minicpmv4_5-lora-sft / trainer_state.json
JuliusFx's picture
Upload folder using huggingface_hub
3ac0b81 verified
Raw
History Blame Contribute Delete
44.3 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 6.0,
"eval_steps": 500,
"global_step": 252,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.023809523809523808,
"grad_norm": 0.28116118907928467,
"learning_rate": 0.0,
"loss": 1.1129,
"step": 1
},
{
"epoch": 0.047619047619047616,
"grad_norm": 0.29382961988449097,
"learning_rate": 1.153846153846154e-06,
"loss": 1.1202,
"step": 2
},
{
"epoch": 0.07142857142857142,
"grad_norm": 0.276279091835022,
"learning_rate": 2.307692307692308e-06,
"loss": 1.1401,
"step": 3
},
{
"epoch": 0.09523809523809523,
"grad_norm": 0.29106155037879944,
"learning_rate": 3.4615384615384617e-06,
"loss": 1.1858,
"step": 4
},
{
"epoch": 0.11904761904761904,
"grad_norm": 0.2803880274295807,
"learning_rate": 4.615384615384616e-06,
"loss": 1.1457,
"step": 5
},
{
"epoch": 0.14285714285714285,
"grad_norm": 0.28456971049308777,
"learning_rate": 5.76923076923077e-06,
"loss": 1.0852,
"step": 6
},
{
"epoch": 0.16666666666666666,
"grad_norm": 0.2712933123111725,
"learning_rate": 6.923076923076923e-06,
"loss": 1.0493,
"step": 7
},
{
"epoch": 0.19047619047619047,
"grad_norm": 0.27702414989471436,
"learning_rate": 8.076923076923077e-06,
"loss": 1.0446,
"step": 8
},
{
"epoch": 0.21428571428571427,
"grad_norm": 0.2814911901950836,
"learning_rate": 9.230769230769232e-06,
"loss": 1.0696,
"step": 9
},
{
"epoch": 0.23809523809523808,
"grad_norm": 0.2962207794189453,
"learning_rate": 1.0384615384615384e-05,
"loss": 1.0993,
"step": 10
},
{
"epoch": 0.2619047619047619,
"grad_norm": 0.3378692865371704,
"learning_rate": 1.153846153846154e-05,
"loss": 1.127,
"step": 11
},
{
"epoch": 0.2857142857142857,
"grad_norm": 0.2905134856700897,
"learning_rate": 1.2692307692307693e-05,
"loss": 1.1312,
"step": 12
},
{
"epoch": 0.30952380952380953,
"grad_norm": 0.31969478726387024,
"learning_rate": 1.3846153846153847e-05,
"loss": 1.1252,
"step": 13
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.2979523241519928,
"learning_rate": 1.5e-05,
"loss": 1.1068,
"step": 14
},
{
"epoch": 0.35714285714285715,
"grad_norm": 0.3173041045665741,
"learning_rate": 1.6153846153846154e-05,
"loss": 1.1149,
"step": 15
},
{
"epoch": 0.38095238095238093,
"grad_norm": 0.32691124081611633,
"learning_rate": 1.7307692307692306e-05,
"loss": 1.146,
"step": 16
},
{
"epoch": 0.40476190476190477,
"grad_norm": 0.3304688334465027,
"learning_rate": 1.8461538461538465e-05,
"loss": 1.1602,
"step": 17
},
{
"epoch": 0.42857142857142855,
"grad_norm": 0.32470542192459106,
"learning_rate": 1.9615384615384617e-05,
"loss": 1.1232,
"step": 18
},
{
"epoch": 0.4523809523809524,
"grad_norm": 0.33648180961608887,
"learning_rate": 2.076923076923077e-05,
"loss": 1.1382,
"step": 19
},
{
"epoch": 0.47619047619047616,
"grad_norm": 0.3396536111831665,
"learning_rate": 2.1923076923076924e-05,
"loss": 1.2481,
"step": 20
},
{
"epoch": 0.5,
"grad_norm": 0.3541926145553589,
"learning_rate": 2.307692307692308e-05,
"loss": 1.111,
"step": 21
},
{
"epoch": 0.5238095238095238,
"grad_norm": 0.36257779598236084,
"learning_rate": 2.423076923076923e-05,
"loss": 1.138,
"step": 22
},
{
"epoch": 0.5476190476190477,
"grad_norm": 0.3602922260761261,
"learning_rate": 2.5384615384615386e-05,
"loss": 1.0467,
"step": 23
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.39060717821121216,
"learning_rate": 2.6538461538461538e-05,
"loss": 1.1,
"step": 24
},
{
"epoch": 0.5952380952380952,
"grad_norm": 0.3968435525894165,
"learning_rate": 2.7692307692307694e-05,
"loss": 1.1138,
"step": 25
},
{
"epoch": 0.6190476190476191,
"grad_norm": 0.41629916429519653,
"learning_rate": 2.884615384615385e-05,
"loss": 1.0513,
"step": 26
},
{
"epoch": 0.6428571428571429,
"grad_norm": 0.4167330861091614,
"learning_rate": 3e-05,
"loss": 1.0623,
"step": 27
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.42382487654685974,
"learning_rate": 2.999855077059572e-05,
"loss": 1.1167,
"step": 28
},
{
"epoch": 0.6904761904761905,
"grad_norm": 0.43757516145706177,
"learning_rate": 2.9994203362418313e-05,
"loss": 1.1078,
"step": 29
},
{
"epoch": 0.7142857142857143,
"grad_norm": 0.4598352909088135,
"learning_rate": 2.998695861552002e-05,
"loss": 1.0382,
"step": 30
},
{
"epoch": 0.7380952380952381,
"grad_norm": 0.4605550467967987,
"learning_rate": 2.9976817929807542e-05,
"loss": 0.9906,
"step": 31
},
{
"epoch": 0.7619047619047619,
"grad_norm": 0.5748752355575562,
"learning_rate": 2.996378326477153e-05,
"loss": 1.0196,
"step": 32
},
{
"epoch": 0.7857142857142857,
"grad_norm": 0.47698867321014404,
"learning_rate": 2.9947857139107964e-05,
"loss": 1.0654,
"step": 33
},
{
"epoch": 0.8095238095238095,
"grad_norm": 0.475506454706192,
"learning_rate": 2.992904263023146e-05,
"loss": 0.9911,
"step": 34
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.4737609326839447,
"learning_rate": 2.990734337368062e-05,
"loss": 1.041,
"step": 35
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.4721866548061371,
"learning_rate": 2.9882763562415518e-05,
"loss": 0.9602,
"step": 36
},
{
"epoch": 0.8809523809523809,
"grad_norm": 0.47256630659103394,
"learning_rate": 2.9855307946007532e-05,
"loss": 0.9134,
"step": 37
},
{
"epoch": 0.9047619047619048,
"grad_norm": 0.49380165338516235,
"learning_rate": 2.982498182972154e-05,
"loss": 0.9797,
"step": 38
},
{
"epoch": 0.9285714285714286,
"grad_norm": 0.4870937168598175,
"learning_rate": 2.9791791073490795e-05,
"loss": 0.9693,
"step": 39
},
{
"epoch": 0.9523809523809523,
"grad_norm": 0.4636681079864502,
"learning_rate": 2.9755742090784617e-05,
"loss": 0.8875,
"step": 40
},
{
"epoch": 0.9761904761904762,
"grad_norm": 0.4943704307079315,
"learning_rate": 2.9716841847369106e-05,
"loss": 0.9243,
"step": 41
},
{
"epoch": 1.0,
"grad_norm": 0.5488421320915222,
"learning_rate": 2.967509785996114e-05,
"loss": 0.9337,
"step": 42
},
{
"epoch": 1.0238095238095237,
"grad_norm": 0.5011991858482361,
"learning_rate": 2.963051819477592e-05,
"loss": 0.9167,
"step": 43
},
{
"epoch": 1.0476190476190477,
"grad_norm": 0.45742085576057434,
"learning_rate": 2.958311146596833e-05,
"loss": 0.866,
"step": 44
},
{
"epoch": 1.0714285714285714,
"grad_norm": 0.47257000207901,
"learning_rate": 2.953288683396841e-05,
"loss": 0.8597,
"step": 45
},
{
"epoch": 1.0952380952380953,
"grad_norm": 0.44165483117103577,
"learning_rate": 2.9479854003711298e-05,
"loss": 0.8339,
"step": 46
},
{
"epoch": 1.119047619047619,
"grad_norm": 0.4865115284919739,
"learning_rate": 2.9424023222761938e-05,
"loss": 0.9022,
"step": 47
},
{
"epoch": 1.1428571428571428,
"grad_norm": 0.4433363974094391,
"learning_rate": 2.9365405279334904e-05,
"loss": 0.857,
"step": 48
},
{
"epoch": 1.1666666666666667,
"grad_norm": 0.4483701288700104,
"learning_rate": 2.930401150020983e-05,
"loss": 0.8316,
"step": 49
},
{
"epoch": 1.1904761904761905,
"grad_norm": 0.45240312814712524,
"learning_rate": 2.9239853748542717e-05,
"loss": 0.7974,
"step": 50
},
{
"epoch": 1.2142857142857142,
"grad_norm": 0.48092514276504517,
"learning_rate": 2.9172944421573587e-05,
"loss": 0.7941,
"step": 51
},
{
"epoch": 1.2380952380952381,
"grad_norm": 0.4520972669124603,
"learning_rate": 2.9103296448230986e-05,
"loss": 0.8205,
"step": 52
},
{
"epoch": 1.2619047619047619,
"grad_norm": 0.43941670656204224,
"learning_rate": 2.9030923286633703e-05,
"loss": 0.809,
"step": 53
},
{
"epoch": 1.2857142857142856,
"grad_norm": 0.4921863079071045,
"learning_rate": 2.8955838921490252e-05,
"loss": 0.8192,
"step": 54
},
{
"epoch": 1.3095238095238095,
"grad_norm": 0.4463539719581604,
"learning_rate": 2.8878057861396606e-05,
"loss": 0.8358,
"step": 55
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.463927686214447,
"learning_rate": 2.8797595136032675e-05,
"loss": 0.8458,
"step": 56
},
{
"epoch": 1.3571428571428572,
"grad_norm": 0.44254210591316223,
"learning_rate": 2.8714466293258142e-05,
"loss": 0.7023,
"step": 57
},
{
"epoch": 1.380952380952381,
"grad_norm": 0.44470295310020447,
"learning_rate": 2.8628687396108107e-05,
"loss": 0.7205,
"step": 58
},
{
"epoch": 1.4047619047619047,
"grad_norm": 0.4261890947818756,
"learning_rate": 2.8540275019689237e-05,
"loss": 0.7707,
"step": 59
},
{
"epoch": 1.4285714285714286,
"grad_norm": 0.47844111919403076,
"learning_rate": 2.8449246247976947e-05,
"loss": 0.7681,
"step": 60
},
{
"epoch": 1.4523809523809523,
"grad_norm": 0.485981285572052,
"learning_rate": 2.835561867051426e-05,
"loss": 0.7262,
"step": 61
},
{
"epoch": 1.4761904761904763,
"grad_norm": 0.47706907987594604,
"learning_rate": 2.825941037901294e-05,
"loss": 0.7139,
"step": 62
},
{
"epoch": 1.5,
"grad_norm": 0.4486680030822754,
"learning_rate": 2.816063996385765e-05,
"loss": 0.7122,
"step": 63
},
{
"epoch": 1.5238095238095237,
"grad_norm": 0.4582575857639313,
"learning_rate": 2.805932651051372e-05,
"loss": 0.6618,
"step": 64
},
{
"epoch": 1.5476190476190477,
"grad_norm": 0.4188469648361206,
"learning_rate": 2.7955489595839228e-05,
"loss": 0.6886,
"step": 65
},
{
"epoch": 1.5714285714285714,
"grad_norm": 0.45904433727264404,
"learning_rate": 2.784914928430218e-05,
"loss": 0.6671,
"step": 66
},
{
"epoch": 1.5952380952380953,
"grad_norm": 0.455920934677124,
"learning_rate": 2.7740326124103416e-05,
"loss": 0.6745,
"step": 67
},
{
"epoch": 1.619047619047619,
"grad_norm": 0.44718676805496216,
"learning_rate": 2.762904114320609e-05,
"loss": 0.6736,
"step": 68
},
{
"epoch": 1.6428571428571428,
"grad_norm": 0.44715356826782227,
"learning_rate": 2.751531584527241e-05,
"loss": 0.6702,
"step": 69
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.46412724256515503,
"learning_rate": 2.7399172205508476e-05,
"loss": 0.6182,
"step": 70
},
{
"epoch": 1.6904761904761905,
"grad_norm": 0.4637816846370697,
"learning_rate": 2.7280632666418013e-05,
"loss": 0.6125,
"step": 71
},
{
"epoch": 1.7142857142857144,
"grad_norm": 0.43852296471595764,
"learning_rate": 2.715972013346576e-05,
"loss": 0.6306,
"step": 72
},
{
"epoch": 1.7380952380952381,
"grad_norm": 0.4781780540943146,
"learning_rate": 2.703645797065147e-05,
"loss": 0.6379,
"step": 73
},
{
"epoch": 1.7619047619047619,
"grad_norm": 0.45325422286987305,
"learning_rate": 2.6910869995995247e-05,
"loss": 0.6164,
"step": 74
},
{
"epoch": 1.7857142857142856,
"grad_norm": 0.4447559416294098,
"learning_rate": 2.678298047693518e-05,
"loss": 0.6486,
"step": 75
},
{
"epoch": 1.8095238095238095,
"grad_norm": 0.4800730347633362,
"learning_rate": 2.6652814125638142e-05,
"loss": 0.5419,
"step": 76
},
{
"epoch": 1.8333333333333335,
"grad_norm": 0.4950225353240967,
"learning_rate": 2.652039609422463e-05,
"loss": 0.6572,
"step": 77
},
{
"epoch": 1.8571428571428572,
"grad_norm": 0.4797077476978302,
"learning_rate": 2.638575196990862e-05,
"loss": 0.5417,
"step": 78
},
{
"epoch": 1.880952380952381,
"grad_norm": 0.5117254257202148,
"learning_rate": 2.624890777005332e-05,
"loss": 0.6147,
"step": 79
},
{
"epoch": 1.9047619047619047,
"grad_norm": 0.49460187554359436,
"learning_rate": 2.6109889937143828e-05,
"loss": 0.5439,
"step": 80
},
{
"epoch": 1.9285714285714286,
"grad_norm": 0.455595999956131,
"learning_rate": 2.5968725333677628e-05,
"loss": 0.5177,
"step": 81
},
{
"epoch": 1.9523809523809523,
"grad_norm": 0.4988625645637512,
"learning_rate": 2.582544123697395e-05,
"loss": 0.4811,
"step": 82
},
{
"epoch": 1.9761904761904763,
"grad_norm": 0.55754154920578,
"learning_rate": 2.568006533390295e-05,
"loss": 0.5444,
"step": 83
},
{
"epoch": 2.0,
"grad_norm": 0.5739458799362183,
"learning_rate": 2.5532625715535733e-05,
"loss": 0.5936,
"step": 84
},
{
"epoch": 2.0238095238095237,
"grad_norm": 0.47513002157211304,
"learning_rate": 2.5383150871716342e-05,
"loss": 0.5242,
"step": 85
},
{
"epoch": 2.0476190476190474,
"grad_norm": 0.5243293642997742,
"learning_rate": 2.5231669685556636e-05,
"loss": 0.5182,
"step": 86
},
{
"epoch": 2.0714285714285716,
"grad_norm": 0.4926926791667938,
"learning_rate": 2.507821142785516e-05,
"loss": 0.4948,
"step": 87
},
{
"epoch": 2.0952380952380953,
"grad_norm": 0.49989083409309387,
"learning_rate": 2.4922805751441174e-05,
"loss": 0.4448,
"step": 88
},
{
"epoch": 2.119047619047619,
"grad_norm": 0.49200865626335144,
"learning_rate": 2.4765482685444786e-05,
"loss": 0.486,
"step": 89
},
{
"epoch": 2.142857142857143,
"grad_norm": 0.5504772067070007,
"learning_rate": 2.460627262949443e-05,
"loss": 0.4386,
"step": 90
},
{
"epoch": 2.1666666666666665,
"grad_norm": 0.5244171023368835,
"learning_rate": 2.4445206347842714e-05,
"loss": 0.4779,
"step": 91
},
{
"epoch": 2.1904761904761907,
"grad_norm": 0.573525071144104,
"learning_rate": 2.428231496342181e-05,
"loss": 0.5095,
"step": 92
},
{
"epoch": 2.2142857142857144,
"grad_norm": 0.5168107151985168,
"learning_rate": 2.4117629951829602e-05,
"loss": 0.4114,
"step": 93
},
{
"epoch": 2.238095238095238,
"grad_norm": 0.5406803488731384,
"learning_rate": 2.395118313524758e-05,
"loss": 0.4327,
"step": 94
},
{
"epoch": 2.261904761904762,
"grad_norm": 0.5871009230613708,
"learning_rate": 2.3783006676291866e-05,
"loss": 0.4387,
"step": 95
},
{
"epoch": 2.2857142857142856,
"grad_norm": 0.50551837682724,
"learning_rate": 2.361313307179837e-05,
"loss": 0.4716,
"step": 96
},
{
"epoch": 2.3095238095238093,
"grad_norm": 0.5389571189880371,
"learning_rate": 2.3441595146543458e-05,
"loss": 0.4019,
"step": 97
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.5192880034446716,
"learning_rate": 2.3268426046901153e-05,
"loss": 0.3883,
"step": 98
},
{
"epoch": 2.357142857142857,
"grad_norm": 0.5577870011329651,
"learning_rate": 2.3093659234438266e-05,
"loss": 0.4168,
"step": 99
},
{
"epoch": 2.380952380952381,
"grad_norm": 0.5412160158157349,
"learning_rate": 2.291732847944861e-05,
"loss": 0.3231,
"step": 100
},
{
"epoch": 2.4047619047619047,
"grad_norm": 0.559502124786377,
"learning_rate": 2.2739467854427512e-05,
"loss": 0.4011,
"step": 101
},
{
"epoch": 2.4285714285714284,
"grad_norm": 0.5564844012260437,
"learning_rate": 2.2560111727488e-05,
"loss": 0.3355,
"step": 102
},
{
"epoch": 2.4523809523809526,
"grad_norm": 0.657665491104126,
"learning_rate": 2.237929475571979e-05,
"loss": 0.3178,
"step": 103
},
{
"epoch": 2.4761904761904763,
"grad_norm": 0.5791051387786865,
"learning_rate": 2.219705187849254e-05,
"loss": 0.3433,
"step": 104
},
{
"epoch": 2.5,
"grad_norm": 0.5594415664672852,
"learning_rate": 2.2013418310704422e-05,
"loss": 0.3279,
"step": 105
},
{
"epoch": 2.5238095238095237,
"grad_norm": 0.5653091073036194,
"learning_rate": 2.1828429535977585e-05,
"loss": 0.3687,
"step": 106
},
{
"epoch": 2.5476190476190474,
"grad_norm": 0.6664767265319824,
"learning_rate": 2.1642121299801594e-05,
"loss": 0.3678,
"step": 107
},
{
"epoch": 2.571428571428571,
"grad_norm": 0.4978804588317871,
"learning_rate": 2.1454529602626336e-05,
"loss": 0.3429,
"step": 108
},
{
"epoch": 2.5952380952380953,
"grad_norm": 0.6083326935768127,
"learning_rate": 2.126569069290562e-05,
"loss": 0.3275,
"step": 109
},
{
"epoch": 2.619047619047619,
"grad_norm": 0.4487161934375763,
"learning_rate": 2.107564106009286e-05,
"loss": 0.3362,
"step": 110
},
{
"epoch": 2.642857142857143,
"grad_norm": 0.5292232036590576,
"learning_rate": 2.0884417427590217e-05,
"loss": 0.3151,
"step": 111
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.5099493265151978,
"learning_rate": 2.0692056745652483e-05,
"loss": 0.2466,
"step": 112
},
{
"epoch": 2.6904761904761907,
"grad_norm": 0.5174189805984497,
"learning_rate": 2.0498596184247196e-05,
"loss": 0.2709,
"step": 113
},
{
"epoch": 2.7142857142857144,
"grad_norm": 0.4637812376022339,
"learning_rate": 2.030407312587224e-05,
"loss": 0.2764,
"step": 114
},
{
"epoch": 2.738095238095238,
"grad_norm": 0.45562928915023804,
"learning_rate": 2.010852515833242e-05,
"loss": 0.2742,
"step": 115
},
{
"epoch": 2.761904761904762,
"grad_norm": 0.49018341302871704,
"learning_rate": 1.9911990067476336e-05,
"loss": 0.3294,
"step": 116
},
{
"epoch": 2.7857142857142856,
"grad_norm": 0.46769657731056213,
"learning_rate": 1.9714505829895004e-05,
"loss": 0.2476,
"step": 117
},
{
"epoch": 2.8095238095238093,
"grad_norm": 0.46229276061058044,
"learning_rate": 1.951611060558363e-05,
"loss": 0.2972,
"step": 118
},
{
"epoch": 2.8333333333333335,
"grad_norm": 0.4360576868057251,
"learning_rate": 1.9316842730567902e-05,
"loss": 0.2707,
"step": 119
},
{
"epoch": 2.857142857142857,
"grad_norm": 0.4538360834121704,
"learning_rate": 1.9116740709496334e-05,
"loss": 0.2914,
"step": 120
},
{
"epoch": 2.880952380952381,
"grad_norm": 0.4812716245651245,
"learning_rate": 1.8915843208199967e-05,
"loss": 0.2864,
"step": 121
},
{
"epoch": 2.9047619047619047,
"grad_norm": 0.40740448236465454,
"learning_rate": 1.8714189046220946e-05,
"loss": 0.2494,
"step": 122
},
{
"epoch": 2.928571428571429,
"grad_norm": 0.4739934206008911,
"learning_rate": 1.851181718931141e-05,
"loss": 0.2906,
"step": 123
},
{
"epoch": 2.9523809523809526,
"grad_norm": 0.3870614469051361,
"learning_rate": 1.830876674190411e-05,
"loss": 0.2701,
"step": 124
},
{
"epoch": 2.9761904761904763,
"grad_norm": 0.4243805706501007,
"learning_rate": 1.8105076939556238e-05,
"loss": 0.2473,
"step": 125
},
{
"epoch": 3.0,
"grad_norm": 0.728119969367981,
"learning_rate": 1.790078714136792e-05,
"loss": 0.233,
"step": 126
},
{
"epoch": 3.0238095238095237,
"grad_norm": 0.5070134401321411,
"learning_rate": 1.769593682237682e-05,
"loss": 0.2477,
"step": 127
},
{
"epoch": 3.0476190476190474,
"grad_norm": 0.39993464946746826,
"learning_rate": 1.7490565565930382e-05,
"loss": 0.2233,
"step": 128
},
{
"epoch": 3.0714285714285716,
"grad_norm": 0.3518825173377991,
"learning_rate": 1.7284713056037074e-05,
"loss": 0.2324,
"step": 129
},
{
"epoch": 3.0952380952380953,
"grad_norm": 0.36889803409576416,
"learning_rate": 1.7078419069698283e-05,
"loss": 0.254,
"step": 130
},
{
"epoch": 3.119047619047619,
"grad_norm": 0.4398171603679657,
"learning_rate": 1.687172346922213e-05,
"loss": 0.2549,
"step": 131
},
{
"epoch": 3.142857142857143,
"grad_norm": 0.4238666594028473,
"learning_rate": 1.6664666194520873e-05,
"loss": 0.2421,
"step": 132
},
{
"epoch": 3.1666666666666665,
"grad_norm": 0.36933112144470215,
"learning_rate": 1.645728725539329e-05,
"loss": 0.2286,
"step": 133
},
{
"epoch": 3.1904761904761907,
"grad_norm": 0.3686385452747345,
"learning_rate": 1.6249626723793572e-05,
"loss": 0.2126,
"step": 134
},
{
"epoch": 3.2142857142857144,
"grad_norm": 0.34437063336372375,
"learning_rate": 1.6041724726088187e-05,
"loss": 0.2521,
"step": 135
},
{
"epoch": 3.238095238095238,
"grad_norm": 0.36857160925865173,
"learning_rate": 1.5833621435302247e-05,
"loss": 0.2482,
"step": 136
},
{
"epoch": 3.261904761904762,
"grad_norm": 0.41980957984924316,
"learning_rate": 1.5625357063356825e-05,
"loss": 0.2647,
"step": 137
},
{
"epoch": 3.2857142857142856,
"grad_norm": 0.3371034264564514,
"learning_rate": 1.541697185329881e-05,
"loss": 0.2137,
"step": 138
},
{
"epoch": 3.3095238095238093,
"grad_norm": 0.33477044105529785,
"learning_rate": 1.5208506071524727e-05,
"loss": 0.1929,
"step": 139
},
{
"epoch": 3.3333333333333335,
"grad_norm": 0.48055484890937805,
"learning_rate": 1.5e-05,
"loss": 0.2787,
"step": 140
},
{
"epoch": 3.357142857142857,
"grad_norm": 0.40071988105773926,
"learning_rate": 1.4791493928475275e-05,
"loss": 0.2275,
"step": 141
},
{
"epoch": 3.380952380952381,
"grad_norm": 0.3763311207294464,
"learning_rate": 1.4583028146701191e-05,
"loss": 0.2081,
"step": 142
},
{
"epoch": 3.4047619047619047,
"grad_norm": 0.317967027425766,
"learning_rate": 1.437464293664318e-05,
"loss": 0.2001,
"step": 143
},
{
"epoch": 3.4285714285714284,
"grad_norm": 0.35785412788391113,
"learning_rate": 1.4166378564697757e-05,
"loss": 0.1762,
"step": 144
},
{
"epoch": 3.4523809523809526,
"grad_norm": 0.34185105562210083,
"learning_rate": 1.3958275273911812e-05,
"loss": 0.1976,
"step": 145
},
{
"epoch": 3.4761904761904763,
"grad_norm": 0.367768257856369,
"learning_rate": 1.375037327620643e-05,
"loss": 0.2354,
"step": 146
},
{
"epoch": 3.5,
"grad_norm": 0.3221169710159302,
"learning_rate": 1.3542712744606712e-05,
"loss": 0.2075,
"step": 147
},
{
"epoch": 3.5238095238095237,
"grad_norm": 0.332815945148468,
"learning_rate": 1.3335333805479126e-05,
"loss": 0.1951,
"step": 148
},
{
"epoch": 3.5476190476190474,
"grad_norm": 0.3435097932815552,
"learning_rate": 1.3128276530777874e-05,
"loss": 0.2328,
"step": 149
},
{
"epoch": 3.571428571428571,
"grad_norm": 0.35089072585105896,
"learning_rate": 1.292158093030172e-05,
"loss": 0.2263,
"step": 150
},
{
"epoch": 3.5952380952380953,
"grad_norm": 0.2681962847709656,
"learning_rate": 1.2715286943962925e-05,
"loss": 0.1824,
"step": 151
},
{
"epoch": 3.619047619047619,
"grad_norm": 0.3577785789966583,
"learning_rate": 1.2509434434069625e-05,
"loss": 0.2258,
"step": 152
},
{
"epoch": 3.642857142857143,
"grad_norm": 0.4009724259376526,
"learning_rate": 1.2304063177623182e-05,
"loss": 0.2548,
"step": 153
},
{
"epoch": 3.6666666666666665,
"grad_norm": 0.35427093505859375,
"learning_rate": 1.2099212858632083e-05,
"loss": 0.1965,
"step": 154
},
{
"epoch": 3.6904761904761907,
"grad_norm": 0.3180360496044159,
"learning_rate": 1.1894923060443763e-05,
"loss": 0.201,
"step": 155
},
{
"epoch": 3.7142857142857144,
"grad_norm": 0.35209810733795166,
"learning_rate": 1.169123325809589e-05,
"loss": 0.1746,
"step": 156
},
{
"epoch": 3.738095238095238,
"grad_norm": 0.2801775336265564,
"learning_rate": 1.1488182810688593e-05,
"loss": 0.1782,
"step": 157
},
{
"epoch": 3.761904761904762,
"grad_norm": 0.2826492190361023,
"learning_rate": 1.1285810953779057e-05,
"loss": 0.1849,
"step": 158
},
{
"epoch": 3.7857142857142856,
"grad_norm": 0.3703887164592743,
"learning_rate": 1.1084156791800036e-05,
"loss": 0.2118,
"step": 159
},
{
"epoch": 3.8095238095238093,
"grad_norm": 0.3333575129508972,
"learning_rate": 1.0883259290503665e-05,
"loss": 0.2544,
"step": 160
},
{
"epoch": 3.8333333333333335,
"grad_norm": 0.3313210904598236,
"learning_rate": 1.0683157269432097e-05,
"loss": 0.2226,
"step": 161
},
{
"epoch": 3.857142857142857,
"grad_norm": 0.3558778464794159,
"learning_rate": 1.0483889394416373e-05,
"loss": 0.2326,
"step": 162
},
{
"epoch": 3.880952380952381,
"grad_norm": 0.29440999031066895,
"learning_rate": 1.0285494170104996e-05,
"loss": 0.1599,
"step": 163
},
{
"epoch": 3.9047619047619047,
"grad_norm": 0.29854339361190796,
"learning_rate": 1.0088009932523664e-05,
"loss": 0.2412,
"step": 164
},
{
"epoch": 3.928571428571429,
"grad_norm": 0.28367212414741516,
"learning_rate": 9.891474841667585e-06,
"loss": 0.1735,
"step": 165
},
{
"epoch": 3.9523809523809526,
"grad_norm": 0.32328376173973083,
"learning_rate": 9.695926874127765e-06,
"loss": 0.2121,
"step": 166
},
{
"epoch": 3.9761904761904763,
"grad_norm": 0.33302921056747437,
"learning_rate": 9.501403815752813e-06,
"loss": 0.2368,
"step": 167
},
{
"epoch": 4.0,
"grad_norm": 0.4951595664024353,
"learning_rate": 9.307943254347521e-06,
"loss": 0.1148,
"step": 168
},
{
"epoch": 4.023809523809524,
"grad_norm": 0.2874068021774292,
"learning_rate": 9.115582572409789e-06,
"loss": 0.1935,
"step": 169
},
{
"epoch": 4.0476190476190474,
"grad_norm": 0.3146367073059082,
"learning_rate": 8.92435893990714e-06,
"loss": 0.1743,
"step": 170
},
{
"epoch": 4.071428571428571,
"grad_norm": 0.3270380198955536,
"learning_rate": 8.734309307094382e-06,
"loss": 0.2156,
"step": 171
},
{
"epoch": 4.095238095238095,
"grad_norm": 0.3079623281955719,
"learning_rate": 8.545470397373665e-06,
"loss": 0.2299,
"step": 172
},
{
"epoch": 4.119047619047619,
"grad_norm": 0.3418531119823456,
"learning_rate": 8.357878700198407e-06,
"loss": 0.2609,
"step": 173
},
{
"epoch": 4.142857142857143,
"grad_norm": 0.3398002088069916,
"learning_rate": 8.171570464022419e-06,
"loss": 0.1882,
"step": 174
},
{
"epoch": 4.166666666666667,
"grad_norm": 0.2681673765182495,
"learning_rate": 7.986581689295577e-06,
"loss": 0.2207,
"step": 175
},
{
"epoch": 4.190476190476191,
"grad_norm": 0.4535702168941498,
"learning_rate": 7.802948121507462e-06,
"loss": 0.196,
"step": 176
},
{
"epoch": 4.214285714285714,
"grad_norm": 0.30191582441329956,
"learning_rate": 7.620705244280208e-06,
"loss": 0.1812,
"step": 177
},
{
"epoch": 4.238095238095238,
"grad_norm": 0.32031571865081787,
"learning_rate": 7.439888272512004e-06,
"loss": 0.174,
"step": 178
},
{
"epoch": 4.261904761904762,
"grad_norm": 0.3073733448982239,
"learning_rate": 7.260532145572487e-06,
"loss": 0.2141,
"step": 179
},
{
"epoch": 4.285714285714286,
"grad_norm": 0.29829952120780945,
"learning_rate": 7.082671520551391e-06,
"loss": 0.232,
"step": 180
},
{
"epoch": 4.309523809523809,
"grad_norm": 0.3112286627292633,
"learning_rate": 6.906340765561734e-06,
"loss": 0.1621,
"step": 181
},
{
"epoch": 4.333333333333333,
"grad_norm": 0.3731706142425537,
"learning_rate": 6.731573953098851e-06,
"loss": 0.2016,
"step": 182
},
{
"epoch": 4.357142857142857,
"grad_norm": 0.27395114302635193,
"learning_rate": 6.558404853456545e-06,
"loss": 0.1835,
"step": 183
},
{
"epoch": 4.380952380952381,
"grad_norm": 0.35525375604629517,
"learning_rate": 6.38686692820163e-06,
"loss": 0.2645,
"step": 184
},
{
"epoch": 4.404761904761905,
"grad_norm": 0.3270358443260193,
"learning_rate": 6.2169933237081386e-06,
"loss": 0.2678,
"step": 185
},
{
"epoch": 4.428571428571429,
"grad_norm": 0.2867538034915924,
"learning_rate": 6.048816864752422e-06,
"loss": 0.1942,
"step": 186
},
{
"epoch": 4.4523809523809526,
"grad_norm": 0.24296227097511292,
"learning_rate": 5.882370048170403e-06,
"loss": 0.1636,
"step": 187
},
{
"epoch": 4.476190476190476,
"grad_norm": 0.30451998114585876,
"learning_rate": 5.71768503657819e-06,
"loss": 0.1886,
"step": 188
},
{
"epoch": 4.5,
"grad_norm": 0.3037683665752411,
"learning_rate": 5.55479365215729e-06,
"loss": 0.209,
"step": 189
},
{
"epoch": 4.523809523809524,
"grad_norm": 0.3124084770679474,
"learning_rate": 5.393727370505569e-06,
"loss": 0.2097,
"step": 190
},
{
"epoch": 4.5476190476190474,
"grad_norm": 0.249891996383667,
"learning_rate": 5.234517314555213e-06,
"loss": 0.1664,
"step": 191
},
{
"epoch": 4.571428571428571,
"grad_norm": 0.3419524133205414,
"learning_rate": 5.077194248558827e-06,
"loss": 0.2217,
"step": 192
},
{
"epoch": 4.595238095238095,
"grad_norm": 0.29300788044929504,
"learning_rate": 4.921788572144841e-06,
"loss": 0.1702,
"step": 193
},
{
"epoch": 4.619047619047619,
"grad_norm": 0.29377758502960205,
"learning_rate": 4.768330314443367e-06,
"loss": 0.1883,
"step": 194
},
{
"epoch": 4.642857142857143,
"grad_norm": 0.31856659054756165,
"learning_rate": 4.616849128283658e-06,
"loss": 0.1415,
"step": 195
},
{
"epoch": 4.666666666666667,
"grad_norm": 0.27666354179382324,
"learning_rate": 4.4673742844642716e-06,
"loss": 0.2132,
"step": 196
},
{
"epoch": 4.690476190476191,
"grad_norm": 0.2768075168132782,
"learning_rate": 4.319934666097055e-06,
"loss": 0.2197,
"step": 197
},
{
"epoch": 4.714285714285714,
"grad_norm": 0.3824225664138794,
"learning_rate": 4.174558763026048e-06,
"loss": 0.1516,
"step": 198
},
{
"epoch": 4.738095238095238,
"grad_norm": 0.2814243733882904,
"learning_rate": 4.031274666322372e-06,
"loss": 0.1632,
"step": 199
},
{
"epoch": 4.761904761904762,
"grad_norm": 0.320901483297348,
"learning_rate": 3.8901100628561755e-06,
"loss": 0.163,
"step": 200
},
{
"epoch": 4.785714285714286,
"grad_norm": 0.30128878355026245,
"learning_rate": 3.7510922299466818e-06,
"loss": 0.186,
"step": 201
},
{
"epoch": 4.809523809523809,
"grad_norm": 0.3085160255432129,
"learning_rate": 3.6142480300913805e-06,
"loss": 0.1951,
"step": 202
},
{
"epoch": 4.833333333333333,
"grad_norm": 0.2675895392894745,
"learning_rate": 3.4796039057753703e-06,
"loss": 0.1754,
"step": 203
},
{
"epoch": 4.857142857142857,
"grad_norm": 0.315141886472702,
"learning_rate": 3.3471858743618615e-06,
"loss": 0.2127,
"step": 204
},
{
"epoch": 4.880952380952381,
"grad_norm": 0.30165189504623413,
"learning_rate": 3.217019523064825e-06,
"loss": 0.1749,
"step": 205
},
{
"epoch": 4.904761904761905,
"grad_norm": 0.2545117735862732,
"learning_rate": 3.089130004004754e-06,
"loss": 0.1414,
"step": 206
},
{
"epoch": 4.928571428571429,
"grad_norm": 0.26600050926208496,
"learning_rate": 2.96354202934853e-06,
"loss": 0.1706,
"step": 207
},
{
"epoch": 4.9523809523809526,
"grad_norm": 0.32053714990615845,
"learning_rate": 2.8402798665342412e-06,
"loss": 0.175,
"step": 208
},
{
"epoch": 4.976190476190476,
"grad_norm": 0.33824998140335083,
"learning_rate": 2.7193673335819893e-06,
"loss": 0.2015,
"step": 209
},
{
"epoch": 5.0,
"grad_norm": 0.5689420104026794,
"learning_rate": 2.600827794491524e-06,
"loss": 0.1378,
"step": 210
},
{
"epoch": 5.023809523809524,
"grad_norm": 0.2932104170322418,
"learning_rate": 2.4846841547275916e-06,
"loss": 0.2041,
"step": 211
},
{
"epoch": 5.0476190476190474,
"grad_norm": 0.30341067910194397,
"learning_rate": 2.3709588567939118e-06,
"loss": 0.2191,
"step": 212
},
{
"epoch": 5.071428571428571,
"grad_norm": 0.27707991003990173,
"learning_rate": 2.259673875896585e-06,
"loss": 0.1355,
"step": 213
},
{
"epoch": 5.095238095238095,
"grad_norm": 0.2761556804180145,
"learning_rate": 2.150850715697823e-06,
"loss": 0.2066,
"step": 214
},
{
"epoch": 5.119047619047619,
"grad_norm": 0.30768999457359314,
"learning_rate": 2.044510404160774e-06,
"loss": 0.209,
"step": 215
},
{
"epoch": 5.142857142857143,
"grad_norm": 0.3525684177875519,
"learning_rate": 1.9406734894862848e-06,
"loss": 0.2322,
"step": 216
},
{
"epoch": 5.166666666666667,
"grad_norm": 0.26923832297325134,
"learning_rate": 1.8393600361423534e-06,
"loss": 0.1825,
"step": 217
},
{
"epoch": 5.190476190476191,
"grad_norm": 0.2886522710323334,
"learning_rate": 1.7405896209870665e-06,
"loss": 0.2271,
"step": 218
},
{
"epoch": 5.214285714285714,
"grad_norm": 0.27028337121009827,
"learning_rate": 1.6443813294857452e-06,
"loss": 0.1579,
"step": 219
},
{
"epoch": 5.238095238095238,
"grad_norm": 0.22688105702400208,
"learning_rate": 1.550753752023053e-06,
"loss": 0.1498,
"step": 220
},
{
"epoch": 5.261904761904762,
"grad_norm": 0.4237247705459595,
"learning_rate": 1.459724980310767e-06,
"loss": 0.216,
"step": 221
},
{
"epoch": 5.285714285714286,
"grad_norm": 0.3102908134460449,
"learning_rate": 1.3713126038918977e-06,
"loss": 0.1534,
"step": 222
},
{
"epoch": 5.309523809523809,
"grad_norm": 0.2860654294490814,
"learning_rate": 1.2855337067418576e-06,
"loss": 0.1518,
"step": 223
},
{
"epoch": 5.333333333333333,
"grad_norm": 0.30412977933883667,
"learning_rate": 1.2024048639673225e-06,
"loss": 0.2061,
"step": 224
},
{
"epoch": 5.357142857142857,
"grad_norm": 0.261602520942688,
"learning_rate": 1.1219421386033958e-06,
"loss": 0.183,
"step": 225
},
{
"epoch": 5.380952380952381,
"grad_norm": 0.31666767597198486,
"learning_rate": 1.0441610785097471e-06,
"loss": 0.2057,
"step": 226
},
{
"epoch": 5.404761904761905,
"grad_norm": 0.26625874638557434,
"learning_rate": 9.690767133662976e-07,
"loss": 0.1723,
"step": 227
},
{
"epoch": 5.428571428571429,
"grad_norm": 0.36171892285346985,
"learning_rate": 8.967035517690147e-07,
"loss": 0.2047,
"step": 228
},
{
"epoch": 5.4523809523809526,
"grad_norm": 0.28423306345939636,
"learning_rate": 8.270555784264167e-07,
"loss": 0.2082,
"step": 229
},
{
"epoch": 5.476190476190476,
"grad_norm": 0.3602757751941681,
"learning_rate": 7.601462514572877e-07,
"loss": 0.1992,
"step": 230
},
{
"epoch": 5.5,
"grad_norm": 0.32955092191696167,
"learning_rate": 6.959884997901705e-07,
"loss": 0.2127,
"step": 231
},
{
"epoch": 5.523809523809524,
"grad_norm": 0.2506847679615021,
"learning_rate": 6.345947206650981e-07,
"loss": 0.1743,
"step": 232
},
{
"epoch": 5.5476190476190474,
"grad_norm": 0.28736478090286255,
"learning_rate": 5.759767772380648e-07,
"loss": 0.1955,
"step": 233
},
{
"epoch": 5.571428571428571,
"grad_norm": 0.25602585077285767,
"learning_rate": 5.201459962886995e-07,
"loss": 0.1329,
"step": 234
},
{
"epoch": 5.595238095238095,
"grad_norm": 0.28149425983428955,
"learning_rate": 4.6711316603159084e-07,
"loss": 0.1319,
"step": 235
},
{
"epoch": 5.619047619047619,
"grad_norm": 0.2809825539588928,
"learning_rate": 4.1688853403167194e-07,
"loss": 0.1838,
"step": 236
},
{
"epoch": 5.642857142857143,
"grad_norm": 0.3335093855857849,
"learning_rate": 3.6948180522408006e-07,
"loss": 0.2569,
"step": 237
},
{
"epoch": 5.666666666666667,
"grad_norm": 0.2959223687648773,
"learning_rate": 3.2490214003885963e-07,
"loss": 0.1872,
"step": 238
},
{
"epoch": 5.690476190476191,
"grad_norm": 0.32720711827278137,
"learning_rate": 2.831581526308935e-07,
"loss": 0.2051,
"step": 239
},
{
"epoch": 5.714285714285714,
"grad_norm": 0.30836525559425354,
"learning_rate": 2.4425790921538404e-07,
"loss": 0.137,
"step": 240
},
{
"epoch": 5.738095238095238,
"grad_norm": 0.22278806567192078,
"learning_rate": 2.0820892650920686e-07,
"loss": 0.2019,
"step": 241
},
{
"epoch": 5.761904761904762,
"grad_norm": 0.2525181472301483,
"learning_rate": 1.7501817027846255e-07,
"loss": 0.1822,
"step": 242
},
{
"epoch": 5.785714285714286,
"grad_norm": 0.27369964122772217,
"learning_rate": 1.4469205399246844e-07,
"loss": 0.1069,
"step": 243
},
{
"epoch": 5.809523809523809,
"grad_norm": 0.3073241710662842,
"learning_rate": 1.1723643758448144e-07,
"loss": 0.2186,
"step": 244
},
{
"epoch": 5.833333333333333,
"grad_norm": 0.2990500330924988,
"learning_rate": 9.265662631938399e-08,
"loss": 0.2293,
"step": 245
},
{
"epoch": 5.857142857142857,
"grad_norm": 0.26736584305763245,
"learning_rate": 7.095736976853895e-08,
"loss": 0.1718,
"step": 246
},
{
"epoch": 5.880952380952381,
"grad_norm": 0.28982308506965637,
"learning_rate": 5.214286089203546e-08,
"loss": 0.1902,
"step": 247
},
{
"epoch": 5.904761904761905,
"grad_norm": 0.3653714656829834,
"learning_rate": 3.621673522847035e-08,
"loss": 0.2339,
"step": 248
},
{
"epoch": 5.928571428571429,
"grad_norm": 0.3287780284881592,
"learning_rate": 2.3182070192460104e-08,
"loss": 0.1513,
"step": 249
},
{
"epoch": 5.9523809523809526,
"grad_norm": 0.2435285449028015,
"learning_rate": 1.3041384479981489e-08,
"loss": 0.151,
"step": 250
},
{
"epoch": 5.976190476190476,
"grad_norm": 0.3266546428203583,
"learning_rate": 5.796637581689246e-09,
"loss": 0.2409,
"step": 251
},
{
"epoch": 6.0,
"grad_norm": 0.5456566214561462,
"learning_rate": 1.4492294042839361e-09,
"loss": 0.1587,
"step": 252
},
{
"epoch": 6.0,
"step": 252,
"total_flos": 9.225268114559155e+16,
"train_loss": 0.45595223279226393,
"train_runtime": 1969.0507,
"train_samples_per_second": 2.014,
"train_steps_per_second": 0.128
}
],
"logging_steps": 1,
"max_steps": 252,
"num_input_tokens_seen": 0,
"num_train_epochs": 6,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 9.225268114559155e+16,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}