{
"best_metric": 1.8370285034179688,
"best_model_checkpoint": "miner_id_24/checkpoint-200",
"epoch": 0.02939231390991256,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0001469615695495628,
"grad_norm": 1.3901768922805786,
"learning_rate": 1e-05,
"loss": 2.1107,
"step": 1
},
{
"epoch": 0.0001469615695495628,
"eval_loss": 2.4028263092041016,
"eval_runtime": 239.1891,
"eval_samples_per_second": 47.916,
"eval_steps_per_second": 11.982,
"step": 1
},
{
"epoch": 0.0002939231390991256,
"grad_norm": 1.6152689456939697,
"learning_rate": 2e-05,
"loss": 2.2268,
"step": 2
},
{
"epoch": 0.00044088470864868834,
"grad_norm": 1.7118444442749023,
"learning_rate": 3e-05,
"loss": 2.28,
"step": 3
},
{
"epoch": 0.0005878462781982512,
"grad_norm": 1.6755433082580566,
"learning_rate": 4e-05,
"loss": 2.199,
"step": 4
},
{
"epoch": 0.000734807847747814,
"grad_norm": 1.6319252252578735,
"learning_rate": 5e-05,
"loss": 2.3356,
"step": 5
},
{
"epoch": 0.0008817694172973767,
"grad_norm": 1.361087679862976,
"learning_rate": 6e-05,
"loss": 2.2309,
"step": 6
},
{
"epoch": 0.0010287309868469395,
"grad_norm": 1.3249088525772095,
"learning_rate": 7e-05,
"loss": 2.1566,
"step": 7
},
{
"epoch": 0.0011756925563965023,
"grad_norm": 1.266327142715454,
"learning_rate": 8e-05,
"loss": 2.0796,
"step": 8
},
{
"epoch": 0.0013226541259460651,
"grad_norm": 1.32285737991333,
"learning_rate": 9e-05,
"loss": 2.1211,
"step": 9
},
{
"epoch": 0.001469615695495628,
"grad_norm": 1.3480980396270752,
"learning_rate": 0.0001,
"loss": 1.9753,
"step": 10
},
{
"epoch": 0.0016165772650451908,
"grad_norm": 1.3346507549285889,
"learning_rate": 9.999316524962345e-05,
"loss": 1.9767,
"step": 11
},
{
"epoch": 0.0017635388345947534,
"grad_norm": 1.3289947509765625,
"learning_rate": 9.997266286704631e-05,
"loss": 2.019,
"step": 12
},
{
"epoch": 0.0019105004041443162,
"grad_norm": 1.2248716354370117,
"learning_rate": 9.993849845741524e-05,
"loss": 2.0302,
"step": 13
},
{
"epoch": 0.002057461973693879,
"grad_norm": 1.2586091756820679,
"learning_rate": 9.989068136093873e-05,
"loss": 1.9915,
"step": 14
},
{
"epoch": 0.002204423543243442,
"grad_norm": 1.3422409296035767,
"learning_rate": 9.98292246503335e-05,
"loss": 2.0409,
"step": 15
},
{
"epoch": 0.0023513851127930046,
"grad_norm": 1.2683439254760742,
"learning_rate": 9.975414512725057e-05,
"loss": 2.0911,
"step": 16
},
{
"epoch": 0.0024983466823425672,
"grad_norm": 1.327904462814331,
"learning_rate": 9.966546331768191e-05,
"loss": 2.2103,
"step": 17
},
{
"epoch": 0.0026453082518921303,
"grad_norm": 1.352623701095581,
"learning_rate": 9.956320346634876e-05,
"loss": 2.092,
"step": 18
},
{
"epoch": 0.002792269821441693,
"grad_norm": 1.326048493385315,
"learning_rate": 9.944739353007344e-05,
"loss": 1.8681,
"step": 19
},
{
"epoch": 0.002939231390991256,
"grad_norm": 1.2482925653457642,
"learning_rate": 9.931806517013612e-05,
"loss": 2.1253,
"step": 20
},
{
"epoch": 0.0030861929605408185,
"grad_norm": 1.2479603290557861,
"learning_rate": 9.917525374361912e-05,
"loss": 1.7682,
"step": 21
},
{
"epoch": 0.0032331545300903815,
"grad_norm": 1.3151602745056152,
"learning_rate": 9.901899829374047e-05,
"loss": 1.8526,
"step": 22
},
{
"epoch": 0.003380116099639944,
"grad_norm": 1.2352029085159302,
"learning_rate": 9.884934153917997e-05,
"loss": 1.8497,
"step": 23
},
{
"epoch": 0.0035270776691895067,
"grad_norm": 1.5575189590454102,
"learning_rate": 9.86663298624003e-05,
"loss": 1.9744,
"step": 24
},
{
"epoch": 0.00367403923873907,
"grad_norm": 1.486586570739746,
"learning_rate": 9.847001329696653e-05,
"loss": 1.9153,
"step": 25
},
{
"epoch": 0.0038210008082886324,
"grad_norm": 1.6065561771392822,
"learning_rate": 9.826044551386744e-05,
"loss": 2.1219,
"step": 26
},
{
"epoch": 0.003967962377838195,
"grad_norm": 1.379812240600586,
"learning_rate": 9.803768380684242e-05,
"loss": 1.8542,
"step": 27
},
{
"epoch": 0.004114923947387758,
"grad_norm": 1.4364986419677734,
"learning_rate": 9.780178907671789e-05,
"loss": 1.8688,
"step": 28
},
{
"epoch": 0.004261885516937321,
"grad_norm": 1.5832526683807373,
"learning_rate": 9.755282581475769e-05,
"loss": 1.8082,
"step": 29
},
{
"epoch": 0.004408847086486884,
"grad_norm": 1.55263352394104,
"learning_rate": 9.729086208503174e-05,
"loss": 1.8751,
"step": 30
},
{
"epoch": 0.004555808656036446,
"grad_norm": 1.5681918859481812,
"learning_rate": 9.701596950580806e-05,
"loss": 1.9212,
"step": 31
},
{
"epoch": 0.004702770225586009,
"grad_norm": 1.5252001285552979,
"learning_rate": 9.672822322997305e-05,
"loss": 1.6804,
"step": 32
},
{
"epoch": 0.004849731795135572,
"grad_norm": 1.8776458501815796,
"learning_rate": 9.642770192448536e-05,
"loss": 1.7954,
"step": 33
},
{
"epoch": 0.0049966933646851345,
"grad_norm": 1.6283355951309204,
"learning_rate": 9.611448774886924e-05,
"loss": 1.7365,
"step": 34
},
{
"epoch": 0.0051436549342346975,
"grad_norm": 1.767272710800171,
"learning_rate": 9.578866633275288e-05,
"loss": 1.9975,
"step": 35
},
{
"epoch": 0.0052906165037842606,
"grad_norm": 1.9818291664123535,
"learning_rate": 9.545032675245813e-05,
"loss": 1.8694,
"step": 36
},
{
"epoch": 0.005437578073333824,
"grad_norm": 1.893059492111206,
"learning_rate": 9.509956150664796e-05,
"loss": 1.61,
"step": 37
},
{
"epoch": 0.005584539642883386,
"grad_norm": 1.8232024908065796,
"learning_rate": 9.473646649103818e-05,
"loss": 1.7359,
"step": 38
},
{
"epoch": 0.005731501212432949,
"grad_norm": 1.8515968322753906,
"learning_rate": 9.43611409721806e-05,
"loss": 1.7162,
"step": 39
},
{
"epoch": 0.005878462781982512,
"grad_norm": 1.9406285285949707,
"learning_rate": 9.397368756032445e-05,
"loss": 1.7408,
"step": 40
},
{
"epoch": 0.006025424351532074,
"grad_norm": 2.118736743927002,
"learning_rate": 9.357421218136386e-05,
"loss": 1.7957,
"step": 41
},
{
"epoch": 0.006172385921081637,
"grad_norm": 1.9897270202636719,
"learning_rate": 9.316282404787871e-05,
"loss": 1.6664,
"step": 42
},
{
"epoch": 0.0063193474906312,
"grad_norm": 2.087229013442993,
"learning_rate": 9.273963562927695e-05,
"loss": 1.8842,
"step": 43
},
{
"epoch": 0.006466309060180763,
"grad_norm": 2.176999092102051,
"learning_rate": 9.230476262104677e-05,
"loss": 1.7334,
"step": 44
},
{
"epoch": 0.006613270629730325,
"grad_norm": 2.1053779125213623,
"learning_rate": 9.185832391312644e-05,
"loss": 1.695,
"step": 45
},
{
"epoch": 0.006760232199279888,
"grad_norm": 2.430379629135132,
"learning_rate": 9.140044155740101e-05,
"loss": 1.8374,
"step": 46
},
{
"epoch": 0.006907193768829451,
"grad_norm": 2.8678176403045654,
"learning_rate": 9.093124073433463e-05,
"loss": 1.6383,
"step": 47
},
{
"epoch": 0.0070541553383790135,
"grad_norm": 2.872281074523926,
"learning_rate": 9.045084971874738e-05,
"loss": 1.5129,
"step": 48
},
{
"epoch": 0.0072011169079285765,
"grad_norm": 2.9773223400115967,
"learning_rate": 8.995939984474624e-05,
"loss": 1.5495,
"step": 49
},
{
"epoch": 0.00734807847747814,
"grad_norm": 4.707193374633789,
"learning_rate": 8.945702546981969e-05,
"loss": 1.8149,
"step": 50
},
{
"epoch": 0.00734807847747814,
"eval_loss": 2.041609048843384,
"eval_runtime": 240.6928,
"eval_samples_per_second": 47.617,
"eval_steps_per_second": 11.907,
"step": 50
},
{
"epoch": 0.007495040047027703,
"grad_norm": 1.8746854066848755,
"learning_rate": 8.894386393810563e-05,
"loss": 2.2614,
"step": 51
},
{
"epoch": 0.007642001616577265,
"grad_norm": 2.163214683532715,
"learning_rate": 8.842005554284296e-05,
"loss": 2.3014,
"step": 52
},
{
"epoch": 0.007788963186126828,
"grad_norm": 1.9265990257263184,
"learning_rate": 8.788574348801675e-05,
"loss": 2.3692,
"step": 53
},
{
"epoch": 0.00793592475567639,
"grad_norm": 1.5063375234603882,
"learning_rate": 8.73410738492077e-05,
"loss": 2.2857,
"step": 54
},
{
"epoch": 0.008082886325225953,
"grad_norm": 1.1396548748016357,
"learning_rate": 8.678619553365659e-05,
"loss": 2.136,
"step": 55
},
{
"epoch": 0.008229847894775516,
"grad_norm": 1.0275360345840454,
"learning_rate": 8.622126023955446e-05,
"loss": 2.0337,
"step": 56
},
{
"epoch": 0.008376809464325079,
"grad_norm": 1.0990287065505981,
"learning_rate": 8.564642241456986e-05,
"loss": 2.2272,
"step": 57
},
{
"epoch": 0.008523771033874642,
"grad_norm": 1.1163265705108643,
"learning_rate": 8.506183921362443e-05,
"loss": 2.0692,
"step": 58
},
{
"epoch": 0.008670732603424205,
"grad_norm": 1.1702890396118164,
"learning_rate": 8.44676704559283e-05,
"loss": 1.9862,
"step": 59
},
{
"epoch": 0.008817694172973768,
"grad_norm": 1.1845659017562866,
"learning_rate": 8.386407858128706e-05,
"loss": 2.0729,
"step": 60
},
{
"epoch": 0.00896465574252333,
"grad_norm": 1.1590219736099243,
"learning_rate": 8.32512286056924e-05,
"loss": 2.0301,
"step": 61
},
{
"epoch": 0.009111617312072893,
"grad_norm": 1.128943681716919,
"learning_rate": 8.262928807620843e-05,
"loss": 1.9868,
"step": 62
},
{
"epoch": 0.009258578881622456,
"grad_norm": 1.0628398656845093,
"learning_rate": 8.199842702516583e-05,
"loss": 1.911,
"step": 63
},
{
"epoch": 0.009405540451172019,
"grad_norm": 1.0641523599624634,
"learning_rate": 8.135881792367686e-05,
"loss": 1.9107,
"step": 64
},
{
"epoch": 0.009552502020721582,
"grad_norm": 1.119457483291626,
"learning_rate": 8.07106356344834e-05,
"loss": 2.0692,
"step": 65
},
{
"epoch": 0.009699463590271145,
"grad_norm": 1.098373293876648,
"learning_rate": 8.005405736415126e-05,
"loss": 1.9411,
"step": 66
},
{
"epoch": 0.009846425159820708,
"grad_norm": 1.118204116821289,
"learning_rate": 7.938926261462366e-05,
"loss": 1.8076,
"step": 67
},
{
"epoch": 0.009993386729370269,
"grad_norm": 1.1174434423446655,
"learning_rate": 7.871643313414718e-05,
"loss": 1.8735,
"step": 68
},
{
"epoch": 0.010140348298919832,
"grad_norm": 1.1931475400924683,
"learning_rate": 7.803575286758364e-05,
"loss": 1.8814,
"step": 69
},
{
"epoch": 0.010287309868469395,
"grad_norm": 1.2271125316619873,
"learning_rate": 7.734740790612136e-05,
"loss": 1.8281,
"step": 70
},
{
"epoch": 0.010434271438018958,
"grad_norm": 1.19547700881958,
"learning_rate": 7.66515864363997e-05,
"loss": 1.8666,
"step": 71
},
{
"epoch": 0.010581233007568521,
"grad_norm": 1.3930631875991821,
"learning_rate": 7.594847868906076e-05,
"loss": 1.8231,
"step": 72
},
{
"epoch": 0.010728194577118084,
"grad_norm": 1.2994108200073242,
"learning_rate": 7.52382768867422e-05,
"loss": 1.8487,
"step": 73
},
{
"epoch": 0.010875156146667647,
"grad_norm": 1.3405499458312988,
"learning_rate": 7.452117519152542e-05,
"loss": 1.8949,
"step": 74
},
{
"epoch": 0.011022117716217208,
"grad_norm": 1.2351549863815308,
"learning_rate": 7.379736965185368e-05,
"loss": 1.874,
"step": 75
},
{
"epoch": 0.011169079285766772,
"grad_norm": 1.181893229484558,
"learning_rate": 7.30670581489344e-05,
"loss": 1.7945,
"step": 76
},
{
"epoch": 0.011316040855316335,
"grad_norm": 1.2529072761535645,
"learning_rate": 7.233044034264034e-05,
"loss": 1.7906,
"step": 77
},
{
"epoch": 0.011463002424865898,
"grad_norm": 1.3317253589630127,
"learning_rate": 7.158771761692464e-05,
"loss": 1.9559,
"step": 78
},
{
"epoch": 0.01160996399441546,
"grad_norm": 1.4497019052505493,
"learning_rate": 7.083909302476453e-05,
"loss": 1.8443,
"step": 79
},
{
"epoch": 0.011756925563965024,
"grad_norm": 1.3831413984298706,
"learning_rate": 7.008477123264848e-05,
"loss": 1.9285,
"step": 80
},
{
"epoch": 0.011903887133514587,
"grad_norm": 1.3806034326553345,
"learning_rate": 6.932495846462261e-05,
"loss": 1.7688,
"step": 81
},
{
"epoch": 0.012050848703064148,
"grad_norm": 1.3095686435699463,
"learning_rate": 6.855986244591104e-05,
"loss": 1.7448,
"step": 82
},
{
"epoch": 0.012197810272613711,
"grad_norm": 1.4687689542770386,
"learning_rate": 6.778969234612584e-05,
"loss": 1.7486,
"step": 83
},
{
"epoch": 0.012344771842163274,
"grad_norm": 1.4057215452194214,
"learning_rate": 6.701465872208216e-05,
"loss": 1.5815,
"step": 84
},
{
"epoch": 0.012491733411712837,
"grad_norm": 1.5889978408813477,
"learning_rate": 6.623497346023418e-05,
"loss": 1.6297,
"step": 85
},
{
"epoch": 0.0126386949812624,
"grad_norm": 1.569990873336792,
"learning_rate": 6.545084971874738e-05,
"loss": 1.6882,
"step": 86
},
{
"epoch": 0.012785656550811963,
"grad_norm": 1.5704505443572998,
"learning_rate": 6.466250186922325e-05,
"loss": 1.711,
"step": 87
},
{
"epoch": 0.012932618120361526,
"grad_norm": 1.4888503551483154,
"learning_rate": 6.387014543809223e-05,
"loss": 1.6201,
"step": 88
},
{
"epoch": 0.013079579689911087,
"grad_norm": 1.7181146144866943,
"learning_rate": 6.307399704769099e-05,
"loss": 1.6051,
"step": 89
},
{
"epoch": 0.01322654125946065,
"grad_norm": 1.6352920532226562,
"learning_rate": 6.227427435703997e-05,
"loss": 1.3659,
"step": 90
},
{
"epoch": 0.013373502829010214,
"grad_norm": 2.094393730163574,
"learning_rate": 6.147119600233758e-05,
"loss": 1.8436,
"step": 91
},
{
"epoch": 0.013520464398559777,
"grad_norm": 1.9679774045944214,
"learning_rate": 6.066498153718735e-05,
"loss": 1.8226,
"step": 92
},
{
"epoch": 0.01366742596810934,
"grad_norm": 1.8467638492584229,
"learning_rate": 5.985585137257401e-05,
"loss": 1.7656,
"step": 93
},
{
"epoch": 0.013814387537658903,
"grad_norm": 1.8387916088104248,
"learning_rate": 5.90440267166055e-05,
"loss": 1.5562,
"step": 94
},
{
"epoch": 0.013961349107208466,
"grad_norm": 1.8392918109893799,
"learning_rate": 5.8229729514036705e-05,
"loss": 1.6858,
"step": 95
},
{
"epoch": 0.014108310676758027,
"grad_norm": 2.010154962539673,
"learning_rate": 5.74131823855921e-05,
"loss": 1.4479,
"step": 96
},
{
"epoch": 0.01425527224630759,
"grad_norm": 2.9376213550567627,
"learning_rate": 5.6594608567103456e-05,
"loss": 1.7915,
"step": 97
},
{
"epoch": 0.014402233815857153,
"grad_norm": 2.68959379196167,
"learning_rate": 5.577423184847932e-05,
"loss": 1.3952,
"step": 98
},
{
"epoch": 0.014549195385406716,
"grad_norm": 2.9025957584381104,
"learning_rate": 5.495227651252315e-05,
"loss": 1.5607,
"step": 99
},
{
"epoch": 0.01469615695495628,
"grad_norm": 4.17297887802124,
"learning_rate": 5.4128967273616625e-05,
"loss": 1.8803,
"step": 100
},
{
"epoch": 0.01469615695495628,
"eval_loss": 1.9667052030563354,
"eval_runtime": 240.5684,
"eval_samples_per_second": 47.641,
"eval_steps_per_second": 11.913,
"step": 100
},
{
"epoch": 0.014843118524505842,
"grad_norm": 1.3378857374191284,
"learning_rate": 5.330452921628497e-05,
"loss": 2.1213,
"step": 101
},
{
"epoch": 0.014990080094055405,
"grad_norm": 1.5491584539413452,
"learning_rate": 5.247918773366112e-05,
"loss": 2.3245,
"step": 102
},
{
"epoch": 0.015137041663604967,
"grad_norm": 1.5535106658935547,
"learning_rate": 5.165316846586541e-05,
"loss": 2.14,
"step": 103
},
{
"epoch": 0.01528400323315453,
"grad_norm": 1.4448267221450806,
"learning_rate": 5.0826697238317935e-05,
"loss": 2.2545,
"step": 104
},
{
"epoch": 0.015430964802704093,
"grad_norm": 1.299795389175415,
"learning_rate": 5e-05,
"loss": 2.0789,
"step": 105
},
{
"epoch": 0.015577926372253656,
"grad_norm": 1.127762794494629,
"learning_rate": 4.917330276168208e-05,
"loss": 2.0942,
"step": 106
},
{
"epoch": 0.01572488794180322,
"grad_norm": 1.1137936115264893,
"learning_rate": 4.834683153413459e-05,
"loss": 2.1587,
"step": 107
},
{
"epoch": 0.01587184951135278,
"grad_norm": 1.0205131769180298,
"learning_rate": 4.7520812266338885e-05,
"loss": 2.1194,
"step": 108
},
{
"epoch": 0.016018811080902345,
"grad_norm": 1.0848978757858276,
"learning_rate": 4.669547078371504e-05,
"loss": 1.7718,
"step": 109
},
{
"epoch": 0.016165772650451906,
"grad_norm": 1.0921462774276733,
"learning_rate": 4.5871032726383386e-05,
"loss": 1.9232,
"step": 110
},
{
"epoch": 0.01631273422000147,
"grad_norm": 1.0820595026016235,
"learning_rate": 4.504772348747687e-05,
"loss": 2.049,
"step": 111
},
{
"epoch": 0.016459695789551032,
"grad_norm": 1.1296908855438232,
"learning_rate": 4.4225768151520694e-05,
"loss": 2.0442,
"step": 112
},
{
"epoch": 0.016606657359100597,
"grad_norm": 1.1160941123962402,
"learning_rate": 4.3405391432896555e-05,
"loss": 2.0111,
"step": 113
},
{
"epoch": 0.016753618928650158,
"grad_norm": 1.146777868270874,
"learning_rate": 4.2586817614407895e-05,
"loss": 1.934,
"step": 114
},
{
"epoch": 0.01690058049819972,
"grad_norm": 1.0680501461029053,
"learning_rate": 4.17702704859633e-05,
"loss": 1.8576,
"step": 115
},
{
"epoch": 0.017047542067749284,
"grad_norm": 1.0956605672836304,
"learning_rate": 4.095597328339452e-05,
"loss": 1.784,
"step": 116
},
{
"epoch": 0.017194503637298846,
"grad_norm": 1.131801724433899,
"learning_rate": 4.0144148627425993e-05,
"loss": 2.0918,
"step": 117
},
{
"epoch": 0.01734146520684841,
"grad_norm": 1.1751807928085327,
"learning_rate": 3.933501846281267e-05,
"loss": 2.0375,
"step": 118
},
{
"epoch": 0.01748842677639797,
"grad_norm": 1.2810145616531372,
"learning_rate": 3.852880399766243e-05,
"loss": 1.9278,
"step": 119
},
{
"epoch": 0.017635388345947536,
"grad_norm": 1.0781248807907104,
"learning_rate": 3.772572564296005e-05,
"loss": 1.7968,
"step": 120
},
{
"epoch": 0.017782349915497098,
"grad_norm": 1.1969642639160156,
"learning_rate": 3.6926002952309016e-05,
"loss": 2.0,
"step": 121
},
{
"epoch": 0.01792931148504666,
"grad_norm": 1.2677849531173706,
"learning_rate": 3.612985456190778e-05,
"loss": 1.8359,
"step": 122
},
{
"epoch": 0.018076273054596224,
"grad_norm": 1.416784405708313,
"learning_rate": 3.533749813077677e-05,
"loss": 1.9841,
"step": 123
},
{
"epoch": 0.018223234624145785,
"grad_norm": 1.1759361028671265,
"learning_rate": 3.4549150281252636e-05,
"loss": 1.9378,
"step": 124
},
{
"epoch": 0.01837019619369535,
"grad_norm": 1.2099229097366333,
"learning_rate": 3.3765026539765834e-05,
"loss": 1.952,
"step": 125
},
{
"epoch": 0.01851715776324491,
"grad_norm": 1.198412299156189,
"learning_rate": 3.298534127791785e-05,
"loss": 1.7128,
"step": 126
},
{
"epoch": 0.018664119332794476,
"grad_norm": 1.3011282682418823,
"learning_rate": 3.221030765387417e-05,
"loss": 1.6071,
"step": 127
},
{
"epoch": 0.018811080902344037,
"grad_norm": 1.4033522605895996,
"learning_rate": 3.144013755408895e-05,
"loss": 2.0028,
"step": 128
},
{
"epoch": 0.0189580424718936,
"grad_norm": 1.3320146799087524,
"learning_rate": 3.0675041535377405e-05,
"loss": 1.7789,
"step": 129
},
{
"epoch": 0.019105004041443163,
"grad_norm": 1.356528639793396,
"learning_rate": 2.991522876735154e-05,
"loss": 1.8743,
"step": 130
},
{
"epoch": 0.019251965610992725,
"grad_norm": 1.3291962146759033,
"learning_rate": 2.916090697523549e-05,
"loss": 1.9648,
"step": 131
},
{
"epoch": 0.01939892718054229,
"grad_norm": 1.4146960973739624,
"learning_rate": 2.8412282383075363e-05,
"loss": 1.8739,
"step": 132
},
{
"epoch": 0.01954588875009185,
"grad_norm": 1.2984075546264648,
"learning_rate": 2.766955965735968e-05,
"loss": 1.812,
"step": 133
},
{
"epoch": 0.019692850319641415,
"grad_norm": 1.5239055156707764,
"learning_rate": 2.693294185106562e-05,
"loss": 1.76,
"step": 134
},
{
"epoch": 0.019839811889190977,
"grad_norm": 1.6176416873931885,
"learning_rate": 2.6202630348146324e-05,
"loss": 1.7613,
"step": 135
},
{
"epoch": 0.019986773458740538,
"grad_norm": 1.637717604637146,
"learning_rate": 2.547882480847461e-05,
"loss": 1.7462,
"step": 136
},
{
"epoch": 0.020133735028290103,
"grad_norm": 1.305309534072876,
"learning_rate": 2.476172311325783e-05,
"loss": 1.5191,
"step": 137
},
{
"epoch": 0.020280696597839664,
"grad_norm": 1.460196852684021,
"learning_rate": 2.405152131093926e-05,
"loss": 1.5225,
"step": 138
},
{
"epoch": 0.02042765816738923,
"grad_norm": 1.530012845993042,
"learning_rate": 2.3348413563600325e-05,
"loss": 1.7372,
"step": 139
},
{
"epoch": 0.02057461973693879,
"grad_norm": 1.5662577152252197,
"learning_rate": 2.2652592093878666e-05,
"loss": 1.5763,
"step": 140
},
{
"epoch": 0.020721581306488355,
"grad_norm": 1.7299176454544067,
"learning_rate": 2.196424713241637e-05,
"loss": 1.7782,
"step": 141
},
{
"epoch": 0.020868542876037916,
"grad_norm": 1.8284093141555786,
"learning_rate": 2.128356686585282e-05,
"loss": 1.8607,
"step": 142
},
{
"epoch": 0.021015504445587477,
"grad_norm": 2.0168917179107666,
"learning_rate": 2.061073738537635e-05,
"loss": 1.3412,
"step": 143
},
{
"epoch": 0.021162466015137042,
"grad_norm": 2.0473945140838623,
"learning_rate": 1.9945942635848748e-05,
"loss": 1.682,
"step": 144
},
{
"epoch": 0.021309427584686604,
"grad_norm": 1.908447265625,
"learning_rate": 1.928936436551661e-05,
"loss": 1.2712,
"step": 145
},
{
"epoch": 0.02145638915423617,
"grad_norm": 1.8296012878417969,
"learning_rate": 1.8641182076323148e-05,
"loss": 1.5448,
"step": 146
},
{
"epoch": 0.02160335072378573,
"grad_norm": 2.143108367919922,
"learning_rate": 1.800157297483417e-05,
"loss": 1.538,
"step": 147
},
{
"epoch": 0.021750312293335294,
"grad_norm": 2.5496554374694824,
"learning_rate": 1.7370711923791567e-05,
"loss": 1.5385,
"step": 148
},
{
"epoch": 0.021897273862884856,
"grad_norm": 2.8672902584075928,
"learning_rate": 1.6748771394307585e-05,
"loss": 1.4678,
"step": 149
},
{
"epoch": 0.022044235432434417,
"grad_norm": 3.4696292877197266,
"learning_rate": 1.6135921418712956e-05,
"loss": 1.6712,
"step": 150
},
{
"epoch": 0.022044235432434417,
"eval_loss": 1.84725821018219,
"eval_runtime": 240.2018,
"eval_samples_per_second": 47.714,
"eval_steps_per_second": 11.932,
"step": 150
},
{
"epoch": 0.022191197001983982,
"grad_norm": 0.8707069754600525,
"learning_rate": 1.553232954407171e-05,
"loss": 1.8186,
"step": 151
},
{
"epoch": 0.022338158571533543,
"grad_norm": 0.9638893604278564,
"learning_rate": 1.4938160786375572e-05,
"loss": 1.9337,
"step": 152
},
{
"epoch": 0.022485120141083108,
"grad_norm": 1.0150665044784546,
"learning_rate": 1.435357758543015e-05,
"loss": 1.9687,
"step": 153
},
{
"epoch": 0.02263208171063267,
"grad_norm": 0.9966422915458679,
"learning_rate": 1.3778739760445552e-05,
"loss": 2.0731,
"step": 154
},
{
"epoch": 0.022779043280182234,
"grad_norm": 0.9775243401527405,
"learning_rate": 1.3213804466343421e-05,
"loss": 2.0418,
"step": 155
},
{
"epoch": 0.022926004849731795,
"grad_norm": 1.0116099119186401,
"learning_rate": 1.2658926150792322e-05,
"loss": 1.9921,
"step": 156
},
{
"epoch": 0.023072966419281356,
"grad_norm": 1.0379962921142578,
"learning_rate": 1.2114256511983274e-05,
"loss": 2.0333,
"step": 157
},
{
"epoch": 0.02321992798883092,
"grad_norm": 1.0433306694030762,
"learning_rate": 1.157994445715706e-05,
"loss": 2.118,
"step": 158
},
{
"epoch": 0.023366889558380483,
"grad_norm": 1.0108063220977783,
"learning_rate": 1.1056136061894384e-05,
"loss": 1.9515,
"step": 159
},
{
"epoch": 0.023513851127930047,
"grad_norm": 1.0546578168869019,
"learning_rate": 1.0542974530180327e-05,
"loss": 2.043,
"step": 160
},
{
"epoch": 0.02366081269747961,
"grad_norm": 1.0310964584350586,
"learning_rate": 1.0040600155253765e-05,
"loss": 2.0187,
"step": 161
},
{
"epoch": 0.023807774267029173,
"grad_norm": 1.082858681678772,
"learning_rate": 9.549150281252633e-06,
"loss": 2.0717,
"step": 162
},
{
"epoch": 0.023954735836578735,
"grad_norm": 1.0323563814163208,
"learning_rate": 9.068759265665384e-06,
"loss": 1.8542,
"step": 163
},
{
"epoch": 0.024101697406128296,
"grad_norm": 1.0657505989074707,
"learning_rate": 8.599558442598998e-06,
"loss": 2.028,
"step": 164
},
{
"epoch": 0.02424865897567786,
"grad_norm": 1.1028742790222168,
"learning_rate": 8.141676086873572e-06,
"loss": 2.112,
"step": 165
},
{
"epoch": 0.024395620545227422,
"grad_norm": 1.1056249141693115,
"learning_rate": 7.695237378953223e-06,
"loss": 1.8936,
"step": 166
},
{
"epoch": 0.024542582114776987,
"grad_norm": 1.1669912338256836,
"learning_rate": 7.260364370723044e-06,
"loss": 2.1979,
"step": 167
},
{
"epoch": 0.024689543684326548,
"grad_norm": 1.0488593578338623,
"learning_rate": 6.837175952121306e-06,
"loss": 2.0042,
"step": 168
},
{
"epoch": 0.024836505253876113,
"grad_norm": 1.1563531160354614,
"learning_rate": 6.425787818636131e-06,
"loss": 1.9956,
"step": 169
},
{
"epoch": 0.024983466823425674,
"grad_norm": 1.2589812278747559,
"learning_rate": 6.026312439675552e-06,
"loss": 2.0068,
"step": 170
},
{
"epoch": 0.025130428392975235,
"grad_norm": 1.4134507179260254,
"learning_rate": 5.6388590278194096e-06,
"loss": 2.0624,
"step": 171
},
{
"epoch": 0.0252773899625248,
"grad_norm": 1.169216513633728,
"learning_rate": 5.263533508961827e-06,
"loss": 1.8322,
"step": 172
},
{
"epoch": 0.02542435153207436,
"grad_norm": 1.3565715551376343,
"learning_rate": 4.900438493352055e-06,
"loss": 1.8057,
"step": 173
},
{
"epoch": 0.025571313101623926,
"grad_norm": 1.2308727502822876,
"learning_rate": 4.549673247541875e-06,
"loss": 1.8049,
"step": 174
},
{
"epoch": 0.025718274671173488,
"grad_norm": 1.3102329969406128,
"learning_rate": 4.2113336672471245e-06,
"loss": 2.0119,
"step": 175
},
{
"epoch": 0.025865236240723052,
"grad_norm": 1.210121750831604,
"learning_rate": 3.885512251130763e-06,
"loss": 1.795,
"step": 176
},
{
"epoch": 0.026012197810272614,
"grad_norm": 1.204565405845642,
"learning_rate": 3.5722980755146517e-06,
"loss": 1.7286,
"step": 177
},
{
"epoch": 0.026159159379822175,
"grad_norm": 1.3018378019332886,
"learning_rate": 3.271776770026963e-06,
"loss": 1.8865,
"step": 178
},
{
"epoch": 0.02630612094937174,
"grad_norm": 1.4115206003189087,
"learning_rate": 2.9840304941919415e-06,
"loss": 1.6768,
"step": 179
},
{
"epoch": 0.0264530825189213,
"grad_norm": 1.2980129718780518,
"learning_rate": 2.7091379149682685e-06,
"loss": 1.9056,
"step": 180
},
{
"epoch": 0.026600044088470866,
"grad_norm": 1.4762133359909058,
"learning_rate": 2.4471741852423237e-06,
"loss": 1.8847,
"step": 181
},
{
"epoch": 0.026747005658020427,
"grad_norm": 1.5837223529815674,
"learning_rate": 2.1982109232821178e-06,
"loss": 1.6046,
"step": 182
},
{
"epoch": 0.026893967227569992,
"grad_norm": 1.3648614883422852,
"learning_rate": 1.962316193157593e-06,
"loss": 1.9248,
"step": 183
},
{
"epoch": 0.027040928797119553,
"grad_norm": 1.3527555465698242,
"learning_rate": 1.7395544861325718e-06,
"loss": 1.732,
"step": 184
},
{
"epoch": 0.027187890366669114,
"grad_norm": 1.4113436937332153,
"learning_rate": 1.5299867030334814e-06,
"loss": 1.6009,
"step": 185
},
{
"epoch": 0.02733485193621868,
"grad_norm": 1.5120950937271118,
"learning_rate": 1.333670137599713e-06,
"loss": 1.7918,
"step": 186
},
{
"epoch": 0.02748181350576824,
"grad_norm": 1.4823616743087769,
"learning_rate": 1.1506584608200367e-06,
"loss": 1.6311,
"step": 187
},
{
"epoch": 0.027628775075317805,
"grad_norm": 1.4522048234939575,
"learning_rate": 9.810017062595322e-07,
"loss": 1.4494,
"step": 188
},
{
"epoch": 0.027775736644867367,
"grad_norm": 1.6617823839187622,
"learning_rate": 8.247462563808817e-07,
"loss": 1.7989,
"step": 189
},
{
"epoch": 0.02792269821441693,
"grad_norm": 1.531654715538025,
"learning_rate": 6.819348298638839e-07,
"loss": 1.6397,
"step": 190
},
{
"epoch": 0.028069659783966493,
"grad_norm": 1.7335494756698608,
"learning_rate": 5.526064699265753e-07,
"loss": 1.7845,
"step": 191
},
{
"epoch": 0.028216621353516054,
"grad_norm": 1.8958544731140137,
"learning_rate": 4.367965336512403e-07,
"loss": 1.7091,
"step": 192
},
{
"epoch": 0.02836358292306562,
"grad_norm": 1.8163591623306274,
"learning_rate": 3.3453668231809286e-07,
"loss": 1.6346,
"step": 193
},
{
"epoch": 0.02851054449261518,
"grad_norm": 2.151160955429077,
"learning_rate": 2.458548727494292e-07,
"loss": 1.6286,
"step": 194
},
{
"epoch": 0.028657506062164745,
"grad_norm": 2.189847946166992,
"learning_rate": 1.7077534966650766e-07,
"loss": 1.7776,
"step": 195
},
{
"epoch": 0.028804467631714306,
"grad_norm": 3.335634469985962,
"learning_rate": 1.0931863906127327e-07,
"loss": 1.876,
"step": 196
},
{
"epoch": 0.02895142920126387,
"grad_norm": 2.7531423568725586,
"learning_rate": 6.150154258476315e-08,
"loss": 1.4536,
"step": 197
},
{
"epoch": 0.029098390770813432,
"grad_norm": 2.587663412094116,
"learning_rate": 2.7337132953697554e-08,
"loss": 1.6882,
"step": 198
},
{
"epoch": 0.029245352340362993,
"grad_norm": 3.4245827198028564,
"learning_rate": 6.834750376549792e-09,
"loss": 1.966,
"step": 199
},
{
"epoch": 0.02939231390991256,
"grad_norm": 4.0924530029296875,
"learning_rate": 0.0,
"loss": 1.7434,
"step": 200
},
{
"epoch": 0.02939231390991256,
"eval_loss": 1.8370285034179688,
"eval_runtime": 240.4445,
"eval_samples_per_second": 47.666,
"eval_steps_per_second": 11.92,
"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": 4.0038624657408e+16,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}