meta_chat_reasoning_25_75_system / trainer_state.json
ryanmarten's picture
End of training
49d0624 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": 8.922756620697132,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.0635,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 7.462758378733854,
"learning_rate": 6.451612903225807e-07,
"loss": 0.993,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.639574983040356,
"learning_rate": 9.67741935483871e-07,
"loss": 0.9218,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 13.767638438345536,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.0637,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 6.1214759734762625,
"learning_rate": 1.6129032258064516e-06,
"loss": 0.9747,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 5.385968869738027,
"learning_rate": 1.935483870967742e-06,
"loss": 0.9286,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 4.79449858911541,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.8852,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 5.231271989291822,
"learning_rate": 2.580645161290323e-06,
"loss": 0.9517,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 3.238194040844263,
"learning_rate": 2.903225806451613e-06,
"loss": 0.8344,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 3.768487673008824,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9031,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.6689020897085847,
"learning_rate": 3.548387096774194e-06,
"loss": 0.7785,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 2.166772476797284,
"learning_rate": 3.870967741935484e-06,
"loss": 0.7712,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 2.7670268864044516,
"learning_rate": 4.193548387096774e-06,
"loss": 0.8125,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 3.1355458816867308,
"learning_rate": 4.516129032258065e-06,
"loss": 0.7915,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 2.721387607102408,
"learning_rate": 4.838709677419355e-06,
"loss": 0.7755,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.555650766609947,
"learning_rate": 5.161290322580646e-06,
"loss": 0.7878,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.4283299598857555,
"learning_rate": 5.483870967741935e-06,
"loss": 0.755,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.321575626114488,
"learning_rate": 5.806451612903226e-06,
"loss": 0.759,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.545004599525289,
"learning_rate": 6.129032258064517e-06,
"loss": 0.7314,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.9315535995243904,
"learning_rate": 6.451612903225806e-06,
"loss": 0.6934,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 2.4139391845956872,
"learning_rate": 6.774193548387097e-06,
"loss": 0.7331,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.6948593211591683,
"learning_rate": 7.096774193548388e-06,
"loss": 0.686,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.782492482615504,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.716,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.698942092825406,
"learning_rate": 7.741935483870968e-06,
"loss": 0.7236,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.394499786417354,
"learning_rate": 8.064516129032258e-06,
"loss": 0.6698,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.3558932941480286,
"learning_rate": 8.387096774193549e-06,
"loss": 0.6567,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 1.1726283188152606,
"learning_rate": 8.70967741935484e-06,
"loss": 0.6809,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.3225224528390767,
"learning_rate": 9.03225806451613e-06,
"loss": 0.6932,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 1.1929553484783963,
"learning_rate": 9.35483870967742e-06,
"loss": 0.6139,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 1.356530737240821,
"learning_rate": 9.67741935483871e-06,
"loss": 0.6279,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 1.343278677110846,
"learning_rate": 1e-05,
"loss": 0.6669,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 1.0693287752322858,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.6125,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 1.6360257444223718,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.6523,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 1.0791921968062388,
"learning_rate": 1.096774193548387e-05,
"loss": 0.6402,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 1.1391965125331969,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.5775,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 1.1511729642188147,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.6499,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.9284074873278193,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.6208,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 1.0068759841938995,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.6107,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.9727752700633634,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.5964,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 1.074593921999239,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.6202,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 1.044374512409201,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6278,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.9047898046445844,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.6056,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.937928443334301,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.5483,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 1.0652336350224825,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.6004,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.8522041471466061,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.5885,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 1.751407529834859,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6156,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 1.1313751687496432,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.586,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.9002401915419473,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6089,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 1.3711495183992402,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.595,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 1.0275423372792705,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6069,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 1.305165520438895,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.598,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.9881501313968295,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.5956,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 1.0460247777325016,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.5956,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.984134294688186,
"learning_rate": 1.741935483870968e-05,
"loss": 0.6175,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.786109018649995,
"learning_rate": 1.774193548387097e-05,
"loss": 0.6226,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 1.3507753061994094,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6269,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.9003919455334077,
"learning_rate": 1.838709677419355e-05,
"loss": 0.5605,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.9541505733286052,
"learning_rate": 1.870967741935484e-05,
"loss": 0.5991,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 1.0610748167959207,
"learning_rate": 1.903225806451613e-05,
"loss": 0.6041,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 1.051494618409292,
"learning_rate": 1.935483870967742e-05,
"loss": 0.5984,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 1.5878708960485302,
"learning_rate": 1.967741935483871e-05,
"loss": 0.5941,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.8737557494208722,
"learning_rate": 2e-05,
"loss": 0.574,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 1.1178219939761165,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.6078,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.9937363864768363,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.5801,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 1.0437054630427203,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.5982,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 1.0640161835786457,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.6112,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.972948894948768,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6305,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.9680596266500782,
"learning_rate": 2.193548387096774e-05,
"loss": 0.608,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.9359879154469849,
"learning_rate": 2.225806451612903e-05,
"loss": 0.5944,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.9410175872021985,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.568,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 1.0785167850950501,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.5956,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.2249994760166656,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.6101,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.8303432915233646,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.577,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.7815227546857115,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.588,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.8313376418168616,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.545,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.8838435021137563,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.5876,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.1066535644084792,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.5906,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.8169487414335258,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.5679,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 152.18825285998898,
"learning_rate": 2.548387096774194e-05,
"loss": 0.5697,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.7782246328019985,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.5584,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.7904815517037597,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.5603,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.8896542863044067,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.5483,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.7337134441164757,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.5847,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.9873589292933541,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.5744,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.2780823510748216,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.6416,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.1292526590370748,
"learning_rate": 2.774193548387097e-05,
"loss": 0.6099,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.090580019115647,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.5716,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.9075716254868689,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.5844,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.8091478069538228,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.5779,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.9187885852042238,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.6279,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.9655955272270498,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6002,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.1739838491142596,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.5925,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.7797133494383088,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.5702,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.88262614831411,
"learning_rate": 3.032258064516129e-05,
"loss": 0.5763,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.094407392256796,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.6092,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.7546188161358619,
"learning_rate": 3.096774193548387e-05,
"loss": 0.5845,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.8886087516207225,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.6063,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 4.113138028921058,
"learning_rate": 3.161290322580645e-05,
"loss": 0.5696,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.009244302458443,
"learning_rate": 3.193548387096774e-05,
"loss": 0.6002,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 0.7605483280751203,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.5827,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.7487948798471358,
"learning_rate": 3.258064516129033e-05,
"loss": 0.5571,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.7529037121885555,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.5979,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.8987801633791043,
"learning_rate": 3.322580645161291e-05,
"loss": 0.5895,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.8060442860704915,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.5817,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.9411271209384184,
"learning_rate": 3.387096774193549e-05,
"loss": 0.5901,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.6909247589809421,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.5673,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.6407979341823886,
"learning_rate": 3.451612903225807e-05,
"loss": 0.5866,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.7666336039898485,
"learning_rate": 3.483870967741936e-05,
"loss": 0.5819,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.7815428253906812,
"learning_rate": 3.516129032258065e-05,
"loss": 0.6018,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.007887497339938,
"learning_rate": 3.548387096774194e-05,
"loss": 0.5849,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.1691790317355635,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.5964,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.8856232048712602,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6005,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.7672033510464971,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.5741,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.7750384406876565,
"learning_rate": 3.67741935483871e-05,
"loss": 0.5748,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 0.881516993766239,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6078,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 2.4909205571611612,
"learning_rate": 3.741935483870968e-05,
"loss": 0.596,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.8417657677410285,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.5736,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.8623899642002635,
"learning_rate": 3.806451612903226e-05,
"loss": 0.6106,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.3740630063980765,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.5966,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.8145228241956015,
"learning_rate": 3.870967741935484e-05,
"loss": 0.5491,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 0.8438621260089683,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.6227,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 0.8197790889174071,
"learning_rate": 3.935483870967742e-05,
"loss": 0.6126,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.929517945151773,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.6431,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.4728269059482775,
"learning_rate": 4e-05,
"loss": 0.6475,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.0335016648059694,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.5895,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.8931659214045833,
"learning_rate": 3.999968016170525e-05,
"loss": 0.6145,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.3071255331339113,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.5906,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.8617790232783059,
"learning_rate": 3.999872065705065e-05,
"loss": 0.5775,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.8639038277682108,
"learning_rate": 3.999800103862945e-05,
"loss": 0.5876,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.0841336136487778,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.6444,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.767960666077071,
"learning_rate": 3.999608209836944e-05,
"loss": 0.5663,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.206720431522116,
"learning_rate": 3.999488279187443e-05,
"loss": 0.5886,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.1220194646713841,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.6657,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.1381398192673944,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.5842,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.7646253351101681,
"learning_rate": 3.999032564583976e-05,
"loss": 0.5832,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.0310165846571635,
"learning_rate": 3.998848689546591e-05,
"loss": 0.5818,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.9251828011193141,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.5845,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.7240392555121958,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.5789,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 1.0169736158161264,
"learning_rate": 3.998201174509306e-05,
"loss": 0.5635,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.6879797922605607,
"learning_rate": 3.997953378607961e-05,
"loss": 0.5432,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.868255704066541,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.6016,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.8725655650520259,
"learning_rate": 3.997409862168454e-05,
"loss": 0.6073,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.7417609442676335,
"learning_rate": 3.997114145976237e-05,
"loss": 0.6255,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.9376889538115811,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.5719,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.7382797614419818,
"learning_rate": 3.996474809469545e-05,
"loss": 0.5777,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.8104247498258861,
"learning_rate": 3.996131194267189e-05,
"loss": 0.61,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 1.9599462078857346,
"learning_rate": 3.995771618052995e-05,
"loss": 0.6394,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.8371852933201331,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.5597,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.7719788836089364,
"learning_rate": 3.995004594217343e-05,
"loss": 0.6129,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.8803822413080876,
"learning_rate": 3.994597152728987e-05,
"loss": 0.5963,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 1.1631601847778035,
"learning_rate": 3.994173762494947e-05,
"loss": 0.6086,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.7961879835667748,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.5465,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.8495546108064649,
"learning_rate": 3.993279149458988e-05,
"loss": 0.6165,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.7368312685206785,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.5482,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.7769872311502489,
"learning_rate": 3.992320783722616e-05,
"loss": 0.5714,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.7196952035439147,
"learning_rate": 3.991817703090965e-05,
"loss": 0.5857,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 1.3965632951423828,
"learning_rate": 3.991298695938038e-05,
"loss": 0.6144,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.6618955239977684,
"learning_rate": 3.9907637664138e-05,
"loss": 0.546,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 1.4698885235557029,
"learning_rate": 3.990212918795534e-05,
"loss": 0.5741,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.7217634782817546,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.6169,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.8907187202862861,
"learning_rate": 3.989063487022416e-05,
"loss": 0.5889,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 1.1205805131867252,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.6345,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.6660590913734938,
"learning_rate": 3.987850437381915e-05,
"loss": 0.5797,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.7532903774511963,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6068,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.8678423139275117,
"learning_rate": 3.986573808672002e-05,
"loss": 0.5928,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.675779107290499,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.5735,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.6793146807514133,
"learning_rate": 3.985233641724152e-05,
"loss": 0.5589,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.909812948998432,
"learning_rate": 3.984539744726554e-05,
"loss": 0.6061,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.6938084141004376,
"learning_rate": 3.983829979402038e-05,
"loss": 0.6277,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.9890972945984424,
"learning_rate": 3.983104351425871e-05,
"loss": 0.5917,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.7675933573818587,
"learning_rate": 3.982362866600154e-05,
"loss": 0.5384,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.680255278624252,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.5803,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.8001354375772757,
"learning_rate": 3.980832350242388e-05,
"loss": 0.6154,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.7448484198666343,
"learning_rate": 3.980043330948306e-05,
"loss": 0.5986,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.6187874279017339,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.5463,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.7142458062159085,
"learning_rate": 3.978417801674579e-05,
"loss": 0.6061,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 1.0291454836673994,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.5641,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.8029128011503961,
"learning_rate": 3.976728995023254e-05,
"loss": 0.5616,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.7058019066394278,
"learning_rate": 3.975860879481514e-05,
"loss": 0.5737,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.7273926087887463,
"learning_rate": 3.974976965008833e-05,
"loss": 0.584,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.8402773873517891,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.5618,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.728120270798104,
"learning_rate": 3.973161767667946e-05,
"loss": 0.597,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.7480718263561341,
"learning_rate": 3.972230499314009e-05,
"loss": 0.5603,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.8142768054002116,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.626,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.8634136322383541,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.6669,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.9854107457939656,
"learning_rate": 3.969342105253098e-05,
"loss": 0.5871,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 1.024115584956698,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.6002,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.8161708647158957,
"learning_rate": 3.967337762346568e-05,
"loss": 0.5726,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.6906966482250105,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.568,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.7471301563362426,
"learning_rate": 3.965270496444528e-05,
"loss": 0.6867,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.7730401889828464,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.5807,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 1.2443863798700365,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.5618,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.9241524439554057,
"learning_rate": 3.962051762160151e-05,
"loss": 0.6366,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.7543008544051113,
"learning_rate": 3.960947462140638e-05,
"loss": 0.5845,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 1.2285805855988228,
"learning_rate": 3.959827482437472e-05,
"loss": 0.6213,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.7910303884841902,
"learning_rate": 3.958691832005981e-05,
"loss": 0.5958,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.9100499059193136,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.625,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.8595605099467174,
"learning_rate": 3.956373555405774e-05,
"loss": 0.5822,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.6759365175093333,
"learning_rate": 3.955190947773937e-05,
"loss": 0.5462,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 1.5377319254005983,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.6351,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 1.33103965805549,
"learning_rate": 3.952778841127214e-05,
"loss": 0.6035,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.7941105657031724,
"learning_rate": 3.95154936139947e-05,
"loss": 0.5708,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.9539927683194035,
"learning_rate": 3.950304277135034e-05,
"loss": 0.614,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 1.3400844457991532,
"learning_rate": 3.94904359828957e-05,
"loss": 0.6307,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.7716314816352692,
"learning_rate": 3.94776733494343e-05,
"loss": 0.642,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.6905408731527112,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.6151,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 1.54928177590148,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.6055,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.8017062270830766,
"learning_rate": 3.943845140573213e-05,
"loss": 0.6249,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.6618636859068463,
"learning_rate": 3.942506642518952e-05,
"loss": 0.5754,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.6131080795441196,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.579,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 1.4161903517771508,
"learning_rate": 3.939783060543167e-05,
"loss": 0.5803,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 1.1342315930940667,
"learning_rate": 3.938397998399333e-05,
"loss": 0.5814,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.8179100776422281,
"learning_rate": 3.936997436876756e-05,
"loss": 0.5749,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.6059636822815604,
"learning_rate": 3.935581387174289e-05,
"loss": 0.5721,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.9430649495726574,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.5909,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.6847997692524418,
"learning_rate": 3.932702868644224e-05,
"loss": 0.6055,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.8255787389658483,
"learning_rate": 3.931240422833181e-05,
"loss": 0.6168,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.8141140702255021,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.6021,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.5887793380304753,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.5374,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6269334859827742,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.5313,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.6331616487694336,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.5165,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 1.2975743405250237,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.668,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.9312479322617329,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.5835,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 1.1445942785562024,
"learning_rate": 3.92057159243286e-05,
"loss": 0.6123,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 1.1201126354981454,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.5765,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.9585244467481518,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.6065,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.8049453465139287,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.5488,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.590280692577845,
"learning_rate": 3.914136997982874e-05,
"loss": 0.5606,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.7101046704569106,
"learning_rate": 3.912490053436079e-05,
"loss": 0.5898,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.64852268924167,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.5544,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.6574002805926922,
"learning_rate": 3.909150300975178e-05,
"loss": 0.6058,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.8035760709584319,
"learning_rate": 3.907457519765642e-05,
"loss": 0.5944,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.6651480715843158,
"learning_rate": 3.905749486576607e-05,
"loss": 0.5628,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.6095254121545584,
"learning_rate": 3.90402621506546e-05,
"loss": 0.5477,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.8355962153609019,
"learning_rate": 3.902287719011435e-05,
"loss": 0.5927,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.6785183423820359,
"learning_rate": 3.900534012315499e-05,
"loss": 0.5596,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.8613013385187801,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.5436,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 1.532817333631006,
"learning_rate": 3.896981023209777e-05,
"loss": 0.6,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 1.1387786540523168,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.5767,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 1.3367020848938775,
"learning_rate": 3.893367361386491e-05,
"loss": 0.609,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 1.037080069190835,
"learning_rate": 3.891537814248417e-05,
"loss": 0.6283,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.5928927721144301,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5606,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.7837811130013674,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.5728,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.9755794870054264,
"learning_rate": 3.885958483839049e-05,
"loss": 0.5858,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.7282136857495322,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.5728,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.7531014142710862,
"learning_rate": 3.882163505079171e-05,
"loss": 0.6212,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 2.422039239207708,
"learning_rate": 3.880243433489132e-05,
"loss": 0.6183,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.7678357992349311,
"learning_rate": 3.878308327522702e-05,
"loss": 0.4272,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.9517458612372994,
"learning_rate": 3.876358202652936e-05,
"loss": 0.4457,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.7042111709946467,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.4413,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.8517098223138486,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.4443,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.5793707766949321,
"learning_rate": 3.870417871154801e-05,
"loss": 0.4183,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.6163547637841102,
"learning_rate": 3.868407827802192e-05,
"loss": 0.418,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.6016614064777838,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.4535,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.7271255007122116,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.4554,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.6845976909879619,
"learning_rate": 3.862288124195319e-05,
"loss": 0.4448,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.5916539279866084,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.424,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.7453101878562176,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.4509,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.8170556642149521,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.4601,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 1.4034690011044606,
"learning_rate": 3.853920126717703e-05,
"loss": 0.4411,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.7901009956239896,
"learning_rate": 3.851791025492574e-05,
"loss": 0.4527,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.5349357403754708,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.4213,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.6312542029736854,
"learning_rate": 3.847488419569866e-05,
"loss": 0.4395,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.6019927203718833,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.4276,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.6507580699739721,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.4498,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.6776924623611542,
"learning_rate": 3.840923760917199e-05,
"loss": 0.4224,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.6295983748714553,
"learning_rate": 3.838706077964473e-05,
"loss": 0.4172,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.6715114892654235,
"learning_rate": 3.836473692766944e-05,
"loss": 0.4321,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.6096275174599162,
"learning_rate": 3.834226623174705e-05,
"loss": 0.4102,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.7976080309030267,
"learning_rate": 3.831964887155264e-05,
"loss": 0.4621,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.9810447584931207,
"learning_rate": 3.829688502793403e-05,
"loss": 0.4352,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.6659615470383815,
"learning_rate": 3.827397488291029e-05,
"loss": 0.511,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.5371270095336858,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.4073,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.5441004893233986,
"learning_rate": 3.822771642257145e-05,
"loss": 0.4247,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 4.405731249973416,
"learning_rate": 3.820436847713776e-05,
"loss": 0.4208,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.8364360653178303,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.4167,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.5762388790653856,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.4277,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.60433798039635,
"learning_rate": 3.813345202354145e-05,
"loss": 0.4397,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.9975496996973136,
"learning_rate": 3.810952296329563e-05,
"loss": 0.4288,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.8315438983844141,
"learning_rate": 3.808544909978676e-05,
"loss": 0.3997,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.7502697215406875,
"learning_rate": 3.806123062550882e-05,
"loss": 0.462,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.7934376458553692,
"learning_rate": 3.803686773411208e-05,
"loss": 0.4724,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.7923105566706807,
"learning_rate": 3.801236062040158e-05,
"loss": 0.4401,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.6862711101448248,
"learning_rate": 3.798770948033553e-05,
"loss": 0.4185,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.5784145550637363,
"learning_rate": 3.796291451102381e-05,
"loss": 0.4055,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.6037623202781082,
"learning_rate": 3.793797591072631e-05,
"loss": 0.4297,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.7044860219012635,
"learning_rate": 3.791289387885144e-05,
"loss": 0.4335,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.6081026746969189,
"learning_rate": 3.788766861595443e-05,
"loss": 0.4339,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.7860036046935399,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.455,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.7964287580672423,
"learning_rate": 3.783678920503983e-05,
"loss": 0.4314,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.6870927426022054,
"learning_rate": 3.781113546385265e-05,
"loss": 0.42,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.7837570924541487,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.4413,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.9609566280057812,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.4416,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.7992675936335638,
"learning_rate": 3.77333205623025e-05,
"loss": 0.4379,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.5205175189749793,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.4183,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.6519783892256845,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.4207,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.7163900938992708,
"learning_rate": 3.765422951112704e-05,
"loss": 0.4542,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.6144855625376888,
"learning_rate": 3.762758321970216e-05,
"loss": 0.4236,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.5443708346456452,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.4081,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.6824740847365902,
"learning_rate": 3.757386800198538e-05,
"loss": 0.4337,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.628400341998299,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.4169,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.6129853973650182,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.4083,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.5372197443001991,
"learning_rate": 3.749224181796305e-05,
"loss": 0.3965,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.6575500709146053,
"learning_rate": 3.746475306375545e-05,
"loss": 0.436,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.5247529252838592,
"learning_rate": 3.743712466184774e-05,
"loss": 0.4202,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.7492478634745412,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.4285,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.6043149511623532,
"learning_rate": 3.73814497997107e-05,
"loss": 0.4173,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.7587933478844103,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.425,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.49505461717332283,
"learning_rate": 3.732521901224724e-05,
"loss": 0.4242,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.5984786318672968,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.4235,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.5763664892855912,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.4472,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 1.0071613348029302,
"learning_rate": 3.723983441007882e-05,
"loss": 0.4181,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.6230316570338719,
"learning_rate": 3.721109687296781e-05,
"loss": 0.4449,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.6376191206956247,
"learning_rate": 3.718222171638484e-05,
"loss": 0.4209,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.6227077737529094,
"learning_rate": 3.715320917121488e-05,
"loss": 0.4256,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.5776467005040263,
"learning_rate": 3.712405946944149e-05,
"loss": 0.4324,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.715162288380283,
"learning_rate": 3.709477284414488e-05,
"loss": 0.455,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.9254118886684941,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.4513,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 1.8517899807007157,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.5213,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.6376850060453758,
"learning_rate": 3.70060937743295e-05,
"loss": 0.412,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.6652425359524794,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.4561,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.803432792934158,
"learning_rate": 3.694629409915555e-05,
"loss": 0.4687,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.5218712868181832,
"learning_rate": 3.691619088858401e-05,
"loss": 0.4278,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.6706544107801329,
"learning_rate": 3.68859524166009e-05,
"loss": 0.4205,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.5510831075884195,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.4172,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.5524411977800691,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.4361,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.5310214055107949,
"learning_rate": 3.67944278554384e-05,
"loss": 0.406,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 1.2869269307694025,
"learning_rate": 3.676365076645557e-05,
"loss": 0.4202,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 1.0530073308573875,
"learning_rate": 3.673273963576791e-05,
"loss": 0.4388,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.6148466591054766,
"learning_rate": 3.670169471054001e-05,
"loss": 0.4225,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.5180254041364384,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.4229,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.7507052522723787,
"learning_rate": 3.663920447046888e-05,
"loss": 0.4416,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.5892223492868366,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.4254,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.5608892195290113,
"learning_rate": 3.657618204491937e-05,
"loss": 0.4228,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.9691732078060527,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.4287,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.6788711024033722,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.4286,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.6866298585314433,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.4682,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.5859716374115704,
"learning_rate": 3.644854871713611e-05,
"loss": 0.416,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.4646573046314648,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.3855,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.7567622373599121,
"learning_rate": 3.638394189710493e-05,
"loss": 0.4522,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.7331640674866137,
"learning_rate": 3.635144184933412e-05,
"loss": 0.4303,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5407174133508206,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.4481,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.4386080849115827,
"learning_rate": 3.628604977762752e-05,
"loss": 0.4611,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.8082944701050998,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.4134,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.5268481069687946,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.4352,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.6145775622595361,
"learning_rate": 3.618698565901727e-05,
"loss": 0.4343,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.6071391585198108,
"learning_rate": 3.615370507164654e-05,
"loss": 0.4711,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.7052943344906565,
"learning_rate": 3.61202953196805e-05,
"loss": 0.4325,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 1.3384212696165634,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.4486,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.5720702272306822,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.4189,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.5435181107217375,
"learning_rate": 3.601929375279662e-05,
"loss": 0.4307,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.7317349074676738,
"learning_rate": 3.598537002418012e-05,
"loss": 0.437,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.7933595180482337,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.4398,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.6061597145262321,
"learning_rate": 3.591713938344416e-05,
"loss": 0.4283,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.5910225596605109,
"learning_rate": 3.588283301689507e-05,
"loss": 0.4095,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.5489836148695966,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.3659,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.5421297268050872,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.3922,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.6766837832511393,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.4108,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.643715935216453,
"learning_rate": 3.574434032194994e-05,
"loss": 0.4511,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.5558977149670012,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.4502,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.5405367162068344,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.4069,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.7127948766984166,
"learning_rate": 3.563914797637216e-05,
"loss": 0.4979,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.5413272736897238,
"learning_rate": 3.560383338592649e-05,
"loss": 0.4176,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.6547130638522803,
"learning_rate": 3.556839402764487e-05,
"loss": 0.4482,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.7280530237356369,
"learning_rate": 3.553283018489946e-05,
"loss": 0.4321,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.6241177021191937,
"learning_rate": 3.54971421420578e-05,
"loss": 0.4328,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.5109183502025175,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.4032,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.8232458972163672,
"learning_rate": 3.54253945985191e-05,
"loss": 0.3971,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.5686405328834092,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.4135,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.6847845112381481,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.4511,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.7359329757283671,
"learning_rate": 3.531684894865866e-05,
"loss": 0.4391,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.6316733197445737,
"learning_rate": 3.528042173241128e-05,
"loss": 0.4537,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.5455522307544448,
"learning_rate": 3.524387233431892e-05,
"loss": 0.4478,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.5884017893830273,
"learning_rate": 3.520720104662958e-05,
"loss": 0.4402,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.5301676893283479,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.4331,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.5898586203918397,
"learning_rate": 3.513349397632281e-05,
"loss": 0.4353,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.6630662587566013,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.4141,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.5719861171191336,
"learning_rate": 3.505930287892533e-05,
"loss": 0.4197,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.6157590759853384,
"learning_rate": 3.502202656100099e-05,
"loss": 0.4344,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 1.878688437785235,
"learning_rate": 3.498463012735256e-05,
"loss": 0.4636,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.5471388971182282,
"learning_rate": 3.494711387700094e-05,
"loss": 0.4247,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.5905956239527337,
"learning_rate": 3.490947810992505e-05,
"loss": 0.4355,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 1.3223961200029826,
"learning_rate": 3.48717231270595e-05,
"loss": 0.4084,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.6751041705771126,
"learning_rate": 3.483384923029211e-05,
"loss": 0.4322,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.6020684344178646,
"learning_rate": 3.479585672246156e-05,
"loss": 0.4389,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.7274204382916282,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.4285,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.517954522929833,
"learning_rate": 3.471951708970527e-05,
"loss": 0.4019,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.6442480580145598,
"learning_rate": 3.468117057518921e-05,
"loss": 0.413,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.9053435436133861,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.4381,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.5954800195334677,
"learning_rate": 3.460412568296731e-05,
"loss": 0.4638,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.6985540582989047,
"learning_rate": 3.456542792131039e-05,
"loss": 0.4265,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.6125233067127686,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.424,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.648632629656704,
"learning_rate": 3.448768331403348e-05,
"loss": 0.4158,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.6327887263790825,
"learning_rate": 3.44486370900573e-05,
"loss": 0.4388,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.5395939829375749,
"learning_rate": 3.440947533516396e-05,
"loss": 0.4171,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.5722923369932603,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.425,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 1.0189231076638545,
"learning_rate": 3.43308064860925e-05,
"loss": 0.4662,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.49347659940509275,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.4295,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.6008814102429345,
"learning_rate": 3.425167928295015e-05,
"loss": 0.4581,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.6014456668515523,
"learning_rate": 3.421194458890411e-05,
"loss": 0.4894,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.5160595058382019,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.4035,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.6050961244873906,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.4426,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.5415138299899115,
"learning_rate": 3.409205995219561e-05,
"loss": 0.4208,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.5382346312145968,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.4247,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.6933021092533758,
"learning_rate": 3.40115729298209e-05,
"loss": 0.4248,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.6220269906597455,
"learning_rate": 3.397116120327025e-05,
"loss": 0.4458,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.6327089686586256,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.4364,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.5017653562842671,
"learning_rate": 3.389000293509524e-05,
"loss": 0.4281,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.6010778960603278,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.4584,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.5725285435252481,
"learning_rate": 3.380840041143494e-05,
"loss": 0.4133,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.632758518607424,
"learning_rate": 3.376743336885792e-05,
"loss": 0.4415,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.5331618046231585,
"learning_rate": 3.372635624225055e-05,
"loss": 0.4194,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.5259816154084318,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.4184,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.6612223315644383,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.4235,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.5368585289569562,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.4466,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.6678036989495196,
"learning_rate": 3.356095347769799e-05,
"loss": 0.5431,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.604194801663076,
"learning_rate": 3.351933087522552e-05,
"loss": 0.4246,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.5815165961548651,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.4953,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.6069310072025264,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.4214,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.5988228666351716,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.4452,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.8155327339283576,
"learning_rate": 3.335176280432308e-05,
"loss": 0.4254,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.6034407908670272,
"learning_rate": 3.330960304618295e-05,
"loss": 0.4279,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.7828950918841847,
"learning_rate": 3.326733686481153e-05,
"loss": 0.4129,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.5825106529087656,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.4395,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.5410500395755904,
"learning_rate": 3.318248658506012e-05,
"loss": 0.4213,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.5654282489788992,
"learning_rate": 3.313990316514069e-05,
"loss": 0.4416,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.5318128789774709,
"learning_rate": 3.309721467890571e-05,
"loss": 0.4072,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.4590239123048617,
"learning_rate": 3.305442146769115e-05,
"loss": 0.4575,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.5741229946512377,
"learning_rate": 3.30115238736704e-05,
"loss": 0.3912,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.5153483287078461,
"learning_rate": 3.296852223985148e-05,
"loss": 0.4501,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 1.1350040990533432,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.4622,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.5537764694851404,
"learning_rate": 3.288220822900796e-05,
"loss": 0.4212,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.47817220808057553,
"learning_rate": 3.283889654214788e-05,
"loss": 0.4089,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.5719484011100011,
"learning_rate": 3.279548219581317e-05,
"loss": 0.4169,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.6547683876582869,
"learning_rate": 3.275196553714379e-05,
"loss": 0.4453,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.5852429155537745,
"learning_rate": 3.270834691409778e-05,
"loss": 0.4058,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.547394498836765,
"learning_rate": 3.266462667544849e-05,
"loss": 0.4267,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.5107219616736967,
"learning_rate": 3.262080517078178e-05,
"loss": 0.4477,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.5579045632116383,
"learning_rate": 3.257688275049323e-05,
"loss": 0.4188,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.669259164572059,
"learning_rate": 3.253285976578535e-05,
"loss": 0.4392,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.4928580867575044,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.4345,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.6665384918426878,
"learning_rate": 3.244451351193935e-05,
"loss": 0.4397,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 1.055953737663882,
"learning_rate": 3.240019094921551e-05,
"loss": 0.4447,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.7188816449368186,
"learning_rate": 3.235576923489527e-05,
"loss": 0.4194,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.5361507477965242,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.4179,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.692129543338782,
"learning_rate": 3.226662977303494e-05,
"loss": 0.4263,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.5878832433093457,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.432,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.47244158869586983,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.4226,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.8497095701664864,
"learning_rate": 3.21321858487569e-05,
"loss": 0.4577,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.5796222715818109,
"learning_rate": 3.208717671149931e-05,
"loss": 0.4053,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.5680860932845522,
"learning_rate": 3.204207092549882e-05,
"loss": 0.4056,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.6072539050018658,
"learning_rate": 3.19968688514201e-05,
"loss": 0.4144,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.61582120466893,
"learning_rate": 3.195157085069774e-05,
"loss": 0.4206,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.642000365497325,
"learning_rate": 3.190617728553332e-05,
"loss": 0.4569,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.9029567148495458,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.4617,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.5423148322076903,
"learning_rate": 3.181510491450254e-05,
"loss": 0.4227,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.6069614254867896,
"learning_rate": 3.176942683684842e-05,
"loss": 0.4445,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.5796363572863613,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.4255,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.7072927604471487,
"learning_rate": 3.167778872346327e-05,
"loss": 0.3954,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.7624442534972488,
"learning_rate": 3.163182942046815e-05,
"loss": 0.418,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.4966293401393492,
"learning_rate": 3.158577710967494e-05,
"loss": 0.4334,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.561259389533278,
"learning_rate": 3.153963215931669e-05,
"loss": 0.3932,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.5097240389489114,
"learning_rate": 3.14933949383672e-05,
"loss": 0.444,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.5498552269980426,
"learning_rate": 3.144706581653805e-05,
"loss": 0.3849,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.6194828200698852,
"learning_rate": 3.140064516427566e-05,
"loss": 0.4215,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.6629220545696763,
"learning_rate": 3.135413335275833e-05,
"loss": 0.4225,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.6468530111677213,
"learning_rate": 3.130753075389327e-05,
"loss": 0.4344,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.6353982193586561,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.4277,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.5184423748312045,
"learning_rate": 3.121405468537547e-05,
"loss": 0.4423,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.5039500201576953,
"learning_rate": 3.116718196315488e-05,
"loss": 0.4037,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.739915742049295,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.42,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.922395692088925,
"learning_rate": 3.107316901675252e-05,
"loss": 0.4338,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.7526942661399864,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.4302,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.6038399171708805,
"learning_rate": 3.097880190800057e-05,
"loss": 0.417,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.6101595336697264,
"learning_rate": 3.093148648549788e-05,
"loss": 0.4596,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.5823276345644798,
"learning_rate": 3.088408365512055e-05,
"loss": 0.4307,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.5646393321768425,
"learning_rate": 3.083659379590034e-05,
"loss": 0.4411,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.5799043699813998,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.4288,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.5549474093359122,
"learning_rate": 3.074135451053475e-05,
"loss": 0.4083,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.6317123965071052,
"learning_rate": 3.069360584592015e-05,
"loss": 0.4264,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.4693608513407897,
"learning_rate": 3.064577167551816e-05,
"loss": 0.4075,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.5241414849013282,
"learning_rate": 3.059785238180953e-05,
"loss": 0.4306,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.6340518598940883,
"learning_rate": 3.054984834795566e-05,
"loss": 0.4233,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.5535833266953609,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.4246,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.5874094412232571,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.4325,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.4841209574619167,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.4489,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.5225066204361923,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.4425,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.47549109183908267,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.4143,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.7627497534045743,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.4328,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 1.4407795209791165,
"learning_rate": 3.021147971398632e-05,
"loss": 0.4535,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.6498965813242454,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.4276,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.5203055542653752,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.4193,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.5980019496882648,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.4131,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.6657079507827348,
"learning_rate": 3.001632138223332e-05,
"loss": 0.4271,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.49008940464494816,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.3827,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.45608311905042914,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.2796,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.5692851663484676,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.3361,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.7187358364236799,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.3529,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.6025197147291143,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.2944,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.5272713574403348,
"learning_rate": 2.972118904869376e-05,
"loss": 0.3107,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.5573415835768539,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.3247,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.4670432999106424,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.3045,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.49400658949867465,
"learning_rate": 2.95725682050846e-05,
"loss": 0.3099,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.46261495913588657,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.329,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.5776265273171929,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.3289,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.5217041200839099,
"learning_rate": 2.942325848714505e-05,
"loss": 0.3226,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.4452561400377221,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.3036,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.5656907886603654,
"learning_rate": 2.932334127129686e-05,
"loss": 0.3233,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.4764205586002139,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.2987,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.44914222072224125,
"learning_rate": 2.922312585929131e-05,
"loss": 0.3051,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.7082260792491232,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.3185,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.4726421507119465,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.3092,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.45101135993348146,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.3114,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.4245976877532271,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.3132,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.42743431286518613,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.3012,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.4483165706891017,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.3187,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.5744893854098472,
"learning_rate": 2.887006998351372e-05,
"loss": 0.3215,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.4189430103469416,
"learning_rate": 2.881934649605535e-05,
"loss": 0.3119,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.5085278362943932,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.3089,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.5345752507541971,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.3069,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.4532000968253981,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.2999,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.45274268023731357,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.3331,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.4550080663868162,
"learning_rate": 2.856467940121025e-05,
"loss": 0.3136,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.5378172684975311,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.3061,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.4294037495940703,
"learning_rate": 2.846233032790149e-05,
"loss": 0.3053,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.5393426583255807,
"learning_rate": 2.84110540896642e-05,
"loss": 0.3179,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.5194818169805515,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.3114,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.5402786311119351,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.2774,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.7385848069274665,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.3291,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.4558436616815717,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.3268,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.6304379522529843,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.3031,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.41643336456788305,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.3132,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.4774522865776435,
"learning_rate": 2.805026032173773e-05,
"loss": 0.2939,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.45475253921979814,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.3018,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.5074866281847359,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.3184,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.45501060799827103,
"learning_rate": 2.7894660620408e-05,
"loss": 0.2935,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.7413737713278893,
"learning_rate": 2.784266724883173e-05,
"loss": 0.3009,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.45799201162386266,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.3063,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.5056291360985579,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.3136,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.5188137515806808,
"learning_rate": 2.768631254100716e-05,
"loss": 0.3226,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.4407394798496406,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.3194,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.7619819265510853,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.3298,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.4638633662652221,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.308,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.5625726232153286,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.3097,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.7415156736621548,
"learning_rate": 2.742449775134085e-05,
"loss": 0.3166,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.47197120039467005,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.2965,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.40866871009760153,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.3015,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.48539633855304326,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.3319,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.4584238097587312,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.3198,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.505956839767554,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.2973,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.4384442717169746,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.3171,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.4470033751713966,
"learning_rate": 2.705547548999996e-05,
"loss": 0.3365,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.4728675709328957,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.3167,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.8090572898444159,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.3427,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.5145518142066062,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.3091,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.4607195752310096,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.297,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.5042893200059623,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.3155,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.544392847033265,
"learning_rate": 2.67369650552155e-05,
"loss": 0.317,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.43014992687680415,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.294,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.6184145034117693,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.3262,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.42372750458059943,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.3156,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.5135902753902686,
"learning_rate": 2.652354173593614e-05,
"loss": 0.3265,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.4518918476831004,
"learning_rate": 2.647005443331317e-05,
"loss": 0.3237,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.44097854649346047,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.3641,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.4470089627140868,
"learning_rate": 2.636292505301542e-05,
"loss": 0.3017,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.8700017414215735,
"learning_rate": 2.630928383194431e-05,
"loss": 0.3115,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.48124024263605103,
"learning_rate": 2.625559216200781e-05,
"loss": 0.2934,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.43493514713578857,
"learning_rate": 2.620185047252308e-05,
"loss": 0.3003,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.46626441448846545,
"learning_rate": 2.614805919320723e-05,
"loss": 0.3038,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.5343581221208537,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.3355,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.41958854127444095,
"learning_rate": 2.60403295859298e-05,
"loss": 0.3729,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.4347803124865119,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.2999,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.4070353216611326,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.322,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.4348972403186984,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.3047,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.5337129899631197,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.3121,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.4099859308141371,
"learning_rate": 2.577016790136453e-05,
"loss": 0.3025,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.42541669079850025,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.3209,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.8006784069685919,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.3212,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.5315471183450885,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.3041,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.6714019475046087,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.3095,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.4261553078197934,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.3239,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.46200711794569904,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.3281,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.49823287934771454,
"learning_rate": 2.539001591449889e-05,
"loss": 0.3172,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.3985044736032715,
"learning_rate": 2.53355326151989e-05,
"loss": 0.2977,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.40720644373302683,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.3205,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.46226479292024863,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.3363,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.41620271053480223,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.2959,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.539060841363732,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.3111,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.4372525731760258,
"learning_rate": 2.506248490185627e-05,
"loss": 0.2779,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.4395026736983067,
"learning_rate": 2.500775217241923e-05,
"loss": 0.2968,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.6446621866278701,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.3103,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.4238003646002282,
"learning_rate": 2.4898167025948e-05,
"loss": 0.2992,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.5209701015053085,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.3256,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.5679405033134387,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.3056,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.7875896098582856,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.331,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.4148430841964563,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.3156,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 1.1181712837967281,
"learning_rate": 2.46235264422994e-05,
"loss": 0.3456,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.5210921388326624,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.3185,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.4225072433762623,
"learning_rate": 2.451340834511592e-05,
"loss": 0.2924,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.4306909245558404,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.3184,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.5122841301501496,
"learning_rate": 2.440314589184958e-05,
"loss": 0.3129,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.4573164508372005,
"learning_rate": 2.434796163357541e-05,
"loss": 0.3169,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.43140776628617916,
"learning_rate": 2.429274260911588e-05,
"loss": 0.3527,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.46710017407548654,
"learning_rate": 2.423748926000084e-05,
"loss": 0.322,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.4420438281884562,
"learning_rate": 2.418220202803459e-05,
"loss": 0.3122,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.44584804523820015,
"learning_rate": 2.412688135529237e-05,
"loss": 0.3199,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.7271505199281973,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.3186,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.4497489158859775,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.2947,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.397912611925959,
"learning_rate": 2.396072311715187e-05,
"loss": 0.334,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.3855393618675577,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.3249,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.45973804554348946,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.3014,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.42379242382447235,
"learning_rate": 2.379427985200223e-05,
"loss": 0.2855,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.4162237574218144,
"learning_rate": 2.373873749394972e-05,
"loss": 0.3212,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.4066895552794778,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.303,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.7383035305521636,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.3356,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.6860851799528164,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.3231,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.4572965931810166,
"learning_rate": 2.351627355800233e-05,
"loss": 0.304,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.45987865593376703,
"learning_rate": 2.346058617160256e-05,
"loss": 0.323,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.43726963957480314,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.3034,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.4668717977186601,
"learning_rate": 2.334912883203469e-05,
"loss": 0.3241,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.4337402440078607,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.2957,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.48569347079934516,
"learning_rate": 2.323756437450136e-05,
"loss": 0.3281,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.4051265140781894,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.3062,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.3672248457476319,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.3212,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.41790008509217036,
"learning_rate": 2.307002464849052e-05,
"loss": 0.3121,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.42418115942342377,
"learning_rate": 2.30141283818846e-05,
"loss": 0.3093,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.5050354049177646,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.2915,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.4290144104985669,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.3019,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.416058578026031,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.3028,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 1.092712331559632,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.3737,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.4521988057893872,
"learning_rate": 2.273429448107257e-05,
"loss": 0.3133,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.45575833779231845,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.3016,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.40987596323590897,
"learning_rate": 2.262220474347102e-05,
"loss": 0.2966,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.405776228330502,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.3119,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.409197264051615,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.3069,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.40767945745722883,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.3001,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.4276119429120177,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.3192,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.37813587794180087,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.3093,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.37953561528200597,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.3152,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.5267167103704734,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.3149,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.45160608298448646,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.376,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.43339390160515934,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.2945,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.42856315092771036,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.285,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.3907937733099373,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.3295,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.44284802180317945,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.312,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.45250790623261605,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.2797,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.4443201769539544,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.284,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.45218741829815606,
"learning_rate": 2.17791064266051e-05,
"loss": 0.306,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 1.9092865081842738,
"learning_rate": 2.172276926861741e-05,
"loss": 0.373,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.46091007658149796,
"learning_rate": 2.166641833541255e-05,
"loss": 0.3274,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.4541256641163881,
"learning_rate": 2.161005407757109e-05,
"loss": 0.3155,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.5039514881461518,
"learning_rate": 2.155367694578013e-05,
"loss": 0.3253,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.44354440158026537,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.3012,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.6570968351011919,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.3322,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.35353268987841413,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.3238,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.5165110710288772,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.3224,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.4018810005190142,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.3379,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.40411236822059693,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.312,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.39957060709233927,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.3436,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.4660591036027492,
"learning_rate": 2.110225054402733e-05,
"loss": 0.3363,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.36935677977172165,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.3325,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.441813141075504,
"learning_rate": 2.09892967807724e-05,
"loss": 0.3078,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.3742890737407314,
"learning_rate": 2.093280780774835e-05,
"loss": 0.3331,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.3933321801568322,
"learning_rate": 2.087631137601793e-05,
"loss": 0.3232,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.5421931598796745,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.3077,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.7448632351388639,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.3166,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.48956091742702706,
"learning_rate": 2.070678184630458e-05,
"loss": 0.3253,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.46032422702652337,
"learning_rate": 2.06502600977305e-05,
"loss": 0.3014,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.43019353759060647,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.3278,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.440253261364569,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.3247,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.6124660301549056,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.2934,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 1.0768069034396848,
"learning_rate": 2.042412562887594e-05,
"loss": 0.3439,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.43181288292493186,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.3062,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.48139109904664024,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.3261,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.41112487115637036,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.3026,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.3805256347204249,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.2828,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.43064320310356935,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.3049,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.4380188661604868,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.3441,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.4461294495573458,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.3078,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.4308789479869072,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.3201,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.47391756456882195,
"learning_rate": 1.99151687695709e-05,
"loss": 0.3127,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.4590556312306975,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.3343,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.4288946010736675,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.3189,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.4403664524066101,
"learning_rate": 1.974551241345447e-05,
"loss": 0.3147,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.4875275918627585,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.3042,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.4608221719386246,
"learning_rate": 1.963241759676897e-05,
"loss": 0.3019,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.3928904121491965,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.2912,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.40596811689522116,
"learning_rate": 1.951933453677638e-05,
"loss": 0.3088,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.8211431351926621,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.3198,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.4004938745467522,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.2974,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.40934126961399697,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.3015,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.6786373519832286,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.3083,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.4286031363425062,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.2904,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.4136935040316442,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.305,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.35443952509017207,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.3243,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.5172393720208145,
"learning_rate": 1.906719219225165e-05,
"loss": 0.2919,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.43880500018685864,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.3086,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.41019882316334966,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.3465,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.37396672768441097,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.3251,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.4791999420002742,
"learning_rate": 1.884128556891707e-05,
"loss": 0.3036,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.379409947614861,
"learning_rate": 1.878483094691119e-05,
"loss": 0.308,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.4799961163912734,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.2903,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.4572943773424488,
"learning_rate": 1.867195130360947e-05,
"loss": 0.3099,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.58531284292641,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.3096,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.5566028406510289,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.311,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.45626639537110913,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.3146,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.588039093774563,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.3083,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 1.074591408854621,
"learning_rate": 1.838994592242891e-05,
"loss": 0.3041,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.5282634786306994,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.3565,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.39060985466318227,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.3299,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.45108843035560114,
"learning_rate": 1.82208935733949e-05,
"loss": 0.3376,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.42864218510468655,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.2914,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.4719263018638575,
"learning_rate": 1.810826238483892e-05,
"loss": 0.3029,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.34699786519362186,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.3072,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.4723073883381442,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.3085,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.5401158588766362,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.3092,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.37515751242440804,
"learning_rate": 1.788318512320838e-05,
"loss": 0.3054,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.4061286057256452,
"learning_rate": 1.782695699829051e-05,
"loss": 0.2989,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.4487689621333657,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.3099,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.4056624476324874,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.3096,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.4137691996913999,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.3083,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.5865620782487434,
"learning_rate": 1.76022227484387e-05,
"loss": 0.3228,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.38039561995648763,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.3153,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.5060267081863444,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.3125,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.3909128993619498,
"learning_rate": 1.743387180005809e-05,
"loss": 0.3176,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.4047861753322662,
"learning_rate": 1.737779525652899e-05,
"loss": 0.2965,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.47196231723930676,
"learning_rate": 1.732173968007914e-05,
"loss": 0.327,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.4959990236618709,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.3365,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.4217200246704095,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.2921,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.4781495745468262,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.2909,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.4483764421382982,
"learning_rate": 1.709773600686015e-05,
"loss": 0.3103,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.3910744358783246,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.3402,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.41757988682786984,
"learning_rate": 1.69858716181154e-05,
"loss": 0.3105,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.5419206303938837,
"learning_rate": 1.692997535150948e-05,
"loss": 0.3146,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.5545513750688471,
"learning_rate": 1.687410363273884e-05,
"loss": 0.2939,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.5484440999027419,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.3242,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.362491257663092,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.3273,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.42883329500811873,
"learning_rate": 1.67066402299235e-05,
"loss": 0.3068,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.38242725589114873,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.3078,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.4701106552944446,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.3227,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.8476405047255288,
"learning_rate": 1.653941382839745e-05,
"loss": 0.3225,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.4165329565099254,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.2361,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.35423393879654574,
"learning_rate": 1.642806717162757e-05,
"loss": 0.2318,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.33745352271167633,
"learning_rate": 1.637243646233718e-05,
"loss": 0.2431,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.4691593127255064,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.2768,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.39664020035761643,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.2168,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.3586948690881302,
"learning_rate": 1.620572014799777e-05,
"loss": 0.2431,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.4030614214985918,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.2564,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.7231859076327536,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.2824,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.3582714159082952,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.2489,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.34519295024928526,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.2594,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.44744008923507106,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.2706,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.37729328990853817,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.2396,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.40837665226908143,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.2188,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.3525685362979502,
"learning_rate": 1.576251073999917e-05,
"loss": 0.2216,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.3561369802320159,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.243,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.3476451701959481,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.2185,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.36245289527471747,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.2235,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.3447066080539056,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.2431,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.37119371545246516,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.2272,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.42288722018626773,
"learning_rate": 1.543151434154755e-05,
"loss": 0.24,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.3480376054553034,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.2488,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.36128843038107594,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.2074,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.34460193319842336,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.2315,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.30852164040219493,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.2748,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.35447931140859995,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.2358,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.3428615723091298,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.2438,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.34629098696671823,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.2388,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.47495397105199255,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.2471,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.32955335029106236,
"learning_rate": 1.493751509814374e-05,
"loss": 0.2842,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.3466406802506899,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.2472,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.30565132349977203,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.2565,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.3179322882554552,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.2178,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.37882691234362526,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.2228,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.36355238287853625,
"learning_rate": 1.466446738480111e-05,
"loss": 0.217,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.3024549340269593,
"learning_rate": 1.460998408550112e-05,
"loss": 0.2298,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.36318587817666936,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.2428,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.29234286896561323,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.2395,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.37423672894819665,
"learning_rate": 1.444679451900437e-05,
"loss": 0.2371,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.33327593843394676,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.2128,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.3257824660660848,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.2396,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.31644236922484803,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.2145,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.9593488823142466,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.2666,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.3267445742500735,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.2552,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.3362519088260353,
"learning_rate": 1.412162598317602e-05,
"loss": 0.2449,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.33332142522099056,
"learning_rate": 1.406759321421907e-05,
"loss": 0.2341,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.34719251773596366,
"learning_rate": 1.401360788062871e-05,
"loss": 0.2404,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.4327421075927299,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.2609,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.29609979128483516,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.2686,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.33717741726727957,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.248,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.33305391956280656,
"learning_rate": 1.379814952747693e-05,
"loss": 0.2556,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.30851417879007154,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.2652,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.33316066163416036,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.2272,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.2981457701630897,
"learning_rate": 1.363707494698459e-05,
"loss": 0.2408,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.2975708681083881,
"learning_rate": 1.358348460369265e-05,
"loss": 0.2137,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.5806920769103417,
"learning_rate": 1.352994556668684e-05,
"loss": 0.2513,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.3684479381603638,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.2627,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.31785764098176933,
"learning_rate": 1.342302312350678e-05,
"loss": 0.255,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.3070952624001418,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.2299,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.30318753890760564,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.1905,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.33919169915118946,
"learning_rate": 1.326303494478451e-05,
"loss": 0.2362,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.3188786520115454,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.2462,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.29845903071155694,
"learning_rate": 1.315664479122898e-05,
"loss": 0.2526,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.3281845893672677,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.2659,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.29726998654773956,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.2593,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.479063237793849,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.2524,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.32831217034654264,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.2373,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.3446711429419755,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.2567,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.3131318092960437,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.2666,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.30337333387129706,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.2374,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.29543737706391315,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.246,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.4081061514213887,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.2273,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.5189959448650874,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.2652,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.3445859353218213,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.3241,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.3040632241450436,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.2374,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.31702509302942816,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.2072,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.3030418095769276,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.2168,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.3134119359502183,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.2636,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.35515535217170613,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.2482,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.3737679414763495,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.2276,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.3442956008569726,
"learning_rate": 1.220938883250288e-05,
"loss": 0.2017,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.3094678047892662,
"learning_rate": 1.215733275116827e-05,
"loss": 0.2276,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.32005233642131986,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.2468,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.37877616777656886,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.2423,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.3477929845126638,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.2353,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.31390001796070544,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.2358,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.33162555061944987,
"learning_rate": 1.189800129803139e-05,
"loss": 0.2446,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.3371057662294509,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.2571,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.3272348670276923,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.2461,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.5165360348833223,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.2272,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.31078493068051766,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.2468,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.32794426677802363,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.2211,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.3739670352710367,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.2658,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.3448482322641542,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.2497,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.2978500891408339,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.2117,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.28165039672663955,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.2632,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.32690008682331395,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.2212,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.2766341976669344,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.2577,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.31512077879477673,
"learning_rate": 1.128231163049258e-05,
"loss": 0.2281,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.32274771736931346,
"learning_rate": 1.123144751066259e-05,
"loss": 0.2238,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.29711500480704717,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.2221,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.39067037426489426,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.2244,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.30886056509559595,
"learning_rate": 1.107927745387114e-05,
"loss": 0.2428,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.3137868186857913,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.228,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.3004951683718494,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.2205,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.2950807446684163,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.2276,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.29626262515075585,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.2445,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.3177500084133719,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.2428,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.29638946353721096,
"learning_rate": 1.077687414070869e-05,
"loss": 0.239,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.29041192928437365,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.2242,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.31585652643926804,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.2286,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.29847404498732477,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.2352,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.2995385804528724,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.2105,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.3357611521079923,
"learning_rate": 1.052689572753247e-05,
"loss": 0.2256,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.2922894715408855,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.2415,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.29776881521295656,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.2241,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.27723775539082024,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.2464,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.29559787848579905,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.2378,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.35030799433907023,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.262,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.3057837513361621,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.2454,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.26995231080247784,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.2319,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.3255067200054481,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.2397,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.3036415350890184,
"learning_rate": 1.008173988718005e-05,
"loss": 0.2384,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.3966588765324388,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.2687,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.3130328265830377,
"learning_rate": 9.983678617766677e-06,
"loss": 0.2411,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.422944006846728,
"learning_rate": 9.934767922553788e-06,
"loss": 0.2326,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.4250271890379803,
"learning_rate": 9.885937708668371e-06,
"loss": 0.2206,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.2960913328638998,
"learning_rate": 9.837188366555524e-06,
"loss": 0.2389,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.32595737742383735,
"learning_rate": 9.788520286013691e-06,
"loss": 0.2511,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.29061656344675757,
"learning_rate": 9.739933856191537e-06,
"loss": 0.2357,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.2998431196875516,
"learning_rate": 9.691429465584866e-06,
"loss": 0.2637,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.32155716123986744,
"learning_rate": 9.643007502033492e-06,
"loss": 0.2533,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.29013398861841794,
"learning_rate": 9.594668352718152e-06,
"loss": 0.2064,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.338266165869751,
"learning_rate": 9.546412404157383e-06,
"loss": 0.2101,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.29559588085002875,
"learning_rate": 9.498240042204471e-06,
"loss": 0.2765,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.3288454011909329,
"learning_rate": 9.450151652044339e-06,
"loss": 0.2709,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.27353795036137096,
"learning_rate": 9.402147618190472e-06,
"loss": 0.2518,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.36645209589890626,
"learning_rate": 9.354228324481841e-06,
"loss": 0.2396,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.28783422780802015,
"learning_rate": 9.306394154079852e-06,
"loss": 0.3226,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.2854776948760246,
"learning_rate": 9.258645489465254e-06,
"loss": 0.239,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.29702492138918324,
"learning_rate": 9.21098271243509e-06,
"loss": 0.2006,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.2919064468613261,
"learning_rate": 9.163406204099659e-06,
"loss": 0.2331,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.33550060637514934,
"learning_rate": 9.115916344879452e-06,
"loss": 0.2389,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.3102570770505113,
"learning_rate": 9.068513514502126e-06,
"loss": 0.2422,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.32651405515860804,
"learning_rate": 9.021198091999439e-06,
"loss": 0.2775,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.3436777953406692,
"learning_rate": 8.973970455704251e-06,
"loss": 0.2334,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.32352110679959295,
"learning_rate": 8.926830983247495e-06,
"loss": 0.2739,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.30098628280551026,
"learning_rate": 8.879780051555116e-06,
"loss": 0.2232,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.30271116331151976,
"learning_rate": 8.83281803684512e-06,
"loss": 0.2301,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.3258748151588678,
"learning_rate": 8.785945314624532e-06,
"loss": 0.2315,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.282512839525901,
"learning_rate": 8.739162259686385e-06,
"loss": 0.2617,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.30194783722117263,
"learning_rate": 8.692469246106734e-06,
"loss": 0.2462,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.4730708029120556,
"learning_rate": 8.645866647241677e-06,
"loss": 0.2422,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.403467655288548,
"learning_rate": 8.599354835724348e-06,
"loss": 0.2363,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.31650042267625633,
"learning_rate": 8.552934183461958e-06,
"loss": 0.2577,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.36338404922510736,
"learning_rate": 8.506605061632805e-06,
"loss": 0.2373,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.2876914633788526,
"learning_rate": 8.460367840683308e-06,
"loss": 0.3262,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.2957117938102586,
"learning_rate": 8.414222890325065e-06,
"loss": 0.2143,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3578045389314102,
"learning_rate": 8.368170579531855e-06,
"loss": 0.2378,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3184864380769367,
"learning_rate": 8.322211276536733e-06,
"loss": 0.2346,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.2706182480107469,
"learning_rate": 8.27634534882907e-06,
"loss": 0.2215,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.3320596930600255,
"learning_rate": 8.230573163151585e-06,
"loss": 0.2573,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.3103365235058734,
"learning_rate": 8.184895085497466e-06,
"loss": 0.2236,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.6620600695342523,
"learning_rate": 8.139311481107401e-06,
"loss": 0.2367,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.3180413519233412,
"learning_rate": 8.093822714466686e-06,
"loss": 0.2425,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.29319135281794767,
"learning_rate": 8.048429149302271e-06,
"loss": 0.2492,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.2915441361953479,
"learning_rate": 8.003131148579902e-06,
"loss": 0.2514,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.3217756130273758,
"learning_rate": 7.957929074501182e-06,
"loss": 0.2328,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.2793884948375591,
"learning_rate": 7.912823288500695e-06,
"loss": 0.2159,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.3227847979125366,
"learning_rate": 7.867814151243102e-06,
"loss": 0.2358,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.3498129609429656,
"learning_rate": 7.822902022620267e-06,
"loss": 0.2261,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.32207579598578445,
"learning_rate": 7.778087261748377e-06,
"loss": 0.2338,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.3087726116623244,
"learning_rate": 7.733370226965056e-06,
"loss": 0.2446,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.5166123674048811,
"learning_rate": 7.688751275826527e-06,
"loss": 0.2316,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.32411764228462975,
"learning_rate": 7.644230765104735e-06,
"loss": 0.2098,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.3165389765455102,
"learning_rate": 7.599809050784502e-06,
"loss": 0.2293,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.2816011643011378,
"learning_rate": 7.555486488060661e-06,
"loss": 0.2524,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.33361657503571907,
"learning_rate": 7.511263431335249e-06,
"loss": 0.2393,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.3319632251696948,
"learning_rate": 7.467140234214656e-06,
"loss": 0.2571,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.30161044241803175,
"learning_rate": 7.423117249506777e-06,
"loss": 0.2342,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.37000826998992686,
"learning_rate": 7.379194829218231e-06,
"loss": 0.2351,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.30319602514030375,
"learning_rate": 7.335373324551518e-06,
"loss": 0.2416,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.2677738429639176,
"learning_rate": 7.291653085902224e-06,
"loss": 0.2477,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.35806651492196057,
"learning_rate": 7.248034462856213e-06,
"loss": 0.2513,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.32585835307330796,
"learning_rate": 7.20451780418683e-06,
"loss": 0.2508,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.31123400282161184,
"learning_rate": 7.161103457852128e-06,
"loss": 0.2336,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.28066674561892585,
"learning_rate": 7.117791770992046e-06,
"loss": 0.2568,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.31790480089052525,
"learning_rate": 7.074583089925688e-06,
"loss": 0.2177,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.31986538916124013,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.2221,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.2928344364520107,
"learning_rate": 6.988476126329604e-06,
"loss": 0.2287,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.2781369146195698,
"learning_rate": 6.945578532308857e-06,
"loss": 0.2235,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.3245396099408783,
"learning_rate": 6.902785321094301e-06,
"loss": 0.265,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.31672986803139724,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.2089,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.2779429682691372,
"learning_rate": 6.817513414939892e-06,
"loss": 0.267,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.3169531714947628,
"learning_rate": 6.775035401831924e-06,
"loss": 0.2314,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.2998657926146226,
"learning_rate": 6.732663135188473e-06,
"loss": 0.2343,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.27500443703197724,
"learning_rate": 6.690396953817055e-06,
"loss": 0.2552,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.28476571020356295,
"learning_rate": 6.648237195676927e-06,
"loss": 0.2197,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.3090249183577863,
"learning_rate": 6.606184197876395e-06,
"loss": 0.2439,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.2934753536853391,
"learning_rate": 6.564238296670109e-06,
"loss": 0.2311,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.4578436849641745,
"learning_rate": 6.52239982745638e-06,
"loss": 0.2276,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.331212623631764,
"learning_rate": 6.480669124774484e-06,
"loss": 0.2771,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.3056071101730326,
"learning_rate": 6.439046522302013e-06,
"loss": 0.2327,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.2646811555082932,
"learning_rate": 6.397532352852191e-06,
"loss": 0.2309,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.29406628400192414,
"learning_rate": 6.356126948371202e-06,
"loss": 0.2429,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.3060480398573419,
"learning_rate": 6.314830639935559e-06,
"loss": 0.2516,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.46760124814508686,
"learning_rate": 6.273643757749448e-06,
"loss": 0.2435,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.27562911435928567,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.2589,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.29954345577735175,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.2273,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.311163144834058,
"learning_rate": 6.150742957589784e-06,
"loss": 0.2282,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.29084662887089424,
"learning_rate": 6.109997064904774e-06,
"loss": 0.2288,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.29351908284553446,
"learning_rate": 6.069362236313092e-06,
"loss": 0.2235,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.28166388439116313,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.2352,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.2843901544975278,
"learning_rate": 5.988427070179108e-06,
"loss": 0.2405,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.29164468015771444,
"learning_rate": 5.948127379792234e-06,
"loss": 0.2501,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.29101998001907026,
"learning_rate": 5.90794004780439e-06,
"loss": 0.2198,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.3020559217615913,
"learning_rate": 5.867865395552405e-06,
"loss": 0.2003,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.3300111369514149,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.2557,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.2715540136828848,
"learning_rate": 5.788055411095892e-06,
"loss": 0.2555,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.29437033365902743,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.2406,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.4331236549424227,
"learning_rate": 5.708699979051613e-06,
"loss": 0.2291,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.28825731082087,
"learning_rate": 5.66919351390751e-06,
"loss": 0.2249,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.4576182929139844,
"learning_rate": 5.629801637510188e-06,
"loss": 0.2416,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.31449821456354793,
"learning_rate": 5.590524664836043e-06,
"loss": 0.2357,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.38834265331158535,
"learning_rate": 5.551362909942708e-06,
"loss": 0.2115,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.3136570747596069,
"learning_rate": 5.51231668596653e-06,
"loss": 0.2102,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.2712481018181437,
"learning_rate": 5.47338630512007e-06,
"loss": 0.2591,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.30565857456296114,
"learning_rate": 5.434572078689617e-06,
"loss": 0.2401,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.29737958978914997,
"learning_rate": 5.395874317032694e-06,
"loss": 0.2633,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 1.5395036220736737,
"learning_rate": 5.357293329575573e-06,
"loss": 0.268,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.28232967601321146,
"learning_rate": 5.318829424810797e-06,
"loss": 0.2637,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.3060964724067651,
"learning_rate": 5.280482910294728e-06,
"loss": 0.2465,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.42267689414048687,
"learning_rate": 5.242254092645078e-06,
"loss": 0.2404,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.46591754076894504,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.2257,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.3381427819739966,
"learning_rate": 5.166150769707892e-06,
"loss": 0.2288,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.3077976401553716,
"learning_rate": 5.128276872940503e-06,
"loss": 0.2416,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.28325930451530523,
"learning_rate": 5.090521890074953e-06,
"loss": 0.2604,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.3003891239976868,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.2449,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.2788895438683762,
"learning_rate": 5.015369872647442e-06,
"loss": 0.2255,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.3111679594426875,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.2162,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.28600637563941944,
"learning_rate": 4.940697121074675e-06,
"loss": 0.2212,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.3077646348721957,
"learning_rate": 4.903541216934864e-06,
"loss": 0.2574,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.33375462005707845,
"learning_rate": 4.8665060236772e-06,
"loss": 0.2484,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.3651659291518854,
"learning_rate": 4.829591837434102e-06,
"loss": 0.2485,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.2748371653729288,
"learning_rate": 4.792798953370424e-06,
"loss": 0.2382,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.28323766217376145,
"learning_rate": 4.756127665681083e-06,
"loss": 0.2657,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.333664578045552,
"learning_rate": 4.719578267588725e-06,
"loss": 0.2484,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.2962054937568731,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.218,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.334907620018295,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.2196,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.2783428602995866,
"learning_rate": 4.610664328486485e-06,
"loss": 0.1903,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.24387652103406596,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1937,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.2303990480360872,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.226,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.23355830508392866,
"learning_rate": 4.502857857942204e-06,
"loss": 0.2057,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.28770021696757514,
"learning_rate": 4.467169815100549e-06,
"loss": 0.2057,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.31616008731518297,
"learning_rate": 4.431605972355137e-06,
"loss": 0.23,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.26885779176380786,
"learning_rate": 4.396166614073516e-06,
"loss": 0.2035,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.25070236584593064,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1831,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.24288080938908835,
"learning_rate": 4.325662483392656e-06,
"loss": 0.2256,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.2723291615490844,
"learning_rate": 4.29059827474257e-06,
"loss": 0.2314,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.7369798443550938,
"learning_rate": 4.255659678050063e-06,
"loss": 0.2191,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.24446979011194067,
"learning_rate": 4.22084697268323e-06,
"loss": 0.1926,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.24210628510972781,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1817,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.24690811732895793,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1825,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.2652643544850087,
"learning_rate": 4.11716698310493e-06,
"loss": 0.2011,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.262085687844557,
"learning_rate": 4.082860616555848e-06,
"loss": 0.1996,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.2751508463730161,
"learning_rate": 4.048681523029125e-06,
"loss": 0.1808,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.26440839889067164,
"learning_rate": 4.014629975819882e-06,
"loss": 0.1975,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.24996874866930333,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.204,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.24116999648724544,
"learning_rate": 3.94691060843287e-06,
"loss": 0.226,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.2631178885354638,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1932,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.2538717151461219,
"learning_rate": 3.879704680319503e-06,
"loss": 0.1738,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.2623203883356945,
"learning_rate": 3.846294928353471e-06,
"loss": 0.2254,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.2959004639396224,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.2051,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.23522244118945543,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.2108,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.2264717988625861,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.2127,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.22800878052371495,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1892,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.22696330108435117,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.209,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.24959481472245368,
"learning_rate": 3.648558150665886e-06,
"loss": 0.2033,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.24062307521518123,
"learning_rate": 3.616058102895075e-06,
"loss": 0.1665,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.23268771721614814,
"learning_rate": 3.583689060687079e-06,
"loss": 0.2049,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.23231558092785085,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.2011,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.23247791652579766,
"learning_rate": 3.51934502719796e-06,
"loss": 0.2095,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.23500983731351877,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.1934,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.26844975112731,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.2362,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.23237318135143814,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.1789,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.2311727337529998,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.2135,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.24624161057953423,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.2281,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.26875966557818437,
"learning_rate": 3.32948376099375e-06,
"loss": 0.1998,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.220143679050013,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.2048,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.26832424777555325,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.2412,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.23549490682775928,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1955,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.23540920534639068,
"learning_rate": 3.205572144561604e-06,
"loss": 0.2213,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.2391407423725077,
"learning_rate": 3.174929343376374e-06,
"loss": 0.2074,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.26246479074442114,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.1904,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.2361818877228161,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.2272,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.23777978008607806,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.2039,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.24784627666570033,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.2386,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.24008541916987308,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1968,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.24967516576999518,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.2231,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.22525281863086394,
"learning_rate": 2.964210239224696e-06,
"loss": 0.2117,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.22045562293971216,
"learning_rate": 2.934650470499869e-06,
"loss": 0.1937,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.23114604996160723,
"learning_rate": 2.905227155855128e-06,
"loss": 0.1884,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.260577828553986,
"learning_rate": 2.875940530558519e-06,
"loss": 0.2339,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.235181293573089,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.2586,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.23775892068123983,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.1919,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.23527307133812642,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.2272,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.2245308197801299,
"learning_rate": 2.760165589921184e-06,
"loss": 0.2199,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.24242516880482454,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.2225,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.2420144741499501,
"learning_rate": 2.703104292151164e-06,
"loss": 0.2094,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.22579041136365988,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.2081,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.23158911248052447,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.2036,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.22590695535155472,
"learning_rate": 2.618550200289303e-06,
"loss": 0.22,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.22894785839927373,
"learning_rate": 2.59064316684412e-06,
"loss": 0.2163,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.2389729728990816,
"learning_rate": 2.562875338152264e-06,
"loss": 0.1799,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.5584308302370931,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.212,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.28014026285008564,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.2161,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.26221514872703516,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1758,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.23977735723433208,
"learning_rate": 2.45320049480108e-06,
"loss": 0.2517,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.23130059243926465,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.2044,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.22969760494031596,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1928,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.24499284616216666,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.2064,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.22873493803406858,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.2024,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.2460118589040839,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1656,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.22924098524560685,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.2108,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.38488923753866955,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.2074,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.24314694656969701,
"learning_rate": 2.24059906434998e-06,
"loss": 0.2167,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.22755012422402532,
"learning_rate": 2.214660694699091e-06,
"loss": 0.2371,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.2519970227810416,
"learning_rate": 2.188864536147355e-06,
"loss": 0.2008,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.22797182895308446,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.2235,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.23444715655705975,
"learning_rate": 2.137699676264171e-06,
"loss": 0.2062,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2384290316222321,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.204,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.23900418838669635,
"learning_rate": 2.08710612114857e-06,
"loss": 0.187,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.22727021130408617,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.2236,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.2307784729261204,
"learning_rate": 2.0370854889762e-06,
"loss": 0.1679,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.21833402228504864,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.206,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.2410105756629979,
"learning_rate": 1.987639379598427e-06,
"loss": 0.2092,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.2584471993991531,
"learning_rate": 1.963132265887919e-06,
"loss": 0.1913,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.2507319704794648,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.1716,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.23702186883916876,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.1886,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.22386652550905292,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.1829,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.24584861465014698,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.2047,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.2621240762354824,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1941,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.2469318490663371,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.2074,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.2521861830722741,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.2033,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.25895737256855805,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.2025,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.25415407931953826,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.2625,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.2234118599200071,
"learning_rate": 1.726025117089718e-06,
"loss": 0.2115,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.26669480712687205,
"learning_rate": 1.70311497206598e-06,
"loss": 0.2073,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.28854647387742044,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.1923,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.21634458214333746,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.2029,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.24819751142807803,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.2046,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.25735658599213757,
"learning_rate": 1.612939220355272e-06,
"loss": 0.2218,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.2458429219558819,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1906,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.2514844484681909,
"learning_rate": 1.568732761074121e-06,
"loss": 0.1962,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.26359442908447017,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.2092,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.21980431043641316,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.2129,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.23822318369060652,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.1812,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.22126903161081196,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.2054,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.22367805250833708,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.1987,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.22793909964810724,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.22,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.2287803624968138,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.1902,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.2174185009968746,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.2106,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.23636838829579584,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.2203,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.2160967938752529,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.1954,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.23299816216165203,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.2095,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.22701634029385415,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.2342,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.21532968317209827,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.2219,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.2174013438343391,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1643,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.22933877760408103,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.1989,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.23522291640417478,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1971,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.23710185687708932,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1928,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.2289143570310709,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1765,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.230924949520026,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.1832,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.39833165290626227,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.2123,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.23355051335659893,
"learning_rate": 1.140415161609505e-06,
"loss": 0.2202,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.25641947432487555,
"learning_rate": 1.121666393356604e-06,
"loss": 0.2209,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.218083326714615,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.2154,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.2176402813397855,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.2215,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.23626082678042842,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1827,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.21940734907738726,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.2072,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.22462568631402965,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.2369,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.24821343437083618,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1884,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.22044497278050065,
"learning_rate": 9.9465987684501e-07,
"loss": 0.1971,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.22967414822981633,
"learning_rate": 9.771228098856534e-07,
"loss": 0.2003,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.21787457647014347,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1935,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.24142703456258133,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1737,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.23347064286696775,
"learning_rate": 9.254248023435864e-07,
"loss": 0.2006,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.24030856337228257,
"learning_rate": 9.084969902482288e-07,
"loss": 0.1779,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.25197333761013024,
"learning_rate": 8.917218333021993e-07,
"loss": 0.2646,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.2362362457662725,
"learning_rate": 8.750994656392109e-07,
"loss": 0.1724,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.23168912090410212,
"learning_rate": 8.58630020171265e-07,
"loss": 0.1793,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.2311265750581956,
"learning_rate": 8.423136285876099e-07,
"loss": 0.2208,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.22592106987126484,
"learning_rate": 8.26150421353682e-07,
"loss": 0.2012,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.22300154023674806,
"learning_rate": 8.101405277100549e-07,
"loss": 0.2123,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.23001702500848611,
"learning_rate": 7.942840756714077e-07,
"loss": 0.191,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.2449530039026689,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1764,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.2316336468580194,
"learning_rate": 7.630320023322024e-07,
"loss": 0.2411,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.2489280258951386,
"learning_rate": 7.476366309223903e-07,
"loss": 0.2249,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.22893982469368104,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1933,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.233686716989286,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1802,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.28469853537524276,
"learning_rate": 7.023746512481989e-07,
"loss": 0.2185,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.23064771362232014,
"learning_rate": 6.875957716681902e-07,
"loss": 0.208,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.5722747600310555,
"learning_rate": 6.729713135577687e-07,
"loss": 0.2259,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.23004396239074806,
"learning_rate": 6.585013938537143e-07,
"loss": 0.2079,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.21785547702350166,
"learning_rate": 6.441861282571115e-07,
"loss": 0.2116,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.24833563667231873,
"learning_rate": 6.300256312324471e-07,
"loss": 0.2305,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.25349730590114855,
"learning_rate": 6.160200160066776e-07,
"loss": 0.2234,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.23555955421749372,
"learning_rate": 6.021693945683327e-07,
"loss": 0.2099,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.24432736666318072,
"learning_rate": 5.884738776666044e-07,
"loss": 0.1954,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.3707585882037652,
"learning_rate": 5.749335748104878e-07,
"loss": 0.1855,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.233564985003395,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1842,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.2406242442411204,
"learning_rate": 5.483190430647289e-07,
"loss": 0.1923,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.23345587557599137,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1988,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.28305145440530743,
"learning_rate": 5.22326650565701e-07,
"loss": 0.2104,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.23505012019790672,
"learning_rate": 5.095640171043048e-07,
"loss": 0.2172,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.2237947546058349,
"learning_rate": 4.969572286496594e-07,
"loss": 0.2288,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.22330291755892143,
"learning_rate": 4.845063860053079e-07,
"loss": 0.2168,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.24794710273429038,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1887,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.21648945396137342,
"learning_rate": 4.600729351261857e-07,
"loss": 0.2469,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.23301288354050018,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.2097,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.21380796834766738,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1809,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.2703583274562674,
"learning_rate": 4.245948007320588e-07,
"loss": 0.1811,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.24442979751610855,
"learning_rate": 4.130816799401971e-07,
"loss": 0.2123,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.23337314451402694,
"learning_rate": 4.017251756252827e-07,
"loss": 0.1934,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.24012514359963644,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.2199,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.23071013576237787,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.2337,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.24684817631911143,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1998,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.2083477303857567,
"learning_rate": 3.578671204615347e-07,
"loss": 0.207,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.22443980035137398,
"learning_rate": 3.472950355547267e-07,
"loss": 0.181,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.2323396411999352,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.199,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.23218790792754287,
"learning_rate": 3.266223765343224e-07,
"loss": 0.2185,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.22718975083508444,
"learning_rate": 3.16521967718757e-07,
"loss": 0.2001,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.24132044488357368,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.2034,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.2309197497880529,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.2019,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.22192056700857932,
"learning_rate": 2.871653894244597e-07,
"loss": 0.1748,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.22643181484276684,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.196,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.23127531276345437,
"learning_rate": 2.683823233205485e-07,
"loss": 0.2006,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.21847461771662519,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1917,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.23090672788716912,
"learning_rate": 2.502303499116754e-07,
"loss": 0.1881,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.2364552083744487,
"learning_rate": 2.413912051848666e-07,
"loss": 0.1953,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.21589900695051165,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.2391,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.23186910405866165,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.1951,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.21125341233664266,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.2083,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.22800038795023547,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1819,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.21112482309452488,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.2017,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.2322163827419028,
"learning_rate": 1.916764975761254e-07,
"loss": 0.215,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.22576518888655658,
"learning_rate": 1.839446914621923e-07,
"loss": 0.1855,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.21441985339197292,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1942,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.22477783864414128,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.199,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.23115449637785146,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.1786,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.23476722554336105,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.2094,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.23596997320005414,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.2008,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.22072890722101515,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.2083,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.2293409162280026,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.2038,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.232885585912813,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1872,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.2449498273510922,
"learning_rate": 1.214956261808542e-07,
"loss": 0.2185,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.22956021764259568,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1931,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.23925270734919707,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.229,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.22875497673367048,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.2228,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2302673378630637,
"learning_rate": 9.787081204466253e-08,
"loss": 0.2297,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.23308279134442947,
"learning_rate": 9.236233586200627e-08,
"loss": 0.2004,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2383705050216258,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1976,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.22558234820338938,
"learning_rate": 8.182296909035047e-08,
"loss": 0.2034,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.2239786270476794,
"learning_rate": 7.679216277384439e-08,
"loss": 0.2252,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.21834897346583804,
"learning_rate": 7.19206618963053e-08,
"loss": 0.2001,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.23338427818200963,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1801,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.3080039496840972,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1894,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.22317598495441168,
"learning_rate": 5.826237505053067e-08,
"loss": 0.2121,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.21831521769487966,
"learning_rate": 5.4028472710137e-08,
"loss": 0.206,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.24475162403654244,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.2251,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.22544963788953729,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.1838,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.22465279975027436,
"learning_rate": 4.228381947004812e-08,
"loss": 0.2166,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.21610667233813188,
"learning_rate": 3.868805732811343e-08,
"loss": 0.213,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.2302219256576141,
"learning_rate": 3.525190530455636e-08,
"loss": 0.2004,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.24895423994541624,
"learning_rate": 3.197539087475887e-08,
"loss": 0.2158,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.2315486495179633,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.2031,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.22347250694632176,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.1777,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.21804877749366836,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.224,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.23026189807169534,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.1785,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.22816060600922533,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1687,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.22286622125422223,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.2235,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.21003234627564252,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.1995,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.2319928451004343,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.2105,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.2288891023636212,
"learning_rate": 9.67435416023843e-09,
"loss": 0.2185,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.22238692245174582,
"learning_rate": 7.995445897566178e-09,
"loss": 0.1807,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.21582892521708108,
"learning_rate": 6.476393170578021e-09,
"loss": 0.2234,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.22958203026285853,
"learning_rate": 5.11720812557348e-09,
"loss": 0.2021,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.2327403060433384,
"learning_rate": 3.917901630563492e-09,
"loss": 0.1971,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.22551341244118314,
"learning_rate": 2.878483275170485e-09,
"loss": 0.2011,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.2422481242955349,
"learning_rate": 1.998961370555108e-09,
"loss": 0.2313,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.2361431477937266,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.1938,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.23302061794178197,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1917,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.2264724615669771,
"learning_rate": 3.198382947511292e-10,
"loss": 0.2011,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.21792269779599655,
"learning_rate": 7.995973352770137e-11,
"loss": 0.2092,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.21131196075574488,
"learning_rate": 0.0,
"loss": 0.2246,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 2.0080798451032392e+18,
"train_loss": 0.36193929355636784,
"train_runtime": 24487.1768,
"train_samples_per_second": 6.452,
"train_steps_per_second": 0.05
}
],
"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": 2.0080798451032392e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}