b2_code_askllm / trainer_state.json
neginr's picture
End of training
2f8d0bd verified
{
"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": 6.365932069908804,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.0183,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 6.344555285224076,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0131,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.414121681053585,
"learning_rate": 9.67741935483871e-07,
"loss": 1.0222,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.374985055080899,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.045,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 5.949131265098381,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.0083,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 5.8062597256903805,
"learning_rate": 1.935483870967742e-06,
"loss": 1.0061,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 4.601490280703054,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.9898,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.31518678584465,
"learning_rate": 2.580645161290323e-06,
"loss": 0.979,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.8212093264879137,
"learning_rate": 2.903225806451613e-06,
"loss": 0.9371,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.6261792094946705,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9335,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.4748741132429304,
"learning_rate": 3.548387096774194e-06,
"loss": 0.9272,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.942978757215843,
"learning_rate": 3.870967741935484e-06,
"loss": 0.9245,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.282669491198364,
"learning_rate": 4.193548387096774e-06,
"loss": 0.9149,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 4.060789468493325,
"learning_rate": 4.516129032258065e-06,
"loss": 0.8993,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.6956654476626394,
"learning_rate": 4.838709677419355e-06,
"loss": 0.8906,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.774662628146736,
"learning_rate": 5.161290322580646e-06,
"loss": 0.8596,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.6038794828165392,
"learning_rate": 5.483870967741935e-06,
"loss": 0.8468,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 1.999971452622986,
"learning_rate": 5.806451612903226e-06,
"loss": 0.8244,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.5317242376306208,
"learning_rate": 6.129032258064517e-06,
"loss": 0.8066,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.2037656238524517,
"learning_rate": 6.451612903225806e-06,
"loss": 0.7959,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.1498584778478609,
"learning_rate": 6.774193548387097e-06,
"loss": 0.7977,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.209130253983395,
"learning_rate": 7.096774193548388e-06,
"loss": 0.7786,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.0826886953438783,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.7719,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 0.9388630096442024,
"learning_rate": 7.741935483870968e-06,
"loss": 0.7577,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.9544711288499865,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7658,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.917185436740453,
"learning_rate": 8.387096774193549e-06,
"loss": 0.7526,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.8441163504949355,
"learning_rate": 8.70967741935484e-06,
"loss": 0.741,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.711976009928759,
"learning_rate": 9.03225806451613e-06,
"loss": 0.7325,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.6834464026307774,
"learning_rate": 9.35483870967742e-06,
"loss": 0.7328,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.7769231050391968,
"learning_rate": 9.67741935483871e-06,
"loss": 0.7278,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.7570294281509553,
"learning_rate": 1e-05,
"loss": 0.7154,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.6777233759159024,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.7178,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.6650249849708503,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.7085,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.7374440391015477,
"learning_rate": 1.096774193548387e-05,
"loss": 0.7084,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.7354590166290027,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.7216,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.6332544965589864,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.691,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.7398560628129567,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.7021,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.534231898872087,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.6906,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.6526983571916968,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.6948,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.5772711790915424,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.6894,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.5053938486325497,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6873,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.5986159965286046,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.6801,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.5767733647441332,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.6983,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.6416847287935327,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.6912,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.5156245893394785,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.6696,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.5233501879603067,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6878,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.476285301568637,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.6653,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.4479261740286652,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6868,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.5559021947661293,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.679,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.45688732420387623,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6681,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.5331703893493608,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.676,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.5455912169456584,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6593,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.5282062901804013,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.694,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.560482015279053,
"learning_rate": 1.741935483870968e-05,
"loss": 0.6695,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.5441386356963763,
"learning_rate": 1.774193548387097e-05,
"loss": 0.667,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.5499308279947338,
"learning_rate": 1.806451612903226e-05,
"loss": 0.655,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.48209591094965126,
"learning_rate": 1.838709677419355e-05,
"loss": 0.6595,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.6193931315155232,
"learning_rate": 1.870967741935484e-05,
"loss": 0.6499,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.785202367868218,
"learning_rate": 1.903225806451613e-05,
"loss": 0.6576,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.7611582365109675,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6464,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.6264082565264656,
"learning_rate": 1.967741935483871e-05,
"loss": 0.659,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.672452177055808,
"learning_rate": 2e-05,
"loss": 0.6549,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.8364723038226105,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.6531,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.8782631585005579,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.6476,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.6998931724706275,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.6389,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.6277206489788999,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.6603,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.8170822394375091,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6547,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.955200502518709,
"learning_rate": 2.193548387096774e-05,
"loss": 0.6571,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.7790409174434272,
"learning_rate": 2.225806451612903e-05,
"loss": 0.6375,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.6024659771590815,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.6432,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.8075517150634371,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6421,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.9736813401021198,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.6429,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 1.174660377090872,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.6529,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.9123466910129738,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.6461,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.7724217266801572,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.6428,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.664761455041938,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.6446,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.876328923956443,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.6333,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.6779916624273734,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.6321,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.0468930674636825,
"learning_rate": 2.548387096774194e-05,
"loss": 0.6395,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.2499615067917944,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.6503,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.8482038019427856,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.6473,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.0919540952012097,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.6423,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.3431609307213608,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6283,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.9153560338469113,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.648,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.9834930216381309,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.639,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.9119720776789618,
"learning_rate": 2.774193548387097e-05,
"loss": 0.6378,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.9864436497889687,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.6534,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.055435330580234,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.6408,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.7238921425463404,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.6223,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.2418432928987555,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.6123,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.1253700098689394,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6304,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.9053423012502596,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.6328,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.0504249928079226,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.6337,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.5625966512250586,
"learning_rate": 3.032258064516129e-05,
"loss": 0.6324,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.6198442488762552,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.6173,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.1514848228600076,
"learning_rate": 3.096774193548387e-05,
"loss": 0.6352,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.5039142730760393,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.6492,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.9725826474837616,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6254,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.47417561781738,
"learning_rate": 3.193548387096774e-05,
"loss": 0.627,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.0599310053692177,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.6352,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.8157723100856893,
"learning_rate": 3.258064516129033e-05,
"loss": 0.6405,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.118298409841623,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.6343,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 2.327399245958186,
"learning_rate": 3.322580645161291e-05,
"loss": 0.6452,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 2.0681848451706903,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.638,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.6988221610151262,
"learning_rate": 3.387096774193549e-05,
"loss": 0.6302,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.4132252868136983,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.6135,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.7545347048283029,
"learning_rate": 3.451612903225807e-05,
"loss": 0.6167,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.0234462821153367,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6326,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.7739141949460313,
"learning_rate": 3.516129032258065e-05,
"loss": 0.645,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.2314055622931084,
"learning_rate": 3.548387096774194e-05,
"loss": 0.6266,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.5231141611970918,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.6272,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.4766720794586021,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6213,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.3363196073834194,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.621,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.3126435538505712,
"learning_rate": 3.67741935483871e-05,
"loss": 0.6265,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.0979805761541852,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6375,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.2190266678096948,
"learning_rate": 3.741935483870968e-05,
"loss": 0.6269,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 1.0883978637003502,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.6223,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.933619610974311,
"learning_rate": 3.806451612903226e-05,
"loss": 0.6273,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.31266810651966,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.6335,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.8057147751540732,
"learning_rate": 3.870967741935484e-05,
"loss": 0.6037,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.2691919605028887,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.642,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 0.7920267610775633,
"learning_rate": 3.935483870967742e-05,
"loss": 0.6228,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.2632034105259216,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.6203,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.8681353330406992,
"learning_rate": 4e-05,
"loss": 0.6203,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.139388962213116,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.6252,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.1412272329006083,
"learning_rate": 3.999968016170525e-05,
"loss": 0.6254,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.9327312298966958,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.6199,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 1.1077506028676176,
"learning_rate": 3.999872065705065e-05,
"loss": 0.6275,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.2462599008179664,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6149,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.6763520943041988,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.6288,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.0270730469470335,
"learning_rate": 3.999608209836944e-05,
"loss": 0.6288,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.7610817908891001,
"learning_rate": 3.999488279187443e-05,
"loss": 0.6249,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.8546000226917787,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.6296,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 0.9152582396355989,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.6166,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.7101934942943119,
"learning_rate": 3.999032564583976e-05,
"loss": 0.6179,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.7557413367492869,
"learning_rate": 3.998848689546591e-05,
"loss": 0.6164,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.7957920680035879,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.6169,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.5791105663993673,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.6342,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.8257362748592121,
"learning_rate": 3.998201174509306e-05,
"loss": 0.6231,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.76834505096332,
"learning_rate": 3.997953378607961e-05,
"loss": 0.6238,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.8100043931517156,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.6166,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.9528498707483495,
"learning_rate": 3.997409862168454e-05,
"loss": 0.6059,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.8526402316185355,
"learning_rate": 3.997114145976237e-05,
"loss": 0.6214,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.639590800170498,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.6057,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.6845011192516964,
"learning_rate": 3.996474809469545e-05,
"loss": 0.6144,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.5499823667753353,
"learning_rate": 3.996131194267189e-05,
"loss": 0.6139,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.5897197315357695,
"learning_rate": 3.995771618052995e-05,
"loss": 0.6059,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.443043660399808,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6133,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.537147475302268,
"learning_rate": 3.995004594217343e-05,
"loss": 0.6102,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.6148441249576211,
"learning_rate": 3.994597152728987e-05,
"loss": 0.6271,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.5153083591318546,
"learning_rate": 3.994173762494947e-05,
"loss": 0.6107,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.6778682212522537,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.6182,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.5924356360533881,
"learning_rate": 3.993279149458988e-05,
"loss": 0.6112,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.6245306111165102,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.6111,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.583722294203964,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6087,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.7115397027482657,
"learning_rate": 3.991817703090965e-05,
"loss": 0.6019,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.9727313575685181,
"learning_rate": 3.991298695938038e-05,
"loss": 0.623,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 1.0961071330702266,
"learning_rate": 3.9907637664138e-05,
"loss": 0.6209,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.8734159617315277,
"learning_rate": 3.990212918795534e-05,
"loss": 0.6232,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.8694812496817182,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.6059,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.6178134891910002,
"learning_rate": 3.989063487022416e-05,
"loss": 0.6,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.6869849991828404,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.6107,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.7324977224570017,
"learning_rate": 3.987850437381915e-05,
"loss": 0.6185,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.6963552491488963,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6043,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.7289493893372786,
"learning_rate": 3.986573808672002e-05,
"loss": 0.6006,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.5993441533236873,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.616,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.786211426208705,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6101,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.6543153346942473,
"learning_rate": 3.984539744726554e-05,
"loss": 0.6143,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.5234038453788133,
"learning_rate": 3.983829979402038e-05,
"loss": 0.6138,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.6683907436225927,
"learning_rate": 3.983104351425871e-05,
"loss": 0.6178,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.7373373473803458,
"learning_rate": 3.982362866600154e-05,
"loss": 0.6112,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.539667021047024,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.6079,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.5701389222303728,
"learning_rate": 3.980832350242388e-05,
"loss": 0.6074,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.6693716050485724,
"learning_rate": 3.980043330948306e-05,
"loss": 0.6007,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.5131670096722384,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6032,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.9093308551515558,
"learning_rate": 3.978417801674579e-05,
"loss": 0.6085,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.8968288054694145,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.6042,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.8003767108115308,
"learning_rate": 3.976728995023254e-05,
"loss": 0.5964,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.5906927463516863,
"learning_rate": 3.975860879481514e-05,
"loss": 0.6127,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.6498911744825713,
"learning_rate": 3.974976965008833e-05,
"loss": 0.6081,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.7250681767176947,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.6095,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.7881194488444259,
"learning_rate": 3.973161767667946e-05,
"loss": 0.6118,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.4310529285716044,
"learning_rate": 3.972230499314009e-05,
"loss": 0.6092,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.7301439048555269,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.6045,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.617725302446067,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.5998,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.5772728458269555,
"learning_rate": 3.969342105253098e-05,
"loss": 0.602,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.5843572025594007,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.6163,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.524372956797119,
"learning_rate": 3.967337762346568e-05,
"loss": 0.609,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.6925242670822201,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.6108,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.7603967627324496,
"learning_rate": 3.965270496444528e-05,
"loss": 0.5888,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.601381175541688,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.6025,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.6110968544458918,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.5969,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.38749556406210245,
"learning_rate": 3.962051762160151e-05,
"loss": 0.6025,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.48749036215278707,
"learning_rate": 3.960947462140638e-05,
"loss": 0.6084,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.5099516966588166,
"learning_rate": 3.959827482437472e-05,
"loss": 0.5932,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.6301909525011713,
"learning_rate": 3.958691832005981e-05,
"loss": 0.6051,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.6066888196952726,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.5854,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.4481360998105811,
"learning_rate": 3.956373555405774e-05,
"loss": 0.5967,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.4405784045203193,
"learning_rate": 3.955190947773937e-05,
"loss": 0.5887,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.5586533275891751,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.6039,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.5826351167361246,
"learning_rate": 3.952778841127214e-05,
"loss": 0.6025,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.5331732542287889,
"learning_rate": 3.95154936139947e-05,
"loss": 0.6041,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.5067883723374395,
"learning_rate": 3.950304277135034e-05,
"loss": 0.6059,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.4022284620660996,
"learning_rate": 3.94904359828957e-05,
"loss": 0.5951,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.5767523802515382,
"learning_rate": 3.94776733494343e-05,
"loss": 0.6097,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.48919752482521073,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.5877,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.4585393708177223,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.6034,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.444215832230063,
"learning_rate": 3.943845140573213e-05,
"loss": 0.596,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.5196924557957775,
"learning_rate": 3.942506642518952e-05,
"loss": 0.5863,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.6451630280760132,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.5998,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.6664598196354372,
"learning_rate": 3.939783060543167e-05,
"loss": 0.6052,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.7804916159986954,
"learning_rate": 3.938397998399333e-05,
"loss": 0.5943,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.5373882372210023,
"learning_rate": 3.936997436876756e-05,
"loss": 0.6023,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.572070332984624,
"learning_rate": 3.935581387174289e-05,
"loss": 0.5959,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.54052506155776,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.6002,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.44374348020523724,
"learning_rate": 3.932702868644224e-05,
"loss": 0.5931,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.732124229848298,
"learning_rate": 3.931240422833181e-05,
"loss": 0.5971,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.6313026548897355,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.591,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.4650977331834361,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.6063,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6076013441700673,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.5977,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.49165059802112193,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.5994,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.5832092592502193,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.5846,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.5689240282184826,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.6001,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.5160798909086958,
"learning_rate": 3.92057159243286e-05,
"loss": 0.5879,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.6750620987491187,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.5939,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.7146770150607378,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.6055,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.6574518167632445,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.5867,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.5150493607680833,
"learning_rate": 3.914136997982874e-05,
"loss": 0.6061,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.6267917532000336,
"learning_rate": 3.912490053436079e-05,
"loss": 0.5975,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.8956815654908546,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.5833,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.8468393624133076,
"learning_rate": 3.909150300975178e-05,
"loss": 0.5955,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.6688427779519989,
"learning_rate": 3.907457519765642e-05,
"loss": 0.5943,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.5627388004827203,
"learning_rate": 3.905749486576607e-05,
"loss": 0.5907,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.7107005727274045,
"learning_rate": 3.90402621506546e-05,
"loss": 0.6001,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.7451630005569735,
"learning_rate": 3.902287719011435e-05,
"loss": 0.6011,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.7609054054058997,
"learning_rate": 3.900534012315499e-05,
"loss": 0.6017,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5314532568777672,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.581,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.473602587924691,
"learning_rate": 3.896981023209777e-05,
"loss": 0.6034,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.616691988987422,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.6021,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.6656483820062816,
"learning_rate": 3.893367361386491e-05,
"loss": 0.6,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.5576953010302002,
"learning_rate": 3.891537814248417e-05,
"loss": 0.5779,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.49770594415947844,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5806,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.5435301599467727,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.5864,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.47879322401574786,
"learning_rate": 3.885958483839049e-05,
"loss": 0.5847,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.4448769429861536,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.5864,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.48321247748536594,
"learning_rate": 3.882163505079171e-05,
"loss": 0.5919,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.4366653812767668,
"learning_rate": 3.880243433489132e-05,
"loss": 0.583,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.5669663050990057,
"learning_rate": 3.878308327522702e-05,
"loss": 0.5575,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.5300987136642419,
"learning_rate": 3.876358202652936e-05,
"loss": 0.5493,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.5710901220152013,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.5372,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.49970870838683545,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.5547,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.6108388630630515,
"learning_rate": 3.870417871154801e-05,
"loss": 0.5476,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.5168990061025039,
"learning_rate": 3.868407827802192e-05,
"loss": 0.5364,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.5846628137053826,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.5386,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.48395107789128083,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.5366,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.438719661049156,
"learning_rate": 3.862288124195319e-05,
"loss": 0.5441,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.5707725869691566,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.5403,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.5748801856919622,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.5385,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.4755911396300031,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.5485,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.5977273355134618,
"learning_rate": 3.853920126717703e-05,
"loss": 0.5515,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.6328509074513967,
"learning_rate": 3.851791025492574e-05,
"loss": 0.5326,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.5624355784444264,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.537,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.5429464329653789,
"learning_rate": 3.847488419569866e-05,
"loss": 0.5428,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.6164192446413845,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.5521,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.4887304196638157,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.5377,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.4829260599763923,
"learning_rate": 3.840923760917199e-05,
"loss": 0.5506,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.4693219347744369,
"learning_rate": 3.838706077964473e-05,
"loss": 0.5344,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.5687081050354862,
"learning_rate": 3.836473692766944e-05,
"loss": 0.5314,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.46793017246201113,
"learning_rate": 3.834226623174705e-05,
"loss": 0.5383,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.60264583380415,
"learning_rate": 3.831964887155264e-05,
"loss": 0.54,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.5389143964375449,
"learning_rate": 3.829688502793403e-05,
"loss": 0.5372,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.6984298378168058,
"learning_rate": 3.827397488291029e-05,
"loss": 0.5447,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.6390217609694538,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.5338,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.6380342929287056,
"learning_rate": 3.822771642257145e-05,
"loss": 0.549,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.688260367831964,
"learning_rate": 3.820436847713776e-05,
"loss": 0.5493,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.6047700901110771,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.5336,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.6163712529652209,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.5361,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.7163690515587048,
"learning_rate": 3.813345202354145e-05,
"loss": 0.5341,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.6345073700826702,
"learning_rate": 3.810952296329563e-05,
"loss": 0.5404,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.468868223737519,
"learning_rate": 3.808544909978676e-05,
"loss": 0.5373,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.5557704345307953,
"learning_rate": 3.806123062550882e-05,
"loss": 0.5521,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.6280499083420101,
"learning_rate": 3.803686773411208e-05,
"loss": 0.5352,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.5295264185035361,
"learning_rate": 3.801236062040158e-05,
"loss": 0.5612,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.4635220026380816,
"learning_rate": 3.798770948033553e-05,
"loss": 0.5406,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.4080892461307958,
"learning_rate": 3.796291451102381e-05,
"loss": 0.5452,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.5479396539876062,
"learning_rate": 3.793797591072631e-05,
"loss": 0.5327,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.484539360483584,
"learning_rate": 3.791289387885144e-05,
"loss": 0.5576,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.534274328072437,
"learning_rate": 3.788766861595443e-05,
"loss": 0.5527,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.4714873830624942,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.5428,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.40242245818006944,
"learning_rate": 3.783678920503983e-05,
"loss": 0.5524,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.4410586534238311,
"learning_rate": 3.781113546385265e-05,
"loss": 0.5353,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.49565634104878314,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.5352,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.4173680714196532,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.5509,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.4853769999472759,
"learning_rate": 3.77333205623025e-05,
"loss": 0.5467,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.6106042923895957,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.5554,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.538223759769638,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.5373,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.46099786485139255,
"learning_rate": 3.765422951112704e-05,
"loss": 0.5455,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.4843675280857737,
"learning_rate": 3.762758321970216e-05,
"loss": 0.5299,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.43117311437630035,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.5416,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.583647793491073,
"learning_rate": 3.757386800198538e-05,
"loss": 0.5524,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.5139995840221059,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.5355,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.4618485340001734,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.5278,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.5587303509765852,
"learning_rate": 3.749224181796305e-05,
"loss": 0.5477,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5173343817583607,
"learning_rate": 3.746475306375545e-05,
"loss": 0.5364,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.3626026217462254,
"learning_rate": 3.743712466184774e-05,
"loss": 0.5435,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.500390013101597,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.5518,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.6356181392270599,
"learning_rate": 3.73814497997107e-05,
"loss": 0.543,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.42247652818230164,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.5322,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.45822289724594006,
"learning_rate": 3.732521901224724e-05,
"loss": 0.5367,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.4303050888310358,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.54,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.5443356495994559,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.5491,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.5351763210519035,
"learning_rate": 3.723983441007882e-05,
"loss": 0.5243,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.4029325292124675,
"learning_rate": 3.721109687296781e-05,
"loss": 0.5257,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.47946310351455634,
"learning_rate": 3.718222171638484e-05,
"loss": 0.5375,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.43433824321640335,
"learning_rate": 3.715320917121488e-05,
"loss": 0.5342,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.5456464003768422,
"learning_rate": 3.712405946944149e-05,
"loss": 0.5504,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.4962370803611603,
"learning_rate": 3.709477284414488e-05,
"loss": 0.5392,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.38032547133390127,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.5583,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.4563738689400018,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.5242,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.3947718235160555,
"learning_rate": 3.70060937743295e-05,
"loss": 0.5488,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.41052033174567154,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.5594,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.4868289808247716,
"learning_rate": 3.694629409915555e-05,
"loss": 0.5446,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.5738444749830752,
"learning_rate": 3.691619088858401e-05,
"loss": 0.544,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.46009192732605014,
"learning_rate": 3.68859524166009e-05,
"loss": 0.549,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.5053082330317,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.5295,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.5081886666543652,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.5438,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.5660319003437974,
"learning_rate": 3.67944278554384e-05,
"loss": 0.5422,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.5247289466386763,
"learning_rate": 3.676365076645557e-05,
"loss": 0.5559,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.4600536906874467,
"learning_rate": 3.673273963576791e-05,
"loss": 0.5416,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.3968708046291205,
"learning_rate": 3.670169471054001e-05,
"loss": 0.5355,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.4949508533980698,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.5407,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.4815832457165761,
"learning_rate": 3.663920447046888e-05,
"loss": 0.5349,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.40137739026733266,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.5214,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.52621896697138,
"learning_rate": 3.657618204491937e-05,
"loss": 0.5444,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.3813661398251934,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.5449,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.4253907018111229,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.5369,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.45823898083116643,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.5382,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.43876746146375567,
"learning_rate": 3.644854871713611e-05,
"loss": 0.534,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.36989817834232935,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.5342,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.44406538944461127,
"learning_rate": 3.638394189710493e-05,
"loss": 0.5457,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.4050207865117354,
"learning_rate": 3.635144184933412e-05,
"loss": 0.5383,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.3985355842815619,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.5406,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.4426549924874509,
"learning_rate": 3.628604977762752e-05,
"loss": 0.5404,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.5204271429574534,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.5375,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.40546424210380716,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.5444,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.7042207737656191,
"learning_rate": 3.618698565901727e-05,
"loss": 0.5414,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.494263649608411,
"learning_rate": 3.615370507164654e-05,
"loss": 0.5399,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.597293613514033,
"learning_rate": 3.61202953196805e-05,
"loss": 0.5443,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.6691636971996244,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.5266,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.43414671377846864,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.5405,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.5433236757088997,
"learning_rate": 3.601929375279662e-05,
"loss": 0.552,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.4721730194093908,
"learning_rate": 3.598537002418012e-05,
"loss": 0.5208,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.43892109138790775,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.5403,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.42952421793802253,
"learning_rate": 3.591713938344416e-05,
"loss": 0.5351,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.4847790171501976,
"learning_rate": 3.588283301689507e-05,
"loss": 0.5353,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.42897951340646173,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.5332,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.5088359392932278,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.5361,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.44722790356381403,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.5409,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.5000527238432252,
"learning_rate": 3.574434032194994e-05,
"loss": 0.529,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.4523297930844468,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.5395,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.4390719401048928,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.5279,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.45825527578861414,
"learning_rate": 3.563914797637216e-05,
"loss": 0.5307,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.4218500754173183,
"learning_rate": 3.560383338592649e-05,
"loss": 0.5295,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.4142242314509717,
"learning_rate": 3.556839402764487e-05,
"loss": 0.5569,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.45199658041728336,
"learning_rate": 3.553283018489946e-05,
"loss": 0.539,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.3498595589306179,
"learning_rate": 3.54971421420578e-05,
"loss": 0.5389,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.3979995173824424,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.5263,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.3934707834251953,
"learning_rate": 3.54253945985191e-05,
"loss": 0.5257,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.420028895957417,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.5606,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.4620506910718832,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.5424,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.3744708358248262,
"learning_rate": 3.531684894865866e-05,
"loss": 0.5396,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.42114804193332656,
"learning_rate": 3.528042173241128e-05,
"loss": 0.5461,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.4115225951970625,
"learning_rate": 3.524387233431892e-05,
"loss": 0.5393,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.3692224605522494,
"learning_rate": 3.520720104662958e-05,
"loss": 0.5463,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.37697025900142767,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.5498,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.4037406955307903,
"learning_rate": 3.513349397632281e-05,
"loss": 0.5217,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.40583397609324307,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.5408,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.4561958870612415,
"learning_rate": 3.505930287892533e-05,
"loss": 0.5423,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.3694509810305884,
"learning_rate": 3.502202656100099e-05,
"loss": 0.5511,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.37129173722891023,
"learning_rate": 3.498463012735256e-05,
"loss": 0.5401,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.36737065602832775,
"learning_rate": 3.494711387700094e-05,
"loss": 0.5326,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.3443113856144028,
"learning_rate": 3.490947810992505e-05,
"loss": 0.5437,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.4055428784756743,
"learning_rate": 3.48717231270595e-05,
"loss": 0.5429,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.30580856211619106,
"learning_rate": 3.483384923029211e-05,
"loss": 0.5415,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.375645311078997,
"learning_rate": 3.479585672246156e-05,
"loss": 0.5382,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.3905767332854254,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.5301,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.34376216287366945,
"learning_rate": 3.471951708970527e-05,
"loss": 0.5384,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.37835608453451025,
"learning_rate": 3.468117057518921e-05,
"loss": 0.5439,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.4217883315499564,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.5379,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.4371566837606642,
"learning_rate": 3.460412568296731e-05,
"loss": 0.5337,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.3640377206983907,
"learning_rate": 3.456542792131039e-05,
"loss": 0.5466,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.416764396773084,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.531,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.458980238916017,
"learning_rate": 3.448768331403348e-05,
"loss": 0.5362,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.4533228355039324,
"learning_rate": 3.44486370900573e-05,
"loss": 0.5331,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.4779918207965055,
"learning_rate": 3.440947533516396e-05,
"loss": 0.5462,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.4549280040806854,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.5308,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.6134765568690345,
"learning_rate": 3.43308064860925e-05,
"loss": 0.5313,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.5199197491300807,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.5424,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.458789650164132,
"learning_rate": 3.425167928295015e-05,
"loss": 0.5305,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.43836478031187004,
"learning_rate": 3.421194458890411e-05,
"loss": 0.531,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.4644682080933282,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.5372,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.4638341633310999,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.5448,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.4760886991282732,
"learning_rate": 3.409205995219561e-05,
"loss": 0.545,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.45885768279351785,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.5281,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.5147363956888559,
"learning_rate": 3.40115729298209e-05,
"loss": 0.5315,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.5412955198370947,
"learning_rate": 3.397116120327025e-05,
"loss": 0.5359,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.39591564264934437,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.5382,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.4561488439526256,
"learning_rate": 3.389000293509524e-05,
"loss": 0.5317,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.46218121597884876,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.543,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.3746117124138126,
"learning_rate": 3.380840041143494e-05,
"loss": 0.539,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.4737603906647078,
"learning_rate": 3.376743336885792e-05,
"loss": 0.5301,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.48837095006034653,
"learning_rate": 3.372635624225055e-05,
"loss": 0.5444,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.45280394513912253,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.5403,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.5163509296144241,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.5361,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.48134682539509066,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.5383,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.41955776056191574,
"learning_rate": 3.356095347769799e-05,
"loss": 0.5314,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.4167257234049371,
"learning_rate": 3.351933087522552e-05,
"loss": 0.5095,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.38883904822284104,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.5424,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.3516986831326666,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.5309,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.363148990087454,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.53,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.39372251087794286,
"learning_rate": 3.335176280432308e-05,
"loss": 0.5308,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.3479528723364203,
"learning_rate": 3.330960304618295e-05,
"loss": 0.5361,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.3910048575019358,
"learning_rate": 3.326733686481153e-05,
"loss": 0.5342,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.389312224546066,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.5311,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.36814119788233546,
"learning_rate": 3.318248658506012e-05,
"loss": 0.5372,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.4105109968287385,
"learning_rate": 3.313990316514069e-05,
"loss": 0.5257,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.3885841847524013,
"learning_rate": 3.309721467890571e-05,
"loss": 0.548,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.4095184510589649,
"learning_rate": 3.305442146769115e-05,
"loss": 0.5253,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.43924238202463023,
"learning_rate": 3.30115238736704e-05,
"loss": 0.5562,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.35628490569196014,
"learning_rate": 3.296852223985148e-05,
"loss": 0.5358,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.3851770879955261,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.543,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.39786115488382096,
"learning_rate": 3.288220822900796e-05,
"loss": 0.5354,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.3854007564686984,
"learning_rate": 3.283889654214788e-05,
"loss": 0.5264,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.3629307437763341,
"learning_rate": 3.279548219581317e-05,
"loss": 0.5212,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.3689713781151831,
"learning_rate": 3.275196553714379e-05,
"loss": 0.5496,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.37073176755443044,
"learning_rate": 3.270834691409778e-05,
"loss": 0.5378,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.4215209104012034,
"learning_rate": 3.266462667544849e-05,
"loss": 0.5242,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.3229941300698283,
"learning_rate": 3.262080517078178e-05,
"loss": 0.5339,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.40030177577107007,
"learning_rate": 3.257688275049323e-05,
"loss": 0.532,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.496334686312838,
"learning_rate": 3.253285976578535e-05,
"loss": 0.5496,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.3262860414695905,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.5465,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.3916641708102616,
"learning_rate": 3.244451351193935e-05,
"loss": 0.5354,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.4173637383536746,
"learning_rate": 3.240019094921551e-05,
"loss": 0.5313,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.3531929402164988,
"learning_rate": 3.235576923489527e-05,
"loss": 0.5275,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.38223038391783537,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.5388,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.38003457026340076,
"learning_rate": 3.226662977303494e-05,
"loss": 0.542,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.34026479812941185,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.5533,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.44252923918901366,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.5348,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.33859800542785295,
"learning_rate": 3.21321858487569e-05,
"loss": 0.5426,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.37335544922770547,
"learning_rate": 3.208717671149931e-05,
"loss": 0.5355,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.3678704656416537,
"learning_rate": 3.204207092549882e-05,
"loss": 0.5274,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.31120502852009124,
"learning_rate": 3.19968688514201e-05,
"loss": 0.5331,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.3296008625447762,
"learning_rate": 3.195157085069774e-05,
"loss": 0.5315,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.3183731213610909,
"learning_rate": 3.190617728553332e-05,
"loss": 0.5479,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.34227184651906756,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.5334,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.3621775921163093,
"learning_rate": 3.181510491450254e-05,
"loss": 0.546,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.36741074731888207,
"learning_rate": 3.176942683684842e-05,
"loss": 0.5345,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.3739512103120928,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.5373,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.38065499248246715,
"learning_rate": 3.167778872346327e-05,
"loss": 0.5317,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.3614333523819024,
"learning_rate": 3.163182942046815e-05,
"loss": 0.5302,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.32400899527874333,
"learning_rate": 3.158577710967494e-05,
"loss": 0.5368,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.36843879932089657,
"learning_rate": 3.153963215931669e-05,
"loss": 0.5104,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.3487272672657628,
"learning_rate": 3.14933949383672e-05,
"loss": 0.5395,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.3661106141652723,
"learning_rate": 3.144706581653805e-05,
"loss": 0.5245,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.42187293931732794,
"learning_rate": 3.140064516427566e-05,
"loss": 0.5541,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.3603309399259442,
"learning_rate": 3.135413335275833e-05,
"loss": 0.533,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.3417070254213681,
"learning_rate": 3.130753075389327e-05,
"loss": 0.5376,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.3662181921719701,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.5296,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.39132544373291145,
"learning_rate": 3.121405468537547e-05,
"loss": 0.5333,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.35840904876822127,
"learning_rate": 3.116718196315488e-05,
"loss": 0.5333,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.377621434814045,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.5197,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.38528583576797665,
"learning_rate": 3.107316901675252e-05,
"loss": 0.5421,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.43051080298110067,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.5433,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.3802368467719489,
"learning_rate": 3.097880190800057e-05,
"loss": 0.531,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.423477943236433,
"learning_rate": 3.093148648549788e-05,
"loss": 0.5325,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.7258360253346451,
"learning_rate": 3.088408365512055e-05,
"loss": 0.5227,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.47462507273800514,
"learning_rate": 3.083659379590034e-05,
"loss": 0.5344,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.5224949941638604,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.5322,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.6909001518241397,
"learning_rate": 3.074135451053475e-05,
"loss": 0.5292,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.40551292048688686,
"learning_rate": 3.069360584592015e-05,
"loss": 0.5223,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.4666925611016581,
"learning_rate": 3.064577167551816e-05,
"loss": 0.5322,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.4625397556286767,
"learning_rate": 3.059785238180953e-05,
"loss": 0.5229,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.3613975742040179,
"learning_rate": 3.054984834795566e-05,
"loss": 0.5362,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.4412437997945228,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.5402,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.476789568423606,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.5276,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.28914764172407303,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.5421,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.4034948147427383,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.5369,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.39488141354379425,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.5362,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.34695933087823316,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.5301,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.48999824783956025,
"learning_rate": 3.021147971398632e-05,
"loss": 0.5222,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.40380822283185386,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.5329,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.3765920569471513,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.5224,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.4097112795333291,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.5419,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.4106121818172213,
"learning_rate": 3.001632138223332e-05,
"loss": 0.5318,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.5900395184113225,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.4715,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.710369335978151,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.4666,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.5207773785866732,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.4669,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.5011967678179439,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.4622,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.42553914844631374,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.4651,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.5101211207100113,
"learning_rate": 2.972118904869376e-05,
"loss": 0.4757,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.5542315290998314,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.4612,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.48088127660475743,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.4667,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.4510885206756992,
"learning_rate": 2.95725682050846e-05,
"loss": 0.4712,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.4512897065113964,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.4661,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.4833220987216583,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.4664,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.4727651187969584,
"learning_rate": 2.942325848714505e-05,
"loss": 0.4652,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.42261510581393974,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.4776,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.4693447427390034,
"learning_rate": 2.932334127129686e-05,
"loss": 0.472,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.40586910088062494,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.4812,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.38490788861983155,
"learning_rate": 2.922312585929131e-05,
"loss": 0.4599,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.40415558746094626,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.4597,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.3240634845638582,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.4631,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.39040474131813996,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.4708,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.3841046257531689,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.4582,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.40029127421561345,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.4501,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.4155857963631504,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.4528,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.42479639092486754,
"learning_rate": 2.887006998351372e-05,
"loss": 0.4489,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.4306637585774127,
"learning_rate": 2.881934649605535e-05,
"loss": 0.469,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.38639008562733657,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.4635,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.41082172091820024,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.463,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.36827561600435976,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.4708,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.4005165290147431,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.4663,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.3637924675981945,
"learning_rate": 2.856467940121025e-05,
"loss": 0.4736,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.39404997885200926,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.4769,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.4072953048499291,
"learning_rate": 2.846233032790149e-05,
"loss": 0.4638,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.2952597549171135,
"learning_rate": 2.84110540896642e-05,
"loss": 0.4691,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.4052937150979547,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.4528,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.3194878928969691,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.4579,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.3904746116052017,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.4536,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.35111411652681923,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.4701,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.4033259415428442,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.466,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.39053769675727096,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.4643,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.37844340176851005,
"learning_rate": 2.805026032173773e-05,
"loss": 0.4681,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.3994251878740622,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.4674,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.31763053085511805,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.4623,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.3939092796242195,
"learning_rate": 2.7894660620408e-05,
"loss": 0.4597,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.3109629932120362,
"learning_rate": 2.784266724883173e-05,
"loss": 0.469,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.3747832198313615,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.4619,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.3247977860425188,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.4559,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.33192908813104977,
"learning_rate": 2.768631254100716e-05,
"loss": 0.4695,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.32606313783564417,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.4619,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.3865387601371181,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.459,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.3247489752074455,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.4703,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.39332754706874873,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.4823,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.3451412759490647,
"learning_rate": 2.742449775134085e-05,
"loss": 0.4585,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.3731163245223199,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.4561,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.37297145406486376,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.4602,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.4212924490030045,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.4568,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.3984139198497382,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.4664,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.3730267820651919,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.471,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.42944060562778896,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.4596,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.36003781907037635,
"learning_rate": 2.705547548999996e-05,
"loss": 0.4641,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.3238806225873347,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.4771,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.4369000181515545,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.4744,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.36432127183550683,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.4666,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.3331865100998883,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.4738,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.3613175203015621,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.4574,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.31396999760186683,
"learning_rate": 2.67369650552155e-05,
"loss": 0.4593,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.4266113026829062,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.4751,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.3436288698356712,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.4725,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.3626137614300752,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.4685,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.35470869624621193,
"learning_rate": 2.652354173593614e-05,
"loss": 0.4644,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.3666483661697568,
"learning_rate": 2.647005443331317e-05,
"loss": 0.4749,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.3039187266964638,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.4664,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.3421015060977144,
"learning_rate": 2.636292505301542e-05,
"loss": 0.4587,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.3377250832612359,
"learning_rate": 2.630928383194431e-05,
"loss": 0.4744,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.32237127786805064,
"learning_rate": 2.625559216200781e-05,
"loss": 0.465,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.35861780255912645,
"learning_rate": 2.620185047252308e-05,
"loss": 0.4796,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.28758230028629833,
"learning_rate": 2.614805919320723e-05,
"loss": 0.484,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.34569312637804567,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.47,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.3072065820628191,
"learning_rate": 2.60403295859298e-05,
"loss": 0.465,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.28114238935268965,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.4644,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.3110482514040902,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.4685,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.2987325462787782,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.4667,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.32119898322439366,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.4605,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.3099267473019894,
"learning_rate": 2.577016790136453e-05,
"loss": 0.4617,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.3471929404457688,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.4598,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.32506041551510806,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.4637,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.32812354775658736,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.4636,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.32426445441831414,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.4568,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.3044533295538876,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.4711,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.31695430856109225,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.4665,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.30643915563626073,
"learning_rate": 2.539001591449889e-05,
"loss": 0.4739,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.34674035203036613,
"learning_rate": 2.53355326151989e-05,
"loss": 0.4646,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.371548388317074,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.4666,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.32354516068744243,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.4618,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.46541858130083824,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.4695,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.2954473959324836,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.4791,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.46462700255505107,
"learning_rate": 2.506248490185627e-05,
"loss": 0.474,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.3040441241735997,
"learning_rate": 2.500775217241923e-05,
"loss": 0.4858,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.4151559930521032,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.4519,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.31841226267556744,
"learning_rate": 2.4898167025948e-05,
"loss": 0.472,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.42326975723682975,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.4703,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.35658646607810773,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.4748,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.3662947618742797,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.4745,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.4130008030456172,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.4795,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.47449089416848017,
"learning_rate": 2.46235264422994e-05,
"loss": 0.4708,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.3962116125574668,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.4722,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.2825175872954479,
"learning_rate": 2.451340834511592e-05,
"loss": 0.4644,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.33078855910210836,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.4656,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.36441387654111274,
"learning_rate": 2.440314589184958e-05,
"loss": 0.48,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.3242143025837691,
"learning_rate": 2.434796163357541e-05,
"loss": 0.4598,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.40188493888839416,
"learning_rate": 2.429274260911588e-05,
"loss": 0.4609,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.4183800428888049,
"learning_rate": 2.423748926000084e-05,
"loss": 0.4673,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.32838288936230764,
"learning_rate": 2.418220202803459e-05,
"loss": 0.465,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.4646548468490286,
"learning_rate": 2.412688135529237e-05,
"loss": 0.4637,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.36255903277668927,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.4796,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.4448925146447858,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.4729,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.31290733500200507,
"learning_rate": 2.396072311715187e-05,
"loss": 0.4797,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.4056683554136508,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.4852,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.33601182726395035,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.4673,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.3424870590593991,
"learning_rate": 2.379427985200223e-05,
"loss": 0.4698,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.30416635808154435,
"learning_rate": 2.373873749394972e-05,
"loss": 0.4792,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.304410200438657,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.4614,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.3076647812053155,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.4639,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.28561388565604784,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.4758,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.3158323663087668,
"learning_rate": 2.351627355800233e-05,
"loss": 0.4732,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.28908826122454884,
"learning_rate": 2.346058617160256e-05,
"loss": 0.4815,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.2933880407213429,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.4622,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.3032050093219066,
"learning_rate": 2.334912883203469e-05,
"loss": 0.4687,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.2994804602793366,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.479,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.3026253959948541,
"learning_rate": 2.323756437450136e-05,
"loss": 0.4725,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.2953222664819672,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.4665,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.31547245671314805,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.4544,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.2981557979745489,
"learning_rate": 2.307002464849052e-05,
"loss": 0.4527,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.29689026415482345,
"learning_rate": 2.30141283818846e-05,
"loss": 0.4706,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.27700692233138796,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.4763,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.2618212327050192,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.4452,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.31861960694763697,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.4622,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.28236771634372815,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.4795,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.33276251891291386,
"learning_rate": 2.273429448107257e-05,
"loss": 0.4564,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.28794482143206085,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.4691,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.3407352805715064,
"learning_rate": 2.262220474347102e-05,
"loss": 0.4665,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.28512334430536107,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.4625,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.3399038919524068,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.4559,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.28388919420437914,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.478,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.27224938200122734,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.4748,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.27906870493186153,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.4761,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.2566281609553861,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.4872,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.3223052876109759,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.4629,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.2774658873994465,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.4777,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.2941107921800589,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.4704,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.31681761111004725,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.4793,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.27507787418994956,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.4806,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.3450839435651159,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.4628,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.2853538791743948,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.469,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.33714800174532383,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.4735,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.2862243928069121,
"learning_rate": 2.17791064266051e-05,
"loss": 0.4605,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.3418939340290592,
"learning_rate": 2.172276926861741e-05,
"loss": 0.4753,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.3179184222089352,
"learning_rate": 2.166641833541255e-05,
"loss": 0.457,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.3501687847233519,
"learning_rate": 2.161005407757109e-05,
"loss": 0.4658,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.318284642383645,
"learning_rate": 2.155367694578013e-05,
"loss": 0.4656,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.3055925124496277,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.4656,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.3275049310988119,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.4563,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.32355786267385095,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.4675,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.3146654422246908,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.4766,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.39811989076183124,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.4728,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.2922183689808242,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.448,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.4555238532700317,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.474,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.2898527276106555,
"learning_rate": 2.110225054402733e-05,
"loss": 0.4656,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.3630931861356178,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.4592,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.3026631116174656,
"learning_rate": 2.09892967807724e-05,
"loss": 0.4628,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.3622362989355077,
"learning_rate": 2.093280780774835e-05,
"loss": 0.4705,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.3030379810815664,
"learning_rate": 2.087631137601793e-05,
"loss": 0.4642,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.3509163250995733,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.4779,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.3433933097592043,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.46,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.31047235897847497,
"learning_rate": 2.070678184630458e-05,
"loss": 0.4704,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.30585469105354324,
"learning_rate": 2.06502600977305e-05,
"loss": 0.4673,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.3060479910309725,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.457,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.28727637971808956,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.4574,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.3335919408645046,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.4635,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.2671367705026204,
"learning_rate": 2.042412562887594e-05,
"loss": 0.4651,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.35731721338616373,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.4652,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.2895551252166923,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.4678,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.34815292780576335,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.4498,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.293153763295397,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.4555,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.2961301819533101,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.4695,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.2783158044705876,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.4728,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.27654474244484745,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.4688,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.28318424911785944,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.4422,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.30449709429272676,
"learning_rate": 1.99151687695709e-05,
"loss": 0.4481,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.2837681375750007,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.4686,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.3091283394188403,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.4792,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.2975042384447033,
"learning_rate": 1.974551241345447e-05,
"loss": 0.4766,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.3078037421501495,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.4604,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.34143269877509064,
"learning_rate": 1.963241759676897e-05,
"loss": 0.4661,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.30339668690392946,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.4571,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.29895231871189837,
"learning_rate": 1.951933453677638e-05,
"loss": 0.4664,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.3362588705033712,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.4668,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.31023004714244135,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.4732,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.3382494600849527,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.4738,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.3078920451867756,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.4744,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3321726580418784,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.4556,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.31706871771484835,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.4618,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.3393664753266988,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.4763,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.2886659061336517,
"learning_rate": 1.906719219225165e-05,
"loss": 0.4694,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.3453598556629049,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.4735,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.2948426448122251,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.4687,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.3128902453514714,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.4573,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.29298062160958904,
"learning_rate": 1.884128556891707e-05,
"loss": 0.4628,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.31732089580555245,
"learning_rate": 1.878483094691119e-05,
"loss": 0.4775,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.2793617032268231,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.4688,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.33819181360404027,
"learning_rate": 1.867195130360947e-05,
"loss": 0.4663,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.30150501934482843,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.4498,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.31869153942902223,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.4696,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.2874300348228681,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.4629,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.2873274754054015,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.4625,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.27705503376002444,
"learning_rate": 1.838994592242891e-05,
"loss": 0.452,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.283636951735376,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.4813,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.2813792254744984,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.4603,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.31018399262686314,
"learning_rate": 1.82208935733949e-05,
"loss": 0.4787,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.2990297831338351,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.4699,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.30119766740216214,
"learning_rate": 1.810826238483892e-05,
"loss": 0.4727,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.3081798856956723,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.4736,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.28771444422154324,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.4753,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.30473255928702664,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.4692,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.3284964674938114,
"learning_rate": 1.788318512320838e-05,
"loss": 0.4544,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.3138746197485922,
"learning_rate": 1.782695699829051e-05,
"loss": 0.4817,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.3219575102947003,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.4673,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.31129172736545546,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.4671,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.28257522048999584,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.4746,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.2869451995264101,
"learning_rate": 1.76022227484387e-05,
"loss": 0.4663,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.2937763527079879,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.4573,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.293709060889098,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.4741,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.30912739462668154,
"learning_rate": 1.743387180005809e-05,
"loss": 0.4604,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.3126191657998514,
"learning_rate": 1.737779525652899e-05,
"loss": 0.4681,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.3149439512708344,
"learning_rate": 1.732173968007914e-05,
"loss": 0.4707,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.3356901499246364,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.4546,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.33125925347928037,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.4719,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.2900005617527874,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.4717,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.3478453828405883,
"learning_rate": 1.709773600686015e-05,
"loss": 0.4728,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.2745739217504584,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.4826,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.3609423535462701,
"learning_rate": 1.69858716181154e-05,
"loss": 0.445,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.34697677372330465,
"learning_rate": 1.692997535150948e-05,
"loss": 0.4741,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.2941037866047232,
"learning_rate": 1.687410363273884e-05,
"loss": 0.4721,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.3915146664747339,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.4651,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.32480439189285915,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.4608,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.3288053362736928,
"learning_rate": 1.67066402299235e-05,
"loss": 0.4549,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.32571153793089624,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.4537,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.29753739127771534,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.477,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.31329029095192673,
"learning_rate": 1.653941382839745e-05,
"loss": 0.4629,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.4835614969450231,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.4137,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.3586686257305305,
"learning_rate": 1.642806717162757e-05,
"loss": 0.4131,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.6614006859261516,
"learning_rate": 1.637243646233718e-05,
"loss": 0.3894,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.44450446154355777,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.4162,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.48679280619653104,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.3938,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.4005055014379787,
"learning_rate": 1.620572014799777e-05,
"loss": 0.3988,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.4212960875185841,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.3942,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.5143393873179192,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.4012,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.38723040481086657,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.3975,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.4194410547575954,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.3998,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.417327510187978,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.394,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.38278319230005975,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.4017,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.4173777015585436,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.3862,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.3953705861527238,
"learning_rate": 1.576251073999917e-05,
"loss": 0.4036,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.3604806255980559,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.3861,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.37984082948327197,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.4072,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.39263864694921935,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.415,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.3600369289339034,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.4099,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.39737829095291705,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.4156,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.3513735011305332,
"learning_rate": 1.543151434154755e-05,
"loss": 0.4164,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.35084852052116505,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.4069,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.3823804847984202,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.4115,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.3312180375460843,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.3973,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.3619115268772131,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.3958,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.29957453817483215,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.3871,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.2965166135383191,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.4025,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.2949853320580859,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.3994,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.2756075809660654,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.4065,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.2906666053035034,
"learning_rate": 1.493751509814374e-05,
"loss": 0.3907,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.27301765677963946,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.3968,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.30236971974643256,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.3885,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.2830993061941779,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.3951,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.2892785919770606,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.4019,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.2690110800902146,
"learning_rate": 1.466446738480111e-05,
"loss": 0.404,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.29796519815346983,
"learning_rate": 1.460998408550112e-05,
"loss": 0.4005,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.2614789462762425,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.4068,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.29046279311013656,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.4081,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.27502990200019345,
"learning_rate": 1.444679451900437e-05,
"loss": 0.3808,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.26932013922478665,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.4097,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.2748873718132537,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.3934,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.2684307839097541,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.3914,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.2653665541136886,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.3821,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.2689320496085573,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.4045,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.2904947564510507,
"learning_rate": 1.412162598317602e-05,
"loss": 0.3929,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.2709984015074161,
"learning_rate": 1.406759321421907e-05,
"loss": 0.4006,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.29413265292187896,
"learning_rate": 1.401360788062871e-05,
"loss": 0.3932,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.333579156027734,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.4066,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.27656000574933565,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.3775,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.26251754262991334,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.4117,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.2899553302645965,
"learning_rate": 1.379814952747693e-05,
"loss": 0.4249,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.26036793962304006,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.4111,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3003798311401155,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.4009,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.2952386948180941,
"learning_rate": 1.363707494698459e-05,
"loss": 0.406,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.2712000911409667,
"learning_rate": 1.358348460369265e-05,
"loss": 0.4114,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.2837256079212392,
"learning_rate": 1.352994556668684e-05,
"loss": 0.4099,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.26443540082530825,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.403,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.29012426113876083,
"learning_rate": 1.342302312350678e-05,
"loss": 0.4078,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.26005503920240935,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.397,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.3158345421634542,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.3971,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.27304677511630054,
"learning_rate": 1.326303494478451e-05,
"loss": 0.4108,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.2762953278081361,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.3864,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.26489999329766345,
"learning_rate": 1.315664479122898e-05,
"loss": 0.403,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.271486485641988,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.3951,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.2613884689900172,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.3865,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.27148263366427544,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.4094,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.2718059428804698,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.3918,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.25834628964263584,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.3984,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.26180017781083703,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.4201,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.2680869687808338,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.4055,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.29330047521126595,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.3975,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.2598128079951678,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.4008,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.2658567397612568,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.4036,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.2717226502498962,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.41,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.2766890735784523,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.3952,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.24866115821921655,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.392,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.3117713548793464,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.4001,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.26946690654816025,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.4,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.28147078224533034,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.4032,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.2850808456520689,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.3947,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.24977272814046977,
"learning_rate": 1.220938883250288e-05,
"loss": 0.428,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.2816402698286664,
"learning_rate": 1.215733275116827e-05,
"loss": 0.4126,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.2540034960742461,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.407,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.2736751270402989,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.4012,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.2609268405453056,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.4092,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.27281805370998125,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.4098,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.2560773455738514,
"learning_rate": 1.189800129803139e-05,
"loss": 0.3917,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.26098414437893774,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.4149,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.27043173767962697,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.4058,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.25228506144888974,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.405,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.2933198580304236,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.3938,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.2524752323466704,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.397,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.2652388139599213,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.3925,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.263866852493644,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.3993,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.2520264779898306,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.4143,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.25526696147522265,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.4138,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.2529774746219158,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.3975,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.24598837442408517,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.3975,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.25221431897832297,
"learning_rate": 1.128231163049258e-05,
"loss": 0.4032,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.24415635896546284,
"learning_rate": 1.123144751066259e-05,
"loss": 0.4106,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.25597387345908446,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.3882,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.24494831179606436,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.4126,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.25914862014888285,
"learning_rate": 1.107927745387114e-05,
"loss": 0.4016,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.24774764604810134,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.4181,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.2651430631445135,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.3985,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.29756600356842455,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.4092,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.23375242926921108,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.4044,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.26027807834679534,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.4037,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.25941446541912305,
"learning_rate": 1.077687414070869e-05,
"loss": 0.4072,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.2384482893091086,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.4088,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.2616096708300904,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.407,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.26214862343932593,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.4121,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.2735061517682652,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.4014,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.2407057506119702,
"learning_rate": 1.052689572753247e-05,
"loss": 0.3873,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.2502113882764415,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.4073,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.2489088837451034,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.4007,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.22554703838451348,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.4028,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.2531023248900002,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.4146,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.23468995999054965,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.4041,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.27893576544074533,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.4029,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.25035906760489407,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.4028,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.25871535898771886,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.4075,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.2617913336435063,
"learning_rate": 1.008173988718005e-05,
"loss": 0.3869,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.24223836606533689,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.4047,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.25267892159476896,
"learning_rate": 9.983678617766677e-06,
"loss": 0.4033,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.2511236910801193,
"learning_rate": 9.934767922553788e-06,
"loss": 0.4005,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.24233328963616885,
"learning_rate": 9.885937708668371e-06,
"loss": 0.39,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.26144390804316,
"learning_rate": 9.837188366555524e-06,
"loss": 0.4295,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.2566935426591645,
"learning_rate": 9.788520286013691e-06,
"loss": 0.4112,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.24041815344342848,
"learning_rate": 9.739933856191537e-06,
"loss": 0.4044,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.2730196847337532,
"learning_rate": 9.691429465584866e-06,
"loss": 0.4109,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.25928398876314745,
"learning_rate": 9.643007502033492e-06,
"loss": 0.4148,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.2281992260840117,
"learning_rate": 9.594668352718152e-06,
"loss": 0.4066,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.23845969995278662,
"learning_rate": 9.546412404157383e-06,
"loss": 0.3887,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.24340618173574782,
"learning_rate": 9.498240042204471e-06,
"loss": 0.4111,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.23404258867807398,
"learning_rate": 9.450151652044339e-06,
"loss": 0.4085,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.25343625982097173,
"learning_rate": 9.402147618190472e-06,
"loss": 0.4194,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.2283834949448838,
"learning_rate": 9.354228324481841e-06,
"loss": 0.409,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.28088359687476594,
"learning_rate": 9.306394154079852e-06,
"loss": 0.3861,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.2531343089236657,
"learning_rate": 9.258645489465254e-06,
"loss": 0.3828,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.24926100300827123,
"learning_rate": 9.21098271243509e-06,
"loss": 0.397,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.25444943049539204,
"learning_rate": 9.163406204099659e-06,
"loss": 0.3985,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.23965293094764542,
"learning_rate": 9.115916344879452e-06,
"loss": 0.405,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.2449002900370905,
"learning_rate": 9.068513514502126e-06,
"loss": 0.3901,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.250756028859042,
"learning_rate": 9.021198091999439e-06,
"loss": 0.4042,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.2366598373169564,
"learning_rate": 8.973970455704251e-06,
"loss": 0.3884,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.250803613337877,
"learning_rate": 8.926830983247495e-06,
"loss": 0.3967,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.2518012652800586,
"learning_rate": 8.879780051555116e-06,
"loss": 0.409,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.2500981418522944,
"learning_rate": 8.83281803684512e-06,
"loss": 0.4113,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.23340035809059193,
"learning_rate": 8.785945314624532e-06,
"loss": 0.4024,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.2263541183652927,
"learning_rate": 8.739162259686385e-06,
"loss": 0.424,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.2655318847563107,
"learning_rate": 8.692469246106734e-06,
"loss": 0.3735,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.2285379460594085,
"learning_rate": 8.645866647241677e-06,
"loss": 0.3906,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.241138289661892,
"learning_rate": 8.599354835724348e-06,
"loss": 0.4057,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.25337131019054965,
"learning_rate": 8.552934183461958e-06,
"loss": 0.3956,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.2476971319115545,
"learning_rate": 8.506605061632805e-06,
"loss": 0.4097,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.25470352079503794,
"learning_rate": 8.460367840683308e-06,
"loss": 0.4043,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.24730221661664573,
"learning_rate": 8.414222890325065e-06,
"loss": 0.4041,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.23031880053204198,
"learning_rate": 8.368170579531855e-06,
"loss": 0.4191,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.24152204185464263,
"learning_rate": 8.322211276536733e-06,
"loss": 0.4153,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.25578564152349476,
"learning_rate": 8.27634534882907e-06,
"loss": 0.4021,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.25058938341199616,
"learning_rate": 8.230573163151585e-06,
"loss": 0.3871,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.23099906950802732,
"learning_rate": 8.184895085497466e-06,
"loss": 0.4212,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.25791036714376936,
"learning_rate": 8.139311481107401e-06,
"loss": 0.4079,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.2509658579578589,
"learning_rate": 8.093822714466686e-06,
"loss": 0.396,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.23824794337987287,
"learning_rate": 8.048429149302271e-06,
"loss": 0.3942,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.2466813246247566,
"learning_rate": 8.003131148579902e-06,
"loss": 0.4089,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.22619433465429237,
"learning_rate": 7.957929074501182e-06,
"loss": 0.4142,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.25628175132985587,
"learning_rate": 7.912823288500695e-06,
"loss": 0.3976,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.2516119847478111,
"learning_rate": 7.867814151243102e-06,
"loss": 0.4128,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.2236316556851319,
"learning_rate": 7.822902022620267e-06,
"loss": 0.407,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.2708438476761306,
"learning_rate": 7.778087261748377e-06,
"loss": 0.3975,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.24615284622257888,
"learning_rate": 7.733370226965056e-06,
"loss": 0.4064,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.22490610973359249,
"learning_rate": 7.688751275826527e-06,
"loss": 0.4079,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.25949865421835233,
"learning_rate": 7.644230765104735e-06,
"loss": 0.4067,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.24029267412879052,
"learning_rate": 7.599809050784502e-06,
"loss": 0.3992,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.2545082902845871,
"learning_rate": 7.555486488060661e-06,
"loss": 0.3902,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.24689983596172074,
"learning_rate": 7.511263431335249e-06,
"loss": 0.3867,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.24632756210785856,
"learning_rate": 7.467140234214656e-06,
"loss": 0.4074,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.25674438468267813,
"learning_rate": 7.423117249506777e-06,
"loss": 0.3911,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.23916615853162745,
"learning_rate": 7.379194829218231e-06,
"loss": 0.4088,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.2566896385159918,
"learning_rate": 7.335373324551518e-06,
"loss": 0.4041,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.22159900431780855,
"learning_rate": 7.291653085902224e-06,
"loss": 0.4236,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.2383147510298567,
"learning_rate": 7.248034462856213e-06,
"loss": 0.3938,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.2311904732603452,
"learning_rate": 7.20451780418683e-06,
"loss": 0.3997,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.24446762306024492,
"learning_rate": 7.161103457852128e-06,
"loss": 0.4073,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.23424708711075806,
"learning_rate": 7.117791770992046e-06,
"loss": 0.4068,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.2297650251179049,
"learning_rate": 7.074583089925688e-06,
"loss": 0.3776,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.24389251970475284,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.4133,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.23175721165780086,
"learning_rate": 6.988476126329604e-06,
"loss": 0.3942,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.21588276745861326,
"learning_rate": 6.945578532308857e-06,
"loss": 0.3996,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.21927267316542415,
"learning_rate": 6.902785321094301e-06,
"loss": 0.4218,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.25072172237683604,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.4045,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.21345424965494042,
"learning_rate": 6.817513414939892e-06,
"loss": 0.4078,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.23118630398101445,
"learning_rate": 6.775035401831924e-06,
"loss": 0.407,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.2260571941216985,
"learning_rate": 6.732663135188473e-06,
"loss": 0.4015,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.22339567806720065,
"learning_rate": 6.690396953817055e-06,
"loss": 0.3931,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.2461841762706571,
"learning_rate": 6.648237195676927e-06,
"loss": 0.4029,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.22119362650664628,
"learning_rate": 6.606184197876395e-06,
"loss": 0.3925,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.22965683213084986,
"learning_rate": 6.564238296670109e-06,
"loss": 0.4015,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.24970623158282815,
"learning_rate": 6.52239982745638e-06,
"loss": 0.404,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.24289581529713722,
"learning_rate": 6.480669124774484e-06,
"loss": 0.3931,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.22747381563600352,
"learning_rate": 6.439046522302013e-06,
"loss": 0.405,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.2291563432363056,
"learning_rate": 6.397532352852191e-06,
"loss": 0.398,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.2467647425172062,
"learning_rate": 6.356126948371202e-06,
"loss": 0.4045,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.22529783977426926,
"learning_rate": 6.314830639935559e-06,
"loss": 0.3935,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.23835655070040643,
"learning_rate": 6.273643757749448e-06,
"loss": 0.3974,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.250727699608511,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.4028,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.23315906464722339,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.3916,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.22727433081291648,
"learning_rate": 6.150742957589784e-06,
"loss": 0.3927,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.2524540652511112,
"learning_rate": 6.109997064904774e-06,
"loss": 0.368,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.23890812408149628,
"learning_rate": 6.069362236313092e-06,
"loss": 0.385,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.25537329124633107,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.388,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.2465454828487374,
"learning_rate": 5.988427070179108e-06,
"loss": 0.406,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.23838580487191244,
"learning_rate": 5.948127379792234e-06,
"loss": 0.4055,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.2310808799490194,
"learning_rate": 5.90794004780439e-06,
"loss": 0.397,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.24200653884699427,
"learning_rate": 5.867865395552405e-06,
"loss": 0.4082,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.25088567437955017,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.3828,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.23879179176913923,
"learning_rate": 5.788055411095892e-06,
"loss": 0.3971,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.25733956379740763,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.3777,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.23174155565306062,
"learning_rate": 5.708699979051613e-06,
"loss": 0.3965,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.2475288065474328,
"learning_rate": 5.66919351390751e-06,
"loss": 0.4101,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.234846841440531,
"learning_rate": 5.629801637510188e-06,
"loss": 0.3966,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.24814955954096107,
"learning_rate": 5.590524664836043e-06,
"loss": 0.402,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.23350842230142366,
"learning_rate": 5.551362909942708e-06,
"loss": 0.3967,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.23543775756138957,
"learning_rate": 5.51231668596653e-06,
"loss": 0.4005,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.22914293906098274,
"learning_rate": 5.47338630512007e-06,
"loss": 0.4084,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.2433146790190593,
"learning_rate": 5.434572078689617e-06,
"loss": 0.386,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.24434136838406292,
"learning_rate": 5.395874317032694e-06,
"loss": 0.3959,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.22718461464523712,
"learning_rate": 5.357293329575573e-06,
"loss": 0.3946,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.22181293708482694,
"learning_rate": 5.318829424810797e-06,
"loss": 0.3864,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.22996334959163175,
"learning_rate": 5.280482910294728e-06,
"loss": 0.3839,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.2308855756397987,
"learning_rate": 5.242254092645078e-06,
"loss": 0.3886,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.2220845474572386,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.4053,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.23795134339730004,
"learning_rate": 5.166150769707892e-06,
"loss": 0.399,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.21306631066226905,
"learning_rate": 5.128276872940503e-06,
"loss": 0.4088,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.21788438421046044,
"learning_rate": 5.090521890074953e-06,
"loss": 0.4045,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.2324673183803766,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.3937,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.26243761048636094,
"learning_rate": 5.015369872647442e-06,
"loss": 0.3959,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.2288703741242363,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.3998,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.22353723167821896,
"learning_rate": 4.940697121074675e-06,
"loss": 0.3842,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.2425995634257285,
"learning_rate": 4.903541216934864e-06,
"loss": 0.4131,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.26623965748838724,
"learning_rate": 4.8665060236772e-06,
"loss": 0.3985,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.23708548708685082,
"learning_rate": 4.829591837434102e-06,
"loss": 0.3879,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.22418779861816707,
"learning_rate": 4.792798953370424e-06,
"loss": 0.3816,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.22656105811322838,
"learning_rate": 4.756127665681083e-06,
"loss": 0.3916,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.2431260539704468,
"learning_rate": 4.719578267588725e-06,
"loss": 0.3974,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.2598212661419699,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.4069,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.4032886301706836,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.348,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.2988227858899047,
"learning_rate": 4.610664328486485e-06,
"loss": 0.3466,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.31139101436090794,
"learning_rate": 4.5746054014809e-06,
"loss": 0.3672,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.4307976499940234,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.3585,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.3547719618309115,
"learning_rate": 4.502857857942204e-06,
"loss": 0.3698,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.3344239442188738,
"learning_rate": 4.467169815100549e-06,
"loss": 0.3545,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.3747544264550817,
"learning_rate": 4.431605972355137e-06,
"loss": 0.361,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.31919703325646165,
"learning_rate": 4.396166614073516e-06,
"loss": 0.3561,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.2723510905558984,
"learning_rate": 4.360852023627846e-06,
"loss": 0.3646,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.29971413845636524,
"learning_rate": 4.325662483392656e-06,
"loss": 0.3742,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.3558826600527984,
"learning_rate": 4.29059827474257e-06,
"loss": 0.3487,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.28187385285739125,
"learning_rate": 4.255659678050063e-06,
"loss": 0.3724,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.27673832922433295,
"learning_rate": 4.22084697268323e-06,
"loss": 0.3585,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.269313019512948,
"learning_rate": 4.186160437003535e-06,
"loss": 0.3426,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.29472649351925045,
"learning_rate": 4.151600348363582e-06,
"loss": 0.3625,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.2665027144145087,
"learning_rate": 4.11716698310493e-06,
"loss": 0.3644,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.26291922503847887,
"learning_rate": 4.082860616555848e-06,
"loss": 0.3555,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.2484052710927546,
"learning_rate": 4.048681523029125e-06,
"loss": 0.3453,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.26653937324340865,
"learning_rate": 4.014629975819882e-06,
"loss": 0.3446,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.27324916003511157,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.3566,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.2313259326427848,
"learning_rate": 3.94691060843287e-06,
"loss": 0.3732,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.24760812940143095,
"learning_rate": 3.913243329737353e-06,
"loss": 0.3443,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.24296986650546906,
"learning_rate": 3.879704680319503e-06,
"loss": 0.3747,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.22169076961278453,
"learning_rate": 3.846294928353471e-06,
"loss": 0.3633,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.23322494651419157,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.3544,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.2394656341976172,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.3557,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.23471717738263317,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.3616,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.23226245591010475,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.3612,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.22834738533833956,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.3413,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.2392170362147929,
"learning_rate": 3.648558150665886e-06,
"loss": 0.3594,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.22237215837659682,
"learning_rate": 3.616058102895075e-06,
"loss": 0.356,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.21508709162356826,
"learning_rate": 3.583689060687079e-06,
"loss": 0.3548,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.22987195269731878,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.3502,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.23004933433649946,
"learning_rate": 3.51934502719796e-06,
"loss": 0.3581,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.2237680147514658,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.354,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.22089291901264188,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.3614,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.21419656382720162,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.3541,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.21545483867240872,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.3601,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.21137327535570577,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.3478,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.2169739133051747,
"learning_rate": 3.32948376099375e-06,
"loss": 0.3512,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.21676280062324232,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.3572,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.2303618113911263,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.3494,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.21680710903405104,
"learning_rate": 3.236349233544438e-06,
"loss": 0.3685,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.21361136817144996,
"learning_rate": 3.205572144561604e-06,
"loss": 0.3654,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.21778992248588605,
"learning_rate": 3.174929343376374e-06,
"loss": 0.347,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.21933128177328015,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.3686,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.22288870803025873,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.3512,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.20701498068249752,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.3632,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.22229633556532694,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.3431,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.2165299700666236,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.3641,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.2235328401707794,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.3415,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.21495102560327922,
"learning_rate": 2.964210239224696e-06,
"loss": 0.3472,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.22607414852502117,
"learning_rate": 2.934650470499869e-06,
"loss": 0.3595,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.23992302554898062,
"learning_rate": 2.905227155855128e-06,
"loss": 0.3509,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.21792496953070303,
"learning_rate": 2.875940530558519e-06,
"loss": 0.3296,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.21266685275320169,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.3639,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.21462253359666178,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.3598,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.2177691340084782,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.3647,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.2140466539066708,
"learning_rate": 2.760165589921184e-06,
"loss": 0.3661,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.2182512452997086,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.3556,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.21659945646857587,
"learning_rate": 2.703104292151164e-06,
"loss": 0.3683,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.20789343306278776,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.3589,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.21315833224429415,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.365,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.22243143339634028,
"learning_rate": 2.618550200289303e-06,
"loss": 0.3723,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.22197681288412313,
"learning_rate": 2.59064316684412e-06,
"loss": 0.3529,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.22884736532643302,
"learning_rate": 2.562875338152264e-06,
"loss": 0.3541,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.20426024307174293,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.3692,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.21426225450164363,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.3528,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.2120340205120408,
"learning_rate": 2.480409295328803e-06,
"loss": 0.3702,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.20884759774415157,
"learning_rate": 2.45320049480108e-06,
"loss": 0.362,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.20608242003329813,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.3798,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.21339340567341622,
"learning_rate": 2.399204021408421e-06,
"loss": 0.3729,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.21509420485154726,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.3637,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.22350697267097247,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.3553,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.2188420608588161,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.3306,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.21628534052837015,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.3401,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.2169672409384836,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.3446,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.22531857621476684,
"learning_rate": 2.24059906434998e-06,
"loss": 0.3243,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.2156930066481178,
"learning_rate": 2.214660694699091e-06,
"loss": 0.3514,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.22104944387804062,
"learning_rate": 2.188864536147355e-06,
"loss": 0.3566,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.20684730068035417,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.3734,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.21564934572855443,
"learning_rate": 2.137699676264171e-06,
"loss": 0.3364,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.20257372027273815,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.368,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.22032668121821714,
"learning_rate": 2.08710612114857e-06,
"loss": 0.3493,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.2188127149535317,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.3507,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.20990843737591522,
"learning_rate": 2.0370854889762e-06,
"loss": 0.3577,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.21040428556278445,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.3559,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.20651040293554934,
"learning_rate": 1.987639379598427e-06,
"loss": 0.3622,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.20672229984142296,
"learning_rate": 1.963132265887919e-06,
"loss": 0.3504,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.19674898135913454,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.3625,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.21551538723091326,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.3563,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.22473486238545487,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.3365,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.21432964216414116,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.335,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.20490479899758793,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.3502,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.20866004936143034,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.3501,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.20795659206504286,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.3609,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.21136587306498072,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.3565,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.21306613993748968,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.3517,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.20103339224668504,
"learning_rate": 1.726025117089718e-06,
"loss": 0.364,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.20758510940178287,
"learning_rate": 1.70311497206598e-06,
"loss": 0.343,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.2034923345705462,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.3659,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.2080519572135497,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.3648,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.21594816680155227,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.3484,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.20681105667067767,
"learning_rate": 1.612939220355272e-06,
"loss": 0.3546,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.21620603788580184,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.3507,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.21012122281314138,
"learning_rate": 1.568732761074121e-06,
"loss": 0.3484,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.21617274806312442,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.3521,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.2037223851200216,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.368,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.2119317710056052,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.3518,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.20843202954393678,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.3537,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.20991277441912395,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.3726,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.21572474956451593,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.3382,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.21387860942169024,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.3451,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.22390837156771023,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.3486,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.2228150794461855,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.3258,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.20419325535599003,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.3562,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.1990894499567938,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.3634,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.20875376250102726,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.3382,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.2141717865075067,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.3255,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.21668829417260507,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.3385,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.22206426442328667,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.3588,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.22591803387230242,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.3407,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.2100514381685964,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.3505,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.21071628332148876,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.368,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.20664780627085674,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.3474,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.21250615637544149,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.3471,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.216898657068745,
"learning_rate": 1.140415161609505e-06,
"loss": 0.3583,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.21321687866446729,
"learning_rate": 1.121666393356604e-06,
"loss": 0.3522,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.21181722950236692,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.3471,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.21233376275300841,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.3622,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.20391862568578842,
"learning_rate": 1.066326386135097e-06,
"loss": 0.3693,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.21777086212703048,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.339,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.2038688404002597,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.3321,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.22065122574784687,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.3564,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.2056502930066932,
"learning_rate": 9.9465987684501e-07,
"loss": 0.3671,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.21606368147204277,
"learning_rate": 9.771228098856534e-07,
"loss": 0.3464,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.20081236530192464,
"learning_rate": 9.597378493454013e-07,
"loss": 0.3522,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.22189243238851233,
"learning_rate": 9.425051342339353e-07,
"loss": 0.3448,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.20152808130480682,
"learning_rate": 9.254248023435864e-07,
"loss": 0.3553,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.20339600177803557,
"learning_rate": 9.084969902482288e-07,
"loss": 0.3679,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.21372770464103055,
"learning_rate": 8.917218333021993e-07,
"loss": 0.3546,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.2093671856404566,
"learning_rate": 8.750994656392109e-07,
"loss": 0.3642,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.2162830536763631,
"learning_rate": 8.58630020171265e-07,
"loss": 0.3499,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.2069134318767335,
"learning_rate": 8.423136285876099e-07,
"loss": 0.3468,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.20374311598806902,
"learning_rate": 8.26150421353682e-07,
"loss": 0.3349,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.21044240787744717,
"learning_rate": 8.101405277100549e-07,
"loss": 0.3462,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.20680878937613706,
"learning_rate": 7.942840756714077e-07,
"loss": 0.3716,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.21313185029567183,
"learning_rate": 7.785811920255093e-07,
"loss": 0.3512,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.2137915816254422,
"learning_rate": 7.630320023322024e-07,
"loss": 0.3466,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.20907271062801355,
"learning_rate": 7.476366309223903e-07,
"loss": 0.3568,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.20771882851563095,
"learning_rate": 7.323952008970537e-07,
"loss": 0.3506,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.21221930747856915,
"learning_rate": 7.173078341262618e-07,
"loss": 0.3335,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.20082502214667317,
"learning_rate": 7.023746512481989e-07,
"loss": 0.3673,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.2054210691154485,
"learning_rate": 6.875957716681902e-07,
"loss": 0.3701,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.19679634776318816,
"learning_rate": 6.729713135577687e-07,
"loss": 0.3548,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.21657765756245873,
"learning_rate": 6.585013938537143e-07,
"loss": 0.3359,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.21006555969758564,
"learning_rate": 6.441861282571115e-07,
"loss": 0.3521,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.2152971827995608,
"learning_rate": 6.300256312324471e-07,
"loss": 0.3519,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.20618361410936814,
"learning_rate": 6.160200160066776e-07,
"loss": 0.3692,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.20643912621743712,
"learning_rate": 6.021693945683327e-07,
"loss": 0.3422,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.20023885979081965,
"learning_rate": 5.884738776666044e-07,
"loss": 0.3548,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.20609282098328743,
"learning_rate": 5.749335748104878e-07,
"loss": 0.3542,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.20785386431778285,
"learning_rate": 5.615485942678778e-07,
"loss": 0.3343,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.20351637547299448,
"learning_rate": 5.483190430647289e-07,
"loss": 0.3555,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.20798640075110775,
"learning_rate": 5.352450269841747e-07,
"loss": 0.3606,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.20844979735383737,
"learning_rate": 5.22326650565701e-07,
"loss": 0.3473,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.21341000050245984,
"learning_rate": 5.095640171043048e-07,
"loss": 0.3622,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.19558812882182072,
"learning_rate": 4.969572286496594e-07,
"loss": 0.3557,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.19192014235116914,
"learning_rate": 4.845063860053079e-07,
"loss": 0.3977,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.20119831437172989,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.3438,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.21025679761977328,
"learning_rate": 4.600729351261857e-07,
"loss": 0.3457,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.20303983607742943,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.3541,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.20595638159044127,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.3465,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.21200735345637245,
"learning_rate": 4.245948007320588e-07,
"loss": 0.3518,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.200791108438132,
"learning_rate": 4.130816799401971e-07,
"loss": 0.3637,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.2121099669601256,
"learning_rate": 4.017251756252827e-07,
"loss": 0.3258,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.20653279262310947,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.3561,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.20292219213881255,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.3724,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.20295347659559781,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.3584,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.20522879326665122,
"learning_rate": 3.578671204615347e-07,
"loss": 0.3606,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.2082752391188304,
"learning_rate": 3.472950355547267e-07,
"loss": 0.3487,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.200603803147777,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.3772,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.19874093036274434,
"learning_rate": 3.266223765343224e-07,
"loss": 0.366,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.20491940223914334,
"learning_rate": 3.16521967718757e-07,
"loss": 0.3535,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.20532476262899452,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.3484,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.20252785965433953,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.3675,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.20408688929195515,
"learning_rate": 2.871653894244597e-07,
"loss": 0.3517,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.19367049601414843,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.3731,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.19981255396003017,
"learning_rate": 2.683823233205485e-07,
"loss": 0.3522,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.2102768908466816,
"learning_rate": 2.592274132703287e-07,
"loss": 0.354,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.19883883206649414,
"learning_rate": 2.502303499116754e-07,
"loss": 0.3779,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.20132448436206837,
"learning_rate": 2.413912051848666e-07,
"loss": 0.3544,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.19870917940871308,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.3752,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.19435147527066396,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.3766,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.2070136423038842,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.3533,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.2002245365408372,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.3512,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.20398638884066714,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.3464,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.21875669445803833,
"learning_rate": 1.916764975761254e-07,
"loss": 0.363,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.2068633041535305,
"learning_rate": 1.839446914621923e-07,
"loss": 0.3643,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.20430416261738138,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.3592,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.20257477948466093,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.3729,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.21239697971845703,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.3371,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.20148727568784283,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.3647,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.20692147930324709,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.327,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.20934961219776363,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.3574,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.19447226305379503,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.3674,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.20195402140414273,
"learning_rate": 1.277993209368744e-07,
"loss": 0.3542,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.19182430011111617,
"learning_rate": 1.214956261808542e-07,
"loss": 0.3762,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.204698316615029,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.3761,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.2021628996443327,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.3508,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.20782891615662868,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.3598,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.20507510920682148,
"learning_rate": 9.787081204466253e-08,
"loss": 0.3582,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.21799601869527868,
"learning_rate": 9.236233586200627e-08,
"loss": 0.3141,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.19827277542486393,
"learning_rate": 8.70130406196279e-08,
"loss": 0.3481,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.1975468813629135,
"learning_rate": 8.182296909035047e-08,
"loss": 0.3565,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.20795159937961083,
"learning_rate": 7.679216277384439e-08,
"loss": 0.3326,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.2068644501146018,
"learning_rate": 7.19206618963053e-08,
"loss": 0.3505,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.2001922952918142,
"learning_rate": 6.720850541012347e-08,
"loss": 0.3791,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.2026862933346359,
"learning_rate": 6.265573099357714e-08,
"loss": 0.3575,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.19448927274916245,
"learning_rate": 5.826237505053067e-08,
"loss": 0.3587,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.1999504241323524,
"learning_rate": 5.4028472710137e-08,
"loss": 0.3479,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.1994117371430948,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.3496,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.2060563956493721,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.3493,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.2115590677031768,
"learning_rate": 4.228381947004812e-08,
"loss": 0.3382,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.18903828548157367,
"learning_rate": 3.868805732811343e-08,
"loss": 0.3702,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.20609304655674743,
"learning_rate": 3.525190530455636e-08,
"loss": 0.3431,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.19873320963859592,
"learning_rate": 3.197539087475887e-08,
"loss": 0.3433,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.2063174174363604,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.374,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.19799686894789337,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.3613,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.19304214383324225,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.3594,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.20083567337985145,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.3488,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.20489241522213364,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.3346,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.2036346939907363,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.3524,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.20055229096064034,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.3668,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.20978145419335423,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.3455,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.20158642733310161,
"learning_rate": 9.67435416023843e-09,
"loss": 0.3629,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.1983578529796547,
"learning_rate": 7.995445897566178e-09,
"loss": 0.354,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.20313586192500663,
"learning_rate": 6.476393170578021e-09,
"loss": 0.3608,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.20414886972316862,
"learning_rate": 5.11720812557348e-09,
"loss": 0.3418,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.1918089411113897,
"learning_rate": 3.917901630563492e-09,
"loss": 0.3711,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.20176530063870468,
"learning_rate": 2.878483275170485e-09,
"loss": 0.3395,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.20150460494310404,
"learning_rate": 1.998961370555108e-09,
"loss": 0.3505,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.19337686089703265,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.3667,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.20908780405234206,
"learning_rate": 7.196337656112384e-10,
"loss": 0.3697,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.2102398570712456,
"learning_rate": 3.198382947511292e-10,
"loss": 0.3375,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.20799017102005446,
"learning_rate": 7.995973352770137e-11,
"loss": 0.349,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.19751276449881888,
"learning_rate": 0.0,
"loss": 0.3471,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 4.760225726924325e+18,
"train_loss": 0.4833199501520226,
"train_runtime": 29358.4045,
"train_samples_per_second": 5.382,
"train_steps_per_second": 0.042
}
],
"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": 4.760225726924325e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}