ko120's picture
Initial upload
c75757c verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 7.0,
"eval_steps": 500,
"global_step": 140,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.05128205128205128,
"grad_norm": 3.0017077922821045,
"learning_rate": 0.0,
"loss": 0.5219,
"step": 1
},
{
"epoch": 0.10256410256410256,
"grad_norm": 3.0514369010925293,
"learning_rate": 7.142857142857143e-07,
"loss": 0.5409,
"step": 2
},
{
"epoch": 0.15384615384615385,
"grad_norm": 3.4082279205322266,
"learning_rate": 1.4285714285714286e-06,
"loss": 0.5321,
"step": 3
},
{
"epoch": 0.20512820512820512,
"grad_norm": 2.9593334197998047,
"learning_rate": 2.1428571428571427e-06,
"loss": 0.5692,
"step": 4
},
{
"epoch": 0.2564102564102564,
"grad_norm": 3.162719964981079,
"learning_rate": 2.8571428571428573e-06,
"loss": 0.5689,
"step": 5
},
{
"epoch": 0.3076923076923077,
"grad_norm": 2.942983388900757,
"learning_rate": 3.5714285714285718e-06,
"loss": 0.5833,
"step": 6
},
{
"epoch": 0.358974358974359,
"grad_norm": 2.6459100246429443,
"learning_rate": 4.2857142857142855e-06,
"loss": 0.5508,
"step": 7
},
{
"epoch": 0.41025641025641024,
"grad_norm": 2.5366854667663574,
"learning_rate": 5e-06,
"loss": 0.5117,
"step": 8
},
{
"epoch": 0.46153846153846156,
"grad_norm": 2.084428071975708,
"learning_rate": 5.7142857142857145e-06,
"loss": 0.5472,
"step": 9
},
{
"epoch": 0.5128205128205128,
"grad_norm": 1.9129199981689453,
"learning_rate": 6.4285714285714295e-06,
"loss": 0.5,
"step": 10
},
{
"epoch": 0.5641025641025641,
"grad_norm": 0.9752050042152405,
"learning_rate": 7.1428571428571436e-06,
"loss": 0.5004,
"step": 11
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.8704648613929749,
"learning_rate": 7.857142857142858e-06,
"loss": 0.5277,
"step": 12
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.8690164685249329,
"learning_rate": 8.571428571428571e-06,
"loss": 0.5162,
"step": 13
},
{
"epoch": 0.717948717948718,
"grad_norm": 1.0232652425765991,
"learning_rate": 9.285714285714288e-06,
"loss": 0.4785,
"step": 14
},
{
"epoch": 0.7692307692307693,
"grad_norm": 1.0574804544448853,
"learning_rate": 1e-05,
"loss": 0.4788,
"step": 15
},
{
"epoch": 0.8205128205128205,
"grad_norm": 1.4246594905853271,
"learning_rate": 9.998445910004082e-06,
"loss": 0.4292,
"step": 16
},
{
"epoch": 0.8717948717948718,
"grad_norm": 1.6748791933059692,
"learning_rate": 9.993784606094612e-06,
"loss": 0.4618,
"step": 17
},
{
"epoch": 0.9230769230769231,
"grad_norm": 1.3139477968215942,
"learning_rate": 9.986018985905901e-06,
"loss": 0.4347,
"step": 18
},
{
"epoch": 0.9743589743589743,
"grad_norm": 1.5158166885375977,
"learning_rate": 9.975153876827008e-06,
"loss": 0.4892,
"step": 19
},
{
"epoch": 1.0,
"grad_norm": 1.5158166885375977,
"learning_rate": 9.961196033000862e-06,
"loss": 0.4701,
"step": 20
},
{
"epoch": 1.0512820512820513,
"grad_norm": 0.5997151732444763,
"learning_rate": 9.944154131125643e-06,
"loss": 0.4485,
"step": 21
},
{
"epoch": 1.1025641025641026,
"grad_norm": 0.4533652067184448,
"learning_rate": 9.924038765061042e-06,
"loss": 0.4405,
"step": 22
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.27620309591293335,
"learning_rate": 9.900862439242719e-06,
"loss": 0.4603,
"step": 23
},
{
"epoch": 1.205128205128205,
"grad_norm": 0.3358033299446106,
"learning_rate": 9.874639560909118e-06,
"loss": 0.4521,
"step": 24
},
{
"epoch": 1.2564102564102564,
"grad_norm": 0.39922070503234863,
"learning_rate": 9.84538643114539e-06,
"loss": 0.4317,
"step": 25
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.4135103225708008,
"learning_rate": 9.81312123475006e-06,
"loss": 0.4635,
"step": 26
},
{
"epoch": 1.358974358974359,
"grad_norm": 0.34654220938682556,
"learning_rate": 9.777864028930705e-06,
"loss": 0.4313,
"step": 27
},
{
"epoch": 1.4102564102564101,
"grad_norm": 0.3868102431297302,
"learning_rate": 9.73963673083566e-06,
"loss": 0.4259,
"step": 28
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.6085906028747559,
"learning_rate": 9.698463103929542e-06,
"loss": 0.45,
"step": 29
},
{
"epoch": 1.5128205128205128,
"grad_norm": 0.28642383217811584,
"learning_rate": 9.654368743221022e-06,
"loss": 0.4133,
"step": 30
},
{
"epoch": 1.564102564102564,
"grad_norm": 0.4036538898944855,
"learning_rate": 9.60738105935204e-06,
"loss": 0.4445,
"step": 31
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.31555476784706116,
"learning_rate": 9.557529261558367e-06,
"loss": 0.4104,
"step": 32
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.26336777210235596,
"learning_rate": 9.504844339512096e-06,
"loss": 0.452,
"step": 33
},
{
"epoch": 1.717948717948718,
"grad_norm": 0.26181158423423767,
"learning_rate": 9.449359044057344e-06,
"loss": 0.4007,
"step": 34
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.23928822576999664,
"learning_rate": 9.391107866851143e-06,
"loss": 0.3953,
"step": 35
},
{
"epoch": 1.8205128205128205,
"grad_norm": 0.3171135187149048,
"learning_rate": 9.330127018922195e-06,
"loss": 0.4395,
"step": 36
},
{
"epoch": 1.8717948717948718,
"grad_norm": 0.2389429360628128,
"learning_rate": 9.266454408160779e-06,
"loss": 0.399,
"step": 37
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.24025550484657288,
"learning_rate": 9.200129615753858e-06,
"loss": 0.4082,
"step": 38
},
{
"epoch": 1.9743589743589745,
"grad_norm": 0.262188196182251,
"learning_rate": 9.131193871579975e-06,
"loss": 0.4198,
"step": 39
},
{
"epoch": 2.0,
"grad_norm": 0.44420406222343445,
"learning_rate": 9.059690028579285e-06,
"loss": 0.4012,
"step": 40
},
{
"epoch": 2.051282051282051,
"grad_norm": 0.34896019101142883,
"learning_rate": 8.985662536114614e-06,
"loss": 0.3907,
"step": 41
},
{
"epoch": 2.1025641025641026,
"grad_norm": 0.3515949249267578,
"learning_rate": 8.90915741234015e-06,
"loss": 0.4016,
"step": 42
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.3820473551750183,
"learning_rate": 8.83022221559489e-06,
"loss": 0.4069,
"step": 43
},
{
"epoch": 2.2051282051282053,
"grad_norm": 0.31491103768348694,
"learning_rate": 8.748906014838672e-06,
"loss": 0.4069,
"step": 44
},
{
"epoch": 2.2564102564102564,
"grad_norm": 0.3246411085128784,
"learning_rate": 8.665259359149132e-06,
"loss": 0.3576,
"step": 45
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.2691456973552704,
"learning_rate": 8.579334246298593e-06,
"loss": 0.3707,
"step": 46
},
{
"epoch": 2.358974358974359,
"grad_norm": 0.29816141724586487,
"learning_rate": 8.491184090430365e-06,
"loss": 0.347,
"step": 47
},
{
"epoch": 2.41025641025641,
"grad_norm": 0.34237486124038696,
"learning_rate": 8.400863688854598e-06,
"loss": 0.3955,
"step": 48
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.3234226703643799,
"learning_rate": 8.308429187984298e-06,
"loss": 0.4257,
"step": 49
},
{
"epoch": 2.5128205128205128,
"grad_norm": 0.33046239614486694,
"learning_rate": 8.213938048432697e-06,
"loss": 0.3709,
"step": 50
},
{
"epoch": 2.564102564102564,
"grad_norm": 0.31583672761917114,
"learning_rate": 8.117449009293668e-06,
"loss": 0.4144,
"step": 51
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.29995957016944885,
"learning_rate": 8.019022051627387e-06,
"loss": 0.3747,
"step": 52
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.32305797934532166,
"learning_rate": 7.918718361173951e-06,
"loss": 0.3781,
"step": 53
},
{
"epoch": 2.717948717948718,
"grad_norm": 0.3537105917930603,
"learning_rate": 7.81660029031811e-06,
"loss": 0.3862,
"step": 54
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.2909030616283417,
"learning_rate": 7.712731319328798e-06,
"loss": 0.3903,
"step": 55
},
{
"epoch": 2.8205128205128203,
"grad_norm": 0.32513272762298584,
"learning_rate": 7.607176016897491e-06,
"loss": 0.3992,
"step": 56
},
{
"epoch": 2.871794871794872,
"grad_norm": 0.3026527166366577,
"learning_rate": 7.500000000000001e-06,
"loss": 0.3822,
"step": 57
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.3264043629169464,
"learning_rate": 7.391269893106592e-06,
"loss": 0.3985,
"step": 58
},
{
"epoch": 2.9743589743589745,
"grad_norm": 0.5407824516296387,
"learning_rate": 7.281053286765816e-06,
"loss": 0.3865,
"step": 59
},
{
"epoch": 3.0,
"grad_norm": 0.5407824516296387,
"learning_rate": 7.169418695587791e-06,
"loss": 0.3956,
"step": 60
},
{
"epoch": 3.051282051282051,
"grad_norm": 0.18096458911895752,
"learning_rate": 7.056435515653059e-06,
"loss": 0.348,
"step": 61
},
{
"epoch": 3.1025641025641026,
"grad_norm": 0.20220977067947388,
"learning_rate": 6.942173981373474e-06,
"loss": 0.319,
"step": 62
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.16257236897945404,
"learning_rate": 6.8267051218319766e-06,
"loss": 0.3357,
"step": 63
},
{
"epoch": 3.2051282051282053,
"grad_norm": 0.20337192714214325,
"learning_rate": 6.710100716628345e-06,
"loss": 0.3396,
"step": 64
},
{
"epoch": 3.2564102564102564,
"grad_norm": 0.15443633496761322,
"learning_rate": 6.592433251258423e-06,
"loss": 0.3366,
"step": 65
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.18854251503944397,
"learning_rate": 6.473775872054522e-06,
"loss": 0.3506,
"step": 66
},
{
"epoch": 3.358974358974359,
"grad_norm": 0.21565477550029755,
"learning_rate": 6.354202340715027e-06,
"loss": 0.3491,
"step": 67
},
{
"epoch": 3.41025641025641,
"grad_norm": 0.211973637342453,
"learning_rate": 6.233786988451468e-06,
"loss": 0.3385,
"step": 68
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.15138986706733704,
"learning_rate": 6.112604669781572e-06,
"loss": 0.3571,
"step": 69
},
{
"epoch": 3.5128205128205128,
"grad_norm": 0.2490689605474472,
"learning_rate": 5.990730715996989e-06,
"loss": 0.3509,
"step": 70
},
{
"epoch": 3.564102564102564,
"grad_norm": 0.20323821902275085,
"learning_rate": 5.8682408883346535e-06,
"loss": 0.3439,
"step": 71
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.21185290813446045,
"learning_rate": 5.745211330880872e-06,
"loss": 0.3858,
"step": 72
},
{
"epoch": 3.6666666666666665,
"grad_norm": 0.1838921755552292,
"learning_rate": 5.621718523237427e-06,
"loss": 0.3272,
"step": 73
},
{
"epoch": 3.717948717948718,
"grad_norm": 0.19704340398311615,
"learning_rate": 5.497839232979084e-06,
"loss": 0.3108,
"step": 74
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.18711262941360474,
"learning_rate": 5.373650467932122e-06,
"loss": 0.3632,
"step": 75
},
{
"epoch": 3.8205128205128203,
"grad_norm": 0.2144545167684555,
"learning_rate": 5.249229428303486e-06,
"loss": 0.3405,
"step": 76
},
{
"epoch": 3.871794871794872,
"grad_norm": 0.1933801919221878,
"learning_rate": 5.1246534586903655e-06,
"loss": 0.3469,
"step": 77
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.17201346158981323,
"learning_rate": 5e-06,
"loss": 0.3327,
"step": 78
},
{
"epoch": 3.9743589743589745,
"grad_norm": 0.19057320058345795,
"learning_rate": 4.875346541309637e-06,
"loss": 0.2937,
"step": 79
},
{
"epoch": 4.0,
"grad_norm": 0.38898980617523193,
"learning_rate": 4.750770571696514e-06,
"loss": 0.3462,
"step": 80
},
{
"epoch": 4.051282051282051,
"grad_norm": 0.28359612822532654,
"learning_rate": 4.626349532067879e-06,
"loss": 0.3026,
"step": 81
},
{
"epoch": 4.102564102564102,
"grad_norm": 0.31366801261901855,
"learning_rate": 4.502160767020918e-06,
"loss": 0.3221,
"step": 82
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.2640708386898041,
"learning_rate": 4.3782814767625755e-06,
"loss": 0.3084,
"step": 83
},
{
"epoch": 4.205128205128205,
"grad_norm": 0.28817296028137207,
"learning_rate": 4.254788669119127e-06,
"loss": 0.3569,
"step": 84
},
{
"epoch": 4.256410256410256,
"grad_norm": 0.29913732409477234,
"learning_rate": 4.131759111665349e-06,
"loss": 0.2827,
"step": 85
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.2851954400539398,
"learning_rate": 4.009269284003014e-06,
"loss": 0.3338,
"step": 86
},
{
"epoch": 4.358974358974359,
"grad_norm": 0.28056982159614563,
"learning_rate": 3.887395330218429e-06,
"loss": 0.3198,
"step": 87
},
{
"epoch": 4.410256410256411,
"grad_norm": 0.26367899775505066,
"learning_rate": 3.7662130115485317e-06,
"loss": 0.3084,
"step": 88
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.27056267857551575,
"learning_rate": 3.6457976592849753e-06,
"loss": 0.3086,
"step": 89
},
{
"epoch": 4.512820512820513,
"grad_norm": 0.28712373971939087,
"learning_rate": 3.526224127945479e-06,
"loss": 0.3375,
"step": 90
},
{
"epoch": 4.564102564102564,
"grad_norm": 0.28978201746940613,
"learning_rate": 3.4075667487415785e-06,
"loss": 0.3119,
"step": 91
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.2983953356742859,
"learning_rate": 3.289899283371657e-06,
"loss": 0.2993,
"step": 92
},
{
"epoch": 4.666666666666667,
"grad_norm": 0.34065279364585876,
"learning_rate": 3.173294878168025e-06,
"loss": 0.2931,
"step": 93
},
{
"epoch": 4.717948717948718,
"grad_norm": 0.30179089307785034,
"learning_rate": 3.057826018626527e-06,
"loss": 0.3212,
"step": 94
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.2840173840522766,
"learning_rate": 2.9435644843469434e-06,
"loss": 0.3295,
"step": 95
},
{
"epoch": 4.82051282051282,
"grad_norm": 0.26852840185165405,
"learning_rate": 2.83058130441221e-06,
"loss": 0.3035,
"step": 96
},
{
"epoch": 4.871794871794872,
"grad_norm": 0.30153700709342957,
"learning_rate": 2.718946713234185e-06,
"loss": 0.3392,
"step": 97
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.32083266973495483,
"learning_rate": 2.608730106893411e-06,
"loss": 0.3324,
"step": 98
},
{
"epoch": 4.9743589743589745,
"grad_norm": 0.26210591197013855,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.3356,
"step": 99
},
{
"epoch": 5.0,
"grad_norm": 0.26210591197013855,
"learning_rate": 2.39282398310251e-06,
"loss": 0.3178,
"step": 100
},
{
"epoch": 5.051282051282051,
"grad_norm": 0.21580009162425995,
"learning_rate": 2.2872686806712037e-06,
"loss": 0.3246,
"step": 101
},
{
"epoch": 5.102564102564102,
"grad_norm": 0.21004189550876617,
"learning_rate": 2.1833997096818897e-06,
"loss": 0.284,
"step": 102
},
{
"epoch": 5.153846153846154,
"grad_norm": 0.19976164400577545,
"learning_rate": 2.081281638826052e-06,
"loss": 0.3327,
"step": 103
},
{
"epoch": 5.205128205128205,
"grad_norm": 0.16318979859352112,
"learning_rate": 1.980977948372612e-06,
"loss": 0.2896,
"step": 104
},
{
"epoch": 5.256410256410256,
"grad_norm": 0.21000930666923523,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.2783,
"step": 105
},
{
"epoch": 5.3076923076923075,
"grad_norm": 0.2006593644618988,
"learning_rate": 1.7860619515673034e-06,
"loss": 0.2768,
"step": 106
},
{
"epoch": 5.358974358974359,
"grad_norm": 0.20942449569702148,
"learning_rate": 1.6915708120157042e-06,
"loss": 0.3332,
"step": 107
},
{
"epoch": 5.410256410256411,
"grad_norm": 0.20298852026462555,
"learning_rate": 1.5991363111454023e-06,
"loss": 0.2723,
"step": 108
},
{
"epoch": 5.461538461538462,
"grad_norm": 0.16583769023418427,
"learning_rate": 1.5088159095696365e-06,
"loss": 0.3119,
"step": 109
},
{
"epoch": 5.512820512820513,
"grad_norm": 0.2339308112859726,
"learning_rate": 1.4206657537014078e-06,
"loss": 0.2772,
"step": 110
},
{
"epoch": 5.564102564102564,
"grad_norm": 0.18335282802581787,
"learning_rate": 1.3347406408508695e-06,
"loss": 0.2719,
"step": 111
},
{
"epoch": 5.615384615384615,
"grad_norm": 0.1852513551712036,
"learning_rate": 1.2510939851613285e-06,
"loss": 0.305,
"step": 112
},
{
"epoch": 5.666666666666667,
"grad_norm": 0.15820161998271942,
"learning_rate": 1.1697777844051105e-06,
"loss": 0.2776,
"step": 113
},
{
"epoch": 5.717948717948718,
"grad_norm": 0.19469872117042542,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.3214,
"step": 114
},
{
"epoch": 5.769230769230769,
"grad_norm": 0.15481670200824738,
"learning_rate": 1.0143374638853892e-06,
"loss": 0.2907,
"step": 115
},
{
"epoch": 5.82051282051282,
"grad_norm": 0.1693563610315323,
"learning_rate": 9.403099714207175e-07,
"loss": 0.2971,
"step": 116
},
{
"epoch": 5.871794871794872,
"grad_norm": 0.23726695775985718,
"learning_rate": 8.688061284200266e-07,
"loss": 0.2962,
"step": 117
},
{
"epoch": 5.923076923076923,
"grad_norm": 0.16373063623905182,
"learning_rate": 7.99870384246143e-07,
"loss": 0.2969,
"step": 118
},
{
"epoch": 5.9743589743589745,
"grad_norm": 0.1886770874261856,
"learning_rate": 7.33545591839222e-07,
"loss": 0.2807,
"step": 119
},
{
"epoch": 6.0,
"grad_norm": 0.42370569705963135,
"learning_rate": 6.698729810778065e-07,
"loss": 0.3277,
"step": 120
},
{
"epoch": 6.051282051282051,
"grad_norm": 0.26122286915779114,
"learning_rate": 6.088921331488568e-07,
"loss": 0.2549,
"step": 121
},
{
"epoch": 6.102564102564102,
"grad_norm": 0.25389620661735535,
"learning_rate": 5.506409559426573e-07,
"loss": 0.308,
"step": 122
},
{
"epoch": 6.153846153846154,
"grad_norm": 0.2587980628013611,
"learning_rate": 4.951556604879049e-07,
"loss": 0.3121,
"step": 123
},
{
"epoch": 6.205128205128205,
"grad_norm": 0.24804840981960297,
"learning_rate": 4.4247073844163434e-07,
"loss": 0.2875,
"step": 124
},
{
"epoch": 6.256410256410256,
"grad_norm": 0.24467125535011292,
"learning_rate": 3.9261894064796136e-07,
"loss": 0.3057,
"step": 125
},
{
"epoch": 6.3076923076923075,
"grad_norm": 0.2763948440551758,
"learning_rate": 3.4563125677897936e-07,
"loss": 0.2734,
"step": 126
},
{
"epoch": 6.358974358974359,
"grad_norm": 0.24583575129508972,
"learning_rate": 3.015368960704584e-07,
"loss": 0.3049,
"step": 127
},
{
"epoch": 6.410256410256411,
"grad_norm": 0.21760328114032745,
"learning_rate": 2.6036326916434153e-07,
"loss": 0.2916,
"step": 128
},
{
"epoch": 6.461538461538462,
"grad_norm": 0.2987053394317627,
"learning_rate": 2.2213597106929608e-07,
"loss": 0.2784,
"step": 129
},
{
"epoch": 6.512820512820513,
"grad_norm": 0.27152180671691895,
"learning_rate": 1.8687876524993987e-07,
"loss": 0.2616,
"step": 130
},
{
"epoch": 6.564102564102564,
"grad_norm": 0.24838034808635712,
"learning_rate": 1.5461356885461077e-07,
"loss": 0.3087,
"step": 131
},
{
"epoch": 6.615384615384615,
"grad_norm": 0.3044689893722534,
"learning_rate": 1.253604390908819e-07,
"loss": 0.3187,
"step": 132
},
{
"epoch": 6.666666666666667,
"grad_norm": 0.2829769253730774,
"learning_rate": 9.913756075728088e-08,
"loss": 0.2477,
"step": 133
},
{
"epoch": 6.717948717948718,
"grad_norm": 0.2556080222129822,
"learning_rate": 7.59612349389599e-08,
"loss": 0.2936,
"step": 134
},
{
"epoch": 6.769230769230769,
"grad_norm": 0.24655228853225708,
"learning_rate": 5.584586887435739e-08,
"loss": 0.3051,
"step": 135
},
{
"epoch": 6.82051282051282,
"grad_norm": 0.24615831673145294,
"learning_rate": 3.8803966999139686e-08,
"loss": 0.2631,
"step": 136
},
{
"epoch": 6.871794871794872,
"grad_norm": 0.25910142064094543,
"learning_rate": 2.4846123172992953e-08,
"loss": 0.2717,
"step": 137
},
{
"epoch": 6.923076923076923,
"grad_norm": 0.27064117789268494,
"learning_rate": 1.3981014094099354e-08,
"loss": 0.3052,
"step": 138
},
{
"epoch": 6.9743589743589745,
"grad_norm": 0.23766082525253296,
"learning_rate": 6.215393905388278e-09,
"loss": 0.3186,
"step": 139
},
{
"epoch": 7.0,
"grad_norm": 0.23766082525253296,
"learning_rate": 1.5540899959187727e-09,
"loss": 0.344,
"step": 140
},
{
"epoch": 7.0,
"step": 140,
"total_flos": 1.609991857138303e+18,
"train_loss": 0.3682906585080283,
"train_runtime": 4813.8883,
"train_samples_per_second": 0.904,
"train_steps_per_second": 0.029
}
],
"logging_steps": 1,
"max_steps": 140,
"num_input_tokens_seen": 0,
"num_train_epochs": 7,
"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": 1.609991857138303e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}