nemo-novel-magnum-1ep / trainer_state.json
taozi555's picture
Upload folder using huggingface_hub
9cd8068 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 225,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0044444444444444444,
"grad_norm": 7.197563171386719,
"learning_rate": 2.5000000000000004e-07,
"loss": 1.3915,
"step": 1
},
{
"epoch": 0.008888888888888889,
"grad_norm": 7.682923793792725,
"learning_rate": 5.000000000000001e-07,
"loss": 1.3854,
"step": 2
},
{
"epoch": 0.013333333333333334,
"grad_norm": 9.152216911315918,
"learning_rate": 7.5e-07,
"loss": 1.3159,
"step": 3
},
{
"epoch": 0.017777777777777778,
"grad_norm": 9.388493537902832,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.3236,
"step": 4
},
{
"epoch": 0.022222222222222223,
"grad_norm": 12.188794136047363,
"learning_rate": 1.25e-06,
"loss": 1.3502,
"step": 5
},
{
"epoch": 0.02666666666666667,
"grad_norm": 12.245216369628906,
"learning_rate": 1.5e-06,
"loss": 1.404,
"step": 6
},
{
"epoch": 0.03111111111111111,
"grad_norm": 19.210506439208984,
"learning_rate": 1.75e-06,
"loss": 1.4199,
"step": 7
},
{
"epoch": 0.035555555555555556,
"grad_norm": 9.991368293762207,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.3316,
"step": 8
},
{
"epoch": 0.04,
"grad_norm": 6.724348545074463,
"learning_rate": 2.25e-06,
"loss": 1.4482,
"step": 9
},
{
"epoch": 0.044444444444444446,
"grad_norm": 8.962543487548828,
"learning_rate": 2.5e-06,
"loss": 1.29,
"step": 10
},
{
"epoch": 0.04888888888888889,
"grad_norm": 7.128359317779541,
"learning_rate": 2.7500000000000004e-06,
"loss": 1.3139,
"step": 11
},
{
"epoch": 0.05333333333333334,
"grad_norm": 11.580962181091309,
"learning_rate": 3e-06,
"loss": 1.3238,
"step": 12
},
{
"epoch": 0.057777777777777775,
"grad_norm": 8.380171775817871,
"learning_rate": 3.2500000000000002e-06,
"loss": 1.3434,
"step": 13
},
{
"epoch": 0.06222222222222222,
"grad_norm": 5.92144250869751,
"learning_rate": 3.5e-06,
"loss": 1.2896,
"step": 14
},
{
"epoch": 0.06666666666666667,
"grad_norm": 8.112162590026855,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.2117,
"step": 15
},
{
"epoch": 0.07111111111111111,
"grad_norm": 3.5406394004821777,
"learning_rate": 4.000000000000001e-06,
"loss": 1.2094,
"step": 16
},
{
"epoch": 0.07555555555555556,
"grad_norm": 22.047605514526367,
"learning_rate": 4.25e-06,
"loss": 1.2027,
"step": 17
},
{
"epoch": 0.08,
"grad_norm": 6.793862819671631,
"learning_rate": 4.5e-06,
"loss": 1.2434,
"step": 18
},
{
"epoch": 0.08444444444444445,
"grad_norm": 9.382174491882324,
"learning_rate": 4.75e-06,
"loss": 1.3183,
"step": 19
},
{
"epoch": 0.08888888888888889,
"grad_norm": 5.871135234832764,
"learning_rate": 5e-06,
"loss": 1.2432,
"step": 20
},
{
"epoch": 0.09333333333333334,
"grad_norm": 17.656475067138672,
"learning_rate": 5.2500000000000006e-06,
"loss": 1.1843,
"step": 21
},
{
"epoch": 0.09777777777777778,
"grad_norm": 10.098402976989746,
"learning_rate": 5.500000000000001e-06,
"loss": 1.3401,
"step": 22
},
{
"epoch": 0.10222222222222223,
"grad_norm": 23.264266967773438,
"learning_rate": 5.75e-06,
"loss": 1.3025,
"step": 23
},
{
"epoch": 0.10666666666666667,
"grad_norm": 4.140642166137695,
"learning_rate": 6e-06,
"loss": 1.2141,
"step": 24
},
{
"epoch": 0.1111111111111111,
"grad_norm": 6.262714862823486,
"learning_rate": 6.25e-06,
"loss": 1.2665,
"step": 25
},
{
"epoch": 0.11555555555555555,
"grad_norm": 6.946407794952393,
"learning_rate": 6.5000000000000004e-06,
"loss": 1.1989,
"step": 26
},
{
"epoch": 0.12,
"grad_norm": 2.5363879203796387,
"learning_rate": 6.750000000000001e-06,
"loss": 1.1771,
"step": 27
},
{
"epoch": 0.12444444444444444,
"grad_norm": 3.964658737182617,
"learning_rate": 7e-06,
"loss": 1.2362,
"step": 28
},
{
"epoch": 0.1288888888888889,
"grad_norm": 2.513711452484131,
"learning_rate": 7.25e-06,
"loss": 1.2203,
"step": 29
},
{
"epoch": 0.13333333333333333,
"grad_norm": 2.6071319580078125,
"learning_rate": 7.500000000000001e-06,
"loss": 1.167,
"step": 30
},
{
"epoch": 0.13777777777777778,
"grad_norm": 3.0134849548339844,
"learning_rate": 7.75e-06,
"loss": 1.2139,
"step": 31
},
{
"epoch": 0.14222222222222222,
"grad_norm": 2.4809670448303223,
"learning_rate": 8.000000000000001e-06,
"loss": 1.1603,
"step": 32
},
{
"epoch": 0.14666666666666667,
"grad_norm": 3.1294925212860107,
"learning_rate": 8.25e-06,
"loss": 1.1922,
"step": 33
},
{
"epoch": 0.1511111111111111,
"grad_norm": 2.703998565673828,
"learning_rate": 8.5e-06,
"loss": 1.1858,
"step": 34
},
{
"epoch": 0.15555555555555556,
"grad_norm": 4.653899192810059,
"learning_rate": 8.750000000000001e-06,
"loss": 1.2678,
"step": 35
},
{
"epoch": 0.16,
"grad_norm": 2.976705312728882,
"learning_rate": 9e-06,
"loss": 1.2004,
"step": 36
},
{
"epoch": 0.16444444444444445,
"grad_norm": 3.332261323928833,
"learning_rate": 9.250000000000001e-06,
"loss": 1.2128,
"step": 37
},
{
"epoch": 0.1688888888888889,
"grad_norm": 2.7455172538757324,
"learning_rate": 9.5e-06,
"loss": 1.2348,
"step": 38
},
{
"epoch": 0.17333333333333334,
"grad_norm": 2.5534517765045166,
"learning_rate": 9.75e-06,
"loss": 1.1752,
"step": 39
},
{
"epoch": 0.17777777777777778,
"grad_norm": 2.2505180835723877,
"learning_rate": 1e-05,
"loss": 1.1248,
"step": 40
},
{
"epoch": 0.18222222222222223,
"grad_norm": 3.083693265914917,
"learning_rate": 9.999853218975136e-06,
"loss": 1.1466,
"step": 41
},
{
"epoch": 0.18666666666666668,
"grad_norm": 2.871704339981079,
"learning_rate": 9.99941288451841e-06,
"loss": 1.1622,
"step": 42
},
{
"epoch": 0.19111111111111112,
"grad_norm": 2.9086086750030518,
"learning_rate": 9.998679022482916e-06,
"loss": 1.1584,
"step": 43
},
{
"epoch": 0.19555555555555557,
"grad_norm": 2.8458251953125,
"learning_rate": 9.997651675955467e-06,
"loss": 1.1556,
"step": 44
},
{
"epoch": 0.2,
"grad_norm": 3.2609152793884277,
"learning_rate": 9.99633090525405e-06,
"loss": 1.1991,
"step": 45
},
{
"epoch": 0.20444444444444446,
"grad_norm": 14.039426803588867,
"learning_rate": 9.9947167879243e-06,
"loss": 1.1318,
"step": 46
},
{
"epoch": 0.2088888888888889,
"grad_norm": 3.4546499252319336,
"learning_rate": 9.992809418734932e-06,
"loss": 1.1987,
"step": 47
},
{
"epoch": 0.21333333333333335,
"grad_norm": 3.3863046169281006,
"learning_rate": 9.99060890967219e-06,
"loss": 1.171,
"step": 48
},
{
"epoch": 0.21777777777777776,
"grad_norm": 2.217785596847534,
"learning_rate": 9.988115389933263e-06,
"loss": 1.124,
"step": 49
},
{
"epoch": 0.2222222222222222,
"grad_norm": 3.8226356506347656,
"learning_rate": 9.985329005918702e-06,
"loss": 1.2048,
"step": 50
},
{
"epoch": 0.22666666666666666,
"grad_norm": 4.5612688064575195,
"learning_rate": 9.982249921223833e-06,
"loss": 1.1899,
"step": 51
},
{
"epoch": 0.2311111111111111,
"grad_norm": 3.4833076000213623,
"learning_rate": 9.978878316629132e-06,
"loss": 1.1745,
"step": 52
},
{
"epoch": 0.23555555555555555,
"grad_norm": 2.259718894958496,
"learning_rate": 9.975214390089637e-06,
"loss": 1.1016,
"step": 53
},
{
"epoch": 0.24,
"grad_norm": 2.242814779281616,
"learning_rate": 9.9712583567233e-06,
"loss": 1.1984,
"step": 54
},
{
"epoch": 0.24444444444444444,
"grad_norm": 4.208090305328369,
"learning_rate": 9.967010448798376e-06,
"loss": 1.1899,
"step": 55
},
{
"epoch": 0.24888888888888888,
"grad_norm": 2.8487319946289062,
"learning_rate": 9.962470915719775e-06,
"loss": 1.1735,
"step": 56
},
{
"epoch": 0.25333333333333335,
"grad_norm": 2.4001946449279785,
"learning_rate": 9.957640024014426e-06,
"loss": 1.1199,
"step": 57
},
{
"epoch": 0.2577777777777778,
"grad_norm": 4.717513561248779,
"learning_rate": 9.952518057315624e-06,
"loss": 1.1234,
"step": 58
},
{
"epoch": 0.26222222222222225,
"grad_norm": 2.7141494750976562,
"learning_rate": 9.947105316346372e-06,
"loss": 1.1619,
"step": 59
},
{
"epoch": 0.26666666666666666,
"grad_norm": 2.610142946243286,
"learning_rate": 9.941402118901743e-06,
"loss": 1.2141,
"step": 60
},
{
"epoch": 0.27111111111111114,
"grad_norm": 3.9055914878845215,
"learning_rate": 9.9354087998302e-06,
"loss": 1.1778,
"step": 61
},
{
"epoch": 0.27555555555555555,
"grad_norm": 3.3144619464874268,
"learning_rate": 9.929125711013952e-06,
"loss": 1.1574,
"step": 62
},
{
"epoch": 0.28,
"grad_norm": 2.429931640625,
"learning_rate": 9.922553221348281e-06,
"loss": 1.1524,
"step": 63
},
{
"epoch": 0.28444444444444444,
"grad_norm": 2.495755195617676,
"learning_rate": 9.915691716719899e-06,
"loss": 1.1692,
"step": 64
},
{
"epoch": 0.28888888888888886,
"grad_norm": 1.96566903591156,
"learning_rate": 9.908541599984276e-06,
"loss": 1.0634,
"step": 65
},
{
"epoch": 0.29333333333333333,
"grad_norm": 2.161886215209961,
"learning_rate": 9.901103290941996e-06,
"loss": 1.1318,
"step": 66
},
{
"epoch": 0.29777777777777775,
"grad_norm": 2.408609390258789,
"learning_rate": 9.893377226314113e-06,
"loss": 1.118,
"step": 67
},
{
"epoch": 0.3022222222222222,
"grad_norm": 2.3472368717193604,
"learning_rate": 9.885363859716497e-06,
"loss": 1.1064,
"step": 68
},
{
"epoch": 0.30666666666666664,
"grad_norm": 4.991092681884766,
"learning_rate": 9.877063661633213e-06,
"loss": 1.1629,
"step": 69
},
{
"epoch": 0.3111111111111111,
"grad_norm": 2.2165753841400146,
"learning_rate": 9.868477119388897e-06,
"loss": 1.0692,
"step": 70
},
{
"epoch": 0.31555555555555553,
"grad_norm": 3.0987343788146973,
"learning_rate": 9.859604737120131e-06,
"loss": 1.1878,
"step": 71
},
{
"epoch": 0.32,
"grad_norm": 3.5139403343200684,
"learning_rate": 9.850447035745868e-06,
"loss": 1.1443,
"step": 72
},
{
"epoch": 0.3244444444444444,
"grad_norm": 2.9309511184692383,
"learning_rate": 9.841004552936817e-06,
"loss": 1.1986,
"step": 73
},
{
"epoch": 0.3288888888888889,
"grad_norm": 2.642705202102661,
"learning_rate": 9.831277843083904e-06,
"loss": 1.1076,
"step": 74
},
{
"epoch": 0.3333333333333333,
"grad_norm": 2.53140926361084,
"learning_rate": 9.821267477265705e-06,
"loss": 1.1432,
"step": 75
},
{
"epoch": 0.3377777777777778,
"grad_norm": 3.0417535305023193,
"learning_rate": 9.810974043214923e-06,
"loss": 1.1889,
"step": 76
},
{
"epoch": 0.3422222222222222,
"grad_norm": 2.1667232513427734,
"learning_rate": 9.800398145283874e-06,
"loss": 1.1088,
"step": 77
},
{
"epoch": 0.3466666666666667,
"grad_norm": 2.451881170272827,
"learning_rate": 9.789540404409017e-06,
"loss": 1.125,
"step": 78
},
{
"epoch": 0.3511111111111111,
"grad_norm": 5.966605186462402,
"learning_rate": 9.778401458074482e-06,
"loss": 1.2031,
"step": 79
},
{
"epoch": 0.35555555555555557,
"grad_norm": 2.767612934112549,
"learning_rate": 9.766981960274653e-06,
"loss": 1.1299,
"step": 80
},
{
"epoch": 0.36,
"grad_norm": 2.8305821418762207,
"learning_rate": 9.755282581475769e-06,
"loss": 1.106,
"step": 81
},
{
"epoch": 0.36444444444444446,
"grad_norm": 2.256387710571289,
"learning_rate": 9.74330400857655e-06,
"loss": 1.1994,
"step": 82
},
{
"epoch": 0.3688888888888889,
"grad_norm": 2.9722063541412354,
"learning_rate": 9.731046944867883e-06,
"loss": 1.1182,
"step": 83
},
{
"epoch": 0.37333333333333335,
"grad_norm": 3.528620958328247,
"learning_rate": 9.718512109991516e-06,
"loss": 1.2216,
"step": 84
},
{
"epoch": 0.37777777777777777,
"grad_norm": 2.0519301891326904,
"learning_rate": 9.705700239897809e-06,
"loss": 1.1228,
"step": 85
},
{
"epoch": 0.38222222222222224,
"grad_norm": 2.4317550659179688,
"learning_rate": 9.692612086802536e-06,
"loss": 1.164,
"step": 86
},
{
"epoch": 0.38666666666666666,
"grad_norm": 3.0390419960021973,
"learning_rate": 9.679248419142704e-06,
"loss": 1.2052,
"step": 87
},
{
"epoch": 0.39111111111111113,
"grad_norm": 2.4871633052825928,
"learning_rate": 9.665610021531447e-06,
"loss": 1.1605,
"step": 88
},
{
"epoch": 0.39555555555555555,
"grad_norm": 3.762112617492676,
"learning_rate": 9.651697694711959e-06,
"loss": 1.1504,
"step": 89
},
{
"epoch": 0.4,
"grad_norm": 18.623851776123047,
"learning_rate": 9.637512255510475e-06,
"loss": 1.173,
"step": 90
},
{
"epoch": 0.40444444444444444,
"grad_norm": 2.8003592491149902,
"learning_rate": 9.623054536788315e-06,
"loss": 1.1239,
"step": 91
},
{
"epoch": 0.4088888888888889,
"grad_norm": 2.560473918914795,
"learning_rate": 9.608325387392988e-06,
"loss": 1.1203,
"step": 92
},
{
"epoch": 0.41333333333333333,
"grad_norm": 1.885018229484558,
"learning_rate": 9.593325672108352e-06,
"loss": 1.142,
"step": 93
},
{
"epoch": 0.4177777777777778,
"grad_norm": 4.514256000518799,
"learning_rate": 9.578056271603837e-06,
"loss": 1.1398,
"step": 94
},
{
"epoch": 0.4222222222222222,
"grad_norm": 2.7508609294891357,
"learning_rate": 9.562518082382751e-06,
"loss": 1.1355,
"step": 95
},
{
"epoch": 0.4266666666666667,
"grad_norm": 2.800530433654785,
"learning_rate": 9.546712016729625e-06,
"loss": 1.1512,
"step": 96
},
{
"epoch": 0.4311111111111111,
"grad_norm": 3.151304006576538,
"learning_rate": 9.530639002656665e-06,
"loss": 1.1215,
"step": 97
},
{
"epoch": 0.43555555555555553,
"grad_norm": 2.7462916374206543,
"learning_rate": 9.514299983849267e-06,
"loss": 1.1412,
"step": 98
},
{
"epoch": 0.44,
"grad_norm": 1.8580820560455322,
"learning_rate": 9.497695919610595e-06,
"loss": 1.088,
"step": 99
},
{
"epoch": 0.4444444444444444,
"grad_norm": 2.2820510864257812,
"learning_rate": 9.480827784805278e-06,
"loss": 1.1844,
"step": 100
},
{
"epoch": 0.4488888888888889,
"grad_norm": 1.8663792610168457,
"learning_rate": 9.463696569802163e-06,
"loss": 1.1218,
"step": 101
},
{
"epoch": 0.4533333333333333,
"grad_norm": 2.3818678855895996,
"learning_rate": 9.446303280416167e-06,
"loss": 1.186,
"step": 102
},
{
"epoch": 0.4577777777777778,
"grad_norm": 1.8850212097167969,
"learning_rate": 9.428648937849227e-06,
"loss": 1.0865,
"step": 103
},
{
"epoch": 0.4622222222222222,
"grad_norm": 2.012317419052124,
"learning_rate": 9.410734578630344e-06,
"loss": 1.0468,
"step": 104
},
{
"epoch": 0.4666666666666667,
"grad_norm": 2.9131033420562744,
"learning_rate": 9.392561254554712e-06,
"loss": 1.1631,
"step": 105
},
{
"epoch": 0.4711111111111111,
"grad_norm": 3.2192587852478027,
"learning_rate": 9.374130032621993e-06,
"loss": 1.1361,
"step": 106
},
{
"epoch": 0.47555555555555556,
"grad_norm": 2.818880319595337,
"learning_rate": 9.355441994973639e-06,
"loss": 1.1578,
"step": 107
},
{
"epoch": 0.48,
"grad_norm": 3.6746394634246826,
"learning_rate": 9.336498238829383e-06,
"loss": 1.1542,
"step": 108
},
{
"epoch": 0.48444444444444446,
"grad_norm": 3.4095306396484375,
"learning_rate": 9.317299876422797e-06,
"loss": 1.142,
"step": 109
},
{
"epoch": 0.4888888888888889,
"grad_norm": 2.7571191787719727,
"learning_rate": 9.297848034936007e-06,
"loss": 1.1133,
"step": 110
},
{
"epoch": 0.49333333333333335,
"grad_norm": 2.1086196899414062,
"learning_rate": 9.278143856433503e-06,
"loss": 1.1164,
"step": 111
},
{
"epoch": 0.49777777777777776,
"grad_norm": 2.419229507446289,
"learning_rate": 9.258188497795093e-06,
"loss": 1.1713,
"step": 112
},
{
"epoch": 0.5022222222222222,
"grad_norm": 2.4053385257720947,
"learning_rate": 9.237983130647973e-06,
"loss": 1.1185,
"step": 113
},
{
"epoch": 0.5066666666666667,
"grad_norm": 2.489058017730713,
"learning_rate": 9.217528941297942e-06,
"loss": 1.168,
"step": 114
},
{
"epoch": 0.5111111111111111,
"grad_norm": 2.2254421710968018,
"learning_rate": 9.196827130659752e-06,
"loss": 1.1715,
"step": 115
},
{
"epoch": 0.5155555555555555,
"grad_norm": 2.17338228225708,
"learning_rate": 9.175878914186591e-06,
"loss": 1.1486,
"step": 116
},
{
"epoch": 0.52,
"grad_norm": 3.704568386077881,
"learning_rate": 9.154685521798736e-06,
"loss": 1.2477,
"step": 117
},
{
"epoch": 0.5244444444444445,
"grad_norm": 2.340566396713257,
"learning_rate": 9.13324819781133e-06,
"loss": 1.1748,
"step": 118
},
{
"epoch": 0.5288888888888889,
"grad_norm": 2.876037359237671,
"learning_rate": 9.111568200861324e-06,
"loss": 1.1003,
"step": 119
},
{
"epoch": 0.5333333333333333,
"grad_norm": 1.6906325817108154,
"learning_rate": 9.089646803833589e-06,
"loss": 1.0845,
"step": 120
},
{
"epoch": 0.5377777777777778,
"grad_norm": 2.1755309104919434,
"learning_rate": 9.067485293786173e-06,
"loss": 1.2046,
"step": 121
},
{
"epoch": 0.5422222222222223,
"grad_norm": 2.471137046813965,
"learning_rate": 9.045084971874738e-06,
"loss": 1.1749,
"step": 122
},
{
"epoch": 0.5466666666666666,
"grad_norm": 2.135450839996338,
"learning_rate": 9.022447153276169e-06,
"loss": 1.1702,
"step": 123
},
{
"epoch": 0.5511111111111111,
"grad_norm": 2.0650486946105957,
"learning_rate": 8.99957316711135e-06,
"loss": 1.1009,
"step": 124
},
{
"epoch": 0.5555555555555556,
"grad_norm": 2.5196189880371094,
"learning_rate": 8.976464356367133e-06,
"loss": 1.1063,
"step": 125
},
{
"epoch": 0.56,
"grad_norm": 2.1774826049804688,
"learning_rate": 8.95312207781749e-06,
"loss": 1.1902,
"step": 126
},
{
"epoch": 0.5644444444444444,
"grad_norm": 2.327805757522583,
"learning_rate": 8.929547701943849e-06,
"loss": 1.1152,
"step": 127
},
{
"epoch": 0.5688888888888889,
"grad_norm": 3.271408796310425,
"learning_rate": 8.905742612854628e-06,
"loss": 1.1846,
"step": 128
},
{
"epoch": 0.5733333333333334,
"grad_norm": 2.898059844970703,
"learning_rate": 8.881708208203977e-06,
"loss": 1.16,
"step": 129
},
{
"epoch": 0.5777777777777777,
"grad_norm": 2.6147210597991943,
"learning_rate": 8.857445899109716e-06,
"loss": 1.12,
"step": 130
},
{
"epoch": 0.5822222222222222,
"grad_norm": 3.8312957286834717,
"learning_rate": 8.832957110070482e-06,
"loss": 1.1535,
"step": 131
},
{
"epoch": 0.5866666666666667,
"grad_norm": 3.3056952953338623,
"learning_rate": 8.808243278882094e-06,
"loss": 1.0797,
"step": 132
},
{
"epoch": 0.5911111111111111,
"grad_norm": 3.1664161682128906,
"learning_rate": 8.783305856553143e-06,
"loss": 1.1774,
"step": 133
},
{
"epoch": 0.5955555555555555,
"grad_norm": 2.9546732902526855,
"learning_rate": 8.758146307219793e-06,
"loss": 1.213,
"step": 134
},
{
"epoch": 0.6,
"grad_norm": 1.9610460996627808,
"learning_rate": 8.732766108059814e-06,
"loss": 1.167,
"step": 135
},
{
"epoch": 0.6044444444444445,
"grad_norm": 2.216212749481201,
"learning_rate": 8.707166749205867e-06,
"loss": 1.0953,
"step": 136
},
{
"epoch": 0.6088888888888889,
"grad_norm": 5.400648593902588,
"learning_rate": 8.681349733658002e-06,
"loss": 1.085,
"step": 137
},
{
"epoch": 0.6133333333333333,
"grad_norm": 2.8066205978393555,
"learning_rate": 8.65531657719542e-06,
"loss": 1.1447,
"step": 138
},
{
"epoch": 0.6177777777777778,
"grad_norm": 3.9573774337768555,
"learning_rate": 8.629068808287476e-06,
"loss": 1.1372,
"step": 139
},
{
"epoch": 0.6222222222222222,
"grad_norm": 2.352977991104126,
"learning_rate": 8.602607968003935e-06,
"loss": 1.1005,
"step": 140
},
{
"epoch": 0.6266666666666667,
"grad_norm": 3.9835476875305176,
"learning_rate": 8.575935609924505e-06,
"loss": 1.1974,
"step": 141
},
{
"epoch": 0.6311111111111111,
"grad_norm": 2.4645981788635254,
"learning_rate": 8.549053300047602e-06,
"loss": 1.0944,
"step": 142
},
{
"epoch": 0.6355555555555555,
"grad_norm": 1.7835776805877686,
"learning_rate": 8.521962616698428e-06,
"loss": 1.1516,
"step": 143
},
{
"epoch": 0.64,
"grad_norm": 2.5801022052764893,
"learning_rate": 8.494665150436288e-06,
"loss": 1.1421,
"step": 144
},
{
"epoch": 0.6444444444444445,
"grad_norm": 2.424285650253296,
"learning_rate": 8.467162503961209e-06,
"loss": 1.147,
"step": 145
},
{
"epoch": 0.6488888888888888,
"grad_norm": 2.7262933254241943,
"learning_rate": 8.439456292019849e-06,
"loss": 1.1269,
"step": 146
},
{
"epoch": 0.6533333333333333,
"grad_norm": 1.9822264909744263,
"learning_rate": 8.411548141310683e-06,
"loss": 1.1161,
"step": 147
},
{
"epoch": 0.6577777777777778,
"grad_norm": 1.773859977722168,
"learning_rate": 8.383439690388491e-06,
"loss": 1.0834,
"step": 148
},
{
"epoch": 0.6622222222222223,
"grad_norm": 2.605008363723755,
"learning_rate": 8.35513258956817e-06,
"loss": 1.1789,
"step": 149
},
{
"epoch": 0.6666666666666666,
"grad_norm": 2.594944477081299,
"learning_rate": 8.326628500827826e-06,
"loss": 1.1489,
"step": 150
},
{
"epoch": 0.6711111111111111,
"grad_norm": 2.652006149291992,
"learning_rate": 8.297929097711207e-06,
"loss": 1.1225,
"step": 151
},
{
"epoch": 0.6755555555555556,
"grad_norm": 3.7691638469696045,
"learning_rate": 8.269036065229426e-06,
"loss": 1.2454,
"step": 152
},
{
"epoch": 0.68,
"grad_norm": 2.865312337875366,
"learning_rate": 8.239951099762058e-06,
"loss": 1.1437,
"step": 153
},
{
"epoch": 0.6844444444444444,
"grad_norm": 2.1165366172790527,
"learning_rate": 8.210675908957513e-06,
"loss": 1.1353,
"step": 154
},
{
"epoch": 0.6888888888888889,
"grad_norm": 2.63763427734375,
"learning_rate": 8.1812122116328e-06,
"loss": 1.1393,
"step": 155
},
{
"epoch": 0.6933333333333334,
"grad_norm": 3.102932929992676,
"learning_rate": 8.151561737672591e-06,
"loss": 1.1366,
"step": 156
},
{
"epoch": 0.6977777777777778,
"grad_norm": 3.0296387672424316,
"learning_rate": 8.12172622792767e-06,
"loss": 1.0632,
"step": 157
},
{
"epoch": 0.7022222222222222,
"grad_norm": 2.647352933883667,
"learning_rate": 8.091707434112717e-06,
"loss": 1.0779,
"step": 158
},
{
"epoch": 0.7066666666666667,
"grad_norm": 3.4690728187561035,
"learning_rate": 8.061507118703456e-06,
"loss": 1.187,
"step": 159
},
{
"epoch": 0.7111111111111111,
"grad_norm": 2.914712905883789,
"learning_rate": 8.031127054833192e-06,
"loss": 1.1118,
"step": 160
},
{
"epoch": 0.7155555555555555,
"grad_norm": 2.348532199859619,
"learning_rate": 8.000569026188684e-06,
"loss": 1.1566,
"step": 161
},
{
"epoch": 0.72,
"grad_norm": 2.2149953842163086,
"learning_rate": 7.969834826905441e-06,
"loss": 1.1446,
"step": 162
},
{
"epoch": 0.7244444444444444,
"grad_norm": 1.8673402070999146,
"learning_rate": 7.938926261462366e-06,
"loss": 1.1512,
"step": 163
},
{
"epoch": 0.7288888888888889,
"grad_norm": 2.2096667289733887,
"learning_rate": 7.90784514457583e-06,
"loss": 1.0488,
"step": 164
},
{
"epoch": 0.7333333333333333,
"grad_norm": 2.886953592300415,
"learning_rate": 7.876593301093104e-06,
"loss": 1.1158,
"step": 165
},
{
"epoch": 0.7377777777777778,
"grad_norm": 3.801467180252075,
"learning_rate": 7.845172565885237e-06,
"loss": 1.1106,
"step": 166
},
{
"epoch": 0.7422222222222222,
"grad_norm": 2.3027687072753906,
"learning_rate": 7.813584783739314e-06,
"loss": 1.1146,
"step": 167
},
{
"epoch": 0.7466666666666667,
"grad_norm": 3.938884973526001,
"learning_rate": 7.78183180925015e-06,
"loss": 1.1328,
"step": 168
},
{
"epoch": 0.7511111111111111,
"grad_norm": 2.19864821434021,
"learning_rate": 7.749915506711402e-06,
"loss": 1.1701,
"step": 169
},
{
"epoch": 0.7555555555555555,
"grad_norm": 4.781134128570557,
"learning_rate": 7.717837750006106e-06,
"loss": 1.1709,
"step": 170
},
{
"epoch": 0.76,
"grad_norm": 2.3147037029266357,
"learning_rate": 7.685600422496666e-06,
"loss": 1.0917,
"step": 171
},
{
"epoch": 0.7644444444444445,
"grad_norm": 3.0294220447540283,
"learning_rate": 7.653205416914267e-06,
"loss": 1.1209,
"step": 172
},
{
"epoch": 0.7688888888888888,
"grad_norm": 1.97251296043396,
"learning_rate": 7.620654635247762e-06,
"loss": 1.0584,
"step": 173
},
{
"epoch": 0.7733333333333333,
"grad_norm": 4.131389141082764,
"learning_rate": 7.587949988631982e-06,
"loss": 1.171,
"step": 174
},
{
"epoch": 0.7777777777777778,
"grad_norm": 2.2693028450012207,
"learning_rate": 7.555093397235553e-06,
"loss": 1.1358,
"step": 175
},
{
"epoch": 0.7822222222222223,
"grad_norm": 2.0202383995056152,
"learning_rate": 7.5220867901481335e-06,
"loss": 1.1823,
"step": 176
},
{
"epoch": 0.7866666666666666,
"grad_norm": 2.4606690406799316,
"learning_rate": 7.488932105267171e-06,
"loss": 1.0837,
"step": 177
},
{
"epoch": 0.7911111111111111,
"grad_norm": 1.997691035270691,
"learning_rate": 7.455631289184117e-06,
"loss": 1.1091,
"step": 178
},
{
"epoch": 0.7955555555555556,
"grad_norm": 2.143990993499756,
"learning_rate": 7.422186297070136e-06,
"loss": 1.1174,
"step": 179
},
{
"epoch": 0.8,
"grad_norm": 3.189858913421631,
"learning_rate": 7.388599092561315e-06,
"loss": 1.1757,
"step": 180
},
{
"epoch": 0.8044444444444444,
"grad_norm": 1.796143889427185,
"learning_rate": 7.3548716476433756e-06,
"loss": 1.1114,
"step": 181
},
{
"epoch": 0.8088888888888889,
"grad_norm": 1.725069284439087,
"learning_rate": 7.32100594253589e-06,
"loss": 1.0613,
"step": 182
},
{
"epoch": 0.8133333333333334,
"grad_norm": 2.555976152420044,
"learning_rate": 7.2870039655760186e-06,
"loss": 1.0673,
"step": 183
},
{
"epoch": 0.8177777777777778,
"grad_norm": 2.573438882827759,
"learning_rate": 7.252867713101772e-06,
"loss": 1.1355,
"step": 184
},
{
"epoch": 0.8222222222222222,
"grad_norm": 2.54061222076416,
"learning_rate": 7.218599189334799e-06,
"loss": 1.1363,
"step": 185
},
{
"epoch": 0.8266666666666667,
"grad_norm": 2.8169195652008057,
"learning_rate": 7.184200406262717e-06,
"loss": 1.0974,
"step": 186
},
{
"epoch": 0.8311111111111111,
"grad_norm": 10.035176277160645,
"learning_rate": 7.149673383520978e-06,
"loss": 1.1138,
"step": 187
},
{
"epoch": 0.8355555555555556,
"grad_norm": 3.066096544265747,
"learning_rate": 7.115020148274294e-06,
"loss": 1.1197,
"step": 188
},
{
"epoch": 0.84,
"grad_norm": 2.641267776489258,
"learning_rate": 7.080242735097622e-06,
"loss": 1.1566,
"step": 189
},
{
"epoch": 0.8444444444444444,
"grad_norm": 3.2539026737213135,
"learning_rate": 7.045343185856701e-06,
"loss": 1.0788,
"step": 190
},
{
"epoch": 0.8488888888888889,
"grad_norm": 3.5505993366241455,
"learning_rate": 7.01032354958817e-06,
"loss": 1.1023,
"step": 191
},
{
"epoch": 0.8533333333333334,
"grad_norm": 2.408964157104492,
"learning_rate": 6.975185882379272e-06,
"loss": 1.1087,
"step": 192
},
{
"epoch": 0.8577777777777778,
"grad_norm": 2.967325210571289,
"learning_rate": 6.939932247247126e-06,
"loss": 1.0942,
"step": 193
},
{
"epoch": 0.8622222222222222,
"grad_norm": 1.98670494556427,
"learning_rate": 6.9045647140176145e-06,
"loss": 1.1288,
"step": 194
},
{
"epoch": 0.8666666666666667,
"grad_norm": 2.2912962436676025,
"learning_rate": 6.869085359203844e-06,
"loss": 1.1148,
"step": 195
},
{
"epoch": 0.8711111111111111,
"grad_norm": 2.9244866371154785,
"learning_rate": 6.833496265884241e-06,
"loss": 1.1047,
"step": 196
},
{
"epoch": 0.8755555555555555,
"grad_norm": 3.550187349319458,
"learning_rate": 6.79779952358024e-06,
"loss": 1.1299,
"step": 197
},
{
"epoch": 0.88,
"grad_norm": 2.448188066482544,
"learning_rate": 6.76199722813361e-06,
"loss": 1.0937,
"step": 198
},
{
"epoch": 0.8844444444444445,
"grad_norm": 2.392951488494873,
"learning_rate": 6.726091481583396e-06,
"loss": 1.1075,
"step": 199
},
{
"epoch": 0.8888888888888888,
"grad_norm": 1.764351487159729,
"learning_rate": 6.690084392042514e-06,
"loss": 1.0864,
"step": 200
},
{
"epoch": 0.8933333333333333,
"grad_norm": 2.1497156620025635,
"learning_rate": 6.653978073573962e-06,
"loss": 1.0698,
"step": 201
},
{
"epoch": 0.8977777777777778,
"grad_norm": 2.4440653324127197,
"learning_rate": 6.6177746460667124e-06,
"loss": 1.1207,
"step": 202
},
{
"epoch": 0.9022222222222223,
"grad_norm": 5.237727165222168,
"learning_rate": 6.581476235111244e-06,
"loss": 1.2957,
"step": 203
},
{
"epoch": 0.9066666666666666,
"grad_norm": 2.54868745803833,
"learning_rate": 6.545084971874738e-06,
"loss": 1.1072,
"step": 204
},
{
"epoch": 0.9111111111111111,
"grad_norm": 2.269233465194702,
"learning_rate": 6.508602992975963e-06,
"loss": 1.096,
"step": 205
},
{
"epoch": 0.9155555555555556,
"grad_norm": 1.835715413093567,
"learning_rate": 6.472032440359817e-06,
"loss": 1.0722,
"step": 206
},
{
"epoch": 0.92,
"grad_norm": 2.5577235221862793,
"learning_rate": 6.43537546117158e-06,
"loss": 1.1194,
"step": 207
},
{
"epoch": 0.9244444444444444,
"grad_norm": 2.0571846961975098,
"learning_rate": 6.398634207630841e-06,
"loss": 1.1089,
"step": 208
},
{
"epoch": 0.9288888888888889,
"grad_norm": 3.0140180587768555,
"learning_rate": 6.361810836905138e-06,
"loss": 1.2144,
"step": 209
},
{
"epoch": 0.9333333333333333,
"grad_norm": 2.731870412826538,
"learning_rate": 6.32490751098331e-06,
"loss": 1.1493,
"step": 210
},
{
"epoch": 0.9377777777777778,
"grad_norm": 2.427889347076416,
"learning_rate": 6.287926396548556e-06,
"loss": 1.0504,
"step": 211
},
{
"epoch": 0.9422222222222222,
"grad_norm": 2.487750291824341,
"learning_rate": 6.250869664851226e-06,
"loss": 1.0908,
"step": 212
},
{
"epoch": 0.9466666666666667,
"grad_norm": 2.3912594318389893,
"learning_rate": 6.213739491581347e-06,
"loss": 1.1216,
"step": 213
},
{
"epoch": 0.9511111111111111,
"grad_norm": 3.4675991535186768,
"learning_rate": 6.176538056740871e-06,
"loss": 1.0171,
"step": 214
},
{
"epoch": 0.9555555555555556,
"grad_norm": 1.81716787815094,
"learning_rate": 6.139267544515689e-06,
"loss": 1.044,
"step": 215
},
{
"epoch": 0.96,
"grad_norm": 2.6640546321868896,
"learning_rate": 6.101930143147395e-06,
"loss": 1.1029,
"step": 216
},
{
"epoch": 0.9644444444444444,
"grad_norm": 2.501460313796997,
"learning_rate": 6.064528044804805e-06,
"loss": 1.1141,
"step": 217
},
{
"epoch": 0.9688888888888889,
"grad_norm": 2.3960464000701904,
"learning_rate": 6.0270634454552494e-06,
"loss": 1.0526,
"step": 218
},
{
"epoch": 0.9733333333333334,
"grad_norm": 2.989225387573242,
"learning_rate": 5.989538544735644e-06,
"loss": 1.1437,
"step": 219
},
{
"epoch": 0.9777777777777777,
"grad_norm": 1.963538646697998,
"learning_rate": 5.951955545823342e-06,
"loss": 1.0998,
"step": 220
},
{
"epoch": 0.9822222222222222,
"grad_norm": 2.2665040493011475,
"learning_rate": 5.9143166553067846e-06,
"loss": 1.1062,
"step": 221
},
{
"epoch": 0.9866666666666667,
"grad_norm": 2.201535701751709,
"learning_rate": 5.87662408305594e-06,
"loss": 1.042,
"step": 222
},
{
"epoch": 0.9911111111111112,
"grad_norm": 2.513442039489746,
"learning_rate": 5.8388800420925616e-06,
"loss": 1.0733,
"step": 223
},
{
"epoch": 0.9955555555555555,
"grad_norm": 2.0468199253082275,
"learning_rate": 5.801086748460255e-06,
"loss": 1.1171,
"step": 224
},
{
"epoch": 1.0,
"grad_norm": 3.2198939323425293,
"learning_rate": 5.763246421094373e-06,
"loss": 1.1027,
"step": 225
}
],
"logging_steps": 1,
"max_steps": 450,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 225,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 8.193854425949798e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}