d1_code_long_paragraphs / trainer_state.json
neginr's picture
End of training
0a42fe4 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 6.315344804493503,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.0153,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 6.550338022416612,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0505,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 6.512881279766042,
"learning_rate": 9.67741935483871e-07,
"loss": 1.0537,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 6.3351344391816165,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.0436,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 5.909020918589431,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.037,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 5.720518494050111,
"learning_rate": 1.935483870967742e-06,
"loss": 1.0057,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 4.531097253096675,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.9783,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 4.15627199556777,
"learning_rate": 2.580645161290323e-06,
"loss": 0.9793,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.735309115677909,
"learning_rate": 2.903225806451613e-06,
"loss": 0.9648,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.5311423524368597,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9614,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.543943733944322,
"learning_rate": 3.548387096774194e-06,
"loss": 0.9572,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.71479521838602,
"learning_rate": 3.870967741935484e-06,
"loss": 0.9485,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 4.098746883205381,
"learning_rate": 4.193548387096774e-06,
"loss": 0.9314,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 3.9089267232260014,
"learning_rate": 4.516129032258065e-06,
"loss": 0.9269,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 3.6204526204689786,
"learning_rate": 4.838709677419355e-06,
"loss": 0.9229,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.7040343639093742,
"learning_rate": 5.161290322580646e-06,
"loss": 0.898,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.4304650597026907,
"learning_rate": 5.483870967741935e-06,
"loss": 0.8641,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 2.026195969551545,
"learning_rate": 5.806451612903226e-06,
"loss": 0.862,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 1.4599636716987738,
"learning_rate": 6.129032258064517e-06,
"loss": 0.8212,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.0659253174242778,
"learning_rate": 6.451612903225806e-06,
"loss": 0.8151,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.024877402771985,
"learning_rate": 6.774193548387097e-06,
"loss": 0.8071,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.1394547056067608,
"learning_rate": 7.096774193548388e-06,
"loss": 0.7908,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.0508897424680632,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.8072,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 0.9345834788518447,
"learning_rate": 7.741935483870968e-06,
"loss": 0.793,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.9262369881779331,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7629,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.031791179198799,
"learning_rate": 8.387096774193549e-06,
"loss": 0.7824,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.8178587789084945,
"learning_rate": 8.70967741935484e-06,
"loss": 0.774,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.7775254109643744,
"learning_rate": 9.03225806451613e-06,
"loss": 0.7733,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.8990159330168151,
"learning_rate": 9.35483870967742e-06,
"loss": 0.7711,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.8541258800430523,
"learning_rate": 9.67741935483871e-06,
"loss": 0.7421,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.678031068635989,
"learning_rate": 1e-05,
"loss": 0.7415,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.7560371160958362,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.7554,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.6672303575336098,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.7448,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.635004017399332,
"learning_rate": 1.096774193548387e-05,
"loss": 0.7287,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.6554664495782486,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.7428,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.5118975692517076,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.7245,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.6254529771722961,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.7114,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.5475668620517357,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.7192,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.5487832631953162,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.7309,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.5312529139292682,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.6911,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.549754109178147,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.7194,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.5643178608779346,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.713,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.49033660950364344,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.7052,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.5797220801772299,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.7006,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.4213856080741056,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.7007,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.5406990322246001,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6848,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.5256415092081611,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.7012,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.4951821055672577,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6857,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.5395282226808027,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.6952,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.452084985267733,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6851,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.4131420176850552,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.6881,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.5201171168320949,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6797,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.5499695958362775,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.7115,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.5545525431432802,
"learning_rate": 1.741935483870968e-05,
"loss": 0.6908,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.5649740171659292,
"learning_rate": 1.774193548387097e-05,
"loss": 0.6884,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.4440233162945597,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6784,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.5110220044884946,
"learning_rate": 1.838709677419355e-05,
"loss": 0.6834,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.4579574745620592,
"learning_rate": 1.870967741935484e-05,
"loss": 0.6811,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.5824974353005503,
"learning_rate": 1.903225806451613e-05,
"loss": 0.692,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.5350887863523411,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6855,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.5356460113607747,
"learning_rate": 1.967741935483871e-05,
"loss": 0.68,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.5509217349665791,
"learning_rate": 2e-05,
"loss": 0.6768,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.5867061411799527,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.6784,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.47148095946941704,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.6737,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.6192097804965417,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.6584,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.6359930420208773,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.6754,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.6821998979016293,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6786,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.9827229147888139,
"learning_rate": 2.193548387096774e-05,
"loss": 0.6685,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 1.4532336509263055,
"learning_rate": 2.225806451612903e-05,
"loss": 0.6612,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.7633837294601843,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.6618,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.908728189450159,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6617,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.1803168465767158,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.6653,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.7157801299503778,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.6625,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.8605578546013016,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.6691,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.0393919658222548,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.6486,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.9152924136080443,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.6719,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.7866639434511055,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.6517,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.0061124358802018,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.6803,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.09932508899296,
"learning_rate": 2.548387096774194e-05,
"loss": 0.6586,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.1881326424887722,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.6549,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.9814567378912181,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.6683,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.0863137531252542,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.6745,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.142705339957574,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6668,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.9233789277811821,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.6513,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.6331794948442986,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.6578,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.8661989417176686,
"learning_rate": 2.774193548387097e-05,
"loss": 0.643,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.9047380043405302,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.647,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.106355770729105,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.661,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 2.1106345081318216,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.6587,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.8969974467777355,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.653,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.582135803093039,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6533,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.777299619349364,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.6617,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.3739607775338774,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.661,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.4998178167394507,
"learning_rate": 3.032258064516129e-05,
"loss": 0.6805,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.3454681727245221,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.6519,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.4285908570899157,
"learning_rate": 3.096774193548387e-05,
"loss": 0.6517,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.2164160589809216,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.6498,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.3750758664770255,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6529,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.0776548605518967,
"learning_rate": 3.193548387096774e-05,
"loss": 0.6521,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.134435345478007,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.6442,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.9884712985453011,
"learning_rate": 3.258064516129033e-05,
"loss": 0.6553,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.4437205148969223,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.6387,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.09048556133505,
"learning_rate": 3.322580645161291e-05,
"loss": 0.6541,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.9118404013282653,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.655,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.0211265725462333,
"learning_rate": 3.387096774193549e-05,
"loss": 0.639,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.1508372580683752,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.6449,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.0893210333630714,
"learning_rate": 3.451612903225807e-05,
"loss": 0.6483,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 1.1909139632196595,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6451,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.45120060067361,
"learning_rate": 3.516129032258065e-05,
"loss": 0.6441,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.6932984830052681,
"learning_rate": 3.548387096774194e-05,
"loss": 0.6205,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.1372057561805617,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.66,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.084810847366894,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6522,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 1.382461103752375,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.6412,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.8465623766777975,
"learning_rate": 3.67741935483871e-05,
"loss": 0.6389,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.0210182555379022,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6374,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 1.318327894401307,
"learning_rate": 3.741935483870968e-05,
"loss": 0.653,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.99691430306203,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.6205,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.8907870122814721,
"learning_rate": 3.806451612903226e-05,
"loss": 0.6167,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 1.085251854015405,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.6318,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.8757032346342265,
"learning_rate": 3.870967741935484e-05,
"loss": 0.622,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.1040658478482388,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.6377,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.4791795077949847,
"learning_rate": 3.935483870967742e-05,
"loss": 0.6545,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.9910514661306425,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.6507,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.175779196030108,
"learning_rate": 4e-05,
"loss": 0.6561,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.1937986552070659,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.6563,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.294797277876138,
"learning_rate": 3.999968016170525e-05,
"loss": 0.6497,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.7430085594242907,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.6331,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.9897268906319332,
"learning_rate": 3.999872065705065e-05,
"loss": 0.6411,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 1.71312152165083,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6316,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.134920455491779,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.6289,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.7501403016850074,
"learning_rate": 3.999608209836944e-05,
"loss": 0.6491,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.375220579786531,
"learning_rate": 3.999488279187443e-05,
"loss": 0.6274,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.3382392698840457,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.6419,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.343712167228504,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.6382,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.1535884537775063,
"learning_rate": 3.999032564583976e-05,
"loss": 0.6397,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.3898408080913458,
"learning_rate": 3.998848689546591e-05,
"loss": 0.6415,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.7292894688708798,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.6372,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.3274780435575062,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.6326,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.9227455540374597,
"learning_rate": 3.998201174509306e-05,
"loss": 0.6303,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 1.1572237759425517,
"learning_rate": 3.997953378607961e-05,
"loss": 0.6355,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.9036561234082887,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.6274,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.9246772473838042,
"learning_rate": 3.997409862168454e-05,
"loss": 0.6334,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.8165069002039731,
"learning_rate": 3.997114145976237e-05,
"loss": 0.6266,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.7137437675166645,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.6312,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.7955846933546589,
"learning_rate": 3.996474809469545e-05,
"loss": 0.6248,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.693750549050478,
"learning_rate": 3.996131194267189e-05,
"loss": 0.6236,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.5681170403726109,
"learning_rate": 3.995771618052995e-05,
"loss": 0.6318,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.8557826738516338,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6385,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.7355079073132349,
"learning_rate": 3.995004594217343e-05,
"loss": 0.6203,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.660293176396107,
"learning_rate": 3.994597152728987e-05,
"loss": 0.6315,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.5220566987016094,
"learning_rate": 3.994173762494947e-05,
"loss": 0.6167,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.5331118380606101,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.622,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.5667155758190681,
"learning_rate": 3.993279149458988e-05,
"loss": 0.6038,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.6314429327912492,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.6183,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.540554835602311,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6202,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.6004543358870873,
"learning_rate": 3.991817703090965e-05,
"loss": 0.622,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.4731169202112667,
"learning_rate": 3.991298695938038e-05,
"loss": 0.6243,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.7267570828911265,
"learning_rate": 3.9907637664138e-05,
"loss": 0.628,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.7664721903024793,
"learning_rate": 3.990212918795534e-05,
"loss": 0.6194,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.6524112498725948,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.6388,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.5500418130428069,
"learning_rate": 3.989063487022416e-05,
"loss": 0.6161,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.5398216049746369,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.6252,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.5091902035206076,
"learning_rate": 3.987850437381915e-05,
"loss": 0.6227,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.4557677616402488,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6176,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.5690103018219557,
"learning_rate": 3.986573808672002e-05,
"loss": 0.6293,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.5668792523454736,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.6193,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.5093969465257118,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6271,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.6337280073132896,
"learning_rate": 3.984539744726554e-05,
"loss": 0.6168,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.6046597944944554,
"learning_rate": 3.983829979402038e-05,
"loss": 0.6248,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.5530489871851663,
"learning_rate": 3.983104351425871e-05,
"loss": 0.6162,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.5720503562661912,
"learning_rate": 3.982362866600154e-05,
"loss": 0.6216,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.5585632901614648,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.6302,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.5631502832971399,
"learning_rate": 3.980832350242388e-05,
"loss": 0.6217,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.7771089797925732,
"learning_rate": 3.980043330948306e-05,
"loss": 0.6183,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.8518110382083331,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6158,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 1.0121868575029893,
"learning_rate": 3.978417801674579e-05,
"loss": 0.6276,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 1.0238450301278594,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.6192,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.8059330417700897,
"learning_rate": 3.976728995023254e-05,
"loss": 0.62,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.7745926630503348,
"learning_rate": 3.975860879481514e-05,
"loss": 0.6264,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.8219279826945453,
"learning_rate": 3.974976965008833e-05,
"loss": 0.634,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.7435438125667699,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.6252,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.6877637644122355,
"learning_rate": 3.973161767667946e-05,
"loss": 0.617,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.5734141180932284,
"learning_rate": 3.972230499314009e-05,
"loss": 0.6249,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.501253074974105,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.6236,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.7891533070464258,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.6105,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.6286704835885626,
"learning_rate": 3.969342105253098e-05,
"loss": 0.6028,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.5444498161498965,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.6125,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.6613528540126358,
"learning_rate": 3.967337762346568e-05,
"loss": 0.607,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.6489882850792098,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.6082,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.4838820709071552,
"learning_rate": 3.965270496444528e-05,
"loss": 0.6125,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.6397799678701908,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.6312,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.6247734653180871,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.6272,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.4258746668727437,
"learning_rate": 3.962051762160151e-05,
"loss": 0.6088,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.5905377316927359,
"learning_rate": 3.960947462140638e-05,
"loss": 0.5922,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.6076558006816525,
"learning_rate": 3.959827482437472e-05,
"loss": 0.6042,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.6145035334672493,
"learning_rate": 3.958691832005981e-05,
"loss": 0.6138,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.566762408801833,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.6193,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.5575349494146696,
"learning_rate": 3.956373555405774e-05,
"loss": 0.6064,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.5074062278239697,
"learning_rate": 3.955190947773937e-05,
"loss": 0.6061,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.46118082881475225,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.6244,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.4238651789576494,
"learning_rate": 3.952778841127214e-05,
"loss": 0.6006,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.45086004497181853,
"learning_rate": 3.95154936139947e-05,
"loss": 0.6021,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.44527788865356,
"learning_rate": 3.950304277135034e-05,
"loss": 0.6123,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.4892768382624844,
"learning_rate": 3.94904359828957e-05,
"loss": 0.5886,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.6329115221309641,
"learning_rate": 3.94776733494343e-05,
"loss": 0.6167,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.5462279018147309,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.6284,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.5279980434051793,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.6052,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.4644128110532906,
"learning_rate": 3.943845140573213e-05,
"loss": 0.6117,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.6321887138452011,
"learning_rate": 3.942506642518952e-05,
"loss": 0.603,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.6100845627332797,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.6026,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.5676311427198942,
"learning_rate": 3.939783060543167e-05,
"loss": 0.6012,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.5959581908381649,
"learning_rate": 3.938397998399333e-05,
"loss": 0.6213,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.5478369467777332,
"learning_rate": 3.936997436876756e-05,
"loss": 0.6134,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.4634584422802504,
"learning_rate": 3.935581387174289e-05,
"loss": 0.6113,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.4531918345431539,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.6012,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.49088750946038057,
"learning_rate": 3.932702868644224e-05,
"loss": 0.6041,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.49648585429575676,
"learning_rate": 3.931240422833181e-05,
"loss": 0.6227,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.6241262724357991,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.617,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.7225709268725832,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.6102,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.6350235110253573,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.6069,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.5790956679985777,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.6073,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.5382821200814115,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.6089,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.45893709321226345,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.6085,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.634905365782114,
"learning_rate": 3.92057159243286e-05,
"loss": 0.5971,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.8834580711773048,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.6182,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.7725152294444944,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.6266,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.456833558050999,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.6014,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.43471302128002165,
"learning_rate": 3.914136997982874e-05,
"loss": 0.5938,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.4682870810864501,
"learning_rate": 3.912490053436079e-05,
"loss": 0.6095,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.5439059557475846,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.6167,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.5802810782719365,
"learning_rate": 3.909150300975178e-05,
"loss": 0.609,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.6147571735648176,
"learning_rate": 3.907457519765642e-05,
"loss": 0.5974,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.5156030360336709,
"learning_rate": 3.905749486576607e-05,
"loss": 0.6048,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.6820273936039182,
"learning_rate": 3.90402621506546e-05,
"loss": 0.6006,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.6617945877972128,
"learning_rate": 3.902287719011435e-05,
"loss": 0.6054,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.5131325929285164,
"learning_rate": 3.900534012315499e-05,
"loss": 0.6107,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.5342112094728354,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.6007,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.43269563279407847,
"learning_rate": 3.896981023209777e-05,
"loss": 0.5974,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.574006401116357,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.594,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.6311608779319255,
"learning_rate": 3.893367361386491e-05,
"loss": 0.6072,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.4865480543454508,
"learning_rate": 3.891537814248417e-05,
"loss": 0.6043,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.5503091667863292,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5952,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.6889312683862858,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.6017,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.5926882080628791,
"learning_rate": 3.885958483839049e-05,
"loss": 0.6031,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.6673239327720968,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.596,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.696415780140904,
"learning_rate": 3.882163505079171e-05,
"loss": 0.5817,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.48951801411182955,
"learning_rate": 3.880243433489132e-05,
"loss": 0.607,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.8567892719188328,
"learning_rate": 3.878308327522702e-05,
"loss": 0.5629,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.8573714017783893,
"learning_rate": 3.876358202652936e-05,
"loss": 0.55,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.6874427730707529,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.5479,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.6723248821546056,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.5573,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.7797748333712637,
"learning_rate": 3.870417871154801e-05,
"loss": 0.5601,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.6596331914866355,
"learning_rate": 3.868407827802192e-05,
"loss": 0.5641,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.6098811067406051,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.561,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.7665113346352799,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.5747,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.7368688725563961,
"learning_rate": 3.862288124195319e-05,
"loss": 0.5499,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.5556525934658203,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.565,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.5869026007529993,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.5482,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.6503606047909136,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.5705,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.8651940490678236,
"learning_rate": 3.853920126717703e-05,
"loss": 0.553,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.9380623288727348,
"learning_rate": 3.851791025492574e-05,
"loss": 0.5599,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.5621769444387076,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.5529,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.641429888047757,
"learning_rate": 3.847488419569866e-05,
"loss": 0.5465,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.8294093509064677,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.5392,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.5581143557023232,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.548,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.5216428655341537,
"learning_rate": 3.840923760917199e-05,
"loss": 0.5467,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.7206651723494089,
"learning_rate": 3.838706077964473e-05,
"loss": 0.5627,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.6597112214820025,
"learning_rate": 3.836473692766944e-05,
"loss": 0.5754,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.5131749009701888,
"learning_rate": 3.834226623174705e-05,
"loss": 0.5594,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.6166128749563262,
"learning_rate": 3.831964887155264e-05,
"loss": 0.5558,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.6561632694487464,
"learning_rate": 3.829688502793403e-05,
"loss": 0.5529,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.8649743301615463,
"learning_rate": 3.827397488291029e-05,
"loss": 0.572,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.7572950030348651,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.5415,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.44503899739126873,
"learning_rate": 3.822771642257145e-05,
"loss": 0.5546,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.7786038098754372,
"learning_rate": 3.820436847713776e-05,
"loss": 0.5606,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.6908623469586844,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.5457,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.460011132173869,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.5515,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.7932718724334049,
"learning_rate": 3.813345202354145e-05,
"loss": 0.5447,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.6300451325872004,
"learning_rate": 3.810952296329563e-05,
"loss": 0.5537,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.5095487274161112,
"learning_rate": 3.808544909978676e-05,
"loss": 0.5498,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.5259118406367365,
"learning_rate": 3.806123062550882e-05,
"loss": 0.5427,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.5942445177680632,
"learning_rate": 3.803686773411208e-05,
"loss": 0.5444,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.5450521091275004,
"learning_rate": 3.801236062040158e-05,
"loss": 0.5467,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.5117581275851685,
"learning_rate": 3.798770948033553e-05,
"loss": 0.5533,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.4895300062331844,
"learning_rate": 3.796291451102381e-05,
"loss": 0.5618,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.3857875259007961,
"learning_rate": 3.793797591072631e-05,
"loss": 0.5508,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.3540414193557973,
"learning_rate": 3.791289387885144e-05,
"loss": 0.5466,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.43442445652208606,
"learning_rate": 3.788766861595443e-05,
"loss": 0.5426,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.40843041370513383,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.5582,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.4412896968132787,
"learning_rate": 3.783678920503983e-05,
"loss": 0.5501,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.45801628074475204,
"learning_rate": 3.781113546385265e-05,
"loss": 0.5411,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.3936280889111345,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.5601,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.45926539347087547,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.5522,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.3987478929596095,
"learning_rate": 3.77333205623025e-05,
"loss": 0.5511,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.4390210810020236,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.5558,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.4643150113658209,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.5467,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.42540931178778196,
"learning_rate": 3.765422951112704e-05,
"loss": 0.556,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.42699193579481187,
"learning_rate": 3.762758321970216e-05,
"loss": 0.5547,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.4308917461479461,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.576,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.39958398321228167,
"learning_rate": 3.757386800198538e-05,
"loss": 0.5506,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.4526211306195955,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.5424,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.3733872582096588,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.5454,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.3775130061713638,
"learning_rate": 3.749224181796305e-05,
"loss": 0.546,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.4605706786951504,
"learning_rate": 3.746475306375545e-05,
"loss": 0.5402,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.35704857863447564,
"learning_rate": 3.743712466184774e-05,
"loss": 0.5505,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.4099671618575976,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.5446,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.35137585369931057,
"learning_rate": 3.73814497997107e-05,
"loss": 0.55,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.3853892078853167,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.5555,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.40945227904964526,
"learning_rate": 3.732521901224724e-05,
"loss": 0.5557,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.4062229213260818,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.5483,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.4423325645507449,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.5529,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.38957525744908034,
"learning_rate": 3.723983441007882e-05,
"loss": 0.5497,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.4522777825365163,
"learning_rate": 3.721109687296781e-05,
"loss": 0.5499,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.4921783032387999,
"learning_rate": 3.718222171638484e-05,
"loss": 0.556,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.44091995808783024,
"learning_rate": 3.715320917121488e-05,
"loss": 0.5479,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.40002340643337464,
"learning_rate": 3.712405946944149e-05,
"loss": 0.5579,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.5142543781483211,
"learning_rate": 3.709477284414488e-05,
"loss": 0.5548,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.45736228344712015,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.5427,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.3768860541448736,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.5604,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.3909234819415993,
"learning_rate": 3.70060937743295e-05,
"loss": 0.5495,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.40058266641263146,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.5614,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.45685529260677904,
"learning_rate": 3.694629409915555e-05,
"loss": 0.5455,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.4194615552744798,
"learning_rate": 3.691619088858401e-05,
"loss": 0.5506,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.5087543635530766,
"learning_rate": 3.68859524166009e-05,
"loss": 0.5537,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.409162621715833,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.552,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.4911848547859652,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.5493,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.4482650506914698,
"learning_rate": 3.67944278554384e-05,
"loss": 0.5454,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.40787879546906025,
"learning_rate": 3.676365076645557e-05,
"loss": 0.5553,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.4467475405614192,
"learning_rate": 3.673273963576791e-05,
"loss": 0.554,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.44535967210330835,
"learning_rate": 3.670169471054001e-05,
"loss": 0.5479,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.4276397591222824,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.5453,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.46821358358583576,
"learning_rate": 3.663920447046888e-05,
"loss": 0.5519,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.4628872687302214,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.5528,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.38580827132087386,
"learning_rate": 3.657618204491937e-05,
"loss": 0.5618,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.43974572338245244,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.5497,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.41170192241190956,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.5637,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.426150389276357,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.5475,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.42101637311352896,
"learning_rate": 3.644854871713611e-05,
"loss": 0.5541,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.39499437813327076,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.5432,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.3861199713430869,
"learning_rate": 3.638394189710493e-05,
"loss": 0.5504,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.42095727271117106,
"learning_rate": 3.635144184933412e-05,
"loss": 0.5489,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.35043108827700176,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.5472,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.4335905337570274,
"learning_rate": 3.628604977762752e-05,
"loss": 0.5565,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.419759294634855,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.5568,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.40806138852458285,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.5443,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.35149390105916695,
"learning_rate": 3.618698565901727e-05,
"loss": 0.5567,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.3868481217444394,
"learning_rate": 3.615370507164654e-05,
"loss": 0.5406,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.35415349255152573,
"learning_rate": 3.61202953196805e-05,
"loss": 0.5478,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.32100258273803883,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.5384,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.4067417362728998,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.559,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.37174515841009703,
"learning_rate": 3.601929375279662e-05,
"loss": 0.5461,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.3464332228404925,
"learning_rate": 3.598537002418012e-05,
"loss": 0.5492,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.39496566086494145,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.5644,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.3602754870807741,
"learning_rate": 3.591713938344416e-05,
"loss": 0.5428,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.40883127565834,
"learning_rate": 3.588283301689507e-05,
"loss": 0.5393,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.301050173078219,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.5486,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.37866314866460304,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.5463,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.35397851174852574,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.5432,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.33002625832682636,
"learning_rate": 3.574434032194994e-05,
"loss": 0.5438,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.34020052777399884,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.5468,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.3717542849954809,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.5442,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.43223443380703697,
"learning_rate": 3.563914797637216e-05,
"loss": 0.534,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.45186822153190787,
"learning_rate": 3.560383338592649e-05,
"loss": 0.5472,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.46253239799001694,
"learning_rate": 3.556839402764487e-05,
"loss": 0.5344,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.5610813587040556,
"learning_rate": 3.553283018489946e-05,
"loss": 0.5399,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.38675276747419574,
"learning_rate": 3.54971421420578e-05,
"loss": 0.5213,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.5039560139778603,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.5481,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.4859258541353042,
"learning_rate": 3.54253945985191e-05,
"loss": 0.5607,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.4558187362350024,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.5564,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.5575264727517055,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.5461,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.3835477867227507,
"learning_rate": 3.531684894865866e-05,
"loss": 0.5352,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.44796052386027324,
"learning_rate": 3.528042173241128e-05,
"loss": 0.5575,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.5295461240498046,
"learning_rate": 3.524387233431892e-05,
"loss": 0.5406,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.3938858892300531,
"learning_rate": 3.520720104662958e-05,
"loss": 0.5473,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.39606188173519774,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.5505,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.4598409937841456,
"learning_rate": 3.513349397632281e-05,
"loss": 0.5595,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.4064363638956567,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.5545,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.4494051205612022,
"learning_rate": 3.505930287892533e-05,
"loss": 0.5424,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.3499583964582132,
"learning_rate": 3.502202656100099e-05,
"loss": 0.5482,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.41159040956660264,
"learning_rate": 3.498463012735256e-05,
"loss": 0.5529,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.3462640380827727,
"learning_rate": 3.494711387700094e-05,
"loss": 0.5424,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.4128848264646138,
"learning_rate": 3.490947810992505e-05,
"loss": 0.5488,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.3496155700958841,
"learning_rate": 3.48717231270595e-05,
"loss": 0.5604,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.3948749151856486,
"learning_rate": 3.483384923029211e-05,
"loss": 0.5413,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.4975596803092942,
"learning_rate": 3.479585672246156e-05,
"loss": 0.547,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.3471370615783147,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.546,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.5123379253720602,
"learning_rate": 3.471951708970527e-05,
"loss": 0.5373,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.43004900091358306,
"learning_rate": 3.468117057518921e-05,
"loss": 0.5592,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.4320671835826538,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.559,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.48602325253728856,
"learning_rate": 3.460412568296731e-05,
"loss": 0.5489,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.45387045787335223,
"learning_rate": 3.456542792131039e-05,
"loss": 0.5355,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.4915049197856742,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.5444,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.4872917095574931,
"learning_rate": 3.448768331403348e-05,
"loss": 0.5363,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.41198337126896983,
"learning_rate": 3.44486370900573e-05,
"loss": 0.547,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.4183216891145548,
"learning_rate": 3.440947533516396e-05,
"loss": 0.5597,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.5172163041779689,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.5368,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.4130340972950618,
"learning_rate": 3.43308064860925e-05,
"loss": 0.5456,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.3945917207894784,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.5428,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.44274829240546465,
"learning_rate": 3.425167928295015e-05,
"loss": 0.5451,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.42951625641992425,
"learning_rate": 3.421194458890411e-05,
"loss": 0.5543,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.4223762875441522,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.5407,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.3776399331463046,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.557,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.422110076250925,
"learning_rate": 3.409205995219561e-05,
"loss": 0.5512,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.4644628987293047,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.5533,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.4593650516062552,
"learning_rate": 3.40115729298209e-05,
"loss": 0.5352,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.4081552207835842,
"learning_rate": 3.397116120327025e-05,
"loss": 0.5536,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.46390821919352304,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.5488,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.48006420156734697,
"learning_rate": 3.389000293509524e-05,
"loss": 0.5565,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.4009530408779364,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.5334,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.347584356059142,
"learning_rate": 3.380840041143494e-05,
"loss": 0.5436,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.4644286018919832,
"learning_rate": 3.376743336885792e-05,
"loss": 0.5482,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.3946242989687989,
"learning_rate": 3.372635624225055e-05,
"loss": 0.5338,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.37538994395020603,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.5427,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.3513395367883207,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.5457,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.3474769708730885,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.5508,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.3500200763656646,
"learning_rate": 3.356095347769799e-05,
"loss": 0.5546,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.31344947330665496,
"learning_rate": 3.351933087522552e-05,
"loss": 0.54,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.33519383912596035,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.5462,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.31597342178523496,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.5584,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.36534733881037756,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.5483,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.37880815948683116,
"learning_rate": 3.335176280432308e-05,
"loss": 0.5432,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.34032985750991906,
"learning_rate": 3.330960304618295e-05,
"loss": 0.555,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.38066777605505797,
"learning_rate": 3.326733686481153e-05,
"loss": 0.5546,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.4529085395725005,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.5517,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.34542066140601835,
"learning_rate": 3.318248658506012e-05,
"loss": 0.5393,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.4000665348836789,
"learning_rate": 3.313990316514069e-05,
"loss": 0.5399,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.3884695029820167,
"learning_rate": 3.309721467890571e-05,
"loss": 0.5396,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.4335213158107321,
"learning_rate": 3.305442146769115e-05,
"loss": 0.5513,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.3713376240327559,
"learning_rate": 3.30115238736704e-05,
"loss": 0.5522,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.3882721675980082,
"learning_rate": 3.296852223985148e-05,
"loss": 0.553,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.3837278862143727,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.5427,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.3163337433824195,
"learning_rate": 3.288220822900796e-05,
"loss": 0.5413,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.4398928553357752,
"learning_rate": 3.283889654214788e-05,
"loss": 0.5402,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.3732474597321962,
"learning_rate": 3.279548219581317e-05,
"loss": 0.5209,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.46868624574754797,
"learning_rate": 3.275196553714379e-05,
"loss": 0.5551,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.4512038045423586,
"learning_rate": 3.270834691409778e-05,
"loss": 0.5447,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.5318064892492224,
"learning_rate": 3.266462667544849e-05,
"loss": 0.552,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.5034309799925539,
"learning_rate": 3.262080517078178e-05,
"loss": 0.5418,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.4646460223068323,
"learning_rate": 3.257688275049323e-05,
"loss": 0.5608,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.49420896864573616,
"learning_rate": 3.253285976578535e-05,
"loss": 0.5517,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.4263467887245604,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.5468,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.3567891171768097,
"learning_rate": 3.244451351193935e-05,
"loss": 0.5571,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.38292514381202,
"learning_rate": 3.240019094921551e-05,
"loss": 0.5415,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.4105959269727847,
"learning_rate": 3.235576923489527e-05,
"loss": 0.5337,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.49537536958558376,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.5345,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.37675699771433485,
"learning_rate": 3.226662977303494e-05,
"loss": 0.5468,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.5149527014246056,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.5433,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.5127484945382321,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.5511,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.32825291947554175,
"learning_rate": 3.21321858487569e-05,
"loss": 0.5376,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.500971338414554,
"learning_rate": 3.208717671149931e-05,
"loss": 0.5463,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.48928953785109147,
"learning_rate": 3.204207092549882e-05,
"loss": 0.5364,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.36495838264098474,
"learning_rate": 3.19968688514201e-05,
"loss": 0.5544,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.48505888922684526,
"learning_rate": 3.195157085069774e-05,
"loss": 0.5535,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.33826304448506544,
"learning_rate": 3.190617728553332e-05,
"loss": 0.552,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.3659638128189783,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.5378,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.39688016767094286,
"learning_rate": 3.181510491450254e-05,
"loss": 0.5489,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.34537552592006976,
"learning_rate": 3.176942683684842e-05,
"loss": 0.5398,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.3289600515288395,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.5339,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.35978508391951114,
"learning_rate": 3.167778872346327e-05,
"loss": 0.5403,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.4133842045490602,
"learning_rate": 3.163182942046815e-05,
"loss": 0.55,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.4305689642129004,
"learning_rate": 3.158577710967494e-05,
"loss": 0.5479,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.3476649836154382,
"learning_rate": 3.153963215931669e-05,
"loss": 0.5485,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.4593112919341221,
"learning_rate": 3.14933949383672e-05,
"loss": 0.5415,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.324220350258683,
"learning_rate": 3.144706581653805e-05,
"loss": 0.545,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.4200339298811297,
"learning_rate": 3.140064516427566e-05,
"loss": 0.5535,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.2925070098443886,
"learning_rate": 3.135413335275833e-05,
"loss": 0.5494,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.39026978325851763,
"learning_rate": 3.130753075389327e-05,
"loss": 0.5459,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.39653741357854394,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.5442,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.332638494940465,
"learning_rate": 3.121405468537547e-05,
"loss": 0.5411,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.41344241319186487,
"learning_rate": 3.116718196315488e-05,
"loss": 0.5251,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.30611630098202414,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.5375,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.40155184261660404,
"learning_rate": 3.107316901675252e-05,
"loss": 0.5503,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.32753033672367304,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.5587,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.4186838042945674,
"learning_rate": 3.097880190800057e-05,
"loss": 0.5327,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.3392295005777781,
"learning_rate": 3.093148648549788e-05,
"loss": 0.5293,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.4402194207404536,
"learning_rate": 3.088408365512055e-05,
"loss": 0.5554,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.3670570483290637,
"learning_rate": 3.083659379590034e-05,
"loss": 0.5615,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.4054953942387685,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.537,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.37507724265473,
"learning_rate": 3.074135451053475e-05,
"loss": 0.5505,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.3783147867537531,
"learning_rate": 3.069360584592015e-05,
"loss": 0.5561,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.4455556574151963,
"learning_rate": 3.064577167551816e-05,
"loss": 0.5313,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.36672487784200686,
"learning_rate": 3.059785238180953e-05,
"loss": 0.5455,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.4320033757231638,
"learning_rate": 3.054984834795566e-05,
"loss": 0.5536,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.3505940485826497,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.5358,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.4280513111160789,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.532,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.35853288245611564,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.5515,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.3825579346616797,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.5278,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.3273710573630122,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.5371,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.36898893801472143,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.5452,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.3559780310460431,
"learning_rate": 3.021147971398632e-05,
"loss": 0.5469,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.34034642888662375,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.5353,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.3965019423450832,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.5408,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.3418690333771321,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.5415,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.33028339024503234,
"learning_rate": 3.001632138223332e-05,
"loss": 0.5329,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.4690172729536988,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.4799,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.36608415898933006,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.4823,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.535214982289163,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.4822,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.4872561927165149,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.4807,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.45253547034758235,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.4712,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.40466821934344727,
"learning_rate": 2.972118904869376e-05,
"loss": 0.4701,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.47146389507771114,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.4862,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.3561040327317459,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.4693,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.3797748931957262,
"learning_rate": 2.95725682050846e-05,
"loss": 0.4813,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.3610524070410098,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.4772,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.40352740040884855,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.4797,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.4139535237581799,
"learning_rate": 2.942325848714505e-05,
"loss": 0.4756,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.36724653293986975,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.4665,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.4521771685394572,
"learning_rate": 2.932334127129686e-05,
"loss": 0.483,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.4503900130775678,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.4769,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.3796709090333619,
"learning_rate": 2.922312585929131e-05,
"loss": 0.4806,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.34172247353633206,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.4616,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.33983786985242637,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.4794,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.3903145596718878,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.4733,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.31374447087368595,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.4655,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.3626646955657353,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.4649,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.3646303000532483,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.477,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.3164606995215515,
"learning_rate": 2.887006998351372e-05,
"loss": 0.476,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.38836334907586184,
"learning_rate": 2.881934649605535e-05,
"loss": 0.4904,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.3299664834830835,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.473,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.34405336070284825,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.4792,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.34085866204574,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.4808,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.33036022932489234,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.4799,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.37789951712021486,
"learning_rate": 2.856467940121025e-05,
"loss": 0.4747,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.37841116518712475,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.4751,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.31510504073593937,
"learning_rate": 2.846233032790149e-05,
"loss": 0.4879,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.31135599510117323,
"learning_rate": 2.84110540896642e-05,
"loss": 0.4709,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.29695573595743124,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.4631,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.32427402999798255,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.4686,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.353708472079373,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.4758,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.3306744478007084,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.476,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.3233642684129719,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.4588,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.3757005217313458,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.4685,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.33642052432333264,
"learning_rate": 2.805026032173773e-05,
"loss": 0.4714,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.36922002186982295,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.4592,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.30906048903008737,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.4731,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.29172038737662054,
"learning_rate": 2.7894660620408e-05,
"loss": 0.4779,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.3229490428883993,
"learning_rate": 2.784266724883173e-05,
"loss": 0.4843,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.313426269763496,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.4789,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.3247958362514588,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.4792,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.35621808872829713,
"learning_rate": 2.768631254100716e-05,
"loss": 0.4773,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.3448703296002475,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.4763,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.32325971698360423,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.485,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.32699383766223117,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.4638,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.34883325178473457,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.4802,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.304552169878538,
"learning_rate": 2.742449775134085e-05,
"loss": 0.4647,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.36119098513846026,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.484,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.2975974305812074,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.4846,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.34365237267538457,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.4703,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.354788526864924,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.4734,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.3187671555942889,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.4783,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.3159479237702368,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.4836,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.30932076950495657,
"learning_rate": 2.705547548999996e-05,
"loss": 0.4726,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.3496375358118754,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.4841,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.31261006421340465,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.4799,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.31146301665781956,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.4785,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.3584025940609912,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.4787,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.30342929339014896,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.4726,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.33726399932476303,
"learning_rate": 2.67369650552155e-05,
"loss": 0.4753,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.32062877367538695,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.4791,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.31552920381762223,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.4807,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.2779962253134438,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.4833,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.3307867655606134,
"learning_rate": 2.652354173593614e-05,
"loss": 0.4762,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.3076320668297416,
"learning_rate": 2.647005443331317e-05,
"loss": 0.4772,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.31197074523113716,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.4871,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.2819214949937856,
"learning_rate": 2.636292505301542e-05,
"loss": 0.4722,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.33630374534682345,
"learning_rate": 2.630928383194431e-05,
"loss": 0.4756,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.30799708778286916,
"learning_rate": 2.625559216200781e-05,
"loss": 0.4832,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.30529590722609257,
"learning_rate": 2.620185047252308e-05,
"loss": 0.4663,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.32382944700987654,
"learning_rate": 2.614805919320723e-05,
"loss": 0.4934,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.29574422798520167,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.4649,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.29851914550007635,
"learning_rate": 2.60403295859298e-05,
"loss": 0.481,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.37750698064451077,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.4907,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.32330013742136465,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.4814,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.3123708253008509,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.475,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.3433242567053657,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.4814,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.40564679055624225,
"learning_rate": 2.577016790136453e-05,
"loss": 0.4767,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.3127235884739865,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.4719,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.43953678643214955,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.4842,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.3153250596194981,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.4822,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.4259711510551156,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.4713,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.3589084250041222,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.4813,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.32207968700137235,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.4872,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.3358315225119104,
"learning_rate": 2.539001591449889e-05,
"loss": 0.4783,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.39059800382284826,
"learning_rate": 2.53355326151989e-05,
"loss": 0.4745,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.29962401571216596,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.4737,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.4302533963106725,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.4827,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.4057723968456833,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.4643,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.34954130570983344,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.4761,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.44649865938475014,
"learning_rate": 2.506248490185627e-05,
"loss": 0.4692,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.3588678393094076,
"learning_rate": 2.500775217241923e-05,
"loss": 0.4903,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.364131064270581,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.4824,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.3984606039107438,
"learning_rate": 2.4898167025948e-05,
"loss": 0.4807,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.37473084674110874,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.4792,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.39368105387424435,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.4754,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.3889298159911366,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.4787,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.3824055553397547,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.4934,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.3891334837454286,
"learning_rate": 2.46235264422994e-05,
"loss": 0.4779,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.35160506513974854,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.4778,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.32480714421096923,
"learning_rate": 2.451340834511592e-05,
"loss": 0.4871,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.3534834817596485,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.4837,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.28128581366325967,
"learning_rate": 2.440314589184958e-05,
"loss": 0.4803,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.38151389222284443,
"learning_rate": 2.434796163357541e-05,
"loss": 0.4853,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.30728827275585047,
"learning_rate": 2.429274260911588e-05,
"loss": 0.4785,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.327007149543431,
"learning_rate": 2.423748926000084e-05,
"loss": 0.4864,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.3782833659550479,
"learning_rate": 2.418220202803459e-05,
"loss": 0.4811,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.28013958100636444,
"learning_rate": 2.412688135529237e-05,
"loss": 0.4811,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.3687053220204744,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.4892,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.32888564535976467,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.4829,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.332521971586647,
"learning_rate": 2.396072311715187e-05,
"loss": 0.4786,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.3059582772249945,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.4723,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.305225491723913,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.4824,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.3498416147443178,
"learning_rate": 2.379427985200223e-05,
"loss": 0.468,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.2837079263203963,
"learning_rate": 2.373873749394972e-05,
"loss": 0.4796,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.366607209797723,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.4884,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.3029420327159463,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.4742,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.38194232535026823,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.4793,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.29884697224490336,
"learning_rate": 2.351627355800233e-05,
"loss": 0.4844,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.36081961435548315,
"learning_rate": 2.346058617160256e-05,
"loss": 0.4871,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.35061373923187805,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.4861,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.3289292822111215,
"learning_rate": 2.334912883203469e-05,
"loss": 0.47,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.3691210793139029,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.4804,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.30199227859313205,
"learning_rate": 2.323756437450136e-05,
"loss": 0.4717,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.3490918036791618,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.4858,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.33455152592212906,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.4718,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.33215448002072157,
"learning_rate": 2.307002464849052e-05,
"loss": 0.472,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.2783535190401024,
"learning_rate": 2.30141283818846e-05,
"loss": 0.476,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.3138164579455423,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.4816,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.30463698352251584,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.4888,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.3304651097807202,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.4671,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.2752709524733647,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.4915,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.36325005802843185,
"learning_rate": 2.273429448107257e-05,
"loss": 0.4828,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.297548296135387,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.4731,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.3360388594610977,
"learning_rate": 2.262220474347102e-05,
"loss": 0.4676,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.30986675504687394,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.4759,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.3245255861287697,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.4796,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.30411255457430586,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.4837,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.2885807310350221,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.4752,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.32825935261925465,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.4644,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.26551812424121507,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.4907,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.2823145638058377,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.4733,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.26164324087892615,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.4786,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.2968136434772259,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.4772,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.30929508356267926,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.4988,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.29169339448219167,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.4741,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.27515367690766523,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.4836,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.2875550472878563,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.4935,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.2856574224592378,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.4694,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.2630856528667692,
"learning_rate": 2.17791064266051e-05,
"loss": 0.4738,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.3158765106618256,
"learning_rate": 2.172276926861741e-05,
"loss": 0.4884,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.2474833488606744,
"learning_rate": 2.166641833541255e-05,
"loss": 0.4827,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.3248701150981445,
"learning_rate": 2.161005407757109e-05,
"loss": 0.4792,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.26126596007902786,
"learning_rate": 2.155367694578013e-05,
"loss": 0.4828,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.35843505611019494,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.4736,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.29139419663152466,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.4786,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.3333723557221308,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.4723,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.2937798252393051,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.4815,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.3439389038259579,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.4784,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.3002970995038096,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.481,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.29870611103092914,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.4998,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.28554130362427005,
"learning_rate": 2.110225054402733e-05,
"loss": 0.4744,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.28351583475043385,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.4823,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.3352879482994599,
"learning_rate": 2.09892967807724e-05,
"loss": 0.4741,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.25854596926557905,
"learning_rate": 2.093280780774835e-05,
"loss": 0.4946,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.3141512470462282,
"learning_rate": 2.087631137601793e-05,
"loss": 0.4858,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.2816693653702389,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.474,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.26149709269073346,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.4661,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.2593079669846544,
"learning_rate": 2.070678184630458e-05,
"loss": 0.4723,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.24412013454548268,
"learning_rate": 2.06502600977305e-05,
"loss": 0.4755,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.27087590122684235,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.4797,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.30680643567948584,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.478,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.2779081810487316,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.4896,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.3221557215580396,
"learning_rate": 2.042412562887594e-05,
"loss": 0.4936,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.2926262374956299,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.4737,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.3059643493178367,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.4809,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.2714811117906573,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.4836,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.3119380422208538,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.4587,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.27904575537296417,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.4749,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.3080022008134408,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.4767,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.2739979025766608,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.4854,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.30407110300163176,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.4788,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.26646934503544406,
"learning_rate": 1.99151687695709e-05,
"loss": 0.464,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.29583327716179947,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.4753,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.2612471150462115,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.4845,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.28070379617845387,
"learning_rate": 1.974551241345447e-05,
"loss": 0.4838,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.2833900297210306,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.4801,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.36582064712260637,
"learning_rate": 1.963241759676897e-05,
"loss": 0.5039,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.27743678882409323,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.4678,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.3521632171002781,
"learning_rate": 1.951933453677638e-05,
"loss": 0.4836,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.3236268436393915,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.4806,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.2545076430959982,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.472,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.32714708334211334,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.4661,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.2820409185781744,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.4808,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.34340583061738195,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.4764,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.28290912223738435,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.4775,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.3140401605472174,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.4748,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.2866242480857415,
"learning_rate": 1.906719219225165e-05,
"loss": 0.4773,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.29688545734839306,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.4769,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.2717323327684143,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.483,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.30182125004886956,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.4778,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.3030275748545609,
"learning_rate": 1.884128556891707e-05,
"loss": 0.4784,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.27889031801806036,
"learning_rate": 1.878483094691119e-05,
"loss": 0.4919,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.3172946145599113,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.492,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.2556649896069282,
"learning_rate": 1.867195130360947e-05,
"loss": 0.4867,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.30119033704852866,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.4986,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.2467640266673773,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.4819,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.305351815947829,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.4755,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.29751826550780447,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.4707,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.30168619941749636,
"learning_rate": 1.838994592242891e-05,
"loss": 0.468,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.27864872983905226,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.4754,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.29571673092897577,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.4804,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.28221519233998005,
"learning_rate": 1.82208935733949e-05,
"loss": 0.4766,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.2839859072032392,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.4654,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.26560713891583854,
"learning_rate": 1.810826238483892e-05,
"loss": 0.4788,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.26412800966540745,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.4808,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.24175997554037276,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.4758,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.2703803862709751,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.4663,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.27745426780994825,
"learning_rate": 1.788318512320838e-05,
"loss": 0.4807,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.25138851490402403,
"learning_rate": 1.782695699829051e-05,
"loss": 0.4657,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.2853132374611635,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.4747,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.2919385359896836,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.4871,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.27512977663192145,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.4733,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.2728060340448131,
"learning_rate": 1.76022227484387e-05,
"loss": 0.4666,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.27561853650318746,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.4743,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.2500096703902725,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.481,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.29327140681373676,
"learning_rate": 1.743387180005809e-05,
"loss": 0.465,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.26526587406864954,
"learning_rate": 1.737779525652899e-05,
"loss": 0.4721,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.27219693007300155,
"learning_rate": 1.732173968007914e-05,
"loss": 0.4812,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.2692853852070794,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.4671,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.2813602635980238,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.4778,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.28648519068350875,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.4696,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.28567041060464204,
"learning_rate": 1.709773600686015e-05,
"loss": 0.4667,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.2981332333724621,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.477,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.27633689727752775,
"learning_rate": 1.69858716181154e-05,
"loss": 0.4784,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.31335253333134155,
"learning_rate": 1.692997535150948e-05,
"loss": 0.4764,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.32352071612674266,
"learning_rate": 1.687410363273884e-05,
"loss": 0.4732,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.24973887606529926,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.4699,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.38417312740918524,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.485,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.24594624194024362,
"learning_rate": 1.67066402299235e-05,
"loss": 0.4791,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.32581969763630547,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.4671,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.26519188017288997,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.4744,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.29663891818158217,
"learning_rate": 1.653941382839745e-05,
"loss": 0.4698,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.46212894990721537,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.4062,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.42247373114845527,
"learning_rate": 1.642806717162757e-05,
"loss": 0.4246,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.5091849946869025,
"learning_rate": 1.637243646233718e-05,
"loss": 0.4295,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.537834888762331,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.4118,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.4236263762867585,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.4153,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.4236652206067095,
"learning_rate": 1.620572014799777e-05,
"loss": 0.4186,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.40347667766343837,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.4074,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.37257669282929934,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.4095,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.3427231041669938,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.4067,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.37297648898485525,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.4111,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.30942575027921215,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.4168,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.3364619055836672,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.4069,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.3278794455845886,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.4252,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.3237115566313643,
"learning_rate": 1.576251073999917e-05,
"loss": 0.4279,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.30962681708601775,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.4137,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.30021154428356395,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.4277,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.2896788849525453,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.4119,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.2854540565709141,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.4096,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.28313032712344044,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.4086,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.2879107479115269,
"learning_rate": 1.543151434154755e-05,
"loss": 0.4103,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.29125626746714894,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.4094,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.308282412155216,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.4257,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.3082715352534945,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.4245,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.32285859359168695,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.3969,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.28533611498355066,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.404,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.2809242884399436,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.4048,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.29575284003356217,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.4101,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.261768773778476,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.4236,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.2613837960781988,
"learning_rate": 1.493751509814374e-05,
"loss": 0.4249,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.2725981851920068,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.4151,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.2588838129594178,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.3985,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.26351092903469897,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.4176,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.28437524631845973,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.4195,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.24828969524453592,
"learning_rate": 1.466446738480111e-05,
"loss": 0.4075,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.2620932462606984,
"learning_rate": 1.460998408550112e-05,
"loss": 0.412,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.2643228537199386,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.4198,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.24878857311283495,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.415,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.24435090375706914,
"learning_rate": 1.444679451900437e-05,
"loss": 0.3986,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.2756873683049054,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.4044,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.260685053186765,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.4235,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.27923385503796183,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.4041,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.2585469147224591,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.4128,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.26769899419127,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.4146,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.2696745455109515,
"learning_rate": 1.412162598317602e-05,
"loss": 0.4177,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.2752118257607324,
"learning_rate": 1.406759321421907e-05,
"loss": 0.4217,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.26383086152609625,
"learning_rate": 1.401360788062871e-05,
"loss": 0.4436,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.27548940447253967,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.4012,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.26150267077747846,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.3797,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.27945497587706203,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.4032,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.26849281001110487,
"learning_rate": 1.379814952747693e-05,
"loss": 0.4173,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.27963352434220023,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.4065,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.250404124248925,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.412,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.2866109499788337,
"learning_rate": 1.363707494698459e-05,
"loss": 0.4221,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.277245774873839,
"learning_rate": 1.358348460369265e-05,
"loss": 0.4093,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.2419593887266395,
"learning_rate": 1.352994556668684e-05,
"loss": 0.4194,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.27149589595920065,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.3973,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.2535249507199965,
"learning_rate": 1.342302312350678e-05,
"loss": 0.4014,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.2797902468536024,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.4011,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.2769255160537979,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.399,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.2459674855959307,
"learning_rate": 1.326303494478451e-05,
"loss": 0.421,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.28765522994752923,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.4163,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.2613451333757003,
"learning_rate": 1.315664479122898e-05,
"loss": 0.4056,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.28249264396633994,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.4185,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.2615845107411125,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.422,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.26724163076624896,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.4116,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.27866606616749817,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.4116,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.26254503751819297,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.4052,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.25564978750209927,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.4257,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.2753877579174224,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.4018,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.26888416558488937,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.4153,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.27462250457959336,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.4244,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.2763201316493202,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.4196,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.2650175462382448,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.3975,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.27559144112699857,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.4113,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.2583006610309122,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.4085,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.27176426848872937,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.4183,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.27090134078524997,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.4079,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.25131983618547526,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.3946,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.2512938414451212,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.414,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.2824713568357215,
"learning_rate": 1.220938883250288e-05,
"loss": 0.4127,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.2791777252767588,
"learning_rate": 1.215733275116827e-05,
"loss": 0.4203,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.2608057813154562,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.3995,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.258646568355715,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.4197,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.255935384994262,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.3991,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.2772220863941059,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.4141,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.2906783254948936,
"learning_rate": 1.189800129803139e-05,
"loss": 0.4254,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.2637512135653491,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.4224,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.2501103344566086,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.4236,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.26108334556783547,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.4177,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.23690471733921595,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.4205,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.2846511337860584,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.4372,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.2726968527521934,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.4132,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.2544681925333692,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.4155,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.2865385909141826,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.3929,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.24507201850944968,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.4273,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.2702068861794847,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.4111,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.24168324519170156,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.4082,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.24970511134578297,
"learning_rate": 1.128231163049258e-05,
"loss": 0.4035,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.23384272565894684,
"learning_rate": 1.123144751066259e-05,
"loss": 0.4169,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.25287037003995416,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.4156,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.26948751202757826,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.4037,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.2511182088268803,
"learning_rate": 1.107927745387114e-05,
"loss": 0.421,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.22710956573981803,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.4192,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.2593332974303366,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.4165,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.235832665034044,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.4116,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.25818182248591476,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.4259,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.23895871459091272,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.4323,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.23683754501140006,
"learning_rate": 1.077687414070869e-05,
"loss": 0.3976,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.2408684017018197,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.4081,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.23096485417175316,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.4225,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.2429563254984418,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.4261,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.23145653641150138,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.4198,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.24026379099222855,
"learning_rate": 1.052689572753247e-05,
"loss": 0.4086,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.23384209256344032,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.4175,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.22271639125151008,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.4127,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.24857123891225302,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.4106,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.23976237719588717,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.4088,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.2665465763778325,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.4,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.22537145088639948,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.4163,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.24214161123990185,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.4046,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.22700607101139725,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.4098,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.23288940568661937,
"learning_rate": 1.008173988718005e-05,
"loss": 0.4132,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.23572302498849448,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.4147,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.22807310290895674,
"learning_rate": 9.983678617766677e-06,
"loss": 0.4149,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.25660202970878854,
"learning_rate": 9.934767922553788e-06,
"loss": 0.414,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.24819479516162188,
"learning_rate": 9.885937708668371e-06,
"loss": 0.424,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.22804645600823154,
"learning_rate": 9.837188366555524e-06,
"loss": 0.4006,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.2504391620083499,
"learning_rate": 9.788520286013691e-06,
"loss": 0.4034,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.2406176720516951,
"learning_rate": 9.739933856191537e-06,
"loss": 0.4201,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.2604489043668825,
"learning_rate": 9.691429465584866e-06,
"loss": 0.4207,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.23534806066906125,
"learning_rate": 9.643007502033492e-06,
"loss": 0.4146,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.27849608770199713,
"learning_rate": 9.594668352718152e-06,
"loss": 0.4113,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.24549656587047236,
"learning_rate": 9.546412404157383e-06,
"loss": 0.4038,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.2337568797497819,
"learning_rate": 9.498240042204471e-06,
"loss": 0.4092,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.26536811641619246,
"learning_rate": 9.450151652044339e-06,
"loss": 0.4253,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.22547782558464918,
"learning_rate": 9.402147618190472e-06,
"loss": 0.4164,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.23098439735990858,
"learning_rate": 9.354228324481841e-06,
"loss": 0.4318,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.23827309909441813,
"learning_rate": 9.306394154079852e-06,
"loss": 0.402,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.24225298836025827,
"learning_rate": 9.258645489465254e-06,
"loss": 0.429,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.22575382573444558,
"learning_rate": 9.21098271243509e-06,
"loss": 0.4112,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.23312754718245454,
"learning_rate": 9.163406204099659e-06,
"loss": 0.4177,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.24150278143005124,
"learning_rate": 9.115916344879452e-06,
"loss": 0.4193,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.23154893338102145,
"learning_rate": 9.068513514502126e-06,
"loss": 0.4225,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.22588305771559986,
"learning_rate": 9.021198091999439e-06,
"loss": 0.417,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.2290382070585622,
"learning_rate": 8.973970455704251e-06,
"loss": 0.4211,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.2413127364569342,
"learning_rate": 8.926830983247495e-06,
"loss": 0.4101,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.2338283193899335,
"learning_rate": 8.879780051555116e-06,
"loss": 0.4234,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.22897804147166984,
"learning_rate": 8.83281803684512e-06,
"loss": 0.4074,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.2326001347841395,
"learning_rate": 8.785945314624532e-06,
"loss": 0.4135,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.23026078314636295,
"learning_rate": 8.739162259686385e-06,
"loss": 0.4073,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.22473651332868777,
"learning_rate": 8.692469246106734e-06,
"loss": 0.4176,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.2443783361362979,
"learning_rate": 8.645866647241677e-06,
"loss": 0.4108,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.23002169990766075,
"learning_rate": 8.599354835724348e-06,
"loss": 0.4116,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.23268128242547734,
"learning_rate": 8.552934183461958e-06,
"loss": 0.3984,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.2224927264395771,
"learning_rate": 8.506605061632805e-06,
"loss": 0.4167,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.23101976210461447,
"learning_rate": 8.460367840683308e-06,
"loss": 0.4202,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.21712611256209488,
"learning_rate": 8.414222890325065e-06,
"loss": 0.4133,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.21756982915767442,
"learning_rate": 8.368170579531855e-06,
"loss": 0.4213,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.2209265515373188,
"learning_rate": 8.322211276536733e-06,
"loss": 0.4157,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.23420271135517912,
"learning_rate": 8.27634534882907e-06,
"loss": 0.4093,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.2304021023437358,
"learning_rate": 8.230573163151585e-06,
"loss": 0.4075,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.2341416351430568,
"learning_rate": 8.184895085497466e-06,
"loss": 0.4217,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.21541524778968704,
"learning_rate": 8.139311481107401e-06,
"loss": 0.4125,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.24357309262982635,
"learning_rate": 8.093822714466686e-06,
"loss": 0.4035,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.22594737977404436,
"learning_rate": 8.048429149302271e-06,
"loss": 0.4204,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.23350312242870855,
"learning_rate": 8.003131148579902e-06,
"loss": 0.4112,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.21496136318907585,
"learning_rate": 7.957929074501182e-06,
"loss": 0.4175,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.21634916136906276,
"learning_rate": 7.912823288500695e-06,
"loss": 0.4009,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.22439313812949593,
"learning_rate": 7.867814151243102e-06,
"loss": 0.4117,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.21106458876182177,
"learning_rate": 7.822902022620267e-06,
"loss": 0.4097,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.2239151770160289,
"learning_rate": 7.778087261748377e-06,
"loss": 0.4201,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.23800318924957842,
"learning_rate": 7.733370226965056e-06,
"loss": 0.4245,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.2236619322311424,
"learning_rate": 7.688751275826527e-06,
"loss": 0.4026,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.22295583035322739,
"learning_rate": 7.644230765104735e-06,
"loss": 0.4284,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.23891133234645762,
"learning_rate": 7.599809050784502e-06,
"loss": 0.4077,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.23098270106614083,
"learning_rate": 7.555486488060661e-06,
"loss": 0.4142,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.24876440065136285,
"learning_rate": 7.511263431335249e-06,
"loss": 0.4222,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.2410553548747029,
"learning_rate": 7.467140234214656e-06,
"loss": 0.4186,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.2358495320192969,
"learning_rate": 7.423117249506777e-06,
"loss": 0.4091,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.23630193973592392,
"learning_rate": 7.379194829218231e-06,
"loss": 0.4079,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.2439011022411804,
"learning_rate": 7.335373324551518e-06,
"loss": 0.4059,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.22534943320328846,
"learning_rate": 7.291653085902224e-06,
"loss": 0.4217,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.23553770628617307,
"learning_rate": 7.248034462856213e-06,
"loss": 0.4056,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.2473300318678417,
"learning_rate": 7.20451780418683e-06,
"loss": 0.4176,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.22323434010865567,
"learning_rate": 7.161103457852128e-06,
"loss": 0.4241,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.2632135352296226,
"learning_rate": 7.117791770992046e-06,
"loss": 0.4107,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.24222304718371251,
"learning_rate": 7.074583089925688e-06,
"loss": 0.4029,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.21831243640942655,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.4226,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.24260393418471354,
"learning_rate": 6.988476126329604e-06,
"loss": 0.4062,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.24220807096110208,
"learning_rate": 6.945578532308857e-06,
"loss": 0.4142,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.2172131480934278,
"learning_rate": 6.902785321094301e-06,
"loss": 0.4114,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.23787859324843463,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.4012,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.24666465988379774,
"learning_rate": 6.817513414939892e-06,
"loss": 0.4183,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.2141431854876769,
"learning_rate": 6.775035401831924e-06,
"loss": 0.418,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.2518472261514643,
"learning_rate": 6.732663135188473e-06,
"loss": 0.3999,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.22966553336759604,
"learning_rate": 6.690396953817055e-06,
"loss": 0.4275,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.221286295367878,
"learning_rate": 6.648237195676927e-06,
"loss": 0.4165,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.24631994228635434,
"learning_rate": 6.606184197876395e-06,
"loss": 0.4258,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.24352784994048748,
"learning_rate": 6.564238296670109e-06,
"loss": 0.3994,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.22241816679368606,
"learning_rate": 6.52239982745638e-06,
"loss": 0.4092,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.2305695018072674,
"learning_rate": 6.480669124774484e-06,
"loss": 0.4142,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.23196094164528994,
"learning_rate": 6.439046522302013e-06,
"loss": 0.416,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.21685440854228646,
"learning_rate": 6.397532352852191e-06,
"loss": 0.4056,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.2383922100855665,
"learning_rate": 6.356126948371202e-06,
"loss": 0.4138,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.232575928806162,
"learning_rate": 6.314830639935559e-06,
"loss": 0.4094,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.21646461385766977,
"learning_rate": 6.273643757749448e-06,
"loss": 0.4197,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.2333131619871786,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.4171,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.23062076093536288,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.4216,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.2109922821932709,
"learning_rate": 6.150742957589784e-06,
"loss": 0.4125,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.2182514395397677,
"learning_rate": 6.109997064904774e-06,
"loss": 0.4274,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.23335923435743444,
"learning_rate": 6.069362236313092e-06,
"loss": 0.3939,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.21300296409427907,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.4262,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.22615463591608181,
"learning_rate": 5.988427070179108e-06,
"loss": 0.4064,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.2326776682470719,
"learning_rate": 5.948127379792234e-06,
"loss": 0.4274,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.25045264685590574,
"learning_rate": 5.90794004780439e-06,
"loss": 0.409,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.2153490980612771,
"learning_rate": 5.867865395552405e-06,
"loss": 0.4165,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.22103535934998259,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.4004,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.2180002774548252,
"learning_rate": 5.788055411095892e-06,
"loss": 0.4149,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.2385268622228181,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.4054,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.22364044639864258,
"learning_rate": 5.708699979051613e-06,
"loss": 0.4172,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.21566111880087405,
"learning_rate": 5.66919351390751e-06,
"loss": 0.4267,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.21162797426216315,
"learning_rate": 5.629801637510188e-06,
"loss": 0.4152,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.24617069774505523,
"learning_rate": 5.590524664836043e-06,
"loss": 0.4157,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.21811703248357764,
"learning_rate": 5.551362909942708e-06,
"loss": 0.4165,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.20929947200073917,
"learning_rate": 5.51231668596653e-06,
"loss": 0.4095,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.2070012034538388,
"learning_rate": 5.47338630512007e-06,
"loss": 0.4257,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.23427390155664682,
"learning_rate": 5.434572078689617e-06,
"loss": 0.4073,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.21471735899249653,
"learning_rate": 5.395874317032694e-06,
"loss": 0.4103,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.21678814945254746,
"learning_rate": 5.357293329575573e-06,
"loss": 0.3996,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.23340263963521438,
"learning_rate": 5.318829424810797e-06,
"loss": 0.4176,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.2162874815139532,
"learning_rate": 5.280482910294728e-06,
"loss": 0.4157,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.22339215875882196,
"learning_rate": 5.242254092645078e-06,
"loss": 0.4133,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.21935991782793254,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.4021,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.2149330227657988,
"learning_rate": 5.166150769707892e-06,
"loss": 0.404,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.2145717450210439,
"learning_rate": 5.128276872940503e-06,
"loss": 0.396,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.24191749283780567,
"learning_rate": 5.090521890074953e-06,
"loss": 0.3967,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.22865579608560596,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.4115,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.2163577214366134,
"learning_rate": 5.015369872647442e-06,
"loss": 0.4092,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.24010967513327772,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.4102,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.21489762049834785,
"learning_rate": 4.940697121074675e-06,
"loss": 0.4215,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.22753544862841632,
"learning_rate": 4.903541216934864e-06,
"loss": 0.4122,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.20794872178003584,
"learning_rate": 4.8665060236772e-06,
"loss": 0.4126,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.21293096308687784,
"learning_rate": 4.829591837434102e-06,
"loss": 0.4113,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.22136108376988034,
"learning_rate": 4.792798953370424e-06,
"loss": 0.3999,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.22197928419093232,
"learning_rate": 4.756127665681083e-06,
"loss": 0.4108,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.220542314385165,
"learning_rate": 4.719578267588725e-06,
"loss": 0.3991,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.22381399194416057,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.4069,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.41990410208493173,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.3603,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.2846313002791846,
"learning_rate": 4.610664328486485e-06,
"loss": 0.3645,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.3518970662046642,
"learning_rate": 4.5746054014809e-06,
"loss": 0.3731,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.48722615992784146,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.3821,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.3022965106506803,
"learning_rate": 4.502857857942204e-06,
"loss": 0.3587,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.3669773130269681,
"learning_rate": 4.467169815100549e-06,
"loss": 0.3642,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.41460914371006374,
"learning_rate": 4.431605972355137e-06,
"loss": 0.3688,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.3018738705514469,
"learning_rate": 4.396166614073516e-06,
"loss": 0.3699,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.2462077568681124,
"learning_rate": 4.360852023627846e-06,
"loss": 0.367,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.31794615220711214,
"learning_rate": 4.325662483392656e-06,
"loss": 0.3701,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.35365490821530077,
"learning_rate": 4.29059827474257e-06,
"loss": 0.3892,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.2992181732707704,
"learning_rate": 4.255659678050063e-06,
"loss": 0.3707,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.2547241838184468,
"learning_rate": 4.22084697268323e-06,
"loss": 0.3631,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.3158804992602464,
"learning_rate": 4.186160437003535e-06,
"loss": 0.358,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.31715550719909613,
"learning_rate": 4.151600348363582e-06,
"loss": 0.362,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.2308248557505715,
"learning_rate": 4.11716698310493e-06,
"loss": 0.3605,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.2647111234849541,
"learning_rate": 4.082860616555848e-06,
"loss": 0.3702,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.29694106361902867,
"learning_rate": 4.048681523029125e-06,
"loss": 0.3737,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.2584909825229887,
"learning_rate": 4.014629975819882e-06,
"loss": 0.3673,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.25061613765345925,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.3723,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.2424683718616947,
"learning_rate": 3.94691060843287e-06,
"loss": 0.3633,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.23978916687788301,
"learning_rate": 3.913243329737353e-06,
"loss": 0.3533,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.26992429707618276,
"learning_rate": 3.879704680319503e-06,
"loss": 0.3759,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.23413954299329837,
"learning_rate": 3.846294928353471e-06,
"loss": 0.3575,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.23916001057959713,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.3653,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.24175942457552543,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.3671,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.22315740706365356,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.3634,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.2332601031508653,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.3904,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.2331101962909002,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.3576,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.2366476444253031,
"learning_rate": 3.648558150665886e-06,
"loss": 0.371,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.21935432597740612,
"learning_rate": 3.616058102895075e-06,
"loss": 0.374,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.24006551523553918,
"learning_rate": 3.583689060687079e-06,
"loss": 0.371,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.2196032025164845,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.3705,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.23514203818796536,
"learning_rate": 3.51934502719796e-06,
"loss": 0.3727,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.23364166938139527,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.3776,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.2326541009931241,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.3563,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.23137610626787197,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.3444,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.22662777234085926,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.3678,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.22386101661633637,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.3579,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.21939029259850715,
"learning_rate": 3.32948376099375e-06,
"loss": 0.3584,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.2269387260168193,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.3596,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.22001985930391788,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.3479,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.2133249866081245,
"learning_rate": 3.236349233544438e-06,
"loss": 0.3722,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.22282431767251853,
"learning_rate": 3.205572144561604e-06,
"loss": 0.3681,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.22108047683381493,
"learning_rate": 3.174929343376374e-06,
"loss": 0.3638,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.21036938039492986,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.3809,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.21495310485676417,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.3702,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.21831037843753306,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.3678,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.20220346168426853,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.3854,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.21294703635667386,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.3671,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.2088083219710041,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.3622,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.22469548017350555,
"learning_rate": 2.964210239224696e-06,
"loss": 0.3558,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.20881870401171027,
"learning_rate": 2.934650470499869e-06,
"loss": 0.3761,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.2095862260230288,
"learning_rate": 2.905227155855128e-06,
"loss": 0.3858,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.20533670009184238,
"learning_rate": 2.875940530558519e-06,
"loss": 0.3886,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.2050845697572956,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.3742,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.2177856023987455,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.3795,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.2210359884701473,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.3684,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.21083445634828366,
"learning_rate": 2.760165589921184e-06,
"loss": 0.3611,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.20772099591204485,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.358,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.22462541754795778,
"learning_rate": 2.703104292151164e-06,
"loss": 0.3779,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.21592148084499152,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.3712,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.20923692191479656,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.3576,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.20628045337400902,
"learning_rate": 2.618550200289303e-06,
"loss": 0.3637,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.21899646591589092,
"learning_rate": 2.59064316684412e-06,
"loss": 0.3608,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.20848962415181024,
"learning_rate": 2.562875338152264e-06,
"loss": 0.3577,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.20543920475340757,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.364,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.20503341299675748,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.3724,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.20163893039465708,
"learning_rate": 2.480409295328803e-06,
"loss": 0.378,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.21328422527898044,
"learning_rate": 2.45320049480108e-06,
"loss": 0.3658,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.2226241924065024,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.354,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.203283101728901,
"learning_rate": 2.399204021408421e-06,
"loss": 0.3635,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.20081783229720987,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.3816,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.19804533068254646,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.3722,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.21323268338210974,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.3442,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.21970528642864318,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.3676,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.2076048569589428,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.3769,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.21331906498321923,
"learning_rate": 2.24059906434998e-06,
"loss": 0.3715,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.21043330932708856,
"learning_rate": 2.214660694699091e-06,
"loss": 0.3781,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.20098057416207832,
"learning_rate": 2.188864536147355e-06,
"loss": 0.3773,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.19944101479688023,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.3937,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.20574450301065694,
"learning_rate": 2.137699676264171e-06,
"loss": 0.3632,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.21394563676710998,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.3524,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.1963216556738751,
"learning_rate": 2.08710612114857e-06,
"loss": 0.3619,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.19816152527158684,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.3673,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.20809845900631813,
"learning_rate": 2.0370854889762e-06,
"loss": 0.3687,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.20971845531213001,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.3659,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.21185427846240582,
"learning_rate": 1.987639379598427e-06,
"loss": 0.3699,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.20292905244231785,
"learning_rate": 1.963132265887919e-06,
"loss": 0.378,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.20175846781379592,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.3734,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.20238668609790197,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.3631,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.21421147361222254,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.3578,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.20730764677390007,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.3678,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.2075773955230489,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.358,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.20554370745800202,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.3528,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.2137210431302318,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.3634,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.21112513861398044,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.3645,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.20491443466835957,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.3734,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.19491676633151628,
"learning_rate": 1.726025117089718e-06,
"loss": 0.3729,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.20623860157783805,
"learning_rate": 1.70311497206598e-06,
"loss": 0.357,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.19999844034197756,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.3759,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.21164825836882653,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.3847,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.21852369210202754,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.3605,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.20583442367828989,
"learning_rate": 1.612939220355272e-06,
"loss": 0.3573,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.2060195894314662,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.3653,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.19183635222573836,
"learning_rate": 1.568732761074121e-06,
"loss": 0.3706,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.20957887255986124,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.3607,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.2000018866542812,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.3593,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.2049369384661975,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.3674,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.2051675189844366,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.3772,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.2054811627132044,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.3613,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.20041842629438972,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.3802,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.20255137230200237,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.3745,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.1998521020986781,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.358,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.19709926512736992,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.3652,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.2002833206439517,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.3679,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.19606343693903563,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.3827,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.20017121995639542,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.3573,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.20623587429240745,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.3673,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.20752894629647722,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.3612,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.20251203185347544,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.3558,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.19281456711485753,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.3799,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.21548738131382678,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.3524,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.19891618486057355,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.367,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.20241956573420672,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.3641,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.19994872978641956,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.3592,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.2063967795515686,
"learning_rate": 1.140415161609505e-06,
"loss": 0.3704,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.1991198013448001,
"learning_rate": 1.121666393356604e-06,
"loss": 0.3638,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.19649377629970372,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.3789,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.20021775357533622,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.376,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.19567535855471496,
"learning_rate": 1.066326386135097e-06,
"loss": 0.3884,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.20528821830817376,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.3656,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.19928374675256444,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.3687,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.19818631405318057,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.3771,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.2029150942092355,
"learning_rate": 9.9465987684501e-07,
"loss": 0.3659,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.2036202366995783,
"learning_rate": 9.771228098856534e-07,
"loss": 0.365,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.20857784869213425,
"learning_rate": 9.597378493454013e-07,
"loss": 0.3529,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.20375189892660048,
"learning_rate": 9.425051342339353e-07,
"loss": 0.3636,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.20021588426667247,
"learning_rate": 9.254248023435864e-07,
"loss": 0.3671,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.19750633968707412,
"learning_rate": 9.084969902482288e-07,
"loss": 0.3708,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.20241345851860265,
"learning_rate": 8.917218333021993e-07,
"loss": 0.3851,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.21134070514187947,
"learning_rate": 8.750994656392109e-07,
"loss": 0.34,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.20140515878320991,
"learning_rate": 8.58630020171265e-07,
"loss": 0.3784,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.20377128061776306,
"learning_rate": 8.423136285876099e-07,
"loss": 0.3506,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.20485794308925587,
"learning_rate": 8.26150421353682e-07,
"loss": 0.3504,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.2000477388498111,
"learning_rate": 8.101405277100549e-07,
"loss": 0.3705,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.2050988073300908,
"learning_rate": 7.942840756714077e-07,
"loss": 0.3715,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.20106734896698994,
"learning_rate": 7.785811920255093e-07,
"loss": 0.3753,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.20115012850771896,
"learning_rate": 7.630320023322024e-07,
"loss": 0.3685,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.19866374969424744,
"learning_rate": 7.476366309223903e-07,
"loss": 0.3655,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.18936429933378507,
"learning_rate": 7.323952008970537e-07,
"loss": 0.3712,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.20915292865121363,
"learning_rate": 7.173078341262618e-07,
"loss": 0.3636,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.20523214581575067,
"learning_rate": 7.023746512481989e-07,
"loss": 0.3603,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.1979853652779165,
"learning_rate": 6.875957716681902e-07,
"loss": 0.3622,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.196621974204872,
"learning_rate": 6.729713135577687e-07,
"loss": 0.3794,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.1970125961253692,
"learning_rate": 6.585013938537143e-07,
"loss": 0.3655,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.19791248022514119,
"learning_rate": 6.441861282571115e-07,
"loss": 0.3492,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.2021441503471418,
"learning_rate": 6.300256312324471e-07,
"loss": 0.3468,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.19342044089152963,
"learning_rate": 6.160200160066776e-07,
"loss": 0.3798,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.20547524570843428,
"learning_rate": 6.021693945683327e-07,
"loss": 0.3649,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.20548174449956483,
"learning_rate": 5.884738776666044e-07,
"loss": 0.3681,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.1965293322504822,
"learning_rate": 5.749335748104878e-07,
"loss": 0.384,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.2002045495098851,
"learning_rate": 5.615485942678778e-07,
"loss": 0.3566,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.19722652555829734,
"learning_rate": 5.483190430647289e-07,
"loss": 0.3633,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.19812197519623653,
"learning_rate": 5.352450269841747e-07,
"loss": 0.3738,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.19834958884973036,
"learning_rate": 5.22326650565701e-07,
"loss": 0.3665,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.19780691182497906,
"learning_rate": 5.095640171043048e-07,
"loss": 0.3569,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.20426847488635339,
"learning_rate": 4.969572286496594e-07,
"loss": 0.3466,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.20017462354285764,
"learning_rate": 4.845063860053079e-07,
"loss": 0.3674,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.20175752306966438,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.3613,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.1980569419071521,
"learning_rate": 4.600729351261857e-07,
"loss": 0.3674,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.19965913041133676,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.3774,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.19529118872053333,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.3761,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.2015538426563906,
"learning_rate": 4.245948007320588e-07,
"loss": 0.3675,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.20389754583493486,
"learning_rate": 4.130816799401971e-07,
"loss": 0.3633,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.19930404361626008,
"learning_rate": 4.017251756252827e-07,
"loss": 0.375,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2048613160383962,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.3595,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.1971859822484076,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.3773,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.1973401464009615,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.3577,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.19711588716595396,
"learning_rate": 3.578671204615347e-07,
"loss": 0.363,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.1935014873680179,
"learning_rate": 3.472950355547267e-07,
"loss": 0.3607,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.19570172237540753,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.3873,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.19813314654717876,
"learning_rate": 3.266223765343224e-07,
"loss": 0.3497,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.19315160883448024,
"learning_rate": 3.16521967718757e-07,
"loss": 0.374,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.19675456815731834,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.3749,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.19409983221912222,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.3767,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.19600634886875637,
"learning_rate": 2.871653894244597e-07,
"loss": 0.3691,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.19778563586024653,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.3646,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.19981629930055383,
"learning_rate": 2.683823233205485e-07,
"loss": 0.3574,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.19878735679974358,
"learning_rate": 2.592274132703287e-07,
"loss": 0.3655,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.20672016043766658,
"learning_rate": 2.502303499116754e-07,
"loss": 0.3702,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.19500059090105432,
"learning_rate": 2.413912051848666e-07,
"loss": 0.3594,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.20152881812635645,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.3747,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.20001428102709265,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.37,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.20279016569453043,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.3745,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.1912364391881628,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.3727,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.19417346455203527,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.3748,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.19955141694172623,
"learning_rate": 1.916764975761254e-07,
"loss": 0.3695,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.20408763553906922,
"learning_rate": 1.839446914621923e-07,
"loss": 0.3887,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.20260168479981014,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.3728,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.202404052620326,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.3726,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.1985784366559,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.3559,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.2007705485209311,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.3722,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.19990162748803394,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.3508,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.19103273358317058,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.3751,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.19805239711085088,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.3668,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.19950237228613302,
"learning_rate": 1.277993209368744e-07,
"loss": 0.3784,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.1960848414275156,
"learning_rate": 1.214956261808542e-07,
"loss": 0.3761,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.20409226225799393,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.3742,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.20190762383171812,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.3644,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.2008830510374153,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.3618,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.2015642582992315,
"learning_rate": 9.787081204466253e-08,
"loss": 0.3585,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.200337024148203,
"learning_rate": 9.236233586200627e-08,
"loss": 0.3649,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.20216490119510958,
"learning_rate": 8.70130406196279e-08,
"loss": 0.3534,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.19807600329723068,
"learning_rate": 8.182296909035047e-08,
"loss": 0.3708,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.1992917666630579,
"learning_rate": 7.679216277384439e-08,
"loss": 0.3664,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.19950219543559608,
"learning_rate": 7.19206618963053e-08,
"loss": 0.367,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.19627907924336976,
"learning_rate": 6.720850541012347e-08,
"loss": 0.3712,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.19239160939995406,
"learning_rate": 6.265573099357714e-08,
"loss": 0.3779,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.19659005352363035,
"learning_rate": 5.826237505053067e-08,
"loss": 0.3791,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.20239587822533653,
"learning_rate": 5.4028472710137e-08,
"loss": 0.3583,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.1937931557117631,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.3742,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.19524122686233425,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.3637,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.20192118873749879,
"learning_rate": 4.228381947004812e-08,
"loss": 0.3579,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.19733989201596763,
"learning_rate": 3.868805732811343e-08,
"loss": 0.3592,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.19305015351314248,
"learning_rate": 3.525190530455636e-08,
"loss": 0.3678,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.19525704329565102,
"learning_rate": 3.197539087475887e-08,
"loss": 0.364,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.19560428897725776,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.3713,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.19636360248064622,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.3685,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.19577666384959366,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.3742,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.1943038219112364,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.371,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.1996087118781684,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.3654,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.18986010265663797,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.3651,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.19537342685909986,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.3672,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.20320417679939393,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.3551,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.19889551308801154,
"learning_rate": 9.67435416023843e-09,
"loss": 0.352,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.18680970283587553,
"learning_rate": 7.995445897566178e-09,
"loss": 0.3706,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.2100211114936319,
"learning_rate": 6.476393170578021e-09,
"loss": 0.3496,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.1971852641189796,
"learning_rate": 5.11720812557348e-09,
"loss": 0.3606,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.19282427439779737,
"learning_rate": 3.917901630563492e-09,
"loss": 0.3666,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.19272500290775516,
"learning_rate": 2.878483275170485e-09,
"loss": 0.3596,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.18987418792031807,
"learning_rate": 1.998961370555108e-09,
"loss": 0.3774,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.2026764277086141,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.3641,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.19564385973295234,
"learning_rate": 7.196337656112384e-10,
"loss": 0.3637,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.19897868638013028,
"learning_rate": 3.198382947511292e-10,
"loss": 0.3567,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.19506782460794858,
"learning_rate": 7.995973352770137e-11,
"loss": 0.3739,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.1974464457437831,
"learning_rate": 0.0,
"loss": 0.3597,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 5.085521666277114e+18,
"train_loss": 0.4956043636026653,
"train_runtime": 29487.8841,
"train_samples_per_second": 5.358,
"train_steps_per_second": 0.042
}
],
"logging_steps": 1,
"max_steps": 1235,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.085521666277114e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}