoch": 0.4161533196440794,
"grad_norm": 1534.3037109375,
"learning_rate": 1.4938160786375572e-05,
"loss": 22.2431,
"step": 152
},
{
"epoch": 0.4188911704312115,
"grad_norm": 1850.4324951171875,
"learning_rate": 1.435357758543015e-05,
"loss": 22.1803,
"step": 153
},
{
"epoch": 0.4216290212183436,
"grad_norm": 2013.541015625,
"learning_rate": 1.3778739760445552e-05,
"loss": 21.0909,
"step": 154
},
{
"epoch": 0.4243668720054757,
"grad_norm": 1362.583984375,
"learning_rate": 1.3213804466343421e-05,
"loss": 20.5832,
"step": 155
},
{
"epoch": 0.4271047227926078,
"grad_norm": 1862.6373291015625,
"learning_rate": 1.2658926150792322e-05,
"loss": 21.6435,
"step": 156
},
{
"epoch": 0.42984257357973993,
"grad_norm": 1716.3819580078125,
"learning_rate": 1.2114256511983274e-05,
"loss": 21.5584,
"step": 157
},
{
"epoch": 0.432580424366872,
"grad_norm": 2431.09765625,
"learning_rate": 1.157994445715706e-05,
"loss": 21.4107,
"step": 158
},
{
"epoch": 0.4353182751540041,
"grad_norm": 1921.3095703125,
"learning_rate": 1.1056136061894384e-05,
"loss": 21.5094,
"step": 159
},
{
"epoch": 0.4380561259411362,
"grad_norm": 2468.018310546875,
"learning_rate": 1.0542974530180327e-05,
"loss": 20.9342,
"step": 160
},
{
"epoch": 0.4407939767282683,
"grad_norm": 2785.00732421875,
"learning_rate": 1.0040600155253765e-05,
"loss": 20.8618,
"step": 161
},
{
"epoch": 0.44353182751540043,
"grad_norm": 1683.6207275390625,
"learning_rate": 9.549150281252633e-06,
"loss": 21.5471,
"step": 162
},
{
"epoch": 0.4462696783025325,
"grad_norm": 2032.5487060546875,
"learning_rate": 9.068759265665384e-06,
"loss": 20.6241,
"step": 163
},
{
"epoch": 0.4490075290896646,
"grad_norm": 1672.19873046875,
"learning_rate": 8.599558442598998e-06,
"loss": 21.4172,
"step": 164
},
{
"epoch": 0.4517453798767967,
"grad_norm": 1837.2938232421875,
"learning_rate": 8.141676086873572e-06,
"loss": 20.6563,
"step": 165
},
{
"epoch": 0.4544832306639288,
"grad_norm": 3977.8662109375,
"learning_rate": 7.695237378953223e-06,
"loss": 20.1397,
"step": 166
},
{
"epoch": 0.45722108145106094,
"grad_norm": 3161.969482421875,
"learning_rate": 7.260364370723044e-06,
"loss": 20.9031,
"step": 167
},
{
"epoch": 0.459958932238193,
"grad_norm": 2307.077880859375,
"learning_rate": 6.837175952121306e-06,
"loss": 21.5183,
"step": 168
},
{
"epoch": 0.46269678302532513,
"grad_norm": 2354.14697265625,
"learning_rate": 6.425787818636131e-06,
"loss": 21.1961,
"step": 169
},
{
"epoch": 0.4654346338124572,
"grad_norm": 2700.833740234375,
"learning_rate": 6.026312439675552e-06,
"loss": 20.6793,
"step": 170
},
{
"epoch": 0.4681724845995893,
"grad_norm": 2215.9658203125,
"learning_rate": 5.6388590278194096e-06,
"loss": 21.1577,
"step": 171
},
{
"epoch": 0.47091033538672145,
"grad_norm": 3172.705810546875,
"learning_rate": 5.263533508961827e-06,
"loss": 20.1834,
"step": 172
},
{
"epoch": 0.4736481861738535,
"grad_norm": 2935.427001953125,
"learning_rate": 4.900438493352055e-06,
"loss": 21.1244,
"step": 173
},
{
"epoch": 0.47638603696098564,
"grad_norm": 3998.6240234375,
"learning_rate": 4.549673247541875e-06,
"loss": 21.0286,
"step": 174
},
{
"epoch": 0.4791238877481177,
"grad_norm": 2781.40625,
"learning_rate": 4.2113336672471245e-06,
"loss": 19.8333,
"step": 175
},
{
"epoch": 0.48186173853524983,
"grad_norm": 3193.22509765625,
"learning_rate": 3.885512251130763e-06,
"loss": 21.8278,
"step": 176
},
{
"epoch": 0.48459958932238195,
"grad_norm": 3767.453369140625,
"learning_rate": 3.5722980755146517e-06,
"loss": 21.2134,
"step": 177
},
{
"epoch": 0.487337440109514,
"grad_norm": 3105.34033203125,
"learning_rate": 3.271776770026963e-06,
"loss": 21.0267,
"step": 178
},
{
"epoch": 0.49007529089664614,
"grad_norm": 3699.114501953125,
"learning_rate": 2.9840304941919415e-06,
"loss": 20.2173,
"step": 179
},
{
"epoch": 0.4928131416837782,
"grad_norm": 2566.81201171875,
"learning_rate": 2.7091379149682685e-06,
"loss": 21.2091,
"step": 180
},
{
"epoch": 0.49555099247091033,
"grad_norm": 2592.570556640625,
"learning_rate": 2.4471741852423237e-06,
"loss": 21.1579,
"step": 181
},
{
"epoch": 0.49828884325804246,
"grad_norm": 4382.9091796875,
"learning_rate": 2.1982109232821178e-06,
"loss": 21.8012,
"step": 182
},
{
"epoch": 0.5010266940451745,
"grad_norm": 3561.0341796875,
"learning_rate": 1.962316193157593e-06,
"loss": 20.7861,
"step": 183
},
{
"epoch": 0.5037645448323066,
"grad_norm": 2896.11474609375,
"learning_rate": 1.7395544861325718e-06,
"loss": 21.4137,
"step": 184
},
{
"epoch": 0.5065023956194388,
"grad_norm": 4552.0712890625,
"learning_rate": 1.5299867030334814e-06,
"loss": 20.9947,
"step": 185
},
{
"epoch": 0.5092402464065708,
"grad_norm": 3301.87353515625,
"learning_rate": 1.333670137599713e-06,
"loss": 21.8624,
"step": 186
},
{
"epoch": 0.5119780971937029,
"grad_norm": 7506.54345703125,
"learning_rate": 1.1506584608200367e-06,
"loss": 21.8487,
"step": 187
},
{
"epoch": 0.5147159479808351,
"grad_norm": 4901.4677734375,
"learning_rate": 9.810017062595322e-07,
"loss": 21.3934,
"step": 188
},
{
"epoch": 0.5174537987679672,
"grad_norm": 7550.8125,
"learning_rate": 8.247462563808817e-07,
"loss": 21.7613,
"step": 189
},
{
"epoch": 0.5201916495550992,
"grad_norm": 6221.083984375,
"learning_rate": 6.819348298638839e-07,
"loss": 20.6848,
"step": 190
},
{
"epoch": 0.5229295003422314,
"grad_norm": 5665.91650390625,
"learning_rate": 5.526064699265753e-07,
"loss": 21.8913,
"step": 191
},
{
"epoch": 0.5256673511293635,
"grad_norm": 4058.3818359375,
"learning_rate": 4.367965336512403e-07,
"loss": 22.0122,
"step": 192
},
{
"epoch": 0.5284052019164955,
"grad_norm": 4175.564453125,
"learning_rate": 3.3453668231809286e-07,
"loss": 21.4304,
"step": 193
},
{
"epoch": 0.5311430527036276,
"grad_norm": 4648.82568359375,
"learning_rate": 2.458548727494292e-07,
"loss": 21.2028,
"step": 194
},
{
"epoch": 0.5338809034907598,
"grad_norm": 4365.23583984375,
"learning_rate": 1.7077534966650766e-07,
"loss": 22.0969,
"step": 195
},
{
"epoch": 0.5366187542778919,
"grad_norm": 3843.424072265625,
"learning_rate": 1.0931863906127327e-07,
"loss": 21.2312,
"step": 196
},
{
"epoch": 0.5393566050650239,
"grad_norm": 7738.5390625,
"learning_rate": 6.150154258476315e-08,
"loss": 22.2269,
"step": 197
},
{
"epoch": 0.5420944558521561,
"grad_norm": 3570.88232421875,
"learning_rate": 2.7337132953697554e-08,
"loss": 20.8003,
"step": 198
},
{
"epoch": 0.5448323066392882,
"grad_norm": 4533.88525390625,
"learning_rate": 6.834750376549792e-09,
"loss": 21.8544,
"step": 199
},
{
"epoch": 0.5475701574264202,
"grad_norm": 6495.978515625,
"learning_rate": 0.0,
"loss": 21.446,
"step": 200
},
{
"epoch": 0.5475701574264202,
"eval_loss": 5.247466564178467,
"eval_runtime": 2.1294,
"eval_samples_per_second": 288.817,
"eval_steps_per_second": 72.322,
"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": 337932019826688.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}