checkpoint-170 / trainer_state.json
XuyaoWang's picture
Upload folder using huggingface_hub
4c20b9a verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 170,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.029411764705882353,
"grad_norm": 5.318241315636508,
"learning_rate": 5.882352941176471e-07,
"loss": 1.2574,
"step": 1
},
{
"epoch": 0.058823529411764705,
"grad_norm": 5.457430759320893,
"learning_rate": 1.1764705882352942e-06,
"loss": 1.2653,
"step": 2
},
{
"epoch": 0.08823529411764706,
"grad_norm": 5.122370897835391,
"learning_rate": 1.7647058823529414e-06,
"loss": 1.2381,
"step": 3
},
{
"epoch": 0.11764705882352941,
"grad_norm": 4.942746835965387,
"learning_rate": 2.3529411764705885e-06,
"loss": 1.1866,
"step": 4
},
{
"epoch": 0.14705882352941177,
"grad_norm": 4.38157157995271,
"learning_rate": 2.9411764705882355e-06,
"loss": 1.1738,
"step": 5
},
{
"epoch": 0.17647058823529413,
"grad_norm": 3.9400034435555047,
"learning_rate": 3.529411764705883e-06,
"loss": 1.1143,
"step": 6
},
{
"epoch": 0.20588235294117646,
"grad_norm": 2.453696487436283,
"learning_rate": 4.11764705882353e-06,
"loss": 1.1093,
"step": 7
},
{
"epoch": 0.23529411764705882,
"grad_norm": 2.230814314539596,
"learning_rate": 4.705882352941177e-06,
"loss": 1.0627,
"step": 8
},
{
"epoch": 0.2647058823529412,
"grad_norm": 2.1481810875048635,
"learning_rate": 5.294117647058824e-06,
"loss": 1.0824,
"step": 9
},
{
"epoch": 0.29411764705882354,
"grad_norm": 2.8344316487833536,
"learning_rate": 5.882352941176471e-06,
"loss": 1.1828,
"step": 10
},
{
"epoch": 0.3235294117647059,
"grad_norm": 2.4435138857634113,
"learning_rate": 6.470588235294119e-06,
"loss": 1.1263,
"step": 11
},
{
"epoch": 0.35294117647058826,
"grad_norm": 2.2241980945396187,
"learning_rate": 7.058823529411766e-06,
"loss": 1.1104,
"step": 12
},
{
"epoch": 0.38235294117647056,
"grad_norm": 2.3189562810467175,
"learning_rate": 7.647058823529411e-06,
"loss": 1.0238,
"step": 13
},
{
"epoch": 0.4117647058823529,
"grad_norm": 2.3264631769942863,
"learning_rate": 8.23529411764706e-06,
"loss": 1.1413,
"step": 14
},
{
"epoch": 0.4411764705882353,
"grad_norm": 1.6883233037452305,
"learning_rate": 8.823529411764707e-06,
"loss": 1.0494,
"step": 15
},
{
"epoch": 0.47058823529411764,
"grad_norm": 1.43620603692946,
"learning_rate": 9.411764705882354e-06,
"loss": 1.1637,
"step": 16
},
{
"epoch": 0.5,
"grad_norm": 1.6156970232652779,
"learning_rate": 1e-05,
"loss": 1.1306,
"step": 17
},
{
"epoch": 0.5294117647058824,
"grad_norm": 1.6157860447579155,
"learning_rate": 9.998945997517957e-06,
"loss": 0.8976,
"step": 18
},
{
"epoch": 0.5588235294117647,
"grad_norm": 1.2741201300727243,
"learning_rate": 9.99578443444032e-06,
"loss": 1.0067,
"step": 19
},
{
"epoch": 0.5882352941176471,
"grad_norm": 1.287709582244065,
"learning_rate": 9.990516643685222e-06,
"loss": 1.0362,
"step": 20
},
{
"epoch": 0.6176470588235294,
"grad_norm": 1.2670402021830112,
"learning_rate": 9.983144846158472e-06,
"loss": 1.0597,
"step": 21
},
{
"epoch": 0.6470588235294118,
"grad_norm": 1.1468391105481275,
"learning_rate": 9.973672149817232e-06,
"loss": 0.9466,
"step": 22
},
{
"epoch": 0.6764705882352942,
"grad_norm": 1.2405872721167117,
"learning_rate": 9.96210254835968e-06,
"loss": 1.0102,
"step": 23
},
{
"epoch": 0.7058823529411765,
"grad_norm": 1.0405159898011183,
"learning_rate": 9.948440919541277e-06,
"loss": 1.013,
"step": 24
},
{
"epoch": 0.7352941176470589,
"grad_norm": 0.988631297941715,
"learning_rate": 9.932693023118299e-06,
"loss": 1.1261,
"step": 25
},
{
"epoch": 0.7647058823529411,
"grad_norm": 1.0106547826681243,
"learning_rate": 9.91486549841951e-06,
"loss": 1.0004,
"step": 26
},
{
"epoch": 0.7941176470588235,
"grad_norm": 0.9186647940375248,
"learning_rate": 9.894965861547023e-06,
"loss": 1.0681,
"step": 27
},
{
"epoch": 0.8235294117647058,
"grad_norm": 0.9101484018960783,
"learning_rate": 9.873002502207502e-06,
"loss": 0.9657,
"step": 28
},
{
"epoch": 0.8529411764705882,
"grad_norm": 0.9351878539175523,
"learning_rate": 9.848984680175049e-06,
"loss": 1.002,
"step": 29
},
{
"epoch": 0.8823529411764706,
"grad_norm": 0.9103433041975992,
"learning_rate": 9.822922521387277e-06,
"loss": 1.0379,
"step": 30
},
{
"epoch": 0.9117647058823529,
"grad_norm": 0.8488744288909514,
"learning_rate": 9.794827013676206e-06,
"loss": 0.9999,
"step": 31
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.8547096036420078,
"learning_rate": 9.764710002135784e-06,
"loss": 1.107,
"step": 32
},
{
"epoch": 0.9705882352941176,
"grad_norm": 0.8832505570569958,
"learning_rate": 9.732584184127973e-06,
"loss": 0.9733,
"step": 33
},
{
"epoch": 1.0,
"grad_norm": 0.8177547975613774,
"learning_rate": 9.698463103929542e-06,
"loss": 0.9918,
"step": 34
},
{
"epoch": 1.0294117647058822,
"grad_norm": 0.7788115313989687,
"learning_rate": 9.66236114702178e-06,
"loss": 0.8969,
"step": 35
},
{
"epoch": 1.0588235294117647,
"grad_norm": 0.8247936301981957,
"learning_rate": 9.62429353402556e-06,
"loss": 1.001,
"step": 36
},
{
"epoch": 1.088235294117647,
"grad_norm": 0.7273884864758107,
"learning_rate": 9.584276314284316e-06,
"loss": 0.9413,
"step": 37
},
{
"epoch": 1.1176470588235294,
"grad_norm": 0.7371763227953478,
"learning_rate": 9.542326359097619e-06,
"loss": 0.8537,
"step": 38
},
{
"epoch": 1.1470588235294117,
"grad_norm": 0.8141348723393264,
"learning_rate": 9.498461354608228e-06,
"loss": 0.8322,
"step": 39
},
{
"epoch": 1.1764705882352942,
"grad_norm": 0.7149246736159542,
"learning_rate": 9.452699794345583e-06,
"loss": 0.9635,
"step": 40
},
{
"epoch": 1.2058823529411764,
"grad_norm": 0.821497729928177,
"learning_rate": 9.405060971428924e-06,
"loss": 0.9376,
"step": 41
},
{
"epoch": 1.2352941176470589,
"grad_norm": 0.851059564164393,
"learning_rate": 9.355564970433288e-06,
"loss": 0.9014,
"step": 42
},
{
"epoch": 1.2647058823529411,
"grad_norm": 0.7213011278134036,
"learning_rate": 9.30423265892184e-06,
"loss": 0.8558,
"step": 43
},
{
"epoch": 1.2941176470588236,
"grad_norm": 0.7497486871587185,
"learning_rate": 9.251085678648072e-06,
"loss": 0.8632,
"step": 44
},
{
"epoch": 1.3235294117647058,
"grad_norm": 0.7922961890052337,
"learning_rate": 9.196146436431635e-06,
"loss": 0.9467,
"step": 45
},
{
"epoch": 1.3529411764705883,
"grad_norm": 0.815700313616905,
"learning_rate": 9.13943809471159e-06,
"loss": 0.9879,
"step": 46
},
{
"epoch": 1.3823529411764706,
"grad_norm": 0.7713577153963318,
"learning_rate": 9.08098456178111e-06,
"loss": 0.9351,
"step": 47
},
{
"epoch": 1.4117647058823528,
"grad_norm": 0.715482838419465,
"learning_rate": 9.020810481707709e-06,
"loss": 0.8108,
"step": 48
},
{
"epoch": 1.4411764705882353,
"grad_norm": 1.021755928580874,
"learning_rate": 8.958941223943292e-06,
"loss": 0.9631,
"step": 49
},
{
"epoch": 1.4705882352941178,
"grad_norm": 0.6996310156748503,
"learning_rate": 8.895402872628352e-06,
"loss": 0.9475,
"step": 50
},
{
"epoch": 1.5,
"grad_norm": 0.741220539859401,
"learning_rate": 8.83022221559489e-06,
"loss": 0.9227,
"step": 51
},
{
"epoch": 1.5294117647058822,
"grad_norm": 0.8915705207795342,
"learning_rate": 8.763426733072624e-06,
"loss": 0.9589,
"step": 52
},
{
"epoch": 1.5588235294117647,
"grad_norm": 0.8999233325654797,
"learning_rate": 8.695044586103297e-06,
"loss": 0.9648,
"step": 53
},
{
"epoch": 1.5882352941176472,
"grad_norm": 0.7584193738597184,
"learning_rate": 8.625104604667965e-06,
"loss": 0.8171,
"step": 54
},
{
"epoch": 1.6176470588235294,
"grad_norm": 0.9706095097818166,
"learning_rate": 8.553636275532236e-06,
"loss": 0.8898,
"step": 55
},
{
"epoch": 1.6470588235294117,
"grad_norm": 0.7765692251989467,
"learning_rate": 8.480669729814635e-06,
"loss": 0.8235,
"step": 56
},
{
"epoch": 1.6764705882352942,
"grad_norm": 0.9119169437229295,
"learning_rate": 8.40623573028327e-06,
"loss": 1.0011,
"step": 57
},
{
"epoch": 1.7058823529411766,
"grad_norm": 0.8846080365528076,
"learning_rate": 8.330365658386252e-06,
"loss": 0.9523,
"step": 58
},
{
"epoch": 1.7352941176470589,
"grad_norm": 0.7810465665711979,
"learning_rate": 8.25309150102121e-06,
"loss": 0.9193,
"step": 59
},
{
"epoch": 1.7647058823529411,
"grad_norm": 0.6840392531400987,
"learning_rate": 8.174445837049614e-06,
"loss": 0.9383,
"step": 60
},
{
"epoch": 1.7941176470588234,
"grad_norm": 0.7186778991653127,
"learning_rate": 8.094461823561473e-06,
"loss": 0.9286,
"step": 61
},
{
"epoch": 1.8235294117647058,
"grad_norm": 0.7086123945614116,
"learning_rate": 8.013173181896283e-06,
"loss": 0.9594,
"step": 62
},
{
"epoch": 1.8529411764705883,
"grad_norm": 0.8777147669457216,
"learning_rate": 7.930614183426074e-06,
"loss": 0.8592,
"step": 63
},
{
"epoch": 1.8823529411764706,
"grad_norm": 0.7021315919788967,
"learning_rate": 7.846819635106569e-06,
"loss": 0.8997,
"step": 64
},
{
"epoch": 1.9117647058823528,
"grad_norm": 0.7669181704786203,
"learning_rate": 7.76182486480253e-06,
"loss": 0.8705,
"step": 65
},
{
"epoch": 1.9411764705882353,
"grad_norm": 0.7415265424714063,
"learning_rate": 7.675665706393502e-06,
"loss": 0.8394,
"step": 66
},
{
"epoch": 1.9705882352941178,
"grad_norm": 0.6616287892937567,
"learning_rate": 7.588378484666214e-06,
"loss": 0.7859,
"step": 67
},
{
"epoch": 2.0,
"grad_norm": 0.6308863230543723,
"learning_rate": 7.500000000000001e-06,
"loss": 0.9024,
"step": 68
},
{
"epoch": 2.0294117647058822,
"grad_norm": 0.751568095417785,
"learning_rate": 7.4105675128517456e-06,
"loss": 0.9108,
"step": 69
},
{
"epoch": 2.0588235294117645,
"grad_norm": 0.7056219692093446,
"learning_rate": 7.320118728046818e-06,
"loss": 0.8944,
"step": 70
},
{
"epoch": 2.088235294117647,
"grad_norm": 0.6840607587078892,
"learning_rate": 7.2286917788826926e-06,
"loss": 0.8664,
"step": 71
},
{
"epoch": 2.1176470588235294,
"grad_norm": 0.7686260079297964,
"learning_rate": 7.136325211051905e-06,
"loss": 0.8624,
"step": 72
},
{
"epoch": 2.1470588235294117,
"grad_norm": 0.6880672461301692,
"learning_rate": 7.043057966391158e-06,
"loss": 0.899,
"step": 73
},
{
"epoch": 2.176470588235294,
"grad_norm": 0.6818887749750221,
"learning_rate": 6.948929366463397e-06,
"loss": 0.882,
"step": 74
},
{
"epoch": 2.2058823529411766,
"grad_norm": 0.653556285710504,
"learning_rate": 6.8539790959798045e-06,
"loss": 0.7438,
"step": 75
},
{
"epoch": 2.235294117647059,
"grad_norm": 0.75928710261434,
"learning_rate": 6.758247186068684e-06,
"loss": 0.7844,
"step": 76
},
{
"epoch": 2.264705882352941,
"grad_norm": 0.6167440243179065,
"learning_rate": 6.6617739973982985e-06,
"loss": 0.7404,
"step": 77
},
{
"epoch": 2.2941176470588234,
"grad_norm": 0.8360406813068504,
"learning_rate": 6.5646002031607726e-06,
"loss": 0.8031,
"step": 78
},
{
"epoch": 2.323529411764706,
"grad_norm": 0.7649411038527903,
"learning_rate": 6.466766771924231e-06,
"loss": 0.7445,
"step": 79
},
{
"epoch": 2.3529411764705883,
"grad_norm": 0.73162135940835,
"learning_rate": 6.368314950360416e-06,
"loss": 0.7933,
"step": 80
},
{
"epoch": 2.3823529411764706,
"grad_norm": 0.7584559428273998,
"learning_rate": 6.269286245855039e-06,
"loss": 0.8263,
"step": 81
},
{
"epoch": 2.411764705882353,
"grad_norm": 0.7004962389388844,
"learning_rate": 6.169722409008244e-06,
"loss": 0.8664,
"step": 82
},
{
"epoch": 2.4411764705882355,
"grad_norm": 0.5958812852500521,
"learning_rate": 6.0696654160324875e-06,
"loss": 0.847,
"step": 83
},
{
"epoch": 2.4705882352941178,
"grad_norm": 0.7521440026657608,
"learning_rate": 5.9691574510553505e-06,
"loss": 0.7147,
"step": 84
},
{
"epoch": 2.5,
"grad_norm": 0.7266061294049975,
"learning_rate": 5.8682408883346535e-06,
"loss": 0.9203,
"step": 85
},
{
"epoch": 2.5294117647058822,
"grad_norm": 0.7265821759271923,
"learning_rate": 5.766958274393428e-06,
"loss": 0.8468,
"step": 86
},
{
"epoch": 2.5588235294117645,
"grad_norm": 0.6841951835147332,
"learning_rate": 5.66535231008227e-06,
"loss": 0.823,
"step": 87
},
{
"epoch": 2.588235294117647,
"grad_norm": 0.6741663308556166,
"learning_rate": 5.5634658325766066e-06,
"loss": 0.8129,
"step": 88
},
{
"epoch": 2.6176470588235294,
"grad_norm": 0.6983390361641961,
"learning_rate": 5.46134179731651e-06,
"loss": 0.7445,
"step": 89
},
{
"epoch": 2.6470588235294117,
"grad_norm": 0.6425709318152082,
"learning_rate": 5.359023259896638e-06,
"loss": 0.8594,
"step": 90
},
{
"epoch": 2.6764705882352944,
"grad_norm": 0.6701317358512593,
"learning_rate": 5.2565533579139484e-06,
"loss": 0.7333,
"step": 91
},
{
"epoch": 2.7058823529411766,
"grad_norm": 0.6692919862632479,
"learning_rate": 5.153975292780852e-06,
"loss": 0.7438,
"step": 92
},
{
"epoch": 2.735294117647059,
"grad_norm": 0.7321294278976899,
"learning_rate": 5.05133231151145e-06,
"loss": 0.8914,
"step": 93
},
{
"epoch": 2.764705882352941,
"grad_norm": 0.6522585141065795,
"learning_rate": 4.948667688488552e-06,
"loss": 0.8122,
"step": 94
},
{
"epoch": 2.7941176470588234,
"grad_norm": 0.6263777175964031,
"learning_rate": 4.846024707219149e-06,
"loss": 0.7541,
"step": 95
},
{
"epoch": 2.8235294117647056,
"grad_norm": 0.6673242834932129,
"learning_rate": 4.7434466420860515e-06,
"loss": 0.7659,
"step": 96
},
{
"epoch": 2.8529411764705883,
"grad_norm": 0.6739327377674906,
"learning_rate": 4.640976740103363e-06,
"loss": 0.7908,
"step": 97
},
{
"epoch": 2.8823529411764706,
"grad_norm": 0.6654953214889198,
"learning_rate": 4.53865820268349e-06,
"loss": 0.7284,
"step": 98
},
{
"epoch": 2.911764705882353,
"grad_norm": 0.5502642793517006,
"learning_rate": 4.436534167423395e-06,
"loss": 0.7609,
"step": 99
},
{
"epoch": 2.9411764705882355,
"grad_norm": 0.6020194378784254,
"learning_rate": 4.334647689917734e-06,
"loss": 0.8331,
"step": 100
},
{
"epoch": 2.9705882352941178,
"grad_norm": 0.7424629295713845,
"learning_rate": 4.233041725606573e-06,
"loss": 0.7938,
"step": 101
},
{
"epoch": 3.0,
"grad_norm": 0.6629484785609059,
"learning_rate": 4.131759111665349e-06,
"loss": 0.7711,
"step": 102
},
{
"epoch": 3.0294117647058822,
"grad_norm": 0.6026051956900621,
"learning_rate": 4.03084254894465e-06,
"loss": 0.8273,
"step": 103
},
{
"epoch": 3.0588235294117645,
"grad_norm": 0.658236487732468,
"learning_rate": 3.930334583967514e-06,
"loss": 0.8171,
"step": 104
},
{
"epoch": 3.088235294117647,
"grad_norm": 0.5636951073854605,
"learning_rate": 3.8302775909917585e-06,
"loss": 0.7215,
"step": 105
},
{
"epoch": 3.1176470588235294,
"grad_norm": 0.598401075644721,
"learning_rate": 3.730713754144961e-06,
"loss": 0.7915,
"step": 106
},
{
"epoch": 3.1470588235294117,
"grad_norm": 0.6412364136287656,
"learning_rate": 3.6316850496395863e-06,
"loss": 0.66,
"step": 107
},
{
"epoch": 3.176470588235294,
"grad_norm": 0.7569448576920692,
"learning_rate": 3.5332332280757706e-06,
"loss": 0.7335,
"step": 108
},
{
"epoch": 3.2058823529411766,
"grad_norm": 0.5829777613775653,
"learning_rate": 3.4353997968392295e-06,
"loss": 0.7929,
"step": 109
},
{
"epoch": 3.235294117647059,
"grad_norm": 0.6108245548786231,
"learning_rate": 3.3382260026017027e-06,
"loss": 0.7261,
"step": 110
},
{
"epoch": 3.264705882352941,
"grad_norm": 0.6404400909526411,
"learning_rate": 3.241752813931316e-06,
"loss": 0.7325,
"step": 111
},
{
"epoch": 3.2941176470588234,
"grad_norm": 0.6152524258702544,
"learning_rate": 3.1460209040201967e-06,
"loss": 0.7889,
"step": 112
},
{
"epoch": 3.323529411764706,
"grad_norm": 0.668579758825565,
"learning_rate": 3.0510706335366034e-06,
"loss": 0.5724,
"step": 113
},
{
"epoch": 3.3529411764705883,
"grad_norm": 0.595881338350268,
"learning_rate": 2.956942033608843e-06,
"loss": 0.7365,
"step": 114
},
{
"epoch": 3.3823529411764706,
"grad_norm": 0.6930967395148463,
"learning_rate": 2.863674788948097e-06,
"loss": 0.7538,
"step": 115
},
{
"epoch": 3.411764705882353,
"grad_norm": 0.7260532087043763,
"learning_rate": 2.771308221117309e-06,
"loss": 0.6676,
"step": 116
},
{
"epoch": 3.4411764705882355,
"grad_norm": 0.6216831086120721,
"learning_rate": 2.6798812719531843e-06,
"loss": 0.7755,
"step": 117
},
{
"epoch": 3.4705882352941178,
"grad_norm": 0.5497545572594094,
"learning_rate": 2.5894324871482557e-06,
"loss": 0.6816,
"step": 118
},
{
"epoch": 3.5,
"grad_norm": 0.5913808667188861,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.7633,
"step": 119
},
{
"epoch": 3.5294117647058822,
"grad_norm": 0.6385542994883869,
"learning_rate": 2.411621515333788e-06,
"loss": 0.7819,
"step": 120
},
{
"epoch": 3.5588235294117645,
"grad_norm": 0.5677339327237313,
"learning_rate": 2.324334293606499e-06,
"loss": 0.8098,
"step": 121
},
{
"epoch": 3.588235294117647,
"grad_norm": 0.6153392042757299,
"learning_rate": 2.238175135197471e-06,
"loss": 0.7186,
"step": 122
},
{
"epoch": 3.6176470588235294,
"grad_norm": 0.6583273380044558,
"learning_rate": 2.1531803648934333e-06,
"loss": 0.7443,
"step": 123
},
{
"epoch": 3.6470588235294117,
"grad_norm": 0.6429891399875521,
"learning_rate": 2.069385816573928e-06,
"loss": 0.7676,
"step": 124
},
{
"epoch": 3.6764705882352944,
"grad_norm": 0.5722615124025376,
"learning_rate": 1.9868268181037186e-06,
"loss": 0.6626,
"step": 125
},
{
"epoch": 3.7058823529411766,
"grad_norm": 0.5583251229881766,
"learning_rate": 1.9055381764385272e-06,
"loss": 0.8078,
"step": 126
},
{
"epoch": 3.735294117647059,
"grad_norm": 0.7782753111255696,
"learning_rate": 1.8255541629503865e-06,
"loss": 0.6978,
"step": 127
},
{
"epoch": 3.764705882352941,
"grad_norm": 0.6136307715058319,
"learning_rate": 1.746908498978791e-06,
"loss": 0.766,
"step": 128
},
{
"epoch": 3.7941176470588234,
"grad_norm": 0.6017776226087933,
"learning_rate": 1.6696343416137495e-06,
"loss": 0.8564,
"step": 129
},
{
"epoch": 3.8235294117647056,
"grad_norm": 5.645152034778923,
"learning_rate": 1.5937642697167288e-06,
"loss": 0.7281,
"step": 130
},
{
"epoch": 3.8529411764705883,
"grad_norm": 0.5843508931001282,
"learning_rate": 1.5193302701853674e-06,
"loss": 0.7521,
"step": 131
},
{
"epoch": 3.8823529411764706,
"grad_norm": 0.6807269875732602,
"learning_rate": 1.4463637244677648e-06,
"loss": 0.7041,
"step": 132
},
{
"epoch": 3.911764705882353,
"grad_norm": 0.6608937838150601,
"learning_rate": 1.374895395332037e-06,
"loss": 0.7889,
"step": 133
},
{
"epoch": 3.9411764705882355,
"grad_norm": 0.5595353160353056,
"learning_rate": 1.3049554138967052e-06,
"loss": 0.7123,
"step": 134
},
{
"epoch": 3.9705882352941178,
"grad_norm": 0.5524171514160743,
"learning_rate": 1.2365732669273778e-06,
"loss": 0.7066,
"step": 135
},
{
"epoch": 4.0,
"grad_norm": 0.5179948723237111,
"learning_rate": 1.1697777844051105e-06,
"loss": 0.7855,
"step": 136
},
{
"epoch": 4.029411764705882,
"grad_norm": 0.5460401923029647,
"learning_rate": 1.1045971273716476e-06,
"loss": 0.6795,
"step": 137
},
{
"epoch": 4.0588235294117645,
"grad_norm": 0.5757389756240946,
"learning_rate": 1.0410587760567104e-06,
"loss": 0.6672,
"step": 138
},
{
"epoch": 4.088235294117647,
"grad_norm": 0.5086031243463442,
"learning_rate": 9.791895182922911e-07,
"loss": 0.6856,
"step": 139
},
{
"epoch": 4.117647058823529,
"grad_norm": 0.5562476722385284,
"learning_rate": 9.190154382188921e-07,
"loss": 0.7066,
"step": 140
},
{
"epoch": 4.147058823529412,
"grad_norm": 0.5522802181901002,
"learning_rate": 8.605619052884106e-07,
"loss": 0.6206,
"step": 141
},
{
"epoch": 4.176470588235294,
"grad_norm": 0.5449924779039531,
"learning_rate": 8.03853563568367e-07,
"loss": 0.7474,
"step": 142
},
{
"epoch": 4.205882352941177,
"grad_norm": 0.5653638609920547,
"learning_rate": 7.489143213519301e-07,
"loss": 0.741,
"step": 143
},
{
"epoch": 4.235294117647059,
"grad_norm": 0.5361598210599389,
"learning_rate": 6.957673410781617e-07,
"loss": 0.8035,
"step": 144
},
{
"epoch": 4.264705882352941,
"grad_norm": 0.5215258721188443,
"learning_rate": 6.444350295667112e-07,
"loss": 0.7208,
"step": 145
},
{
"epoch": 4.294117647058823,
"grad_norm": 0.4854309284329868,
"learning_rate": 5.949390285710777e-07,
"loss": 0.8025,
"step": 146
},
{
"epoch": 4.323529411764706,
"grad_norm": 0.5317945849506307,
"learning_rate": 5.473002056544191e-07,
"loss": 0.6857,
"step": 147
},
{
"epoch": 4.352941176470588,
"grad_norm": 0.5817093639383343,
"learning_rate": 5.015386453917742e-07,
"loss": 0.675,
"step": 148
},
{
"epoch": 4.382352941176471,
"grad_norm": 0.49920997210952905,
"learning_rate": 4.576736409023813e-07,
"loss": 0.7376,
"step": 149
},
{
"epoch": 4.411764705882353,
"grad_norm": 0.6146123548983002,
"learning_rate": 4.15723685715686e-07,
"loss": 0.7194,
"step": 150
},
{
"epoch": 4.4411764705882355,
"grad_norm": 0.5311364766405502,
"learning_rate": 3.7570646597444196e-07,
"loss": 0.7464,
"step": 151
},
{
"epoch": 4.470588235294118,
"grad_norm": 0.5378627651929206,
"learning_rate": 3.3763885297822153e-07,
"loss": 0.7662,
"step": 152
},
{
"epoch": 4.5,
"grad_norm": 0.5148080302477278,
"learning_rate": 3.015368960704584e-07,
"loss": 0.6089,
"step": 153
},
{
"epoch": 4.529411764705882,
"grad_norm": 0.5437158601735758,
"learning_rate": 2.6741581587202747e-07,
"loss": 0.6111,
"step": 154
},
{
"epoch": 4.5588235294117645,
"grad_norm": 0.5462747342783272,
"learning_rate": 2.3528999786421758e-07,
"loss": 0.6778,
"step": 155
},
{
"epoch": 4.588235294117647,
"grad_norm": 0.5403089338530145,
"learning_rate": 2.0517298632379445e-07,
"loss": 0.7225,
"step": 156
},
{
"epoch": 4.617647058823529,
"grad_norm": 0.5445908692397746,
"learning_rate": 1.770774786127244e-07,
"loss": 0.6131,
"step": 157
},
{
"epoch": 4.647058823529412,
"grad_norm": 0.5446950433509009,
"learning_rate": 1.510153198249531e-07,
"loss": 0.7269,
"step": 158
},
{
"epoch": 4.676470588235294,
"grad_norm": 0.5503240223281384,
"learning_rate": 1.2699749779249926e-07,
"loss": 0.7369,
"step": 159
},
{
"epoch": 4.705882352941177,
"grad_norm": 0.6220986168710282,
"learning_rate": 1.0503413845297739e-07,
"loss": 0.7636,
"step": 160
},
{
"epoch": 4.735294117647059,
"grad_norm": 0.5050515958802209,
"learning_rate": 8.513450158049109e-08,
"loss": 0.7519,
"step": 161
},
{
"epoch": 4.764705882352941,
"grad_norm": 0.4984049802546836,
"learning_rate": 6.730697688170251e-08,
"loss": 0.8161,
"step": 162
},
{
"epoch": 4.794117647058823,
"grad_norm": 0.5089245402255683,
"learning_rate": 5.155908045872349e-08,
"loss": 0.6582,
"step": 163
},
{
"epoch": 4.823529411764706,
"grad_norm": 0.5550922648989051,
"learning_rate": 3.7897451640321326e-08,
"loss": 0.7096,
"step": 164
},
{
"epoch": 4.852941176470588,
"grad_norm": 0.5577193029080871,
"learning_rate": 2.6327850182769065e-08,
"loss": 0.6734,
"step": 165
},
{
"epoch": 4.882352941176471,
"grad_norm": 0.5403837151621453,
"learning_rate": 1.6855153841527915e-08,
"loss": 0.8296,
"step": 166
},
{
"epoch": 4.911764705882353,
"grad_norm": 0.5647120937538365,
"learning_rate": 9.48335631477948e-09,
"loss": 0.7355,
"step": 167
},
{
"epoch": 4.9411764705882355,
"grad_norm": 0.6037964283320547,
"learning_rate": 4.2155655596809455e-09,
"loss": 0.6639,
"step": 168
},
{
"epoch": 4.970588235294118,
"grad_norm": 0.5310856829158342,
"learning_rate": 1.054002482043237e-09,
"loss": 0.6789,
"step": 169
},
{
"epoch": 5.0,
"grad_norm": 0.48150867750313575,
"learning_rate": 0.0,
"loss": 0.7042,
"step": 170
}
],
"logging_steps": 1.0,
"max_steps": 170,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 30,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 215979230625792.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}