gsmyrnis's picture
End of training
0caa436 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": 7.365854730794966,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.0588,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 7.435104204110287,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0849,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 7.174382615981859,
"learning_rate": 9.67741935483871e-07,
"loss": 1.0561,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 7.1946965754183845,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.0645,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 6.8123394997516415,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.0643,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 6.697784302841602,
"learning_rate": 1.935483870967742e-06,
"loss": 1.0631,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 5.066356415957141,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.9828,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.881299157461293,
"learning_rate": 2.580645161290323e-06,
"loss": 1.0096,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 3.194893103698737,
"learning_rate": 2.903225806451613e-06,
"loss": 0.9556,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.8497093187777427,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9456,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.76701267381977,
"learning_rate": 3.548387096774194e-06,
"loss": 0.9505,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 4.089878340236963,
"learning_rate": 3.870967741935484e-06,
"loss": 0.9185,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.376464870174611,
"learning_rate": 4.193548387096774e-06,
"loss": 0.9105,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 4.283666001763946,
"learning_rate": 4.516129032258065e-06,
"loss": 0.9069,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.634802806774234,
"learning_rate": 4.838709677419355e-06,
"loss": 0.8825,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.5383507116746444,
"learning_rate": 5.161290322580646e-06,
"loss": 0.8504,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 3.003684746052102,
"learning_rate": 5.483870967741935e-06,
"loss": 0.8285,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.404184094866108,
"learning_rate": 5.806451612903226e-06,
"loss": 0.8334,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.7769575160146607,
"learning_rate": 6.129032258064517e-06,
"loss": 0.81,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.3943167808332046,
"learning_rate": 6.451612903225806e-06,
"loss": 0.7858,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.2480960516714292,
"learning_rate": 6.774193548387097e-06,
"loss": 0.7654,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.295701117922533,
"learning_rate": 7.096774193548388e-06,
"loss": 0.7735,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.1225528799979796,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.7449,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 0.818726512477109,
"learning_rate": 7.741935483870968e-06,
"loss": 0.7473,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.0136926170931375,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7305,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.0102445816585612,
"learning_rate": 8.387096774193549e-06,
"loss": 0.7241,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.9187984755176175,
"learning_rate": 8.70967741935484e-06,
"loss": 0.7341,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.8431220951338042,
"learning_rate": 9.03225806451613e-06,
"loss": 0.7102,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.8797480640288946,
"learning_rate": 9.35483870967742e-06,
"loss": 0.7009,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 1.043043567947307,
"learning_rate": 9.67741935483871e-06,
"loss": 0.7132,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.8226824837708274,
"learning_rate": 1e-05,
"loss": 0.6807,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.8730274002205809,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.6915,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.7627673387961271,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.6687,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.8762681820564673,
"learning_rate": 1.096774193548387e-05,
"loss": 0.6979,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.6674011348283742,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.6851,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.833459984663004,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.6548,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.7415216244099724,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.66,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6755570757050359,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.6621,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.7274679006404473,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.6579,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.6869580828984803,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.653,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.6605153966557434,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6345,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.5627370243740394,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.6572,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.7299693329811974,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.653,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.6512397726449258,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.6555,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.7063058177695943,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.6537,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.7162699484499521,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6391,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.6983548241090217,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.6449,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.6244122766346908,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6532,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.840453715155802,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.624,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.7481905818749721,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6282,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.6641260739141847,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.6181,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.6563212657535579,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6239,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.7468076842323307,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.6308,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.6211161420351672,
"learning_rate": 1.741935483870968e-05,
"loss": 0.6246,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.6695081108348522,
"learning_rate": 1.774193548387097e-05,
"loss": 0.6252,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.6751568307197949,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6212,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.5832604217325091,
"learning_rate": 1.838709677419355e-05,
"loss": 0.6142,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.7103501420103268,
"learning_rate": 1.870967741935484e-05,
"loss": 0.6164,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.7514961775563496,
"learning_rate": 1.903225806451613e-05,
"loss": 0.6065,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.768739531651892,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6182,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.8452041712176581,
"learning_rate": 1.967741935483871e-05,
"loss": 0.6444,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.8461771758118967,
"learning_rate": 2e-05,
"loss": 0.6035,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.8795205226588066,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.6231,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.895679422509461,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.6064,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.8123927476449917,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.6259,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 1.1679413307715416,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.5937,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.8115996035293659,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6183,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.7336950524357856,
"learning_rate": 2.193548387096774e-05,
"loss": 0.6074,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.9892876201696593,
"learning_rate": 2.225806451612903e-05,
"loss": 0.6002,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 1.0212297935552404,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.611,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.773560281010534,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6186,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.0480247964759537,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.5954,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.9288566777069883,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.5948,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.334995266640031,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.5994,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.8088295658757351,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.6176,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.9928028503584028,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.5952,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.4568635703627786,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.6194,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.844217899236601,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.6077,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.456631449471167,
"learning_rate": 2.548387096774194e-05,
"loss": 0.6052,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.8991542866742706,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.6168,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.0916940258582257,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.5974,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.9573095889076995,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.593,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.3434435566687113,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6123,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.9009011265596587,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.5985,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.3385942963399977,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.5851,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.5486354027893792,
"learning_rate": 2.774193548387097e-05,
"loss": 0.5943,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.2204380622468818,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.5954,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.5195774078190034,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.5987,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.229986376674695,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.5795,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.2859231311758594,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.5882,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.135676066636644,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.5837,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.0855628571890246,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.5767,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.2026900087818058,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.596,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.3777276663428097,
"learning_rate": 3.032258064516129e-05,
"loss": 0.6093,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.9910439990055603,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.5987,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.9285303856757625,
"learning_rate": 3.096774193548387e-05,
"loss": 0.5965,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.5246489388583582,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.5904,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.8895968667660066,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6058,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.8195516771563582,
"learning_rate": 3.193548387096774e-05,
"loss": 0.5913,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.3448083163244307,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.5916,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.1465862399736315,
"learning_rate": 3.258064516129033e-05,
"loss": 0.5851,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.3509075333442542,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.6037,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.6374855340713583,
"learning_rate": 3.322580645161291e-05,
"loss": 0.5788,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.0761252083921398,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.592,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.9091135487481137,
"learning_rate": 3.387096774193549e-05,
"loss": 0.5915,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.2937585924402224,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.5832,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.8934998484818437,
"learning_rate": 3.451612903225807e-05,
"loss": 0.5962,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.6745657326213779,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6079,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.6289164991703275,
"learning_rate": 3.516129032258065e-05,
"loss": 0.5787,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.1818548090934269,
"learning_rate": 3.548387096774194e-05,
"loss": 0.5646,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.9143139166232501,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.5852,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.4745703119716707,
"learning_rate": 3.612903225806452e-05,
"loss": 0.5803,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.839735296134254,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.5873,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.3391954892907576,
"learning_rate": 3.67741935483871e-05,
"loss": 0.5892,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.5689771668654544,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.5788,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.1162181750412974,
"learning_rate": 3.741935483870968e-05,
"loss": 0.5984,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 1.7063744402174528,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.5929,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.3604746636851932,
"learning_rate": 3.806451612903226e-05,
"loss": 0.5751,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.6890556080400716,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.5969,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.470791773151716,
"learning_rate": 3.870967741935484e-05,
"loss": 0.5939,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.4811442479368848,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.5723,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.1686381740223633,
"learning_rate": 3.935483870967742e-05,
"loss": 0.5772,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.4061997030813163,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.5692,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.9659031695336677,
"learning_rate": 4e-05,
"loss": 0.5882,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.3742207194487774,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.5695,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.4638548556436077,
"learning_rate": 3.999968016170525e-05,
"loss": 0.5843,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.9575220951448068,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.5688,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 1.8169115487734537,
"learning_rate": 3.999872065705065e-05,
"loss": 0.5927,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.2312199739032885,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6029,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.488177819547375,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.5702,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.1155290052634357,
"learning_rate": 3.999608209836944e-05,
"loss": 0.5879,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.4057646382922655,
"learning_rate": 3.999488279187443e-05,
"loss": 0.579,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.9025119636551316,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.5839,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.02798429398462,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.5844,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.8598461963650137,
"learning_rate": 3.999032564583976e-05,
"loss": 0.5765,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.0622428386823448,
"learning_rate": 3.998848689546591e-05,
"loss": 0.5771,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.8831794816225582,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.5806,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.9389501348795023,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.5678,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.8756257820091032,
"learning_rate": 3.998201174509306e-05,
"loss": 0.5882,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.9822884828064798,
"learning_rate": 3.997953378607961e-05,
"loss": 0.5686,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.7402496371962597,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.5774,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.7422095410187902,
"learning_rate": 3.997409862168454e-05,
"loss": 0.5817,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.7634450598570579,
"learning_rate": 3.997114145976237e-05,
"loss": 0.5865,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.7113606941512401,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.583,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.8299491936328591,
"learning_rate": 3.996474809469545e-05,
"loss": 0.5795,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.9988768355691081,
"learning_rate": 3.996131194267189e-05,
"loss": 0.5729,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.7142068801507949,
"learning_rate": 3.995771618052995e-05,
"loss": 0.5748,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.9126153709504702,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.5788,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.7390315804432509,
"learning_rate": 3.995004594217343e-05,
"loss": 0.5854,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.9943222174395583,
"learning_rate": 3.994597152728987e-05,
"loss": 0.5877,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.7425408521356281,
"learning_rate": 3.994173762494947e-05,
"loss": 0.5791,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.0611539519037008,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.5712,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 1.1149702428786463,
"learning_rate": 3.993279149458988e-05,
"loss": 0.5839,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.7051971429185496,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.5802,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.9583625036017844,
"learning_rate": 3.992320783722616e-05,
"loss": 0.5737,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.8165633051626792,
"learning_rate": 3.991817703090965e-05,
"loss": 0.5743,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.5910991702209677,
"learning_rate": 3.991298695938038e-05,
"loss": 0.5703,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.6696420374417866,
"learning_rate": 3.9907637664138e-05,
"loss": 0.5564,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.741539102766627,
"learning_rate": 3.990212918795534e-05,
"loss": 0.569,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.8965482414379039,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.5676,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.6890589410907708,
"learning_rate": 3.989063487022416e-05,
"loss": 0.5776,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 1.0314692858402486,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.5734,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.9903720662170363,
"learning_rate": 3.987850437381915e-05,
"loss": 0.5688,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.7270271178392912,
"learning_rate": 3.987220067906313e-05,
"loss": 0.5786,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.861926952286642,
"learning_rate": 3.986573808672002e-05,
"loss": 0.562,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.848844897553498,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.5795,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.6578775182735873,
"learning_rate": 3.985233641724152e-05,
"loss": 0.5732,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.777688746436825,
"learning_rate": 3.984539744726554e-05,
"loss": 0.5584,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.8602445686306917,
"learning_rate": 3.983829979402038e-05,
"loss": 0.574,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 1.178452143804664,
"learning_rate": 3.983104351425871e-05,
"loss": 0.5575,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.7045837224155359,
"learning_rate": 3.982362866600154e-05,
"loss": 0.535,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 1.0237223298934797,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.5682,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.8113959087722853,
"learning_rate": 3.980832350242388e-05,
"loss": 0.5483,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.8337064150463924,
"learning_rate": 3.980043330948306e-05,
"loss": 0.5638,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.8173916013489356,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.5707,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.6542855213353214,
"learning_rate": 3.978417801674579e-05,
"loss": 0.5768,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.7622364322413046,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.5765,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.6787321440150551,
"learning_rate": 3.976728995023254e-05,
"loss": 0.5675,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.744949047190797,
"learning_rate": 3.975860879481514e-05,
"loss": 0.5545,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.6353329321695053,
"learning_rate": 3.974976965008833e-05,
"loss": 0.5726,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.7298436734078604,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.559,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.5742265896314972,
"learning_rate": 3.973161767667946e-05,
"loss": 0.5553,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.6867962199286674,
"learning_rate": 3.972230499314009e-05,
"loss": 0.5575,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.7073059301009924,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.5448,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.7368223997101905,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.5575,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.6909516538711786,
"learning_rate": 3.969342105253098e-05,
"loss": 0.5491,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.6045572519070597,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.5553,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.5857085158105796,
"learning_rate": 3.967337762346568e-05,
"loss": 0.5478,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.6750488378433451,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.5572,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.6311073630890582,
"learning_rate": 3.965270496444528e-05,
"loss": 0.5729,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.5805965650687086,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.548,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.7077733992885723,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.5578,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.48118525112292093,
"learning_rate": 3.962051762160151e-05,
"loss": 0.5571,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.6311709047432276,
"learning_rate": 3.960947462140638e-05,
"loss": 0.5435,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.5205411937174941,
"learning_rate": 3.959827482437472e-05,
"loss": 0.5628,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.5059298192447107,
"learning_rate": 3.958691832005981e-05,
"loss": 0.5529,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.531578458329531,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.5823,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.4891131123844724,
"learning_rate": 3.956373555405774e-05,
"loss": 0.5761,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.5432583736129524,
"learning_rate": 3.955190947773937e-05,
"loss": 0.5413,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.5251988376687431,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.5697,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.49375114275403065,
"learning_rate": 3.952778841127214e-05,
"loss": 0.5463,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.5784860303046007,
"learning_rate": 3.95154936139947e-05,
"loss": 0.5534,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.5188055220105146,
"learning_rate": 3.950304277135034e-05,
"loss": 0.5608,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.6516566595206597,
"learning_rate": 3.94904359828957e-05,
"loss": 0.5619,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.5774651095594212,
"learning_rate": 3.94776733494343e-05,
"loss": 0.5675,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.6676022818907797,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.5639,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.49374598578128726,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.5451,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.5848929930800226,
"learning_rate": 3.943845140573213e-05,
"loss": 0.5443,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.5446273769258022,
"learning_rate": 3.942506642518952e-05,
"loss": 0.5504,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.7299647985239084,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.5656,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.5878177209909801,
"learning_rate": 3.939783060543167e-05,
"loss": 0.5651,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.5687425031162314,
"learning_rate": 3.938397998399333e-05,
"loss": 0.5578,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.6620658423409145,
"learning_rate": 3.936997436876756e-05,
"loss": 0.5539,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.6016918401827115,
"learning_rate": 3.935581387174289e-05,
"loss": 0.5549,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.5758384756758861,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.5505,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.5721392962126582,
"learning_rate": 3.932702868644224e-05,
"loss": 0.5532,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.5513294822121878,
"learning_rate": 3.931240422833181e-05,
"loss": 0.5529,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.5571694749494083,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.5507,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.594037963262179,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.5607,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.5238472984868597,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.5488,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.6050635226391278,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.5436,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.6491766636199613,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.5366,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.6065405523017935,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.5522,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.6458703483962422,
"learning_rate": 3.92057159243286e-05,
"loss": 0.5422,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.5678484202331869,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.5651,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.5955258999988293,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.5518,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.6172470100455575,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.5387,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.4642305986765359,
"learning_rate": 3.914136997982874e-05,
"loss": 0.5463,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.6675365628883995,
"learning_rate": 3.912490053436079e-05,
"loss": 0.5545,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.5525136348830739,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.5556,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.560835494694375,
"learning_rate": 3.909150300975178e-05,
"loss": 0.5497,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.5439452736838635,
"learning_rate": 3.907457519765642e-05,
"loss": 0.5515,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.6966888945216803,
"learning_rate": 3.905749486576607e-05,
"loss": 0.572,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.5058195354466648,
"learning_rate": 3.90402621506546e-05,
"loss": 0.5393,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.5055097863575556,
"learning_rate": 3.902287719011435e-05,
"loss": 0.5352,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.5728632336996909,
"learning_rate": 3.900534012315499e-05,
"loss": 0.5524,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5314224265880464,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.5545,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.5758763588424354,
"learning_rate": 3.896981023209777e-05,
"loss": 0.553,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.5793890110068655,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.5416,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.6282796525742659,
"learning_rate": 3.893367361386491e-05,
"loss": 0.5389,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.5342119982924454,
"learning_rate": 3.891537814248417e-05,
"loss": 0.5391,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.6774184928712229,
"learning_rate": 3.889693142424384e-05,
"loss": 0.562,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.5943843398679732,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.5306,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.6554778341437831,
"learning_rate": 3.885958483839049e-05,
"loss": 0.5472,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.7170817256066074,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.5478,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.644797870463,
"learning_rate": 3.882163505079171e-05,
"loss": 0.5309,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.6324941941352765,
"learning_rate": 3.880243433489132e-05,
"loss": 0.5477,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.5607290234828152,
"learning_rate": 3.878308327522702e-05,
"loss": 0.4907,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.6461952286610885,
"learning_rate": 3.876358202652936e-05,
"loss": 0.5139,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.6039279356068958,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.4914,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.5093265488783933,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.4881,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.6828050894576737,
"learning_rate": 3.870417871154801e-05,
"loss": 0.4903,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.5020187806867761,
"learning_rate": 3.868407827802192e-05,
"loss": 0.4902,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.6153433160541816,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.5117,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.5401675499427212,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.5071,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.5149189247990918,
"learning_rate": 3.862288124195319e-05,
"loss": 0.4927,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.5288650557585627,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.5092,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.4863451229879114,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.5133,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.4925285657322285,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.5033,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.5616399031448933,
"learning_rate": 3.853920126717703e-05,
"loss": 0.4939,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.49780497190573797,
"learning_rate": 3.851791025492574e-05,
"loss": 0.4967,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.6120290468395212,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.4917,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.5496477036920224,
"learning_rate": 3.847488419569866e-05,
"loss": 0.5057,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.5320621150039541,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.4839,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.471897855924614,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.502,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.5284650054495513,
"learning_rate": 3.840923760917199e-05,
"loss": 0.4934,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.5580485766256089,
"learning_rate": 3.838706077964473e-05,
"loss": 0.4808,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.520216241808935,
"learning_rate": 3.836473692766944e-05,
"loss": 0.4838,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.5604739401331407,
"learning_rate": 3.834226623174705e-05,
"loss": 0.4905,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.5506086510783618,
"learning_rate": 3.831964887155264e-05,
"loss": 0.5034,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.44708184169175585,
"learning_rate": 3.829688502793403e-05,
"loss": 0.5033,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.5500512374852573,
"learning_rate": 3.827397488291029e-05,
"loss": 0.5068,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.4616337588796714,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.5056,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.5146027674146528,
"learning_rate": 3.822771642257145e-05,
"loss": 0.4716,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.5271016532607421,
"learning_rate": 3.820436847713776e-05,
"loss": 0.4789,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.5492312893104668,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.4694,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.5975478279334647,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.5031,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.46021931019271534,
"learning_rate": 3.813345202354145e-05,
"loss": 0.509,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.5231163492285292,
"learning_rate": 3.810952296329563e-05,
"loss": 0.4915,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.5194039788113349,
"learning_rate": 3.808544909978676e-05,
"loss": 0.4956,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.5881171464864303,
"learning_rate": 3.806123062550882e-05,
"loss": 0.5094,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.4438287406777171,
"learning_rate": 3.803686773411208e-05,
"loss": 0.4992,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.5609681535886908,
"learning_rate": 3.801236062040158e-05,
"loss": 0.5011,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.45580886399998777,
"learning_rate": 3.798770948033553e-05,
"loss": 0.477,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.5699515494870474,
"learning_rate": 3.796291451102381e-05,
"loss": 0.4834,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.4928373173167812,
"learning_rate": 3.793797591072631e-05,
"loss": 0.4816,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.5288426657960312,
"learning_rate": 3.791289387885144e-05,
"loss": 0.4915,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.5063208610325264,
"learning_rate": 3.788766861595443e-05,
"loss": 0.5024,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.545745890849546,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.4877,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.6021786197572858,
"learning_rate": 3.783678920503983e-05,
"loss": 0.4802,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.5612588565823254,
"learning_rate": 3.781113546385265e-05,
"loss": 0.5084,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.5294563296787831,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.4921,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.4347644015320258,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.4888,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.510207808888941,
"learning_rate": 3.77333205623025e-05,
"loss": 0.5002,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.46199453548009944,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.5067,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.6231082189457711,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.5063,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.6217269238845342,
"learning_rate": 3.765422951112704e-05,
"loss": 0.5018,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.48861517286563666,
"learning_rate": 3.762758321970216e-05,
"loss": 0.4834,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.5655351416471153,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.5128,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.5070330223954254,
"learning_rate": 3.757386800198538e-05,
"loss": 0.4904,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.520494699017952,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.5002,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.6881679666985382,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.4989,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.37360474045450004,
"learning_rate": 3.749224181796305e-05,
"loss": 0.5001,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.6213256269677001,
"learning_rate": 3.746475306375545e-05,
"loss": 0.5011,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.5127699631530193,
"learning_rate": 3.743712466184774e-05,
"loss": 0.4933,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.464479492987708,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.4969,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.5817324618786768,
"learning_rate": 3.73814497997107e-05,
"loss": 0.5056,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.540963534538154,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.496,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.43262432202699086,
"learning_rate": 3.732521901224724e-05,
"loss": 0.4842,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.6304171991488219,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.4989,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.49841157339820913,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.4844,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.4692064012828636,
"learning_rate": 3.723983441007882e-05,
"loss": 0.4896,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.591481753840897,
"learning_rate": 3.721109687296781e-05,
"loss": 0.496,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.44301795478239875,
"learning_rate": 3.718222171638484e-05,
"loss": 0.4955,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.6101822766984679,
"learning_rate": 3.715320917121488e-05,
"loss": 0.4972,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.5446793812673162,
"learning_rate": 3.712405946944149e-05,
"loss": 0.5145,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.49830633380111883,
"learning_rate": 3.709477284414488e-05,
"loss": 0.5013,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.5099847389531773,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.4965,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.40697325352114216,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.4748,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.446332894914272,
"learning_rate": 3.70060937743295e-05,
"loss": 0.5169,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.43764641376903896,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.5073,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.38331329604880704,
"learning_rate": 3.694629409915555e-05,
"loss": 0.4889,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.43905299714763935,
"learning_rate": 3.691619088858401e-05,
"loss": 0.4836,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.3817993622953091,
"learning_rate": 3.68859524166009e-05,
"loss": 0.4871,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.46290017123819616,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.4968,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.3891433105259083,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.4996,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.45571000605834583,
"learning_rate": 3.67944278554384e-05,
"loss": 0.503,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.366334718700721,
"learning_rate": 3.676365076645557e-05,
"loss": 0.4929,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.4786112660777718,
"learning_rate": 3.673273963576791e-05,
"loss": 0.4685,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.455624313024752,
"learning_rate": 3.670169471054001e-05,
"loss": 0.4913,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.43107459269622844,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.4793,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.43836575212460893,
"learning_rate": 3.663920447046888e-05,
"loss": 0.5055,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.46036834672489163,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.4888,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.4600242152346772,
"learning_rate": 3.657618204491937e-05,
"loss": 0.4927,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.4340672844827833,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.4847,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.4405058495017093,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.5049,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.5093753672816718,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.5083,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.45154997360921895,
"learning_rate": 3.644854871713611e-05,
"loss": 0.5028,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.47084553494813103,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.4813,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.46057868011793085,
"learning_rate": 3.638394189710493e-05,
"loss": 0.4842,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.5545743793282565,
"learning_rate": 3.635144184933412e-05,
"loss": 0.4868,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5927627944563303,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.5018,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.4345605236435263,
"learning_rate": 3.628604977762752e-05,
"loss": 0.4939,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.5801586767986322,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.4951,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.4991502902330786,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.5081,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.5834362003806903,
"learning_rate": 3.618698565901727e-05,
"loss": 0.5081,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.6092467781877209,
"learning_rate": 3.615370507164654e-05,
"loss": 0.5153,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.43356584427223077,
"learning_rate": 3.61202953196805e-05,
"loss": 0.5017,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.5101758839179795,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.4984,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.4854807476151392,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.4991,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.5344736090576728,
"learning_rate": 3.601929375279662e-05,
"loss": 0.5074,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.6042054055676381,
"learning_rate": 3.598537002418012e-05,
"loss": 0.4873,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.3799037318747127,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.4918,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.564125136274142,
"learning_rate": 3.591713938344416e-05,
"loss": 0.5133,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.45987056061307086,
"learning_rate": 3.588283301689507e-05,
"loss": 0.4867,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.4736169799296141,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.4891,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.5402528099468081,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.4854,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.42632205437005655,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.491,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.5189662904439428,
"learning_rate": 3.574434032194994e-05,
"loss": 0.4901,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.4937457778779865,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.4933,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.5262247134473184,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.5091,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.5502021404145703,
"learning_rate": 3.563914797637216e-05,
"loss": 0.5015,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.479142897381768,
"learning_rate": 3.560383338592649e-05,
"loss": 0.4873,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.5782995829484542,
"learning_rate": 3.556839402764487e-05,
"loss": 0.4821,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.4690424982640206,
"learning_rate": 3.553283018489946e-05,
"loss": 0.4829,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.5095988714552322,
"learning_rate": 3.54971421420578e-05,
"loss": 0.4818,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.48692103914304796,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.4922,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.4572253413378329,
"learning_rate": 3.54253945985191e-05,
"loss": 0.4984,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.404689171748191,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.4813,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.4531068366632024,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.497,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.45767790549208276,
"learning_rate": 3.531684894865866e-05,
"loss": 0.5112,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.4600838020525436,
"learning_rate": 3.528042173241128e-05,
"loss": 0.4941,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.452237689009667,
"learning_rate": 3.524387233431892e-05,
"loss": 0.5033,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.5214275483971059,
"learning_rate": 3.520720104662958e-05,
"loss": 0.4942,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.4042736836156371,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.4918,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.4092534779728241,
"learning_rate": 3.513349397632281e-05,
"loss": 0.4856,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.4361894738604706,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.5062,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.3998864090193819,
"learning_rate": 3.505930287892533e-05,
"loss": 0.491,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.40580539943989163,
"learning_rate": 3.502202656100099e-05,
"loss": 0.4892,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.4755123989524639,
"learning_rate": 3.498463012735256e-05,
"loss": 0.4845,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.4182510986371007,
"learning_rate": 3.494711387700094e-05,
"loss": 0.4896,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.5210469290268023,
"learning_rate": 3.490947810992505e-05,
"loss": 0.4771,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.40798737322421935,
"learning_rate": 3.48717231270595e-05,
"loss": 0.5017,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.5755817829975204,
"learning_rate": 3.483384923029211e-05,
"loss": 0.4796,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.5534051671264112,
"learning_rate": 3.479585672246156e-05,
"loss": 0.5016,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.4484933808541243,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.4937,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.6099585236665457,
"learning_rate": 3.471951708970527e-05,
"loss": 0.495,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.4573829906955328,
"learning_rate": 3.468117057518921e-05,
"loss": 0.4764,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.58972048350147,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.5017,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.5025314160435018,
"learning_rate": 3.460412568296731e-05,
"loss": 0.4818,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.4759826418628189,
"learning_rate": 3.456542792131039e-05,
"loss": 0.4815,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.5520205436462825,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.4956,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.4917457784643788,
"learning_rate": 3.448768331403348e-05,
"loss": 0.4874,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.5981550230438653,
"learning_rate": 3.44486370900573e-05,
"loss": 0.4928,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.5229733008024208,
"learning_rate": 3.440947533516396e-05,
"loss": 0.4943,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.5524059022553847,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.5162,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.5810094703894411,
"learning_rate": 3.43308064860925e-05,
"loss": 0.4976,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.5114130028455287,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.4863,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.5868140817528339,
"learning_rate": 3.425167928295015e-05,
"loss": 0.5039,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.5145035999155259,
"learning_rate": 3.421194458890411e-05,
"loss": 0.4902,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.605498836188728,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.487,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.5772834831494518,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.4935,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.43014655831623866,
"learning_rate": 3.409205995219561e-05,
"loss": 0.499,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.4511689093372272,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.4874,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.4541573526107167,
"learning_rate": 3.40115729298209e-05,
"loss": 0.4906,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.4431485294368503,
"learning_rate": 3.397116120327025e-05,
"loss": 0.4868,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.462152686070644,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.4915,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.5315134794961228,
"learning_rate": 3.389000293509524e-05,
"loss": 0.4799,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.4320718336959199,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.5004,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.4888854453752876,
"learning_rate": 3.380840041143494e-05,
"loss": 0.4963,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.44532700150516347,
"learning_rate": 3.376743336885792e-05,
"loss": 0.5034,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.47553315358389675,
"learning_rate": 3.372635624225055e-05,
"loss": 0.4889,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.41932388703863577,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.4987,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.4978214390920301,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.4723,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.46719861185124384,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.4856,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.5066709549823435,
"learning_rate": 3.356095347769799e-05,
"loss": 0.4778,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.4212480049624941,
"learning_rate": 3.351933087522552e-05,
"loss": 0.4887,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.41909115503886346,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.4784,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.4264559274004611,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.5001,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.44648882858233646,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.4859,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.4421960987561675,
"learning_rate": 3.335176280432308e-05,
"loss": 0.5041,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.4694292001860267,
"learning_rate": 3.330960304618295e-05,
"loss": 0.498,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.4185499823535191,
"learning_rate": 3.326733686481153e-05,
"loss": 0.4866,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.41989484843493263,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.4927,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.4444758506579028,
"learning_rate": 3.318248658506012e-05,
"loss": 0.5078,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.40893239189338465,
"learning_rate": 3.313990316514069e-05,
"loss": 0.5031,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.40260914559640837,
"learning_rate": 3.309721467890571e-05,
"loss": 0.4709,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.40874541395221664,
"learning_rate": 3.305442146769115e-05,
"loss": 0.4791,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.41833089330333295,
"learning_rate": 3.30115238736704e-05,
"loss": 0.4919,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.4096199136424183,
"learning_rate": 3.296852223985148e-05,
"loss": 0.4814,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.4074649220945212,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.4847,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.4242732067159921,
"learning_rate": 3.288220822900796e-05,
"loss": 0.4956,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.46826065702525926,
"learning_rate": 3.283889654214788e-05,
"loss": 0.4957,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.4083927607792412,
"learning_rate": 3.279548219581317e-05,
"loss": 0.4972,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.4665437396178256,
"learning_rate": 3.275196553714379e-05,
"loss": 0.5005,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.40295216446292637,
"learning_rate": 3.270834691409778e-05,
"loss": 0.4755,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.3954941491796749,
"learning_rate": 3.266462667544849e-05,
"loss": 0.4927,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.44714890929498874,
"learning_rate": 3.262080517078178e-05,
"loss": 0.4906,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.42513629042495776,
"learning_rate": 3.257688275049323e-05,
"loss": 0.4828,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.38579976511488906,
"learning_rate": 3.253285976578535e-05,
"loss": 0.495,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.44430239490153606,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.4901,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.47608522633961353,
"learning_rate": 3.244451351193935e-05,
"loss": 0.5088,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.3399552654306246,
"learning_rate": 3.240019094921551e-05,
"loss": 0.4932,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.4661423496735538,
"learning_rate": 3.235576923489527e-05,
"loss": 0.4881,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.40707957061944916,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.5072,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.45168027818469686,
"learning_rate": 3.226662977303494e-05,
"loss": 0.499,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.41084267015339815,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.5045,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.4225373999988794,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.4908,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.39665031604092915,
"learning_rate": 3.21321858487569e-05,
"loss": 0.487,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.42475306653643563,
"learning_rate": 3.208717671149931e-05,
"loss": 0.485,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.36372997396821577,
"learning_rate": 3.204207092549882e-05,
"loss": 0.4928,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.45137793710127083,
"learning_rate": 3.19968688514201e-05,
"loss": 0.4956,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.36790324890490245,
"learning_rate": 3.195157085069774e-05,
"loss": 0.4826,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.40600962212588304,
"learning_rate": 3.190617728553332e-05,
"loss": 0.4933,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.3617224010274562,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.487,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.40443857919727694,
"learning_rate": 3.181510491450254e-05,
"loss": 0.4859,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.3850682415119401,
"learning_rate": 3.176942683684842e-05,
"loss": 0.4958,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.3983571746069637,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.4792,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.34738883381153784,
"learning_rate": 3.167778872346327e-05,
"loss": 0.4785,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.40251296259859737,
"learning_rate": 3.163182942046815e-05,
"loss": 0.4908,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.37048530080388925,
"learning_rate": 3.158577710967494e-05,
"loss": 0.4901,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.42999468870556284,
"learning_rate": 3.153963215931669e-05,
"loss": 0.4834,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.35963159393214894,
"learning_rate": 3.14933949383672e-05,
"loss": 0.4882,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.4049547176536214,
"learning_rate": 3.144706581653805e-05,
"loss": 0.4845,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.40983907766131217,
"learning_rate": 3.140064516427566e-05,
"loss": 0.4959,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.3969469847975174,
"learning_rate": 3.135413335275833e-05,
"loss": 0.4984,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.4431355290055679,
"learning_rate": 3.130753075389327e-05,
"loss": 0.4894,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.4383084352923244,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.4902,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.5336665923561199,
"learning_rate": 3.121405468537547e-05,
"loss": 0.489,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.38538194236117357,
"learning_rate": 3.116718196315488e-05,
"loss": 0.4577,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.6309092693634908,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.4689,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.45129510229635006,
"learning_rate": 3.107316901675252e-05,
"loss": 0.4881,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.6184773618708512,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.4826,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.37686818649363646,
"learning_rate": 3.097880190800057e-05,
"loss": 0.5021,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.504908377319058,
"learning_rate": 3.093148648549788e-05,
"loss": 0.4842,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.36298875894774024,
"learning_rate": 3.088408365512055e-05,
"loss": 0.4853,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.4305509851495135,
"learning_rate": 3.083659379590034e-05,
"loss": 0.4893,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.4103838493876109,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.475,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.45789767256488056,
"learning_rate": 3.074135451053475e-05,
"loss": 0.4915,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.3812738047332791,
"learning_rate": 3.069360584592015e-05,
"loss": 0.489,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.4246528008481879,
"learning_rate": 3.064577167551816e-05,
"loss": 0.4733,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.4123797150551038,
"learning_rate": 3.059785238180953e-05,
"loss": 0.4943,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.3573996493788802,
"learning_rate": 3.054984834795566e-05,
"loss": 0.4941,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.5249436295584182,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.4919,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.39767403118872935,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.4912,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.5450581833588878,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.4895,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.3713476495173732,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.4868,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.43641493990497526,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.4766,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.3982613900069206,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.4857,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.3911193456264521,
"learning_rate": 3.021147971398632e-05,
"loss": 0.464,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.4173500777953056,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.4941,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.4437872622698337,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.4928,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.4119646620739995,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.4663,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.46281829210077524,
"learning_rate": 3.001632138223332e-05,
"loss": 0.4648,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.5608118036635582,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.4202,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.4458535873518934,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.4149,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.6514115638189755,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.4157,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.42110999334540106,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.4153,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.517018113621199,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.4114,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.48270360123888956,
"learning_rate": 2.972118904869376e-05,
"loss": 0.3872,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.46948695260785966,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.4157,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.45640321914740456,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.4088,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.4683490255130862,
"learning_rate": 2.95725682050846e-05,
"loss": 0.4243,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.40944294376107804,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.4135,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.46716025852423404,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.4095,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.4154502175914396,
"learning_rate": 2.942325848714505e-05,
"loss": 0.4053,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.46113023551787385,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.4024,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.39052904899897717,
"learning_rate": 2.932334127129686e-05,
"loss": 0.4109,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.3849518888086147,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.4177,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.5059776858241279,
"learning_rate": 2.922312585929131e-05,
"loss": 0.4186,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.37891702004264344,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.418,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.4258790712183089,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.4247,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.43145545276896685,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.4159,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.37781192219729376,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.4256,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.4658961336931615,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.4166,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.37467971393518584,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.4126,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.3653580572162924,
"learning_rate": 2.887006998351372e-05,
"loss": 0.422,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.4051008742876637,
"learning_rate": 2.881934649605535e-05,
"loss": 0.4185,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.34646353511146816,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.4091,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.4600665718897344,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.4137,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.43612768978869226,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.4181,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.45417990130628993,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.4058,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.5088641674140494,
"learning_rate": 2.856467940121025e-05,
"loss": 0.4129,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.37383306358354806,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.4116,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.5135567172737797,
"learning_rate": 2.846233032790149e-05,
"loss": 0.4121,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.341427852257275,
"learning_rate": 2.84110540896642e-05,
"loss": 0.4131,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.5268304698887463,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.4222,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.40115906928325806,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.4092,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.6083760770678344,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.4073,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.4599710044996645,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.4143,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.44391232575610656,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.4068,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.3861065886745017,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.4081,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.45861945150952194,
"learning_rate": 2.805026032173773e-05,
"loss": 0.4085,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.3980419874235137,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.4015,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.43504763293319865,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.4266,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.39162273239865364,
"learning_rate": 2.7894660620408e-05,
"loss": 0.4159,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.43791765994848375,
"learning_rate": 2.784266724883173e-05,
"loss": 0.4212,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.36692122494825363,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.4203,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.4894797171945178,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.4147,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.40199805515471415,
"learning_rate": 2.768631254100716e-05,
"loss": 0.4244,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.447868928727005,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.4161,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.44978699144158024,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.3981,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.4616510147540796,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.4163,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.43745377496142146,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.403,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.43819191270884206,
"learning_rate": 2.742449775134085e-05,
"loss": 0.4014,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.4017583421854867,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.4123,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.48920217435567037,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.4161,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.3768463673930125,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.4182,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.48107183228837713,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.4078,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.3662001837937039,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.4101,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.416970625578487,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.4163,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.3901298807307542,
"learning_rate": 2.705547548999996e-05,
"loss": 0.4318,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.44206275902684,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.4146,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.33509418226725546,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.4164,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.382968268115161,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.4254,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.4105647800777282,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.4146,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.3676754496935559,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.4184,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.38594449939037007,
"learning_rate": 2.67369650552155e-05,
"loss": 0.3988,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.3795209211340006,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.4118,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.41522701765259457,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.398,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.41163846018577704,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.4082,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.378242230249019,
"learning_rate": 2.652354173593614e-05,
"loss": 0.4158,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.418437399449027,
"learning_rate": 2.647005443331317e-05,
"loss": 0.4119,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.36181030041110357,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.411,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.41025405593000747,
"learning_rate": 2.636292505301542e-05,
"loss": 0.4248,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.35888286934042957,
"learning_rate": 2.630928383194431e-05,
"loss": 0.4142,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.3787430076221679,
"learning_rate": 2.625559216200781e-05,
"loss": 0.4044,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.36057866906254465,
"learning_rate": 2.620185047252308e-05,
"loss": 0.4189,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.3534758584378497,
"learning_rate": 2.614805919320723e-05,
"loss": 0.4251,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.348422146623841,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.4229,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.39032379822867636,
"learning_rate": 2.60403295859298e-05,
"loss": 0.411,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.3687205778463392,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.4202,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.3730194107381572,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.4147,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.4069840354977184,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.4,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.3591657475372222,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.4111,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.4092602459653616,
"learning_rate": 2.577016790136453e-05,
"loss": 0.4081,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.34937124470275394,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.409,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.3963982303393558,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.42,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.3470927640294334,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.4088,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.36559813934038804,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.4062,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.35407878652036984,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.4303,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.3915960173096782,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.4214,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.35215495685608167,
"learning_rate": 2.539001591449889e-05,
"loss": 0.4041,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.41655744469771405,
"learning_rate": 2.53355326151989e-05,
"loss": 0.4299,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.3910333241052514,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.4035,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.3966133167083973,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.4091,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.3513061233413095,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.4197,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.3948743276854339,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.4308,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.3862797790668624,
"learning_rate": 2.506248490185627e-05,
"loss": 0.4239,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.3380672796669441,
"learning_rate": 2.500775217241923e-05,
"loss": 0.4372,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.36810656620969945,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.4019,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.36138316832391654,
"learning_rate": 2.4898167025948e-05,
"loss": 0.4306,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.3607711966819219,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.4272,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.3640897102902797,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.413,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.39217385056728626,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.4266,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.3494040138065032,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.419,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.3682511726725405,
"learning_rate": 2.46235264422994e-05,
"loss": 0.4248,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.37644446228181566,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.4189,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.35673579537354083,
"learning_rate": 2.451340834511592e-05,
"loss": 0.417,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.35426739064574736,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.4019,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.37648749336126236,
"learning_rate": 2.440314589184958e-05,
"loss": 0.4029,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.35282135915106366,
"learning_rate": 2.434796163357541e-05,
"loss": 0.3997,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.37712956017802773,
"learning_rate": 2.429274260911588e-05,
"loss": 0.4216,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.37180678306403026,
"learning_rate": 2.423748926000084e-05,
"loss": 0.4208,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.4100971043060173,
"learning_rate": 2.418220202803459e-05,
"loss": 0.4148,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.3524794589225651,
"learning_rate": 2.412688135529237e-05,
"loss": 0.4097,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.42807267368905083,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.4096,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.3490320109031565,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.437,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.41908541569820557,
"learning_rate": 2.396072311715187e-05,
"loss": 0.4422,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.36397796503860347,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.4018,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.5039856983582545,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.4133,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.36470139272104224,
"learning_rate": 2.379427985200223e-05,
"loss": 0.4187,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.4211471057062955,
"learning_rate": 2.373873749394972e-05,
"loss": 0.4167,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.37648148071407855,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.4118,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.4265557117377515,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.4092,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.36928474074539575,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.4208,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.3862309412595783,
"learning_rate": 2.351627355800233e-05,
"loss": 0.3928,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.39546972442026573,
"learning_rate": 2.346058617160256e-05,
"loss": 0.3934,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.48717135441456266,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.4263,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.3793262192692635,
"learning_rate": 2.334912883203469e-05,
"loss": 0.4141,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.42551478558380257,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.4227,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.4182183004761265,
"learning_rate": 2.323756437450136e-05,
"loss": 0.4195,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.3614110077625091,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.3938,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.47404083340431596,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.4135,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.35249372889293573,
"learning_rate": 2.307002464849052e-05,
"loss": 0.4053,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.43170561492997306,
"learning_rate": 2.30141283818846e-05,
"loss": 0.4173,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.340975080135885,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.4276,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.35794270260392946,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.3887,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.35614776694427575,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.4145,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.34714524375109473,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.4054,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.38670671913151533,
"learning_rate": 2.273429448107257e-05,
"loss": 0.4169,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.3485501463920474,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.4126,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.3673961433979642,
"learning_rate": 2.262220474347102e-05,
"loss": 0.4312,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.3653093329171211,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.4228,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.37025732962381713,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.4225,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.38563366680354877,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.4116,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.34548161117166176,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.4381,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.38903380074794214,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.4245,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.3910986482705796,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.4097,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.42199682869221594,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.4368,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.36821891876755103,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.4139,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.4488610394316327,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.4251,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.38969979131435223,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.4233,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.3916647509783997,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.4058,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.38892138924482106,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.4266,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.4122437416766424,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.3979,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.4245722266225139,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.403,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.43002680503856733,
"learning_rate": 2.17791064266051e-05,
"loss": 0.4209,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.4235778917874476,
"learning_rate": 2.172276926861741e-05,
"loss": 0.396,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.43834316573134274,
"learning_rate": 2.166641833541255e-05,
"loss": 0.4127,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.4109037998758103,
"learning_rate": 2.161005407757109e-05,
"loss": 0.4157,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.4160048353040478,
"learning_rate": 2.155367694578013e-05,
"loss": 0.4218,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.37567388159783316,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.4049,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.39376421272445533,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.4007,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.34027513928203984,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.4226,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.39974250061183736,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.4174,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.33092472526991257,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.4086,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.3770263116865114,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.421,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.3687926675684961,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.413,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.32990062226923417,
"learning_rate": 2.110225054402733e-05,
"loss": 0.4034,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.3711156964469027,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.4029,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.3864618294289979,
"learning_rate": 2.09892967807724e-05,
"loss": 0.432,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.39126511204695774,
"learning_rate": 2.093280780774835e-05,
"loss": 0.4221,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.40135389760999696,
"learning_rate": 2.087631137601793e-05,
"loss": 0.4209,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.3149968376483946,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.4187,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.3607669183472611,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.4276,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.36591799139261033,
"learning_rate": 2.070678184630458e-05,
"loss": 0.4092,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.3603255796103461,
"learning_rate": 2.06502600977305e-05,
"loss": 0.4132,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.34715628733335,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.4194,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.3616457982370737,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.4239,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.32147728004315596,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.4213,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.35918901173969936,
"learning_rate": 2.042412562887594e-05,
"loss": 0.4139,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.35905594640227745,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.4226,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.33643504644530065,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.4179,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.3206539670194943,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.4037,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.34931099800315557,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.4112,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.3265797768244479,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.4263,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.34176641662674495,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.4205,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.32057019141434545,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.4078,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.324576032157484,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.4161,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.32858178136933347,
"learning_rate": 1.99151687695709e-05,
"loss": 0.4121,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.32302920417457537,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.4213,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.31749664330644606,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.4141,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.34241009310243836,
"learning_rate": 1.974551241345447e-05,
"loss": 0.4264,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.3195955951665441,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.4252,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.3460153054060508,
"learning_rate": 1.963241759676897e-05,
"loss": 0.4089,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.33662167790608,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.4081,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.32590340735171036,
"learning_rate": 1.951933453677638e-05,
"loss": 0.4231,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.354594897338742,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.4151,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.3131939367066724,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.404,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.3348923609627802,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.4174,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.333866654155564,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.4225,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.3212720155854923,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.4231,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.3240872145658091,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.422,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.3083033633410398,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.4159,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.34968731874075776,
"learning_rate": 1.906719219225165e-05,
"loss": 0.4115,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.3301617310074452,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.4165,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.7032042121333083,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.4274,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.39860637418514683,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.4177,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.32449601966522135,
"learning_rate": 1.884128556891707e-05,
"loss": 0.4319,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.4179231028838326,
"learning_rate": 1.878483094691119e-05,
"loss": 0.3995,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3721112169042911,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.4237,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.3408119200052768,
"learning_rate": 1.867195130360947e-05,
"loss": 0.4143,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.42118411434390157,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.4274,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.3546566122659199,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.4212,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.3683303670938878,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.4058,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.41662719921111796,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.415,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.37171145311642584,
"learning_rate": 1.838994592242891e-05,
"loss": 0.4089,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.39770324912347516,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.4073,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.3685614243478002,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.418,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.349410256232596,
"learning_rate": 1.82208935733949e-05,
"loss": 0.4189,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.35901511834660343,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.4244,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.36335362287989464,
"learning_rate": 1.810826238483892e-05,
"loss": 0.4122,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.38246827289363855,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.4109,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.3295667497694108,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.4273,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.35931433955103476,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.4057,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.35289385722138855,
"learning_rate": 1.788318512320838e-05,
"loss": 0.3967,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.3784482680446265,
"learning_rate": 1.782695699829051e-05,
"loss": 0.3988,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.3620653412335145,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.4175,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.3529295801548278,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.413,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.3801403216930023,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.4023,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.3738436198021569,
"learning_rate": 1.76022227484387e-05,
"loss": 0.4227,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.33819683355265445,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.4155,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.40226896433129616,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.421,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.3174643686127531,
"learning_rate": 1.743387180005809e-05,
"loss": 0.4243,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.332971988582128,
"learning_rate": 1.737779525652899e-05,
"loss": 0.399,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.37359847432583476,
"learning_rate": 1.732173968007914e-05,
"loss": 0.422,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.32234023544942636,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.4132,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.3379126960287671,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.4179,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.3072805685273048,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.4044,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.3186345421224538,
"learning_rate": 1.709773600686015e-05,
"loss": 0.4042,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.32537377690996694,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.4213,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.32939576737655235,
"learning_rate": 1.69858716181154e-05,
"loss": 0.4127,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.32594363571587254,
"learning_rate": 1.692997535150948e-05,
"loss": 0.4219,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.31774933066533523,
"learning_rate": 1.687410363273884e-05,
"loss": 0.4014,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.34632446975695524,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.4038,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.3054560974644175,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.411,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.33542038344385894,
"learning_rate": 1.67066402299235e-05,
"loss": 0.4129,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.3681104298257328,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.4122,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.3178424677273126,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.4104,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.39625151863556485,
"learning_rate": 1.653941382839745e-05,
"loss": 0.4039,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.5029186515334423,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.3499,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.39572223211517715,
"learning_rate": 1.642806717162757e-05,
"loss": 0.3489,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.7091970073900354,
"learning_rate": 1.637243646233718e-05,
"loss": 0.3411,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.3966780251785902,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.3377,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.5810449823994424,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.3414,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.41490804812907117,
"learning_rate": 1.620572014799777e-05,
"loss": 0.3322,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.41013556909011967,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.325,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.44577881961399773,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.3328,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.4619816354362409,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.3466,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.39899279344402777,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.332,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.5098192946612119,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.3403,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.4427904083873376,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.3411,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.4112285040801413,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.3231,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.49001467066386667,
"learning_rate": 1.576251073999917e-05,
"loss": 0.325,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.3784272928584796,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.3182,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.5013706416371262,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.3134,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.4382179223147644,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.3127,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.3703928941544541,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.3248,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.48963100044410307,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.3127,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.408531249874786,
"learning_rate": 1.543151434154755e-05,
"loss": 0.3349,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.36073981704696817,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.3229,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.4303499701239575,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.3361,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.365369007923413,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.3287,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.3410697208910492,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.326,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.4429600956553487,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.3348,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.3337403410576911,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.3369,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.3750183895215754,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.3299,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.3844123379612027,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.3331,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.3669319301383747,
"learning_rate": 1.493751509814374e-05,
"loss": 0.3256,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.3627181274923972,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.331,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.36953548841514144,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.3408,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.3662959790849556,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.3262,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.3711091167071754,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.354,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.3724293545511927,
"learning_rate": 1.466446738480111e-05,
"loss": 0.3241,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.3387954656678085,
"learning_rate": 1.460998408550112e-05,
"loss": 0.3169,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.3783775982172417,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.323,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.34133962138064033,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.3207,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.34991340280071287,
"learning_rate": 1.444679451900437e-05,
"loss": 0.3299,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.3644884255960819,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.3341,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.32879365076218625,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.3301,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.34696496661863724,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.33,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.3514828702974075,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.3196,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.3538997341168854,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.3493,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.3335190721829741,
"learning_rate": 1.412162598317602e-05,
"loss": 0.3157,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.32636170300704576,
"learning_rate": 1.406759321421907e-05,
"loss": 0.3412,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.3417585349609768,
"learning_rate": 1.401360788062871e-05,
"loss": 0.3514,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3158527724250949,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.3345,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.3686341451900863,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.3396,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.35265291213399286,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.3143,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.35763758622693603,
"learning_rate": 1.379814952747693e-05,
"loss": 0.3439,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.3376475150161031,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.32,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3476850234253549,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.3438,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.3472392731455894,
"learning_rate": 1.363707494698459e-05,
"loss": 0.3283,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.36067550401026555,
"learning_rate": 1.358348460369265e-05,
"loss": 0.3447,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.3524359300060277,
"learning_rate": 1.352994556668684e-05,
"loss": 0.3353,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.3235813002732538,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.3347,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.3383054216478946,
"learning_rate": 1.342302312350678e-05,
"loss": 0.3281,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.3656077638145719,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.3243,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.3681349613134353,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.322,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.33683352098790587,
"learning_rate": 1.326303494478451e-05,
"loss": 0.3276,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.34925352787995173,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.3368,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.3645573510800313,
"learning_rate": 1.315664479122898e-05,
"loss": 0.3327,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.3504017220141857,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.3319,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.41018040184275656,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.3252,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.3322618205360091,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.3559,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.33093382848756553,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.3256,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.3924533051534646,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.3247,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.3654717180608553,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.3299,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.38621531679999677,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.3387,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.37687182575257827,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.3412,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.3639545080428123,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.3235,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.3652217937022056,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.3459,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.3533332265558164,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.3397,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.3717518942297021,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.3284,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.3328037405171679,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.3377,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.35523716023391294,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.3245,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.34784149046125823,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.3355,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.3292285481758225,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.321,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.4000846378414402,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.3175,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.3284774722695516,
"learning_rate": 1.220938883250288e-05,
"loss": 0.3283,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.33198136982969567,
"learning_rate": 1.215733275116827e-05,
"loss": 0.3399,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.35733909852096984,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.3255,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.32358362753434444,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.328,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.33552335323126525,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.3379,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.3544454724656476,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.3165,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.30606606030437516,
"learning_rate": 1.189800129803139e-05,
"loss": 0.331,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.3477262480075034,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.3289,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.32470761818376337,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.3261,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.34516010503624556,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.3343,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.3254778246721633,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.3331,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.3460124439067998,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.3274,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.32069440155387974,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.3523,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.3321201263000401,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.3198,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.30470406807787204,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.3437,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.33360290328735087,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.343,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.3477409390254483,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.3415,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.3147776515758168,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.3495,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.33397572965758504,
"learning_rate": 1.128231163049258e-05,
"loss": 0.3288,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.30904310695887516,
"learning_rate": 1.123144751066259e-05,
"loss": 0.3309,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.3328474450875233,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.3363,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.31631398650014614,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.342,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.30995897635290404,
"learning_rate": 1.107927745387114e-05,
"loss": 0.3168,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.32566206321859426,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.3409,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.30789443387211174,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.3438,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.3271873960288959,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.3253,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.34127545148022487,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.3367,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.33800391470722513,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.3473,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.33559939280928613,
"learning_rate": 1.077687414070869e-05,
"loss": 0.3257,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.34414582217803397,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.3441,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.3507201836500248,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.3218,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.34452370652255554,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.3434,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.3338328165253491,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.3387,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.4067386088278003,
"learning_rate": 1.052689572753247e-05,
"loss": 0.3388,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.337065369678276,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.3134,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.36298516861396807,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.3301,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.3448452396969425,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.3314,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.330570025064138,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.3356,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.33054336319528327,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.3366,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.32343708065280363,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.3434,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.3226448210811945,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.3277,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.31410711234534827,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.3322,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.3393301858476013,
"learning_rate": 1.008173988718005e-05,
"loss": 0.3345,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.32360178733573736,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.3265,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.30868679048303255,
"learning_rate": 9.983678617766677e-06,
"loss": 0.3416,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.3407040837357195,
"learning_rate": 9.934767922553788e-06,
"loss": 0.3324,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.32272001773062375,
"learning_rate": 9.885937708668371e-06,
"loss": 0.3279,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.3010603374025253,
"learning_rate": 9.837188366555524e-06,
"loss": 0.3319,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.32508012773662,
"learning_rate": 9.788520286013691e-06,
"loss": 0.3406,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.32457886328617824,
"learning_rate": 9.739933856191537e-06,
"loss": 0.3468,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.30422960954755984,
"learning_rate": 9.691429465584866e-06,
"loss": 0.3345,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.32576866361511747,
"learning_rate": 9.643007502033492e-06,
"loss": 0.3345,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.31400607980579,
"learning_rate": 9.594668352718152e-06,
"loss": 0.3235,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.30991179343441916,
"learning_rate": 9.546412404157383e-06,
"loss": 0.3328,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.2988968470360409,
"learning_rate": 9.498240042204471e-06,
"loss": 0.3405,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.31957850072392285,
"learning_rate": 9.450151652044339e-06,
"loss": 0.3449,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.3162108859259316,
"learning_rate": 9.402147618190472e-06,
"loss": 0.3355,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.3192203561884435,
"learning_rate": 9.354228324481841e-06,
"loss": 0.3267,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.32119314748231825,
"learning_rate": 9.306394154079852e-06,
"loss": 0.3292,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.3285796327579748,
"learning_rate": 9.258645489465254e-06,
"loss": 0.3376,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.31536266756813147,
"learning_rate": 9.21098271243509e-06,
"loss": 0.3515,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.3165285523457979,
"learning_rate": 9.163406204099659e-06,
"loss": 0.3423,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.3153009990512768,
"learning_rate": 9.115916344879452e-06,
"loss": 0.3434,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.3355280942032933,
"learning_rate": 9.068513514502126e-06,
"loss": 0.3242,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.3134906236013576,
"learning_rate": 9.021198091999439e-06,
"loss": 0.3344,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.3079949955714399,
"learning_rate": 8.973970455704251e-06,
"loss": 0.3402,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.33955743148936346,
"learning_rate": 8.926830983247495e-06,
"loss": 0.3235,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.2964411235457803,
"learning_rate": 8.879780051555116e-06,
"loss": 0.3492,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.30902266747583584,
"learning_rate": 8.83281803684512e-06,
"loss": 0.3366,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.33702689973246935,
"learning_rate": 8.785945314624532e-06,
"loss": 0.3302,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.30714355031503754,
"learning_rate": 8.739162259686385e-06,
"loss": 0.3276,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.32770186602824963,
"learning_rate": 8.692469246106734e-06,
"loss": 0.331,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.3456035930331097,
"learning_rate": 8.645866647241677e-06,
"loss": 0.3321,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.29713565106945256,
"learning_rate": 8.599354835724348e-06,
"loss": 0.3303,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.33116111673295284,
"learning_rate": 8.552934183461958e-06,
"loss": 0.3366,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.31984931245608655,
"learning_rate": 8.506605061632805e-06,
"loss": 0.3323,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.3101081764455109,
"learning_rate": 8.460367840683308e-06,
"loss": 0.3195,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.31907802487591536,
"learning_rate": 8.414222890325065e-06,
"loss": 0.3246,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3349208492981717,
"learning_rate": 8.368170579531855e-06,
"loss": 0.3122,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.30531875207837905,
"learning_rate": 8.322211276536733e-06,
"loss": 0.3269,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.3321423390699121,
"learning_rate": 8.27634534882907e-06,
"loss": 0.3366,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.3185494793599894,
"learning_rate": 8.230573163151585e-06,
"loss": 0.3235,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.3117539710363005,
"learning_rate": 8.184895085497466e-06,
"loss": 0.3212,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.30709061457875325,
"learning_rate": 8.139311481107401e-06,
"loss": 0.3435,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.34035430810795214,
"learning_rate": 8.093822714466686e-06,
"loss": 0.3247,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.32473673001374503,
"learning_rate": 8.048429149302271e-06,
"loss": 0.3301,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.3064874607916501,
"learning_rate": 8.003131148579902e-06,
"loss": 0.3226,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.33260800177413496,
"learning_rate": 7.957929074501182e-06,
"loss": 0.339,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.3278639767006492,
"learning_rate": 7.912823288500695e-06,
"loss": 0.3422,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.32341295681190213,
"learning_rate": 7.867814151243102e-06,
"loss": 0.3392,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.28610120203973965,
"learning_rate": 7.822902022620267e-06,
"loss": 0.336,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.3457446142226605,
"learning_rate": 7.778087261748377e-06,
"loss": 0.3113,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.3400814274247059,
"learning_rate": 7.733370226965056e-06,
"loss": 0.339,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.31172792809683025,
"learning_rate": 7.688751275826527e-06,
"loss": 0.3314,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.31847633119795304,
"learning_rate": 7.644230765104735e-06,
"loss": 0.3255,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.3103259760068463,
"learning_rate": 7.599809050784502e-06,
"loss": 0.3186,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.31468259885549427,
"learning_rate": 7.555486488060661e-06,
"loss": 0.3257,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.30569252619166604,
"learning_rate": 7.511263431335249e-06,
"loss": 0.3192,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.3324676987129961,
"learning_rate": 7.467140234214656e-06,
"loss": 0.3448,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.3316351468020031,
"learning_rate": 7.423117249506777e-06,
"loss": 0.3401,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.3000345154218382,
"learning_rate": 7.379194829218231e-06,
"loss": 0.3474,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.31570730741693576,
"learning_rate": 7.335373324551518e-06,
"loss": 0.3356,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.35391333576607925,
"learning_rate": 7.291653085902224e-06,
"loss": 0.3242,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.3252309672894141,
"learning_rate": 7.248034462856213e-06,
"loss": 0.3338,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.29343298724551675,
"learning_rate": 7.20451780418683e-06,
"loss": 0.3206,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.31301893796595875,
"learning_rate": 7.161103457852128e-06,
"loss": 0.3169,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.34643403939623607,
"learning_rate": 7.117791770992046e-06,
"loss": 0.3225,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.30430111718334857,
"learning_rate": 7.074583089925688e-06,
"loss": 0.3174,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.29257035230637235,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.3271,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.32770707936164795,
"learning_rate": 6.988476126329604e-06,
"loss": 0.3128,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.3074253451573704,
"learning_rate": 6.945578532308857e-06,
"loss": 0.3348,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.3087901898069212,
"learning_rate": 6.902785321094301e-06,
"loss": 0.3254,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.2954130526800374,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.3308,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.3095726986650257,
"learning_rate": 6.817513414939892e-06,
"loss": 0.328,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.3141949034708538,
"learning_rate": 6.775035401831924e-06,
"loss": 0.3422,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.32163789980887286,
"learning_rate": 6.732663135188473e-06,
"loss": 0.3265,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.28437143657276936,
"learning_rate": 6.690396953817055e-06,
"loss": 0.3449,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.29328628045488325,
"learning_rate": 6.648237195676927e-06,
"loss": 0.3338,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.2892332085556652,
"learning_rate": 6.606184197876395e-06,
"loss": 0.3405,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.3027509560897138,
"learning_rate": 6.564238296670109e-06,
"loss": 0.3302,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.2971138218921373,
"learning_rate": 6.52239982745638e-06,
"loss": 0.3167,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.29742243767278326,
"learning_rate": 6.480669124774484e-06,
"loss": 0.3237,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.3129180707850192,
"learning_rate": 6.439046522302013e-06,
"loss": 0.3139,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.31585330684856044,
"learning_rate": 6.397532352852191e-06,
"loss": 0.3234,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.3039413078319902,
"learning_rate": 6.356126948371202e-06,
"loss": 0.3221,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.3107190569216704,
"learning_rate": 6.314830639935559e-06,
"loss": 0.3225,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.3050010975801508,
"learning_rate": 6.273643757749448e-06,
"loss": 0.3318,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.3177608699996375,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.3268,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.31245660064580355,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.3116,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.30307273108739663,
"learning_rate": 6.150742957589784e-06,
"loss": 0.3372,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.3132363336365843,
"learning_rate": 6.109997064904774e-06,
"loss": 0.3243,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.3033163829649213,
"learning_rate": 6.069362236313092e-06,
"loss": 0.3342,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.3166105454262171,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.3317,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.29168432104225733,
"learning_rate": 5.988427070179108e-06,
"loss": 0.3459,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.28992794511544395,
"learning_rate": 5.948127379792234e-06,
"loss": 0.33,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.2943736335960581,
"learning_rate": 5.90794004780439e-06,
"loss": 0.3416,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.2971050808968395,
"learning_rate": 5.867865395552405e-06,
"loss": 0.3237,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.30516834369846496,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.3227,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.2990191751565575,
"learning_rate": 5.788055411095892e-06,
"loss": 0.3415,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.300847766896369,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.3273,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.31833446298175005,
"learning_rate": 5.708699979051613e-06,
"loss": 0.3324,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.3051825377407012,
"learning_rate": 5.66919351390751e-06,
"loss": 0.3263,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.3036430889023723,
"learning_rate": 5.629801637510188e-06,
"loss": 0.3229,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.33219673847435166,
"learning_rate": 5.590524664836043e-06,
"loss": 0.3115,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.2986253681125616,
"learning_rate": 5.551362909942708e-06,
"loss": 0.3262,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.31008456237046855,
"learning_rate": 5.51231668596653e-06,
"loss": 0.3281,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.30891874462891966,
"learning_rate": 5.47338630512007e-06,
"loss": 0.3331,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.30541724332595954,
"learning_rate": 5.434572078689617e-06,
"loss": 0.3338,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.31521074476210204,
"learning_rate": 5.395874317032694e-06,
"loss": 0.3228,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.3029011039050774,
"learning_rate": 5.357293329575573e-06,
"loss": 0.3341,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.3072020051055282,
"learning_rate": 5.318829424810797e-06,
"loss": 0.3447,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.3076240542547005,
"learning_rate": 5.280482910294728e-06,
"loss": 0.3306,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.3185283276457241,
"learning_rate": 5.242254092645078e-06,
"loss": 0.3186,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.29506504215199536,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.3273,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.2915801641946289,
"learning_rate": 5.166150769707892e-06,
"loss": 0.3268,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.33921791288153536,
"learning_rate": 5.128276872940503e-06,
"loss": 0.3125,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.3323432794799164,
"learning_rate": 5.090521890074953e-06,
"loss": 0.3442,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.28894954958597113,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.3315,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.29108572105889646,
"learning_rate": 5.015369872647442e-06,
"loss": 0.3261,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.29788590090879674,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.321,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.3028383472714517,
"learning_rate": 4.940697121074675e-06,
"loss": 0.343,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.2885158687635852,
"learning_rate": 4.903541216934864e-06,
"loss": 0.343,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.30052894901131866,
"learning_rate": 4.8665060236772e-06,
"loss": 0.3112,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.29523771020669665,
"learning_rate": 4.829591837434102e-06,
"loss": 0.3306,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.30079340907893665,
"learning_rate": 4.792798953370424e-06,
"loss": 0.328,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.31174751275309587,
"learning_rate": 4.756127665681083e-06,
"loss": 0.3194,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.3076835949043808,
"learning_rate": 4.719578267588725e-06,
"loss": 0.3291,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.2784998003576208,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.328,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.6140565373456089,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.2652,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.4404917915008651,
"learning_rate": 4.610664328486485e-06,
"loss": 0.2601,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.33242975985809214,
"learning_rate": 4.5746054014809e-06,
"loss": 0.2797,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.5729727070992469,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.2801,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.5714506146915019,
"learning_rate": 4.502857857942204e-06,
"loss": 0.259,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.374158841958129,
"learning_rate": 4.467169815100549e-06,
"loss": 0.2628,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.4595767426493985,
"learning_rate": 4.431605972355137e-06,
"loss": 0.2675,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.5329420883230571,
"learning_rate": 4.396166614073516e-06,
"loss": 0.2775,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.4443457196075269,
"learning_rate": 4.360852023627846e-06,
"loss": 0.2632,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.33203485838183494,
"learning_rate": 4.325662483392656e-06,
"loss": 0.2793,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.4007456485777789,
"learning_rate": 4.29059827474257e-06,
"loss": 0.2692,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.45679048014069856,
"learning_rate": 4.255659678050063e-06,
"loss": 0.2795,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.38742853005921074,
"learning_rate": 4.22084697268323e-06,
"loss": 0.2623,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.3382116976302307,
"learning_rate": 4.186160437003535e-06,
"loss": 0.2668,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.3627945288674671,
"learning_rate": 4.151600348363582e-06,
"loss": 0.2492,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.34723177427087576,
"learning_rate": 4.11716698310493e-06,
"loss": 0.2697,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.3671334172798468,
"learning_rate": 4.082860616555848e-06,
"loss": 0.252,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.342919655988502,
"learning_rate": 4.048681523029125e-06,
"loss": 0.2743,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.34443974430125784,
"learning_rate": 4.014629975819882e-06,
"loss": 0.2655,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.3264506755544591,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.2767,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.3442934359408843,
"learning_rate": 3.94691060843287e-06,
"loss": 0.2639,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.3615714661962304,
"learning_rate": 3.913243329737353e-06,
"loss": 0.2463,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.33926654279544677,
"learning_rate": 3.879704680319503e-06,
"loss": 0.2501,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.317193496712978,
"learning_rate": 3.846294928353471e-06,
"loss": 0.2705,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.3258446018286081,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.2519,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.3139323095852117,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.2588,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.3239752853986446,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.2781,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.3339907781947759,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.2612,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.31420930602735514,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.2545,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.3251131891972874,
"learning_rate": 3.648558150665886e-06,
"loss": 0.2555,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.3146127044796944,
"learning_rate": 3.616058102895075e-06,
"loss": 0.2805,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.3213313556624278,
"learning_rate": 3.583689060687079e-06,
"loss": 0.2672,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.32174430439895996,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.255,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.3312047415136393,
"learning_rate": 3.51934502719796e-06,
"loss": 0.2396,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.3198013664493681,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.2743,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.3254012649997914,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.2519,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.3111204766325758,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.2543,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.3075583723807622,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.2662,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.3129452023140067,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.2722,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.3260230934094378,
"learning_rate": 3.32948376099375e-06,
"loss": 0.2689,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.3036872387310707,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.2562,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.306148477397116,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.264,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.31992958894125717,
"learning_rate": 3.236349233544438e-06,
"loss": 0.2878,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.3152479470262191,
"learning_rate": 3.205572144561604e-06,
"loss": 0.2484,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.29730413457694793,
"learning_rate": 3.174929343376374e-06,
"loss": 0.2655,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.31587752713982564,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.2509,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.3091157478901203,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.2513,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.30851703057305907,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.2618,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.29875605755787343,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.2621,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.3078303332732579,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.2738,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.3248252383999141,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.2459,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.31070842369334706,
"learning_rate": 2.964210239224696e-06,
"loss": 0.2363,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.3234938216747012,
"learning_rate": 2.934650470499869e-06,
"loss": 0.2637,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.3082408582269153,
"learning_rate": 2.905227155855128e-06,
"loss": 0.2724,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.3075794215828019,
"learning_rate": 2.875940530558519e-06,
"loss": 0.2525,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.3023259553605899,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.2689,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.3117993069154726,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.2559,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.30601871316214896,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.27,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.3057030523897095,
"learning_rate": 2.760165589921184e-06,
"loss": 0.2672,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.3045889154343344,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.2477,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.2945113307957829,
"learning_rate": 2.703104292151164e-06,
"loss": 0.2673,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.3137405204565335,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.2625,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.3080974901674334,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.2521,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.30491967134738257,
"learning_rate": 2.618550200289303e-06,
"loss": 0.2621,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.32086254347056353,
"learning_rate": 2.59064316684412e-06,
"loss": 0.2516,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.3089702464596926,
"learning_rate": 2.562875338152264e-06,
"loss": 0.2538,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.31796103068945136,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.2573,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.3255553993158029,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.2392,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.31091472493304184,
"learning_rate": 2.480409295328803e-06,
"loss": 0.2581,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.2964206949591107,
"learning_rate": 2.45320049480108e-06,
"loss": 0.266,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.3143068809842156,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.2668,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.2977285665500604,
"learning_rate": 2.399204021408421e-06,
"loss": 0.261,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.30469795988388154,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.2641,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.3008316385326882,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.258,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.2977931634002165,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.2683,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.31886237465556,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.255,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.30859525046213315,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.2544,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.30163983878105377,
"learning_rate": 2.24059906434998e-06,
"loss": 0.2727,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.2990959439028651,
"learning_rate": 2.214660694699091e-06,
"loss": 0.2851,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.3052236677442969,
"learning_rate": 2.188864536147355e-06,
"loss": 0.2761,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.3123503494418708,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.247,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.30678791860057364,
"learning_rate": 2.137699676264171e-06,
"loss": 0.2583,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2945378333281356,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.2772,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.3189987499590066,
"learning_rate": 2.08710612114857e-06,
"loss": 0.2553,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.3134839996749211,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.2484,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.3068287392675579,
"learning_rate": 2.0370854889762e-06,
"loss": 0.2634,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.2962417917886189,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.2736,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.314188304975855,
"learning_rate": 1.987639379598427e-06,
"loss": 0.2717,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.30469472062051733,
"learning_rate": 1.963132265887919e-06,
"loss": 0.266,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.29821991637721396,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.2674,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.3118450250574211,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.2632,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.2930927765056414,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.2676,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.3109385051003846,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.2638,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.31087185225371866,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.2666,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.3100970646148723,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.2493,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.3059168146761295,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.2574,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.3047177072848232,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.25,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.3046307223380484,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.258,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.3245007678603466,
"learning_rate": 1.726025117089718e-06,
"loss": 0.2558,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.3091564071503492,
"learning_rate": 1.70311497206598e-06,
"loss": 0.263,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.3057487782277161,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.2717,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.28714370193195937,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.273,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.29532998968350915,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.2658,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.31337184582381544,
"learning_rate": 1.612939220355272e-06,
"loss": 0.2487,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.2929292315651962,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.2803,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.2951322963780015,
"learning_rate": 1.568732761074121e-06,
"loss": 0.2764,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.31408270087866064,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.2721,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.30246470238512774,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.2552,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.3132947978054432,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.2476,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.31333587115587047,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.2597,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.3123623818776228,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.263,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.2967773080358545,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.2573,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.29800992981816965,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.2789,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.3103293230410108,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.2669,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.30613635757337876,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.2721,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.30027549152893535,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.2566,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.29531100927679543,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.2618,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.2945728086639293,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.2644,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.3033716695580158,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.2504,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.2898391768345869,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.2617,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.30258078677416506,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.2685,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.29759768091456706,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.2646,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.3036969156477461,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.2605,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.3119177087163645,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.2704,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.3020863744763518,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.2599,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.2871446876504259,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.2571,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.297081416590952,
"learning_rate": 1.140415161609505e-06,
"loss": 0.2752,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.29449765628329483,
"learning_rate": 1.121666393356604e-06,
"loss": 0.2729,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.2973221294548874,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.2629,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.2920250122607044,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.2497,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.2907058360248583,
"learning_rate": 1.066326386135097e-06,
"loss": 0.2611,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.30901537519175937,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.2712,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.3031519069416129,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.2574,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.3104977070164243,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.251,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.29086985768136564,
"learning_rate": 9.9465987684501e-07,
"loss": 0.2633,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.296360472112165,
"learning_rate": 9.771228098856534e-07,
"loss": 0.2755,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.29144325648137087,
"learning_rate": 9.597378493454013e-07,
"loss": 0.2836,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.2954952484820719,
"learning_rate": 9.425051342339353e-07,
"loss": 0.2548,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.33545429669991816,
"learning_rate": 9.254248023435864e-07,
"loss": 0.251,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.3206372625562427,
"learning_rate": 9.084969902482288e-07,
"loss": 0.2536,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.30071554013115326,
"learning_rate": 8.917218333021993e-07,
"loss": 0.2576,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.30288058711672405,
"learning_rate": 8.750994656392109e-07,
"loss": 0.2576,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.29471964711625503,
"learning_rate": 8.58630020171265e-07,
"loss": 0.2529,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.30130451493584776,
"learning_rate": 8.423136285876099e-07,
"loss": 0.2635,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.29654768954530514,
"learning_rate": 8.26150421353682e-07,
"loss": 0.2528,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.30779687734059674,
"learning_rate": 8.101405277100549e-07,
"loss": 0.2546,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.2916264893480381,
"learning_rate": 7.942840756714077e-07,
"loss": 0.2656,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.288884269221097,
"learning_rate": 7.785811920255093e-07,
"loss": 0.2659,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.28985676226753754,
"learning_rate": 7.630320023322024e-07,
"loss": 0.267,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.2942409295948515,
"learning_rate": 7.476366309223903e-07,
"loss": 0.2783,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.29346053872012656,
"learning_rate": 7.323952008970537e-07,
"loss": 0.2679,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.292866247409529,
"learning_rate": 7.173078341262618e-07,
"loss": 0.2555,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.3008885885184485,
"learning_rate": 7.023746512481989e-07,
"loss": 0.2531,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.3047815091657204,
"learning_rate": 6.875957716681902e-07,
"loss": 0.2594,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.2887158070362815,
"learning_rate": 6.729713135577687e-07,
"loss": 0.2621,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.2958341501912404,
"learning_rate": 6.585013938537143e-07,
"loss": 0.2624,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.3047540887738058,
"learning_rate": 6.441861282571115e-07,
"loss": 0.2501,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.29348507489654785,
"learning_rate": 6.300256312324471e-07,
"loss": 0.2703,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.30189123739962553,
"learning_rate": 6.160200160066776e-07,
"loss": 0.2653,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.2927692482947967,
"learning_rate": 6.021693945683327e-07,
"loss": 0.2427,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.3004466422065112,
"learning_rate": 5.884738776666044e-07,
"loss": 0.2716,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.3078124666914736,
"learning_rate": 5.749335748104878e-07,
"loss": 0.2655,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.3117921166156773,
"learning_rate": 5.615485942678778e-07,
"loss": 0.2518,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.3023320754428296,
"learning_rate": 5.483190430647289e-07,
"loss": 0.27,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.28275652106879773,
"learning_rate": 5.352450269841747e-07,
"loss": 0.2789,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.3086847533118679,
"learning_rate": 5.22326650565701e-07,
"loss": 0.2476,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.2922637865455216,
"learning_rate": 5.095640171043048e-07,
"loss": 0.2703,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.3115260916155263,
"learning_rate": 4.969572286496594e-07,
"loss": 0.2532,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.2857180591173472,
"learning_rate": 4.845063860053079e-07,
"loss": 0.2858,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.2994817280730904,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.263,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.29697142253551545,
"learning_rate": 4.600729351261857e-07,
"loss": 0.258,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.28421885371973865,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.2645,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.30689377263715245,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.257,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.2834208614661261,
"learning_rate": 4.245948007320588e-07,
"loss": 0.2522,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.31410593279192567,
"learning_rate": 4.130816799401971e-07,
"loss": 0.2686,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.3071632608280912,
"learning_rate": 4.017251756252827e-07,
"loss": 0.2498,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2855503989218659,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.2781,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.3083096206543167,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.2602,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.3010255092065491,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.2669,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.2887031154382034,
"learning_rate": 3.578671204615347e-07,
"loss": 0.2653,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.286698597114262,
"learning_rate": 3.472950355547267e-07,
"loss": 0.2635,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.29957616352935695,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.2644,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.30791673671079167,
"learning_rate": 3.266223765343224e-07,
"loss": 0.2532,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.29340590965163854,
"learning_rate": 3.16521967718757e-07,
"loss": 0.2701,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.30299503276076256,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.2724,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.29317308084337607,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.2462,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.2995584329412469,
"learning_rate": 2.871653894244597e-07,
"loss": 0.2468,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.29227925090666224,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.2656,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.31129626195865495,
"learning_rate": 2.683823233205485e-07,
"loss": 0.2578,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.29686766257182173,
"learning_rate": 2.592274132703287e-07,
"loss": 0.264,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.29520567605785686,
"learning_rate": 2.502303499116754e-07,
"loss": 0.2571,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.2903230254841975,
"learning_rate": 2.413912051848666e-07,
"loss": 0.2628,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.29373805754943694,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.2556,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.30077233826941213,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.2584,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.29548522362045165,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.2709,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.2801726276417024,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.2767,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.2970290828424477,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.2807,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.29197239001385206,
"learning_rate": 1.916764975761254e-07,
"loss": 0.264,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.2926238710679969,
"learning_rate": 1.839446914621923e-07,
"loss": 0.2488,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.31388677709688395,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.2656,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.2863799130586825,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.2714,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2944434082071575,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.264,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.2957063840994442,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.275,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.2884347098073302,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.2677,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.30015090781473003,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.2698,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.2883875531478792,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.2741,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.3167857116021705,
"learning_rate": 1.277993209368744e-07,
"loss": 0.2523,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.30029690369530765,
"learning_rate": 1.214956261808542e-07,
"loss": 0.2902,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.31831392524630203,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.2614,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.28014302273862224,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.283,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.29621194569182624,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.258,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2946727174209939,
"learning_rate": 9.787081204466253e-08,
"loss": 0.2529,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.2947596375569267,
"learning_rate": 9.236233586200627e-08,
"loss": 0.2578,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.29617971979305174,
"learning_rate": 8.70130406196279e-08,
"loss": 0.249,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.28096566413336116,
"learning_rate": 8.182296909035047e-08,
"loss": 0.2855,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.28870032530220624,
"learning_rate": 7.679216277384439e-08,
"loss": 0.2584,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.29138658081800806,
"learning_rate": 7.19206618963053e-08,
"loss": 0.2682,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.2921334756952463,
"learning_rate": 6.720850541012347e-08,
"loss": 0.2702,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.2899424899071104,
"learning_rate": 6.265573099357714e-08,
"loss": 0.2664,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.276793538974755,
"learning_rate": 5.826237505053067e-08,
"loss": 0.2632,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.30018116093265595,
"learning_rate": 5.4028472710137e-08,
"loss": 0.2635,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.28115117650916954,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.275,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.2913098644778604,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.2602,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.2890311825948716,
"learning_rate": 4.228381947004812e-08,
"loss": 0.2626,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.288498863754315,
"learning_rate": 3.868805732811343e-08,
"loss": 0.2668,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.291165493758173,
"learning_rate": 3.525190530455636e-08,
"loss": 0.2646,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.28599576522666065,
"learning_rate": 3.197539087475887e-08,
"loss": 0.2616,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.29334860988566935,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.2547,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.3020971635435551,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.2665,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.2826854090113753,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.2731,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.32604943678894666,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.245,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.3010114174238244,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.2584,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.2986141424941457,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.2686,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.28248778719093803,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.2781,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.29365559930521984,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.2511,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.2886088534644375,
"learning_rate": 9.67435416023843e-09,
"loss": 0.2638,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.29608124566070454,
"learning_rate": 7.995445897566178e-09,
"loss": 0.2607,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.2940889228304093,
"learning_rate": 6.476393170578021e-09,
"loss": 0.2625,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.3106884280694384,
"learning_rate": 5.11720812557348e-09,
"loss": 0.2468,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.2948541264516829,
"learning_rate": 3.917901630563492e-09,
"loss": 0.2557,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.2871798628546014,
"learning_rate": 2.878483275170485e-09,
"loss": 0.2547,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.29718274268973505,
"learning_rate": 1.998961370555108e-09,
"loss": 0.2629,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.2882190055952896,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.2706,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.28499526351012416,
"learning_rate": 7.196337656112384e-10,
"loss": 0.2642,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.2862297551980583,
"learning_rate": 3.198382947511292e-10,
"loss": 0.2614,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.2956108557778653,
"learning_rate": 7.995973352770137e-11,
"loss": 0.2751,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.28808446109010327,
"learning_rate": 0.0,
"loss": 0.2679,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 1750256009674752.0,
"train_loss": 0.42420748597697205,
"train_runtime": 24391.1253,
"train_samples_per_second": 6.478,
"train_steps_per_second": 0.051
}
],
"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": 1750256009674752.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}