gsmyrnis's picture
End of training
cc8e432 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": 8.601307335783071,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.0671,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 8.328370218623881,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0801,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 8.043152061335471,
"learning_rate": 9.67741935483871e-07,
"loss": 1.0749,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 8.056314507487304,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.0638,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 7.331645722018016,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.04,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 7.6190157099752405,
"learning_rate": 1.935483870967742e-06,
"loss": 1.0719,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 5.496804780294433,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.9953,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 5.054216713327257,
"learning_rate": 2.580645161290323e-06,
"loss": 1.0036,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 3.3047701600265222,
"learning_rate": 2.903225806451613e-06,
"loss": 0.9263,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 3.5597607205806927,
"learning_rate": 3.225806451612903e-06,
"loss": 0.949,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 3.184074850648674,
"learning_rate": 3.548387096774194e-06,
"loss": 0.9372,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.171492214529249,
"learning_rate": 3.870967741935484e-06,
"loss": 0.8798,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 3.323450108254412,
"learning_rate": 4.193548387096774e-06,
"loss": 0.8801,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 2.742977648125394,
"learning_rate": 4.516129032258065e-06,
"loss": 0.8508,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 2.363565284710602,
"learning_rate": 4.838709677419355e-06,
"loss": 0.8573,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 3.1029446518359243,
"learning_rate": 5.161290322580646e-06,
"loss": 0.8182,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 3.4546613462867235,
"learning_rate": 5.483870967741935e-06,
"loss": 0.8013,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.8140387415461303,
"learning_rate": 5.806451612903226e-06,
"loss": 0.7732,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.8065206409965802,
"learning_rate": 6.129032258064517e-06,
"loss": 0.7499,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.415974319647387,
"learning_rate": 6.451612903225806e-06,
"loss": 0.7559,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.8264372805863862,
"learning_rate": 6.774193548387097e-06,
"loss": 0.7403,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.938778693596728,
"learning_rate": 7.096774193548388e-06,
"loss": 0.7354,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.458755968577317,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.7195,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.0820896807225167,
"learning_rate": 7.741935483870968e-06,
"loss": 0.7093,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.2494882201856516,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7161,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.1170946589710122,
"learning_rate": 8.387096774193549e-06,
"loss": 0.7072,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.9500861634451244,
"learning_rate": 8.70967741935484e-06,
"loss": 0.697,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.0213395502456828,
"learning_rate": 9.03225806451613e-06,
"loss": 0.6992,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.9236590295045541,
"learning_rate": 9.35483870967742e-06,
"loss": 0.6714,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.962466841098055,
"learning_rate": 9.67741935483871e-06,
"loss": 0.6813,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.919342335159254,
"learning_rate": 1e-05,
"loss": 0.6829,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.8845150355507606,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.6629,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.8768891517103876,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.6653,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.9049846526026932,
"learning_rate": 1.096774193548387e-05,
"loss": 0.6754,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.7500430156563704,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.6663,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.8832752949815681,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.6767,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.9602701756034921,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.6971,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.7585068773827992,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.6689,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.8367884514697835,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.6586,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 1.031371406577607,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.6576,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.8263000135369607,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6529,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 1.065208013543696,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.6508,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.7316306276541995,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.6558,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.9743216409324986,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.6625,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.8066544692846043,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.6537,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 1.1779755296318692,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.636,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.8500973316240698,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.6385,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 1.24991285098212,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6522,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 1.051399527983429,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.6327,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.8691659533758651,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6485,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.9969101538778252,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.6426,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.8151555990287668,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6451,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.7649109348101558,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.6197,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.9194285126383531,
"learning_rate": 1.741935483870968e-05,
"loss": 0.6241,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.8624652111595815,
"learning_rate": 1.774193548387097e-05,
"loss": 0.6252,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.8619342215258212,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6142,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.8811579237018927,
"learning_rate": 1.838709677419355e-05,
"loss": 0.6287,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.8964789202934302,
"learning_rate": 1.870967741935484e-05,
"loss": 0.6336,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.8321002563034355,
"learning_rate": 1.903225806451613e-05,
"loss": 0.6336,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.9816360565025389,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6236,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.9888155180024727,
"learning_rate": 1.967741935483871e-05,
"loss": 0.603,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 1.0072807046042542,
"learning_rate": 2e-05,
"loss": 0.6441,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.9636459296131132,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.642,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.8266736900572581,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.6279,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 1.221832665610348,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.615,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.7514633111550074,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.6104,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 1.0498923058409986,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6216,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.9000162776673665,
"learning_rate": 2.193548387096774e-05,
"loss": 0.6339,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 1.106733641568249,
"learning_rate": 2.225806451612903e-05,
"loss": 0.6176,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.8731572354355446,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.6268,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 1.1722193111162509,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6359,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.8386394965208652,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.6224,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.944348629301325,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.6153,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.9247452125739438,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.6253,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.7944107315509182,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.6123,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.1484937949634482,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.6189,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.9292476078137539,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.6104,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.0891516634670004,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.612,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.0125931901836607,
"learning_rate": 2.548387096774194e-05,
"loss": 0.6044,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.7704913609319698,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.5991,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.9555539036395295,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.6237,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.0094721432498157,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.6218,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.8256578400915043,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6101,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.911329995727236,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.6347,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.8892559012159675,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.6258,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.8863803677637514,
"learning_rate": 2.774193548387097e-05,
"loss": 0.6242,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.8672542928040163,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.6181,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.984864314974834,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.6331,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.0363748291362564,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.6209,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.9720532895439381,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.6129,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.8661027613049201,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6177,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.0482097807387907,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.6139,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.9410708549220071,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.6193,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.8944470381558258,
"learning_rate": 3.032258064516129e-05,
"loss": 0.6083,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.121175719026002,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.61,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.900691155368688,
"learning_rate": 3.096774193548387e-05,
"loss": 0.615,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.9733588489046303,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.6063,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.2563452964080144,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6081,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.8362013959150261,
"learning_rate": 3.193548387096774e-05,
"loss": 0.6159,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 0.9751651255922443,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.5984,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.9930672720335189,
"learning_rate": 3.258064516129033e-05,
"loss": 0.6158,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.9225936323313001,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.6258,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.1183664839762288,
"learning_rate": 3.322580645161291e-05,
"loss": 0.6236,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.0188776994360207,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.6071,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.035057561608788,
"learning_rate": 3.387096774193549e-05,
"loss": 0.6065,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.0008249201593546,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.6316,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.0393220510318477,
"learning_rate": 3.451612903225807e-05,
"loss": 0.5935,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.2235117623381178,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6176,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.8197791471783691,
"learning_rate": 3.516129032258065e-05,
"loss": 0.609,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.2136247487824305,
"learning_rate": 3.548387096774194e-05,
"loss": 0.6029,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 0.81439119522686,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.609,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.0793357211175787,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6157,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.8955382385617229,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.621,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.8424280995987296,
"learning_rate": 3.67741935483871e-05,
"loss": 0.6091,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 0.9614429601443664,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6309,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.12659146242878,
"learning_rate": 3.741935483870968e-05,
"loss": 0.6073,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.7893001176231549,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.6088,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.1824239430768966,
"learning_rate": 3.806451612903226e-05,
"loss": 0.6019,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.8193426858894743,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.6171,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 1.192684937240083,
"learning_rate": 3.870967741935484e-05,
"loss": 0.6094,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.0214472681728577,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.6086,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.2537400776746723,
"learning_rate": 3.935483870967742e-05,
"loss": 0.6147,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.0086032835353171,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.6008,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.9983731516329845,
"learning_rate": 4e-05,
"loss": 0.6149,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.963867530815525,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.6135,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.861531919065205,
"learning_rate": 3.999968016170525e-05,
"loss": 0.5962,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.8339398719901872,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.6047,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.8691596479262259,
"learning_rate": 3.999872065705065e-05,
"loss": 0.6077,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.9456271777651757,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6251,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.8795833614452091,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.6079,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.9481809567196432,
"learning_rate": 3.999608209836944e-05,
"loss": 0.6143,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.035128391280372,
"learning_rate": 3.999488279187443e-05,
"loss": 0.6174,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.9741649169985407,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.6097,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.009564516628795,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.6028,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.8189473860614097,
"learning_rate": 3.999032564583976e-05,
"loss": 0.6109,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.7711583404659397,
"learning_rate": 3.998848689546591e-05,
"loss": 0.6079,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.8888806005595777,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.6282,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.6855818646280398,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.6197,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.9845009570528365,
"learning_rate": 3.998201174509306e-05,
"loss": 0.6146,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.7283683559550065,
"learning_rate": 3.997953378607961e-05,
"loss": 0.6115,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.7141649503512427,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.6105,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.8195682712170989,
"learning_rate": 3.997409862168454e-05,
"loss": 0.6211,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.773070729113917,
"learning_rate": 3.997114145976237e-05,
"loss": 0.6059,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.7371190396324736,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.588,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.8955694082272588,
"learning_rate": 3.996474809469545e-05,
"loss": 0.5998,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.692910928608666,
"learning_rate": 3.996131194267189e-05,
"loss": 0.6177,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.7677870181624947,
"learning_rate": 3.995771618052995e-05,
"loss": 0.6035,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.7130497389772649,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6022,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.7683241501045616,
"learning_rate": 3.995004594217343e-05,
"loss": 0.6101,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.7437949807864661,
"learning_rate": 3.994597152728987e-05,
"loss": 0.6052,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.8791243101429292,
"learning_rate": 3.994173762494947e-05,
"loss": 0.6085,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.8637420760083566,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.6149,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.7553427621159046,
"learning_rate": 3.993279149458988e-05,
"loss": 0.6131,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.8584041741348365,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.6046,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.8032089812907629,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6017,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.8528505915523048,
"learning_rate": 3.991817703090965e-05,
"loss": 0.6173,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.8462879635101395,
"learning_rate": 3.991298695938038e-05,
"loss": 0.6158,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.6113525435262692,
"learning_rate": 3.9907637664138e-05,
"loss": 0.5845,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.8677667310980961,
"learning_rate": 3.990212918795534e-05,
"loss": 0.6182,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.6387710399147766,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.6067,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.944527687580318,
"learning_rate": 3.989063487022416e-05,
"loss": 0.6028,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.879696954251426,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.6069,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.727850105013305,
"learning_rate": 3.987850437381915e-05,
"loss": 0.5987,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.8640683605530534,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6048,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.8283137417028975,
"learning_rate": 3.986573808672002e-05,
"loss": 0.6088,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.6716052429497308,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.6048,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.8230659869941046,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6163,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.8026917825618893,
"learning_rate": 3.984539744726554e-05,
"loss": 0.6013,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 1.670052493143577,
"learning_rate": 3.983829979402038e-05,
"loss": 0.6091,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.9524521045858511,
"learning_rate": 3.983104351425871e-05,
"loss": 0.5733,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 1.1814152660515107,
"learning_rate": 3.982362866600154e-05,
"loss": 0.591,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.9192533436866611,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.6205,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 1.3460285467122992,
"learning_rate": 3.980832350242388e-05,
"loss": 0.6023,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 1.019056204112095,
"learning_rate": 3.980043330948306e-05,
"loss": 0.6212,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 1.1926252765488825,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6179,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 901.3045577952346,
"learning_rate": 3.978417801674579e-05,
"loss": 0.6155,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 1.4240448374323886,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.6156,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 1.7804920428194349,
"learning_rate": 3.976728995023254e-05,
"loss": 0.5928,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 1.560708891107332,
"learning_rate": 3.975860879481514e-05,
"loss": 0.5923,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 1.0005273237372652,
"learning_rate": 3.974976965008833e-05,
"loss": 0.5949,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.7631891438095894,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.6058,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 1.0082936060384486,
"learning_rate": 3.973161767667946e-05,
"loss": 0.5942,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.5865662053584003,
"learning_rate": 3.972230499314009e-05,
"loss": 0.6025,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 1.205951780749423,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.5923,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.697247446063407,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.6026,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 1.0293974139409006,
"learning_rate": 3.969342105253098e-05,
"loss": 0.5944,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.7663210855188518,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.602,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.7614527030127093,
"learning_rate": 3.967337762346568e-05,
"loss": 0.6071,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.7131382891493514,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.605,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.7041669630271111,
"learning_rate": 3.965270496444528e-05,
"loss": 0.5844,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.9125353793424347,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.5922,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.8033636504723382,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.603,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.6301037464713812,
"learning_rate": 3.962051762160151e-05,
"loss": 0.6076,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.8076536990414389,
"learning_rate": 3.960947462140638e-05,
"loss": 0.6082,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.6034196150146528,
"learning_rate": 3.959827482437472e-05,
"loss": 0.5885,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.6592354583519062,
"learning_rate": 3.958691832005981e-05,
"loss": 0.6039,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.6255889855795861,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.5909,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.7034480739826989,
"learning_rate": 3.956373555405774e-05,
"loss": 0.6004,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.6870558396033561,
"learning_rate": 3.955190947773937e-05,
"loss": 0.5867,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.640746034962403,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.5891,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.7006353613946127,
"learning_rate": 3.952778841127214e-05,
"loss": 0.5934,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.602140518409273,
"learning_rate": 3.95154936139947e-05,
"loss": 0.593,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.7015094919483297,
"learning_rate": 3.950304277135034e-05,
"loss": 0.5944,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.6385220150574151,
"learning_rate": 3.94904359828957e-05,
"loss": 0.6058,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.6169385010060486,
"learning_rate": 3.94776733494343e-05,
"loss": 0.5971,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.652906831428837,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.601,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.6690024567436169,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.6262,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.5929507519833446,
"learning_rate": 3.943845140573213e-05,
"loss": 0.5871,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.6339309739992114,
"learning_rate": 3.942506642518952e-05,
"loss": 0.597,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.7543601584351635,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.6054,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.5538456158966861,
"learning_rate": 3.939783060543167e-05,
"loss": 0.5899,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.6481027453508507,
"learning_rate": 3.938397998399333e-05,
"loss": 0.5931,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.6312718550334367,
"learning_rate": 3.936997436876756e-05,
"loss": 0.5899,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.62579831735874,
"learning_rate": 3.935581387174289e-05,
"loss": 0.5934,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.7384698457219656,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.5967,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.9254692852206298,
"learning_rate": 3.932702868644224e-05,
"loss": 0.5996,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.631310813298911,
"learning_rate": 3.931240422833181e-05,
"loss": 0.5765,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.8925672576895171,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.5767,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.7788672562044252,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.5799,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6881796270332703,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.5924,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.6587487992177504,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.6005,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.6692836222755285,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.5927,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.6969851294838456,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.5834,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.534439946632832,
"learning_rate": 3.92057159243286e-05,
"loss": 0.5913,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.6472051815976058,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.5924,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.6113294241987056,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.6043,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.6270290839195711,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.5805,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.5748844146975585,
"learning_rate": 3.914136997982874e-05,
"loss": 0.6002,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.5528872957966856,
"learning_rate": 3.912490053436079e-05,
"loss": 0.5738,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.6707942417974989,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.5897,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.5171394530888248,
"learning_rate": 3.909150300975178e-05,
"loss": 0.5756,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.58583187405586,
"learning_rate": 3.907457519765642e-05,
"loss": 0.5778,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.5824803868778964,
"learning_rate": 3.905749486576607e-05,
"loss": 0.5922,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.5379734255882475,
"learning_rate": 3.90402621506546e-05,
"loss": 0.5871,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.5727707409165514,
"learning_rate": 3.902287719011435e-05,
"loss": 0.6104,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.5308799490549538,
"learning_rate": 3.900534012315499e-05,
"loss": 0.5945,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5842235621064716,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.6006,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.564412832000257,
"learning_rate": 3.896981023209777e-05,
"loss": 0.5771,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.6364897325848058,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.589,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.5253107111205129,
"learning_rate": 3.893367361386491e-05,
"loss": 0.5955,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.6039192982841366,
"learning_rate": 3.891537814248417e-05,
"loss": 0.5868,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.6480139536377179,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5936,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.6442760988981961,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.564,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.7831585109477994,
"learning_rate": 3.885958483839049e-05,
"loss": 0.589,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.643972752960703,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.5785,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.8035995934415533,
"learning_rate": 3.882163505079171e-05,
"loss": 0.592,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.6641631808309529,
"learning_rate": 3.880243433489132e-05,
"loss": 0.5745,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.8228945087595148,
"learning_rate": 3.878308327522702e-05,
"loss": 0.5085,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.5978767260782761,
"learning_rate": 3.876358202652936e-05,
"loss": 0.4815,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.7983360766497524,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.5106,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.7224818163439991,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.4836,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.6980232987595988,
"learning_rate": 3.870417871154801e-05,
"loss": 0.4785,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.7394136149932103,
"learning_rate": 3.868407827802192e-05,
"loss": 0.4962,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.7271251478760489,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.4863,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.6693180954366243,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.4908,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.6547117881914994,
"learning_rate": 3.862288124195319e-05,
"loss": 0.4721,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.6104682243610603,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.5071,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.6526499845199218,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.4628,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.7732752071114543,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.488,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.6432842485185836,
"learning_rate": 3.853920126717703e-05,
"loss": 0.4928,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.6212357969865219,
"learning_rate": 3.851791025492574e-05,
"loss": 0.479,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.764953056065485,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.4664,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.6982014473424545,
"learning_rate": 3.847488419569866e-05,
"loss": 0.4877,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.6836946284425846,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.4887,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.6673965025437539,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.5012,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.6931748715256172,
"learning_rate": 3.840923760917199e-05,
"loss": 0.4739,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.7516993733787104,
"learning_rate": 3.838706077964473e-05,
"loss": 0.4885,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.756131617096004,
"learning_rate": 3.836473692766944e-05,
"loss": 0.4884,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.6076676606272249,
"learning_rate": 3.834226623174705e-05,
"loss": 0.4984,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.9336193776143785,
"learning_rate": 3.831964887155264e-05,
"loss": 0.492,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.5892957068673769,
"learning_rate": 3.829688502793403e-05,
"loss": 0.4654,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.8592674691512096,
"learning_rate": 3.827397488291029e-05,
"loss": 0.5004,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.6026900919257342,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.4789,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.7198578587621196,
"learning_rate": 3.822771642257145e-05,
"loss": 0.4923,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.6300906722700229,
"learning_rate": 3.820436847713776e-05,
"loss": 0.4692,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.6410679188416422,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.4867,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.5685420929295426,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.4908,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.6515367145027284,
"learning_rate": 3.813345202354145e-05,
"loss": 0.4843,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.6595372258897648,
"learning_rate": 3.810952296329563e-05,
"loss": 0.5113,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.7222576385468328,
"learning_rate": 3.808544909978676e-05,
"loss": 0.4892,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.5670203849913091,
"learning_rate": 3.806123062550882e-05,
"loss": 0.4843,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.7228049683525365,
"learning_rate": 3.803686773411208e-05,
"loss": 0.4865,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.542549494689836,
"learning_rate": 3.801236062040158e-05,
"loss": 0.4914,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.6780996382678266,
"learning_rate": 3.798770948033553e-05,
"loss": 0.4844,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.5583943328749598,
"learning_rate": 3.796291451102381e-05,
"loss": 0.4972,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.6284270960749772,
"learning_rate": 3.793797591072631e-05,
"loss": 0.4834,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.7369690868299322,
"learning_rate": 3.791289387885144e-05,
"loss": 0.4982,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.6566992092009426,
"learning_rate": 3.788766861595443e-05,
"loss": 0.4851,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.6585364685057487,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.4899,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.6161008720858371,
"learning_rate": 3.783678920503983e-05,
"loss": 0.4848,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.6891372529477686,
"learning_rate": 3.781113546385265e-05,
"loss": 0.4837,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.6132773094257938,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.4892,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.632519128347971,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.4951,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.6177101316598756,
"learning_rate": 3.77333205623025e-05,
"loss": 0.4876,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.606537078795999,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.4956,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.5762153104414762,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.4712,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.6276081469134208,
"learning_rate": 3.765422951112704e-05,
"loss": 0.4837,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.7556339916883433,
"learning_rate": 3.762758321970216e-05,
"loss": 0.4859,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.5144316680764498,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.4913,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.712838822967591,
"learning_rate": 3.757386800198538e-05,
"loss": 0.4911,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.6049263387532265,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.4887,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.6433024287695573,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.5059,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.6579892835474357,
"learning_rate": 3.749224181796305e-05,
"loss": 0.486,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5854203012511091,
"learning_rate": 3.746475306375545e-05,
"loss": 0.5032,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.5890754963445658,
"learning_rate": 3.743712466184774e-05,
"loss": 0.4837,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.5598235114509308,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.4874,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.6016393197031823,
"learning_rate": 3.73814497997107e-05,
"loss": 0.4967,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.5623012870702787,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.4671,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.6600071920698355,
"learning_rate": 3.732521901224724e-05,
"loss": 0.4837,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.5407412999361597,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.4871,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.6870966552006237,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.4808,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.6801531356608841,
"learning_rate": 3.723983441007882e-05,
"loss": 0.4941,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.6070741582846388,
"learning_rate": 3.721109687296781e-05,
"loss": 0.4916,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.7572192087409956,
"learning_rate": 3.718222171638484e-05,
"loss": 0.491,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.5020775146001032,
"learning_rate": 3.715320917121488e-05,
"loss": 0.5041,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.7062152909365481,
"learning_rate": 3.712405946944149e-05,
"loss": 0.4891,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.552553623435224,
"learning_rate": 3.709477284414488e-05,
"loss": 0.4986,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.8008713464037737,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.486,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.6398280795379085,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.4711,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.7698424296945966,
"learning_rate": 3.70060937743295e-05,
"loss": 0.4997,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.5598380783636175,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.496,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.7457705384609147,
"learning_rate": 3.694629409915555e-05,
"loss": 0.4815,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.6156345358354545,
"learning_rate": 3.691619088858401e-05,
"loss": 0.4851,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.6978018253155585,
"learning_rate": 3.68859524166009e-05,
"loss": 0.469,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.5648085114782528,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.4778,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.6136679833220141,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.4815,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.5474239702918084,
"learning_rate": 3.67944278554384e-05,
"loss": 0.4837,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.5798018765707861,
"learning_rate": 3.676365076645557e-05,
"loss": 0.4923,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.5396607529455884,
"learning_rate": 3.673273963576791e-05,
"loss": 0.4796,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.5947797428016354,
"learning_rate": 3.670169471054001e-05,
"loss": 0.5054,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.744928538336682,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.4941,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.6094963766791355,
"learning_rate": 3.663920447046888e-05,
"loss": 0.4794,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.6892468510743713,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.488,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.6061729582793888,
"learning_rate": 3.657618204491937e-05,
"loss": 0.4911,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.8640771489711204,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.506,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.5826197758385832,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.4914,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.7498014961420068,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.4795,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.5091754812595436,
"learning_rate": 3.644854871713611e-05,
"loss": 0.4875,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.6840410204118376,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.4698,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.5224257562777955,
"learning_rate": 3.638394189710493e-05,
"loss": 0.4887,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.6911352590962296,
"learning_rate": 3.635144184933412e-05,
"loss": 0.4803,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.6511541559309822,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.4813,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.5053766865354835,
"learning_rate": 3.628604977762752e-05,
"loss": 0.4937,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.614956008400144,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.4955,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.5517758580494271,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.4849,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.6494857412686765,
"learning_rate": 3.618698565901727e-05,
"loss": 0.4972,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.6612740612246343,
"learning_rate": 3.615370507164654e-05,
"loss": 0.5028,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.5974163367124855,
"learning_rate": 3.61202953196805e-05,
"loss": 0.4936,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.6541985175762443,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.4954,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.5517220273714586,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.4794,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.7452847492495228,
"learning_rate": 3.601929375279662e-05,
"loss": 0.5018,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.5236370594969174,
"learning_rate": 3.598537002418012e-05,
"loss": 0.4909,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.8028919189264919,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.48,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.4844645228932066,
"learning_rate": 3.591713938344416e-05,
"loss": 0.4889,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.8036720892580059,
"learning_rate": 3.588283301689507e-05,
"loss": 0.4837,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.5739506388500862,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.483,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.6808228674347147,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.4767,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.6952582092409157,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.4795,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.6143809124379651,
"learning_rate": 3.574434032194994e-05,
"loss": 0.5006,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.6166451676168143,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.4813,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.5340257791904888,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.4971,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.6965499200086757,
"learning_rate": 3.563914797637216e-05,
"loss": 0.4895,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.5263755468576894,
"learning_rate": 3.560383338592649e-05,
"loss": 0.5025,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.6138314396879392,
"learning_rate": 3.556839402764487e-05,
"loss": 0.5171,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.5389448323983257,
"learning_rate": 3.553283018489946e-05,
"loss": 0.4706,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.5782793932529388,
"learning_rate": 3.54971421420578e-05,
"loss": 0.4884,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.6671233514106224,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.477,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.6557662288210675,
"learning_rate": 3.54253945985191e-05,
"loss": 0.5056,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.5246419994007278,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.4818,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.6000319881890753,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.4938,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.5320966750609725,
"learning_rate": 3.531684894865866e-05,
"loss": 0.4817,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.5781884822890083,
"learning_rate": 3.528042173241128e-05,
"loss": 0.4656,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.6141250725950506,
"learning_rate": 3.524387233431892e-05,
"loss": 0.5014,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.5567906150235905,
"learning_rate": 3.520720104662958e-05,
"loss": 0.4899,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.5639093560922038,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.4696,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.5754814140134213,
"learning_rate": 3.513349397632281e-05,
"loss": 0.489,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.5187040104389421,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.4973,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.6103174280142268,
"learning_rate": 3.505930287892533e-05,
"loss": 0.481,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.6370677372784128,
"learning_rate": 3.502202656100099e-05,
"loss": 0.487,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.5911543675462141,
"learning_rate": 3.498463012735256e-05,
"loss": 0.4998,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.6001147758676448,
"learning_rate": 3.494711387700094e-05,
"loss": 0.4812,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.5404320724215192,
"learning_rate": 3.490947810992505e-05,
"loss": 0.4806,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.5627701773940565,
"learning_rate": 3.48717231270595e-05,
"loss": 0.4983,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.5233927664582952,
"learning_rate": 3.483384923029211e-05,
"loss": 0.486,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.5738192765291384,
"learning_rate": 3.479585672246156e-05,
"loss": 0.4848,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.5045154616058208,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.4763,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.5291603416845545,
"learning_rate": 3.471951708970527e-05,
"loss": 0.4845,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.5155334220428768,
"learning_rate": 3.468117057518921e-05,
"loss": 0.4757,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.5697326555231722,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.5037,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.5237498235254767,
"learning_rate": 3.460412568296731e-05,
"loss": 0.4901,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.6694985291691952,
"learning_rate": 3.456542792131039e-05,
"loss": 0.4749,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.6227297876909916,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.464,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.6665222289997121,
"learning_rate": 3.448768331403348e-05,
"loss": 0.492,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.5668376111805614,
"learning_rate": 3.44486370900573e-05,
"loss": 0.4711,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.5890330439913507,
"learning_rate": 3.440947533516396e-05,
"loss": 0.4794,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.6371735367790188,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.5217,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.5103169905516324,
"learning_rate": 3.43308064860925e-05,
"loss": 0.5048,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.5475262694964187,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.4884,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.587799112372174,
"learning_rate": 3.425167928295015e-05,
"loss": 0.5059,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.5716258905274335,
"learning_rate": 3.421194458890411e-05,
"loss": 0.4743,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.5214241335130699,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.4925,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.6052373536226633,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.4935,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.5528055739119957,
"learning_rate": 3.409205995219561e-05,
"loss": 0.4809,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.5977141626993169,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.4834,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.4907498686275603,
"learning_rate": 3.40115729298209e-05,
"loss": 0.4929,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.6640014169319526,
"learning_rate": 3.397116120327025e-05,
"loss": 0.4887,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.49496149210417634,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.4956,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.6563460655408488,
"learning_rate": 3.389000293509524e-05,
"loss": 0.4867,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.5069168499837369,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.4921,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.6795708441533735,
"learning_rate": 3.380840041143494e-05,
"loss": 0.4977,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.5272060546895645,
"learning_rate": 3.376743336885792e-05,
"loss": 0.4826,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.5831835501373667,
"learning_rate": 3.372635624225055e-05,
"loss": 0.5033,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.5973081692200758,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.4977,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.6433162027632308,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.4848,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.5758143511845033,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.4895,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.5380238966496841,
"learning_rate": 3.356095347769799e-05,
"loss": 0.4854,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.6169917175673243,
"learning_rate": 3.351933087522552e-05,
"loss": 0.4781,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.5417804655216881,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.4994,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.563809685669335,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.4962,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.5828072103469902,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.498,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.610024602322831,
"learning_rate": 3.335176280432308e-05,
"loss": 0.4755,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.589713927379954,
"learning_rate": 3.330960304618295e-05,
"loss": 0.475,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.5575099707765646,
"learning_rate": 3.326733686481153e-05,
"loss": 0.4665,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.6222957009854766,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.4759,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.5178795310704521,
"learning_rate": 3.318248658506012e-05,
"loss": 0.4771,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.5553212774984989,
"learning_rate": 3.313990316514069e-05,
"loss": 0.4831,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.5088919519878332,
"learning_rate": 3.309721467890571e-05,
"loss": 0.4826,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.5653409035849475,
"learning_rate": 3.305442146769115e-05,
"loss": 0.494,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.5264212921362773,
"learning_rate": 3.30115238736704e-05,
"loss": 0.4923,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.597019720425593,
"learning_rate": 3.296852223985148e-05,
"loss": 0.4895,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.5955269156933591,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.4891,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.5417937864138593,
"learning_rate": 3.288220822900796e-05,
"loss": 0.481,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.5462290805702545,
"learning_rate": 3.283889654214788e-05,
"loss": 0.4753,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.58503113852544,
"learning_rate": 3.279548219581317e-05,
"loss": 0.4815,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.5205249882087247,
"learning_rate": 3.275196553714379e-05,
"loss": 0.4915,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.6060885895533507,
"learning_rate": 3.270834691409778e-05,
"loss": 0.4834,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.5268512254297043,
"learning_rate": 3.266462667544849e-05,
"loss": 0.4749,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.5585192504980981,
"learning_rate": 3.262080517078178e-05,
"loss": 0.4838,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.543861883008989,
"learning_rate": 3.257688275049323e-05,
"loss": 0.5013,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.5239420002058491,
"learning_rate": 3.253285976578535e-05,
"loss": 0.4916,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.49830298593617506,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.4989,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.5581028558782758,
"learning_rate": 3.244451351193935e-05,
"loss": 0.5025,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.5076243021179011,
"learning_rate": 3.240019094921551e-05,
"loss": 0.4691,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.5122576391483217,
"learning_rate": 3.235576923489527e-05,
"loss": 0.4898,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.52489476226409,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.4965,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.5321744053653977,
"learning_rate": 3.226662977303494e-05,
"loss": 0.4972,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.5175836328642819,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.4781,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.4900548948539292,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.4849,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.4989285594957582,
"learning_rate": 3.21321858487569e-05,
"loss": 0.497,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.471767246957906,
"learning_rate": 3.208717671149931e-05,
"loss": 0.4772,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.5274486199001006,
"learning_rate": 3.204207092549882e-05,
"loss": 0.4801,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.4549013863531439,
"learning_rate": 3.19968688514201e-05,
"loss": 0.4783,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.5259626920048207,
"learning_rate": 3.195157085069774e-05,
"loss": 0.4886,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.4834074290661155,
"learning_rate": 3.190617728553332e-05,
"loss": 0.5075,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.5919932754691668,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.4987,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.46627093661125424,
"learning_rate": 3.181510491450254e-05,
"loss": 0.4956,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.5352421403542948,
"learning_rate": 3.176942683684842e-05,
"loss": 0.4846,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.5046890474270365,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.4868,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.48909655392274665,
"learning_rate": 3.167778872346327e-05,
"loss": 0.5024,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.46612866931762825,
"learning_rate": 3.163182942046815e-05,
"loss": 0.4724,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.49261596915816663,
"learning_rate": 3.158577710967494e-05,
"loss": 0.4749,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.4787420278825731,
"learning_rate": 3.153963215931669e-05,
"loss": 0.4937,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.510976905692121,
"learning_rate": 3.14933949383672e-05,
"loss": 0.4926,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.5228517803056089,
"learning_rate": 3.144706581653805e-05,
"loss": 0.4766,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.50250740283449,
"learning_rate": 3.140064516427566e-05,
"loss": 0.4944,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.6558287012183577,
"learning_rate": 3.135413335275833e-05,
"loss": 0.4908,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.5073631020705645,
"learning_rate": 3.130753075389327e-05,
"loss": 0.4871,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.5875574801778404,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.4617,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.4916975444859896,
"learning_rate": 3.121405468537547e-05,
"loss": 0.4761,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.5295943852005267,
"learning_rate": 3.116718196315488e-05,
"loss": 0.4867,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.5112609374417397,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.4722,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.4973755779047426,
"learning_rate": 3.107316901675252e-05,
"loss": 0.4835,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.5570433462603714,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.4681,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.4813958731159753,
"learning_rate": 3.097880190800057e-05,
"loss": 0.5077,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.5906911421022619,
"learning_rate": 3.093148648549788e-05,
"loss": 0.4843,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.476975281849464,
"learning_rate": 3.088408365512055e-05,
"loss": 0.4927,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.6229598715094249,
"learning_rate": 3.083659379590034e-05,
"loss": 0.4914,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.4848637910001278,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.5004,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.6813398815776044,
"learning_rate": 3.074135451053475e-05,
"loss": 0.4771,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.5136919708072767,
"learning_rate": 3.069360584592015e-05,
"loss": 0.4918,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.5498952618927222,
"learning_rate": 3.064577167551816e-05,
"loss": 0.4751,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.505155303608714,
"learning_rate": 3.059785238180953e-05,
"loss": 0.4791,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.5317149729867828,
"learning_rate": 3.054984834795566e-05,
"loss": 0.4975,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.5063416001575178,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.4759,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.5389402056576629,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.4863,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.5187044973370801,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.5083,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.5251642770331854,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.4818,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.5297244496216363,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.4902,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.5343947040641502,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.4997,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.5450038612549828,
"learning_rate": 3.021147971398632e-05,
"loss": 0.4954,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.4742594256669605,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.4789,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.5028007788458656,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.4949,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.46522096891090975,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.4758,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.5120500751606087,
"learning_rate": 3.001632138223332e-05,
"loss": 0.4749,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.6620065339782488,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.3438,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 1.1335099414020762,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.3639,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.649594561895499,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.3486,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.7344045927476281,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.3606,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.5945472446106459,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.3468,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.5861706653530124,
"learning_rate": 2.972118904869376e-05,
"loss": 0.3497,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.5751248877236812,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.3477,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.564287972385071,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.326,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.5780458363907993,
"learning_rate": 2.95725682050846e-05,
"loss": 0.3488,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.5692195826308518,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.3448,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.5694023240230277,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.3227,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.5485018468112803,
"learning_rate": 2.942325848714505e-05,
"loss": 0.3319,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.5086240029531337,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.3425,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.6162739636563547,
"learning_rate": 2.932334127129686e-05,
"loss": 0.3568,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.5135838162207526,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.3266,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.6404134788033442,
"learning_rate": 2.922312585929131e-05,
"loss": 0.3394,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.5001539242146019,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.3178,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.586445301059122,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.3478,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.546638615045592,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.3354,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.5392335225733723,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.3305,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.5453272133923841,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.3331,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.5433973209665044,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.3567,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.5073524123066606,
"learning_rate": 2.887006998351372e-05,
"loss": 0.3447,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.56190925475955,
"learning_rate": 2.881934649605535e-05,
"loss": 0.3391,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.5146886317595698,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.3458,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.568200298912905,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.3491,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.5058099584831386,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.3381,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.604137119261305,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.3415,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.5187538060525041,
"learning_rate": 2.856467940121025e-05,
"loss": 0.3435,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.5210570724324869,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.3509,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.5537573837980728,
"learning_rate": 2.846233032790149e-05,
"loss": 0.3364,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.5376404470948608,
"learning_rate": 2.84110540896642e-05,
"loss": 0.3253,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.6323006373519799,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.3387,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.5026151278345214,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.3409,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.6339522041731952,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.3475,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.5508365760483179,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.3425,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.5359994467642364,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.3478,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.5588262056644517,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.3316,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.5792712309712402,
"learning_rate": 2.805026032173773e-05,
"loss": 0.3531,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.5434268454193882,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.3421,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.6482515775534247,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.3655,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.6255218519421463,
"learning_rate": 2.7894660620408e-05,
"loss": 0.3425,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.5617401151771508,
"learning_rate": 2.784266724883173e-05,
"loss": 0.3488,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.5266741056650176,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.3379,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.5378441349914665,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.3515,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.47795795728537244,
"learning_rate": 2.768631254100716e-05,
"loss": 0.3352,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.5110025977591313,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.3388,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.4740713594937846,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.3489,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.476110669147143,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.321,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.48764454310226396,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.3511,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.5402964950713682,
"learning_rate": 2.742449775134085e-05,
"loss": 0.3328,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.4686102620879234,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.3496,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.5624507801219046,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.338,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.5527720951951689,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.3512,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.5286014798310574,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.3377,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.6156875559416892,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.3467,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.52399719197147,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.3459,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.5748827306544487,
"learning_rate": 2.705547548999996e-05,
"loss": 0.3436,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.4920149465218204,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.3536,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.49830151168771303,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.3601,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.5747673521381317,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.3615,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.4840757113520605,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.3573,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.5372752152054912,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.3537,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.5109314832457056,
"learning_rate": 2.67369650552155e-05,
"loss": 0.3582,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.5218068357252481,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.3423,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.5218095131905658,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.3424,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.5303107842447794,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.3428,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.5401824759164451,
"learning_rate": 2.652354173593614e-05,
"loss": 0.3504,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.46900054195379814,
"learning_rate": 2.647005443331317e-05,
"loss": 0.3471,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.581378212128943,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.3468,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.49604226584927624,
"learning_rate": 2.636292505301542e-05,
"loss": 0.3435,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.5298163103458983,
"learning_rate": 2.630928383194431e-05,
"loss": 0.3357,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.5094043891988934,
"learning_rate": 2.625559216200781e-05,
"loss": 0.342,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.5034565061416533,
"learning_rate": 2.620185047252308e-05,
"loss": 0.3301,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.5576346925884523,
"learning_rate": 2.614805919320723e-05,
"loss": 0.3399,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.499356192132404,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.3466,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.5702988162276218,
"learning_rate": 2.60403295859298e-05,
"loss": 0.349,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.4743354431540526,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.3512,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.5806825774883326,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.3455,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.48447501661842957,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.3461,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.5017926914208367,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.3538,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.5374702701924794,
"learning_rate": 2.577016790136453e-05,
"loss": 0.3549,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.503251694337065,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.3411,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.5953260367525278,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.3356,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.5023231634075906,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.3396,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.5695745835693699,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.3451,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.5067379090824179,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.3302,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.5539328091324991,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.3484,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.5292518565583003,
"learning_rate": 2.539001591449889e-05,
"loss": 0.3629,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.4866006579071995,
"learning_rate": 2.53355326151989e-05,
"loss": 0.3405,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.5021038187055724,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.334,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.502925869229295,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.3339,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.4693164328758545,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.3516,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.5175915148616609,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.3366,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.4995939957318962,
"learning_rate": 2.506248490185627e-05,
"loss": 0.3302,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.4984370196565979,
"learning_rate": 2.500775217241923e-05,
"loss": 0.3566,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.4690245810114145,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.3434,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.5327994958630091,
"learning_rate": 2.4898167025948e-05,
"loss": 0.3343,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.5226607203363901,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.3685,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.5071810509104674,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.3566,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.47140344461019146,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.3437,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.4896122510385392,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.361,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.5462374372763148,
"learning_rate": 2.46235264422994e-05,
"loss": 0.3295,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.48548366079928174,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.3408,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.5262147829859056,
"learning_rate": 2.451340834511592e-05,
"loss": 0.3547,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.5114993096245907,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.3477,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.552496484750578,
"learning_rate": 2.440314589184958e-05,
"loss": 0.3403,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.5180487229045235,
"learning_rate": 2.434796163357541e-05,
"loss": 0.3334,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.48650128704138657,
"learning_rate": 2.429274260911588e-05,
"loss": 0.34,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.5320292898143328,
"learning_rate": 2.423748926000084e-05,
"loss": 0.3497,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.5305792610743953,
"learning_rate": 2.418220202803459e-05,
"loss": 0.3451,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.5275289277695794,
"learning_rate": 2.412688135529237e-05,
"loss": 0.3564,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.525049647979377,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.3576,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.5089531147977431,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.3512,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.4915918868750602,
"learning_rate": 2.396072311715187e-05,
"loss": 0.3424,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.48883243947107274,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.324,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.5683223670885122,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.3584,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.47654849237888486,
"learning_rate": 2.379427985200223e-05,
"loss": 0.3428,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.49183030138051714,
"learning_rate": 2.373873749394972e-05,
"loss": 0.3465,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.4801889006637134,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.3603,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.4696748656971157,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.3474,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.46776433727366273,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.3636,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.5242592134496635,
"learning_rate": 2.351627355800233e-05,
"loss": 0.3353,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.48214165492236405,
"learning_rate": 2.346058617160256e-05,
"loss": 0.3337,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.46996159284955596,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.3578,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.509202045962556,
"learning_rate": 2.334912883203469e-05,
"loss": 0.3492,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.47650126133707643,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.3322,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.46713271149477587,
"learning_rate": 2.323756437450136e-05,
"loss": 0.3483,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.48847681225209566,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.3576,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.44385653312253187,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.353,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.47755583213847363,
"learning_rate": 2.307002464849052e-05,
"loss": 0.3392,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.48010538865775787,
"learning_rate": 2.30141283818846e-05,
"loss": 0.3361,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.4695089055266705,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.3511,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.4675031547422366,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.3371,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.5019400371204509,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.3443,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.47211875805851217,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.3509,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.48709303239746155,
"learning_rate": 2.273429448107257e-05,
"loss": 0.342,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.4863460169440333,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.3402,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.511602308570585,
"learning_rate": 2.262220474347102e-05,
"loss": 0.3392,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.4764790098776345,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.3464,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.4872978765818646,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.3407,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.5080862929114529,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.3365,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.4938618031957982,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.3473,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.5550922235142868,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.3414,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.5709457205367366,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.3426,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.5644149816160069,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.3388,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.49686313269316845,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.3479,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.5482924203445034,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.364,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.49286204320710447,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.3502,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.4973314687273963,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.3662,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.47144470477118594,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.3334,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.47343892596986076,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.3508,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.5484879475039547,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.3381,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.47291397921572503,
"learning_rate": 2.17791064266051e-05,
"loss": 0.3228,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.5267830784468017,
"learning_rate": 2.172276926861741e-05,
"loss": 0.3577,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.5461056752854688,
"learning_rate": 2.166641833541255e-05,
"loss": 0.3348,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.4355395305130441,
"learning_rate": 2.161005407757109e-05,
"loss": 0.3367,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.5483996198389219,
"learning_rate": 2.155367694578013e-05,
"loss": 0.35,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.5188820568716528,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.3411,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.49994696383174936,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.3391,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.4724118210701204,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.356,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.4901351277996137,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.3453,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.4596612850697064,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.3475,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.4744158373190256,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.3388,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.48289786802312956,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.3461,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.5100459296906398,
"learning_rate": 2.110225054402733e-05,
"loss": 0.3478,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.43991132716762327,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.344,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.5155537226035414,
"learning_rate": 2.09892967807724e-05,
"loss": 0.3459,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.4773438500958207,
"learning_rate": 2.093280780774835e-05,
"loss": 0.3456,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.4813941742015794,
"learning_rate": 2.087631137601793e-05,
"loss": 0.3433,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.4520806936965106,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.3618,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.47107123105527593,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.3587,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.4571423799739353,
"learning_rate": 2.070678184630458e-05,
"loss": 0.3483,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.4653983963132758,
"learning_rate": 2.06502600977305e-05,
"loss": 0.333,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.5117586595657848,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.3405,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.5301392312444082,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.3564,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.46921912018024925,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.3337,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.547498151252228,
"learning_rate": 2.042412562887594e-05,
"loss": 0.335,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.4618292680127445,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.3434,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.5007272471781119,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.3468,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.44709427258369594,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.3473,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.514013141278628,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.3489,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.4660880617335241,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.34,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.5032599714371393,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.3245,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.5015658262799269,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.336,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.45437390246772474,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.35,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.5118575908143941,
"learning_rate": 1.99151687695709e-05,
"loss": 0.3488,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.4463646708950846,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.3458,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.49908556902663753,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.3458,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.4612416031407178,
"learning_rate": 1.974551241345447e-05,
"loss": 0.3574,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.4714333163012124,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.3348,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.46450420176618273,
"learning_rate": 1.963241759676897e-05,
"loss": 0.3518,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.4672212953889052,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.3304,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.48313989729031515,
"learning_rate": 1.951933453677638e-05,
"loss": 0.3395,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.492855898656183,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.3357,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.46839894954705247,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.3451,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.518710428049196,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.3279,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.4495514909222075,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.3529,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.4777887413750957,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.3528,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.45078700934386634,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.3512,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.46625717406214745,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.3436,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.4516147724987438,
"learning_rate": 1.906719219225165e-05,
"loss": 0.3281,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.45045593752738455,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.352,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.45131328301627177,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.3227,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.5027361770317768,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.3357,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.45814822053348087,
"learning_rate": 1.884128556891707e-05,
"loss": 0.3482,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.4759742162195423,
"learning_rate": 1.878483094691119e-05,
"loss": 0.3539,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.5301390370627922,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.3461,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.4706890629539606,
"learning_rate": 1.867195130360947e-05,
"loss": 0.3441,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.491854366852641,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.3495,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.5258327110234694,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.3424,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.45539207788298086,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.3495,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.46018136546168764,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.3516,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.4968459089252833,
"learning_rate": 1.838994592242891e-05,
"loss": 0.3536,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.4861870955860603,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.3317,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.4853351695515632,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.3128,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.5393291275856827,
"learning_rate": 1.82208935733949e-05,
"loss": 0.346,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.4489411066275738,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.3416,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.49082302857636056,
"learning_rate": 1.810826238483892e-05,
"loss": 0.3315,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.49889363381291657,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.3532,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.45542325127055855,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.3475,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.45212851778981583,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.3511,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.4636742415650472,
"learning_rate": 1.788318512320838e-05,
"loss": 0.3465,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.4477608669963357,
"learning_rate": 1.782695699829051e-05,
"loss": 0.3531,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.4572833130497358,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.3328,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.47922909902053673,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.3399,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.48205980994572006,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.3355,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.49248968793797165,
"learning_rate": 1.76022227484387e-05,
"loss": 0.3323,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.46316893632095973,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.3575,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.49292361316567523,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.3361,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.4532124804482044,
"learning_rate": 1.743387180005809e-05,
"loss": 0.3453,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.5023844096924255,
"learning_rate": 1.737779525652899e-05,
"loss": 0.3424,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.4830443685014541,
"learning_rate": 1.732173968007914e-05,
"loss": 0.3445,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.4972658158832811,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.3563,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.44566323974421446,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.3524,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.4667388796124335,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.3378,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.46972554215369033,
"learning_rate": 1.709773600686015e-05,
"loss": 0.3528,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.4486630939276104,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.3397,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.4343203924927934,
"learning_rate": 1.69858716181154e-05,
"loss": 0.3451,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.4695216936185801,
"learning_rate": 1.692997535150948e-05,
"loss": 0.3354,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.5035399923425705,
"learning_rate": 1.687410363273884e-05,
"loss": 0.3596,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.4786138752004407,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.3496,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.44928687374127657,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.3479,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.4822958367598581,
"learning_rate": 1.67066402299235e-05,
"loss": 0.3538,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.4583602282370906,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.3541,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.44630544524786137,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.341,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.48622201456381786,
"learning_rate": 1.653941382839745e-05,
"loss": 0.3492,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.715595649870702,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.2225,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.44906959320899015,
"learning_rate": 1.642806717162757e-05,
"loss": 0.2347,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.7948680350470639,
"learning_rate": 1.637243646233718e-05,
"loss": 0.2372,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.5730935953531575,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.2035,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.5623620784071174,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.2328,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.618901036179523,
"learning_rate": 1.620572014799777e-05,
"loss": 0.2292,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.5532952555102304,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.231,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.47542482532401376,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.2368,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.5160367355413007,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.1979,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.4969796489568406,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.2101,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.49725627115990395,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.2252,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.4821346548455903,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.2077,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.473169806877361,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.2191,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.4731908915125825,
"learning_rate": 1.576251073999917e-05,
"loss": 0.2033,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.4655913428679166,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.1971,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.4755171501176994,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.2007,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.4455363173464476,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.2177,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.49351197110379436,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.207,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.44504510381535867,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.2174,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.4449760212698243,
"learning_rate": 1.543151434154755e-05,
"loss": 0.2234,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.41968167500618453,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.225,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.4669264750805211,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.2171,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.479771458939276,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.2071,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.44739228893717226,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.2124,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.4764542437007891,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.2158,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.45954145326082263,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.2032,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.48219747030849575,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.219,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.45906274897506544,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.2068,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.44726629876806945,
"learning_rate": 1.493751509814374e-05,
"loss": 0.2115,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.46338955737745396,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.228,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.46305270045319324,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.2184,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.44884673623518645,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.2046,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.4514354065141202,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.2117,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.4485092071978243,
"learning_rate": 1.466446738480111e-05,
"loss": 0.2163,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.447401505370989,
"learning_rate": 1.460998408550112e-05,
"loss": 0.1967,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.46466066237822634,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.2046,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.4441932890204616,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.2186,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.44413364808083244,
"learning_rate": 1.444679451900437e-05,
"loss": 0.2175,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.42920685231115474,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.2214,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.4615080750319658,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.2195,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.4616449386966017,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.2113,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.44585657554173785,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.2185,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.4425892660279131,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.2183,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.44309206354448893,
"learning_rate": 1.412162598317602e-05,
"loss": 0.2105,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.4356565235551029,
"learning_rate": 1.406759321421907e-05,
"loss": 0.2325,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.4447737198221389,
"learning_rate": 1.401360788062871e-05,
"loss": 0.2155,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.43904056530966123,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.2321,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.45726350215740974,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.207,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.47578084869610193,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.2237,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.44630391223037097,
"learning_rate": 1.379814952747693e-05,
"loss": 0.214,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.45530541790894236,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.2162,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.4528628913669272,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.2064,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.46008788883808804,
"learning_rate": 1.363707494698459e-05,
"loss": 0.21,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.4774380298050798,
"learning_rate": 1.358348460369265e-05,
"loss": 0.2065,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.4309069166978016,
"learning_rate": 1.352994556668684e-05,
"loss": 0.2083,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.4452205033657286,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.2212,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.48351216699054,
"learning_rate": 1.342302312350678e-05,
"loss": 0.2088,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.4367123386362796,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.2237,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.4547146932302134,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.194,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.46103740352693506,
"learning_rate": 1.326303494478451e-05,
"loss": 0.2156,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.44826627279128956,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.1983,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.41568613071192756,
"learning_rate": 1.315664479122898e-05,
"loss": 0.2094,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.4799551944179489,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.2124,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.443693352276919,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.2187,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.4608991900966141,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.1892,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.44595891772797175,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.223,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.433840920352011,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.2111,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.449964319802965,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.2193,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.4520819384110775,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.2077,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.43851894824516685,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.2134,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.46849965541588734,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.2187,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.45870106635955454,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.2082,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.4581388230336086,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.1949,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.7830113538630371,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.2177,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.47754944460361026,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.22,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.4539887664013291,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.2179,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.4468899637585918,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.2245,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.47249279474867556,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.1982,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.45553915350750396,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.2255,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.4436845313679605,
"learning_rate": 1.220938883250288e-05,
"loss": 0.2402,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.46157100680928986,
"learning_rate": 1.215733275116827e-05,
"loss": 0.2047,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.4523192654500016,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.2202,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.4835794680182389,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.2042,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.45723052524904595,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.1938,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.4752193307610503,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.2287,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.49624065395302674,
"learning_rate": 1.189800129803139e-05,
"loss": 0.2017,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.4412120757915916,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.2129,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.44650092590157225,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.2173,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.47574472757891156,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.2341,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.44025196273814793,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.209,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.4538575059024162,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.2127,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.4652143031883985,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.204,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.4605045294926187,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.1954,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.4441735797188403,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.234,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.4062339232938583,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.2261,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.4499193968387828,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.202,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.46245716062133285,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.2104,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.43653439042592784,
"learning_rate": 1.128231163049258e-05,
"loss": 0.2288,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.4512247862299901,
"learning_rate": 1.123144751066259e-05,
"loss": 0.2048,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.4744896913566861,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.2122,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.42814535347832094,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.213,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.44442884743480754,
"learning_rate": 1.107927745387114e-05,
"loss": 0.1972,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.4463074695409709,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.2093,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.43934298856946485,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.2153,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.4300823647620388,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.2125,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.44657684281321225,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.1979,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.4560286512990139,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.2135,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.5764419771170909,
"learning_rate": 1.077687414070869e-05,
"loss": 0.1946,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.44235879321563937,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.1925,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.4564356815010931,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.2191,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.4513031814958925,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.208,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.4396570323305813,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.2103,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.442114734111564,
"learning_rate": 1.052689572753247e-05,
"loss": 0.2058,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.4390321241839422,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.2124,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.44659043665471476,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.2036,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.4237111918926671,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.1992,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.45913341729932317,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.2202,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.4563123297863082,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.2259,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.4400732485922006,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.1951,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.44657831122563196,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.2004,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.4363284847813842,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.2199,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.4543361712047089,
"learning_rate": 1.008173988718005e-05,
"loss": 0.2113,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.4443257567076496,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.2238,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.4449361698272137,
"learning_rate": 9.983678617766677e-06,
"loss": 0.1986,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.417695811426032,
"learning_rate": 9.934767922553788e-06,
"loss": 0.2048,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.43308115752499116,
"learning_rate": 9.885937708668371e-06,
"loss": 0.2047,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.45916969118161555,
"learning_rate": 9.837188366555524e-06,
"loss": 0.2095,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.44410937542820095,
"learning_rate": 9.788520286013691e-06,
"loss": 0.1946,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.42262756800574314,
"learning_rate": 9.739933856191537e-06,
"loss": 0.2092,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.434859142707035,
"learning_rate": 9.691429465584866e-06,
"loss": 0.2051,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.43391532053394455,
"learning_rate": 9.643007502033492e-06,
"loss": 0.1979,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.4217011953270413,
"learning_rate": 9.594668352718152e-06,
"loss": 0.2069,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.44280734187109844,
"learning_rate": 9.546412404157383e-06,
"loss": 0.2137,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.43118963701516,
"learning_rate": 9.498240042204471e-06,
"loss": 0.2155,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.44486384142448315,
"learning_rate": 9.450151652044339e-06,
"loss": 0.201,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.4505422117819547,
"learning_rate": 9.402147618190472e-06,
"loss": 0.2145,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.4625615397102745,
"learning_rate": 9.354228324481841e-06,
"loss": 0.2207,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.44577437704371353,
"learning_rate": 9.306394154079852e-06,
"loss": 0.1992,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.4379821652979295,
"learning_rate": 9.258645489465254e-06,
"loss": 0.2024,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.45336303690795393,
"learning_rate": 9.21098271243509e-06,
"loss": 0.2061,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.4362069957589626,
"learning_rate": 9.163406204099659e-06,
"loss": 0.2176,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.4301225969080896,
"learning_rate": 9.115916344879452e-06,
"loss": 0.2197,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.43241920765162783,
"learning_rate": 9.068513514502126e-06,
"loss": 0.2014,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.44096769632081206,
"learning_rate": 9.021198091999439e-06,
"loss": 0.2181,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.42326562781339716,
"learning_rate": 8.973970455704251e-06,
"loss": 0.2076,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.4532953449817347,
"learning_rate": 8.926830983247495e-06,
"loss": 0.1984,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.4509587825116214,
"learning_rate": 8.879780051555116e-06,
"loss": 0.192,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.4395810639244242,
"learning_rate": 8.83281803684512e-06,
"loss": 0.2079,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.45678029531426956,
"learning_rate": 8.785945314624532e-06,
"loss": 0.19,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.43962115269041324,
"learning_rate": 8.739162259686385e-06,
"loss": 0.2079,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.4273511326962007,
"learning_rate": 8.692469246106734e-06,
"loss": 0.2091,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.42754177152149503,
"learning_rate": 8.645866647241677e-06,
"loss": 0.2029,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.4301206450038315,
"learning_rate": 8.599354835724348e-06,
"loss": 0.2104,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.4425119266725608,
"learning_rate": 8.552934183461958e-06,
"loss": 0.2114,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.43242173817078966,
"learning_rate": 8.506605061632805e-06,
"loss": 0.2002,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.433985700656666,
"learning_rate": 8.460367840683308e-06,
"loss": 0.2359,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.4430371816615738,
"learning_rate": 8.414222890325065e-06,
"loss": 0.2201,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.43731705784155855,
"learning_rate": 8.368170579531855e-06,
"loss": 0.2056,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.41069342890753535,
"learning_rate": 8.322211276536733e-06,
"loss": 0.2279,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.4292697606025428,
"learning_rate": 8.27634534882907e-06,
"loss": 0.2075,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.4415771165212286,
"learning_rate": 8.230573163151585e-06,
"loss": 0.1961,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.40582537371351024,
"learning_rate": 8.184895085497466e-06,
"loss": 0.2173,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.4300118408439246,
"learning_rate": 8.139311481107401e-06,
"loss": 0.2314,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.4637670786031287,
"learning_rate": 8.093822714466686e-06,
"loss": 0.2027,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.42307913929593244,
"learning_rate": 8.048429149302271e-06,
"loss": 0.2201,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.42650428516066535,
"learning_rate": 8.003131148579902e-06,
"loss": 0.2149,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.42750411082100004,
"learning_rate": 7.957929074501182e-06,
"loss": 0.2064,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.4190031378255664,
"learning_rate": 7.912823288500695e-06,
"loss": 0.2004,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.4313024244721021,
"learning_rate": 7.867814151243102e-06,
"loss": 0.2214,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.4646973538047448,
"learning_rate": 7.822902022620267e-06,
"loss": 0.2005,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.41654562937774664,
"learning_rate": 7.778087261748377e-06,
"loss": 0.2092,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.4419995622937455,
"learning_rate": 7.733370226965056e-06,
"loss": 0.203,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.42178763920519474,
"learning_rate": 7.688751275826527e-06,
"loss": 0.2181,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.43858389442180445,
"learning_rate": 7.644230765104735e-06,
"loss": 0.1846,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.4236142351051262,
"learning_rate": 7.599809050784502e-06,
"loss": 0.2211,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.4196916343741567,
"learning_rate": 7.555486488060661e-06,
"loss": 0.2229,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.4368162606613971,
"learning_rate": 7.511263431335249e-06,
"loss": 0.2091,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.4303990500331555,
"learning_rate": 7.467140234214656e-06,
"loss": 0.2211,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.41305452206179866,
"learning_rate": 7.423117249506777e-06,
"loss": 0.2234,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.42920388211184407,
"learning_rate": 7.379194829218231e-06,
"loss": 0.2056,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.4154079180079675,
"learning_rate": 7.335373324551518e-06,
"loss": 0.2066,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.4306317504253931,
"learning_rate": 7.291653085902224e-06,
"loss": 0.2335,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.44048787171655585,
"learning_rate": 7.248034462856213e-06,
"loss": 0.2003,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.44787855666212706,
"learning_rate": 7.20451780418683e-06,
"loss": 0.2055,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.4504520001994488,
"learning_rate": 7.161103457852128e-06,
"loss": 0.1995,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.44725091419325236,
"learning_rate": 7.117791770992046e-06,
"loss": 0.2043,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.44194617138222175,
"learning_rate": 7.074583089925688e-06,
"loss": 0.1918,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.42363612885599156,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.2284,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.42873814458247933,
"learning_rate": 6.988476126329604e-06,
"loss": 0.2079,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.4290590861825237,
"learning_rate": 6.945578532308857e-06,
"loss": 0.2024,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.4018892222984421,
"learning_rate": 6.902785321094301e-06,
"loss": 0.2175,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.4260239420236521,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.2199,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.4222400881611877,
"learning_rate": 6.817513414939892e-06,
"loss": 0.2153,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.45271476327838484,
"learning_rate": 6.775035401831924e-06,
"loss": 0.2086,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.4122522795179344,
"learning_rate": 6.732663135188473e-06,
"loss": 0.2194,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.44803985928300305,
"learning_rate": 6.690396953817055e-06,
"loss": 0.1992,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.42835169541785184,
"learning_rate": 6.648237195676927e-06,
"loss": 0.1877,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.4033879333587172,
"learning_rate": 6.606184197876395e-06,
"loss": 0.2238,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.42279810616551544,
"learning_rate": 6.564238296670109e-06,
"loss": 0.2037,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.43373278619876643,
"learning_rate": 6.52239982745638e-06,
"loss": 0.2126,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.44734336263432484,
"learning_rate": 6.480669124774484e-06,
"loss": 0.21,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.4084681665114773,
"learning_rate": 6.439046522302013e-06,
"loss": 0.2021,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.41946372613485633,
"learning_rate": 6.397532352852191e-06,
"loss": 0.2051,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.40392013010755307,
"learning_rate": 6.356126948371202e-06,
"loss": 0.232,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.42286872143486515,
"learning_rate": 6.314830639935559e-06,
"loss": 0.2147,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.41379767608992424,
"learning_rate": 6.273643757749448e-06,
"loss": 0.1971,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.42427290194518474,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.1862,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.4320054076510458,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.1979,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.42411657482704784,
"learning_rate": 6.150742957589784e-06,
"loss": 0.1954,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.4246430836182492,
"learning_rate": 6.109997064904774e-06,
"loss": 0.2148,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.43602526825038074,
"learning_rate": 6.069362236313092e-06,
"loss": 0.1922,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.42319128744495654,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.2029,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.3917570592832134,
"learning_rate": 5.988427070179108e-06,
"loss": 0.2049,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.45207486831314064,
"learning_rate": 5.948127379792234e-06,
"loss": 0.2254,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.4131726397583107,
"learning_rate": 5.90794004780439e-06,
"loss": 0.2147,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.41120554013799493,
"learning_rate": 5.867865395552405e-06,
"loss": 0.2243,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.4190719994712867,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.2108,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.4205272714353976,
"learning_rate": 5.788055411095892e-06,
"loss": 0.2052,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.4057302977771638,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.218,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.4208812333692303,
"learning_rate": 5.708699979051613e-06,
"loss": 0.1894,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.44414068817529484,
"learning_rate": 5.66919351390751e-06,
"loss": 0.1835,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.4105298460870282,
"learning_rate": 5.629801637510188e-06,
"loss": 0.1996,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.4173343285103676,
"learning_rate": 5.590524664836043e-06,
"loss": 0.2053,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.4222585606641241,
"learning_rate": 5.551362909942708e-06,
"loss": 0.2139,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.4283114864753907,
"learning_rate": 5.51231668596653e-06,
"loss": 0.2158,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.40657339125837394,
"learning_rate": 5.47338630512007e-06,
"loss": 0.2199,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.41493176788101394,
"learning_rate": 5.434572078689617e-06,
"loss": 0.2244,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.4259702742555211,
"learning_rate": 5.395874317032694e-06,
"loss": 0.2232,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.4130855103603863,
"learning_rate": 5.357293329575573e-06,
"loss": 0.2259,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.40829552183895795,
"learning_rate": 5.318829424810797e-06,
"loss": 0.2179,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.4441264962047589,
"learning_rate": 5.280482910294728e-06,
"loss": 0.2189,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.43112958908005533,
"learning_rate": 5.242254092645078e-06,
"loss": 0.1846,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.40529098391171553,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.2066,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.42597902571938495,
"learning_rate": 5.166150769707892e-06,
"loss": 0.2084,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.421287072975049,
"learning_rate": 5.128276872940503e-06,
"loss": 0.1972,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.4126470597443414,
"learning_rate": 5.090521890074953e-06,
"loss": 0.2005,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.4064143329588678,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.203,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.44044147048037197,
"learning_rate": 5.015369872647442e-06,
"loss": 0.2018,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.4223526965080778,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.2078,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.41918470076873127,
"learning_rate": 4.940697121074675e-06,
"loss": 0.2056,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.4040885810940347,
"learning_rate": 4.903541216934864e-06,
"loss": 0.2041,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.4264523780479886,
"learning_rate": 4.8665060236772e-06,
"loss": 0.1938,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.432423751717881,
"learning_rate": 4.829591837434102e-06,
"loss": 0.208,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.42733370463331555,
"learning_rate": 4.792798953370424e-06,
"loss": 0.2109,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.4209529598170311,
"learning_rate": 4.756127665681083e-06,
"loss": 0.1956,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.40675336911429394,
"learning_rate": 4.719578267588725e-06,
"loss": 0.2021,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.41637471049681507,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.1838,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.5785491914423727,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.1443,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.5230152584280603,
"learning_rate": 4.610664328486485e-06,
"loss": 0.1225,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.4180179077957587,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1351,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.39546203761735615,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.1365,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.44794999675161906,
"learning_rate": 4.502857857942204e-06,
"loss": 0.1258,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.5506229632972923,
"learning_rate": 4.467169815100549e-06,
"loss": 0.1252,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.5343121665966415,
"learning_rate": 4.431605972355137e-06,
"loss": 0.1165,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.5193684864587037,
"learning_rate": 4.396166614073516e-06,
"loss": 0.1148,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.42804430511944325,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1166,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.42649624661163266,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1229,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.46335947866265736,
"learning_rate": 4.29059827474257e-06,
"loss": 0.1313,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.49400173752096505,
"learning_rate": 4.255659678050063e-06,
"loss": 0.1435,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.4354705448540904,
"learning_rate": 4.22084697268323e-06,
"loss": 0.1375,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.39802643003267424,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1311,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.37577790201129296,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1314,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.38730025484217784,
"learning_rate": 4.11716698310493e-06,
"loss": 0.1367,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.39339048855221276,
"learning_rate": 4.082860616555848e-06,
"loss": 0.1144,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.45706689078657126,
"learning_rate": 4.048681523029125e-06,
"loss": 0.1471,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.4107689194923916,
"learning_rate": 4.014629975819882e-06,
"loss": 0.1165,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.39618745122658955,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1244,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.40077888998030314,
"learning_rate": 3.94691060843287e-06,
"loss": 0.1462,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.376169926435522,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1274,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.37953534706670694,
"learning_rate": 3.879704680319503e-06,
"loss": 0.1198,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.38073142310911695,
"learning_rate": 3.846294928353471e-06,
"loss": 0.1231,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.3851587272420345,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.1239,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.3746902670369432,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.1309,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.3839729548126052,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.1167,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.37967857606728816,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1296,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.37816394452959534,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.1302,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.38389132015571076,
"learning_rate": 3.648558150665886e-06,
"loss": 0.1249,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.371512029945898,
"learning_rate": 3.616058102895075e-06,
"loss": 0.1402,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.3704488857034918,
"learning_rate": 3.583689060687079e-06,
"loss": 0.1293,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.37739137308583415,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.1171,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.36323796310433787,
"learning_rate": 3.51934502719796e-06,
"loss": 0.109,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.3881181307698154,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.1323,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.3607433494884644,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1319,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.37988158471398603,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.121,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.3690598021458168,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.1254,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.35378708772885814,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.1208,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.36870479275785645,
"learning_rate": 3.32948376099375e-06,
"loss": 0.139,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.36332821223640693,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.1144,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.3730588861002674,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.1257,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.3636602338866966,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1289,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.36408729086599634,
"learning_rate": 3.205572144561604e-06,
"loss": 0.1451,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.3470879924323275,
"learning_rate": 3.174929343376374e-06,
"loss": 0.1299,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.34968272689855945,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.1474,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.37500939749019757,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.1173,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.3483737962033575,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.1341,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.36478074075648975,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.1202,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.36668056902945734,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1371,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.38715498063896225,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.1352,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.36882261275630196,
"learning_rate": 2.964210239224696e-06,
"loss": 0.1212,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.3749896262058112,
"learning_rate": 2.934650470499869e-06,
"loss": 0.1292,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.3578793350309039,
"learning_rate": 2.905227155855128e-06,
"loss": 0.124,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.36267403299864825,
"learning_rate": 2.875940530558519e-06,
"loss": 0.124,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.3694378102554738,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.1343,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.3712024081097994,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.1229,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.38893502850106393,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1227,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.3655957417569477,
"learning_rate": 2.760165589921184e-06,
"loss": 0.1304,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.3569742904599074,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.1367,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.36452390521953704,
"learning_rate": 2.703104292151164e-06,
"loss": 0.129,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.360785634192623,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.1172,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.3596961784309719,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.1198,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.36747168090318016,
"learning_rate": 2.618550200289303e-06,
"loss": 0.111,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.38125416293319164,
"learning_rate": 2.59064316684412e-06,
"loss": 0.0994,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.3662363584370369,
"learning_rate": 2.562875338152264e-06,
"loss": 0.1229,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.36847211535641544,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1127,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.37180106567876753,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.1088,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.3712096382226989,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1365,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.3673125722202832,
"learning_rate": 2.45320049480108e-06,
"loss": 0.1233,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.3678508102790805,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.1225,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.35483694048045356,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1283,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.34403345882403985,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.1379,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.36869285306496663,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.1399,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.3498480377414309,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1319,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.3500531315728084,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.122,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.3720473745210559,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.1128,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.37187564936541173,
"learning_rate": 2.24059906434998e-06,
"loss": 0.1281,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.345240132278537,
"learning_rate": 2.214660694699091e-06,
"loss": 0.1225,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.36143355970648433,
"learning_rate": 2.188864536147355e-06,
"loss": 0.1224,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.3764320489931499,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.1188,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.3841058278999449,
"learning_rate": 2.137699676264171e-06,
"loss": 0.1337,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.36698799785543434,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.1355,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.3650998796971638,
"learning_rate": 2.08710612114857e-06,
"loss": 0.1104,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.3735390360192394,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.1129,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.35881754445257313,
"learning_rate": 2.0370854889762e-06,
"loss": 0.116,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.3813737829008485,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.1186,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.35596820522993017,
"learning_rate": 1.987639379598427e-06,
"loss": 0.12,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.37001967581772194,
"learning_rate": 1.963132265887919e-06,
"loss": 0.1233,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.358600388621373,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.1376,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.3936588178301786,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.122,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.3627038131626256,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.1172,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.37217070367564137,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.1249,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.35706334595800926,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1117,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.3802091966809692,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.1255,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.37905782424631396,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.12,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.3888719308821773,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.1143,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.3609704335466115,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.1134,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.3552743648025839,
"learning_rate": 1.726025117089718e-06,
"loss": 0.1235,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.3662711734008893,
"learning_rate": 1.70311497206598e-06,
"loss": 0.1186,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.35320921150890244,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.1138,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.3764819421493319,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.1157,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.38118320844834724,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.1137,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.3657700493128824,
"learning_rate": 1.612939220355272e-06,
"loss": 0.1274,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.3792604281642914,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1276,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.36834717819967255,
"learning_rate": 1.568732761074121e-06,
"loss": 0.1222,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.3795116774238506,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1281,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.36503317055211315,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.1103,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.3723509853254097,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.1044,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.3645015483368837,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.1197,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.3708941638516995,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.1236,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.35621076581606365,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.1246,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.37038930911471635,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.1262,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.3875826981405282,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.1073,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.3834955678767935,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.1225,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.362395608501794,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.1243,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.3776150930313762,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1143,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.3718640026769463,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.1218,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.3531387846728781,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.1166,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.37705271061173873,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1216,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.3761277359844243,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.1215,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.3704433783991667,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1335,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.36822901121585583,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1289,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.3556534903353581,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1384,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.3489627111036211,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.1149,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.37361401167785624,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.1179,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.3477847283669575,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1288,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.3768841276136566,
"learning_rate": 1.121666393356604e-06,
"loss": 0.132,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.36947345521480396,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.1351,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.3667356079655377,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.1372,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.3644424767261504,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1289,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.3841346165866125,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.1118,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.37567207232271976,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1283,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.37045413349083467,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1235,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.37190128000204636,
"learning_rate": 9.9465987684501e-07,
"loss": 0.1099,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.3385611677740173,
"learning_rate": 9.771228098856534e-07,
"loss": 0.0945,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.3720518544216269,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1336,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.35209142461026993,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1294,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.3549936320625652,
"learning_rate": 9.254248023435864e-07,
"loss": 0.1344,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.36559209182475627,
"learning_rate": 9.084969902482288e-07,
"loss": 0.1167,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.37253114888737093,
"learning_rate": 8.917218333021993e-07,
"loss": 0.1081,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.35745856529277037,
"learning_rate": 8.750994656392109e-07,
"loss": 0.1203,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.37694518551794526,
"learning_rate": 8.58630020171265e-07,
"loss": 0.1061,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.35831520945224965,
"learning_rate": 8.423136285876099e-07,
"loss": 0.1349,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.36882552365998555,
"learning_rate": 8.26150421353682e-07,
"loss": 0.1025,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.35794257202128194,
"learning_rate": 8.101405277100549e-07,
"loss": 0.1225,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.35433875494034367,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1361,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.37243231351523676,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1159,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.3696616748710537,
"learning_rate": 7.630320023322024e-07,
"loss": 0.1208,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.361646770556294,
"learning_rate": 7.476366309223903e-07,
"loss": 0.132,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.3774389003430557,
"learning_rate": 7.323952008970537e-07,
"loss": 0.117,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.36041600216136127,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1222,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.3690238857874115,
"learning_rate": 7.023746512481989e-07,
"loss": 0.1314,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.3544606324553892,
"learning_rate": 6.875957716681902e-07,
"loss": 0.1159,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.35308908076155904,
"learning_rate": 6.729713135577687e-07,
"loss": 0.1256,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.3554687469511732,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1375,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.35879326106870585,
"learning_rate": 6.441861282571115e-07,
"loss": 0.1243,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.3550521838189161,
"learning_rate": 6.300256312324471e-07,
"loss": 0.1278,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.36914390223865917,
"learning_rate": 6.160200160066776e-07,
"loss": 0.113,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.35669960410089047,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1159,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.36193292946392636,
"learning_rate": 5.884738776666044e-07,
"loss": 0.1288,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.3534226495040384,
"learning_rate": 5.749335748104878e-07,
"loss": 0.1049,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.3703927323004323,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1257,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.367693204672864,
"learning_rate": 5.483190430647289e-07,
"loss": 0.1113,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.3619696625423461,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1342,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.36499072751275313,
"learning_rate": 5.22326650565701e-07,
"loss": 0.1094,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.3552069633297617,
"learning_rate": 5.095640171043048e-07,
"loss": 0.1205,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.3629596405273338,
"learning_rate": 4.969572286496594e-07,
"loss": 0.1376,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.3706853279723823,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1258,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.3764543778082175,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1249,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.3688421343545412,
"learning_rate": 4.600729351261857e-07,
"loss": 0.1237,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.36879143344885873,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.12,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.3610190921828175,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1318,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.3527689577768133,
"learning_rate": 4.245948007320588e-07,
"loss": 0.1219,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.3620032405952352,
"learning_rate": 4.130816799401971e-07,
"loss": 0.1201,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.3666755524955893,
"learning_rate": 4.017251756252827e-07,
"loss": 0.1354,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.36280954111510666,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.1157,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.36056039988426947,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.1225,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.3527879709530482,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1288,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.35634441501360514,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1337,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.3759705596735846,
"learning_rate": 3.472950355547267e-07,
"loss": 0.1209,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.35080035568715057,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.1228,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.3629974330139459,
"learning_rate": 3.266223765343224e-07,
"loss": 0.1147,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.3586690991712178,
"learning_rate": 3.16521967718757e-07,
"loss": 0.1223,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.36084868237334244,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.1279,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.3569332643786679,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.1097,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.36983543590059154,
"learning_rate": 2.871653894244597e-07,
"loss": 0.1211,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.35628336999936894,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.1132,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.3730820293722409,
"learning_rate": 2.683823233205485e-07,
"loss": 0.1235,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.3498128993122315,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1144,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.37145418673689584,
"learning_rate": 2.502303499116754e-07,
"loss": 0.1279,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.36941539609368823,
"learning_rate": 2.413912051848666e-07,
"loss": 0.1313,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.3500125921475765,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.1166,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.35708019020748727,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.1294,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.3579435975907873,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.1255,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.3661449153591382,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1248,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.36489562136937526,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.1286,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.36997413657746375,
"learning_rate": 1.916764975761254e-07,
"loss": 0.1358,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.3499045849532955,
"learning_rate": 1.839446914621923e-07,
"loss": 0.1371,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.35284564225000786,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1185,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.3506309579811897,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.1226,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.35436050260995844,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.1289,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.34667100721268335,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.1266,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.3808740343316525,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.1119,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.3577670509963671,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.117,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.37088733684639125,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.1312,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.3649000012817196,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1204,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.4138928981970615,
"learning_rate": 1.214956261808542e-07,
"loss": 0.1102,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.38490399588173874,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1221,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.3608057538733826,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.1272,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.3804299099481242,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.1296,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.3663604727593268,
"learning_rate": 9.787081204466253e-08,
"loss": 0.1321,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.3541541909858556,
"learning_rate": 9.236233586200627e-08,
"loss": 0.1347,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.3604359711203267,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1255,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.36093812543058873,
"learning_rate": 8.182296909035047e-08,
"loss": 0.1222,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.3642544914487702,
"learning_rate": 7.679216277384439e-08,
"loss": 0.1242,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.3599134798867866,
"learning_rate": 7.19206618963053e-08,
"loss": 0.1179,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.34610251996371433,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1289,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.3542658824261793,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1195,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.35653260569397993,
"learning_rate": 5.826237505053067e-08,
"loss": 0.1172,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.35327647416204555,
"learning_rate": 5.4028472710137e-08,
"loss": 0.1245,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.3603370561795942,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.1346,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.3524055581817428,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.1199,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.36295155275168556,
"learning_rate": 4.228381947004812e-08,
"loss": 0.1088,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.36484906053508975,
"learning_rate": 3.868805732811343e-08,
"loss": 0.1223,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.36541197901389677,
"learning_rate": 3.525190530455636e-08,
"loss": 0.1236,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.3577164674668402,
"learning_rate": 3.197539087475887e-08,
"loss": 0.1292,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.3709872996980265,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.1223,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.3588799381698973,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.1298,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.355894376154264,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.1295,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.36594209923186094,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.121,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.3540113679719508,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1264,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.3689466868286068,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.1317,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.3599407203594357,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.1216,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.3720998555244566,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.1251,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.3623827627977052,
"learning_rate": 9.67435416023843e-09,
"loss": 0.1321,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.37635072148018583,
"learning_rate": 7.995445897566178e-09,
"loss": 0.1252,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.3559132279702116,
"learning_rate": 6.476393170578021e-09,
"loss": 0.1264,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.3841132401370309,
"learning_rate": 5.11720812557348e-09,
"loss": 0.1028,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.3630237376406681,
"learning_rate": 3.917901630563492e-09,
"loss": 0.1453,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.35684292472692136,
"learning_rate": 2.878483275170485e-09,
"loss": 0.1317,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.3694829073262199,
"learning_rate": 1.998961370555108e-09,
"loss": 0.1299,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.3662592482784979,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.1252,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.37278248317026647,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1223,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.368994110423858,
"learning_rate": 3.198382947511292e-10,
"loss": 0.1187,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.37424640774162293,
"learning_rate": 7.995973352770137e-11,
"loss": 0.1129,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.359864459285011,
"learning_rate": 0.0,
"loss": 0.1126,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 812982844620800.0,
"train_loss": 0.361616323924499,
"train_runtime": 10075.5296,
"train_samples_per_second": 15.682,
"train_steps_per_second": 0.123
}
],
"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": 812982844620800.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}