aleegis09's picture
Training in progress, step 200, checkpoint
d8b0359 verified
{
"best_metric": 1.45925772190094,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 0.37019898195279966,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0018509949097639982,
"grad_norm": 59.89948654174805,
"learning_rate": 1e-05,
"loss": 10.1586,
"step": 1
},
{
"epoch": 0.0018509949097639982,
"eval_loss": 3.6028523445129395,
"eval_runtime": 68.5961,
"eval_samples_per_second": 13.266,
"eval_steps_per_second": 3.324,
"step": 1
},
{
"epoch": 0.0037019898195279964,
"grad_norm": 100.43816375732422,
"learning_rate": 2e-05,
"loss": 11.8233,
"step": 2
},
{
"epoch": 0.005552984729291994,
"grad_norm": 60.06867980957031,
"learning_rate": 3e-05,
"loss": 10.9429,
"step": 3
},
{
"epoch": 0.007403979639055993,
"grad_norm": 53.39894104003906,
"learning_rate": 4e-05,
"loss": 9.6525,
"step": 4
},
{
"epoch": 0.00925497454881999,
"grad_norm": 67.93931579589844,
"learning_rate": 5e-05,
"loss": 10.0067,
"step": 5
},
{
"epoch": 0.011105969458583989,
"grad_norm": 47.26503372192383,
"learning_rate": 6e-05,
"loss": 8.465,
"step": 6
},
{
"epoch": 0.012956964368347987,
"grad_norm": 38.43549346923828,
"learning_rate": 7e-05,
"loss": 8.9691,
"step": 7
},
{
"epoch": 0.014807959278111986,
"grad_norm": 40.666404724121094,
"learning_rate": 8e-05,
"loss": 7.4453,
"step": 8
},
{
"epoch": 0.016658954187875982,
"grad_norm": 29.77450942993164,
"learning_rate": 9e-05,
"loss": 8.4803,
"step": 9
},
{
"epoch": 0.01850994909763998,
"grad_norm": 29.103349685668945,
"learning_rate": 0.0001,
"loss": 7.4746,
"step": 10
},
{
"epoch": 0.02036094400740398,
"grad_norm": 30.80104637145996,
"learning_rate": 9.999316524962345e-05,
"loss": 8.5573,
"step": 11
},
{
"epoch": 0.022211938917167977,
"grad_norm": 30.97075653076172,
"learning_rate": 9.997266286704631e-05,
"loss": 8.1872,
"step": 12
},
{
"epoch": 0.024062933826931976,
"grad_norm": 29.791688919067383,
"learning_rate": 9.993849845741524e-05,
"loss": 7.0701,
"step": 13
},
{
"epoch": 0.025913928736695974,
"grad_norm": 46.57557678222656,
"learning_rate": 9.989068136093873e-05,
"loss": 8.5949,
"step": 14
},
{
"epoch": 0.027764923646459973,
"grad_norm": 31.162527084350586,
"learning_rate": 9.98292246503335e-05,
"loss": 8.3614,
"step": 15
},
{
"epoch": 0.02961591855622397,
"grad_norm": 28.787185668945312,
"learning_rate": 9.975414512725057e-05,
"loss": 7.7322,
"step": 16
},
{
"epoch": 0.03146691346598797,
"grad_norm": 26.941804885864258,
"learning_rate": 9.966546331768191e-05,
"loss": 6.6127,
"step": 17
},
{
"epoch": 0.033317908375751965,
"grad_norm": 26.8864688873291,
"learning_rate": 9.956320346634876e-05,
"loss": 7.7146,
"step": 18
},
{
"epoch": 0.035168903285515966,
"grad_norm": 25.0333194732666,
"learning_rate": 9.944739353007344e-05,
"loss": 6.7545,
"step": 19
},
{
"epoch": 0.03701989819527996,
"grad_norm": 22.383333206176758,
"learning_rate": 9.931806517013612e-05,
"loss": 7.2769,
"step": 20
},
{
"epoch": 0.03887089310504396,
"grad_norm": 20.333555221557617,
"learning_rate": 9.917525374361912e-05,
"loss": 5.9189,
"step": 21
},
{
"epoch": 0.04072188801480796,
"grad_norm": 22.781673431396484,
"learning_rate": 9.901899829374047e-05,
"loss": 6.9341,
"step": 22
},
{
"epoch": 0.04257288292457196,
"grad_norm": 24.01888084411621,
"learning_rate": 9.884934153917997e-05,
"loss": 7.1848,
"step": 23
},
{
"epoch": 0.044423877834335955,
"grad_norm": 22.655099868774414,
"learning_rate": 9.86663298624003e-05,
"loss": 6.002,
"step": 24
},
{
"epoch": 0.04627487274409996,
"grad_norm": 24.61270523071289,
"learning_rate": 9.847001329696653e-05,
"loss": 6.2469,
"step": 25
},
{
"epoch": 0.04812586765386395,
"grad_norm": 25.12575912475586,
"learning_rate": 9.826044551386744e-05,
"loss": 6.8516,
"step": 26
},
{
"epoch": 0.04997686256362795,
"grad_norm": 22.560014724731445,
"learning_rate": 9.803768380684242e-05,
"loss": 5.8165,
"step": 27
},
{
"epoch": 0.05182785747339195,
"grad_norm": 22.28982162475586,
"learning_rate": 9.780178907671789e-05,
"loss": 6.3579,
"step": 28
},
{
"epoch": 0.053678852383155944,
"grad_norm": 24.577585220336914,
"learning_rate": 9.755282581475769e-05,
"loss": 5.9195,
"step": 29
},
{
"epoch": 0.055529847292919945,
"grad_norm": 30.741830825805664,
"learning_rate": 9.729086208503174e-05,
"loss": 7.2511,
"step": 30
},
{
"epoch": 0.05738084220268394,
"grad_norm": 26.450328826904297,
"learning_rate": 9.701596950580806e-05,
"loss": 7.2489,
"step": 31
},
{
"epoch": 0.05923183711244794,
"grad_norm": 24.98292350769043,
"learning_rate": 9.672822322997305e-05,
"loss": 6.9673,
"step": 32
},
{
"epoch": 0.06108283202221194,
"grad_norm": 21.95580291748047,
"learning_rate": 9.642770192448536e-05,
"loss": 5.2675,
"step": 33
},
{
"epoch": 0.06293382693197594,
"grad_norm": 23.09197235107422,
"learning_rate": 9.611448774886924e-05,
"loss": 4.9874,
"step": 34
},
{
"epoch": 0.06478482184173993,
"grad_norm": 24.217933654785156,
"learning_rate": 9.578866633275288e-05,
"loss": 6.4793,
"step": 35
},
{
"epoch": 0.06663581675150393,
"grad_norm": 25.319543838500977,
"learning_rate": 9.545032675245813e-05,
"loss": 5.6718,
"step": 36
},
{
"epoch": 0.06848681166126794,
"grad_norm": 20.073984146118164,
"learning_rate": 9.509956150664796e-05,
"loss": 4.0343,
"step": 37
},
{
"epoch": 0.07033780657103193,
"grad_norm": 25.865989685058594,
"learning_rate": 9.473646649103818e-05,
"loss": 4.9742,
"step": 38
},
{
"epoch": 0.07218880148079593,
"grad_norm": 26.664836883544922,
"learning_rate": 9.43611409721806e-05,
"loss": 6.4406,
"step": 39
},
{
"epoch": 0.07403979639055992,
"grad_norm": 21.804039001464844,
"learning_rate": 9.397368756032445e-05,
"loss": 4.4743,
"step": 40
},
{
"epoch": 0.07589079130032392,
"grad_norm": 25.674467086791992,
"learning_rate": 9.357421218136386e-05,
"loss": 6.674,
"step": 41
},
{
"epoch": 0.07774178621008793,
"grad_norm": 22.556793212890625,
"learning_rate": 9.316282404787871e-05,
"loss": 5.0478,
"step": 42
},
{
"epoch": 0.07959278111985192,
"grad_norm": 22.43199920654297,
"learning_rate": 9.273963562927695e-05,
"loss": 4.8246,
"step": 43
},
{
"epoch": 0.08144377602961592,
"grad_norm": 22.04460906982422,
"learning_rate": 9.230476262104677e-05,
"loss": 6.7395,
"step": 44
},
{
"epoch": 0.08329477093937991,
"grad_norm": 27.928298950195312,
"learning_rate": 9.185832391312644e-05,
"loss": 5.0662,
"step": 45
},
{
"epoch": 0.08514576584914392,
"grad_norm": 21.5211181640625,
"learning_rate": 9.140044155740101e-05,
"loss": 4.8276,
"step": 46
},
{
"epoch": 0.08699676075890792,
"grad_norm": 30.689226150512695,
"learning_rate": 9.093124073433463e-05,
"loss": 5.0762,
"step": 47
},
{
"epoch": 0.08884775566867191,
"grad_norm": 24.589027404785156,
"learning_rate": 9.045084971874738e-05,
"loss": 5.8773,
"step": 48
},
{
"epoch": 0.0906987505784359,
"grad_norm": 25.96907615661621,
"learning_rate": 8.995939984474624e-05,
"loss": 4.7589,
"step": 49
},
{
"epoch": 0.09254974548819991,
"grad_norm": 31.189998626708984,
"learning_rate": 8.945702546981969e-05,
"loss": 4.126,
"step": 50
},
{
"epoch": 0.09254974548819991,
"eval_loss": 2.353039264678955,
"eval_runtime": 69.882,
"eval_samples_per_second": 13.022,
"eval_steps_per_second": 3.263,
"step": 50
},
{
"epoch": 0.09440074039796391,
"grad_norm": 110.58844757080078,
"learning_rate": 8.894386393810563e-05,
"loss": 15.2761,
"step": 51
},
{
"epoch": 0.0962517353077279,
"grad_norm": 58.721832275390625,
"learning_rate": 8.842005554284296e-05,
"loss": 11.9424,
"step": 52
},
{
"epoch": 0.0981027302174919,
"grad_norm": 30.038488388061523,
"learning_rate": 8.788574348801675e-05,
"loss": 9.9225,
"step": 53
},
{
"epoch": 0.0999537251272559,
"grad_norm": 20.106815338134766,
"learning_rate": 8.73410738492077e-05,
"loss": 8.2082,
"step": 54
},
{
"epoch": 0.1018047200370199,
"grad_norm": 23.7856388092041,
"learning_rate": 8.678619553365659e-05,
"loss": 8.4048,
"step": 55
},
{
"epoch": 0.1036557149467839,
"grad_norm": 25.19913101196289,
"learning_rate": 8.622126023955446e-05,
"loss": 8.0735,
"step": 56
},
{
"epoch": 0.10550670985654789,
"grad_norm": 25.34491539001465,
"learning_rate": 8.564642241456986e-05,
"loss": 7.4242,
"step": 57
},
{
"epoch": 0.10735770476631189,
"grad_norm": 23.657678604125977,
"learning_rate": 8.506183921362443e-05,
"loss": 8.2696,
"step": 58
},
{
"epoch": 0.1092086996760759,
"grad_norm": 23.854257583618164,
"learning_rate": 8.44676704559283e-05,
"loss": 8.2411,
"step": 59
},
{
"epoch": 0.11105969458583989,
"grad_norm": 25.120820999145508,
"learning_rate": 8.386407858128706e-05,
"loss": 7.9448,
"step": 60
},
{
"epoch": 0.11291068949560389,
"grad_norm": 22.30246925354004,
"learning_rate": 8.32512286056924e-05,
"loss": 7.962,
"step": 61
},
{
"epoch": 0.11476168440536788,
"grad_norm": 22.673364639282227,
"learning_rate": 8.262928807620843e-05,
"loss": 7.8381,
"step": 62
},
{
"epoch": 0.11661267931513189,
"grad_norm": 26.36172103881836,
"learning_rate": 8.199842702516583e-05,
"loss": 8.6518,
"step": 63
},
{
"epoch": 0.11846367422489588,
"grad_norm": 22.343236923217773,
"learning_rate": 8.135881792367686e-05,
"loss": 6.7058,
"step": 64
},
{
"epoch": 0.12031466913465988,
"grad_norm": 29.06987190246582,
"learning_rate": 8.07106356344834e-05,
"loss": 7.2401,
"step": 65
},
{
"epoch": 0.12216566404442387,
"grad_norm": 25.895959854125977,
"learning_rate": 8.005405736415126e-05,
"loss": 8.3782,
"step": 66
},
{
"epoch": 0.12401665895418787,
"grad_norm": 24.11650848388672,
"learning_rate": 7.938926261462366e-05,
"loss": 7.4369,
"step": 67
},
{
"epoch": 0.12586765386395188,
"grad_norm": 22.802165985107422,
"learning_rate": 7.871643313414718e-05,
"loss": 5.9717,
"step": 68
},
{
"epoch": 0.12771864877371586,
"grad_norm": 21.314449310302734,
"learning_rate": 7.803575286758364e-05,
"loss": 7.2408,
"step": 69
},
{
"epoch": 0.12956964368347987,
"grad_norm": 18.106534957885742,
"learning_rate": 7.734740790612136e-05,
"loss": 6.1463,
"step": 70
},
{
"epoch": 0.13142063859324388,
"grad_norm": 21.958393096923828,
"learning_rate": 7.66515864363997e-05,
"loss": 6.8193,
"step": 71
},
{
"epoch": 0.13327163350300786,
"grad_norm": 19.48069953918457,
"learning_rate": 7.594847868906076e-05,
"loss": 6.5375,
"step": 72
},
{
"epoch": 0.13512262841277187,
"grad_norm": 23.702938079833984,
"learning_rate": 7.52382768867422e-05,
"loss": 7.0712,
"step": 73
},
{
"epoch": 0.13697362332253588,
"grad_norm": 22.461809158325195,
"learning_rate": 7.452117519152542e-05,
"loss": 5.9379,
"step": 74
},
{
"epoch": 0.13882461823229986,
"grad_norm": 24.132768630981445,
"learning_rate": 7.379736965185368e-05,
"loss": 6.974,
"step": 75
},
{
"epoch": 0.14067561314206387,
"grad_norm": 25.628990173339844,
"learning_rate": 7.30670581489344e-05,
"loss": 7.1283,
"step": 76
},
{
"epoch": 0.14252660805182785,
"grad_norm": 24.302547454833984,
"learning_rate": 7.233044034264034e-05,
"loss": 7.334,
"step": 77
},
{
"epoch": 0.14437760296159186,
"grad_norm": 18.943586349487305,
"learning_rate": 7.158771761692464e-05,
"loss": 5.2304,
"step": 78
},
{
"epoch": 0.14622859787135586,
"grad_norm": 18.917469024658203,
"learning_rate": 7.083909302476453e-05,
"loss": 5.6487,
"step": 79
},
{
"epoch": 0.14807959278111985,
"grad_norm": 21.635984420776367,
"learning_rate": 7.008477123264848e-05,
"loss": 6.0559,
"step": 80
},
{
"epoch": 0.14993058769088385,
"grad_norm": 21.822101593017578,
"learning_rate": 6.932495846462261e-05,
"loss": 6.1391,
"step": 81
},
{
"epoch": 0.15178158260064784,
"grad_norm": 19.244054794311523,
"learning_rate": 6.855986244591104e-05,
"loss": 5.7351,
"step": 82
},
{
"epoch": 0.15363257751041184,
"grad_norm": 22.16774559020996,
"learning_rate": 6.778969234612584e-05,
"loss": 5.5628,
"step": 83
},
{
"epoch": 0.15548357242017585,
"grad_norm": 19.686809539794922,
"learning_rate": 6.701465872208216e-05,
"loss": 5.6571,
"step": 84
},
{
"epoch": 0.15733456732993983,
"grad_norm": 18.753782272338867,
"learning_rate": 6.623497346023418e-05,
"loss": 5.5124,
"step": 85
},
{
"epoch": 0.15918556223970384,
"grad_norm": 22.904891967773438,
"learning_rate": 6.545084971874738e-05,
"loss": 6.0632,
"step": 86
},
{
"epoch": 0.16103655714946785,
"grad_norm": 20.548492431640625,
"learning_rate": 6.466250186922325e-05,
"loss": 6.0648,
"step": 87
},
{
"epoch": 0.16288755205923183,
"grad_norm": 21.983842849731445,
"learning_rate": 6.387014543809223e-05,
"loss": 5.0114,
"step": 88
},
{
"epoch": 0.16473854696899584,
"grad_norm": 18.991622924804688,
"learning_rate": 6.307399704769099e-05,
"loss": 5.3445,
"step": 89
},
{
"epoch": 0.16658954187875982,
"grad_norm": 20.161306381225586,
"learning_rate": 6.227427435703997e-05,
"loss": 5.8939,
"step": 90
},
{
"epoch": 0.16844053678852383,
"grad_norm": 16.23657989501953,
"learning_rate": 6.147119600233758e-05,
"loss": 4.8956,
"step": 91
},
{
"epoch": 0.17029153169828784,
"grad_norm": 15.523630142211914,
"learning_rate": 6.066498153718735e-05,
"loss": 3.6043,
"step": 92
},
{
"epoch": 0.17214252660805182,
"grad_norm": 17.354217529296875,
"learning_rate": 5.985585137257401e-05,
"loss": 4.0083,
"step": 93
},
{
"epoch": 0.17399352151781583,
"grad_norm": 18.608896255493164,
"learning_rate": 5.90440267166055e-05,
"loss": 5.0076,
"step": 94
},
{
"epoch": 0.1758445164275798,
"grad_norm": 18.4306583404541,
"learning_rate": 5.8229729514036705e-05,
"loss": 4.9611,
"step": 95
},
{
"epoch": 0.17769551133734382,
"grad_norm": 17.808902740478516,
"learning_rate": 5.74131823855921e-05,
"loss": 4.66,
"step": 96
},
{
"epoch": 0.17954650624710783,
"grad_norm": 20.695646286010742,
"learning_rate": 5.6594608567103456e-05,
"loss": 5.8069,
"step": 97
},
{
"epoch": 0.1813975011568718,
"grad_norm": 20.554363250732422,
"learning_rate": 5.577423184847932e-05,
"loss": 4.5419,
"step": 98
},
{
"epoch": 0.18324849606663582,
"grad_norm": 20.178359985351562,
"learning_rate": 5.495227651252315e-05,
"loss": 4.041,
"step": 99
},
{
"epoch": 0.18509949097639983,
"grad_norm": 18.658552169799805,
"learning_rate": 5.4128967273616625e-05,
"loss": 1.9184,
"step": 100
},
{
"epoch": 0.18509949097639983,
"eval_loss": 1.8544436693191528,
"eval_runtime": 69.8668,
"eval_samples_per_second": 13.025,
"eval_steps_per_second": 3.263,
"step": 100
},
{
"epoch": 0.1869504858861638,
"grad_norm": 41.62975311279297,
"learning_rate": 5.330452921628497e-05,
"loss": 10.6419,
"step": 101
},
{
"epoch": 0.18880148079592782,
"grad_norm": 29.18929672241211,
"learning_rate": 5.247918773366112e-05,
"loss": 10.8545,
"step": 102
},
{
"epoch": 0.1906524757056918,
"grad_norm": 20.762163162231445,
"learning_rate": 5.165316846586541e-05,
"loss": 9.604,
"step": 103
},
{
"epoch": 0.1925034706154558,
"grad_norm": 17.0522403717041,
"learning_rate": 5.0826697238317935e-05,
"loss": 7.3362,
"step": 104
},
{
"epoch": 0.19435446552521982,
"grad_norm": 15.715831756591797,
"learning_rate": 5e-05,
"loss": 8.2619,
"step": 105
},
{
"epoch": 0.1962054604349838,
"grad_norm": 17.21978187561035,
"learning_rate": 4.917330276168208e-05,
"loss": 7.198,
"step": 106
},
{
"epoch": 0.1980564553447478,
"grad_norm": 17.804431915283203,
"learning_rate": 4.834683153413459e-05,
"loss": 7.937,
"step": 107
},
{
"epoch": 0.1999074502545118,
"grad_norm": 17.663671493530273,
"learning_rate": 4.7520812266338885e-05,
"loss": 7.3717,
"step": 108
},
{
"epoch": 0.2017584451642758,
"grad_norm": 19.943225860595703,
"learning_rate": 4.669547078371504e-05,
"loss": 7.8254,
"step": 109
},
{
"epoch": 0.2036094400740398,
"grad_norm": 19.38444709777832,
"learning_rate": 4.5871032726383386e-05,
"loss": 8.0488,
"step": 110
},
{
"epoch": 0.20546043498380379,
"grad_norm": 20.862642288208008,
"learning_rate": 4.504772348747687e-05,
"loss": 7.8827,
"step": 111
},
{
"epoch": 0.2073114298935678,
"grad_norm": 21.540983200073242,
"learning_rate": 4.4225768151520694e-05,
"loss": 7.5021,
"step": 112
},
{
"epoch": 0.2091624248033318,
"grad_norm": 24.776857376098633,
"learning_rate": 4.3405391432896555e-05,
"loss": 8.8133,
"step": 113
},
{
"epoch": 0.21101341971309578,
"grad_norm": 18.81482696533203,
"learning_rate": 4.2586817614407895e-05,
"loss": 6.2371,
"step": 114
},
{
"epoch": 0.2128644146228598,
"grad_norm": 19.74003791809082,
"learning_rate": 4.17702704859633e-05,
"loss": 7.1802,
"step": 115
},
{
"epoch": 0.21471540953262377,
"grad_norm": 23.482675552368164,
"learning_rate": 4.095597328339452e-05,
"loss": 7.1844,
"step": 116
},
{
"epoch": 0.21656640444238778,
"grad_norm": 21.858055114746094,
"learning_rate": 4.0144148627425993e-05,
"loss": 6.6821,
"step": 117
},
{
"epoch": 0.2184173993521518,
"grad_norm": 23.546022415161133,
"learning_rate": 3.933501846281267e-05,
"loss": 6.0886,
"step": 118
},
{
"epoch": 0.22026839426191577,
"grad_norm": 20.7333927154541,
"learning_rate": 3.852880399766243e-05,
"loss": 7.5445,
"step": 119
},
{
"epoch": 0.22211938917167978,
"grad_norm": 22.107837677001953,
"learning_rate": 3.772572564296005e-05,
"loss": 5.8253,
"step": 120
},
{
"epoch": 0.22397038408144376,
"grad_norm": 20.686031341552734,
"learning_rate": 3.6926002952309016e-05,
"loss": 5.9198,
"step": 121
},
{
"epoch": 0.22582137899120777,
"grad_norm": 22.019136428833008,
"learning_rate": 3.612985456190778e-05,
"loss": 6.5592,
"step": 122
},
{
"epoch": 0.22767237390097178,
"grad_norm": 18.745075225830078,
"learning_rate": 3.533749813077677e-05,
"loss": 5.8053,
"step": 123
},
{
"epoch": 0.22952336881073576,
"grad_norm": 19.928760528564453,
"learning_rate": 3.4549150281252636e-05,
"loss": 5.6467,
"step": 124
},
{
"epoch": 0.23137436372049977,
"grad_norm": 23.67180824279785,
"learning_rate": 3.3765026539765834e-05,
"loss": 6.1394,
"step": 125
},
{
"epoch": 0.23322535863026378,
"grad_norm": 21.75117301940918,
"learning_rate": 3.298534127791785e-05,
"loss": 5.6863,
"step": 126
},
{
"epoch": 0.23507635354002776,
"grad_norm": 19.03370475769043,
"learning_rate": 3.221030765387417e-05,
"loss": 5.2691,
"step": 127
},
{
"epoch": 0.23692734844979177,
"grad_norm": 22.11434555053711,
"learning_rate": 3.144013755408895e-05,
"loss": 7.5865,
"step": 128
},
{
"epoch": 0.23877834335955575,
"grad_norm": 20.700796127319336,
"learning_rate": 3.0675041535377405e-05,
"loss": 4.9502,
"step": 129
},
{
"epoch": 0.24062933826931976,
"grad_norm": 22.236141204833984,
"learning_rate": 2.991522876735154e-05,
"loss": 6.357,
"step": 130
},
{
"epoch": 0.24248033317908377,
"grad_norm": 19.655351638793945,
"learning_rate": 2.916090697523549e-05,
"loss": 4.5109,
"step": 131
},
{
"epoch": 0.24433132808884775,
"grad_norm": 20.140209197998047,
"learning_rate": 2.8412282383075363e-05,
"loss": 5.5753,
"step": 132
},
{
"epoch": 0.24618232299861176,
"grad_norm": 19.17459487915039,
"learning_rate": 2.766955965735968e-05,
"loss": 5.437,
"step": 133
},
{
"epoch": 0.24803331790837574,
"grad_norm": 19.464014053344727,
"learning_rate": 2.693294185106562e-05,
"loss": 5.243,
"step": 134
},
{
"epoch": 0.24988431281813975,
"grad_norm": 18.852859497070312,
"learning_rate": 2.6202630348146324e-05,
"loss": 6.4638,
"step": 135
},
{
"epoch": 0.25173530772790376,
"grad_norm": 14.234766006469727,
"learning_rate": 2.547882480847461e-05,
"loss": 3.8942,
"step": 136
},
{
"epoch": 0.25358630263766774,
"grad_norm": 15.460630416870117,
"learning_rate": 2.476172311325783e-05,
"loss": 3.9669,
"step": 137
},
{
"epoch": 0.2554372975474317,
"grad_norm": 19.800920486450195,
"learning_rate": 2.405152131093926e-05,
"loss": 5.8222,
"step": 138
},
{
"epoch": 0.25728829245719576,
"grad_norm": 17.38718605041504,
"learning_rate": 2.3348413563600325e-05,
"loss": 6.0962,
"step": 139
},
{
"epoch": 0.25913928736695974,
"grad_norm": 15.048430442810059,
"learning_rate": 2.2652592093878666e-05,
"loss": 3.9541,
"step": 140
},
{
"epoch": 0.2609902822767237,
"grad_norm": 16.55597496032715,
"learning_rate": 2.196424713241637e-05,
"loss": 4.1839,
"step": 141
},
{
"epoch": 0.26284127718648775,
"grad_norm": 16.657272338867188,
"learning_rate": 2.128356686585282e-05,
"loss": 4.9786,
"step": 142
},
{
"epoch": 0.26469227209625174,
"grad_norm": 21.036558151245117,
"learning_rate": 2.061073738537635e-05,
"loss": 5.0662,
"step": 143
},
{
"epoch": 0.2665432670060157,
"grad_norm": 17.173782348632812,
"learning_rate": 1.9945942635848748e-05,
"loss": 4.0465,
"step": 144
},
{
"epoch": 0.26839426191577975,
"grad_norm": 18.073938369750977,
"learning_rate": 1.928936436551661e-05,
"loss": 4.8604,
"step": 145
},
{
"epoch": 0.27024525682554373,
"grad_norm": 14.32249641418457,
"learning_rate": 1.8641182076323148e-05,
"loss": 3.3291,
"step": 146
},
{
"epoch": 0.2720962517353077,
"grad_norm": 20.76031494140625,
"learning_rate": 1.800157297483417e-05,
"loss": 5.7931,
"step": 147
},
{
"epoch": 0.27394724664507175,
"grad_norm": 15.284453392028809,
"learning_rate": 1.7370711923791567e-05,
"loss": 3.1407,
"step": 148
},
{
"epoch": 0.27579824155483573,
"grad_norm": 23.898174285888672,
"learning_rate": 1.6748771394307585e-05,
"loss": 5.8988,
"step": 149
},
{
"epoch": 0.2776492364645997,
"grad_norm": 20.32386589050293,
"learning_rate": 1.6135921418712956e-05,
"loss": 3.6415,
"step": 150
},
{
"epoch": 0.2776492364645997,
"eval_loss": 1.531545639038086,
"eval_runtime": 69.842,
"eval_samples_per_second": 13.029,
"eval_steps_per_second": 3.265,
"step": 150
},
{
"epoch": 0.2795002313743637,
"grad_norm": 21.202672958374023,
"learning_rate": 1.553232954407171e-05,
"loss": 8.2199,
"step": 151
},
{
"epoch": 0.28135122628412773,
"grad_norm": 22.77549934387207,
"learning_rate": 1.4938160786375572e-05,
"loss": 8.2585,
"step": 152
},
{
"epoch": 0.2832022211938917,
"grad_norm": 21.384912490844727,
"learning_rate": 1.435357758543015e-05,
"loss": 7.9309,
"step": 153
},
{
"epoch": 0.2850532161036557,
"grad_norm": 18.97314453125,
"learning_rate": 1.3778739760445552e-05,
"loss": 8.0805,
"step": 154
},
{
"epoch": 0.28690421101341973,
"grad_norm": 18.570676803588867,
"learning_rate": 1.3213804466343421e-05,
"loss": 7.4273,
"step": 155
},
{
"epoch": 0.2887552059231837,
"grad_norm": 20.62229347229004,
"learning_rate": 1.2658926150792322e-05,
"loss": 7.6848,
"step": 156
},
{
"epoch": 0.2906062008329477,
"grad_norm": 16.83143424987793,
"learning_rate": 1.2114256511983274e-05,
"loss": 6.4836,
"step": 157
},
{
"epoch": 0.29245719574271173,
"grad_norm": 21.52330207824707,
"learning_rate": 1.157994445715706e-05,
"loss": 7.286,
"step": 158
},
{
"epoch": 0.2943081906524757,
"grad_norm": 21.631942749023438,
"learning_rate": 1.1056136061894384e-05,
"loss": 7.57,
"step": 159
},
{
"epoch": 0.2961591855622397,
"grad_norm": 20.015647888183594,
"learning_rate": 1.0542974530180327e-05,
"loss": 7.5558,
"step": 160
},
{
"epoch": 0.2980101804720037,
"grad_norm": 20.935726165771484,
"learning_rate": 1.0040600155253765e-05,
"loss": 9.1985,
"step": 161
},
{
"epoch": 0.2998611753817677,
"grad_norm": 20.2431583404541,
"learning_rate": 9.549150281252633e-06,
"loss": 6.8121,
"step": 162
},
{
"epoch": 0.3017121702915317,
"grad_norm": 20.497529983520508,
"learning_rate": 9.068759265665384e-06,
"loss": 6.5842,
"step": 163
},
{
"epoch": 0.30356316520129567,
"grad_norm": 19.90401840209961,
"learning_rate": 8.599558442598998e-06,
"loss": 6.6889,
"step": 164
},
{
"epoch": 0.3054141601110597,
"grad_norm": 20.3807430267334,
"learning_rate": 8.141676086873572e-06,
"loss": 5.8842,
"step": 165
},
{
"epoch": 0.3072651550208237,
"grad_norm": 19.59682846069336,
"learning_rate": 7.695237378953223e-06,
"loss": 6.7799,
"step": 166
},
{
"epoch": 0.30911614993058767,
"grad_norm": 16.78765296936035,
"learning_rate": 7.260364370723044e-06,
"loss": 5.4901,
"step": 167
},
{
"epoch": 0.3109671448403517,
"grad_norm": 16.29376983642578,
"learning_rate": 6.837175952121306e-06,
"loss": 5.9158,
"step": 168
},
{
"epoch": 0.3128181397501157,
"grad_norm": 18.27606201171875,
"learning_rate": 6.425787818636131e-06,
"loss": 6.7209,
"step": 169
},
{
"epoch": 0.31466913465987967,
"grad_norm": 17.894317626953125,
"learning_rate": 6.026312439675552e-06,
"loss": 6.3001,
"step": 170
},
{
"epoch": 0.3165201295696437,
"grad_norm": 21.63971710205078,
"learning_rate": 5.6388590278194096e-06,
"loss": 7.1544,
"step": 171
},
{
"epoch": 0.3183711244794077,
"grad_norm": 21.629539489746094,
"learning_rate": 5.263533508961827e-06,
"loss": 6.2639,
"step": 172
},
{
"epoch": 0.32022211938917167,
"grad_norm": 19.84263801574707,
"learning_rate": 4.900438493352055e-06,
"loss": 6.8295,
"step": 173
},
{
"epoch": 0.3220731142989357,
"grad_norm": 20.238880157470703,
"learning_rate": 4.549673247541875e-06,
"loss": 5.8894,
"step": 174
},
{
"epoch": 0.3239241092086997,
"grad_norm": 17.637775421142578,
"learning_rate": 4.2113336672471245e-06,
"loss": 4.9432,
"step": 175
},
{
"epoch": 0.32577510411846367,
"grad_norm": 20.498844146728516,
"learning_rate": 3.885512251130763e-06,
"loss": 5.9794,
"step": 176
},
{
"epoch": 0.32762609902822765,
"grad_norm": 17.650970458984375,
"learning_rate": 3.5722980755146517e-06,
"loss": 4.5091,
"step": 177
},
{
"epoch": 0.3294770939379917,
"grad_norm": 21.161008834838867,
"learning_rate": 3.271776770026963e-06,
"loss": 6.4282,
"step": 178
},
{
"epoch": 0.33132808884775566,
"grad_norm": 19.688852310180664,
"learning_rate": 2.9840304941919415e-06,
"loss": 5.8306,
"step": 179
},
{
"epoch": 0.33317908375751965,
"grad_norm": 17.909574508666992,
"learning_rate": 2.7091379149682685e-06,
"loss": 6.0146,
"step": 180
},
{
"epoch": 0.3350300786672837,
"grad_norm": 19.697717666625977,
"learning_rate": 2.4471741852423237e-06,
"loss": 5.26,
"step": 181
},
{
"epoch": 0.33688107357704766,
"grad_norm": 22.01299476623535,
"learning_rate": 2.1982109232821178e-06,
"loss": 6.6505,
"step": 182
},
{
"epoch": 0.33873206848681164,
"grad_norm": 16.821945190429688,
"learning_rate": 1.962316193157593e-06,
"loss": 5.1342,
"step": 183
},
{
"epoch": 0.3405830633965757,
"grad_norm": 17.76853370666504,
"learning_rate": 1.7395544861325718e-06,
"loss": 4.8088,
"step": 184
},
{
"epoch": 0.34243405830633966,
"grad_norm": 20.280498504638672,
"learning_rate": 1.5299867030334814e-06,
"loss": 5.978,
"step": 185
},
{
"epoch": 0.34428505321610364,
"grad_norm": 20.38216209411621,
"learning_rate": 1.333670137599713e-06,
"loss": 5.6034,
"step": 186
},
{
"epoch": 0.3461360481258677,
"grad_norm": 17.165733337402344,
"learning_rate": 1.1506584608200367e-06,
"loss": 4.2564,
"step": 187
},
{
"epoch": 0.34798704303563166,
"grad_norm": 17.380369186401367,
"learning_rate": 9.810017062595322e-07,
"loss": 4.6483,
"step": 188
},
{
"epoch": 0.34983803794539564,
"grad_norm": 20.93096351623535,
"learning_rate": 8.247462563808817e-07,
"loss": 5.1795,
"step": 189
},
{
"epoch": 0.3516890328551596,
"grad_norm": 17.947641372680664,
"learning_rate": 6.819348298638839e-07,
"loss": 5.088,
"step": 190
},
{
"epoch": 0.35354002776492366,
"grad_norm": 17.104299545288086,
"learning_rate": 5.526064699265753e-07,
"loss": 3.7657,
"step": 191
},
{
"epoch": 0.35539102267468764,
"grad_norm": 19.836400985717773,
"learning_rate": 4.367965336512403e-07,
"loss": 4.1091,
"step": 192
},
{
"epoch": 0.3572420175844516,
"grad_norm": 16.77385902404785,
"learning_rate": 3.3453668231809286e-07,
"loss": 5.1161,
"step": 193
},
{
"epoch": 0.35909301249421566,
"grad_norm": 18.861099243164062,
"learning_rate": 2.458548727494292e-07,
"loss": 4.392,
"step": 194
},
{
"epoch": 0.36094400740397964,
"grad_norm": 15.38021469116211,
"learning_rate": 1.7077534966650766e-07,
"loss": 3.6737,
"step": 195
},
{
"epoch": 0.3627950023137436,
"grad_norm": 21.193696975708008,
"learning_rate": 1.0931863906127327e-07,
"loss": 6.081,
"step": 196
},
{
"epoch": 0.36464599722350766,
"grad_norm": 20.111827850341797,
"learning_rate": 6.150154258476315e-08,
"loss": 5.7416,
"step": 197
},
{
"epoch": 0.36649699213327164,
"grad_norm": 18.77097511291504,
"learning_rate": 2.7337132953697554e-08,
"loss": 4.9755,
"step": 198
},
{
"epoch": 0.3683479870430356,
"grad_norm": 26.088537216186523,
"learning_rate": 6.834750376549792e-09,
"loss": 5.5487,
"step": 199
},
{
"epoch": 0.37019898195279966,
"grad_norm": 25.18060302734375,
"learning_rate": 0.0,
"loss": 4.2396,
"step": 200
},
{
"epoch": 0.37019898195279966,
"eval_loss": 1.45925772190094,
"eval_runtime": 69.8425,
"eval_samples_per_second": 13.029,
"eval_steps_per_second": 3.264,
"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": 2.862000849813504e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}