prxy5605's picture
Training in progress, step 200, checkpoint
c2d7ea3 verified
{
"best_metric": 0.38890671730041504,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 0.1262427015938141,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006312135079690705,
"grad_norm": 2.510498523712158,
"learning_rate": 5e-06,
"loss": 1.3575,
"step": 1
},
{
"epoch": 0.0006312135079690705,
"eval_loss": 0.5412741303443909,
"eval_runtime": 238.8352,
"eval_samples_per_second": 11.171,
"eval_steps_per_second": 5.585,
"step": 1
},
{
"epoch": 0.001262427015938141,
"grad_norm": 3.5491857528686523,
"learning_rate": 1e-05,
"loss": 1.7241,
"step": 2
},
{
"epoch": 0.0018936405239072117,
"grad_norm": 3.036754846572876,
"learning_rate": 1.5e-05,
"loss": 1.6489,
"step": 3
},
{
"epoch": 0.002524854031876282,
"grad_norm": 2.9480931758880615,
"learning_rate": 2e-05,
"loss": 1.7361,
"step": 4
},
{
"epoch": 0.0031560675398453527,
"grad_norm": 2.8215577602386475,
"learning_rate": 2.5e-05,
"loss": 1.7123,
"step": 5
},
{
"epoch": 0.0037872810478144234,
"grad_norm": 2.637902021408081,
"learning_rate": 3e-05,
"loss": 1.6068,
"step": 6
},
{
"epoch": 0.004418494555783494,
"grad_norm": 2.700479745864868,
"learning_rate": 3.5e-05,
"loss": 1.6763,
"step": 7
},
{
"epoch": 0.005049708063752564,
"grad_norm": 2.643176794052124,
"learning_rate": 4e-05,
"loss": 1.6168,
"step": 8
},
{
"epoch": 0.005680921571721635,
"grad_norm": 2.6839067935943604,
"learning_rate": 4.5e-05,
"loss": 1.6763,
"step": 9
},
{
"epoch": 0.006312135079690705,
"grad_norm": 2.832852363586426,
"learning_rate": 5e-05,
"loss": 1.6942,
"step": 10
},
{
"epoch": 0.006943348587659776,
"grad_norm": 2.4192471504211426,
"learning_rate": 5.500000000000001e-05,
"loss": 1.7016,
"step": 11
},
{
"epoch": 0.007574562095628847,
"grad_norm": 2.353262424468994,
"learning_rate": 6e-05,
"loss": 1.7291,
"step": 12
},
{
"epoch": 0.008205775603597917,
"grad_norm": 2.393162965774536,
"learning_rate": 6.500000000000001e-05,
"loss": 1.7481,
"step": 13
},
{
"epoch": 0.008836989111566988,
"grad_norm": 2.343433380126953,
"learning_rate": 7e-05,
"loss": 1.7567,
"step": 14
},
{
"epoch": 0.009468202619536058,
"grad_norm": 2.266066312789917,
"learning_rate": 7.500000000000001e-05,
"loss": 1.6315,
"step": 15
},
{
"epoch": 0.010099416127505128,
"grad_norm": 2.1219441890716553,
"learning_rate": 8e-05,
"loss": 1.522,
"step": 16
},
{
"epoch": 0.0107306296354742,
"grad_norm": 2.1620051860809326,
"learning_rate": 8.5e-05,
"loss": 1.6477,
"step": 17
},
{
"epoch": 0.01136184314344327,
"grad_norm": 2.0980477333068848,
"learning_rate": 9e-05,
"loss": 1.9465,
"step": 18
},
{
"epoch": 0.011993056651412341,
"grad_norm": 2.2929768562316895,
"learning_rate": 9.5e-05,
"loss": 1.7456,
"step": 19
},
{
"epoch": 0.01262427015938141,
"grad_norm": 2.2193870544433594,
"learning_rate": 0.0001,
"loss": 1.7356,
"step": 20
},
{
"epoch": 0.01325548366735048,
"grad_norm": 2.251990556716919,
"learning_rate": 9.999238475781957e-05,
"loss": 1.6963,
"step": 21
},
{
"epoch": 0.013886697175319552,
"grad_norm": 2.1346232891082764,
"learning_rate": 9.99695413509548e-05,
"loss": 1.5908,
"step": 22
},
{
"epoch": 0.014517910683288622,
"grad_norm": 2.163315534591675,
"learning_rate": 9.99314767377287e-05,
"loss": 1.6235,
"step": 23
},
{
"epoch": 0.015149124191257694,
"grad_norm": 2.047973155975342,
"learning_rate": 9.987820251299122e-05,
"loss": 1.7013,
"step": 24
},
{
"epoch": 0.015780337699226762,
"grad_norm": 2.0005741119384766,
"learning_rate": 9.980973490458728e-05,
"loss": 1.7541,
"step": 25
},
{
"epoch": 0.016411551207195833,
"grad_norm": 2.1012520790100098,
"learning_rate": 9.972609476841367e-05,
"loss": 1.6195,
"step": 26
},
{
"epoch": 0.017042764715164905,
"grad_norm": 2.241985321044922,
"learning_rate": 9.962730758206611e-05,
"loss": 1.7854,
"step": 27
},
{
"epoch": 0.017673978223133976,
"grad_norm": 1.977676510810852,
"learning_rate": 9.951340343707852e-05,
"loss": 1.7176,
"step": 28
},
{
"epoch": 0.018305191731103045,
"grad_norm": 1.9915571212768555,
"learning_rate": 9.938441702975689e-05,
"loss": 1.8064,
"step": 29
},
{
"epoch": 0.018936405239072116,
"grad_norm": 2.151484489440918,
"learning_rate": 9.924038765061042e-05,
"loss": 2.0075,
"step": 30
},
{
"epoch": 0.019567618747041188,
"grad_norm": 1.9701048135757446,
"learning_rate": 9.908135917238321e-05,
"loss": 1.5496,
"step": 31
},
{
"epoch": 0.020198832255010256,
"grad_norm": 2.0728065967559814,
"learning_rate": 9.890738003669029e-05,
"loss": 1.7886,
"step": 32
},
{
"epoch": 0.020830045762979327,
"grad_norm": 2.117295265197754,
"learning_rate": 9.871850323926177e-05,
"loss": 1.6142,
"step": 33
},
{
"epoch": 0.0214612592709484,
"grad_norm": 1.9997711181640625,
"learning_rate": 9.851478631379982e-05,
"loss": 1.6704,
"step": 34
},
{
"epoch": 0.022092472778917467,
"grad_norm": 2.1985790729522705,
"learning_rate": 9.829629131445342e-05,
"loss": 1.9918,
"step": 35
},
{
"epoch": 0.02272368628688654,
"grad_norm": 2.025857925415039,
"learning_rate": 9.806308479691595e-05,
"loss": 1.5683,
"step": 36
},
{
"epoch": 0.02335489979485561,
"grad_norm": 2.131359338760376,
"learning_rate": 9.781523779815179e-05,
"loss": 1.8734,
"step": 37
},
{
"epoch": 0.023986113302824682,
"grad_norm": 2.0778703689575195,
"learning_rate": 9.755282581475769e-05,
"loss": 1.7188,
"step": 38
},
{
"epoch": 0.02461732681079375,
"grad_norm": 2.165473461151123,
"learning_rate": 9.727592877996585e-05,
"loss": 1.7564,
"step": 39
},
{
"epoch": 0.02524854031876282,
"grad_norm": 2.105788230895996,
"learning_rate": 9.698463103929542e-05,
"loss": 1.707,
"step": 40
},
{
"epoch": 0.025879753826731893,
"grad_norm": 2.1210577487945557,
"learning_rate": 9.667902132486009e-05,
"loss": 1.724,
"step": 41
},
{
"epoch": 0.02651096733470096,
"grad_norm": 2.0978639125823975,
"learning_rate": 9.635919272833938e-05,
"loss": 1.751,
"step": 42
},
{
"epoch": 0.027142180842670033,
"grad_norm": 2.273977279663086,
"learning_rate": 9.602524267262203e-05,
"loss": 1.6593,
"step": 43
},
{
"epoch": 0.027773394350639104,
"grad_norm": 2.122015953063965,
"learning_rate": 9.567727288213005e-05,
"loss": 1.756,
"step": 44
},
{
"epoch": 0.028404607858608173,
"grad_norm": 2.081787109375,
"learning_rate": 9.53153893518325e-05,
"loss": 1.5773,
"step": 45
},
{
"epoch": 0.029035821366577244,
"grad_norm": 2.3352270126342773,
"learning_rate": 9.493970231495835e-05,
"loss": 1.6289,
"step": 46
},
{
"epoch": 0.029667034874546316,
"grad_norm": 2.268566370010376,
"learning_rate": 9.45503262094184e-05,
"loss": 1.7125,
"step": 47
},
{
"epoch": 0.030298248382515387,
"grad_norm": 2.2713985443115234,
"learning_rate": 9.414737964294636e-05,
"loss": 1.5243,
"step": 48
},
{
"epoch": 0.030929461890484455,
"grad_norm": 2.297013282775879,
"learning_rate": 9.373098535696979e-05,
"loss": 1.4457,
"step": 49
},
{
"epoch": 0.031560675398453523,
"grad_norm": 2.792546510696411,
"learning_rate": 9.330127018922194e-05,
"loss": 1.4837,
"step": 50
},
{
"epoch": 0.031560675398453523,
"eval_loss": 0.4193588197231293,
"eval_runtime": 240.4268,
"eval_samples_per_second": 11.097,
"eval_steps_per_second": 5.548,
"step": 50
},
{
"epoch": 0.0321918889064226,
"grad_norm": 1.4785792827606201,
"learning_rate": 9.285836503510562e-05,
"loss": 0.9662,
"step": 51
},
{
"epoch": 0.03282310241439167,
"grad_norm": 1.5581904649734497,
"learning_rate": 9.24024048078213e-05,
"loss": 1.4336,
"step": 52
},
{
"epoch": 0.03345431592236074,
"grad_norm": 1.3941839933395386,
"learning_rate": 9.193352839727121e-05,
"loss": 1.2795,
"step": 53
},
{
"epoch": 0.03408552943032981,
"grad_norm": 1.5098243951797485,
"learning_rate": 9.145187862775209e-05,
"loss": 1.5065,
"step": 54
},
{
"epoch": 0.03471674293829888,
"grad_norm": 1.6207574605941772,
"learning_rate": 9.09576022144496e-05,
"loss": 1.5554,
"step": 55
},
{
"epoch": 0.03534795644626795,
"grad_norm": 1.558283805847168,
"learning_rate": 9.045084971874738e-05,
"loss": 1.6396,
"step": 56
},
{
"epoch": 0.03597916995423702,
"grad_norm": 1.633166790008545,
"learning_rate": 8.993177550236464e-05,
"loss": 1.4533,
"step": 57
},
{
"epoch": 0.03661038346220609,
"grad_norm": 1.6480422019958496,
"learning_rate": 8.940053768033609e-05,
"loss": 1.6044,
"step": 58
},
{
"epoch": 0.037241596970175164,
"grad_norm": 1.5684033632278442,
"learning_rate": 8.885729807284856e-05,
"loss": 1.6098,
"step": 59
},
{
"epoch": 0.03787281047814423,
"grad_norm": 1.6380794048309326,
"learning_rate": 8.83022221559489e-05,
"loss": 1.7237,
"step": 60
},
{
"epoch": 0.0385040239861133,
"grad_norm": 1.626305103302002,
"learning_rate": 8.773547901113862e-05,
"loss": 1.6403,
"step": 61
},
{
"epoch": 0.039135237494082376,
"grad_norm": 1.6016772985458374,
"learning_rate": 8.715724127386972e-05,
"loss": 1.6659,
"step": 62
},
{
"epoch": 0.039766451002051444,
"grad_norm": 1.6089552640914917,
"learning_rate": 8.656768508095853e-05,
"loss": 1.5893,
"step": 63
},
{
"epoch": 0.04039766451002051,
"grad_norm": 1.551110029220581,
"learning_rate": 8.596699001693255e-05,
"loss": 1.5991,
"step": 64
},
{
"epoch": 0.04102887801798959,
"grad_norm": 1.6614254713058472,
"learning_rate": 8.535533905932738e-05,
"loss": 1.6066,
"step": 65
},
{
"epoch": 0.041660091525958655,
"grad_norm": 1.6592720746994019,
"learning_rate": 8.473291852294987e-05,
"loss": 1.78,
"step": 66
},
{
"epoch": 0.04229130503392772,
"grad_norm": 1.694076418876648,
"learning_rate": 8.409991800312493e-05,
"loss": 1.7028,
"step": 67
},
{
"epoch": 0.0429225185418968,
"grad_norm": 1.687516689300537,
"learning_rate": 8.345653031794292e-05,
"loss": 1.6425,
"step": 68
},
{
"epoch": 0.043553732049865866,
"grad_norm": 1.8176817893981934,
"learning_rate": 8.280295144952536e-05,
"loss": 1.7688,
"step": 69
},
{
"epoch": 0.044184945557834934,
"grad_norm": 1.764524221420288,
"learning_rate": 8.213938048432697e-05,
"loss": 1.7655,
"step": 70
},
{
"epoch": 0.04481615906580401,
"grad_norm": 1.7133554220199585,
"learning_rate": 8.146601955249188e-05,
"loss": 1.7212,
"step": 71
},
{
"epoch": 0.04544737257377308,
"grad_norm": 1.6277440786361694,
"learning_rate": 8.07830737662829e-05,
"loss": 1.5719,
"step": 72
},
{
"epoch": 0.04607858608174215,
"grad_norm": 1.6812453269958496,
"learning_rate": 8.009075115760243e-05,
"loss": 1.8084,
"step": 73
},
{
"epoch": 0.04670979958971122,
"grad_norm": 1.72732412815094,
"learning_rate": 7.938926261462366e-05,
"loss": 1.7182,
"step": 74
},
{
"epoch": 0.04734101309768029,
"grad_norm": 1.7740263938903809,
"learning_rate": 7.86788218175523e-05,
"loss": 1.7463,
"step": 75
},
{
"epoch": 0.047972226605649364,
"grad_norm": 1.6268154382705688,
"learning_rate": 7.795964517353735e-05,
"loss": 1.511,
"step": 76
},
{
"epoch": 0.04860344011361843,
"grad_norm": 1.5691125392913818,
"learning_rate": 7.723195175075136e-05,
"loss": 1.7255,
"step": 77
},
{
"epoch": 0.0492346536215875,
"grad_norm": 1.7060211896896362,
"learning_rate": 7.649596321166024e-05,
"loss": 1.7701,
"step": 78
},
{
"epoch": 0.049865867129556575,
"grad_norm": 1.7170287370681763,
"learning_rate": 7.575190374550272e-05,
"loss": 1.8898,
"step": 79
},
{
"epoch": 0.05049708063752564,
"grad_norm": 1.6707334518432617,
"learning_rate": 7.500000000000001e-05,
"loss": 1.6891,
"step": 80
},
{
"epoch": 0.05112829414549471,
"grad_norm": 1.708794116973877,
"learning_rate": 7.424048101231686e-05,
"loss": 1.5987,
"step": 81
},
{
"epoch": 0.051759507653463786,
"grad_norm": 1.7319235801696777,
"learning_rate": 7.347357813929454e-05,
"loss": 1.5944,
"step": 82
},
{
"epoch": 0.052390721161432854,
"grad_norm": 1.7364600896835327,
"learning_rate": 7.269952498697734e-05,
"loss": 1.686,
"step": 83
},
{
"epoch": 0.05302193466940192,
"grad_norm": 1.8175327777862549,
"learning_rate": 7.191855733945387e-05,
"loss": 1.6516,
"step": 84
},
{
"epoch": 0.053653148177371,
"grad_norm": 1.837198257446289,
"learning_rate": 7.113091308703498e-05,
"loss": 1.6562,
"step": 85
},
{
"epoch": 0.054284361685340066,
"grad_norm": 1.8270783424377441,
"learning_rate": 7.033683215379002e-05,
"loss": 1.8419,
"step": 86
},
{
"epoch": 0.054915575193309134,
"grad_norm": 1.8131033182144165,
"learning_rate": 6.953655642446368e-05,
"loss": 1.5259,
"step": 87
},
{
"epoch": 0.05554678870127821,
"grad_norm": 1.8537405729293823,
"learning_rate": 6.873032967079561e-05,
"loss": 1.8515,
"step": 88
},
{
"epoch": 0.05617800220924728,
"grad_norm": 1.7982525825500488,
"learning_rate": 6.7918397477265e-05,
"loss": 1.7434,
"step": 89
},
{
"epoch": 0.056809215717216345,
"grad_norm": 1.9493149518966675,
"learning_rate": 6.710100716628344e-05,
"loss": 1.7293,
"step": 90
},
{
"epoch": 0.05744042922518542,
"grad_norm": 1.804345726966858,
"learning_rate": 6.627840772285784e-05,
"loss": 1.5746,
"step": 91
},
{
"epoch": 0.05807164273315449,
"grad_norm": 1.8308101892471313,
"learning_rate": 6.545084971874738e-05,
"loss": 1.4938,
"step": 92
},
{
"epoch": 0.05870285624112356,
"grad_norm": 1.9381098747253418,
"learning_rate": 6.461858523613684e-05,
"loss": 1.662,
"step": 93
},
{
"epoch": 0.05933406974909263,
"grad_norm": 1.9891799688339233,
"learning_rate": 6.378186779084995e-05,
"loss": 1.5954,
"step": 94
},
{
"epoch": 0.0599652832570617,
"grad_norm": 1.9715620279312134,
"learning_rate": 6.294095225512603e-05,
"loss": 1.3628,
"step": 95
},
{
"epoch": 0.060596496765030775,
"grad_norm": 2.135878562927246,
"learning_rate": 6.209609477998338e-05,
"loss": 1.658,
"step": 96
},
{
"epoch": 0.06122771027299984,
"grad_norm": 1.9819763898849487,
"learning_rate": 6.124755271719325e-05,
"loss": 1.4918,
"step": 97
},
{
"epoch": 0.06185892378096891,
"grad_norm": 2.1434459686279297,
"learning_rate": 6.0395584540887963e-05,
"loss": 1.614,
"step": 98
},
{
"epoch": 0.062490137288937986,
"grad_norm": 2.191251516342163,
"learning_rate": 5.9540449768827246e-05,
"loss": 1.4206,
"step": 99
},
{
"epoch": 0.06312135079690705,
"grad_norm": 2.3877665996551514,
"learning_rate": 5.868240888334653e-05,
"loss": 1.2033,
"step": 100
},
{
"epoch": 0.06312135079690705,
"eval_loss": 0.40419265627861023,
"eval_runtime": 240.7348,
"eval_samples_per_second": 11.083,
"eval_steps_per_second": 5.541,
"step": 100
},
{
"epoch": 0.06375256430487612,
"grad_norm": 1.128631830215454,
"learning_rate": 5.782172325201155e-05,
"loss": 0.9465,
"step": 101
},
{
"epoch": 0.0643837778128452,
"grad_norm": 1.3276664018630981,
"learning_rate": 5.695865504800327e-05,
"loss": 1.3996,
"step": 102
},
{
"epoch": 0.06501499132081427,
"grad_norm": 1.2908798456192017,
"learning_rate": 5.6093467170257374e-05,
"loss": 1.2058,
"step": 103
},
{
"epoch": 0.06564620482878333,
"grad_norm": 1.2702507972717285,
"learning_rate": 5.522642316338268e-05,
"loss": 1.3292,
"step": 104
},
{
"epoch": 0.06627741833675241,
"grad_norm": 1.4327809810638428,
"learning_rate": 5.435778713738292e-05,
"loss": 1.7259,
"step": 105
},
{
"epoch": 0.06690863184472148,
"grad_norm": 1.3425438404083252,
"learning_rate": 5.348782368720626e-05,
"loss": 1.4477,
"step": 106
},
{
"epoch": 0.06753984535269054,
"grad_norm": 1.397606611251831,
"learning_rate": 5.26167978121472e-05,
"loss": 1.4911,
"step": 107
},
{
"epoch": 0.06817105886065962,
"grad_norm": 1.3934853076934814,
"learning_rate": 5.174497483512506e-05,
"loss": 1.3199,
"step": 108
},
{
"epoch": 0.0688022723686287,
"grad_norm": 1.4510592222213745,
"learning_rate": 5.0872620321864185e-05,
"loss": 1.6421,
"step": 109
},
{
"epoch": 0.06943348587659776,
"grad_norm": 1.425351619720459,
"learning_rate": 5e-05,
"loss": 1.5701,
"step": 110
},
{
"epoch": 0.07006469938456683,
"grad_norm": 1.503740906715393,
"learning_rate": 4.912737967813583e-05,
"loss": 1.4684,
"step": 111
},
{
"epoch": 0.0706959128925359,
"grad_norm": 1.4400185346603394,
"learning_rate": 4.825502516487497e-05,
"loss": 1.423,
"step": 112
},
{
"epoch": 0.07132712640050497,
"grad_norm": 1.5304445028305054,
"learning_rate": 4.738320218785281e-05,
"loss": 1.6053,
"step": 113
},
{
"epoch": 0.07195833990847404,
"grad_norm": 1.4800596237182617,
"learning_rate": 4.6512176312793736e-05,
"loss": 1.6218,
"step": 114
},
{
"epoch": 0.07258955341644312,
"grad_norm": 1.630185604095459,
"learning_rate": 4.564221286261709e-05,
"loss": 1.8255,
"step": 115
},
{
"epoch": 0.07322076692441218,
"grad_norm": 1.5210514068603516,
"learning_rate": 4.477357683661734e-05,
"loss": 1.5067,
"step": 116
},
{
"epoch": 0.07385198043238125,
"grad_norm": 1.559752345085144,
"learning_rate": 4.390653282974264e-05,
"loss": 1.5384,
"step": 117
},
{
"epoch": 0.07448319394035033,
"grad_norm": 1.586678147315979,
"learning_rate": 4.3041344951996746e-05,
"loss": 1.5304,
"step": 118
},
{
"epoch": 0.07511440744831939,
"grad_norm": 1.4632230997085571,
"learning_rate": 4.2178276747988446e-05,
"loss": 1.5951,
"step": 119
},
{
"epoch": 0.07574562095628846,
"grad_norm": 1.5816446542739868,
"learning_rate": 4.131759111665349e-05,
"loss": 1.7446,
"step": 120
},
{
"epoch": 0.07637683446425754,
"grad_norm": 1.559669852256775,
"learning_rate": 4.045955023117276e-05,
"loss": 1.5426,
"step": 121
},
{
"epoch": 0.0770080479722266,
"grad_norm": 1.6283562183380127,
"learning_rate": 3.960441545911204e-05,
"loss": 1.8152,
"step": 122
},
{
"epoch": 0.07763926148019568,
"grad_norm": 1.5507304668426514,
"learning_rate": 3.875244728280676e-05,
"loss": 1.5015,
"step": 123
},
{
"epoch": 0.07827047498816475,
"grad_norm": 1.5461734533309937,
"learning_rate": 3.790390522001662e-05,
"loss": 1.5926,
"step": 124
},
{
"epoch": 0.07890168849613381,
"grad_norm": 1.6004427671432495,
"learning_rate": 3.705904774487396e-05,
"loss": 1.5735,
"step": 125
},
{
"epoch": 0.07953290200410289,
"grad_norm": 1.6594659090042114,
"learning_rate": 3.6218132209150045e-05,
"loss": 1.6118,
"step": 126
},
{
"epoch": 0.08016411551207196,
"grad_norm": 1.65187406539917,
"learning_rate": 3.5381414763863166e-05,
"loss": 1.6812,
"step": 127
},
{
"epoch": 0.08079532902004102,
"grad_norm": 1.6164668798446655,
"learning_rate": 3.4549150281252636e-05,
"loss": 1.6003,
"step": 128
},
{
"epoch": 0.0814265425280101,
"grad_norm": 1.607835054397583,
"learning_rate": 3.372159227714218e-05,
"loss": 1.526,
"step": 129
},
{
"epoch": 0.08205775603597917,
"grad_norm": 1.707564353942871,
"learning_rate": 3.289899283371657e-05,
"loss": 1.7708,
"step": 130
},
{
"epoch": 0.08268896954394823,
"grad_norm": 1.6170071363449097,
"learning_rate": 3.2081602522734986e-05,
"loss": 1.702,
"step": 131
},
{
"epoch": 0.08332018305191731,
"grad_norm": 1.6684552431106567,
"learning_rate": 3.12696703292044e-05,
"loss": 1.5426,
"step": 132
},
{
"epoch": 0.08395139655988638,
"grad_norm": 1.6673823595046997,
"learning_rate": 3.046344357553632e-05,
"loss": 1.6907,
"step": 133
},
{
"epoch": 0.08458261006785545,
"grad_norm": 1.6971157789230347,
"learning_rate": 2.9663167846209998e-05,
"loss": 1.6752,
"step": 134
},
{
"epoch": 0.08521382357582452,
"grad_norm": 1.725826382637024,
"learning_rate": 2.886908691296504e-05,
"loss": 1.6407,
"step": 135
},
{
"epoch": 0.0858450370837936,
"grad_norm": 1.7856934070587158,
"learning_rate": 2.8081442660546125e-05,
"loss": 1.6094,
"step": 136
},
{
"epoch": 0.08647625059176266,
"grad_norm": 1.669678807258606,
"learning_rate": 2.7300475013022663e-05,
"loss": 1.6688,
"step": 137
},
{
"epoch": 0.08710746409973173,
"grad_norm": 1.6383804082870483,
"learning_rate": 2.6526421860705473e-05,
"loss": 1.4347,
"step": 138
},
{
"epoch": 0.08773867760770081,
"grad_norm": 1.760483741760254,
"learning_rate": 2.575951898768315e-05,
"loss": 1.6037,
"step": 139
},
{
"epoch": 0.08836989111566987,
"grad_norm": 1.814774513244629,
"learning_rate": 2.500000000000001e-05,
"loss": 1.8616,
"step": 140
},
{
"epoch": 0.08900110462363894,
"grad_norm": 1.8109049797058105,
"learning_rate": 2.4248096254497288e-05,
"loss": 1.6744,
"step": 141
},
{
"epoch": 0.08963231813160802,
"grad_norm": 1.82987380027771,
"learning_rate": 2.350403678833976e-05,
"loss": 1.7784,
"step": 142
},
{
"epoch": 0.0902635316395771,
"grad_norm": 1.7920920848846436,
"learning_rate": 2.2768048249248648e-05,
"loss": 1.701,
"step": 143
},
{
"epoch": 0.09089474514754615,
"grad_norm": 1.8180540800094604,
"learning_rate": 2.2040354826462668e-05,
"loss": 1.5727,
"step": 144
},
{
"epoch": 0.09152595865551523,
"grad_norm": 1.8675249814987183,
"learning_rate": 2.132117818244771e-05,
"loss": 1.4887,
"step": 145
},
{
"epoch": 0.0921571721634843,
"grad_norm": 2.0394644737243652,
"learning_rate": 2.061073738537635e-05,
"loss": 1.6597,
"step": 146
},
{
"epoch": 0.09278838567145337,
"grad_norm": 1.9944242238998413,
"learning_rate": 1.9909248842397584e-05,
"loss": 1.6877,
"step": 147
},
{
"epoch": 0.09341959917942244,
"grad_norm": 1.9664406776428223,
"learning_rate": 1.9216926233717085e-05,
"loss": 1.3912,
"step": 148
},
{
"epoch": 0.09405081268739152,
"grad_norm": 2.1415047645568848,
"learning_rate": 1.8533980447508137e-05,
"loss": 1.4632,
"step": 149
},
{
"epoch": 0.09468202619536058,
"grad_norm": 2.2631072998046875,
"learning_rate": 1.7860619515673033e-05,
"loss": 1.3699,
"step": 150
},
{
"epoch": 0.09468202619536058,
"eval_loss": 0.3925303518772125,
"eval_runtime": 240.523,
"eval_samples_per_second": 11.092,
"eval_steps_per_second": 5.546,
"step": 150
},
{
"epoch": 0.09531323970332965,
"grad_norm": 0.9829754829406738,
"learning_rate": 1.7197048550474643e-05,
"loss": 0.9275,
"step": 151
},
{
"epoch": 0.09594445321129873,
"grad_norm": 1.0895384550094604,
"learning_rate": 1.6543469682057106e-05,
"loss": 1.1512,
"step": 152
},
{
"epoch": 0.09657566671926779,
"grad_norm": 1.1453516483306885,
"learning_rate": 1.5900081996875083e-05,
"loss": 1.2615,
"step": 153
},
{
"epoch": 0.09720688022723686,
"grad_norm": 1.2454159259796143,
"learning_rate": 1.526708147705013e-05,
"loss": 1.4501,
"step": 154
},
{
"epoch": 0.09783809373520594,
"grad_norm": 1.273112177848816,
"learning_rate": 1.4644660940672627e-05,
"loss": 1.3519,
"step": 155
},
{
"epoch": 0.098469307243175,
"grad_norm": 1.3017650842666626,
"learning_rate": 1.4033009983067452e-05,
"loss": 1.4724,
"step": 156
},
{
"epoch": 0.09910052075114407,
"grad_norm": 1.2280182838439941,
"learning_rate": 1.3432314919041478e-05,
"loss": 1.2936,
"step": 157
},
{
"epoch": 0.09973173425911315,
"grad_norm": 1.365878701210022,
"learning_rate": 1.2842758726130283e-05,
"loss": 1.4915,
"step": 158
},
{
"epoch": 0.10036294776708221,
"grad_norm": 1.3801015615463257,
"learning_rate": 1.22645209888614e-05,
"loss": 1.4224,
"step": 159
},
{
"epoch": 0.10099416127505129,
"grad_norm": 1.328890085220337,
"learning_rate": 1.1697777844051105e-05,
"loss": 1.3808,
"step": 160
},
{
"epoch": 0.10162537478302036,
"grad_norm": 1.4287265539169312,
"learning_rate": 1.1142701927151456e-05,
"loss": 1.6477,
"step": 161
},
{
"epoch": 0.10225658829098942,
"grad_norm": 1.3774094581604004,
"learning_rate": 1.0599462319663905e-05,
"loss": 1.5419,
"step": 162
},
{
"epoch": 0.1028878017989585,
"grad_norm": 1.4486933946609497,
"learning_rate": 1.006822449763537e-05,
"loss": 1.5195,
"step": 163
},
{
"epoch": 0.10351901530692757,
"grad_norm": 1.4375733137130737,
"learning_rate": 9.549150281252633e-06,
"loss": 1.7109,
"step": 164
},
{
"epoch": 0.10415022881489663,
"grad_norm": 1.4317312240600586,
"learning_rate": 9.042397785550405e-06,
"loss": 1.4966,
"step": 165
},
{
"epoch": 0.10478144232286571,
"grad_norm": 1.5089269876480103,
"learning_rate": 8.548121372247918e-06,
"loss": 1.5649,
"step": 166
},
{
"epoch": 0.10541265583083478,
"grad_norm": 1.5436687469482422,
"learning_rate": 8.066471602728803e-06,
"loss": 1.6303,
"step": 167
},
{
"epoch": 0.10604386933880385,
"grad_norm": 1.5016186237335205,
"learning_rate": 7.597595192178702e-06,
"loss": 1.6437,
"step": 168
},
{
"epoch": 0.10667508284677292,
"grad_norm": 1.5110390186309814,
"learning_rate": 7.1416349648943894e-06,
"loss": 1.5768,
"step": 169
},
{
"epoch": 0.107306296354742,
"grad_norm": 1.5523266792297363,
"learning_rate": 6.698729810778065e-06,
"loss": 1.7035,
"step": 170
},
{
"epoch": 0.10793750986271106,
"grad_norm": 1.5230318307876587,
"learning_rate": 6.269014643030213e-06,
"loss": 1.5579,
"step": 171
},
{
"epoch": 0.10856872337068013,
"grad_norm": 1.508470892906189,
"learning_rate": 5.852620357053651e-06,
"loss": 1.5116,
"step": 172
},
{
"epoch": 0.1091999368786492,
"grad_norm": 1.504398226737976,
"learning_rate": 5.449673790581611e-06,
"loss": 1.4858,
"step": 173
},
{
"epoch": 0.10983115038661827,
"grad_norm": 1.4883153438568115,
"learning_rate": 5.060297685041659e-06,
"loss": 1.7057,
"step": 174
},
{
"epoch": 0.11046236389458734,
"grad_norm": 1.4853483438491821,
"learning_rate": 4.684610648167503e-06,
"loss": 1.5148,
"step": 175
},
{
"epoch": 0.11109357740255642,
"grad_norm": 1.553637981414795,
"learning_rate": 4.322727117869951e-06,
"loss": 1.6167,
"step": 176
},
{
"epoch": 0.11172479091052548,
"grad_norm": 1.578129529953003,
"learning_rate": 3.974757327377981e-06,
"loss": 1.5491,
"step": 177
},
{
"epoch": 0.11235600441849455,
"grad_norm": 1.5624874830245972,
"learning_rate": 3.6408072716606346e-06,
"loss": 1.6006,
"step": 178
},
{
"epoch": 0.11298721792646363,
"grad_norm": 1.5715187788009644,
"learning_rate": 3.3209786751399187e-06,
"loss": 1.6976,
"step": 179
},
{
"epoch": 0.11361843143443269,
"grad_norm": 1.6272014379501343,
"learning_rate": 3.0153689607045845e-06,
"loss": 1.6757,
"step": 180
},
{
"epoch": 0.11424964494240177,
"grad_norm": 1.546974539756775,
"learning_rate": 2.724071220034158e-06,
"loss": 1.5974,
"step": 181
},
{
"epoch": 0.11488085845037084,
"grad_norm": 1.6230578422546387,
"learning_rate": 2.4471741852423237e-06,
"loss": 1.6203,
"step": 182
},
{
"epoch": 0.11551207195833992,
"grad_norm": 1.5810596942901611,
"learning_rate": 2.1847622018482283e-06,
"loss": 1.5336,
"step": 183
},
{
"epoch": 0.11614328546630898,
"grad_norm": 1.7239481210708618,
"learning_rate": 1.9369152030840556e-06,
"loss": 1.676,
"step": 184
},
{
"epoch": 0.11677449897427805,
"grad_norm": 1.6532574892044067,
"learning_rate": 1.70370868554659e-06,
"loss": 1.8418,
"step": 185
},
{
"epoch": 0.11740571248224713,
"grad_norm": 1.6952757835388184,
"learning_rate": 1.4852136862001764e-06,
"loss": 1.5633,
"step": 186
},
{
"epoch": 0.11803692599021619,
"grad_norm": 1.7153465747833252,
"learning_rate": 1.2814967607382432e-06,
"loss": 1.6121,
"step": 187
},
{
"epoch": 0.11866813949818526,
"grad_norm": 1.8238195180892944,
"learning_rate": 1.0926199633097157e-06,
"loss": 1.8494,
"step": 188
},
{
"epoch": 0.11929935300615434,
"grad_norm": 1.7845555543899536,
"learning_rate": 9.186408276168013e-07,
"loss": 1.6834,
"step": 189
},
{
"epoch": 0.1199305665141234,
"grad_norm": 1.8193535804748535,
"learning_rate": 7.596123493895991e-07,
"loss": 1.624,
"step": 190
},
{
"epoch": 0.12056178002209247,
"grad_norm": 1.7187602519989014,
"learning_rate": 6.15582970243117e-07,
"loss": 1.6485,
"step": 191
},
{
"epoch": 0.12119299353006155,
"grad_norm": 1.9166899919509888,
"learning_rate": 4.865965629214819e-07,
"loss": 1.8563,
"step": 192
},
{
"epoch": 0.12182420703803061,
"grad_norm": 1.856611728668213,
"learning_rate": 3.7269241793390085e-07,
"loss": 1.7069,
"step": 193
},
{
"epoch": 0.12245542054599969,
"grad_norm": 1.9729714393615723,
"learning_rate": 2.7390523158633554e-07,
"loss": 1.6582,
"step": 194
},
{
"epoch": 0.12308663405396876,
"grad_norm": 1.8488212823867798,
"learning_rate": 1.9026509541272275e-07,
"loss": 1.5962,
"step": 195
},
{
"epoch": 0.12371784756193782,
"grad_norm": 1.8829165697097778,
"learning_rate": 1.2179748700879012e-07,
"loss": 1.6283,
"step": 196
},
{
"epoch": 0.1243490610699069,
"grad_norm": 2.0195155143737793,
"learning_rate": 6.852326227130834e-08,
"loss": 1.5913,
"step": 197
},
{
"epoch": 0.12498027457787597,
"grad_norm": 1.976454257965088,
"learning_rate": 3.04586490452119e-08,
"loss": 1.4226,
"step": 198
},
{
"epoch": 0.12561148808584505,
"grad_norm": 2.081446647644043,
"learning_rate": 7.615242180436522e-09,
"loss": 1.3406,
"step": 199
},
{
"epoch": 0.1262427015938141,
"grad_norm": 2.4766805171966553,
"learning_rate": 0.0,
"loss": 1.5761,
"step": 200
},
{
"epoch": 0.1262427015938141,
"eval_loss": 0.38890671730041504,
"eval_runtime": 240.7623,
"eval_samples_per_second": 11.081,
"eval_steps_per_second": 5.541,
"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": 3.194703055064924e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}