8688 lines
210 KiB
JSON
8688 lines
210 KiB
JSON
|
|
{
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 5.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 1235,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.004048582995951417,
|
||
|
|
"grad_norm": 6.529969639004119,
|
||
|
|
"learning_rate": 3.2258064516129035e-07,
|
||
|
|
"loss": 1.0591,
|
||
|
|
"step": 1
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.008097165991902834,
|
||
|
|
"grad_norm": 6.6200958359971525,
|
||
|
|
"learning_rate": 6.451612903225807e-07,
|
||
|
|
"loss": 1.0685,
|
||
|
|
"step": 2
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.012145748987854251,
|
||
|
|
"grad_norm": 6.5053460954628495,
|
||
|
|
"learning_rate": 9.67741935483871e-07,
|
||
|
|
"loss": 1.0446,
|
||
|
|
"step": 3
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.016194331983805668,
|
||
|
|
"grad_norm": 6.38964653410593,
|
||
|
|
"learning_rate": 1.2903225806451614e-06,
|
||
|
|
"loss": 1.0575,
|
||
|
|
"step": 4
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.020242914979757085,
|
||
|
|
"grad_norm": 6.256048954955403,
|
||
|
|
"learning_rate": 1.6129032258064516e-06,
|
||
|
|
"loss": 1.0547,
|
||
|
|
"step": 5
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.024291497975708502,
|
||
|
|
"grad_norm": 5.906526094468779,
|
||
|
|
"learning_rate": 1.935483870967742e-06,
|
||
|
|
"loss": 1.0443,
|
||
|
|
"step": 6
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.02834008097165992,
|
||
|
|
"grad_norm": 4.720924779310831,
|
||
|
|
"learning_rate": 2.2580645161290324e-06,
|
||
|
|
"loss": 1.0039,
|
||
|
|
"step": 7
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.032388663967611336,
|
||
|
|
"grad_norm": 4.3557410458205155,
|
||
|
|
"learning_rate": 2.580645161290323e-06,
|
||
|
|
"loss": 1.004,
|
||
|
|
"step": 8
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03643724696356275,
|
||
|
|
"grad_norm": 2.8932334090195897,
|
||
|
|
"learning_rate": 2.903225806451613e-06,
|
||
|
|
"loss": 0.9558,
|
||
|
|
"step": 9
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04048582995951417,
|
||
|
|
"grad_norm": 2.652247270259386,
|
||
|
|
"learning_rate": 3.225806451612903e-06,
|
||
|
|
"loss": 0.9629,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.044534412955465584,
|
||
|
|
"grad_norm": 2.545365253066352,
|
||
|
|
"learning_rate": 3.548387096774194e-06,
|
||
|
|
"loss": 0.9542,
|
||
|
|
"step": 11
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.048582995951417005,
|
||
|
|
"grad_norm": 3.9315528894192817,
|
||
|
|
"learning_rate": 3.870967741935484e-06,
|
||
|
|
"loss": 0.94,
|
||
|
|
"step": 12
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05263157894736842,
|
||
|
|
"grad_norm": 4.137544437321004,
|
||
|
|
"learning_rate": 4.193548387096774e-06,
|
||
|
|
"loss": 0.9349,
|
||
|
|
"step": 13
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05668016194331984,
|
||
|
|
"grad_norm": 3.9465378672538676,
|
||
|
|
"learning_rate": 4.516129032258065e-06,
|
||
|
|
"loss": 0.9312,
|
||
|
|
"step": 14
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06072874493927125,
|
||
|
|
"grad_norm": 3.595489793720538,
|
||
|
|
"learning_rate": 4.838709677419355e-06,
|
||
|
|
"loss": 0.9129,
|
||
|
|
"step": 15
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06477732793522267,
|
||
|
|
"grad_norm": 2.6397721210601084,
|
||
|
|
"learning_rate": 5.161290322580646e-06,
|
||
|
|
"loss": 0.8783,
|
||
|
|
"step": 16
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06882591093117409,
|
||
|
|
"grad_norm": 2.4240178771196343,
|
||
|
|
"learning_rate": 5.483870967741935e-06,
|
||
|
|
"loss": 0.8682,
|
||
|
|
"step": 17
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0728744939271255,
|
||
|
|
"grad_norm": 2.00999422861566,
|
||
|
|
"learning_rate": 5.806451612903226e-06,
|
||
|
|
"loss": 0.8575,
|
||
|
|
"step": 18
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07692307692307693,
|
||
|
|
"grad_norm": 1.5223510919211352,
|
||
|
|
"learning_rate": 6.129032258064517e-06,
|
||
|
|
"loss": 0.8403,
|
||
|
|
"step": 19
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08097165991902834,
|
||
|
|
"grad_norm": 1.189574399590261,
|
||
|
|
"learning_rate": 6.451612903225806e-06,
|
||
|
|
"loss": 0.795,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08502024291497975,
|
||
|
|
"grad_norm": 1.124160330311177,
|
||
|
|
"learning_rate": 6.774193548387097e-06,
|
||
|
|
"loss": 0.8068,
|
||
|
|
"step": 21
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08906882591093117,
|
||
|
|
"grad_norm": 1.2321755393793086,
|
||
|
|
"learning_rate": 7.096774193548388e-06,
|
||
|
|
"loss": 0.8264,
|
||
|
|
"step": 22
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0931174089068826,
|
||
|
|
"grad_norm": 1.1157221564859934,
|
||
|
|
"learning_rate": 7.4193548387096784e-06,
|
||
|
|
"loss": 0.7984,
|
||
|
|
"step": 23
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09716599190283401,
|
||
|
|
"grad_norm": 0.9234055073791403,
|
||
|
|
"learning_rate": 7.741935483870968e-06,
|
||
|
|
"loss": 0.7919,
|
||
|
|
"step": 24
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10121457489878542,
|
||
|
|
"grad_norm": 0.8824709909470099,
|
||
|
|
"learning_rate": 8.064516129032258e-06,
|
||
|
|
"loss": 0.7575,
|
||
|
|
"step": 25
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10526315789473684,
|
||
|
|
"grad_norm": 0.8927990673095986,
|
||
|
|
"learning_rate": 8.387096774193549e-06,
|
||
|
|
"loss": 0.7722,
|
||
|
|
"step": 26
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10931174089068826,
|
||
|
|
"grad_norm": 0.7904767128873359,
|
||
|
|
"learning_rate": 8.70967741935484e-06,
|
||
|
|
"loss": 0.7682,
|
||
|
|
"step": 27
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11336032388663968,
|
||
|
|
"grad_norm": 0.7835737436161865,
|
||
|
|
"learning_rate": 9.03225806451613e-06,
|
||
|
|
"loss": 0.7439,
|
||
|
|
"step": 28
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11740890688259109,
|
||
|
|
"grad_norm": 0.7039714072286063,
|
||
|
|
"learning_rate": 9.35483870967742e-06,
|
||
|
|
"loss": 0.7359,
|
||
|
|
"step": 29
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1214574898785425,
|
||
|
|
"grad_norm": 0.742602582354179,
|
||
|
|
"learning_rate": 9.67741935483871e-06,
|
||
|
|
"loss": 0.7288,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12550607287449392,
|
||
|
|
"grad_norm": 0.6895474616730898,
|
||
|
|
"learning_rate": 1e-05,
|
||
|
|
"loss": 0.7433,
|
||
|
|
"step": 31
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12955465587044535,
|
||
|
|
"grad_norm": 0.67546853744356,
|
||
|
|
"learning_rate": 1.0322580645161291e-05,
|
||
|
|
"loss": 0.715,
|
||
|
|
"step": 32
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13360323886639677,
|
||
|
|
"grad_norm": 0.6837612229305043,
|
||
|
|
"learning_rate": 1.0645161290322582e-05,
|
||
|
|
"loss": 0.7312,
|
||
|
|
"step": 33
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13765182186234817,
|
||
|
|
"grad_norm": 0.6563589423404679,
|
||
|
|
"learning_rate": 1.096774193548387e-05,
|
||
|
|
"loss": 0.7278,
|
||
|
|
"step": 34
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1417004048582996,
|
||
|
|
"grad_norm": 0.6861089450244338,
|
||
|
|
"learning_rate": 1.1290322580645164e-05,
|
||
|
|
"loss": 0.7237,
|
||
|
|
"step": 35
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.145748987854251,
|
||
|
|
"grad_norm": 0.64416899060065,
|
||
|
|
"learning_rate": 1.1612903225806453e-05,
|
||
|
|
"loss": 0.7165,
|
||
|
|
"step": 36
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14979757085020243,
|
||
|
|
"grad_norm": 0.6726397275319622,
|
||
|
|
"learning_rate": 1.1935483870967743e-05,
|
||
|
|
"loss": 0.6971,
|
||
|
|
"step": 37
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15384615384615385,
|
||
|
|
"grad_norm": 0.6120205136061637,
|
||
|
|
"learning_rate": 1.2258064516129034e-05,
|
||
|
|
"loss": 0.7136,
|
||
|
|
"step": 38
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15789473684210525,
|
||
|
|
"grad_norm": 0.5842112714653325,
|
||
|
|
"learning_rate": 1.2580645161290324e-05,
|
||
|
|
"loss": 0.7263,
|
||
|
|
"step": 39
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16194331983805668,
|
||
|
|
"grad_norm": 0.5985131754938803,
|
||
|
|
"learning_rate": 1.2903225806451613e-05,
|
||
|
|
"loss": 0.7265,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1659919028340081,
|
||
|
|
"grad_norm": 0.5421734335942299,
|
||
|
|
"learning_rate": 1.3225806451612903e-05,
|
||
|
|
"loss": 0.7021,
|
||
|
|
"step": 41
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1700404858299595,
|
||
|
|
"grad_norm": 0.5439477578841015,
|
||
|
|
"learning_rate": 1.3548387096774194e-05,
|
||
|
|
"loss": 0.7127,
|
||
|
|
"step": 42
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17408906882591094,
|
||
|
|
"grad_norm": 0.548772122979161,
|
||
|
|
"learning_rate": 1.3870967741935486e-05,
|
||
|
|
"loss": 0.6885,
|
||
|
|
"step": 43
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17813765182186234,
|
||
|
|
"grad_norm": 0.5438852285801518,
|
||
|
|
"learning_rate": 1.4193548387096776e-05,
|
||
|
|
"loss": 0.6976,
|
||
|
|
"step": 44
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18218623481781376,
|
||
|
|
"grad_norm": 0.703789779018483,
|
||
|
|
"learning_rate": 1.4516129032258066e-05,
|
||
|
|
"loss": 0.6799,
|
||
|
|
"step": 45
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1862348178137652,
|
||
|
|
"grad_norm": 0.5839323106634435,
|
||
|
|
"learning_rate": 1.4838709677419357e-05,
|
||
|
|
"loss": 0.6994,
|
||
|
|
"step": 46
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1902834008097166,
|
||
|
|
"grad_norm": 0.5047233313587509,
|
||
|
|
"learning_rate": 1.5161290322580646e-05,
|
||
|
|
"loss": 0.6885,
|
||
|
|
"step": 47
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19433198380566802,
|
||
|
|
"grad_norm": 0.5466473189578837,
|
||
|
|
"learning_rate": 1.5483870967741936e-05,
|
||
|
|
"loss": 0.6747,
|
||
|
|
"step": 48
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19838056680161945,
|
||
|
|
"grad_norm": 0.5003352871369224,
|
||
|
|
"learning_rate": 1.5806451612903226e-05,
|
||
|
|
"loss": 0.6864,
|
||
|
|
"step": 49
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20242914979757085,
|
||
|
|
"grad_norm": 0.6332807756583119,
|
||
|
|
"learning_rate": 1.6129032258064517e-05,
|
||
|
|
"loss": 0.6924,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20647773279352227,
|
||
|
|
"grad_norm": 0.6354224122058385,
|
||
|
|
"learning_rate": 1.6451612903225807e-05,
|
||
|
|
"loss": 0.7048,
|
||
|
|
"step": 51
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21052631578947367,
|
||
|
|
"grad_norm": 0.6383184074404532,
|
||
|
|
"learning_rate": 1.6774193548387098e-05,
|
||
|
|
"loss": 0.6775,
|
||
|
|
"step": 52
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2145748987854251,
|
||
|
|
"grad_norm": 0.6105784591765013,
|
||
|
|
"learning_rate": 1.7096774193548388e-05,
|
||
|
|
"loss": 0.6836,
|
||
|
|
"step": 53
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21862348178137653,
|
||
|
|
"grad_norm": 0.7397161991819674,
|
||
|
|
"learning_rate": 1.741935483870968e-05,
|
||
|
|
"loss": 0.6806,
|
||
|
|
"step": 54
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22267206477732793,
|
||
|
|
"grad_norm": 0.6293965991496899,
|
||
|
|
"learning_rate": 1.774193548387097e-05,
|
||
|
|
"loss": 0.6801,
|
||
|
|
"step": 55
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22672064777327935,
|
||
|
|
"grad_norm": 0.6033220591385403,
|
||
|
|
"learning_rate": 1.806451612903226e-05,
|
||
|
|
"loss": 0.6763,
|
||
|
|
"step": 56
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23076923076923078,
|
||
|
|
"grad_norm": 0.6757597984809728,
|
||
|
|
"learning_rate": 1.838709677419355e-05,
|
||
|
|
"loss": 0.6671,
|
||
|
|
"step": 57
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23481781376518218,
|
||
|
|
"grad_norm": 0.6453830325096583,
|
||
|
|
"learning_rate": 1.870967741935484e-05,
|
||
|
|
"loss": 0.6653,
|
||
|
|
"step": 58
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2388663967611336,
|
||
|
|
"grad_norm": 0.6320988356582008,
|
||
|
|
"learning_rate": 1.903225806451613e-05,
|
||
|
|
"loss": 0.6728,
|
||
|
|
"step": 59
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.242914979757085,
|
||
|
|
"grad_norm": 0.7455217279066642,
|
||
|
|
"learning_rate": 1.935483870967742e-05,
|
||
|
|
"loss": 0.6864,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24696356275303644,
|
||
|
|
"grad_norm": 0.6021375449643891,
|
||
|
|
"learning_rate": 1.967741935483871e-05,
|
||
|
|
"loss": 0.6658,
|
||
|
|
"step": 61
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25101214574898784,
|
||
|
|
"grad_norm": 0.6516520040260566,
|
||
|
|
"learning_rate": 2e-05,
|
||
|
|
"loss": 0.6849,
|
||
|
|
"step": 62
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2550607287449393,
|
||
|
|
"grad_norm": 0.7753318890823696,
|
||
|
|
"learning_rate": 2.0322580645161292e-05,
|
||
|
|
"loss": 0.6655,
|
||
|
|
"step": 63
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2591093117408907,
|
||
|
|
"grad_norm": 0.866848128282218,
|
||
|
|
"learning_rate": 2.0645161290322582e-05,
|
||
|
|
"loss": 0.6771,
|
||
|
|
"step": 64
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2631578947368421,
|
||
|
|
"grad_norm": 0.5729195151500129,
|
||
|
|
"learning_rate": 2.0967741935483873e-05,
|
||
|
|
"loss": 0.6453,
|
||
|
|
"step": 65
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26720647773279355,
|
||
|
|
"grad_norm": 0.7500011761258638,
|
||
|
|
"learning_rate": 2.1290322580645163e-05,
|
||
|
|
"loss": 0.6723,
|
||
|
|
"step": 66
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27125506072874495,
|
||
|
|
"grad_norm": 0.7930700971827921,
|
||
|
|
"learning_rate": 2.161290322580645e-05,
|
||
|
|
"loss": 0.6622,
|
||
|
|
"step": 67
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27530364372469635,
|
||
|
|
"grad_norm": 0.7044288300845005,
|
||
|
|
"learning_rate": 2.193548387096774e-05,
|
||
|
|
"loss": 0.6652,
|
||
|
|
"step": 68
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2793522267206478,
|
||
|
|
"grad_norm": 0.7470327200358218,
|
||
|
|
"learning_rate": 2.225806451612903e-05,
|
||
|
|
"loss": 0.6665,
|
||
|
|
"step": 69
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2834008097165992,
|
||
|
|
"grad_norm": 0.8061535159957012,
|
||
|
|
"learning_rate": 2.2580645161290328e-05,
|
||
|
|
"loss": 0.6716,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2874493927125506,
|
||
|
|
"grad_norm": 0.7038774097264121,
|
||
|
|
"learning_rate": 2.2903225806451615e-05,
|
||
|
|
"loss": 0.6728,
|
||
|
|
"step": 71
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.291497975708502,
|
||
|
|
"grad_norm": 0.7066161590002146,
|
||
|
|
"learning_rate": 2.3225806451612906e-05,
|
||
|
|
"loss": 0.6559,
|
||
|
|
"step": 72
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29554655870445345,
|
||
|
|
"grad_norm": 0.6054155989171657,
|
||
|
|
"learning_rate": 2.3548387096774196e-05,
|
||
|
|
"loss": 0.6538,
|
||
|
|
"step": 73
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29959514170040485,
|
||
|
|
"grad_norm": 0.6915675088727053,
|
||
|
|
"learning_rate": 2.3870967741935486e-05,
|
||
|
|
"loss": 0.6485,
|
||
|
|
"step": 74
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30364372469635625,
|
||
|
|
"grad_norm": 0.8360497101963237,
|
||
|
|
"learning_rate": 2.4193548387096777e-05,
|
||
|
|
"loss": 0.6642,
|
||
|
|
"step": 75
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3076923076923077,
|
||
|
|
"grad_norm": 0.8600219267094722,
|
||
|
|
"learning_rate": 2.4516129032258067e-05,
|
||
|
|
"loss": 0.6621,
|
||
|
|
"step": 76
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3117408906882591,
|
||
|
|
"grad_norm": 0.8595703863626754,
|
||
|
|
"learning_rate": 2.4838709677419358e-05,
|
||
|
|
"loss": 0.6592,
|
||
|
|
"step": 77
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3157894736842105,
|
||
|
|
"grad_norm": 1.1595341999461948,
|
||
|
|
"learning_rate": 2.5161290322580648e-05,
|
||
|
|
"loss": 0.6716,
|
||
|
|
"step": 78
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31983805668016196,
|
||
|
|
"grad_norm": 1.0409445503540686,
|
||
|
|
"learning_rate": 2.548387096774194e-05,
|
||
|
|
"loss": 0.6612,
|
||
|
|
"step": 79
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32388663967611336,
|
||
|
|
"grad_norm": 1.2267989679211295,
|
||
|
|
"learning_rate": 2.5806451612903226e-05,
|
||
|
|
"loss": 0.661,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32793522267206476,
|
||
|
|
"grad_norm": 0.7779818541728512,
|
||
|
|
"learning_rate": 2.6129032258064516e-05,
|
||
|
|
"loss": 0.6682,
|
||
|
|
"step": 81
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3319838056680162,
|
||
|
|
"grad_norm": 0.6786494700253207,
|
||
|
|
"learning_rate": 2.6451612903225806e-05,
|
||
|
|
"loss": 0.6401,
|
||
|
|
"step": 82
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3360323886639676,
|
||
|
|
"grad_norm": 0.9364366485836457,
|
||
|
|
"learning_rate": 2.6774193548387097e-05,
|
||
|
|
"loss": 0.6528,
|
||
|
|
"step": 83
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.340080971659919,
|
||
|
|
"grad_norm": 0.9358927507972675,
|
||
|
|
"learning_rate": 2.7096774193548387e-05,
|
||
|
|
"loss": 0.6716,
|
||
|
|
"step": 84
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3441295546558704,
|
||
|
|
"grad_norm": 1.186303936950258,
|
||
|
|
"learning_rate": 2.7419354838709678e-05,
|
||
|
|
"loss": 0.6498,
|
||
|
|
"step": 85
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3481781376518219,
|
||
|
|
"grad_norm": 1.0176956447006573,
|
||
|
|
"learning_rate": 2.774193548387097e-05,
|
||
|
|
"loss": 0.6483,
|
||
|
|
"step": 86
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3522267206477733,
|
||
|
|
"grad_norm": 0.763434015594322,
|
||
|
|
"learning_rate": 2.8064516129032262e-05,
|
||
|
|
"loss": 0.6478,
|
||
|
|
"step": 87
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3562753036437247,
|
||
|
|
"grad_norm": 0.8412683277547806,
|
||
|
|
"learning_rate": 2.8387096774193552e-05,
|
||
|
|
"loss": 0.6637,
|
||
|
|
"step": 88
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3603238866396761,
|
||
|
|
"grad_norm": 1.1471166103609722,
|
||
|
|
"learning_rate": 2.8709677419354843e-05,
|
||
|
|
"loss": 0.6422,
|
||
|
|
"step": 89
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3643724696356275,
|
||
|
|
"grad_norm": 1.02560876253542,
|
||
|
|
"learning_rate": 2.9032258064516133e-05,
|
||
|
|
"loss": 0.6543,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3684210526315789,
|
||
|
|
"grad_norm": 1.0261130982502162,
|
||
|
|
"learning_rate": 2.9354838709677423e-05,
|
||
|
|
"loss": 0.6482,
|
||
|
|
"step": 91
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3724696356275304,
|
||
|
|
"grad_norm": 0.9985680238163215,
|
||
|
|
"learning_rate": 2.9677419354838714e-05,
|
||
|
|
"loss": 0.6501,
|
||
|
|
"step": 92
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3765182186234818,
|
||
|
|
"grad_norm": 0.822642203018212,
|
||
|
|
"learning_rate": 3.0000000000000004e-05,
|
||
|
|
"loss": 0.6552,
|
||
|
|
"step": 93
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3805668016194332,
|
||
|
|
"grad_norm": 1.3385269411223348,
|
||
|
|
"learning_rate": 3.032258064516129e-05,
|
||
|
|
"loss": 0.6581,
|
||
|
|
"step": 94
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38461538461538464,
|
||
|
|
"grad_norm": 1.0104061576274967,
|
||
|
|
"learning_rate": 3.0645161290322585e-05,
|
||
|
|
"loss": 0.6574,
|
||
|
|
"step": 95
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38866396761133604,
|
||
|
|
"grad_norm": 1.395722260369119,
|
||
|
|
"learning_rate": 3.096774193548387e-05,
|
||
|
|
"loss": 0.6683,
|
||
|
|
"step": 96
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.39271255060728744,
|
||
|
|
"grad_norm": 0.7630998511082215,
|
||
|
|
"learning_rate": 3.1290322580645166e-05,
|
||
|
|
"loss": 0.6435,
|
||
|
|
"step": 97
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3967611336032389,
|
||
|
|
"grad_norm": 0.9598538021255822,
|
||
|
|
"learning_rate": 3.161290322580645e-05,
|
||
|
|
"loss": 0.6502,
|
||
|
|
"step": 98
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4008097165991903,
|
||
|
|
"grad_norm": 1.2624432100322756,
|
||
|
|
"learning_rate": 3.193548387096774e-05,
|
||
|
|
"loss": 0.6579,
|
||
|
|
"step": 99
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4048582995951417,
|
||
|
|
"grad_norm": 1.3947492509385464,
|
||
|
|
"learning_rate": 3.2258064516129034e-05,
|
||
|
|
"loss": 0.6422,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4089068825910931,
|
||
|
|
"grad_norm": 0.7815112312968828,
|
||
|
|
"learning_rate": 3.258064516129033e-05,
|
||
|
|
"loss": 0.6481,
|
||
|
|
"step": 101
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.41295546558704455,
|
||
|
|
"grad_norm": 1.376172482674097,
|
||
|
|
"learning_rate": 3.2903225806451614e-05,
|
||
|
|
"loss": 0.6518,
|
||
|
|
"step": 102
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.41700404858299595,
|
||
|
|
"grad_norm": 0.9860394802734346,
|
||
|
|
"learning_rate": 3.322580645161291e-05,
|
||
|
|
"loss": 0.6502,
|
||
|
|
"step": 103
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.42105263157894735,
|
||
|
|
"grad_norm": 0.9483153291680856,
|
||
|
|
"learning_rate": 3.3548387096774195e-05,
|
||
|
|
"loss": 0.6305,
|
||
|
|
"step": 104
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4251012145748988,
|
||
|
|
"grad_norm": 0.8395001349312757,
|
||
|
|
"learning_rate": 3.387096774193549e-05,
|
||
|
|
"loss": 0.6524,
|
||
|
|
"step": 105
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4291497975708502,
|
||
|
|
"grad_norm": 1.217875542596544,
|
||
|
|
"learning_rate": 3.4193548387096776e-05,
|
||
|
|
"loss": 0.6257,
|
||
|
|
"step": 106
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4331983805668016,
|
||
|
|
"grad_norm": 0.8987334812224276,
|
||
|
|
"learning_rate": 3.451612903225807e-05,
|
||
|
|
"loss": 0.6413,
|
||
|
|
"step": 107
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.43724696356275305,
|
||
|
|
"grad_norm": 0.9749608839246254,
|
||
|
|
"learning_rate": 3.483870967741936e-05,
|
||
|
|
"loss": 0.6672,
|
||
|
|
"step": 108
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.44129554655870445,
|
||
|
|
"grad_norm": 1.1130462256658686,
|
||
|
|
"learning_rate": 3.516129032258065e-05,
|
||
|
|
"loss": 0.6421,
|
||
|
|
"step": 109
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.44534412955465585,
|
||
|
|
"grad_norm": 1.0313733954709317,
|
||
|
|
"learning_rate": 3.548387096774194e-05,
|
||
|
|
"loss": 0.635,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4493927125506073,
|
||
|
|
"grad_norm": 1.3343313614801513,
|
||
|
|
"learning_rate": 3.5806451612903225e-05,
|
||
|
|
"loss": 0.6354,
|
||
|
|
"step": 111
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4534412955465587,
|
||
|
|
"grad_norm": 0.9136566659216148,
|
||
|
|
"learning_rate": 3.612903225806452e-05,
|
||
|
|
"loss": 0.6434,
|
||
|
|
"step": 112
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4574898785425101,
|
||
|
|
"grad_norm": 0.8559261120779332,
|
||
|
|
"learning_rate": 3.6451612903225805e-05,
|
||
|
|
"loss": 0.6371,
|
||
|
|
"step": 113
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.46153846153846156,
|
||
|
|
"grad_norm": 1.1620663232974304,
|
||
|
|
"learning_rate": 3.67741935483871e-05,
|
||
|
|
"loss": 0.6251,
|
||
|
|
"step": 114
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.46558704453441296,
|
||
|
|
"grad_norm": 0.9172092098695629,
|
||
|
|
"learning_rate": 3.7096774193548386e-05,
|
||
|
|
"loss": 0.6353,
|
||
|
|
"step": 115
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.46963562753036436,
|
||
|
|
"grad_norm": 0.817297456923834,
|
||
|
|
"learning_rate": 3.741935483870968e-05,
|
||
|
|
"loss": 0.644,
|
||
|
|
"step": 116
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.47368421052631576,
|
||
|
|
"grad_norm": 0.7972250031264954,
|
||
|
|
"learning_rate": 3.7741935483870974e-05,
|
||
|
|
"loss": 0.6488,
|
||
|
|
"step": 117
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4777327935222672,
|
||
|
|
"grad_norm": 1.1227209164357583,
|
||
|
|
"learning_rate": 3.806451612903226e-05,
|
||
|
|
"loss": 0.6593,
|
||
|
|
"step": 118
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4817813765182186,
|
||
|
|
"grad_norm": 0.9952477658398028,
|
||
|
|
"learning_rate": 3.8387096774193555e-05,
|
||
|
|
"loss": 0.6397,
|
||
|
|
"step": 119
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.48582995951417,
|
||
|
|
"grad_norm": 1.1809215504883686,
|
||
|
|
"learning_rate": 3.870967741935484e-05,
|
||
|
|
"loss": 0.6481,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4898785425101215,
|
||
|
|
"grad_norm": 1.438907188392669,
|
||
|
|
"learning_rate": 3.9032258064516135e-05,
|
||
|
|
"loss": 0.6309,
|
||
|
|
"step": 121
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4939271255060729,
|
||
|
|
"grad_norm": 0.7663114723438924,
|
||
|
|
"learning_rate": 3.935483870967742e-05,
|
||
|
|
"loss": 0.6389,
|
||
|
|
"step": 122
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4979757085020243,
|
||
|
|
"grad_norm": 1.5680395279138002,
|
||
|
|
"learning_rate": 3.9677419354838716e-05,
|
||
|
|
"loss": 0.6432,
|
||
|
|
"step": 123
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5020242914979757,
|
||
|
|
"grad_norm": 1.0327193880827006,
|
||
|
|
"learning_rate": 4e-05,
|
||
|
|
"loss": 0.6348,
|
||
|
|
"step": 124
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5060728744939271,
|
||
|
|
"grad_norm": 1.1687309323783,
|
||
|
|
"learning_rate": 3.9999920040266474e-05,
|
||
|
|
"loss": 0.638,
|
||
|
|
"step": 125
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5101214574898786,
|
||
|
|
"grad_norm": 1.0910714606480205,
|
||
|
|
"learning_rate": 3.999968016170525e-05,
|
||
|
|
"loss": 0.6616,
|
||
|
|
"step": 126
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5141700404858299,
|
||
|
|
"grad_norm": 1.1076017858087872,
|
||
|
|
"learning_rate": 3.9999280366234394e-05,
|
||
|
|
"loss": 0.6304,
|
||
|
|
"step": 127
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5182186234817814,
|
||
|
|
"grad_norm": 0.9923133005646836,
|
||
|
|
"learning_rate": 3.999872065705065e-05,
|
||
|
|
"loss": 0.6288,
|
||
|
|
"step": 128
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5222672064777328,
|
||
|
|
"grad_norm": 1.0804508855939996,
|
||
|
|
"learning_rate": 3.999800103862945e-05,
|
||
|
|
"loss": 0.6365,
|
||
|
|
"step": 129
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5263157894736842,
|
||
|
|
"grad_norm": 0.817665032911421,
|
||
|
|
"learning_rate": 3.9997121516724834e-05,
|
||
|
|
"loss": 0.6314,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5303643724696356,
|
||
|
|
"grad_norm": 0.9352471916864006,
|
||
|
|
"learning_rate": 3.999608209836944e-05,
|
||
|
|
"loss": 0.6323,
|
||
|
|
"step": 131
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5344129554655871,
|
||
|
|
"grad_norm": 0.8817695822294224,
|
||
|
|
"learning_rate": 3.999488279187443e-05,
|
||
|
|
"loss": 0.6424,
|
||
|
|
"step": 132
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5384615384615384,
|
||
|
|
"grad_norm": 0.8755749602851766,
|
||
|
|
"learning_rate": 3.9993523606829425e-05,
|
||
|
|
"loss": 0.6364,
|
||
|
|
"step": 133
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5425101214574899,
|
||
|
|
"grad_norm": 0.8934221664965026,
|
||
|
|
"learning_rate": 3.9992004554102435e-05,
|
||
|
|
"loss": 0.6324,
|
||
|
|
"step": 134
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5465587044534413,
|
||
|
|
"grad_norm": 1.0782795203324302,
|
||
|
|
"learning_rate": 3.999032564583976e-05,
|
||
|
|
"loss": 0.6321,
|
||
|
|
"step": 135
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5506072874493927,
|
||
|
|
"grad_norm": 1.1322382934999973,
|
||
|
|
"learning_rate": 3.998848689546591e-05,
|
||
|
|
"loss": 0.6235,
|
||
|
|
"step": 136
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5546558704453441,
|
||
|
|
"grad_norm": 0.8422117252027796,
|
||
|
|
"learning_rate": 3.9986488317683496e-05,
|
||
|
|
"loss": 0.6497,
|
||
|
|
"step": 137
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5587044534412956,
|
||
|
|
"grad_norm": 0.832426124213949,
|
||
|
|
"learning_rate": 3.9984329928473067e-05,
|
||
|
|
"loss": 0.6413,
|
||
|
|
"step": 138
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.562753036437247,
|
||
|
|
"grad_norm": 1.0340569952005274,
|
||
|
|
"learning_rate": 3.998201174509306e-05,
|
||
|
|
"loss": 0.6367,
|
||
|
|
"step": 139
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5668016194331984,
|
||
|
|
"grad_norm": 0.8527185628226626,
|
||
|
|
"learning_rate": 3.997953378607961e-05,
|
||
|
|
"loss": 0.646,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5708502024291497,
|
||
|
|
"grad_norm": 0.9958832726645142,
|
||
|
|
"learning_rate": 3.9976896071246403e-05,
|
||
|
|
"loss": 0.6422,
|
||
|
|
"step": 141
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5748987854251012,
|
||
|
|
"grad_norm": 1.2128829428930699,
|
||
|
|
"learning_rate": 3.997409862168454e-05,
|
||
|
|
"loss": 0.6372,
|
||
|
|
"step": 142
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5789473684210527,
|
||
|
|
"grad_norm": 0.689582620288129,
|
||
|
|
"learning_rate": 3.997114145976237e-05,
|
||
|
|
"loss": 0.6338,
|
||
|
|
"step": 143
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.582995951417004,
|
||
|
|
"grad_norm": 1.062778190567828,
|
||
|
|
"learning_rate": 3.9968024609125245e-05,
|
||
|
|
"loss": 0.6401,
|
||
|
|
"step": 144
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5870445344129555,
|
||
|
|
"grad_norm": 0.90248862633625,
|
||
|
|
"learning_rate": 3.996474809469545e-05,
|
||
|
|
"loss": 0.6324,
|
||
|
|
"step": 145
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5910931174089069,
|
||
|
|
"grad_norm": 1.043632473165674,
|
||
|
|
"learning_rate": 3.996131194267189e-05,
|
||
|
|
"loss": 0.638,
|
||
|
|
"step": 146
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5951417004048583,
|
||
|
|
"grad_norm": 0.7850950492848126,
|
||
|
|
"learning_rate": 3.995771618052995e-05,
|
||
|
|
"loss": 0.6395,
|
||
|
|
"step": 147
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5991902834008097,
|
||
|
|
"grad_norm": 0.7894320986198138,
|
||
|
|
"learning_rate": 3.9953960837021256e-05,
|
||
|
|
"loss": 0.6326,
|
||
|
|
"step": 148
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6032388663967612,
|
||
|
|
"grad_norm": 0.803501288869217,
|
||
|
|
"learning_rate": 3.995004594217343e-05,
|
||
|
|
"loss": 0.6211,
|
||
|
|
"step": 149
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6072874493927125,
|
||
|
|
"grad_norm": 0.9247824586670753,
|
||
|
|
"learning_rate": 3.994597152728987e-05,
|
||
|
|
"loss": 0.6295,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.611336032388664,
|
||
|
|
"grad_norm": 0.8989681645653814,
|
||
|
|
"learning_rate": 3.994173762494947e-05,
|
||
|
|
"loss": 0.6443,
|
||
|
|
"step": 151
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6153846153846154,
|
||
|
|
"grad_norm": 0.722741778431614,
|
||
|
|
"learning_rate": 3.9937344269006424e-05,
|
||
|
|
"loss": 0.6322,
|
||
|
|
"step": 152
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6194331983805668,
|
||
|
|
"grad_norm": 0.7436875364196052,
|
||
|
|
"learning_rate": 3.993279149458988e-05,
|
||
|
|
"loss": 0.6451,
|
||
|
|
"step": 153
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6234817813765182,
|
||
|
|
"grad_norm": 0.7578110975199271,
|
||
|
|
"learning_rate": 3.9928079338103695e-05,
|
||
|
|
"loss": 0.6351,
|
||
|
|
"step": 154
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6275303643724697,
|
||
|
|
"grad_norm": 0.7050694314401702,
|
||
|
|
"learning_rate": 3.992320783722616e-05,
|
||
|
|
"loss": 0.6286,
|
||
|
|
"step": 155
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.631578947368421,
|
||
|
|
"grad_norm": 0.7067647872750286,
|
||
|
|
"learning_rate": 3.991817703090965e-05,
|
||
|
|
"loss": 0.6463,
|
||
|
|
"step": 156
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6356275303643725,
|
||
|
|
"grad_norm": 0.983934515966914,
|
||
|
|
"learning_rate": 3.991298695938038e-05,
|
||
|
|
"loss": 0.6305,
|
||
|
|
"step": 157
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6396761133603239,
|
||
|
|
"grad_norm": 0.7954101351871492,
|
||
|
|
"learning_rate": 3.9907637664138e-05,
|
||
|
|
"loss": 0.6347,
|
||
|
|
"step": 158
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6437246963562753,
|
||
|
|
"grad_norm": 0.775081487114188,
|
||
|
|
"learning_rate": 3.990212918795534e-05,
|
||
|
|
"loss": 0.6328,
|
||
|
|
"step": 159
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6477732793522267,
|
||
|
|
"grad_norm": 1.5872758288659552,
|
||
|
|
"learning_rate": 3.9896461574878036e-05,
|
||
|
|
"loss": 0.6487,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6518218623481782,
|
||
|
|
"grad_norm": 0.7168847242793819,
|
||
|
|
"learning_rate": 3.989063487022416e-05,
|
||
|
|
"loss": 0.6371,
|
||
|
|
"step": 161
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6558704453441295,
|
||
|
|
"grad_norm": 1.6134037960682746,
|
||
|
|
"learning_rate": 3.9884649120583903e-05,
|
||
|
|
"loss": 0.6339,
|
||
|
|
"step": 162
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.659919028340081,
|
||
|
|
"grad_norm": 0.9904786856875362,
|
||
|
|
"learning_rate": 3.987850437381915e-05,
|
||
|
|
"loss": 0.6167,
|
||
|
|
"step": 163
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6639676113360324,
|
||
|
|
"grad_norm": 1.8512232041435623,
|
||
|
|
"learning_rate": 3.987220067906313e-05,
|
||
|
|
"loss": 0.6282,
|
||
|
|
"step": 164
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6680161943319838,
|
||
|
|
"grad_norm": 1.8384581741287656,
|
||
|
|
"learning_rate": 3.986573808672002e-05,
|
||
|
|
"loss": 0.6395,
|
||
|
|
"step": 165
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6720647773279352,
|
||
|
|
"grad_norm": 0.8481826015397634,
|
||
|
|
"learning_rate": 3.9859116648464534e-05,
|
||
|
|
"loss": 0.6232,
|
||
|
|
"step": 166
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6761133603238867,
|
||
|
|
"grad_norm": 1.0936447673758132,
|
||
|
|
"learning_rate": 3.985233641724152e-05,
|
||
|
|
"loss": 0.6275,
|
||
|
|
"step": 167
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.680161943319838,
|
||
|
|
"grad_norm": 0.7977314365370172,
|
||
|
|
"learning_rate": 3.984539744726554e-05,
|
||
|
|
"loss": 0.6373,
|
||
|
|
"step": 168
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6842105263157895,
|
||
|
|
"grad_norm": 0.9136182673626502,
|
||
|
|
"learning_rate": 3.983829979402038e-05,
|
||
|
|
"loss": 0.6088,
|
||
|
|
"step": 169
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6882591093117408,
|
||
|
|
"grad_norm": 0.8314997365759503,
|
||
|
|
"learning_rate": 3.983104351425871e-05,
|
||
|
|
"loss": 0.6277,
|
||
|
|
"step": 170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6923076923076923,
|
||
|
|
"grad_norm": 1.044219320494823,
|
||
|
|
"learning_rate": 3.982362866600154e-05,
|
||
|
|
"loss": 0.6312,
|
||
|
|
"step": 171
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6963562753036437,
|
||
|
|
"grad_norm": 0.700018469077164,
|
||
|
|
"learning_rate": 3.9816055308537806e-05,
|
||
|
|
"loss": 0.6408,
|
||
|
|
"step": 172
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7004048582995951,
|
||
|
|
"grad_norm": 1.1693680118216576,
|
||
|
|
"learning_rate": 3.980832350242388e-05,
|
||
|
|
"loss": 0.6228,
|
||
|
|
"step": 173
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7044534412955465,
|
||
|
|
"grad_norm": 0.6002979966906169,
|
||
|
|
"learning_rate": 3.980043330948306e-05,
|
||
|
|
"loss": 0.6116,
|
||
|
|
"step": 174
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.708502024291498,
|
||
|
|
"grad_norm": 1.0373869288521915,
|
||
|
|
"learning_rate": 3.9792384792805124e-05,
|
||
|
|
"loss": 0.6232,
|
||
|
|
"step": 175
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7125506072874493,
|
||
|
|
"grad_norm": 0.6923956546670341,
|
||
|
|
"learning_rate": 3.978417801674579e-05,
|
||
|
|
"loss": 0.6192,
|
||
|
|
"step": 176
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7165991902834008,
|
||
|
|
"grad_norm": 0.7859532332776078,
|
||
|
|
"learning_rate": 3.9775813046926244e-05,
|
||
|
|
"loss": 0.6183,
|
||
|
|
"step": 177
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7206477732793523,
|
||
|
|
"grad_norm": 0.707810864869097,
|
||
|
|
"learning_rate": 3.976728995023254e-05,
|
||
|
|
"loss": 0.6072,
|
||
|
|
"step": 178
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7246963562753036,
|
||
|
|
"grad_norm": 0.620891261793413,
|
||
|
|
"learning_rate": 3.975860879481514e-05,
|
||
|
|
"loss": 0.6153,
|
||
|
|
"step": 179
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.728744939271255,
|
||
|
|
"grad_norm": 0.651372927484864,
|
||
|
|
"learning_rate": 3.974976965008833e-05,
|
||
|
|
"loss": 0.6368,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7327935222672065,
|
||
|
|
"grad_norm": 0.47855093913379376,
|
||
|
|
"learning_rate": 3.9740772586729674e-05,
|
||
|
|
"loss": 0.6107,
|
||
|
|
"step": 181
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7368421052631579,
|
||
|
|
"grad_norm": 0.731646773128479,
|
||
|
|
"learning_rate": 3.973161767667946e-05,
|
||
|
|
"loss": 0.6216,
|
||
|
|
"step": 182
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7408906882591093,
|
||
|
|
"grad_norm": 0.5982659171536929,
|
||
|
|
"learning_rate": 3.972230499314009e-05,
|
||
|
|
"loss": 0.6309,
|
||
|
|
"step": 183
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7449392712550608,
|
||
|
|
"grad_norm": 0.5151942725007864,
|
||
|
|
"learning_rate": 3.9712834610575545e-05,
|
||
|
|
"loss": 0.6123,
|
||
|
|
"step": 184
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7489878542510121,
|
||
|
|
"grad_norm": 0.5892170734133515,
|
||
|
|
"learning_rate": 3.9703206604710746e-05,
|
||
|
|
"loss": 0.6106,
|
||
|
|
"step": 185
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7530364372469636,
|
||
|
|
"grad_norm": 0.4537942669553407,
|
||
|
|
"learning_rate": 3.969342105253098e-05,
|
||
|
|
"loss": 0.6229,
|
||
|
|
"step": 186
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.757085020242915,
|
||
|
|
"grad_norm": 0.5616845843720286,
|
||
|
|
"learning_rate": 3.9683478032281246e-05,
|
||
|
|
"loss": 0.6015,
|
||
|
|
"step": 187
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7611336032388664,
|
||
|
|
"grad_norm": 0.4466693217963082,
|
||
|
|
"learning_rate": 3.967337762346568e-05,
|
||
|
|
"loss": 0.6124,
|
||
|
|
"step": 188
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7651821862348178,
|
||
|
|
"grad_norm": 0.5419606067309186,
|
||
|
|
"learning_rate": 3.9663119906846885e-05,
|
||
|
|
"loss": 0.6182,
|
||
|
|
"step": 189
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7692307692307693,
|
||
|
|
"grad_norm": 0.516175412259924,
|
||
|
|
"learning_rate": 3.965270496444528e-05,
|
||
|
|
"loss": 0.6229,
|
||
|
|
"step": 190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7732793522267206,
|
||
|
|
"grad_norm": 0.4996499636275476,
|
||
|
|
"learning_rate": 3.9642132879538465e-05,
|
||
|
|
"loss": 0.6182,
|
||
|
|
"step": 191
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7773279352226721,
|
||
|
|
"grad_norm": 0.46417953170161513,
|
||
|
|
"learning_rate": 3.9631403736660566e-05,
|
||
|
|
"loss": 0.6093,
|
||
|
|
"step": 192
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7813765182186235,
|
||
|
|
"grad_norm": 0.5342907677014899,
|
||
|
|
"learning_rate": 3.962051762160151e-05,
|
||
|
|
"loss": 0.6153,
|
||
|
|
"step": 193
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7854251012145749,
|
||
|
|
"grad_norm": 0.5001495397084043,
|
||
|
|
"learning_rate": 3.960947462140638e-05,
|
||
|
|
"loss": 0.6135,
|
||
|
|
"step": 194
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7894736842105263,
|
||
|
|
"grad_norm": 0.5729588664989923,
|
||
|
|
"learning_rate": 3.959827482437472e-05,
|
||
|
|
"loss": 0.616,
|
||
|
|
"step": 195
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7935222672064778,
|
||
|
|
"grad_norm": 0.4419488962650638,
|
||
|
|
"learning_rate": 3.958691832005981e-05,
|
||
|
|
"loss": 0.6284,
|
||
|
|
"step": 196
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7975708502024291,
|
||
|
|
"grad_norm": 0.4486218672121873,
|
||
|
|
"learning_rate": 3.9575405199267944e-05,
|
||
|
|
"loss": 0.6242,
|
||
|
|
"step": 197
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8016194331983806,
|
||
|
|
"grad_norm": 0.43348975514551746,
|
||
|
|
"learning_rate": 3.956373555405774e-05,
|
||
|
|
"loss": 0.617,
|
||
|
|
"step": 198
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.805668016194332,
|
||
|
|
"grad_norm": 0.48273858803181113,
|
||
|
|
"learning_rate": 3.955190947773937e-05,
|
||
|
|
"loss": 0.6241,
|
||
|
|
"step": 199
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8097165991902834,
|
||
|
|
"grad_norm": 0.4424487924793021,
|
||
|
|
"learning_rate": 3.9539927064873824e-05,
|
||
|
|
"loss": 0.6196,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8137651821862348,
|
||
|
|
"grad_norm": 0.5422773138347089,
|
||
|
|
"learning_rate": 3.952778841127214e-05,
|
||
|
|
"loss": 0.6354,
|
||
|
|
"step": 201
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8178137651821862,
|
||
|
|
"grad_norm": 0.42729713498861527,
|
||
|
|
"learning_rate": 3.95154936139947e-05,
|
||
|
|
"loss": 0.6195,
|
||
|
|
"step": 202
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8218623481781376,
|
||
|
|
"grad_norm": 0.5091026290897528,
|
||
|
|
"learning_rate": 3.950304277135034e-05,
|
||
|
|
"loss": 0.6086,
|
||
|
|
"step": 203
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8259109311740891,
|
||
|
|
"grad_norm": 0.48614170839755616,
|
||
|
|
"learning_rate": 3.94904359828957e-05,
|
||
|
|
"loss": 0.6229,
|
||
|
|
"step": 204
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8299595141700404,
|
||
|
|
"grad_norm": 0.42978490241726597,
|
||
|
|
"learning_rate": 3.94776733494343e-05,
|
||
|
|
"loss": 0.6165,
|
||
|
|
"step": 205
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8340080971659919,
|
||
|
|
"grad_norm": 0.4414966377823553,
|
||
|
|
"learning_rate": 3.9464754973015826e-05,
|
||
|
|
"loss": 0.6148,
|
||
|
|
"step": 206
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8380566801619433,
|
||
|
|
"grad_norm": 0.42144526997964776,
|
||
|
|
"learning_rate": 3.9451680956935275e-05,
|
||
|
|
"loss": 0.6282,
|
||
|
|
"step": 207
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8421052631578947,
|
||
|
|
"grad_norm": 0.4971978379430151,
|
||
|
|
"learning_rate": 3.943845140573213e-05,
|
||
|
|
"loss": 0.6183,
|
||
|
|
"step": 208
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8461538461538461,
|
||
|
|
"grad_norm": 0.4397336850109413,
|
||
|
|
"learning_rate": 3.942506642518952e-05,
|
||
|
|
"loss": 0.6182,
|
||
|
|
"step": 209
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8502024291497976,
|
||
|
|
"grad_norm": 0.4259875817327147,
|
||
|
|
"learning_rate": 3.9411526122333396e-05,
|
||
|
|
"loss": 0.6043,
|
||
|
|
"step": 210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.854251012145749,
|
||
|
|
"grad_norm": 0.5635474056854293,
|
||
|
|
"learning_rate": 3.939783060543167e-05,
|
||
|
|
"loss": 0.6301,
|
||
|
|
"step": 211
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8582995951417004,
|
||
|
|
"grad_norm": 0.5603256300857928,
|
||
|
|
"learning_rate": 3.938397998399333e-05,
|
||
|
|
"loss": 0.6092,
|
||
|
|
"step": 212
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8623481781376519,
|
||
|
|
"grad_norm": 0.44306757287931325,
|
||
|
|
"learning_rate": 3.936997436876756e-05,
|
||
|
|
"loss": 0.609,
|
||
|
|
"step": 213
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8663967611336032,
|
||
|
|
"grad_norm": 0.4364476799153564,
|
||
|
|
"learning_rate": 3.935581387174289e-05,
|
||
|
|
"loss": 0.6085,
|
||
|
|
"step": 214
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8704453441295547,
|
||
|
|
"grad_norm": 0.5045660091969703,
|
||
|
|
"learning_rate": 3.9341498606146295e-05,
|
||
|
|
"loss": 0.6242,
|
||
|
|
"step": 215
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8744939271255061,
|
||
|
|
"grad_norm": 0.5647741955357021,
|
||
|
|
"learning_rate": 3.932702868644224e-05,
|
||
|
|
"loss": 0.6042,
|
||
|
|
"step": 216
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8785425101214575,
|
||
|
|
"grad_norm": 0.4570840163078956,
|
||
|
|
"learning_rate": 3.931240422833181e-05,
|
||
|
|
"loss": 0.6344,
|
||
|
|
"step": 217
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8825910931174089,
|
||
|
|
"grad_norm": 0.5370484336906498,
|
||
|
|
"learning_rate": 3.9297625348751806e-05,
|
||
|
|
"loss": 0.6313,
|
||
|
|
"step": 218
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8866396761133604,
|
||
|
|
"grad_norm": 0.713468321749162,
|
||
|
|
"learning_rate": 3.9282692165873744e-05,
|
||
|
|
"loss": 0.6287,
|
||
|
|
"step": 219
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8906882591093117,
|
||
|
|
"grad_norm": 0.7233019925145787,
|
||
|
|
"learning_rate": 3.9267604799102945e-05,
|
||
|
|
"loss": 0.62,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8947368421052632,
|
||
|
|
"grad_norm": 0.5684226820308063,
|
||
|
|
"learning_rate": 3.9252363369077615e-05,
|
||
|
|
"loss": 0.6062,
|
||
|
|
"step": 221
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8987854251012146,
|
||
|
|
"grad_norm": 0.47880428862025937,
|
||
|
|
"learning_rate": 3.9236967997667796e-05,
|
||
|
|
"loss": 0.6211,
|
||
|
|
"step": 222
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.902834008097166,
|
||
|
|
"grad_norm": 0.7613900856575858,
|
||
|
|
"learning_rate": 3.9221418807974496e-05,
|
||
|
|
"loss": 0.6253,
|
||
|
|
"step": 223
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9068825910931174,
|
||
|
|
"grad_norm": 0.8939981732774428,
|
||
|
|
"learning_rate": 3.92057159243286e-05,
|
||
|
|
"loss": 0.5996,
|
||
|
|
"step": 224
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9109311740890689,
|
||
|
|
"grad_norm": 0.6481498236160634,
|
||
|
|
"learning_rate": 3.9189859472289956e-05,
|
||
|
|
"loss": 0.615,
|
||
|
|
"step": 225
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9149797570850202,
|
||
|
|
"grad_norm": 0.4572178581636744,
|
||
|
|
"learning_rate": 3.9173849578646324e-05,
|
||
|
|
"loss": 0.6171,
|
||
|
|
"step": 226
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9190283400809717,
|
||
|
|
"grad_norm": 0.6300277681213433,
|
||
|
|
"learning_rate": 3.9157686371412395e-05,
|
||
|
|
"loss": 0.6078,
|
||
|
|
"step": 227
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9230769230769231,
|
||
|
|
"grad_norm": 0.5861102168625972,
|
||
|
|
"learning_rate": 3.914136997982874e-05,
|
||
|
|
"loss": 0.5878,
|
||
|
|
"step": 228
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9271255060728745,
|
||
|
|
"grad_norm": 0.6130034914826623,
|
||
|
|
"learning_rate": 3.912490053436079e-05,
|
||
|
|
"loss": 0.6049,
|
||
|
|
"step": 229
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9311740890688259,
|
||
|
|
"grad_norm": 0.49515242299378187,
|
||
|
|
"learning_rate": 3.9108278166697805e-05,
|
||
|
|
"loss": 0.6119,
|
||
|
|
"step": 230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9352226720647774,
|
||
|
|
"grad_norm": 0.597966386055594,
|
||
|
|
"learning_rate": 3.909150300975178e-05,
|
||
|
|
"loss": 0.5956,
|
||
|
|
"step": 231
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9392712550607287,
|
||
|
|
"grad_norm": 0.55660854181991,
|
||
|
|
"learning_rate": 3.907457519765642e-05,
|
||
|
|
"loss": 0.6237,
|
||
|
|
"step": 232
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9433198380566802,
|
||
|
|
"grad_norm": 0.5553598038284859,
|
||
|
|
"learning_rate": 3.905749486576607e-05,
|
||
|
|
"loss": 0.6038,
|
||
|
|
"step": 233
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9473684210526315,
|
||
|
|
"grad_norm": 0.5471646390468593,
|
||
|
|
"learning_rate": 3.90402621506546e-05,
|
||
|
|
"loss": 0.6098,
|
||
|
|
"step": 234
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.951417004048583,
|
||
|
|
"grad_norm": 0.5513647965488957,
|
||
|
|
"learning_rate": 3.902287719011435e-05,
|
||
|
|
"loss": 0.6132,
|
||
|
|
"step": 235
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9554655870445344,
|
||
|
|
"grad_norm": 0.5875164911765222,
|
||
|
|
"learning_rate": 3.900534012315499e-05,
|
||
|
|
"loss": 0.6143,
|
||
|
|
"step": 236
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9595141700404858,
|
||
|
|
"grad_norm": 0.6352542152027957,
|
||
|
|
"learning_rate": 3.8987651090002454e-05,
|
||
|
|
"loss": 0.6078,
|
||
|
|
"step": 237
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9635627530364372,
|
||
|
|
"grad_norm": 0.5679110920722806,
|
||
|
|
"learning_rate": 3.896981023209777e-05,
|
||
|
|
"loss": 0.626,
|
||
|
|
"step": 238
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9676113360323887,
|
||
|
|
"grad_norm": 0.7078123352220793,
|
||
|
|
"learning_rate": 3.8951817692095966e-05,
|
||
|
|
"loss": 0.6116,
|
||
|
|
"step": 239
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.97165991902834,
|
||
|
|
"grad_norm": 0.481187098076771,
|
||
|
|
"learning_rate": 3.893367361386491e-05,
|
||
|
|
"loss": 0.6156,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9757085020242915,
|
||
|
|
"grad_norm": 0.5952843121024332,
|
||
|
|
"learning_rate": 3.891537814248417e-05,
|
||
|
|
"loss": 0.6068,
|
||
|
|
"step": 241
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.979757085020243,
|
||
|
|
"grad_norm": 0.5998410954147507,
|
||
|
|
"learning_rate": 3.889693142424384e-05,
|
||
|
|
"loss": 0.6081,
|
||
|
|
"step": 242
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9838056680161943,
|
||
|
|
"grad_norm": 0.5469912261406582,
|
||
|
|
"learning_rate": 3.8878333606643405e-05,
|
||
|
|
"loss": 0.6124,
|
||
|
|
"step": 243
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9878542510121457,
|
||
|
|
"grad_norm": 0.4684319128726047,
|
||
|
|
"learning_rate": 3.885958483839049e-05,
|
||
|
|
"loss": 0.6111,
|
||
|
|
"step": 244
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9919028340080972,
|
||
|
|
"grad_norm": 0.455722128318724,
|
||
|
|
"learning_rate": 3.8840685269399786e-05,
|
||
|
|
"loss": 0.6314,
|
||
|
|
"step": 245
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9959514170040485,
|
||
|
|
"grad_norm": 0.5133103364417629,
|
||
|
|
"learning_rate": 3.882163505079171e-05,
|
||
|
|
"loss": 0.6119,
|
||
|
|
"step": 246
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"grad_norm": 0.5734870359919972,
|
||
|
|
"learning_rate": 3.880243433489132e-05,
|
||
|
|
"loss": 0.5973,
|
||
|
|
"step": 247
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0040485829959513,
|
||
|
|
"grad_norm": 0.540210858366143,
|
||
|
|
"learning_rate": 3.878308327522702e-05,
|
||
|
|
"loss": 0.5576,
|
||
|
|
"step": 248
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.008097165991903,
|
||
|
|
"grad_norm": 0.5227357169726952,
|
||
|
|
"learning_rate": 3.876358202652936e-05,
|
||
|
|
"loss": 0.5447,
|
||
|
|
"step": 249
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0121457489878543,
|
||
|
|
"grad_norm": 0.7520944057826915,
|
||
|
|
"learning_rate": 3.8743930744729835e-05,
|
||
|
|
"loss": 0.5751,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0161943319838056,
|
||
|
|
"grad_norm": 0.540377666500093,
|
||
|
|
"learning_rate": 3.8724129586959535e-05,
|
||
|
|
"loss": 0.5575,
|
||
|
|
"step": 251
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0202429149797572,
|
||
|
|
"grad_norm": 0.607476450642687,
|
||
|
|
"learning_rate": 3.870417871154801e-05,
|
||
|
|
"loss": 0.549,
|
||
|
|
"step": 252
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0242914979757085,
|
||
|
|
"grad_norm": 0.5392363836243653,
|
||
|
|
"learning_rate": 3.868407827802192e-05,
|
||
|
|
"loss": 0.5483,
|
||
|
|
"step": 253
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0283400809716599,
|
||
|
|
"grad_norm": 0.5395653224036908,
|
||
|
|
"learning_rate": 3.8663828447103804e-05,
|
||
|
|
"loss": 0.5558,
|
||
|
|
"step": 254
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0323886639676114,
|
||
|
|
"grad_norm": 0.6204120058255576,
|
||
|
|
"learning_rate": 3.8643429380710765e-05,
|
||
|
|
"loss": 0.5571,
|
||
|
|
"step": 255
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0364372469635628,
|
||
|
|
"grad_norm": 0.48328284319753095,
|
||
|
|
"learning_rate": 3.862288124195319e-05,
|
||
|
|
"loss": 0.5507,
|
||
|
|
"step": 256
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.040485829959514,
|
||
|
|
"grad_norm": 0.5189758746634813,
|
||
|
|
"learning_rate": 3.8602184195133464e-05,
|
||
|
|
"loss": 0.56,
|
||
|
|
"step": 257
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0445344129554657,
|
||
|
|
"grad_norm": 0.5111870246043166,
|
||
|
|
"learning_rate": 3.8581338405744604e-05,
|
||
|
|
"loss": 0.5328,
|
||
|
|
"step": 258
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.048582995951417,
|
||
|
|
"grad_norm": 0.5080575325509256,
|
||
|
|
"learning_rate": 3.8560344040468996e-05,
|
||
|
|
"loss": 0.5478,
|
||
|
|
"step": 259
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0526315789473684,
|
||
|
|
"grad_norm": 0.5095334983196335,
|
||
|
|
"learning_rate": 3.853920126717703e-05,
|
||
|
|
"loss": 0.5457,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.05668016194332,
|
||
|
|
"grad_norm": 0.507676212117483,
|
||
|
|
"learning_rate": 3.851791025492574e-05,
|
||
|
|
"loss": 0.5674,
|
||
|
|
"step": 261
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0607287449392713,
|
||
|
|
"grad_norm": 0.5191765008859061,
|
||
|
|
"learning_rate": 3.8496471173957515e-05,
|
||
|
|
"loss": 0.5565,
|
||
|
|
"step": 262
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0647773279352226,
|
||
|
|
"grad_norm": 0.4282035210013262,
|
||
|
|
"learning_rate": 3.847488419569866e-05,
|
||
|
|
"loss": 0.5518,
|
||
|
|
"step": 263
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0688259109311742,
|
||
|
|
"grad_norm": 0.4889287480748198,
|
||
|
|
"learning_rate": 3.8453149492758077e-05,
|
||
|
|
"loss": 0.5296,
|
||
|
|
"step": 264
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0728744939271255,
|
||
|
|
"grad_norm": 0.47003278405863397,
|
||
|
|
"learning_rate": 3.8431267238925886e-05,
|
||
|
|
"loss": 0.553,
|
||
|
|
"step": 265
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0769230769230769,
|
||
|
|
"grad_norm": 0.5036742644253677,
|
||
|
|
"learning_rate": 3.840923760917199e-05,
|
||
|
|
"loss": 0.546,
|
||
|
|
"step": 266
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0809716599190284,
|
||
|
|
"grad_norm": 0.6419475051987283,
|
||
|
|
"learning_rate": 3.838706077964473e-05,
|
||
|
|
"loss": 0.5558,
|
||
|
|
"step": 267
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0850202429149798,
|
||
|
|
"grad_norm": 0.4399528869308145,
|
||
|
|
"learning_rate": 3.836473692766944e-05,
|
||
|
|
"loss": 0.5435,
|
||
|
|
"step": 268
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0890688259109311,
|
||
|
|
"grad_norm": 0.4231051807662026,
|
||
|
|
"learning_rate": 3.834226623174705e-05,
|
||
|
|
"loss": 0.5594,
|
||
|
|
"step": 269
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0931174089068827,
|
||
|
|
"grad_norm": 0.5069959107564458,
|
||
|
|
"learning_rate": 3.831964887155264e-05,
|
||
|
|
"loss": 0.55,
|
||
|
|
"step": 270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.097165991902834,
|
||
|
|
"grad_norm": 0.4453717423833374,
|
||
|
|
"learning_rate": 3.829688502793403e-05,
|
||
|
|
"loss": 0.5395,
|
||
|
|
"step": 271
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1012145748987854,
|
||
|
|
"grad_norm": 0.6655348812599078,
|
||
|
|
"learning_rate": 3.827397488291029e-05,
|
||
|
|
"loss": 0.5651,
|
||
|
|
"step": 272
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1052631578947367,
|
||
|
|
"grad_norm": 0.5010083669051043,
|
||
|
|
"learning_rate": 3.8250918619670343e-05,
|
||
|
|
"loss": 0.5547,
|
||
|
|
"step": 273
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1093117408906883,
|
||
|
|
"grad_norm": 0.519276120776747,
|
||
|
|
"learning_rate": 3.822771642257145e-05,
|
||
|
|
"loss": 0.5585,
|
||
|
|
"step": 274
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1133603238866396,
|
||
|
|
"grad_norm": 0.5029795532137715,
|
||
|
|
"learning_rate": 3.820436847713776e-05,
|
||
|
|
"loss": 0.5588,
|
||
|
|
"step": 275
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.117408906882591,
|
||
|
|
"grad_norm": 0.5002262377924468,
|
||
|
|
"learning_rate": 3.8180874970058826e-05,
|
||
|
|
"loss": 0.5429,
|
||
|
|
"step": 276
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1214574898785425,
|
||
|
|
"grad_norm": 0.5680291049643177,
|
||
|
|
"learning_rate": 3.8157236089188103e-05,
|
||
|
|
"loss": 0.5617,
|
||
|
|
"step": 277
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.125506072874494,
|
||
|
|
"grad_norm": 0.4702917372790098,
|
||
|
|
"learning_rate": 3.813345202354145e-05,
|
||
|
|
"loss": 0.5435,
|
||
|
|
"step": 278
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1295546558704452,
|
||
|
|
"grad_norm": 0.5847386721530332,
|
||
|
|
"learning_rate": 3.810952296329563e-05,
|
||
|
|
"loss": 0.5556,
|
||
|
|
"step": 279
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1336032388663968,
|
||
|
|
"grad_norm": 0.4393713600650091,
|
||
|
|
"learning_rate": 3.808544909978676e-05,
|
||
|
|
"loss": 0.5508,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1376518218623481,
|
||
|
|
"grad_norm": 0.5808942464909159,
|
||
|
|
"learning_rate": 3.806123062550882e-05,
|
||
|
|
"loss": 0.5498,
|
||
|
|
"step": 281
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1417004048582995,
|
||
|
|
"grad_norm": 0.4993187940081679,
|
||
|
|
"learning_rate": 3.803686773411208e-05,
|
||
|
|
"loss": 0.5475,
|
||
|
|
"step": 282
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.145748987854251,
|
||
|
|
"grad_norm": 0.5509555642971127,
|
||
|
|
"learning_rate": 3.801236062040158e-05,
|
||
|
|
"loss": 0.5521,
|
||
|
|
"step": 283
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1497975708502024,
|
||
|
|
"grad_norm": 0.6525608738472736,
|
||
|
|
"learning_rate": 3.798770948033553e-05,
|
||
|
|
"loss": 0.5354,
|
||
|
|
"step": 284
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1538461538461537,
|
||
|
|
"grad_norm": 0.5127217866220706,
|
||
|
|
"learning_rate": 3.796291451102381e-05,
|
||
|
|
"loss": 0.5496,
|
||
|
|
"step": 285
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1578947368421053,
|
||
|
|
"grad_norm": 0.45509269290060966,
|
||
|
|
"learning_rate": 3.793797591072631e-05,
|
||
|
|
"loss": 0.5485,
|
||
|
|
"step": 286
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1619433198380567,
|
||
|
|
"grad_norm": 0.4826078903938804,
|
||
|
|
"learning_rate": 3.791289387885144e-05,
|
||
|
|
"loss": 0.5614,
|
||
|
|
"step": 287
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.165991902834008,
|
||
|
|
"grad_norm": 0.6419715492761251,
|
||
|
|
"learning_rate": 3.788766861595443e-05,
|
||
|
|
"loss": 0.5398,
|
||
|
|
"step": 288
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1700404858299596,
|
||
|
|
"grad_norm": 0.6379950497762314,
|
||
|
|
"learning_rate": 3.7862300323735835e-05,
|
||
|
|
"loss": 0.5482,
|
||
|
|
"step": 289
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.174089068825911,
|
||
|
|
"grad_norm": 0.5062083950047841,
|
||
|
|
"learning_rate": 3.783678920503983e-05,
|
||
|
|
"loss": 0.5643,
|
||
|
|
"step": 290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1781376518218623,
|
||
|
|
"grad_norm": 0.6625968553598032,
|
||
|
|
"learning_rate": 3.781113546385265e-05,
|
||
|
|
"loss": 0.5786,
|
||
|
|
"step": 291
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1821862348178138,
|
||
|
|
"grad_norm": 0.4994277022519586,
|
||
|
|
"learning_rate": 3.7785339305300916e-05,
|
||
|
|
"loss": 0.5533,
|
||
|
|
"step": 292
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1862348178137652,
|
||
|
|
"grad_norm": 0.4819023491870224,
|
||
|
|
"learning_rate": 3.7759400935650025e-05,
|
||
|
|
"loss": 0.553,
|
||
|
|
"step": 293
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1902834008097165,
|
||
|
|
"grad_norm": 0.5441594571533597,
|
||
|
|
"learning_rate": 3.77333205623025e-05,
|
||
|
|
"loss": 0.5337,
|
||
|
|
"step": 294
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.194331983805668,
|
||
|
|
"grad_norm": 0.4895437341365659,
|
||
|
|
"learning_rate": 3.7707098393796295e-05,
|
||
|
|
"loss": 0.5469,
|
||
|
|
"step": 295
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1983805668016194,
|
||
|
|
"grad_norm": 0.4865308090361683,
|
||
|
|
"learning_rate": 3.7680734639803185e-05,
|
||
|
|
"loss": 0.5497,
|
||
|
|
"step": 296
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2024291497975708,
|
||
|
|
"grad_norm": 0.5864893603557435,
|
||
|
|
"learning_rate": 3.765422951112704e-05,
|
||
|
|
"loss": 0.5628,
|
||
|
|
"step": 297
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2064777327935223,
|
||
|
|
"grad_norm": 0.591303872390734,
|
||
|
|
"learning_rate": 3.762758321970216e-05,
|
||
|
|
"loss": 0.5549,
|
||
|
|
"step": 298
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2105263157894737,
|
||
|
|
"grad_norm": 0.53310021325941,
|
||
|
|
"learning_rate": 3.7600795978591584e-05,
|
||
|
|
"loss": 0.5493,
|
||
|
|
"step": 299
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.214574898785425,
|
||
|
|
"grad_norm": 0.5224984563250727,
|
||
|
|
"learning_rate": 3.757386800198538e-05,
|
||
|
|
"loss": 0.562,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2186234817813766,
|
||
|
|
"grad_norm": 0.5507639513338117,
|
||
|
|
"learning_rate": 3.7546799505198925e-05,
|
||
|
|
"loss": 0.5597,
|
||
|
|
"step": 301
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.222672064777328,
|
||
|
|
"grad_norm": 0.4517072122304285,
|
||
|
|
"learning_rate": 3.7519590704671197e-05,
|
||
|
|
"loss": 0.5524,
|
||
|
|
"step": 302
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2267206477732793,
|
||
|
|
"grad_norm": 0.5586001959862958,
|
||
|
|
"learning_rate": 3.749224181796305e-05,
|
||
|
|
"loss": 0.5554,
|
||
|
|
"step": 303
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2307692307692308,
|
||
|
|
"grad_norm": 0.5335778652193891,
|
||
|
|
"learning_rate": 3.746475306375545e-05,
|
||
|
|
"loss": 0.562,
|
||
|
|
"step": 304
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2348178137651822,
|
||
|
|
"grad_norm": 0.611563231236552,
|
||
|
|
"learning_rate": 3.743712466184774e-05,
|
||
|
|
"loss": 0.5567,
|
||
|
|
"step": 305
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2388663967611335,
|
||
|
|
"grad_norm": 0.44935695850558155,
|
||
|
|
"learning_rate": 3.7409356833155885e-05,
|
||
|
|
"loss": 0.5658,
|
||
|
|
"step": 306
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.242914979757085,
|
||
|
|
"grad_norm": 0.6270737796700282,
|
||
|
|
"learning_rate": 3.73814497997107e-05,
|
||
|
|
"loss": 0.5548,
|
||
|
|
"step": 307
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2469635627530364,
|
||
|
|
"grad_norm": 0.7223834929063898,
|
||
|
|
"learning_rate": 3.7353403784656095e-05,
|
||
|
|
"loss": 0.5536,
|
||
|
|
"step": 308
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2510121457489878,
|
||
|
|
"grad_norm": 0.6307477544218978,
|
||
|
|
"learning_rate": 3.732521901224724e-05,
|
||
|
|
"loss": 0.549,
|
||
|
|
"step": 309
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2550607287449393,
|
||
|
|
"grad_norm": 0.721289777737997,
|
||
|
|
"learning_rate": 3.7296895707848844e-05,
|
||
|
|
"loss": 0.5731,
|
||
|
|
"step": 310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2591093117408907,
|
||
|
|
"grad_norm": 0.5434580430976932,
|
||
|
|
"learning_rate": 3.7268434097933275e-05,
|
||
|
|
"loss": 0.546,
|
||
|
|
"step": 311
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.263157894736842,
|
||
|
|
"grad_norm": 0.5682339502589898,
|
||
|
|
"learning_rate": 3.723983441007882e-05,
|
||
|
|
"loss": 0.5437,
|
||
|
|
"step": 312
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2672064777327936,
|
||
|
|
"grad_norm": 0.49879264820585584,
|
||
|
|
"learning_rate": 3.721109687296781e-05,
|
||
|
|
"loss": 0.5522,
|
||
|
|
"step": 313
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.271255060728745,
|
||
|
|
"grad_norm": 0.42957942474280564,
|
||
|
|
"learning_rate": 3.718222171638484e-05,
|
||
|
|
"loss": 0.5627,
|
||
|
|
"step": 314
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2753036437246963,
|
||
|
|
"grad_norm": 0.43551789799851554,
|
||
|
|
"learning_rate": 3.715320917121488e-05,
|
||
|
|
"loss": 0.5508,
|
||
|
|
"step": 315
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2793522267206479,
|
||
|
|
"grad_norm": 0.4169534808214485,
|
||
|
|
"learning_rate": 3.712405946944149e-05,
|
||
|
|
"loss": 0.5627,
|
||
|
|
"step": 316
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2834008097165992,
|
||
|
|
"grad_norm": 0.5227318569046997,
|
||
|
|
"learning_rate": 3.709477284414488e-05,
|
||
|
|
"loss": 0.5493,
|
||
|
|
"step": 317
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2874493927125505,
|
||
|
|
"grad_norm": 0.4375006421267927,
|
||
|
|
"learning_rate": 3.7065349529500136e-05,
|
||
|
|
"loss": 0.5511,
|
||
|
|
"step": 318
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.291497975708502,
|
||
|
|
"grad_norm": 0.5327518828699664,
|
||
|
|
"learning_rate": 3.7035789760775305e-05,
|
||
|
|
"loss": 0.5548,
|
||
|
|
"step": 319
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2955465587044535,
|
||
|
|
"grad_norm": 0.4793560806529308,
|
||
|
|
"learning_rate": 3.70060937743295e-05,
|
||
|
|
"loss": 0.5613,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2995951417004048,
|
||
|
|
"grad_norm": 0.4422973110007042,
|
||
|
|
"learning_rate": 3.6976261807611045e-05,
|
||
|
|
"loss": 0.5589,
|
||
|
|
"step": 321
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3036437246963564,
|
||
|
|
"grad_norm": 0.5255791899941813,
|
||
|
|
"learning_rate": 3.694629409915555e-05,
|
||
|
|
"loss": 0.5485,
|
||
|
|
"step": 322
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3076923076923077,
|
||
|
|
"grad_norm": 0.47677190019246984,
|
||
|
|
"learning_rate": 3.691619088858401e-05,
|
||
|
|
"loss": 0.5454,
|
||
|
|
"step": 323
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.311740890688259,
|
||
|
|
"grad_norm": 0.44723231231376087,
|
||
|
|
"learning_rate": 3.68859524166009e-05,
|
||
|
|
"loss": 0.5529,
|
||
|
|
"step": 324
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3157894736842106,
|
||
|
|
"grad_norm": 0.5169572952108423,
|
||
|
|
"learning_rate": 3.6855578924992226e-05,
|
||
|
|
"loss": 0.5475,
|
||
|
|
"step": 325
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.319838056680162,
|
||
|
|
"grad_norm": 0.6009692364899125,
|
||
|
|
"learning_rate": 3.6825070656623626e-05,
|
||
|
|
"loss": 0.5486,
|
||
|
|
"step": 326
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3238866396761133,
|
||
|
|
"grad_norm": 0.42519216712380825,
|
||
|
|
"learning_rate": 3.67944278554384e-05,
|
||
|
|
"loss": 0.5589,
|
||
|
|
"step": 327
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3279352226720649,
|
||
|
|
"grad_norm": 0.4811007271686678,
|
||
|
|
"learning_rate": 3.676365076645557e-05,
|
||
|
|
"loss": 0.5496,
|
||
|
|
"step": 328
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3319838056680162,
|
||
|
|
"grad_norm": 0.6209529255724376,
|
||
|
|
"learning_rate": 3.673273963576791e-05,
|
||
|
|
"loss": 0.5515,
|
||
|
|
"step": 329
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3360323886639676,
|
||
|
|
"grad_norm": 0.4700061665762509,
|
||
|
|
"learning_rate": 3.670169471054001e-05,
|
||
|
|
"loss": 0.5575,
|
||
|
|
"step": 330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3400809716599191,
|
||
|
|
"grad_norm": 0.5037104146644436,
|
||
|
|
"learning_rate": 3.6670516239006254e-05,
|
||
|
|
"loss": 0.5674,
|
||
|
|
"step": 331
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3441295546558705,
|
||
|
|
"grad_norm": 0.6790236698248087,
|
||
|
|
"learning_rate": 3.663920447046888e-05,
|
||
|
|
"loss": 0.5472,
|
||
|
|
"step": 332
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3481781376518218,
|
||
|
|
"grad_norm": 0.5424416726637894,
|
||
|
|
"learning_rate": 3.6607759655295954e-05,
|
||
|
|
"loss": 0.5639,
|
||
|
|
"step": 333
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3522267206477734,
|
||
|
|
"grad_norm": 0.6090700756392909,
|
||
|
|
"learning_rate": 3.657618204491937e-05,
|
||
|
|
"loss": 0.5476,
|
||
|
|
"step": 334
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3562753036437247,
|
||
|
|
"grad_norm": 0.5451550927564242,
|
||
|
|
"learning_rate": 3.6544471891832875e-05,
|
||
|
|
"loss": 0.5648,
|
||
|
|
"step": 335
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.360323886639676,
|
||
|
|
"grad_norm": 0.48567533213127456,
|
||
|
|
"learning_rate": 3.6512629449589983e-05,
|
||
|
|
"loss": 0.5646,
|
||
|
|
"step": 336
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3643724696356276,
|
||
|
|
"grad_norm": 0.43598024622999454,
|
||
|
|
"learning_rate": 3.6480654972802045e-05,
|
||
|
|
"loss": 0.5732,
|
||
|
|
"step": 337
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.368421052631579,
|
||
|
|
"grad_norm": 0.5054750326986709,
|
||
|
|
"learning_rate": 3.644854871713611e-05,
|
||
|
|
"loss": 0.5765,
|
||
|
|
"step": 338
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3724696356275303,
|
||
|
|
"grad_norm": 0.42163280134862907,
|
||
|
|
"learning_rate": 3.6416310939312924e-05,
|
||
|
|
"loss": 0.5645,
|
||
|
|
"step": 339
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.376518218623482,
|
||
|
|
"grad_norm": 0.47045530112085526,
|
||
|
|
"learning_rate": 3.638394189710493e-05,
|
||
|
|
"loss": 0.552,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3805668016194332,
|
||
|
|
"grad_norm": 0.5550586320534833,
|
||
|
|
"learning_rate": 3.635144184933412e-05,
|
||
|
|
"loss": 0.5495,
|
||
|
|
"step": 341
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3846153846153846,
|
||
|
|
"grad_norm": 0.5780287615962894,
|
||
|
|
"learning_rate": 3.6318811055869994e-05,
|
||
|
|
"loss": 0.5488,
|
||
|
|
"step": 342
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3886639676113361,
|
||
|
|
"grad_norm": 0.4774294949658317,
|
||
|
|
"learning_rate": 3.628604977762752e-05,
|
||
|
|
"loss": 0.5688,
|
||
|
|
"step": 343
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3927125506072875,
|
||
|
|
"grad_norm": 0.41444601369997114,
|
||
|
|
"learning_rate": 3.6253158276565004e-05,
|
||
|
|
"loss": 0.5455,
|
||
|
|
"step": 344
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3967611336032388,
|
||
|
|
"grad_norm": 0.5013850214901393,
|
||
|
|
"learning_rate": 3.6220136815682016e-05,
|
||
|
|
"loss": 0.5696,
|
||
|
|
"step": 345
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4008097165991904,
|
||
|
|
"grad_norm": 0.3679283745536597,
|
||
|
|
"learning_rate": 3.618698565901727e-05,
|
||
|
|
"loss": 0.5619,
|
||
|
|
"step": 346
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4048582995951417,
|
||
|
|
"grad_norm": 0.3897272243538813,
|
||
|
|
"learning_rate": 3.615370507164654e-05,
|
||
|
|
"loss": 0.5493,
|
||
|
|
"step": 347
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.408906882591093,
|
||
|
|
"grad_norm": 0.3832925323325651,
|
||
|
|
"learning_rate": 3.61202953196805e-05,
|
||
|
|
"loss": 0.5515,
|
||
|
|
"step": 348
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4129554655870447,
|
||
|
|
"grad_norm": 0.46367799002036797,
|
||
|
|
"learning_rate": 3.6086756670262654e-05,
|
||
|
|
"loss": 0.5461,
|
||
|
|
"step": 349
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.417004048582996,
|
||
|
|
"grad_norm": 0.39284924559677564,
|
||
|
|
"learning_rate": 3.6053089391567135e-05,
|
||
|
|
"loss": 0.5476,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4210526315789473,
|
||
|
|
"grad_norm": 0.4952992287311946,
|
||
|
|
"learning_rate": 3.601929375279662e-05,
|
||
|
|
"loss": 0.559,
|
||
|
|
"step": 351
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.425101214574899,
|
||
|
|
"grad_norm": 0.4610452474949315,
|
||
|
|
"learning_rate": 3.598537002418012e-05,
|
||
|
|
"loss": 0.5369,
|
||
|
|
"step": 352
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4291497975708503,
|
||
|
|
"grad_norm": 0.46389667312705835,
|
||
|
|
"learning_rate": 3.5951318476970885e-05,
|
||
|
|
"loss": 0.546,
|
||
|
|
"step": 353
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4331983805668016,
|
||
|
|
"grad_norm": 0.4431678938980406,
|
||
|
|
"learning_rate": 3.591713938344416e-05,
|
||
|
|
"loss": 0.5291,
|
||
|
|
"step": 354
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4372469635627532,
|
||
|
|
"grad_norm": 0.4831802326494412,
|
||
|
|
"learning_rate": 3.588283301689507e-05,
|
||
|
|
"loss": 0.5495,
|
||
|
|
"step": 355
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4412955465587045,
|
||
|
|
"grad_norm": 0.5268832881017466,
|
||
|
|
"learning_rate": 3.5848399651636424e-05,
|
||
|
|
"loss": 0.5442,
|
||
|
|
"step": 356
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4453441295546559,
|
||
|
|
"grad_norm": 0.38585200877783915,
|
||
|
|
"learning_rate": 3.5813839562996473e-05,
|
||
|
|
"loss": 0.545,
|
||
|
|
"step": 357
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4493927125506074,
|
||
|
|
"grad_norm": 0.46686962852077835,
|
||
|
|
"learning_rate": 3.5779153027316775e-05,
|
||
|
|
"loss": 0.5459,
|
||
|
|
"step": 358
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4534412955465588,
|
||
|
|
"grad_norm": 0.5341180163010323,
|
||
|
|
"learning_rate": 3.574434032194994e-05,
|
||
|
|
"loss": 0.5578,
|
||
|
|
"step": 359
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.45748987854251,
|
||
|
|
"grad_norm": 0.4801419466950967,
|
||
|
|
"learning_rate": 3.5709401725257436e-05,
|
||
|
|
"loss": 0.5578,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4615384615384617,
|
||
|
|
"grad_norm": 0.49864545695542234,
|
||
|
|
"learning_rate": 3.5674337516607346e-05,
|
||
|
|
"loss": 0.5506,
|
||
|
|
"step": 361
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.465587044534413,
|
||
|
|
"grad_norm": 0.48229097959718886,
|
||
|
|
"learning_rate": 3.563914797637216e-05,
|
||
|
|
"loss": 0.5545,
|
||
|
|
"step": 362
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4696356275303644,
|
||
|
|
"grad_norm": 0.590027701630367,
|
||
|
|
"learning_rate": 3.560383338592649e-05,
|
||
|
|
"loss": 0.5516,
|
||
|
|
"step": 363
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4736842105263157,
|
||
|
|
"grad_norm": 0.3973589361909056,
|
||
|
|
"learning_rate": 3.556839402764487e-05,
|
||
|
|
"loss": 0.5618,
|
||
|
|
"step": 364
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4777327935222673,
|
||
|
|
"grad_norm": 0.5573343027370713,
|
||
|
|
"learning_rate": 3.553283018489946e-05,
|
||
|
|
"loss": 0.5426,
|
||
|
|
"step": 365
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4817813765182186,
|
||
|
|
"grad_norm": 0.413026150092314,
|
||
|
|
"learning_rate": 3.54971421420578e-05,
|
||
|
|
"loss": 0.5292,
|
||
|
|
"step": 366
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.48582995951417,
|
||
|
|
"grad_norm": 0.4307883449695307,
|
||
|
|
"learning_rate": 3.5461330184480524e-05,
|
||
|
|
"loss": 0.5586,
|
||
|
|
"step": 367
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4898785425101215,
|
||
|
|
"grad_norm": 0.4400574114423602,
|
||
|
|
"learning_rate": 3.54253945985191e-05,
|
||
|
|
"loss": 0.5403,
|
||
|
|
"step": 368
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4939271255060729,
|
||
|
|
"grad_norm": 0.39465493810840213,
|
||
|
|
"learning_rate": 3.5389335671513524e-05,
|
||
|
|
"loss": 0.538,
|
||
|
|
"step": 369
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4979757085020242,
|
||
|
|
"grad_norm": 0.4874948301856071,
|
||
|
|
"learning_rate": 3.5353153691789986e-05,
|
||
|
|
"loss": 0.5517,
|
||
|
|
"step": 370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5020242914979756,
|
||
|
|
"grad_norm": 0.39455894043145295,
|
||
|
|
"learning_rate": 3.531684894865866e-05,
|
||
|
|
"loss": 0.5541,
|
||
|
|
"step": 371
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5060728744939271,
|
||
|
|
"grad_norm": 0.46882120257103344,
|
||
|
|
"learning_rate": 3.528042173241128e-05,
|
||
|
|
"loss": 0.5421,
|
||
|
|
"step": 372
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5101214574898787,
|
||
|
|
"grad_norm": 0.45718890645236365,
|
||
|
|
"learning_rate": 3.524387233431892e-05,
|
||
|
|
"loss": 0.5417,
|
||
|
|
"step": 373
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5141700404858298,
|
||
|
|
"grad_norm": 0.4803488825701153,
|
||
|
|
"learning_rate": 3.520720104662958e-05,
|
||
|
|
"loss": 0.567,
|
||
|
|
"step": 374
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5182186234817814,
|
||
|
|
"grad_norm": 0.4467902505737919,
|
||
|
|
"learning_rate": 3.5170408162565904e-05,
|
||
|
|
"loss": 0.5522,
|
||
|
|
"step": 375
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.522267206477733,
|
||
|
|
"grad_norm": 0.4170937118219365,
|
||
|
|
"learning_rate": 3.513349397632281e-05,
|
||
|
|
"loss": 0.5414,
|
||
|
|
"step": 376
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.526315789473684,
|
||
|
|
"grad_norm": 0.42477100779857063,
|
||
|
|
"learning_rate": 3.5096458783065145e-05,
|
||
|
|
"loss": 0.5602,
|
||
|
|
"step": 377
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5303643724696356,
|
||
|
|
"grad_norm": 0.4241683836722998,
|
||
|
|
"learning_rate": 3.505930287892533e-05,
|
||
|
|
"loss": 0.5548,
|
||
|
|
"step": 378
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5344129554655872,
|
||
|
|
"grad_norm": 0.38051661807760684,
|
||
|
|
"learning_rate": 3.502202656100099e-05,
|
||
|
|
"loss": 0.5603,
|
||
|
|
"step": 379
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5384615384615383,
|
||
|
|
"grad_norm": 0.5357352899755009,
|
||
|
|
"learning_rate": 3.498463012735256e-05,
|
||
|
|
"loss": 0.5569,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.54251012145749,
|
||
|
|
"grad_norm": 0.482019382832418,
|
||
|
|
"learning_rate": 3.494711387700094e-05,
|
||
|
|
"loss": 0.5544,
|
||
|
|
"step": 381
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5465587044534415,
|
||
|
|
"grad_norm": 0.3938424753573089,
|
||
|
|
"learning_rate": 3.490947810992505e-05,
|
||
|
|
"loss": 0.552,
|
||
|
|
"step": 382
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5506072874493926,
|
||
|
|
"grad_norm": 0.5174118429863914,
|
||
|
|
"learning_rate": 3.48717231270595e-05,
|
||
|
|
"loss": 0.5527,
|
||
|
|
"step": 383
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5546558704453441,
|
||
|
|
"grad_norm": 0.5394486595950968,
|
||
|
|
"learning_rate": 3.483384923029211e-05,
|
||
|
|
"loss": 0.5484,
|
||
|
|
"step": 384
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5587044534412957,
|
||
|
|
"grad_norm": 0.41399024778091725,
|
||
|
|
"learning_rate": 3.479585672246156e-05,
|
||
|
|
"loss": 0.558,
|
||
|
|
"step": 385
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5627530364372468,
|
||
|
|
"grad_norm": 0.542223157798853,
|
||
|
|
"learning_rate": 3.4757745907354924e-05,
|
||
|
|
"loss": 0.5536,
|
||
|
|
"step": 386
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5668016194331984,
|
||
|
|
"grad_norm": 0.5654299434172781,
|
||
|
|
"learning_rate": 3.471951708970527e-05,
|
||
|
|
"loss": 0.5368,
|
||
|
|
"step": 387
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5708502024291497,
|
||
|
|
"grad_norm": 0.6319379175885185,
|
||
|
|
"learning_rate": 3.468117057518921e-05,
|
||
|
|
"loss": 0.5569,
|
||
|
|
"step": 388
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.574898785425101,
|
||
|
|
"grad_norm": 0.6374862750444101,
|
||
|
|
"learning_rate": 3.4642706670424426e-05,
|
||
|
|
"loss": 0.5497,
|
||
|
|
"step": 389
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5789473684210527,
|
||
|
|
"grad_norm": 0.6211163536487391,
|
||
|
|
"learning_rate": 3.460412568296731e-05,
|
||
|
|
"loss": 0.5443,
|
||
|
|
"step": 390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.582995951417004,
|
||
|
|
"grad_norm": 0.5075433409846783,
|
||
|
|
"learning_rate": 3.456542792131039e-05,
|
||
|
|
"loss": 0.5487,
|
||
|
|
"step": 391
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5870445344129553,
|
||
|
|
"grad_norm": 0.4881890963846757,
|
||
|
|
"learning_rate": 3.4526613694879936e-05,
|
||
|
|
"loss": 0.5608,
|
||
|
|
"step": 392
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.591093117408907,
|
||
|
|
"grad_norm": 0.5134647230223943,
|
||
|
|
"learning_rate": 3.448768331403348e-05,
|
||
|
|
"loss": 0.5494,
|
||
|
|
"step": 393
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5951417004048583,
|
||
|
|
"grad_norm": 0.4850033851457585,
|
||
|
|
"learning_rate": 3.44486370900573e-05,
|
||
|
|
"loss": 0.5649,
|
||
|
|
"step": 394
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5991902834008096,
|
||
|
|
"grad_norm": 0.4366799252060436,
|
||
|
|
"learning_rate": 3.440947533516396e-05,
|
||
|
|
"loss": 0.5433,
|
||
|
|
"step": 395
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6032388663967612,
|
||
|
|
"grad_norm": 0.4915409095825524,
|
||
|
|
"learning_rate": 3.4370198362489816e-05,
|
||
|
|
"loss": 0.5626,
|
||
|
|
"step": 396
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6072874493927125,
|
||
|
|
"grad_norm": 0.45940535841031205,
|
||
|
|
"learning_rate": 3.43308064860925e-05,
|
||
|
|
"loss": 0.5435,
|
||
|
|
"step": 397
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6113360323886639,
|
||
|
|
"grad_norm": 0.37867939112393256,
|
||
|
|
"learning_rate": 3.4291300020948393e-05,
|
||
|
|
"loss": 0.564,
|
||
|
|
"step": 398
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6153846153846154,
|
||
|
|
"grad_norm": 0.46982559972310506,
|
||
|
|
"learning_rate": 3.425167928295015e-05,
|
||
|
|
"loss": 0.5558,
|
||
|
|
"step": 399
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6194331983805668,
|
||
|
|
"grad_norm": 0.3447582771059348,
|
||
|
|
"learning_rate": 3.421194458890411e-05,
|
||
|
|
"loss": 0.539,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.623481781376518,
|
||
|
|
"grad_norm": 0.4739530629561756,
|
||
|
|
"learning_rate": 3.4172096256527865e-05,
|
||
|
|
"loss": 0.5512,
|
||
|
|
"step": 401
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6275303643724697,
|
||
|
|
"grad_norm": 0.35737477422788033,
|
||
|
|
"learning_rate": 3.4132134604447595e-05,
|
||
|
|
"loss": 0.5498,
|
||
|
|
"step": 402
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.631578947368421,
|
||
|
|
"grad_norm": 0.37833474400556966,
|
||
|
|
"learning_rate": 3.409205995219561e-05,
|
||
|
|
"loss": 0.5461,
|
||
|
|
"step": 403
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6356275303643724,
|
||
|
|
"grad_norm": 0.42136837893130685,
|
||
|
|
"learning_rate": 3.4051872620207765e-05,
|
||
|
|
"loss": 0.5257,
|
||
|
|
"step": 404
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.639676113360324,
|
||
|
|
"grad_norm": 0.47416360503414695,
|
||
|
|
"learning_rate": 3.40115729298209e-05,
|
||
|
|
"loss": 0.575,
|
||
|
|
"step": 405
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6437246963562753,
|
||
|
|
"grad_norm": 0.41500293083067213,
|
||
|
|
"learning_rate": 3.397116120327025e-05,
|
||
|
|
"loss": 0.5518,
|
||
|
|
"step": 406
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6477732793522266,
|
||
|
|
"grad_norm": 0.5123692778577028,
|
||
|
|
"learning_rate": 3.3930637763686916e-05,
|
||
|
|
"loss": 0.5393,
|
||
|
|
"step": 407
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6518218623481782,
|
||
|
|
"grad_norm": 0.4512501757856307,
|
||
|
|
"learning_rate": 3.389000293509524e-05,
|
||
|
|
"loss": 0.5477,
|
||
|
|
"step": 408
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6558704453441295,
|
||
|
|
"grad_norm": 0.3927144558433259,
|
||
|
|
"learning_rate": 3.3849257042410225e-05,
|
||
|
|
"loss": 0.5478,
|
||
|
|
"step": 409
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6599190283400809,
|
||
|
|
"grad_norm": 0.45171782322140364,
|
||
|
|
"learning_rate": 3.380840041143494e-05,
|
||
|
|
"loss": 0.5555,
|
||
|
|
"step": 410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6639676113360324,
|
||
|
|
"grad_norm": 0.38609462614884377,
|
||
|
|
"learning_rate": 3.376743336885792e-05,
|
||
|
|
"loss": 0.5512,
|
||
|
|
"step": 411
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6680161943319838,
|
||
|
|
"grad_norm": 0.3855332775234726,
|
||
|
|
"learning_rate": 3.372635624225055e-05,
|
||
|
|
"loss": 0.5569,
|
||
|
|
"step": 412
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6720647773279351,
|
||
|
|
"grad_norm": 0.5152862683549998,
|
||
|
|
"learning_rate": 3.3685169360064445e-05,
|
||
|
|
"loss": 0.5653,
|
||
|
|
"step": 413
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6761133603238867,
|
||
|
|
"grad_norm": 0.532655027125735,
|
||
|
|
"learning_rate": 3.3643873051628804e-05,
|
||
|
|
"loss": 0.5608,
|
||
|
|
"step": 414
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.680161943319838,
|
||
|
|
"grad_norm": 0.6108427499529075,
|
||
|
|
"learning_rate": 3.3602467647147814e-05,
|
||
|
|
"loss": 0.5735,
|
||
|
|
"step": 415
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6842105263157894,
|
||
|
|
"grad_norm": 0.5015537247732341,
|
||
|
|
"learning_rate": 3.356095347769799e-05,
|
||
|
|
"loss": 0.5481,
|
||
|
|
"step": 416
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.688259109311741,
|
||
|
|
"grad_norm": 0.5348150300321532,
|
||
|
|
"learning_rate": 3.351933087522552e-05,
|
||
|
|
"loss": 0.5545,
|
||
|
|
"step": 417
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6923076923076923,
|
||
|
|
"grad_norm": 0.41510643025274463,
|
||
|
|
"learning_rate": 3.3477600172543626e-05,
|
||
|
|
"loss": 0.5493,
|
||
|
|
"step": 418
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6963562753036436,
|
||
|
|
"grad_norm": 0.5123200661215432,
|
||
|
|
"learning_rate": 3.3435761703329894e-05,
|
||
|
|
"loss": 0.5587,
|
||
|
|
"step": 419
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7004048582995952,
|
||
|
|
"grad_norm": 0.3938895165544414,
|
||
|
|
"learning_rate": 3.3393815802123606e-05,
|
||
|
|
"loss": 0.5395,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7044534412955465,
|
||
|
|
"grad_norm": 0.48836833526973966,
|
||
|
|
"learning_rate": 3.335176280432308e-05,
|
||
|
|
"loss": 0.5428,
|
||
|
|
"step": 421
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.708502024291498,
|
||
|
|
"grad_norm": 0.41700771118523683,
|
||
|
|
"learning_rate": 3.330960304618295e-05,
|
||
|
|
"loss": 0.5381,
|
||
|
|
"step": 422
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7125506072874495,
|
||
|
|
"grad_norm": 0.42392661138671284,
|
||
|
|
"learning_rate": 3.326733686481153e-05,
|
||
|
|
"loss": 0.5408,
|
||
|
|
"step": 423
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7165991902834008,
|
||
|
|
"grad_norm": 0.4498850780548398,
|
||
|
|
"learning_rate": 3.3224964598168084e-05,
|
||
|
|
"loss": 0.5589,
|
||
|
|
"step": 424
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7206477732793521,
|
||
|
|
"grad_norm": 0.4389911378602445,
|
||
|
|
"learning_rate": 3.318248658506012e-05,
|
||
|
|
"loss": 0.5524,
|
||
|
|
"step": 425
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7246963562753037,
|
||
|
|
"grad_norm": 0.45545630188421915,
|
||
|
|
"learning_rate": 3.313990316514069e-05,
|
||
|
|
"loss": 0.5608,
|
||
|
|
"step": 426
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.728744939271255,
|
||
|
|
"grad_norm": 0.3724498302675261,
|
||
|
|
"learning_rate": 3.309721467890571e-05,
|
||
|
|
"loss": 0.5456,
|
||
|
|
"step": 427
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7327935222672064,
|
||
|
|
"grad_norm": 0.4433652919655176,
|
||
|
|
"learning_rate": 3.305442146769115e-05,
|
||
|
|
"loss": 0.554,
|
||
|
|
"step": 428
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.736842105263158,
|
||
|
|
"grad_norm": 0.3882061134747794,
|
||
|
|
"learning_rate": 3.30115238736704e-05,
|
||
|
|
"loss": 0.5607,
|
||
|
|
"step": 429
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7408906882591093,
|
||
|
|
"grad_norm": 0.4120555730353867,
|
||
|
|
"learning_rate": 3.296852223985148e-05,
|
||
|
|
"loss": 0.5324,
|
||
|
|
"step": 430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7449392712550607,
|
||
|
|
"grad_norm": 0.36729827490763023,
|
||
|
|
"learning_rate": 3.2925416910074315e-05,
|
||
|
|
"loss": 0.5477,
|
||
|
|
"step": 431
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7489878542510122,
|
||
|
|
"grad_norm": 0.45722175229791023,
|
||
|
|
"learning_rate": 3.288220822900796e-05,
|
||
|
|
"loss": 0.5404,
|
||
|
|
"step": 432
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7530364372469636,
|
||
|
|
"grad_norm": 0.3776632127113273,
|
||
|
|
"learning_rate": 3.283889654214788e-05,
|
||
|
|
"loss": 0.5311,
|
||
|
|
"step": 433
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.757085020242915,
|
||
|
|
"grad_norm": 0.4056120433513921,
|
||
|
|
"learning_rate": 3.279548219581317e-05,
|
||
|
|
"loss": 0.5457,
|
||
|
|
"step": 434
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7611336032388665,
|
||
|
|
"grad_norm": 0.4105112997028296,
|
||
|
|
"learning_rate": 3.275196553714379e-05,
|
||
|
|
"loss": 0.5512,
|
||
|
|
"step": 435
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7651821862348178,
|
||
|
|
"grad_norm": 0.3680625477691142,
|
||
|
|
"learning_rate": 3.270834691409778e-05,
|
||
|
|
"loss": 0.5433,
|
||
|
|
"step": 436
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7692307692307692,
|
||
|
|
"grad_norm": 0.45468805123926964,
|
||
|
|
"learning_rate": 3.266462667544849e-05,
|
||
|
|
"loss": 0.5473,
|
||
|
|
"step": 437
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7732793522267207,
|
||
|
|
"grad_norm": 0.3546369201109977,
|
||
|
|
"learning_rate": 3.262080517078178e-05,
|
||
|
|
"loss": 0.5556,
|
||
|
|
"step": 438
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.777327935222672,
|
||
|
|
"grad_norm": 0.4720267497109987,
|
||
|
|
"learning_rate": 3.257688275049323e-05,
|
||
|
|
"loss": 0.5466,
|
||
|
|
"step": 439
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7813765182186234,
|
||
|
|
"grad_norm": 0.34544705728811137,
|
||
|
|
"learning_rate": 3.253285976578535e-05,
|
||
|
|
"loss": 0.5443,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.785425101214575,
|
||
|
|
"grad_norm": 0.45228105252286954,
|
||
|
|
"learning_rate": 3.2488736568664756e-05,
|
||
|
|
"loss": 0.5436,
|
||
|
|
"step": 441
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7894736842105263,
|
||
|
|
"grad_norm": 0.41616520171122157,
|
||
|
|
"learning_rate": 3.244451351193935e-05,
|
||
|
|
"loss": 0.5475,
|
||
|
|
"step": 442
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7935222672064777,
|
||
|
|
"grad_norm": 0.4657120830205619,
|
||
|
|
"learning_rate": 3.240019094921551e-05,
|
||
|
|
"loss": 0.5509,
|
||
|
|
"step": 443
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7975708502024292,
|
||
|
|
"grad_norm": 0.35807530886934,
|
||
|
|
"learning_rate": 3.235576923489527e-05,
|
||
|
|
"loss": 0.5535,
|
||
|
|
"step": 444
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8016194331983806,
|
||
|
|
"grad_norm": 0.42910499118022677,
|
||
|
|
"learning_rate": 3.2311248724173476e-05,
|
||
|
|
"loss": 0.5468,
|
||
|
|
"step": 445
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.805668016194332,
|
||
|
|
"grad_norm": 0.4138754244853176,
|
||
|
|
"learning_rate": 3.226662977303494e-05,
|
||
|
|
"loss": 0.5594,
|
||
|
|
"step": 446
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8097165991902835,
|
||
|
|
"grad_norm": 0.36266840152979907,
|
||
|
|
"learning_rate": 3.2221912738251626e-05,
|
||
|
|
"loss": 0.5506,
|
||
|
|
"step": 447
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8137651821862348,
|
||
|
|
"grad_norm": 0.4526986373398701,
|
||
|
|
"learning_rate": 3.2177097977379736e-05,
|
||
|
|
"loss": 0.5301,
|
||
|
|
"step": 448
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8178137651821862,
|
||
|
|
"grad_norm": 0.3585237262414563,
|
||
|
|
"learning_rate": 3.21321858487569e-05,
|
||
|
|
"loss": 0.5551,
|
||
|
|
"step": 449
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8218623481781377,
|
||
|
|
"grad_norm": 0.47654069174534053,
|
||
|
|
"learning_rate": 3.208717671149931e-05,
|
||
|
|
"loss": 0.5506,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.825910931174089,
|
||
|
|
"grad_norm": 0.40839194924916145,
|
||
|
|
"learning_rate": 3.204207092549882e-05,
|
||
|
|
"loss": 0.5404,
|
||
|
|
"step": 451
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8299595141700404,
|
||
|
|
"grad_norm": 0.3998278469136544,
|
||
|
|
"learning_rate": 3.19968688514201e-05,
|
||
|
|
"loss": 0.5473,
|
||
|
|
"step": 452
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.834008097165992,
|
||
|
|
"grad_norm": 0.44899481722834267,
|
||
|
|
"learning_rate": 3.195157085069774e-05,
|
||
|
|
"loss": 0.5478,
|
||
|
|
"step": 453
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8380566801619433,
|
||
|
|
"grad_norm": 0.4075145680490999,
|
||
|
|
"learning_rate": 3.190617728553332e-05,
|
||
|
|
"loss": 0.5648,
|
||
|
|
"step": 454
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8421052631578947,
|
||
|
|
"grad_norm": 0.39224400573681795,
|
||
|
|
"learning_rate": 3.1860688518892606e-05,
|
||
|
|
"loss": 0.5631,
|
||
|
|
"step": 455
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8461538461538463,
|
||
|
|
"grad_norm": 0.35881012324622164,
|
||
|
|
"learning_rate": 3.181510491450254e-05,
|
||
|
|
"loss": 0.554,
|
||
|
|
"step": 456
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8502024291497976,
|
||
|
|
"grad_norm": 0.4070925198407618,
|
||
|
|
"learning_rate": 3.176942683684842e-05,
|
||
|
|
"loss": 0.5436,
|
||
|
|
"step": 457
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.854251012145749,
|
||
|
|
"grad_norm": 0.3995158115344002,
|
||
|
|
"learning_rate": 3.1723654651170934e-05,
|
||
|
|
"loss": 0.5485,
|
||
|
|
"step": 458
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8582995951417005,
|
||
|
|
"grad_norm": 0.46543775323672254,
|
||
|
|
"learning_rate": 3.167778872346327e-05,
|
||
|
|
"loss": 0.5261,
|
||
|
|
"step": 459
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8623481781376519,
|
||
|
|
"grad_norm": 0.4314581370809254,
|
||
|
|
"learning_rate": 3.163182942046815e-05,
|
||
|
|
"loss": 0.5611,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8663967611336032,
|
||
|
|
"grad_norm": 0.3423538756451831,
|
||
|
|
"learning_rate": 3.158577710967494e-05,
|
||
|
|
"loss": 0.5492,
|
||
|
|
"step": 461
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8704453441295548,
|
||
|
|
"grad_norm": 0.48678469136820907,
|
||
|
|
"learning_rate": 3.153963215931669e-05,
|
||
|
|
"loss": 0.542,
|
||
|
|
"step": 462
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.874493927125506,
|
||
|
|
"grad_norm": 0.3394810968245217,
|
||
|
|
"learning_rate": 3.14933949383672e-05,
|
||
|
|
"loss": 0.5456,
|
||
|
|
"step": 463
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8785425101214575,
|
||
|
|
"grad_norm": 0.3937261361856436,
|
||
|
|
"learning_rate": 3.144706581653805e-05,
|
||
|
|
"loss": 0.5541,
|
||
|
|
"step": 464
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.882591093117409,
|
||
|
|
"grad_norm": 0.3592123099391868,
|
||
|
|
"learning_rate": 3.140064516427566e-05,
|
||
|
|
"loss": 0.5531,
|
||
|
|
"step": 465
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8866396761133604,
|
||
|
|
"grad_norm": 0.4167953601339203,
|
||
|
|
"learning_rate": 3.135413335275833e-05,
|
||
|
|
"loss": 0.5567,
|
||
|
|
"step": 466
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8906882591093117,
|
||
|
|
"grad_norm": 0.35623651252377,
|
||
|
|
"learning_rate": 3.130753075389327e-05,
|
||
|
|
"loss": 0.531,
|
||
|
|
"step": 467
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8947368421052633,
|
||
|
|
"grad_norm": 0.5205198585236963,
|
||
|
|
"learning_rate": 3.1260837740313625e-05,
|
||
|
|
"loss": 0.5476,
|
||
|
|
"step": 468
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8987854251012146,
|
||
|
|
"grad_norm": 0.40651845114732427,
|
||
|
|
"learning_rate": 3.121405468537547e-05,
|
||
|
|
"loss": 0.553,
|
||
|
|
"step": 469
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.902834008097166,
|
||
|
|
"grad_norm": 0.41841636366429635,
|
||
|
|
"learning_rate": 3.116718196315488e-05,
|
||
|
|
"loss": 0.5376,
|
||
|
|
"step": 470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9068825910931175,
|
||
|
|
"grad_norm": 0.37450354897551613,
|
||
|
|
"learning_rate": 3.1120219948444894e-05,
|
||
|
|
"loss": 0.5449,
|
||
|
|
"step": 471
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9109311740890689,
|
||
|
|
"grad_norm": 0.478571281268759,
|
||
|
|
"learning_rate": 3.107316901675252e-05,
|
||
|
|
"loss": 0.5446,
|
||
|
|
"step": 472
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9149797570850202,
|
||
|
|
"grad_norm": 0.3431961573217063,
|
||
|
|
"learning_rate": 3.1026029544295745e-05,
|
||
|
|
"loss": 0.538,
|
||
|
|
"step": 473
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9190283400809718,
|
||
|
|
"grad_norm": 0.4569109066299344,
|
||
|
|
"learning_rate": 3.097880190800057e-05,
|
||
|
|
"loss": 0.5458,
|
||
|
|
"step": 474
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9230769230769231,
|
||
|
|
"grad_norm": 0.3797637854649633,
|
||
|
|
"learning_rate": 3.093148648549788e-05,
|
||
|
|
"loss": 0.5444,
|
||
|
|
"step": 475
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9271255060728745,
|
||
|
|
"grad_norm": 0.4850407895914578,
|
||
|
|
"learning_rate": 3.088408365512055e-05,
|
||
|
|
"loss": 0.5352,
|
||
|
|
"step": 476
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.931174089068826,
|
||
|
|
"grad_norm": 0.39154724201976876,
|
||
|
|
"learning_rate": 3.083659379590034e-05,
|
||
|
|
"loss": 0.5555,
|
||
|
|
"step": 477
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9352226720647774,
|
||
|
|
"grad_norm": 0.4059931816105655,
|
||
|
|
"learning_rate": 3.0789017287564913e-05,
|
||
|
|
"loss": 0.5389,
|
||
|
|
"step": 478
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9392712550607287,
|
||
|
|
"grad_norm": 0.39198406824890436,
|
||
|
|
"learning_rate": 3.074135451053475e-05,
|
||
|
|
"loss": 0.5506,
|
||
|
|
"step": 479
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9433198380566803,
|
||
|
|
"grad_norm": 0.4309495555395892,
|
||
|
|
"learning_rate": 3.069360584592015e-05,
|
||
|
|
"loss": 0.5471,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9473684210526314,
|
||
|
|
"grad_norm": 0.36268280729713287,
|
||
|
|
"learning_rate": 3.064577167551816e-05,
|
||
|
|
"loss": 0.5418,
|
||
|
|
"step": 481
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.951417004048583,
|
||
|
|
"grad_norm": 0.5410017886574541,
|
||
|
|
"learning_rate": 3.059785238180953e-05,
|
||
|
|
"loss": 0.5464,
|
||
|
|
"step": 482
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9554655870445345,
|
||
|
|
"grad_norm": 0.38013406316511567,
|
||
|
|
"learning_rate": 3.054984834795566e-05,
|
||
|
|
"loss": 0.5498,
|
||
|
|
"step": 483
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9595141700404857,
|
||
|
|
"grad_norm": 0.3926074445531767,
|
||
|
|
"learning_rate": 3.0501759957795537e-05,
|
||
|
|
"loss": 0.5363,
|
||
|
|
"step": 484
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9635627530364372,
|
||
|
|
"grad_norm": 0.474164894760192,
|
||
|
|
"learning_rate": 3.0453587595842627e-05,
|
||
|
|
"loss": 0.5439,
|
||
|
|
"step": 485
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9676113360323888,
|
||
|
|
"grad_norm": 0.4137540319481608,
|
||
|
|
"learning_rate": 3.0405331647281856e-05,
|
||
|
|
"loss": 0.5575,
|
||
|
|
"step": 486
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.97165991902834,
|
||
|
|
"grad_norm": 0.3592404831855967,
|
||
|
|
"learning_rate": 3.0356992497966508e-05,
|
||
|
|
"loss": 0.5299,
|
||
|
|
"step": 487
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9757085020242915,
|
||
|
|
"grad_norm": 0.43995117174321957,
|
||
|
|
"learning_rate": 3.0308570534415137e-05,
|
||
|
|
"loss": 0.5448,
|
||
|
|
"step": 488
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.979757085020243,
|
||
|
|
"grad_norm": 0.3846267866467673,
|
||
|
|
"learning_rate": 3.0260066143808467e-05,
|
||
|
|
"loss": 0.5359,
|
||
|
|
"step": 489
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9838056680161942,
|
||
|
|
"grad_norm": 0.3822557457112387,
|
||
|
|
"learning_rate": 3.021147971398632e-05,
|
||
|
|
"loss": 0.5584,
|
||
|
|
"step": 490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9878542510121457,
|
||
|
|
"grad_norm": 0.43765737969487994,
|
||
|
|
"learning_rate": 3.0162811633444478e-05,
|
||
|
|
"loss": 0.5409,
|
||
|
|
"step": 491
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9919028340080973,
|
||
|
|
"grad_norm": 0.4185551888312873,
|
||
|
|
"learning_rate": 3.0114062291331635e-05,
|
||
|
|
"loss": 0.5546,
|
||
|
|
"step": 492
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9959514170040484,
|
||
|
|
"grad_norm": 0.4504733570689978,
|
||
|
|
"learning_rate": 3.0065232077446217e-05,
|
||
|
|
"loss": 0.5449,
|
||
|
|
"step": 493
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"grad_norm": 0.3380077122309689,
|
||
|
|
"learning_rate": 3.001632138223332e-05,
|
||
|
|
"loss": 0.5235,
|
||
|
|
"step": 494
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0040485829959516,
|
||
|
|
"grad_norm": 0.5289131107391366,
|
||
|
|
"learning_rate": 2.9967330596781565e-05,
|
||
|
|
"loss": 0.4834,
|
||
|
|
"step": 495
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0080971659919027,
|
||
|
|
"grad_norm": 0.6022235810466267,
|
||
|
|
"learning_rate": 2.9918260112819952e-05,
|
||
|
|
"loss": 0.4836,
|
||
|
|
"step": 496
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0121457489878543,
|
||
|
|
"grad_norm": 0.4400790952976979,
|
||
|
|
"learning_rate": 2.9869110322714778e-05,
|
||
|
|
"loss": 0.4674,
|
||
|
|
"step": 497
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.016194331983806,
|
||
|
|
"grad_norm": 0.545815817391453,
|
||
|
|
"learning_rate": 2.9819881619466447e-05,
|
||
|
|
"loss": 0.4882,
|
||
|
|
"step": 498
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.020242914979757,
|
||
|
|
"grad_norm": 0.496900921892097,
|
||
|
|
"learning_rate": 2.9770574396706364e-05,
|
||
|
|
"loss": 0.4661,
|
||
|
|
"step": 499
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0242914979757085,
|
||
|
|
"grad_norm": 0.456382911999509,
|
||
|
|
"learning_rate": 2.972118904869376e-05,
|
||
|
|
"loss": 0.4593,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.02834008097166,
|
||
|
|
"grad_norm": 0.5263819962981636,
|
||
|
|
"learning_rate": 2.9671725970312563e-05,
|
||
|
|
"loss": 0.4656,
|
||
|
|
"step": 501
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.032388663967611,
|
||
|
|
"grad_norm": 0.4230056736391131,
|
||
|
|
"learning_rate": 2.9622185557068238e-05,
|
||
|
|
"loss": 0.471,
|
||
|
|
"step": 502
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0364372469635628,
|
||
|
|
"grad_norm": 0.46179584479083574,
|
||
|
|
"learning_rate": 2.95725682050846e-05,
|
||
|
|
"loss": 0.4839,
|
||
|
|
"step": 503
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0404858299595143,
|
||
|
|
"grad_norm": 0.48439366788063903,
|
||
|
|
"learning_rate": 2.9522874311100684e-05,
|
||
|
|
"loss": 0.4726,
|
||
|
|
"step": 504
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0445344129554655,
|
||
|
|
"grad_norm": 0.41547617409636994,
|
||
|
|
"learning_rate": 2.9473104272467526e-05,
|
||
|
|
"loss": 0.4942,
|
||
|
|
"step": 505
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.048582995951417,
|
||
|
|
"grad_norm": 0.594966581072272,
|
||
|
|
"learning_rate": 2.942325848714505e-05,
|
||
|
|
"loss": 0.4567,
|
||
|
|
"step": 506
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0526315789473686,
|
||
|
|
"grad_norm": 0.4294633248047652,
|
||
|
|
"learning_rate": 2.9373337353698804e-05,
|
||
|
|
"loss": 0.4957,
|
||
|
|
"step": 507
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0566801619433197,
|
||
|
|
"grad_norm": 0.526729042254394,
|
||
|
|
"learning_rate": 2.932334127129686e-05,
|
||
|
|
"loss": 0.4727,
|
||
|
|
"step": 508
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0607287449392713,
|
||
|
|
"grad_norm": 0.45056322101181373,
|
||
|
|
"learning_rate": 2.9273270639706547e-05,
|
||
|
|
"loss": 0.456,
|
||
|
|
"step": 509
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.064777327935223,
|
||
|
|
"grad_norm": 0.39405012393372085,
|
||
|
|
"learning_rate": 2.922312585929131e-05,
|
||
|
|
"loss": 0.4702,
|
||
|
|
"step": 510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.068825910931174,
|
||
|
|
"grad_norm": 0.4204333811534276,
|
||
|
|
"learning_rate": 2.9172907331007478e-05,
|
||
|
|
"loss": 0.4899,
|
||
|
|
"step": 511
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0728744939271255,
|
||
|
|
"grad_norm": 0.3887012620747415,
|
||
|
|
"learning_rate": 2.9122615456401053e-05,
|
||
|
|
"loss": 0.4654,
|
||
|
|
"step": 512
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.076923076923077,
|
||
|
|
"grad_norm": 0.35990481638091615,
|
||
|
|
"learning_rate": 2.9072250637604535e-05,
|
||
|
|
"loss": 0.4571,
|
||
|
|
"step": 513
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.080971659919028,
|
||
|
|
"grad_norm": 0.4002959330102329,
|
||
|
|
"learning_rate": 2.9021813277333667e-05,
|
||
|
|
"loss": 0.4686,
|
||
|
|
"step": 514
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.08502024291498,
|
||
|
|
"grad_norm": 0.39423234242883837,
|
||
|
|
"learning_rate": 2.8971303778884244e-05,
|
||
|
|
"loss": 0.466,
|
||
|
|
"step": 515
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0890688259109313,
|
||
|
|
"grad_norm": 0.3681086322766949,
|
||
|
|
"learning_rate": 2.8920722546128863e-05,
|
||
|
|
"loss": 0.4848,
|
||
|
|
"step": 516
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0931174089068825,
|
||
|
|
"grad_norm": 0.38978246302612757,
|
||
|
|
"learning_rate": 2.887006998351372e-05,
|
||
|
|
"loss": 0.4658,
|
||
|
|
"step": 517
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.097165991902834,
|
||
|
|
"grad_norm": 0.3532291335669879,
|
||
|
|
"learning_rate": 2.881934649605535e-05,
|
||
|
|
"loss": 0.4925,
|
||
|
|
"step": 518
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1012145748987856,
|
||
|
|
"grad_norm": 0.38186745331201294,
|
||
|
|
"learning_rate": 2.8768552489337408e-05,
|
||
|
|
"loss": 0.4711,
|
||
|
|
"step": 519
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1052631578947367,
|
||
|
|
"grad_norm": 0.36906415232922274,
|
||
|
|
"learning_rate": 2.8717688369507425e-05,
|
||
|
|
"loss": 0.4843,
|
||
|
|
"step": 520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1093117408906883,
|
||
|
|
"grad_norm": 0.36645053349597984,
|
||
|
|
"learning_rate": 2.8666754543273535e-05,
|
||
|
|
"loss": 0.4474,
|
||
|
|
"step": 521
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.11336032388664,
|
||
|
|
"grad_norm": 0.37114722122388955,
|
||
|
|
"learning_rate": 2.8615751417901273e-05,
|
||
|
|
"loss": 0.4555,
|
||
|
|
"step": 522
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.117408906882591,
|
||
|
|
"grad_norm": 0.372911001636862,
|
||
|
|
"learning_rate": 2.856467940121025e-05,
|
||
|
|
"loss": 0.4807,
|
||
|
|
"step": 523
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1214574898785425,
|
||
|
|
"grad_norm": 0.4071572798214776,
|
||
|
|
"learning_rate": 2.8513538901570964e-05,
|
||
|
|
"loss": 0.464,
|
||
|
|
"step": 524
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.125506072874494,
|
||
|
|
"grad_norm": 0.35632640658907105,
|
||
|
|
"learning_rate": 2.846233032790149e-05,
|
||
|
|
"loss": 0.4636,
|
||
|
|
"step": 525
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1295546558704452,
|
||
|
|
"grad_norm": 0.407421736752148,
|
||
|
|
"learning_rate": 2.84110540896642e-05,
|
||
|
|
"loss": 0.4676,
|
||
|
|
"step": 526
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.133603238866397,
|
||
|
|
"grad_norm": 0.3600858946924453,
|
||
|
|
"learning_rate": 2.8359710596862555e-05,
|
||
|
|
"loss": 0.4615,
|
||
|
|
"step": 527
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1376518218623484,
|
||
|
|
"grad_norm": 0.3887966984904334,
|
||
|
|
"learning_rate": 2.8308300260037734e-05,
|
||
|
|
"loss": 0.4578,
|
||
|
|
"step": 528
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1417004048582995,
|
||
|
|
"grad_norm": 0.42801194935906345,
|
||
|
|
"learning_rate": 2.8256823490265424e-05,
|
||
|
|
"loss": 0.4675,
|
||
|
|
"step": 529
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.145748987854251,
|
||
|
|
"grad_norm": 0.3950771679844169,
|
||
|
|
"learning_rate": 2.8205280699152515e-05,
|
||
|
|
"loss": 0.4745,
|
||
|
|
"step": 530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1497975708502026,
|
||
|
|
"grad_norm": 0.3585003723819789,
|
||
|
|
"learning_rate": 2.8153672298833777e-05,
|
||
|
|
"loss": 0.4628,
|
||
|
|
"step": 531
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1538461538461537,
|
||
|
|
"grad_norm": 0.41410722244982173,
|
||
|
|
"learning_rate": 2.8101998701968615e-05,
|
||
|
|
"loss": 0.485,
|
||
|
|
"step": 532
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1578947368421053,
|
||
|
|
"grad_norm": 0.3434360443862549,
|
||
|
|
"learning_rate": 2.805026032173773e-05,
|
||
|
|
"loss": 0.4815,
|
||
|
|
"step": 533
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.161943319838057,
|
||
|
|
"grad_norm": 0.39743634784732035,
|
||
|
|
"learning_rate": 2.7998457571839832e-05,
|
||
|
|
"loss": 0.4746,
|
||
|
|
"step": 534
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.165991902834008,
|
||
|
|
"grad_norm": 0.3813098043806024,
|
||
|
|
"learning_rate": 2.7946590866488322e-05,
|
||
|
|
"loss": 0.4696,
|
||
|
|
"step": 535
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1700404858299596,
|
||
|
|
"grad_norm": 0.3983542399307604,
|
||
|
|
"learning_rate": 2.7894660620408e-05,
|
||
|
|
"loss": 0.459,
|
||
|
|
"step": 536
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.174089068825911,
|
||
|
|
"grad_norm": 0.41987785835931407,
|
||
|
|
"learning_rate": 2.784266724883173e-05,
|
||
|
|
"loss": 0.4947,
|
||
|
|
"step": 537
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1781376518218623,
|
||
|
|
"grad_norm": 0.40241111496012827,
|
||
|
|
"learning_rate": 2.7790611167497128e-05,
|
||
|
|
"loss": 0.4789,
|
||
|
|
"step": 538
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.182186234817814,
|
||
|
|
"grad_norm": 0.39023673319231317,
|
||
|
|
"learning_rate": 2.7738492792643224e-05,
|
||
|
|
"loss": 0.4792,
|
||
|
|
"step": 539
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1862348178137654,
|
||
|
|
"grad_norm": 0.3793270718490223,
|
||
|
|
"learning_rate": 2.768631254100716e-05,
|
||
|
|
"loss": 0.4493,
|
||
|
|
"step": 540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1902834008097165,
|
||
|
|
"grad_norm": 0.37919787964995894,
|
||
|
|
"learning_rate": 2.7634070829820836e-05,
|
||
|
|
"loss": 0.4643,
|
||
|
|
"step": 541
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.194331983805668,
|
||
|
|
"grad_norm": 0.464876555099051,
|
||
|
|
"learning_rate": 2.7581768076807592e-05,
|
||
|
|
"loss": 0.4636,
|
||
|
|
"step": 542
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1983805668016196,
|
||
|
|
"grad_norm": 0.34831664505059756,
|
||
|
|
"learning_rate": 2.7529404700178832e-05,
|
||
|
|
"loss": 0.4663,
|
||
|
|
"step": 543
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2024291497975708,
|
||
|
|
"grad_norm": 0.473538749300893,
|
||
|
|
"learning_rate": 2.7476981118630735e-05,
|
||
|
|
"loss": 0.4647,
|
||
|
|
"step": 544
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2064777327935223,
|
||
|
|
"grad_norm": 0.41532908723824047,
|
||
|
|
"learning_rate": 2.742449775134085e-05,
|
||
|
|
"loss": 0.4744,
|
||
|
|
"step": 545
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2105263157894735,
|
||
|
|
"grad_norm": 0.4325420713415665,
|
||
|
|
"learning_rate": 2.7371955017964788e-05,
|
||
|
|
"loss": 0.4667,
|
||
|
|
"step": 546
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.214574898785425,
|
||
|
|
"grad_norm": 0.3712572618100617,
|
||
|
|
"learning_rate": 2.7319353338632845e-05,
|
||
|
|
"loss": 0.4729,
|
||
|
|
"step": 547
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2186234817813766,
|
||
|
|
"grad_norm": 0.4192323550670153,
|
||
|
|
"learning_rate": 2.7266693133946645e-05,
|
||
|
|
"loss": 0.4784,
|
||
|
|
"step": 548
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2226720647773277,
|
||
|
|
"grad_norm": 0.3760390778470649,
|
||
|
|
"learning_rate": 2.7213974824975785e-05,
|
||
|
|
"loss": 0.4648,
|
||
|
|
"step": 549
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2267206477732793,
|
||
|
|
"grad_norm": 0.4085235609646252,
|
||
|
|
"learning_rate": 2.7161198833254464e-05,
|
||
|
|
"loss": 0.4748,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.230769230769231,
|
||
|
|
"grad_norm": 0.36440628001217223,
|
||
|
|
"learning_rate": 2.7108365580778095e-05,
|
||
|
|
"loss": 0.4786,
|
||
|
|
"step": 551
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.234817813765182,
|
||
|
|
"grad_norm": 0.36064660679079313,
|
||
|
|
"learning_rate": 2.705547548999996e-05,
|
||
|
|
"loss": 0.4814,
|
||
|
|
"step": 552
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2388663967611335,
|
||
|
|
"grad_norm": 0.34405519777520305,
|
||
|
|
"learning_rate": 2.7002528983827817e-05,
|
||
|
|
"loss": 0.4652,
|
||
|
|
"step": 553
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.242914979757085,
|
||
|
|
"grad_norm": 0.3561085501893368,
|
||
|
|
"learning_rate": 2.6949526485620517e-05,
|
||
|
|
"loss": 0.4698,
|
||
|
|
"step": 554
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.246963562753036,
|
||
|
|
"grad_norm": 0.33308767185517657,
|
||
|
|
"learning_rate": 2.6896468419184623e-05,
|
||
|
|
"loss": 0.4774,
|
||
|
|
"step": 555
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.251012145748988,
|
||
|
|
"grad_norm": 0.35670569942609115,
|
||
|
|
"learning_rate": 2.6843355208771024e-05,
|
||
|
|
"loss": 0.4769,
|
||
|
|
"step": 556
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2550607287449393,
|
||
|
|
"grad_norm": 0.33158505889937534,
|
||
|
|
"learning_rate": 2.6790187279071536e-05,
|
||
|
|
"loss": 0.474,
|
||
|
|
"step": 557
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2591093117408905,
|
||
|
|
"grad_norm": 0.38381653261218773,
|
||
|
|
"learning_rate": 2.67369650552155e-05,
|
||
|
|
"loss": 0.475,
|
||
|
|
"step": 558
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.263157894736842,
|
||
|
|
"grad_norm": 0.3422734312753868,
|
||
|
|
"learning_rate": 2.6683688962766407e-05,
|
||
|
|
"loss": 0.4613,
|
||
|
|
"step": 559
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2672064777327936,
|
||
|
|
"grad_norm": 0.38188805603816467,
|
||
|
|
"learning_rate": 2.6630359427718468e-05,
|
||
|
|
"loss": 0.4667,
|
||
|
|
"step": 560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2712550607287447,
|
||
|
|
"grad_norm": 0.46925988661233364,
|
||
|
|
"learning_rate": 2.6576976876493225e-05,
|
||
|
|
"loss": 0.4689,
|
||
|
|
"step": 561
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2753036437246963,
|
||
|
|
"grad_norm": 0.37793021251002573,
|
||
|
|
"learning_rate": 2.652354173593614e-05,
|
||
|
|
"loss": 0.4797,
|
||
|
|
"step": 562
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.279352226720648,
|
||
|
|
"grad_norm": 0.4605648510080909,
|
||
|
|
"learning_rate": 2.647005443331317e-05,
|
||
|
|
"loss": 0.4563,
|
||
|
|
"step": 563
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.283400809716599,
|
||
|
|
"grad_norm": 0.43125599069077186,
|
||
|
|
"learning_rate": 2.6416515396307354e-05,
|
||
|
|
"loss": 0.4742,
|
||
|
|
"step": 564
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2874493927125505,
|
||
|
|
"grad_norm": 0.46432052133395,
|
||
|
|
"learning_rate": 2.636292505301542e-05,
|
||
|
|
"loss": 0.4898,
|
||
|
|
"step": 565
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.291497975708502,
|
||
|
|
"grad_norm": 0.44733140224716633,
|
||
|
|
"learning_rate": 2.630928383194431e-05,
|
||
|
|
"loss": 0.4562,
|
||
|
|
"step": 566
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2955465587044532,
|
||
|
|
"grad_norm": 0.48860470102474335,
|
||
|
|
"learning_rate": 2.625559216200781e-05,
|
||
|
|
"loss": 0.4753,
|
||
|
|
"step": 567
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.299595141700405,
|
||
|
|
"grad_norm": 0.36079111143282705,
|
||
|
|
"learning_rate": 2.620185047252308e-05,
|
||
|
|
"loss": 0.4859,
|
||
|
|
"step": 568
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3036437246963564,
|
||
|
|
"grad_norm": 0.3702166687096711,
|
||
|
|
"learning_rate": 2.614805919320723e-05,
|
||
|
|
"loss": 0.485,
|
||
|
|
"step": 569
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3076923076923075,
|
||
|
|
"grad_norm": 0.35874775260123865,
|
||
|
|
"learning_rate": 2.6094218754173895e-05,
|
||
|
|
"loss": 0.4859,
|
||
|
|
"step": 570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.311740890688259,
|
||
|
|
"grad_norm": 0.42129381075605,
|
||
|
|
"learning_rate": 2.60403295859298e-05,
|
||
|
|
"loss": 0.4669,
|
||
|
|
"step": 571
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3157894736842106,
|
||
|
|
"grad_norm": 0.38504775067315683,
|
||
|
|
"learning_rate": 2.5986392119371297e-05,
|
||
|
|
"loss": 0.4648,
|
||
|
|
"step": 572
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3198380566801617,
|
||
|
|
"grad_norm": 0.3891173682528562,
|
||
|
|
"learning_rate": 2.5932406785780928e-05,
|
||
|
|
"loss": 0.4579,
|
||
|
|
"step": 573
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3238866396761133,
|
||
|
|
"grad_norm": 0.3426575406941934,
|
||
|
|
"learning_rate": 2.5878374016823987e-05,
|
||
|
|
"loss": 0.4579,
|
||
|
|
"step": 574
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.327935222672065,
|
||
|
|
"grad_norm": 0.4511644725966341,
|
||
|
|
"learning_rate": 2.5824294244545045e-05,
|
||
|
|
"loss": 0.485,
|
||
|
|
"step": 575
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.331983805668016,
|
||
|
|
"grad_norm": 0.29572287582732354,
|
||
|
|
"learning_rate": 2.577016790136453e-05,
|
||
|
|
"loss": 0.4831,
|
||
|
|
"step": 576
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3360323886639676,
|
||
|
|
"grad_norm": 0.3313896962120594,
|
||
|
|
"learning_rate": 2.5715995420075225e-05,
|
||
|
|
"loss": 0.4766,
|
||
|
|
"step": 577
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.340080971659919,
|
||
|
|
"grad_norm": 0.3315563690434995,
|
||
|
|
"learning_rate": 2.5661777233838864e-05,
|
||
|
|
"loss": 0.4603,
|
||
|
|
"step": 578
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3441295546558703,
|
||
|
|
"grad_norm": 0.3373110078275791,
|
||
|
|
"learning_rate": 2.5607513776182616e-05,
|
||
|
|
"loss": 0.4538,
|
||
|
|
"step": 579
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.348178137651822,
|
||
|
|
"grad_norm": 0.3477669631255805,
|
||
|
|
"learning_rate": 2.5553205480995633e-05,
|
||
|
|
"loss": 0.4641,
|
||
|
|
"step": 580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3522267206477734,
|
||
|
|
"grad_norm": 0.3440817599220889,
|
||
|
|
"learning_rate": 2.5498852782525604e-05,
|
||
|
|
"loss": 0.4634,
|
||
|
|
"step": 581
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3562753036437245,
|
||
|
|
"grad_norm": 0.37659900133153484,
|
||
|
|
"learning_rate": 2.5444456115375252e-05,
|
||
|
|
"loss": 0.4497,
|
||
|
|
"step": 582
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.360323886639676,
|
||
|
|
"grad_norm": 0.3602350626809477,
|
||
|
|
"learning_rate": 2.539001591449889e-05,
|
||
|
|
"loss": 0.4553,
|
||
|
|
"step": 583
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3643724696356276,
|
||
|
|
"grad_norm": 0.39741638228239157,
|
||
|
|
"learning_rate": 2.53355326151989e-05,
|
||
|
|
"loss": 0.4776,
|
||
|
|
"step": 584
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3684210526315788,
|
||
|
|
"grad_norm": 0.32748728437779256,
|
||
|
|
"learning_rate": 2.5281006653122288e-05,
|
||
|
|
"loss": 0.4677,
|
||
|
|
"step": 585
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3724696356275303,
|
||
|
|
"grad_norm": 0.3861007820195246,
|
||
|
|
"learning_rate": 2.5226438464257217e-05,
|
||
|
|
"loss": 0.4601,
|
||
|
|
"step": 586
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.376518218623482,
|
||
|
|
"grad_norm": 0.36997494102133976,
|
||
|
|
"learning_rate": 2.5171828484929447e-05,
|
||
|
|
"loss": 0.4839,
|
||
|
|
"step": 587
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.380566801619433,
|
||
|
|
"grad_norm": 0.37491367405944975,
|
||
|
|
"learning_rate": 2.5117177151798934e-05,
|
||
|
|
"loss": 0.4591,
|
||
|
|
"step": 588
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3846153846153846,
|
||
|
|
"grad_norm": 0.3233274364368284,
|
||
|
|
"learning_rate": 2.506248490185627e-05,
|
||
|
|
"loss": 0.4802,
|
||
|
|
"step": 589
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.388663967611336,
|
||
|
|
"grad_norm": 0.35083045483335773,
|
||
|
|
"learning_rate": 2.500775217241923e-05,
|
||
|
|
"loss": 0.4739,
|
||
|
|
"step": 590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3927125506072873,
|
||
|
|
"grad_norm": 0.326044765223085,
|
||
|
|
"learning_rate": 2.4952979401129266e-05,
|
||
|
|
"loss": 0.4779,
|
||
|
|
"step": 591
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.396761133603239,
|
||
|
|
"grad_norm": 0.34841866611269434,
|
||
|
|
"learning_rate": 2.4898167025948e-05,
|
||
|
|
"loss": 0.4512,
|
||
|
|
"step": 592
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4008097165991904,
|
||
|
|
"grad_norm": 0.31477475403259236,
|
||
|
|
"learning_rate": 2.4843315485153703e-05,
|
||
|
|
"loss": 0.476,
|
||
|
|
"step": 593
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4048582995951415,
|
||
|
|
"grad_norm": 0.3482626468712851,
|
||
|
|
"learning_rate": 2.4788425217337856e-05,
|
||
|
|
"loss": 0.4807,
|
||
|
|
"step": 594
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.408906882591093,
|
||
|
|
"grad_norm": 0.32164393241425837,
|
||
|
|
"learning_rate": 2.4733496661401572e-05,
|
||
|
|
"loss": 0.4709,
|
||
|
|
"step": 595
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4129554655870447,
|
||
|
|
"grad_norm": 0.3970875387432804,
|
||
|
|
"learning_rate": 2.4678530256552114e-05,
|
||
|
|
"loss": 0.4894,
|
||
|
|
"step": 596
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.417004048582996,
|
||
|
|
"grad_norm": 0.34086951114197794,
|
||
|
|
"learning_rate": 2.46235264422994e-05,
|
||
|
|
"loss": 0.4649,
|
||
|
|
"step": 597
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4210526315789473,
|
||
|
|
"grad_norm": 0.35673013344339477,
|
||
|
|
"learning_rate": 2.4568485658452452e-05,
|
||
|
|
"loss": 0.4714,
|
||
|
|
"step": 598
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.425101214574899,
|
||
|
|
"grad_norm": 0.3323238295952803,
|
||
|
|
"learning_rate": 2.451340834511592e-05,
|
||
|
|
"loss": 0.4692,
|
||
|
|
"step": 599
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.42914979757085,
|
||
|
|
"grad_norm": 0.39056065594393535,
|
||
|
|
"learning_rate": 2.4458294942686528e-05,
|
||
|
|
"loss": 0.4698,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4331983805668016,
|
||
|
|
"grad_norm": 0.303973712665435,
|
||
|
|
"learning_rate": 2.440314589184958e-05,
|
||
|
|
"loss": 0.4784,
|
||
|
|
"step": 601
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.437246963562753,
|
||
|
|
"grad_norm": 0.38368627274800393,
|
||
|
|
"learning_rate": 2.434796163357541e-05,
|
||
|
|
"loss": 0.4841,
|
||
|
|
"step": 602
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4412955465587043,
|
||
|
|
"grad_norm": 0.35763831536025387,
|
||
|
|
"learning_rate": 2.429274260911588e-05,
|
||
|
|
"loss": 0.4631,
|
||
|
|
"step": 603
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.445344129554656,
|
||
|
|
"grad_norm": 0.37175290371079334,
|
||
|
|
"learning_rate": 2.423748926000084e-05,
|
||
|
|
"loss": 0.4458,
|
||
|
|
"step": 604
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4493927125506074,
|
||
|
|
"grad_norm": 0.3585467420104883,
|
||
|
|
"learning_rate": 2.418220202803459e-05,
|
||
|
|
"loss": 0.4796,
|
||
|
|
"step": 605
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4534412955465585,
|
||
|
|
"grad_norm": 0.3655276274995093,
|
||
|
|
"learning_rate": 2.412688135529237e-05,
|
||
|
|
"loss": 0.471,
|
||
|
|
"step": 606
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.45748987854251,
|
||
|
|
"grad_norm": 0.4176348120637941,
|
||
|
|
"learning_rate": 2.4071527684116807e-05,
|
||
|
|
"loss": 0.4792,
|
||
|
|
"step": 607
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4615384615384617,
|
||
|
|
"grad_norm": 0.3582853844595898,
|
||
|
|
"learning_rate": 2.4016141457114372e-05,
|
||
|
|
"loss": 0.479,
|
||
|
|
"step": 608
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.465587044534413,
|
||
|
|
"grad_norm": 0.36586477669709805,
|
||
|
|
"learning_rate": 2.396072311715187e-05,
|
||
|
|
"loss": 0.4681,
|
||
|
|
"step": 609
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4696356275303644,
|
||
|
|
"grad_norm": 0.36553198190210434,
|
||
|
|
"learning_rate": 2.3905273107352863e-05,
|
||
|
|
"loss": 0.4639,
|
||
|
|
"step": 610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.473684210526316,
|
||
|
|
"grad_norm": 0.31806306388717664,
|
||
|
|
"learning_rate": 2.3849791871094157e-05,
|
||
|
|
"loss": 0.4759,
|
||
|
|
"step": 611
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.477732793522267,
|
||
|
|
"grad_norm": 0.3343846510914838,
|
||
|
|
"learning_rate": 2.379427985200223e-05,
|
||
|
|
"loss": 0.494,
|
||
|
|
"step": 612
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4817813765182186,
|
||
|
|
"grad_norm": 0.3509763594646804,
|
||
|
|
"learning_rate": 2.373873749394972e-05,
|
||
|
|
"loss": 0.4719,
|
||
|
|
"step": 613
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.48582995951417,
|
||
|
|
"grad_norm": 0.38163474874396014,
|
||
|
|
"learning_rate": 2.3683165241051836e-05,
|
||
|
|
"loss": 0.48,
|
||
|
|
"step": 614
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4898785425101213,
|
||
|
|
"grad_norm": 0.34888297907860066,
|
||
|
|
"learning_rate": 2.3627563537662823e-05,
|
||
|
|
"loss": 0.47,
|
||
|
|
"step": 615
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.493927125506073,
|
||
|
|
"grad_norm": 0.5043817163040952,
|
||
|
|
"learning_rate": 2.3571932828372433e-05,
|
||
|
|
"loss": 0.4701,
|
||
|
|
"step": 616
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4979757085020244,
|
||
|
|
"grad_norm": 0.31986595378205324,
|
||
|
|
"learning_rate": 2.351627355800233e-05,
|
||
|
|
"loss": 0.4578,
|
||
|
|
"step": 617
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5020242914979756,
|
||
|
|
"grad_norm": 0.44809394604156066,
|
||
|
|
"learning_rate": 2.346058617160256e-05,
|
||
|
|
"loss": 0.4724,
|
||
|
|
"step": 618
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.506072874493927,
|
||
|
|
"grad_norm": 0.3803406268645901,
|
||
|
|
"learning_rate": 2.3404871114447976e-05,
|
||
|
|
"loss": 0.4675,
|
||
|
|
"step": 619
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5101214574898787,
|
||
|
|
"grad_norm": 0.38727070679134573,
|
||
|
|
"learning_rate": 2.334912883203469e-05,
|
||
|
|
"loss": 0.4691,
|
||
|
|
"step": 620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.51417004048583,
|
||
|
|
"grad_norm": 0.4085269350736793,
|
||
|
|
"learning_rate": 2.3293359770076505e-05,
|
||
|
|
"loss": 0.4735,
|
||
|
|
"step": 621
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5182186234817814,
|
||
|
|
"grad_norm": 0.30185259471088327,
|
||
|
|
"learning_rate": 2.323756437450136e-05,
|
||
|
|
"loss": 0.4864,
|
||
|
|
"step": 622
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.522267206477733,
|
||
|
|
"grad_norm": 0.3894773932462322,
|
||
|
|
"learning_rate": 2.3181743091447754e-05,
|
||
|
|
"loss": 0.4577,
|
||
|
|
"step": 623
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.526315789473684,
|
||
|
|
"grad_norm": 0.32351716330654307,
|
||
|
|
"learning_rate": 2.3125896367261164e-05,
|
||
|
|
"loss": 0.4764,
|
||
|
|
"step": 624
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5303643724696356,
|
||
|
|
"grad_norm": 0.34173411462586567,
|
||
|
|
"learning_rate": 2.307002464849052e-05,
|
||
|
|
"loss": 0.4718,
|
||
|
|
"step": 625
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.534412955465587,
|
||
|
|
"grad_norm": 0.3516830125710817,
|
||
|
|
"learning_rate": 2.30141283818846e-05,
|
||
|
|
"loss": 0.4489,
|
||
|
|
"step": 626
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5384615384615383,
|
||
|
|
"grad_norm": 0.3289270653436247,
|
||
|
|
"learning_rate": 2.2958208014388462e-05,
|
||
|
|
"loss": 0.4603,
|
||
|
|
"step": 627
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.54251012145749,
|
||
|
|
"grad_norm": 0.3359630658424367,
|
||
|
|
"learning_rate": 2.2902263993139857e-05,
|
||
|
|
"loss": 0.4811,
|
||
|
|
"step": 628
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5465587044534415,
|
||
|
|
"grad_norm": 0.34174714321643945,
|
||
|
|
"learning_rate": 2.2846296765465708e-05,
|
||
|
|
"loss": 0.4679,
|
||
|
|
"step": 629
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5506072874493926,
|
||
|
|
"grad_norm": 0.30507686335405604,
|
||
|
|
"learning_rate": 2.2790306778878465e-05,
|
||
|
|
"loss": 0.4698,
|
||
|
|
"step": 630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.554655870445344,
|
||
|
|
"grad_norm": 0.3136405593559288,
|
||
|
|
"learning_rate": 2.273429448107257e-05,
|
||
|
|
"loss": 0.4813,
|
||
|
|
"step": 631
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5587044534412957,
|
||
|
|
"grad_norm": 0.332931904968984,
|
||
|
|
"learning_rate": 2.2678260319920868e-05,
|
||
|
|
"loss": 0.4629,
|
||
|
|
"step": 632
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.562753036437247,
|
||
|
|
"grad_norm": 0.311658805962014,
|
||
|
|
"learning_rate": 2.262220474347102e-05,
|
||
|
|
"loss": 0.4585,
|
||
|
|
"step": 633
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5668016194331984,
|
||
|
|
"grad_norm": 0.3117286819046009,
|
||
|
|
"learning_rate": 2.2566128199941918e-05,
|
||
|
|
"loss": 0.4712,
|
||
|
|
"step": 634
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.57085020242915,
|
||
|
|
"grad_norm": 0.36469265045480437,
|
||
|
|
"learning_rate": 2.2510031137720104e-05,
|
||
|
|
"loss": 0.4598,
|
||
|
|
"step": 635
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.574898785425101,
|
||
|
|
"grad_norm": 0.2839508147983164,
|
||
|
|
"learning_rate": 2.2453914005356206e-05,
|
||
|
|
"loss": 0.5082,
|
||
|
|
"step": 636
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5789473684210527,
|
||
|
|
"grad_norm": 0.3621309008665885,
|
||
|
|
"learning_rate": 2.2397777251561302e-05,
|
||
|
|
"loss": 0.4797,
|
||
|
|
"step": 637
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.582995951417004,
|
||
|
|
"grad_norm": 0.3233280989264786,
|
||
|
|
"learning_rate": 2.2341621325203393e-05,
|
||
|
|
"loss": 0.4907,
|
||
|
|
"step": 638
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5870445344129553,
|
||
|
|
"grad_norm": 0.32922250296269107,
|
||
|
|
"learning_rate": 2.2285446675303764e-05,
|
||
|
|
"loss": 0.471,
|
||
|
|
"step": 639
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.591093117408907,
|
||
|
|
"grad_norm": 0.3046512033855943,
|
||
|
|
"learning_rate": 2.2229253751033423e-05,
|
||
|
|
"loss": 0.4732,
|
||
|
|
"step": 640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5951417004048585,
|
||
|
|
"grad_norm": 0.2831733484243294,
|
||
|
|
"learning_rate": 2.2173043001709492e-05,
|
||
|
|
"loss": 0.478,
|
||
|
|
"step": 641
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.5991902834008096,
|
||
|
|
"grad_norm": 0.3504749377354086,
|
||
|
|
"learning_rate": 2.2116814876791627e-05,
|
||
|
|
"loss": 0.467,
|
||
|
|
"step": 642
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.603238866396761,
|
||
|
|
"grad_norm": 0.29875773004011347,
|
||
|
|
"learning_rate": 2.2060569825878413e-05,
|
||
|
|
"loss": 0.464,
|
||
|
|
"step": 643
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6072874493927127,
|
||
|
|
"grad_norm": 0.32165733067425256,
|
||
|
|
"learning_rate": 2.2004308298703777e-05,
|
||
|
|
"loss": 0.4876,
|
||
|
|
"step": 644
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.611336032388664,
|
||
|
|
"grad_norm": 0.3007203918461871,
|
||
|
|
"learning_rate": 2.1948030745133396e-05,
|
||
|
|
"loss": 0.4706,
|
||
|
|
"step": 645
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6153846153846154,
|
||
|
|
"grad_norm": 0.3135960643322974,
|
||
|
|
"learning_rate": 2.1891737615161083e-05,
|
||
|
|
"loss": 0.4626,
|
||
|
|
"step": 646
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.619433198380567,
|
||
|
|
"grad_norm": 0.32427753547893956,
|
||
|
|
"learning_rate": 2.1835429358905218e-05,
|
||
|
|
"loss": 0.4672,
|
||
|
|
"step": 647
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.623481781376518,
|
||
|
|
"grad_norm": 0.33476942408537713,
|
||
|
|
"learning_rate": 2.17791064266051e-05,
|
||
|
|
"loss": 0.4645,
|
||
|
|
"step": 648
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6275303643724697,
|
||
|
|
"grad_norm": 0.31649300179640893,
|
||
|
|
"learning_rate": 2.172276926861741e-05,
|
||
|
|
"loss": 0.4743,
|
||
|
|
"step": 649
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6315789473684212,
|
||
|
|
"grad_norm": 0.31768589472707026,
|
||
|
|
"learning_rate": 2.166641833541255e-05,
|
||
|
|
"loss": 0.468,
|
||
|
|
"step": 650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6356275303643724,
|
||
|
|
"grad_norm": 0.33028944430403434,
|
||
|
|
"learning_rate": 2.161005407757109e-05,
|
||
|
|
"loss": 0.4532,
|
||
|
|
"step": 651
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.639676113360324,
|
||
|
|
"grad_norm": 0.302483299095317,
|
||
|
|
"learning_rate": 2.155367694578013e-05,
|
||
|
|
"loss": 0.4819,
|
||
|
|
"step": 652
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6437246963562755,
|
||
|
|
"grad_norm": 0.3536366573491848,
|
||
|
|
"learning_rate": 2.1497287390829712e-05,
|
||
|
|
"loss": 0.477,
|
||
|
|
"step": 653
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6477732793522266,
|
||
|
|
"grad_norm": 0.30598265942096503,
|
||
|
|
"learning_rate": 2.1440885863609208e-05,
|
||
|
|
"loss": 0.4803,
|
||
|
|
"step": 654
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.651821862348178,
|
||
|
|
"grad_norm": 0.30999528367794305,
|
||
|
|
"learning_rate": 2.1384472815103737e-05,
|
||
|
|
"loss": 0.458,
|
||
|
|
"step": 655
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6558704453441297,
|
||
|
|
"grad_norm": 0.350195897974722,
|
||
|
|
"learning_rate": 2.1328048696390534e-05,
|
||
|
|
"loss": 0.4815,
|
||
|
|
"step": 656
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.659919028340081,
|
||
|
|
"grad_norm": 0.37060919417675453,
|
||
|
|
"learning_rate": 2.1271613958635342e-05,
|
||
|
|
"loss": 0.4867,
|
||
|
|
"step": 657
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6639676113360324,
|
||
|
|
"grad_norm": 0.36560319023462984,
|
||
|
|
"learning_rate": 2.1215169053088825e-05,
|
||
|
|
"loss": 0.4569,
|
||
|
|
"step": 658
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.668016194331984,
|
||
|
|
"grad_norm": 0.3314019334883962,
|
||
|
|
"learning_rate": 2.1158714431082937e-05,
|
||
|
|
"loss": 0.4758,
|
||
|
|
"step": 659
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.672064777327935,
|
||
|
|
"grad_norm": 0.46988782171655563,
|
||
|
|
"learning_rate": 2.110225054402733e-05,
|
||
|
|
"loss": 0.4712,
|
||
|
|
"step": 660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6761133603238867,
|
||
|
|
"grad_norm": 0.3410482456079,
|
||
|
|
"learning_rate": 2.1045777843405747e-05,
|
||
|
|
"loss": 0.4875,
|
||
|
|
"step": 661
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6801619433198383,
|
||
|
|
"grad_norm": 0.34162144130301464,
|
||
|
|
"learning_rate": 2.09892967807724e-05,
|
||
|
|
"loss": 0.4761,
|
||
|
|
"step": 662
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6842105263157894,
|
||
|
|
"grad_norm": 0.298288139149004,
|
||
|
|
"learning_rate": 2.093280780774835e-05,
|
||
|
|
"loss": 0.4768,
|
||
|
|
"step": 663
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.688259109311741,
|
||
|
|
"grad_norm": 0.3385298260242562,
|
||
|
|
"learning_rate": 2.087631137601793e-05,
|
||
|
|
"loss": 0.4754,
|
||
|
|
"step": 664
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6923076923076925,
|
||
|
|
"grad_norm": 0.2888870921061015,
|
||
|
|
"learning_rate": 2.0819807937325104e-05,
|
||
|
|
"loss": 0.464,
|
||
|
|
"step": 665
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6963562753036436,
|
||
|
|
"grad_norm": 0.34346896754454553,
|
||
|
|
"learning_rate": 2.0763297943469847e-05,
|
||
|
|
"loss": 0.4704,
|
||
|
|
"step": 666
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.700404858299595,
|
||
|
|
"grad_norm": 0.3038289102570644,
|
||
|
|
"learning_rate": 2.070678184630458e-05,
|
||
|
|
"loss": 0.4587,
|
||
|
|
"step": 667
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7044534412955468,
|
||
|
|
"grad_norm": 0.34339827099837295,
|
||
|
|
"learning_rate": 2.06502600977305e-05,
|
||
|
|
"loss": 0.4677,
|
||
|
|
"step": 668
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.708502024291498,
|
||
|
|
"grad_norm": 0.32201713444714464,
|
||
|
|
"learning_rate": 2.0593733149694008e-05,
|
||
|
|
"loss": 0.4618,
|
||
|
|
"step": 669
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7125506072874495,
|
||
|
|
"grad_norm": 0.36296637557038713,
|
||
|
|
"learning_rate": 2.0537201454183074e-05,
|
||
|
|
"loss": 0.4563,
|
||
|
|
"step": 670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.716599190283401,
|
||
|
|
"grad_norm": 0.3456821093327759,
|
||
|
|
"learning_rate": 2.0480665463223626e-05,
|
||
|
|
"loss": 0.4869,
|
||
|
|
"step": 671
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.720647773279352,
|
||
|
|
"grad_norm": 0.3272477673973111,
|
||
|
|
"learning_rate": 2.042412562887594e-05,
|
||
|
|
"loss": 0.4728,
|
||
|
|
"step": 672
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7246963562753037,
|
||
|
|
"grad_norm": 0.3866003204906215,
|
||
|
|
"learning_rate": 2.0367582403231033e-05,
|
||
|
|
"loss": 0.4723,
|
||
|
|
"step": 673
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7287449392712553,
|
||
|
|
"grad_norm": 0.3073713438860689,
|
||
|
|
"learning_rate": 2.0311036238407023e-05,
|
||
|
|
"loss": 0.4609,
|
||
|
|
"step": 674
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7327935222672064,
|
||
|
|
"grad_norm": 0.3922461608910711,
|
||
|
|
"learning_rate": 2.0254487586545534e-05,
|
||
|
|
"loss": 0.4887,
|
||
|
|
"step": 675
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.736842105263158,
|
||
|
|
"grad_norm": 0.27893091137961484,
|
||
|
|
"learning_rate": 2.0197936899808088e-05,
|
||
|
|
"loss": 0.4528,
|
||
|
|
"step": 676
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7408906882591095,
|
||
|
|
"grad_norm": 0.3678379085622595,
|
||
|
|
"learning_rate": 2.0141384630372466e-05,
|
||
|
|
"loss": 0.4748,
|
||
|
|
"step": 677
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7449392712550607,
|
||
|
|
"grad_norm": 0.3057727233857769,
|
||
|
|
"learning_rate": 2.0084831230429106e-05,
|
||
|
|
"loss": 0.4779,
|
||
|
|
"step": 678
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.748987854251012,
|
||
|
|
"grad_norm": 0.36732655096292677,
|
||
|
|
"learning_rate": 2.0028277152177492e-05,
|
||
|
|
"loss": 0.4965,
|
||
|
|
"step": 679
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.753036437246964,
|
||
|
|
"grad_norm": 0.33240567014777495,
|
||
|
|
"learning_rate": 1.9971722847822518e-05,
|
||
|
|
"loss": 0.4667,
|
||
|
|
"step": 680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.757085020242915,
|
||
|
|
"grad_norm": 0.3303802378860808,
|
||
|
|
"learning_rate": 1.99151687695709e-05,
|
||
|
|
"loss": 0.4688,
|
||
|
|
"step": 681
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7611336032388665,
|
||
|
|
"grad_norm": 0.37329682185170615,
|
||
|
|
"learning_rate": 1.9858615369627537e-05,
|
||
|
|
"loss": 0.4756,
|
||
|
|
"step": 682
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.765182186234818,
|
||
|
|
"grad_norm": 0.3850595273440119,
|
||
|
|
"learning_rate": 1.9802063100191916e-05,
|
||
|
|
"loss": 0.4567,
|
||
|
|
"step": 683
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.769230769230769,
|
||
|
|
"grad_norm": 0.3663096864871412,
|
||
|
|
"learning_rate": 1.974551241345447e-05,
|
||
|
|
"loss": 0.4661,
|
||
|
|
"step": 684
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7732793522267207,
|
||
|
|
"grad_norm": 0.3599165352209678,
|
||
|
|
"learning_rate": 1.9688963761592984e-05,
|
||
|
|
"loss": 0.4504,
|
||
|
|
"step": 685
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7773279352226723,
|
||
|
|
"grad_norm": 0.38821124338314106,
|
||
|
|
"learning_rate": 1.963241759676897e-05,
|
||
|
|
"loss": 0.4732,
|
||
|
|
"step": 686
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7813765182186234,
|
||
|
|
"grad_norm": 0.358350604687425,
|
||
|
|
"learning_rate": 1.9575874371124066e-05,
|
||
|
|
"loss": 0.4799,
|
||
|
|
"step": 687
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.785425101214575,
|
||
|
|
"grad_norm": 0.40827396803548827,
|
||
|
|
"learning_rate": 1.951933453677638e-05,
|
||
|
|
"loss": 0.4805,
|
||
|
|
"step": 688
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7894736842105265,
|
||
|
|
"grad_norm": 0.35373790989653736,
|
||
|
|
"learning_rate": 1.9462798545816932e-05,
|
||
|
|
"loss": 0.454,
|
||
|
|
"step": 689
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7935222672064777,
|
||
|
|
"grad_norm": 0.363751919325277,
|
||
|
|
"learning_rate": 1.9406266850305996e-05,
|
||
|
|
"loss": 0.4596,
|
||
|
|
"step": 690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7975708502024292,
|
||
|
|
"grad_norm": 0.33994129204769535,
|
||
|
|
"learning_rate": 1.9349739902269503e-05,
|
||
|
|
"loss": 0.481,
|
||
|
|
"step": 691
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.801619433198381,
|
||
|
|
"grad_norm": 0.34565772212921586,
|
||
|
|
"learning_rate": 1.9293218153695425e-05,
|
||
|
|
"loss": 0.4643,
|
||
|
|
"step": 692
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.805668016194332,
|
||
|
|
"grad_norm": 0.3109370153332686,
|
||
|
|
"learning_rate": 1.9236702056530153e-05,
|
||
|
|
"loss": 0.4778,
|
||
|
|
"step": 693
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8097165991902835,
|
||
|
|
"grad_norm": 0.3357299663575633,
|
||
|
|
"learning_rate": 1.9180192062674907e-05,
|
||
|
|
"loss": 0.4853,
|
||
|
|
"step": 694
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.813765182186235,
|
||
|
|
"grad_norm": 0.29742860144277455,
|
||
|
|
"learning_rate": 1.9123688623982076e-05,
|
||
|
|
"loss": 0.4704,
|
||
|
|
"step": 695
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.817813765182186,
|
||
|
|
"grad_norm": 0.35922528314987906,
|
||
|
|
"learning_rate": 1.906719219225165e-05,
|
||
|
|
"loss": 0.4758,
|
||
|
|
"step": 696
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8218623481781377,
|
||
|
|
"grad_norm": 0.32391005534615974,
|
||
|
|
"learning_rate": 1.9010703219227603e-05,
|
||
|
|
"loss": 0.4697,
|
||
|
|
"step": 697
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8259109311740893,
|
||
|
|
"grad_norm": 0.33210164499127826,
|
||
|
|
"learning_rate": 1.8954222156594256e-05,
|
||
|
|
"loss": 0.4752,
|
||
|
|
"step": 698
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8299595141700404,
|
||
|
|
"grad_norm": 0.31726860344691454,
|
||
|
|
"learning_rate": 1.8897749455972673e-05,
|
||
|
|
"loss": 0.4657,
|
||
|
|
"step": 699
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.834008097165992,
|
||
|
|
"grad_norm": 0.3306097933209381,
|
||
|
|
"learning_rate": 1.884128556891707e-05,
|
||
|
|
"loss": 0.4658,
|
||
|
|
"step": 700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8380566801619436,
|
||
|
|
"grad_norm": 0.3424920513114646,
|
||
|
|
"learning_rate": 1.878483094691119e-05,
|
||
|
|
"loss": 0.476,
|
||
|
|
"step": 701
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8421052631578947,
|
||
|
|
"grad_norm": 0.28904704909201984,
|
||
|
|
"learning_rate": 1.8728386041364664e-05,
|
||
|
|
"loss": 0.4692,
|
||
|
|
"step": 702
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8461538461538463,
|
||
|
|
"grad_norm": 0.3595502806011488,
|
||
|
|
"learning_rate": 1.867195130360947e-05,
|
||
|
|
"loss": 0.466,
|
||
|
|
"step": 703
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.850202429149798,
|
||
|
|
"grad_norm": 0.2730135535030052,
|
||
|
|
"learning_rate": 1.8615527184896263e-05,
|
||
|
|
"loss": 0.4747,
|
||
|
|
"step": 704
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.854251012145749,
|
||
|
|
"grad_norm": 0.33496161092061855,
|
||
|
|
"learning_rate": 1.8559114136390795e-05,
|
||
|
|
"loss": 0.4866,
|
||
|
|
"step": 705
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8582995951417005,
|
||
|
|
"grad_norm": 0.2877227589279155,
|
||
|
|
"learning_rate": 1.8502712609170298e-05,
|
||
|
|
"loss": 0.4789,
|
||
|
|
"step": 706
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.862348178137652,
|
||
|
|
"grad_norm": 0.29099491510428,
|
||
|
|
"learning_rate": 1.8446323054219876e-05,
|
||
|
|
"loss": 0.4855,
|
||
|
|
"step": 707
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.866396761133603,
|
||
|
|
"grad_norm": 0.29740243907780906,
|
||
|
|
"learning_rate": 1.838994592242891e-05,
|
||
|
|
"loss": 0.4663,
|
||
|
|
"step": 708
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8704453441295548,
|
||
|
|
"grad_norm": 0.29543102669769844,
|
||
|
|
"learning_rate": 1.8333581664587453e-05,
|
||
|
|
"loss": 0.4395,
|
||
|
|
"step": 709
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8744939271255063,
|
||
|
|
"grad_norm": 0.3204411434370082,
|
||
|
|
"learning_rate": 1.8277230731382593e-05,
|
||
|
|
"loss": 0.4738,
|
||
|
|
"step": 710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8785425101214575,
|
||
|
|
"grad_norm": 0.28886011765893554,
|
||
|
|
"learning_rate": 1.82208935733949e-05,
|
||
|
|
"loss": 0.4846,
|
||
|
|
"step": 711
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.882591093117409,
|
||
|
|
"grad_norm": 0.9417814970165729,
|
||
|
|
"learning_rate": 1.8164570641094793e-05,
|
||
|
|
"loss": 0.5128,
|
||
|
|
"step": 712
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8866396761133606,
|
||
|
|
"grad_norm": 0.27661003241316195,
|
||
|
|
"learning_rate": 1.810826238483892e-05,
|
||
|
|
"loss": 0.4742,
|
||
|
|
"step": 713
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8906882591093117,
|
||
|
|
"grad_norm": 0.3086585071380245,
|
||
|
|
"learning_rate": 1.8051969254866608e-05,
|
||
|
|
"loss": 0.4818,
|
||
|
|
"step": 714
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8947368421052633,
|
||
|
|
"grad_norm": 0.30078715055061295,
|
||
|
|
"learning_rate": 1.7995691701296226e-05,
|
||
|
|
"loss": 0.4446,
|
||
|
|
"step": 715
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.898785425101215,
|
||
|
|
"grad_norm": 0.2871803436954661,
|
||
|
|
"learning_rate": 1.7939430174121593e-05,
|
||
|
|
"loss": 0.4532,
|
||
|
|
"step": 716
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.902834008097166,
|
||
|
|
"grad_norm": 0.3138947010184722,
|
||
|
|
"learning_rate": 1.788318512320838e-05,
|
||
|
|
"loss": 0.4508,
|
||
|
|
"step": 717
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9068825910931175,
|
||
|
|
"grad_norm": 0.30361044615173394,
|
||
|
|
"learning_rate": 1.782695699829051e-05,
|
||
|
|
"loss": 0.4843,
|
||
|
|
"step": 718
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.910931174089069,
|
||
|
|
"grad_norm": 0.2808852401394827,
|
||
|
|
"learning_rate": 1.7770746248966587e-05,
|
||
|
|
"loss": 0.4582,
|
||
|
|
"step": 719
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.91497975708502,
|
||
|
|
"grad_norm": 0.3083753302687146,
|
||
|
|
"learning_rate": 1.7714553324696243e-05,
|
||
|
|
"loss": 0.475,
|
||
|
|
"step": 720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.919028340080972,
|
||
|
|
"grad_norm": 0.3132897444909797,
|
||
|
|
"learning_rate": 1.7658378674796614e-05,
|
||
|
|
"loss": 0.469,
|
||
|
|
"step": 721
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9230769230769234,
|
||
|
|
"grad_norm": 0.36205543815132457,
|
||
|
|
"learning_rate": 1.76022227484387e-05,
|
||
|
|
"loss": 0.4875,
|
||
|
|
"step": 722
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9271255060728745,
|
||
|
|
"grad_norm": 0.31189433187662546,
|
||
|
|
"learning_rate": 1.7546085994643807e-05,
|
||
|
|
"loss": 0.4628,
|
||
|
|
"step": 723
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.931174089068826,
|
||
|
|
"grad_norm": 0.2874349653410532,
|
||
|
|
"learning_rate": 1.7489968862279902e-05,
|
||
|
|
"loss": 0.4773,
|
||
|
|
"step": 724
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9352226720647776,
|
||
|
|
"grad_norm": 0.29349611573681844,
|
||
|
|
"learning_rate": 1.743387180005809e-05,
|
||
|
|
"loss": 0.4726,
|
||
|
|
"step": 725
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9392712550607287,
|
||
|
|
"grad_norm": 0.29408824436113395,
|
||
|
|
"learning_rate": 1.737779525652899e-05,
|
||
|
|
"loss": 0.4672,
|
||
|
|
"step": 726
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9433198380566803,
|
||
|
|
"grad_norm": 0.35048544272501203,
|
||
|
|
"learning_rate": 1.732173968007914e-05,
|
||
|
|
"loss": 0.4683,
|
||
|
|
"step": 727
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9473684210526314,
|
||
|
|
"grad_norm": 0.29914971824174563,
|
||
|
|
"learning_rate": 1.7265705518927437e-05,
|
||
|
|
"loss": 0.4802,
|
||
|
|
"step": 728
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.951417004048583,
|
||
|
|
"grad_norm": 0.3410707511238311,
|
||
|
|
"learning_rate": 1.7209693221121542e-05,
|
||
|
|
"loss": 0.4601,
|
||
|
|
"step": 729
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9554655870445345,
|
||
|
|
"grad_norm": 0.33866593860371036,
|
||
|
|
"learning_rate": 1.7153703234534302e-05,
|
||
|
|
"loss": 0.4719,
|
||
|
|
"step": 730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9595141700404857,
|
||
|
|
"grad_norm": 0.31519873092215756,
|
||
|
|
"learning_rate": 1.709773600686015e-05,
|
||
|
|
"loss": 0.4903,
|
||
|
|
"step": 731
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9635627530364372,
|
||
|
|
"grad_norm": 0.3119387947740441,
|
||
|
|
"learning_rate": 1.7041791985611548e-05,
|
||
|
|
"loss": 0.4584,
|
||
|
|
"step": 732
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.967611336032389,
|
||
|
|
"grad_norm": 0.344297030115303,
|
||
|
|
"learning_rate": 1.69858716181154e-05,
|
||
|
|
"loss": 0.4544,
|
||
|
|
"step": 733
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.97165991902834,
|
||
|
|
"grad_norm": 0.27305573370869207,
|
||
|
|
"learning_rate": 1.692997535150948e-05,
|
||
|
|
"loss": 0.4801,
|
||
|
|
"step": 734
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9757085020242915,
|
||
|
|
"grad_norm": 0.30897170847665006,
|
||
|
|
"learning_rate": 1.687410363273884e-05,
|
||
|
|
"loss": 0.4801,
|
||
|
|
"step": 735
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.979757085020243,
|
||
|
|
"grad_norm": 0.3323742547274772,
|
||
|
|
"learning_rate": 1.6818256908552257e-05,
|
||
|
|
"loss": 0.4516,
|
||
|
|
"step": 736
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.983805668016194,
|
||
|
|
"grad_norm": 0.2855963734008308,
|
||
|
|
"learning_rate": 1.6762435625498646e-05,
|
||
|
|
"loss": 0.4433,
|
||
|
|
"step": 737
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9878542510121457,
|
||
|
|
"grad_norm": 0.2742376352429609,
|
||
|
|
"learning_rate": 1.67066402299235e-05,
|
||
|
|
"loss": 0.4622,
|
||
|
|
"step": 738
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9919028340080973,
|
||
|
|
"grad_norm": 0.32590838548268597,
|
||
|
|
"learning_rate": 1.6650871167965313e-05,
|
||
|
|
"loss": 0.4532,
|
||
|
|
"step": 739
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9959514170040484,
|
||
|
|
"grad_norm": 0.29238168793636055,
|
||
|
|
"learning_rate": 1.6595128885552028e-05,
|
||
|
|
"loss": 0.4704,
|
||
|
|
"step": 740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0,
|
||
|
|
"grad_norm": 0.3063776998072053,
|
||
|
|
"learning_rate": 1.653941382839745e-05,
|
||
|
|
"loss": 0.4558,
|
||
|
|
"step": 741
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0040485829959516,
|
||
|
|
"grad_norm": 0.47020268979759683,
|
||
|
|
"learning_rate": 1.6483726441997673e-05,
|
||
|
|
"loss": 0.4145,
|
||
|
|
"step": 742
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0080971659919027,
|
||
|
|
"grad_norm": 0.35013469844452316,
|
||
|
|
"learning_rate": 1.642806717162757e-05,
|
||
|
|
"loss": 0.3894,
|
||
|
|
"step": 743
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0121457489878543,
|
||
|
|
"grad_norm": 0.6038473772462262,
|
||
|
|
"learning_rate": 1.637243646233718e-05,
|
||
|
|
"loss": 0.3944,
|
||
|
|
"step": 744
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.016194331983806,
|
||
|
|
"grad_norm": 0.5928692805303908,
|
||
|
|
"learning_rate": 1.6316834758948174e-05,
|
||
|
|
"loss": 0.4361,
|
||
|
|
"step": 745
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.020242914979757,
|
||
|
|
"grad_norm": 0.467563311823661,
|
||
|
|
"learning_rate": 1.6261262506050282e-05,
|
||
|
|
"loss": 0.4121,
|
||
|
|
"step": 746
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0242914979757085,
|
||
|
|
"grad_norm": 0.35909206816783784,
|
||
|
|
"learning_rate": 1.620572014799777e-05,
|
||
|
|
"loss": 0.4079,
|
||
|
|
"step": 747
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.02834008097166,
|
||
|
|
"grad_norm": 0.4176368143252143,
|
||
|
|
"learning_rate": 1.6150208128905857e-05,
|
||
|
|
"loss": 0.409,
|
||
|
|
"step": 748
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.032388663967611,
|
||
|
|
"grad_norm": 0.4340286108130099,
|
||
|
|
"learning_rate": 1.6094726892647147e-05,
|
||
|
|
"loss": 0.3952,
|
||
|
|
"step": 749
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0364372469635628,
|
||
|
|
"grad_norm": 0.39798184382786383,
|
||
|
|
"learning_rate": 1.6039276882848135e-05,
|
||
|
|
"loss": 0.3916,
|
||
|
|
"step": 750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0404858299595143,
|
||
|
|
"grad_norm": 0.4211512391371717,
|
||
|
|
"learning_rate": 1.5983858542885635e-05,
|
||
|
|
"loss": 0.3825,
|
||
|
|
"step": 751
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0445344129554655,
|
||
|
|
"grad_norm": 0.36513888965892505,
|
||
|
|
"learning_rate": 1.5928472315883203e-05,
|
||
|
|
"loss": 0.4021,
|
||
|
|
"step": 752
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.048582995951417,
|
||
|
|
"grad_norm": 0.3828145516699195,
|
||
|
|
"learning_rate": 1.5873118644707633e-05,
|
||
|
|
"loss": 0.4034,
|
||
|
|
"step": 753
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0526315789473686,
|
||
|
|
"grad_norm": 0.3625023427209613,
|
||
|
|
"learning_rate": 1.5817797971965413e-05,
|
||
|
|
"loss": 0.3802,
|
||
|
|
"step": 754
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0566801619433197,
|
||
|
|
"grad_norm": 0.3307646942092969,
|
||
|
|
"learning_rate": 1.576251073999917e-05,
|
||
|
|
"loss": 0.3992,
|
||
|
|
"step": 755
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0607287449392713,
|
||
|
|
"grad_norm": 0.3764630940923384,
|
||
|
|
"learning_rate": 1.5707257390884126e-05,
|
||
|
|
"loss": 0.3819,
|
||
|
|
"step": 756
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.064777327935223,
|
||
|
|
"grad_norm": 0.3536487231964897,
|
||
|
|
"learning_rate": 1.5652038366424595e-05,
|
||
|
|
"loss": 0.3835,
|
||
|
|
"step": 757
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.068825910931174,
|
||
|
|
"grad_norm": 0.341091257672784,
|
||
|
|
"learning_rate": 1.5596854108150424e-05,
|
||
|
|
"loss": 0.3882,
|
||
|
|
"step": 758
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0728744939271255,
|
||
|
|
"grad_norm": 0.4418188726507605,
|
||
|
|
"learning_rate": 1.5541705057313476e-05,
|
||
|
|
"loss": 0.4071,
|
||
|
|
"step": 759
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.076923076923077,
|
||
|
|
"grad_norm": 0.33226439855907913,
|
||
|
|
"learning_rate": 1.5486591654884086e-05,
|
||
|
|
"loss": 0.4138,
|
||
|
|
"step": 760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.080971659919028,
|
||
|
|
"grad_norm": 0.34958945516212986,
|
||
|
|
"learning_rate": 1.543151434154755e-05,
|
||
|
|
"loss": 0.4037,
|
||
|
|
"step": 761
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.08502024291498,
|
||
|
|
"grad_norm": 0.3316440411170271,
|
||
|
|
"learning_rate": 1.5376473557700615e-05,
|
||
|
|
"loss": 0.4006,
|
||
|
|
"step": 762
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0890688259109313,
|
||
|
|
"grad_norm": 0.32339105456377293,
|
||
|
|
"learning_rate": 1.5321469743447893e-05,
|
||
|
|
"loss": 0.3926,
|
||
|
|
"step": 763
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0931174089068825,
|
||
|
|
"grad_norm": 0.2965961077665869,
|
||
|
|
"learning_rate": 1.5266503338598434e-05,
|
||
|
|
"loss": 0.4211,
|
||
|
|
"step": 764
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.097165991902834,
|
||
|
|
"grad_norm": 0.2934465581260307,
|
||
|
|
"learning_rate": 1.5211574782662149e-05,
|
||
|
|
"loss": 0.4107,
|
||
|
|
"step": 765
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1012145748987856,
|
||
|
|
"grad_norm": 0.30748642889930955,
|
||
|
|
"learning_rate": 1.5156684514846307e-05,
|
||
|
|
"loss": 0.3854,
|
||
|
|
"step": 766
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1052631578947367,
|
||
|
|
"grad_norm": 0.2857971484072039,
|
||
|
|
"learning_rate": 1.5101832974052012e-05,
|
||
|
|
"loss": 0.4037,
|
||
|
|
"step": 767
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1093117408906883,
|
||
|
|
"grad_norm": 0.3036256448093267,
|
||
|
|
"learning_rate": 1.5047020598870737e-05,
|
||
|
|
"loss": 0.3925,
|
||
|
|
"step": 768
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.11336032388664,
|
||
|
|
"grad_norm": 0.3044484387216606,
|
||
|
|
"learning_rate": 1.4992247827580778e-05,
|
||
|
|
"loss": 0.3882,
|
||
|
|
"step": 769
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.117408906882591,
|
||
|
|
"grad_norm": 0.2820156848268788,
|
||
|
|
"learning_rate": 1.493751509814374e-05,
|
||
|
|
"loss": 0.4064,
|
||
|
|
"step": 770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1214574898785425,
|
||
|
|
"grad_norm": 0.2964875179324039,
|
||
|
|
"learning_rate": 1.4882822848201073e-05,
|
||
|
|
"loss": 0.4025,
|
||
|
|
"step": 771
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.125506072874494,
|
||
|
|
"grad_norm": 0.2639983591779373,
|
||
|
|
"learning_rate": 1.4828171515070553e-05,
|
||
|
|
"loss": 0.4315,
|
||
|
|
"step": 772
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1295546558704452,
|
||
|
|
"grad_norm": 0.2737789307748836,
|
||
|
|
"learning_rate": 1.4773561535742793e-05,
|
||
|
|
"loss": 0.3868,
|
||
|
|
"step": 773
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.133603238866397,
|
||
|
|
"grad_norm": 0.2895731291526676,
|
||
|
|
"learning_rate": 1.4718993346877715e-05,
|
||
|
|
"loss": 0.3979,
|
||
|
|
"step": 774
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1376518218623484,
|
||
|
|
"grad_norm": 0.26142750104972096,
|
||
|
|
"learning_rate": 1.466446738480111e-05,
|
||
|
|
"loss": 0.4104,
|
||
|
|
"step": 775
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1417004048582995,
|
||
|
|
"grad_norm": 0.26481475798300586,
|
||
|
|
"learning_rate": 1.460998408550112e-05,
|
||
|
|
"loss": 0.3934,
|
||
|
|
"step": 776
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.145748987854251,
|
||
|
|
"grad_norm": 0.2547449336718836,
|
||
|
|
"learning_rate": 1.4555543884624751e-05,
|
||
|
|
"loss": 0.4024,
|
||
|
|
"step": 777
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1497975708502026,
|
||
|
|
"grad_norm": 0.2906738492510942,
|
||
|
|
"learning_rate": 1.4501147217474402e-05,
|
||
|
|
"loss": 0.4004,
|
||
|
|
"step": 778
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1538461538461537,
|
||
|
|
"grad_norm": 0.2694246620778903,
|
||
|
|
"learning_rate": 1.444679451900437e-05,
|
||
|
|
"loss": 0.4176,
|
||
|
|
"step": 779
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1578947368421053,
|
||
|
|
"grad_norm": 0.2714479652863721,
|
||
|
|
"learning_rate": 1.4392486223817397e-05,
|
||
|
|
"loss": 0.3982,
|
||
|
|
"step": 780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.161943319838057,
|
||
|
|
"grad_norm": 0.28778595555389214,
|
||
|
|
"learning_rate": 1.4338222766161143e-05,
|
||
|
|
"loss": 0.3952,
|
||
|
|
"step": 781
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.165991902834008,
|
||
|
|
"grad_norm": 0.2691997521461367,
|
||
|
|
"learning_rate": 1.4284004579924777e-05,
|
||
|
|
"loss": 0.3992,
|
||
|
|
"step": 782
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1700404858299596,
|
||
|
|
"grad_norm": 0.2559117840207078,
|
||
|
|
"learning_rate": 1.4229832098635479e-05,
|
||
|
|
"loss": 0.402,
|
||
|
|
"step": 783
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.174089068825911,
|
||
|
|
"grad_norm": 0.28294512331136434,
|
||
|
|
"learning_rate": 1.4175705755454963e-05,
|
||
|
|
"loss": 0.4207,
|
||
|
|
"step": 784
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1781376518218623,
|
||
|
|
"grad_norm": 0.28035673170635383,
|
||
|
|
"learning_rate": 1.412162598317602e-05,
|
||
|
|
"loss": 0.3993,
|
||
|
|
"step": 785
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.182186234817814,
|
||
|
|
"grad_norm": 0.27910519069089856,
|
||
|
|
"learning_rate": 1.406759321421907e-05,
|
||
|
|
"loss": 0.3799,
|
||
|
|
"step": 786
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1862348178137654,
|
||
|
|
"grad_norm": 0.2673901381311783,
|
||
|
|
"learning_rate": 1.401360788062871e-05,
|
||
|
|
"loss": 0.3924,
|
||
|
|
"step": 787
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1902834008097165,
|
||
|
|
"grad_norm": 0.2884832420537452,
|
||
|
|
"learning_rate": 1.3959670414070204e-05,
|
||
|
|
"loss": 0.3947,
|
||
|
|
"step": 788
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.194331983805668,
|
||
|
|
"grad_norm": 0.2788663496080745,
|
||
|
|
"learning_rate": 1.3905781245826108e-05,
|
||
|
|
"loss": 0.3833,
|
||
|
|
"step": 789
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1983805668016196,
|
||
|
|
"grad_norm": 0.35329624051570796,
|
||
|
|
"learning_rate": 1.3851940806792778e-05,
|
||
|
|
"loss": 0.4058,
|
||
|
|
"step": 790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2024291497975708,
|
||
|
|
"grad_norm": 0.2882563916985464,
|
||
|
|
"learning_rate": 1.379814952747693e-05,
|
||
|
|
"loss": 0.3767,
|
||
|
|
"step": 791
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2064777327935223,
|
||
|
|
"grad_norm": 0.2856549012844441,
|
||
|
|
"learning_rate": 1.3744407837992193e-05,
|
||
|
|
"loss": 0.4097,
|
||
|
|
"step": 792
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2105263157894735,
|
||
|
|
"grad_norm": 0.26966913831050154,
|
||
|
|
"learning_rate": 1.3690716168055692e-05,
|
||
|
|
"loss": 0.4014,
|
||
|
|
"step": 793
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.214574898785425,
|
||
|
|
"grad_norm": 0.31155437315143175,
|
||
|
|
"learning_rate": 1.363707494698459e-05,
|
||
|
|
"loss": 0.3754,
|
||
|
|
"step": 794
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2186234817813766,
|
||
|
|
"grad_norm": 0.27011940396193224,
|
||
|
|
"learning_rate": 1.358348460369265e-05,
|
||
|
|
"loss": 0.4053,
|
||
|
|
"step": 795
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2226720647773277,
|
||
|
|
"grad_norm": 0.2864982194889137,
|
||
|
|
"learning_rate": 1.352994556668684e-05,
|
||
|
|
"loss": 0.3932,
|
||
|
|
"step": 796
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2267206477732793,
|
||
|
|
"grad_norm": 0.25478520159468765,
|
||
|
|
"learning_rate": 1.3476458264063865e-05,
|
||
|
|
"loss": 0.3993,
|
||
|
|
"step": 797
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.230769230769231,
|
||
|
|
"grad_norm": 0.26975633302096436,
|
||
|
|
"learning_rate": 1.342302312350678e-05,
|
||
|
|
"loss": 0.4102,
|
||
|
|
"step": 798
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.234817813765182,
|
||
|
|
"grad_norm": 0.26810245424854356,
|
||
|
|
"learning_rate": 1.3369640572281537e-05,
|
||
|
|
"loss": 0.4061,
|
||
|
|
"step": 799
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2388663967611335,
|
||
|
|
"grad_norm": 0.2640966133451815,
|
||
|
|
"learning_rate": 1.3316311037233596e-05,
|
||
|
|
"loss": 0.4124,
|
||
|
|
"step": 800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.242914979757085,
|
||
|
|
"grad_norm": 0.27329588045000036,
|
||
|
|
"learning_rate": 1.326303494478451e-05,
|
||
|
|
"loss": 0.4187,
|
||
|
|
"step": 801
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.246963562753036,
|
||
|
|
"grad_norm": 0.270120015801755,
|
||
|
|
"learning_rate": 1.3209812720928472e-05,
|
||
|
|
"loss": 0.4038,
|
||
|
|
"step": 802
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.251012145748988,
|
||
|
|
"grad_norm": 0.25361627090680017,
|
||
|
|
"learning_rate": 1.315664479122898e-05,
|
||
|
|
"loss": 0.4036,
|
||
|
|
"step": 803
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2550607287449393,
|
||
|
|
"grad_norm": 0.27696163182826794,
|
||
|
|
"learning_rate": 1.3103531580815378e-05,
|
||
|
|
"loss": 0.3836,
|
||
|
|
"step": 804
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2591093117408905,
|
||
|
|
"grad_norm": 0.2741932454538595,
|
||
|
|
"learning_rate": 1.3050473514379488e-05,
|
||
|
|
"loss": 0.3911,
|
||
|
|
"step": 805
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.263157894736842,
|
||
|
|
"grad_norm": 0.2940740189934853,
|
||
|
|
"learning_rate": 1.2997471016172188e-05,
|
||
|
|
"loss": 0.3971,
|
||
|
|
"step": 806
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2672064777327936,
|
||
|
|
"grad_norm": 0.2806074580169743,
|
||
|
|
"learning_rate": 1.2944524510000042e-05,
|
||
|
|
"loss": 0.3678,
|
||
|
|
"step": 807
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2712550607287447,
|
||
|
|
"grad_norm": 0.29640537891218965,
|
||
|
|
"learning_rate": 1.2891634419221908e-05,
|
||
|
|
"loss": 0.3799,
|
||
|
|
"step": 808
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2753036437246963,
|
||
|
|
"grad_norm": 0.2872446355915579,
|
||
|
|
"learning_rate": 1.2838801166745545e-05,
|
||
|
|
"loss": 0.3976,
|
||
|
|
"step": 809
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.279352226720648,
|
||
|
|
"grad_norm": 0.29459332147371187,
|
||
|
|
"learning_rate": 1.2786025175024216e-05,
|
||
|
|
"loss": 0.4159,
|
||
|
|
"step": 810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.283400809716599,
|
||
|
|
"grad_norm": 0.30407622562278297,
|
||
|
|
"learning_rate": 1.2733306866053357e-05,
|
||
|
|
"loss": 0.3765,
|
||
|
|
"step": 811
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2874493927125505,
|
||
|
|
"grad_norm": 0.2848231519218241,
|
||
|
|
"learning_rate": 1.2680646661367163e-05,
|
||
|
|
"loss": 0.4107,
|
||
|
|
"step": 812
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.291497975708502,
|
||
|
|
"grad_norm": 0.2997670595564209,
|
||
|
|
"learning_rate": 1.2628044982035219e-05,
|
||
|
|
"loss": 0.3982,
|
||
|
|
"step": 813
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2955465587044532,
|
||
|
|
"grad_norm": 0.28633408701217816,
|
||
|
|
"learning_rate": 1.2575502248659155e-05,
|
||
|
|
"loss": 0.3932,
|
||
|
|
"step": 814
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.299595141700405,
|
||
|
|
"grad_norm": 0.2813981315800815,
|
||
|
|
"learning_rate": 1.2523018881369269e-05,
|
||
|
|
"loss": 0.4075,
|
||
|
|
"step": 815
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3036437246963564,
|
||
|
|
"grad_norm": 0.2881658757982672,
|
||
|
|
"learning_rate": 1.2470595299821176e-05,
|
||
|
|
"loss": 0.4075,
|
||
|
|
"step": 816
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3076923076923075,
|
||
|
|
"grad_norm": 0.27549271302449885,
|
||
|
|
"learning_rate": 1.2418231923192416e-05,
|
||
|
|
"loss": 0.374,
|
||
|
|
"step": 817
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.311740890688259,
|
||
|
|
"grad_norm": 0.2995541806438367,
|
||
|
|
"learning_rate": 1.2365929170179165e-05,
|
||
|
|
"loss": 0.4074,
|
||
|
|
"step": 818
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3157894736842106,
|
||
|
|
"grad_norm": 0.2710618269456132,
|
||
|
|
"learning_rate": 1.2313687458992848e-05,
|
||
|
|
"loss": 0.3898,
|
||
|
|
"step": 819
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3198380566801617,
|
||
|
|
"grad_norm": 0.2903411492108107,
|
||
|
|
"learning_rate": 1.2261507207356785e-05,
|
||
|
|
"loss": 0.4019,
|
||
|
|
"step": 820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3238866396761133,
|
||
|
|
"grad_norm": 0.27852378785558235,
|
||
|
|
"learning_rate": 1.220938883250288e-05,
|
||
|
|
"loss": 0.4107,
|
||
|
|
"step": 821
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.327935222672065,
|
||
|
|
"grad_norm": 0.273773599420098,
|
||
|
|
"learning_rate": 1.215733275116827e-05,
|
||
|
|
"loss": 0.4021,
|
||
|
|
"step": 822
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.331983805668016,
|
||
|
|
"grad_norm": 0.2775478592107353,
|
||
|
|
"learning_rate": 1.2105339379592005e-05,
|
||
|
|
"loss": 0.4037,
|
||
|
|
"step": 823
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3360323886639676,
|
||
|
|
"grad_norm": 0.2589429630173823,
|
||
|
|
"learning_rate": 1.2053409133511681e-05,
|
||
|
|
"loss": 0.3831,
|
||
|
|
"step": 824
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.340080971659919,
|
||
|
|
"grad_norm": 0.2765149231633141,
|
||
|
|
"learning_rate": 1.2001542428160173e-05,
|
||
|
|
"loss": 0.412,
|
||
|
|
"step": 825
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3441295546558703,
|
||
|
|
"grad_norm": 0.28328996539229645,
|
||
|
|
"learning_rate": 1.1949739678262278e-05,
|
||
|
|
"loss": 0.3993,
|
||
|
|
"step": 826
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.348178137651822,
|
||
|
|
"grad_norm": 0.26801889221001546,
|
||
|
|
"learning_rate": 1.189800129803139e-05,
|
||
|
|
"loss": 0.4167,
|
||
|
|
"step": 827
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3522267206477734,
|
||
|
|
"grad_norm": 0.2781858550205906,
|
||
|
|
"learning_rate": 1.1846327701166228e-05,
|
||
|
|
"loss": 0.3948,
|
||
|
|
"step": 828
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3562753036437245,
|
||
|
|
"grad_norm": 0.2819650988515181,
|
||
|
|
"learning_rate": 1.1794719300847493e-05,
|
||
|
|
"loss": 0.3961,
|
||
|
|
"step": 829
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.360323886639676,
|
||
|
|
"grad_norm": 0.2691519302094065,
|
||
|
|
"learning_rate": 1.1743176509734582e-05,
|
||
|
|
"loss": 0.4062,
|
||
|
|
"step": 830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3643724696356276,
|
||
|
|
"grad_norm": 0.29610933766207376,
|
||
|
|
"learning_rate": 1.1691699739962275e-05,
|
||
|
|
"loss": 0.41,
|
||
|
|
"step": 831
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3684210526315788,
|
||
|
|
"grad_norm": 0.276292334108597,
|
||
|
|
"learning_rate": 1.1640289403137456e-05,
|
||
|
|
"loss": 0.3984,
|
||
|
|
"step": 832
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3724696356275303,
|
||
|
|
"grad_norm": 0.27321465201751177,
|
||
|
|
"learning_rate": 1.1588945910335802e-05,
|
||
|
|
"loss": 0.4019,
|
||
|
|
"step": 833
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.376518218623482,
|
||
|
|
"grad_norm": 0.29173995779685047,
|
||
|
|
"learning_rate": 1.1537669672098517e-05,
|
||
|
|
"loss": 0.3922,
|
||
|
|
"step": 834
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.380566801619433,
|
||
|
|
"grad_norm": 0.2563678019392041,
|
||
|
|
"learning_rate": 1.1486461098429036e-05,
|
||
|
|
"loss": 0.4198,
|
||
|
|
"step": 835
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3846153846153846,
|
||
|
|
"grad_norm": 0.2603386173751867,
|
||
|
|
"learning_rate": 1.1435320598789747e-05,
|
||
|
|
"loss": 0.4096,
|
||
|
|
"step": 836
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.388663967611336,
|
||
|
|
"grad_norm": 0.29885623995862615,
|
||
|
|
"learning_rate": 1.1384248582098736e-05,
|
||
|
|
"loss": 0.403,
|
||
|
|
"step": 837
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.3927125506072873,
|
||
|
|
"grad_norm": 0.26228948582261363,
|
||
|
|
"learning_rate": 1.1333245456726467e-05,
|
||
|
|
"loss": 0.4015,
|
||
|
|
"step": 838
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.396761133603239,
|
||
|
|
"grad_norm": 0.2707014181849456,
|
||
|
|
"learning_rate": 1.128231163049258e-05,
|
||
|
|
"loss": 0.3957,
|
||
|
|
"step": 839
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4008097165991904,
|
||
|
|
"grad_norm": 0.28603557246858824,
|
||
|
|
"learning_rate": 1.123144751066259e-05,
|
||
|
|
"loss": 0.4027,
|
||
|
|
"step": 840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4048582995951415,
|
||
|
|
"grad_norm": 0.267053574727869,
|
||
|
|
"learning_rate": 1.1180653503944657e-05,
|
||
|
|
"loss": 0.3793,
|
||
|
|
"step": 841
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.408906882591093,
|
||
|
|
"grad_norm": 0.26974157313035957,
|
||
|
|
"learning_rate": 1.1129930016486287e-05,
|
||
|
|
"loss": 0.397,
|
||
|
|
"step": 842
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4129554655870447,
|
||
|
|
"grad_norm": 0.2697248631250802,
|
||
|
|
"learning_rate": 1.107927745387114e-05,
|
||
|
|
"loss": 0.4009,
|
||
|
|
"step": 843
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.417004048582996,
|
||
|
|
"grad_norm": 0.2773343238615587,
|
||
|
|
"learning_rate": 1.1028696221115766e-05,
|
||
|
|
"loss": 0.3693,
|
||
|
|
"step": 844
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4210526315789473,
|
||
|
|
"grad_norm": 0.2702112505643063,
|
||
|
|
"learning_rate": 1.0978186722666341e-05,
|
||
|
|
"loss": 0.397,
|
||
|
|
"step": 845
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.425101214574899,
|
||
|
|
"grad_norm": 0.2679778046710898,
|
||
|
|
"learning_rate": 1.0927749362395475e-05,
|
||
|
|
"loss": 0.4073,
|
||
|
|
"step": 846
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.42914979757085,
|
||
|
|
"grad_norm": 0.28056735226034046,
|
||
|
|
"learning_rate": 1.0877384543598954e-05,
|
||
|
|
"loss": 0.4185,
|
||
|
|
"step": 847
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4331983805668016,
|
||
|
|
"grad_norm": 0.25814114028190616,
|
||
|
|
"learning_rate": 1.0827092668992529e-05,
|
||
|
|
"loss": 0.414,
|
||
|
|
"step": 848
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.437246963562753,
|
||
|
|
"grad_norm": 0.2701765872611739,
|
||
|
|
"learning_rate": 1.077687414070869e-05,
|
||
|
|
"loss": 0.3893,
|
||
|
|
"step": 849
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4412955465587043,
|
||
|
|
"grad_norm": 0.2587460379987047,
|
||
|
|
"learning_rate": 1.0726729360293451e-05,
|
||
|
|
"loss": 0.3827,
|
||
|
|
"step": 850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.445344129554656,
|
||
|
|
"grad_norm": 0.2588484943755391,
|
||
|
|
"learning_rate": 1.0676658728703147e-05,
|
||
|
|
"loss": 0.4035,
|
||
|
|
"step": 851
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4493927125506074,
|
||
|
|
"grad_norm": 0.2846609630093237,
|
||
|
|
"learning_rate": 1.0626662646301198e-05,
|
||
|
|
"loss": 0.3918,
|
||
|
|
"step": 852
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4534412955465585,
|
||
|
|
"grad_norm": 0.24811677295859325,
|
||
|
|
"learning_rate": 1.0576741512854958e-05,
|
||
|
|
"loss": 0.4099,
|
||
|
|
"step": 853
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.45748987854251,
|
||
|
|
"grad_norm": 0.27466256426218233,
|
||
|
|
"learning_rate": 1.052689572753247e-05,
|
||
|
|
"loss": 0.4111,
|
||
|
|
"step": 854
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4615384615384617,
|
||
|
|
"grad_norm": 0.2888514039023773,
|
||
|
|
"learning_rate": 1.0477125688899328e-05,
|
||
|
|
"loss": 0.3946,
|
||
|
|
"step": 855
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.465587044534413,
|
||
|
|
"grad_norm": 0.2850114987091914,
|
||
|
|
"learning_rate": 1.0427431794915406e-05,
|
||
|
|
"loss": 0.4205,
|
||
|
|
"step": 856
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4696356275303644,
|
||
|
|
"grad_norm": 0.2764173276711337,
|
||
|
|
"learning_rate": 1.0377814442931769e-05,
|
||
|
|
"loss": 0.4037,
|
||
|
|
"step": 857
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.473684210526316,
|
||
|
|
"grad_norm": 0.25564512891009095,
|
||
|
|
"learning_rate": 1.0328274029687439e-05,
|
||
|
|
"loss": 0.3926,
|
||
|
|
"step": 858
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.477732793522267,
|
||
|
|
"grad_norm": 0.2626487150494378,
|
||
|
|
"learning_rate": 1.0278810951306253e-05,
|
||
|
|
"loss": 0.3938,
|
||
|
|
"step": 859
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4817813765182186,
|
||
|
|
"grad_norm": 0.2638548708002062,
|
||
|
|
"learning_rate": 1.0229425603293647e-05,
|
||
|
|
"loss": 0.399,
|
||
|
|
"step": 860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.48582995951417,
|
||
|
|
"grad_norm": 0.26785696189459274,
|
||
|
|
"learning_rate": 1.0180118380533558e-05,
|
||
|
|
"loss": 0.3991,
|
||
|
|
"step": 861
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4898785425101213,
|
||
|
|
"grad_norm": 0.24716388565742417,
|
||
|
|
"learning_rate": 1.0130889677285225e-05,
|
||
|
|
"loss": 0.4139,
|
||
|
|
"step": 862
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.493927125506073,
|
||
|
|
"grad_norm": 0.264117042678668,
|
||
|
|
"learning_rate": 1.008173988718005e-05,
|
||
|
|
"loss": 0.39,
|
||
|
|
"step": 863
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4979757085020244,
|
||
|
|
"grad_norm": 0.28146011721560793,
|
||
|
|
"learning_rate": 1.0032669403218439e-05,
|
||
|
|
"loss": 0.4028,
|
||
|
|
"step": 864
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5020242914979756,
|
||
|
|
"grad_norm": 0.25381746306604436,
|
||
|
|
"learning_rate": 9.983678617766677e-06,
|
||
|
|
"loss": 0.3973,
|
||
|
|
"step": 865
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.506072874493927,
|
||
|
|
"grad_norm": 0.248414384918097,
|
||
|
|
"learning_rate": 9.934767922553788e-06,
|
||
|
|
"loss": 0.3919,
|
||
|
|
"step": 866
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5101214574898787,
|
||
|
|
"grad_norm": 0.261041913037476,
|
||
|
|
"learning_rate": 9.885937708668371e-06,
|
||
|
|
"loss": 0.3756,
|
||
|
|
"step": 867
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.51417004048583,
|
||
|
|
"grad_norm": 0.2521268466552633,
|
||
|
|
"learning_rate": 9.837188366555524e-06,
|
||
|
|
"loss": 0.3939,
|
||
|
|
"step": 868
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5182186234817814,
|
||
|
|
"grad_norm": 0.2544659403691464,
|
||
|
|
"learning_rate": 9.788520286013691e-06,
|
||
|
|
"loss": 0.3843,
|
||
|
|
"step": 869
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.522267206477733,
|
||
|
|
"grad_norm": 0.2519326730914753,
|
||
|
|
"learning_rate": 9.739933856191537e-06,
|
||
|
|
"loss": 0.3877,
|
||
|
|
"step": 870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.526315789473684,
|
||
|
|
"grad_norm": 0.2601352550985588,
|
||
|
|
"learning_rate": 9.691429465584866e-06,
|
||
|
|
"loss": 0.4062,
|
||
|
|
"step": 871
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5303643724696356,
|
||
|
|
"grad_norm": 0.2452523864712829,
|
||
|
|
"learning_rate": 9.643007502033492e-06,
|
||
|
|
"loss": 0.3972,
|
||
|
|
"step": 872
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.534412955465587,
|
||
|
|
"grad_norm": 0.2580080844174789,
|
||
|
|
"learning_rate": 9.594668352718152e-06,
|
||
|
|
"loss": 0.4287,
|
||
|
|
"step": 873
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5384615384615383,
|
||
|
|
"grad_norm": 0.2774683350412787,
|
||
|
|
"learning_rate": 9.546412404157383e-06,
|
||
|
|
"loss": 0.3883,
|
||
|
|
"step": 874
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.54251012145749,
|
||
|
|
"grad_norm": 0.26582335181234895,
|
||
|
|
"learning_rate": 9.498240042204471e-06,
|
||
|
|
"loss": 0.4031,
|
||
|
|
"step": 875
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5465587044534415,
|
||
|
|
"grad_norm": 0.27102193629605836,
|
||
|
|
"learning_rate": 9.450151652044339e-06,
|
||
|
|
"loss": 0.4046,
|
||
|
|
"step": 876
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5506072874493926,
|
||
|
|
"grad_norm": 0.2621535176741002,
|
||
|
|
"learning_rate": 9.402147618190472e-06,
|
||
|
|
"loss": 0.4092,
|
||
|
|
"step": 877
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.554655870445344,
|
||
|
|
"grad_norm": 0.25287360382794427,
|
||
|
|
"learning_rate": 9.354228324481841e-06,
|
||
|
|
"loss": 0.4164,
|
||
|
|
"step": 878
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5587044534412957,
|
||
|
|
"grad_norm": 0.25996917355505217,
|
||
|
|
"learning_rate": 9.306394154079852e-06,
|
||
|
|
"loss": 0.3979,
|
||
|
|
"step": 879
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.562753036437247,
|
||
|
|
"grad_norm": 0.257356902067772,
|
||
|
|
"learning_rate": 9.258645489465254e-06,
|
||
|
|
"loss": 0.4112,
|
||
|
|
"step": 880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5668016194331984,
|
||
|
|
"grad_norm": 0.2625344106358796,
|
||
|
|
"learning_rate": 9.21098271243509e-06,
|
||
|
|
"loss": 0.3858,
|
||
|
|
"step": 881
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.57085020242915,
|
||
|
|
"grad_norm": 0.2718294980979648,
|
||
|
|
"learning_rate": 9.163406204099659e-06,
|
||
|
|
"loss": 0.3978,
|
||
|
|
"step": 882
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.574898785425101,
|
||
|
|
"grad_norm": 0.27805947441109446,
|
||
|
|
"learning_rate": 9.115916344879452e-06,
|
||
|
|
"loss": 0.3915,
|
||
|
|
"step": 883
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5789473684210527,
|
||
|
|
"grad_norm": 0.2605246498114428,
|
||
|
|
"learning_rate": 9.068513514502126e-06,
|
||
|
|
"loss": 0.4132,
|
||
|
|
"step": 884
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.582995951417004,
|
||
|
|
"grad_norm": 0.2636106593688388,
|
||
|
|
"learning_rate": 9.021198091999439e-06,
|
||
|
|
"loss": 0.4039,
|
||
|
|
"step": 885
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5870445344129553,
|
||
|
|
"grad_norm": 0.25161807643977996,
|
||
|
|
"learning_rate": 8.973970455704251e-06,
|
||
|
|
"loss": 0.3799,
|
||
|
|
"step": 886
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.591093117408907,
|
||
|
|
"grad_norm": 0.2525315152985144,
|
||
|
|
"learning_rate": 8.926830983247495e-06,
|
||
|
|
"loss": 0.3963,
|
||
|
|
"step": 887
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5951417004048585,
|
||
|
|
"grad_norm": 0.26246867649983074,
|
||
|
|
"learning_rate": 8.879780051555116e-06,
|
||
|
|
"loss": 0.4025,
|
||
|
|
"step": 888
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.5991902834008096,
|
||
|
|
"grad_norm": 0.24695841781571237,
|
||
|
|
"learning_rate": 8.83281803684512e-06,
|
||
|
|
"loss": 0.3938,
|
||
|
|
"step": 889
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.603238866396761,
|
||
|
|
"grad_norm": 0.24609400849220273,
|
||
|
|
"learning_rate": 8.785945314624532e-06,
|
||
|
|
"loss": 0.403,
|
||
|
|
"step": 890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6072874493927127,
|
||
|
|
"grad_norm": 0.2581802374962549,
|
||
|
|
"learning_rate": 8.739162259686385e-06,
|
||
|
|
"loss": 0.3839,
|
||
|
|
"step": 891
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.611336032388664,
|
||
|
|
"grad_norm": 0.25075357340334853,
|
||
|
|
"learning_rate": 8.692469246106734e-06,
|
||
|
|
"loss": 0.4109,
|
||
|
|
"step": 892
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6153846153846154,
|
||
|
|
"grad_norm": 0.26237016496273113,
|
||
|
|
"learning_rate": 8.645866647241677e-06,
|
||
|
|
"loss": 0.4004,
|
||
|
|
"step": 893
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.619433198380567,
|
||
|
|
"grad_norm": 0.2882246362187001,
|
||
|
|
"learning_rate": 8.599354835724348e-06,
|
||
|
|
"loss": 0.3782,
|
||
|
|
"step": 894
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.623481781376518,
|
||
|
|
"grad_norm": 0.25659285881051,
|
||
|
|
"learning_rate": 8.552934183461958e-06,
|
||
|
|
"loss": 0.4162,
|
||
|
|
"step": 895
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6275303643724697,
|
||
|
|
"grad_norm": 0.26088902735066194,
|
||
|
|
"learning_rate": 8.506605061632805e-06,
|
||
|
|
"loss": 0.4129,
|
||
|
|
"step": 896
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6315789473684212,
|
||
|
|
"grad_norm": 0.26965931170838897,
|
||
|
|
"learning_rate": 8.460367840683308e-06,
|
||
|
|
"loss": 0.4011,
|
||
|
|
"step": 897
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6356275303643724,
|
||
|
|
"grad_norm": 0.26530896527002007,
|
||
|
|
"learning_rate": 8.414222890325065e-06,
|
||
|
|
"loss": 0.421,
|
||
|
|
"step": 898
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.639676113360324,
|
||
|
|
"grad_norm": 0.24627037772337812,
|
||
|
|
"learning_rate": 8.368170579531855e-06,
|
||
|
|
"loss": 0.4003,
|
||
|
|
"step": 899
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6437246963562755,
|
||
|
|
"grad_norm": 0.2808326149734353,
|
||
|
|
"learning_rate": 8.322211276536733e-06,
|
||
|
|
"loss": 0.3927,
|
||
|
|
"step": 900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6477732793522266,
|
||
|
|
"grad_norm": 0.2810110062329657,
|
||
|
|
"learning_rate": 8.27634534882907e-06,
|
||
|
|
"loss": 0.3905,
|
||
|
|
"step": 901
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.651821862348178,
|
||
|
|
"grad_norm": 0.2846287059994868,
|
||
|
|
"learning_rate": 8.230573163151585e-06,
|
||
|
|
"loss": 0.3978,
|
||
|
|
"step": 902
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6558704453441297,
|
||
|
|
"grad_norm": 0.2776981963791679,
|
||
|
|
"learning_rate": 8.184895085497466e-06,
|
||
|
|
"loss": 0.4108,
|
||
|
|
"step": 903
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.659919028340081,
|
||
|
|
"grad_norm": 0.25479335375747225,
|
||
|
|
"learning_rate": 8.139311481107401e-06,
|
||
|
|
"loss": 0.3974,
|
||
|
|
"step": 904
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6639676113360324,
|
||
|
|
"grad_norm": 0.26138374170115236,
|
||
|
|
"learning_rate": 8.093822714466686e-06,
|
||
|
|
"loss": 0.4153,
|
||
|
|
"step": 905
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.668016194331984,
|
||
|
|
"grad_norm": 0.25970177380514753,
|
||
|
|
"learning_rate": 8.048429149302271e-06,
|
||
|
|
"loss": 0.4007,
|
||
|
|
"step": 906
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.672064777327935,
|
||
|
|
"grad_norm": 0.26833712196057713,
|
||
|
|
"learning_rate": 8.003131148579902e-06,
|
||
|
|
"loss": 0.4388,
|
||
|
|
"step": 907
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6761133603238867,
|
||
|
|
"grad_norm": 0.26967763660794486,
|
||
|
|
"learning_rate": 7.957929074501182e-06,
|
||
|
|
"loss": 0.3878,
|
||
|
|
"step": 908
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6801619433198383,
|
||
|
|
"grad_norm": 0.24779583699298474,
|
||
|
|
"learning_rate": 7.912823288500695e-06,
|
||
|
|
"loss": 0.398,
|
||
|
|
"step": 909
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6842105263157894,
|
||
|
|
"grad_norm": 0.25156114361962584,
|
||
|
|
"learning_rate": 7.867814151243102e-06,
|
||
|
|
"loss": 0.3911,
|
||
|
|
"step": 910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.688259109311741,
|
||
|
|
"grad_norm": 0.26425518726978486,
|
||
|
|
"learning_rate": 7.822902022620267e-06,
|
||
|
|
"loss": 0.4054,
|
||
|
|
"step": 911
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6923076923076925,
|
||
|
|
"grad_norm": 0.2730049375494375,
|
||
|
|
"learning_rate": 7.778087261748377e-06,
|
||
|
|
"loss": 0.4013,
|
||
|
|
"step": 912
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.6963562753036436,
|
||
|
|
"grad_norm": 0.24850133788333098,
|
||
|
|
"learning_rate": 7.733370226965056e-06,
|
||
|
|
"loss": 0.3852,
|
||
|
|
"step": 913
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.700404858299595,
|
||
|
|
"grad_norm": 0.2892570968437071,
|
||
|
|
"learning_rate": 7.688751275826527e-06,
|
||
|
|
"loss": 0.4235,
|
||
|
|
"step": 914
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7044534412955468,
|
||
|
|
"grad_norm": 0.2608505844113899,
|
||
|
|
"learning_rate": 7.644230765104735e-06,
|
||
|
|
"loss": 0.3952,
|
||
|
|
"step": 915
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.708502024291498,
|
||
|
|
"grad_norm": 0.25553764028268006,
|
||
|
|
"learning_rate": 7.599809050784502e-06,
|
||
|
|
"loss": 0.3686,
|
||
|
|
"step": 916
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7125506072874495,
|
||
|
|
"grad_norm": 0.25154241452219767,
|
||
|
|
"learning_rate": 7.555486488060661e-06,
|
||
|
|
"loss": 0.3891,
|
||
|
|
"step": 917
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.716599190283401,
|
||
|
|
"grad_norm": 0.27048754388083324,
|
||
|
|
"learning_rate": 7.511263431335249e-06,
|
||
|
|
"loss": 0.3934,
|
||
|
|
"step": 918
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.720647773279352,
|
||
|
|
"grad_norm": 0.2579563080100022,
|
||
|
|
"learning_rate": 7.467140234214656e-06,
|
||
|
|
"loss": 0.3993,
|
||
|
|
"step": 919
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7246963562753037,
|
||
|
|
"grad_norm": 0.2458306996708746,
|
||
|
|
"learning_rate": 7.423117249506777e-06,
|
||
|
|
"loss": 0.405,
|
||
|
|
"step": 920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7287449392712553,
|
||
|
|
"grad_norm": 0.24330160175726903,
|
||
|
|
"learning_rate": 7.379194829218231e-06,
|
||
|
|
"loss": 0.4003,
|
||
|
|
"step": 921
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7327935222672064,
|
||
|
|
"grad_norm": 0.2510017842896356,
|
||
|
|
"learning_rate": 7.335373324551518e-06,
|
||
|
|
"loss": 0.4023,
|
||
|
|
"step": 922
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.736842105263158,
|
||
|
|
"grad_norm": 0.2453276727063968,
|
||
|
|
"learning_rate": 7.291653085902224e-06,
|
||
|
|
"loss": 0.3999,
|
||
|
|
"step": 923
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7408906882591095,
|
||
|
|
"grad_norm": 0.5212947883385897,
|
||
|
|
"learning_rate": 7.248034462856213e-06,
|
||
|
|
"loss": 0.397,
|
||
|
|
"step": 924
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7449392712550607,
|
||
|
|
"grad_norm": 0.2467623662008714,
|
||
|
|
"learning_rate": 7.20451780418683e-06,
|
||
|
|
"loss": 0.3836,
|
||
|
|
"step": 925
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.748987854251012,
|
||
|
|
"grad_norm": 0.25487743709850136,
|
||
|
|
"learning_rate": 7.161103457852128e-06,
|
||
|
|
"loss": 0.4147,
|
||
|
|
"step": 926
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.753036437246964,
|
||
|
|
"grad_norm": 0.25464076806229746,
|
||
|
|
"learning_rate": 7.117791770992046e-06,
|
||
|
|
"loss": 0.4041,
|
||
|
|
"step": 927
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.757085020242915,
|
||
|
|
"grad_norm": 0.25167927177436367,
|
||
|
|
"learning_rate": 7.074583089925688e-06,
|
||
|
|
"loss": 0.4059,
|
||
|
|
"step": 928
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7611336032388665,
|
||
|
|
"grad_norm": 0.2655923532043895,
|
||
|
|
"learning_rate": 7.0314777601485155e-06,
|
||
|
|
"loss": 0.3856,
|
||
|
|
"step": 929
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.765182186234818,
|
||
|
|
"grad_norm": 0.2576473066804947,
|
||
|
|
"learning_rate": 6.988476126329604e-06,
|
||
|
|
"loss": 0.3942,
|
||
|
|
"step": 930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.769230769230769,
|
||
|
|
"grad_norm": 0.2501951106942732,
|
||
|
|
"learning_rate": 6.945578532308857e-06,
|
||
|
|
"loss": 0.383,
|
||
|
|
"step": 931
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7732793522267207,
|
||
|
|
"grad_norm": 0.2356251784291764,
|
||
|
|
"learning_rate": 6.902785321094301e-06,
|
||
|
|
"loss": 0.4194,
|
||
|
|
"step": 932
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7773279352226723,
|
||
|
|
"grad_norm": 0.25810031591063787,
|
||
|
|
"learning_rate": 6.8600968348593115e-06,
|
||
|
|
"loss": 0.4156,
|
||
|
|
"step": 933
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7813765182186234,
|
||
|
|
"grad_norm": 0.2548572712801417,
|
||
|
|
"learning_rate": 6.817513414939892e-06,
|
||
|
|
"loss": 0.3837,
|
||
|
|
"step": 934
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.785425101214575,
|
||
|
|
"grad_norm": 0.24568361082556728,
|
||
|
|
"learning_rate": 6.775035401831924e-06,
|
||
|
|
"loss": 0.3952,
|
||
|
|
"step": 935
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7894736842105265,
|
||
|
|
"grad_norm": 0.24790324486991364,
|
||
|
|
"learning_rate": 6.732663135188473e-06,
|
||
|
|
"loss": 0.4004,
|
||
|
|
"step": 936
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7935222672064777,
|
||
|
|
"grad_norm": 0.2580268284209731,
|
||
|
|
"learning_rate": 6.690396953817055e-06,
|
||
|
|
"loss": 0.3757,
|
||
|
|
"step": 937
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7975708502024292,
|
||
|
|
"grad_norm": 0.2661571470471815,
|
||
|
|
"learning_rate": 6.648237195676927e-06,
|
||
|
|
"loss": 0.4105,
|
||
|
|
"step": 938
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.801619433198381,
|
||
|
|
"grad_norm": 0.2540927957213218,
|
||
|
|
"learning_rate": 6.606184197876395e-06,
|
||
|
|
"loss": 0.3982,
|
||
|
|
"step": 939
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.805668016194332,
|
||
|
|
"grad_norm": 0.23852402240223206,
|
||
|
|
"learning_rate": 6.564238296670109e-06,
|
||
|
|
"loss": 0.4053,
|
||
|
|
"step": 940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8097165991902835,
|
||
|
|
"grad_norm": 0.24463113602140202,
|
||
|
|
"learning_rate": 6.52239982745638e-06,
|
||
|
|
"loss": 0.3795,
|
||
|
|
"step": 941
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.813765182186235,
|
||
|
|
"grad_norm": 0.26859166521212224,
|
||
|
|
"learning_rate": 6.480669124774484e-06,
|
||
|
|
"loss": 0.4261,
|
||
|
|
"step": 942
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.817813765182186,
|
||
|
|
"grad_norm": 0.2493362043402934,
|
||
|
|
"learning_rate": 6.439046522302013e-06,
|
||
|
|
"loss": 0.3982,
|
||
|
|
"step": 943
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8218623481781377,
|
||
|
|
"grad_norm": 0.23372007207470108,
|
||
|
|
"learning_rate": 6.397532352852191e-06,
|
||
|
|
"loss": 0.3849,
|
||
|
|
"step": 944
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8259109311740893,
|
||
|
|
"grad_norm": 0.2384052036329297,
|
||
|
|
"learning_rate": 6.356126948371202e-06,
|
||
|
|
"loss": 0.3915,
|
||
|
|
"step": 945
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8299595141700404,
|
||
|
|
"grad_norm": 0.2660293047055337,
|
||
|
|
"learning_rate": 6.314830639935559e-06,
|
||
|
|
"loss": 0.3829,
|
||
|
|
"step": 946
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.834008097165992,
|
||
|
|
"grad_norm": 0.2426288318934874,
|
||
|
|
"learning_rate": 6.273643757749448e-06,
|
||
|
|
"loss": 0.4,
|
||
|
|
"step": 947
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8380566801619436,
|
||
|
|
"grad_norm": 0.23072334367483865,
|
||
|
|
"learning_rate": 6.2325666311420856e-06,
|
||
|
|
"loss": 0.4108,
|
||
|
|
"step": 948
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8421052631578947,
|
||
|
|
"grad_norm": 0.23566078085542072,
|
||
|
|
"learning_rate": 6.1915995885650675e-06,
|
||
|
|
"loss": 0.4187,
|
||
|
|
"step": 949
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8461538461538463,
|
||
|
|
"grad_norm": 0.2533004766523023,
|
||
|
|
"learning_rate": 6.150742957589784e-06,
|
||
|
|
"loss": 0.3939,
|
||
|
|
"step": 950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.850202429149798,
|
||
|
|
"grad_norm": 0.255026363909699,
|
||
|
|
"learning_rate": 6.109997064904774e-06,
|
||
|
|
"loss": 0.3945,
|
||
|
|
"step": 951
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.854251012145749,
|
||
|
|
"grad_norm": 0.2516959153765303,
|
||
|
|
"learning_rate": 6.069362236313092e-06,
|
||
|
|
"loss": 0.3969,
|
||
|
|
"step": 952
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8582995951417005,
|
||
|
|
"grad_norm": 0.23944396762387107,
|
||
|
|
"learning_rate": 6.0288387967297565e-06,
|
||
|
|
"loss": 0.3967,
|
||
|
|
"step": 953
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.862348178137652,
|
||
|
|
"grad_norm": 0.2563244427632202,
|
||
|
|
"learning_rate": 5.988427070179108e-06,
|
||
|
|
"loss": 0.3933,
|
||
|
|
"step": 954
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.866396761133603,
|
||
|
|
"grad_norm": 0.24314778546685753,
|
||
|
|
"learning_rate": 5.948127379792234e-06,
|
||
|
|
"loss": 0.3855,
|
||
|
|
"step": 955
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8704453441295548,
|
||
|
|
"grad_norm": 0.23067182066503353,
|
||
|
|
"learning_rate": 5.90794004780439e-06,
|
||
|
|
"loss": 0.428,
|
||
|
|
"step": 956
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8744939271255063,
|
||
|
|
"grad_norm": 0.23327483579024305,
|
||
|
|
"learning_rate": 5.867865395552405e-06,
|
||
|
|
"loss": 0.4099,
|
||
|
|
"step": 957
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8785425101214575,
|
||
|
|
"grad_norm": 0.2352275179837997,
|
||
|
|
"learning_rate": 5.8279037434721364e-06,
|
||
|
|
"loss": 0.4091,
|
||
|
|
"step": 958
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.882591093117409,
|
||
|
|
"grad_norm": 0.23797554785516736,
|
||
|
|
"learning_rate": 5.788055411095892e-06,
|
||
|
|
"loss": 0.3972,
|
||
|
|
"step": 959
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8866396761133606,
|
||
|
|
"grad_norm": 0.240505128143109,
|
||
|
|
"learning_rate": 5.7483207170498625e-06,
|
||
|
|
"loss": 0.3765,
|
||
|
|
"step": 960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8906882591093117,
|
||
|
|
"grad_norm": 0.24245588301650184,
|
||
|
|
"learning_rate": 5.708699979051613e-06,
|
||
|
|
"loss": 0.393,
|
||
|
|
"step": 961
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8947368421052633,
|
||
|
|
"grad_norm": 0.24322487305380647,
|
||
|
|
"learning_rate": 5.66919351390751e-06,
|
||
|
|
"loss": 0.3911,
|
||
|
|
"step": 962
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.898785425101215,
|
||
|
|
"grad_norm": 0.2632892867939988,
|
||
|
|
"learning_rate": 5.629801637510188e-06,
|
||
|
|
"loss": 0.3941,
|
||
|
|
"step": 963
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.902834008097166,
|
||
|
|
"grad_norm": 0.26557379365147565,
|
||
|
|
"learning_rate": 5.590524664836043e-06,
|
||
|
|
"loss": 0.3802,
|
||
|
|
"step": 964
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9068825910931175,
|
||
|
|
"grad_norm": 0.2388672706173119,
|
||
|
|
"learning_rate": 5.551362909942708e-06,
|
||
|
|
"loss": 0.4009,
|
||
|
|
"step": 965
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.910931174089069,
|
||
|
|
"grad_norm": 0.24146410913985683,
|
||
|
|
"learning_rate": 5.51231668596653e-06,
|
||
|
|
"loss": 0.3877,
|
||
|
|
"step": 966
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.91497975708502,
|
||
|
|
"grad_norm": 0.23118435148167393,
|
||
|
|
"learning_rate": 5.47338630512007e-06,
|
||
|
|
"loss": 0.4022,
|
||
|
|
"step": 967
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.919028340080972,
|
||
|
|
"grad_norm": 0.22922462083807338,
|
||
|
|
"learning_rate": 5.434572078689617e-06,
|
||
|
|
"loss": 0.4164,
|
||
|
|
"step": 968
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9230769230769234,
|
||
|
|
"grad_norm": 0.24877936048119406,
|
||
|
|
"learning_rate": 5.395874317032694e-06,
|
||
|
|
"loss": 0.387,
|
||
|
|
"step": 969
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9271255060728745,
|
||
|
|
"grad_norm": 0.26055430526319345,
|
||
|
|
"learning_rate": 5.357293329575573e-06,
|
||
|
|
"loss": 0.3783,
|
||
|
|
"step": 970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.931174089068826,
|
||
|
|
"grad_norm": 0.22105744471340408,
|
||
|
|
"learning_rate": 5.318829424810797e-06,
|
||
|
|
"loss": 0.4327,
|
||
|
|
"step": 971
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9352226720647776,
|
||
|
|
"grad_norm": 0.2449757871570657,
|
||
|
|
"learning_rate": 5.280482910294728e-06,
|
||
|
|
"loss": 0.3859,
|
||
|
|
"step": 972
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9392712550607287,
|
||
|
|
"grad_norm": 0.24018477802916766,
|
||
|
|
"learning_rate": 5.242254092645078e-06,
|
||
|
|
"loss": 0.3954,
|
||
|
|
"step": 973
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9433198380566803,
|
||
|
|
"grad_norm": 0.54641945163986,
|
||
|
|
"learning_rate": 5.2041432775384445e-06,
|
||
|
|
"loss": 0.3981,
|
||
|
|
"step": 974
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9473684210526314,
|
||
|
|
"grad_norm": 0.2443746903062744,
|
||
|
|
"learning_rate": 5.166150769707892e-06,
|
||
|
|
"loss": 0.396,
|
||
|
|
"step": 975
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.951417004048583,
|
||
|
|
"grad_norm": 0.2555352085521793,
|
||
|
|
"learning_rate": 5.128276872940503e-06,
|
||
|
|
"loss": 0.3953,
|
||
|
|
"step": 976
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9554655870445345,
|
||
|
|
"grad_norm": 0.2391802374854783,
|
||
|
|
"learning_rate": 5.090521890074953e-06,
|
||
|
|
"loss": 0.4082,
|
||
|
|
"step": 977
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9595141700404857,
|
||
|
|
"grad_norm": 0.2644542339847005,
|
||
|
|
"learning_rate": 5.0528861229990675e-06,
|
||
|
|
"loss": 0.3768,
|
||
|
|
"step": 978
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9635627530364372,
|
||
|
|
"grad_norm": 0.24523603226311658,
|
||
|
|
"learning_rate": 5.015369872647442e-06,
|
||
|
|
"loss": 0.4261,
|
||
|
|
"step": 979
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.967611336032389,
|
||
|
|
"grad_norm": 0.23307816128507172,
|
||
|
|
"learning_rate": 4.9779734389990195e-06,
|
||
|
|
"loss": 0.4064,
|
||
|
|
"step": 980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.97165991902834,
|
||
|
|
"grad_norm": 0.2434032550275147,
|
||
|
|
"learning_rate": 4.940697121074675e-06,
|
||
|
|
"loss": 0.417,
|
||
|
|
"step": 981
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9757085020242915,
|
||
|
|
"grad_norm": 0.24908357483438634,
|
||
|
|
"learning_rate": 4.903541216934864e-06,
|
||
|
|
"loss": 0.407,
|
||
|
|
"step": 982
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.979757085020243,
|
||
|
|
"grad_norm": 0.2378276060516935,
|
||
|
|
"learning_rate": 4.8665060236772e-06,
|
||
|
|
"loss": 0.4032,
|
||
|
|
"step": 983
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.983805668016194,
|
||
|
|
"grad_norm": 0.2512193602980041,
|
||
|
|
"learning_rate": 4.829591837434102e-06,
|
||
|
|
"loss": 0.3942,
|
||
|
|
"step": 984
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9878542510121457,
|
||
|
|
"grad_norm": 0.2369980926472287,
|
||
|
|
"learning_rate": 4.792798953370424e-06,
|
||
|
|
"loss": 0.4004,
|
||
|
|
"step": 985
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9919028340080973,
|
||
|
|
"grad_norm": 0.23480574859245698,
|
||
|
|
"learning_rate": 4.756127665681083e-06,
|
||
|
|
"loss": 0.3799,
|
||
|
|
"step": 986
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.9959514170040484,
|
||
|
|
"grad_norm": 0.23241652156430984,
|
||
|
|
"learning_rate": 4.719578267588725e-06,
|
||
|
|
"loss": 0.4183,
|
||
|
|
"step": 987
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.0,
|
||
|
|
"grad_norm": 0.25062131131933907,
|
||
|
|
"learning_rate": 4.6831510513413525e-06,
|
||
|
|
"loss": 0.3991,
|
||
|
|
"step": 988
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.004048582995951,
|
||
|
|
"grad_norm": 0.3783650838000581,
|
||
|
|
"learning_rate": 4.6468463082100176e-06,
|
||
|
|
"loss": 0.386,
|
||
|
|
"step": 989
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.008097165991903,
|
||
|
|
"grad_norm": 0.2920697730941661,
|
||
|
|
"learning_rate": 4.610664328486485e-06,
|
||
|
|
"loss": 0.3684,
|
||
|
|
"step": 990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.012145748987854,
|
||
|
|
"grad_norm": 0.22950016611971452,
|
||
|
|
"learning_rate": 4.5746054014809e-06,
|
||
|
|
"loss": 0.3463,
|
||
|
|
"step": 991
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.016194331983805,
|
||
|
|
"grad_norm": 0.43840434835816955,
|
||
|
|
"learning_rate": 4.5386698155194766e-06,
|
||
|
|
"loss": 0.3548,
|
||
|
|
"step": 992
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.020242914979757,
|
||
|
|
"grad_norm": 0.33888835470517975,
|
||
|
|
"learning_rate": 4.502857857942204e-06,
|
||
|
|
"loss": 0.3403,
|
||
|
|
"step": 993
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.0242914979757085,
|
||
|
|
"grad_norm": 0.26320838133436825,
|
||
|
|
"learning_rate": 4.467169815100549e-06,
|
||
|
|
"loss": 0.3397,
|
||
|
|
"step": 994
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.02834008097166,
|
||
|
|
"grad_norm": 0.27813571661970427,
|
||
|
|
"learning_rate": 4.431605972355137e-06,
|
||
|
|
"loss": 0.3676,
|
||
|
|
"step": 995
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.032388663967612,
|
||
|
|
"grad_norm": 0.3158905375825157,
|
||
|
|
"learning_rate": 4.396166614073516e-06,
|
||
|
|
"loss": 0.357,
|
||
|
|
"step": 996
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.036437246963563,
|
||
|
|
"grad_norm": 0.28364267632515516,
|
||
|
|
"learning_rate": 4.360852023627846e-06,
|
||
|
|
"loss": 0.3532,
|
||
|
|
"step": 997
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.040485829959514,
|
||
|
|
"grad_norm": 0.2543842301000708,
|
||
|
|
"learning_rate": 4.325662483392656e-06,
|
||
|
|
"loss": 0.3813,
|
||
|
|
"step": 998
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.044534412955466,
|
||
|
|
"grad_norm": 0.26105290830466216,
|
||
|
|
"learning_rate": 4.29059827474257e-06,
|
||
|
|
"loss": 0.3411,
|
||
|
|
"step": 999
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.048582995951417,
|
||
|
|
"grad_norm": 0.2868179658328338,
|
||
|
|
"learning_rate": 4.255659678050063e-06,
|
||
|
|
"loss": 0.351,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.052631578947368,
|
||
|
|
"grad_norm": 0.2675746084353129,
|
||
|
|
"learning_rate": 4.22084697268323e-06,
|
||
|
|
"loss": 0.3461,
|
||
|
|
"step": 1001
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.05668016194332,
|
||
|
|
"grad_norm": 0.2443820660376726,
|
||
|
|
"learning_rate": 4.186160437003535e-06,
|
||
|
|
"loss": 0.3639,
|
||
|
|
"step": 1002
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.060728744939271,
|
||
|
|
"grad_norm": 0.2616353649512723,
|
||
|
|
"learning_rate": 4.151600348363582e-06,
|
||
|
|
"loss": 0.337,
|
||
|
|
"step": 1003
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.064777327935222,
|
||
|
|
"grad_norm": 0.25914622000301085,
|
||
|
|
"learning_rate": 4.11716698310493e-06,
|
||
|
|
"loss": 0.3564,
|
||
|
|
"step": 1004
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.068825910931174,
|
||
|
|
"grad_norm": 0.2873524988949604,
|
||
|
|
"learning_rate": 4.082860616555848e-06,
|
||
|
|
"loss": 0.357,
|
||
|
|
"step": 1005
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.0728744939271255,
|
||
|
|
"grad_norm": 0.22891956962770094,
|
||
|
|
"learning_rate": 4.048681523029125e-06,
|
||
|
|
"loss": 0.3405,
|
||
|
|
"step": 1006
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.076923076923077,
|
||
|
|
"grad_norm": 0.24571382770822517,
|
||
|
|
"learning_rate": 4.014629975819882e-06,
|
||
|
|
"loss": 0.3497,
|
||
|
|
"step": 1007
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.080971659919029,
|
||
|
|
"grad_norm": 0.24843214042062728,
|
||
|
|
"learning_rate": 3.9807062472033855e-06,
|
||
|
|
"loss": 0.3546,
|
||
|
|
"step": 1008
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.08502024291498,
|
||
|
|
"grad_norm": 0.2465607959774885,
|
||
|
|
"learning_rate": 3.94691060843287e-06,
|
||
|
|
"loss": 0.3603,
|
||
|
|
"step": 1009
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.089068825910931,
|
||
|
|
"grad_norm": 0.2675122961627523,
|
||
|
|
"learning_rate": 3.913243329737353e-06,
|
||
|
|
"loss": 0.3562,
|
||
|
|
"step": 1010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.093117408906883,
|
||
|
|
"grad_norm": 0.2424809729186539,
|
||
|
|
"learning_rate": 3.879704680319503e-06,
|
||
|
|
"loss": 0.3466,
|
||
|
|
"step": 1011
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.097165991902834,
|
||
|
|
"grad_norm": 0.23108768484925987,
|
||
|
|
"learning_rate": 3.846294928353471e-06,
|
||
|
|
"loss": 0.3757,
|
||
|
|
"step": 1012
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.101214574898785,
|
||
|
|
"grad_norm": 0.2303927276614939,
|
||
|
|
"learning_rate": 3.8130143409827345e-06,
|
||
|
|
"loss": 0.3579,
|
||
|
|
"step": 1013
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.105263157894737,
|
||
|
|
"grad_norm": 0.2503006712404116,
|
||
|
|
"learning_rate": 3.7798631843179887e-06,
|
||
|
|
"loss": 0.3425,
|
||
|
|
"step": 1014
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.109311740890688,
|
||
|
|
"grad_norm": 0.23587877517272107,
|
||
|
|
"learning_rate": 3.7468417234349973e-06,
|
||
|
|
"loss": 0.3546,
|
||
|
|
"step": 1015
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.113360323886639,
|
||
|
|
"grad_norm": 0.21503001494073443,
|
||
|
|
"learning_rate": 3.7139502223724823e-06,
|
||
|
|
"loss": 0.3359,
|
||
|
|
"step": 1016
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.117408906882591,
|
||
|
|
"grad_norm": 0.2289005650196009,
|
||
|
|
"learning_rate": 3.6811889441300095e-06,
|
||
|
|
"loss": 0.3499,
|
||
|
|
"step": 1017
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.1214574898785425,
|
||
|
|
"grad_norm": 0.23211782207857637,
|
||
|
|
"learning_rate": 3.648558150665886e-06,
|
||
|
|
"loss": 0.3553,
|
||
|
|
"step": 1018
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.125506072874494,
|
||
|
|
"grad_norm": 0.23029248826189908,
|
||
|
|
"learning_rate": 3.616058102895075e-06,
|
||
|
|
"loss": 0.3594,
|
||
|
|
"step": 1019
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.129554655870446,
|
||
|
|
"grad_norm": 0.2216727178175573,
|
||
|
|
"learning_rate": 3.583689060687079e-06,
|
||
|
|
"loss": 0.3403,
|
||
|
|
"step": 1020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.133603238866397,
|
||
|
|
"grad_norm": 0.22664469690604203,
|
||
|
|
"learning_rate": 3.5514512828639023e-06,
|
||
|
|
"loss": 0.3364,
|
||
|
|
"step": 1021
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.137651821862348,
|
||
|
|
"grad_norm": 0.23222039032627587,
|
||
|
|
"learning_rate": 3.51934502719796e-06,
|
||
|
|
"loss": 0.361,
|
||
|
|
"step": 1022
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.1417004048583,
|
||
|
|
"grad_norm": 0.22928636739656733,
|
||
|
|
"learning_rate": 3.4873705504100185e-06,
|
||
|
|
"loss": 0.3537,
|
||
|
|
"step": 1023
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.145748987854251,
|
||
|
|
"grad_norm": 0.22560270439305877,
|
||
|
|
"learning_rate": 3.4555281081671345e-06,
|
||
|
|
"loss": 0.3481,
|
||
|
|
"step": 1024
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.149797570850202,
|
||
|
|
"grad_norm": 0.22532252455367618,
|
||
|
|
"learning_rate": 3.4238179550806326e-06,
|
||
|
|
"loss": 0.3388,
|
||
|
|
"step": 1025
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.153846153846154,
|
||
|
|
"grad_norm": 0.2258203766081333,
|
||
|
|
"learning_rate": 3.3922403447040497e-06,
|
||
|
|
"loss": 0.3457,
|
||
|
|
"step": 1026
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.157894736842105,
|
||
|
|
"grad_norm": 0.2457246908515629,
|
||
|
|
"learning_rate": 3.3607955295311247e-06,
|
||
|
|
"loss": 0.3272,
|
||
|
|
"step": 1027
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.161943319838056,
|
||
|
|
"grad_norm": 0.2195767860941197,
|
||
|
|
"learning_rate": 3.32948376099375e-06,
|
||
|
|
"loss": 0.3372,
|
||
|
|
"step": 1028
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.165991902834008,
|
||
|
|
"grad_norm": 0.21990430120043997,
|
||
|
|
"learning_rate": 3.2983052894599976e-06,
|
||
|
|
"loss": 0.37,
|
||
|
|
"step": 1029
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.17004048582996,
|
||
|
|
"grad_norm": 0.23571372424783638,
|
||
|
|
"learning_rate": 3.2672603642320945e-06,
|
||
|
|
"loss": 0.3546,
|
||
|
|
"step": 1030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.174089068825911,
|
||
|
|
"grad_norm": 0.22263428561959228,
|
||
|
|
"learning_rate": 3.236349233544438e-06,
|
||
|
|
"loss": 0.3274,
|
||
|
|
"step": 1031
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.178137651821863,
|
||
|
|
"grad_norm": 0.2244410793184272,
|
||
|
|
"learning_rate": 3.205572144561604e-06,
|
||
|
|
"loss": 0.3485,
|
||
|
|
"step": 1032
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.182186234817814,
|
||
|
|
"grad_norm": 0.22446774577553197,
|
||
|
|
"learning_rate": 3.174929343376374e-06,
|
||
|
|
"loss": 0.3531,
|
||
|
|
"step": 1033
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.186234817813765,
|
||
|
|
"grad_norm": 0.23610655476947706,
|
||
|
|
"learning_rate": 3.1444210750077777e-06,
|
||
|
|
"loss": 0.3692,
|
||
|
|
"step": 1034
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.190283400809717,
|
||
|
|
"grad_norm": 0.23756303888407893,
|
||
|
|
"learning_rate": 3.1140475833991046e-06,
|
||
|
|
"loss": 0.3571,
|
||
|
|
"step": 1035
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.194331983805668,
|
||
|
|
"grad_norm": 0.22231894655105244,
|
||
|
|
"learning_rate": 3.0838091114159895e-06,
|
||
|
|
"loss": 0.3634,
|
||
|
|
"step": 1036
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.198380566801619,
|
||
|
|
"grad_norm": 0.22448325256959328,
|
||
|
|
"learning_rate": 3.0537059008444524e-06,
|
||
|
|
"loss": 0.3511,
|
||
|
|
"step": 1037
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.202429149797571,
|
||
|
|
"grad_norm": 0.2335532285443408,
|
||
|
|
"learning_rate": 3.0237381923889563e-06,
|
||
|
|
"loss": 0.3504,
|
||
|
|
"step": 1038
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.206477732793522,
|
||
|
|
"grad_norm": 0.5793528331954596,
|
||
|
|
"learning_rate": 2.9939062256704997e-06,
|
||
|
|
"loss": 0.3726,
|
||
|
|
"step": 1039
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.2105263157894735,
|
||
|
|
"grad_norm": 0.22423547369345884,
|
||
|
|
"learning_rate": 2.964210239224696e-06,
|
||
|
|
"loss": 0.3604,
|
||
|
|
"step": 1040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.2145748987854255,
|
||
|
|
"grad_norm": 0.23161100538736404,
|
||
|
|
"learning_rate": 2.934650470499869e-06,
|
||
|
|
"loss": 0.3536,
|
||
|
|
"step": 1041
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.218623481781377,
|
||
|
|
"grad_norm": 0.2163494154930157,
|
||
|
|
"learning_rate": 2.905227155855128e-06,
|
||
|
|
"loss": 0.3367,
|
||
|
|
"step": 1042
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.222672064777328,
|
||
|
|
"grad_norm": 0.2205296998945866,
|
||
|
|
"learning_rate": 2.875940530558519e-06,
|
||
|
|
"loss": 0.3674,
|
||
|
|
"step": 1043
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.22672064777328,
|
||
|
|
"grad_norm": 0.23574239307053088,
|
||
|
|
"learning_rate": 2.8467908287851154e-06,
|
||
|
|
"loss": 0.3434,
|
||
|
|
"step": 1044
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.230769230769231,
|
||
|
|
"grad_norm": 0.2225426660179318,
|
||
|
|
"learning_rate": 2.8177782836151603e-06,
|
||
|
|
"loss": 0.3593,
|
||
|
|
"step": 1045
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.234817813765182,
|
||
|
|
"grad_norm": 0.21403984761568692,
|
||
|
|
"learning_rate": 2.7889031270321874e-06,
|
||
|
|
"loss": 0.3576,
|
||
|
|
"step": 1046
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.238866396761134,
|
||
|
|
"grad_norm": 0.23335424224168902,
|
||
|
|
"learning_rate": 2.760165589921184e-06,
|
||
|
|
"loss": 0.3566,
|
||
|
|
"step": 1047
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.242914979757085,
|
||
|
|
"grad_norm": 0.21839872353101586,
|
||
|
|
"learning_rate": 2.7315659020667305e-06,
|
||
|
|
"loss": 0.3727,
|
||
|
|
"step": 1048
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.246963562753036,
|
||
|
|
"grad_norm": 0.21536077129215667,
|
||
|
|
"learning_rate": 2.703104292151164e-06,
|
||
|
|
"loss": 0.3703,
|
||
|
|
"step": 1049
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.251012145748988,
|
||
|
|
"grad_norm": 0.21915686633686665,
|
||
|
|
"learning_rate": 2.6747809877527607e-06,
|
||
|
|
"loss": 0.3641,
|
||
|
|
"step": 1050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.255060728744939,
|
||
|
|
"grad_norm": 0.23552742930679932,
|
||
|
|
"learning_rate": 2.6465962153439082e-06,
|
||
|
|
"loss": 0.3594,
|
||
|
|
"step": 1051
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.2591093117408905,
|
||
|
|
"grad_norm": 0.2296164934760995,
|
||
|
|
"learning_rate": 2.618550200289303e-06,
|
||
|
|
"loss": 0.3354,
|
||
|
|
"step": 1052
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.2631578947368425,
|
||
|
|
"grad_norm": 0.22687092005409798,
|
||
|
|
"learning_rate": 2.59064316684412e-06,
|
||
|
|
"loss": 0.3605,
|
||
|
|
"step": 1053
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.267206477732794,
|
||
|
|
"grad_norm": 0.21758148571738767,
|
||
|
|
"learning_rate": 2.562875338152264e-06,
|
||
|
|
"loss": 0.3465,
|
||
|
|
"step": 1054
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.271255060728745,
|
||
|
|
"grad_norm": 0.21873330025515378,
|
||
|
|
"learning_rate": 2.5352469362445552e-06,
|
||
|
|
"loss": 0.3517,
|
||
|
|
"step": 1055
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.275303643724697,
|
||
|
|
"grad_norm": 0.23071892573407807,
|
||
|
|
"learning_rate": 2.5077581820369523e-06,
|
||
|
|
"loss": 0.3481,
|
||
|
|
"step": 1056
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.279352226720648,
|
||
|
|
"grad_norm": 0.22942132465826762,
|
||
|
|
"learning_rate": 2.480409295328803e-06,
|
||
|
|
"loss": 0.3536,
|
||
|
|
"step": 1057
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.283400809716599,
|
||
|
|
"grad_norm": 0.22170699369129182,
|
||
|
|
"learning_rate": 2.45320049480108e-06,
|
||
|
|
"loss": 0.3499,
|
||
|
|
"step": 1058
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.287449392712551,
|
||
|
|
"grad_norm": 0.2295617065259299,
|
||
|
|
"learning_rate": 2.4261319980146293e-06,
|
||
|
|
"loss": 0.3467,
|
||
|
|
"step": 1059
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.291497975708502,
|
||
|
|
"grad_norm": 0.22497332961078273,
|
||
|
|
"learning_rate": 2.399204021408421e-06,
|
||
|
|
"loss": 0.3431,
|
||
|
|
"step": 1060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.295546558704453,
|
||
|
|
"grad_norm": 0.2292723559038031,
|
||
|
|
"learning_rate": 2.3724167802978395e-06,
|
||
|
|
"loss": 0.3501,
|
||
|
|
"step": 1061
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.299595141700405,
|
||
|
|
"grad_norm": 0.21960199376120665,
|
||
|
|
"learning_rate": 2.3457704888729625e-06,
|
||
|
|
"loss": 0.357,
|
||
|
|
"step": 1062
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.303643724696356,
|
||
|
|
"grad_norm": 0.21196127660437591,
|
||
|
|
"learning_rate": 2.3192653601968162e-06,
|
||
|
|
"loss": 0.3697,
|
||
|
|
"step": 1063
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.3076923076923075,
|
||
|
|
"grad_norm": 0.23493728730326113,
|
||
|
|
"learning_rate": 2.2929016062037057e-06,
|
||
|
|
"loss": 0.3385,
|
||
|
|
"step": 1064
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.3117408906882595,
|
||
|
|
"grad_norm": 0.22362764822046213,
|
||
|
|
"learning_rate": 2.2666794376975055e-06,
|
||
|
|
"loss": 0.3465,
|
||
|
|
"step": 1065
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.315789473684211,
|
||
|
|
"grad_norm": 0.22643559066713614,
|
||
|
|
"learning_rate": 2.24059906434998e-06,
|
||
|
|
"loss": 0.3246,
|
||
|
|
"step": 1066
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.319838056680162,
|
||
|
|
"grad_norm": 0.21694790043613899,
|
||
|
|
"learning_rate": 2.214660694699091e-06,
|
||
|
|
"loss": 0.3411,
|
||
|
|
"step": 1067
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.323886639676114,
|
||
|
|
"grad_norm": 0.23120501051576178,
|
||
|
|
"learning_rate": 2.188864536147355e-06,
|
||
|
|
"loss": 0.3483,
|
||
|
|
"step": 1068
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.327935222672065,
|
||
|
|
"grad_norm": 0.22398304071067057,
|
||
|
|
"learning_rate": 2.1632107949601757e-06,
|
||
|
|
"loss": 0.3605,
|
||
|
|
"step": 1069
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.331983805668016,
|
||
|
|
"grad_norm": 0.2326243643738351,
|
||
|
|
"learning_rate": 2.137699676264171e-06,
|
||
|
|
"loss": 0.3446,
|
||
|
|
"step": 1070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.336032388663968,
|
||
|
|
"grad_norm": 0.22215961851764462,
|
||
|
|
"learning_rate": 2.1123313840455738e-06,
|
||
|
|
"loss": 0.3472,
|
||
|
|
"step": 1071
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.340080971659919,
|
||
|
|
"grad_norm": 0.21620112433540514,
|
||
|
|
"learning_rate": 2.08710612114857e-06,
|
||
|
|
"loss": 0.3602,
|
||
|
|
"step": 1072
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.34412955465587,
|
||
|
|
"grad_norm": 0.22509800012653997,
|
||
|
|
"learning_rate": 2.0620240892736953e-06,
|
||
|
|
"loss": 0.363,
|
||
|
|
"step": 1073
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.348178137651822,
|
||
|
|
"grad_norm": 0.21705431774253145,
|
||
|
|
"learning_rate": 2.0370854889762e-06,
|
||
|
|
"loss": 0.3698,
|
||
|
|
"step": 1074
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.352226720647773,
|
||
|
|
"grad_norm": 0.22491323102615568,
|
||
|
|
"learning_rate": 2.0122905196644726e-06,
|
||
|
|
"loss": 0.3388,
|
||
|
|
"step": 1075
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.3562753036437245,
|
||
|
|
"grad_norm": 0.22184473402586508,
|
||
|
|
"learning_rate": 1.987639379598427e-06,
|
||
|
|
"loss": 0.344,
|
||
|
|
"step": 1076
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.3603238866396765,
|
||
|
|
"grad_norm": 0.21353558784278348,
|
||
|
|
"learning_rate": 1.963132265887919e-06,
|
||
|
|
"loss": 0.3593,
|
||
|
|
"step": 1077
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.364372469635628,
|
||
|
|
"grad_norm": 0.2123526897186212,
|
||
|
|
"learning_rate": 1.9387693744911784e-06,
|
||
|
|
"loss": 0.3598,
|
||
|
|
"step": 1078
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.368421052631579,
|
||
|
|
"grad_norm": 0.2160864041827765,
|
||
|
|
"learning_rate": 1.9145509002132368e-06,
|
||
|
|
"loss": 0.3545,
|
||
|
|
"step": 1079
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.372469635627531,
|
||
|
|
"grad_norm": 0.21291893043423948,
|
||
|
|
"learning_rate": 1.8904770367043723e-06,
|
||
|
|
"loss": 0.3551,
|
||
|
|
"step": 1080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.376518218623482,
|
||
|
|
"grad_norm": 0.22510111085313814,
|
||
|
|
"learning_rate": 1.8665479764585504e-06,
|
||
|
|
"loss": 0.3682,
|
||
|
|
"step": 1081
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.380566801619433,
|
||
|
|
"grad_norm": 0.22336477231529045,
|
||
|
|
"learning_rate": 1.8427639108118977e-06,
|
||
|
|
"loss": 0.3403,
|
||
|
|
"step": 1082
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.384615384615385,
|
||
|
|
"grad_norm": 0.2137247678596304,
|
||
|
|
"learning_rate": 1.8191250299411755e-06,
|
||
|
|
"loss": 0.374,
|
||
|
|
"step": 1083
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.388663967611336,
|
||
|
|
"grad_norm": 0.2166068504125183,
|
||
|
|
"learning_rate": 1.7956315228622446e-06,
|
||
|
|
"loss": 0.345,
|
||
|
|
"step": 1084
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.392712550607287,
|
||
|
|
"grad_norm": 0.22511242742682394,
|
||
|
|
"learning_rate": 1.7722835774285553e-06,
|
||
|
|
"loss": 0.3364,
|
||
|
|
"step": 1085
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.396761133603239,
|
||
|
|
"grad_norm": 0.2214638720990216,
|
||
|
|
"learning_rate": 1.7490813803296624e-06,
|
||
|
|
"loss": 0.3504,
|
||
|
|
"step": 1086
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.40080971659919,
|
||
|
|
"grad_norm": 0.253537346844459,
|
||
|
|
"learning_rate": 1.726025117089718e-06,
|
||
|
|
"loss": 0.3249,
|
||
|
|
"step": 1087
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.4048582995951415,
|
||
|
|
"grad_norm": 0.2150184483673469,
|
||
|
|
"learning_rate": 1.70311497206598e-06,
|
||
|
|
"loss": 0.3504,
|
||
|
|
"step": 1088
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.4089068825910935,
|
||
|
|
"grad_norm": 0.2157399759446024,
|
||
|
|
"learning_rate": 1.6803511284473638e-06,
|
||
|
|
"loss": 0.3477,
|
||
|
|
"step": 1089
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.412955465587045,
|
||
|
|
"grad_norm": 0.22610928635292016,
|
||
|
|
"learning_rate": 1.6577337682529538e-06,
|
||
|
|
"loss": 0.3423,
|
||
|
|
"step": 1090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.417004048582996,
|
||
|
|
"grad_norm": 0.21510373934092863,
|
||
|
|
"learning_rate": 1.6352630723305617e-06,
|
||
|
|
"loss": 0.354,
|
||
|
|
"step": 1091
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.421052631578947,
|
||
|
|
"grad_norm": 0.220831958681427,
|
||
|
|
"learning_rate": 1.612939220355272e-06,
|
||
|
|
"loss": 0.3385,
|
||
|
|
"step": 1092
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.425101214574899,
|
||
|
|
"grad_norm": 0.21331794750385474,
|
||
|
|
"learning_rate": 1.5907623908280112e-06,
|
||
|
|
"loss": 0.3481,
|
||
|
|
"step": 1093
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.42914979757085,
|
||
|
|
"grad_norm": 0.21965358589691067,
|
||
|
|
"learning_rate": 1.568732761074121e-06,
|
||
|
|
"loss": 0.3385,
|
||
|
|
"step": 1094
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.433198380566802,
|
||
|
|
"grad_norm": 0.21808252432907507,
|
||
|
|
"learning_rate": 1.5468505072419237e-06,
|
||
|
|
"loss": 0.3517,
|
||
|
|
"step": 1095
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.437246963562753,
|
||
|
|
"grad_norm": 0.219725370488758,
|
||
|
|
"learning_rate": 1.5251158043013448e-06,
|
||
|
|
"loss": 0.3399,
|
||
|
|
"step": 1096
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.441295546558704,
|
||
|
|
"grad_norm": 0.21447630744324067,
|
||
|
|
"learning_rate": 1.5035288260424885e-06,
|
||
|
|
"loss": 0.3448,
|
||
|
|
"step": 1097
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.445344129554655,
|
||
|
|
"grad_norm": 0.23811093204087833,
|
||
|
|
"learning_rate": 1.4820897450742623e-06,
|
||
|
|
"loss": 0.356,
|
||
|
|
"step": 1098
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.449392712550607,
|
||
|
|
"grad_norm": 0.2963749157288049,
|
||
|
|
"learning_rate": 1.4607987328229767e-06,
|
||
|
|
"loss": 0.3598,
|
||
|
|
"step": 1099
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.4534412955465585,
|
||
|
|
"grad_norm": 0.22526454186464548,
|
||
|
|
"learning_rate": 1.4396559595310056e-06,
|
||
|
|
"loss": 0.3336,
|
||
|
|
"step": 1100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.4574898785425106,
|
||
|
|
"grad_norm": 0.21931253242822588,
|
||
|
|
"learning_rate": 1.4186615942553973e-06,
|
||
|
|
"loss": 0.369,
|
||
|
|
"step": 1101
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.461538461538462,
|
||
|
|
"grad_norm": 0.21424967558111693,
|
||
|
|
"learning_rate": 1.3978158048665424e-06,
|
||
|
|
"loss": 0.3327,
|
||
|
|
"step": 1102
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.465587044534413,
|
||
|
|
"grad_norm": 0.2174484158413736,
|
||
|
|
"learning_rate": 1.3771187580468115e-06,
|
||
|
|
"loss": 0.3584,
|
||
|
|
"step": 1103
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.469635627530364,
|
||
|
|
"grad_norm": 0.21771367345686166,
|
||
|
|
"learning_rate": 1.3565706192892392e-06,
|
||
|
|
"loss": 0.3695,
|
||
|
|
"step": 1104
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.473684210526316,
|
||
|
|
"grad_norm": 0.21968952269674819,
|
||
|
|
"learning_rate": 1.3361715528961993e-06,
|
||
|
|
"loss": 0.3375,
|
||
|
|
"step": 1105
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.477732793522267,
|
||
|
|
"grad_norm": 0.21393503845515646,
|
||
|
|
"learning_rate": 1.3159217219780817e-06,
|
||
|
|
"loss": 0.3629,
|
||
|
|
"step": 1106
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.481781376518219,
|
||
|
|
"grad_norm": 0.21907466640012907,
|
||
|
|
"learning_rate": 1.2958212884519949e-06,
|
||
|
|
"loss": 0.3595,
|
||
|
|
"step": 1107
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.48582995951417,
|
||
|
|
"grad_norm": 0.21248830945237449,
|
||
|
|
"learning_rate": 1.2758704130404675e-06,
|
||
|
|
"loss": 0.3682,
|
||
|
|
"step": 1108
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.489878542510121,
|
||
|
|
"grad_norm": 0.24613964059771007,
|
||
|
|
"learning_rate": 1.2560692552701738e-06,
|
||
|
|
"loss": 0.362,
|
||
|
|
"step": 1109
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.493927125506072,
|
||
|
|
"grad_norm": 0.21006573306691365,
|
||
|
|
"learning_rate": 1.2364179734706382e-06,
|
||
|
|
"loss": 0.3706,
|
||
|
|
"step": 1110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.497975708502024,
|
||
|
|
"grad_norm": 0.21927485168668118,
|
||
|
|
"learning_rate": 1.2169167247729852e-06,
|
||
|
|
"loss": 0.3552,
|
||
|
|
"step": 1111
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.502024291497976,
|
||
|
|
"grad_norm": 0.22757958778961074,
|
||
|
|
"learning_rate": 1.1975656651086864e-06,
|
||
|
|
"loss": 0.3568,
|
||
|
|
"step": 1112
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.506072874493928,
|
||
|
|
"grad_norm": 0.2258873096014452,
|
||
|
|
"learning_rate": 1.1783649492082926e-06,
|
||
|
|
"loss": 0.3458,
|
||
|
|
"step": 1113
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.510121457489879,
|
||
|
|
"grad_norm": 0.21963859396377536,
|
||
|
|
"learning_rate": 1.1593147306002183e-06,
|
||
|
|
"loss": 0.3475,
|
||
|
|
"step": 1114
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.51417004048583,
|
||
|
|
"grad_norm": 0.21342866870982344,
|
||
|
|
"learning_rate": 1.140415161609505e-06,
|
||
|
|
"loss": 0.3566,
|
||
|
|
"step": 1115
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.518218623481781,
|
||
|
|
"grad_norm": 0.2164689381124055,
|
||
|
|
"learning_rate": 1.121666393356604e-06,
|
||
|
|
"loss": 0.3708,
|
||
|
|
"step": 1116
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.522267206477733,
|
||
|
|
"grad_norm": 0.22600934975751422,
|
||
|
|
"learning_rate": 1.1030685757561632e-06,
|
||
|
|
"loss": 0.3359,
|
||
|
|
"step": 1117
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.526315789473684,
|
||
|
|
"grad_norm": 0.23203643357959783,
|
||
|
|
"learning_rate": 1.0846218575158373e-06,
|
||
|
|
"loss": 0.3365,
|
||
|
|
"step": 1118
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.530364372469636,
|
||
|
|
"grad_norm": 0.2210541520003719,
|
||
|
|
"learning_rate": 1.066326386135097e-06,
|
||
|
|
"loss": 0.3318,
|
||
|
|
"step": 1119
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.534412955465587,
|
||
|
|
"grad_norm": 0.3468166668970051,
|
||
|
|
"learning_rate": 1.0481823079040398e-06,
|
||
|
|
"loss": 0.3576,
|
||
|
|
"step": 1120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.538461538461538,
|
||
|
|
"grad_norm": 0.21424482433826317,
|
||
|
|
"learning_rate": 1.0301897679022344e-06,
|
||
|
|
"loss": 0.3634,
|
||
|
|
"step": 1121
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.5425101214574894,
|
||
|
|
"grad_norm": 0.22063064037182897,
|
||
|
|
"learning_rate": 1.0123489099975491e-06,
|
||
|
|
"loss": 0.3519,
|
||
|
|
"step": 1122
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.5465587044534415,
|
||
|
|
"grad_norm": 0.21828075641418282,
|
||
|
|
"learning_rate": 9.9465987684501e-07,
|
||
|
|
"loss": 0.3576,
|
||
|
|
"step": 1123
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.550607287449393,
|
||
|
|
"grad_norm": 0.21720568793672695,
|
||
|
|
"learning_rate": 9.771228098856534e-07,
|
||
|
|
"loss": 0.3366,
|
||
|
|
"step": 1124
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.554655870445345,
|
||
|
|
"grad_norm": 0.21237924750192497,
|
||
|
|
"learning_rate": 9.597378493454013e-07,
|
||
|
|
"loss": 0.3446,
|
||
|
|
"step": 1125
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.558704453441296,
|
||
|
|
"grad_norm": 0.21966762364519243,
|
||
|
|
"learning_rate": 9.425051342339353e-07,
|
||
|
|
"loss": 0.3161,
|
||
|
|
"step": 1126
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.562753036437247,
|
||
|
|
"grad_norm": 0.21776030085005246,
|
||
|
|
"learning_rate": 9.254248023435864e-07,
|
||
|
|
"loss": 0.3387,
|
||
|
|
"step": 1127
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.566801619433198,
|
||
|
|
"grad_norm": 0.2158183267551532,
|
||
|
|
"learning_rate": 9.084969902482288e-07,
|
||
|
|
"loss": 0.3541,
|
||
|
|
"step": 1128
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.57085020242915,
|
||
|
|
"grad_norm": 0.20854892839872805,
|
||
|
|
"learning_rate": 8.917218333021993e-07,
|
||
|
|
"loss": 0.358,
|
||
|
|
"step": 1129
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.574898785425101,
|
||
|
|
"grad_norm": 0.20522235613228879,
|
||
|
|
"learning_rate": 8.750994656392109e-07,
|
||
|
|
"loss": 0.3697,
|
||
|
|
"step": 1130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.578947368421053,
|
||
|
|
"grad_norm": 0.7308803631927825,
|
||
|
|
"learning_rate": 8.58630020171265e-07,
|
||
|
|
"loss": 0.3578,
|
||
|
|
"step": 1131
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.582995951417004,
|
||
|
|
"grad_norm": 0.21438119122643884,
|
||
|
|
"learning_rate": 8.423136285876099e-07,
|
||
|
|
"loss": 0.3368,
|
||
|
|
"step": 1132
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.587044534412955,
|
||
|
|
"grad_norm": 0.21078722284785054,
|
||
|
|
"learning_rate": 8.26150421353682e-07,
|
||
|
|
"loss": 0.3507,
|
||
|
|
"step": 1133
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.5910931174089065,
|
||
|
|
"grad_norm": 0.2241967793900079,
|
||
|
|
"learning_rate": 8.101405277100549e-07,
|
||
|
|
"loss": 0.3232,
|
||
|
|
"step": 1134
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.5951417004048585,
|
||
|
|
"grad_norm": 0.24736717702231711,
|
||
|
|
"learning_rate": 7.942840756714077e-07,
|
||
|
|
"loss": 0.3439,
|
||
|
|
"step": 1135
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.59919028340081,
|
||
|
|
"grad_norm": 0.2256851105132317,
|
||
|
|
"learning_rate": 7.785811920255093e-07,
|
||
|
|
"loss": 0.3239,
|
||
|
|
"step": 1136
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.603238866396762,
|
||
|
|
"grad_norm": 0.21713672643091528,
|
||
|
|
"learning_rate": 7.630320023322024e-07,
|
||
|
|
"loss": 0.3432,
|
||
|
|
"step": 1137
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.607287449392713,
|
||
|
|
"grad_norm": 0.21469515378147339,
|
||
|
|
"learning_rate": 7.476366309223903e-07,
|
||
|
|
"loss": 0.3548,
|
||
|
|
"step": 1138
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.611336032388664,
|
||
|
|
"grad_norm": 0.20722760757546388,
|
||
|
|
"learning_rate": 7.323952008970537e-07,
|
||
|
|
"loss": 0.3457,
|
||
|
|
"step": 1139
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.615384615384615,
|
||
|
|
"grad_norm": 0.21059549574085148,
|
||
|
|
"learning_rate": 7.173078341262618e-07,
|
||
|
|
"loss": 0.3455,
|
||
|
|
"step": 1140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.619433198380567,
|
||
|
|
"grad_norm": 0.2178516769467103,
|
||
|
|
"learning_rate": 7.023746512481989e-07,
|
||
|
|
"loss": 0.355,
|
||
|
|
"step": 1141
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.623481781376518,
|
||
|
|
"grad_norm": 0.22997604619876846,
|
||
|
|
"learning_rate": 6.875957716681902e-07,
|
||
|
|
"loss": 0.3422,
|
||
|
|
"step": 1142
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.62753036437247,
|
||
|
|
"grad_norm": 0.21809168795733133,
|
||
|
|
"learning_rate": 6.729713135577687e-07,
|
||
|
|
"loss": 0.347,
|
||
|
|
"step": 1143
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.631578947368421,
|
||
|
|
"grad_norm": 0.20557633917638404,
|
||
|
|
"learning_rate": 6.585013938537143e-07,
|
||
|
|
"loss": 0.3612,
|
||
|
|
"step": 1144
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.635627530364372,
|
||
|
|
"grad_norm": 0.2201300927637923,
|
||
|
|
"learning_rate": 6.441861282571115e-07,
|
||
|
|
"loss": 0.3432,
|
||
|
|
"step": 1145
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.6396761133603235,
|
||
|
|
"grad_norm": 0.20951324776043373,
|
||
|
|
"learning_rate": 6.300256312324471e-07,
|
||
|
|
"loss": 0.3543,
|
||
|
|
"step": 1146
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.6437246963562755,
|
||
|
|
"grad_norm": 0.2127126175388503,
|
||
|
|
"learning_rate": 6.160200160066776e-07,
|
||
|
|
"loss": 0.358,
|
||
|
|
"step": 1147
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.647773279352227,
|
||
|
|
"grad_norm": 0.21574183445888204,
|
||
|
|
"learning_rate": 6.021693945683327e-07,
|
||
|
|
"loss": 0.335,
|
||
|
|
"step": 1148
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.651821862348179,
|
||
|
|
"grad_norm": 0.22151779935166038,
|
||
|
|
"learning_rate": 5.884738776666044e-07,
|
||
|
|
"loss": 0.351,
|
||
|
|
"step": 1149
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.65587044534413,
|
||
|
|
"grad_norm": 0.21472485408796055,
|
||
|
|
"learning_rate": 5.749335748104878e-07,
|
||
|
|
"loss": 0.3494,
|
||
|
|
"step": 1150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.659919028340081,
|
||
|
|
"grad_norm": 0.2274685289308812,
|
||
|
|
"learning_rate": 5.615485942678778e-07,
|
||
|
|
"loss": 0.3417,
|
||
|
|
"step": 1151
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.663967611336032,
|
||
|
|
"grad_norm": 0.2232214155589449,
|
||
|
|
"learning_rate": 5.483190430647289e-07,
|
||
|
|
"loss": 0.3303,
|
||
|
|
"step": 1152
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.668016194331984,
|
||
|
|
"grad_norm": 0.21684150531046809,
|
||
|
|
"learning_rate": 5.352450269841747e-07,
|
||
|
|
"loss": 0.3532,
|
||
|
|
"step": 1153
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.672064777327935,
|
||
|
|
"grad_norm": 0.2248154804634695,
|
||
|
|
"learning_rate": 5.22326650565701e-07,
|
||
|
|
"loss": 0.3392,
|
||
|
|
"step": 1154
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.676113360323887,
|
||
|
|
"grad_norm": 0.22466369098426495,
|
||
|
|
"learning_rate": 5.095640171043048e-07,
|
||
|
|
"loss": 0.3481,
|
||
|
|
"step": 1155
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.680161943319838,
|
||
|
|
"grad_norm": 0.20727811265249846,
|
||
|
|
"learning_rate": 4.969572286496594e-07,
|
||
|
|
"loss": 0.3448,
|
||
|
|
"step": 1156
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.684210526315789,
|
||
|
|
"grad_norm": 0.2112213551031676,
|
||
|
|
"learning_rate": 4.845063860053079e-07,
|
||
|
|
"loss": 0.3514,
|
||
|
|
"step": 1157
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.6882591093117405,
|
||
|
|
"grad_norm": 0.21480913344922994,
|
||
|
|
"learning_rate": 4.7221158872785997e-07,
|
||
|
|
"loss": 0.3413,
|
||
|
|
"step": 1158
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.6923076923076925,
|
||
|
|
"grad_norm": 0.21471149284077426,
|
||
|
|
"learning_rate": 4.600729351261857e-07,
|
||
|
|
"loss": 0.3631,
|
||
|
|
"step": 1159
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.696356275303644,
|
||
|
|
"grad_norm": 0.2153034048725801,
|
||
|
|
"learning_rate": 4.4809052226063355e-07,
|
||
|
|
"loss": 0.3216,
|
||
|
|
"step": 1160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.700404858299595,
|
||
|
|
"grad_norm": 0.22747359845273807,
|
||
|
|
"learning_rate": 4.3626444594226046e-07,
|
||
|
|
"loss": 0.3254,
|
||
|
|
"step": 1161
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.704453441295547,
|
||
|
|
"grad_norm": 0.20540732174402357,
|
||
|
|
"learning_rate": 4.245948007320588e-07,
|
||
|
|
"loss": 0.3673,
|
||
|
|
"step": 1162
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.708502024291498,
|
||
|
|
"grad_norm": 0.2152998120592181,
|
||
|
|
"learning_rate": 4.130816799401971e-07,
|
||
|
|
"loss": 0.3494,
|
||
|
|
"step": 1163
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.712550607287449,
|
||
|
|
"grad_norm": 0.21346712908090132,
|
||
|
|
"learning_rate": 4.017251756252827e-07,
|
||
|
|
"loss": 0.3398,
|
||
|
|
"step": 1164
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.716599190283401,
|
||
|
|
"grad_norm": 0.21132438841224435,
|
||
|
|
"learning_rate": 3.9052537859362473e-07,
|
||
|
|
"loss": 0.3586,
|
||
|
|
"step": 1165
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.720647773279352,
|
||
|
|
"grad_norm": 0.2012849616835743,
|
||
|
|
"learning_rate": 3.7948237839849687e-07,
|
||
|
|
"loss": 0.3664,
|
||
|
|
"step": 1166
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.724696356275303,
|
||
|
|
"grad_norm": 0.21325554419180645,
|
||
|
|
"learning_rate": 3.6859626333943797e-07,
|
||
|
|
"loss": 0.3402,
|
||
|
|
"step": 1167
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.728744939271255,
|
||
|
|
"grad_norm": 0.22467465337471185,
|
||
|
|
"learning_rate": 3.578671204615347e-07,
|
||
|
|
"loss": 0.3482,
|
||
|
|
"step": 1168
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.732793522267206,
|
||
|
|
"grad_norm": 0.21957208624605082,
|
||
|
|
"learning_rate": 3.472950355547267e-07,
|
||
|
|
"loss": 0.3568,
|
||
|
|
"step": 1169
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.7368421052631575,
|
||
|
|
"grad_norm": 0.21790040803264907,
|
||
|
|
"learning_rate": 3.3688009315312244e-07,
|
||
|
|
"loss": 0.3436,
|
||
|
|
"step": 1170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.7408906882591095,
|
||
|
|
"grad_norm": 0.21071110166407558,
|
||
|
|
"learning_rate": 3.266223765343224e-07,
|
||
|
|
"loss": 0.3639,
|
||
|
|
"step": 1171
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.744939271255061,
|
||
|
|
"grad_norm": 0.2173513694277339,
|
||
|
|
"learning_rate": 3.16521967718757e-07,
|
||
|
|
"loss": 0.3534,
|
||
|
|
"step": 1172
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.748987854251012,
|
||
|
|
"grad_norm": 0.20986188074936768,
|
||
|
|
"learning_rate": 3.0657894746902726e-07,
|
||
|
|
"loss": 0.3737,
|
||
|
|
"step": 1173
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.753036437246964,
|
||
|
|
"grad_norm": 0.2105350349001971,
|
||
|
|
"learning_rate": 2.9679339528925654e-07,
|
||
|
|
"loss": 0.3611,
|
||
|
|
"step": 1174
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.757085020242915,
|
||
|
|
"grad_norm": 0.20841721906603652,
|
||
|
|
"learning_rate": 2.871653894244597e-07,
|
||
|
|
"loss": 0.3651,
|
||
|
|
"step": 1175
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.761133603238866,
|
||
|
|
"grad_norm": 0.20946338290949917,
|
||
|
|
"learning_rate": 2.7769500685991266e-07,
|
||
|
|
"loss": 0.3586,
|
||
|
|
"step": 1176
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.765182186234818,
|
||
|
|
"grad_norm": 0.20818293808999547,
|
||
|
|
"learning_rate": 2.683823233205485e-07,
|
||
|
|
"loss": 0.3795,
|
||
|
|
"step": 1177
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.769230769230769,
|
||
|
|
"grad_norm": 0.21714903566281354,
|
||
|
|
"learning_rate": 2.592274132703287e-07,
|
||
|
|
"loss": 0.3367,
|
||
|
|
"step": 1178
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.77327935222672,
|
||
|
|
"grad_norm": 0.22896158253435145,
|
||
|
|
"learning_rate": 2.502303499116754e-07,
|
||
|
|
"loss": 0.3455,
|
||
|
|
"step": 1179
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.777327935222672,
|
||
|
|
"grad_norm": 0.21682902898881515,
|
||
|
|
"learning_rate": 2.413912051848666e-07,
|
||
|
|
"loss": 0.352,
|
||
|
|
"step": 1180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.781376518218623,
|
||
|
|
"grad_norm": 0.21977143136584582,
|
||
|
|
"learning_rate": 2.3271004976746613e-07,
|
||
|
|
"loss": 0.3502,
|
||
|
|
"step": 1181
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.7854251012145745,
|
||
|
|
"grad_norm": 0.21766349727526496,
|
||
|
|
"learning_rate": 2.2418695307376393e-07,
|
||
|
|
"loss": 0.356,
|
||
|
|
"step": 1182
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.7894736842105265,
|
||
|
|
"grad_norm": 0.20907183893598175,
|
||
|
|
"learning_rate": 2.1582198325420968e-07,
|
||
|
|
"loss": 0.3387,
|
||
|
|
"step": 1183
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.793522267206478,
|
||
|
|
"grad_norm": 0.21138155191339106,
|
||
|
|
"learning_rate": 2.0761520719488447e-07,
|
||
|
|
"loss": 0.3368,
|
||
|
|
"step": 1184
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.797570850202429,
|
||
|
|
"grad_norm": 0.2124284279326404,
|
||
|
|
"learning_rate": 1.9956669051694798e-07,
|
||
|
|
"loss": 0.3521,
|
||
|
|
"step": 1185
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.801619433198381,
|
||
|
|
"grad_norm": 0.21694717104780725,
|
||
|
|
"learning_rate": 1.916764975761254e-07,
|
||
|
|
"loss": 0.3668,
|
||
|
|
"step": 1186
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.805668016194332,
|
||
|
|
"grad_norm": 0.2170397663825815,
|
||
|
|
"learning_rate": 1.839446914621923e-07,
|
||
|
|
"loss": 0.3409,
|
||
|
|
"step": 1187
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.809716599190283,
|
||
|
|
"grad_norm": 0.21748643805786635,
|
||
|
|
"learning_rate": 1.7637133399845962e-07,
|
||
|
|
"loss": 0.3339,
|
||
|
|
"step": 1188
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.813765182186235,
|
||
|
|
"grad_norm": 0.20707902210633883,
|
||
|
|
"learning_rate": 1.6895648574129397e-07,
|
||
|
|
"loss": 0.3363,
|
||
|
|
"step": 1189
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.817813765182186,
|
||
|
|
"grad_norm": 0.2194633874502687,
|
||
|
|
"learning_rate": 1.6170020597962465e-07,
|
||
|
|
"loss": 0.3255,
|
||
|
|
"step": 1190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.821862348178137,
|
||
|
|
"grad_norm": 0.2087898244726411,
|
||
|
|
"learning_rate": 1.5460255273447078e-07,
|
||
|
|
"loss": 0.3619,
|
||
|
|
"step": 1191
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.825910931174089,
|
||
|
|
"grad_norm": 0.21490707710056509,
|
||
|
|
"learning_rate": 1.4766358275847715e-07,
|
||
|
|
"loss": 0.3486,
|
||
|
|
"step": 1192
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.82995951417004,
|
||
|
|
"grad_norm": 0.21347088097556302,
|
||
|
|
"learning_rate": 1.4088335153546573e-07,
|
||
|
|
"loss": 0.345,
|
||
|
|
"step": 1193
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.834008097165992,
|
||
|
|
"grad_norm": 0.2076237152665109,
|
||
|
|
"learning_rate": 1.3426191327998496e-07,
|
||
|
|
"loss": 0.3633,
|
||
|
|
"step": 1194
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.838056680161944,
|
||
|
|
"grad_norm": 0.2174825648484177,
|
||
|
|
"learning_rate": 1.277993209368744e-07,
|
||
|
|
"loss": 0.3644,
|
||
|
|
"step": 1195
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.842105263157895,
|
||
|
|
"grad_norm": 0.22262434135765247,
|
||
|
|
"learning_rate": 1.214956261808542e-07,
|
||
|
|
"loss": 0.3364,
|
||
|
|
"step": 1196
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.846153846153846,
|
||
|
|
"grad_norm": 0.20514308901313075,
|
||
|
|
"learning_rate": 1.1535087941609846e-07,
|
||
|
|
"loss": 0.3688,
|
||
|
|
"step": 1197
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.850202429149798,
|
||
|
|
"grad_norm": 0.20592342617789813,
|
||
|
|
"learning_rate": 1.0936512977583802e-07,
|
||
|
|
"loss": 0.3601,
|
||
|
|
"step": 1198
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.854251012145749,
|
||
|
|
"grad_norm": 0.22715270838435855,
|
||
|
|
"learning_rate": 1.0353842512196732e-07,
|
||
|
|
"loss": 0.359,
|
||
|
|
"step": 1199
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.8582995951417,
|
||
|
|
"grad_norm": 0.20917532052739576,
|
||
|
|
"learning_rate": 9.787081204466253e-08,
|
||
|
|
"loss": 0.352,
|
||
|
|
"step": 1200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.862348178137652,
|
||
|
|
"grad_norm": 0.20905358860043996,
|
||
|
|
"learning_rate": 9.236233586200627e-08,
|
||
|
|
"loss": 0.3567,
|
||
|
|
"step": 1201
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.866396761133603,
|
||
|
|
"grad_norm": 0.21305061918848994,
|
||
|
|
"learning_rate": 8.70130406196279e-08,
|
||
|
|
"loss": 0.3563,
|
||
|
|
"step": 1202
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.870445344129554,
|
||
|
|
"grad_norm": 0.2065338874580827,
|
||
|
|
"learning_rate": 8.182296909035047e-08,
|
||
|
|
"loss": 0.3413,
|
||
|
|
"step": 1203
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.874493927125506,
|
||
|
|
"grad_norm": 0.21448653176165883,
|
||
|
|
"learning_rate": 7.679216277384439e-08,
|
||
|
|
"loss": 0.3573,
|
||
|
|
"step": 1204
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.8785425101214575,
|
||
|
|
"grad_norm": 0.22006429040638045,
|
||
|
|
"learning_rate": 7.19206618963053e-08,
|
||
|
|
"loss": 0.3393,
|
||
|
|
"step": 1205
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.882591093117409,
|
||
|
|
"grad_norm": 0.20715854975505488,
|
||
|
|
"learning_rate": 6.720850541012347e-08,
|
||
|
|
"loss": 0.3631,
|
||
|
|
"step": 1206
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.886639676113361,
|
||
|
|
"grad_norm": 0.21297644455013515,
|
||
|
|
"learning_rate": 6.265573099357714e-08,
|
||
|
|
"loss": 0.3473,
|
||
|
|
"step": 1207
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.890688259109312,
|
||
|
|
"grad_norm": 0.21646831391396468,
|
||
|
|
"learning_rate": 5.826237505053067e-08,
|
||
|
|
"loss": 0.3424,
|
||
|
|
"step": 1208
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.894736842105263,
|
||
|
|
"grad_norm": 0.21129211798763098,
|
||
|
|
"learning_rate": 5.4028472710137e-08,
|
||
|
|
"loss": 0.3413,
|
||
|
|
"step": 1209
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.898785425101215,
|
||
|
|
"grad_norm": 0.21609170971481384,
|
||
|
|
"learning_rate": 4.9954057826571146e-08,
|
||
|
|
"loss": 0.3464,
|
||
|
|
"step": 1210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.902834008097166,
|
||
|
|
"grad_norm": 0.21080543543007416,
|
||
|
|
"learning_rate": 4.6039162978743776e-08,
|
||
|
|
"loss": 0.3451,
|
||
|
|
"step": 1211
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.906882591093117,
|
||
|
|
"grad_norm": 0.2069308289313909,
|
||
|
|
"learning_rate": 4.228381947004812e-08,
|
||
|
|
"loss": 0.3551,
|
||
|
|
"step": 1212
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.910931174089069,
|
||
|
|
"grad_norm": 0.20960865982845941,
|
||
|
|
"learning_rate": 3.868805732811343e-08,
|
||
|
|
"loss": 0.3666,
|
||
|
|
"step": 1213
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.91497975708502,
|
||
|
|
"grad_norm": 0.21861468388063918,
|
||
|
|
"learning_rate": 3.525190530455636e-08,
|
||
|
|
"loss": 0.3675,
|
||
|
|
"step": 1214
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.919028340080971,
|
||
|
|
"grad_norm": 0.21330046643890457,
|
||
|
|
"learning_rate": 3.197539087475887e-08,
|
||
|
|
"loss": 0.3519,
|
||
|
|
"step": 1215
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.923076923076923,
|
||
|
|
"grad_norm": 0.22121906183317566,
|
||
|
|
"learning_rate": 2.8858540237641786e-08,
|
||
|
|
"loss": 0.3569,
|
||
|
|
"step": 1216
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.9271255060728745,
|
||
|
|
"grad_norm": 0.21833650582428296,
|
||
|
|
"learning_rate": 2.5901378315458247e-08,
|
||
|
|
"loss": 0.3439,
|
||
|
|
"step": 1217
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.931174089068826,
|
||
|
|
"grad_norm": 0.19587236502502833,
|
||
|
|
"learning_rate": 2.3103928753598347e-08,
|
||
|
|
"loss": 0.369,
|
||
|
|
"step": 1218
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.935222672064778,
|
||
|
|
"grad_norm": 0.22091116653653184,
|
||
|
|
"learning_rate": 2.0466213920393718e-08,
|
||
|
|
"loss": 0.3478,
|
||
|
|
"step": 1219
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.939271255060729,
|
||
|
|
"grad_norm": 0.22495205921700442,
|
||
|
|
"learning_rate": 1.7988254906942115e-08,
|
||
|
|
"loss": 0.3645,
|
||
|
|
"step": 1220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.94331983805668,
|
||
|
|
"grad_norm": 0.20674120912861857,
|
||
|
|
"learning_rate": 1.5670071526936447e-08,
|
||
|
|
"loss": 0.3702,
|
||
|
|
"step": 1221
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.947368421052632,
|
||
|
|
"grad_norm": 0.20410245486620435,
|
||
|
|
"learning_rate": 1.3511682316509344e-08,
|
||
|
|
"loss": 0.3617,
|
||
|
|
"step": 1222
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.951417004048583,
|
||
|
|
"grad_norm": 0.2099043633781449,
|
||
|
|
"learning_rate": 1.1513104534086605e-08,
|
||
|
|
"loss": 0.3642,
|
||
|
|
"step": 1223
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.955465587044534,
|
||
|
|
"grad_norm": 0.21209144325685536,
|
||
|
|
"learning_rate": 9.67435416023843e-09,
|
||
|
|
"loss": 0.3345,
|
||
|
|
"step": 1224
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.959514170040486,
|
||
|
|
"grad_norm": 0.20383189291337764,
|
||
|
|
"learning_rate": 7.995445897566178e-09,
|
||
|
|
"loss": 0.3534,
|
||
|
|
"step": 1225
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.963562753036437,
|
||
|
|
"grad_norm": 0.21166245571885528,
|
||
|
|
"learning_rate": 6.476393170578021e-09,
|
||
|
|
"loss": 0.3538,
|
||
|
|
"step": 1226
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.967611336032388,
|
||
|
|
"grad_norm": 0.21973251934476282,
|
||
|
|
"learning_rate": 5.11720812557348e-09,
|
||
|
|
"loss": 0.383,
|
||
|
|
"step": 1227
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.97165991902834,
|
||
|
|
"grad_norm": 0.21187157850872798,
|
||
|
|
"learning_rate": 3.917901630563492e-09,
|
||
|
|
"loss": 0.3391,
|
||
|
|
"step": 1228
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.9757085020242915,
|
||
|
|
"grad_norm": 0.20778380586964065,
|
||
|
|
"learning_rate": 2.878483275170485e-09,
|
||
|
|
"loss": 0.3758,
|
||
|
|
"step": 1229
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.979757085020243,
|
||
|
|
"grad_norm": 0.20770995650646387,
|
||
|
|
"learning_rate": 1.998961370555108e-09,
|
||
|
|
"loss": 0.3351,
|
||
|
|
"step": 1230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.983805668016195,
|
||
|
|
"grad_norm": 0.21165800109265864,
|
||
|
|
"learning_rate": 1.2793429493518361e-09,
|
||
|
|
"loss": 0.3562,
|
||
|
|
"step": 1231
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.987854251012146,
|
||
|
|
"grad_norm": 0.20757703136943653,
|
||
|
|
"learning_rate": 7.196337656112384e-10,
|
||
|
|
"loss": 0.3503,
|
||
|
|
"step": 1232
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.991902834008097,
|
||
|
|
"grad_norm": 0.2051477933500928,
|
||
|
|
"learning_rate": 3.198382947511292e-10,
|
||
|
|
"loss": 0.3646,
|
||
|
|
"step": 1233
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.995951417004049,
|
||
|
|
"grad_norm": 0.2082728527602629,
|
||
|
|
"learning_rate": 7.995973352770137e-11,
|
||
|
|
"loss": 0.3765,
|
||
|
|
"step": 1234
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 5.0,
|
||
|
|
"grad_norm": 0.21010748134866428,
|
||
|
|
"learning_rate": 0.0,
|
||
|
|
"loss": 0.3403,
|
||
|
|
"step": 1235
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 5.0,
|
||
|
|
"step": 1235,
|
||
|
|
"total_flos": 4.3365838011792097e+18,
|
||
|
|
"train_loss": 0.4892277446353001,
|
||
|
|
"train_runtime": 29008.2393,
|
||
|
|
"train_samples_per_second": 5.447,
|
||
|
|
"train_steps_per_second": 0.043
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 1,
|
||
|
|
"max_steps": 1235,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 5,
|
||
|
|
"save_steps": 500,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": true
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 4.3365838011792097e+18,
|
||
|
|
"train_batch_size": 1,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|